A major world concern in fulfilling the food need is to supply the protein required by the whole population now and in the future. Legumes such as peas, beans, and lentils are high in protein content and have been used as an inexpensive protein source in the diets where animal proteins are either unaffordable or are considered detrimental to the health and nutrition conscience population (Nutrition News, 1991). Oil seeds such as cotton seed, canola seed, sunflower seed, also has been investigated as a source for human food protein.

Legumes and seeds as protein sources, are used as flours in products such as baby formula or supplemental diet for preschool children (Jansen and Harper, 1980; Akinyele et al., 1988), baking products (Mustafa et al, 1986; Guy, 1984), pastas (Bahnassey et al, 1986; Molina et al, 1982) or extruded products (Aguilera and Kosikowski, 1976; Ringe and Love, 1988; Likimani et al, 1991). Since the flour form is needed in larger amount to reach the same protein level than if using concentrate or isolate form, there had been an attempt to extract protein from legumes and seeds. This extracted protein generally has a high protein concentration, provides nutritional quality due to lower antinutritional factors with minimal off odor and color. The concentrate and isolate forms are mostly used widely as an ingredient due to their functional properties (Kinsella, 1979) in food systems. Since every storage protein of seeds has its unique characteristics, these functional properties can be optimized by knowing the chemical and physicochemical characteristics of these proteins. Characterization is the initial step for modification and manipulation according to the purpose and need.

Uses Of Vegetable Protein

Soybean and other legume proteins have been used widely in replacing parts of wheat flour in bread and cookies. Sathe et al (1981) used Northern bean flour or bean protein concentrates in bread mixtures. Results indicated that addition of up to 10% protein gave better dough and bread baking quality than that with bean flour, due to higher water absorption and less mixing time.

Thompson (1977) evaluated the fortification of mungbean flour and proteins isolate into wheat flour bread. Additions of 10% isolate which gave 20.5% protein level still produce acceptable bread. To reach the same protein level, mungbean flour had to be added at 40% and produced unacceptable bread. Addition of stearoyl-2-lactylate could improve loaf volume; however, texture and color were still inferior to the one with isolate.

Yellow pea, lentil, and faba bean protein isolate both from germinated and ungerminated seeds, had been used to replace 3, 5, or 8% wheat flour in bread making (Hsu et al, 1982). All products were acceptable, except the ones with 8% protein isolate. Bread flour mixed with 12% soy flour and 1-3% whey solid (Guy, 1984) had an acceptable product and had 2 weeks long shelf life, after incorporation of 0.5% sodium steroyl lactylate in the mixture. Fortified bread and cookies with cowpea flour or its protein isolates at 10 or 20%, respectively, gave acceptable products, though 20% isolate gave high spreading ratio (Mustata et al, 1986).

Lorimer et al (1991) studied the effect of globular proteins from navy bean (5 and 10%) substitution on wheat flour dough. The results showed that both high protein fraciton bean flour and globular proteins increased absorption and arrival time in farinograph measures, but decreased stability. Combined with scanning electron micrograph evaluation and SH-SS interchange, there was still not enough evidence that bean flour or globular prtein disrupted the S-S interchange.

Defatted soybean and soy isolate has been added into wheat flour tortilla up to 15.5% protein content (Gonzalez-Agramon and Serna-Saldivar, 1988). Protein Efficiency Ratio was increased and the products had higher caloric energy than the control. Fortification caused increased water absorption and decreased dough elasticity, but no adverse effect on product quality.

Spaghetti has also been fortified with edible roasted and unroasted legumes, and their protein concentrates, such as navy bean, pinto bean, and lentils (Bahnassey et al, 1986). The fortified products had better balance of lysine, SAA, and mineral content than the control. Addition up to 10% legumes decreased yellow color but was still acceptable. Pasta made of semolina-corn mixes was fortified with soy flour at 8% or 0.3%, respectively (Molina et al, 1982). Resistance to disintegration and sensory evaluation indicated that quality decreased significantly when semolina in the blend was lower than 40%.

Extruded rice flour fortified with soy isolate (Noguchi et al, 1981) had lower water absorption capacity, but less protein insolubilization compared to nonfortified extrusions. This was likely because of a sparing effect between rice flour and soy protein isolate when processed simultaneously. While the protein isolate formed new noncovalent interactions and disulfide bonding when extruded by itself resulting in decreased protein solubility. Densitograms of PAGE and SEM photographs revealed that 7S subunit had increased noncovalent interactions and disulfide bonding and formed fine strings in the starch matrix, resulting in decreased of protein solubility, compared to 11S fraction.

In cowpea- corn flour extruded mixes (Ringe and Love, 1988), there was increase in lysine loss with increased storage temperature and time for all water activity levels. Losses were detected before the non-enzymatic brown pigment formed to be measured analytically. Nutrient composition of dry extruded cowpea products was studied (Akinyele et al, 1988) in an infant formula. Blended cowpea with rice or corn, or both, with addition of banana to form the baby formula, had average energy content 440 kcal/100 g with protein content 20%. Trypsin inhibitor activity was reduced by 63-84%, and had low oligosaccharide content.

Chickpea protein concentrate has been studied in development of special infant formula (Ulloa et all, 1988) with and without methionine supplementation. The formula had higher Relative Net Protein ratio (83.6%) compared to that of protein concentrate itself. This indicated no apparent change in protein quality during processing. In fact, extrusion cooking may improve nutritional and functional quality of the protein (Camire, 1991) through modification of the extrusion conditions. Antinutritional factors and undesirable compounds can be reduced, while improving flavors and odors.

Navy bean protein concentrate was added to retail ground beef as meat extender (Duszkiewics-Reinhard et al, 1988). Addition of protein concentrate lowered the pH of the mixture and production of lactic acid resulted in lower pathogenic or spoilage aerobic plate count and coliforms. The fat of comminuted meat-type product or bologna was substituted with tofu or soy protein concentrate (Jeng et al, 1988). Bologna with tofu had higher emulsifying capacity, tougher texture, less moisture stability and lower fat and caloric level than control or soy concentrate-bologna. Soy concentrate-bologna had more beany flavor and was more tender.

Rice et al. (1989) evaluated the effects of soy protein blends added to beef patties. They reported that combinations of textured and powdered soy proteins increased yields and nutrient retention of beef patties. Such products have been used to provide better balance of nutrients for school lunch (Engineered, 1991; Duxbury, 1991).

Seafood was extended with soy protein concentrate, structured protein fiber, and textured vegetable protein (Beuchat and Jones, 1979). In a simple system of liquid culture media, there was enhanced growth of V. parahaemolyticus by treatments with protein extender. However, V. parahaemolyticus was retarded in the ground flounder, shrimp, and crab, which were extended with 5-25% soy concentrate or structured protein fiber. Texturized vegetable protein promoted the rate of microbial growth when added to seafood at level 5 and 10%. These inhibitory and stimulatory effects were attributed to possible differences in types of competitive microflora that might be indigenous to the soy extenders and to the chemical composition and physical form.

Peanut protein isolates from 3 different varieties (Conkerton and Ory, 1976) were used as protein supplementation in pine apple juice, as an acid type beverage, at 1% level. There was no difference on flavor, texture, or aroma, although the turbidity was slightly increased.

Methods of Protein Extraction

Many factors affect the extractability of protein, including particle size and quality of flour, solvent to flour ratio, pH and temperature during extraction, ionic strength or addition of the salts into extractant (Kinsella, 1979). It is important to recover as much protein as possible during extraction, to get maximum protein content in the concentrate or isolate products. A large number of reports (Appendix 1), have evaluated extraction techniques and processing conditions in a variety of legumes and seeds. The studies analyzed yield and quality of certain fractions of proteins, besides considering cost and extraction time factor.

Effect of size and treatment of the particles on extraction yield had been studied (Aguilera and Garcia, 1989). Smaller particle size increased protein yeild, while flaked or exploded material had higher diffusion coefficients which gave higher yield. The extraction yield of whole and dehulled yellow pea flour were studied (Sumner et al., 1981). Results showed that dehulled pea flour gave lower crude fiber and higher protein contents.

A positive relationship between protein curd yield with protein recovery during solvent extraction (Wang et al., 1983), but not with the protein content of the soybeans has been reported. Alkali extraction at pH 7.0-12.0, followed by precipitation at pH 4.5 gave protein yield about 60-92% (Ulloa et al, 1988) in chick peas. Yield was highest at pH 12.0 but this required large amounts of NaOH that may cause extreme changes in environment. The NaOH can also be detrimental to protein quality.

Water and other solvent extractions (Table 1) at pH 7.0 could produce 80% or more protein yield.

Combination of low ionic strength and different pH level of extractant produced different protein fractions (Brooks and Morr, 1984; Honig et al., 1984). Low temperature (0-10oC) during extraction is important if the protein is to be evaluated for its biological activity or its intact storage protein. Most reported extractions (Appendix 1) were done at room temperature (20-25oC).

Ultrafiltration, diafiltration, and ion exchange were applied following the extraction and precipitation of isolates (Appendix 1) to produce proteins that are free from or low in glucosinolates, phytates, fiber, tannin, or other impurities, and produced the yield with very high solubility indices (Kinsella, 1979). Co-extraction and co-ultrafiltration of diary and vegetable protein blends has been used (Nichols and Cheryan, 1982) to produce milk protein replacer that has good functional properties. Industrial membrane systems (Manak et al., 1980) have been used to produce protein isolates from soy, cotton seed, and peanut. The isolates contain >90% protein with good functional properties. Temperature sensitive gels of poly iso propyl acrylamide has been used (Trank et al., 1989) to recover essentially all of the proteins, remove undesirable components, and improve nutritional and functional quality. Most extractions used flour to solvent ratios of 1:5 to 1:30, with 1 to 4 times repetitive extractions. It was revealed that 2 extractions were generally adequate (Rhee et al., 1972) since the third and fourth did not significantly increase yield. The optimum ratio reported for rape seed (Tzeng et al., 1988) or peanut (Rhee et al., 1972) was 1:20. This made it easier to handle the total amount of solvent and reduced water volume.

Protein curd was produced from field peas (Pisum sativum var. Trapper) (Gebre-Egziabher and Sumner, 1983) using 1:5 flour to water ratio at pH 8.8-9.0. This was adjusted using 0.2% CaO. The pea milk was then coagulated using 2% CaSO4 until reaching 0.54% concentration. As expected the curd from CaSO4 coagulant had a high ash content. Such coagulation would be a good source of Ca in countries where the milk supply is low or expensive.

The effects of drying method in preparing protein isolate from smooth seeded yellow pea (Pisum sativum var. Trapper) on chemical and functional properties was evaluated (Sumner et al, 1981). Alkaline extraction was employed followed by precipitation at the isoelectric point, producing isolate with 90% protein content. Drum drying, compared to freeze drying and spray drying, decreased nitrogen solubility index and increased water absorption. This was likely due to protein denaturation during increased heating. Freeze dried and spray dried isolates had the highest emulsification and water absorption values. Spray drying produced the best foaming, color, and flavor properties. The spray dried protein content may have had less Maillard reaction or polyphenol oxidation.

Characterization of Protein

Chemical characteristics and physical behavior of protein extracts include amino acid composition, protein conformation, charge distribution, molecular size, the extent of inter- and intra-molecular bonding, and are influenced by the environment such as temperature, pH and ionic strength. Functional properties of proteins are the reflection of those intrinsic physicochemical attributes and environment conditions. High quality protein ingredients must fulfill criteria, such as (Fennemma, 1985) contain high protein concentration, provide nutritional quality, have minimal off odor and color, are low amounts of toxic or antinutritional factors, are readily available at low cost, have compatibility with other ingredients under the processing conditions, and have desirable functional properties.

Chemical characterization: Amino acid composition is important to nutritional and functional quality of protein. Amino acids will contribute to the essential amino acid content of the protein, which will determine its nutritional value; or to total charge and disulfide bonds within subunit or between subunits to form polymer with flexible or rigid structure.

Soybean globulins were quantitatively analyzed (Sato et al., 1986) using densitometry in SDS-polyacrylamide slab gel and coomassie blue as a staining reagent. Conglycinin or 7S globulins contained 27.8% beta, 6.2% gamma, and 3.0% basic-conglycinin and 2.9% trypsin inhibitor. Glycinin or 11S globulins consisted of 36.5%, and other proteins were 23.6%.

Size heterogeneity of cotton seed storage protein has been characterized by SDS-PAGE gels (Marshall et al., 1984). There were 6 fractions, with the larger subunits covalently bonding, and smaller subunits of globulin fractions forming aggregates of quaternary structure that were stabilized by hydrophobic interactions. Molecular size of yellow pea proteins had been investigated in the preliminary study by this author using SDS-PAGE gel electrophoresis with 12% acrylamide. The results indicated that this protein extract consisted of 6 fractions with sharp bands and 7 fractions with lighter bands, molecular weight range from 22 kd to 81 kd. Water absorption of smaller molecular weight (15 kD) as result of bromelin treatment on the soy protein (Mohri and Matsushita, 1984) increased by 2-2.5 times, due to degradation of the 11s globulin. This degradation allows for new associations through hydrophobic and disulfide bonds. Water imbibing (Yao et al., 1988) of soy protein decreased with increasing maturity, which correlated to the increase in the ratio of 7S to 11S.

Isolation of 7S storage protein of cotton seed has been done (Zarins et al., 1984) using electrofocusing in polyacrylamide gels. The results showed that 7S (molecular weight range 15000.17000 d) consisted of basic units composed 8 components differing in charges, while larger subunits consisted of 2 or more of these subunits. Protein modification in the form of molecular size reduction in yellow pean, lentils, and faba bean after germination was evaluated using SDS-PAGE and densitometric scanner (Hsu et al., 1982). There were increases in small subunit proteins after germination. Protein isolate from those germinated legumes had higher emulsion capacity, increased foam expansion, decreased foam stability, lower viscosity, formed soft and smooth curd; except protein isolate from germinated pea or lentils, which gave severe syneresis.

Physicochemical or Functional Characterization:

Functional characteization of protein can be generalized as hydration, emulsification, textural, and rheological (Fenemma, 1985). Those characteritics can be measured through their nitrogen solubility, water absorption, viscosity, swelling, gelation, fat adsorption, foaming, whipping, adhesion, fiber/texture, aggregation, dough formation, and extrudability. Some of the measures will be discussed. Nitrogen solubilty is one aspect of hydration, which is the most important characteristic in evaluating protein quality since many functional properties of protein depend upon their capacity to go into solution initially. Solubility is affected by many factors, such as pH during extraction or solubilization. Sathe et al. (1982) reported that lupin protein extracted at alkaline pH had greater dispersibility at pH >3.0. Highest solubility at pH 12.0 was reported on winged bean protein (Okezie and Bello, 1988); while the lowest solubiltity of protein from yellow pea, lentiles, and faba bean were at pH 4.5-5.0 (Hsu et al., 1982) which were not affected by molecular size reduction due to germination.

Enzymatic hydrolysis increased the solubility of soy protein isolate (Kim et al., 1990) due to breakdown of the oligomeric structure of 7S globulin which caused alteration of charge properties. Highly denatured soy protein isolate had lower solubility due to increased free sulfhydryl content (Wagner and Anon., 1990), compared to isolate with a low degree of denaturation which had more native protein.

Gelation is another aspect of hydration and of textural and rheological properties of protein. It is defined as the formation of three dimension intermolecular networks through hydrogen, hydrophobic, and disulfide bonds that entrapped water solvent and other ingredients. This entrapment contributes to texture and chewiness of the food products (Furukawa and Ohta, 1982; Utsumi and Kinsella, 1985). The important initial step in heat-induced gelation of globular proteins is heating of the protein solution above the denaturation temperature to expose the functional groups, so that the intermolecular network can be produced. Additionally, high numbers of intermolecular disulfide bonds increased water holding capacity, and, as a result, it increased gel hardness.

The structural state of globular proteins from bovine serum albumin and soy protein in gels was studied (Wang and Damodaran,1991). The results indicated that gelation of BSA involved transconformation of alpha-helix and aperiodic structures into beta-sheet conformation. Conditions that decreased the formation of beta-sheet structure decreased gel strength. Soy proteins, which contain mainly beta-sheet and aperiodic structures in their native state, only showed a reduction in beta-sheet and an increase in aperiodic structure content in the gel state. It is hypothesized that in globular protein gels, intermolecular H-bonding between segments of beta-sheets oriented either in parallel or in antiparallel configurations may act as junction zones in the gel network.

Soybean protein consists of 4 main components; 2S, 7S, 11S, and 15S. The 7S and 11S fractions are main storage proteins and account for about 70% of the total protein content. Saio and Watanabe (1978) studied the differences in gel formation of these 7S and 11S proteins. They reported tat the 11S fraction formed harder gels with S-S bonds as the predominant bonding force and precipitated faster and formed larger aggregates. Gels made of 11S protein had higher water holding capacity, higher tensile volume, higher hardness, and expanded more on heating.

The study on proteolytic degradation of soy protein isolate (Kim et al., 1990) showed there was more than reduction of molecular size that increased thermal aggregation. 11S protein was reported as the most responsible for thermal aggregation, and especially because basic polypeptide precipitated almost quantitatively upon heating, whereas 7S globulin prevent their thermal aggregation through the formation of a soluble complex.

The effect of temperature, protein concentration, and proportion of glycinin to conglycinin on the mechanical parameters of soy protein gels were reported (Kang et al., 1991). The gels formed at higher heating temperature and protein concentration were firm, tough, and unfracturable. Elasticities were similar at all concentrations and lower when heated at higher temperature. Heating above 93C was necessary for formation of rigid gels. Glycinin to beta-conglycinin ratios affected the texture; low glycinin gels were more elastic than regular gels, thus indicating that beta-conflycinin largely contributes to elasticity.

Yang and Taranto (1982) evaluated the textural properties of mozarella cheese analogs manufactured from soybean protein. The results showed that physical and functional properties of the final product, which were prepared from soy protein, gelatin, and gum arabic, were found to be similar to those of natural mozarella cheese. Stretchability of the product is related to the amount of soy protein. Addition of salts, such as calcium chloride, was found to improve the stretchyability of the progel.

Foaming property is very important to improve texture, consistency, and appearance of food; such as baked and confectionery goods. Foam ability or foaming power corresponded to the ratio of gas volume to liquid volume in foams. There was significant correlation (Kitabatake and Doi, 1982) between the rate constant of protein solution surface tension decay and foam ability, but not with its absolute surface tension. In this study, soy protein has lower foaming power but higher stability than casein as could be seen from its smaller size and compact foams.

Emulsification. properties include fat adsorption, foaming, and whipping abilities. The oil-in-water emulsion stabilized by protein was represented by a bilayer model (Elizalde et al., 1988). protein diffused and reoriented at the oil-water interface, the hydrophilic groups oriented towards the water phase and the hydrophobic groups towards the oil. The thickness of the interfacial bilayer depended on the water and oil absorption capacity of the proteins and on the concentration of the protein. Protein-protein interactions were important to enhance cohesive forces between the proteins and hence the rigidity and mechanical strength of the bilayer film to prevent rupture and coalescence of oil globules.

Emulsion stability is enhanced by high protein and oil concentration (Elizalde et al., 1991), and these factors are highly interrelaated. Temperature variation, 37, 45, and 60C, had little effect on emulsion stability as measured by the energy activation. They also reported that emulsion stability depended primarily upon the WAC (water absorption activity) and OAC (oil absorption activity) of protein. Proteins with low WAC and WOAI (Water-oil absorption index) are not suitable for formulation of stable emulsions, even if using high concentrations or formulation with high oil concentration.

Emulsion properties of yellow pea flour with different particle size were studied (Horvath et al., 1989). They reported that flour with higher protein content had greater emulsifying capacity. Fine particles had high oil absorption capacity and good emulsifying capacity and stability, while coarse particles had high WA capacities. All samples had high NSI and were considered to have very promising functional properties.

Emulsion capacity measurement for nonmeat proteins was discussed (Amundson and Sebranek, 1990). Addition of NaCl or added protein to water, as opposed to oil, decreased the end point of soy protein isolate emulsion. On the other hand, a low NaCl levels increased the end point of Na-caseinate emulsion, and the order of wataer or oil addition had no effect.

Textural property of texturized plant protein is very important, as it needs to resemble meat texture. To produce a texturized product, plant protein must comply with certain minimum prerequisites. Kazemzadeh et al (Kazemzadeh et al., 1982), states as -a minimum protein content of 50% -a nitrogen solubility index of 50-70 -a maximum soluble carbohydrate content of 30%, and less than 1% fat is generally needed for textured products. Defatted soy meal can be converted into a fibrous, porous product with meat-like texture by use of high temperature, high pressure, and short time extrusion (Stanley and deMan, 1978). The possible mechanism is due to the dissociation of protein in high heat and pressure into subunits and followed by forming high molecular aggregates through intermolecular amide bonds.

Protein level and temperature had positive relationship with the shear force as texture indicator (Kazemzadeh et al., 1986). According to mechanical and microscopic evaluations, high protein content is essential for well-developed texture in soybase mixtures, with the assumption that proteins contribute to the skeletal structure in which carbohbydrates are dispersed. There was a pronounced increase in all of the measured parameters between 40 and 70% soy protein isolate content.

The possible role of crude fiber of soy flour in texture formation during nonextrusion texturization processing (Taranto et al., 1981) was studied. The results indicated that crude fiber may control the type of alveolation developed and the type of cuticle morphology exhibited in the products alveoli as a result of its plastic response to deformation. it seems probable that the presence of the soluble carbohydrate fraction is not a prerequisite for the formation of the alveolate morphology. The structural ingegrity of the final texturized product was lowered as the amount of crude fiber present in the soy flour mixture was increased from 5 to 12%.

Updated: Wednesday, June 20, 2007.