Industrial Process food

propyl side chain to the amino group, rotation of the N-Ca bond is not possible, and therefore the f angle has a fixed value of 70°. In addition, since there is no hydrogen at the nitrogen atom, it cannot form hydrogen bonds. Because of these two attributes, segments containing proline residues cannot form a-helices. In fact, proline is considered to be an a-helix breaker. Proteins containing high levels of proline residues tend to assume a random or aperiodic structure. For example, proline residues constitute about 17% of the total amino acid residues in b-casein, and 8.5% of the residues in as1-casein, and because of the uniform distribution of these residues in the primary structures of these proteins, a-helices are not present and the proteins have random structures. However, polyproline is able to form two types of helical structures, termed polyproline I and polyproline II. In polyproline I, the peptide bonds are in the cis-configuration, and in polyproline II they are in trans. Other geometric characteristics of these helices are given in Table 7. Collagen, which is the most abundant animal protein, exists as polyproline IItype helix. In collagen, every third residue is a glycine, which is preceded usually by a proline residue. Three polypeptide chains are entwined to form a triple helix, and the stability of the triple helix is maintained by interchain hydrogen bonds. b-Sheet Structure The b-sheet structure is an extended structure with specific geometries given in Table 7. In this extended form, the C=O and NH groups are oriented perpendicular to the direction of the chain, and therefore hydrogen bonding is possible only between segments (i.e., intersegment), and not within a segment (i.e., intrasegment). The b-strands are usually about 5–15 amino acid residues long. In proteins, two b-strands of the same molecule interact via hydrogen bonds, forming a sheet-like structure known as b-pleated sheet. In the sheet-like structure, the side chains are Pag e 341 oriented perpendicular (above and below) to the plane of the sheet. Depending on the N C directional orientations of the strands, two types of b-pleated sheet structures, namely parallel b-sheet and antiparallel b-sheet, can form (Figure 5). In parallel b-sheet the directions of the b-strands run parallel to each other, whereas in the other they run opposite to each other. These differences in chain directions affect the geometry of hydrogen bonds. In antiparallel b-sheets the N-H … O atoms lie in a straight line (zero H-bond angle), which enhances the stability of the hydrogen bond, whereas in parallel b-sheets they lie at an angle, which reduces the stability of the hydrogen bonds. Antiparallel b-sheets are, therefore, more stable than parallel b-sheets. The binary code that specifies formation of b-sheet structures in proteins is -N-P-N-P-N-P-N-P-. Clearly, polypeptide segments containing alternating polar and nonpolar residues have a high propensity to form b-sheet structures. Segments rich in bulky hydrophobic side chains, such as Val and Ile, also have a tendency to form a b-sheet structures. As expected, some variation in the code is tolerated. The b-sheet structure is generally more stable than the a-helix. Proteins that contain large fractions of b-sheet structure usually exhibit high denaturation temperatures. Examples are b-lactoglobulin (51% b-sheet) and soy 11S globulin (64% b-sheet), which have thermal denaturation temperatures of 75.6 and 84.5°C, respectively. On the other hand, the denaturation temperature of bovine serum albumin, which has about 64% a-helix structure, is only about 64°C [19,20]. When solutions of a-helix-type proteins are heated and cooled, the a-helix is usually converted to b-sheet [19]. However, conversion from b-sheet to a-helix has not been observed in proteins. Another common structural feature found in proteins is the b-bend or b-turn. This arises as a result of 180° reversal of the polypeptide chain involved in b-sheet formation (Fig. 6). The hairpin bend is the result of antiparellel b-sheet formation, and the crossover bend is the result of parallel b-sheet formation. Usually a b-bend involves a four residue segment folding FIGURE 5 Anti-parallel (left) and parallel (rig ht) b-sheets. The dotted lines represent hydrog en bonds between peptide g roups. The arrows indicate N C chain direction. The side chains at Ca atoms are oriented perpendicular (up or down) to the direction of the backbone. (From Ref. 99; courtesy of Spring er-Verlag New York, Inc.) Pag e 342 FIGURE 6 Conformations of type I (a) and type II (b) b-turns. (From Ref. 93; courtesy of Plenum Publishing Corp.) back on itself, and the bend is stabilized by a hydrogen bond. The amino acid residues Asp, Cys, Asn, Gly, Tyr, and Pro are common in b-bends. The secondary structure contents of several proteins are given in Table 8. 6.3.1.3 Tertiary Structure Tertiary structure refers to the spatial arrangement attained when a linear protein chain with secondary structure segments folds further into a compact three-dimensional form. The tertiary structures of b-lactoglobulin and phaseolin (the storage protein in kidney beans) are shown in Figure 7. Transformation of a protein from a linear configuration into a folded tertiary structure is a complex process. At the molecular level, the details for formation of a protein structure are present in its amino acid sequence. From an energetics viewpoint, formation of tertiary structure involves optimization of various interactions (hydrophobic, electrostatic, and van der Waals), and hydrogen bonding between various groups in protein, so that the free energy of the molecule is reduced to the minimum value possible. The most important geometric rearrangement that accompanies the reduction in free energy during formation of tertiary structure is the relocation of most of the hydrophobic residues at the interior of the protein structure and relocation of most of the hydrophilic residues, especially charged residues, at the protein-water interface. Although there is a strong general tendency for hydrophobic residues to be buried in the protein interior, this often can be accomplished only partially. In fact, in most globular proteins, about 40–50% of the water accessible surface is occupied by nonpolar residues [63]. Also, some polar groups are inevitably buried in the interior of proteins; however, these buried polar groups are invariably hydrogen bonded to other polar groups, such that their free energies are minimized in the apolar environment of the protein interior. Pag e 343 Table 8 Secondary Structure Content of Selected Proteins Protein a-Helix (%) b-Sheet (%) b-Turns (%) Aperiodic (%) Deoxyhemog lobin 85.7 0 8.8 5.5 Bovine serum albumin 67.0 0 0 33.0 Chymotrypsinog en 11.0 49.4 21.2 18.4 Immunog lobulin G 2.5 67.2 17.8 12.5 Insulin (dimer) 60.8 14.7 10.8 15.7 Bovine trypsin inhibitor 25.9 44.8 8.8 20.5 Ribonuclease A 22.6 46.0 18.5 12.9 Lysozyme 45.7 19.4 22.5 12.4 Papain 27.8 29.2 24.5 18.5 a-Lactalbumin 26.0 14.0 — 60.0 b-Lactog lobulin 6.8 51.2 10.5 31.5 Soy 11S 8.5 64.5 0 27.0 Soy 7S 6.0 62.5 2.0 29.5 Phaseolin 10.5 50.5 11.5 27.5 Note: Compiled from various sources. The values represent percent of total number of amino acid residues. The folding of a protein from a linear structure to a folded tertiary structure is accompanied by a reduction in interfacial area. The accessible interfacial area of a protein is defined as the total interfacial area of a three-dimensional space occupied by the protein, as determined by figuratively rolling a spherical water molecule of radius 1.4Å over the entire surface of the protein molecule. For native globular proteins, the accessible interfacial area (in Å2 ) is a simple function of their molecular weight, as given by the equation [71] As = 6.3M0.73 (22) The total accessible interfacial area of a nascent polypeptide in its extended linear state (i.e., fully stretched molecule with no secondary, tertiary, or quaternary structure) is also correlated to its molecular weight, M, by the equation [71] At = 1.48M + 21 (23) The initial area (Ab) of a protein that has folded during formation of a globular tertiary structure (i.e., buried area) can be estimated from Equations 22 and 23. The fraction and distribution of hydrophilic and hydrophobic residues in the primary structure affects several physicochemical properties of the protein. For instance, the shape of a protein molecule is dictated by its amino acid sequence. If a protein contains a large number of hydrophilic residues distributed uniformly in its sequence, it will assume an elongated or rod-like shape. This is because, for a given mass, an elongated shape has a large surface are to volume ratio so that more hydrophilic residues can be placed on the surface. On the other hand, if a protein contains a large number of hydrophobic residues, it will assume a globular (roughly spherical) shape. This minimizes the surface-area-to-volume ratio, enabling more hydrophobic residues to be buried in the protein interior. Among globular proteins, it is generally found that larger molecules contain larger fractions of nonpolar amino acids than do smaller molecules. The tertiary structures of several single polypeptide proteins are made up of domains. Domains are defined as those regions of the polypeptide sequence that fold up into a tertiary form independently. These are, in essence, miniproteins within a single protein. The structural Pag e 344 FIGURE 7 Tertiary structures of (A) phaseolin subunit and (B) b-lactog lobulin. The arrows indicate b-sheet strands, and the cylinders indicate a-helix (From Ref. 62a and Ref. 85, respectively). stability of each domain is largely independent of the others. In most single-chain proteins, the domains fold independently and then interact with each other to form the unique tertiary structure of the protein. In some proteins, as in the case of phaseolin (Fig. 7), the tertiary structure may contain two or more distinct domains (structural entities) connected by a segment of the polypeptide chain. The number of domains in a protein usually depends on its molecular weight. Small proteins (e.g., lysozyme, b-lactoglobulin, and a-lactalbumin) with 100–150 amino acid residues usually form a single domain tertiary structure. Large proteins, such as immunoglobulin, contain multiple domains. The light chain of immunoglobulin G contains two domains, and the heavy chain contains four domains. The size of each of these domains is about 120 amino acid residues. Human serum albumin, which is made up of 585 amino acid residues, has three homologous domains, and each domain contains two subdomains [47]. 6.3.1.4 Quaternary Structure Quaternary structure refers to the spatial arrangement of a protein when it contains more than one polypeptide chain. Several biologically important proteins exist as dimers, trimers, tetramers, etc. Any of these quaternary complexes (also referred to as oligomers) can be made up of protein subunits (monomers) that are the same (homogeneous) or different (heterogeneous). For example, b-lactoglobulin exits as a dimer in the pH range 5–8, as an octomer in the pH range 3–5, and as a monomer above pH 8, and the monomeric units of these complexes are identical. On the other hand, hemoglobin is a tetramer made up of two different polypeptide chains that is, a and b chains. Formation of oligomeric structures is the result of specific protein-protein interactions. These are primarily noncovalent interactions, such as hydrogen bonds, and both hydrophobic and electrostatic interactions. The fraction of hydrophobic amino acids appears to influence the tendency to form oligomeric proteins. Proteins that contain more than 30% hydrophobic amino acid residues exhibit a greater tendencey to form oligomeric structures than do those that contain fewer hydrophobic amino acid residues. Formation of quaternary structure is primarily driven by the thermodynamic requirement to bury exposed hydrophobic surfaces of subunits. When the hydrophobic amino acid content of a protein is greater than 30%, it is physically impossible to form a structure that will bury all of the nonpolar residues. Consequently, there is a greater likelihood of hydrophobic patches to exist on the surface, and interaction of these patches between adjacent monomers can lead to the formation of dimers, trimers, etc. (Fig. 8). Many food proteins, especially cereal proteins, exist as oligomers of different polypeptides. As would be expected, these proteins typically contain more than 35% hydrophobic amino acid residues (Ile, Leu, Trp, Tyr, Val, Phe, and Pro). In addition, they also contain 6–12% proline [12]. As a consequence, cereal proteins exist in complex oligomeric structures. The major storage proteins of soybean, namely, b-conglycinin and glycinin, contain about 41% and 39% hydrophobic amino acid residues, respectively. b-Conglycinin is a trimeric protein made up of three different subunits, and it exhibits complex associationdissociation phenomenon as a FIGURE 8 Schematic representation of formation of dimers and olig omers in proteins. Pag e 346 function of ionic strength and pH [76]. Glycinin is made up of 12 subunits, six of the subunits being acidic and the others basic. Each basic subunit is cross-linked to an acid subunit via a disulfide bond. The six acidic-basic pairs are held together in the oligomeric state by non-covalent interactions. Glycinin also exhibits complex association-dissociation behavior as a function of ionic strength [76]. In oligomeric proteins, the accessible surface area, AS, is correlated to the molecular weight of the oligomer [71] by the equation AS = 5.3M0.76 (24) This relationship is different from that which applies to monomeric proteins. The surface area buried when the native oligomeric structure is formed from its constituent polypeptide subunits can be estimated from the equation Ab = At – AS = (1.48M + 21) – 5.3M0.76 (25) where At is the total accessible interfacial area of the nascent polypeptide subunits in their linear state. 6.3.2 Forces Involved in the Stability of Protein Structure The process of folding of a random polypeptide chain into a unique three-dimensional structure is quite complex. As mentioned earlier, the basis for the biologically native conformation is encoded in the amino acid sequence of the protein. In the 1960s, Anfinsen and co-workers showed that when denatured ribonuclease was added to a physiological buffer solution, it refolded to its native conformation and regained almost 100% of its biological activity. A majority of enzymes have been subsequently shown to exhibit similar propensity. The slow but spontaneous transformation of an unfolded state to a folded state is facilitated by several intramolecular noncovalent interactions. The native conformation of a protein is a thermodynamic state in which various favorable interactions are maximized and the unfavorable ones are minimized such that the overall free energy of the protein molecule is at the lowest possible value. The forces that contribute to protein folding may be grouped into two categories: (a) intramolecular interactions emanating from forces intrinsic to the protein molecule, and (b) intramolecular interactions affected by the surrounding solvent. van der Waals and steric interactions belong the former, and hydrogen bonding, electrostatic, and hydrophobic interactions belong to the latter. Steric Strains Although the f and y angles theoretically have a 360° rotational freedom, their values are very much restricted because of steric hindrance from side-chain atoms. Because of this, segments of a polypeptide chain can assume only a limited number of configurations. Distortions in the planar geometry of the peptide unit, or stretching and bending of bonds, will cause an increase in the free energy of the molecule. Therefore, folding of the polypeptide chain can occur only in such a way that deformation of bond lengths and bond angles are avoided. van der Waals Interactions These are dipole-induced dipole and induced dipole-induced dipole interactions between neutral atoms in protein molecules. When two atoms approach each other, each atom induces a dipole in the other via polarization of the electron cloud. The interactions between these induced dipoles have an attractive as well as a repulsive component. The magnitudes of these forces are dependent on the interatomic distance. The attractive energy is inversely proportional to the sixth Pag e 347 power of the interatomic distance, and the repulsive interaction is inversely proportional to the twelfth power of this distance. Therefore, at a distance r, the net interaction energy between two atoms is given by the potential energy function (26) where A and B are constants for a given pair of atoms, and Ea and Er are the attractive and repulsive interaction energies, respectively. van der Waals interactions are very weak, decrease rapidly with distance, and become negligible beyond 6 Å. The van der Waals interaction energy for various pairs of atoms ranges from -0.17 to -0.8 kJ/mol. In proteins, however, since numerous pairs of atoms are involved in van der Waals interactions, the sum of its contribution to protein folding and stability is very significant. Hydrogen Bonds The hydrogen bond involves the interaction of a hydrogen atom that is covalently attached to an electronegative atom (such as N, O, or S) with another electronegative atom. Schematically, a hydrogen bond may be represented as D–H … A, where D and A are, respectively, the donor and acceptor electronegative atoms. The strength of a hydrogen bond ranges between 8.4 to 33 kJ/mol, depending on the pair of electronegative atoms involved and the bond angle. Proteins contain several groups capable of forming hydrogen bonds. Some of the possible candidates are shown in Figure 9. Among these groups, the greatest number of hydrogen bonds are formed between the N-H and C=O groups of the peptide bonds in a-helix and b-sheet structures. FIGURE 9 H-bonding g roups in proteins. (From Ref. 98.) Pag e 348 The peptide hydrogen bond can be considered as a strong permanent dipole-dipole interaction between the Nd- -Hd+ and Cd+ = Od- dipoles as shown: (27) The strength of the hydrogen bond is given by the potential energy function (28) where m1 and m2 are the dipole moments, e is the dielectric constant of the medium, r is the distance between the electronegative atoms, and q is the hydrogen bond angle. The hydrogen bond energy is directly proportional to the product of the dipole moments and to the cosine of the bond angle, and is inversely proportional to the third power of the N … O distance and to the dielectric constant of the medium. The strength of the hydrogen bond is maximum when q is zero (cos 0 = 1), and it is zero when q is 90°. The hydrogen bonds in a-helix and antiparallel b-sheet structures have a q value very close to zero, whereas those in parallel b-sheets have larger q values. The optimum N … O distance for maximum hydrogen bond energy is 2.9 Å. At shorter distances the electrostatic repulsive interaction between the Nd- and Od- atoms causes a significant decrease in the strength of the hydrogen bond. At longer distances weak dipole-dipole interaction between the N-H and C=O groups decreases the strength of the hydrogen bond. The strength of N-H … O=C hydrogen bonds in proteins is typically about 18.8 kJ/mol. The “strength” refers to the amount of energy needed to break the bond. The existence of hydrogen bonds in proteins is well established. Since each hydrogen bond decreases the free energy of the protein by about -18.8 kJ/mol, it is commonly believed that they may act not only as the driving force for protein folding but also may contribute enormously to the stability of the native structure. However, this is not a valid assumption. Because water can compete for hydrogen bonding with N-H and C=O groups in proteins, hydrogen bonding between these groups cannot occur spontaneously, nor can formation of N-H … O=C hydrogen bonds be the driving force for formation of a-helix and b-pleated sheets in proteins. The hydrogen bonding interactions in a-helix and b-sheets are, therefore, a consequence of other favorable interactions that drive formation of these secondary hydrogen-bonded structures. The hydrogen bond is primarily an ionic interaction. Like other ionic interctions, its stability also depends upon the dielectric constant of the environment. The stability of hydrogen bonds in secondary structures is mainly due to a local environment with a low permittivity (low dielectric constant) created by interaction between nonpolar residues. These bulky side chains prevent access of water to the N-H … O=C hydrogen bonds. They are only stable as long as they are protected from water. Electrostatic Interactions As noted earlier, proteins contain several amino acid residues with ionizable groups. At neutral pH, Asp and Glu residues are negatively charged, and Lys, Arg, and His are positively charged. At alkaline pH, Cys and Tyr residues assume a negative charge. Depending upon the relative number of negatively and positively charged residues, Pag e 349 proteins assume either a net negative or a net positive charge at neutral pH. The pH at which the net charge is zero is called the isoelectric pH (pI). The isoelectric pH is different from the isoionic point. The isoionic point is the pH of the protein solution in the absence of electrolytes. The isoelectric pH of a protein can be estimated from its amino acid composition and the pKa values of the ionizable groups using the Hendersen-Hasselbach equation (Eq. 5). With few exceptions, almost all charged groups in proteins are distributed on the surface of the protein molecule. Since at neutral pH proteins assume either a net positive or a net negative charge, one might expect that the net repulsive interaction between like charges would destabilize protein structure. It is also reasonable to assume that attractive interactions between oppositely charged groups at certain critical locations might contribute to the stability of the protein structure. In reality, however, the strength of these repulsive and attractive forces is minimized in aqueous solutions because of the high permittivity of water. The electrostatic interaction energy between two fixed charges q1 and q2 separated by distance r is given by (29) where e is the permittivity of the medium. In vacuum or air (e = 1), the electrostatic interaction energy between two charges at a distance of 3 to 5 Å is about ±460 to ±277 kJ/mol. In water, however, this interaction energy is reduced to ±5.8 to ±3.5 kJ/mol, which is of the order of the thermal kinetic energy (RT) of the protein molecule at 37°C. Therefore, the attractive and repulsive electrostatic interactions between charges located on the protein surface do not contribute significantly to protein stability. However, charged groups partially buried in the protein interior, where the permittivity is lower than that of water, usually form salt bridges with strong interaction energy. The electrostatic interaction energy ranges between ±3.5 and ±460 kJ/mol depending on the distance and the local permittivity. Although electrostatic interactions may not act as the primary force for protein folding, their penchant to remain exposed to the aqueous environment certainly would influence the folding pattern. Hydrophobic Interactions It is obvious from the foregoing discussions that in aqueous solutions the hydrogen bonding and electrostatic interactions between various polar groups in a polypeptide chain do not posses sufficient energy to act as driving forces for protein folding. These polar interactions in proteins are not very stable, and their stabilities depend on maintenance of an apolar environment. The major force driving protein folding comes from hydrophobic interactions among nonpolar groups. In aqueous solutions, the hydrophobic interaction between nonpolar groups is the result of thermodynamically unfavorable interaction between water and nonpolar groups. When a hydrocarbon is dissolved in water, the free energy change (DG) is positive and the volume (DV) and enthalpy change (DH) are negative. Even though DH is negative, meaning that there is favorable interaction between water and the hydrocarbon, DG is positive. Since DG = DH – T DS (where T is the temperature and DS is the entropy change), the positive change in DG must result from a large negative change in entropy, which offsets the favorable change in DH. The decrease in entropy is caused by formation of a clathrate or cage-like water structure around the hydrocarbon. Because of the net positive change in DG, interaction between water and nonpolar groups is highly restricted. Consequently, in aqueous solutions, nonpolar groups tend to aggregate, so that the area of direct contact with water is minimized (see Chap. 2). This water structure-induced interaction between nonpolar groups in aqueous solutions is known as hydrophobic interaction. In proteins, hydrophobic interaction between nonpolar side chains of amino acid residues is the Pag e 350 major reason that proteins fold into unique tertiary structures in which a majority of the nonpolar groups are removed from the aqueous environment. Since the hydrophobic interaction is the antithesis of solution of nonpolar groups in water, DG for hydrophobic interaction is negative, and DV, DH, and DS are positive. Unlike other noncovalent interactions, hydrophobic interactions are endothermic; that is, hydrophobic interactions are stronger at high temperatures and weaker at low temperatures (opposite to that for hydrogen bonds). The variation of hydrophobic free energy with temperature usually follows a quadratic function, that is, DGHf = a + bT + cT2 (30) where a, b, and c are constants, and T is absolute temperature. The hydrophobic interaction energy between two spherical nonpolar molecules can be estimated from the potential energy equation [50] (31) where R1 and R2 are the radii of the nonpolar molecules, D is the distance (nm) between the molecules, and D0 is the decay length (1 nm). Unlike electrostatic, hydrogen bonding, and van der Waals interactions, which follow a power-law relationship with distance between interacting groups, the hydrophobic interaction follows an exponential relationship with distance between interacting groups. Thus, it is effective over relatively long distances, such as 10 nm. The hydrophobic free energy of proteins cannot be quantified using the preceding equation because of involvement of several nonpolar groups. It is possible, however, to estimate the hydrophobic free energy of a protein using other empirical correlations. The hydrophobic free energy of a molecule is directly proportional to the nonpolar surface area that is accessible to water (Fig. 10). The proportionality constant, that is, the slope, varies between 92 J mol-1 Å-2 FIGURE 10 The relationship between hydrophobicity and accessible surface area of amino acid side chains (open circles) and hydrocarbons (filled circles); 1 kcal = 4.18 kJ. (From Ref. 92; courtesy of Annual Reviews, Inc.) Pag e 351 for Ala, Val, Leu, and Phe and 109 J mol-1 Å-2 for Ser, Thr, Trp, and Met. On average, the hydrophobicity of apolar groups in amino acids or amino acid residues is about 100 J mol-1 Å-2. This is close to the 104.5 J mol-1 Å-2 value for alkanes. What this means is that for the removal of every 1 Å2 area of nonpolar surface from the water environment, a protein will decrease its free energy by about 100 J/mol. Thus, the hydrophobic free energy of a protein can be estimated simply by multiplying the total buried surface area by 100 J mol-1 Å-2 . The buried surface area in several globular proteins and the estimated hydrophobic free energies are shown in Table 9. It is evident that hydrophobic free energy contributes significantly to the stability of protein structure. The average hydrophobic free energy per amino acid residue in globular proteins amounts to about 10.45 kJ/mol. Disulfide Bonds Disulfide bonds are the only covalent side-chain cross-links found naturally in proteins. They can occur both intramolecularly and intermolecularly. In monomeric proteins, disulfide bonds are formed as a result of protein folding. When two Cys residues are brought into proximity with proper orientation, oxidation of the sulfhydryl groups by molecular oxygen results in disulfide bond formation. Once formed, disulfide bonds help stabilize the folded structure of proteins. Protein mixtures containing cystine and Cys residues are able to undergo sulfhydryl-disulfide interchange reactions as shown: (32) This interchange reaction also can occur within a single protein if it contains a free sulfhydryl group and a disulfide bond. The interchange reaction often leads to a decrease in stability of the protein molecule. TABLE 9 Accessible Surface Area ( A s), Buried Surface Area ( A b), and Hydrophobic Free Energ y of Proteins Protein Molecular mass (daltons) As (Å2 ) Ab (Å2 ) DG Hf (kJ/mol) Parvalbumin 11,450 5,930 11,037 1108 Cytochrome c 11,930 5,570 12,107 1212 Ribonuclease A 13,960 6,790 13,492 1354 Lysozyme 14,700 6,620 15,157 1521 Myog lobin 17,300 7,600 18,025 1810 Retinol binding protein 20,050 9,160 20,535 2061 Papain 23,270 9,140 25,535 2541 Chymotrypsin 25,030 10,440 26,625 2671 Subtilsin 27,540 10,390 30,390 3047 Carbonic anhydrase B 28,370 11,020 30,988 3110 Carboxypeptidase A 34,450 12,110 38,897 3900 Thermolysin 34,500 12,650 38,431 3854 Note: As values are from Ref. 71. A b was calculated from Equations 22 and 23. Pag e 352 In summary, the formation of a unique three-dimensional protein structure is the net result of various repulsive and attractive noncovalent interactions and a few covalent disulfide bonds. 6.3.3 Conformational Stability and Adaptability of Proteins The stability of the native protein structure is defined as the difference in free energy between the native and denatured (or unfolded) states of the protein molecule. This is usually denoted as DGD. All of the noncovalent interactions discussed already, except the repulsive electrostatic interactions, contribute to the stability of the native protein structure. The stabilizing influence on the native structure of the total free energy changes attributed to these interactions amounts to hundreds of kilojoules per mole. However, the DGD of the majority of proteins is in the range of 20–85 kJ/mol. The major force tending to destabilize the native structure is the conformational entropy of the polypeptide chain. When a random-state polypeptide is folded into a compact state, the loss of translational, rotational, and vibrational motions of various groups of the protein molecule results in a decrease in conformational entropy. The entropy-derived increase in free energy as the protein is folded into its native state is more than offset by favorable noncovalent interactions, resulting in a net decrease in free energy. Thus, the difference in free energy between the native and denatured states can be expressed as (33) where DGH-bond, DGele, DGHf, and DGvdW, respectively, are free energy changes for hydrogen bonding, electrostatic, hydrophobic, and van der Waals interactions, and DSconf is the conformaTABLE 10 DG D Values for Selected Proteins Protein pH T (°C) DG D (kJ/mol) a-Lactalbumin 7 25 18.0 Bovine b-lactog lobulin A + B 7.2 25 31.3 Bovine b-lactog lobulin A 3.15 25 42.2 Bovine b-lactog lobulin B 3.15 25 48.9 T4 Lysozyme 3.0 37 19.2 Hen eg g -white lysozyme 7.0 37 50.2 Gactin 7.5 25 26.7 Lipase (from asperg illus) 7.0 — 46.0 Troponin 7.0 37 19.6 Ovalbumin 7.0 25 24.6 Cytochrome c 5.0 37 32.6 Ribonuclease 7.0 37 33.4 a-Chymotrypsin 4.0 37 33.4 Trypsin — 37 54.3 Pepsin 6.5 25 45.1 Growth hormone 8.0 25 58.5 Insulin 3.0 20 26.7 Alkaline phosphatase 7.5 30 83.6 Note:DG D represents G U – GN, where G U and G N are free energ ies of the denatured and native states, respectively, of a protein molecule. Compiled from several sources. Pag e 353 tional entropy of the polypeptide chain. The DSconf of a protein in the unfolded state is about 8–42 J mol-1 K-1 per residue. Usually, an average value of 21.7 J mol-1 K-1 per residue is assumed. A protein with 100 amino acid residues at 310 K will have a conformational entropy of about 21.7 × 100 × 310 = 672.7 kJ/mol. This destabilizing conformational energy will reduce the net stability of the native structure resulting from noncovalent interactions. The DGD values, that is, energy required to unfold, of various proteins are presented in Table 10. These values clearly indicate that in spite of numerous intramolecular interactions, proteins are only marginally stable. For example, DGD values of most proteins correspond to an energy equivalent of one to three hydrogen bonds or about two to five hydrophobic interactions, suggesting that breakage of a few noncovalent interactions would destabilize the native structure of many proteins. Conversely, it appears that proteins are not designed to be rigid molecules. They are highly flexible, their native state is in a metastable state, and breakage of one to three hydrogen bonds or a few hydrophobic interactions can easily cause a conformational change in proteins. Conformational adaptability to changing solution conditions is necessary to enable proteins to carry out several critical biological functions. For example, efficient binding of substrates or prosthetic ligands to enzymes invariably involves reorganization of polypeptide segments at the binding sites. Proteins that require high structural stability to function as catalysts usually are stabilized by intramolecular disulfide bonds, which effectively reduce conformational entropy (i.e., the tendency of the polypeptide chain to unfold). 6.4 Protein Denaturation The native structure of a protein is the net result of various attractive and repulsive interactions emanating from various intramolecular forces as well as interaction of various protein groups with surrounding solvent water. However, native structure is largely the product of the protein’s environment. The native state (of a single protein molecule) is thermodynamically the most stable with lowest feasible free energy at physiological conditions. Any change in its environment, such as pH, ionic strength, temperature, solvent composition, etc., will force the molecule to assume a new equilibrium structure. Subtle changes in structure, which do not drastically alter the molecular architecture of the protein, are usually regarded as “conformational adaptability,” whereas major changes in the secondary, tertiary, and quaternary structures without cleavage of backbone peptide bonds are regarded as “denaturation.” From a structural point of view, the denatured state of a protein molecule is an ill-defined state. “A major change” in structure may mean an increase in a-helix and b-sheet structure at the expense of random structure or vice versa. However, in most instances, denaturation involves a loss of ordered structure. Depending on the conditions of denaturation, proteins may assume several “denatured states,” each differing only slightly in free energy. This is shown schematically in Figure 11. Some denatured states possess more residual folded structure than others. When fully denatured, globular proteins resemble a random coil. The intrinsic viscosity of a fully denatured protein is a function of the number of amino acid residues, and is expressed by the equation [105] [h] = 0.716n 0.66 (34) where n is the number of amino acid residues in the protein. Often denaturation has a negative connotation, because it indicates loss of some properties. For example, many biologically active proteins lose their activity upon denaturation. In the case of food proteins, denaturation usually causes insolublization and loss of some functional properties. In some instances, however, protein denaturation is desirable. For example, thermal denaturation of trypsin inhibitors in legumes markedly improves digestibility and biological availability of legume proteins when consumed by some animal species. Partially denatured proteins are more digestible and have better foaming and emulsifying properties than do native proteins. Thermal denaturation is also a prerequisite for heat-induced gelation of food proteins. 6.4.1 Thermodynamics of Denaturation Denaturation is a phenomenon that involves transformation of a well-defined, folded structure of a protein, formed under physiological conditions, to an unfolded state under nonphysiological conditions. Since structure is not an easily quantifiable parameter, direct measurement of the fractions of native and denatured protein in a solution is not possible. However, conformational changes in proteins invariably affect several of its chemical and physical properties, such as ultraviolet (UV) absorbance, fluorescence, viscosity, sedimentation coefficient, optical rotation, circular dichroism, reactivity of sulfhydryl groups, and enzyme activity. Thus, protein denaturation can be studied by monitoring changes in these physical and chemical properties. When changes in a physical or chemical property, y, are monitored as a function of denaturant concentration or temperature, many monomeric globular proteins exhibit denaturation profiles as shown in Figure 12. The terms yN and yD are y values for the native and denatured states, respectively, of the protein. For most proteins, as denaturant concentration (or temperature) is increased, the value of y remains unchanged initially, and above a critical point its value changes abruptly from yN to yD within a narrow range of denaturant concentration or temperature. The steepness of the transition curve observed for a majority of monomeric globular proteins indicates that protein denaturation is a cooperative process. That is, once a protein molecule begins to unfold, or once a few interactions in the protein are broken, the whole molecule completely unfolds with a further slight increase in denaturant concentration or temperature. This cooperative nature of unfolding suggests that globular proteins can exist only in the native and denatured states; that Pag e 355 FIGURE 12 Typical protein denaturation curves; y represents any measurable physical or chemical property of the protein molecule that varies with protein conformation; yN and yD are the values of y for the native and denatured states, respectively. is, intermediate states are not possible. This is known as a “two-state transition” model. For this two-state model, the equilibrium between the native and the denatured state in the cooperative transition region can be expressed as (35) where KD is the equilibrium constant. Since the concentration of denatured protein molecules in the absence of a denaturant (or critical input of heat) is extremely low (about 1 in 109 ), estimation of KD is not possible. However, in the transition region, that is, at sufficiently high denaturant concentration (or sufficiently high temperature), an increase in the population of the denatured protein molecule permits determination of the apparent equilibrium constant, Kapp. In the transition region, where both native and denatured protein molecules are present, the value of y is given by y = ƒNyN + ƒDyD (36) where ƒN and ƒD are the fractions of the protein in the native and denatured states, and yN and yD are y values for the native and denatured states, respectively. From Figure 12, ƒN = (yD – y)/(yD – yN) (37) ƒD = (y – yN)/(yD – yN) (38) The apparent equilibrium constant is given by Pag e 356 Kapp = ƒD/ƒN = (y – yN)/(y – yD) (39) and the free energy of denaturation is given by DGapp = -RT ln kapp (40) A plot of -RT ln Kapp versus denaturant concentration in the transition region results in a straight line. The KD and DGD of the protein in pure water (or in buffer in the absence of denaturant) are obtained from the y-intercept. The enthalpy of denaturation, DHD, is obtained from variation of the free energy change with temperature using the van’t Hoff equation, (41) Monomeric proteins that contain two or more domains with different structural stabilities usually exhibit multiple transition steps in the denaturation profile. If the transition steps are well separated, the stabilities of each domain can be obtained from the transition profile by using the preceding two-state model. Denaturation of oligomeric proteins proceeds via dissociation of subunits, followed by denaturation of the subunits. Protein denaturation is reversible. When the denaturant is removed from the protein solution (or the sample is cooled), most monomeric proteins (in the absence of aggregation) refold to their native conformation under appropriate solution conditions, such as pH, ionic strength, redox potential, and protein concentration. Many proteins refold when the protein concentration is below 1 mM. Above 1 mM protein concentration, refolding is partially inhibited because of greater intermolecular interaction at the cost of intramolecular interactions. A redox potential comparable to that of biological fluid facilitates formation of the correct pairs of disulfide bonds during refolding. 6.4.2 Denaturing Agents 6.4.2.1 Physical Agents Temperature and Denaturation Heat is the most commonly used agent in food processing and preservation. Proteins undergo varying degrees of denaturation during processing. This can affect their functional properties in foods, and it is therefore important to understand the factors affecting protein denaturation. When a protein solution is gradually heated above a critical temperature, it undergoes a sharp transition from the native state to the denatured state. The temperature at the transition midpoint, where the concentration ratio of native and denatured states is 1, is known either as the melting temperature Tm, or the denaturation temperature Td. The mechanism of temperature-induced denaturation is highly complex and involves primarily destabilization of the major noncovalent interactions. Hydrogen bonding, electrostatic, and van der Waals interactions are exothermic (enthalpy driven) in nature. Therefore, they are destabilized at high temperatures and stabilized at low temperatures. However, since peptide hydrogen bonds in proteins are mostly buried in the interior, they remain stable over a wide range of temperature. On the other hand, hydrophobic interactions are endothermic (entropy driven). They are stabilized at high temperatures and destabilized at low temperatures. Therefore, as the temperature is increased, the changes in the stabilities of these two groups of noncovalent interactions oppose each other. However, the stability of hydrophobic interactions cannot increase infinitely with increasing temperature, because above a certain temperature, gradual breakdown of water structure will Pag e 357 eventually destabilize hydrophobic interactions as well. The strength of hydrophobic interactions reaches a maximum at about 60–70°C[9]. Another major force that affects conformational stability of proteins is the conformational entropy, -T DSConf, of the polypeptide chain. As temperature is increased, the increase in thermal kinetic energy of the polypeptide chain greatly facilitates unfolding of the polypeptide chain. The relative contributions of the major forces to stability of a protein molecule as a function of temperature are depicted in Figure 13. The temperature at which the sum of the free energies is zero (i.e., KD = 1) is the denaturation temperature of the protein. The Td values of some proteins are listed in Table 11. It is often assumed that the lower the temperature, the greater will be the stability of a protein. This is not always true. For example (Fig. 14), the stability of lysozyme increases with lowering of temperature, whereas those of myoglobin and a mutant phage T4 lysozyme show maximum stability at about 30 and 12.5°C, respectively. Below and above these temperatures, myoglobin and T4 lysozyme are less stable. When stored below 0°C, these two proteins undergo cold-induced denaturation. The temperature of maximum stability (minimum free energy) depends on the relative magnitude of contributions from polar and nonpolar interactions. Those proteins in which polar interactions dominate over nonpolar interactions are more stable at or below refrigeration temperatures than they are at higher temperatures. On the other hand, proteins that are primarily stabilized by hydrophobic interactions are more stable at about ambient temperature than they are at refrigeration temperature. Several food proteins undergo reversible dissociation and denaturation at low temperature. FIGURE 13 Relative chang es in free energ y contributions by hydrog en bonding , hydrophobic interactions, and conformational entropy to the stability of proteins as a function of temperature. Pag e 358 TABLE 11 Thermal Denaturation Temperatures ( T d) and Mean Hydrophobicities of Proteins Protein Td Mean hydrophobicity (kJ mol-1 residue-1) Trypsinog en 55 3.68 Chymotrypsinog en 57 3.78 Elastase 57 Pepsinog en 60 4.02 Ribonuclease 62 3.24 Carboxypeptidase 63 Alcohol dehydrog enase 64 Bovine serum albumin 65 4.22 Hemog lobin 67 3.98 Lysozyme 72 3.72 Insulin 76 4.16 Egg albumin 76 4.01 Trypsin inhibitor 77 Myog lobin 79 4.33 a-Lactalbumin 83 4.26 Cytochrome c 83 4.37 b-Lactog lobulin 83 4.50 Avidin 85 3.81 Soy g lycinin 92 Broadbean 11S protein 94 Sunflower 11S protein 95 Oat g lobulin 108 Source: Data were compiled from Ref. 11. Glycinin, one of the storage proteins of soybean, aggregates and precipitates when stored at 2°C [58], then becomes soluble when returned to ambient temperature. When skim milk is stored at 4°C, b-casein dissociates from casein micelles, and this alters the physicochemical and rennetting properties of the micelles. Several oligomeric enzymes, such as lactate dehydrogenase and glyceraldehyde phosphate dehydrogenase, lose most of their enzyme activity when stored at 4°C, and this has been attributed to dissociation of the subunits. However, when warmed to and held at ambient temperature for a few hours, they reassociate and completely regain their activity [111]. The amino acid composition affects thermal stability of proteins. Proteins that contain a greater proportion of hydrophobic amino acid residues, especially Val, Ile, Leu, and Phe, tend to be more stable than the more hydrophilic proteins [118]. Proteins of thermophilic organisms usually contain large amounts of hydrophobic amino acid residues. However, this positive correlation between mean hydrophobicity and thermal denaturation temperature of proteins is only an approximate one (Table 11), suggesting that other factors, such as disulfide bonds and the presence of salt bridges buried in hydrophobic clefts, may also contribute to thermostability. A strong positive correlation also exists between thermostability and the number percent of certain amino acid residues. For example, statistical analysis of 15 different proteins has shown that thermal denaturation temperatures of these proteins are positively correlated (r = 0.98) to the number percent of Asp, Cys, Glu, Lys, Leu, Arg, Trp, and Tyr residues. On the other hand, thermal denaturation temperatures of the same set of proteins are negatively correlated Pag e 359 FIGURE 14 Variation of protein stability ( DG D) with temperature for myog lobin (…), ribonuclease A (—-), and a mutant ot T4 phag e lysozyme (O). K is the equilibrium constant. (Compiled from Refs. 15 and 62.) (r = -0.975) to the number percent of Ala, Asp, Gly, Gln, Ser, Thr, Val, and Tyr (Fig. 15) [88]. Other amino acid residues have little influence on Td. The underlying causes of these correlations are not clear. It seems, however, that thermostability of proteins is not simply dependent on either polar or nonpolar content, but on an optimum distribution of these two groups in the protein structure. An optimum distribution may maximize intramolecular interactions, decrease chain flexibility, and thus enhance thermostability. Thermostability is inversely correlated with protein flexibility [107]. Thermal denaturation of monomeric globular proteins is mostly reversible. For example, when many monomeric enzymes are heated above their denaturation temperatures, or even briefly held at 100°C, and then are immediately cooled to room temperature, they fully regain their activities [62]. However, thermal denaturation can become irreversible when the protein is heated at 90–100°C for a prolonged period even at neutral pH [2]. This irreversibility occurs because of several chemical changes in the protein, such as deamidation of Asn residues, cleavage of peptide bonds at Asp residues, destruction of Cys and cystine residues, and aggregation [2,109]. Water greatly facilitates thermal denaturation of proteins [37,95]. Dry protein powders are extremely stable to thermal denaturation. The value of Td decreases rapidly as the water content is increased from 0 to 0.35 g water/g protein (Fig. 16). An increase in water content from 0.35 to 0.75 g water/g protein causes only a marginal decrease in Td. Above 0.75 g water/g protein, the Td of the protein is the same as in a dilute protein solution. The effect of hydration on thermostability is fundamentally related to protein dynamics. In the dry state, proteins have a static structure, that is, the mobility of polypeptide segments is restricted. As the water content is increased, hydration and partial penetration of water into surface cavities cause swelling of Pag e 360 FIGURE 15 Group correlations of amino acid residues to thermal stability of g lobular proteins. Group X1 represents Asp, Cys, Glu, Lys, Leu, Arg , Trp, and Tyr. Group X2 represents Ala, Asp, Gly, Gln, Ser, Thr, Val, and Tyr. (Adapted from Ref. 88.) FIGURE 16 Influence of water content on the temperature ( T d) and enthalpy (DH D) of denaturation of ovalbumin. (From Ref. 37.) Pag e 361 the protein. This swollen state presumably reaches a maximum value at a water content of 0.3–0.4 g/g protein at room temperature. The swelling of the protein increases chain mobility and flexibility, and the protein molecule assumes a more dynamic molten structure. When heated, this dynamic flexible structure provides greater access of water to salt bridges and peptide hydrogen bonds than is possible in the dry state, resulting in a lower Td. Additives such as salts and sugars affect thermostability of proteins in aqueous solutions. Sugars such as sucrose, lactose, glucose, and glycerol stabilize proteins against thermal denaturation [3,44]. Addition of 0.5 M NaCl to proteins such as blactoglobulin, soy proteins, serum albumin, and oat globulin significantly increases their Td [19,20,44]. Hydrostatic Pressure and Denaturation One of the thermodynamic variables that affects conformation of proteins is hydrostatic pressure. Unlike temperature-induced denaturation, which usually occurs in the range of 40–80°C at 1 atmospheric pressure (atm); pressure-induced denaturation can occur at 25°C if the pressure is sufficiently great. Most proteins undergo pressure-induced denaturation in the range of 1–12 kbar as evidenced from changes in their spectral properties. The midpoint of pressure-induced transition occurs at 4–8 kbar [48,112]. Pressure-induced denaturation of proteins occurs mainly because proteins are flexible and compressible. Although amino acid residues are densely packed in the interior of globular proteins, some void spaces invariably exist and this leads to compressibility. The average partial specific volume of globular proteins in the hydrated state, vº, is about 0.74 ml/g. The partial specific volume can be considered as the sum of three components: vº=VC + VCav + DVSol (42) where VC is the sum of the atomic volumes, VCav is the sum of the volumes of the void spaces in the interior of the protein, and DVSol is the volume change due to hydration [38]. The larger the vº of a protein, the larger is the contribution of void spaces to partial specific volume, and the more unstable the protein will be when pressurized. Fibrous proteins are mostly devoid of void spaces, and hence they are more stable to hydrostatic pressure than globular proteins. Pressure-induced denaturation of globular proteins is usually accompanied by a reduction in volume of about 30–100 ml/mol. This decrease in volume is caused by two factors: elimination of void spaces as the protein unfolds, and hydration of the nonpolar amino acid residues that become exposed during unfolding. The later event results in a decrease in volume (see Sec. 6.3.2). The volume change is related to the free energy change by the expression DV = d(DG)/dp (43) where p is the hydrostatic pressure. If a globular protein completely unfolds during pressurization, the volume change should be about 2%. However, the experimental value of 30–100 ml/mol volume decrease obtained with pressurized globular proteins corresponds to only about 0.5% volume decrease. This indicates that proteins only partially unfold even at hydrostatic pressure as high as 10 kbar. Pressure-induced protein denaturation is highly reversible. Most enzymes, in dilute solutions, regain their activity once the pressure is decreased to atmospheric pressure. However, regeneration of near complete activity usually takes several hours. In the case of pressure-denatured oligomeric proteins and enzymes, subunits first dissociate at 0.001–2 kbar, and then subunits denature at higher pressures [111]; removal of pressure results in subunit reassociation and almost complete restoration of enzyme activity after several hours. Pag e 362 High hydrostatic pressures are being investigated as a food processing tool, for example, for microbial inactivation or gelation. Since high hydrostatic pressure (2–10 kbar) irreversibly damages cell membranes and causes dissociation of organelles in microorganisms, it will inactivate vegetative microorganisms [46]. Pressure gelation of egg white, 16% soy protein solution, or 3% actomyosin solution can be achieved by application of 1–7 kbar hydrostatic pressure for 30 min at 25°C. These pressureinduced gels are softer than thermally induced gels [82]. Also, exposure of beef muscle to 1–3 kbar hydrostatic pressure causes partial fragmentation of myofibrils, which may be useful as a means of tenderizing meat [102]. Pressure processing, unlike thermal processing, does not harm essential amino acids or natural color and flavor, nor does it cause toxic compounds to develop. Thus, processing of foods with high hydrostatic pressure may prove advantageous (except for cost) for certain food products. Shear and Denaturation High mechanical shear generated by shaking, kneading, whipping, etc. can cause denaturation of proteins. Many proteins denature and precipitate when they are vigorously agitated [81]. In this circumstance, denaturation occurs because of incorporation of air bubbles and adsorption of protein molecules to the air-liquid interface. Since the energy of the air-liquid interface is greater than that of the bulk phase, proteins undergo conformational changes at the interface. The extent of conformational change depends on the flexibility of the protein. Highly flexible proteins denature more readily at an air-liquid interface than do rigid proteins. The nonpolar residues of denatured protein orient toward the gas phase and the polar residues orient toward the aqueous phase. Several food processing operations involve high pressure, shear, and high temperature, for example, extrusion, high-speed blending, and homogenization. When a high shear rate is produced by a rotating blade, subsonic pulses are created and cavitation also occurs at the trailing edges of the blade. Both of these events contribute to protein denaturation. The greater the shear rate, the greater is the degree of denaturation. The combination of high temperature and high shear force causes irreversible denaturation of proteins. For example, when a 10–20% whey protein solution at pH 3.5–4.5 and at 80–120°C is subjected to a shear rate of 7,500–10,000/sec, it forms insoluble spherical macrocolloidal particles of about 1 mm diameter. A hydrated material produced under these conditions, “Simplesse,” has a smooth, emulsion-like organoleptic character [101]. 6.4.2.2 Chemical Agents pH and Denaturation Proteins are more stable against denaturation at their isoelectric point than at any other pH. At neutral pH, most proteins are negatively charged, and a few are positively charged. Since the net electrostatic repulsive energy is small compared to other favorable interactions, most proteins are stable at around neutral pH. However, at extreme pH values, strong intramolecular electrostatic repulsion caused by high net charge results in swelling and unfolding of the protein molecule. The degree of unfolding is greater at extreme alkaline pH values than it is at extreme acid pH values. The former behavior is attributed to ionization of partially buried carboxyl, phenolic, and sulfhydryl groups, which cause unraveling of the polypeptide chain as they attempt to expose themselves to the aqueous environment. pH-induced denaturation is mostly reversible. However, in some cases, partial hydrolysis of peptide bonds, deamidation of Asn and Gln, destruction of sulfhydryl groups at alkaline pH, or aggregation can result in irreversible denaturation of proteins. Pag e 363 Organic Solvents and Denaturation Organic solvents affect the stability of protein hydrophobic interactions, hydrogen bonding, and electrostatic interactions in different ways. Since nonpolar side chains are more soluble in organic solvents than in water, hydrophobic interactions are weakended by organic solvents. On the other hand, since the stability and formation of peptide hydrogen bonds are enhanced in a low-permittivity environment, certain organic solvents may actually strengthen or promote formation of peptide hydrogen bonds. For example, 2-chloroethanol causes an increase in a-helix content in globular proteins. The action of organic solvents on electrostatic interactions is twofold. By decreasing permittivity, they enhance electrostatic interactions between oppositely charged groups and also enhance repulsion between groups with like charge. The net effect of an organic solvent on protein structure, therefore, usually depends on the magnitude of its effect on various polar and nonpolar interactions. At low concentration, some organic solvents can stabilize several enzymes against denaturation [5]. At high concentrations, however, all organic solvents cause denaturation of proteins because of their solubilizing effect on nonpolar side chains. Organic Solutes and Denaturation Organic solutes, notably urea and guanidine hydrochloride (GuHCl), induce denaturation of proteins. For many globular proteins the midpoint of transition from the native to denatured state occurs at 4–6 M urea and 3–4 M GuHCl at room temperature. Complete transition often occurs in 8 M urea and in about 6 M GuHCl. GuHCl is a more powerful denaturant than urea because of its ionic character. Many globular proteins do not undergo complete denaturation even in 8 M urea, whereas in 8 M GuHCl they usually exist in a random coil state (completely denatured). Denaturation of proteins by urea and GuHCl is thought to involve two mechanisms. The first mechanism involves preferential binding of urea and GuHCl to the denatured protein. Removal of denatured protein as a protein-denaturant complex shifts the N D equilibrium to the right. As the denaturant concentration is increased, continuous conversion of the protein to proteindenaturant complex eventually results in complete denaturation of the protein. Since binding of denaturant to denatured protein is very weak, a high concentration of denaturant is needed to cause complete denaturation. The second mechanism involves solubilization of hydrophobic amino acid residues in urea and GuHCl solutions. Since urea and GuHCl have the potential to form hydrogen bonds, at high concentration these solutes break down the hydrogen-bonded structure of water. This destructuring of solvent water makes it a better solvent for nonpolar residues. This results in unfolding and solubilization of apolar residues from the interior of the protein molecule. Urea or GuHCl-induced denaturation can be reversed by removing the denaturant. However, complete reversibility of protein denaturation by urea is sometimes difficult. This is because some urea converts to cyanate and ammonia. Cyanate reacts with amino groups and alters the charge of the protein. Detergents and Denaturation Detergents, such as sodium dodecyl sulfate (SDS), are powerful protein denaturing agents. SDS at 3–8 mM concentration denatures most globular proteins. The mechanism involves preferential binding of detergent to the denatured protein molecule. This causes a shift in equilibrium between the native and denatured states. Unlike urea and GuHCl, detergents bind strongly to denatured proteins, which is the reason complete denaturation occurs at a relatively low detergent concentration of 3–8 mM. Because of this strong binding, detergent-induced denaturation is irreversible. Globular proteins denatured by SDS do not exist in a random coil state; instead, they assume an a-helical rod shape in SDS solutions. This rod shape is properly regarded as denatured. Pag e 364 Chaotropic Salts and Denaturation Salts affect protein stability in two different ways. At low concentrations, ions interact with proteins via nonspecific electrostatic interactions. This electrostatic neutralization of protein charges usually stabilizes protein structure. Complete charge neutralization by ions occurs at or below 0.2 ionic strength, and it is independent of the nature of the salt. However, at higher concentrations (> 1 M), salts have ion specific effects that influence the structural stability of proteins. Salts such as Na2SO4 and NaF enhance, whereas NaSCN and NaClO4 weaken it. Protein structure is influenced more by anions than by cations. For example, the effect of various sodium salts on the thermal denaturation temperature of b-lactoglobulin is shown in Figure 17. At equal ionic strength, Na2SO4 and NaCl increase Td, whereas NaSCN and NaClO4 decrease it. Regardless of their chemical makeup and conformational differences, the structural stability of macromolecules, including DNA, is adversely affected by high concentrations of salts [108]. NaSCN and NaClO4 are strong denaturants. The relative ability of various anions at isoionic strength to influence the structural stability of protein (and DNA) in general follows the series . This ranking is known as the Hofmeister series or chaotropic series. Fluoride, chloride, and sulfate salts are structure stabilizers, whereas the salts of other anions are structure destabilizers. The mechanism by which salts affect the structural stability of proteins is not well understood; however, their relative ability to bind to and alter hydration properties of proteins is probably involved. Salts that stabilize proteins enhance hydration of proteins and bind weakly, whereas salts that destabilize proteins decrease protein hydration and bind strongly [4]. These effects are primarily the consequence of energy perturbations at the protein-water FIGURE 17 Effects of various sodium salts on the temperature of denaturation, T d, of b-lactog lobulin at pH 7.0. , NaCl; , NaBr; , NaClO4; , NaSCN; , urea. (From Ref. 20.) Pag e 365 interface. On a more fundamental level, protein stabilization or denaturation by salts is related to their effect on bulk water structure. Salts that stabilize protein structure also enhance the hydrogen-bonded structure of water, and salts that denature proteins also break down bulk water structure and make it a better solvent for apolar molecules. In other words, the denaturing effect of chaotropic salts might be related to destabilization of hydrophobic interactions in proteins. 6.5 Functional Properties of Proteins Food preferences by human beings are based primarily on sensory attributes such as texture, flavor, color, and appearance. The sensory attributes of a food are the net effect of complex interactions among various minor and major components of the food. Proteins generally have a great influence on the sensory attributes of foods. For example, the sensory properties of bakery products are related to the viscoelastic and dough-forming properties of wheat gluten; the textural and succulence characteristics of meat products are largely dependent on muscle proteins (actin, myosin, actomyosin, and several water-soluble meat proteins); the textural and curd-forming properties of dairy products are due to the unique colloidal structure of casein micelles; and the structure of some cakes and the whipping properties of some dessert products depend on the properties of egg-white proteins. The functional roles of various proteins in different food products are listed in Table 12. “Functionality” of food proteins is defined as “those physical and chemical properties which affect the behavior of proteins in food systems during processing, storage, prepartion and consumption”[55]. The sensory attributes of foods are achieved by complex interactions among various functional ingredients. For instance, the sensory attributes of a cake emanate from gelling/heat-setting, foaming, and emulsifying properties of the ingredients used. Therefore, for a protein to be useful as an ingredient in cakes and other such products, it must possess multiple functionalities. Proteins of animal origin, such as milk (caseins), egg and meat proteins, are widely used in fabricated foods. These proteins are mixtures of several proteins with wide-ranging physico-chemical properties, and they are capable of performing multiple functions. For example, egg white possesses multiple functionalities such as gelation, emulsification, foaming, water binding, and heat coagulation, which makes it a highly desirable protein in many foods. The multiple functionalities of egg white arise from complex interactions among its protein constituents, namely, ovalbumin, conalbumin, lysozyme, ovomucin, and other albumintype proteins. Plant proteins (e.g., soy and other legume and oilseed proteins), and other proteins, such as whey proteins, are used to a limited extent in conventional foods. Even though these proteins are also mixtures of several proteins, they do not perform as well as animal proteins in most food products. The exact molecular properties of proteins that are responsible for the various desirable functionalities in food are poorly understood. The physical and chemical properties that govern protein functionality include size; shape; amino acid composition and sequence; net charge and distribution of charges; hydrophobicity/hydrophilicity ratio; secondary, tertiary, and quaternary structures; molecular flexibility/rigidity; and ability to interact/react with other components. Since proteins possess a multitude of physical and chemical properties, it is difficult to delineate the role of each of these properties with respect to a given functional property. On an empirical level, the various functional properties of proteins can be viewed as manifestations of two molecular aspects of proteins: (a) hydrodynamic properties, and (b) protein surface-related properties [23]. The functional properties such as viscosity (thickening), gelation, and texturization are related to the hydrodynamic properties of proteins, which depend on size, shape, and molecular flexibility. Functional properties such as wettability, dispersibility, solubil- Pag e 366 TABLE 12 Functional Roles of Food Proteins in Food Systems Function Mechanism Food Protein type Solubility Hydrophilicity Beverag es W hey proteins Viscosity W ater binding , hydrodynamic size and shape Soups, g ravies, and salad dressing s, desserts Gelatin W ater binding Hydrog en bonding , ionic hydration Meat sausag es, cakes, and breads Muscle proteins, eg g proteins Gelation W ater entrapment and immobilization, network formation Meats, g els, cakes, bakeries, cheese Muscle proteins, eg g and milk proteins Cohesionadhesion Hydrophobic, ionic, and hydrog en bonding Meats, sausag es, pasta, baked g oods Muscle proteins, eg g proteins, whey proteins Elasticity Hydrophobic bonding , disulfide cross-links Meats, bakery Muscle proteins, cereal proteins Emulsification Adsorption and film formation at interfaces Sausag es, bolog na, soup, cakes, dressing s Muscle proteins, eg g proteins, milk proteins Foaming Interfacial adsorption and film formation W hipped topping s, ice cream, cakes, desserts Eg g proteins, milk proteins Fat and flavor binding Hydrophobic bonding , entrapment Low-fat bakery products, doug hnuts Milk proteins, eg g proteins, cereal proteins Source: Ref. 56. ity, foaming, emulsification, and fat and flavor binding are related to the chemical and topo-graphical properties of the protein surface. Although much is known about the physicochemical properties of several food proteins, prediction of functional properties from their molecular properties has not been successful. A few empirical correlations between molecular properties and certain functional properties in model protein systems have been established [74]. However, behavior in model systems often is not the same as behavior in real food products. This is attributable, in part, to denaturation of proteins during food fabrication. The extent of denaturation depends on pH, temperature, other processing conditions, and product characteristics. In addition, in real foods, proteins interact with other food components, such as lipids, sugars, polysaccharides, and minor components, and this modifies their functional behavior. Despite these inherent difficulties, considerable progress has been made toward understanding the relationship between various physicochemical properties of protein molecules and their functional properties. 6.5.1 Protein Hydration Water is an essential constituent of foods. The rheological and textural properties of foods depend on the interaction of water with other food constituents, especially with macromolecules, such as proteins and polysaccharides. Water modifies the physicochemical properties of proteins. For example, the plasticizing effect of water on amorphous and semicrystalline food proteins changes their glass transition temperature (see Chap. 2) and TD. The glass transition temperature Pag e 367 refers to the conversion of a brittle amorphous solid (glass) to a flexible rubbery state, whereas the melting temperature refers to transition of a crystalline solid to a disordered structure. Many functional properties of proteins, such as dispersibility, wettability, swelling, solubility, thickening/viscosity, water-holding capacity, gelation, coagulation, emulsification, and foaming, depend on water-protein interactions. In low and intermediate moisture foods, such as bakery and comminuted meat products, the ability of proteins to bind water is critical to the acceptability of these foods. The ability of a protein to exhibit a proper balance of protein-protein and protein-water interactions is critical to their thermal gelation. Water molecules bind to several groups in proteins. These include charged groups (ion-dipole interactions); backbone peptide groups; the amide groups of Asn and Gln; hydroxyl groups of Ser, Thr, and Tyr residues (all dipole-dipole interactions); and nonpolar residues (dipole-induced dipole interaction, hydrophobic hydration). The water binding capacity of proteins is defined as grams of water bound per gram of protein when a dry protein powder is equilibrated with water vapor at 90–95% relative humidity. The water binding capacities (also sometimes called hydration capacity) of various polar and nonpolar groups of proteins are given in Table 13. Amino acid residues with charged groups bind about 6 mol water/mol residue, the uncharged polar residues bind about 2 mol/mol residue, and the nonpolar groups bind about 1 mol/mol residue. The hydration capacity of a protein therefore is related, in part, to its amino acid composition—the greater the number of charged residues, TABLE 13 Hydration Capacitiesª of Amino Acid Residues Amino acid residue Hydration (moles H2O/mol residue) Polar Asn 2 Gln 2 Pro 3 Ser, The 2 Trp 2 Asp (unionized) 2 Glu (unionized) 2 Tyr 3 Arg (unionized) 3 Lys (unionized) 4 Ionic Asp- 6 Glu- 7 Tyr- 7 Arg + 3 His+ 4 Lys + 4 Nonpolar Ala 1 Gly 1 Phe 0 Val, Ile, Leu, Met 1 ªRepresents unfrozen water associated with amino acid residues based on nuclear mag netic resonance studies of polypeptide. Source: Ref. 59. Pag e 368 the greater is the hydration capacity. The hydration capacity of a protein can be calculated from its amino acid composition using the empirical equation [59] a = ƒC + 0.4ƒP + 0.2ƒN (44) where a is g water/g protein, and ƒC, ƒP, and ƒN are the fractions of the charged, polar, and nonpolar residues, respectively, in the protein. The experimental hydration capacities of several monomeric globular proteins agree very well with those calculated from the preceding equation. This, however, is not true for oligomeric proteins. Since oligomeric structures involve partial burial of the protein surface at the subunit-subunit interface, calculated values are usually greater than experimental values. On the other hand, the experimental hydration capacity of casein micelles (~4 g water/g protein) is much larger than that predicted by the preceding equation. This is because of the enormous amount of void space within the casein micelle structure, which imbibes water through capillary action and physical entrapment. On a macroscopic level, water binding to proteins occurs in a stepwise process. The high-affinity ionic groups are solvated first at low water activity, followed by polar and nonpolar groups. The sequence of steps involved at increasing water activity is presented in Figure 18 (see also Chap 2). The sorption isotherm of proteins, that is, the amount of water bound per gram of protein as a function of relative humidity, is invariably a sigmoidal curve (see Chap. 2). FIGURE 18 Sequence of steps involved in hydration of a protein. (A) Unhydrated protein. (B) Initial hydration of charg ed g roups. (C) W ater cluster formation near polar and charg ed sites. (D) Completion of hydration at the polar surfaces. (E) Hydrophobic hydration of nonpolar patches; completion of monolayer coverag e. (F) Bridg ing between protein-associated water and bulk water. (G) Completion of hydrodynamic hydration. (From Ref. 96.) Pag e 369 For most proteins, so-called monolayer coverage occurs at a water activity (aW) of 0.05–0.3, and multilayers of water are formed in the water activity range of 0.3–0.7. Water present in the monolayer associates primarily with ionic groups. This water is unfreezable, does not take part as a solvent in chemical reactions, and is often referred to as “bound” water, which should be understood to mean water with “hindered” mobility. The free energy change for desorption of water (i.e., for transfer from the protein surface to the bulk phase) in the monolayer hydration range of 0.07–0.27 g water/g protein is only about 0.75 kJ/mol at 25°C. Since the thermal kinetic energy of water at 25°C is about 2.5 kJ/mol (which is greater than the free energy change for desorption), water molecules in the monolayer are reasonably mobile. At aW = 0.9, proteins bind about 0.3–0.5 g water/g protein (Table 14). Much of this water is unfreezable at 0°C. At aW > 0.9, liquid (bulk) water condenses into the clefts and crevices of protein molecules, or in the capillaries of insoluble protein systems, such as myofibrils. The properties of this water are similar to those of bulk water. This water is known as hydrodynamic water, and moves with the protein molecule. Several environmental factors, such as pH, ionic strength, type of salts, temperature, and protein conformation, influence the water binding capacity of proteins. Proteins exhibit the least hydration at their isoelectric pH, where enhanced protein-protein interactions result in minimal interaction with water. Above and below the isoelectric pH, because of the increase in the net charge and repulsive forces, proteins swell and bind more water. The water binding capacity of most proteins is greater at pH 9–10 than at any other pH. This is due to ionization of sulfhydryl and tyrosine residues. Above pH 10, the loss of positively charged e-amino groups of lysyl residues results in reduced water binding. At low concentrations (< 0.2 M), salts increase the water binding capacity of proteins. This is because hydrated salt ions bind (weakly) to charged groups on proteins. At this low concentration, binding of ions to proteins does not affect the hydration shell of the charged TABLE 14 Hydration Capacities of Various Proteins Protein Hydration capacity (g water/g protein) Pure proteinsª Ribonuclease 0.53 Lysozyme 0.34 Myog lobin 0.44 b-Lactog lobulin 0.54 Chymotrypsinog en 0.23 Serum albumin 0.33 Hemog lobin 0.62 Collag en 0.45 Casein 0.40 Ovalbumin 0.30 Commercial protein preparations b W hey protein concentrates 0.45–0.52 Sodium caseinate 0.38–0.92 Soy protein 0.33 ªAt 90% relative humidity. bAt 95% relative humidity. Source: Refs. 57 and 60. Pag e 370 groups on the protein, and the increase in water binding essentially comes from water associated with the bound ions. However, at high salt concentrations, much of the existing water is bound to salt ions, and this results in dehydration of the protein. The water binding capacity of proteins generally decreases as the temperature is raised, because of decreased hydrogen bonding and decreased hydration of ionic groups. The water binding capacity of a denatured protein is generally about 10% greater than that of the native protein. This is due to an increase in surface area to mass ratio with exposure of some previously buried hydrophobic groups. However, if denaturation leads to aggregation of the protein, then its water binding capacity may actually decrease because of the protein-protein interactions. It should be pointed out that most denatured food proteins exhibit very low solubility in water. Their water binding capacities, however, are not drastically different from those in the native state. Thus, water binding capacity cannot be used to predict the solubility characteristics of proteins. In other words, the solubility of a protein is dependent not only on water binding capacity, but also on other thermodynamic factors. In food applications, the water-holding capacity of a protein preparation is more important than the water binding capacity. Water-holding capacity refers to the ability of the protein to imbibe water and retain it against gravitational force within a protein matrix, such as protein gels or beef and fish muscle. This water refers to the sum of the bound water, hydrodynamic water, and the physically entrapped water. The contribution of the physically entrapped water to water holding capacity is much larger than those of the bound and hydrodynamic water. However, studies have shown that the water-holding capacity of proteins is positively correlated with water binding capacity. The ability of protein to entrap water is associated with juiciness and tenderness of comminuted meat products and desirable textural properties of bakery and other gel-type products. 6.5.2 Solubility The functional properties of proteins are often affected by protein solubility, and those most affected are thickening, foaming, emulsifying, and gelling. Insoluble proteins have very limited uses in food. The solubility of a protein is the thermodynamic manifestation of the equilibrium between protein-protein and protein-solvent interactions. Protein-Protein + Solvent-Solvent Protein-Solvent (45) The major interactions that influence the solubility characteristics of proteins are hydrophobic and ionic in nature. Hydrophobic interactions promote protein-protein interactions and result in decreased solubility, whereas ionic interactions promote proteinwater interactions and result in increased solubility. Ionic residues introduce two kinds of repulsive forces between protein molecules in solution. The first involves electrostatic repulsion between protein molecules owing to a net positive or negative charge at any pH other than the isoelectric pH; the second involves repulsion between hydration shells around ionic groups. Bigelow [6] proposed that the solubility of a protein is fundamentally related to the average hydrophobicity of the amino acid residues and the charge frequency. The average hydrophobicity is defined as (46) where Dgresidue is the hydrophobicity of each amino acid side chain obtained from the free energy change for transfer from ethanol to water, and n is the total number of residues in the protein. The charge frequency is defined as Pag e 371 s = (n + + n - )/n (47) where n + and n - are the total number of positively and negatively charged residues, respectively, and n is the total number of residues. According to Bigelow [6], the smaller the average hydrophobicity and the larger the charge frequency, the greater will be the solubility of the protein. Although this empirical correlation is true for most proteins, it is not an absolute one. The Bigelow approach is faulty because the solubility of a protein is dictated by the hydro-philicity and hydrophobicity of the protein surface that contacts with the surrounding water, rather than the average hydrophobicity and charge frequency of the molecule as a whole. The fewer the number of surface hydrophobic patches, the greater the solubility. Based on solubility characteristics, proteins are classified into four categories. Albumins are those that are soluble in water at pH 6.6 (e.g., serum albumin, ovalbumin, and a-lactalbumin), globulins are those that are soluble in dilute salt solutions at pH 7.0 (e.g., glycinin, phaseolin, and b-lactoglobulin), glutelins are those that are soluble only in acid (pH 2) and alkaline (pH 12) solutions (e.g., wheat glutelins), and prolamines are those soluble in 70% ethanol (e.g., zein and gliadins). Both prolamines and glutelins are highly hydrophobic proteins. In addition to these intrinsic physicochemical properties, solubility is influenced by several solution conditions, such as pH, ionic strength, temperature, and the presence of organic solvents. pH and Solubility At pH values below and above the isoelectric pH, proteins carry a net positive or a net negative charge, respectively. Electrostatic repulsion and hydration of charged residues promote solubilization of the protein. When solubility is plotted against pH, most food proteins exhibit a U-shaped curve. Minimum solubility occurs at about the isoelectric pH of proteins. A majority of food proteins are acidic proteins; that is, the sum of Asp and Glu residues is greater than the sum of Lys, Arg, and His residues. Therefore, they exhibit minimum solubility at pH 4–5 (isoelectric pH) and maximum solubility at alkaline pH. The occurrence of minimum solubility near the isoelectric pH is due primarily to the lack of electrostatic repulsion, which promotes aggregation and precipitation via hydrophobic interactions. Some food proteins, such as b-lactoglobulin (pI 5.2) and bovine serum albumin (pI 5.3), are highly soluble at their isoelectric pH. This is because these proteins contain a large ratio of surface hydrophilic residues to surface nonpolar groups. It should be remembered that even though a protein is electrically neutral at its pI, it is still charged, but the positive and negative charges on the surface are equal. If the hydrophilicity and the hydration repulsion forces arising from these charged residues are greater than the protein-protein hydrophobic interactions, then the protein will still be soluble at the pI. Since most proteins are highly soluble at alkaline pH (8–9), protein extraction from plant sources, such as soybean flour, is carried out at this pH. The protein is then recovered from the extract by isoelectric precipitation at pH 4.5–4.8. Heat denaturation changes the pH-solubility profile of proteins (Fig. 19). Native whey protein isolate is completely soluble in the pH range 2–9, but when heated at 70°C for 1 min to 10 min a typical U-shaped solubility profile develops with a solubility minimum at pH 4.5. The change in the solubility profile upon heat denaturation is due to an increase in the hydrophobicity of the protein surface as a consequence of unfolding. Unfolding alters the balance between protein-protein and protein-solvent interactions in favor of the former. Ionic Strength and Solubility The ionic strength of a salt solution is given bywhere Ci is concentration of an ion, and Zi is its valance. At low ionic strength (< 0.5), ions neutralize charges at the surface of proteins. This charge screening affects solubility in two different ways, depending on the characteristics of the protein surface. Solubility decreases for those proteins that contain a high incidence of nonpolar patches, and it increases for those that don't. The former behavior is typical for soy proteins, and the latter behavior is exhibited by b-lactoblobulin. While the decrease in solubility is caused by enhanced hydrophobic interactions, the increase in solubility is caused by a decrease in the ionic activity of the protein macroion. At ionic strength > 1.0, salts have ion specific effects on protein solubility. As salt concentration is increased up to m = 1; sulfate and fluoride salts progressively decrease solubility (salting out), whereas thiocyanate and perchlorate salts increase solubility (salting in). At constant m, relative effectiveness of various ions on solubility follows the Hofmeister series, with anions promoting solubility in the order , and cations decreasing solubility in the order . This behavior is analogous to the effects of salts on the thermal denaturation temperature of proteins (see Sec. 6.4). Generally, solubility of proteins in salt solutions follows the relation log(S/S0) = b – KSCS (49) where S and S0 are solubilities of the protein in the salt solution and in water, respectively, KS is the salting out constant, CS is molar concentration of salt, and b is a constant. KS is positive for salting-out type of salts and negative for salting-in type of salts. Temperature and Solubility At constant pH and ionic strength, the solubility of most proteins generally increases with temperature between 0 and 40°C. Exceptions occur with highly hydrophobic proteins, such as b-casein and some cereal proteins, which exhibit a negative relationship with temperature. Pag e 373 Above 40°C, the increase in thermal kinetic energy causes protein unfolding (denaturation), exposure of nonpolar groups, aggregation, and precipitation, that is, decreased solubility. Organic Solvents and Solubility Addition of organic solvents, such as ethanol or acetone, lowers the permittivity of an aqueous medium. This increases intra- and intermolecular electrostatic forces, both repulsive as well as attractive. The repulsive intramolecular electrostatic interactions cause unfolding of the protein molecule. In the unfolded state, the permittivity promotes intermolecular hydrogen bonding between the exposed peptide groups and attractive intermolecular electrostatic interactions between oppositely charged groups. These intermolecular polar interactions lead to precipitation of the protein in organic solvents or reduced solubility in an aqueous medium. The role of hydrophobic interactions in causing precipitation in organic solvents is minimal because of the solubilizing effect of organic solvents on nonpolar residues. However, even in aqueous media containing a low concentration of organic solvent, hydrophobic interactions between exposed residues may also contribute to insolublization. Since solubility of proteins is intimately related to their structural states, it is often used as a measure of the extent of denaturation during extraction, isolation, and purification processes. It is also used as an index of the potential applications of proteins. Commercially prepared protein concentrates and isolates show a wide range of solubility. The solubility characteristics of these protein preparations are expressed as protein solubility index (PSI) or protein dispersibility index (PDI). Both of these terms express the percentage of soluble protein present in a protein sample. The PSI of commercial protein isolates varies from 25 to 80%. 6.5.3 Interfacial Properties of Proteins Several natural and processed foods are either foam- or emulsion-type products. These types of dispersed systems are unstable unless a suitable amphiphilic substance is present at the interface between the two phases (see Chap. 3). Proteins are amphiphilic molecules, and they migrate spontaneously to an air-water interface or an oil-water interface. This spontaneous migration of proteins from a bulk liquid to an interface indicates that the free energy of proteins is lower at the interface than it is in the bulk aqueous phase. Thus, when equilibrium is established, the concentration of protein in the interfacial region is always much greater than it is in the bulk aqueous phase. Unlike low-molecular-weight surfactants, proteins from a highly viscoelastic film at an interface, which has the ability to withstand mechanical shocks during storage and handling. Thus, protein-stabilized foams and emulsions are more stable than those prepared with low-molecular-weight surfactants, and because of this, proteins are extensively used for these purposes. Although all proteins are amphiphilic, they differ significantly in their surface-active properties. The differences in the surfaceactive properties among proteins cannot be attributed to differences in the ratio of hydrophobic to hydrophilic residues. If a large hydrophobicity/hydrophilicity ratio were the primary determinant of the surface activity of proteins, then plant proteins, which contain more than 40% hydrophobic amino acid residues, should be better surfactants than albumin-type proteins, such as ovalbumin and bovine serum albumin, which contain less than 30% hydrophobic amino acid residues. On the contrary, ovalbumin and serum albumin are better emulsifying and foaming agents than are soy proteins and other plant proteins. Furthermore, average hydrophobicities of most proteins fall within a narrow range [7], yet they exhibit remarkable differences in their surface activity. It must be concluded, therefore, that differences in surface activity are related primarily to differences in protein conformation. The conformational factors of importance include stability/flexibility of the polypeptide chain, ease of adaptability to changes in the environment, and distribution pattern of hydrophilic and Pag e 374 hydrophobic groups on the protein surface. All of these conformational factors are inter-dependent, and they collectively have a large influence on the surface activity of proteins. It has been shown that desirable surface-active proteins have three attributes: (a) ability to rapidly adsorb to an interface, (b) ability to rapidly unfold and reorient at an interface, and (c) an ability, once at the interface, to interact with the neighboring molecules and form a strong cohesive, viscoeleastic film that can withstand thermal and mechanical motions [26,87]. The first and most critical event in the creation of stable foams and emulsions during whipping or homogenization is the spontaneous and rapid adsorption of proteins at the newly created interface. The rapidity with which a protein can adsorb to airwater or oil-water interfaces depends on the distribution pattern of hydrophobic and hydrophilic patches on its surface. If the protein surface is extremely hydrophilic and contains no discernable hydrophobic patches, adsorption probably will not take place because its free energy will be lower in the aqueous phase than either at the interface or in the nonpolar phase. As the number of hydrophobic patches on the protein surface is increased, spontaneous adsorption to an interface becomes more probable (Fig. 20). Single hydrophobic residues randomly distributed on the protein surface do not constitute a hydrophobic patch, nor do they possess sufficient interaction energy to strongly anchor the protein at an interface. Even though more than 40% of a protein’s overall accessible surface is covered with nonpolar residues, they will not enhance protein adsorption unless they exist as segregated regions or patches. In other words, the molecular characteristics of the protein surface have an enormous influence on whether a protein will spontaneously adsorb to an interface and how effective it will be as a stabilizer of dispersions. The mode of adsorption of proteins at an interface is different from that of low-molecular-weight surfactants. In the case of lowmolecular-weight surfactants, such as phospholipids and monoacylglycerols, conformational constraints for adsorption and orientation do not exist because hydrophilic and hydrophobic moieties are present at the ends of the molecule. In the case of proteins, however, the distribution pattern of hydrophobic and hydrophilic patches on the surface and the structural rigidity of the molecule cause constraints to adsorption and orientation. Because of the bulky, folded nature of proteins, once adsorbed, a large portion of the molecule remains in the bulk phase and only a small portion is anchored at the interface. The tenacity with FIGURE 20 Schematic representation of the role of surface hydrophobic patches on the probability of adsorption of proteins at the air-water interface. (From Ref. 22.) Pag e 375 which this small portion of the protein molecule remains attached at the interface depends on the number of peptide segments anchored to the interface, and the energetics of interaction between these segments and the interface. The protein will be retained at the interface only when the sum of negative free energy changes of segment interactions is much greater than the thermal kinetic energy of the protein molecule. The number of peptide segments anchored at the interface depends, in part, on the conformational flexibility of the molecule. Highly flexible molecules, such as caseins, can undergo rapid conformational changes once they are adsorbed at the interface, enabling additional polypeptide segments to bind to the interface. On the other hand, rigid globular proteins, such as lysozyme and soy proteins, cannot undergo extensive conformational changes at the interface. At interfaces, polypeptide chains assume one or more of three distinct configurations: trains, loops, and tails (Fig. 21). The trains are segments that are in direct contact with the interface, loops are segments of the polypeptide that are suspended in the aqueous phase, and tails are N- and C-terminal segments of the protein that are usually located in the aqueous phase. The relative distribution of these three configurations depends on the conformational characteristics of the protein. The greater the proportion of polypeptide segments in a train configuration, the stronger is the binding and the lower is the interfacial tension. The mechanical strength of a protein film at an interface depends on cohesive inter-molecular interactions. These include attractive electrostatic interactions, hydrogen bonding, and hydrophobic interactions. Interfacial polymerization of adsorbed proteins via disulfide-sulfhydryl interchange reactions also increase its viscoelastic properties. The concentration of protein in the interfacial film is about 20–25% (w/v), and the protein exists in almost a gel-like state. The balance of various noncovalent interactions is crucial to the stability and viscoelastic properties of this gel-like film. For example, if hydrophobic interactions are too strong, this can lead to interfacial aggregation, coagulation, and eventual precipitation of the protein, which will be detrimental to film integrity. If repulsive electrostatic forces are much stronger than attractive interactions, this may prevent formation of a thick, cohesive film. Therefore, a proper balance of attractive, repulsive, and hydration forces is required to form a stable viscoelastic film. The basic principles involved in formation and stability of emulsions and foams are very similar. However, since the energetics of these interfaces are different, the molecular requireFIGURE 21 The various config urations of a flexible polypeptide at an interface. (From Ref. 22.) Pag e 376 ments for protein functionality at these interfaces are not the same. In other words, a protein that is a good emulsifier may not be a good foaming agent. It should now be clear that the behavior of proteins at interfaces is very complex and not well understood. Therefore, the following discussion of the emulsifying and foaming properties of food proteins will be largely qualitative in nature. 6.5.3.1 Emulsifying Properties The physical chemistry of emulsion formation and the factors affecting creaming, flocculation, coalescence, and stability were reviewed in Chapter 3. Several natural and processed foods, such as milk, egg yolk, coconut milk, soy milk, butter, margarine, mayonnaise, spreads, salad dressings, frozen desserts, frankfurter, sausage, and cakes, are emulsion-type products where proteins play an important role as an emulsifier. In natural milk, the fat globules are stabilized by a membrane composed of lipoproteins. When milk is homogenized, the lipoprotein membrane is replaced by a protein film comprised of casein micelles and whey proteins. Homogenized milk is more stable against creaming than is natural milk because the casein micelle-whey protein film is stronger than the natural lipoprotein membrane. 6.5.3.2 Methods for Determining the Emulsifying Properties of Proteins The emulsifying properties of food proteins are evaluated by several methods such as size distribution of oil droplets formed, emulsifying activity, emulsion capacity, and emulsion stability. Emulsifying Activity Index The physical and sensory properties of a protein-stabilized emulsion depend on the size of the droplets formed and the total interfacial area created. The average droplet size of emulsions can be determined by several methods, such as light microscopy (not very reliable), electron microscopy, light scattering (photon correlation spectroscopy), or use of a Coulter counter. Knowing mean droplet size, total interfacial area can be obtained from the relation (50) where f is the volume fraction of the dispersed phase (oil) and R is the mean radius of the emulsion particles. If m is the mass of the protein, then the emulsifying activity index (EAI), that is, the interfacial area created per unit of mass protein, is (51) Another simple and more practical method to determine EAI of proteins is the turbidimetric method [86]. The turbidity of an emulsion is given by (52) where A is absorbance and l is path length. According to the Mie theory of light scattering, the interfacial area of an emulsion is two times its turbidity. If f is the volume fraction of the oil and C is the weight of protein per unit volume of the aqueous phase, then the EAI of the protein is given by Pag e 377 (53) It should be mentioned that in the original article [86], f instead of (1 – f) was used in the denominator of this equation. The expression as given in Equation 53 is the correct one because f is defined as the oil volume fraction, and thus (1 – f)C is the total mass of protein in a unit volume of the emulsion [13]. Although this method is simple and practical, the main drawback is that it is based on measurement of turbidity at a single wavelength, 500 nm. Since the turbidity of food emulsions is wavelength dependent, the interfacial area obtained from turbidity at 500 nm is not very accurate. Therefore, use of this equation to estimate mean particle diameter or the number of emulsion particles present in the emulsion gives results that are not very reliable. However, the method can be used for qualitative comparison of emulsifying activities of different proteins, or changes in the emulsifying activity of a protein after various treatments. Protein Load The amount of protein adsorbed at the oil-water interface of an emulsion has a bearing on its stability. To determine the amount of protein adsorbed, the emulsion is centrifuged, the aqueous phase is separated, and the cream phase is repeatedly washed and centrifuged to remove any loosely adsorbed proteins. The amount of protein adsorbed to the emulsion particles is determined from the difference between the total protein initially present in the emulsion and the amount present in the wash fluid from the cream phase. Knowing the total interfacial area of the emulsion particles, the amount of protein adsorbed per square meter of the interfacial area can be calculated. Generally, the protein load is in the range of about 1–3 mg/m2 of interfacial area. As the volume fraction of the oil phase is increased the protein load decreases at constant protein content in the total emulsion. For high-fat emulsions and small-sized droplets, more protein is obviously needed to adequately coat the interfacial area and stabilize the emulsion. Emulsion Capacity Emulsion capacity (EC) is the volume (ml) of oil that can be emulsified per gram of protein before phase inversion (a change from oil-in-water emulsion to water-in-oil) occurs. This method involves addition of oil or melted fat at a constant rate and temperature to an aqueous protein solution that is continuously agitated in a food blender. Phase inversion is detected by an abrupt change in viscosity or color (usually a dye is added to the oil), or by an increase in electrical resistance. For a proteinstabilized emulsion, phase inversion usually occurs when f is about 0.65–0.85. Inversion is not instantaneous, but is preceded by formation of a water-in-oil-in-water double emulsion. Since emulsion capacity is expressed as volume of oil emulsified per gram protein at phase inversion, it decreases with increasing protein concentration once a point is reached where unadsorbed protein accumulates in the aqueous phase. Therefore, to compare emulsion capacities of different proteins, EC versus protein concentration profiles should be used instead of EC at a specific protein concentration. Emulsion Stability Protein-stabilized emulsions are often stable for days. Thus, a detectable amount of creaming or phase separation is usually not observed in a reasonable amount of time when samples are stored at atmospheric conditions. Therefore, drastic conditions, such as storage at elevated temperature or separation under centrifugal force, is often used to evaluate emulsion stability. If centrifugation is used, stability is then expressed as percent decrease in interfacial area (i.e., turbidity) of the emulsion, or percent volume of cream separated, or as the fat content of the cream layer. More often, however, emulsion stability is expressed as Pag e 378 (54) where the volume of the cream layer is measured after a standardized centrifugation treatment. A common centrifugation technique involves centrifugation of a known volume of emulsion in a graduated centrifuge tube at 1300 × g for 5 min. The volume of the separated cream phase is then measured and expressed as a percentage of the total volume. Sometimes centrifugation at a relatively low gravitational force (180 × g) for a longer time (15 min) is used to avoid coalescence of droplets. The turbidimetric method (discussed earlier) can also be used to evaluate emulsion stability. In this case stability is expressed as emulsion stability index (ESI), which is defined as the time to achieve a turbidity of the emulsion that is one-half of the original value. The methods used to determine emulsion stability are very empirical. The most fundamental quantity related to stability is the change in interfacial area with time, but this is difficult to measure directly. 6.5.3.3 Factors Influencing Emulsification The properties of protein-stabilized emulsions are affected by several factors. These include intrinsic factors, such as pH, ionic strength, temperature, presence of low-molecular-weight surfactants, sugars, oil-phase volume, type of protein, and the melting point of the oil used; and extrinisic factors, such as type of equipment, rate of energy input, and rate of shear. Standardized methods for systematically evaluating the emulsifying properties of proteins have not been agreed upon. Therefore, results among laboratories cannot be accurately compared, and this has hampered the understanding of the molecular factors that affect emulsifying properties of proteins. The general forces involved in the formation and stabilization of emulsion were discussed in Chapter 3. Therefore, only the molecular factors that affect protein-stabilized emulsions need be discussed here. Solubility plays a role in emulsifying properties, but 100% solubility is not an absolute requirement. While highly insoluble proteins do not perform well as emulsifiers, no reliable relationship exists between solubility and emulsifying properties in the 25–80% solubility range [64]. However, since the stability of a protein film at the oil-water interface is dependent on favorable interactions with both the oil and aqueous phases, some degree of solubility is likely to be necessary. The minimum solubility requirement for good performance may vary among proteins. In meat emulsions, such as in sausage and frankfurter, solubilization of myofibrillar proteins in 0.5 M NaCl enhances their emulsifying properties. Commercial soy protein isolates, which are isolated by thermal processing, have poor emulsifying properties because of their very low solubility. The formation and stability of protein-stabilized emulsions are affected by pH. Several mechanisms are involved. Generally, proteins that have high solubility at the isoelectric pH (e.g., serum albumin, gelatin, and egg-white proteins) show maximum emulsifying activity and emulsion capacity at that pH. The lack of net charge and electrostatic repulsive interactions at the isoelectric pH helps maximize protein load at the interface and promotes formation of a highly viscoelastic film, both of which contribute to emulsion stability. However, the lack of electrostatic repulsive interactions among emulsion particles can, in some instances, promote flocculation and coalescence, and thus decrease emulsion stability. On the other hand, if the protein is highly hydrated at the isoelectric pH (unusual), then hydration repulsion forces between emulsion particles may prevent flocculation and coalescence, and thus stabilize the emulsion. Because most food proteins (caseins, commercial whey proteins, meat proteins, soy proteins) at their isoelectric pH are sparingly soluble, poorly hydrated, and lack electrostatic Pag e 379 repulsive forces, they are generally poor emulsifiers at this pH. These proteins may, however, be effective emulsifiers when moved away from their isoelectric pH. The emulsifying properties of proteins show a weak positive correlation with surface hydrophobicity, but not with mean hydrophobicity (i.e., Jmol-1residue-1). The ability of various proteins to decrease interfacial tension at the oil-water interface and to increase the emulsifying activity index is related to their surface hydrophobicity values (Fig. 22). However, this relationship is by no means perfect. The emulsifying properties of several proteins, such as b-lactoglobulin, a-lactalbumin, and soy proteins, do not show a strong correlation with surface hydrophobicity. The surface hydrophobicity of proteins is usually determined by measuring the amount of a hydrophobic fluorescent probe, such as cis-parinaric acid, that can bind to the protein [53]. Although this method provides some information on the nonpolarity of the protein surface, it is questionable whether the measured value truly reflects the “hydrophobicity” of the protein surface. The true definition of surface hydrophobicity is that portion of the nonpolar surface of the protein that makes contact with the surrounding bulk water. However, cis-parinaric acid is capable of binding only to hydrophobic cavities formed at the surface by association of nonpolar residues. These protein cavities are accessible to nonpolar ligands, but they are not accessible to water and may not be accessible to either phase in an oil-water emulsion, unless the protein is able to undergo rapid conformational rearrangement at the interface. The poor correlation of surface hydrophobicity (as measured by cis-parinaric acid binding) with the emulsifying properties of some proteins may be related to the fact that cis-parinaric acid provides no indication of molecular flexibility. Molecular flexibility at the oil-water interface may be the most important determinant of the emulsifying properties of proteins. Partial denaturation (unfolding) of proteins prior to emulsification, which does not result FIGURE 22 Correlations of surface hydrophobicity of various proteins with (A) oil/water interfacial tension, and (B) emulsifying activity index. Surface hydrophobicity was determined from the amount of hydrophobic fluorescent probe bound per unit weig ht of protein. The numbers in the plots represent (1) bovine serum albumin; (2) b-lactoblobulin; (3) trypsin; (4) ovalbumin; (5) conalbumin; (6) lysozyme; (7) k-casein; (8–12) ovalbumin denatured by heating at 85°C for 1, 2, 3, 4, or 5 min, respectively; (13–18) lysozyme denatured by heating at 85°C for 1, 2, 3, 4, 5, or 6 min, respectively; (19–23) ovalbumin bound to 0.2, 0.3, 1.7, 5.7, or 7.9 mol dodecyl sulfate/mol protein, respectively; (24–28) ovalbumin bound to 0.3, 0.9, 3.1, 4.8 or 8.2 mol linoleate/mol protein, respectively. (From Ref. 53.) Pag e 380 in insolublization, usually improves their emulsifying properties. This is due to increased molecular flexibility and surface hydrophobicity. The rate of unfolding at an interface depends on the flexibility of the original molecule. In the unfolded state, proteins containing free sulfhydryl groups and disulfide bonds undergo slow polymerization via disulfide-sulfhydryl interchange reaction [27]. This leads to formation of a highly viscoelastic film at the oil-water interface. Heat denaturation that is sufficient to cause insolublization impairs emulsifying properties of proteins. 6.5.3.4 Foaming Properties Foams consist of an aqueous continuous phase and a gaseous (air) dispersed phase. Many processed foods are foam-type products. These include whipped cream, ice cream, cakes, meringue, bread, souffles, mousses, and marshmallow. The unique textural properties and mouthfeel of these products stem from the dispersed tiny air bubbles. In most of these products, proteins are the main surface-active agents that help in the formation and stabilization of the dispersed gas phase. Generally, protein-stabilized foams are formed by bubbling, whipping, or shaking a protein solution. The foaming property of a protein refers to its ability to form a thin tenacious film at gas-liquid interfaces so that large quantities of gas bubbles can be incorporated and stabilized. Foaming properties are evaluated by several means. The foamability or foaming capacity of a protein refers to the amount of interfacial area that can be created by the protein. It can be expressed in several ways, such as overrun(or steady-state foam value), or foaming power (or foam expansion). Overrun is defined as (55) (56) Foaming power generally increases with protein concentration until a maximum value is attained. It is also affected by the method used for foam formation. FP at a given protein concentration is often used as a basis for comparing the foaming properties of various proteins. The foaming powers of various proteins at pH 8.0 are given in Table 15. Foam stability refers to the ability of protein to stabilize foam against gravitational and mechanical stresses. Foam stability is often expressed as the time required for 50% of the liquid to drain from a foam or for a 50% reduction in foam volume. These are very empirical methods, and they do not provide fundamental information about the factors that affect foam stability. The most direct measure of foam stability is the reduction in foam interfacial area as a function of time. This can be done as follows. According to the Laplace principle, the internal pressure of a bubble is greater than the external (atmospheric) pressure, and under stable conditions the pressure difference, DP, is (57) where Pi and Po are the internal and external pressures, respectively, r is radius of the foam bubble, and g is surface tension. According to this equation the pressure inside a closed vessel containing foam will increase when the foam collapses. The net change in where V is the total volume of the system, DP is the pressure change, and DA is the net change in interfacial area resulting from the fraction of collapsed foam. The initial interfacial area of the foam is given by (59) where dP is the net pressure change when the entire foam is collapsed. The A0 value is a measure of foamability, and the rate of decrease of A with time can be used as a measure of foam stability. This approach has been used to study the foaming properties of food proteins [117,119]. The strength or stiffness of the foam refers to the maximum weight a column of foam can withstand before it collapses. This property is also assessed by measuring foam viscosity. 6.5.3.5 Environmental Factors Influencing Foam Formation and Stability pH Several studies have shown that protein-stabilized foams are more stable at the isoelectric pH of the protein than at any other pH, provided there is no insolublization of the protein at pI. At or near the isoelectric pH region, the lack of repulsive interactions promotes favorable protein-protein interactions and formation of a viscous film at the interface. In addition, an increased amount of protein is adsorbed to the interface at the pI because of lack of repulsion between the interface and adsorbing molecules. These two factors improve both foamability and foam stability. If the protein is sparingly soluble at pI, as most food proteins are, then only the soluble protein fraction will be involved in foam formation. Since the concentration of this soluble fraction is very low, the amount of foam formed will be less, but the stability will be high. Although the insoluble fraction does not contribute to foamability, adsorption of these insoluble protein particles may stabilize the foam, probably by increasing cohesive forces in the protein film. Generally, adsorption of hydrophobic particles increases the stability of foams. At pH other than Pag e 382 pI, foamability of proteins is often good, but foam stability is poor. Egg-white proteins exhibit good foaming properties at pH 8–9 and at their isoelectric pH 4–5. Salts The effects of salts on the foaming properties of proteins depends on the type of salt and the solubility characteristics of the protein in that salt solution. The foamability and foam stability of most globular proteins, such as bovine serum albumin, egg albumin, gluten, and soy proteins, increase with increasing concentration of NaCl. This behavior is usually attributed to neutralization of charges by salt ions. However, some proteins, such as whey proteins, exhibit the opposite effect: Both foamability and foam stability decrease with increasing concentration of NaCl (Table 16) [120]. This is attributed to salting-in of whey proteins, especially b-lactoglobulin, by NaCl. Proteins that are salted out in a given salt solution generally exhibit improved foaming properties, whereas those that are salted in generally exhibit poor foaming properties. Divalent cations, such as Ca2+ and Mg2+, dramatically improve both foamability and foam stability at 0.02–0.4 M concentration. This is primarily due to crosslinking of protein molecules and creation of films with better viscoelastic properties [121]. Sugars Addition of sucrose, lactose and other sugars to protein solutions often impairs foamability, but improves foam stability. The positive effect of sugars on foam stability is due to increased bulk-phase viscosity, which reduces the rate of drainage of the lamella fluid. The depression in foam overrum is mainly due to enhanced stability of protein structure in sugar solutions. Because of this, the protein molecule is less able to unfold upon adsorption at the interface. This decreases the ability of the protein to produce large interfacial areas and large foam volume during whipping. In sugar-containing, foam-type dessert products, such as meringues, souffles, and cakes, it is preferable to add sugar after whipping when possible. This will enable the protein to adsorb, unfold, and form a stable film, and then the added sugar will increase foam stability by increasing the viscosity of the lamella fluid. Lipids Lipids, especially phospholipids, when present at concentrations greater than 0.5%, markedly impair the foaming properties of proteins. Because lipids are more surface-active than proteins, they readily adsorb at the air-water interface and inhibit adsorption of proteins during foam formation. Since lipid films lack the cohesive and viscoelastic properties necessary to withstand TABLE 16 Effect of NaCl on Foamability and Stability of W hey Protein Isolate NaCl concentration (M) Total interfacial area (cm2 /ml of foam) Time for 50% collapse of initial area (sec) 0.00 333 510 0.02 317 324 0.04 308 288 0.06 307 180 0.08 305 165 0.10 287 120 0.15 281 120 Source: Compiled from Ref. 120. Pag e 383 the internal pressure of the foam bubbles, the bubbles rapidly expand, then collapse during whipping. Thus, lipid-free whey protein concentrates and isolates, soy proteins, and egg proteins without egg yolk display better foaming properties than do lipid-contaminated preparations. Protein Concentration Several properties of foams are influenced by protein concentration. The greater the protein concentration, the stiffer the foam. Foam stiffness results from small bubble size and high viscosity. The stability of the foam is enhanced by greater protein concentration because this increases viscosity and facilitates formation of a multilayer, cohesive protein film at the interface. Foamability generally reaches a maximum value at some point during an increase in protein concentration. Some proteins, such as serum albumin, are able to form relatively stable foams at 1% protein concentration, whereas proteins such as whey protein isolate and soy conglycinin require a minimum of 2–5% to form a relatively stable foam. Generally, most proteins display maximum foamability at 2–8% concentration. The interfacial concentration of proteins in foams is about 2–3 mg/m2 . Partial heat denaturation improves the foaming properties of proteins. For instance, heating of whey protein isolate (WPI) at 70°C for 1 min improves, whereas heating at 90°C for 5 min decreases foaming properties even though the heated proteins remain soluble in both instances [119]. The decrease in foaming properties of WPI heated at 90°C was attributed to extensive polymerization of the protein via disulfide-sulfhydryl interchange reactions. These very highmolecular-weight polymers are unable to adsorb to the air-water interface during foaming. The method of foam generation influences the foaming properties of proteins. Air introduction by bubbling or sparging usually results in a “wet” foam with a relatively large bubble size. Whipping at moderate speed generally results in foam with small-sized bubbles because the shearing action results in partial denaturation of the protein before adsorption occurs. However, whipping at high shear rate or “overbeating” can decrease foaming power because of aggregation and precipitation of proteins. Some foam-type food products, such as marshmallow, cakes, and bread, are heated after the foam is formed. During heating, expansion of air and decreased viscosity can cause bubble rupture and collapse of the foam. In these instances, the integrity of the foam depends on gelation of the protein at the interface so sufficient mechanical strength is developed to stabilize the foam. Gelatin, gluten, and egg white, which display good foaming and gelling properties, are highly suitable for this purpose. 6.5.3.6 Molecular Properties Influencing Foam Formation and Stability For a protein to perform effectively as a foaming agent or an emulsifier it must meet the following basic requirements: (a) it must be able to rapidly adsorb to the air-water interface, (b) it must readily unfold and rearrange at the interface, and (c) it should be able to form a viscous cohesive film through intermolecular interactions. The molecular properties that affect foaming properties are molecular flexibility, charge density and distribution, and hydrophobicity. The free energy of an air-water interface is significantly greater than that of an oil-water interface. Therefore, to stabilize an airwater interface, the protein must have the ability to rapidly adsorb to the freshly created interface, and instantaneously decrease the interfacial tension to a low level. The lowering of interfacial tension is dependent on the ability of the protein to rapidly unfold, rearrange, and expose hydrophobic groups at the interface. Studies have shown that b-casein, which is a random-coil-type protein, performs in this manner. On the other hand, lysozyme, which is a tightly folded globular protein with four intramolecular Pag e 384 disulfide bonds, adsorbs very slowly, only partially unfolds, and reduces the surface tension only slightly (Fig. 23) [115]. Lysozyme is, therefore, a poor foaming agent. Thus, molecular flexibility at the interface is quintessential for good performance as a foaming agent. Apart from molecular flexibility, hydrophobicity plays an important role in foamability of proteins. The foaming power of proteins is positively correlated with mean hydrophobicity FIGURE 23 Variation of surface concentration ( ) and surface pressure ( ) with time during adsorption of b-casein (A) and eg g -white lysozyme (B) at the air-water interface from a 1.5-mg /ml protein solution. (From Ref. 115.) Pag e 385 (Fig. 24A). However, the foaming power of proteins varies curvilinearly with surface hydrophobicity, and a significant correlation does not exist between these two properties at hydrophobicity values of greater than 1000 [54]. This indicates that a surface hydrophobicity of at least 1000 is needed for initial adsorption of proteins at the air-water interface, whereas, once adsorbed, the ability of the protein to create more interfacial area during foam formation depends on the mean hydrophobicity of the protein. A protein that displays good foamability need not be a good foam stabilizer. For example, although b-casein exhibits excellent foamability, the stability of the foam is poor. On the other hand, lysozyme exhibits poor foamability, but its foams are very stable. Generally, proteins that possess good foaming power do not have the ability to stabilize a foam, and proteins that produce stable foams often exhibit poor foaming power. It appears that foamability and stability are influenced by two different sets of molecular properties of proteins that are often antagonistic. Whereas foamability is affected by rate of adsorption, flexibility, and hydrophobicity, stability depends on the rheological properties of the protein film. The rheological properties of films depend on hydration, thickness, protein concentration, and favorable intermolecular interactions. Proteins that only partially unfold and retain some degree of folded structure usually form thicker, denser films, and more stable foams (e.g., lysozyme and serum albumin) than do those that completely unfold (e.g., b-casein) at the air-water interface. In the former case, the folded structure extends into the subsurface in the form of loops. Noncovalent interactions, and possibly disulfide cross-linking, between these loops promote formation of a gel network, which has excellent viscoelastic and mechanical properties. For a protein to possess good foamability and foam stability it should have an appropriate balance between flexibility and rigidity, should easily undergo unfolding, and should engage in abundant cohesive interactions at the interface. However, what extent of unfolding is desirable for a given protein is difficult, if not impossible, to predict. In addition to these factors, foam stability usually exhibits an inverse relationship with the charge density of proteins (Fig. 24B). High charge density apparently interferes with formation of cohesive film. Most food proteins are mixtures of various proteins, and therefore their foaming properties are influenced by interaction between the protein components at the interface. The excellent whipping properties of egg white are attributed to interactions between its protein components, such as ovalbumin, conalbumin, and lysozyme. Several studies have indicated that the foaming properties of acidic proteins can be improvd by mixing them with basic proteins, such as lysozyme and clupeine [89]. This enhancing effect seems to be related to the formation of an electrostatic complex between the acidic and basic proteins. Limited enzymatic hydrolysis of proteins generally improves their foaming properties. This is because of increased molecular flexibility and greater exposure of hydrophobic groups. However, extensive hydrolysis impairs foamability because lowmolecular-weight peptides cannot form a cohesive film at the interface. 6.5.4 Flavor Binding Proteins themselves are odorless. However, they can bind flavor compounds, and thus affect the sensory properties of foods. Several proteins, especially oilseed proteins and whey protein concentrates, carry undesirable flavors, which limits their usefulness in food applications. These off flavors are due mainly to aldehydes, ketones, and alcohols generated by oxidation of unsaturated fatty acids. Upon formation, these carbonyl compounds bind to proteins and impart characteristic off flavors. For example, the beany and grassy flavor of soy protein preparations is attributed to the presence of hexanal. The binding affinity of some of these carbonyls is so strong that they resist even solvent extraction. A basic understanding of the mechanism of Pag e 386 FIGURE 24 Correlations between foaming power and averag e hydrophobicity (A), and foam stability and charg e density (B) of proteins. See orig inal article for identities of the proteins; 1 kcal = 4. 18 kJ. (From Ref. 106.) Pag e 387 binding of off flavors to proteins is needed so that appropriate methods can be developed for their removal. The flavor-binding property of proteins also has desirable aspects, because they can be used as flavor carriers or flavor modifiers in fabricated foods. This is particularly useful in meat analogues containing plant proteins, where successful simulation of a meat-like flavor is essential for consumer acceptance. In order for a protein to function as a good flavor carrier, it should bind flavors tightly, retain them during processing, and release them during mastication of food in the mouth. However, proteins do not bind all flavor compounds with equal affinity. This leads to uneven and disproportionate retention of some flavors and undesirable losses during processing. Because protein-bound flavorants do not contribute to taste and aroma unless they are released readily in the mouth, knowledge of the mechanisms of interaction and binding affinity of various flavorants is essential if effective strategies for producing flavor-protein products or for removing off flavors from protein isolates are to be devised. 6.5.4.1 Thermodynamics of Protein-Flavor Interactions In water-flavor model systems, addition of proteins has been shown to reduce the headspace concentration of flavor compounds [35]. This is attributed to binding of flavors to proteins. The mechanism of flavor binding to proteins depends upon the moisture content of the protein sample, but interactions are normally noncovalent. Dry protein powders bind flavors mainly via van der Waals, hydrogen bonding, and electrostatic interactions. Physical entrapment within capillaries and crevices of dry protein powders may also contribute to flavor properties of dry protein powders. In liquid or high-moisture foods, the mechanism of flavor binding by proteins primarily involves interaction of the nonpolar ligand with hydrophobic patches or cavities on the protein surface. In addition to hydrophobic interactions, flavor compounds with polar head groups, such as hydroxyl and carboxyl groups, may also interact with proteins via hydrogen bonding and electrostatic interactions. After binding to the surface hydrophobic regions, aldehydes and ketones may be able to diffuse into the hydrophobic interior of the protein molecule. Flavor interactions with proteins are usually completely reversible. However, aldehydes can covalently bind to the amino group of lysyl side chains, and this interaction is nonreversible. Only the noncovalently bound fraction can contribute to aroma and taste of the protein product. The extent of flavor binding by hydrated proteins depends on the number of hydrophobic binding regions available on the protein surface. The binding sites are usually made up of groups of hydrophobic residues segregated in the form of a well-defined cavity. Single nonpolar residues on the protein surface are less likely to act as binding sites. Under equilibrium conditions, the reversible noncovalent binding of a flavor compound with proteins follows the Scatchard equation, (60) where n is moles of ligand bound per mole of protein, n is the total number of binding sites per mole of protein, [L] is the free ligand concentration at equilibrium, and K is the equilibrium binding constant (M-1). According to this equation, a plot of n/[L] versus n will be a straight line; the values of K and n can be obtained from the slope and the intercept, respectively. The free energy change for binding of ligand to protein is obtained from the equation DG = -RT ln K, where R is the gas constant and T is absolute temperature. The thermodynamic constants for the binding of carbonyl compounds to various proteins are presented in Table 17. The binding constant increases by about threefold for each methylene group increment in chain length, with Pag e 388 TABLE 17 Thermodynamic Constants for Binding of Carbonyl Compounds to Proteins Protein Carbonyl compound n (mol/mol) K (M-1) DG (kJ/mol) Serum albumin 2-Nonanone 6 1800 -18.4 2-Heptanone 6 270 -13.8 b-Lactog lobulin 2-Heptanone 2 150 -12.4 2-Octanone 2 480 -15.3 2-Nonanone 2 2440 -19.3 Soy protein Native 2-Heptanone 4 110 -11.6 2-Octanone 4 310 -14.2 2-Nonanone 4 930 -16.9 5-Nonanone 4 541 -15.5 Nonanal 4 1094 -17.3 Partially denatured 2-Nonanone 4 1240 -17.6 Succinylated 2-Nonanone 2 850 -16.7 Note: n, number of binding sites in native state; K, equilibrium binding constant. Source: Refs. 24, 25, 83. a corresponding free energy change of -2.3 kJ/mol per CH2 group. This indicates that the binding is hydrophobic in nature. It is assumed in the Scatchard relationship that all ligand binding sites in a protein have the same affinity, and that no conformational changes occur upon binding of the ligand to these sites. Contrary to the latter assumption, proteins generally do undergo a conformational change upon binding of flavor compounds. Diffusion of flavor compounds into the interior of the protein may disrupt hydrophobic interactions between protein segments, and thus destabilize the protein structure. Flavor ligands with reactive groups, such as aldehydes, can covalently bind the e-amino groups of lysyl residues, change the net charge of the protein, and thus cause protein unfolding. Unfolding generally results in exposure of new hydrophobic sites for ligand binding. Because of these structural changes, Scatchard plots for protein are generally curvilinear. In the case of oligomeric proteins, such as soy proteins, conformational changes may involve both dissociation and unfolding of subunits. Denatured proteins generally exhibit a large number of binding sites with weak association constants. Methods for measuring flavor binding can be found in References 24 and 25. 6.5.4.2 Factors Influencing Flavor Binding Since volatile flavors interact with hydrated proteins mainly via hydrophobic interactions, any factor that affects hydrophobic interactions or surface hydrophobicity of proteins will influence flavor binding. Temperature has very little effect on flavor binding, unless significant thermal unfolding of the protein occurs. This is because the association process is primarily entropy driven, not enthalpy driven. Thermally denatured proteins exhibit increased ability to bind flavors; however, the binding constant is usually low compared to that of native protein. The effect of salts on flavor binding are related to their salting-in and salting-out properties. Salting-in-type salts, which destabilize hydrophobic interactions, decrease flavor binding, whereas salting-out-type salts increase flavor binding. The effect of pH on flavor binding is generally related to pH-induced conformational changes in proteins. Flavor binding is Pag e 389 usually enhanced more at alkaline pH that at acid pH; this is because proteins tend to denature more extensively at alkaline pH than at acid pH. Breakage of protein disulfide bonds, which causes unfolding of proteins, usually increases flavor binding. Extensive proteolysis, which disrupts and decreases the number of hydrophobic regions, decreases flavor binding. This can be used as a way of removing off flavors from oilseed proteins. However, protein hydrolysis sometimes liberates bitter peptides. Bitterness of peptides is often related to hydrophobicity. Peptides with a mean hydrophobicity of less than 5.3 kJ/mol do not have a bitter taste. On the other hand, peptides with a mean hydrophobicity of greater than 5.85 kJ/mol often are bitter. Formation of bitter peptides in protein hydrolysates depends on the amino acid composition and sequence, and the type of enzyme used. Caseins and soy proteins hydrolyzed with several commercial proteases result in bitter peptides. The bitterness can be reduced or eliminated by using endo- and exo-peptidases, which further breakdown bitter peptides into fragments that have less than 5.3 kJ/mol mean hydrophobicity. 6.5.5 Viscosity The consumer acceptability of several liquid and semisolid-type foods (e.g., gravies, soups, beverages, etc.) depends on the viscosity or consistency of the product. The viscosity of a solution relates to its resistance to flow under an applied force (or shear stress). For an ideal solution, the shear stress (i.e, force per unit area, F/A) is directly proportional to the shear rate (i.e., the velocity gradient between the layers of the liquid, dv/dr). This is expressed as (61) The proportionality constant h is known as the viscosity coefficient. Fluids that obey this expression are called Newtonian fluids. The flow behavior of solutions is greatly influenced by solute type. High-molecular-weight soluble polymers greatly increase viscosity even at very low concentrations. This again depends on several molecular properties such as size, shape, flexibility, and hydration. Solutions of randomly coiled macromolecules display greater viscosity than do solutions of compact folded macromolecules of the same molecular weight. Most macromolecular solutions, including protein solutions, do not display Newtonian behavior, especially at high protein concentrations. For these systems, the viscosity coefficient decreases when the shear rate increases. This behavior is known as pseudoplastic or shear thinning, and follows the relationship (62) where m is the consistency coefficient and n is an exponent known as the “flow behavior index.” The pseudoplastic behavior of protein solutions arises because of the tendency of protein molecules to orient their major axes in the direction of flow. Dissociation of weakly held dimers and oligomers into monomers also contributes to shear thinning. When shearing or flow is stopped, the viscosity may or may not return to the original value, depending on the rate of relaxation of the protein molecules to random orientation. Solutions of fibrous proteins, such as gelatin and actomyosin, usually remain oriented, and thus do not quickly regain their original viscosity. On the other hand, solutions of globular proteins, such as soy proteins and whey proteins, rapidly regain their viscosity when flow is stopped. Such solutions are called thixotropic. Pag e 390 The viscosity (or consistency) coefficient of most protein solutions follows an exponential relationship with protein concentration because of both protein-protein interactions and interactions between the hydration spheres of protein molecules. An example involving soy protein fractions is shown in Figure 25 [91]. At high protein concentrations or in protein gels, where proteinprotein interactions are numerous and strong, proteins display plastic viscoelastic behavior. In these cases, a specific amount of force, known as “yield stress,” is required to initiate flow. The viscosity behavior of proteins is a manifestation of complex interactions among several variables, including size, shape, protein-solvent interactions, hydrodynamic volume, and molecular flexibility in the hydrated state. When dissolved in water, proteins absorb water and swell. The volume of the hydrated molecules, that is, their hydrodynamic size or volume, is much larger than their unhydrated size or volume. The protein-associated water induces long-range effects on the flow behavior of the solvent. The dependence of viscosity on shape and size of protein molecules follows the relationship hsp = bC(n2 + d1n1) (63) where hsp is specific viscosity, b is the shape factor, C is concentration, n2 and n1 are specific volumes of unhydrated protein and solvent, respectively, and d1 is grams of water bound per gram of protein. Here, n2 is also related to molecular flexibility; the greater the specific volume of the protein, the greater its flexibility. The viscosity of dilute protein solutions is expressed in several ways. Relative viscosity, hrel refers to the ratio of viscosity of the protein solution to that of the solvent. It is measured in an Ostwald-Fenske type capillary viscometer, and is expressed as (64) where r and ro are densities of protein solution and solvent, respectively, and t and to are times of flow for a given volume of protein solution and solvent, respectively, through the capillary. Other forms of expressing viscosity can be obtained from the relative viscosity. Specific viscosity is defined as hsp = hrel – 1 (65) Reduced viscosity is (66) where C is the protein concentration, and intrinsic viscosity is (67) The intrinsic viscosity, [h], is obtained by extrapolating a plot of reduced viscosity versus protein concentration to zero protein concentration (Lim). Since protein-protein interactions are non-existent at infinite dilution, intrinsic viscosity accurately depicts the effects of shape and size on the flow behavior of individual protein molecules. Changes in the hydrodynamic shape of proteins that result from heat and pH treatment can be studied by measuring their intrinsic viscosities. 6.5.6 Gelation A gel is an intermediate phase between a solid and a liquid. Technically, it is defined as “a substantially diluted system which exhibits no steady state flow” [33]. It is made up of polymers cross-linked via either covalent or noncovalent bonds to form a network that is capable of entrapping water and other low-molecular-weight substances (see Chap. 3). Protein gelation refers to transformation of a protein from the “sol” state to a “gel-like” state. This transformation is facilitated by heat, enzymes, or divalent cations under appropriate conditions. All these agents induce formation of a network structure; however, the types of covalent and noncovalent interactions involved, and the mechanism of network formation can differ considerably. Most food protein gels are prepared by heating a protein solution. In this mode of gelation, the protein in a sol state is first transformed into a “progel” state by denaturation. The progel state is usually a viscous liquid state in which some degree of protein polymerization has already occurred. This step causes unfolding of the protein and exposure of a critical number of functional groups, such as hydrogen bonding and hydrophobic groups, so that the second stage, formation of a protein network, can occur. Creation of the progel is irreversible because many protein-protein interactions occur between the unfolded molecules. When the progel is cooled to ambient or refrigeration temperature, the decrease in the thermal kinetic energy facilitates Pag e 392 formation of stable noncovalent bonds among exposed functional groups of the various molecules and this constitutes gelation. The interactions involved in network formation are primarily hydrogen bonds, and hydrophobic and electrostatic interactions. The relative contributions of these forces vary with the type of protein, heating conditions, the extent of denaturation, and environmental conditions. Hydrogen bonding and hydrophobic interactions contribute more than electrostatic interactions to network formation except when multivalent ions are involved in cross-linking. Some proteins generally carry a net charge, electrostatic repulsion occurs among protein molecules and this is not usually conducive to network formation. However, charged groups are essential for maintaining protein-water interactions and water holding capacity of gels. Gel networks that are sustained primarily by noncovalent interactions are thermally reversible; that is, upon reheating they will melt to a progel state, as is commonly observed with gelatin gels. This is especially true when hydrogen bonds are the major contributor to the network. Since hydrophobic interactions are strong at elevated temperatures, gel networks formed by hydrophobic interactions are irreversible, such as egg-white gels. Proteins that contain both cysteine and cystine groups can undergo polymerization via sulfhydryl-disulfide interchange reactions during heating, and form a continuous covalent network upon cooling. Such gels are usually thermally irreversible. Examples of gels of this type are ovalbumin, b-lactoglobulin, and whey protein gels. Proteins form two types of gels, namely, coagulum (opaque) gels and translucent gels. The type of gel formed by a protein is dictated by its molecular properties and solution conditions. Proteins containing large amounts of nonpolar amino acid residues undergo (68) (PN is native state, PD is unfolded state, and n is the number of protein molecules taking part in cross-linking) hydrophobic aggregation upon denaturation. These insoluble aggregates then randomly associate and set into an irreversible coagulum-type gel. Since the rate of aggregation and network formation is faster than the rate of denaturation, proteins of this type readily set into a gel network even while being heated. The opaqueness of these gels is due to light scattering caused by the unordered network of insoluble protein aggregates. Proteins that contain small amounts of nonpolar amino acid residues form soluble complexes upon denaturation. Since the rate of association of these soluble complexes is slower than the rate of denaturation, and the gel network is predominantly formed by hydrogen bonding interactions, they often do not set into a gel until heating followed cooling has occurred (8–12% protein concentration assumed). Upon cooling, the slow rate of association of the soluble complexes facilitates formation of an ordered translucent gel network. At the molecular level, coagulum-type gels tend to form when the sum of Val, Pro, Leu, Ile, Phe, and Trp residues of the protein exceeds 31.5 mole percent [100]. Those that contain less Pag e 393 than 31.5 mole percent of the above hydrophobic residues usually form translucent gels when water is the solvent. There are several exceptions to this empirical rule. For example, the hydrophobic amino acid content of b-lactoblobulin is 32 mole percent, yet it forms a translucent gel in water. However, when NaCl is included, it forms a coagulum-type gel even when the salt concentration is as low as 50 mM. This occurs because of charge neutralization by NaCl, which promotes hydrophobic aggregation upon heating. Thus, gelation mechanism and gel appearance are fundamentally controlled by the balance between attractive hydrophobic interactions and repulsive electrostatic interactions. These two forces in effect control the balance between protein-protein and protein-solvent interactions in a gelling system. If the former is much greater than the latter, a precipitate or a coagulum is likely to form. If protein-solvent interactions predominate, the system may not gel. A coagulum gel or a translucent gel results when the hydrophobic and hydrophilic forces are somewhere in-between these two extremes. Protein gels are highly hydrated systems, containing up to 98% water. The water entrapped in these gels has chemical potential (activity) similar to that in dilute aqueous solutions, but lacks fluidity and cannot be easily squeezed out. The mechanisms by which liquid water can be held in a nonflowable semisolid state in gels are not well understood. The fact that translucent gels, formed primarily by hydrogen-bonding interactions, hold more water than coagulum-type gels and are less prone to syneresis suggests that much of the water is hydrogen bonded to C=O and N-H groups of the peptide bonds, is associated with charged groups in the form of hydration shells, and/or exists in extensively hydrogen-bonded water-water networks. It is also possible that within the restricted environment of each cell of the gel structure, water may act as an hydrogen-bonding cross-linker between C=O and N-H groups of peptide segments (see Chap. 2). This may restrict the flowability of water within each cell, the more so as the cell size decreases. It is also likely that some water may be held as capillary water in the pores of the gel structure, especially in coagulum gels. The stability of a gel network against thermal and mechanical forces is dependent on the number of types of cross-links formed per monomer chain. Thermodynamically, a gel network would be stable only when the sum of the interaction energies of a monomer in the gel network is greater than its thermal kinetic energy. This is dependent on several intrinsic (such as the size, net charge, etc.) and extrinsic factors (such as pH, temperature, ionic strength, etc.). The square root of the hardness of protein gels exhibits a linear relationship with molecular weight [110]. Globular proteins with molecular weight < 23,000 cannot form a heatinduced gel at any reasonable protein concentration, unless they contain at least one free sulfhydryl group or a disulfide bond. The sulfhydryl groups and disulfide bonds facilitate polymerization, and thus increase the effective molecular weight of polypeptides to > 23,000. Gelatin preparations with effective molecular weights of less than 20,000 cannot form a gel. Another critical factor is protein concentration. To form a self-standing gel network, a minimum protein concentration, known as least concentration endpoint (LCE), is required. The LCE is 8% for soy proteins, 3% for egg albumin, and about 0.6% for gelatin. Above this mimimum concentration, the relationship between gel strength, G, and protein concentration, C, usually follows a power law, (69) where Co is the LCE. For proteins, the value of n varies from 1 to 2. Several environmental factors, such as pH, salts, and other additives, also affect gelation of proteins. At or near isoelectric pH, proteins usually form coagulum-type gels. At extremes of pH, weak gels are formed because of strong electrostatic repulsion. The optimum pH for gel formation is about 7–8 for most proteins. Pag e 394 Formation of protein gels can sometimes be facilitated by limited proteolysis. A well-known example is cheese. Addition of chymosin (rennin) to casein micelles in milk results in the formation of a coagulum-type gel. This is achieved by cleavage of kcasein, a micelle component, causing release of a hydrophilic portion, known as the glycomacropeptide. The remaining so-called para-casein micelles possess a highly hydrophobic surface that facilitates formation of a weak gel network. Enzymic cross-linking of proteins at room temperature can also result in formation of a gel network. Transglutaminase is the enzyme often used to prepare these gels. This enzyme catalyzes formation of e-(g-glutamyl)lysyl cross-links between the glutamine and lysyl groups of protein molecules [77]. Using this enzymic cross-linking method, highly elastic and irreversible gels can be formed even at low protein concentration. Divalent cations, such as Ca2+ and Mg2+, can also be used to form protein gels. These ions form cross-links between negatively charged groups of protein molecules. A good example of this type of gel is tofu from soy proteins. Alginate gels also can be formed by this means. 6.5.7 Dough Formation [67,68] Among food proteins, wheat protein is very unique because of its exceptional ability to form a viscoelastic dough. When a mixture of wheat flour and water (about 3:1 ratio) is kneaded, a dough with viscoelastic properties forms that is suitable for making bread and other bakery products. These unusual dough characteristics are mainly attributable to the proteins in wheat flour. Wheat flour contains soluble and insoluble protein fractions. The soluble proteins, comprising about 20% of the total proteins, are primarily albumin- and globulin-type enzymes and certain minor glycoproteins. These proteins do not contribute to the dough-forming properties of wheat flour. The major storage protein of wheat is gluten. Gluten is a heterogeneous mixture of proteins, mainly gliadins and glutenins, with limited solubility in water. The formation of a viscoelastic dough capable of entrapping gas during fermentation is attributed entirely to the gluten proteins. Gluten has a unique amino acid composition, with Glu/Gln and Pro accounting for more than 50% of the amino acid residues. The low water solubility of gluten is attributable to its low content of Lys, Arg, Glu, and Asp residues, which together amount to less than 10% of the total amino acid residues. About 30% of gluten’s amino acid residues are hydrophobic, and the residues contribute greatly to its ability to form protein aggregates by hydrophobic interactions and to bind lipids and other nonpolar substances. The high glutamine and hydroxyl amino acid (~10%) contents of gluten is responsible for its water binding properties. In addition, hydrogen bonding between glutamine and hydroxyl residues of gluten polypeptides contributes to its cohesion-adhesion properties. Cysteine and cystine residues account for 2–3% of gluten’s total amino acid residues. During formation of the dough, these residues undergo sulfhydryl-disulfide interchange reactions resulting in extensive polymerization of gluten proteins. Several physical and chemical transformations occur during mixing and kneading of a mixture of wheat flour and water. Under the applied shear and tensile forces, gluten proteins absorb water and partially unfold. The partial unfolding of protein molecules facilitates hydrophobic interactions and sulfhydryl-disulfide interchange reactions, which result in formation of thread-like polymers. These linear polymers in turn are believed to interact with each other, presumably via hydrogen bonding, hydrophobic associations, and disulfide cross-linking, to form a sheet-like film capable of entrapping gas. Because of these transformations in gluten, the resistance of the dough increases with time until a maximum is reached, and this is followed by a decrease in resistance, indicative of a breakdown in the network structure. The breakdown involves alignment of polymers in the direction of shear and some scission of disulfide Pag e 395 cross-links, which reduces the polymer size. The time it takes to reach maximum dough strength during kneading is used as a measure of wheat quality for bread making—a longer time indicating better quality. The development of a viscoelastic dough is thought to be related to the extent of sulfhydryl-disulfide interchange reactions. This is supported by the fact that when reductants, such as cysteine, or sulfhydryl blocking agents, such as N-ethylmaleimide, are added to dough, viscoelasticity decreases greatly. On the other hand, addition of oxidizing agents, such as iodates and bromates, increase the elasticity of the dough. This implies that wheat varieties that are rich is SH and S-S groups might possess superior bread-making qualities, but this relationship is unreliable. Thus, the role of sulfhydryl-disulfide interchange reactions in the development of viscoelastic doughs is poorly understood. Differences in bread-making qualities of different wheat cultivars may be related to the differences in the structure of gluten itself. As mentioned earlier, gluten is made up of gliadins and glutenins. Gliadins are comprised of four groups, namely a, b, g, and wgliadins. In gluten these exist as single polypeptides with molecular weight ranging from 30,000 to 80,000. Although gliadins contain about 2–3% half-cystine residues, they apparently do not undergo extensive polymerization via sulfhydryl-disulfide interchange reactions. The disulfide bonds appear to remain as intramolecular disulfides during dough making. Dough made from isolated gliadins and starch is viscous but not viscoelastic. Glutenins, on the other hand, are heterogeneous polypeptides with molecular weight ranging from 12,000 to 130,000. These are further classified into high-molecular-weight (M.W. >90,000, HMW) and low-molecular-weight (M.W. <90,000, LMW) glutenins. In gluten, these glutenin polypeptides are present as polymers joined by disulfide cross-links, with molecular weights ranging into the millions. Because of their ability to polymerize extensively via sulfhydryl-disulfide interchange reactions, glutenins contribute greatly to the elasticity of dough. Therefore, an optimum ratio of gliadins and glutenins seems to be necessary to form a viscoelastic dough. Some studies have shown a significant positive correlation between HMW glutenin content and breadmaking quality in some wheat varieties, but not in others. Available information indicates that a specific pattern of disulfide crosslinked association among LMW and HMW glutenins in gluten structure may be far more important to bread quality than the amount of this protein. For example, association/polymerization among LMW glutenins gives rise to a structure similar to that formed by HMW gliadin. This type of structure contributes to viscosity of the dough, but not to elasticity. In contrast, if LMW glutenins link to HMW glutenins via disulfide cross-links (in gluten), then this is believed to contribute to dough elasticity. It is possible that in good-quality wheat varieties, more of the LMW glutenins may polymerize with HMW, whereas in poor-quality wheat varieties, most of the LMW glutenins may polymerize among themselves. These differences in associated states of glutenins in gluten of various wheat varieties may be related to differences in their conformational properties, such as surface hydrophobicity, and reactivity of sulfhydryl and disulfide groups. In summary, hydrogen bonding among amide and hydroxyl groups, hydrophobic interactions, and sulfhydryl-disulfide interchange reactions all contribute to the development of the unique viscoelastic properties of wheat dough. However, culmination of these interactions into good dough-making properties may depend on the structural properties of each protein and the proteins with which it associates in the overall gluten structure. Because polypeptides of gluten, especially the glutenins, are rich in proline, they have very little folded structure. Whatever folded structure initially exists in gliadins and glutenins is lost during mixing and kneading. Therefore, no additional unfolding occurs during baking. Supplementation of wheat flour with albumin- and globulin-type proteins, such as whey proteins and soy proteins, adversely affects the viscoelastic properties and baking quality of thedough. These proteins decrease bread volume by interfering with formation of the gluten network. Addition of phospholipids or other surfactants to dough counters the adverse effects of foreign proteins on loaf volume. In this case, the surfactant/protein film compensates for the impaired gluten film. Although this approach results in acceptable loaf volume, the textural and sensory qualities of the bread are less desirable than normal. Isolated gluten is sometimes used as a protein ingredient in nonbakery products. Its cohesion-adhesion properties make it an effective binder in comminuted meat and surimitype products. 6.6 Nutritional Properties of Proteins Proteins differ in their nutritive value. Several factors, such as content of essential amino acids and digestibility, contribute to these differences. The daily protein requirement therefore depends on the type and composition of proteins in a diet. 6.6.1 Protein Quality The “quality” of a protein is related mainly to its essential amino acid composition and digestibility. High-quality proteins are those that contain all the essential amino acids at levels greater than the FAO/WHO/UNU [30] reference levels, and a digestibility comparable to or better than those of egg-white or milk proteins. Animal proteins are better “quality” than plant proteins. Proteins of major cereals and legumes are often deficient in at least one of the essential amino acids. While proteins of cereals, such as rice, wheat, barley, and maize, are very low in lysine and rich in methionine, those of legumes and oilseeds are deficient in methionine and rich or adequate in lysine. Some oilseed proteins, such as peanut protein, are deficient in both methionine and lysine contents. The essential amino acids whose concentrations in a protein are below the levels of a reference protein are termed limiting amino acids. Adults consuming only cereal proteins or legume proteins have difficulty maintaining their health; children below 12 years of age on diets containing only one of these protein sources cannot maintain a normal rate of growth. The essential amino acid contents of various food proteins are listed in Table 18. Both animal and plant proteins generally contain adequate or more than adequate amounts of His, Ile, Leu, Phe + Tyr, and Val. These amino acids usually are not limiting in staple foods. More often, either Lys, Thr, Trp, or the sulfur-containing amino acids are the limiting amino acids. The nutritional quality of a protein that is deficient in an essential amino acid can be improved by mixing it with another protein that is rich in that essential amino acid. For example, mixing of cereal proteins with legume proteins provides a complete and balanced level of essential amino acids. Thus, diets containing appropriate amounts of cereals and legumes (pulses) and otherwise nutritionally complete are often adequate to support growth and maintenance. The poor-quality protein also can be nutritionally improved by supplementing it with essential free amino acids that are under-represented. Supplementation of legumes with Met and of cereals with Lys usually improves their quality. The nutritional quality of a protein or protein mixture is ideal when it contains all of the essential amino acids in proportions that produce optimum rates of growth and/or optimum maintenance capability. The ideal essential amino acid patterns for children and adults are given in Table 19. However, because actual essential amino acid requirements of individuals in a given population vary depending on their nutritional and physiological status, the essential amino acid requirements of preschool children (age 2–5) are generally recommended as a safe level for all age groups [31]. Overconsumption of any particular amino acid can lead to “amino acid antagonism” or Pag e 397 TABLE 18 Essential Amino Acid Contents and Nutritional Value of Proteins from Various Sources (mg /g Protein) Protein source Property (mg /g protein) Egg Cow's milk Beef Fish W heat Rice Maize Barley Soybean Field bean (boiled) Pea Peanut French bean Amino acid concentration (mg /g protein) His 22 27 34 35 21 21 27 20 30 26 26 27 30 Ile 54 47 48 48 34 40 34 35 51 41 41 40 45 Leu 86 95 81 77 69 77 127 67 82 71 70 74 78 Lys 70 78 89 91 23ª 34ª 25ª 32ª 68 63 71 39ª 65 Met + Cys 57 33 40 40 36 49 41 37 33 22b 24b 32 26 Phe + Tyr 93 102 80 76 77 94 85 79 95 69 76 100 83 Thr 47 44 46 46 28 34 32b 29b 41 33 36 29b 40 Trp 17 14 12 11 10 11 6b 11 14 8ª 9ª 11 11 Val 66 64 50 61 38 54 45 46 52 46 41 48 52 Total essential amino acids 512 504 480 485 336 414 422 356 466 379 394 400 430 Protein content, % 12 3.5 18 19 12 7.5 — — 40 32 28 30 30 Chemical score (%) (based on FAO/W HO [30] pattern) 100 100 100 100 40 59 43 55 100 73 82 67 PER 3.9 3.1 3.0 3.5 1.5 2.0 — — 2.3 — 2.65 — — BV (on rats) 94 84 74 76 65 73 — — 73 — — — — NPU 94 82 67 79 40 70 — — 61 — — — — Note: Chemical score is defined as the ratio of the amount of a limiting essential amino acid in 1 g of a test protein to the amount of the same amino acid in 1 g of a reference protein. PER, protein efficiency ratio; BV, biolog ical value; NPU, net protein utilization. ªPrimary limiting acid. bSecond limiting acid. Source: Refs. 28 and 30. Pag e 398 TABLE 19 Recommended Essential Amino Acid Pattern for Food Proteins Recommended pattern (mg /g protein) Amino acid Infant (2–5 years) Preschool child (10–12 years) Preschool child Adult Histidine 26 19 19 16 Isoleucine 46 28 28 13 Leucine 93 66 44 19 Lysine 66 58 44 16 Met + Cys 42 25 22 17 Phe + Tyr 72 63 22 19 Threonine 43 34 28 9 Tryptophan 17 11 9 5 Valine 55 35 25 13 Total 434 320 222 111 Source: Ref. 30. toxicity. Excessive intake of one amino acid often results in an increased requirement for other essential amino acids. This is due to competition among amino acids for absorption sites on the intestinal mucosa. For example, when the Leu level is relatively high, this decreases absorption of Ile, Val, and Tyr even if the dietary levels of these amino acids are adequate. This leads to an increased dietary requirement for the latter three amino acids. Overconsumption of other essential amino acids also can inhibit growth and induce pathological conditions. 6.6.2 Digestibility Digestibility is defined as the proportion of food nitrogen that is absorbed after ingestion. Although the content of essential amino acids is the primary indicator of protein quality, true quality also depends on the extent to which these amino acids are utilized in the body. Thus, digestibility of amino acids can affect the quality of proteins. Digestibilities of various proteins in humans are listed in Table 20. Food proteins of animal origin are more completely digested than those of plant origin. Several factors affect digestibility of proteins. TABLE 20 Dig estibility of Various Food Proteins in Humans Protein source Dig estibility (%) Protein source Dig estibility (%) Egg 97 Millet 79 Milk, cheese 95 Peas 88 Meat, fish 94 Peanut 94 Maize 85 Soy flour 86 Rice (polished) 88 Soy protein isolate 95 W heat, whole 86 Beans 78 W heat flour, white 96 Corn, cereal 70 W heat g luten 99 W heat, cereal 77 Oatmeal 86 Rice cereal 75 Source: Ref. 30. Pag e 399 1. Protein conformation: The structural state of a protein influences its hydrolysis by proteases. Native proteins are generally less completely hydrolyzed than partially denatured ones. For example, treatment of phaseolin (a protein from kidney beans) with a mixture of proteases results only in limited cleavage of the protein, resulting in liberation of a 22,000 MW polypeptide as the main product. When heat-denatured phaseolin is treated under similar conditions, it is completely hydrolyzed to amino acids an dipeptides. Generally, insoluble fibrous proteins and extensively denatured globular proteins are difficult to hydrolyze. 2. Antinutritional factors: Most plant protein isolates and concentrates contain trypsin and chymotrypsin inhibitors (Kunitz type and Bowman-Birk type) and lectins. These inhibitors impair complete hydrolysis of legume and oilseed proteins by pancreatic proteases. Lectins, which are glycoproteins, bind to intestinal mucosa cells and interfere with absorption of amino acids. Lectins and Kunitz-type protease inhibitors are thermolabile, whereas the Bowman-Birk type inhibitor is stable under normal thermal processing conditions. Thus, heat-treated legume and oilseed proteins are generally more digestible than native protein isolates (despite some residual Bowman-Birk type inhibitor). Plant proteins also contain other antinutritional factors, such as tannins and phytate. Tannins, which are condensed products of polyphenols, covalently react with e-amino groups of lysyl residues. This inhibits trypsin-catalyzed cleavage of the lysyl peptide bond. 3. Binding: Interaction of proteins with polysaccharides and dietary fiber also reduces the rate and completeness of hydrolysis. 4. Processing: Proteins undergo several chemical alterations involving lysyl residues when exposed to high temperatures and alkaline pH. Such alterations reduce their digestibility. Reaction of reducing sugars with e-amino groups also decreases digestibility of lysine. 6.6.3 Evaluation of Protein Nutritive Value Since the nutritional quality of proteins can vary greatly and is affected by many factors, it is important to have procedures for evaluating quality. Quality estimates are useful for (a) determining the amount required to provide a safe level of essential amino acids for growth and maintenance, and (b) monitoring changes in the nutritive value of proteins during food processing, so that processing conditions that minimize quality loss can be devised. The nutritive quality of proteins can be evaluated by several biological, chemical, and enzymatic methods. Biological Methods Biological methods are based on weight gain or nitrogen retention in test animals when fed a protein-containing diet. A proteinfree diet is used as the control. The protocol recommended by FAO/WHO [31] is generally used for evaluating protein quality. Rats are the usual test animal, although humans are sometimes used. A diet containing about 10% protein on a dry weight basis is used to ensure that the protein intake is below daily requirements. Adequate energy is supplied in the diet. Under these conditions, protein in the diet is used to the maximum possible extent for growth. The number of test animals used must be sufficient to assure results that are statistically reliable. A test period of 9 days is common. During each day of test period, the amount (g) of diet consumed is tabulated for each animal, and the feces and urine are collected for nitrogen analysis. The data from animal feeding studies are used in several ways to evaluate protein quality

Leave a Reply

Your email address will not be published.