J . Chem. Tech. Biotechnof..

1984,34B, 182-189

Functional Properties of Proteins in Foods“

Peter Wilding, Peter J. Lillford and Joe M. Regenstein

Unilever Research, Colworth House, Sharnbrook, Bedford MK44 1LQ

(Manuscript received 16 May 1984 and accepted 24 May 1984)

The measurement of functionality of protein food ingredients has developed

somewhat haphazardly, probably due to the wide range of proteins used as
ingredients and the diversity of foods. Studies of the physiochemical properties of
proteins should enable prediction of a proteins response to process environments
and prove more fruitful than many of the empirical measurements of functionality.
The effects of p H , salt type and concentration on the phase behaviour of the
oilseed globulin and arachin, demonstrates the complexity of protein solubility and the
inadequacies of simple tests that have arisen. Studies of the effects of salts and
conditioning on meat fibres, coupled with measurement of the location of water in
pellets from water holding tests enable the latter to be applied with increased
confidence. Comparison of the endothermic transitions observed on heating with
the development of storage and loss moduli allow the contributions of domains of
skeletal muscle myosin to gel structure to be investigated.
Keywordr: Functionality; protein; solubility; water holding; gelation; arachin; meat;
myosin.

1. Introduction
The term ‘functionality’ implies different concepts to different research scientists and depends
upon their respective field of interest. The term is often used to denote any property of a protein
(or proteins) which affects its use, either as a processing aid, or as a direct contributor of product
attributes. Functionality has been defined as: ‘Any property of a food ingredient, except its
nutritional value, that effects its utilisation’.’ The functional requirements of a protein food
ingredient depend both on the desired product’s attributes and on the ingredient’s behaviour
during the chosen process route.
The functional properties normally associated with proteins as food ingredients are: colour,
flavour, texture, smoothness, turbidity, solubility, swelling, gelling, water holding, syneresis,
viscosity, emulsification, foaming, stabilisation, elasticity, grittiness, chewiness, adhesion, fibre
formation and extrudability.
The descriptions are far from scientific: most relate to perceived properties which cannot be
related easily to the physical or chemical properties of the protein ingredient. Most research into
protein functionality is directed within commodity or product types, rather than in respect to
general function. Thus, comparison across systems is generally difficult. The functionality of a
protein in a food system must be a composite expression of its physicochemical properties. These
in turn arise from the protein’s amino acid composition, amino acid sequence, secondary, tertiary
and quaternary structure. Some physicochemical properties of proteins are surface charge,
hydrophobicity, thermal stability, molecular weight, shape, and association/dissociation be-
haviour. Measurement of a range of physicochemical parameters of a protein should allow
important aspects of its behaviour under process conditions to be predicted.
’ Based on a paper presented at the symposium New rechniques for the processing and modification of proteim on 11 April 1984,
organised by the Biotechnology Group, Society of Chemical Industry.

182
Fimctfonal properties of proteins in foods 183

2. Functional properties of proteins

2.1. Solubility of proteins
Solubility is probably the most important of protein functional properties. Many of the other
functional properties (for example, foaming, emulsification, gelation), are affected by solubility.
Protein solubility is dependent upon the molecular weight and therefore associationtdissociation
behaviour, the number of exposed ionic sidechains and hydrophobicity:

A(25 c, atmos press) ==== A(25 c, atmos press, ph, I)

crystal solution

Thermodynamic solubility is independent of the route to achieve it, but over practical timescales
this is not so for proteins. Alternative definitions, such as the protein dispersibility index and the
nitrogen solubility index have been suggested.
The term solubility implies criteria about the nature of the solid phase. Generally for protein
solubility there is always a protein poor phase (the solution) and a protein rich phase, which may
take a variety of forms.
2.1.1. Solubility of the oil-seed globulin, arachin
Solutions of arachin show liquid-liquid phase separations under certain conditions of ionic
strength and p R 2 Pseudo-equilibrium phase diagrams show upper, lower and closed boundaries
corresponding to salting in, salting out and an unexplained solubility behaviour. Figures 1 and 2
show the effect of ionic strength on phase behaviour and composition of arachin solutions.
Coincident studies using gel permeation chromatography and sedimentation velocity studies
demonstrate that arachin undergoes a concentration-dependent self-association, and produces
oligomers that may confer thermodynamic instability, thus bringing about phase separation.

3.0

0 10 20 30 40 50 0 10 20 30 40 50
Protein concentration ( g 100 g - ’ in phase)

Figure 1. The effect of ionic strength on phase hehaviour and composition of arachin solutions. (a) Protein concentration
in the phases as a function of ionic strength due to sodium choride ( X ) and sodium sulphate (0)at pH 5.5 Ringed point
indicates a single phase system. (b) Effect of pH in sodium sulphate; pH 6.0 (O),pH 6.5 ( X ) . (c) Variation of phase composition
with ionic strength in sodium chloride at pH 6.9 ( x ) and pH 7.5 (0).Ringed points indicate single phase systems. (d)
Results at pH 6.0 (0).The broken line (---) indicates failure to reach equilibrium. A duplicate determination (0) gave a
single phase system. From Tombs et al.’
184 P. Wilding ef al.

Gravitational
separation
/ Collision

p'6 . .o. Disproportionation I

1 Coo lescence

Figure 2. Basic processes in the breakdown of emulsions: a diagrammatic representation. From H a l l i ~ ~ g . ~

Unfortunately, the full situation is complex. Self-association behaviour can be influenced by

the history of exposure to conformational perturbants. The solubility of globulins from the soya
bean demonstrates this3 Soya protein isolates are often prepared by aqueous extraction at about
pH 7, followed by subsequent concentration by isoelectric precipitation. Typically only about
50% of this precipitate is soluble in salt, however, if the protein is treated with salt prior to
acidification, the solubility is maintained. Table 1 shows the effect of order of salting and
acidification on the solubility of soya bean globulins. The losss of solubility was shown to occur
by acid-induced non-covalent intermolecular self association.
2.2 Emulsification
This is an area that is extremely difficult to investigate experimentally, particularly in
compositions relevant to food. The information available derives from detailed studies on model
systems of very dilute proteins, and from purely empirical studies using functional tests such as
emulsifying capacity and stability. Often it is difficult to separate the effects of protein on
emulsification, and the stability of the protein to heat, shear, etc. It is generally agreed that
soluble proteins are surface active, because of their hydrophilic and hydrophobic side chains, and
because proteins exhibit aqueous solubility, they promote oil-in-water emulsions. It would be
reasonable to expect the ability of proteins to form or to stabilise emulsions to be an expression
of their physico-chemical make up: in other words, of their amino acid composition, sequence,
hydrophobicity and conformational stability. The latter might be expected to affect the rate and
extent of denaturation at watertoil interfaces.

Table 1. The effect of order of addition of reagents on the solubility of soya bean
globulins in 0.65~ NaCl
Acidification to NaCl to 0.65~
Aqueous pH 4.5 then NaCl then acidification
extract to 0.65~ to pH 4.5
Soluble fraction 23.1 11.6 23.1
Insoluble fcaction - 11.0 -
Solubility is expressed as mg protein soluble per lOOmg extracted flour.
Functional properties of proteins in foods 185

-
.-C
0)
+ I5Pb’
En
Figure 3. (a) Dispersibility curves of the 11s globu-
lins determined by the method of Fan and So~ulski’~
with slight modifications. 0: glycinin; W: legumin
(pea); A: legumin (field bean). (b) Emulsifying capac- 5
ity of the 11s globulins at pH 7.5 obtained by a
modification of the method of Inklaar and Fortuin.”
From Wright’ (reproduced with permission).
Soya Field Pea
bean

Figure 3 shows basic processes involved in the breakdown of emulsions. Dissolved or

suspended protein can affect all rate processes via modification of bulk rheology. This is
particularly so for meat systems: because of the high concentrations of protein employed, the
aqueous phase often exhibits yield stress which imparts droplet immobility and thus stabilises the
system. Suspended particles transmit shear during mixing, and therefore reduce oil droplet sizes,
thus apparently increasing stability. Probably the most popular functional tests for emulsification
has been the emulsion capacity (EC) test. In this, oil is added to a protein solution or dispersion
with rapid stirring, until an end point is reached. This is normally defined as a sudden decrease in
viscosity, visual change or as a drop in conductivity. E C is defined as volume of oil per unit
protein weight, at the end point. In fact, oil phase volumes are always found to be in the region
of 0.65 to 0.85. This corresponds to the phase volumes where droplet distortion must occur to
enable further droplet inclusion. A coherent protein film requires about 5mg protein per ml.
Using this value, careful study of reported values show most can be recalculated to show the
disperse phase volume is in the range 0.65-0.85. The wide variation in reported EC are then
usually due to arbitrarily selected protein concentration^.^
A better approach is to relate similarities in functional performance to similarities in origin or
molecular structure. The 11s globulins from soya, field bean and pea have similar structural and
 P. Wilding et ul.

‘-Or
I
t
r. I Figure 4. Effect of post-

1
mortem storage (at 4”C), on
I 0.2501 potassium iodide in-
duced swelling of individual
fibres of rabbit longksimur

t
dorsi. Fibres diameters are ex-
pressed as diameter in 0 . 2 5 ~
potassium iodide relative to
8. the original diameter in 0.15~
1.01 H
i potassium chloride. The verti-
cal bar represents the mean
I I I I I I I I I. I standard deviation.
0 10 20 30 40 50 60 70 80 90 100
Time post rnortem ( h )

physicochemical proper tie^,^ amino acid composition, thermal stability, hydrophobicity and
associationldissociation behaviour. They would therefore be expected to have similar emulsifying
capacities and other functional properties. Wright’ demonstrated this to be so. Figure 4 shows
the dispersibility curves and EC values of the three purified proteins.
2.3. Water-holding in meat
In the meat science area, many tests have been developed to determine the function of water
binding or water holding.6 By studying the physicochemical behaviour of the muscle protein
complex, it is possible to demonstrate the fallibilities of many of the empirical tests, and in fact to
predict the response to process conditions. One widely-used approach is to determine the
salt-induced water holding capacity.’ This test involves comminution of the meat, salt treatment
and centrifuging, with determination of change in mass of the meat test material.
Transverse proton relaxation n.m.r.’ has indicated that salt-induced swelling of meat occurred
via imbibition of water into the fibres. Tables 2 and 3 show water-holding potential mesurements
(by the method of Regenstein and Rank Stamm7) and the relaxation times of water in the

Table 2. Salt swelline of muscle tisue

Salt swelling (%)

Species Water 0 . 2 5 ~KCI 0 . 2 5 ~KI

Cod 107 89 148
Chicken 88 96 109
Beef 69 74 81
Values quoted are water holding potential measurements by
the method of Regenstein and Rank Stamm.’

Table 3. Proton spin-spin relaxation times of water in swollen tissue

Relaxation time (msec)

Species Control Water 0 . 2 5 ~KCI 0 . 2 5 ~KI

Cod 52 73 13 157
Chicken 49 - 63 67
Beef 53 - 58 73
Proton spin-spin relaxation times were obtained by the method of Lillford,
Clark and Jones.8
Functional properties of proteins in foods 187

Figure 5. Proton spin-spin

relaxation time distributions*
of water in pellets from salt-
induced water holding tests’ on ( d )
haddock mince. All four sam-
ples had water holding poten-
tial values of 80+_7%. (a) and
(b) 0 . 2 5 ~KCI induced; 0 and
22 days storage at -15°C. (c)
and (d) 0 . 2 5 ~KI induced; 0
and 22 days storage at -15°C.

swollen tissues respectively. Microscopy has confirmed this. The present authors found, using
rabbit longissirnus dorsi fibres, that post-mortem conditioning greatly influenced the response of
fibres to salt solutions. Figure 5 depicts the effect of post-mortem storage on the response of
fibres to salt. Periods exist when hypertonic salt solutions do not induce swelling of intact fibres.
The action of collagenase, or mechanical damage, eliminated the early no-swelling period,
indicating that swelling is controlled by the mechanical distensibility of the collagenous
endomysium. Determination of the myofilament lattice spacings by low angle X-ray diffraction
showed that changes in these dimensions could account for the increased volume in salt solution.
It is thus apparent that the classical water-holding or swelling tests of muscle reflect both the
ability of the myofibrils to imbibe water and the degree of mechanical damage imparted during
the comminution stage of the test procedure. Salt-induced water holding can therefore be
represented by the following relationship:
W~Sf(t).V+Smf(~).(l-v)
where: S’(t) is the swelling ability of fibres at time t post rnortern; Smf(t) is the swelling ability of
myofibrils and damaged fibres at time t ; V is the fraction of meat present as fibres in the
comminute.
The general case can be represented by:
w =ZSi(t>.Vi
where i is a measure of breakdown in the comminute.
Water-holding tests, involving centrifugal or pressure expression of water, are also liable to
misguide the investigator, as no allowance for the mechanism of waterholding is incorporated in
it. While studying the water holding of fish flesh using T2 n.m.r., the authors discovered another
mechanism by .which water may be retained in the water-holding pellet. Figure 6 shows the T2
distributions in the 0 . 2 5 ~KI water-holding pellets of fresh and frozen deteriorated cod. They
have similar water holding values, although other techniques (solubility, thermal stability)
indicated a marked difference in their response to salt. The water distribution shows that the way
188 P. Wilding et al.

1.85mg rabbit skeletal muscle myosin (13% by wt in 0 . 9 6 ~

0 025-
KCI,0 . 0 5 ~potassium phosphate pH 6.0) was heated at 10°C
min-' in a Perkin-Elmer DSC I1 by the method of Wright and
I I I I I Wilding.'3

Temperature ("C)

the deteriorated sample retained water in the centrifuged pellet, was not by expansion of the
myofilament lattice, but in interfragment spaces. These spaces were maintained by mechanical
resistance to compressive forces or by the ability to recover from compression. Hence,
water-holding tests should always assess the mechanism by which the water is held in relation to
the structure. Provided these factors are considered, better identification of the value of
ingredients should be obtained.
2.4. Gelation
Gelation is one of the most important of functional properties in foods, and apart from a few
examples (i.e. gelatin), one of the least .understood. Gelation is the formation of a three-
dimensional molecular network, which confers solid-like properties (i.e. high storage modulus).
At high concentrations (about 20%), all protein dispersions and some solutions exhibit this
phenomenon. Two stages are known to be involved in heat gelation of proteins:
1. Thermal denaturation; this does not involve complete unfolding, only 'loosening' of the
structure.
2 . Reassociation into strands: multiple bond formation is involved, and the balance is highly
specific to protein type. Ribonuclease only gels after prolonged heating above its denaturation
temperature, when disulphide bonds form. Ovalbumin, serum albumin and soya glycinin gels are
dominated by weak forces. These have marked pH and ionic strength dependence which affects
kinetics of strand formation and final gel properties. Information on the mechanism of network
formation is now Information is also emerging on the gelation of more complex

Figure 7. Effect of temperature on storage (G') and loss

( G ) moduli of myosin (34mg ml-' in 0 . 9 6 ~potassium
chloride, 0 . 0 5 ~potassium phosphate pH 6.0). Results were
determined using a Rheometrics Mechanical Spectrometer, at
a heating rate of 1°C min-', frequency 10 rad s-', strain 10%.
30 40 50 60 70 80 90
Temperature ("C
Functional properties of proteins im f d s 189

proteins such as skeletal muscle, myosin and actomyosin.l1*l2The thermal unfolding of domains
of the myosin molecule have been followed using differential scanning ca10rimetry.l~ At pH 6
and an ionic strength of 1.0, myosin undergoes three endothermic transitions on heating (Figure
7). These have been assigned to the helical tail, the globular head and the so-called hinge
region.13 Comparison of the endothermic transitions observed with the development of storage
and loss moduli (Figure 8), allow the contribution of the myosin domains to gel structure to be
investigated.

References
1. Kinsella, J. E. Crit. Rev. Food Sci. Nutr. 1976, 7, 219-280.
2. Tombs, M. P.; Newsom, B. G.; Wilding, P. Int. J. Pept. Res. 1974, 6, 253-277.
3. Lillford, P. J.; Wright, D. J. J . Sci. Food Agric. 1981, 32, 315-327.
4. Halling, P. J. Crit. Rev. Food Sci. Nufr. 1981, 15, 155-203.
5. Wright, D. J. Qual. Plant. Foods Hum. Nutr. 1983, 32, 389-400.
6. Hamm, R . A d v . Food Res. 1960. 10, 355.
7. Regenstein, J. M. and Rank Stamm, J. J. Food Biochem. 1979,3, 212-221.
8. Lillford, P. J.; Clark, A. H.; Jones, D. V. ACS (Am. Chem. SOC.) Symp. Ser. 1980, 127, 117-195.
9. Clarke, A. H.; Judge, F.J.; Richard, J. B.; Stubbs,, J. M.; Sugget, A. Int. J . Peptide Protein Res. 1981, 17, 380-392.
10. Clark, A. H. Prog. Food Nutr. Sci. 1982, 6, 149.
11. Ishioroshi, M.; Samejima, K.; Yasui, T. J. Food Sci. 1979, 44, 1280-1284.
12. Yasui, T.; Ishioroshi, M.; Nankano, H.; Samejima, K. J . Food Sci. 1979, 44, 1210-1211.
13. Wright, D. J.; Wilding, P. J. Sci. Food Agric. In press.
14. Fan, T. Y.; Sosulski, F. W. Can. Inst. Food Sci. Technol. 1974, 7,256-259.
15. Inklaar, P. A,; Fortuin, J. Food Technol., 1969, 23, 103-107.