ted labuza, peng zhou**, xiaoming liu**laurie davis* & amy

Protein aggregation and hardening of whey based nutritional bars during storage

Ted Labuza, Peng Zhou**, Xiaoming Liu**Laurie Davis* & Amy TranDepartment of Food Science & Nutrition

University of Minnesota**Jiangnan Univ. PRC

This research was supported by the USDA, DMI

& the Davisco Co* LeSueur MN

QuickTime™ and a decompressor

are needed to see this picture.

Typical Protein Bars

òNutrition Factsè150 – 300 caloriesè25 – 40g proteinè10 – 30g CHOè0 – 5g fat

èBodybuildersèSpecial protein blend

òNutrition Factsè150 – 300 caloriesè25 – 40g proteinè10 – 30g CHOè0 – 5g fat

èBodybuildersèSpecial protein blend

ò Est. $1 billion in sales by 2010

ò 35% sales growth from 2001-2006è Initial success followed

by maturation

U.S. sales of nutrition and energy bars, 2001-06

U.S. sales of nutrition and energy bars, 2001-06

Example protein nutritional bar

ò Moisture ~ 15%(WB) 18%(DB)ò aw ~ 0.55

ò Proteinò Up to 40% of total weight

ò Protein sourcesò Whey protein (WPI/WPC/WPH)ò Caseinate/caseinò Milk protein (MPI/MPC) ò Soy protein (SPI/SPC)ò Gelatin or hydrolysates

ò Sugar/sugar alcoholsò Glycerol (glycerine)ò Maltitol and maltitol syrupò Oligofructoseò maltodextrinò Xylitolò Sorbitol

òMechanisms of quality loss during storageèTexture changes - hardening (major complaint)èMaillard ReactionèTaste and flavorèLoss of nutritional value

òMechanisms of quality loss during storageèTexture changes - hardening (major complaint)èMaillard ReactionèTaste and flavorèLoss of nutritional value

è Bar model A test formula4 35% WPI4 25% Corn syrup4 25% HFCS4 10% Peanut butter4 5% glycerol

è Moisture: ~15%è aw: ~0.55è Q10 ~ 3.6 è Note initial jump presumed due

to redistribution of water& humectants into protein particles as aw equilibration occurs but continues at slower rate in long storage

0

2000

4000

6000

8000

10000

12000

0 5 10 15 20 25 30Storage (days)

35 C

23 C

Unknown 1: What hardness level is unacceptable ?

Unknown 2: What causes the hardening ?

Problem - hardening of nutritional bar during storage

Possible mechanisms for hardening

òPossible mechanisms for hardeningè1. Moisture & humectant redistribution into

protein particles.

è2. Protein-protein interactionsàaggregation by S-S

è3. Humectant effects on mobility, local viscosity/glass transition, and protein interaction

è4. Maillard reaction effects on color & texture

òPossible mechanisms for hardeningè1. Moisture & humectant redistribution into

protein particles.

è2. Protein-protein interactionsàaggregation by S-S

è3. Humectant effects on mobility, local viscosity/glass transition, and protein interaction

è4. Maillard reaction effects on color & texture

Why does hardening occur?

0

3000

6000

9000

12000

0 5 10 15 20 25 30Storage (days)

A B Cè Stage A: Redistribution of

water/humectants in matrix and particles after mixing

è EVIDENCE : slight reduction in water activity over 1st 3 to 5 days

Commercial bar model A stored at 35 degree C

QuickTime™ and aTIFF (LZW) decompressor

are needed to see this picture.

Possible solution: Grind protein particles to as fine as possible (danger - explosion possibility)

1. Initial redistribution of water/humectants

What do we know about protein aggregation?

ie proteins are used as inhalation drugs and emulsifers

Aggregation reactions have Q10 ~ 3 to 4

Example rbST stability @ 47°CHageman et al J Ag. Food Chem 40:342 1992

Note 1. rate 5 g/100 g = 0.5%/day @ 47°C; est. <0.11% /day @23°C or ~40% in one year for very dry system based on Q14 =4.7

Note 2. suggested lysine -NH2 on one chain reacting with -CHO of carbonyl of asparagine/glutamine on another chain --> insoluble

Insulin aggregation as a function of water sorption of the protein Langer’s group @ MIT

ò Figure (A).The sorption isotherm at 50°C (m0 is 4.6 g water/100 g protein)

ò Figure (B). Aggregation of insulin after 24 hr at 50°C as f (%RH)

ò Rate at 65% RH &50°C =~20%/day Est. ~ 3.8% loss /day at 35°C and 1%/day @ RT or ~90% in 3 months

ò Constantino proposed both disulfide bond formation and other mechanisms

Costantino, Langer and Klibanov 1994, Pharm Res 11(1): 21

IMF range

Note: In 2007 FDA asked 10 Pharma groups to recall inhalation insulin Stone & Feldman Drug Delivery Technol 2007 7:86-92

2. Protein-protein interactions in whey

ò Driving forcesè Covalent bonding

4Disulfide bonds4NEB like reactions

è Non-covalent interactions4H-bonding, hydrophobic interactions, electrostatic

interactions

ò Driving forcesè Covalent bonding

4Disulfide bonds4NEB like reactions

è Non-covalent interactions4H-bonding, hydrophobic interactions, electrostatic

interactions

β-Lactoglobulin α-Lactalbumin BSA

% of wheyproteins SH S-S

β-Lactoglobulin ~50 1 2

α-Lactalbumin ~20 0 4

BSA ~10 1 17

Example BSA % solubility loss @ 37°Liu et al Biotech Bioeng 37:177 1991

% s

olub

ility

loss

in 2

4 hr

aw ~ 0.8 ~25g H20/100 g solids

aw ~ 0.25 aw ~ 0.6 ~14 g H20/100 g solids

Note 1. water addition done by mixing in directly

Note 2 most bars aw > 0.65 Rate37°C @ 5gH2O/100g = 2%/day Est. Rate 23°C = 0.43%/day or 100% in 8 months

aw ~ 0.9 ~67g H20/100 g solids Rate = 60%/day

Dickinson & McClements work on emulsifier properties of whey

J.Biol Macro 13:26 (1991); J. Food Sci. 58 (1):1037 (1993); JAFC 41:1826 (1993)NEM N-ethylmalemide (blocks -S-S- formation) reducing flocculation

Simplified bar model system WB

òSimplified bar model system (WB): WPI and buffer systemèBioPRO whey protein isolate from Davisco

4Protein, 97.4% on dry basis4Lactose, < 1% of dry basis (minimal Maillard

reaction)4Fat, 0.3% of dry basis ( lipid oxidation minimal)

èPhosphate buffer (10 mM, pH 7)èWPI/buffer: 3/2 (WPI, 60% of the total weight)

40.67:1 water to protein ratio4aw ~ 40% moisture wet basis Upper limit of

semimoist range4 Accelerated shelf life test

èSodium Azide (0.05% of the total weight)

òSimplified bar model system (WB): WPI and buffer systemèBioPRO whey protein isolate from Davisco

4Protein, 97.4% on dry basis4Lactose, < 1% of dry basis (minimal Maillard

reaction)4Fat, 0.3% of dry basis ( lipid oxidation minimal)

èPhosphate buffer (10 mM, pH 7)èWPI/buffer: 3/2 (WPI, 60% of the total weight)

40.67:1 water to protein ratio4aw ~ 40% moisture wet basis Upper limit of

semimoist range4 Accelerated shelf life test

èSodium Azide (0.05% of the total weight)

Extent of whey protein aggregation as f(t,T) in Model WB bar system

0

10

20

30

40

50

60

70

0 20 40 60 80 100 120Storage time (days)

45 degree C

34 degree C

23 degree C

4 degree C

0

10

20

30

40

50

60

70

0 20 40 60 80 100 120Storage time (days)

45 degree C

34 degree C

23 degree C

4 degree C

Measured Q10 aggregation ~ 3.3 from 23 to 45 °C or 14 fold factor

So 1 month at 45°C = 14 months at 23°C thus ~ 2 weeks at 45 °C = ~ 6 months at RT

Rate estimated at 37°C ~0.2%/day and 0.04%/day at 23°C or ~15% loss in one year but bar would need antimicrobials

Aggregation Rate

Aggregate particle formation by SEM

Fresh control Storage at 45 °C for 3 months

(Particle diameter 50~100 nm)

Note interparticle agglomeration

Changes in the conformation of whey protein molecules by DSC

Confirmation by FTIR for WB system

Mechanisms of proteinaggregation

n Solubility of aggregates in various solutions

Solutions Aggregate solubility(%)

Buffer (10 mM, pH 7) 4.4 ± 0.6Buffer with 0.1% SDS 7.2 ± 0.5Buffer with 6 M guanidine HCl 10.9 ± 0.7Buffer with 8 M urea 11.6 ± 1.7Buffer with 10 mM dithiotreitol 92.2 ± 0.9Buffer with SDS and dithiotreitol 97.1 ±1.7

Non-covalent

Covalent(Disulfide bond)

Note: This confirms work of Langer’s group @ MIT

ò Texture measurementèTA-XT2 texture analyzerèPlunger: 3mm diameterèCompression speed: 1mm/sèDeformation strain: 50%èHardness is recorded as the

maximum force during the compression

ò Texture measurementèTA-XT2 texture analyzerèPlunger: 3mm diameterèCompression speed: 1mm/sèDeformation strain: 50%èHardness is recorded as the

maximum force during the compression

Changes in the texture of whey protein bar model system WB

Aggregation and Hardness of whey system WB as f(t)

1

10

100

1 10 100Storage time (days)

1

10

100

1 10 100Storage time (days)

1

10

100

1 10 100Storage time (days)

45 degree C34 degree C23 degree C

(b)

Hardening in textureFormation of aggregates

Fresh bar base

Formation of separate aggregates between tiny particles

Formation of aggregate network

0

10

20

30

40

50

60

70

0 20 40 60 80Protein aggregation (%)

45 C, t(hard)=3 days, 21% aggregates

34 C, t(hard)=30days, 15% aggregates

23 C, t(hard)=60 days, 12% aggregates

Separate aggregates Formation of aggregates network

Suggested primary mechanism for texture change in whey bars due to disulfide bond interactions

3. Potential influence of sugars/polyols (humectants)on hardening in whey bar systems

ò Bar systems need an aw of < 0.75 to prevent microbial growth otherwise antimicrobial agents required

ò Accomplished by replacing water with sugars or sugar alcohols (polyols) as plasticizers in whey systemè Lowers aw and will influence:

4Local viscosity of liquid phase which controls mobility & thus reaction rates (find maxima in aw 0.6 to 0.8 range)

4Tg of system which affects molecular mobility and texture

4Protein conformation4Maillard reaction (browning) if humectant has

reducing groups (HFCS)4Crystallization (graining) if use sucrose to control

Maillard4Sugars as humectants also sweeten the product

masking Maillard reaction flavors

ò Bar systems need an aw of < 0.75 to prevent microbial growth otherwise antimicrobial agents required

ò Accomplished by replacing water with sugars or sugar alcohols (polyols) as plasticizers in whey systemè Lowers aw and will influence:

4Local viscosity of liquid phase which controls mobility & thus reaction rates (find maxima in aw 0.6 to 0.8 range)

4Tg of system which affects molecular mobility and texture

4Protein conformation4Maillard reaction (browning) if humectant has

reducing groups (HFCS)4Crystallization (graining) if use sucrose to control

Maillard4Sugars as humectants also sweeten the product

masking Maillard reaction flavors

0 °C

Te

Tg

Glassy state

rubberystate

Crystal melt line

Tm

Solution

freezing lineice

super sat solution

0 % solids 85% 100 %1 water activity 0.8 0.7 0.6 0.5 0.3 0

State Diagram

Tg dry

-140°C

and

boiling line

vapor

A= soft @ 15%wb

H= very soft @ 10% wb

H

A

T = 23°C

0 0.2 0.4 0.6 0.8 1Water activity

Reaction rate & mobility as f(T-Tg)

Added humectant

Note: softness f(T-Tg) and aw is lowered to reduce reaction rates

Modified from Roos and KarelFood Tech 45(12): 66 1991

0.0

0.2

0.4

0.6

0.8

1.0

0 20 40 60 80 100[Polyol/(Polyol+H2O)]%

GlycerolSorbitolXylitolMaltitolFructoseGlucoseSucrose

Critical micro level

Note 1. PG cannot be measured but slightly better than glycerolNote 2. Organoleptic problems (sweetness, metallic) at those levels Note 3. Physiological problems (anal leakage, Heinz body formation)

PG estimate

Effects of sugar/polyols on lowering of awSystem WB’ (6:4.5 whey:water+polyol ratio)

0

30

60

90

120

150

180

0 20 40 60 80[Polyol/(Polyol+Water)]%

Propylene glycolGlycerolSorbitolMaltitol

(B) 7 days

Effects on texture after 1 week @ 45°C System WB’ (6:4.5 Protein:Polyol+ water ratio)

0

20

40

60

80

100

0 20 40 60 80 100[Polyol/(Polyol+H2O)]%

GlycerolPropylene glycolSorbitolXylitolMaltitolFructoseGlucoseSucrose

Note 1. PG causes significant amounts of protein aggregation Note 2. Other sugar/polyols decrease the protein aggregation

Effects of sugar/polyols on whey protein aggregation after 1 week @ 45°C Model system WB’

Effect of PG on aggregate solubility

Note: suggests some other mechanism possible

Solubility of protein aggregates in WB’

w 30% glycerol

Buffer (10 mM, pH 7) 2.5 ± 0.4 1.7 ± 0.8

Bufferwith0.1% SDS 6.8 ± 0.7 1.7 ± 0.2

Bufferwith6 M guanidineHCl 8.7 ± 0.9 6.2 ± 1.1

Bufferwith8 M urea 11.3 ± 0.5 10.3 ± 0.3

Bufferwith10 mM DTT 83.1 ± 3.3 29.1 ± 1.7

Bufferwith0.1% SDS, 8M ureaand 10 mM DTT 99.1 ± 0.4 97.9 ± 4.0

Solubility of protein aggregates in WB

% soluble

DSC of protein bar WB’ w/wo added humectants

(I = 1/3 of humectant in 4.5 grams plasticizer II = 50%)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

15 25 35 45 55 65 75 85 95 105

Temperture (Degree C)

WB

WG-I

WG-II

WM-I

WM-II

WS-I

WS-II

WP-I

WP-II

Peak 1 Peak 2

Peak 2'

Peak 2'

(A)

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

15 25 35 45 55 65 75 85 95 105

Temperture (Degree C)

WB

WG-I

WG-II

WM-I

WM-II

WS-I

WS-II

WP-I

WP-II(B)

Freshly prepared

Prop Glycol

sorbitol

maltitol

glycerol

control

Prop Glycol

sorbitol

maltitol

glycerol

control

* Note propylene glycol 50% system stored at 23 *C was also fully denatured

stored one week @ 45 °C

FTIR of control and bar with PG Model WB

160016201640166016801700Wavenumber (cm-1)

WB of 0 dayWB of 7 days at 45 CWP-II of 0 dayWP-II of 7 days at 45 C

Beta-strandBeta-sheetAlpha-helix

Intermolecular aggregation

Control- freshControl 7 days @ 45°C

Control- freshControl 7 days @ 45°C

PG - freshPG 7 days @ 45°C

4. Maillard reaction in the whey bar

ò Driving forces èWhey proteins are rich in lysine (>10 g/100 g protein) è The presence of reducing sugars eg. fructose/glucoseè The aw of most nutritional bars is between 0.65 ~ 0.75,

in the reactive range for Maillard reaction

0 0.2 0.4 0.6 0.8 1Water activity

Early Study

òModel system :òModel system :

ò Initial water activity: 0.78ò FDNB methodò Half life at 35°C for lysine loss = 20

days

ò Initial water activity: 0.78ò FDNB methodò Half life at 35°C for lysine loss = 20

days

18.53g/100 g

solidWater

20Microcrystalline

cellulose

20Apiezon B oil

30Whey protein

20Glycerol

10Glucose

0.3K-sorbate

%Ingredient

Schnickels, Warmbier, Labuza: Effect of Protein Substitution on Nonenzymatic Browning in an Intermediate Moisture Food System, Journal of Agricultural and Food Chem33istry, 1976

Non enzymatic browning in protein bars with HFCS and different WPI/Soy ratios at aw ~ 0.65

Soy has less available lysine

But we need an all dairy protein solution

System A

35% protein

25% corn syrup

25% HFCS

Peanut butter 10%

Glycerol 5%

Choice of humectant System WB’ASLT at 45 °C for 7 Days

Imperative: Don’t use reducing plasticizers

Maillard browning problem: reducing sugars

Pre-trial – color changes

òHFCS-CS bars stored at 35 °C & aw =0.65ò Model system A’ 35% WPI 25% corn syrup 25% HFCS 5% glycerol & 10% shortening or

same but Maltitol substituted for CS-HFCS

òHFCS-CS bars stored at 35 °C & aw =0.65ò Model system A’ 35% WPI 25% corn syrup 25% HFCS 5% glycerol & 10% shortening or

same but Maltitol substituted for CS-HFCS

0 1 2 3 4 5

Weeks of storage time for CS and HFCS system L value 89 73 62.5 56.5 52.5 48

Note that the maltitol system after 18 weeks reached an L value of 80

HFCS Model system C Q10=3.6

Shelf Life Plot Q10 = 3.6

A’

Model A’ Maltitol System No HFCS Color

Model System A’ HFCS Texture Q10= 3.6

Model System A’ HFCS Texture Q10= 3.6

Model System A’ Maltitol Texture Change

NEB effects on protein quality @ aw = 0.65 & 35°Cchemical(OPA ortho-phthaldialdehyde) vs biological

method comparison

OPA

Tetrahymena assay

Lysine Loss in Model C

HOW to make and keep a bar soft and maintain high nutritional quality ?

ò Potential solutions:

è1. Add reducing reagents and thiol-blocking reagents if allowed

è2. Add whey protein hydrolysates (Lowering the Tg of bar system)

è3. Control the types and ratio of humectants

ò Potential solutions:

è1. Add reducing reagents and thiol-blocking reagents if allowed

è2. Add whey protein hydrolysates (Lowering the Tg of bar system)

è3. Control the types and ratio of humectants

ò Try to slow down protein aggregation and hardeningè Reducing reagents: Cysteine, glutathioneè Thiol blocking reagents: N-ethylmaleimide

Model ID Protein (6g) Buffers (4g) FunctionB WPI 10 mM phosphate buffer ControlC-L WPI 0.06 M L-cysteine ReducingC-H WPI 0.45 M L-cysteine ReducingG-L WPI 0.06 M L-glutathione ReducingG-H WPI 0.45 M L-glutathione ReducingE-L WPI 0.06 M N-ethylmaleimide Thiol blockingE-H WPI 0.45 M N-ethylmaleimide Thiol blocking

System B 6:4 protein:buffer ratio no humectant aw ~ 0.97

1 . Reducing or thiol blocking reagents

30 day @ 45°C equivalent to ~ 1.3 year at 23 °C

Storage of whey system B WPI bar at 45 °C

2. Addition of protein hydrolysates

0

0.1

0.2

0.3

0.4

0 0.2 0.4 0.6 0.8 1

Water activity

WPIGAB line of WPIH1GAB line of H1H2GAB line of H2H3GAB line of H3

(A) 23 degree C

0

0.1

0.2

0.3

0.4

0 0.2 0.4 0.6 0.8 1Water activity

WPIGAB line of WPIH1GAB line of H1H2GAB line of H2H3GAB line of H3

(B) 45 degree C

H1= 5.2% hydrolyzed H2= 8.8% H3= 14.9%

Hydrolysate effectiveness is not related to greater water holding capacity

Hardness development at 45 °C Day 7 vs Day 0

Whey Hydrolysate model system B aw ~ 0.97

Hydrolysate substitution lowers Tgso softer based on T-Tg

WPI: y = -47.051Ln(x) + 205.86R2 = 0.9643

H1: y = -54.61Ln(x) + 199.22R2 = 0.9948

H2: y = -65.562Ln(x) + 212.89R2 = 0.9975

H3: y = -64.534Ln(x) + 200.85R2 = 0.996

-60

-20

20

60

100

140

1 10 100Water content (g H2O / 100 g solid)

WPIH1H2H3Log. (WPI)Log. (H1)Log. (H2)Log. (H3)

Data below for pure proteinswith water only

Degree of hydrolysis 5.2% 8.8% 14.9%

ò Bar model system C (25% replacement of the WPI with whey protein hydrolysates) aw ~0.6

H=1

H=9L=47

H=1H=1 H=1

H=3L=38

H=5L=41

H=8L=46

0%

H = hardness in Newtons

All had L ~ 90 at day 0

Day 0

Day 7

@ 45°C

System C 35% WPI, 30% corn syrup, 30% HFCS, no fat, 5% glycerol

Whey protein hydrolysates(Davisco)

Lend = 76

Lend = 75 Lend = 82

L0 = 90

Texture Change Model System A’ Maltitol+25% of protein as Hydrolyzed whey protein

System A’

Solution # 3 Replace HFCSwith maltitol

Effect of denaturation Model C

Day 0

Day 7 @45°C

Native Denatured

H=5

H=1

H=30

H=1

Denaturation by mixing 60% Whey:40% water into a dough bake @ 85°C for 1 hr; freeze dry, grind to powder

Makes -SH more exposed for interaction

Solutions reviewed

ò Use ASLT of 4 weeks @ 45°C ~ 1 year @ 23 °C

ò Add reducing agents or thiol blockers to inhibit S-S formationè Work is in progress

ò Partially replace WPI with 25% whey protein hydrolysatesè Makes a softer bar initially thus significantly slowing down

the hardening

ò Control type and ratio of humectantsè No propylene glycolè Eliminate HFCS, use sucrose instead for sweetness or

artificial sweetenerè Use glycerol in combo with sorbitol, maltitol, and xylitolè no browning and good plasticizers

ò Use ASLT of 4 weeks @ 45°C ~ 1 year @ 23 °C

ò Add reducing agents or thiol blockers to inhibit S-S formationè Work is in progress

ò Partially replace WPI with 25% whey protein hydrolysatesè Makes a softer bar initially thus significantly slowing down

the hardening

ò Control type and ratio of humectantsè No propylene glycolè Eliminate HFCS, use sucrose instead for sweetness or

artificial sweetenerè Use glycerol in combo with sorbitol, maltitol, and xylitolè no browning and good plasticizers

Questions and/or comments?

Contact: Dr Ted LabuzaDepartment of Food Science and NutritionUniversity of Minnesota St Paul 55108 USA612-624-9701 fax 625-5272 tplabuza@umn.eduhttp://www.ardilla.umn.edu/Ted_Labuza