cheese science

657

Upload: bernolim

Post on 29-Nov-2014

276 views

Category:

Documents


18 download

TRANSCRIPT

Page 1: Cheese Science
Page 2: Cheese Science

iii This page has been reformatted by Knovel to provide easier navigation.

Contents

Preface ................................................................................................... ix

1. Cheese: Historical Aspects .......................................................... 1 1.1 Introduction .......................................................................................... 1 1.2 Cheese Production and Consumption ................................................ 5 1.3 Cheese Science and Technology ....................................................... 7

2. Overview of Cheese Manufacture ................................................ 10 2.1 Selection of Milk ................................................................................... 10 2.2 Standardization of Milk Composition ................................................... 10 2.3 Heat Treatment of Milk ........................................................................ 12 2.4 Cheese Color ....................................................................................... 13 2.5 Conversion of Milk to Cheese Curd .................................................... 14 2.6 Ripening ............................................................................................... 17 2.7 Processed Cheese Products ............................................................... 17 2.8 Whey and Whey Products ................................................................... 17

3. Chemistry of Milk Constituents .................................................... 19 3.1 Introduction .......................................................................................... 19 3.2 Lactose ................................................................................................ 20 3.3 Milk Lipids ............................................................................................ 25 3.4 Milk Proteins ........................................................................................ 31 3.5 Milk Salts .............................................................................................. 39 3.6 pH of Milk ............................................................................................. 41 3.7 Physicochemical Properties of Milk ..................................................... 43

4. Bacteriology of Cheese Milk ......................................................... 45 4.1 Contamination of Raw Milk .................................................................. 45 4.2 Pasteurization ...................................................................................... 47

Page 3: Cheese Science

iv Contents

This page has been reformatted by Knovel to provide easier navigation.

4.3 Alternatives to Heat Treatment ............................................................ 49 4.4 Prematuration ...................................................................................... 53

5. Starter Cultures .............................................................................. 54 5.1 Introduction .......................................................................................... 54 5.2 Types of Cultures ................................................................................ 54 5.3 Taxonomy ............................................................................................ 62 5.4 Metabolism of Starters ......................................................................... 69 5.5 Plasmids .............................................................................................. 82 5.6 Inhibition of Acid Production ................................................................ 83 5.7 Bacteriophage ...................................................................................... 83 5.8 Bacteriocins ......................................................................................... 92 5.9 Production of Starters in Cheese Plants ............................................. 94 5.10 Measurement of Generation Times .................................................... 96

6. Enzymatic Coagulation of Milk ..................................................... 98 6.1 The Primary Phase of Rennet Coagulation ........................................ 98 6.2 Rennet ................................................................................................. 101 6.3 Factors That Affect the Hydrolysis of κ-Casein and the Primary

Phase of Rennet Coagulation ............................................................. 102 6.4 The Secondary (Nonenzymatic) Phase of Coagulation and Gel

Assembly ............................................................................................. 103 6.5 Factors That Affect the Nonenzymatic Phase of Rennet

Coagulation .......................................................................................... 108 6.6 Measurement of Rennet Coagulation Properties ............................... 109 6.7 Factors That Affect Rennet Coagulation ............................................. 120 6.8 Rennet Substitutes .............................................................................. 130 6.9 Immobilized Rennets ........................................................................... 135

7. Post-Coagulation Treatment of Renneted Milk Gel .................... 138 7.1 Introduction .......................................................................................... 138 7.2 Methods for Measuring Syneresis ....................................................... 139 7.3 Influence of Compositional Factors on Syneresis ............................... 139 7.4 Influence of Processing Variables on Syneresis ................................. 140 7.5 Kinetics and Mechanism of Syneresis ................................................ 145

Page 4: Cheese Science

Contents v

This page has been reformatted by Knovel to provide easier navigation.

7.6 Textured Cheese ................................................................................. 145 7.7 Molding and Pressing of Cheese Curd ............................................... 150 7.8 Packaging ............................................................................................ 151

8. Salting of Cheese Curd ................................................................. 153 8.1 Introduction .......................................................................................... 153 8.2 Salting of Cheese Curd ....................................................................... 155 8.3 Effect of Salt on Cheese Composition ................................................ 162 8.4 Effect of NaCl on the Microbiology of Cheese .................................... 163 8.5 Influence of NaCl on Enzymes in Cheese .......................................... 165 8.6 Effect of Salt on Cheese Quality ......................................................... 166 8.7 Nutritional Aspects of NaCl in Cheese ................................................ 167

9. Cheese Yield .................................................................................. 169 9.1 Introduction .......................................................................................... 169 9.2 Definition of Cheese Yield ................................................................... 169 9.3 Measurement of Cheese Yield and Efficiency .................................... 171 9.4 Prediction of Cheese Yield .................................................................. 173 9.5 Factors That Affect Cheese Yield ....................................................... 174 9.6 Conclusion ........................................................................................... 202

10. Microbiology of Cheese Ripening ................................................ 206 10.1 General Features ................................................................................. 206 10.2 Microbial Activity during Ripening ....................................................... 207 10.3 Growth of Starter Bacteria in Cheese ................................................. 213 10.4 Growth of Nonstarter Lactic Acid Bacteria in Cheese ........................ 215 10.5 Other Microorganisms in Ripening Cheese ........................................ 217 10.6 Examples of Microbial Growth in Cheese ........................................... 226 10.7 Microbial Spoilage of Cheese .............................................................. 232

11. Biochemistry of Cheese Ripening ................................................ 236 11.1 Introduction .......................................................................................... 236 11.2 Ripening Agents in Cheese ................................................................. 236 11.3 Contribution of Individual Agents to Ripening ..................................... 237 11.4 Glycolysis and Related Events ............................................................ 238

Page 5: Cheese Science

vi Contents

This page has been reformatted by Knovel to provide easier navigation.

11.5 Citrate Metabolism ............................................................................... 248 11.6 Lipolysis and Related Events .............................................................. 249 11.7 Proteolysis ........................................................................................... 255 11.8 Characterization of Proteolysis in Cheese .......................................... 268 11.9 Catabolism of Amino Acids and Related Events ................................ 274 11.10 Conclusion ........................................................................................... 278

12. Cheese Flavor ................................................................................ 282 12.1 Introduction .......................................................................................... 282 12.2 Analytical Methods .............................................................................. 284 12.3 Contribution of the Aqueous Phase of Cheese to Flavor ................... 288 12.4 Contribution of Volatile Compounds to Cheese Flavor ....................... 292 12.5 Off-Flavors in Cheese .......................................................................... 293 12.6 Formation of Flavor Compounds ......................................................... 297 12.7 Intervarietal and Intravarietal Comparison of Cheese Ripening ......... 300 12.8 Conclusion ........................................................................................... 303

13. Cheese Rheology and Texture ..................................................... 305 13.1 Introduction .......................................................................................... 305 13.2 Cheese Microstructure ........................................................................ 306 13.3 Rheological Characteristics of Cheese ............................................... 311 13.4 Cheese Texture ................................................................................... 333

14. Factors That Affect Cheese Quality ............................................. 341 14.1 Introduction .......................................................................................... 341 14.2 Milk Supply .......................................................................................... 341 14.3 Coagulant (Rennet) ............................................................................. 343 14.4 Starter .................................................................................................. 343 14.5 Nonstarter Lactic Acid Bacteria (NSLAB) ........................................... 344 14.6 Cheese Composition ........................................................................... 345 14.7 Ripening Temperature ......................................................................... 347 14.8 Conclusion ........................................................................................... 347

15. Acceleration of Cheese Ripening ................................................. 349 15.1 Introduction .......................................................................................... 349

Page 6: Cheese Science

Contents vii

This page has been reformatted by Knovel to provide easier navigation.

15.2 Elevated Temperature ......................................................................... 350 15.3 Exogenous Enzymes ........................................................................... 351 15.4 Selected, Activated, or Modified Starters ............................................ 355 15.5 Adjunct Starters ................................................................................... 358 15.6 Secondary Cultures ............................................................................. 359 15.7 Enzyme-Modified Cheese ................................................................... 359 15.8 Addition of Amino Acids to Cheese Curd ............................................ 360 15.9 Prospects for Accelerated Ripening .................................................... 360

16. Fresh Acid-Curd Cheese Varieties ............................................... 363 16.1 Introduction .......................................................................................... 363 16.2 Overview of the Manufacturing Process for Fresh Acid-Curd

Cheese Products ................................................................................. 363 16.3 Principles of Acid Milk Gel Formation ................................................. 364 16.4 Prerequisites for Gel Formation .......................................................... 368 16.5 Effect of Gel Structure on Quality ........................................................ 369 16.6 Factors That Influence the Structure of Acid Gels and the

Quality of Fresh Cheese Products ...................................................... 374 16.7 Treatments of the Separated Curd ...................................................... 378 16.8 Major Fresh Acid-Curd Cheese Varieties ........................................... 379

17. Principal Families of Cheese ........................................................ 388 17.1 Introduction .......................................................................................... 388 17.2 Rennet-Coagulated Cheeses .............................................................. 392 17.3 Acid-Coagulated Cheeses ................................................................... 422 17.4 Heat/Acid-Coagulated Cheeses .......................................................... 422 17.5 Concentration and Crystallization ....................................................... 423 17.6 Ultrafiltration Technology in Cheesemaking ....................................... 425 Appendix 17-A: Compositions of Selected Cheese Varieties ....................... 428

18. Processed Cheese and Substitute or Imitation Cheese Products ......................................................................................... 429 18.1 Introduction .......................................................................................... 429 18.2 Pasteurized Processed Cheese Products .......................................... 429 18.3 Imitation and Substitute Cheese Products .......................................... 443

Page 7: Cheese Science

viii Contents

This page has been reformatted by Knovel to provide easier navigation.

19. Cheese as a Food Ingredient ........................................................ 452 19.1 Introduction .......................................................................................... 452 19.2 Overview of the Requirements of Cheese as an Ingredient ............... 452 19.3 Functional Properties of Cheese as an Ingredient .............................. 456 19.4 Dried Cheese Products ....................................................................... 475 19.5 Conclusion ........................................................................................... 482

20. Pathogens and Food-Poisoning Bacteria in Cheese .................. 484 20.1 Introduction .......................................................................................... 484 20.2 Pathogens in Raw Milk ........................................................................ 487 20.3 Pathogens in Cheese .......................................................................... 488 20.4 Listeriosis ............................................................................................. 489 20.5 Pathogenic Escherichia Coli ................................................................ 490 20.6 Growth of Pathogens during Cheese Manufacture ............................ 491 20.7 Growth of Pathogens in Cheese during Ripening .............................. 493 20.8 Raw Milk Cheeses ............................................................................... 498 20.9 Control of the Growth of Pathogens .................................................... 499 20.10 Enterococci .......................................................................................... 500 20.11 Biogenic Amines .................................................................................. 501

21. Nutritional Aspects of Cheese ...................................................... 504 21.1 Introduction .......................................................................................... 504 21.2 Fat and Cholesterol ............................................................................. 504 21.3 Protein and Carbohydrate ................................................................... 506 21.4 Vitamins and Minerals ......................................................................... 506 21.5 Additives in Cheese ............................................................................. 508 21.6 Cheese and Dental Caries .................................................................. 509 21.7 Mycotoxins ........................................................................................... 509 21.8 Biogenic Amines in Cheese ................................................................ 512

22. Whey and Whey Products ............................................................. 514 22.1 Introduction .......................................................................................... 514 22.2 Clarification of Whey ............................................................................ 515 22.3 Concentrated and Dried Whey Products ............................................ 516 22.4 Lactose ................................................................................................ 516

Page 8: Cheese Science

Contents ix

This page has been reformatted by Knovel to provide easier navigation.

22.5 Whey Proteins ..................................................................................... 517 22.6 Whey Cheese ...................................................................................... 519 22.7 Fermentation Products ........................................................................ 522 22.8 Conclusion ........................................................................................... 522

23. Analytical Methods for Cheese ..................................................... 523 23.1 Introduction .......................................................................................... 523 23.2 Methods of Sampling Cheese ............................................................. 523 23.3 Compositional Analysis ....................................................................... 525 23.4 Biochemical Assessment of Cheese Ripening ................................... 527 23.5 Techniques to Study Volatile Flavor Compounds ............................... 536 23.6 Microbiological Analysis of Cheese ..................................................... 536 23.7 Objective Assessment of Cheese Texture .......................................... 540 23.8 Sensory Analysis of Cheese Flavor and Texture ................................ 540 23.9 Detection of Interspecies Adulteration of Milks and Cheeses ............ 543

Table of Sources ................................................................................... 545

Index ...................................................................................................... 559

Page 9: Cheese Science

Fundamentals of Cheese Science

Patrick F. Fox, PhD5 DScProfessor, Food Chemistry

Food Science and TechnologyUniversity College, Cork

Cork, Ireland

Timothy P. Guinee, PhDSenior Research Officer

Dairy Products Research CentreTeagasc

Cork, Ireland

Timothy M. Cogan, PhDSenior Principal Research OfficerDairy Products Research Centre

TeagascVisiting Professor

University College, CorkCork, Ireland

Paul L. H. McSweeney, PhDStatutory Lecturer, Food Chemistry

Food Science and TechnologyUniversity College, Cork

Cork, Ireland

AN ASPEN PUBLICATIONAspen Publishers, Inc.

Gaithersburg, Maryland2000

Page 10: Cheese Science

The authors have made every effort to ensure the accuracy of the information herein. However, appropriateinformation sources should be consulted, especially for new or unfamiliar procedures. It is the responsibility ofevery practitioner to evaluate the appropriateness of a particular opinion in the context of actual clinical situationsand with due considerations to new developments. The author, editors, and the publisher cannot be held respon-sible for any typographical or other errors found in this book.

Aspen Publishers, Inc., is not affiliated with the American Society of Parenteral and Enteral Nutrition.

Library of Congress Cataloging-in-Publication Data

Fundamentals of Cheese SciencePatrick F. Fox. . . [etal.].

p. cm.Includes bibliographical references.

ISBN 0-8342-1260-9!.Cheese. I. Fox, P.F.

SF271.F862000637'.3—dc21

99-053386

Copyright © 2000 by Aspen Publishers, Inc.A Walters Kluwer Company

www.aspenpublishers.comAll rights reserved.

Aspen Publishers, Inc., grants permission for photocopying for limited personal or internal use.This consent does not extend to other kinds of copying, such as copying for general distribution,

for advertising or promotional purposes, for creating new collective works, or for resale.For information, address Aspen Publishers, Inc., Permissions Department,

200 Orchard Ridge Drive, Suite 200, Gaithersburg, Maryland 20878.

Orders: (800) 638-8437Customer Service: (800) 234-1660

About Aspen Publishers • For more than 40 years, Aspen has been a leading professional publisher ina variety of disciplines. Aspen's vast information resources are available in both print and electronicformats. We are committed to providing the highest quality information available in the most appropri-ate format for our customers. Visit Aspen's Internet site for more information resources, directories,articles, and a searchable version of Aspen's full catalog, including the most recent publications:www.aspenpublishers.com

Aspen Publishers, Inc. • The hallmark of quality in publishingMember of the worldwide Wolters Kluwer group.

Editorial Services: Jane ColillaLibrary of Congress Catalog Card Number: 99-053386

ISBN: 0-8342-1260-9Printed in the United States of America

1 2 3 4 5

Page 11: Cheese Science

Preface

Cheese, which has been produced for about5,000 years, is one of the classical fabricatedfoods in the human diet. During its long history,the volume and diversity of cheese productionhave increased such that today annual produc-tion is about 15 x 106 tonnes (representing about35% of total world milk production) in at least500 varieties. Cheese is one of our most complexand dynamic food products, and its study in-volves a wide range of disciplines, especiallyanalytical and physical chemistry, biochemistry,microbiology, rheology, and sensory science.Not surprisingly, a large and diverse literatureon the science and technology of cheese has ac-cumulated over the past 100 years. Like in thecase of the other great fermented foods, wineand beer, the epicurean attributes of cheese at-tract the attention of consumers and endow itwith a certain social status. Cheese is a highlynutritious food with a very positive image. It isthe quintessential consumer-ready food, yet it isone of the most flexible food ingredients. Inmany respects, cheese is the ideal food: nutri-tious, flexible in use and application, and senso-rially appealing to a wide range of consumers.

Cheese has been the subject of a considerablenumber of books, but most of these were writtenwith the general reader in mind (see "SuggestedReadings" at the end of Chapter 1). There are atleast three books on cheese technology, but thescientific aspects of cheese have been less well

covered, the only exception being the two-vol-ume set Cheese: Chemistry, Physics and Micro-biology, edited by P.P. Fox. That set, though, as-sumes a substantial background knowledge onthe part of its readers.

Fundamentals of Cheese Science providescomprehensive coverage of the scientific as-pects of cheese, appropriate for anyone workingwith cheese, from researchers and professionalsto undergraduate and graduate students in foodscience and technology. The book assumes fa-miliarity with biochemistry, microbiology, anddairy chemistry, and it emphasizes fundamentalprinciples rather than technological aspects.

The book is divided into 23 chapters that dealwith the chemistry and microbiology of milk forcheesemaking, starter cultures, coagulation ofmilk by enzymes or by acidification, the micro-biology and biochemistry of cheese ripening, theflavor and rheology of cheese, processed cheese,cheese as a food ingredient, public health andnutritional aspects of cheese, and various meth-ods used for the analysis of cheese. The bookcontains copious references to other texts andreview articles, but references to the primary lit-erature are kept to a minimum to facilitate easypresentation.

Finally, the authors would like to express theirappreciation for the highly skilled and enthusi-astic assistance of Ms. Anne Cahalane in thepreparation of the manuscript.

Page 12: Cheese Science

Fundamentals of Cheese Science

Patrick F. Fox, PhD5 DScProfessor, Food Chemistry

Food Science and TechnologyUniversity College, Cork

Cork, Ireland

Timothy P. Guinee, PhDSenior Research Officer

Dairy Products Research CentreTeagasc

Cork, Ireland

Timothy M. Cogan, PhDSenior Principal Research OfficerDairy Products Research Centre

TeagascVisiting Professor

University College, CorkCork, Ireland

Paul L. H. McSweeney, PhDStatutory Lecturer, Food Chemistry

Food Science and TechnologyUniversity College, Cork

Cork, Ireland

AN ASPEN PUBLICATIONAspen Publishers, Inc.

Gaithersburg, Maryland2000

Page 13: Cheese Science

The authors have made every effort to ensure the accuracy of the information herein. However, appropriateinformation sources should be consulted, especially for new or unfamiliar procedures. It is the responsibility ofevery practitioner to evaluate the appropriateness of a particular opinion in the context of actual clinical situationsand with due considerations to new developments. The author, editors, and the publisher cannot be held respon-sible for any typographical or other errors found in this book.

Aspen Publishers, Inc., is not affiliated with the American Society of Parenteral and Enteral Nutrition.

Library of Congress Cataloging-in-Publication Data

Fundamentals of Cheese SciencePatrick F. Fox. . . [etal.].

p. cm.Includes bibliographical references.

ISBN 0-8342-1260-9!.Cheese. I. Fox, P.F.

SF271.F862000637'.3—dc21

99-053386

Copyright © 2000 by Aspen Publishers, Inc.A Walters Kluwer Company

www.aspenpublishers.comAll rights reserved.

Aspen Publishers, Inc., grants permission for photocopying for limited personal or internal use.This consent does not extend to other kinds of copying, such as copying for general distribution,

for advertising or promotional purposes, for creating new collective works, or for resale.For information, address Aspen Publishers, Inc., Permissions Department,

200 Orchard Ridge Drive, Suite 200, Gaithersburg, Maryland 20878.

Orders: (800) 638-8437Customer Service: (800) 234-1660

About Aspen Publishers • For more than 40 years, Aspen has been a leading professional publisher ina variety of disciplines. Aspen's vast information resources are available in both print and electronicformats. We are committed to providing the highest quality information available in the most appropri-ate format for our customers. Visit Aspen's Internet site for more information resources, directories,articles, and a searchable version of Aspen's full catalog, including the most recent publications:www.aspenpublishers.com

Aspen Publishers, Inc. • The hallmark of quality in publishingMember of the worldwide Wolters Kluwer group.

Editorial Services: Jane ColillaLibrary of Congress Catalog Card Number: 99-053386

ISBN: 0-8342-1260-9Printed in the United States of America

1 2 3 4 5

Page 14: Cheese Science

Preface

Cheese, which has been produced for about5,000 years, is one of the classical fabricatedfoods in the human diet. During its long history,the volume and diversity of cheese productionhave increased such that today annual produc-tion is about 15 x 106 tonnes (representing about35% of total world milk production) in at least500 varieties. Cheese is one of our most complexand dynamic food products, and its study in-volves a wide range of disciplines, especiallyanalytical and physical chemistry, biochemistry,microbiology, rheology, and sensory science.Not surprisingly, a large and diverse literatureon the science and technology of cheese has ac-cumulated over the past 100 years. Like in thecase of the other great fermented foods, wineand beer, the epicurean attributes of cheese at-tract the attention of consumers and endow itwith a certain social status. Cheese is a highlynutritious food with a very positive image. It isthe quintessential consumer-ready food, yet it isone of the most flexible food ingredients. Inmany respects, cheese is the ideal food: nutri-tious, flexible in use and application, and senso-rially appealing to a wide range of consumers.

Cheese has been the subject of a considerablenumber of books, but most of these were writtenwith the general reader in mind (see "SuggestedReadings" at the end of Chapter 1). There are atleast three books on cheese technology, but thescientific aspects of cheese have been less well

covered, the only exception being the two-vol-ume set Cheese: Chemistry, Physics and Micro-biology, edited by P.P. Fox. That set, though, as-sumes a substantial background knowledge onthe part of its readers.

Fundamentals of Cheese Science providescomprehensive coverage of the scientific as-pects of cheese, appropriate for anyone workingwith cheese, from researchers and professionalsto undergraduate and graduate students in foodscience and technology. The book assumes fa-miliarity with biochemistry, microbiology, anddairy chemistry, and it emphasizes fundamentalprinciples rather than technological aspects.

The book is divided into 23 chapters that dealwith the chemistry and microbiology of milk forcheesemaking, starter cultures, coagulation ofmilk by enzymes or by acidification, the micro-biology and biochemistry of cheese ripening, theflavor and rheology of cheese, processed cheese,cheese as a food ingredient, public health andnutritional aspects of cheese, and various meth-ods used for the analysis of cheese. The bookcontains copious references to other texts andreview articles, but references to the primary lit-erature are kept to a minimum to facilitate easypresentation.

Finally, the authors would like to express theirappreciation for the highly skilled and enthusi-astic assistance of Ms. Anne Cahalane in thepreparation of the manuscript.

Page 15: Cheese Science

1.1 INTRODUCTION

Cheese is the generic name for a group of fer-mented milk-based food products producedthroughout the world in a great diversity of fla-vors, textures, and forms. Sandine and Elliker(1970) suggest that there are more than 1,000varieties of cheese. Walter and Hargrove (1972)describe about 400 varieties and list the namesof a further 400, while Burkhalter (1981) classi-fies 510 varieties.

It is commonly believed that cheese evolvedin the Fertile Crescent between the Tigris andEuphrates rivers, in Iraq, some 8,000 years ago,during the so-called Agricultural Revolution,when certain plants and animals were domesti-cated as sources of food. Among the earliest ani-mals domesticated were goats and sheep; beingsmall, gregarious, and easily herded, these wereused to supply meat, milk, hides, and wool.Cattle were more difficult to domesticate; wildcattle were much larger and more ferocious thanmodern cattle and were also less well adjusted tothe arid Middle East than goats and sheep. Ap-parently, cattle were used mainly as work ani-mals (as they still are) and did not become a ma-jor source of milk until relatively recently. Mansoon recognized the nutritive value of milk pro-duced by domesticated animals, and milk anddairy products became important components ofthe human diet.

Milk is also a rich source of nutrients forbacteria that contaminate the milk and grow

well under ambient conditions. Some contami-nating bacteria utilize milk sugar, lactose, as asource of energy, producing lactic acid as abyproduct; these bacteria, known as lactic acidbacteria (LAB), include the genera Lacto-coccus, Lactobacillus, Streptococcus, Entero-coccus, Leuconostoc, and Pediococcus. LABare used in the production of a wide range offermented milk, meat, and vegetable products.They are generally considered to be beneficialto human health and have been studied exten-sively (see Chapter 5).

Bacterial growth and acid production wouldhave occurred in milk during storage or duringattempts to dry milk in the prevailing warm, dryclimate of the Middle East to produce a morestable product; air-drying of meat, and probablyfruits and vegetables, appears to have been prac-ticed as a primitive form of food preservation atthis period of human evolution. When sufficientacid is produced, the principal proteins in milk,the caseins, coagulate at ambient temperature(210C) in the region of their isoelectric points(~ pH 4.6) to form a gel in which the fat andaqueous phases of milk are entrapped. Thus, thefirst fermented dairy foods were probably pro-duced accidentally. Numerous, basically similarproducts are produced in various regions of theworld by artisanal methods probably little differ-ent from those used several thousand years ago.Some descendants of these ancient fermentedmilks are now produced by scientifically basedtechnology in sophisticated factories.

Cheese: Historical Aspects

CHAPTER 1

Page 16: Cheese Science

The first fermented dairy foods were pro-duced by a fortuitous combination of events—the growth in milk of a group of lactic acid bac-teria produced just enough acid to reduce the pHof the milk to the isoelectric point of the caseins,causing these proteins to coagulate. Neither thelactic acid bacteria nor the caseins were "de-signed" for this function. The caseins were de-signed to be coagulated enzymatically in thestomach of neonatal mammals, the gastric pH ofwhich is around 6 (i.e., very much higher thanthe isoelectric point of the caseins). The abilityof LAB to ferment lactose, a sugar specific tomilk, is frequently encoded on plasmids, sug-gesting that this characteristic was acquired rela-tively recently in the evolution of these bacteria.Their natural habitats are vegetation, from whichthey presumably colonized the teats of mammalscontaminated with lactose-containing milk.

An acid-induced milk gel is quite stable if leftundisturbed, but if it is broken, either acciden-tally (e.g., by movement of the storage vessels)or intentionally, it separates into curds andwhey. It would have soon been realized that theacid whey is a pleasant, refreshing drink for im-mediate consumption, whereas the curds couldbe consumed fresh or stored for future use. Itwas probably also realized that the shelf-life ofthe curds could be greatly extended by dehydra-tion and/or by adding salt; heavily salted cheesevarieties (e.g., Feta and Domiati) are still wide-spread throughout the Middle East and theBalkans, where the ambient temperature is high.Air- or sun-dried varieties of cheese are lesscommon today, but numerous examples survivethroughout the hot, dry areas of North Africa andthe Middle East.

Today, acid-coagulated cheeses, which in-clude Cottage cheese, Cream cheese, Quarg,Fromage frais, and some varieties of Quesobianco, represent approximately 25% of totalcheese production and in some countries are theprincipal varieties. They are consumed fresh(not ripened) and are widely used in other prod-ucts (e.g., cheesecake, cheese-based dips, andsauces).

An alternative mechanism for coagulatingmilk was also discovered at an early date. Manyproteolytic enzymes can modify the milk proteinsystem, causing it to coagulate under certain cir-cumstances. Enzymes capable of causing thistransformation are widespread and are found inbacteria, molds, plants, and animal tissues, butthe most obvious source would have been ani-mal stomachs. It would have been observed thatthe stomachs of young slaughtered animals fre-quently contained curds, especially if the ani-mals had suckled shortly before slaughter; curdswould also have been observed in the vomit ofhuman infants. Before the development of pot-tery (about 5,000 B.C.), storage of milk in bagsmade from animal skins was probably common(it is still practiced in many countries). Stomachsfrom slaughtered animals provided ready-made,easily sealed containers; if stored in such con-tainers, milk would extract coagulating enzymes(referred to as rennets) from the stomach tissue,leading to its coagulation during storage. Theproperties of rennet-coagulated milk curds arevery different from those of curds produced byisoelectric (acid) precipitation. For example,they have better syncretic (curd-contracting)properties, making it possible to produce low-moisture cheese curd without hardening. Ren-net-coagulated curds can, therefore, be con-verted to more stable low-moisture productsthan can acid curd, and rennet coagulation hasbecome the principal mechanism for milk co-agulation in cheese manufacture. Most moderncheese varieties and approximately 75% of totalworld production of cheese are produced by thismechanism.

During the storage of rennet-coagulatedcurds, various bacteria grow, and the enzymes inrennet continue to act. Thus, the flavor and tex-ture of the cheese curds change during storage.When controlled, this process is referred to asripening (maturation), during which a great di-versity of characteristic flavors and textures de-velop. Although animal rennets were probablythe first enzyme coagulants used, rennets pro-duced from a range of plant species (e.g., figs

Page 17: Cheese Science

and thistle) appear to have been common in Ro-man times. However, plant rennets are not suit-able for the manufacture of long-ripened cheesevarieties, and gastric proteinases from younganimals became the standard rennets until ashortage of supply made it necessary to intro-duce rennet substitutes, which are discussed inChapter 6.

While the coagulation of milk by the in situproduction of lactic acid was, presumably, acci-dental, the use of rennets to coagulate milk wasintentional. It was, in fact, quite an ingeniousdevelopment—if the conversion of milk tocheese by the use of rennets was discovered to-day, it would be hailed as a major biotechnologi-cal discovery! The use of rennets in cheese manu-facture is probably the oldest and is still one ofthe principal industrial applications of enzymes.

The advantages accruing from the ability toconvert the principal constituents of milk tocheese would have been readily apparent: stor-age stability, ease of transport, and diversifica-tion of the human diet. Cheese manufacture ac-companied the spread of civilization throughoutthe Middle East, Egypt, Greece, and Rome.There are several references in the Old Testa-ment to cheese, such as in Job (1520 B.C.) andSamuel (1170-1017 B.C.); on the walls of An-cient Egyptian tombs; and in classical Greek lit-erature, including Homer (12th century B.C.),Herodotus (484-408 B.C.), and Aristotle (384-322 B.C.). Cheese manufacture was well estab-lished at the time of the Roman Empire, andcheese was included in the rations of Roman sol-diers. The demand for cheese in Rome musthave exceeded supply, since the EmperorDiocletian (A.D. 284-305) fixed a maximumprice for cheese. Many Roman writers, includ-ing Cato (about 150 B.C.), Varro (about 40 B.C.),Columella (A.D. 50), and Pliny (A.D. 23-89), de-scribed the manufacture, quality attributes, andculinary uses of cheese. Columella, in particular,gave a detailed account of cheese manufacture inhis treatise on agriculture, De Re Rustica.

Movements of Roman armies and administra-tors contributed to the spread of cheese through-

out the then known world. Although archeologi-cal evidence suggests that cheese may have beenmanufactured in pre-Roman Britain, the first un-equivocal evidence credits the Romans with theestablishment of cheesemaking in Britain.Palladius wrote a treatise on Roman-Britishfarming in the 4th century A.D., which included adescription of and advice on cheesemaking.Cheesemaking practice appears to have changedlittle from the time of Columella and Palladiusuntil the 19th century.

The great migrations of peoples throughoutEurope after the fall of the Roman Empire prob-ably promoted the spread of cheese manufac-ture, as did the Crusaders and pilgrims of theMiddle Ages. However, the most important con-tributors to the development of cheese technol-ogy and to the evolution of cheese varieties dur-ing the Middle Ages were the monasteries andfeudal estates. In addition to their roles in thespread of Christianity and in the preservationand expansion of knowledge during the DarkAges, the monasteries were major contributorsto the advancement of agriculture in Europe andto the development and improvement of foodcommodities, notably wine, beer, and cheese.Many current cheese varieties were developed inmonasteries, such as Wenslydale (Rievaulx Ab-bey, Yorkshire), Port du Salut or St. Paulin(Monastery of Notre Dame du Port du Salut,Laval, France), Fromage de Tamie (Abbey ofTamie, Lac d'Annecy, France), Maroilles (Ab-bey of Maroilles, Avesnes, France), and Trappist(Maria Stern Monastery, Banja Luka, Bosnia).The intermonastery movement of monks prob-ably contributed to the spread of cheese varietiesand to the development of new hybrid cheeses.

The great feudal estates of the Middle Ageswere self-contained communities that, in the ab-sence of an effective transport system, relied ona supply of locally produced foods. Surplus foodwas produced in summer and preserved to meetthe requirements of the community throughoutthe year. Especially in cool, wet Europe, fermen-tation and salting were the most effective meth-ods of food preservation. Well-known examples

Page 18: Cheese Science

of products preserved by these methods includefermented and salted meat, salted fish, beer,wine, fermented vegetables, and cheese (themanufacture of which exploits both fermenta-tion and salting). Cheese probably representedan item of trade when amounts beyond local re-quirements were available.

Within large estates, individuals acquiredspecial skills that were passed on to succeedinggenerations. The feudal estates evolved intovillages and some into larger communities.Because monasteries and feudal estates wereessentially self-contained communities withlimited intercommunity travel, it is readily ap-parent how several hundred distinct varieties ofcheese could have evolved from essentially thesame raw material. Traditionally, cheese variet-ies were produced in quite limited geographicalregions, especially in mountainous areas. Thelocalized production of certain varieties is stillapparent and indeed is preserved through thedesignation of Appellation d'Origine Controlee.The regionalization of certain cheese varieties isstill particularly marked in Spain, Italy, andFrance, where the production of many varietiesis restricted to very limited, sometimes legallydefined regions. Almost certainly, most cheesevarieties evolved by accident because of par-ticular local circumstances (e.g., a local speciesor breed of dairy animal or a peculiarity of thelocal milk supply with respect to chemical com-position or microflora) or because of an unin-tended event during the manufacture or storageof the cheese (e.g., growth of molds or other mi-croorganisms). Presumably, the accidents thatled to desirable changes in the quality of thecheese were incorporated into the manufactur-ing protocol; each variety would thus have un-dergone a series of evolutionary changes and re-finements.

The final chapter in the spread of cheesethroughout the world resulted from the coloniza-tion of North and South America, Oceania, andAfrica by European settlers who carried theircheesemaking skills with them. Cheese has be-come an item of major economic importance insome of these non-European countries, notably

the United States, Canada, Australia, and NewZealand, but the varieties produced are mainlyof European origin, modified in some cases tomeet local conditions and requirements. There isno evidence that cheese was produced in theAmericas or Oceania prior to colonization; infact, animals had not been domesticated for milkproduction in these countries.

Cheesemaking remained a craft until rela-tively recently. With the gradual acquisition ofknowledge about the chemistry and microbiol-ogy of milk and cheese, it became possible togain more control over the cheesemaking pro-cess. Few new varieties have evolved as a resultof the increased knowledge, but existing variet-ies have become better defined and their qualityhas become more consistent. Although thenames of many current varieties were introducedseveral hundred years ago (Table 1-1), it is verylikely that those cheeses were not comparable totheir modern counterparts. Cheesemaking wasnot standardized until relatively recently. Forexample, the first attempt to standardize thewell-known English varieties, Cheddar andCheshire, was made by John Harding in themiddle of the 19th century. Prior to that, "Ched-dar cheese" was cheese produced around the vil-lage of Cheddar, in Somerset, England, and

Table 1-1 First Recorded Date for Some MajorCheese Varieties

Variety Year

Gorgonzola 897Schabzieger 1000Roquefort 1070Maroilles 1174Schwangenkase 1178Grana 1200Taleggio 1282Cheddar 1500Parmesan 1579Gouda 1697Gloucester 1783Stilton 1785Camembert 1791St. Paulin 1816

Page 19: Cheese Science

probably varied considerably depending on thecheesemaker and other factors. Cheese manu-facture during most of its history was a farm-stead enterprise. The first cheese factory in theUnited States was established near Rome, NewYork, in 1851, and the first in Britain atLongford, Derbyshire, in 1870. There werethousands of small-scale cheese manufacturers,and there must have been great variation withinany one general type of cheese. When one con-siders the very considerable interfactory and in-deed intrafactory variation in quality and charac-teristics that still occurs today in well-definedvarieties (e.g., Cheddar) in spite of the very con-siderable scientific and technological advances,one can readily appreciate the variation thatmust have existed in earlier times.

Some major new varieties, notably Jarlsbergand Maasdamer, have been developed recentlyas a consequence of scientific research. Manyother varieties have evolved considerably, evento the extent of becoming new varieties, as aconsequence of scientific research and the de-velopment of new technology. Notable ex-amples are Queso bianco as produced in theUnited States, Feta-type cheese produced fromultrafiltered milk, and various forms of Quarg.There has been a marked resurgence of farm-house cheesemaking in recent years. Many ofthe cheeses being produced on farms are notstandard varieties, and it will be interesting tosee if some of these evolve to become major newvarieties.

A main cause of variation in the characteris-tics of cheese is the species from which the milkis produced. Although milks from several spe-cies are used by humans, the cow is by far theprincipal producer. Worldwide, 85% of com-mercial milk is bovine. However, goats, sheep,and water buffalo are significant producers ofmilk in certain regions (e.g., the Mediterraneanbasin and India). Goats and sheep are especiallyimportant in cheese production, since the milksof these species are used mainly for the produc-tion of fermented milks and cheese. Manyworld-famous cheeses are produced fromsheep's milk (e.g., Roquefort, Feta, Romano,

and Manchego). Traditional Mozzarella is madefrom the milk of the water buffalo. As discussedin Chapter 3, there are very significant inters-pecies differences in the composition of milkthat are reflected in the characteristics of thecheeses produced from them. There are also sig-nificant differences in milk composition be-tween breeds of cattle, and these influencecheese quality, as do variations due to seasonal,lactational, and nutritional factors and of coursethe methods of milk production, storage, andcollection.

1.2 CHEESE PRODUCTION ANDCONSUMPTION

World production of cheese is roughly 15 x106 tonnes per annum (about 35% of total milkproduction) and has increased at an average an-nual rate of about 4% over the past 30 years. Eu-rope, with a production of roughly 8 x 106 tonnesper annum, is by far the largest producing block(Table 1-2). Thus, while cheese manufacture ispracticed worldwide, it is apparent from Table1-2 that cheese is primarily a product of Euro-pean countries and those populated mostly byEuropean immigrants.

Cheese consumption varies widely betweencountries, even within Europe (Table 1-3).Cheese consumption in most countries for whichdata are available has increased consistentlyover many years. Along with fermented milks,cheese is the principal growth product within thedairy sector. There are many reasons for theincreased consumption of cheese, including apositive dietary image, convenience and flex-ibility in use, and the great diversity of flavorsand textures. Cheese can be regarded as thequintessential convenience food: it can be usedas a major component of a meal, as a dessert, as acomponent of other foods, or as a food ingredi-ent; it can be consumed without preparation orsubjected to various cooking processes. Themost rapid growth in cheese consumption hasoccurred in its use as a food component or ingre-dient; these applications are discussed in Chap-ter 19.

Page 20: Cheese Science

Table 1-2 World Production of Cheese, 1994

CountryCheese

Production(1,000 Tonnes)

World

AfricaAlgeriaAngolaBotswanaEgyptEthiopiaMauritaniaMoroccoNigerNigeriaSouth AfricaSudanTanzaniaTunisiaZambiaZimbabwe

North and CentralAmericaCanadaCosta RicaCubaDominican

RepublicEl SalvadorGuatemalaHondurasMexicoNicaraguaPanamaUnited States

South AmericaArgentinaBoliviaBrasilChile

15,084

511112

349327

147

37762612

4,130323

615

33

118

12367

3,627

677405

76051

CountryCheese

Production(1,000 Tonnes)

ColombiaEcuadorPeruUruguayVenezuela

AsiaAfghanistanArmeniaAzerbaijanBangladeshChinaCyprusGeorgiaIranIraqIsraelJapanJordanKazakhstanKyrgyzstanLebanonMongoliaMyanmarSyriaTajikistanTurkeyTurkmenistanUzbekistanYemen

EuropeAlbaniaAustriaBelarusBelgium-

LuxembourgBosnia-

Herzgovina

5176

2370

1,01816971

20653

1971792

1144

345

151

2986

1139

131410

8,2011

10239

75

14

CountryCheese

Production(1,000 Tonnes)

BulgariaCroatiaCzech RepublicDenmarkEstoniaFinlandFranceGermanyGreeceHungaryIcelandIrelandItalyLatviaLithuaniaMacedonia,

FYR ofMoldova

RepublicNetherlandsNorwayPolandPortugalRomaniaRussian

FederationSlovakiaSlovaniaSpainSwedenSwitzerlandUnited KingdomUkraineYugoslavia, FR

OceaniaAustraliaNew Zealand

7221

139291

1889

1,6051,569

216883

911,017

1127

1

3688

893976542

4774416

1631281333857114

544270274

/Vote: The following countries were included by the Food and Agriculture Organization in 1997, but no data for cheese productionare available: Burkina Faso, Madagascar, Somalia, Jamaica, Trinidad and Tobago, India, Indonesia, Republic of Korea, Malaysia,Eritrea, Kenya, Namibia, Bhutan, Oman, Malta, Nepal, Pakistan, Philippines, Saudi Arabia, Thailand, Sri Lanka, United Arab Emir-ates, and Fiji.

Page 21: Cheese Science

Table 1-3 Consumption of Cheese, 1993 (Kilograms per Person per Year)

Country

FranceGreeceItalyBelgiumGermanyLithuaniaIcelandSwitzerlandSwedenLuxembourgNetherlandsDenmarkFinlandNorwayCanadaUnited StatesAustriaCzech and Slovak RepublicsEstoniaAustraliaUnited KingdomNew ZealandHungaryRussiaSpainIrelandChileSouth AfricaJapanIndia

Fresh

7.50.26.74.78.0

11.65.22.80.95.01.70.92.30.20.91.33.94.05.6

3.32.8

2.00.10.20.2

Ripened

15.521.813.415.110.56.8

11.913.615.511.314.114.512.014.012.411.97.56.64.4

4.64.9

2.01.51.2

Total

22.822.020.119.818.518.417.116.416.416.315.815.414.314.213.313.211.410.610.08.88.38.17.97.77.05.64.01.61.40.2

1.3 CHEESE SCIENCE ANDTECHNOLOGY

Cheese is the most diverse group of dairyproducts and is, arguably, the most academicallyinteresting and challenging. While many dairyproducts, if properly manufactured and stored,are biologically, biochemically, and chemicallystable, cheeses are, in contrast, biologically andbiochemically active and consequently undergochanges in flavor, texture, and functionality dur-ing storage. Throughout manufacture and ripen-ing, cheese production represents a finelyorchestrated series of consecutive and concom-

itant biochemical events that, if synchronizedand balanced, lead to products with highly desir-able aromas and flavors, but, if unbalanced, re-sult in off-flavors and off-odors. Consideringthat a basically similar raw material (milk from avery limited number of species) is subjected to amanufacturing protocol, the general principlesof which are common to most cheese varieties, itis fascinating that such a diverse range of prod-ucts can be produced. No two batches of thesame variety are identical.

A further important facet of cheese produc-tion is the range of scientific disciplines in-volved: the study of cheese manufacture and rip-

Page 22: Cheese Science

ening encompasses the chemistry and biochem-istry of milk constituents, the chemical charac-terization of cheese constituents, microbiology,enzymology, molecular genetics, flavor chemis-try, rheology, and chemical engineering. It is notsurprising, therefore, that many scientists havebecome involved in the study of cheese manu-facture and ripening. A voluminous scientific

REFERENCES

Burkhalter, G. (1981). Catalogue of cheese (Bulletin 141).Brussels: International Dairy Federation.

Sandine, W.E., & Elliker, P.R. (1970). Microbially inducedflavors and fermented foods: Flavor in fermented dairy

SUGGESTED READINGS

Andrews, A.T., & Varley, J. (1994). Biochemistry of milkand milk products. Cambridge: Royal Society of Chemis-try.

Anifantakis, E.M. (1991). Greek cheeses: A tradition of cen-turies. Athens: National Dairy Committee of Greece.

Berger, W., Klostermeyer, H., Merkenich, K., & Uhlmann,G. (1989). Die Schmelzkdselerstellung. Ledenburg, Ger-many: Benckiser-Knapsack GmbH.

Cantin, C. (1976). Guide pratique des frontages. Paris: SolarEditeur.

Cheke, V. (1959). The story of cheesemaking in Britain.London: Routledge & Kegan Paul.

Davis, J.G. (1965-1967). Cheese (VoIs. 1-4). London:Churchill Livingstone.

DOC cheeses of Italy: A great heritage. (1992). Milan:Franco Angeli.

Eck, A., & Gilles, J.C. (1997). Le frontage (2d ed.). Paris:Technique et Documentation (Lavoisier).

Eekhof-Stork, N. (1976). World atlas of cheese. London:Paddington Press.

Fox, P.F. (1993). Cheese: Chemistry, physics and microbiol-ogy (2d ed., VoIs. 1, 2). London: Chapman & Hall.

Glynn Christian's world guide to cheese. (1984). (S. Harris,trans.). London: Ebury Press.

Gonzalez, M.A., & del Cerro, C.G. (1988). Quesos deEspana. Madrid: Espasa-Calpe.

Kosikowski, F.V., & Mistry, V.V. (1997). Cheese and fer-mented milks (3d ed., VoIs. 1, 2). Westport, CT: F.V.Kosikowski LLC.

and technological literature has accumulated, in-cluding several textbooks (see "SuggestedReadings") and chapters in many others. Manyof these textbooks deal mainly with cheese tech-nology or assume an overall knowledge ofcheese. The present book is intended to providea fairly comprehensive treatment of the scien-tific aspects of cheese.

products. Journal of Agricultural and Food Chemistry,18, 557-566.

Walter, H.E., & Hargrove, R.C. (1972). Cheeses of theworld. New York: Dover.

Kosikowski, F.V., & Mocquot, G. (1958). Advances incheese technology. (Food & Agriculture OrganizationStudy No. 38). Rome: Food & Agriculture Organization.

Law, B. A. (1997). Microbiology and biochemistry of cheeseand fermented milks (2d ed.). London: Chapman & Hall.

Layton, J.A. (1973). The cheese handbook. New York: Do-ver.

Lembo, P., & Spedicato, E. (1992). / prodotti caseari delmezzogiorno. Rome: Consiglio Nazionale delle Ricerche.

Mair-Waldburg, H. (1974). Handbook of cheese; Cheeses ofthe world A to Z. Kempten Allgan, Germany: VoIk-wertschaftlecher Verlag GmbH.

Malin, E.L., & Tunick, M.H. (1995). Chemistry of structure-function relationships in cheese. New York: PlenumPress.

Masui, K., & Yamada, T. (1996). French cheeses. London:Dorling Kindersley.

Meyer, A. (1973). Processed cheese manufacture. London:Food Trade Press.

Robinson, R.K. (1993). Modern dairy technology (2d ed.,Vol. 2). London: Elsevier Applied Science.

Robinson, R.K. (1995). A colour guide to cheese and fer-mented milks. London: Chapman & Hall.

Robinson, R.K., & Tamime, A.Y. (1991). Feta and relatedcheeses. London: Ellis Horwood.

Robinson, R.K., & Wilbey, R.A. (1998). Cheesemakingpractice (3d ed.). Gaithersburg, MD: Aspen Publishers.

Sammis, J.L. (1948). Cheesemaking. Madison, WI: Cheese-maker Book Co.

Page 23: Cheese Science

Scott, R. (1986). Cheesemaking practice (2d ed.). London:Elsevier Applied Science.

Simon, A.L. (1956). Cheeses of the world. London: Faber &Faber.

Squire, E.H. (1937). Cheddar gorge: A book of Englishcheeses. London: Collins,

van Slyke, L.L., & Price, W.V. (1949). Cheese. New York:Orange Judd.

Page 24: Cheese Science

The production of all varieties of cheese in-volves a generally similar protocol (Figure 2-1),various steps of which are modified to give aproduct with the desired characteristics. Theprincipal steps will be described in individualchapters. The objective of this chapter is topresent a very brief description of the principaloperations so that the operations described in thefollowing chapters can be seen in an overall con-text.

2.1 SELECTION OF MILK

The composition of cheese is strongly influ-enced by the composition of the cheese milk, es-pecially the content of fat, protein, calcium, andpH. The constituents of milk, which are de-scribed in Chapter 3, are influenced by severalfactors, including the species, breed, individual-ity, nutritional status, health, and stage of lacta-tion of the producing animal. Owing to majorcompositional abnormalities, milk from cows inthe very early or late stages of lactation andthose suffering from mastitis should be ex-cluded. Somatic cell (leucocyte) count is a use-ful index of quality. Some genetic polymorphsof the milk proteins have a significant effect oncheese yield and quality, and there is increasinginterest in breeding for certain polymorphs. Themilk should be free of chemical taints and freefatty acids, which cause off-flavors in thecheese, and antibiotics, which inhibit bacterialcultures.

The milk should be of good microbiologicalquality, as contaminating bacteria will be con-centrated in the cheese curd and may cause de-fects or public health problems (see Chapter 4).

2.2 STANDARDIZATION OF MILKCOMPOSITION

Milk for cheese is subjected to a number ofpretreatments, with various objectives.

Different cheese varieties have a certain fat-in-dry-matter content, in effect, a certainfatprotein ratio, and this content has legal statusin the "Standards of Identity" for many cheesevarieties. While the moisture content of cheese,and hence the level of fat plus protein, is deter-mined mainly by the manufacturing protocol,the fatiprotein ratio is determined mainly by thefatcasein ratio in the cheese milk. Depending onthe ratio required, it can be modified by

• removing some fat by natural creaming, as inthe manufacture of Parmigiano-Reggiano,or centrifugation

• adding skim milk• adding cream• adding milk powder, evaporated milk, or

ultrafiltration retentate (such additions alsoincrease the total solids content of the milkand hence cheese yield, as discussed inChapter 9)

Calcium plays a major role in the coagulationof milk by rennet and the subsequent processing

Overview of Cheese Manufacture

CHAPTER 2

Page 25: Cheese Science

of the coagulum, and hence it is common prac-tice to add CaCl2 (e.g., 0.01%) to cheese milk.

The pH of milk is a critical factor in cheese-making. The pH is inadvertently adjusted by theaddition of 1.5-2% starter culture, which re-duces the pH of the milk immediately by about

0.1 unit. Starter concentrates (sometimes calleddirect-to-vat starters), which are sometimesused, have no immediate acidifying effect.

Previously, it was standard practice to add thestarter to the cheese milk 30-60 min before ren-net addition. During this period, the starter be-

Figure 2-1 General protocol for cheese manufacture.

Mature Cheese

Salting (most varieties)Ripening (most rennet-coagulated cheeses)

Fresh Cheese

AcidificationSpecial operations(e.g., cheddaring, stretching)Salting (some varieties)MoldingPressing (some varieties)

Curds

Cut coagulumStirHeatAcidification (rennet-coagulated cheeses)Separation of curds from whey

Coagulum (gel)

Addition of:starter culture (acidification)color (optional)CaCl2 (optional)

Coagulation (rennet or acid [produced in situ orpre-formed] or heat/acid)

Cheese milk

SelectionPre-treatmentStandardization

Milk

Page 26: Cheese Science

gan to grow and produce acid, a process referredto as ripening. Ripening served a number offunctions:

• It allowed the starter bacteria to enter theirexponential growth phase and hence to behighly active during cheesemaking; this isnot necessary with modern high-qualitystarters.

• The lower pH was more favorable for ren-net action and gel formation.

However, the practice increases the risk ofbacteriophage infection of the starter as phagebecome distributed throughout the liquid milkbut are locked in position after it has coagulated(see Chapter 5). Although ripening is still prac-ticed for some varieties, it has been discontinuedfor most varieties.

The pH of milk on reception at the dairy ishigher today than it was previously owing to im-proved hygiene during milking and the wideruse of refrigeration at the farm and factory. Inthe absence of acid production by contaminatingbacteria, the pH of milk increases slightly duringstorage due to the loss of CO2 to the atmosphere.The natural pH of milk is about 6.6 but variessomewhat (e.g., it increases in late lactation andduring mastitic infection).

To offset these variations and to reduce thepH as an alternative to ripening, the preacid-iflcation of milk by 0.1-0.2 pH units is recom-mended, either through the use of the acidogengluconic acid-5-lactone, or by limited growth ofa lactic acid starter, followed by pasteurization(referred to as prematuratiori). Such a practiceresults in a gel with more uniform characteris-tics, reflected in the production of cheese ofmore uniform quality.

2.3 HEAT TREATMENT OF MILK

Traditionally, all cheese was made from rawmilk, a practice that remained widespread untilthe 1940s. Even today, significant amounts ofcheese are made from raw milk in Europe. Theuse of raw milk is undesirable for two reasons:

1. dangers to public health2. the presence of undesirable microorgan-

isms, which may cause defects in flavorand/or texture

When cheese was produced from fresh milkon farms or in small local factories, the growthof contaminating microorganisms was minimal,but as cheese factories became larger, storage ofmilk for longer periods became necessary, andhence the microbiological quality of the milkvaried. For public health reasons, it became in-creasingly popular from the beginning of thiscentury to pasteurize milk for liquid consump-tion.

The pasteurization of cheese milk becamewidespread about 1940, primarily for publichealth reasons but also to provide a milk supplyof more uniform bacteriological quality and toimprove its keeping quality. Although a consid-erable amount of cheese is still produced fromraw milk, on both an artisanal and factory scale,especially in southern Europe (including suchfamous varieties as Swiss Emmental, Gruyere,Comte, Parmigiano-Reggiano, and GranoPadano), pasteurized milk is now generallyused, especially in large factories. Aspects ofpasteurization are discussed in Chapter 4.

There are four alternatives to pasteurizationfor reducing the number of microorganisms inmilk:

1. treatment with H2O2

2. activation of the lactoperoxidase-H2O2-thiocyanate system

3. bactofugation4. microfiltration

These processes are also discussed briefly inChapter 4.Gluconic acid 8-lactone Gluconic acid

Page 27: Cheese Science

2.4 CHEESE COLOR

Color is a very important attribute of foodsand serves as an index of quality, although insome cases, it is merely cosmetic. The principalindigenous pigments in milk are carotenoidswhich are obtained from the animal's diet, espe-cially from fresh grass and clover.

The carotenoids are secondary pigments in-volved in photosynthesis (the structure ofP-carotene is shown in Figure 2-2). Owing tothe conjugated double-bond system, they absorbultraviolet and visible light, giving them colorsranging from yellow to red. They are responsiblefor the color of many foods (e.g., carrots,squashes, peppers, and corn). They are alsopresent in the leaves of plants, in which theircolor is masked by the green chlorophylls. Somecarotenoids have pro-vitamin A activity andmay be converted to retinol (Figure 2-2) in thebody.

Animals do not synthesize carotenoids but ab-sorb them from plant materials in their diet. Inaddition to serving as pro-vitamin A, some ani-mals store carotenoids in their tissues, whichthen acquire a color (e.g., salmon, cooked lob-ster, and egg yolk). Cattle transfer carotenoids toadipose tissue and milk but goats, sheep, andbuffalo do not. Therefore, bovine milk and milkproducts are yellow to an extent dependent onthe carotenoid content of the animal's diet. Prod-

ucts such as butter and cheese made from sheep,goat, or buffalo milk are very white in compari-son with counterparts made from bovine milk.This yellowish color may make products pro-duced from cow milk less acceptable than prod-ucts produced from sheep, goat, or buffalo milkin Mediterranean countries, where the latter aretraditional.

The carotenoids in bovine milk can bebleached by treatment with H2O2 or benzoyl per-oxide or masked by chlorophyll or titanium ox-ide (TiO2), although such practices are not per-mitted in all countries.

At the other end of the spectrum are individu-als who prefer highly colored cheese, butter, oregg yolk. Intense colors may be obtained by add-ing carotenoids (synthetic or natural extracts). Inthe case of cheese and dairy products, annatto,extracted from the pericarp of the seeds of theannatto plant (Bixa orelland), a native of Brasil,is used most widely. Annatto contains two apo-carotenoid pigments, bixin and norbixin (Figure2-3). By suitable modification, the annatto pig-ments can be made fat soluble, for use in butteror margarine, or water soluble, for use in cheese.

Initially, annatto may have been used incheese manufacture to give the impression of ahigh fat content in partially skimmed cheese, butsome people believe that colored ("red") cheesetastes better than its white counterpart of equiva-lent quality.

Figure 2-2 Structures of P-carotene and retinol.

Retinol

p-Carotene

Cleavage at this point resultsin two molecules of Vitamin A

Page 28: Cheese Science

2.5 CONVERSION OF MILK TO CHEESECURD

After the milk has been standardized, pasteur-ized, or otherwise treated, it is transferred to vats(or kettles). These vats are of various shapes(hemispherical, rectangular, vertical cylindrical,and horizontal cylindrical), may be open orclosed, and may range in size from a few hun-dred liters to 30,000 liters. Here, it is convertedto cheese curd, a process that involves three ba-sic operations: acidification, coagulation, anddehydration.

2.5.1 Acidification

Acidification is usually achieved through thein situ production of lactic acid through the fer-mentation of the milk sugar lactose by lactic acidbacteria. Initially, the indigenous milk micro-flora was relied upon to produce acid, but sincethis microflora was variable, the rate and extentof acidification were variable, resulting incheese of variable quality. Cultures of lactic acidbacteria for cheesemaking were introduced com-mercially about 100 years ago, and these havebecome increasingly improved and refined. The

science and technology of starters is described inChapter 5. The acidification of curd for someartisanal cheeses still relies on the indigenousmicroflora.

Direct acidification using an acid (usually lac-tic or hydrochloric acid) or an acidogen (usuallygluconic acid-5-lactone) may be used as an alter-native to biological acidification. It is used com-mercially to a significant extent in the manufac-ture of Cottage cheese, Quarg, and Feta-typecheese from ultrafiltration-concentrated milkand Mozzarella.

Direct acidification is more controllable thanbiological acidification, and, unlike starters, it isnot susceptible to phage infection. However, inaddition to acidification, the starter bacteriaserve very important functions in cheese ripen-ing (see Chapters 10 and 11), and hence chemi-cal acidification is used mainly for cheese variet-ies for which texture is more important thanflavor.

The rate of acidification is fairly characteristicof the variety, and its duration ranges from 5-6hr for Cheddar and Cottage cheese to 10-12 hrfor Dutch and Swiss types. The rate of acidifica-tion, which depends on the amount of starterused and on the temperature profile of the curd,

Bixin

Norbixin

Figure 2-3 Structures of cis-bixin and norbixin, the apocarotenoid pigments in annatto.

Page 29: Cheese Science

has a major effect on the texture of cheese,mainly through its solubilizing effect on colloi-dal calcium phosphate (see Chapter 13).

Regardless of the rate of acidification, the ul-timate pH of the curd for most hard cheese vari-eties is in the range 5.0-5.3 but it is 4.6 for thesoft, acid-coagulated varieties (e.g., Cottagecheese, Quarg, and Cream cheese) and somerennet-coagulated varieties (e.g., Camembertand Brie).

The production of acid at the appropriate rateand time is critical for the manufacture of goodquality cheese. Acid production affects severalaspects of cheese manufacture, many of whichare discussed in more detail in later chapters:

• coagulant activity during coagulation• denaturation and retention of the coagulant

in the curd during manufacture and hencethe level of residual coagulant in the curd(this influences the rate of proteolysis dur-ing ripening and may affect cheese quality)

• curd strength, which influences cheeseyield

• gel syneresis, which controls cheese mois-ture and hence regulates the growth of bac-teria and the activity of enzymes in thecheese (consequently, it strongly influencesthe rate and pattern of ripening and thequality of the finished cheese)

• the extent of dissolution of colloidal cal-cium phosphate, which modifies the sus-ceptibility of the caseins to proteolysis dur-ing ripening and influences the rheologicalproperties of the cheese (e.g., compare thetexture of Emmental, Gouda, Cheddar, andCheshire cheese) (see Chapter 13)

• the growth of many nonstarter bacteria incheese, including pathogenic, food-poison-ing, and gas-producing microorganisms(properly made cheese is a very safe prod-uct from a public health standpoint) (seeChapter 20)

The level and time of salting have a major in-fluence on pH changes in cheese. The concentra-tion of NaCl in cheese (commonly 0.7-4%,which is equivalent to 2-10% salt in the mois-ture phase) is sufficient to halt the growth of

starter bacteria. Some varieties, mostly of Brit-ish origin, are salted by mixing dry salt with thecurd toward the end of manufacture, and hencethe pH of curd for these varieties must be closeto the ultimate value (~ pH 5.1) at salting. How-ever, most varieties are salted by immersion inbrine or by surface application of dry salt. Saltdiffusion in cheese moisture is a relatively slowprocess and thus there is ample time for the pHto decrease to about 5.0 before the salt concen-tration becomes inhibitory throughout the inte-rior of the cheese. The pH of the curd for mostcheese varieties (e.g., Swiss, Dutch, Tilsit, andBlue cheese) is 6.2-6.5 at molding and pressingbut decreases to around 5-5.2 during or shortlyafter pressing and before salting. The signifi-cance of various aspects of the concentration anddistribution of NaCl in cheese is discussed inChapter 8.

In a few special cases (e.g., Domiati), a highlevel of NaCl (10-12%) is added to the cheesemilk, traditionally to control the growth of theindigenous microflora. This concentration ofNaCl has a major influence not only on acid de-velopment but also on rennet coagulation, gelstrength, and curd syneresis.

2.5.2 Coagulation

The essential step in the manufacture of allcheese varieties involves coagulation of thecasein component of the milk protein system toform a gel that entraps the fat, if present. Coagu-lation may be achieved by

• limited proteolysis by selected proteinases(rennets)

• acidification to pH 4.6• acidification to a pH value greater than 4.6

(perhaps ~ 5.2) in combination with heatingto roughly 9O0C

The vast majority of cheese varieties (repre-senting about 75% of total production) are pro-duced by rennet coagulation, but some acid-co-agulated varieties, such as Quarg and Cottagecheese, are of major importance. The acid-heat-coagulated cheeses are of relatively minor im-portance. They are usually produced from whey

Page 30: Cheese Science

or a blend of whey and skim milk and probablyevolved as a useful means for recovering the nu-tritionally valuable whey proteins. Their proper-ties are very different from those of rennet- oracid-coagulated cheeses, and they are usuallyused as food ingredients. Important varieties areRicotta and related varieties (indigenous toItaly), Anari (Cyprus), and Manouri (Greece)(see Chapters 17 and 19).

The coagulation of milk by rennets and thecoagulation of milk by acid are discussed inChapters 6 and 16, respectively.

A fourth (minor) group of cheeses is pro-duced, not by coagulation, but by thermal evapo-ration of water from a mixture of whey and skimmilk, whole milk, or cream and crystallization oflactose. Varietal names include Mysost andGjetost. These cheeses, which are almost exclu-sive to Norway, bear little resemblance to ren-net- or acid-coagulated cheeses and probablyshould be classified as whey products rather thancheese.

2.5.3 Postcoagulation Operations

Rennet- or acid-coagulated milk gels are quitestable if maintained under quiescent conditions,but if cut or broken, they rapidly undergo syner-esis, expelling whey. Syneresis essentially con-centrates the fat and casein of milk by a factor of6-12, depending on the variety. In the dairy in-dustry, concentration is normally achievedthrough thermal evaporation of water and morerecently by removing water through semiperme-able membranes. The syneresis of rennet- oracid-coagulated milk gels is thus an unusualmethod of dehydration.

The rate and extent of syneresis are influ-enced, inter alia, by the milk composition, espe-cially the concentrations of Ca2+ and casein; thepH of the whey; the cooking temperature; therate of stirring of the curd-whey mixture; and,of course, time (see Chapter 7). The composi-tion of the finished cheese is, to a very large de-gree, determined by the extent of syneresis, andsince this is readily under the control of thecheesemaker, it is here that the differentiation of

the individual cheese varieties really begins, al-though the type and composition of the milk,the amount and type of starter, and the amountand type of rennet are also significant in this re-gard.

A more or less unique protocol has been de-veloped for the manufacture of each cheese vari-ety. Such protocols differ mainly with respect tothe syneresis process. The protocols for themanufacture of the principal families of cheeseare summarized in Chapter 17.

2.5.4 Salting

Salting is the last manufacturing operation.Salting promotes syneresis, but it is not a satis-factory method for controlling the moisture con-tent of cheese curd, which is best achieved byensuring that the degree of acidification, heat-ing, and stirring in the cheese vat are appropriateto the particular variety. Salt has several func-tions in cheese, and these are described in Chap-ter 8. Although salting should be a very simpleoperation, quite frequently it is not performedproperly, with adverse effects.

2.5.5 Ultrafiltration

As indicated previously, cheese manufactureis essentially a dehydration process. With thedevelopment of ultrafiltration as a concentrationprocess, it was obvious that this process wouldhave applications in cheese manufacture, notonly for standardizing cheese milk with respectto fat and casein but more importantly for thepreparation of a concentrate with the composi-tion of the finished cheese, commonly referredto aspre-cheese. Standardization of cheese milkby adding ultrafiltration concentrate (retentate)is now common in some countries, but themanufacture of pre-cheese has to date been suc-cessful commercially for only certain cheese va-rieties, most notably ultrafiltration Feta andQuarg. Undoubtedly, the use of ultrafiltrationwill become much more widespread in cheesemanufacture (see Chapter 17).

Page 31: Cheese Science

2.6 RIPENING

Fresh cheeses constitute a major proportion ofthe cheese consumed in some countries (seeTable 1-3). Most of these cheeses are producedby acid coagulation and are described in Chapter16. Although rennet-coagulated cheese varietiesmay be consumed at the end of manufacture anda few are (e.g., Burgos), most rennet-coagulatedcheeses are ripened (cured, matured) for a periodranging from about 3 weeks to more than 2years. Generally, the duration of ripening is in-versely related to the moisture content of thecheese. Many varieties may be consumed at anyof several stages of maturity, depending on theflavor preferences of consumers and economicfactors.

Although curds for different cheese varietiesare recognizably different at the end of manufac-ture (mainly as a result of compositional and tex-tural differences arising from differences in milkcomposition and processing factors), the uniquecharacteristics of the individual cheeses developduring ripening as a result of a complex set ofbiochemical reactions. The changes that occurduring ripening—and hence the flavor, aroma,and texture of the mature cheese—are largelypredetermined by the manufacturing process,that is, by the composition (especially moisture,NaCl, and pH), by the level of residual coagulantactivity, by the type of starter, and in many casesby secondary inocula added to or gaining accessto the milk or curd.

The biochemical changes that occur duringripening are caused by one or more of the fol-lowing agents:

• the coagulant• indigenous milk enzymes, especially pro-

teinase and lipase, which are particularlyimportant in cheese made from raw milk

• starter bacteria and their enzymes• secondary microorganisms and their en-

zymes

The secondary microflora may arise from in-digenous microorganisms that survive pasteur-ization or gain entry to the milk after pasteuriza-

tion (e.g., Lactobacillus, Pediococcus, and Mi-crococcus). They may also be added as a sec-ondary starter, such as citrate-positive Lacto-coccus or Leuconostoc spp. in Dutch-typecheese, Propionibacterium in Swiss cheese,Penicillium roqueforti in Blue varieties, P.camemberti in Camembert or Brie, or Brevi-bacterium linens in surface smear-ripened vari-eties (e.g., Tilsit and Limburger). In many cases,the characteristics of the finished cheese aredominated by the metabolic activity of these sec-ondary microorganisms.

The primary biochemical changes involveglycolysis, lipolysis, and proteolysis, but theseare followed and overlapped by a host of sec-ondary catabolic changes to the compounds pro-duced in these primary pathways, including de-amination, decarboxylation, and desulfurylationof amino acids, (3-oxidation of fatty acids, andeven some synthetic reactions (e.g., esterifica-tion).

Although it is not yet possible to fully de-scribe the biochemistry of cheese ripening, veryconsiderable progress has been made on eluci-dating the primary reactions, and these are dis-cussed in Chapter 11.

2.7 PROCESSED CHEESE PRODUCTS

Depending on culinary traditions, a variableproportion of mature cheese is consumed ascheese (such cheese is often referred to as tablecheese). A considerable amount of naturalcheese is used as an ingredient in other foods(e.g., Parmesan or Grana on pasta products,Mozzarella on pizza, Quarg in cheesecake,Ricotta in ravioli). A third major outlet forcheese is in the production of a broad range ofprocessed cheese products, which in turn have arange of applications, especially as spreads,sandwich fillers, or food ingredients. Theseproducts are discussed in Chapters 18 and 19.

2.8 WHEY AND WHEY PRODUCTS

Only about 50% of the solids in milk are incor-porated into cheese; the remainder (90% of the

Page 32: Cheese Science

1 actose,« 20% of the protein, and «10% of the fat)are present in the whey. Until recently, whey wasregarded as an essentially useless byproduct to bedisposed of as cheaply as possible. However, inthe interest of reducing environmental pollution,

but also because it is now possible to producevaluable food products from whey, whey pro-cessing has become a major facet of the totalcheese industry. The principal aspects of wheyprocessing are discussed in Chapter 22.

Page 33: Cheese Science

3.1 INTRODUCTION

Milk is a fluid secreted by the female of allmammals, of which there are more than 4,000species, for the primary function of meeting thecomplete nutritional requirements of the neonateof the species. It must supply energy (mainlyfrom fats and sugar [lactose]), amino acids (fromproteins), vitamins, and atomic elements (com-monly but inaccurately referred to as minerals).In addition, several physiological functions areperformed by milk constituents, including anti-microbial substances (immunoglobulins, lacto-peroxidase, and lactotransferrin), enzymes andenzyme inhibitors, vitamin-binding carrier pro-teins, and cell growth and control factors. Be-cause the nutritional and physiological require-ments of each species are more or less unique,the composition of milk shows very markedinterspecies differences. The milks of only about180 species have been analyzed, and the data foronly about 50 of these species are considered tobe reliable (sufficient number of samples, repre-sentative sampling, and adequate coverage ofthe lactation period). Not surprisingly, the milksof the principal dairying species (i.e., cow, goat,sheep, and buffalo) and Homo sapiens areamong those that are well characterized. Thegross composition of milks from selected spe-cies is summarized in Table 3-1. Very extensivedata on the composition of bovine and humanmilks are compiled by Jensen (1995).

In addition to the principal constituents listedin Table 3-1, milk contains several hundred mi-nor constituents, many of which have a majorimpact on the nutritional, technological, andsensoric properties of milk and dairy products(e.g., vitamins, small inorganic and organic ions,and flavor compounds).

Milk is a very variable biological fluid. In ad-dition to interspecies differences, the milk of anyparticular species varies with the breed (in thecase of commercial dairying species), health,nutritional status, stage of lactation, and age ofthe animal, the interval between milkings, and soon. In a bulked factory milk supply, variabilitydue to many of these factors is reduced, butsome variability persists, and it can even be quitelarge in situations where milk production is sea-sonal. In addition to variations in the concentra-tions of the principal and minor constituents dueto the above factors, the chemistry of some ofthe constituents also varies (e.g., the fatty acidprofile is strongly influenced by diet). Some ofthe variability in the composition and constitu-ents of milk can be adjusted or counteracted byprocessing technology but some differencesmay persist. As will become apparent in laterchapters, the variability of milk compositionposes major problems in cheese production.

Physicochemically, milk is a very complexfluid. The constituents of milk occur in threephases. Most of the mass of milk is an aqueoussolution of lactose, organic and inorganic salts,

Chemistry of Milk Constituents

CHAPTER 3

Page 34: Cheese Science

vitamins, and other small molecules. In thisaqueous solution are dispersed proteins, some atthe molecular level (whey proteins), others aslarge colloidal aggregates ranging in diameterfrom 50 to 600 nm (the caseins), and lipids,which exist in an emulsified state as globulesranging in diameter from 0.1 to 20 jam. Thus,colloidal chemistry is important in the study ofmilk, for example, in the context of surfacechemistry, light scattering, and rheology.

Milk is a dynamic system owing to the insta-bility of many of its structures (e.g., the milk fatglobule membrane); changes in the solubility ofmany constituents, especially the inorganic saltsand proteins, with temperature and pH; the pres-ence of various enzymes that can modify con-stituents through lipolysis, proteolysis, or oxida-tion-reduction; the growth of microorganisms,which can cause major changes either directlythrough their growth (e.g., changes in pH or re-dox potential [Eh]), or indirectly through en-zymes they excrete; and the interchange of gaseswith the atmosphere (e.g., CO2). Milk was in-tended to be consumed directly from the mam-mary gland and to be expressed from the gland atfrequent intervals. However, in dairy operations,milk is stored for various periods, ranging from afew hours to several days, during which it iscooled (and perhaps heated) and agitated to vari-

ous degrees. These treatments will cause somephysical changes and permit some enzymaticand microbiological changes that may alter theprocessing properties of milk. It may be possibleto counteract some of these changes.

Although many of the minor constituents ofmilk are important from a nutritional standpoint,the technological properties of milk are deter-mined mainly by its macroconstituents (pro-teins, lipids, and lactose) and some of its lowmolecular mass species, especially calcium,phosphate, citrate, and pH. The properties ofthese constituents, with emphasis on their sig-nificance in cheesemaking, are discussed brieflyin this chapter. For a more thorough discussion,the reader is referred to Cayot and Lorient(1998); Fox (1982, 1983, 1985, 1989, 1992,1995, 1997); Fox and McSweeney (1998);Jenness and Patton (1959); Walstra and Jenness(1984); Webb and Johnson (1965); Webb,Johnson, and Alford (1974); and Wong, Jenness,Keeney, and Marth (1988).

3.2 LACTOSE

Lactose is the principal carbohydrate in themilk of all mammals, which is the only source.Milk contains only trace amounts of other sug-ars, including glucose, fructose, glucosamine,

Species

HumanCowGoatSheepPigHorseDonkeyReindeerDomestic rabbitBisonIndian elephantPolar bearGrey seal

Total Solids (%)

12.212.712.319.318.811.211.733.132.814.631.947.667.7

Fat (%)

3.83.74.57.46.81.91.4

16.918.33.5

11.633.153.1

Protein (%)

1.03.42.94.54.82.52.0

11.511.94.54.9

10.911.2

Lactose (%)

7.04.84.14.85.56.27.42.82.15.14.70.30.7

Ash (%)

0.20.70.81.0

0.50.5

1.80.80.71.4

Table 3-1 Composition of the Milks of Some Species

Page 35: Cheese Science

galactosamine, neuraminic acid, and neutral andacidic oligosaccharides.

The concentration of lactose in milk varieswidely between species (Table 3-2). The lactosecontent of cow milk varies with the breed ofcow, individuality factors, udder infection, andespecially stage of lactation. The concentrationof lactose decreases progressively and signifi-cantly during lactation (Figure 3-1); this trendcontrasts with the lactational trends for lipidsand proteins, which, after decreasing duringearly lactation, increase strongly during the sec-ond half of lactation. Lactose and soluble ions(e.g., Na+, K+, and Cl~) are the compounds

Table 3-2 Concentration of Lactose in the Milksof Selected Species

Species Lactose (%)

California sea lion 0.0Hooded seal 0.0Black bear 0.4Dolphin 0.6Echidna 0.9Blue whale 1.3Rabbit 2.1Red deer 2.6Grey seal 2.6Rat (Norwegian) 2.6Mouse (house) 3.0Guinea pig 3.0Dog (domestic) 3.1Sika deer 3.4Goat 4.1Elephant (Indian) 4.7Cow 4.8Sheep 4.8Water buffalo 4.8Cat (domestic) 4.8Pig 5.5Horse 6.2Chimpanzee 7.0Rhesus monkey 7.0Human 7.0Donkey 7.4Zebra 7.4Green monkey 10.2

mainly responsible for the osmotic pressure ofmilk. During mastitis, the concentration of NaClin milk increases, resulting in an increase in os-motic pressure. This increase is compensated forby a decrease in the lactose content; that is, thereis an inverse relationship between the concentra-tion of NaCl and lactose in milk, which partlyexplains why certain milks with a high lactosecontent have a low ash content and vice versa(see Table 3-1). The inverse relationship be-tween the concentration of lactose and chlorideis the basis of the Koestler's chloride-lactose testfor abnormal milk:

percentage of chloride x 100Koestler number = percentage of lactose

A Koestler number less than 2 indicates normalmilk while a value greater than 3 is consideredabnormal.

Lactose plays an important role in milk andmilk products:

• It is essential in the production of fermenteddairy products, including cheese.

• It contributes to the nutritive value of milkand its products. However, many non-Europeans have limited or zero ability todigest lactose in adulthood, leading to asyndrome known as lactose intolerance.Mature cheese is free of lactose, and hencecheese is suitable for inclusion in the diet oflactose-intolerant individuals.

• It affects the texture of certain concentratedand frozen products.

• It is involved in heat-induced changes in thecolor and flavor of highly heated milk prod-ucts.

3.2.1 Structure of Lactose

Lactose is a disaccharide consisting of galac-tose and glucose, linked by a P1-4 glycosidicbond (Figure 3-2). Its systematic name is O- (3-D-galactopyranosyl-(l-4)-a-D-glucopyranose(a-lactose) or 0-(3-D-galactopyranosyl-(l^)-p-D-glucopyranose ((i-lactose). The hemiacetalgroup of the glucose moiety is potentially free

Page 36: Cheese Science

(i.e., lactose is a reducing sugar) and may existas an a- or p-anomer. In the structural formulaof the a-form, the hydroxyl group on the Ci ofglucose is cis to the hydroxyl group at C2 (ori-ented downward) and vice versa for the P-form(oriented upward).

3.2.2 Biosynthesis of Lactose

Lactose is essentially unique to mammary se-cretions. It is synthesized from glucose absorbedfrom blood. One molecule of glucose is con-verted to UDP-galactose via the 4-enzymeLeloir pathway (Figure 3-3). UDP-galactose isthen linked to another molecule of glucose in areaction catalyzed by the enzyme lactose syn-thetase, a 2-component enzyme. Component Ais a nonspecific galactosyl transferase that trans-fers the galactose from UDP-galactose to a num-ber of acceptors. In the presence of the B compo-nent, which is the whey protein oc-lactalbumin,the transferase becomes highly specific for glu-cose (its KM decreases 1,000-fold), leading to thesynthesis of lactose. Thus, oc-lactalbumin is an

enzyme modifier, and its concentration in themilk of several species is directly related to theconcentration of lactose in those milks; the milksof some marine mammals contain neither oc-lac-talbumin nor lactose.

The presumed significance of this controlmechanism is to enable mammals to terminatethe synthesis of lactose when necessary, that is,to regulate and control osmotic pressure whenthere is an influx of NaCl, such as during masti-tis or in late lactation (milk is isotonic withblood, the osmotic pressure of which is essen-tially constant). The ability to control osmoticpressure is sufficiently important to justify anelaborate control mechanism and "wastage" ofthe enzyme modifier.

3.2.3 Lactose Equilibrium in Solution

The configuration around the anomeric Ci ofthe glucose moiety of lactose is not stable and canreadily change (mutarotate) from the a- to thep-form and vice versa when the sugar is in solu-tion as a consequence of the fact that the hemiac-

Week of lactation

Figure 3-1 Changes in the concentrations of fat (•), protein (•), and lactose (O) in milk during lactation.

Per

cent

Page 37: Cheese Science

Galactose Glucose(a)

(b)

Anomeric carbon

(c)

(d)

Figure 3-2 Structural formulae of a- and (3-lactose: (a) open chains, (b) Fischer projection, (c) Haworth projec-tion, and (d) conformational formula.

Page 38: Cheese Science

etal form is in equilibrium with the open-chainaldehyde form, which can be converted to eitherof the two isomeric forms (see Figure3-2). When either isomer is dissolved in water,there is a gradual change from one form to theother until equilibrium is established (i.e., mu-tarotation). These changes are reflected bychanges in optical rotation from +89.4° for a-lactose or +35° for p-lactose to a value of+55.4°at equilibrium. These values for specific rotationindicate that at equilibrium, a solution of lactoseconsists of 62.7% p anomer and 37.3% a anomer.

The a and p anomers of lactose differ mark-edly with respect to solubility, crystal shape, hy-dration of the crystals, hygroscopicity, specificrotation, and sweetness.

oc-Lactose is soluble to around 7 g/100 mlH2O at 2O0C, and the solubility of p-lactose isaround 50 g/100 ml. However, the solubility ofoc-lactose is more temperature dependent thanthat of (J-lactose, and their solubility curves in-tersect at about 940C (Figure 3-4). Thus, oc-lac-tose is the form normally produced by crystalli-zation. oc-Lactose crystallizes as a monohydrate,whereas crystals of P-lactose are anhydrous. Al-though lactose has low solubility in comparisonwith other sugars, once in solution it crystallizes

slowly, and precautions must be taken in themanufacture of concentrated and dehydratedproducts; otherwise, hygroscopicity, caking, anda grainy texture (due to the slow growth of lac-tose crystals to a size greater than 15 [\m) willensue. These physicochemical properties of lac-tose are of major concern to manufacturers ofconcentrated, dehydrated, and frozen dairyproducts, but problems can be avoided by propermanufacturing procedures. Such properties areof no consequence in cheese, in which all thelactose is utilized either during manufacture orearly ripening; fresh curd contains about 1% lac-tose. The behavior of lactose is of major concernin the manufacture of whey powders, sincearound 70% of the total solids in whey are lac-tose, and hence the properties of whey concen-trates and powders are strongly influenced bythe properties of lactose.

In cheese, lactose is fermented to lactic acidby lactic acid bacteria, a process which has ma-jor, indeed vital, consequences for the manufac-ture and quality of cheese, as is discussed inChapters 5, 10, and 11.

For further information on the properties ofand the problems caused by lactose, the readeris referred to Fox (1985, 1997), Fox and

Figure 3-3 Pathway for lactose synthesis.

Glucose

LACTOSEa-lactalbwnin

galactosyltransferase

ATPUDP

ADPUTP

Glucose-1 -phosphate

Glucose-6-phosphate

Phosphoglucomutase

hexokinase

ADPATP

P-P

UDP glucose

GLUCOSE

UDP-glucosepyrophosphorylase

UDP glucose-4-epimerase

UDP-galactose

Page 39: Cheese Science

McSweeney (1998), Walstra and Jenness(1984), and Wong et al. (1988).

3.3 MILK LIPIDS

The lipid content of milk varies more widelythan any other constituent; concentrations rangefrom around 2% to more than 50% (Table 3-3).The average fat content of cow, goat, sheep, andbuffalo milk is 3.5, 3.5, 6.5, and 7 g/L, respec-tively. Within any particular species, there areconsiderable variations due to breed, individual-ity, stage of lactation, age, animal health, nutri-tional status, interval between milking, and soon. Among the common breeds of dairy cattle,Jersey cows produce milk with the highest fatcontent (6-7%) and Holstein/Friesian, the low-est. Within any breed, there are considerable in-dividual cow variations. The fat content of milkdecreases for several weeks after parturition andthen increases, especially toward the end of lac-tation (see Figure 3-1). If the interval betweenmilkings is not equal, the milk obtained after the

shorter interval has the higher fat content. Thesynthesis of all milk constituents, including fat,decreases during a mastitic infection, and the fatcontent of milk decreases slightly as the animalages.

The lipids in milk are predominantly triglyc-erides (triacylglycerols), which make up about98% of the total lipid fraction; the remaining 2%comprises diglycerides, monoglycerides, fattyacids, phospholipids, sterols (principally choles-terol), and trace amounts of fat-soluble vitamins(A, D, E, and K). Typical values for the concen-tration of the various lipids in milk are given inTable 3-4.

3.3.1 Fatty Acid Composition

Ruminant milk fats contain a greater diversityof fatty acids than other fats; about 400 fatty ac-ids have been identified in bovine milk fat. Thepredominant fatty acids have a straight carbonchain with an even number of carbon atoms andmay be saturated or unsaturated (1,2, or 3 C = C

Figure 3-4 Solubility of lactose in water.

Temperature, 0C

Initial solubility of (3-lactose

Initial solubility of a-lactose

Final solubility at equilibrium

Usual range ofsupersaturation

Solu

bilit

y, g

anh

ydro

us la

ctos

e /1

00 g

wat

er

Page 40: Cheese Science

Table 3-3 Fat Content of Milks of Various Spe-cies

Species Fat Content (g/L)

Cow 33-47Buffalo 47Sheep 40-99Goat 41-45Musk ox 109DaII sheep 32-206Moose 39-105Antelope 93Elephant 85-190Human 38Horse 19Monkeys 10-51Lemurs 8-33Pig 68Marmoset 77Rabbit 183Guinea pig 39Snowshoe hare 71Muskrat 110Mink 134Chinchilla 117Rat 103Red kangaroo 9-119Dolphin 62-330Manatee 55-215Pygmy sperm whale 153Harp seal 502-532Bear (four species) 108-331

double bonds). There are smaller amounts offatty acids with an uneven number of carbon at-oms, branched or cyclic hydrocarbon chains, orhydroxyl or keto groups. The principal fatty ac-ids in the milk fat of a selection of species arelisted in Table 3-5.

The fatty acid profile of ruminant milk fatshas a number of interesting features:

• Ruminant milk fats contain a considerableamount of butanoic acid (C4:0) and are infact the only fats that contain this acid. Thehigh content of butanoic acid is due to thesynthesis of 3-hydroxybutanoic acid ((3-hy-

droxybutyric acid) and its reduction to bu-tanoic acid by bacteria in the rumen. Thehigh concentration of butanoic acid in rumi-nant milk fats provides the basis for themethod commonly used to detect and quan-tify the adulteration of milk fat with otherfats, that is, the Reichert-Meissl number (mlof 0.1 IyI KOH required to neutralize thevolatile water-soluble fatty acids releasedfrom 5 g fat upon hydrolysis).

• Ruminant milk fats, in general, and ovinemilk fat in particular, contain relativelyhigh concentrations of middle chain fattyacids (hexanoic [C6:0] to decanoic [Ci0:o]).This is due to high thioacylhydrolase activ-ity in the fatty acid synthetase complex,which causes the early release of fatty acidsduring the chain elongation process.

• The short and middle chain acids (C410-Cio:o) are relatively volatile and watersoluble and have a relatively low flavorthreshold. They are esterified predomi-nantly at the sn3 position of glycerol andhence are selectively released by lipases,especially by the indigenous lipoprotein Ii-pase in milk. In milk and butter, the releaseof these highly flavored short chain fattyacids gives rise to off-flavors, referred to ashydrolytic rancidity. However, whenpresent at an appropriate level, these shortchain acids contribute positively to the fla-vor of cheese, especially hard Italian andblue-mold varieties.

• Ruminant milk fats contain low levels ofpolyunsaturated fatty acids (PUFAs; Ci8:2,Ci8:3), which are considered to be nutrition-ally desirable. However, the low level ofPUFAs makes milk fat relatively resistantto oxidative rancidity. The low concentra-tion of PUFAs in ruminant milk fats is dueto the hydrogenation of dietary fatty acidsby bacteria in the rumen, although ruminantfeed usually contains high levels of PUFAs.On the positive side, biohydrogenation ofPUFAs results in lower levels of trans iso-mers than chemical hydrogenation, such asis practiced in the processing of vegetable

Page 41: Cheese Science

oils. Trans fatty acids are considered to benutritionally undesirable.

The concentration of PUFAs in ruminant milkfats can be increased by including protected lip-ids in the animal's diet. This involves encapsu-lating the dietary lipids in a layer of polymerizedprotein or using crushed vegetable seeds. Encap-sulation protects the PUFAs against hydrogena-tion in the rumen, but the capsule is digested inthe abomasum, liberating the encapsulated lip-ids, which are then metabolized, as in non-ruminants. Fat has a major effect on the rheo-logical properties of cheese. Polyunsaturatedlipids, which have a low melting point, have anundesirable effect on cheese texture, but a lowlevel is acceptable.

Although phospholipids are present at verylow concentrations in milk, they play an impor-tant role in the emulsification of fat in milk. Milkcontains a relatively low concentration of cho-lesterol, a high level of which in the diet is con-sidered to be nutritionally undesirable. The cowtransfers dietary carotenoids to its milk, andhence its milk fat has a yellow color, the inten-sity of which depends on the concentration ofcarotenoids in the animal's feed—fresh grassand especially clover and lucerne are rich incarotenoids (see Fox and McSweeney, 1998, forthe structures of the principal phospholipids,cholesterol, and fat-soluble vitamins). Sheepand goats do not transfer dietary carotenoids totheir milk, and consequently their milk fat and

fat-containing products (including cheese) madefrom ovine or caprine milk are much whiter thantheir bovine counterparts. Products traditionallymade from ovine or caprine milk may be unac-ceptable when made from bovine milk, owing totheir yellow color, especially if the cows are fedon fresh grass. However, it is possible to bleachor mask the color of carotenoids (e.g., usingH2O2, benzoyl peroxide, TiO2, or chlorophyll).

Some carotenoids are precursors of vitamin A(retinol). Milk contains low levels of vitamin D,and liquid milk products are commonly fortifiedwith vitamin D. Milk contains a substantialamount of vitamin E (tocopherols), which is apotent antioxidant. The tocopherol content ofmilk may be increased by supplementing theanimal's diet with tocopherols; this may be donefor nutritional or stability reasons. However, lipidoxidation is not a problem in cheese, probablybecause of its low redox potential (Eh: -150 mV).

3.3.2 Milk Fat as an Emulsion

Lipids are insoluble in and less dense thanwater (the specific gravity of fat and skim milk isaround 0.9 and 1.036, respectively), and hencethey would be expected to form a layer on thesurface of milk. Lipids in general can be madecompatible with water by forming an emulsionthrough homogenization, in which the fat is dis-persed as small globules, each of which is sur-rounded by a layer of emulsifier. An emulsion isdefined as a two-phase system, one phase (the

Table 3-4 Composition of Individual Simple Lipids and Total Phospholipids in Milks of Various Spe-cies (Percentage of Total Lipids by Weight)

LIpId Class

TriacylglycerolsDiacylglycerolsMonoacylglycerolsCholesteryl estersCholesterolFree fatty acidsPhospholipids

* T = trace

Cow

97.50.360.027T0.310.0270.6

Buffalo

98.6

0.10.30.50.5

Human

98.20.7T*T0.250.40.26

Pig

96.80.70.10.060.60.21.6

Rat

87.52.90.4

1.63.10.7

Mink

81.31.7TTT1.3

15.3

Page 42: Cheese Science

Table 3-5 Principal Fatty Acids in Milk Triacylglycerols or Total Lipids of Various Species (Percentage of Total by Weight)

C/20— U£218:318:218:118:016:116:014:012:010:08:06:04:0Species

T*T

0.42.6

T

1.1T7.0T0.4T

0.112.20.2

17.30.44.5

31.229.311.39.5

0.82.51.4

3.04.13.7

0.71.4

1.30.60.5

12.60.70.85.70.94.49.81.71.52.92.17.60.10.22.20.60.9

0.42.3

2.42.52.12.22.74.0

20.23.33.0

13.0

14.537.66.6

14.911.916.318.410.914.024.710.614.926.810.45.47.91.21.80.61.21.91.25.6

29.828.720.028.527.223.121.219.217.346.4

26.022.725.720.935.226.733.629.613.612.714.436.135.237.222.757.723.147.046.621.541.630.130.2

14.610.19.0

12.523.015.54.55.50.55.9

4.94.21.02.33.56.52.93.43.83.02.9

10.9

6.33.97.03.30.57.44.93.6

13.920.4

2.33.43.42.21.72.44.35.73.05.7

6.71.29.67.8

11.31.92.45.52.01.05.05.2

6.813.93.4

13.311.69.1

17.45.7

16.83.2

26.330.425.424.619.523.028.439.312.620.2

21.416.527.123.832.922.631.318.114.218.724.826.130.031.219.814.121.120.227.613.614.218.516.4

9.58.7

11.810.36.2

10.62.0

11.55.35.1

2.81.3

15.05.74.08.22.67.71.72.05.33.33.02.71.60.93.26.33.65.32.63.92.7

3.12.05.43.32.31.80.63.5

18.33.1

4.42.3

10.56.20.57.5

8.52.93.85.50.5T0.1

0.10.34.0

0.50.1

3.01.99.08.44.74.95.56.5

29.41.3

11.07.91.95.10.77.0

8.020.114.317.7

3.5

T

1.31.12.72.71.90.28.42.70.3T

5.95.10.21.8

1.1

22.49.6

10.9

0.6

1.61.62.82.90.90.3T6.0

T

0.60.4

T

T

T

T

TT

3.33.64.02.6T0.60.46.77.4

0.4

CowBuffaloSheepGoatMusk oxDaII sheepMooseBlackbuck antelopeElephantHumanMonkey (mean

of six species)BaboonLemur macacoHorsePigRatGuinea pigMarmosetRabbitCottontail rabbitEuropean hareMinkChinchillaRed kangarooPlatypusNumbatBottle-nosed dolphinManateePygmy sperm whaleHarp sealNorthern elephant sealPolar bearGrizzly bear

* T = trace.

Page 43: Cheese Science

discontinuous, dispersed, phase) being dispersedin the other (the continuous phase) and separatedby a layer of emulsifier. In milk and cream, fat isthe emulsified phase and water (or, more cor-rectly, skim milk) is the continuous phase (i.e.,milk is an oil-in-water emulsion). In butter (andmargarine), the situation is reversed: water drop-lets are dispersed in a continuous oil/fat phase(i.e., butter is a water-in-oil emulsion).

Emulsifiers are amphipathic molecules, withhydrophobic (lipophilic, fat-loving) and hydro-philic (water-loving) domains. The principalnatural emulsifiers are polar lipids and proteins;in addition, numerous synthetic emulsifiers areavailable and are used widely in the manufactureof high-fat foods.

In milk, the fat exists as globules, 0.1-20 jamin diameter (mean diameter, 3-4 urn). Numeri-cally, most of the globules have a diameter lessthan 1 jam, but these small globules representonly a small fraction of the mass of milk fat. Theglobules are surrounded by a structured mem-brane, referred to as the milk fat globule mem-brane (MFGM), consisting mainly of phospho-lipids and proteins; the approximate compositionof the MFGM is summarized in Table 3-6. Theinner layers of the membrane are acquired withinthe secretory cell (mammocyte) as the fat glob-ules move from the site of biosynthesis (i.e., therough endoplasmic reticulum located toward thebase of the cell) toward the apical membrane,through which they are expressed into the lumenof the mammary alveoli by exocytosis. Duringexocytosis, the fat globules become surrounded

by the apical cell membrane, which thereforeforms the outer layer of the MFGM of freshlysecreted milk fat globules. However, much ofthis membrane, which has a typical trilaminarfluid mosaic structure, is lost as the milk ages,and much of it accumulates as lipoprotein par-ticles, sometimes referred to as microsomes, inthe skim milk phase.

Many of the indigenous enzymes in milk areconstituents of the MFGM; consequently, iso-lated membrane (prepared by de-emulsification[churning] and washing) serves as the sourcematerial for the isolation of many indigenous milkenzymes. Xanthine oxidase is one of the principalproteins of the MFGM. Two notable exceptionsare the principal indigenous proteinase plasminand lipoprotein lipase (LPL), which are associ-ated mainly with the casein micelles. The MFGMisolates and protects the triglycerides from LPLbut if the membrane is damaged (e.g., by agita-tion), the enzyme and its substrate come into con-tact, and lipolysis and hydrolytic rancidity ensue,with undesirable consequences for the organolep-tic quality of milk and many dairy products.

Although the milk fat emulsion is stable tophase separation, it does exhibit rapid creamingowing to the difference in density between thephases (i.e., the fat globules rise to the surfacebut remain discrete and can be redispersed bygentle agitation). The rate of creaming is gov-erned by Stokes' law:

Table 3-6 Gross Composition of the Milk Fat Globule Membrane

Component

ProteinPhospholipidCerebrosidesCholesterolNeutral glyceridesWaterTotal

Amount inFat Globule(mgWOg-1)

900600

8040

300280

2,200

Amount inFat Globule

Surface (mg rrr2)

4.53.00.40.21.51.4

11.0

Percentage of TotalMembrane by Weight

412732

1413

100

Page 44: Cheese Science

where v is the velocity of particle movement; r isthe radius of the globules; p1 and p2 are the den-sities of the continuous and dispersed phases, re-spectively; g is the acceleration due to gravity(9.8 m s~2); and T] is the viscosity coefficient ofthe emulsion.

For milk, the parameters of Stokes' law wouldsuggest that a cream layer would form afterabout 60 hr, but in fact it forms in about 30 min.This large discrepancy between the actual andthe predicted rate of creaming is due to floccula-tion of the fat globules: the large globules risefaster than and collide with smaller globules,and the globules form clusters owing to the ag-glutinating action of immunoglobulin M. Thisprotein is referred to as a cryoglobulin, since itadsorbs onto the fat globules as the temperatureis reduced. The cluster then rises as a unit, col-liding with other globules as it does so and there-fore rising at an accelerating rate. The cryo-globulins solubilize as the temperature isincreased and are fully soluble above 370C.Consequently, creaming is promoted by lowtemperatures and is very slow above 370C.Cryoglobulins are denatured and inactivated byheating at time-temperature treatments greaterthan 740C x 15 s; hence, severely pasteurizedmilk creams poorly or not at all. Sheep, goat, andbuffalo milks are devoid of cryoglobulins andhence cream very slowly.

If the MFGM is physically damaged by hightemperatures and/or agitation, the globules coa-lesce, and eventually phase inversion will occur(i.e., an oil-in-water emulsion is converted to awater-in-oil emulsion). Free (nonglobular) fatwill float on the surface. Such damage occurs toat least some extent during cheesemaking; thefree fat is not incorporated into the coagulumand floats as quite large masses on the surface ofthe whey and is lost to the cheese. About 10% ofthe fat in milk is normally lost in this way. It canbe recovered from the whey by centrifugationand made into whey butter or other products.

Milk for many dairy products is "homog-enized," usually by using a valve homogenizer.Homogenization reduces the size of the fat glob-ules (average diameter, less than 1 |um) and dena-tures the cryoglobulins, and homogenized milk

does not cream owing to the combined effects ofglobule size reduction and denaturation of cryo-globulins. The membrane on the fat globules inhomogenized milk is mainly casein and does notprotect the triglycerides against lipolysis. Ho-mogenized milk must therefore be pasteurizedbefore or immediately after homogenization toprevent the occurrence of hydrolytic rancidity.

Milk for cheesemaking is not normally ho-mogenized, because homogenized milk forms arennet coagulum (gel) with a lower tendency toundergo syneresis upon cutting or stirring thanthat from nonhomogenized milk. Homogeniza-tion results in cheese with a higher moisture con-tent. This situation arises because the casein-coated fat globules behave somewhat like caseinmicelles, but they limit the contraction of thecasein matrix. It may be advantageous to ho-mogenize milk for low-fat cheese so as to obtaina higher moisture content and thus soften thetexture of the cheese. In some cases, milk forBlue cheese is separated and the cream homog-enized to promote lipolysis (which is desirablein Blue cheese). The lipolysed cream and skimmilk are then combined and pasteurized beforecheese manufacture. Milk for yogurt and creamcheese is also homogenized to

• prevent creaming during the relatively longgelation period

• increase the effective protein concentrationby converting the fat globules to pseudo-protein particles, thereby giving a firmergel for a given level of protein

• minimize syneresis

Fat plays an essential role in cheese quality:

• It acts as a plasticizer and affects cheesetexture (low-fat cheese has a hard, crumblytexture).

• It serves as a source of fatty acids, whichhave a direct effect on cheese flavor and arechanged to other flavor compounds (e.g.,carbonyls, lactones, esters, and thioesters).

• It acts as a solvent for flavor compoundsproduced from lipids, proteins, or lactose.

With the objective of reducing the calorificcontent of cheese, there is considerable commer-

Page 45: Cheese Science

Table 3-7 Protein Content in the Milk of VariousSpecies

Casein Whey TotalSpecies (%) Protein (%) (%)

Bison 3.7 0.8 4.5Black bear 8.8 5.7 14.5Camel (bactrian) 2.9 1.0 3.9Cat - - 11.1Cow 2.8 0.6 3.4Domestic rabbit 9.3 4.6 13.9Donkey 1.0 1.0 2.0Echidna 7.3 5.2 12.5Goat 2.5 0.4 2.9Grey seal - - 11.2Guinea pig 6.6 1.5 8.1Hare - - 19.5Horse 1.3 1.2 2.5House mouse 7.0 2.0 9.0Human 0.4 0.6 1.0Indian elephant 1.9 3.0 4.9Pig 2.8 2.0 4.8Polar bear 7.1 3.8 10.9Red kangaroo 2.3 2.3 4.6Reindeer 8.6 1.5 10.1Rhesus monkey 1.1 0.5 1.6Sheep 4.6 0.9 5.5White-tailed

jack rabbit 19.7 4.0 23.7

cial interest in the production of low-fat cheeses,but the quality of such cheeses is reduced, andconsequently they have had only limited market-ability.

3.4 MILK PROTEINS

From a cheesemaking standpoint, the proteinsof milk are its most important constituents. Theprotein content of milk shows large interspeciesdifferences, ranging from about 1% for humanmilk to more than 20% for the milk of smallmammals such as mice and rats (Table 3-7).There is a good correlation between the proteincontent of milk and the growth rate of the neo-nate of that species (Figure 3-5).

The proteins of milk belong to two main cat-egories that can be separated based on their solu-bility at pH 4.6 at 2O0C. Under these conditions,

one of the groups precipitates; these are knownas caseins. The proteins that remain soluble un-der these conditions are known as serum orwhey proteins. Approximately 80% of the totalnitrogen in bovine, ovine, caprine, and buffalomilks is casein, but casein constitutes only about40% of the protein in human milk. Both caseinsand whey proteins are heterogeneous and havevery different molecular and physicochemicalproperties.

3.4.1 Caseins

Bovine casein consists of four types of proteinwith substantially different properties: asl-,ocS2-> P~» and K~; these make up approximately38%, 10%, 34%, and 15%, respectively, of wholecasein. The caseins are well characterized at themolecular level (some of the major properties aresummarized in Table 3-8), and the amino acidsequences are known (Figures 3-6 to 3-9). Someof the more important properties of the caseins areas follows:

• They are quite small molecules, with mo-lecular masses of 20-25 kDa.

• All are phosphorylated. Most molecules ofocsi-casein contain 8 mol PO4/mol of pro-tein, but some contain 9 mol PO4/mol.(3-Casein molecules usually contain 5 molPO4/mol, but some contain 4 mol PO4/mol.ocs2-Casein molecules contain 10, 11, 12, or13 mol PO4/mol. Most molecules ofK-casein contain 1 mol PO4/mol, but somecontain 2 or perhaps 3 mol PO4AnOl.

• The phosphate groups are esterified as mo-noesters of serine and most occur as clus-ters. The phosphate groups bind polyvalentcations strongly, causing charge neutraliza-tion and precipitation of ocsr, ocs2-, and (3-caseins at greater than 6 mM Ca2+ at 3O0C.K-Casein, which usually contains only 1mol PO4/mol, binds cations weakly and isnot precipitated by them. It can stabilize upto 10 times its weight of calcium-sensitivecaseins via the formation of micelles (seeSection 3.4.2). In milk, the principal cationbound is calcium.

Page 46: Cheese Science

• Only ocs2- and K-caseins contain cysteine,which normally exists as intermoleculardisulphide bonds. as2-Casein usually occursas disulphide-linked dimers, but up to atleast 10 K-casein molecules may be di-sulphide linked. The absence of cysteine orcystine in asr and p-caseins increases theflexibility of these molecules.

• All the caseins, especially P-casein, containrelatively high levels of proline. Inp-casein, 35 of the 209 residues are proline,and these are uniformly distributed through-out the molecule. The presence of a highlevel of proline prevents the formation ofsecondary structures (ex-helices, p-sheets).

• Experimental techniques indicate that thecaseins have low levels of secondary andtertiary structures, although theoretical cal-culations indicate that they do have somedegree of higher structure. It has been sug-gested that, rather than lacking secondary

structures, the caseins have very flexiblestructures, which have been described asrheomorphic. The lack of stable secondaryand tertiary structures renders the caseinsstable to denaturing agents (e.g., heat orurea); contributes to their surface activityproperties; and makes them readily suscep-tible to proteolysis, which is important incheese ripening.

• The caseins are relatively hydrophobic buthave high surface hydrophobicity (owing totheir open structures) rather than a high to-tal hydrophobicity. The hydrophobic, polar,and charged residues are not uniformly dis-tributed throughout the molecular se-quences but occur as hydrophobic or hydro-philic patches (Figure 3-10), giving thecaseins strongly amphiphatic structures thatmake them highly surface active. TheN-terminal 2/3 of K-casein, which is par-ticularly significant in cheese manufacture,

Days

Figure 3-5 Relationship between the growth rate (days to double birthweight) of the young of some species ofmammal and the protein content (expressed as the percentage of total calories derived from protein) of the milk ofthat species.

Cal

orie

s fr

om p

rote

in, %

Man

BuffaloReindeer

Horse

CowSheep

Goat

Pig

RabbitRat

Dog

Cat

Page 47: Cheese Science

Table 3-8 Amino Acid Composition of the Principal Proteins in Milk

a-LactAlbumin B

p-LactoGlobulin A

73_

Casein Ay2-

Casein A271-

Casein A2K-

Casein BP-

Casein A2OC52-

Casein AGVr

Casein BAminoAcid

91277O852638618

13444

1231O

12314,174

4.68

11587O

16983

145

104

1022442

1523O

16218,362

5.03

2147O4

112122O

1043

14351332O

10211,557

6.29

2147O4

112122O

1043

14351442O

1041 1 ,822

6.23

438

101

11213445O

1767

19491

1052O

18120,520

5.85

47

14121

1214202

152

112

1389419O^

51

20923,980

5.58

459

115

18213555O

196

1022491

1154O

16919,005

5.12

414156

11251510282

145

11131262

2436O

20725,228

4.64

78588

24151799O

115

11171082

1456O

19923,612

4.89

AspAsnThrSerSerPGIuGInProGIyAla1/2 CysVaIMetHeLeuTyrPheTrpLysHisArgPyroGlu

Total residuesMolecular weightHOave (kJ/residue)

Page 48: Cheese Science

is strongly hydrophobic, whereas the C-ter-minal 1/3 is strongly hydrophilic. The hy-drophobicity of the caseins explains whytheir hydrolysates have a high propensity tobitterness, which is one of the principal de-fects in many cheese varieties.

• K-Casein is glycosylated (asr, as2-, and p-caseins are not). It contains galactose, ga-lactosamine, and 7V-acetylneuraminic acid(sialic acid), which occur as either trisac-charides or tetrasaccharides attached tothreonine residues in the C-terminal re-gion. K-Casein may contain O to 4 tri- ortetrasaccharides moieties (i.e., 10 variantsof K-casein exist). The presence of oligo-

saccharides attached to the C-terminal ofK-casein increases the hydrophilicity ofthat region.

• All the caseins exhibit genetic polymor-phism that involves the substitution of 1 or2 amino acids and rarely the deletion of asegment. The variant or variants present inmilk are determined by simple Mendeliangenetics. The presence of certain geneticvariants in milk has a significant effect onthe cheesemaking properties of the milk.

The preceding indicates that the casein systemis extremely heterogeneous and that a logicalnomenclature system is necessary. The follow-

1 Glu-Val-Leu-Asn-Glu-Asn-Leu-H. Arg-Pro-Lys-His-Pro-Ile-Lys-His-Gln-Gly-Leu-Pro-Gln-

21Leu-Arg-Phe-Phe-Val-Ala-(Variants B, C, D, E)

-Pro-Phe-Pro-Glu-Val-Phe-Gly-Lys-Glu-Lys-Val-Asn-Glu-Leu(Variant A)

41 Ala (Variants A, B, C, E) GIn (Variants A, B, C, D)Ser-Lys-Asp-Ile-Gly-StrP-Glu-SerP-Thr-Glu-Asp-Gln- -Met-GIu-Asp-Ile-Lys- -Met

ThrP (Variant D) GIu (Variant E)61Glu-Ala-Glu-SerP-Ile-ScrP-SerP-ScrP-Glu-Glu-Ile-Val-Pro-Asn-SerP-Val-Glu-Gln-Lys-His-

81Ile-Gln-Lys-Glu-Asp-Val-Pro-Ser-Glu-Arg-Tyr-Leu-Gly-Tyr-Leu-Glu-Gln-Leu-Leu-Arg-

101Leu-Lys-Lys-Tyr-Lys-Val-Pro-Gln-Leu-Glu-Ile-Val-Pro-Asn-SerP-Ala-Glu-Glu-Arg-Leu-

121His-Ser-Met-Lys-Glu-Gly-Ile-His-Aia-Gln-Gln-Lys-Glu-Pro-Met-Ile-Gly-Val-Asn-Gln-

141Glu-Leu-Ala-Tyr-Phe-Tyr-Pro-Glu-Leu-Phe-Arg-Gln-Phe-Tyr-Gln-Leu-Asp-Ala-Tyr-Pro-

161Ser-Gly-Ala-Trp-Tyr-Tyr-Val-Pro-Leu-Gly-Thr-Gln-Tyr-Thr-Asp-Ala-Pro-Ser-Phe-Ser-

181 GIu (Variant A, B, D) 199Asp-Ile-Pro-Asn-Pro-Ile-Gly-Ser-Glu-Asn-Ser- -Lys-Thr-Thr-Met-Pro-Leu-Trp. OH

GIy (Variant C, E)

Figure 3-6 Amino acid sequence of asl-casein, showing the amino acid substitutions or deletions in the principalgenetic variants.

Page 49: Cheese Science

ing nomenclature has been adopted: The caseinfamily is indicated by a Greek letter with a sub-script, if necessary: ocsr, ocs2-, P-, K-. The Greekletter is followed by CN, and the genetic variantis indicated by a Latin letter (A, B, C, etc.) with asuperscript, if necessary: asrCN B, p-CN A1.Finally, the number of phosphate residues is in-dicated: OC81-CN B-8P, p-CN Al-5P.

Minor components of the casein system arethe y-caseins, which are C-terminal fragments ofp-casein produced through the action of the in-digenous proteinase plasmin. The N-terminalfragments are included in the so-called proteosepeptone fraction of milk protein. These peptidesare summarized in Figure 3-11.

Ovine, caprine, and buffalo caseins generallyexhibit heterogeneity similar to that of the bo-vine caseins.

3.4.2 Casein Micelles

As mentioned earlier, ocsr, ocs2-, and p-caseins,which together constitute about 85% of wholecasein, are precipitated by concentrations of Cagreater than 6 mM. Since bovine milk containsabout 30 mM Ca, it might be expected that thesecaseins would precipitate in milk. However, K-casein, which contains only 1 mol PCVmol, isinsensitive to Ca2+ and, moreover, can stabilizeup to 10 times its weight of the Ca-sensitivecaseins against precipitation by Ca2+. It does thisvia the formation of a type of quaternary struc-ture, referred to as the casein micelle.

The principal properties of the casein micelleare summarized in Table 3-9. Many attemptshave been made to elucidate the structure of themicelle. The most widely accepted view is that

1H. Lys-Asn-Thr-Met-Glu-His-Val-SerP-SerP-SerP-Glu-Glu-Ser-Ile-Ile-SerP-Gln-Glu-Thr-Tyr-

21Lys-Gln-Glu-Lys-Asn-Met-Ala-Ile-Asn-Pro-Ser-Lys-Glu-Asn-Leu-Cys-Ser-Thr-Phe-Cys-

41Lys-Glu-VaI-Val-Arg-Asn-Ala-Asn-Glu-Glu-Glu-Tyr-Ser-Ile-Gly-SerP-SerP-SerP-Glu-Glu-

61SerP-Ala-Glu-Val-Ala-Thr-Glu-Glu-Val-Lys-Ile-Thr-Val-Asp-Asp-Lys-His-Tyr-Gln-Lys-

81Ala-Leu-Asn-Glu-Ile-Asn-Gln-Phe-Tyr-Gln-Lys-Phe-Pro-Gln-Tyr-Leu-Gln-Tyr-Leu-Tyr-

101Gln-Gly-Pro-Ile-Val-Leu-Asn-Pro-Trp-Asp-Gin-Val-Lys-Arg-Asn-Ala-Val-Pro-Ile-Thr-

121Pro-Thr-Leu-Asn-Arg-Glu-Gln-Leu-SerP-Thr-ScrP-Glu-Glu-Asn-Ser-Lys-Lys-Thr-Val-Asp-

141Met-Glu-Sei-P-Thr-Glu-Val-Phe-Thr-Lys-Lys-Thr-Lys-Leu-Thr-Glu-Glu-Glu-Lys-Asn-Arg-

161Leu-Asn-Phe-Leu-Lys-Lys-Ile-Ser-Gln-Arg-Tyr-Gln-Lys-Phe-Ala-Leu-Pro-Gln-Tyr-Leu-

181Lys-Thr-Val-Tyr-Gln-His-Gln-Lys-Ala-Met-Lys-Pro-Trp-Ile-Gln-Pro-Lys-Thr-Lys-Val-

(Leu)201 207Ile-Pro-Tyr-Val-Arg-Tyr-Leu. OH

Figure 3-7 Amino acid sequence of bovine ocs2-casein A, showing 11 of the 13 potential phosphorylation sites.

Page 50: Cheese Science

the micelles are composed of submicelles ofmass around 5 x 106 kDa. The core of thesubmicelles is considered to consist of the Ca-sensitive asr, as2-, and p-caseins, with variableamounts of K-casein located principally on thesurface of the submicelles. The K-casein-defi-cient submicelles are located in the center of themicelles, and the K-casein-rich submicelles areconcentrated at the surface. The hydrophobic N-terminal segment of K-casein is considered to in-teract hydrophobically with the Ca-sensitivecaseins, with the hydrophilic C-terminal seg-ment protruding from the surface, giving the

whole micelle a hairy appearance (Figures 3-12and 3-13). The colloidal stability of the micellesis attributed to a zeta potential of about -20 mVat 2O0C and the steric stabilization provided bythe protruding hairs. The submicelles are consid-ered to be held together by microcrystals of cal-cium phosphate and perhaps hydrophobic andhydrogen bonds.

Although this model of the casein micelle isnot universally accepted, it is adequate to ex-plain many of the technologically importantproperties of the micelles, including rennet co-agulation, which follows the specific hydrolysis

iH.Arg-Glu-Leu-Glu-Glu-LeuAsn-Val-Pro-Gly-GIu-Ile-VaI"GIu-5erP-Leu5er/>-5erP-5erP-GIu-

21 *—^ Yi -caseins (Variant C)

GIu-Ser-Ile-Thr-Arg-Ile-Asn-Lyskys-Ile-Glu-Lys-Phe-Gln-Ser -Glu-LyS-GIn-Gln-Gln-SerP GIu

(Variants A, B)41Thr-Glu-Asp-Glu-Leu-Gln-Asp-Lys-Ile-His-Pro-Phe-Ala-GIn-Thr-Gln-Ser-Leu-Val-Tyr-

61 Pro (Variants A2, A3)Pro-Phe-Pro-Gly-Pro-Ile- -Asn-Ser-Leu-Pro-GIn-Asn-Ile-Pro-Pro-Leu-Thr-Gln-Thr

His (Variants C, A1' and B)

81

Pro- VaI-VaI- Val-Pro-Pro-Phe-Leu-Gln-Pro-Glu- Val-Met-Gly- Val-Ser-Lys- Val-Lys-Glu-

^ Y3-caseins101(Variants A1, A2, B, C) His IAla-Met-Ala-Pro-Lvsr -LyslGlu-Met-Pro-Phe-Pro-Lys-Tyr-Pro-Val-Glu-Pro-Phe-Thr-

(VariantA5)|Gln121 Ser (Variants A, Q ^2 -caseins

GIu- -Gln-Ser-Leu-Thr-Leu-Thr-Asp-Val-Glu-Asn-Leu-His-Leu-Pro-Leu-Pro-Leu-Leu-Arg (Variant B)

141

Gln-Ser-Trp-Met-His-GIn-Pro-His-Gln-Pro-Leu-Pro-Pro-Thr-Val-Met-Phe-Pro-Pro-Gln-

161

Ser-Val-Leu-Ser-Leu-Ser-Gln-Ser-Lys-Val-Leu-Pro-Val-Pro-Gln-Lys-AIa-Val-Pro-Tyr-

181

Pro-Gln-Arg-Asp-Met-Pro-Ile-Gln-Ala-Phe-Leu-Leu-Tyr-GIn-Glu-Pro-Val-Leu-Gly-Pro-

201 209

Val-Arg-Gly-Pro-Phe-Pro-Ile-Ile-VaLOH

Figure 3-8 Amino acid sequence of bovine p-casein, showing the amino acid substitutions in the genetic vari-ants and the principal plasmin cleavage sites (T).

Page 51: Cheese Science

of the micelle-stabilizing K-casein, as a result ofwhich the stabilizing surface layer is lost.

As far as is known, the structure of the caseinmicelles in bovine, ovine, caprine, and buffalomilks is essentially similar.

3.4.3 Whey Proteins

The whey protein fraction of bovine, ovine,caprine, and buffalo milk contains four mainproteins: (3-lactoglobulin (P-Ig, 50%), oc-lactal-bumin (oc-la, 20%), blood serum albumin (BSA,10%), and immunoglobulins (Ig, 10%; mainlyIgGi, with lesser amounts of IgG2, IgA, andIgM). Human milk contains no p-lg and the prin-cipal Ig is IgA.

The principal properties of the whey proteinsare listed in Table 3-8. In contrast to the caseins,the whey proteins possess high levels of second-ary, tertiary, and quaternary structures. They aretypical globular proteins and are denatured upon

heating (e.g., completely at 9O0C x 10 min).They are not phosphorylated and are insensitiveto Ca2+. All whey proteins contain intramolecu-lar disulfide bonds that stabilize their structure.P-Lg contains one sulfydryl group that undercertain conditions can undergo sulfydryl-disul-fide interactions with other proteins; the mostimportant of these interactions, with K-casein,occurs upon heating at about 750C x 15 s. Thelatter can markedly impair the rennet coagula-tion properties of milk and alter the gel structureand rheological and synertic properties of acidgel-based products such as yogurt and freshcheeses.

The whey proteins are not directly involved incheese manufacture. However, they are indi-rectly involved, as in these examples:

• Heat-induced interaction of whey proteinswith K-casein has undesirable effects onrennet coagulation.

1Pyro-Glu-Glu-Gln-Asn-Gln-Glu-Gln-Pro-Ile-Arg-Cys-Glu-Lys-Asp-Glu-Arg-Phe-Phe-Ser-Asp-

21Lys-Ile-Ala-Lys-Tyr-Ile-Pro-Ile-Gln-Tyr-Val-Leu-Ser-Arg-Tyr-Pro-Ser-Tyr-Gly-Leu-

41Asn-Tyr-Tyr-Gln-Gln-Lys-Pro-Val-Ala-Leu-Ile-Asn-Asn-Gln-Phe-Leu-Pro-Tyr-Pro-Tyr-

61Tyr-Ala-Lys-Pro-Ala-Ala-Val-Arg-Ser-Pro-Ala-Gln-Ile-Leu-Gln-Trp-Gln-Val-Leu-Ser-

81Asn-Thr-Val-Pro-Ala-Lys-Ser-Cys-Gln-Ala-Gln-Pro-Thr-Thr-Met-Ala-Arg-His-Pro-His-

101 105|l06Pro-His-Leu-Ser-Ph^Met-Ala-Ile-Pro-Pro-Lys-Lys-Asn-Gln-Asp-Lys-Thr-Glu-Ile-Pro-

121 He (Variant B)Thr-Ile-Asn-Thr-Ile-Ala-Ser-Gly-Glu-Pro-ITir-Ser-T/ir-Pro-TTzr- -Glu-Ala-Val-Glu-

Thr (Variant A)

141 Ala (Variant B)Ser-Thr-Val-Ala-Thr-Leu-Glu- - SerP-Pro-Glu-Val-Ile-Glu-Ser-Pro-Pro-Glu-Ile-Asn-

Asp (Variant A)

161 169Thr-Val-Gln-Val-Thr-Ser-Thr-Ala-VaLOH

Figure 3-9 Amino acid sequence of bovine K-casein, showing the amino acid substitutions in genetic poly-morphs A and B and the chymosin cleavage site, (T). The sites of posttransitional phosphorylation orglycosylation are italicized.

Page 52: Cheese Science

Figure 3-10 Schematic representation of the distribution of hydrophobic and charged residues in the principal milk proteins.

Page 53: Cheese Science

• Whey proteins are incorporated into cheesemade from milk concentrated by ultrafiltra-tion.

• Whey proteins are heat-denatured in themanufacture of some Quarg products.

• Valuable functional proteins are recoveredfrom whey.

3.4.4 Minor Proteins

Milk contains numerous minor proteins. Theseare found mainly in the whey, but some are alsofound in the fat globule membrane. These minorproteins include enzymes (perhaps 60), enzymeinhibitors, metal-binding proteins (especiallylactoferrin and osteopontin), vitamin-bindingproteins, and several growth factors. As far as isknown, most of these are of no consequence incheese. Some of the indigenous enzymes are ac-tive in cheese during ripening, especially plasminand xanthine oxidase and possibly acid phos-phatase. Lipoprotein lipase is probably quite im-portant in raw milk cheese and perhaps even inpasteurized milk cheese, since some probablypartially survives HTST (high temperature, shorttime) pasteurization. The significance of other

indigenous enzymes in cheese has not been inves-tigated and perhaps warrants study.

3.5 MILK SALTS

After milk has been heated in a muffle furnaceat around 60O0C for 5 hr, a residue (ash), repre-senting roughly 0.7 g/100 ml of the mass of themilk sample, remains. The ash contains the inor-ganic salts present in the original milk plus someelements, especially phosphorus, present origi-nally in organic molecules, especially proteinsand phospholipids, and lesser amounts of sugarphosphates and high-energy phosphates. The el-ements in the ash are changed from their originalform; they are present, not as their original salts,but as oxides and carbonates. Organic salts, themost important of which is citrate, are lost onashing. Fresh milk does not contain lactic acid,but lactic acid may be present in stored milk as aresult of microbial growth. Although the salts ofmilk are quantitatively minor constituents, theyare of major significance to its technologicalproperties.

The typical concentration of the principal ele-ments or compounds that constitute the salts of

Figure 3-11 Principal peptides produced from p-casein by plasmin.

Page 54: Cheese Science

milk are summarized in Table 3-10. Some ofthe salts are present in milk at concentrationsbelow their solubility limit and are thereforefully soluble. However, others, especially cal-cium phosphate, exceed their solubility and oc-cur partly in solution and partly in the colloidalphase, associated mainly with the casein mi-celles. These salts are collectively referred to asmicellar or colloidal calcium phosphate (CCP),although several other elements or ions arepresent also. Several elements are also presentin the MFGM, mainly as constituents of en-

zymes. There are several techniques for parti-tioning the colloidal and soluble salts (see Fox& McSweeney, 1998). Typical distributions areindicated in Table 3-10.

It is possible to either determine experimen-tally or to calculate (after making certain assump-tions) the concentration of the principal ions inmilk; these are also indicated in Table 3-10.

From a cheesemaking viewpoint, the mostimportant salts or ions are calcium, phosphate,and, to a lesser extent, citrate. As shown in Table3-10, bovine milk contains about 1200 mg Ca/L

Figure 3-12 Submicelle model of the casein micelle.

Submicelle

Protrudingchain

Calciumphosphate

Table 3-9 Average Characteristics of Casein Micelles

Characteristic

DiameterSurface areaVolumeDensity (hydrated)MassWater contentHydrationVoluminosltyMolecular weight (hydrated)Molecular weight (dehydrated)Number of peptide chainsNumber of particles per milliliter of milkSurface area of micelles per milliliter of milkMean free distance

Value

120 nm (range: 50-500 nm)8 x 10-10cm2

2.1 x10-15cm3

1 .0632 g cm-3

2.2x10-15g63%3.7 g H2O g~1 protein4.4 cm3 g-1

1.3 x 1O9Da5 x 1O8Da104

1014-1016

5x104cm2

240 nm

Page 55: Cheese Science

(i.e., 30 mM). About 30% is soluble, most ofwhich occurs as un-ionized salts of citrate, butabout 30% exists as Ca2+, which means that 10%of the total calcium exists as Ca2+ (2-3 mM). Al-though present at low concentrations, Ca2+ are ofmajor significance in various aspects of the ren-net coagulation of milk (see Chapter 6). The[Ca2+] is inversely related to the citrate concen-tration.

The insoluble calcium occurs mainly associ-ated with the casein micelles, either as colloidalcalcium phosphate (CCP) or casein Ca. CCPplays a major role in micellar integrity and has a

very significant role in rennet coagulation. Theprecise composition and structure of CCP arenot known. The simplest possible structure istertiary phosphate, Ca3(PO4)2, but the form forwhich the best experimental evidence exists isbrushite, CaHPO4.2H2O, which forms micro-crystals with organic casein phosphate.

3.6 pH OF MILK

As will become apparent in subsequent chap-ters, pH is a critical factor in several aspects ofthe manufacture and ripening of cheese curd.

Figure 3-13 Model of the casein micelle.

Page 56: Cheese Science

Table 3-11 Some Physical Properties of Milk

Property

Osmotic pressureWater activity, aw

Boiling pointFreezing pointRedox potential, Eh (in equilibrium with air

at 250C and pH 6.6)Refractive index, nD

20

Specific refractive indexDensity (2O0C)Specific gravity (2O0C)Specific conductanceIonic strengthSurface tension (2O0C)Coefficient of viscosityThermal conductivity (2.9% fat)Thermal diffusivity (15-2O0C)Specific heatpH (at 250C)Titratable acidity

Coefficient of cubic expansion (273-333 K)

Value

- 700 kPa« 0.993-100.150C-0.5220C (approximately)+0.25 to +0.35 V

1 .3440 to 1 .3485- 0.2075-1 030 kg x nrr3

-1.0321- 0.0050 ohm-1 crrr1

- 0.07 M- 52 N m-1

2.127 mPaxs- 0.559 Wm-1' K-1

-1.25x10-7m2 -s-1

- 3.931 kJ • kg-1 • K~1

-6.61 .3-2.0 meq OH- per 100 ml(0.14-0.16% as lactic acid)0.0008 m3 • m-3 • K-1

Table 3-10 Concentration and Partition of Milk Salts

Soluble

Species

SodiumPotassiumChlorideSulphatePhosphate

Citrate

Calcium

Magnesium

Concentration (mg/L)

5001,4501,200

100750

1,750

1,200

130

Percentage

9292

10010043

94

34

67

Form

Completely ionizedCompletely ionizedCompletely ionizedCompletely ionized10% bound to Ca and Mg51% H2PO4-39% HPO4

2-85% bound to Ca and Mg14%Citrate3-1%H.citrate2-35% Ca2+

55% bound to citrate10% bound to phosphateProbably similar to calcium

Colloidal (%)

88

57

66

33

Page 57: Cheese Science

The pH of milk at 250C is usually in the range6.5 to 7.0, with a mean value of 6.6. pH in-creases with advancing lactation and may ex-ceed 7.0 in very late lactation; colostrum canhave a pH as low as 6.0. The pH increases dur-ing mastitic infection owing to the increasedpermeability of the mammary gland mem-branes, which permits greater influx of bloodconstituents into the milk (the pH of cow'sblood is 7.4). The difference in pH betweenblood and milk results from the active transportof various ions into the milk; precipitation ofCCP, which results in the release of H+ duringthe synthesis of casein micelles; higher concen-trations of acidic groups in milk; and the rela-tively low buffering capacity of milk betweenpH 6.0 and 8.0.

One of the key events during the manufactureof cheese is the production of lactic acid fromlactose by lactic acid bacteria (see Chapter 5).Consequently, the pH decreases to about 5.0.While lactic acid is primarily responsible for thedecrease in pH, the actual pH attained isstrongly affected by the buffering capacity ofthe milk and curd.

Milk contains a range of groups that are effec-tive in buffering over a wide pH range. The prin-cipal buffering compounds in milk are its salts(particularly soluble phosphate, citrate, and bi-carbonate) and acidic and basic amino acid side-

REFERENCES

Cayot, P., & Lorient, D. (1998). Structures et techno-fonctions desproteins du lait. Paris: Lavoisier.

Fox, P.P. (Ed.). (1982). Developments in dairy chemistry:Vol. L Proteins. London: Applied Science Publishers.

Fox, P.F. (Ed.). (1983). Developments in dairy chemistry:Vol. 2. Lipids. London: Applied Science Publishers.

Fox, P.F. (Ed.). (1985). Developments in dairy chemistry:Vol. 3. Lactose and minor constituents. London: ElsevierApplied Science Publishers.

Fox, P.F. (Ed.). (1989). Developments in dairy chemistry:Vol. 4. Functional proteins. London: Elsevier AppliedScience Publishers.

Fox, P.F. (Ed.). (1992). Advanced dairy chemistry, Vol. 1.Milk proteins. London: Elsevier Applied Science Pub-lishers.

chains of proteins (particularly the caseins). Thecontribution of these components to the buffer-ing of milk is discussed in detail by Singh,McCarthy, and Lucey (1997).

The buffering capacity of milk and curd is ofsignificance during cheesemaking, since it is thefactor that determines the rate of decrease in pHcaused by the production of lactic acid by thestarter. The buffering capacity of milk is lownear its natural pH but increases rapidly to amaximum at about pH 5.1. This means that,given a steady rate of acidification by the starter,the pH of milk decreases rapidly initially andlater slows down. Since all of the soluble andsome of the colloidal calcium phosphate are lostin the whey, it is not surprising that the bufferingproperties of cheese differ from those of milk.Cheddar and Emmental cheeses have maximumbuffering capacities at around pH 4.8.

3.7 PHYSICOCHEMICAL PROPERTIESOF MILK

Information on the physicochemical proper-ties of milk is important when developing andprocessing dairy products, designing processingequipment, and using dairy products in foodproducts. Some of the principal physicochemi-cal properties are summarized in Table 3-11.

Fox, P.F. (Ed.). (1995). Advanced dairy chemistry, Vol. 2.Lipids (2d ed.). London: Chapman & Hall.

Fox, P.F. (Ed.). (1997). Advanced dairy chemistry, Vol. 3.Lactose, water, salts and vitamins. London: Chapman &Hall.

Fox, P.F., & McSweeney, P.L.H. (1998). Dairy chemistryand biochemistry. London: Chapman & Hall.

Jenness, R., & Patton, S. (1959). Principles of dairy chemis-try. New York: Wiley.

Jensen, R.G. (Ed.). (1995). Handbook of milk composition.San Diego, CA: Academic Press.

Singh, H., McCarthy, O.J., & Lucey, J.A. (1997). Physico-chemical properties of milk. In P.F. Fox (Ed.), Advanceddairy chemistry: Vol. 3. Lactose, water, salts and vita-mins. London: Chapman & Hall.

Page 58: Cheese Science

Walstra, P., & Jenness, R. (1984). Dairy chemistry and phys-ics. New York: Wiley.

Webb, B.H., & Johnson, A.H. (Eds.). (1965). Fundamentalsof dairy chemistry. Westport, CT: AVI Publishing.

Webb, B.H., Johnson, A.H., & Alford, J.A. (Eds.). (1974).

Fundamentals of dairy chemistry (2d ed.). Westport, CT:AVI Publishing.

Wong, N.P., Jenness, R., Keeney, M., & Marth, E.H. (Eds.).(1988). Fundamentals of dairy chemistry (3d ed.).Westport, CT: AVI Publishing.

Page 59: Cheese Science

4.1 CONTAMINATION OF RAW MILK

The pH of milk (around 6.6), its temperaturein the udder (around 380C), and its high nutri-tional value are ideal for the growth of bacteria.However, bacteria growth does not usually oc-cur because milk in the udder is sterile, unlessthe udder is infected. Bacteria can colonize theteat canal but are expelled in the first few squirtsof milk. However, during milking, the milkbecomes contaminated with microorganisms,mainly from the milking equipment, and it will,if maintained at a temperature above 150C forseveral hours, coagulate due to the production ofacid by adventitious bacteria, such as lactic acidbacteria (LAB) and coliforms. Therefore, greatcare must be taken to ensure that milk is pro-duced hygienically. Today, there is no difficultyin producing milk with less than 5,000 colony-forming units (cfu) per ml, whereas 30 years agoit was difficult to produce it with less than100,000 cfu/ml. These improvements in the mi-crobial quality of raw milk are due to better hy-giene during milking; improved design of milk-ing equipment, making it easier to clean; coolingof the milk to a temperature less than 50C withina few hours of production; and holding the milkat less than 50C in easily cleaned, stainless steelbulk-storage tanks until it is collected and trans-ported to the factory.

The sources of microorganisms in milk in-clude the udder (unhealthy animals and the teatcanal), the exterior of the udder (outside surfaces

of the teats and udder), the bedding on which thecow lies, food eaten by the cow, the milker, theair, water used to wash the udder, and the milk-ing and storage equipment.

Cows suffering from diseases like salmonel-losis, tuberculosis, and brucellosis may shed thebacteria that cause these diseases into their milk.Normally, milk from cows infected with thesediseases is not a major source of bacteria in rawmilk. Mastitis is a bacterial infection of themammary gland and is common in dairy cows. Itcan occur in subclinical (the majority of out-breaks) or clinical forms and is caused mainly byStaphylococcus aureus, although Streptococcusdysgalactiae, Sc. agalactiae, Escherichia coli,and Corynebacterium spp. may also be respon-sible. S. aureus is a Gram-positive coccus, andmany strains produce heat-stable toxins, termedenterotoxins, which can cause food poisoning.Generally, growth to around 106 cfu/ml is neces-sary before sufficient toxin is produced to causefood poisoning. In subclinical mastitis, no physi-cal change or abnormality is evident in the milk,whereas in clinical mastitis, large clots con-sisting of a mixture of milk, somatic cells, andbacteria are produced. Subclinical mastitis isgenerally manifested by increased numbers ofpolymorphonuclear leucocyte (PMNs) cells inthe milk. In clinical mastitis, the number of bac-teria in the milk varies depending on whetherphagocytosis (i.e., engulfment by the PMNs) hasoccurred. At the beginning of infection, beforephagocytosis has occurred, several million bac-

Bacteriology of Cheese Milk

CHAPTER 4

Page 60: Cheese Science

teria per ml may be present, but as phagocytosisdevelops, the numbers present will decrease rap-idly to perhaps less than 1,000 bacteria per ml. Inother words, bacterial numbers are high at thebeginning of infection but very low as the infec-tion progresses.

Small numbers of microorganisms can enterthe teat canal of healthy animals from the out-side of the teat but are generally washed out inthe first squirts of milk. Less than 200 cm/ml areprobably added to the milk from the teat canalduring milking.

During milking, milk can also become con-taminated with bacteria from the air, the outsideof the udder, the bedding, the feed, and themilker, but these are generally minor sources ofcontamination. However, extremely dirty uddersmay contaminate milk with up to 105 cfu/ml.Dirty udders are more likely to occur in winter,when the cows are housed, than in summer,when the cows are on pasture. Therefore, it isimportant to wash the udders and teats thor-oughly before milking.

The major source of contamination of rawmilk is improperly cleaned milking equipment.For this reason, considerable emphasis is placedon the satisfactory cleaning of the milking ma-chine, its associated rubber hoses and pipework,and the bulk-storage tank. The machine shouldbe cleaned after each milking, and the bulk-stor-age tank after it has been emptied. Hot and colddetergent washes are used and generally a hotacid rinse is given once a week to prevent thebuild up of "milk stone," which can harbor bac-teria and make the equipment difficult to clean.Milk stone is composed mainly of calcium phos-phate, but sufficient nutrients may also be pres-ent to allow significant microbial growth be-tween milkings if the ambient temperature isgreater than 150C. Only very heavily contami-nated milking equipment will cause a markedincrease in the bacterial count in the raw milk.For example, 1 million organisms are required toincrease the bacterial count of 1,000 L of milkby 1 bacterium per ml; therefore, to increase thecount by 10,000/ml would require the addition

of 1010 bacteria. The milking machine and its as-sociated pipe lines and rubber hoses have a largesurface area and may harbor such large numbersof bacteria if they are not adequately cleaned.Residues of milk left on the equipment after in-adequate cleaning may contain sufficient nu-trients to sustain bacterial growth at ambienttemperatures. Immediately after milking, goodquality milk produced using properly cleanedmilking machines and bulk-storage tanks shouldhave a count of less than 5,000 cfu/ml.

Gram-positive bacteria (e.g., Micrococcus,Corynebacterium, Microbacterium, Lactobacil-lus, Lactococcus, Enterococcus, etc.) and Gram-negative bacteria (Pseudomonas, Achromobac-ter, Enterobacter, Escherichia, Flavobacterium,etc.) are found in milk immediately after milk-ing. In the past, milk was either not cooled at allor cooled to ambient temperature with water.Under these conditions, the growth of Gram-positive bacteria, particularly LAB, such asLactobacillus, Lactococcus, Enterococcus, andStreptococcus spp., was more common than thegrowth of Gram-negative microorganisms.Many of the Gram-positive genera includecheese starter bacteria. Nowadays, milk is nor-mally cooled to less than 50C within 1-2 hr ofmilking, and the flora has changed from onedominated by Gram-positive bacteria to onedominated by Gram-negative, psychrotrophicbacteria, particularly of the genera Pseudomo-nas andAchromobacter. Psychrotrophs are gen-erally defined as bacteria capable of growing attemperatures under 70C.

Cooling significantly slows down the rate ofmultiplication of bacteria in raw milk (Figure4-1). However, slow growth of bacteria, particu-larly psychrotrophs, still occurs at 40C, and sig-nificant numbers (e.g., 106 or 107 cfu/ml) can of-ten be reached in 3 or 4 days' storage on the farm.Raw milk may also be stored in silos for 1 or 2days at the factory before use, during which fur-ther growth of psychrotrophs will occur. It ismore important to use properly cleaned milkingequipment than to cool the milk rapidly. In otherwords, rapid cooling of milk will not compensate

Page 61: Cheese Science

for improperly cleaned milking machines andstorage tanks, either on the farm or at the factory.

The growth of bacteria in four raw milks duringstorage at 50C is shown in Figure 4-2. Little or nogrowth occurred during the first 2 days of storage,after which two of the milks showed a significantincrease in bacterial numbers and the other twodid not. This difference was probably due to dif-ferences in the species of organisms present andtheir ability to grow at 50C. It is also noticeablethat the initial level of contamination had littleeffect on the subsequent rates of bacterial growth.After 4 days at 50C, counts in some cases ex-ceeded 107 cfu/ml. Counts of this magnitudewould be totally unacceptable in milk for cheese-making—or indeed for making any dairy product.

Today, most raw milk received by cheese fac-tories in developed countries has a viable countof less than 50,000 cfu/ml. The milk is oftenstored in large silos for perhaps 24 hr or longer,and therefore further growth and contaminationfrom improperly cleaned silos can occur, so thatmilk for cheesemaking may have counts in ex-cess of 105 cfu/ml. Such counts, while high, willnot have a major effect on cheese quality. How-ever, counts greater than 106 cfu/ml in the milkbefore pasteurization could affect cheese qual-ity, because many psychrotrophs, especiallyPseudomonas spp., produce heat-stable lipases

and proteinases, which withstand heating to10O0C for 30 min, although the bacteria that pro-duce these enzymes are killed. These enzymesmay be retained in the curd during cheese-making and cause off-flavors to develop duringripening, especially in semi-hard and hard vari-eties, which are ripened for a long time (e.g.,Cheddar, Gouda, Comte, etc.).

4.2 PASTEURIZATION

Pasteurization of milk for cheesemaking be-came widespread after about 1940, primarily forpublic health reasons (Mycobacterium tubercu-losis, the organism that causes tuberculosis, iskilled by pasteurization) but also to provide amilk supply of more uniform bacteriologicalquality and to increase the keeping quality of thecheese. Batch pasteurization (low temperature,long time [LTLT] pasteurization; 63-650C x 30min) was used initially but was replaced by con-tinuous high temperature, short time (HTST)pasteurization (720C x 15 s). A line diagram ofan HTST pasteurizer is shown in Figure 4-3.

Most (> 99.9%) of the bacteria found in rawmilk are heat labile and are killed by pasteuriza-tion at 720C for 15 s; most milk for cheese-making is subjected to this heat treatment. Pas-teurization kills all potential pathogens that

Time,h

Figure 4-1 Effect of temperature on the growth of bacteria in a sample of raw milk.

cfu/

ml

3OC

2OC

1OC

5 C

Page 62: Cheese Science

might be present in the milk, but spores ofClostridium and Bacillus are not killed by thistreatment. In addition, organisms like Micrococ-cus, Microbacterium, and Enterococcus spp.,

which can withstand this heat treatment, arefound in raw milk. These are called thermoduricbacteria and invariably come from improperlycleaned equipment. Generally, thermoduric bac-

Ti me, days

Figure 4-2 Growth of bacteria in four raw milks incubated at 50C.

cfu/

ml

A

C

O

F

Figure 4-3 Layout of HTST pasteurizer (the insert shows a schematic diagram of the heat exchange sections):(A) feed tank, (B) balance tank, (C) feed pump, (D) flow controller, (E) filter, (P) product, (S) steam injection (hotwater section), (V) flow diversion valve, (MW) mains water cooling, (CW) chilled water, (TC) temperaturecontroller, (1) regeneration, (2) hot water section, (3) holding tube, (4) mains cooling water, and (5) chilled watercooling.

TC

Page 63: Cheese Science

teria grow only slowly, if at all, in raw milk (En-terococcus spp. are exceptions), so that counts ofthermoduric bacteria, even within 24 hr of milk-ing, are a useful indicator of how well the milk-ing equipment was cleaned. These bacteria canbe enumerated by heat-treating the milk at 630Cfor 30 min before plating. In cheese factories,another thermoduric bacterium, Streptococcusthermophilus, can grow as a biofilm in the re-generation section of the pasteurizer during longruns. This organism grows rapidly in milk and isalso used as a starter culture (see Chapter 5).

Pasteurization also inactivates several en-zymes in the milk, including lipase and alkalinephosphatase. Lack of alkaline phosphatase ac-tivity in milk indicates that the milk has beenproperly pasteurized.

In some countries, such as Canada, signifi-cant amounts of cheese are made from milkheat-treated to a temperature lower than pas-teurization. This treatment is called thermiza-tion and generally involves heating the milk to630C for 10-15 s. This treatment results in lessinactivation of enzymes and nonstarter lacticacid bacteria (NSLAB), which may be impor-tant in developing cheese flavor. Only somepathogenic and food-poisoning microorganismsare killed by thermization, and the milk must besubsequently fully pasteurized to meet publichealth regulations. The purpose of thermizationis to kill psychrotrophs, which dominate themicroflora of refrigerated milk and excrete po-tent proteinases and lipases, which may causeflavor and textural defects in cheese. Such treat-ments are also used in Europe to reduce thenumber of microorganisms in the raw milk as itis taken into the factory and thereby to prolongthe keeping quality of raw milk, but in thesecases the milk is subsequently pasteurized be-fore cheesemaking.

While pasteurization reduces the risk of pro-ducing low-quality cheese resulting from thegrowth of undesirable bacteria and kills patho-gens and food-poisoning microorganisms, pas-teurization of cheese milk may damage itscheesemaking properties if the heat treatment istoo severe (owing to heat-denaturation of the

whey proteins and their interaction withK-casein; see Chapter 6). The extent of this dam-age is negligible under HTST pasteurizationconditions. The flavor of cheese made from pas-teurized milk develops more slowly and is lessintense than that made from raw milk, appar-ently because certain components of the micro-flora of raw milk contribute positively to cheeseflavor. To overcome this deficiency, adjunctcultures of selected NSLAB are being recom-mended for use in the manufacture of long-ripened, low-moisture cheese made from pas-teurized milk. This topic is discussed fully inChapters 11 and 15.

Many cheeses, including such famous va-rieties as Emmental, Gruyere, Comte, andParmigiano-Reggiano, are still produced, at bothfactory and farmhouse level, from raw milk. Oneof the important safety factors in these cheeses isthat all are cooked to a high temperature(> 5O0C) for up to 1 hr, which kills some of thebacteria in the raw milk. Legislation in manycountries requires that cheese be made fromHTST pasteurized milk; that it be aged for atleast 60 days, during which food-poisoning orpathogenic bacteria die; or the cheese itself bepasteurized (i.e., converted to processedcheese). The public health aspects of cheese arediscussed in Chapter 20.

4.3 ALTERNATIVES TO HEATTREATMENT

There are at least four alternatives to heattreatment for reducing the numbers of bacteria incheese milk:

• treatment with hydrogen peroxide (H2O2)• activation of the lactoperoxidase-H2O2-thi-

ocyanate system• bactofugation• microfiltration

4.3.1 Treatment with Hydrogen Peroxide

Hydrogen peroxide (H2O2) is a very effectivebactericidal agent, the use of which is permitted

Page 64: Cheese Science

in some countries, including the United States.The excess H2O2 is usually destroyed by addingcatalase. The use OfH2O2 to treat cheese milk isnot practiced commercially to any great extent.

4.3.2 Lactoperoxidase-H2O2-Thiocyanate

Lactoperoxidase (LPO), an indigenous en-zyme in milk, reduces H2O2 in the presence of asuitable reducing agent:

H2O2 + 2HA -> 2H2O + 2A

One such reducing agent is the thiocyanate anionSCN~, which is oxidized to various species (e.g.,OSCN") that are strongly bactericidal. Milk con-tains a low concentration of indigenous SCN~,arising from the catabolism of glucosinolates(from members of the Cruciferae family) bybacteria in the rumen (Figure 4-4). Milk con-tains no H2O2, which must be added or producedin situ via oxidation of glucose by glucose oxi-dase (Figure 4-5) or from xanthine by xanthine

oxidase (Figure 4-6); or by starter culturesgrown in the presence of O2.

The LPO system is very effective for the"cold pasteurization" of milk. The process hasbeen patented but has attracted limited, if any,interest in developed dairy countries, possiblyfor economic reasons. It is practiced to a smallextent in developing countries.

4.3.3 Bactofugation

A high percentage (98-99%) of somatic andbacterial cells and bacterial spores in milk can beremoved by centrifugation at high gravitationalforces using a special centrifuge called a bac-tofuge. The cells and spores are more dense thanmilk serum and are concentrated in the sludgeduring bactofugation. The sludge is subse-quently sterilized to kill the spores and bacteriaand is then added back to the milk. Bactofu-gation is not widely used in general cheese-making but is commonly used to remove

Glucosinolate

thioglucosidase

a thiocyanate

p-glucose an isothiocyanate a nitrile

Figure 4—4 Enzymatic degradation of a glucosinolate.

Page 65: Cheese Science

Clostridium tyrobutyricum from milk intendedfor Dutch- and Swiss-type cheeses; the latter un-dergo a propionic acid fermentation. Cheesesthat undergo a propionic acid fermentation aregenerally ripened at below 130C for a fewweeks, after which the temperature is increasedto around 220C for 3-4 weeks to promote the fer-mentation. Clostridia, especially Cl tyrobutyri-cum, can also grow under these circumstancesand produce late gas (see Chapter 11), but bacto-fugation is very effective in eliminating sporesof Cl. tyrobutyricum from the milk.

Cl. tyrobutyricum is an obligate anaerobe thatcan ferment lactic acid, the principal acid incheese, and the anaerobic cheese environment isideal for its growth. The major source of the or-ganism in raw milk is improperly fermented si-lage. For this reason, feeding of silage to cows isprohibited in the areas of Switzerland whereEmmental cheese is made. Contamination withspores is much greater in winter, when the cowsare fed indoors with silage, than in summer,when they are on pasture. The vegetative cellsare probably killed by pasteurization (althoughthere is no proof of this), but the spores are heatresistant, requiring several minutes at 10O0C tokill them.

When silage contaminated with clostridia iseaten by cows, the spores pass through their gas-trointestinal tract. The teats and udders of cowslying in their own feces become contaminatedwith fecal material containing clostridial spores,which then contaminate the milk. Proper clean-ing of the teats and udders will reduce contami-nation from this source. However, less than 10spores per 100 ml of milk are sufficient to causelate gas production in Dutch-type cheeses. The

critical number is lower in cheeses that undergoa propionic acid fermentation (i.e., Swiss-typecheeses) because of the higher ripening tempera-ture used for these cheeses.

The design of the bactofuge (Figure 4-7) isessentially similar to that of separators used toseparate the fat from milk but is modified in sucha way that only bacteria and spores, which aremore dense than skim milk, are forced outwardand move down along the lower side of the up-per member of a pair of disks and eventuallythrough orifices in the bowl of the centrifuge as abacterial concentrate called the bactofugate (thisrepresents about 3%, v/v, of milk). Some large,dense casein micelles, perhaps as much as 6% ofthe total casein, are also removed by this pro-cess. The loss of casein will cause a decrease incheese yield that may be avoided by heat-steril-izing the bactofugate and returning it to the milkor by otherwise supplementing the casein con-tent (e.g., by adding ultrafiltration retentate).

4.3.4 Microflltration

Microfiltration is a membrane separation pro-cess, in principle like reverse osmosis, nano-filtration or ultrafiltration, except that largepore-size membranes (0.8-1.4 jam) are used. Thesemipermeable membranes used retain the bac-teria but allow milk constituents, including mostof the casein micelles, to pass through in the per-meate. The process can only be applied to skimmilk, as the fat-globules in whole milk block thepores of the membrane and reduce its efficiency.Therefore, the cream must be separated from themilk, pasteurized, and added back to the micro-filtered skim milk before cheesemaking.

Figure 4-5 Formation OfH2O2 upon oxidation of glucose by glucose oxidase (GO).

gluconic acid

gluconic acid-6-lactoneGlucose

Page 66: Cheese Science

Microfiltration is very efficient at removingbacterial cells (> 99%) and is being used in-creasingly in the dairy industry, such as in theproduction of extra-long life pasteurized milk.It is not yet widely used for cheese milk exceptfor the removal of spores from milk for Swissand similar cheeses. The technique has beenvery useful in studying the effect of the indig-

enous raw-milk microflora and of enzymes inac-tivated by pasteurization on cheese flavor. Thequality of Cheddar and Comte cheeses madefrom microfiltered milk is similar to that madefrom pasteurized milk and is different from thatmade from raw milk, which indicates that thedifferences in flavor between raw and pasteur-ized milk cheeses are due principally to the in-

Uric acid

Figure 4-6 Diagram of catabolism of purine nucleotides, showing the production OfH2O2 from hy-poxanthine and xanthine by xanthine oxidase.

xanthineoxidase

Xanthine(enol form)

guaninedeaminase

xanthineoxidase

Hypoxanthine(keto form)

Ribose

Guanine

nucleosidase

Inosine

adenosinedeaminase

Adenosine

5-nucleotidase

AMP

GMP

5-nucleotidase

Guanosine

nucleosidaseRibose

Page 67: Cheese Science

digenous microorganisms that are efficiently re-moved by microfiltration or killed by pasteur-ization rather than to the inactivation of indig-enous enzymes or other heat-induced changes.

A microfiltration system known as the Bacto-catch System has been developed by the AlfaLaval Company (Sweden) for the decontamina-tion of milk as an alternative to pasteurization.

4.4 PREMATURATION

In some countries, particularly France, lacticcultures are added to the cheese milk, which isthen incubated at 8-1O0C for 12-15 hr (over-

SUGGESTED READINGS

Bramley, A.J., & McKinnon, CH. (1990). The microbiol-ogy of raw milk. In R.K. Robinson (Ed.), Dairy microbi-ology (2d ed., Vol. 1). London: Elsevier Applied Science.

International Dairy Federation. (1996). Symposium on Bac-teriological Quality of Raw Milk. Wolfpassing, Austria,13-15 March 1996. Brussels: International Dairy Federa-tion.

Jelen, P. (1985). Introduction to food processing. Engle-wood Cliffs, NJ: Prentice Hall.

Ledford, R.A. (1998). Raw milk and fluid milk products. In

night), during which time a slight drop in pH oc-curs (ApH of 0.1 units). This process is calledprematuration, and often the milk is pasteurizedbefore the cheesemaking starter is added. Pre-maturation may suppress the growth of psychro-trophs during storage and/or produce com-pounds that stimulate the growth of thecheesemaking starter. In addition, the drop inpH stimulates rennet action. Alternatively, aproportion of the milk is inoculated with cultureand grown until a pH of 4.6 is attained, and asufficient volume is mixed with fresh bulk milkto reduce the pH of the fresh milk by 0.1 pHunits.

E.H. Marth & J.H. Steele (Eds.), Applied dairy microbiol-ogy. New York: Marcel Dekker.

Palmer, J. (No date). Hygienic milk production and equip-ment cleaning. Dublin: Teagasc.

Palmer, J. (1980). Contamination of milk from the milkingenvironment. International Dairy Federation Bulletin,120, 16-21.

Robinson, R.K. (Ed.). (1994). Modern dairy technology (2ded., Vol. 1). London: Chapman & Hall.

Figure 4-7 A cutaway of a Westfalia bacteria clarifier showing milk inlet (1), milk outlet (9), and bacteriastream (concentrate) outlet (12).

Page 68: Cheese Science

5.1 INTRODUCTION

In the manufacture of most cheeses, carefullyselected strains of different species of lactic acidbacteria (LAB) are added to the milk shortly be-fore renneting. Their major function is to pro-duce lactic acid and, in some cases, flavor com-pounds, particularly acetic acid, acetaldehyde,and diacetyl. Acid production, in turn, has threefunctions: it promotes rennet activity; aids theexpulsion of whey from the curd, thus reducingthe moisture content of the cheese; and helps toprevent the growth of undesirable bacteria in thecheese. These cultures are called starters be-cause they initiate (start) the production of acid.They are also called lactic cultures because theyproduce lactic acid. If raw milk is incubated at atemperature in the range 20-4O0C, it will coagu-late within 10-24 hr. This physical transforma-tion is due to the growth of and acid productionby adventitious LAB present in the raw milk. Inthe past, such coagulated milks were used asstarter cultures for cheese manufacture and areprobably the source of many starter strains in usetoday. Starter cultures are not used for manycheeses made in Mediterranean countries (e.g.,La Serena and artisanal production of Manchegocheese); instead, the cheesemaker relies on theadventitious LAB present in the milk to growduring cheesemaking and produce the necessaryacid.

5.2 TYPES OF CULTURES

Starter cultures are commonly divided intomesophilic cultures (with an optimum tempera-ture of about 3O0C) and thermophilic cultures(with an optimum temperature of about 420C).Each group of starters can be further subdividedinto defined- and mixed-strain cultures. De-fined-strain cultures are pure cultures, the physi-ological characteristics of which are known andidentifiable. Such cultures are used commer-cially in most large cheesemaking plants in Aus-tralia, New Zealand, the United States, theUnited Kingdom, and Ireland. They have beenisolated mainly from mixed-strain cultures (seebelow) but also from fermented products madeby indigenous LAB and plants. Before beingused commercially, the strains are screened forimportant technological properties, such as theirsalt tolerance and their ability to grow and pro-duce lactic acid in milk, resist attack by bacte-riophage, utilize citrate, and produce good-qual-ity cheese.

Mixed-strain cultures contain unknown num-bers of strains of the same species. In addition,many of them also often contain bacteria fromdifferent genera of LAB, including Lactococcusand Leuconostoc spp. in the case of mesophilic,mixed-strain cultures and Sc. thermophilus andLb. delbrueckii subsp. bulgaricus, Lb. del-brueckii subsp. lactis, and Lb. helveticus in the

Starter Cultures

CHAPTER 5

Page 69: Cheese Science

case of thermophilic, mixed-strain cultures.Many mixed cultures in use today are subcul-tures of coagulated milks that produced good-quality cheese at the turn of the century whencheese was beginning to be produced on a largescale. A further complication is that defined cul-tures are used rarely as pure cultures or singlestrains but as mixtures of 2 to 6 strains, whichmeans that they can also, with some degree ofjustification, be called (defined) mixed cultures.Mixed-strain cultures generally contain phage(see later) and are generally used by small-scaleproducers, while defined cultures are usuallyused by large-scale producers. Some of the im-portant distinguishing characteristics of the bac-teria found in defined and mixed cultures aresummarized in Table 5-1.

The main species of Lactococcus in mixed-strain mesophilic cultures is Lc. lactis subsp.cremoris, although Lc. lactis subsp. lactis is alsofound. These two subspecies are differentiatedfrom each other by their growth response at4O0C and the ability to produce NH3, ornithine,and citrulline from arginine (Table 5-1). In addi-tion, Lc. lactis subsp. lactis contains glutamatedecarboxylase, which produces y-aminobutyricacid from glutamate, while Lc. lactis subsp.cremoris does not (Nomura, Kimoto, Someya,& Suzuki, 1999). The exact species of Leu-conostoc found in starter cultures is not clear,but it is likely that both Ln. mesenteroides subsp.cremoris and Ln. lactis are present. The functionof the Leuconostoc spp. is to metabolize citrateto CO2, diacetyl, and acetate. CO2 is responsiblefor eye formation in Edam and Gouda cheeses,while diacetyl and acetate are important flavorcomponents of Quarg, Fromage frais, and Cot-tage cheese. For this reason, the citrate utilizersare often called aroma producers. Some of theLactococcus strains found in mixed cultures areable to utilize citrate (Cit+), and some are not(Qt-). CiIr lactococci dominate these culturesand generally make up about 90% of the organ-isms present, whereas the Cit+ lactococci andleuconostoc make up the rest. Depending on thearoma producers present in the culture, meso-philic, mixed-strain starters containing only Qt+

leuconostoc are called L cultures (L, the first let-ter of Leuconostoc), and those containing onlyCit+ lactococci are called D cultures (D, fromStreptococcus diacetilactis, the old name forCit+ lactococcus). Cultures containing both Qt+

lactococci and leuconostoc are called DL cul-tures, and those containing no aroma producer(i.e., only CiIr lactococci are present) are calledO cultures.

How do we know that mixed cultures containdifferent strains? This question is not easy to an-swer. Isolates from mixed cultures produce acidat widely different rates, e.g., pH values withinthe range 5-6.5 in milk after 6 hr of incubationare common (this may be related to lack of pro-teinase activity), have different phage-host pat-terns, and show several different plasmid pro-files. Recently, a detailed analysis of 113 isolatesfrom a mixed-strain DL culture (Flora Danica)was reported (Lodics & Steenson, 1990). This isa well-known commercial starter, commonlyused in the manufacture of soft cheeses. Seventyisolates were identified as Lc. lactis subsp.cremoris, 2 as CiIrLc. lactis subsp. lactis, 21 asCit+ Lc. lactis subsp. lactis, 18 as Ln. mesen-teroides subsp. cremoris, and 1 isolate could notbe classified. Twenty different plasmid profileswere found, examples of 17 of which are shownin Figure 5-1. Most of the Lc. lactis subsp.cremoris isolates (58) fell into 12 groups (lanes Ato L inclusive), while the 25 isolates of Cit+ Lc.lactis subsp. lactis fell into 3 groups (lanes M, N,and O), and all of the Ln. mesenteroides subsp.cremoris isolates (18) belonged to the same plas-mid group (lane Q). Twenty strains coagulatedreconstituted skim milk (11% solids) in less than8 hr at 210C, 2 strains in 18 to 24 hr, and theremainder in more than 24 hr. Seven phages wereisolated from the starter, and 51 isolates were in-sensitive to these phages. The remaining isolatesshowed various sensitivities to the 7 phages iso-lated from the culture itself and to 10 relativelywell-known phages. These data support the viewthat genuinely different strains are present inmixed cultures.

Thermophilic cultures almost always consistof two organisms, Streptococcus thermophilus

Page 70: Cheese Science

Table 5-1 Some Distinguishing Characteristics of the Lactic Acid Bacteria Found in Commercial and Natural or Artisanal Starter Cultures

Growth at

450C4O0C150C1O0CNH3 fromArginine

Metabolismof Citrate

Percentage ofLactic Acid

Produced in Milk*3ShapeType3Name

+++

+

+

+

+++

+++

+

+

+

+

+++

+

+++

++

+

++++++

+

+++

+

+

++

±

+

+

+++

+

±++

±

+

+

0.62.01.8

1.8

0.8

0.8<0.5

0.2

CoccusRodRod

Rod

Coccus

CoccusCoccusCoccus

RodRod

Rod

RodRodRodRod

CoccusCoccus

TTT

T

M

MMM

Commercial

Streptococcus thermophilusLactobacillus helveticusLactobacillus delbrueckiisubsp.

bulgaricusLactobaccilus delbrueckiisubsp.

lactisLactococcus lactis subsp.

cremorisLactococcus lactis subsp. lactisLeuconostoc lactisLeuconostoc mesenteroides

subsp. cremoris

Natural or artisanal, above plus

Lactobacillus case/ subsp. easelLactobacillus paracasei subsp.

paracaseiLactobacillus paracasei subsp.

toleransLactobacillus rhamnosusLactobacillus plantarumLactobacillus curvatusLactobacillus fermentumEnterococcus faecalisEnterococcus faecium

Page 71: Cheese Science

Fermentation ofd

MtIRafManRhmLacGaIGIuAllosteric LactateDehydrogenase0

lsomer ofLactate

Fermentationof SugarName

++

++

±

+

+

+

++

+

±

++

++

++

+

+

++

+

++++

++

+

++±+++

++

+++±

++

+

++++++

+ -++

+

++++

++

+

++++++

++

++

+

+

+

LDLD

D

LLDD

LL

L

LDLDLDLLL

HomoHomoHomo

Homo

HomoHomoHeteroHetero

HomoHomo

Homo

HomoHomoHomoHeteroHomoHomo

CommercialStreptococcus thermophilusLactobacillus helveticusLactobacillus delbrueckii subsp.

bulgaricusLactobaccilus delbrueckii subsp.

lactisLactococcus lactis subsp. cremorisLactococcus lactis subsp. lactisLeuconostoc lactisLeuconostoc mesenteroides subsp.

cremoris

Natural or artisanal, above plusLactobacillus easel subsp. easelLactobacillus paracasei subsp.

paracaseiLactobacillus paracasei subsp.

tolerans?Lactobacillus rhamnosusLactobacillus plantarumLactobacillus curvatusLactobacillus fermentumEnterococcus faecalisEnterococcus faecium

3 T = thermophilic; M = mesophilic.b Approximate values; individual strains vary.c Activated by fructose-1 ,6-phosphate and in lactobacilli also by Mn2+.d GIu = glucose; GaI = galactose; Lac = lactose; Rhm = rhamnose; Man = mannose; Raf = raffinose; MtI = manitol.6 Survives heating at 720C for 40 s.

Page 72: Cheese Science

and either Lactobacillus helveticus, Lb. del-brueckii subsp. lactis, or Lb. delbrueckii subsp.bulgaricus. These are often referred to as thecoccus and rod, respectively. Bulk starters of therod and coccus are generally grown individuallyfor cheese manufacture, but they are grown to-gether for yogurt production. In the latter case,they are genuinely mixed cultures, but defmed-strain thermophilic cultures are also used. Likemesophilic mixed cultures, thermophilic mixedcultures may contain several strains of each spe-cies. For some products (e.g., Mozzarellacheese), the rod:coccus ratio is important, and itis much easier to control this ratio by growingthe cultures separately.

The rod and coccus produce much more acidwhen they are grown together than when theyare grown separately (Figure 5-2). Acid produc-tion is an excellent indicator of the growth ofLAB (see Section 5.10). The improved growth is

due to the production of amino acids, particu-larly leucine, isoleucine, and valine, from caseinin the milk by the proteolytic system of the Lac-tobacillus, which stimulates the growth of Sc.thermophilus. The Streptococcus, in turn, pro-duces small amounts of CO2 and formic acidfrom lactose, which stimulate the growth of theLactobacillus. Thus, the relationship is genu-inely symbiotic. The CO2 produced is not suffi-cient to be apparent in the milk. However, thegrowth of the rod and coccus are more difficultto control when they are grown together.

Whether symbiosis occurs between the Leu-conostoc and the Lactococcus in mesophilic cul-tures is not clear; there is some evidence in theold literature that a certain amount of symbiosisoccurs but the exact nature of the interaction hasnot been determined. A certain amount of sym-biosis must occur, since many lactococci inmixed cultures do not have a proteinase system

Figure 5-1 Plasmid profiles of representatives of different plasmid groups (lanes A to Q) isolated from a FloraDanica culture. Lane S, plasmid standard containing Salmonella typhimurium LT2 (60 MDa), Escherichia collV517 (32, 5.2, 3.5, 3.0, 2.2, 1.7, 1.5. 1.2 MDa), pSA (23 MDa), and pSA3 (6.8 MDa) DNA.

A B C D E F G H I J K L M N O P Q S

Page 73: Cheese Science

Time, h

Figure 5-2 Acid production by Lactobacillus hel-veticus 243, Streptococcus thermophilus 302, or theircombination in milk at 420C,

(Prtr) and depend on the Prt+ strains to producethe amino acids they require for growth (dis-cussed later). Examples of the cultures used inthe production of different cheeses are shown inTable 5-2.

5.2.1 Artisanal (Natural) Cultures

In many countries, especially Italy, France,Switzerland, and Greece, other types of mixedcultures, called artisanal cultures, are used.These are derived mainly from the practice of"back-slopping," where some of the previousbatch of cheese is used as the inoculum for thenew batch. An example of this is the Greekcheese Kopanisti. More commonly, whey andmilk from today's cheesemaking are incubatedat a high temperature (45-520C) for an extendedperiod (4.5-18 hr), depending on the culture, foruse in tomorrow's cheesemaking. The cheeses inwhich these cultures are used are made from rawmilk. Such cultures depend on the presence ofLAB in the raw milk, and the cultures are calledwhey cultures or natural milk cultures. The tem-

perature of incubation and the pH exert selectivepressure on the types of bacteria that grow underthese conditions. The composition of these cul-tures is extremely complex and very variableand can include Lb. delbrueckii subsp. lactis, Lb.delbrueckii subsp. bulgaricus, Lb. helveticus,Lb. plantarum, Lb. casei, Lb. paracasei, Lc.lactis, Sc. thermophilus, Enterococcus faecalis,Ec. faecium, and Leuconostoc spp. (see Table5-1). These cultures have many of the attributesof mixtures of both mesophilic and thermophiliccultures, and many of the bacteria in them arelysogenic (see Section 5.7.2) and, as such, are apotential source of phage in factories; however,they can be very inconsistent in performance.

5.2.2 Adjunct Cultures

Much hard cheese made commercially todayis thought to lack flavor. The probable reasonsfor this are very low bacterial numbers in theraw milk and, more importantly, vastly im-proved hygiene in cheese factories. Because ofthis, various methods for improving flavor havebeen developed. Traditionally, only mesophiliccultures were used in the production of Ched-dar and other low-cooked cheeses. However, inrecent years, thermophilic starters, particularlySc. thermophilus and Lb. helveticus, have beenincluded in the starter and are thought to im-prove the flavor of the cheese. These organismsare very resistant to the cook temperature(~38°C) used for those cheeses. However, theirgrowth in such cheeses is limited to tempera-tures above 250C. At temperatures below this,little growth and consequently little acid pro-duction occurs.

Carefully selected strains of mesophilic lacto-bacilli, particularly Lb. paracasei and Lb. casei,are also used to improve the flavor of somecheeses, especially Cheddar. The basis for theiruse is the fact that large numbers of these bacte-ria (108 cfu/g) are found in ripened cheese, and itis assumed they must have some role. However,despite extensive research over several decades,the role of these bacteria in cheese flavor devel-opment is still unclear (see Chapter 10).

Lac

tic a

cid,

g/lO

Om

l

Page 74: Cheese Science

Cheese

Emmental cheese

Mozzarella and otherItalian cheese

Cheddar cheese

Edam and Goudacheese

Camembert and Briecheese

Tilset, Limburger, andMunster cheese

Yogurt

Fromage frais andQuarg

Starter Cultures

Sc. thermophilus and Lb.helveticus', galactose-positive Lb. delbrueckiisubsp. lactis may alsobe used

Sc. thermophilus and Lb.helveticus or a mixed-strain thermophilicculture

Defined strains of Lc.lactis subsp. lactis or O,L, or DL mesophilicmixed cultures;sometimes thermophiliccultures are included

Mainly DL mesophilicmixed cultures

O, L, or DL mesophilicmixed cultures

O, L, or DL mesophilicmixed cultures

Mainly thermophilic mixedcultures; defined strainsof Sc. thermophilus, Lb.delbrueckii subsp.bulgaricus, and Lb.delbrueckii subsp. lactismay also be used

O, L, or DL mesophilicmixed cultures

Other Cultures

Propionibacteriumfreudenreichii

Penicillium camemberti,Geotrichum candidum,Candida utilis

Brevibacterium linens,Geotrichum candidum,Candida utilis

Important ProductsOther Than Lactic

Acid

CO2, propionate, andacetate

CO2 and acetate

Sulphur compounds(e.g., methional)

Acetaldehyde

Diacetyl and acetate

5.2.3 Effect of Temperature on the Growthof Cultures

Temperature has a major effect on the growthof starter bacteria. In Figure 5-3, two different

measurements, the generation time and the de-crease in pH after 5.5 hr, are used as indicators ofgrowth. The latter is quite acceptable, as the de-crease in pH is a measure of the amount of acidproduced by the culture, and acid production is

Table 5-2 Starter Cultures Used in the Manufacture of Different Cheeses

Page 75: Cheese Science

<!

t^

Ow

^

- ^

3

O*

Q-

ga

ri

O'

^^

se

cD

E?

Ig-g

-ar* I

a 8

~ I

d'

< S?-

IT s

r S £

:«•*£

§ S

J 8

! H

S

§

2

~-S

Sf^ -

-I Ii

»8-1

1 S

T 3

8 I

8 §

5" S

" 3

g

-P"

g.

w |.

e

§. M

f?

^ 8

g-

«-f

Co

1- o

O

l •

°

ll|o

P|^P

8 S i

5 6

11 I' I'

11-1 i

Pf I

.-1

o:l !

Hf^

B-

P^

l^o

si?

!^

lle^

flS

P ^

^ ?

^?

?• ? cheese, whereas the growth of most strains of

Lc. lactis subsp. cremoris is markedly inhibitedat this temperature. A common cooking tem-perature for Swiss-type cheese is 540C, andstrains of Sc. thermophilus and ZZ?. helveticusproduce little acid at this temperature (Figure5-3). However, they can withstand this tempera-ture and begin to grow again when the tempera-ture decreases to 470C and below.

Figure 5-3 Effect of temperature on the growth of (A) two strains of Sc. thermophilus, (B) two strains of Lb.helveticus in milk, (C) Lc. lactis subsp. lactis 10 and Lc. lactis subsp. cremoris 3 and 9, and (D) three strains of Ln.mesenteroides subsp. cremoris in complex medium. Note differences in j; axes.

C D

Temperature, 0CTemperature, 0C

Gen

erat

ion

time,

h

Gen

erat

ion

time,

h

Temperature, 0C Temperature, 0C

pH a

t 5 h

pH a

t 5

h

BA

Page 76: Cheese Science

5.3 TAXONOMY

There are 12 genera of lactic acid bacteria:Aerococcus, Alloiococcus, Carnobacterium,Enterococcus, Lactobacillus, Lactococcus, Leu-conostoc, Pediococcus Streptococcus, Tetra-genococcus, Vagococcus, and Weissella. Onlyfive genera, however, contain organisms used ascheese starter cultures: Lactococcus, Enterococ-cus, Leuconostoc, Streptococcus, and Lactoba-cillus. All the lactic acid bacteria used in startercultures are Gram-positive, catalase-negative,nonmotile, non-spore-forming bacteria.

The taxonomy of lactic acid bacteria, includ-ing those used in starter cultures, has gonethrough several major revisions during the past30 years as a result of more detailed and sophis-ticated analyses. These include DNA:DNA andDNA:RNA hybridizations, comparative oligo-nucleotide cataloging and sequencing of the!6SrDNA gene, and serological studies with super-oxide dismutase. The current names are used inTable 5-1. The lactococci have probably gonethrough the greatest number of changes; theirtaxonomy is reviewed by Schleifer and Kilpper-BaIz (1987).

5.3.1 Lactococcus

Lc. lactis was isolated from soured milk in1873 by Lister and named Bacterium lactis(Latin for bacterium of milk). It was the firstbacterium isolated in pure culture. In 1909, itwas renamed Streptococcus lactis by Lohnis andplaced in the genus Streptococcus. In the 189Os,Storch isolated a very similar organism fromcream, which OrIa-Jensen called Sc. cremoris(of cream). In 1937, Sherman divided the genusStreptococcus into four groups—pyogenic, lac-tic, faecal and viridans—and placed Sc. lactisand Sc. cremoris in the lactic group. The groupswere relatively easily distinguished from eachother on the basis of growth under different con-ditions (Table 5-3). Serology (Lancefieldgrouping), based on the presence of differentcarbohydrates in the cell wall, was also used atthat time to separate the streptococci into differ-

ent groups; for example, the lactic group reactedonly with Group N antiserum and the faecalgroup only with Group D antisera. However,later studies showed that serology was not espe-cially useful, since many streptococci isolatedsubsequently did not possess group-specific an-tigens, and the group D antigen was also presentin organisms from the viridans group (e.g., Sc.bovis).

Homology studies of the DNA of Sc. lactisand Sc. cremoris have shown that these two or-ganisms are closely related to each other, and in1982 they were reclassified as subspecies of Sc.lactis and named Sc. lactis subsp. lactis and Sc.lactis subsp. cremoris, respectively (Garvie &Farrow, 1982). In 1985, it was realized that theseorganisms were only distantly related to the ge-nus Streptococcus, and therefore they weretransferred to a new genus, Lactococcus, as Lc.lactis subsp. lactis and Lc. lactis subsp.cremoris, respectively (Schleifer et al., 1985).The genus also includes four new species, Lc.raffinolactis, Lc. garvieae, Lc. plantarum, andLc. piscium, which were isolated from spontane-ously soured raw milk, a cow suffering frommastitis, frozen peas, and fish, respectively, anda subspecies, Lc. lactis subsp. hordniae, isolatedfrom the leaf hopper insect. None of the newspecies grows well in milk and hence they are oflittle value as starter cultures.

In 1936, an organism similar to Sc. lactis wasisolated from fermenting potatoes and named Sc.diacetilactis. The latter organism was laterrenamed Sc. lactis subsp. diacetylactis. Diaceti-lactis was the spelling used when the organismwas considered to be a species, but the spellingchanged to diacetylactis when the organism wasgiven subspecies status. It differed from Sc.lactis only in its ability to metabolize citrate, andsince the transport of citrate is plasmid encoded,it was renamed Lc. lactis subsp. lactis biovar.diacetylactis. Since biovar is not an acceptedtaxonomic epithet, this organism is now calledcitrate-utilizing (Qt+) Lc. lactis subsp. lactis orQt+ Lactococcus to distinguish it from the vastmajority of lactococci unable to metabolize cit-rate (Cit).

Page 77: Cheese Science

Lc. lactis subsp. lactis and Lc. lactis subsp.cremoris are the main species isolated from me-sophilic starters and raw milk soured at 18-3O0C, and it is generally believed that Lc. lactissubsp. cremoris gives a better flavored cheesethan Lc. lactis subsp. lactis. Lc. lactis subsp.lactis is able to grow at 4O0C and, in the presenceof 4% NaCl, produce NH3 from arginine and fer-ment maltose, whereas Lc. lactis subsp.cremoris cannot (see Table 5-1). However, thelatter organism can grow in the presence of 2%NaCl. The concentration of salt in cheese variesfrom 4% to 6% in the moisture of the cheese andis therefore inhibitory to most strains of lacto-cocci. Lc. lactis subsp. hordniae is similar to Lc.lactis subsp. cremoris, except that it does notferment lactose, galactose, maltose, or ribose.Lc. rqffinolactis is also like Lc. lactis subsp.lactis but ferments melibiose and raffinose anddoes not grow at 4O0C or produce NH3 fromarginine. Lactococci ferment sugars by the gly-colytic pathway to L lactate.

Lactococci are spherical or ovoid cells thatoccur singly, in pairs, or in chains elongated inthe direction of the chain, which can sometimescause them to be misidentified as Leuconostocspp. They grow at 1O0C but not at 450C or in thepresence of 6.5% NaCl. They are nonmotile.Motile strains of Lc. lactis that have the Group Nantigen have recently been transferred to the ge-nus Vagococcus.

5.3.2 Enterococcus

There is considerable debate as to whether en-terococci should be considered as starter cul-

tures, since the main source of many of them isfecal material. Because of this, they are oftenused as indicators of fecal pollution. Occasion-ally, they cause endocarditis and urinary tractinfections. Some of them, especially Ec.faecalis, are promiscuous and can easily pick upantibiotic resistance genes, especially for vanco-mycin, from plasmids or transposons. They areincluded here because they are common in arti-sanal cultures and are considered to impart desir-able flavors to cheeses made with these cultures.Enterococci occur in pairs or chains and are saltand heat tolerant and generally grow in the pres-ence of 6.5% NaCl and at 450C. These proper-ties make them ideal starters for cheesemaking,but their use has been questioned because theyare used as indicators of fecal contamination offood. Unlike lactococci, enterococci are notkilled by pasteurization to any great extent andcan be found in large numbers in many cheeses.

In 1984, DNA hybridization studies showedthat Sc. faecalis and Sc. faecium, which werethen classified in the faecal group of Streptococ-cus, were only distantly related to the strepto-cocci, and they were transferred to the new ge-nus, Enterococcus, as Ec. faecalis and Ec.faecium, respectively (for a review, see Schleifer& Kilpper-Balz, 1987). Since then, 17 specieshave been added, and these are divided into indi-vidual species and 3 "species" groups, based onthe sequences of the 16S rRNA. Group 1 con-tains Ec. durans, Ec. faecium, Ec. mundtii, andEc. hirae (98.7-99.7% similarity); group 2 con-tains Ec. avium, Ec. raffinosus, Ec. malodoratus,and Ec. pseudoavium (99.3-99.7% similarity);and group 3 contains Ec. gallinarum and Ec.

Table 5-3 Some Distinguishing Characteristics of the Different Groups of Streptococci

Growth at or in the Presence of

Group

PyogenicViridansLacticFaecal

1O0C

++

45 0C

+

+

pH9.6

+

6. 5% NaCl

+

0.1%MethyleneBlue

++

Page 78: Cheese Science

casseliflavus (99.8% similarity). Ec. dispar, Ec.saccharolyticus, Ec. sulfureus, Ec. columbae,Ec. cecorum, and Ec. faecalis form individuallines of descent. Not all enterococci are of fecalorigin. For example, Ec. mundtii, Ec. sulfureus,and Ec. casseliflavus have been isolated fromplants; Ec. malodoratus from Gouda cheese (inwhich it caused off-flavor development); Ec.durans from milk and meat; Ec. pseudoaviumfrom a cow with mastitis but not from bovine fe-ces; and Ec. raffinosus from a clinical source (itshabitat is not known). Therefore, their useful-ness as indicators of fecal pollution is question-able.

Unfortunately, there are no simple biochemi-cal or physiological tests that will categoricallyseparate Lactococcus from Enterococcus. Thiscan be done only by sophisticated techniques,such as protein profiling of extracts from wholecells by SDS-PAGE or genus-specific DNAprobes. Ec. faecalis and Ec faecium are com-monly found in the feces of humans and animals.These two species are easily separated fromLactococcus by their ability to grow at pH 9.6, at1O0C and at 450C, and in the presence of 6.5%NaCl. The more recently recognized Enterococ-cus species do not give positive results withsome of these tests. For example, Ec. dispar andEc. sulfureus do not grow at 450C, and Ec.cecorum and Ec. columbae do not grow at 1O0Cor in the presence of 6.5% salt and therefore maybe confused with Lactococcus spp. In addition, afew, mainly from humans, isolates of Lacto-coccus grow at 450C and in the presence of 6.5%NaCl. However, the likelihood of isolating thesespecies from starters and cheese is small. En-terococci ferment sugars by the glycolytic path-way to L lactate.

In the 193Os, Lancefield introduced a method,based on the antigenic structure of the cell wall,to help in identifying what were then Streptococ-cus spp. Prior to the division of Streptococcusinto Lactococcus, Enterococcus, and Strepto-coccus, the terms fecal streptococci, Group Dstreptococci, and enterococci were used inter-changeably. Group D species are found in bothEnterococcus (e.g., Ec. faecalis, Ec. faecium,

Ec. durans) and Streptococcus (e.g., Sc. bovisand Sc. equinus), and therefore the Group D de-scriptor is illogical and the term fecal strepto-cocci is now defunct. Lactococcus spp. reactswith Group N antiserum. The newer species ofEnterococcus, e.g., Ec. dispar, Ec. cecorum, andEc. columbae, do not have a Lancefield antigen.Based on these findings, Lancefield groupingsare probably of little relevance today.

5.3.3 Streptococcus

These are also spherical to ovoid cells, ar-ranged in pairs or chains. Currently, 39 speciesof Streptococcus are recognized but only one ofthem, Sc. thermophilus, is used as a starter cul-ture. It grows at 450C but not at 1O0C and in thepresence of 2.5% but not 4% NaCl, and it wasincluded in the viridans group by Sherman(1937). It is closely related to Sc. salivarius, aninhabitant of the mouth. A few years ago, it wasrenamed Sc. salivarius subsp. thermophilus, butit has recently been restored to full species sta-tus. It ferments sugars by the glycolytic pathwayto L lactate and does not have a Lancefield anti-gen.

5.3.4 Leuconostoc

These are spherical or lenticular cells that oc-cur in pairs and chains and are commonly foundin mesophilic cultures. Therefore, they can beconfused with lactococci and (heteroferment-ative) lactobacilli. They differ from lactococci inthree fundamental respects:

1. They ferment sugars heterofermenta-tively rather than homofermentatively,producing equimolar amounts of lactate,ethanol, and CO2. Small amounts of ac-etate may also be produced.

2. They produce the D rather than the L iso-mer of lactate.

3. With the exception of Ln. lactis, theyshow no visual evidence of growth in lit-mus milk, unless yeast extract (0.3g/100ml) is added.

Page 79: Cheese Science

Currently, the following species of Leu-conostoc are recognized: Ln. lactis, Ln.citreum, Ln. pseudomes enter oides, Ln. argen-tinum, Ln. fallax, Ln. amelibiosum, Ln.gelidum, Ln. carnosum, Ln. mesenteroidessubsp. mesenteroides, Ln. mesenteroides subsp.dextranicum, and Ln. mesenteroides subsp.cremoris. Despite the fact that leuconostocswere first identified in mixed-strain starter cul-tures in 1919, the exact species found in startersis still not clear. In the interim, different specieshave been implicated, but they have never beenidentified taxonomically, as researchers weremore interested in how they behaved in milkfermentations than in their taxonomy. Ln.mesenteroides subsp. cremoris is certainly in-volved, since starter cultures are the onlyknown source of this microorganism. It is alsolikely that Ln. lactis is involved. Ln. mesen-teroides subsp. cremoris is unusual in that itferments only lactose and its component mono-saccharides, glucose, and galactose. Leuconos-toc spp. do not hydrolyze arginine (Table 5-1).

5.3.5 Lactobacillus

These are rod-shaped cells that may be longand slender or short and sometimes bent, andthey often occur in chains. Currently, about 64different species of Lactobacillus are recog-nized. The genus is divided into three groups:obligately homofermentative, facultatively het-erofermentative, and obligately heteroferment-ative, depending on whether they containaldolase and phosphoketolase. The obligateheterofermenters can be coccobacillary inshape and may be confused with Leuconostoc.

The obligate homofermenters (Group 1) con-tain aldolase but not phosphoketolase andhence cannot ferment pentoses or gluconate;they ferment hexoses exclusively by the glyco-lytic (homofermentative) pathway to DL, L,and/or D lactate. This group includes all thethermophilic lactobacilli found in starter cul-tures (Lb. helveticus, Lb. delbrueckii subsp.bulgaricus, and Lb. delbrueckii subsp. lactis).In addition, all strains of Lb. delbrueckii subsp.

bulgaricus and most strains of Lb. delbrueckiisubsp. lactis excrete galactose in proportion tothe amount of lactose taken up by the cell (seeSection 5.4.3). Lb. helveticus produces DL lac-tate while Lb. delbrueckii produces only the Disomer.

The facultative heterofermenters (Group 2)contain both aldolase and phosphoketolase, andtherefore they ferment hexoses homofermenta-tively to lactate and pentoses and gluconateheterofermentatively to lactate and acetate.Growth on glucose represses the formation ofphosphoketolase. This group includes severalof the lactobacilli found in artisanal culturesand mature cheeses (e.g., Lb. easel, Lb.paracasei, Lb. plantarum, and Lb. curvatus).These are generally referred to as the non-starter lactic acid bacteria and are also calledmesophilic lactobacilli. Nonstarter LAB countsof 107 to 109 cfu/g can be found in Cheddarcheese within 2 months of ripening (see Chap-ters 10 and 11).

The obligate heterofermenters (Group 3)possess phosphoketolase but not aldolase, andhence, like Leuconostoc, they ferment sugarsheterofermentatively to equimolar concentra-tions of lactate, ethanol, and CO2. Smallamounts of acetate may be produced also. Al-most invariably, members of this group pro-duce NH3 from arginine. The only Group 3 lac-tobacilli reported in cheese are Lb. brevis andLb. fermentum, and these are also considered tobe nonstarter bacteria.

Generally, Group 1 lactobacilli do not growat 150C but do grow at 450C, while those inGroups 2 and 3 grow at 150C but not at 450C.This is not an absolute rule but generally ap-plies to most of the lactobacilli found in startersand cheese. Lb. fermentum is an exception, as itis the only Group 3 Lactobacillus that grows at450C. The lactobacilli found in starter culturesare often called thermophilic because their opti-mum growth temperature is around 420C. Theyare not true thermophiles, as they do not growat 550C. They are, however, able to withstandthe cooking temperature (540C) used in Swisscheese manufacture.

Page 80: Cheese Science

5.3.6 Differentiation of Lactic Acid Bacteriain Starters

A relatively simple scheme involving only afew criteria including Gram reaction, catalase,shape, growth at different temperatures and inthe presence of different concentrations of NaCl,and the pathway by which sugar is fermentedcan be used to identify the LAB found in starters(Table 5-^). The scheme is not definitive butprovides good practical guidance.

During the past 20 years, several new tech-niques for identifying LAB have been devel-oped, including methods for analyzing theamino acid and menaquione content of the cellwall, PAGE of the whole cell proteins, 16SrDNA sequencing, and randomly amplifiedpolymorphic DNA (RAPD) techniques. Allthese techniques are very sophisticated and areeither too slow or require standardized condi-tions or elaborate equipment for routine use.Species-specific probes based on 16S rDNA se-quences and amplification by polymerase chainreaction (PCR) have been or are being devel-oped for all LAB. These methods are relativelysimple to use and will probably become routinetools for identification within a few years.

Recently, Taillez, Tremblay, Ehrlich, andChopin (1998) used a RAPD technique withthree different primers to study the relationshipsbetween 113 strains of lactococci from a culturecollection. The strains could be divided into twogroups, Gl and G2 plus G3, based on the ge-nomic analysis, and they showed excellentagreement with the phenotypic analysis. GroupsGl and G3 contained only the lactis subspecies,whereas group G2 contained almost exclusivelydairy strains of the cremoris subspecies. It wassuggested that Group G2 evolved from the G3strains through the loss of specific characteris-tics in the dairy environment.

Scanning electron micrographs of some spe-cies of LAB found in starter cultures are shownin Figure 5—4. The considerable variation thatoccurs in the shape of these bacteria is evident.A microcolony of Lb. helveticus in Grana cheeseis also included.

5.3.7 Phylogeny

There is sound scientific evidence that all liv-ing organisms, including animals, plants, andmicroorganisms, evolved from a common an-cestor. The study of the evolutionary history ofbacteria is called phylogeny and it has receivedmuch attention during the past 20 years in at-tempts to understand the relationships betweendifferent microorganisms. To study phylogeny,one selects and compares the sequence of a mac-romolecule that is present and has the samefunction in all cells. Such molecules have beencalled evolutionary chronometers. The 16SrDNA gene is probably most widely used forthis purpose, because it is a relatively large mol-ecule (~ 15,000 nucleotides) and some se-quences are highly conserved while others arevariable. A comparison of the sequences has al-lowed scientists to construct the universal tree oflife. Bacteria form one of the 3 major domains(Archea and Eucarya are the others), and withinthe bacterial domain there are at least 12 distinctphylogenetic lineages, of which the Gram-posi-tive bacteria are one. Gram-negative bacteria donot form a single group. The Gram-positive bac-teria are divided into two branches, the clos-tridial branch, with a % mol guanine + cytosine(GC) content of less than 50 and the actino-mycete branch, with a % mol GC greater than55. All starter bacteria and the bacteria found inripening cheese are Gram-positive. Sc. ther-mophilus and the species of starter LAB found inthe genera Lactococcus, Lactobacillus, Enter o-coccus, and Leuconosoc belong to the clostridialsubdivision, whereas other bacteria commonlyfound in ripening cheese (Micrococcus, Coryne-bacterium, Brevibacterium, and Arthrobacter)belong to the actinomycete branch of the Gram-positive bacteria (see Chapter 10).

Computer analysis of the 16S rRNA mol-ecules has revealed short oligonucleotide se-quences, called signature sequences, that areunique to each bacterial species and that enablethe construction of species-specific nucleic acidprobes usable for identifying microorganisms.These complement the biochemical and physi-

Page 81: Cheese Science

Table 5-4 Criteria for Discriminating between the Different Groups of Lactic Acid Bacteria Found in Cultures

Growth atGrowth in NaCI

450C150CW0C65g/L40g/L20g/LFermentation

of GlucoseGram's StainCatalaseShape

+

++

+++

9

+

+

+

+

?

?

+

+/-

+/-

+/-

+

+/-

+

+

+

+

+

+

HomoHeteroHomo

HomoHomo

Hetero or homo

+++

+++

CocciCocciCocci

CocciRodsRods

LactococcusLeuconostooEnterococcus

Sc. thermophilusThermophilic (Group 1) LactobacillusMesophilic (Group 2 & 3) Lactobacillus

Page 82: Cheese Science

Figure 5-4 Scanning electron micrographs of some bacteria found in starter cultures. (A) Lactoba-cillus helveticus (13,00Ox); (B) Streptococcus thermophilus (9,00Ox), (C) Lactococcus lactis(13,00Ox); (D) a microcolony of Lactobacillus helveticus in Grana cheese (20,00Ox); (E) a mixedculture of Streptococcus thermophilus and Lactobacillus delbrueckii subsp. bulgaricus, used in yo-gurt manufacture (3,75Ox). The degree of magnification is shown in parenthesis.

E

C

D

B

A

Page 83: Cheese Science

ological tests normally used to identify bac-teria.

The phylogenetic relationships of starter andsome other LAB are outlined in Figure 5-5.Such analyses show interesting relationships.Both Ec.faecalis and Str. bovis are found in bo-vine feces and both react with Group D antigen,yet it is clear that Enterococcus is only distantlyrelated to Streptococcus and Lactococcus. Al-though not shown, Pediococcus spp. are foundin the Lb. casei/Lb. parcacasei group eventhough pediococci (tetrads) are morphologicallyquite distinct from lactobacilli (rods). In addi-tion, the heterofermentative thermophile, Lb.fermentum, is relatively closely related to thefacultatively heterofermentative mesophile, Lb.casei, and not to the other thermophilic lactoba-cilli. Recently, several heterofermentative lacto-bacilli, Lb. viridescens, Lb. confusus, and Lb.halotolerans, which are coccobacillary in shape,and the heterofermentative Leuc. parames-enteroides have been transferred to a new genuscalled Weisella. These considerations suggestthat neither shape, nor type of fermentation, norgrowth at 1O0C, 150C, or 450C give absolute in-formation on the relationship of LAB with eachother.

The 16S rDNA sequences have also shownthat Enterococci are more closely related toCarnobacterium and Vagococcus than to Lac-tococcus and Streptococcus.

5.4 METABOLISM OF STARTERS

5.4.1 Proteolysis

All LAB are auxotrophic and require severalamino acids and vitamins for growth. Specificstrains of LAB are still used to assay foods forvitamins (e.g., Lb. delbruckii ATCC 7830 for vi-tamin B12 and Ec.faecalis ATCC 8043 for folicacid). The requirements for amino acids arestrain specific and vary from as few as 4 to per-haps 12 or more. Glutamic acid, methionine, va-line, leucine, isoleucine, and histidine are re-quired by most lactococci, and many strainshave additional requirements for phenylalanine,

tyrosine, lysine, and alanine. The amino acid re-quirements of Sc. thermophilus andLeuconostocspp. are similar to those of the lactococci. Onlyone strain of Lb. helveticus has been studied; itrequired all the amino acids except glycine, ala-nine, serine, and cysteine. Lactococci possessmany of the genes of the amino acid biosyntheticpathways in their chromosomes, and the aminoacid requirements are probably the result of mi-nor deletions in the nucleotide sequences. This isprobably also true for the other starter LAB, butthe question has not been studied.

Fully grown milk cultures of starter bacteriacontain around 109cfu/ml. The concentrations ofamino acids and peptides in milk are low andsufficient to sustain only about 25% of the maxi-mum number of starter cells present in a fullygrown starter culture. Therefore, starter bacteriamust have a proteolytic system to hydrolyze themilk proteins to the amino acids required forgood growth in milk. The proteolytic system ofLactococcus involves a cell wall-associatedproteinase, amino acid and peptide transportsystems, and peptidases (for a review see Kunji,Mierau, Hagting, Poolman, & Konings, 1996).The generally accepted view of the system isshown in Figure 5-6.

The lactococcal proteinase (PrtP) is one of themost intensively studied enzymes of starter bac-teria, but it is likely that other starter bacteriahave similar systems. It is a serine proteinasethat is synthesized in the cell as a pre-pro-pro-teinase and that is transformed into the mature,active proteinase by a process not yet completelyunderstood. It is not a truly extracellular enzymebut is anchored to the cell membrane by its ex-tremely hydrophobic C-terminal region. Themature proteinase, called PrtP, contains about1,800 amino acid residues, has a molecular massof about 185 kDa, and has an optimum pH ofabout 6. The gene encoding the proteinase (prtP)in several strains of lactococci has been clonedand sequenced. The nucleotide sequences arevery similar (98% identical), implying that thereis only one proteinase, but the amino acid se-quences are sufficiently different to result in dif-ferent activities on the various caseins. Two dif-

Next Page

Page 84: Cheese Science

ological tests normally used to identify bac-teria.

The phylogenetic relationships of starter andsome other LAB are outlined in Figure 5-5.Such analyses show interesting relationships.Both Ec.faecalis and Str. bovis are found in bo-vine feces and both react with Group D antigen,yet it is clear that Enterococcus is only distantlyrelated to Streptococcus and Lactococcus. Al-though not shown, Pediococcus spp. are foundin the Lb. casei/Lb. parcacasei group eventhough pediococci (tetrads) are morphologicallyquite distinct from lactobacilli (rods). In addi-tion, the heterofermentative thermophile, Lb.fermentum, is relatively closely related to thefacultatively heterofermentative mesophile, Lb.casei, and not to the other thermophilic lactoba-cilli. Recently, several heterofermentative lacto-bacilli, Lb. viridescens, Lb. confusus, and Lb.halotolerans, which are coccobacillary in shape,and the heterofermentative Leuc. parames-enteroides have been transferred to a new genuscalled Weisella. These considerations suggestthat neither shape, nor type of fermentation, norgrowth at 1O0C, 150C, or 450C give absolute in-formation on the relationship of LAB with eachother.

The 16S rDNA sequences have also shownthat Enterococci are more closely related toCarnobacterium and Vagococcus than to Lac-tococcus and Streptococcus.

5.4 METABOLISM OF STARTERS

5.4.1 Proteolysis

All LAB are auxotrophic and require severalamino acids and vitamins for growth. Specificstrains of LAB are still used to assay foods forvitamins (e.g., Lb. delbruckii ATCC 7830 for vi-tamin B12 and Ec.faecalis ATCC 8043 for folicacid). The requirements for amino acids arestrain specific and vary from as few as 4 to per-haps 12 or more. Glutamic acid, methionine, va-line, leucine, isoleucine, and histidine are re-quired by most lactococci, and many strainshave additional requirements for phenylalanine,

tyrosine, lysine, and alanine. The amino acid re-quirements of Sc. thermophilus andLeuconostocspp. are similar to those of the lactococci. Onlyone strain of Lb. helveticus has been studied; itrequired all the amino acids except glycine, ala-nine, serine, and cysteine. Lactococci possessmany of the genes of the amino acid biosyntheticpathways in their chromosomes, and the aminoacid requirements are probably the result of mi-nor deletions in the nucleotide sequences. This isprobably also true for the other starter LAB, butthe question has not been studied.

Fully grown milk cultures of starter bacteriacontain around 109cfu/ml. The concentrations ofamino acids and peptides in milk are low andsufficient to sustain only about 25% of the maxi-mum number of starter cells present in a fullygrown starter culture. Therefore, starter bacteriamust have a proteolytic system to hydrolyze themilk proteins to the amino acids required forgood growth in milk. The proteolytic system ofLactococcus involves a cell wall-associatedproteinase, amino acid and peptide transportsystems, and peptidases (for a review see Kunji,Mierau, Hagting, Poolman, & Konings, 1996).The generally accepted view of the system isshown in Figure 5-6.

The lactococcal proteinase (PrtP) is one of themost intensively studied enzymes of starter bac-teria, but it is likely that other starter bacteriahave similar systems. It is a serine proteinasethat is synthesized in the cell as a pre-pro-pro-teinase and that is transformed into the mature,active proteinase by a process not yet completelyunderstood. It is not a truly extracellular enzymebut is anchored to the cell membrane by its ex-tremely hydrophobic C-terminal region. Themature proteinase, called PrtP, contains about1,800 amino acid residues, has a molecular massof about 185 kDa, and has an optimum pH ofabout 6. The gene encoding the proteinase (prtP)in several strains of lactococci has been clonedand sequenced. The nucleotide sequences arevery similar (98% identical), implying that thereis only one proteinase, but the amino acid se-quences are sufficiently different to result in dif-ferent activities on the various caseins. Two dif-

Previous Page

Page 85: Cheese Science

ferent specificities are found, one of which, PIproteinase, hydrolyzes principally (3-casein andto some extent K-casein, while the other, PIIIproteinase, acts efficiently on asi-, (3-, and K-caseins. Pi-type proteinases hydrolyze the C-ter-minal region of p-casein, which is quite hydro-phobic, producing bitter peptides that may beresponsible for the development of bitter-fla-vored cheese. One way of preventing this is to

use only strains that produce Pill-type protein-ase as starters.

Only limited information is available on theproteinases of the other starter bacteria, but theavailable data indicate that they are similar to thelactococcal proteinases (e.g., the PrtP of Lc.lactis and Lb. pararcasei are 95% similar). Un-usually, no proteolytic activity has been detectedin Sc. thermophilus except in the so-called Asian

Figure 5-5 Phylogenetic tree showing the relationships among some starter lactic acid bacteria. E, Enterococ-cus; L, Lactobacillus; Lc, Lactococcus; L, Leuconostoc; S, Streptococcus; O, Oenoccos; W, Weissella.

Page 86: Cheese Science

strains, which were isolated in outer Mongolia,India, and Japan. This lack of proteolytic activitymay explain the symbiotic relationship betweenLb. delbrueckii subsp. bulgaricus and Sc. ther-mophilus in yogurt cultures (Figure 5-2).

There are four different caseins in milk, asr,ocS2-, P"» and K-, which occur in the ratio ofroughly 4:1:3:1 and make up about 80% of thetotal protein in milk (see Chapter 3). Hydrolysisof these proteins by lactococcal proteinases re-sults in production of numerous oligopeptides ofdifferent sizes. For example, hydrolysis of (3-casein, which contains 209 amino acid residues,by the PI proteinase results in the production of100 peptides, the majority of which contain be-tween 4 and 30 amino acid residues. Peptidescontaining up to 8 amino acid residues can betransported across the cell membrane into thecell. In the lactococci, various transport systems,including an oligopeptide transport system, a di-and tripetide transport system, and at least 10amino acid transport systems, which have a highspecificity for structurally related amino acids(e.g., Glu/Gln, Leu/Ile/Val) have been de-

scribed. The driving forces for transport includethe proton motive force, antiport and symportsystems, and ATP hydrolysis (see Kunji et al.,1996, for a review).

Inside the cell, the peptides are hydrolyzed bypeptidases to the individual amino acids neces-sary for the synthesis of the proteins required forcell growth. Numerous peptidases have beenidentified in starter LAB, including at least threeaminopeptidases (Pep N, Pep A, and Pep C), twotripeptidases (Pep T and Pep 53), and two dipep-tidases (Pep V and Pep D) that release singleamino acids from the N-terminal end of the rel-evant substrates. In Lc. lactis subsp. cremoris,two different endopeptidases (Pep O and Pep F)have been identified that hydrolyze internal pep-tide bonds in peptides but not in the intactcaseins. The proline content of casein is quitehigh, and, because of its structure, specific pepti-dases, called prolidases (Pep Q), aminopepti-dase P (Pep P), X-prolyl-dipeptidyl aminopepti-dase (Pep X), prolinase (Pep R), and prolineiminopeptidase (Pep I), are required to hydro-lyze it from peptides. Some of the important

Figure 5-6 The proteolytic system of Lactococcus.

MilkCellwall

Cellmembrane Cytoplasm

Amino acids

Dipeptides

Tripeptides

Peptidases

Oligopeptides

Proteinase

Casein

Oligopeptides

Tripeptides

Dipeptides

Amino acids

Oligopeptidetransporter

Di-and tri-peptidetransporters

Amino acidtransporters

Page 87: Cheese Science

Table 5-5 Properties of the Various Peptidases Found in Starter Lactic Acid Bacteria

Location13Optimum pHType3MoI. Wt. (kDa)OrganismSubstraten= 1,2,3...NamePeptidase

II

CW, IIIII

CWI

I

III

CW5 IIIII

CWI

77

6.577

7.55.88

7.586

7-8

8

7-8.56.5-7

6.57.5

6.5

MMMTTTMMMMMT

MMM

SSS

SSS

959597505250

46-52>2350515054

424143

59-9082-9588-95

3530-50

3334

Lc. lactisLb. delbrueckiiLb. helveticusLc. lactisLb. delbrueckiiLb. helveticusLc. lactisLc. lactisLc. lactisLb. delbrueckiiLb. helveticusLb. helveticus

Lc. lactisLb. delbrueckiiLc. lactis

Lc. lactisLb. delbrueckiiLb. helveticusLb. helveticusLc. lactisLb. delbrueckiiLb. helveticus

X * (X)n

X * (X)n

X ^ X - X

X ^ X

X ^ X

X * Pro

X *> Pr0-(X)n

X Pro-^ (X)n

Pro J- XPro * X-(X)n

Pep N

Pep C

Pep TPep 53Pep V

Pep D

Pep Q

Pep P

Pep X

Pep RPep I

GeneralAminopeptidase N

Aminopeptidase C

Tripeptidase

Dipeptidase

Proline SpecificProlidase

Aminopeptidase PX-prolyl-dipeptidylaminopeptidase

ProlinaseProline iminopeptidase

Note: Many of the enzymes were isolated from several strains of the species, and the data presented are a summary.3 M = metalloenzyme; S = serine peptidase; T = thiol peptidase.b I = intracellular; CW = cell wall.

Page 88: Cheese Science

properties of the peptidases are summarized inTable 5-5. Generally, the peptidases were iso-lated from several strains of the same species,and the data shown are a summary. Pep R hasbeen found only in Lb. helevticus. No carboxy-peptidase has been found in LAB.

The peptidases are either serine-, metallo-, orthiol-enzymes and have pH optima in the range6.0 to 8.0. All of them are located inside the celland, acting together, they hydrolyze the peptidestransported into the cell by the oligopeptide andthe di- and tripeptide transport systems to theirconstituent amino acids for use in protein syn-thesis. They are also important in the ripening ofcheese. During ripening, the starter bacteriagradually die and lyse and release their intracel-lular peptidases, which then act on any peptidespresent around the cell. The amino acids pro-duced are considered to be the precursors of theflavor compounds necessary for the develop-ment of good-flavored cheese (see Chapter 11).

5.4.2 Arginine Metabolism

Many LAB produce NH3 from arginine(Table 5-1) by the arginine deiminase pathwayand simultaneously produce 1 mol of ATP permol of arginine metabolized (Figure 5-7). Argi-nine is first hydrolyzed to NH3 and citrulline byarginine deiminase. Ornithine carbamyltrans-ferase then catalyses the phosphorolysis of cit-rulline to ornithine and carbamyl phosphate; thelatter is then hydrolyzed to NH3 and CO2 by car-bamate kinase, with the concomitant productionof ATP. The uptake of arginine is driven by anantiport transport system in which ornithine isexchanged for arginine.

Lc. lactis subsp. lactis produces NH3 fromarginine via this pathway, whereas Lc. lactissubsp. cremoris does not, owing to the lack of atleast one of the three enzymes of the pathway.

5.4.3 Lactose Metabolism

Lactose, the principal sugar found in milk, is adisaccharide composed of one glucose and one

Figure 5-7 Arginine/ornithine antiport and the argi-nine deiminase pathway. Accumulation of ornithine(lysine) via the Dp-driven lysine transport system isalso shown. ADI, arginine deiminase; OCT, ornithinecarbamoyltransferase; CK, carbamate kinase.

galactose residue linked by a pi-4 bond (seeChapter 3). The major function of starter LAB incheesemaking is the production of lactic acidfrom the fermentation of lactose. Since starterbacteria do not contain a functional cytochromesystem, their metabolism of sugars is fermenta-tive rather than respirative, and ATP is producedby substrate-level phosphorylation rather thanby oxidative phosphorylation. Despite their lackof a cytochrome system, most LAB are aero-tolerant and grow quite well in air. They cannotgrow without a sugar that serves as an energysource. Fermentation of sugars occurs by theglycolytic (Figure 5-8) or phosphoketolase (PK)(Figure 5-9) pathways. Lactate is the end-prod-uct of glycolysis, while lactate, ethanol, and CO2

are the end products of the PK pathway.

Arginine Deiminase Pathway

Arginine

Ornithine

Carbamoylphosphate

OmithineLysine

Citrulline

Arginine

alkaline acid

Page 89: Cheese Science

Figure 5-8 Glycolytic pathway of lactose metabolism in lactic acid bacteria.

EXTERNALENVIRONMENT L-LACTATE

CYTOPLASM

LACTOSE LACTOSE EXTERNALENVIRONMENT

CYTOPLASMLACTOSELACTOSE-P

PEP/PTS permease

LELOIRPATHWAY

GALACTOSE

GALACTOSE-I-P

GLUCOSE-I-P

GLUCOSEATP

GALACTOSE-6-P GLUCOSE-6-P

FRUCTOSE-6-P

FRU CTOSE-1,6-biP

TAGATOSE-6-P

TAGATOSE-l,6-biP

DIHYDROXYACETONE-P GLYCERALDEHYDE-3-P

TAGATOSEPATHWAY

1,3-DIPHOSPHOGLYCERATE

3-PHOSPHOGLYCERATE

2-PHOSPHOGLYCERATE

PHOSPBOENOLPYRUVATE

PYRUVATE

L-LACTATE

MEMBRANE

MEMBRANE

NADH

NAD

ADP

ADP

ATP

ATP

NADH

Page 90: Cheese Science

Figure 5-9 Phosphoketolase pathway of lactose metabolism in lactic acid bacteria.

D-LACTATE

PYRUVATE

PHOSPHOENOLPYRUVATE

2-PHOSPHOGLYCERATE

3-PHOSPHOGLYCERATE

1,3-DIPHOSPHOGLYCERATE

GLYCERALDEHYDE-3-P

XYLULOSE-5-P

RIBULOSE-S-P

ACETYL-P

ACETYL-CoAACETATE

ACETALDEHYDE

ETHANOL

6-PHOSPHOGLUCONATE

GLUCOSE-6-PGLUCOSE-I-P

GALACTOSE-I-P

GLUCOSEGALACTOSELELOIR

PATHWAY

LACTOSE CYTOPLASM

LACTOSE EXTERNALENVIRONMENT

MEMBRANEpermease

Page 91: Cheese Science

Lactic acid contains an asymmetric carbonand hence can exist as D and L isomers:

COOH COOHI I

H-C-OH HO-C-HI I

CH3 CH3

D-Iactic acid L-lactic acid

The isomer of lactate produced is useful in theidentification of the various genera and speciesof starter LAB. For example, Leuconostoc andLb. delbrueckii produce only the D isomer,Lactococcus and Sc. thermophilus produce onlythe L form, and Lb. helveticus produces a mix-ture of the D and L isomers, owing to the pres-ence of two lactate dehydrogenases in the cell,one of which is specific for the L and the otherfor the D isomer.

Before lactose can be fermented, it must betransported into the cell. Starter LAB use twodifferent systems to transport lactose, the per-mease and the phosphotransferase (PTS), bothof which require energy. Energy for the per-mease system is derived from ATP, and lactoseis transported without being transformed. In thePTS system, the energy is derived from phos-phoenolpyruvate (PEP) in a complex series ofreactions involving two enzymes, EI and EIII,and a heat-stable protein, HPr. During transport,the high-energy phosphate in PEP is eventuallytransferred to lactose to form lactose-phosphaterather than lactose (Figure 5-10). Lactococci usethe PTS system to transport lactose, while allother starter LAB use the permease system.

The initial enzyme involved in the metabo-lism of lactose depends on the transport systemused. In the PTS system, lactose-P is formedduring transport and is hydrolyzed by phospho-(3-galactosidase (Ppgal) to glucose and galac-tose-6-P, while in the permease transport sys-tem, lactose is transported intact and ishydrolyzed by p-galactosidase (pgal) to glucoseand galactose (Table 5-6).

In the lactococci, the subsequent fermentationof glucose is via the glycolytic pathway, and ga-

lactose-6-P is metabolized through several taga-tose derivatives to glyceraldehyde-3-P anddihydroxy acetone phosphate (Figure 5-8).Tagatose is a stereoisomer of fructose:

CH2OH CH2OHI IC=O C=O

I IHO-C-H HO-C-H

I IHC-OH HO-CH

I IHC-OH HC-OH

I ICH2OH CH2OH

D-fructose D-tagatose

Fermentation of glucose and galactose by ther-mophilic starter LAB also occurs via glycolysisbut some of these bacteria, including Sc.thermophilus, Lb. delbrueckii subsp. bulgaricus,and some strains of Lb. delbrueckii subsp. lactis,ferment only the glucose moiety of lactose andexcrete galactose in proportion to the amount oflactose transported. Initially, this was thought tobe due to the lack of galactokinase, the first en-zyme of the Leloir pathway, but more recentstudies have suggested that galactose is excretedas an exchange molecule in the transport of lac-tose into the cell, in which process a single trans-porter, lactose permease, simultaneously trans-ports lactose into and galactose out of the cell.Leuconostoc spp. ferment the glucose and galac-tose by the phosphoketolase (PK) pathway; thegalactose is first transformed to glucose-1-P viathe Leloir pathway (Figure 5-9).

The products of both pathways are quite dif-ferent. In glycolysis, 1 mole of lactose is trans-formed to 4 moles of lactic acid (or 2 in the caseof Sc. thermophilus, Lb. delbrueckii subsp.bulgaricus, and the strains of Lb. delbrueckiisubsp. lactis, which excrete galactose), whereasin the PK pathway it is transformed to 2 moleseach of lactate, ethanol, and CO2.

The reason for the production of lactate inboth pathways and of ethanol in the PK pathwayis the need to reoxidize the NADH and NADPHproduced in the earlier steps of the pathway to

Page 92: Cheese Science

allow fermentation to continue. The purpose ofthe fermentations is to produce sufficient ATP tosustain growth. Production of ATP by fermenta-tion is much less efficient than the production ofATP by respiration (e.g., in glycolysis, 4 molesof ATP are produced per mole of lactose fer-mented, compared with a possible 76 moles byrespiration). Therefore, to produce the sameamount of ATP by fermentation as by respira-tion, a large amount of sugar must be fermented,and consequently large amounts of lactic acidare produced by LAB.

The growth of some strains of Lactococcus ongalactose or low levels of glucose leads to theproduction of other compounds, besides lactate,from pyruvate, such as ethanol, acetate, and ac-etaldehyde (Figure 5-11). In these bacteria, lac-tic dehydrogenase (LDH) and pyruvate-formatelyase (PFL) are allosteric enzymes. LDH re-quires the presence of fructose-1,6-bisphosphatefor activity, whereas pyruvate-formate lyase(PF) activity is inhibited by triose-phosphates.Normally, both effectors are present at high con-centrations in the cell, favoring LDH activityand lactate production. Growth at low sugar con-centrations results in lower intracellular concen-trations of the effectors, and therefore LDH ac-tivity is reduced and PFL activity increased,which allows pyruvate to be channelled to etha-

nol and acetate rather than lactate. Normally,during anaerobic growth, PFL activity is favoredover PDH activity in the initial formation ofacetyl CoA.

The end-products of lactose fermentation aremainly acidic and will, unless excreted, acidifythe cell cytoplasm. LAB have two mechanismsfor excreting lactate and protons. One involvesthe transmembrane reversible F0F1-ATPaSe andis responsible for the secretion of protons. Thesecond involves the simultaneous secretion oflactate anions and protons in symport with eachother. This mechanism occurs especially whenthe external concentration of lactate is low andthe internal concentration is high. Energy canalso be derived from this reaction through thecreation of a proton-motive force.

5.4.4 Citrate

Cit+ Lactococcus and Leuconostoc spp. pres-ent in mesophilic starters metabolize citrate toacetate, CO2, diacetyl, acetoin, and 2,3-bu-tanediol by the pathway shown in Figure5-12. Citrate is not used as an energy source butis co-metabolized with lactose or some othersugar. The organisms involved are Cit+ lacto-cocci, Ln. mesenteroides subsp. cremoris, andLn. lactis. Citrate is not metabolized by thermo-

Table 5-6 Salient Features of Lactose Metabolism in Starter Organisms

Organism

Lactococcus

Leuconostoc

Sc. thermophilus

Lb. delbrueckii

Lb. helveticus

Transport3

PEP-PTS

Permease

Permease

Permease

Permease

Pathway13

GLY

PK

GLY

GLY

GLY

CleavageEnzyme

Ppgal

pgal

Pgal

pgai

Pgal

Products(mol/mol lactose)

4 lactate

2 lactate + ethanol+ 2CO2

2 lactate0

2 lactate0

4 lactate

lsomer ofLactate

L

D

L

D

DL3PTS = phosphotransferase system.bGLY = glycolysis; PK = phosphoketolase.cThese species metabolize only the glucose moiety of lactose.

Page 93: Cheese Science

philic starters. The CO2 produced is responsiblefor the small eyes in Edam and Gouda cheese,while diacetyl and acetate are important con-tributors to the flavor of many fermented prod-ucts, including Quarg, Fromage frais, and Cot-tage cheese. Diacetyl is produced in only smallamounts (< 10 mg/ml or 0.11 mM in milk), andacetoin production is generally 10-50 timesgreater than that of diacetyl. One mole of acetateis produced from each mole of citrate used, butrecent studies suggest that about 1.2 moles ofacetate are produced per mole of citrate used.This is probably due to the production of smallamounts of acetate from sugar metabolism.There is very little information on the concentra-tion of 2,3-butanediol produced by starters.

There is still controversy on how diacetyl isactually produced. One of the reasons for this isthat the putative enzyme, diacetyl synthase, hasnever been categorically found in LAB. An-other, and probably more important, reason isthat acetolactate (AL) is very unstable andreadily autodecarboxylates nonoxidatively toacetoin or oxidatively to diacetyl. AL is so un-stable that it is available commercially only as adouble ester to protect it from autodecar-boxylation; it is normally hydrolyzed with 2equivalents of NaOH just before use. Acetoin isproduced by AL decarboxylase activity, but it isgenerally believed that diacetyl is producedchemically rather than enzymatically from AL.Despite this, it is difficult to see how diacetylcan be produced oxidatively from AL by startercultures, which essentially grow anaerobically

and produce very low Eh values (== -250 mV).Acetolactate is not normally found as an end-product unless the strain lacks acetolactate de-carboxylase activity.

Growth and metabolite production by a D anda DL mesophilic mixed-strain culture in milk at210C are shown in Figure 5-13. The cell numbersand the production of end-products are plottedsemi-logarithmically because bacteria grow andproduce end-products exponentially. Arithmeticplots are quite different. The graphs show ex-amples, and it is important to remember that indi-vidual cultures will exhibit some variation. Pro-duction of the various metabolites is in thefollowing order: lactate > acetate > acetoin »diacetyl. Citrate utilization is generally slower inL than in D or DL cultures because of the slowergrowth of the Cit+ Leuconostoc (compared withCit+ Lactococcus). Production of diacetyl andacetoin ceases as soon as all the citrate has beenused, after which the levels of acetoin and diacetylmay decrease due to the activity of acetoin dehy-drogenase. The same enzyme is probably respon-sible for the reduction of diacetyl to acetoin and ofacetoin to 2,3-butanediol. Therefore, to retain themaximum amount of diacetyl in unripenedcheese, the product should be cooled as soon aspossible after citrate utilization is complete.

Little, if any, acetolactate (AL) accumulates incultures, because most Ck+ lactococci contain anactive AL decarboxylase (ALD). However, theCit+ lactococci in the D culture in Figure 5-13contain an inactive ALD and hence accumulateAL. Nucleotide sequences of the aid gene in this

Figure 5-10 Phosphoenolpyruvate-phosphotransferase system for transport of sugar in Lactococcus lactis,

SUGAR

SUGAR-P

PYRUVATE

PEP

OUTIN

Page 94: Cheese Science

strain and in a strain containing ALD activityshowed only one substitution (cytosine for thym-ine at position 659), which resulted in a change ofhistidine to tyrosine in a motif conserved in allALDs. It is thought that this change is sufficientto lead to the synthesis of an inactive enzyme(Goupil, Cormier, Ehrlich, & Renault, 1996).Autodecarboxylation of AL to acetoin (mainly)and diacetyl probably occurs throughout growth,but is observed only as soon as AL productionceases (when all the citrate is used). The level ofdiacetyl produced from autodecarboxylation canbe increased by aeration at acid pH, and this prop-erty is used to produce diacetyl in the manufac-

ture of lactic butter by this particular culture.Cit~ lactococci are the dominant bacteria in

mesophilic mixed-strain cultures. The levels ofthe Cit+ lactococci and leuconostocs in these cul-tures vary but generally comprise less than 10%of the total number of bacteria present. It is obvi-ous from Figure 5-13 that lactic acid productionfollows the increase in cell numbers very closely.This explains why the amount of lactic acid pro-duced is a good indicator of the growth of LAB.As lactic acid is produced, the pH decreases, sopH is also a good indicator of growth. However,the relationship between the increase in acid pro-duction (or cell numbers) and pH is not linear.

Figure 5-11 Pathways of pyruvate metabolism in lactic acid bacteria.

ETHANOL ACETATE

ACETALDEHYDE ACETYL-P

FORMATE

ACETYL-CoA

L-LACTATE

ACETOIN

a ACETOLACTATE PYRUVATE

LACTOSE

Page 95: Cheese Science

Figure 5-12 Citric acid metabolism in lactic acid bacteria.

2,3-BUTANEDIOL

DIACETYL

ACETATE

ACETYLPHOSPHATE ACETOIN

ct-ACETOLACTATE

ACETYL-COA

ACETALDEHYDE-TPP

PYRUVATELACTATE

OXALACETATE LACTOSE

CITRIC ACID

ACETATE

Page 96: Cheese Science

Figure 5-13 Growth and metabolite production by a D (4/25), a DL (1362), and an L (FrS) culture in sterile 10% (w/v) reconstituted skim milk at 210C. TheDL and L cultures are commercial cheese cultures; the D culture is used in the production of lactic butter.

Time, h

Culture 1362

Time, h

Time, h

Culture 4/25

Time, h

Time, h

Time, h

Page 97: Cheese Science

Nevertheless, pH is easy to measure and is themost widely used indicator of starter growth inthe cheese industry.

Pure cultures of Ck+ Lactococcus and Leu-conostoc differ in the products produced duringco-metabolism of citrate and lactose. The formerproduce lactate, acetoin, and CO2 in a mannersimilar to that of mixed-strain mesophilic cul-tures. In contrast, Cit+ Leuconostoc spp. produceno acetoin or diacetyl. Instead, they produce in-creased amounts of lactate and acetate. Pyruvateis an intermediate in the metabolism of both lac-tose and citrate. In leuconostocs, the pyruvateproduced from the metabolism of both sub-strates is converted to lactic acid. This relievesthe cells from forming ethanol to regenerate thepyridine nucleotides. Instead, the acetyl-P pro-duced from lactose is used to produce acetateand ATP:

Acetyl-P + ADP -> Acetate + ATP

This extra production of ATP also results inmuch faster growth of the organism. However,leuconostocs will produce diacetyl and acetoinfrom citrate in the absence of an energy source,and proportionately more of the citrate is con-verted to these compounds as the pH decreases.The question then may be asked, how do leu-conostocs produce diacetyl and acetoin in mixedcultures? The answer is not clear, but it is prob-ably due to the fact that leuconostocs cannot takeup much lactose at pH values below 5.5.

5.4.5 Acetaldehyde

Acetaldehyde is produced by both mesophilicand thermophilic cultures and is an importantcomponent of the flavor of yogurt. The amountproduced varies but can reach 30 mg/ml. It isgenerally considered to be a product of carbohy-drate (pyruvate) metabolism (Figure 5-12), butit can also be produced from threonine via threo-nine aldolase activity:

Threonine —» glycine + acetaldehyde

This is the mechanism by which acetaldehydeis produced by lactococci. The only strain ofLactococcus that has a requirement for glycine

also lacks threonine aldolase, implying that thephysiological function of this reaction is the pro-vision of glycine for growth. It is not clear howacetaldehyde is produced by thermophilic cul-tures.

Both types of bacteria present in yogurt cul-tures produce acetaldehyde, but Sc. thermo-philus produces greater amounts than Lb. del-brueckii subsp. bulgaricus, and both typesproduce more when grown together than whengrown individually, due to symbiosis betweenthe two.

In some fermented dairy products, particu-larly those prepared with mesophilic cultures,the ratio of diacetyl to acetaldehyde is important.The optimum ratio is 4:1, but when the ratio fallsto 3:1, a "green" off-flavor defect, reminiscentof yogurt, develops. One of the functions ofLeu-conostoc spp. in mixed cultures is to reduce theacetaldehyde produced by Lactococcus to etha-nol, which, at the concentrations found in cul-tures, has no effect on their flavor.

5.5 PLASMIDS

Plasmids are extrachromosomal pieces ofDNA that are much smaller than the chromo-some; their molecular mass ranges from about 2to 100 kDa. Several of the commercially impor-tant properties of starters—including the pro-teinase production, transport of citrate, severalof the enzymes involved in transport andmetabolism of lactose, exopolysaccharide pro-duction, bacteriocin production (Section 5.8),and resistance to phage (Section 5.7.5)—arecommonly encoded on plasmids. The lactoseplasmid encodes the enzymes of the tagatosepathway (galactose-6-phosphate isomerase,tagatose-6-phosphate kinase, and tagatose-1,6-bisphosphate aldolase), P(3gal, and enzyme I andenzyme III of the PTS system. Exopoly-saccharide production by starter LAB is respon-sible for thickening several Scandinavian fer-mented milk products (e.g., Taette and Skyr).

Plasmids are easily lost upon subculturing,and once this occurs, the properties encoded bythe genes on that plasmid are also lost. For thisreason, subculturing of starters should be lim-

Page 98: Cheese Science

ited. Instead, several aliquots of the starter canbe frozen. When required, an aliquot is thawed,and, after two or three subcultures, it is dis-carded and replaced by another frozen aliquot.

Cells that lose the proteinase plasmid becomeproteinase negative (Pit-) and consequently growpoorly in milk. Those that lose the lactose plas-mid are unable to metabolize lactose and there-fore cannot grow in milk, whereas those that losethe citrate transport plasmid are unable to me-tabolize citrate, even though they contain the nec-essary enzymes. In some strains of starters, theproteinase and lactose genes are encoded on thesame plasmid, but in other strains two separateplasmids are involved. Prtr strains are often iso-lated from mesophilic mixed-strain starters. Inthese cultures, Prtr strains rely on Prt+ strains toproduce the amino acids and peptides requiredfor growth. Conjugative plasmids (i.e., plasmidsthat can mediate their own transfer to other cells)played a major role in our understanding of thegenetics and metabolism of LAB.

5.6 INHIBITION OF ACID PRODUCTION

Slow acid development during cheesemakingis an important cause of poor quality cheese.There are four main causes of slow acid produc-tion: natural inhibitors, the presence of antibiot-ics in the milk, bacteriophage, and bacteriocins.Of these, bacteriophage is the most important.

Milk contains natural inhibitors, calledlactenins, that inhibit the growth of some strainsof starter bacteria. The lactenins have been iden-tified as immunoglobulins and lactoperoxidase(LP). The immunoglobulins cause susceptiblestarter bacteria to aggregate. This causes local-ized acid production and precipitation, and insevere cases the aggregates settle on the bottom ofthe cheese vat. The starters still continue to grow,but localized acid production is so great that theyeventually inhibit themselves. Immunoglobulinsare inactivated (denatured) by pasteurization.

LP requires H2O2 and thiocyanate (SCN-) foractivity. SCN" is normally present in milk, and theconcentration is higher in milk from cows fedBrassica plants (cabbage and kale), while theH2O2 can be produced by the starter bacteria dur-

ing growth or through xanthine oxidase or glu-cose oxidase activity (see Chapter 4). The threecomponents are required together to inhibit thegrowth of starters. The actual inhibitor has notbeen identified but is thought to be OSCN-. LP isquite heat resistant and is not inactivated byHTST pasteurization, but it is inactivated by heat-ing to 8O0C for a few seconds (so-called flashpasteurization). Inhibition of starters by lacteninsis unusual in modern cheesemaking because thestrains used have been selected so as not to beaffected to any great extent by lactenins.

Antibiotic residues occur in milk because oftheir use to control mastitis in dairy cows. Aneffective way of curing mastitis is to infuse thecow's udder with antibiotics, especially penicil-lin or its derivatives. This results in contamina-tion of the milk with antibiotics. The concentra-tion of antibiotics in the milk decreases with eachmilking, and generally all the antibiotics will beexcreted within 72 hr, depending on the prepara-tion used. Milk from cows treated with antibioticsshould be withheld for the length of time pre-scribed for the antibiotic preparation. The sensi-tivity of starter cultures to various antibiotics usedin mastitis treatment is shown in Table 5-7.

Thermophilic cultures are much more sensi-tive to penicillin and more resistant to streptomy-cin than mesophilic cultures. In the past, antibi-otic residues were a major cause of slow acidproduction in cheese manufacture, but nowadays,with the availability of simple and sensitive testsfor the detection of antibiotic residues in milk andwith better education of farmers, problems due toantibiotic residues in milk are rare.

5.7 BACTERIOPHAGE

The major cause of slow acid production incheese plants today is bacteriophage (phage).This can significantly upset manufacturingschedules and, in extreme cases, result in com-plete failure of acid production or "dead vats."Phage for Lactococcus were first reported inNew Zealand in 1935 and since then they havebeen described for all starter LAB.

Phage are viruses that can multiply onlywithin a bacterial cell. They are ubiquitous in

Page 99: Cheese Science

nature and are so small that they can be "seen"only with the electron microscope. They have ahead, which contains the DNA, and a tail, whichis composed of protein. A photomicrograph of aphage for Lc. lactis is shown in Figure 5-14.Morphologically, three types of phage for lacto-cocci can be distinguished: small isometric-headed (spherical-headed) phage (the most com-mon), prolate-headed (oblong-headed) phage,and large isometric-headed phage. The tail var-ies in length from 20 to 500 nm. Other features,such as collars between the head and the tail,base plates at the end of the tail, and fibers on thebase plates, may also be present. Phage multipli-cation occurs in one or two ways, called the lyticand lysogenic cycles. Lytic and lysogenicphages are sometimes called virulent and tem-perate, respectively.

5.7.1 The Lytic Cycle

In the lytic cycle (Figure 5-15), the first stepinvolves adsorption of the phage onto special at-tachment sites, called phage receptors, on thecell surface of the host. This step requires Ca2+,and prevention of this step is the basis for the useof phosphate and citrate as chelators in phage-inhibitory media. Phage adsorb to the cellthrough their tails. Once a phage has attached tothe receptors, it injects its DNA into the host

cell. Immediately, phage DNA and phage pro-teins are produced rather than host cell DNA andproteins. The phage DNA is packaged in a con-centrated form in the phage head, and whenphage synthesis is completed, the cell lyses, re-leasing new phage particles, which start the pro-cess again. Lysis is caused by a lytic enzyme,called lysin, which is encoded in the phage DNAand hydrolyzes the cell wall of the host cell.

The growth of virulent phage in a sensitivehost is characterized by both a latent period anda burst size, which are determined in one-stepgrowth experiments (Figure 5-16). For such ex-periments, phage and whole cells are mixed sothat the ratio of cells to phage (the multiplicity ofinfection) is low, and the number of phage ismonitored periodically during incubation. At thebeginning, the number of phage remains low,since new phage are being synthesized inside thecell. This is called the latent period and spans thetime from the initial adsorption of the phage tothe host cell until the detection of phage progenyafter cell lysis. The suddenly increased numberof phage, called the burst size, is caused by lysisof the host cells by phage lysin. For lactococcalphage, the latent period varies from 10 to 140min and the burst size varies from about 10 to300 phage. Compared with starters, phage multi-plication is very rapid. Assuming a latent periodof 1 hr and a burst size of 150,1 phage will result

Table 5-7 Concentration of Some Antibiotics That Cause 50% Inhibition of Growth of Starter Bacteriain Milk

Antibiotic fag/ml + SD)

Organism

Lc. lactis subsp. cremoris

Lc. lactis subsp. lactis

Sc. thermophilus

Lb. delbrueckii subsp.bulgaricus

Lb. delbrueckii subsp. lactis

No. of Strains

4

4

3

2

1

Penicillin

0.11 ±0.028

0.1 2 ±0.025

0.01 ± 0.002

0.03 ± 0.006

0.024

Cloxacillin

1 .69 ± 0.38

2.16 ±0.41

0.42 ± 0.07

0.29 ± 0.04

0.24

Tetracycline

0.14 ±0.02

0.15 ±0.05

0.19 ±0.06

0.37 ± 0.04

0.60

Streptomycin

0.67 ±0.15

0.53 ±0.1 8

10.5 ±0.29

3.0 ±2.0

2.29

Page 100: Cheese Science

in the production of 22,500 phage (150 x 150) ina little over 2 hr. In 3 hr, the number of phagewill increase to 3.4 x 106. In 3 hr, a Lactococcuscell will multiply three times, producing only 8cells. Thus, the phage will vastly outnumber thebacterial cells very quickly. This clearly indi-cates the serious problems that occur followingcontamination with phage. Phage multiplicationduring cheesemaking is shown in Figure 5-17.The initial number of phage was 100/ml (i.e., log2), and within a few hours multiplication to 107/ml (equivalent to log 7) had occurred.

5.7.2 The Lysogenic Cycle

The second phage multiplication process iscalled the lysogenic cycle. Adsorption and injec-tion of DNA occur as in the lytic cycle, but, in-stead of phage multiplication, the phage DNA isinserted into the bacterial chromosome and mul-tiplies with the chromosome. Under these condi-tions, the phage is called a temperate phage or aprophage, and the cells are considered to be

lysogenized. Most strains of LAB are lysogenic,and in this state the host cell is immune to attackby its own phage. Generally, the prophage alsoimmunizes the host cell to closely related strainsof phage. This is called superimmunity.

In certain circumstances, some temperatephage can be induced, become lytic, and multi-ply. The host cells in which these phage multiplyare called indicator strains. The conditions thatcause induction in commercial practice are un-clear, but in the laboratory UV light and the anti-biotic mitomycin C are used. In many bacteria,lysogenic (temperate) phage are considered tobe the source of phage, but this has not beenshown for starter LAB, except for the strain ofLb. casei used in the production of the Japanesefermented milk Yakult.

5.7.3 Pseudolysogeny

Many mixed-strain starters are permanentlyinfected with a low number of virulent phage.These are called "own" phage to distinguish

Figure 5-14 Schematic drawing (left) and corresponding electron micrograph (right) of a bacteriophage ofLactococcus lactis. The important structural components of the phage particle are shown in the drawing.

Head

DNA

Collar

Tail

Base plate

Spikes

Page 101: Cheese Science

Figure 5-15 Schematic drawing of the propagation of virulent (right branch) and temperate (left branch) bacte-riophage in a host cell. The cell is shown with its chromosomal DNA. The phage DNA is indicated by dottedlines.

Time after infection, minFigure 5-16 Results of a one-step growth experiment on Lactococcus lactis infected with a lytic phage. Therelease of phage (A) began 25 min (latent period) after infection (burst size = 124). The increase of free phage (asplaque-forming units [pfu]) was accompanied by a decrease in culture turbidity (O) due to cell lysis.

Turb

idity

, O

D 6O

O

Pha

ge li

ter,

p

.f.u

. pe

r in

fect

ed c

ell

Integration ofphage DMA intothe chromosome

Adsorptionofphage

Injection ofphage DNA

Induction ofprophage Maturation

of phage

Lysis ofcell

Page 102: Cheese Science

them from lytic or "disturbing" phage and theymultiply on any phage-sensitive strains presentin the culture. Normal growth of the mixed cul-ture is unaffected by own phage because of thepresence of large numbers of acid-producing,phage-insensitive cells. This phenomenon iscalled pseudolysogeny, and the phage insensi-tivity of the whole culture is stable as long as noinfection with disturbing phage occurs.

5.7.4 Classification of Phage

Phage cannot multiply outside of their hosts,and therefore classical bacterial identificationmethods cannot be used to identify them. Othertechniques have been developed, and some ofthese have already been mentioned (e.g., phagemorphology). Phage protein composition, hostrange, serology, and DNA homology are alsoused. Host range is of considerable practical sig-nificance, since starter cultures attacked by thesame phage should not be used in rotations. Hostranges can be broad (one phage attacks severalstrains) or narrow (the phage attacks only one ortwo strains). DNA homology has resulted inlactococcal phage being divided into 12 "spe-cies," 3 of which, P36 (isometric), P335 (isomet-

ric), and C2 (prolate), are commonly found incheese plants.

5.7.5 Phage-Resistance Mechanisms

Several phage-resistance mechanisms, in-cluding inhibition of phage adsorption, restric-tion-modification mechanisms, and abortive in-fection mechanisms, are found in LAB. All ofthese are commonly encoded on plasmids.

In adsorption inhibition, the receptor sites forthe phage on the cell surface are masked, so thephage cannot attach to the cell and hence are un-able to multiply. In Lc. lactis subsp. cremorisSKl 10, adsorption inhibition has been shown tobe plasmid encoded, and the masking agent hasbeen identified as a galactose-containing lipo-teichoic acid.

Restriction-modification involves two en-zymes, one of which, the restriction enzyme, hy-drolyzes the phage DNA. The other (modifica-tion) enzyme modifies the host cell DNA,usually by methylation of some of the nucleicacid bases, in such a way that the restriction en-zyme cannot hydrolyze it. This mechanism isoperative only after adsorption and injection ofphage DNA.

Time, h

Figure 5-17 Phage multiplication (plaque-forming units [pfu]) during Cheddar cheese manufacture in two vats.

Log

pfu

/ml

VatlVat 2

Page 103: Cheese Science

Abortive infection (obi) is the term used forphage-resistance mechanisms that do not involveeither inhibition of adsorption or restriction-modification systems. Generally, a total loss ofplaque formation (see Section 5.7.6) or a reduc-tion in plaque size occurs, as a result of a reduc-tion in both the latent period and burst size, al-though in some cases only one of these is affected.

Many phage-resistance plasmids are conjuga-tive, and this fact has been used to improve thephage resistance of phage-sensitive commercialcultures. The technique is relatively simple (Fig-ure 5-18): Lac~ mutants of the strain harboringthe phage-resistance plasmid are isolated (lac~phager) and mixed with lac+ phage8 recipients.Lac+ phager transconjugants are selected in thepresence of excess virulent phage on lactose agar,which contains a dye to indicate acid production.The lac~ phager cells do not grow very well on thismedium, and they are destroyed by the phagepresent in the agar. The lac+ phager transcon-jugants are then isolated and checked for theirability to produce acid and for the presence of thephage-resistance plasmid. This is a totally naturalprocess that does not involve genetic engineeringtechniques, and it has been used in the productionof phage-resistant strains for commercial use. Anew strategy involving the sequential use of thesame strain of starter containing different phagedefense mechanisms has also been suggested(Sing & Klaenhammer, 1993).

5.7.6 Detection of Phage

Phage are much smaller than their hosts andcan be easily separated from them by filtrationthrough a 0.45 jam filter. The host cells are re-tained by the filter while the phage particles aresmall enough to pass through into the filtrate.Phage are easily detected. A small volume (e.g.,0.1 ml) of a filter-sterilized (host-free) sample ofmaterial suspected of containing phage is addedto 10 ml of milk that has been inoculated with asusceptible host. After incubation at the opti-mum temperature of the host for 6-10 hr, the pHis measured. A difference of more than 0.3 unitsbetween the pH of host grown in the absence orpresence of the material suspected of containing

phage indicates the presence of phage (Figure 5—19). The decrease in pH can also be visualizedby adding a suitable pH indicator (bromocresolpurple) to the milk. In broths, measurement ofthe optical density at 600 nm is used. The opticaldensity increases for a little while as the hostgrows but then decreases as the cells Iyse. Theseprocedures give no indication of the number ofphage present.

The number of phage in a sample can becounted relatively easily. The material suspectedof containing phage is filter sterilized, and 10-fold dilutions of the filtrate are made. Then 1 mlof each dilution is mixed with 0.1 ml of the hostculture (containing 108 cells), 0.1 ml of 0.185 MCa2+, and 2.5 ml of a suitable molten mediumcontaining 0.7% agar tempered to 450C. (Thismedium is referred to as "sloppy agar," and itsfunction is to allow the phage to diffuse andform fairly large plaques.) The mixture is thenpoured immediately onto a prehardened plate ofthe same medium containing the normal amountof agar. After incubation at the optimum tem-perature of the host for several hours, clearzones, called plaques, are seen in the back-ground lawn of bacterial growth if phage arepresent (Figure 5-20). Each plaque is consideredto have arisen from one phage, and counting thenumber of plaques and multiplying by the dilu-tion factor gives the number of plaque-formingunits (pfu) per ml.

5.7.7 Source of Phage

To control phage, it is important to identifytheir source. However, the ultimate source ofphage for LAB is still unclear. In most bacteria,lysogenic phage are considered to be the source oflytic phage. Many mixed cultures contain lyso-genic phage, but no DNA homology has beenfound between these phage and lytic phage thatattack these cultures. Whey contains a large num-ber of phage, and aerosols produced during sepa-ration of fat from whey are probably the primarysource of phage in cheese plants. Raw milk is alsoconsidered to be an important source of phage,but only relatively small numbers of lactococcalphage have been isolated from it.

Page 104: Cheese Science

Figure 5-18 Schematic representation of a geneticstrategy for the construction of phage-insensitivestarter cultures by conjugation. A phage-resistanceplasmid with an abortive infection determinant istransferred from a donor lacking the ability to metabo-lize lactose (Lac~) to a recipient culture that can me-tabolize lactose (Lac+). Phage-resistant transconju-gants are selected on lactose indicator agar in thepresence of the virulent phage. Under these condi-tions, the Lac~ donor cells do not grow well. Finally, ahybridization experiment is carried out to distinguishtrue transconjugants from phage-resistant mutants.

5.7.8 Control of Phage

Starters are often produced in large volumes;a 500 L tank of starter contains around 5 x 1014

cells. One phage getting into such a tank couldhave serious consequences. Therefore, the mostimportant factor in producing good qualitycheese is to ensure that the bulk starter is free ofphage. This point cannot be overstressed, but itis often overlooked in commercial practice.

Daily determination of phage levels in startersand cheese whey should be an integral part ofany well-designed quality control scheme in amodern cheese factory. The important factors incontrolling phage are

• use of a limited number of cultures• aseptic inoculation• use of phage-inhibitory media heat-treated

to 850C for 30 min• rotation of phage-unrelated strains• physical separation of starter and cheese

production areas (there should be no directaccess between the starter unit and cheeseproduction)

• addition of starter and rennet together• chlorination of vats between fills• use of closed vats• physical separation of the cheese produc-

tion area from the separator used to recoverfat from whey

• heat treatment (> 9O0C for 45 min) of startercontaining phage before discarding it

Many of these factors are common sense but afew need further elaboration. The greater thenumber of strains used in a plant, the greater isthe likelihood of a phage attack. Therefore, de-fined strain cultures should be used instead ofmixed cultures. Phage are quite resistant to heat.Some phage withstand heating at 750C for a fewminutes, and hence the medium used for grow-ing the starter must be heat-treated at a high tem-perature (e.g., 850C for 30 min) to inactivate anyphage that might be present. The medium shouldbe heated in the tank in which the culture is to begrown. Some cheese manufacturers heat-treatthe milk in a pasteurizer and then fill the startertank with the heated milk. This practice is notrecommended because of the danger that phagemight be present in an otherwise clean startertank. One phage in a tank can be sufficient tocause a reduction in the activity of a culture. Theheat treatment has other effects beside the inacti-vation of phage: it improves the nutritional valueof the milk as a culture medium because it inacti-vates the natural inhibitors and causes slight hy-drolysis of the proteins. Bulk cultures contami-

Addition of viratent phagefor recipient & sectionon lactose indicator agar

Conjugation

Donor Recipient

Page 105: Cheese Science

Time, h

Figure 5-19 Effect of phage on acid production byLactococcus lactis in 10% (w/v) reconstituted skimmilk at 3O0C; control, A; infected with phage, •.

nated with phage should be heat-treated (> 9O0Cfor 45 min) to prevent their spread.

In the past, milk was the medium used forgrowth, but nowadays phage-inhibitory mediaare used (see Section 5.9). Inoculation should bedone as aseptically as possible; an aseptic inocu-lation device has been developed and is com-monly used in the Netherlands for starter pro-duction (Figure 5-21).

During cooling, the air entering the startertank should be filtered through a high-efficiencyparticulate air filter to prevent phage in the airfrom entering the tank. Transfer of airbornephage into the starter tank during incubationmay be eliminated by a slight positive air pres-sure in the tank and starter room.

Defined cultures are also very useful in con-trolling phage. Such cultures contain only asmall number of phage-unrelated strains andhelp to reduce the numbers of different phagepresent in factories. The same strains are usedeach day and wheys from bulk starter andcheesemaking are checked daily for phageagainst all the strains in use. Bulk starter wheyshould always be clear of phage. If a phage in-fection does occur in the bulk starter culture, theculture should not be used, even if the phagelevel is low, as phage will multiply more rapidly

than bacteria when subcultured in the cheesemilk. Instead, the bulk culture should be re-placed by a bacteriophage-insensitive mutant(BIM) or another phage-unrelated strain. Thebulk culture should be heat-treated at higher than850C for 30 min before discarding it to preventthe spread of phage throughout the factory. Onecan tolerate the presence of phage in the cheesewhey and, indeed, it is probably inevitable, but ifthe level increases to more than 106 pfu per ml ofwhey, then the infected strain should be re-moved and replaced with a phage-unrelated cul-ture or a BIM.

BIMs are relatively easily isolated. Thephage-sensitive strain is grown with the phage inmilk until the milk coagulates. Coagulation isdue to the growth of a small number of phage-insensitive strains that are present in almost ev-ery strain of starter. The coagulated culture isthen plated on a suitable medium and severalcolonies examined for their ability to coagulatemilk rapidly, for their resistance to the particularphage, and for their ability to grow in milk with-out causing off-flavors. It is sometimes not pos-sible to isolate BIMs from cultures; presumably,such cultures do not contain phage insensitivecells.

Addition of rennet as soon as possible afterthe starter has been added to the milk in thecheese vat also helps, because the coagulum willphysically separate phage-infected cells fromnoninfected cells, and the phage are unable topenetrate the curd to locate noninfected cells.

As little as 1 ml of residual whey in a cheese vatcan be a potent source of lytic phage. Therefore,cheese vats should be cleaned routinely and chlo-rinated between fills. Chlorination is a very effec-tive phagicide. The final step in cleaning equip-ment, such as vats and filling lines, should includea chlorination component. Exposure to 100 jig ofavailable or "active" chlorine per ml for 10 min isusually sufficient to inactivate all phage on equip-ment surfaces. The residual chlorine should notbe rinsed from the equipment to prevent furthercontamination with phage from hoses or water.Any residual chlorine on the equipment is imme-diately inactivated when it comes in contact with

pH

Page 106: Cheese Science

Figure 5-20 Agar plates with a lawn of Lactococcus lactis cells showing clear zones (plaques) due to phageinfections. Each plaque is the result of the infection of a single cell with a single phage. The progeny phagemultiply subsequently on neighboring cells.

Page 107: Cheese Science

organic material, like milk. The use of closed vatsis also recommended to prevent contaminationfrom whey aerosols.

5.8 BACTERIOCINS

Many bacteria produce proteins that inhibitthe growth of other bacteria, and these proteinsare called bacteriocins. Generally, they have anarrow host range, inhibiting only closely re-lated bacteria, but bacteriocins with broad hostranges are not unknown. Bacteriocins producedby Gram-positive bacteria usually do not inhibitGram-negative bacteria and vice versa. Usually,

they are of small molecular mass, but bacterio-cins of high molecular mass also occur. Theirproteinaceous nature and their relatively narrowhost range distinguish them from antibiotics.During the past decade, food safety has becomea major issue, and a concerted effort has beenmade to identify bacteriocins that inhibit patho-gens and food spoilage organisms. LAB areideal for this purpose, because they are generallyregarded as safe organisms. Bacteriocins that in-hibit Listeria are particularly useful, since L.monocytogenes in soft cheeses has been incrimi-nated as a cause of human listeriosis (see Chap-ter 20). For use in foods, the bacteriocin should

Figure 5-21 An aseptic inoculation device for starter tanks. The device is opened using the screw (6). Cartons(3) containing the frozen inoculum are inverted and placed on the holder (4). The device is assembled and filledwith chlorinated water through the pipe (5). This helps to thaw the culture, and as soon as thawing begins, thechlorinated water is run to waste through the valve (2). The screw (6) is then turned, causing the prongs (7) topuncture the foil lid. The cartons empty and their contents (the inoculum) are added to the starter tank through thevalve (1). Any residual chlorine has no effect on the starter, as it will be diluted and inactivated on contact with thefluid in which the starter cells are suspended.

Page 108: Cheese Science

• be heat stable• be acid stable• be resistant to potential proteinases present

in the food• be active over a prolonged period• be active at the pH of food (4.5 to 7.0)• have a bactericidal rather than a bacterio-

static mode of action• have a broad host range, inhibiting several

pathogens and spoilage microorganisms

In assaying bacteriocins produced by LAB, itis important to distinguish between bacteriocinsper se and other potential inhibitors that can beproduced by these bacteria, including H2O2, lac-tate, and acetate. This can be done by determin-ing the effect of catalase on the activity and byneutralizing culture supernatants before assay-ing them.

Bacteriocin production by LAB is common.The best known is nisin, which is produced bysome strains of Lc. lactis subsp. lactis. Nisin hasa molecular weight of 3,353 Da, contains 34amino acids, and normally occurs as dimers andtetramers. It is soluble only at low pH, which re-duces its potential use significantly. Its heat sta-bility depends very much on pH; for example, itis stable to autoclaving at 1150C at pH 2 butloses 40% of its activity at pH 5. Nisin has abroad spectrum of activity, inhibiting Bacillus,Clostridium, Staphylococcus, Listeria, andStreptococcus. It is particularly active againstspores of Cl tyrobutyricum, which is the causeof late-gas production in hard natural and pro-cessed cheese. In the case of spores, nisin acts bypreventing spore germination, but in preventingthe growth of bacteria, it acts by creating poresin the cell membrane of sensitive cells, destroy-ing the proton-motive force and permitting therelease of cytoplasmic components from thecell. Nisin can be used as a replacement for NO3

in foods. Nisin is synthesized as a 57 amino acidpeptide that is posttranslationally modified to a34 amino acid peptide containing unusual aminoacids: dehydroalanine, dehydrobutyrine, lanthi-onine (Ala-S-Ala), and (3-methylanthionine(Ala-S-Aba [aminobutyric acid]). Nisin is calleda !antibiotic because it contains lanthionine and

(J-methylanthionine. Besides nisin, LAB pro-duce other !antibiotics, such as lactocin 481,which is produced by Lc. lactis subsp. lactisCNRZ 481, and lactocin S, which is produced byLactobacillus sake L45. Other starter bacteria,including Lb. helveticus and Ln. mesenteroides,have been shown to produce broad-spectrumbacteriocins, called helveticin J and mesen-tericin, respectively. Helveticin is temperaturesensitive and is a large, complex molecule,which limits its usefulness in foods.

There are several forms of nisin arising fromamino acid substitutions; for example, nisin Zdiffers from nisin A in having asparagine insteadof histidine at position 27. Because of its greatersolubility, nisin Z also has greater inhibitory ac-tivity than nisin A.

Bacteriocins may also be complexed withother macromolecules, including lipids andpolysaccharides. They are encoded either on thechromosome or on plasmids. Bacteriocin-pro-ducing bacteria must also have a gene that en-codes immunity to it.

The bacteriocins produced by LAB can be di-vided into 3 groups:

Class I: !antibioticsClass II: small heat-stable nonlantibiotics

Ua: single-peptide bacteriocins(e.g., pediocin pAH)lib: two-peptide bacteriocins(e.g., lactoccin G and plantaricin E)Uc: sec-dependent secreted bacte-riocins

Class III: large heat-labile proteins (e.g., hel-veticin J and caseicin 80)

A fourth class containing complex bacterio-cins composed of proteins, lipids, and carbohy-drates (e.g., plantaricin S and lactocin 27) hasbeen proposed, but the evidence suggests thatthese complex molecules are artifacts caused byinteraction between cell constituents or thegrowth medium and the bacteriocin. Class Uabacteriocins have been isolated from species ofseveral genera, including Pediococcus (afterwhich they are named), Leuconostoc, Lactoba-cillus, and Enterococcus, and they share consid-

Page 109: Cheese Science

erable amino acid sequence homology (38-55%).

It is difficult to compare the reported hostranges of bacteriocin producers because differentmethods and, more importantly, different strainshave been used. Most bacteriocins produced byLAB have a bactericidal mode of action, due tothe destruction of the proton motive force in-volved in transport, which also results in the re-lease of intracellular compounds. Lactocin 27,produced by Lb. helveticus, is an exception and isbacteriostatic. Bacteriocin-producing strains, ifpresent in mixed cultures, will reduce the numberof strains in these cultures upon subculturing,eventually leading to a mixture of perhaps 1 or 2strains. Such cultures will also be much moreprone to attack by phage than similar mixed cul-tures that do not contain bacteriocin producers.

5.9 PRODUCTION OF STARTERS INCHEESE PLANTS

Until perhaps 30 years ago, milk was the me-dium used for starter production in cheese fac-tories. The milk was selected from cows notsuffering from mastitis and so was free of antibi-otics. Spray-dried skim milk powder replacedcheese milk when it became more readily avail-able; the solids concentration used was 10-12%(w/v). Today, phage-inhibitory media have re-placed skim milk powder for the production ofbulk cultures. These are generally carefully for-mulated proprietary media that contain milk orwhey solids, phosphate and/or citrate, and yeastextract. The function of the citrate and phos-phate is to chelate Ca2+, which are essential forphage multiplication, and the function of theyeast extract is to stimulate growth of the starter.

The medium is generally heated at 850C orhigher for 30 min in the tank in which the starteris to be grown, and it is cooled to the incubationtemperature of about 420C, in the case of ther-mophilic cultures, or 210C, in the case of meso-philic cultures. It is then inoculated with about1% (v/v) of the culture, and after incubation for8-10 hr, in the case of thermophilic cultures, orovernight (16 hr), in the case of mesophilic cul-tures, the culture is fully grown.

In the past, the inoculum for bulk cultureswas built up progressively from a small volumeof mother culture to larger volumes over sev-eral days (Figure 5-22). Minimal subculturingof starters is desirable to prevent loss of plas-mids and to maintain the balance betweenstrains in mixed cultures. Nowadays, inoculafor bulk cultures are generally obtained fromspecialized laboratories in which the startersare grown under optimum conditions (e.g., pH6.3 and 280C, in the case of lactococci) in anappropriate medium, the composition of whichis proprietary. After growth, the cells are har-vested by ultrafiltration or centrifugation andfrozen in liquid N or freeze-dried in sufficientvolumes to inoculate 300, 500, or 1,000 L ofmedium directly. Such cultures generally con-tain about 5 x 109 cfu/g, or roughly 5 timesmore than a normal milk-grown culture. Cryo-protectants (e.g., glycerol, sucrose, or monoso-dium glutamate) are often added to protect thecells during freezing or freeze-drying. Wherepure cultures (single strains) are used, it ismuch more economical to start from a smallvolume and build up the necessary inoculumover a few days. Superconcentrated cultures arealso commercially available to inoculate thecheese milk in the vat directly. These are oftencalled DVS (direct-vat-set) or DVI (direct-vat-inoculation) cultures. Such cultures are expen-sive and are often kept as standby cultures foruse in a crisis, such as a phage outbreak.

Thermophilic cultures are usually grown attheir optimum temperature (~ 420C), but meso-philic cultures are grown at 210C, which is about90C below their optimum temperature. The rea-son for the lower temperature of incubation inthe case of mesophilic cultures is that if the milkis inoculated at, say, 3 PM, the cultures are fullygrown 16 hr later, at 7 AM the following morning,which is often the starting time for cheese-making in small cheese plants.

Fully grown cultures of Lactococcus and Sc.thermophilus generally reduce the pH of milkfrom its initial value of 6.6 to 4.6 and produceabout 90 mM (0.8%, w/v) lactic acid, whereasmany thermophilic lactobacilli can produce upto 200 mM (~ 2%, w/v) lactic acid and reduce

Page 110: Cheese Science

the pH to around 3.0. The actual amount of acidproduced depends on the buffering capacity ofthe medium. Such cultures contain about 109

cm/ml. However, in cheese factories, the lacto-bacilli are usually grown only to a pH of around4.0 (equivalent to 1%, w/v, lactic acid). Culturesare then generally cooled to either 40C or 1O0C,in the case of mesophilic and thermophilic cul-tures, respectively, and checked for activity (i.e.,their ability to produce lactic acid). Activity isnormally assessed by measuring acid productionor the decrease in the pH of the culture grownunder standardized conditions, such as in 10%(w/v) reconstituted skim milk after 6 hr at 3O0Cfor mesophiles or 5 hr at 4O0C for thermophilesusing a standardized inoculum, generally 1%(v/v). Sometimes incubation is carried out overthe cheese temperature profile, to simulate theirpotential activity in cheese. For day-to-day com-parisons between cultures, it is important tostandardize the inoculum and the incubationconditions very carefully. Under the inoculationand incubation conditions just described, Lc.lactis subsp. lactis, Sc. thermophilus, Lb. hel-veticus, and Lb. delbrueckii subsp. lactis will re-duce the pH from 6.6 to less than 5.3 (equivalentto ~ 0.5%, w/v, lactic acid). Both pH and theamount of acid produced (i.e., the titratable acid-

ity) are used to monitor growth. pH is mucheasier and faster to measure, and automatedequipment able to measure the pH of up to 24samples continuously is commercially available.When cooled to 40C, mesophilic cultures willretain activity for 2-3 days, and thermophiliccultures for up to 12 days.

The pH of the medium used in the produc-tion of bulk cultures is often controlled duringgrowth. Control can be external or internal. Ex-ternal control involves the use of a pH controlunit and a neutralizer (e.g., NH4OH), and inter-nal control involves the use of an insolublebuffer that dissolves as lactic acid is producedand maintains the pH above 5.3. pH control in-creases the number of cells per unit volume andtherefore reduces the volume of starter requiredfor cheesemaking. Neutralization of the pH ofthe bulk culture to 6.5 after growth and furtherincubation for a few hours will also result in anincrease in the number of cells in the culture.

The exact amount of bulk starter made eachyear throughout the world is difficult to calculateexactly, but the following three assumptions al-low us to estimate it:

1. Annual cheese production is about 15million tonnes (i.e., 15 x 109 kg).

Figure 5-22 A possible protocol for scaling up intermediate cultures for bulk production of Lactococcus startersin cheese factories.

Mother Culture

100 ml RSM

(U-24 h at 21C]

If or ImI

VIAL

Intermediate Culture

1« Lttres RSM

[14-14 a at 21C]

Bulk Culture

10OO Litres RSM

[14-K h at 21C)

Jacket

Agitator

SterileairInoculation

port

Page 111: Cheese Science

2. Ten liters of milk are required to make 1kg of cheese.

3. An inoculation rate of 1.0% (v/v) is used.

The second assumption is only valid for hardcheeses like Cheddar. Soft and semi-soft cheesesrequire less milk (in the case of Quarg, about 5liters per kg); while Parmigiano-Reggiano, avery hard Italian cheese, requires about 12 litersper kg). This assumption also does not take intoaccount the differences in solids levels of cow,goat, sheep, and buffalo milk. The third assump-tion also only applies to some cheeses. An in-oculation rate of 1.5 to 1.8% (v/v) is used inCheddar and perhaps less than 0.3% (v/v) inEmmental. These considerations mean thatabout 15 x 1010 liters of milk are used to makecheese annually which requires production of 15XlO 8 liters of starter. Assuming that starters con-tain about 109 cells per ml, this amounts to pro-duction of 15 x 1020 starter cells per year.

5.10 MEASUREMENT OF GENERATIONTIMES

If bacterial growth is balanced, the followingdifferential equation describes the increase incell mass with time:

dx/dt = kxwhere ;c is the cell mass in dry weight/g, t is timeand k is the growth rate constant. Since lacticacid production is proportional to cell mass,

d(lactic acid)/dt = k (lactic acid)

Integrating we get

In (lactic acidt) - In (lactic acid0) = k (t —10)

REFERENCES

Garvie, E.I., & Farrow, J.A.E. (1982). Streptococcus lactissubsp. cremoris (OrIa-Jensen) comb. nov. and Strepto-coccus lactis subsp. diacetilactis (Matuszewski et al.)nom. rev. comb. nov. International Journal of SystematicBacteriology, 32, 453^55.

where lactic acido is the amount of lactic acidpresent at time0, lactic acid, is the amount of lac-tic acid present at time t, and k is the growth rate.Converting to logio we get

logio(lactic acidt) - Iogi0(lactic acido) = k(t - to)/2.303

Rearranging we get

logioOactic acidt) = k(t -10)/2.303 + Iog10(lactic acido)

This equation is that of a straight line (y = mx +c)9 where m = &/2.303 ork = mx 2.303.

The generation time (g) is the time requiredfor the amount of lactic acid to double. Substitut-ing in this equation gives us

Ioglo2 = k(g)/2.303 + loglol

Rearranging we get

g = log,0 2 x 2.303/k + Iog10 1

And by substituting for k we get

g = log2/m + 0 = 0.301/m

The slope, m, is easily calculated by regressionanalysis of the data. It is important to correcteach value for the inherent acidity of the un-inoculated milk. The data in Figure 5-2 are plot-ted correctly (i.e., semi-logarithmically), but thegeneration time cannot be calculated accuratelybecause the acidity of the uninoculated milk wasnot reported. However, the data for the culturesshown in Figure 5-13 were corrected for theamount of lactic acid in the uninoculated milk.Based on lactic acid production at 210C, thesecultures had a generation time of around 2.2 hr;at 3O0C, the generation time would be 1 hr. Thedata need to be transformed to logio before theregression equation is calculated.

Goupil, N., Cormier, G., Ehrlich, S.D., & Renault, P. (1996).Imbalance of leucine flux in Lactococcus lactis and its usefor the isolation of diacetyl overproducing strains. Ap-plied and Environmental Microbiology, 62, 2636—2640.

Page 112: Cheese Science

Kunji, E.R.S., Mierau, L, Hagting, A., Poolman, B., &Konings, W.N. (1996). The proteolytic systems of lacticacid bacteria. Antonie van Leewenhoek, 70, 187-221.

Lodics, T., & Steenson, L. (1990). Characterisation of bacte-riophages and bacteria indigenous to a mixed-straincheese starter. Journal of Dairy Science, 73, 2685-2696.

Nomura, M., Kimoto, H., Someya, Y., & Suzuki, I. (1999).Novel characteristic for distinguishing Lactococcus lactissubsp. lactis from subsp. cremoris. International Journalof Systematic Bacteriology, 49, 163-166.

Schleifer, K.H., & Kilpper-Balz, R. (1987). Molecular andchemotaxonomic approaches to the classification ofstreptococci, enterococci and lactococci: A review. Sys-tematic and Applied Microbiology, 10, 1-9.

Schleifer, K.H., Kraus, J., Dvorak, C., Kilpper-Balz, R.,Collins, M.D., & Fischer, W. (1985). Transfer ofStrepto-

SUGGESTED READINGS

Board, R.G., Jones, D., & Jarvis, B. (Eds.). (1995). Micro-bial fermentations: Beverages, foods and feeds [Specialissue]. Journal of Applied Bacteriology, 79, 1S-139S.

Chapman, H.R., & Sharpe, M.E. (1990). Microbiology ofcheese. In R.K. Robinson (Ed.), Dairy microbiology (vol.2). London: Elsevier Applied Science Publishers.

Cocaign-Bousquet, M., Garriguess, C., Loubiere, P., &Lindley, N.D. (1996). Physiology of pyruvate metabolismin Lactococcus lactis. Antonie van Leewenhoek, 70, 253—267.

Cogan, T.M., & Accolas, J.P. (1996). Dairy starter cultures.New York: VCH.

Cogan, T.M., & Hill, C. (1993). Cheese starter cultures. InP.P. Fox (Ed.), Cheese: Physics, chemistry and microbi-ology (2d ed., VoI 1). London: Chapman & Hall.

The Dairy Leuconostoc [Symposium]. (1994). Journal ofDairy Science, 77, 2704-2737.

Daly, C., Fitzgerald, G.F., & Davis, R. (1996). Biotechnol-ogy of lactic acid bacteria with special reference to bacte-riophage. Antonie van Leeuwenhoek, 70, 99-110.

De Vuyst, L., & Vandamme, EJ. (1994). Bacteriocins oflactic acid bacteria. Glasgow: Blackie Academic andProfessional.

Fifth Symposium on Lactic Acid Bacteria: Genetics, Me-tabolism and Applications (1996). Antonie van Leeuwen-hoek, 12, 1-271.

Fourth Symposium on Lactic Acid Bacteria: Genetics, Me-tabolism and Applications. (1993). FEMS MicrobiologyReviews, 12, 1-272.

Gasson, MJ., & de Vos, W.M. (1994). Genetics and bio-

coccus lactis and related streptococci to the genus Lac-tococcus gen. nov. Systematic and Applied Microbiology,6, 183-195.

Sherman, J.M. (1937). The streptococci. Bacteriology Re-views, 1, 3-97.

Sing, W.D., & Klaenhammer, T.R. (1993). A strategy forrotation of different bacteriophage defences in a lacto-coccal single-strain starter culture system. Applied andEnvironmental Microbiology, 59, 365-372.

Taillez, P., Tremblay, J., Ehrlich, S.D., & Chopin, A. (1998).Molecular diversity and relationships within Lactococcuslactis as revealed by randomly amplified polymorphicDNA (RAPD). Systematic and Applied Microbiology, 21,530-538.

technology of lactic acid bacteria. London: Chapman &Hall.

Hoover, D.G., & Steenson, L.R. (1993). Bacteriocins of lac-tic acid bacteria. New York: Academic Press.

Hugenholtz, J. (1993). Citrate metabolism in lactic acid bac-teria. FEMS Microbiology Reviews, 12, 165-178.

Lawrence, R.C., & Heap, H.A. (1986). The New Zealandstarter system (Bulletin No. 199). Brussels: InternationalDairy Federation.

Lodics, T., & Steenson, L. (1993). Phage-host interactions incommercial mixed-strain cultures: Practical significance,A review. Journal of Dairy Science, 76, 2380-2391.

Nes, I.F., Diep, D.B., Haverstein, L.S., Bruberg, M.R.,Eijsink, V., & HoIo, H. (1996). Biosynthesis of bac-tericins in lactic acid bacteria. Antonie van Leeuwenhoek,70, 113-128.

Pritchard, G.G., & Coolbear, T. (1993). The physiology andbiochemistry of the proteolytic system in lactic acid bac-teria. FEMS Microbiology Reviews, 12, 179-206.

Second Symposium on Lactic Acid Bacteria: Genetics, Me-tabolism and Applications. (1987). FEMS MicrobiologyReviews, 46, 201-379.

Stadhouders, J. (1986). The control of starter activity. Neth-erlands Milk and Dairy Journal, 40, 155-173.

Stiles, M.E. (1996). Biopreservation by lactic acid bacteria.Antonie van Leeuwenhoek, 70, 331-335.

Third Symposium on Lactic Acid Bacteria: Genetics, Me-tabolism and Applications. (1990). FEMS MicrobiologyReviews, 87, 1-188.

Page 113: Cheese Science

As discussed in Chapter 2, the milk for mostcheese varieties is coagulated through the actionof selected proteinases, called rennets. The ren-net-induced coagulation of milk is in fact a two-stage process (Figure 6-1). The primary phaseinvolves the specific enzymatic modification ofthe casein micelles to produce paracasein mi-celles that aggregate in the presence of Ca2+ attemperatures above about 2O0C. Aggregation ofthe rennet-altered micelles is referred to as thesecondary phase of coagulation. The primaryphase of rennet action is well characterized, butthe secondary phase is less clear. The subject hasbeen reviewed by Dalgleish (1992, 1993); Fox(1984); Fox and McSweeney (1997); Fox andMulvihill (1990); and Fox, O'Connor,McSweeney, Guinee, and O'Brien (1996).

6.1 THE PRIMARY PHASE OF RENNETCOAGULATION

As discussed in Chapter 3, the caseins exist asmicelles stabilized by a surface layer of K-casein. Following the isolation of K-casein in1956, it was shown that this protein is the mi-celle-stabilizing protein and that its stabilizingproperties are destroyed on renneting. Shortlyafterwards, it was shown that K-casein is theonly protein hydrolyzed during the rennet co-agulation of milk and that it is hydrolyzed spe-cifically at the bond Phei05-Meti06 (Figure 6-2).The N-terminal part of the molecule, K-CN fl-105, referred to as para-K-casein, remains at-

tached to the casein micelle, whereas the C-ter-minal part, referred to as the (caseino) macro-peptide (CMP; or glycomacropeptide, since itcontains the carbohydrate moieties of K-casein)is lost in the surrounding aqueous medium. It hasbeen recognized since the end of the 19th cen-tury that small peptides are produced upon ren-neting. As discussed in Chapter 3, there areabout 10 forms of K-casein that differ in sugarcontent; hence, 10 CMPs are produced. All theCMPs are soluble in 2% trichloroacetic acid(TCA) but only the glycosylated forms are sol-uble at higher concentrations of TCA. Thus,TCA-soluble N, or more specifically TCA-soluble sugars (e.g., N-acetyl neuramic acid),can be used to monitor the primary phase of ren-net coagulation (Figure 6-3).

The unique sensitivity of the Phe-Met bond ofK-casein has aroused interest. The dipeptideH.Phe-Met.OH is not hydrolyzed, nor are tri- ortetrapeptides containing a Phe-Met bond. How-ever, this bond is hydrolyzed in the pentapeptideH.Ser-Leu-Phe-Met-AIa-OMe, and reversingthe positions of serine and leucine, to give thecorrect sequence of K-casein, increases the sus-ceptibility of the Phe-Met bond to chymosin.Both the length of the peptide and the sequencearound the Phe-Met bond are important determi-nants of enzyme-substrate interaction. Serinei04appears to be particularly important, and its re-placement by Ala or even L-Ser in the abovepentapeptide renders the Phe-Met bond very re-sistant to hydrolysis by chymosin. Extension of

Enzymatic Coagulation of Milk

CHAPTER 6

Page 114: Cheese Science

the pentapeptide H.Ser.Phe.Met.Ala.Ile.OH

(i.e., K-CN f!04-108) from the N- and/or C-ter-

minal to reproduce the sequence of K-casein

around the chymosin-susceptible bond increases

the efficiency with which the Phe-Met bond is

hydrolyzed by chymosin (Table 6-1). The se-

quence K-CN f98—111 includes all the residues

necessary to render the Phe-Met bond as suscep-

tible to hydrolysis by chymosin at pH 4.7 as it is

in intact K-casein; it is hydrolyzed around

66,000 times faster than the parent pentapeptide

(K-CN f 104-108), with a kcat/KM of about 2 M-1

S"1, which is similar to that for intact K-casein. K-

Casein and the peptide K-CN f98-lll are also

readily hydrolyzed at pH 6.6, but smaller pep-

tides are not.

Figure 6-1 Summary of the rennet coagulation of milk. The primary phase involves enzymatic hydrolysis of K-casein, while the secondary stage involves aggregation of the rennet-altered (paracasein) micelle into a three-dimensional gel network or coagulum.

Casein Para-casein + small peptides(primary, enzymatic phase)

(secondary,non-enzymatic phase)

Coagulum (gel)

rennet

iPyro Glu-Glu-Gln-Asn-Gln-Glu-Gln-Pro-Ile-Arg-Cys-Glu-Lys-Asp-Glu-Arg-Phe-Phe-Ser-Asp-

21Lys-Ile-Ala-Lys-Tyr-Ile-Pro-Ile-Gln-Tyr-Val-Leu-Ser-Arg-Tyr-Pro-Ser-Tyr-Gly-Leu-

41Asn-Tyr-Tyr-Gln-Gln-Lys-Pro-Val-Ala-Leu-Ile-Asn-Asn-Gln-Phe-Leu-Pro-Tyr-Pro-Tyr-

61Tyr-Ala-Lys-Pro-Ala-Ala-Val-Arg-Ser-Pro-Ala-Gln-Ile-Leu-Gln-Trp-Gln-Val-Leu-Ser-

81Asn-Thr-Val-Pro-Ala-Lys-Ser-Cys-Gln-Ala-Gln-Pro-Thr-Thr-Met-Ala-Arg-His-Pro-His-

101 1051 106Pro-His-Leu-Ser-PhetMet-Ala-Ile-Pro-Pro-Lys-Lys-Asn-Gln-Asp-Lys-Thr-Glu-Ile-Pro-

121 He (Variant B)Thr-Ile-Asn-Thr-Ile-Ala-Ser-Gly-Glu-Pro-77zr- Ser-Thr -Pro-Thr - -GIu-Ala-Val-Glu-

Thr (Variant A)141 Ala (Variant B)Ser-Tfer-Val-Ala-Thr-Leu-Glu- -SerP - Pro-Glu-Val-Ile-Glu-Ser-Pro-Pro-Glu-Ile-Asn-

Asp (Variant A)161 169Thr-Val-Gln-Val-Thr-Ser-Thr-Ala-Val.OH

Figure 6-2 Amino acid sequence of K-casein, showing the principal chymosin cleavage site (i); oligosaccha-rides are attached at some or all of the threonine residues shown in italics.

Page 115: Cheese Science

The Phe and Met residues in the chymosin-sensitive bond of K-casein are not intrinsicallyessential for chymosin action. There are numer-ous Phe and a substantial number of Met resi-dues in all milk proteins. In porcine and humanK-caseins, the chymosin-sensitive bond is Phe-Ue, while in rat and mouse K-caseins, it is Phe-Leu; yet, these proteins are readily hydrolyzedby calf chymosin, although more slowly thanbovine K-casein. In contrast, porcine milk is co-agulated more effectively than bovine milk byporcine chymosin, indicating that unidentifiedsubtle structural features influence chymosin ac-tion. Peptides in which Phe is replaced by Phe(NO2) or cyclohexylamine are also hydrolyzedby chymosin, although less effectively thanthose with a Phe-Met bond. Oxidation of Met106reduces kcat/KM roughly tenfold, but substitutionof He for Met increases it about threefold.

A genetically engineered mutant of K-caseinin which Met106 was replaced by Phei06 (i«e., thechymosin-sensitive bond was changed fromPheio5-Met106 to Pheios-Pheioe) was hydrolyzed1.8 times faster by chymosin than naturalK-casein. These findings suggest that the se-quence around the Phe-Met bond, rather than theresidues in the bond itself, contains the impor-tant determinants of hydrolysis by chymosin.The particularly important residues are Seri04,

the hydrophobic residues Leuioa and Ileios* atleast one of the three histidines (residues 98,100, or 102, as indicated by the inhibitory effectof photo-oxidation), and LySm. Studies onchemically or enzymatically modified peptideanalogues of K-CN f98-112 indicated the rela-tive importance of residues in the sequences of98-102 and 111-112. It has been suggested thatthe sequence Leuios to Ilcios of K-casein, whichprobably exists as an extended (3-structure, fitsinto the active site cleft of acid proteinases. Thehydrophobic residues Leuios, Prices, Metioe, andHe iog are directed toward hydrophobic pocketsalong the active site cleft, while the hydroxylgroup of Serio4 forms part of a hydrogen bondwith some counterpart in the enzyme. It hasbeen proposed that the sequences 98-102 and109-111 form (3-turns around the edges of theactive site cleft of the enzyme. This conforma-tion is stabilized by Pro residues at positions 99,101, 109, and 110. The three His residues at po-sitions 98, 100, 102, and LySm are probably in-volved in electrostatic bonding between enzymeand substrate; none appears to have a predomi-nant role. Lys 112 appears not to be important inenzyme-substrate binding as long as Lysm ispresent.

The significance of electrostatic interactionsin chymosin-substrate complex formation is in-

Time from rennet addition

Figure 6-3 Release of nitrogen soluble in 2% (O) or 12% (•) TCA by rennet from casein in milk.

TC

A-s

olub

leN

Page 116: Cheese Science

dicated by the effect of added NaCl on the rennetcoagulation time (RCT) of milk. Addition ofNaCl up to 3 mM reduces RCT but higher con-centrations have an inhibitory effect. It is claimedthat the effect of NaCl is on the primary, enzy-matic phase rather than on the aggregation of ren-net-altered micelles. Increasing ionic strength(0.01-0.11) reduced the rate of hydrolysis of K-CN f HiS98-LySi i m u in a model system; the effectbecame more marked as the reaction pH was in-creased, but it was independent of ion type.

As well as serving to elucidate the importanceof certain residues in the hydrolysis of K-caseinby chymosin, small peptides that mimic or areidentical to the sequence of K-casein around thePhe-Met bond are very useful substrates for de-termining the activity of rennets in absoluteunits, that is, independent of variations in thenonenzymatic phase of coagulation of differentmilks. Standard methods for such quantificationhave been developed and chromogenic peptidesubstrates are available commercially. Since thespecific activity of different rennets on thesepeptides varies, methods for quantifying the pro-

portions of acid proteinases in commercial ren-nets have been proposed.

6.2 RENNET

Several proteinases will coagulate milk undersuitable conditions but most are too proteolyticrelative to their milk-clotting activity (MCA).Consequently, they hydrolyze the caseins in thecoagulum too quickly, causing a reduced cheeseyield. (MCA is the inverse of RCT; i.e., MCA =1/RCT.) Excessive proteolysis or incorrectspecificity may also lead to defects in the flavor(especially bitterness) and texture of the cheese.Although plant proteinases appear to have beenused as rennets since prehistoric times, gastricproteinases from calves, kids, or lambs havebeen traditionally used as rennets, with very fewexceptions.

Animal rennets are prepared by extracting thedried (usually) or salted gastric tissue (referredto as veils) with 10% NaCl and activating andstandardizing the extract. Standard calf rennetcontains about 60-70 RU/ml and is preserved by

Table 6-1 Kinetic Parameters for Hydrolysis of K-Casein Peptides by Chymosin at pH 4.7

Peptide

S.F.M.A.I.S.F.M.A.I.P.

S.F.M.A.I. P.P.S.F.M.A.I. P.P.K.

LS.F.MAI.LS.F.M.A.I.P.LS.F.M.A.I.P.P.

LS.F.MAI. P.P.K.LS.F.M.A.I.P.P.K.K.

H.LS.F.M.A.I.P.H.LS.F.M.A.I

H.P.H.P.H.LS.F.M.A.I.P.P.K.

K-Caseinb

LS.F.(NO2)Nle A.L.OMe

a pH 6.6.b pH 4.6.

Sequence

104-108104-109104-110104-111103-108103-109103-110103-111103-112102-108101-10898-1 1 198-1 11a

kcat (S-1)

0.331.051.570.75

18.338.143.333.630.216.033.566.246.2a

2-2012.0

MmM)

8.509.206.803.200.850.690.410.430.460.520.340.0260.029a

0.001-0.0050.95

kcat/KM(s-1mM-1)

0.0380.1140.2310.239

21.655.1

105.178.365.330.8

100.225091621a

200-2,00012.7

Page 117: Cheese Science

making the extract up to 20% NaCl and addingsodium benzoate or sodium propionate. A rennetunit (RU) is the amount of rennet activity thatwill coagulate 10 ml of milk (usually low-heatskim milk powder reconstituted in 0.01% CaCl2

and perhaps adjusted to pH 6.5) in 100 s.Chymosin (an aspartyl acid proteinase, i.e., aproteinase with two aspartic acid residues at theactive site and with a pH optimum of 2-4) repre-sents more than 90% of the MCA of good-qual-ity veal rennet, the remaining activity being dueto pepsin. As the animal ages, especially whenfed solid food, the secretion of chymosin de-clines and that of pepsin increases.

Like many other animal proteinases, chy-mosin is secreted as its zymogen, prochymosin,which is autocatalytically activated on acidifica-tion to pH 2-^ by removal of a 44-residue pep-tide from the N-terminal of the zymogen (seeFoltmann, 1993).

Chymosin is well characterized at the molecu-lar level (see Chitpinityol & Crabbe, 1998;Foltmann, 1993). The enzyme, which was crys-tallized in the 1960s, is a single-chain polypep-tide that contains about 323 amino acid residuesand has a molecular mass of 35,600 Da. Its pri-mary structure has been established, and a con-siderable amount of information is available onits secondary and tertiary structures. The mol-ecule exists as two domains separated by an ac-tive site cleft in which the two catalyticallyactive aspartyl groups (Asp32 and Asp215) are lo-cated.

Calf rennet contains three chymosin isoen-zymes, principally A and B, with lesser amountsof C. Chymosins A and B are produced from thecorresponding zymogens, prochymosins A andB, but chymosin C appears to be a degradationproduct of chymosin A that lacks three residues,ASp244-PhC246. The specific activity of chymosinA, B, and C is 120, 100, and 50 RU/mg, respec-tively. Chymosins A and B differ by a singleamino acid substitution, Asp and GIy, respec-tively, at position 244 and have an optimum pHat 4.2 and 3.7, respectively.

The properties of different rennets are dis-cussed in Section 6.8.

6.3 FACTORS THAT AFFECT THEHYDROLYSIS OF K-CASEIN AND THEPRIMARY PHASE OF RENNETCOAGULATION

The hydrolysis of K-casein is influenced bymany factors, some of which are discussed be-low. While many factors influence both the pri-mary and secondary stages, the effects on eachare discussed separately.

• The pH optimum for chymosin and bovinepepsin on small synthetic peptides is about4.7 but is 5.3-5.5 on K-CN f HiS98-LySm71I2.Chymosin hydrolyzes insulin, acid-dena-tured hemoglobin and Na-caseinate opti-mally at pH 4.0, 3.5, and 3.5, respectively.The pH optimum for the first stage of rennetaction in milk is about 6.0 with 40C or3O0C.

• The influence of ionic strength on the pri-mary phase of rennet coagulation is dis-cussed in Section 6.1.

• The optimum temperature for the coagula-tion of milk by calf rennet at pH 6.6 isaround 450C. Presumably, the optimum forthe hydrolysis of K-casein is around thisvalue. The temperature coefficient (Q10) forthe hydrolysis of K-casein in solutions ofNa-caseinate is about 1.8; the activation en-ergy, Ea, is about 40,000 J-mol'1; and theactivation entropy, AS, is about -90!•Kr^moH. Generally similar values havebeen reported for the hydrolysis of isolatedK-casein by chymosin.

• Heat treatment of milk at temperaturesabove 650C adversely affects its rennet co-agulability. If the heat treatment is verysevere (> 9O0C for 10 min), the milk failsto coagulate upon renneting. Althoughchanges in salts equilibria are contributoryfactors, the principal causative factor is in-termolecular disulphide bond formation be-tween K-casein and (3-lactoglobulin and/ora-lactalbumin. Both the primary and espe-cially the secondary phases of rennet actionare inhibited in heated milk, as reflected by

Page 118: Cheese Science

the marked decrease in the curd-firmingrate and in the strength of the resulting gel.The adverse effects of heating can be re-versed by acidification to pH values in theregion 6.6-6.0 before or after heating or bythe addition of CaCl2 (which causes a re-duction in pH). The secondary, rather thanthe primary, phase of rennet action prob-ably benefits from these treatments.

6.4 THE SECONDARY(NONENZYMATIC) PHASE OFCOAGULATION AND GEL ASSEMBLY

Hydrolysis of K-casein by chymosin or simi-lar enzymes during the primary phase of rennetaction releases the highly charged, hydrophilicC-terminal segment of K-casein (macropeptide),as a result of which the zeta potential of thecasein micelles is reduced from -10/-20 to -5/-7 mV and the protruding peptides (hairs) areremoved from their surfaces, thus destroying theprincipal micelle-stabilizing factors (electro-static and steric) and their colloidal stability.When roughly 85% of the total K-casein hasbeen hydrolyzed, the stability of the micelles isreduced to such an extent that when they collide,they remain in contact and eventually build intoa three-dimensional network, referred to as a co-agulum or gel (Figure 6—4). Gel formation is ac-companied by sharp increases in viscosity andelastic shear modulus, G', which is a measure ofgel firmness (Figure 6-5; see Section 6.6.2). Re-ducing the pH or increasing the temperaturefrom the normal values (~ 6.6 and ~ 310C, re-spectively) permits coagulation at a lower levelof K-casein hydrolysis. Although the precise re-actions involved in aggregation are not known,the kinetics of aggregation have been described.

The assembly of rennet-altered micelles into agel has been studied using various forms of vis-cometry, electron microscopy, and light scat-tering. Viscosity measurements show that theviscosity of renneted milk remains constant ordecreases slightly during a period equivalent toroughly 60% of the visually observed RCT (Fig-

ures 6-4 and 6-5). It has been suggested that thedecrease in viscosity is due to a decrease in thevoluminosity of the casein micelles followingrelease of the macropeptides, which form a"hairy layer" about 10 nm thick (Figure 6-4).The decrease in micelle size has been confirmedby quasi-elastic light scattering.

It is generally agreed that following the initiallag period, the viscosity increases exponentiallyup to the onset of visual coagulation or gelation,that is, 100% RCT (Figure 6-5). The gelationprocess, generally referred to as the secondaryphase of rennet coagulation, involves initiallythe formation of chains and clumps of micellesand leads eventually to the formation of a net-work of partly fused micelles. During the first60% of the visually observed RCT, the micellesexist as individual particles; the primary enzy-matic reaction is about 85% complete at 60% ofthe visual RCT. Between 60% and 80% of theRCT, the rennet-altered micelles begin to aggre-gate steadily, with no sudden change in the typeor extent of aggregation. Small chain-like aggre-gates, rather than clumps, form initially (Figure6-6). At 100% of the RCT, most of the micelleshave aggregated into short chains, which thenbegin to aggregate to form a network. Initially,most micelles are linked by bridges (655 nmlong and 40 nm wide) and do not touch. An un-expectedly large proportion of the surface of theparticipating micelles is involved in the bridg-ing, indicating a large amount of material, theorigin of which is unknown. Although the mi-celles themselves appear the most probablesource, if they are the source, micellar rearrange-ment would be necessary. No change is ob-served in the size, surface structure, or generalappearance of the micelles up to 60% of the RCT(i.e., lag phase). Thus, if a micellar rearrange-ment is a prerequisite for aggregation, it mustoccur during the latter half of the RCT. Linkageof the rennet-altered micelles probably occurs atdefinite surface sites.

Aggregation of the rennet-altered micellescan be described by the von Smoluchowskitheory for diffusion-controlled aggregation ofhydrophobic colloids when allowance is made

Page 119: Cheese Science

for the need to produce, by enzymatic hydroly-sis, a sufficient concentration of particles ca-pable of aggregating (i.e., casein micelles inwhich > 97% of the K-casein has been hydro-Iyzed). The diffusion of the particles is rate lim-

iting and is determined by the random fruitfulcollision of particles (rennet-altered micelles).The rate of aggregation is not consistent with abranching process model, since the micellarfunctionality is 1.8 whereas an average function-

Figure 6—4 Schematic representation of the rennet coagulation of milk, (a) Casein micelles with intact K-caseinlayer being attacked by the chymosin (C); (b) micelles partially denuded of K-casein; (c) extensively denudedmicelles in the process of aggregation; (d) release of macropeptides (^) and changes in relative viscosity (Q)during the course of rennet coagulation.

% of visually observed clotting time

Rfil

fiasf

t n

f m

ar.Y

rmfim

iilfis

n xei

a b c

d

Page 120: Cheese Science

ality greater than 2 is required for network for-mation.

According to Dalgleish (1980), the overallrennet coagulation of milk can be described bycombining three factors:

1. proteolysis of K-casein, which may be de-scribed by Michaelis-Menten kinetics

2. the requirement that around 97% of the K-casein on a micelle be hydrolyzed beforeit can participate in aggregation

3. aggregation of paracasein micelles via avon Smoluchowski process

The overall clotting time tc is the sum of theenzymatic phase and the aggregation phase:

.=W"^4^} + S0+^(^-l)^max V 1 ac J %iax ZKs^Q \ 1V10 )

where: Km and Vmax are the Michaelis-Menten

Time from rennet addition, min

Figure 6-5 Rheological changes in milk during rennet coagulation under quiescent conditions. (A) Phase angle(•) and viscosity (O); (B) elastic modulus (•) and loss modulus (Q).

Gel

mod

uli,

PaPh

ase

angl

e, 6

, °

Visc

osity

, Pa

s

(A)

(B)

Page 121: Cheese Science

parameters, occ is the extent of K-casein hydroly-sis, S0 is the initial concentration of K-casein, ks

is the rate constant for aggregation, C0 is the con-centration of aggregating material, Mcrit is theweight average molecular weight at tc («10 mi-cellar units), and M0 is the weight average mo-lecular weight at t0.

Darling and van Hooydonk (1981) proposedan alternative model for rennet coagulation,again by combining Michaelis-Menten enzymekinetics with von Smoluchowski aggregationkinetics. The stability factor in vonSmoluchowski's theory is considered as a vari-able determined by the concentration of un-hydrolyzed surface K-casein. The coagulationtime, tc, is given by

,.^^^.^^^±-±]

where V is the velocity of enzymatic hydrolysisof K-casein, S0 is the initial concentration of K-casein, Cm is a constant relating the stability ofthe casein micelle to K-casein concentration, W0

is the initial stability factor for casein micelles,W0 is the initial concentration of casein micelles,and nc is the concentration of casein aggregatesat the observed clotting time tc. It is claimed thatthis theoretical model explains the experimen-tally observed influence of protein concentra-tion, enzyme concentration, and temperature onRCT and the occurrence of a lag phase equal to60% of RCT.

Figure 6-6 Schematic representation of the progress of micelle aggregation during the rennet coagulation ofmilk. Aggregation of rennet-altered micelles results in the formation aggregates that fuse to form chain-likestructures, and these eventually overlap and cross-link to form a three-dimension casein network or gel after atime greater than the visual rennet coagulation time.

Gel suitable forcutting

% of visual rennet coagulation time

Deg

ree

of c

oagu

latio

n

Page 122: Cheese Science

With milk of normal concentration, about 90%of the micelles are incorporated into the curd at100% of the visual RCT, but only about 50% areincorporated in a fourfold concentrate when thesame level of rennet is used. The micelles that are"free" at or after the RCT may react differentlyfrom those free prior to RCT. That is, before vi-sual coagulation, all micelles are freely dispersedin the serum and can aggregate randomly, butonce a gel matrix has started to form, free micellesmay react either with the gel matrix or with otherfree micelles. Therefore, a gel assembly may beregarded as a two-stage process, and the proper-ties of the final curd may be affected considerablyby the amount of casein free during the RCT.Since this amount is particularly high in concen-trated milks, it may explain the coarser structureof curd made from these milks.

Measurements made using the Instron Uni-versal Testing machine showed that the rate offirming of a renneted milk gel as a function oftime has two maxima (Stony & Ford, 1982): un-der the experimental conditions used, firmingwas first observed about 2.5 min after the visualassessment of clotting, and the firming rate in-creased to a maximum 12-15 min after clotting.The rate decreased over the next 10-15 min toabout 80% of the maximum value, after which iteither remained constant or increased slightlyfor a further 15 min and thereafter decreasedsteadily. However, this two-stage gel firmingprocess has not been observed when dynamicrheometry, a very sensitive technique, is used.

Based on viscometric data, Tuszynski (1971)suggested that gel assembly is a two-stage pro-cess consisting of what he called "flocculation"and "gelation," but the meaning of these terms isnot clear. Turbidity experiments also suggest atwo-stage gel assembly process (Surkov,Klimovskii, & Krayushkin, 1982), althoughagain the terminology is not very clear:

where E is the enzyme, S is the substrate, P is thereaction product, P* is the paracasein micellewith transformed quaternary structure, and Pn isthe gelled micelle aggregate. The first two stepsare the Michaelis-Menten model for the primary,enzymatic phase and are essentially as proposedby Payens, Wiersma, and Brinkhuis, 1977):

k, k2

E + S " •- ES >• E + P1 + Mk_,

where P\ is para-K-casein and M is amacropeptide. Payens et al. (1977) suggestedthat the second, nonenzymatic phase may berepresented by

Jl_

where i is any number of the aggregating particlePi.

Surkov et al. (1982) suggested that the en-zyme-altered micelles (paracasein micelle P)undergo a cooperative transition in quaternarystructure to yield clot-forming particles (P*)with a rate constant kc and with E0 = 191 kJmoh1 and QIQ-C = 16. These values are close tothose reported by Tuszynski (1971).

The sites involved in the aggregation processare not known. Following reduction of the mi-cellar zeta potential by proteolysis of K-casein,linkage of particles is facilitated. Interparticlelinkage could be via calcium bridges and/or hy-drophobic interactions (which the marked tem-perature-dependence of the secondary phasewould indicate). Changes in the surface hydro-phobicity of casein micelles during rennetinghas been demonstrated through changes in thebinding of the fluorescent marker 8-anilinonaphthalene-1-sulfonate (Iametti, Giangiacomo,Messina, & Bonomi, 1993; Peri, Pagliarini,Iametti, & Bonomi, 1990). The hydrophobicamino terminal segment (residues 14-24) of

Page 123: Cheese Science

(Xsi-casein appears to be important in the estab-lishment of a rennet curd structure. It has beensuggested that the matrix of young cheese curdconsists of a network of ocsi-casein moleculeslinked together via hydrophobic patches that ex-tends throughout the cheese structure. The soft-ening of texture during the early stages of ripen-ing is considered to be due to breaking of thenetwork upon hydrolysis of the Phe23-Phe24 bondof ocsi-casein. Modification of histidyl, lysyl, andarginyl residues in K-casein inhibits the second-ary phase of rennet coagulation, suggesting thata positively-charged cluster on para-K-casein in-teracts electrostatically with unidentified nega-tive sites. In native micelles, this positive sitemay be masked or covered by the macropeptidesegment of K-casein but becomes exposed andreactive when this peptide is released.

Following the RCT, the network appearancebecomes gradually more apparent. The strandsof the network, which are more or less in paral-lel, are roughly 5 micelles thick and 10 micellediameters apart. The bridges between the mi-celles contract slowly, forcing the micelles intocontact and eventually causing them to fuse par-tially. The fate of the bridging material upon mi-celle fusion has not been explained, but it maybe that its disappearance is responsible for thesecond maximum in the firming rate time curveand for the reported flocculation and gelationstages in the gel assembly process.

Normally, the rate of an enzymatic reactionincreases linearly with enzyme concentration,within certain limits. In the case of rennet coagu-lation, RCT is inversely related to enzyme con-centration, as expressed by the formula

Et0 = k

where E = enzyme concentration and tc is theRCT.

This equation, which assumes that visuallyobserved coagulation is dependent only on theenzymatic process, has been modified to takeaccount of the duration of the secondary, nonen-zymatic phase:

E(tc-x) = k

where x is time required for the coagulation ofthe enzymatically altered casein micelles andtc-x is the time required for the enzymatic stage.Rearrangement of this equation results in a moreconvenient form:

tc = k ( l / E ) + x

which is valid within a certain range of rennetconcentrations and under certain conditions oftemperature and pH. A very good linear relation-ship exists between clotting time and the recip-rocal of enzyme concentration (Figure 6-7).

The coagulation equations developed byDalgleish (1980) and by Darling and vanHooydonk (1981) might be regarded as greatlyrefined versions of these simpler equations andreduce to them on first approximations. Rennetclotting time (tc) has also been expressed(Payens and Wiersma, 1977) by the equation:

, ^ 1 2c \\kVV 7S'max

where ks, the diffusion-controlled flocculationrate constant according to von Smoluchowski'stheory (nonenzymatic phase), is proportional tothe concentration of reactive (coagulable) par-ticles (proteolyzed micelles) and hence to en-zyme concentration; Fmax is the maximum veloc-ity in Michaelis-Menten kinetics (enzymaticphase) and is proportional to enzyme concentra-tion; and c is a constant that depends on the ex-perimental conditions.

6.5 FACTORS THAT AFFECT THENONENZYMATIC PHASE OF RENNETCOAGULATION

The coagulation of renneted micelles is verytemperature dependent (Q10 ~ 16), and bovinemilk does not coagulate at less than around 180Cunless the Ca2+ concentration is increased. Themarked difference between the temperature de-pendence of the enzymatic and nonenzymatic

Page 124: Cheese Science

1/E

Figure 6-7 Relationship between enzyme con-centration (E) and rennet coagulation time.

phases of rennet coagulation has been exploitedin studies on the effects of various factors on therennet coagulation of milk—studies done as partof attempts to develop a system for the continu-ous coagulation of milk for cheese or rennetcasein manufacture and to discover possible ap-plications of immobilized rennets. The very hightemperature dependence of rennet coagulationsuggests that hydrophobic interactions are im-portant.

Coagulation of rennet-altered micelles de-pends on a critical concentration of Ca2+, whichmay act by cross-linking rennet-altered micelles,possibly via serine phosphate residues or simplyby charge neutralization. Colloidal calciumphosphate is also essential for coagulation butcan be replaced by increased Ca2+. Partial enzy-matic dephosphorylation of casein, which re-duces micellar charge, reduces coagulability; in-teraction of casein micelles with various cationicspecies predisposes them to coagulation by ren-net and may even coagulate unrenneted mi-celles. Chemical modification of histidine,lysine, or arginine residues inhibits coagulation,presumably by reducing micellar positivecharge.

The apparent importance of micellar charge inthe coagulation of rennet-altered micelles sug-gests that pH should have a major influence on

the secondary phase of coagulation. Reductionin pH in the range 6.6-6.0 is accompanied by in-creases in the rates of the enzymic and coagula-tion reactions, reductions in gelation time (timefor gel onset) and the degree of K-casein hy-drolysis necessary for the onset of gelation (e.g.,from « 97% to « 80% of total K-casein), and in-creases in the curd-firming rate and firmness af-ter a given renneting time. Although it is claimedthat pH has essentially no effect on the coagula-tion process, the rate of firming of the resultantgel is significantly increased upon reducing thepH (Figure 6-8).

The rate of firming of renneted milk gels isinfluenced by the type of rennet, especially un-der unfavorable conditions, such as high pH orlow Ca2+. Perhaps differences in the firming ratereflect the effect of pH on rennet activity or per-haps some general proteolysis by rennet substi-tutes.

Heat treatment of milk under conditions thatdenature p-lactoglobulin and promote its inter-action with K-casein via sulfydryl-disulfide in-teraction adversely affects all aspects of rennetcoagulation but especially the buildup of a gelnetwork (Figure 6-9) (van Hooydonk, deKoster,& Boerrigter, 1987). Presumably, the attach-ment of denatured p-lactoglobulin to the surfaceof the casein micelles (as is evident from elec-tron micrographs of casein micelles) preventstheir aggregation in a form capable of buildingup a gel network.

6.6 MEASUREMENT OF RENNETCOAGULATION PROPERTIES

The rennet gelation of milk under quiescentconditions involves the conversion of milk froma colloidal dispersion of stable micelles to a net-work (gel) of aggregated paracasein micelles,which forms a continuous phase, entrappingmoisture and fat globules in its pores. The gelbecomes more elastic and firm with time (i.e.,upon aging). The transformation is accompaniedby a number of physicochemical changes, in-cluding hydrolysis of K-casein, with a concomi-tant increase in the concentration of the gly-

Ren

net c

oagu

latio

n tim

e

Page 125: Cheese Science

comacropeptide; aggregation of the sensitizedparacasein micelles; increases in viscosity andelasticity; and a decrease in the ratio of the vis-cous to elastic character of the milk. Suchchanges may also alter some of the physicalproperties of the milk, such as light reflectanceand thermal conductivity.

Numerous methods, the principles of whichare based on detection of one or more of theabove changes, have been developed to measurethe rennet coagulation characteristics of milk orthe activity of rennets. Owing to the commercialimportance of gel formation in milk as a meansof recovering milk fat and casein in the form ofcheese curd, most methods measure gel forma-tion (also referred to as curd formation or rennetcoagulability)—that is, the combined first andsecond stages—but some specifically monitorthe hydrolysis of K-casein. Various terms or de-scriptors, some of which are used interchange-ably, are employed to describe the rennet coagu-lation of milk. These are defined below:

Aggregation. The joining of particles, suchas micelles, by various types of electrostaticor hydrophobic bonds. The aggregates arevisible by electron microscopy.Coagulation or flocculation. The collisionand joining of aggregates, especially undernonquiescent conditions, to form floes vis-ible to the naked eye.Gelation. The aggregation of particles (e.g.,micelles or aggregates of micelles) to formparticulate strands whose particles undergolimited touching and that eventually form agel network.Elasticity. The ability of the gel to recover,instantaneously, its original shape and di-mensions after removal of an applied stress.Viscoelastic materials, such as a rennetedmilk gel, are elastic at relatively smallstrains (e.g., 0.025, which is much less thanfracture strain). In this region of strain,known as the linear viscoelastic stress-strain region, the strain is directly propor-

Time after rennet addition, min

Figure 6-8 Development of elastic shear modulus (G', equivalent to curd firmness) in rennet-treated, high-protein (-180 g/kg) milk retentate obtained from skim milk heated to 10O0C for 12Os and renneted at pH 6.67(O), 6.55 (•), 6.45 (Q), 6.3 (•), 6.15 (A), and 6.0 (A).

Elas

tic s

hear

mod

ulus

, G1, k

Pa

Page 126: Cheese Science

tional to the applied stress, and the material(e.g., a section of a gel strand that bears theapplied stress and is strained) recovers itsoriginal dimensions immediately upon re-moval of the stress.

• Viscosity. The physical property of a gelgiven by the ratio between the stress andstrain rate.

• Curd firmness, curd strength, or curd ten-sion. The stress required to cause a givenstrain or deformation. (Curd tension is aterm frequently used to express the firm-ness of formed gels.)

Types of measurement used to evaluate gel-forming characteristics include

• measurement of flocculation time undernonquiescent conditions (e.g., rennet co-agulation test)

dynamic measurement of the viscous dragof gelling milk by suspending a pendulumin the milk and determining the tilt of thependulum over time using instruments suchas the Thromboelastrograph, Formagraph,and Gelometerdynamic measurement of the ability of gel-ling milk to transmit a pressure by using, forexample, a hydraulically operated oscillat-ing diaphragm apparatusmeasurement of the apparent viscosity ofgelled milk after a given time at a fixedshear rate (e.g., by using various types ofrotational viscometers) or, alternatively,measurement of the firmness of the gel us-ing various types of penetrometerdynamic measurement of parameters suchas viscosity, elastic shear modulus (G % lossmodulus (G"), and phase angle (8) by ap-plying a low-amplitude oscillating strain or

Time after rennet addition, min

Figure 6-9 Release of casein macropeptides (•, Q) and changes in the viscosity (A, A) and curd firmness (•,O) in skim milk unheated (closed symbols) or heated at 950C for 15 s (open symbols). Arrows indicate the timeat which cutting is initiated during cheese manufacture.

Arbi

trary

uni

ts

Page 127: Cheese Science

stress to the milk sample using, for ex-ample, controlled strain or controlled stressrheometers

• dynamic measurement of some physicalproperties of the gelling milk in the cheesevat using special probes, including thermalconductivity (using a hot wire probe) andreflectance of near infrared light (using anear infrared diffuse reflectance probe)

Some of the more commonly used laboratoryand online methods for monitoring the rennet-induced gelation of milk are described below.

6.6.1 Measurement of the Primary Phase ofRennet Coagulation

The primary phase of rennet action may bemonitored by measuring the formation of eitherproduct, that is, para-K-casein or the CMP. Para-K-casein may be measured by SDS-polyacryla-mide gel electrophoresis (PAGE), which is slowand cumbersome, or by ion-exchange high-per-formance liquid chromatography (HPLC). TheCMP is soluble in TCA (2-12% depending onits carbohydrate content) and may be quantifiedby the Kjeldahl method or more specifically bydetermining the concentration of N-acetylneura-minic acid (Figure 6-3) or by RP-HPLC. Theactivity of rennets can be determined easily us-ing chromogenic peptide substrates, a number ofwhich are available. The latter method is gener-ally used as a research tool to study rennet char-acteristics and/or the kinetics of the primaryphase of rennet coagulation.

6.6.2 Methods for Assessing Coagulation,Gel Formation, and/or Curd Tension

Measurement of Rennet CoagulationTime

The simplest laboratory method for measur-ing the overall rennet coagulation process is tomonitor the time between the addition of a mea-sured amount of diluted rennet to a sample ofmilk in a temperature-controlled water-bath at,

for example, 3O0C and the onset of visual coagu-lation. If the coagulating activity of a rennetpreparation is to be determined, a "reference"milk, such as low-heat milk powder reconsti-tuted in 0.01% CaCl2 and perhaps adjusted to acertain pH (e.g., 6.5), should be used. A standardmethod has been published (International DairyFederation [IDF], 1992), and a reference milkpowder may be obtained from Instirut Nationalde Ia Recherche Agronomique, Poligny, France.If the coagulability of a particular milk is to bedetermined, the pH may or may not be adjustedto a standard value (e.g., 6.55) to reflect thatwhich is typical at setting (rennet addition) dur-ing cheese manufacture.

The coagulation point may be determined byplacing the milk sample in a bottle or tube that isrotated in a water-bath (Figure 6-10); the fluidmilk forms a film on the inside of the rotatingbottle or tube, but floes of protein form in thefilm upon coagulation. The rennet coagulationtime (RCT) provides a very good index of thegelation potential of milk; a low RCT usuallyindicates potentially good gel formation andhigh gel strength after a given renneting time.The method is simple and enables the accuratemeasurement of several samples simulta-neously. It has been used to accumulate much ofthe extensive information reported in the scien-tific literature on the effects of various process-ing parameters on the rennet coagulability ofmilk. However, in contrast to cheese manufac-ture, where milk is renneted under quiescentconditions to ensure gel formation, this methoddetermines the time for coagulation (i.e., aggre-gation and flocculation) of the paracasein underagitation.

Nondynamic Assessment of ViscosityCurd Firmness Tests

The apparent viscosity and firmness of the co-agulum may be measured using various types ofviscometers and penetrometers, respectively.However, use of these instruments permits mea-surement at only a single point in time, which isa serious limitation in kinetic studies, and it also

Page 128: Cheese Science

requires meticulous test conditions, since curdstrength increases with time after renneting.

Dynamic Curd Firmness Tests

Various instruments, involving different prin-ciples, have been developed to monitor changescontinuously throughout the gelation process.These are discussed below.

Formagraph. The most popular of the dy-namic measuring instruments, although it is notwidely used, is the Formagraph (e.g., Type1170, Foss Electric, Denmark), a diagram ofwhich is shown in Figure 6-11. The apparatusconsists of

• an electrically heated metal block• a sample rack with cavities (usually 10) into

which sample cuvettes fit• a set of pendulums, each having an arm

with an attached mirror• an optical source that beams a light ray onto

the mirror attached to each pendulum• a chart recorder with photosensitive paper

onto which the reflected light from eachmirror is directed

Samples of milk to be analyzed are placed inthe cuvettes and tempered to the desired tem-perature (typically 310C) in the heating block.Rennet is then added, and the cuvettes are re-placed in the instrument so that a loop-shapedpendulum is suspended in each sample. Themetal block is moved back and forth, creating a"drag" on the pendulum in the milk. A flashinglight is directed onto the mirror on the arm ofeach pendulum and reflected onto photosensi-tive paper, creating a mark. While the milk isfluid, the viscosity is low and the drag on thependulum is slight. It scarcely moves from itsvertically suspended "zero-time" position, andhence a single straight line appears on the paper.As the milk coagulates, its viscosity increasesand the pendulum is dragged from its zero-timeposition, resulting in bifurcation of the trace.The rate at and extent to which the arms of thetrace diverge are indicators of the gel-formingcharacteristics of the milk. A typical trace (seeFigure 6-11) may be used to calculate the fol-lowing parameters:

• The rennet coagulation time (r) is the time(in minutes) from rennet addition to the on-

Figure 6-10 Schematic of apparatus for visual determination of the rennet coagulation time of milk.

Milk sample

Water-bath at 3O0C

Page 129: Cheese Science

set of gelation (i.e., point where the tracebegins to fork).

• &2o, essentially the inverse of the curd-firm-ing rate, is the time from the onset of gela-tion until a firmness corresponding to atrace width of 20 mm is obtained.

• ah which represents the curd firmness attime t after rennet addition, is the tracewidth (in millimeters) at time t.

Good gel-forming properties are character-ized by a relatively rapid coagulation time (low r

Figure 6-11 (a) Schematic representation of the Formagraph apparatus for determining the rennet coagulationof milk, (b) Typical formagram. The asterisk represents the point of rennet addition, r is the rennet coagulationtime, k2Q is the time required from coagulation for the arms of the formagram to bifurcate by 20 mm, and ^30 is theextent of bifurcation 30 min after rennet addition (the approximate time at which the coagulum is cut incheesemaking).

Oscillating heating block

Damper

Mirror

Lighl Hash

Photographicchart paper

(a)

(b)

MILK

Page 130: Cheese Science

value), high-curd firming rate (low k20 value),and a high-curd firmness or strength after agiven renneting time (high ^30 value). Typicalvalues for these parameters for a pasteurizedmidlactation milk (3.3%, w/w, protein) rennetedunder normal conditions (rennet dosage, «16RU/L; pH, 6.55; temperature, 310C) are r = 5.5min, A2O = 11 min, and ^30 = 48.5 mm.

While the latter parameters have no preciserheological significance, the Formagraphmethod offers many advantages over the RCTtest:

• The method simulates gel formation duringcheesemaking.

• The results of the assay are less subjective,being independent of operator judgment.

• The test parameters provide more informa-

tion on the changes in curd strength overtime.

However, production of the instrument hasbeen discontinued.

Hydraulic ally oscillating diaphragm. In ahydraulically operated oscillating diaphragm ap-paratus, a sample of milk is placed between twodiaphragms (Figure 6-12) and rennet is added.One diaphragm (the transmitting diaphragm) ismade to vibrate through the cyclical applicationof hydraulic pressure. When the milk is liquid,the effect of the vibration is dissipated rapidlyand does not affect the receiving diaphragm.When a gel is formed, the vibrations emitted bythe transmitting diaphragm reach the receivingdiaphragm, causing it to vibrate. These vibra-tions are detected and quantified by a suitable

Pulsed pressure

Oscillating diaphragm

Sensor

Receiver Transmitter

firmness suitablefor cutting

Coagulation time

Time from rennet addition

Figure 6-12 Schematic representation of a pressure transmission apparatus for measuring the rennet coagulationtime of milk and the strength of the resulting gel.

Am

plitu

de o

f pu

lse

(a)

(b)

Page 131: Cheese Science

sensing device. An output generally similar tothat of the Formagraph is obtained, from whichthe coagulation time and a measure of gelstrength can be determined.

Low-Amplitude Stress or Strain Rheometry.Since the 1980s, several controlled-strain rhe-ometers (e.g., Bohlin VOR, Bohlin Rheologi,Sweden; Rotovisco RV 100/CV 100, HaakeBucchler Instruments, United States) and con-trolled-stress rheometers (e.g., Bohlin CS,Bohlin Rheologi, Sweden; Cari-med CSL2, TAInstruments, United States; Rheometric Scien-tific SR5, Rheometric Scientific Inc, UnitedStates) have been used increasingly as researchtools for the continuous measurement of the vis-coelastic properties of renneted milks as a func-tion of time from rennet addition. A rheometer(Figure 6-13) essentially consists of:

• a DC motor• a gear box• an electromagnetic clutch• a direct drive position servo actuator,

which, combined with the clutch, enableslow-amplitude angular deflection (andhence strain) to be transferred from the mo-tor via the gear box to the stalk of the outercylinder of the sample cell

• a sample cell, which can be of different ge-ometries but typically consists of two co-axial cylinders—an outer cup (e.g., internaldiameter, 27.5 mm), into which the milksample is placed, and an inner bob (e.g., in-ternal diameter, 25 mm); the strain on thesample results in a stress that, when re-leased, creates a strain on or angular dis-placement of the bob)

• a temperature sensor, which assists in accu-rate temperature control within the heating/cooling chamber surrounding the outer cyl-inder of the sample cell and hence withinthe sample

• a frictionless air bearing, which ensures thatthe strain or torque created on the bob bythe sample is transferred to a torsion bar

• a torque-measuring transducer, which mea-

sures the torque, digitizes it, and relays it tothe software of the interfacing computer

Dynamic measurements are performed by ap-plying a low-amplitude oscillating shear stress(a) or shear strain (y), depending on the type ofrheometer, to the milk sample via oscillations ofthe outer cylinder. The value of G or y is main-tained sufficiently low in order to stay within thelinear viscoelastic limits of the sample (i.e., theregion where G and y are directly proportional).Hence, the terms low-amplitude stress oscilla-tion and low-amplitude strain oscillation. (Am-plitude refers to the maximum displacement ofany point on the oscillating cup [and hence in themilk sample or on the inner bob] from its mean[or "zero"] position.) Under these conditions,the gel strands of the gelling milk are strained toa fixed displacement (within their elastic limit)and recover instantaneously when the stress isremoved. The stress required to achieve a fixedstrain (e.g., displacement of a gel strand at agiven position) increases as the gel strands be-come more elastic and firm; hence, measure-ment of stress energy provides a measure of thegel strength.

When a controlled-strain rheometer is used,the sample of renneted milk is subjected to anharmonic, low-amplitude shear strain, (y) of an-gular frequency co:

y =y0cos cot

where: y0 is the shear strain amplitude, co is theangular frequency (i.e., 27U)), and cos cot is aterm of the simple harmonic function. The shearstrain results in an oscillating shear stress (a) onthe milk that is of the same angular frequencybut is out of phase by the angle 8:

G = G0 cos (cot + 8)

where: G0 is the stress amplitude and G is thephase angle between the shear stress and shearstrain oscillations, the magnitude of which de-

Page 132: Cheese Science

pends on the viscoelasticity (ratio of viscous toelastic properties) of the gelling system. Thephase angle ranges from 0° for an elastic solid to90° for a Newtonian liquid and has intermediatevalues for viscoelastic materials (Figure 6-6).The rheological parameters a, G', and G" arecomputed continually from the measurement ofstress energy over time. The first, a, is of coursethe phase angle. G' is the storage or elastic shearmodulus, which represents elastically storedstress energy and thus gel elasticity or firmness.It is given by the equation:

G", the viscous or loss modulus, which repre-sents energy dissipated in flow, is given by theequation:

Typical changes in the above rheological pa-rameters from the time of rennet addition arepresented in Figure 6-5. The onset of gelation ismarked by sharp increases in G' and G" and adecrease in 5, whose abrupt decline from about80° in milk to about 10° marks the transitionfrom a viscoelastic material that is largely vis-cous (i.e., milk) to a gel that is largely elastic incharacter. Structural elements that impart elas-ticity include the relatively weak, continuousparacasein gel (formed from overlapping strandsof aggregated paracasein micelles), while thosethat contribute to viscosity include aggregatesand/or short dangling strands not yet connectedto the main network. Following the onset of ge-lation, the gel undergoes a progressive increase

Figure 6-13 Schematic representation of the main parts of a controlled-strain rheometer: DC motor (1), gear box(2), electromagnetic clutch (3), direct drive position servo actuator (4), sample cell (5), temperature sensor (6),frictionless air bearing (7), torsion bar (8), and torque-measuring transducer (9).

water

Page 133: Cheese Science

in firmness over time. However, the rate of curdfirming increases initially to a maximum andthen decreases (as reflected by the changes inslope of the G'/time curve). The increase in curdfirmness ensues from the following changes:

• the inclusion of free aggregates or danglingstrands into the network

• the increase in contact surface area andnumber of attractions between neighboringmicelles in the gel (Walstra & van Vliet,1986), which result in thicker strands withhigher stress-bearing capacity and thusmore elasticity

G" and 8 are useful parameters for monitor-ing the viscoelastic changes in the gel during ag-ing but are not directly related to gel strength andare thus not discussed further. In contrast, G' is adirect measure of curd firmness and is thus ofsignificance in cheese manufacture. Various ob-jective rennet coagulation parameters pertinentin cheesemaking may be derived from the G'time curve, upon modelling, as described below(Guinee, O'Callaghan, Mulholland, Pudja, &O'Brien, 1996):

• gel time, defined as the time at which G'reaches a threshold value, Gg, arbritarily setat 0.2 Pa

• the firmness after a fixed renneting time(e.g., 30 or 60 min, G'30 or G «,)

• maximum curd-firming rate, defined as themaximum slope, *Smax, of the GY graph

• set-to-cut time (SCT), which is the time be-tween rennet addition and curd cutting at asuitable firmness (e.g., 40Pa, SCT40Pa)

Low-strain or -stress rheometry gives param-eters that are Theologically precise and accu-rately quantify the dynamic rheological changesthat occur during renneting without altering theprocess of gel formation. Hence, it accuratelyreflects the changes in curd firmness that occurupon renneting milk in the cheese vat underquiescent conditions. However, limitationscompared to other instruments, such as theFormagraph, include the high level of operator

skill required for accurate measurement and thefact that only one sample can be analyzed at anyone time. Moreover, instruments currently avail-able cannot be used for online measurement ofgel formation in the cheese vat.

Online Sensors for Predicting Curd Firmnessand Cutting Time during Cheese Manufac-ture. Following the onset of gelation, there is aprogressive increase in curd firmness, and the geleventually attains an optimum firmness (e.g., 40Pa after 40-50 min, depending on milk composi-tion and renneting conditions), which, for a givenvat design, allows it to withstand the mechanicalaction of the cutting knives in the cheese vat with-out shattering. Curd shattering corresponds to thefracturing of the individual curd particles (e.g., bythe cheese knife and by impact with other curdparticles and/or the knife and walls of the cheesevat), especially if the coagulum is soft at cutting.Shattering results in an excessively large curdparticle surface area, through which fat is lost intothe cheese whey, and it is conducive to the forma-tion of very small curd particles (i.e., curd fines< 1 mm), which are also lost in the whey. Thus,cutting at the optimum firmness and the rate ofcurd firming are crucial for obtaining the correctparticle size, minimizing the losses of fat andfines in the whey, and maximizing cheese yield(Figure 6-14).

The firmness of the coagulum after a givenset-to-cut time is influenced by many factors,such as the concentrations of fat and casein inthe milk, the stage of lactation and the diet of thecow, the milk pH, the starter activity (which in-fluences pH at set and at cut time), and therennetcasein ratio. Variation in the firmness atcutting can, in turn, lead to variations in cheesecomposition (especially moisture), yield, andquality. Hence, standardizing and optimizing thefirmness at cutting is essential for consistentlyensuring high cheese yields and good-qualitycheese.

Formerly, the time at which the coagulum wascut was usually determined by the cheese maker,who subjectively assessed firmness by variousmeans, such as by cutting a small portion with a

Page 134: Cheese Science

Set-to-cut time, % of control

Figure 6-14 Effect of set-to-cut time (and hencefirmness at cutting) and healing time on moisture-ad-justed Cheddar cheese yield. Healing times were Omin (O), 5 min (•), and 10 min (Q).

hand knife and observing the "cleanness" of thecut and the clarity of the exuding whey. How-ever, in large modern factories, conditions arenot conducive to testing gel firmness in cheesevats from separate milk silos or from an indi-vidual day's milk, because of the large scale ofoperations (frequently more than 106 L of milkare processed per day) and the use of pre-programmed vats with limited operator access.Hence, most of the cheesemaking operations areperformed on the basis of a preset time schedulerather than on the basis of objective criteria, suchas gel firmness at cutting or pH at whey drain-age. The criteria for determining gel cut timesare probably based on one or more of the fol-lowing:

• data from previous production years thatsuggest that milk at different periods of theyear requires different set-to-cut times (e.g.,due to differences in milk composition)

• data from laboratory analysis from a previ-ous day's production showing the level offat and fines in cheese whey

• recovery of fat and casein in cheese

The above methods are not sufficiently pre-cise to allow cutting at a constant gel firmness inevery vat. The limitations of these methods haveled to the development of in-vat gel-firmnesssensors, which dynamically monitor milk co-agulation and, when incorporated into an inte-grated system, activate the curd-knives to cut thegel when it has attained the desired firmness(strength). The mechanisms employed to date indesigning in-vat sensors to monitor the develop-ment of curd firmness over time include moni-toring related changes in

• convective heat transfer from a probe (a"hot wire") to the surrounding milk, as inhot wire probe sensors (Bellon, Quiblier,Durier, & Noel, 1988; Hori, 1985; LeFevre& Richardson, 1990)

• turbidity (McMahon, Brown, & Ernstrom,1984)

• diffuse reflectance of visible (e.g., A,, 660nm), near infrared (e.g., K9 820 nm), or in-frared light (e.g., X, 950 nm), as in variousfiber-optic probes such as the CoAguLitefiber-optic sensor and the Omron E3XA(Payne, 1995)

• near-infrared light transmissions as in vari-ous infrared probes such as TxPro andGelographNT (O'Callaghan, O'Donnell, &Payne, in press)

• absorption and attenuation of ultrasoundwaves, or pulses, of different frequencies(e.g., > 0.5 MHz) passed through the milk(Benguigui, Emerery, Durand, & Busnel,1994; Gunasekaran & Ay, 1994).

The hot wire and fiber-optic probes are manu-factured commercially and are being used incheese factories as online curd-firmness sensors;these are discussed briefly below. Sensors basedon measuring changes in tubidity or ultrasoundattenuation have not yet, to the authors' knowl-

Moi

stur

e-ad

just

ed c

hees

e yi

eld,

kg/1

00 k

g m

ilk

Page 135: Cheese Science

edge, been successfully developed for use incheese plants.

The Hot Wire Probe. The hot wire probe hasachieved the furthest development toward com-mercial application of all curd firmness sensors,with most of the research occurring at the SnowBrand Dairy Products Company (Japan) and theNational Institute for Agronomic Research(France). Stoelting Inc. (Kiel, Wisconsin) hascommercialized the Optiset II hot wire sensor(from Snow Brand), which is now operating inseveral cheesemaking plants in the UnitedStates.

The principle of measurement is based onchanges in heat transfer from a hot wire to themilk. A thin platinum wire probe is immersed inthe milk. A constant current is passed throughthe wire, generating heat, which is dissipatedreadily, by convection currents near the wire,while the milk is liquid. As the milk coagulates,its viscosity increases, and generated heat is nolonger readily dissipated. The temperature of thewire increases, causing an increase in its resis-tance. The resistance and temperature of thewire are dynamically measured by monitoringchanges in voltage across the wire, giving a con-tinuous output signal.

A typical output signal is shown in Figure6-15. The peak in the first derivative of the out-put signal corresponds to the onset of gelation.The instrument does not detect a gel cuttingtime; the increase in gel firmness beyond the on-set of gelation (i.e., the gel point) is not readilydetected by the hot wire, as it has a relativelysmall effect on heat dissipation (compared withthe transition from a liquid to a gel). However,empirical equations have been developed to re-late the gel point, as detected by the hot wire, tocut times (at a particular firmness, e.g., 40 Pa) asdetermined by low-amplitude strain oscillationrheometry, the Formagraph, or other laboratorymethods.

Diffuse Reflectance Fiber-Optic Probe. Aninfrared diffuse reflectance probe, designed at theUniversity of Kentucky, was installed in twocheese plants in the United States in 1993. Theprinciple of measurement is based on changes in

the light-scattering properties of milk. Infraredlight is emitted from an LED and transmittedthrough one branch of a bifurcated cable contain-ing optical fibers to the tip of a probe in contactwith the renneted milk (Figures 6-16 through6-18). Light reflected by both the fat globules andcasein micelles is detected by the optical fibers inthe other branch of the bifurcated cable and trans-mitted to a photodetector. As the milk coagulates,more light is reflected (due to the aggregation ofthe paracasein micelles) and transmitted to thephotodetector, the output signal from which isdirectly proportional to the amount of light re-ceived. As in the case of the hot wire probe, thepeak in the first derivative of the output signalcorresponds to the onset of gelation, which is thenrelated to the cut time at a given firmness as deter-mined by laboratory instruments.

6.7 FACTORS THAT AFFECT RENNETCOAGULATION

The strength of the resulting gel (curd tension)is as important as the coagulation time, if notmore so, especially from the point of view ofcheese yield. The gel assembly process is quiteslow (see Figure 6-6), and in the case of mostcheese varieties a period roughly equal to theRCT is allowed from the onset of visual coagu-lation for the gel to become sufficiently firmprior to cutting. If the gel is too soft when cut, fatand casein losses in the whey will be high (seeBynum & Olson, 1982, for a description of theinfluence of curd firmness on cheese yield andfor references on this subject). In general, thereis an inverse relationship between RCT and curdtension, which means any factor that reducesRCT increases curd tension and vice versa. Theeffects of various compositional and environ-mental factors on the primary and secondaryphases of rennet coagulation are summarized inTable 6-2.

6.7.1 Milk Protein Level

The coagulation time of milk decreases mark-edly with protein (and thus casein) content, in

Next Page

Page 136: Cheese Science

edge, been successfully developed for use incheese plants.

The Hot Wire Probe. The hot wire probe hasachieved the furthest development toward com-mercial application of all curd firmness sensors,with most of the research occurring at the SnowBrand Dairy Products Company (Japan) and theNational Institute for Agronomic Research(France). Stoelting Inc. (Kiel, Wisconsin) hascommercialized the Optiset II hot wire sensor(from Snow Brand), which is now operating inseveral cheesemaking plants in the UnitedStates.

The principle of measurement is based onchanges in heat transfer from a hot wire to themilk. A thin platinum wire probe is immersed inthe milk. A constant current is passed throughthe wire, generating heat, which is dissipatedreadily, by convection currents near the wire,while the milk is liquid. As the milk coagulates,its viscosity increases, and generated heat is nolonger readily dissipated. The temperature of thewire increases, causing an increase in its resis-tance. The resistance and temperature of thewire are dynamically measured by monitoringchanges in voltage across the wire, giving a con-tinuous output signal.

A typical output signal is shown in Figure6-15. The peak in the first derivative of the out-put signal corresponds to the onset of gelation.The instrument does not detect a gel cuttingtime; the increase in gel firmness beyond the on-set of gelation (i.e., the gel point) is not readilydetected by the hot wire, as it has a relativelysmall effect on heat dissipation (compared withthe transition from a liquid to a gel). However,empirical equations have been developed to re-late the gel point, as detected by the hot wire, tocut times (at a particular firmness, e.g., 40 Pa) asdetermined by low-amplitude strain oscillationrheometry, the Formagraph, or other laboratorymethods.

Diffuse Reflectance Fiber-Optic Probe. Aninfrared diffuse reflectance probe, designed at theUniversity of Kentucky, was installed in twocheese plants in the United States in 1993. Theprinciple of measurement is based on changes in

the light-scattering properties of milk. Infraredlight is emitted from an LED and transmittedthrough one branch of a bifurcated cable contain-ing optical fibers to the tip of a probe in contactwith the renneted milk (Figures 6-16 through6-18). Light reflected by both the fat globules andcasein micelles is detected by the optical fibers inthe other branch of the bifurcated cable and trans-mitted to a photodetector. As the milk coagulates,more light is reflected (due to the aggregation ofthe paracasein micelles) and transmitted to thephotodetector, the output signal from which isdirectly proportional to the amount of light re-ceived. As in the case of the hot wire probe, thepeak in the first derivative of the output signalcorresponds to the onset of gelation, which is thenrelated to the cut time at a given firmness as deter-mined by laboratory instruments.

6.7 FACTORS THAT AFFECT RENNETCOAGULATION

The strength of the resulting gel (curd tension)is as important as the coagulation time, if notmore so, especially from the point of view ofcheese yield. The gel assembly process is quiteslow (see Figure 6-6), and in the case of mostcheese varieties a period roughly equal to theRCT is allowed from the onset of visual coagu-lation for the gel to become sufficiently firmprior to cutting. If the gel is too soft when cut, fatand casein losses in the whey will be high (seeBynum & Olson, 1982, for a description of theinfluence of curd firmness on cheese yield andfor references on this subject). In general, thereis an inverse relationship between RCT and curdtension, which means any factor that reducesRCT increases curd tension and vice versa. Theeffects of various compositional and environ-mental factors on the primary and secondaryphases of rennet coagulation are summarized inTable 6-2.

6.7.1 Milk Protein Level

The coagulation time of milk decreases mark-edly with protein (and thus casein) content, in

Previous Page

Page 137: Cheese Science

the range 2.0-3.0% (w/w), when rennet is addedon a volume basis (Figures 6-19 and 6-20). Fur-ther increases in milk protein level (i.e., > 3.0%,w/w) result in a slight increase in gelation time,an effect attributable to the decreasing ren-

net: casein ratio, which necessitates an increasein the time required to generate sufficient hy-drolysis of K-casein to induce aggregation ofparacasein micelles. At a constant rennetcaseinratio, the RCT decreases with increasing casein

(b) Time after rennet addition

Figure 6-15 (a) Hot wire sensor for objectively measuring the rennet coagulation of milk, (b) Changes in thetemperature of the hot wire during the course of the rennet coagulation of milk.

Non-enzymaticcoagulation

Enzymaticreaction

Coagulation

MilkData acquisition

Wire

Tem

pera

ture

of

hot w

ire

Page 138: Cheese Science

concentration, such as obtained by ultrafiltra-tion, and vice versa. From a practical viewpoint,a minimum protein level of 2.5-3.0% (w/w) isnecessary for gel formation in cheese manufac-ture (i.e., within 40-60 min). The maximumcurd-firming rate (Smax) and curd firmness (G increase more than proportionally with proteinlevel (Figure 6-18), with a power law depen-dence of the latter parameters and protein con-

centration (i.e., Smax <*p«i and G' <*Pn2, where nland n2 > 1.0, typically ~ 2.0 [Guinee et al.,1996]). Hence, small variations in the proteincontent of milk, as can occur throughout thecheesemaking season, exert a relatively large ef-fect on the coagulation properties of rennet. Thepositive effects of the higher milk protein con-tent on the rennet coagulation properties prob-ably ensue from the higher level of gel-forming

Figure 6-16 Schematic representation of the fiber-optic sensor developed to measure the diffuse reflectance ofcoagulating milk. LED = light emitting diode.

Detector

LEDFibre cable

Probe

Milk vat

Figure 6-17 Schematic representation of the diffuse reflectance probe of the fiber-optic sensor.

Reflected light

Light fromemitter

Fibre opticprobe

Casein micelles

Wall of cheese vat

Page 139: Cheese Science

protein, which increases the proximity of caseinmicelles and thus augments the rate of caseinaggregation.

One of the economic attractions in using ultra-filtration-concentrated milk in cheesemaking isthe savings that accrue from using less rennet.Cheese made from ultrafiltration-concentratedmilk ripens more slowly than normal, due partlyto slower proteolysis, for which there may be anumber of causes, including the lower ratio ofrennet to casein.

6.7.2 Milk Fat Level

Increasing fat content in the range 0.1-10%(w/w) while maintaining the protein level con-stant (e.g., at 3.3%, w/w) enhances the rennetcoagulation properties, as reflected by decreasesin coagulation time and set-to-cut time and

higher values for Smax and G' (Figure 6-20).However, the positive effects are much lower

than those obtained by increasing protein con-tent in the same range. Indeed, in a milk in whichthe level of fat plus protein is maintained con-stant, increasing the fat level results in signifi-cant decreases in 5max and G'. In commercialcheese manufacture, where standardization ofmilk protein to a fixed level (e.g., by ultrafiltra-tion of skim milk) is not normally practiced, Smax

and G' increase progressively upon addingcream to a fat level of about 4% (w/w) and de-crease rapidly thereafter. The decrease is due tothe dilution effect on the protein, which eventu-ally offsets the benefits of increasing the fat con-tent. From physical and structural consider-ations, the effect of increasing the fat level in amilk where the absolute level of gel-formingprotein is constant is probably twofold:

Time from rennet addition, min

Figure 6-18 Typical diffuse reflectance profile of rennet-treated milk showing the various stages of rennet co-agulation process: induction (rennet hydrolysis of K-casein), sigmoidal (aggregation of paracasein micelles andgel formation), and logarithmic (continued fusion of paracasein micelles and curd firming). The reflectance slopeis obtained from the first derivative of the reflectance ratio; the maximum slope (arrow a) corresponds to the timeof the maximum aggregation rate; arrow b indicates the time at which the gel has become sufficiently firm forcutting.

Out

put s

igna

l, R

efle

ctiv

e ra

tio

S to

pe of

the

refle

ctan

ce ra

tio c

urve

induction sigmoidal logarithmic

Page 140: Cheese Science

Table 6-2 Effects of Some Compositional and Processing Factors on Various Aspects of the Rennet Coagulation of Milk

Overall

Set-to-Cut Time

Curd Firmnessafter a Fixed

Renneting TimeCurd

Firming RateGTRCTSecondPhase

FirstPhaseFactor

U

Il

ft

NDftU

UftUIiU

ft

ft

Ii

NDUft

ftUftftft

ft

ft

U

NDU

NE

ftIiftftft

U

U

ftNDftU

U

UUU

ND

ND

ND

Iift

ND

UNDUUU

+++

+

+++

ND

+

++++

ND

ND

ND

++

ND

NEND+

++++++

Increasing the casein levelwhen rennet is added ona volume basis

Increasing the fat content of milkwhen casein level is constantwhen fat plus casein levels

are constantPasteurization temperature:

6O0C x 15 s> 720C x 15 s

Milk homogenizationAdded CaCI2

0.2-1OmM> 1OmM

Gelation temperature (4->35°C)Decreasing gelation pH (6.6^6.0)Rennet concentration

Key: RCT = rennet coagulation time as determined by the rennet coagulation time assay; GT = gelation time as determined by dynamic methods such as the Formagraph method orlow-amplitude strain oscillation rheometry; NE = no effect; ND = no data available; - = slight negative effect; -- = moderate negative effect; + = slight positive effect; ++ = moderatepositive effect; +++ = large positive effect; ft = magnitude of the rennet coagulation parameter increases; ii = magnitude of rennet coagulation parameter decreases.

Page 141: Cheese Science

1. The concomitant increase in viscositywith fat content probably restricts themovement of gel strands and thereby con-tributes to a higher gel rigidity.

2. Simultaneously, the increasing number offat globules causes the gel strands to be-come more elongated to surround and oc-clude the obstructing fat globules; this re-sults in thinner and weaker gel strands.

6.7.3 Pasteurization Temperature

Preheating milk up to about 650C has a benefi-cial effect on rennet coagulation, owing to heat-induced precipitation of calcium phosphate and aconcomitant decrease in pH. These changes occuralso at higher temperatures, but their beneficialeffects on rennet coagulation are offset and even-tually overridden by the combined effects of

Figure 6—20 Effect of increasing level of protein (•) in skim milk or fat (O) in milks containing 33 g/kg proteinon rennet coagulation properties: maximum curd-firming rate (A), curd firmness at 40 min after rennet addition(B), and set-to-cut time at a firmness of 20 Pa (C).

Milk protein or fat, g/100g Milk protein or fat, g/100g Milk protein or fat, g/10Og

Max

imum

cur

d fir

min

g ra

te, m

Pas'

1

Cur

d firm

ness

at 4

0 m

in, P

a

Set-t

o-cu

t tim

e at

20

Pa (m

in)

(A) (B) (C)

Time from rennet addition, min

Figure 6-19 Effect of milk protein level (30 [A], 35 [B], 45 [C], 69 [D], or 82 [E] g/kg) on the elastic modulusof rennet-coagulated milk. Milks B-E were prepared by ultrafiltration of milk A. Coagulation parameters thatmay be derived from the curve are gelation time (point at which G'begins to increase, curd-firming rate (slope ofG' time curve in the linear region), and curd firmness (the value of G' at a given time from rennet addition).

Elas

tic s

hear

mod

ulus

, Pa

Page 142: Cheese Science

• v/hey protein denaturation and the interac-tion of denatured (3-lactoglobulin with mi-cellar K-casein

• the deposition of heat-induced, insolublecalcium phosphate and the consequent re-duction, upon subsequent cooling, in theconcentration of native micellar calciumphosphate, which is important for cross-linking paracasein micelles, and hence ag-gregation, during gel formation

The complexation of denatured whey proteinwith K-casein adversely affects both the en-zymatic and nonenzymatic phases of rennet co-agulation but especially the latter. Increasing theextent of whey protein denaturation (as a per-centage of total) to a level greater than 15% byhigh heat treatment (e.g., > 8O0C x 15 s) impairsthe rennet coagulation characteristics to such anextent that the milk is unsuitable for commercialcheese manufacture (Figure 6-21). Very se-verely heated milk (e.g., 9O0C x 10 min, 80-90% total whey protein denatured) is not coagu-lable by rennet.

If heated milk is cooled, the RCT increasesfurther (Figure 6-22), a phenomenon referred toas rennet hysteresis. The effect can be explainedas follows: the adverse influence of the interac-tion of P-Ig with K-CN on rennet coagulation isoffset to some extent by the beneficial effect ofheat-precipitated calcium phosphate and re-duced pH. However, heat-induced changes incalcium phosphate are at least partially revers-ible upon cooling, and hence the full adverse ef-fects of the protein interaction become fully ap-parent upon cooling. In practice, milk should bepasteurized immediately before cheesemakingand should not be cold-stored before use.

6.7.4 Cooling and Cold Storage of Milk

Cooling and cold storage of milk (raw orheated) have adverse effects on the cheese-making properties of milk. Apart from thegrowth of psychrotrophs, two undesirablechanges occur:

1. Some indigenous colloidal calcium phos-phate dissolves, with a concomitant in-crease in pH.

2. Some proteins, especially (3-casein, dis-sociate from the micelles.

These changes are reversed by HTST pasteur-ization or by heating at a lower temperature,such as 310C, for a longer period.

6.7.5 Milk Homogenization

Homogenization of milk is practiced in themanufacture of some cheese varieties in whichlipolysis is important for flavor development,such as Blue cheese. The objective is to in-crease the accessibility of the fat to fungal Ii-pases and thus to increase the formation of fattyacids and their derivatives (e.g., methyl ke-tones). Moreover, homogenization is a centralpart of the manufacturing process for cheesesmade from recombined milks. Homogenizationreduces fat globule size and increases the inter-facial area of the fat surface by a factor of 5-6.Simultaneously, the fat globules become coatedwith a protein layer consisting of casein mi-celles, micelle subunits, and whey proteins.Hence, the newly formed fat globules behaveas pseudoprotein particles and are able to be-come part of the gel network. Numerous stud-ies have been undertaken to evaluate the effectof homogenization under different tempera-tures, pressures, and/or milk fat levels. Whilesome discrepancies exist between the results ofthese studies, the main trends indicate that ho-mogenization lowers the gelation time slightly,has no effect on the curd-firming rate, andcauses a slight increase in G'. However, thehigher moisture content of cheese made fromhomogenized milk, compared with that madefrom nonhomogenized milk, suggests that ho-mogenization may alter the rate of casein ag-gregation during the later stages of cheesemanufacture (i.e., after cutting).

6.7.6 Renneting (Set) Temperature

The principal effect of set temperature is onthe secondary, nonenzymatic phase of coagula-tion, which does not occur at temperatures be-

Page 143: Cheese Science

low around 180C. Above this temperature, thecoagulation time decreases to a broad minimumat 40-450C and then increases again as the en-zyme becomes denatured. In cheesemaking, ren-net coagulation normally occurs at around 310C,well below the optimum temperature. The lowertemperature is necessary to optimize the growthof mesophilic starter bacteria, which have an op-timum growth temperature of about 27-280Cand will not grow, nor perhaps even survive,above 4O0C. In addition, the structure of the co-agulum is improved at the lower temperature,

which is therefore used even for cheeses madeusing thermophilic cultures.

6.7.7 pH

Due to the effect of pH on the activity of theenzyme, the rennet coagulation time increaseswith increasing pH, especially above pH 6.4(Figure 6-23). The sensitivity to pH depends onthe rennet used. Porcine pepsin is particularlysensitive, while the microbial rennets are rela-tively insensitive. Owing to the pH dependenceof the rennet coagulation of milk, factors that

Figure 6-21 Effect of pasteurization temperature (for 15 s) on the level of whey protein denaturation (A), andeffect of the level of whey protein denaturation on maximum firming rate (B), curd firmness 60 min after rennetaddition (C), and set-to-curd time at a firmness of 20 Pa (D).

Whey protein denaturation, % Whey protein denaturation, %

Curd

firm

ness

at 6

0 m

in, P

a

Set-t

o-cu

t tim

e at

20

Pa, m

in(C) (D)

Pasteurization temperature, 0C Whey protein denaturation, %

(A)(B)

Whe

y pr

otei

n de

natu

ratio

n, %

Max

imum

cur

d fir

min

g ra

te, m

Pa/s

Page 144: Cheese Science

might affect the pH of milk (e.g., amount andform of starter added, addition of CaCl2, ripen-ing of milk, pH adjustment by addition of acid oracidogen, mastitis, and stage of lactation) willaffect rennet coagulability. Curd firmness in-creases markedly with decreasing pH to a maxi-mum atpH 5.9-6.0. The decrease in curd tensionat lower pH values may be due in part to thesolubilization of colloidal calcium phosphate asthe pH is reduced. The pH of milk increasesmarkedly in response to mastitic infection andmay exceed 7.0 (i.e., it approaches the pH ofblood, which is around 7.4). Mastitic milk has alonger RCT and lower curd tension than milkfrom healthy cows, probably owing to a combi-nation of factors, such as high pH, low caseincontent, and the generally high somatic cellcount (e.g., > 106 cells/ml) and associated pro-teolytic activity (which causes extensive hy-drolysis of asi- and p-caseins). The pH of firstcolostrum is around 6; the pH increases to thenormal value (6.7) within about 1 week and then

remains relatively constant for the main part oflactation, before increasing substantially (to pH7 or even higher) at the end of lactation.

6.7.8 Added CaCl2

The addition of CaCl2 to milk, which is com-mon practice, promotes rennet coagulation viathree beneficial changes:

1. an increase in [Ca2+]2. an increase in the concentration of colloi-

dal calcium phosphate3. a concomitant decrease in pH (the addi-

tion OfCaCl2 to 0.02%, i.e., 1.8 mM Ca,reduces the pH by ~ 0.05-0.1 units, de-pending on protein level)

Hence, the addition OfCaCl2 (to 0.02 g/L, i.e.,~ 2 mM Ca) enhances the rennet coagulationproperties as reflected by a reduction in gel timeand by increases in the curd-firming rate andcurd firmness (Figure 6-23). However, at addi-

hysteresiseffect

Time after heating

Figure 6-22 Schematic representation of the hysteresis effect on the rennet coagulation time (RCT) of heatedmilk. The symbols represent the RCT of raw milk (•), milk immediately after pasteurization (O), and milk 6 hrafter pasteurization (X).

Ren

net c

oagu

latio

n tim

e

Page 145: Cheese Science

2!±

O

P

O

P

K

* £±

< ^ ^

CJQ

-5 O

8 <L

& &

?

e.™

§3

'^

2. S

* iv

S ft

s*^

oS

S^

o^

^^

g^tr

^ag

/TH

-P ^

P.

P

Q

tr g

cT

p

co

ilc

lHl

E 0

§

^P;O

Q 3.

«^

PO

o-

CD

M

1

O

g"

3-2

3 H

| 5 o

Mfi

^I"

?^§

lff«

g3

3 %

a- S

o

F Q

Il P

^J fr

|es

^I's

-s-^

s-JS

:!i

-t

fD

mf

Dr

-^

Qf

Do

.

ES

-oD

ffiS

'-Q

A,

OO

pf

Dt

ic

/J

-^

CD

ti

*

I ??

U^

I?

the positive charge on the casein, making it lessprone to aggregation. As expected, the additionof calcium chelators (e.g., EDTA and sodiumphosphates) reduces gel firmness. The additionof NaCl or KCl increases gel firmness up to 100mM but markedly decreases it at higher concen-trations, possibly via displacement of micellarCa.

Figure 6-23 Effect of various factors on the rennet coagulation time of milk.

1/Rennet % Protein

RCT RC

T

RCT RC

T

RC

T

RC

T

Ca 0C

PH0C

(d)(c)

(O(e)

(a) (b)

Factor

TemperaturepHCaPre-heatingRennet concentrationProtein concentration

First phase

++++

+++++++

Second phase

++

+++++++

++++

Overall effect,see oanel

abC

def

Page 146: Cheese Science

6.7.9 Rennet Concentration

Obviously, the rate of the enzymatic phase ofrennet coagulation is directly related to theamount of rennet used; there is a good linear rela-tionship between enzyme concentration andmilk-clotting activity (MCA) (Figure 6-23).However, there are discrepancies as to the effectof rennet level on the curd-firming rate and curdfirmness, with some studies showing increases inthe latter parameters and others no effect or slightdecreases, depending on the stage of lactation.

In cheesemaking, the amount of rennet addedis sufficient to coagulate the milk in 30-40 min(200-220 ml of standard calf rennet [« 60 RU/ml] per 1,000 L of milk). This level of rennet istraditional and is presumably based on experi-ence. From a strictly coagulation viewpoint,more or less rennet could probably be used with-out adverse effects (other than the change inRCT). However, the amount of rennet retainedin the curd is proportional to the amount of ren-net added to the milk (at least for calf rennet),and this has a major effect on the rate of pro-teolysis during ripening. The retention of gastricrennets (e.g., calf chymosin and bovine pepsin)increases with decreases in pH at gel cutting andat whey drainage. On the other hand, the reten-tion of R. miehei (Rennilase) and R. pusillus(Emporase) proteinases is not influenced by pHat cutting or at whey drainage.

There are numerous reports that curd tensionis strongly influenced by the type of rennet used:calf chymosin gives a more rapid increase incurd tension than microbial rennets, although thesubstrate on which the rennets were standard-ized for clotting activity is of some significance.The fact that rennets standardized to equal clot-ting activity give different rates of curd firmingand behave differently in response to composi-tional factors, such as Ca2+, suggests possibledifferences in the extent and/or specificity ofproteolysis during the primary, enzymatic phaseof rennet coagulation. As far as is known, theprimary phase of coagulation by all the principalcoagulants involves cleavage of the PhC105-

Metioe bond except the enzyme from C. para-sitica, which hydrolyzes Seri04-Phei05. Possibly,other bonds are also hydrolyzed by microbialrennets, although this is not obvious from gelelectrophoretic studies. Further studies on thispoint are required.

The amount of rennet used seems to be op-tional, but the strategy of increasing concomi-tantly both the level of rennet used and startercell numbers does not seem to have been investi-gated as a possible means of accelerating cheeseripening.

6.7.10 Other Factors

The rennet coagulation properties of milkmay be influenced by stage of lactation and diet,which cause changes in milk composition (i.e.,casein, fat, mineral, and pH levels); degree ofcasein hydrolysis (e.g., as influenced by plasminand other proteinases); and the health of the cow.These effects tend to be more marked in coun-tries, such as Ireland, New Zealand, and Austra-lia, where milk is largely collected from spring-calving herds fed predominantly on pasture.Late lactation milk, especially when the lactoselevel is below 4.1% (w/w), is frequently associ-ated with long coagulation times and low curdfirmness. These defects may be alleviated bydrying off cows at milk yields of above 8 L/d,improving the quality of the feed, blending latelactation milk with early lactation milk, andstandardizing the cheesemaking process (e.g.,pH at set and the renneticasein ratio).

6.8 RENNET SUBSTITUTES

Owing to the increasing world production ofcheese (roughly 2-3% per annum over the past30 years) and the reduced supply of calf veils(due to a decrease in calf numbers and a ten-dency to slaughter calves at an older age), thesupply of calf rennet has been inadequate formany years. This has led to an increase in theprice of veal rennet and to a search for rennetsubstitutes. Despite the availability of numerous

Page 147: Cheese Science

potentially useful milk coagulants, only six ren-net substitutes (all aspartyl proteinases) havebeen found to be more or less acceptable forcheese production: bovine, porcine, and chickenpepsins and the acid proteinases from Rhizo-mucor miehei, R. pusillus, and Cryphonectriaparasitica. (Rhizomucor and Cryphonectriawere previously known as Mucor and Endothia,respectively.)

In addition to fulfilling the criteria laid downby legislative agencies regarding purity, safety,and absence of antibiotics (IDF, 1990), rennetsubstitutes must possess the following charac-teristics (Guinee & Wilkinson, 1992):

• A high MCA:proteolytic activity ratio. Ahigh ratio, as for example with calf rennet,prevents excessive nonspecific proteolysisduring manufacture and hence protectsagainst a weak gel structure, high losses ofprotein and fat in the whey, and reducedyields of cheese solids. Moreover, it avoidsexcessive proteolysis during maturationand thus ensures the correct balance of pep-tides of different molecular weights andhence desirable flavor, body, and functionalcharacteristics in the ripened cheese, mak-ing it suitable for certain applications (e.g.,processed cheese products and cheese pow-der). Excessive proteolysis, especially of (3-casein, is associated with the developmentof a bitter flavor.

• An MCA that is not very pH dependent inthe region 6.5-6.9. A sharp decrease inMCA combined with increasing pH maylead to slow gelation and a low curd tensionat cutting, especially if the milk pH at set-ting is high (e.g., 6.7-6.8, as may occur inlate lactation) or when the casein concentra-tion is low (e.g., < 2.4%, w/w). These con-ditions are conducive to low recovery of fatand reduced cheese yield and can occur inlarge factories, where the duration of milkripening is short (especially with the use ofdirect vat starters) and production steps (in-cluding cutting) are generally carried out

according to a fixed time schedule. The ad-dition OfCaCl2Or acidulants (e.g., gluconicacid-5-lactone) may overcome the latterproblems.

• Thermostability comparable to that of calfrennet at the pH values and temperaturesused during cheesemaking. This can mark-edly influence the level of residual rennet inhigh-cook cheeses such as Emmental,Romano, Provolone, and low-moistureMozzarella and hence the level of proteoly-sis, texture, and functionality of the cheeseduring maturation (see Chapters 11 and 19).

• Low thermostability of rennets duringwhey processing. A heat-stable rennet inthe whey (« 90% of that added to the cheesemilk) may lead to coagulation of formu-lated milks, which normally include whey(e.g., infant formulae and calf milk replac-ers) upon reconstruction.

• The ability to impart desired flavor, body,and texture characteristics to the finishedcheese.

Chicken pepsin is the least suitable of thecommercial rennet substitutes and was usedwidely only in Israel, where it has now been re-placed by microbial chymosin. Owing to itslow MCA:proteolytic activity ratio, chickenpepsin promotes extensive degradation of bothocsi- and (3-caseins in Cheddar cheese, leadingto the development of flavor defects (e.g., bit-terness) and textural defects (soft body andgreasiness) during maturation. Bovine pepsin isprobably the most satisfactory. Good qualityveal rennet contains about 10% bovine pepsin,and many commercial "calf rennets" containabout 50%. Its proteolytic specificity is similarto that of calf chymosin, and it gives generallysatisfactory results with respect to cheese yieldand quality. The activity of porcine pepsin isvery sensitive to pH greater than 6.6, and itmay be denatured extensively during cheese-making, impairing proteolysis during cheeseripening. A 50:50 mixture of porcine pepsinand calf rennet gave generally acceptable re-

Page 148: Cheese Science

suits, but porcine pepsin has been withdrawnfrom most markets.

Although the proteolytic specificity of thethree commonly used fungal rennets is consider-ably different from that of calf chymosin, theyhave given generally satisfactory results whenused in the manufacture of most cheese variet-ies. However, the proteolytic activity of all therennet substitutes is higher than that of calfchymosin, resulting in higher levels of protein inthe cheese whey and lower cheese yields (Figure

6-24). Prior to the introduction of geneticallyengineered chymosin, microbial rennets wereused widely in the United States but not in mostEuropean countries, Australia, or New Zealand.The extensive literature on rennet substitutes hasbeen reviewed (see Fox & McSweeney, 1997,for references).

Like chymosin, all commercially successfulrennet substitutes are acid (aspartyl) proteinases.The molecular and catalytic properties of theprincipal rennet substitutes are generally similar

(A)

(B)

BP CR/SP RP RM CP BP

BP CR/SP RP RM CP BP

Coagulant

Estim

ated

redu

ctio

n in

che

ese

yield

, %In

crea

se in

leve

l of w

hey

N c

ompa

red

toca

lf re

nnet

, %

Figure 6-24 Increase in the level of N in whey, expressed as percentage protein (A), and estimated decrease inmoisture-adjusted (370 g/kg) yield of Cheddar cheese (B), compared with calf rennet (94% chymosin and 6%bovine pepsin) when using different coagulants. BP, bovine pepsin (91% bovine pepsin + 9% calf chymosin);CR/SP, 50:50 blend of calf rennet and swine pepsin; RP, Rhizomucor pusillus\ RM, R. miehei; CP, Cryphonectriaparasitica\ BP, Bacillus polymyxa.

Page 149: Cheese Science

to those of chymosin (see Chitpinityol &Crabbe, 1998; Foltmann, 1993). Acid protein-ases have a relatively narrow specificity, with apreference for peptide bonds to which a bulkyhydrophobic residue supplies the carboxylgroup. Their narrow specificity is significant forthe success of these enzymes in cheese manufac-ture. The fact that the pH of cheese is far re-moved from their optima (~ 2 for porcine pep-sin) is probably also significant. However, notall acid proteinases are suitable as rennets, be-cause they are too active even under the prevail-ing relatively unfavorable conditions in milk andcheese. The specificity of porcine and bovinepepsins on asi- and p-caseins is quite similar tothat of chymosin, but the specificity of the fun-gal rennet substitutes is quite different (seeChapter 11). Like chymosin, the Phei05-Met106bond of K-casein is also preferentially hydro-lyzed by pepsins and the acid proteinases ofRhizomucor miehei and R. pusillus, but the acidproteinase of Cryphonectria parasitica prefer-entially cleaves the SCr1 Q4-PhC105 bond. However,unlike chymosin, the Rhizomucor and Crypho-nectria parasitica proteinases also cleave sev-eral other bonds in K-casein.

The MCA of commercial rennets (calf rennet,R. miehei, R. pusillus, and C. parasitica) in-creases with temperature in the range 28-360C.The MCA of porcine pepsin, calf rennet, andbovine pepsin at pH 6.6 increases with tempera-ture up to 440C, 450C and 520C, respectively.The fungal enzymes (R. miehei, R. pusillus, andC. parasitica) lose activity at 470C, 570C, and570C, respectively. The MCA of the pepsins, es-pecially porcine pepsin, is more pH dependentthan that of chymosin, while that of the fungalrennets is less sensitive in the pH region 6.2-6.8(Figure 6-25). The coagulation of milk by C.parasitica proteinase is also less sensitive toadded Ca2+ than coagulation by calf rennet, butcoagulation by Rhizomucor proteinases is moresensitive. For a given MCA, the rate of gel firm-ing depends on the rennet used; this aspect ofmilk coagulation should be independent of ren-net type and may indicate nonspecific proteoly-sis by the fungal enzymes.

The thermal stability of rennets differs consid-erably (Figure 6-26). Thermal stability is impor-tant when the whey is to be used in food process-ing. The early fungal rennets were considerablymore thermostable than chymosin or pepsins, butthe present products have been modified (by oxi-dation of methionine residues in the molecule)and have thermal stability similar to that ofchymosin. The thermal stability of C. parasiticaproteinase is less than that of chymosin at pH 6.6.The thermal stability of all rennets increasesmarkedly with decreasing pH (Figure 6-26)(Thunell, Duersch, & Ernstrom, 1979).

Although they are relatively cheap, rennetsrepresent the largest single industrial applica-tion of enzymes, with a world market of about25 x 106 L of standard rennet per annum.Therefore, rennets have attracted the attentionof industrial enzymologists and biotechnol-ogists. The gene for prochymosin has beencloned in E. coli, Saccharomyces cerevisiae,Kluyveromyces marxianus var. lactis, Aspergil-lus nidulans, A. niger, and Tricoderma reesei(see Foltmann, 1993, and Pitts et al., 1992, forreferences). The enzymatic properties of the re-combinant enzymes are indistinguishable fromthose of calf chymosin, although they may con-tain only one of the isoenzymes, A or B. Thecheesemaking properties of recombinant chy-mosins have been assessed on many cheese va-rieties, always with very satisfactory results(see review by Fox & Stepaniak, 1993). Re-combinant chymosins have been approved forcommercial use in many, but not all, countries.Three recombinant chymosins are now mar-keted commercially: Maxiren, secreted by K.marxianus var. lactis and produced by GistBrocades (the Netherlands); Chymogen, se-creted by A. niger and produced by Hansen's(Denmark); and Chymax secreted by E. coli,and developed by Pfizer (United States). Thegenes for Maxiren and Chymogen were isolatedfrom calf abomasum, while that used forChymax was synthesized. Microbial chymosinshave taken market share from both calf rennetand especially fungal rennets and now representabout 35% of the total market.

Page 150: Cheese Science

The recombinant chymosins currently avail-able are identical, or nearly so, to calf chymosin,but there are several published studies on engi-neered chymosins (Fox & McSweeney, 1997).At present, attention is focused on elucidatingthe relationship between enzyme structure and

function, but this work may lead to rennets withimproved MCA or modified general proteolyticactivity (i.e., on asr and/or p-casein). The natu-ral function of chymosin is to coagulate milk inthe stomach of the neonate. It was not intendedfor cheesemaking, and the wild-type enzyme

PH

Figure 6-25 Effect of pH on the rennet coagulation time (RCT) of milk using (A) calf chymosin (Q), bovinepepsin (•), ovine pepsin (O), and porcine pepsin (•); (B) calf rennet (Q) and Rhizomucor miehei (^), R. pusillus(O), Cryphonectria parasitica (A), and Bacillus polymyxa (A) proteinases.

Fold

incr

ease

in R

CT

B

A

Fold

incr

ease

in R

CT

Page 151: Cheese Science

Figure 6-26 Effect of heating time at 68.30C on theresidual activity of various coagulants in whey at pH5.2: calf rennet (Q), bovine pepsin (•), porcine pepsin(A), Rhizomucor miehei (•), R. pusillus (O), andCryphonectria parasitica (A) proteinases.

probably is not the most efficient or effectiveproteinase for catalyzing proteolysis in cheeseduring ripening. Therefore, it may be possible tomodify chymosin so as to accelerate its actionon specific bonds of casein during ripening and/or to reduce its activity on others, hydrolysis ofwhich may lead to undesirable consequences,such as bitterness. To date, the pH optimum,thermal stability, kcat, and KM of synthetic pep-tides have been modified through genetic engi-neering. We are not aware of any cheesemakingstudies using engineered chymosins, and ap-proval has not been obtained for their use.

The gene for R. miehei proteinase has beencloned in and expressed by A. oryzae (NovoNordisk A/S, Denmark). It is claimed that thisnew rennet (Marzyme GM) is free of other pro-teinase or peptidase activities that are present infungal rennets and may reduce cheese yield. Ex-cellent cheesemaking results with Marzyme GMhave been reported. Cloning of the gene for R.miehei proteinase has created the possibility forsite-directed mutagenesis of the enzyme.

6.9 IMMOBILIZED RENNETS

Most of the rennet added to cheese milk is lostin the whey (more than 90% in the case of Ched-dar). The loss of rennet represents an economicloss and creates potential problems for wheyprocessors. Both problems could be solvedthrough the use of immobilized rennets. A fur-ther incentive for immobilizing rennets is thepossibility of producing cheese curd continu-ously by using a cold renneting technique (i.e.,renneting at around 1O0C, which allows the pri-mary phase but not the secondary phase to oc-cur), which should facilitate process control.The feasibility of continuous coagulation usingcold renneting principles has been demon-strated, but the technique has not been com-mercially successful to date. As discussed inChapter 11, the chymosin (or rennet substitute)retained in cheese curd plays a major role incheese ripening. Consequently, if an immobi-lized rennet was used to coagulate milk, it wouldbe necessary to add some chymosin (or similarproteinase) to the curd, and uniform incorpora-tion of any such enzyme would be problematic,as has been demonstrated by the use of exog-enous proteinases to accelerate cheese ripening(see Chapter 15).

In modern cheesemaking, most operations arecontinuous or nearly so. The actual coagulationstep is the only major batch operation remaining,although the use of small "batches" of milk, as inthe Alpma process for Camembert, makes co-agulation, in effect, a continuous process. How-

Heating time at 68.30C, min

% O

rigin

al a

ctiv

ity

Page 152: Cheese Science

ever, in large modern Cheddar and Goudacheese factories, very large vats are used(20,000-30,000 L).

There is interest in the manufacture of ren-net-free curd for studies on the contribution ofenzymes from different sources to cheese rip-ening. A number of approaches have been usedto produce rennet-free curd (see Fox, Law,McSweeney, & Wallace, 1993), but an effec-tive completely immobilized rennet would bevery useful.

Several investigators have immobilized dif-ferent rennets on a range of supports and haveclaimed that these can coagulate milk. However,it appears that in such studies some enzymeleached from the support and that this solubi-lized enzyme was responsible for coagulation.An irreversibly immobilized rennet was unableto coagulate milk although it could hydrolyzenonmicellar casein. Presumably, the K-casein onthe surface of casein micelles is unable to enterthe active site cleft of the immobilized enzymeowing to steric factors.

Even if immobilized rennets could coagulatemilk, they may not be cost competitive (rennetsare relatively cheap) and would be difficult to

REFERENCES

Bellon, J.L., Quiblier, J.P., Durier, C., & Noel, Y. (1988). Unnoveau capteur industrial de mesure du temps de coagula-tion du lait: Le coagulometre. Technique Latiere andMarketing, 1031, 29-32. Cited from Dairy Science Ab-stracts, 50, 775.

Benguigui, L., Emerery, J., Durand, D., & Busnel, J. (1994).Ultrasonic study of milk clotting. Le Lait, 74, 197-206.

Bynum, D.G., & Olson, N.F. (1982). Influence of curd firm-ness at cutting on Cheddar cheese yield and recovery ofmilk constituents. Journal of Dairy Science, 65, 2290—2291.

Chitpinityol, S., & Crabbe, M.J.C. (1998). Chymosin andaspartic proteinases. Food Chemistry, 61, 395^18.

Dalgleish, D.G. (1980). Effect of milk concentration on therennet coagulation time. Journal of Dairy Research, 47,231-235.

Dalgleish, D.G. (1992). The enzymatic coagulation of milk.In P.P. Fox (Ed.), Advanced dairy chemistry: Vol. 1. Pro-teins. London: Elsevier Applied Science Publishers.

use in factory situations. The strategy envisagedfor their use involves the passage of cold milk(e.g., 1O0C) through a column of immobilizedenzyme where the enzymatic phase of rennetingwould occur without coagulation (owing to thelow temperature). The rennet-altered micelleswould then be coagulated by heating the milkexiting the column to about 3O0C. Heatingwould have to be conducted under quiescentconditions to ensure the formation of a good geland to minimize losses of fat and protein; quies-cent heating may be difficult on an industrialscale (e.g., many cheese factories process 106 Lof milk per day). Hygiene and phage-relatedproblems may present serious difficulties sincecleaning the column by standard regimes wouldinactivate the enzyme. Plugging of the columnand loss of activity have been problems even ona laboratory scale, and power cuts long enoughto lead to an increase in temperature would bedisastrous, as the column reactors would becomeplugged with cheese curd that would be difficultor impossible to remove. In short, the prospectsfor the use of immobilized rennets on a commer-cial scale are not bright, and such rennets are notcurrently being used.

Dalgleish, D.G. (1993). The enzymatic coagulation of milk,In P.P. Fox (ed.), Cheese: Chemistry, physics and micro-biology (2d ed., Vol. 1). London: Chapman & Hall.

Darling, D.F., & van Hooydonk, A.C.M. (1981). Derivationof a mathematical model for the mechanism of casein mi-celle coagulation by rennet. Journal of Dairy Research,48, 189-200.

Foltmann, B. (1993). General and molecular aspects of ren-nets. In P.P. Fox (Ed.), Cheese: Chemistry, physics andmicrobiology (2d ed., Vol. 1). London: Chapman & Hall.

Fox, P.P. (1984). Proteolysis and protein-protein interac-tions in cheese manufacture. In B.J.F. Hudson (Ed.), De-velopments in food proteins (Vol. 3). London: ElsevierApplied Science Publishers.

Fox, P.P., Law, J., McSweeney, P.L.H., & Wallace, J.(1993). Biochemistry of cheese ripening, In P.P. Fox(Ed.), Cheese: Chemistry, physics and microbiology (2ded., Vol. 1). London: Chapman & Hall.

Fox, P.P., & McSweeney, P.L.H. (1997). Rennets: Their rolein milk coagulation and cheese ripening. In B.A. Law

Page 153: Cheese Science

(Ed.), Microbiology and biochemistry of cheese and fer-mented milk (2d ed.). London: Chapman & Hall.

Fox, P.P., & Mulvihill, D.M. (1990). Casein. In P. Harris(Ed.), Food gels. London: Elsevier Applied Science Pub-lishers.

Fox, P.F., O'Connor, T.P., McSweeney, P.L.H., Guinee,T.P., & O'Brien, N. (1996). Cheese: Physical, biochemi-cal and nutritional aspects. Advances in Food Science andNutrition, 39, 163-328.

Fox, P.F., & Stepaniak, L. (1993). Enzymes in cheese tech-nology. International Dairy Journal, 3, 509-530.

Guinee, T.P., O'Callaghan, D.J., Mulholland, E.O., Pudja,P.D., & O'Brien, N. (1996). Rennet coagulation proper-ties of retentates obtained by ultrafiltration of skim milksheated to different temperatures. International DairyJournal, 6, 581-596.

Guinee, T.P., & Wilkinson, M.G. (1992). Rennet coagula-tion and coagulants in cheese manufacture. Journal of theSociety of Dairy Technology, 45, 94-104.

Gunasekaran, S., & Ay, C. (1994). Evaluating milk coagula-tion with ultrasonics. Food Technology, 48, 774-780.

Hori, T. (1985). Objective measurement of the process ofcurd formation during rennet treatment of milks by the hotwire method. Journal of Food Science, 50, 911-917.

Iametti, S., Giangiacomo, R., Messina, G., & Bonomi, F.(1993). Influence of processing on the molecular modifi-cations of milk proteins in the course of enzymic coagula-tion. Journal of Dairy Research, 60, 151-159.

International Dairy Federation. (1990). Use of enzymes incheesemaking (Bulletin No. 247). Brussels: Author.

International Dairy Federation. (1992). Bovine rennets. De-termination of total milk-clotting activity (ProvisionalStandard 157). Brussels: Author.

LeFevre, M.J., & Richardson, G.H. (1990). Monitoringcheese manufacture using a hot wire probe. (Abstract).Journal of Dairy Science, 73 (Suppl. 1), 74.

McMahon, D.J., Brown, R.J., & Ernstrom, C.A. (1984). En-zymic coagulation of milk. Journal of Dairy Science, 67,745-748.

O'Callaghan, D.J., O'Donnell, C.P., & Payne, F.A. (inpress). Effect of protein content of milk on the storage and

loss moduli in renneting milk gels. Journal of Food Pro-cess Engineering.

Payens, T.A.J., & Wiersma, A. (1977). On enzymic clottingprocesses: V. Rate equations for the case of arbitrary rateof production on the clotting species. Biophysical Chem-istry, 11, 137-146.

Payens, T.A.J., Wiersma, A., & Brinkhuis, J. (1977). On en-zymic clotting processes: I. Kinetics of enzyme-triggeredcoagulation reactions. Biophysical Chemistry, 6, 253-261.

Payne, F.A. (1995). Automatic control of coagulum cuttingin cheese manufacture. Applied Engineering in Agricul-ture, 77,691-697.

Peri, C., Pagliarini, E., Iametti, S., & Bonomi, F. (1990). Astudy of surface hydrophobicity of milk proteins duringenzymic coagulation and curd hardening. Journal ofDairy Research, 57, 101-108.

Pitts, J.E., Dhanaraj, V., Dealwics, C.G., Mantafounis, D.,Nugent, P., Orprayoon, P., Cooper, J.B., Newman, M., &Blundel, T.L. (1992). Multidisciplinary cycles for proteinengineering: Site-directed mutagenesis and X-ray struc-tural studies on aspartic proteinases. Scandinavian Jour-nal of Clinical and Laboratory Investigation, 52 (Suppl.210), 39-50.

Storry, J.E., & Ford, G.D. (1982). Development of coagulumfirmness in renneted milk: A two-phase process. Journalof Dairy Research, 49, 343-346.

Surkov, B.A., Klimovskii, I.I., & Krayushkin, V.A. (1982).Turbidimetric study of kinetics and mechanism of milkclotting by rennet. Milchwissenschaft, 37, 393-395.

Thunell, R.K., Duersch, J.W., & Ernstrom, C.A. (1979).Thermal inactivation of residual milk clotting enzymes inwhey. Journal of Dairy Science, 62, 373-377.

Tuszynski, W.B. (1971). A kinetic model of the clotting ofcasein by rennet. Journal of Dairy Research, 38, 115-125.

van Hooydonk, A.C.M., de Koster, P.G., & Boerrigter, IJ.(1987). The renneting properties of heated milk. Nether-lands Milk and Dairy Journal, 41, 3-18.

Walstra, P., & van Vliet, T. (1986). The physical chemistryof curd making. Netherlands Milk and Dairy Journal, 40,241-259.

Page 154: Cheese Science

7.1 INTRODUCTION

The rennet coagulation process is essentiallysimilar for all cheese varieties and the structureof the coagulum (gel) is also similar. The gel issubjected to a series of treatments (see Chapter2), the principal object of which is to removewhey from the gel and effectively concentratethe casein and fat to the degree characteristic ofthe variety. The principal treatments are de-scribed in this chapter. A summary on the manu-facturing protocol for each of a number ofcheeses is given in Chapter 17.

Rennet- or acid-coagulated milk gels are quitestable if left undisturbed, but when they are cutor broken or subjected to external pressure, theparacasein matrix contracts, expressing theaqueous phase of the gel (known as whey). Thisprocess, known as syneresis, enables the cheesemaker to control the moisture content of thecheese; hence the activity of microorganismsand enzymes in the cheese, and hence the bio-chemistry of ripening and the stability and qual-ity of the finished cheese. The higher the mois-ture content of cheese, the faster it will maturebut the less stable it will be. High-moisturecheeses have a much greater propensity to de-velop off-flavors than low-moisture varieties.Although the starter and adventitious microfloraof cheese have a major impact on the biochemis-try of cheese ripening, they do so only in as far asthe composition of the cheese curd permits. Sy-neresis is under the control of the cheese maker,

and, via syneresis, so is the composition andquality of the cheese.

Many of the treatments to which rennet-co-agulated milk gels are subjected may be classi-fied generically as dehydration. Cheese manu-facture essentially involves concentrating the fatand casein of milk approximately tenfold, andremoving lactose, whey proteins, and solublesalts in the whey. Although there are certaincommon features, the factors that promote andregulate syneresis (dehydration) in a cheese va-riety or family of varieties are specific to thatvariety or family. In the case of Cheddar- andSwiss-type cheeses, dehydration is accom-plished mainly in the cheese vat by fine cuttingthe coagulum, extensive "cooking" of the curds-whey mixture (to ~ 4O0C for Cheddar-type and« 550C for Swiss-type cheeses) and vigorousagitation during cooking. For the softer (high-moisture) varieties, the gel may be scooped di-rectly into the molds without cutting or cooking,and whey explosion occurs mainly in the moldsas the pH decreases. Curds for some varieties(e.g., Cheddar and Swiss) are subjected to con-siderable pressure in the molds to aid whey re-moval, while curds for the softer varieties arepressed only under their own weight.

Most of the published studies on syneresishave been concerned mainly with the factorsthat affect it during the early stages of dehydra-tion in Cheddar- and Dutch-type cheeses, thatis, mainly during cooking, but it is assumedthat basically the same mechanisms operate in

Postcoagulation Treatment ofRenneted Milk Gel

CHAPTER 7

Page 155: Cheese Science

all varieties throughout the dehydration pro-cess.

Despite its accepted importance in the controlof cheese moisture, the mechanism of syneresisof renneted milk gels is not well understood.There is a considerable amount of empirical in-formation on factors that influence syneresis butthe actual mechanism of syneresis has receivedvery little study. Poor methodology is mainly re-sponsible for the lack of information; the num-ber of principles exploited in methods used tomeasure syneresis attests to their unsuitability.Some authors have attempted to simulate cheesemanufacture, such as by stirring, observing acooking profile, even adding starter, but the ac-curacy and precision of many of the methods arepoor. Many of the methods have been used onlyby the original investigator.

The literature on the syneresis of milk gels hasbeen reviewed by Green and Grandison (1993)and Walstra (1993) and is summarized here.

7.2 METHODS FOR MEASURINGSYNERESIS

A variety of methods have been employed toquantify syneresis. These include

• measuring the volume of whey expressedfrom curd pieces under standard conditions,following cutting of the gel

• measuring changes in the moisture content,volume, or density of curd pieces over time

• using tracers or markers to indirectly mea-sure whey volume

• measuring changes in the electrical resis-tance of the curd

Techniques for measuring the volume of ex-pressed whey are simple and straightforward toexecute, but complete recovery of whey is diffi-cult, and syneresis continues during any separa-tion process. Similar constraints apply to meth-ods that depend on the volume or composition ofthe curd, and, in addition, the actual analyticalstep may be difficult while avoiding continuingsyneresis. Methods based on the use of a traceror marker involve adding a small volume of a

solution of some tracer (e.g., a dye) to the systemat the start of syneresis; as the volume of freewhey increases, the concentration of tracer in thesolution decreases. The principal problems to beavoided are diffusion of the tracer into or its ad-sorption onto the curd particles. The oppositestrategy has also been used—the placing of asmall amount of clarified whey on top of the cutgel. Whey expressed from the curd is turbid ow-ing to the presence of fat globules, and thereforethe turbidity of the free whey increases as syner-esis progresses. As the moisture content of curddecreases, its electrical conductivity decreases.As with many other methods, clean separation ofcurd particles from the whey without concomi-tant changes is difficult to achieve.

However, data from studies using these meth-ods and from actual cheesemaking experimentshave helped clarify the influence of several fac-tors on syneresis, at least in general terms.

7.3 INFLUENCE OF COMPOSITIONALFACTORS ON SYNERESIS

The syneresis of renneted milk gels is influ-enced by milk composition, which in turn is af-fected by the feed, stage of lactation, and healthof the animals from which the milk is obtained.Fat tends to reduce syneresis and increase thewater-holding capacity of cheese curd, and in-creasing the fat content of cheese milk increasescheese yield by about 1.2 times the weight of theadditional fat. However, syneresis tends to bedirectly related to casein concentration, which isto say that good syneresis occurs at high caseinlevels. Since the fat and casein levels in milktend to change in parallel, they have offsettingeffects on syneresis. Concentration of milk sup-presses syneresis, possibly because of its effecton gel strength, although the rigidity modulus(see Chapter 6) at the time of cutting appears tohave little effect on syneresis.

The rate of syneresis is directly related to theacidity and therefore is inversely related to pH; itis optimal at the isoelectric point (i.e., pH 4.6-4.7). The addition OfCaCl2 to milk promotes sy-neresis, but the effect appears to be less than

Page 156: Cheese Science

might be expected and may be negative at cer-tain pH values and at high calcium concentra-tions, especially if the gel is held for a long pe-riod before cutting. The adverse effect of a highconcentration of calcium has been attributed tointeraction of Ca2+ with the aspartate and glu-tamate groups of proteins, leading to an in-creased net positive charge, swelling of the pro-tein, and suppression of syneresis. It is likelythat a firmer gel, such as would be obtained onlonger holding, would also be more resistant tosyneresis. The influence of colloidal calciumphosphate on syneresis does not appear to havebeen investigated. Addition of low levels ofNaCl increases the rate of syneresis but higherlevels retard it.

7.4 INFLUENCE OF PROCESSINGVARIABLES ON SYNERESIS

The extent of syneresis, and hence the mois-ture content of cheese, is influenced by variousfactors associated with cheesemaking proce-dures. Many of these are exploited by cheesemakers to control cheese composition and thusits flavor and texture. The principal factors aredescribed below.

7.4.1 Size of the Curd Particles

Everything else being equal, the smaller thecurd pieces, the faster the rate and the greater theextent of syneresis, reflecting the greater surfacearea available for loss of whey. For some high-moisture cheeses, the coagulum may not be cutbut scooped, unbroken, into cheese molds. ForCheddar and Dutch-type cheeses the coagulumis cut into cubes of about 1 cm size using kniveswith vertical or horizontal wires or bars (Figure7-1). Traditionally, the coagulum for manySwiss or Italian varieties was cut with a harp or aspino, respectively (Figure 7-1), which is usedin a swirling action around the hemispherical orconically shaped vats used traditionally for thesevarieties. In the large modern vats used forCheddar, Dutch, and other varieties, the cuttingknives are fixed in the vats and serve to cut the

coagulum and agitate the curd-whey mixtureduring cooking (Figure 7-2).

7.4.2 Cook Temperature

Heating the curds-whey mixture (a processreferred to as scalding or cooking) promotes sy-neresis (Figure 7-3). The cook temperature ischaracteristic of the variety. For example, thecook temperature is 310C for high-moisture va-rieties such as Camembert (in effect, no cook-ing), 360C for Gouda and Edam, 38-4O0C forCheddar, and 52-550C for Emmental andParmesan. The cook temperature must match thethermal stability of the starter. Acid productionby some Lactococcus strains is stopped around350C, but other strains withstand cooking at 40-420C, and cooking cheese curds to a temperaturethat inhibits the culture may have a negative ef-fect on syneresis owing to the reduced rate ofacidification. A cook temperature up to 550Cmay be used when a thermophilic starter is used.Such starters survive but do not grow at 550C,and hence syneresis depends on temperaturerather than on pH. In fact, temperature and pHare complementary: syneresis of low-acid curds(e.g., Emmental) depends mainly on tempera-ture, while in high-acid curds (e.g., Camembert)temperature is of little consequence.

For most varieties, cooking is done by circu-lating hot water or steam through the jacket ofthe cheese vat (steam is preferable because it canbe shut off more readily than hot water, facilitat-ing better control of temperature). Before theavailability of hot water or steam for cookingcurds and whey, cooking was done over an openfire and would have been difficult to control pre-cisely (open-fire cooking is still practiced inartisanal farmhouse cheesemaking). Illustrationsof old cheese factories suggest that jacketed vats(and cooking by hot water or steam) were usedfrom the start of industrialized cheesemaking.For Dutch-type cheeses and a number of othervarieties, cooking is done by removing part ofthe whey (30^0%) and replacing it with warmwater to give a blend of the desired temperature.This method was probably developed for

Page 157: Cheese Science

A B

C D

Figure 7-1 Examples of tools used to cut renneted milk gels: (A) vertical knife, (B) horizontal knife, (C) harp,and (D) spino.

Page 158: Cheese Science

Figure 7-2 Illustrations of cheese vats showing the curd knife-stirrer blades. The curd knife-stirrer blades areshown arranged vertically in a Double-O Multicurd vat (A); arranged horizontally in a staggered mode along acentral horizontal shaft in a cylindrical OST vat (B); and located toward the end of the cutting cycle in a horizon-tal cheese vat (C). The blades are tapered and move in the direction of the sharp edge (knife) when cutting and inthe reverse direction (blunt edge, stirrer) when stirring.

C

A B

Page 159: Cheese Science

(a)

volu

me

of w

hey

Time after cutting

(b)

Vol

ume

of

whe

y

Time after cutting

Figure 7-3 Effect of temperature (a) and pH (b) on the rate and extent of syneresis in cut or broken renneted milkgels.

Page 160: Cheese Science

cheesemaking on farms, which would havelacked the means to circulate hot water or steamthrough a jacketed vat. It was probably used formany other varieties, but it is now mainly re-stricted to Dutch-type cheeses, where its mainfunction is to reduce the lactose content of thecheese curd and thereby control the pH of thecheese.

The rate of cooking is characteristic of the va-riety (see Chapter 17). If the rate of cooking istoo fast, especially during the early stages, ex-cessive dehydration will occur at the curd sur-face, leading to the formation of a skin (casehardening), which will retard syneresis and theremoval of whey from the interior of the curdpieces and result in a high-moisture content.

7.4.3 Rate of Acid Development

The lower the pH, the faster the rate and thegreater the extent of syneresis (see Figure 7-3).Presumably, this relationship reflects the re-duced negative charge on the casein moleculesas the isoelectric point is approached.

7.4.4 Stirring of the Curd-Whey Mixture

During cooking, the curd-whey mixture isstirred, which serves a number of functions:

• It facilitates cooking.• It prevents the curd pieces from matting

(which would have a strong negative effecton syneresis).

• It promotes syneresis via collisions be-tween curd pieces and between curd piecesand the vat wall.

Everything else being equal, syneresis is di-rectly proportional to the intensity of stirring.Initially, the curd is very soft, and gentle stirringshould be used. A period of 5-10 min is allowedfor the cut surfaces of the curd pieces to "heal,"and vigorous agitation of the curd during thisperiod will cause extensive losses of fat and pro-tein into the whey and a decrease in cheese yield(see Chapter 9).

For some varieties, the curd is held in thewhey until a certain pH is reached (e.g., 6.2 forCheddar), after which the curds and whey areseparated, usually using metal screens. For othervarieties (e.g., Gouda), the curds and whey areseparated after holding at the desired cook tem-perature for a defined period. For yet other vari-eties (e.g., Parmesan), the curds and some wheyare scooped from the vat using a cheesecloth andtransferred to perforated molds, where much ofthe whey drainage occurs. If whey separationoccurs in the cheese vat, stirring of the curd dur-ing draining (referred to as dry stirring) is a use-ful way of promoting syneresis, but this methodis not applicable to most varieties.

7.4.5 Pressing

After removal of the whey, the curds mat toform a continuous mass. Treatment of this massof curd is characteristic of the variety and mayinvolve inverting the mass of curd in the molds,turning and piling blocks of curd in the vat (tra-ditional "cheddaring"), and in many cases press-ing (see Chapter 17). Syneresis occurs duringthese operations but is not easily controlled.With the exception of Cheddar-type cheeses,acidification occurs mainly after molding, andthis promotes considerable syneresis of thecurds. For Cheddar-type cheeses, acidificationoccurs mainly during "cheddaring" in the vats,and relatively little syneresis occurs after mold-ing.

7.4.6 Salting

As discussed in Chapter 8, all cheeses aresalted at the end of manufacture. Salting causesthe loss of moisture from the curd (~ 2 kg H2Oare lost per kg of salt absorbed). However, salt-ing should not be relied upon as a means of con-trolling cheese moisture.

7.4.7 Milk Quality and Pretreatment

Heating milk under conditions that causewhey protein denaturation and interaction with

Page 161: Cheese Science

casein micelles reduces the tendency of rennetedmilk gels to synerese. Homogenization of wholemilk has a similar effect. It is reported that thegrowth of psychrotrophs in milk reduces itssyncretic properties (i.e., leads to a high mois-ture content). Increased levels of rennet slightlyincrease the rate of syneresis, while plasmin ac-tivity in milk is reported to reduce syneresis.

7.5 KINETICS AND MECHANISM OFSYNERESIS

Data from various investigations indicate thatsyneresis is initially a first-order reaction, that is,the rate of syneresis depends on the amount ofwhey remaining in the curd. Within the tempera-ture range 16-450C at constant pH, curd vol-ume, Vt9 as a function of time can be expressedby the formula:

2.3 logV, = 2.3 logV0-tk(T-T0)

where / is time (min), T is temperature, and k andT0 are constants having values of 0.225 x 10~3

and 16 under the experimental conditions.It is generally assumed that syneresis is due to

protein-protein interactions and may be re-garded as a continuation of the gel assembly pro-cess during rennet coagulation. The inhibitoryeffects of high concentrations of salts (CaCl2,NaCl, KCl) on syneresis suggest that ionic at-tractions are involved. Urea promotes syneresis,suggesting that hydrogen bonds are not in-volved. The effectiveness of pH in promotingsyneresis is probably due to a reduction of over-all charge as the isoelectric point is approached.Studies of artificial milk systems implicate the e-NH2 group of the lysine residues in p-casein insyneresis. The apparent importance of 8-NH2

groups in the second phase of rennet coagulationwas discussed in Chapter 6. Lysozyme, whichreacts with casein micelles, reducing theircharge and rennet coagulation time, also acceler-ates syneresis when added to milk.

Some authors have concluded that curd-firm-ing and syneresis are different aspects of thesame phenomenon, but opinions are not unani-

mous. As discussed in Chapter 6, electron mi-croscopy studies have shown that the aggrega-tion of casein micelles to form a gel is followedby increasingly closer contact between the mi-celles, leading to fusion. The syncretic pressurein an uncut gel is very small (~ 1 Pa), but whenthe coagulum is cut, the whey leaks out. Syner-esis is initially a first-order reaction, because thepressure depends on the amount of whey in thecurd. Holding curd in whey retards syneresisowing to back pressure of the surrounding whey,while removing whey promotes syneresis. Whenthe curd is reduced to roughly 70% of its initialvolume, syneresis becomes dependent on factorsother than the volume of residual whey in thecurd. It has been proposed that hydrophobic andionic interactions within the casein network areprobably responsible for the advanced stages ofsyneresis. This view is in accord with the promo-tion of syneresis by reduced pH and low levelsof CaCl2, which reduce micellar charge and in-crease hydrophobicity, and by increased tem-perature, which increases hydrophobic interac-tions.

The foregoing discussion of syneresis per-tains especially to the syneresis occurring in thecheese vat, that is, mainly to hard and semi-hardvarieties of cheese. Syneresis continues afterhooping (molding), and syneresis during this pe-riod represents the major part of syneresis in softvarieties. Presumably, the mechanism of syner-esis in the molds is the same as in the cheese vat,although the range of treatments that can be ap-plied at this stage is rather restricted. Externalpressure is applied to the curds for many cheesevarieties after molding and makes a significantcontribution to whey removal. In general, thedrier the cheese curd at hooping, the higher thepressure applied, which is probably a reflectionof the greater difficulty in ensuring fusion oflow-moisture curds.

7.6 TEXTURED CHEESE

The development of a recognizably fibroustexture is part of the manufacturing procedure

Page 162: Cheese Science

for a small number of cheese varieties, and thistexture was traditionally regarded as an essentialorganoleptic characteristic of these cheeses.Texturized cheeses belong to two classes: Ched-dar and some closely related varieties, in which afibrous texture is developed prior to pressing,and pasta filata types, such as Mozzarella,Kashkaval, and Provolone, in which textur-ization is accomplished by heating, stretching,and kneading the curd.

In traditional Cheddar manufacture, thedrained curds are piled along the sides of the vat,with the result that matting (fusion) of curd par-ticles occurs. To enable faster turnover of thecheese vats, it became common practice in the1960s to transfer the curd-whey mixture aftercooking to cheaper cheddaring "sinks," wherewhey drainage and cheddaring occurred. Thepiles of curd are cut into blocks (30 x 10 cm),which are inverted frequently and piled over aperiod of about 2 hr. This operation, known ascheddaring, was considered by many research-ers and cheese makers as the most characteristicpart of the Cheddar cheese manufacturing pro-cess. During cheddaring, the curd flows under itsown weight, leading to fusion and deformationof curd particles, which was believed to be re-sponsible for the "chicken breast meat" structureof fresh Cheddar curd and for the characteristictexture of mature Cheddar cheese. Cheddaringpromotes a number of physicochemical condi-tions that are conducive to curd flow and tex-turization:

• Solubilization of micellar calcium, which isbound to the casein and acts as a cementingagent between the casein micelles/submi-celles

• A decrease in the concentration of micellarCa, resulting in an increase in the ratio ofsoluble to casein-bound Ca (soluble Ca as apercentage of total Ca in the curd increasesfrom ~ 5% to 40% as the pH decreases from6.15 to 5.2)

• An increase in paracasein hydration, whichincreases with decreasing pH in the range6.6-5.15

• An increase in the viscous character of thecurd

The increase in casein hydration with decreasingpH is a probably a consequence of the increasein the ratio of soluble to micellar Ca. It has beenfound in model casein systems that casein hydra-tion is inversely related to the concentration ofcasein-bound Ca (Sood, Sidhu, & Dewan,1980).

As a consequence of the decrease in casein-bound Ca and the increase in casein hydration,the viscoelastic casein matrix, with occludedliquid fat and moisture phases, flows if unre-stricted, especially when piled and pressed underits own weight. The flow of curd gives the de-sired planar orientation of the strands of theparacasein network (Figure 7—4). The physico-chemical changes in curd during cheddaring aresummarized in Figure 7-5. Note that there islittle scientific support for the necessity ofcheddaring. On the contrary, there is strong evi-dence that cheddaring is of no consequence toCheddar cheese quality and serves only to allowthe desired degree of acid development and sy-neresis to occur.

Various forms of restricted flow under differ-ent degrees of external pressure result in Ched-dar cheese with a lower moisture content thancurd cheddared in the traditional manner. Differ-ences in the extent of curd deformation causedby modified cheddaring processes diminish dur-ing milling, salting, and pressing and have littleeffect on the flavor and textural characteristicsof the final cheese. The development of a fibroustexture results in loss of micelle structure, butthis change in structure is not essential, as theamount of deformation is very small and is prob-ably altered by the subsequent and more exten-sive deformation during pressing.

In modern practice, most Cheddar cheese curdis manufactured in continuous, mechanicalcheddaring systems in which little flow occurs incomparison with traditional methods. Indeed,matting is prevented in the manufacture of someCheddar-type cheeses, such as stirred-curdCheddar. The textural quality of Cheddar cheese

Page 163: Cheese Science

produced by these systems is acceptable, indi-cating that flow during manufacture is not essen-tial.

Presumably, the various interactions, ionicand/or hydrophobic, that are considered to be re-sponsible for syneresis continue during thecheddaring process, but there appear to havebeen no studies on this aspect of cheesemaking.

In the manufacture of Mozzarella and otherpasta-filata (stretched-curd) cheeses, the ched-dared curd is shredded, heated to around 58-6O0C by kneading in hot water (« 780C), andstretched in equipment designed to promote ex-tension of the hot molten curd (Figure 7-6).The process by which the curd is converted into

a plastic molten mass is referred to as plas-ticization and was originally developed in hotclimates as a means of pasteurizing and henceextending the shelf-life of curd of poor micro-biological status. Successful plasticization ofthe curd requires that the viscoelastic para-casein matrix undergoes limited flow andstretches into hot molten sheets without break-ing. Plasticization is accompanied by micro-structural changes in the cheddared curd, in-cluding further linearization of the paracaseinmatrix into fibers and coalescence of fat intoelongated pools trapped between and showingthe same orientation as the protein fibers (Fig-ure 7-4).

Figure 7—4 Confocal laser scanning micrographs of Mozzarella cheese curd at various stages of manufacture:after cutting (A), at whey drainage (B), after cheddaring (C), and after plasticization (D). The white-gray areasrepresent the paracasein matrix, and the black areas represent the occluded fat and moisture phases. Bar = 25 um.

Page 164: Cheese Science

The physicochemical changes responsible forplasticization of the curd have not been fully elu-cidated. However, based on the behavior of curdsof different composition and pH when subjectedto texturization (Guinee, unpublished results),the microstructural changes that accompanyplacticization (e.g., Figure 7-4), and the vis-coelastic changes in curd when heated to tem-peratures similar to those during the plasticiza-tion process (Guinee et al., 1998), it may bespeculated that successful texturization is a con-sequence of

• an adequate degree of casein hydration inthe cheddared curd, which is controlled byits pH, its total calcium content, and the ra-tio of soluble to micellar Ca

• heat-induced coalescence of free fat(formed as a consequence of shearing of thefat globule membrane), which lubricatesthe flow of the paracasein matrix

• extension and shear stresses applied to thecurd, which assist in the displacement ofcontiguous planes of the paracasein matrix(see Chapter 13)

The relationship between paracasein hydra-tion and pH may be explained by the dominanceof two opposing factors over the pH range 6.0-5.0:

1. neutralization of negative charges, whichleads to contraction of the paracasein andthereby limits hydration and impedes theflow of the paracasein matrix

2. solubilization of micellar calcium, whichis conducive to casein hydration and pro-motes the flow of the paracasein matrix

At pH values in the range 6.0-5.2, solubiliza-tion of micellar calcium appears to be dominant,as decreasing pH results in an increase in hydra-tion of paracasein. In contrast, the reduction in pHis the dominant factor at pH 5.2-4.6, as decreas-ing pH results in a marked decrease in paracaseinhydration. The total calcium content of the curd,which is controlled mainly by the pH of the milkat setting and that of the curd at whey drainage,determines the curd pH at which plasticization ispossible. In the normal manufacture of Mozza-rella, the milk is typically set at pH 6.55, the wheyis drained at pH 6.15, and the ideal pH for plasti-

Figure 7-5 Physicochemical changes in cheese curd during cheddaring.

Cheddared curd

Limited flow of para-casein matrix

Increase in para-casein hydration

Solubilization of micellar calcium phosphateCa9(PCXO6 -^- 9Ca2++ 6PO4

3'Increase in ratio of soluble to micellar CaDecrease in casein-bound Ca

Pressure on curd dueto its own weight

Piling of curdLactose - - lactic aciddecrease in pH, 6.15 —^* 5.15

Curd (viscoelastic solid)

Page 165: Cheese Science

Figure 7-6 Plasticization of low-moisture Mozzarella. (A) Kneading-plasticizing equipment (from Costruzioni Meccaniche e Tecnologia S.p.a.,Perveragno, Cuneo, Italy) consisting of a hot water heating unit (1); a cheese-shredding unit (2); a plasticization chamber, where the curd is kneaded andstretched in hot water by toothed arms that oscillate backwards and forwards in opposite directions (3); and an auger (4), which conveys the plasticizedcheese to the molding unit (5). (B) Initial stages of plasticization; shredded curds have not yet fused. (C) Midstage of plasticization; shredded curds havebegun to fuse but plasticization is not yet complete, as shown by the presence of lumps in the curd mass. (D) Fully plasticized molten curd mass, whichexhibits a long consistency and an oily surface sheen.

Page 166: Cheese Science

cizing cheddared curd is about 5.15. At this pH,the concentration of calcium in the curd (~ 27 mg/g protein) and the proportion of soluble calcium(~ 40% of total) ensure that the paracasein is suf-ficiently hydrated to enable successful plasticiza-tion. At increasingly higher curd pH, the curdbecomes progressively less smooth after plastici-zation, reflecting the decrease in paracasein hy-dration resulting from the reduced ratio of solublecalcium to micellar calcium. Similarly, in pro-cessed cheese manufacture, heating of cheese isaccompanied by aggregation of the protein andexudation of moisture and free fat unless emulsi-fying salts (e.g., sodium orthophosphates) areadded to chelate the micellar Ca (see Chapter 18).At a curd pH greater than 5.4, the curd fails toplasticize correctly. Instead, a nonplastic masswith a rough, dull, short, lumpy consistency isobtained. However, successful plasticizationmay be achieved at a higher curd pH (e.g., 5.6-5.8) if the Ca level of the curd is sufficiently low(e.g., < 18 mg/g protein), as in the case of directlyacidified Mozzarella.

In the manufacture of directly acidified Moz-zarella, acidification is achieved by the additionof food-grade organic acids rather than the con-version of lactose to lactic acid by starter culture.The milk pH is typically adjusted to about 5.6prior to rennet addition, and no further change inpH occurs during curd manufacture, which isotherwise similar to that for conventional Moz-zarella made using a starter culture. Followingwhey drainage, the curd, typically with a pH ofaround 5.6, plasticizes readily upon heating andstretching. The ability of curd made by directacidification to plasticize at a higher than normalpH can be explained on the basis of the interac-tive effects of total curd calcium and soluble Ca:micellar Ca ratio (which changes with pH) onparacasein hydration. While soluble Ca as a per-centage of total Ca decreases from around 40%to 20% as the pH is increased from 5.15 to 5.6,the total concentration of calcium is lower, andhence the level of micellar Ca is probably simi-lar to that obtained at pH 5.15 in conventionalcheese manufacture. Consequently, as there is

an inverse relationship between casein-boundCa and casein-bound moisture, the degree ofparacasein hydration obtained in directly acidi-fied Mozzarella curd at pH 5.6 is similar to orsomewhat greater than that in conventionallyproduced Mozzarella curd at pH 5.3. Indeed,comparative studies have shown that the water-binding capacity of directly acidified Mozza-rella cheese curd (pH 5.6) is higher than that ofconventionally produced Mozzarella curd (pH5.2) during the first 3 weeks of aging (Kindstedt& Guo, 1997).

7.7 MOLDING AND PRESSING OFCHEESE CURD

At some stage in the manufacturing process(e.g., just after coagulation for Camembert, aftercooking for Emmental, or after acidification forCheddar), the curds are transferred to molds of thecheese's characteristic shape and size. The prin-cipal purpose of molding is to allow the curd toform a continuous mass; matting of high-mois-ture curds occurs readily under their own weightbut pressing is required for low-moisture cheese.It is important that the curds are warm duringpressing, especially for low-moisture cheeses.

Various pressing systems have been devel-oped, ranging from the very simple to the con-tinuous-pressing systems. In modern Cheddarcheese factories, the salted curds are formed andpressed under their own weight and under aslight vacuum in towers (Wincanton towers) forabout 30 min. Upon exiting the tower, the col-umn of curd is cut into 20 kg blocks by a guillo-tine, and the blocks are placed in plastic bagssealed under vacuum. For examples of cheese-pressing systems, the reader is referred toKosikowski and Mistry (1997) or Robinson andWilbey (1998) or to any text on cheese technol-ogy referenced in the "Suggested Readings" sec-tion of Chapter 1.

Cheeses are made up in characteristic shapesand sizes (see Chapter 17). At first glance, it mightappear that the shape and size of a cheese is cos-metic. While this may be so in many cases, size

Page 167: Cheese Science

and shape are very significant in some varieties.For example, surface-ripened cheeses (mold orsmear) are made up in small, low cylinders toallow ripening from the surface toward the center.If such a cheese were made large, the surfacewould become overripe while the center re-mained unripe. For cheeses with large eyes (e.g.,Emmental), a large size is required, as otherwisethe leakage of CO2 through the surface would beexcessive and the pressure of gas within thecheese would not build up to the level needed toform eyes.

7.8 PACKAGING

Like other sectors of the food industry, in-deed, of industry in general, packaging has be-come a major feature of cheese production, dis-tribution, and retailing. Kosikowski and Mistry(1997) include a useful chapter on various as-pects of the packaging of cheese and fermentedmilks. Kadoya (1990) provides a more generaldiscussion of food packaging, including a chap-ter on cheese and fermented milks. The scienceand technology of packaging are specializedsubjects that will not be discussed here.

The objectives of cheese packaging, as offood packaging generally, are as follows:

• To protect the cheese against physical,chemical, or microbial contamination.Mold growth is of particular concern. Sincemolds are aerobic, their growth can be pre-vented by covering the cheese with wax orplasticote or vacuum-packing it in plasticfilm with low permeability to oxygen andfree of pin holes. By preventing contamina-tion, packaging serves a public health func-tion as well as reduces losses due to spoil-age.

• To reduce loss of moisture from the surfaceand therefore increase economic return. Toachieve this, the packaging material shouldhave low permeability to moisture.

• To prevent physical deformation of thecheese, especially soft cheeses, and thus fa-

cilitate stacking during ripening, transport,and retailing.

• To allow for product labelling and brandidentification, which in turn provides an op-portunity for advertising and the provisionof nutritional information.

After salting (see Chapter 8), those cheeseson which the growth of molds (surface or inter-nal) or of a surface smear is encouraged aretransferred to a room at a controlled temperature(« 15PC) and humidity (90-95% equilibriumrelative humidity). Even at this high humidity,some loss of moisture from the surface occurs,but the loss is insufficient to create a rind (low-moisture surface layer). After adequate growthof the mold or smear has occurred, such cheesesmay be wrapped in foil or grease-proof paper toavoid further loss of moisture.

Traditionally, the development of a rind wasencouraged on internally bacterial-ripenedcheese by controlled drying of the surface. Ifproperly formed, the rind effectively sealed offthe interior of the cheese, preventing excessiveloss of moisture and the growth of microorgan-isms on the surface. To further stabilize the sur-face of such cheeses, they were rubbed with oil(e.g., butter oil or olive oil) or coated with par-affin wax. Sometimes wax of a particular colorwas used (e.g., red for Edam, black for extra-mature Manchego and Cheddar). The color ofthe wax was characteristic of the variety or ofits maturity and was recognized by the con-sumer as an index of variety or quality.

Today, many internally bacterial-ripenedcheeses are packaged in plastic bags of low gaspermeability or coated with film-forming plasticmaterial. A variety of plastic packaging materi-als are used for cheese, including cellophane,cellophane-polyethylene, polyvinyl chloride,polyvinylidene chloride, polystyrene, polypro-pylene, ethylene vinyl acetate, co-extrudedpolyolifm, metal foils, and paper.

Gases such as CO2 and H2S are produced inmany cheeses during ripening. CO2 will causebulging of the package, while H2S has an obnox-

Page 168: Cheese Science

ious aroma that will render the cheese unaccept-able. To avoid such problems, the packageshould be permeable to these gases.

Packaging is particularly important for softcheeses, such as Cottage cheese, Quarg, Creamcheese, and processed cheeses. Metal foils arewidely used for consumer or catering packagesof processed cheese. Much processed cheese iscommercialized as individual slices wrapped inplastic material. High-moisture fresh cheeses arecommercialized in plastic tubs; plastic-, wax-, or

REFERENCES

Green, M.L., & Grandison, A.S. (1993). Secondary (non-en-zymatic) phase of rennet coagulation and post-coagula-tion phenomena. In P.P. Fox (Ed.), Cheese: Chemistry,physics and microbiology (2d ed., Vol. 1). London:Chapman & Hall.

Guinee, T.P., Auty, M.A.E., Harrington, D., Corcoran,M.O., Mullins, C., & Mulholland, E.G. (1998). Character-istics of different cheeses used in pizza pie (Abstract).Australian Journal of Dairy Technology, 53, 109.

Kadoya, T. (1990). Food packaging. San Diego: AcademicPress.

Kindstedt, P.S., & Guo, M.R. (1997). Chemically-acidifiedpizza cheese production and functionality. In T.M.

foil-lined cardboard containers; and plasticpackages.

Metal cans or glass jars may be used to pack-age natural and processed cheese to offer a nov-elty presentation feature or, in the case of cans,to provide extra physical protection during dis-tribution and storage.

As with other foods, the packaging of cheesehas led to the development of specialized pack-aging equipment, much of which is highly auto-mated and computerized.

Cogan, P.P. Fox, & R.P. Ross (Eds.), Proceedings of theFifth Cheese Symposium. Dublin: Teagasc.

Kosikowski, F.V., & Mistry, V.V. (1997). Cheese and fer-mented milk foods (3d ed.). Westport, CT: F.V.Kosikowski, LLC.

Robinson, R.K., & Wilbey, R.A. (1998). Cheesemakingpractice (3d ed.). Gaithersburg, MD: Aspen Publishers.

Sood, S.M., Sidhu, K.S., & Dewan, R.K. (1980). Volumin-osity and hydration of casein micelles from abnormalmilks. New Zealand Journal of Dairy Science and Tech-nology, 15, 29-35.

Walstra, P. (1993). The syneresis of curd. In P.F. Fox (Ed.),Cheese: Chemistry, physics and microbiology (2d ed.,Vol. 1). London: Chapman & Hall.

Page 169: Cheese Science

8.1 INTRODUCTION

The use of salt (NaCl) as a food preservativedates from prehistoric times and, together withfermentation and dehydration by exposure tolow-humidity air, is one of the classical methodsof food preservation. Until the development inthe 19th century of modern methods such as pas-teurization or sterilization, chilling or freezing,and "hot air" drying, salting was probably themost widely used method for the long-term pres-ervation of many foods. Salt was a highly valueditem of trade and was exchanged for goods andservices. One can readily envisage how fermen-tation and "natural" dehydration could havebeen discovered by accident, but the use of saltas a preservative required direct intervention andwas a very significant discovery at an early stageof human civilization. The three classical meth-ods of food preservation, fermentation, salting,and dehydration, along with refrigeration, areused to preserve cheese and/or control its matu-ration.

The preservative action of NaCl is due to itseffect on the water activity (aw) of the medium:

aw=p/p0

where/* and/?0 are the vapor pressure of the wa-ter in a system and of pure water, respectively. Ifthe system is at equilibrium with its gaseous at-mosphere, then aw = ERHI100, where ERH is theequilibrium relative humidity.

Due to the presence of various solutes infoods, the vapor pressure of water in a food sys-tem is always less than that of pure water (i.e., aw

< 1.0). The relationship between water activityand the moisture content of food is shown (Fig-ure 8-1). Three zones are usually evident:

• Zone I represents monolayer water that istightly bound to polar groups in the food,such as the -OH group of carbohydrates andthe -NH3

+ and -COO" groups of proteins.• Zone II consists of multilayer water in addi-

tion to the monolayer water.• Zone III contains bulk phase water in addi-

tion to monolayer and multilayer water.

General discussions on the general concept ofwater activity in relation to foods are providedby Fennema (1996), Rockland and Beuchat(1987), and Rockland and Stewart (1981). Morespecific aspects of water activity in relation todairy products are discussed by Kinsella and Fox(1986) and Roos (1997).

The water activity of food depends on itsmoisture content and the concentration of lowmolecular mass solutes. The water activity ofyoung cheese is determined almost entirely bythe concentration of NaCl in the aqueous phase:

0 W = 1 - 0.033 [NaCl01] = 1 - 0.00565 [NaCl]

where [NaCl1n] is the molality of NaCl (i.e.,moles of NaCl per liter OfH2O) and [NaCl] is theconcentration of NaCl as g/100 g cheese mois-ture (Marcos, 1993).

Salting of Cheese Curd

CHAPTER 8

Page 170: Cheese Science

The salt content of cheese varies from about1.0% for Emmental to about 5% for Domiati(Table 8-1). Typical values for the aw of somecheese varieties are shown in Table 8-2. Othercompounds besides NaCl including lactic andother acids, amino acids, very small peptides,and calcium phosphate (from milk), contributeto the depression of aw, especially in extra ma-ture cheeses.

Salt increases the osmotic pressure of theaqueous phase of foods, causing dehydration ofbacterial cells and killing them or at least pre-venting their growth. It will be apparent fromFigure 8-2 that the water activity of most cheesevarieties is not low enough to prevent thegrowth of yeasts and molds, but together with a

low pH it is quite effective in controlling bacte-rial growth. Enzyme activity is also affectedstrongly by water activity (Figure 8-2).

Measurement of the salt content of cheese isan important quality control step in cheese pro-duction. As described above, the water activityof cheese can be calculated from its composi-tion, but it can also be determined experimen-tally (see Marcos, 1993, and Chapter 23).

The concentration and distribution of salt incheese have a major influence on various aspectsof cheese quality. Among the principal effects ofsalt are the following:

• Salt inhibits or retards the growth and activ-ity of microorganisms, including patho-

awFigure 8-1 Idealized relationship between the water activity (aw) of the food and its water content.

MO

IST

UR

E C

ON

TE

NT

(g

H20

/g d

ry m

atte

r)

Page 171: Cheese Science

genie and food-poisoning microorganisms(Table 8-2), and hence increases the safetyof cheese (see Chapter 20).

• It inhibits the activity of various enzymes incheese.

• It affects the syneresis of cheese curd, re-sulting in whey expulsion and thus in a re-duction in the moisture of cheese, whichalso influences the activity of microorgan-isms and enzymes.

• It causes changes in cheese proteins that in-fluence cheese texture, protein solubility,and probably protein conformation.

• It affects cheese flavor directly and indi-rectly via its influence on microorganismsand enzymes in cheese.

• High levels of salt in cheese may have un-desirable nutritional effects (see Chapter21).

Various aspects of the significance of saltin cheese are discussed comprehensively byGuinee and Fox (1993) and are summarized be-low.

8.2 SALTING OF CHEESE CURD

Cheese is salted by one of four methods. Themethod used is characteristic of the variety:

1. Dry salt is added to and mixed with smallpieces of fresh cheese curd prepared bymilling or breaking larger blocks ofcheese (as in Cheddar-type cheeses) priorto molding and pressing. This method iscommonly referred to as dry salting.

2. Dry salt or a salt slurry is rubbed on thesurface of the molded cheese. Thismethod is used for some Blue cheeses.

3. The molded cheese is immersed in a con-centrated NaCl brine (15-23%). Edam,Gouda, Emmental, and Camembert aresalted in this manner.

4. A combination of two of these methods isused for quite a few varieties. For ex-ample, milled (broken) Mozzarella curdmay be partially dry-salted before stretch-

ing and molding, then subjected to brin-ing or the surface application of dry salt.

8.2.1 Dry-Salting of Cheese

Dry-salting is used mainly for British variet-ies of cheese, such as Cheddar, Cheshire, andStilton, and American Cottage cheese. Tradi-tionally, the salt was added to the milled curdand mixed manually, but the process is nowdone mechanically, permitting better control ofthe level of salt added and its distribution. Whensmall curd pieces are salted, the concentration ofNaCl rapidly reaches a level throughout thepieces that is sufficiently high to retard thegrowth of the starter culture, and hence the pHof the curd at salting is close to the final desiredvalue in the freshly made cheese (e.g., 5.4 vs.5.1 for Cheddar curd). However, althoughgrowth of the starter ceases shortly after salting,metabolism of lactose continues, and the con-centration of lactic acid typically increases, forexample, from 0.7% to 1.5% during the first 24hr post salting. The decrease in pH during thisperiod is much smaller (e.g., from 5.4 to 5.2-5.1) than would be expected from the amount oflactic acid produced, owing to the fact that thebuffering capacity of cheese curd has a maxi-mum at about pH 5.2 and to the fact that pH islogarithmic.

It should be possible to achieve very preciseand uniform control of salt concentration in dry-salted milled cheese curd. When dry-salting isproperly executed, it is possible to get to within± 0.1% of the desired salt concentration. How-ever, if salt distribution is poor initially, uniformdistribution of salt throughout the cheese is notattained during the life of the cheese. This is be-cause each piece of curd behaves as a mini-cheese in which salt equilibrium is attainedwithin about 24 hr. Thereafter, there may be sev-eral centers of high or low salt concentrationthroughout the cheese and hence a very lowdriving force for the attainment of overall equi-librium throughout the cheese.

When dry-salting milled curd, it is easy to cal-culate the correct amount of salt that should beadded. When a continuous mechanized mixing

Page 172: Cheese Science

system is used, localized variations will occurunless adequate metering and mixing systemsare used. An effective system involves the use ofan oscillating boom that releases salt in responseto a signal indicating either the depth or weightof the bed of curd beneath the boom. However,the temperature and humidity conditions of stor-age and transportation of salt to the oscillatoryboom are critically important for uniform salt

deposition and distribution. Adequate mixing ofthe salted curd chips can be achieved using rotat-ing pegged rollers or a rotating drum. Althoughthe latter should give more uniform mixing, theformer is more common in commercial cheeseproduction.

Even when properly measured and mixed,variations in salt concentration still occur, due tovariability in the uptake of salt added to the curd.

Table 8-2 Water Activity (aw) of Some Cheese Varieties

a w Cheese

1.00 Cheese curd, Whey cheese0.99 Beaumont, Cottage cheese, Fresh cheese, Quarg0.98 Belle des Champs, Munster, Pyrenees, Processed, Taleggio0.97 Brie, Camembert, Emmental, Fontina, Limburger, St. Paulin, Serra da Estrela0.96 Appenzeller, Chaumes, Edam, Fontal, Havarti, Mimolette, Norvegia, Samso, Tilsit0.95 Bleu de Bresse, Cheddar, Gorgonzola, Gouda, Gruyere, Manchego0.94 Idiazabal, Majorero, Mozzarella, Norzola, Raclette, Romano, Sbrinz, Stilton0.93 Danablu, Edelpilzkase, Normanna, Torta del Casar0.92 Castellano, Parmesan, Roncal, Zamorano0.91 Provolone, Roquefort0.90 Cabrales, Gamalost, Gudbrandsdalsost, Primost

Table 8-1 Typical Composition of Major Cheeses

Cheese (%) Total Fat (%) Total Solids (%) Protein (%) Salt(%) Ash (%) pH (%)

Blue 29.0 58.0 21.0 4.5 6.0 6.5Brick 30.0 60.0 22.5 1.9 4.4 6.4Bulgarian White 32.3 68.0 22.0 3.5 5.3 5.0Camembert 23.0 47.5 18.5 2.5 3.8 6.9Cheddar 32.0 63.0 25.0 1.5 4.1 5.5Edam 24.0 57.0 26.1 2.0 3.0 5.7Emmental 30.5 64.5 27.5 1.2 3.5 5.6Gouda 28.5 59.0 26.5 2.0 3.0 5.8Grana (Parmesan) 25.0 69.0 36.0 2.6 5.4 5.4Gruyere 30.0 66.5 30.0 1.1 4.1 5.7Limburger 28.0 55.0 22.0 2.0 4.8 6.8Muenster 29.0 57.0 23.0 1.8 4.4 6.2Provolone 27.0 57.5 25.0 3.0 4.0 5.4Romano 24.0 77.0 35.0 5.5 10.5 5.4Roquefort 31.0 60.0 21.5 3.5 6.0 6.4Domiati3 25.0 45.0 12.0 4.8 - 4.6Feta 26.0 47.0 16.7 3.0 - 4.5

amade from buffalo milk

Page 173: Cheese Science

The principal factors that affect the uptake ofsalt by Cheddar cheese curd are summarized inFigure 8-3.

• There is nonlinear relationship between saltuptake and the amount of salt added (Figure8-4).

• Salt is lost in whey expressed from the curddue to syneresis caused by salt. The volumeof whey expressed from the curd increasesalmost linearly with the amount of saltadded, and approximately 2 kg of H2O arelost per kg of salt absorbed.

• The amount of salt lost in the whey in-creases slightly with increasing temperatureat salting.

• Increasing the duration of the salt and curdmixing has little effect on the volume ofwhey released but reduces the amount ofsalt lost and hence increases the salt andsalt-in-moisture content of the cheese.

• Salt losses are reduced substantially—andhence the salt and salt-in-moisture percent-

ages are increased—by extending the pre-pressing holding period.

• Upon salting, about 0.25 kg of fat are lostper 100 kg cheese. Loss of fat increasesmarkedly with increasing temperature.

• The level of salt lost increases with increas-ing moisture content of the cheese curd.

• Higher losses of salt occur from high-acid-ity than from low-acidity curd.

• The size of the salt crystals has little effecton salt retention. Salt retention increases asthe size of the curd particles is reduced.

• Increasing the depth of the bed of saltedcurd during holding prior to molding re-duces the level of moisture, salt, and salt-in-moisture in the cheese.

8.2.2 Brine-Salting of Cheese

When cheese is placed in brine, there is a netmovement OfNa+ and Ch from the brine into thecheese as a consequence of the difference in os-

Water activity

Figure 8-2 Generalized deterioration reaction rates in food systems as a function of water activity (aw) at roomtemperature.

Rel

ativ

e ra

te

Structural transformationsstickinesscakingcollapselactose crystallization

Growth ofmoulds

Yeasts

Bacteria

Oxidation

Diffusion-limited reactionsnon-enzymatic browning

Enzyme activityLoss of lysine

Critical aw

Page 174: Cheese Science

Table 8-3 Minimum Water Activity (aw) for theGrowth of Pathogenic Bacteria in Foods

Pathogen Minimum aw

Aeromonas hydrophila 0.970Bacillus cereus 0.930Campylobacter jejuni 0.990Clostridium botulinum A 0.940Clostridium botulinum B 0.940Clostridium botulinum E 0.965Clostridium botulinum G 0.965Clostridium perfringens 0.945Escherichia coll 0.935Listeria monocytogenes 0.920Salmonella spp. 0.940Shigella spp. 0.960Staphylococcus

aureus (anaerobic) 0.910Staphylococcus

aureus (aerobic) 0.860Vibrio parahaemolyticus 0.936Yersinia enterocolitica 0.960

motic pressure between the moisture phase ofthe cheese and the brine. Water in the cheese dif-fuses out through the cheese matrix to establishosmotic equilibrium. Cheese can be viewed as aspongelike matrix consisting of strands of fusedparacasein micelles. The properties of the inter-stitial fluid are generally not appreciably differ-ent from those of corresponding solutions.Hence, the diffusion of salt from the brinethrough the moisture phase of cheese would beexpected not to differ significantly from that ofsalt molecules through pure water (e.g., wherepure water and the salt solution are separated bya semipermeable membrane). However, modelexperiments on the brining of cheese designed toobey Pick's laws for unidimensional flow haveshown that the rate of diffusion of salt in cheesemoisture is much lower than that in pure water.The diffusion coefficient (D) of salt in cheesemoisture at 120C is roughly 0.1-0.2 cmVday,compared with 1.0 cmVday for salt in pure wa-ter. The difference is due to a number of factorsassociated with the cheese that retard and/or im-

pede the movement of salt. The factors includethese:

• The narrowness of the pores of the para-casein matrix retard the movement of Na+

and Cl" passing through them.• Fat globules and casein particles in the

cheese obstruct the passage of salt mol-ecules/ions. This circuitous route taken bythe ions increases the effective distanceover which they must move.

• The higher apparent viscosity of the cheesemoisture compared to that of pure water.

Thus, there is a strong salt concentration gra-dient within a cheese after salting, but thisgradually disappears during storage, and if rip-ening is long enough, equilibrium will be estab-lished throughout the cheese (Figure 8-5).

The absorption and diffusion of salt in cheesecurd are affected by several compositional andenvironmental factors (Figure 8-6). These arediscussed next.

8.2.3 Concentration Gradient

Although the diffusion coefficient for NaCl incheese is essentially independent of brine con-centration, the rate of NaCl uptake increases at adecreasing rate with increasing brine concentra-tion (Figure 8-7).

8.2.4 Salting Time

The quantity of salt absorbed increases withsalting time, but the rate of salt absorption de-creases with time, owing to a decrease in theNaCl concentration gradient between the cheesemoisture and the brine. The quantity of NaCltaken up by a cheese is proportional to the squareroot of brining time, t. The theoretical relation-ship for the quantity of salt absorbed through aflat surface as a function of brining time is:

M1 = 2(C-C0)(D*t/7i)1/2 Vw

where Mt is the quantity of salt absorbed over time(g NaCl/cm2), C is the concentration of NaCl inthe brine (g NaCl/ml), C0 is the original salt con-

Page 175: Cheese Science

Figure 8-3 Principal factors that affect the uptake of salt by Cheddar curd.

Salting rate, g/100g

Figure 8-4 The relationship between the salt content (•) and salt-in-moisture level ( O) of batches of curd thatwere from the same vat but salted at different levels.

Sal

t in

chee

se,

g/10

0 g S

alt-i

n-m

oist

ure,

g/1

00 g

Moisturecontentof curd

Salt uptake

Salt-in-moisture

Curd acidityat salting

Manufacturing conditions

Curd particlesize at milling

Saltinglevel

Extent ofmixing

Method ofsalt addition

Page 176: Cheese Science

centration in the cheese (g/ml cheese moisture),D* is the pseudodiffusion coefficient of NaCl incheese moisture (cm2/d), t is the duration of salt-ing period (days), and Vw is the average watercontent throughout the cheese at time t (g/g).

8.2.5 Cheese Size and Geometry

The rate of salt absorption increases with anincreasing surface area:volume ratio for thecheese. Thus, for cheeses of the same shape and

relative dimensions, the mean salt content willdecrease with increasing size after brining forequal intervals. The quantity of salt absorbedfrom the brine by a cheese depends on its shape.The quantity of NaCl absorbed per cm2 of cheesesurface is greater for a flat slab than for a sphere(and the relative reduction increases with the de-gree of curvature and duration of brining) andfor a rectangular-shaped cheese (three effectivedirections of salt penetration) than for a cylindri-cal cheese (two effective directions).

Distance from cheese centre, cm

Figure 8-5 The mean salt-in-moisture level throughout cylindrical Romano-type cheese salted in 19.5% NaClbrine at 230C for 1 day (•), 3 days (O), or 5 days (•) or salted for 5 days and stored wrapped at 1O0C for 30 days(Q) and 83 days ( A).

Sal

t-in-

moi

stur

e, g

/100

g

Page 177: Cheese Science

Manufacturing conditions

Curd dimensionsat salting

Shape Surface areato volumeratio

Size

Curd acidityat salting

Moisturecontentof curd

Salt uptake

Salt-in-moisture

Brining conditions(e.g., temperature,% NaCl)

Figure 8-6 Principal factors that affect salt uptake by brine-salted cheeses.

Sal

t-in-

moi

stur

e, g

/100

g

Brine concentration, g/100 g

Figure 8-7 Salt level in cheese slices (7 cm diameter, 0.5 cm thick) salted in brines of different concentration for200 min at 2O0C.

Page 178: Cheese Science

8.2.6 Temperature of Curd and Brine

Model brining experiments with small cheeses(curd chips) have shown that salt uptake by curdtempered at any temperature in the range 27-430C increases with increasing brining tempera-ture in the same range. However, curd temperedto 320C absorbs salt less readily than curd tem-pered to a lower or higher temperature. This ef-fect has been attributed to a layer of exuded fat onthe surface of curd particles at 320C that impedessalt uptake; less fat is exuded at temperaturesbelow 320C, while at higher temperatures ex-uded fat is liquid and disperses in the brine. In-creasing the brining temperature increases themobility of NaCl and the amount absorbed,partly because of an increase in true diffusion andpartly because of an increase in the effectivewidth of the pores in the protein matrix as non-solvent water decreases with increasing tempera-ture. The reverse situation generally occurs whencurd chips are dry-salted (as in Cheddar), wherean increase in curd temperature in the range 24-410C is paralleled by a decrease in the salt andsalt-in-moisture in the final cheese. This effect isdue to the greater expulsion of whey from thecurd as the temperature is increased, which inturn causes more salt to be lost in the whey, mak-ing less available for absorption by the curd. Incontrast, in brine-salting the cheese is immersedin brine, which is usually agitated; hence the saltavailable for absorption is not limiting.

8.2.7 Curd pH

Curd salted at a low acidity retains more saltthan more acidic curd. The opposite might beexpected, since low acidity curd normally con-tains more moisture than acidic curd and hencewill undergo greater syneresis, resulting in ahigher loss of salt. However, everything else be-ing equal, the rate of salt uptake and diffusionare higher in high-moisture curd. In addition, theprotein in high pH curd may be more salt-soluble than that in more acid curd, which mightimprove salt retention.

8.2.8 Moisture Content of Cheese Curd

The diffusion coefficient and the quantity ofsalt absorbed upon brine-salting increases as themoisture content of the curd increases (Figure8-8), which has been attributed to an increase inthe relative pore size in the protein matrix, re-sulting in less retardation of the diffusing saltmolecules/ions.

The reverse situation occurs upon dry-saltingmilled Cheddar curd. That is, as the initial mois-ture content increases, the level of salt and salt-in-moisture in the cheese decreases for a fixedsalting level. This effect has been attributed togreater losses of whey and salt from the high-moisture curds. Thus, while the diffusion of saltwithin each curd particle increases with mois-ture content, less salt is available.

8.3 EFFECT OF SALT ON CHEESECOMPOSITION

For any particular variety, there is an inverserelationship between the levels of moisture andsalt in cheese, everything else being equal. In thecase of dry-salted cheese, this reflects the syner-esis of the curd upon salting, the extent of whichis directly related to the amount of salt added tothe curd. The relative fluxes of NaCl and H2O inunidimensional brine-salted cheese are relatedas -AWx ~ p ASx, where AWx and ASx are thechanges (from the unsalted cheese) in g H2O andg NaCl, respectively, per 100 g nonsalt cheesesolids in planes of the cheese x cm from thecheese-brine interface. The experimental valueof p is about 2 (Figure 8-9). That is, the weightof water lost is about twice that of salt taken up,but the value varies from 1.5 to 2.34 (or from < 1to 3.75 in another study), depending on the dis-tance from the surface. Changes in the textureand appearance of the cheese can be seen as thesalt front moves through the cheese.

Uptake of salt during brining is sometimes ac-companied by an increase in moisture content inthe vicinity of the cheese-brine interface, espe-cially in calcium-free weak brines (< 10%, w/v,

Page 179: Cheese Science

NaCl) (Figure 8-10). Such an effect is respon-sible for the "soft-rind" defect and swelling incheese, and it is attributed to "salting-in" (solu-bilization) of the cheese protein in dilute NaClsolutions.

Higher salt concentrations are associated withincreased levels of fat and protein due to the lossof water from the cheese. There is a significantinverse relationship between fat and moisturelevels in mature Cheddar cheese.

The concentrations of lactose and lactic acidin and the pH of cheese depend on the continuedactivity of the starter and hence on salt content.

8.4 EFFECT OF NaCl ON THEMICROBIOLOGY OF CHEESE

The concentration of salt-in-cheese moisturehas a major effect on the growth of microorgan-

isms in and on the cheese (see Chapter 10).Probably the most extreme example of the use ofsalt to control bacterial growth is the Domiati-type cheese, where 8-15% NaCl is added to thecheese milk to inhibit bacterial growth andmaintain milk quality. In the manufacture ofmost cheese varieties, salt is added after curdformation.

The growth of Lactococcus strains used asstarters is stimulated by low levels of NaCl but isstrongly inhibited at levels above 5% NaCl. Indry-salted cheeses, the concentration of NaClrapidly reaches an inhibitory level throughoutthe cheese. In Cheddar-type cheese, startergrowth ceases shortly after salting, but metabo-lism of lactose continues unless the level of salt-in-moisture exceeds about 5% (Figures 8-11and 8-12). Strains of Lc. lactis ssp. lactis aremore salt tolerant than Lc. lactis ssp. cremoris:the former grow in the presence of 4% NaCl

Moisture, g/100g

Figure 8-8 Dependence of the pseudodiffusion coefficient of salt in cheese moisture (D*) on the initial moisturecontent of cheese salted in « 20% NaCl brine at 15-160C. Blue cheese (A, B), Gouda (C, D), Romano (E),Jarlsberg (F), Emmental (G, I), unsalted milled Cheddar (H).

D*.

cm2/d

ay

Page 180: Cheese Science

whereas the latter grow in the presence of 2%but not 4% NaCl. However, there is considerableinterstrain variation in both subspecies. A lowlevel of salt-in-moisture may lead to high num-bers of starter cells in Cheddar cheese, whichmay lead to bitterness.

If starter activity is inhibited after manufac-ture owing to an excessively high level of salt-in-moisture, residual lactose will be metabolizedrelatively late in ripening, when the number ofnonstarter lactic acid bacteria (NSLAB) is high.In modern Cheddar cheese, the number ofNSLAB is low initially (< 100 cfu/g) and theygrow at a rate that is largely determined by therate of cooling of the pressed curd and the ripen-ing temperature. NSLAB vary in their ability togrow in the presence of NaCl. Most strains cangrow in the presence of 6% but not 8% NaCl,and those strains that can grow in the presence of

8% NaCl are inhibited by 10% NaCl. Normally,the number of NSLAB is too low to rapidly me-tabolize the residual lactose, unless it persists forseveral weeks (see Chapter 10).

In brine-salted or dry surface-salted cheese,NaCl diffuses slowly from the surface to thecenter, and even in small cheeses salt probablydoes not reach an inhibitory concentration in theinterior until starter growth has ceased, owing todepletion of lactose or perhaps low pH. Thus,although thermophilic starters, Streptococcusthermophilus and Lactobacillus spp., are moresensitive to NaCl than Lactococcus spp., this isprobably not significant in cheese acidification,since cheeses made using these starters are usu-ally brine-salted.

Although data on the salt sensitivity ofPropi-onibacterium spp. are variable, most studiesshow that they are quite salt sensitive and cannot

Distance from cheese surface, cm

Figure 8-9 Experimental (•) and theoretical (1-4) moisture levels and experimental salt-in-moisture level (O)in a full-fat Gouda cheese (pH 5.64) after brining for about 8 days as a function of distance from the cheesesurface in contact with the brine (20.5 g NaCl/100 g H2O; temperature, 12.60C). Theoretical moisture levels werecalculated using the relationship AWx = PASx, where W and S are g H2O and g NaCl, respectively, per 100 gnonsalt cheese solids, x is the distance (cm) from the cheese surface in contact with the brine, and/? is a coeffi-cient denoted as the flux ratio. The theoretical moisture levels were calculated for/? = 2.5(1);/? varying from 1.7at the salt front (i.e., maximum distance to which salt had penetrated) to 2.9 at the cheese surface (2); p = 1 (3);and p = O (4).

Sal

t-in-

moi

stur

e, g

/10O

g H

2O

Moi

stur

e co

nten

t of c

hees

e, g

/1 O

O g

Page 181: Cheese Science

grow in the presence of 3% NaCl. The salt-in-moisture in Emmental is only around 2%.

Blue cheeses, at 3-5% NaCl, are among themost heavily salted varieties. Germination of P.roqueforti spores is stimulated by 1% NaCl butinhibited by 3-6% NaCl, depending on thestrain. Germinated spores can grow in the pres-ence of up to 10% NaCl. It is fairly commoncommercial practice to add 1% NaCl directly toBlue cheese curd, perhaps to stimulate sporegermination, although it also serves to give thecheese a more open structure, which facilitatesmold growth. Since most Blue cheeses are drysurface-salted, a salt gradient from the surfaceto the center exists for a considerable period af-ter manufacture, and the salt-in-moisture per-centage in the outer layer of the cheese may behigh enough to inhibit spore germination duringa critical period, resulting in a mold-free zone onthe outside of the cheese.

Growth of P. camemberti is also stimulatedby low levels of NaCl. Mold growth on Cam-

embert cheese is poor and patchy at levels below0.8% NaCl.

Since smear-ripened cheeses are brine-salted,a salt gradient from the surface to the centerexists initially. However, most of these cheesesare relatively small and have a relatively highmoisture content. Therefore, salt equilibratesthroughout the cheese relatively quickly. Thesecheeses are also rubbed with brine occasionallyto distribute the microorganisms evenly over thesurface. The surface microflora of these cheesesis very complex, but the principal microorgan-isms are yeasts and coryneform bacteria, both ofwhich are quite salt tolerant (see Chapter 10).

8.5 INFLUENCE OF NaCl ON ENZYMESIN CHEESE

8.5.1 Coagulant

With the exception of high-cooked cheeses,such as Emmental and Parmesan, in which the

Distance from cheese surface, cm

Figure 8-10 Moisture content (open symbols) and salt-in-moisture concentration (black symbols) in Goudacheese as a function of distance from the salting surface after brine-salting for 4 days at 2O0C in 5 (^, ), 12 (A,A), 20 (Q, •), and 24.8 (O, •) percent NaCl solution (without calcium).

Sal

t-in-

chee

se m

oist

ure,

g/1

00 g

Moi

stur

e, g

/10O

g

Page 182: Cheese Science

rennet is denatured extensively during cooking,primary proteolysis is catalyzed mainly by theresidual coagulant (see Chapter 11). Althoughchymosin, pepsins, and Rhizomucor proteinasesreadily hydrolyze (3-casein in solution, ocsr

casein is the principal substrate in cheese.P-Casein is less susceptible to hydrolysis by thecoagulant, probably mainly because of hydro-phobic interactions between adjacent C-terminalregions, which contain the primary chymosin-susceptible bonds (see Chapter 11); these inter-actions are intensified at high ionic strength. Theconcentration of salt in cheese has a large effecton the rate of proteolysis (see Chapter 11).

8.5.2 Milk Proteinases

The principal indigenous proteinase in milk,plasmin, contributes to proteolysis in all cheesevarieties that have been studied, as indicated bythe formation of y-caseins. It is a major contribu-tor to proteolysis in high-cooked cheeses, owing

to partial or complete inactivation of the coagu-lant. Plasmin is associated with the casein mi-celles in milk and is incorporated into cheesecurd. The activity of plasmin in cheese is stimu-lated by low levels of NaCl, up to a maximum at2%, but it is inhibited by higher concentrations,although some activity remains at 8% NaCl. Theinfluence of NaCl on the activity of the indig-enous acid milk proteinase (cathepsin D) has notbeen investigated.

8.5.3 Microbial Enzymes

The effect of NaCl on the stability and activityof microbial enzymes, especially in the cheeseenvironment, has received little attention andappears to warrant research.

8.6 EFFECT OF SALT ON CHEESEQUALITY

Considering that salt has a major influence onthe microbiology, enzymology, pH, and mois-

Salt-in-moisture, g/100 g

Figure 8-11 Effect of salt-in-moisture concentration on the concentration of lactose (A) and pH (•) within asingle block of Cheddar cheese analyzed 14 days after manufacture.

Lact

ose,

g/1

00 g

che

ese

PH

Page 183: Cheese Science

ture content of cheese, it is not surprising that theconcentration of salt in cheese has a major effecton its quality (Figure 8-12). Ripening is retardedat high salt concentrations, whereas defects suchas bitterness are common at low concentrations.The optimum concentration for Cheddar is about5% salt-in-moisture. Although the impact ofNaCl concentration on cheese quality is wellrecognized, its effects at the molecular level arenot known. It is likely that high concentrationsof NaCl retard ripening through a general inhibi-tory effect on several enzymes in cheese. Highconcentrations of NaCl (e.g., > 8% salt-in-mois-ture in Cheddar) probably inhibit the growth ofNSLAB, but concentrations in the range nor-mally encountered in Cheddar appear to havelittle or no effect. Flavor defects encountered at

low salt concentrations probably arise from ex-cessive or unbalanced enzyme activity. For ex-ample, bitterness can occur in Dutch-typecheeses if excessive proteolysis of (3-casein bychymosin occurs, which releases bitter C-termi-nal peptides (e.g., (3-CN fl93-209).

NaCl makes a direct positive contribution tocheese flavor, as most consumers appreciate asalty taste in foods. Salt-free cheese has a ratherinsipid, watery taste; 0.8% NaCl is sufficient toovercome this defect.

8.7 NUTRITIONAL ASPECTS OF NaCl INCHEESE

High intake of salt in the diet is undesirable,since it increases hypertension and the risk of

Salt-in-moisture, g/100 g

Figure 8-12 Effect of salt-in-moisture levels on pH (O) at 8 weeks and the total grade score (maximum 30) (•)for Cheddar cheese made from curd from the same vat but salted at different levels.

Tota

l gra

de s

core

(max

imum

30

poin

ts)

pH

Page 184: Cheese Science

osteoporosis via increased excretion of calcium.Sodium rather than chloride is the responsibleagent (see Chapter 21). Even in countries with ahigh consumption of cheese, cheese contributesonly about 5% of total sodium intake. Neverthe-less, there has been considerable interest in theproduction of reduced-sodium cheeses. Ap-proaches used include

• reducing the level of salt added (the degreeof reduction is limited owing to the devel-opment of off-flavors in low-salt cheese)

• replacing some of the NaCl by KCl, MgCl2,

REFERENCES

Fennema, O.R. (Ed.). (1996). Food chemistry (3d ed.). NewYork: Marcel Dekker.

Guinee, T.P., & Fox, P.P. (1993). Salt in cheese: Physical,chemical and biological aspects. In P.P. Fox (Ed.),Cheese: Chemistry, physics and microbiology (2d ed.,Vol. 1). London: Chapman & Hall.

Kinsella, J.E., & Fox, P.P. (1986). Water sorption by pro-teins: Milk and whey proteins. CRC Critical Reviews inFood Science and Nutrition, 24, 91-139.

Marcos, A. (1993). Water activity in cheese. In P.P. Fox

or CaCl2 (about 50% of the NaCl may bereplaced by KCl without undesirable conse-quences, but higher levels of replacementlead to bitterness caused by KCl)

• use of flavor enhancers to mask defects

Processed cheeses and cheese products con-tain much higher levels of Na than naturalcheeses owing to the addition of sodium-rich"emulsifying" salts (see Chapter 18). Therewould appear to be greater opportunities to re-duce the concentration of Na in these productsthan in natural cheeses.

(Ed.), Cheese: Chemistry, physics and microbiology (2ded., Vol. 1). London: Chapman & Hall.

Rockland, L.B., & Beuchat, L.R. (Eds.). (1987). Water activ-ity: Theory and applications to food. New York: MarcelDekker.

Rockland, L.B., & Stewart, G.F. (1981). Water activity: In-fluences on food quality. New York: Academic Press.

Roos, Y. (1997). Water activity in milk products. In P.P. Fox(Ed.), Advanced dairy chemistry: Vol. 3. Lactose, water,salts and vitamins. London: Chapman & Hall.

Page 185: Cheese Science

9.1 INTRODUCTION

World production of cheese is 15 milliontonnes per annum, with an estimated value of$55 billion. Approximately 7% of world cheeseis traded on the global market, the major suppli-ers being the European Union (~ 50%), NewZealand (« 16%), and Australia (« 11%). Theyield of cheese and its control are of great eco-nomic importance, determining the profit ofcheese plants and the price of milk accruing tofarmers. Owing to its economic importance,cheese yield and the factors that affect it havebeen investigated extensively, and several com-prehensive reviews on the subject have beenpublished (International Dairy Federation [IDF]1991, 1994; Lucey & Kelly, 1994).

9.2 DEFINITION OF CHEESE YIELD

The definition of cheese yield is important fortwo main applications:

1. measuring the efficiency of and determin-ing the economic viability of a cheese-making operation

2. measuring the results of experiments,which is essential for evaluating the po-tential usefulness of a particular processor change in technology

Yield may be expressed in many ways, as dis-cussed below. The format used is generally de-

termined by the needs of the particular situation.Cheese yield may be expressed simply as thequantity of cheese of a given dry matter pro-duced from a given quantity of milk with a de-fined protein and fat content (kg/100 kg milk).Actual cheese yield (Ya) is often loosely ex-pressed as the "kilogram of cheese per 100 kilo-grams of milk" or "percent yield." An alternate,frequently used way of stating cheese yield is"the number of liters of milk (of a given compo-sition) required to manufacture one tonne ofcheese," which in the case of Cheddar cheese isaround 10,000 liters. Such a definition of cheeseyield is suitable only when one variety of cheeseis being manufactured to a relatively constantcomposition from milk with a relatively constantcomposition. However, the composition of milkand hence the yield of cheese from a given quan-tity of milk can vary depending on several fac-tors, including species (e.g., cow, goat, orsheep), breed, stage of lactation, plane of nutri-tion, lactation number, and animal health. More-over, the composition of cheese milk may be al-tered by technological interventions, includingthe following:

• Standardization. Standardization is a pro-cess whereby the casein:fat (CF) ratio is ad-justed to produce cheese of the requiredcomposition. As milk is generally standard-ized to a different CF ratio for each cheesevariety, it is necessary to state the milkcomposition when expressing the yield of

Cheese Yield

CHAPTER 9

Page 186: Cheese Science

different varieties from a given quantity ofmilk. Hence, the yield of Cheddar may beexpressed as kg/100 kg of milk with a fatcontent of 33 g/kg and a protein content of,for example, 31.7 g/kg.

• Low concentration factor ultrafiltration(LCF-UF) or fortification with reconsti-tuted extra-low heat skim milk powder. Ul-trafiltration and fortification are undertakento increase and maintain the level of caseinat a fixed value (e.g., 38 g/kg) throughoutthe year. LCF-UF, which is widely prac-ticed, is a very effective method for prepar-ing milk of a more uniform compositionand hence producing cheese of more con-sistent quality, especially in regions wherelarge variations in milk composition occurthroughout the manufacturing season.

• Pasteurization practices. For a given levelof protein in the standardized milk, the ef-fective protein level (i.e., the level of gel-forming protein that is potentially recover-able in cheese curd) may be increased bychanging the heat treatment of the milk. Atnormal pasteurization temperature (720C x15 s), a low level of whey protein is dena-tured (~ 5% of total), depending on the lev-els and proportions of different whey pro-teins present. It is generally assumed thatthe denatured whey proteins complex withK-casein and are retained in the cheese curd(see Chapter 6). As the severity of the heattreatment is increased, the extent of wheyprotein denaruration and hence the effectivemilk protein level increase and contribute toincreased cheese yield.

Likewise, the composition of cheese differswith variety. Hence, a more precise definition ofcheese yield is "kilograms of actual cheese type(e.g., Cheddar) per 100 kilograms of milk con-taining specified levels of fat and protein (orpreferably casein)." When comparing actualcheese yield from milks of different compositionusing the latter definition, yield may be definedas "kilograms of cheese type (e.g., Cheddar) per100 kilograms of milk adjusted to a standard

concentration of casein (or protein) plus fat" (ab-breviated YaCFAM).

The composition of most cheese varieties mustfall within certain specifications prescribed bynational or international standards of identity(e.g., for Cheddar cheese in the United States, fat> 330 g/kg; moisture < 390 g/kg). However,intravarietal differences in composition areusual. The moisture content of Cheddar cheese,for instance, typically varies from around 350 to380 g/kg. Table cheese and most cheese suppliedto the food service and catering industries are soldon the basis of total weight. Hence, increasing themoisture content to maximize cheese yield is de-sirable, provided that the composition of thecheese is within legal specifications and thatquality is not impaired. Generally, for mostcheese varieties, there are defined bands for thevarious compositional parameters (moisture,moisture in nonfat substances, fat in dry matter,pH, and Ca) that give optimum cheese quality(see Chapter 14). In some applications, cheese isincorporated, along with water (and other op-tional ingredients), as a formulation ingredient inthe preparation of another food, such as pasteur-ized processed cheese products and cheese pow-ders. In these applications, a low moisture con-tent (e.g., 330-340 g/kg for Cheddar) is oftenpreferred and specified, as it leads to a significantreduction in transport costs, especially when thecheese is transported over long distances. Theratio of moisture to cheese solids in the final prod-uct is easily regulated by adding water during theformulation of these products. Hence, for theseapplications, buyers prefer to purchase cheese ona dry weight basis and cheese yield is best ex-pressed as moisture-adjusted cheese yield, that is,kilograms of moisture-adjusted cheese type (e.g.,380 g/kg moisture Cheddar) per 1OO kilograms ofmilk adjusted for casein (e.g., 25 g/kg) and fat(e.g., 33.3 g/kg fat), or as dry matter cheese yield,that is, kilograms of cheese solids per 100 kilo-grams of milk adjusted for casein (e.g., 25 g/kg)and fat (e.g., 33.3 g/kg fat).

The percentage recovery of a particular com-ponent (e.g., milk fat) in cheese affects cheeseyield and determines the efficiency of the

Page 187: Cheese Science

cheesemaking operation. Moreover, informationon the recovery of fat and casein is useful, as itmay help trace the cause(s) of high fat losses,such as inadequate curd firmness at cutting andblunt curd knives. Hence, the amounts of milkfat, protein, and/or casein recovered are used asindirect measures of yield. Alternatively, thelevels of fat and curd fines in the whey, whichrepresent unrecovered milk solids, may also beused as an index of cheesemaking efficiency andprovide indirect information on cheese yield.Curd fines (curd dust) are fragments of curd bro-ken off the curd particles during cutting and/orthe initial phases of stirring.

From the foregoing, it is clear that cheeseyield may be expressed in different ways, de-pending on milk composition and the use of thefinal cheese. The definition of yield for a par-ticular cheese plant should be chosen to ensuremaximum profitability. However, there may bedifferences between cheese yield and profitabil-ity, depending on plant type and cheese variety.Maximization of cheese yield is profitable onlyif the cost of implementing new procedures isreasonable and savings are significant. Hence,increasing fat recovery improves cheese yieldbut scarcely affects the profitability of cheese-making, as whey cream, recovered by separationof whey, has almost the same value as sweetcream. Similarly, improving the recovery offines from whey may improve cheese yield, butthe capital cost of new equipment may be toohigh to improve profitability, at least in the shortterm. Formulae for the determination of cheeseyield and recoveries are described in the nextsection.

9.3 MEASUREMENT OF CHEESE YIELDAND EFFICIENCY

The determination of actual cheese yield re-quires measurement of the weight of all inputsand outputs. A typical mass balance for the ex-perimental production of Cheddar is presentedin Table 9-1.

In pilot-scale cheesemaking experiments, anaccurate mass balance is easily achievable be-

Table 9-1 Typical Mass Balance for a Full-FatCheddar Cheese

Weight (kg)

Inputs

Pasteurized milk 454.8Starter 6.37Rennet solution 1.00Salt 1.44Fat in cheese milk + starter 16.41Protein in cheese milk + starter 16.46Total weight of inputs 463.61Weight of fat + protein 32.87

Outputs

Cheese 46.98Bulk wheya 409.86White wheyb 5.97Fat in cheese 14.55Fat in bulk whey 1.71Fat in white whey 0.14Protein in cheese 12.46Protein in bulk whey 3.92Protein in white whey 0.06Total weight of outputs 462.81Weight of fat + protein 32.84

a Bulk whey is whey removed at whey drainage and duringcheddaring.

b White whey is whey expressed during salting and pressing.

cause of the batch nature and small scale, whichfacilitate weighing of all materials. The fitting ofload cells to pilot-scale cheese vats further sim-plifies and increases the accuracy of mass bal-ances in experimental cheesemaking. In con-trast, achieving a mass balance (especially forindividual vats) in commercial cheesemaking ismore difficult because of the continuous natureof the different operations (e.g., overlapping ofcurds from two or more vats on the drainage beltor cheddaring tower of a Cheddar Master sys-tem). Hence, at the industrial level, a mass bal-ance tends to be performed on a day's produc-tion. Quantities of milk, starter, and whey areusually measured using on-line flow meters.

The actual cheese yield may be then calcu-lated using Equation 9.1:

Page 188: Cheese Science

[Equation 9.1]

A i • u /v \ innf Weight of chcCSC ^Actual yield (Y3) = 100^ weight of milk + starter culture + salt J

The units of actual yield are usually kg/100kg. An alternate term is percent yield.

While each cheese variety has a maximumpermitted moisture content (e.g., 390 g/kg forCheddar), variations in moisture content for agiven variety are common. Comparison of theactual yield of a given variety from milk of agiven composition may thus reflect differencesin both moisture content and/or recovery of milkconstituents. Hence, comparison of actualyields, when moisture levels are different, mayconceal inefficiencies in the recovery of milkcomponents such as fat or protein. However, it issometimes desired to compare the yield of twoor more batches of a given variety of cheese butwith a different moisture content and from milkof a given composition. In this situation, adjust-ing the moisture content of the different batchesto a reference or desired value eliminates the ef-fects of variations in moisture content on yield.The resulting yield expression is termed "mois-ture-adjusted cheese yield" (MACY):

[Equation 9.2]

MACY (kg/100 kg) =

, . ../ 100-actual cheese moisture content )(actual yield)

1^100 - reference cheese moisture content^

Many factors affect the yield of a particularvariety of cheese, including the milk composi-tion, the moisture content of the cheese, thecheesemaking process, and the type of plantequipment, as discussed in Section 9.5. The lat-ter two factors influence cheese yield, since theyaffect the recovery of milk fat and protein in thecheese. The actual yield is also influenced by thevariety of cheese being manufactured, which de-termines the moisture content, and the numberof operations where fat and protein may be lost.In the manufacture of Cheddar cheese, the saltedcurd is vacuum packed in polyethylene linersfollowing pressing, and little or no further loss ofmoisture, fat, or protein occurs during storage.In contrast, brine-salted cheeses lose moisture

during salting. The weight of moisture lost isabout twice that of salt absorbed. Moreover,other constituents, e.g., soluble nitrogenous ma-terial, Ca and fat, may be lost to a greater orlesser degree, depending on the temperature andthe duration of brining. For Gouda cheese, thereis an approximate net weight loss of about 3%during salting. Moreover, moisture may be lostfrom brine-salted cheeses during storage, de-pending on shape and size, the type of packagingmaterial (e.g., natural rind, plasticoat, or wax),storage humidity, and temperature. In wheel-shaped Romano-type cheese («2.8 kg) with anatural rind, the moisture content decreases byabout 50 g/kg during storage (70C, 85% relativehumidity [RH]) for 120 days.

The recovery of milk components (fat andprotein or casein) may also be determined whentheir concentrations in the inputs (milk andstarter) and outputs (cheese and whey) areknown. This enables the recovery or loss of fat,casein, nonfat solids, and/or protein to be calcu-lated using Equations 9.3 and 9.4.

[Equation 9.3]

Percentage of fat recovered in cheese =

(weight of cheese)(fat content of cheese) 1IUU

(weight of milk)(fat content of milk + starter)]

[Equation 9.4]

Percentage of fat lost in whey =

F (weight of whey)(fat content of whey) "I[(weight of milk + starter)(fat content of milk + starter)]

The recovery of fat and protein for a particularvariety of cheese is influenced by many factors,as described in Section 9.5. Fat recovery is verydependent on the cheese variety being manufac-tured, which determines the number and types ofoperations where fat may be lost. In industrialCheddar cheese manufacture, about 8.5% of to-tal fat is lost in the whey, of which about 76%,17%, 5%, and 2% are lost in the whey from thecheese vat, cheddaring tower, salting belt, andblock former, respectively. Similar levels of fatloss have been reported during the commercialmanufacture of Gouda and Emmental. In batch

Page 189: Cheese Science

pilot-scale manufacture of low-moisture Mozza-rella cheese curd, a higher percentage milk fat islost than for Cheddar (« 20 vs. - 8.5%) due tohigh losses during kneading and stretching of thecurd in hot water (about 8O0C); « 60% of the to-tal fat lost during the manufacture of low-mois-ture Mozzarella cheese curd occurs in the stretchwater. Typically, 25-27% of total protein is lostduring cheese manufacture, about 71%, 14%,and 14% of which is whey protein, casein, andnonprotein nitrogen (expressed as protein), re-spectively. Loss of casein (~ 5% of total casein)ensues mainly from the glycomacropepeptide,which is soluble in whey following its releasefrom the K-casein by rennet.

It may be of interest to compare the Ya orMACY obtained on different days or differentyears. Such comparison enables a cheese com-pany to monitor its efficiency over time. How-ever, when the composition of milk changes, es-pecially the levels of fat and casein, during thecheesemaking season, then comparison of Ya orMACY for a particular variety at different timeshas little value as an indicator of the efficiencyof cheesemaking. The differences in yield due todifferent recoveries (e.g., of fat) could be com-pletely masked by those that occur as a result ofdifferences in milk composition. In this situa-tion, if the composition of the cheesemilk (pro-tein and fat) and the cheese (protein, fat, andsalt) at the different manufacturing times isknown, the yield of cheese at the different timesmay be meaningfully compared by adjusting theprotein and fat content of the milk to referencevalues, as in Equation 9.5. The resulting yieldexpression is termed the moisture-adjustedcheese yield/100 kg milk adjusted for proteinand fat (MACYPFAM):

[Equation 9.5]

MACYPFAM (kg/100 kg) =

MACY 7 mI 100 x (actual content of protein and fat in milk + starter)^ content of protein and fat in reference milk + starter J

In Equation 9.5, it is assumed that casein, as apercentage of total protein, does not change overtime. However, if casein concentration changes

over time, substitution of casein for protein inEquation 9.5 allows comparison of milks withdifferent levels of casein.

9.4 PREDICTION OF CHEESE YIELD

Predictive yield formulae are used to estimatethe yield of a particular variety of cheese frommilk of a given composition. Prediction ofcheese yield is useful in that it allows a cheeseplant to

• measure its efficiency by comparing actualand predicted yields

• plan production (i.e., capacity and technol-ogy) in the event of an anticipated increasein milk supply

• plan product mix and milk-pricing schemes

Predictive yield formulae for a particular vari-ety are compiled on the basis of information ob-tained from

• cheese yield experiments in which yieldand component recovery are related to milkcomposition

• theoretical consideration of the cheese-making process's influence on the partitionof the various components of milk (e.g.,loss of glycomacropeptide, milk salts, andfat) between the curd and whey

Predictive yield formulae and their applica-tion to different types of cheese, especiallyCheddar, have been reviewed recently (see IDF,1994). Predictive formulae are of the followinggeneral type: Y= aF + bC or Y= aF + bC + k,where y = yield; F and C are the fat content andcasein contents of the milk (with added starterculture), respectively; A: is a constant, the magni-tude of which depends on the loss of casein andthe levels of nonfat, noncasein solids in thecheese; and a and b are coefficients, the magni-tude of which depends on the contributions of fatand casein to yield. Probably the simplest andmost widely applied formula for predicting theyield of different cheese varieties is that of vanSlyke, which was developed for Cheddar cheese

Page 190: Cheese Science

in 1936. The van Slyke yield formulae for actualand moisture-adjusted cheese yields are

[Equation 9.6]

[FJC (%FR/100) -C-a] x b

[actual moisture I{ Too J

[Equation 9.7]

_ [FX (%FR 1100) - C - a] x b^MACY = 7 ~ T~~ \

I reference moisture\ 1 0 0 J

where, F and C are the fat content and caseincontent of the cheesemilk (with added starterculture), %FR is the fat recovery, a is the coeffi-cient for casein loss, and b is the coefficient toaccount for cheese solids nonfat, nonprotein(SNFC). The values for %FR/1009 a, and b forCheddar cheese, as predicted by van Slyke, are0.93,0.1, and 1.09, respectively. These formulaecan be rewritten in the format Y = aF + bC,where the values of coefficients a and b are 1.66and 1.78, respectively, for Cheddar cheese con-taining 39Og moisture/kg. The van Slyke for-mula has been modified for other cheese types,based on results from cheesemaking experi-ments. For low-moisture Mozzarella, reportedvalues of %FR/WQ, a, and b are typically 0.86,0.36, and 1.09, respectively. The lower values of%FR/100 and a for low-moisture Mozzarellacompared with Cheddar reflect the higher lossesof fat and casein in the hot water used to heat thecurd during stretching. For commercial FinnishEdam and Emmental cheeses, the mean valuesof %FR/\00 and the coefficient a were 88.7 and88.1, and 0.13 and 0.15, respectively. There isconsiderable intravarietal variation in the re-ported values of %FR/IQQ and the coefficients aand b. For example, the %FR/IQO for Cheddarcheese has been found to range from about 83 to93. Discrepancies between studies undoubtedlyreflect differences in milk composition, milkquality and storage conditions, milk heat treat-

ment, cheesemaking conditions, and cheese-making technology (see Section 9.5).

Variations in the coefficients also occur be-tween cheese plants due to the above factors,which cause interplant differences in efficiency.Moreover, deviations between the actual andpredicted yields may vary between plants.Hence, the application of a generic predictivecheese yield formula for a given cheese varietymay not accurately predict cheese yield in allplants. As an alternative to generic predictiveformulae, plant-specific formulae may be devel-oped for each factory. Plant-specific yield for-mulae may be developed by the statistical analy-sis of historical data (collected weekly ormonthly) on milk composition, milk quality(e.g., somatic cell count), fat and protein recov-ery, and cheese moisture. Plant-specific formu-lae tend to give very accurate predictions ofcheese yield because they reflect, more than ge-neric prediction formulae, the composition andquality of the milk, and the actual cheesemakingconditions used in the plant. The formulae areuseful in that they compare the actual yield withthat expected from a given weight of milk of agiven composition using a given process. If ac-tual yield is less than the predicted yield, reme-dial action is taken to redress process inefficien-cies. However, a close relationship betweenactual and predicted yields does not indicate thata cheese plant is operating at maximum effi-ciency. Hence, a plant-specific yield formulashould be updated regularly (e.g., annually) toreflect improvements in milk quality andcheesemaking technology, factors that influencecheesemaking efficiency (see Section 9.5).

9.5 FACTORS THAT AFFECT CHEESEYIELD

The principal factors that influence cheeseyield are discussed below.

9.5.1 Milk Composition

The single most important factor affectingcheese yield is the composition of the milk, par-ticularly the concentrations of fat and casein,

Next Page

Page 191: Cheese Science

in 1936. The van Slyke yield formulae for actualand moisture-adjusted cheese yields are

[Equation 9.6]

[FJC (%FR/100) -C-a] x b

[actual moisture I{ Too J

[Equation 9.7]

_ [FX (%FR 1100) - C - a] x b^MACY = 7 ~ T~~ \

I reference moisture\ 1 0 0 J

where, F and C are the fat content and caseincontent of the cheesemilk (with added starterculture), %FR is the fat recovery, a is the coeffi-cient for casein loss, and b is the coefficient toaccount for cheese solids nonfat, nonprotein(SNFC). The values for %FR/1009 a, and b forCheddar cheese, as predicted by van Slyke, are0.93,0.1, and 1.09, respectively. These formulaecan be rewritten in the format Y = aF + bC,where the values of coefficients a and b are 1.66and 1.78, respectively, for Cheddar cheese con-taining 39Og moisture/kg. The van Slyke for-mula has been modified for other cheese types,based on results from cheesemaking experi-ments. For low-moisture Mozzarella, reportedvalues of %FR/WQ, a, and b are typically 0.86,0.36, and 1.09, respectively. The lower values of%FR/100 and a for low-moisture Mozzarellacompared with Cheddar reflect the higher lossesof fat and casein in the hot water used to heat thecurd during stretching. For commercial FinnishEdam and Emmental cheeses, the mean valuesof %FR/\00 and the coefficient a were 88.7 and88.1, and 0.13 and 0.15, respectively. There isconsiderable intravarietal variation in the re-ported values of %FR/IQQ and the coefficients aand b. For example, the %FR/IQO for Cheddarcheese has been found to range from about 83 to93. Discrepancies between studies undoubtedlyreflect differences in milk composition, milkquality and storage conditions, milk heat treat-

ment, cheesemaking conditions, and cheese-making technology (see Section 9.5).

Variations in the coefficients also occur be-tween cheese plants due to the above factors,which cause interplant differences in efficiency.Moreover, deviations between the actual andpredicted yields may vary between plants.Hence, the application of a generic predictivecheese yield formula for a given cheese varietymay not accurately predict cheese yield in allplants. As an alternative to generic predictiveformulae, plant-specific formulae may be devel-oped for each factory. Plant-specific yield for-mulae may be developed by the statistical analy-sis of historical data (collected weekly ormonthly) on milk composition, milk quality(e.g., somatic cell count), fat and protein recov-ery, and cheese moisture. Plant-specific formu-lae tend to give very accurate predictions ofcheese yield because they reflect, more than ge-neric prediction formulae, the composition andquality of the milk, and the actual cheesemakingconditions used in the plant. The formulae areuseful in that they compare the actual yield withthat expected from a given weight of milk of agiven composition using a given process. If ac-tual yield is less than the predicted yield, reme-dial action is taken to redress process inefficien-cies. However, a close relationship betweenactual and predicted yields does not indicate thata cheese plant is operating at maximum effi-ciency. Hence, a plant-specific yield formulashould be updated regularly (e.g., annually) toreflect improvements in milk quality andcheesemaking technology, factors that influencecheesemaking efficiency (see Section 9.5).

9.5 FACTORS THAT AFFECT CHEESEYIELD

The principal factors that influence cheeseyield are discussed below.

9.5.1 Milk Composition

The single most important factor affectingcheese yield is the composition of the milk, par-ticularly the concentrations of fat and casein,

Previous Page

Page 192: Cheese Science

which together represent around 94% of the drymatter of Cheddar cheese. Yield increases lin-early as the concentrations of fat and casein areincreased (Figures 9-1 and 9-2). In the range 4-33 g fat/kg, the yield of Cheddar cheese typicallyincreases by 1.16 kg cheese/kg milk fat. Simi-larly, cheese yield increases with casein level inthe range 20-30 g/kg, typically by 2.39 kg/kgcasein for Cheddar.

Formulae of the general type Y = aF + bC,which relate cheese yield to the concentrationsof milk fat (F) and casein (Q, have been devel-oped for prediction of cheese yield (see Section9.4). These formulae were derived from experi-mental cheesemaking data and consideration ofmany factors, including the chemistry of theconversion of milk to cheese curd, milk compo-sition, cheese composition, retention of fat andcasein, and partition of components such as milksalts and lactose between the whey and thecheese curd.

The values of the coefficients a and b havebeen found to range from about 1.47-1.6 and1.44-1.9, respectively, for Cheddar cheese. Arecent pilot-scale (500 L) study (M.A. Fenelonand T.P. Guinee, unpublished results) showed

that the actual (Ya) and moisture-adjusted(MACY) yields of Cheddar cheese were accu-rately described by Equation 9.8 (Figure 9-3):

[Equation 9.8]

7cheddar = 1.56F + 1.71C

Undoubtedly, the values of the coefficientsdepend on the composition of the milk, the makeprocedure, and the equipment design, which in-fluence cheese composition and/or the retentionof fat and casein. While formulae of the type Y=aF + bC are empirical, they indicate that in gen-eral casein contributes significantly more toCheddar cheese yield than fat. The separate ef-fects of increasing levels of casein and fat inmilk on cheese yield are illustrated in Figures 9—1 and 9-2. The greater contribution of casein isexpected, as it forms the continuous paracaseinspongelike network that occludes the fat andmoisture (serum) phases. In contrast, fat on itsown has little water-holding capacity. Occludedmoisture contributes directly to cheese yield andindirectly owing to the presence of dissolvedsolids, including whey proteins, K-casein glyco-macropeptide, lactate, and soluble milk salts. In

Concentration of milk fat, g/kg

Figure 9-1 Effect of milk fat concentration on actual (O) and dry matter (•) cheese yield at a constant milkcasein level of 25.5 g/kg.

Che

ese

yiel

d, k

g/10

0 kg

Page 193: Cheese Science

Milk casein plus fat (g/kg)

Figure 9-3 Comparison of actual (•) and moisture-adjusted (to 380g/kg; O) Cheddar cheese yield, with theyield predicted using the formula Y= 1.56F +1.71C, where F and C are the concentrations of casein and fat,respectively, in the pasteurized cheese milk. The regression line for predicted yield and the means from sixreplicate trials are shown.

Che

ese

yiel

d, k

g/10

0 kg

Concentration of casein, g/kg

Figure 9-2 Effect of casein concentration on the actual (O) and dry matter (•) yield of skim milk cheese.

Che

ese

yiel

d, k

g/10

0kg

Page 194: Cheese Science

milk, micellar Ca and PO4, which are associatedwith the casein micelles, are present at concen-trations of around 21 and 16 mM, respectively,and around 90% Ca and 98% PO4 are retainedduring the manufacture of Cheddar curd. Thecombined contribution of dissolved solids andmicellar Ca and PO4 to the yield of Cheddarcheese ranges from around 9% for full-fat Ched-dar (330 g/kg) to around 14% for low-fat Ched-dar (60 g/kg) (Fenelon and Guinee, unpublishedresults).

Fat generally contributes more than its ownweight to Cheddar-type cheese (yield increasesby about 1.16 kg/kg milk fat). This greater thanpro-rata increase is due to the increase in thelevel of moisture in nonfat substance as the fatcontent of the cheese increases. Fat is occludedin the pores of the paracasein network of thecheese and impedes syneresis. The occluded fatglobules physically limit aggregation of the sur-rounding paracasein network and therefore re-duce the degree of matrix contraction and mois-ture expulsion. Hence, as the fat content of thecurd is increased, it becomes more difficult toexpel moisture, and the moisture'.protein ratioincreases. However, if the moisture in nonfatsubstance is maintained constant (e.g., by pro-cess modifications such as reduction of curd par-ticle size and slight elevation of the scald tem-perature), fat contributes less than its ownweight to cheese yield (~ 0.9 kg/ kg), owing tothe fact that about 8-10% of the milk fat is nor-mally lost in the whey.

The contribution of fat and casein to the yieldof a particular cheese type, such as Cheddar or

Cream cheese, is critically dependent on thecasein:fat ratio to which the milk is standard-ized. Other factors that affect the retention of fatand casein, including equipment design and op-eration and make procedure, also have an influ-ence. For any cheese variety, increasing thecasein:fat ratio (e.g., by reducing the level of fat)results in a higher moisture content (and dis-solved solids) and, apart from acid-curd cheesessuch as Quarg, in higher levels of ash. Con-versely, reducing the casein:fat ratio (e.g., bymaintaining the casein level constant and in-creasing the fat content) increases the level of fatand decreases the level of moisture in the cheese.Similarly, the contributions of fat and casein tothe yield of different varieties of cheese dependon the ratio of protein (mainly casein) to fat inthe cheese, which is controlled by standardiza-tion to the desired casein:fat ratio in the milk.Thus, only the casein is important in determin-ing the yield of skim milk cheeses, such asQuarg and Cottage cheese, whereas fat is muchmore important than casein in the yield of Creamcheese.

Many factors contribute to variations in boththe concentration and state (e.g., level of caseinhydrolysis and free fat) of casein and fat in milk.These variations are, in turn, associated withvariations in the rennet coagulation properties,cheese composition, cheese yield, and quality.

Species and Breed

The species of animal has a major influenceon the concentrations of fat and casein and theconcentrations of different caseins (Table 9-2).

Table 9-2 Mean Gross Composition and Estimated Yield of Cheddar Cheese from Standardized Milks(Casein:Fat = 0.75) from Different Species

Species

SheepWater buffaloCowGoat

Fatrm

(9/10Og)

7.27.43.94.5

Fatsm

(9/10Og)

5.204.263.463.46

Casein(9/10Og)

3.93.22.62.6

Lactose(9/10Og)

4.84.84.64.3

Ash(9/10Og)

0.90.80.70.8

Cheese Yield(kg/1 OO kg)

14.7812.119.869.84

Key: Fatrm = fat in raw milk; fatsm = fat in milk standardized to a casein:fat ratio of 0.725.

Page 195: Cheese Science

The estimated yield of cheese from bovine milkis lower than that from sheep or water buffalomilk and similar to that from goat milk.

The breed of cow has a marked influence onmilk composition and its cheese-yielding capac-ity, with breeds having higher protein and fatlevels also having higher cheese yields. Jerseycow milk, which contains 52.2 g/kg fat and 92.7g/kg nonfat solids, yields 13.48 kg Cheddarcheese/100 kg milk compared to 10.1 kg/100 kgfor Friesian milk, which contains 38.0 g/kg fatand 89.0 g/kg nonfat solids. However, the aver-age weight of milk solids (fat and protein) perlactation is higher for Friesian than for Jerseycows, indicating a higher cheese-yielding poten-tial of the latter over the course of a lactation.

Low Concentration FactorUltrafiltration (LCF-UF)

Low concentration factor (1.5-2.0 x) UF iswidely practiced in the manufacture of rennet-curd cheeses as a means of standardizing theprotein level in milk and thereby obtaining moreconsistent cheese yield and quality. Variations ingel strength at cutting, buffering capacity, andthe rennet:casein ratio are minimized. However,when using conventional cheesemaking vats,concentration is limited to a maximum concen-

tration factor of about 1.5, or 40-55 g protein/kg.Increasing the protein level results in faster curd-firming rates and higher curd firmness after agiven renneting time (Figure 9—4). Owing to therapid curd-firming rate, it becomes increasinglydifficult to cut the coagulum cleanly (withouttearing) before the end of cutting, especially ifthe cut program is relatively long (e.g., > 10min). Reflecting the tearing of the coagulum andconsequent shattering of curd particles, fatlosses in the whey are markedly higher thanthose predicted on the basis of volume reduction(due to UF) for milks with a protein concentra-tion greater than 45 g/kg (Figure 9-5), and theyield decreases. The poorer fat-retaining abilityof the coagulum from high-protein milk (> 45 g/kg) may also be due in part to a coarser, moreporous protein network.

Reduction of the set temperature normalizesthe rate of aggregation of rennet-altered mi-celles, gel-firming rate, and set-to-cut time forhigh-protein milks. Hence, at 270C, the gel-firming rate and set-to-cut time of milk contain-ing 45 g/L protein are similar to those for milkwith a protein content of 33 g/L renneted ataround 310C. Under these conditions, increasingthe concentration of protein in the milk in therange 30-45 g/kg protein has no effect on the

Time from rennet addition, min

Figure 9 1 Effect of milk protein level (30 g/kg [A], 35 g/kg [B], 45 g/kg [C], 69 g/kg [D], and 82 g/kg [E]) onthe rennet coagulation properties of milk. Milks B-E were prepared by ultrafiltration of milk A. The rennetcoagulation properties obtained from the curve are gelation time (point at which G' begins to increase), curd-firming rate (slope of G'/time curve in the linear region), and curd firmness (value of G' at a given time afterrennet addition).

Elas

tic s

hear

mod

ulus

, G',

curd

firm

ness

, Pa

Page 196: Cheese Science

percentage recovery of fat and protein. More-over, the moisture-adjusted Cheddar cheeseyield increases with milk protein concentrationat a rate similar to that predicted by the vanSlyke cheese yield equation (Equation 9.7) onthe basis of milk fat and casein (Figure 9-6).However, because of the inverse relationship be-tween the concentration of protein in the milkand the moisture content of the cheese, the actualcheese yield increases less than predicted on thebasis of casein and fat (Equation 9.6). The mois-ture content of cheese made from high-proteinmilk (> 30 g protein/kg) may be increased to thatof the control milk (3Og protein/kg) by manipu-lating the cheesemaking process, such as by cut-ting at a higher gel firmness, increasing the cutsize, lowering the scald temperature, and/or in-creasing the rate of cooking. With such alter-ations, the increases in actual and predictedcheese yields with milk protein level in the range30-45 g/kg are similar.

9.5.2 Somatic Cell Count and Mastitis

The influence of somatic cells and mastitis onthe composition of milk and its suitability forcheese manufacture has been studied exten-sively. There are three main types of somatic

cells: lymphocytes (L), phagocytes, and mam-mary gland epithelial cells (E) (Burvenich,Guidry, & Paape, 1995). Lymphocytes functionin humoral and cell-mediated immunity whilephagocytes, of which there are two types, poly-morphonuclear leucocytes (PMN) and macro-phages (M4,), ingest and kill pathogenic microor-ganisms that invade the mammary gland. Lownumbers of somatic cells (e.g., < 100,000/ml)are present in normal milk from healthy animalsduring mid lactation, with M<j>, L, PMN, and Ecells typically in a ratio of roughly 2.1:1.0:0.4:0.2. Somatic cells are released from theblood to combat udder infection and thereby pre-vent or reduce inflammation (mastitis). Factorsthat contribute to increases in the somatic cellcount (SCC) of bulk manufacturing milk includesubclinical mastitis, advanced stage of lactation,lactation number, stress, and poor nutrition. Dur-ing clinical mastitis, there is a rapid increase inSCC, due primarily to PMN. Depending on thetype and extent of bacterial infection, milk frominfected (mastitic) quarters of the udder mayhave a SCC of 200-5,000 x lOVml. However,the milk from animals suffering from clinicalmastitis is excluded from the commercial milksupply. Such milk frequently forms clots, whichare a mixture of somatic cells and precipitated

Milk protein level (g/kg)

Figure 9-5 Effect of milk protein concentration, varied by ultrafiltration, on actual (A) and predicted (B) fatlevels in bulk Cheddar cheese whey. Predicted values were calculated by adjusting the value for the control (3Ogprotein/kg) for the volume reduction due to ultrafiltration.

Fat i

n bu

lk c

hees

e w

hey

(g/l)

A

B

Page 197: Cheese Science

milk proteins, within the udder. In severe masti-tis, these clots block the drainage ductules andducts in the mammary gland, thereby preventingmilk drainage. The initial stage of mastitic infec-tion is subclinical, with inflammation so slightthat it is not detectable by visual examination.Hence, the milk from cows suffering from sub-clinical mastitis becomes part of bulk herd milkand bulk manufacturing milk unless individualcows are tested routinely at farm level for sub-clinical mastitis (e.g., SCC), which is seldom thecase. While bulking dilutes such milk, subclini-cal mastitis may contribute to an increased SCCin the milk and thereby reduce its suitability forcheese manufacture.

An increased SCC in milk is associated withmarked changes in both the concentration andthe state (e.g., degree of hydrolysis) of milk con-stituents that lead to a deterioration in rennet co-agulation properties, curd syneresis, and re-duced cheese yield. Increasing SCC in the range105-106/ml results in progressive increases in

the concentration of whey proteins, especiallyIg, and decreases in the levels of casein and fat(Figure 9-7). Elevated SCC results in a markeddecrease in p-casein as a proportion of totalcasein and concomitant increases in y-caseins,proteose peptones, and the ratio of soluble tomicellar casein. These changes ensue from hy-drolysis of P- and as2 caseins by the elevated ac-tivity of plasmin (and probably other protein-ases) in the milk that parallels increased SCC;asi- and K-caseins are hydrolyzed more slowlyby plasmin than p- and as2-caseins. These pro-teinases may come directly from the blood (e.g.,plasmin and its zymogen plasminogen) or fromthe somatic cells. The types of casein-derivedpeptides produced by the two types of protein-ases differ somewhat.

Increasing SCC in the range 105 to 6 x lOYmlresults in an increase in rennet coagulation timeand a decrease in curd-firming rate (reciprocal ofk20; see Section 6.6.2) and curd firmness(Politis & Ng-Kwai-Hang, 1988b). Typical

Milk protein concentration, g/kg

Figure 9-6 Effect of milk protein level on the yield of full-fat Cheddar cheese: moisture-adjusted (to 380 g/kg)yield (•, A), moisture-adjusted (to 380 g/kg) yield predicted using the van Slyke cheese yield formula (•, B;Equation 9.7), and actual yield (O, C).

Che

ddar

che

ese

yiel

d, k

g/10

0 kg

AB

C

Page 198: Cheese Science

trends in rennet coagulation properties with SCCare shown in Figure 9-8. Losses of fat and pro-tein during Cheddar cheese manufacture in-crease, more or less linearly, by about 0.7% and2.5%, respectively, with SSC in the range 105-106/ml (Politis & Ng-Kwai-Hang, 1988a). Anincrease in SCC from 105 to 6 x lOVml resultedin an 11% decrease in moisture-adjusted (370g/kg) Cheddar cheese yield (Figure 9-9). It isnoteworthy that there was a relatively large de-crease in yield (~ about 0.4 kg/100 kg milk) onincreasing the SCC from 105 to 2 x 105, a rangethat would be considered relatively low for goodquality bulk milk. Other studies have reportedsimilar trends. For example, Auldist et al. (1996)found that an increase in SCC from above 3 x105 to above 5 x 105 in late lactation (220 d) re-sulted in a 9.3% decrease in moisture-adjusted(to 355 g/kg) yield of Cheddar and decreases inthe recovery of fat (from 90.1% to 86.6%) andprotein (from 78.3% to 74.4%). A decrease of4.3% has been reported in the yield of un-creamed Cottage cheese upon increasing theSCC from 8.3 x 104 to 8.7 x 105 cells/ml.

The negative impact of SCC on cheese yieldand recovery is due largely to the increased pro-teolysis of ocs-caseins (asr + as2-) and (3-caseinsto products that are soluble in the serum and are

not recoverable in the cheese. Moreover, thelower effective concentration of gel-formingprotein results in a slower curd-firming rate andhence a lower degree of casein-casein interac-tion in the gel following cutting (at a given firm-ness) and during early stirring. A gel with thelatter characteristics exhibits

• greater susceptibility to shattering duringcutting and the early stages of stirring, re-sulting in higher losses of curd fines andmilk fat

• an impaired syncretic capacity, with a con-sequent increase in moisture level

A high SCC may also inhibit the activity ofsome strains of lactococci during cheese manu-facture, an effect expected to further impair thecurd-firming rate and reduce firmness at cut-ting. In commercial practice, the gel is generallycut, not on the basis of firmness, but rather onthe basis of a preset renneting time, based oncurd firmness within the acceptable range fornormal milk. In large modern factories, condi-tions are not conducive to assessing the firm-ness of gels in individual vats because of thelarge scale of operations (more than a million li-ters per day) and the use of preprogrammed vatswith limited operator access. In such factories,

Var

iabi

lity

in m

ilk c

ompo

sitio

n, %

Somatic cell count, 1O3CeIIs/ml

Figure 9-7 Variability in the composition of milks from individual cows with somatic cell count. Casein (O), fat(•), total protein (•), and whey protein (Q).

Page 199: Cheese Science

Ren

net c

oagu

latio

n tim

e, m

inC

urd

firm

ing

rate

, 1/

k20,

min

-C

urd

firm

ness

at 3

0 m

in, A

30, m

in

Somatic cell count, cells/ml x 103

Figure 9-8 Effect of somatic cell count on the rennet coagulation properties of milks from individual cows.

Page 200: Cheese Science

the effects of increased SCC may be accentu-ated, as the slower than normal curd-firmingrate is conducive to lower than optimum firm-ness at cutting. Another factor that contributesto poor yield from high SCC milks is the in-creased susceptibility of the milk fat globulemembrane to damage, which increases the sus-ceptibility of the fat to lipolysis by indigenousmilk lipoprotein lipase and lipases from psych-rotrophic bacteria.

In conclusion, high SCC is detrimental tocheese yield and the profitability of cheese-making. It is estimated that the monetary loss re-sulting from a 2% decrease in cheese yield dueto increasing the SCC from 105 to 5 x 105 cells/ml is about $6,000 per day in a Cheddar cheeseplant processing 1 million liters per day, basedon a value of $3 per kilogram of curd. The SCCin milk is being lowered through the use of goodon-farm practices, such as reducing the percent-age of animals with subclinical mastitis, and isbeing driven by new regulations for lower SCC.For example, in the European Union, the permit-ted maximum of SCC in milk for the manufac-

ture of dairy products bearing the health markwas reduced in 1998 from 5 x 105 cells/ml to 4 x105 cells/ml (EU directive 92/46/EEC).

9.5.3 Stage of Lactation and Season

Marked changes occur in the composition ofmilk throughout the year, especially when milk isproduced mainly from spring-calving herds fedpredominantly on pasture, as in Ireland, NewZealand, and parts of Australia. These changes inmilk composition are due principally to thephysiologically induced changes in the biosyn-thetic performance of the mammary gland asinfluenced by stage of lactation and diet (seeChapter 3). Lactation-related changes in milkcomposition may be defined as those that occurduring the period of milk production, betweenparturition and drying off, mainly as a result ofphysiological changes in the mammary gland ofhealthy cows fed on a standard good quality diet.On the other hand, seasonal changes in the milkfrom individual cows may be defined as thosearising due to lactation and the superimposed ef-

Moi

stur

e-ad

just

ed c

hees

e yi

eld,

kg/

100

kg

Somatic cell count, x 103cells/ml

Figure 9-9 Influence of somatic cell count on the moisture-adjusted (to 370 g/kg) Cheddar cheese yield frommilks from individual cows.

Page 201: Cheese Science

fects of other factors, such as variation in diet,illness (e.g., mastitis), environmental factors, andclimate. However, in practice, individual-cowmilks are bulked at the farm level, and the ensuingherd milks are further bulked during the transportto and assimilation at the manufacturing plant.

Seasonal changes in milk composition, whichare most pronounced at the extremes of lactation(Figure 9-10), result in variations in rennet co-agulation properties, cheese composition, recov-ery of fat and casein, cheese yield, and quality.Seasonal studies in Ireland during the early1980s showed that between March and Novem-ber protein increased from around 30 to 42 g/kg,casein from 22 to around 30 g/kg, and proteosepeptone from around 16 to 53 mg/100 g. The es-timated yield of Cheddar cheese varied fromabout 8.5 to 11.5 kg/100 kg milk. Similar studiesin Scotland reported smaller seasonal variationsin milk composition and cheese yield: proteinranged from around 30.9 to 35.6 g/kg, caseinfrom 24.1 to 27.4 g/kg, fat from 35.6 to 41.3g/kg, and Cheddar cheese yield from 9.3 to 10.5kg/100 kg. Yield studies in the United Stateshave shown even smaller variations in the com-position of commercial milk. For example, in

New York, the range for casein was 23-25 g/kg;for milk fat, 34-38 g/kg; for casein number,around 76.5 to 81.7; and for actual Cheddarcheese yield, around 9.1 to 9.76 kg/100 kg.Marked seasonal variations in the actual yield ofEmmental cheese have been observed in Ger-many (« about 8.94-9.45 kg/100 kg) and Fin-land (« about 10.8-12.0 kg/100 kg). Seasonalvariations in milk composition are paralleled byvariations in the recovery of fat and/or casein,reflecting variations in curd-firming rates andthe susceptibility of the resulting curds to shat-tering during cutting and stirring (see Section9.5.12).

In addition to changes in gross composition,the relative concentrations of the individualcaseins as a percentage of total casein show sea-sonal changes, especially in factory milk ob-tained largely from spring-calving herds (Figure9-10). These changes probably occur as a resultof an increase in plasmin activity in milk overthe course of lactation (Figure 9-11). (3-Caseinand as2-casein are readily cleaved by plasmin,whereas K-casein and ocsrcasein are more resis-tant. The authors are not aware of any studiesreporting the effects of variations in the propor-

A B

DC

J F M A M J J A S O N D J F M A M J J A S O N D

% o

f who

le c

asei

n

Tota

l pro

tein

, g/k

g%

of w

hole

cas

ein

% o

f who

le c

asei

n

J F M A M J J A S O N D J F M A M J J A S O N D

Figure 9-10 Seasonal changes in the level of total protein in winter-spring-calving herd milk (A) and in theproportions of ocs-casein (B), (3-casein (C), and y-casein (D) in Irish bulk creamery milk, mainly from winter-spring-calving herds (B-D).

Month Month

Page 202: Cheese Science

tions of different caseins on cheese yield. How-ever, considering the different properties ofcaseins (see Chapter 3), it is likely that variationsin the ratio of the caseins would influence rennetcoagulation properties, recovery of components,and cheese composition and yield.

Numerous reports have shown that the rennetcoagulation and curd-syncretic properties of latelactation milk are inferior to those of mid-lactation milk. Consequently, cheese made fromlate lactation milk tends to have a high moisturecontent and to be of inferior quality. All otherfactors being equal, the gel-firming rate is posi-tively correlated with the milk casein level (seeChapter 6). Considering its high casein content,late lactation milk would be expected to havegood cheesemaking properties and to give highyields. However, the reverse is observed, possi-bly owing to a high pH, changes in the propor-tions of the caseins (Figure 9-10), proteolysis byplasmin, and perhaps unidentified factors.Cheese made from late lactation milk tends tohave a high moisture content, reflecting poorsyncretic properties, which have not been fullyexplained. The high moisture content of cheesefrom late lactation milk contributes to a higheractual cheese yield.

Seasonal changes in milk composition can beminimized and its suitability for cheese manu-facture increased through good husbandry and

milk-handling practices, such as

maintenance of a high-energy diet (e.g., drymatter intake of > 17 kg/cow/day, with adry matter digestibility of 820 g/kg organicmatter) by good grassland managementpractices and concentrate supplementationof herbage-based diets, especially whengrass is in short supplyelimination of extreme late lactation milkby drying off cows at a suitable stage (e.g.,at a milk yield > 8 kg/cow/day)reducing stress and infection

9.5.4 Genetic Polymorphism of MilkProteins

All the major proteins in milk (i.e., as]-, as2-,P-, and K-caseins, p-lactoglobulin, and a-lactal-bumin, exhibit genetic polymorphism (seeChapter 3).

The genetic variants that have been investi-gated most thoroughly for their effects on therennet coagulation and cheesemaking character-istics of milk are those of K-casein and p-lacto-globulin. Compared to the AA variants, the BBgenotypes of p-lactoglobulin and K-casein aregenerally associated with higher concentrationsof casein and superior rennet coagulation prop-erties, as reflected by higher gel-firming rates

Plas

min

con

cent

ratio

n, m

g/l

Jan-Feb Mar-Apr May-Jun Jul-Aug Sept Oct-Dec

Month

Figure 9-11 Effect of stage of lactation on the concentration of plasmin in milk from individual cows.

Page 203: Cheese Science

and gel firmness after a given renneting time(Figure 9-12). The BB variants of K-casein andp-lactoglobulin are also associated with superiorcheesemaking properties, as reflected by ahigher recovery of fat, lower levels of curd finesin cheese whey, and higher actual and moisture-adjusted cheese yields for a range of varieties,including Cheddar, Sveciaost, Parmigiano-Reggiano, Edam, Gouda, and low-moistureMozzarella. Reported increases in moisture-ad-justed yield with the K-casein BB variant rangefrom about 3% to 8%, depending on milk com-position and cheese type. The superior rennetcoagulation and cheese-yielding characteristicsof K-casein BB, compared with the AA variant,probably result from the higher level of K-caseinas a percentage of total casein, its smaller mi-celles, and its lower negative charge. Theseproperties are conducive to a higher degree ofcasein aggregation and a more compact arrange-ment of the sensitized paracasein micelles,which in turn favors more intermicellar bondsduring gel formation. Model rennet coagulationstudies have shown that, for a given casein con-centration, the curd-firming rate of renneted mi-celles is inversely proportional to the cube of themicelle diameter. The generally higher level of

casein accompanying K-casein BB, comparedwith the AA variant, also contributes to its supe-rior rennet coagulation and cheese-yieldingproperties. Milk containing K-casein AB gener-ally exhibits rennet coagulation and cheese-yielding characteristics intermediate betweenthose of K-casein AA and BB.

Compared with OC8I-CN BB, ocsi-CN BC caseinis associated with higher levels of ocsl-casein, totalcasein, and total protein and higher estimatedyields of Parmesan cheese. However, owing tothe higher yields of milk, milk fat, and milk pro-tein in milks containing asrCN BB casein over acomplete lactation, cows producing this variantgive higher cheese yields during a lactation thanthose producing (X8I-CN BC casein.

In conclusion, the BB variants of K-casein andp-lactoglobulin enhance cheese yield and haveno adverse effects on cheese quality or on func-tionality in the case of low-moisture Mozzarellacheese (Walsh et al., 1998).

9.5.5 Cold Storage of Milk

In modern farm and milk collection, milk iscooled rapidly to below 80C following milking,and milk collection from the farm often occurs

Ela

stic

she

ar m

odul

us,

G',

Pa

Time from rennet addition, min

Figure 9-12 The development of elastic shear modulus (curd firmness) of rennet-treated milk for different K-casein variants. K-casein AA, 34.4 g protein/kg (O); K-casein BB, 35.3 g protein/kg (•).

Page 204: Cheese Science

every second or third day. Moreover, cold milkis hauled over long distances and is often cold-stored at the cheese plant for 1-3 days, depend-ing on time of year and the manufacturingschedules. Consequently, milk can be cold-stored for 2-5 days prior to processing. Duringthis period, the cold milk is subjected to varyingdegrees of shear from pumping, flow in pipe-lines, and agitation.

Cold storage and shearing result in a numberof physicochemical changes that may alter thecheesemaking properties of milk, including

• solubilization of micellar caseins, espe-cially p-casein, and colloidal calcium phos-phate, leading to an increase in serumcasein

• increased susceptibility of serum casein tohydrolysis by proteinases from psychro-trophic bacteria or somatic cells and/orplasmin and the concomitant increase innonprotein N (Figure 9-13)

• damage to the milk fat globule membraneand hydrolysis of free fat by Upases frompsychrotrophic bacteria and/or milk, result-ing in a decrease in the level of fat (Figure9-13)

Cold storage impairs rennet coagulation prop-erties, reduces the recovery of protein and fat,

and reduces cheese yield (Figure 9-13 and Table9-3). However, there is disagreement betweenreported studies as to the magnitude of the effectof cold storage (cold aging). Discrepancies be-tween reports may be attributed to variations inexperimental conditions, such as prehandlingand temperature history of milk prior to experi-mentation, milk pH, somatic cell count, bacterialcount, and species/strains of psychrotrophic bac-teria in milk, storage temperature and time, andcheesemaking conditions. It is generally agreedthat, at a level of less than 106 cfu/ml, psychro-trophs have little effect on the cheesemakingproperties of milk.

On storage at 40C for 48 hr, the increase in therennet coagulation time (RCT) of milk from indi-vidual cows ranged from 10% to 200% and thatof bulk milk from 9% to 60% (Fox, 1969). A di-rect relationship was found between the increasein the RCT of cold-aged milk and the initial RCTbefore aging. The large variation in the RCT ofcold-aged milk from individual cows is probablya consequence of differences in composition,microbiological status, and somatic cell count.The chemical changes (i.e., the increases in se-rum casein, in the ratio of soluble Ca to micellarCa, and in RCT associated with cold storage) arealmost complete after 24 hr and are largely re-versed by pasteurization (720C x 15 s) or milder

Figure 9-13 Effect of type and population of psychrotrophic bacteria on milk composition (a, b) and the drymatter yield of experimental cheese curd (c). Pasteurized milk was inoculated at various levels (10°-106 cfu/ml)with Bacillus (O) or Pseudomonas (•) species isolated from raw milk (stored at 70C for 3 days) and incubatedfor 6 days (Pseudomonas) or 10 days (Bacillus) before cheese manufacturing. Data are the means of eight trialswith each microorganism.

Log psychrotrophic count, cfu/ml Log psychrotrophic count, cfu/m! Log psychrotrophic count, cfu/ml

Non

-pro

tein

N in

milk

, m

g/m

l

Red

uctio

n in

leve

l of m

ilk fa

t,g/1

00g

Dry

mat

ter

chee

se y

ield

,kg

/1 O

O kg

(a) (b) (c)

Page 205: Cheese Science

heat treatment (e.g., 5O0C x 300 s).The increase in RCT—and consequently the

longer time required for curd formation duringcheese manufacture—that generally accompa-nies cold storage beyond 24 hr appears to be due,at least in part, to enzymatic degradation ofcasein. Proteolysis reduces the concentration ofgel-forming casein to an extent dependent on theproteolytic activity in the milk. Peptides, whichare soluble in the serum phase (as nonproteinN),do not coagulate upon renneting and are largelylost in the whey. Moreover, the reduced caseinlevel results in slow curd formation and a softcoagulum at cutting (see Chapter 6), a situationconducive to curd shattering, high losses of fatin the whey, and reduced cheese yields. In com-mercial practice, the coagulum is usually cut at afixed time after rennet addition rather than at agiven firmness.

In the European Union, the permitted maxi-mum total bacteria count (TBC) in milk for themanufacture of dairy products was reduced from4 x 105 cfu/ml to 1 x 105 cfu/ml in 1998 (EUdirective 92/46/EEC). Improved dairy hus-bandry practices, combined with the more strin-gent standards for TBC and SCC, should reducethe level of storage-related proteolytis and Ii-polysis in milk. The fact that the chemicalchanges that occur during cold storage are re-versed by pasteurization suggests that, withmodern milk production practices, cold storageof milk for several days probably has little influ-ence on its cheesemaking properties.

9.5.6 Thermization of Milk

Prolonged holding of milk prior to processingcan occur in the factory when milk is in oversup-

ply, such as during the spring flush, or on thefarm when daily production is low, such as dur-ing winter. Extended cold storage may lead tothe development of high psychrotrophic popula-tions that produce proteolytic and lipolytic en-zymes resistant to pasteurization. These en-zymes can reduce cheese yield (as discussed inSection 9.5.5) and adversely affect product qual-ity. Thermization of milk at a subpasteurizationtemperature (e.g., 57-680C for 10-15 s) on re-ception at cheese factories is widely practiced.Thermization reduces the number of psychro-trophs in milk to an extent dependent on the heattreatment. Studies on Cottage cheese showedsignificantly higher actual and moisture-ad-justed (to 820 g/kg) yields from milk heated at740C x 10 s prior to storage than from theunpreheated control (16.85 vs. 16.0 kg/100 kg),when milk was cold stored at 30C for 7 days.Hence, it has been suggested that when milk isstored for a long period on farms, on-farmthermization may prove advantageous forcheese yield.

9.5.7 Pasteurization of Milk andIncorporation of Whey Proteins

Pasteurization of milk (720C x 15 s) resultsin a low level of denaturation of whey proteins(< 5% of total), which complex with K-caseinand are retained in the cheese curd, where theycontribute to a yield increase of about 0.1-0.4%. However, most of the native whey pro-teins (~ 94-97%, depending on the cheesemoisture level), which account for 20% of thetotal milk protein, are lost in the whey. Unlikecasein, native whey proteins are stable to rennet

Table 9-3 Effect of Cold-Storing Milk at 30C on the Recovery of Milk Fat and Solids and on ActualCheese Yield

Recovery (%)

Storage Time (Days)

O35

Fat (%)

89.287.786.0

Solids (%)

57.049.248.3

Yield (kg/100 kg)

9.79.19.0

Page 206: Cheese Science

treatment and acidification to pH 4.6 and thusremain soluble in whey during the manufactureof rennet- and acid-curd cheeses. Theoretically,if all whey proteins were retained without ad-versely affecting cheese moisture or quality, ayield increase of approximately 12% (10.7 vs.9.54 kg/100 kg) would be achievable for Ched-dar cheese with around 380 g/kg moisture. In-deed, since an increase in the severity of milkheat treatment is paralleled by an increase incheese moisture (Figure 9-14), the increase inactual yield would be even higher (e.g., -15%following pasteurizing at 880C x 15 s). How-ever, inclusion of high levels of whey protein(> 35% of total whey protein), either in dena-tured or native form, can adversely affect rennetcoagulation properties (see Chapter 6), cheeserheology (e.g., reduced elasticity and firmness),functionality (e.g., reduced flowability), and theoverall quality of most rennet-curd cheeses (seeIDF, 1991). Owing to an effect on cheese rheol-ogy, high levels of denatured whey proteins incheese milk may be exploited as a means of im-proving the texture (reducing the firmness andelasticity) of low-fat cheeses, which tend to betoo firm and rubbery (chewy), compared withtheir full-fat equivalents. Inclusion of high lev-els of whey proteins in some fresh acid-curdcheeses (e.g., Quarg, Cream cheese), while al-tering the textural properties, generally does notimpair eating quality. However, even here func-tionality may be altered. In acid-heat coagu-lated cheese types (e.g., Ricotta, Paneer, sometypes of Queso bianco), the incorporation of ahigh level of denaturated whey protein is a fea-ture of the manufacturing process and may im-part certain desirable attributes, such as flow re-sistance and lack of elasticity. Whey proteinsmay be incorporated into cheese in a number ofways:

• Whey proteins may be denatured in situ byhigh heat treatment (HHT) of the cheesemilk (e.g., ~ 65% of total whey proteins at10O0Cx 12Os).

• High concentration factor ultrafiltration(HCF-UF) can be used, with or withoutHHT of the milk before UF or HHT of the

rententate after UF, as in the production ofpre-cheese.

• Partially denatured whey protein concen-trates, prepared by high heat treatment and/or acidification of cheese whey, can beadded to the cheese curd. Commercialwhey protein preparations of this type (e.g.,Dairy Lo and Simplesse 100) have beenused to improve the textural characteristicsof reduced-fat cheese.

High Heat Treatment of Cheese Milk

In situ denaturation of whey proteins by HHTis widely used in the manufacture of fresh acid-curd cheeses, such as Quarg, Fromage frais, andCream cheese, with typical heat treatments rang-ing from 720C x 15 s to 950C x 120-300 s. Theextent of whey protein denaturation and theyield of Quarg (18% moisture) from milk sub-jected to these treatments are about 3% and 18.6kg/100 kg and about 70% and 21.3 kg/100 kg,respectively. Increasing the level of whey pro-tein denaturation from 0% to 50% by HHT ofmilk increases protein recovery and actual andmoisture-adjusted cheese yield for a range ofhard and semi-hard rennet curd cheeses (Table9-4). The recovery of fat also increases as thelevel of whey protein denaturation increases (atleast to « 21% of total) provided that thecheesemaking process is altered to restore thegel-forming properties of the HHT milk to thoseof the unheated milk, for example, by extendingthe set-to-cut time when the pH of milk is nor-mal, by adding CaCl2, and/or by lowering the setpH by the addition of food grade acid oracidogen.

The effects of HHT on the rennet coagulationproperties of milk are discussed in Chapter 6.These effects are attributable mainly to the in-crease in the concentration of the gel-formingprotein and the reduction in the porosity of themore finely structured gel produced from theHHT milk (see Chapter 16), which increases thewater- and fat-holding capacities of the gel. Thecomplex between K-casein and denaturedp-lactoglobulin results in the formation of fila-mentous appendages that protrude from the sur-

Page 207: Cheese Science

face of the casein micelles. These appendagesrender the K-casein less susceptible to hydrolyisby rennet and probably limit the degree of aggre-gation of the paracasein micelles. The rate of ag-gregation is also impaired as a consequence ofthe reduction in the level of micellar calciumphosphate, which is the intramicellar "cement-ing agent." The more finely structured gel (com-pared to that from pasteurized milk) has im-paired syncretic properties, and consequentlythe moisture content (Figure 9-14) and actualcheese yield are increased. However, rennet-in-duced gels from HHT milks with a level of wheyprotein denaturation (WPD) above 30% tend tobe very fragile and prone to shattering upon cut-ting and stirring, even when the milk has beenacidified to pH 6.0 prior to setting. Hence, highlosses of fat (up to 15%) and curd fines (up to800 mg/kg) in the whey are possible unless thegel is treated very gently. The susceptibility ofcurd particles from HHT milk to fracture persistsfor a longer time than normal into the stirringperiod, because of the lower tendency to syn-

erese and the slower rate of firming of the curdparticles.

Addition of Denatured Whey Protein toCheese Milk

An alternative to in situ denaturation of wheyprotein in milk is the addition of denatured wheyprotein (recovered from whey) to the cheesemilk. Potential advantages of this method are asfollows:

• The curd-forming properties of the caseinare not (or only slightly) impaired.

• The cheese yield is increased but the textureis not altered to the same extent as in curdfrom HHT milk with an equivalent level ofdenatured whey protein.

Potential disadvantages of adding denaturedwhey proteins include these:

• The addition of a milk protein fraction (i.e.,whey protein) may not comply with the Ie-

Pasteurization temperature, 0C

Figure 9-14 Effect of milk pasteurization temperature (for 15 s) on whey protein denaturation (•) and themoisture content of half-fat Cheddar cheese (O).

% W

hey

prot

ein

dena

tura

tion

Che

ese

moi

stur

e, g

/kg

Page 208: Cheese Science

Table 9-4 Effect of Whey Protein Incorporation by in situ Denaturation at High Temperature on the Recovery of Components and Yield of Different Cheese

Percentage Increase inYield Relative to Control3Recovery

Types

Heat Treatment WPD Cheese Cheese Moisture Fat Protein Dry Matter SNF Moisture DryReferencedMatterAdjustedActual(%)(%)(%)(%)(g/kg)Type(%)(°Cxs)

1

2

3

4d

2.1

NPNP

NP

NP

2.0

NPNP

4.5

4.7

0.78

1.63.4

NP

11.9

NPNP

32.533.234.3NP

NPNP

NPNPNPNPNPNP

NP (4.5)NPNP

74.575.176.077.479.5NP

NP (6.7)77.886.2

92.493.597.498.196.3NP

NP (0.7)c

94.395.8

363371515516496NPNP371387

Cheddar

Mozzarella

Cheshire

Cheddar

O53

21323?323?50

C: O x O (raw)T: 63 x 1800C: 80 x 2T1: 110x2T2:130x2C: 72 x 17T: 97 x 15C: 72 x 16T: 110x60

Key: SNF = solids-not-fat; WPD = whey protein denaturation; C = control; T = treatment; NP = data not presented; ? = data not presented butvalue estimated.

b References: 1. K.Y. Lau, D. M. Barbano, & R.D. Rasmussen, Influence of pasteurization on fat and nitrogen recoveries and Cheddar cheese yield, Journal of Dairy Science, 73(1990), 561-570. 2. H.W. Shafer & N. F. Olson, Characteristics of Mozzarella cheese made by direct acidification from ultra-high-temperature processed milk, Journal of Dairy Science,58 (1974), 494-501. 3. RJ. Marshall, Increasing cheese yield by high heat treatment of milk, Journal of Dairy Research, 53 (1986), 31 3-322. 4. J.M. Banks, G. Stewart, D. D. Muir, &I.G.West, Increasing the yield of Cheddar cheese by the acidification of milk containing heat denatured whey protein, Milchwissenschaft, 42 (1987), 212-215.

c Values in parentheses for recoveries denote presented data for the percentage increases in fat, protein, and dry matter.d In this experiment, the milk was heated at 1 1 00C x 60 s, cooled to 3O0C, acidified to pH 5.8, set and cut after a 1 5 min set-to-cut time. Cutting the gel from the treated milk after a time

similar to that for the control (45 min) resulted in a reduced fat recovery (8.4%).

Page 209: Cheese Science

gal standard of identity of the cheese in allcountries.

• It may be too expensive.

Several processes have been developed forthe recovery of protein from sweet whey, mostof which involve flocculation, by heat denatur-ation and/or acification, and recovery by cen-trifugation. The Centriwhey process yields con-centrates containing around 12Og solids/L andthe Lactal process yields concentrates contain-ing around 180 solids g/L. These concentratesare added at the desired level to the cheese milk,which is then treated as in normal cheese manu-facture. Commercial partially denatured wheyprotein may also be prepared by concentratingsweet whey by ultrafiltration and heat-treatingand shearing the retentate to give a controlledlevel of denaturation. The heat-treated retentatemay be supplied as a viscous, wet slurry or spraydried.

Variable results have been obtained with theaddition of partially denatured whey proteinconcentrate (PDWPC, prepared by the Centri-whey, Lactal, or ultrafiltration processes) tomilk for the manufacture of hard and semi-hardcheeses, such as Cheddar and Gouda. There isgeneral agreement that the addition of PDWPCincreases the moisture content, actual yield, andmoisture-adjusted yield, with the extent of theincreases depending on the amount of PDWPCadded and the degree of whey protein denatur-ation of the concentrate. However, the additionof PDWPC generally leads to defective bodycharacteristics (i.e., greasy, soft) and flavorcharacteristics (i.e., unclean, astringent) inGouda and Cheddar cheese.

Commercial PDWPCs (e.g., Simplesse 100 orDairy Lo) have been used as fat mimetics to im-prove the texture of reduced-fat cheeses, whichtend to be more firm and elastic than their full-fat equivalents. These materials are added at lev-els of around 10-20 g/kg to increase the level ofmilk protein by about 3.5 g/kg. Their use in half-fat Cheddar cheese gives a higher level of cheesemoisture (e.g., 450 vs. 427 g/kg) and higher ac-tual cheese yields (e.g., 7.7 vs. 7.2 kg/100 kg)and moisture-adjusted cheese yields (Fenelon &

Guinee, 1997; IDF, 1994). These materials alsoimprove the texture of reduced-fat Cheddar (byreducing fracture stress, fracture strain, andfirmness) and have little influence on flavor oraroma.

High Concentration FactorUltrafiltration

High concentration factor ultrafiltration(HCF-UF) has been widely used as an alterna-tive to centrifugation, especially in Europe sincethe 1980s, for concentrating the gelled milk inthe commercial production of fresh acid-curdcheeses, such as Quarg and Cream cheese. Thismethod, which concentrates the milk sixfold ormore, allows complete recovery of whey pro-teins and gives higher yields of cheese of veryacceptable quality. The use of HCF-UF insteadof centrifugation for the manufacture of freshacid-curd cheeses has necessitated slightchanges in the manufacturing procedure to givecomparable sensory quality (see Chapter 16).Functionality is sometimes a quality attribute inCream cheese (e.g., desired degree of flow-ability in cordon bleu poultry). While no pub-lished information is available on this aspect ofCream cheese, it is envisaged that Cream cheeseproduced from HCF-UF retentate would be re-sistant to flow upon cooking owing to the ther-mal gelation of the whey proteins at the highcook temperature (e.g., 90-10O0C). In contrast,the production of Cream cheese with customizedflowability is easily achieved by manipulatingthe manufacturing procedure when using cen-trifugation to concentrate the curd.

HCF-UF has also been used for the com-mercial manufacture of rennet-curd cheeses, in-cluding cast Feta in Denmark and Cheddar inAustralia made using the Siro-Curd process.Manufacture involves HCF-UF of the milk toproduce a retentate or liquid pre-cheese with ahigh dry matter content (e.g., 400-500 g/kg),which has a composition close to that of the fin-ished cheese. Rennet and starter cultures areadded to the pre-cheese, which is then treated inthe normal manner, except that little (e.g., Ched-dar) or no (cast Feta) expulsion of whey occurs.Since the upper limit for concentration by UF is

Page 210: Cheese Science

about 7:1 for whole milk, it is not possible toachieve the dry matter level required for hardcheeses, such as Cheddar or Gouda. Hence, fur-ther whey expulsion from the pre-cheese occursafter coagulation and cutting on conventionalequipment (e.g., drainage belts, finishing vats)or on specialized equipment (e.g., Siro-Curd).

The attraction of HCF-UF for the manufac-ture of rennet-curd cheese is the increased yieldowing to the high level of retention of whey pro-teins and glycomacropeptides. The exact degreeof retention depends on

the heat treatment of the milk prior to ren-neting, which determines the extent ofwhey protein denaturation and hence wheyprotein solubilitythe level of whey expulsion from the pre-cheese following coagulation and cutting(native whey proteins and glycomacropep-tides are soluble and are lost in the whey)

When complete recovery of the native wheyproteins and glycomacropeptides is achievable(as in cast Feta), the estimated saving in skimmilk is about 9%, which makes the process eco-nomically viable (see IDF, 1994).

Extensive research has been undertaken on thecharacteristics of different hard cheeses pro-duced at pilot or laboratory scale from HCF-UFretentate. The sensory characteristics of HCF-UFcheeses tend to differ from those of traditionalcheese to a degree dependent on the level of wheyprotein included, the extent of whey protein de-naturation, and other manufacturing conditionsthat influence the composition and rate of pro-teolysis and/or lipolysis during maturation.

9.5.8 Homogenization and Microfluidization

Homogenization of milk reduces fat globulesize and increases the surface area of the fat by afactor of 5 to 6. The fat globules become coatedwith a protein layer consisting of casein micelles,submicelles, and to a lesser extent whey proteins.Hence, the newly formed fat globules behave aspseudocasein particles and exhibit the ability tobecome part of the gel network formed upon

renneting or acidification. Homogenization ofmilk or cream is not widely practiced in the manu-facture of rennet-curd cheeses, as it tends to pro-duce cheese with a high moisture content and al-tered texture (e.g., lower elasticity and firmness),altered flavor (e.g., hydrolytic rancidity), andaltered functionality (e.g., reduced flow) to a de-gree dependent on milk composition and homog-enization temperature and pressure. The reduc-tion in firmness has been exploited as a means ofimproving the texture of reduced-fat cheeses,which tend to be excessively firm and elastic as aconsequence of their relatively high protein:fatratio. Cheese produced from homogenized milkis whiter than that from unhomogenized milk, anattribute that may be desirable (e.g., in the case ofBlue cheese or Mozzarella) or undesirable (e.g.,in the case of Swiss-type cheese). The main appli-cations of homogenization in cheese manufactureare these:

Cheeses made from recombined milk areformed by homogenizing oils (butter oiland/or vegetable oils) in aqueous dis-persions of milk protein (e.g., reconstitutedor reformed skim milks). This method isused in countries where the demand forcheesemaking exceeds the local supply offresh milk.In the case of fresh acid-curd cheeses (e.g.,Cream cheese), especially when the fat con-tent of the milk is high (e.g., 10%, w/w),homogenization contributes to product ho-mogeneity, as it retards creaming of the fatglobules during the relatively long gelationperiod (-12 hr) and improves producttexture by increasing the level of effectiveprotein, as the casein-coated fat globulesbecome part of the gel rather than being oc-cluded within the gel, as with native fatglobules in milk.In Blue cheese manufacture, the casein-based fat globule membrane allows accessfor Upases from the mold to the fat andthereby enhances the formation of free fattyacids, the main substrates for the produc-tion of methyl ketones, which are very im-portant for flavor.

Page 211: Cheese Science

The effects of homogenization on composi-tion, quality, and yield have been investigatedfor many rennet-curd cheeses (Jana &Upadhyay, 1992). Homogenization of milk orcream increases the yield of practically all vari-eties, the effect being more pronounced with

increasing homogenization pressure (in therange 5-25 MPa)homogenization of standardized cheesemilk rather than homogenization of thecream used for standardization

The reported yield increase, which rangesfrom 2.8% to 6.3% for Cheddar, is attributed tothe increased moisture content (e.g., 20-30 g/kgfor Cheddar, depending on homogenization andcheesemaking conditions) and the lower lossesof fat in whey (e.g., 2-6% milk fat) and/or instretch water in the case of pasta-filata cheeses.However, the efficiency of recovery of fat fromthe whey from homogenized milk is only about30^40% of that from nonhomogenized milk us-ing centrifugal separators.

Microfluidization is a relatively new technol-ogy that has been applied in the health care in-dustry since the late 1980s for the reduction ofmean particle size in the manufacture of prod-ucts such as antibiotic dispersions, parenteralemulsions, and diagnostics. Microfluidizationdiffers from homogenization in the types offorces applied to the fluid and also in the sizedistribution and mean diameter of the resultingparticles.

It is generally accepted that, given the appli-cation of equivalent pressures to the milk, mi-crofluidization results in a lower mean fat glob-ule diameter (e.g., 0.03-0.3 um vs. 0.5-1.0 |um)and a narrower size distribution than homogeni-zation. Moreover, the fat globule membrane inmicrofluidized milk has a high proportion offragmented casein micelles and little or no wheyprotein compared with that in homogenizedmilk. Microfluidization of milk (at 7 MPa) orcream (at 14 or 69 MPa) results in a higher mois-ture content in Cheddar, higher retention of milkfat, and higher actual and moisture-adjustedyields (Table 9-5).

9.5.9 Type of Starter Culture and GrowthMedium

Casein is a major contributor to cheese yield,directly and indirectly, as it forms the matrix thatoccludes the fat and moisture; the latter containsdissolved substances, including native wheyproteins, nonprotein nitrogen (NPN), lactate,and soluble salts. All other conditions beingequal, the recovery of casein from cheese milk istheoretically higher when cheese is made by di-rect acidification through the addition of acidand/or acidogen (e.g., gluconic acid-8-lactone)rather than by a starter culture. Starter cultureshydrolyze casein to varying degrees, dependingon their proteolytic activity, during preparationof the bulk culture and during curd manufactureto small peptides that are ingested by the cellsfor growth (see Chapter 5). Model studies, usingskim milk acidified by starter culture or directacidification, have shown that starter culturesproduce significantly higher losses of casein(0.7-6.6%) than direct acidification, dependingon the proteolytic activity of the culture. Yetstarter cultures are generally used in preferenceto direct acidification as a means of acidifica-tion, principally because of their contribution tocheese flavor development (see Chapters 11 and12). In addition, the starter culture ferments re-sidual lactose in cheese curd to lactic acid,thereby removing it as a carbon source for thegrowth of nonstarter LAB (NSLAB). However,direct acidification is sometimes used as theprincipal or sole means of acidification in themanufacture of some cheeses, for example,where flavor resulting from starter activity is nota major quality attribute (e.g., low-moistureMozzarella, Cream cheese, Queso bianco,Ricotta, and Paneer) or where flavor can be pro-vided by alternative means, such as flavoreddressing (e.g., creamed Cottage cheese).

Proteinase-negative single-strain starters gen-erally give higher dry matter yields of Cheddarcheese than the corresponding proteinase-posi-tive starters, with the yield increase rangingfrom 1.4% to 2.4%, depending on the starterstrain. Similarly, in model acidified skim milksystems, the recovery of casein with proteinase-

Page 212: Cheese Science

negative starter is significantly higher (3.1%)than with the corresponding proteinase-positivestrain and only marginally lower (-0.69%) thanthat obtained using direct acidification. Protein-ase-negative strains, which rely on the indig-enous amino acids and small peptides in the milkfor growth, reproduce very slowly and thereforereduce the pH too slowly for cheese manufac-ture. Moreover, their use may lead to slow pro-teolysis and flavor development during matura-tion. Hence, proteinase-negative strains aregenerally not used alone but rather in blendswith proteinase-positive strains. Such blends arecommonly used as cheese cultures.

There is disagreement in the literature on therelative effects of direct-vat starter versus bulkstarter on the recovery of milk components andcheese yield. Discrepancies between results mayarise from several sources, including

the method of calculating yield (i.e.,whether starter solids are included in theyield calculation)the starter medium used in the preparationof the bulk starter (i.e., whether skim milkbased or whey based)the temperature-time treatment of the bulkstarter medium prior to inoculationmilk composition, pH at set, firmness atcutting, and proteolytic activity of thestarter strains in direct-vat and bulk starterscheese moisture

Differences in these factors can influence thelevel of whey protein denaturation, level of re-coverable solids, degree of casein hydrolysisduring manufacture, and susceptibility of the co-agulum to shattering at cutting. Approximately40% of the bulk starter solids (casein plus dena-tured whey proteins) are retained in Cheddarcheese when the starter medium is 10% reconsti-tuted low-heat skim milk powder heated at 85-9O0C for 1 hr (Banks, Tamime, & Muir, 1985).Some authors have reported increases in actualyield (e.g., ~ 2%) and recovery of milk solids(e.g., 49.2% vs. 48.5%) in Cheddar cheese whendirect-vat starter rather than bulk starter wasused (Salji & Kroger, 1981). In contrast, a Scot-tish study showed that the use of bulk startergrown in 10% reconstituted skim milk powderresulted in higher actual (1.1%) and moisture-adjusted (1.05%) cheese yields and higher reten-tion of milk solids (i.e., 53.1% vs. 52.6%) com-pared to DVS (Banks et al., 1985). Moreover,further studies by the same group (Banks &Muir, 1985) showed that, compared with a com-mercial casein-free starter medium, the use ofreconstituted skim milk powder as starter me-dium resulted in higher actual yield (1.8%) andrecovery of milk solids (1.7%). The greater re-covery of solids was attributed to the retention ofcasein and denatured whey protein from thestarter, which are coagulable following acidifi-cation to pH 4.6 in the bulk starter and re-neutralization to around pH 6.5, when the starter

Table 9-5 Effects of Microfluidization on the Yield of Cheddar Cheese

Milk Cream

OMPa 7MPa 14MPa 69 MPa

Cheese compositionMoisture (g/kg) 350 376 384 393Fat(g/kg) 344 340 334 330

Bulk whey compositionFat (g/kg) 5.2 2.3 1.7 1.3Curd fines (g/kg) 0.9 1.0 1.1 1.2

Cheese yieldActual (kg/1 OO kg) 9.4 10.2 10.4 10.6Moisture-adjusted (kg/1 OO kg) 9.7 10.1 10.2 10.2

Page 213: Cheese Science

is added to the milk. Obviously, further studiesin which factors such as starter strain(s), milkcomposition, set pH, and gel firmness at cuttingare standarized are required to clarify the com-parative effects of bulk and direct-vat starters onCheddar cheese yield.

Significant increases in Cottage cheese yieldhave been reported when whey-based media withexternal pH control (i.e., pH constantly main-tained at the initial value, such as 6.6, by additionof base) were used instead of conventional skimmilk-based media without pH control. In the au-thors' experience, it is important that when a di-rect-vat starter is used, the pH at set should bestandardized to that normally obtained with abulk starter (e.g., to 6.55) by using CaCl2 or al-lowing sufficient ripening time, especially whenthe level of casein in milk is low. Otherwise, gelfirmness after a specified set-to-cut time may below, resulting in shattering of the curds and re-duction in cheese yield due to high losses of fatand curd fines. The addition of direct-vat starterhas little or no immediate effect on milk pH,whereas the addition of bulk starter gives an im-mediate decrease in pH of around 0.1. Moreover,during the ripening or vat filling, the decrease inpH is greater when bulk starter is used.

9.5.10 Addition of CaCl2

The addition OfCaCl2 at a level of about 0.02g/L (i.e., ~ 2 mM Ca) to milk is common com-mercial practice, especially when using late lac-tation milk. Addition of CaCl2 generally im-proves the rennet coagulation properties, aneffect attributable to the reduction in pH and theincrease in the concentration OfCa2+ (see Chap-ter 6). While the effect OfCaCl2 addition on ren-net coagulation properties has been studiedextensively, comparatively few studies haveconsidered its effect on cheese yield. An investi-gation on the commercial manufacture of Swiss-type cheese showed that the addition of CaCl2

(0.1 g/L) produced insignificant increases in themean recovery of milk fat (85.3% vs. 84.7%)and nonfat milk solids (33.85% vs. 33.75%) anda significant increase in the mean cheese yield(0.038 kg/100 kg) (Wolfschoon-Pombo, 1997).

The proportion of large curd particles (5.5-7.5mm) was increased, and the proportion of smallparticles (< 3.5 mm) was reduced. These trendssuggest that the positive effects of CaCl2 on therecovery of fat and protein and cheese yieldprobably ensue from an enhanced degree ofcasein aggregation, which reduces the suscepti-bility of the curd to fracturing during cutting andthe initial phase of stirring (see Section 9.5.12).

9.5.11 Rennet Type

Ideally, rennets should hydrolyze only thePheios-Metloe bond of K-casein during milk co-agulation, with further cleavage of caseins oc-curring only after complete removal of whey. Inthis situation, the recovery of casein is maxi-mized and cheese yield increased. The variousrennets used in cheesemaking differ in their milkclottingiproteolytic activity ratio and thus hy-drolyze casein to a greater or lesser degree dur-ing cheese manufacture, depending on the lengthof time the curd is in contact with the whey andthe curd pH at whey drainage. Some breakdownproducts of casein are soluble in whey and areremoved and lost in the whey at whey drainage.Calf chymosin is the least proteolytic of the gas-tric proteinases, the proteolytic activity of whichdecreases in the following general order:chicken pepsin > porcine pepsin > ovine pepsin> bovine pepsin > calf rennet (chymosin) ~ fer-mentation-produced chymosin. Microbial ren-nets are also more proteolytic than calf chy-mosin, with proteolytic activity being in thefollowing order: Cryphonectria parasitica pro-teinase » Rhizomucor miehei > R. pusillus >calf chymosin.

However, the relative proteolytic activity ofdifferent rennets is not always reflected incheese yield. Indeed, many discrepancies existbetween reported results on the effects of coagu-lant on cheese yield. In a laboratory-scale Cana-dian cheesemaking study (Emmons & Beckett,1990), the increase in nonprotein N level (ex-pressed as whey protein) in Cheddar cheesewhey, compared with calf rennet, ranged from0.006% for bovine pepsin to 0.19% (w/w) forBacillus poly myxa proteinase (Figure 9-15).

Page 214: Cheese Science

The corresponding estimated reduction in mois-ture-adjusted cheese yield ranged from 0.16% to4.5%, respectively. In a pilot-scale study involv-ing 9 to 12 replicate trials, Barbano andRasmussen (1994) found that the losses of fatand protein in whey obtained with calf rennet orrecombinant chymosin were similar and lowerthan those with R. miehei, R. pusillus, or bovinerennet (78% bovine pepsin, 22% chymosin).The moisture-adjusted Cheddar yield was high-est for calf rennet and recombinant chymosinand lowest for R. pusillus (Figure 9-16). A simi-lar study by Ustanol and Hicks (1990) showedthat the above coagulants had no significant ef-fect on N or fat levels in whey or on dry mattercheese yield. However, compared with the otherrennets, C. parasitica proteinase produced sig-nificantly higher levels of N and fat in the wheyand a lower dry matter cheese yield. The de-crease in dry matter yield of Cheddar cheese thatresulted from using a C. parasitica coagulantwas eliminated upon the addition of CaCl2 at alevel of 0.02%. The discrepancy between studiesmay be attributed in part to differences in curdpH at whey drainage (e.g., 5.85 vs. 6.1). At thelower pH, all coagulants are more proteolytic,especially those with a low ratio of milk clottingto proteolytic activity, and therefore producemore soluble peptides, which are lost in thewhey. Hence, cheese yield decreases as the pHat whey drainage decreases.

Generally, no significant differences havebeen reported between the commercially avail-able recombinant chymosins and calf rennet inrelation to fat and N levels in the whey and mois-ture-adjusted cheese yield.

In conclusion, the extent of casein hydrolysisduring the manufacture of cheese curd is lowestwith calf rennet and recombinant chymosins, in-termediate with bovine pepsin and Rhizomucorrennets, and highest with C. parasitica and Ba-cillus polymyxa proteinases. Whether these dif-ferences in proteolytic activity impact signifi-cantly on cheese yield probably depends largelyon the pH at whey drainage. Rennets with a highlevel of proteolytic activity (compared with calfrennet) probably have little effect on yield whenthe pH at whey drainage is high (e.g., > 6.15), as

in the case of Cheddar, Gouda, and Emmental,but reduce yield when the pH at drainage is be-low 6.0, as in the case of Blue cheese and Cam-embert. The thermostability of the different ren-nets at the cook temperature for a given varietyprobably also determines how differences inproteolytic activity impact yield.

9.5.12 Firmness at Cutting

Cutting the gel is a central part of cheesemanufacture, being the first step in the dehydra-tion process by which the colloidal constituentsof milk (fat, casein, and micellar salts) are con-centrated to form cheese curd. The effect offirmness at cutting on cheese yield has been asubject of extensive research in recent years ow-ing to the increasing competitiveness of cheesemanufacture.

During gel formation, firmness increases pro-gressively from the onset of gelation as a conse-quence of ongoing aggregation of paracaseinmicelles (see Chapter 6). Eventually, the gelreaches a firmness that allows it to withstandmechanical cutting by the knives in the cheesevat without shattering. Traditionally, in com-mercial cheese manufacture, and still in most re-ported experimental studies, the curd particlesare allowed to sit quiescently in the whey aftercutting. During this period, referred to as heal-ing, syneresis proceeds rapidly and the curd par-ticles heal, that is, become firmer and develop asurface film, which is essentially an outer layerwith a higher casein:fat ratio than the interior.The combined effects of the film and the cush-ioning effect of the expressed whey limit thedamage inflicted on the curd particles by impactwith the agitators and vat surfaces and by thevelocity gradients during the initial phases ofstirring. Hence, healing reduces the tendency ofcurd particles to shatter (i.e., fracture along theirweakest points into smaller particles with jaggededges). The surface film becomes progressivelystronger as a consequence of the dehydrating ef-fects of heat, acid, and stirring (which createspressure gradients over the surface, forcing newaggregation sites in the interior of the curd par-ticle), and it seals the fat and casein within thecurd particles. The skin develops into curd gran-

Page 215: Cheese Science

Rennet type

Figure 9-15 Effect of different rennets on the level of nonprotein nitrogen, expressed as percentage protein, inbulk Cheddar cheese wheys. Calf rennet (A), R. miehei (B), R. pusillus (C), C. parasitica (D), and B. polymyxa(E).

Non

pro

tein

N in

whe

y, e

xpre

ssed

as

prot

ein,

g/1

00 g

A B C D E

RC CR BR RM RP

Rennet type

Figure 9-16 Effect of coagulant type on moisture-adjusted Cheddar cheese yield. Coagulants: RC, recombinantchymosin (Chymax); CR, calf rennet (94% chymosin); BR, bovine rennet (78% bovine pepsin, 22% chymosin);RM, Rhizomucor miehei (Morcurd plus); RP, Rhizomucorpusillus (EMPORASE sf 100).

Moi

stur

e-ad

just

ed C

hedd

ar c

hees

e y

ield

, kg

/100

kg

Page 216: Cheese Science

ule junctions in the molded cheese curd, whichare readily recognizable upon microstructuralanalysis of the cheese.

In large modern factories, the curds are notgiven a defined period for healing. Instead, thecurds may heal during the cutting program, to agreater or lesser degree depending on the pro-gram, which determines the number and dura-tion of alternate cutting and rest cycles. In prac-tice, cutting cycles for Cheddar may range fromaround 10 to 20 min, depending on knife speedduring successive cutting cycles.

Shattering is undesirable, as it results in an in-crease in the surface area through which fatglobules can escape from the surface of the curdparticle, along with the outflow of whey imme-diately after cutting and during the early stagesof stirring. Moreover, curd shattering results inthe formation of curd fines (curd particles lessthan 1 mm), which may be lost from the curdmass to a greater or lesser degree depending ondownstream curd-handling equipment (ex-vat).

Consideration of the micro structure of therenneted milk gel and its interaction with thecheese knives suggests that cutting when the gelis too soft or too firm increases the propensity toshattering. When the gel is too soft, the gel struc-ture is insufficiently developed and fractureseven under the small strains applied upon gentlecutting. In an overfirm gel, the degree of caseinaggregation is relatively high, and the gel tendsto be brittle (low fracture strain) and susceptibleto breakage. Observations at the practical levellargely substantiate this analysis. In large mod-ern cheese plants, the gel is generally cut after aspecified set-to-cut time (e.g., 40 min) to con-form to factory schedules. However, many fac-tors that affect gel firmness do not remain con-stant throughout the cheesemaking season.Hence, the firmness at cutting can vary, resultingin variations in cheese yield. Several factors in-fluence the firmness of the rennet-induced milkgel after specified times (see Chapter 6), includ-ing milk composition, stage of lactation, somaticcell count, milk heat treatment, culture type, andpH. Hence, much attention has been focused onquantifying the effect of cutting the gel after dif-

ferent set times or at different firmness (whencurd firmness sensors were available) on cheese-making efficiency.

Increasing the set-to-cut time from 80% to200% of the control optimum value (a value de-termined subjectively by the cheesemaker)while maintaining a constant heal time results inhigher levels of cheese moisture and moisture-adjusted Cheddar cheese yield and greater reten-tion of milk fat and solids (Figure 9-17). Thesetrends have been attributed to the more completedevelopment of the gel structure at cutting,which enhances the retention of milk solids.However, there is an interactive effect betweenfirmness at cutting and heal time (Figure 9-18).The positive effect of healing was small at set-to-cut times greater than 85% of the control butbecame increasingly greater at set-to-cut timesless than 85% of the control. As the healing timeis reduced, the yield-enhancing effects of in-creasing firmness at cutting become markedlygreater, especially when the curd is underset atcutting, that is, at set-to-cut times less than 95%of the control (Figure 9-18). Factors that con-tribute to undersetting for a given set-to-cut timeinclude increases in SCC and milk pH and a de-crease in milk protein level. Bynum and Olson(1982) reported that the effects of increasing set-to-cut time on fat recovery and Cheddar cheeseyield depended on vat size. Increasing curd firm-ness had no effect on fat recovery or moisture-adjusted Cheddar yield when small experimen-tal vats (460 L) were used and had a positiveeffect (i.e., resulted in higher yield and higherrecovery of fat and casein) when larger vats(2,400 L) were used. The observed differencesbetween large and small vats may be attributedto differences in stirrer design and to how rap-idly clumps of curd particles are broken up dur-ing the initial phase of stirring. In large vats,clumps of curd particles tend to disintegratemore slowly, which has the effect of increasinghealing time. Other studies (Banks & Muir,1984) have shown that varying set-to-cut times(to produce underset and overset gels at cutting)had no significant effect on moisture-adjustedCheddar cheese yield.

Page 217: Cheese Science

Set-to-cut time, % of control

Figure 9-18 Effect of set-to-cut time and healing time on moisture-adjusted Cheddar cheese yield. Healing timewas O min (•), 5 min (•), and 10 min (Q) min.

Moi

stur

e-ad

just

ed c

hees

e yi

eld,

kg/1

OO

kg m

ilk

Figure 9-17 Effect of firmness at cutting (varied by changing the set-to-cut time) on the efficiency of Cheddarcheese manufacture.

Set-to-cut time, % of control Set-to-cut time, % of control Set-to-cut time, % of control

Moi

stur

e-ad

just

ed c

hees

e yi

eld,

kg/1

00kg

Fat r

ecov

ery,

% o

f tot

al

Milk

sol

ids

reco

very

,%

of t

otal

Page 218: Cheese Science

9.5.13 Particle Size

The curd particle size distribution (CPSd),during the initial phase of stirring affects yieldefficiency, as it determines the surface areathrough which fat escapes into the whey. How-ever, it is not possible to measure the CPSdat thisstage of cheese manufacture, as the curd par-ticles are still very fragile and would fractureduring the sieving process involved in the deter-mination. Moreover, such a measurement is notvery relevant, as the curd particles fracture dur-ing the initial stages of stirring to a greater orlesser degree, depending on their size, the speedof stirring, and the vat design (Johnston, Dunlop,& Lawson, 1991). In practice, curd particle sizeis measured just prior to whey drainage, whenthe curd particles are resilient enough to resistfracture during assay.

Johnston et al. (1991) investigated the effectsof variations in the speed and duration of cuttingin 20,000 L Damrow (variable speed motor) vatson the efficiency of commercial Cheddar manu-facture. Cut programs with a short duration ofcutting at slow speeds produced a low %CPSd

and relatively high fat losses in the whey where%CPSd was defined as the percentage of totalcurd particles with a size < 7.5 mm. These cut-ting conditions resulted in large curd particles(i.e., high %CPSd), which shattered quickly dur-ing subsequent stirring, thereby increasing thesurface area for loss of fat. A maximum %CPSd

and a minimum fat level were reached after theknives in the Damrow vat had completed 37-40rotations (Figure 9-19). Hence, the time re-quired to obtain the maximum %CPSd decreasedas knife speed increased (e.g., -18 min at 2 rpmand ~ 8 min at 5 rpm). Fat losses in whey de-creased and hence cheese yield increased as thenumber of revolutions completed by the knivesreached 37-40 (Figure 9-19). A further increasein the number of knife revolutions had little ef-fect on fat losses. This study indicates that for aparticular vat and knife design, the %CPSd andhence fat losses in whey are influenced by acombination of the speed and duration of cuttingand the subsequent speed of stirring prior tocooking. For a given vat design, proper mainte-

nance of knives (i.e., edge and knife angle) isessential to enable clean cutting and thereby re-duce the risk of tearing the curds and causinghigh fat losses in cheese whey.

9.5.14 Design and Operation of the CheeseVat

The design and operation of the cheese vathave a large influence on cheesemaking effi-ciency, cheese composition, and cheese quality.In all mechanized vats, the coagulum moves to agreater or lesser degree as soon as the knives areswitched on, with a firm coagulum movingfaster than a soft gel. Cheese vats of differentdesign and mode of operation that enable theknives to cut the moving curd efficiently havebeen developed. Design features that allow theknives to "catch up" with the coagulum movingbefore it include

side-mounted baffles that push the curdonto the knivescontinuously variable-speed knife drivescapable of speeds up to 12 rpmintermittent cutting cycles, followed by restperiods, which permit the knives to cutthrough the settling curd layers

A survey (Phelan, 1981) of Irish commercialCheddar-manufacturing facilities showed thatthe fat content of whey varied markedly with thetype of vat used and season of the year (Figure9-20). The increased level of fat in the whey asthe cheesemaking season progressed was attrib-uted to the increase in the level of milk fat andnot to variations in the percentage of milk fat re-covered, which remained relatively constantthroughout the season. In a series of Dutch stud-ies (during the period 1976-1987), 6 six differ-ent vat types were compared. The level of fat inwhey ranged from 5.7% to 7.2% of the total andcurd fines from 97 to 179 mg/kg whey during theproduction of Gouda cheese. A further Irishstudy (see IDF, 1994) showed that Tebel Ost IVvats and APV OCT vats gave similar mean fatlevels in the whey (0.30%, w/w), levels signifi-cantly lower than that for a W-vat (0.47%, w/w).These studies indicate that under normal operat-

Page 219: Cheese Science

ing conditions some vat designs are inherentlybetter than others in maximizing the retention offat in cheese curd, probably as a consequence ofa more favorable curd particle size distribution(CPSd). However, for any vat design, maximumefficiency requires in-plant studies to optimizethe interactive effects of coagulum firmness,cutting program, and stirring speed so as toachieve the best CPSd and fat retention results.

9.5.15 Curd-Handling Systems

Most of the losses during cheese manufactureoccur in the cheese vat. For example, about 6.5%of milk fat and about 4-5% of the casein (owingto loss of the glycomacropeptides) are lost dur-ing commercial Cheddar manufacture (Table9-6). The losses that occur ex-vat are compara-tively small (e.g.,« 0% of the casein and 2.0% ofthe total milk fat.) Nonetheless these losses areimportant determinants of cheesemaking effi-ciency. Following removal of most of the wheyat drainage—on drainage belts with overheadstirrers, which agitate the curd—the curd is sub-jected to variety-specific conveying and han-

dling processes, such as stirring, cheddaring,milling, salting, and prepressing. During theseoperations, moisture and fat are lost to varyingdegrees, thereby affecting actual cheese yield.Milling of cheddared curd exposes fresh sur-faces from which fat is lost to an increasing ex-tent in the whey (so-called salt whey) with el-evation of temperature, reduction in chip size,and severity of mechanical squeezing either inthe mill or on worm conveyors. It is also con-ceivable that the level of fat in whey from blockformers increases with the level of vacuum. Ow-ing to the general nonavailability and high costof downstream pilot-scale curd-handling sys-tems, little published information is available onthe comparative effects of different systems(e.g., CheddarMaster vs. Alfamatic or strainervat vs. Casomatic) under different operatingconditions on cheese yield efficiency.

9.6 CONCLUSION

Cheese is a very important trade item in thedairy industry (~ 35% of total milk production).In the production of commodity cheeses, such as

Total revolutions of knife

Figure 9-19 Effect of number of cheese knife revolutions (rpm x min) on the amount of curd particles smallerthan 7.5 mm as percentage of total (O) and on the fat level in cheese whey (•).

% C

urd

parti

cles

< 7

.5 m

m

Fat l

evel

in w

hey

, g/

100

g

Page 220: Cheese Science

Cheddar and Gouda, increasing the scale of pro-duction and cheesemaking efficiency are essen-tial for reducing production costs and ensuringcompetitiveness and survival in the marketplace.Hence, there is much interest in cheese yield at thecommercial level, as it determines the profits thataccrue to a cheese plant and the price it can affordto pay for milk. The plant's profits are also deter-mined by its cost-effectiveness in recovery fromby-streams (e.g., cheese whey and stretch water)and the production of whey byproducts. Theseaspects are discussed in Chapter 22.

Cheese yield is influenced by many factors,including the composition and quality of the rawmilk, milk-handling and -storage practices, milkpretreatments (e.g., the pasteurization tempera-ture), the cheesemaking process (i.e., the make

procedure, equipment, and technology), andcheese composition (e.g., moisture). Maximiza-tion of cheese yield requires a comprehensiveknowledge of milk composition, the factors thatinfluence it, gelation, and the influence of thecheesemaking process on the gel. Measurementof cheesemaking efficiency is essential so thatinefficiencies can be redressed. Indices ofcheesemaking efficiency include cheese yieldand/or recovery of components, especiallycasein and fat. Comparison of actual and pre-dicted yields allows a cheese plant to monitor itsefficiency over time. However, agreement be-tween actual and predicted yields does not implythat the yield is at a maximum. Moreover, theuse of plant-specific yield prediction formulaedoes not allow comparison of yields from an in-

Month

Figure 9-20 Effect of three types of vat on the level of fat in Cheddar cheese whey from Irish factories.

April May June July Aug Sept Oct

Fat i

n w

hey,

g/1

00g

Page 221: Cheese Science

dividual plant with those from other plants orpublished reports. In modern cheesemaking, fullyield potential is not yet achievable, as fat andprotein recovery is still less than 100%. Un-doubtedly, in the future a closer realization ofmaximum yield will ensue from continued im-provements in a number of areas, including

• milk quality• seasonal variations in the composition of

cheese milk (minimized through improved

REFERENCES

Auldist, M.J., Coats, S., Sutherland, B.J., Mayes, J.,McDowell, G.H., & Rogers, G.L. (1996). Effects of so-matic cell count and stage of lactation on raw milk com-position and the yield and quality of Cheddar cheese.Journal of Dairy Research, 63, 269-280.

Banks, J.M., & Muir, D.D. (1984). Coagulum strength andcheese yield. Dairy Industries International, 49(9}, 17-21,36.

Banks, J.M., & Muir, D.D. (1985). Incorporation of the pro-tein from starter growth medium in curd during manufac-ture of Cheddar cheese. Milchwissenschaft, 4O9 209-212.

Banks, J.M., Tamime, A.Y., & Muir, D.D. (1985). The effi-ciency of recovery of solids from bulk starter in Cheddarcheesemaking. Dairy Industries International, 50(1), 11—13,21.

Barbano, D.M., & Rasmussen, R.R. (1994). Cheese yieldperformance of various coagulants. In Cheese yield andfactors affecting its control [Proceedings of IDF Seminar,Cork, Ireland, 1993]. Brussels: International Dairy Fed-eration.

milk production practices and/or standard-ization to a consistent casein level usingLCF-UF)

• standardization of casein and fat in milk(using online methods)

• starter cultures (development of culturesthat cause less hydrolysis of casein duringmanufacture)

• firmness at cutting (through use of in-vatcurd firmness sensors)

• equipment and process design

Burvenich, C., Guidry, A.J., & Paape, MJ. (1995). Naturaldefence mechanisms of the lactating and dry mammarygland. In A. Saran & S. Soback (Eds.), Proceedings of theThird IDF International Mastitis Seminar. Haifa, Israel:M. Lachmann.

Bynum, D.G., & Olson, N.F. (1982). Influence of cut firm-ness on Cheddar cheese yield and recovery of milk con-stituents. Journal of Dairy Science, 65, 2281-2290.

Emmons, D.B., & Beckett, D.C. (1990). Milk clotting en-zymes: 1. Proteolysis during cheesemaking in relation toestimated losses of yield. Journal of Dairy Science, 73,8-16.

Fenelon, M.A., & Guinee, T.P. (1997). The compositional,textural and maturation characteristics of reduced-fatCheddar made from milk containing added Dairy-Lo™.Milchwissenschaft, 52, 385-389.

Fox, P.F. (1969). Effect of cold-ageing on the rennet coagu-lation time of milk. Irish Journal of Agricultural Re-search^, 175-182.

Table 9-6 Mass Balance of Fat during Cheddar Cheese Manufacture

Constituent Weight (kg) Fat (kg) Fat (% Total Fat)

Milk 10,000 335 100Cheese 927 306.5 91.5Whey

Drain whey 8,750 21.9 6.50White whey

Cheddaring tower 227 4.6 1.37Salting belt 68 1.4 0.42Block former 28 0.6 0.18

Page 222: Cheese Science

International Dairy Federation. (1991). Factors affecting theyield of cheese (Special Issue No. 9301). Brussels: Au-thor.

International Dairy Federation. (1994). Cheese yield andfactors affecting its control [Proceedings of IDF Seminar,Cork, Ireland, 1993]. Brussels: Author.

Jana, A.H., & Upadhyay, K.G. (1992). Homogenisation ofmilk for cheesemaking [A review]. Australian Journal ofDairy Technology, 47, 72-79.

Johnston, K.A., Dunlop, P.P., & Lawson, M.F. (1991). Ef-fects of speed and duration of cutting in mechanisedCheddar cheesemaking on curd particle size and yield.Journal of Dairy Research, 58, 345—354.

Lucey, J., & Kelly, J. (1994). Cheese yield. Journal of theSociety of Dairy Technology, 47, 1—14.

Phelan, J.A. (1981). Standardisation of milk for cheese-making at factory level. Journal of the Society of DairyTechnology, 34, 152-156.

Politis, L, & Ng-Kwai-Hang, K.F. (1988a). Association be-tween somatic cell count of milk and cheese-yielding ca-pacity. Journal of Dairy Science, 71, 1720-1727.

Politis, L, & Ng-Kwai-Hang, K.F. (1988b). Effects of so-matic cell counts and milk composition on the coagulat-ing properties of milk. Journal of Dairy Science, 71,1740-1746.

Salji, J.P., & Kroger, M. (1981). Effect of using frozen con-centrated direct-to-the-vat culture on the yield and qualityof Cheddar cheese. Journal of Food Science, 48, 920-924.

Ustanol, Z., & Hicks, C.L. (1990). Effect of milk clottingenzymes on cheese yield. Journal of Dairy Science, 73,8-16.

Walsh, C.D., Guinee, T.P., Reville, W.D., Harrington, D.,Murphy, JJ., O'Kennedy, B.T., & Fitzgerald, RJ.(1998). Influence of k-casein genetic variant on rennet gelmicrostructure, Cheddar cheesemaking properties andcasein micelle size. International Dairy Journal, 8, 707-714.

Wolfschoon-Pombo, A.F. (1997). Influence of calcium chlo-ride addition to milk on the cheese yield. InternationalDairy Journal, 7, 249-254.

Page 223: Cheese Science

10.1 GENERAL FEATURES

The initial objective of cheesemaking was toconserve the principal constituents of milk, andhence any changes that occurred during storagewere unintentional. Cheese curd contains a di-versity of microorganisms and enzymes, andtherefore biological, biochemical, and probablychemical changes can be expected to occur un-less the preservative factors are sufficient to pre-vent them. In most cases, the cheese environ-ment is not sufficiently severe to prevent theactivity of enzymes originating from the milk,the coagulant, and the starter and nonstarter mi-croorganisms. The action of these enzymes andof the secondary microflora induces changes inflavor and texture, a process referred to as ripen-ing or maturation. The changes that occur duringripening are responsible for the characteristicflavor (taste and aroma), texture, and in mostcases appearance (e.g., the formation of eyes,growth of molds) of the individual varieties. Thechanges range from the very limited (e.g., Moz-zarella) to the very extensive (e.g., Bluecheeses). The duration of ripening ranges fromabout 3 weeks (e.g., Mozzarella) to 2 or moreyears (e.g., Parmesan and extra mature Ched-dar). The rate of ripening is directly related tothe moisture content of the cheese and inverselyrelated to its salt content.

Although some high-moisture, rennet-coagu-lated curd is consumed fresh (e.g., the SpanishBurgos cheese), most rennet-coagulated cheeses

are ripened to at least some extent. In contrast,all acid-coagulated cheeses are consumed fresh.Why acid-coagulated cheeses are not ripened isnot clear. As discussed in Chapter 16, acid-co-agulated curds do not synerese as well as rennet-coagulated curds, and consequently they have ahigh moisture content and would ripen and dete-riorate very rapidly; however, moisture contentcould probably be reduced, if desired. In order toreduce the moisture content of acid-coagulatedcurds, a high cook temperature is used (e.g.,550C for Cottage and - 6O0C for thermo Quarg).As a result, significant numbers of the starterbacteria are killed and their enzymes are exten-sively denatured. The heat/acid-coagulatedcheeses are subjected to a high temperature andare consumed fresh. Hence, only rennet-coagu-lated cheeses are ripened.

The quality of cheese is determined mainly byits flavor and texture, and thus considerable ef-fort has been devoted to elucidating the principalmicrobiological and biochemical changes thatoccur during ripening. The appearance of many,perhaps most, varieties changes during ripening.These changes include the formation of holes,called eyes, in Swiss-type and to a lesser extentin Dutch-type cheeses, the growth of mold onthe surface (e.g., Brie and Camembert) or inte-rior (Blue varieties), or the growth of microor-ganisms on the surface (smear-ripened cheeses).Since the changes in appearance are visuallyperceptible, they are the criteria by which theconsumer initially judges cheese quality and

Microbiology of Cheese Ripening

CHAPTER 10

Page 224: Cheese Science

hence are of major significance. These changesare not just cosmetic—they are visual evidencethat the flavor and texture are satisfactory. Theabsence of eyes in Swiss cheese indicates thatthe propionic acid fermentation has not occurredsatisfactorily and that therefore the flavor is un-likely to be satisfactory. The absence of mold inmold-ripened cheese clearly indicates unaccept-able quality, and of course the growth of moldson nonmold varieties indicates spoilage.

Traditionally, the surface of cheese was ex-posed to the atmosphere, and hence loss of mois-ture occurred. The loss of moisture is especiallycritical for varieties in which the growth of mi-croorganisms on the surface is a key feature ofripening, and thus such varieties are ripened inhigh-humidity environments, traditionally incaves with naturally high humidity and fre-quently now in environments with artificiallycontrolled humidity.

Loss of some moisture through evaporationfrom the surface is not critical for internally bac-terial-ripened cheeses, and if controlled, it leadsto the formation of a rind (a low-moisture sur-face layer), which effectively seals the cheese,restricting continued loss of water and prevent-ing the growth of microorganisms on the surface(owing to the low water activity in the rind). Inrinded cheeses, there is a moisture gradientwithin the cheese, which affects ripening.

As discussed in Chapter 8, initially there is asalt gradient in surface-salted cheese, but equi-librium is established gradually throughout thecheese.

10.2 MICROBIAL ACTIVITY DURINGRIPENING

The factors controlling the growth of microor-ganisms in cheese include water activity, con-centration of salt, oxidation-reduction potential,pH, NO3~, ripening temperature, and the pres-ence or absence of bacteriocins (produced bysome starters). Individually, the effect of thesefactors may not be very great, but their joint im-pact as so-called hurdles is the real controllingfactor. Other compounds produced during curd

manufacture and ripening (e.g., H2O2 and fattyacids) also inhibit microbial growth, but the con-centrations of these compounds produced by thestarters in cheese are not sufficiently high tohave a significant effect on microbial growth.

10.2.1 Water Activity

All microorganisms require water for growth,but it is the availability of the water rather thanthe total amount present that is the importantfactor. Water availability is expressed in termsof water activity (aw), which is defined as the ra-tio of the vapor pressure over the cheese (p) tothe vapor presence of pure water (/?0) at that tem-perature:

aw=~Po

The value of aw ranges from O to 1.0.Cheese, unless vacuum packed, loses mois-

ture by evaporation during ripening. The pro-teins in cheese are hydrated, and this "bound"water is not available for bacterial growth. Thehydrolysis of proteins to peptides and amino ac-ids and of lipids to glycerol and fatty acids dur-ing ripening reduces the availability of water,since one molecule of water is added at eachbond hydrolyzed. In addition, the salt and or-ganic acids (lactate, acetate, and propionate) aredissolved in the moisture of the cheese and re-duce the vapor pressure. Each of these factorsreduces the aw of cheese during ripening.

Yeast grow at a lower aw than bacteria, andmolds at still lower values. Most bacteria requirea minimum aw of around 0.92 for growth. Thelimit for most yeast is around 0.83, but osmo-philic yeast grow at aw values below 0.60, whilemolds have a lower aw limit of around 0.75.Growth of microorganisms at low aw is charac-terized by a long lag phase, a slow rate of growth(i.e., a long generation time), and a reduction inthe maximum number of cells produced. Each ofthese factors helps to limit the number of cellsproduced. Lactic acid bacteria (LAB) generallyhave higher minimum aw values than other bac-

Page 225: Cheese Science

teria. The minimum aw for Lc. lactis, Sc.thermophilus, Lb. helveticus, and P. freuden-reichii is about 0.93, 0.98, 0.96, and 0.96, re-spectively. The influence of aw on the growth ofsome other microorganisms associated withcheese is shown in Table 10-1. Penicilliumcamemberti is the mold responsible for the whitecoating on Camembert and Brie cheese, andBrevibacterium linens and Debaryomyceshansenii are important microorganisms on thesurface of smear-ripened cheeses. P. camem-berti, B. linens, and D. hansenii can grow slowlyin the presence of 10%, 12%, and 15% NaCl, re-spectively. Staphylococcus aureus and micro-cocci can grow quite well in the presence of6.5% NaCl, which is equivalent to an aw value of0.96. Compared with other fungi, Geotrichumcandidum is very sensitive to aw whereas B. lin-ens is quite resistant. Propionibacteria are alsoparticularly sensitive to aw. Facultative anaer-obes have different minimum aw values depend-ing on whether the organisms are growing aero-bically or anaerobically. For example, in thepresence of O2, S. aureus has a minimum aw of0.86, but in the absence of O2, the minimum is0.91.

Evaporation of water from the cheese surfaceduring ripening also contributes to the reductionof the aw of cheese (examples for Emmental andGruyere are shown in Figure 10-1). The fasterrate of decrease in the aw of Gruyere may be dueto the surface salting of Gruyere during ripening.In addition, the aw of cheese can vary throughoutits mass (Figure 10-2). Variations are muchgreater in large cheeses, like Emmental (50-60kg), than in small cheeses, like Appenzeller (6-8kg). This is due to several factors, including thetemperature gradient in the cheese during theearly stages of the fermentation, the loss ofmoisture during ripening, the NaCl gradient inthe cheese, and microbial activity on the rind.These factors must be taken into account in de-termining the significance of aw, especially inlarge cheeses. Typical aw values for cheese arelisted in Table 10-2. As a comparison, the aw ofmilk is 0.995. Since the aw of cheese decreasesduring ripening, some of these values must be

interpreted with care; however, they are usefulas a guide. Except for the soft cheeses like Brieand Camembert, most of these values are closeto the minima for starter growth cited above.

10.2.2 Salt

The use of NaCl to prevent microbial spoilageof food is probably as old as food production it-self. The concentration required depends on thenature of the food, its pH, and its moisture con-tent, but generally less than 10% is sufficient.The major inhibitory factor is probably the re-duction in aw that occurs when salt (or any sol-ute) is dissolved in water. The relationship be-tween salt concentration and aw is shown inFigure 10-3 and is almost, but not quite, linear.The linear equation is

aw = -0.0007x+1.0042

and describes the relationship very well, sincethe r2 value is 0.997, and x = concentration ofNaCl (g/kg). It is generally considered that an aw

value of less than 0.92 is necessary to preventbacterial growth; this is equivalent to a salt con-centration of 12.4%. In cheese, the salt concen-tration varies from 0.7% to 7%. Therefore, otherfactors are involved in preventing bacterialgrowth in cheese, such as pH and temperature.The ions themselves are also important (e.g.,Na+ is a much more effective inhibitor than K+).In calculating the inhibitory effect of salt incheese, the concentration of salt dissolved in thewater of the cheese rather than the actual con-centration of salt is the important parameter. Forexample, in a Cheddar cheese with 38g mois-ture/lOOg and 1.9g salt/lOOg, the salt-in-waterpercentage is 5%. Generally, the salt-in-waterpercentage in Cheddar cheese varies from 4% to6%.

Cheese is either dry-salted (e.g., Cheddar) orbrine-salted (most cheeses). In brine-saltedcheeses, the salt concentration is influenced di-rectly by the size of the cheese, the concentrationof salt in the brine, the temperature of the brine,and the length of time the cheese is immersed inthe brine (see Chapter 8). Data for the effect of

Page 226: Cheese Science

brining time on the salt concentration and the aw

of Camembert cheese are shown in Figure 10—4.The brine normally used contains about 20%NaCl, has a pH of about 5.2 (adjusted with lacticacid), and a Ca content of 0.2% (adjusted withCaCl2). The pH and Ca concentration simulatethe levels in cheese and help to prevent theefflux of lactate and Ca from the cheese.

10.2.3 Oxidation-Reduction Potential

Oxidation-reduction potential (Eh) is a mea-sure of the ability of chemical or biochemical

systems to oxidize (lose electrons) or reduce(gain electrons). Eh is generally measured usinga platinum electrode coupled with a calomel ref-erence electrode and is expressed in mV. It canalso be estimated using indicator dyes thatchange color at different redox potentials. Apositive value indicates an oxidized state and anegative value indicates a reduced state.

The Eh of milk is about +150 mV whereas thatof cheese is about -250 mV. The exact mecha-nism by which the Eh of cheese is reduced is notclear but is almost certainly related to the fer-mentation of lactose to lactic acid by the starter

Table 10-1 Influence of Water Activity (aw) on the Growth of Different Microorganisms in NutrientBroth, pH 6.6 after 10 Days at 250C (Results are Expressed as Percentage of Maximum Development)

NaCI Concentration (g/100 ml)

O 5 10 15 20

aw

0.992 0.975 0.947 0.916 0.880

MoldsMucor mucedo 540 100.0 47.4 11.6Penicillium candidum 53Il 100.0 80.9 36.4 4.1 1.1Cladosporium herbarum 53b 82.6 100.0 62.4 13.9 3.5Scopulariopsis fusca? 53L1 100.0 78.4 76.9 65.4 14.6

YeastRhodotorulaspp.44a 100.0 69.5 21.8 1.0Debaryomyces spp.54/c 100.0 49.7 30.2 10.5 8.2Geotrichum candidum 53aa 100.0 46.9 -Trichsporon 57k 100.0 62.6 - - -

BacteriaMicrococcus saprophyticus? 55a 100.0 96.3 67.2 19.1 -Micrococcus saprophyticus? 56b 84.7 100.0 61.2 16.7Micrococcus lactis 57h 100.0 45.9 -Brevibacterium linens 58a 100.0 44.1 29.9 13.9 4.1Brevibacterium linens BL107 100.0 67.0 30.0 15.6 3.2Arthrobacter citrans KR3 100.0 19.4 7.0Escherichia coli Strain 54i 100.0 23.4 -Escherichia coli Strain SL 100.0 19.9

Note: - indicates no growth

a These are the names given in the original reference; Micrococcus saprophyticus is likely to be Staphylococcus saprophyticus.

Page 227: Cheese Science

during growth and is probably related to the re-duction of the small amounts of O2 in the milk toH2O (or to H2O2 and then to H2O). Because ofthese reactions, cheese is essentially an an-aerobic system, in which only facultatively orobligately anaerobic microorganisms can grow.Obligate aerobes, like Pseudomonas spp., Brevi-bacterium spp., and Micrococcus spp., will notgrow within the cheese, even when other condi-tions for growth are favorable. Eh is thereforeimportant in determining the types of microor-ganisms that grow in cheese. The bacteria thatdevelop on the surface of cheese are mainly obli-gate aerobes and are unable to grow within theanaerobic cheese environment.

10.2.4 pH and Organic Acids

Most bacteria require a neutral pH value foroptimum growth and grow poorly at pH valuesbelow 5.0. The pH of cheese curd after manufac-ture generally lies within the range 4.5-5.3, sopH is also a significant factor in controlling bac-terial growth in cheese. LAB, especially lactoba-cilli, generally have pH optima below 7, andLactobacillus spp. can grow at pH 4.0. Mostyeast and molds can grow at pH values below

3.0, although their optima range is from 5 to 7.B. linens, an important organism in smear-rip-ened cheese, cannot grow below pH 6.0. Micro-coccus spp., which are commonly found on thesurface of soft cheeses, cannot grow at pH 5 andonly slowly at pH 5.5.

The efficacy of organic acids as inhibitors ofmicrobial growth is thought to depend on theamount of undissociated acid present and there-fore on the dissociation constant (pKa) and pH.The pKas of propionic, acetic, and lactic acids,the principal acids found in cheese, are 4.87,4.75, and 3.08, respectively, so at the same con-centration lactic acid is the least and propionicacid the most effective inhibitor. Propionic acidis very effective at repressing the growth ofmolds. However, the concentration of the acid isalso important, and lactate is invariably presentat much greater concentrations in cheese andcheese curd than either of the other two acids.The pH of mold- and smear-ripened cheesescharacteristically increases during ripening, par-ticularly on the surface, due to the growth ofyeast and molds. Sometimes, it is thought thatthe difference between pH 5.2, the pH of a well-made cheese, and pH 5.4, the pH of a poorlymade cheese, is not very great. However, this is

Days ripening

Figure 10-1 Decrease in the aw of Emmental and Gruyere cheese during ripening. The #wat time zero (0.995)corresponds to that of milk.

Emmental

Gruyere

aw

Page 228: Cheese Science

not so. pH is a log scale, and a difference of 1 pHunit is equivalent to a tenfold difference in theH+ concentration. The difference in [H+] be-tween 5.2 and 5.4 is twofold.

10.2.5 Nitrate

NO3-, as KNO3 (saltpeter) or NaNO3, is addedto the milk (20g/100L) for some cheeses, espe-cially Dutch-type cheeses like Gouda and Edam,to prevent the production of early and late gas by

coliform and Clostridium tyrobutyricum, respec-tively. Much of the NO3- is lost in the whey. Themaximum amount of NO3- permitted in cheese is50 mg/kg, calculated as NaNO3. The real inhibi-tor is NO2-, which is formed from NO3- by theindigenous xanthine oxidase present in the milkor curd. How NO2" acts in preventing microbialgrowth is not clear. NO2- can also react with aro-matic amino acids in cheese to produce nitro-samines, many of which are carcinogenic (seeChapter 21).

Figure 10-2 Typical variations in the aw of slices, from the center to the surface, of (a) Emmental, (b) Sbrinz,(c) Gruyere, and (d) Appenzeller cheese. The cheeses were about 5 months old, and the aw of the rinds was (a)0.90-0.95, (b) 0.80-0.90, and (c, d) 0.92-0.98.

a

b

c

d

Page 229: Cheese Science

Table 10-2 Typical Water Activity (aw) Values for Various Cheeses

Range of Typical TypicalType Typical aw* SD aw in Rind Moisture (%) Salt (%)

Appenzeller 0.960 0.011 0.97-0.98Brie 0.980 0.006 0.98-0.99 48.4 1.91Camembert 0.982 0.008 0.98-0.99 51.8 2.5Cheddar 0.950 0.010 0.94-0.95 36.8 1.5Cottage cheese 0.988 0.006 82.5 1.0Edam 0.960 0.008 0.92-0.94 41.5 2.0Emmentalt 0.972 0.007 0.90-0.95 37.2 1.2Fontal 0.962 0.010 0.93-0.96Gorgonzola 0.970 0.017 0.97-0.99 3.5Gouda 0.950 0.009 0.94-0.95 41.4 2.0Gruyeret 0.948 0.012 0.92-0.98 34.5 1.06Limburger 0.974 0.015 0.96-0.98 48.4 2.74Munster 0.977 0.011 0.96-0.98 41.8 1.8St. Paulin 0.968 0.007 0.96-0.97Parmesan 0.917 0.012 0.85-0.88 29.2 2.67Quarg 0.990 0.005 79.0 0.70Sbrinzt 0.940 0.011 0.80-0.90 42.9 1.90Tilsiter 0.962 0.014 0.92-0.96 2.63Processed cheese 0.975 0.010

* Measured at 250C.

+Values for Emmental, Gruyere, and Sbrinz were measured after ripening for 4-5,6-7, and 10-11 months, respectively. The othervalues were determined in commercial samples of unknown age.

NaCl, g/kg

Figure 10-3 Effect of salt concentration on the aw of brine.

»w

Page 230: Cheese Science

10.2.6 Temperature

Generally, the optimum temperature for thegrowth of bacteria is around 350C for mesophilesand around 550C for thermophiles. Thermophilicstarters have an optimum temperature of around420C. Psychrophilic bacteria have an optimumtemperature below 2O0C, but true psychrophilesare not found in cheese. At temperatures belowthe optimum, growth is retarded. The tempera-ture at which cheese is ripened is dictated by twoopposing requirements—on the one hand, theneed to control the growth of potential spoilageand pathogenic bacteria and, on the other, theneed to promote the ripening reactions and thegrowth of the secondary microflora (in the case ofsoft and Swiss-type cheeses). Higher tempera-tures promote faster ripening by the starter andnonstarter microorganisms but also allow thegrowth of spoilage and pathogenic bacteria. Gen-erally, Cheddar cheese is ripened at 6-80C whileCamembert and other mold and bacterial smear-ripened cheeses are ripened at 10-150C. Em-mental cheese is ripened initially for 2-3 weeks ata low temperature (~ 120C), after which the tem-perature is increased to 20-240C for 2-A weeks topromote the growth of propionic acid bacteriaand the fermentation of lactate to propionate, ac-etate, and CO2. The temperature is then reduced

again to around 40C. For soft cheeses, the humid-ity of the environment is also controlled to pre-vent excessive evaporation of moisture from thecheese surface.

Traditional Emmental cheese is made fromraw milk, and because of the relatively high tem-perature of ripening for this cheese, great atten-tion must be paid to the microbial quality of theraw milk. The need for such attention is miti-gated to some extent by the higher cooking tem-perature (~ 520C) and longer cooking time (60-90 min) to which Emmental curd is subjectedduring manufacture, compared with other hardcheeses like Cheddar, for which the maximumcooking temperature is around 380C.

Increasing the temperature of ripening isprobably the simplest and most cost-effectiveway of accelerating the ripening of cheese (seeChapter 15), but it also increases the rate ofgrowth of other bacteria that may be present.

10.3 GROWTH OF STARTER BACTERIAIN CHEESE

The initial number of starter bacteria in cheesemilk ranges from about 105 to 107 cfu/ml anddepends on the level of inoculation. Subsequentgrowth of the starter and syneresis (contraction)of the curd during manufacture results in starter

Brining time, min

Figure 10-4 Influence of the duration of brining at 140C in 20% NaCl brine on the aw of Camembert cheese. TheNaCl concentration and aw levels were determined 15 days after manufacture.

NaO

, g/lO

Og

chee

se m

oist

ure

*w

NaCl

aw

Page 231: Cheese Science

counts of about 109cfu/g in almost all cheeseswithin 1 day of manufacture. During ripening,starter organisms dominate the microflora ofcheeses but most die off and lyse relatively rap-idly. This is shown in Figure 10-5 for 5 strainsof Lactococcus in Cheddar cheese and in Figure10-6 for Sc. thermophilus and Lb. helveticus inComte cheese. In the case of Cheddar cheese,the rate of death depends on the strain, and in thecase of Comte cheese, the rate of death of Sc.thermophilus is faster than that of the thermo-philic lactobacilli Lb. helveticus and Lb. del-brueckii subsp. lactis. Many artisanal cheeses,especially Spanish varieties, are made withoutthe deliberate addition of a starter. In thesecheeses, lactococci also make up the major partof the microflora and, except for La Serena, alsoshow significant rates of death during ripening(Figure 10-7). The reason for the slow rate ofdeath of lactococci in La Serena cheese may bedue to its relatively low salt concentration duringthe early weeks of ripening.

Once the starter counts begin to decrease, ly-sis usually occurs and intracellular enzymes,particularly peptidases, are released, which, to-gether with chymosin and the starter proteinase,hydrolyze the caseins to peptides and amino ac-ids, which are the precursors of the flavor com-pounds in cheese (see Chapters 11 and 12).Starters vary in their ability to lyse: some strainslyse relatively quickly while others hardly lyseat all. Lysis is caused by an intracellularmuramidase which hydrolyzes the cell wall pep-tidoglycan. This enzyme is under stringent regu-lation; otherwise the cells would not grow. Gen-erally, strains of Lc. lactis subsp. cremoris lysefaster than strains of Lc. lactis subsp. lactis,which may partly explain why the former arethought to produce a better-flavored cheese thanthe latter. Lysis is influenced by several factors,including the level of salt and the presence ofprophage, which is thought to be induced bycooking. The presence of small numbers of lyticphage may also have a role in lysis. Cheese

Cou

nt, l

og c

fii/g

Figure 10-5 Numbers of different strains of Lactococcus in Cheddar cheese during ripening.

Weeks

AMI

AM2

HP

MLl

Z8

Page 232: Cheese Science

made with a fast-lysing starter will ripen morerapidly than those made using slow-lysing cul-tures.

10.4 GROWTH OF NONSTARTERLACTIC ACID BACTERIA IN CHEESE

Most, if not all, cheeses, whether made fromraw or pasteurized milk, contain adventitious

nonstarter LAB. These bacteria are mainly fac-ultatively heterofermentative lactobacilli (groupII), such as Lb. casei, Lb. pararcasei, Lb. plan-tarum, and Lb. curvatus, but Pediococcus spp.and obligately heterofermentative Lactobacillusspp. (group III), such as Lb. brevis and Lb. fer-mentum, are also found occasionally. The spe-cies of group II and III lactobacilli found incheese are referred to as mesophilic to distin-

Figure 10-6 Numbers of Streptococcus thermophilus (St) and Lactobacillus helveticus (Lh) in Comte cheeseduring ripening.

Weeks

Cou

nt, l

og c

fii/g

Lh

St

Lac

toco

cci,

log

cfii/

g

Days

Figure 10-7 Changes in the number of lactococci in several artisanal Spanish cheeses during ripening.

Casar de Caceres

La Serena

Afuegal Pitu

Cabrales

Page 233: Cheese Science

guish them from the thermophilic lactobacilliused as starters. The sources of these bacteriaare the raw milk and the factory environment.Small numbers of some lactobacilli survive pas-teurization and the cooking temperature (520C)used for hard cheeses, like Emmental, which istraditionally made from raw milk. All of thesebacteria are salt and acid tolerant and are facul-tative anaerobes, and therefore they grow quitewell in cheese. Many of the nonstarter lactoba-cilli and pediococci found in cheese can growin the presence of 10% NaCl. They need a fer-mentable carbohydrate for energy production,but the energy source used by them in cheese isunclear, since at the time of exponential growthof nonstarter LAB (NSLAB) no lactose is pres-ent. Possible substrates include citrate and/oramino acids. In model systems, mesophilic lac-tobacilli can utilize the sugar of the glycopro-teins of the milk fat globule membrane as anenergy source.

In contrast to starter cells, the initial numberof NSLAB in cheese varies considerably, fromabout 100 cfu/g in Cheddar cheese to 106 cfu/g inCasar de Caceres (Figure 10-8). Within the firstfew weeks of ripening, however, they grow rela-tively quickly to high numbers (~ 108 cfu/g) inall cheese at a rate that depends primarily on theripening temperature. Their generation time inCheddar cheese ripened at 60C is 8.5 days.NSLAB grow much more rapidly in Casar deCaceres and La Serena cheese than in Comte orCheddar cheese (Figure 10-8). The difference isdue to the higher moisture content of the firsttwo cheeses, which is around 50% at day 4 butsubsequently decreases to about 35% and 45%after 45 days ripening for Casar de Caceres andLa Serena, respectively; the moisture content ofComte and Cheddar cheese is around 38% fromthe beginning of ripening. The faster growthrates in Casar de Caceres and La Serena cheesemay also reflect the higher ripening temperatureused for these cheeses. In addition, Cheddar isthe only one of the cheeses in Figure 10-8 that ismade from pasteurized milk, which probably ex-plains the low initial number of NSLAB in thischeese. The higher rate of growth of NSLAB in

Comte cheese, compared with Cheddar, is likelyto be due to the higher ripening temperature ofComte (3 weeks at 140C, followed by 9 weeks at180C, at which time the temperature is reducedto 70C); Cheddar is kept at 60C throughout rip-ening. A higher temperature is used in the ripen-ing of Comte to promote the growth of propionicacid bacteria. In raw milk cheese, the number ofNSLAB in the curd is higher, they grow faster,and the population is more heterogeneous thanfor pasteurized milk cheeses.

Despite extensive study, the role of NSLAB inthe development of cheese flavor is not clear. Incontrast to starter cells, mesophilic lactobacillidie off very slowly in hard cheese (Figure 10-8).They appear to lack a cell envelope-bound pro-teinase, and since they die off only slowly incheese, their intracellular enzymes are probablynot released into the cheese matrix. Nevertheless,cells of NSLAB are viable and, at the high celldensities found in cheese, exhibit considerablemetabolic activity. NSLAB do contribute to rip-ening, but the significance of this contribution isopen to question. At least in Cheddar cheese, theytransform L lactate to D lactate; racemases are notinvolved. It is likely that L lactate is oxidized topyruvate, which is then reduced to D lactate. Aracemic mixture of both isomers is formed even-tually. Some NSLAB can also oxidize lactate toacetate and CO2 on the cheese surface in the pres-ence ofO2, sharpening the taste of the cut surfacesof the cheese, especially if the cut surfaces remainuncovered for several hours. Pediococci aremuch more active than lactobacilli in formingacetate from lactate.

Cheddar cheese is one of the few varietieswithout a deliberately added secondary micro-flora, but there is considerable interest in inocu-lating milk for Cheddar with selected mesophiliclactobacilli with the objective of accelerating rip-ening and/or intensifying its flavor (see Chapter15).

In virtually all artisanal cheeses, Leuconostocand Enterococcus spp. are also found in largenumbers and contribute to flavor development.Further information on enterococci is given inChapter 20.

Page 234: Cheese Science

10.5 OTHER MICROORGANISMS INRIPENING CHEESE

Many cheese varieties contain a secondary,non-LAB microflora, the function of which is toproduce some specific characteristic change inthe cheese, such as surface growth in the case ofbacterial-ripened (smear-ripened) and mold-ripened cheeses and the production of CO2, pro-pionate, and acetate in the case of some Swissvarieties (e.g., Emmental and Comte). CO2 is re-sponsible for eye formation in these cheeses. Inall of these cheeses, flavor development is domi-nated by the metabolic activity of the secondaryflora.

Several microorganisms are involved, includ-ing bacteria (Arthrobacter, Brevibacterium,Brachybacterium, Corynebacterium, Microbac-terium, Propionibacterium, and Micrococcusspp.), yeasts (Kluyveromyces marxianus andDebaryomyces hansenii), and molds (Geo-trichum candidum, Penicillium camemberti, andP. roqueforti). They are all involved in ripeningand, except for Propionibacterium spp. and P.roqueforti, develop only on the cheese surface.

The surface microflora has two important func-tions in ripening:

1. production of enzymes2. deacidification of the cheese surface

The enzymes include Upases, proteinases, andpeptidases. The lipases and proteinases hydro-lyze the fat and protein to fatty acids and pep-tides, respectively, while the peptidases hydro-lyze the smaller peptides to amino acids. Bothfatty acids and amino acids are the precursors ofmany of the flavor compounds in mold- andsmear-ripened cheese (see Chapters 11 and 12).

During the first few days of ripening smear-and mold-ripened cheese, yeast and molds growon the cheese surface and deacidify it by oxidiz-ing the lactate to H2O and CO2. In turn, thiscauses the pH of the surface to increase from aninitial value of about 4.8 to 5.8 or higher. Thesurface bacteria grow poorly, if at all, at the lowinitial pH, and the increase in pH promotes theirgrowth considerably.

The presence of molds and yeasts on thecheese surface is to be expected, since cheesehas a relatively low pH (both types of microor-

Days

Figure 10-8 Growth of mesophilic (mainly facultatively heterofermentative) lactobacilli in Casar de Caceres,La Serena, Comte, and Cheddar cheese during ripening.

Cou

nt, l

og c

fii/g

Casar de Caceres

La Serena

Comte

Cheddar

Page 235: Cheese Science

ganism can grow at pH values below 3), areadily fermentable substrate (lactate) and arelatively low aw. Traditionally, cheese becamecontaminated by these microorganisms from theenvironment, and their growth was promoted byhigh relative humidity and/or high temperature(12-240C) in the ripening rooms or caves.Wooden shelves are extremely porous and are alikely source of contamination. Generally, thebacteria and molds are added deliberately to themilk or cheese, but the yeasts are adventitiouscontaminants of the cheese surface. In thosecheeses in which they are found, yeasts grow to106/cm2 of surface within a few days of ripening,after which they generally decrease and stabilizeat 104 to lOVcm2. Several methods are used toinoculate the cheese with molds, including addi-tion to the curd before molding, addition to thebrine, dusting spores on the surface of the curd,and smearing with an aqueous suspension ofmold spores. In modern practice, the milk formold-ripened varieties is inoculated with a pureculture of P. roqueforti, in the case of Bluecheeses, or P. camemberti, in the case of Cam-embert and Brie, at the same time as the starters.The curd for Blue cheese is subsequently piercedto allow limited entry of O2 to promote growthof P. roqueforti. Surface- or smear-ripenedcheeses, like Tilsit, Munster, and Limburger, aredipped, sprayed, or brushed with aqueous sus-pensions of G. candidum and B. linens as soon asthe cheeses are removed from the brine. Bothmold- and bacterial-ripened cheese are then rip-ened at 10-150C at a high relative humidity toprevent the loss of moisture from the cheese sur-face. Traditionally, natural contamination of themilk was relied upon as the source of propionicacid bacteria in the case of Emmental and Comtecheeses, but nowadays these bacteria generallyare added deliberately to the milk with the startercultures.

For a long time, B. linens was thought to bethe most important bacterium growing on thesurface of smear-ripened cheeses. Recent evi-dence (Table 10-3) shows that other bacteria,particularly Arthrobacter globiformis, A. nico-tianae, Cory neb acterium ammoniagenes, C.

variabilis, Microbacterium imperiale, andRhodococcus facians are also important. Itshould be noted that many of the bacteria havenot been identified, implying that the surfacemicroflora is very complicated. In addition, Mi-crococcus spp. are found on the surface ofRoquefort and Comte cheese and probably mostother cheese, but information on the species in-volved is very limited. Brachybacterium alimen-tarium and Br. tyrofermentans are also found onComte and Beaufort cheese. All of the smearbacteria are salt-tolerant (the surface layer ofsurface-ripened cheese can contain up to 15%NaCl), aerobic, or facultatively anaerobic mi-croorganisms and hence grow easily at the highsalt level in the surface layer of these brine-salted cheeses. B. linens does not grow at pHvalues below 6.0. This is also probably true ofthe other bacteria found on the surface of cheese.

Arthrobacter, Brevibacterium, Brachybac-terium, Corynebacterium, and Microbacteriumare generically called coryneform bacteria. Allof them are Gram-positive, catalase-positive,non-spore-forming, and generally nonmotilerods. A major feature of their growth is that ex-ponential-phase cells are pleomorphic, showingthe presence of irregularly shaped rods, includ-ing wedge, club, V, and curved shapes. In addi-tion, Arthrobacter, Brevibacterium, and Brachy-bacterium spp. go through a marked rod-coccuscycle during growth, with rod forms dominatingthe exponential phase of growth (1-2 days) andcoccal forms dominating the stationary phase(5-7 days). Except for Corynebacterium andBrachybacterium, metabolism of sugars bycoryneforms, if it occurs, is respiratory, althoughthe evidence for acid production from glucoseby Micrococcus spp. is conflicting. Cell-wallcomposition (particularly the amino acids [Iy-sine, ornithine, or 2,6-diaminopimelic acid] andsugars [arabinose or galactose] found in the pep-tidoglycan), the presence and type of mena-quinones, the presence or absence of mycolicacids, and whether they go through a rod-coccuscycle during growth are important criteria inidentifying coryneform bacteria. Despite this,their taxonomy is very confusing. All belong,

Page 236: Cheese Science

however, to the actinomycete branch of theGram-positive bacteria. There is now over-whelming biochemical and genetic evidencethat Corynebacterium sensu stricto, which con-tain mycolic acids, are quite unrelated to thecoryneform bacteria (Arthrobacter, Brevibac-terium, and Microbacterium), which lack my-colic acids.

A brief discription of the above genera and ofother microorganisms found in cheese is givenbelow.

10.5.1 Arthrobacter

These are Gram-positive, catalase-positive,nonmotile, obligately aerobic rods that go

through a marked rod-coccus cycle duringgrowth. Their peptidoglycan contains lysine.They have nonexacting nutritional require-ments; generally, biotin is the only vitamin re-quired. Their habitat is soil, and they do notwithstand HTST pasteurization of milk.

10.5.2 Brachybacterium

These are Gram-positive, facultatively an-aerobic short rods that exhibit a rod-coccusgrowth cycle. Their optimum temperature isaround 3O0C. They contain meso-diammo-pimelic acid and glucose, galactose, and rham-nose, but not mycolic acids, in their cell walls.Five species are recognized and two of these, Br.

Table 10-3 Species of Bacteria Found in Smear-Ripened Cheeses

Limburger, Romadour,Weinkase, and Harzer Tilsit Cheese Mainly*

Species Cheese5 (6 Plants) (15 Plants)

Arthrobacter citreus 1 19Arthrobacter globiformis 102C

Arthrobacter nicotianae 14 10Brevibacterium fermentans 3Brevibacterium imperiale 8Brevibacterium linens 25 77d

Brevibacterium fuscum 2Brevibacterium oxydans 1 3e

Brevibacterium helvolum 1Corynebacterium ammoniagenes 36 53Corynebacterium betae 4Corynebacterium insidiosum 8Corynebacterium variabilis 12 14Curtobacterium poinsettiae 12Microbacterium imperiale 5 8Rhodococcus fascians 15Total number identified 112 321Total number not identified 36 73

a N. Valdes-Stauber, S. Scherer, & H. Seller, Identification of yeasts and coryneform bacteria from the surface microflora of brickcheeses, International Journal of Food Microbiology, 34 (1997), 115-119.

bF. Eliskases-Lechner and W. Ginzinger, The bacterial flora of surface-ripened cheeses with special regard to corneforms, Lait, 75(1995), 571-584.

c Seventy-two from 1 plant.d From 11 of 15 plants.e Only from 1 plant.

Page 237: Cheese Science

alimentarium and Br. tyrofermentans, have beenisolated from Comte and Beaufort cheese, re-spectively (Schubert et al., 1996). Their nutri-tional characteristics do not appear to have beenstudied, but Br. alimentarium and Br. tyrofer-mentans can grow in the presence of 14% and16% NaCl, respectively.

10.5.3 Brevibacterium

These are Gram-positive, catalase-positive,nonmotile, obligately aerobic rods that gothrough a marked rod-coccus cycle duringgrowth. Their cell walls contain meso-diamino-pimelic acid, and they metabolize sugars by res-piration. There are four species: B. linens, B.casei, B. iodinum, and B. epidermidis. The firsttwo species have been isolated from cheese, thethird from milk, and the fourth from skin. B. lin-ens produces yellow, orange, red, or brown colo-nies, while those of B. iodinum are purple, due tothe production of a phenazine derivative. Theother two species produce gray-white colonies.Brevibacteria grow poorly, if at all, at 50C, havean optimum temperature of 20-250C, and growin the presence of a high concentration of NaCl.B. linens and B. iodinum grow in the presence of8-10% NaCl, and B. easel and B. epidermidis inthe presence of 15% NaCl. Their nutrition hasnot been studied in depth, but most strains of B.linens require amino acids and vitamins forgrowth. Their metabolism is respiratory, andthey do not produce acid from glucose. They areeasily confused with Arthrobacter spp. B. linensmetabolizes methionine to methional, which isthought to be responsible from the characteristic"dirty sock" odor of smear-ripened cheeses. B.linens can be determined specifically throughthe production of a stable pink color within 2min following treatment of a small amount of acolony with a drop of 5 M KOH or 5 M NaOH ora salmon pink color within 1 min followingtreatment with glacial acetic acid. Brevibacteriaare acid sensitive and will not grow at a pH valuebelow 6.0. Their major habitats are dairy prod-ucts, especially cheese, activated sludge, and hu-man skin.

10.5.4 Corynebacterium

These are Gram-positive, catalase-positive,nonmotile, facultatively anaerobic slightlycurved rods with tapered ends; club-shapedforms may be found also. Currently, 16 species ofCorynebacterium are recognized. A rod-coccuscycle does not occur. Metachromatic granules,which stain deeply with methylene blue, areformed. Meso-diaminopimelic acid and short-chain mycolic acids (22-36 C atoms) are found intheir cell walls. They are nutritionally exacting,requiring several vitamins, amino acids, purines,and pyrimidines for growth. Two bacteria,Microbacterium flavum and Caseobacter poly-morphus, which were isolated from cheese, havebeen reclassified as C. flavescens and C. vari-abilis, respectively. C. variabilis was isolatedfrom the surface of Dutch smear-ripened cheese.This organism produces gray-white, slightlypink, or slightly red colonies.

10.5.5 Microbacterium

These are small, Gram-positive, nonmotile ormotile rods that do not go through a rod-coccuscycle. However, in older cultures (3-7 days), therods are short and a proportion may be coccoid.Currently, there are 13 species, only one ofwhich, M. lacticum, has been found in milk.Their optimum temperature is 3O0C. Coloniesvary in color from gray-white to pale green oryellow. Their cell wall peptidoglycan containsIysine. Generally, their metabolism is respira-tory, but acid is produced from glucose andsome other sugars in peptone-containing media.Most strains require biotin, pantothenic acid,and thiamine for growth. The main speciesfound in milk is M. lacticum, which is thermodu-ric and survives heating at 630C for 30 min. Theorganism is not found in aseptically drawn milk,and there is strong evidence that the majorsource of contamination of milk with this organ-ism is improperly cleaned dairy equipment.

10.5.6 Rhodococcus

These are Gram-positive, catalase-positive,aerobic rods that usually produce an extensive

Page 238: Cheese Science

mycelium, which may fragment into rods andcocci in older cultures. Most grow well on nor-mal laboratory media at 3O0C, producing or-ange, pink, red, or brown colonies. Their cellwalls contain raes-odiamnopimelic acid, arabi-nose, galactose, and mycolic acids having 32-66carbon atoms and up to 4 double bonds. They areclosely related to Corynebacterium and arefound extensively in soil and dung.

10.5.7 Propionibacterium

These are Gram-positive, nonmotile, pleo-morphic rods, which may be coccoid, bifid, orsometimes branched in shape. They may occursingly, in pairs, in short chains, or in clumpswith "Chinese lettering" arrangements. Coloniesvary in color and can be white, gray, pink, red,yellow, or orange. Despite the fact that these or-ganisms are catalase positive, they are essen-tially anaerobic or microaerophilic bacteria. Thegenus Propionibacterium is divided into twogroups, the classical group and the acnes group.The classical propionibacteria are found mostlyin dairy products, particularly cheese, althoughthey are also found in silage and olive fermenta-tions. The acnes group are found on human skin.The classical group is divided into four species:P. freudenreichii, which is the most commonspecies; P. jensenii; P. thoenii', and P. acidi-propionici. The peptidoglycan of P. freuden-reichii contains meso-diaminopimilic acid,while the L isomer is found in the other threespecies. P. freudenreichii was considered to ex-ist as two subspecies, P. freudenreichii subsp.freudenreichii and P. freudenreichii subsp. sher-manii. Genetic studies have shown that bothsubspecies are identical; the only phenotypicdifference is that P. freudenreichii subsp. freud-enreichii is able to ferment lactose while theother subspecies cannot. P. freudenreichii and P.jensenii produce cream-colored colonies, whileP. thoenii and P. acidipropionici produce red-brown and cream to orange-yellow colonies,respectively. P. thoenii is p-hemolytic. Propi-onic acid bacteria have relatively simple nutri-tional requirements, although they generally re-

quire pantothenic acid, biotin, or thiamine forgrowth, and many of them can use NO3- as thesole source of N.

10.5.8 Pediococcus

These are Gram-positive, catalase-negativecocci that occur in tetrads. Tetrad formation isdue to cell division in two directions in a singleplane and is characteristic of this genus. Cur-rently, eight species are recognized. They arehomo fermentative, producing either oL or L lac-tate from sugars, and most strains can grow inthe presence of 6.5% NaCl. They generally havecomplex nutritional requirements. Most pedio-cocci do not ferment lactose and therefore growpoorly in milk. Those strains that can metabolizelactose may lack a proteinase to hydrolyze themilk protein to the amino acids and peptides re-quired for growth in milk.

Some pediococci can grow at pH 8.5 and 4.2,and some metabolize citrate to acetate and for-mate rather than diacetyl and acetoin. Pedio-cocci are found occasionally as a minor part ofthe NSLAB flora in some hard cheeses, but theirinfluence on the production of cheese flavor isnot clear.

10.5.9 Micrococcus

Micrococci are Gram-positive, catalase-posi-tive, strictly aerobic, nonmotile cocci (0.2-2.0mm in diameter) that occur in pairs, clusters, ortetrads. Division occurs in several planes, result-ing in formation of regular and irregular clusters.Their natural habitat is skin, and currently 17species are recognized. All grow in the presenceof 5% NaCl and many in the presence of 10-15% NaCl. Many species produce yellow, or-ange, or red colonies. Their nutritional require-ments are variable. M. luteus, the type species,produces yellow colonies and grows on glu-tamate as the sole source of C and N in the pres-ence of thiamin and/or biotin. Some species canutilize ammonium phosphate as a N source, butmany species have complex nutritional require-ments. They are commonly found on the surface

Page 239: Cheese Science

of smear-ripened cheese but the species foundare not clear. They dominate the surface ofComte and Blue cheese.

10.5.10 Staphylococcus

These are Gram-positive, catalase-positive,facultatively anaerobic, nonmotile cocci (0.5 to1.5 jLim in diameter), which characteristically di-vide in more than one plane to form clusters.They can also occur in pairs and tetrads. Cur-rently, 19 species of staphylococci are recog-nized and many produce yellow or orange colo-nies. They are facultative anaerobes and growbetter aerobically than anaerobically. Moststrains grow in the presence of 10% NaCl andbetween 1O0C and 4O0C. Acid is produced an-aerobically from several sugars, including glu-cose and lactose. They are fastidious, requiring5-12 amino acids and several B vitamins forgrowth. S. aureus causes mastitis in cows andboils and carbuncles in humans and is consid-ered to be a pathogen. Many strains of S. aureusproduce a heat-stable enterotoxin that causesfood poisoning. Growth to about 106 cfu/g infood is necessary to produce sufficient toxin(0.1-1.0 mg/kg) to cause food intoxication. En-terotoxins are difficult to measure. Coagulase isaccepted as the indicator of pathogenicity in sta-phylococci, and S. aureus, S. intermedius, and S.hyicus produce it. S. intermedius has been foundin the nasal passages of horses, dogs, mink, andfoxes, and S. hyicus has been found on the skinof pigs and less frequently on the skin and in themilk of cows. Major habitats of staphylococciinclude the nasal membranes, skin, and the gas-trointestinal and genital tracts of warm-bloodedanimals.

Traditionally, Micrococcus and Staphylococ-cus have been placed in the family Micrococ-caceae, indicating that they are closely related.However, phylogenetic studies show that theyare quite distant from each other; Staphylococ-cus spp. belong to the Clostridium branch of theGram-positive bacteria, and Micrococcus be-long to the actinomycete branch. This is also re-flected in the guanine plus cytosine levels, with

staphylococci containing 30-39% and micro-cocci 63-73%. Phylogenetically, Micrococcusis closely related to Arthrobacter and may be adegenerate form of this genus.

It is relatively easy to distinguish micrococcifrom staplylococci. The simplest way is to checkfor acid production from glucose under aerobicand anaerobic conditions. Staphylococci pro-duce acid from glucose aerobically and an-aerobically, whereas micrococci either do notproduce acid or produce it only aerobically. Inaddition, micrococci are resistant to lysostaphin,a cell-wall degrading enzyme, and are sensitiveto erythromycin (0.04 mg/ml) whereas staphylo-cocci have the opposite reactions.

10.5.11 Yeasts and Molds

Yeasts and molds are generally not nutrition-ally demanding and are larger and grow moreslowly than bacteria. Therefore, they do notcompete with bacteria in environments in whichbacteria grow, for example, at pH values around7. However, they grow quite well at pH values of2 to 4, where bacteria either do not grow or growonly very poorly. The low pH of freshly madecheese is therefore partially selective for theirgrowth. Yeast and molds are eukaryotes—thatis, they contain a clearly identifiable nucleus—and most of them also contain chitin, a (3-1,4polymer of N-acetylglucosamine, which is re-sponsible for their rigid structure.

Colonies of yeast generally have a soft consis-tency, while those of molds are hard and largeand often exhibit several different colors. In ad-dition, they look quite different under the micro-scope: yeast are generally round and pear-shaped whereas molds show a mycelial networkof filamentous hyphae. Some fungi are dimor-phic, producing hyphae under one set of circum-stances and yeastlike cells under another. Thehuman pathogen Candida albicans is the bestexample of dimorphism; it grows like a yeast inbody fluids but develops hyphae to invade tis-sue.

Both yeast and molds are classified as fungiand are divided into three major groups: Asco-

Page 240: Cheese Science

mycetes, Zygomycetes, and Deuteromycetes.Classification of fungi is very complex, and onlya few important criteria are mentioned here.These include whether cells in the mycelium areseptate (possess cross-walls) or nonseptate (lackcross-walls), the type of spores and how they areproduced, and whether reproduction is sexual orasexual. Ascomycetes and Zygomycetes areseptate, and Deuteromycetes are nonseptate.The spores produced by Ascomycetes areformed in a sac called the ascus (for this reasonthese spores are called ascospores) and are in-volved in sexual reproduction. The spores pro-duced by Deuteromycetes and Zygomycetes arecalled conidia (see below) and sporangiophores,respectively, and are not involved in sexual re-production. Yeast generally multiply by bud-ding: a protuberance is formed on the wall of thecell and eventually breaks off to form a new cell,in which further budding occurs. Sometimes,several buds are produced by the same cell andremain attached to it. Some yeast (Schizo-saccharomyces spp.) multiply by binary fission.Sexual reproduction is given the generic nameteleomorph, and asexual reproduction is calledanamorph. The same fungus has often beengiven different names, depending on the type ofreproduction. Some examples of this are shownin Table 10-4. Taxonomically, the teleomorphicname is normally used, but there are exceptions.For instance, the anamorphic name Geotrichumcandidum is more commonly used than the tel-eomorphic name Galactomyces candidum.

The species of yeast found on the surface ofdifferent cheeses show considerable diversity(Table 10-4). The dominant species in allcheeses, except Romadour from one plant, is D.hansenii. Kluvyeromyces spp. are also dominantin the French cheeses (Roquefort, Camembert,and St. Nectaire) and in the Spanish cheese(Cabrales) but appears to be absent from theGerman and Austrian cheeses (Weinkase, Lim-burg, Romadour, and Tilsit). Saccharomycescerevisiae and Pichia spp. are also important inCamembert and Cabrales cheese. All theseyeasts are members of the Ascomycetes group.S. cerevisiae is also involved in wine, beer, and

bread making. Very few yeast are capable of fer-menting lactose, but Kluvyeromyces lactis is anexception. This may be one reason for its domi-nance in some surface-ripened cheese. Whethervariation occurs within the same cheese has notbeen studied to any great extent. The evidence inTable 10-4 suggests that it does occur, at least inLimburg. Both cheeses examined contained D.hansenii and Geotrichum candidum in signifi-cant numbers, but in addition Torulaspora del-brueckii was found in one cheese and Yarrowialipolytica in the other.

The most important molds in cheese are P.camemberti, P. roqueforti, and G. candidum', allare members of the Deuteromyces group. P.camemberti is responsible for the white growthon the surface of Camembert and Brie, and P.roqueforti is responsible for the blue veins inRoquefort and other Blue cheeses. It is generallythought that G. candidum is present on the sur-face of most mold- and bacterial-ripened cheese.The results in Table 10-4 suggest that it is foundonly in Weinkase, Romadour, Limburg, andTilsit. Scanning electron micrographs of Cam-embert cheese show the presence of G. candi-dum, and it is likely that the reason it was notreported to be present in the other cheeses inTable 10 1 is that the various workers involvedconsidered it to be a mold rather than a yeast.

Microscopic observation is very important inclassifying fungi because their various struc-tures can be seen clearly. Both P. camembertiand P. roqueforti reproduce asexually fromconidia (spores) extruded from a flask-shapedcell called aphialide, which is borne on the con-idiophore or spore-bearing hyphae (Figure10-9). The multiplication of G. candidum isquite different. The hyphae grow to a consider-able extent, then stop, and septa are formedtransversely, separating the hyphae into shortcompartments that eventually fragment intoseparate conidia, which start the reproductiveprocess again. Many molds produce toxins thatare carcinogenic (e.g., the aflatoxins producedby Aspergillus JIavus). However, the strains in-volved in cheese do not produce toxins. Thephysiological conditions for the production of

Page 241: Cheese Science

Table 10-4 Species of Yeast Found in Different Cheeses

Limburg1Romadour1Weinkase1

St. Nectaire6CamemberPCabrales4Roquefort3Tilsit2Factory CFactory AFactory AFactory DFactory CFactory BFactory AAnamorphTeleomorph

86

101

6

6

529

3

8

4

3015

7

166

21

131

1

11

16

356

15

75

5

2

79

5

7

12

85

2

64

17

19

55

21

24

10

3

87

3

69

6

22

22

95

1

86

4

3

Candida famataGeotrichum capitatumGeotrichum candidumCandida sphaericaCandida kefyrCandida scottiiCandida pelliculosaCandida lambica

Candida valida

Candida robusta

Candida colliculosa

Candida lipolytica

Candida catenulataCandida intermediaCandida mogiiCandida rugosaCandida saitoanaCandida versatilisCryptococcus flavusDebaryomyces hanseniiDipodascus capitatusGalactomyces geotrichumKluyveromyces lactisKluyveromyces marxianusLeucosporidium scottiiPichia anomalaPichia fermentansPichia kluyveriPichia membranaefaciensRhodotorula spp.Saccharomyces cerevisiaeSaccharomyces unisporusTorulaspora delbrueckiiTrichosporon beigeliiYarrowia lipolyticaZygosaccharomyces rouxii

Note: Results from reference 1 are as a percentage of the surface yeast microflora; other results are as a percentage of the number of strains isolated and/or identified. Factories A and C produced more than onevariety of cheese.

1 N. Valdes-Stauber, S. Scherer, & H. Seller, Identification of yeasts and coryneform bacteria from the surface microflora of brick cheeses, International Journal of Food Microbiology, 34 (1 997), 1 1 5-1 29.2 E. Eliskases-Lechner & W. Ginzinger, The yeast flora of surface-ripened cheese, Milchwissenschaft, 50 (1995), 458-462.3 JJ. Devoyod & D. Sponem, La flore microbiennee du fromage de Roquefort. 6: Les levures, Lait, 50 (1970), 524-543.4 M. Nunez, M. Medina, P. Gaya, & C. Dias-Amado, Les levures et les moissures dans Ie fromage bleu de Cabrales, Lait, 61 (1981), 62-79.5 C. Baroiller & J. Schmidt, Contribution a !'etude de I'origine des levures du fromage Camembert, Lait, 70(1980), 67-84.6 J. Vergeade, J. Guiraud, JP. Larpent, & P. Galzy, Etude de Ia flore de levure du Saint-Nectaire, Lait, 56(1976), 275-285.

Page 242: Cheese Science

toxins by microorganisms are generally muchnarrower than those for growth.

G. candidum has characteristics of both a yeastand a mold, and in the past it was often called ayeastlike fungus. It was initially called Oidiumlactis, later called Oospora lactis, and then givenits present name. It is commonly known as thedairy mold. Its natural habitat is soil, where it isinvolved in the decay of organic matter.

Most fungi grow quite well at the pH ofcheese, and most of those found in cheese arealso quite tolerant to salt. For example, thegrowth of P. camemberti is largely unaffectedby 10% NaCl (Table 10-1), and some strains ofP. roqueforti can tolerate 20% NaCl. An excep-tion is G. candidum, which is quite sensitive tosalt. A slight reduction in its growth occurs inthe presence of 1% NaCl, and it is completely

Figure 10-9 Major features of Penicillium roqueforti, Penicillium camemberti, and Geotrichum candidum.

Geotrichum candidum

Conidia

Penicillium camemberti

Conidiophore

Penicillium roqueforti

Conidia

Phialides

Page 243: Cheese Science

inhibited at about 6%. Therefore, too much brin-ing will prevent its growth on the cheese surface.Perhaps its intolerance to salt explains why it issometimes deliberately added in the manufac-ture of some surface-ripened cheeses, the hopebeing that some cells will grow.

Generally, yeast are facultative anaerobeswhereas molds are considered to be obligate aer-obes. However, P. roqueforti can grow in thepresence of limited levels of O2, as demonstratedby its growth throughout the mass of Bluecheese. Yeast and molds are generally heat sen-sitive and are killed by pasteurization.

Occasionally, yeast have been incriminated inthe spoilage of cheese, either through the pro-duction of gas (CO2) or the development of off-flavors. Unripened cheeses (e.g., Cottage cheeseand Quarg), which can contain a high level oflactose, are particularly prone to spoilage byyeast. Reviews of the role of yeast in dairy prod-ucts include those of Fleet (1990) and Jakobsenand Narvhus( 1996).

10.6 EXAMPLES OF MICROBIALGROWTH IN CHEESE

Cheese is a very complex microbiological eco-system in which molds, yeasts, and bacteria coex-ist and multiply. Some examples are given below.

10.6.1 Cheddar

This is a hard, dry-salted cheese made with amesophilic starter, which grows very rapidly inthe cheese from an initial level of about 107/g to108 or 109/g at salting (~ 5.5 hr after inoculation).Mesophilic starters generally die out relativelyrapidly during the first few weeks of ripening(Figure 10-5), but the rate is strain dependentand probably reflects the ability of the strain towithstand the cooking temperature of the cheeseand its ability to lyse. Phage may also be in-volved in reducing the number of cells.

Normally, the fermentation of lactose by thestarter LAB in cheese is rapid and is completewithin 1 day. However, in dry-salted cheeses,like Cheddar, a relatively large amount of lac-

tose (~ lOg/kg of cheese) is present in the curdafter overnight pressing. This is due to inhibitionof the metabolism of the starter cultures by thesalt and the relatively low pH. The S/M in Ched-dar cheese determines the subsequent rate of lac-tose fermentation by the starter. High S/M levelsreduce the rate and low levels increase it (Figure10-10). For example, at an S/M of 4.1 the fer-mentation is virtually complete in 7 or 8 days,while at a 6% level it takes more than 50 days.These values are only indicative and vary de-pending on the sensitivity of the particular cul-ture to salt. Lc. lactis subsp. cremoris is muchmore sensitive to salt than Lc. lactis subsp. lactisstrains. The former cannot grow in the presenceof 4% salt whereas the latter can. Salt can alsouncouple acid production from growth.

NSLAB, particularly facultative heterofer-menters like Lb. paracasei and Lb. casei, arefacultative anaerobes and are also acid and salttolerant. They can grow at pH 4.5 and in thepresence of more than 6% salt. In fact, many ofthem can grow in the presence of 8 or 10% salt,so salt is of little consequence in preventing theirgrowth in cheese. They grow relatively rapidlyin Cheddar cheese during ripening from low ini-tial numbers (~102/g) to final numbers of 107 to108/g. Generation times of 8.5 days have beenreported for Cheddar cheese ripened at 60C.Such high numbers of lactobacilli must havesome role in flavor development in the cheese,but that role is unclear. They do transform L-lac-tate to o-lactate, eventually forming a racemicmixture, but this transformation has no effect onthe flavor of the cheese. However, Ca o-lactateis very insoluble and can form small crystalsthroughout the cheese. An example of thegrowth of lactobacilli and the racemization ofthe L-lactate in Cheddar cheese during ripeningis shown in Figure 10-11.

10.6.2 Emmental and Comte

Both of these cheeses are made with thermo-philic cultures, and little acid is produced in thevat during their manufacture. During the initialhours in the press, most of the lactic acid is pro-

Page 244: Cheese Science

Time, days

Figure 10-10 Effect of salt-in-moisture (S/M) on lactose metabolism in Cheddar cheese made with Lc. lactisssp. cremoris C13 and 266 and ripened at 12 ° C.

Lac

tose

, g/lO

Og

Lactate, lactose, g/lO

OgN

SLA

B, l

og c

fti/m

l or

g

Ripening time, days

Figure 10-11 Relationship between nonstarter lactic acid bacteria (NSLAB; •), metabolism of lactose (A), andproduction of L-lactate (•) and D-lactate (Q) in Cheddar cheese (4.1% salt-in-moisture) during ripening at 120C.

NSLAB

Llactate

D lactate

Lactose

Page 245: Cheese Science

duced by Sc. thermophilus, but, as the tempera-ture and pH decrease, Lb. helveticus begins togrow, reaching maximum numbers 12-20 hr af-ter the addition of starter. Counts of both Sc.thermophilus and Lb. helveticus in Comtecheese, and presumably Emmental also, arehigher at the periphery than at the center (Fig-ure 10-12). In both cheeses, growth of thestarter is limited by the high cooking tempera-ture (52-540C), but growth begins again assoon as the temperature decreases. The tem-perature falls more rapidly at the periphery ofthe cheese than in the center, and hence greaterbacterial growth (and acid production) occurs atthe periphery.

Sc. thermophilus metabolizes only the glu-cose moiety of lactose and excretes galactose,which, along with residual lactose, is metabo-lized by Lb. helveticus. All the lactose is fer-mented during the first 10 or 12 hr of manufac-ture. The L isomer of lactate is produced by bothSc. thermophilus and Lb. helveticus, while the D

isomer is produced only by the latter organism.After several weeks ripening at a low tempera-ture (4-140C), the cheese is placed in a "warmroom" at 18-240C, during which the propionicacid bacteria grow and transform the lactate topropionate, acetate, and CO2, which is respon-sible for eye formation. The eyes in Emmentalcheese are much larger than in Comte cheesebecause Emmental is ripened at 220C and Comteat 180C. In traditionally made Comte andEmmental, the propionic acid bacteria are natu-ral contaminants of the raw milk, but, in the in-dustrial production of Emmental, they are nor-mally added deliberately to the milk to giveinitial counts of about 103 to lOVml.

The pathway of lactate fermentation by propi-onic acid bacteria is complicated (Figure 10-13)and involves two separate cycles, one in whichpropionate is produced and the other in whichacetate is produced. The lactate is first oxidizedto pyruvate, two moles of which are converted topropionate and one mole to acetate. ATP is only

Cou

nt, l

og c

fti/g

Time,h

Figure 10-12 Growth of Streptococcus thermophilus (A, A) and Lactobacillus helveticus (Q, •) at the center(open symbols) and periphery (closed symbols) of Gruyere cheese during manufacture.

S. thermophilus, centre

S. thermophilus, periphery

L. helveticus, centre

L. helveticus, periphery

Page 246: Cheese Science

generated in the production of acetate. The over-all stoichiometry is

3 Lactate -> 2 Propionate + 1 Acetate + 1 CO2 + 1 H2O

Transcarboxylase is the key enzyme in the pro-duction of propionate and requires biotin for ac-tivity. Generally, propionic acid bacteria areable to metabolize both isomers of lactate, but,in a mixture of the two, they preferentially me-tabolize the L rather than the D isomer. The com-plex interrelationships between lactose and lac-tate utilization and production of propionate and

acetate by the propionic acid bacteria in Em-mental cheese are shown in Figure 10-14.

Comte cheese is also covered by an orangesmear, called the morge, composed mainly ofcorynebacteria, micrococci, and yeast. About1010 microorganisms/cm2 are present on the sur-face of ripened cheese, and it has been calculatedthat the total number of bacteria in the smear ofComte cheese is equal to the total number in thecheese mass, but, except for Brachybacterium,the species involved do not appear to have beenidentified.

Figure 10-13 Cycles involved in the propionic acid fermentation (LDH, lactose dehydrogenase; PoIy-Pn,polyphosphate; PPi, pyrophosphate). For reasons of clarity, only the pyrophosphate-dependent conversion offructose-6-P to fructose-1,6-diP is shown, and the generation of ATP by the electron transfer system is omitted.All the reactions are directed toward propionate production, even though the reactions are reversible.

Propionate

Propionyl CoA

Succinyl CoA

Succinate

Fumarate

Methyl malonyl CoA

Acetate

Lactate Pyruvate

CitrateATP

ADPMaI ate

acetyl-PAcetyl CoA. Oxaloacetate

P-enolpyruvate

3-P-gly cerate

Glyceraldehyde-3-P

Fructose-1,6-diP

Fructose-6-P

Glucose-6-P

Glucose

6-P-gluconate

Pentose-P

ot-Ketoglutarate

Page 247: Cheese Science

10.6.3 Camembert

This is a mold-ripened cheese with a rela-tively high moisture content (~ 50%) and shortripening time. Spores of P. camemberti are ei-ther added to the milk with the mesophilic starteror are inoculated directly on the surface of thecheese after salting. This organism is an obligateaerobe and grows only on the cheese surface.Adventitious acid-tolerant yeast and G. can-didum also grow on the surface of Camembertcheese within a few days of manufacture, whileP. camemberti is generally not visible on the sur-face until about the 6th day of ripening. Scan-ning electron micrographs of the development ofthese on the surface of the cheese have been pub-lished (Rousseau, 1984). The yeast and moldsoxidize lactate to CO2 and H2O, causing a sig-nificant increase in the pH of the cheese, particu-larly on the surface. Significant proteolysis alsooccurs on the surface, and production of NH3

through deamination of the resulting amino ac-

ids may also be involved in increasing the pH ofthe cheese. Some microbiological changes thatoccur on the surface and in the interior of Cam-embert cheese are shown in Figure 10-15. Num-bers of Lactococcus reach about 109/g at the be-ginning of ripening and remain at this level bothin the interior and on the surface throughout rip-ening. In contrast, growth of yeast and Micro-coccus is much greater on the surface than in theinterior of the cheese. S. cerevisiae is an impor-tant yeast in Camembert cheese (Table 10-4),and yeast counts on the surface can reach 108/gwithin the first 2 weeks of ripening, after whichthey began to decrease very slowly.

10.6.4 Cabrales

Cabrales is a Spanish Blue cheese made fromraw milk without the deliberate addition of start-ers or molds. Adventitious LAB in the milk areresponsible for acid production during manufac-ture and ripening. The coagulum is cut 2 hr after

hours daysTime from start of manufacture

Figure 10-14 Relationship between lactose and lactate metabolism, growth of propionibacteria, and productionof propionate and acetate in Swiss-type cheese.

Log

prop

ioni

bact

erla

/g c

hees

e

Con

cent

ratio

n (g

/100

g ch

eese

)

Propionibacteria

Lactose

L-lactate^

Galactose

D-lactate

Propionate

Acetate

Page 248: Cheese Science

Days Days

Figure 10-15 Growth of lactococci, micrococci, and yeast and changes in the pH on the surface (Q) and in the interior (•) of Camembert cheese.

Days Days

Page 249: Cheese Science

addition of rennet and is then scooped intomolds, which are held at 16-180C for 48 hr toallow whey drainage to occur. Then the curd isremoved from the mold, covered with coarsesalt, held for a further 48 hr at 16-180C, ripenedat 10-120C for 10-15 days, and transferred tocaves for further ripening at 9-1O0C and at 90-95% relative humidity.

The microbiological changes that occur in thischeese during ripening are shown in Figure 1O-16. In each graph, the first and second points re-fer to the counts in the milk and the curd at 1-2days, respectively. Growth of lactococci is rela-tively rapid during the first few days, after whichtheir number decreases (more rapidly on the sur-face than in the interior of the cheese). The num-ber of mesophilic lactobacilli remains more orless constant at 106/g on the surface but increasesto 108/g in the interior, after which it decreases.Micrococcus spp. show the opposite trend, beingmore abundant on the surface than in the inte-rior, and yeast show significant growth on thesurface of the cheese early in ripening, afterwhich they decrease. Coliforms grow duringcheesemaking, but their number decreases rap-idly over the next 2 weeks. This is probably dueto the very rapid decrease in pH, which reachesabout 5.0 in 48 hr due to the growth of thelactococci, after which it increases to 6.5 in theinterior and to 7 on the surface due to metabo-lism of lactate to CO2 by the yeast and molds.

10.6.5 Tilsit

Tilsit, a smear-ripened cheese made with amesophilic starter, is particularly popular in Ger-many, Austria, and Switzerland. Sometimescommercial cultures containing some or all of thefollowing microorganisms are used to smear thecheese: B. linens, G. candidum, Candida utilis, D.hansenii, and K. lactis. In Austria, only B. linensis deliberately inoculated onto the surface of thecheese. The yeast are natural contaminants of thecheese and arise from the milk, air, equipment,brine, and smear water. In Germany, the youngcheese is smeared with so-called old smear fromwell-ripened cheese. This can be problematic,because such smear can be contaminated with

pathogens, particularly Listeria monocytogenes(see Chapter 20), which serve as an inoculum forthe new cheese. The function of the yeast is tometabolize the lactate on the surface, which re-sults in an increase in pH to a point where thebacteria, particularly B. linens and corynebacte-ria, can grow. The microbiological changes inTilsit cheese are shown in Figure 10-17.

The pH increases steadily from about 5.5 to7.5 during the first 2 weeks of ripening, afterwhich it remains more or less constant. Simulta-neously, the number of yeast and salt-tolerantbacteria (coryne forms) also increase, with thebacteria increasing more rapidly than the yeast.In Figure 10-17, bacterial numbers are plottedper cm2; counts/g of surface would be 100 timesgreater. Despite the fact that deliberate additionof B. linens is used to inoculate the surface, thereis a high diversity in the different species of bac-teria and yeast on the surface of Tilsit cheese(Tables 10-3 and 10-4). D. hansenii and Y.lipolytica are the dominant yeasts, and A. globi-formis, A. citreus, B. ammoniagenes, and B. lin-ens are the dominant bacteria.

10.7 MICROBIAL SPOILAGE OFCHEESE

The most common microbial defects ofcheese are early and late gas, both of which arerelatively uncommon in cheese today, mainlybecause of better hygiene and better quality con-trol in cheese plants.

Early gas generally occurs within 1 or 2 daysafter manufacture. It is characterized by the ap-pearance of many small holes and is caused bycoliform bacteria and/or yeast. The gas producedby coliform is mainly H2, which is produced byformic hydrogenylase activity from formate, aproduct of lactose metabolism. It is more prob-lematic in soft and semi-soft cheese than in hardcheese because of higher aw in the former. Aneffective way of controlling early gas is to addKNO3 OrNaNO3 at low levels (0.2%) to the milk.NO3 does not prevent the growth of coliform butacts as an alternative electron acceptor, allowingcomplete oxidation of lactose to CO2 and H2O

Page 250: Cheese Science

Lac

toco

cci,

log

cfu/

g

Lac

toba

cilli

, log

cfii

/g

Mic

roco

cci,

log

cfli/

g

Yea

st &

mou

ld, l

og c

fu/g

Col

iform

, log

cft

i/g

PH

Days Days

Days Days

Days Days

Figure 10-16 Growth of different organisms on and changes in the pH of Cabrales cheese during ripening.Surface (Q), interior (•).

Page 251: Cheese Science

rather than fermentation to formate, thus effec-tively reducing the production of H2 from for-mate. Early gas production by yeast is due to theproduction of CO2 from lactose or lactate.

Late gas formation, or late blowing, does notoccur until late in ripening. It is due to fermenta-tion of lactate to butyrate and the production ofcopious amounts of H2 by Clostridium tyro-butyricum and Cl butyricum. Consequently,large holes are generally produced. The butyrateis responsible for off-flavor development in thecheese. Late gas can be particularly prevalent inSwiss-type cheese, where clostridia can growwith the propionic acid bacteria during the "hotroom" ripening period. Silage is a potent sourceof these bacteria, and for this reason it is forbid-den in Switzerland to feed it to cows whose milkis intended for cheesemaking. In addition, manythermophilic cultures are thought to stimulatethe growth of clostridia through the productionof peptides and amino acids. Late gas productioncan be controlled by bactofugation of the milk(see Chapter 4), but this often results in inferiorquality cheese.

The bacteriocin nisin, which is produced bysome strains of Lc. lactis subsp. lactis, is effec-tive in controlling the growth of clostridia and isused for this purpose in processed cheese. How-ever, it is not suitable for use in natural cheese,

because many starters are sensitive to it. Increas-ing the level of salt, lowering the pH of thecheese rapidly through the use of an activestarter, adding NO3-, or adding lysozyme canalso be effective in preventing late gas produc-tion. Lysozyme, which is found in milk, saliva,tears, and other body fluids, hydrolyzes the cellwalls of sensitive bacteria (e.g., Cl tyrobutyri-cum), causing them to lyse. Commonly used inItaly, it is added to the milk with the starter at alevel of 25 mg/L. It is generally considered tohave no effect on the growth of starters, althoughsome strains in Italian natural whey cultures areinhibited by it.

NO3- is an effective inhibitor of clostridiathough not coliform. It was initially thoughtthat NO2- combined with some of the enzymesinvolved in respiration, particularly those con-taining an -SH group. However, this cannot betrue if NO3" is effective against an obligate an-aerobe such as Cl tyrobutyricum. It is possiblethat NO3- interacts with the Fe in the ferredoxinprotein involved in oxidation-reduction reac-tions in clostridia, but this hypothesis has notbeen proven.

Other microorganisms have occasionallybeen implicated as spoilage organisms. Citrate-metabolizing lactobacilli have been incrimi-nated as the cause of open texture in Cheddar

Weeks

Figure 10-17 Growth of yeast and salt-tolerant bacteria (STB) (on plate count agar containing 8% salt) andchanges in the pH on the surface of Tilsit cheese during ripening.

Log

cfii

/cm

2

Hd

Page 252: Cheese Science

cheese due to the production of CO2 from citrate.The optimum pH for uptake of citrate rangesfrom 4 to 5, and significant metabolism of citrateoccurs in the absence of an energy source at pH5.2, the pH of many semi-hard and hard cheeses.Ec. malodoratus, which, as its name implies,causes the production of bad flavors, has beenfound in Gouda cheese. The surface of cheese,especially when it is moist (e.g., unwrapped softor semi-soft cheese), is an ideal environment forthe growth of molds and yeast. These cause littledamage to the cheese but are unsightly. They canbe washed off the cheese surface with a dilutebrine solution.

Yeasts and molds are occasionally incrimi-nated as spoilage organisms in cheese. Sorbicacid is allowed in Italy as a preservative for hard,fresh, and processed cheese. Sorbate-resistantmolds, Paecilomyces variotti, and the yeast D.hansenii from Crescenza and Provolone cheeseare able to grow in the presence of 3 mg of sorbic

REFERENCES

Fleet, G.H. (1990). Yeasts in dairy products. Journal of Ap-plied Bacteriology, 68, 199-211.

Hocking, A.D., & Faedo, M. (1992). Furgi causing threadmould spoilage of vacuum packaged Cheddar cheese dur-ing maturation. InternationalJournal of Food Microbiol-ogy, 16, 123-130.

Jakobsen, H., & Narvhus, J. (1996). Yeasts and their pos-sible beneficial and negative effects on the quality ofdairy products. International Dairy Journal, 6, 755-768.

Lund, F., Filtenberg, O., & Frisvad, J.C. (1995). Associatedmycoflora of cheese. Food Microbiology, 12, 173-180.

Rousseau, M. (1984). Study of the surface flora of traditionalCamembert cheese by scanning electron microscopy.Milchwissenschaft, 39, 129-135.

SUGGESTED READINGS

Deacon, J.W. (1997). Modern mycology (3d ed.). Oxford:Blackwell Science.

Eck, A. (1986). Cheesemaking, science and technology(English ed.). Paris: Lavoisier.

acid/g and are also able to transform sorbic acidto £rans-l,3-pentadiene, which has a taste andodor like kerosene (Sensidini, Rondinini, Peres-sini, Maifreni, & Bartolomeazzi, 1994).

Cladosporium cladosporioides, Penicilliumcommune, C. herbarum, P. glabrum, and aPhoma species are responsible for the "threadmold" defect of Cheddar cheese (Hocking &Faedo, 1992). This defect occurs as a black, darkbrown, or dark green spot or thread in the folds,creases, and gusset ends of the plastic bags usedto wrap Cheddar cheese during ripening. It canoccur on the cheese surface but is more often as-sociated with free whey drawn from the freshcheese block during vacuum packaging. Thesemolds are obviously able to grow in the presenceof low levels of O2.

Growth of P. commune, which is closely re-lated to P. camemberti, can result in discolora-tion of cheese surfaces and the production of off-flavors (Lund, Filtenberg, & Frisvad, 1995).

Schubert, K., Ludwig, W., Springer, N., Kroppenstedt, R.M.,Accolas, J.P., & Fiedler, F. (1996). Two coryneform bac-teria isolated from the surface of French Gruyere or Beau-fort cheeses are new species of the genus Brachybac-terium: Brachybacterium alimentarium sp. nov andBrachybacterium tyrofermentair sp. nov. InternationalJournal of Systematic Bacteriology, 46, 81-87.

Sensidini, A., Rondinini, G., Peressini, D., Maifreni, M., &Bartolomeazzi, R. (1994). Presence of an off-flavour as-sociated with the use of sorbates in cheese and margarine.Italian Journal of Food Science, 6, 237-242.

Kurtzman, C.P., & Fell, J.W. (1998). The yeasts: A taxo-nomic study (4th ed.). Amsterdam: Elsevier Science.

Pill, J.I., & Hocking, A.D. (1997). Fungi and food spoilage(2d ed.). London: Blackie Academic and Professional.

Page 253: Cheese Science

11.1 INTRODUCTION

As discussed in Chapter 10, the original ob-jective of cheese production was to conserve theprincipal constituents of milk—the lipids andcaseins. However, although well-made cheese isa hostile environment for microbial growth andpossesses several preservative hurdles, thesehurdles are not sufficient to prevent the growthof certain microorganisms (see Chapter 10) andthe activity of those enzymes from varioussources that may be present. These microorgan-isms and enzymes catalyze a complex series ofbiochemical reactions, which, if unbalanced,cause off-flavors and textural defects but, ifproperly controlled and balanced, lead to the de-sirable and characteristic flavors and textures ofthe numerous cheese varieties. Although thebiochemistry of cheese ripening is not yet fullycharacterized, a considerable body of informa-tion is now available. The objective of this chap-ter is to describe the principal biochemical reac-tions that occur in cheese during ripening.

Cheese curd is a relatively simple mixture ofcasein (and very little whey proteins except incheese made from milk concentrated by ultrafil-tration), lipids, a little lactose (~ 1% at pressing),lactic acid (« 1%), citric acid (~ 0.2%), NaCl(0.7 to ~ 6%), and water. Not surprisingly, theprimary features of ripening involve the twoprincipal organic constituents, proteins and lip-ids. However, the metabolism of lactose and cit-

rate, although they are present at low concentra-tions, is important in all varieties and critical insome. Most of the primary reactions are wellcharacterized. Many of the products of the pri-mary reactions undergo further modifications,which are not fully understood but which areprobably responsible for the characteristic flavorof cheese.

11.2 RIPENING AGENTS IN CHEESE

The ripening of cheese is catalyzed by themetabolic activity of living organisms and en-zymes from these organisms or from othersources:

• Coagulant. The coagulant usually contrib-utes chymosin or other suitable proteinase(see Chapter 6), but rennet paste used insome Italian varieties contributes both pro-teinase and lipase. In high-cooked cheeses(e.g., Emmental and Parmesan), enzymes inthe coagulant are extensively or completelydenatured.

• Milk. As discussed in Chapter 3, milk con-tains about 60 indigenous enzymes, at leastsome of which are significant in cheese rip-ening, including proteinases, especiallyplasmin; lipase; acid phosphatase; and xan-thine oxidase. Most of the indigenous en-zymes are quite heat stable and fully or par-tially survive pasteurization. Furthermore,

Biochemistry of Cheese Ripening

CHAPTER 11

Page 254: Cheese Science

they are either associated with the caseinmicelles or present in the fat globule mem-brane and are therefore incorporated into thecheese curd. Enzymes present in the serumphase are largely lost in the whey and thusare of little importance in cheese ripening.

• Starter culture. Starter cultures were dis-cussed in Chapter 5. Live starter cells prob-ably contribute little to cheese ripening butthey possess a diversity of enzymes (seeChapter 10 and Section 11.7.4), which arelocated mainly intracellularly and are re-leased upon cell death and lysis. Thesestarter enzymes are major contributors toripening.

• Secondary micro/lorn. Many cheese variet-ies contain a secondary (nonstarter) micro-flora (see Chapter 10), the function ofwhich is not acid production but rathersome specific secondary function. In manycases, flavor development is dominated bythe metabolic activity of the secondary cul-ture. The microorganisms involved includepropionic acid bacteria, coryneform bacte-ria, yeasts, and molds. In addition to these,the growth of which is characteristic andencouraged, cheeses contain adventitiousnonstarter lactic acid bacteria (NSLAB)that originate from the milk or the environ-ment. Owing to the selective nature of theinterior of cheese (see Chapter 10), this ad-ventitious microflora is composed mainlyof mesophilic lactobacilli and, to a lesserextent, pediococci.

• Exogenous enzymes. With the objective ofaccelerating ripening, cheese makers haveexperimented with adding exogenous en-zymes, usually proteinases and perhapspeptidases and Upases, to the cheese curd(see Chapter 15).

11.3 CONTRIBUTION OF INDIVIDUALAGENTS TO RIPENING

The role of the individual ripening agents incheese has been studied using model cheesesystems in which the action of one or more of

the ripening agents is eliminated (see Fox, Law,McSweeney, & Wallace, 1993). Pioneeringstudies on this subject used milk obtained fromselected cows by aseptic milking techniques.However, in our experience, bulk herd milkwith a total bacteria count (TBC) less than 103

cfu/ml can be obtained from a healthy commer-cial herd using good but not special milkingpractices.

Heat treatment of aseptically drawn milk isnecessary to further reduce bacterial counts.Batch pasteurization (680C x 5 min or 630C x 30min), HTST pasteurization (« 72 or 770C x 15s), or ultra-high temperature treatment have beenused. A heat treatment of 830C x 15 s or 720C x58 s is necessary to ensure an 8-log reduction inthe bacterial population, which is deemed neces-sary to produce cheese with a nonstarter countbelow 10 cfu/kg cheese for milk with an initialTBC of 103 cfu/ml. HTST pasteurization (720Cx 15 s) is sufficient for milk with an initial countof 10 cfu/ml.

To avoid contamination from the environ-ment, cheese is manufactured under aseptic con-ditions, which can be achieved using enclosedcheese vats, a sterile room with a filtered air sup-ply, or a laminar air-flow unit (the latter is prob-ably the simplest of these techniques). Usingaseptic conditions, it is relatively easy to pro-duce curd free of NSLAB, but in our experienceNSLAB always grow in such cheese, sometimesonly after a long lag period (e.g., 100 days). Acocktail of antibiotics (penicillin, streptomycin,and nisin) extends the lag period and reduces thefinal number of NSLAB. The growth of NSLABis strongly retarded, essentially prevented, byripening at about 10C, although the whole ripen-ing process is also retarded.

The acidifying role of starter can be simulatedclosely using an acidogen, usually gluconicacid-8-lactone (GDL), although the rate of acidi-fication is faster than occurs in biologicallyacidified cheese. Incremental additions of lacticacid and GDL give the best results, but precisecontrol of pH is difficult.

To study the role of the coagulant in cheeseripening, it is necessary to inactivate the rennet

Page 255: Cheese Science

Gluconic arid-6-lactone Gluconic acid

after coagulation, for which four techniqueshave been developed. One approach involvesseparating the first and second stages of rennetaction. Milk depleted OfCa2+ and Mg2+ by treat-ment with an ion-exchange resin is renneted (butdoes not coagulate), heated (720C x 20 s) to in-activate the rennet, and cooled to below 150C;CaCl2 is added. The renneted milk is then heateddielectrically to induce coagulation. Porcinepepsin may be inactivated after coagulation ofthe milk by adjusting the pH of the curd-wheymixture to 7.0. Piglet gastric proteinase, whichhydrolyzes bovine K-casein but has little effecton OC8I- or p-casein, has been used to prepare ren-net-free curd in small-scale cheesemaking trials.Chymosin and all commercial rennet substitutesare aspartyl proteinases and are inhibited bypepstatin. The effectiveness of adding pepstatinto Cheddar cheese curd at salting has been dem-onstrated on a small scale.

6-Aminohexanoic acid (AHA), a noncom-petitive inhibitor of plasmin, has been used tostudy the significance of plasmin in cheese rip-ening. It is necessary to use a high concentrationof AHA, which affects curd syneresis and themoisture content of the cheese. Also, since AHAcontains N, the background level of soluble N isincreased greatly. Plasmin is inhibited by severalproteins, including soybean trypsin inhibitor,which may be suitable for the inhibition of plas-min in cheese. It is also specifically inhibited bydichloroisocoumarin, but neither it nor the in-hibitory proteins have been investigated incheesemaking. Plasmin activity is increased byhigh cooking temperatures, probably owing toinactivation of the indigenous inhibitors of plas-min or of plasminogen activators. The high heat

stability of plasmin suggests that it may be pos-sible to develop a model system that is based onaseptic curd in which the rennet is denatured bya suitable cook temperature and acidified byGDL and that can be used to assess plasmin ac-tivity alone.

Some or all of these techniques, in variouscombinations, have been used to study the con-tribution of various agents to cheese ripening,especially proteolysis and flavor development.

The biochemistry of cheese ripening will beconsidered in three principal sections based onthe principal biochemical events: glycolysis, Ii-polysis, and proteolysis.

11.4 GLYCOLYSIS AND RELATEDEVENTS

The primary glycolytic event, the conversionof lactose to lactate, is normally mediated by thestarter culture during curd preparation or theearly stages of ripening. In cases where glycoly-sis has not been completed by the starter,NSLAB may contribute. The metabolism of lac-tose by LAB was discussed in Chapter 5.

Approximately 96% of the lactose in milk isremoved in the whey as lactose or lactate. How-ever, fresh curd contains a considerable amountof lactose, the fermentation of which has a sig-nificant effect on cheese quality. Obviously, theconcentration of lactose in fresh curd depends onits moisture content, the extent of fermentationprior to molding, and whether the curd is washedwith water or not. Cheddar curd, which is exten-sively drained and has reached a pH of about 5.4at milling, contains 0.8-1.0% lactose. In themanufacture of Dutch-type cheese, part of thewhey is removed and replaced by water, but thecurd is subjected to less syneresis and the pH ishigh (~ 6.2-6.3) at molding. Hence, Goudacheese curd contains around 3.0% lactose atmolding. Although Emmental cheese curd iscooked to a high temperature (52-550C) andhence undergoes extensive syneresis in the vat,it is transferred with the whey to molds at around

Page 256: Cheese Science

pH 6.4 and contains around 2% lactose at mold-ing.

Compared with other varieties, the residuallactose in Cheddar is fermented relativelyslowly at a rate and to an extent dependent on thepercentage salt-in-moisture (S/M) in the curd(see Figure 8-11). At low S/M concentrationsand low populations of NSLAB, residual lactoseis converted mainly to L-lactate by the starter. Ata high population of NSLAB (e.g., at a high stor-age temperature), a considerable amount of D-lactate is formed, partly by fermentation of re-sidual lactose and partly by isomerization of L-to o-lactate. At high S/M levels (e.g., 6%) andlow NSLAB populations, the concentration oflactose decreases slowly and changes in lactateare slight. The quality of Cheddar cheese isstrongly influenced by the fermentation of re-sidual lactose: the pH decreases after salting atS/M levels below 5%, owing primarily to thecontinued action of the starter, but at higher lev-els of S/M, starter activity decreases abruptly, asindicated by a high level of residual lactose and ahigh pH, accompanied by a sharp decrease incheese quality (see Figure 8-12).

Dutch-type cheese contains about 3.0% lac-tose at pressing, but the amount of lactose de-creases to undetectable levels within about 12hr.

Typically, Emmental cheese curd containsroughly 1.7% lactose 30 min after molding. Nei-ther Streptococcus thermophilus nor starter Lac-tobacillus spp. grow in Emmental curd duringcooking owing to the high cook temperature(52-550C). The curds are transferred to moldswhile still hot, but as they cool in the molds, Sc.thermophilus begins to grow and metabolize lac-tose. Only the glucose moiety of lactose is me-tabolized by Sc. thermophilus, and consequentlygalactose accumulates to a maximum of around0.7% at about 10 hr, when the galactose-positivelactobacilli begin to multiply. These metabolizegalactose and residual lactose to a mixture of D-and L-lactate, which reach around 0.35% and1.2%, respectively, at 14 days, when all the sug-ars have been metabolized (see Figure 10-13).

11.4.1 Effect of Lactose Concentration onCheese Quality

Although the concentration of lactose in milkdecreases with advancing lactation (see Figure 3-1), its concentration in bulked factory suppliesfrom cows on a staggered calving pattern is essen-tially constant. However, when synchronizedcalving is practiced, as in New Zealand, Ireland,and Australia, substantial seasonal changes occurin the concentration of lactose in milk and conse-quently in fresh cheese curd. Variations in theconcentration of lactose in cheese curd probablyaffect the final pH of the cheese, which, in turn,affects cheese texture, enzyme activity, and per-haps the nonstarter microflora. Cheese flavor islikely to vary owing to variations in the concen-tration of lactic and acetic acids and to variationsin the metabolic activity of the cheese microflora.

The concentration of lactose in cheese curd isinfluenced by some features of the manufactur-ing process. As far as is known, the concentra-tion of lactose in cheese curd is not increased in-tentionally for any variety of rennet-coagulatedcheese. However, curd made from milk concen-trated to a high factor by ultrafiltration (i.e., pre-cheese) contains a high level of lactose owing tothe lack of syneresis, and lactose may have to bereduced to an appropriate level by diafiltration.If the concentration of lactose in cheese curd istoo high, the concentration of D-Ca-lactate willexceed its solubility and crystallize on the sur-face of the cheese. The concentration of lactosein the curd of several varieties, including Goudaand Edam, is reduced by replacing part of thewhey by warm water. This process, which wasprobably introduced as a simple method forcooking the curds on farms lacking jacketedcheese vats, effectively controls the pH of thecheese. In these cheeses, the level of wash wateradded is based on the concentrations of lactoseand casein in the milk. This washing protocolminimizes variations in the pH of cheese curdexpress that might otherwise occur due to sea-sonal variations in the lactate:casein ratio. Curdsare washed in the washed-curd variants of Ched-

Page 257: Cheese Science

dar cheese and perhaps in the production of low-fat cheese to increase its moisture content.

The effect of variations in the concentration oflactose in cheese curd on the quality of the ma-ture cheese has received minimal attention. Inan attempt to vary the concentration of lactosein Cheddar cheese curd, Huffman and Kristof-fersen (1984) added lactose to the curd-wheymixture after cutting the coagulum, but owing tothe strong outflow of whey from the curd at thatstage, due to syneresis, the achieved increase inthe concentration of lactose within the curd wasquite small. Waldron and Fox (unpublishedstudy) reduced the lactose content of Cheddarcheese curd by replacing 35^45% of the wheyshortly after cutting the coagulum by an equalvolume of warm water. The curd contained0.25% lactose compared with about 1% in thecontrol. A curd containing only 0.03% lactosewas obtained by repeating the whey removal-re-placement treatment. The lactose level was alsoreduced by washing the curd with water prior tosalting, but this was less effective than whey re-placement, possibly because little syneresis oc-curred at this late stage of curd production. Thelactose content of other batches of curd was in-creased by using lactose-supplemented milk(6.4% or 8.4% lactose) for curd manufacture.Overall, the concentration of lactose in the 1-day-old cheese ranged from 0.03% to 2.5%.

Changes in the concentration of lactose in andthe pH of the cheese during ripening are shownin Figure 11—1. The lactose in both types ofwashed-curd cheese was completely metabo-lized within about 2 weeks, but it persisted in thehigh-lactose cheeses throughout ripening. Notsurprisingly, the pH of the cheeses was inverselyproportional to the concentration of lactose inthe curd. The pH of high-lactose cheeses contin-ued to decrease (to -4.8) throughout ripening,whereas in the washed-curd cheeses the pH in-creased once the lactose had been exhausted.This increase in pH is common in many varietiesof cheese, probably due to proteolysis and theproduction OfNH3 from amino acids. Flavor de-velopment was substantially faster in the high-lactose cheese than in the washed-curd cheeses,

although it was considered to be rather harsh,perhaps due to the low pH. The flavor of thelow-lactose cheeses was clean and mild. Therate of growth and the final number of NSLABwere not affected by the concentration of lac-tose, suggesting that NSLAB do not depend onlactose as a growth substrate. (Factors affectingthe growth of NSLAB in cheese are discussed inChapter 10 and by Fox, McSweeney, & Lynch,1998.)

The results of this study suggest that the con-centration of lactose in cheese curd has a sub-stantial effect on the quality of Cheddar andprobably other cheeses. Replacing some of thewhey by water or washing the cheese curd mightbe considered when a mild, clean flavor is de-sired. Normal variations in the lactose content ofmilk from mixed-calving herds are probably notsignificant but may have a substantial effectwhen synchronized calving is practiced. Undersuch circumstances, the pH of the cheese mayvary in an undesirable manner. The use of lowconcentration factor ultrafiltration (LCF-UF)milk for cheese manufacture is not expected toinfluence cheese pH, as the lactose content ofcheese is increased only very slightly (by~ 0.25% when milk is concentrated 1.5-fold).The problem is more serious when high concen-tration factor ultrafiltration retentate is used anddiafiltration of the cheese milk or washing of thecurd would appear to be desirable or essential.Cheese made from milk with a high content offat and casein may have a reduced lactose con-tent.

The presence of residual lactose or its compo-nent monosaccharides in cheese may lead toMaillard (nonenzymatic) browning, especially ifthe cheese is heated (e.g., as a food ingredient).Lactose per se is unlikely to be a problem exceptwhen very young cheese curd is used in pro-cessed cheese. Browning is most likely to beproblematic in cheeses made with thermophiliccultures. As discussed in Chapters 5 and 10, Sc.thermophilus is unable to metabolize the galac-tose moiety of lactose, which it excretes. If a ga-lactose-positive strain of Lactobacillus is used,the galactose will be metabolized to L- or DL-lac-

Page 258: Cheese Science

Res

idua

l L

acto

se (

%)

A

Time (Weeks)

pH

B

Time (weeks)

Figure 11-1 Changes in the concentration of lactose (A) and in the pH (B) of cheese made from curd withmodified levels of lactose during ripening. Control cheese, (O); 35% whey-replaced cheese, (Q); washed-curdcheese, (•); and lactose-enriched cheese 6.4%, A or 8.4%, (A).

Page 259: Cheese Science

fate, but most strains of Lactobacillus del-orueckii and Lb. lactis are galactose-negative,and therefore galactose accumulates. The re-sidual galactose may cause undesirable brown-ing in many cheeses but is a particularly seriousproblem in Mozzarella, which is subjected toconsiderable heating during the cooking ofpizza. Browning may also be problematic incheeses that are typically grated, such as Par-mesan and Grana, which have a low-moisturecontent and approach a water activity level thatis optimal for Maillard browning (~ 0.6). Sincegrating cheeses are extensively ripened, othercarbonyls (e.g., diacetyl and glyoxals), whichare very active in Maillard browning, may con-tribute to browning.

11.4.2 Modification and Catabolism ofLactate

The fate of lactic acid during cheese ripeninghas some significance in all varieties and is ofmajor consequence in some types. Lactic acidhas a direct effect on the taste of cheese, espe-cially young cheese, which lacks other flavorcompounds. Obviously, lactic acid affects thepH of cheese and consequently its texture (seeChapter 13). pH affects the solubility ofCa3(PO4)2 and hence also indirectly affectscheese texture. Perhaps most importantly, lacticacid is an important substrate for microbialgrowth in many cheese varieties, and its catabo-lism has major effects on cheese flavor (see Fox,Lucey, & Cogan, 1990).

Typical concentrations of lactate in Cam-embert, Swiss, Romano, and Cheddar are 1.0%,1.4%, 1.0%, and 1.5%, respectively. The fate oflactic acid in cheese depends on the variety. Ini-tially, Cheddar contains only L(+) lactic acid, butas the cheese matures the concentration of D-lac-tate increases, and eventually a racemic mixtureis formed (see Figure 10-11). o-Lactate could beformed from residual lactose by lactobacilli or byracemization of L-lactate by NSLAB, includingpediococci. Except in cases where the post-milling activity of the starter is suppressed (e.g.,by S/M > 6%), racemization is probably the prin-

cipal mechanism. Racemization of L-lactate ap-pears to occur in several cheese varieties (Tho-mas & Crow, 1983), and a racemic mixture willbe formed if the duration of ripening is longenough (Table 11-1). Racemization is not sig-nificant from the flavor viewpoint, but calcium D-lactate, which is less soluble than L-lactate, maycrystallize in cheese, especially on the surface,causing undesirable white specks.

Lactate in cheese may be oxidized to acetate.Pediococci produce 1 mole of acetate and 1 moleof CO2 and consume 1 mole of O2 per mole oflactate utilized. Lactate is not oxidized until allsugars have been exhausted. The oxidation oflactate to acetate in cheese depends on theNSLAB population and on the availability of O2,which is determined by the size of the block andthe oxygen permeability of the packaging mate-rial (Figure 11-2). Acetate, which may also beproduced by starter bacteria from lactose or cit-rate or from amino acids by starter bacteria andlactobacilli, is usually present at fairly high con-centrations in Cheddar cheese and is consideredto contribute to cheese flavor, although highconcentrations may cause off-flavors. Thus, theoxidation of lactate to acetate probably contrib-utes to Cheddar cheese flavor.

In Romano cheese, L-lactate predominatesinitially, reaching a maximum of around 1.9%at 1 day. The concentration of L-lactate beginsto decrease after 10 days, reaching 0.2-0.6% af-ter 150-240 days of ripening. Some of the de-crease is due to racemization to o-lactate, whichreaches a maximum (up to 0.6% in somecheeses) at around 90 days and then declinessomewhat. In some cheeses, acetate reachesvery high levels (1.2%) at around 30 days, butdecreases thereafter. The agents responsible forthe metabolism of acetate have not been identi-fied, but yeasts (Debaryomyces hansenii) maybe involved. Presumably, the oxidation of lac-tate to acetate also occurs in other hard andsemi-hard cheeses, but studies in this area arelacking.

The metabolism of lactate is very extensive insurface mold-ripened varieties, such as Cam-embert and Brie. The concentration of lactic acid

Page 260: Cheese Science

in these cheeses is around 1.0% at 1 day. Thelactic acid is produced exclusively by the meso-philic starter and hence is L-lactate. Secondaryorganisms quickly colonize and dominate thesurface of these cheeses, initially Geotrichumcandidum and yeasts, followed by Penicilliumcamemberti, and, in traditional manufacture, byBrevibacterium linens and other coryneformbacteria (see Chapter 10). G. candidum and P.camemberti rapidly metabolize lactate to CO2

and H2O, causing an increase in pH (Figure 11-3). Deacidification occurs initially at the surface,resulting in a pH gradient from the surface to thecenter and causing lactate to diffuse outward.When the lactate has been exhausted, P. cam-emberti metabolizes proteins, producing NH3,which diffuses inward, further increasing thepH. The concentration of calcium phosphate atthe surface exceeds its solubility at the increasedpH, and it precipitates as a layer of Ca3(PO4)2 atthe surface, thereby causing a calcium phosphategradient within the cheese (Figure 11-4). Theelevated pH stimulates the action of plasmin,which contributes significantly to proteolysis.Although surface microorganisms secrete very

potent proteinases, they diffuse into the cheeseto only a very limited extent; however, pep tidesproduced at the surface may diffuse into thecheese. The combined action of increased pH,loss of calcium (necessary for the integrity of theprotein network), and proteolysis is necessaryfor the very extensive softening of the body ofBrie and Camembert (Karahadian & Lindsay,1987; Lenoir, 1984).

The pH of Blue cheese also increases substan-tially during ripening (Figure 11-5), but in con-trast to surface mold-ripened cheeses, the extentof the increase is greater at the center than at thesurface. One would expect that catabolism oflactic acid would be responsible for the increasein pH, but the only published data available sug-gest that Blue cheese contains a high concentra-tion of lactic acid (^ 1.2%, see Table 11-1). Per-haps the increase in pH is due to the productionof NH3 on the catabolism of amino acids.Among Blue cheeses, the increase in pH appearsto be least in Danablue, in which low levels ofNH3 are produced.

Large changes in pH also occur in surfacesmear-ripened cheeses, especially at the surface

Table 11-1 Concentration of L(+)- and o(-)-Lactate in Various Cheeses

Lactate (g/100 g Cheese)

Cheese Type Age (Weeks) L(+) D(-)

Cheshire 15 0.75 0.6623 0.74 0.73

Colby 16 0.71 0.680.57 0.51

Egmont 20 0.62 0.37Gouda 10 0.84 0.31

15 0.66 0.550.69 0.42

Blue 8 0.65 0.610.74 0.43

Camembert 6 0.17 0.028 0.04 0.013 0.57 0.02

Feta - 0.97 0.88

Page 261: Cheese Science

(Figure 11-6). In these cheeses, lactate in thesurface layer is catabolized by yeasts, which arethe first microorganisms to colonize the surface.The increase in pH at the surface is a critical fac-tor in the ripening of these cheeses, since Brevi-bacterium linens, the characteristic microorgan-ism in the smear, does not grow at a pH below5.8 (see Chapter 10).

The catabolism of lactic acid is also criticalin Swiss-type cheeses, but the causative agentsand effects are different from those in surface

mold-ripened and smear-ripened cheeses. Ontransfer to the warm room, Propionibacteriumsp., the characteristic microorganisms in Swiss-type cheeses, multiply by 2-3 log cycles andmetabolize lactate, preferentially the L-isomer,to propionate, acetate, and CO2 (Figures 10-13and 10-14):

outside Distance (mm) outsidesurface surface

Figure 11-2 Acetate concentration gradients in 20 kg blocks of Cheddar cheese inoculated with P. pentosaceus1220 and ripened at 120C and 85-90% RH for 3, 6, and 10 months. The cheeses were wrapped in plastic film withan oxygen transmission rate (O2TR) of 27 ml O2 m~2 24 hr1 (black symbols) or 112 ml O2 nr2 24 hr1 (whitesymbols).

Ace

tate

con

cent

ratio

n (g

/100

g c

hees

e)

10 months

6 months

3 months

Lactic acid Propionic acid Acetic acid

Page 262: Cheese Science

Figure 11-3 Rate of lactic acid metabolism (•), ammonia production (•), and pH changes (+) in Brie cheesesampled at the center (A), surface (B), and corner (C) during ripening.

DAYS

PE

RC

EN

T

AM

MO

NIA

C

DAYS

•Lactic Acid•Ammonia«pH

B

PE

RC

EN

T

AM

MO

NIA

•Lactic Acid•Ammonia4pH

•Lactic Acid•Ammonia•pH

A

PE

RC

EN

T

AM

MO

NIA

DAYS

PE

RC

EN

T

LA

CT

IC

AC

IDP

ER

CE

NT

L

AC

TIC

AC

IDP

ER

CE

NT

L

AC

TIC

A

CID

Page 263: Cheese Science

Figure 11-4 Schematic representation of the gradients of calcium, phosphate, lactic acid, pH, and ammonia inripening of Camembert cheese.

Cross-SectionalView

CheeseExterior with

Surface Microflora

Ammonia Produced

Lactate MetabolizedCa3(PO4J2

Precipitated

Ch

eese

In

teri

or

>(Hi9her) Soluble Ca/P04/Lactate (Lower)

Concentration Gradient

(Lower)pH Gradient

(Higher)

(Lower) Ammonium Ion (Higher)Concentration Gradient

pH

Ripening period, days

Figure 11-5 Changes in the pH of two batches (A, A) of an Irish Blue cheese during ripening.

Page 264: Cheese Science

The CO2 generated is responsible for eye devel-opment, a characteristic feature of these variet-ies. Most of the CO2 produced diffuses throughthe curd and is lost, but if the growth of Propi-onibacterium sp. is adequate, sufficient CO2 isproduced to induce good eye formation. Theacetic acid and especially the propionic acid pro-duced in this fermentation contribute to the fla-vor of Swiss-type cheeses.

A common defect in many cheeses arisesfrom the metabolism of lactate (or glucose) byClostridium spp. to butyrate, H2, and CO2 (Fig-ure 11-7). This reaction leads to late gas blow-ing and off-flavors in many cheese varieties un-less precautions are taken, such as good hygiene,addition of NaNO3 or lysozyme, bactofugation,or microfiltration.

The significance of the primary fermentationof lactose to L-lactate in cheese manufacture iswell recognized (see Chapters 5, 6, and 10). Theforegoing discussion indicates that the metabo-lism of lactose and lactate in cheese during rip-ening is well understood. Quantitatively, thesechanges are among the principal metabolicevents in most cheese varieties. In comparisonwith other biochemical changes during cheeseripening, however, the conversion of lactose tolactate may have relatively little direct effect onthe flavor of mature cheese. Nonetheless, since itdetermines the pH of cheese, it is of major sig-nificance in regulating the various biochemicalreactions that occur in cheese during ripening.The isomerization of lactate probably has littleimpact on cheese flavor, but its conversion to

Time (days)

Figure 11-6 Changes in the pH in the surface (SI), middle (^), and core (Q) layers of Taleggio cheese duringripening.

PH

Page 265: Cheese Science

propionate and/or acetate is probably signifi-cant, and when it occurs, the metabolism of lac-tate to butyrate has a major adverse effect oncheese quality.

11.5 CITRATE METABOLISM

The relatively low concentration of citrate inmilk (~ 8 mM) belies the importance of its me-tabolism in some cheeses made using meso-philic cultures (for review, see Cogan & Hill,1993). Most of the starters used in cheese pro-duction—Lc. lactis subsp. lactis, Lc. lactissubsp. cremoris, Lactobacillus ssp. and Sc. ther-mophilus—do not metabolize citrate, but a mi-nor component of mixed-strain mesophilic start-ers, such as those used for Dutch-type cheeses,contain strains of Lc. lactis subsp. lactis andLeuconostoc spp., which metabolize citrate todiacetyl in the presence of a fermentable sugar

during manufacture and early ripening (see Fig-ure 5-12). The CO2 produced is responsible forthe small eyes characteristic of Dutch-typecheeses.

Diacetyl is a very significant compound forthe aroma and flavor of unripened cheeses, in-cluding Cottage cheese and Quarg, and manyfermented milks. It contributes to the flavor ofDutch-type cheeses and possibly also Cheddar.Acetate may also contribute to the flavor ofthese cheeses.

Approximately 90% of the citrate in milk issoluble and is lost in the whey. However, theconcentration of citrate in the aqueous phase ofcheese is roughly 3 times that in whey, reflectingthe concentration of colloidal citrate. Cheddarcheese contains 0.2-0.5% (w/w) citrate, whichdecreases to 0.1 % at 6 months through the meta-bolic activity of some mesophilic lactobacilli,many of which catabolize citrate to ethanol, ac-etate, and formate (see Figure 5-11) late in the

Figure 11-7 Metabolism of glucose or lactic acid by Clostridium tyrobutyricum, with the production of butyricacid, CO2, and H2.

Page 266: Cheese Science

ripening when numbers of NSLAB have in-creased sufficiently.

11.6 LIPOLYSIS AND RELATED EVENTS

Pure triglycerides elicit an oily sensation inthe mouth but are devoid of flavor stricto senso.However, lipids have a major effect on the flavorand texture of foods, including cheese. The in-fluence of lipids on cheese texture is discussedin Chapter 13.

Lipids contribute to cheese flavor in threeways:

1. They are a source of fatty acids, espe-cially short-chain fatty acids, which havestrong and characteristic flavors. Fattyacids are produced through the action oflipases in a process referred to as lipoly-sis. In some varieties, the fatty acids maybe converted to other sapid and aromaticcompounds, especially methyl ketonesand lactones.

2. Fatty acids, especially polyunsaturatedfatty acids, undergo oxidation, leading tothe formation of various unsaturated alde-hydes that are strongly flavored and causea flavor defect referred to as oxidativerancidity. Lipid oxidation appears to bevery limited in cheese, probably owing toits low redox potential (-250 mV).

3. Lipids function as solvents for sapid andaromatic compounds produced not onlyfrom lipids but also from proteins and lac-tose. Lipids may also absorb from the en-vironment compounds that cause off-fla-vors.

Of the various possible contributions of lipidsto cheese flavor, lipolysis and modification ofthe resultant fatty acids are the most significant.

The degree of lipolysis in cheese varieswidely between varieties, from about 6 mEq freefatty acids in Gouda to 45 mEq/100 g fat in Dan-ish Blue (Gripon, 1987,1993). Lipases in cheeseoriginate from milk, rennet preparation (paste),starter, adjunct starter, or nonstarter bacteria. Li-polysis in internally bacteria-ripened varieties,

such as Gouda, Cheddar, and Swiss, is generallylow but is extensive in mold-ripened and someItalian varieties. In general, in those varieties inwhich extensive lipolysis occurs, lipases origi-nate from the coagulant (rennet paste, whichcontains pre-gastric esterase, as used in someItalian varieties) or from the adjunct culture(Penicillium spp., which produce a number of li-pases [Gripon, 1987, 1993] in mold-ripened va-rieties).

11.6.1 Lipases and Lipolysis

Lipases are hydrolases that hydrolyze estersof carboxylic acids (EC group 3.1.1):

/P /£>R1—C— OR2 + H2O ^-R1C—OH + R2OH

Lipases have little or no effect on soluble es-ters, and they prefer to act at the oil-water inter-face of emulsified esters. Thus, lipases are dis-tinguished by the physical state of the substraterather than by the type of bond hydrolyzed.

Lipases exhibit various types of specificity:

• They are usually specific for the outer esterbonds of tri- or diglycerides (i.e., snl andsn3 positions). Thus, initially they hy-drolyze triglycerides to 1,2- and 2,3-di-glycerides and later to 2-monoglycerides.The fatty acid at the sn2 position migratesto the vacant snl or sn3 position and is thenreleased by lipase (Figure 11-8). Therefore,lipases eventually hydrolyze triglyceridesto glycerol and 3 fatty acids. In mostcheeses, lipolysis probably does not go be-yond the first step.

• Lipases usually exhibit specificity for fattyacids of certain chain length.

• Some lipases exhibit specificity for satu-rated or unsaturated fatty acids.

Although some lipases may be optimally ac-tive at neutral or acid pH values, most have analkaline pH optimum. Lipases are inactivated byfatty acids and therefore are activated by Ca2+,which precipitates the fatty acids as insoluble

Page 267: Cheese Science

Triglyceride 1,2-Diglyceride 2,3-Diglyceride

1-Monoglyceride 3-Monoglyceride 2-Monoglyceride

Glycerol

Figure 11-8 Hyydrolysis of a triglyceride by a lipase.

soaps and removes them from the reaction envi-ronment. The whey proteins (3-lactoglobulin andblood serum albumin bind fatty acids and stimu-

late lipase activity. Lipases are activated by bilesalts, which emulsify the triglyceride substrate.Some lipases, referred to as lipoprotein Upases,

Page 268: Cheese Science

are also stimulated by lipoproteins, which pro-mote the adsorption of the enzyme at the oil-wa-ter interface. An example is blood serum lipase,which is the indigenous lipase in milk.

11.6.2 Indigenous Milk Lipase

Milk contains an indigenous lipoprotein li-pase (LPL) that is well characterized(Olivecrona & Bengtsson-Olivecrona, 1991;Olivecrona, Vilaro, & Bengtsson-Olivecrona,1992). The enzyme enters milk as a result ofleakage through the mammary cell membranefrom the blood where it is involved in the me-tabolism of plasma triglycerides. Bovine milkcontains 12 mg lipase/L (10-20 nM). Under op-timum conditions, it has a turnover of 3,000 s"1

and could theoretically release sufficient fattyacids in 10 s to cause hydrolytic rancidity. How-ever, most of the lipase (> 90%) is associatedwith the casein micelles, and the fat occurs inglobules surrounded by a lipoprotein membrane(the milk fat globule membrane [MFGM]).Thus, the substrate and enzyme are compart-mentalized, and lipolysis does not occur unlessthe MFGM is damaged by agitation, foaming,freezing, or homogenization, for example. Inap-propriate milking and milk-handling techniquesat the farm and/or factory may cause sufficientdamage to the MFGM to permit significant li-polysis and thus off-flavors in cheese and otherdairy products.

LPL is rather nonspecific for the type of fattyacid but is specific for the snl and sn3 positionsof mono-, di-, and triglycerides. Therefore, li-polysis in milk leads to preferential release ofshort- and medium-chain acids, which in milktriglycerides are esterified predominantly at thesn3 position. Since more than 90% of the LPL inbovine milk is associated with the casein mi-celles, it is incorporated into cheese curd. LPL isrelatively heat labile and is extensively inacti-vated by HTST pasteurization although heatingat the equivalent of 780C for 10 s is required forcomplete inactivation. Significantly more li-polysis occurs in raw milk cheese than in pas-teurized milk cheese (Figure 11-9). Milk LPL

probably contributes to this difference but theNSLAB microflora of raw and pasteurized milkcheeses also differ markedly.

11.6.3 Lipases from Rennet

Good quality rennet extract contains no Ii-polytic activity. However, rennet paste used inthe manufacture of hard Italian varieties (e.g.,Romano, Provolone) contains a potent lipase,pregastric esterase (PGE), which is responsiblefor the extensive lipolysis in and the characteris-tic "piccante" flavor of such varieties. The litera-ture on PGE was comprehensively reviewed byNelson, Jensen, and Pitas (1977) and updated byFox and Stepaniak (1993).

PGE, also called lingual or oral lipase, is se-creted by glands at the base of the tongue. Suck-ling stimulates the secretion of PGE, which issubsequently washed into the abomasum bymilk and siliva. Rennet paste is prepared fromthe abomasa of calves, kids, or lambs slaugh-tered after suckling. The abomasa are partiallydried and ground into a paste, which is slurriedin milk or water before being added to cheesemilk. Rennet pastes are considered unhygienic,and their use is not permitted in several coun-tries, including the United States. Instead, par-tially purified PGEs are used.

Calf, kid, and lamb PGEs have been partiallypurified from commercial preparations and calfPGE from oral tissue. The enzyme is a glycopro-tein with an isoelectric point of 7.0 and a mo-lecular weight of about 49 kDa. The gene for ratlingual lipase has been cloned and sequenced,and the primary structure of the enzyme hasbeen deduced. PGE is highly specific for short-chain acids esterified at the sn3 position andtherefore releases high concentrations of highlyflavored short- and medium-chain acids frommilk fat. The specificity of calf, lamb, and kidPGEs differ slightly, and consequently the flavorcharacteristics of Italian cheese differ slightly,depending on the source of the PGE used. Mostother lipase s are unsuitable for the manufactureof Italian cheese because of incorrect specificity,but it has been claimed that certain fungal Ii-

Page 269: Cheese Science

pases may be acceptable alternatives (see Fox &Stepaniak, 1993). The use of PGE to acceleratethe ripening of other cheese varieties is dis-cussed in Chapter 15.

11.6.4 Microbial Lipases

Lactococcus spp. and Lactobacillus spp. havelower levels of lipolytic activity than other bac-teria (e.g., Pseudomonas) and molds. However,in the absence of strongly lipolytic agents and

when present at high numbers over a long pe-riod, as in ripening cheese, lipases and esterasesof lactococci and lactobacilli are probably theprincipal lipolytic agents in Cheddar and Dutch-type cheeses made from pasteurized milk. Asep-tic cheeses acidified with GDL instead of starterhave low concentrations of free fatty acids thatdo not increase during ripening. The lipase andesterase activity of LAB appears to be entirelyintracellular. Cell-free extracts of various dairyLAB are most active on tributyrin at pH 6-8 and

Weeks

Figure 11-9 Liberation of free fatty acids in Cheddar cheese made from raw milk (•), pasteurized milk (•), and(A) microfiltered milk.

mEq

Pal

mita

te/k

g C

hees

e

Page 270: Cheese Science

at 370C. They have little or no activity on trig-lycerides of long-chain fatty acids (e.g., > Ci0).There appears to be considerable interstrainvariation in esterase and lipase activity, andsome strains appear to possess two esterases.Starter bacteria can liberate free fatty acids frommono- and diglycerides produced in milk byother lipases (e.g., milk LPL or Upases fromgram-negative bacteria).

The intracellular esterase and lipase of twoLactococcus strains have been isolated and char-acterized (Chich, Marchesseau, & Gripon, 1997;Holland & Coolbear, 1996). At present, little isknown about the genetics of these enzymes. Iso-lation of lipase- and esterase-negative variantsof Lactococcus would permit the significance ofthese enzymes in cheese ripening to be assessed.

Both mesophilic and thermophilic lactobacillipossess (mainly) intracellular esterolytic and Ii-polytic activity (Gobbetti, Fox, & Stepaniak,1996; Khalid, El-Soda, & Marth, 1990). Theesterolytic and lipolytic activity in cellhomogenates of a number of lactobacilli wascharacterized by El-Soda, Abd El-Wahab,Ezzat, Desmazeaud, and Ismail (1986), and anintracellular lipase and an intracellular esterasefrom Lb. plantarum were purified and character-ized by Gobbetti, Fox, and Stepaniak (1997) andGobbetti, Fox, Smacchi, Stepaniak, andDamiani(1997).

Micrococcus, which constitutes part of thenonstarter microflora of cheese, especiallythe surface microflora, produces lipasesthat may contribute to lipolysis during ripening(Bhowmik & Marth, 199Oa, 199Ob). The non-starter microflora of cheese may also includePediococcus spp., which are weakly esterolyticand lipolytic (Bhowmik & Marth, 1989;Tzanetakis & Litopoulou-Tzanetaki, 1989).

An intracellular lipase of Propionibacteriumshermanii was partially characterized byOterholm, Ordal, and Witter (1970); it probablycontributes to lipolysis in Swiss varieties. Brevi-bacterium linens, a major component of the sur-face of smear-ripened cheeses, possesses intrac-ellular lipases and esterases. An intracellularesterase has been purified and characterized(Rattray & Fox, 1997).

Extensive lipolysis occurs in mold-ripenedcheese, particularly Blue varieties. In somecases, up to 25% of the total fatty acid may befree (see Gripon, 1987, 1993). However, the im-pact of free fatty acid on the flavor of Blue mold-ripened cheeses is less than in hard Italian variet-ies, possibly owing to neutralization as the pHincreases during ripening and to the dominantinfluence of methyl ketones on the flavor ofBlue cheese. Lipolysis in mold-ripened varietiesis due primarily to the lipases of Penicilliumroqueforti or P. camemberti, which secrete po-tent, well-characterized extracellular lipases(see Gripon, 1993). P. camemberti appears toexcrete only one lipase, which is optimally ac-tive at around pH 9.0 and 350C. P. roquefortiexcretes two lipases, one with a pH optimum ataround 8.0, the other at around 6.0. The acid andalkaline lipases exhibit different specificities.Geotrichum candidum produces two lipaseswith different substrate specificities (seeCharton, Davies, & McCrae, 1992; Sidebottometal, 1991).

Psychrotrophs, which usually dominate themicroflora of refrigerated milk, are a potentiallyimportant source of potent lipases in cheese butare considered not to be very important unlesstheir numbers exceed 107 cfu/ml. Many psy-chrotroph lipases are heat stable and thus maycause rancidity in cheese over the course of along ripening period. The subject of psychro-troph enzymes in cheese was discussed byMortar (1989). Unlike psychrotroph proteinases,which are largely water-soluble and are lost inthe whey, psychrotroph lipases adsorb onto thefat globules and are therefore concentrated incheese.

11.6.5 Pattern and Levels of Lipolysis inSelected Cheeses

Lipolysis is considered to be undesirable inmost cheese varieties. Cheddar, Gouda, andSwiss-type cheeses containing even a moderatelevel of free fatty acids would be consideredrancid. However, certain cheese varieties arecharacterized by extensive lipolysis (e.g.,Romano, Parmesan, and Blue cheeses). Only

Page 271: Cheese Science

small qualitative and quantitative differences infree fatty acids (C2:o-Ci8:3) occur between Ched-dar cheeses differing widely in flavor. The pro-portions of free fatty acids (C6:0-Ci8:3) in cheeseare similar to those in milk fat, indicating thatthey are released in a nonspecific manner. How-ever, free butyric acid is usually present at ahigher concentration than can be explained byits proportion in milk fat, suggesting that it isliberated selectively. Lipolysis in hard Italianvarieties is extensive and due primarily to theaction of PGE in the rennet paste used in themanufacture of these cheeses. Lipolysis in Bluecheese varieties is extensive owing to the actionof Upases from Penicillium spp. The free fattyacid levels in a number of cheese varieties arelisted in Table 11-2.

11.6.6 Catabolism of Fatty Acids

The taste and aroma of Blue cheese is domi-nated by saturated n-methyl ketones, a homolo-gous series which, containing an odd number ofcarbon atoms from C3 to Q7, is present. Concen-trations of methyl ketones in Blue cheese fluctu-ate, presumably due to reduction to secondaryalcohols, but heptan-2-one, nonan-2-one, andundecan-2-one dominate.

The metabolism of fatty acids in cheese byPenicillium spp. involves four main steps (Fig-ure 11-10; see Kinsella & Hwang, 1976):

1. release of fatty acids by the lipolytic sys-tems (see Sections 11.6.1 to 11.6.5)

2. oxidation of p-ketoacids3. decarboxylation to methyl ketone with

one less carbon atom4. reduction of methyl ketones to the corre-

sponding secondary alcohol (this step isreversible under aerobic conditions)

The concentration of methyl ketones is re-lated to lipolysis. Methyl ketones can also beformed by the action of the mold on the keto-acids naturally present at low concentrations inmilk fat (~ 1% of total fatty acids). They couldalso be formed by the oxidation of mono-unsaturated acids, but evidence for such a path-way is equivocal.

Table 11-2 Typical Concentrations of FreeFatty Acids (FFAs) in Some Cheese Varieties

Variety FFA (mg/kg)

Sapsago 211Edam 356Mozzarella 363Colby 550Camembert 681Port Salut 700Monterey Jack 736Cheddar 1,028Gruyere 1,481Gjetost 1,658Provolone 2,118Brick 2,150Limburger 4,187Goat's milk cheese 4,558Parmesan 4,993Romano 6,754Blue (US) 32,230Roquefort 32,453

A number of factors affect the rate of methylketone production, including temperature, pH,physiological state of the mold, and the ratio ofthe concentration of fatty acids to the dry weightof spores. Both resting spores and fungal myce-lium are capable of producing methyl ketones.The rate of production of methyl ketones doesnot depend directly on the concentrations of freefatty acid precursors. Indeed, high concentra-tions of free fatty acids are toxic to P. roqueforti.

Lactones are cyclic esters resulting from theintramolecular esterification of a hydroxyacidthrough the loss of water to form a ring structure:

A fatty acid A6-ketone

Page 272: Cheese Science

a-Lactones and (3-lactones are highly reactiveand are used, or occur, as intermediates in or-ganic synthesis; y- or 8-lactones are stable andhave been found in cheese. Lactones possess astrong aroma, which, although not specificallycheese-like, may be important in the overall fla-vor of cheese.

y-Lactones and 8-lactones in freshly secretedmilk probably originate from the correspondingy- and 8-hydroxyacids following release fromtriglycerides; they are formed spontaneouslyfollowing release of the corresponding hydroxy-acid. Lactones could also be produced from ketoacids released by lipolysis, followed by reduc-tion to hydroxyacids. It has been reported thatthe mammary gland of ruminants has a 8-oxida-tion system for fatty acid catabolism, and thusoxidation within the mammary gland may be theprimary source of lactone precursors. The poten-tial for lactone production depends on such fac-tors as feed, season, stage of lactation, andbreed.

8-Lactones have very low flavor thresholds.y-C12, y-C14, y-C16, 8-C10, 8-C12, S-C14, 8-C15, 8-C16, and 8-C18 lactones have been identified inCheddar cheese, and their concentration corre-lates with age and flavor intensity, suggestingthat certain lactones are significant in Cheddarcheese flavor, although this has not been con-firmed.

The concentration of lactones in Blue cheeseis higher than in Cheddar, probably reflectingthe extensive lipolysis that occurs in Bluecheese. The principal lactones in Blue cheese areS-C14 and S-C16.

11.7 PROTEOLYSIS

11.7.1 Introduction

Of the three primary biochemical events (gly-colysis, lipolysis, and proteolysis) that occur incheese during ripening, proteolysis is the mostcomplex and, in the view of most investigators,

Figure 11-10 p-Oxidation of fatty acids to methyl ketones by Penicillium roqueforti and subsequent reductionto secondary alcohols.

Secondary alcohol (Cn.])

Redufiase

Methylketone(C,,.,) + CO2

Keto acyl CoA

Saturated fatty acids (C11)

p-Kcioacyldccarboxylase

Krcbs Cycle

ThiolaseThiohydrolasc

CoA-SH + p-Keto acid Acetyl CoA + Acyl CoA (C2n.2)

Next Page

Page 273: Cheese Science

a-Lactones and (3-lactones are highly reactiveand are used, or occur, as intermediates in or-ganic synthesis; y- or 8-lactones are stable andhave been found in cheese. Lactones possess astrong aroma, which, although not specificallycheese-like, may be important in the overall fla-vor of cheese.

y-Lactones and 8-lactones in freshly secretedmilk probably originate from the correspondingy- and 8-hydroxyacids following release fromtriglycerides; they are formed spontaneouslyfollowing release of the corresponding hydroxy-acid. Lactones could also be produced from ketoacids released by lipolysis, followed by reduc-tion to hydroxyacids. It has been reported thatthe mammary gland of ruminants has a 8-oxida-tion system for fatty acid catabolism, and thusoxidation within the mammary gland may be theprimary source of lactone precursors. The poten-tial for lactone production depends on such fac-tors as feed, season, stage of lactation, andbreed.

8-Lactones have very low flavor thresholds.y-C12, y-C14, y-C16, 8-C10, 8-C12, S-C14, 8-C15, 8-C16, and 8-C18 lactones have been identified inCheddar cheese, and their concentration corre-lates with age and flavor intensity, suggestingthat certain lactones are significant in Cheddarcheese flavor, although this has not been con-firmed.

The concentration of lactones in Blue cheeseis higher than in Cheddar, probably reflectingthe extensive lipolysis that occurs in Bluecheese. The principal lactones in Blue cheese areS-C14 and S-C16.

11.7 PROTEOLYSIS

11.7.1 Introduction

Of the three primary biochemical events (gly-colysis, lipolysis, and proteolysis) that occur incheese during ripening, proteolysis is the mostcomplex and, in the view of most investigators,

Figure 11-10 p-Oxidation of fatty acids to methyl ketones by Penicillium roqueforti and subsequent reductionto secondary alcohols.

Secondary alcohol (Cn.])

Redufiase

Methylketone(C,,.,) + CO2

Keto acyl CoA

Saturated fatty acids (C11)

p-Kcioacyldccarboxylase

Krcbs Cycle

ThiolaseThiohydrolasc

CoA-SH + p-Keto acid Acetyl CoA + Acyl CoA (C2n.2)

Previous Page

Page 274: Cheese Science

the most important. It is primarily responsiblefor textural changes—in hardness, elasticity, co-hesiveness, fracturability, stretchability, melt-ability, adhesiveness, and emulsifying prop-erties (see Chapter 13)—and makes a majorcontribution to cheese flavor and the perceptionof flavor (through release of sapid compounds).Unfortunately, some small peptides are bitterand, if present at sufficient concentrations, willcause bitterness, a common flavor defect incheese (see Chapter 12).

Proteolysis during maturation is essential inmost cheese varieties. The extent of proteolysisvaries from very limited (e.g., Mozzarella) tovery extensive (e.g., Blue varieties), and theproducts range in size from large polypeptidesonly slightly smaller than the intact caseinsthrough a succession of medium and small pep-tides to free amino acids. Clearly, no one pro-teolytic agent is responsible for such a widerange of products. Small peptides and amino ac-ids contribute directly to cheese flavor, and thelatter may be catabolized to a range of sapid andaromatic compounds that are major contributorsto cheese flavor (e.g., amines, acids, carbonyls,and sulfur-containing compounds). Althoughthe catabolism of amino acids is not proteolysis,it is dependent on the formation of amino acidsand will be treated in this section.

11.7.2 Assessment of Proteolysis

Proteolysis is routinely monitored in studieson cheese ripening and is a useful index ofcheese maturity and quality. Considering thecomplexity of proteolysis, a variety of methodsmay be used, depending on the depth of informa-tion required. These methods fall into two gen-eral classes: specific and nonspecific. The latterinclude determination of nitrogen soluble in orextractable by one of a number of solvents orprecipitants (e.g., water, pH 4.6 buffers, NaCl,ethanol, trichloroacetic acid, phosphotungsticacid, and sulfosalicylic acid) or permeablethrough ultrafiltration membranes and quanti-fied by any of several methods (e.g., Kjeldahl,biuret, Lowry, Hull, absorbance at 280 nm) or bythe formation of reactive a-amino groups quan-

tified by reaction with one of several reagents(e.g., trinitrobenzene sulphonic acid [TNBS], o-phthaldialdehyde [OPA], fluorescamine, Cd-ninhydrin, and Li-ninhydrin). Such methods arevaluable for assessing the overall extent of pro-teolysis and the general contribution of each pro-teolytic agent. Nonspecific techniques are rela-tively simple and are valuable for the routineassessment of cheese maturity, since soluble ni-trogen correlates well with cheese age and to alesser extent with quality.

Specific techniques involve the use of chroma-tography and/or electrophoresis, which resolveindividual peptides. They permit monitoring pro-teolysis of the individual caseins and identifica-tion of the peptides formed. Various forms ofchromatography have been used to study peptidesin cheese, including paper, thin-layer, ion-ex-change, gel permeation, and metal chelate tech-niques as well as, more recently, a variety of high-performance techniques, especially reverse-phase high-performance liquid chromatography.Electrophoresis is a very effective and populartechnique for assessing primary proteolysis incheese, especially alkaline urea-PAGE, but SDS-PAGE and isoelectric focusing are also used. Gelelectrophoretograms are not easy to quantify ac-curately, which is a major limitation of these tech-niques. In recent years, capillary electrophoresisis being applied increasingly to the analysis ofpeptides in cheese and has given very satisfactoryquantifiable results.

Techniques for assessing proteolysis incheese during ripening have been the subject ofa number of recent reviews, including Fox(1989); Fox, McSweeney, and Singh (1995);Grappin, Rank, and Olson (1985); InternationalDairy Federation (1991); McSweeney and Fox(1993, 1997); and Rank, Grappin, and Olson(1985). See Chapter 23 for further details.

11.7.3 Proteolytic Agents in Cheese andTheir Relative Importance

Cheese contains proteolytic enzymes fromrennet, milk, starter lactic acid bacteria (LAB),adventitious nonstarter LAB (NSLAB), and, inmost varieties, secondary cultures.

Page 275: Cheese Science

Several studies using the model cheese sys-tems described in Section 11.3, especially stud-ies on Cheddar and Gouda, have shown that en-zymes in rennet (chymosin or rennet substitute)are mainly responsible for initial proteolysis andthe production of most of the water-soluble orpH 4.6-soluble N (Figure 11-11). However, theproduction of small peptides and free amino ac-ids is due primarily to the action of enzymesfrom starter bacteria. y-Caseins, formed from (3-caseins by plasmin, have been found in allcheese varieties that have been studied, indicat-ing plasmin activity, and such activity has beenconfirmed in cheese supplemented with plasminor containing a plasmin inhibitor. Nonetheless,plasmin activity is probably not necessary forsatisfactory cheese ripening. The rennet en-zymes are extensively, probably completely, de-natured in high-cooked cheeses, such as Mozza-rella, Parmesan, and Emmental, and thereforethe contribution of plasmin to primary proteoly-sis is considerably higher in these varieties thanin Cheddar- and Dutch-type cheeses. The pHoptimum for plasmin is about 7.5, and hencecheese (~ pH 5.2) is not a very suitable substrate.However, the pH of many cheeses increases dur-ing ripening (e.g., the pH of Camembert in-creases to ~ 7), and these therefore are moreamenable to plasmin action.

Although NSLAB can dominate the micro-flora of Cheddar-type cheese during much of itsripening, their influence on proteolysis in cheeseis relatively limited and mainly at the level offree amino acid formation.

Some adjunct or secondary cultures are veryproteolytic (e.g., P. roqueforti, P. camemberti,and B. linens). Consequently, these microorgan-isms make a major contribution to proteolysis inthose cheeses in which they are used, especiallyin Blue cheeses, in which extensive mold growthoccurs throughout the cheese.

The extent and specificity of proteolysis inrepresentatives of the principal groups of cheesehave been characterized (and are described inSection 11.8). The specificity of the principalproteinases and peptidases on the individualcaseins in cheese has been established and canbe related to proteolysis in cheese. The specific-

ity of the principal proteolytic enzymes found incheese is described briefly below, and actualproteolysis in some varieties is discussed in Sec-tion 11.8.

11.7.4 Specificity of Cheese-relatedProteinases

Coagulant

As discussed in Chapter 6, the principal andessential role of the coagulant in cheesemakingis the specific hydrolysis of K-casein, as a resultof which the colloidal stability of the casein mi-celles is destroyed and coagulation occurs undersuitable conditions of temperature and calciumconcentration. Most of the rennet added tocheesemilk is lost in the whey or denatured as aresult of cooking the curd-whey mixture to ahigh temperature. Typical values for the percent-age of added rennet retained in the curd rangefrom about 0% for high-cook cheese (e.g., Moz-zarella, Parmesan, and Emmental) through 6%for Cheddar to about 20% for high-moisture,low-cook cheeses (e.g., Camembert).

Chymosin (EC 3.4.23.4), the principal pro-teinase in traditional rennets used for cheese-making, is an aspartyl proteinase of gastric ori-gin, secreted by young mammals. Its action onthe B-chain of insulin indicates that chymosin isspecific for hydrophobic and aromatic aminoacid residues. Chymosin is weakly proteolytic.Indeed, limited proteolysis is one of the charac-teristics to be considered when selecting protein-ases for use as rennet substitutes.

The primary chymosin cleavage site in themilk protein system is the Phei05-Meti06 bond inK-casein, which is many times more susceptibleto chymosin than any other bond in milk pro-teins, and its hydrolysis leads to coagulation ofthe milk (see Chapter 6). Cleavage of K-casein atPheios-Metloe yields para-K-casein (K-CN fl-105) and glycomacropeptides (GMP; K-CNf 106-169). Most of the glycomacropeptides arelost in the whey, but the para-K-casein remainsattached to the casein micelles and is incorpo-rated into the cheese. ocsi-casein, ocs2-casein, andP-casein are not hydrolyzed during milk coagu-

Page 276: Cheese Science

lation but may be hydrolyzed in cheese duringripening.

In solution, chymosin cleaves the LeUi92-Ty^93

bond of p-casein very rapidly; the larger peptide,p-CN f 1-192, is commonly referred to as P-I-casein. At very low ionic strength (e.g., distilledwater), this bond is the second most susceptiblebond to hydrolysis in the caseins, after the PhCi05-Metio6 bond of K-casein, with a KM and kcat of0.075 mM and 1.54 S"1 for the micellar protein and0.007 mM and 0.56s"1 for the monomeric protein.However, its hydrolysis is strongly inhibitedwhen the ionic strength is increased, even in 50mM phosphate buffer; it is strongly inhibited by5% NaCl and completely by 10% NaCl. The bondAlai89-Phei90 and bonds in the region of residuesLeUi65 and LeUi40 of p-casein are hydrolyzed lessrapidly. The peptide p-CN f 1-189 has the samemobility as p-CN fl-192 on urea-PAGE at pH 9and is also referred to as p-I-casein. The peptidesP-CN fl-165 and p-CN fl-140, referred to as p-II- and p-III-casein, respectively, are readily re-solved by urea-PAGE.

p-Casein undergoes very little proteolysis bychymosin in cheese. Undoubtedly, NaCl is an

inhibitory factor (Cheddar cheese contains4-6% salt-in-moisture), but even in salt-freecheese proteolysis is very limited. Perhaps theconcentration of milk salts is sufficient to causeinhibition, and protein-protein interactions mayalso contribute to the low level of proteolysis(the C-terminal region of p-casein is very hydro-phobic, and intermolecular hydrophobic interac-tions may cause the chymosin-susceptible bondsto become inaccessible). The small peptide P-CN f 193-209 and ^agments thereof are bitterand hence even limited hydrolysis of p-casein bychymosin may cause bitterness. Because theconcentration of NaCl in the interior of mostbrine-salted cheeses increases slowly due to dif-fusion from the surface (see Chapter 8), suffi-cient proteolysis of p-casein by chymosin mayoccur in Dutch-type and other low-cookedcheese to cause bitterness.

The primary site of chymosin action on ocsi-casein is PhC23-PhC24. Cleavage of this bond isbelieved to be responsible for softening ofcheese texture, and the small peptide asrCN fl-23 is rapidly hydrolyzed by starter proteinases.The hydrolysis of asrcasein in solution by

Age

Figure 11-11 Formation of water-soluble nitrogen (WSN) in Cheddar cheese that (A) has controlled microflora(is free of nonstarter bacteria); (B) has controlled microflora and is chemically acidified (starter free); (C) hascontrolled microflora and is rennet free; and (D) has controlled microflora and is rennet and starter free.

WSN

as

% o

f to

tal N A

B

C

D

Page 277: Cheese Science

chymosin is influenced by pH and ionicstrength. In 0,1 M phosphate buffer, pH 6.5,chymosin cleaves ocsl-casein at PhC23-PhC24,Phe28-Pro29, Leu40-Ser4i, LeUi49-PhCi50, Phe153-Tyri54, LeUi56-ASp157, Tyr159-Proi6o, and Trp164-Tyr165. These bonds are also hydrolyzed at pH5.2 in the presence of 5% NaCl (i.e., conditionssimilar to those in cheese), and in addition Leun-PrOi2, PhC32-GIy33, LeU10I-LyS102, LeU142-AIa144,and Phe179-Ser180 are hydrolyzed (McSweeney,Olson, Fox, Healy, & Hojmp, 1993a). The rateat which many of these bonds are hydrolyzeddepends on the ionic strength and pH, particu-larly LeU101-LyS102, which is cleaved far faster atthe lower pH. The kcat and KM for the hydrolysisof Phe23-Phe24 bond of asl-casein by chymosin is0.7 s-1 and 0.37 mM, respectively.

Primary proteolysis (i.e., the formation oflarge peptides) in low-cooked cheese, includingCheddar, in which the chymosin is not inacti-vated during cooking, is due mainly tochymosin, and as discussed in Section 11.8, thecleavage sites correspond to those cleaved in(X81-CN in solution at pH 5.2.

as2-Casein appears to be relatively resistant toproteolysis by chymosin. Cleavage sites are re-stricted to the hydrophobic regions of the mol-ecule (i.e., residues 90-120 and 160-207). Thebonds Phe88-Tyr89, Tyr95-Leu96, Gln97-Tyr98,Tyr98-Leu99, Phe163-Leu164, PhC174-AIa175, andTyr179-Leu180 were reported by McSweeney,Olson, Fox, and Healy (1994) to be the primarycleavage sites. Although para-K-casein has sev-eral potential chymosin cleavage sites, it doesnot appear to be hydrolyzed either in solution orin cheese.

Good-quality calf (veal) rennet contains about10% bovine pepsin (EC 3.4.23.1), but values upto 50% have been reported in "calf rennet. Theprincipal peptides produced from Na-caseinateby bovine pepsin are similar to those producedby chymosin, but the specificity of bovine orporcine pepsins on bovine caseins has not beenrigorously determined. The LeU109-GIu110 bondof asl-CN appears to be resistant to chymosinbut is relatively susceptible to pepsin. The pep-tide OC81-CN fl 10-199 is quite pronounced in

electrophoretograms of cheese made with com-mercial calf rennet but not of cheese made usingmicrobial chymosin, which has the same speci-ficity as purified calf chymosin.

The specificity of the microbial rennet substi-tutes is quite different from that of chymosin(Figure 11-12); C. parasitica proteinase is muchmore active on P-casein than chymosin. Theprincipal cleavage sites for R. miehei proteinasein ocsl-casein are Phe23-Phe24, PhC24-VaI25,Met123-Lys124, and Tyr165-Tyr166. The principalsites on p-casein are GIu31-LyS32, VaI58-VaI59,Met93-Gly94, and Phe190-Leu191.

Indigenous Milk Proteinases

Milk contains several indigenous proteinases.Plasmin is the principal indigenous proteinase,but a low level of cathepsin D is also present.Several other proteinases have been reported inmilk but are believed to be of little significancein milk and dairy products.

Plasmin. Plasmin (fibrinolysin, EC 3.4.21.7)has been the subject of much study (for reviewssee Bastian & Brown, 1996; Grufferty & Fox,1988). It is a component of blood, where itsphysiological role is solubilization of fibrinclots. Plasmin is a component of a complex sys-tem consisting of the active enzyme, its zy-mogen (plasminogen), and activators and inhibi-tors of the enzyme and of its activators (Figure11-13), all of which are present in milk. Thecomponents of the plasmin system enter milk viadefective mammocyte membranes and are el-evated in late lactation and during mastitic infec-tion. Plasmin, plasminogen, and plasminogenactivators are associated with the casein micellesin milk and consequently are incorporated intocheese curd, while the inhibitors of both plasminand plasminogen activators are in the serumphase and are lost in the whey.

Plasmin is a trypsin-like serine proteinasewith a pH optimum at about 7.5 and a high speci-ficity for peptide bonds containing lysine at theN-terminal side. It is active on all caseins but es-pecially on ocs2- and p-caseins. Plasmin cleavesp-casein at three primary sites: LyS28-LyS29,LyS105-HiS106, and LyS107-GIu108, with the forma-

Page 278: Cheese Science

tion of the polypeptides p-CN f29-209 (Yi-CN),f 106-209 (Y2-CN), and fl 08-209 (Y3-CN); (3-CNf 1-105 and fl-107 (proteose peptone 5); p-CNf29-105 and f29-107 (proteose peptone 8-slow);and (3-CN fl-28 (proteose peptone 8-fast) (seeFigure 3-11). Additional cleavage sites includeLyS29-Ue30, Lysii3-TyrU4, and Arg183-Aspi84.

Plasmin cleaves as2-casein in solution at eightsites—LyS2I-GIn22, Lys24-Asn25, Argn4-Asnii5,LySi49-LyS150, Lysi50-Thr15i, Lys18i-Thr182,Lys!88-Alai89, and Lys197-Thr198—and therebyproduces about 14 peptides, 3 of which are po-tentially bitter.

Although plasmin is less active on asrcaseinthan on ocs2- and p-caseins, it hydrolyzes ocsl-casein in solution at the bonds Arg22-Phe23,ATg90-TyT9I, LyS102-LyS103, LyS103-TyT104, LyS105-VaI106, LyS124-GIu125, and Arg151-Gln152 (Mc-

Sweeney et al., 1993b). The formation of A,-casein, a minor casein component, has been at-tributed to its action on ocsl-casein.

Although K-casein contains several potentialsites, it is very resistant to plasmin, and the prod-ucts have not been identified. Even though thepH of cheese is quite far removed from the pHoptimum of plasmin, the hydrolysis of p-caseinin cheese is due mainly to plasmin. Being quiteheat stable, plasmin and plasminogen surviveHTST pasteurization and cheese cooking, andplasmin is principally responsible for primaryproteolysis in high-cooked cheeses, such as Par-mesan and Emmental.

Cathepsin D. The indigenous acid proteinasein milk, cathepsin D (EC 3.4.23.5), has receivedlittle attention. It is relatively heat labile (com-pletely inactivated by 7O0C x 10 min) and has a

Figure 11-12 Urea-PAGE of Na-caseinate hydrolyzed with different rennets (0.04 RU/ml) at pH 5.2 for 30 min.Lane 1: sodium casemate; lanes 2, 5, 8, 11, and 14: sodium casemate containing O, 1, 2.5, 5, and 10% NaClhydrolyzed by chymosin; lanes 3, 6, 9, 12, and 15: sodium casemate containing O, 1, 2.5, 5, and 10% NaClhydrolyzed by R. miehei proteinase; lanes 4, 7, 10, 13, and 16: sodium caseinate containing O, 1, 2.5, 5, and 10%NaCl hydrolyzed by C. parasitica proteinase.

Page 279: Cheese Science

pH optimum of 4.0. The specificity of cathepsinD on the caseins has not been determined, al-though electrophoretograms of hydrolyzates in-dicate that its specificity is similar to that ofchymosin; surprisingly, it is unable to coagulatemilk (McSweeney, Fox, & Olson, 1995).

The contribution of cathepsin to cheese ripen-ing is unclear but is very likely to be less thanthat of chymosin, which is present at a higherconcentration and has a similar specificity.

Other Indigenous Milk Proteinases. Thepresence of other proteolytic enzymes in milkhas been reported, including thrombin, a lysineaminopeptidase, and proteinases from leuco-cytes, but they are considered not to be signifi-cant in cheese (see Grufferty & Fox, 1988).

Proteolytic Enzymes from Starter

Although LAB are weakly proteolytic, theydo possess a proteinase and a wide range of pep-tidases, which are principally responsible for theformation of small peptides and amino acids incheese. The proteolytic system of Lactococcushas been studied thoroughly at the molecular,biochemical, and genetic levels. The system ofLactobacillus spp. is less well characterized butthe systems of both genera are generally similar.Sc. thermophilus is less proteolytic than Lacto-coccus or Lactobacillus and has been the subjectof little research. The extensive literature on theproteolytic systems of LAB has been compre-hensively reviewed by Kunji, Mierau, Hagting,Poolman, and Konings (1996); Law andHaandrikman (1997); Monnet, Chapot-Chartier,

and Gripon (1993); Tan, Poolman, and Konings(1993); Thomas and Pritchard (1987); andVisser(1993).

The proteinase in LAB is anchored to the cellmembrane and protrudes through the cell wall,giving it ready access to extracellular proteins;all the peptidases are intracellular. The oligo-peptides produced by the proteinase are activelytransported into the cell, where they are hydro-lyzed further by the battery of peptidases (seeChapter 5).

Cell envelope-associated proteinases (CEPs)of Lactococcus have been classified into threegroups: P1-, Pnr, and mixed-type proteinases. P1-type proteinases degrade (3- but not ocsi-casein ata significant rate, while Pm-type proteinases rap-idly degrade both asr and (3-caseins. The nucle-otide sequences of the genes for both P1- and P1n-type proteinases are very similar. Based on theirspecificity on ocsi-CN fl-23 (a peptide producedvery rapidly from ocsi-CN by chymosin), lacto-coccal proteinases can be classified into sevengroups, a-g (Figure 11-14).

The specificity of the CEPs from severalLactococcus strains on asr, as2-, P-, and K-caseins and short peptide substrates has been es-tablished (see Fox & McSweeney, 1996; Fox,O'Connor, McSweeney, Guinee, & O'Brien,1996; Fox, Singh, & McSweeney, 1994). Thelactococcal CEP is involved in the formation ofmany of the small peptides in cheese, and thepeptidases are responsible for the release ofamino acids.

Lactococcus spp. contain at least the follow-ing intracellular proteolytic enzymes:

Plasminogen(casein micelles)

Plasminogen activator(s)(casein micelles)

Inhibitors of plasminogenactivators

(milk serum)

Plasmin inhibitors(milk serum)

Plasmin(casein micelles)

Figure 11-13 Schematic representation of the plasmin system in milk.

Page 280: Cheese Science

Amino acid substitutions at positions relevant for substrate binding

763748747177166144142138131

Asn

Asn

Asn

His

Asn

His

His

Asn

Lys

Lys

Lys

Lys

Thr

Thr

Thr

Thr

Thr

Arg

Arg

Arg

Arg

Leu

Leu

Leu

GIn

GIy

Leu

Leu

He

Leu

Leu

Leu

He

Leu

Asn

Asp

Asp

Asp

Asp

Asp

Asp

Asp

VaI

VaI

Leu

Leu

Leu

Leu

Leu

Leu

Leu

He

Ala

Ala

Ala

Ala

Ser

Ala

Asp

Ala

Asp

Lys

Thr

Thr

Thr

Thr

Thr

Thr

Thr

GIy

Ser

Thr

Thr

Thr

Thr

Thr

Thr

Thr

Ser

Cleavage sites in 0,,-casein fragment 1-23Substr.

«.„ P, K

ex.,, p, K

a.,, P, K

P. K

0.,,P, K

P, K

Strains

L lactls

AM1.SK11.US3

AM2

E8

NCD0763, UC317

WG2.C13.KH

Z8, H61.TR, FD27

HP

Lb. paracasei

NCDO151

Lb. bulgaricus

NCDO1489

ZJb. helveticus

L89

Group

a

b

C

d

e

f

g

Figure 11-14 Classification of cell envelope-associated proteinases of lactic acid bacteria according to their specificity toward ocsl-casein fragment 1-23.

Page 281: Cheese Science

• Proteinases capable of hydrolyzing caseinswith a specificity different from that of theCEPs. Their activity in cheese has not beendemonstrated.

• Four endopeptidases (PepOi, PepO2, PepFb

PepF2) that are unable to hydrolyze intactcaseins but can hydrolyze internal peptidebonds in large casein-derived peptides (upto 30 amino acid residues). The specificityof these endopeptidases on several casein-derived peptides and on synthetic peptideshas been established, but their activity incheese has not been demonstrated.

• Four aminopeptides (PepN, PepA, PepC,PepP). PepN is a broad specificity metallo-enzyme, PepC is a broad specificity thiolenzyme, PepA is a metalloenzyme withhigh specificity for N-terminal GIu or Aspresidues, and PepP is a metalloenzyme thatreleases the N-terminal residue from pep-tides with Pro as the penultimate residue.

• An iminopeptidase (Pepl) that releases N-terminal Pro.

• A dipeptidyl aminopeptidase (PepX) thathas a high but not absolute specificity forpeptides with Pro at the penultimate posi-tion, releasing X-Pro dipeptides, where Xmay be one of several residues.

• A pyrrolidone carboxylyl peptidase (PCP)that releases a cyclic pyroglutamic acidfrom the N-terminal.

• A tripeptidase (PepT).• A number of dipeptidases, including a gen-

eral dipeptidase (PepV); PepL, which pref-erentially hydrolyzes dipeptides and sometripeptides containing an N-terminal Leu;and proline-specific dipeptidases, prolinase(PepR) and prolidase (PepQ), which hydro-lyze Pro-X and X-Pro dipeptidases, respec-tively.

Most of these peptidases have been isolatedfrom at least one strain of Lactococcus and char-acterized (Table 11-3).

The activity and stability of these peptidasesin cheese has not been established, but at leastsome are active, as indicated by the presence of

relatively high concentrations of certain pep-tides and amino acids in cheese. The proteolyticsystem is capable of hydrolyzing casein com-pletely to free amino acids. The sequential ac-tion of the peptidase system is shown schemati-cally in Figure 11-15. This complex proteolyticsystem is required by LAB for growth to highnumbers in milk that contains a low concentra-tion of small peptides and free amino acids.

Thermophilic obligately homofermentativeLactobacillus spp. (Lb. helveticus, Lb. del-brueckii spp. bulgaricus, Lb. delbrueckii ssp.acidophilus), alone or paired with Sc. ther-mophilus, are used as starters for high-cookedcheeses. The proteolytic system of the thermo-philic lactobacilli is generally similar to that ofthe lactococci and includes a CEP, the specific-ity of which has not been determined. Thesebacteria die and lyse relatively rapidly incheese (see Chapter 10), releasing intracellularpeptidases, which explains the high level ofamino acids in cheese made with thermophilicstarters.

Sc. thermophilus is weakly proteolytic (noproteinase has yet been isolated), but it pos-sesses substantial peptidase activity. Its contri-bution to proteolysis in cheese is probably lessthan that of the thermophilic lactobacilli, but de-finitive studies are lacking.

Proteolytic System ofNonstarterMicroflora

The starter cells, both Lactococcus and ther-mophilic Lactobacillus, reach maximum num-bers shortly after manufacture and then die offand lyse. In contrast, NSLAB grow from verylow initial numbers (< 50 cfu/g) to about 107-108

cfu/g within about 3 months and dominate themicroflora of long-ripened cheeses for most oftheir ripening period. As discussed in Chapter10, the interior of cheese is a hostile environ-ment for bacteria. It has a relatively low pH(=5), has a relatively high salt content (2-4%),lacks a fermentable carbohydrate, is anaerobic,and may contain bacteriocins produced bystarter bacteria. Hence, cheese is highly selec-tive, and the NSLAB microflora is dominated by

Page 282: Cheese Science

Table 11-3 Peptidases of Lactic Acid Bacteria

Optimal Activity

ClassSubunits0CpHMoI. Wt. (kDa)Principal Assay

SubstrateOrganism

Oligoendopeptidases (PepO, PepF, LEP, MEP, NOP)

MMMMNMMMM

121

>211111

4037

30-38423540

4047

7-7.5,6

6-6.5,8-96-77.56.08.07.7

499880701807070707070

PeptidesPeptidesO81-CN 11-23Met-enkephalin(X31-CNf 1-23O8I-CNf 1-23O81-CNf 1-23BradykininBradykininMet-enkephalin

Lc. lactis ssp. lactis CNRZ 267Lc. lactis ssp. cremoris H61Lc. lactis ssp. cremoris H61Lc. lactis ssp. cremoris Wg2Lc. lactis ssp. cremoris HPLc. lactis ssp. cremoris C1 3Lc. lactis ssp. lactis MG 1363Lc. lactis ssp. cremoris SK1 1Lc. lactis ssp. lactis NCDO763Lb. delbr. ssp. bulgaricus B14

Aminopeptidases

Aminopeptidase N (general aminopeptidase, AMP, PepN)

MM

M, -SHMMMMMMMMMMM

M1-SH

continues

111

11111111

354040

47.5

5037394050353635

6.577

6.2-7.2

7776

6.56.57.0

6.9-7.0

853695

78-913895979592879897899796

Lys-p-NALys-p-NALys-p-NALys-p-NALys-p-NALys-p-NALys-p-NALys-p-NALeu-p-NALeu-p-NALys-p-NALys-p-NALys-p-NALys-p-NALys-p-NA

Lc. bv. diacetylactis CNRZ 267Lc. lactis ssp. cremoris AC1Lc. lactis ssp. cremoris WG2

Lb. delbrueckii ssp. lactis 1 183Lb. acidophilus R-26Lb. delbr. ssp. bulgaricus CNRZ 397Lb. helveticus CNRZ 32Lb. delbrueckii ssp. bulgaricus B14Lb. helveticus LME-51 1Lb. easel ssp. case/ LLGLb. delbr. ssp. bulgar. ACA-DC233Lfc. helveticus ITGL1Sfr. sa/. ssp. thermophilus CNRZM 99Sc. sa/. ssp. thermophilus CNRZ302Sc. sa/. ssp. thermophilus NCDO573

Page 283: Cheese Science

Aminopeptidase A (glutamyl aminopeptidase, GAP, PepA)

50-55 3 M8 65 6 M8 50 -10 M

G M

130245520240

Glu-/Asp-p-NAGlu-p-NAGlu-p-NAAsp-p-NA

Lc. lactis ssp. cremoris HPLc. lactis ssp. lactis NCDO 712Lc. lactis ssp. cremoris HPLc. lactis ssp. cremoris AM2

Aminopeptidase C (thiol aminopeptidase, PepC)

-SH-SH

64

4050

76.5-7

300220

His-p-NALeu-Gly-Gly

Lc. lactis ssp. cremoris AM2Lb. delbrueckiissp. bulgaricus B14

Pyrrolidone carboxylyl peptidase (pyroglutamyl aminopeptidase, PCP)

S2378-8.580Pyr-p-NAPyr-p-NA

Lc. lactis ssp. cremoris HPLc. lactis ssp. cremoris HP

X-Prolyldipeptidyl aminopeptidase (XPDA, PPDA, XAP, PepX)

SSSSSSSSSSSSS

continues

22

21

223111

45-5040-45

50-554050454550

46-505050

78.56-96-9777

6.56.56.57.06.57.0

1801901171501657282

170-200170-200

270959079

X-Pro-p-NAAla-Pro-p-NAGly-Pro-NH-MecX-Pro-p-NAX-Pro-p-NAX-Pro-p-NAX-Pro-p-NAAla-Pro-p-NAAla-Pro-p-NAGly-Pro-p-NAAla-Pro-p-NAGly-Pro-p-NAGly-Pro-p-NA

Lc. lactis ssp. cremoris P8-2-47Lc. lactis ssp. lactis NCDO 763Lc. lactis ssp. cremoris AM2Lc. lactis ssp. lactis H1Lb. delbrueckii ssp. lactisLb.helveticusCNRZ32Lb. delbr. ssp. bulgaricus CNRZ 397Lb. delbrueckii ssp. bulgaricus B1 4Lb. acidophilus 357Lb. delbr. ssp. bulgaricus LBU- 147Lb. delbr. ssp. lactis DSM7290Lb.helveticusLHE-5MLb. easel ssp. case/ LLG

Page 284: Cheese Science

Table 11 -3 continued

Optimal Activity

ClassSubunits0CpHMoI. Wt. (kDa)Principal Assay

SubstrateOrganism

MS

-SH

MMMMM

MM

MMM

-SHM

231

1

11

223

>1

374040

30

5050

3555

3340

8.56-77.5

77.5887

6.5-7.57.35-9.0

77.58.65.86.0

10010046

25,34511004951

4342

75103-105

1057285

Pro-Gly-GlyPro-p-NAPro-AMC

DipeptidesLeu-LeuLeu-GlyDipeptidesDipeptides

Leu-ProLeu-Pro

TripeptidesLeu-Leu-LeuTripeptidesLeu-Leu-LeuLeu-Gly-Gly

Proline iminopeptidase (PIP, Pepl)

Pr. freud. ssp. shermanii 13673Lc. lactis ssp. cremoris HPLb. delbr. ssp. bulgar. CNRZ 397Lb. case/ ssp. easel LLG

Dipeptidases (DIP, PepV, PepD)

Lactococcus spp.Lc. bv. diacetylactis CNRZ 267Lc. lactis ssp. cremoris H61Lc. lactis ssp. cremoris Wg2Lb. delbr. ssp. bulgaricus B14

Prolidase (PRD, PepQ)

Lc. lactis ssp. cremoris H61Lc. lactis ssp. cremoris AM2

Tripeptidases (TRP, PepT)

Lc. bv. diacetylactis CNRZ 267Lc. lactis ssp. cremoris Wg2Lc. lactis ssp. cremoris AM2Lc. lactis ssp. cremor/slMN-C12Lb. delbrueckiissp. bulgaricus B1 4

Key: AMC = aminomethyl coumarin; M = metallo; p-NA = p-nitroanalide; S = serine; -SH = thiol.

Page 285: Cheese Science

Figure 11-15 Schematic representation of (a) the hydrolysis of casein by lactococcal cell envelope-associatedproteinase (CEP) and (b) the degradation of a hypothetical dodecapeptide by the combined action of lactococcalpeptidases: oligopeptidase (PepO), various aminopeptidases (PCP, PepN, Pep A, PepX), tripeptidase (PepT),prolidase (PepQ), and dipeptidase (PepV).

Leu-Leu

PepV

Leu-Leu-Leu

PepTPepQ

Gly-Pro

Gly-Pro-Leu-Leu-Leu

PepXGlx-Gly-Pro-Leu-Leu-Leu

PepA

Ala-Glx-Gly-Pro-Leu-Leu-Leu

PepN

Lys-Ala-Glx-Gly-Pro-Leu-Leu-Leu

PepN

/ryro-Glu-Lys-Ala-Glx-Gly-Pro-Leu-Leu-Leu

PCP

(b)#yr0-Glu-Lys-Ala-Glx-Gly-Pro-Leu-Leu-Leu-Pro-His-Phe

PepO

His^Phe

PepV

Pro-THis-Phe

Pepl

PEPTIDES

CEP

(a) CASEINS, CASEIN-DERIVED PEPTIDES

Page 286: Cheese Science

a few species of mesophilic lactobacilli (Lb.casei, Lb. paracasei, Lb. plantarum, and Lb.curvatus). Some authors have reported the pres-ence of pediococci in cheese, but recent studieshave failed to find them in significant numbers.

The proteolytic system of these mesophiliclactobacilli is not as well studied as that of thelactococci or thermophilic lactobacilli. Sincethey do not grow well in milk in the absence ofan added source of small peptides or amino ac-ids, they may lack a CEP. Also, the method usedto release CEPs from lactococci—washing cellswith a calcium-free buffer—fails to releaseCEPs from mesophilic lactobacilli. These latterbacteria do possess a range of intracellular pepti-dases, but few of these have been studied.

NSLAB appear to contribute little to primaryproteolysis in Cheddar cheese but do contributeto the release of free amino acids (see Fox,McSweeney, & Lynch, 1998). As discussed inChapter 15, mesophilic lactobacilli have beenused as adjunct starters in Cheddar cheese, inwhich they are reported to modify and perhapsimprove flavor.

There are few reports on the proteolytic activ-ity of pediococci (see Fox et al., 1996) and theirsignificance in cheese ripening is unknown.

Proteinases from Secondary Starter

Most cheese varieties have a secondary mi-croflora, the function of which is other than acidproduction. Originally, this microflora was ad-ventitious, and its development occurred as a re-sult of selective conditions (e.g., pH, humidity,temperature, and aw). Today, the secondary mi-croflora may be adventitious, but in many casesthe milk or curd is inoculated with selected mi-croorganisms. The principal secondary microor-ganisms are Penicillium roqueforti (Blue moldcheese); P. camemberti (surface mold cheese,such as Camembert and Brie); Brevibacteriumlinens, Arthrobacter, and other coryneform bac-teria (surface smear-ripened cheese); Propioni-bacterium freudenreichii subsp. shermanii(Swiss-type cheese); and several species ofyeasts (see Chapter 10). Most of these microor-ganisms are metabolically very active and con-

sequently may dominate the ripening of cheesesin which they occur.

P. roqueforti and P. camemberti secrete as-partyl and metalloproteinases, which have beenfairly well characterized, including their speci-ficity on OC8I- and (3-caseins (see Gripon, 1993).Intracellular acid proteinase(s) and exopepti-dases (amino and carboxy) are also produced byP. roqueforti and P. camemberti but have notbeen well studied (see Gripon, 1993). Tri- anddipeptidases and proline-specific peptidasesfrom P. roqueforti and P. camemberti do not ap-pear to have been studied.

The proteinase of Br. linens ATCC 9174 hasbeen purified and characterized, including itsspecificity on as!- and p-casein (see Rattray &Fox, 1998). An extracellular aminopeptidaseand an intracellular aminopeptidase have alsobeen purified.

Propionibacterium spp. are weakly proteo-lytic but strongly peptidolytic. They are particu-larly active on proline-containing peptides, andconsequently Swiss-type cheeses contain highconcentrations of proline, which may contributeto the characteristic flavor of these cheeses.Aminopeptidase, iminopeptidase, and X-prolyldipeptidylaminopeptidase has been isolatedfrom at least one strain of P. freudenreichiisubsp. shermanii (see Fernandez-Espla & Fox,1997).

11.8 CHARACTERIZATION OFPROTEOLYSIS IN CHEESE

The extent of proteolysis in cheese rangesfrom very limited (e.g., Mozzarella) to very ex-tensive (e.g., Blue-mold varieties). PAGE showsthat the proteolytic pattern, as well as its extent,exhibits marked intervarietal differences; thePAGE patterns of both the water-insoluble andwater-soluble fractions are, in fact, quite charac-teristic of the variety, as shown in Figures 11-16and 11-17. RP-HPLC of the water-soluble frac-tion or subfractions thereof also show varietalcharacteristics (Figures 11-18 and 11-19). Boththe PAGE and HPLC patterns vary and becomemore complex as the cheese matures, and they

Page 287: Cheese Science

are in fact very useful indices of cheese maturityand to a lesser extent of cheese quality. There-fore, they have potential as tools in the objectiveassessment of cheese quality. Urea-PAGE pat-terns of Cheddar cheeses at various stages ofmaturity are shown in Figure 11-20.

Complete characterization of proteolysis incheese requires isolation and identification of theindividual peptides. A comprehensive fraction-ation protocol is shown in Figure 11—21. Many ofthe water-insoluble and water-soluble peptides inCheddar cheese have been isolated and identifiedby amino acid sequencing and mass spectrom-etry; these are summarized in Figures 11-22 and

11-23. All the principal water-insoluble peptidesare produced either from ocsi-casein by chymosinor from (3-casein by plasmin and represent the C-terminal fragments of these proteins (Figure11-22). In mature Cheddar (> 6 months old), allof the ocsi-casein is hydrolyzed by chymosin atPhC23-PhC24. The peptide ocsl-CN fl-23 does notaccumulate but is hydrolyzed rapidly at GIn9-Glyio, GInn-GlUi4, and/or LeUi6-ASnn by thelactococcal cell wall proteinase, depending on itsspecificity (see Figure 11-14). A significantamount of the larger peptide (O8I-CN f24-199) ishydrolyzed at Leui0i-Lysi02. In 6-month-oldCheddar, about 50% of the (J-casein is hydro-

Figure 11-16 Urea-PAGE of the water-insoluble fraction of a selection of cheese varieties: 1, Na-caseinate; 2-4, Cheddar; 5-7, Emmental; 8, Maasdamer; 9, Jarlsberg; 10-12, Edam; 13-15, Gouda.

Page 288: Cheese Science

lyzed, mainly by plasmin, to y-casein ((3-CN £29-209, f 105-209, and f 107-209) and proteose pep-tones (p-CN fl-28, fl-104, fl-106, £29-104, £29-106). These polypeptides do not appear to behydrolyzed by chymosin or lactococcal protein-ase. Although as2-casein gradually disappearsfrom PAGE patterns of cheese during ripening,few polypeptides produced from it have beenidentified. Para-K-casein is quite resistant to pro-teolysis, and no peptides produced from it havebeen identified.

Most of the water-soluble peptides are de-rived from the N-terminal half of ocsl- and (3-caseins (Figure 11-23). The N-terminal of manyof these peptides corresponds to a chymosin(Osi-CN) or plasmin (|3-CN) cleavage site, but

some appear to arise from the action of the lac-tococcal CEP. However, the N-terminal and es-pecially the C-terminal of many peptides do notcorrespond precisely to the known cleavage sitesof chymosin, plasmin, or lactococcal proteinase.The discrepancy in N-terminal suggests the ac-tion of bacterial aminopeptidases. Carboxypep-tidase activity would explain why the C-terminalof many peptides does not correspond to knownproteinase cleavage sites, but this activity hasnot been reported in Lactococcus spp. It must bepresumed that another proteinase, perhaps fromNSLAB or endopeptidases (PepO or PepFtypes) from starter and NSLAB, are involved, orperhaps other cleavage sites for lactococcal cellwall proteinase remain to be identified.

Figure 11-17 Urea-PAGE of the water-soluble fraction of a selection of cheese varieties: 1, Na-caseinate; 2—4,Cheddar; 5-7, Emmental; 8, Maasdamer; 9, Jarlsberg; 1-12, Edam; 13-15, Gouda.

Page 289: Cheese Science

The N-terminal sequence of asrCN fl-9 andfl-13 is RPKHPIK, which should be susceptibleto PepX. The accumulation of these peptides inCheddar and the apparent absence of peptides

with a sequence commencing at Lys3 of otsl-CNsuggest that PepX is not active in cheese.

A number of authors have shown that the verysmall peptides (< 500 Da) make a significant

Figure 11-18 RP-HPLC profiles of the 70% ethanol-soluble fractions of Cheddar (1), Parmesan (2), Emmental(3), Leerdammer (4), Edam (5), and Gouda (6) cheese.

Page 290: Cheese Science

contribution to Cheddar flavor but only a few ofthese peptides have been identified.

A large number of 12% TCA soluble and in-soluble peptides in the water-soluble extract of

Parmesan have been identified by fast atombombardment mass spectrometry (Addeo,Chianese, Sacchi, et al., 1994; Addeo, Chianese,Salzano, et al., 1992). Although Parmesan un-

Figure 11-19 RP-HPLC profiles of the 70% ethanol-insoluble fractions of Cheddar (1), Parmesan (2),Emmental (3), Leerdammer (4), Edam (5), and Gouda (6) cheese.

Page 291: Cheese Science

dergoes extensive proteolysis and has a veryhigh concentration of free amino acids, it con-tains low concentrations of medium-size pep-tides.

Although very extensive proteolysis occurs inBlue cheeses, and some of the larger peptidesdetectable by PAGE have been partially identi-fied (see Gripon, 1993), very little work hasbeen done on the small pH 4.6-soluble peptides.Some of the peptides resulting from the cleavageof OC8I-CN fl-23 (produced by chymosin) bylactococcal CEP have been identified in Gouda.

Proteolysis in Swiss-type cheeses has been stud-ied using PAGE and RP-HPLC, but small pep-tides have not been isolated and characterized.

Significant concentrations of amino acids, thefinal products of proteolysis, occur in all cheesesthat have been investigated (see Fox & Wallace,1997). Relative to the level of water-soluble N,Cheddar contains a low concentration of aminoacids (see Figure 12-14). The principal aminoacids in Cheddar are GIu, Leu, Arg, Lys, Phe,and Ser (Figure 11-24) (see Fox & Wallace,1997, for a comprehensive compilation of data

Figure 11-20 Urea-polyacrylamide gel electrophoretograms of Cheddar cheese after ripening for 0,1,2, 3,4, 5,6, 8, 10, 12, 14, 16, 18, and 20 weeks (lanes 1-14). C, sodium casemate.

Page 292: Cheese Science

for amino acids, in various cheeses). Parmesancontains a very high concentration of amino ac-ids, which appear to make a major contributionto the characteristic flavor of this cheese. Thepresence of amino acids in cheeses clearly indi-cates aminopeptidase activity. Since these en-zymes are intracellular, their action indicateslactococcal cell lysis. Based on the presumptionthat amino acids contribute to cheese flavor, asearch is now on for fast-lysing lactococcalstrains susceptible to heat-, phage-, or bacterio-cin-induced lysis (see Chapter 15). Amino acidshave characteristic flavors (see Chapter 12), andalthough none has a cheeselike flavor, it is be-

lieved that they contribute to the savory taste ofmature cheese.

11.9 CATABOLISM OF AMINO ACIDSAND RELATED EVENTS

Catabolism of free amino acids probablyplays some role in all cheese varieties but is par-ticularly significant in mold- and smear-ripenedvarieties. Catabolism involves decarboxylation,deamination, transamination, desulfuration, andhydrolysis of amino acid side chains leading tothe production of a wide array of compounds,

Figure 11-21 Scheme for the fractionation of cheese nitrogen.

RP-HPLCRP-HPLC

Urea-PAGEElectroblotting SepPakC8orC18

Amino acids

FractionsI -Vm

DEAE-cellulose

Permeate

Sephadex G-25

Fractionsi, n, HI, iv, v, vi, vn, vm, DC

Retentate

WISN

Re extract as above UF, 10 kDa cut off membranes

Fat,WSEWISN

Centrifuge (10,000 g x 30 min)

Water: Cheese, 2:1Homogenize (Stomacher or similar apparatus)

Grated Cheese

Page 293: Cheese Science

including carboxylic acids, amines, NH3, CO2,aldehydes, alcohols, thiols, and other sulfurcompounds, phenols, and hydrocarbons. Gen-eral pathways of amino acid catabolism are sum-marized in Figure 11-25. The catabolism ofamino acids has been reviewed by Fox andWallace (1997); Hemme, Bouillanne, Metro,and Desmazeaud (1982); and Law (1987).

Decarboxylation involves the conversion ofamino acid to the corresponding amine, with theloss OfCO2. The presence of primary amines incheese can be explained in terms of simple de-carboxylation, although the formation of sec-ondary and tertiary amines is more difficult toexplain. The principal amine in cheese istyramine. A number of amines produced incheese are biologically active (see Chapter 21).

Deamination results in the formation of NH3

and a-ketoacids. Ammonia is an important con-stituent of many cheeses, such as Camembert,Gruyere, and Comte. Ammonia can also beformed by oxidative deamination of amines,

yielding aldehydes. Transamination results inthe formation of other amino acids by the actionof transaminases. Aldehydes formed by theabove processes can then be oxidized to acids orreduced to the corresponding alcohols.

Amino acid side chains can also be modifiedin cheese. Hydrolases can release ammonia fromAsn and GIn or by the partial hydrolysis of theguanidino group of Arg, forming citrulline or,through degradation, ornithine. Phenol and in-dole can be produced by the action of C-C lyaseson Tyr and Trp.

Volatile sulfur compounds, including hydro-gen sulfide (H2S), dimethylsulfide [(CH3)2-S],dimethyldisulfide (CH3-S-S-CH3), and methan-ethiol (CH3SH), are found in most cheeses andcan be important flavor constituents. Sulfur-con-taining compounds are produced mainly frommethionine, since Cys is rare in the caseins (itoccurs at low levels in only ocs2- and K-caseins,which are not extensively hydrolyzed in cheese).Methanethiol and related compounds are

Figure 11-22 Schematic representation of the principal water-insoluble peptides isolated from Cheddar cheeseand identified. ocsl = casein (A), (3 = casein (B).

Page 294: Cheese Science

Figure 11-23 Water-soluble peptides derived from ocsl

Cheddar cheese and identified. The principal chymosin,age sites are indicated by arrows.

;asein (A), ocs2-casein (B), and p-casein (C) isolated from3lasmin, and lactococcal cell-envelope proteinase cleav-

C

B

A

Page 295: Cheese Science

Figure 11-24 Typical concentrations of amino acids in Cheddar, Gouda, Emmental, and Parmigiano-Reggiano.

Cys Asp Thr Ser GIu Pro GIy Ala VaI Met lie Leu Tyr Phe His Lys Arg

EmmentalParmigiano

CheddarGouda

Page 296: Cheese Science

thought to be particularly important in the flavorof Cheddar cheese.

Although enzymes capable of catalyzing mostof the catabolic reactions described above havebeen identified, few of them have been isolatedand characterized from cheese-related microor-ganisms, especially LAB. Perhaps this is be-cause more complex assay methods are neededto detect these enzymes when present at low lev-els. Since the products of amino acid catabolismprobably contribute to the finer points of cheeseflavor, it is expected that research will focus onthis area in the immediate future.

11.10 CONCLUSION

Cheese ripening involves a very complex se-ries of biochemical reactions catalyzed by livingmicroorganisms or by enzymes from severalsources. The primary events—glycolysis, Ii-polysis, and proteolysis—have been describedrather thoroughly. The fermentation of lactose,mainly to lactic acid, is caused by living micro-organisms, principally the starter culture (or ad-ventitious bacteria in the case of artisanalcheeses). The fermentation of lactose has beendescribed thoroughly. Lipolysis is quite limitedin most cheese varieties, and in those varieties inwhich it is important, lipolysis has been wellcharacterized in terms of extent and the enzymesinvolved. The initial steps in proteolysis and theenzymes responsible have been established forthe principal varieties. Secondary proteolysis ischaracterized to some extent in a few varieties,but proteolysis is so complex and variable, bothbetween and within varieties, that it is probablynot possible to characterize it in full detail.

Many of the enzymes responsible for primaryripening have been isolated and characterized.However, the stability and activity of these en-zymes in the cheese environment have receivedlittle attention, although they appear to warrantresearch. The primary reactions are probably re-sponsible for changes in cheese texture, such asan increase in pH (due to the catabolism of lacticacid and/or production OfNH3 from deaminationof amino acids) or hydrolysis of the protein ma-

trix. With the exception of fatty acids, the prod-ucts of primary reactions are relatively minorcontributors to cheese flavor.

Modifications of the primary products of gly-colysis and lipolysis are fairly well character-ized. The catabolism of fatty acids to methyl ke-tones via (3-oxidation and decarboxylation is amajor contributor to the characteristic flavor ofBlue-mold cheeses but is not significant in mostvarieties. The catabolism of lactic acid has a mi-nor influence on the flavor of most cheeses. Anexception is Swiss-type cheeses, but even inthese cheeses the catabolism of lactic acid is lessimportant for the flavor than for the productionof CO2 for eye formation. The catabolism ofamino acids is the least well characterized aspectof cheese ripening. It is very likely that the prod-ucts of amino acid catabolism are major con-tributors to the flavor of many cheese varieties.The ammonia produced in many of these reac-tions contributes to the pH of cheese during rip-ening, and this change in pH affects the textureof the cheese and probably affects the stabilityand activity of many enzymes, which in turnprobably influence flavor development. It isvery likely that future research on cheese ripen-ing will focus on amino acid catabolism.

Since the biochemistry of cheese ripening isresponsible for its flavor, texture, and appear-ance, elucidation of these biochemical reactionsis clearly a prerequisite for controlling andmodifying cheese ripening. Such knowledge isessential for the selection of primary and sec-ondary cultures and for their genetic modifica-tion. Without this knowledge, selection of startercultures will be empirical.

The stability and activity of microorganismsand enzymes in cheese depend on its composi-tion. Although the composition of cheese pro-duced in modern factories using modern tech-nology is controlled within quite narrow limits,the quality of the resultant cheese is somewhatvariable, even when the cheese is made frompasteurized milk essentially free of indigenousbacteria and using high-quality rennet andstarter. This suggests that slight variations incurd composition are important, perhaps owing

Page 297: Cheese Science

to their effect on the stability and activity of keyenzymes. To date, most studies on these aspectsof cheese ripening have been performed onmodel systems that are well controlled andoversimplified.

REFERENCES

Addeo, F., Chianese, L., Sacchi, R., Musso, S.P., Ferranti, P.,& Molorni, A. (1994). Characterization of the oligo-peptides of Parmigiano-Reggiano cheese soluble in 12Ogtrichloroacetic acid/1. Journal of Dairy Research, 61,365-374.

Addeo, F., Chianese, L., Salzano, A., Sacchi, R., Cappuccio,U., Ferranti, P., & Molorni, A. (1992). Characterization ofthe 12% trichloroacetic acid insoluble oligopeptides ofParmigiano-Reggiano cheese. Journal of Dairy Research,59, 401-411.

Bastian, E.D., & Brown, RJ. (1996). Plasmin in milk anddairy products [An update]. International Dairy Journal,6, 435^57.

Bhowmik, T., & Marth, E.H. (1989). Esterolytic activities ofPediococcus species. Journal of Dairy Science, 72,2869—2872.

Bhowmik, T., & Marth, E.H. (199Oa). Esterases of Micro-coccus species: Identification and partial characterization.Journal of Dairy Science, 73, 33^40.

Bhowmik, T., & Marth, E.H. (199Ob). Role of Micrococcusand Pediococcus species in cheese ripening [A review].Journal of Dairy Science, 73, 859-866.

Charton, E., Davies, C., & McCrae, A.R. (1992). Use of spe-cific polyclonal antibodies to detect heterogeneous Ii-pases from Geotrichum candidum. Biochimica Biophys-icaActa,ll27, 191-198.

Chich, J.-F., Marchesseau, K., & Gripon, J.-C. (1997). Inter-cellular esterase from Lactococcus lactis subsp. lactisNCDO 763: Purification and characterization. Interna-tional Dairy Journal, 7, 169-174.

Cogan, T.M., & Hill, C. (1993). Cheese starter cultures. InP.F. Fox (Ed.), Cheese: Chemistry, physics and microbi-ology (2d ed., Vol. 1). London: Chapman & Hall.

El-Soda, M., Abd El-Wahab, H., Ezzat, N., Desmazeaud,M.J., & Ismail, A. (1986). The esterolytic and lipolyticactivities of the lactobacilli. II. Detection of esterase sys-tem of Lactobacillus helveticus, Lactobacillus bulgari-cus, Lactobacillus lactis and Lactobacillus acidophilus.Lait, 66, 431^43.

Fernandez-Espla, M.D., & Fox, P.F. (1997). Purification andcharacterization of X-prolyl dipeptidyl aminopeptidasefrom Propionibacterium shermani NCDO 853. Interna-tional Dairy Journal, 7, 23-29.

Very considerable advances in the biochemis-try of cheese ripening have been made duringthe past 20 years. The general features have beenelucidated, but the details remain to be estab-lished.

Fox, P.F. (1989). Proteolysis during cheese manufacture andripening. Journal of Dairy Science, 72, 1379-1400.

Fox, P.P., Law, J., McSweeney, P.L.H., & Wallace, J.(1993). Biochemistry of cheese ripening. In P.F. Fox(Ed.), Cheese: Chemistry, physics and microbiology (2ded., Vol. 1). London: Chapman & Hall.

Fox, P.F., Lucey, J.A., & Cogan, T.M. (1990). Glycolysisand related reactions during cheese manufacture and rip-ening. CRC Critical Reviews in Food Science and Nutri-tion, 29, 237-253.

Fox, P.F., & McSweeney, P.L.H. (1996). Proteolysis incheese during ripening. Food Reviews International, 12,457-509.

Fox, P.P., McSweeney, P.L.H., & Lynch, C.M. (1998). Sig-nificance of non-starter lactic acid bacteria in Cheddarcheese. Australian Journal of Dairy Technology, 53, 83-89.

Fox, P.P., McSweeney, P.L.H., & Singh, T.K. (1995). Meth-ods for assessing proteolysis in cheese during ripening. InE.L. Malin & M.H. Tunick (Eds.), Chemistry of struc-ture-function relationships in cheese. New York: PlenumPress.

Fox, P.P., O'Connor, T.P., McSweeney, P.L.H., Guinee,T.P., & O'Brien, N.M. (1996). Cheese: Physical, bio-chemical and nutritional aspects. Advances in Food andNutrition Research, 39, 163-328.

Fox, P.F., Singh, T.K., & McSweeney, P.L.H. (1994). Pro-teolysis in cheese during ripening. In A.T. Andrews & J.Varley (Eds.), Biochemistry of milk products. Cam-bridge: Royal Society of Chemistry.

Fox, P.P., & Stepaniak, L. (1993). Enzymes in cheese tech-nology. International Dairy Journal, 3, 609.

Fox, P.P., & Wallace, J.M. (1997). Formation of flavourcompounds in cheese. Advances in Applied Microbiol-ogy, 45, 17-85.

Gobbetti, M., Fox, P.F., Smacchi, E., Stepaniak, L., &Damiani, P. (1997). Purification and characterization of alipase from Lactobacillus plantarum 2739. Journal ofFood Biochemistry, 20, 227-246.

Gobbetti, M., Fox, P.F., & Stepaniak, L. (1996). Esterolyticand lipolytic activities of mesophilic and thermophiliclactobacilli. Italian Journal of Food Science, 8, 127-136.

Page 298: Cheese Science

Gobbetti, M., Fox, P.P., & Stepaniak, L. (1997). Isolationand characterization of a tributyrin esterase from Lacto-bacillus plantarum 2739. Journal of Dairy Science, 80,3099-3106.

Grappin, R., Rank, T.C., & Olson, N.F. (1985). Primary pro-teolysis of cheese proteins during ripening. Journal ofDairy Science, 68, 531-540.

Gripon, J.-C. (1987). Mould-ripened cheeses. In P.P. Fox(Ed.), Cheese: Chemistry, physics and microbiology(Vol. 2). London: Elsevier Science Publishers.

Gripon, J.-C. (1993). Mould-ripened cheeses. In P.P. Fox(Ed.), Cheese: Chemistry, physics and microbiology (2ded., Vol. 2). London: Chapman & Hall.

Grufferty, M.B., & Fox, P.P. (1988). Milk alkaline protein-ase [A review]. Journal of Dairy Research, 55, 609-630.

Hemme, D., Bouillanne, C., Metro, P., & Desmazeaud, MJ.(1982). Microbial catabolism of amino acids duringcheese ripening. Science des Aliments, 2, 113-123.

Holland, R., & Coolbear, T. (1996). Purification of tributyrinesterase from Lactococcus lactis subsp. cremoris E8.Journal of Dairy Research, 63, 131-140.

Huffman, L.M., & Kristoffersen, T. (1984). Role of lactosein Cheddar cheese manufacturing and ripening. NewZealand Journal of Dairy Science and Technology, 19,151-162.

International Dairy Federation. (1991). Chemical methodsfor evaluation ofproteolysis in cheese maturation [Bulle-tin No. 261]. Brussels: Author.

Karahadian, C., & Lindsay, R.C. (1987). Integrated roles oflactate, ammonia, and calcium in texture development ofmold surface-ripening cheese. Journal of Dairy Science,70,909-9IS.

Khalid, N.M., El-Soda, M., & Marth, E.H. (1990). Esteraseof Lactobacillus helveticus and Lactobacillus delbrueckiissp. bulgaricus. Journal of Dairy Science, 73, 2711—2719.

Kinsella, J.E., & Hwang, D.H. (1976). Enzymes ofPenicil-lium roqueforti involved in the biosynthesis of cheese fla-vor. CRC Critical Reviews of Food Science and Nutri-tion, 8, 191-228.

Kunji, E.R.S., Mierau, I., Hagting, A., Poolman, B., &Konings, W.N. (1996). The proteolytic system of lacticacid bacteria. Antonie van Leeuwenhoek, 70, 187-221.

Law, B.A. (1987). Proteolysis in relation to normal and ac-celerated cheese ripening. In P.P. Fox (Ed.), Cheese:Chemistry, physics and microbiology (Vol. 1). London:Elsevier Science Publishers.

Law, J., & Haandrikman, A. (1997). Proteolytic enzymes oflactic acid bacteria. International Dairy Journal, 7, 1—11.

Lenoir, J. (1984). The surface flora and its role in the ripen-ing of cheese [Bulletin No. 171]. Brussels: InternationalDairy Federation.

McSweeney, P.L.H., & Fox, P.P. (1993). Cheese: Methodsof chemical analysis. In P.P. Fox (Ed.), Cheese: chemis-try, physics and microbiology (2d ed., Vol. 1). London:Chapman & Campbell.

McSweeney, P.L.H., & Fox, P.P. (1997). Chemical methodsfor the characterization of proteolysis in cheese duringripening. LaIt9 77, 41-76.

McSweeney, P.L.H., Fox, P.P., & Olson, N.F. (1995). Pro-teolysis of bovine casein by cathepsin D: Preliminary ob-servations and comparison with chymosin. InternationalDairy Journal, 5, 321-336.

McSweeney, P.L.H., Olson, N.F., Fox, P.P., & Healy, A.(1994). Proteolysis of bovine ocs2-casein by chymosin. Z.Zeitschrift fur Lebensmittel Untersuchung undForschung 119, 429^32.

McSweeney, P.L.H., Olson, N.F., Fox, P.P., Healy, A., &Hojrup, P. (1993a). Proteolytic specificity of chymosinon bovine ocsi-casein. Journal of Dairy Research, 60,401-412.

McSweeney, P.L.H., Olson, N.F., Fox, P.P., Healy, A., &Hojrup, P. (1993b). Proteolytic specificity of plasmin onbovine ocsl-casein. Food Biotechnology, 7, 143-158.

Monnet, V., Chapot-Chartier, M.P., & Gripon, J.-C. (1993).Les peptidases des lactocoques. LaIt9 73, 97-108.

Mottar, J.F. (1989). Effect on the quality of dairy products.In R.C. McKellar (Ed.), Enzymes ofpsychrotrophs of rawfoods. Boca Raton, FL: CRC Press.

Nelson, J.H., Jensen, R.G., & Pitas, R.E. (1977). Pregastricesterase and other oral lipases [A review]. Journal ofDairy Science, 60, 327-362.

Olivecrona, T., & Bengtsson-Olivecrona, G. (1991). Indig-enous enzymes in milk: Lipase. In P.P. Fox (Ed.), FoodEnzymology (Vol. 1). London: Elsevier Science Publish-ers.

Olivecrona, T., Vilaro, S., & Bengtsson-Olivecrona, G.(1992). Indigenous enzymes in milk. II. Lipases in milk.In P.P. Fox (Ed.), Advanced Dairy Chemistry (Vol. 1).London: Elsevier Science Publishers.

Oterholm, A., Ordal, Z.J., & Witter, L.D. (1970). Purifica-tion and properties of glycerol ester hydrolase (lipase)from Propionibacterium shermanii. Applied Microbiol-ogy, 20, 16-22.

Rank, T.C., Grappin, R., & Olson, N.F. (1985). Secondaryproteolysis of cheese during ripening [A review]. Journalof Dairy Science, 68, 801-805.

Rattray, P.P., & Fox, P.P. (1997). Purification and character-ization of an intracellular esterase from Brevibacteriumlinens ATCC 9174. International Dairy Journal, 7, 273-278.

Rattray, P.P., & Fox, P.P. (1998). Recently identified en-zymes of Brevibacterium linens ATCC 9174 [A review].Journal of Food Biochemistry, 22, 353-373.

Page 299: Cheese Science

Sidebottom, C.M., Charton, E., Dunn, P.P.J., Mycock, G.,Davies, C., Sutton, J.L., MacCrae, A.R., & Slabas, A.R.(1991). Geotrichum candidum produces several Upaseswith markedly different substrate specificities. EuropeanJournal of Biochemistry, 202, 485-491.

Tan, P.S.T., Poolman, B., & Konings, W.N. (1993). Pro-teolytic enzymes of Lactococcus lactis. Journal of DairyResearch, 60, 269-286.

Thomas, T.D., & Crow, V.L. (1983). Mechanism of D(->lactic acid formation in Cheddar cheese. New ZealandJournal of Dairy Science and Technology, 18, 131-141.

Thomas, T.D., & Pritchard, G.C. (1987). Proteolytic en-zymes in dairy starter cultures. Federation of EuropeanMicrobiological Associations Microbiology Review, 46,245-268.

Tzanetakis, N., & Litopoulou-Tzanetaki, E. (1989). Bio-chemical activities of Pediococcus pentosaceus isolatesof dairy origin. Journal of Dairy Science, 72, 859-863.

Visser, S. (1993). Proteolytic enzymes and their relation tocheese ripening and flavor [An overview]. Journal ofDairy Science, 76, 329-350.

Page 300: Cheese Science

12.1 INTRODUCTION

The quality of cheese is determined by its fla-vor (taste and aroma), texture (hardness, crum-bliness, cohesiveness, stretchability, sliceability,etc.), and appearance (color, uniformity, eyesand other fissures, and presence or absence ofmolds). When a consumer selects cheese, its ap-pearance is the first and perhaps the only crite-rion of quality applied. For example, an Em-mental without eyes, a Gouda with patches ofmold, or a Roquefort without mold will not bepurchased. Today, cheese offered for sale atreputable outlets is unlikely to be defective inappearance. The relative importance of flavorand texture depends on the variety. For some va-rieties, such as Mozzarella, which has a verymild flavor, the textural attributes of meltabilityand stretchability are paramount, whereas flavoris the most important characteristic of Bluecheese varieties. While the nutritional and safetyaspects of cheese are essential, most consumers,at least in developed countries, consume cheesemainly for its sensory attributes, which are,therefore, the raison d'etre of cheesemaking.

Various aspects of cheese texture are dis-cussed in Chapter 13. Aspects of the flavor ofcheese are discussed in this chapter.

Cheese flavor has been the subject of scien-tific investigation since the beginning of thiscentury. Initially, it was believed that a singlecompound or small group of compounds might

be responsible for cheese flavor, but as data ac-cumulated it became apparent that this was notso and that a large number of sapid and aromaticcompounds are present in cheese. This led to the"component balance theory," developed in the1950s, which proposed that cheese flavor resultsfrom the correct balance and concentration ofnumerous sapid and aromatic compounds. Dur-ing the intervening 40 years, there has been ex-tensive research on the flavor of several cheesevarieties, but complete information is not yetavailable on the flavor chemistry of any specificvariety. The extensive literature on cheese flavorhas been reviewed by Adda, Gripon, and Vassal(1982); Aston and Dulley (1982); Bosset andGauch (1993); Fox, Singh, and McSweeney(1995); Fox and Wallace (1997); Imhof andBosset (1994); Kristoffersen (1973, 1985);McGugan (1975); McSweeney, Nursten, andUrbach (1997); Olson (1990); Reiter, Fryer,Sharpe, and Lawrence (1966); and Urbach(1993,1995,1997). Imhof and Bosset (1994) list57 reviews on various aspects of cheese flavor.

In contrast to the attempts at chemically defin-ing desirable cheese flavor, efforts to identifythe compounds responsible for off-flavors havebeen moderately successful, because off-flavorsarise from a disproportionally high concentra-tion of certain compounds or groups of com-pounds. For example, bitterness is due mainly tohydrophobic peptides, rancidity to fatty acids,and fruitiness to esters.

Cheese Flavor

CHAPTER 12

Page 301: Cheese Science

Although it is not possible to describe theflavor of cheese in precise chemical terms, veryconsiderable progress has been made regardingthe identification of flavor compounds incheese and the elucidation of the biochemicalpathways by which these compounds are pro-duced. It is generally recognized that the aromaof cheese is in the volatile fraction and taste islargely in the aqueous phase; until recently,most researchers focused on the volatile frac-tion. Intervarietal comparisons should be valu-able for identifying key flavor compounds. Al-though several comparative studies on thevolatile compounds have been reported, therehave been relatively few comparative studieson the aqueous phase.

One of the major problems encountered in re-search on cheese flavor is defining what the typi-cal flavor should be. Within any variety, a fairlywide range of flavor and textural characteristicsis acceptable. This is particularly true for Ched-dar, which makes it especially difficult tochemically define its flavor. In cheese factories,wholesale or retail outlets, and research labora-tories, somebody, perhaps a single grader whoseviews may not be typical, decides what consti-tutes desirable and undesirable flavor. System-atic attempts to objectively describe the sensoryattributes of cheese have been made only re-cently, such as Hirst, Muir, and Naes (1994);McEwan, Moore, and Colwill (1989); and Muirand Hunter (1992). An international study underthe European Union's FLAIR-SENS program(FLAIR Conceited Action No. 2, Cost 902, Re-lating Instrumental, Sensory and ConsumerData) had a similar objective, especially forcheese varieties with Appellation d'OrigineControlee status. Progress in developing an in-ternational protocol for sensory profiling of hardcheese is reported by Nielsen et al. (1998). Anagreed vocabulary is essential if the results ofinstrumental studies are to be related to the sen-sory attributes and quality of cheese. The vo-cabulary developed by Nielsen et al. (1998) isshown in Exhibit 12-1 and the aroma-taste pro-files for 12 hard cheeses using this vocabularyare shown in Figure 12-1.

Exhibit 12-1 Descriptors Selected To DescribeCheese Flavor

Smell

CreamyYogurtCitrus fruit; other fruit; nuttyGrassCowshedCaramelSourAmmonia

Aroma-Taste

Creamy; yogurtGrassCitrus fruit; other fruit; nuttyCowshedCaramelSourPungentAmmoniaSweetSaltyAcidBitter

Texture

ElasticFirmnessCrumblinessCoatingDrynessMelting/solubilityGrainy

The texture of cheese has a major impact onflavor perception. For example, it has been sug-gested that the main contribution of proteolysisto cheese flavor is due to its effect on cheesetexture, which affects the release of sapid com-pounds during mastication of the cheese. There-fore, these attributes should be considered to-gether. However, they rarely are, and cheesetexture is even less well understood at the mo-lecular level than cheese flavor.

Page 302: Cheese Science

12.2 ANALYTICAL METHODS

12.2.1 Nonvolatile Compounds

Although studies on cheese flavor date fromthe beginning of this century, the techniquesavailable prior to the development of gas chro-matography (GC) in the 1950s were inadequateto permit significant progress. Early investiga-tors recognized the important contribution ofproteolysis and lipolysis to cheese ripening.Studies on proteolysis relied on changes in pro-tein and peptide solubility (e.g., in water, pH 4.6buffers, TCA, and ethanol). Such techniques,which have been reviewed by Fox (1989),International Dairy Federation (1991), andMcSweeney and Fox (1993, 1997) and are de-scribed briefly in Chapter 23, are still widelyused as indices of cheese maturity but are me-diocre as indices of cheese quality. Since the

products of proteolysis are nonvolatile, theycontribute to cheese taste but not to its aroma.

More specific studies on proteolysis becamepossible with the development of various typesof chromatography (paper, ion exchange, gelpermeation, and especially reversed phase high-performance liquid chromatography [RP-HPLC]). Large water-soluble and -insolublepeptides are best characterized by electrophore-sis, especially polyacrylamide gel electrophore-sis (PAGE), and, recently, capillary electro-phoresis. Isoelectric focusing has had limitedapplication in studies on cheese ripening. Morethan 200 peptides have been isolated from Ched-dar cheese and identified by N-terminal se-quencing and mass spectrometry (see Chapter11). Amino acids are usually quantified by ionexchange HPLC with postcolumn derivitizationusing ninhydrin or by separation of fluorescentamino acid derivatives by RP-HPLC.

Figure 12-1 Profiles of aroma-taste for 12 hard cheeses.

Sbrinz Svenbo

Emmental

Caerphilly Cheddar

Fontina

Comt£

Gouda

Danbo

Jarlsberg

Edam

Parmigiano-Reggiano

Fruity ToastedCreamy

Animal Pungent

SaltyAmmonia

IntensitySweet

AcidBitter

Star Key

Page 303: Cheese Science

The total concentration of free fatty acids isusually determined by extraction or titrationmethods or spectrophotometrically as Cu-soaps.Early attempts to quantify the concentration ofindividual short-chain fatty acids involvedsteam distillation and adsorption chromatogra-phy. Complete separation and quantitation offree fatty acids can be achieved by GC (usuallyas their methyl esters, for which several prepara-tive techniques have been published) or by RP-HPLC (see Chapter 23). Free fatty acids are ma-jor contributors to the flavor of some varieties,such as Romano, Feta, and Blue cheeses. In Bluecheeses, up to 25% of the total fatty acids may befree. Short-chain fatty acids are important con-tributors to cheese aroma, whereas longer chainacids contribute to taste. Excessive concentra-tions of either cause off-flavors (rancidity), andthe critical concentration is quite low in manyvarieties, such as Cheddar and Gouda.

Several other organic acids, especially lacticacids, are present in cheese and probably contrib-ute to flavor. These acids are routinely analyzedby HPLC. Enzymatic methods are available forseveral acids, such as D- and L-lactic acids (theenzymatic method is the easiest for distinguish-ing between the isomers) as well as acetic, pyru-vic, and succinic acids (see Chapter 23).

12.2.2 Volatile Compounds

The early GC instruments used packed col-umns, which gave relatively poor resolution, andthermal conductivity detectors, which lackedsensitivity. Volatiles were stripped by vacuumdistillation at, for example, 7O0C, and trapped incold fingers cooled by, for example, liquid nitro-gen. Such techniques may cause heat-inducedartifacts, and very volatile compounds may belost. Compounds were identified by comparingtheir retention times with those of standard com-pounds. The introduction of flame ionization de-tectors and capillary columns and the interfacingof GC with mass spectrometry (MS) greatly im-proved resolution and increased sensitivity. Se-lective detectors are available for sulfur com-pounds.

The problem of separating the volatile com-pounds from cheese remains. Vacuum distilla-tion of whole cheese or the fat fraction using im-proved apparatus (Figure 12-2) is still widelyused. In some cases, steam distillation is used.To avoid the generation of artifacts, the analysisof headspace volatiles has been used frequently.One form of this technique involves removing aplug from a block of cheese and covering theopening of the resulting hole with a septum (Fig-ure 12-3). After a period of time, a sample of gasin the cavity is withdrawn using an airtight sy-ringe and injected into the GC column. The sen-sitivity of headspace analysis can be improvedby trapping headspace volatiles in a Tenax trap,which can be inserted directly into the port of theGC. Alternatively, a sample of grated cheesemay be placed in a glass vessel closed with aseptum (Figure 12^4). A sample of headspacegas is withdrawn by syringe through the septum.The flask and contents may be heated to increasethe release of volatiles, if desired. A more effec-tive apparatus for collecting and concentratingheadspace gas is shown in Figure 12-5. Gratedcheese is placed in a long tube that is flushed foran extended period with an inert gas (e.g., he-lium), and the volatiles are collected in a Tenaxtrap. This approach permits a large volume ofheadspace gas to be trapped, thereby increasingsensitivity. The tube containing the sample maybe heated, if desired.

Some authors have used solvent extraction(e.g., with dichloromethane) or a combination ofsteam distillation and solvent extraction. Ultra-pure solvents are required to avoid artifacts, andthe solvent peak may mask peaks due to compo-nents of the sample.

A major challenge facing researchers study-ing cheese flavor is identifying the key com-pounds responsible for cheese flavor—many ofthe volatile compounds identified by GC-MSmay not contribute to flavor. A common ap-proach adopted to identify key aroma com-pounds is to sniff the eluate from the GC col-umn, a technique now referred to as gaschromatography/olfactometry (GCO). GCO ishard to quantify, and it is difficult to establish

Page 304: Cheese Science

Figure 12-2 Apparatus for primary vacuum distillation of cheese volatiles.

Cold-trap 4 Cold-trap 3 Cold-trap 2 Cold-trap 1

Dcwar flaskcontaining

liquid nitrogen

Ice water or water at 350C

4L Round-bottomedflnsk containingcheese sample

Rotary evaporatorThermostat

Vacuum pump

Cold-trap 5

TaplTap 2Tap 3

Tap 6Tap5

Vacuum gauges

Page 305: Cheese Science

whether a particular odorous compound is a ma-jor contributor to the flavor of the food under in-vestigation. To overcome these problems and toobtain information on the odor activities of theodorants detected during sniffing, two methodshave been developed to make GCO semi-quantitative: CHARM-analysis and aroma ex-tract dilution analysis (AEDA). Both methodsinvolve diluting the volatile fraction stepwisewith a suitable solvent and evaluating each dilu-tion by GCO. This procedure is performed untilno odorant is perceivable in the GC eluate. Thehighest dilution at which a compound can besmelled is defined as its flavor dilution (FD) fac-tor. The FD factor is the same for CHARJVI andAEDA, but the former also takes the persistenceof smell into account. FD factors may be plotted

against retention time (Rt) to give an FD-chro-matogram (Figure 12-6). CHARM or AEDAmay be applied to the analysis of volatile distil-lates, solvent extracts, or headspace gases.

Developments in the analysis of volatileflavor compounds have been reviewed bySchieberle (1995).

The human nose is extremely sensitive to aro-matic compounds—it is reported that some com-pounds can be detected at concentrations as lowas 0.02 |ig/kg. This very high sensitivity,coupled with the very large number of odorousmolecules in foods (e.g., > 600 odorous com-pounds have been detected in coffee), makes itvery expensive and very difficult or impossibleto detect complex odors by conventional analyti-cal methods. Consequently, traditional sensory

Figure 12-3 Arrangement for obtaining a sample of Figure 12-4 Arrangement for obtaining a sample ofheadspace gas from an opening cut in a block of headspace gas from grated cheese held in a glass ves-cheese. sel.

2 cm

Cheese

10 cm

Septum

Syringe

Page 306: Cheese Science

(organoleptic) methods have survived. Recently,attempts have been made to use arrays of sensorsthat are sensitive to certain compounds and givean electronic response in a manner somewhatanalogous to the nose (the device has been re-ferred to as the "electronic nose"). The mostcommonly used sensors are metal oxides, quartzresonators, lipid layers, phthalocyanines, andconducting polymers. The mechanism of opera-tion of these sensors and the electronic nose willnot be discussed here; the reader is referred toBartlett, Elliott, and Gardner (1997) andGardner (1996). The electronic nose could beused continuously, for example, as a quality con-trol device, and, unlike GC-MS, it does not re-quire sample preparation, which may generateartifacts. Such instruments have been used todistinguish between different coffees, differentbrands of sausage, different alcoholic beverages,and different types of meat or fish and to detectspoiled meat or fish. There is interest in the ap-plication of the electronic nose to the study ofcheese flavor, but the authors are not aware ofpublished data.

12.3 CONTRIBUTION OF THE AQUEOUSPHASE OF CHEESE TO FLAVOR

The ultrafiltration (UF) permeate (obtainedusing 10 kDa cutoff membranes) of the water-soluble fraction of cheese has a taste essentiallysimilar to that of the original cheese. This sug-gests that the basic taste of cheese is due to amixture of water-soluble compounds on which

other flavors are superimposed. The UF perme-ate contains small peptides, amino acids, short-chain fatty acids (C4-Ci0), other acids (acetic,succinic, especially lactic, and in some varietiespropionic acid), and inorganic salts, especiallyNaCl. There is a positive, although not a verystrong, correlation between the concentration ofphosphotungstic acid (PTA)-soluble N (verysmall peptides and amino acids) and the inten-sity of cheese flavor (Figure 12-7). On fraction-ation of the UF permeate of cheese by gel per-meation on a column of Sephadex G-25, thefraction containing very small peptides has a sa-vory cheeselike taste. However, no single pep-tide with a cheeselike taste has been isolated. Asdiscussed in Chapter 11, about 200 peptideshave been isolated from cheese, especiallyCheddar and Parmesan, but the very small pep-tides remain to be identified (the smallest pep-tides identified to date contained 5-6 amino ac-ids). There are probably very many di-, tri-, andtetrapeptides in mature cheese, but these are dif-ficult to isolate and identify. As discussed inSection 12.5.1, some peptides may have a bittertaste. If such peptides accumulate to an exces-sively high level, they lead to bitterness, a com-mon defect in many cheeses. However, probablyall cheeses contain bitter peptides, and whenpresent at certain concentrations and balancedby other sapid compounds, they probably makea positive contribution to cheese flavor.

The exact role of the medium and small pep-tides in cheese flavor is not clear, although it islikely that they contribute to the background fla-

Figure 12-5 Apparatus for the dynamic headspace extraction of solid foods: (1) carrier gas (He) inlet, (2) belttransmission, (3) glass cylinder, (4) Tenax trap, (5) thermostatted bath.

Page 307: Cheese Science

vor of Cheddar, at least toward a brothy or sa-vory flavor.

y-Glutamyl peptides, which contain a peptidebond between the y-carboxylic acid group of L-glutamic acid and another amino acid, have beenimplicated in cheese flavor. Although no y-glutamyl bonds occur in the caseins, y-Glu-Phe,y-Glu-Leu, and y-Glu-Tyr (at 9, 20, and 70 mg/kg, respectively) have been isolated fromGruyere de Comte cheese. Presumably, they areformed by y-glutamyl transferase (GGT, EC2.3.2.2), an indigenous enzyme in bovine milk.GGT catalyzes the transfer of the y-glutamylresidue from a y-glutamyl-containing peptide(e.g., glutathione, which occurs naturally inmilk) to an acceptor amino acid:

y-glutamyl-peptide + amino acid ^peptide + y-glutamyl-amino acid

It has also been postulated that the action of cer-tain synthetases, such as y-Glu-Cys synthetase(EC 6.3.2.2), perhaps in combination with GGT,may be responsible for the formation y-glutamylpeptides in cheese.

Some y-glutamyl peptides (e.g., y-Glu-Glu, y-GIu-Asp, y-Glu-Ala, and y-Glu-Gly) have a sourtaste, while the taste of y-Glu-Phe has been de-scribed as umami, slightly sour, salty, or metal-lic. The concentrations of the 3 y-glutamyl pep-tides found in Comte (9-70 mg/kg) are wellbelow the taste threshold of these compounds insolution (200-500 mg/kg). The presence of y-glutamyl peptides in Comte may be related tothe fact that it is a raw milk cheese; GGT activityin milk is attenuated by pasteurization (~ 80%loss on heating at 750C x 30 s). To date, there areno reports of the presence of y-glutamyl peptidesin other varieties of cheese. Further work on the

Figure 12-6 Example of a flavor dilution chromatogram.

Retention time [min]

Num

ber

of

as

se

ss

ors

Page 308: Cheese Science

occurrence of these peptides in other cheese va-rieties appears to be warranted.

Amino acids are produced in cheese by theaction of microbial exopeptidases on small andmedium-size peptides. The concentration ofamino acids in cheese varies with variety andage (see Fox & Wallace, 1997). Some amino ac-ids are bitter (Table 12-1), mainly those with

nonpolar or hydrophobic side chains, such as He,although Lys (usually charged) and Tyr (polarbut normally uncharged) are also considered tobe bitter. Pro and Lys are reported to be bitter-sweet. Arg is bitter, although it has low hydro-phobicity. Ala, GIy, Ser, and Thr are sweet,while GIu, His, and Asp are sour. Asp and GIuhave the lowest taste thresholds of the amino ac-

PTA-sol amino N ( p moles amino acid g ~1)

Figure 12-7 Correlations between 5% phosphotungstic acid (PTA)-soluble N (a measure of amino acids andshort peptides) and (a) total flavor, (b) mature flavor, and (c) age of Cheddar cheese.

Est

imat

ed A

ge (

mon

ths)

Mat

ure

Fla

vour

Sco

reT

otal

Fla

vour

Sco

re

Page 309: Cheese Science

ids (30 and 50 mg/L, respectively). The concen-trations of several amino acids in a water-solubleextract of mature Cheddar cheese are above theirflavor thresholds.

The term "sweet flavor" when applied tocheese often implies a lack of acidity but doesnot imply fruity aroma, which is due principallyto esters. Sweetness per se is regarded as a defectin many cheese varieties, although sweet flavornotes are desirable in some varieties (e.g.,Swiss). The sweet flavor in Swiss cheese, whichis concentrated in the water-soluble, nonvolatilefraction, is not due to sugars (lactose, glucose,and galactose), which are essentially absentfrom mature Swiss cheese, but to products ofproteolysis, especially proline. It has been sug-gested that the sweet flavor of Swiss cheese re-sults from the interaction OfCa2+ and Mg2+ withamino acids and small peptides. However, 4-hy-droxy-2,5-dimethyl-2(2H)-furanone and 5-

ethyl-4-hydroxy-3(2H)-furanone are also con-sidered to contribute to the sweet note in thearoma profile of Emmental cheese.

Thus, it appears that amino acids potentiallycontribute directly to cheese flavor and quality,although their exact role is not clear. However,products arising from the degradation of aminoacids through transamination, deamination, de-carboxylation, and other mechanisms (see Chap-ter 11) are very important to the development ofcheese aroma. These compounds include ammo-nia, amines, acids, carbonyls, sulfur-containingcompounds (e.g., H2S, methional, dimethyl sul-fide), and complex products produced via theMaillard and Strecker reactions. Most of thesecompounds are volatile and are discussed in Sec-tion 12.4.

Acid taste is caused by H+, the taste thresholdof which is about 2 mM. The principal acid incheese is lactic acid. Its concentration (and con-

Table 12-1 Taste Descriptor and Threshold Values of Amino Acids

Amino Acids Taste Threshold Value (mg/L) Average Hydrophobicity (cal/mol)

Histidine Bitter 200 500Methionine Bitter 300 1,300Valine Bitter 400 1,500Arginine Bitter 500 750lsoleucine Bitter 900 2,950Phenylalanine Bitter 900 2,500Tryptophan Bitter 900 3,400Leucine Bitter 1,900 1,800Tyrosine Bitter ND 2,300Alanine Sweet 600 500Glycine Sweet 1,300 OSerine Sweet 1,500 -300Threonine Sweet 2,600 400Lysine Sweet and bitter 3,000 1,500Proline Sweet and bitter 3,000 2,600Aspartic acid Sour 30 OGlutamic acid Sour 50 OAsparagine Sour 1,000 OGlutamine Flat ND OCysteine - ND 1,000Glutamate Na Umami 300Aspartate Na Umami 1,000

Page 310: Cheese Science

sequently the pH of cheese) varies considerablywith variety as influenced by its initial produc-tion by the starter, the extent of loss in the whey,and its metabolism by the secondary microfloraof the cheese. Total lactate concentration maynot be a good index of cheese acidity, since incertain varieties (e.g., mold-ripened cheeses) thepH increases during ripening, as a result of am-monia released by deamination of amino acidsand/or by the metabolism of lactate (see Chap-ters 10 and 11). The perception of acidity incheese can be influenced by NaCl concentration.Several other acids have been identified incheese (principally acetic, propionic, and bu-tyric acids) and presumably contribute to acid-ity, although they principally affect cheesearoma.

A number of short-chain fatty acids areformed in cheese during ripening. Some (e.g.,butanoic acid) are produced principally throughlipolysis while others (e.g., propanoic, acetic,and formic acids) result from the action ofcheese microflora on lactose, lactate, citrate, andamino acids. Short and medium-chain fatty acids(C4 to C10) contribute to the acid taste of cheese,although they contribute principally to its aroma.The acid taste of Swiss cheese correlates withthe concentration of di- and tripeptides (r2 =0.81) and amino acids (r2=.80) but not withcheese pH or the concentration of lactate.

Salt (NaCl) is an important contributor to thetaste of the water-soluble, nonvolatile fraction ofcheese. Salty taste is stimulated by small inor-ganic ions. The taste of chlorides of group I ele-ments varies from acid (HCl) through salty (e.g.,NaCl) to bitter (e.g., KCl and CsCl). The taste ofmost high molecular weight salts is bitter ratherthan salty, and cheeses salted using KCl, CaCl2,or MgCl2 are extremely bitter.

In the context of dairy foods, the compoundresponsible for salty taste is almost always NaCl.The threshold for NaCl is 30 mM (unspecifictaste response) and 40-50 mM for its salty taste.Bovine milk typically contains 25 mM Na+ and29 mM Cb, so the salty taste of dairy productsarises mainly from NaCl added during manufac-

ture. NaCl is particularly important to the flavorand quality of cheese and butter. Salt was origi-nally added to cheese as a preservative, acting toreduce water activity (aw) (see Chapters 8 and10). Although its preservative effect is still im-portant, the taste of the final product is now alsoconsidered when determining the amount of saltto be added during processing. Reduction in thelevel of NaCl in cheese (by adding a reduced levelor by partial replacement with KCl) for nutri-tional reasons can lead to bitterness due to exces-sive or unbalanced proteolysis (see Chapter 11).

Typical values for the concentration of NaClin cheese range from 0.7% for Quarg to 5.5% forRomano. The majority of cheeses contain 1-3%(w/w) NaCl (see Chapters 8, 17, and 21). Varia-tions in the salty taste of cheese are likely to berelated to processing parameters, such as theamount of salt added and its distributionthroughout the cheese. Consumers appear to beunable to detect differences in flavor and texturebetween cheeses containing 1.12% versus1.44% NaCl, but differences are apparent be-tween cheeses containing 0.75% versus 1.12%NaCl. Low NaCl concentrations facilitate pro-teolysis and the growth of lactic acid bacteria.

The apparent saltiness of cheese increaseswith maturity, increasing NaCl concentration,and decreasing pH. Grated cheese is perceivedas being more salty than the corresponding intactproduct but less salty than an aqueous NaCl so-lution of the same concentration.

12.4 CONTRIBUTION OF VOLATILECOMPOUNDS TO CHEESE FLAVOR

Compounds responsible for cheese aroma arevolatile to at least some degree. While somework on the volatile constituents of cheese wasdone before 1960 (e.g., on short-chain fatty ac-ids and amines), significant progress was notpossible until the development of GC. GC wasfirst applied to the study of Cheddar cheesevolatiles in 1962. Volatiles were stripped fromthe sample by vacuum distillation and con-densed in cold traps. They were then resolved by

Page 311: Cheese Science

GC on packed columns and located by thermalconductivity detectors. The introduction offlame ionization detectors, capillary columns,and interfacing GC with mass spectrometry(MS) markedly increased sensitivity and greatlyextended the number of compounds detected(> 200 for Cheddar).

Based on a survey of the published literature,Maarse and Visscher (1989) listed 213 volatilecompounds that had been identified in 40 studieson Cheddar cheese, including 33 hydrocarbons,24 alcohols, 13 aldehydes, 17 ketones, 42 acids,30 esters, 12 lactones, 18 amines, 7 sulfur com-pounds, 5 halogens, 6 nitriles and amides, 4phenols, 1 ether, and 1 pyran. The concentra-tions of many of these compounds were re-ported. The principal volatile compounds identi-fied in Cheddar are listed in Exhibit 12-2.

Thus, a great diversity of potentially sapidand/or aromatic compounds have been identi-fied in one or more varieties. These includesmall peptides (200 or more) and amino acidsand more than 200 volatile compounds (fatty ac-ids, other acids, carbonyls, amines, sulfur com-pounds, and hydrocarbons). Most cheese variet-ies contain similar volatile compounds but indifferent proportions. A comparison of the GC-MS profiles of six cheeses is shown in Figure12-8. The concentrations and proportion ofvolatile and nonvolatile flavor compounds areprobably responsible for the specific flavor ofeach cheese variety. However, the flavor ofsome varieties is dominated by a particular com-pound or class of compounds, such as short-chain fatty acids for Romano or methyl ketonesfor Blue cheeses. Although many of the flavorcompounds that have been identified in cheeseare present at concentrations well below theirflavor threshold, they may still modify the over-all flavor impact. The compounds considered byUrbach (1997) to be important or key contribu-tors to the flavor of a number of cheese varietiesare listed in Table 12-2. Although several hun-dred volatile and nonvolatile flavor compoundshave been identified in Cheddar cheese, Urbach(1997) and many earlier workers consider

methanethiol (CH3SH, derived from methion-ine) to be the key characteristic flavor compoundin Cheddar, although alone it does not have acheesy aroma. Obviously, the aroma of methan-ethiol is modified in cheese by the presence ofmany other compounds.

12.5 OFF-FLAVORS IN CHEESE

Avoiding off-flavors is generally more impor-tant than producing a cheese of exceptionallygood quality—most consumers will object tocheese with an off-flavor but relatively few willappreciate an excellent cheese. Consequently,cheese produced in large, well-managed facto-ries rarely suffers serious off-flavors. Since off-flavors usually have a specific cause, it is easierto control them than to develop good flavors,which have very complex causes.

Many off-flavors are due to the presence ofdisproportionately high concentrations of cer-tain compounds that may contribute positivelyto flavor at lower concentrations. Since specificcompounds are, in most cases, responsible foroff-flavors, they are generally understood at themolecular level. Some specific off-flavors areconsidered in the following sections.

12.5.1 Bitterness

Bitterness is probably the principal taste de-fect in cheese. Although amino acids, amines,amides, substituted amides, long-chain ketones,some monoglycerides, N-acyl amino acids, anddiketopiperazines may contribute to bitterness,this defect in cheese usually results from the ac-cumulation of hydrophobic peptides.

Bitterness and Hydrophobicity

Ney (1979) suggested that the mean hydro-phobicity (Q) of a peptide, expressed as

e = Mn

where, A/ is side chain hydrophobicity and n isthe number of residues, is the principal determi-

Page 312: Cheese Science

nant of bitterness rather than any particularamino acid sequence. Although further worksuggested that the nature of the terminal aminoacids and certain steric parameters influence theperception of bitterness (see Lemieux & Simard,1991, 1992), the mean hydrophobicity of a pep-tide appears to be the single most important fac-tor determining its bitterness. Peptides with amolecular weight of 0.1 to 10 kDa and Q lessthan 1,300 cal per residue are not bitter whilethose with Q greater than 1,400 cal per residueand a molecular weight of 0.1 to 6 kDa are bitter;above 6 kDa, even peptides with Q greater than1,400 cal per residue are not bitter. Hydrolysatesof proteins with a high mean hydrophobicity arelikely to contain bitter peptides, although thedistribution of hydrophobic residues along thepolypeptide and the specificity of the proteinaseused to prepare the hydrolysate also influencethe development of bitterness. Since the caseins,especially p-casein, are quite hydrophobic and

the hydrophobic residues are clustered (seeChapter 3), casein hydrolysates have a high pro-pensity to bitterness.

As discussed in Chapter 11, cheese contains agreat diversity of proteinases and peptidaseswith different and complementary specificities.Although detailed kinetic studies are lacking, atleast some of the peptides in cheese are transient,and hence bitterness may wax and wane as bitterpeptides are formed and hydrolyzed or maskedby other sapid compounds. It is very likely thatall cheeses contain bitter peptides, which prob-ably contribute positively to the overall desir-able flavor. Bitterness becomes a problem onlywhen bitter peptides accumulate to an excessive,unbalanced level. Although bitter peptides canoriginate from either asr or P-casein, the actionof chymosin and/or lactococcal cell envelope-associated proteinase (CEP) on the very hydro-phobic C-terminal region of P-casein may resultin the production of bitter peptides. This action

Exhibit 12-2 Volatile Compounds Identified in Cheddar Cheese

acetaldehydeacetoinacetoneacetophenonep-angelicalatone1,2-butanedioln-butanol2-butanolbutanonen-butyl acetate2-butyl acetaten-butyl butyraten-butyric acidcarbon dioxide/>-cresoly-decalactone8-decalactonen-decanoic aciddiacetyldiethyl ether

dimethyl sulfidedimethyl disulfidedimethyl trisulfide8-dodecalactoneethanolethyl acetate2-ethyl butanolethyl butyrateethyl hexanoate2-heptanonen-hexanaln-hexanoic acidn-hexanol2-hexanonehexanethiol2-hexenalisobutanolisohexanalmethane thiolmemional

methyl acetate2-methylbutanol3-memylbutanol3 -methyl-2-butanone3 -methy!butyric acid2-nonanone8-octalactonen-octanoic acid2-octanol2,4-pentanedioln-pentanoic acid2-pentanolpentan-2-onen-propanolpropanalpropenaln-propyl buryratetetrahydrofuranthiophen-2-aldehyde2-tridecanone2-undecanone

Page 313: Cheese Science

is very dependent on salt concentration (seeChapter 11). The peptides produced initiallyfrom ocsi-casein are from the hydrophilic N-ter-minal region and therefore are less likely to bebitter. The production of bitter peptides also de-pends on the specificity of the lactococcal CEP.For example, Pm-type CEP (Lc. lactis ssp.cremoris AMi) produces less bitter casein hy-drolysates than Pi-type CEP (Lc. lactis ssp.cremoris HP), perhaps because more peptidesare initially released from the hydrophobic C-terminal region of p-casein by the latter. Theconcentration of bitter peptides depends on the

rate at which they are degraded by lactococcalpeptidases and perhaps, in the case of larger bit-ter peptides, by the CEP. The debittering effectof aminopeptidase N on a tryptic digest of P-casein has been demonstrated.

Factors Affecting the Development ofBitterness in Cheese

Certain starters have a propensity to cause bit-terness. Nonbitter cheese can be made usingthese strains provided they are combined with"nonbitter" strains. Since the coagulant maycause the release of bitter peptides, factors that

Time (min)

Figure 12-8 GC-MS chromatograms of the headspace volatiles for six cheese varieties.

Abu

ndan

ce (

arbi

trar

y un

it)

PARMlGlANOREGGIANO

MAHON

FONTINA

COMTE

BEAUFORT

APPENZELLER

Page 314: Cheese Science

Table 12-2 Important Flavor Compounds in Various Cheeses

Cheese

Cheddar

Camembert

Emmental

Romano

Parmesan

Provolone

Goat milk cheese

Pecorino Romano

Limburg

Surface-ripened cheeses

Pont-l'Eveque

Vacherin

Roquefort

Livarot

Munster

Trappist

Blue cheeses

Brie

Carre de I'Est

Epoisses

Maroilles

Langres

Buffalo Mozzarella

Bovine Mozzarella

Compounds

Methanethiol

Nonan-2-one, oct-1-en-3-ol, N-isobutylacetamide, 2-phenylethanol, 2-phenylethyl acetate, 2-heptanol, 2-nonanol, NH3, isovaleric acid, isobu-tyrlc acid, hydroxybenzoic acid, hydroxyphenylacetic acid

Methional, 4-hydroxy-2,5-dimethyl-3(2H)-furanone (Furaneol), 5-ethyl-4-hydroxy-2-methyl-3 (2H)-furanone

Butanoic, hexanoic, and octanoic acids

Butanoic, hexanoic, and octanoic acids, ethyl butyrate, ethyl hexanoate,ethyl acetate, ethyl octanoate, ethyl decanoate, methyl hexanoate

Butanoic, hexanoic, and octanoic acids

4-Methyloctanoic acid, 4-ethyloctanoic acid

4-Methyloctanoic and 4-ethyloctanoic acids, p-cresol, /n-cresol, 3,4-dimethylphenol

Methanethiol, methyl thioacetate

Methyl thioesters

/V-lsobutylacetamide, phenol, isobutyric acid, 3-methylvaleric acid, isova-leric acid, heptan-2-one, nonan-2-one, acetophenone, 2-phenylethanol,indole

Acetophenone, phenol, dimethyl disulfide, indole, terpineol, isoborneol,linalool

Oct-1-en-3-ol, methyl ketones

Phenol, m- and p-cresols, dimethyl disulfide, isobutyric acid,3-methylvaleric acid, isovaleric acid, benzoic acid, phenylacetic acid,nonan-2-one, acetophenone, 2-phenylethanol, 2-phenylethyl acetate,dimethyl disulfide, indole

Dimethyl disulfide, isobutyric acid, 3-methylvaleric acid, isovaleric acid,benzoic acid, phenylacetic acid

H2S, methanethiol

Heptan-2-one, nonan-2-one, methyl esters of C4i6j8,io,i2 acids, ethyl estersof C1>2j4,6,8,io acids

Isobutyric acid, isovaleric acid, methyl ketones, sulfur compounds, oct-1-en-3-ol

Isobutyric acid, isovaleric acid, 3-methylvaleric acid

2-Phenylethanol

2-Phenylethanol, nonan-2-one, acetophenone, phenol, indole

2-Phenylethanol, dimethyl disulfide, styrene, indole

Oct-1-en-3-ol, nonanal, indole, component Rl 975 with odor of truffles

Ethyl isobutanoate, ethyl 3-methylbutanoate

Page 315: Cheese Science

affect its retention in cheese curd (type andquantity used in cheesemaking, pH at wheydrainage, and cook temperature) influence thedevelopment of bitterness. Certain rennet substi-tutes produce bitter cheese, owing to excessivelyhigh activity and/or incorrect specificity. The pHof cheese also influences the activity of residualcoagulant and other enzymes. Cheese with a lowsalt concentration is very prone to bitterness,perhaps because the susceptibility of p-casein tohydrolysis by chymosin, with the production ofthe bitter peptide p-CN f 193-209, is strongly af-fected by the NaCl concentration in cheese. Saltalso inhibits lactococcal CEP and may promotethe aggregation of large, nonbitter hydrophobicpeptides that would otherwise be degraded tobitter peptides. Bitterness can be particularlyproblematic in low-fat cheeses, perhaps owingto reduced partitioning of hydrophobic peptidesinto the fat phase.

Bitter Peptides Isolated from Cheese

Bitter peptides that have been isolated fromcheese are summarized in Figures 12-9 and12-10. As expected, bitter peptides originateprincipally from hydrophobic regions of thecaseins, including sequences 14 to 34, 91 to 101,and 143 to 151 of asl-casein and 46 to 90 or 190to 209 of p-casein. As discussed by McSweeneyet al. (1997), the majority of these peptides showevidence of some degradation by lactococcalproteinases and/or peptidases.

12.5.2 Astringency

Astringency is a taste-related phenomenonperceived as a dry feeling in the mouth and apuckering of the oral tissue. In many foods, as-tringency involves interaction between tannins(polyphenols) and proteins in saliva, but this isprobably not the cause of astringency in cheese.An astringent fraction was extracted from Ched-dar cheese using chloroform-methanol, but itwas not characterized. Since the abstract ab-sorbed light at 280 nm, it probably containedpeptides. The aqueous fraction of Comte cheese

contains N-propionyl methionine, which isslightly bitter, astringent, and pungent.

12.5.3 Fruitiness

The principal compounds responsible forfruitiness in cheese are esters, especially ethylbutyrate and ethyl hexanoate, formed by esterifi-cation of fatty acids with ethanol. Production ofethanol appears to be the limiting factor, as fattyacids are present in cheese at relatively high con-centrations. Ethyl esters are present at low con-centrations in nonfruity cheeses, and the fruitydefect occurs as a result of excessive productionof ethanol or its precursors.

12.5.4 "Unclean" Off-Flavors

Phenylacetaldehyde, which is produced viathe Strecker reaction, causes "unclean" and re-lated flavors in Cheddar cheese. At higher con-centrations (> 500 |Hg/kg), phenylacetaldehydeimparts astringent, bitter, and stinging flavors tocheese. /?-Cresol imparts a "utensil-type" flavorwhen present at high concentrations. Short-chain fatty acids potentiate the flavor impact of/?-cresol. Phenol contributes to the unclean fla-vor of cheese but enhances the sharpness ofCheddar flavor.

12.6 FORMATION OF FLAVORCOMPOUNDS

Some of the flavor compounds in cheese, in-cluding lactic acid, diacetyl, and methyl ketones,are produced by the metabolic activity of livingmicroorganisms. A few are produced by chemi-cal reactions, for example, via the Maillard andStrecker reactions between amino acids andvarious carbonyls, especially dicarbonyls, e.g.,diacetyl, glyoxal, or methyl glyoxal. Some ex-amples of the Strecker reaction are shown inFigure 12-11. However, most flavor compoundsare produced by enzymes that are either indig-enous to milk (especially plasmin and lipase),added to the milk (especially chymosin or rennet

Page 316: Cheese Science

RESIDUE NUMBER

Figure 12-9 Hydrophobicity of bovine asl-casein plotted as a function of amino acid residue. The locations of bitter peptides originating from ocsl-casein andisolated from cheese or casein hydrolysates (as listed by Lemieux and Simard, 1992) are shown approximately to scale.

Peptides isolated from Cheddar

Peptides isolated from casein hydrolysates

Peptide isolated from Alpkase

Page 317: Cheese Science

RESIDUE NUMBER

Figure 12-10 Hydrophobicity of bovine p-casein plotted as a function of amino acid residue. The locations of bitter peptides originating from p-casein andisolated from cheese or casein hydrolysates or synthesized (as listed by Lemieux and Simard, 1992) are shown approximately to scale.

Page 318: Cheese Science

substitute and, in some varieties, pregastric es-terase), secreted by microorganisms (especiallymolds and other microorganisms in the surfacesmear), or released from microbial cells follow-ing cell death and lysis.

The pathways leading to the formation ofmost of the flavor compounds in cheese areknown and are described in Chapter 11. The sub-ject has been reviewed by Fox and Wallace(1997). The principal gap in our knowledge onflavor generation is the ability to precisely bal-ance and control the various reactions. Many ofthe compounds that are considered to be mostimportant in cheese flavor are present at verylow concentrations and, in many varieties, are

produced mainly via the catabolism of aminoacids. The enzymes involved are intracellularbacterial (starter and nonstarter) enzymes pres-ent at low levels and are often relatively difficultto assay. Therefore, they have to date beenlargely neglected, although they are now attract-ing some attention.

12.7 INTERVARIETAL ANDINTRAVARIETAL COMPARISON OFCHEESE RIPENING

As discussed in Section 12.4, most ripenedcheeses contain essentially the same sapid andaromatic compounds but at different concentra-

Diacety!(butanedione)

Valine

lsobutyraldehyde

Tetramethylpyrazine

Figure 12-11 Example of the Strecker degradation reaction.

Page 319: Cheese Science

tions and proportions. Therefore, it appears rea-sonable to presume that inter- and intravarietalcomparison, especially of closely related variet-ies, might help to identify compounds mostlikely to contribute to characteristic cheese fla-vors. However, although both the water-solublenonvolatile and volatile fractions of severalcheese varieties have been analyzed, there arerelatively few intervarietal comparisons, espe-cially of the nonvolatile fractions. In this section,the results of some such studies are discussed.

12.7.1 Gel Electrophoresis

Gel electrophoresis, especially alkaline (pH9.0) urea-PAGE, is the best method for charac-terizing the large water-insoluble peptides and isalso useful for characterizing the larger water-soluble peptides.

Urea-PAGE of water-insoluble and -solublefractions of cheese indicate variety-specific pep-tide patterns (Figures 11-16 and 11-17), reflect-ing mainly the relative activities of chymosinand plasmin (see Chapter 11). The peptides de-tectable by urea-PAGE are too large to affectcheese flavor, but large peptides signal the pres-ence of certain small peptides that may affectflavor or perhaps, more correctly, off-flavor, es-pecially bitterness. For instance, the presence ofa band corresponding to (3-CN fl-192 indicatesthe presence of p-CN f 193-209 or fractionsthereof, which are bitter.

Urea-PAGE of the water-insoluble and-soluble fractions of cheeses made using singlestrain starters show very little difference be-tween cheeses differing substantially in quality.This is not surprising, since starter proteinasescontribute little to primary proteolysis in cheese,as detected by PAGE. Urea-PAGE also fails toshow substantial differences between cheesesmade from pasteurized or raw milk, although theflavor of these cheeses differ markedly. Urea-PAGE of the water-insoluble fraction generallyreflects the age of the cheese (if ripened at a par-ticular temperature) and is a useful index of itstextural quality (if the age is known) but not ofits flavor.

12.7.2 High Performance LiquidChromatography

RP-HPLC is more effective than PAGE(which does not detect peptides less than 3,000Da) for analysis of the water-soluble peptides.RP-HPLC profiles of the UF permeate andretentate of cheese become more complex as thecheese matures (see O'Shea, Uniacke-Lowe, &Fox, 1996). Unfortunately, it is not possible atpresent to relate cheese flavor or texture toHPLC chromatograms. HPLC of the 70% etha-nol-soluble and -insoluble fractions of the water-soluble peptides shows clear varietal character-istics (see Figures 11-18 and 11-19). Theconcentration of free amino acids (measured byreaction with Cd-Ninhydrin) in Cheddar ishighly correlated with age (Figure 12-12) andhence with the intensity of cheese flavor. Theratio of free amino acids to water-soluble N ap-pears to be characteristic of the variety (Figure12-13). Cheddar appears to contain a very lowlevel of free amino acids relative to small pep-tides. Parmesan contains a particularly high con-centration of amino acids, which have a majoreffect on its flavor, but it contains a low concen-tration of small peptides (Figures 11-18 and 11-19).

The water-soluble fraction also containsshort-chain fatty acids (< C9:0), which impart a"cheesy" aroma. The few available studies indi-cate that there are substantial intervarietal differ-ences with respect to short-chain fatty acids.Further work in this area appears to be war-ranted.

12.7.3 Cheese Volatiles

A number of intra- and intervarietal compari-sons of cheese volatiles have been published. Anearly example is the study of Manning andMoore (1979), who analyzed headspace vola-tiles of nine fairly closely related varieties. Con-siderable intervarietal differences were evident,but the four samples of "Cheddar" that were ana-lyzed also differed markedly. The intensity ofcheese flavor was reported to be related to the

Page 320: Cheese Science

mg

leu/

g pe

rmea

te

Age (months)

Figure 12-12 Total free amino acid concentrations (Cd-Ninhydrin assay) in Cheddar cheese as a function of age.

Cheddar

Swiss type

Ecfem

Gouda

% water soluble N

Figure 12-13 Relation between the concentration of amino acids (Cd-Ninhydrin assay) and percentage of water-soluble N in a selection of cheese varieties.

A5

07

(Cd

-nin

hy

dri

n)

Page 321: Cheese Science

concentration of sulfur compounds; 2-pentanonewas also considered to be important for Cheddarcheese flavor.

A comprehensive study on Cheddar, Gouda,Edam, Swiss, and Parmesan (total of 82 sam-ples) was reported by Aishima and Nakai(1987). The volatiles were extracted by CH2Cl2

and analyzed by GC. More than 200 peaks wereresolved in all chromatograms, 118 of whichwere selected as variables for discriminativeanalysis. Expression of the area of each of the118 peaks as a percentage of total chromatogramarea clearly permitted classification of the fivevarieties. The compounds likely to be respon-sible for the characteristic flavor of each varietywere not discussed.

Bosset and Gauch (1993) concentrated theheadspace volatiles from six cheese varieties bya "purge and trap" method for analysis by GC-MS. A total of 81 compounds were isolated andidentified (Figure 12-8), 20 of which werefound in all six varieties and a further nine in fiveof the six varieties. The authors concluded that"practically all types of cheese analysed containmore or less the same volatiles, but at differentconcentrations. Thus, the flavour of these

REFERENCES

Adda, J., Gripon, J.-C., & Vassal, L. (1982). The chemistryof flavour and texture development in cheese. FoodChemistry, 9, 115-129.

Aishima, T., & Nakai, S. (1987). Pattern recognition of GCprofiles for classification of cheese variety. Journal ofFood Science, 52, 939-942.

Aston, J.W., & Dulley, J.R. (1982). Cheddar cheese flavour.Australian Journal of Dairy Technology, 37, 59—64.

Bartlett, P.N., Elliott, J.M., & Gardner, J.W. (1997). Elec-tronic noses and their application in the food industry.Food Technology, 51(12), 44^8.

Bosset, J.O., & Gauch, R. (1993). Comparison of the volatileflavour in six European AOC cheeses by using a new dy-namic headspace GC-MS method. International DairyJournal, 3, 359-377.

Fox, P.P. (1989). Proteolysis in cheese during manufacturingand ripening. Journal of Dairy Science, 72, 1379-1400.

Fox, P.P., Singh, T.K., & McSweeney, P.L.H. (1995). Bio-genesis of flavour compounds in cheese. In E.L. Malin &

cheeses seems to depend not on any particularkey compound, but rather on a critical balance,or a weighted concentration ratio of all compo-nents present" (p. 366).

12.8 CONCLUSION

The component balance theory of cheese fla-vor still applies. While analytical techniqueshave improved greatly during the last 40 yearsand data are available on the concentrations ofmany flavor compounds in cheese, the criticalcompounds, if any, are unknown in most cases.Further intervarietal comparisons may be usefulif quantitative data are obtained, although fre-quently they are not. Perhaps it would be fruit-ful to reinvestigate cheeses with a controlledmicroflora. The last studies on such systemswere done in the 1970s and those were con-cerned mainly or totally with proteolysis. Itwould seem to be particularly useful to combinestudies on controlled microflora cheese withintervarietal comparisons, but perhaps such anundertaking is beyond the capabilities of asingle laboratory.

M.H. Tunick (Eds.), Chemistry of structure-function re-lationships in cheese. New York: Plenum Press.

Fox, P.P., & Wallace, J.M. (1997). Formation of flavourcompounds in cheese. Advances in Applied Microbiol-ogy, 45, 17-85.

Gardner, J. (1996). An introduction to electronic nose tech-nology. Essex, UK: Neotronics Scientific Ltd.

Hirst, D., Muir, D.D., & Naes, T. (1994). Definition of thesensory properties of hard cheese: A collaborative studybetween Scottish and Norwegian panels. InternationalDairy Journal, 4, 743-761.

Imhof, R., & Bosset, J.O. (1994). Relationships between mi-croorganisms and formation of aroma compounds in fer-mented dairy products. Zeitschrift fur Untersuchung undForschung, 198, 267-276.

International Dairy Federation. (1991). Chemical methodsfor evaluation of proteolysis in cheese maturation [Bulle-tin No. 261]. Brussels; Author.

Page 322: Cheese Science

Kristoffersen, T. (1973). Biogenesis of cheese flavour. Jour-nal of Agricultural and Food Chemistry, 21, 573-575.

Kristoffersen, T. (1985). Development of flavour in cheese.Milchwissenschaft, 40, 147-199.

Lemieux, L., & Simard, R.E. (1991). Bitter flavour in dairyproducts. I. A review of the factors likely to influence itsdevelopment, mainly in cheese manufacture. Lait, 71,599-636.

Lemieux, L., & Simard, R.E. (1992). Bitter flavour in dairyproducts. II. A review of bitter peptides from the caseins:their formation, isolation and identification, structuremasking and inhibition. Lait, 72, 335-382.

Maarse, H., & Visscher, C.A. (1989). Volatile compounds infood (6th ed.). Zeist, The Netherlands: TNO Division forNutrition and Food Research.

Manning, D.J., & Moore, C. (1979). Headspace analysis ofhard cheese. Journal of Dairy Research, 46, 539-545.

McEwan, J.A., Moore, J.D., & Colwill, J.S. (1989). The sen-sory characteristics of Cheddar cheese and their relation-ship with acceptability. Journal of the Society of DairyTechnology, 4, 112-117.

McGugan, W.A. (1975). Cheddar cheese flavour: A reviewof current progress. Journal of Agricultural and FoodChemistry, 23, 1047-1050.

McSweeney, P.L.H., & Fox, P.F. (1993). Cheese: Methodsof chemical analysis. In P.F. Fox (Ed.), Cheese: Chemis-try, physics and microbiology (2d ed., Vol. 1). London:Chapman & Hall.

McSweeney, P.L.H., & Fox, P.F. (1997). Chemical methodsfor characterization of proteolysis in cheese during ripen-ing. Lait, 77,41-76.

McSweeney, P.L.H., Nursten, H.E., & Urbach, G. (1997).Flavours and off-flavours in milk and dairy products. InP.F. Fox (Ed.), Advanced Dairy Chemistry (2d ed., Vol.3). London: Chapman & Hall.

Muir, D.D., & Hunter, E.A. (1992). Sensory evaluation ofCheddar cheese: The relation of sensory properties to per-ception of maturity. Journal of the Society of Dairy Tech-nology, 45, 23-30.

Ney, K.H. (1979). Bitterness of peptides: Amino acid com-position and chain length. In J.C. Boudreau (Ed.), Foodtaste chemistry (ACS Symposium Series 115). Washing-ton, DC: American Chemical Co.

Nielsen, R.G., Zannoni, M., Beriodier, F., Lavanchy, P.,Muir, D.D., & Siverten, H.K. (1998). Progress in devel-oping an international protocol for sensory profiling ofhard cheese. International Journal of Dairy Technology,51, 57-64.

Olson, N.F. (1990). The impact of lactic acid bacteria oncheese flavour. Federation of European MicrobiologicalSocieties Microbiological Reviews, 43, 497—499.

O'Shea, B.A., Uniacke-Lowe, T., & Fox, P.F. (1996). Ob-jective assessment of Cheddar cheese quality. Interna-tional Dairy Journal, 6, 1135—1147.

Reiter, B., Fryer, T.F., Sharpe, M.E., & Lawrence, R.C.(1966). Studies on cheese flavour. Journal of AppliedBacteriology, 29, 231-242.

Schieberle, P. (1995). New developments in methods foranalysis of volatile flavor compounds and their precur-sors. In A.G. Gaonkar (Ed.), Characterization of food:Emerging methods. Amsterdam: Elsevier Science.

Urbach, G. (1993). Relations between cheese flavour andchemical composition. International Dairy Journal, 3,389^22.

Urbach, G. (1995). Contribution of lactic acid bacteria toflavour compound formation in dairy products. Interna-tional Dairy Journal, 5, 877-903.

Urbach, G. (1997). The flavour of milk and dairy products.II. Cheese: Contribution of volatile compounds. Interna-tional Journal of Dairy Technology, 50, 79-89.

Page 323: Cheese Science

13.1 INTRODUCTION

Rheology involves the study of the deforma-tion and flow of materials when subjected to astress or strain. The rheological properties ofcheese are those that determine its response to astress or strain (e.g., compression, shearing, orcutting) that is applied during processing (e.g.,portioning, slicing, shredding, or grating) andconsumption (slicing, spreading, masticating,and chewing). These properties include intrinsiccharacteristics—such as elasticity, viscosity,and viscoelasticity—that are related primarily tothe composition, structure, and strength of theattractions between the structural elements ofthe cheese. The rheological characteristics aredetermined by the application of a fixed stress orstrain to a sample of cheese under defined ex-perimental conditions. The relationship betweenstress and strain may be described using variousrheological terms, including bulk modulus, elas-tic modulus, shear modulus, fracture stress orstrain, and firmness. In lay language, the behav-ior of cheese when subjected to a stress or straincan be described by terms such as hardness,firmness, springiness, crumbliness, and adhe-siveness. The rheological properties of cheeseare of considerable importance, since they affectits

• handling, portioning, and packing charac-teristics

• texture and eating quality, which are deter-mined by the effort required to masticate

the cheese or, alternatively, the level ofmastication achieved for a given level ofchewing (the latter may, in turn, influenceits flavor and aroma and the suitability ofthe cheese for different consumer groups,such as children or aged people)

• use as an ingredient, since they determineits behavior when subjected to different sizereduction methods (such as shredding, grat-ing, and shearing) and how the cheese inter-acts with other ingredients in foods

• ability to retain a given shape at a giventemperature or when stacked

• ability to retain gas and hence to form eyesor cracks or to swell

In short, the rheological properties of cheese arequality attributes that are important to the manu-facturer, packager, distributor, retailer, indus-trial user, and consumer.

The rheology of cheese is a function of itscomposition, microstructure (i.e., the structuralarrangement of its components), the physico-chemical state of its components, and its macro-structure, which reflects the presence of hetero-geneities such as curd granule junctions, cracks,and fissures. The physicochemical properties in-clude parameters such as the level of fat coales-cence, solid fatliquid fat ratio, degree of hy-drolysis and hydration of the paracasein matrix,and level of intermolecular attractions betweenparacasein molecules. Hence, the rheologicalcharacteristics differ markedly with the cheese

Cheese Rheology and Texture

CHAPTER 13

Page 324: Cheese Science

variety and its age. The effect of variety on thetheological properties is readily apparent whenone compares an almost flowable mature Cam-embert with a firm, brittle Parmesan or com-pares a crumbly Cheshire cheese with a springyEmmental-type cheese or String cheese. Simi-larly, the influence of age is shown by the differ-ence between a young (< 1-2 months) rubberyCheddar with a fully mature pliable Cheddar.

Cheese texture may be defined as a compositesensory attribute resulting from a combinationof physical properties and perceived by thesenses of sight, touch, and hearing (Brennan,1988). The properties of cheese that contributeto its texture may be divided into three principalcategories (Szczesniak, 1963): (1) mechanical,(2) geometrical, and (3) others.

Mechanical properties are manifested by thereaction of food to a stress applied during con-sumption (e.g., squeezing between the fingers,manual cutting, and mastication). They com-prise the following characteristics: hardness, co-hesiveness, viscosity, springiness, chewiness,brittleness, and gumminess. The mechanicalproperties are measured organoleptically by thepressure exerted on the cheese by the teeth,tongue, and roof of the mouth during eating.Geometrical properties include the size distribu-tion, shape, and orientation of particles within afood (particles of the whole food material or ofcomponents such as occluded air or fat). Theyare reflected mainly in the visual characteristicsof the cheese (e.g., granularity in Cottage cheeseor the fibrous nature of String cheese). However,the geometrical characteristics may be suffi-ciently pronounced to affect the mechanicalproperties of the cheese. The other propertiesthat contribute to cheese texture include charac-teristics such as greasiness, oiliness, succulence,and mouth-coating associated with the presenceof fat and moisture within the cheese.

An alternative scheme to that of Szczesniak(1963) for classifying textural properties wasproposed by Sherman (1969). In this scheme,textural properties are considered to be primary(or fundamental), secondary, or tertiary. The pri-mary characteristics—from which the secondary

and tertiary characteristics are derived—includethe composition of the food, its micro- and mac-rostructure, and its molecular properties. Thetextural characteristics are further subdividedinto two broad groups based on sensory percep-tions of the material before and during consump-tion (Figure 13-1). Characteristics that contrib-ute to the initial perception of cheese texture,before eating, include

• visual appearance (e.g., presence of holes,eyes, or granules and surface roughness orsmoothness)

• crumbliness, springiness, stickiness, andslicing characteristics

• spreading characteristics (important forpasteurized processed cheese spreads)

The secondary and tertiary categories of texturalproperties include many characteristics (e.g.,hardness, brittleness, softness, and springiness)that are directly related to the intrinsic rheologi-cal properties (e.g., elasticity and viscosity) thatdetermine the cheese's response to the stressesapplied during biting, chewing, and salivation.Hence, cheese texture and cheese rheology areclosely related, in that many of the textural prop-erties of cheese are determined by its rheologicalproperties.

Cheese rheology, which has been studied andreviewed extensively (Culioli & Sherman, 1976;Prentice, Langley, & Marshall, 1993; Rao, 1992;Sherman, 1969, 1988; van Vliet, 1991; Visser,1991) is the focus of this chapter. For detailedinformation on cheese texture, the reader is re-ferred to a number of studies and reviews(Brandt, Skinner, & Coleman, 1963; Brennan,1988; Jack, Paterson, & Piggott, 1993, 1995;Sherman, 1969; Szczesniak, 1963).

13.2 CHEESE MICROSTRUCTURE

The micro- and macrostructure of cheese aremajor determinants of its rheological and tex-tural properties and are discussed briefly below.Natural rennet-curd cheese is essentially a cal-cium phosphate-paracasein matrix composed of

Page 325: Cheese Science

Initial perception(before placing inmouth)

Initial perceptionon palate

Mastication(high shearingstress)

Residualmasticatoryimpression

Figure 13-1 Textural characteristics of foods.

Primary characteristics

Secondary characteristics

Tertiary characteristics

Visual appearanceSampling and slicing characteristicsSpreading, creaming characteristics, pourability

Chemical compositionParticle size, shape, and size distributionOil content, size, shape, and size distribution of oildroplets

Elasticity, cohesionViscosityAdhesion (to palate)

Hard, softBrittle, plastic, crisp, rubbery, spongySmooth, coarse, powdery, lumpy, pastyCreamy, watery, soggySticky, tacky

Greasy, gummy, stringyMelt-down properties on palate

Page 326: Cheese Science

overlapping and cross-linked strands of partiallyfused paracasein aggregates (in turn formedfrom fused paracasein micelles) (Figure 13-2).The integrity of the matrix is maintained by vari-

ous intra- and interaggergate hydrophobic andelectrostatic attractions. The matrix occludes,within its pores, fat globules (in varying degreesof coalescence), moisture, and dissolved sub-

Figure 13-2 Scanning electron micrographs of Cheddar cheese, showing the continuous paracasein matrix (ar-row heads) permeated by holes and fissures corresponding to discreet fat globules and/or pools of clumped orcoalesced fat globules (solid arrows). Bar in (A) equals 5 um and in (B) 1 um.

A

B

Page 327: Cheese Science

stances (minerals, lactic acid, peptides, andamino acids), and enzymes (e.g., residual rennetand proteinases from starter and nonstarter mi-croorganisms). The cheese matrix also containsvarious microorganisms (in most cases starterand nonstarter bacteria and in some cases molds,yeasts, and other bacteria on the surface of thecheese) and their enzymes (e.g., proteinases,peptidases, and lipases), which are released intothe cheese matrix at various rates during matura-tion. A dynamic equilibrium exists between theconcentrations of Ca and inorganic phosphate inthe paracasein matrix and the cheese serum. Theequilibrium is influenced by pH and other fac-tors such as the concentration of Na+ in the se-rum phase.

The paracasein network is essentially con-tinuous, extending in all directions, althoughsome discontinuities in the matrix may exist dueto the presence of curd granule junctions and/orcurd chip junctions (e.g., as occur in Cheddarand related dry-salted varieties). Microscopicalexamination of low-moisture Mozzarella cheesereveals the presence of well-defined curd gran-ule junctions (~ 4 jam wide) (Figure 13-3), whichappear as veins running along the perimeters ofneighboring curd particles. Unlike the interior ofthe curd particles, the junctions are almost de-void of fat, owing to the expression of fat glob-ules from the surfaces of the curd particles intothe whey during cutting of the coagulum andcooking. Similarly, chip junctions in Cheddarand related dry-salted varieties have a highercasein:fat ratio than the interior and are clearlydiscernible upon examination of the cheese bylight microscopy (Figure 13-4) or by visual in-spection of cheese, especially in the case ofcheeses exhibiting the defect of seaminess. Dif-ferences in cheese composition between junc-tions and the interior of the curd particlesprobably lead to differences in the molecular at-tractions between the casein molecules in the in-terior and exterior of curd particles.

Various physicochemical changes occur inthe structural components of the matrix duringmaturation. These changes are mediated by theresidual rennet, microorganisms and their en-

zymes (see Chapter 11), and changes in mineralequilibria between the serum and paracaseinmatrix. The type and extent of physicochemicalchanges depend on the cheese variety, cheesecomposition, and ripening conditions. Thechanges include the following:

• The conversion of residual lactose to lacticacid and/or acetic and propionic acids.

• Hydrolysis of the caseins to peptides of vary-ing molecular weights and amino acids andcatabolism of amino acids to amines, alde-hydes, alcohols, and NH3 (Fox, O'Connor,McSweeney, Guinee, & O'Brien, 1996;Chapter 11).

• Hydrolysis of triglycerides to free fatty ac-ids, which may be degraded further to ke-tones and alcohols.

• Increased hydration of the paracasein medi-ated by factors such as its hydrolysis, theincrease in cheese pH, and solubilization ofcasein bound calcium. Solubilization ofCa2+ attached to the casein occurs whenthey are partially replaced by Na+, espe-cially when the concentrations of Na+ andCa2+ in the cheese moisture (serum) are low(i.e., 30 g/L and 4 g/L, respectively). Theincrease in casein hydration is paralleled bya physical expansion or swelling of theparacasein matrix that is clearly observedupon examination of the cheese by confocallaser scanning microscopy at various inter-vals during storage (see Figure 19-7).

• Coalescence of fat globules resulting in theformation of fat pools. This appears to oc-cur in all cheeses, as reflected by increasesin the level of fat that can be expressed fromthe cheese when subjected to hydraulicpressure or centrifugation (see Figures19-7 and 19-14) or exudes from the cheeseupon baking (Guinee, Mulholland, Mullins,& Corcoran, 1997; Kindstedt, 1995). Theincrease in free fat during maturation maybe due to the physical swelling of the pro-tein phase into spaces previously occupiedby fat, which forces the partially denudedfat globules closer together.

Page 328: Cheese Science

Figure 13-3 Light micrographs of Mozzarella (A), Gouda (B), and Edam (C) cheese showing curd granulejunctions, which appear as dark lines. Bar equals 5 mm.

C

B

A

Page 329: Cheese Science

Hence, cheese is, chemically, biologically,and biochemically, a dynamic system in whichthe structural components undergo storage-re-lated physicochemical and microstructuralchanges such as hydrolysis and hydration of thecasein, matrix swelling, fat coalescence, and/orfat hydrolysis. These changes aid in the conver-sion of fresh curd to a mature cheese and mark-edly influence its rheological, textural, func-tional, and flavor characteristics (see Chapters12 and 19). Thus, a ripening period is generallyrequired for all natural cheeses, apart from freshacid-curd varieties, before they attain the desired

attributes for the particular variety (e.g., flavor,aroma, and degree of meltability).

13.3 RHEOLOGICALCHARACTERISTICS OF CHEESE

13.3.1 Definitions of Different Types ofRheological Behavior Based on Creep andRecovery Experiments

The application of stress to a solid materialresults in deformation and strain. The appliedstress (denoted a or T) is the force per unit sur-

Figure 13-4 Light micrographs of Cheddar cheeses subjected to pressing in a Wincanton block former (A) or ina mold (B), as in traditional Cheddar manufacture. Curd chip junctions appear as heavy dark lines, and thejunctions of curd particles within the curd chips are also discernible as fine black lines. Bar equals 10 mm.

Page 330: Cheese Science

face area of the material. It may be applied to thesurface in a normal, perpendicular direction (a),resulting in compression (e.g., when a weight isplaced on top of a piece of material), or in exten-sion (e.g., when a weight is hung from the mate-rial) (Figure 13-5, diagram A). Alternatively,the stress may be applied tangentially, or paral-lel, to the direction of the surface (T), causingcontiguous parts of the material to slide relativeto each other in a direction parallel to their planeof contact (Figure 13-5, diagram B). The defor-mation is the change in distance between twopoints within the cheese mass as stress is ap-plied, and the strain is the fractional change in aparticular dimension (e.g., height or length).When the stress is applied normally, the ensuingstrain (denoted by 8) may be defined as Ah/h orA///, where h and / correspond to the originalheight or length of the sample and A/z and A/ tothe change in height or length. When the stress isapplied tangentially, the ensuing shear or shearstrain (denoted y) may be defined as the distance(A/) through which the point of applicationmoves divided by the distance (/) between themoving and stationary planes of the sample (Fig-ure 13-5, diagram B).

A solid material is described as being per-fectly elastic (or Hookean) if the strain is di-rectly proportional to the applied stress, with thea/8 curve passing through the origin (Figure13-6). Depending on how the stress is applied,two types of moduli (the proportionality con-stant between a and 8 or the slope of the a/8curve) may be obtained for a Hookean solid:

• The modulus of elasticity or Young's mod-ulus (E), where the force is normal to thearea bearing the stress, is expressed as a = Ex 8, with 8 = A///.

• The elastic shear modulus, also termed themodulus of rigidity (G or G% where theforce is parallel to the area bearing thestress, is expressed as T = G x y, with y =A///.

The moduli E and G for a perfectly elasticmaterial are independent of the rate at which thestress is applied. Hence, the stress-strain curve isalways linear. A plot, known as a creep curve,shows the variation of 8 with time upon the ap-plication and removal of a fixed stress (a) to anideal elastic material (where a is low enough notto fracture the material). An ideal elastic mate-

Figure 13-5 Application of stress (i.e., force F per unit area A) to a solid material. Stress may be applied (a) in adirection normal to the surface, resulting in uniaxial compression, or (b) parallel to the surface, resulting in shearcompression. In both cases, the deformation is A/ and the strain is A///.

(a) (b)

Page 331: Cheese Science

rial deforms instantly upon the application of aand recovers instantly to its original shape anddimensions when a is removed (Figure 13-7,diagram A). During the application of a, stressenergy is absorbed or stored by the structural el-ements of the material. Upon removal of a, thestored energy is released and the material re-gains its original dimensions instantly. Thisrheological behavior, which is a hypotheticalone, may be represented mechanically by asimple spring (Figure 13-7, diagram B), wherethe degree of extension (i.e., the strain) is di-rectly proportional to the mass hanging from thespring (and hence the force per unit area).

In contrast to an elastic solid, a fluid does notsupport a stress. Hence, the strain changes con-stantly as long as the stress is maintained. A ma-terial is defined as an ideal viscous (Newtonian)fluid if the rate of change of the strain (i.e., dy/dt,usually denoted y) is directly proportional to theapplied (shear) stress, a, with the a/y curve pass-ing through the origin (Figure 13-8). The pro-portionality constant—the slope between stressand strain rate (more usually denoted as shearrate)—is known as the coefficient of viscosity orjust viscosity (r|, where r| = a/y). A typical creepcurve for a Newtonian fluid shows that it startsto flow instantly at a fixed rate upon the applica-

tion of a constant shear stress and ceases to flowimmediately (but is permanently deformed)upon its removal (Figure 13-9, diagram A). Un-like a solid material, the stress energy is notstored but is dissipated in the form of flow, withthe strain being proportional to the time overwhich the stress is applied. This rheological be-havior may be represented mechanically by adashpot (i.e., a piston enclosed in a cylinderfilled with a viscous liquid; Figure 13-9, dia-gram B), where the rate of movement of the pis-ton (strain rate) is directly proportional to theapplied stress.

Like most solid and semisolid foods, cheeseexhibits characteristics of both an elastic solidand a Newtonian fluid and therefore is termedviscoelastic. The relationship between stress andstrain for these materials is not linear except atvery low strains, as discussed below. The stressincreases less than proportionally with strain, re-sulting in a curve that is concave downward(Figure 13-10). The rheological properties (i.e.,E, G, Tj) of viscoelastic materials differ fromthose of perfectly elastic or viscous materials inthat they are dependent on time (they are a func-tion of the time over which a fixed stress orstrain is applied) and the magnitude of the stress.However, upon the application of a strain that is

Strain, 8, arbitrary units

Figure 13-6 Relationship between stress (a) and strain (e) for an elastic solid.

Stre

ss, a

, arb

itrar

y un

its

Page 332: Cheese Science

sufficiently small so as not to induce permanentdamage or fracturing of the microstructure, vis-coelastic solid and semisolid food materials ex-hibit elastic behavior. The strain at which linear-ity between stress and strain is lost is referred toas the critical strain, which for most foods, in-cluding cheese, is relatively small (e.g., 0.02-0.05). At a strain less than the critical strain, therheological properties of viscoelastic materialsare time dependent. The typical change in strainwith time upon the application of a constantstress to a 3-month-old Cheddar cheese is shownin Figure 13-11. The curve is termed a creepcompliance-time behavior curve, where creepcompliance (J) is the ratio of the strain to the ap-plied constant stress. Three characteristic re-gions are evident in the creep curve:

1. elastic deformation (region AB), wherethe strain is instantaneous and fully re-versible (the strain is referred to as elasticcompliance [J0])

2. viscoelastic deformation (region BC),

where the strain is partly elastic andpartly viscous (the strain is referred to asretarded elastic compliance [JR], and theelastic component of the strain recoversslowly upon removal of the stress)

3. viscous deformation (region beyond C),where the deformation increases linearlywith time and is permanent (the strain,which is referred to as Newtownian com-pliance [JN], is not recoverable)

Upon removal of the strain at point D, the re-covery follows a similar sequence to straincreep, with three regions evident: an instanta-neous elastic recovery (DE), a delayed elasticrecovery (EF), and an eventual flattening. Thevertical distance from the flat portion of thecurve to the time axis is the nonrecoverablestrain per unit stress, which is related to theamount of structural damage to the sample dur-ing the test.

In the elastic region, the strands of the cheesematrix absorb and store the applied stress en-

Figure 13-7 Time-related change in the strain of an elastic solid upon the application or removal of a fixed stress(force [F] per unit area) (a). The behavior is mechanically represented by a single spring (b).

(a) (b)

Time

Stra

in, 8

Application of stress Removal of stress

F

Page 333: Cheese Science

Stress, a, arbitrary units

Figure 13-8 Relationship between shear stress (a) and strain rate(y) for an ideal viscous (Newtonian) liquid.

Shea

r rat

e, y

, arb

itrar

y un

its

Shea

r rat

e, y

TimeF

(a) (b)

Figure 13-9 Time-related change in the shear of an ideal viscous liquid upon the application or removal of afixed shear stress (force [F] per unit area) (a). The behavior is mechanically represented by a dashpot (b).

Application of stress Removal of stress

Page 334: Cheese Science

Figure 13-10 Schematic of time-related change in the deformation of a piece of cheese after subjecting it to aconstant stress at time tQ (—) and deformation recovery upon removal of the stress after a fixed time t} (—). Threetypes of deformation are apparent during the application of stress: elastic (AB), viscoelastic (BC), and viscous(beyond C). Similarly, upon removal of the stress (at D) three regions of deformation recovery are evident:elastic, viscoelastic (delayed), and lasting deformation as a result of viscous deformation.

to tl Time

delayedrecovery

elasticrecovery

Def

orm

atio

nC

D

lasting deformation

Stressapplied

Stressremoved

Creep Recovery

Cre

ep c

ompl

ianc

e, J

, Pa

"1 x 1

0"6

Time, s

Figure 13-11 Creep compliance and recovery of a 3-month-old Cheddar cheese. The various terms are dis-cussed in Section 13.3.

Page 335: Cheese Science

ergy, which is instantly released upon removalof the stress, enabling the cheese to regain itsoriginal dimensions. The extent and duration ofthe elastic region depend on the magnitude ofthe applied stress and the structural and compo-sitional characteristics of the cheese. The strainat which elastic behavior is lost for cheese de-pends on the variety but is generally in the orderof 0.02-0.05. At strains greater than the criticalstrain, the structure of the cheese is altered dueto breaking of bonds between structural ele-ments, which are stressed beyond their elasticlimit. Eventually, when the stress-bearing struc-tural casein matrix has fractured completely, thecheese flows.

Viscoelastic behavior may be represented bymechanical models that contain different ar-rangements of dashpots and springs in seriesand/or in parallel, such as Maxwell (spring anddashpot in series), Kelvin (spring and dashpot inparallel), or Burger bodies. The shape of thecreep curve for the latter model, which consistsof a Maxwell and a Kelvin body in series (Figure13-12, diagram C), indicates that it gives a muchcloser approximation of the rheological behav-ior of cheese upon the application of a small loador stress than that obtained using the model ofeither an elastic solid or a Newtonian fluid on itsown (see Figures 13-7, 13-9, and 13-12).

The viscoelastic nature of cheese implies thatthe ratio of elastic to viscous properties dependson the time-scale over which the deformation isapplied. At short time scales, cheese is essen-tially elastic, whereas after a long deformationtime, cheese flows, although very slowly in thecase of hard cheeses. However, even hardcheeses flow eventually when stressed and willnot recover completely upon removal of thestress. Failure to appreciate this characteristiccan often lead to alteration of shape, especiallybulging, during distribution and retailing, whencheeses of different consistencies are often hap-hazardly laid upon each other.

The flow of rigid materials is not alwaysreadily apparent because of the relatively longtime required to produce a notable deformation,and the notion may even appear somewhat ab-

stract. However, there are many natural ex-amples that reveal the slow flow of rigid materi-als over very long time periods, equivalent tocreep experiments where the time approachesinfinity—for example, the flow of glass in win-dow panes due to the force of gravity, as indi-cated by the increase in the thickness of indi-vidual window panes of old buildings withincreasing distance from the top of the pane.

13.3.2 Measurement of the RheologicalBehavior of Cheese

Rheological measurements on cheese may beclassified into three main types:

1. Empirical tests. In these tests, the testconditions are arbitrary and the aim is toobtain a number that gives a vague indi-cation of the textural characteristics of thecheese (e.g., its hardness). Often used bycheese graders, empirical tests may bebased on the subjection of cheese to astress or strain that results in visual frac-ture (e.g., rubbing cheese between the fin-gers until it becomes pliable or cutting thecheese with a knife). Alternatively, thecheese may be subjected to a stress orstain that causes no visible fracture (e.g.,pressing one's thumb or a ball into cheeseand recording the resistance either men-tally or by an instrument [see Chapter23]).

2. Creep and stress relaxation experiments.These allow the quantitative determina-tion of precise rheological characteristics,as described in Section 13.3.1.

3. Large-scale deformation tests. These al-low determination of the force required tofracture the cheese at given deformationrates. The test conditions may be de-signed to simulate the conditions towhich the cheese is subjected in practice,such as during mastication or during sizereduction processes (e.g., portioning). Inthis case, the test is described as beingquantitative in relation to the process it is

Page 336: Cheese Science

attempting to simulate. Alternatively, thetest conditions may not simulate thestress or strain to which the cheese is sub-jected in a particular application. Al-though physical quantities (e.g., fracturestress or fracture strain) may be measuredduring the latter test, the test is describedas being nonquantitative for that applica-tion, and the quantities obtained are oflittle use in trying to predict how thecheese will behave Theologically in thatapplication.

Empirical Tests

These tests involve subjecting a cheesesample to a stress or strain by various techniques(e.g., inserting a penetrometer). The differenttypes include

• penetration tests, where the force requiredto insert a probe a given distance into thecheese or, alternatively, the depth of pen-etration by a probe under a fixed load for a

given time is measured (e.g., penetrometertest)

• compression tests, where the extent of com-pression under a constant load for a speci-fied time is measured (e.g., ball compressortest)

The penetrometer test measures the depth towhich a penetrometer (e.g., needle or cone) canbe forced into a cheese under a constant stress.As the needle or cone penetrates the cheese, thecheese in its path is fractured and forced apart.The progress of the penetrometer is retarded toan extent that depends on the hardness of thecheese in its path, the adhesion of the cheese toits surface (which increases with the depth ofpenetration into the cheese), and its surface areaof contact with the cheese (regulated by thethickness of the needle or angle of the coneused). Eventually, the retardation stresses be-come equal to the applied stress and penetrationceases. The test, which is used to provide an in-dex of hardness (i.e., resistance of a surface to

(A) (B) (C)Figure 13—12 Mechanical models (bodies) representing viscoelastic behavior and their corresponding creep andrecovery curves: Maxwell body (A), Kelvin body (B), and Burger body (C).

Time Time Time

Stra

in, E

Stra

in, e

Stra

in, e

W WW

Page 337: Cheese Science

penetration), is suitable for closed-texturedcheeses such as Gouda and Mozzarella, whichare macroscopically homogeneous. Conversely,it is unsuitable for open-textured cheeses withsmall mechanical openings or eyes (e.g., Tilsitand Gruyere) or cheeses that are macroscopi-cally nonuniform owing to the presence of chipboundaries (e.g., Cheddar).

The ball compressor test measures the depthof indentation after a given time made by a smallball or hemisphere when placed under a givenload (stress) on the cheese surface (Figure13-13). Hence, the test simulates the action of agrader who, in the course of examination,presses the ball of his or her thumb into thecheese. The depth of penetration has been useddirectly as an index of firmness. Alternatively,by making a number of simple assumptions,testers may use it to calculate a modulus, analo-gous to an elastic modulus (G, given by theequation G = 3M[16(&D3)]1/2, where M is theapplied force and R and D are the radius anddepth of the indentation, respectively). Theabove tests, which are described in detail byPrentice et al. (1993), are generally nondestruc-tive and can be performed on the whole cheese atseveral locations. However, they provide only

an overall measure of the many different facetsof rheological behavior.

Quantitative Tests (Creep and StressRelaxation)

In quantitative tests, the cheese is subjected toa very small stress or strain so as to minimizestructural alteration. The ensuing stress or strainis measured dynamically over time, and the re-sultant curves may be used for the measurementof precise rheological properties such as defor-mation (at a given stress), modulus of elasticity(E), and relaxation time required for stress alle-viation to a certain value of the maximum stressapplied. Thus, they provide information on theviscoelasticity and structure of the cheese. Ahigh value for E suggests that the cheese matrixis elastic and continuous, with strong intermo-lecular attractions, whereas a low value of E in-dicates that the matrix is less elastic and weakeras a consequence of proteolysis and casein hy-dration, for example.

The viscoelastic behavior of cheese is mea-sured using two quasi-static tests, creep and stressrelaxation. In a creep test, as described in Section13.3.1, the sample is instantaneously subjected toa low, constant stress (e.g., 10-50 Pa), and the

Depth gauge

WEIGHT

hemisphere

Cheese surface

cheese interior

Figure 13-13 Schematic diagram of the ball compressor. A hemisphere placed on the surface of the cheese isforced into the cheese surface under the action of a weight.

Page 338: Cheese Science

ensuing increase in deformation or strain is mea-sured over time (see Figure 13-11). The constantstress may be applied using a control stress rhe-ometer (see Chapter 6), where the test involvesplacing a cheese sample between two parallelplates and applying a load to the top plate. A stressrelaxation test involves instantaneous applicationof a low constant deformation or strain to a cheesesample (e.g., < 0.05) and measurement of thedecrease in stress overtime. The constant strain isusually applied by placing the sample betweentwo parallel plates of an instrument such as theInstron Universal Testing Machine and allowingthe top plate to compress the sample to a particu-lar deformation (Figure 13-14).

Force Compression Tests betweenParallel Plates

In practice, cheese is subjected to largestresses and strains (i.e., > 0.05), which usuallyresult in fracture. The use of cheese as an ingre-dient may involve precutting of large blocksand/or comminution (e.g., by a conveying augercrushing and forcing precut cheese through dieplates with narrow apertures). Other industrialoperations that result in the fracture of cheeseinclude shredding (e.g., cutting into thin, narrowcylindrical pieces or shreds, e.g., 2.5 cm inlength and 0.4 cm in diameter), dicing (cuttinginto very small cubes, e.g., 0.4 cm3), or cubing(2.5 cm3). In the home, slicing of table cheeseideally results in a clean cut along the path of theknife blade. Also, cheese, when eaten, is submit-ted to a number of fracture strains that reduce itto a pulp capable of being swallowed. First thecheese is cut by the incisors, then it is com-pressed by the molars during chewing andsheared between the palate and the tongue. Mostof these actions involve a combination of com-pressive and shear forces. Thus, prediction ofhow cheese behaves under large stresses andstrains is desirable.

Force-compression tests measure the dynamicrheological behavior of cheese as the strain isincreased over time to values that generally re-sult in fracture and flow of the cheese beingtested. The tests, which are today commonplace

for the measurement of fracture properties, aremore or less empirical, depending on the magni-tude of the stresses and strains involved. At verylow strains (< 0.02), the structure of the cheese iseffectively unaltered during the measurement,and the values obtained may be then used toquantify the elastic modulus (usually denotedcompression modulus). Compression tests usu-ally involve large strains (up to 0.8) that causecomplete breakdown of the cheese structure. In-cidentally, these strains simulate the masticationof cheese, where compression of the cheese be-tween the molars is typically about 70%. Thelarge-strain compression test is useful in that itprovides a measure of the stresses and strains atwhich cheese is effectively elastic, at which itfractures, and at which it flows. These data maybe useful for designing size-reduction equip-ment, optimizing the settings of a piece of equip-ment for a particular type or batch of cheese, orgauging how difficult or how easily a piece ofcheese is degraded when compressed under con-ditions similar to those in the mouth.

In a large-strain force-compression test, a cy-lindrical or cubical cheese sample of fixed di-mensions and at a fixed temperature is placedbetween the two parallel plates of an instrumentsuch as the Universal Testing Machine (Figure13-14). One plate, denoted the base plate, isfixed; the other, denoted the cross-head, is pro-grammed to move at a fixed rate (typically 50cm/min) and compresses the cheese sample to apredetermined level (typically to 25% of itsoriginal height). The force required to compressthe sample to a given level (or percent compres-sion) is a measure of two parameters: the forcerequired to overcome the surface friction of thesample and the force required to actually com-press the cheese sample. The force (F) de-veloped during compression is recorded as afunction of distance or displacement (AL); alter-natively, the force may be converted into stress(a) and the displacement into strain (e) as fol-lows: a (Pa) =FIA (NIm2), where A is the surfacearea of the cheese sample in square meters, and 8= AL/Z, where AL (m) is the distance throughwhich the cross-head moves.

Page 339: Cheese Science

A typical force (stress)/displacement (strain)curve for mature Cheddar reveals a number ofdistinct regions (Figure 13-15):

• Region AB. The stress increases propor-tionally with e. If compression is stopped inthis range, the cheese returns to its originaldimensions. The slope of this linear regioncorresponds to the compression modulus, E(i.e., E = a/e), which is an indication of thetexture profile descriptors springiness andelasticity. Alternative (rarely used) expres-sions for E are yield stress and yield strain(the stress and strain at which elastic recov-ery is lost). The value of E decreases withassay temperature and moisture content and

increases upon elevation of fat and salt-in-moisture levels and with maturity (Luyten,1988).

• Region BC. The stress increases less thanproportionally with 8 (i.e., the curve is con-cave downward). The slightly lower slopeof the curve in this region compared withthat in region AB is probably due to theformation of microcracks that allow somestress to be dissipated but do not spreadthroughout the sample. However, the curveis still relatively linear, indicating that thecheese is still largely elastic and recoversalmost completely when the compressionceases.

Figure 13-14 Compression instrument (Instron Universal Testing Machine, Model 112). A cylindrical sampleof cheese is compressed at a fixed rate to a preset deformation between two parallel plates, the bottom (base) plateand the top plate (cross-head). During compression, the force exerted by the cheese on the moving top plate ismeasured by a load cell interfaced with a personal computer for continuous data capture.

Page 340: Cheese Science

• Region CD. The slope of the a/£ curve de-creases markedly. The cheese begins tofracture at C, as cracks grow and spreadthroughout the sample at an increasing rate.Eventually, at D, the rate of collapse of thestress-supporting paracasein matrix over-takes the build-up of stress within the ma-trix through further compression, and apeak stress, denoted the fracture stress, isreached. The fracture stress, a/, and strain,8/, are measures of the stress and strain, re-spectively, required to cause complete frac-ture of the sample into several individualpieces. If the compression is stopped at C,the sample may recover partially and maystill be in the form of one mass.

• Region DE. The stress decreases with fur-ther compression, reflected by the negativeslope of the curve, owing to the collapse ofthe stress-bearing structure. The samplefractures into pieces that spread, to a greateror lesser degree, over the base plate of theinstrument. Hence, the force per unit sur-face area decreases.

• Region EF. The stress increases as thecross-head begins to compress the frag-mented pieces of cheese. The stress at theend of the compression (point F) is a mea-sure of firmness, as judged in the first biteduring mastication.

The various quantities obtained from theforce-displacement curve and their interpreta-tion are given in Table 13—1.

13.3.3 Relationships Between RheologicalQuantities and Cheese Characteristics

The relationships between the outputs fromvarious rheological measurements (e.g., curvesfrom creep, recovery, and large-scale compres-sion tests) and rheological characteristics, as de-noted by textural descriptors used in practice,are shown in Table 13-2. An increase in E im-plies that the cheese is more elastic, springy, orrubbery at low a. Likewise, a cheese that dis-plays a high a/may be referred to as tough, rub-bery, firm, or strong. Conversely, a cheese with

Compression, %

Figure 13-15 Force compression curve for a 6-month-old mature Cheddar cheese compressed to 75% of origi-nal height at a rate of 5 cm/min. Several regions are identifiable: AB, BC, CD, DE, EF. (See Section 15.3 fordetails.)

Forc

e, N

Page 341: Cheese Science

a low a/could be described as soft or weak-bod-ied. Cheeses that exhibit a high £/are frequentlyreferred to as being "long," while those with lowEf are described as being "short" or brittle. Acheese that requires a high a to achieve a given eis described as firm.

13.3.4 Factors That Influence theRheological Characteristics of Cheese asMeasured Using Force-Compression Tests

The exact shape of the curve and the magni-tude of the various rheological quantities ob-tained from force-deformation curves depend ontest conditions and cheese variety, composition,and degree of maturity (Culioli & Sherman,1976; Dickinson & Goulding, 1980; Prentice etal., 1993; Shama & Sherman, 1973; Sherman,1988; Vernon Carter & Sherman, 1978; Visser,1991).

Test Conditions

Increasing the compression or strain rate (i.e.,speed of the cross-head) in the range 5-100 cm/min results in progressive increases in E (a/), thestress required to achieve a given level of com-pression and firmness (Figure 13-16). More-over, the magnitude of these increases is varietydependent. For example, a/ increases more rap-idly with the compression rate for Cheddar thanfor Cheshire (Figure 13-17). Hence, compara-tive studies on two or more cheeses can produceincorrect conclusions if the compressions areperformed at different compression rates. How-ever, if the effect of the strain rate for differentvarieties under standard conditions (e.g., tem-perature and sample dimensions) is known, thenit is possible to compare results obtained at dif-ferent strain rates by adjusting the results to acommon strain rate. The strain rate and the level

Table 13-1 Rheological Parameters Derived from a Stress-Strain Curve Obtained from a Force Com-pression Test

Rheological Quantity

Compression modulus

Fracture stress (yieldstress)

Fracture strain (compres-sion strain, fracturestrain, yield displace-ment)

Maximum strain (maxi-mum stress, maximumload)

Abbreviation

E

GI

Zf

^max

Interpretation

Measure of true elasticityat a low strain

Stress required to causefracture and collapsethe cheese matrixbeyond point ofrecovery

Strain at which cheesecollapses completely

Stress required to reach agiven deformation

Rheological Characteristicto Which Quantity Is

Related

Degree of elasticity orspringiness,rubberiness

Strength, toughness,brittleness

Shortness, longness,crumbliness

Firmness, hardness

Note: Refer to Figure 13-15 for details.

Page 342: Cheese Science

Rheological Property

Elasticity (rubberiness)

Springiness

Elastic fracturability

Brittleness

Firmness (hardness)

Longness

Toughness(chewiness)

Softness

Plastic fracturability

Shortness

Adhesiveness(stickiness)

Crumbliness

Shear thickening

Shear thinning

Definition

Tendency of cheese to recover its originalshape and dimensions upon removal of anapplied stress

Tendency to recover from large deformation(strain) after removal of deforming stress

Tendency of hard cheese to crack, with verylimited flow (only near the crack) of thecheese (after fracture, the broken surfacescan be fitted closely to each other)

Tendency of (hard) cheese to fracture at arelatively low permanent deformation

High resistance to deformation by applied stress

The failure of cheese to fracture until a relativelylarge deformation is attained

A high resistance to breakdown upon mastica-tion

Low resistance to deformation by applied force

The tendency of cheese to fracture accompa-nied by flow

The tendency to plastic fracture at a smalldeformation; low resistance to breakdownupon mastication

The tendency to resist separation from anothermaterial it contacts (e.g., another ingredient ora surface such as a knife blade or palate)

The tendency to break down easily into small,irregular shaped particles (e.g., by rubbing)

The tendency to increase in apparent viscositywhen subjected to an increasing shear rate(especially upon heating)

The tendency to exhibit a decrease in apparentviscosity when subjected to an increasingshear rate

Cheese Type DisplayingProperty

Swiss-type cheese, low-moisture Mozzarella

Swiss-type cheese, low-moisture Mozzarella

Parmesan, Romano,Gruyere

Romano, Parmesan,Gruyere

Cheddar, Swiss-typecheese, Romano,Parmesan, Gouda

Mozzarella, Swiss

Mozzarella, String cheese

Blue cheese, Brie, Creamcheese

Mature Cheddar, Bluecheese, Appenzeller,Chaumes, Raclette

Camembert, Brie

Mature Camembert

Cheshire, Wensyledale,Blue cheese, Stilton,Feta

Cream cheese (whenheated)

Quarg (especially at lowtemperatures, i.e.,<4°C)

Table 13-2 Rheological Properties of Raw Cheese and Their Definitions, Showing the Relationship toRheological Quantities as Measured lnstrumentally

Page 343: Cheese Science

Forc

e, N

Relative compression, %

Figure 13-16 Force as a function of compression for samples of Gloucester cheese compressed at 5 (Q), 20 (•),50 (O), and 100 (•) cm/min.

Stre

ss,

a, K

Pa

Compression rate, s"025

Figure 13-17 Effect of rate of compression on fracture stress (Oy) for different cheese types: Double Gloucester(•), Cheddar (O), Leicester (•), and Cheshire (Q).

Page 344: Cheese Science

of compression that are chosen depend on whatthe test is designed to simulate. Comparison ofinstrumental force-compression data obtained atdifferent compression rates with consumers'evaluations, both orally and using the fingers, ofa wide range of solid foods suggests that lowcompression levels (< 50%) correspond to mea-surement of firmness or hardness by squeezingthe cheese between the fingers. Higher compres-sion levels (> -70%) correspond to firmness as

judged by chewing the cheese. Evaluation offood by compressing between the fingers de-pends more on the elasticity of the cheese,whereas in the mouth it depends on the cumula-tive contributions of elasticity, fracture, andflow.

Increasing the height of cylindrical samplesinto the range 0.75-3.5 cm results in reductionsin Of and the force required to achieve a givencompression and an increase in 8/(Figure

Forc

e, N

(A)

(B)

For

ce, N

Compression, %

Figure 13-18 Effect of sample height on the force-compression behavior of cylindrical samples (2.5 cm diam-eter) of Gouda cheese compressed at 5 cm/min (A) and 50 cm/min (B). Sample height: 0.75 (•), 1.5 (O), 2.5 (A),and 3.5 (A) cm.

Page 345: Cheese Science

13-18). The effects are more pronounced as thecompression rate is increased into the range5-50 cm/min. Similarly, the force required tocause a given deformation increases as thediametenheight ratio of the cylindrical samplesis increased, a trend that is more accentuatedwith the level of compression (Figure 13-19).The influence of sample dimensions is related tosurface friction (between sample and instrumentplates), which is associated with barrel deforma-tion of the cheese as it is compressed. Barrel de-formation refers to the shape frequently assumedby a cylindrical sample during compression—progressive increase in diameter from the flatsurfaces (in contact with the plates) to a maxi-mum at the central region of the cylinder (Figure13-20). Experiments with cylindrical samples ofGouda cheese (height, 2.5 cm; diameter, 2.5 cm)showed that barelling, which was first observed

at 20% compression, increased with level ofcompression in the range 20-80%, especially athigh compression rates (> 50 cm/min). Barreldeformation restricts the expansion of the flatfaces of the sample upon compression and re-sults in the development of a transverse com-pressive stress in addition to the axial compres-sive stress and thereby contributes to surfacefriction. In the absence of surface friction (whichcan be reduced to nothing by lubricating the flatsurfaces of the cheese sample with mineral oilprior to compression), the surface area of the flatsections increases more than in the presence ofsurface friction. Hence, as the height of thesample decreases during compression, the actualsurface area over which the force is applied in-creases in the absence of surface friction; conse-quently, the actual stress does not increase asrapidly (Sherman, 1988). The transverse stress

Forc

e, N

Relative compression, %

Figure 13-19 Effect of diameter:height ratio on the force-compression behavior of cylindrical samples (1.64 cmdiameter) of German loaf cheese. Sample height (cm) and diameter: height ratio are 1.65 and 1.0 (A), 1.3 and 1.25(O), and 1.1 and 1.5 (•).

Page 346: Cheese Science

associated with barreling decreases with dis-tance from the flat surface and reaches zero at adistance that is approximately equal to thesample diameter. Hence, the compressive stressdecreases as the diameteriheight ratio (andhence height) of cylindrical samples is reduced.

Sample shape (i.e., cubical or cylindrical) alsohas an effect on the shape of the force-compres-sion behavior. Generally, cubes require a higherstress to achieve a given percent compressionthan cylinders of comparable volume, especiallyat compression levels greater than 50%. Thistrend may be related to frictional effects.

Increasing the sample temperature results in amarked decrease in the elastic modulus, fracturestress, and firmness (Figure 13-21). This effectis attributed to liquefaction of the fat fraction,which contributes to lubrication of fracture sur-faces, and a probable reduction in the surfacefriction between the sample and the instrumentplates as liquid fat is exuded at the flat surfaces.

Cheese Structure, Composition, andMaturity

The viscoelasticity of cheese results from theinteractive rheological contributions of its indi-vidual constituents—protein, fat, and moisture.

The structure, which represents the way inwhich the individual constituents coexist, andthe physical nature of the constituents (i.e.,whether solid or liquid) are major determinants.

Upon the application of a stress to a cheeseproduct, the matrix will at first limit the defor-mation. As the concentration of casein in thecheese matrix increases, the intra- and inter-strand linkages become more numerous and thematrix displays greater elasticity and is more dif-ficult to deform (Figure 13-22). Hence, re-duced-fat Cheddar, which contains a higher con-centration of structural matrix per unit volumethan full-fat Cheddar, is firmer and has a highera/than the latter (Figure 13-23). Factors thatpromote weakening of the casein matrix reducethe stress required to cause a given deformation.Hence, the a/ and firmness of cheese generallydecrease with ripening time owing to hydrolysisand hydration of the casein, both of which pro-cesses contribute to disintegration of the caseinmatrix (Figure 13-24). The structure of cheese ismarkedly weakened by the early hydrolysis ofOC8I-casein by residual chymosin at the Phe2s-Phe24 bond. The sequence 1-23 of asi-casein isstrongly hydrophobic and interacts with the hy-drophobic regions of other asr and (3-casein

Figure 13-20 (a) Schematic representation of barrel compression of a cylindrical cheese sample during com-pression, showing a progressive increase in diameter from the surface to the center of the sample. Friction be-tween the cheese and the compression plate (cross-head) prevents the cheese surface from spreading, (b) Sche-matic representation of ideal compression.

(b)

(a)

Increasing % compression

Page 347: Cheese Science

Compression, %

Figure 13-21 Effect of temperature on the force compression behavior of Gouda cheese compressed at 50 cm/min at 1O0C (•), 150C (O), and 2O0C (A).

Forc

e, N

Stre

ss,

a, k

Pa

Protein ,g/10Og

Figure 13-22 Relationship between protein content and cheese firmness. Data derived from the testing of 10different types of hard cheese and a pasteurized processed Cheddar cheese.

Page 348: Cheese Science

molecules and thus contributes to the overallcontinuity and integrity of the matrix (Creamer& Olson, 1982).

The contribution of fat to the rheologicalproperties of cheese depends on its physical state

and therefore on the temperature, which controlsthe ratio of solid fatliquid fat. At low tempera-tures (< 50C), where the milk fat is predomi-nantly solid, the fat adds to the elasticity of thecasein matrix. The solid fat globules limit defor-

Ripening time, day

Figure 13-23 Effect of fat level [6 (•), 17 (O), 22 (A), and 33 (A) g/100 g] on the firmness (A) and fracturestress (B) of Cheddar cheese compressed at 5 cm/min.

Frac

ture

stre

ss, a

fl kP

aFi

rmne

ss, N

(A)

(B)

Page 349: Cheese Science

mation of the casein matrix, as the deformationof the latter also requires deformation of the fatglobules enmeshed within its pores. As the pro-portion of liquid fat increases, the fat behavesmore like a fluid and confers viscosity ratherthan elasticity or rigidity on the cheese. More-over, liquid fat acts as a lubricant on fracture sur-faces of the casein matrix and thereby reducesthe stress required to fracture the matrix. Hence,for a given fat content, increasing the assay tem-perature during compression results in a markeddecrease in the elastic modulus, fracture stress,and firmness (Figure 13-21). An increase in thefat-in-dry-matter level of cheese (while retainingthe other compositional parameters constant) isparalleled by a decrease in a/, with the effect be-coming more pronounced as the temperature isincreased. Generally, an increase in the fat con-tent of cheese is accompanied by decreases inthe levels of protein and moisture as well as de-creases in the a/and firmness (Figure 13-23).

The third major component of cheese is mois-ture, which acts as a plasticizer in the proteinmatrix, thereby making it less elastic and moresusceptible to fracture upon compression. Thus,increasing the moisture content of cheese resultsin decreases in E, a/, and firmness (Figure13-25). The e/ increases slightly with moisturecontent to an extent dependent on cheese pH andmaturity (Creamer & Olson, 1982; Luyten,1988).

Other compositional factors also influence therheology of cheese. Increasing the pH into therange 5.0-5.2 reduces both E and a/ (Figure13-26). In contrast, increasing the pH into therange 5.2-5.6 results in a marked increase in a/(to values much higher than those at pH < 5.2)and a slight increase in E. The e/for young (1-week-old) Gouda cheese was at its maximum atpH 5.2 and decreased upon lowering or raisingthe pH to 4.8 and 5.6. However, the pH at which8/is maximal increases with ripening time (e.g.,

Strain, e, -

Figure 13-24 Effect of age (1 week, O; 7.5 month, •) on the force-compression behavior of Gouda cheese(moisture, 41.6 g/100 g).

Stre

ss,

a, K

Pa

Page 350: Cheese Science

from ~ 5.2 for 1-week-old Gouda to ~ 5.4 for 3-month-old Gouda cheese). Differences in pHmay help to explain the different rheologicalcharacteristics exhibited by some commoncheeses upon compression. Low pH cheeses(e.g., Cheshire and Feta) generally tend to havelow Of and e/ values and to crumble into manypieces upon fracturing, whereas relatively highpH cheeses (e.g., Emmental and Gouda) exhibithigher a/and e/ values and tend to fracture intolarger pieces (Prentice et al., 1993). The effect ofpH probably ensues from its influence on

• the ratio of soluble calcium to colloidal cal-cium

• the degree of paracasein hydration, which ismaximal at about pH 5.2

• the types of intra- and interaggregate bonds

Moreover, the effect of pH appears to be relatedto other factors, such as the levels of moistureand salt in the cheese and the degree of proteoly-sis (Walstra and van Vliet, 1982).

The salt content of rennet-curd cheeses variesfrom about 0.79 g/100 g in Emmental to about 6

g/100 g in Feta; however, because salt is dis-solved in the moisture phase, the effective con-centration is much higher (~ 2 and 12 g/100 g,respectively). Increasing the salt-in-moisture inGouda-type cheese into the range 0-12 g/100 gwhile maintaining the other compositional pa-rameters relatively constant is associated withincreases in E and a/ (which is relatively constantin the range 3-7 g/100 g S/M) (Figure 13-27,graph A). The fracture strain, £/, increases slightlyto a maximum at an S/M of 4 g/100 g, then de-creases sharply to about half its maximum value,and thereafter plateaus at this value for S/M in therange 5-12 g/100 g (Figure 13-27, graph B).

The rheological characteristics of cheese mayalso be influenced indirectly by many other fac-tors that influence its composition and/or struc-ture, some of which are listed below:

• Seasonal changes in milk composition andthe state of its components. These changesare associated with the stage of lactation,quality of the cows' diet, health status of thecows, and on-farm husbandry practices.

Strain, e, -

Figure 13-25 Effect of moisture content [32.3 (•), 35.6 (O), 38.6 (A), 41.4 (A), 43.9 (•), and 46.2 (Q) g/100 g]on the force-compression behavior of 7.5-month-old Gouda-type cheese.

Stre

ss,

a, k

Pa

Page 351: Cheese Science

Hence, poor-quality, late lactation milkfrom cows on a poor-quality diet often re-sults in cheese with a high-moisture contentand low firmness.

• Cheesemaking conditions. For a givencheese variety, composition may vary withcheesemaking conditions, especially if themilk supply is predominantly from spring-calving cows fed on pasture. In the lattersituation, the level of casein in milk mayvary markedly throughout the cheesemak-ing season (e.g., in Ireland from ~ 22 g/kgin March to 31 g/kg in October). Hence,when rennet and starter are added on a vol-ume basis rather than on a casein basis, andcertain cheesemaking steps (e.g., setting,cutting, cut programs) are undertaken onthe basis of time rather than on the basis ofdefined criteria (e.g., pH and gel firmness),variations in cheese composition (espe-cially moisture content) may be expected.

Increasing the casein content of milk (e.g.,by ultrafiltration) enhances its curd-form-ing characteristics and reduces cheesemoisture. Hence, the fracture stress ofCheddar cheese increases linearly with in-creasing protein content of cheese milk(Figure 13-28).

• Genetic variants of milk protein, such as K-casein AA or BB variants, owing to theireffect on cheese composition (Walsh et al.,1998).

• Cheese-ripening conditions (e.g., tempera-ture and relative humidity, which influenceparameters such as the rate of casein degra-dation and/or moisture loss).

13.4 CHEESE TEXTURE

Cheese texture is a sensory characteristic andtherefore is directly measurable only by sensoryanalysis. Texture perceptions arise from a corn-

Strain, e, -

Figure 13-26 Effect of pH [5.02 (•), 5.1 (O), 5.2 (A), 5.43 (Q), and 5.58 (A)] on the force-compression behav-ior of 1-week-old Gouda cheese when composition was otherwise similar.

Stre

ss, a

, KPa

Page 352: Cheese Science

plex array of sensory inputs that occur both priorto and during food consumption. Consumptionof a piece of cheese involves the following seriesof events:

• Visual assessment (e.g., for eye distributionin a Swiss-type cheese, granularity of Cot-tage cheese, whiteness of Feta cheese). Thevisual perception may create an important

Salt-in-moisture, g/100 g

Figure 13-27 Effect of salt-in-moisture level [0.4 (A), 3.3 (A), 7.3 (O), 11.3 (•) g/100 g] on the force compres-sion-behavior (A) and fracture strain (B) of Gouda cheese when composition was otherwise similar.

Frac

ture

stra

in, s

f

Strain, e, -

(B)

(A)

Stre

ss, C

T, KP

a

Page 353: Cheese Science

first impression about the anticipated tasteand texture of the cheese. For instance, thegranular, dry (nonglossy) appearance ofParmesan may create the perception that itwill likely have a dry, grainy mouthfeel; thesurface sheen of a freshly sliced Cam-embert, or its absence, may be taken as evi-dence of its state of maturity, flavor, andpalatability.

• Assessment by touch (e.g., assessment ofthe resistance of a piece of String cheese totouch or that of a piece of Cheddar orCamembert when it is cut with a knife orpunctured by a fork).

• Eating, which occurs in four phases:1. Placement in mouth, including contact

with nerve endings on the tongue andcheeeks that contribute to sensations col-lectively referred to as somaesthesis(e.g., sensations of touch, pain, warmth,

and cold). The ingested cheese is com-pressed by the various parts of the oralcavity (palate and inside of lips andcheeks), and concomitantly the counter-stress exerted by the cheese is detectedby nerve endings.

2. An initial bite by the teeth (resistance tocutting by the incisors may be involved).

3. Chewing and mastication. The cheese iscompressed repeatedly, mainly betweenthe molars, moved toward the teeth bythe tongue, and mixed with saliva. Thisresults in diminution of the cheese to apasty bolus ready for swallowing. Evi-dence suggests that the tooth socket iselastic, permitting vertical and horizon-tal movement of the teeth (Brennan,1988). The degree of deformation of theperiodontal membrane (the structure thatsurrounds the tooth in the socket and at-

MiIk protein level, g/100 g

Figure 13-28 Effect of milk protein level on the fracture stress of 6-month-old Cheddar cheese.

Frac

ture

stre

ss,

afl k

Pa

Page 354: Cheese Science

taches it to the jaw bone) is the stimulusfor detection of the mechanical proper-ties of the cheese being eaten.

4. Swallowing of the cheese, during whichthe bolus is discharged out of the oralcavity.

In the process of eating, the cheese is sub-jected to cutting, shearing, and compressionforces that fracture and reduce it to a state readyfor swallowing. The objective of sensory textureevaluation is to translate the cumulative sensa-tions perceived into descriptors or scores.

13.4.1 Sensory Measurement of CheeseTexture

The sensory methods employed to measurefood texture are of three general types (Jack,Paterson, & Piggott, 1995; Powers, 1984):

1. attribute (or profiling) methods2. difference methods3. preference methods

Attribute methods, which are used mostwidely include

• the texture profile method (Brandt et al.,1963)

• descriptive analysis• free choice profiling

In the texture profile method, three categoriesof texture characteristics were proposed bySzczesniak (1963): (1) mechanical characteris-tics, which are related to the reaction of food tostress; (2) geometrical characteristics, which arerelated to size, shape, and orientation of particleswithin the food; and (3) other characteristics,which are related to the perception of the levelsof moisture and fat in the food. The mechanicalcharacteristics in turn were divided into primarycharacteristics, such as hardness, cohesiveness,viscosity, elasticity, and adhesiveness, and sec-ondary characteristics, such as brittleness,chewiness, and gumminess. The definitions ofthe mechanical characteristics are summarizedin Exhibit 13-1. Each of the above attributes is

scored on a nine-point equidistant scale relativeto a standard reference cheese.

Descriptive textural analysis involves thescoring of samples in terms of intensity using afixed consensus vocabulary of textural descrip-tors that are agreed upon jointly by the sensorypanel and the trainers during training sessions.Standard products (of the same generic type asthe product being evaluated) considered to ex-hibit one or more particular descriptors may beused to define characteristics, with the panelmembers being repeatedly exposed to the stan-dard products during training.

The free choice profiling method of textureevaluation is similar to descriptive analysis, ex-cept that the sensory descriptors are proposed bythe individual assessor, who quantifies the inten-sity of a particular attribute by assigning a scoreon a line scale.

13.4.2 Instrumental Methods for Evaluationof Cheese Texture

Cheese texture, being a sensory property, isultimately expressed in sensory terms or de-scriptors. However, trained texture panels maybe difficult and costly to establish and maintain.Moreover, instrumental methods are easier toperform, standardize, and reproduce and requirethe involvement of fewer trained people. Hence,much research effort has focused on the mea-surement of textural properties using instrumen-tal methods. Textural evaluation by instrumentalmethods is generally based on force-compres-sion tests that are designed to simulate the com-pression of cheese between the molars duringchewing. The first apparatus of this type, devel-oped for foods in general, was the forced-com-pression test using the General Foods Textur-ometer (Bourne, 1978; Friedman, Whitney, &Szczesniak, 1963), which was designed to simu-late the compression of food between the teeth(Figure 13-29). Essentially, a food sample isloaded onto a plate attached to a beam and thensubjected to a deforming force provided by atooth-shaped plunger moved by a wheel devicein a motion designed to simulate the vertical ac-tion of the jaw. When the plunger deforms the

Page 355: Cheese Science

sample, strain gauges attached to the sample-holding beam detect the movement of the beam,and a force-time trace is recorded. The sample issubjected to two successive deformations (re-ferred to as bites) so as to simulate more closelyrepetitive chewing action (Figure 13-30). Therelationships between the trace and textural de-scriptors, as described by Szczesniak (1963), areoutlined in Exhibit 13-2.

More recently developed instruments (e.g.,the Stevens Compressor Response Analyzer,Instron Universal Testing Machine, andTA.XT2 and TA.HD Texture Analyzers) alsouse the principle of compression for texturalevaluation. In addition, these instruments have arange of attachments (e.g., cones, needles,knives, and plungers) that facilitate the determi-nation of the response of the cheese to otherforces, such as cutting, at various deformationrates.

Effective simulation of sensory texture evalu-ation by instrumental compression tests necessi-tates test conditions (e.g., compression rates)that resemble those during consumption and thatare carefully standardized between samples(e.g., temperature, sample size and shape, andsize and shape of the compression plates). Re-search has revealed that most people compressfood to around 70% and chew at a rate of 40-80masticatory strokes per minute, indicating thatthe duration of the downstroke in a bite is around5 s, though it can be more or less depending onthe food (Sherman, 1988). Yet in many studiesthat attempted to relate instrumental and sensoryevaluations of textural attributes such as firm-ness or brittleness, lower levels of compressionand compression rates were used. As discussedin Section 13.3.4, the quantities derived fromcompression tests are markedly influenced bycompression rate.

Mechanical Characteristics

Primary properties

Hardness:

Cohesiveness:Springiness (elasticity):

Adhesiveness:

Secondary properties

Fractur ability (brittleness):

Chewiness:

Gumminess:

The force required to compress a cheese between the molar teeth (e.g.,hard and semi-hard cheese) or between the tongue and palate (e.g., forsoft spreadable cheese and process cheese spreads) to a given deforma-tion or to the point of penetration.The extent to which a cheese can be deformed before it ruptures.The degree of recovery of a deformed (strained) piece of cheese after thedeforming force is removed.The force required to remove cheese that adheres to the mouth (generallythe palate) during the normal eating process.

The force at which a cheese crumbles, cracks, or shatters when de-formed. Fracturability is the result of a high degree of hardness and alow degree of adhesiveness.The length of time or the number of chews required to masticate a cheeseto a state ready for swallowing. Chewiness is the product of hardness,cohesiveness, and springiness.A denseness that persists throughout mastication; energy required to dis-integrate a piece of cheese to a state ready for swallowing. Gumminess isa product of a low degree of hardness and a high degree of cohesiveness.

Exhibit 13-1 Definitions of Mechanical Properties of Cheese Using the General Foods Texture Profile.

Page 356: Cheese Science

eccentric rotatingwheel

beam

plunger

strain gaugessample

platebeam

Figure 13-29 Schematic representation of the structure of the General Foods Texturometer. The various termsare discussed in Section 13.4.

Figure 13-30 A typical texture profile of cheese obtained using the General Foods Texturometer. See Exhibit13-2 for interpretation of curve.

Time

Stre

ss, a

Page 357: Cheese Science

13.4.3 Comparison of Instrumental andSensory Evaluation of TextureNumerous comparative studies on the evalua-

tion of texture by compression and sensorymethods have been undertaken (Green,Marshall, & Brooker, 1985; Jack et al., 1993;Lee, Imoto, & Rha, 1978). Owing to differencesin test conditions (temperature, strain rate, level

of compression) and sample dimensions, the re-sults from the different studies have varied, in-cluding the degree of correlation between thesensory parameters (e.g., firmness and springi-ness) and the instrumental quantities. However,good correlations have been reported, in general,for hardness, chewiness, adhesiveness, andspringiness (Table 13-3).

Table 13-3 Relationships between Textured Parameters Obtained from Sensory Evaluation andRheological Parameters from a Universal lnstron Testing Machine for a Range of Different CheeseTypes

Sensory Parameter

Rheological Parameters Hardness Chewiness Springiness Adhesiveness

Compression stress (firmness) 0.95* 0.86* 0.27* 0.59Fracture stress 0.70* 0.82* 0.80* -0.80*Elastic recovery 0.17 0.36 0.97* -0.27Cohesiveness 0.87* 0.80* 0.21 0.44Adhesiveness 0.70* 0.82* 0.80* -0.80*

Note: Cheese types with a wide variety of textures were used: Cream cheese, Camembert, Mozzarella, Muenster, processedCheddar, Swiss cheese, and Cheddar cheeses of different degrees of maturity. Cheese samples (height, 1.0 cm; diameter, 1.5 cm)were compressed to 20% of original height (80% deformation) at a cross-head speed of 50 mm/min at room temperature, with 2consecutive cycles applied to the cheese sample (e.g., as in Figure 13-30). The Theological parameters are defined as follows:compression stress is the stress required for a given deformation; elastic recovery is the recovery in the height of the cheese cylinderafter the first bite and before the second bite (equivalent to distance B in Figure 13-30); cohesiveness is the ratio of the areas peak1 to peak 2 (A1/A2); and adhesive force is the force exerted on the ascending motion of the cross-head after the first bite.

*p<0.01.

Fracturability: Height of the first peak (Hl) in the first bite (Al).

Hardness: Height of the second peak (H2) in the first bite (Al).

Cohesiveness: Ratio of area on second bite to that on first bite (A2/A1).

Adhesiveness: Area (A3) of the negative peak formed when the plunger is pulled from the sample afterfirst bite, due to cheese adhering to the plunger.

Springiness: Difference between distance B (measured from the initial point of contact of the plungerwith the sample in bite 1 to contact with the sample in bite 2) and distance C (the samemeasurement made on a completely inelastic material such as clay) (B-C).

Chewiness: Hardness x cohesiveness x springiness (Al x [A2/A1] x [B- C]).

Gumminess: Hardness x cohesiveness x 100 (Al x [A2/A1] x 100).

Exhibit 13-2 Interpretation of the Force-Compression Curve from the General Foods Texrurometer

Page 358: Cheese Science

REFERENCES

Bourne, M.C. (1978). Texture profile analysis. Food Tech-nology, 32(2), 63-66, 72.

Brandt, M.A., Skinner, E.Z., & and Coleman, J.A. (1963).Texture profile method. Journal of Food Science. 4, 404-409.

Brennan, J.G. (1988). Texture perception and measurement.In J.R. Piggott (Ed.), Sensory analysis of foods (2d ed.).London: Elsevier Applied Science.

Creamer, L.K., & Olson, N.F. (1982). Rheological evalua-tion of maturing Cheddar cheese. Journal of Food Sci-ence, 47, 631-636, 646.

Culioli, J., & Sherman, P. (1976). Evaluation of Goudacheese firmness by compression tests. Journal of TextureStudies, 7, 353-372.

Dickinson, E., & Goulding, LC. (1980). Yield behavior ofcrumbly English cheeses in compression. Journal of Tex-ture Studies, 11, 51-63.

Fox, P.P., O'Connor, T.P., McSweeney, P., Guinee, T.P., &O'Brien, N. (1996). Cheese, physical, biochemical andnutritional aspects. Advances in Food and Nutrition Re-search, 39, 163-329.

Friedman, H.H., Whitney, J.E., & Szczesniak, A.S. (1963).The texturometer: A new instrument for objective texturemeasurement. Journal of Food Science, 28, 390-396.

Green, M.L., Marshall, R.J., & Brooker, B.E. (1985). Instru-mental and sensory texture assessment of Cheddar andCheshire cheeses. Journal of Texture Studies, 16, 351—364.

Guinee, T.P., Mulholland, E.O., Mullins, C., & Corcoran,M.O. (1997). Functionality of low-moisture Mozzarellacheese during ripening. In T.M. Cogan, P.F. Fox, & P.Ross (Eds.), Proceedings of the 5th Cheese Symposium.Dublin: Teagasc.

Jack, F.R., Paterson, A., & Piggott, J.R. (1993). Relation-ships between rheology and composition of Cheddarcheeses and texture as perceived by consumers. Interna-tional Journal of Food Science and Technology, 28, 293-302.

Jack, F.R., Paterson, A., & Piggott, J.R. (1995). Perceivedtexture: Direct and indirect methods for use in productdevelopment. International Journal of Food Science andTechnology, 30, 1-12.

Kindstedt, P.S. (1995). Factors affecting the functional char-acteristics of unmelted and melted cheese. In E.L. Malin& M.H. Tunick (Eds.), Chemistry of structure-functionrelationships in cheese. New York: Plenum Press.

Lee, C.-H., Imoto, E.M., & Rha, C. (1978). Evaluation ofcheese texture. Journal of Food Science, 43, 1600-1605.

Luyten, H. (1988). The rheological and fracture propertiesof gouda cheese. Unpublished doctoral dissertation,Wageningen Agricultural University, Wageningen, TheNetherlands.

Powers, J.M. (1984). Current practices and applications ofdescriptive methods. In RJ. Piggott (Ed.), Sensory analy-sis of foods. London: Elsevier Applied Science.

Prentice, J.H., Langley, K.R., & Marshall, RJ. (1993).Cheese rheology. In P.F. Fox (Ed.), Cheese: Chemistry,physics and microbiology (2d ed., Vol. 1). London:Chapman & Hall.

Rao, M.O. (1992). Classification, description and measure-ment of viscoelastic properties of solid foods. In M.O.Rao & J.F. Steffe (Eds.), Viscoelastic properties of foods.London: Elsevier Applied Science.

Shama, F., & Sherman, P. (1973). Evaluation of some tex-tural properties of foods with the Instron Universal Test-ing Machine. Journal of Texture Studies, 4, 344-353.

Sherman, P. (1969). A texture profile of foodstuffs basedupon well-defined rheological properties. Journal ofFood Science, 34, 458-462.

Sherman, P. (1988). Rheological evaluation of the texturalproperties of foods. Progress and Trends in Rheology, 11,44-53.

Szczesniak, A.S. (1963). Classification of the textural char-acteristics. Journal of Food Science, 28, 385-389.

van Vliet, T. (1991). Terminology to be used in cheese rheol-ogy. In Rheological and fracture properties of cheese[Bulletin No. 268]. Brussels: International Dairy Federa-tion.

Vernon Carter, EJ., & Sherman, P. (1978). Evaluation of thefirmness of Leicester cheese by compression tests withthe Universal Testing Machine. Journal of Texture Stud-ies^, 311-324.

Visser, J. (1991). Factors affecting the rheological and frac-ture properties of hard and semi-hard cheese. In Rheo-logical and fracture properties of cheese [Bulletin No.268]. Brussels: International Dairy Federation.

Walsh, C.D., Guinee, T.P., Reville, W.D., Harrington, D.,Murphy, JJ., O'Kennedy, B.T., & Fitzgerald, R.F.(1998). Influence of K-casein genetic variant on rennetgel microstructure, Cheddar cheesemaking properties andcasein micelle size. International Dairy Journal, 8, 707—714.

Walstra, P., & van Vliet, T. (1982). Rheology of cheese. InBulletin No. 153. Brussels: International Dairy Federa-tion.

Page 359: Cheese Science

14.1 INTRODUCTION

As discussed in Chapter 11, the ripening ofcheese, and hence its quality, is due to the ac-tivity of microorganisms and enzymes fromfour or five sources. Therefore, it seems like itshould be possible to produce premium-qualitycheese consistently by controlling these agents.However, in spite of considerable research andquality control efforts, it is not yet possible todo so.

A very wide and diverse range of factors in-teract to affect the composition of cheese curdand hence the quality of the final cheese; wehave attempted to summarize these in Figure14-1. Some of these factors or agents can bemanipulated easily and precisely, whereas oth-ers are more difficult, or perhaps impossible, tocontrol. It should be possible to apply the prin-ciples of hazard analysis critical control point(HACCP) to cheese production. However, theprecise influence of many of the factors includedin Figure 14-1 on cheese ripening and qualityare not known precisely, and many of the factorsare interactive. The interactions between theprincipal factors that affect cheese quality werereviewed by Lawrence and Gilles (1980). Figure14-1, which is more comprehensive than thescheme presented by Lawrence and Gilles(1980), is not presented as definitive, but ourhope is that it may stimulate others to applyHACCP principles to cheesemaking.

14.2 MILK SUPPLY

It is well recognized that the quality of themilk supply has a major impact on the quality ofthe resultant cheese. Three aspects of qualitymust be considered: microbiological, enzymatic,and chemical.

14.2.1 Microbiology

In countries with a developed dairy industry,the quality of the milk supply has improvedmarkedly during the past 30 years. Total bacte-rial counts are now usually below 20,000 cfu/ml ex-farm. The total bacterial count probablyincreases during transport and storage at thefactory, but growth can be minimized by ther-mization (650C x 15 s) of the milk supply,which is standard practice in some countries(see Chapter 4).

Although many cheeses are made from rawmilk, most cheese is made from milk pasteurizedat or close to 720C x 15 s. If produced fromgood-quality raw milk and subsequently han-dled under hygienic conditions, pasteurized milkshould have a very low total bacterial count andtherefore represents a very uniform raw materialfrom a microbiological viewpoint.

14.2.2 Indigenous Enzymes

Milk contains as many as 60 indigenous en-zymes (see Andrews et al., 1992), but the signifi-

Factors That Affect Cheese Quality

CHAPTER 14

Page 360: Cheese Science

X*"*"" "" -v Species^^_ ^^^ ^Composition) Breed

X^ ""X ^ .^^ Stage of Lactationf Somatic cell count) ^^ Plane of NutritionN*- ^,,^S ^\^^ Animal Health

/C7. ,. i . , . T^X N. *""**- Composition^Microbiological QuahtJ) ^^ ^ ^ ^ -x^v. . casein

^v_ ^^^ """ 1 IJA w IV/TT IT I -^ Public Health and Safety) ^NSSsvx -fat^ RAWMILK I-*A V -calcium/^ X -^ PH

I ^1 1 fnatural creaming " enzymes

Thermization I centrifugalOn-farm C hygiene Standardization -^-4 milk powder

Pasteurization I TJF

Transport, temPerature f t ^^^__^__^time [CIlEESE MILKl | Selection cntena

In-factory ' 1 ' Acid production^ Colour phaSe sensitivity

CaQ Ripening propertiesQTJL R&te °f ^ysig

Starter culture^^Secondary/adjunct culture.^—^fype^Rennet ^-—^

Vs^^AmoMw^xGel strength -«^—— i i ^^ ^^subjective/objective assessment | COAGULUM |

I - cheese yield ~ ^ - cheese yieldX - curd composition * - curd composition

Chee!q»ality |CURDS/WHE^ Mg^Sta,^^j^ \ - acidification

- Cook *^^^ I - retention of coagulant- syneresis ^- curd composition . Agitate,.- curd structure , -^ Acidification - —^curd syneresis- retention of coagulant D ^ curd C0mposiiton-solubilizationofCCP ^s i m

• T . ^^ wheyICURDS I

- Acidification- Dehydration- Texturization?- Salting?- Moulding- Pressing?

I UNRIPENED/FRESH CHEESE I

- Salting^^^ ^^^^^ - Special secondary cultures

/^"^Rennel ^\ -Coating/packagingf Milk enzymes \ " SMoistur\( Starter enzymes ) I RIPENING I A pJ?ri JV Secondary culture / L ^ ' J^ V NaCl /

X,, ^ Adventitious microflora ^/ ^>^ - Composition \Fat^S^"^^^^ ^^^^^ \ -Temperature ^*™-*<^

"*•*—• \ -Humidity) -Time

- Proteolysis /- Lipolysis S \ r- Glycolysis ^ -^ T- Secondary changes [MA CHRESE [ ^ Flavour

Texture^

TFunctional properties

Figure 14-1 Factors that affect the quality of cheese. The figure is intended to show the multiplicity of factorsthat impact, directly or indirectly, on the quality of rennet-coagulated cheeses. It is possible to standardize andcontrol many of the factors involved. Knowledge of the factors that affect cheese quality should enable a HACCPapproach to be applied to cheese production.

Page 361: Cheese Science

cance of these to cheese quality has not yet beenresearched adequately. Several of these enzymeshave the potential to affect cheese quality, espe-cially lipase, proteinase(s), acid phosphatase,and perhaps xanthine oxidase, sulfydryl oxidase,lactoperoxidase, and y-ghitamyl transpeptidase.Many of these survive high-temperature, short-time (HTST; 720C x 15 s) pasteurization to agreater or lesser extent, and at least some (e.g.,plasmin, acid phosphatase, and xanthine oxi-dase) are active during cheese ripening (seeChapter 11).

Somatic cells are an important source of en-zymes, particularly proteinases, in milk. So-matic cell count (SCC) is negatively correlatedwith cheese yield (see Chapter 9) and quality. ASCC of less than 300,000/ml is recommended.

Although precise information is lacking, it isunlikely that indigenous milk enzymes are a ma-jor cause of variability in cheese quality. Some ofthese enzymes contribute to cheese ripening andmay contribute to the superior quality of raw milkcheese, a possibility that warrants investigation.

14.2.3 Chemical Composition

The chemical composition of milk, especiallythe concentrations of casein, fat, calcium, andpH, has a major influence on several aspects ofcheese manufacture, especially rennet coagula-bility, gel strength, curd syneresis, and hencecheese composition and cheese yield. When sea-sonal milk production is practiced, as in NewZealand, Ireland, and Australia, milk composi-tion varies widely, and there is some variabilityeven with random calving patterns due to nutri-tional factors. It is possible to reduce, but noteliminate, the variability in the principal milkconstituents by standardizing the concentrationsof fat and casein, not just the ratio (protein con-tent can be standardized by adding UF reten-tate), the pH (using gluconic acid-8-lactone),and calcium content (by adding CaCl2), as dis-cussed in Chapter 2.

14.3 COAGULANT (RENNET)

It is generally accepted that calf chymosinproduces the best-quality cheese. An adequate

supply of chymosin from genetically engineeredmicroorganisms is now available (although itsuse is not permitted in all countries), and there-fore rennet quality should not be a cause of vari-ability in cheese quality.

As discussed in Chapter 11, the proportion ofadded rennet retained in cheese curd varies withrennet type, cook temperature, and pH at drain-ing. These variables should be standardized ifcheese of consistent quality is to be produced.Increased retention of the coagulant in the curdresults in greater initial hydrolysis of asi-casein,although this does not appear to be reflected insensory assessment of cheese texture and flavor.It has been suggested that the activity of chy-mosin in cheese curd is the limiting factor incheese ripening. However, excessive rennet ac-tivity leads to bitterness. There have been rela-tively few studies on how chymosin activity af-fects cheese quality, an issue that appears towarrant further research.

14.4 STARTER

Since the starter plays a key role in cheesemanufacture and ripening, it seems that differ-ences between the enzyme profiles of starterstrains should affect cheese quality. Modernsingle-strain starters produce acid very repro-ducibly and, if properly managed, show goodphage resistance. Lactococcus strains have beenselected mainly on the basis of acid-producingability, phage resistance, and compatibility.Based on pilot-scale studies and commercial ex-perience, strains that produce unsatisfactory, es-pecially bitter, cheese have been identified andexcluded from commercial usage. However,systematic studies on strains with positivecheesemaking attributes are lacking. This prob-ably reflects the lack of information on preciselywhat attributes of a starter are desirable from aflavor-generating viewpoint. Studies on geneti-cally engineered strains that superproduce pro-teinase and/or the general aminopeptidase PepNshowed that cheese quality was not improved,although proteolysis was accelerated. Since alllactococcal enzymes, except the cell wall-asso-ciated proteinase, are intracellular, the cells must

Page 362: Cheese Science

lyse before these enzymes can participate in rip-ening. Therefore, the rate of lysis of Lactococcusstrains is being studied with the objective of se-lecting strains with improved cheesemakingproperties.

Sulfur compounds have long been consideredas contributors to the flavor of many cheese vari-eties. Some strains of Lc. lactis spp. cremoris(but not Lc. lactis spp. lactis) can absorb glu-tathione (y-Glu.Cys.Gly; GSH) from the growthmedium. Release of GSH into the cheese uponcell lysis may affect the redox potential (Eh) ofcheese and hence the concentration of thiol com-pounds. Comparative cheesemaking studies us-ing starter strains that accumulate glutathioneand those that do not are warranted.

Although considerable information is avail-able on the individual enzymes of Lactococcusand, to a lesser extent, of Lactobacillus, espe-cially on the glycolytic and proteolytic systems,there have been few studies on the proportionsof different enzyme activities in starter strains.There have been even fewer studies on the rela-tionship between different starter enzyme pro-files and cheese quality. It would appear to behighly desirable that studies should be under-taken to relate cheese quality to the natural en-zyme profile of starter strains or genetically en-gineered starters. The availability of starterstrains deficient in or overproducing one or moreenzymes will facilitate such studies.

It is very likely that the desirable cheese-making properties of starters are due to a balancebetween certain, perhaps secondary, enzymaticactivities that have not yet been identified.

14.5 NONSTARTER LACTIC ACIDBACTERIA (NSLAB)

The significance of lactobacilli for Cheddarand Dutch cheese quality is controversial (seeChapters 11 and 15). Many researchers considertheir contribution to be negative (in the Nether-lands, a maximum of 2 x 106 NSLAB/g is speci-fied for Gouda). Although there are several stud-ies on controlled microflora cheeses, we are not

aware of studies in which cheese free of NSLABwas compared with "control" cheeses contain-ing "wild" NSLAB. Several comparative studieson cheese made under aseptic or nonaseptic con-ditions using Lactococcus starter alone or withselected Lactobacillus adjuncts indicate that in-oculation of cheese milk with selected strains ofLactobacillus improves cheese flavor and possi-bly accelerates ripening. Thermophilic Lactoba-cillus spp. are more effective as adjuncts thanmesophilic lactobacilli, probably because theydie rapidly in cheese, lyse, and release intracel-lular enzymes. Both mesophilic and thermo-philic lactobacilli and Sc. thermophilus are beingused commercially as adjunct cultures for Ched-dar cheese and possibly for other varieties.

Since the numbers and strains of NSLAB incheese are uncontrolled, it is likely that they con-tribute to variability in cheese quality. It is im-possible to eliminate NSLAB completely, evenunder experimental conditions. Therefore, it ap-pears worthwhile to determine what factors af-fect their growth. The number of NSLAB inCheddar is strongly influenced by the rate atwhich the curd is cooled and subsequently rip-ened. Rapid cooling of the curd after molding isthe most effective way of retarding the growth ofNSLAB, although they will grow eventually toabout 107 cfu/g. The growth of NSLAB can beprevented by ripening at about 10C, but all ripen-ing reactions are retarded. The growth ofNSLAB does not appear to be influenced by thecomposition of cheese (moisture, salt, or pH)within the ranges normally found in commercialcheese.

NSLAB grow mainly after the lactose hasbeen metabolized by residual starter activity. Al-though the growth substrates in cheese for Lac-tobacillus are not known, it is likely that they arelimited (NSLAB normally plateau at about 107

cfu/g), and hence it might be possible to out-compete wild NSLAB by adding selected strainsof Lactobacillus to cheese milk, thereby offeringbetter control of the ripening process. NSLABmay also be controlled by including a broadspectrum bacteriocin-producing strain in thestarter culture.

Page 363: Cheese Science

14.6 CHEESE COMPOSITION

The quality of cheese is influenced by itscomposition, especially moisture content, NaClconcentration (preferably expressed as salt-in-moisture [S/M]), pH, moisture in nonfat sub-stances (MNFS; essentially the ratio of proteinto moisture), and percentage fat in dry matter(FDM). At least five studies (Fox, 1975; Gilles& Lawrence, 1973; Lelievre & Gilles, 1982;O'Connor, 1971; Pearce & Gilles, 1979) haveattempted to relate the quality of Cheddar cheeseto its composition. While these authors agreethat moisture content, S/M, and pH are the keydeterminants of cheese quality, they disagreeabout the relative importance of these param-eters (see Figure 14-2).

O'Connor (1971) found that flavor, texture,and total score were not correlated with moisturecontent but were significantly correlated withthe percentage of NaCl and particularly with pH.Salt content and pH were themselves stronglycorrelated with each other, as were salt andmoisture.

Based on the results of a study on experimen-tal and commercial cheeses in New Zealand,Gilles and Lawrence (1973) proposed a grading(selection) scheme that has since been appliedcommercially in New Zealand for young (14-day-old) Cheddar cheese. The standards pre-scribed for premium grade were pH, 4.95-5.10;S/M, 4.0-6.02%; MNFS, 52-56%; and FDM,52-55%. The corresponding values for first-grade cheeses were 4.85-5.20%, 2.5-6%, 50-57%, and 50-56%. Young cheese with a compo-sition outside these ranges was consideredunlikely to develop into good-quality maturecheese. Quite wide ranges of FDM are accept-able. Lawrence and Gilles (1980) suggested thatsince relatively little lipolysis occurs in Cheddarcheese, fat content plays a minor role in deter-mining cheese quality but if FDM falls belowabout 48%, the cheese is noticeably firmer andless attractive in flavor. Pearce and Gilles (1979)found that the grade of young (14-day-old)cheeses produced at the New Zealand Dairy Re-search Institute was most highly correlated with

moisture content. The optimum compositionalranges were MNFS, 52-54%; S/M, 4.2-5.2%;and pH, 4.95-5.15.

Fox (1975) reported weak correlations be-tween grade and moisture and between salt andpH for Irish Cheddar cheeses, but a high percent-age of cheeses with compositional extremeswere downgraded, especially those with low salt(< 1.4%), high moisture (> 38%), or high pH(> pH 5.4). Salt concentration seemed to exer-cise the strongest influence on cheese quality,and the lowest percentage of downgradedcheeses can be expected in the salt range 1.6—1.8% (4.0-4.9% S/M). Apart from the upper ex-tremes, pH and moisture appear to exercise littleinfluence on quality. High salt levels tend tocause curdy textures, probably due to insuffi-cient proteolysis. A pasty body, often accompa-nied by off-flavors, is associated with low saltand high moisture levels. In the same study, thecomposition of extra mature cheeses was foundto vary less, and the mean moisture content was1% lower than that of regular cheeses.

A very extensive study of the relationship be-tween the composition and quality of nearly10,000 cheeses produced at five commercialNew Zealand factories was reported by Lelievreand Gilles (1982). As in previous studies, con-siderable compositional variation was evident,but the variation was less for some factories thanothers. While the precise relationship betweenquality and composition varied between plants,certain generalizations emerged:

• Within the compositional range suggestedby Gilles and Lawrence (1973) for pre-mium quality cheese, composition does nothave a decisive influence on grade, whichdecreases outside this range.

• Composition alone does not provide a basisfor grading as currently acceptable in NewZealand.

• MNFS was again found to be the principalfactor affecting quality.

• Within the recommended compositionalbands, grades declined marginally asMNFS increased from 51% to 55% and

Page 364: Cheese Science

Figure 14-2 Relationships between composition (determined at various stages during ripening) and the quality of mature Cheddar cheese (moisture-in-nonfat substances [MNFS]; fat-in-dry-matter [FDM], and salt-in-moisture [S/M]).

Pearce and Gilles (1979)Composition of cheeses wasdetermined at 14 days and related toquality of Cheddar cheese.

Fox (1975)Relationship between the qualityand composition of 10-week-oldCheddar cheese.

Gilles and Lawrence (1973)Composition of cheeses was determinedat 14 days and related to quality of matureCheddar cheese.

PremiumQuality /PremiumN

Quality/PremiumX

Quality N

pH 4.95-5.15Salt>1.4%A

FDM 50-56%pH 4.85-5.20

S/M 4.2-5.2^MNFS 52-54%^ pH <5.4MNFS 52-56% S/M 4.0-6.0%

S/M 2.5-6.0%MNFS 50-57%

Moisture <38%

pH 4.95-5.10 FDM 52-55%

Page 365: Cheese Science

they increased slightly as S/M decreasedfrom 6 to 4. pH had no consistent effectwithin the range 4.9-5.2, and FDM had noinfluence in the range 50-57%.

The authors stress that, because specific inter-plant relationships exist between grade and com-position, each plant should determine the com-positional parameters that are optimum for itself.

The results of the foregoing investigations in-dicate that high values for moisture and pH and alow salt level lead to flavor and textural defects.The desired ranges suggested by Gilles andLawrence (1973) appear to be reasonable, atleast for New Zealand conditions, but within theprescribed zones composition is not a good pre-dictor of Cheddar cheese quality. Presumably,several other factors—such as microflora, activ-ity of indigenous milk enzymes, relatively smallvariations in cheese composition, and probablyother unknown factors—influence cheese qual-ity but become dominant only under conditionswhere the principal determinants (moisture, salt,and pH) are within appropriate limits.

Although the role of calcium concentrationin cheese quality has received occasional men-tion, its significance was largely overlookeduntil the work of Lawrence and Gilles (1980),who pointed out that the concentration of cal-cium in cheese curd determines the cheese ma-trix and, together with pH, indicates whetherproper procedures were used to manufacture aspecific cheese variety. As the pH decreasesduring cheese manufacture, colloidal calciumphosphate dissolves and is removed in thewhey. The whey removed after cooking com-prises 90-95% of the total whey lost duringcheesemaking, and this whey contains, undernormal conditions, about 85% of the calciumand about 90% of the phosphorus lost from thecheese curd. Thus, the calcium content ofcheese reflects the pH of the curd at wheydrainage. There are strong correlations betweenthe calcium content of cheese and the pH at 1day and between pH at 14 days and the amountof starter used (see Lawrence, Heap, & Gilles,1984). Since the pH of cheese increases during

ripening, the pH of mature cheese may be apoor index of the pH of the young cheese.Therefore, calcium concentration is probably abetter record of the history of a cheese with re-spect to the rate of acidification than the finalpH. Reduction in calcium phosphate concentra-tion by excessively rapid acid development alsoreduces the buffering capacity of cheese, andhence the pH of the cheese will fall to a lowervalue for any particular level of acid develop-ment. Unfortunately, no recent work on thelevel and significance of calcium in Cheddarcheese appears to be available.

14.7 RIPENING TEMPERATURE

The final factor known to influence the rate ofripening and cheese quality is ripening tempera-ture. Ripening at an elevated temperature is nor-mally done with the objective of acceleratingripening, but it also affects cheese quality. Theliterature on the accelerated ripening of cheese isdiscussed in Chapter 15.

14.8 CONCLUSION

Through increased knowledge of the chemis-try, biochemistry, and microbiology of cheese, itis now possible to consistently produce cheeseof an acceptable quality, although this accept-ability is not always achieved, owing to failureto control one or more of the key parameters thataffect cheese composition and ripening. Milk isa variable raw material, and although it is pos-sible to eliminate major variations in the princi-pal milk constituents, some variation persists.Variability in milk composition can also be com-pensated for by manipulating some process pa-rameters in the cheesemaking process. Mostlarge factories operate on a strict time schedule,and hence subtle process manipulation on an in-dividual vat basis may not be possible. There-fore, strict control of milk composition andstarter activity are critical.

From a microbiological viewpoint, the milksupplied to modern cheese factories is of veryhigh quality and, after pasteurization, is essen-

Page 366: Cheese Science

tially free of bacteria. In modern factories, whereenclosed vats and other equipment is used, thelevel of contamination from the environment isvery low. Cheese curd containing fewer than 103

NSLAB/g at day 1 is normal. However, theseadventitious NSLAB grow to about 107 cfu/gand dominate the microflora of long-ripenedcheese. Since the adventitious NSLAB growslowly in cheese, they are most significant inlong-ripened cheese. Although the significanceof the adventitious NSLAB in long-ripenedcheese is unclear, it would appear to be desirableto control them, either by eliminating them orstandardizing them. It is not possible to elimi-nate NSLAB, even in cheese made on a pilotscale under aseptic conditions. Their growth canbe prevented by ripening at about I0C, but theoverall ripening process then becomes unaccept-ably lengthened. Outcompeting indigenousNSLAB by an adjunct Lactobacillus culture,which does not have to contribute to ripening, is

REFERENCES

Andrews, A.T., Olivecrona, T., Vilaro, S., Bengtsson-Olivecrona, G., Fox, P.P., Bjorck, L., & Farkye, N. Y.(1992). Indigenous enzymes in milk. In P.P. Fox (Ed.),Advanced dairy chemistry (V ol. 1). London: Elsevier Ap-plied Science.

Fox, P.F. (1975). Influence of cheese composition on qual-ity. Irish Journal of Agricultural Research, 14, 33—42.

Gilles, J., & Lawrence, R.C. (1973). The assessment ofcheese quality by compositional analysis. New ZealandJournal of Dairy Science and Technology, 8, 148-151.

Lawrence, R.C., & Gilles, J. (1980). The assessment of thepotential quality of young Cheddar cheese. New ZealandJournal of Dairy Science and Technology, 15, 1-12.

a possibility, but this approach has not been in-vestigated.

Although it is now possible to avoid majordefects in cheese produced using modern tech-nology, further research on the biochemistry ofcheese ripening is required to enable the processof cheese manufacture and ripening to be refinedto an extent that will allow the consistent pro-duction of premium quality cheese.

The key to successful cheesemaking is a goodreliable starter, both from the viewpoint of re-producible acid production and subsequent rip-ening. If properly managed, modern starters aregenerally satisfactory, and their performance isbeing improved progressively.

The use of starter adjuncts, usually mesophiliclactobacilli, for some varieties, especially Ched-dar, is increasing, with the objective of intensify-ing and modifying cheese flavor, acceleratingripening, and perhaps controlling adventitiousNSLAB and thus standardizing quality.

Lawrence, R.C., Heap, H.A., & Gilles, J. (1984). A con-trolled approach to cheese technology. Journal of DairyScience, 67, 1632-1645.

Lelievre, J., & Gilles, J. (1982). The relationship between thegrade (product value) and composition of young commer-cial Cheddar cheese. New Zealand Journal of Dairy Sci-ence and Technology, 49, 1098-1101.

O'Connor, C.B. (1971). Composition and quality of somecommercial Cheddar cheese. Irish Agricultural andCreamery Review, 26(10), 5-6.

Pearce, K.N., & Gilles, J. (1979). Composition and grade ofCheddar cheese manufactured over three seasons. NewZealand Journal of Dairy Science and Technology, 14,63-71.

Page 367: Cheese Science

15.1 INTRODUCTION

The original objective of cheese manufacturewas to conserve the principal nutrients in milk(i.e., lipids and proteins). This was achieved by acombination of acidification, dehydration, lowredox potential, and salting. Although a few mi-nor cheese varieties are dehydrated sufficientlyor contain a sufficiently high level of NaCl toprevent microbiological and/or enzymaticchanges during storage, the composition of mostvarieties permits biological and enzymatic activ-ity to occur, which causes numerous changesduring storage. These changes are referred to asripening (maturation) and were described inChapter 11.

Cheese ripening can be a slow process, rang-ing from about 3 weeks for Mozzarella to 2 ormore years for Parmesan and extra mature Ched-dar. In general, the rate of ripening is directly re-lated to the moisture content of the cheese, withlow-moisture cheeses ripening more slowly. Atime-consuming process, cheese ripening is alsoan expensive one, owing to inventory costs, theneed for controlled-atmosphere ripening rooms,and the risk of defects. It is estimated that thecost of ripening cheese is about $100 per tonneper month. Cheese ripening is also a rather un-controlled process, although increased knowl-edge of the microbiology and biochemistry ofripening has increased the probability of produc-ing a good-quality cheese. Consequently, thereare economic and technological incentives for

accelerating the cheese-ripening process, pro-vided that the flavor and textural properties ofthe cheese are not altered.

As discussed in Chapter 11, glycolysis of theresidual lactose in cheese curd is completewithin, at most, a few weeks. The racemizationof L-lactic acid to DL-lactic acid, as occurs inCheddar and Dutch-type cheeses, has no effecton the flavor or texture of cheese, while its ca-tabolism in mold-ripened and Swiss-typecheeses occurs relatively quickly—within a fewweeks. Hence, it is not necessary to acceleratethe metabolism of lactose or lactic acid.

Only limited lipolysis occurs in most cheesevarieties. Major exceptions are Blue and someItalian varieties, such as Romano and Provalone.Blue cheeses ripen relatively rapidly (less than 4months), and the principal lipase is that secretedby P. roqueforti. The characteristic flavor ofBlue cheeses is due to methyl ketones, producedby p-oxidation of free fatty acids by the mold.Lipolysis in the above Italian cheeses is duemainly to an exogenous lipase, pregastric es-terase, added to the cheese milk. Consequently,the rate and extent of lipolysis can be readily al-tered if desired.

Proteolysis occurs in all cheese varieties,ranging from limited, such as Mozzarella, tovery extensive, such as Blue, Parmesan, and ex-tra mature Cheddar. Proteolysis is largely re-sponsible for the textural changes in most variet-ies. It also makes a direct contribution to flavor(e.g., peptides and amino acids), produces sub-

Acceleration of Cheese Ripening

CHAPTER 15

Page 368: Cheese Science

strates (amino acids) for the generation of sapidcompounds (e.g., amines, acids, thiols, andthioesters), and facilitates the release of sapidcompounds from the cheese mass during masti-cation. Proteolysis is perhaps the most importantreaction during cheese ripening, but in Blue andItalian varieties lipolysis and fatty acid oxidationare also very important. Proteolysis appears tobe rate limiting in the maturation of most cheesevarieties and hence has been the focus of mostresearch on the acceleration of ripening, whichis most pertinent for low-moisture, slow-ripen-ing varieties (most published work has been onCheddar). Techniques for the acceleration of rip-ening are also applicable to low-fat cheeses,which ripen more slowly than their full-fat coun-terparts.

Thus, the objective of accelerating cheese rip-ening should be to accelerate the proteolytic pro-cess and related events that occur in naturallyripened cheese as closely as possible. As de-scribed in Chapter 11, proteolysis in Cheddarand related cheeses is now fairly well under-stood at the molecular level, and hence it shouldbe feasible to accelerate the process. Several ap-proaches have been adopted. An extensive lit-erature on the acceleration of cheese ripeninghas accumulated and has been the subject of sev-eral reviews (see Fox et al., 1996; Wilkinson,1993).

The methods used to accelerate ripening fallinto six categories:

1. elevated ripening temperature2. exogenous enzymes3. chemically or physically modified bacte-

rial cells4. genetically modified starters5. adjunct cultures6. cheese slurries and enzyme-modified

cheeses

These methods, each of which has advantagesand limitations (Table 15-1), aim to accelerateripening either by increasing the level of puta-tive key enzymes or by making the conditionsunder which the "indigenous" enzymes incheese operate more favorable for their activity.

15.2 ELEVATED TEMPERATURE

Traditionally, cheese was ripened in caves orcellars, probably at 15-2O0C for much of theyear. Since the introduction of mechanical re-frigeration for cheese-ripening rooms in the1940s, the use of a controlled ripening tempera-ture has become normal practice in modern fac-tories. Typical ripening temperatures are as fol-lows: Emmental, 22-240C (for part of ripening,i.e., the critical "hot room" period); mold andsmear-ripened cheeses, 12-150C; Dutch variet-ies, 12-140C; and Cheddar, 6-80C (the ripeningtemperature for Cheddar is exceptionally low).The ripening temperature for most varieties isprofiled. The above temperatures are the "maxi-mum" in the profiles and are usually maintainedfor 4-6 weeks, usually to induce the growth of adesired secondary microflora, after which thecheese is transferred to a much lower tempera-ture (e.g., 40C). Again, Cheddar is an exception,since it is normally kept at 6-80C throughout theripening process.

About 2O0C is probably the upper limit forcheese ripening. Above this temperature, thecheese is very soft and deforms readily. Exuda-tion of fat and excessive evaporation of moisturemay also occur in cheeses that are not filmwrapped (some brine-salted varieties). Thus, thescope for accelerating the ripening of mostcheese varieties by increasing the ripening tem-perature is quite limited. However, this approachhas potential for Cheddar and offers the simplestand cheapest method for accelerating ripening:no additional costs are involved (indeed, savingsmay accrue from reduced refrigeration costs),and there are no legal barriers. However, consid-ering the numerous complex biochemical reac-tions that occur during ripening, it is unlikelythat all reactions will be accelerated equally atelevated temperatures, and unbalanced flavor oroff-flavors may result. We are not aware of dataon the effect of ripening temperature on the rela-tive rates of individual reactions during ripening.The growth of nonstarter bacteria is acceleratedat elevated temperatures, but at least in the caseof Cheddar, the final number of nonstarter lactic

Page 369: Cheese Science

acid bacteria (NSLAB) is independent of ripen-ing temperature, at least up to 2O0C, providedthat composition and microbiological status aresatisfactory. Furthermore, the species of NSLABpresent may depend on the ripening temperature,but we are not aware of data on this, apart fromthe fact that the growth of heterofermentativelactobacilli is promoted by high temperaturesduring the early phase of ripening.

At least three studies on the ripening of Ched-dar cheese at elevated temperatures (up to2O0C), alone or in combination with other agents(e.g., exogenous proteinases and/or lactose-negative starter supplements), have been pub-lished (see Fox et al., 1996). These studies agreethat ripening at an elevated temperature is thesingle most effective method for acceleratingripening. For example, the duration of ripeningcan be reduced by 50% by increasing the ripen-ing temperature from 60C to 130C, without ad-verse effects. The highest temperature that canbe used continuously is about 160C, although2O0C could be used for a short period; 12-140Cis probably optimal. Ripening can be acceleratedor delayed by raising or reducing the tempera-ture at any stage during the process, thus en-

abling the supply of mature cheese to be regu-lated. Cheese intended for ripening at an el-evated temperature should be of good chemicalcomposition (cheese with a high pH, low NaCl,or high moisture content is unsuitable) and havea low initial NSLAB count (< 103 cfu/g at theend of manufacture has been suggested), whichcan be controlled by cooling the cheese curd rap-idly after pressing.

The quality of Cheddar cheese produced inmodern factories is, in most cases, sufficientlyhigh to withstand ripening at 12-140C withoutproblems. Except in cases where it is desired toretard ripening, for whatever reason, there is nojustification for ripening at a temperature as lowas 20C, which is sometimes used. The ripeningof all hard and semi-hard cheeses could probablybe conducted successfully at 12-140C. A tem-perature of 160C was used successfully to accel-erate the ripening of Manchego cheese.

15.3 EXOGENOUS ENZYMES

Since enzymes are directly responsible formost of the changes that occur during ripening, it

Table 15-1 Principal Methods Used To Accelerate Cheese Ripening

Method Advantages Disadvantages

Elevated temperature No legal barriers; technically Nonspecific action; increased risk ofsimple; no cost (perhaps saving) spoilage

Exogenous enzymes Low cost; specific action, choice of Limited choice of useful enzymes;flavor options possible legal barriers; difficult to

ensure uniform incorporation ofenzymes; risk of overripening

Modified starter cells Easy to incorporate; natural Technically complex; rather expen-enzyme balance retained sive

Genetically engineered Easy to incorporate; choice of Possible legal barriers; may experi-starters options ence consumer resistance

Cheese slurries and Very rapid flavor development High risk of microbial spoilage; finalenzyme-modified product requires processingcheese

Page 370: Cheese Science

seems reasonable to assume that ripening couldbe accelerated by adding exogenous enzymes tothe cheese curd. A number of options have beeninvestigated.

15.3.1 Coagulant

Since the coagulant is principally responsiblefor primary proteolysis in most cheese varieties(see Chapter 11), it might be expected that ripen-ing could be accelerated by increasing the levelor activity of rennet in the cheese curd. This canbe done by increasing the level of rennet addedto the cheese milk (which would necessitate pro-cess modification), renneting at a lower pH (theproportion of gastric and fermentation rennetsretained in the curd increases as the pH of themilk is reduced), or adding powdered rennet tothe cheese curd (e.g., at salting, in the case ofCheddar). Although chymosin appears to be thelimiting proteolytic agent in the production ofsoluble N in cheese, increasing the level of ren-net in cheese curd does not accelerate ripeningand may cause bitterness. Large chymosin-pro-duced polypeptides accumulate in cheese duringripening, but many of the smaller peptides arefurther hydrolyzed by microbial proteinasesand/or peptidases, and these may in fact be rate-limiting with respect to flavor development.Therefore, it might be necessary to increase thestarter population or its proteolytic activity toexploit the benefits that might accrue fromhigher rennet activity. As far as we are aware,this approach has not been investigated.

The natural function of chymosin is to coagu-late milk in the neonatal stomach, thereby in-creasing the efficiency of digestion. It is fortu-itous that chymosin is not only the most efficientmilk coagulant but also gives best results incheese ripening. However, it seems probablethat the efficiency of chymosin in cheese ripen-ing could be improved by protein engineering.As discussed in Chapter 6, some modifiedchymosins have been produced through geneticengineering, but the cheesemaking properties ofthese have not been studied to date.

As discussed in Chapter 11, chymosin hasvery little activity on (3-casein in cheese, prob-

ably because the principal chymosin-susceptiblebond in (3-casein, Leui92-Tyr193, is in the hydro-phobic C-terminal region of the molecule, whichis inaccessible in (3-casein because of its hydro-phobicity. However, C. parasitica proteinasepreferentially hydrolyzes (3-casein in cheesewithout causing flavor defects (perhaps its pre-ferred cleavage sites are in the hydrophilic N-terminal region, but its specificity is unknown).A rennet containing chymosin and C. parasiticaproteinase might be useful for accelerating rip-ening, but, as far as we are aware, this has notbeen investigated.

15.3.2 Plasmin

Plasmin contributes to proteolysis in cheese,especially in high-cooked varieties in whichchymosin is extensively or totally inactivated(see Chapter 11). Plasmin is associated with thecasein micelles in milk, which can bind at least10 times the amount of plasmin normally presentin milk, and is totally and uniformly incorpo-rated into cheese curd, thus overcoming one ofthe major problems encountered with the use ofexogenous enzymes to accelerate cheese ripen-ing.

Addition of exogenous plasmin to cheesemilk accelerates the ripening of Cheddar cheesemade from that milk, without off-flavors. Atpresent, plasmin is too expensive for use incheese on a commercial scale. Perhaps the genefor plasmin can be cloned in and expressed by asuitable bacterial host, which would reduce itscost. It may also be possible to clone the plasmingene in Lactococcus, which could be used as astarter that would lyse and release plasmin dur-ing ripening. A nonstarter mesophilic Lactoba-cillus might also be a suitable host, but thesebacteria lyse to only a limited extent duringcheese ripening, and therefore the plasminwould either have to be excreted or extracellu-larly located.

Since milk normally contains four times asmuch plasminogen as plasmin, an alternativestrategy might be to activate indigenous plasmi-nogen by adding a plasminogen activator (e.g.,urokinase) that also associates with the casein

Page 371: Cheese Science

micelles. Recent work in the authors' laboratoryhas shown that this approach does work, but itmay be too expensive for commercial use.

Preliminary reports have indicated thattrypsin, which is relatively cheap and readilyavailable and has a specificity similar to that ofplasmin, may also have potential for accelerat-ing cheese ripening. This hypothesis requiresconfirmation.

15.3.3 Exogenous Proteinases

The possibility of accelerating ripeningthrough the use of exogenous proteinases otherthan those in rennet has attracted considerableattention for more than 20 years. The principalproblems with this approach are that uniformdistribution of the enzyme(s) in the curd is diffi-cult to achieve and that exogenous enzymes areprohibited in many countries.

Most proteinases can accelerate proteolysis incheese, but most have given unsatisfactory re-sults, probably due to inappropriate specificity.The neutral proteinase "neutrase," from Bacillussubtilis, has been found to be the most effective,and when used alone or in combination with an-other means, such as elevated temperature or alactose-negative culture, it is claimed to acceler-ate ripening substantially. Initially, neutrase wasused alone, but more recently a cocktail of pro-teinase and peptidases and in some cases a lipasehas been used. Commercial preparations includeNaturAge, Accelase, FlavourAge, and DCA50.Some investigators have claimed positive resultswith these preparations, but others have foundno substantial acceleration of flavor develop-ment and in some cases have observed flavorand textural defects. It is not known how widelythese preparations are used commercially.

With the exception of rennet and plasmin(which adsorbs on casein micelles), the incorpo-ration and uniform distribution of exogenousproteinases throughout the cheese matrix posesseveral problems:

• Since most proteinases are water soluble,most of the enzyme added to cheese milk islost in the whey, which increases cost.

• Enzyme-contaminated whey must be heat-treated to inactivate the proteinase if thewhey proteins are recovered for use as func-tional proteins. Therefore, the enzyme mustbe inactivated at a temperature below thatwhich causes denaturation of whey proteins.

• To ensure a sufficient level of enzyme inthe curd, a very high level of enzyme mustbe added to the milk, which causes exten-sive early proteolysis, leading to a loss ofcasein-derived peptides in the whey and areduction in cheese yield.

Consequently, most investigators have addedenzyme preparations, usually diluted with salt tofacilitate mixing, to the curd at salting. Since thediffusion of large molecules, like proteinasesand Upases, in the cheese matrix is very slow oroccurs not at all, this method is applicable onlyto Cheddar-type cheeses, which are salted aschips at the end of manufacture. This method isnot suitable for surface-salted cheeses, which in-clude most varieties. Even with Cheddar-typecheeses, the enzyme will be concentrated at thesurface of chips, and uneven mixing of the salt-enzyme mixture with the curds may cause "hotspots" in which excessive proteolysis, with con-comitant off-flavors, may occur.

Microencapsulation is a fairly widely usedtechnology in the pharmaceutical industry. Thecompound of interest is encapsulated in sometype of membrane, usually with the objective ofprotecting it in a particular environment. Whenthe microcapsules reach the target site, themembrane dissolves or disintegrates, releasingthe entrapped compound. Enzymes have beenencapsulated for a number of applications andhave attracted the attention of cheese technolo-gists as a means of adding exogenous enzymesto cheese. Usually, the enzyme, encapsulated ina lipid or phospholipid membrane, is added tothe cheese milk. The microcapsules are effi-ciently entrapped in the coagulum formed uponrenneting and are retained in the curd. Duringcooking, the lipid membrane melts, releasingthe encapsulated enzymes (Figure 15-1).

Several studies on the microencapsulation ofenzymes for use in cheese have been reported.

Page 372: Cheese Science

Although microcapsules added to milk are in-corporated efficiently into cheese curd, the effi-ciency of enzyme encapsulation is low, thus in-creasing cost. At present, encapsulated enzymesare not being used commercially in cheese pro-duction.

15.3.4 Exogenous Lipases

Lipolysis is a major contributor, directly orindirectly, to flavor development in strong-fla-

vored cheeses, such as Romano, Provolone, andBlue varieties. Rennet paste or crude prepara-tions of pregastric esterase (PGE) are normallyused in the production of Romano, Provolone,and some other Italian cheeses. R. miehei lipasemay also be used for Italian cheeses, although itis less effective than PGE. It has been reportedthat lipases from P. roqueforti or P. candidummay be satisfactory also. The ripening of Bluecheese may be accelerated and quality improvedby the addition of lipases.

Figure 15-1 Schematic representation of the incorporation of microencapsulated exogenous enzymes intocheese curd.

Mature cheese - disrupted liposoraes

Hooping

Cheddaring - intact liposomes

Curd formation

Cheese vat

Liposome-bound proteinase

'Liberated' proteinase

Page 373: Cheese Science

Although Cheddar-type and Dutch-typecheeses undergo little lipolysis during ripening,it has been claimed that the addition of rennetpaste improves the flavor of Cheddar cheese, es-pecially that made from pasteurized milk. Sev-eral authors have claimed that Upases improvethe flavor of "American" or "processed Ameri-can" cheese. In contrast, a number of authorshave reported that the addition of PGE or R.miehei lipase, with or without neutase, to Ched-dar cheese curd has a negative effect on flavorquality.

A strain of A. oryzae secretes a lipase that hasan exceptionally high specificity for C6-C8 acidsand forms micelles in aqueous media, as a resultof which about 94% of the enzyme added to milkis retained in the cheese curd. This lipase, com-mercialized as Flavour Age (Chr. Hansen's, Mil-waukee, Wisconsin), is claimed to accelerate theripening of Cheddar cheese. The formation ofshort-chain fatty acids parallels flavor intensityin Cheddar cheese. In contrast to the free fattyacid profile caused by PGE, which liberates highconcentrations of butanoic acid, the profile incheese treated with FlavourAge is similar to thatin the control cheese, except that the levels offree fatty acids are much higher.

Feta-type cheese produced from cow milkwith a starter containing Lc. lactis and Lb. caseiand a blend of kid and lamb PGEs developed abody, flavor, and texture similar to those of au-thentic Feta cheese. It has been reported that theflavor of Ras and Domiati cheeses can be im-proved and flavor development accelerated bythe addition of low levels of PGE or Upases fromR. miehei or R. pusillus.

15.4 SELECTED, ACTIVATED, ORMODIFIED STARTERS

Since the starter bacteria are mainly respon-sible for the formation of small peptides andamino acids and for flavor development incheese (Chapter 11), it seems obvious to exploitcertain characteristics of the starter to accelerateripening. At least four approaches have beenadopted.

15.4.1 Selected Starters

The primary function of starters is to produceacid at a reliable and adequate rate. As discussedin Chapter 5, highly refined starters, containingonly one or a few selected strains, are nowwidely used, especially by large cheese manu-facturers. The principal criteria for the selectionof single-strain starters are phage-unrelatedness,the ability to grow well and produce acid overthe temperature profile used in cheesemaking,and interstrain compatibility. The selection pro-tocol does not include specific criteria for theability to produce high-quality cheese, butstrains with undesirable cheesemaking proper-ties (e.g., bitterness) have been excluded, andstrains that more or less consistently producehigh-quality cheese have been selected on thebasis of commercial experience. The scientificselection of starter strains with desirablecheesemaking properties is hampered by thelack of precise knowledge as to which enzymesare most important. However, there is strongevidence that Lactococcus spp. differ markedlywith respect to the activity and specificity of cellwall-associated proteinase and the activity ofvarious intracellular exopeptidases, acid phos-phomonoesterase, and esterase. There is alsostrong evidence that the cheesemaking proper-ties of starter strains differ markedly, but thecheesemaking properties and enzyme activitieshave not been correlated. Research in this areaappears to be warranted. Information is also re-quired on the activity of enzymes involved in thecatabolism of amino acids, which is believed tobe important in flavor development.

Since the cell envelope-associated proteinaseis the only extracellularly located enzyme inLactococcus and Lactobacillus and the cellscease to grow in cheese within about 1 day andare therefore unable to transport compounds intothe cell, the cells must lyse so that their intracel-lular enzymes (exopeptidases, esterases, phos-phatase, etc.) can encounter their substrates.Thus, the faster cells lyse, the sooner their intra-cellular enzymes can become involved in cheeseripening. There is natural variation in the sus-

Page 374: Cheese Science

ceptibility of Lactococcus strains to lysis, andhence the selection of fast-lysing strains may bea useful approach to accelerating ripening. Lysisof some Lactococcus strains is thermoinducible(e.g., occurs during cooking). It has been sug-gested that lysis might be induced via a con-trolled phage infection, but this would probablybe unacceptable to cheese manufacturers owingto the difficulty of controlling the level of infec-tion.

Some strains of Lactococcus produce bacte-riocins that cause the lysis of other Lactococcusstrains. By including an appropriate level of abacteriocin-producing strain in the starter, it ap-pears to be possible to induce early lysis of thestarter cells without interfering with acid pro-duction. The results of preliminary experimentssuggest that this approach accelerates ripeningand improves flavor, but larger-scale studies arerequired.

15.4.2 Attenuated Starters

Since the starter plays a key role in cheese rip-ening, it might be expected that increasing cellnumbers would accelerate ripening. At least inthe case of Cheddar, high numbers of startercells have been associated with bitterness. How-ever, not all authors agree that bitterness is re-lated simply to starter cell numbers. Some sug-gest that too much proteolytic activity or thewrong specificity is responsible, for example,too little peptidase activity relative to proteinaseactivity. In fact, a number of authors reportedthat stimulating starter growth accelerates ripen-ing. Perhaps the significance of starter cellnumbers for cheese ripening should be rein-vestigated.

An alternative to the use of high starter cellnumbers is the addition of attenuated starter cellsto the cheese milk. The rationale behind thismethod is that it would likely destroy the acid-producing ability of the starter (since exces-sively rapid acid development is undesirable)but cause as little denaturation of the cells' en-zymes as possible. The discussion in the preced-ing paragraph suggests that adding attenuated

cells might cause bitterness, but this has not beenreported to be a problem with the use of attenu-ated starters; in fact, the opposite has usuallybeen reported. However, most studies on the useof attenuated starters have been on varietiesother than Cheddar.

Five alternative approaches to the productionof attenuated starters have been investigated;these are discussed below.

Lysozyme Treatment

Lysozyme is a widely distributed enzyme thathydrolyzes the cell wall of certain bacteria, in-cluding Lactococcus and Lactobacillus, causingthe cells to lyse unless the osmotic pressure ofthe surrounding medium is high. Lysozyme-treated Lactococcus cells do not lyse in milk orunsalted cheese curd but do lyse when the curd issalted, releasing their intracellular enzymes intothe cheese matrix (if the lysozyme-treated cellslysed in the milk, most of the intracellular en-zymes would be lost in the whey). Limited stud-ies on the addition of lysozyme-treated cells tomilk for Cheddar cheese indicate that, althoughthe addition of the equivalent of 1010 cells/gcheese accelerated proteolysis, flavor develop-ment was not accelerated significantly. Lyso-zyme is rather expensive, and a cheaper source isrequired before this approach would be commer-cially successful.

Heat- or Freeze-Shocked Cells

The lactic acid-producing ability of lactic acidbacteria can be markedly reduced by a sublethalheat treatment (e.g., 60-7O0C for 15 s) withoutreducing proteinase and peptidase activity to anysignificant extent. Heat-shocked cells added as aconcentrate to cheese milk are entrapped in thecurd (~ 90% retention). A number of indepen-dent studies on several types of cheese haveshown that this approach accelerates ripeningand intensifies flavor. However, addition of alarge number of cells (e.g., 109 cfu/g) is neces-sary to achieve a significant impact, and the costmay be prohibitive.

Freezing and thawing also kill lactic acid bac-teria (LAB) without inactivating their enzymes.

Page 375: Cheese Science

Although this technique would appear to beeasier to execute than heat-shocking and appearsto be equally effective, it has been used to a verylimited extent. As far as we know, neither heat-shocked nor freeze-thaw-shocked cells are usedcommercially. Lactose-negative (Lac~) strainsof Lactococcus, which do not grow in milk,would appear to be a more attractive alternativeto heat-shocked or freeze-shocked cells (see"Mutant Starters" below).

Solvent-Treated Cells

Treatment of starter cells with n-butanol acti-vates some membrane-bound proteinases andpeptidases, presumably by increasing accessibil-ity to the substrate. The addition of butanol-treated cells to cheese milk accelerated ripeningslightly and reduced the intensity of bitter flavor.The use of solvent-treated cells in cheesemakingmay not be permitted by regulatory authorities.

Mutant Starters

Lactose-negative (Lac~) cells cannot grow inmilk, probably because they lack the lactosetransport system (see Chapter 5), and hence theycan be added to cheese milk without interferingwith the rate of acid production. The Lac~ cellscontain the full complement of enzymes andthus serve as a package of enzymes that is en-trapped in the cheese curd and lyses slowly, re-leasing enzymes. Since the Lac gene is encodedon a plasmid and plasmids are easily lost, Lac-mutants occur naturally and are easily isolated.The principal characteristics required in a starter(phage resistance, reliable acid production, andstrain compatibility) are not important in Lac-adjuncts, and hence a wider range of strains maybe suitable, such as those with high proteinaseand/or peptidase activity. Strains may also beengineered to have very high levels of certainenzyme activities that are considered to be im-portant in cheese ripening. The ability of Lac~ orLac-Prtr mutants to accelerate ripening and/orintensify flavor has been assessed in severalstudies, dating back to 1983. In general, theyhave given satisfactory results. Lac~ Lactococ-cus strains with high exopeptidase activity are

commercially available and have given satisfac-tory results in pilot-scale and commercial-scalestudies. The cultures are available as frozen con-centrates or freeze-dried preparations. Althoughsuch cultures are relatively expensive (the rec-ommended level of usage costs « $96 per tonneof cheese), the extra cost is affordable if ripeningtime is reduced by more than 20% and/or cheeseflavor is intensified. The extent of their commer-cial use is not known.

Other Bacteria as Additives

LAB are weakly proteolytic in comparisonwith other microorganisms. It might be possibleto accelerate or modify ripening by adding cellsof other genera with a very high level of generalor specific peptidolytic activity. The cells, whichserve as packages of enzymes, when added tothe milk will be efficiently entrapped and uni-formly distributed in the curd. As discussed inChapter 10, cheese is a very selective environ-ment in which few bacterial genera can grow orsurvive.

Preliminary studies have been reported on theuse of Pseudomonas, Brevibacterium, and Pro-pionibacterium to accelerate or modify ripening.The first two of these bacteria are strictly aerobicand therefore will not grow in the interior ofcheese or on its surface when vacuum packed.Significant modification of ripening (i.e., accel-erated release of amino acids and slightly accel-erated flavor development) was found onlywhen high numbers (108-109/g) of washedPseudomonas or Brevibacterium cells werepresent in the cheese. Such high numbers maynot be economical, especially if the effect ob-tained is rather small.

Propionibacterium freudenreichii subsp.shermanii, the characteristic organism in Swiss-type cheese, is anaerobic but is strongly inhib-ited by NaCl and does not grow below about180C. When it was added to milk for Cheddarcheese, it had a very significant effect on the fla-vor of the cheese at numbers above 108 cfu/g.Proteolysis was accelerated slightly, and the re-sultant cheese had a very pronounced Swiss-type flavor. Although different from that of

Page 376: Cheese Science

Cheddar, the flavor generated was very attrac-tive. In fact, a new type of cheese was produced(i.e., the ripening of the cheese was modifiedrather than accelerated). Further work with suchgenera, especially species of Propionibacte-rium, appears warranted.

15.4.3 Genetically Engineered Starters

Considerable knowledge is now available onthe genetics of the cell envelope-associated pro-teinase (CEP) and of many of the intracellularpeptidases of Lactococcus spp. and to a lesserextent ofLactobacillus. Thus, it may be possibleto specifically modify their proteolytic system.

Mutants deficient in CEP or in one or moreintracellular peptidases are available and havebeen used to study the significance of these en-zymes for cell growth. The significance of CEPin cheese ripening has also been assessed usingPrtr or CEP superproducing mutants. Proteolysisin the cheese made using the Prtr mutant wasslower than in the control, and the cheese lackedflavor, but the superproducing mutant did notaccelerate proteolysis or flavor development,suggesting that CEP is not the limiting factor incheese ripening.

The gene for the neutral proteinase (neutrase)of B. subtilis has been cloned in Lc. lactisUC317. Cheddar cheese manufactured with thisengineered culture as the sole starter underwentvery extensive proteolysis, and the texture be-came very soft within 2 weeks at 80C. Since thegenetically modified cells were not food grade,the cheese was not tasted, but its aroma was sat-isfactory. By using a blend of unmodified andneutrase-producing cells as starter, a more con-trolled rate of proteolysis was obtained and rip-ening was accelerated. An 80:20 blend ofunmodified:modified cells gave best results. Theresults appear sufficiently interesting to warrantfurther investigation when a food-grade mutantbecomes available.

Since free amino acids are widely believed tomake a major contribution, directly or indirectly,to flavor development in cheese, a starter with

increased aminopeptidase activity would appearto be attractive. Two studies on a starter geneti-cally engineered to superproduce the generalaminopeptidase PepN have been reported. Al-though the release of total amino acids was ac-celerated, the rate of flavor development and fla-vor intensity were not, suggesting that therelease of total amino acids is not rate limiting.

15.5 ADJUNCT STARTERS

The fourth group of contributors to the ripen-ing of cheese are NSLAB, which may originatein the milk, especially if raw milk is used, or thecheesemaking environment (equipment, air, andpersonnel). Their most likely source is the milk.As discussed in Chapter 10, the interior ofcheese is a hostile environment for bacteria, andconsequently few genera of bacteria can growwithin cheese. The principal NSLAB are meso-philic lactobacilli. Cheddar cheese made fromgood-quality pasteurized milk in modern en-closed equipment with a good active starter con-tains very few NSLAB initially (< 50 cfu/g), butthese multiply to roughly 107 cfu/g within about3 months. The NSLAB population of pasteur-ized milk cheese is dominated by a few species,usually Lb. casei and/or Lb. paracasei. TheNSLAB population in raw milk Cheddar cheeseusually exceeds 108 cfu/g and is more heteroge-neous.

There is a widely held view, substantiated bycomparative studies on cheese made from raw,pasteurized, or microfiltered milk, that cheesemade from raw milk ripens faster and develops amore intense (although not always typical or de-sirable) flavor than cheese made from pasteur-ized milk and that the indigenous microflora isresponsible. The results of these studies havestimulated interest in the selection of Lactoba-cillus cultures for addition to pasteurized milk tosimulate the quality of raw milk cheese. Suchcultures are now available from commercialstarter suppliers. Several studies in which com-mercial or noncommercial Lactobacillus ad-juncts were used have been published (see Fox

Page 377: Cheese Science

et al., 1996; Fox, McSweeney, & Lynch, 1998).In all of these studies, low numbers of selectedmesophilic lactobacilli were added to the cheesemilk. There is general agreement that the lacto-bacilli modify proteolysis; in particular, they re-sult in a higher concentration of free amino acidsand improve the sensoric quality.

In contrast to mesophilic lactobacilli, thermo-philic lactobacilli die rapidly in cheese, lyse, andrelease their intracellular enzymes. Conse-quently, cheeses made with thermophilic Lacto-bacillus spp. as starters contain high concentra-tions of amino acids (the concentrations areparticularly high in Parmesan cheese). Althoughthermophilic lactobacilli will not grow in Ched-dar cheese, their inclusion as a starter adjunctmarkedly intensifies the flavor of Cheddar. Ad-juncts of thermophilic lactobacilli and Sc. ther-mophilus are available commercially.

There appears to be strong evidence that se-lected lactobacilli have the potential to improvecheese flavor and to accelerate flavor develop-ment. It is likely that research on this subject willcontinue and that improved strains ofLactobacil-lus will be isolated. It is also likely that a cocktailof strains will be more effective than individualstrains. It appears that the principal contributionof adjunct NSLAB is to the formation of aminoacids. Perhaps superproducing adjuncts can bedeveloped through genetic engineering once thekey enzymes have been identified.

15.6 SECONDARY CULTURES

Secondary cultures are involved in the ripen-ing of many cheese varieties, including Propi-onibacterium, Brevibacterium, Penicillium, andsome yeasts. As discussed in Chapters 10 and11, these cultures play key and characterizingroles in the ripening of cheeses in which they areused. With the exception of some Swiss variet-ies, cheeses in which secondary cultures areused have relatively short ripening times, due totheir relatively high moisture content and thevery high level of activity of the secondarystarter.

Apart from the production of enzyme-modi-fied cheeses (discussed in Section 15.7), therehas been little work on accelerating the ripeningof cheeses using a secondary starter.

15.7 ENZYME-MODIFIED CHEESE

An extreme form of accelerated ripening ispracticed in the production of enzyme-modifiedcheese (EMC), which has been reviewed byKilcawley, Wilkinson, and Fox (1998). The ba-sic steps involved in the production of EMCs areshown in Figure 15-2. Fresh curd or youngcheese is homogenized (dispersed) and pasteur-ized, and a cocktail of enzymes (proteinases,peptidases, lipases, and perhaps bacterial cul-tures) is added. The mixture is incubated for therequisite period, depending on the activity of theenzymes added, and then repasteurized to termi-nate the microbiological and enzymatic reac-tions. The product may be spray dried or com-mercialized as a paste (see Chapter 19).

Although their flavor does not resemble oreven approximate that of natural cheeses, EMCshave the ability to potentiate cheese flavor invarious food products, including processedcheese, cheese analogues, cheese sauces, cheesedips, and products incorporating cheese, such ascrackers and crisps. For such applications,EMCs may be able to replace 20-50 times theirweight of natural cheese and therefore provide acheaper alternative to natural cheese for impart-ing cheese flavor to formulated foods. Typicalcosts of natural Cheddar cheese and CheddarEMC are $3.75 and $12 per kg, respectively.Cheddar EMCs are the most important commer-cially, but EMCs that simulate several varietieshave been developed, such as Blue cheese,Swiss, and Romano (see Chapter 19).

EMCs are based on "cheese slurries," whichwere developed in the 1970s. Cheese slurrieshave served as model cheese systems in which tostudy the pathways involved in cheese ripening,for the selection of enzymes to accelerate theripening of cheese, and more recently for the se-lection of starter cultures with superior cheese-

Page 378: Cheese Science

Figure 15-2 Typical protocol for the manufacture ofenzyme-modified cheese.

making properties. It is claimed that cheese slur-ries may be added to natural cheese to accelerateripening. Presumably, they act as cultures of de-sirable secondary bacteria. It is also claimed thatcheese slurries develop a characteristic flavor inas little as 4-5 days. Again, most of the work oncheese slurries has been related to Cheddar

cheese, but slurries mimicking other varietieshave also been produced.

15.8 ADDITION OF AMINO ACIDS TOCHEESE CURD

Since amino acids are considered to be impor-tant contributors, either directly or indirectly, tocheese flavor and their production in cheese isrelatively slow, it seems reasonable to assumethat the addition of amino acids to cheese curdmight accelerate flavor development. Glutamicacid, leucine, and methionine are considered tobe the most important amino acids with respectto cheese flavor. Glutamic acid has a brothy fla-vor, while methionine is the precursor of severalsulfur compounds considered to be very impor-tant contributors to the flavor of many cheeses(see Chapter 12).

Preliminary studies by Wallace and Fox(1997) have shown that the addition of interme-diate levels of free amino acids to Cheddarcheese curd at salting (5-6 g/kg curd) had a ben-eficial effect on the development of cheese fla-vor. Amino acids appear to stimulate proteoly-sis, particularly secondary proteolysis involvingthe breakdown of small peptides to amino acids,either due to the activation of peptidases, in-creased cell lysis, or perhaps increased growthof NSLAB, which was not studied. The productsof amino acid catabolism were also not studied,but they may merit study, as they are thought tobe major contributors to cheese flavor. The eco-nomics of incorporating amino acids into cheesecurd also requires evaluation.

15.9 PROSPECTS FOR ACCELERATEDRIPENING

There is undoubtedly an economic incentivefor accelerating the ripening of low-moisture,highly flavored, long-ripened cheeses. Althoughconsumer preferences are tending toward moremild flavored cheeses, there are considerableniche markets for highly flavored products.While the ideal might be to have cheese readyfor consumption within a few days, this is un-

EMCPASTE

REPASTEURIZE(66-720C, 4-8 min)

INCUBATE(4O0C, 48 h)

COOL(e.g., to 4O0C)

Enzyme preparationproteinase(s)peptidase(s)lipase(s)(perhaps bacterial cells)

PASTEURIZE66-720C, 4-6 min

CHEESESLURR?(45% solids)

MIX/homogenise

Disodium phosphateWater

SHREDDED CHEESE(55% solids)

Page 379: Cheese Science

likely to be attained, and in any case it would benecessary to stabilize the product after it reachesoptimum quality, such as by heat treatment (as isused in the production of EMCs).

Although the possibility of using exogenous(nonrennet) proteinases and in some cases pepti-dases attracted considerable attention for a pe-riod, this approach has not been commerciallysuccessful, for which a number of factors may beresponsible:

• Although a necessary prerequisite, primaryproteolysis is probably not the rate-limitingreaction in flavor development.

• The use of exogenous enzymes in cheese isprohibited in several countries.

• Uniform incorporation of enzymes is stillproblematic, and the use of encapsulatedenzymes is not viable at present.

Plasmin may have potential as a cheese-ripeningaid because it can be easily incorporated intocheese curd, is an indigenous enzyme active innatural cheese, and has narrow specificity, pro-ducing nonbitter pep tides. At present, it is tooexpensive, but its cost may be reduced via ge-netic engineering.

Attenuated cells appear to have given usefulresults in pilot-scale experiments, but, consider-ing the mass of cells required, the cost of suchcells would appear to be prohibitive for commer-cial use, except perhaps in special circum-stances. Selected peptidase-rich LacVPrt- Lac-tococcus cells added as adjuncts have givenpromising results, but further work is required,and they may not be cost effective.

The selection of starter strains according toscientific principles holds considerable poten-tial. Such selection is hampered by the lack ofinformation on the key enzymes involved in rip-ening. Preliminary studies on the significance ofearly cell lysis have given promising results, andfurther studies are warranted. Bacteriocin-in-duced lysis appears to be particularly attractive.

The ability to genetically modify startersholds enormous potential, but results to date us-ing genetically engineered starters have beendisappointing. Identifying the key enzymes inripening is essential for the success of this ap-proach. It is hoped that current research oncheese ripening will identify the key sapid com-pounds in cheese and hence the critical rate-lim-iting enzymes. Genetic manipulation of Lac~/Prt adjunct Lactococcus will also be possiblewhen key limiting enzymes have been identi-fied. We believe that adjunct starters, especiallylactobacilli, hold considerable potential. It ap-pears to be possible to produce cheese of ac-ceptable quality without lactobacilli, but thesebacteria do intensify (Cheddar) cheese flavorand offer flavor options. The volume of litera-ture published on starter adjuncts has beenrather limited to date. Further work will almostcertainly lead to the development of superioradjuncts. There is the obvious possibility oftransferring desirable enzymes from lactobacillito starter lactococci.

At present, an elevated ripening temperature(~ 150C) offers the most effective and certainlythe simplest and cheapest method for accelerat-ing the ripening of Cheddar, which is usuallyripened at an unnecessarily low temperature.However, this approach is less applicable tomost other varieties, for which relatively highripening temperatures are used at present.

The key to accelerating ripening ultimatelyrests on identifying the key sapid compounds incheese. This, so far, has been an intractable prob-lem. Work on the subject commenced nearly 100years ago and has been quite intense since 1960,when gas chromatography was developed. Al-though as many as 400 compounds that might beexpected to influence cheese taste and aroma havebeen identified, it is not possible to describecheese flavor precisely (see Chapter 12). Untilsuch information is available, attempts to acceler-ate ripening will be speculative and empirical.

Page 380: Cheese Science

REFERENCES

Fox, P.P., McSweeney, P.L.H., & Lynch, C.M. (1998). Sig-nificance of non-starter lactic acid bacteria in Cheddarcheese. Australian Journal of Dairy Technology, 53, 83—89.

Fox, P.P., Wallace, J.M., Morgan, S., Lynch, C.M., Niland,EJ., & Tobin, J. (1996). Acceleration of cheese ripening.Antonie von Leeuwenhoek, 70, 271-297.

Kilcawley, K.N., Wilkinson, M.G., & Fox, P.F. (1998). En-

zyme-modified cheese [A review]. International DairyJournal. 8, 1-10.

Wallace, J., & Fox, P.F. (1997). Effect of adding free aminoacids to Cheddar cheese curd on proteolysis and flavourdevelopment. International Dairy Journal, 7, 157-167.

Wilkinson, M.G. (1993). Acceleration of cheese ripening. InP.F. Fox (Ed.), Cheese: Chemistry, physics and microbi-ology (2d ed., Vol. 1). London: Chapman & Hall.

Page 381: Cheese Science

16.1 INTRODUCTION

Fresh acid-curd cheeses comprise those vari-eties that are produced by the coagulation ofmilk, cream, or whey via acidification or a com-bination of acid and heat and that are ready forconsumption once the manufacturing operationsare complete (Figure 16-1). They differ fromrennet-curd cheeses, for which coagulation is in-duced by the action of rennet at pH 6.4-6.6, inthat coagulation occurs close to the isoelectricpH of casein (i.e., pH 4.6) or at a higher valuewhen a higher temperature is used (e.g., pH 6.0at 8O0C for Ricotta). While a very small amountof rennet may be used in the production ofQuarg, Cottage cheese, and Fromage frais togive firmer coagula and to minimize caseinlosses during subsequent whey separation, itsaddition is not essential.

Annual world production of fresh acid-curdcheeses amounts to about 3.5 million tonnes,which is equivalent to roughly 23% of totalcheese production (S0rensen, 1997). Quarg, Cot-tage cheese, Cream cheese, Fromage frais, andRicotta are commercially the most importanttypes. Consumption grew by about 2.5% per an-num during the 1987-1996 period. Factors con-tributing to this increase include these:

• They offer a large variety of consistenciesand flavors, made possible by changes incheesemaking protocols; blending of one ormore cheese types to create new products;

and the addition of sugar, fruit purees,spices, and condiments.

• Their soft, ingestible consistency makesthem safe for and attractive to very youngchildren.

• They are perceived as healthy by diet-con-scious consumers. In general, the fat contentof these cheeses is lower than that of rennet-curd cheeses. Double Cream cheese, an ex-ception in the group, has a fat content (-330g/kg) similar to that of Cheddar. However,the cheeses are relatively low in calcium(typically < 0.8 g/kg) compared with rennet-curd cheeses such as Cheddar («7.5 g/kg) orSwiss (« 9.5 g/kg) (Table 16-1).

16.2 OVERVIEW OF THEMANUFACTURING PROCESS FORFRESH ACID-CURD CHEESEPRODUCTS

Production generally involves pretreatment ofmilk (standardization, pasteurization, and per-haps homogenization), slow quiescent acidifica-tion, gel formation, dehydration of the gel (wheyseparation), and in some cases further treatmentsof the curd (pasteurization; shearing; addition ofsalt, condiments, and stabilizers; and homogeni-zation) (Figure 16-2). Acidification is generallyslow, 12-16 hr at 21-230C (long set) or 4-6 hr at3O0C (short set), and is usually brought about bythe in situ conversion of lactose to lactic acid, byan added starter culture and/or by the addition of

Fresh Acid-Curd Cheese Varieties

CHAPTER 16

Page 382: Cheese Science

food-grade acid (e.g., lactic or citric) or acid-ogen, such as gluconic acid-8-lactone (whichhydrolyzes to gluconic acid). The structure ofthe gel has a major effect on the texture (e.g.,spreadability and firmness) and sensory attri-butes (smoothness) of the final product and itsphysicochemical stability (i.e., stability towheying-off and/or to the development of achalky/grainy mouthfeel) during storage. Thestructure of the gel is influenced by many pro-cessing factors, such as the protein level, themilk pasteurization treatment, homogenization,the temperature during acidification, and the pHat which the gel is broken and subjected to dehy-dration. The effect of gel structure on productquality is most pronounced in products whosecurd, following whey separation and concentra-tion, is not treated further (e.g., Quarg andFromage frais). In hot-pack products, such asCream cheeses and some fresh cheese prepara-

tions, curd treatments (pasteurization, homog-enization, and hydrocolloid addition) have a ma-jor impact on the quality of the final product(Guinee, Pudja, & Farkye, 1993).

16.3 PRINCIPLES OF ACID MILK GELFORMATION

Slow acidification of milk under quiescentconditions is accompanied by two opposing setsof physicochemical changes:

1. a tendency toward disaggregation of thecasein micelles into a more disorderedsystem as a result of• solubilization of the internal micellar

cementing agent, colloidal calciumphosphate (CCP), which, at 20-3O0C,is fully soluble at about pH 5.2-5.3(Figure 16-3)

Whey-based

RicottoneAnari

Queso biancoRicottaMascarpone

Acid-heat coagulated

Milk/cream-based

• Quarg-type—Skim milk Quarg—Full-fat Quarg—Tvorog

FromageLabnehLabanehFresh cheese preparationsCream cheese-type—Double/single Cream cheese—Petit Suisse—NeufchatelCottage cheese type—Low/high-fat Cottage cheese—Baker's cheese

Acid coagulated

Figure 16-1 Fresh acid-curd cheese varieties.

Page 383: Cheese Science

• a pH- and temperature-dependent dis-sociation of individual caseins, espe-cially p-casein, from the micelles, witha concomitant increase in the level ofserum casein (casein dissociation de-creases with decreasing pH to about pH6.2, then increases to a maximum at pH5.3-5.6, depending on the temperature,and thereafter decreases to a minimumat the isoelectric pH [Figure 16 ])

• an increase in casein solvation with pHreduction in the range 6.7-5.3 (Figure16-5)

2. a tendency for the casein micelles to ag-gregate into a more ordered system due to• the reduction of the negative surface

charge on the casein micelles andhence of intermicellar repulsive forcesdue to the production of lactic acid

• a decrease in casein hydration in the pHrange 5.3-A6 (Figure 16-5)

• the increase in the ionic strength of themilk serum (due to the increased con-centrations of calcium and phosphateions), which has a shrinking effect onthe casein micelles

At pH values greater than that at the onset ofgelation (« 5.1-5.3 at 20-3O0C), disaggregatingforces predominate and hence a gel is notformed. At lower pH values, forces that pro-mote aggregation of the casein micelles prevailand gel formation begins. Electron microscopicexamination has revealed the presence of a het-erogeneous size distribution of casein aggre-gates (composed of fused casein micelles) atthe onset of gelation. Further reduction of pH isparalleled by a touching of aggregates that ini-tiates the formation of loose porous strands.Eventually, as the pH of the milk approachesthe isoelectric point, dangling strands touch andcross-link to form a three-dimensional particu-

Table 16-1 Approximate Composition of Various Fresh Cheeses

DAy Matter Fat Protein Lactose (Lactate) Salt CaVariety (%, w/w) (%, w/w) (%, w/w) (%, w/w) (%, w/w) (mg/100 g) pH

Cream cheeseDouble 40 30 8-10 2-3 0.75 80 4.6Single 30 14 20 3.5 0.75 100 4.6

Neufchatel 35 20 10-12 2-3 0.75 75 4.6Labneh 25 11.6 8.4 4.4 - - 4.2Quarg

Skim milk 18 0.5 13 3-4 - 120 4.5Full fat 27 12 10 2-3 - 100 4.6

Cottage cheeseLow-fat 21 2 14 - 90 4.8Creamed 21 5 13 - 60 4.8

Fromage fraisSkim milk 1 4 1 8 3 . 5 - 0.15 4 . 4

Queso bianco 49 15 23 1.8 3.9 - 5.4Ricotta

Whole milk 28 13 11.5 3.0 - 200 5.8Part skim 25 8 12 3.6 - 280 5.8

Ricottone 18 0.5 11 5.2 - 400 5.3Brunost

Gubrandsdalsost 82 30 11 38 - 400Flotemyost 80 19 11 46 - - -

Page 384: Cheese Science

late casein network (or gel), which extends con-tinuously throughout the serum phase. The gelis described as particulate because, whenviewed by scanning electron microscopy, theindividual gel strands are found to be composedof particles (casein aggregates) that undergo

limited touching (over part of their surfaces)and are linked together, rather like the beads ina necklace.

Gel formation is accompanied by a markedincrease in the elastic shear modulus (index ofcurd firmness), which increases progressively

Figure 16-2 Generalized production protocol for fresh cheese products.

Heat, Blend, Homogenize

Hot Blend Hot Pack Fresh CheesePreparations

Other Fresh Cheeses,Cream and/orYoghurt and/orCondiments

Hot Treated Curd Hot-Pack Products:Cream cheeseNeufchatel

Pasteurization,Hydrocolloid andCondiment Addition,Homogenization

Whey/Permeate Curd Cold-Pack Products:QuargFromage fraisCottage cheese

Separation (Dehydration)

Gelled, Acidified MilkpH4.6

Rennet (0.5-1 ml/1001)

Incubation (Quiescent)

Starter (~ 1%)

Cooling - 22-3O0C

Pre-treatment- Pasteurization- Homogenization- Partial Acidification

Standardized milk

Page 385: Cheese Science

Milk pH

Figure 16-3 Micellar calcium (•) and inorganic phosphate (O) in skim milk as a function of pH at 3O0C (deter-mined by ultracentrifugation at 88,000 g for 1.5 hr and expressed as percentage of total concentration in milk.

% o

f tot

alF

ract

ion

of c

asei

n di

ssoc

iate

d

Figure 16-4 Serum (nonsedimentable) casein in skim milk as a function of pH at 40C (•), 2O0C (O), and 3O0C(A), determined by centrifugation at 70,000 g for 4,2, and 1.75 hr, respectively, and expressed as a percentage oftotal casein.

Milk pH

Page 386: Cheese Science

with further pH reduction (to 4.6) and casein ag-gregation (Figure 16-6).

The physicochemical and micro structuralchanges accompanying the conversion of milkto an acid gel are summarized in Figure 16-7.

16.4 PREREQUISITES FOR GELFORMATION

Acidification of milk may result in the forma-tion of a gel or a precipitate, depending on therate and extent of casein aggregation. Gelationoccurs when forces that promote aggregation ofthe casein micelles slowly overcome those thatpromote repulsion of the micelles. These condi-tions result in the formation of relatively loose,porous, hydrated aggregates of casein micelles,which are only slightly more dense than the se-rum phase in which they are dispersed. Owing tothe relatively small density gradient between theaggregates and the serum phase, the aggregateshave sufficient time to link together, via strandformation, to form a continuous casein network,which physically entraps the serum (whey)phase. In contrast, the casein micelles aggregatemore rapidly and undergo a high degree of fu-sion (i.e., touch neighboring micelles over a

much larger part of their surface) to formsmaller, less porous, and less hydrated aggre-gates when conditions that promote aggregationof the casein micelles are more extreme, e.g.,rapid acidification under nonquiescent condi-tions at a high temperature, as in the manufac-ture of acid casein. Owing to their relatively highdensity, these aggregates sediment as a precipi-tate. Compared to a gel, the casein in a precipi-tate is highly aggregated, has a very low water-holding capacity, and occupies a much lowerspecific volume (i.e., has a low voluminosity).Although the production of both acid casein andfresh acid-curd cheeses, such as Quarg, involvesthe acidification of skim milk, the conditions ofacidification differ markedly (Figure 16-8) andresult in the production of two very differenttypes of product—a precipitate from which themoisture is expelled rapidly, enabling the recov-ery of casein as a food ingredient (as in acidcasein), and a cheese (gel), which has superiorwater-holding capacity.

To obtain a gel rather than a precipitate, thenumber of attractive forces and hence the sur-face area of contact between the dispersed par-ticles (casein micelles) must be limited. Condi-tions conducive to limited aggregation include a

Figure 16-5 Solvation of casein micelles as a function of pH at 2O0C for skim milk (•) and rennet-treated skimmilk (O).

pH

Solv

atio

n of

cas

ein

mic

elle

s,g

H2O

/g d

ry c

asei

n p

elle

t

Page 387: Cheese Science

slow rate of acidification under quiescent condi-tions. Slow acidification is promoted by the useof a starter culture and a relatively low tempera-ture (e.g., 22-3O0C) during acidification. Duringthe production of fresh acid-curd cheeses, condi-tions that promote a greater degree of casein at-traction and fusion (e.g., when the rate of acidifi-cation is increased) lead to the formation of a gelthat is less voluminous and closer to a precipi-tate. The latter type of gel, which has a lowerwater-holding capacity, is said to be coarser. Al-ternatively, conditions that promote a lower de-gree of casein aggregation (e.g., a slow rate ofacidification) result in a finer gel network, whichhas a relatively high water-holding capacity. Thestructures of fine and coarse gels and of a pre-cipitate in which the concentration of gel-form-ing protein is equal are illustrated schematicallyin Figure 16-9.

In extreme situations (e.g., acidification oc-curs very slowly and it takes more than 16 hr forthe pH to fall from ~ 6.6 to 4.6), if the number ofinterparticle attraction sites is lower than opti-mum, slowly forming aggregates may have suf-ficient time to precipitate before fusing and link-ing with neighboring aggregates to form into anetwork. An example of the latter is the defect inCottage cheese production known as "majorsludge formation," whereby phage infection ofthe starter, after acid development has pro-gressed to an advanced stage (~ pH 5.2-5.3),leads to casein precipitation rather than gelation.

16.5 EFFECT OF GEL STRUCTURE ONQUALITY

Gel structure is a major determinant of qualityattributes, such as the mouthfeel (smoothness or

Time from starter addition, h

Figure 16-6 Development of elastic shear modulus (G *) during fermentation of skim milk pasteurized at 720C x15 s (O) and at 9O0C x 300 s (•); see Table 16-2.

Elas

tic s

hear

mod

ulus

, G

', Pa

Page 388: Cheese Science

Milkcolloidal dispersionof casein micelles

Acidificationreduction in:

- micelle charge- micelle hydration- inter micellar

repulsionlimited aggregationof casein micellesacid gel formation

Acid gelNetwork of uniformlydistributedcross-linked strandscomposed of aggregatedcasein micelles

Figure 16-7 Schematic representation of the conversion of milk to a gel by slow quiescent acidification using astarter culture.

SKIM MILK

Skim milk (220C)

Starter culture

Slow conversion oflactose to lactic acid

GelationatpH4.6

Quarg

Skim milk (< 1O0C)

Addition of HCl

pH4.6

Heat rapidly to 550C

Precipitate

Acid Casein

Figure 16-8 Acidification conditions for the production of Quarg and acid casein.

Page 389: Cheese Science

chalkiness), appearance (coarseness or smooth-ness), and physicochemical stability (absence ofwheying-off and graininess during storage) offresh acid-curd cheese products, especially cold-pack varieties (e.g., Quarg and Fromage frais),where the curd, following whey separation, isnot further treated. Cold-pack varieties are dis-cussed below in more detail.

16.5.1 Syneresis

Syneresis, or whey expulsion, of acid milkgels is necessary for dehydration of the gel dur-ing cheese manufacture and may be achieved bysubjecting the gel, following incubation, to con-centration (i.e., whey removal) by cutting,stirring, cooking, whey drainage, and/or mech-anical centrifugation. On the other hand,wheying-off in the final fresh cheese productduring storage is undesirable, as it leads to theformation of a whey layer, which consumersview as undesirable. However, syneresis fre-quently occurs in acid gel-based products (e.g.,yogurt and fresh cheese) because of the rela-tively high moisture :protein ratio compared withrennet-curd cheeses (e.g., ~ 17.6 g H2O/g proteinin yogurt compared with ~ 1.44g H2O/g proteinin Cheddar cheese). Moreover, owing to the factthe casein is at its isoelectric pH, the water in

acid-curd cheeses is mainly physically imbibedrather than chemically bound by the protein. Incontrast, about 15% of the moisture in young(i.e., < 1 week) rennet-curd cheeses is chemi-cally bound by the paracasein, and the level ap-pears to increase during ripening.

Syneresis of acid (and rennet) milk gels re-quires rearrangement and shrinkage of thecasein matrix (gel) into a more compact struc-ture. However, the gel is generally unable tocontract to any appreciable degree when left un-der quiescent conditions, owing to its rigidity.Thus, the initiation of extreme syneresis (e.g., asrequired to separate the whey and recover thecurd during manufacture) necessitates the appli-cation of a stress to the gel to break the gelstrands. Breaking of the gel strands and hencethe gel as a whole enhances syneresis by

• permitting a large portion of the entrappedwhey to escape, via the surfaces of thenewly created curd particles, from the ma-trix

• allowing the broken strands to come intocloser proximity, enabling them to reknitinto a more compact arrangement

• facilitating the physical expulsion of whey,due to the rearrangement and shrinkage ofthe protein phase, which has the effect of"squeezing out" the entrapped whey

Figure 16-9 Schematic representation of a fine-structured (A) and coarse-structured (B) acid milk gel and aprecipitate (C). The progression from a fine gel through a coarse gel to a precipitate is paralleled by an increasingdegree of fusion of the milk protein (dark areas). Simultaneously, the protein network in the acidified milk be-comes more open and porous.

(A) (B) (C)

Page 390: Cheese Science

Moreover, the prevailing conditions of low pHand relatively high temperature (usually > 220Cduring separation) are conducive to casein dehy-dration and aggregation. The stress required toinitiate syneresis during the manufacture offresh acid-curd cheeses is applied in the form ofexternal pressure via cutting and/or stirring thegel. The broken gel is then subjected to centrifu-gal force (e.g., in a nozzle separator) or gravita-tional force (e.g., by pouring the broken gel ontomuslin bags suspended on a frame).

Acid milk gels formed in situ in the package,such as set natural yogurt, show little tendencyto synerese if left undisturbed. However, even inthis situation, spontaneous syneresis may even-tually occur to a greater or lesser degree, de-pending on the level of fortification and process-ing conditions, such as preheating of milk,which causes differences in the porosity andstructure of the gel. The syneresis may be partlydue to slow proteolysis of the casein caused byenzymes of starter bacteria, a decrease in pH,and temperature fluctuations. Hydrolysis ofcasein influences its hydration and hence its ten-dency to aggregate. The change in the degree ofcasein aggregation results in internal stresseswithin the gel and rearrangement of the caseinmatrix, which in turn leads to syneresis. It hasbeen suggested that casein hydrolysis may be re-sponsible for the widely different practical expe-rience (day-to-day in-factory and interfactoryinconsistencies) regarding syneresis in set fer-mented milk products (Walstra, van Dijk, &Geurts, 1985). Changes in pH and temperatureduring storage, which would alter the state ofcasein aggregation, probably also contribute tospontaneous syneresis (shrinkage) or wheying-off.

Shrinkage of the casein particles of the net-work caused, for example, by a reduction in pHand/or an increase in temperature following gelformation enhances both induced and spontane-ous syneresis.

Once syneresis has started, the outward flowof whey through the gel matrix becomes increas-ingly impeded over time by the sieve effect ex-

erted on it by the relatively narrow pores of thegel, the porosity of which depends on the gelstructure. The impedance of the pores to the out-ward migration of whey (syneresis) increaseswith the duration of syneresis, owing to the pro-gressive contraction of the matrix as it loseswhey and the reduction in the size of its pores.Hence, once a gel is broken, the rate of outwardmigration of whey decreases with time.

The structure of a gel has a marked effect onits ability to synerese or undergo wheying-off.For unidimensional flow through a porous me-dium, such as an acid milk gel, the rate of syner-esis, v, may be expressed by Darcy's law:

V = BbPIhI

where v is the whey flux (i.e., the volume flowrate in the direction of flow divided by the cross-sectional area perpendicular to this directionthrough which the whey flows, measured inmeters per second); B is the permeability coeffi-cient of the gel matrix, which corresponds to theaverage cross-sectional area of the gel pores; h isthe viscosity of the whey flowing through thematrix; AP is the pressure gradient arising fromsyneretic pressure exerted on the entrapped se-rum by the matrix; and / is the distance overwhich the serum flows.

The permeability coefficient B depends on thevolume fraction of the protein matrix and thespatial distribution of the matrix strands (i.e., gelfineness or coarseness). For a given synereticpressure, the resistance to the passage of wheythrough the gel decreases as the permeability co-efficient increases. Hence, a fine gel structurehas a relatively low porosity and a lower perme-ability to outflowing whey than its coarse-struc-tured counterpart. Therefore, while it is moredifficult to remove whey from fine-structuredgels during manufacture, they are much lessprone to wheying-off during storage.

The influence of gel structure on its suscepti-bility to wheying-off may be easily explained byreference to Figure 16-9, which depicts thestructural differences between fine and coarse

Page 391: Cheese Science

gels in which the concentration of gel-formingprotein is equal. In the fine gel, A, the micelleshave formed into thin strands (chains), resultingin a highly branched, continuous gel network.Conversely, in the coarse gel, B, the micelleshave fused to a much greater degree to producethicker gel strands and a less continuous, moreporous structure, which is more susceptible tosyneresis during storage (whey drains easilythrough the large open channels between the net-work strands). The gel-forming protein in A ismore uniformly distributed, and the gel hassmaller interstitial spaces or pores. The rela-tively low porosity of gel A retards the outflowof whey and so endows the gel with better water-holding capacity than B and a low tendency to-ward syneresis.

16.5.2 Rheology

The structure of a gel also has a major influ-ence on its rheological properties. For gels ofsimilar composition, gel strength is primarilydependent on the homogeneity of the gel, whichdetermines the number of stress-bearing strandsper unit area of the gel. In the case of a gel towhich a relatively small stress (i.e., much lessthan yield stress) is applied in direction jc, theelastic shear modulus (G', i.e., ratio of shearstress to shear strain, a/y), which is an index ofelasticity or strength of the gel, can be related tothe number of strands per unit area according tothis equation (Walstra & van Vliet, 1986):

G'= CNxd2F/dx2

where N is the number of stress-bearing strandsper unit area of the gel in a cross section perpen-dicular to Jt; C is a coefficient related to the char-acteristic length determining the geometry of thenetwork; and dF is the change in Gibb's free en-ergy when the aggregates in the strands aremoved apart by a distance dx upon the applica-tion of the stress. The number of strands per unitarea of a gel is determined by its fineness orcoarseness, with a fine gel network having agreater number of stress-bearing strands than a

coarse gel. The thickness and hence the strengthof the stress-bearing strands is, on average,greater in the coarser gel because of the greaternumber of attractions between the aggregates.However, within the normal parameters of freshacid-curd cheese manufacture, a fine gel gener-ally has a greater gel firmness than a coarser gelwith a similar composition and concentration ofgel-forming protein. Compared to a gel, a pre-cipitate (and its accompanying expressed whey)with the same level of gel-forming protein has amuch lower G' value, as the rheological contri-bution ensues mainly from the continuous wheyphase.

16.5.3 Sensory Attributes

The structure of the gel may also influence thesensory characteristics of fresh fermented prod-ucts, especially in cold-pack products where,following its formation, the gel is subjected tolittle further processing (e.g., Quarg andFromage frais) or none (e.g., set yogurt, wherethe gel is formed in its package). A smoothmouthfeel is generally an indicator of good qual-ity in fresh cheese products. Cottage cheese is anexception, in that granularity, as imparted by the"chewy" curd particles, is an indicator of qual-ity. Common sensory defects in fresh cheeseproducts include "chalkiness" (perceived as adry or powdery mouthfeel), grittiness, andgraininess. Electron microscopic analyses ofcheeses with these defects have revealed thepresence of large protein conglomerates (massesof highly fused casein aggregates), suggestingthat the defects ensue from excessive protein ag-gregation during gel formation and/or duringwhey separation. These defects are more likelyto occur in products made from coarse-struc-tured gels than fine-structured-gels, owing to thehigher level of casein aggregation in the former.The defects are more prevalent in productswhere the gel, following fermentation, is con-centrated and/or heated (e.g., Cream cheese);these conditions are conducive to protein dehy-dration and hence aggregation.

Page 392: Cheese Science

16.6 FACTORS THAT INFLUENCE THESTRUCTURE OF ACID GELS ANDTHE QUALITY OF FRESH CHEESEPRODUCTS

Many factors influence the structure of acidmilk gels and hence impact the rheology, sus-ceptibility to wheying-off, and the mouthfeelcharacteristics of fresh acid-curd cheeses. Theprincipal compositional and processing factorsthat influence these are discussed below.

16.6.1 Level of Gel-Forming Protein

Higher concentrations of gel-forming proteingenerally result in denser gel matrices, whichhave a higher number of strands per unit volumeand are more highly branched and less porous.The resultant gels are generally less prone to sy-neresis and are firmer and more elastic (i.e.,higher G% The lower susceptibility to syneresisis particularly desirable in products where thegel is essentially the final product, such as setand stirred curd yogurts. Wheying-off is a com-mon defect in these products, especially thosemade without the addition of stabilizers, such ashydrocolloids. A high gel firmness is also desir-able in yogurt, as it conveys the impression ofbeing creamier, more viscous, and richer to theconsumer. Owing to the fact that fresh cheesesare usually manufactured with a standard levelof protein in the final product, it is envisagedthat increasing the level of milk protein has aless dramatic impact on their rheological proper-ties (e.g., fracture stress and firmness) than it hason those of yogurt, although the authors are notaware of any studies on this aspect of fresh acid-curd cheese. However, a higher gel firmness(due to a higher level of milk protein) is desir-able in the manufacture of cold-pack freshcheeses, as it reduces the susceptibility of the gelto shattering during subsequent whey separa-tion, minimizes the loss of curd fines in thewhey, and contributes positively to cheese yield.Thus, it is widespread practice in the commer-cial manufacture of yogurt and cold-pack fresh

cheese products to increase the level of milk pro-tein prior to acidification by, for example, ultra-filtration of the milk or the addition of dairy in-gredients, such as skim milk powder, wheyprotein concentrate, or blends of dairy ingredi-ents.

The structure of the gel is also markedly influ-enced by the ratio of casein to whey protein inthe milk. Reducing the ratio from 4.6:1 to 3.2:1results in set yogurt that has a finer, more highlybranched, and less porous matrix; a smootherconsistency; and a lower susceptibility towheying-off during storage. Moreover, for agiven concentration of protein and a given wheyproteinxasein ratio in the milk, the syncretic andrheological characteristics of the resultant gelmay differ markedly, depending on the type ofingredient used to increase the level of milk pro-tein. Stirred curd yogurt made from milk stan-dardized to 5% protein with whey protein con-centrate (75% protein, WPC 75) is markedlymore viscous and less prone to syneresis thanthat made from milk standardized to 5% proteinwith WPC 35 (Guinee, Mullins, Reville, & Cot-ter, 1995). Differences in the performance ofprotein ingredients with the same type of pro-tein, whether casein or whey protein, may be re-lated to differences in the degree of whey proteindenaturation, the type and level of minerals, andthe level of other materials, such as lactose.

The effective concentration of milk proteinmay also be increased, while maintaining the ac-tual protein level constant, by

• homogenization of the milk (as practicedin the production of yogurt and Creamcheese), which converts fat globules topseudoprotein particles

• high heat treatment of the milk (e.g., 950C x5 min), which causes denaturation andbinding of whey proteins to casein micelles(the denatured whey proteins become partof the ensuing gel, but undenatured wheyproteins are soluble at their isoelectric pH,around 4.6, and do not participate in gel for-mation)

Page 393: Cheese Science

16.6.2 Heat Treatment of Milk

High heat treatment of milk (e.g., 9O0C x 5 mincompared to 720C x 15 s) prior to culturing offermented products, such as yogurt and Quarg,gives a smoother and firmer consistency (Figure16-6). This effect is due to extensive denaturationof the whey proteins (e.g., > 70% of total) andtheir binding (especially of p-lactoglobulin), viadisulfide interaction, to K-casein; the denaturedwhey proteins subsequently become part of thegel. This interaction results in a higher effectiveconcentration of gel-forming protein and a finer-structured gel with a lower permeability coeffi-cient (and hence lower porosity) and a reducedpropensity to spontaneous wheying-off (Table16-2 and Figures 16-10 and 16-11). Electronmicroscopic analysis of high heat-treated milkshows that these complexes result in the forma-tion of filamentous appendages that protrudefrom the surface of the micelles and prevent theclose approach, and hence extensive fusion, ofmicelles upon subsequent acidification.

For similar levels of whey protein denatur-ation, the type of heat treatment has a significantinfluence on the textural parameters of fer-mented milks. Thus, the viscosity of natural yo-gurt produced from ultrahigh temperature-treated milk (130-15O0C x 2-15 s) is lower thanthat of yogurt produced from high temperature-short time-treated milk (« 80-9O0C x 0.5-5min), which in turn is lower than that of yogurtproduced from batch-heated milk (63-8O0C x10-40 min). The differences in viscosity andfirmness at similar levels of whey protein dena-turation may be associated with different typesof denaturation (e.g., level of unfolding) and/orbinding of denatured whey proteins to thecasein, which alters the structure (i.e., coarse-ness or fineness) of the ensuing gel.

16.6.3 Incubation Temperature and Rate ofAcidification

Increasing the acidification temperature ofmilk, in the range 20-430C, results in

• the onset of gelation at a higher pH value(e.g., pH 5.5 at 430C compared with pH 5.1at 3O0C)

• a coarser gel structure that is firmer (moreelastic) and more prone to wheying-off dur-ing storage

These effects are thought to be associatedwith the faster rate of acidification (when usingthermophilic cultures) and the reduced degree ofcasein dissociation from the micelles at thehigher incubation temperature (see Section16.3).

In an extreme situation, rapid acidification topH 4.6 promotes rapid aggregation of casein andthe formation of large dense aggregates that pre-cipitate rather than form a gel. Gel formation byrapid acidification is, however, possible whenthe tendency of micelles to coagulate is reducedby acidifying to about pH 4.6 at a low tempera-ture (0-40C) and then heating the milk up slowly(~ 0.5°C/min) under quiescent conditions toabout 3O0C.

16.6.4 pH of the Gel

The firmness (elastic shear modulus) of acidmilk gels increases with decreasing pH towardthe isoelectric point and is maximal at aroundpH 4.5. This effect is due to a greater degree ofcasein aggregation and a concomitant reductionin negative charge. Lowering the pH of acidmilk gels at cutting (e.g., from 4.92 to 4.59 inCottage cheese gels) reduces the level of syner-esis. The latter effect may be attributed to thefact that the gel strands are more rigid at thelower pH and hence are less susceptible tobreakage upon cutting. A lower degree of strandbreakage affords less potential for new bondingsites, and matrix contraction is thus less severethan otherwise. Consequently, the pressure ex-erted on the entrapped serum by the matrix isrelatively low and less wheying-off ensues. As acorollary, a decrease in pH during syneresis ofan acid milk gel results in a higher level of syner-esis than if the gel is brought to the same pH be-fore cutting (Walstra et al., 1985).

Page 394: Cheese Science

16.6.5 Rennet Addition

It is common practice during the manufactureof some fresh cheese products, such as Quargand Cottage cheese, to add a small quantity ofrennet to the milk shortly after culture addition(e.g., ~ 1-2 hr), when the pH is roughly 6.1-6.3.Typical levels of addition are 30-60 RU (or 0.5-1.0 ml single strength rennet) per 100 L. Therennet hydrolyzes some K-casein, and there areconcomitant decreases in

• the negative charge on the micelles (i.e.,^-potential)

• casein dissociation from the micelles• casein hydration over the pH region 6.6-4.6

(Figure 16-5)

These changes contribute to enhanced aggrega-tion of the micelles, and gelation begins at ahigher pH value than otherwise. Hence, a gel suf-ficiently firm for cutting and whey separation isobtained at a higher pH value (e.g., 4.8 comparedwith 4.6). In the absence of added rennet, cuttingis performed at about pH 4.6 so as to prevent ex-cessive loss of fines upon whey separation.

16.6.6 Added Stabilizers

A wide variety of plant- and animal-derivedhydrocolloids (including pectins, pregelatinizedstarch, cellulose derivatives, alginates, carrag-eenans, and gelatin) are added to the milk priorto fermentation (e.g., in yogurt) or to the sepa-rated curd (e.g., Cream cheese and fresh cheeseproducts) to immobilize water and reduce syner-esis. Although these additives are very effective,their inclusion in the product detracts from thenatural image, and some may have adverse ef-fects on flavor and consistency. Recent studieshave shown that fortification of yogurt milk withvarious dairy-based proteins (or blends) can beas effective as adding hydrocolloids in retardingwheying-off. The use of slime-producing cul-tures in yogurt has also been found to reduce sy-neresis.

16.6.7 Packaging and Retailing

Disturbance of set fermented milk products,such as by movement during cartoning andtransport, creates stress for bond breakage and

Table 16-2 Effect of Heat Treatment on the Level of Whey Protein Denaturation in Skim Milk and thePermeability Coefficient of the Resultant Skim Milk Gels

Heat Treatment

720C x 15 s 90°Cx5min

Milk compositionDry matter (g/kg) 98.6 98.4Total protein (g/kg) 36.6 36.5Casein number 75.2 87.2NPN (% total N) 7.3 7.4Whey protein denaturation (% total) 2.5 70.0

Gel CharacteristicsElastic shear modulus, G' at 16 hr (Pa) 100.0 20Permeability coefficient, B (m2) 2.56 x 10~13 1.61 x 10~13

Note: The data presented in Figures 16-6 and 16-11 are for gels obtained from the above milks.

Page 395: Cheese Science

A

B

Figure 16-10 Confocal laser scanning micrographs of gels from skim milk pasteurized at 720C x 15 s (A) and at9O0C x 5 min (B) (described in Table 16-2). Bar equals 10 urn.

Page 396: Cheese Science

matrix rearrangement and may initiate or accen-tuate syneresis. For a given level of syncreticpressure, syneresis increases with increasingsurface area to volume ratio of the gel. Theshape of the package containing the gel may alsoinfluence syneresis. For example, in a packagewith sloping walls, the gel may have a tendencyto detach from the walls, which leads to a stressin the gel and breakage. The breakage of the gelstrands provides new bonding sites and rear-rangement of the gel matrix into a more compactstructure.

16.7 TREATMENTS OF THESEPARATED CURD

In the production of many fresh cheese prod-ucts, the gel produced upon acidification is sub-jected to a number of further processing steps,such as stirring, whey separation or concentra-tion, heating, homogenization, agitation, andcooling (Figure 16-2). Various materials, such

as cream, sugar, salt, fruit purees, and hydrocol-loids, may be added to the curd. Such treatmentsinfluence the structural, rheological, andsyncretic properties of the final product. The ef-fects of various processing steps are discussedbelow.

Cutting of the gel into cubes, as in Cottagecheese, initiates syneresis, which is enhanced bycooking and stirring, as in the manufacture ofrennet-curd cheese. Stirring of the gel (as inQuarg, Cream cheese, and Fromage frais)breaks the matrix strands to an extent that de-pends on the severity of agitation. For a givendegree of agitation, cooling of the gel to below2O0C (e.g., to retard a further decrease in pH be-fore whey separation) may result in more de-struction of the gel matrix. The contribution ofhydrophobic bonds to the integrity of the caseinmatrix decreases upon reducing the temperature(Hayakawa & Nakai, 1985; Kinsella, 1984), In-creasing the temperature of the gel (e.g., in therange 25-850C) enhances whey separation. Anyfactor that increases the firmness of the gel at

Centrifugation force, g

Figure 16-11 Level of syneresis from acid-coagulated gels formed from skim milk pasteurized at 720C x 15 s(•) and at 9O0C x 5 min (Q) (described in Table 16-2). After fermentation of the skim milk by a starter culture at220C, the gels (pH 4.6) were stirred gently and samples were weighed in centrifuge tubes and held at 80C for 36-48 hr. The samples were then centrifuged at 2,000 or 3,000 g. The weight of whey expelled was expressed as apercentage of the original sample weight.

Syn

eres

is (

% w

/w)

Page 397: Cheese Science

separation (e.g., proximity to the isoelectric pH,rennet addition, milk homogenization, higherlevels of gel-forming protein, and increasedtemperature at gelation) renders the gel less sus-ceptible to fracture and disintegration at a givenshear. Whey separation, which may be achievedby pouring the hot fluid onto cheese cloth, ultra-filtration, or centrifugation, results in concentra-tion and aggregation of the broken pieces of gelto a greater or lesser degree. Collision of thepieces of gel during concentration forces theminto close proximity and thus contributes to fur-ther casein aggregation. The moisture content ofthe curd is inversely related to the degree of ag-gregation. Factors that enhance casein aggrega-tion and hence casein dehydration (e.g., highertemperature and higher centrifugation force) re-duce the moisture content. Indeed, in the manu-facture of many fresh acid-curd cheese varieties(such as Cream cheese and Cottage cheese),heating of the curds-whey mixture, after stirringof the gel, is done to induce whey separationand permit efficient recovery of curds with thedesired dry matter level.

Homogenization or shearing of the dehy-drated gel results in destruction of casein con-glomerates to an extent dependent on the magni-tude of the shear and thereby contributes to amore homogeneous size and spatial distributionof the matrix-forming material. The holding ofhot Cream cheese at 75-850C (e.g., in a surgetank prior to packaging) may result in a markedincrease in elasticity. The thickening of the con-sistency is rather similar to the thickening pro-cess (often referred to as "creaming") observedin pasteurized processed cheese products whenheld for a long time at a high temperature (e.g.,> 750C) during processing. However, while im-parting a more elastic character, holding a prod-uct at a high temperature may also lead to thedevelopment of chalkiness, powdery mouthfeel,grittiness, or graininess. These quality defectsmay be attributed to acute protein dehydrationand the consequent formation of compact pro-tein conglomerates. It is noteworthy that a hightemperature and low pH (generally in the range4.5-4.8 for fresh acid-curd products) are very

conducive to casein aggregation. Hence, as fre-quently observed in the commercial productionof Cream cheese, prolonged holding of thecheese at a high temperature (> 750C) causes thecheese to be notably more brittle and firmer andto have a tendency toward elastic fracture. Slowcooling probably accentuates this defect, as pro-tein aggregation has more time to proceed un-hindered before being arrested by the lower tem-perature. The matrix of the cooled Cream cheeseis more or less continuous, with the degree ofcontinuity being governed by the size and spatialdistribution of the matrix-forming material be-fore cooling and the rate of cooling. A finer ma-trix manifests itself in a product that has asmoother appearance and mouthfeel and that isless susceptible to spontaneous wheying-offduring storage. The addition of hydrocolloids tothe curd also minimizes syneresis. Stabilizersthat interact with casein, particularly K-carrag-eenan, may interrupt matrix formation and yielda smoother, softer product.

16.8 MAJOR FRESH ACID-CURDCHEESE VARIETIES

16.8.1 Quarg and Related Varieties

Also referred to as Tvorog in some Europeancountries, Quarg is a cheese of major commer-cial significance in Germany, where annual percapita consumption is about 7.1 kg. Quarg is asoft, homogeneous, mildly supple white cheesewith a smooth mouthfeel and a clean, refreshing,mildly acidic flavor. The product is shelf-stablefor 2-4 weeks at below 80C. Stability refers tothe absence of bacteriological deterioration,wheying-off (syneresis), and the development ofgraininess and overacid or bitter flavors duringstorage.

Quarg is sometimes loosely referred to as theGerman equivalent of Cottage cheese. However,while these cheeses are related, in the sense thatboth are fresh acid-curd products of similar com-position, they are quite different from a produc-tion viewpoint and in sensory aspects. Cottagecheese is a (dressed) granular cheese, and its

Page 398: Cheese Science

granules ideally have a chewy, meat-like tex-ture.

Quarg is normally made from pasteurized(72-850C x 15 s) skim milk cooled to 20-230Cand inoculated with an O-type culture. The milk,at 20-230C, is held for 14-18 hr until the desiredpH of 4.6^.8 is reached. Shortly after cultureaddition (e.g., 1-2 hr), a small quantity of rennet(30-60 RU/100 L) is added when the pH isaround 6.3-6.1. Rennet gives a firmer coagulumat a higher pH, and its addition minimizes caseinloss upon subsequent whey separation and re-duces the risk of overacidity (in the absence ofrennet, a lower pH is required to obtain the samedegree of curd firmness). The fermented gelledmilk is stirred gently (100-200 rpm) into asmooth flowable consistency and pumped to anozzle centrifuge, where it is separated intocurds (Quarg) and whey, containing 0.65%whey protein and 0.19% nonprotein N. TheQuarg is cooled immediately (< 1O0C) en routeto the buffer tank feeding the packaging ma-chine.

Various methods have been employed to re-duce the loss of whey proteins and to increaseyield:

• In the Westfalia thermoprocess, the milk ispasteurized at 95-980C x 2-3 min and thegelled milk (pH 4.6) is heated to 6O0C x « 3min and then cooled to the separating tem-perature (250C). In this process, 50-60% ofthe whey proteins are recovered in thecheese.

• In the centriwhey process, the whey fromthe separator is heated to 950C to precipi-tate the whey proteins. The denatured wheyproteins are recovered by centrifugation inthe form of a concentrate (~ 12-14% drymatter), which is added to the milk for thenext batch of Quarg.

• In the lactal process, the whey from theseparator is heated to 950C to precipitatethe whey proteins, which are allowed tosettle. A concentrated whey (~ 7-8% drymatter) is obtained upon partial decantationof the whey. A whey Quarg (17-18% sol-

ids), which is blended at a level of about10% with regular Quarg, is produced uponfurther concentration using a nozzle centri-fuge.

• Ultrafiltration of the gelled milk is now be-ing used on a large scale for the commercialproduction of Quarg and other fresh cheesevarieties. This method gives full recoveryof whey proteins in the cheese. However,the nonprotein nitrogen fraction, whichamounts to 2-3 g/kg of milk, is not concen-trated and passes into the permeate.

Quarg cheeses that are made from milk pas-teurized at a similar temperature (720C x 15 s)and have the same level of protein (140 g/kg)and dry matter (180 g/kg) may have differentlevels of casein and whey protein, depending onthe curd separation technique. Hence, the levelsof casein and whey protein in skim milk Quargproduced by the separator (centrifuge) or ultra-filtration techniques are approximately 134 and6 g/kg and 110 and 30 g/kg, respectively. How-ever, whey proteins in the native state do not gelunder the cheesemaking conditions used forQuarg made by either the ultrafiltration or cen-trifugation techniques. Hence, while the levelsof total protein in the products produced by thesemethods are similar, the concentrations of gel-forming protein differ. Therefore, suppliers ofultrafiltration units to the Quarg industry recom-mend a high milk pasteurization treatment (950Cx 3-5 min). The high heat treatment results inbinding of denatured whey proteins with thecasein and thereby increases the level of gel-forming protein in ultrafiltration-producedQuarg to the same level as in separator-producedQuarg. Otherwise, ultrafiltration-producedQuarg, although containing the correct level oftotal protein, has a relatively thin consistencydue to the lower level of matrix-building protein(110 g/kg compared to 134 g/kg). Quarg pro-duced by the recommended ultrafiltration proce-dure (i.e., high heat milk treatment prior to cul-turing) has sensory characteristics similar tothose of Quarg produced using the standardseparator process.

Page 399: Cheese Science

Owing to its relatively high moisture (820g/kg) and low protein (140 g/kg) levels, theshelf-life of Quarg is 2-4 weeks at below 80Cowing to microbial growth, syneresis, and off-flavor defects (especially bitterness). Microbio-logical quality can be improved by variousmethods, including the addition of sorbates,modified atmosphere packaging, thermization(58-6O0C) of the broken gel prior to separation,and high heat treatment of the product (contain-ing hydrocolloids). Addition of excessive rennet(> 78 RU/100 L), while increasing yield, leads tobitterness in Quarg after storage at 5-1O0C for 4weeks. Addition of rennet at a level of about 39RU/100 L has been found to give the best com-promise between yield and lack of bitter flavor.It is easier to prevent bitter flavor developmentin ultrafiltration-produced Quarg as the additionof rennet is not necessary, since yield is not af-fected by its addition. Quarg produced from lac-tose-hydrolyzed milk is sweeter and has a yel-lower color than that produced from normalmilk.

Further processing (e.g., heating, homogeni-zation, and/or aeration) and the addition of vari-ous ingredients (e.g., spices, herbs, fruit purees,cream, sugar, other fresh fermented products ofdifferent fat levels, and hydrocolloids) give riseto a range of Quarg-based products such as half-fat (20% FDM) and full-fat (40% FDM) Quarg,fruit and savory Quargs, Shrikhand, dairy des-serts, and fresh cheese preparations.

Labneh, Labeneh, Ymer, and Fromage fraisare similar to Quarg. These products are ver-sions of concentrated natural stirred-curd yogurtand represent the interface between the classicalfresh cheeses (i.e., standard separator Quarg andCream cheese) and yogurt. As for yogurt, themilk is subjected to a high heat treatment(~ 950C x 5 min) in order to cause a high degreeof p-lactoglobulin-K-casein interaction, whichin turn leads to a finer gel network, manifestingitself in the form of a product with a smoothermouthfeel and the ability to occlude more water.Unlike in the case of yogurt, the milk is not nor-mally fortified, and the coagulated milk is con-centrated by various means (pouring into cloth

bags, as is done in traditional Labeneh manufac-ture, or use of a Quarg-type separator or ultrafil-tration).

Production of these products generally in-volves

• standardization and heat treatment of themilk

• acidification by a yogurt-type starter cul-ture to pH 4.6

• concentration of the coagulated milk• homogenization of the curd

They may be flavored by the addition of sugar,fruit purees, or other condiments, which areblended in prior to homogenization. While ac-ceptable products in their own right, they are,like Quarg and Cream cheese, often blendedwith yogurt and other fresh cheeses for the pro-duction of "new" fresh cheese preparations withdifferent compositional, textural, and flavor at-tributes.

16.8.2 Cream Cheese and Related Varieties

Cream cheese (hot pack) is a cream-colored,clean, and slightly acid tasting product with amild diacetyl flavor. Its consistency ranges frombrittle (especially double Cream cheese) tospreadable (e.g., single Cream cheese). Theproduct, which is most popular in NorthAmerica, has a shelf-life of around 3 months atbelow 80C.

Cream cheese is produced from standardized,homogenized, pasteurized (72-750C x 30-90 s)milk (typically with a fatprotein ratio of 2.85:1for double and 1.2:1 for single Cream cheese).Homogenization is important for the followingreasons:

• It reduces creaming of fat during the fer-mentation or acidification stage and there-fore prevents compositional heterogeneityof the resultant gel.

• It reduces fat losses upon subsequent wheyseparation.

• It converts, via the coating of fat with caseinand whey protein, naturally emulsified fat

Page 400: Cheese Science

globules to pseudoprotein particles, whichparticipate in gel formation upon subse-quent acidification. The incorporation of fatinto the gel structure by this means gives asmoother, firmer curd and therefore is espe-cially important to the quality of cold-packCream cheese, in which the curd is nottreated further.

Following pasteurization, the milk is cooled(20-3O0C), inoculated with a D-type starter cul-ture, and held at this temperature until the de-sired pH (~ 4.5^4.8) is reached. The resulting gelis agitated gently, heated, and concentrated byvarious methods:

• draining through muslin bags at 60-9O0Cover 12-16 hr, as in the traditional batchmethod

• continuous concentration using a centrifu-gal curd separator at 70-850C

• ultrafiltration at 50-550C

In the batch method, the curd is cooled toaround 1O0C, and salt (5-10 g/kg) and hydrocol-loid (< 5 g/kg; e.g., sodium alginate and carrag-eenan) are added. Then the treated curd may bepackaged directly as cold-pack Cream cheese,which has a somewhat spongy, aerated consis-tency and a coarse appearance, or heated (70-850C) and sheared by batch cooking (in a pro-cess cheese-type cooker at a relatively highshear rate for 4-15 min) or continuous cooking(in scraped-surf ace heat exchangers). The de-gree of heat and shear and the duration of cook-ing have a major influence on the consistency ofthe final product. Increasing the latter two pa-rameters while keeping the temperature constantgenerally results in an increasingly more elasticand brittle texture. The hot, molten product,known as hot-pack Cream cheese, has a shelf-life of about 3 months at 4-80C.

In the continuous production method, curdfrom the separator is treated continuously withstabilizer via an online metering and mixing de-vice, pumped through a scraped-surface heat ex-changer, homogenized online, and fed to thebuffer tank feeding the packaging machine.

Owing to the thick, viscous consistency of thecurd, concentration by ultrafiltration necessi-tates a two-stage process in order to maintainsatisfactory flux rates and obtain the correct drymatter level. Stage 1 involves standard moduleswith centrifugal or positive displacementpumps, and stage 2 involves high-flow moduleswith positive displacement pumps.

The flavor diversity of Cream cheese may beincreased by adding various flavors, spices,herbs, and sterilized, slurried, deboned fish.Cream cheese-type products may also be pre-pared by blending two or more acid-curd prod-ucts (e.g., fermented cream, Ricotta, Quarg, andcultured buttermilk), then pasteurizing and ho-mogenizing the blend and hot packing. Thesecheeses compare well with commercial doubleCream cheese in all quality aspects.

The manufacture and sensory attributes ofother cream cheese-type products, such asNeufchatel and Petit Suisse, are similar todouble Cream cheese. They differ mainly withrespect to composition. Mascarpone, however,differs from other Cream cheese-type productsin that acidification and coagulation are broughtabout by a combination of chemical acidifica-tion (using food-grade organic acids, such aslactic or citric) to about pH 5.0-5.6 and heat(90-950C) rather than by starter fermentation at20-450C. The hot, acidified cream (400-500g/kg fat), which is Mascarpone cheese, is packedin cartons or tubs and stored at around 50C. Theproduct, which has a shelf-life of 1-3 weeks, hasa soft homogeneous texture and a slightly but-tery, slightly tangy flavor.

16.8.3 Cottage Cheese

Cottage cheese is a soft granular unripenedcheese in which the curd granules are lightlycoated with a salted cream dressing. The flavorranges from a cream-like blandness to mildlyacidic with overtones of diacetyl.

Cottage cheese is made from skimmed, pas-teurized (720C x 15 s) milk inoculated with alactic acid-producing starter at a level depend-ing on the set time. Long-set Cottage cheese is

Page 401: Cheese Science

inoculated with 5-10 g/kg starter and incubatedat 21-230C for 14-15 hr, whereas short-set Cot-tage cheese is inoculated with 5-10 g/kg starterand incubated at 30-320C for 4-5 hr. Thestarter normally consists of lactic acid-produc-ing bacteria (Lc. lactis subsp. cremoris or Lc.lactis subsp. lactis) and flavor-producing bacte-ria (citrate-positive lactococci or Leuconostocmesenteroides subsp. cremoris). The metabo-lism of citrate by the latter results in the pro-duction of the flavor compounds diacetyl andacetate as well as CO2, which vaporizes uponsubsequent cooking (~ 550C) and forms gasbubbles that tend to cause floating of curd par-ticles to the top of the whey. Excessive CO2

production gives rise to the defect known as"floating curd," which reduces the yield. Thecurd is fragile and shatters upon cutting and stir-ring to give fines that are lost in the whey andwash water. However, selection of a suitablestarter with the correct balance of acid and fla-vor producers gives cheese of a satisfactory fla-vor while avoiding the above defect. Much ofthe diacetyl produced (> 3.2 mg/kg) is lost inthe whey. The risk of floating curd is minimizedby removing the flavor-producing strains fromthe culture. Instead, diacetyl may be added di-rectly to the cream dressing or the creamingmixture may be cultured with a diacetyl-pro-ducing starter.

Another starter-related problem in Cottagecheese manufacture is agglutination, and its as-sociated defect is known as "minor sludge for-mation" (i.e., the formation of a layer of fragile,discolored [yellowish] material at the bottom ofthe vat during acidification). Agglutination ofstarter lactococci is caused by immunoglobulinsthat occur naturally in milk as part of the wheyprotein fraction and are at a particularly highlevel in colostrum and mastitic milk. Upon ag-glutination, the starter bacteria clump togetherand settle to the bottom of the cheese vat. Lacticacid production becomes localized, and a pHdifference of around 0.5 unit between the milk atthe top and the bottom of the vat occurs afterabout 4 hours of incubation. Consequently, pre-cipitation of casein (~ 4-8% of the total casein)

results in the formation of a sludge, which shat-ters upon subsequent cutting and stirring to pro-duce fines that are lost during whey drainage andwashing. The risk of starter agglutination is re-duced by homogenizing the skim milk (e.g., at apressure of- 155 bar) or the bulk culture (~ 176bar) or by the addition of lecithin to the bulk cul-ture. Homogenization of skim milk destroysagglutins, while homogenization or addition oflecithin to the culture causes fragmentation ofstarter chains without affecting cell numbers oracid production. The defect known as "majorsludge formation," in which all the casein formsa precipitate (which cannot be made into satis-factory cheese) rather than a gel during acidifi-cation, is thought to be due to phage infection ofstarter after acid development is well advanced(i.e., at pH 5.2-4.9).

As in the manufacture of Quarg, a smallamount of rennet is added to the milk when thepH reaches 6.3 with the aim of increasing gelfirmness at cutting. The optimum pH at cuttingis about 4.8. At a constant cooking temperaturein the range 50-6O0C, the firmness and dry mat-ter of the final Cottage cheese increase with pHat cutting in the range 4.6—4.9. At a cutting pHabove 4.8, the curds tend to mat upon cooking,giving rise to clumping of the curd granules.However, if the heat treatment is more severethan normal pasteurization (i.e., > 720C), thecurds synerese poorly during subsequent stirringand cooking, resulting in a soft, mushy product.Increasing the cutting pH to above 4.9 (e.g., to5.1) and increasing the level of rennet added en-hances the syncretic properties of curd from highheat-treated milk without the risk of matting. Inaddition to heat treatment of the milk, other fac-tors that influence the firmness and syncreticproperties of the curd during cooking determinethe desired cutting pH:

• Milk composition (stage of lactation). At agiven cutting pH, higher casein levels givefirmer gels, which synerese better thanthose from low-protein milk.

• Level of added rennet. Higher levels of ren-net addition promote firmer gels at a given

Page 402: Cheese Science

pH. However, as in Quarg, excess rennetcauses bitterness.

• Grain size. Large curd granules, becausethey require a longer time to dehydrateupon cooking and stirring, tend to be moresusceptible than smaller grains to shatteringand therefore generally necessitate a highercutting pH and hence a coagulum that firmsmore rapidly upon cooking.

All factors being equal, reducing the pH from4.8 to 4.6 tends to give a softer, more fragile co-agulum, an effect that may be due to increasedloss of casein-bound calcium, which impairs theability of the curd to synerese and become firmduring stirring and cooking. The cut size de-pends on whether a large curd (~ 2 cm cut) orsmall curd (~ 1 cm cut) end product is desired.

After cutting, the curds are allowed to settlefor 5-15 min (depending on their firmness) inorder to undergo "healing" of the cut surfaces,then gently agitated and cooked slowly (at therate of increase of I0C per 5 min to 4O0C and20C per 5 min to 550C). Higher cooking tem-peratures enhance syneresis and consequentlygive a product in which the curd granules aremore defined and stronger and have a morechewy, meatlike texture. When the curd par-ticles have acquired the correct degree of resil-ience and firmness, the whey is drained off andthe curds are washed to

• prevent them from matting together• remove lactose and minimize the growth of

spoilage bacteria in the final product• remove lactic acid and therefore prevent the

likelihood of overly acid tasting cheese

Washing involves the addition of water (at a vol-ume equivalent to the volume of whey drained)at about 250C to the curds, stirring for 2-3 min,and then draining. This process is repeated twoor three times using ice-water chlorinated to alevel of 5-25 mg/kg. The wash water is drained,and the cooled curd grains are trenched and al-lowed to stand for at least 1 hr until all wash wa-ter has drained away.

A homogenized, pasteurized, salted (10-40g/kg salt) cream dressing (90-180 g/kg fat) ismixed with the curds in the proper amount tocreate a finished product with the desired levelof salt (8-10 g/kg) and fat. In dry-curd Cottagecheese, the level of fat is below 5 g/kg; in low-fatCottage cheese, it is between 5 g/kg and 20 g/kg;and in Cottage cheese, it is not less than 40 g/kg.The dressing may be cultured or contain addedstarter distillate. The use of such a dressing isadvocated when a plant is experiencing produc-tion difficulties as a result of "floating curd," asdiscussed above.

Recent developments in Cottage cheese man-ufacture include direct-acid-set coagulation us-ing food-grade acid or acidogen and a transitionfrom all-cheese-vat batch operations to continu-ous production systems, where washing, cool-ing, and draining (pressing) of the curds occuron rotating belts.

16.8.4 Ricotta and Ricottone

Ricotta is a soft, cream-colored unripenedcheese with a sweet-cream and somewhat nuttyor caramel flavor and a delicate aerated texture.The cheese, which was produced traditionally inItaly from cheese whey from ewe milk, now en-joys more widespread popularity, particularly inNorth America and Western Europe, where it isproduced mainly from whole or partly skimmedbovine milk or whey-skim milk mixtures.

In the traditional batch production method,the milk or milk-whey blend is directly acidifiedto around pH 5.9-6.0 by the addition of food-grade acid (e.g., acetic, citric, or lactic acid),starter culture (~ 200 g/kg inoculum), or acidwhey powder (~ 25% addition). Heating of themilk to about 8O0C by direct steam injection in-duces coagulation of the casein and whey pro-teins and thus results in the formation of curdfloes in the whey after about 30 min, at whichpoint direct steam heating is discontinued. Thecurd particles, now under quiescent conditions,begin to coalesce and float to the surface, wherethey form into a layer. Indirect steam (applied tothe vat jacket), together with manual movement

Page 403: Cheese Science

of curd from the vat walls toward the center, ini-tiates the process of "rolling," whereby the curdsroll from the walls toward the center of the vatand there form into a layer that is easily recov-ered by scooping (using perforated scoops). Thecurds are filled into perforated molds and al-lowed to drain for 4-6 hr at below 80C.

The above procedure gives only partial recov-ery of the whey proteins. A secondary precipita-tion, in which the whey from Ricotta cheesemanufacture is acidified to pH 5.4 with citricacid, heated to 8O0C, and treated as for Ricotta,is therefore sometimes practiced in order to re-cover remaining whey proteins in the form ofRicottone cheese. Ricottone has a relatively hardand tough consistency and therefore is normallyblended with Ricotta in an attempt to moderateits undesirable features.

Owing to its relatively high pH, its high mois-ture content (Table 16-1), and the manualmethod of filling, Ricotta produced by the tradi-tional method is very susceptible to spoilage byyeasts, molds, and bacteria and hence has a rela-tively short shelf-life—1 to 3 weeks at 40C. How-ever, significant advances have been made in theautomation of Ricotta cheese production with theobjective of improving curd separation, cheeseyield, and shelf-life. Excellent quality Ricotta hasbeen produced using an ultrafiltration-based pro-duction method. Whole milk is acidified withacid whey powder to pH 5.9 and ultrafiltered at550C to 11.6% protein (« 29% dry matter). Theretentate is heated batchwise at 8O0C for 2 min toinduce coagulation (without whey separation).The coagulum is hot packed and has a shelf-life ofat least 9 weeks at 90C. In another process basedon ultrafiltration, milk and/or whey is standard-ized, pasteurized at pH 6.3, cooled to 5O0C, andultrafiltered to 30% dry matter. The retentate isheated to 9O0C and continuously acidified to pH5.75-6.0 at a pressure of 1.0-1.5 bar. The pres-sure is reduced to induce coagulation withoutwhey separation, and the curds are cooled to 7O0Cand hot packed. In a process developed byModler (1988), a 20:80 blend of whole milk andconcentrated whey (neutralized to pH 6.9-7.1) isheated from 40C to 920C, pumped to a 10 min

holding tube (to induce whey protein denatur-ation), and acidified, by on-line dosing with citricacid (250 g/kg), to induce coagulation. The curdsare separated from the "deproteinated" whey on anylon conveyor belt. This process gave excellentrecoveries of fat and protein (99.6 and 99.5 g/kg,respectively). Other methods employed to in-crease yield and automate the production ofRicotta include filtration of whey after curd re-moval and the use of perforated tubes or basketsin the bottom of the curd-forming vat to collectthe curds after whey drainage.

Ricotta cheese, in addition to being an accept-able product itself, has many applications, in-cluding use as a base for whipped dairy desserts,Cream cheese, and pasteurized processed cheeseproducts and use in confectionery fillings andcheesecake.

16.8.5 Queso Blanco

Queso bianco (white cheese) is the genericname for white, semi-hard cheeses produced inCentral and South America. These cheeses canbe consumed fresh but some cheeses may beheld for 2-8 weeks before consumption (Torres& Chandan, 198Ia, 198Ib). Elsewhere in theworld, similar cheeses include Chhana andPaneer in India, Armavir in the WesternCaucasus, Zsirpi in the Himalayas, and low-salt(< 10 g/kg), high-moisture (> 600 g/kg), unrip-ened cheeses in the Balkans (e.g., BeIi sir types).BeIi sir-type cheeses may also be salted and rip-ened in brine for up to 2 months to give whitepickled cheeses usually known by local names,such as Travnicki sir and Sjenicki sir.

In Latin America, Queso bianco covers manywhite cheese varieties, which differ from eachother by the method of production (i.e., acid/heator rennet coagulated), composition, size, shape,and region of production. Examples includeQueso de Cincho, Queso del Pais, and QuesoLlanero, which are acid/heat coagulated, andQueso de Matera and Queso Pasteurizado,which are rennet coagulated. The use of a hightemperature (80-9O0C) during the production ofacid/heat-coagulated white cheeses was, tradi-

Page 404: Cheese Science

tionally, very effective for improving the keep-ing quality in warm climates.

In general, Queso bianco-type cheeses arecreamy, highly salted, and acid in flavor. Theirtexture and body resemble those of very younghigh-moisture Cheddar, and they have good slic-ing properties. The average composition of afresh cheese is 40-50% moisture, 22-25% pro-tein, and 15-20% fat (Kosikowski & Mistry,1997; Torres & Chandan, 198Ia).

The production method for acid/heat-coagu-lated Queso bianco varies but generally involvesthe following steps:

• Standardization of milk to the required pro-tein: fat ratio to achieve the desired end-product composition.

• Heat treatment of the milk to about 82-850C, followed by holding for about 5 min.This heat treatment achieves partial dena-turation (~ 600-700 g/kg) of whey proteins,which complex with the caseins and aretherefore recovered with the casein uponsubsequent coagulation.

• Acidification of the hot milk to pH 5.3 byadding food-grade acid (acetic, citric, ortartaric acid, lime juice, or lactic culture) tothe milk while stirring gently. Citric andacetic acid are used most frequently, some-times jointly. The acids are diluted prior toaddition, typically to a concentration of 50-100 g/L, to facilitate dispersion and preventlocalized coagulation.

• Curd formation. Protein aggregation occursrapidly under nonquiescent conditions ow-ing to the low pH and high temperature ofthe milk, resulting in the formation of curdparticles and whey.

• Curd recovery, salting, molding, and press-ing. The curd particles are separated fromthe whey and dry stirred, dry-salted (at alevel sufficient to result in ~ 20-40 g/kg inthe final cheese), and pressed. The pressedcheese is cut into consumer-size portions,which are vacuum packed and stored at4-80C. The product is shelf-stable at thistemperature for 2-3 months.

Queso bianco is traditionally consumed freshbecause, as a result of high heat treatment duringcurd formation, very few biochemical changesoccur during storage. However, starter bacteria(Lactobacillus spp) and/or exogenous Upasesmay be added to the curd before salting and press-ing to improve the flavor of the cheese duringstorage. Major volatile compounds that contrib-ute to the flavor and aroma of Queso bianco in-clude acetaldehyde, acetone, isopropanol, bu-tanol and formic, acetic, propionic, and butyricacids. The pH of Queso bianco decreases fromabout 5.2 to 4.9 during ripening, an effect thatmay be due to fermentation of residual lactose tolactic acid by heat-stable indigenous bacteria inmilk that survive cheesemaking or by post-manufacture contaminating bacteria (Torres &Chandan, 198Ib).

One of the interesting properties of the cheeseis its flow resistance upon heating, owing to theinclusion of whey proteins that gel upon heating.This enables the cheese to be deep-fat fried inthe preparation of many savory snack foods,such as cheese sticks in batter.

16.8.6 Whey Cheeses

Brunost, meaning "brown cheese," refers to adistinctive group of Norwegian unripened"cheese" varieties made from sweet whey (ren-net casein or cheese whey) or skim milk, towhich cream may be added (Kosikowski &Mistry, 1997; Otterholm, 1984). The best knownmembers of the group include Mysost andGudbrandsdalsost (> 350 g/kg fat-in-dry-matter[FDM]; from bovine and caprine milk), Ek-tegeitost (> 33% FDM; from goat milk compo-nents), Flotemysost (> 33% FDM; from bovinemilk), and Primost.

In the classical sense, Mysost is not a cheesebut rather a fat-protein-enriched concentratedheated whey. However, being unripened, it maybe defined as a "fresh cheese." The cheese, char-acterized by a light golden to a dark brown color,is produced using these steps:

• In this step, known as "standardization," thewhey, milk, and/or cream are blended to

Page 405: Cheese Science

give the correct end-product FDM. Thewhey is first filtered or decanted to removecasein particles, which otherwise occur inthe product as black-brown specks.

• The standardized whey is preconcentratedin a multistage film evaporator to 50-60%total solids. The viscous concentrate is thentransferred to special steam-jacketed, coni-cal kettles in which it is further concen-trated to 800-820 g/kg dry matter by being

REFERENCES

Guinea, T.P., Mullins, C.G., Reville, W.J., & Cotter, M.P.(1995). Physical properties of stirred-curd unsweetenedyoghurts stabilized with different dairy ingredients.Milchwissenschaft, 50, 196-200.

Guinee, T.P., Pudja, P.O., & Farkye, N. Y. (1993). Freshacid-curd cheese varieties. In P.P. Fox (Ed.), Cheese:Chemistry, physics and microbiology (2d ed., Vol. 1).London: Chapman & Hall.

Hayakawa, S., & Nakai, S. (1985). Relationships of hydro-phobicity and net charge to the solubility of milk and soyproteins. Journal of Food Science, 50, 486^91.

Kinsella, I.E. (1984). Milk proteins: Physicochemical andfunctional properties. Critical Reviews in Food Scienceand Nutrition, 21, 197-262.

Kosikowski, F.V., & Mistry, V. V. (1977). Cheese and fer-mented milk food: Vol. 1. Origins and principles.Westport, CT: F.V. Kosikowski, LLC.

Modler, H.W. (1988). Development of a continuous process

heated, while agitated vigorously, under avacuum.

• The molten viscous mass is transferred to avessel with a strong rotary, swept metalagitator. Kneading over a 20 min periodwhile slow atmospheric cooling occurshelps to give the product its butterlike,plastic consistency and prevents the forma-tion of large lactose crystals (and thereforegrittiness).

for the production of Ricotta cheese. Journal of DairyScience, 71, 2003-2009.

Oterholm, A. (1984). Cheesemaking in Norway [BulletinNo. 171]. Brussels: International Dairy Federation.

S0rensen, H.H. (1997). The world market for cheese [Bulle-tin No. 326]. Brussels: International Dairy Federation.

Torres, N., & Chandan, R.C. (198Ia). Latin American whitecheese [A review]. Journal of Dairy Science, 64', 552-557.

Torres, N., & Chandan, R.C. (198Ib). Flavor and texture de-velopment in Latin American white cheese. Journal ofDairy Science, 64, 2161-2169.

Walstra, P., van Dijk, H.J.M., & Geurts, TJ. (1985). The sy-neresis of curd: 1. General considerations and literaturereview. Netherlands Milk and Dairy Journal, 39, 209-246.

Walstra, P., & van Vliet, T. (1986). The physical chemistryof curd-making. Netherlands Milk and Dairy Journal, 40,241-259.

Page 406: Cheese Science

17.1 INTRODUCTION

The diversity of cheese types is truly breath-taking. Despite the limited range of raw materi-als (bovine, ovine, caprine, or buffalo milk), ap-proximately 500 varieties of cheese recognizedby the International Dairy Federation (Burk-halter, 1981) are produced. Numerous other mi-nor local cheeses are also manufactured.

In order to facilitate their study, a number ofattempts have been made to classify cheese vari-eties into meaningful groups or families. As dis-cussed by Fox (1993a), traditional classificationschemes have been based principally on mois-ture content, such as hard, semi-hard, or soft.Although this is a widely used basis for classifi-cation, it suffers from a serious drawback: itgroups together cheeses with widely differentcharacteristics and manufacturing protocols. Forexample, Cheddar, Parmesan, and Emmental areoften grouped together as hard cheeses, althoughthey have quite different flavors and the methodsfor their manufacture are quite different. At-tempts have been made to make this schememore discriminating by including factors such asorigin of the cheese milk, moisture content, tex-ture, principal ripening microorganisms, andcooking temperature (Exhibit 17-1). A classifi-cation scheme proposed by Walstra (see Fox,1993a) differentiates between varieties based onthe type of primary and secondary starter usedand moisture content (expressed as the mois-ture:protein ratio). Walter and Hargrove (1972),

who classified cheeses on the basis of manufac-turing technique, suggested that there are only18 distinct types of natural cheese, which theygrouped into 8 families under the headings veryhard, hard, semi-soft, and soft:

1. Very hard (grating)1.1 Ripened by bacteria (e.g., Parmesan)

2. Hard2.1 Ripened by bacteria, without eyes

(e.g., Cheddar)2.2 Ripened by bacteria, with eyes (e.g.,

Emmental)3. Semi-soft

3.1 Ripened principally by bacteria(e.g., Gouda)

3.2 Ripened by bacteria and surface mi-croorganisms (e.g., Limburger)

3.3 Ripened principally by blue mold inthe interior (e.g., Roquefort)

4. Soft4.1 Ripened (e.g., Brie)4.2 Unripened (e.g., Cottage)

Unfortunately, none of these schemes is com-pletely satisfactory and thus none is universallyaccepted. Fox (1993a) proposed a number of"superfamilies" into which all cheeses would begrouped based on the method of milk coagula-tion:

• rennet-coagulated cheeses (most major in-ternational cheese varieties)

• acid-coagulated cheeses (e.g., Cottage andQuarg)

Principal Families of Cheese

CHAPTER 17

Page 407: Cheese Science

1.6 FreshFresh, acidL 4 Fresh, rennet 1.51.3 Soft (> 55% H2O)1.2 Semi-hard/semi-softsoft (43-5 5% H2O)

1. COWMILK

1.1 Hard (< 42% H2O)

1.3.1 Blue veined

1.3.2 White surface mold

1.3.3 Bacterial surface smear

1.3.4 No rind

1.2.1 Small round openings

1.2.2 Irregular openings

1.2.3 No openings

1.2.4 Blue veined

1.1.1 Grating cheese (extra hard)

1.1.2 Large round openings

1.1.3 Medium round openings

1 . 1 .4 Small round openings

1.1.5 Irregular openings

1.1.6 No openings

2. SHEEP MILKHard; semi-hard; soft;blue-veined; fresh

3. GOATMILK

4. BUFFALO MILK

Note: Unless otherwise stated, the cheeses are internally bacterial ripened.

Exhibit 17-1 Classification of Cheese According to Source of Milk, Moisture Content, Texture, and Ripening Agents

Page 408: Cheese Science

• heat/acid coagulated (e.g., Ricotta)• concentration/crystallization (e.g., Mysost)

All ripened cheeses are coagulated by rennet(~ 75% of total world production). Acid-curdcheeses (see Chapter 16) are the next most im-portant group. Coagulation by a combination ofheat and acid is used for a few minor varieties,including Ricotta. Concentration/crystallizationis used in Norway to produce "whey cheeses"(e.g., Mysost).

There is a great diversity of rennet-coagu-lated cheeses, and therefore they must be classi-fied further. One such classification scheme isproposed in Figure 17-1. Rennet-coagulated va-rieties are subdivided into relatively homoge-neous groups based on the characteristic ripen-ing agents and/or manufacturing technology.The most diverse family of rennet-coagulatedcheeses is that containing the internal bacteri-ally ripened varieties, which include most hardand semi-hard cheeses. The term internally bac-terially ripened may be somewhat misleading,since indigenous milk enzymes and residual co-agulant also play important roles in the ripeningof these cheese varieties. This group may besubdivided based on moisture content (extrahard, hard, and semi-hard) and on whether thecheese has eyes. Many varieties produced on alarge industrial scale are included in this group.Parmesan (extra hard) is used as a gratingcheese, and its manufacture is characterized bya high cook temperature. Cheddar and Britishterritorial varieties (for which the curds are of-ten textured and dry-salted) are classified ashard or semi-hard internal bacterially ripenedcheeses. Internal bacterially ripened cheeseswith eyes are further subdivided on the basis ofmoisture content into hard varieties (e.g., Em-mental), in which the numerous large eyes areformed by CO2 produced on fermentation oflactate by Propionibacterium freudenreichiisubsp. shermanii, and semi-hard varieties (e.g.,Edam and Gouda), in which a few small eyesdevelop due to the formation of CO2 by fermen-tation of citrate by a component of the starter(see Chapters 10 and 11).

Most of the varieties classified in groups otherthan internal bacterially ripened cheeses are softor semi-hard. Pasta filata cheeses (e.g., Mozza-rella) are characterized by stretching in hot water,which texturizes the curd prior to salting. Mold-ripened cheeses are subdivided into surfacemold-ripened varieties (e.g., Camembert andBrie), in which ripening is characterized by thegrowth ofPenicillium camemberti on the surface,and internal mold-ripened cheeses (Bluecheeses), in which P. roqueforti grows in fissuresthroughout the mass of the cheese. Surfacesmear-ripened cheeses are characterized by thedevelopment of a complex microflora consistinginitially of yeasts and ultimately of bacteria (par-ticularly coryneforms) on the cheese surface dur-ing ripening. White, brined cheeses, includingFeta and Domiati, are ripened under brine andhave a high salt content, and consequently theyare grouped together in a separate category.

The classification scheme proposed in Figure17-1 is not without inconsistencies. A cursoryglance will show that cheeses made from themilk of different species are grouped together(e.g., Roquefort and Gorgonzola are Bluecheeses but the former is made from sheep milkand the later from cow milk) and that the subdi-vision between hard and semi-hard cheeses israther arbitrary. There is also some crossoverbetween categories. Gruyere is classified as aninternal bacterially ripened variety with eyes,but it is also characterized by the growth of a sur-face microflora, while some cheeses classifiedas surface ripened (e.g., Havarti and Port duSalut) are often produced without a surface floraand therefore are, in effect, soft internal bacteri-ally ripened varieties. Likewise, Pasta filata andhigh-salt varieties are considered as separatefamilies because of their unique technologies(stretching and ripening under brine, respec-tively), but they are actually ripened by the sameagents as internal bacterially ripened cheeses.However, we believe that the scheme proposedin Figure 17-1 is a useful basis for classification,and therefore the diversity of cheeses will bediscussed under these headings. Ultrafiltrationtechnology is used for the manufacture of some

Page 409: Cheese Science

ACID COAGULATEDCottageCreamQuargQueso Blanco

CHEESE

RENNET COAGULATED

Internal Bacterially-Ripened

Surface Mould(usually P. camemberti)

BrieCamembertCarre de VEst

Mould-Ripened

HEAT/ACID COAGULATIONRicotta

CONCENTRATION/CRYSTALLIZATIONMysost

Surface-RipenedBrickHavartiLimburgerMunsterPort du SalutTrappistTaleggioTilsit

Internal Mould(usually P. roqueforti)

RoquefortDanabluGorgonzolaStilton

Extra-Hard

Grana PadanoParmesanAsiagoSbrinz

Hard

CheddarCheshireGravieraRas

Semi-Hard

CaerphillyMahonMonterey Jack Swiss-type

(Lactate metabolismby Propionibacterium spp.)

EmmentalGruyereMaasdam

Cheeses with eyes

Dutch-typeEyes caused by

citrate metabolism

EdamGouda

High Salt VarietiesDomiatiFeta

Pasta-Filata Varieties

MozzarellaKashkavalProvolone

Figure 17-1 The diversity of cheese. Cheese varieties are classified into superfamilies based on the method of coagulation and further subdivided based onthe principal ripening agents and/or characteristic technology.

Page 410: Cheese Science

cheese varieties, which are discussed separately(Section 17.6). The following discussion isbased largely on descriptions of cheeses given inFox (1993b), Kosikowski and Mistry (1997),Robinson (1995), Robinson and Wilbey (1998),and Scott (1986), to whom the reader is referredfor detailed manufacturing protocols. In somecases, the manufacturing protocols described inthese sources for certain varieties are inconsis-tent, and therefore should be treated with duecaution. This divergence is particularly true forminor varieties, which are probably ill definedand variable in any case. The typical composi-tion of a selection of cheeses is shown in Appen-dix 17-A.

A number of cheese varieties have ProtectedDesignation of Origin (PDO) (or "Appellationd'Origine Controlee") status, which recognizesa specific heritage and provides consumers witha guarantee of authenticity. Unlike commercialtrademarks, PDO denomination reflects a col-lective heritage and may be used by all produc-ers of a particular cheese in a particular geo-graphical area. PDO cheeses are protected by theEuropean Union under various internationalagreements (Bertozzi & Pandri, 1993). PDO de-nomination also certifies that the cheese hasbeen made using specified (usually traditional)technology. A list of cheeses with PDO status isgiven in Table 17—1.

17.2 RENNET-COAGULATED CHEESES

17.2.1 Internal Bacterially Ripened Varieties

Internal bacterially ripened cheese varietiesform a very diverse group of cheeses character-ized by the absence of a surface microflora orinternal mold growth. The agents that contributeto the ripening of internal bacterially ripened va-rieties originate from the milk (plasmin andother enzymes), the rennet (chymosin and/orother proteinases and, in certain cases, Upases),and the internal bacterial microflora (starter andnonstarter bacteria in all cases and adjunctstarter bacteria in some cheeses, particularlythose in which eyes develop). Some internal

bacterially ripened varieties are easily classifiedinto homogeneous groups based on some dis-tinctive technology (e.g., Pasta filata cheeses orcheeses ripened under brine). However, the clas-sification used here for most varieties is basedon texture (extra hard, hard, and semi-hard) andis therefore somewhat arbitrary.

Extra Hard Varieties

The majority of extra hard internal bacteriallyripened cheese varieties originated in Italy.Guinee and Fox (1987) grouped extra hard Ital-ian-type cheeses into 3 subcategories: Parmesanand related varieties, Asiago, and Romano.These cheeses are usually matured for a long pe-riod (2 years or more) and often have a hard,grainy texture. They may be consumed as tablecheeses when young or in grated form when ma-ture. The hard texture of these cheeses resultsfrom the use of semi-skimmed milk in theirmanufacture, a high cooking temperature, andevaporation of moisture during ripening.

Parmigiano-Reggiano and Grana Padano areimportant members of this group. They are pro-duced from raw milk in the Po valley in North-ern Italy and are protected by designation of ori-gin. A grainy texture in the mature cheese isdesirable (hence the name "Grana"; Figure17-2). Parmesan-type cheeses are made world-wide, particularly in the United States, from pas-teurized milk. These cheeses are often smallerthan Italian Grana-type cheeses, are cooked to alower temperature (~ 5O0C) than the traditionalproduct (540C), are salted more heavily, are rip-ened for a shorter period, and are usually used ingrated form.

The manufacturing protocol for Parmigiano-Reggiano is summarized in Figure 17-3. Themajor features of its manufacture are the use ofsemi-skimmed raw milk produced by gravitycreaming. In traditional manufacture, theevening milk is creamed in shallow vats and thelower skimmed milk layer is drawn off andmixed with whole morning milk. Traditionally,the milk was coagulated and the curds cooked incopper vats that were shaped like an inverted belland heated by a fire underneath. Modern vats are

Next Page

Page 411: Cheese Science

cheese varieties, which are discussed separately(Section 17.6). The following discussion isbased largely on descriptions of cheeses given inFox (1993b), Kosikowski and Mistry (1997),Robinson (1995), Robinson and Wilbey (1998),and Scott (1986), to whom the reader is referredfor detailed manufacturing protocols. In somecases, the manufacturing protocols described inthese sources for certain varieties are inconsis-tent, and therefore should be treated with duecaution. This divergence is particularly true forminor varieties, which are probably ill definedand variable in any case. The typical composi-tion of a selection of cheeses is shown in Appen-dix 17-A.

A number of cheese varieties have ProtectedDesignation of Origin (PDO) (or "Appellationd'Origine Controlee") status, which recognizesa specific heritage and provides consumers witha guarantee of authenticity. Unlike commercialtrademarks, PDO denomination reflects a col-lective heritage and may be used by all produc-ers of a particular cheese in a particular geo-graphical area. PDO cheeses are protected by theEuropean Union under various internationalagreements (Bertozzi & Pandri, 1993). PDO de-nomination also certifies that the cheese hasbeen made using specified (usually traditional)technology. A list of cheeses with PDO status isgiven in Table 17—1.

17.2 RENNET-COAGULATED CHEESES

17.2.1 Internal Bacterially Ripened Varieties

Internal bacterially ripened cheese varietiesform a very diverse group of cheeses character-ized by the absence of a surface microflora orinternal mold growth. The agents that contributeto the ripening of internal bacterially ripened va-rieties originate from the milk (plasmin andother enzymes), the rennet (chymosin and/orother proteinases and, in certain cases, Upases),and the internal bacterial microflora (starter andnonstarter bacteria in all cases and adjunctstarter bacteria in some cheeses, particularlythose in which eyes develop). Some internal

bacterially ripened varieties are easily classifiedinto homogeneous groups based on some dis-tinctive technology (e.g., Pasta filata cheeses orcheeses ripened under brine). However, the clas-sification used here for most varieties is basedon texture (extra hard, hard, and semi-hard) andis therefore somewhat arbitrary.

Extra Hard Varieties

The majority of extra hard internal bacteriallyripened cheese varieties originated in Italy.Guinee and Fox (1987) grouped extra hard Ital-ian-type cheeses into 3 subcategories: Parmesanand related varieties, Asiago, and Romano.These cheeses are usually matured for a long pe-riod (2 years or more) and often have a hard,grainy texture. They may be consumed as tablecheeses when young or in grated form when ma-ture. The hard texture of these cheeses resultsfrom the use of semi-skimmed milk in theirmanufacture, a high cooking temperature, andevaporation of moisture during ripening.

Parmigiano-Reggiano and Grana Padano areimportant members of this group. They are pro-duced from raw milk in the Po valley in North-ern Italy and are protected by designation of ori-gin. A grainy texture in the mature cheese isdesirable (hence the name "Grana"; Figure17-2). Parmesan-type cheeses are made world-wide, particularly in the United States, from pas-teurized milk. These cheeses are often smallerthan Italian Grana-type cheeses, are cooked to alower temperature (~ 5O0C) than the traditionalproduct (540C), are salted more heavily, are rip-ened for a shorter period, and are usually used ingrated form.

The manufacturing protocol for Parmigiano-Reggiano is summarized in Figure 17-3. Themajor features of its manufacture are the use ofsemi-skimmed raw milk produced by gravitycreaming. In traditional manufacture, theevening milk is creamed in shallow vats and thelower skimmed milk layer is drawn off andmixed with whole morning milk. Traditionally,the milk was coagulated and the curds cooked incopper vats that were shaped like an inverted belland heated by a fire underneath. Modern vats are

Previous Page

Page 412: Cheese Science

Table 17-1 Cheeses with Protected Designations of Origin

Milk Variety

FranceBovine Abondance

BeaufortBleu d'AuvergneBleu des GaussesBleu de Gex-Haut Jura-SeptmoncelBrie de MeauxBrie de MelunCamembert de NormandieCantalChaourceComteEpoisse de BourgogneFourme de'Ambert ou MontbrisonLaguioleLangresLivarotMaroilles or MarchesMont d'Or/Vacherin du Haut DoubsMunster or Munster GeromeNeufchatelPont I'EvequeReblochon and Petit ReblochonSaint NectaireSalers

Caprine Chabichou du PoitouCrottin de ChavignolPicodon de PArdeche/DromePouligny Saint PierreSainte Maure de TouraineSelles sur Cher

Ovine Ossau-lraty-Brebis-PyreneesRoquefort

Caprine-ovine whey Brocciu Corse or Brocciu

SpainBovine Cantabria

Mahon

Ovine IdiazabalManchegoRoncal

continues

Page 413: Cheese Science

Table 17-1 continued

Milk Variety

Mixed (cow, goat, ewe) CabralesLiebana

PortugalBovine San Jorge

Ovine AzeitaoSerpaSerra da EstrelaCastelo BlancoPicante da Beira Baixa Amarelo

ItalyBovine Asiago

BraCastelmagnoFontinaFormai de MutGorgonzolaGrana PadanoMontasioMurazzanoParmigiano-ReggianoRascheraRobiola di RoccaveranoTaleggio

Ovine Canestrato PuglieseCasciotta di UrbinoFiore SardoPecorino RomanoPecorino SicilianoPecorino ToscanoPecorino Sardo

often made from copper-plated steel and havesteam jackets for heating, but the traditionalshape is retained. Acidification is by a whey cul-ture prepared by incubating whey from the previ-ous day's manufacture. Calf rennet is used tocoagulate the milk, and the coagulum is brokenby means of a wire basketlike implement (spino)(see Chapter 7). The curds are cooked to 53-550C in 10-12 min and then transferred to a mold

large enough to produce a cheese of 25-^0 kg.The cheeses may be subjected to light pressureand turned frequently to encourage whey expul-sion. The cheeses are brine-salted for 20-23 daysand ripened at 16-180C and 85% equilibriumrelative humidity (ERH) for 18-24 months. Therind of these cheeses is cleaned frequently.

The manufacturing protocol for GranaPadano cheese is generally similar to that for

Page 414: Cheese Science

Parmigiano-Reggiano, except that it is madefrom raw milk from a single batch of milk that ispartially skimmed after creaming for about 8hours. Grana Padano is brine-salted for about 25days and ripened for 14-16 months.

Asiago is produced in the province of Vicenza,Italy, from partly skimmed raw cow milk. Rennetpaste is used to coagulate the milk, and a naturalwhey starter is added for acidification. Thecheeses, weighing 8-12 kg, are matured for vari-ous lengths of time, depending on the intensity offlavor desired. Mature Asiago (-12 months old)has a hard, granular texture. Montasio, whichoriginated in northeastern Italy, is made from raw

cow milk and coagulated using calf rennet ex-tract. Montasio may be consumed as a tablecheese after 2-3 months or it may be matured fora longer period (14-18 months), during whichtime the cheese hardens and becomes suitable forgrating. Sbrinz is a hard grating cheese that origi-nated in Switzerland. It is made from full-fat cowmilk using rennet extract and a natural thermo-philic whey starter. The curds are cooked to about570C, placed in molds, pressed for 2-3 days, andeither dry- or brine-salted. The cheese may beconsumed as a table cheese after a short ripeningperiod or matured for up to 3 years and used forgrating.

Figure 17-2 Grana cheese.

Page 415: Cheese Science

Romano-type cheeses are important membersof the extra hard group. Cheese manufactured inSardinia is called Sardo and that in Sicily iscalled Siciliano. The adjectives Pecorino,Vacchino, and Caprino indicate whether the

cheese was made from ewe, cow, or goat milk,respectively. Italian Pecorino Romano cheese ismade from sheep milk using a thermophilicstarter (commercial or whey based). Rennetpaste is used as coagulant. The high lipolytic ac-

Figure 17-3 Manufacturing protocol for Parmigiano-Reggiano, an extra hard Italian cheese variety.

Parmigiano-Reggiano Cheese

2 years

Ripening

Brine salting

Further whey drainage, light pressure

Curds placed in mould

Whey

Cook to 53-550C

Coagulum broken

Coagulum

Calf rennetAcidification by natural whey cultures

Copper-plated conical vat

Semi-skimmed milkMorning milk

Raw Cows1 Milk

CreamGravity creaming overnight

Evening milk

Raw Cows' Milk

Page 416: Cheese Science

tivity of the rennet paste results in the develop-ment of a strong, slightly rancid flavor in themature cheese (the fat in ewe milk has a highcontent of short-chain, middle-chain, andbranched fatty acids, which give it a characteris-tic flavor). The curds-whey mixture is cooked to45-480C, and the whey is then drained off.Blocks of curd are placed in molds and pressedlightly before brine- or dry-salting. The cheesesare ripened for about 8 months. PecorinoRomano cheese is usually grated and used as acondiment.

Hard Varieties

Hard pressed varieties include some of themost commercially important cheeses producedworldwide (e.g., Cheddar; Figure 17-4). Thereis some heterogeneity in manufacturing technol-ogy for cheeses within this group, and whether avariety should be considered a hard cheese is notalways clear. However, hard varieties usuallyhave a moisture content in the range 30^5%and are subjected to high pressure during manu-facture to give a hard, uniform, close texture.According to Robinson (1995), the manufactureof these cheeses has a number of features incommon, including renneting at about 3O0C,cutting the coagulum into small pieces and cook-ing to 39-4O0C, and whey drainage. In the caseof some hard cheeses, such as Cheddar and otherBritish varieties, the curds are textured in the vat(cheddared) and are milled and dry-salted whensufficient acidity has developed. The saltedcurds are then molded and pressed at a high pres-sure for 12-16 hr or longer and matured for 3-12months. Hard cheese varieties include Cheddar;Cheshire, Derby, Gloucester, and Leicester(British Territorial varieties); Cantal (French);Friesian Clove cheese and Leiden (The Nether-lands); Graviera and Kefalotiri (Greece);Manchego, Idiazabal, Roncal, and La Serena(Spain); and Ras (Egypt).

Cheddar cheese originated around the villageof Cheddar, England, and is now one of the mostimportant cheese varieties worldwide. Cheddaris produced on a large scale in most English-speaking countries, particularly in the United

States, United Kingdom, Australia, NewZealand, Canada, and Ireland. Cheddar cheese isusually made from pasteurized whole cow milkstandardized to a caseiir.fat ratio of 0.67-0.72:1and coagulated using calf rennet or a rennet sub-stitute. The starter used is Lc. lactis subsp.cremoris or Lc. lactis subsp. lactis. Definedstrain starter systems are now used in largeCheddar factories in New Zealand, Australia,Ireland, and the United States, but mixed, unde-fined cultures are also used (see Chapter 5). Themilk is renneted at about 3O0C, and the coagu-lum is cut and cooked to 37-390C over 30 minand held at this temperature for about 1 hr. Thewhey is then drained and the curds are ched-dared. The traditional cheddaring process in-volves piling blocks of curd on top of each other,with regular turning and stacking of the curdblocks. The cheddaring process allows time foracidity to develop in the curd (pH decreasesfrom around 6.1 to 5.4) and subjects the curds togentle pressure, which assists in whey drainage.During the cheddaring process, the curd gran-ules fuse and the texture changes from soft andfriable to quite tough and pliable. The curdshould have a texture similar to cooked chickenbreast meat at the end of cheddaring. When thepH has reached around 5.4, the curd blocks aremilled into small pieces and dry-salted. A "mel-lowing" period follows, during which the saltdissolves in moisture on the surface of the curdchips. The curds are then molded and pressedovernight at up to 200 kN/m2. Cheddar is ma-tured at 6-1O0C for a period ranging from 3-4months to 2 years, depending on the maturitydesired.

Although this traditional manufacturing pro-cess (Figure 17-5) is still practiced on a farm-house scale, most Cheddar cheese is now manu-factured in highly automated factories (e.g.,Figure 17-6). The principal features of auto-mated Cheddar production include the use of anumber of cheese vats in which cheesemakingcommences at 30 min intervals to provide asemicontinuous supply of curd. Whey drainageis mechanized and automated, as is the ched-daring process, which occurs as the curd passes

Page 417: Cheese Science

continuously through a large tower in which thecurds are pressed gently by the weight of the col-umn of curds above or, alternatively, on a beltsystem. The curds are then milled and saltedmechanically on a belt system. Pressing andmolding are done automatically by pneumati-cally conveying the salted curds to the top of ablock former, which is a large (Wincanton)tower in which the curds are compressed by theirown weight. A close texture is ensured by apply-ing a vacuum. As the curds exit the blockformer, 20 kg blocks are cut off by a guillotineand vacuum packaged in plastic bags, placed incardboard boxes, stacked on a pallet, and trans-ferred to the cheese store. In many large facto-ries, the boxed cheeses are cooled rapidly by be-ing passed through a forced air cooling tunnelbefore palleting. The objective is to retard thegrowth of nonstarter lactic acid bacteria(NSLAB), which may cause defects in flavorand texture. Most Cheddar is now produced in

block form, although traditional Cheddar has acylindrical shape. Annatto (see Chapter 2) or an-other colorant may be added to milk for Cheddarcheese; the resulting product is known as RedCheddar.

Cheshire cheese is a British Territorial variety(such varieties originated in various parts ofBritain) with a hard texture and perhaps somemechanical openings. Its manufacture is charac-terized by rapid acidification and a lower cooktemperature (32-350C) than Cheddar, which re-sults in a lower pH, a higher moisture content,and a shorter manufacturing time. The curd matthat develops while the curd mass stands on thebottom of the vat after whey drainage is brokenfrequently to prevent the development of an ex-tensive structure in the curd mass. Cheshirecurds are dry-salted, placed in molds, allowed todrain overnight, and then pressed. The cheesesare packaged and matured at 6-80C. Leicester issimilar to Cheshire but is normally colored with

Figure 17-4 Cheddar cheese.

Page 418: Cheese Science

Figure 17-5 Traditional protocol for the manufacture of Cheddar cheese.

Cheddar Cheese

Ripening 0.5-2 years at 6-80C

Moulding and Pressing

Dry Salting

NaCl, (-2%, w/w)

pH -5.2Milling of the Curd

Cheddaring of the Curd

Whey Drainage

Cook

-6 mm cubes

Raise temperature: 3O0C to 37-390Cover -30 min.

Cook at 37-390C for-Ih

Cut the Coagulum

Coagulum

Rennet (1:15,000)(CaCl2, 0.02%, w/v)

Starter (1-2%, v/v):Lactococcus lactis ssp. cremorisand/orLactococcus lactis ssp. lactis

310C

Pasteurized Cows' Milk

Page 419: Cheese Science

Figure 17-6 Large-scale Cheddar cheesemaking factory incorporating the CheddarMaster 3 system with a Cheddaring tower. (1) Pasteurization and fatstandardization, (2) protein standardization using UF, (3) cheesemaking, (4) draining conveyor, (5) cheddaring tower, (6) salting/mellowing conveyor, (7)block former, (8) vacuum packaging, (9) cheese block packing, and (10) main process control panel.

Page 420: Cheese Science

annatto. It is made from cow milk using a meso-philic starter and is cooked at about 370C. Thecurds are pressed after whey drainage, andblocks of curd are placed on a draining rack andturned and cut to promote further whey drain-age. The cheeses are dry-salted and matured for4-8 months at 10-150C. Derby is similar to Le-icester and is somewhat softer and flakier thanCheddar. Gloucester, another British Territorialvariety, has a cylindrical shape and is about 40cm in diameter. Single Gloucester is 6-8 cmhigh while a cylinder of Double Gloucester is15-20 cm high. The manufacturing procedurefor Gloucester is similar to that for Cheddar, andannatto is added to the milk to color the curds,which are cooked to 35-380C. The curds are tex-tured, milled twice, dry-salted, pressed, and ma-tured for 4-6 months.

Cantal is a hard French cheese from theAuvergne region and is manufactured by a pro-cess somewhat similar to that used for BritishTerritorial varieties. The milk is coagulated usingstandard calf rennet and acidified by a mesophiliclactic starter. The curds-whey mixture is notcooked, but the whey is drained and the curds arecheddared. Weights may be placed on the bed ofcurd to assist in whey drainage. The blocks ofcurd are then milled, dry-salted, molded, pressed,and matured at 8-1O0C for 3-6 months.

Kefalotiri is a Greek cheese made from pas-teurized sheep or goat milk standardized toabout 6.0% fat. The milk is inoculated with athermophilic culture (usually Sc. thermophilusand Lb. delbrueckii subsp. bulgaricus) and co-agulated by calf rennet. The curds are cooked to43-450C, transferred to molds lined withcheesecloth and subjected to a low pressure,which is increased slowly. Upon removal fromthe molds, the cheeses are dried overnight andbrine-salted. After brining, dry salt is rubbedonto the surface of the cheese over the next fewdays to give a final salt content close to 4%. Dur-ing this time, the cheeses are washed with abrine-soaked cloth to control microbial growthon the surface and ripened for about 3 months.According to Robinson (1995), Kefalotiri has ahard texture and a strong, salty flavor.

Graviera is a relatively recently developedGreek variety made principally from ewe milk.It is acidified by a mixed mesophilic culture(1%) containing Lc. lactis subsp. lactis or Lc.lactis subsp. cremoris and a smaller amount(0.1%) of a thermophilic culture containing Sc.thermophilus and Lb. helveticus. The curds arecooked first to about 5O0C, and after whey drain-age the curds are molded and pressed at an in-creasing pressure. The cheeses are then salted byfrequent application of dry salt to the surface for2-3 weeks and ripened for 3-4 months. In somefactories, the early stages of dry-salting may bereplaced by brining.

The principal hard cheese produced in Egyptis Ras, the production process for which is simi-lar in many respects to that for Kefalotiri. It ismade from cow milk standardized to 3% fat. Thecurds are cooked to 450C, salted at a level of 1%after whey drainage, molded, and pressed in amanner similar to Kefalotiri. The cheeses arebrined for 24 hr and rubbed with a small quantityof dry salt daily for several weeks. During thisperiod, the cheeses are also washed with brine.

Manchego cheese, probably the most impor-tant Spanish variety, is made from ewe milk, al-though generally similar cheeses (without PDOstatus) are manufactured from milks of otherspecies. Two types of Manchego are produced:artisanal (made from raw milk without cultureaddition) and commercial (made from pasteur-ized milk inoculated with a mesophilic starter;see Chapter 5). The milk is coagulated with stan-dard calf rennet. The curds are cooked to about380C, transferred to molds, and pressed for 12-16 hr. Manchego has characteristic side mark-ings made by binding the cheeses in basketworkwrappings for about 30 min after pressing. Thecheeses are then brine-salted and matured for atleast 2 months at 10-150C and 85% ERH.

Idiazabal cheese is produced in the Basque re-gion of northern Spain from raw ewe milk co-agulated at 380C. The coagulum is allowed tocool to 250C and then broken, and the curds areladled into molds. The cheeses are salted bybrining or by the application of dry salt and ma-tured in caves for 2 months. They are then

Page 421: Cheese Science

smoked in beechwood kilns and further maturedfor up to 1 year. Roncal is made in northernSpain from raw ewe milk and acidified by theindigenous flora of the milk. The curd-wheymixture is cooked to about 37-4O0C, and thecurds are then allowed to settle to the bottom ofthe vat. The whey is removed slowly, and thecurds are pressed against the sides of the vat,molded, and pressed before being dry-salted.Roncal is smoked and ripened at 6-80C and100% ERH for 45-50 days. La Serena is a hardcheese made in western Spain from ewe milk.Traditionally, raw milk is used, although pas-teurized milk is used in large-scale production.The milk is coagulated with rennet extractedfrom the cardoon thistle (Cynara cardunculus).

Other hard cheese varieties include Leidenand Friesian Clove cheeses from the Nether-lands, which are characterized by the addition ofcumin seeds (Leiden) or cloves and cumin seeds(Freisian Clove).

Semi-Hard Varieties

The description of a cheese as semi-hard isarbitary. The semi-hard group of cheeses is thusheterogeneous, and the distinction between thisand other groups of cheeses (e.g., hard cheeses,smear-ripened varieties, and Pasta filatacheeses) may not be clear. Semi-hard cheesesinclude Colby and Monterey (stirred-curd Ched-dar-type cheeses), a number of British Territo-rial varieties (Caerphilly, Lancashire, andWensleydale), and cheeses such as Bryndza(Slovakia) and Mahon and Majorero (Spain).

Caerphilly, which originated in Wales, is acrumbly acid cheese. It is made from pasteurizedcow milk using calf rennet and a mesophilicstarter. The curds are cooked to 32-340C andheld at this temperature for about 1 hr. The wheyis drawn off and the curds are collected at thebottom of the vat, where rapid acid productionoccurs. Some dry salt (1%) is added to the curdsbefore molding and pressing overnight. Thepressed curds are then brine-salted for 24 hr andpackaged. Caerphilly matures rapidly and isready for sale after 10-14 days. Lancashire, an-

other British Territorial variety, is made fromcow milk using rennet and a mesophilic starter.The curds and whey are not cooked, but afterdraining, the curds are cheddared and held over-night, during which time extensive acid produc-tion occurs. The next day, fresh curds are mixedwith the acidified curds, and the mixture ismilled to ensure homogeneity. The curds aredry-salted, placed in molds overnight at roomtemperature, and pressed for 3 days. Lancashireis ripened at 13-180C for 3-12 weeks.

Wensleydale cheese, which originated inYorkshire, England, is made from cow milk in-oculated with a mesophilic starter. The curds arecooked at 32-340C. The whey is drained off, andthe curd mat is broken into pieces to assist wheydrainage. The pieces of curd are dry-salted andmolded, held overnight at about 210C withoutpressing, and then pressed lightly for 5 hr.Wensleydale, which is matured for about 1month, has mechanical openings and a mild,acidic taste.

Stirring Cheddar-type cheese curd inhibits thedevelopment of curd structure and results in acheese with a higher moisture content and thus asofter texture. Two stirred-curd variants ofCheddar cheese are recognized: Colby andMonterey. The manufacture of Colby, whichoriginated in the United States, follows a proto-col similar to Cheddar until after cooking, whensome whey is removed and replaced by cold wa-ter. The curds and whey are stirred, and most ofthe whey is removed. The curds and remainingwhey are stirred vigorously. The remainingwhey is then drained off, and the curds arestirred further. Stirring prevents the develop-ment of an extensive structure while the curdsare in the vat. Salt is added to the curds, whichare molded and pressed. Colby is ripened for 2-3months at 3-40C. Monterey (Monterey Jack)cheese was first made in California and is simi-lar to Colby. The whey is removed and the curdsare left on the bottom of the vat, with occasionalstirring until the pH reaches 5.3. The curds aredry-salted and pressed lightly overnight. Thecheeses are allowed to form a rind before wax-ing or packaging in films. Monterey, which is

Page 422: Cheese Science

ripened for 5-7 weeks, has many mechanicalopenings.

Bryndza is made from sheep milk coagulatedwith rennet (sometimes with significant lipaseactivity) and acidified by the indigenous micro-flora of the milk or by a mesophilic starter. Thecurds are allowed to settle to the bottom of thevat, most of the whey is removed, and the curdsare consolidated into lumps by hand. The lumpsof curd are placed into cloth bags and stored for3 days while sufficient acidity develops. At thisstage, the cheese is known as Hrudka and maybe sold locally. However, most Hrudka is trans-ported to a central factory, where it is brokeninto pieces, salted, and passed between graniterollers to produce a smooth paste, which isplaced in polyethylene-lined wooden tubs andmatured.

Mahon is produced in the Balearic island ofMinorca from raw cow milk acidified by its in-digenous microflora. Mahon is brine-salted andripened at 180C for about 2 months. Althoughthe texture of Mahon is semi-hard, its moisturecontent is reported to be roughly 32%. Majorejois made from goat milk on Fuerteventura Island,one of the Canary Islands. In commercial prac-tice, milk is acidified by a mesophilic starter andcoagulated using rennet, although artisanalcheesemakers rely on the indigenous microfloraof the milk and use rennet paste as a coagulant.The curds are molded in braided palm leaves(which impart to its surface a characteristic pat-tern), pressed lightly, and dry-salted. Majorejodevelops a strong flavor during ripening.

Cheeses with Eyes

Mechanical openings, resulting from the in-complete fusion of curd pieces, are common inmany cheese varieties and may be considereddesirable (e.g., Monterey) or a defect (e.g.,Cheddar). However, some internal bacteriallyripened varieties are characterized by the devel-opment of eyes caused by the entrapment withinthe curd of gas produced by bacterial metabo-lism. The development of eyes in cheese is gov-erned by the rate of gas production by bacteriaand the ability of the curd to retain the gas. There

are two main families of cheese with eyes:Dutch types (Edam, Gouda [Figure 17-7], andrelated varieties), which have small eyes, andSwiss types, which are characterized by largeeyes. In the case of Edam and Gouda, CO2 isproduced from citrate by the DL culture (seeChapter 5), while in Swiss varieties, CO2 is pro-duced by Propionibacteriumfreudenreichii spp.shermanii from lactate during ripening (seeChapters 10 and 11).

Gouda (Figure 17-8) originated in the Nether-lands but is now produced worldwide from pas-teurized cow milk coagulated using calf rennetand acidified by a mesophilic DL starter. Nitratemay be added to the milk to suppress the growthof Clostridium spp., which produce gas (H2 andCO2) from lactate during ripening, causing a de-fect known as "late gas blowing." Butanoic acidis also produced from lactate, and it causes off-flavors. The coagulum is cut and the curd-wheymixture is stirred for 20-30 min before a portion(about 30%) of the whey is drained off and re-placed by hot water, which raises the tempera-ture to 36-380C. This washing step removessome of the lactose and consequently reducesthe development of acidity after the curds aremolded. The curds are cooked at this tempera-ture and then allowed to settle to the bottom ofthe vat where they are pressed under the whey.The bed of curd is cut into blocks, which aretransferred to molds (wheel-shaped or rectangu-lar, producing a cheese of 4-20 kg) and pressedfor 5-6 hr. Gouda is brine-salted, coated withyellow wax (traditionally), and ripened at 150Cfor 2-3 months or longer (up to 2 years for extramature cheese).

Edam, a Dutch variety similar to Gouda butwith a distinctive spherical shape, is coated withred wax and is made from semi-skimmed milk(~ 2.5% fat). A few small eyes develop in Edam,which may be sold after ripening for 6-30 weeks.Maribo is a similar variety produced in Denmarkfrom cow milk, and, in addition to eyes caused bya mesophilic DL starter, it has numerous me-chanical openings. Other Dutch-type cheeses in-clude Danbo (Denmark), Colonia and Hollanda(Argentina), and Svecia(ost) (Sweden).

Page 423: Cheese Science

Swiss-type cheeses are characterized by largeeyes produced by P. freudenreichii subsp. sher-manii, which metabolizes lactate to propionate,acetate, and CO2 (see Chapters 10 and 11). Pro-pionibacteria do not grow in the milk duringcheesemaking but grow in the cheese duringmaturation, when it is transferred to a hot room(« 20-220C). The curd of these cheeses is quiterubbery and is able to trap the CO2 (which mi-grates through the curd until it reaches a fissureor weakness, at which place an eye develops).The texture of these cheeses is influenced by ahigh cook temperature (~ 550C), which inacti-vates most of the coagulant, and a high pH atdraining (which leads to a high concentration ofcalcium in the curd). Swiss-type cheeses are tra-ditionally made as large wheels and are brine-salted. The relatively slow diffusion of salt in a

large cheese allows the salt-sensitive propioni-bacteria time to grow.

Emmental (Figure 17-9) is a typical Swisscheese variety that is now made worldwide. Tra-ditional Emmental is made from raw cow milkthat is acidified with a mixed thermophilicstarter consisting of Sc. thermophilus and a Lac-tobacillus species (Figure 17-10). Lb. helveticuswas used traditionally but Lb. delbrueckii subsp.lactis is now common also. Propionibacteriamay be added to the milk or may contaminatemilk from the environment. The milk, at 3O0C, iscoagulated using calf rennet, and the coagulumis cut into small pieces and cooked to about 550Cuntil the curd grains are of the desired firmness.The curds and whey are transferred to molds,where the whey is separated. The molds are suf-ficiently large to give a wheel of cheese weigh-

Figure 17-7 Gouda cheese.

Page 424: Cheese Science

ing up to 100 kg and as much as 1 m in diameter.The size of the Emmental wheel is significantbecause it determines the rate of cooling of thecurd (and thus the activity of the starter; seeChapter 10), determines the diffusion of salt

throughout the cheese mass, and helps to trapgas within the cheese. Over the next 1-2 days,the wheels are pressed and turned frequently.During this time, the curds cool and acid produc-tion by the starter organisms (which were dor-

Figure 17-8 Manufacturing protocol for Gouda cheese.

Gouda Cheese

2-3 months at ca. 150C

Ripen

Wax

3-5 days

Brine salt

5-6 hours

Curds placed in moulds

Press under whey

Stir for ca. 30 min

ca. 360C

Hot waterWhey

Coagulum

Rennet(Mesophilic) DL starterNitrate3O0C

Pasteurized Cow's Milk

Page 425: Cheese Science

mant during cooking) recommences. Completefermentation of lactose and its constituentmonosaccharides in Emmental takes about 24hr. After pressing, the cheese wheels are brine-salted, stored in a cool room (10-150C, 90%ERH) for 10-14 days, and brushed, dry-salted,and turned daily until a smooth rind develops.The cheeses are then transferred to a hot room(20-220C, 80-83% ERH) and held there for 3-6weeks, until adequate eye formation has oc-curred (which is indicated by a drum-like soundwhen a cheese is tapped with a cheese trier). Thecheeses are then matured at about 70C for a fur-ther 1 or 2 months. Rindless Emmental in blockform is produced by a protocol generally similarto that for Emmental, but milk with a lower fatcontent and a lower cooking temperature areused. The cheese is wrapped in a plastic film,and therefore no rind develops during matura-tion.

Gruyere is another popular Swiss-typecheese. It differs from Emmental in being

smaller and having a somewhat stronger flavorand fewer eyes, and it is characterized by the de-velopment of a surface flora (similar to thatwhich develops on smear-ripened varieties, Sec-tion 17.2.3). The surface flora is encouraged byripening for 2-3 weeks at 1O0C and then for 2-3months at 15-180C and 90-95% ERH, duringwhich time the cheeses are rubbed with a brine-soaked cloth. Further ripening at 12-150C is re-quired before sale at 8-12 months of age.

Similar varieties include Raclette (which ismanufactured from raw milk and is acidified bythe indigenous milk microflora) and Gruyere deComte, which is produced in eastern Francefrom raw bovine milk. Beaufort is a French vari-ety similar to but larger than (~ 45 kg) Gruyere,but only Lb. helveticus is used as starter. Ap-penzeller, which originated in Switzerland, is asmall cheese (~ 30 cm in diameter) with a softtexture. It undergoes propionic acid fermenta-tion, resulting in the development of a few eyes.The curds are cooked at 43^50C. Appenzeller

Figure 17-9 Emmental cheese.

Page 426: Cheese Science

Figure 17-10 Manufacturing protocol for Emmental, a Swiss-type cheese.

Cowfs Milk

30-320CRennet,Thermophilic starter,iPropionibacteria

Coagulum

Cut coagulum,Cook to ca. 550C

Curds-whey pitched into moulds

Whey

Pressing

Dry cheese surfaceSalt applied to the surface

1-2 days

Brine salting

ca 2 days, 8-1O0C

Cool storageDry rind forms

10-14 day s10-150C, 90% ERH

Hot room20-240C, 80-83% erh

3 weeks-2 monthsEye development

Cool storage

1-2 months70C

Emmental Cheese

Page 427: Cheese Science

is immersed in cider or spiced wine or rubbedwith a mixture of salt and spices during ripening,which impart a distinctive flavor to the cheese.Maasdammer, a variety developed recently inthe Netherlands, is characterized by the use of amesophilic starter and extensive propionic acidfermentation, which causes large eyes and givesa domed appearance to the cheese wheels. Jarls-berg is a Swiss-type cheese produced in Norwayusing a mesophilic starter.

High-Salt Varieties

White brined cheeses originated around theeastern Mediterranean. According to Robinson

(1995), the original characteristic features of themanufacture of these cheeses were the use ofsheep milk (which yields a very white curd), ahigh ambient temperature, and storage in brine(leading to a high salt content) for purposes ofpreservation. The principal white brined variet-ies today are Feta (Figure 17-11), Telemes(Greece), and Domiati (Egypt). White brinedcheeses are now made worldwide and are majorindustrial products.

Feta cheese is made from sheep milk or mixedsheep and goat milk. Strenuous efforts on thepart of the Greek government have resulted inPDO status for Feta, although similar cheeses

Figure 17-11 Feta cheese.

Page 428: Cheese Science

are manufactured worldwide. Milk for most Fetais pasteurized and standardized to a casein:fatratio of about 0.7-0.8:1 (Figure 17-12). A ther-mophilic or mesophilic starter culture is added toensure rapid acidification. The rennet used is of-

ten a mixture of standard calf rennet and a localrennet extract with some lipase activity. CaCl2

may be added to the milk before renneting. Therennet-induced coagulum is cut (into 2-3 cmcubes), and the soft curds are ladled directly into

Sheep's milkSheep's/Goat's milk

(-6% fat)

RennetCaCl2

Starter

Coagulum

Cut

T

MouldMoulds inverted

Whey

Curd cut into blocks

Dry salt

Curds placed in container

ca. 1 days

Container filled with brine

14-160C

ca. pH 4.5

3-40C>2 months

Feta Cheese

Figure 17-12 Manufacturing protocol for Greek Feta cheese.

Page 429: Cheese Science

molds. The curd-whey mixture is not cooked.Whey drainage occurs in the molds, which areinverted after 2-3 hr. The curd mass is then firmenough to be removed from the molds and is cutinto blocks, which are dry-salted (or brined) be-fore being transferred to a barrel or other con-tainer. The container is filled with brine (-14%NaCl) after around 7 days and held at 14-160Cuntil the pH of the cheese has decreased to aboutpH 4.5, at which point the cheeses are stored at3-^0C for at least 2 months.

Feta-type cheese is now also manufacturedindustrially from cow milk concentrated by ul-trafiltration (concentrated approximately five-fold using membranes with a cutoff of 10-20kDa). Less rennet is used in the manufacture ofFeta-type cheese from ultrafiltration retentate,and the yield is higher than for Feta made fromunconcentrated milk owing to the incorporationof whey proteins into the curd (see also Section17.6).

Telemes cheese is made by a protocol similarto that for Feta, except that some pressure is ap-plied to the curds in the molds to aid in the ex-pulsion of whey and the curd blocks are brinedafter removal from the molds.

Domiati is made from cow or buffalo milk ora mixture of both containing 2%, 4%, or 8% fat(giving reduced- or full-fat cheese). Domiatimay be made from pasteurized milk to whichNaCl is added to a level of 5-15%. Alterna-tively, about one-third of the milk may be heatedto around 8O0C and salt added to the remainder.At the level used, NaCl has a strong antibacterialeffect, and halotolerant lactobacilli are used asstarter. Domiati may be consumed fresh or rip-ened in brine for a number of months.

Halloumi is a brined cheese made in Cyprusfrom sheep milk. The curds are cooked to 38-420C during manufacture and the cheesespressed. Blocks of curd are scalded by immer-sion in hot whey (90-920C) for 30 min but arenot stretched. The cheeses are dry-salted andconsumed fresh or after storage in brine.

Other white pickled cheeses include Lightvan(Iran), Beda (Egypt), and Bulgarian white brinedcheeses.

Pasta Filata Varieties

Pasta filata cheeses are semi-hard varieties,the curds for which are heated to 550C and aboveand mechanically stretched during manufacture.Stretching causes the curds to become fibrousand malleable. Most Pasta filata cheeses origi-nated in the Mediterranean region.

By far the most important member of thisgroup is Mozzarella, which originated in south-ern Italy and was originally manufactured frombuffalo milk. Mozzarella di bufala (Figure17-13) is hand-molded into round pieces (100-300 g) during manufacture. This cheese is stillmanufactured in Italy, but the type of Mozza-rella now widely manufactured around the worldis made from pasteurized, partly skimmed cowmilk (Figure 17-14) and is often referred to aspizza cheese or, in the United States, low-mois-ture, part-skimmed Mozzarella. This type ofMozzarella has a higher salt concentration (1.5-1.7%) than Mozzarella di bufala (<1.0%). Theproduction of low-moisture, part-skimmedMozzarella cheese has increased greatly in re-cent years as a result of the increased popularityof pizza pie. Some is consumed as a table cheeseor as a component of salads.

The manufacturing process for Mozzarellafor use as pizza topping (Figure 17-14) involvesstandardizing pasteurized cow milk to around1.8% fat. A higher fat content (« 3.6%) is usedfor Mozzarella intended to be consumed as atable cheese. A thermophilic starter (1-2%)containing a combination of Lactobacillus spp.and Sc. thermophilus is used in the manufactureof pizza cheese. The Lactobacillus is oftenomitted when Mozzarella is intended as a tablecheese, since the rate of acidification need notbe as fast as in pizza cheese. Proteolytic en-zymes of the Lactobacillus may make a minorcontribution to the functionality of the finalproduct by causing slight hydrolysis of thecaseins. The milk is renneted after some acidityhas developed, and the coagulum is cut andcooked to around 410C. The whey is then usu-ally drained off, and texture is developed in thecurds (usually by cheddaring) until the pH drops

Page 430: Cheese Science

to around pH 5.1-5.3. Because the productionand the treatment of the curds are quite similarto those used for Cheddar up to this stage, Ched-dar plants may be easily modified to produceMozzarella by altering the starter and tempera-ture profile used and by the use of an appropri-ate stretcher. The next stages in Mozzarellamanufacture are stretching and kneading, whichare characteristic of pasta filata varieties. Thecheddared curds are placed in hot water(~ 7O0C) and kneaded, stretched, and folded un-til the desired texture has been developed. The

curds for pizza cheese are stretched more exten-sively than those for table Mozzarella. Theformer may also be salted during the stretchingand forming stages. The hot, plastic curds aremolded (usually into rectangular blocks) andcooled quickly in cold water or brine, and if saltwas not added during the cooking and stretchingprocess, the cheeses are then brine-salted. Moz-zarella is usually consumed within a few weeksof manufacture. Extensive ripening is undesir-able, since the functional properties of thecheese deteriorate.

Figure 17-13 Mozzarella di bufala.

Page 431: Cheese Science

So-called string cheese is produced fromMozzarella, Cheddar, or similar cheese curd bycooking and extruding the plastic curd as longrods (1-2 cm in diameter) and then brining therods (brining occurs rapidly because of the smallcross-sectional area of the rods). The rods arethen cut into convenient lengths and packaged.Strings of cheese may be torn from these rods,and this novelty feature is the major selling pointfor string cheese, whose target market consistsof young children.

Kashkaval is a stretched-curd variety from theBalkans and was traditionally made from sheep

milk, although cow milk Kashkaval is now com-mon. The cheese is usually brine-salted, butsome is dry-salted or stored in brine. Typically,the cheese is matured for 2-3 months beforeconsumption. Kasseri, a Greek cheese similar toKashkaval, is made from sheep milk or a mix-ture of sheep and goat milk. Provolone (Figure17-15), which is characteristically pear shaped,originated in southern Italy, where it is madefrom cow milk. Rennet paste may be used in itsmanufacture, giving the resulting cheese (Provo-lone piccanti) a stronger flavor than normal Pro-volone (Provolone dulce), which is manufac-

Figure 17-14 Manufacturing protocol for low-moisture, part-skimmed Mozzarella (pizza) cheese.

Mozzarella Cheese(Low Moisture, Part-Skimmed)

Short ripening, <1 month

Brine salting

Place plastic curds in mouldscool until solid

Heat, stretch at ~65°C

pH 5.1-5.3

Condition (cheddar) curd

Whey

Cut coagulum,Cook to 410C

Coagulum

Rennet Thermophilic starter

Pasteurized Cow's Milk(Part-skimmed, -1.8 % fat)

Page 432: Cheese Science

tured using rennet extract. Provolone is ripenedfor 2-6 months. Caciovallo is a hard Italiancheese manufactured from cow milk by a pro-cess somewhat similar to that used for Provo-lone. The curds are stretched in hot water andbrine-salted. Caciovallo is ripened for 3-4months—or longer (>12 months) if the cheese isto be grated. Ostiepok, a stretched-curd cheesefrom central Europe (Czech Republic andSlovakia), is made from sheep milk (althoughcow milk is used sometimes), brine-salted, andsmoked heavily.

17.2.2 Mold-Ripened Cheeses

Cheese varieties on which molds grow duringripening fall into two broad categories: surfacemold-ripened cheeses (e.g., Brie and Camem-

bert; Figure 17-16), in which the mold grows asa mat on the surface, and blue-veined varieties(Figure 17-17), which are characterized by thegrowth of P. roqueforti in fissures throughoutthe cheese. Although these two groups havemold growth in common, the methods used fortheir manufacture and the flavor and texture ofthe mature cheese are quite different. Hybridmold-ripened cheeses are also produced (e.g.,Cambazola); these have a white mold growth onthe surface and blue mold in the interior.

Surface Mold-Ripened Varieties

Surface mold-ripened cheeses are generallysoft varieties characterized by the growth of thewhite mold Penicillium camemberti on the sur-face of the cheese. The surface flora is oftenmore complex, particularly in cheeses made

Figure 17-15 Provolone cheese.

Page 433: Cheese Science

Figure 17-17 Blue-veined cheese.

Figure 17-16 Camembert cheese.

Page 434: Cheese Science

from raw milk by traditional technology. Thesecurds are acidified to around pH 4.6 duringmanufacture using a mesophilic starter. Lacticacid produced by the starter is metabolized bythe mold, which also produces ammonia, andtherefore the pH of the surface layer of thecheese increases to around 7.0. If present, yeastalso catabolize lactic acid. An important conse-quence of the increase in the pH of these cheesesis that there is considerable migration of calciumphosphate to the surface layer (which contains ~80% of the calcium and ~ 55% of the phospho-rus of the mature cheese). As in all cheeses con-taining active rennet, the coagulant plays a rolein the development of texture. However, in sur-face mold-ripened cheeses, the role of rennet isrelatively minor. These cheeses soften from thesurface toward the center during ripening, ow-ing, not to proteolysis by mold enzymes, whichdiffuse only a few millimeters into the cheese,but to the establishment of pH and calcium phos-phate gradients. Softening in these cheeses is of-ten quite extensive, leading to a spreadable, al-most liquid consistency.

Many surface mold-ripened varieties origi-nated in France. They are usually manufacturedfrom cow milk acidified by a mesophilic starter.The first microorganisms to become establishedon the surface are yeasts, including Debar-omyces spp. and Kluyveromyces spp. Geo-trichum candidum also becomes established atthis time, although its growth may be limited ifthe level of salt is high. P. camemberti is ob-served after 6-7 days of ripening and forms thecharacteristic mat on the cheese surface. Oncethe surface of the cheese has been neutralized(after 15-20 days), aerobic bacteria (particularlymicrococci and coryneforms), which are inhib-ited by the low initial pH, begin to grow.

Camembert, the most important surfacemold-ripened variety, originated in Normandy,France, in the 18th century. It is a small cheese(-10 cm diameter and 200-250 g) manufacturedfrom cow milk (Figure 17-18). Raw milk is usedtraditionally but industrial Camembert is nowproduced from pasteurized milk. A mesophilicstarter (~ 0.1%) is used, and when the pH of the

milk has fallen to about 6.1, rennet is added. Tra-ditionally, the coagulum is not cut but is ladledinto molds, where drainage occurs. To facilitatemanufacture, the coagulum for industrialCamembert is first cut into large cubes and thentransferred to molds without cooking. Tradition-ally, the cheeses are dry-salted and P. camem-berti spores are sprayed on the surface, althoughit is now industrial practice to inoculate the milkwith mold spores and to brine-salt the cheeses.The surface of the cheese is allowed to dry atambient temperature in a well-ventilated room,after which the cheeses are transferred to a store-room at about 120C for 10-12 days for mold de-velopment. The cheeses are then packaged inwaxed paper and placed in wooden or cardboardboxes prior to final ripening at 70C for 7-10days.

Brie is a flat, cylindrical surface mold-rip-ened cheese with a larger diameter than Cam-embert, which it resembles closely in flavor, tex-ture, and manufacturing protocol. Carre de 1'Estis a square surface mold-ripened cheese thatoriginated in eastern France. Neufchatel origi-nated near Rouen, France, where it is still pro-duced, mainly on farms. Raw cow milk is inocu-lated with a small quantity of an artisanal starterand renneted. The coagulum that forms over-night is transferred to muslin bags throughwhich the whey is drained. After most of thewhey has drained, the curds are removed fromthe bags, mixed, salted, and placed in molds.When the curds are firm enough to remove fromthe molds, their surfaces are salted, dusted withP. camemberti spores, and ripened for 2-3weeks. In addition to the above-mentioned vari-eties, St. Marcellin (which was manufacturedoriginally from goat milk but is now made fromcow or sheep milk) and a number of minor goatmilk cheeses also develop a surface mold growthduring ripening. The microflora of these minorvarieties is often uncontrolled.

Blue-Veined Cheeses

Blue-veined cheeses are characterized by thegrowth of P. roqueforti in fissures throughoutthe cheese (Figure 17-17). These cheeses usu-

Page 435: Cheese Science

ally have a soft texture and a flavor dominatedby alkan-2-ones (methyl ketones), which areproduced from free fatty acids by the mold viathe [i-oxidation pathway (see Chapter 11). ThepH of blue mold cheeses increases during ripen-ing, from 4.6-5.0 after molding to 6.0-6.5 whenmature.

Since P. roqueforti requires O2 for growth,the manufacture of Blue cheese is dominated bythe need to provide a suitable environment for

its growth. This is achieved by encouraginglarge mechanical openings in the cheese (by notpressing the cheese when molded) and by pierc-ing the cheese to allow air into its center and toallow CO2 produced by the mold to escape. Li-polysis is often encouraged in Blue cheeses byseparation of the raw cheese milk and homog-enization of the cream, which encourages theaction of the indigenous milk lipase. The levelof starter used (usually a DL starter) is normally

Raw1 or Pasteurized2 Cow's Milk

DL starter,Rennet,Spores of Penicillium camemberti23O0C

CurdSaddled into moulds without cutting, or

2cut into large cubes

Whey drainage in moulds overnight

Whey

Dry1 or brine2 salting(1SpOrCS of P. camemberti added to the dry salt)

ambient temperature

Surface of cheese dry

120C10-12 days

Mould development

Packaging70C7-10 day s

traditional protocolIndustrial protocolCamembert Cheese

Figure 17-18 Manufacturing protocol for Camembert, a surface mold-ripened cheese.

Page 436: Cheese Science

very low, and in some cases the curds are acidi-fied by the indigenous microflora, since thecurds are usually held for an extended period inmolds or special drainers, during which timeacidity develops. Spores of P. roqueforti areadded either to the milk or to the curds duringmanufacture. The curds for Blue cheese arecooked at a low temperature and transferred todrainers or molds, where the whey is separatedfrom the curds. Blue cheeses are dry-salted, ei-ther by the application of salt to the surface ofthe cheese or by milling the curds after wheydrainage and mixing with salt prior to molding.The cheeses are ripened under conditions oftemperature (usually 10-120C) and relative hu-midity that favor mold growth. They are piercedduring ripening to facilitate uniform moldgrowth, they are turned, and their surfaces arecleaned regularly.

Roquefort is a Blue cheese with PDO statusmanufactured from raw ewe milk. Cheeses mustbe ripened in caves in a defined area of south-eastern France. The absence of carotenoids fromewe milk results in a very white cheese thathighlights the contrast in color between the curdand the mold. The manufacturing protocol forRoquefort is shown in Figure 17-19. Usually, nostarter is added to the milk, which is acidified byits indigenous microflora. In addition to homo-fermentative lactococci, this microflora containsleuconostocs and other heterofermentative lacticacid bacteria (LAB) that produce CO2 as a by-product. The CO2 causes small openings in thecurd that favor the growth of the mold. Lambrennet is added to the cheese milk, and coagula-tion is complete in about 2 hr, when the coagu-lum is cut. The curds are not cooked but aremixed with spores of P. roqueforti and placed inperforated metal molds for whey drainage. Thewhey is allowed to drain for 4-5 days, duringwhich the cheeses are inverted periodically andacidity develops. The cheeses are then removedfrom the molds and dry-salted over a period of 1week, after which they are pierced and placed inlimestone caves that have the correct tempera-ture and relative humidity to encourage moldgrowth. The cheeses are matured for 3-5

months, during which their surface is cleaned toremove adventitious molds or smear-formingbacteria.

Bleu d'Auvergne is a Blue cheese manufac-tured in France from cow milk. The SpanishBlue cheese, Cabrales, is usually made from cowmilk and is ripened in local caves. Cabralescheeses are often covered with sycamore (Acerpseudoplatanus) leaves to retain humidityaround the cheese. The cheeses are not inocu-lated with P. roqueforti spores but become con-taminated with mold spores from the environ-ment during ripening. Thus, the degree of molddevelopment in this variety can be variable.

Danablu (Danish blue) is perhaps the mostcommercially important Blue cheese. It is manu-factured from cow milk, the cream from which ishomogenized (to encourage lipolysis) prior topasteurization. Pasteurized skim milk and creamare mixed and inoculated with starter and moldspores. The manufacturing protocol for Danabluis broadly similar to that for Roquefort, althoughchlorophyll is sometimes added to mask the yel-low color of the carotenoids in cow milk, thusgiving a whiter cheese. Danablu is ripened undercontrolled temperature (14-180C) and relativehumidity (90-95%) for up to 3 months. Edel-pilkase is a Blue cheese produced in Austria andGermany. Mycella, a Blue cheese larger thanDanablu produced in Denmark, is characterizedby a yellow-white color and intense moldgrowth. Gorgonzola, the traditional Blue cheeseof Italy, is also manufactured from cow milk.The traditional protocol for the manufacture ofGorgonzola involves the separate production ofcurds from evening and morning milk. P.roqueforti spores are added to both batches ofcurd. The evening curds are stored overnight incloth bags and used for cheesemaking the fol-lowing morning. The morning curds are placedon the bottom and around the sides of the moldwhile still warm, and the cool evening curds areplaced in the center. The top of the mold is thenfilled with warm morning curds. This layeringencourages the development of mechanicalopenings in the cheese and yields a cheese with asmooth, firm surface.

Page 437: Cheese Science

Stilton is a Blue cheese that legally can onlybe produced in the English counties of Leicester-shire, Derbyshire, and Nottinghamshire frompasteurized cow milk. A mesophilic DL starter(< 0.04%) and P. roqueforti spores are added tothe milk. After renneting, the curds are allowedto settle on the bottom of the vat, and the whey iswithdrawn slowly over the next 12-18 hr. Thecurd mass is cut to facilitate drainage, and thecurd pieces are milled, dry-salted, and placed inmolds. Whey drainage continues for about 7days and is facilitated by frequent turning. Dur-ing this time, the cheeses are kept warm (26-3O0C, 90% relative humidity) so that the starterbacteria can produce sufficient acid in the cheese

curd. The cheeses are then placed in a coolerroom (13-150C, 85-90% relative humidity) for6-7 weeks, during which time the cheeses cooland a rind develops on the surface. The cheesesare pierced, and after sufficient mold growth hasoccurred (2-3 weeks), they are moved to a coldstoreroom at 50C.

17.2.3 Surface Smear-Ripened Cheeses

Cheeses ripened with a mixed surface micro-flora perhaps constitute the most hetereo-geneous group of rennet-coagulated cheeses. Al-though most varieties in this group are soft orsemi-hard, a surface flora may also develop on

Figure 17-19 Manufacturing protocol for Roquefort, an internal mold-ripened (Blue) cheese.

Roquefort Cheese

3-5 months5-1O0C, 95% erhSurface cleaned every 2-3 weeks

Transfer to limestone caves

Pierce cheese

Dry salting for 1 week

Whey

Natural drainage for 4-5 days

Whey

Perforated metal moulds

Spores of Penicillium roqueforti

Coagulum

Rennet3O0C

Raw Sheep's Milk

Page 438: Cheese Science

hard cheeses such as Gruyere. However, in thelatter case, the contribution of the surface florato cheese ripening is relatively minor.

The distinguishing feature of surface-ripenedcheeses is the development of a mixed microfloraon the cheese surface, forming a red-orange smear. These cheeses are manufacturedusing a mesophilic starter (most varieties) or athermophilic starter (Gruyere and similarcheeses) and are usually brine-salted. Manufac-turing protocols usually result in curds with a highmoisture content. After manufacture, a range ofsalt-tolerant yeasts (Kluyveromyces, Debaro-myces, Saccharomyces, Candida, Pichia,Hansenula, and Rhodotorula), together withGeotrichum candidum, become established onthe cheese surface, where they metabolize lactateto CO2 and H2O. This change in environment fa-vors the growth of other microorganisms. Themicroflora is complex and consists of Gram-posi-tive bacteria, including Micrococcus, Staphylo-coccus, and various coryneform bacteria (whichare responsible for the color of the smear). Onecomponent of the surface smear is the coryneformBrevibacterium linens, which is widely used insmear inocula, although recent research has sug-gested that it may be only a relatively minor com-ponent of the complex surface flora.

The surface microflora may reach 1011 cfu/cm2 and thus the enzymatic activities of thesmear microorganisms contribute significantlyto the flavor of the cheese. Since enzymes do notdiffuse through cheese curd, patterns of pro-teolysis in smear-ripened cheeses are similar tothose in internal bacterially ripened varieties ex-cept at the cheese surface. However, products ofthe metabolic activities of the smear diffuse intothe cheese and influence its flavor. Soft surface-ripened cheeses usually have a strong flavorwhereas semi-hard varieties have a milderflavor.

Soft surface-ripened cheeses mature quiterapidly. The rate at which they ripen is governedby the size of the cheese, its moisture content,the ripening conditions, and the composition ofthe surface microflora. The high moisture con-tent of these cheeses accrues from cutting the

coagulum into large pieces and cooking thecurds at a low temperature. Much whey is re-tained, and therefore the curds have a relativelyhigh lactose content, favoring the growth of thestarter, which acidifies the cheese to a low pH(~ 5). The ratio of surface area to volume (andthus the size and shape of the cheese) is veryimportant in surface-ripened varieties. Thesmaller the cheese (and thus the greater this ra-tio), the greater the influence of the surface floraon the flavor of the cheese. The relative humid-i t y(> 95%) and temperature (12-2O0C) of ripeningrooms are controlled so as to favor the growth ofthe surface smear. In some cases, cheeses areheld in an environment with a lower relative hu-midity to encourage the development of a rind.

The smear develops initially as a series ofcolonies on the surface of the cheese. During rip-ening, the surface of these cheeses may be "mas-saged" (washed) with a brine solution, whichdistributes the microorganisms evenly over thesurface of the cheese and results in the develop-ment of a uniform smear. Although it is commonpractice to inoculate cheeses with Br. linens, theprincipal source of the surface microflora is thecheesemaking environment. To encourage de-velopment of the microflora, young cheeses areoften smeared with the same brine used previ-ously to smear older, high-quality cheeses.

The ripening period for smear-ripenedcheeses depends on the desired flavor intensitybut is usually relatively short. For some cheeses(e.g., Brick), the smear is washed from the sur-face and the cheeses are coated with protectivematerial before being transferred to a ripeningroom at a lower temperature (~ 1O0C) for furthermaturation. The degree to which the smear ispermitted to develop varies greatly among vari-eties. In some cheeses, smear development isdesired principally to color the cheese surfaceand is therefore very limited. These cheeses ma-ture in a similar manner to internal bacteriallyripened varieties. In some cases, the surface ispainted with dye to mimic smear color.

The majority of bacterial surface-ripenedcheeses originated in northern Europe, but many

Page 439: Cheese Science

are now produced worldwide. Limburger (Fig-ure 17-20), one of the most important smear-ripened varieties, is named after the region ofLimburg, now divided between the Netherlandsand Belgium, but the cheese is currently manu-factured widely in Germany and North America.The manufacturing protocol (Figure 17-21) issimilar to those for other smear-ripened variet-ies. Limburger cheese is produced from pasteur-ized cow milk acidified using a mesophilic DLstarter and coagulated using calf rennet. Aftercutting the coagulum, the curds and whey arecooked slowly to around 370C. Most of thewhey is then drained off and in some cases re-

placed by dilute brine, which firms the curds andremoves lactose, thereby reducing the level ofacid produced. The curds are then transferred toblock-shaped molds, and whey drainage is as-sisted by frequent turning or the application oflow pressure. During this time, the pH decreasesand the curds mat sufficiently to retain the shapeof the cheese when it is removed from the mold.The cheese is then either salted by the applica-tion of dry salt to the surface or by immersion inbrine (10-150C) for 1-2 days. The saltedcheeses are transferred to ripening rooms kept at10-150C and above 95% relative humidity,where the characteristic surface microflora de-

Figure 17-20 Limburger cheese.

Page 440: Cheese Science

velops during the next 2-3 weeks. During thistime, the cheeses are turned frequently andsmeared with a brine solution containing desir-able microorganisms. After the development ofthe surface microflora, the cheeses are wrappedin foil and ripened for a further 3-8 weeks ataround 40C to complete the development of fla-vor. Limburger is a strong-flavored, soft,rindless cheese with mechanical openings. Anumber of cheeses similar to Limburger are pro-duced, including low-fat Limburger, Romadour,and Weisslacker.

Pont 1'Eveque is a square (10-11 cm sides)smear-ripened cheese that originated in

Normandy and is manufactured from pasteur-ized cow milk. Curds are placed in cloth bags forinitial draining and then in metal molds standingon rush mats. The cheeses are dry-salted. Thesurface flora is dominated initially by G.candidum, which gives the cheese a white ap-pearance. Excessive mold growth is preventedby daily turning and washing with dilute brine.Later during ripening, Pont 1'Eveque develops abacterial microflora characteristic of smear-rip-ened cheeses. Port du Salut and related varieties(e.g., Saint Paulin) are semi-hard cheeses withan elastic body. They are made from cow milk,ripened at a relative humidity above 90%, and

Pasteurized Cow's Milk

32-340CRennet,Mesophilic starter

Coagulum

Cook, 370CDilute brine

Whey

Curds placed in block-shaped moulds

Natural drainage and frequent turning of mouldsor light pressure

Blocks, ~1 kgDry salting or immersion in brine for 1 -2 days

2 - 3 weeks 10 - 150C, 95% erhFrequent smearing and turning of cheese

Surface microflora developedCheese wrapped in foil

Low temperature store 6 -8 weeks

Limburger Cheese

Figure 17-21 Manufacturing protocol for Limburger, a bacterial surface-ripened cheese.

Page 441: Cheese Science

washed periodically with brine to restrict the de-velopment of the smear. Brick is a semi-hardsmear-ripened cheese that originated in theUnited States. Smear development on thischeese is terminated after about 2 weeks by wax-ing or wrapping in film, and ripening is com-pleted at a low temperature during a further 2-3months. Butterkase is another surface-ripenedvariety with limited growth of smear. Thecheese is brine-salted (which encourages the de-velopment of a rind) and smear growth is en-couraged for 2-3 weeks. As suggested by itsname, Butterkase is a soft cheese with a butter-like consistency.

Trappist cheese is reputed to have originatedin a Trappist monastery in Bosnia. Smeargrowth is limited by the short ripening time (2-3weeks). Tilsit is an important smear-ripened va-riety that originated in Prussia. The cheese issomewhat similar to Limburger, but its texture isfirmer and there are more mechanical openings.The cheeses are brine-salted, and a strong smeardevelops during maturation. Taleggio is a softsmear-ripened variety that originated in Lom-bardy in the 1920s and has a characteristicsquare shape. Serra da Estrela is manufacturedin Portugal from sheep milk coagulated using anextract from cardoon flowers of the thistle,Cynara cardunculus. After manufacture,cheeses are first ripened at a high humidity,which promotes the development of a yeastysmear ("reima"). The smear is removed about 14days after manufacture and the cheeses are thenripened at a lower humidity to promote rind de-velopment. Minister cheese, which originated inGermany, is brine-salted, and smear growth isencouraged to give a color to the surface of thecheese. Livarot is produced in Normandy. Someis sold as fresh cheese but most is matured. Dur-ing ripening an intense smear develops. Charac-teristic reed bands are placed around the cheese.Havarti, which originated in Denmark, has nu-merous irregular openings. The cheese is brine-salted, and a dry surface develops when cheesesare held at room temperature for 1-2 days. Asurface flora is then encouraged to contribute toflavor development.

Variants of some of the above smear-ripenedcheeses (e.g., Havarti and Saint Paulin) are alsoproduced in which a smear is not allowed to de-velop. These cheeses are semi-soft and inter-nally bacterially ripened. Such cheeses aresometimes covered with a red or orange coatingto give the impression of smear growth.

17.3 ACID-COAGULATED CHEESES

Acid-coagulated cheeses (e.g., Cream cheese,Cottage cheese, Quarg, and some types of Quesobianco) are produced from milk or cream byacidification to around pH 4.6, which causes thecaseins to coagulate at their isoelectric point(~ pH 4.6). Acid-curd cheeses were probably thefirst cheeses produced, since such products mayresult from the natural souring of milk. Acidifi-cation is usually achieved by the action of a me-sophilic starter, but direct acidification is alsopracticed. A small amount of rennet may be usedin certain varieties (e.g., Cottage cheese andQuarg) but is not essential and serves to increasethe firmness of the coagulum and to minimizecasein losses in the whey. The coagulum may ormay not be cut or cooked during manufacture,but the curds are not pressed. Acid-coagulatedcheeses are characterized by a high moisturecontent and are usually consumed soon aftermanufacture. They are discussed in detail inChapter 16.

17.4 HEAT/ACID-COAGULATEDCHEESES

A small group of cheeses are coagulated by acombination of heat and acid. The most impor-tant member of this subgroup is Ricotta, an Ital-ian cheese originally produced from whey.Ricotta (the name derives from the Italianricottura, meaning "reheated") was producedoriginally from cheese (Mozzarella or Provo-lone) whey, perhaps with some milk added, byheat-induced coagulation (85-9O0C) and someacidifying agent (e.g., lemon juice or vinegar).Ricotta curds were then transferred to molds sur-rounded by ice, where drainage occurred.

Next Page

Page 442: Cheese Science

washed periodically with brine to restrict the de-velopment of the smear. Brick is a semi-hardsmear-ripened cheese that originated in theUnited States. Smear development on thischeese is terminated after about 2 weeks by wax-ing or wrapping in film, and ripening is com-pleted at a low temperature during a further 2-3months. Butterkase is another surface-ripenedvariety with limited growth of smear. Thecheese is brine-salted (which encourages the de-velopment of a rind) and smear growth is en-couraged for 2-3 weeks. As suggested by itsname, Butterkase is a soft cheese with a butter-like consistency.

Trappist cheese is reputed to have originatedin a Trappist monastery in Bosnia. Smeargrowth is limited by the short ripening time (2-3weeks). Tilsit is an important smear-ripened va-riety that originated in Prussia. The cheese issomewhat similar to Limburger, but its texture isfirmer and there are more mechanical openings.The cheeses are brine-salted, and a strong smeardevelops during maturation. Taleggio is a softsmear-ripened variety that originated in Lom-bardy in the 1920s and has a characteristicsquare shape. Serra da Estrela is manufacturedin Portugal from sheep milk coagulated using anextract from cardoon flowers of the thistle,Cynara cardunculus. After manufacture,cheeses are first ripened at a high humidity,which promotes the development of a yeastysmear ("reima"). The smear is removed about 14days after manufacture and the cheeses are thenripened at a lower humidity to promote rind de-velopment. Minister cheese, which originated inGermany, is brine-salted, and smear growth isencouraged to give a color to the surface of thecheese. Livarot is produced in Normandy. Someis sold as fresh cheese but most is matured. Dur-ing ripening an intense smear develops. Charac-teristic reed bands are placed around the cheese.Havarti, which originated in Denmark, has nu-merous irregular openings. The cheese is brine-salted, and a dry surface develops when cheesesare held at room temperature for 1-2 days. Asurface flora is then encouraged to contribute toflavor development.

Variants of some of the above smear-ripenedcheeses (e.g., Havarti and Saint Paulin) are alsoproduced in which a smear is not allowed to de-velop. These cheeses are semi-soft and inter-nally bacterially ripened. Such cheeses aresometimes covered with a red or orange coatingto give the impression of smear growth.

17.3 ACID-COAGULATED CHEESES

Acid-coagulated cheeses (e.g., Cream cheese,Cottage cheese, Quarg, and some types of Quesobianco) are produced from milk or cream byacidification to around pH 4.6, which causes thecaseins to coagulate at their isoelectric point(~ pH 4.6). Acid-curd cheeses were probably thefirst cheeses produced, since such products mayresult from the natural souring of milk. Acidifi-cation is usually achieved by the action of a me-sophilic starter, but direct acidification is alsopracticed. A small amount of rennet may be usedin certain varieties (e.g., Cottage cheese andQuarg) but is not essential and serves to increasethe firmness of the coagulum and to minimizecasein losses in the whey. The coagulum may ormay not be cut or cooked during manufacture,but the curds are not pressed. Acid-coagulatedcheeses are characterized by a high moisturecontent and are usually consumed soon aftermanufacture. They are discussed in detail inChapter 16.

17.4 HEAT/ACID-COAGULATEDCHEESES

A small group of cheeses are coagulated by acombination of heat and acid. The most impor-tant member of this subgroup is Ricotta, an Ital-ian cheese originally produced from whey.Ricotta (the name derives from the Italianricottura, meaning "reheated") was producedoriginally from cheese (Mozzarella or Provo-lone) whey, perhaps with some milk added, byheat-induced coagulation (85-9O0C) and someacidifying agent (e.g., lemon juice or vinegar).Ricotta curds were then transferred to molds sur-rounded by ice, where drainage occurred.

Previous Page

Page 443: Cheese Science

However, much Ricotta cheese is now pro-duced from full-cream or skim milk. The milk isacidified to about pH 6.0 by the addition of alarge amount (typically > 20%) of bulk starter.Unlike in the case of other varieties, the starterdoes not produce acid in the cheese vat and isused simply as a source of preformed lactic acid.Alternatively, the milk may be acidified withacetic acid (white wine vinegar) or citric acid.The acidified milk is heated to about 8O0C bydirect steam injection, during which time NaCl(0.2%, w/v) and stabilizers may be added to themilk. Precipitation occurs in about 30 min atabout 8O0C, after which the curds and whey areheld for 15-20 min. During holding, the curdparticles become firm, coalesce, and float to thesurface due to entrapped air. They are scoopedinto perforated molds, which are cooled withcrushed ice. Ricotta has a high moisture content(~ 73%) and thus has a short shelf-life. It is nor-mally consumed soon after manufacture as atable cheese or as an ingredient in lasagne,ravioli, or desserts.

Ricottone cheese is manufactured roughlylike Ricotta but from sweet whey to whichmilk, skim milk, or buttermilk is added. Sincethe pH of sweet cheese whey is about 6.2, littleadditional acidification is usually necessary.Dry Ricotta, which is a grating cheese, is pro-duced by pressing Ricottone curd and dryingthe cheese at 10-160C for several months (or 4weeks at 210C). Mascarpone is similar toRicotta except that cream is added to the milk,and a slightly higher cooking temperature isused. The resulting cheese is more creamy thanRicotta and is usually salted at a low level,whipped, and formed into a cylindrical shape.Impastata is made like Ricotta except that thecurds are agitated gently as they form, whichcauses them to sink to the bottom of the vat,where they are cooked more efficiently thanRicotta curd, which remains on the surface. Theresult is a drier cheese, with a coarse texture,which is often ground to give a smooth, dough-like texture. Impastata is used mainly in theconfectionery industry as an ingredient inpastry.

Other recooked cheese varieties and theircountry of origin include Bruscio (Corsica),Cacio-ricotta (Malta), Mizthra and Manouri(Greece), and Ziger (Yugoslavia).

17.5 CONCENTRATION ANDCRYSTALLIZATION

A few Norwegian cheeses are produced fromwhey by concentration and crystallization of lac-tose and concentration of other solids in thewhey. One could argue that such varieties arenot cheeses at all but rather byproducts of cheesemanufacture made from whey. These cheeses(Brunost, meaning "brown cheese") are charac-terized by a smooth creamy body and a sweet,caramel-like flavor. Sweet whey is the usualstarting material, although acid whey may beused for some varieties. Sometimes skim milk orcream is added to the whey to give a lighterproduct (which would otherwise be dark brown).

The manufacturing protocol for Primost isshown in Figure 17-22. Primost ("premiumquality cheese") differs from the otherwisesimilar Gjetost by the addition of cream towhey derived from a mixture of goat and cowmilk. The whey-cream mixture is first concen-trated to around 60% total solids (often in amultistage vacuum evaporator). A second con-centration step (to > 80% total solids) follows,and it requires a higher vacuum. The resultingplastic mass is heated to around 950C. TheMaillard reaction is encouraged during themanufacture of these cheeses and is importantfor the final color and flavor. The concentrate isthen cooled, kneaded, and packaged. Crystalli-zation of lactose is controlled so as to avoidsandiness in the product. Several varieties (in-cluding Mysost, Gjetost, Niesost, F10temyost,and Gudbrandsdalsost) are produced using thisbasic process. Differences arise from the originof the whey (cow or goat milk); the addition ofskim milk, milk, or cream to the mix; and theuse of sweet or acid whey. These cheeses havea high total solids content (< 18% moisture),are high in calories, and are characterized by along shelf-life.

Page 444: Cheese Science

Figure 17-22 Protocol for the manufacture of Primost from whey by concentration and crystallization.

Primost Cheese

Blocks lightly waxed

Butter-type printer

Fine-grain, butter-like texture

Remove from kettle;Plastic mass kneaded for -20 min

Release vacuum, heat to -950C

Evaporate in vacuum kettle to 80-84% TS

-60% Total solids (TS)

Evaporate

Cream

Sweet whey

CaseinRennet

Skim milk

SeparateCream

Goats' (88%) and Cows' (12%) Milk

Page 445: Cheese Science

17.6 ULTRAFILTRATIONTECHNOLOGY IN CHEESEMAKING

Ultrafiltration as a technology for cheesemanufacture was introduced in the early 1970sand has been investigated extensively and re-viewed (Ernstrom & Anis, 1985; Lawrence,1989; Lelievre & Lawrence, 1988; Ottosen,1988; Spangler, Jensen, Amundson, Olson, &Hill, 1991; ZaIl, 1985). It has attracted the atten-tion of cheese and equipment manufacturers,primarily because of its potential to increaseyield through the recovery of whey proteins inthe cheese curd (see Chapter 9). Other advan-tages include its potential to reduce productioncosts and to produce new cheese varieties withdifferent textural and functional characteristics.In this section, some of the more important as-pects of ultrafiltration in cheesemaking are high-lighted.

The most successful commercial applicationsof ultrafiltration in cheese manufacture to datehave been in the production of cast Feta in Den-mark, the production of fresh acid-curd varieties(Quarg, Ricotta, and Cream cheeses) in Ger-many and other European countries, and thestandardization of milk protein, to 4-5%, for theproduction of Camembert and other varieties.

Based on the degree of concentration andwhether whey expulsion following concentra-tion is necessary, ultrafiltration in cheesemakingmay be classified into three general areas:

1. Low concentration factor ultrafiltration,followed by cheesemaking and whey re-moval using conventional equipment.The main application of this type of ultra-filtration is the standardization of milk toa fixed protein level to obtain a more con-sistent end product. Variations in gelstrength at cutting, buffering capacity,and rennetcasein ratio are minimized.However, when using conventionalcheesemaking vats, concentration ap-pears limited to a maximum concentra-tion factor of around 1.5, or 4-5% pro-tein, because of difficulties in handlingthe curd and yield losses.

2. Medium concentration factor ultrafiltra-tion (2-6 x) to the final solids content ofthe cheese without whey expulsion. Themain attraction of this type of technologyis the increased cheese yield associatedwith retention of whey proteins and theincreased moisture when whey proteinsare denatured prior to ultrafiltration. Themain commercial application is in theproduction of high-moisture, unripenedcheeses (e.g., Quarg and Cream cheese)and of cheeses that are not very depen-dent on proteolysis during ripening forflavor development (e.g., Feta). Feta pro-duced by this method (by addition of ren-net to a concentrate, i.e., pre-cheese,without cutting the coagulum) has asmoother, more homogeneous texturethan the more "curdy textured" traditionalproduct, hence the name "cast" Feta.

There are numerous reports on usingultrafiltration concentration to attain thefinal cheese dry matter level for the pro-duction of soft or semi-hard rennet-curdcheeses, including Camembert, Bluecheese, Havarti, and Mozzarella. Manu-facture essentially involves preacidifi-cation, ultrafiltration or diafiltration,starter addition, rennet addition, coagula-tion and automated cutting of the coagu-lum using specialized equipment (e.g.,Al-Curd or Ost Retentate coagulators),molding, pressing, and brining. To date,the use of ultrafiltration technology bythe dairy industry for the production ofthe latter cheeses has been limited. Apartfrom uncertainties concerning the regula-tory status of such cheeses and the rela-tively small reported increases in yield,the main drawbacks include changes incheese texture, flavor, and functionality(i.e., meltability and stretchability).

3. High concentration factor ultrafiltrationfollowed by whey expulsion in novelequipment. Because the upper limit ofconcentration by ultrafiltration is about7:1 for whole milk, it is not possible to

Page 446: Cheese Science

achieve the dry matter levels required forhard cheeses such as Cheddar and Gouda.Hence, further whey must be expelledfollowing coagulation of the retentate andcutting of the coagulum. Owing to thehigh curd:whey ratio, efficient curd han-dling (i.e., stirring and heat transfer) is notfeasible in conventional systems. Theonly continuous system capable of han-dling such concentrates is the Siro-Curd,which was used for the production ofCheddar cheese in Australia during the1980s, although its use has been discon-tinued. The cheese produced by this pro-cess, which gives a yield increase ofaround 4-6%, was claimed to be indistin-guishable from Cheddar manufacturedusing standard equipment.

When renneting is done at a fixed dosagelevel, increasing the protein level in milk resultsin a reduced rennet coagulation time, an increasein the level of soluble (nonaggregated) casein atthe point of gelation, an increased rate of curdfirming, a reduced set-to-cut time when cuttingat a given curd strength, a decrease in the degreeof aggregation at cutting, and a coarser gel net-work. Micelles that are not modified, or aggre-gated, at the onset of gelation are presumablymodified later and incorporated into the gel to agreater or lesser degree.

Owing to the rapid rate of curd firming, it be-comes increasingly difficult, as the milk proteinlevel is increased, to cut the coagulum cleanly(without tearing) before the end of the cuttingcycle. Reflecting the tearing of the coagulumand consequent shattering of curd particles, fatlosses in the whey are greater than those pre-dicted on the basis of volume reduction (due toultrafiltration) for milks with protein concentra-tions less than 5%. Similar findings have beenattributed partly to the poorer fat-retaining abil-ity of higher protein curds, which have coarser,more porous protein networks. Reducing the set-ting temperature, in the range 31-270C, and re-

ducing the level of rennet added results in set-to-cut times and curd-firming rates for concen-trated milk closer to those for the control milk.

Increasing the concentration of protein alsoresults in slower proteolysis during ripeningwhen equal quantities of rennet on a milk vol-ume basis are used. The slower rate of proteoly-sis in cheeses made from ultrafiltered milks maybe attributed to a number of factors, including

• the lower effective rennet concentration(i.e., rennetcasein ratio) and hence activityin the cheese

• the inhibition of the indigenous milk pro-teinase plasmin by retained p-lactoglobulinin cheeses containing a significant quantityof whey proteins

• the concentration during ultrafiltration ofindigenous proteinase or peptidase inhibi-tors

• the resistance of undenatured whey proteinsto proteolysis in cheese in which they repre-sent a substantial portion (~ 18%) of theproteins

However, at equal rennet:casein ratios, thelevel of ocsi -casein hydrolysis is higher in controlCheddar cheese than in that made from milkconcentrated fivefold by ultrafiltration. The re-duced surface area:volume ratio (SA:V) of theprotein network in cheeses made from concen-trated milks, resulting from their coarser net-work, may also contribute to the observed reduc-tion in proteolysis. It is conceivable that for agiven level of enzyme activity in the cheesecurd, casein degradation decreases as the SA:Vof the matrix decreases.

Cheese becomes progressively firmer (i.e., re-quires a higher compression force to induce frac-ture), more cohesive, mealier, and drier and thestructure of the protein matrix becomes coarserand more compact (fused) with an increasingconcentration factor. The reduced rate of pro-teolysis results in slower softening and flavordevelopment during maturation.

Page 447: Cheese Science

REFERENCES

Bertozzi, L., & Panari, G. (1993). Cheeses with Appellationd'Origine Controlee (AOC): Factors that affect quality.International Dairy Journal, 3, 297-312.

Burkhalter, G. (1981). Catalogue of cheeses [Document No.141]. Brussels: International Dairy Federation.

Ernstrom, C. A., & Anis, S.K. (1985). Properties of productsfrom ultrafiltered whole milk. In New dairy products vianew technology [Proceedings of International Dairy Fed-eration Seminar, Atlanta]. Brussels: International DairyFederation.

Fox, P.F. (1993a). Cheese: An overview. In P.F. Fox (Ed.),Cheese: chemistry, physics and microbiology (2d ed.,Vol. 1). London: Chapman & Hall.

Fox, P.F. (Ed.) (1993b). Cheese: Chemistry, physics and mi-crobiology: Vol. 2. Major cheese groups (2d ed.). Lon-don: Chapman & Hall.

Guinee, T.P., & Fox, P.F. (1987). Salt in cheese: Physical,chemical and biological aspects. In P.F. Fox (Ed.),Cheese: Chemistry, physics and microbiology (Vol. 1).London: Elsevier Applied Science.

Kosikowski, F. V., & Mistry, V. V. (1997). Cheese and fer-mented milk foods (3d ed., VoIs. 1, 2). Westport, CT:F.V. Kosikowski, LLC.

Lawrence, R.C. (1989). The use ofultraflltration technology

in cheesemaking [Bulletin No. 24O]. Brussels: Interna-tional Dairy Federation.

Lelievre, J., & Lawrence, R.C. (1988). Manufacture ofcheese from milk concentrated by ultrafiltration. Journalof Dairy Research, 55, 465^78.

Ottosen, N. (1988). Protein standardization: Technical in-formation. Silkeborg, Denmark: APV Pasilac.

Robinson, R.K. (Ed.). (1995). A colour guide to cheese andfermented milks. London: Chapman & Hall.

Robinson, R.K., & Wilbey, R.A. (1998). Cheesemakingpractice (3d ed.). Gaithersburg, MD: Aspen Publishers.

Scott, R. (1986). Cheesemaking practice. London: ElsevierApplied Science.

Spangler, P.L., Jensen, L.A., Amundson, C.H., Olson, N.F.,& Hill, G.G., Jr. (1991). Ultrafiltered Gouda cheese: Ef-fects of preacidification, diafiltration, rennet and starterconcentration and time to cut. Journal of Dairy Science,74,2809-2819.

Walter, H.E., & Hargrove, R.C. (1972). Cheeses of theworld. New York: Dover.

ZaIl, R.R. (1985). On-farm ultrafiltration. In New dairyproducts via new technology [Proceedings of Interna-tional Dairy Federation Seminar, Atlanta]. Brussels: In-ternational Dairy Federation.

Page 448: Cheese Science

Appendix 17-A

Compositions of Selected Cheese Varieties

Cheese Fat (%) Total Solids (%) Total Protein (%) Salt (%) Ash (%) pH

Asiago 30.8 72.5 30.9 3.6 6.6 5.3Blue 29.0 58.0 21.0 4.5 6.0 6.5Blue Stilton 33.0 61.7 24.8 3.5 3.2 5.2Brick 30.0 60.0 22.5 1.9 4.4 6.4Bulgarian White 32.3 68.0 22.0 3.5 5.3 5.0Caciocavallo Siciliano 27.5 70.9 33.1 4.0 7.0 6.0Caerphilly 34.0 67.7 27.2 1.5 3.4 5.4Camembert 23.0 47.5 18.5 2.5 3.8 6.9Cheddar (American) 32.0 63.0 25.0 1.5 4.1 5.5Cheshire 33.0 66.7 26.7 1.8 3.9 5.3Comte 30.0 66.5 30.0 1.1 4.1 5.7Cottage 4.2 21.0 14.0 1.0 1.0 5.0Cream 33.5 50.0 10.0 0.75 1.3 4.6Edam 24.0 57.0 26.1 2.0 3.0 5.7Emmental (Swiss) 30.5 64.5 27.5 1.2 3.5 5.6Feta 20.3 40.3 13.4 2.2 2.3 4.2Gouda 28.5 59.0 26.5 2.0 3.0 5.8Grana (Parmesan) 25.0 69.0 36.0 2.6 5.4 5.4Gruyere 30.0 66.5 30.0 1.1 4.1 5.7Havarti 26.5 56.5 24.7 2.2 2.8 5.9Leicester 33.0 64.7 25.5 1.6 3.5 6.5Limburger 28.0 55.0 22.0 2.0 4.8 6.8Manchego 25.9 62.1 28.1 1.5 3.6 5.8Mitzithra 25.0 56.3 18.4 1.6 2.5 5.0Mozzarella 18.0 46.0 22.1 0.7 2.3 5.2Mozzarella-low moisture 23.7 53.0 21.0 1.0 3.0 5.2Muenster 29.0 57.0 23.0 1.8 4.4 6.2PontL'Eveque 25.8 57.2 26.5 2.8 2.4 7.0Provolone 27.0 57.5 25.0 3.0 4.0 5.4Quarg 0.2 21.0 15.0 0.70 1.0 4.5Queso bianco 15.0 49.0 22.9 2.0-3.9 5.4 5.3Ricottone (whey Ricotta) 0.5 27.5 11.0 <0.5 4.0 4.9Ricotta 12.7 28.0 11.2 <0.5 - 5.9Romano 24.0 77.0 35.0 5.5 10.5 5.4Roquefort 31.0 60.0 21.5 3.5 6.0 6.4Samsoe 27.0 59.9 26.5 1.8 3.7 5.5Serra da Estrela 27.5 51.3 21.3 1.9 2.8 6.5Svecia 28.3 56.0 21.8 2.5 4.1 5.5

Page 449: Cheese Science

18.1 INTRODUCTION

Products in this group of cheese products dif-fer from natural cheeses in that they are notmade directly from milk (or dehydrated milk)but rather from various ingredients such as natu-ral cheese, skim milk, water, butter oil, casein,casemates, other dairy ingredients, vegetableoils, vegetable proteins, and minor ingredients.The two main categories, namely pasteurizedprocessed cheese products and substitute or imi-tation products, may be further subdivided de-pending on composition and the types and levelsof ingredients used (Figure 18-1). The indi-vidual categories are discussed below.

18.2 PASTEURIZED PROCESSEDCHEESE PRODUCTS

Pasteurized processed cheese products(PCPs) are produced by comminuting, melting,and emulsifying into a smooth homogeneousmolten blend one or more natural cheeses andoptional ingredients using heat, mechanicalshear, and (usually) emulsifying salts. Optionalingredients permitted are determined by theproduct type (processed cheese, processedcheese food, and processed cheese spread) andinclude dairy ingredients, vegetables, meats, sta-bilizers, emulsifying salts, flavors, colors, pre-servatives, and water (Table 18-1 and Exhibit18-1).

Although a product of recent origin (firstdeveloped in 1911), processed cheese productshave had a growth rate similar to that of naturalcheese (~ 2.3% per annum) in Europe andNorth America during the period 1987-1996(S0rensen, 1997). Current global production is1.9 million tonnes per annum, equal to 17% ofnatural cheese production (S0rensen, 1997).Factors contributing to the continued growthand success of these products are as follows:

• They offer almost unlimited variety in fla-vor, consistency, functionality (e.g., slice-ability, meltability, flowability), and con-sumer appeal as a result of differences informulation, processing conditions, andpackaging into various shapes and sizes.

• They cost less than natural cheese becausethey incorporate low-grade natural cheeseand cheaper noncheese dairy ingredients.

• They are adaptable to the fast-food trade.The most notable example is the inclusionof cheese slices in burgers and the use ofdried processed cheeses (cheese powder) assnack and popcorn coatings.

• They have a relatively long shelf-life, andwaste is minimal.

• Some companies specialize in the manufac-ture of equipment, emulsifying salts, andother ingredients tailor-made to industry'sneed for new products and consistent qual-ity.

Processed Cheese and Substitute orImitation Cheese Products

CHAPTER 18

Page 450: Cheese Science

18.2.1 Classification of Processed CheeseProducts

The named categories of PCPs and their stan-dards of identity (composition and levels andtypes of permitted ingredients) vary somewhatbetween countries. Hence, in the United King-dom there are two categories of PCPs, namely,processed cheese and cheese spread (Cheeseand Cream Regulations 1995 SI 1995/3240),whereas in Germany there are four categories,namely, Schmelzkase (processed cheese),Schmelzkasezubereitung (processed cheese

preparation), Kasezubereitung (cheese prepara-tion), and Kasekomposition (cheese composi-tion), as described in the Deutsche Kasever-ordnung of 12 November 1990. Currently, theInternational Dairy Federation, under the aus-pices of the Codex Alimentarius Commission, isendeavoring to draft a single standard for pas-teurized processed cheese products that will beaccepted globally. It is expected that a Codexstandard will assume increased importance be-cause such a standard will be used by the WorldTrade Organization in the resolution of tradedisputes. In this chapter, the scheme used in the

Pasteurized Processed Cheese and Substitute/Imitation Cheese Products

Pasteurized ProcessedCheese Products

Substitute/ImitationCheese Products

manufactured by blending,heating and shearing mixturesof ingredients mainly of dairyoriginnatural cheese as an ingredientis present at level equivalent to> 510 g/kg final product

manufactured by blendingheating and shearingmixtures of ingredientsof dairy and/or vegetableoriginneed not contain naturalcheese as ingredientnatural cheese may beadded in small quantity(e.g.,< 50 g/kg finalproduct) to impartflavour or to complywith labellingrequirements

manufactured by coagulationof filled milk or soya milk;dehydration of gel usingmethods similar to those fornatural cheese; pressing,salting and/or ripening ofcurd

Categories- Processed cheese- Processed cheese food- Processed cheese spread- Blended cheese

Categories- Dairy analogue cheese- Part-dairy analogue cheese- Non-dairy analogue cheese

Categories- Filled cheese-Tofu

Figure 18-1 Generalized classification scheme for pasteurized processed cheese and substitute or imitationcheese products based on manufacturing procedure and ingredients used.

Page 451: Cheese Science

United States to classify PCPs will be used. Un-der this system, which is described in the Codeof Federal Regulations, Food and Drugs, Part133 (Edition 4-1-93), four main categories ofPCPs are identified:

1. pasteurized process cheese2. pasteurized process cheese food3. pasteurized process cheese spread4. pasteurized blended cheese

A fifth category comprises nonstandarizedpasteurized cheese-type products such as dipsand sauces. The criteria for classification includepermitted ingredients and compositional param-eters. The main aspects of the different catego-ries are summarized in Tables 18-1 and 18-2.

Pasteurized process cheese is usually sold inthe form of sliceable blocks (e.g., processed

Cheddar) or slices. Spreads and foods may be inthe form of blocks, slices, spreads, dips, sauces,or pastes (e.g., in tubes). Pasteurized blendedcheese, which is the least common type, is usu-ally sold in forms that give a natural cheese im-age.

18.2.2 Manufacturing Protocol for ProcessedCheese Products

Manufacture involves the following steps(Figure 18-2):

• formulation of blend, which involves selec-tion of the correct type and quantity of natu-ral cheeses, emulsifying salts, water, andoptional ingredients

• shredding or comminuting of cheese andblending with optional ingredients

Table 18-1 Permitted Ingredients in Pasteurized Processed Cheese Products

Product Ingredients

Pasteurized blended cheese Cheese; cream, anhydrous milk fat, dehydrated cream (inquantities such that the fat derived from them is less than5% [w/w] in finished product); water; salt; food-grade colors,spices, and flavors; mold inhibitors (sorbic acid, potassium/sodium sorbate, and/or sodium/calcium propionates), atlevels < 0.2% (w/w) of finished product

Pasteurized process cheese As for pasteurized blended cheese, but with the following extraoptional ingredients: emulsifying salts (sodium phosphates,sodium citrates; < 3% [w/w] of finished product); food-gradeorganic acids (e.g., lactic, acetic, or citric) at levels such thatpH of the finished product is > 5.3

Pasteurized process cheese foods As for pasteurized process cheese, but with the following extraoptional dairy ingredients: milk, skim milk, buttermilk, cheesewhey, whey proteins—in wet or dehydrated forms

Pasteurized process cheese spread As for pasteurized process cheese food but with the followingextra optional ingredients: food-grade hydrocolloids (e.g.,carob bean gum, guar gum, xanthan gums, gelatin, car-boxymethylcellulose, and/or carrageenan) at levels < 0.8%(w/w) of finished products; food-grade sweetening agents(e.g., sugar, dextrose, corn syrup, glucose syrup, hydrolyzedlactose)

Note: For more detail, see Code of Federal Regulations (1988).

Page 452: Cheese Science

Exhibit 18-1 Optional Ingredients Permitted in Pas-teurized Processed Cheese Products

Dairy ingredients• Anhydrous milk fat, cream, milk, skim

milk solids, whey solids, milk proteins,co-precipitates, milk ultraflltrates

Stabilizers• Emulsifying salts, including sodium

phosphates and sodium citrates• Hydrocolloids: guar gum, xanthan gum,

carrageenans• Organic emulsiflers: lecithin, mono- and

diglyceridesAcidifying agents

• Various food-grade organic acids,including lactic, acetic, phosphoric, andcitric acids

Sweetening agents• Sucrose, dextrose, corn syrup, hydro-

lyzed lactoseFlavors

• Enzyme modified cheese (EMC),artificial flavors, smoke extracts, starterdistillate

Flavor enhancers• NaCl, autolyzed yeast extract

Colors• Annatto, oleoresin, paprika, artificial

colorsPreservatives

• Potassium sorbate, calcium/sodiumpropionates, nisin

Condiments• Cooked meats/fish• Cooked or dried fruit or vegetables

• processing of the blend• homogenization of the hot molten blend

(this step is optional and implementationdepends on the fat content of the blend, typeof cooker used, and body characteristics ofthe end product)

• packaging and cooling

Processing refers to the heat treatment of theblend by direct or indirect steam, with constantagitation. Application of a partial vacuum duringcooking is optional. It may be used to regulate

moisture content when direct steam injection isused, and it is also beneficial in removing air andthus preventing air openings in the finished setproduct. In batch processing, the temperature-time combination varies (70-950C for 4-15min), depending on the formulation; extent ofagitation; and desired product texture, body, andshelf-life characteristics. At a given tempera-ture, the processing time generally decreaseswith agitation rate, which may vary, dependingon the kettle (cooker) type, from 50 to 3000 rpm.In continuous cookers, which are used mainlyfor dips and sauces, the blend is mixed andheated to 80-9O0C in a vacuum mixer, fromwhich it is pumped through a battery of tubularheat exchangers and heated to 130-1450C for afew seconds and then flash cooled to 9O0C. Thecooked product is then pumped to a surge tankthat feeds the packaging machine.

In the manufacture of slices, the hot moltencheese is pumped through a manifold with 8-12nozzles that extrude ribbons of cheese onto thefirst of 2 or 3 counter-rotating chill rolls, whichcool the ribbons from 70-8O0C to 3O0C. The rib-bons are cut automatically into slices, which arestacked and packed.

18.2.3 Principles of Manufacture forProcessed Cheese Products

Micro structurally, natural cheese may beviewed as a three-dimensional paracasein net-work composed of overlapping and cross-linkedchains of partially fused aggregates (in turnformed from fused paracasein micelles). Mois-ture and fat, in the form of discrete or coalescedglobules, are entrapped within the pores of thenetwork, which visually resembles a loose semi-rigid sponge (Figure 18-3). The integrity of theparacasein network is maintained by various in-ter- and intra-aggregate bonds, including hydro-phobic and electrostatic attractions (e.g., cal-cium cross-linking via casein phosphoserine andionized carboxyl residues).

Application of heat (70-9O0C) and mechani-cal shear to natural cheese, as in processing, inthe absence of stabilizers, usually results in a

Page 453: Cheese Science

Table 18-2 Compositional Specifications for Pasteurized Processed Cheese Products

Product Category Moisture (%, w/w) Fat (%, w/w) Fat in Dry Matter (%, w/w)

Pasteurized blended cheese < 43 - > 47

Pasteurized process cheese < 43 - > 47

Pasteurized process cheese foods < 44 > 23 -

Pasteurized process cheese spread 40-60 > 20 -

Note: Minimum temperature and time specified for processing is 65.50C for 30 s. The compositional specifications for pasteurizedprocess cheese may differ from those given, depending on the type of product. For more detail, see the Code of Federal Regulations(1988).

Figure 18-2 Batch process for the manufacture of pasteurized processed cheese products.

Cold storage

Hot fill,Portioning,Packaging

Processing:Heat to ~ 85 0CShear continuously

(cooker)

Blending(mixer)

Formulation

Cheese:clean,remove rind,comminute

Optionalingredients

Emulsifying salts,water

Page 454: Cheese Science

heterogeneous, gummy, puddinglike mass thatundergoes extensive oiling-off and exudation ofmoisture during manufacture, especially uponcooling. These defects arise from

• the coalescence of liquified fat due to shear-ing of the fat globule membranes

• partial dehydration or aggregation andshrinkage of the paracasein matrix inducedby the relatively low pH of cheese (for mostcheeses < 5.7) and the high temperature ap-plied during processing

The modified structure, consisting of a shrunkenparacasein matrix with large pools of free oil andfree moisture, has an impaired ability to occludefat and free moisture. Consequently, free mois-ture and de-emulsified liquified fat seep throughthe more porous, modified structure.

The addition of emulsifying salts (10-30g/kg) during processing promotes emulsificationof free fat and rehydration of protein and thus

contributes greatly to the formation of a smooth,homogeneous, stable product. The emulsifyingsalts most commonly used for pasteurized pro-cessed cheese manufacture include sodium cit-rates, sodium orthophosphates, sodium pyro-phosphates, sodium tripolyphosphates, sodiumpolyphosphates (e.g., Calgon), basic sodium alu-minum phosphates (e.g., Kasal), and phosphateblends (e.g., Joha, Solva blends). These salts gen-erally have a mono valent cation (i.e., sodium) anda polyvalent anion (e.g., phosphate). While thesesalts are not emulsifiers, they promote, with theaid of heat and shear, a series of concerted physi-cochemical changes in the cheese blend that resultin rehydration of the aggregated paracasein ma-trix and its conversion into an active emulsifyingagent. These changes include calcium sequestra-tion, upward adjustment and stabilization (buffer-ing) of pH, paracasein hydration (solvation) anddispersal, emulsification of free fat, and structureformation. They are discussed briefly below.

Figure 18-3 Schematic representation of the structure of natural cheese, showing network of fused paracaseinmicelles (O) that occludes within its pores the fat phase, consisting of individual and partly coalesced globules(•) and moisture (- -).

Page 455: Cheese Science

Calcium Sequestration

This involves the exchange of the divalentCa2+ of the paracasein matrix (attached tocasein via the carboxyl groups of acidic aminoacids and/or by phosphoseryl residues) for themonovalent Na+ of the emulsifying salt. The re-moval of Ca2+, which is referred to as calciumsequestration or chelation, results in

• partial breakdown of the paracasein matrixdue to disintegration of the intra- and inter-aggregate bonds and consequently of thelinks between the strands of the paracaseinmatrix

• an ensuing conversion of the calcium para-casein gel network into a sodium phosphateparacaseinate dispersion (sol), to a greateror lesser degree, depending on the process-ing conditions and type of salt (calciumchelating strength, pH, and buffering ca-pacity)

Displacement and Stabilization(Buffering) ofpH

The use of the correct blend of emulsifyingsalts usually shifts the pH of the cheese upward,typically from around 5.0-5.5 in the naturalcheese to 5.6-5.9 in the processed cheese product,and stabilizes it by virtue of the high bufferingcapacity of the salts. This change contributes tothe formation of a stable product by increasing

• the calcium-sequestering ability of theemulsifying salts

• the negative charge on the paracaseinate,which in turn promotes further disintegra-tion of the calcium paracasein network anda more open, reactive paracaseinate confor-mation, with superior water-binding andemulsifying properties

Hence, the extent of pH buffering is a criticalfactor controlling the textural attributes of pro-cessed cheese products.

Dispersion and Water Binding ofParacasein

Dispersion of paracasein, also referred to aspeptization or swelling, involves the disintegra-tion of the cheese matrix and conversion of thecalcium paracasein into a charged, hydrated so-dium (phosphate) paracaseinate; it is caused bythe above-mentioned emulsifying salt-inducedchanges in combination with the mechanical andthermal energy inputs of processing.

The conversion of calcium paracasein to so-dium (phosphate) paracaseinate during process-ing is the major factor affecting the water-bind-ing capacity of the protein. The increase incasein hydration during processing is consistentwith the inverse relationship found betweencasein-bound calcium and casein hydration.

Emulsification

Under the conditions of cheese processing,the dispersed hydrated paracaseinate contributesto

• emulsification by coating the surfaces ofdispersed free fat droplets, resulting in theformation of recombined fat globule mem-branes

• emulsion stability by immobilization of alarge amount of free water

18.2.4 Structure Formation upon Cooling

During the cooling of processed cheese prod-ucts, the homogeneous, molten, viscous masssets to form a characteristic body, which, de-pending on the blend formulation, processingconditions, and cooling rate, may vary fromfirm and sliceable to semi-soft and spreadable.Factors that contribute to structure formation(setting) during cooling include solidification(crystallization) of fat and protein-protein inter-actions, which result in the formation of a newmatrix. It is envisaged that the newly formedemulsified fat globules become an integral partof the matrix owing to interaction of their para-

Page 456: Cheese Science

caseinate membrane with the paracaseinatematrix.

Electron microscopy studies of processedcheese products indicate the following:

• The protein phase exists in the form of rela-tively short strands connected to varyingdegrees, resulting in a matrix with differentdegrees of continuity depending on producttype. The matrix strands are much finerthan those in natural cheese.

• The fat globules are uniformly distributed(unlike natural cheese) within the proteinmatrix and generally range from 0.3 to 5.0jam in diameter. Fat globule size varies withthe degree of emulsiflcation, which in turnis regulated by the formulation (type andquantity of emulsifying salt and other ingre-dients and age of the cheese) and process-ing conditions (shear rate, temperature, andtime).

• The paracaseinate membranes of the emul-sified fat globules attach to the matrixstrands. The ensuing anchoring of the rela-tively short strands by the recombined fatglobules probably contributes to the conti-nuity and elasticity of the matrix in thecooled product.

The size of the fat globules is important, as itinfluences the firmness of the final product andthe ability of the fat to become free and contrib-ute to oiling-off when the processed cheeseproduct is subsequently cooked (e.g., processedcheese slices on cheeseburgers and processedcheese insets in burgers). When cheese is bakedor grilled, some oiling-off is desirable, as it lim-its drying out of the cheese and thus contributesto the desired flowability, succulence, and sur-face sheen of the melted product (see Chapter19). Generally, for a given formulation, a reduc-tion in the mean diameter of the emulsified fatglobules results in processed cheese productsthat are firmer and exhibit a low tendency to oil-off and poor flowability upon cooking. Com-parative studies on the effect of different emulsi-

fying salts indicate that, for a given processingtime, the mean fat globule diameter is generallysmallest when tetrasodium pyrophosphate(TSPP) is used, largest with basic sodium alumi-num phosphate (SALP), and intermediate withtrisodium citrate (TSC) or disodium phosphate(DSP). Hence, SALP is generally claimed toproduce processed cheeses with good meltingproperties. Increasing the concentration of emul-sifying salt (10^10 g/kg) and processing tem-perature (80-14O0C) results in a progressive de-crease in mean fat globule diameter and aconcomitant increase in firmness. Increasing theprocessing time for a given formulation resultsin final products that are firmer, more elastic,and less flowable (Rayan, Kalab, & Ernstrom,1980). These trends undoubtedly reflect de-creases in the mean fat globule diameter and thelevel of paracasein hydration (or alternatively anincrease in protein aggregation) upon prolongedholding or shearing of the hot molten blend at ahigh temperature. The degree of paracasein hy-dration in cheese is a major factor influencing itsrheology and functionality upon cooking (seeChapter 19).

18.2.5 Properties of Emulsifying Salts

The emulsifying salts most commonly usedare sodium citrates, sodium hydrogen ortho-phosphates, sodium polyphosphates, and so-dium aluminum phosphates (Table 18-3). Otherpotential emulsifying agents include gluconates,lactates, malates, ammonium salts, gluconicacid, lactones, and tartarates. Nowadays, emul-sifying salts are generally supplied as blends ofphosphates (e.g., Joha C special and Solva 35S)or phosphates and citrates (e.g., Solva NZ 10),tailor-made to impart certain functionalities(e.g., different degrees of meltability, slice-ability, spreadibility) to different pasteurizedproducts (e.g., blocks, slices, spreadable prod-ucts) manufactured under different conditions(e.g., with cheeses of varying degrees of matu-rity, in cookers with varying degrees of shear in-put). The properties of different emulsifying

Page 457: Cheese Science

salts have been studied and reviewed exten-sively (Caric & Kalab, 1993; Cavalier-Salou &Cheftel, 1991; Fox, O'Connor, McSweeney,Guinee, & O'Brien, 1996; Rayan, Kalab, &Ernstrom, 1980; Scharf, 1971; Tanaka et al.,1986; van Wazar, 1971).

Sodium citrates are used widely. While theuse of potassium citrates for the manufacture oflow-sodium PCPs has been reported, they arenot normally used commercially, as they tend toimpart a bitter flavor (relative to sodium citrate).Trisodium citrate is used most commonly. Themono- and disodium forms (NaH2C6H5O7 andNa2HC6H5O7), when used alone, tend to result inoveracidic processed cheese products that aremealy, acidic, and crumbly and show a tendencytoward oiling-off due to poor emulsification.The dissociation constants (pKa's) of citric acid,at the ionic strength of milk, are 3.0,4.5, and 4.9.Owing to their acidic properties, mono- and di-sodium citrates may be used to correct the pH ofa processed cheese blend, for example, when ahigh proportion of very mature, high-pH cheeseor skim milk solids is used.

The phosphates used in cheese processing in-clude sodium monophosphates (sodium ortho-phosphates), which contain 1 P atom (n= 1), andlinear condensed phosphates such as pyrophos-phates (n = 2), and polyphosphates (n = 3-25;e.g., tripolyphosphate, n = 3). Of the orthophos-phates, disodium hydrogen orthophosphate(Na2HPO4) is the form normally used. Whenused alone, the mono- and trisodium salts tend toproduce overacidic and underacidic products,respectively. Comparative studies have shownthat the potassium salts of orthophosphates, py-rophosphates, and citrates produce processedcheeses with textural properties similar to thosemade with the equivalent sodium salts at similarconcentrations. Hence, potassium emulsifyingsalts may have potential in the preparation of re-duced-sodium formulations. Owing to their alu-minum content and the possible association ofaluminum with Alzheimer's disease, sodiumaluminum phosphates (e.g., kasal) are used toonly a limited extent.

The effectiveness of the different salts in pro-moting the various physicochemical changesthat occur during processing has been studiedextensively in both pasteurized processedcheese products and analogue cheeses. Discrep-ancies exist between these studies as regards theinfluence of emulsifying salts on different physi-cochemical changes, probably due to differencesin product formulation (e.g., levels of total pro-tein and intact protein, pH), the level of emulsi-fying salts added, and processing conditions(e.g., cooker type, degree of shear, and time-temperature treatment). However, these studiesindicate definite trends, which are summarizedin Table 18-4, and discussed below.

Calcium Sequestration

Ion exchange is best accomplished by saltsthat contain a monovalent cation and a polyva-lent anion, and effectiveness generally increaseswith the valency of the anion. The general rank-ing of the calcium sequestration ability of thecommon emulsifying salts used in cheese pro-cessing is in the following order: polyphos-phates > pyrophosphates > orthophosphates >sodium aluminum phosphates ~ citrates. How-ever, the sequestering ability, especially of theshorter chain phosphates, is strongly influencedby pH. The increased ion exchange function athigher pH values is attributed to more completedissociation of the sodium phosphate molecules,resulting in the formation of a higher valencyanion. Thus, for the shorter chain phosphates,calcium binding increases in the following or-der: NaH2PO4 < Na2HPO4 < Na2H2P2O7 <Na3HP2O7 < Na4P2O7 < Na4P2O7.

Displacement and Buffering ofpH

The buffering capacity of sodium phosphatesin the pH range normally encountered in pro-cessed cheese products (5.5-6.0) decreases withincreasing chain length and is effectively zerofor the longer chain phosphates (n > 10). Thisdecrease in buffering capacity with chain lengthis due to the corresponding reduction in thenumber of acid groups per molecule, which oc-

Page 458: Cheese Science

Table 18-3 Properties of Emulsifying Salts for Processed Cheese Products

pH Value (1% Solution)Solubility at 2O0C (%)FormulaEmulsifying SaltGroup

6.23-6.26

4.0-4.28.9-9.1

4.0-4.56.7-7.5

10.2-10.4

9.3-9.5

9.0-9.56.0-7.5

8.0

High

4018

10.732.0

10-12

14-15

14-15Very high

2Na3C6H5O7JH2O

NaH2PO4.2H2ONa3HPO4-1 2H2O

Na2H2P2O7

Na3HP2O7.9H2ONa4P2O7-IOH2O

Na5P3Oi0

Na6P4O13

Nan+2PnO3n+1 (n= 10-25)

NaH14AI3(PO4)8.4H8O

Trisodium citrate

Monosodium phosphateDisodium phosphate

Disodium pyrophosphateTrisodium pyrophosphateTetrasodium pyrophosphate

Pentasodium tripolyphos-phate

Sodium tetrapolyphosphateSodium hexametaphosphate

(Graham's salt)

Sodium aluminum phosphate

Citrates

Orthophosphates

Pyrophosphates

Polyphosphates

Aluminum phosphates

Page 459: Cheese Science

Table 18-4 General Properties of Emulsifying Salts in Relation to Cheese Processing

AluminumPhosphatePolyphosphatesPyrophosphatesOrthophosphatesCitratesProperty

Low

Very low

High-very high

Low-very low

Very high

Very high (n = 3-1 0)

High-very high

Moderate

Moderate

High

Very high

High

Low

High

Low

Low

Low

Low

High

Low

Low

Nil

Ion exchange (calcium sequestration)

Buffering action in the pH range 5.3-6.0

Paracaseinate dispersion (peptization)

Emulsification

Bacteriostatic effects

Page 460: Cheese Science

cur singly at each end of the polyphosphatechain. The ortho- and pyrophosphates possesshigh buffering capacities in the pH ranges 2-3,5.5-7.5, and 10-12. Thus, in cheese processingthey are suitable not only as buffering agents butalso as pH correction agents. Within the citrategroup, only the trisodium salt has buffering ca-pacity in the pH range 5.3-6.0. The more acidicmono- and disodium citrates produce over-acidic, crumbly cheese with a propensity to oil-ing-off.

The pH of processed cheese products is re-lated to the pH of the solution of emulsifying saltand to its buffering capacity. The pH of analoguecheese made with trisodium citrate or differentsodium phosphate emulsifying salts, at equalconcentrations, decreases in the following order:tetrasodium pyrophosphate «trisodium citrate ~pentasodium tripolyphosphate > disodium hy-drogen phosphate > sodium polyphosphate. ThepH of processed cheese generally increases lin-early with emulsifying salt concentration in therange 0-30 g/kg for trisodium citrate, tetra-sodium pyrophosphate, sodium tripolyphos-phate, and disodium hydrogen phosphate.

Hydration and Dispersion of Casein

The ability of the different groups of emulsi-fying salts to promote protein hydration and dis-persion during cheese processing is in the fol-lowing general order: polyphosphates (n = 3-10)> pyrophosphates > monophosphates ~ citrates.The greater hydrating effect of polyphosphatesover citrates and monophosphates can be ex-plained in terms of the greater calcium-seques-tering ability of the former.

Ability To Promote Emulsification

The effectiveness of different emulsifyingsalts in promoting emulsification, as indicatedby electron microscopy and oiling-off studies, inprocessed cheese is in the following general or-der: sodium tripolyphosphates > pyrophosphates> polyphosphates (P > 10) > citrates ~ ortho-phosphates ~ basic sodium aluminum phos-phates. Their ability to promote emulsification

generally parallels their effectiveness in promot-ing hydration of the paracaseinate complex.

Hydrolysis (Stability)

During processing and storage of processedcheese products, linear condensed phosphatesundergo varying degrees of hydrolysis to ortho-phosphates. The extent of degradation increaseswith processing time and temperature, productstorage time and temperature, and phosphatechain length. Other influencing factors includethe type of cheese, quantity of emulsifying salt,and type of product being produced. In experi-ments with pasteurized processed Emmental,the level of polyphosphate (n > 4) breakdownduring melting at 850C varied from 7% forblock cheese (processed for 4 min) to 45% forspreadable cheese (processed for 10 min).While the breakdown of condensed phosphatesto monophosphates was complete in the spread-able cheese after 7 weeks, low levels were de-tectable in block processed cheese even after 12weeks. The greater degradation of polyphos-phates in spreadable processed cheeses is alsobelieved to be due to their higher pH and mois-ture content.

Bacteriostatic Effects

Cheese processing normally involves tem-peratures (70-950C) that are lower than thoseused for sterilization. Thus, processed cheeseproducts may contain viable spores, especiallyof the genus Clostridium, which originate in theraw materials. Germination of spores duringstorage often leads to problems such as blowingof cans, protein putrefaction, and off-flavors.While bacterial spoilage is minimized throughthe addition of preservatives, some of the emul-sifying salts also possess bacteriostatic prop-erties. Polyphosphates inhibit many microor-ganisms, including Staphylococcus aureus,Bacillus subtilis, Clostridium sporogenes, andvarious Salmonella species. Orthophosphateshave been found to inhibit the growth of CLbotulinum in processed cheese. Citrates possessno bacteriostatic effects and may even be de-graded by bacteria, thus reducing product-keep-

Page 461: Cheese Science

ing quality (Caric & Kalab, 1993). The inhibi-tory effect of sodium orthophosphates on CLbotulinum, which has been found to be superiorto that of sodium citrates in pasteurized proc-essed cheese spreads with moisture levels in therange of 520-580 g/k, depends on the levels ofmoisture and NaCl and the pH of the processedcheese product (Tanaka et al., 1986). The gen-eral bacteriostatic effect of phosphates, whichincreases with chain length, may be attributed totheir interactions with bacterial proteins and se-questration of calcium, which generally servesas an important cellular cation and cofactor forsome microbial enzymes (Stanier, Ingraham,Wheelis, & Painter, 1987).

Flavor Effects

It is generally recognized that sodium citratesimpart a "clean" flavor, whereas phosphatesmay impart off-flavors such as soapiness (in thecase of orthophosphates) and bitterness. Potas-sium citrates also tend to cause bitterness.

18.2.6 Influence of Various Parameters onthe Consistency of Processed CheeseProducts

Numerous investigations have been under-taken to assess the effects of different variables(e.g., levels and types of ingredients, changes inprocessing conditions, and composition) on thetextural and functional characteristics of PCPs.Some discrepancies occur with regard to theconclusions of different studies in which similarvariables were investigated, probably due, inpart, to differences in formulation and process-ing conditions. However, certain trends emerge,which are discussed below.

Blend Ingredients

Cheese is the major constituent of processedcheese products. Its proportion ranges from aminimum of around 51% in cheese spreads andcheese foods to around 98% in processedcheeses. Hence, both the type and degree of ma-turity of the cheese used have a major influenceon the consistency of the product. Block pro-

cessed cheese with good sliceability and elastic-ity requires predominantly young cheese(75-90% intact casein), whereas predominantlymedium-ripe cheese (60-75% intact casein) isrequired for spreads.

There is an inverse relationship between theage (and hence degree of proteolysis) of thecheese and its emulsifying capacity. Therefore,it is not surprising that the meltability and firm-ness of PCPs generally increase and decrease,respectively, with the maturity of the cheeseblend, since there is a positive relationship be-tween the degree of emulsification and firmnessor hardness. Owing to intervarietal differencesin microstructure, composition, and level of pro-teolysis, different types of cheese give processedproducts different consistency characteristics. Itis generally recognized that hard and semi-hardcheese varieties, such as Cheddar, Gouda, andEmmental, which have a relatively high level ofintact casein, give firmer, longer bodied (highfracture strain), more elastic processed productsthan mold-ripened varieties, such as Camembertand Blue cheese. The latter cheeses undergomore extensive proteolyis during ripening andhave a low Ca:casein ratio.

Rework refers to processed cheese that is notpackaged for sale. It is obtained from "leftovers"in the filling and cooking machines—damagedpacks and batches that have overthickened andare too viscous to pump. When added at a maxi-mum level of about 200 g/kg, rework cheese in-creases the viscosity of the molten blend duringprocessing, especially in blends with a highmoisture content (e.g., cheese spreads) or a highproportion of overripe cheese. Overripe cheesetends to give poor emulsification due to the verylow level of intact casein. Addition of reworkcheese generally produces PCPs that are firmerand less spreadable and have poor flowabilitywhen remelted.

Cheese base is being used increasingly as acheese substitute in processed cheese manufac-ture. Its main advantages are its lower cost andmore consistent quality (i.e., intact casein con-tent). Production generally involves ultrafiltra-tion and diafiltration of milk, inoculation of the

Page 462: Cheese Science

retentate with a lactic culture, incubation to a setpH (5.2-5.8), pasteurization, and scraped-sur-face evaporation to 600 g/kg dry matter. Increas-ing the level of cheese substitution with cheesebase generally results in PCPs that are longerbodied, firmer, and less flowable upon remelt-ing. However, the effects vary depending on themethod of cheese base preparation and subse-quent heat treatment during processing:

• Decreasing the pH of milk, in the range6.6-5.2, prior to ultrafiltration gives a lowerCa concentration in the cheese base andyields PCPs with improved meltability.

• Rennet treatment of the ultrafiltration re-tentate results in poorer flowability of thePCPs, an effect that may be attributed togreater interaction of (3-lactoglobulin withpara-K-casein than with native casein dur-ing subsequent processing. This interactioncontributes to the gelation of whey proteins,which impairs the flowability of the PCPupon remelting.

• Treatment of the retentate with other pro-teinases (e.g., Savourase-A, proteinasesfromAspergillus oryzae and Candida cylin-draced) leads to higher levels of proteolysisin the cheese base, which in turn yieldsPCPs that are softer and more flowable thanthose containing untreated base.

• For a given level of cheese base inclusion inthe processed cheese blend, increasing theprocessing temperature in the range66-820C results in PCPs with reducedflowability, an effect attributed to the heatgelation of whey proteins at the higher tem-peratures, especially when rennet-treatedbase is used. In this context, it is noteworthythat flow-resistant PCPs may be preparedby adding a heat-coagulable protein (30-70g/kg whey protein or egg albumin), at atemperature below 7O0C, to the cheeseblend upon completion of processing.

Noncheese dairy ingredients may account fora maximum of about 150 g/kg of the PCP blend.Addition of skim milk powder at a level of 30-

50 g/kg of the blend results in a softer, morespreadable PCP but increases the propensity toundergo nonenzymatic browning during stor-age. Higher levels (70-100 g/kg) are conduciveto the development of textural defects, such ascrumbliness.

The addition of milk protein coprecipitates(produced by high heat treatment of milk fol-lowed by acidification and calcium addition), atlevels up to 50 g/kg of the blend, yields pasteur-ized processed Cheddar that is firmer and lessflowable upon remelting. However, the level atwhich flowability becomes noticeably impairedvaries from 2.5 to 30 g/kg and varies with thesource of the coprecipitate.

Hydrocolloids, including carob bean gum,guar gum, carrageenan, sodium alginate, karayagum, pectins, and carboxymethylcellulose, arepermitted in pasteurized processed cheesespreads at a maximum level of 8 g/kg. Owing totheir water-binding and/or gelation properties,they increase viscosity and thus find applicationin PCPs that have a high water content or a highproportion of overripe cheese. These materials,along with polysaccharides and polysaccharidederivatives (e.g., inulin), are finding increasingapplication as fillers and texturizers in the manu-facture of low-fat products, including PCPs.

Processing Conditions

Increases in the shear and temperature (in therange 70-9O0C) during processing generally re-sult in a higher degree of emulsification andPCPs that are firmer, less spreadable, and lessflowable on remelting. Hence, high moistureformulations, such as processed cheese spreads,are generally subjected to conditions (highertemperature and more vigorous agitation) thatpromote a higher degree of emulsification thanblock processed cheese.

Composition

Although the rheological attributes of pro-cessed cheese products with the same moisturelevel can differ significantly owing to variationsin blend composition and processing conditions,increasing the moisture content generally yields

Page 463: Cheese Science

products that are softer, less elastic, and moreadhesive and spreadable. Product pH has a ma-jor effect on texture. Low pH (4.8-5.2), caused,for example, by use of monosodium citrate,monosodium phosphate, or sodium hexameta-phosphate alone, produces short, dry, crumblycheeses with a tendency toward oiling-off. HighpH values (> 6.0) yield products that tend to besoft and exhibit excessive flow upon remelting.

18.3 IMITATION AND SUBSTITUTECHEESE PRODUCTS

Cheese substitutes or imitation cheeses maybe generally defined as products that are in-tended to partly or wholly substitute for or imi-tate cheese and in which milk fat, milk protein,or both are partially or wholly replaced by non-milk-based alternatives, principally of vegetableorigin. However, their designations and labelingshould by law clearly distinguish them fromcheese or PCPs. The labeling requirements forimitation and substitute cheeses have been re-viewed by McCarthy (1991). In the UnitedStates, "an imitation cheese is defined as a prod-uct which is a substitute for, and resembles, an-other cheese but is nutritionally inferior, wherenutritional inferiority implies a reduction in thecontent of an essential nutrient(s) present in ameasurable amount but does not include a reduc-tion in the caloric or fat content" (Food and DrugAdministration Regulation 101.3, Identity La-beling of Food in Packaged Form (e)). A substi-tute cheese is defined as a product that is a sub-stitute for and resembles another cheese and isnot nutritionally inferior. Outside the UnitedStates, there is little specific legislation coveringimitation or substitute cheeses. For pertinent in-formation regarding designation and labeling,the reader should consult International DairyFederation (1989), McCarthy (1991), andcurrent national regulations and the CodexAlimentarius.

There are few, if any, standards relating topermitted ingredients or manufacturing proce-dures for imitation cheese products. The prod-

ucts may be arbitrarily classified into three cat-egories—analogue cheeses, filled cheeses, andtofu—based on the ingredients used and themanufacturing procedure (Figure 18-1). The ef-fects of various ingredients, various processingconditions, and low temperature storage on thequality of imitation cheese products have beenreported extensively (see Cavalier-Salou &Cheftel, 1991; Marshall, 1990; Mulvihill &McCarthy, 1994; Yang & Taranto, 1982). How-ever, in many of these studies, model productformulations that bear little resemblance to thoseused in commercial practice have been used.Despite the fact that substitute Mozzarellacheese is the principal substitute cheese productused commercially, research has concentratedprincipally on composition and texture, withlittle focus on functionality and viscoelasticproperties during melting and comparison withthose of natural cheese. Therefore, most of therelevant information on these products is of aproprietary nature. Pertinent reviews includeShaw (1984) and International Dairy Federation(1989). The individual products are discussedbelow.

18.3.1 Cheese Analogues

Analogue cheeses, which were introduced inthe United States in the early 1970s, constituteby far the largest group of imitation or substitutecheese products. The manufacture of analoguesof a wide variety of natural cheeses (e.g., Ched-dar, Monterey Jack, Mozzarella, Parmesan,Romano, Blue cheese, and Cream cheese) andpasteurized processed cheese products has beenreported in the trade literature. Based on feed-back from the marketplace, current annual pro-duction of analogue cheese in the United States,the major producing region, amounts to around300,000 tonnes. The major products are substi-tutes for or imitations of low-moisture Mozza-rella, Cheddar, and pasteurized processed Ched-dar. These products find application mainly ascheese topping for frozen pizza pie and as slicesin beef burgers. Other applications include usein salads, sandwiches, spaghetti sprinkling,

Page 464: Cheese Science

cheese sauces, cheese dips, and ready-mademeals. European production is estimated to berelatively small (~ 20,000 tonnes/annum), a factthat may be attributed to the lack of a commoneffective legislation policy, the efforts of groupswith the objective of protecting the designationof milk and dairy products, the lower level ofpizza consumption compared with the UnitedStates, and the fact that the flavor systems usedfor analogues are still not developed to a pointwhere the analogues have the same flavors andtextures as the corresponding table cheese prod-ucts. The success of analogue cheese products inthe United States may be attributed to a numberof factors:

• Natural cheeses cost more than substitutes.The low cost of analogues is due to the lowcost of vegetable oils compared with butter-fat, the low cost of casein imported fromNew Zealand and Europe (price-subsi-dized), the absence of a maturation period

(for natural cheeses maturation costs aboutUS$1.4/week), and the relatively low costof manufacturing equipment comparedwith that required for natural cheese.

• They offer a diverse functionality range(e.g., flowability, melt resistance, andshredability) made possible by tailor-madeformulations, and they exhibit high func-tional stability during storage.

• Fast food and ready-made meals have be-come extremely popular.

• They can be designed to meet special di-etary needs through formulation changes(e.g., products can be lactose free, low calo-rie, low in saturated fat, and vitamin en-riched).

Classification

Cheese analogues may be arbitrarily catego-rized as dairy, partial dairy, or nondairy, depend-ing on whether the fat and/or protein compo-nents are from dairy or vegetable sources

Figure 18—4 Classification of analogue cheeses based on the sources of proteins and oils used in product formu-lation.

No Ripening

CaseinCasematesButter oil

CaseinCasematesVegetable oils

Vegetable proteinsVegetable oils

Dairy Partial Dairy Nondairy(vegetable)

Processed cheese manufacturing methods

Cheese Analogues

Page 465: Cheese Science

(Figure 18-5). Partial dairy analogues, in whichthe fat is mainly vegetable oil (e.g., soya oil,palm oil, rapeseed, and their hydrogenatedequivalents) and the protein is dairy based (usu-ally rennet casein and/or casemate) are the mostcommon. Nondairy analogues, in which both fat

and protein are vegetable derived, have little orno commercial significance and to the authors'knowledge are not commercially available. Thepreparation of experimental substitute or imita-tion cheese products (e.g., Mozzarella cheeseand PCPs) from various vegetable proteins (e.g.,

Mix for ~ 1 min at room temperature in cheese cooker

Formulation of blend

Process: heat to ~ 85 0C9 shear continuously

Oil

Water

Dry ingredients- casein,- starch- emulsifying

salts

Process: heat to ~ 85 0C, shear continously

Homogeneous molten masspH~8.5

Flavors Acid regulator.

Homogeneous molten masspH~5.7-6.0

Mold and hot pack

Place molded cheese at 4 to -4 0C

Figure 18-5 Typical manufacturing procedure for analogue pizza cheese.

Page 466: Cheese Science

peanut and soya proteins or blends of these pro-teins with casein) has generally shown that thesubstitution of casein by vegetable proteins re-sults in impaired texture. Dairy analogues arenot produced in large quantities because theircost is prohibitive. The following discussionpertains to partial dairy analogue cheese (e.g.,low-moisture Mozzarella type) for use in pizzapie (analogue pizza cheese, APC).

Analogue Pizza Cheese

Manufacturing Protocol. The manufacture ofanalogue pizza cheese (APC), which is similarto the manufacture of PCPs, involves the formu-lation, processing, and packing of a hot moltenproduct. A typical formulation (Table 18-5)shows that it differs from that for PCPs in thatcheese is not normally included, although somecheese may be introduced as a flavoring agent oras required by customer specifications for labeldeclaration. While production methods varysomewhat, a typical manufacturing procedure(Figure 18-6) involves

• simultaneous addition of the requiredamount of water and of dry ingredients(e.g., casein and emulsifying salts)

Table 18-5 Typical Formulation for AnaloguePizza Cheese

Level AddedIngredient (g/1 OO g Blend)

Casein and caseinates 23.00Vegetable oil 25.00Starch 2.00Emulsifying salts 2.00Flavor 2.00Flavor enhancer 2.00Acid regulator 0.40Color 0.04Preservative 0.10Water 38.50Condensate3 7.00

a Upon cooking the blend to about 850C using direct steaminjection, condensate equivalent to about 7.0 g is absorbed bythe blend.

• addition of oil and cooking to about 850Cusing direct steam injection while continu-ously shearing until a uniform homoge-neous molten mass is obtained (typically 5-8 min)

• addition of flavoring materials (e.g., en-zyme-modified cheese or starter distillate)and acid regulator (e.g., citric acid) andblending the mixture for a further 1-2 min

• packing the hot molten blend

Horizontal twin screw cookers operating attypical screw speeds of 40 rpm are used in themanufacture of APC. This design of the cookerensures adequate blending and a relatively lowdegree of mechanical shear, compared with thehomogenizing effects of some processed cheesecookers. These process conditions, together withthe correct formulation, promote a low degree offat dispersion and hence a relatively large fatglobule size (e.g., 5-25 jam). Upon subsequentbaking of the analogue cheese on pizza pie, therelatively large fat globules ensure a sufficientdegree of oiling-off, limit dehydration of thecheese topping, and thereby are conducive toachieving the desired flow and succulence char-acteristics. It is noteworthy that there is gener-ally an inverse relationship between the degreeof fat emulsification and the flowability ofPCPs.

The addition of flavors toward the end of themanufacturing process minimizes the loss of fla-vor volatiles in the dissipating steam. In themanufacture of PCPs, the pH of the final product(e.g., 5.5-5.9) is regulated by adding the correctblend of emulsifying salts, which adjust andbuffer the pH of the blend during processing tothis pH value. In contrast, the addition of food-grade acids toward the end of the process (fol-lowing casein hydration and oil dispersion oremulsification) is the normal procedure used toadjust the pH of the cooked APC to that requiredin the finished product. This protocol is essentialto ensure a high pH (typically > 7.0) during pro-cessing of the product when rennet casein is themajor protein ingredient. A high pH at this stagegives a higher negative charge to the casein and

Page 467: Cheese Science

is conducive to greater calcium sequestration bysodium phosphate emulsifying salts. Both fac-tors contribute to the efficient hydration of ren-net casein and hence the emulsification of addedvegetable oil. Two factors necessitate this proce-dure:

• Rennet casein has a higher (e.g., 35 mg/gcasein) calcium content than natural cheese(e.g., ~ 28 mg/g casein for Cheddar).

• Rennet casein is in a dehydrated statewhereas the casein in natural cheese is hy-drated to a degree dependent on the extentof proteolysis, pH, and concentrations ofNa and Ca in the moisture phase.

Principles of Manufacture. The principles ofmanufacture of APC from rennet casein aresimilar to those for PCPs. The combined effectsof emulsifying salts, heat, and shear promote se-questration of Ca from the rennet casein (dehy-drated paracasein, which in effect is equivalentto dehydrated cheese protein), pH adjustmentand buffering of the blend, casein hydration, fatdispersion and fat emulsification by the hydrated

paracaseinate, and setting of the molten massupon cooling.

Composition and Functionality. Comparisonof the mean composition of commercial samplesof low-moisture Mozzarella cheese (LMMC)and APC (Table 18-6) indicates that while manyof the gross compositional parameters of APCare similar to those of LMMC, the latter gener-ally has a lower level of protein and higher levelsof fat-in-dry-matter, Ca, and P. Although intra-varietal differences in composition occur forboth cheese types, they are more pronounced inAPC. Moreover, the sum of the mean values formoisture, fat, protein, and ash account for onlyabout 930 g/kg dry matter (compared with ~ 990g/kg in LMMC), suggesting the addition of car-bohydrate-based ingredients (e.g., lactose, mal-todextrins, starch) during formulation. Thesematerials may be added to impart certain func-tional characteristics to the end product and/or aspartial substitutes for rennet casein, thereby re-ducing formulation costs. The relatively largecompositional variations exhibited by APCsprobably reflect deliberate differences in formu-

Figure 18-6 Confocal laser scanning micrographs showing the microstructures of commercial samples of low-moisture Mozzarella (a) and analogue pizza cheese (b). The bar = 25 jam; protein is shown in black, and fat isshown in white/gray. In the low-moisture Mozzarella, the protein is in the form of elongated fibers and the fat isin the form of pools trapped between the protein fibers. In the analogue pizza cheese, the protein is not organizedinto libers and the fat is mainly in the form of discrete globules.

A B

Page 468: Cheese Science

lation so as to achieve customized function-alities in the finished products.

Some important functional attributes ofmelted cheese on a cooked pizza pie are as fol-lows:

• Melt time. An index of how rapidly theshredded cheese on a pizza pie melts andflows into a homogeneous molten massshowing no traces of shred identity.

• Flowability. A measure of the degree offlow.

• Stretchability. A measure of the tendency toform cohesive strings or sheets when ex-tended.

• Apparent viscosity. A measure of chew-iness (Table 18-7 and Chapter 13).

Upon baking, a good-quality pizza cheesemelts relatively quickly, flows adequately togive the desired degree of surface coverage, andpossesses the desired degrees of chewiness andStretchability, which, perhaps more than otherfunctional properties, endow pizza pie with itsunique culinary qualities (see Chapter 19). Com-parison of the functional characteristics of com-mercial LMMC and APC indicates that bothcheeses have similar mean values for melt time,flowability, and apparent viscosity. However,the Stretchability of APC is generally inferior tothat of LMMC. The differences in Stretchabilitybetween LMMC and APC may be related prima-

rily to differences in the degree of aggregationand microstructure of the paracasein caused bydifferences in the procedures used to manufac-ture the two products. During the manufacture ofLMMC and other Pasta filata cheeses, such asProvolone or Kashkaval, the cheese curd, ataround pH 5.15, is subjected to a plasticizationprocess, which involves heating to around 55-6O0C and kneading the curd in hot (e.g., 7O0C)water or dilute brine. These conditions promotea limited degree of aggregation and contractionof the paracasein gel matrix and thereby lead tothe formation of paracasein fibers with a hightensile strength (see Figure 18-7 and Figure19-7). The cheese fat is physically entrappedbetween the paracasein fibers. In contrast, theconditions used in the manufacture of APC aredesigned to disaggregate and hydrate the para-casein aggregates of rennet casein and casemate.The hydrated paracaseinate immobilizes largequantities of added water and emulsifies theadded vegetable oil, thereby contributing to for-mation and physicochemical stability of theproduct. Hence, unlike low-moisture Mozza-rella, the casein in the APC is in the form of apartially hydrated dispersion rather than para-casein fibers (Figure 18-7). Both LMMC andAPC exhibit marked intravarietal differences at-tributable to designed differences in formulation(for APC) or processing conditions and degreeof maturity (for LMMC). The intravarietal dif-

Table 18-6 Typical Compositions of Low-Moisture Mozzarella and Analogue Pizza Cheese

Low-Moisture Mozzarella Analogue Pizza Cheese

Moisture (g/100 g) 46.4 48.8Protein (g/1 OO g) 26.0 18.5Fat (g/1 OO g) 23.2 25.0Fat-in-dry-matter (g/1 OO g) 44.6 49.0Salt-in-moisture (g/100 g) 3.1 3.5Ash (g/1 OO g) 3.9 4.2Ca(mg/100g) 27.5 34.4pH 5.5 6.1

Note: Values presented are means of 8 samples of each cheese type sourced in Ireland, the United Kingdom, and/or Denmark.

Page 469: Cheese Science

ferences in functionality allow manufacturers tocustomize their cheeses to the requirements ofdifferent pizzerias.

Functional Stability during Storage. Thefunctional properties of LMMC change mark-edly during storage at 40C (see Chapter 19). Ini-tially, during the first 1-5 days, depending onmanufacturing procedure and composition, thecheese is nonfunctional and burns or crusts dur-ing baking. After 5-10 days of ripening, thecheese acquires functionality, as reflected by thedecreases in melt time and apparent viscosity(chewiness) and the increases in flowability andstretchability (see Chapter 19). Thereafter, thesechanges occur more slowly, and the cheese re-tains desirable functionality for around 40-50days. However, prolonged aging of LMMC (e.g.,to 75 days) is associated with excessive flow-ability, loss of chewiness, and a "soupy" consis-tency in the grilled or baked cheese. Moreover,the uncooked shredded cheese develops an in-creased susceptibility to clumping or sticking, anundesirable change, as it leads to the blocking ofcheese-dispensing units on pizza pie productionlines and to nonuniform distribution of thecheese topping on the pizza pie. The functionalchanges that occur during ripening are mediatedby proteolysis of paracasein (by plasmin andpossibly residual coagulant), hydration andswelling of the paracasein matrix, and coales-cence of the fat phase (see Chapters 11 and 19).The increase in the hydration of the paracasein isthought to result from a number of factors: pro-teolysis and the concomitant increase in the num-ber of amino and carboxyl groups, the increase in

pH during early ripening, and the low concentra-tion of NaCl in the moisture phase of the cheese(which is conducive to hydration of the para-casein). Few studies have considered thechanges in casein-based APCs during ripening.Mulvihill and McCarthy (1994) reported a pro-gressive increase in proteolysis (e.g., pH 4.6-soluble N increased from ~ 35 at O days to 195 g/kg total N after 51 weeks) and decreases in elas-ticity and chewiness during storage at 40C for 51weeks. However, the changes during the first 6weeks were relatively small. Normally, ana-logues are used within 1 month after manufac-ture. Kiely, McConnell, and Kindstedt (1991)reported that casein-based APCs were morefunctionally stable than LMMCs during storageat 40C for 28 days. In the authors' experience,casein-based analogues containing a high levelof starch (> 40 g/kg) may lose their functionalityrelatively rapidly (e.g., after 4 weeks) duringstorage at 40C, an effect possibly associated withthe retrogradation of amylose. The loss of func-tionality is reflected by the increase in loosemoisture upon shredding, the loss of meltabilityand flowability, and burning or crusting uponbaking. Added starch may undergo postmanu-facture retrogradation during cold storage toform gels to an extent depending on processingconditions (e.g., the level of heat and shear) andthe level and type of starch used (e.g., amyl-ose ramylopectin ratio and whether the starch isnative or modified). These gels contract duringstorage, resulting in the expulsion of moisture. Itis envisaged that these changes in the starch com-ponent of APCs during storage result in products

Table 18-7 Functionality of Low-Moisture Mozzarella and Analogue Pizza Cheese

Functional Attributes Low-Moisture Mozzarella Analogue Pizza Cheese

Aggregation index 3.95 3.74Melt time (seconds) 108 105Flowability (%) 53 42Stretchability (cm) 87 28Apparent viscosity (Pa x s) 630 650

Page 470: Cheese Science

with higher levels of unbound water and impedeflowability upon remelting.

18.3.2 Filled Cheeses

Filled cheeses generally differ from naturalcheeses in that the milk fat is partly or fully re-placed by vegetable oils, which may be partiallyhydrogenated to impart a melting profile similarto that of milk fat. However, filled cheeses maybe categorized into two types depending onwhether the base material is native skim milk orreformed skim milk; the latter is prepared by dis-persing dairy ingredients, such as whey and totalmilk protein, in water. In all cases, preparationof the filled milk involves dispersion of the veg-etable oil in the native or reformed skim milkusing a high speed mixer and subsequent ho-mogenization of the blend. Dispersion and ho-mogenization ensure emulsification of the addedvegetable oils and thus prevent phase separationand/or excessive creaming during cheese-making.

The filled milk is then subjected to the con-ventional in-vat cheesemaking procedure usedfor the variety being substituted for or imitated.

Homogenization of milk generally results incurd that synereses poorly and therefore tends toyield cheeses with a higher moisture content,lower yield stress and firmness, and lowerflowability upon remelting than those made

REFERENCES

Abou El-Ella, W.M. (1980). Hard cheese substitute from soymilk. Journal of Food Science, 45, 1777-1778.

Caric, M., & Kalab, M. (1993). Processed cheese products.In P.P. Fox (Ed.), Cheese: Chemistry, physics and micro-biology (2d ed., Vol. 2). London: Chapman & Hall.

Cavalier-Salou, C., & Cheftel, J.C. (1991). Emulsifyingsalts' influence on characteristics of cheese analogs fromcalcium caseinate. Journal of Food Science, 56, 1542—1547, 1551.

Code of Federal Regulations. (1988). Part 133: Cheese andrelated products. In Food and Drugs 21. Code of FederalRegulations, Parts 100 to 169. Washington, D.C.: USGovernment Printing Office.

from nonhomogenized milk (Fenelon andGuinee, unpublished results).

18.3.3 Tofu or Soybean Cheeses

Torn, a stable food in the Orient for centuries,is a tough, rubbery curd made from soya (soy-bean) milk. Manufacture essentially involvessoaking the soybeans in water for a long period(during which they swell), adding extra water,grinding and milling the bean-water mixture intoa smooth slurry, and filtering the slurry to obtainsoya milk. The soya milk is boiled to induce p-rotein denaturation, cooled to about 370C, andcoagulated by adding a divalent salt (e.g., cal-cium lactate) and adjusting the pH to 4.5-5.0(McCarthy, 1991; Tharp, 1986). Following co-agulation, the whey is drained off and the curd ismolded and lightly pressed to give tofu, in whichthe levels of dry matter, protein, fat, and carbo-hydrate are typically 152, 77,42, and 24, respec-tively. The molded curd may be subjected to ahigh pressure and brine-salted to yield soybeancheeses with a higher dry-matter level than Tofu(e.g., 530 g/kg dry matter; Abou El-Ella, 1980).Ras cheese made from soya milk was found tohave a higher moisture level and received lowersensory scores for color, flavor, and body andtexture characteristics than that made from cowmilk by conventional cheesemaking procedures(Abou El-Ella, 1980).

Fox, P.P., O'Connor, T.P., McSweeney, P.L.H., Guinee,T.P., & O'Brien, N.M. (1996). Cheese, physical, bio-chemical and nutritional aspects. Advances in Food andNutrition Research, 39, 163-329.

International Dairy Federation. (1989). The present and fu-ture importance of imitation dairy products. [Bulletin No.239]. Brussels: Author.

Kiely, LJ., McConnell, S.L., & Kindstedt, P.S. (1991). Ob-servations on the melting behaviour of imitation mozza-rella cheese. Journal of Dairy Science, 74, 3568-3592.

Marshall, RJ. (1990). Composition, structure, rheologicalproperties and sensory texture of processed cheese ana-logues. Journal of the Science of Food and Agriculture,50, 237-252.

Page 471: Cheese Science

McCarthy, J. (1991). Imitation cheese products [BulletinNo. 249]. Brussels: International Dairy Federation.

Mulvihill, D.M., & McCarthy, A. (1994). Proteolytic andrheological changes during ageing of cheese analoguesmade from rennet casein. International Dairy Journal, 4,15-23.

Rayan, A.A., Kalab, M., & Ernstrom, C.A. (1980). Micro-structure and rheology of process cheese. Scanning Elec-tron Microscopy, 3, 635—643.

Scharf, L.G., Jr. (1971). The use of phosphates in cheese pro-cessing. In J.M. Deman & P. Melnchyn (Eds.), Phos-phates in food processing. Westport, CT: AVI PublishingCo.

Shaw, M. (1984). Cheese substitutes: Threat or opportunity?Journal of the Society of Dairy Technology, 37, 27-31.

S0rensen, H.H. (1997). The world market for cheese [Bulle-tin No. 326]. Brussels: International Dairy Federation.

Stanier, R.Y., Ingraham, J.L., Wheelis, M.L., & Painter, P.R.(1987). General microbiology (5th ed.). London:Macmillan Press.

Tanaka, N., Traisman, E., Plantinga, P., Finn, L., Flom, W.,Meske, L., & Guggisberg, J. (1986). Evaluation of factorsinvolved in antibotulinal properties of pasteurized presscheese spreads. Journal of Food Protection, 49, 526-531.

Tharp, B. (1986, September). Frozen desserts containingtofu. Dairy Field, pp. 38^2, 59.

van Wazer, J.R. (1971). Chemistry of the phosphates andcondensed phosphates. In J.M. Deman & P. Melnchyn(Eds.), Phosphates in food processing. Westport, CT:AVI Publishing Co.

Yang, C.S.T., & Taranto, M.V. (1982). Textural propertiesof Mozzarella cheese analogs manufactured from soyabeans. Journal of Food Science, 47, 906-910.

Page 472: Cheese Science

19.1 INTRODUCTION

Owing to its numerous varieties, cheese offersthe consumer a very wide diversity of flavors,aromas, and textures. Hence, as a product,cheese has been enjoyed since antiquity, as testi-fied by the numerous references to it in earlywritings (see Chapter 1).

While it is generally assumed that cheese wasoriginally eaten on its own or with bread, hu-mans probably soon realized that it enhanced theorganoleptic qualities of other foods to which itwas added. The Romans were the first to recordthe use of cheese as an ingredient (Ridgway,1986). Typical applications included the blend-ing of hard cheese with oil, flour, and eggs in thepreparation of cakes and the mixing of softcheeses with meat or fish, boiled eggs, and herbsin the making of pies. Cheese has long been usedas a culinary ingredient, along with other foodsand condiments, to create an extensive array ofdishes (Figure 19-1). Today, natural cheese con-tinues to be used as a major ingredient in the ho-tel and catering industry. Typical cheese dishesinclude omelets, quiches, sauces, chicken cor-don-bleu, and pasta. Natural cheese is also usedextensively by the industrial catering sector inthe mass production of ready-to-use gratedcheeses, shredded cheeses, cheese blends, com-bination products, and cheese-based ingredients,such as pasteurized processed cheese products(PCPs), cheese powders, and enzyme-modifiedcheeses (EMCs). Combination products contain

two or more types of food, such as cheese, meat,fish, and vegetables, each of which retains itsown distinct identity (e.g., as a layer in the prod-uct). They are generally produced by coextru-sion of the different foods or by the dipping ofone food (e.g., cooled salami) into a hot moltenform of another food (e.g., pasteurized proc-essed cheese). Commercially, cheese-based in-gredients are used by the catering industry (e.g.,burger outlets, pizzerias, and restaurants) and bythe manufacturers of formulated foods such assoups, sauces, and ready-made meals.

In this chapter, the functional properties ofnatural cheese, cheese powders, and enzyme-modified cheese as an ingredient are discussed.Cheese base, a concentrated cultured ultra-filtered milk retentate (-550 g/kg dry matter;pH ~ 5.3), is an important cheeselike ingredientthat finds applications as a substitute for youngcheese in PCPs. The use of cheese base in PCPsis discussed in Chapter 18.

19.2 OVERVIEW OF THEREQUIREMENTS OF CHEESE AS ANINGREDIENT

When used as an ingredient in food applica-tions, cheese is required to perform one or morefunctions, some of which are listed in Table19-1. In its natural state, cheese is required toexhibit a number of rheological properties so asto facilitate its use in the primary stages ofpreparation of various dishes, such as the ability

Cheese as a Food Ingredient

CHAPTER 19

Page 473: Cheese Science

INDUSTRIALINDUSTRIALINDUSTRIALINDUSTRIAL

PCPsReady mealsImitation cheeseDressingsDipsSeasonings

ENZVMEMODIFIEDCHEESE

CHEESEPOWDERS

SauceSoupSnack coatingsCake mixesReady mealsGratinsBiscuitsPizza toppingsDipsCrisps

PASTEURIZEDPROCESSEDCHEESEPRODUCTS(PCPs)

Cheese burgersSalamiLuncheon rollsCo-extruded productsSaucesDips

SHREDDED,DICED,CRUMBLED,DRIEDCHEESES

Pizza toppingsReady mealsFillingsSprinklingsFilled sandwiches

CATERING

COOKED,BAKED

SaucesFonduesSoupsGratinsRacletteRarebitLasagnePizza breadQuicheOmeletSoufflesPasta dishes

DessertsSaladsCheesecakes

UNBOOKED

CHEESEDISHES

Cheese boardAccompaniment to

bread/crackersSandwiches

TABLECHEESE

COMMODITY

CATERINGHOlVE

MAJOR USER: HOMEMNORUSER: CATERING

Figure 19-1 Uses of cheese as an ingredient.

INGREDIENTS

Page 474: Cheese Science

Table 19-1 Typical Requirements of Cheese as a Food Ingredient

Examples of Cheese or Cheese-Based IngredientExamples of Food ApplicationsRequirement

Feta, Cheshire, StiltonStilton

Swiss-type, Gouda, EdamSwiss-type, Cheddar, MozzarellaCheddarCheddar

Swiss-type, Cheddar, MozzarellaMozzarella, Provolone, Cheddar, analogue pizza

cheese, MontereyCheddar, Romano, Provolone

Grated Parmesan and RomanoCheese powdersCheese powders, enzyme-modified cheeses

Quarg, Fromage frais, Cream cheese

Cream cheese, RicottaMascarponeCream cheese

Dried cheeses, especially rennet-curd varieties(high in calcium)

Mozzarella, Cheddar, Raclette, Swiss, Romano,analogue pizza cheese, PCPs

Mixed saladsSoup

Filled cheese rolls (finger foods)Sandwiches (filled, open, toasted)Cheese slices in burgersCheese slices on crackers

Consumer packs of shredded cheesePizza pie (frozen/fresh baked)

Pasta dishes (lasagne, macaroni and cheese)

Cheese sprinklings (e.g., on lasagne)Snack coating (e.g., popcorn)Dry soup/sauce mixes

Fresh cheese desserts

CheesecakeTiramisuHome-made desserts

Baby foods

All cooked dishes (including sauces, fondues,pizza pie)

Ability to crumble when rubbed

Sliceability

Shreddability

Free-flow when shaken

Ability to flow when blended withother materials in the raw state

Ability to "cream" or to form a pastewhen sheared

Nutritional value

Meltability upon grilling/baking

Page 475: Cheese Science

Mozzarella, Cheddar, Swiss, analogue pizzacheeses

PCPs, Cream cheese

PCPs, analogue pizza cheese, custom-madeMozzarella or string cheese

PCPs, analogue pizza cheesePaneer, acid-coagulated Queso bianco

Mozzarella, Kashkaval, young Cheddar, analoguepizza cheese

Halloumi, Mozzarella, Provolone, Kashkaval, youngCheddar

Mozzarella, Kashkaval

Cheddar, RomanoCheddar, Romano, ParmesanMozzarella, analogue pizza cheese

Cheese powders, PCPsCheese powders, Cheddar, Blue cheese, PCPsCream cheese

Cheddar, Romano, Swiss-type, ParmesanCheese powders, enzyme-modified cheesesCheese powdersCheese powdersDried cheesesCheese powders

Most cooked dishes (e.g., pizza pie, cheeseslices on burgers)

Chicken cordon-bleu

Deep-fried breaded cheese sticks

Deep-fried burgers containing cheese insetsFried cheese dishes

Pizza pie

Pizza pie

Pizza pie

Macaroni and cheeseLasagnePizza pie

SoupsSaucesCheesecake

Most cheese dishes, soupsBaked productsSnack coatingsDressingsBaby foodReady-made meals

Flowability upon grilling/baking

Flow resistance upon deep-frying

Stretchability when baked/grilled

Chewiness when baked/grilled

Limited oiling-off when baked/grilled

Limited browning when baked/grilled

Viscosity

Flavor

Key: PCPs = pasteurized processed cheese products.

Page 476: Cheese Science

to crumble easily, to slice or to shred cleanly, orto bend when in sliced form. The rheologicalproperties also determine the textural propertiesof the cheese during mastication. Cheese is gen-erally required to contribute to the organolepticcharacteristics (flavor, aroma, and texture) of thefood in which it is an ingredient. Upon grillingor baking, the cheese may be required to melt,flow, stretch, brown, blister, oil-off, and/orstretch to varying degrees. The baked cheesemay also be expected to be chewy (as in pizzapie) and contribute to certain mouth-coatingcharacteristics (as in sauces and soups). In manydishes, including sauces, the cheese is requiredto be able to interact with other food componentssuch as water, carbohydrates, proteins, and fatsduring food preparation.

19.3 FUNCTIONAL PROPERTIES OFCHEESE AS AN INGREDIENT

The ability of cheese to fulfill its expected re-quirements as an ingredient is related to its func-tional properties. These may be defined as thoserheological, physicochemical, microstructural,and organoleptic properties that affect the be-havior of the cheese in food systems duringpreparation, processing, storage, cooking, and/or consumption and that therefore contribute tothe quality and organoleptic attributes of thefood in which cheese is included.

The functional properties of cheese can beclassified into four main types:

1. Rheology-related properties of the rawcheese. These properties are exhibitedwhen the cheese is subjected to a stress(such as cutting, shearing, and mastica-tion) or strain (compression and exten-sion). They may include sensoric proper-ties such as consistency, fracturability,crumbliness, stickiness, firmness, hard-ness, and softness (see Chapter 13).

2. Rheology-related properties of the heatedcheese. These properties are exhibitedwhen the cheese is subjected to a stressthroughout its mass as a result of heat-in-

duced physicochemical and microstruc-tural changes such as liquefaction of thefat, protein dehydration, fat coalescence,and matrix collapse. Included amongthem is the ability of the cheese to melt,flow, and stretch.

3. Physicochemical and microstructuralproperties induced by heating. Theseproperties include oiling-off, browning,blistering, fat coalescence and exudationor separation, interaction of free aminogroups with reducing sugars, moistureevaporation, and paracasein aggregationand precipitation.

4. Flavor- and aroma-related properties.These include properties such as ched-dariness, saltiness, fruitiness, piquancy,and sweetness.

19.3.1 Functional Properties of Raw Cheese

The functional properties of raw cheese arerelated to its taste and aroma and its rheologicalcharacteristics, discussed in Chapters 12 and 13,respectively. Although the rheological proper-ties do not directly affect the taste and aroma,their influence on the rate and extent of break-down during mastication may alter the lattercharacteristics indirectly. For example, a cheesewith low values for fracture stress and strain isexpected to deform rapidly and release its fatmore quickly after a given residence time in themouth. Free liquid oil quickly coats parts of themouth, allowing the aroma and taste of its vola-tile and nonvolatile flavor compounds to be per-ceived rapidly.

Rheology-Related Characteristics

The primary stage of preparation of any foodcontaining cheese requires that the cheese massbe reduced in size so as to facilitate dispersion,mixing, and/or layering onto the food. Size re-duction is usually achieved by cutting into rela-tively large pieces and then crumbling, slicing,shredding, dicing, grating, and/or shearing.These actions usually involve a combination ofcompressive and shear stresses. The behavior of

Page 477: Cheese Science

the cheese when subjected to different size re-duction methods constitutes a group of impor-tant functional properties, which are listed inTable 19-2. These properties are related to therheological characteristics of the cheese, whichdetermine the magnitude of the strain (e.g.,change in dimensions and fracture) upon appli-cation of the stresses applied by the differentsize-reduction methods.

The rheology-related functional properties ofthe raw cheese determine its suitability for aparticular application (Table 19-1), as shownby the examples below. Mature Camembert andChaumes, which are soft, short, and adhesive(see Chapter 13), are not used in shredded ordiced cheese applications, such as pizza pie, be-cause of their tendency to ball and clump. How-ever, the ability of these cheeses to undergoplastic fracture and flow under shear (i.e.,spread) makes them ideal for blending withother materials such as butter, milk, or flour inthe preparation of fondues and sauces (Tables19-1 and 19-2). The brittleness and tendency ofhard cheeses, such as Parmesan and Romano,with low levels of moisture and fat-in-dry-mat-ter (FDM) to undergo elastic fracturability en-dow them with excellent gratability, such aswhen crushed between rollers, and they are suit-able for sprinkling onto dishes such as spaghettiBolognase. However, these properties renderthe latter cheeses unsuitable in food applica-tions that require cheese slices, such as filledsandwiches or cheeseburgers. Conversely, otherhard cheeses, such as Cheddar and Gouda, areunsuitable for grating owing to their lack ofbrittleness and to their elasticity and relativelyhigh fracture stress and strain, which enables arelatively high degree of recovery to their origi-nal shape and dimensions following crushing.Moreover, the relatively high moisture and fat-in-dry matter levels of the latter cheeses areconducive to a higher degree of flow followingfracture than Romano or Parmesan and hence tothe development of tackiness following crush-ing. Owing to its springiness, elasticity, and"long" body, Swiss-type cheese is ideal for slic-ing very thinly and therefore is particularly well

suited for applications such as filled sandwichesand rolled cheese slices containing fillings.Similarly, the springiness of low-moisture Moz-zarella cheese (LMMC) endows it with goodshreddability (a low tendency to fracture andform fines or curd dust) and nonstick propertiesand facilitates uniform distribution on the sur-face of pizza pies (Tables 19-1 and 19-2). Ow-ing to their crumbliness, cheeses such as Feta,Cheshire, and Caerphilly are particularly wellsuited for inclusion in mixed salads. While itcould be argued that shreddable cheese such asMozzarella or Gouda could also be included inmixed salads, crumbly cheeses are more desir-able, as the irregularly shaped, curdlike par-ticles are more visually appealing to the con-sumer than cheese shreds (they convey animage of "real" cheese).

Factors That Influence theFunctionality of Raw Cheese

Little or no information is available on factorsthat influence the various rheology-related func-tional properties of raw cheese, apart fromshreddability.

The suitability of cheese for shredding may bequantitatively assessed by determining the ten-dency of the shredded cheese to aggregate orclump when vibrated under controlled condi-tions similar to those used on commercial pizzaproduction lines. Cheese, after storing at 40C for12 hr or more, is cut into cubes of fixed dimen-sions (e.g., 2.5 cm), and a fixed weight (W\) ofshredded cheese is placed immediately on thetop sieve of a stack of sieves ranging in aperturefrom 9.5 to 1 mm. The stack is vibrated at a fixedamplitude for a given time, resulting in thecheese shreds passing through the stack to a de-gree dependent on their susceptibility to stick orclump on the one hand or fracture on the other.The cheese on each sieve is then weighed (W)and an aggregation index (AGI) is calculated:

AGI = 2 (Wx SA)IW1

where SA is the sieve aperture. A higher AGIvalue corresponds to a higher susceptibility to

Page 478: Cheese Science

Table 19-2 Functional Properties of Raw Cheese That Influence Its Functionality as an Ingredient

Positively Associated RheologicalParameters

Cheeses Generally Displaying thePropertyDefinitionProperty

Elasticity, springiness, firmness,longness

Elasticity, springiness, firmness,longness

Elastic fracturability

Plastic fracturability, softness,adhesiveness, shortness

Elastic fracturability at low defor-mation, low cohesiveness

Low-moisture Mozzarella, Swiss-typecheese, medium-aged Cheddar,Gouda, Provolone

Low-moisture Mozzarella, Swiss-typecheese, Provolone, analogue pizzacheese, PCPs (some)

Parmesan, Romano

Mature Camembert, Cream cheese,mature Blue cheese

Blue cheese, Cheshire

The ability of a cheese block to shredinto thin strips of uniform dimensions,resist fracture during shredding, andresist clumping/balling duringshredding

The ability to be cut cleanly into thinslices without fracturing or crumblingor sticking to cutting implement

The ability to fracture (elastically) intosmall particles, with a low tendency tostick, upon shearing and crushing

The ability to spread easily whensubjected to a shear stress

The ability to break down into smallirregular shaped pieces when rubbed(at low deformation)

Shreddability

Sliceability

Gratability

Spreadability

Crumbliness

Page 479: Cheese Science

aggregation and clumping. Similar approachesare also used commercially (e.g., hand vibratingcheese in a colander and determining the quan-tity retained after a fixed time).

Young Mozzarella (1-5 days old) generallydoes not shred well owing to the large amount offree moisture on its surface and within the bodyof the cheese (Kindstedt, 1995). Thereafter,shreddability improves, and it becomes optimalafter about 3 weeks' storage at 40C, but it dete-riorates progressively during further storage,and the cheese becomes soft and sticky. Freewater and stickiness are undesirable, as they pro-mote clumping of the cheese shreds, which leadsto blocking of cheese-dispensing units on pizzapie production lines, poor distribution of cheeseon pizza pies, and matting of shredded cheesewhen placed in retail packs.

The trend in the shreddability of LMMC as afunction of maturation time is associated withincreases in the levels of primary proteolysis,the water-binding capacity of the cheese pro-teins, and the free (nonglobular) fat in thecheese during ripening (see Section 19.3). Theeffect of free fat on shreddability depends onthe ratio of solid to liquid fat, which decreasesas the temperature is raised. At temperatureswhere milk fat is largely in the liquid state (60%of total fat at 2O0C), free fat exudes to the sur-face of the cheese shreds, where it acts as an ad-hesive for other shreds. This defect is com-pounded by temperature fluctuations duringstorage. For example, cooling after holding atambient temperature (~ 2O0C) leads to solidifi-cation of exuded fat, which makes the breakingup of clumps of cheese shreds more difficult.Such a problem is sometimes encountered insmall retail units where stacked packs of gratedcheese are subjected to temperature fluctua-tions, leading to clumping of the entire contentsof the packs, especially those at the bottom ofthe pile. Hence, in practice, cheeses are cooledto about 20C prior to shredding and distributingon pizza pies to avoid blockage of cheese shred-ding and dispensing machines. Other factorsthat are conducive to clumping of shreddedcheese include

• longer shred length and shred diameter,which increase the chance of shred en-tanglement

• increasing moisture content, although theeffect appears to be related to the method ofMozzarella production (e.g., whether acidi-fied by a starter culture or food-grade acid),composition, and level of primary proteoly-sis

• increasing fat content (in the range 50-330g/kg for Cheddar cheese)

The AGI value for a range of commercialcheeses indicates that cheese variety has amarked influence on shreddability (Figure19-2). No intervarietal correlation was foundbetween AGI and individual gross composi-tional parameters, pH, or level of primary pro-teolysis (as measured by N solubility at pH 4.6).

Although little or no published information isavailable on factors that affect the other rheol-ogy-related functional properties of raw cheese,it may be assumed that these are influenced byconditions that impact the rheological character-istics. The factors that influence the latter arediscussed in Chapter 13 and include

• cheese macrostructure, which determinesthe discontinuity of the cheese matrix dueto curd granule junctions, chip boundaries,cracks and fissures, gas holes, and eyes

• cheese composition (e.g., the levels ofmoisture, protein, fat, salt, and pH)

• temperature of the cheese, which influencesthe ratio of solid fat to liquid fat

• the extent of cheese maturation, whichcauses certain physical changes in the struc-tural components during ripening (e.g.,changes in the ratio of intact casein to hy-drolyzed casein, the level of casein hydra-tion, and/or the degree of fat coalescence)

Organoleptic Characteristics

Cheese is generally required to contribute tothe organoleptic characteristics of most foods inwhich it is incorporated by possessing certaintaste, aroma, texture, and/or mouth-coatingcharacteristics. The importance of the contribu-

Page 480: Cheese Science

tion of the cheese ingredient to overall flavor ishighlighted by the development of and growth inthe use of highly flavored enzyme-modifiedcheeses for a range of products, such as PCPs,imitation cheese products, and cheese powders.These products, in turn, are used to impartcheesy flavor to other products, such as ready-made meals, snacks, soups, and sauces. The in-creasing importance of cheese flavor is alsohighlighted by the increasing use of cheeses,such as mature Cheddar and Colby, which havepoor stretchability compared to LMMC, in pizzacheese toppings. Cheese flavor is discussed inChapter 12.

19.3.2 Functional Properties of HeatedCheese

Comparison of Different Cheese Types

The properties displayed by cheese uponcooking are of importance in many applicationsof cheese, especially grilled cheese sandwiches,pizza pie, cheeseburgers, pasta dishes, andsauces. The various terms used to describe the

functional properties of baked or grilled cheeseare defined in Table 19-3. Upon grilling or bak-ing, the cheese may be required to melt, flow,stretch, brown, blister, oil-off, and/or stretch tovarying degrees; it may also be expected to bechewy and contribute to certain mouth-coatingcharacteristics. The cooking application deter-mines whether one or more of these functions isnecessary. Despite the increasing use of differ-ent cheese varieties in cooked dishes, surpris-ingly little is known about the functionality ofcheeses other than LMMC, which has been stud-ied extensively and reviewed (Guinee, MuI-holland, Mullins, & Corcoran, 1997; Kindstedt,1995).

Guinee (unpublished study) compared thefunctional properties of commercial samples ofdifferent natural cheese varieties and analoguepizza cheeses upon heating. There were large in-ter- and intravarietal differences in melt time,flowability, stretchability, and apparent viscos-ity, which is an index of chewiness (Table19-4). This trend undoubtedly reflects inter-varietal differences in the conditions of manu-

A B C D E F G H I

Cheese code

Figure 19-2 Susceptibility of different types of shredded cheese to clumping (as measured by aggregation in-dex): Gruyere (A), Emmental (B), Appenzeller (C), analogue pizza cheese (D), Kashkaval (E), low-moistureMozzarella (F), Cheddar (G), Tetilla (H), and Fontina (I).

Agg

rega

tion

inde

x

Page 481: Cheese Science

Table 19-3 Functional Properties of Grilled or Baked Cheese That Influence Its Functionality as an Ingredient

Property Related toPhysicochemical State

Cheeses GenerallyDisplaying This PropertyDefinitionProperty

Fat liquifaction, fat coalescence

Fat liquifaction, casein hydration, highdegree of fat coalescence, limitedoiling-off

Moderate degree of casein hydrationand casein aggregation, level andtype of molecular attractions betweenparacasein molecules

Absence of fat coalescence, heat-induced gelation of particularcomponent(s) (e.g., whey proteins),thermo-irreversibility of gel system inthe uncooked product upon heating

As for stretchability

Relatively high level of casein hydration

Limited degree of fat coalescence

Adequate degrees of casein hydrationand of oiling off during baking; lowlevel of residual reducing sugars(e.g., lactose galactose) and Maillardbrowning

Most cheeses after a given storageperiod, PCPs, APCs, Creamcheese

Most cheeses after a given storageperiod, PCPs, OACs, Creamcheese

Low-moisture Mozzarella,Kashkaval, young Cheddar (15days)

Paneer, PCPs, OACs, naturalcheeses made from high heat-treated milk

Low-moisture Mozzarella,Kashkaval, young Cheddar (15days)

Mature Cheddar, aged Mozzarella,Cream cheese, PCPs, OACs

Most natural cheeses (if not verymature or very young), PCPs,APCs

Mature Cheddar, aged Mozzarella,Cream cheese, PCPs, OACs(depending on formulation)

The ability of cheese to soften to amolten cohesive mass on heating

The ability of the melted cheese to flow

The ability of the melted cheese to formcohesive fibers, strings, or sheetswhen extended uniaxially

The resistance to flow of melted cheese

High resistance to breakdown uponmastication

Low resistance of melted cheese tobreakdown upon mastication

Ability of cheese to express a little freeoil upon heating so as to reducecheese dehydration and thereby tomaintain succulence of and impartsurface sheen to melted cheese

Desired degree of surface sheen withfew, if any, dry, scorched black orbrown patches

Meltability

Flowability

Stretchability

Flow resistance(often referred to asmelt-resistance)

Chewiness (rubbery,tough, elastic)

Viscous (soupy)

Limited oiling off

Desirable surfaceappearance

Key: PCP = processed cheese product; APC = analogue pizza cheese; OAC = other analogue cheese.

Page 482: Cheese Science

facture, composition (Table 19-5), degree ofmaturity, and/or formulation (e.g., levels andtypes of added ingredients) in the case of theAPCs.

The pasta filata cheeses were differentiatedfrom all other varieties by their superior stretch-ability (Figure 19-3), relatively high apparentviscosity (Figure 19-4), moderate flowability(Figure 19-5), and melt time (Figure 19-6).These functional properties endow these cheeseswith attributes typically associated with themelted cheese on pizza pie; sufficiently rapidmelt and desirable levels of stringiness, chewi-ness, and flow. Some of the pasta filata cheeses(Provolone dolce and string cheese) had a veryhigh viscosity, which would undoubtedly be as-sociated with overchewiness on pizza pie. Thefunctional requirements of the global pizza mar-ket tend to be region specific. However, whilethe functional requirements of different pizze-rias vary somewhat within and between coun-tries, cheese with the following characteristics isgenerally acceptable: melt time less than 120 s;flowability, 40-55%; stretchability, greater than75 cm; apparent viscosity, 800^00 Pa x s; AGI,3.5^.5. Hence, the higher flow and/or apparentviscosity of Provolone dolce and string cheese

may be better suited to the customized require-ments of local pizza markets. While the degreeof browning required on pizza pie appears to de-pend very strongly on the pizzeria, generally alow degree is more desirable.

The superior stringiness of pasta filatacheeses upon baking, compared with other natu-ral cheeses, may be attributed primarily to plasti-cization of the curd during the kneading andstretching process. In this process, which isunique to the manufacture of pasta filata-typecheeses, the milled curd is heated to about 57-6O0C and kneaded in hot water or brine at about78-820C when the curd pH is about 5.2. Thecombined effect of the high temperature and lowpH during kneading results in casein aggrega-tion and contraction of the strands of the para-casein gel, resulting in the formation of para-casein fibers of high tensile strength. The curdkneading and stretching operations of the plasti-cization process, which are carried out in equip-ment that simulates the traditional hand stretch-ing (Figure 7-6), give the newly formed curdfibers a linear orientation (parallel to the direc-tion of stretching). Confocal laser scanning mi-crographs of the curd before and after textur-ization clearly demonstrate the formation and

Table 19-4 Functional Characteristics of Different Types of Natural Cheeses

Sample Melt Flowability Stretchability ApparentCheese Type Size Time(s) (%) (cm) Viscosity (Pa x s)

Pasta filata-typeLow-moisture 8 108(6) 53(8) 83(21) 623(303)

MozzarellaKashkaval 2 96(11) 67(4) 87(13) 522(330)Provolone dolce 3 86(6) 64(21) 80(13) 950 (-)Provolone fumica 1 92 71 76Provolone 1 92 71 76

Cheese with eyesGruyere 1 105 78 67 391Jarlsberg 1 82 52 35 371Emmental 1 81 74 35 269

Cheddar 8 100(7) 69(9) 23(10) 349(129)Analogue pizza cheese 8 105(13) 42(19) 27(8) 668(307)

Note: Where the sample size > 2, mean values are presented. Values in parentheses are standard deviations.

Page 483: Cheese Science

Table 19-5 Compositional Analyses of Different Commercial Cheese Varieties

pH

PTAN(9/10Ogtotal N)

pH4.6SN(9/10Ogtotal N)

P(mg/g cheese

protein)

Ca(mg/g cheese

protein)AshS/M

(9/kg)MNFS(9/kg)

FDM(9/kg)

Fat(9/kg)

Protein(9/kg)

Moisture(9/kg)

SampleSizeSourceCheese Type

5.53

5.37 (0.02)

5.54(0.14)

5.29

5.29

5.83

5.72

5.64

5.14(0.12)

6.3(0.1)

0.5

0.7 (0.2)

4.6 (2.5)

0.2(0.1)

4.7

6.1 (3.9)

10.2(9.8)

7.5

10.8

10.8

8.6

9.2

20.3 (3.7)

2.3 (0.8)

20.6

22.2(18)

20.6(1.3)

28.1 (44)

27.2

31 .0 (5)

28.0(1.4)

34.4(21)

38

43(0)

37(3)

42(3)

31

49(2)

57(7)

47

44

49

28

13

45(7)

35(4)

604

596 (4)

546(17)

573

562

540

587

553

556(10)

651 (44)

446

458(17)

477 (36)

428

476

558

524

578

526 (24)

490 (34)

232

256(12)

294 (37)

243

285

368

313

380

331 (20)

250(19)

260

253 (3)

276 (5)

278

281

277

277

249

254 (8)

184(19)

464

441 (7)

386 (32)

434

402

341

404

343

372(11)

489 (37)

8

2

2

1

1

1

1

1

8

8

Ireland, UK,Denmark

Yugoslavia

Italy

Italy

Italy

Switzerland

Norway

Switzerland

Ireland

Ireland

Pasta filata typeLow moisture

Mozzarella

Kashkaval

Provolone dolce

Provolone fumica

Provolone

Cheese with eyesGruyere

Jarlsberg

Emmental

Cheddar

Analogue pizzacheese

Note: Where sample size > 2, mean values are presented. Values in parenthesis are standard deviations.

Key: FDM = fat-in-dry matter; MNFS = moisture-in-non-fat substance; S/M = salt-in-moisture; pH4.6 SN = nitrogen soluble at pH 4.6; PTAN = nitrogen soluble in 5% phosphotungsticacid.

Page 484: Cheese Science

A B C D E F G H I J

Cheese code

Figure 19-4 Apparent viscosity, at 7O0C, of different cheese types: low-moisture Mozzarella (A), Kashkaval(B), Provolone (C), string cheese (D), Emmental (E), analogue pizza cheese (F), Cheddar (G), Parmesan (H),Raclette (I), and Appenzeller (J).

App

aren

t vis

cosi

ty, P

as

A B C D E F G H I JCheese code

Figure 19-3 Stretchability of different cheese types after baking at 28O0C for 4 min: low-moisture Mozzarella(A), Kashkaval (B), Provolone (C), string cheese (D), Emmental (E), analogue pizza cheese (F), Cheddar (G),Parmesan (H), Raclette (I), and Appenzeller (J).

Stre

tcha

bilit

y, c

m

Page 485: Cheese Science

Flow

abili

ty, %

A B C E F G I J

Cheese code

Figure 19-5 Flowability of different cheese types after baking at 28O0C for 4 min: low-moisture Mozzarella (A),Kashkaval (B), Provolone (C), Emmental (E), analogue pizza cheese (F), Cheddar (G), Parmesan (H), Raclette(I), and Appenzeller (J).

Mel

t tim

e, s

A B C D E F G H I J

Cheese code

Figure 19-6 Melt time of different cheese varieties at 28O0C: low-moisture Mozzarella (A), Kashkaval (B),Provolone (C), string cheese (D), Emmental (E), analogue pizza cheese (F), Cheddar (G), Parmesan (H), Raclette(I), and Appenzeller (J).

Page 486: Cheese Science

Figure 19-7 Confocal laser scanning micrographs of low-moisture Mozzarella cheese at different stages ofproduction and storage at 40C. (a) Before plasticization: salted and milled curd showing paracasein matrix strands(long arrows) and void spaces containing fat globules (short arrows), (b) After plasticization (24 hr storage at40C): stretched curd showing extensive linearization of paracasein into fibers (long arrows) and fat globules orpools (short arrows), (c) After plasticization (43 days storage at 40C): aged cheese showing hydrated swollenparacasein fibers forming a continuous protein phase occluding fat mainly in the form of pools. Bar = 25 jam.

C

B

A

Page 487: Cheese Science

linearization of protein fibers (Figure 19-7).In contrast to the pasta filata cheeses, most

other cheese types, including analogue pizzacheeses, Cheddar, and Emmental, had relativelylow stretchability, low apparent viscosity (< 400Pa x s) and in some cases excessive (> 55%) orimpaired (19%) flowability characteristics. Ifsuch cheeses were used on pizza pie, the meltedcheese would lack the desired stringiness, wouldflow excessively or insufficiently, and wouldlack the desired chewiness. Conversely, stringi-ness, which is typical of LMMC and other pastafilata cheeses such as Kashkaval and Provolone,is an undesirable attribute for applications suchas sauces, fondues, and toasted sandwiches.Cheeses such as mature Cheddar, Emmental,Raclette, and Gouda are more satisfactory be-cause of their excellent flowability and flavorand the absence of stringiness upon baking orgrilling.

The findings of the above study indicate that,whereas many cheeses have good flow and vis-cosity characteristics, few, apart from LMMCand other pasta filata cheeses, possess the goodstretchability associated with cheese topping onpizza pie. However, other cheeses, such asCheddar, are being used increasingly on pizzapie to impart more cheese flavor. Relativelylittle is known on how the blending of LMMCwith other cheeses affects the functionality ofthe cheese topping on pizza pie. The inter-varietal differences in functional propertiesprobably arise from variations in proteolysis, thewater-binding capacity of the cheese proteins,and the concentrations of structural components,namely, fat, protein, and moisture (as discussedin the following section). The preliminary find-ings suggest that more knowledge is required toidentify the effects of the different factors thatcontribute to the functionality and viscoelastic-ity of baked or grilled cheeses to enable an effi-cient approach to the blending of differentcheese varieties for pizza cheese toppings.

In other applications, melt is essential butvery limited flow is required so as to preservethe shape and identity of the cheese. Examplesof the latter include fried Paneer, grilled or friedburgers containing cheese insets, and deep-fried

breaded cheese sticks. Most mature naturalcheeses are unsuitable for these applications ow-ing to excessive flow and oiling-off during cook-ing. In the case of cheese insets in deep-friedburgers, such attributes would cause the cheeseto permeate the interstices of the coarse meatemulsion and thereby cause the cooked cheese tolose its shape and visual effect. For these appli-cations, pasteurized processed cheese products,which are specially formulated to endow themwith varying degrees of flow resistance, aremore suitable than most natural cheeses (Figure19-8). Selective manipulation of formulationand processing conditions enables the structuraland physicochemical properties of naturalcheese to be modified readily by heating andshearing processes and by blending with otheringredients, such as emulsifying salts and wheyproteins (see Chapter 18). However, some natu-ral cheeses, such as Queso bianco and Creamcheeses (manufactured using a particular proto-col), exhibit excellent flow resistance and are,therefore, often used in applications such asdeep-fried battered cheese sticks or cheese insetsin various dishes. The manufacture of thesecheeses generally involves a high heat treatmentof the milk and/or the curd (e.g., > 850C for 5-15min) following whey drainage, as in Creamcheese produced by the ultrafiltration method.This results in high levels of whey protein dena-turation (e.g., ~ 600 g/kg total whey protein) andthe complexation of whey protein with casein.Owing to the heat gelation of the included wheyproteins during subsequent baking or grilling ofthe cheese, they impede flow of the cheese.

In other meat products, such as cordon-bleupoultry products, extensive flow is required,with little or no oiling-off or stringiness.

Factors That Influence theFunctionality of Cheeses uponCooking

Little information is available on the factorsthat affect the functional characteristics of natu-ral cheeses during cooking, apart from LMMC.Most natural cheeses are consumed principallyas table cheeses, and the functionality of thecooked cheese is of only secondary importance

Page 488: Cheese Science

compared to the taste and texture of the naturalcheese.

19.3.3 Ripening Time

Changes in the functional characteristics ofLMMC during storage have been reported ex-tensively (Guinee et al., 1997; Kindstedt, 1995).Comparatively little information is available onthe age-related changes in other cheeses such asother pasta filata cheeses, Cheddar, Parmesan, orEmmental. LMMC made by the conventionalprocedure (see Chapter 17) generally is non-functional upon cooking during the first 5-10days of storage at 40C after manufacture. This isreflected by drying out and crusting of thecheese on the pizza pie during baking owing tothe low water-binding capacity of paracasein(and the concomitant extensive dehydration ofthe cheese during baking) and the low propen-sity of the cheese to express free oil. Both factors

are conducive to excessive evaporation of mois-ture at the high temperature (typically ~ 9O0C) inthe mass of melting cheese when heated in aconvection oven at 28O0C for 4 min). The dried-out, crusted cheese lacks succulence and fails tomelt, flow, or stretch. Moreover, it is extremelytough and chewy, as reflected by a high apparentviscosity of more than 1,000 Pa x s.

The functionality of LMMC improves mark-edly during the first 2 weeks of ripening at 40C, asindicated by decreases in melt time and apparentviscosity and increases in flowability and stretch-ability. This status is maintained until about 40-50 days (Figures 19-9 to 19-12). The improvedfunctionality may be attributed to increases inprotein hydration and free fat during aging of thecheese (Figures 19-13 and 19-14) (Guinee et al.,1997; Kindstedt & Guo, 1997). The water vaporpressure of water bound by the paracasein islower than that of free water and thus has a lowerpropensity to evaporate during baking. The exu-

Figure 19-8 Pasteurized processed Cheddar cheeses exhibiting varying levels of flowability in grilled cheese-meat burgers.

Page 489: Cheese Science

App

aren

t vis

cosi

ty,

Pas

Ripening time, days

Figure 19-10 Typical changes during storage at 40C in the apparent viscosity of low-moisture Mozzarellacheese heated at 7O0C.

Ripening time, days

Figure 19-9 Typical changes during storage at 40C in the melt time of low-moisture Mozzarella cheese baked at28O0C for 4 min.

Mel

t tim

e, s

Page 490: Cheese Science

Ripening time, days

Figure 19-11 Typical changes during storage at 40C in the flowability of low-moisture Mozzarella cheese bakedat 28O0C for 4 min.

Flow

abili

y, %

Stre

tcha

bilit

y, c

m

Ripening time, days

Figure 19-12 Typical changes during storage at 40C in the Stretchability of low-moisture Mozzarella cheesebaked at 280 0C for 4 min.

Page 491: Cheese Science

Exp

ress

ible

ser

um,

g/kg

tota

l m

oist

ure

Ripening time, days

Figure 19-13 Typical changes during storage at 40C in the level of serum expressed from low-moisture Mozza-rella upon hydraulic pressing (3.2 MPa at 2O0C for 3 hr). A decrease in the level of expressible serum is an indexof increased water-binding capacity, or hydration, of the paracasein in the cheese.

Exp

ress

ible

fat

,g/

kg to

tal

fat

Ripening time, days

Figure 19-14 Typical changes during storage at 40C in the level of oil expressed from low-moisture Mozzarellaupon hydraulic pressing (3.2 MPa at 2O0C for 3 hr). An increase in expressible oil is an index of the increasedpotential of the cheese to oil-off during baking.

Page 492: Cheese Science

dation of free oil from the shredded cheese duringbaking also limits dehydration; the free oil formsan apolar surface layer, which impedes the escapeof water vapor. The changes in protein hydrationand free oil formation during storage are medi-ated by the following:

• The small increase in pH (from -5.15 be-fore plasticization to ~ 5.35 at 5 days). Theincrease in pH during plasticization is dueto the loss of lactic acid and soluble calciumphosphate in the water used to heat the curdduring plasticization. The loss of soluble Caresults in the solubilization of colloidal cal-cium phosphate in the curd during subse-quent cooling of the curd, to restore equilib-rium between the soluble and colloidalphases. The resulting phosphate anions(PO4

3-) scavenge free H+ from the aqueousphase and thereby reduce the [H+] (increasepH). Paracasein has maximum hydration ataround pH 5.35 (see Chapter 16).

• The increase in primary proteolysis (Figure19-15) of paracasein by residual rennetand/or plasmin, as reflected by the increasein pH 4.6 soluble N. Compared with othervarieties, such as Cheddar and Gouda,LMMC has a low level of proteolysis, ap-parently because of the relatively low levelof active rennet in the curd due to its ther-mal inactivation during plasticization (seeChapter 11). A comparative study of 12-week-old Cheddar, Gouda, and LMMCshowed that the latter had the highest levelof intact asi-casein and intermediate levelsof OC5I-CN G4-199 and y-caseins. The pres-ence of these breakdown products suggestscontributions from residual chymosin andplasmin to primary proteolysis in LMMC.

• The solubilization of casein-bound calciumthat occurs when the calcium attached tothe casein matrix is partially replaced bysodium. Conditions in the cheese conduciveto this ion exchange effect are 2.0-4.0%NaCl and 0.4% Ca in the serum phase(Kindstedt & Guo, 1997).

• The increased propensity of the cheese tooil-off upon cooking, which may be associ-

ated with the age-related degradation of thefat globule membrane and/or the continuedcoalescence of the partially denuded fatglobules after plasticization. The latter ef-fect may result from the hydration and con-comitant physical swelling of the proteinmatrix, an occurrence that physically forcesthe partially denuded globules into closeproximity.

During prolonged storage (up to 75 days), theunbaked cheese generally becomes too soft andsticky whereas the baked cheese becomes exces-sively flowable and "soupy" and lacks the de-sired chewiness, reflected by the relatively lowapparent viscosity. These changes in functional-ity are attributed to excessive proteolyis. How-ever, stretchability remains relatively constanteven when the product is stored for up to 4months at 40C, suggesting that the level of pri-mary proteolysis in the cheese at this time (i.e.,pH 4.6-soluble N « 120 g/kg total N) is insuffi-cient to significantly impair stretchability. In-deed, preliminary experiments show that youngCheddar cheese (i.e., 15-35 days, with a pH 4.6-soluble N level of < 120 g/kg total N) has goodstretchability, similar to that of LMMC (Guinee,unpublished results). In contrast, Cheddar (witha pH 4.6-soluble N level > 150 g/kg total N) hasinferior stretchability compared with LMMC.

Maillard browning on pizza pie results fromheat-induced reactions between the carbonylgroup of reducing sugars (lactose and galactose)and the amino groups of peptides and aminogroups. The degree of browning is related to thesugar-fermenting and proteolytic characteristicsof the starter culture used (Kindstedt, 1993).Most strains of Streptococcus thermophilus andLacobacillus delbrueckii spp. bulgaricus, whichare commonly used in the manufacture ofLMMC, are unable to metabolize galactose, andhence cheese made solely with such cultures issusceptible to browning. However, Lb. hel-veticus, which is frequently used as a componentof the starter culture, ferments lactose and galac-tose (resulting from the incomplete metabo-lization of lactose by the former cultures) com-pletely to lactic acid (see Chapter 5). Attempts to

Page 493: Cheese Science

control the level of browning in pizza cheese in-clude control of residual sugars and/or proteoly-sis products via

• adjustment of the ratio of Lb. helveticus toSc. thermophilus in the starter culture

• use of galactose-positive strains of Lb. del-brueckii spp. bulgaricus

• the use of proteinase-negative starter strainsto limit the formation of free amino groups

• the use of curd washing to remove lactosefrom the cheese curd

The propensity of LMMC to brown upon bak-ing changes markedly during ripening. Freshcheese curd (< 2-5 days) shows a high propen-sity to brown. Its propensity to brown decreasesmarkedly during the first few weeks of ripeningdue to the metabolism of lactose and/or galac-tose by galactose-fermenting starters but in-creases progressively thereafter owing to the ac-cumulation of small peptides and amino acids.

19.3.4 Cheese Composition and Proteolysis

The composition of cheese has a major effecton its functionality, as shown by inter- and intra-varietal correlations between composition andfunctionality. A recent study involving commer-cial samples of 21 different natural hard and

semi-hard cheese varieties showed severalacross-varietal correlations between cheesecomposition and functionality upon heating(Table 19-6). Stretchability was negatively cor-related with the levels of fat, FDM, and pH 4.6-soluble N and positively correlated with mois-ture content. In contrast, flowability wasnegatively correlated with the levels of moistureand moisture-in-non-fat-substance (MNFS). Asimilar across-varietal relationship betweenflowability and MNFS has been found for arange of hard and semi-hard cheeses. LikeStretchability, the apparent viscosity was nega-tively correlated with the level of pH 4.6-solubleN. Hence, Cheddar and other cheeses that hadhigher levels of pH 4.6-soluble N and relativelylower levels of moisture than pasta filata cheeseshad significantly higher flowability and lowerapparent viscosity and Stretchability than theformer.

The effects of intravarietal compositional dif-ferences on the functionality of LMMC havebeen reviewed by Kindstedt (1993). Increasingthe moisture content of LMMC results in lowerapparent viscosity and higher flowability. Re-ducing the level of fat in LMMC results in alower level of free oil and flowability upon bak-ing. Reducing the levels of calcium and phos-phate, by reducing the pH at coagulation, results

pH 4

.6 s

olub

le N

9

g/kg

tota

l N

Ripening time, days

Figure 19-15 Typical changes in pH 4.6-soluble N in low-moisture Mozzarella cheese during storage at 40C.

Page 494: Cheese Science

in higher flowability in LMMC made by directacidification, that is, by the addition of food-grade acid rather than acid development by astarter culture. Increases in the levels of primaryand secondary proteolysis in LMMC, throughthe use of a more proteolytic coagulant thanchymosin (e.g., C. parasitica proteinase), re-duces the apparent viscosity and increases freeoil and flowability. Increasing the salt content ofLMMC from 11 to 17.8 g/kg results in cheesethat is less stringy and less flowable and has ahigher apparent viscosity.

Increasing the fat content of Cheddar cheesefrom 50 to 330 g/kg is associated with shortermelt times, greater flowability, greater stretch-ability, and lower apparent viscosity.

19.3.5 Cheesemaking Conditions

The composition and functionality of LMMCduring ripening may be altered by varyingcheesemaking conditions and factors, including

• the type of coagulant• pH at whey drainage, which influences the

level of rennet retained in the curd

• temperature of the curd during cooking andstretching, which influences the stability ofrennets

• pH at stretching, which influences the levelof casein-bound calcium and its water-binding capacity

• salting method

Hence, Mozzarella cheese with customizedfunctionalities may be produced by varying thecheesemaking conditions. Other factors that in-fluence the composition and functionality ofLMMC include

• the stage of lactation and diet of the cows• homogenization of cheese milk• design and type of stretching equipment• whether a starter or direct acidification is

used

The composition and functionality of Mozza-rella acidified by the addition of acids is in-fluenced markedly by the calcium-chelatingproperties of the acids used and the pH andtemperature at coagulation. The latter param-eters affect the structure of the gel, its ability tosynerese following cutting or cooking, the level

Table 19-6 Relationships between Compositional and Functionality Variables for Different NaturalCommercial Cheese Types

Compositional Degrees of CorrelationFunctional Variable Variable Freedom Coefficient (r) Significance

Raw cheeseAggregation index pH 4.6SN 27 +0.36 p < 0.05

Cooked or baked cheeseStretchability (cm) Fat 36 -0.66 p < 0.001

FDM 36 -0.57 p< 0.001Moisture 36 +0.68 p < 0.001

pH4.6SN 34 -0.67 p< 0.001Flowability (%) MNFS 34 -0.53 p< 0.001

Moisture 32 -0.40 p < 0.02Melt time (s) pH 38 -0.45 p<0.05Apparent viscosity (Pa x s) pH 4.6 SN 28 -0.43 p < 0.02

Note: Correlation coefficients obtained by linear regression of the data. Data refer to cheeses in Tables 19-4 and 19-5.

Key: FDM = fat-in-dry-matter (g/kg); pH 4.6 SN = nitrogen soluble at pH 4.6 (g/100g total N); MNFS = moisture-in-non-fat-sub-stance (g/kg).

Page 495: Cheese Science

of casein-bound calcium, and the water-bindingcapacity of the casein matrix during subsequentstretching.

19.4 DRIED CHEESE PRODUCTS

Dehydrated cheese products are industriallyproduced cheese-based ingredients that were de-veloped for the U.S. Army during World War IIas a means of preserving cheese solids underconditions to which natural cheese would notnormally be subjected, such as temperatureabove 210C for long periods. Since then, theyhave become ingredients of major economic im-portance owing to their widespread use as fla-voring agents and/or nutritional supplements ina wide range of foods. These include bakeryproducts, biscuits, dehydrated salad dressings,sauces, snack coatings, soups, pasta dishes, sa-vory baby meals, cheese dips, au gratin potatoes,and ready-made dinners. They are also includedin processed and analogue cheese products asflavoring agents or as functional ingredients inpowdered instant cheese preparations, whichcan be reconstituted by the consumer for thepreparation of instant functional cheeses, such aspizza-type cheese, for domestic use. Advantagesover natural cheeses as an ingredient in theabove applications include these:

• Convenience of use in fabricated foods.Cheese powders can be easily applied to thesurface of snack foods, such as popcorn,potato crisps, and nachos, or easily incorpo-rated into food formulations by dry mixingwith other dry ingredients such as skimmilk powder (e.g., dried soup, sauce, andcake mixes) or blending into wet formula-tions. In contrast, natural cheeses requiresize reduction prior to their use in these ap-plications.

• Longer shelf-life because of their lower wa-ter activity (aw) than natural cheese. Thewater activity (aw) of natural cheeses rangesfrom about 0.99 for Quarg to 0.917 forParmesan (see Chapter 8), it ranges fromabout 0.93 to 0.97 for processed cheese

products, but it only ranges from 0.2 to 0.3for various dairy powders. Owing to theirrelatively high stability, cheese powdersmay be stored for a long period without al-teration or deterioration of quality. In con-trast, the changes that occur in naturalcheese during storage may influence itsprocessability (e.g., the ease with which itcan be size-reduced or its interaction withother ingredients) and its flavor profile andintensity. Hence, cheese powders are moreamenable to inventory management and setmanufacturing methods, and they yield endproducts with more consistent quality inlarge-scale manufacturing operations thannatural cheese.

• Greater diversity of flavor and functionalcharacteristics. The wide range of charac-teristics is made possible by the use of dif-ferent types of cheese, EMCs, and other in-gredients in the preparation of cheesepowders.

Dehydrated cheese products may be classifiedinto four categories, depending on the ingredi-ents used:

1. dried grated cheeses, such as Parmesanand Romano

2. natural cheese powders, made using natu-ral cheeses, emulsifying salts, and naturalcheese flavors (optional)

3. extended cheese powders, incorporatingnatural cheese and other ingredients suchas dairy ingredients (e.g., skim milk sol-ids, whey, lactose), starches, maltodex-trins, flavors, flavor enhancers, and/orcolors

4. dried EMCs

19.4.1 Dried Grated Cheeses

Dried grated cheeses are normally used ashighly flavored sprinklings, on pasta dishes, forexample, and in bakery products (e.g., biscuits).Essentially, the production of these products in-volves finely grating hard cheeses and drying theground cheese, usually in a fluidized bed drier

Page 496: Cheese Science

by exposure to low humidity air (15-20% rela-tive humidity) at an inlet temperature below3O0C. Under these conditions, the cheese is de-hydrated rapidly and evaporatively cooled,thereby reducing the risk of fat exudation andthe tendency to ball or clump. The dried, gratedcheese (typically containing 17% moisture) isusually pulverized and packed under nitrogen toreduce the risk of oxidative rancidity during dis-tribution and retailing.

Certain properties are required for the produc-tion of dried grated cheeses, such as relativelylow levels of moisture (300-340 g/kg) and fat-in-dry-matter (FDM; 390 g/kg), brittleness, andelastic fracture characteristics. These propertiesfoster efficient size reduction upon grinding,minimal susceptibility to fat exudation andsticking of the cheese particles, and efficientdrying to a homogeneous product free ofclumps. An intense cheese flavor is also a desir-able characteristic. Generally, dried gratedcheeses are used in small quantities, as sprin-klings, to impart strong cheese notes to pastadishes, soups, and casseroles. The cheeses thatbest meet these criteria are Parmesan and Ro-mano, because of their composition; fractureproperties; and strong, piquant, lipolyzed flavor.The flavor of Romano-type cheeses resultsmainly from the addition of pregastric esterase(from kid, goat, or lamb) to the cheese milk,which preferentially hydrolyzes the short-chainfatty acids (especially butanoic acid) from milkfat triglycerides during maturation. A high levelof butanoic acid (1,500-2,000 mg/kg cheese) isresponsible for the peppery, piquant flavor ofRomano cheese.

Owing to their lower firmness and higher lev-els of moisture and FDM, cheeses such as ma-ture Cheddar (moisture ~ 370 g/kg and FDM ~520 g/kg) or Gouda (moisture -410 g/kg andFDM -480 g/kg) are unsuitable for drying.These characteristics render the cheese suscep-tible to fat exudation and clumping during grind-ing and drying. However, they can be dried ifthey are first shredded and blended withParmesan- or Romano-type cheeses beforegrinding. The moisture content of dehydrated

grated cheese may be reduced further by usingthe Sander's drying process. The grated cheesepowder is placed on trays that are conveyedthrough a drying tunnel, where it is exposed tohot air. The hot air heats the cheese particles to630C and reduces the moisture content from 170g/kg to about 35 g/kg. High moisture cheeses(820 g/kg), such as Cottage cheese, may also bedried directly to a 30-^40 g/kg moisture level byfirst pulverizing and then drying them in special-ized spray driers (e.g., silo spray drying usingthe Birs Dehydration Process; Kosikowski &Mistry, 1997). These low-moisture, dried natu-ral cheeses are generally used for nutritionalsupplementation of foods (e.g., dried babymeals).

19.4.2 Cheese Powders

Manufacture

The manufacture of cheese powders essen-tially involves the production of a pasteurizedprocessed cheese slurry (400^50 g/kg dry mat-ter), which is then spray-dried (Figure 19-16).The production steps include formulation of theblend, processing of the blend to form a slurry,homogenization, and drying of the slurry.

The blend usually consists of comminutednatural cheese, water, emulsifying salts, flavor-ing agents, flavor enhancers, colors, antioxi-dants, and perhaps filling materials, such aswhey or skim milk solids, starches, maltodex-trins, and milk fat. The type of cheese powder(e.g., natural or extended), flavor required, andapplication (e.g., whether intended for use in asauce, soup, snack coating, or cheese dip) deter-mine the type of ingredients included. Antioxi-dants, such as propyl gallate and butylated hy-droxyanisole, may be added at a level of 0.5-1.0g/kg fat to retard oxidative rancidity. Typicalformulations of the slurries required for the pro-duction of natural and extended cheese powderswith different levels of cheese solids are given inTable 19-7.

The flavor profile and intensity of the finalcheese powder is determined by the types of

Page 497: Cheese Science

Figure 19-16 Production process for cheese powder.

Cheese powder(> 960 g/kg dry matter)

Spray dry

Atomization

Homogenization

Hot molten slurry(350-450 g/kg dry matter;

75-85 0C)

Blend and shear

Dissolving tank(cooker)

Steam

Formulation

Flavour enhancers:NaClmonosodium glutamateautolysed yeast extract

ColoursAntioxidantsFree flow agents.

Cheese

ComminutionWater

Milk solids:wheybuttermilkskim milkcaseinate

MaltodextrinStarches

Emulsifyingsalts

Flavours:EMCNature identical

Page 498: Cheese Science

cheese used and the types and levels of other fla-voring agents (e.g., EMC, hydrolyzed milk fat,and starter distillate) and flavor enhancers (e.g.,NaCl, monosodium glutamate, and autolyzedyeast extract). Generally, mature cheese with anintense flavor is used so as to impart a strong fla-vor to the final product. Apart from their lack offlavor-imparting properties, young cheeses witha high level of intact casein are unsuitable, asthey result in slurries that are very viscous andare difficult to atomize and dry efficiently. Fill-ing materials in extended cheese powders areusually added to replace cheese solids and re-duce formulation costs. However, they may in-fluence flavor, wettability, and mouth-coatingcharacteristics of the product in which thecheese powder is used.

Processing principles and technology aresimilar to those used for the manufacture of pro-cessed cheese products. Processing involvesheating the blend to around 8O0C in a processedcheese-type cooker using direct steam injectionor in large "dissolving tanks" (e.g., 5,000 L)containing shearing blades using indirect steaminjection and continuous shearing at 1,500-3,000 rpm. The blend is worked until the hotmolten slurry is homogeneous in color and con-sistency and free of lumps or nonhydrated mate-rial. The maximum processing temperature

should be maintained below 850C to minimizethe loss of volatile flavor compounds in the dis-sipating steam and to minimize the risk ofbrowning, especially for formulations contain-ing a high level of lactose or high dextrose (glu-cose) equivalent (DE) maltodextrins.

The viscosity of the cheese slurry has a majorinfluence on its tendency to foam and thereforeon the level of air in the resultant powder. Owingto its effects on the air content of the powder, theviscosity of the cheese slurry influences thephysical characteristics (bulk density and wet-tability) of the resultant powder and its suscepti-bility to oxidative rancidity and flavor deteriora-tion during storage. High-viscosity slurries (>3.0 Pa x s) have a lower propensity to foam andtherefore yield a powder with a lower level of airthan low-viscosity slurries (< 0.3 Pa x s). Theviscosity of the cheese slurry is determined byits dry matter content and the characteristics ofits ingredients, such as density of the differentingredients, levels of fat and protein, pH, anddegree of ingredient hydration. The air contentof the cheese powder is also influenced by thelevels of formulation ingredients that tend topromote foaming of the slurry during prepara-tion and drying (undenatured whey proteins) orto depress such foaming (fat and food-gradeantifoaming agents).

Table 19-7 Typical Formulations of Cheese Slurries for the Production of Cheese Powder with Differ-ent Levels of Cheese Solids

Loiv Medium High(262g/kg) (530 g/kg) (950 g/kg)

Medium aged Cheddar - 190 -Mature Cheddar 200 170 635EMC paste 5 2 10EMC powder 5 20 -Whey powder 120 50Skim milk powder 80 38 -Maltodextrin (DE 17) 165 110Emulsifying salts 15 25 15Butylated hydroxyanisole (BHA) 0.5 0.5 0.5Sodium chloride 15 10 5Water and condensate 390 385 335

Key: EMC = enzyme-modified cheese.

Page 499: Cheese Science

Homogenization of the slurry is optional butis commonly practiced to ensure homogeneity ofthe slurry. The pressures applied (typically 15and 5 MPa in the first and second stages, respec-tively) have a major effect on the viscosity of theslurry, with higher pressures generally impartinghigher viscosity for a given level of dry matter.

Several spray-drying processes (e.g., singlestage or two stage) and dryer configurations(e.g., tall-form, filtermat, silo-form) may beused. The design and operation of the dryer (e.g.,atomizer type) and the pressure, direction of airflow, air inlet and outlet temperatures, and airhumidity influence the physical characteristics(e.g., bulk density, wettability, and solubility)and the flavor characteristics of the cheese pow-der. The physical properties are important in ap-plications that require reconstitution of thecheese powder, such as ready-made soups,sauces, and baby foods.

In all cases, the homogenized cheese slurry ispumped to the dryer, where it is atomized anddried, typically at an inlet air temperature of180-20O0C and an outlet air temperature of 85-9O0C, depending on the type of dryer. The pow-der is then cooled from about 550C to 2O0C,separated from the air, and packaged. The mois-ture content of the dried powder is typically 30-40 g/kg and generally decreases with increasingoutlet air temperature. However, an elevatedoutlet temperature, above 950C, for example,may be detrimental to product quality owing to

• increased Maillard browning• reduced product wettability and solubility

(due to denaturation of ingredients)• loss of volatile flavor compounds• greater susceptibility to oiling-off and free-

fat formation (free fat in the cheese powderleads to lumpiness, flow problems, and fla-vor deterioration)

Commercially, cheese powders are normallymanufactured using two-stage drying systems.Filtermat (box) dryers are used frequently in theUnited States, whereas tall-form dryers with anintegrated fluidized bed are widely used in Eu-rope. While the operating conditions of thesedryers influence the quality of the cheese pow-

der, the tall-form dryer is generally consideredto give better flavor retention, larger powderparticles, and better powder flowability. Con-ventional single-stage tall-form dryers are rarelyused because of the high outlet air temperature(e.g., > 950C) necessary to achieve the lowmoisture content required. However, single-stage silo dryers (with a 70 m drying tower,compared with the 10m tower used in tall-formdryers) may be used, as in the Birs DehydrationProcess. In this process, the drying air is dehu-midified but not heated. The main advantagesover conventional two-stage drying are im-proved color stability and enhanced flavor reten-tion, especially in mildly flavored products, theflavor of which is dominated by a few com-pounds (e.g., Cottage cheese).

Composition

The composition of cheese powders variesconsiderably, depending on the formulation in-gredients; typical values are shown in Table19-8.

Applications

Cheese powders are generally used as flavor-ing ingredients in a wide variety of foods, espe-cially snack coatings (e.g., popcorn, nachos, tor-tilla shells), cheese sauces, soups, savorydressings, and savory biscuits. In snack foods,the powder is dusted after the snack has beensprayed lightly with vegetable oil. In cheesesauces, the level of cheese powder is typically50-100 g/kg, depending on the flavor intensityof the cheese powder and the types and levels ofother flavoring ingredients in the formulation.Generally, at these levels, the cheese powder haslittle influence on the rheological properties ofthe sauces, which are controlled mainly by thetypes and levels of starchy materials used(Guinee, O'Brien, & Rawle, 1994).

19.4.3 Enzyme-Modified Cheeses

Enzyme-modified cheeses (EMCs) are usedprincipally as flavoring agents in industriallyproduced cheese products and ingredients, suchas pasteurized processed cheese products,

Page 500: Cheese Science

cheese substitutes or imitations, cheese pow-ders, and ready-made meals. Natural cheeseshave certain limitations as flavor ingredients:

• low flavor stability due to ongoing bio-chemical and microbiological changes dur-ing storage

• flavor inconsistency, due to changes, forexample, in cheese composition

• insufficient flavor intensity (large quanti-ties are required to impart a strong cheeseflavor)

• high cost, due to the relatively long ripeningtime required for most cheese varieties

• the need to comminute cheese prior to itsincorporation into foods and the fact thatcomminuted cheese is not suited to the bak-ery and snack food industries, which arelarge users of cheese

These limitations led to the development, in the1960s, of enzyme-modified cheeses with flavors5-20 times more intense than those of the corre-sponding natural cheeses.

The production of EMCs generally involvesthe following production steps (Figure 19-17):

• Production of a cheese curd, as for conven-tional cheese.

• Formation of a paste (typically 400-500g/kg dry matter) by blending the curd withwater and emulsifying salts. The addition ofemulsifying salts assists in adjusting the pHof the paste to a value that is optimal forsubsequent enzymatic reactions.

• Pasteurization of the cheese paste to inacti-vate the cheese microflora and enzymaticactivities. Pasteurization reduces the risk offlavor inconsistencies due to variations instrain composition and populations ofstarter and nonstarter lactic acid bacteriaand variations in enzyme (e.g., residualchymosin) activity in curd obtained fromdifferent suppliers.

• Treatment of the pasteurized curd with thedesired cocktail of enzymes and perhapsstarters to produce the required flavor pro-file and intensity. Added enzymes may in-clude proteinases, peptidases, and lipaseschosen based on knowledge of the enzy-mology and flavor profile of the cheese be-ing simulated. Some commercial EMCs areprepared using a combination of added en-zymes and bacterial culture systems. Theadvantage of using starter culture systems isthat each starter cell is essentially a sack ofenzymes that are known to contribute tobalanced flavor production in any givencheese variety. Hence, it is generally easierto simulate a particular cheese flavor by us-ing cultures than enzyme cocktails.

• Incubation of the slurry at 30-4O0C for24-72 hr. During this period, the added en-zymes, or those released from starter cellsduring growth and/or autolysis, act on thecasein and fat in the paste to produce thecorrect balance of peptides, amino acids,amines, aldehydes, ammonia, fatty acids,ketones, and alcohols.

Table 19-8 Composition of Cheese Powders with Different Levels of Cheese Solids

Low (262 g/kg) Medium (530 g/kg) High (950 g/kg)

Dry matter (g/kg) 970 970 960Protein (g/kg) 201 230 361Fat (g/kg) 145 219 388Lactose (g/kg) 264 123 3Ash (g/kg) 89 104 104pH 6.35 6.50 6.3

Page 501: Cheese Science

Figure 19—17 Production process for enzyme-modified cheese.

EMC powdei

Spray dry

EMC paste

Pasteurization(e.g., 72 0C for 4-8 min)

Incubation(24-72 h)

cool(30-40 0C)

Enzymes:proteinasespeptidasesUpases

Starter culture,culture adjunct,Attenuated starter culture^

Water

Pasteurize and shear(e.g., 720C for 4-8 min)

Shred

Fresh Cheese Curd

Conventional Cheesemaking Process

Whey

Rennet Starter culture

Standardized Cheese Milk

Emulsifyingsalts

Page 502: Cheese Science

• Pasteurization of the enzyme-treated pasteto inactivate enzymes and thereby preservethe flavor characteristics generated withminimum change during storage.

• Homogenization of the hot pasteurizedpaste to reduce the risk of phase separationduring storage and ensure product homoge-neity. The homogenized cheese paste,known as EMC paste, may be packaged andstored at refrigeration temperatures, usuallyin opaque materials to minimize the risk ofoxidative rancidity and off-flavor develop-ment.

• Drying. The paste may be dried to producean EMC powder, which has a longer shelf-life than the paste and is better suited forapplications involving dry-blending withother ingredients.

EMC variants of many natural cheeses, suchas Cheddar, Blue cheese, Romano, and Em-mental, are commercially available (seeKilcawley, Wilkinson, & Fox, 1998). Little in-formation has been published on the productionof EMCs, as most of it is proprietary andguarded by the manufacturers. The developmentof EMCs requires elucidation of the flavor com-pounds and/or enzyme activities in the cheesebeing simulated, followed by testing of the vari-ous enzyme cocktails under various conditions(e.g., various levels of paste dry matter, pH, andincubation temperature) until the desired flavorcharacteristics are obtained. The composition,flavor-forming reactions, and flavor compo-nents of EMCs have been reviewed extensively(see Kilcawley et al., 1998).

19.5 CONCLUSION

Cheese is a highly versatile dairy ingredientthat can be used directly in an array of culinarydishes, formulated food products, and ready-made meals. In these applications, added cheeseperforms a number of functions; it contributes tostructure, texture, flavor, mouth-feel, melt prop-erties, and nutrition. However, natural cheese isan expensive ingredient, may be functionally

unstable, and may have somewhat variable func-tionality. Moreover, the consistency of naturalcheese is unsuitable for some applications, suchas those involving dry-blending with other in-gredients in the manufacture of cake mixes,dried soups, and baby meals. Hence, cheese maybe dried in the original or an extended form toyield cheese powder or be used as a substrate forthe development of EMCs.

Little information is available on the quanti-ties of natural cheese and cheese products con-sumed as ingredients in other foods. Sutherland(1991) reported that about 30% of Australiancheese is used by the food industry and food ser-vice sectors and predicted that the consumptionof cheese as an ingredient of other foods willgrow rapidly because of the greater demand forprepared meals. The Australian trend probablyreflects trends in the United States and Europe.In addition, cheese is used as an ingredient in thehome. Indeed, it is estimated that about 25% oftotal cheese consumed is incorporated into vari-ous homemade dishes and that, in developedcountries, about 50% of cheese is used as an in-gredient by home cooks and by the food indus-try. It is envisaged that the industrial use ofcheese as an ingredient will become a majordriving force in increasing the per capita con-sumption of cheese in developed countries,where the demand for convenience and preparedfoods is growing.

In cheese-containing prepared foods, thecheese is expected to exhibit functional charac-teristics such as flowability, mouth-feel, flavor,and stretchability. As the production and con-sumption of "fast foods" and ready-made mealsgrow, the demand for greater functionality andfor customized cheeses is increasing. Function-ality is a major factor contributing to the in-crease in cheese consumption, which is clearlyreflected by the recent dramatic growth in theconsumption of pizza cheese, especially in theUnited States, where current annual productionis about 1 million tonnes (Lavoie & Mertz,1995; Sorensen, 1997). Production of pizzacheese is also increasing in Europe (100,000tonnes per annum) and New Zealand and Aus-

Page 503: Cheese Science

tralia (100,000 tonnes per annum), where con-sumption is following the number of outlets ofthe major pizzeria chains. Moreover, it is be-coming apparent that the taste and functionalityexpected of pizza cheese show regional varia-

REFERENCES

Guinee, T.P., Mulholland, E.G., Mullins, C., & Corcoran,M.O. (1997). Functionality of low moisture Mozzarellacheese during ripening. In T.M. Cogan, P.F. Fox, & R.P.Ross (Eds.), Proceedings of the Fifth Cheese Symposium.Dublin: Teagasc.

Guinee, T.P., O'Brien, N., & Rawle, D.F. (1994). The vis-cosity of cheese sauces with different starch systems andcheese powders. Journal of the Society of Dairy Technol-ogy, 47, 132-138.

Kilcawley, K.N., Wilkinson, M.G., & Fox. P.P. (1998). En-zyme-modified cheese [A review]. International DairyJournal, 8, 1-10.

Kindstedt, P.S. (1993). Effect of manufacturing factors,composition, and proteolysis on the functional character-istics of Mozzarella cheese. Critical Reviews in Food Sci-ence and Nutrition, 32, 167-187.

Kindstedt, P.S. (1995). Factors affecting the functional char-acteristics of unmelted and melted Mozzarella cheese. In

tions. Hence, it is conceivable that as pizzeriachains expand into new markets, there will be ademand for cheeses with different flavors andfunctionality than those currently required in or-der to meet local taste preferences.

E.L. Malin & M.H. Tunick (Eds.), Chemistry of struc-ture-function relationships in cheese. New York: PlenumPress.

Kindstedt, P.S., & Guo, M.R. (1997). Recent developmentsin the science and technology of pizza cheese. AustralianJournal of Dairy Technology, 52, 41- 13.

Kosikowski, F.V., & Mistry, V. V. (1997). Drying and freez-ing of cheese. In Cheese and fermented milk foods: Vol. 1.Origins and principles. Westport, CT: F. V. Kosikowski,LLC.

Lavoie, C., & Mertz, T. (1995). The U.S. market for pizza[Report]. Commack, NY: Business Trend Analysts, Inc.

Ridgway, J. (1986). The complete cheese book. London:Judy Piatkus Ltd.

S0rensen, H.H. (1997). The world market for cheese [Bulle-tin No. 326]. Brussels: International Dairy Federation.

Sutherland, BJ. (1991). New cheese products as food ingre-dients. Food Research Quarterly, 51(1, 2), 114-119.

Page 504: Cheese Science

20.1 INTRODUCTION

Milk is a highly nutritious medium of almostneutral pH and, therefore, many bacteria, includ-ing spoilage and pathogenic bacteria, can growin it. Numerous outbreaks of food poisoninghave been traced to milk. Although cheese isequally nutritious, it has been shown to be re-sponsible for relatively few food-poisoning out-breaks. These are summarized in Table 20-1.There were 21 confirmed outbreaks of food poi-soning in Western Europe during the years1970-1997, 7 outbreaks in the United Statesfrom 1948 to 1997, and 4 outbreaks in Canadafrom 1970 to 1997 due to consumption of cheese("Food Safety and Cheese," 1998; Johnson,Nelson, & Johnson, 199Oa, 199Ob, 199Oc). Dur-ing the period 1970-1997, an estimated235,000,000 tonnes of cheese were produced inWestern Europe, the United States, and Canada.Such data imply that cheese is a relatively safefood product, but 28% of the outbreaks involvedcheeses made from raw milk.

Several organisms were involved in cheese-related outbreaks of food poisoning, but Salmo-nella spp., Staphylococcus aureus, and Listeriamonocytogenes were the most common (Table20-1). The species of organism involved in theEuropean outbreaks were more diverse thanthose in the United States, which may reflect thegreater diversity of cheese varieties made in Eu-rope compared with the United States and

Canada. The primary reasons for these cheese-related food-poisoning outbreaks were poorstarter activity (due to phage, antibiotic residuesin the milk, etc.), poor hygiene in the plant, grossenvironmental contamination, and faulty pas-teurization.

In recent years, the most important pathogensfound in cheese have been L. monocytogenesand enteropathogenic Escherichia coll. Theformer is the more important since the outbreaksinvolved several deaths. The outbreaks due to E.coll 0157 involved cheese made from raw milk,and only a few cases were involved. Neverthe-less, a major outbreak due to E, coli 0157 hasthe potential to be very serious. The majorcheeses involved were soft surface-ripenedcheese and cheese with a low acidity (e.g., Mexi-can-style cheese). There is no indication thatMycobacterium bovis and Brucella abortus andthe so-called emerging pathogens, Campylo-bacter jejunii, Yersinia entercolitica, andAeromonas hydrophilia grow during manufac-ture of cheese. M. bovis, which causes tubercu-losis in cows and sometimes in humans, and B.abortus, which causes abortion in cows and un-dulant fever in humans, are sometimes excretedin the milk of infected cows. These microorgan-isms were probably important causes of humandisease in the past but are not important now,since TB and brucellosis are controlled in almostall dairy herds in developed countries. An out-break of food poisoning in Canada involving

Pathogens and Food-PoisoningBacteria in Cheese

CHAPTER 20

Page 505: Cheese Science

Table 20-1 Food-Poisoning Outbreaks Associated with Cheese in the United States (1948-1997), Canada (1970-1997), and Europe (1970-1997)

ReferenceCausative

Organism(s)No. ofDeaths

No. ofCasesYear

Varietyof Cheese

Countryof Origin

Allen and Stovall, 1960Hendricksetal., 1959Zehren and Zehren, 1968

Fontaine etal., 1980Sharpe, 1987Linnan etal., 1988Hedberg etal., 1992

Eckman, 1975

Toddetal., 1979, 1981bToddetal., 1981 aSharpe, 1987D'Aoustetal., 1985; D'Aoust, 1994Sharpe, 1987Marieretal., 1973

MacDonaldetal., 1985Desenclos etal., 1996Gouletetal., 1995Public Health Laboratory Service,

London, 1994

Kauf et al., 1974; Sebald et al.,1974

Sharpe, 1987

Maguire etal., 1992continues

Staph. aureusStaph. aureusStaph. aureus

SalmonellaSc. zooepidemicusL. monocytogenesSalmonella javiana;Salmonella oranienberg

Brucella

Staph. aureusStaph. aureusSalmonellaSalmonella typhimuriumSalmonellaE. co//O124:B17

E. co//O27:H20Salmonella paratyphi BListeria monocytogenesVeratoxic E. coll

Cl. botulinum

Brucella

Salmonella dublin

48O

1

O

141

O

O

200200

>33916

142164

3

1262

>2700

35387

>300027320NR

77

23

42

195819581965

1976198319851989

1975

1977198019821984

19831971

1983199319951992

1974

1983

1989

ColbyCheddarCheddar, Kuninost, and

MontereyCheddarHomemadeMexican styleMozzarella

Mexican style3

EmmentalCheese curdCheddar, other typesCheddar

FarmhouseCamembert, Brie,

Coulommiersa

Brieb

Goat milk cheese0

Brie de Meauxc

Fromage fraisc

Soft cheese

Homemade, unripened

Soft cheese0

USUSUS

USUSUSUS

Mexico

CanadaCanadaCanadaCanada

FinlandFrance

FranceFranceFranceFrance

France, Switzerland

Greece

Ireland

Page 506: Cheese Science

Table 20-1 continued

ReferenceCausative

Organism(s)No. ofDeaths

No. ofCasesYear

Varietyof Cheese

Countryof Origin

Aureli etal., 1996

Public Health Laboratory Service,London, 1995

Sharpe, 1987

BiIIe, 1990; Malinverni et al., 1985

SadiketaL, 1986Valliant et al., 1996

Sharpe, 1987Public Health Laboratory Service,

London, 1997aBone etal., 1989Public Health Laboratory Service,

Edinburgh, 1994Public Health Laboratory Service,

London, 1997b

C. botulinum

Br. melitensis

Bacillus; Shigella sonnei

L monocytogenes

Salmonella typhimuriumSalmonella dublin

Staph. aureusSalmonella gold-coast

Staph. aureusE co// O1 57

E. co// O1 57

1

1

34

O5

O

OO

O

8

135

>50

122

>4025

2>84

>13>20

2

1996

1995

1982

1983/87

19851995

19831994

19841992

1997

Mascarpone

Soft cheese

Brie

Vacherin Mont d'Or0

Vacherin Mont d'Or°Doubsc

CheddarCheddar

Sheep milk cheese0

Farmhouse cheese

Lancashire0

Italy

Malta

Scandinavia

Switzerland

Switzerland, FranceSwitzerland, France

UKUK

UKUK

UK

a Consumed in the United States.b Consumed in the United States, Sweden, and The Netherlands.c These cheeses were made from raw milk.

Page 507: Cheese Science

Cheddar cheese made from raw milk was tracedto a farm where one cow was shedding about200 cfu of salmonella per milliliter of milk.

The infective dose of pathogens can be low.For instance, S. heidelberg has an infective doseof 100-500 cells, and the infective dose for E.coli O157:H7 is thought to be 10 cells. In con-trast, the infective dose for S. aureus is high be-cause the actual cause of the food poisoning isnot the organism itself but a number of closelyrelated, heat-stable protein toxins (able to with-stand 10O0C for > 30 min) produced by it. Thesetoxins are called staphylococcal enterotoxin A(SEA), SEB, SEC, etc., and some of them (e.g.,SEC) are further subdivided into SECl, SEC2,SEC3, etc. The amino acid sequences of all theenterotoxins are very similar. Some of them(e.g., SEA and SEE) are produced during expo-nential growth, while others (e.g., SEB andSEC) are produced mainly during the stationaryphase of growth. Those produced in the expo-nential phase are the more common causes ofstaphylococcal food poisoning. The minimumnumber of staphylococci and the level of toxinrequired to cause food poisoning are thought tobe 105 cells/g and 1 ng/g of food ingested. Thestrains of S. aureus present in raw milk are pri-marily those that cause mastitis, and about 20%of these strains produce enterotoxin.

20.2 PATHOGENS IN RAW MILK

Recent surveys of raw milk quality show thatthe incidence of pathogens in raw milk is low.For example, about 5% of individual Irish, En-glish, and French farm milks are contaminatedwith Listeria and less than 1 % with Salmonella(Desmasures, Bazin, & Gueguen, 1977; Minis-try of Agriculture, Fisheries and Food, 1997;O'Donnell, 1995; Rea, Cogan, & Tobin, 1992).The microbiological quality of farm milks inNormandy (France), around the area where rawmilk Camembert is made, is generally verygood; 83% of 69 samples had total bacterialcounts below 20,000 cfu/ml, and the sampleshad an average somatic cell count of 176,000/ml(Table 20-2). The average number of coliforms,enterococci, and S. aureus was 77, 79, and 350per milliliter, respectively, in summer milk.Winter-produced milk was also good, withcounts of 57, 74, and 450 per milliliter for theabove bacterial groups, respectively. The inci-dence of Yersinia entercolitica was relativelyhigh and that of Campylobacter low. All thesedata suggest that hygiene and cooling of milkwere effective, although 14 of the 43 milks ex-amined did not meet the European Union crite-rion for S. aureus in milk destined for cheese-making (< 500/ml) (Table 20-3).

Table 20-2 Microbiological Quality (cfu/ml) of Milk Produced in Normandy, France

n Winter n Spring/Summer

Total count 39 71,000 ± 27,000 30 86,000 ±21,000Enterococci 37 74 ± 150 25 79 ± 400Coliforms 29 57 ±2,400 19 77 ±5,000S. aureus 25 450 ±1,700 18 350 ±280L monocytogenes 39 4 positive samples 30 No positive samplesSalmonella 39 1 positive sample 30 1 positive sampleY. entercolitica* 39 19 positive samples 30 6 positive samplesCampylobacter 39 1 positive sample 30 No positive samples

3OnIy 1 of 61 isolates was a potential pathogen.

Page 508: Cheese Science

20.3 PATHOGENSINCHEESE

Soft cheeses contain high levels of moisture,and the pH, particularly at the surface, increasesduring ripening. In addition, some of them aremade from raw milk, and smear cheeses are fre-

quently handled during ripening. Consequently,soft cheeses are more prone to microbial growththan hard or semi-hard varieties. A recent surveyof the microbiological quality of soft cheeses inthe United Kingdom market is summarized inTable 20-4. Cheeses from several countries, in-

Table 20-3 European Union Guidelines (Standards) for Milk for Cheesemaking and Cheese

Soft Cheese Soft Cheesefrom Raw Milk from Pasteurized Milk Raw Milk

Listeria Absent in 25 g* Absent in 25 gm 5 5c O O

Salmonella Absent in 25g Absent in 25 gm 5 5c O O

S. aureusmt 1,000/g 100/g 500/ml*M 10,000/g 1000/g 2,000/mln 5 5 5c 2 2 2

Conformsm 10,000/gM 100,000/gn 5c 2

E. collm 10,000/g 100/gM 100,000/g 1,000/gn 5 5c 2 2

Total plate count <100,000/ml§Somatic cells <400,000/ml

Key: n = number of samples, m = threshold value; the result is satisfactory if the number of bacteria in all sample units does notexceed m. M = maximum value; the result is unsatisfactory if the number of bacteria in > 1 sample units exceeds M. c = number ofsamples for which the counts may lie between m and M; the sample is considered acceptable if the numbers are < m in the othersample units.

* The 25 g sample should consist of five 5 g portions taken from different parts of the same product.

t For products labeled "made from raw milk."

* Geometric mean of 2 results, with > 2 samples per unit.

§ Geometric mean of 3 results, with > 1 sample per unit.

Page 509: Cheese Science

eluding Scandinavia, France, Germany, Greece,Cyprus, Italy, and the United Kingdom, wereanalyzed. Salmonella were not found in any ofthe 1,437 cheeses examined. Overall, the micro-biological quality of both raw and pasteurizedmilk cheese was quite good. One (1.4%) of theraw milk cheeses failed to meet the EuropeanUnion criteria (Table 20-3) for Listeria in rawmilk cheese, and 2 (3%) failed to meet the crite-ria for E. coli. There are no standards for coli-form numbers in raw milk cheese, but 28 of the405 (7%) samples of pasteurized milk cheeseexamined failed the coliform standard, and 7(2%) failed the Listeria standard (Table 20-3).Coliforms are killed by pasteurization, and thehigh numbers in some of the pasteurized milkcheeses probably reflect the spread of contami-nation during the smearing of surface-ripenedcheese (see Chapter 10).

20.4 LISTERIOSIS

Listeriosis is caused by L. monocytogenes, es-pecially serotypes l/2a and 4b, and it affectsmainly pregnant women, immunocompromisedpeople (e.g., those who are HIV positive or arerecovering from chemotherapy after treatmentfor cancer), and the elderly. Prominent symp-toms include vomiting and diarrhea, which maylead to meningitis and bacteremia. Infection ofthe blood stream, the central nervous system, thefetus in utero, and infants by mothers who showno obvious signs of infection during birth are themost common forms of listeriosis.

Two relatively recent major outbreaks of lis-terosis have been traced to cheese, one in theUnited States, involving Mexican-style cheese,and one in Switzerland, involving VacherinMont d'Or cheese, which is a soft cheese made

Table 20-4 Microbiological Quality of Soft Cheese Made from Raw and Pasteurized Milk

10 to 102 to 103 to 104 toND* <1& <1& <1& <704 <1& >105

Raw milk cheeses (72)Coliforms 32 8 5 3 9 2 13E.co// 4 8 7 1 0 4 1 0 2S . aureus 7 0 O 1 1 0 0 0L . monocytogenes 7 1 1 O O O O OOther Listeria spp. 6 8 3 O 0 1 0 0

Pasteurized Milk Cheeses (405)Coliforms 284 13 38 19 23 9 19E . coli 3 8 3 7 9 4 1 0 2S . aureus 4 0 1 O 2 O 1 1 0L . monocytogenes 4 0 3 2 0 0 0 0 0Other Listeria spp. 4 0 0 5 0 0 0 0 0

Unlabeled cheeses (960)Coliforms 611 37 71 54 96 30 61E . coli 8 9 1 2 2 2 5 9 6 2 5S . aureus 9 5 8 O O 1 1 0 0L . monocytogenes 9 4 7 13 O O O O OOther Listeria spp. 9 2 6 3 0 O 1 1 0 2

* Not detected.

+Total per gram.

Page 510: Cheese Science

from raw milk. Both outbreaks involved fatali-ties, 48 in the United States (20 fetuses, 10 in-fants, and 18 nonpregnant adults) and 34 inSwitzerland (Table 20-1). Poor hygiene was themajor factor in both outbreaks, and improperpasteurization was also implicated in the case ofthe Mexican-style cheese. However, the fact thatboth cheeses also had low salt levels and that theMexican-styIe product is a low-acid cheese,made without the deliberate addition of startercultures, while Vacherin is a surface-ripened va-riety, in which the pH increases during ripening,contributed to the outbreaks.

L. monocytogenes is a Gram-positive rod thatcan grow at temperatures from -0.40C to 450C,at pH values of 4.4 to 9.4, and in the presence of10% NaCl. Generation times at O0C and I0C are131 and 62 hr, respectively, and the lag times are33 and 3 days, respectively. Therefore, holdingcheese as close to O0C as possible will help toprevent the growth of listeria. The optimum pHand temperature are 7.0 and 370C, respectively.These properties make this microorganism par-ticularly problematical. It is an ubiquitous or-ganism and is found in soil, water, silage, and soon. Sometimes the organism is found insidephagocytes (neutrophils and macrophages) inmilk, and this location was thought to protect thecells from inactivation during pasteurization.The general consensus now, however, is that theorganism is inactivated by pasteurization whe-ther the cells are inside phagocytes or not.

20.5 PATHOGENIC ESCHERICHIA COLI

The normal habitat of E. coll is animal (andhuman) feces whence it can contaminate rawmilk, particularly if the animals have been lyingin their own dung and the udders have not beenproperly washed before milking. Most strains ofE. coli are harmless, commensal organisms, butsome are pathogenic. They are differentiatedfrom each other on the basis of the serologicaldetection of somatic (O), flagellae (H), and cap-sular (K) antigens. To date, 174 O antigens, 56 Hantigens, and 80 K antigens have been detected.Pathogenic strains are generally subdivided into

enteropathogenic (EPEC), enterotoxigenic(ETEC), enteroinvasive (EIEC), and enterohem-orrhagic (EHEC), depending on how they causeinfection.

Five outbreaks of foodborne disease due topathogenic E. coli have been traced to the con-sumption of soft cheeses. These involved ETECO27:H20, EIEC O124:B17, and EHEC O157(Table 20-1). The outbreak due to E. coliO124:B17 occurred in the United States but in-volved French Camembert, Brie, and Coulom-miers cheese made in the same plant over a 2-day period. Despite this, the cheese was widelydistributed, because outbreaks occurred in 14states from Connecticut in the east to Californiain the west; 389 people developed food poison-ing, but no deaths were reported. In the outbreakinvolving E. coli O27:H20, the cheese involvedwas from two different lots made 46 days apart,suggesting that contamination was intermittent.It is not clear if the cheeses in these outbreakswere made from raw or pasteurized milk. E. coliO157:H7 is an emerging food pathogen that hasbeen associated with several severe food-poi-soning outbreaks involving meat products andalso a small outbreak involving a raw milkcheese (Table 20-1).

The symptoms of the E. coli 0124 infectionincluded diarrhea, fever, and nausea; cramps,chills, vomiting, aches, and headaches were lesscommon. The median time before onset was 18hr (range, 2^48 hr) and the median duration ofthe illness was 2 days (range, less than 1 day to15 days). The symptoms of the E. coli O27:H20infection were fairly similar; the average timebefore onset was 44 hr (range, 6-144 hr), and thesymptoms lasted for 4.4 days (range, 1-14days). The typical symptoms of O157:H7 foodpoisoning include production of stools contain-ing blood and mucus, as a result of hemorrhagiccolitis, and acute renal failure, particularly inchildren. Generally, a long incubation period (upto 9 days, average 4 days) and duration of illness(up to 9 days, average 4 days) elapse before theonset of symptoms and the occurrence of deaths.In contrast, in food poisoning due to other patho-genic strains of E. coli, the symptoms occur

Page 511: Cheese Science

within 2 days of ingestion of the contaminatedfood, and no deaths have been reported. Oralchallenges using human volunteers suggest that10M010 cells of EPEC, 108-1010 cells of ETEC,and 108 cells of EIEC are required to cause diar-rhea. In contrast, only 10-100 cells of EHEC arerequired to cause illness.

The precise mechanism by which E. coliO157:H7 causes disease has not been fully eluci-dated, but isolates produce 1 or 2 toxins that arecytotoxic to Vero cells, an African green mon-key kidney cell line. For this reason, E. coliO157:H7 is also called verotoxigenic E. coli(VTEC). One of these toxins is structurally andimmunologically indistinguishable from theshiga toxin of Shigella dysenteriae. How E. coliO157:H7 acquired the shiga toxin is not clear.

There is relatively little information on the in-cidence in E. coli in raw milk or cheese, but theevidence suggests that it is low. A relatively re-cent survey of milks throughout the productionseason (Rea, Cogan, & Tobin, 1992) showedthat more than 60-100% of samples containedfewer than 10 E. coli per ml, depending on thedate of sampling (Figure 20-1). Very few milkscontained more than 100 E. coli per mililiter.None of 568 raw milks examined in the UnitedKingdom contained E. coli O157:H7 (Neaves,Deacon, & Bell, 1994). In a limited survey, E.coli O157:H7 was not detected in 19 soft andsemi-soft American-made cheese, while otherstrains of E. coli were found in 11 samples(Ansay & Casper, 1997). In a Spanish study of221 raw milk cheeses and 75 pasteurized milkcheeses, 3 cheeses (1.4%), each of which hadbeen produced from raw milk, showed the pres-ence of toxigenic E. coli (Quinto & Cepeda,1997).

E. coli O157:H7 has a minimum growth tem-perature of 80C, an optimum of 370C, a maxi-mum of 440C, and it does not withstand pasteur-ization. E. coli strains normally do not toleratelow pH values, but E. coli O157:H7 is an excep-tion and can grow at pH 4.5 in media adjustedwith HCl—but not if the pH is adjusted with lac-tic acid. The organism does not grow in cheeseat pH values at or below 5.4.

20.6 GROWTH OF PATHOGENS DURINGCHEESE MANUFACTURE

The major factors responsible for the controlof microbial growth in cheese have been de-scribed in Chapter 10. These factors are also in-volved in controlling the growth of pathogens.The main reason for the low incidence of food-poisoning outbreaks caused by cheese is thatmost milk for cheesemaking is pasteurized,which kills all pathogens in the raw milk. This isprobably the most important factor in control-ling the growth of potential pathogens in cheese.There is some evidence that phagocytosis of L.monocytogenes by somatic cells present in rawmilk increases the heat resistance of L. mono-cytogenes, but the consensus is that this organ-ism is still killed by normal pasteurization. Sig-nificant amounts of cheese are made from rawmilk in countries bordering the Mediterranean(~ 15% of all French cheese is made from rawmilk), and in some cases no starter is used (e.g.,artisanal production of the Spanish cheesesManchego and Cabrales), which implies thatcheese made from raw milk is also safe, sincefew outbreaks have been attributed to such prod-ucts. Pasteurization is normally carried out at720C for 15s, but lower heat treatments (e.g.,650C for 16-18 s) will destroy all the likelypathogenic microorganisms commonly found inmilk except, perhaps, L. monocytogenes. Sub-pasteurization heat treatments are also used insome countries (e.g., Canada) for milk forcheesemaking. The reason for this practice isthat stronger flavored cheeses are produced frommilk that has not been pasteurized at all or hasbeen subpasteurized than from fully pasteurizedmilk (see Chapter 15).

M. paratuberculosis causes paratuberculosis(Johne's disease) in cattle, and there is some evi-dence that this organism also may be involved inthe etiology of Crohn's disease in humans. Be-cause of this evidence, milk has been suggestedas a possible vehicle for the transmission of theorganism from cattle to man. Whether the organ-ism withstands normal pasteurization is impor-tant. There is conflicting evidence on this point

Page 512: Cheese Science

in the literature, and some cheese manufacturershave recently increased the temperature of pas-teurization by a few degrees and/or the time ofpasteurization by a few seconds to ensure thatthe milk is free of this organism. These more se-vere conditions may damage the rennetability ofthe milk.

The presence and survival of pathogens incheese is influenced by several factors, includ-ing

• species• initial numbers• physiological condition of the microorgan-

ism• rate of acid production by the starter and

consequent decrease in pH• tolerance of the pathogen to acid and salt• tolerance to the cooking temperature to

which the cheese curd is subjected• postmanufacturing contamination

• biochemical changes that occur in thecheese during ripening

• ripening and storage temperature of thecheese

• composition of the cheese

The time-temperature profile during cheese-making, the initial low rate of decrease in pH,and the cooking temperature, which can varyfrom 330C for Camembert (essentially no cook-ing) to 360C for Dutch-type cheese, 380C forCheddar, and 540C for many Swiss and Italiancheeses, play major roles in promoting thegrowth of pathogens in cheese during manufac-ture. A temperature of 25-4O0C is conducive tothe growth of pathogens, if they are present. Thecooking temperature (540C) and the length oftime for which Swiss-type cheese curds are heldat this temperature (~ 60 min) will kill most, ifnot all, pathogens. In soft cheese, a cooking tem-perature of 350C or lower is used, which is ideal

Week of year

Figure 20-1 Incidence ofEscherichia coli in raw milks during the year.

% o

f sam

ples

<10 du/ml

10-<100 du/ml

>100 cfu/ml

Page 513: Cheese Science

for the growth of pathogens. Obviously, if thestarter is active, the pH will decrease quickly andthe growth of the pathogens will be retarded.The reverse also occurs. That is, if growth of thestarter is slow due to phage contamination and/or antibiotic residues in the milk, considerablegrowth of pathogens can occur. Therefore, a fastacid-producing starter is one of the best meansof controlling the growth of pathogens in cheese.

In studies on the growth of pathogens incheese, the milk is normally inoculated with thepathogens, pregrown under favorable condi-tions. The numbers found in naturally contami-nated milk would be much lower and probablyin a more stressed physiological state than thosegrown in laboratory media. In such a stressedstate, bacteria are probably less resistant to theeffects of pH and temperature, but experimentalevidence for this hypothesis is not available, pri-marily because of the difficulty of obtainingnaturally contaminated milk. Many raw milksare, in fact, free of pathogenic microorganisms.

Data for the growth of E. coli O157:H7, Lmonocytogenes, and S. aureus in Cheddarcheese curd during manufacture are shown inFigure 20-2. E. coli and S. aureus multipliedduring manufacture but L. monocytogenes didnot. When interpreting these data, one must re-member that the moisture content in the curd de-creases at each stage of manufacture, resulting inan apparent increase in bacterial numbers due tothe concentration effect. Based on this phenom-enon, it appears that a small decrease in thenumber of listeria occurred during manufacture.Considerable growth of E. coli occurred be-tween the beginning of manufacture and cuttingthe coagulum, when little acid production wouldhave occurred. No data were reported for S.aureus between these two stages, but growth ofthis organism is also likely to occur. These datarefer to Cheddar cheese curds, in which acid pro-duction is relatively rapid (acid production willalso have a major rate-limiting effect ongrowth). Growth of pathogens in the curd ofmost other varieties would probably be greaterthan in Cheddar because of slower acid pro-duction.

The growth of most pathogens will slow downand eventually cease owing to the decrease inpH (as significant amounts of lactic acid are pro-duced) and also to the increase in temperatureduring cooking.

20.7 GROWTH OF PATHOGENS INCHEESE DURING RIPENING

What happens during ripening of a cheese de-pends on the variety. Each cheese is a uniquemicrobial ecosystem and should probably beconsidered individually. Nevertheless, broadgeneralizations can be made. Hard and semi-hard cheeses, if made properly, are safe, sincealmost all pathogens die off during ripening; incontrast, significant growth of pathogens can oc-cur in soft cheese.

20.7.1 Hard and Semi-Hard Cheeses

Coliform bacteria die off at a rate of 0.3 logper week in Cheddar and 0.7 log per week inGouda. The fate of several pathogens in Em-mental and Cheddar cheese is shown in Figures20-3 and 20-4. Both Emmental and Cheddarcheeses are hard cheeses with similar pH values(~ 5.2) immediately after manufacture. None ofthe pathogens, except S. aureus at very low lev-els, was detected in the Emmental cheese within1 day of manufacture. This is most likely a resultof the high cook temperature (~ 520C) used inthe manufacture of this cheese. In Cheddarcheese, S. aureus, E, faecalis, E. coli, and a Sal-monella species all decreased during ripening,and the Gram-negative bacteria decreased at afaster rate than the Gram-positive organisms(Figure 20-4). One of the problems with S.aureus is that, even though the numbers of theorganism decrease significantly during ripening,sufficiently high numbers may have been pres-ent during the early stages of ripening to producethe small amounts of enterotoxin necessary tocause food poisoning. Therefore, it is possiblethat a cheese with a low level of S. aureus maycontain a high level of enterotoxin. Enterotoxinsare proteins, and whether they are hydrolyzed by

Page 514: Cheese Science

chymosin or bacterial proteinases during cheeseripening does not appear to have been studied. Inthe United States, storage of cheese at 20C for 60days may be used instead of pasteurization.

The number of L. monocytogenes in Cheddarcheese also decreases during ripening (Figure20-5), but some variations in individual trialsoccur. There was also some variation in the rateof die-off of L. monocytogenes in cheese ripenedat 130C and in cheese ripened at 60C, but gener-ally the differences were small. E. coli O157:H7died off relatively rapidly (2 log cycles in 25days) in Cheddar cheese during ripening at6.50C (Figure 20-6).

The effect of the pH of the cheese is also criti-cal. Data for Salmonella in Cheddar cheese areshown in Figure 20-7. At pH 5.03 and 5.23, theydied off quite quickly, but at pH 5.7 they did notdie at all. A pH of 5.23 is typical of a well-made

Cheddar, and a pH of 5.7 indicates poor starteractivity, either as a result of phage contamina-tion or antibiotic residues in the milk.

In Tilsit, a semi-hard cheese, the numbers ofall the pathogens tested decreased during ripen-ing, except L. monocytogenes, which remainedfairly constant during ripening for 30 days, afterwhich a gradual decrease of about 1 log cycleoccurred over the following 2 months (Figure20-3). The stability of L. monocytogenes in thischeese was attributed to the relatively low cook-ing temperature and short cooking time (420Cfor 15 min), which were bacteriostatic ratherthan bactericidal. pH may also be important. ThepH of the Tilsit cheese increased during ripeningfrom 5.2 at day 1 to 5.8 at day 90. Commercialsamples normally have a pH of about 6.2 at 90days. L. monocytogenes can grow over a widerange of temperature, from -I0C to 450C, and

Figure 20-2 Growth of E. coli 0157, L. monocytogenes, and S. aureus in Cheddar cheese curd during manufac-ture.

Cfu

/ml o

r g

Beg

inni

ng o

fm

anuf

actu

re

Cur

d af

ter C

uttin

g

Cur

d af

ter

Hea

ting

Cur

d af

ter

Coo

king

Cur

d af

ter

Salti

ng

Cur

d at

Pre

ssin

g

E. coli

L. monocytogenes

S. aureus

Page 515: Cheese Science

Emmental cheeseL

og c

fti/m

l or

g

L. monocytogenes

E. coll

S. aureus

Y. enterocolitica

P. aeruginosa

S. typhimurium

Milk Curdafter

cooking

Cheeselday

Cheese7 days

Cheese30 days

Cheese60 days

Cheese90 days

Tilsit cheese L. monocytogenes

E. coli

S. aureus

Y. enterolytica

P. aeruginosa

S. typhimurium

A. hydrophila

C. jejuni

Log

cfu

/ml o

r g

Milk Curdafter

cooking

Cheeselday

Cheese7 days

Cheese30 days

Cheese60 days

Cheese90 days

Figure 20-3 Growth of L. monocytogenes, E. coli, S. aureus, Y. enterocolitica, P. aeruginosa, S. typhimurium,Aeromonas hydrophila, and Campylobacter jejuni in Emmental and Tilsit cheese during ripening.

Page 516: Cheese Science

Time, days

Figure 20-5 Growth of L. monocytogenes Scott A in three trials of Cheddar cheese during ripening at 60C. Thevariation that occurred in different trials is clearly seen.

Cou

nt, c

fa/g

Trial 1Trial 2Trials

Days

Figure 20—4 Decrease in numbers of S. aureus, E. faecalis, E. coli, and Salmonella in Cheddar cheese duringripening at 120C.

this ability may be important for its survival inTilsit cheese.

20.7.2 Soft Cheeses

The situation in soft mold- and smear-ripenedvarieties like Camembert, Brie, and Limburger

is quite different, and many pathogens can growreadily in such cheeses. The reasons for this areas follows:

• These cheeses have a relatively high mois-ture content.

• They are ripened at a temperature (10—150C) at which bacterial growth can occur.

Log

cfti

/g

S. aureus

E. faecalis

E. coli

Salmonella

Page 517: Cheese Science

Days

Figure 20-6 Survival of E. coli O157:H7 in Cheddarcheese during ripening.

• The pH increases during ripening, espe-cially at the surface, owing to metabolismof lactate by Geotrichum candidium and thePenicillium spp., to a point where growth ofbacterial contaminants can occur (seeChapter 10).

The increase in the pH of soft cheese duringripening is significantly greater at the surface

than in the interior of the cheese (see Chapter 10)and is conducive to the growth of some patho-gens. Growth of L. monocytogenes Scott A, E.coli B2C, Enterobacter aerogenes MFl, andHafnia strain 14-1 occurred in Camembertcheese during manufacture. The strain of E. coliused was an enterotoxigenic strain. Hafnia areclosely related to coliform bacteria, and only onespecies, H. alvei, appears to occur in water andthe feces of humans and animals. What happensto these organisms during ripening is shown inFigure 20-8. The number of L. monocytogenesdecreased initially during ripening but increasedagain once the pH rose above 6. The numberalso increased in the core but not to the same ex-tent, probably because the pH increased moreslowly. In contrast, the number of E. coli andEnt. aerogenes increased during manufacturebut began to decrease once the pH of the curdreached 5.0, and the number continued to de-crease during ripening. This pattern of responseprobably characterizes all coliforms, withHafnia strain 14-1 as an exception to the rule.The number of Hafnia strain 14-1 increased untilthe pH fell to around 5; then it remained constantand decreased to 10 cfu/g during the first week

Salm

onel

la c

ount

, cfu

/g

Time, days

Figure 20-7 Effect of pH on the survival of Salmonella in Cheddar cheese during ripening.

pH 5.03

pH 5.25

pH 5.53

pH 5.70

Log

, cf

u/g

Page 518: Cheese Science

of ripening. The number began to increase againas soon as the pH began to increase, reaching afinal cell number of 108 cfu/g. The rate of in-crease in the pH of the three Camembert cheesesvaried. This was probably due to differences inthe manufacturing procedures and in the rate ofgrowth of the different strains of yeast and P.camemberti used. Of course, it is the combinedeffect of the temperature during ripening, thesalt concentration, and the decrease in pH thatreally determine the extent of growth of patho-gens.

Blue cheese is also a soft variety, but L. mono-cytogenes dies out during ripening even thoughthe pH increases from about 4.6 at day 1 to 6.2after 10 days. The death of Listeria in Bluecheese has been attributed to inhibition of their

growth by the high level of salt (« 10% salt-in-moisture) in these cheeses. The growth of otherpathogens on the surface of soft cheeses does notappear to have been investigated.

Significant differences in pH between thecore and the surface of soft cheese develop dur-ing ripening. This difference creates problems inobtaining representative samples of thesecheeses for analysis. Wedge-shaped samples arethe most representative of soft cheese (see Chap-ter 23).

20.8 RAW MILK CHEESES

Cheese made from raw milk has a much stron-ger flavor than the same cheese made from pas-teurized milk, and this is an important marketing

Time, weeks Time, weeks

Figure 20-8 Growth of L monocytogenes, E. coli, Enterobacter aerogenes, and Hafnia of strain 14-1 and theincrease in pH in Camembert cheese during ripening.

log

cfu/

g

Hd

log

cfu/

g

H«*

Hafnia, strain 14-1

Time, weeks Time, weeks

E. coli B2C

log

cfu/

g

Hd

Log

cel

l no

Hd

L. monocytogenes Scott A £>tf. aerogenes MFl

Page 519: Cheese Science

advantage for raw milk cheeses. Nevertheless,from the foregoing it is clear that soft cheesescan be problematic, and those made from rawmilk are particularly so. S. aureus is a commoncause of mastitis in dairy cows and therefore isprobably present in most raw milks. Approxi-mately 20% of the S. aureus strains present inraw milk produce enterotoxin. Growth of suchstrains to high numbers could therefore causeproblems in cheeses made from raw milk. It isalso likely that E. coli 0157:H7 is present in rawmilk, as its major source is bovine feces. Despitethis, only two food-poisoning outbreaks—andthese were small ones—have been traced tocheese containing E. coli O157:H7 (Table 20-1). Small numbers of L. monocytogenes mayalso be present in raw milk and may subse-quently grow in the cheese. S. aureus and E. coligrow during cheese manufacture and are poten-tial problems in cheeses made from raw milk. Inaddition, soft mold- and smear-ripened cheeseswith a high moisture content and in which thepH increases, especially at the surface, duringripening are potentially hazardous, especiallywhen they are made from raw milk.

To produce a good-quality raw milk cheese,free from pathogens, the raw milk should

• be of good quality, with bacterial countsbelow 20,000 cfu/ml and somatic cellcounts below 400,000/ml

• be produced under extremely hygienic con-ditions

• be free of pathogens• be held at a low temperature (< 40C)• be made with an active (fast acid-produc-

ing) starter• have good quality control procedures in

place, including hazard analysis criticalcontrol points (HACCP)

HACCP was developed by the U.S. space pro-gram to ensure the microbiological safety offoods for astronauts. It involves identifying themicrobiological hazards and preventive mea-sures that can be taken at each step in the manu-facture of the product. The points at which con-trol is critical to managing the safety of the

product are then identified and limits are set.These critical control points are then monitoredand recorded during the manufacture of eachsubsequent batch of product. In cheese manufac-ture, pasteurization and the pH of the curd at apredetermined time after addition of the starter,which estimates the rapidity of acid develop-ment, are obvious critical control points.

The European Union standards for differentpathogens in raw milk and in soft cheese madefrom raw or pasteurized milk are shown in Table20-3. Listeria and Salmonella must be absent in25 g, and a distinction is made between softcheeses produced from raw milk and those madefrom pasteurized milk, with more stringent stan-dards being set for the former. This reflects thepropensity of pathogens to grow in raw milkcheeses during manufacture.

20.9 CONTROL OF THE GROWTH OFPATHOGENS

To prevent the growth of pathogens, it is im-perative to prevent contamination of the milkand cheese and to be meticulously hygienic. To-day, much cheese is made in automated systems,but small-scale artisanal production involvesmanual manipulation of the curd during manu-facture, molding, and ripening. Good hygiene iscritical at each of these steps. Implementation ofHACCP systems is also very effective in pre-venting the growth of pathogens in cheese. Anactive, phage-free starter and pasteurization aremajor critical control points. The activity of thestarter should be assessed by determining the pHof every batch of cheese at a preset time afterstarter inoculation, such as 10 hr and 24 hr in thecase of soft cheeses and Cheddar, respectively.Comparisons of the data on a daily basis will in-dicate if starter activity is normal. Soft cheesesare small and will cool quickly. Therefore, keep-ing the ambient temperature high is importantwhen the curds are in the molds.

Good hygiene is particularly important in themanufacture of smear cheeses, especially whereold smear is used to inoculate the fresh cheeses.Old smear may be contaminated with L. mono-

Page 520: Cheese Science

cytogenes and may infect all cheeses. The use ofold smear is traditional in the production of thesecheeses, and efforts are being made to developdefined-strain smear starters to overcome theproblem. For example, much attention is beingfocused on identifying smear bacteria that pro-duce bacteriocins active against L. mono-cytogenes. Application of such cultures to thesurface of a cheese should be very useful in help-ing to prevent the growth of Listeria spp. Thedirect application of bacteriocins produced bylactic acid bacteria (LAB) that inhibit Listeria onthe surface of cheese is also being advocated asan effective method for controlling listerialgrowth on cheese.

As already indicated, several factors are in-volved in controlling the growth of pathogens(and other organisms) in cheese: pH, tempera-ture, and the level of salt are probably the mostimportant. The combined effect of these factorson growth is much greater than each factor's in-dividual effect, and in recent years there hasbeen a major effort to develop models for pre-dicting the growth of pathogens in foods basedon their growth responses to combinations ofsalt, temperature, andpH. These predictive mod-els have been developed mainly from experi-ments carried out in complex media, and the re-sults are felt to reflect the worst-case scenario infoods, since growth in foods at the same tem-perature, salt concentration, and pH value is gen-erally less than in model systems. For example,soft cheeses often have a salt concentration of1.5%, have a pH of 6.5, and are stored at 50C.Using these parameters, the model predicts thatan initial level of 10 L monocytogenes cells/gwould multiply to 10,000/g in 10 days (a genera-tion time of about 1 day).

20.10 ENTEROCOCCI

Enterococci are found at high numbers inmany cheeses, particularly those made aroundthe Mediterranean. Many of these are artisanalraw milk cheeses made at the farmhouse level.Enterococci are considered to be important inthe development of flavor in these cheeses,

where they comprise a significant proportion ofthe microflora of the raw milk and starter. Theirability to metabolize lactose and their toleranceto salt and heat make them ideal candidates asstarters. Enterococci at levels above 107/g havebeen found in such cheeses, and these high num-bers would almost certainly play a role in flavordevelopment.

There is considerable debate as to whether en-terococci should be considered pathogens. Dur-ing the past few decades, they have been impli-cated in several diseases, including bacteremia,urinary tract infections, and endocarditis. Manystrains are very promiscuous and easily pick upplasmids that encode vancomycin resistance.Many of these plasmids also are conjugative andcan be transferred naturally from cell to cell bysexual combination. Vancomycin is a glycopep-tide antibiotic that acts by inhibiting cell wallbiosynthesis. The incidence of vancomycin-re-sistant enterococci (VRE) in hospitals has in-creased dramatically. The use of avoparcin,which is also a glycopeptide antibiotic, as agrowth promoter in animal feed has been impli-cated in the increased occurrence of VREs infarm animals, including pigs and poultry. Be-cause of this, the use of avoparcin has beenbanned recently in several European countries.Many VREs are difficult to deal with becausethey are also resistant to other therapeutic antibi-otics, implying that alternative antibiotic therapymay not be available. However, many bacteria,including starter LAB like Lactobacillus, Pedio-coccus, and Leuconostoc, are intrinsically resis-tant to vancomycin.

There is little information on how rapidly En-ter'ococcus spp. grow in milk, but in Cheddarcheese they remain fairly constant during ripen-ing. Data for other commercial cheeses aresparse in the literature, but data for someartisanal Spanish and Italian cheeses are shownin Figure 20-9. Casar de Caceres and La Serenaare made from raw ewe milk and Afuega'l Pitufrom raw cow milk. No starters are used for anyof the three cheeses. Pecorino Umbro is madefrom pasteurized ewe milk and a mesophilicstarter is also used. A surface microflora devel-

Page 521: Cheese Science

ops on some of these cheeses, but the counts inFigure 20-9 are for the internal part of eachcheese. The first point on each line in the figureis the number of enterococci present in the milkat the beginning of manufacture. The data showthat considerable growth occurred during manu-facture and during the first days of ripening, af-ter which the numbers remained constant, exceptfor Afuega'l Pitu cheese, in which the numberdecreased. The numbers of enterococci in Casarde Caceres and La Serena cheese were well inexcess of 106/g and probably contribute to ripen-ing. Sometimes, it is difficult to distinguish be-tween lactococci and enterococci. However, the

counts of enterococci shown in Figure 20-9 arereliable, as selective media were used to enu-merate them.

20.11 BIOGENIC AMINES

Biogenic amines can be formed through de-carboxylation of amino acids by some strains ofnonstarter LAB, particularly Lb. buchneri, dur-ing cheese ripening. Tyramine, produced fromtyrosine, is probably the most important. Theseamines, which can cause food intoxicationwithin a few hours of ingestion, are discussed inChapter 21.

Ent

eroc

occi

, log

cfu

/g

Casar de Caceres

La Serena

Afuega'l Pitu

Pecorino Umbro

Time, days

Figure 20-9 Growth of enterococci in Casar de Caceres (+), La Serena (•), Afuega'l Pitu (A), and PecorinoUmbro (•) cheeses during ripening. The first point on each line is the count in the milk.

Page 522: Cheese Science

REFERENCES

Allen, V.D., & Stovall, W.D. (1960). Laboratory aspects ofstaphylococcal food poisoning from Colby cheese. Jour-nal of Milk and Food Technology, 23, 271—274.

Ansay, S.E., & Casper, C.W. (1997). Survey of retailcheeses, dairy processing environments and raw milk forEscherichia coli O157:H7. Letters in Applied Microbiol-ogy, 25, 111-114.

Aureli, P., Franciosa, G., & Pourshaban, M. (1996). Food-borne botulism in Italy. Lancet, 348, 1594.

Bille, J. (1990). Epidemiology of human listeriosis in Europewith special reference to the Swiss outbreak. In AJ.Miller, J.L. Smith, & G.A. Somkuti (Eds.), Foodbornelisterosis. Amsterdam: Elsevier.

Bone, F.J., Bogie, D., & Morgan-Jones, S.C. (1989). Staphy-lococcal food poisoning from sheep milk cheese. Epide-miology and Infection, 103, 449^58.

D'Aoust, J.Y. (1994). Salmonella and international trade.International Journal of Food Microbiology, 24, 11—31.

D'Aoust, J.Y., Warburton D.W., & Sewell, A.M. (1985).Salmonella typhimurium phage type 10 from Cheddarcheese implicated in a major Canadian foodborne out-break. Journal of Food Protection, 48, 1062-1066.

Desenclos, J.C., Bouvet, P., Benz-Lemoine, E., Grimont, F.,Desqueyroux, H., Rebiere, L, & Grimont, P.A. (1996).Large outbreak of Salmonella enterica serotype para-typhi B infection caused by a goats' milk cheese, France,1993: A case finding and epidemiological study. BritishMedical Journal, 312, 91-94.

Desmasures, N., Bazin, F., & Gueguen, M. (1997). Micro-biological composition of raw milk from selected farmersin the Camembert region of Normandy. Journal of Ap-plied Microbiology, 83, 53-58.

Eckman, M.R. (1975). Brucellosis linked to Mexican cheese.Journal of the American Medical Association, 232, 636—637.

Fontaine, R.E., Cohen, M.L., Martin, W.T., & Vernon, T.M.(1980). Epidemic salmonellosis from Cheddar cheese:Surveillance and prevention. American Journal of Epide-miology, 111, 247-253.

Food safety and cheese. (1998). International Food SafetyNews, 7, 6-10.

Goulet, V., Jacquet, C., Vaillent, V., Rebiere, L, Mouret, E.,Lorente, C., Maillot, E., Stanier, F., & Rocourt, F. (1995).Listeriosis from consumption of raw-milk cheese. Lan-cet, 345, 1581-1582.

Hedberg, C.W., Korlath, J.A., D'Aoust, J.Y., White, K.E.,Schell, W.L., Miller, M.R., Cameron, D.N., MacDonald,K.L., & Osterholm, M.T. (1992). A multistate outbreak ofSalmonella javiana and Salmonella oranienberg infec-tions due to contaminated cheese. Journal of the Ameri-can Medical Association, 268, 3203-3207.

Hendricks, S.L., Belknap, R.A., & Hausler, WJ. (1959). Sta-phylococcal food intoxication due to Cheddar cheese: 1.Epidemiology. Journal of Milk and Food Technology, 22,313-317.

Johnson, E.A., Nelson, J.H., & Johnson, M. (199Oa). Micro-biological safety of cheese made from heat-treated milk.1. Executive summary, introduction and history. Journalof Milk and Food Technology, 53, 441^52.

Johnson, E.A., Nelson, J.H., & Johnson, M. (199Ob). Micro-biological safety of cheese made from heat-treated milk:2. Microbiology. Journal of Milk and Food Technology,53,519-540.

Johnson, E.A., Nelson, J.H., & Johnson, M. (199Oc). Micro-biological safety of cheese made from heat-treated milk:3. Technology, discussions, recommendations, bibliogra-phy. Journal of Milk and Food Technology, 53, 610—623.

Kauf, C., Lorent, J.P., Mosimann, J., Schlatter, L, Somanini,B., & Velvart. J. (1974). Botulismusepidemic vom typeB. Schweizerische Medizinische Wochenschrift, 104,677-685.

Linnan, MJ., Mascola, M., Lou, X.O., Goulet, V., May, S.,Salminen, C., Hird, D.W., Yonekura, L., Hayes, P.,Weaver, R., Andurier, A., Plikaytis, B.D., Fannin, S.L.,Kicks, A., & Broome, C.V. (1988). Epidemic listeriosisassociated with Mexican-style cheese. New EnglandJournal of Medicine, 319, 823-828.

MacDonald, K.L., Eidson, M., Strohmeyer, C., Levy, E.,Wells, J.G., Puhr, N.D., Wachsmuth, K., Nargett, N.T., &Cohen, M.L. (1985). A multistate outbreak of gastrointes-tinal illness caused by enterotoxigenic Escherichia coli inimported semisoft cheese. Journal of Infectious Diseases,757,716-720.

Maguire, H.C.F., Boyle, M., Lewis, MJ., Pankhurst, J.,Wieneke, A.A., Jacob, M., Bruce, J., & O'Mahony, M.(1992). An outbreak of Salmonella dublin infection inEngland and Wales associated with a soft, unpasteurisedcow's milk cheese. Epidemiology and Infection, 109,389-396.

Malinverni, R., Bille, J., Perret, C., Regli, F., Tanner, F., &Glauser, M.P. (1985). Listeriose epidemique. Schweizer-ische Medizinische Wochenschrift, 115, 2-10.

Marier, R., Wells, J.G., Swanson, R.C., Callahan, W., &Mehlman, IJ. (1973, December 15). An outbreak of en-teropathogenic Escherichia coli foodborne disease tracedto imported French cheese. Lancet, 2, 1376-1378.

Ministry of Agriculture, Fisheries and Food. (1997). Micro-biological survey: Raw cow's milk on retail sale. FoodSafety Information Bulletin, 12-13.

Neaves, P., Deacon, J., & Bell, C. (1994). A survey of theincidence of E. coli 0157 in the UK dairy industry. Inter-national Dairy Journal, 4, 679—696.

Page 523: Cheese Science

O'Donnell, E.T. (1995). The incidence of Salmonella andListeria in raw milk from farm bulk tanks in England andWales. Journal of the Society of Dairy Technology, 48,25-29.

Public Health Laboratory Service, Edinburgh. (1994). E. coli0157 phage type 28 infections in Grampian. Communi-cable Diseases and Environmental Health, 28 (No. 94/46), 1.

Public Health Laboratory Service, London. (1994). Twoclusters of haemolytic uremic syndrome in France. Com-municable Disease Report, 4(1}, 29.

Public Health Laboratory Service, London. (1995). Brucel-losis associated with unpasteurised milk products abroad.Communicable Disease Report, 5(32), 151.

Public Health Laboratory Service, London. (1997a). Salmo-nella gold-coast and Cheddar cheese [Update]. Commu-nicable Disease Report, 7(11), 93, 96.

Public Health Laboratory Service, London. (1997b). Verocytotoxin producing Escherichia coli 0157. Communi-cable Disease Report, 7(46), 409, 412.

Quinto, E.J., & Cepada, A. (1997). Incidence of toxigenicEscherichia coli in soft cheese made from raw milk. Let-ters in Applied Microbiology, 24, 291-295.

Rea, M.C., Cogan, T.M., & Tobin, S. (1992). Incidence ofpathogenic bacteria in raw milk in Ireland. Journal ofApplied Bacteriology, 73, 331-336.

Sadik, C., Krending, MJ., Mean, F., Aubort, J.D.,Schneider, P.A., & Roussianos, D. (1986). An epidemic-logical investigation following an infection by Salmo-nella typhimurium due to the ingestion of cheese madefrom raw milk. In Proceedings of the Second World Con-

SUGGESTED READINGS

Bell, C., & Kyriakides, A. (1998a). E. coli: A practical ap-proach to the organism and its control in foods. London:Blackie Academic and Professional.

Bell, C., & Kyriakides, A. (1998b). Listeria: A practical ap-proach to the organism and its control in foods. London:Blackie Academic and Professional.

Doyle, M.P., Beuchat, L.R., & Montville, TJ. (1997). Food

gress on Foodborne Infections and Intoxications (Vol. 1).Berlin.

Sebald, M., Jouglard, J., & Gilles, G. (1974). Type B botu-lism in man due to cheese. Annales de I 'Institut Pasteur,125A(3), 349-357.

Sharpe, J.C.M. (1987). Infections associated with milk anddairy products in Europe and North America, 1980-85.Bulletin of the World Health Organization, 65, 397^06.

Todd, E.C.D., Shelley, D., Szabo, R., Robem, H., Gleeson,T., Durante, A., Marcoux, A., Entis, P., Morrison, D.,Purvis, U., Foster, R., Burgener, D.M., Wright, W.W.,Maharaja, R.S., Brodsky, M., Magus, M., Ruf, F.W., &Shab, H.W. (1979). Staphyloccal intoxication fromSwiss-type cheese: Quebec and Ontario. Canadian Dis-eases Weekly Report, 5(26), 110-112.

Todd, E.C.D., Szabo, R., Gardiner, M.A., Akhtar, M.,Delorme, L., Tourillon, P., Moisan, S., Rochefort, J.,Roy, D., Loit, A., Lamontagne, Y., Gosselin, L.,Martineau, G., & Breton, J.P. (198Ia). Staphylococcalintoxication from cheese curds: Quebec. Canadian Dis-eases Weekly Report, 7(34), 171-172.

Todd, E.C.D., Szabo, R., Robern, H., Gleeson, T., Park, C.,& Clark, D.S. (198Ib). Variation in counts, enterotoxinlevels and TNase in Swiss-type cheese contaminated withStaphylococcus aureus. Journal of Food Protection, 44,839-846,851-852,856.

Vaillant, V., Haeghebaert, S., & Desenclos, J.C. (1996). Out-break of Salmonella dublin infection in France, Novem-ber-December 1995. Eurosurveillance, 1, 2, 9-10.

Zehren, V.L., & Zehren, V.F. (1968). Examination of largequantities of cheese for Staphylococcal enterotoxin A.Journal of Dairy Science, 51, 635-644.

microbiology. Washington, DC: American Society forMicrobiology.

Microorganisms in foods 5. (1996). In Microbiologicalspecifications of food pathogens. London: Blackie Aca-demic and Professional.

Mortimore, S., & Wallace, C. (1994). HACCP: A practicalapproach. London: Chapman & Hall.

Page 524: Cheese Science

21.1 INTRODUCTION

While the nutritional merits of any foodshould be considered in the context of overalldietary intake, nevertheless, it is accurate to de-scribe cheese as a nutritious and versatile foodthat can play an important role in a healthy diet,as outlined in current guidelines (Exhibit 21-1).Although per capita consumption of most dairyproducts has declined worldwide, cheese is anotable exception. Current consumption in anumber of countries is shown in Table 1-3. Thepopularity of cheese is enhanced by its healthyand positive image, the variety of cheeses avail-able, and the compatibility of cheese and cheese-containing products with modern trends toward

Exhibit 21-1 Dietary Guidelines for Americans Is-sued by the U.S. Departments of Agriculture andHealth and Human Services

• Eat a variety of foods.• Balance the food you eat with physical ac-

tivity; maintain or reduce your weight.• Choose a diet containing plenty of grain

products, vegetables, and fruits.• Choose a diet low in fat, saturated fat, and

cholesterol.• Choose a diet moderate in sugars.• Choose a diet moderate in salt and sodium.• If you drink alcoholic beverages, do so in

moderation.

greater consumption of convenience and pre-pared foods.

Cheese is a nutrient-dense food, the precisenutritional composition of which is determinedby multifactorial parameters, including the typeof milk used (species, breed, stage of lactation,and fat content) and the manufacturing and rip-ening procedures. In general, cheese is rich inthe fat and casein constituents of milk, which areretained in the curd during manufacture, and itcontains relatively small amounts of the water-soluble constituents (whey proteins, lactose, andwater-soluble vitamins), which partition mainlyinto the whey. The composition of selectedcheeses is shown in Table 21-1.

21.2 FAT AND CHOLESTEROL

Fat plays several important functions incheese: it affects, for example, cheese firmness,adhesiveness, mouth-feel, and flavor (see Chap-ters 12 and 13). It also contributes significantlyto the nutritional properties of cheese, as mostcheeses contain significant amounts of fat. Forexample, a 50 g serving of Cheddar cheese pro-vides 17 g fat, in which approximately 66% ofthe fatty acids are saturated, 30% are mono-unsaturated, and 4% are polyunsaturated. Atypical Western diet providing 2,000 kcal (8,400kJ) per day, with 40% of energy derived fromfat, contains approximately 88 g fat. Thus,cheese contributes a significant amount of bothsaturated fat and total fat to the diet.

Nutritional Aspects of CheeseThomas P. O'Connor and Nora M. O'Brien

CHAPTER 21

Page 525: Cheese Science

Many expert groups worldwide have issueddietary guidelines recommending reductions inthe intake of both total and saturated fat byWestern populations (Exhibit 21-1). Althoughsome experts dispute the merit and efficacy ofthese guidelines, the vast body of expert opinionsupports the concept that dietary intakes do sig-nificantly influence the risk of chronic disease.This message has been, in large part, acceptedby consumers, and the food industry has re-sponded by producing foods low in fat and cho-lesterol to meet market trends. A range of "light"cheese products with a reduced level of fat hasbeen developed.

The cholesterol content of cheese varies fromapproximately 10 to 100 mg/100 g, dependingon the variety (Table 21-1). It is well estab-lished that dietary cholesterol intake exerts a

much smaller influence than the intake of di-etary saturated fat on a person's blood choles-terol level, which is a significant risk indicatorfor coronary heart disease (Keys, 1984). Mostindividuals (80%) show little change in theirblood cholesterol level in response to a changein dietary cholesterol intake in the range 250-800 mg/day. However, a minority of adults doexhibit an increased level of blood cholesterolin response to increased dietary intake of cho-lesterol (McNamara, 1987).

In recent years, there has been considerableresearch interest in the role of ingested choles-terol oxidation products (COPs) in the etiologyof chronic diseases. However, under normalconditions of manufacture, ripening, and stor-age, negligible amounts of COPs are formed incheese.

Table 21-1 Composition of Selected Cheeses

Water Protein Fat Carbohydrate Cholesterol EnergyCheese Type (9/10Og) (g/100g) (g/WOg) (g/100g) (mg/100 g) kcal kJ

Brie 48.6 19.2 26.9 Trace 100 319 1,323Caerphilly 41.8 23.2 31.3 0.1 90 375 1,554Camembert 50.7 20.9 23.7 Trace 75 297 1,232Cheddar (normal) 36.0 25.5 34.4 0.1 100 412 1,708Cheddar (reduced fat) 47.1 31.5 15.0 Trace 43 261 1,091Cheshire 40.6 24.0 31.4 0.1 90 379 1,571Cottage cheese 79.1 13.8 3.9 2.1 13 98 413Cream cheese 45.5 3.1 47.4 Trace 95 439 1,807Danish blue 45.3 20.1 29.6 Trace 75 347 1,437Edam 43.8 26.0 25.4 Trace 80 333 1,382Emmental 35.7 28.7 29.7 Trace 90 382 1,587Feta 56.5 15.6 20.2 1.5 70 250 1,037Fromagefrais 77.9 6.8 7.1 5.7 25 113 469Gouda 40.1 24.0 31.0 Trace 100 375 1,555Gruyere 35.0 27.2 33.3 Trace 100 409 1,695Mozzarella 49.8 25.1 21.0 Trace 65 289 1,204Parmesan 18.4 39.4 32.7 Trace 100 452 1,880Processed cheese8 45.7 20.8 27.0 0.9 85 330 1,367Ricotta 72.1 9.4 11.0 2.0 50 144 599Roquefort 41.3 19.7 32.9 Trace 90 375 1,552Stilton 38.6 22.7 35.5 0.1 105 411 1,701

a Variety not specified.

Page 526: Cheese Science

21.3 PROTEIN AND CARBOHYDRATE

The concentration of protein in cheese variesfrom approximately 3% to 40%, depending onthe variety (Table 21-1). Cheese protein is pre-dominantly casein, as the vast majority of thewhey proteins are lost in the whey. As casein isslightly deficient in sulfur-containing amino ac-ids, the biological value of cheese protein isslightly less than that of total milk protein. If theessential amino acid index of total milk proteinis assigned a value of 100, the correspondingvalue for cheese protein varies from 91 to 97,depending on the variety. If whey proteins areincorporated into cheese, such as by use of ultra-filtration, the biological value of cheese proteinis similar to that of total milk protein.

Cheese ripening typically involves the pro-gressive breakdown of casein by indigenousmilk enzymes, rennet, and bacterial enzymesinto water-soluble and -insoluble peptides andamino acids. This process, which is essential forthe development of flavor and texture (see Chap-ters 11-13), also increases the digestibility ofcheese protein to almost 100%.

Cheese contains only trace amounts of re-sidual carbohydrate, primarily lactose. The re-sidual lactose in cheese curd is, normally, fer-mented to lactic acid by starter bacteria duringmanufacture and ripening. Thus, cheese can besafely consumed by persons deficient in the in-testinal enzyme p-galactosidase, which is in-volved in the digestion of lactose.

21.4 VITAMINS AND MINERALS

Since most of the milk fat is retained in thecheese curd, it follows that the fat-soluble vita-mins in milk also partition into the curd. Mostof the vitamin A in milk fat (80-85%) is presentin cheese fat. Conversely, most of the water-soluble vitamins in milk partition into the wheyduring curd manufacture. However, some mi-crobial synthesis of B vitamins may occur incheese during ripening. Significant quantities ofvitamin B12 are produced in Swiss cheeses by

propionic acid bacteria. The vitamin content ofa range of cheeses is indicated in Table 21-2.In general, most cheeses are good sources ofvitamin A, riboflavin, vitamin Bn, and, to alesser extent, folate. Cheese contains negligibleamounts of vitamin C.

Cheese is also an important source of severalnutritionally important elements, including cal-cium, phosphorus, and magnesium (Table21-3). It is a particularly good source of bio-available calcium, with most hard cheeses con-taining approximately 800 mg calcium/100 gcheese. Acid-coagulated cheeses (e.g., Cottagecheese) contain significantly lower levels of cal-cium than rennet-coagulated varieties. Recker,Bammi, Barger-Lux, and Heaney (1988) re-ported that the bioavailability of calcium fromcheese is comparable to that from milk. Os-teoporosis, which may lead to debilitating bonefractures, is a common condition in Western so-cieties. Although it is a disease of multifactorialetiology, there is widespread agreement thatadequate calcium intake during childhood andteenage years, especially by girls, is important inensuring the development of optimum peakbone mass and reducing the risk of subsequentosteoporotic fractures. Cheese can play a posi-tive role in the context of overall diet in supply-ing highly bioavailable calcium.

As discussed in Chapter 8, sodium chlorideplays several important roles during cheesemanufacture. The amount of salt added duringthe manufacture of different cheese varies sig-nificantly, resulting in large differences in theconcentration of sodium in cheese (Table 21-3).There is substantial evidence that adults in West-ern societies consume, on average, above opti-mum levels of sodium. Elevated sodium intakeis recognized as a risk factor for hypertension,particularly in those members of the populationwho are genetically salt sensitive. Hypertension,in turn, is an important risk factor for coronaryheart disease. Most dietary guidelines world-wide recommend moderate salt intake. How-ever, even among populations with a highcheese intake, cheese contributes only about5-8% to total sodium intake.

Page 527: Cheese Science

Table 21-2 Vitamin Content of Selected Cheeses

Biotin(W/100g)

Pantothenate(mg/100g)

Folatedig/100 g)

Vitamin B12

Qig/100g)Vitamin B6

(mg/100g)Niacin

(mg/100g)Riboflavin(mg/100 g)

Thiamine(mg/100 g)

Vitamin E(mg/100g)

Vitamin D(V0/100g)

Carotene(iag/100g)

Retinoldig/100 g)Cheese Type

5.6

3.5

7.6

3.0

3.8

4.0

3.0

1.6

2.7

1.8

3.0

2.4

N

1.4

1.5

2.2

3.3

2.3

N

2.3

3.6

0.35

0.29

0.36

0.36

0.51

0.31

0.40

0.27

0.53

0.38

0.40

0.36

N

0.32

0.35

0.25

0.43

0.31

N

0.50

0.71

58

50

102

33

56

40

27

11

50

40

20

23

15

43

12

19

12

18

N

45

77

1.0

1.1

1.1

1.1

1.3

0.9

0.7

0.3

1.0

2.1

2.0

1.1

1.4

1.7

1.6

2.1

1.9

0.9

0.3

0.4

1.0

0.15

0.11

0.22

0.10

0.13

0.09

0.08

0.04

0.12

0.09

0.09

0.07

0.10

0.08

0.11

0.09

0.13

0.08

0.03

0.09

0.16

0.43

0.11

0.96

0.07

0.09

0.11

0.13

0.06

0.48

0.07

0.10

0.19

0.13

0.05

0.04

0.08

0.12

0.10

0.09

0.57

0.49

0.43

0.47

0.52

0.40

0.53

0.48

0.26

0.13

0.41

0.35

0.35

0.21

0.40

0.30

0.39

0.31

0.44

0.28

0.19

0.65

0.43

0.04

0.03

0.05

0.03

0.03

0.03

0.03

0.03

0.03

0.03

0.05

0.04

0.04

0.03

0.03

0.03

0.03

0.03

0.02

0.04

0.03

0.84

0.78

0.65

0.53

0.39

0.70

0.08

1.0

0.76

0.48

0.44

0.37

0.02

0.53

0.58

0.33

0.70

0.55

0.03

0.55

0.61

0.20

0.24

0.18

0.26

0.11

0.24

0.03

0.27

0.23

0.19

N

0.50

0.05

0.24

0.25

0.16

0.25

0.21

N

N

0.27

210

210

315

225

100

220

10

220

250

150

140

33

Tr

145

225

170

210

95

92

10

185

285

315

230

325

165

350

44

385

280

175

320

220

100

245

325

240

345

270

185

295

355

Brie

Caerphilly

Camembert

Cheddar(normal)

Cheddar(reduced fat)

Cheshire

Cottage cheese

Cream cheese

Danish blue

Edam

Emmental

Feta

Fromage frais

Gouda

Gruyere

Mozzarella

Parmesan

Processed

cheese3

Ricotta

Roquefort

Stilton

Key: Tr = trace; N = nutrient is present in significant quantities but reliable information on the amount is lacking.a Variety not specified.

Page 528: Cheese Science

Cheese, along with other dairy products, is apoor source of dietary iron. Iron deficiency ane-mia is a major worldwide nutrition-related prob-lem, both in developed and developing coun-tries. In an attempt to alleviate this problem,there is considerable interest in fortifying com-monly consumed foods with iron. Cheddar andprocessed cheeses have been successfully forti-fied with iron.

21.5 ADDITIVES IN CHEESE

Preservatives are added occasionally duringthe manufacture of certain cheeses. Growth ofyeasts and molds on hard and semi-hard cheesesmay be inhibited by the addition of sorbic acid orits salts.

Nitrate may be added to milk prior to themanufacture of certain types of cheese. It is re-duced to nitrite, which inhibits growth of Clos-tridium species, a possible cause of late gasblowing and flavor defects (see Chapters 10 and11). However, nitrite does not persist well incheese, and the contribution from cheese to totalintake of nitrite is negligible.

In recent years, there has been considerableresearch interest in the efficacy of natural bacte-rially produced preservatives in cheese. Most in-terest has focused on the bacteriocin nisin, a pep-tide produced by some strains of Lactococcuslactis. It has been exploited commercially inSwiss-type and processed cheeses to prevent lateblowing by Clostridium species, the spores ofwhich survive pasteurization.

Table 21-3 Mineral Content of Selected Cheeses

Na K Ca Mg P Fe ZnCheese Type (mg/100g) (mg/100g) (mg/100g) (mg/100g) (mg/100g) (mg/100g) (mg/100 g)

BrIe 700 100 540 27 390 0.8 2.2Caerphilly 480 91 550 20 400 0.7 3.3Camembert 650 100 350 21 310 0.2 2.7Cheddar 670 77 720 25 490 0.3 2.3

(normal)Cheddar 670 110 840 39 620 0.2 2.8

(reduced fat)Cheshire 550 87 560 19 400 0.3 3.3Cottage cheese 380 89 73 9 160 0.1 0.6Cream cheese 300 160 98 10 100 0.1 0.5Danish blue 1,260 89 500 27 370 0.2 2.0Edam 1,020 97 770 39 530 0.4 2.2Emmental 45 89 970 35 590 0.3 4.4Feta 1,440 95 360 20 280 0.2 0.9Fromagefrais 31 110 89 8 110 0.1 0.3Gouda 910 91 740 38 490 0.1 1.8Gruyere 670 99 950 37 610 0.3 2.3Mozzarella 610 75 590 27 420 0.3 1.4Parmesan 1,090 110 1,200 45 810 1.1 5.3Processed 1,320 130 600 22 800 0.5 3.2

cheese3

Ricotta 100 110 240 13 170 0.4 1.3Roquefort 1,670 91 530 33 400 0.4 1.6Stilton 930 130 320 20 310 0.3 2.5

a Variety not specified.

Page 529: Cheese Science

21.6 CHEESE AND DENTAL CARIES

Dental caries are a commonly occurring prob-lem that, in simple terms, involves degenerationof tooth enamel due to acid produced by oralmicroorganisms during the metabolism of sug-ars. The progress and extent of dental caries areinfluenced by a variety of dietary parameters andnutrient interactions, including the composition,texture, solubility, and retentiveness of food andby its ability to stimulate saliva flow. In recentyears, considerable work has been conducted onthe cariostatic effects of cheese.

Early work demonstrated that dairy productsreduced the development of dental caries in ratsand also in vitro. These effects were attributed tothe high concentrations of calcium and phos-phate in milk and to the protective effects ofcasein. This early work was supported by furtherdetailed work on rats that indicated that bothcasein and whey proteins had protective effects,the former being the more effective, and alsoconfirmed the protective effects of calcium andphosphate.

The first evidence that cheese had an anti-cariogenic effect in humans was reported byRugg-Gunn, Edgar, Geddes, and Jenkins (1975).Consumption of Cheddar cheese after sweet-ened coffee or a sausage roll increased plaquepH, possibly due to increased output of saliva,which can buffer the effect of acids formed inplaque. Some early work suggested that theconsumption of cheese resulted in reducednumbers of Streptococcus mutans, which is in-volved in acid production, in the mouth. How-ever, later work suggested that the cariostaticeffects of cheese may not be directly related toits effect on Sc. mutans but could be explainedprimarily by mass action effects on soluble ions,particularly calcium and phosphate (Jenkins &Harper, 1983).

Further investigation of the cariostatic effectsof cheese in humans was reported by Silva,Jenkins, Burgess, and Sandham (1986), whomeasured the demineralization and hardness ofenamel slabs fastened to a prosthetic appliancemade specifically for each human subject to re-place a missing lower first permanent molar.

Each subject chewed 5 g of cheese immediatelyafter rinsing his or her mouth with a 10% (w/v)solution of sucrose. Chewing cheese resulted ina 71% decrease in demineralization of theenamel slabs and an increase in plaque pH butdid not significantly affect the oral microflora.

Further trials on humans have confirmed thatthe consumption of hard cheese results in sig-nificant rehardening of softened enamel sur-faces (Gedalia, lonat-Bendat, Ben-Mosheh, &Shapira, 1991; Jenkins & Hargreaves, 1989).While more research is required to define pre-cisely the mechanisms involved in the cario-static effects of cheese, it is reasonable to rec-ommend the consumption of cheese at the endof a meal as an anticaries measure.

21.7 MYCOTOXINS

Mycotoxins (Figure 21-1) are fungal metabo-lites that have been shown to be cytotoxic, mu-tagenic, teratogenic, and carcinogenic in ani-mals. Certain mycotoxins (e.g., aflatoxin) areamong the most potent animal toxins known,hence giving rise to concerns regarding their po-tential effects in the human food supply.

The consumption of aflatoxin-contaminatedfood, particularly in conjunction with hepatitis Binfection, is a key risk factor for liver cancer,which is the principal form of cancer reported inless developed countries. Mycotoxins may bepresent in milk and dairy products, such ascheese, owing to indirect contamination (con-tamination of the cows' feedstuff) or direct con-tamination (growth of mycotoxin-producingfungi in the milk and dairy products).

21.7.1 Indirect Contamination

It has been known for almost 40 years that theintake of feedstuff contaminated with aflatoxinBI by dairy cows may result in the excretion oftoxic factors (principally aflatoxin M1) in theirmilk within a few hours (Allcroft & Carnaghan,1962). On average, 1-2% of ingested aflatoxinBI is excreted in milk as aflatoxin MI. It has beenshown subsequently that indirect contaminationof milk with other mycotoxins, such as ochra-

Page 530: Cheese Science

AflatoxinBl

AflatoxinMl

Ochratoxin A

Zearalenone

Patulin

Figure 21-1 Structures of selected mycotoxins.

Page 531: Cheese Science

toxin A, zearalenone, T-2 toxin, sterigmato-cystin, and deoxynivalenone, does not representa major public health issue (for a comprehensivereview of mycotoxins in dairy products, see vanEgmond, 1989).

Van Egmond (1989) summarized the resultsof surveillance programs in many countries foraflatoxin M1 in milk and milk products. The in-cidence and levels of aflatoxin MI in milk andmilk products have decreased significantly overthe years, which can be attributed primarily toimplementation of regulatory limits on the con-tamination of feedstuff with aflatoxin BI. A no-table exception was the significant increase inthe level of aflatoxin MI in U.S. dairy productsin 1988-89. This resulted from feeding aflatoxinB i-contaminated maize products due to the se-vere drought in the U.S. Midwest in 1988 whichcreated ideal conditions for the growth of andaflatoxin production by the producing organism,Aspergillus flavus. Results reported in surveil-lance programs for cheese have, in general, indi-cated that aflatoxin MI was not detected or oc-curred at concentrations below legal limits(0.2-0.25 jig/kg).

Studies have been conducted on the fate andstability of aflatoxin M1 in milk during cheesemanufacture and ripening. These studies indi-cate that aflatoxin M1 partitions between thecurd and whey in both acid-coagulated and ren-net-coagulated cheeses and that aflatoxin M1 isvery stable during cheese manufacture. The par-tition coefficient for aflatoxin M1 in water sug-gests that most of the toxin should partition intothe whey. This anomaly can be explained byfindings that aflatoxin M1 tends to associatehydrophobically with casein micelles, resultingin a greater than expected partitioning into thecheese curd.

In general, it appears that aflatoxin M1 isstable in several cheese varieties during ripen-ing.

21.7.2 Production of Toxic Metabolites inMold-Ripened Cheese

Penicillium roqueforti and P. camemberti areused in the manufacture of various types of blue-

veined and white surface-mold cheeses. Thesemolds can produce a range of toxic metabolites.Some strains of P. roqueforti can produce PRtoxin, patulin, mycophenolic acid, penicillicacid, roquefortine, cyclopiazonic acid, isofumi-gaclavine A and B, and festuclavine. P. cam-emberti strains produce cyclopiazonic acid,which has been detected in commercial samplesof Camembert and Brie. It occurs primarily inthe rind at levels below 0.5 mg/kg whole cheesebut may occur at levels up to 5 mg/kg if the stor-age temperature is too high. However, evalua-tion of available toxicological data for cyclo-piazonic acid, together with potential humanexposure estimated from consumption data forCamembert and Brie, suggests that this metabo-lite causes no appreciable public health risk(Engel & Teuber, 1989).

P. roqueforti can produce a range of toxins, asoutlined above. Patulin, penicillic acid, and PRtoxin have not been detected in commercialsamples of cheese. Mycophenolic acid has beendetected in commercial cheese samples but atlevels well below those that pose a risk to humanhealth. Roquefortine and isofumigaclavine Aand B have been detected at low levels in com-mercial Blue cheese, and their toxicity is low.Compelling evidence that the consumption ofmold-ripened cheeses is not hazardous to humanhealth was provided by studies on rats and rain-bow trout that consumed levels of mold equiva-lent to a daily human intake of 100 kg cheesewith no apparent signs of toxicity. Mold-ripenedcheeses have been consumed for several hun-dred years without apparent ill effects.

21.7.3 Direct Contamination of Cheese withMycotoxins

Cheese is a good substrate for the growth ofadventitious molds given suitable conditions oftemperature, humidity, and oxygen. Mycotoxin-producing molds require oxygen for growth andhence are very unlikely to grow on vacuum-packed or wax-coated cheese, particularly ifthere is good plant sanitation during manufac-ture and handling and if the storage temperatureis low.

Page 532: Cheese Science

Unintentional growth of molds on cheese dur-ing ripening and/or storage results in financialloss, reduces consumer appeal, and may necessi-tate trimming. However, the production of my-cotoxins may also represent a health risk.Cheeses on which unintentional growth of moldhad occurred have been reported to contain my-cotoxins that are nephrotoxic (ochratoxin A, cit-rinin), teratogenic (ochratoxin A, aflatoxin B1),neurotoxic (penitrem A, cyclopiazonic acid),and carcinogenic (aflatoxin B1 and G1, ochra-toxin A, patulin, penicillic acid, sterigmato-cystin) (Ueno, 1985).

Penicillium species are the mycotoxigenicfungi most frequently isolated from cheese; As-pergillus and other species are encountered onlyoccasionally. However, the presence of moldgrowth does not imply that mycotoxins are pres-ent in cheese. A large body of work (see vanEgmond, 1989) has been conducted on the oc-currence of mycotoxins in moldy cheese. Theoverall incidence was low in most studies. Fur-thermore, less than 50% of Penicillium specieswere toxicogenic in animal studies. Work hasalso been conducted on the incidence of myc-otoxins in cheeses contaminated with Aspergil-lus species. There is very little evidence that sig-nificant levels of aflatoxins are produced incheese by these species.

Work has also been undertaken on the abilityof mycotoxins to migrate from the surface ofcheese into the interior. The data are significantfor decisions on whether to trim or discard mold-contaminated cheese. The Health ProtectionBranch of the Department of Health and Wel-fare, Canada, has recommended that if a hardcheese is contaminated with a patch of mold, thecheese can be salvaged by removing the infectedportion to a depth of 2.5 cm.

21.8 BIOGENIC AMINES IN CHEESE

The term biogenic amines refers to nonvola-tile, low-molecular weight aliphatic, acyclic,and heterocyclic amines, such as histamine,tyramine, tryptamine, putrescine, cadaverine,and phenylethylamine, that may be present in

cheese or other foods. In cheese, biogenicamines are produced by decarboxylation ofamino acids during ripening by enzymes re-leased by the microorganisms present. Levelsproduced vary as a function of ripening periodand microflora, with the highest levels mostlikely in cheeses heavily contaminated withspoilage microorganisms. Renner (1987) re-ported average values of histamine and tyraminein some cheeses (Table 21-4).

Consumption of foods containing significantlevels of biogenic amines may cause food poi-soning. However, for most individuals, con-sumption of even large amounts of biogenicamines does not elicit toxicity symptoms, sincethey are converted rapidly to aldehydes bymono- and diamine oxidases and then to car-boxylic acid by oxidative deamination. How-ever, if these enzymes are impaired, owing to agenetic defect or inhibitory drugs, toxic symp-toms may result.

Histamine is a normal body constituentformed from histidine by a pyridoxal phos-phate-dependent decarboxylase. Its concentra-tion in blood is tightly regulated, and orally ad-ministered histamine results in toxicity onlyfollowing ingestion of a very high dose or im-pairment of histidine metabolism. Toxic symp-toms generally become apparent within 3 hr ofingestion and include, initially, a flushing of theface and neck, followed by an intense, throbbingheadache. Other symptoms are observed occa-sionally, including cardiac palpitations, diz-ziness, faintness, rapid and weak pulse, gas-trointestinal complaints, bronchospasms, andrespiratory distress.

Consumption of fish, particularly of theScombroidae family, has been associated withmost cases of histamine poisoning. However,some instances have been reported to be relatedto cheese consumption. Gouda cheese contain-ing 85 mg histidine/100 g was implicated in anoutbreak in the Netherlands. Incidences in theUnited States resulting from the consumption ofcontaminated Swiss cheese have also been re-ported (Taylor, Kiefe, Windham, & Howell,1982).

Page 533: Cheese Science

Table 21-4 Average Tyramine and HistamineContents of Some Cheese Types

Tyramine HistamineCheese Type fiug/g) fiug/g)

Cheddar 910 110Emmental 190 100Blue 440 400Edam, Gouda 210 35Camembert, Brie 140 30Cottage 5 5

Tyramine is normally present at low levels inthe body. In humans, monoamine oxidase(MAO)-catalyzed oxidative deamination to p-hydrophenylacetic acid is the main degradativepathway for tyramine. However, if a genetic de-

REFERENCES

Allcroft, R., & Carnaghan, R.B.A. (1962). Groundnut toxic-ity: Aspergillus flavus toxin (aflatoxin) in animal prod-ucts. Veterinary Record, 74, 863-864.

Engel, G., & Teuber, M. (1989). Toxic metabolites from fun-gal cheese starter cultures. In H.P. van Egmond (Ed.),Mycotoxins in dairy products. London: Elsevier AppliedScience.

Gedalia, L, lonat-Bendat, D., Ben-Mosheh, S., & Shapira, L.(1991). Tooth enamel softening with a cola type drinkand rehardening with hard cheese or stimulated saliva.Journal of Oral Rehabilitation, 18, 501-506.

Jenkins, G.N., & Hargreaves, J.A. (1989). Effect of eatingcheese on Ca and P concentrations of whole mouth salivaand plaque. Caries Research, 23, 159—164.

Jenkins, G.N., & Harper, D.S. (1983). Protective effect ofdifferent cheeses in an in vitro demineralization system[Abstract]. Journal of Dental Research, 62, 284.

Keys, A. (1984). Serum cholesterol response to dietary cho-lesterol. American Journal of Clinical Nutrition, 40, 351—359.

McNamara, DJ. (1987). Effects of fat-modified diets oncholesterol and lipoprotein metabolism. Annual Reviewof Nutrition, 7, 273-290.

ficiency of MAO exists or if MAO-inhibitorydrugs are administered, toxicity symptoms maybe manifest. These include a hypertensive crisis,often accompanied by severe headache and, incertain cases, intercranial hemorrhage, cardiacfailure, and pulmonary edema. The main dietarysources of tyramine, besides cheese, includemarinated herring, dry sausages, and marmite.The tyramine content of cheese is generallygreater in long-ripened varieties, such as extramature Cheddar, than in young cheese. Patientsprescribed MAO-inhibitory drugs should be ad-vised to avoid intake of tyramine-rich foods.Tyramine poisoning in the absence of MAO-in-hibitory drugs has not been reported. The toxic-ity threshold for tyramine has been estimated at400 mg, and therefore healthy individuals cantolerate intakes of large amounts of tyramine-rich cheese.

Recker, R.R., Bammi, A., Barger-Lux, M.J., & Heaney, R.P.(1988). Calcium absorbability from milk products, animitation milk and calcium carbonate. American Journalof Clinical Nutrition, 47, 93-95.

Renner, E. (1987). Nutritional aspects of cheese. In P.P. Fox(Ed.), Cheese: Chemistry, physics and microbiology(Vol. 1). London: Elsevier Applied Science.

Rugg-Gunn, A.J., Edgar, W.M., Geddes, D.A.M., & Jenkins,G.N. (1975). The effect of different meal patterns uponplaque pH in human subjects. British Dental Journal,739,351-356.

Silva, M.D. de A., Jenkins, G.N., Burgess, R.C., &Sandham, HJ. (1986). Effect of cheese on experimentalcaries in human subjects. Caries Research, 20, 263-269.

Taylor, S.L., Kiefe, T.J., Windham, E.S., & Howell, J.F.(1982). Outbreak of histamine poisoning associated withconsumption of Swiss cheese. Journal of Food Protec-tion, 45, 455-457.

Ueno, Y. (1985). The toxicology of mycotoxins. CRC Criti-cal Review of Toxicology, 14, 99-132.

van Egmond, H.P. (1989). Aflatoxin MI: Occurrence, toxic-ity, regulation. In H.P. van Egmond (Ed.), Mycotoxins indairy products. London: Elsevier Applied Science.

Page 534: Cheese Science

22.1 INTRODUCTION

The liquid remaining after removal of the fatand casein from milk by isoelectric or rennet co-agulation of the casein is called whey. The termmilk serum is used for the supernatant liquid ob-tained by ultracentrifuging skimmed (fat-free)milk; 95% of the casein micelles are sedimentedby ultracentrifugation at 100,000 g for 1 hr. Milkserum represents the aqueous phase of milk, un-changed by the process of separation, although itdoes contain small casein micelles.

The wheys prepared by isoelectric precipita-tion or rennet coagulation are called acid wheyand sweet (rennet) whey, respectively. They dif-fer in composition from each other (Table 22-1)and from milk serum because of changes thatoccur during their preparation. For example,acid whey contains a much higher concentrationof calcium, magnesium, phosphate, and citratethan sweet whey or milk serum owing to the so-lution of the colloidal milk salts upon acidifica-tion. If properly prepared, acid whey should befree of casein. Commercially, acid whey is usu-ally prepared from efficiently skimmed milk inthe manufacture of acid-coagulated cheeses (seeChapter 16) or acid casein and is therefore es-sentially free of fat, although it does containsome phospholipids. Sweet whey is a byproductof the manufacture of rennet-coagulated cheeseor rennet casein and its composition varies de-pending on its source (e.g., pH 6.2-6.6), depend-ing on the extent of acidification that had oc-

curred prior to whey separation (hence the con-centration of some salts varies somewhat). Mostcheeses are made from full-fat or partiallyskimmed milk, and typically about 10% of thefat in such milk is lost in the whey as a result ofthe formation of free (nonglobular) fat duringpasteurization and pumping of the milk and theloss of fat globules from the curd pieces duringcutting and cooking. Rennet casein is producedfrom skimmed milk, and therefore the resultingwhey is essentially fat-free. As discussed inChapter 6, rennet coagulation involves cleavageof K-casein, and the resulting macropeptidesformed are present in rennet whey. Other casein-derived peptides may be present in whey if anexcessively proteolytic rennet substitute is used.If the rennet coagulation process is incompletewhen the gel is cut, the whey may contain someuncoagulated casein as well as small particles ofcurd. The compositional data for acid and rennetwheys shown in Table 22-1 are typical values,although the values vary considerably.

It is apparent from Table 22-1 that whey con-tains about 50% of the total solids of milk, in-cluding essentially all of the lactose and wheyproteins (provided that the whey proteins werenot denatured by heat treatment prior to coagula-tion), 50-100% of the milk salts (depending onwhether it was produced by acid or rennet co-agulation), and some fat (depending on whetherskimmed or whole milk was used). Thus, wheyis a valuable source of food constituents fromwhich numerous food products are now pro-

Whey and Whey Products

CHAPTER 22

Page 535: Cheese Science

duced. In this chapter, the principal productsproduced from whey are described briefly. Thereader should refer to Sienkiewiez and Riedel(1990) for a detailed discussion on whey andwhey utilization. Certain aspects of whey pro-teins are covered in Fox (1992) and certain as-pects of lactose and its derivatives in Fox (1997)and International Dairy Federation (1993).

Traditionally, whey was regarded as a wasteproduct and was disposed of by the cheapestpossible method—fed to animals (especiallypigs), spray irrigated onto land, dumped in wa-terways, or treated as effluent. Some whey isstill disposed of by such methods, but dumpingof whey is unacceptable today for environmentalreasons, and improved technology makes it pos-sible to recover whey constituents in a cost-ef-fective manner. World production of whey isabout 160 million tonnes per annum, represent-ing about 7 million tonnes of lactose and 1 mil-lion tonnes of whey protein.

22.2 CLARIFICATION OF WHEY

The curd fines may be removed using a vi-brating screen separator but are removed more

often using a centrifugal separator (clarifier).The fines are recovered as a concentrate (e.g.,50% dry matter), which is pressed and may beused in processed cheese or similar products.Removal of fines facilitates further processingof whey (e.g., by ultrafiltration).

Fat, which is typically present at a level ofabout 0.3% (w/w) in bulk cheese whey, is recov-ered from the clarified whey using centrifugalseparators, typically to levels of 0.07%, w/w.The resultant whey cream (~ 50% fat) is nor-mally used for the manufacture of whey butter,which is used as a food ingredient in, for ex-ample, processed cheese products.

The phospholipids in whey exist as lipoproteinparticles, which block ultrafiltration membranes,reducing the flux rate of the plant. A number ofmethods have been developed to aggregate thelipoprotein particles, such as by adding CaCl2 andraising the pH to around 7.5. The flocculated cal-cium phosphate-lipoprotein particles may be re-moved by sedimentation, by centrifugation, orpreferably by microfiltration. The lipoproteinshave good emulsification properties and may beused in a number of food applications. The clari-fied whey is subjected to further processing.

Table 22-1 Typical Composition and pH of Whole Milk, Sweet (Rennet Casein and Cheddar Cheese)Wheys, and Acid (Lactic and Mineral Acid) Wheys

Sweet Wheys Acid Wheys

Rennet Cheddar3 Lactic Acid Mineral AcidComponent Casein (g/L) Cheese Casein (g/L) Casein (g/L) Whole Milk (g/L)

Total solids 66.0 67.0 64.0 63.0 122.5Total protein (N x 6.38) 6.6 6.5 6.2 6.1 33.0Nonprotein nitrogen (NPN) 0.37 0.27 0.40 0.30Lactose 52.0 52.0 44.0 47.0 47.0Milk fat 0.20 3.0 0.30 0.30 35.0Minerals (ash) 5.0 5.2 7.5 7.9 7.5Calcium 0.50 0.40 1.6 1.4 1.2Phosphate 1.0 0.50 2.0 2.0 2.0Sodium 0.53 0.50 0.51 0.50 0.5Lactate - 2.0 6.4pH 6.4 5.9 4.6 4.7 6.7

a Unseparated whey.

Page 536: Cheese Science

22.3 CONCENTRATED AND DRIEDWHEY PRODUCTS

Whey powders have been produced for manyyears and have several applications in the foodindustry, including as ingredients for bakery andmeat products and ice cream. The value of wheypowders can be increased and their range of ap-plications extended by one of several processmodifications.

22.3.1 Nonhygroscopic Whey Powder

Lactose, which represents about 70% of thetotal solids in whey, is difficult to crystallize, andif the lactose is not properly crystallized, the wheypowder is hygroscopic, making it unstable duringstorage. Nonhygroscopic whey powder is pro-duced by concentrating the whey to 50-60% totalsolids, seeding the concentrate with lactose crys-tals to induce crystallization, and, when crystalli-zation is complete, drying the concentrate.

22.3.2 Demineralized Whey Powder

One of the important applications of wheysolids is in the manufacture of infant formulae.Human milk contains more lactose (~ 7%) andless casein (~ 1% total protein and a wheyprotein:casein ratio of 60:40, compared with20:80 for bovine milk) than cow milk. Manymodern baby formulae based on cow milk arehumanized; that is, their lactose content andcasein:whey protein ratio are adjusted to ap-proximate those in human milk. This adjustmentis usually made by blending bovine whey andskimmed milk. However, the concentration ofsalts in bovine milk is 3-4 times higher than thatin human milk and places an undesirably highrenal load on the baby. The problem may be re-solved by reducing the concentration of ions inwhey by electrodialysis and/or ion exchangers.

22.3.3 Delactosed and Delactosed-Demineralized Whey Powder

For many food applications, it is desirable touse a whey product with a higher than normal

protein content. This may be achieved by theprocesses described below for the production ofwhey protein products or alternatively by crys-tallizing out some of the lactose. The lattermethod involves concentrating the whey, seed-ing it with lactose to induce crystallization, andremoving the lactose crystals by centrifugationor filtration. The mother liquor may or may notbe demineralized (see Section 22.3.2) and spray-dried to yield a protein-rich whey powder.

22.4 LACTOSE

Lactose is a sugar unique to milk (see Chapter3). Among commercially available sugars, lac-tose has many unusual properties:

• low solubility• difficult to crystalize• a tendency to form supersaturated solutions• low sweetness• low hygroscopicity when properly crystal-

ized• a tendency to adsorb flavors and pigments

These characteristics create problems for thedairy industry, but methods have been devel-oped for managing and controlling the problems.In fact, some of these characteristics are ex-ploited in the production of improved dairyproducts, such as instant milk powders and Io w-hydroscopicity icing sugar mixtures (see Fox,1997). Consequently, a substantial market hasdeveloped for lactose, although very small incomparison with that for sucrose (~ 250,000tonnes per annum of pure lactose in comparisonwith 95 million tonnes per annum of sucrose).

As noted above, lactose is produced by con-centrating whey to 50-60% total solids, seedingit with lactose crystals, and recovering of thecrystals by centrifugation or filtration. If extrahigh purity lactose is required, the first crop ofcrystals may be dissolved and recrystallized.

The market for lactose appears to be relativelylimited, but lactose can be converted, enzymati-cally, chemically, or physically, to a range ofuseful derivatives:

• Lactulose (Figure 22-1). Lactulose isformed when the glucose moiety of lactose

Page 537: Cheese Science

is isomerized to fructose by a mild alkalinetreatment. Lactulose is not hydrolyzed by(3-galactosidase in the human intestine andpasses to the lower intestine, where it mayact as a laxative or promote the growth ofbifidobacteria, which can have beneficialeffects on the microbial ecology of thelower intestine (Figure 22-2).

• Lactitol (Figure 22-3). The carbonyl groupof lactose may be reduced to an alcohol,lactitol, by chemical or electrolytic meth-ods. Lactitol is not hydrolyzed in the humanintestine and hence may be used as anonnutritive sweetener. It is claimed tohave anticholesterolemic and anticario-genic properties. Lactitol may be esterifiedwith one or more fatty acids to produce arange of food-grade emulsiflers.

• Lactobionic acid. Lactose may be oxidizedto lactobionic acid, which may be dehy-drated to lactobionic acid lactone (Figure22—4). Both derivatives have a number offood and industrial applications, but thevolumes used are small.

• Glucose-galactose syrups. Lactose may behydrolyzed by (3-galactosidase (lactase) orby free acid or cation exchangers to pro-duce glucose-galactose syrups, which aresweeter and more soluble than lactose.Such syrups have several applications infood products but in most cases are not costcompetitive with glucose, glucose-fructose,or sucrose.

• Galacto-oligosaccharides. p-galactosidasenormally functions as a hydrolase, but un-der certain conditions it may function as atransferase, with the production of galacto-oligosaccharides (Figure 22-5) involvingvarious bond types not digested in the hu-man intestine. These oligosaccharides passinto the large intestine, where they serve topromote the growth of Bifidobacterium spp.

22.5 WHEY PROTEINS

Bovine whey contains two main proteins, (3-lactoglobulin and a-lactalbumin, with lesseramounts of blood serum albumin and immuno-

globulins (mainly IgGi) and trace amounts ofseveral proteins, especially lactotransferrin, andseveral enzymes (see Chapter 3). Many of theseproteins have desirable nutritional, functional,and, in some cases, pharmaceutical properties.Numerous methods are available for the recov-ery of whey proteins in toto and, more recently,for the isolation of individual proteins.

The first and simplest of these is heat denatur-ation and recovery of the aggregated protein,known as lactalbumin. The product is insoluble,has very poor functional properties and is usedmainly in nutritional fortification of foods.

Whey protein concentrates (WPC, 30-80%protein), prepared by ultrafiltration, are widelyused as functional ingredients, such as for thepreparation of gels, foams, and emulsions (seeFox, 1992).

Products with a higher protein content (up to95%; known as whey protein isolates) have bet-ter functionalities and are produced by ion ex-change chromatography. Alternatively, sepa-rated whey is ultrafiltrated to 15% dry matterand microfiltered to remove fat. The defattedpermeate is further concentrated by ultrafiltra-tion and diafiltration to about 20% dry matterand spray dried. The product typically contains92% protein, less than 0.5% fat, and 96% drymatter. Since production costs are high, wheyprotein isolates are produced on a small scale.

The properties of some whey proteins makethem more suitable for certain applications thanothers. For example, the gelation properties of (3-lactoglobulin are superior to those of a-lactalbu-min, but (3-lactoglobulin is less suitable for thefortification of infant formulae, since it does notoccur in human milk and many human infantsare allergic to it. Several methods have been de-veloped for the fractionation of whey proteins,but most are not amenable to industrial-scaleproduction and none is used on a truly commer-cial scale. Presumably, the efficacy of thesemethods will be improved, costs will probablydecrease, and demand will increase, and conse-quently the fractionation of whey proteins on acommercial scale may become a reality.

Some of the minor whey proteins are poten-tially very valuable as nutraceuticals. Much in-

Page 538: Cheese Science

Figure 22-1 Chemical structure of lactulose.

Pyranose form Furanose form

Page 539: Cheese Science

terest has focused on lactoferrin (lactotrans-ferrin), which is present at a very much higherconcentration in human milk than in bovinemilk. Lactoferrin is a nonheme iron-bindingprotein that has bacteriostatic and other physi-ological properties and serves as a source of bio-logically available iron. Because lactoferrin iscationic at the pH of milk (at which most othermilk proteins are anionic), it can be easily iso-lated from milk or whey and has been used tosupplement infant formulae and other special di-etary products.

Lactoperoxidase is also cationic at the pH ofmilk and may be readily isolated. In the presenceof H2O2 and the thiocyanate anion, lactoperoxi-dase is a very effective bactericidal agent andhas been used as an additive in milk replacers forcalves and piglets.

The (glyco)macropeptide (GMP) producedfrom K-casein upon the renneting of milk con-tains no aromatic amino acids and therefore issuitable for patients suffering from phenylketon-urea. It is also claimed to have several interest-ing physiological effects. Methods have beendeveloped for the preparation of GMP on a pilotscale.

22.6 WHEY CHEESE

The whey proteins are recovered as softcheese by heating a mixture of whey and skim orwhole milk, adjusted to pH 6.0, at 9O0C. Ricottaand variants thereof, Anari and Manouri, are ex-amples of this type of cheese; they are discussedin Chapters 16 and 17.

Not cariogenic

LACTULOSE

Oral intake

Non-absorption and migrationto large intestine

Utilization DV Bifidobacteriwn Favourable change ofintestinal microflora

Suppression of intestinalputrefactive bacteria

Suppression of productionof harmful substances

Vitamin synthesis

Production of organic acidsand lowering of intestinal pH

Moderate faecalexcretion

Suppression ofNH3 production

Lessening burdensto hepatic function

Ensuring intestinalfunction

Figure 22-2 Significance of lactulose in health.

Page 540: Cheese Science

Lactitol, 4-O-p-D-galactopyranosyl-D-sorbitol

Lactitol monoester (Lactyl palmitate)

Figure 22-3 Structure of lactitol and its conversion to lactyl palmitate.

Page 541: Cheese Science

LACTOSEGal ({51-*-4) GIu

I Hydrolysis

Gal + GIu

I Internal rearrangement

TGal (1—^-2) GIu

GaI(I-^-S)GIu

Gal (1-^-6) GIu(Allolactose)

Transglycosylation

IGaI(I-^-S)GaI

Gal (1-^-6) Gal

Gal (1—^6) Gal (1 —^4) GIu (6'galactosyl lactose)

Gal (1—^-3) Gal (1 —^4) GIu (3'galactosyl lactose)

Gal (l-*-6) Gal (1 -*-6) GIu

IGal (1-^-6) Gal (1 -*^6) Gal

Tetrasaccharides

Pentasaccharides

IHexasaccharides

Figure 22-5 Possible reaction products from the ac-tion of p-galactosidase on lactose.

Lactobionic acid-5-lactone

Lactose

Figure 22-4 Structure of lactobionic acid and its 8-lactone.

Page 542: Cheese Science

The whey proteins are also incorporated intosome forms of Queso bianco produced fromwhole milk and acidified to pH 5.4 by heating at9O0C. They may be incorporated into Quarg byusing ultrafiltration technology or the Centri-Whey process (whey is heated at 9O0C to dena-ture the whey proteins, which are then recoveredby centrifugation, added to milk for the nextbatch of cheese, and become incorporated intothe Quarg).

Heating milk damages or destroys its rennetcoagulation properties (see Chapter 6), but theseproperties can be restored by acidifying thecheese milk or supplementing it with CaCl2.This approach has been proposed as a means forincreasing the yield of rennet-coagulated cheesebut is used to a very limited extent.

The whey proteins can be incorporated intorennet-coagulated cheese by preconcentratingthe milk to the total solids content of the particu-lar variety by ultrafiltration and coagulating thepre-cheese by rennet. This technology has beenquite successful for soft cheeses but not, to date,for semi-hard or hard cheeses (see Chapter 17).

Finally, whey, usually mixed with wholemilk, may be concentrated by thermal evapora-tion to about 87% solids to produce a uniquefamily of cheeses, examples of which areMysost and Gjetost (see Chapters 16 and 17).These cheeses are quite different from all othercheeses—they have a sweet taste (owing to thehigh level of lactose), a brown color (due to the

REFERENCES

Fox, P.P. (1992). Advanced dairy chemistry: Vol. 1. Proteins(2d ed.). London: Elsevier Applied Science.

Fox, P.F. (1997). Advanced dairy chemistry: Vol. 3. Lactose,water, salts and vitamins (2d ed.). London: Chapman &Hall.

Maillard reaction between lactose and proteins),and a fudgelike consistency.

22.7 FERMENTATION PRODUCTS

Lactose in whey or, more usually, in ultrafil-tration permeate may be used in various fermen-tation processes. The most widespread of thefermentation products is ethanol, which is beingproduced on a commercial scale in several facto-ries. Other fermentation products include aceticacid (from ethanol), lactic acid, and propionicacid (produced from lactic acid by Propionibac-terium spp.). Lactose may also be used as a moregeneral fermentation substrate but is not cost-competitive with sucrose in the form of molas-ses. The production of yeast biomass from wheyfermentation has been considered but it is noteconomical.

22.8 CONCLUSION

Whey, which contains about 50% of the totalsolids of milk and was regarded as a wastestream until recently, can serve as the raw mate-rial for the production of a wide range of foodproducts and food ingredients. Some of these arebeing produced profitably on a commercialscale. It is likely that as new technologies aredeveloped, new and improved food ingredientsderived from or based on whey will be devel-oped.

International Dairy Federation. (1993). Proceedings of theIDF Workshop on Lactose Hydrolysis [Bulletin No. 289].Brussels: Author.

Sienkiewicz, T., & Riedel, C.-L. (1990). Whey and whey uti-lization (2d ed.). Gelsenkirchen-Buer, Germany: VerlagThomas Mann.

Page 543: Cheese Science

23.1 INTRODUCTION

Cheese is analyzed for a variety of reasons: toascertain its composition for nutritional pur-poses, to ensure compliance with standards ofidentity, to assess the efficiency of production,and to assess the microbial safety of the productor the influence of enzymes and/or microorgan-isms on cheese quality, among others. Reliablemicrobiological, physical, chemical, and sen-sory analyses are of critical importance to thedairy scientist involved in cheese research, toanalysts working on quality assurance, and forregulating the production process.

23.2 METHODS OF SAMPLING CHEESE

Unless the entire cheese is sampled, the reli-ability of the results of any analytical procedureis dependent on how representative the sampleanalyzed is. Correct sampling is, therefore, ofparamount importance, and standard procedureshave been published (International Dairy Fed-eration, 1985). In general, samples should betaken by an experienced and responsible personwho is familiar with the methods. Since trace-ability may be an important consideration in in-dustry, the samples should be sealed, appropri-ately labeled, and accompanied by a samplingreport.

Subsamples must be taken from cheese foranalysis, unless the cheese is sufficiently small(or has been previously subdivided), in order for

the entire cheese to be used as the sample. Thebasic apparatus for sampling most cheese variet-ies is the cheese trier (Figure 23-1), although asuitable knife or a cutting wire may also be used.The sampling procedure recommended by theInternational Dairy Federation (IDF, 1985) in-volves removing by trier a shallow plug ofcheese (15-20 mm deep), which may be retainedas a surface sample, followed by the insertion ofa smaller trier into the resulting hole, throughwhich an internal sample is taken. Another com-mon practice is to take a plug using a trier and toretain the outer 15-20 mm of the plug as the sur-face sample. Cheese triers should be made fromstainless steel and should be sterilized (by dip-ping in 70% [v/v] alcohol and flaming) beforesampling for microbiological or sensory analy-sis. Equipment for obtaining samples for chemi-cal or physical analyses should be clean. Cheesesamples should be stored in a suitable container(e.g., a plastic container or bag or aluminumfoil), and containers for microbiological samplesmust be sterile. In general, duplicate samples(100-200 g) should be taken and stored at 0-40Cuntil analyzed, and the analysis should be per-formed as soon as possible after sampling (pref-erably within 24 hr). Care should also be takenwhen sampling fresh cheeses to avoid wheyseparation. Suitable sampling techniques areshown in Figure 23-2.

Care should be taken when sampling cheeseswith gradients from center to surface (e.g., vari-eties ripened with a surface microflora or brine-

Analytical Methods for Cheese

CHAPTER 23

Page 544: Cheese Science

Front Side

Figure 23-1 Diagram of a trier used to take samples from cheese.

Page 545: Cheese Science

salted cheeses). It is recommended that such va-rieties should not be sampled using a trier butrather by a technique that involves cutting thecheese, because a trier will not give a representa-tive sample from cheeses with radial salt andmoisture gradients. For research purposes, it iscommon to analyze separately surface and inter-nal samples from these cheeses.

23.3 COMPOSITIONAL ANALYSIS

Gross compositional analysis (moisture, ash,protein, fat, acidity, and salt) of cheese is con-ducted according to standard methods publishedby the International Dairy Federation (see IDF,1995) or the Association of Official AnalyticalChemists (Cunniff, 1995). Moisture (total sol-

(g) (h)

(e) (f)

(d)(c)

(a) (b)

Figure 23-2 Suggested sampling techniques for cylindrical (a), cubic (b), block-shaped (c), and spherical (d)cheeses using a trier; suggested sampling technique for cheeses with a circular cross section and a mass greaterthan 2,500 g (e) or between 1,100 and 2,500 g (f); suggested sampling techniques for block-shaped cheeses thelargest face of which is rectangular (g) or square (h).

Page 546: Cheese Science

ids) is usually determined gravimetrically bydrying a sample to constant weight in an oven at10O0C or in a microwave oven with an integralweighing apparatus. Certain cheese varieties(e.g., Blue cheese) contain significant amountsof substances other than water that are volatile atbelow 10O0C. Moisture in such cheeses may bedetermined by reflux distillation in the presenceof an imiscible solvent (e.g., «-amyl alcohol andxylene, 1:2) and collection of the water sepa-rated from the cheese in a calibrated tube. Ash isdetermined gravimetrically by heating a samplein a furnace at or below 55O0C until completelyashed. Protein is estimated by measuring the Ncontent of cheese by the Kjeldahl method andmultiplying by a conversion factor (6.38). Rapiddetermination of fat in cheese is often made bythe Gerber (butyrometric) or Babcock methods,although the standard reference method is basedon the Schmid-Bondzynski-Ratzlaff (SBR)technique. The SBR technique, in which fat isextracted by a mixture of diethyl ether and lightpetroleum after digestion of protein with HCl, isrecommended in preference to the Rose-Gottlieb method, since the NH3 solution used inthe latter does not completely dissolve manytypes of cheese and because free fatty acids re-leased by lipolysis are not fully extracted fromthe ammoniacal solution.

The chloride (and thus NaCl) content ofcheese is usually determined by titration withAgNO3, with potentiometric or colorimetricendpoint determination. There appears to be nostandard method for determining the pH ofcheese. The method used in our laboratories is asfollows: grated cheese (10 g) is thoroughlyblended with 10 ml of H2O using a mortar andpestle, and the pH of the resulting slurry is mea-sured potentiometrically. However, it may bepreferable to measure the pH of grated cheesedirectly (using reinforced electrodes) to mini-mize changes in pH caused by alternating thebalance between colloidal and soluble calciumphosphate upon dispersion in water.

The AOAC standard method for determiningthe titratable acidity of cheese involves mixinggrated cheese with water (4O0C) and filtering

and titrating an aliquot of the filtrate with 0.1 MNaOH, using phenolphthalein as the indicator.Results are usually expressed as lactic acid per-centage (Cunniff, 1995). Although the method isprobably suitable for curd or young cheese, itwould appear to be unsatisfactory for maturecheese since the buffering capacity of the filtrateand hence its titratable acidity will increase asproteolysis progresses. It is even less suitable forcheeses the pH of which increases during ripen-ing (e.g., mold- and smear-ripened varieties),owing to the catabolism of lactic acid and theproduction of ammonia (see Chapter 11). Conse-quently, the concentration of lactic acid de-creases but the titratable acidity may increase asthe level of water-soluble peptides increases.Measurement of titratable acidity is not commonin cheese analysis; it may have potential as anindex of ripening.

Calcium can be quantified by

• titration with ethylenediamine tetraaceticacid (EDTA), using ammonium purpurate(murexide) as the indicator

• precipitation as calcium oxalate and weigh-ing

• atomic absorption spectrophotometry• ion-specific electrodes (it should be noted

that this method measures calcium ion ac-tivity and not total concentration)

The concentration of Na can be quantified spe-cifically using an ion-selective electrode, atomicabsorption spectroscopy, or flame spectropho-tometry. Phosphorus may be determined by acolorimetric assay using molybdovanadate ormolybdate-ascorbate reagents.

The water activity (aw) of cheese can be deter-mined by a variety of methods, including psy-chrometry, cryoscopy, dew-point hygrometry,and isopiestic equilibration. A number of regres-sion equations have been developed to predict aw

from chemical composition for various cheeses(see Chapter 8; Marcos, 1993).

Near infrared reflectance spectroscopy mayalso be used to determine fat, protein, moisture,and moisture in nonfatty substances in a range ofcheese varieties. Infrared transmittance spectro-

Page 547: Cheese Science

photometry of solvent extracts of cheese mayalso be used for gross compositional analysis(see McSweeney & Fox, 1993, for references).

Residual alkaline phosphatase activity incheese is used to indicate the use of raw orunderpasteurized milk for cheese manufacture.Alkaline phosphatase activity may be assayedusing several substrates, including disodiumphenylphosphate,/?-nitrophenyl phosphate, phe-nolphthalein monophosphate, and fluorophos.

Residual coagulant activity, which is impor-tant for the development of flavor precursors andtexture in cheese (see Chapter 11), may be as-sessed by monitoring the production of specificpeptides (e.g., asl-CN f24-199) by gel electro-phoresis (see Chapter 11), by measuring the co-agulation time of milk to which an extract fromthe cheese is added, by diffusion assays oncasein-containing agar, by determining the hy-drolysis of various peptide substrates, and byimmunochemical methods (see Baer & Collin,1993).

23.4 BIOCHEMICAL ASSESSMENT OFCHEESE RIPENING

23.4.1 Determination of Products of Lactose,Lactate, and Citrate Metabolism

The metabolism of lactose to lactic acid by thestarter bacteria is fundamental to most, if not all,varieties of cheese (see Chapters 5, 10, and 11).L-Lactate is produced by mesophilic and somethermophilic starters, while a mixture of the D-and L-isomers is produced by other thermophiles(see Chapter 5). In certain varieties (e.g., Swiss)the lactate serves as a substrate for further mi-crobial metabolism, and in many varieties the L-isomer produced by the starter is converted to aracemic mixture by the nonstarter flora of thecheese (see Chapters 10 and 11). Many of theproducts of sugar metabolism may be quantifiedby enzyme-based assays, and in these cases theenzymatic procedure has replaced earlier meth-ods because of increased sensitivity and speci-ficity. For some other compounds, instrumentalor colorimetric methods are also widely used.

Enzyme assay kits (e.g., from BoehringerMannheim, Mannheim, Germany) for lactose,lactic acid, and citrate are available. Lactose, D-glucose, D-galactose, and lactate may also be de-termined by high-performance liquid chroma-tography (HPLC). Lactose in processed cheesemay be determined by a modified Fehling's ti-tration (Cunniff, 1995). The enzymatic assay forlactate is particularly useful, since it distin-guishes between D- and L-isomers. In this assay,L- and/or o-lactate dehydrogenase (LDH) andglutamic-pyruvic transaminase (GPT) are used.Lactate is oxidized to pyruvate by LDH in thepresence OfNAD+. The equilibrium of this reac-tion normally lies in favor of lactate formation,but the pyruvate formed is trapped by reacting itwith L-glutamate in the presence of GPT, shift-ing the equilibrium in favor of pyruvate andNADH. The concentration of NADH is mea-sured spectrophotometrically (at 334, 340, or365 nm) and is stoichiometrically related to theconcentration of D- or L-lactate present.

The metabolism of citrate leads to the pro-duction of flavor (diacetyl) and nonflavor com-pounds (acetoin and 2,3-butanediol) in somecheeses. Citrate can be measured chemically orenzymatically. The chemical determination ofcitrate in cheese involves dispersing the cheesein NaOH, precipitating the protein by trichloro-acetic acid, filtering the reaction mixture, andreacting an aliquot of the filtrate with pyridineand acetic anhydride. The result is a yellowcomplex that is quantified spectrophotometri-cally at 428 nm (Marier & Boulet, 1958). Thisis a good method, but a standard curve must beincluded with each assay, as the relationship be-tween citrate and color intensity shows day-to-day variation. Also care should be taken, as py-ridine is carcinogenic. In the enzymic assay,citrate is converted to oxaloacetate and acetateby citrate lyase. In the presence of malate dehy-drogenase and L-LDH, oxaloacetate and its de-carboxylation product, pyruvate, are reduced toL-malate and L-lactate, respectively, with theconcomitant oxidation of NADH to NAD+,which is quantified spectrophotometrically at340 nm.

Page 548: Cheese Science

Diacetyl and acetoin can be quantified usingcolorimetric or gas chromatographic (GC) pro-cedures. In the colorimetric procedures, a dis-persion of the cheese is steam distilled and thefirst two 10 ml fractions are collected. Diacetylis estimated in the first 10 ml fraction and ac-etoin in the second (Walsh & Cogan, 1974). Forthe determination of diacetyl, aliquots of thesteam distillate are heated with hydroxylamineto form dimethylglyoxime, which is convertedto a pink ammonoferrous dimethylglyoxmatecomplex upon reaction with FeSO4 in acidic so-lution. The absorbance is then read at 525 nm.Diacetyl can also be quantified by reaction withHCl and 3,3'-diamino-benzidine tetrahydro-chloride but care is required since the latter iscarcinogenic. Headspace analysis by gas chro-matography (GC) can also be used to quantifydiacetyl using an electron capture detector.

Acetoin is generally measured in the second10 ml of distillate by reaction with 2-naphtholand creatine. The red complex formed is quanti-fied at 525 nm. Diacetyl also reacts with this re-agent, but no diacetyl is present in the second 10ml of distillate.

Acetolactate (AL) is produced by some startercultures and will break down to diacetyl and/oracetoin during steam distillation. When AL ispresent, the above method for measuring di-acetyl must be modified, and steam distillates ofsamples cannot be used for determination of ac-etoin. Diacetyl is measured in steam distillatesof samples adjusted to pH 0.5 before distillation.Under these conditions, no breakdown of AL todiacetyl occurs (in fact, it is completely con-verted to acetoin). Acetoin can be measured bysolubilizing the cheese with NaOH, followed bydirect estimation. Care should be taken not to in-crease the temperature to an extent that couldcause the breakdown of AL to acetoin.

To measure AL, samples are adjusted to pH3.5, and CuSO4 is added before steam distilla-tion. Under these circumstances, the AL is sto-ichiometrically converted to diacetyl (Mohr,Rea, & Cogan, 1997). The values obtained mustthen be corrected for the amount of "free" di-acetyl present.

A similar method for measurement of AL wasdescribed by Richelieu, Hoalberg, and Nielsen(1997), except that FeCl3 was used to convertAL to diacetyl at pH 3.1, which was then mea-sured by headspace capillary GC. To measurediacetyl, samples were adjusted to pH 7.0 beforeanalysis.

2,3-Butanediol can be quantified by extrac-tion with methylene chloride. The extract isseparated from the residue and dried with anhy-drous Na2SO4, and its volume is reduced by ro-tary evaporation. Upon equilibrating the extractwith water, the 2,3-butanediol passes into theaqueous phase, which is clarified with a mixtureof BaCl2, NaOH, and ZnSO4; the butanediol isthen estimated by GC.

Propionate and acetate are important inSwiss-type cheeses, which undergo a propionicacid fermentation. These can be extracted fromthe cheese by low concentrations of H2SO4 andanalyzed by HPLC. A better method is toacidify and steam distill samples of cheese andthen analyze the acids by GC or HPLC. Steamdistillation gives good cleanup, but it is impor-tant that standards should be included, as eachacid is recovered to a different extent by distil-lation. Pyruvic, lactic, acetic, and propionic ac-ids may also be quantified in cheese extracts byion-exchange HPLC using an Aminex 87Hcolumn.

23.4.2 Assessment of Lipolysis

The degree of lipolysis in cheese depends onthe variety and ranges from slight to very exten-sive (see Chapter 11). Extensive lipolysis in in-ternal bacterially ripened cheeses (e.g., Cheddar,Gouda, and Swiss) is undesirable, whereas inmold-ripened and some hard Italian cheeses li-polysis is essential for flavor development. Anumber of procedures have been developed toquantify lipolysis. The most effective methodsuse GC or HPLC but are tedious. Two rapidmethods, the copper soaps method and the aciddegree value method, have been used to assesslipolysis in cheese. Unfortunately, both havedrawbacks, and thus there does not appear to be

Page 549: Cheese Science

a simple, rapid method for accurately determin-ing the level of free fatty acids in cheese.

The copper soaps assay is a spectrophotomet-ric method used to estimate free fatty acids inmilk and cheese. The copper soaps of free fattyacids in cheese, formed by reaction withCu(NOs)2, are extracted using a mixture ofchloroform, heptane, and methanol as solvent,followed by centrifugation. An aliquot of thesolvent layer is added to a solution of sodium di-ethyldithiocarbamate in butan-1-ol. The color ofthe resulting complex is measured spectrophoto-metrically at 440 nm. The copper soaps of short-chain fatty acids (< Ci0) may partition betweenthe aqueous and apolar solvent phases and there-fore may not be extracted completely. The tech-nique has been criticized because of its poor re-covery of short-chain fatty acids (which have thegreatest effect on cheese flavor) and because itmay have low reproducibility in the hands of aninexperienced technician.

The acid degree value (ADV) has been usedfor many years as an index of lipolysis in dairyproducts. Fat is released by the combined actionof detergent, ion exchange, and heat and is sepa-rated from the aqueous phase of the cheese bycentrifugation. Aliquots of the fat are weighedand dissolved in solvent, and the free fatty acidsare titrated with alcoholic KOH (~ 0.02 M) usingmethanolic phenolphthalein as the indicator.Variations of this method are also used, includ-ing the Bureau of Dairy Industry (BDI) method.The ADV method is tedious, but it is reliable inthe hands of a competent operator for foods con-taining more than 2% fat.

An extraction and preparation procedure isgenerally necessary for gas chromatographictechniques, and in many cases esters of the freefatty acids are chromatographed. A wide varietyof methods are available. The most critical dif-ferences between them are not the GC condi-tions but rather the extraction and derivitizationprocedures used for sample preparation. Reli-able extraction and derivatization techniques areoften tedious, and their tediousness and the capi-tal cost of GC equipment are disadvantages ofthis technique for routine analysis.

Extraction with aqueous or organic solvents isused but has been criticized owing to its parti-tioning effects and the extraction of compoundsthat interfere with analysis. Fatty acids may alsobe adsorbed on various resins (e.g., AmberlystA-26) during analysis, but such methods mayresult in incomplete recovery, glyceride hy-drolysis, and low throughput. Free fatty acids areusually chromatographed as their methyl or bu-tyl esters and are identified and quantified byreference to standards. Because complex proce-dures are often involved in sample preparation,an internal standard (usually a fatty acid with anodd number of C atoms) should be includedearly in the analytical procedure.

In addition to GC, fatty acids may be quanti-fied by HPLC or, in certain cases, by enzyme-based assays. Specific chromatographic tech-niques are also available for triglycerides andpartial glycerides.

Methyl ketones (alkan-2-ones), the character-istic flavor compounds in Blue cheese, may beanalyzed by GC or their 2,4-dinitrophenyl-hydrazone derivatives may be analyzed by chro-matographic or spectrophotometric techniques.Hydroxyacids, fatty acid lactones, and othervolatile products of fatty acid catabolism mayalso be identified and quantified by GC or GC-mass spectrometry (GC-MS), which is discussedin Section 23.5 and in Chapter 12.

23.4.3 Assessment of Proteolysis

As discussed in Chapter 11, proteolysis is theprincipal and most complex biochemical eventthat occurs during the ripening of most cheesevarieties. Therefore, a wide range of techniquesfor its assessment have been developed (see Fox,McSweeney, & Singh, 1995; McSweeney &Fox, 1993, 1997). Such techniques can begrouped into two main classes: nonspecificmethods (which measure the formation of ni-trogenous compounds soluble in various extrac-tants or precipitants or the liberation of reactivegroups) and specific methods (which resolve in-dividual peptides). Nonspecific techniques arenormally relatively straightforward, and some

Page 550: Cheese Science

are suitable for assessment of ripening in qualitycontrol laboratories. However, more informa-tion about proteolysis is provided by techniquesthat resolve individual peptides (i.e., varioustypes of electrophoresis and chromatography).

The amount of cheese nitrogen soluble in wa-ter or buffers at pH 4.6 is widely used as an indexof proteolysis. These extracts contain numeroussmall and medium-sized peptides, amino acidsand their degradation products, organic acidsand their salts, and NaCl. Extraction with waterefficiently separates the small peptides in cheesefrom proteins and large peptides. In Cheddar andmost other cheeses that are not cooked to a hightemperature, the principal proteolytic agents re-sponsible for the production of water-soluble ni-trogen (WSN) are the coagulant (principallychymosin) and to a lesser extent plasmin. Starterproteinases and peptidases play a relatively mi-nor role in the hydrolysis of intact caseins andlarge polypeptides to water-soluble peptides, al-though they are active on many of the peptidesproduced by chymosin or plasmin.

WSN, usually expressed as a percentage oftotal N, varies with variety and increasesthroughout ripening; a typical value for matureCheddar cheese is about 25% (see Chapter 11).Water is a suitable extractant only for cheesesthe pH of which changes little during ripening.However, many cheese varieties are character-ized by an increase in pH during ripening, andsince the extractability of N varies with pH, de-termination of the amount of N soluble in pH 4.6buffers is preferred (Figure 23-3). Other extrac-tants used to fractionate cheese N include solu-tions of CaCl2 and a mixture of chloroform andmethanol. The latter solvent is particularly suit-able for the extraction of hydrophobic peptides,which are often bitter.

Peptides in the primary extract from Cheddarcheese may be fractionated by a number of re-agents, including 70% (w/v) ethanol or 2% (w/v)trichloroacetic acid (TCA), which precipitateslarge and intermediate-sized peptides; 12% (w/v) TCA, which precipitates all but quite shortpeptides (Figure 23-3); and 5% (w/v) phospho-tungstic acid (PTA), in which only free amino

acids (except lysine and arginine) and very shortpeptides (less than ~ 600 Da) are soluble. PTA(5%)-soluble N is commonly used as an indexof total free amino acids (Figure 23-3). WSNmay also be fractionated by dialysis or by ultra-filtration (e.g., through 10 kDa membranes).

A number of rapid techniques can also beused to assess proteolysis. Cleavage of a peptidebond results in the liberation of an amino (-NH2)and a carboxylic acid group. Amino groups maybe quantified by reaction with trinitrobenzene-sulfonic acid (TNBS), ninhydrin, fluorescamine,or o-phthaldialdehyde (OPA). The Cd-ninhydrinreagent is particularly sensitive for the aminogroup of free amino acids and therefore is usedto estimate the liberation of amino acids incheese (Figure 23-4; Folkertsma & Fox, 1992).Cd-ninhydrin-reactive groups correlate wellwith cheese age (Figure 23-5). Other rapid as-says of proteolysis less widely used includemeasurement of absorbance at 280 nm (and thuspeptides containing tryptophan and/or tyrosineresidues) of TCA-soluble fractions of cheese orthe formation of ammonia or enzyme-based as-says for glutamic acid (see McSweeney & Fox,1997).

Although the above techniques are rapid andsuitable for routine analysis, the information ob-tained from nonspecific assays of proteolysis islimited. Since proteolysis results in the forma-tion of variously sized peptides, techniques thatresolve individual peptides provide much moreinformation. They are also more time consum-ing. Electrophoresis in polyacrylamide gels withalkaline urea-containing buffers (urea-PAGE) isthe most commonly used electrophoretic tech-nique for studying cheese (Figure 23-6). So-dium-dodecylsulfate (SDS)-PAGE, which sepa-rates proteins based on their molecular mass, iswidely used in biochemistry laboratories, butsince the caseins have similar molecular masses,the resolution of intact caseins in cheese is poor.However, SDS-PAGE has been used success-fully to separate the peptides produced from thecaseins by proteolysis. Urea-PAGE with directstaining using Coomassie Brilliant Blue G250 issuitable for assessing the initial hydrolysis of the

Page 551: Cheese Science

caseins by chymosin or plasmin and for monitor-ing the subsequent degradation of the large pep-tides produced. It is possible to approximatelyquantify bands on urea-PAGE by densitometry.Capillary electrophoresis (CE) involves the sep-aration of compounds in a buffer-filled capillaryunder the influence of an electric field (Figure23-7). CE is likely to be a valuable research toolin the future for studying proteolysis in cheeseand is currently being applied by a number oflaboratories.

Chromatography is widely used to resolvepeptides in cheese or cheese extracts. Anion-ex-change chromatography on diethylaminoethyl(DEAE) cellulose in the presence of urea is veryuseful for separating the larger peptides in Ched-dar cheese. High-performance liquid chroma-tography (e.g., FPLC, Pharmacia, Uppsala,Sweden) using a Mono-Q column is equivalentto DEAE cellulose and has the advantage that

peak areas may be quantified easily. Size-exclu-sion chromatography is also used to separateshorter peptides in cheese extracts. Reverse-phase (RP)-HPLC on C8 or Ci8 columns using anacetonitrile or water gradient with trifluoro-acetic acid as the ion-pair reagent and detectionat about 214 nm is very successful and is themethod of choice for separating the smaller wa-ter-soluble peptides from cheese (Figure 23-8).

The isolation of individual peptides fromcheese has necessitated the development ofvarious schemes to fractionate cheese N intorelatively homogeneous subfractions. One suchfractionation scheme, which was developed forCheddar cheese, is shown in Figure 23-9. Theisolation and identification (usually by N-termi-nal sequencing and/or mass spectrometry) ofpeptides from cheese have greatly increased theunderstanding of proteolysis at the molecularlevel in a number of varieties, particularly

Ripening time, weeks

Figure 23-3 Formation of pH 4.6-soluble N (•), 12% trichloroacetic acid-soluble N (•), and 5% phosphotung-stic acid (PTA)-soluble N (A) in an Irish farmhouse Blue cheese and formation of water-soluble N (Q) and PTA-soluble N (O) in Cheddar cheese.

%S

N/T

N

Page 552: Cheese Science

A50

7

Ripening time, weeks

Figure 23-4 Liberation of Cd-ninhydrin-reactive amino groups in Cheddar (•) and an Irish farmhouse Bluecheese (•) during ripening. Water-soluble (Cheddar) and pH 4.6-soluble (Blue cheese) fractions were analyzed.

mg

leu/

g pe

rmea

te

Age (months)

Figure 23-5 Total free amino acid concentrations (Cd-ninhydrin assay of 10 kDa untrafiltration permeates ofwater-soluble extracts) in Cheddar cheese as a function of age.

Page 553: Cheese Science

Electrolyte buffer

Figure 23-7 Schematic diagram of a capillary electrophoresis unit.

Sample

Anode

Capillary

Cathode

Detector

Dataacquisition

Capillary electrophoretogram

Figure 23-6 Urea-polyacrylamide gel electrophoretograms of Na casemate (c) and Cheddar cheeses at 1 day, 3months, and 6 months of ripening.

Page 554: Cheese Science

Minutes

Figure 23-8 Reverse-phase (C8) high-performance liquid chromatograms of the 70% ethanol-soluble (a) and-insoluble (b) fractions of a water-soluble extract from a 9-month-old Cheddar cheese. The peptides a rCN fl-9(X) and f 1 -13 (Y) are indicated.

a

b

Minutes

Page 555: Cheese Science

Cheddar, Gouda, and Parmesan (see Fox &McSweeney, 1996, and Chapter 11).

The ultimate products of proteolysis areamino acids. Many amino acids contribute di-rectly to the flavor of cheese but are perhapsmore important as precursors of other flavorcompounds. As discussed above, total free

amino acids can be estimated by rapid assays,while individual amino acids may be quantifiedby automated amino acid analyzers, which sepa-rate amino acids using an ion-exchange resin,followed by postcolumn derivitization (usuallywith ninhydrin) and spectrophotometric detec-tion. Alternatively, fluorescent amino acid de-

Figure 23-9 Fractionation scheme for cheese N. WSN, water-soluble N; WISN, water-insoluble N; DEAE, ion-exchange chromatography on DEAE-cellulose; PAGE, alkaline urea polyacrylamide gel electrophoresis; HPLC,reverse phase (C8) high-performance liquid chromatography.

HPLC

Amino Acids

HPLC

Electroblotting

Peptides

Sephadex G25

PAGEDEAE

RETENTATEor

INSOLUBLE FRACTION

PERMEATEor

SOLUBLE FRACTION

Diafiltration, 1OkDa membranesor

70% Ethanol

Ion Exchange ChromatographyUrea-PAGE

WISN

WSN (Discard)

FatRe-Extract

Pellet

Water Extraction

GRATED CHEESE

Sep-PakC8orC18 I II III IV V VI VII VIII IX X

Page 556: Cheese Science

rivatives (e.g., dansyl, OPA, or 7V-(9-fluorenyl-methoxycarbonyl)-) may be prepared and thenseparated and quantified by RP-HPLC. GC tech-niques have also been developed to quantify freeamino acids but are rarely used (see McSweeney&Fox, 1997).

Decarboxylation of amino acids leads to theformation of amines, many of which are biologi-cally active and have physiological importance.The principal amines in cheese are histamine,tyramine, tryptamine, putrescine, cadaverine,and phenylethylamine (see Chapter 21). Bio-genie amines may be determined spectrofluoro-metrically or by thin-layer chromatography butare now usually quantified by HPLC (often astheir fluorescent derivatives) or by GC. Volatileproducts of amino acid catabolism (see Chapter12) are usually quantified by GC interfaced withmass spectrometry (GC-MS).

23.5 TECHNIQUES TO STUDYVOLATILE FLAVOR COMPOUNDS

The most important aroma compounds inmost cheeses are volatile (see Chapter 12). Withthe exception of volatile fatty acids and somevolatile products of lactate or citrate metabolism(see Sections 23.4.1 and 23.4.2), cheese volatilesare almost universally quantified by variousforms of GC-MS. GC-MS is a very sensitivetechnique with impressive resolving power (seeFigure 12-8). Unfortunately, it is not suitable foruse in an industrial context owing to the highrunning and capital cost of the instrument andslow throughput of samples. A number of differ-ent techniques are used to prepare samples forGC-MS, including vacuum distillation (at~ 7O0C) and trapping in liquid N2 (an approachthat has been criticized because of artefact for-mation) and by trapping volatiles released intothe cheese headspace.

23.6 MICROBIOLOGICAL ANALYSIS OFCHEESE

Cheese is a complex microbial ecosystem thatmay contain starter bacteria (Lactococcus spp.,Leuconostoc spp., Streptococcus spp., and Lac-

tobacillus spp.), nonstarter lactic acid bacteria(Lactobacillus spp., Pediococcus spp., and En-terococcus spp.), non-lactic-acid bacteria (coli-forms, Staphylococcus aureus, Brevibacteriumspp., Micrococcus spp., Corynebacterium spp.,Microbacterium spp., and Arthrobacter spp.),yeast, and molds. Many cheese varieties containmost of these organisms. Only a limited numberof effective selective media are available forenumerating the various microorganisms incheese. Some of the more common media usedin cheese microbiology are listed in Table 23-1.

Some workers have determined total numbersof bacteria in cheese using plate count agar(PCA). Such procedures are not very useful,since cheese contains many different types ofbacteria, some of which die out (the starter)whereas others (the nonstarter LAB [NSLAB])grow during ripening. In addition, many starterand NSLAB either do not grow or grow onlypoorly on PCA. The reason for this is that thelevel of nutrients in PCA is too low to sustaingood growth of LAB, and its buffering capacityis poor. LAB are quite fastidious and requireseveral amino acids and vitamins for growth.Therefore, media must contain rich sources ofamino acids and peptides (peptones and yeastextract) and vitamins (yeast extract) and highlevels of buffer to neutralize the large amountsof lactate produced from sugar metabolism dur-ing growth.

Many media have been developed to enumer-ate starter bacteria. Today, the medium of choicefor enumerating lactococci is medium 17, con-taining lactose (LM-17), which was initially de-veloped by Terzaghi and Sandine (1975) to esti-mate lactococcal phage. It contains sufficientamounts of all the nutrients necessary to supportthe growth of lactococci and a high concentra-tion of |3-glycerophosphate (19 g/L) as a buffer,which, unlike phosphate, does not chelate theCa2+ required for adsorption of phage to its host.Although this medium is nonselective, it is usedto count lactococci in cheese because these bac-teria outnumber all other microorganisms in thecheese, especially during the early stages of rip-ening. As the cheese ripens, the medium be-comes less selective. This is shown in Table

Page 557: Cheese Science

23-2, based on a study in which the medium wasused to enumerate the lactococci in the Spanishcheese, Armada, during ripening (Tornadijo,Fresno, Bernardo, Martin-Sarmiento, &Carballo, 1995). In a study of the Portuguesecheese, Picante de Beira Baixa, 66% of isolatesfrom LM-17 were identified as enterococci(Freitas, Pais, Malcata, & Hogg, 1996).

Incubation of LM-17 plates at 450C makes itrelatively selective for Sc. thermophilus, sincelactococci do not grow at this temperature andthermophilic lactobacilli grow poorly, if at all, inthis medium. However, enterococci will grow atthis temperature, and colonies should be exam-ined to determine if they are enterococci or Sc.thermophilus.

Thermophilic cultures are often used today asadjuncts in commercial cheeses made with me-sophilic cultures. If high counts of Sc. ther-

mophilus (on LM-17 at 450C; see Table23-2) are found in a cheese made with a meso-philic culture, the count of Lactococcus (on LM-17 at 3O0C) must be adjusted, since Sc. ther-mophilus will also grow on LM-17 at 3O0C.Sometimes, the colonies of Sc. thermophilus aremuch smaller than those of Lactococcus spp.,and the difference in size could be used to differ-entiate between them, but a smaller size is by nomeans an absolute feature of these bacteria.

MRS agar is a general purpose medium devel-oped by de Man, Rogosa, and Sharpe (1960) forthe enumeration of lactobacilli. Most other LABcan grow in it, but reducing the pH to 5.4 andincreasing the temperature of incubation to 450Cmake it more or less selective for the thermo-philic lactobacilli found in starters. The meso-philic lactobacilli found in cheese are usuallycounted on Rogosa agar (RA) (Rogosa,

Table 23-1 Media and Incubation Conditions Used To Enumerate Different Types of Bacteria inCheese

TemperatureGroup Medium (0C) Incubation Time and Conditions

Lactococcus LM-17 Agara 30 3 days; spread or pour plate, aerobicLeuconostoc MRS + 20 ug vancomycin/mlb 30 3 days; spread or pour plate, aerobicSc. thermophilus LM-17 Agara 45 2 days; spread or pour plate, aerobicLb. helveticus MRS agar, pH 5.4 agar0 45 3 days; spread plate, anaerobicLb. lactis MRS agar, pH 5.4 agar 45 3 days; spread plate, anaerobicNonstarter LAB Rogosa agar (RA)C 30 5 days; pour plate, aerobic with

overlayCitrate utilizers Calcium-citrate agar1 (KCA) 30 2-3 days; spread plate,

anaerobicS. aureus Baird Parker agar 37 72 hr; spread plate, aerobicConforms Violet red bile agar 30 18 hr; pour plate, aerobicEnterococcus spp. Kanamycin aesculin azide 37 24 hr; pour plate, aerobic

agarSmear bacteria Plate count agar containing 70 25 7-10 days; spread plate, aerobic

g/L NaCIYeast/molds Yeast extract glucose chloram- 25 3-5 days; pour plate, aerobic

phenicol agara Terzaghi and Sandine (1975).b Mathot, Kihal, Prevost, and Divies (1994).c de Man, Rogosa, and Sharpe (1960).d Nickels and Leesment (1964).e Mossel, Baber, and Eeldering (1978).

Page 558: Cheese Science

Table 23-2 Selectivity of Various Agars for the Isolation of Different Lactic Acid Bacteria (Number of Isolates) from Cheese

Lactose Medium 1 7 agar Mayeaux, Sandine, & Elliker Agar Rogosa Agar Kanamycin Aesculin Agar(KAA)d(RA)P(MSE)b(LM- 17)a

(Weeks)TimeRipening(Weeks)TimeRipening(Weeks)TimeRipening(Weeks)TimeRipening168421168421168421168421Genus

6

2037404038

31383734

4

1

3638

1

31

7

1

1

4

21

6

7

5

15

8

4

1

16

2

15

2

7

13

1

9

3

5

23

5

3

15

7

11

2

25

11

1

3

28

1

1

Lactococcus

Lactobacillus

Leuconostoc

Enterococcus

Non-lactic-acidbacteria

a Terzaghi and Sandine (1975).b Mayeaux, Sandine, and Elliker (1962).c de Man, Rogosa, and Sharpe (1960).d Mossel, Baber, and Eeldering (1978).

Page 559: Cheese Science

Mitchell, & Wiseman, 1954). This medium con-tains a high concentration of acetate (0.225 M)and has a low pH (5.4), which make it quite se-lective for mesophilic lactobacilli. Some leuco-nostocs and pediococci may grow on it, but thethermophilic lactobacilli present in starters gen-erally do not grow on this medium. It is currentlywidely used in cheese microbiology and appearsto be quite selective for cheese containing differ-ent LAB (Table 23-2).

In the past, the medium of Mayeux, Sandine,and Elliker (1962) (MSE) was used to enumerateleuconostocs. This medium contains sucrose asthe energy source, and many authors consider itto be selective for Leuconostoc. This is not so.The medium is nutritionally rich, and many, ifnot all, LAB will grow on it (Table 23-2). It maybe selective if only dextran producers (very largecolonies) are counted, since dextran formation isgenerally confined to Leuconostoc spp. How-ever, not all leuconostocs produce dextrans.Since most leuconostocs are resistant to the anti-biotic vancomycin, the addition of vancomycin(20 (ig/ml) to an otherwise nutritionally ad-equate medium (e.g., MRS) makes it selectivefor these microorganisms. This method is ac-ceptable when applied to starters, but mesophiliclactobacilli and pediococci are also naturally re-sistant to vancomycin. Both mesophilic lactoba-cilli and pediococci are often found in largenumbers in cheese, which limits the usefulnessof the medium, unless colonies are also exam-ined microscopically and any doubtful ones ex-amined by additional tests.

Calcium citrate agar (Nickels & Leesment,1964) is very useful for enumerating citrate uti-lizers present in mesophilic cultures and dairyproducts. This is a differential, nonselective me-dium that is opaque owing to the presence of in-soluble calcium citrate. As the citrate is metabo-lized, a clear halo develops around the colony.This medium can also be used to estimate totalstarter numbers if triphenyltetrazolium chloride(TCC) is added to it (1 ml of a filter-sterilized1% [w/v] TTC solution per 100 ml of medium).TTC is reduced by the starter bacteria from acolorless soluble form to a red or pink insoluble

form that precipitates around the starter colo-nies, making them more readily apparent. Fur-ther details can be found in International DairyFederation (1997). If this medium is used to enu-merate citrate-utilizing bacteria in cheese, colo-nies surrounded by halos should be examinedmicroscopically, since many mesophilic lacto-bacilli, which are found in high numbers in rip-ened cheese, metabolize citrate and will producehalos around the colonies.

Numerous selective media have been pro-posed for enumerating enterococci, but none iscompletely reliable. Three commonly used me-dia are m-Enterococcus, KF, and kanamycinaesculin azide (KAA) agars. Ec. faecalis pro-duces dark red colonies on m-Enterococcusagar owing to reduction of TTC and precip-itation of formazan around the colony. Ec.durans, Ec. hirae, Ec. mundtii, and Streptococ-cus bovis reduce TTC less strongly than Ec.faecalis and produce pale pink colonies, whichare counted as enterococci. KF agar also con-tains TTC as an indicator of growth and NaN3

as a selective agent. Ec. faecalis and Ec.faecium produce red colonies, but many entero-cocci of nonfecal origin (Ec. mundtii, Ec.casseliflavus, Ec. pseudoavium, Ec. malo-doratus, and Ec. rqffinosus) and Sc. bovis alsodo so. KAA is a selective, differential medium.Kanamycin is the selective agent, and the dif-ferential reaction is the hydrolysis of aesculinto glucose and aesculetin, which gives a blackcolor in the presence of ferric citrate. Ec.faecalis and Ec. faecium hydrolyze aesculinand produce black colonies in the medium. Itsselectivity for the newer Enterococcus spp. andfor enterococci of nonfecal origin has not beenstudied. However, some lactobacilli do grow onKAA (Table 23-2).

There are no specific selective media for thebacteria found in the smear on cheese. All thesebacteria are salt tolerant, so the addition of NaCl(e.g., 70 g/L) to an otherwise suitable medium(e.g., plate count agar) will inhibit the starterbacteria without affecting the smear bacteria.These bacteria grow slowly, and therefore platesmust be incubated for at least 7 days.

Page 560: Cheese Science

Good selective media are available for Staph.aureus (Baird Parker agar) and coliforms (violetred bile agar). Yeasts and molds can be enumer-ated on potato dextrose agar acidified to pH 3.5with lactic acid or on yeast glucose chlorotetra-cycline agar. The antibiotic (chlorotetracycline)inhibits the bacteria without affecting yeast andmolds. Yeasts and molds are easily distin-guished, since the latter produce large fluffycolonies rather than the small, opaque, some-times glistening colonies of yeast.

23.7 OBJECTIVE ASSESSMENT OFCHEESE TEXTURE

The structure of cheese is formed by a caseinnetwork in brine containing fat globules (seeChapter 13 and Prentice, Langley, & Marshall,1993). The moisture in cheese acts as a plasti-cizer, and its fat contributes to the rheologicalproperties of cheese to an extent determined bythe liquid fatsolid fat ratio. The uniformity ofstructure varies greatly between cheese varietiesdue to differences in the manufacturing protocol.

Since texture is a mechanical property of thecheese, it is more amenable to instrumentalanalysis than is cheese flavor, which resultsfrom the combination of a large number of sapidand volatile compounds (see Chapter 12). How-ever, any instrumental technique used to mea-sure cheese texture should involve fracture andmimic the mechanisms involved in sensoricgrading or consumption. Obtaining representa-tive samples for physical measurement is a prob-lem when studying cheese texture. Techniquesused to study cheese texture are discussed inChapter 13.

23.8 SENSORY ANALYSIS OF CHEESEFLAVOR AND TEXTURE

The most important indices of cheese ripeningare those on which the consumer decides to pur-chase the product: flavor, texture, and appear-ance. Ultimately, the quality of cheese is bestassessed by sensory analysis (see Delahunty &Murray, 1997, and Muir, Banks, & Hunter,

1995, for reviews). A useful guide for the sen-sory evaluation of the texture of hard cheeseswas prepared by Lavanchy et al. (1994). Textson general sensory assessment of food productsinclude Stone and Sidel (1993) and Piggott(1988).

Sensory analysis of cheese has, traditionally,been performed by individual cheese graders orby small groups of "experts" with the objectiveof assessing the potential of a cheese to developa mature flavor or determining the quality at thepoint of sale. In such systems, marks are nor-mally deducted for defects according to theirintensity. Although characteristics sought by ex-pert judges are usually those desired by consum-ers, grades awarded do not always correlate withconsumer preference. Expert grading systemsare widely used in the cheese industry and are,presumably, sufficiently reliable to ensure con-sistent product quality, but they are less useful inresearch.

To assess whether there is a difference be-tween cheeses, simple difference tests, such asthe paired comparison or triangle tests, may beadequate. However, such tests always leavequestions unanswered about the attributes of thecheese and to what extent they vary. The Inter-national Dairy Federation (1987) method forcheese grading involves comparing the cheese toa reference cheese selected as a "standard" andrating the cheese in terms of a list of defects.However, the selection of the standard cheese issubjective, and it is difficult to maintain a con-sistent reference standard.

Descriptive evaluation by a panel of trainedassessors (10-20 persons) is the method ofchoice for determining the sensory profile of acheese and can determine the influence of pro-cessing changes on individual sensory charac-teristics. Such panels are established by select-ing potential assessors based on their ability toperceive sensory stimuli. Assessors taste a rangeof cheeses and during group discussions agreeon and then define a list of descriptors that en-compass the sensory attributes of the cheese(Exhibit 23-1). In an alternative approach (freechoice profiling), assessors use a personal list of

Page 561: Cheese Science

descriptors. The assessors are then trained toquantify the intensity of each descriptor andscore their perceptions by placing marks on aline scale with defined upper and lower limits.Trained assessors should be able to differentiatebetween attributes, should give reproduciblescores, and should agree with other panelists.Assessors who cannot meet these criteria shouldbe removed from the panel. Descriptive sensory

evaluation generates much data concerning theproduct and computer-assisted data capture istherefore beneficial. Data are analyzed statisti-cally by analysis of variance or, preferably, bymultivariate techniques such as principal com-ponent analysis, which presents data as a multi-dimensional matrix that explains as much of thevariance as possible (Figure 23-10). Consider-able work is involved in establishing a reliable

Cheddary: Flavor you associate with a typical Cheddar

Creamy: Containing cream, resembling cream

Buttery: Of the nature of or containing butter

Pungent: Physically penetrating sensation in the nasal cavity; sharp smelling or tasting irritant

Moldy: The combination of aromatics generally associated with molds; they are usually earthy,dirty, stale, musty, and slightly sour

Caramel: Burnt sugar or syrup; toffee made with sugar that has been melted further

Burnt caramel: Sweet flavor notes, dominated by a taste not unlike burnt milk

Soapy: Detergentlike, similar to that of food tainted with a cleansing agent

Smoky: The penetrating acrid aromatic of charred wood; tainted by exposure to smoke

Fruity: The aromatic blend of different fruity identities

Mushroom: Organic; the aromatics associated with raw mushrooms

Rancid: Sour milk, fatty, oxidized; having the rank unpleasant taste or smell characteristic ofoils and fats when no longer fresh

Nutty: The nonspecific nutlike aromatic that is characteristic of several different nuts (e.g.,peanuts, hazelnuts, pecans)

Sweaty: The aromatics reminiscent of perspiration-generated food odor; sour, stale, slightlycheesy, moist, stained or odorous with sweat

Balanced: Mellow, smooth, clean; in equilibrium, well arranged or disposed, with no constituentlacking or in excess

Processed: Tastes of plastic, packaging, shallow; to taste artificial; made by melting, blending, andfrequently emulsifying other cheeses

Salty: Fundamental taste sensation of which sodium chloride is typical

Sweet: Fundamental taste sensation of which sucrose is typical

Acidic: Sour, tangy, sharp, citruslike; the fundamental taste sensations of which lactic acid andcitric acids are typical

Bitter: Chemical-like, aspirin; taste sensations of which caffeine and quinine are typical

Strength: Intensity of flavor, mildness, and maturity

Astringent: Mouth-drying, harsh; the complex of drying, puckering, and shrinking sensations in thelower cavity causing contraction of the body tissues

Exhibit 23-1 A Descriptive Vocabulary Used To Characterize Cheddar Cheese Flavor

Page 562: Cheese Science

descriptive panel for the sensory analysis ofcheese and the effort involved is often underesti-mated by persons unfamiliar with sensory sci-ence.

The question of which cheeses consumersprefer is a very important but subjective consid-eration and can be answered only by an un-trained consumer panel, representative of thetarget market for the cheese. Since a random anduntrained group of individuals will vary widelyin their preferences and their ability to perceivestimuli, it is essential to have sufficient peopleon such a panel to obtain reliable results. A mini-mum number of 50-60 targeted consumers hasbeen recommended.

The power of any sensory analysis techniquewill be increased if external variables are ex-cluded. Sensory assessment should be per-formed in a sensory laboratory equipped withindividual booths with controlled lighting andventilation. The order in which cheeses aretasted is very important: the first cheese a panel-ist tastes is likely to receive a higher score thansubsequent cheeses and the perception of a mild-flavored cheese could be influenced by a stron-ger-tasting cheese tasted immediately before (orvice versa). Order of tasting effects should beeliminated by randomizing the order in whichthe cheeses are presented and balancing the pre-sentation so that each cheese is tasted an equal

PC 1(45%)Figure 23-10 Principal component scores for 33 cheeses of varying fat content and loadings of descriptors (initalics and plotted on the scale between -0.5 and 0.5) on the first two components from principal componentsanalysis of descriptive sensory data. F, full-fat Cheddar; R, reduced-fat Cheddar; H, half-fat Cheddar; FE, full-fatEdam; HE, half-fat Edam. Cheeses in bold were mature.

processed

sweaty

caramelPC2

(14%

)cheddary

balancedcreamy

strengthbitter

nutty

Page 563: Cheese Science

number of times by the panel. Assessor fatiguemust also be considered; there is a limit (unde-fined) to the number of cheeses that may betasted at one session. Assessors should not haveprior knowledge of the cheeses, nor of the ex-pected or desired results. However, such a situa-tion may be difficult to achieve with in-housepanels.

Obtaining reliable results from sensory analy-sis of cheese is far from straightforward. It is de-ceptively simple to establish a taste panel, butpractical problems may arise that negate the bestefforts of persons inexperienced in sensoryanalysis. It is therefore desirable that a personthoroughly familiar with sensory science and itspitfalls be responsible for the sensory assess-ment of cheese.

23.9 DETECTION OF INTERSPECIESADULTERATION OF MILKS ANDCHEESES

Adulteration of ewe or goat milk for cheese-making with less expensive bovine milk has ledto the development of techniques that permit thedetection of mixed-species milks (seeMcSweeney & Fox, 1993). Various authorshave used differences in fatty acid profiles todetect adulteration. Such methods have limitedsensitivity and will not detect adulteration withskim milk, but they can be applied to maturecheese, since relatively small changes occur infatty acid composition during maturation. Dif-ferences in triglycerides may also be used toidentify milks from different species. The ab-sence of p-carotene from sheep and goat milkmay permit the detection of bovine milk in ca-prine or ovine milk.

REFERENCES

Baer, A., & Collin, J.C. (1993). Determination of residualactivity of milk-clotting enzymes in cheese: Specific iden-tification ofchymosin and its substitutes in cheese [Bulle-tin No. 284], Brussels: International Dairy Federation.

Differences between HPLC profiles of tryptichydrolysates of caseins from different specieshave been demonstrated. However, the effect ofproteolysis during cheese ripening may interferewith this approach. Electrophoretic techniqueshave also been developed to detect mixtures ofmilks from different species. However, pro-teolysis during cheese ripening may interferewith these procedures. Interspecies differencesbetween para-K-caseins (which are degradedonly slightly or not at all during ripening) ory-caseins detected by isoelectric focusing havealso been used as indices of adulteration. Inter-species differences in the electrophoretic mobil-ity of whey proteins may be used to identifymixed-species milks but not cheeses.

The higher xanthine oxidase activity of bo-vine milk has been used to detect the adultera-tion of goat milk with cow milk. The test issimple and rapid and can be applied to raw orpasteurized (< 750C x 20 s) milk or fresh cheese;it allows detection of 2% added bovine milk.Differences in the Ca:Mg ratio in cheeses madefrom cow or sheep milk have also been used asan index of adulteration.

Immunological methods are well suited foranalysis of mixed-species milks owing to theirsensitivity and specificity and their availabilityin kit form. Antisera prepared against proteinsfrom a particular milk may react not only withproteins from that milk but with the milks ofother species also, but techniques have been de-veloped to overcome this. Caseins have rel-atively low antigenicity, but immunodottingtechniques have been used to overcome thisproblem. An indirect enzyme-linked immuno-sorbent assay (ELISA) for bovine caseins hasbeen developed and used to assay for bovinemilk in sheep milk and cheese.

Cunniff, P. (Ed.). (1995). Official methods of analysis ofAOAC International (16th ed., VoIs. 1, 2). Arlington,VA: Association of Official Analytical Chemists Interna-tional.

Page 564: Cheese Science

Delahunty, C.M., & Murray, J.M. (1997). Organolepticevaluation of cheese. In Proceedings of the Fifth CheeseSymposium. Dublin: Teagasc.

de Man, J.C., Rogosa, M., & Sharpe, M.E. (1960). A me-dium for the culturation of lactobacilli. Journal of Ap-plied Bacteriology, 23, 130-135.

Folkertsma, B., & Fox, P.P. (1992). Use of Cd-ninhydrin re-agent to assess proteolysis in cheese during ripening.Journal of Dairy Research, 59, 219-224.

Fox, P.P., & McSweeney, P.L.H. (1996). Proteolysis incheese during ripening. Food Reviews International, 12,457-509.

Fox, P.P., McSweeney, P.L.H., & Singh, T.K. (1995). Meth-ods for assessing proteolysis in cheese during ripening. InE.L. Malin & M.H. Tunick (Eds.), Chemistry of struc-ture-function relationships in cheese. New York: Ple-num.

Freitas, A.C., Pais, C., Malcata, F.X., & Hogg, T.A. (1996).Microbiological characterisation of Picante de BeiraBaixa cheese. Journal of Food Protection, 59, 155-160.

International Dairy Federation. (1985). Milk and milk prod-ucts. Methods of sampling [Standard 5OB]. Brussels: Au-thor.

International Dairy Federation. (1987). Sensory evaluationof dairy products [Standard 99A]. Brussels: Author.

International Dairy Federation. (1995). Standards. In Cata-logue of IDFpublications. Brussels: Author.

International Dairy Federation. (1997). Dairy starter culturesof lactic acid bacteria (LAB): Standard of identity [Stan-dard 149Ax]. Brussels: Author.

Lavanchy, A., Berodier, F., Zannoni, M., Noel, Y., Adamo,C., Sequella, J., & Herrero, L. (1994). A guide to the sen-sory evaluation of texture of hard and semi-hard cheeses.Paris: Institut National de Ia Recherche Agronomique.

Marcos, A. (1993). Water activity in cheese in relation tocomposition, stability and safety. In P.F. Fox (Ed.),Cheese: Chemistry, physics and microbiology (2d ed.,Vol. 1). London: Chapman & Hall.

Marier, J.R., & Boulet, M. (1958). Direct determination ofcitric acid in milk with an improved pyridine-acetic anhy-dride method. Journal of Dairy Science, 41, 1683-1692.

Mathot, A.G., Kihal, M., Prevost, H., & Divies, C. (1994).Selective enumeration of Leuconostoc on vancomycinagar media. International Dairy Journal, 4, 459-469.

Mayeux, J.V., Sandine, W.E., & Elliker, P.R. (1962). A se-lective medium for detecting Leuconostoc organisms inmixed strain starter cultures. Journal of Dairy Science,45, 655-656.

McSweeney, P.L.H., & Fox, P.F. (1993). Cheese: Methodsof chemical analysis. In P.F. Fox (Ed.), Cheese: Chemis-try, physics and microbiology (2d ed., Vol. 1). London:Chapman & Hall.

McSweeney, P.L.H., & Fox, P.F. (1997). Chemical methodsfor the characterization of proteolysis in cheese duringripening. Lait, 77, 41-76.

Mohr, B., Rea, M.C., & Cogan, T.M. (1997). A new methodfor the determination of 2-acetolactic acid in dairy prod-ucts. International Dairy Journal, 7, 701-706.

Mossel, D.A.A., Baber, P.G.H., & Eeldering, J. (1978).Streptokokken der Lancefleld Gruppen D in Lebensmittelund Trinkwasser: Ihre Bedeutung, Erfassung undBekampfung. Archiv fur Lebensmittelhygiene, 29, 121-127.

Muir, D., Banks, J.M., & Hunter, E.A. (1995). Sensory prop-erties of cheese. In Proceedings of the Fourth CheeseSymposium. Dublin: Teagasc.

Nickels, C., & Leesment, H. (1964). Methode zur dif-ferenziearung und quantitativen Bestimmung vonSaurewecberbakterien. Milchwissenschaft, 19, 374-378.

Piggott, J.R. (1988). Sensory analyses of foods (2d ed.). Lon-don: Elsevier Applied Sciences.

Prentice, J.H., Langley, K.R., & Marshall, RJ. (1993).Cheese rheology. In P.F. Fox (Ed.), Cheese: Physics,chemistry and microbiology (2d ed., Vol. 1). London:Chapman & Hall.

Richelieu, M., Hoalberg, U., & Nielsen, J.C. (1997). Deter-mination of a-acetolactic acid and volatile compounds byhead-space gas chromatography. Journal of Dairy Sci-ence, 80, 1918-1925.

Rogosa, M., Mitchell, J.A., & Wiseman, R.F. (1954). A se-lective medium for the isolation and enumeration of orallactobacilli. Journal of Dental Research, 10, 682-689.

Stone, H., & Sidel, J.L. (1993). Sensory evaluation practices(2d ed.). New York: Academic Press.

Terzaghi, B.E., & Sandine, W.E. (1975). Improved mediumfor lactic streptococci and their bacteriophages. Appliedand Environmental Microbiology, 29, 807-813.

Tornadijo, M.E., Fresno, J.M., Bernardo, A., Martin-Sarmiento, R., & Carballo, J. (1995). Microbiologicalchanges throughout the manufacturing and ripening of aSpanish goat's raw milk cheese (Armada variety). Lait,75,551-570.

Walsh, B., & Cogan, T.M. (1974). Separation and estimationof diacetyl and acetoin in milk. Journal of Dairy Re-search, 41, 25-30.

Page 565: Cheese Science

CHAPTER 1

Table 1-1. Source: Data from Scott, 1986.Table 1-2. Courtesy of the Food and Agricul-

tural Organization of the United Nations,1997, Rome, Italy.

Table 1-3. Courtesy of the International DairyFederation, 1995, Brussels, Belgium.

CHAPTER 2

Figure 2-2. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, pp. 226-227,© 1998, Aspen Publishers, Inc.

Figure 2-3. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 460, © 1998,Aspen Publishers, Inc.

CHAPTER 3

Table 3-1. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 2, © 1998, AspenPublishers, Inc.

Table 3-2. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 21, © 1998, AspenPublishers, Inc.

Figure 3-1. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 22, © 1998, AspenPublishers, Inc.

Figure 3-2. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 24, © 1998, AspenPublishers, Inc.

Figure 3-3. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 25, © 1998, AspenPublishers, Inc.

Figure 3-4. Source: Data from Jenness and Patton,1959.

Table 3-3. Source: Reprinted with permission fromW.W. Christie, Composition and Structure ofMilk Lipids, in Advanced Dairy Chemistry, Vol. 2,Lipids, P.F. Fox, ed., pp. 1-36, © 1995, AspenPublishers, Inc.

Table 3-4. Source: Reprinted with permission fromW.W. Christie, Composition and Structure ofMilk Lipids, in Advanced Dairy Chemistry, Vol. 2,Lipids, P.F. Fox, ed., pp. 1-36, © 1995, AspenPublishers, Inc.

Table 3-5. Source: Reprinted with permission fromW.W. Christie, Composition and Structure ofMilk Lipids, in Advanced Dairy Chemistry, Vol. 2,Lipids, P.F. Fox, ed., pp. 1-36, © 1995, AspenPublishers, Inc.

Table 3-6. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 95, © 1998, AspenPublishers, Inc.

Table 3-7. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 148, © 1998, AspenPublishers, Inc.

Figure 3-5. Source: Data from Bernhart, 1961.Table 3-8. Source: Reprinted with permission from

P.F. Fox and P.H.L. McSweeney, Dairy Chemis-

Table of Sources

Page 566: Cheese Science

try and Biochemistry, p. 164, © 1998, AspenPublishers, Inc.

Figure 3-6. Source: Reprinted with permission fromH.E. Swaisgood, Chemistry of the Caseins, inAdvanced Dairy Chemistry, Vol. 1, Proteins, P.P.Fox, ed., pp. 63-110, ©1992.

Figure 3-7. Source: Reprinted with permission fromH.E. Swaisgood, Chemistry of the Caseins, inAdvanced Dairy Chemistry, Vol. 1, Proteins, P.P.Fox, ed., pp. 63-110, © 1992.

Figure 3-8. Source: Reprinted with permission fromH.E. Swaisgood, Chemistry of the Caseins, inAdvanced Dairy Chemistry, Vol. 1, Proteins, P.P.Fox, ed., pp. 63-110, ©1992.

Figure 3-9. Source: Reprinted with permission fromH.E. Swaisgood, Chemistry of the Caseins, inAdvanced Dairy Chemistry, Vol. 1, Proteins, P.P.Fox, ed., pp. 63-110, ©1992.

Figure 3-10. Source: Reprinted with permissionfrom H.E. Swaisgood, Chemistry of the Caseins,in Advanced Dairy Chemistry, Vol. 1, Proteins,P.P. Fox, ed., pp. 63-110, © 1992.

Figure 3—11. Source: Reprinted with permissionfrom P.P. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 161, © 1998,Aspen Publishers, Inc.

Table 3-9. Source: Reprinted with permission fromJ. McMahon and RJ. Brown, Composition andIntegrity of Casein Micelles: A Review, Journalof Dairy Science, Vol. 67, pp. 499-512, © 1984,American Dairy Science Association.

Figure 3-12. Source: P. Walstra and R. Jenness,Dairy Chemistry and Physics, © 1984, JohnWiley and Sons, Inc. Reprinted by permission ofJohn Wiley and Sons, Inc.

Figure 3-13. Source: Data from Holt, 1994.Table 3-10. Source: Reprinted with permission from

P.P. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 254, © 1998, AspenPublishers, Inc.

Table 3—11. Source: Reprinted with permission fromP.P. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 437, © 1998, AspenPublishers, Inc.

CHAPTER 4

Figure 4-2. Source: Reprinted with permission fromAJ. Bramley and C.H. McKinnon, The Microbi-ology of Raw Milk, Dairy Microbiology, 2nd Ed.,Vol.1, pp. 163-208, © 1990.

Figure 4-3. Source: Reprinted with permission fromR.K. Robinson, Modern Dairy Technology 2ndEd., Vol.!,©1994.

Figure 4-6. Source: From Principles of Biochemis-try by Lehninger, Nelson, Cox, © 1993, 1982 byWorth Publishers, Inc. Used with permission.

Figure 4-7. Courtesy of Westfalia Separator, Inc.,Northvale, New Jersey.

CHAPTER 5

Figure 5-1. Source: Reprinted with permission fromT. Lodics and L. Steenson, Characterisation ofBacteriophages and Bacteria Indigenous to aMixed-strain Cheese Starter, Journal of DairyScience, Vol. 73, pp. 2685-2696, © 1990,American Dairy Science Association.

Figure 5-2. Source: Reprinted with permission fromAccolas et al, Etude des Iteractions Entre DiverseBacteries Lactiques Thermophiles et Mesophiles,en Realation avec Ia Fabrication des Fromages aPate Cuite, Le Lait, Vol. 51, pp. 249-272, © 1971,Editions Scientifiques et Medicales Elsevier.

Figure 5-3. Source: Reprinted with permission fromK. Cooper and E.B. Collins, Influence ofTemperature on Growth of Leuconostoc cremoris,Journal of Dairy Science, Vol. 61, pp. 1085-1088,1978; D.A. Lee and E.B. Collins, Influence ofTemperature on Growth of Streptococcuscremoris and Streptococcus lactis, Journal ofDairy Science, Vol. 59, pp. 405-409, © 1976; andF.G. Martley, Temperature Sensitivities ofThermophilic Starter Strains, New ZealandJournal of Dairy Science Technology, Vol. 18, pp.191-196, ©1983.

Figure 5-4. Source: Reprinted with permission fromV. Bottazzi and F. Bianchi, I MicrorganismiLattiero-caseari al Micrscopio Elettronico aScansione, © 1984, Edi-ermes.

Figure 5-5. Courtesy of M. Vancanneyt, Universityof Ghent, Belgium.

Table 5-5. Source: Reprinted from Kunji et al, TheProteolytic Systems of Lactic Acid Bacteria,Antonie van Leeuwenhoek, Vol. 70, pp. 187-221,© 1996, with kind permission from KluwerAcademic Publishers.

Figure 5-7. Source: Reprinted from FEMS Microbi-ology Review, Vol. 12, B. Poolman, EnergyTransduction in Lactic Acid Bacteria, pp. 125-148, © 1993, with permission from ElsevierScience.

Page 567: Cheese Science

Table 5-7. Source: Data from Cogan, 1972.Figure 5-14. Courtesy of H. Neve, Institute of

Microbiology, Federal Dairy Research Center,Kiel, Germany.

Figure 5-15. Courtesy of H. Neve, Institute ofMicrobiology, Federal Dairy Research Center,Kiel, Germany.

Figure 5-16. Courtesy of H. Neve, Institute ofMicrobiology, Federal Dairy Research Center,Kiel, Germany.

Figure 5-17. Source: Reprinted with permissionfrom Pearce et al, Bacteriophage MultiplicationCharacteristics in Cheddar Cheesemaking, NewZealand Journal of Dairy Science Technology,Vol. 5, pp. 145-150, © 1970, New Zealand DairyResearch Institute.

Figure 5-18. Source: Reprinted with permissionfrom R. Klaenhammer, Genetic Characterizationof Multiple Mechanisms of Phage Defense from aPrototype Phage-insensitive Strain, Lactococcuslactis ME2, Journal of Dairy Science, Vol. 72, pp.3429-3443, © 1989, American Dairy ScienceAssociation.

Figure 5-20. Courtesy of H. Neve, Institute ofMicrobiology, Federal Dairy Research Center,Kiel, Germany.

Figure 5-21. Courtesy of the Netherlands Institutefor Dairy Research.

CHAPTER 6

Figure 6-2. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 383, © 1998, AspenPublishers, Inc.

Table 6-1. Source: Reprinted with permission fromP.F .Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 384, © 1998, AspenPublishers, Inc.

Figure 6-4. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 386, © 1998, AspenPublishers, Inc.

Figure 6-8. Source: Reprinted with permission fromGuinee et al, Ultrafiltration in Cheesemaking,Proceedings of the 3rd Cheese Symposium, T.M.Cogan, ed., pp. 49-59, © 1992, Dairy ProductsResearch Center, Moorepark, Fermoy, Ireland.

Figure 6-9. Source: Reprinted from InternationalDairy Journal, Vol. 41, vanHooydonk et al, The

Renneting Properties of Heated Milk, pp. 3-18,© 1987 with permission from Elsevier Science.

Figure 6-10. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 389, © 1998,Aspen Publishers, Inc.

Figure 6-11. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 390, © 1998,Aspen Publishers, Inc.

Figure 6-13. Source: Data from Bohlin RheologicalVOR Manual, Gloucestershire, England.

Figure 6-14. Source: Reprinted with permissionfrom JJ. Mayes and BJ. Sutherland, FurtherNotes on Coagulum Firmness and Yield inCheddar Cheese Manufacture, Australian Journalof Dairy Technology, Vol. 44, pp. 47-48, © 1989.

Figure 6-15. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 390, © 1998,Aspen Publishers, Inc.

Figure 6-16. Source: Reprinted with permissionfrom F.A. Payne, Automatic Control of Coagula-tion Cutting in Cheese Manufacture, AppliedEngineering in Agriculture, Vol. 11, pp. 691-697,© 1995, American Society of AgriculturalEngineers.

Figure 6-17. Source: Reprinted with permissionfrom F.A. Payne, Automatic Control of Coagula-tion Cutting in Cheese Manufacture, AppliedEngineering in Agriculture, © 1995, AmericanSociety of Agricultural Engineers.

Figure 6-18. Source: Reprinted with permissionfrom F. A. Payne, Automatic Control of Coagula-tion Cutting in Cheese Manufacture, AppliedEngineering in Agriculture, © 1995, AmericanSociety of Agricultural Engineers.

Figure 6—19. Source: Guinee et al, Effect of MilkProtein Standardization by Ultrafiltration on theManufacture, Composition and Maturation ofCheddar Cheese, Journal of Dairy Research, Vol.61, pp. 117-131, © 1994. Reprinted with thepermission of Cambridge University Press.

Figure 6—20. Source: Reprinted with permissionfrom Guinee et al, The Effects of Compositionand Some Processing Treatments on the RennetCoagulation Properties of Milk, InternationalJournal of Dairy Technology, Vol. 50, pp. 99-106, © 1997, Society of Dairy Technology.

Figure 6-21. Source: Reprinted with permissionfrom Guinee et al, The Effects of Composition

Page 568: Cheese Science

and Some Processing Treatments on the RennetCoagulation Properties of Milk, InternationalJournal of Dairy Technology, Vol. 50, pp. 99-106, © 1997, Society of Dairy Technology.

Figure 6-23. Source: Data from Fox, 1969 andPhelan, 1973.

Figure 6-24. Source: Reprinted with permissionfrom Emmons et al, Milk Clotting Enzymes. 1.Proteolysis During Cheese Making in Relation toEstimated Losses of Yield, Journal of DairyScience, Vol. 73, pp. 2007-2015, © 1990,American Dairy Science Association.

Figure 6-25. Source: Reprinted from P.F. Fox, Milk-clotting and Proteolytic Activities of Rennet, andof Bovine Pepsin and Porcine Pepsin, Journal ofDairy Research, Vol. 36, pp. 427^33, © 1969.Reprinted with the permission of CambridgeUniversity Press; and J.A. Phelan, Laboratory andField Tests on New Milk Coagulants, DairyIndustries International, Vol. 38, pp. 419—424,© 1973, Wilmington Publishing Ltd.

Figure 6-26. Source: Reprinted with permissionfrom Thunell et al, Thermal Inactivation ofResidual Milk Clotting Enzymes in Whey,Journal of Dairy Science, Vol. 62, pp. 373-377,© 1979, American Dairy Science Association.

CHAPTER 7

Figure 7-3. Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 393, © 1998, AspenPublishers, Inc.

CHAPTER 8

Figure 8-1. Source: Reprinted from O.R. Fennema,Food Chemistry, 3rd Ed., © 1996, by courtesy ofMarcel Dekker, Inc.

Table 8-3. Source: Reprinted with permission fromY. Roos, Water Activity in Milk Products, inAdvanced Dairy Chemistry, Vol. 3, pp. 306-346,© 1997, Aspen Publishers, Inc.

Figure 8-2. Source: Reprinted with permission fromY. Roos, Water Activity in Milk Products, inAdvanced Dairy Chemistry, Vol. 3, pp. 306-346,© 1997, Aspen Publishers, Inc.

Figure 8-4. Source: Reprinted with permission fromR.C. Lawrence and J. Gilles, Factors that

Determine the pH of Young Cheddar Cheese, NewZealand Journal of Dairy Science Technology,Vol. 17, pp. 1-14, © 1982, New Zealand DairyResearch Institute.

Figure 8-5. Source: Reprinted with permission fromT.P. Guinee and P.F. Fox, Influence of CheeseGeometry on the Movement of Sodium Chlorideand Water During Brining, Irish Journal of FoodScience Technology, Vol. 10, pp. 73-96, © 1986a,Teagasc.

Figure 8-7. Source: Reprinted from Food Chemistry,Vol. 19, T.P. Guinee and P.F. Fox, Transport ofSodium Chloride and Water in Romano CheeseSlices During Brining, pp. 49-64, © 1986b, withpermission from Elsevier Science.

Figure 8-8. Source: Reprinted with permission fromH.A. Morris, T.P. Guinee and P.F. Fox, Journal ofDairy Science, Vol. 68, p. 1851, © 1985,American Dairy Science Association.

Figure 8-9. Source: Reprinted from InternationalDairy Journal, Vol. 28, TJ. Geurts, P. Walstra,and H. Mulder, Transport of Salt and WaterDuring Salting of Cheese. 1. Analysis of theProcesses Involved, p. 106, © 1974, withpermission from Elsevier Science.

Figure 8-10. Source: Reprinted with permissionfrom T.P. Guinee, Studies on the Movements ofSodium Chloride and Water in Cheese and theEffects Thereof on Cheese Ripening, PhD. Thesis,© 1985.

Figure 8-11. Source: Data from Thomas and Pearce,1981.

Figure 8-12. Source: Data from O'Connor, 1974.

CHAPTER 9

Figure 9-2. Source: Reprinted with permission fromJ. Gilles and R.C. Lawrence, The Yield of Cheese,New Zealand Journal of Dairy Science Technol-ogy, Vol. 20, pp. 205-214, © 1985, New ZealandDairy Research Institute.

Figure 9-4. Source: Guinee et al, Effect of MilkProtein Standardization by Ultrafiltration on theManufacture, Composition and Maturation ofCheddar Cheese, Journal of Dairy Research, Vol.61, pp. 117-131, © 1994. Reprinted with thepermission of Cambridge University Press.

Figure 9-5. Source: Guinee et al, Effect of MilkProtein Standardization by Ultrafiltration on the

Page 569: Cheese Science

Manufacture, Composition and Maturation ofCheddar Cheese, Journal of Dairy Research, Vol.61, pp. 117-131, © 1994. Reprinted with thepermission of Cambridge University Press.

Figure 9-6. Source: Reprinted from Guinee et al,Milk Protein Standardization by Ultrafiltration forCheddar Cheese Manufacture, Journal of DairyResearch, Vol. 63, pp. 281-293, © 1996.Reprinted with the permission of CambridgeUniversity Press.

Figure 9-7. Source: Reprinted with permission fromI. Politis and K.F. Ng-Kwai-Hang, AssociationBetween Somatic Cell Count of Milk and Cheese-yielding Capacity, Journal of Dairy Science, Vol.71, pp. 1720-1727, © 1988a, American DairyScience Association.

Figure 9-8. Source: Reprinted with permission fromI. Politis and K.F. Ng-Kwai-Hang, Effects ofSomatic Cell Counts and Milk Composition on theCoagulating Properties of Milk, Journal of DairyScience, Vol. 71, pp. 1740-1746, © 1988b,American Dairy Science Association.

Figure 9-9. Source: Reprinted with permission fromI. Politis and K.F. Ng-Kwai-Hang, AssociationBetween Somatic Cell Count of Milk and Cheese-yielding Capacity, Journal of Dairy Science, Vol.71, pp. 1720-1727, © 1988a, American DairyScience Association.

Figure 9-10. Source: WJ. Donnelly and J.G. Barry,Casein Compositional Studies.III. Changes inIrish Milk for Manufacturing and Role of MilkProteinase, Journal of Dairy Research, Vol. 50,pp. 433-441, © 1983. Reprinted with thepermission of Cambridge University Press.

Figure 9-11. Source: Reprinted with permissionfrom Politis et al, Plasmin and Plasminogen inBovine Milk: A Relationship with Involution,Journal of Dairy Science, Vol. 72, pp. 900-906,© 1989, American Dairy Science Association.

Figure 9-12. Source: Reprinted from InternationalDairy Journal, Vol. 8, Influence of K-CaseinGenetic Variant on Rennet Gel Microstructure,Cheddar Cheesemaking Properties and CaseinMicelle Size, pp. 707-714, © 1998, with permis-sion from Elsevier Science.

Figure 9-13. Source: Reprinted with permissionfrom Hicks et al, Psychrotrophic Bacteria ReduceCheese Yield, Journal of Food Protection, Vol.45, pp. 331-334, © 1982, International Associa-tion of Milk, Food and Environmental Sanitarians.

Table 9-3. Source: Data from Weathercup et al,1988.

Figure 9-14. Source: Reprinted with permissionfrom Guinee et al, The Influence of MilkPasteurization Temperature and pH at CurdMilling on the Composition, Texture andMaturation of Reduced-fat Cheddar Cheese,International Journal of Dairy Technology, Vol.51, pp. 1-10, © 1998, Society of Dairy Technol-ogy.

Table 9-5. Source: Data from Lemay et al, 1994.Figure 9-15. Source: Reprinted with permission

from D.B. Emmons and D.C. Beckett, MilkClotting Enzymes. 1 .Proteolysis During Cheese-making in Relation to Estimated Losses of Yield,Journal of Dairy Science, Vol. 73, pp. 8-16,© 1990, American Dairy Science Association.

Figure 9-16. Source: Reprinted with permissionfrom D.M. Barbano and R.R. Rasmussen, CheeseYield Performance of Various Coagulants, CheeseYield and Factors Affecting its Control, pp. 255-259, IDF Seminar, Cork, Ireland, © 1994,International Dairy Federation.

Figure 9-17. Source: Reprinted with permissionfrom J. J. Mayes and B. J. Sutherland, CoagulumFirmness and Yield in Cheddar Cheese Manufac-ture-The Role of the Curd Firmness Instrument inDetermining Cutting Time, Australian Journal ofDairy Technology, Vol. 39, pp. 69-73, © 1984.

Figure 9-18. Source: Reprinted with permissionfrom JJ. Mayes and BJ. Sutherland, FurtherNotes on Coagulum Firmness and Yield inCheddar Cheese Manufacture, Australian Journalof Dairy Technology, Vol. 44, pp. 47-48, © 1989.

Figure 9-19. Source: Johnston et al, Effects ofSpeed and Duration of Cutting in MechanisedCheddar Cheesemaking on Curd Particle Size andYield, Journal of Dairy Research, Vol. 58, pp.345-354, © 1991. Reprinted with the permissionof Cambridge University Press.

Figure 9-20. Source: Reprinted with permissionfrom J. A. Phelan, Standardisation of Milk forCheesemaking at Factory Level, InternationalJournal of Dairy Technology, Vol. 34, pp. 152-156, © 1981, Society of Dairy Technology.

CHAPTER 10

Table 10-1. Source: Reprinted with permission fromJ. Stadhouders and L.P.M. Langeveld, The

Page 570: Cheese Science

Microflora of the Surface of Cheese. FactorsAffecting its Composition, 17th International DairyCongress, © 1966, International Dairy Federation.

Figure 10-1. Source: Reprinted with permissionfrom M. Ruegg and B. Blanc, Influence of WaterActivity on the Manufacture and Aging of Cheese,Water Activity: Influences on Food Quality, L.B.Rockland and G.F. Stewart, eds., pp. 791-811,© 1981, Academic Press, Inc.

Figure 10-2. Source: Reprinted with permissionfrom M. Ruegg and B. Blanc, Influence of WaterActivity on the Manufacture and Aging of Cheese,Water Activity: Influences on Food Quality, L.B.Rockland and G.F. Stewart, eds., pp. 791-811,© 1981, Academic Press, Inc.

Table 10-2. Source: Reprinted with permission fromM. Ruegg and B.Blanc, Influence of WaterActivity on the Manufacture and Aging of Cheese,Water Activity: Influences on Food Quality, L.B.Rockland and G.F. Stewart, eds., pp. 791-811,© 1981, Academic Press, Inc.

Figure 10-4. Source: Reprinted with permissionfrom M. Ruegg and B. Blanc, Influence of WaterActivity on the Manufacture and Aging of Cheese,Water Activity: Influences on Food Quality, L.B.Rockland and G.F. Stewart, eds., pp. 791-811,© 1981, Academic Press, Inc.

Figure 10-5. Source: Reprinted with permissionfrom F.G. Martley and R.C. Lawrence, CheddarCheese Flavour. 2. Characteristics of Single StrainStarters Associated with Good or Poor FlavourDevelopment, New Zealand Journal of DairyScience Technology, Vol. 7, pp. 38^4, © 1972,New Zealand Dairy Research Institute.

Figure 10—7. Source: Data from Nunez, 1978; delPozo et al, 1985; Poullet et al, 1991 and Cuesta etal, 1996.

Figure 10-8. Source: Data from delPozo et al, 1985;Poullet et al, 1991; Demarigny et al, 1996 andBeresford and Cogan, unpublished.

Table 10-3. Source: Data from Valdes-Stauber et al,1997 and Eliskases-Lechner and Gringinger,1995a.

Figure 10-9. Source: Reprinted with permissionfrom Samson et al, Introduction to Food-borneFungi, © 1995, Centraalbureau voor Schmmel-cultures.

Figure 10-10. Source: Reprinted with permissionfrom K.W. Turner and T.D. Thomas, LactoseFermentation in Cheddar Cheese and the Effect ofSalt, New Zealand Journal of Dairy Science

Technology, Vol. 15, pp. 265-276, © 1980, NewZealand Dairy Research Institute.

Figure 10-11. Source: Reprinted with permissionfrom K.W. Turner and T.D. Thomas, LactoseFermentation in Cheddar Cheese and the Effect ofSalt, New Zealand Journal of Dairy ScienceTechnology, Vol. 15, pp. 265-276, © 1980, NewZealand Dairy Research Institute.

Figure 10-12. Source: Reprinted with permissionfrom Accolas et al, Evolution de Ia Flore LactiqueThermophile au Courss du Pressage des Fromagea Pate Cuite, Le Lait, Vol. 58, pp. 118-132,© 1978, Editions Scientifique et MedicalesElsevier.

Figure 10-14. Source: Reprinted with permissionfrom Turner et al, Swiss-type Cheese: II. The Roleof Thermophilic Lactobacilli in Sugar Fermenta-tion, New Zealand Journal of Dairy ScienceTechnology, Vol. 18, pp. 117-124, © 1983.

Figure 10-15. Source: Reprinted with permissionfrom J. Lenoir, La Flore Microbienne du Camem-bert et Son Evolution au Cours de Ia Maturation,Le Lait, Vol. 43, pp. 262-270, © 1963, EditionsScientifiques et Medicales Elsevier.

Figure 10-16. Source: M. Nunez, Microflora ofCabrales Cheese: Changes During Maturation,Journal of Dairy Research, Vol. 45, pp. 501-508,© 1978, Cambridge University Press. Reprintedwith the permission of Cambridge UniversityPress.

Figure 10-17. Source: Reprinted with permissionfrom F. Eliskases-Lechner and W. Ginzinger, TheYeast Flora of Surface-ripened Cheese, Milchwis-senschaft, Vol. 50, pp. 458^62, © 1995 a,b.

CHAPTER 11

Table 11-1. Source: Reprinted with permission fromT.D. Thomas and V.L. Crow, Mechanism ofD (-)-Lactic Acid Formation in Cheddar Cheese,New Zealand Journal of Dairy Science Technol-ogy, Vol. 18, pp. 131-141, © 1983, New ZealandDairy Research Institute.

Figure 11-2. Source: Reprinted with permissionfrom T.D. Thomas, Acetate Production fromLactate and Citrate by Non-starter Bacteria inCheddar Cheese, New Zealand Journal of DairyScience Technology, Vol. 22, pp. 25-38, © 1987,Dairy Technology.

Figure 11-3. Source: Reprinted with permissionfrom C. Karahadian and R. C. Lindsay, Integrated

Page 571: Cheese Science

Roles of Lactate, Ammonia, and Calcium inTexture Development of Mold Surface-ripeningCheese, Journal of Dairy Science, Vol. 70, pp.909-918, © 1987, American Dairy ScienceAssociation.

Figure 11-4. Source: Reprinted with permissionfrom C. Karahadian and R. C. Lindsay, IntegratedRoles of Lactate, Ammonia, and Calcium inTexture Development of Mold Surface-ripeningCheese, Journal of Dairy Science, Vol. 70, pp.909-918, © 1987, American Dairy ScienceAssociation.

Figure 11-5. Source: Reprinted with permissionfrom LV. Zarmpoutis, Proteolysis in Blue-veinedCheese Varieties, MSc Thesis, © 1995, NationalUniversity of Ireland.

Figure 11-6. Source: Reprinted with permissionfrom S. A. Lowney, Characterization of Proteoly-sis in the Italian Smear-ripened Cheese, Taleggio,MSc Thesis, © 1997, National University ofIreland.

Figure 11-7. Source: Reprinted with permissionfrom P.F.Fox and P.H.L McSweeney, DairyChemistry and Biochemistry, p.406, © 1998,Aspen Publishers, Inc.

Figure 11-9. Source: Reprinted with permissionfrom McSweeney et al, Contribution of theIndigenour Microflora to the Maturation ofCheddar Cheese, International Dairy Journal,Vol. 3, © 1993, pp. 613-634, with permissionfrom Elsevier Science.

Table 11-2. Source: Data from Woo et al, 1984,Woo and Lindsay, 1984, and Fox and McSwee-ney, 1998.

Figure 11-10. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 470, © 1998,Aspen Publishers, Inc.

Figure 11-12. Source: Reprinted with permissionfrom D.M. Rea, Comparison of Cheddar Cheesemade with Chymosin, Rhizomucor mieheiProteinase or Chryphonectria parasiticaProteinases, MSc Thesis, © 1997, NationalUniversity of Ireland.

Figure 11-13. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, DairyChemistry and Biochemistry, p. 320, © 1998,Aspen Publishers, Inc.

Figure 11-14. Source: Reprinted with permissionfrom Kunji et al, the Proteolytic System of LacticAcid Bacteria, Antonie van Leeuwenhoek, Vol. 70,

p. 97, Table 2, © 1996, with kind permission fromKluwer Academic Publishers.

Table 11-3. Source: Reprinted with permission fromFox et al, Cheese: Physical, Biochemical andNutritional Aspects, Advances in Food NutritionResearch, Vol. 39, pp. 163-328, © 1996,Academic Press, Inc.

Figure 11-16. Source: Reprinted with permissionfrom M. McGoldrick, MSc Thesis, © 1996,University College, Cork.

Figure 11-17. Source: Reprinted with permissionfrom M. McGoldrick, MSc Thesis, © 1996,University College, Cork.

Figure 11-18. Source: Reprinted with permissionfrom M. McGoldrick, MSc Thesis, © 1996,University College, Cork.

Figure 11-19. Source: Reprinted with permissionfrom Mooney et al, Identification of the PrincipalWater-insoluble Peptides in Cheddar Cheese,International Dairy Journal, Vol. 8, © 1998, pp.813-818, with permission from Elsevier Science.

Figure 11-20. Source: Reprinted with permissionfrom Mooney et al, Identification of the PrincipalWater-insoluble Peptides in Cheddar Cheese,International Dairy Journal, Vol. 8, © 1998, pp.813-818, with permission from Elsevier Science.

Figure 11-22. Source: Reprinted with permissionfrom Mooney et al, Identification of the PrincipalWater-insoluble Peptides in Cheddar Cheese,International Dairy Journal, Vol. 8, © 1998, pp.813-818, with permission from Elsevier Science.

Figure 11-23. Source: Reprinted with permissionfrom P.F. Fox and P.H.L. McSweeney, Proteolysisin Cheese During Ripening, Food ReviewsInternational, Vol. 12, pp. 457-509, MarcelDekker, Inc. New York, © 1996.

Figure 11-24. Source: Reprinted with permissionfrom P.F. Fox and J.M. Wallace, Formation ofFlavour Compounds in Cheese, Advances inApplied Microbiology, Vol. 45, pp. 53-58,© 1997, Academic Press, Inc.

CHAPTER 12

Exhibit 12-1. Source: Reprinted with permissionfrom Nielson et al, Progress in Developing anInternational Protocol for Sensory Profiling ofHard Cheese, International Journal of DairyTechnology, Vol. 51, pp. 57-64, © 1998, Societyof Dairy Technology.

Page 572: Cheese Science

Figure 12-1. Source: Reprinted with permissionfrom Nielson et al, Progress in Developing anInternational Protocol for Sensory Profiling ofHard Cheese, International Journal of DairyTechnology, Vol. 51, pp. 57-64, © 1998,Society of Dairy Technology.

Figure 12-3. Source: DJ. Manning and C. Moore,Headspace Analysis of Hard Cheese, Journal ofDairy Research, Vol. 46, pp. 539-545, © 1979.Reprinted with the permission of CambridgeUniversity Press.

Figure 12-5. Source: Reprinted from P. Schieberle,New Developments in Methods for Analysis ofVolatile Flavor Compounds and Their Precur-sors, Characterization of Food: EmergingMethods, A.G. Gaonkar, ed., © 1995, pp. 403-431, with permission from Elsevier Science.

Figure 12-7. Source: Reprinted with permissionfrom Aston et al, Proteolysis and FlavorDevelopment in Cheddar Cheese, AustralianJournal of Dairy Technology, Vol. 38, pp. 55-65, © 1983, Dairy Industry Association ofAustralia.

Exhibit 12-2. Source: Reprinted with permissionfrom G. Urbach, Relations Between CheeseFlavour and Chemical Composition, Interna-tional Dairy Journal, Vol. 3, © 1993, pp. 389-422, with permission from Elsevier Science.

Table 12-2. Source: Reprinted with permissionfrom G. Urbach, The Flavour of Milk and DairyProducts. II. Cheese: Contribution of VolatileCompounds, International Journal of DairyTechnology, Vol. 50, pp. 79-80, © 1997,Society of Dairy Technology.

Figure 12-8. Source: Reprinted from InternationalDairy Journal, Vol. 3, J.O. Bosset and R.Gauch, Comparison of the Volatile Flavour inSix European AOC Cheeses by Using a NewDynamic Headspace GC-MS Method, pp. 359-377, © 1993, with permission from ElsevierScience.

Figure 12-11. Source: Reprinted with permissionfrom T.P. Coultate, Food: The Chemistry of ItsComponents, 2nd Edition, p. 24, © 1989, TheRoyal Society of Chemistry.

Figure 12-12. Source: Reprinted from Interna-tional Dairy Journal, Vol. 6, O'Shea et al,Objective Assessment of Cheddar CheeseQuality, pp. 1135-1147, © 1996, with permis-sion from Elsevier Science.

CHAPTER 13

Figure 13-2. Source: Reprinted with permissionfrom Guinee et al, The Influence of MilkPasteurization Temperature and pH at CurdMilling on the Composition, Texture andMaturation of Reduced-fat Cheddar Cheese,International Journal of Dairy Technology, Vol.60, pp. 1-12, © 1998, Society of Dairy Technol-ogy.

Figure 13-3. Source: Reprinted with permssion fromM. Kalab, Milk Gel Structure. VI. Cheese Textureand Microstructure, Milchwissenschaft, Vol. 32,pp. 449_457? © 1977, Milk Science International.

Figure 13—4. Source: Reprinted with permissionfrom Lowrie et al, Curd Granule and Milled CurdJunction Patterns in Cheddar Cheese Made byTraditional and Mechanized Processes, Journal ofDairy Science, Vol. 65, pp. 1122-1129, © 1982,American Dairy Science Association.

Figure 13-10. Source: Reprinted with permissionfrom P. Walstra and T. van Vliet, Rheology ofCheese, Bulletin No. 153, pp. 22-27, © 1982,International Dairy Federation.

Figure 13-11. Source: Data from Guinee, unpub-lished results.

Figure 13-13. Source: Reprinted with permissionfrom J.H. Prentice, Cheese Rheology, in Cheese:Chemistry, Physics and Microbiology, Volume 1,General Aspects, P.F. Fox, ed., © 1987.

Table 13-2. Source: Data from van Vliet, 1991 andvan Vliet et al, 1991.

Figure 13-16. Source: Reprinted with permissionfrom F. Shama and P. Sherman, Evaluation ofSome Textural Properties of Foods with theInstron Universal Testing Machine, Journal ofTexture Studies, Vol. 4, pp. 344-353, © 1973,Food and Nutrition Press, Inc.

Figure 13-17. Source: Reprinted with permissionfrom J.H. Prentice, Cheese Rheology, in Cheese:Chemistry, Physics and Microbiology, Volume 1,General Aspects, P.F. Fox, ed., © 1993.

Figure 13-18. Source: Reprinted with permissionfrom J. Culioli and P. Sherman, Evaluation ofGouda Cheese Firmness by Compression Tests,Journal of Texture Studies, Vol. 7, pp. 353-372,© 1976, Food and Nutrition Press.

Figure 13-19. Source: Reprinted with permissionfrom P. Sherman, Rheological Evaluation of theTextural Properties of Foods, Progress and

Page 573: Cheese Science

Trends in Rheology II, pp. 44-53, © 1988,Springer-Verlag New York, Inc.

Figure 13-20. Source: Data from Culioli andSherman, 1976, Vernon Carter and Sherman, 1978and Prentice et al, 1993.

Figure 13-21. Source: Reprinted with permissionfrom J. Culioli and P. Sherman, Evaluation ofGouda Cheese Firmness by Compression Tests,Journal of Texture Studies, Vol. 7, pp. 353-372,© 1976, Food and Nutrition Press, Inc.

Figure 13-22. Source: Reprinted with permissionfrom Chen et al, Texture Analysis of Cheese,Journal of Dairy Science, Vol. 62, pp. 901-907,© 1979, American Dairy Science Association.

Figure 13-23. Source: Reprinted with permissionfrom M.A. Fenelon and T.P. Guinee, Improvingthe Quality of Low-fat Cheddar Cheese, ProjectReport, DPRCNo. 4, © 1999, Dairy ProductsResearch Center, Ireland.

Figure 13-24. Source: Reprinted with permissionfrom J. Visser, Factors Affecting the Rheologicaland Fracture Properties of Hard and Semi-hardCheese, Rheological and Fracture Properties ofCheese-Bulletin No. 268, pp. 49-61, © 1991,International Dairy Federation.

Figure 13-25. Source: Reprinted with permissionfrom J. Visser, Factors Affecting the Rheologicaland Fracture Properties of Hard and Semi-hardCheese, Rheological and Fracture Properties ofCheese-Bulletin No. 268, pp. 49-61, © 1991,International Dairy Federation.

Figure 13-26. Source: Reprinted with permissionfrom J. Visser, Factors Affecting the Rheologicaland Fracture Properties of Hard and Semi-hardCheese, Rheological and Fracture Properties ofCheese-Bulletin No. 268, pp. 49-61, © 1991,International Dairy Federation.

Figure 13-27. Source: Reprinted with permissionfrom J. Visser, Factors Affecting the Rheologicaland Fracture Properties of Hard and Semi-hardCheese, Rheological and Fracture Properties ofCheese-Bulletin No. 268, pp. 49-61, © 1991,International Dairy Federation.

Figure 13-28. Source: Guinee et al, Milk ProteinStandardization by Ultrafiltration for CheddarCheese Manufacture, Journal of Dairy Research,Vol. 63, pp. 281-293, © 1996. Reprinted with thepermission of Cambridge University Press.

Exhibit 13—1. Source: Data from Szczesniak, 1963andBrennan, 1988.

Figure 13-29. Source: Reprinted with permissionfrom J.G. Brennan, Texture Perception andMeasurement, Sensory Analysis of Foods, 2ndEd., J.R. Piggott, ed., pp. 69-101, © 1988.

Figure 13-30. Source: Reprinted with permissionfrom J.G. Brennan, Texture Perception andMeasurement, Sensory Analysis of Foods, 2ndEd., J.R. Piggott, ed., pp. 69-101, © 1988.

Exhibit 13-2. Source: Reprinted with permissionfrom J.G. Brennan, Texture Perception andMeasurement, Sensory Analysis of Foods, 2ndEd., J.R. Piggott, ed., pp. 69-101, © 1988.

Table 13-3. Source: Reprinted with permission fromLee et al, Evaluation of Cheese Texture, Journalof Food Science, Vol. 43, pp. 1600-1605, © 1978,Institute of Food Technologists.

CHAPTER 14

Figure 14-2. Source: Data from Gilles andLawrence, 1973, Fox, 1975, and Pearce andGilles, 1979.

CHAPTER 16

Figure 16-1. Source: Reprinted with permissionfrom T.P. Guinee, P.D. Pudja and N.Y. Farkye,Fresh Acid-curd Cheese Varieties, in Cheese:Chemistry, Physics and Microbiology, Vol. 2Major Cheese Groups, 2nd Ed., P.F. Fox, ed., p.364, © 1993, Aspen Publishers, Inc.

Table 16-1. Source: Reprinted with permission fromT.P. Guinee, P.D. Pudja and N.Y. Farkye, FreshAcid-curd Cheese Varieties, in Cheese: Chemis-try, Physics and Microbiology, Vol. 2 MajorCheese Groups, 2nd Ed., P.F. Fox, ed., p. 367,© 1993, Aspen Publishers, Inc.

Figure 16-2. Source: Reprinted with permissionfrom T.P. Guinee, P.D. Pudja and N.Y. Farkye,Fresh Acid-curd Cheese Varieties, in Cheese:Chemistry, Physics and Microbiology, Vol. 2Major Cheese Groups, 2nd Ed., P.F. Fox, ed., p.365, © 1993, Aspen Publishers, Inc.

Figure 16-3. Source: Reprinted from InternationalDairy Journal, Vol. 40, vanHooydonk et al, pH-induced Physico-chemical Changes in CaseinMicelles in Milk and Their Effect on Ren-neting. 1 .Effect of Acidification on Physico-chemical Properties, pp. 281-296, © 1986, withpermission from Elsevier Science.

Page 574: Cheese Science

Figure 16-4. Source: D.G. Dalgleish and AJ.R.Law, pH Induced Dissociation of Bovine CaseinMicelles.I. Analysis of Liberated Caseins, Journalof Dairy Research, Vol. 55, pp. 529-538, © 1988.Reprinted with the permission of CambridgeUniversity Press.

Figure 16-5. Source: Reprinted with permissionfrom L.K. Creamer, Water Absorption byRenneted Casein Micelles, Milchwissenschaft,Vol. 40, pp. 589-59, © 1985.

Figure 16-6. Source: Data from P. McSweeney, T.P.Guinee and M.G. Wilkinson (unpublished results).

Figure 16-10. Source: Data from Auty, McSweeney,Guinee and Wilkinson (unpublished results).

Figure 16-11. Source: Data from P. McSweeney,T.P. Guinee and M.G. Wilkinson (unpublishedresults).

CHAPTER 17

Exhibit 17-1 Source: Reprinted with permissionfrom P.F. Fox, Cheese: An Overview, Cheese,Chemistry, Physics and Microbiology, 2nd Ed,Vol.1, P.F. Fox, ed., pp. 1-36, © 1993, AspenPublishers, Inc.

Table 17-1 Source: Reprinted with permission fromInternational Dairy Journal, Vol. 3, L. Berozziand G. Panari, Cheeses with Appellationd'Origine Controlee (AOC): Factors That AffectQuality, pp. 297-312, © 1993, with permissionfrom Elsevier Science.

Figure 17-6 Courtesy of APV Nordic Cheese,Denmark.

Figure 17-22 Source: Reprinted with permissionfrom F.V. Kosikowski and V.V. Misrry, Cheeseand Fermented Milk Foods, 3rd Ed., VoIs. 1 and2, © 1997, F.V. Kosikowski, LLC.

Appendix 17-A Source: Reprinted with permissionfrom F.V. Kosikowski and V.V. Mistry, Cheeseand Fermented Milk Foods, 3rd Ed., VoIs. 1 and2, © 1997, F.V. Kosikowski, LLC.

CHAPTER 18

Table 18-1 Source: Reprinted with permission fromFox et al, Cheese, Physical, Biochemical andNutritional Aspects, Advances in Food ScienceNutrition Research, Vol. 39, pp. 163-329,© 1996, Academic Press, Inc.

Exhibit 18-1 Source: Reprinted with permissionfrom Fox et al, Cheese, Physical, Biochemical and

Nutritional Aspects, Advances in Food ScienceNutrition Research, Vol. 39, pp. 163-329,© 1996, Academic Press, Inc.

Table 18-2 Source: Reprinted with permission fromFox et al, Cheese, Physical, Biochemical andNutritional Aspects, Advances in Food ScienceNutrition Research, Vol. 39, pp. 163-329,© 1996, Academic Press, Inc.

Table 18-3 Source: Reprinted with permission fromFox et al, Cheese, Physical, Biochemical andNutritional Aspects, Advances in Food ScienceNutrition Research, Vol. 39, pp. 163-329,© 1996, Academic Press, Inc.

Table 18-4 Source: Reprinted with permission fromFox et al, Cheese, Physical, Biochemical andNutritional Aspects, Advances in Food ScienceNutrition Research, Vol. 39, pp. 163-329,© 1996, Academic Press, Inc.

Table 18-5 Source: Reprinted with permission fromGuinee et al, Characteristics of Different CheesesUsed in Pizza Pie, Australian Journal of DairyTechnology, Vol. 53, p. 109, © 1998.

Table 18-6 Source: Reprinted with permission fromGuinee et al, Characteristics of Different CheesesUsed in Pizza Pie, Australian Journal of DairyTechnology, Vol. 53, p. 109, © 1998.

Figure 18-6 Source: Data from Auty and Guinee(unpublished results).

CHAPTER 19

Figure 19-2 Source: Data from Guinee et al(unpublished results).

Figure 19-3 Source: Data from Guinee et al(unpublished results).

Figure 19-4 Source: Data from Guinee et al(unpublished results).

Figure 19-5 Source: Data from Guinee et al(unpublished results).

Figure 19-6 Source: Data from Guinee et al(unpublished results).

Figure 19-7 Source: Data from Auty and Guinee(unpublished results).

Figure 19-8 Source: Reprinted with permission fromT.P. Guinee and M.O. Corcoran, Expanded Use ofCheese in Processed Meat Products, Farm andFood Research, Vol. 4, No. 1, pp. 25-28, © 1994.

Figure 19-9 Source: Reprinted with permission fromGuinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedings

Page 575: Cheese Science

of the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-10 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-11 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-12 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-13 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-14 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Figure 19-15 Source: Reprinted with permissionfrom Guinee et al, Functionality of Low MoistureMozzarella Cheese During Ripening, Proceedingsof the 5th Cheese Symposium, Cork, T.M. Cogan,ed., © 1997, Teagasc, Ireland.

Table 19-7 Source: Reprinted with permission fromGuinee et al, The Viscosity of Cheese Sauces withDifferent Starch Systems and Cheese Powders,International Journal of Dairy Technology, Vol.47, pp. 132-138, © 1994, Society of DairyTechnology.

Table 19-8 Source: Reprinted with permission fromGuinee et al, The Viscosity of Cheese Sauces withDifferent Starch Systems and Cheese Powders,International Journal of Dairy Technology, Vol.47, pp. 132-138, © 1994, Society of DairyTechnology.

CHAPTER 20

Table 20-2 Source: Reprinted with permission fromDesmasures et al, Microbiological Composition ofRaw Milk from Selected Farmers in the Camem-bert Region of Normandy, Journal of Applied

Microbiology, Vol. 83, pp. 53-58, © 1997,Blackwell Science, Ltd.

Table 20-3 Source: Reprinted from EU Guidelines.Table 20-4 Source: Reprinted with permission from

Nichols et al, The Microbiological Quality of SoftCheese, PHLS Microbiological Digest, Vol. 13,pp. 68-75, © 1996, Public Health LaboratoryServices.

Figure 20-1 Source: Reprinted with permission fromRea et al, Incidence of Pathogenic Bacteria inRaw Milk in Ireland, Journal of Applied Bacteri-ology, Vol. 73, pp. 331-336, © 1992, BlackwellScience, Ltd.

Figure 20—2 Source: Reprinted with permission fromthe Journal of Food Protection. Copyright held bythe International Association of Milk, Food andEnvironmental Sanitarians, Inc., Des Moines,Iowa, USA. CJ. Reitsma and D.R. Henning,Survival of Enterohemmorragic Escherichia collO157:H7 During the Manufacture and Curing ofCheddar Cheese, Vol. 59, pp. 460-464, © 1996;E.T. Ryser and E.H. Marth, Behavior ofListeriamonocytogenes during the Manufacture andRipening of Cheddar Cheese, Vol. 50, pp. 7-13,© 1987; and Tuckey et al, Relation ofCheesemaking Operations to Survival and Growthof Staphloccus aureus in Different Varieties ofCheese, Journal of Dairy Science, Vol. 47, pp.604-611,© 1964.

Figure 20—3 Source: Reprinted with permission fromH.P. Bachmann and U. Spahr, The Fate ofPotentially Pathogenic Bacteria in Swiss Hard andSemi-hard Cheeses Made from Raw Milk, Journalof Dairy Science, Vol. 78, pp. 476-483, © 1995,American Dairy Science Association.

Figure 20-4 Source: L. Bautista and R.G. Kroll,Survival of Some Non-starter Bacteria inNaturally Ripened and Enzyme-acceleratedCheddar Cheese, Journal of Dairy Research, Vol.55, pp. 597-602, © 1988. Reprinted with thepermission of Cambridge University press.

Figure 20-5 Source: Reprinted with permission fromthe Journal of Food Protection. Copyright held bythe International Association of Milk, Food andEnvironmental Sanitarians, Inc., Des Moines,Iowa, USA, E.T. Ryser and E.H. Marth, BehaviorofListeria monocytogenes during the Manufactureand Ripening of Cheddar Cheese, Vol. 50, pp. 7-13, ©1987.

Figure 20-6 Source: Reprinted with permission fromCJ. Reitsma and D.R. Henning, Survival of

Page 576: Cheese Science

Enterohemorragic Escherichia coli O157:H7During the Manufacture and Curing of CheddarCheese, Journal of Food Protection, Vol. 59, pp.460-464, © 1996, International Association ofMilk, Food and Environmental Sanitarians.

Figure 20-7 Source: Reprinted with permissionfrom Hargrove et al, Factors Affecting Survivalof Salmonella in Cheddar and Colby Cheese,Journal of Food Technology, Vol. 32, pp. 580-584, © 1969, International Association of Milk,Food and Environmental Sanitarians.

Figure 20-8 Source: Reprinted with permissionfrom the Journal of Food Protection. Copyrightheld by the International Association of Milk,Food and Environmental Sanitarians, Inc., DesMoines, Iowa, USA, Frank et al, Survival ofEnteropathogenic and Non-pathogenic Escheri-chia coli During the Manufacture of CamembertCheese, Vol. 40, pp. 835-842, © 1977;Rutzinski et al, Behaviour of Enterobacteraerogenes and Hafnia species during theManufacture and Ripening of CamembertCheese, Vol. 42, pp. 790-793, © 1979; and E.T.Ryser and E.H. Marth, Behavior ofListeriamonocytogenes During the Manufacture andRipening of Cheddar Cheese, Vol. 50, pp. 7-13,© 1987.

Figure 20-9 Source: Reprinted with permissionfrom Poullet et al, Microbial Study of Casar deCaceres Cheese Throughout Ripening, Journalof Dairy Research, Vol. 58, pp. 231-238,© 1991; delPozo et al, Changes in the Microf-lora of LaSerena Ewe's Milk Cheese DuringRipening, Journal of Dairy Research, Vol. 55,pp. 449^455, © 1985; Cuesta et al, Evaluationof the Microbiological and BiochemicalCharacteristics of Afuga'l Pitu Cheese DuringRipening, Journal of Dairy Science, Vol. 79, pp.1693-1698, © 1996; and Gobbetti et al,Microbiology and Biochemistry of PecorinoUmbro Cheese During Ripening, Italian Journalof Food Science, Vol. 9, pp. 111-126, © 1997.

CHAPTER 21

Exhibit 21-1 Source: Reprinted from the U.S.Department of Agriculture and the Departmentof Health and Human Services, 1995.

Table 21-1 Source: Reprinted with permissionfrom Holland et al, Milk Products and Eggs, The

Fourth Supplement to McCance and Widdowson 'sThe Composition of Foods, 4th ed., © 1989, RoyalSociety of Chemistry/Ministry of Agriculture,Fisheries and Food, Cambridge, UK. Crowncopyright material is adapted/reproduced with thepermission of the Controller of Her Majesty'sStationery Office.

Table 21-2 Source: Reprinted with permission fromHolland et al, Milk Products and Eggs, TheFourth Supplement to McCance and Widdowson'sThe Composition of Foods, 4th ed., © 1989, RoyalSociety of Chemistry/Ministry of Agriculture,Fisheries and Food, Cambridge, UK. Crowncopyright material is adapted/reproduced with thepermission of the Controller of Her Majesty'sStationery Office.

Table 21-3 Source: Reprinted with permission fromHolland et al, Milk Products and Eggs, TheFourth Supplement to McCance and Widdowson'sThe Composition of Foods, 4th ed., © 1989, RoyalSociety of Chemistry/Ministry of Agriculture,Fisheries and Food, Cambridge, UK. Crowncopyright material is adapted/reproduced with thepermission of the Controller of Her Majesty'sStationery Office.

Table 21-4 Source: Adapted with permission fromE. Renner, Nutritional Aspects of Cheese, Cheese:Chemistry, Physics and Microbiology, Vol.1, 2ndEdition, P.P. Fox, ed., pp. 345-363, © 1987.

CHAPTER 22

Table 22-1 Source: Reprinted with permission fromD.M. Mulvihill, Production, Functional Propertiesand Utilization of Milk Protein Products,Advanced Dairy Chemistry, Volume 1, Proteins,P.F. Fox, ed., pp. 369-404, © 1992.

Figure 22-1 Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 46, © 1998, AspenPublishers, Inc.

Figure 22-1 Courtesy of the International DairyFederation, 1993, Brussels, Belgium.

Figure 22-3 Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 51, © 1998, AspenPublishers, Inc.

Figure 22-4 Source: Reprinted with permission fromP.F. Fox and P.H.L. McSweeney, Dairy Chemis-try and Biochemistry, p. 52, © 1998, AspenPublishers, Inc.

Page 577: Cheese Science

Figure 22-5 Courtesy of the International DairyFederation, © 1993, Brussels, Belgium.

CHAPTER 23

Figure 23-1 Courtesy of the International DairyFederation, Standard 5OB, Milk Products,Methods of Sampling, © 1985, International DairyFederation, Brussels.

Figure 23-2 Courtesy of the International DairyFederation, Standard 5OB, Milk Products,Methods of Sampling, © 1985, International DairyFederation, Brussels.

Figure 23-3 Source: Reprinted with permission fromP.H.L. McSweeney and P.F. Fox, ChemicalMethods for the Characterization of Proteolysis inCheese During Ripening, Le Lait, Vol. 77, pp. 41—76, © 1997, Editions Scientific Elsevier.

Figure 23^ Source: Reprinted with permission fromP.H.L. McSweeney and P.F. Fox, ChemicalMethods for the Characterization of Proteolysis inCheese During Ripening, Le Lait, Vol. 77, pp. 41-76, © 1997, Editions Scientific Elsevier.

Figure 23-5 Source: Reprinted from InternationalDairy Journal, Vol. 6, O'Shea et al, ObjectiveAssessment of Cheddar Cheese Quality, pp. 1135-

1147, © 1996, with permission from ElsevierScience.

Figure 23-7 Source: Reprinted with permission fromP.H.L. McSweeney and P.F. Fox, ChemicalMethods for the Characterization of Proteolysis inCheese During Ripening, Le Lait, Vol. 77, pp. 41—76, © 1997, Editions Scientific Elsevier.

Figure 23-8 Source: Reprinted with permission fromP.H.L. McSweeney and P.F. Fox, ChemicalMethods for the Characterization of Proteolysis inCheese During Ripening, Le Lait, Vol. 77, pp. 41—76, © 1997, Editions Scientific Elsevier.

Figure 23-9 Source: Reprinted with permission fromP.H.L. McSweeney and P.F. Fox, ChemicalMethods for the Characterization of Proteolysis inCheese During Ripening, Le Lait, Vol. 77, pp. 41—76, © 1997, Editions Scientific Elsevier.

Exhibit 23-1 Courtesy of C.M. Delahunty and J.M.Murray, Organoleptic Evaluation of Cheese,Proceedings of the 5th Cheese Symposium, pp.90-97, T.M. Cogan, ed., © 1997, Teagasc,Ireland.

Figure 23-10 Courtesy of C.M. Delahunty and J.M.Murray, Organoleptic Evaluation of Cheese,Proceedings of the 5th Cheese Symposium, pp.90-97, T.M. Cogan, ed., © 1997, Teagasc,Ireland.

Page 578: Cheese Science

559 This page has been reformatted by Knovel to provide easier navigation.

Index

Index terms Links

A Abondance cheese 393

Acceleration of ripening 349 by addition of amino acids to cheese curd 360 by addition of exogenous enzymes 351

coagulant 352 lipases 354 plasmin 352 proteinases 353 354

by adjunct starters 358 by attenuated starters 356 economic incentive for 349 360 by elevated temperature 350 for enzyme-modified cheese 359 by genetically engineered starters 358 methods used for 350 351 prospects for 360 by secondary cultures 359 by selected starters 355

Acetaldehyde 54 82

Acetate 528

Acetic acid 54 210 217 242

Acetoin 527

Acetolactate (AL) 528

Achromobacter 46

Acid degree value (ADV) method 528

Acidification 11 14 determination of titratable acidity of cheese 526 direct 14 150 194 of fresh acid-curd cheese varieties 363 375

Page 579: Cheese Science

560 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Acidification (Continued) by lactic acid bacteria (See Starter cultures) rate of 14

Acidogens 12 14 194 237

Actual cheese yield 171 See also Cheese yield

Adjunct starters 59 358

ADV. See Acid degree value method

AEDA. See Aroma extract dilution analysis

Aerococcus 62

Aeromonas hydrophila 158 484 495

Aflatoxins 223 509 See also Mycotoxins

Afuega’l Pitu cheese 215 500 501

Aggregation index (AGI) 457 474

AHA. See 6-Aminohexanoic acid

AL. See Acetolactate

Alfamatic 202

Alkaline phosphatase 527

Alloiococcus 62

Amino acids 535 addition to cheese curd 360 biogenic amines produced by decarboxylation of 512 in caseins 33 34 catabolism of 274 concentrations in selected cheese varieties 277 contribution to cheese flavor 535 lactic acid bacteria requirements for 69 quantification of 284 530 532 535 taste descriptors and threshold values of 290

6-Aminohexanoic acid (AHA) 238

Aminopeptidase A 263 265

Aminopeptidase C 72 263 265

Aminopeptidase N 72 263 264

Page 580: Cheese Science

561 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Aminopeptidase P 71 72 263

Analogue cheeses 443 applications of 443 classification of 444 444 factors influencing success of 444 pizza cheese 446

aggregation index for 460 composition and functionality of 447 462 463 464 formulation for 446 functional stability during storage of 449 manufacturing protocol for 445 446 principles of manufacture for 447

production of 443

Analytical methods 523 biochemical assessment of cheese ripening 527

lipolysis 285 528 products of lactose, lactate, and citrate metabolism 527 proteolysis 256 529

compositional analysis 525 detection of interspecies adulteration of milks and cheeses 543 microbiological analysis 536 objective assessment of cheese texture 540 sampling methods 523 sensory analysis of cheese flavor and texture 540 techniques to study volatile flavor compounds 285 286 536

Animal sources of milk 5 effect on cheese yield 177

Anion-exchange chromatography on diethylaminoethyl cellulose 531

Annatto pigments 13

Antibiotics 83 84

Appearance of cheese 282

"Appellation d’Origine Contrôlée" 392 393

Appenzeller cheese aggregation index for 460 functional characteristics after cooking 464 465

Page 581: Cheese Science

562 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Appenzeller cheese (Continued) manufacture of 406 rheological properties of 324 volatile flavor compounds in 295 water activity of 156 212

Arginine metabolism 73

Armavir cheese 385

Aroma extract dilution analysis (AEDA) 287

Aroma of cheese 283 284 292

Aroma producers 55

Arthrobacter 211 218 219

Arthrobacter citrans 209

Arthrobacter citreus 219 232

Arthrobacter globiformis 218 219 232

Arthrobacter nicotianae 218 219

Artisanal cultures 56 59

Ash 20 39 526

Asiago cheese 394 classification of 391 392 composition of 428 manufacture of 395

Aspergillus flavus 223 511

Aspergillus nidulans 133

Aspergillus niger 133

Astringency in cheese 297

Attenuated starters 356 heat- or freeze-shocked cells 356 lysozyme treatment 356 mutant starters 357 other bacteria as additives 357 solvent-treated cells 357

Azeitao cheese 394

Page 582: Cheese Science

563 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

B Babcock method 526

Bacillus 48 93 486

Bacillus cereus 158

Bacillus subtilis 358

"Back-slopping" 59

Bacteriocins 92 207 classification of 93 lantibiotics 93 nisin 93 234 508 qualities of 92

Bacteriology of milk 1 10 45 484 contamination of raw milk 45

bacterial growth during storage 47 48 due to mastitis 45 Gram-positive and Gram-negative bacteria 46 from milking equipment 46 pathogens 487 487 488

(See also Pathogens and food-poisoning bacteria in cheese) psychrotrophs 46 47 sources of microorganisms 45 temperature effects on 46 47

European Union standards for 488 499 pasteurization and 47

Bacteriophage 12 83 classification of 87 control of 89 92 detection of 88 90 91 lysogenic cycle 85 lytic cycle 84 86 87 phage-resistance mechanisms 87 89 pseudolysogeny 85 source of 88

Bacteriostatic effects of emulsifying salts 440

Bactofugation 50 53

Page 583: Cheese Science

564 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Beaufort cheese 295 393 406

Beaumont cheese 156

Beda cheese 410

Beli sir-type cheeses 385

Belle des Champs cheese 156

Biochemistry of cheese ripening 17 236 assessment of 527

lipolysis 285 528 products of lactose, lactate, and citrate metabolism 527 proteolysis 256 529

catabolism of amino acids and related events 274 citrate metabolism 248 contributions of individual agents to ripening 237 glycolysis and related events 238 349

effect of lactose concentration of cheese quality 239 modification and catabolism of lactate 242

lipolysis and related events 249 349 catabolism of fatty acids 254 indigenous milk lipase 251 252 lipases and lipolysis 249 lipases from rennet 251 microbial lipases 252 pattern and levels of lipolysis in selected cheeses 253

proteolysis 67 255 349 assessment of 256 529 characterization of 268 275 276 277 proteolytic agents in cheese and their relative importance 256 258 specificity of cheese-related proteinases 257

ripening agents in cheese 236 coagulant 236 exogenous enzymes 237 milk 236 secondary microflora 237 starter culture 237

Biogenic amines 501 512

Page 584: Cheese Science

565 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Bitterness in cheese 293 bitter peptides isolated from cheese 297 298 299 factors affecting development of 295 hydrophobicity and 293

Bixin 13 14

Bleu d’Auvergne cheese 393 417

Bleu de Bresse cheese 156

Bleu de Gex-Haut Jura-Septmoncel cheese 393

Bleu des Gausses cheese 393

Blue cheeses 15 17 30 206 414 415

analysis of methyl ketones in 529 classification of 390 391 composition of 156 428 526 free fatty acids in 253 254 416 lactate concentration in 243 lactones in 255 lipolysis in 253 349 354 manufacture of 416 418 mold inoculation for 218 nutrients in 505

vitamins and minerals 507 508 proteolysis in 349 350 rheological properties of 324 ripening of 206

death of Listeria during 498 duration of 349 pH changes during 243 246

salting of 165 toxic metabolites in 511 volatile flavor compounds in 296 349

Bra cheese 394

Brachybacterium 217 218 229

Brachybacterium alimentarium 218

Brachybacterium tyrofermentans 218

Page 585: Cheese Science

566 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Brevibacterium 210 217 218 220 357 359

Brevibacterium casei 220

Brevibacterium epidermidis 220

Brevibacterium fermentans 219

Brevibacterium fuscum 219

Brevibacterium helvolum 219

Brevibacterium imperiale 219

Brevibacterium iodinum 220

Brevibacterium linens 17 218 219 220 232 243 244 419

lipolytic activity of 253 proteinase of 268 water activity for growth of 208 209

Brevibacterium oxydans 219

Brick cheese classification of 391 composition of 156 428 free fatty acid concentration in 254 manufacture of 422

Brie cheese 17 206 413 classification of 388 390 391 cyclopiazonic acid in 511 lactic acid metabolism in 245 manufacture of 415 mold inoculation for 218 nutrients in 505

vitamins and minerals 507 508 rheological properties of 324 starter cultures for 60 volatile flavor compounds in 296 water activity of 156 212

Brie de Meaux cheese 393

Brie de Melun cheese 393

Brine-salting procedure 157 160 161

Page 586: Cheese Science

567 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

British Territorial cheeses hard varieties 397 semi-hard varieties 402

Brocciu (Brocciu Corse) cheese 393

Brucella 485

Brucella abortus 484

Brucella melitensis 486

Brunost cheese 365 386 423

Bruscio cheese 423

Bryndza cheese 403

Buffering capacity of milk 43

Bulgarian White cheese 156 410 428

Burgos cheese 17

2,3-Butanediol 527

Butanoic acid 26

Butterkäse cheese 422

C Cabrales cheese 215 394

manufacture of 417 microbial growth in 223 230 233 234 water activity of 156

Cacio-ricotta cheese 423

Caciocavallo cheese 413 428

Cadaverine 512

Caerphilly cheese classification of 391 composition of 428 flavor of 284 manufacture of 402 nutrients in 505

vitamins and minerals 507 508

Page 587: Cheese Science

568 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Calcium 40 42 in cheese 506 508

determination of 526 fresh acid-curd cheeses 363 365

effects of added calcium chloride on cheese yield 196 on rennet coagulation 124 128 on syneresis 139

emulsifying salts for calcium sequestration 435 437 micellar, pH influence on solubilization of 148

Calgon 434

Calories in cheese 505

Cambazola cheese 413

Camembert cheese 4 17 206 414 Alpma process for 135 classification of 390 391 composition of 156 428 cook temperature for 140 cyclopiazonic acid in 511 free fatty acid concentration in 254 lactate concentration in 242 243 manufacture of 415 416 microbial growth in 223 230 231 mold inoculation for 218 nutrients in 505

vitamins and minerals 507 508 rheological properties of 324 ripening of 246 starter cultures for 60 temperature for ripening of 213 texture of 306 volatile flavor compounds in 296 water activity of 156 212 213

Camembert de Normandie cheese 393

Campylobacter 487

Page 588: Cheese Science

569 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Campylobacter jejuni 158 484 495

Candida 419

Candida albicans 222

Candida catenulata 224

Candida colliculosa 224

Candida famata 224

Candida intermedia 224

Candida kefyr 224

Candida lambica 224

Candida lipolytica 224

Candida mogii 224

Candida pelliculosa 224

Candida robusta 224

Candida rugosa 224

Candida saitoana 224

Candida scottii 224

Candida sphaerica 224

Candida utilis 232

Candida valida 224

Candida versatilis 224

Canestrato Pugliese cheese 394

Cantabria cheese 393

Cantal cheese 393 397 401

Capillary electrophoresis (CE) 531 533

Carbohydrates in cheese 505 506 in milk 20

(See also Lactose)

Carbon dioxide production 151 217

Cariostatic effects of cheese 509

Carnobacterium 62 69

β-Carotene 13

Carotenoids 13 27

Page 589: Cheese Science

570 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Carré de l'Est cheese 296 391 415

Casar de Cáceres cheese 215 216 217 500

Casciotta di Urbino cheese 394

Casein micelles 35 characteristics of 40 destruction of stabilizing factors of 98 in production of fresh acid-curd cheese varieties 364 367 368 rennet-altered, aggregation of 103

factors affecting 108 structure of 35 40 41

Caseins 20 31 amino acid composition of 33 34 in cheese 506 genetic polymorphs of 10 34

effect on cheese yield 185 hydrolysis of 71 214 267 hydrophobic and charged residues in 32 38 influence on cheese yield 174 influence on rheological characteristics of cheese 328 329 isoelectric point of 1 κ-casein 35

amino acid sequence of 99 hydrolysis during rennet coagulation of milk 98

nomenclature for 34 peptides produced from β-casein by plasmin 35 39 properties of 31 proteolysis by chymosin 257 seasonal changes in level of 184 susceptibility to proteolysis during ripening 15

Casomatic 202

Castellano cheese 156

Castelmagno cheese 394

Castelo Blanco cheese 394

Page 590: Cheese Science

571 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Catabolism of amino acids 274 of fatty acids 254 of lactate 242

Cathepsin D 260

CCP. See Colloidal calcium phosphate

Cell envelope-associated proteinases (CEPs) 261 358

Chabichou du Poitou cheese 393

Chaource cheese 393

CHARM-analysis 287

Chaumes cheese 156 324

Cheddar cheese 4 14 15 52 96 398

age and texture of 306 aggregation index for 460 classification of 388 390 391 composition of 156 170 177 428

463 amino acid concentrations 277 302 cheese quality and 345 free fatty acid concentration 254 lactate concentration 242 moisture content 170 residual lactose 239

cook temperature for 140 dehydration of 138 developing texture of 146 148 flavor of 284 355 functional characteristics after cooking 462 464 465 growth of pathogens during ripening of 493 496 497 high concentration factor ultrafiltration for 192 manufacture of 397 399 400

mass balance of fat during 202 204 microbial growth in 226

nonstarter lactic acid bacteria 216 217 226

Page 591: Cheese Science

572 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Cheddar cheese (Continued) starter bacteria 214 226

microscopic appearance of 308 311 molding and pressing of 150 nutrients in 505

vitamins and minerals 507 508 proteolysis in 269 349

assessment of 530 531 rheological properties of 324 325 329 330

335 ripening of 206

exogenous lipases for acceleration of 355 temperature for 213 350

salting of 155 159 162 167 size of curd particles for 140 starter cultures for 59 60 "thread mold" defect of 235 volatile flavor compounds in 293 294 296 water activity of 156 212 water-insoluble peptides isolated from 269 275 276

Cheddaring 146 148

CheddarMaster 202

Cheese as food ingredient 452 dried cheese products 475

cheese powders 476 enzyme-modified cheeses 479 grated cheeses 475

functional properties of 456 cheese composition and proteolysis 473 cheesemaking conditions 474 for heated cheese 460

comparison of different cheese types 460 468 factors influencing functionality of cheeses upon cooking 467

for raw cheese 456 factors influencing functionality 457 460

Page 592: Cheese Science

573 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Cheese as food ingredient (Continued) organoleptic characteristics 459 rheological properties 456

ripening time 468 history of 452 requirements of 452 uses of 453

Cheese base 441

Cheese powders 476 composition of 479 480 formulations of cheese slurries for 476 478 manufacture of 476 uses of 479

Cheese trier 523 524

Cheese yield 169 calculation of 171

actual cheese yield 171 moisture-adjusted cheese yield 172 moisture-adjusted cheese yield/100 kg milk adjusted for protein and fat 173

definition of 169 factors affecting 174

addition of calcium chloride 196 animal species and breed 177 cold storage of milk 186 curd-handling systems 202 curd particle size 201 202 design and operation of cheese vat 201 203 firmness at cutting 197 200 genetic polymorphism of milk proteins 185 high concentration factor ultrafiltration 192 homogenization and microfluidization 193 195 low concentration factor ultrafiltration 170 178 milk composition 169 174 pasteurization of milk and incorporation of whey proteins 188 rennet type 196 198

Page 593: Cheese Science

574 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Cheese yield (Continued) somatic cell count and mastitis 179 343 stage of lactation and season 183 thermization of milk 188 type of starter culture and growth medium 194

prediction of 173

Cheshire cheese 4 15 397 classification of 391 composition of 428 lactate concentration in 243 manufacture of 398 nutrients in 505

vitamins and minerals 507 508 rheological properties of 324 325 salting of 155 texture of 306

Chhana cheese 385

Cholesterol in cheese 505 in milk 25 27

Cholesterol oxidation products (COPs) 505

Chymax 133

Chymogen 133

Chymosin 98 257 35 See also Rennet coagulation of milk recombinant 133

Citrate in milk 40 42

Citrate metabolism 77 248 determination of products of 527

Citrinin 512

Cladosporium cladosporioides 235

Cladosporium herbarum 209 235

Clarification of whey 515

Page 594: Cheese Science

575 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Classification of cheeses 388 See also Families of cheese pasteurized processed cheese products 430

Cleaning of milking equipment 46

Clostridium 48 93 234 508

Clostridium botulinum 158 485 486

Clostridium perfringens 158

Clostridium tyrobutyricum 50 93 234 248

Coagulant 165 236 257 260 See also Rennet coagulation of milk increasing level to accelerate ripening 352

Coagulum. See Milk gel

Codex Alimentarius 430 443

Colby cheese 243 254 402

Coliforms 45 487 537

Colloidal calcium phosphate (CCP) 15 40 364 367

Colonia cheese 403

Color of cheese 13 14

Commercial importance of cheese 169

Composition of acid and sweet wheys 514 515

Composition of cheese 156 428 463 analysis of 525 effect of salt on 162 164 effect on functional characteristics 473 fresh acid-curd varieties 363 365 influence on cheese quality 345 influence on rheological characteristics 328 334 335 nutrients 505 standards of identity for specific varieties 170

Composition of cheese curd 236

Composition of milk 19 515 changes during storage 20 factors affecting 10 influence on cheese quality 343

Page 595: Cheese Science

576 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Composition of milk (Continued) influence on cheese yield 169 174 influence on syneresis 139 interspecies differences in 5 19 20 lactose 20 lipids 25 pH 11 12 41 physicochemical properties 19 42 43 proteins 31 salts 39 seasonal and lactation-related changes in 21 22 183 332 standardization of 10 169 technological interventions affecting 169 variability of 19

Compression tests 318

Comté cheese 12 49 52 214 216 217 393

composition of 428 flavor of 284 manufacture of 406 microbial growth in 226 morge on 229 volatile flavor compounds in 295

Consumption of cheese 5 7 504 fresh acid-curd varieties 363 phases of 334

Cook temperature 140

Cooked cheese, functional properties of 460 comparison of different cheese types 460 factors influencing functionality of cheeses upon cooking 467

Copper soaps assay 528

COPs. See Cholesterol oxidation products

Corynebacterium 45 46 217 218

Corynebacterium ammoniagenes 218 219

Corynebacterium betae 219

Page 596: Cheese Science

577 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Corynebacterium flavescens 220

Corynebacterium insidiosum 219

Corynebacterium variabilis 218 219 220

Coryneform bacteria 218 419

Cottage cheese 2 14 15 55 78 363 382 422

classification of 388 391 composition of 365 428 cook temperature for 206 383 cutting pH for 383 nutrients in 505

vitamins and minerals 506 507 508 packaging of 152 production of 382

"floating curd" defect 383 "major sludge formation" 383 "minor sludge formation" 383

salting of 155 starter cultures for 382 use of rennet in 363 383 water activity of 156 212

Cow's milk 5 10 13 cheese yield from 177 178 composition of 19 20 177

Cream cheese 2 363 381 422 classification of 391 composition of 363 365 428 flavor of 382 high concentration factor ultrafiltration for 192 high heat treatment of 189 nutrients in 505

vitamins and minerals 507 508 packaging of 152 production of 381 rheological properties of 324

Page 597: Cheese Science

578 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Creep and stress relaxation experiments 311 318 319 See also Rheological properties

Creep compliance-time behavior curve 314

Critical strain 314

Crohn's disease 491

Crottin de Chavignol cheese 393

Cryoglobulins 30

Cryoprotectants 94

Cryphonectria parasitica 131 133 196 198

Cryptococcus flavus 224

Curd 2 addition of amino acids to 360 composition of 236 conversion of milk to 14 lactose concentration in 238 molding and pressing of 144 150 particle size of

effect on cheese yield 201 202 effect on syneresis 140 141 142

pH of 15 plasticization of 147 preservation of 2 shattering of 118 stirring of curd-whey mixture 144 treatments of separated curd for fresh acid-curd cheeses 378

Curd firmness 111 definition of 111 dynamic tests of 113

diffuse reflectance fiber-optic probe 120 122 123 formagraph 113 hot wire probe 120 121 hydraulically oscillating diaphragm 115 low-amplitude stress or strain rheometry 116 online sensors 118

nondynamic tests of 112

Page 598: Cheese Science

579 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Curtobacterium poinsettiae 219

Cyclopiazonic acid 511

D D cultures 59

Dairy Lo 192

Danablu cheese 156 391 417

Danbo cheese 284 403

Debaryomyces 209 415 419

Debaryomyces hansenii 208 217 223 224 232 235 242

Decanoic acid 26

Defined-strain cultures 54

Definition of cheese 1

Dehydrated cheese products 475 advantages of 475 cheese powders 476 classification of 475 enzyme-modified cheeses 479 grated cheeses 475

Delactosed and delactosed-demineralized whey powder 516

Demineralized whey powder 516

Dental caries 509

Deoxynivalenone 511

Derby cheese 397 401

Diacetyl 54 78 248 527

Dietary guidelines 504 See also Nutritional aspects of cheese

Diffuse reflectance fiber-optic probe 120 122 123

Diffusion coefficient for salt in cheese 158

Diglycerols in milk 25 27

Dipeptidase 72

Dipodascus capitatus 224

Direct-to-vat starters 11

Page 599: Cheese Science

580 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Disodium phosphate (DSP) 436

Domiati cheese 15 408 adding exogenous lipases to 355 classification of 391 composition of 156 manufacture of 410 salting of 154 163

Dried cheese products 475 advantages of 475 cheese powders 476 477 478 479 classification of 475 enzyme-modified cheeses 479 grated cheeses 475

Dry-salting procedure 155 159

DSP. See Disodium phosphate

Dutch-type cheeses 14 15 17 51 classification of 391 cook temperature for 140 dehydration of 138 lactose in 238 239 manufacture of 403 ripening temperature for 350 size of curd particles for 140

E Eating cheese 335

Edam cheese 302 classification of 391 composition of 156 428 cook temperature for 140 eyes in 78 flavor of 284 free fatty acid concentration in 254 manufacture of 403 microscopic appearance of 310

Page 600: Cheese Science

581 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Edam cheese (Continued) nutrients in 505

vitamins and minerals 507 508 proteolysis in 269 starter cultures for 55 60 water activity of 156 212

Edelpilzkäse cheese 156 417

Egmont cheese 243

Elastic compliance 314

Electron microscopy 68 308 436

Electronic nose 288

ELISA. See Enzyme-linked immunosorbent assay

EMCs. See Enzyme-modified cheeses

Emmental cheese 12 15 49 96 121 217 406

aggregation index for 460 amino acid concentrations in 277 classification of 388 391 composition of 156 428 463 cook temperature for 140 flavor of 284 functional characteristics after cooking 462 464 465 growth of pathogens during ripening of 493 495 lactose in 238 239 manufacture of 404 407 microbial growth in 226 microbial quality of raw milk used for 213 nutrients in 505

vitamins and minerals 507 508 proteolysis in 269 ripening temperature for 213 350 salt content of 154 starter cultures for 60 texture of 306 volatile flavor compounds in 296

Page 601: Cheese Science

582 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Emmental cheese (Continued) water activity of 156 212

decrease during ripening 208 210 211

Emulsification 435 440

Emulsifying salts 29 434 properties of 436

ability to stimulate emulsification 435 440 bacteriostatic effects 440 calcium sequestration 435 437 displacement and buffering of pH 435 437 flavor effects 441 hydration and dispersion of casein 435 440 hydrolysis 440

Emulsion definition of 27 milk fat as 29

Enterobacter 46 498

Enterobacter aerogenes 497

Enterococcus 1 46 48 49 500

bacteriocin production by 93 differentiation from Lactococcus 64 distinguishing characteristics of 67 growth in artisanal cheeses 216 500 methods for enumeration in cheese 537 538 539 in raw milk 487 taxonomy of 62 vancomycin-resistant 500

Enterococcus avium 63

Enterococcus casseliflavus 63

Enterococcus cecorum 64

Enterococcus columbae 64

Enterococcus dispar 64

Enterococcus durans 63

Page 602: Cheese Science

583 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Enterococcus faecalis 56 59 63 69 ATCC 8043 69 growth in cheese during ripening 493 496

Enterococcus faecium 56 59

Enterococcus gallinarum 63

Enterococcus hirae 63

Enterococcus malodoratus 63 235

Enterococcus mundtii 63

Enterococcus pseudoavium 63

Enterococcus raffinosus 63

Enterococcus saccharolyticus 64

Enterococcus sulfureus 64

Enterotoxins 45

Enzyme-linked immunosorbent assay (ELISA) 543

Enzyme-modified cheeses (EMCs) 359 452 479

Enzymes effect of salt on 165 exogenous, added to accelerate ripening 351 indigenous, of milk 259 341 microencapsulated 353 produced by surface microflora 217 proteolytic 2 101 256

(See also Rennet coagulation of milk) from starter 261

Epoisse de Bourgogne cheese 393

Epoisses cheese 296

Equilibrium relative humidity (ERH) 153

Escherichia coli 45 46 133 490 492

food-poisoning outbreaks involving 484 485 486 490 growth in cheese during ripening 493 495 O157:H7 484 486 491

growth during cheese manufacture 493 494 growth during ripening 494 497

Page 603: Cheese Science

584 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Escherichia coli (Continued) infective dose of 487 in raw milk 499 in raw milk 488 489 499 water activity for growth of 158 209

Evolutionary chronometers 66

Extra hard cheeses 388 391 392

Eye formation in cheese 78 206 278 390 391 403

F Factories for cheese production

history of 5 production of starters in 94

Families of cheese 388 acid-coagulated cheeses 422 classification of 388 heat/acid-coagulated cheeses 422 produced by concentration and crystallization 423 rennet-coagulated cheeses 392

internal bacterially ripened varieties 390 392 mold-ripened varieties 413 surface smear-ripened varieties 418

superfamilies 388

Fat in cheese 504 concentrations of free fatty acids in selected cheeses 253 contribution of fatty acids to cheese aroma and taste 249 285 292 determination of 526 fatiprotein ratio 10 169 low-fat cheeses 31 350 505

Fatty acids catabolism of 254 in milk 25 28

(See also Lipids in milk) quantification of 285 529

Page 604: Cheese Science

585 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Fehling's titration method 527

Fermentation of lactose 1 14 24 73 78 79

Fermented dairy products historical production of 1 whey products 522

Festuclavine 511

Feta cheese 5 14 16 408 adding exogenous lipases to 355 classification of 391 composition of 156 428 high concentration factor ultrafiltration for 192 lactate concentration in 243 manufacture of 408 409 nutrients in 505

vitamins and minerals 507 508 rheological properties of 324

Pick's laws of diffusion 158

Filled cheeses 450

Fiore Sardo cheese 394

Flavobacterium 46

Flavor of cheese 2 7 282 analysis of 284

nonvolatile compounds 256 284 sensory analysis 540 volatile compounds 285 536

"component balance theory" of 282 contribution of aqueous phase to 288 contribution of volatile compounds to 292 294 295 296 descriptive vocabulary for 283 284 541 effect of exogenous lipases on 355 effect of texture on perception of 283 enzyme-modified cheeses as flavor ingredients 479 flavor dilution factor 287 289 formation of flavor compounds 297

Page 605: Cheese Science

586 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Flavor of cheese (Continued) intervarietal and intravarietal comparison of

cheese ripening 300 cheese volatiles 301 gel electrophoresis 301 high performance liquid chromatography 301 302

literature reviews on 282 off-flavors 282 293

astringency 297 bitterness 293 fruitiness 297 "unclean" 297

pasteurization effects on 49 role of fatty acids in 249 285 292 role of peptidases in 73 salt and 155 167 292

Flavour Age 355

"Floating curd" defect in Cottage cheese 383

Flotemysost cheese 365 386 423

Fontal cheese 156 212

Fontina cheese 394 aggregation index for 460 flavor of 284 volatile flavor compounds in 295 water activity of 156

Food poisoning. See Pathogens and food-poisoning bacteria in cheese

Food preservation methods historical 1 3 preservatives in cheese 508 salting 153

Force-compression tests 320 323 to evaluate cheese texture 336 338 339 factors influencing rheological characteristics measured using 323

Page 606: Cheese Science

587 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Force-compression tests (Continued) cheese structure, composition, and maturity 328 test conditions 323

Formagraph 113

Formai de Mut cheese 394

Fourme de'Ambert ou Montbrison cheese 393

Fracture stress 322

Freeze-shocked lactic acid bacteria 356

Fresh acid-curd cheese varieties 2 15 206 363 classification of 388 391 422 composition of 363 365 consumption of 7 Cottage cheese 382 Cream cheese and related varieties 381 difference from rennet-curd cheeses 363 effect of gel structure on quality 369

rheological properties 373 sensory attributes 373 syneresis 371

factors that influence structure of acid gels and quality of 374 added stabilizers 376 heat treatment of milk 375 376 377 378 incubation temperature and rate of acidification 375 level of gel-forming protein 374 packaging and retailing 376 pH of gel 375 rennet addition 376

manufacture of 363 366 422 prerequisites for gel formation 368 370 371 principles of acid milk gel formation 364 Quarg and related varieties 379 Queso Blanco 385 Ricotta and Ricottone cheeses 384 treatments of separated curd 378 water activity of 156

Page 607: Cheese Science

588 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Fresh acid-curd cheese varieties (Continued) whey cheeses 386 world production of 363

Friesian Clove cheese 397 402

Fromage frais 2 55 78 363 381

composition of 365 high heat treatment of 189 nutrients in 505

vitamins and minerals 507 508 starter cultures for 60 use of rennet in 363

Fruitiness in cheese 297

Functional properties of heated cheese 460

comparison of different cheese types 460 factors influencing functionality of cheeses upon cooking 467

of raw cheese 456 factors influencing functionality 457 460 organoleptic characteristics 459 rheological properties 456

Fungi 217 222 See also Molds; Yeasts classification of 222 mycotoxins 223 509 as spoilage organisms 235 water activity for growth of 207 209

G Galacto-oligosaccharides 517 521

Galactomyces geotrichum 224

Gamalost cheese 156

Gas chromatography (GC) 284 292 528

Gas chromatography/olfactometry (GCO) 285

Gas formation 232

Page 608: Cheese Science

589 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

GC. See Gas chromatography

GCO. See Gas chromatography/olfactometry

GDL. See Gluconic acid-δ-lactone

Genetic polymorphs of milk proteins 10 34 effect on cheese yield 185

Genetically engineered starters 358

Geotrichum candidum 217 223 224 225 230 232 243 253 415 419 497

salt sensitivity of 225 water activity for growth of 208 209

Geotrichum capitatum 224

Gerber method 526

German loaf cheese 327

Gjetost cheese 16 254 423 522

Gloucester cheese 4 325 397 401

Gluconic acid-δ-lactone (GDL) 12 14 131 194 237 252 364

Glucose-galactose syrups 517

Glutamic-pyruvic transaminase (GPT) 527

γ-Glutamyl peptides 289

Glutathione (GSH) 344

Glycolysis and related events 238 349 effect of lactose concentration of cheese quality 239 modification and catabolism of lactate 242

(Glyco)macropeptide (GMP) 519

Goat's milk 5 13 cheese yield from 177 composition of 19 20 177 detection of interspecies adulteration of 543

Gorgonzola cheese 4 394 classification of 390 391 manufacture of 417 water activity of 156 212

Page 609: Cheese Science

590 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Gouda cheese 4 15 302 404 amino acid concentrations in 277 classification of 388 391 composition of 156 428 cook temperature for 140 Enterococcus malodoratus in 64 eyes in 78 flavor of 284 lactate concentration in 243 lactose in 238 manufacture of 403 405

"late gas blowing" defect 403 microscopic appearance of 310 nutrients in 505

vitamins and minerals 507 508 proteolysis in 269 Theological properties of 324 326 329 331 starter cultures for 55 60 water activity of 156 212

GPT. See Glutamic-pyruvic transaminase

Grana cheese 4 156 395 428

Grana Padano cheese 12 392 394 classification of 391 manufacture of 394

Grated cheeses, dried 475

Graviera cheese 391 397 401

Gruyère cheese 12 49 aggregation index for 460 classification of 390 391 composition of 156 428 463 free fatty acid concentration in 254 functional characteristics after cooking 462 manufacture of 406 419 microbial growth in 228 nutrients in 505

Page 610: Cheese Science

591 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Gruyère cheese (Continued) vitamins and minerals 507 508

rheological properties of 324 water activity of 156 212

decrease during ripening 208 210 211

GSH. See Glutathione

Gudbrandsdalsost cheese 156 365 386 423

H HACCP. See Hazard analysis critical control point system

Hafnia 497

Hajhia alvei 497

Halloumi cheese 410

Hansenula 419

Hard cheeses 388 391 397 growth of pathogens during ripening of 493

Harp for cutting milk gel 140 141

Harzer cheese 219

Havarti cheese classification of 390 391 composition of 428 manufacture of 422 water activity of 156

Hazard analysis critical control point (HACCP) system 341 499

Heat/acid-coagulated cheeses 422

Heat-resistant bacteria 48

Heat-shocked lactic acid bacteria 356

Heat treatment of milk 12 47 237 alternatives to 12 49

bactofugation 50 53 hydrogen peroxide 49 lactoperoxidase-hydrogen peroxide-thiocyanate 50 microfiltration 51

effect on cheese yield 188 191

Page 611: Cheese Science

592 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Heat treatment of milk (Continued) effect on milk composition 170 effect on quality of fresh acid-curd cheese varieties 375 376 377 378 effect on rennet coagulation 102 124 125 127

128 effect on syneresis 144

Heated cheese, functional properties of 460 comparison of different cheese types 460 factors influencing functionality of cheeses upon cooking 467

Helveticin J 93

Hexanoic acid 26

High concentration factor ultrafiltration 192 425

High-moisture cheeses 138 packaging of 152

High performance liquid chromatography (HPLC) 284 301 302 527 531 534

High-salt cheeses 391 408

Histamine 512 513

History of cheese 1 452

Hollanda cheese 403

Homogenization of milk 30 applications in cheese manufacture 193 effect on cheese yield 193 effect on quality of fresh acid-curd cheese varieties 374 effect on rennet coagulation 124 126 effect on syneresis 145

Hot wire probe 120 121

HPLC. See High performance liquid chromatography

Hrudka cheese 403

Hydraulically oscillating diaphragm 115

Hydrochloric acid 14

Hydrocolloids 442

Hydrogen peroxide 49 83

Hydrolytic rancidity 26

Page 612: Cheese Science

593 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

I IDF. See International Dairy Federation

Idiazábal cheese 393 397 manufacture of 401 water activity of 156

Imitation cheese products. See Substitute/imitation cheese products

Immobilized rennets 135

Immunological methods for detection of interspecies adulteration of milk 543

Infrared transmittance spectrophotometry 527

Internal bacterially ripened cheeses 390 391 392 cheeses with eyes 403 extra hard varieties 392 hard varieties 397 high-salt varieties 408 Pasta filata varieties 410 semi-hard varieties 402

International Dairy Federation (IDF) 388 430 443 523 525

Interspecies adulteration of milks and cheeses 543

Interspecies differences in milk composition 5 19 20

Iron in cheese 508

Isofumigaclavine A and B 511

J Jarlsberg cheese 5 408

composition of 463 flavor of 284 functional characteristics after cooking 462 proteolysis in 269

Joha 434 436

Johne's disease 491

Page 613: Cheese Science

594 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

K Kasal 434 437

Käseverordnung 430

Käsezubereitung 430

Kashkaval cheese 146 aggregation index for 460 classification of 391 composition of 463 functional characteristics after cooking 462 464 465 manufacture of 412

Kefalotiri cheese 397 401

Kilocalories in cheese 505

Kinetics of syneresis 145

Kjeldahl method 526

Kluyveromyces 223 415 419

Kluyveromyces lactis 223 224 232

Kluyveromyces marxianus 217 224 var. lactis 133

Knives for cutting milk gel 140 141

Koestler's chloride-lactose test for abnormal milk 21

Kopanisti cheese 59

L L cultures 59

La Serena cheese 54 214 215 216 217 397

growth of enterococci in 500

LAB. See Lactic acid bacteria

Labeneh cheese 381

Labneh cheese 365 381

β-Lactalbumin 22 517

Lactate dehydrogenase (LDH) 527

Page 614: Cheese Science

595 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Lactation effect on milk composition 21 22 183 milk pH during 43 stage of

effect on cheese yield 183 effect on Theological characteristics of cheese 332

Lactenins 83

Lactic acid bacteria (LAB) 1 14 45 bacteriocins produced by 92 cultures of 14 17 54

(See also Starter cultures) differentiation in starters 66 67 68 distinguishing characteristics of 56 growth in cheese during ripening 213

nonstarter lactic acid bacteria 215 217 starter cultures 213

heat- or freeze-shocked 356 inhibition by salt 163 inhibition of production of 83 84 lysis of 214 lysozyme-treated 356 pH for growth of 210 phylogeny of 66 70 proteinases and peptidases of 71 261 264 in raw milk 46 54 solvent-treated 357 taxonomy of 62 temperature effects on growth of 60 water activity for growth of 207

Lactic acid (lactate) 1 14 39 54 210

conversion of lactose to 238 D and L isomers of 76 242 243 527 determination of products of metabolism of 527 measurement of generation times for 96

Page 615: Cheese Science

596 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Lactic acid (lactate) (Continued) modification and catabolism of 242

Lactitol 517 520

Lactobacillus 1 17 46 62 344 355

bacteriocin production by 93 distinguishing characteristics of 67 lipolytic activity of 252 methods for enumeration in cheese 536 537 538 proteolytic system of 261 salt sensitivity of 164 taxonomy of 65

Lactobacillus acidophilus 264

Lactobacillus brevis 65 215

Lactobacillus buchneri 501

Lactobacillus casei 59 65 69 215 226 358

subsp. casei 56 264

Lactobacillus confusus 69

Lactobacillus curvatus 56 65 215

Lactobacillus delbrueckii ATCC 7830 69 lactose metabolism in 76 77 subsp. bulgaricus 54 56 58 59

65 antibiotic inhibition of growth of 84 electron microscopic appearance of 68 peptidases of 264

subsp. lactis 54 56 58 59 65

antibiotic inhibition of growth of 84

Lactobacillus fermentum 56 65 69 215

Lactobacillus halotolerans 69

Page 616: Cheese Science

597 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Lactobacillus helveticus 54 56 58 59 65

amino acid requirements of 69 bacteriocin production by 93 94 electron microscopic appearance of 68 growth in cheese during ripening 214 215 228 lactose metabolism in 77 peptidases of 264 temperature effects on growth of 61 water activity for growth of 208

Lactobacillus paracasei 59 65 215 226 358

subsp. paracasei 56 subsp. tolerans 56

Lactobacillus plantarum 56 59 65 215 253

Lactobacillus rhamnosus 56

Lactobacillus sake 93

Lactobacillus viridescens 69

Lactobionic acid 517 521

Lactocin 27 93 94

Lactocin 481 93

Lactocin S 93

Lactococcus 1 17 46 54 55 58 343 344 355

amino acid requirements of 69 bacteriophage for 83 85 distinguishing characteristics of 67 growth in cheese during ripening 214 lactose metabolism in 77 lipolytic activity of 252 methods for enumeration in cheese 536 538 proteolytic system of 69 71 261 salt sensitivity of 163

Page 617: Cheese Science

598 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Lactococcus (Continued) taxonomy of 62 temperature effects on growth of 61 61

Lactococcus garvieae 62

Lactococcus lactis 62 344 bacteriophage of 85 electron microscopic appearance of 68 lysis of 214 salt sensitivity of 163 226 subsp. cremoris 55 56 61 73

383 antibiotic inhibition of growth of 84 peptidases of 71 264

subsp. hordniae 62 63 subsp. lactis 55 56 61 73

383 antibiotic inhibition of growth of 84 bacteriocins production by 93 peptidases of 264

UC317 358 water activity for growth of 208

Lactococcus piscium 62

Lactococcus plantarum 62

Lactococcus raffinolactis 62 63

Lactoferrin 519

β-Lactoglobulin 517

Lactones 254

Lactoperoxidase (LPO) 50 83 519

Lactose 20 516 α and β anomers of 21 23 24 biosynthesis of 22 24 determination of products of metabolism of 527 during lactation 21 22 effect of lactose concentration of cheese quality 239 equilibrium in solution 22

Page 618: Cheese Science

599 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Lactose (Continued) fermentation of 1 14 24 73

78 79 238 functions of 21 interspecies differences in concentration of 20 21 market for 516 properties of 516 relationship to salt concentration 21 structure of 21 23 73 useful derivatives of 516 518

Lactulose 516 518 519

Laguiole cheese 393

Lancashire cheese 402

Lancefield antigen 64

Langres cheese 296 393

Lantibiotics 93

"Late gas blowing" defect in Gouda cheese 403

LDH. See Lactate dehydrogenase

Leerdammer cheese 271

Leicester cheese 325 397 398 428

Leiden cheese 397 402

Leloir pathway 76

Leuconostoc 1 17 54 58 59 62

amino acid requirements of 69 bacteriocin production by 93 differentiation from Lactococcus 64 distinguishing characteristics of 67 growth in artisanal cheeses 216 lactose metabolism in 76 77 methods for enumeration in cheese 537 538 539 taxonomy of 64 temperature effects on growth of 61

Leuconostoc amelibiosum 65

Leuconostoc argentinum 65

Page 619: Cheese Science

600 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Leuconostoc carnosum 65

Leuconostoc citreum 65

Leuconostoc fallax 65

Leuconostoc gelidum 65

Leuconostoc lactis 55 56 65 77

Leuconostoc mesenteroides bacteriocin production by 93 subsp. cremoris 55 56 65 77

383 subsp. dextranicum 65 subsp. mesenteroides 65

Leuconostoc paramesenteroides 69

Leuconostoc pseudomesenteroides 65

Leucosporidium scottii 224

Liebana cheese 394

Lightvan cheese 410

Limburger cheese 17 420 bacteria in 218 219 classification of 388 391 composition of 156 428 free fatty acid concentration in 254 manufacture of 420 starter cultures for 60 volatile flavor compounds in 296 water activity of 156 212 yeasts in 224

Lipases 249 definition of 249 exogenous, added to accelerate ripening 354 indigenous milk lipase 251 252 lipoprotein 250 microbial 252 psychrotroph 253 from rennet 251 sources of 249

Page 620: Cheese Science

601 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Limburger cheese (Continued) specificity exhibited by 249

Lipids in milk 25 fatty acids 25 28 influence on cheese yield 174 influence on rennet coagulation 123 influence on rheological characteristics of cheese 328 influence on syneresis 139 interspecies differences in concentration of 20 25 26 27 during lactation 22 milk fat as emulsion 27

cryoglobulins 30 homogenization 30 milk fat globule membrane 29 rate of creaming 29

Lipolysis and related events 249 349 assessment of lipolysis 285 528 catabolism of fatty acids 254 255 indigenous milk lipase 251 252 lipases and lipolysis 249 lipases from rennet 251 microbial lipases 252 pattern and levels of lipolysis in selected cheeses 253

Lipoprotein lipase (LPL) 251

Listeria 93 487 489

Listeria monocytogenes 484 485 486 489 bacteriocin inhibition of 92 control of growth of 499 growth during cheese manufacture 491 493 494 growth in cheese during ripening 494 498 in raw milk 487 488 water activity for growth of 158

Listeriosis 489

Livarot cheese 296 393 422

Low-amplitude stress or strain rheometry 116

Page 621: Cheese Science

602 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Low concentration factor ultrafiltration 170 178 425

Low-fat cheeses 31 350 505

Low-moisture cheeses 138

LPL. See Lipoprotein lipase

LPO. See Lactoperoxidase

Lymphocytes 179

Lysogenic cycle 85

Lysozyme-treated lactic acid bacteria 356

Lytic cycle 84

M Maasdamer cheese 5

classification of 391 manufacture of 408 proteolysis in 269

Macrophages 179

Magnesium in cheese 508

Mahón cheese 393 classification of 391 manufacture of 403 volatile flavor compounds in 295

Maillard reaction 423 522

"Major sludge formation" in Cottage cheese 383

Majorejo cheese 156 403

Mammary gland epithelial cells 179

Manchego cheese 5 54 393 397 composition of 428 manufacture of 401 water activity of 156

Manouri cheese 423

Manufacture of cheese 10 aseptic conditions for 237 cheese color 13 14 cheese powders 476

Page 622: Cheese Science

603 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Manufacture of cheese (Continued) cheese yield 169 concentration and crystallization 423 conversion of milk to cheese curd 14

acidification 14 coagulation 15 98 postcoagulation operations 16 138 salting 16 153 ultrafiltration 16

enzyme-modified cheese 480 fresh acid-curd varieties 363 366 422 general protocol for 11 growth of pathogens during 491 heat/acid-coagulated cheeses 422 heat treatment of milk 12 47

alternatives to 12 49 history of 1 milk selection for 10 processed cheese products 17 429 rennet-coagulated cheeses 392

internal bacterially ripened varieties 390 392 mold-ripened varieties 413 surface smear-ripened varieties 418

ripening 17 acceleration of 349 biochemistry of 236 microbiology of 206

standardization of milk composition 10 169 ultrafiltration technology for 16 425 whey and whey products 17

Maribo cheese 403

Maroilles (Marolles) cheese 4 296 393

Marzyme GM 135

Mascarpone cheese 382

Mass spectrometry (MS) 285 293

Page 623: Cheese Science

604 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Mastitis antibiotic treatment for 83 cheese yield and 179 contamination of raw milk due to 45 microbiology of 45 milk pH during 43

Maxiren 133

MCA. See Milk-clotting activity

Mechanical properties of cheese 337

Media for enumeration of starter bacteria 536

Medium concentration factor ultrafiltration 425

Mesentericin 93

Metabolism biochemical assessment of products of lactose, lactate, and citrate

metabolism 527 of lactate 242 of starter cultures 69

acetaldehyde 82 arginine 73 73 citrate 77 248 lactose 73 78 79 238 proteolysis 69 255

Methyl ketones 254 255 529

MFGM. See Milk fat globule membrane

Michaelis-Menten kinetics 105

Microbacterium 46 48 217 218 220

Microbacterium imperiale 218 219

Microbacterium lacticum 220

Microbial growth during ripening 206 factors affecting 207 microbial spoilage 232 nitrate and 211 nonstarter lactic acid bacteria 215 217 other microorganisms 217

Page 624: Cheese Science

605 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Microbial growth during ripening (Continued) Arthrobacter 219 Brachybacterium 219 Brevibacterium 220 Corynebacterium 220 functions of surface microflora 217 Microbacterium 220 Micrococcus 221 Pediococcus 221 Propionibacterium 221 Rhodococcus 220 sources of contamination 218 Staphylococcus 222 yeasts and molds 222

oxidation-reduction potential and 209 pH and organic acids and 210 salt and 154 158 163 208

212 213 in specific varieties 226

Cabrales 230 233 Camembert 230 231 Cheddar 226 227 Emmental and Comté 226 Tilsit 232 234

starter bacteria 213 temperature and 213 water activity and 158 207 209 210

211 212

Microbial lipases 252

Microbiological quality of cheese 232 488 See also Pathogens and food-poisoning bacteria in cheese analysis of 536

Micrococcus 17 46 48 210 217 218 221 419

lipolytic activity of 253

Page 625: Cheese Science

606 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Micrococcus (Continued) water activity for growth of 208

Micrococcus lactis 209

Micrococcus luteus 221

Micrococcus saprophyticus 209

Microfiltration 51

Microfluidization 194 195

Microstructure of cheese 306

Milk acidification of 11 14 amount required to produce 1 kg of cheese 96 bacteriology of 1 10 45 341

487 488 bactofugation of 50 composition of 5 10 19 contribution to cheese ripening 236 conversion to cheese curd 14 detection of interspecies adulteration of 543 heat treatment of 12 homogenization of 30 hydrogen peroxide treatment of 49 indigenous enzymes of 259 341 influence on cheese quality 341 micro filtration of 51 nutritive value of 1 19 pasteurization of 12 47 pH of 11 12 41 physicochemical properties of 19 42 43 preacidification of 12 prematuration of 12 53 raw 12 rennet coagulation of 1 15 98 selection of 10 thermization of 49

Milk-clotting activity (MCA) 101 131 133

Page 626: Cheese Science

607 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Milk fat globule membrane (MFGM) 29 251

Milk gel 15 acid milk gel formation 364

effect of gel structure on quality of 369 prerequisites for 368 370 371

effect of firmness at cutting on cheese yield 197 200 postcoagulation treatment of 16 138 production of 1 rennet coagulation and assembly of 103 109 stability of 2 16 138 tools for cutting of 141

Milk proteinases 166 259

Milk serum 514

Milk stone 46

Milking equipment, cleaning of 46

Mimolette cheese 156

Minerals in cheese 506

"Minor sludge formation" in Cottage cheese 383

Mixed-strain cultures 54 58

Mizathra cheese 423 428

Moisture-adjusted cheese yield 172 See also Cheese yield

Moisture-adjusted cheese yield/100 kg milk adjusted for protein and fat 173 See also Cheese yield

Moisture content of cheese 138 170 505 See also Water activity determination of 526 influence on rheological characteristics 331 332 salting and 162 163 164

Mold-ripened cheeses 391 413 blue-veined varieties 415 production of toxic metabolites in 511 surface mold-ripened varieties 413

Molding of cheese curd 150

Page 627: Cheese Science

608 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Molds 206 217 222 inoculation with 218 methods for enumeration in cheese 537 salt sensitivity of 165 225 as spoilage organisms 235 water activity for growth of 207 209

Monoacylglycerols in milk 25 27

Mont d'Or/Vacherin du Haut Doubs cheese 393

Montasio cheese 394 395

Monterey (Monterey Jack) cheese 254 391 402

Mozzarella cheese 5 14 131 aggregation index for 460 classification of 390 391 composition of 428 463

compared with analogue pizza cheese 448 developing texture of 147 directly acidified 150 free fatty acid concentration in 254 functional characteristics after cooking 462 464 465 manufacture of 410 412 microscopic appearance of 310 Mozzarella di bufala 410 411 nutrients in 505

vitamins and minerals 507 508 proteolysis in 349 rheological properties of 324 ripening of 206 shreddability of 457 459 starter cultures for 58 60 texture of 282 volatile flavor compounds in 296 water activity of 156

MS. See Mass spectrometry

Mucor mucedo 209

Page 628: Cheese Science

609 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Münster cheese 218 classification of 391 composition of 156 428 manufacture of 422 starter cultures for 60 volatile flavor compounds in 296 water activity of 156 212

Münster Géromé cheese 393

Murazzano cheese 394

Mutant starters 357

Mycella cheese 417

Mycobacterium bovis 484

Mycobacterium paratuberculosis 491

Mycobacterium tuberculosis 47

Mycophenolic acid 511

Mycotoxins 223 509 chemical structure of 510 definition of 509 direct contamination with 511 indirect contamination with 509 production of toxic metabolites in mold-ripened cheese 511

Mysost cheese 16 386 390 391 423 522

N Natural milk cultures 56 59

Near infrared reflectance spectroscopy 526

Neufchatel cheese 365 382 393 415

Niesost cheese 423

Nisin 93 234 508

Nitrate 211 508

Nitrite 508

Nitrogen, water-soluble (WSN) 530 535

Nonhygroscopic whey powder 516

Page 629: Cheese Science

610 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Nonstarter lactic acid bacteria (NSLAB) 358 effect on cheese quality 344 growth in cheese during ripening 215 217 methods for enumeration in cheese 537 proteolytic system of 263 268 ripening temperature and 350

Norbixin 13 14

Normanna cheese 156

Norvegia cheese 156

Norzola cheese 156

NSLAB. See Nonstarter lactic acid bacteria

Nutritional aspects of cheese 504 additives 508 biogenic amines 512 carbohydrates 506 cheese consumption 5 7 dental caries and 509 dietary guidelines 504 fat and cholesterol 504 minerals 506 mycotoxins 509 protein 506 salt 167 vitamins 506 507

O O cultures 59

Ochratoxin A 509

Online sensors 118 119

Organic acids to inhibit microbial growth 210

Organoleptic characteristics of raw cheese 459

Ossau-Iraty-Brebis-Pyrénées cheese 393

Oxaloacetate 527

Oxidation-reduction potential (Eh) 209

Oxidative rancidity 26

Page 630: Cheese Science

611 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

P Packaging 151

of fresh acid-curd cheeses 376

Paecilomyces variotti 235

Paneer cheese 385

Paracasein matrix 308 309

Paratuberculosis in cattle 491

Parmesan cheese 4 classification of 388 391 392 composition of 156 428 cook temperature for 140 free fatty acid concentration in 253 254 functional characteristics after cooking 464 465 nutrients in 505

vitamins and minerals 507 508 proteolysis in 271 349 rheological properties of 324 ripening of 206 texture of 306 volatile flavor compounds in 296 water activity of 156 212

Parmigiano-Reggiano cheese 12 49 96 394 amino acid concentrations in 277 flavor of 284 manufacture of 392 396 volatile flavor compounds in 295

Partially denatured whey protein concentrate (PDWPC) 192

Pasta filata cheeses 390 391 410

Pasteurization of milk 12 47 alternatives to 12 49

bactofugation 50 53 hydrogen peroxide 49 lactoperoxidase-hydrogenperoxide-thiocyanate 50 microfiltration 51

effect on cheese yield 188 191

Page 631: Cheese Science

612 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Pasteurization of milk (Continued) effect on milk composition 170 enzymes inactivated by 49 legislation requiring 49 temperature effects on rennet coagulation 102 124 125 127

128 189 thermoduric bacteria and 48

Pasteurized processed cheese products (PCPs) 17 429 classification of 430 compositional specifications for 433 electron microscopy studies of 436 factors affecting consistency of 441

blend ingredients 441 composition 442 processing conditions 442

ingredients in 429 431 432 manufacturing protocol for 431 433 nutrients in 505

vitamins and minerals 507 508 packaging of 152 principles of manufacture for 432

addition of emulsifying salts 434 calcium sequestration 435 dispersion and water binding of paracasein 435 displacement and stabilization of pH 435 emulsification 435

properties of emulsifying salts 436 rework cheese 441 salt content of 168 structure formation upon cooling 435 water activity of 156 212 world production of 429

Pathogens and food-poisoning bacteria in cheese 484 biogenic amines 501 512 control of growth of 499

Page 632: Cheese Science

613 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Pathogens and food-poisoning bacteria in cheese (Continued) enterococci 500 Escherichia coli 490 food-poisoning outbreaks and causative organisms 484 485 growth during cheese manufacture 491 494 growth during ripening 493

hard and semi-hard cheeses 493 soft cheeses 496

Listeria monocytogenes 489 microbiological quality of cheese 488

analysis of 536 mycotoxins 509 pathogens in raw milk 487 488

(See also Bacteriology of milk) raw milk cheeses 498

Patulin 510 511

PCA. See Plate count agar

PCPs. See Pasteurized processed cheese products

PDO. See Protected Designation of Origin

PDWPC. See Partially denatured whey protein concentrate

Pecorino Romano cheese 296 394 396

Pecorino Sardo cheese 394 396

Pecorino Siciliano cheese 394 396

Pecorino Toscano cheese 394

Pecorino Umbro 500 501

Pediococcus 1 17 62 69 93 215 221 242

Penetration tests 318

Penicillic acid 511

Penicillium 359 497

Penicillium camemberti 17 217 223 225 230 243 253 390 413 498

inoculation with 218 proteinase of 268

Page 633: Cheese Science

614 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Penicillium camemberti (Continued) salt sensitivity of 165 208 225 toxic metabolites of 511

Penicillium candidum 209 354

Penicillium commune 235

Penicillium glabrum 235

Penicillium roqueforti 17 217 223 225 226 253 349 354 390 413 415

inoculation with 218 proteinase of 268 salt sensitivity of 165 225 toxic metabolites of 511

Penitrem A 512

Peptidases 71 214 261 264

Peptization 435

Petit Reblochon cheese 393

Petit Suisse cheese 382

PGE. See Pregastric esterase

pH changes during ripening 243 246 247 of curd for hard cheese varieties 15 effect of starter cultures on 11 94 emulsifying salts for displacement and stabilization of 435 437 for growth of lactic acid bacteria 210 influence on rennet coagulation 127 129 influence on rheological characteristics of cheese 331 333 influence on solubilization of micellar calcium 148 influence on syneresis 139 144 of medium used in production of starter cultures 95 for microbial growth 210 of milk 11 12 41 of milk gel for fresh acid-curd cheese varieties 375 of peptidases in lactic acid bacteria 73 salting and 15 162

Page 634: Cheese Science

615 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Phage. See Bacteriophage

Phagocytes 179

Phagocytosis 45

Phenylacetaldehyde 297

Phenylethylamine 512

Phoma 235

Phosphate in milk 40 42

Phospholipids in milk 25 27

Phosphorus in cheese 508

Phylogeny of lactic acid bacteria 66 70

Physicochemical properties of cheese 305

changes during maturation 309 of milk 19 42 43

Picante da Beira Baixa Amarelo cheese 394

Pichia 223 419

Pichia anomala 224

Pichia fermentans 224

Pichia kluyveri 224

Pichia membranaefaciens 224

Picodon de l’Ardèche/Drome cheese 393

Pizza cheese, analogue 446 aggregation index for 460 composition and functionality of 447 462 463 464 formulation for 446 functional stability during storage of 449 manufacturing protocol for 445 446 principles of manufacture for 447

Plant rennets 2

Plantaricin S 93

Plasmids 82

Plasmin 259 261 effect of stage of lactation on concentration of 184 185 exogenous, added to accelerate ripening 352

Page 635: Cheese Science

616 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Plasmin (Continued) heat stability of 238 inhibition of 238 salting and 166

Plasticization 147 466 472

Plate count agar (PCA) 536

PMNs. See Polymorphonuclear leukocytes

Polyacrylamide gel electrophoresis (PAGE) 284 530 533

Polymorphonuclear leukocytes (PMNs) 45 179

Polyunsaturated fatty acids in milk 26

Pont l'Evêque cheese 296 393 composition of 428 manufacture of 421

Port du Salut cheese classification of 390 391 free fatty acid concentration in 254 manufacture of 421

Postcoagulation operations 16 138 development of textured cheese 145 molding and pressing of cheese curd 150 packaging 151 purpose of 138 syneresis 15 16 30 138

Potassium citrate 437

Potassium in cheese 508

Pouligny Saint Pierre cheese 393

PR toxin 511

Preacidification of milk 12

Pregastric esterase (PGE) 251 254 354 355

Prematuration of milk 12 53

Preservatives in cheese 508

Pressing of cheese curd 150 influence on syneresis 144

Primost cheese 156 386 423 424

Page 636: Cheese Science

617 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Processed cheese. See Pasteurized processed cheese products

Prolidase 71 72 263 266

Prolinase 71 72

Proline iminopeptidase 71 72 263 266

Propionate 528

Propionibacterium 17 217 221 243 357 359

proteinase of 268 salt sensitivity of 164 water activity for growth of 208

Propionibacterium acidipropionici 221

Propionibacterium freudenreichii 221 357 subsp. shermanii 253 266 390 404 water activity for growth of 208

Propionibacterium jensenii 221

Propionibacterium thoenii 221

Propionic acid 51 210 217

Protected Designation of Origin (PDO) 392 393

Protein content of cheese 505 506 determination of 526

Proteinases cell envelope-associated 261 358 cheese-related, specificity of 257 exogenous, to accelerate ripening 353 indigenous milk 259 from secondary starter 268

Proteins, whey 33 37 514 515 517

addition to cheese milk 190 cheese yield and 188 191 denaturation by high heat treatment of milk 189 190 191 incorporation into cheese 189 partially denatured whey protein concentrate 192 whey protein concentrates 517 whey protein isolates 517

Page 637: Cheese Science

618 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Proteins in milk 31 casein micelles 35 caseins 31 during lactation 22 genetic polymorphs of 10 34

effect on cheese yield 185 influence on rennet coagulation 120 124 125 influence on rheological characteristics of cheese 328 329 333 335 influence on syneresis 139 interspecies differences in concentration of 20 31

influence on growth rate of neonates 31 32 minor proteins 39 seasonal changes in 184 whey proteins 37

Proteolysis 69 255 284 349 assessment of 256 529 characterization of 268 effect on functional characteristics of cheese 473 proteolytic agents in cheese and their relative importance 2 101 256 258 range in extent of 256 268 role in cheese texture and flavor 256 salt effects on rate of 166 specificity of cheese-related proteinases 257

coagulant 257 260 indigenous milk proteinases 259 proteinases from secondary starter 268 proteolytic enzymes from starter 261 264 267 proteolytic system of nonstarter microflora 263 268

Provolone cheese 131 146 413 classification of 391 composition of 156 428 463 free fatty acid concentration in 254 functional characteristics after cooking 462 464 465 lipolysis in 349 354 manufacture of 412

Page 638: Cheese Science

619 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Provolone cheese (Continued) volatile flavor compounds in 296 water activity of 156

Pseudolysogeny 85

Pseudomonas 46 47 210 252 357

Pseudomonas aeruginosa 495

Psychrotrophs in milk 46 47 49 influence on syneresis 145 lipases produced by 253

Putrescine 512

Pyrénées cheese 156

Pyrrolidone carboxylyl peptidase 265

Pyruvate 527

Q Quality of cheese

factors affecting 206 282 341 cheese composition 345 milk supply 341 nonstarter lactic acid bacteria 344 rennet 343 ripening temperature 347 starter culture 343

fat content and 30 flavor 282 fresh acid-curd cheese varieties 374 lactose concentration and 239 microbiological 232 488

(See also Pathogens and food-poisoning bacteria in cheese) analysis of 536

salting and 154 166 texture 333

Page 639: Cheese Science

620 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Quarg 2 5 14 39 78 96 363 379 422

classification of 388 391 composition of 365 428 cook temperature for 206 high concentration factor ultrafiltration for 192 high heat treatment of 189 packaging of 152 production processes for 380 protein content of 380 rheological properties of 324 salt content of 292 shelf life of 381 starter cultures for 55 60 use of rennet in 363 381 water activity of 156 212 whey proteins in 522

Queso Blanco 2 5 422 classification of 391 composition of 365 386 428 flavor of 386 production of 385 whey proteins in 522

Queso de Cincho 385

Queso de Matera 385

Queso del Pais 385

Queso Llanero 385

Queso Pasteurizado 385

R Raclette cheese

functional characteristics after cooking 464 465 manufacture of 406 rheological properties of 324

Page 640: Cheese Science

621 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Raclette cheese (Continued) water activity of 156

Rancidity 26 282

Randomly amplified polymorphic DNA (RAPD) techniques 66

Ras cheese 397 adding exogenous lipases to 355 classification of 391 manufacture of 401

Raschera cheese 394

Raw milk cheeses produced from 49

pathogens in 498 contamination of 45 487 488

(See also Bacteriology of milk) European Union standards for pathogens in 488 499 lactic acid bacteria in 46 54 risks associated with use of 12

RCT. See Rennet coagulation time

Reblochon cheese 393

Rennet-coagulated cheeses 392 classification of 390 391 internal bacterially ripened varieties 390 392

cheeses with eyes 403 extra hard varieties 392 hard varieties 397 high-salt varieties 408 Pasta filata varieties 410 semi-hard varieties 402

measuring residual coagulant activity 527 mold-ripened varieties 413 surface smear-ripened varieties 418

Rennet coagulation of milk 15 98 238 definitions related to 110 factors affecting 120

calcium 10 128

Page 641: Cheese Science

622 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Rennet coagulation of milk (Continued) cooling and cold storage of milk 126 milk fat level 123 milk homogenization 126 milk protein level 120 125 other factors 130 pasteurization temperature 102 125 127 128

189 pH 102 127 129 rennet concentration 130 renneting (set) temperature 126

immobilized rennets for 135 measurement of 109

dynamic curd firmness tests 113 nondynamic assessment of viscosity curd firmness tests 112 primary phase 112 rennet coagulation time 112 113 types of 111

postcoagulation treatment of milk gel 16 138 primary phase of 98

factors affecting hydrolysis of κ-casein and 102 secondary (nonenzymatic) phase and gel assembly 103 109

factors affecting 108 110 111

Rennet coagulation time (RCT) 101 103 measurement of 112 113

Rennet hysteresis 126 128

Rennet substitutes 3 130 effect on cheese yield 196 198 milk-clotting activity of 131 133

effect of pH on 133 134 molecular and catalytic properties of 132 proteolytic activity of 131 259 260 recombinant chymosins 133 required characteristics of 131 thermal stability of 131 133 135

Page 642: Cheese Science

623 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Renneting temperature 126

Rennets 1 101 added to fresh acid-curd cheeses 363 376 effect of rennet type on cheese quality 343 effect of rennet type on cheese yield 196 198 immobilized 135 increasing level to accelerate ripening 352 lipases from 251

Retinol 13 27

Reverse phase high performance liquid chromatography (RP-HPLC) 284 301 531 534

Rework cheese 441

Rheological properties 15 305 cheese microstructure 306 definition of 305 determinants of 305 differences according to cheese variety and age 305 factors influencing characteristics measured using force compression tests 323

cheese structure, composition, and maturity 328 test conditions 323

fresh acid-curd varieties 373 importance of 305 measurement of 317

creep and stress relaxation experiments 311 318 319 empirical tests 317 force compression tests between parallel plates 320 323 large-scale deformation tests 317

of processed cheese products 442 of raw cheese used as food ingredient 456 relationship between cheese characteristics and 322 324 terminology for 305 texture 333

(See also Texture of cheese)

Rheometry 116

Rhizomucor miehei 130 131 133 135 196 198 354

Page 643: Cheese Science

624 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Rhizomucor pusillus 130 131 133 196 198

Rhodococcus 220

Rhodococcus fascians 218 219

Rhodotorula 209 224 419

Ricotta cheese 363 384 391 composition of 365 428 manufacture of 422 nutrients in 505

vitamins and minerals 507 508 production of 384 shelf life of 385

Ricottone cheese 365 385 423 428

Ripening 12 17 506 acceleration of 349 biochemistry of 17 236 costs of 349 duration of 17 206 349

effect on functional characteristics of cheese 468 environmental humidity for 207 functions of 12 growth of pathogens during 493 microbial growth during 206 risks of 12 role of peptidases in 73 salt effects on 167 temperature for 213

Ripening agents in cheese 236 coagulant 236 contributions of individual agents to ripening 237 exogenous enzymes 237 milk 236 secondary microflora 237 starter culture 237

Robiola di Roccaverano cheese 394

Page 644: Cheese Science

625 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Romadour cheese 219 223 224 421

Romano cheese 5 131 classification of 392 composition of 156 428 flavor of 285 free fatty acids in 253 254 lactate concentration in 242 lipolysis in 349 354 manufacture of 396 rheological properties of 324 salt content of 292 salting of 160 volatile flavor compounds in 296 water activity of 156

Roncal cheese 156 393 397 402

Roquefort cheese 4 5 393 classification of 388 390 391 composition of 156 428 free fatty acid concentration in 254 manufacture of 417 418 nutrients in 505

vitamins and minerals 507 508 volatile flavor compounds in 296 water activity of 156 yeasts in 223

Roquefortine 511

Röse-Gottlieb method 526

RP-HPLC. See Reverse phase high performance liquid chromatography

S Saccharomyces 419

Saccharomyces cerevisiae 133 223 224 230

Saccharomyces unisporus 224

Saint Marcellin cheese 415

Saint Nectaire cheese 223 393

Page 645: Cheese Science

626 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Saint Paulin cheese 4 156 212 421 422

Sainte Maure de Touraine cheese 393

Salers cheese 393

Salmonella food poisoning outbreaks involving 484 485 growth in cheese during ripening 493 495 496 in raw milk 487 488 water activity for growth of 158

Salmonella dublin 485 486

Salmonella heidelberg 487

Salmonella javiana 485

Salmonella oranienberg 485

Salmonella paratyphi B 485

Salmonella typhimurium 485 486 495

SALP. See Sodium aluminum phosphate

Salt-free cheese 167

Salting 2 16 153 See also Water activity of cheese curd 155

brine-salting 157 160 161 cheese size and geometry and 160 concentration gradient 158 161 curd pH and 15 162 dry-salting 155 159 moisture content of cheese curd and 162 163 salting time 158 temperature of curd and brine 162

determination of salt content of cheese 526 effect on cheese composition 162 164 165 effect on cheese quality 154 166 effect on enzymes in cheese 165

coagulant 165 microbial enzymes 166 milk proteinases 166

Page 646: Cheese Science

627 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Salting (Continued) effect on microbiology of cheese 154 158 163 166

167 effect on rheological characteristics of cheese 332 334 effect on syneresis 144 effect on water activity 208 212 213 nutritional aspects of 167 506 508 preservative action of 153 salt content of specific cheese varieties 154 156 292

fresh acid-curd cheeses 365

Salts of milk 39 42

Sampling methods for cheese 523

Samsoe cheese 156 428

San Jorge cheese 394

Sapsago cheese 254

Sbrinz cheese classification of 391 flavor of 284 manufacture of 395 water activity of 156 212

Schabzieger cheese 4

Schmelzkäse 430

Schmelzkäsezubereitung 430

Schmid-Bondzynski-Ratzlaff technique 526

Schwangenkase cheese 4

Science and technology of cheese production 7

Scopulariopsis fusca 209

SDS. See Sodium-dodecylsulfate

Seasonal changes in milk composition effect on cheese yield 183 effect on rheological characteristics of cheese 332

Secondary cultures 237 to accelerate ripening 359 proteinases of 268

Page 647: Cheese Science

628 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Selles sur Cher cheese 393

Semi-hard cheeses 388 389 390 391 402

growth of pathogens during ripening of 493

Sensory analysis of cheese flavor and texture 336 339 540

Serpa cheese 394

Serra da Estrela cheese 156 394 422 428

Shape of cheeses 150 effect on salt absorption 160

Shattering of cheese curd 118

Sheep's milk 5 13 cheese yield from 177 composition of 19 20 177 detection of interspecies adulteration of 543

Shigella 158

Shigella sonnei 486

Shreddability of cheese 457 460

Simplesse 100 192

Siro-Curd process 192

16S rRNA molecule 66

Size of cheeses 150 effect on salt absorption 160

Sodium aluminum phosphate (SALP) 434 436 437 438 439

Sodium citrates 434 436 437 438 439

Sodium-dodecylsulfate (SDS) 530

Sodium orthophosphates 434 436 437 438 439

Sodium polyphosphates 434 436 437 438 439

Sodium pyrophosphates 434 437 438 439

Sodium tripolyphosphates 434

Page 648: Cheese Science

629 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Soft cheeses 388 389 growth of pathogens during ripening of 496 microbiological quality of 488 pH during ripening of 497 498

Solva 434 436

Solvent-treated lactic acid bacteria 357

Somatic cell count (SSC) 179 343

Sorbic acid 508

Soybean cheese 450

Spanish Burgos cheese 206

Spino for cutting milk gel 140 141

Spoilage, microbial 232

SSC. See Somatic cell count

Stabilizers added to fresh acid-curd cheeses 376 379

Standardization of milk composition 10 169

Standards of identity for cheese varieties 10 170 for pasteurized processed cheese products 430

Staphylococcus 93 222 419

Staphylococcus aureus 45 222 489 bovine mastitis due to 499 enterotoxins produced by 487 food poisoning outbreaks involving 484 485 486 growth during cheese manufacture 493 494 growth in cheese during ripening 493 495 496 infective dose of 487 in raw milk 487 488 499 water activity for growth of 158 208

Starter concentrates 11

Starter cultures 17 54 237 to accelerate ripening 355 adjunct 59 358 artisanal (natural) 56 59 attenuated 356

Page 649: Cheese Science

630 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Starter cultures (Continued) heat- or freeze-shocked cells 356 lysozyme treatment 356 mutant starters 357 other bacteria as additives 357 solvent-treated cells 357

bacteriocins 92 characteristics of lactic acid bacteria in 56 D, L, and O types 59 defined-strain 54 differentiation of lactic acid bacteria in 66 67 68 effect on cheese quality 343 effect on cheese yield 194 effect on pH of milk 11 94 function of 54 genetically engineered 358 growth in cheese during ripening 213 inhibition of acid production 83 84

antibiotics 83 84 bacteriophage 12 83 salt 163

lysis of 214 measurement of generation times 96 mesophilic and thermophilic 54 metabolism of 69

acetaldehyde 82 arginine 73 citrate 77 lactose 73 78 79 proteolysis 69

methods for enumeration of 536 mixed-strain 54 58 mutant 357 plasmids and 82 production in cheese plants 94

Page 650: Cheese Science

631 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Starter cultures (Continued) rod:coccus ratio and acid production 58 59 secondary 237 268 359 selected 355 for specific cheese varieties 60 temperature effects on growth of 60 types of 54

Sterigmatocystin 511

Stilton cheese 4 classification of 391 composition of 428 manufacture of 418 nutrients in 505

vitamins and minerals 507 508 rheological properties of 324 salting of 155 water activity of 156

Stokes' law 29

Storage of cheese 2 7 of milk 47 48

effect on cheese yield 186 effect on rennet coagulation 126

Strecker degradation reaction 297 300

Streptococcus 1 46 62 63 64 93

Streptococcus agalactiae 45

Streptococcus aureus 537 540

Streptococcus bovis 64 69

Streptococcus dysgalactiae 45

Streptococcus equinus 64

Streptococcus salivarius 64

Streptococcus thermophilus 49 54 55 56 58 59 64 358

amino acid requirements of 69

Page 651: Cheese Science

632 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Streptococcus thermophilus (Continued) antibiotic inhibition of growth of 84 Asian strains of 71 distinguishing characteristics of 67 electron microscopic appearance of 68 growth in cheese during ripening 214 215 228 lack of proteolytic activity in 70 lactose metabolism in 76 77 methods for enumeration in cheese 537 peptidases of 264 proteolytic system of 261 salt sensitivity of 164 temperature effects on growth of 61 water activity for growth of 208

Streptococcus zooepidemicus 485

String cheese 412

Substitute/imitation cheese products 443 analogue cheeses 443 classification of 430 definition of 443 filled cheeses 450 labeling requirements for 443 tofu or soybean cheese 450

Sugars in milk 20 See also Lactose

Sun-dried cheese 2

Surface mold-ripened cheeses 413

Surface smear-ripened cheeses 391 418

Svecia cheese 428

Svenbo cheese 284

Svicia(ost) cheese 403

Swiss-type cheeses 14 15 17 302 classification of 391 dehydration of 138 lactate in 242 244

Page 652: Cheese Science

633 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Swiss-type cheeses (Continued) manufacture of 404 rheological properties of 324

Syneresis 15 16 30 138 influence of compositional factors on 139 influence of processing variables on 140

cook temperature 140 milk quality and pretreatment 144 pressing 144 rate of acid development 144 salting 144 size of curd particles 140 141 142 stirring of curd-whey mixture 144

kinetics and mechanism of 145 methods for measurement of 139 pH and calcium effects on rate of 139 for production of fresh acid-curd cheese varieties 371

T T-2 toxin 511

Taleggio cheese 4 394 422 classification of 391 pH changes during ripening 247 water activity of 156

Taxonomy of lactic acid bacteria 62

Telemes cheese 408 410

Temperature cold storage of milk

effect on cheese yield 186 effect on rennet coagulation 126

effect of cooking temperature on syneresis 140 effect on growth of starter cultures 60 effect on microbial growth in milk 46 47 effect on salt absorption 162 incubation temperature for fresh acid-curd cheese varieties 375

Page 653: Cheese Science

634 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Temperature (Continued) ripening temperature

effect on cheese quality 347 effect on microbial growth in cheese 213 elevated, to accelerate ripening 350 for specific cheese varieties 350 upper limit for 350

Tetilla cheese 460

Tetragenococcus 62

Tetrasodium pyrophosphate (TSPP) 436

Texture of cheese 2 7 333 See also Rheological properties acidification and 14 assessment of

comparison of instrumental and sensory evaluation 339 instrumental methods 336 338 339 540 sensory analysis 336 540

definition of 306 definitions of mechanical properties 337 development of textured cheeses 145 effect on perception of flavor 283 fresh acid-curd varieties 373 geometrical properties contributing to 306 mechanical properties contributing to 306 other properties contributing to 306 primary, secondary, and tertiary characteristics 306 307 role of proteolysis in 256

Thermal evaporation of water 16

Thermization of milk 49 effect on cheese yield 188

Thermoduric bacteria 48

Thiocyanate 50 83

Tilsit cheese 15 17 422 classification of 391 growth of pathogens during ripening of 494

Page 654: Cheese Science

635 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Tilsit cheese (Continued) microbial growth in 218 219 232 234 starter cultures for 60 water activity of 156 212

Tocopherols in milk 27

Tofu 450

Torta del Casar cheese 156

Torulaspora delbrueckii 224

Trappist cheese 296 391 422

Triacylglycerols in milk 25 27

Trichosporon 209

Trichosporon beigelii 224

Tricoderma reesei 133

Triglyceride hydrolysis by lipases 249 250

Tripeptidases 72 263 266

Trisodium citrate (TSC) 436 437

TSPP. See Tetrasodium pyrophosphate

Tuberculosis 484

Tvorog. See Quarg

Tyramine 501 512

U Ultrafiltration 16 425

high concentration factor 192 425 low concentration factor 170 178 425 medium concentration factor 425

Urea-PAGE 530 533

V Vacherin cheese 296

Vagococcus 62 63 69

Vancomycin-resistant enterococci (VRE) 500

Page 655: Cheese Science

636 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Varieties of cheese 4 388 classification of 388

(See also Families of cheese) composition of 156 428 463

fatiprotein ratio 10 lactate concentration 242 243 nutrients 505 salt content 154 156 292

evolution of 4 4 fresh acid-curd cheeses 363 involved in food-poisoning outbreaks 485 lipolysis in 253 254 number of 1 388 with Protected Designation of Origin 392 393 standards of identity for 10 170 starter cultures for 60 water activity values for 156 212

Vibrio parahaemolyticus 158

Viscoelasticity cheese 313 319 328 See also Rheological properties

Vitamins in cheese 506 507 in milk 27

Volatile compounds in cheese contribution to flavor 292 294 295 296 inter- and intravarietal comparisons of 301 methods for analysis of 285 536

von Slyke yield formula 174 179 180

von Smoluchowski theory 103 105

VRE. See Vancomycin-resistant enterococci

W Water activity (aw) 153

definition of 207 deterioration rates of food systems as function of 157

Page 656: Cheese Science

637 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Water activity (Continued) determination of 526 for microbial growth 158 207 209 of specific cheese varieties 156 212

reduction during ripening 208 210 211

Water buffalo's milk 5 13 cheese yield from 177 composition of 19 21 177

Water-soluble nitrogen (WSN) 530 535

Weinkase cheese 219 224

Weissella 62 69

Weisslacker cheese 421

Wensyledale cheese 324 402

Westfalia thermoprocess 380

Whey and whey products 2 17 138 514 acid whey and sweet whey 514 515 clarification of whey 515 concentrated and dried products 516

delactosed and delactosed-demineralized whey powder 516 demineralized whey powder 516 nonhygroscopic whey powder 516

disposal of whey 515 fermentation products 522 lactose 516 stirring of curd-whey mixture 144 whey cheese 18 386 390 519

manufacture of 423 water activity of 156

whey proteins 33 37 514 515 517

addition to cheese milk 190 cheese yield and 188 191 denaturation by high heat treatment of milk 189 190 191 incorporation into cheese 189 partially denatured whey protein concentrate 192

Page 657: Cheese Science

638 Index terms Links

This page has been reformatted by Knovel to provide easier navigation.

Whey and whey products (Continued) whey protein concentrates 517 whey protein isolates 517

Whey cultures 59

World production of cheese 5 6 95 169 484

fresh acid-curd varieties 363 pasteurized processed cheese products 429

WSN. See Water-soluble nitrogen

X X-prolyldipentidyl aminopeptidase 71 72 263 265

Y Yarrowia lipolytica 224

Yeasts 217 222 359 415 methods for enumeration in cheese 537 multiplication of 223 as spoilage organisms 235 on surface smear-ripened cheeses 419 water activity for growth of 207 209

Yersinia enterocolitica 232 484 growth in cheese during ripening 495 in raw milk 487 water activity for growth of 158

Yield. See Cheese yield

Ymer cheese 381

Yogurt 60 82

Z Zamorano cheese 156

Zearalenone 510 511

Ziger cheese 423

Zinc in cheese 508

Zsirpi cheese 385 Zygosaccharomyces rouxii 224