FOAMING PROPERTIES OF PROTEINS.pdf

Bài báo khoa học

FOAMING PROPERTIES OF PROTEINS: EVALUATION OF A

-ABSTRACT

A specially designed small-scale foaming apparatus was used to de-

termine dynamic and static foaming properties of proteins Foam

was produced by sparging nitrogen at a known rate through a dilute

protein solution The temperature of the protein solution and pro-

tein foam was maintained by a water-jacketed column Ovalbumin

was used as the reference protein Foaming properties (foaming ca-

pacity, foam strength, and stability) were improved when the pro-

tein concentration was increased (O.Ol-0.1%); when sodium chlo-

ride was added to the protein solution and when the temperature

was decreased from 40 to 2°C Foaming properties were optimum at

pH 3.8-4.0, slightly below the isolelectric point of ovalbumin, and

at a gas flow of 20 mi/min The apparatus permits many of the

variables affecting foaming properties of proteins, i.e pH, tempera-

ture, ions, carbohydrates and surfactants to be controlled while

quantitatively determining foaming properties of small quantities of

protein

INTRODUCTION THE FOOD INDUSTRY is constantly seeking new protein

ingredients for use in food manufacture Many proteins are

used for specific applications because of their particular

functional properties (Kinsella, 1976) An important func-

tional requirement of proteins used in angel food cake,

whipped toppings, divinity and souffle-like products is the

capacity to form stiff stable foams In evaluating new pro-

teins as foaming agents an objective method for measuring

foam formation and foam stability is needed Reports in

the literature describe three dynamic procedures for deter-

mining foaming capacity of proteins viz whipping, shaking

or sparging (Cumper, 1953; Yasumatsu et al., 1972) One

important difference between these methods is the amount

of protein required for foam production The amount of

protein ranges from 3-40% for whipping, around 1% for

shaking and ranges from 0.0 l-2% for gas sparging (Lawhon

and Cater, 1971; Yasumatsu et al., 1972; Buckingham,

1970) Another difference between these methods is the’

manner in which the foam is formed While shear forces are

involved in all three methods, they are most important in

whipping and apparently of little importance in the sparg-

ing process Foam bubbles are formed and broken by shear

in both whipping and shaking In the sparging method once

the foam is formed the rate of rupture of the bubble is a

function of the lamella thickness and interfacial viscoelas-

ticity (Cumper, 1953; Mita et al 1977) Whipping, the most

commonly used method, produces protein foams that can

be measured by the increase in foam volume, specific grav-

ity and/or viscosity (Eldridge et al., 1963; McKeller and

Stadelman, 195 5; Lawhon and Cater, 197 1) Rapid shaking

of a horizontal graduated cylinder containing a protein so-

lution (1%) produces a foam that can be measured by its

volume (Yasumatsu et al., 1972; Graham and Phillips,

Authors Waniska and K’insella are with he Dept of Food Science,

Institute of Food Science, Cornell Univeniry, Ithaca, NY 14853

0022-1147/79/0005-1398$02.25/O

01979 Institute of Food Technologists

Fig l-Lower portion of foaming apparatus illustrating sparging disk (SD), protein solution (PS) and protein foam (PFI in a water jacketed column Buffer solution is being injected in the rubber septum with a graduated hypodermic syringe

1976; Wang and Kinsella, 1976) The sparging of gas into a

amounts of protein (O.Ol-2%) Foaming capacity can be measured by the ratio of the volume of gas in foam to the volume of gas sparged, or by the maximum volume of foam divided by the gas flow rate (Buckingham, 1970; Mangan, 1958; Cumper, 1953; Mita et al., 1977)

The stability of protein foams is usually measured by the volume of liquid drained from a foam during a specific time

at room temperature (Eldridge et al., 1963; McKelier and

APPARA TUS TO DETM PROTEIN FOAMING PROPER TIES

Table 1-A summary of the independent parameters used far experi-

ments with the modified foaming apparatus

Nature

Fixed Diameter of sparging disc 1 O cm

Porosity of sparging disc 4-8~

Volume of protein solution 15.0 ml

Volume of column for foam 70.0 ml

Mass of weight (six-2.4 m m holes) 25.09

Stadelman, 1955; Mita et al., 1977) Occasionally the de-

crease in foam volume over time (Wang and Kinsella, 1976)

is used to measure foam stability Dynamic methods that

have been employed to measure foam stability include the

rate of fall of a perforated weight through a column of

foam (Mangan, 19.58; Buckingham, 1970), the penetration

of a penetrometer cone (McKeller and Stadelman, 1955) or

the ability to support a series of specific weights (IAPI,

1956)

Thus a variety of methods have been used to produce

and characterize protein foams The large amount of pro-

tein required in the whipping method limits its usefulness in

testing experimental proteins and furthermore the incorpor-

ation of air increases the temperature of the precooled pro-

tein solution during whipping This affects observed foam-

ing properties The amount of protein required for shaking

is suitable for most experimental situations but the rapid

motion quickly increases the temperature of the protein

solution and the volume of foam produced is limited by the

container The sparging apparatus of Buckingham (1970)

can be modified to maintain the protein solution and foam

column at the desired temperature and can be adapted to

require small amounts of protein

As is well documented in the literature the foaming char-

acteristics of proteins are markedly influenced by condi-

tions of preparation, measurement, etc (Eldridge et al.,

1955; Yasumatsu et al., 1972; McKeller and Stadelman,

1955; Kinsella, 1976) Because of the variety of methods

employed it is difficult to compare data from different

sources

There is a need for a practical standardized method for

determining foaming properties of food proteins Ideally

such methods should provide a measure of the physico-

chemical properties related to foaming A standard method

for testing foaming properties based on the method of

Foulk and Miller (1931) was proposed by Balmaceda et al

(1976) Using a modification of this apparatus we systemat-

ically studied the effects of temperature, gas flow rate, pro-

tein content, pH, salt and sugar concentration on the foam-

ing properties of ovalbumin under controlled conditions

MATERIALS & METHODS CRUDE OVALBUMIN (Sigma Chem Corp., lot #94(3-0247) was

purified by precipitation from saturated NH, SO,, dialyzed, and

then lyophilized The foaming apparatus was assembled from avail-

able laboratory equipment, a specially designed glass stopper and a

machined brass weight The principle component of the foaming

apparatus was a calibrated water-jacketed glass condenser (85 cm

long; 1.10 cm i.d.) with a 24/40 joint at one end (Fig 1 and 2) A

Forma Constant Temperature Circulator (cat no 2095) maintained

the temperature in the condenser within 0S”C of the desired tem-

perature (2-38°C) A speciaUy designed glass stopper (Fig 1) which

permitted the introduction of gas and liquid was made in the Glass

Shop (Dept of Physics, Cornell Univ., Ithaca, NY) The gas entry

port had a fritted glass disc (1.0 cm diameter; 4-8~ porosity; Ace

Glass Co #9436-10) on the inside and a two-way stopcock on the

outside A Perkin Elmer Flow Controller (Serial No GC1773 1) reg-

ulated the gas flow rate of prepurified nitrogen The liquid port was

Fig 2-Upper portion of foaming apparatus illustrating a column of protein foam (PFI with a perforated brass weight (SW) penetrating the foam

sealed with a rubber septum and protein solutions were introduced with a hypodermic syringe Foam strength was measured by a brass weight (Fig 2) which was made at the machine shop (Dept of Physics, Cornell Univ., Ithaca, NY) The brass weight (25.Og; 1.0 cm diam; 4.6 cm long; six 2.4 m m holes lengthwise) was attached to a nylon monofilament line to retrieve it after passage through the foam

The volume of the sparging chamber (15.0 ml) allowed about 0.4 set for the sparged gas to rise to the air-liquid interface This time is required for the mass transfer i.e adsorption of protein from the bulk solution to the gas/liquid interface (Bull, 1972)

The procedure for measuring foaming properties which takes about 30 min per sample is described The water-jacketed column and buffer solution (50 m M sodium citrate) are equilibrated to the specified temperature (Table 1) The protein is dissolved in the buf- fer solution with stirring for 5 min to give a concentration of O.Ol-1.00% (w/v) The protein solution (15.0 ml) is then injected into the sparging chamber via the septum stoppered inlet Nitrogen gas is sparged into the protein solution until the foam chamber (70 ml) is filled with foam while simultaneously maintaining the volume

of liquid in the sparging chamber by the addition of buffer or pro- tein solution The time required to form 70 ml of foam and the volume of buffer added to the sparging chamber are recorded After

4

,.~~- FOAM STRENGTH ,z ~

GR5 FLOW RRTE <ML/MIN>

Fig 3 Effect of gas flow rate and pH on foam strength (see/ml) at

1s”c

10 mm the volume of liquid drained from the foam is recorded The

strength of the foam is then determined by observing the time and

distance of fall of the brass weight through the foam (Fig 2) The

measured variables were employed to calculate several parameters

(Table 2) used to predict foaming properties

Experimental design

The characteristics of the foaming apparatus were studied in

three experiments In the first experiment three independent para-

meters of the apparatus were varied in a central composite design

(Co&ran and Cox, 1957): gas flow rate (4.0, 10.0,20.0,30.0,36.0

ml/min), temperature (7.5, 15.0, 22.5, 30.0, 37S”C), and pH (2.5,

3.0, 4.5, 6.0, 6.5) The second experiment was conducted to deter-

mine more precisely the effect of pH on foaming properties at the

optimum levels of gas-flow rate and temperature as indicated by the

results from experiment 1: A protein content of 0.10% (w/v) was

used in the first two experiments The third experiment was con-

ducted JO optimize the conditions necessary for determining foam-

ing properties of ovaltamin Five factors were varied in a central

composite design in this experiment Values for these factors were:

protein content [O.OlO, 0.032, 0.100,0.316, 1.00% (w/v)], sucrose

(0, 25, 50, 75, 100 mM), sodium chloride (0, 100, 200, 300, 400

Table Z-Equations employeb to calculate parameters used to pre-

dict foaming orooerties

where

Gi =

FR =

Tf =

Vi =

Vd =

Yr =

DlO r

FS =

Tw =

Did =

Gi = (lOO)(FR)(Tf) 70-Vi

Vr = Vi - Vd

D1O = lloo)w)

Vi FS=Tz

percent of gas initially in 70 ml of foam

(gas Flow Rate) - ml/min of nitrogen

time to fill the column with foam

volume of liquid in the foam initially

volume of liquid drained from the foam after IO min

volume of liquid retained in the foam after 10 min

percent of the liquid drained from the foam after IO min

(Foam Strength) - rate of fall of a weight through the

foam (seclmll

time for the weight to fall through the foam (secl

ml of foam penetrated

mM), temperature (2, 10, 20, 30, 38°C) and pH (3.1, 3:8,4.5,5.2, 5.9) while the gas flow rate (20 ml/min) was held constant Statistical analyses were performed with a Minitab II program (Penn State Univ., 1976) at the Cornell University Computing Cen- ter Statistical significance was determined by “t-test” (t = bij/SD) where SD is standard deviation A probability level of 0.01 was used throughout this study

RESULTS & DISCUSSION

THE FIRST EXPERIMENT was conducted to optimize the independent parameters (Table 1) of the foaming appara- tus Results of statistical analyses provided regression coef- ficients along with their level of significance (Table 3) The effect of two independent parameters, pH and gas flow rate, on foam strength while holding the third parameter constant is illustrated in Figure 3 The optimum foam strength (FS) was observed around a gas flow rate (FR) of

20 ml/min and a pH of 5.0 At least three response surface graphs for each parameter [(percent of gas initially in the foam (Gi), volume of liquid retained in the foam after 10

mm (Vr), percent of the liquid drained from the foam after

10 min (DlO), and foam strength, the rate of fall of a weight through the foam, set/ml (FS)] were plotted and evaluated (graphs not shown) Evaluation of the graphs dnd significant regression coefficients are summarized below Variation in FR (Table 3) significantly affected Gi, Vr and DlO At a fast FR more liquid is initially entrapped in the foam and less gas is lost during sparging The large amount of liquid initially in the foam was attributed to the inadequate time for the foam lamellae to reach equilibrium with the liquid, hence there was increased drainage At a slow FR the foam had a longer period to drain before the column was completely filled with foam, thereby decreas- ing the amount of liquid initially in the foam Also, the longer period allowed more gas to be lost from the foam The optimum gas flow rate (20 ml/min) represents a com- promise between these effects

Variation in pH (Table 3 and Fig 3) significantly af-

slightly below the isoelectric point (IEP) of ovalbumin (IEP 4.6-4.8) Near the IEP of proteins there is a decreased electrostatic repulsion between protein (Kitchener and Cooper, 19.59; Cumper, 1953; Kinsella, 1976) resulting in closer packing of protein at the air/liquid interface and in- creased viscosity of the adsorbed protein layer (Mita et al.,

Table 3-Regression coefficients of surface response model for foaming characteristics of ovalbumin obtained by varying gas flow rate, pH and tempera turea

Model

a Coefficients that differ significantly are labelled + or - to indicate whether the T-ratio was larger or smaller than the absolute value

bf 3.0 (p < 0.01):

bThe numbers refer to subscr/pts of p (estimators) in the surface responsemodel: Y =&+B,x, +&x2+P3x1 +PL,x~,+&zX:t

P ,3~: +p12~,x2 +p13x,x3 +&3~I~J whereO=mean value, 1

= gas flow rate, 2 = pH and 3 = temperature

APPARATUS TO DETM PROTEIN FOAMING PROPERTIES

Fig 6-Effect of pH and protein content on drainage while holding

the temperature at lO”C, sucrose cone at 50 m M and salt cone at

200 mM

“k - llE3 -+ ?ir + 20 Lf0

E R S FLUW RRTE <ML/MIN>

Fig 4-Effect of gas flow rate on the rate constant of drainage of Fig 5-Effect of pH on foam strength at 10°C 0.10% protein and a ovalbumin foam prepared at 22°C and pH 4-5 gas flow rate of 20 ml/min

1977) Thus, increased protein-protein interaction near the

IEP gave stronger protein films that were less permeable to

the entrapped gas

Variations in temperature (Table 3) significantly af-

fected FS Results of the surface response design indicated

that 15’C was the best temperature, but the optimum tem-

perature was probably below this value Improved foaming

properties at a lower temperature may be related to the

increased viscosity of the liquid phase and a decreased rate

of protein denaturation (Buckingham, 1970; Mita et al.,

1977)

The rate of drainage of liquid from many foams obeyed

the equation: V = Vo exp(-kt) where V is the volume of

liquid in the foam and k is the rate constant of drainage

(Mita et al., 1977) When log V vs time is plotted, the slope

5UCRtl’iE !i0+ 5RLT i!00mfi T E M P IEC

.ElEl

P H OF PROTEIN SOLUTIDN

3

EFFECT OF P H ON F O A M STRENGTH

P H OF PROTEIN 6~LlJTI~N

of the straight line is equal to k For gluten foams, the rate

of drainage at the beginning and the end of the foam decay were respectively, greater and less than those calculated by this equation (Mita et al., 1977) The rate constant of drain- age was estimated from Vo and Vr, in experiment 1, to com- pare our results with those of Mita et al (1977) Slower FR’s resulted in protein foams with a lower rate constant (Fig 4), indicating a more stable foam Ovalbumin foams near the IEP had the lowest drainage rate constant and percent drainage after 10 min Only at a FR of 10 ml/min or less did temperature affect the drainage rate constant At a flow rate of 10 ml/min foam stability decreased with a rise in temperature A linear relationship was obtained by plotting log k vs l/T and an activation energy of 1.3 Kcal was calcu- lated A higher activation energy (3.6 Kcal) was calculated from drainage constants of foams prepared from wheat glu- ten in 3M urea (Mita et al., 1977)

Closer evaluation of the effect of pH on foam character-

F O A M STRENGTH

PROTEIN ,316 %

SALT CONC <mu>

I

Fig 7-Effect of salt concentration and pH on foam strength while holding the temperature at 3O”C, sucrose cone ar 25 m M and pro- tein con tent 0.3 76%

istics in the second experiment had slightly different re-

sults The effect of pH on FS at a FR of 20 ml/min, a Table 4-Regression coefficients of surface response model for temperature of 10°C and a protein concentration of 0.10%

foaming characteristics of ovalbumin obtained by varying sucrose cone, salt cone, pH, temperature and protein con ten ta

(w/v) is shown in Figure 5 While a quadratic model fit the Model

had a correlation coefficient of 0.92 However, the opti-

mum value of foam strength was accurately estimated by

both models Optimum values for DlO and Vr were at-

tained around pH 4.0-4.7 while variation in pH did not

affect Gi

The third experiment was designed to determine the ef-

fect of five independent variables on foaming properties of

ovalbumin using a surface response model Results of statis-

tical analyses provided regression coefficients (Table 4) for

the quadratic models along with their level of significance

The effect of pH and protein content on DlO (Fig 6) indi-

cates that at the lower level of pH (3.8) and a higher pro-

tein concentration (0.3 16%) optimum retention of liquid in

the foam occurred The effect of salt concentration and pH

on FS (Fig 7) illustrates that at the higher level of salt (300

mM) and at the lower value of pH (3.8) optimum FS was

observed

Ten graphs (not shown herein) for each foaming charac-

teristic were plotted and evaluated to determine the opti-

mum level for each independent variable The results are

summarized below Variation in sucrose concentration up

to 3.44% (w/v) did not significantly affect any measured

foaming property The data revealed trends which showed

that a sucrose level of 50 mM (1.72%) produced foams with

improved properties

55 -1.085-y -0.417

0.784 -1.285 1.492 0.63 1.432 -0.093 -7.48 -0.507 0.403 2.285 0.110 0.589 -2.13 1.43 0.75

-0.21 -1.27 -2.65 1.27 5.36 -1.99 4.94 0.84 -7.50 -6.22 4.20 10.69+ -6.40

4.99

-7.34 4.43 15.8

a Coefficients that differ significantly are labelled + or -to indicate whether the T-ratio was larger or smaller than the absolute value

of 3.0 (p < 0.01)

The stability of wheat gluten foams was not affected

with sucrose concernrations less than 12% (w/v); however,

increased foam stability resulted with 30 and 50% sucrose

(Mita et al., 1977) To stabilize foams made from egg white

or protein isolates sucrose concentrations as high as 45%

have been utilized (Eldridge et al., 1955; Lawhon and Cater,

1971) Thus for significant effects greater sucrose concen-

trations need to be tested in this system

bThe numbers refer the subscripts of p (estimators) in the surface response model: Y = po + p, x, + &x, + &x3 + p4xq + p,xs +

P I 1 4 +&2x: +&3x: +&4x: +&5x: +P,*xIx* +P,3x,xB +

P 14X1X4 +LJ15x,x5 + P 23%X3 + 024 x2x4 +? 25X2X4 +&4x3% + P35x3x5 +P45 x x5 where 0 = mean value, 1 = sucrose cone, 2 = 4 salt cone, 3 = pH, 4 = temperature and 5 = protein content

Variation in sodium chloride concentration significantly

affected Vr (Table 4 and Fig 7) The interaction of protein

content and salt concentration significantly affected Vr and

FS (Table 4) At higher salt levels (2 300 mM) improve-

ment in foaming properties was probably the result of a

more dense protein film (Mita et al., 1977) In the presence

of salt, protein solutions have a lower surface tension Salt

causes a more compact protein conformation at an air/liq-

uid interface resulting in a minimum surface area of the

adsorbed protein molecules and a viscous, adsorbed film

(Mita et al., 1978; Kitchener and Cooper, 1959)

lustrated for DIO in Figure 5 Similar results were reported

by Buckingham (1970) At a protein level of 0.312% oval- bumin produced foams composed of small bubbles The treatment at 30°C and 300 mM salt at this high protein level showed very good foam properties but after several minutes the foam solidified or gelled The treatment at 10°C and 300 mM salt at the high protein level also showed very good foaming properties Small stable bubbles formed quickly and rose in the foam column; however, there was not a clear liquid/foam boundary, hence the volume of buf- fer added was approximated

Variation in pH significantly (p < 0.05) affected FS(Ta-

ble 4) At pH 3.8-4.0, slightly below the IEP of ovalbumin,

optimum foaming properties were observed (Fig 6 and 7)

Near the IEP proteins lose their effective electrical charge,

resulting in minimal electrostatic repulsion and the most

compact conformation Under these conditions more pro-

tein orients at the air/liquid interface because the molecules

pack closer together The thickness of a film of bovine

serum albumin significantly increased as the pH approached

the IEP (Musselwhite and Litchener, 1967) The surface

viscosity and elastic potential of the protein film also in-

crease in the isoelectric zone (Cumper and Alexander,

1950; Biswas and Haydon, 1962)

From the data obtained in this initial study the foaming apparatus can be used to evaluate the important foaming characteristics of protein i.e the foaming capacity of pro- tein and the stability and strength of the foam Foaming ability i.e Gi denotes the amount of gas retained in the foam reflecting the potential of the protein for foaming The measures of foam stability, Vr and DlO, quantitatively reflects the distribution of liquid in the foam and foam strength FS measures the dynamic stability or resistance of the foam to external force or pressure

SUMMARY

Variations in temperature significantly affected D 10 (Ta-

ble 4) At 10°C less liquid drained from the foam probably

because of a slight increase in viscosity and a decreased rate

of denaturation of ovalbumin (Buckingham, 1970; Mita et

al., 1977; Cumper, 1953)

measured foaming properties As protein content of the

solution increased, each measured property improved as il-

THE MODIFIED APPARATUS described herein produced foam under controlled conditions Small quantities (0.10%

or 15 mg) of protein produced foams that could be validly evaluated for Gi, Vi, VJ, DlO and FS Levels of 20 ml/min gas flow rate and 10 C produced foams with optimum properties The foaming properties of ovaibumin were im- proved by increasing protein content Increased salt concen- tration improved foam strength but was associated with increased drainage Foaming properties were optimum slightly below the,JEP of ovalbumin

WATER MOBILITY IN FLOUR DOUGHS

Table 5-Effect of additive on T, of hard wheat flour dough

Flour

Immobile fraction Mobile fraction

T,, (ms) g H, O/g D S T,h (ms) g H, O/g D S

Effect of additives

Salt (sodium chloride) and potassium iodate have been

shown to affect the dough properties (Bloksma, 1973) The

effect of these two additives on Tz of hard wheat flour

doughs is shown in Table 5 No change in Tza and Tzb of

water was observed with the addition of 50 ppm potassium

iodate On the other hand, the addition of 2% salt short-

ened the relaxation time of the mobile fraction The re-

duced water mobility is expected since salt is known to

stiffen the dough (Bloksma, 1973) Salt might also promote

solubilization of some dough components, resulting in re-

REFERENCES

AOAC 1975 “Official Methods of Analysis.” Association of Offi-

286: 16

tion of Cereal Chemists

Digest 40: 38

differential thermal analysis 1 Dough studies using the boiling

differential thermal analysis 2 Dough studies using the ,el&g

differential thermal analysis 3 Bread studies using the melting

the interaction of water with gluten and starch in bread Cereal

REFERENCES

Agric 21: 441

Interface Sci 41: 305

335, 2nd ed John Wiley & Sons, Inc., NY

Trans Faraday Sot 49: 1360

of proteins Trans Faraday Sot 46: 235

dustr Eng Chem 23: 1283

teins at the air-water interface and their role in stabilizing foams

In “Foams,” Ed R.J Akers p 237 Academic Press, London

Cooke, R and Kuntz, I.D 1974 The properties of water in biologi- cal systems Ann Rev Biophys Bioeng, 3: 95

freezable water in dough Chemy Ind., P 1138

binding J Food Technol 4: 117

Baker’s Digest 46(2): 32

378

Sot: 56: 78A

flour crude pentosans and their relation to mixing characteristics

of dough Cereal Chem 48: 211

DC

Food Sci 41: 297

measuiing nuclear relaxation times Rev SC; Instr 29: 688

Triticale flours Cereal Chem 50: 16

wheat flour doughs J Food Technol 5: 65

sonal Communication

Technol 7: 183

ance relaxation studies of adsorbed water on silica gel J Phys Chem 62: 1157

M S received l/5/79: revised 3117179; accepted 3125179

ter Washington State Univ., Pullman, W A 99164

The authors acknowledge the use of the pulsed N M R spectrome-

Laboratory (USDA) for their cooperation

This research was supported in part by a grant-in-aid from Wash- ineton State Universitv

survey Crit Rev Food Technol 7: 219

and pH of precipitation on the yields and functional properties

372

140

Interface Sci 59: 172

ents J Colloid Interface S C & 64: 143

I55 -

of protein films J Colloid Interface Sci 24: 80

Wang and Kinsella J.E 1976

M S received l/12/79; revised 4/16/79: accepted 4123179