bitter flavour in dairy products. i. a review of the

HAL Id: hal-00929271https://hal.archives-ouvertes.fr/hal-00929271

Submitted on 1 Jan 1991

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Bitter flavour in dairy products. I. A review of thefactors likely to influence its development, mainly in

cheese manufactureL Lemieux, Re Simard

To cite this version:L Lemieux, Re Simard. Bitter flavour in dairy products. I. A review of the factors likely to influenceits development, mainly in cheese manufacture. Le Lait, INRA Editions, 1991, 71 (6), pp.599-636.�hal-00929271�

https://hal.archives-ouvertes.fr/hal-00929271

https://hal.archives-ouvertes.fr

Lait (1991) 71, 599-636© Elsevier/INRA

599

Review article

Bitter flavour in dairy products.1. A review of the factors likely to influence

its development, mainly in cheese manufacture

L Lemieux, RE Simard

Université Laval, pavillon Paul-Comtois, centre de recherche STELA,département de sciences et technologie des aliments, Sainte-Foy, Québec, Canada, G 1K 7P4

(Received 11 May 1990; accepted 23 August 1991)

Summary - Following a definition of bitte mess and a mention of the Iink between the bitterness ofpeptides and their content of hydrophobie amino acid residues, a detailed description is given of thenumerous factors Iikely to influence the development of bitter off-flavour defect in dairy products withemphasis on cheese. Factors such as milk quality, pH, psychrotrophic bacteria, fat and mineraicontent, the cheese manufacturing process including the cooking temperature, salt concentrationand manner of draining the curd, the acidity or the pH of cheese, and sanitary processing andhygienic packaging conditions are less important than bacterial proliferation in the cheese vat andthe type and amount of starter and rennet used. The major roles played by these last three factorsare better understood since the development of the aseptic vat technique. Finally, possible ways arementioned of keeping the concentration of bitter peptides below the threshold level for bitter tastedetection.

bitterness / dairy product / cheese / peptide / hydrophobicity

Résumé - L'amertume dans les produits laltiers.L Une revue des facteurs susceptibles d'in-fluencer son apparition, principalement au cours de la fabrication fromagère. Après avoir défi-ni l'amertume et mentionné la relation existant entre l'amertume des peptides et leur contenu enacides aminés hydrophobes, les nombreux facteurs susceptibles d'influencer l'apparition du défautd'amertume dans les produits laitiers sont décrits en détail. Les facteurs tels que la qualité du lait,son pH, son contenu en bactéries psychrotrophes, en matière grasse et en minéraux, le mode de fa-brication du fromage et la température de cuisson, l'acidité ou le pH du fromage, les conditions sani-taires de transport du lait, de fabrication et d'emballage du produit fini sont moins importantes que laprolifération bactérienne dans le bac à fromage et que le type et la quantité de ferment et de présureutilisés. La compréhension du rôle joué par ces 3 derniers facteurs a été rendue possible avec l'avè-nement de la technique de fabrication aseptique du fromage. Pour terminer, sont mentionnésd'éventuels moyens de conserver la concentration en peptides amers au-dessous du seuil de détec-tion de l'amertume.

amertume / produit laitier / fromage / peptide / hydrophobieité

600 L Lemieux, RE Simard

INTRODUCTION

Taste buds, which are located in papillaeon the tongue, can detect the characteris-tic taste of food determined by the balanceof the primary (sweet, sour, salt, bitter andumami) and/or secondary (astringent, me-tallic and hot) tastes (Nishimura and Kato,1988; Chettel et al, 1977). Umami is thecharacteristic taste of monosodium gluta-mate and 5'-ribonucleotides.

Sometimes desired and regarded as atypical flavour note (eg, in grapefruit) andin some cases indesirable and objected toas off-flavour (eg, in cheese). bitterness offoodstuffs may be caused either by natu-rally occurring bitter compounds or byproducts from chemical reactions that oc-cur during storage or processing.

Characterised by a pleasant and slightlysweet taste, milk is very susceptible to fla-vour defects from a variety of sources (ab-sorption of flavours, bacterial contamina-tion and chemical reactions such aslipolysis (Charalambous, 1986)), and un-less properly heat-treated and refrigerated,it has a very short shelf-life. For centuriescheese-making has been the only meansof preserving the major constituents ofmilk, and among the variety of cheesetypes some can be considered as prod-ucts with real long-term storage capabil-ities. The proteolytic activity of starter cul-tures produces bitterness in cheese. Bythis process, caseins are broken down intopeptides of low molecular weight (-1400Da) with bulky hydrophobic groups to-wards the C-terminal end of the peptide(Matoba and Hata, 1972). Peptide bitter-ness, caused by the hydrophobic propertyof the amine acid side chain, can be pre-dicted from the rule proposed by Ney(1971, 1979).

Bitterness in yoghurt has been shownto be caused by the proteolytic activity of

Lactobacillus bulgaricus during storage(Renz and Puhan, 1975). Gelation and theoccurrence of bitter flavour in UHT milk arecaused by heat-stable proteases and llpo-Iytic enzymes produced by psychrotrophicbacteria. However, milk processing tem-perature is critical; while appreciableamounts of lipase may be inactivated byheating milk at 64 oC for 10 s, heat treat-ments required to inactivate proteases alsodenature the milk proteins (Charalambous,1986).

The greatest contribution to the intensityof cheese flavour has been shown to bepresent in the water-soluble fraction(McGugan et al, 1979). Bitterness is mostintense in that fraction and has been attrib-uted to medium sized (tri- to hexa-) pep-tides (Biede and Hammond, 1979) result-ing from the enzymic digestion of casein.Beta-casein may be the source of bitterpeptides produced by the proteinase(s)present in the cell wall of starter bacteria(Exterkate, 1976), while chymosin (rennet)may produce bitter peptides from ail caseincomponents (Exterkate, 1983).

Following the development of technolo-gy for the acceleration of the cheese ripen-ing process (Fox, 1988-1989), develop-ment of low fat cheeses, improved shelf-life, transformation of milk to new foodproducts and alteration of milk componentsby use of recombinant DNA techniques(Muysson and Verrinder Gibbins, 1989),the dairy industry will have to find ways toavoid this potential defect of bitterness.

From the literature surveyed it is surpris-ing to find that bitterness, although report-ed for ail dairy products, has been studiedmost extensively in Cheddar and Goudacheeses. This research has identified 20factors related to bitter off-flavour whichwill be described in detail in this review. ltis understandable that most of the text willbe devoted to bitterness in cheese.

Bitter flavour in dairy products: part 1 601

BITTERNESS-HYDROPHOBICITY

Bitterness, a flavour defect which is en-countered quite frequently in Cheddar andGouda cheeses, results from the accumu-lation of bitter-tasting peptides formed bythe action of proteolytic enzymes on case-in (Fryer, 1969; Sardinas, 1972; Sullivanand Jago, 1972; Sullivan et al, 1973; Stad-houders, 1974; Stadhouders and Hup,1975; Lawrence et al, 1976; Creamer,1978). The bitter peptides have specialchemical praperties which enable them tointeract with the taste buds at the back ofthe tongue to give the sensation of bitter-ness.

The existence in the bitter peptides of 2bitter taste determinant sites which canbind to the bitter taste receptor has beenproposed by Okai (1977), who also postu-lated that the hydrophobie residue of apeptide acts as a first binding site and inthe presence of the second site, called thestimulating site, a bitter taste is detectable.However, the mechanism of the secondsite has not yet been resolved (Ishibashi etal, 1987).

The concentration of the bitter peptidesin the cheese must exceed a certainthreshold level before bitterness can bedetected. The overall level of bitterness incheese might be determined finally by therelative rates at which bitter peptides areformed and broken down to non-bitterproducts (Jago, 1974).

Early on it was suggested, based on thestudy of Emmons et al (1960b) on 132cheeses, that cheese bitterness was dueto unhydralysed peptides. These authorsobtained a correlation coefficient of 0.90between the TCA-soluble N/amino N ratio,which is an estimate of the average nurn-ber of amino groups per peptide in theTCA soluble extract, and the level of bitter-ness.

It is now accepted that the bitter flavourproduced during enzymic hydrolysis ofcasein is due to particular types of pep-tides (Fujimaki et al, 1970). Ney (1971)was the first to propose the Q-hypothesis,a semi-quantitative relationship betweenthe amino acid composition of a peptideand its bitterness. According to Ney(1979), no particular single amino acid orsequence was needed to impart the bittertaste. However, Japanese workers (eg,Shinoda et al, 1985, 1986a, b) proved lateron, by synthesizing bitter peptides andseveral analogs, that the nature of the ter-minai amino acids and their sterie parame-ters have some significance in the intensityof bitter taste.

BITTER FLAVOUR DEVELOPMENTIN CHEESE

Cheese-making, which is an important wayof preserving the nutritive components ofmilk, is practised in various parts of theworld. That explains why many differentcheese varieties have been developed.However, the results obtained using onecheese variety with a particular milk coagu-lant cannot be extrapolated to anothercheese variety or cheese made with a dif-ferent milk coagulant. Modification of thecheese-making conditions and parameters(choice of starter, rennet concentration,cooking temperature and salt addition(Harwalkar and Elliott, 1971; Creamer,1978) has helped in the study of cheesebitterness.

Bitter peptides are released from thecasein molecules primarily by the action ofthe rennet and bacterial proteinases; acontribution may also be expected frombacterial peptidases, which may reducethe size of peptides initially too large togive a bitter taste (Sullivan and Jago,1972; Sullivan et al, 1973). The production

602 L Lemieux,RESimard

of bitter peptides during the process of en-zymic digestion is not always unfavoura-ble; indeed, a bitter taste is one of the im-portant components of cheese tastequality (Shinoda et al, 1985; 1986b). How-ever, when bitter peptides accumulate,their concentration may exceed the flavourthreshold for bitterness and can limit ac-ceptance of the cheese, which will then berated bitter (Visser et al, 1983b, cl. ltshould be noted that this threshold is con-sidered to increase with the age of acheese because of the increase in otherflavour components (Creamer, 1979).

Bitterness has long been recognised asa major defect in Gouda (Visser et al,1983b) and Cheddar cheeses (Emmons etal, 1962a; Richardson and Creamer, 1973;Hamilton et al, 1974; Crawford, 1977;Champion and Stanley, 1982; Edwardsand Kosikowski, 1983), but it also occursin Camembert and similar types of chees-es produced in France (Pélissier et al,1974; Mourgues et al, 1983). This defecthas also been reported in Swiss mountaincheese (Guigoz and Solms, 1974), But-terkâse (Huber and Klostermeyer, 1974),Japanese yeast-ripened cheese (Kanekoand Yoneda, 1974), Gorgonzola (Delformoand Parpani, 1986) and Cottage cheese(Sandine et al, 1972), indicating the uni-versality of this problem, which is a compli-cated defect affected by many factors. Incheese varieties other than Cottage, bitterflavour development has been associatedwith proteolysis (Stone and Naft, 1967).

Proteolysis in cheese can be consid-ered synonymous with the degradation ofthe caseins to peptides and amine acids.Although analysis of the casein sequencesreveals that they contain many hydropho-bic fragments which are related to the for-mation of a bitter taste during the hydroly-sis of these proteins, identification of bitterpeptides isolated from whole casein andfrom cheese has shown that they originate

mostly from us1- and B-caseln. lt seemsreasonable that us1- and ~-caseins withhigh average hydrophobicities of 1.17 and1.33 kcal/residue, are potential sources ofbitter peptides on hydrolysis (Visser,1977c). Until now, no reports on the identi-fication of bitter x-casein fragments incheese have appeared, although Visser etal (1975) have shown that rennet or puri-fied chymosin could generate bitter peptidematerial from para-x-casein.

The C-terminal portion of ~-casein hasan extremely bitter taste (Visser et al,1983b; Shinoda et al, 1986a) and is theprincipal source of bitter peptides in Goudacheese (Visser et al, 1983b, c). In Cheddarcheese, ~-casein is highly resistant to pro-teolysis while usrcasein is extensively de-graded; the main reason for this resistanceto proteolysis is the formation of B-caselnpolymers that are not readily hydrolysed(Phelan et al, 1973; Creamer, 1975).

Beta-casein was earlier suggested asbeing the main source of bitterness in dairyproducts (Sullivan and Jago, 1972). In-deed, strains which produce bitterness incheese degrade ~-casein at a faster ratethan non-bitter strains (Sullivan and Jago,1972); moreover, processes used for thepreparation of partly digested dairy prod-ucts from whole casein will of necessitygive rise to bitter tasting products unlessthey are capable of reducing the peptiderepresenting the sequence 53-79 of ~-casein to very small fragments (Clegg etal, 1974). However, the occurrence of bit-ter defects was observed to be rare in goatand ewe cheeses compared to cow chees-es; cow's milk contains more uS1-casein,hydrolysates of which are more bitter thanthose of ~-casein, suggesting that thesmaller the amount of usrcasein, the lessbitterness that develops in the cheese(Pélissier and Manchon, 1976).

Bitter peptides produced from casein bythe action of rennet (or chymosin), as weil

Bitterflavourindairyproducts:part 1 603

as proteinases from the cell wall of certainstarter bacteria, are degraded by the ac-tion of proteolytic enzymes from the cyto-plasmic membrane of bacterial cells, eitherin concert with enzymes from the cyto-plasm or without their aid. It is known thatsalt strongly influences the net formationand degradation of bitter peptides (Visser,1977b; Visser et al, 1983a). Althoughmany studies report on bitterness incheese made with various rennet substi-tutes, the action of exopeptidases from mi-crobial rennets may help in preventing theoccurrence of bitterness, and in accelerat-ing the cheese ripening process (Malkki etal, 1978).

THE ASEPTIC TECHNIQUE

Before the development of the aseptic vattechnique (Mabbitt et al, 1959), the bitterflavour defect had been observed by nu-merous workers (Moir, 1930; Price, 1936;Raadsveld, 1953), some of whom hadbeen able to show specific causes for il.Strict bacteriological control of the cheeseduring making was not exercised at thattime, and it was generally believed, fromthe results of cheese-making experimentsthat starter was the main agent involved inthe maturation of Cheddar cheese, whileother organisms, in particular lactobacilli,were found to be necessary for the produc-tion of Cheddar flavour. However, accord-ing to Fryer (1969), the results of such rel-atively uncontrolled experiments mightjustifiably be ignored.

Aseptic curds are obtained using a tech-nique of artificial milk acidification (Mabbittet al, 1955) by the addition of ô-qluconicacid lactone in place of starter (Visser,1977a; Stadhouders et al, 1983). Generalinformation concerning the role of the dif-ferent ingredients used for cheese manu-facture and the general mechanism of pro-

tein breakdown during cheese ripening(LeBars et al, 1975; Desmazeaud andGripon, 1977) have been obtained fromthese aseptic curds.

Aseptic manufacture of cheese hasbeen performed principally with Cheddarand Gouda cheeses. In the former case,research was mainly directed towards thestudy of the typical Cheddar flavour (Reiteret al, 1967; McGugan et al, 1968; Reiterand Sharpe, 1971). The aseptic techniqueof cheese manufacture has conclusivelydemonstrated the central role of starter lac-tococci in the development of Cheddarcheese flavour, and that different f1avourintensities are related to different starters(Reiter et al, 1967). The contribution ofstarter lactococci to bitterness productionwas also studied by Lowrie et al (1974) un-der controlled bacteriological conditions. Ithas been concluded, from the study of therate of acid development during Cheddarcheese manufacture, that the excessiveproteolysis which accompanies rapid acidi-fication does not result in the developmentof bitter flavour (O'Keeffe et al, 1975).

To study aspects of cheese ripening,aseptically made cheeses in which the pro-tein is decomposed during ripeningthrough the separate and combined ac-tions of rennet, starter bacteria and milkprotease must be used (Gripon et al, 1975;Visser, 1977a; Visser and de Groot-Mostert, 1977).

Studies conducted by Visser (1977b, c,d), Visser and de Groot-Mostert (1977)and others in which normal aseptic, asep-tic starter-free, aseptic rennet-free, andaseptic rennet and starter-free Goudacheeses were manufactured, have shownthat rennet and bitter starter bacteria wereable, separately, to produce bitterness incheese, although the starter bacteria ap-peared to do this far more specifically thandid rennet by breaking down para-caseinin peptides of low molecular weight (MW).


Cheeses did not develop bitterness whennon-bitter starters were used. It was alsoobserved that higher levels of free aminoacids accumulated in aseptic rennet-freeand normal aseptic cheeses when non-bitter starters were used instead of bitterones.

Rennet, which is responsible not onlyfor coagulation but also for some aspectsof the subsequent ripening (Scott, 1972;Visser, 1977c), has been shown to Iiberatepeptides of high and low MW but only verysmall amounts of amino acids (AA). In nor-mal cheeses, starter peptidases hydrolysepeptides produced by the action of the ren-net leading to the accumulation of AA andlow-MW « 1400 Da) peptides. Smallamounts of AA and low-MW peptides arealso liberated by milk protease.

An experiment conducted by Scott(1972) on Cheddar cheese has indicatedthat the breakdown of milk constituentswas present in the cheese in spite of irradi-ation destruction of bacteria. Although thistreatment produced an intense rancid/bitter flavour, it has shown that once en-zyme systems are present in the curd,living organisms are not necessary for thedegradation of milk constituents, as occursin cheese during ripening.

FACTORS L1KELY TO INFLUENCEBITTER OFF-FLAVOUR DEVELOPMENTIN DAIRY PRODUCTS

Milk quality

Dairy products possessing good flavourand shelf-life cannot be consistently pro-duced when made from raw milk of poororganoleptic and bacteriological quality(White et al, 1978).

The feeding of good-quality silage tocows usually results in milk of good quality

(Polansky, 1989), while feeding beet andturnip tops and roots to cows has beenfound to cause a bitter flavour in the milk(Bassette et al, 1986). When bitter sneeze-weed (Helenium amarum) - a warm-season annual widely distributed in theeastern and southern United States - iseaten by cows, it imparts a bitter taste tothe milk and renders it unpalatable (Smith,1989). A single dose of 500 9 or more ofgreen Hamarum to a lactating cow was re-quired to cause appreciable bitterness inmilk (Ivie et al, 1975). This off-flavour wasfound to be very persistent and did not dis-appear from one milking to the next.

The intensity and persistence of feedand weed flavours in milk are a function ofthe time interval between feeding and milk-ing, ventilation in the barn, soil fertility, andof the physical condition of the cow. Be-sides the feeding (Delbeke and Naudts,1970), the composition of milk also variesaccording to the season (Lawrence andGilles, 1971), the stage of lactation (Okig-bo, 1986), the breed (Richardson andCreamer, 1973), the health of the animal(Sorokina and Karagueuz, 1978) and eventhe individual (Reiter and Sharpe, 1971).

It is generally suggested that milk withpoor rennet coagulation properties shouldnot be used for the manufacture of cheeseowing to a reduced yield of cheese with ahigh moisture content and a very bitterflavour (Okigbo, 1986). Significant quanti-ties of casein c1eavageproducts are to beexpected in bulk milk supplies containingsubstantial amounts of such milk. How-ever, casein c1eavage products are alsopresent in normal milk, for example in mid-lactation milk, but in limited quantities, withtime, as more plasminogen is activatedand microbial flora increases, higher con-centrations of such products accumulate(Visser etaI, 1983b).

Milk as produced in the mammaryglands is considered to be sterile but mi-


crobial contamination occurs during pas-sage through the udder. Flavour defectssuch as bitterness, resulting from this mi-crobial contaminàtion in raw and pasteur-ised milk, can occur at any stage of pro-duction and processing (Bassette et al,1986).

Psychrotrophic bacteria

Psychrotrophic bacteria are present inmost raw milk supplies and can grow read-ily at refrigeration temperatures, producingproteolytic and lipolytic enzymes. The re-sults obtained by Gebre-Egziabher et al(1980) and many others (eg, Richardsonand Newstead, 1979; Dousset et al, 1988)have shown that ail psychrotrophs werehydrolysing casein in raw and processedmilk, thus causing bitterness problems.

During cold storage of milk at the farmand at the dairy before processing, thegrowth of acid-producing bacteria is sup-pressed while psychrotrophic bacteria, af-ter a certain lag period, start to develop.Several psychrotrophic bacteria (eg, Pseu-domonas f1uorescens P26, B12 and B52) es-pecially if they reach a cell population of106/ml, may produce extracellular lipasesand proteases that can survive the heattreatments encountered in pasteurisation(73 ± 1.0 oC for 16 s) and in the UHT pro-cessing of milk (135-150 "C for 3-10 s),and cause the development of bitter taste(Shipe et al, 1978; Dousset et al, 1988).Both bacterial and native (alkaline) milkproteinases can hydrolyse casein and pro-duce bitter off-flavour in milk and milk prod-ucts with an extended shelf-Iife (White andMarshall, 1973b; Richardson and News-tead, 1979; Visser, 1981; Torrie et al,1983; Dousset et al, 1988). White and Mar-shall (1973a) have shown that contamina-tion of raw milk having a bacterial count of< 33/ml inoculated with 160 cells/ml of

Pseudomonas f1uorescens P26 producedsignificant proteolysis and bitter off-flavourin UHT pasteurised milk stored at 4.4 oCfor up to 20 days. However, even in the ab-sence of post-pasteurisation contaminants,off-flavours may be encountered after pas-teurisation of raw milk containing largepopulations of psychrotrophs due to: 1)end-products of microbial metabolism; 2)constituents of large numbers of heat-inactivated and Iysed bacterial cells; 3)heat-stable microbial enzymes; 4) pres-ence and growth of thermoduric psychro-trophs (Patel and Blankenagel, 1972).

A lipase and severai esterases occurnaturally in milk, the lipase attacks the milktriglycerides, releasing free fatty acids andforming mono- and diglycerides which areresponsible for the bitter flavour that occa-sionaliy accompanies lipolysis (Lebedevand Umanskit, 1981; Choisy et al, 1984;Bassette et al, 1986). Dibutyrin is the bitterflavour principle in homogenised raw milk.Since the growth of lactic streptococci isdepressed in lipolysed milk, it is possiblethat slow lipolysis which occurs in milk heldin bulk tanks could produce enough acidsto slightly inhibit the growth of lactic cul-tures and, as a result, the flavour andodour of milk may be altered and the sur-face and interfacial properties modified(Jensen, 1964). In order to help the milk torecover its physico-chemical and biochemi-cal properties, Desmazeaud (1983) hasproposed the storage of good bacterialquality milk for 12-15 h at 10°C in thepresence of 0.1-0.2% of starter.

McKeliar (1981) was the first to reportan attempt to associate bitterness develop-ment in milk with proteolysis measurement.He found that proteolysis ranging from0.289-0.554 and from 0.499-0.746 umolof trichloroacetic acid-soluble free aminegroups/ml for UHT and pasteurised milk,respectively, was necessary for significantoff-flavour development.


Bitter flavour which frequently devel-oped under commercial handling of Cot-tage cheese was associated with increas-es in soluble nitrogen and psychrophilicbacteria. Also, much of the White cheeseoffered in Egyptian local markets is bitterand highly acidic due to raw milk heavilycontaminated with lactobacilli (Abo-Elnaga, 1974). Similarly, Greek Telemecheeses made with raw milk stored for 5days were unacceptable for body and tex-ture and were rejected because of rancidi-ty, unclean flavour and bitterness due tothe presence of psychrotrophic bacteria(Kalogridou-Vassiliadou and Alichanidis,1984). Other workers (Chapman et al,1976; Law et al, 1976; Hicks et al, 1986)have found similar off-flavours in Cheddarcheese made from cold-stored milk.

Treatment and pH of the cheese milk

As early as 1930, Moir observed thatcheese made from pasteurised milk devel-ops a bitter flavour, a defect more pro-nounced for milk f1ash-pasteurised at85 oC than at 74 oC (20-30 s). As men-tioned by Stadhouders (1962), cheese-making from over-pasteurised milk is stillnot allowed in the Netherlands; indeed, bythis process more rennet is retained andthe cheese has a greater chance of be-coming bitter. This was confirmed for Gou-da and Cheddar cheeses produced frommilk of low pH (pH 6.25) (Stadhouders andHup, 1975). The pattern of proteolysis isaltered during maturation; consequently,there is more intensive breakdown of the<Xs1-caseinand increased bitterness in theCheddar cheese (Creamer et al, 1985).

The lowering of the initial pH of cheesemilk happens when: a), the milk is of inferi-or quality and acid formation has takenplace (Okigbo et al, 1985); b), the milk is

pre-ripened with starter; or cl, a largeamount of starter has been used. A solu-tion to this problem may be achieved by re-duction of rennet levels (Banks, 1988) andmanufacture of Cheddar cheese accordingto the method of Banks et al (1987), whichincludes incorporation of denatured wheyprotein into the curd. However, the devel-opment of high quality Cheddar cheeseflavour was impaired by this technique.

Bacteriophage proliferationin the cheese vat

The presence of antibiotics or bacterio-phages seriously inhibits the starter activityand prevents or reduces the developmentof bitterness (Stadhouders, 1974).

Bacteriophage contamination duringmaking, at levels which restricted startergrowth in the final stages of manufacturewithout markedly affecting acid production,had a striking effect on cheese flavour,especially the intensity of bitterness (Low-rie et al, 1972), and was responsible forlonger cheese-making times on subse-quent days (Lawrence and Gilles, 1973;Lowrie, 1977). The cheeses manufacturedunder controlled bacteriological conditionsboth with the normally bitter strain Laeto-coccus laetis subsp laetis MLa or the slowacid-producing strain Laetoeoeeus laetissubsp eremoris AM2 at a low cooking tem-perature developed an intense bitter fla-vour; however, this defect was almost en-tirely eliminated from cheese made in thepresence of bacteriophage (Lowrie et al,1974). Since significant bacteriophagecontamination may occur without apparenteffect on acid production during 'cheess-making, instances of such bacteriophageinfection must be widespread in commer-cial cheese-making operations, even whenso-called "phage-tolerant" or "phage-resistant" starters are used (Lowrie, 1977).

Bitter f1avour in dairy products: part 1 607

Many years aga, cheese of good fla-vour, as weil as complete control over bac-teriophage could be obtained by rotation ofcarefully selected pairs of starters(Lawrence and Gilles, 1973; Gavron et al,1982). The information that bacteriophagedoes not replicate in non-growing cellswhich still retain the capacity ta fermentlactose ta lactic acid would appear, as sug-gested by Hillier et al (1975), ta form a ba-sis for the control of phage proliferation incheese factories.

Fat content

Cheddar cheese made from pasteurisedmilk with a reduced fat content (1.5 ±0.1%) has been found ta develop an in-creasingly acid and bitter flavour (Deaneand Dolan, 1973). On the other hand, thisoff-flavour defect was also observed in Ital-ian cheeses such as Crescenza, Gorgon-zola and Robiola when they were manu-factured from milk with a high fat content(Delformo and Parpani, 1986).

Hydrolytic cleavage of fatty acids frommilk fat by the lipase, which is secreted bymicrobial contaminants in milk, can pro-duce bitter off-flavour (Sassette et al,1986), as discussed before.

Choice of starter

As observed by Lawrence and Gilles(1969), the most important raie in bitter-ness is played by the starter. Tests byHansen et al (1933) indicated that additionof more than 4% of a culture of Streptococ-cus paracitrovorus ta raw or pasteurisedcheese milk resulted in cheese with a bitterflavour and open texture. This was alsotrue of Steptococcus citrovorus when aninoculum of 1-10% was used, regardless

of the kind of milk (raw or pasteurised). Ad-dition of 0.1 to 0.5% of a culture of Strepto-coccus fiquefaciens ta the regular starterresulted in cheese with a flavour sa bitterthat it became unpalatable (Deane, 1951).Similar results were obtained by Yates etal (1955), who made Cheddar cheese withcultures of Streptococcus faecafis var fi-quefaciens previously isolated from rawmilk cheese or whey. Marked proteolysisand bitterness resulted in cheese preparedwith pasteurised cheese milk, added Strep-tococcus liquefaciens and a lactic starter(Tittsler et al, 1948). Outgrowth and survi-val characteristics of starter bacteria duringcheese ripening have been implicated inthe development of Cheddar cheeseflavour as weil as of bitter flavour (Perryand McGillivray, 1964; Reiter et al, 1967;Martley and Lawrence, 1972; Lowrie et al,1974).

For many years starter lactococci havebeen c1assified as bitter or non-bitter ac-cording ta whether or not they consistentlyproduce bitterness in Cheddar cheese(Emmons et al, 1962a; Lawrence andGilles, 1969). Two simple methods for de-tecting bitter strains have been proposedby Klimovsky et al (1970). The first methodconsists of evaluating the flavour of curdsof skim-milk and of skim-chalk-milk madewith rennet in concentration of 0.7 g/I ofmilk following incubation for 24 h and 7days at 26 oCwith the strains being tested.Selected strains were used in the manu-facture of Kostroma, a Russian cheese.There was a perfect correlation betweenbitterness in the cheese and bitterness inthe 7-day test. The second method isbased on the ability of lactic acid bacteriato form peptides rapidly and ta produceglutamate slowly in milk, following additionof rennet, knowing that the most bitterstrains consistently produce less glutamatethan non-bitter strains, independently ofthe milk or rennet used. Hillier et al (1975)


have shown that the estimation of bacterialgrowth in milk containing 1% yeast extract,at pH 6.3 and 37.5 -c, by a rapid absor-bance method, provides a simple meansof identifying strains likely to produce highor low cell populations in the curd andhence, either a bitter or non-bitter cheese.Certain strains of P roqueforti and P ease-ieolum have been shown to be responsiblefor the development of bitterness in Cam-embert cheese (Adda etet, 1982). Accord-ing to the findings of different workers(Martley and Lawrence, 1972; Sullivan andJago, 1972; Choisy et al, 1978) a fastmethod is proposed, based on the deter-mination of the starter growth rate, proteol-ysis and proline-iminopeptidase activity, tocharacterise laetoeoeeus strains for theircapacity, if any, to develop bitterness incheese.

Bitter and non-bitter strains are sup-posed to differ only in their resistance tothe cooking temperature applied in Ched-dar cheese production (=0 38 OC).At thischeese cooking temperature, since thegrowth of the non-bitter strains is re-pressed while that of the bitter strains isnot, a high final total population of starterorganisms in the curd, irrespective ofstrain, has been thus shown to be the ma-jor factor producing bitterness in cheese(Lowrie et al, 1972). The viability of startercultures in milk incubated at different tem-peratures has been used (Sullivan et al,1974) to indicate the potential of individualstrains to produce bitter or non-bittercheese at specifie cooking temperatures. Ithas been observed (Phillips, 1935) that for3 different cultures of Laetoeoeeus laetissubsp lectis, when the pH value of thecheese was s 5.00 at 4 days, bitter flavoursubsequently developed. As suggested byHarris and Hammer (1940), if a Mieroeoe-eus is to be employed in making Cheddarcheese from pasteurised milk, it should beselected on a strain rather than on a spe-

cies basis because of the differences be-tween cultures apparently belonging to thesame species. As reported by Klimovsky etal (1970), Laetoeoeeus laetis subsp laetisand Laetoeoeeus laetis subsp eremorisstrains may give rise to bitterness in Ched-dar cheese and also in Dutch-type chees-es. Martley and Lawrence (1972) havealso come to the conclusion that it is un-wise to regard ML1, a strain of Laetoeoe-eus eremoris, as a typical non-bitter starterbecause this strain gives slightly bitter andother characteristic off-flavours, variouslydescribed as "burnt" or "malty". Similarly,Lawrence et al (1972) stated that the strainof starter used appeared to be the only fac-tor controlling the acceptability of Cheddarcheese and the development of off-flavours such as bitterness when normalcommercial cheese-making conditionswere used. Their finding is not in full agree-ment with the conclusions of Reiter andSharpe (1971) from trials with cheesemade under carefully controlled conditionsin an aseptic vat, that the chemical compo-sition of the milk (which may be influencedby the feed of the cows and the effect ofthe native milk enzymes, ie lipases andproteases) and the bacteriological qualityof the cheese milk may also influence thedevelopment of cheese flavour.

The culture characteristics play a moredecisive role in bitter flavour productionthan the type of milk-c1otting enzyme(Nelson, 1974). Cheeses made with "slow-er" starter strains (ML1, Es, TR, R1, US3,

SK11, AM1 and AM2 of Laetoeoccus iectissubsp eremoris (Exterkate, 1976)) havebeen found to exhibit good flavour, quitefree from bitterness, while those made with"faster" starter strains (HP, C13, KH, FD27and Wg2 of Lectococcus tectis subsp erem-oris (Exterkate, 1976)) exhibited bitterness.The starter strains HP and Wg2 were foundto be capable of producing bitter cheeseson their own, without any interaction with

Bitterf1avourin dairyproducts:part 1 609

rennet (Visser, 1977b). When used as sup-plemental starters, strains of Streptococ-cus faeca/is were found unsuitable forCheddar cheese manufacture becausethey prodùced more defective f1avoursthan did their controls. However, 2 strainsof Streptococcusdurans (15-20 and 9-20)could be considered for supplemental useas commercial starters because they sig-nificantly and consistently produced chees-es with fewer defects than their controls,which were manufactured only with lacto-cocci (Jensen et al, 1975). As observed byLawrence et al (1976), Lactococcus lactissubsp cremoris strains are less likely thanLactococcus lactis subsp lactis strains togive fruity and bitter defects in Cheddarcheese. The central role of starter bacteriain the ripening of Gouda cheese has beenestablished by Kleter (1976), who foundthat it was possible to make Gouda cheesewith a normal ripening process and normalorganoleptic properties when no bacterialenzymes other than those from a properstarter Lactococcus lactis subsp cremoriswere active in the cheese. Lactococcuslactis subsp lactis C2 has the potential toproduce bitter cheese based onits growthin milk at the normal cooking temperaturesfor Cheddar (Sullivan et al, 1974; McKayand Baldwin, 1978). Improvement of Lacto-coccus lactis subsp lactis C2 as a potentialCheddar cheese starter, by genetic manip-ulation, reduced the proteolytic activity ofthe stabilised transductants as comparedto the parent. This resulted in a reductionin the formation of bitter flavour in Cheddarcheese (Kempler et al, 1979).

Lactic starters prepared from non-bitterstrains such as Lactococcus lactis subsplactis INIA 12, may be empJoyed to mini-mize the risk of bitterness in cheese varie-ties where fast acid production is not es-sential for the manufacturing process, ieSpanish Burgos cheese manufacturedfrom pasteurised milk, but without inocula-tion with lactic starters (Chavarri et al,

1988). However, as reported by Nufiez etal (1982) there is no specifie commercialstarter for the manufacture of Manchegocheese in Spain, and bitter flavour defectsdue to the use of imported starters havebeen frequently observed.

Lowrie et al (1972) have concluded thatculture strains which produce little or nobitterness and good cheese flavour exhibitone or more of the following characteris-tics: 1), a low rate of cell division at normalcheese cooking temperaturesz), poor sur-vival in cheese during curing; 3), low prote-olytic activity; 4), high acid phosphataseactivity (Dinesen, 1974; Nelson, 1974).

Behaviour of bitterand non-bitter strains

Lawrence and Gilles (1969) observed thatcheese was seldom bitter when non-bitterstrains were used as starter, regardless ofthe amount of rennet, salt or moisturepresent, or the rate of acid production inthe curd, or the final pH of the cheese.However, when bitter strains were used,the cheese was invariably bitter exceptwhen the salt-in-rnoisture levels and thepH of the cheese were relatively high.They also found (Lawrence and Gilles,1971) that the bitterness of cheese manu-factured with the bitter strains was directlyproportional to the amount of rennet used.This was in contrast to cheese manufac-tured with the non-bitter strains, since a 2-or 3-fold increase in the amount of rennetused did not produce a bitter cheese(Jago, 1974).

As mentioned by Exterkate andStadhouders (1971), bitter and non-bitterstrains of lactococci differ in their growthcharacteristics: bitter strains grow rapidlyunder normal cheese-making conditionsand reach high populations in the cheesecurd prior to salting; on the other hand, themultiplication of non-bitter strains is inhibit-

610

al, 1962a), strains that produced bittercheese hydrolysed bitter-tasting peptides •less extensively than strains that producednon-bitter cheese. Jago (1962) came to asimilar conclusion, ie that the differencesbetween bitter and non-bitter strains weredue to the inability of bitter strains to hydro-lyse the bitter peptides produced by ren-net. It was suggested that bitterness mayresult from the formation of pyrrolidonecarboxylic acid (PCA) at the N-terminalend of a hydrophobie peptide derived fromcasein during proteolysis. Later, Sullivanand Jago (1970a, b) c1aimedthat the pos-session of a pyrrolidone carboxylyl pepti-dase by non-bitter, but not by bitter start-ers, was the critical difference betweenstrains in their ability to degrade bitter pep-tides. This enzyme specifically hydrolysesthe peptide bond joining PCA residues tothe remainder of peptides and proteins.This was further correlated by the detailedinvestigation of Exterkate and Stadhouders(1971) who showed that the enzyme activi-ty was not only present, but was higher, incell extracts from bitter strains than fromnon-bitter strains.

When the amount of bitter peptidespresent in bitter as weil as in non-bittercheeses exceeds the threshold value forbitter taste perception, the cheese is ratedbitter (Visser et al, 1983b). Two reasonswhy the concentration of bitter peptidesmay never exceed the threshold at whichbitterness can be detected in cheese madewith non-bitter strains were reported byLawrence et al (1972): a), non-bitter start-ers would be expected to degrade highMW peptides at a slower rate since theyexhibit less proteolytic activity in cheesethan bitter starters (Martley and Lawrence,1972); and b), non-bitter strains may havea peptidase activity greater than that of bit-ter strains. The conclusion of Gordon andSpeck (1965) that a bitter strain was moreproteolytic than a non-bitter one in milk cul-


ed at the normal cooking temperature incheese-making. According to their results,ail strains must therefore be potentially ca-pable of contributing directly to the forma-tion of bitter-flavoured components incheese.

If the starter population density in thecheese at salting is allowed to become toohigh, however, flavour defects such as bit-terness or fruitiness are produced and de-tract from or mask cheese flavour (Law-rence et al, 1976). The role of starter maytherefore be merely to provide a suitableenvironment which allows the elaborationof cheese flavour (Lowrie et al, 1974).

As reported by Dunn and Lindsay(1985), even if many strains of lactococci(eg, MLa) produce bitter peptides in moreaged cheeses, they are preferred to non-bitter strains because of their efficiency inthe cheese-making process. Indeed, theselatter strains tend to be sensitive to saltand exhibit poor survival in cheese. lt isthen advisable, in order to avoid the devel-opment of off-flavours and bitterness, toselect the bacterial population of a starteron the basis of its proteolytic activity andits rate of lactic acid production over theentire range of temperature and pH occur-ring during manufacture and maturation ofcheese (Emmons et al, 1962a).

Bitterness in cheeseand degradation of bitter peptides

From their experiments, Emmons et al(1960a, b) concluded that bitterness inCheddar cheese is closely associated withthe strain of starter culture. Their resultssuggested that bitterness was due to a de-ficiency, in the strains used, of proteolyticenzymes capable of hydrolysing bitter pri-mary breakdown products of the cheeseprotein. According to the nitrogen contentof Cheddar cheese samples (Emmons et


ture was later asserted to be a non-generalfeature of ail bitter and non-bitter starterstrains (Martley and Lawrence, 1972). Alithe bitter strains are capable of producingbitter peptides from casein by the action oftheir specific cell wall proteinase(s) (Millsand Thomas, 1980; Exterkate, 1983). How-ever, slow variants of these strains le,those which give no off-flavours under nor-mal conditions (Lawrence and Pearce,1968) have lost this ability. Whole cells ofLactococcus lactis subsp cremoris HP canproduce bitter peptides only from wholeand J3-caseinsand degrade only chymosin-generated bitter peptides from O:sl- and(para)-K-casein (Visser et al, 1983c).

Reduction of bitterness in cheese

As shawn by Sullivan et al (1973), thepower of each strain to remove the bittertaste is directly related to its ability to hy-drolyse the bitter peptides to amino acids.A strain of Pseudomonas fluorescens,VITE 8.7, originally isolated from soil andable to produce a complex of endo- andexopeptidases, has been found to prevent,in the presence of a microbial rennet, theformation of bitter taste in Edam cheese(Mâlkkl et al, 1976, 1978). Indeed, thisstrain produces proteinases and peptidas-es able to convert, under laboratory condi-tions in culture flasks, 60-90% of the milkproteins to free amino acids; therefore, thebreakdown of bitter compounds will allow amodification of taste properties (Mâlkki etal, 1979). Knowing that bitter starters lackthe protease enzyme able to degrade bitterpeptides, Sood and Kosikowski (1979)suggested that addition, at low levels, ofmicrobial proteases such as fungal pro-tease 31000 from Aspergillus oryzae, mi-crobial protease P-53 of Bacillus subtilis,or lipase-MY from Candida cylindracea, tothe bitter starter may lead to cheeses freeof bitterness. According to Lawrence et al

(1976), the use of cultures containing slow-coagulating variants (Stadhouders, 1974)for Gouda cheese, and the use of strainsthat cannot grow at 38 oC (Lowrie andLawrence, 1972; Martley and Lawrence,1972; Jago, 1974) for Cheddar cheesewould seem to be logical approaches tothe prevention of bitterness. A reduction inthe maximum starter population or the de-liberate infection of the starter with homolo-gous phage was shown by Martley (1975b)to result in a marked reduction or elimina-tion of bitterness and in a great improve-ment in the overail flavour acceptability ofCamembert cheese. A better flavour, witha lower incidence of bitterness was alsonoticed for Gouda cheese manufacturedwith thawed single strain lactococci con-centrates inoculated directly to the cheesemilk (Martens, 1974).

The starter culture can be altered by in-clusion of a high level of mutant strains(Prr) with a substantial deficiency in sur-face-associated proteinase and with thesame peptidase activities as the parent cell(Kamaly and Marth, 1988). Because bitterflavours are related to proteinase positive(Prt-) cell activity, use of cultures contain-ing a high proportion of proteinase nega-tive (Prr) cells may represent a logical ap-proach for reducing bitternessdevelopment in cheese (Milis and Thomas,1980, 1982; Richardson et al, 1983). Anon-bitter Cheddar cheese has thus beenproduced using a high level (45-75% Prrcells), of variants of Lactococcus lactissubsp lactis (C2) Lac-Prr used to increasethe normal starter population of experimen-tal cheeses without increasing the rate ofacid production during the manufacturingprocess. The absence of bitterness in thefinal product has then indicated that the in-tracellular peptidase system of the addedmutant strain was efficient in degrading bit-ter peptides (Grieve and Dulley, 1983). Adecrease in bitterness in Cheddar was

6121


also obtained for cheese manufacturedwith starters containing Pit- variants ofLaetoeoeeus laetis subsp eremoris andLaetoeoeeus laetis subsp laetis (Monnet etal, 1986). Bitter cheese will no longer beproduced when an originally bitter strain inwhich the "slow" variants, or those whohave lost the ability of forming bitter pep-tides from casein, have grown to a largeproportion of the culture (Mills and Thom-as, 1980; Stadhouders et al, 1983) is usedas a starter. Exterkate (1976) suggestedthat the presence of proteolytic activity Pliwas correlated with bitter peptide produc-tion; non-bitter peptide producers such asLaetoeoeeus laetis subsp eremoris strainML1 produced only PI and PlU' It was laterdemonstrated by Hugenholtz et al (1984)that strain ML1 contains the same proteo-Iytic system as the bitter peptide producingstrains Wg2, HP and C13. As this strain ex-cretes most of its proteases into the sur-rounding medium, the proteases do not re-main in the curd and do not contribute tobitter peptide formation during cheese ri-pening.

Starter-pairing

It is now recognised that the use of certainstarters results in bitterness in cheese andthat pairing of "bitter" starters with "non-bitter" starters markedly reduces bitternessin Cheddar (Lawrence and Pearce, 1968;Creamer et al, 1970) and Gouda (Visser,1977b) cheesemaking.

Emmons et al (1961, 1962a, b) studiedbitterness in pasteurised milk cheese asinfluenced by different combinations andproportions of paired strains of Laetoeoe-eus laetis subsp eremoris in the starter.They found that the level of bitterness de-creased as the proportion of non-bitterstrains in the starter culture' increased.Non-bitter cheese was made by combining

a bitter and a non-bitter strain in suitableproportions as starter in the vat. In a simi-lar way, Creamer et al (1970) reported theundesirability of using "fast" single strainstarters (HP, P2, MLs), unless these arepaired with a "slow" starter (AM2) in orderto reduce the bitterness to an acceptablelevel; however, the bitterness of the startercannot be suppressed entirely by this pair-ing. The very significant decrease in bitter-ness observed when the mixed starter wasused can be totally explained as a dilutioneffect of the bitter starter by the non-bitterstarter. Experiments have c1early shownthat the proteolytic activity of bitter strainsis different from that of non-bitter strains(Stadhouders, 1974), and that the interac-tion between one strain and another maybe important (Reiter and Sharpe, 1971).The suitable pairing of such strains canthus contribute to considerably reduce bit-ter flavour in cheese.

Factors implicated in starter activityand in starter fai/ure

The temperature of storage has beenfound to be critical for starters. They remainactive for only a few weeks at -20 "O, butlose no activity at -40 "C when stored for6-8 months, a period which would be longenough for successful commercial use.Freezing in liquid N2 (-196 OC)is now themost common practice in dairying coun-tries. Even if storage of starter concentrateat -40 oC results in a slightly lower level ofcell survival compared to storage at -196 oC, the difference is unlikely to be sig-nificant when concentrates are used to in-oculate mother cultures and bulk startervessels, or even for direct vat inoculationin the case of Dutch-type cheeses (Law-rence et al, 1976).

The most important extrinsic factors im-plicated in starter failure are antibiotics and


bacteriophage attack. Since starter bacte-ria are sensitive to very low concentrationsof antibiotic residues, milk containing anti-biotics should not be accepted or used forthe manufacture of cheese. However,Gavron et a/ (1982) were able to manufac-ture non-bitter Gouda cheese in the pres-ence of antibiotics, and bitter cheese in theabsence of antibiotics by using group Nstreptococcus strains which had previouslybeen made resistant to penicillin and bac-teriophage, and which were stable with re-gard to both lactose metabolism and pro-teinase activity.

An alternative method of slow and con-tinuous acidification in,cheese-making bysubstituting o-gluconic acid)actone (GAL)for the starter organism wasdeveloped byMabbitt et a/ (1955). This technique _wasalso used (O'Keeffe et al, 1975) in Ched-dar cheese manufacture to simulate thepH development pattern of starter cheese.These latter workers observed a verymarked increase in proteolysis caused bya rapid early acid development in thecheese, and concluded that bitterness wasnot obviously connected with the level ofproteolysis; in fact the excessive proteoly-sis which accompanies rapid acidificationdoes not result in the development of bitterflavour. The contribution of rennet (or ren-net substitutes) to the ripening processmay be assessed by the use of asepticcheeses in which acidulation is simulatedby this technique (Visser, 1977a;Stadhouders et al, 1983).

Cheddar cheese manufactured undercontrolled bacteriological conditions usingGAL in place of starter was sharp or bitter(Perry and McGillivray, 1964; Reiter et al,1967), and Gouda cheese did not developany cheese taste or flavour at ail. Howev-er, after some months of ripening the latterbecame bitter (Kleter, 1976).

Amount of rennet and starter addedto the cheese milk

According to Stadhouders (1962), 3 mainfactors control the concentration of rennetin cheese: 1), the quantity of rennet usedin cheese-making; 2), the manner in whichthe curd is washed and dried (with shortdrying and washing times, more rennet r~-mains in the cheese); 3), the pH of the rnilkand curd during cheese-making. When thepH of the curd is much below the normalpH of milk (pH 6.5-6.6), the casein retainsmore rennet, which is then found in largerquantity in the cheese.

The rennet concentration used in Ched-dar cheese-making has an important influ-ence on the development of bitterness,especially when temperature-insensitivestrains are used (Scott, 1972; Milis andThomas, 1980). Lawrence et a/ (1972)have shown that with these strains (AM1,

AM2), increased rennet levels gave morebitter cheese. As found by Lawrence andGilles (1971), bitterness scores were di-rectly proportional to the level of rennetused with the fast lactococci starters (HP,MLa, z, BAl, Ea and ML1). However, ~oincrease in bitterness was observed wlth2- to 3-fold increases in rennet used withthe slow streptococci starters (AM1, AM2and US3).

Bitterness in Mozzarella cheese madeby direct acidification (Keller et al, 1974)and in a Japanese yeast-ripened cheese(Kaneko and Yoneda, 1974) may be avoid-ed by reducing the amount of rennet.

Since the relative clottinq power of ren-net is not constant during the normal dairy-ing season (Lawrence and Gilles, 1969,1971), the level of rennet used throughoutthe season could be varied according to itsclottinq power. This would save money by

614

ail cheese varieties (Lawrence et al, 1983).The rennet activity depends on its concen-tration in the cheese (Stadhouders, 1974)and to a great extent on the salt content(Stadhouders, 1962; Fox and Walley,1971). There is more opportunity for thechymosin to be active in the interior ofcheeses which are brined or rubbed withsalt than in cheeses of which the curd ismixed with salt before pressing.

Based on the results obtained withaseptically-made Gouda cheeses, Visser(1977b) concluded that rennet alone hasthe potency to produce bitter peptides and,if retained in high concentrations, to makethe cheese very bitter. lt was also ob-served that the contribution of rennet to thedevelopment of bitterness in Cheddarcheese is not as important as in Goudacheese (Stadhouders and Hup, 1975; Vis-ser, 1977b).

Milk ultra-filtration (UF) can be used incheese-making to produce a retentate withtotal solids at a suitable level; the retentateis thus a concentrate of protein and fat,and the permeate is a solution of lactoseand minerais (Mortensen, 1984). Creameret al (1987) manufactured Cheddar cheesefrom milk concentrated 5-fold by UF, andobserved that despite the relatively high re-sidual rennet concentration in the product(up to 33 RU/1 000 9 cheese), bitternesswas not detected. Rennet activity appearsto be somewhat inhibited in UF Cheddarcheese. Such inhibition is likely to be dueto whey proteins (Creamer et al, 1987).Moreover, incorporation of whey proteinsin the UF cheese-making process also hasan affect on the charactaristics and yield ofproduct and then limits the extension of therange of UF cheeses on the market (Le-lievre and Lawrence, 1988).

Mistry and Kosikowski (1986a, b) haveshown that starter concentration is impor-tant when retentates are used as bulkstarters; beyong 1% starter, higher yield in-

L Lemieux,RESimard

reducing the amount of rennet used, andwould also lower the incidence of bitter-ness in Cheddar cheese when certain"fast" starters are used. Furthermore, it isstill possible to reduce the level of rennetnormally used (70-85 9 per 450 kg of milk)when the setting temperature is raised by0.63-1.26 -c.

The rennet retention is Iinearly relatedto the amount of rennet added to the milk(Visser, 1977a). Retention is greatly in-creased with decreasing milk pH and withgreater starter activity (Stadhouders andHup, 1975; Lawrence et al, 1984). A studyby Vassal and Gripon (1984) on Camem-bert-type soft cheeses, made with Lacto-coccus lactis subsp cremoris and Penicilli-um caseicolum as surface flora, has ledthem to conclude that the pH of the milk atrenneting modified only slightly the quanti-ty of rennet retained in the curd, whereasthe quantity of rennet had a much greaterinfluence. In fact the amount of retainedrennet varied almost in a linear mannerbut did not affect the development of bitter-ness.

The significance of both the starterstrain and type of rennet used in the devel-opment of the bitter flavour defect in themanufacture of Cheddar cheese wasshown by Lawrence et al (1972): 1), thecalf rennet concentration determined boththe acceptability and the intensity of bitter-ness when the 'faster' strains of starterwere used; 2), at any given calf rennetconcentration the strain of starter used de-termined the flavour and acceptability ofthe resulting cheese.

Research in New Zealand has c1earlyestablished that if starter and non-startergrowth is controlled so as not to reach lev-els that produce off-flavours and if as littlerennet as possible is used, the flavourwhich develops in Cheddar cheese is like-Iy to be acceptable to most consumers. Itis probable that a similar situation exists in


creases were possible, but bitterness de-veloped in the cheese stored for 4 monthsat 10°C. In the manufacture of traditionalCheddar, the rates of acid developmentand moisture expulsion are critical. There-fore, the amount of normal starter cultureadded to the cheese milk cannot be greatlyincreased without affecting the quality ofthe final product. Lowrie et al (1974) haveshown that a high viable starter populationin curd leads to the development of bitter-ness. However, Grieve and Dulley (1983)have demonstrated that high starter levelsdo not necessarily produce bitterness and,in their experiments, have even producedCheddar cheese with superior flavourscores. In order to increase the normalstarter population of experimental cheesewithout increasing the rate of acid produc-tion during the manufacturing process,they used frozen mutant concentrates of aLaetoeoeeus laetis subsp laetis C2 Lac-Ptt-.

Calf-rennet substitutes

Supply shortages and variations in thequality of rennet from young calvesspurred efforts to discover suitable rennetsubstitutes (Dinesen, 1974). Since most ofthe proteinases will coagulate milk undersuitable conditions, many proteolytic en-zyme preparations of animal, plant and mi-crobial origin have been assessed as re-placements for calf rennet in themanufacture of many standard types ofcheese. Very substantial numbers of peo-ple in the world are opposed to the use ofcertain animal secretions in cheese pro-duction on grounds of religion, morality, ordiet. Ali these factors have contributed tothe worldwide search for animal rennetsubstitutes of microbial or plant origin. En-zymes from the fig (ficin), pineapple (brom-elain), papaya (papain), fungi (Endothia

parasitiea, and Mueor miehei, Mueor pusil-lus) or bacteria (Bacillus subtilis, B eereus)have been and are used as coagulants inthe cheese process. Unfortunately, manyof the rennet substitutes have undesirableside-effects, eg, bitter flavour. Any substi-tute for calf rennet must not only coagulatemilk, but must have low proteolytic activityand produce cheese with acceptable fla-vour and rheological characteristics(Quarne et al, 1968; Edwards and Kosi-kowski, 1970; Sardinas, 1972; Scott, 1972;Hofi et al, 1976; Rao et al, 1979).

Calf rennet, which contains chymosin(rennin) as the main enzyme component,is weil known as the best milk coagulant.While in the Netherlands no type of rennetother than calf rennet is allowed forcheese-making without special exemption(Visser, 1981), the use of substitutes fromanimal, plant or microbial origin has be-come a more general practice in severalcountries.

Animal origin

Cheddar cheese made using only porcinepepsin tends to have a bitter flavour whenfully ripe (Scott, 1972). However, the useof porcine pepsin in rennet mixtures maybe of advantage because it contributesstrongly to c1otting,while it is largely inacti-vated under the subsequent processingconditions. Chymosin is more specifie thanpepsin, and therefore exhibits a more limit-ed proteolytic behaviour (Visser, 1981).Hence, in admixture with rennet, porcinepepsin has a Iimited proteolytic action dur-ing cheese ripening; this avoids the devel-opment of flavour defects such as bittertaste.

ln a small number of studies, use ofcommercial and pure chicken pepsin prep-arations in the manufacture of Cheddarcheese did not affect the cheese-making

616

uct. The bitter taste, often associated withincreased proteolysis, may result from adifferent pattern of casein breakdown andan increased breakdown of whey proteins(Schulz and Thomasow, 1970).

The criteria that a microbial rennet prep-aration must meet in order to substitutesatisfactorily for animal rennet are the fol-lowing: 1), the preparation must effectivelycoagulate milk without excessively hydro-Iysing the resulting curd during or at matu-ration; 2), it must be non-toxic and devoidof antibiotic activity, as weil as free ofpathogens; 3), preferably, it should bereadily water-soluble and possess accept-able colour and odour; 4), it should exhibitreasonable shelf-life (Sardinas, 1972).

Without giving any details, Windlan andKosikowski (1956) reported that when ren-net-like enzymes from microbial sourceswere used at optimum concentration,smooth curds resulted. Bitter flavours ineither curd or Cheddar cheese were rela-tively insignificant.

Edam, Gouda and Cheddar type chees-es made with microbial rennets obtainedfrom Mucor pusillus Undt and Bacillus pol-ymyxa were bitter. Kikuchi and Toyoda(1970) attributed this off-flavour to the in-herent characteristic of the microbial milkclottinq enzymes and concluded that mi-crobial rennets could replace the conven-tional calf rennet in cheese-making if as-pects of manufacture were modified to suitmicrobial rennets.

Lawrence et al (1972) found that bothRennilase and Meito microbial rennets, ob-tained from Mucor miehei and Mucor pusil-lus var Lindt, respectively, produced con-siderably less bitterness in Cheddarcheese manufactured with a bitter strain(Lactococcus lactis subsp cremoris strainHP) than did calf rennet. However, not ailmicrobial rennets were alike in this re-spect, since a rennet obtained from En-

L Lemieux,RESimard

conditions, yield or cheese composition.However, the final product underwent fast-er proteolysis, more intense flavour, off-flavours and bitterness and was softerthan cheese made with calf rennet (Stan-ley et al, 1980; Green et al, 1984).

Plant origin

As early as 1956, Windlan and Kosikowskiused papain for Cheddar cheese manufac-ture. Unfortunately, bitterness was severeenough to make the cheese unfit for con-sumption.

A vegetable rennet extracted from With-ania coagulans berries has been used toprepare Cheddar cheese from cow andbuffalo milks. The former cheese wasslightly bitter while that made from buffalomilk was inferior in quality and had a bitterflavour (Singh et al, 1973).

A coagulant obtained from the flowersof Cynara cardunculus Land used tradi-tionally in Portugal for the manufacture ofewe cheese of Serra and Serpa types, hasbeen assessed in the manufacture ofCamembert and Gruyère cheeses. Unfor-tunately, these cheeses developed verypronounced bitterness (Barbosa et al,1976).

Microbial origin

Microorganisms represent a large, invitingreservoir of potential animal rennet substi-tutes. However, the majority of microbialrennets are much too proteolytic forcheese-making. Replacement of chymosinby microbial rennets influences cheese ri-pening: proteolysis of <Xs1- and 13-caseinsis increased and retarded, respectively, asobserved by Christensen et al (1989) whosuspected the type of rennet used in thecheese manufacture to have an effect onbitter flavour development in the final prod-

1

1

1

1

1

1

1

1

--- '-------------------------:-----------------

Bitterf1avourin dairyproducts:part 1 617

dothia parasitica produced bitter cheese BacterL rennets from Endothia parasiti-even with a non-bitter starter (Lactococcus ca, Muco" miehei, and Mucor pusillus havelactis subsp cremoris strain AM2). On the proved sufficiently suitable for large-scaleother hand, a "somewhat bitter" flavour ap- comrnerclalisation, even if bitterness doespeared more frequently in Gouda made appear particularly in some long-hoIdwith the microbial rennets, Rennilase and cheeses. Of these rennet substitutes, theFromase, than withcalf rennet (Martens, first 2 had a greater proteolytic action on ~-1973). casein than calf rennet (Edwards and Kosi-

Microbial rennet substitutes such as kowski, 1970). Knowing that a coagulantwhich doés not excessively hydrolyse ~-Meito (from Mucor pusillus), Suparen (from 1

Endothia parasitica) and Milcozyme (from casein is suitable for the production ofcheeses of good quality without bitterness,

Bacillus polymyxa) were also used in com- Kobayashi et al (1985) evaluated the milkparison with rennet for the manufacture of clottinq enzyme from Irpex lacteus (IR) asCheddar-type, Tilsit (a Polish semi-hard a calf renriet substitute in Cheddar cheese-cheese) and Kortowski cheeses. Cheeses making. f-Ithough IR cheese showed amade with Suparen and Meito were slightly slightly higher extent of proteolysis in com-bitter, while Milcozyme was found to be un- parison toil' the calf rennet control during ri-suitable for cheese manufacture (Reps et pening, it did not develop a bitter tasteal, 1974). even afterl6 months of ripening. IR is thus

Bitter flavour developed during aging of a promising rennet substitute for cheese-Cheddar cheese manufactured with milk making. 1

clottinq enzyme of Rhizopus oligosporus The use of mixtures of equal parts of(Rao et al, 1979) a recombinant chymosin calf rennet and microbial rennet (Mucorpurified from Escherichia coli K12 (Hicks et miehei protéase) in Ras cheese-makingal, 1984) or as a rennet substitute. Howev- caused bitterness in this Egyptian harder, further cheesemaking studies conduct- type cheese which persisted until theed with dilute enzyme preparations yielded fourth rnonth and disappeared thereafterto a non-bitter Cheddar cheese which did (Mahran et al, 1976).not retain excessive amounts of protease Use ofl peptidases (eg, Pseudomonas(chymosin) (Hicks et al, 1988). Chymosin peptidase) capable of hydrolysing bitterprepared with the aid of genetically engi- peptides ~as been suggested as a possi-neered microorganisms such as E coli, As- ble mechanisrn for preventing the occur-pergillus niger and Kluyveromyces lactis rence of bltterness, thus enabling the widerhas been shown to be identical to the natu- use of rnicrobial rennets. This option hasrai enzyme and to produce different types been triedlfor Edam cheese manufacturedof cheeses (Cheddar, Colby, Edam, Gou- with a commercial rennet from Mucor mie-da, etc) which were indistinguishable from hei; cheeses were not bitter and proteoly-normal cheese (made with natural chymo- sis was ~ore rapid in the experimentalsin) in regard of cheese yield, texture, than in the control cheese (Mâlkkl et al,smell, taste and ripening (Teuber, 1990). 1978). 1

Since bitterness is more significant in long- Addition of fungal acid proteinases withhoId cheese varieties such as Cheddar, pH optima' close to the pH of cheese hasthen many potential rennet substitutes been shown to cause excessive proteoly-which render cheese bitter will have to be sis, leading to bitterness (Law and Wig-rejected. more, 1982b).

618

ness increased with storage time and en-zyme level (from 0.01-0.1 % of curdweight) (Fedrick et al, 1986). These au-thors also found that Colby cheese pre-pared with a very low amount of RhozymePli developed mature cheese texture with-out bitterness. Cheddar cheese exhibitedmoderate proteolysis with extreme bitter-ness when it was treated with Subtilisin A,the Bacillus Iicheniformis alkaline protei-nase, to accelerate ripening (Law and Wig-more, 1982a).

Bitterness was absent when peptidasesfrom Pseudomonas f1uorescens VTTE 8.7were added to Edam-type cheese in con-junction with a commercial Mucor miehei-type rennet. The same enzyme prepara-tion from Pseudomonas prevented bitter-ness and enhanced the ripening processof Cheddar type cheeses made with calfrennet (Mâlkki, 1978; Mâlkki et al, 1978,1979). A pronounced bitter flavour devel-oped in Ras cheese (an Egyptian hardcheese) when Maxazyme, a proteolyticand Iipolytic enzyme preparation, was add-ed to the curd before moulding to increasethe rate of ripening (El Soda et al, 1985).

Addition of germinated barley extract,Neutrase (an extracellular Bacillus subtilisneutral proteinase) or a combination ofboth enzyme preparations to the Cheddarcheese curd yielded, in ail cases, a bitterflavour defect. Although Frey et al (1986)found this off-flaveur defect to be more im-portant in the cheese made with Neutrasealone, some workers (Law and Wigmore,1982b; Fedrick et al, 1986) observed thatoptimum acceleration of Cheddar cheeseripening using Rhozyme Pli a neutral fun-gal protease, seemed to be accompaniedby higher levels- of bitterness than is thecase with Neutrase. Development of a bit-ter taste was also observed during the ri-pening of a Swedish hard type cheesetreated with Neutrase to accelerate thecasein breakdbwn. It was shown that the


Addition of proteolytic enzymesor heat-treated cells to the cheese milkor cheese slurryto accelerate cheese ripening

The influence of commercial mixtures ofproteolytic and lipolytic microbial and ani-mai enzyme preparations added directly tocheese curd and cheese slurries to accel-erate cheese ripening has been studied bymany workers (Fox, 1988-1989). Whenmicrobial acid proteases were added toCheddar cheese slurries, it led to strongbitterness; however, blending of individuallipases with proteases or peptidases be-fore addition to cheese blends or slurriescreated pronounced cheese flavour withless bitterness and rancidity (Kosikowskiand Iwasaki, 1975). Similarly, the incorpo-ration of fungal protease 31000 from As-pergillus oryzae, microbial protease P-53of Bacillus subtilis; or MY-lipase from Can-dida cylindracea into slurries of saltedCheddar curds did not cause the develop-ment of flavour defects (Sood and Kosi-kowski, 1979); it seems that as soon asbitter peptides were formed they were hy-drolysed into still smaller peptides and freeamino acids.

Different results were obtained whenthe same food-grade enzyme preparationswere added to Cheddar cheese curds aspowders mixed with the salt (Law and Wig-more, 1982c; Fox, 1985). Acid proteinasesinvariably yielded bitter cheeses with anunacceptable soft, crumbly and greasytexture (Law and Wigmore, 1982a). Neu-tral proteinase when added at only 0.001-0.002% (w/w) produced medium-flavouredcheese at 12 "O in 2 instead of 4 months;addition of neutral proteinase at higherconcentration resulted in bitter cheese.When Rhozyme Pli, a neutral proteasefrom Aspergillus oryzae, was added to themilled curd during the manufacture ofCheddar cheese, proteolysis and bitter-


simultaneous addition of heat-treated(67 oC for 10 s) cells of Lactobacillus hel-veticus could eliminate the bitter taste byaccelerating the breakdown of peptides inthe Swedish full-fat (Ardô and Pettersson,1988) or low-fat (Ardô et al, 1989) hardcheeses. Addition of Rulactine (a metallo-protease from Micrococcus caseolyticuswith properties close to those of Neutrase)to Saint-Paulin cheese milk also promotedearly proteolysis and enhanced bitternessdevelopment (Alkhalaf et al, 1987). Howev-er, encapsulation of Neutrase into multila-mellar liposomes partially avoided bitter-ness. Indeed, encapsulation of proteinaseinto liposomes has prevented further low-ering of pH at draining by slowing downthe proteolysis process which was exces-sive for the free proteinase, and thus pro-duced less peptides used as substrates formicrobial proteinases and peptidases toyield small peptides and amine acids (Law-rence et al, 1983; Fox, 1988). Further-more, the enzymes retained in the curdhave been progressively released in thecheese matrix after pressing. This slow butgradually accelerated proteolysis has thenhelped to prevent the accumulation of bit-ter peptides in the cheese. From these re-sults and those obtained by Law and Wig-more (1983) on the action ofexopeptidases, Alkhalaf et al (1988) cameto the conclusion that entrapment of bothproteinase and exopeptidase within re-verse-phase evaporation vesicle lipo-somes could eliminate the potential defectof bitterness. The presence of exopepti-dase could thus enhance the degradationof bitter peptides which released low mo-lecular weight peptides and amino acids.

ln Gouda cheese, Bartels et al (1985a,b) found that a strain of Lactobacillus hel-veticus, when heat-treated at 70 oC for18 s or freeze-shocked (Kim et al, 1987),had a similar debittering property whenadded as a starter adjunct (2%) to milk.

The samel

lphenomenon, but to a lesser ex-

tent, was observed when heat-shockedwhole cells of Laetobaeillus bulgarieuswere added, The reduction in bitterness in-tensity observee in severai trials by Bartelset al (1985a, b; 1987a), and which appar-ently resulted from peptidase activity of theadded cel'rs, disagrees with the observa-tions of El Soda et al (1982). Indeed, thelatter workers observed that incorporationof cell-free extracts from Laetobaeillus hel-vetieus, Lectobecëlue bulgarieus or Laeto-baeillus I~etis subsp laetis into Cheddarcheese cJrd enhanced bitterness. In thestudy of Bkrtels et al (1987a), reduction ofbitterness :in cheese by certain lactobacilliwas related to lower concentrations of bit-ter peptides, separated by gel permeationchromatography. In contrast, bitternesswas not reduced by addition of heat-shocked Streptoeoeeus thermophilus 110or Laetob~eillus bulgaricus subsp jugurtiATCC 122'78. According to El Abboudi etal (1991), àddition of heat-shocked cells oflactobacilli Ito the cheese milk before ren-neting could contribute to the reduction ofbitterness in accelerated ripening Cheddarcheese. 1

Accelerated ripening and reduced bitter-ness were hoticed in Cheddar cheese pre-pared with: Lac- mutant strains, isolatedfrom Laetoeoeeus laetis subsp laetis C2,added as ~ starter adjunct to milk (Dulleyet al, 1978). However, when a cell-free ex-tract from Laetobaeillus easei was addedto the curel during the manufacture of aCheddar-type cheese, the ripening was ac-celerated but, unfortunately, bitter flavourdeveloped ~apidly(El Soda et al, 1981).

1

Before the study of Dulley (1976), at-tempts to accelerate cheese ripening werebased mainly on the premise that proteoly-sis must b1eincreased. However, Dulley(1976) observed that extra proteolysis maynot be a necessary prerequisite for flavourdevelopment, Knowing that the products of

1

620

be due to increased casein and cell densi-ty in the curd.


proteolysis are normally present in excess,he thought about converting them into fla-vour components by adding enzymes tothe cheese in arder to accelerate the ri-pening process. For that purpose Cheddarcheese slurries, ripened for about 1 weekat 30 oC, were added to the cheese curd.Used as a source of enzymes and micro-organisms, the Cheddar cheese slurrieswere shown to shorten the ripening time ofCheddar cheese by about 1,5 month with-out the development of excessive bitter-ness.

Use of commercial starter lactococci inconjunction with Laetobaei/lus strains iso-lated from Cheddar cheese shortened thecheese ripening process (Lee et al, 1990b)and also led to an improvement of Ched-dar flavour (Lemieux et al, 1989). Howev-er, certain strains of Laetobaeillus easei-pseudoplantarum and Laetobaei/lus cesei-rhamnosus caused bitterness in thecheese (Lee et al, 1990a).

The manner of draining the curd

Extension of the contact time between thecurd and the whey reduces the curd's buf-fering capacity by increasing its lactose re-tention and its loss of phosphorus. Ched-dar cheese made from such curd becomeshighly acid at maturity and this leads to thedevelopment of bitterness (Czulak et al,1969).

The production of bitter components byincubation of a bitter and a non-bitterstrain of Lactoeoeeus laetis subsp eremor-is in skim milk and in various media, withand without casein, was studied by Har-walkar and Seitz (1971). They observedthat when curd was separated from wheyafter 18 h ripening, further incubation re-sulted in considerably greater bitterness incurd than in whole ripened milk; this could

Cheese manufacturing process

As early as 1932, Riddet et al concludedthat it was possible to modify the cheeseflavour simply by changing sorne details inthe manufacturing process. Such changesproduce curd of a different nature, andsince the curd is the raw material of the ri-pening process, the whole course of the bi-ochemical changes which subsequentlyoccurs may be modified. A typical exampleof a change in the manufacturing processwhich induces cheese modifications is avariation in acidity at any stage of the pro-cess. Type of acid and pH at curd forma-tion affect moisture, minerai content andfirmness of Mozzarella cheese made bythe direct acidification procedure. In thisprocess, there is no addition of lactic start-er cultures, the pH of milk and curd is ad-justed by various acidulants (hydrochloric,phosphoric, or lactic acids) to a valuewhich remains constant during whey syn-eresis (Keller et al, 1974). It was also re-ported by Quarne et al (1968) that Pizzacheese made by direct acidification usingcalf rennet developed bitterness.

Stadhouders and Hup (1975) haveshown that the cumulative formation of bit-ter peptides was due either to the use ofcertain starters or to sorne method ofcheese-making which resulted in too higha level of rennet retained in the cheese(the lower the pH and the lower the cook-ing temperature, the higher the level ofrennet retained in thecheese (Holmes etal, 1977; Stadhouders et al, 1983; Cream-er et al, 1985)).

Many workers have observed that themanufacturing conditions during cheese-making markedly affect cheese flavour,


and more specifically undesirable defectssuch as bitterness (Emmons et al, 1960a,Lawrence and Gilles, 1969; Delbeke andNaudts, 1970; Lowrie et al, 1972; Jago,1974; O'Keeffe et al, 1975; Visser et al,1983b; Berdagué and Grappin, 1988).

Cooking temperature

The final concentration of starter bacteriain the cheese and salt-in-molsture levelsare known to be altered by the cheese-making cooking temperature (Crawford,1977). Moreover, the rate of cell divisionduring cooking has been found to be a de-cisive factor in bitter flavour production(Dinesen, 1974; Nelson, 1974).

Barton (1957) found that the entire pro-duction of the culture organisms used asstarters in the manufacture of Cheddarcheese from pasteurised milk is governedby the rate of cooking and the final temper-ature reached. Although the classificationof culture strains as "bitter" or "non-bitter"is arbitrary, it is more specifically related tocooking temperature used during cheese-making (Lowrie and Lawrence, 1972). Inail the experiments conducted by Lowrie etal (1972), Cheddar cheese was bitter whenthe standard cooking temperature, re-duced from 37.8 to 33.3 "C, permitted ei-ther bitter or non-bitter strains to grow to ahigh population in the curd. Conversely,when starter growth was limited, bitternesswas absent or of reduced intensity in thecheese. These observations led to the con-clusion that starter bacteria had a directrole in the development of bitterness inCheddar cheese. Later, Jago (1974) sug-gested that the formation of bitter peptidescould proceed at a much faster rate thantheir degradation to non-bitter products.According to Nelson (1974), slower strainsexhibited higher rates of cell division and

yielded a bitter cheese (Lowrie et al, 1974)when used in experimental cheese manu-facture in which the cooking temperaturewas lower than normal. On the other hand,a lower rate of cell division was recordedfor faster growing strains used at a highercooking temperature than normal; the re-sulting cheese exhibited low levels of bitterflavour.

The cooking temperature used duringthe manufacture of Gouda cheese wassuspected by Stadhouders (1974) to con-trol the amount of rennet retained in thecurd and thus to influence the develop-ment of bitterness in the cheese. At cook-ing temperatures above 35 -c (35-39 OC)less rennet was found in Gouda cheese,which then had a reduced bitter flavour(Stadhouders and Hup, 1975). However,according to Martley (1975a) Stadhouder'ssuggestion does not seem to apply toCamembert cheese.

Ripening temperature

As reported by Law et al (1979), matura-tion temperature was the most importantfactor in determining the flavour intensity ofCheddar cheese made with Lactococcuslactis subsp cremoris NCDO 924 or 1986,either in enclosed (excluding non-starterflora) or open vats. Cheeses ripened at13 oC for 6 months developed a strongerflavour with less bitterness than those ri-pened at 6 oC for 9 rnonths, irrespective ofthe starter or vat used.

ln contrast, Gouda cheese had a morebitter flaveur when ripened at 16 than at6 oC (Stadhouders and Hup, 1975). Nor-mal ripening temperatures for Goudacheese being 12-13 oC, the risk of bitter-ness occurring was found to be greater attemperatures of 15.5 and 18 oC (Stad-houders et al, 1983).


Acidity or pH of cheese

Excessive acid production by the culturesused as starters is associated with bitter-ness (Scott, 1986). Phillips (1935) pro-posed to establish a relation between acid-ity and the development of bitter flavour atthe different stages of Cheddar cheeseprocessing. His preliminary resultsseemed to associate an increased bitter-ness in cheeses which develop aciditygreater than pH 5.00 at 4 days after mak-ing. Besides, Czulak (1959) found bitterflavour most frequently in Cheddar cheesehaving a pH value of less than 5.0 in thefirst week. Dawson and Feagan (1960)showed a definite peak of bitterness withina pH range of 5.2-5.3, decreasing atgreater and lower pH values. Similarly,Emmons et al (1962a) confirmed a depen-dence of bitter f1avouron cheese pH andfound it to be similar to the effect of pH onprotein breakdown which leads to the for-mation of a large pool of polypeptides, in-c1uding the bitter-tasting ones. They alsoshowed that some starter strains whichproduced bitterness were not greatly at-fected by pH while others were very signifi-cantlyaffected.

According to Czulak et al (1969), whena high lactic acid level is attained rapidly,the curd does not cheddar weil. Thus,cheeses from "fast-acid" vats have a lowpH, a sour and bitter flavour, a crumblybody and pale colour. However, Lawrenceand Gilles (1969) concluded that a highrate of acid production per se does not re-suit in bitter cheese. From their results it isc1earthat the pH of a cheese at 14 days isnot in itself very significant with respect topotential bitterness, it is of great impor-tance only when the "salt-in-moisture"levels lie between 4.30% and 4.90%.

The results of Sullivan et al (1973) sug-gested that pH is a determining factor inthe removal of bitterness by individual

starter streptococci at the pH of ripeningCheddar cheese (= 5.0). Similarly, Nelson(1974) reported that, in Canadian Cheddarmanufactured with single strains of Lacto-coccus lactis subsp cremoris, bitter flavourwas greater in low (pH 5.63) than in highpH (pH 6.41) cheese.

Salt concentration

A close relationship between a low saltcontent of Cheddar cheese and a charac-teristic bitter flavour was mentioned as ear-Iy as 1940 by Tuckey and Ruehe. Howev-er, more recently Lawrence and Gilles(1969), in agreement with the results ofStadhouders (1962), observed a decreasein bitterness in Cheddar at salt levelsgreater than 4.90%, and more retention ofthe added salt in curd salted at a low acidi-ty than at higher acidity. These observa-tions were confirmed by Phelan et al(1973) and by Golding (1947) for Cheddarcheese prepared under conditions of highmoisture content at salt concentrations< 1.6%. These conditions led to extensiveproteolysis, causing bitter and putrid fla-vours. A slight bitterness in f1avourof Kas-kaval, an Egyptian cheese, was also rec-ognised by Moneib and Safwat (1972) fora salt concentration of < 4%. However, thisdefect was not present when the salt levelwas between 4-5%. In the same way, theresults of Keller et al (1974) suggested thatdevelopment of bitterness could be inhibit-ed by increasing the salt content or by de-veloping a mechanism for infusing the cen-tre of Mozzarella cheese blocks with NaCI.As observed by many workers (Stadhoud-ers and Hup, 1975; Stadhouders et al,1983; Visser et al, 1983a, b), salt also re-duces the intensity of the bitter flavour ofGouda cheese; indeed, the higher the saltcontent, the lower the score for bitter fla-vour.

Bitterflaveurindairypreducts:part 1 623

Being able to retard the appearance ofbitter flavour in cheese and to influence theformation/degradation of bitter peptides(Stadhouders et al, 1983; Visser et al,1983c), salt must have a direct impact onthe proteolytic activity of the cell wall andof the intracellular enzymes (Visser et al,1983a). From the studies of Stadhouderset al (1983) on Gouda cheese, it was con-cluded that salt inhibits the formation of bit-ter peptides rather than masking the bitterflavour.

Since growth of lactococci was marked-Iy inhibited at salt-in-moisture levels >4.90%, Lawrence and Gilles (1969) sus-pected the effect of salt to be more specifieon starter, and showed that high salt-in-moisture levels reduce the incidence of bit-terness in cheese manufactured with a bit-ter starter. This was due to the inhibition,by NaCI, of the degradation of l3-caseinbythe rennet and bacterial proteinases (Foxand Walley, 1971; Sullivan and Jago,1972; Thomas and Pearce, 1981). Exterk-ate (1983) confirmed the essential roleplayed by salt in the degradation of bitterpeptides by the protoplast enzymes of thestarter bacteria. Salt may act on the cellwall and membrane structures, in particu-lar those of bitter strains (thus impairingthe accessibility of starter enzymes),orhave a direct inhibitory effect On variousproteolytic enzymes (Lawrence and Gilles,1969; Visser et al, 1983c). Salt-in-moisturelevels played also a direct role in the devel-opment of bitterness in sterile buffaloesskim milk by altering the final concentrationof S faecalis subsp liquefaciens. Growth ofthis organism was thus shown to be re-stricted at salt levels of 8% (Hegazi, 1989).

According to Stadhouders (1962), ren-net activity is optimal at pH 5.2 and at asalt concentration of '" 3%. However, thisactivity decreases sharply as the salt con-centration is increased to 5%. Since thesolubility of paracasein is increased by the

presence of salt, the substrate is thenmore readily available to the enzyme, andthis may be responsible for the stimulatingeffect of salt on the rennet activity(Stadhouders, 1962). Fox and Walley(1971) have shown that the proteolysis ofl3-caseinby rennet or by pepsin was signifi-cantly reduced by 5%, and completely in-hibited by 10% NaCI. Increments in saltconcentration from 1 to 5% were alsoshown to inhibit the breakdown of l3-caseinby Lactococcus lactis subsp cremoris bitterstrain HP (Sullivan and Jago, 1972). Incontrast, the rate of proteolysis of (XS1-

casein was found to be maximal in thepresence of 5-10% NaCI (Fox and Walley,1971). The choice of brining method great-Iy determines the initial salt concentrationin the interior of the cheese and may there-fore influence the protein degradation pat-tern during the ripening process (Visser.1981).

Mineral content

Mortensen (1984) reports that ultrafiltered(UF) Quark (a non-ripened cheese) pro-duction was initially hampered by bitter-ness problerns, which were due to the highminerai content in the retentate. Bitternesshas also resulted from the presence of ex-cess minerais in UF Cast Feta, a ripenedcheese variety in which factors other thanproteolysis are important for productcharacteristics. However, the high salt con-tent and the addition of lipase could effec-tively mask the bitterness defect (Lelievreand Lawrence, 1988).

Surface moulded soft cheeses(Camembert type)

As long as its growth is not Iimited, Penicil-lium caseicolum appears to be the princi-

624

by the pasteurisation of the skim milk andCottage cheese dressing (Stone andLarge, 1968).

When 20% abnormal milk (from cowssuffering from mastitis) was added to thecheese milk, the resulting cheese was bit-ter, having an unclean taste and smell anda short consistency (Sorokina and Kara-gueuz, 1978).

It has also been reported that poor milk-collecting and transformation hygiene, con-taminated feeds or antibiotics injected intothe cow might result in bitterness incheese; thus, Gorgonzola was found to bebitter when manufactured with milk con-taminated by penicillin (Delformo and Par-pani, 1986).

Packaging conditions may also be re-sponsible for the presence of bitterness indairy products. In fact, Kosikowski andBrown (1973) have shown that gas pack-aging (carbon dioxide and nitrogen) wassuccessful in preventing yeast and mouldgrowth on creamed Cottage cheese andthus in reducing the bitter off-flavour.


pal factor responsible for bitterness of softbody 'cheese such as Camembert. Thereare 2 ways to control the growth of this or-ganism: 1), incubation of cheeses in aslightly ammoniacal atmosphere duringthe first days of ripening leads to a fasterincrease of the pH in the outer layer; and2), some strains of Geotrichum candidum,when present in cheese milk (0.25% of theinoculum), have similar effects in reducingPenicillium growth, acid proteinase actionand bitterness (Mourgues et al, 1983; Ri-badeau-Dumas, 1984; Vassal and Gripon,1984).

A study of the neutral volatile com-pounds in 5 different types of surface ri-pened cheese (Maroilles, Livarot, Pont-L'Évêque, Époisses and Langres) hasshown the cheese to be bitter when indoleproduced from the degradation of trypto-phan by the bacteria was present at a con-centration of 10-3 ppm (Dumont et al,1974).

Sanitary and hygienicpackaging conditions

Bitterness in cheese during the maturationprocess has been reported to result fromoverheating of milk during pasteurisation(Riddet et al, 1932). However, since the in-troduction of rigid hygiene and pasteurisa-tion standards, the incidence of the bitterdefect has still been widespread. Bitter-ness cannot therefore be explained entire-Iy by the sanitary and hygiene conditions(Czulak, 1959).

As mentioned above, the developmentof bitterness in creamed Cottage cheesehas been found to be primarily due togrowth at refrigeration temperatures ofpsychrophilic bacteria. However, the de-struction of these microorganisms and theelimination of bitterness can be achieved

CONTROL OF CHEESE BITTERNESS

The concentration of bitter peptides mustexceed a certain threshold level for the bit-terness to be detected in cheese. Howev-er, this threshold level has been found tobe higher in old than in young cheeses.Control of bitterness in cheese, therefore,hinges on keeping the concentration of bit-ter peptides below the threshold level forbitter taste, either by decreasing the rate offormation of the bitter peptides, and/or byincreasing the rate of their degradation tonon-bitter products (Jago,1974).

ln practice, the control of bitterness incheese can be achieved by a combinationof methods which include: 1), limitation ofthe starter cell population; 2), maintenance


of high-salt-in moisture levels; 3)" starter'pairing; 4), alteration of the starter culture'; ,5), avoidance of the use of excess rennet.'

Some 20 years ago, a deficiency of thecheese grading system was pointed out;cheese made with extended time .in thewhey could be of acceptable quallty whengraded at an early stage, but yet developserious defects on maturing. Since thenstandardisation of the pH or' acidity of ei-ther the whey or the curd alone is not suffi-cient; it is also necessary to control therate' of acid development and the time thecurd is held in the whey (Czulak et al,1969).

Manufacture of cheese by more sophis-ticated methods (eg addition of denaturedwhey proteins) may often lead to a bitterdefect in the final product. Since denaturedwhey proteins have a stimulatory effect onthe proteolytic processes and organolepticproperties of cheese, starters with strainscapable of hydrolysing bitter peptidesshould be used to produce high-qualitycheeses with added denatured whey pro-teins (Krasheninin et al, 1974).

Addition (at the 3% level) of sweet cow'swhey to cheese milk before pasteurisationin order to increase the water holding ca-pacity of buffalo cheese and the extent ofpeptide hydrolysis in slurries resulted inthe disappearance of the bitterness usuallyfound in buffalo cheese (Tuckey and AI-Fayadh, 1985).

Many means have been tried to acceler-ate cheese ripening and to replace the calfrennet due to supply shortages and varia-tions in quality. However, a bitter defect isoften noticed in the product. A controlledmethod for accelerated cheese ripeningachieved without upsetting the flavour bal-ance would therefore give large cost sav-ings to the cheese industry and satisfac-tion to the cheese research workers. Thefreeze-shock treatment of lactic acid bacte-ria which enables addition of large num-

bers of whole cells to the milk seems to bea promising method for increasing proteol-.ysis and flavour development in cheese(Bartels et al, 1987b). Lactobacilli speciesand especially Lactobacillus helveticus canbe applied successfully to cheeses suchas Gouda to enhance flavour developmentat initial stages of ripening without detri-mental effect on the cheese-making proce-dures and cheese quality (Bartels et al,1987a, b). Use of Lactobacillus casei-caseiL2A in conjunction with the starter lacto-cocci was shown to shorten the ripeningperiod and to improve the Cheddar flavour(El Abboudi, 1990; Laleye et al, 1990).However, when compared to other Lactob-acil/us strains used in Cheddar cheesemanufacture (Lemieux et al, 1989), Lactob-acillus casei-casei L2A did not produce bit-ter off-flavour and was then suggested topossess debittering enzymes.

It is recommended to periodically test in-dividual cow's niilk to detect poor chymo-sin-coagulation characteristics; at thesame time abnormal milk, which is poorlycoagulated by chymosin, would also be de-tected. Selective breeding of lactatingcows could guarantee the exclusion ofpoor chymosin-coagulating milk from bulkmilk supplies (Okigbo, 1986).

Good hygiene and storage of milk attemperatures below 4 oCwould help to lirn-it thegrowth of Pseudomonas before UHTtreatment and thus prevent the gelation ofthe product before the expiry date (Bas-sette et al, 1986; Dousset et al, 1988). UI-trasound imaging is a promising non-destructive technique by which microbio-logical deterioration of UHT base materialfor soft ice-cream caused by Bacil/us cere-us and Staphylococcus aureus, and ofUHT milk by Pseudomonas f1uorescens(Ahvenainen et al, 1989a, b, c) can be eas-ily detected. This would be a new way toverity the microbiological sterility of theproduct.


CONCLUSION

Because of the great variety of technologi-cal parameters and the extremely complexsystem of both enzymes and substratespresent in the dairy products and especial-Iy in cheese, the understanding of thecheese bitter flavour defect is still ofpresent interest. Many studies aiming toelucidate the cause of this off-flavour de-fect have been considered invalid becausethey were carried out in relatively uncon-trolled conditions. The development of theaseptic vat technique has opened a doorto research on the bitterness defect by al-lowing the problem to be divided into partswhich can be investigated separately or inappropriate combinations. This kind ofstudy, while bringing light on the role of in-dividual enzyme systems in bitter flavourdevelopment in cheese would possiblylead to a better insight into the mechanismof cheese ripening in general and, morefundamentally, would help to understandthe basics of the cheese industry in the21st century. Indeed, understanding thebasics would allow cheese-makers tomodify the composition, flavour and physi-cal properties of cheese without creatingundesirable side effects.

The second part of this review will bemore concerned about the mechanism offormation of bitter peptides, their isolation,identification, composition and structure. Afinal point will de scribe ways of masking orinhibiting the bitter off-flavour defect indairy products, more specifically in chees-es.

REFERENCES

Abo-Elnaga IG (1974) Degeneration of whitecheese offered in the market. Sudan J FoodSei Teehnol6, 35

Adda J, Gripon JC, Vassal L (1982)The chemis-try of flavour and texture generation incheese. Food Chem 9,115-129

Ahvenainen R, Mattila T, Wirtanen G (1989a) UI-trasound penetration through different pack-aging materials. - A non-destructive methodfor quality control of packed UHT milk. Le-bensm Wiss Teehno/22, 268-272

Ahvenainen R, Wirtanen G, Manninen M(1989b) Ultrasound imaging. - A non-destructive method for control of the microbi-ological quality of aseptically packed food-stuffs. Lebensm Wiss Teehno/22, 273-278

Ahvenainen R, Wirtanen G, Manninen M(1989c) Ultrasound imaging. - A non-destructive method for monitoring the micro-biological quality of aseptically-packed milkproducts. Lebensm Wiss Teehno/22, 382-386

Alkhalaf W, Vassal L, Desmazeaud MJ, GriponJC (1987) Utilisation de Rulactine en tantqu'agent d'affinage dans des fromages àpâte pressée. LaitG7, 173-185

Alkhalaf W, Piard JC, El Soda M, Gripon JC, De-smazeaud M, Vassal L (1988) Liposomes asproteinase carriers for the accelerated ripen-ing of Saint-Paulin type cheese. J Food Sei53,1674-1679

Ardô Y, Pettersson HE (1988) Acceleratedcheese ripening with heat treated cells ofLaetobaeil/us helvetieus and a commercialproteolytic enzyme. J Dairy Res 55, 239-245

Ardô Y, Larsson PO, Lindmark Mansson H, He-denberg A (1989) Studies of peptidolysis dur-ing early maturation and its influence on low-fat cheese quality. Milehwissenschaft 44,485-490

Banks JM (1988) Elimination of the develop-ment of bitter flavour in Cheddar cheesemade from milk containing heat-denaturedwhey protein. J Soc Dairy Teehnol41 , 37-41

Banks JM, Stewart G, Muir DO, West IG (1987)Increasing the yield of Cheddar cheese bythe acidification of milk containing heat-denatured whey protein. Milehwissensehaft42,212-215

Barbosa M, Valles E, Vassal L, Mocquot G(1976) L'utilisation d'extrait de Cynara ear-duneulus L comme agent coagulant en fabri-cation de fromages à pâte molle et à pâtecuite. Lait 56, 1-17


Bartels HJ, Johnson ME, Oison NF (1985a) Ac-celerated ripening of Gouda cheese. 1. Effectand evaluation of thermophilic lactobacilli andstreptococci on proteolysis and flavor devel-opment. J Dairy Sci 68 (suppl 1), 69-70

Bartels HJ, Johnson ME, Oison NF (1985b) Ac-celerated ripening of Gouda cheese. 2. Effectof freeze-treated Lactobacillus helveticus onproteolysis and f1avor development. J DairySci68 (suppl 1),70

Bartels HJ, Johnson ME, Oison NF (1987a) Ac-celerated ripening of Gouda cheese. 1. Effectof heat-shocked thermophilic lactobacilli andstreptococci on proteolysis and flavor devel-opme nt. Mi/chwissenschaft 42, 83-88

Bartels HJ, Johnson ME, Oison NF (1987b) Ac-celerated ripening of Gouda cheese. 2. Effectof freeze-shocked Lactobacillus helveticus onproteolysis and flavour development. Mi/ch-wissenschaft 42, 139-144

Barton LB (1957) A modification of the Cheddarcheese process and its influence on selectedchemical and physical properties of thecheese. Milk Prod J 48, 11-12 & 39-41

Bassette R, Fung DYC, Mantha VR (1986) Off-flavors in milk. GRG Grit Rev Food Sci Nutr24,1-52

Berdagué JL, Grappin R (1988) Affinage et qua-lité du gruyère de Comté. VI. Caractéris-tiques sensorielles des fromages. Lait 68,189-204

Biede SL, Hammond EG (1979) Swiss cheesef1avor. II. Organoleptic analysis. J Dairy Sci62,238-248

Champion HM, Stanley DW (1982) HPLC separ-ation of bitter peptides from Cheddar cheese.Gan Inst Food Sci Technol J 15, 283-288

Chapman HR, Sharpe ME, Law BA (1976)Some effects of low-temperature storage ofmilk on cheese production and Cheddarcheese f1avour. Dairy Ind Int 41, 42-45

Charalambous G (1986) Handbook of Food andBeverage Stability. Ghemical, Biochemical,Microbiological and Nutritional Aspects. Aca-demic Press, St-Louis

Chavarri FJ, de Paz M, Nuriez M (1988) Optimi-zation of fermentation parameters for the pro-duction of concentrated starters from non bit-ter Streptococcus lactis INIA 12. J Food Sci53,1854-1857

Chettel JC, Chettel H, Besançon P (1977) Intro-duction à la Biochimie et à la Technologiedes Aliments. Technique et Documentation,Entreprise Moderne d'Edition, Paris, vol 2

Choisy C, Lenoir J, Tourneur C, Guéguen M, Ju-ras M, Rivault D (1978) Essai de caractérisa-tion des streptocoques lactiques au regardde l'amertume des fromages. In: XX Int DairyGongr (brief communications) E, 498-499

Choisy C, Desmazeaud M, Gripon JC, LamberetG, Lenoir J, Tourneur C (1984) Les phé-nomènes microbiologiques et enzymatiqueset la biochimie de l'affinage. In: Le Fromage(Eck A, ed) Technique et Documentation, Dif-fusion Lavoisier, Paris

Christensen TMIE, Kristiansen KR, Madsen JS(1989) Proteolysis in cheese investigated byhigh performance Iiquid chromatography.J Dairy Res 56, 823-828

Clegg KM, Lim CL, Manson W (1974) The struc-ture of a bitter peptide derived from casein bydigestion with papain. J Dairy Res 41, 283-287

Crawford RJM (1977) Introduction to discussionon bitterness in cheese. lnt Dairy Fed AnnuBull 97

Creamer LK (1975) ~-casein degradation inGouda and Cheddar cheese. J Dairy Sci 58,287-292

Creamer LK (1978) Degradation of casein com-ponents during cheese ripening. In: Fifth In-ternational Gongress of Food Science andTechnology. Kyoto, Japan (abstr), 57

Creamer LK (1979) Chemical aspects of foodquality. In: Developments in Food Science: 2.Proceedings of the Fifth International con-gress of Food Science and Technology, Kyo-to, Japan (Chiba H, Fujimaki M, Iwai K, Mit-suda H, Morita Y, eds) Elsevier, Amsterdam,340-346

Creamer LK, Lawrence RC, Gilles J, HumphriesMA (1970) Bitterness in cheese: role of start-er demonstrated by computed statistics. NZJDairy Sci Techno/5, 35

Creamer LK, Lawrence RC, Gilles J (1985) Ef-fect of acidification of cheese milk on the re-sultant Cheddar cheese. NZJ Dairy Sci Tech-no/20, 185-203

Creamer LK, Iyer M, Lelievre J (1987) Effect ofvarious levels of rennet addition on charac-teristics of Cheddar cheese made from ultra-


filtered milk. NZJ Dairy Sei Teehno/22, 205-214

Czulak J (1959) Bitter f1avourin cheese. Aust JDairyTeehno/14,177-179

Czulak J, Conochie J, Sutherland BJ, van Leeu-wen HJM (1969) Lactose, lactic acid andminerai equilibria in Cheddar cheese manu-facture. J Dairy Res 36, 93-101

Dawson DJ, Feagan JT (1960) Bitter f1avoursincheese. Aust J Dairy Teehno/15, 65-66

Deane DO (1951) Preliminary studies of the ef-fect of acido-proteolytic organisms and tem-peratures of curing on the ripening of Ched-dar cheese made from pasteurized milk.J Dairy Sei 34, 776-783

Deane DO, Dolan ET (1973) Curing of cheesewith reduced fat content. J Dairy Sei 56, 631-633

Delbeke R, Naudts M (1970) La fabrication defromage de Herve à partir de lait cru. RevAgrie 2, 333-355

Delformo G, Parpani G (1986) Bitter f1avoursincheese. Mondo Latte 40, 565-567

Desmazeaud M (1983) Comment les bactérieslactiques se comportent-elles dans le lait ?Tech Lait 976, 11-18

Desmazeaud MJ, Gripon JC (1977) Generalmechanism of protein breakdown duringcheese ripening. Milehwissensehaft 32, 731-734

Dinesen N (1974) The real value of cheese co-agulants. Dairy lee Gream Field 157, 78-80

Dousset X, Demaimay M, Ravaud C, LevesqueA, Pinet X, Kergo Y (1988) Influence de latempérature de réfrigération du lait sur laprotéolyse et l'amertume du lait UHT aucours de son stockage. Lait68, 143-156

Dulley JR (1976) The utilization of cheese slur-ries to accelerate the ripening of Cheddarcheese. Aust J Dairy Teehno/31 , 143-148

Dulley JH, Brooks DEJ, Grieve PA (1978) Utili-satlon possible de levains ne fermentant pasle lactose pour accélérer la maturation dufromage et méthode pour les détecter dansle fromage. In: XX Int Dairy Gongr (brief com-munications) F, 492-493

Dumont JP, Roger S, Adda J (1974) Étude descomposés volatils neutres présents dans lesfromages à pâte molle et à croûte lavée. Lait54,31-43

Dunn HC, Lindsay RC (1985) Evaluation of therole of microbial Strecker-derived aromacompounds in unclean-type f1avorsof Ched-dar cheese. J Dairy Sei 68, 2859·2874

Edwards JL Jr, Kosikowski FV (1970) Electro-phoretic proteolytic patterns in Cheddarcheese by rennet and fungal rennets: theirsignificance to international classification ofcheese varieties. J Dairy Sei 52, 1675-1678

Edwards JL Jr, Kosikowski FV (1983) Bittercompounds from Cheddar cheese. J DairySei 66, 727-734

El Abboudi M (1990) Développement d'un pro-cédé de maturation accélérée du fromageCheddar. Master's thesis, université Laval,Quebec, Canada

El Abboudi M, Pandian S, Trépanier G, SimardRE, Lee BH (1991) Heat-shocked lactobacillifor acceleration of Cheddar cheese ripening.J Food Sei 56, 948-949, 953

El Soda M, Desmazeaud MJ, Aboudonia S, Ka-mal N (1981) Acceleration of cheese ripeningby the addition of whole cells or cell free ex-tracts from Laetobaeillus easei to the cheesecurd. Milehwissensehaft36, 140-142

El Soda M, Desmazeaud MJ, Aboudonia S, Ba-dran A (1982) Acceleration of cheese ripen-ing by the addition of extracts from Laetoba-eillus helvetieus, Laetobaeillus bulgarieus andLaetobaeillus laetis to the cheese curd. Mileh-wissensehaft 37, 325-327

El Soda M, Hussein S, Ezzat N (1985) Accelera-tion of Ras cheese ripening with commercialenzyme preparations. J Dairy Sei 68 (suppl1),71

Emmons DB, McGugan WA, Elliott JA (1960a)Effect of strain of starter culture and manu-facturing procedure on bitterness in Cheddarcheese. J Dairy Sei 43, 861-862

Emmons DB, McGugan WA, Elliott JA (1960b)Nitrogen distribution in Cheddar cheese asrelated to bitterness and strain of starter cul-ture. J Dairy Sei 43,862

Emmons DB, McGugan WA, Elliott JA, MorsePM (1961) Effect of combining single-straincultures as cheese starter on bitter flavor inCheddar cheese at 6 months of age. J DairySei44,1157

Emmons DB, McGugan WA, Elliott JA, MorsePM (1962a) Effect of strain of starter cultureand of manufacturing procedure on bitter-


ness and protein breakdown in Cheddarcheese. J Dairy Sei 45, 332-342

Emmons DB, McGugan WA, Elliott JA, MorsePM (1962b) Effect of combining single-straincultures as cheese starter on bitterness inCheddar cheese at six months of age.J Dairy Sei 45, 595-600

Exterkate FA (1976) Comparison of strains ofStreptoeoeeus eremoris for proteolytic activi-ties associated with the cell wall. Neth MilkDairy J 30,95-105

Exterkate FA (1983) The proteolytic system ofstarter streptococci: the crucial point incheesemaking. Neth Milk DairyJ37, 94-96

Exterkate FA, Stadhouders J (1971) Pyrrolidonecarboxylyl peptidase activity in bitter andnon-bitter strains of Streptoeoeeus eremoris.Neth Milk Dairy J 25, 240-245

Fedrick lA, Aston JW, Nottingham SM, DulleyJR (1986) The effect of a neutral fungal pro-tease on Cheddar cheese ripening. NZJDairy Sei Teehno/21 , 9-19

Fox PF (1985) Developments in Dairy Chemis-try: 3. Elsevier, NY

Fox PF (1988) Rennets and theïr action incheese manufacture and ripening. BioteehnolAppl Bioehem 10, 522-535

Fox PF (1988-1989) Acceleration of cheese ri-pening. Food Bioteehnol (NY) 2, 133-185

Fox PF, Walley BF (1971) Influence of sodiumchloride on the proteolysis of casein by ren-net and by pepsin. J Dairy Res 38, 165-170

Frey JP, Johnson ME, Marth EH (1986) Pepti-dases and proteases in barley extract: a po-tenlial source of enzymes for use in cheeseripening. Milehwissensehaft 41, 488-489

Fryer TF (1969) Microflora of Cheddar cheeseand its influence on cheese flavour. Dairy SeiAbstr31 , 471-490

Fujimaki M, Yamashita M, Okazawa Y, Arai S(1970) Applying proteolytic enzymes on soy-bean. 3. Diffusable bitter peptides and freeamino acids in peptic hydrolysate of soybeanprotein. J Food Sei 35, 215-218

Gavron H, Lück H, Van der Merwe NL (1982)The manufacture of cheese using geneticallymodified starter culture. S Afr J Dairy Tecti-no/14, 153-157

Gebre-Egziabher A, Humbert ES, BlankenagelG (1980) Hydrolysis of milk proteins by mi-crobial enzymes. J Food Prot43, 709-712

Golding NS (1947) Bitter flavor in AmericanCheddar cheese made from pasteurizedmilk. Wash State Coll Inst Dairy Proe 16,146-150

Gordon DF Jr, Speck ML (1965) Bitterness inmilk cultures of Streptoeoeeus eremoris.J Dairy Sei 48, 499-500

Green ML, Valler MJ, Kay J (1984) Assessmentof the suitability for Cheddar cheesemakingof purified and commercial chicken pepsinpreparations. J Dairy Res 51, 331-340

Grieve PA, Dulley JR (1983) Use of Streptoeoe-eus laetis lac- mutants for accelerating Ched-dar cheese ripening. 2. Their effect on therate of proteolysis and flavour development.AustJ Dairy Teehno/38, 49·54

Gripon JC, Desmazeaud MJ, Le Bars D,Bergère JL (1975) Étude du rôle des microor-ganismes et des enzymes au cours de lamaturation des fromages. II. Influence de laprésure commerciale. Lait 55, 502-516

Guigoz Y, Solms J (1974) Isolation of a bittertasting peptide from "Alpkase", a Swiss Mou-tain-cheese. Lebensm Wiss Teehnol7, 356-357

Hamilton JS, Hill RD, Van Leeuwen H (1974) Abitter peptide from Cheddar cheese. AgrieBiol Chem 38, 375-379

Hansen HC, Bendixen HA, Theophilus DR(1933) Influence of different starters on thequality of Cheddar cheese. J Dairy Sei 16,121-127

Harris WC, Hammer BW (1940) Effect of variousbacteria on f1avorof Cheddar cheese madefrom pasteurized milk. J Dairy Sei 23, 701-708

Harwalkar VR, Elliott JA (1971) Isolation of bitterand astringent fractions from Cheddarcheese. J Dairy Sei 54, 8-11

Harwalkar VR, Seitz EW (1971) Production ofbitter f1avor components by lactic cultures.J Dairy Sei 54, 12-14

Hegazi FZ (1989) Some properties of Strepto-coccus faeealis subsp liquefaeiens isolatedfrom cheese with special reference to pro-duction of bitterness. Nahrung 33, 989-997

630

Jensen RG (1964) Lipolysis. J Dairy Sei 47,210-215

Jensen JP, Reinbold GW, Washam CJ, Veda-muthu ER (1975) Role of enterococci inCheddar cheese: organoleptic considera-tions. J Milk Food Teehno/38, 142-145

Kalogridou-Vassiliadou D, Alichanidis E (1984)Effect of refrigerated storage of milk on themanufacture and quality of Teleme cheese.J Dairy Res 51, 629-636

Kamaly KM, Marth EH (1988) Proteinase andpeptidase activities of cell-free extracts frommutant strains of lactic streptococci. J DairySei 71, 2349-2357

Kaneko S, Yoneda Y (1974) Studies on thecause of bitterness of yeast ripened cheese.ln: XIX Int Dairy Congr (brief communica-tions) 1E, 776

Keller B, Oison NF, Richardson T (1974) Miner-ai retention and rheological properties ofMozzarella cheese made by direct acidifica-tion. J Dairy Sei 57,174-180

Kempler GM, Baldwin KA, McKay LL, MorrisHA, Halambeck S, Thorsen G (1979) Use ofgenetic alterations to improve Streptocoeeuslaetis C2 as a potential Cheddar cheese start-er. J Dairy Sei 62, (suppl 1), 42-43

Kikuchi T, Toyoda S (1970) Use of microbialrennets in cheesemaking. In: XVIII Int DairyCongr1E,285

Kim MS, Jensen LA, Oison NF (1987) Enhanc-ing flavor characteristics of Gouda cheesefrom ultrafiltered milk by addition of commer-cial enzyme preparations and freeze-shocked Laetobaeillus helvetieus. J Dairy Sei70 (suppl 1), 66

Kleter G (1976) The ripening of Gouda cheesemade under strict aseptic conditions. 1.Cheese with no other bacterial enzymes thanthose from a starter streptococcus. Neth MilkDairy J 30, 254-270

Klimovsky l, Zvyagintsev V, Gudkov A, Medve-deva Z (1970) Detection of "bitter" strains oflactic acid streptococci. In: XVIII Int Dairy

, Congr1E,119

Kobayashi H, Kusakabe l, Murakami K (1985)Milk-c1otting enzyme Irom Irpex laeteus as acalf rennet substitute for Cheddar cheesemanufacture. Agde Biol Chem 49, 1605-1609


Hicks CL, O'Leary J, Bucy J (1984) Use of re-combinant chymosin in the manufacture ofCheddar cheese. J Dairy Sei 67 (suppl 1), 73

Hicks CL, Onuorah C, O'Leary J, Langlois BE(1986) Effect of milk quality and low tempera-ture storage on cheese yield - a summation.J Dairy Sei 69, 649-657

Hicks CL, O'Leary J, Bucy J (1988) Use of re-combinant chymosin in the manufacture 01Cheddar and Colby cheese. J Dairy Sei 71,1127-1131

Hillier AJ, Kanasaki M, Jago GR (1975) Effect oftemperature on the growth and acid produc-tion of lactic acid bacteria. 3. The influenceof added growth supplement. Aust J DairyTeehno/30, 149-152

Hofi AA, EI-Satty MS, Mahran GA, Khorshid MA(1976) Ripening changes in cephalotrye"Ras" cheese as affected by microbial ren-net. Egypt J Dairy Sei 4, 63-70

Holmes DG, Duersch JW, Ernstrom CA (1977)Distribution of milk clotting enzymes betweencurd and whey and their survival duringCheddar cheese making. J Dairy Sei 60,862-869

Huber L, Klostermeyer H (1974) Isolation of abitter peptide from the cheese "Buttsrkâse"and its identification. Milehwissensehaft 29,449-455

Hugenholtz J, Exterkate F, Konings WN (1984)The proteolytic systems of Streptoeoeeuseremoris: an immunological analysis. ApplEnviron Mierobio/48, 1105-1110

Ishibashi N, Sadamori K, Yamamoto 0, Kanehi-sa H, Kouge K, Kikuchi E, Okai H, Fukui S(1987) Bitterness of phenylalanine- and tyro-sine-containing peptides. Agde Biol Chem51, 3309-3313

Ivie GW, Witzel DA, Rushing DO (1975) Toxicityand milk bittering properties of tenulin, themajor sesquiterpene lactone constituent ofHelenium amarum (bitter sneezeweed). JAgde Food Chem 23, 845-849

Jago GR (1962) Proteolysis of milk proteins inrelation to bitter flavour in cheese. Aust JDairy Teehno/17, 83-85

Jago GR (1974) Control of the bitter flavour de-fect in cheese. Aust J Dairy Teehno/29, 94-96


Kosikowski FV, Brown DP (1973) Influence ofcarbon dioxide and nitrogen on microbialpopulations and shelf Iife of Cottage cheeseand sour cream. J Dairy Sei 56, 12-18

Kosikowski FV, Iwasaki T (1975) Changes inCheddar cheese by commercial enzymepreparations. J Dairy Sei 58, 963-970

Krasheninin PF, Zvyagintsev VI, Nebert VK, Tol-kachev AN (1974) Effect of whey proteins onthe proteolytic process in cheese. In: XIX IntDairy Gongr (brief communications) 1E, 343

Laleye LC, Simard RE, Lee BH, Holley RA(1990) Quality attributes of Cheddar cheesecontaining added lactobacilli. J Food Sei 55,114-118

Law BA, Wigmore AS (1982a) Microbial protei-nases as agents for accelerated cheese ri-pening. J Soc Dairy Teehno/35, 75-76

Law BA, Wigmore AS (1982b) Acceleratedcheese ripening with food grade proteinases.J Dairy Res 49,137-146

Law BA, Wigmore AS (1982c) Use of exoge-nous enzymes for accelerated ripening ofCheddar cheese. In: XXI Int Dairy Gongr(brief communications), 1(2), 505

Law BA, Wigmore AS (1983) Accelerated ripen-ing of Cheddar cheese with a commercialproteinase and intracellular enzymes fromstarter streptococci. J Dairy Res so, 519-525

Law BA, Sharpe ME, Chapman HR (1976) Theeffect of Iipolytic Gram-negative psychro-trophs in stored milk on the development ofrancidity in Cheddar cheese. J Dairy Res 43,459-468

Law BA, Hosking ZD, Chapman HR (1979) Theeffect of some manufacturing conditions onthe development of flavour in Cheddarcheese. J Soc Dairy Teehno/32, 87-90

Lawrence RC, Pearce LE (1968) The caseagainst the use of unpaired single starterstrains. NZJ Dairy Sei Teehno!3, 137-139

Lawrence RC, Gilles J (1969) The formation ofbitterness in cheese: a critical evaluation.NZJ Dairy Sei Teehno/4, 189-196

Lawrence RC, Gilles J (1971) The use of rennetin cheese-making. NZJ Dairy Sei Teehno! 6,30

Lawrence RC, Gilles J (1973) Bacteriophage:the unconsidered factor in Cheddar cheeseflavour. NZJ Dairy Sei Teehno!8, 122-123

Lawrence RC, Creamer LK, Gilles J, Martley FG(1972) Cheddar cheese flavour. 1. The roleof starters and rennets. NZJ Dairy Sei Tech-no! 7, 32-37

Lawrence RC, Thomas TD, Terzaghi BE (1976)Reviews of the progress of dairy science:cheese starters. J Dairy Res 43, 141-193

Lawrence RC, Gilles J, Creamer LK (1983) Therelationship between cheese texture and fla-vour. NZJ Dairy Sei Teehno/18, 175-190

Lawrence RC, Heap HA, Gilles J (1984) A con-trolled approach to cheese technology.J Dairy Sei 67, 1632-1645

Le Bars D, Desmazeaud MJ, Gripon JC,Bergère JL (1975) Étude du rôle des microor-ganismes et de leurs enzymes dans la matu-ration des fromages. 1. Fabrication aseptiqued'un caillé modèle. Lait 55, 377-389

Lebedev AB, Umanskit MS (1981) Hydrolysis ofmilk fat by milk lipoprotein lipase. MoloehnPromst9, 13-14,47

Lee BH, Laleye LC, Simard RE, Holley RA, Em-mons DB, Giroux RN (1990a) Influence ofhomofermentative lactobacilli on physico-chemical and sensory properties of Cheddarcheese. J Food Sei 55, 386-390

Lee BH, Laleye LC, Simard RE, Munsch MH,Holley RA (1990b) Influence of homofermen-tative lactobacilli on the microflora and solu-ble nitrogen components in Cheddar cheese.J Food Sei 55, 391-397

Lelievre J, Lawrence RC (1988) Manufacture ofcheese from milk concentrated by ultrafiltra-tion. J Dairy Res 55, 465-478

Lemieux L, Puchades R, Simard RE (1989)Size-exclusion HPLC separation of bitter andastringent fractions from Cheddar cheesemade with added Laetobaeil/us strains to ac-celerate ripening. J Food Sei 54, 1234-1237

Lowrie RJ (1977) Influence of lactic streptococcion bitter f1avor development in cheese.J Dairy Sei 60, 810-814

Lowrie RJ, Lawrence RC (1972) Cheddarcheese flavour. IV. A new hypothesis to ac-count for the development of bitterness. NZJDairy Sei Teehnol7, 51-53

Lowrie RJ, Lawrence RC, Pearce LE, RichardsEL (1972) Cheddar cheese flavour. III. Thegrowth of lactic streptococci during cheese-


making and the effect on bitterness develop-ment. NZJ Dairy Sei Teehnol7, 44-50

Lowrie RJ, Lawrence RC, Peberdy F (1974)Cheddar cheese f1avour.V. Influence of bac-teriophage and cooking temperature oncheese made under controlled bacteriologi-cal conditions. NZJ Dairy Sei Teehno/9, 116-121

Mabbitt LA, Chapman HR, Berridge NJ (1955)Experiments in cheesemaking without start-er. J Dairy Res 22,365-373

Mabbitt LA, Chapman HR, Sharpe ME (1959)Making Cheddar cheese on a small scale un-der controlled bacteriological conditions.J Dairy Res 26, 105-112

Mahran GA, EI-Satty MS, Abdel-Hamid LB,Khorhid MA (1976) The ripening of Rascheese as affected by the use of mixtures ofanimal and microbial rennets. Egypt J DairySei 4,13-20

Mâlkki Y (1978) Ripening and debittering ofcheeses and protein hydrolysates. In: FifthInt Congr Food Sei Technol (abstr) Kyoto,231

Mâlkki Y, Mattsson R, Rouhiainen L, Karaila P,IImonen T (1976) Properties and use of anew Pseudomonas peptidase. In: Fifth Inter-national Fermentation Symposium (abstr),Berlin, 19.07

Mâlkki Y, Karaila P, Antila V, Witting 6 (1978)The use of peptidases in ripening of cheese.ln: Second Franeo-Finnish Symposium onBioteehnology-Bioeonversion in Food Teeh-nology (proceedings), Helsinki, Finland, 91-95

Mâlkki Y, Mattsson R, Markkanen P, Lindroth S,Karaila P (1979) Production and utilization ofPseudomonas f1uoreseens peptidases. In:Second International Congress on Engineer-ing and Food and the Eight European FoodSymposium (proceedings), Espoo, Finland,282-288

Martens R (1973) Gouda cheese made with mi-crobial rennets derived from Mucor miehei.Milehwissensehaft28,87-91

Martens R (1974) Utilisation de ferments con-centrés congelés dans la fabrication de fro-mage Gouda. Rev Agrie 27, 1439-1449

Martley FG (1975a) The behaviour and role ofstarter streptococci in Camembert cheese-making. NZJ Dairy Sei Teehno/10, 12-17

Martley FG (1975b) Comportement et rôle desstreptocoques lactiques du levain en fabrica-tion de Camembert. Lait 55, 310-323

Martley FG, Lawrence RC (1972) Cheddarcheese flavour. II. Characteristics of singlestrain starters associated with good or poorflavour development. NZJ Dairy Sei Technol7,38-44

Matoba T, Hata T (1972) Relationship betweenbitterness of peptides and their chemicalstructures. Agde Biol Chem 36, 1423-1431

McGugan WA, Howsam SG, Elliott JA, EmmonsDB, Reiter B, Sharpe ME (1968) Neutral vol-atiles in Cheddar cheese made asepticallywith and without starter culture. J Dairy Res35,237-245

McGugan WA, Emmons DB, Larmond E (1979)Influence of volatile and nonvolatile fractionson intensity of Cheddar cheese flavor.J Dairy Sei 62, 398-403

McKay LL, Baldwin KA (1978) Stabilization oflactose metabolism in Streptoeoeeus laetisC2• Appl Environ Mierobio/36, 360-367

McKellar RC (1981) Development of off-f1avorsin ultra-high temperature and pasteurizedmilk as a function of proteolysis. J Dairy Sei64,2138-2145

Milis OE, Thomas TD (1980) Bitterness develop-ment in Cheddar cheese: effect of the level ofstarter proteinase. NZJ Dairy Sei Teehno/15,131-141

Milis OE, Thomas TD (1982) Starter proteinaseand bitterness development in Cheddarcheese. In: XXI Int Dairy Congr 1 (1), 511-512

Mistry VV, Kosikowski FV (1986a) Application ofretentate starter produced from ultrafilteredmilk to the manufacture of Cheddar cheese.J Dairy Sei 69, 1484-1490

Mistry VV, Kosikowski FV (1986b) Use of reten-tate starter produced by ultrafiltration in themanufacture of Cheddar cheese from 1.7:1whole milk retentates. J Dairy Sei 69 (suppl1) 79

Moir GM (1930) Pasteurised milk for Cheddarcheesemaking. 1. A preliminary chemical in-vestigation. J Dairy Res 1, 149-167

Moneib A, Safwat 1 (1972) Studies on improvinglocally manufactured Kaskaval cheese. II. Ef-fect of waxing or packing in polyethylenesacs on the physical and chemical propertiesof the cheese. Agde Res Rev 50, 169-174

Bitter f1avourin dairy products: part 1 633

Monnet V, Le Bars D, Gripon JC (1986) Speci-ficity of a cell wall proteinase from Strepto-coccus lactis NCD0763 towards bovine [3-casein. FEBS Microbiol Lett36, 127-131

Mortensen BK (1984) Recent developments inthe utilization of milk proteins in dairy prod-ucts. In: Proceedings of the InternationalGongress on Milk Proteins, Luxembourg,109-119

Mourgues R, Bergère JL, Vassal L (1983) Possi-bilités d'améliorer les qualités organolep-tiques des fromages de Camembert grâce àl'utilisation de Geotrichum candidum. TechLait 978, 11-15

Muysson DJ, Verrinder Gibbins AM (1989) Thealteration of milk content by genetic engineer-ing and recombinant DNA-mediated selectiontechniques. Gan J Anim Sci 69, 517-527

Nelson JH (1974) The riddle of bitter f1avor incheese. Ital Cheese J 3, 1-4,6

Ney KH (1971) Prediction of bitterness of pep-tides from their amino acid composition.Z Lebensm Unters Forsch 147, 64-68

Ney KH (1979) Bitterness of peptides: aminoacid composition and chain length. In: FoodTaste Chemistry (Boudreau JC, ed) ACSSymp Ser 115, Washington, DC, 149-174

Nishimura T, Kato H (1988) Taste of free aminoacids and peptides. Food Rev Int 4, 175-194

Nunez M, Nuüez JA, Medina AL (1982) Freeze-dried concentrated starters for direct inocula-tion of vat milk in Manchego cheese manu-facture. Milchwissenschaft 37, 328-330

Okai H (1977) Relationship between bitternessand chemical structure in synthetic peptides.ln: Peptide Ghemistry 1976 (Nakajima T, ed)14th Symp Peptide Chemistry (proceedings),Hiroshima, 139-142

O'Keeffe RB, Fox PF, Daly C (1975) Proteolysisin Cheddar cheese: influence of the rate ofacid production during manufacture. J DairyRes 42, 111·122

OkigboLM (1986) Evaluation and improvementof coagulation properties of milk for manufac-turing. PhD thesis. Diss Abstrtm B 46, 2515.Order No DA8523677

Okigbo LM, Richardson GH, Brown RJ, Ern-strom CA (1985) Casein composition of

. cow's milk of different chymosin coagulationproperties. J Dairy Sci 68, 1887-1892

Patel GB, Blankenagel G (1972) Bacterialcounts of raw milk and flavor of the milk afterpasteurization and storage. J Milk FoodTechno/35,203-206

Pélissier JP, Manchon P (1976) Comparativestudy of the bitter taste of enzymic hydroly-sates from cow, ewe and goat caseins. JFood Sci 41, 231-233

Pélissier JP, Mercier JC, Ribadeau-Dumas B(1974) Problème de l'amertume dans lesfromages - résultats d'une enquête. Rev LaitFr 325, 820-821

Perry KD, McGillivray WA (1964) The manufac-ture of "normal" and "starter-free" Cheddarcheese under controlled bacteriological con-ditions. J Dairy Res 31,155-165

Phelan JA, Guiney J, Fox PF (1973) Proteolysisof [3-caseinin Cheddar cheese. J Dairy Res40,105-112

Phillips CA (1935) Bitter flavor in Cheddarcheese made from pasteurized milk. J DairySci 18,471-472

Polansky J (1989) Effect of dairy cow nutritionon certain blood characteristics and parame-ters of milk quality. Zivoeisna Vyroba 34,491-497

Price WV (1936) Bitter f1avor in Cheddarcheese. J Dairy Sci 19, 185-190

Quarne EL, Larson WA, Oison NF (1968) Effectof acidulants and milk-c1otting enzymes onyield, sensory quality, and proteolysis of Piz-za cheese made by direct acidification.J Dairy Sei 51, 848-852

Raadsveld CW (1953) Bitter compounds fromcheese. In: XIII Int Dairy Gongr (proceed-ings), Il (Sect Il), 676-680

Rao KSN, Nand K, Srikanta S, Krishna-SwamyMA, Murthy VS (1979) Changes during man-ufacture and ripening of Cheddar cheese pre-pared with fungal rennet substitute of Rhiio-pus oligosporus. Nahrung 23, 621c~26 ' ,

. '., -';",-., l

Reiter B, Sharpe ME (1971) Relationship,of themicroflora tothe flavour of:Ch,eddarcheese.J Appl Bacterio/34, 63-80

Reiter B, Fryer TF, Pickering A, Chapman ,HR,Lawrence RC, Sharpe ME (1967) Theeffectof the microbial flora on the flavour and freefatty acid composition of Cheddar cheese. JDairy Res 34,257-272


Renz VU, Puhan Z (1975) Contribution taknowledge of factors favoring bitterness inyoghourt. Milchwissenschfat 30, 265-271

Reps A, Poznanski S, Kowalewska J, RoskoszD (1974) Characteristics of ripening in chees-es made using microbial rennet substitutes.Rocz Inst Przem Mlecz 16, 5-20

Ribadeau-Dumas B (1984) Maîtrise de l'affinagedes fromages de type Camembert. Lait 64,448-468

Richardson BC, Creamer LK (1973) Casein pro-teolysis and bitter peptides in Cheddarcheese. NZJ Dairy Sci Techno/8, 46-51

Richardson BC, Newstead DF (1979) Eftect ofheat-stable proteases on the storage life ofUHT milk. NZJ Dairy Sci Techno/14, 273-279

Richardson GH, Ernstrom CA, Kim JM, Daly C(1983) Proteinase negative variants of Strep-tococcus cremoris for cheese starters.J Dairy Sei 66, 2278-2286

Riddet W, Valentine GM, Whitehead HR (1932)A starter producing bitterness in cheese.NZJ Agric 44, 32-34

Sandine WE, Daly C, Elliker PR, VedamuthuER (1972) Causes and control of culture-related flavor defects in cultured dairy prod-ucts. J Dairy Sci 55, 1031-1039

Sardinas JL (1972) Microbial rennets. Adv ApplMicrobio/15, 39-73

Schulz ME, Thomasow J (1970) Suitability ofmilk-coagulating enzymes for cheesemakingin relation ta the cheese variety. In: XVlllintDairy Congr 1E, 321

Scott R (1972) Cheesemaking-enzymology orbacteriology? Process Biochem 7,33-36

Scott R (1986) Cheese Making Practice. Elsevi-er Appl Sei Publ, Essex, UK, 2nd edn

Shinoda l, Tada M, Otagiri K, Okai H (1985) Bit-ter taste of Pro-Phe-Pro-Gly-Pro-lle-Pro cor-responding ta the partial sequence (positions61-67) of bovine ~-casein. In: Peptide Chem-istry 1985: 23rd Symposium on PeptideChemistry (proceedings), Kyoto, 287-290

Shinoda l, Okai H, Fukui S (1986a) Bitter tasteof H-Val-Val-Val-Pro-Pro-Phe-Leu-OH corre-sponding ta the partial sequence (positions82-88) of bovine ~-casein, and related pep-tides. Agric Biol Chem 50, 1255-1260

Shinoda l, Tada M, Okai H, Fukui S (1986b) Bit-ter taste of H-Pro-Phe-Pro-Gly-Pro-lle-Pro-OH corresponding ta the partial sequence(positions 61-67) of bovine ~-casein, and re-lated peptides. Agric Biol Chem 50, 1247-1254

Shipe WF, Bassette R, Deane DD, Dunkley WL,Hammond EG, Harper WJ, Kleyn DH, Mor-gan ME, Nelson JH, Scanlan RA (1978) Oftflavors of milk: nomenclature, standards, andbibliography. J Dairy Sci 61, 855-869

Singh J, Chander H, Bhalerao VR, Dastur NN(1973) Studies on vegetable rennet fromWithania coagulans. J Food Sci Techno/1 0,16-19

Smith AE (1989) The potential allelopathic char-acteristics of bitter sneezeweed (Heleniumamarum). Weed Sci37, 665-669

Sood VK, Kosikowski FV (1979) Ripeningchanges and flavor development in microbialenzyme treated Cheddar cheese slurries.J Food Sci 44, 1690-1694

Sorokina 0, Karagueuz A (1978) Désordresaseptiques non spécifiques de la glandemammaire des vaches et influence sur lacomposition et la qualité du lait. In: XX IntDairy Congr E, 160-161

Stadhouders J (1962) The proteolytic activity ofrennet and starter bacteria in cheese with ref-erence ta bitter flavour. In: XVI Int DairyCongr B (section IV), 353-361

Stadhouders J (1974) Dairy starter cultures.Milchwissenschaft 29, 329-337

Stadhouders J, Hup G (1975) Factors aftectingbitter f1avour in Gouda cheese. Neth MilkDairy J 29, 335-353

Stadhouders J, Hup G, Exterkate FA, Visser S(1983) Bitter flavour in cheese. 1. Mechanismof the formation of the bitter flavour defect incheese. NethMilkDairyJ37, 157-167

Stanley DW, Emmons DB, Modler HW, IrvineDM (1980) Cheddar cheese made with chick-en pepsin. Can Inst Food Sci Technol J 13,97-102

Stone WK, Naft DM (1967) Increases in solublenitrogen and bitter flavor development in Cot-tage cheese. J Dairy Sci 50, 1497-1500

Stone WK, Large PM (1968) Preventing bitterflavor development. Am Dairy Rev 3D, 26-27,85

Bitter f1avourin dairy products: part 1 635

Sullivan JJ, Jago GR (1970a) A model for bitterpeptide formation and degradation in cultureddairy products. Aust J Dairy Techno/25, 111

Sullivan JJ, Jago GR (1970b) Pyrrolidonecar-boxylyl peptidase activity in Streptococcuscremoris ML 1. Aust J Dairy Techno/25, 141

Sullivan JJ, Jago GR (1972) The structure of bit-ter peptides and their formation from casein.Aust J Dairy Techno/27, 98-104

Sullivan JJ, Mou L, Rood JI, Jago GR (1973)The enzymic degradation of bitter peptidesby starter streptococci. Aust J Dairy Technol28,20-26

Sullivan JJ, Schlipalius LE, Mou L, Jago GR(1974) Heat lability of starter cultures and itsuse in selecting potentially non-bitter strains.AustJ Dairy Techno/29, 71-74

Teuber M (1990) Production of chymosin (EC3.4.23.4) by microorganisms and its use forcheesemaking. Int Dairy Fed Annu IDF Bull251,3-15

Thomas TD, Pearce KN (1981) Influence of salton lactose fermentation and proteolysis inCheddar cheese. NZJ Dairy Sci Techno/16,253-259

Tittsler RP, Sanders GP, Lochry HR, Sager OS(1948) The influence of various lactobacilliand certain streptococci on the chemicalchanges, flavor development and quality ofCheddar cheese. J Dairy Sci 31, 716

Torrie JP, Cholette H, Froehlich DA, McKellarRC (1983) Growlh of an extracellular protei-nase-deficient strain of Pseudomonas fluo-rescens on milk and milk proteins. J DairyRes 50,365-374

Tuckey SL, Ruehe HA (1940) Relation of saltcontent to bitterflavor development in Ched-dar cheese. J Dairy Sci23, 517-518

Tuckey NK, AI-Fayadh MH (1985) Role of add-ed dry whey and coagulant in the ripening ofbuffalo milk Cheddar cheese slurry. III. Totalpeptides. J Dairy Sci 68 (suppl 1), 91

Vassal L, Gripon JC (1984) L'amertume desfromages à pâte molle de type Camembert:rôle de la présure et de Penicillium ceseico-lum, moyens de la contrôler. Lait 64, 397-417

Visser FMW (1977a) Contribution of enzymesfrom rennet, starter bacteria and milk to pro-teolysis and flavour development in Goudacheese. 1. Description of cheese and aseptic

cheesemaking techniques. Neth Milk Dairy J31,120-133

Visser FMW (1977b) Contribution of enzymesfrom rennet, starter bacteria and milk to pro-teolysis and flavour development in Goudacheese. 2. Development of bitterness andcheese flavour. Neth Milk Dairy J 31, 188-209

Visser FMW (1977c) Contribution of enzymesfrom rennet, starter bacteria and milk to pro-teolysis and flavour development in Goudacheese. 3. Protein breakdown: analysis ofthe soluble nitrogen and amino acid nitrogenfractions. Neth Milk Dairy J 31, 210-239

Visser FMW (1977d) Contribution of enzymesfrom rennet, starter bacteria and milk to pro-teolysis and flavour development in Goudacheese. 5. Some observations on bitter ex-tracts from aseptically made cheeses. NethMilk Dairy J 31, 265-276

Visser S (1981) Proteolytic enzymes and theiraction on milk proteins. A review. Neth MilkDairy J 35, 65-88

Visser FMW, de Groot-Mostert AEA (1977) Con-tribution of enzymes from rennet, starter bac-teria and milk to proteolysis and f1avourde-velopment in Gouda cheese. 4. Proteinbreakdown: a gel electrophoretical study.Neth Milk Dairy J 31, 247-264

Visser S, Slangen KJ, Hup G (1975) Some bitterpeptides from rennet-treated casein. A meth-od for their purification, utilizing chromato-graphic separation on silica gel. Neth MilkDairy J 29, 319-334

Visser S, Hup G, Exterkate FA, Stadhouders J(1983a) Bitter f1avour in cheese. 2. Modelstudies on the formation and degradation ofbitter peptides by proteolytic enzymes fromcalf rennet, starter ceIls and starter cell frac-tions. Neth Milk Dairy J 37, 169-180

Visser S, Slangen KJ, Hup G, Stadhouders J(1983b) Bitter f1avourin cheese. 3. Compara-tive gel-chromatographic analysis of hydro-phobic peptide fractions from twelve Gouda-type cheeses and identification of bitter pep-tides isolated from a cheese made withStreptococcus cremoris strain HP. Neth MilkDairyJ37,181-192

Visser S, Slangen KJ, Hup G, Exterkate FA,Stadhouders J (1983c) The bitter f1avourde-fect in cheese; some chemical and microbio-

raw milk quality tests. J Food Prot 41, 356-360


logical aspects. Neth Milk Dairy J 37, 250-251

White CH, Marshall RT (1973a) Heat stable pro-tease from Pseudomonas f1uoreseens P26degrades ultra-high temperature pasteurizedmilk. J Dairy Sei 56, 624

White CH, Marshall RT (1973b) Reduction ofshelf-life of dairy produets by a heat-stableprotease from Pseudomonas f1uoreseensP26. J Dairy Sei 56, 849-853

White CH, Gillis WT, Simmler DL, Galal MK,Walsh JR, Adams JT (1978) Evaluation of

Windlan H, Kosikowski FV (1956) The influenceof milk-eoagulating enzymes of nonanimal or-igin upon the f1avorand constituents of Ched-dar cheese. J Dairy Sei 39, 917-918

Yates AR, Irvine OR, Cunningham JD (1955)Chromatographie studies on proteolytie bac-teria in their relationship to f1avour develop-ment in Cheddar eheese. Can J Agrie Sei 35,337-343

bitter flavour in dairy products. i. a review of the

Documents