Increased utilization of legume protein by the food industry, especially soybean protein (Kinsella, 1979), has increased research in the utilization of legume or seed proteins in foods (Kim et al, 1990; Gebre-Egziabher and Sumner, 1983; Thompson, 1977; Rhee et al, 1972). Legume or seed proteins as an ingredient primarily to increase nutritional quality and provide a variety of functional properties, including desirable structure, texture, flavor, and color characteristics in formulated food products.

The knowledge of protein structure and size of different legumes or seeds, or different varieties, will bring an understanding of the protein properties. This will permit manipulation of these properties for food product development (Leterme et al, 1990; Kim et al, 1990; Wang and Damodaran, 1990; Cumming et al, 1973). Nutritional and functional qualities of protein are largely determined by its amino acid content and nitrogen solubility (Kinsella, 1979).

A new variety of yellow pea (Pisum sativum L. variety Miranda) has been successfully grown and yields a nutrient quality comparable to those of soybean when used in animal rations (England et al, 1986; Savage et al, 1986). In Oregon, approximately 15 million pounds are produced annually and sold for human consumption, at relatively low price (Carnes, 1988). If the protein in these yellow pea could be characterized and developed into a viable product, it would be of economic benefit to the farmer.

The objective of this study was to investigate the optimum yield and characteristics of the yellow pea protein extracted by acid and salt coagulations. Central composite rotatable design was used to predict the optimum extraction yields.

Yellow peas (Pisum sativum L. Variety Miranda) of sample grade (based on USDA Standards) were grown in 1989 and provided by International Seeds Inc., Halsey, OR. Thirty.five kilograms of yellow peas were ground into medium particles using Thomas Wiley Lab. Mill Model 4, in Crop Science Dept., OSU, OR.and stored at 3C.

Yellow pea flour was weighed into ziploc-plastic bag for 100 g each, then stored in tightly covered plastic bucket in cold room (30C) until needed.

Chemicals for protein extraction and nitrogen determination, including liquid nitrogen, CaCl2.2H2O(Chem. MFG. Corp., Gardena, CA), MgSO4.anhydrous (EM Industries Inc., Cherry Hill, NJ) and other general reagents were purchased from OSU Chemical Store (Corvallis, OR). Acrylamide, bis-acrylamide, AMPS, TEMED, Na-lauryl sulfate, bromchlorophenol blue, 2-mercaptoethanol, and trizma base were obtained from Sigma Chem. Co. (St. Louis, MO). Coomassie blue R-250 was bought from JT. Baker Inc. (Phillipsburg, NJ).

The slab gel electrophoresis unit, SE 250 Mighty Small II, was ordered from Hoefer Scientific Instruments (San Francisco, CA). Other equipments and apparatus were provided by Nutrition and Food Management Dept., OSU (Corvallis, OR). Black and white prints and color slides were done in Communication and Media Center Dept., OSU (Corvallis,OR).

Methods. Protein Extraction. Proteins of yellow peas were extracted according to the temperature, pH level, and coagulation treatments. Protein extraction procedure is reported in figure 1. Yields were calculated from total weight of freeze-dried products per 100 g yellow pea flour. Protein samples were powdered and kept in glass bottles and refrigerated (3C) until needed for protein characterizations.

For each extraction, 100 g of pea flour was blended for 10 sec at low speed with 600 mL redistilled water in Waring Commercial Blender, and continued liquefying for an additional 1 min. The slurry was transferred and brought to 1 L with redistilled water after pH adjustment.

The slurry was shaken in Gyrotory-shaker (New Brunswick Scientific Co., NJ), at speed 2.5 for 30 min with controlled temperature 250C or above, depending on the treatments assigned. Low temperature extractions (less than 250C) were done in low temperature incubator (Precision Scientific with Freas 815 temperature regulator) using a magnetic stirrer.

Samples were then centrifuged (IEC-International Centrifuge, International Equipment Co., Needham, Mass.) at a speed of 15 for 10 min at room temperature (25+20C). Collected protein was adjusted to pH 8 with 2N HCl or 2N NaOH then drop-frozen into liquid nitrogen and stored ("Revco" ultra low freezer, Revco Inc., West Columbia, S.C.) at -550C until it wasfreeze-vacuum dried (VirTis, Consol 4.5 Model; shelf temperature=400C, vacuum pressure=50 mT). Yields were calculated from freeze-dried products. Each yield was powdered in a Waring Commercial Blender for 10 sec. Protein samples were kept in glass bottles and refrigerated (30C) until needed for protein characterizations.

Proximate Analysis. Yellow pea flour was analyzed for proximate composition by Columbia Lab. Inc. (Corbett, OR), as is shown in Table 3-1. Calcium was analyzed by ashing two g of proteinate at 525oC for 24 hr, dissolved in 3 mL of 3N HCl solution and made to 25 mL volume with redistilled water. The solution was analyzed using a Perkin Elmer 2380 atomic absorption spectrometer.

Nitrogen Determination. This N-determination was done using micro Kjeldahl method (AOAC, 1985), on all 13 treatments for the AP, MAP, and CAP. The results are presented (Table 3-2, 3-3) as g N/100 g sample, with conversion factor to protein=5.7 (Columbia Lab. Inc., Corbett, OR.). Protein recovery was calculated based on total protein in the yield compared to that in 100 g yellow pea flour.

Amino acid composition. The composition was analyzed for AP, MAP, and CAP from extraction at pH 9 and 25C, at the Genetic Biochemistry Lab (Oregon State University, Corvallis, OR). Samples were hydrolyzed iin 6N HCl + !.0% phenol before injecting onto a High Performance Liquid Chromatograph. Data is reported in Table 3-4.

Electrophoresis. Electrophoresis of the proteinate used sodium dodecyl sulfate polyacrylamide gel (SDS-PAGE) according to the procedure of Laemmli (1970), using a protein concentration of 1.25 mg/mL, with 20 uL sample solution per well. Protein markers or molecular weight standard of MW-SDS-70L (Sigma Chem. Co., St.Louis, MO) were used with bromchlorophenol blue as the tracking dye.

SDS-PAGE for all samples was done with and without addition of 2-mercaptoethanol (2-ME) as a reducing agent, using 13% and 12% acrylamide gels, respectively. Electrophoreses were run at constant current, 40 ma or 35 ma per 2 gels for unreduced or reduced samples, respectively. Additionally, undenaturated (without addition of SDA or 2-ME) samples were also electrophoresed to reveal the MW pattern of protein polymers. Electrophoresis was performed at a constant current of 40 ma or 35 ma per 2 gels for unreduced or reduced samples, respectively. Staining was done in 0.125% coomassie brilliant blue R-250, 50% methanol, and 10% acetic acid, for 4 hrs. Destaining was done in 50% methanol, 10% acetic acid, for 2 hrs, then continued destaining in 5% methanol and 7% acetic acid, for 6 hrs. All assays were done at 25+2 0C. All assays were done at 25+or - 2C. Molecular weight of protein subunits were estimated from the plot of log MW vs. the ratio of the distance traveled in comparison with the tracking dye. Gels were stored in 7% glacial acetic acid solution until photographs were taken.

Experimental Design and Statistical Analysis. The experimental design used was a two-factor Central Composite Rotatable Design (CCRD) (Cochran and Cox, 1957), (Table 3-5), to predict maximum yields during protein extractions, and optimum characteristic measurements. Temperature in celcius (T) and pH (P) were independent (X) variables. The results obtained from extractions and protein characterizations were dependent (Y) variables. Statistical Analysis System (SAS Institute Inc., Cary, NC.) and Statgraph (Statistical Graphic Co., Rockville, Md.) programs were used to generate the ANOVA, parameter estimates, response surface analysis, canonical analysis, and ridge maximum of the responses. Differences were considered statistically significant at P <0.1. the following model of quadratic polynomial regression was assumed for evaluating the individual y-variables :

Quadratic models were used to plot 3-dimensional response surfaces. Response surface analysis facilitated a understanding of the nature of responses obtained, by graphically indicating maximums, minimums, or saddles. All replications and measurements were done in duplicate.

RESULTS AND DISCUSSION

Table 3-2, and 3-3 present the proximate composition of yellow pea flour and calcium content of proteinates, and protein recovery during extration, respectively. Table 3-4 presents parameter estimates for fitting the quadratic models for AP, MAP, and CAP on their extraction yields and nitrogen content.

There are significant differences (P<0.001) on extraction yields in both AP and CAP, but not in MAP. Nitrogen content in all type proteinate was not significantly different. Amino acid composition of three proteinate can be found in Table 3-5.

Protein extraction yields. The yield of acid-coagulated proteinate was significantly affected by changes in temperature and pH during extraction, as is demonstrated in Figure 3-2a. The RSA has hill shape with the maximum yield 20.2%, at treatment combination T=31.60C and P=9.8, while the yield of calcium-coagulated proteinate was affected (P <.001) only to pH change, and the RSA has a saddle shape ( Figure 3-2c) with maximum yield 19.0% at treatment combination T=27.0 0C and P=10.8. On the other hand, magnesium coagulation was not significantly affected by both factors. Its' RSA has a saddle shape ( Figure 2b), with a tendency to high yield with treatment combinations of high T and high P combination Its optimum treatment combination was T=25.1 0C and P=10.9 producing a yield 16.8%. Significant positive relationship between pH and protein recovery also reported by Shen et al (1991) on soymilk coagulated with glucono d-lactone or CaSO4.

Both salt-coagulated proteinates had 6-16% lower yield compared to acid-coagulated proteinates(Table 3-3). This might be because acid coagulation was done at pH 4, the lowest solubility of yellow pea protein, based on the preliminary work. Salt coagulations have a maximum yield at pH 6.5, since at this pH the carboxyl groups of aspartic and glutamic residues, and the imidazole groups of histidine residues are deprotonated (Kroll, 1984). Those groups bind with calcium or magnesium ions and cause the protein to coagulate. Brooks and Morr (1984) reported that precipitation of soy protein extract at pH 4 caused precipitation of 11S and 7S, while precipitation at pH 6.5 will only precipitate 11S . 7S remained soluble until precipitated at pH 4.

According to Lu et al (1980), the salt coagulated protein is actually the protein isolate which is precipitated at pH higher than its pI, which explains the higher yield of acid-coagulated compared to salt-coagulated proteinate. Both acid and salt coagulations have higher yield compared to that reported by Gebre-Egziabher and Sumner (1983), but lower than that reported by Sumner et al (1981). These might due to differences in variety and processing conditions.

Nitrogen content. There were maximum values of 71.2, 72.4, and 69.5% protein ((Table 3-3) for AP, MAP and CAP, respectively. None of the coagulation methods produced significant changes in nitrogen content, with the changes of temperature and pH during extraction. This likely indicates that the protein being coagulated has similar purity and extraction and coagulation conditions had achieved maximum levels for protein recovery.

As explained by Wang et al (1983) and Aguilera and Garcia (1989) that protein yield was significantly related to protein content in the extract solution. Nitrogen content is about 3.7 times of that in yellow pea flour. Lu et al (1980) also reported no difference in protein content for soy protein curds that were coagulated by acid, glucono lactone, or calcium salts.

Amino Acids. Amino acid compositions of three proteinate are presented in Table 3-5. The amino acid composition was similar for all three proteinates with a hydophilic:hydrophobic ratio of 60:40. All proteinates were low in cysteine and methionine, besides tryptophan as had been reported by other (Leterme et al., 1990). A similar distribution of amino acids also reported by Okezie and Bello (1988) on winged bean protein isolate extracted at pH 10 and 12.

Electrophoresis. Representations of SDS-PAGE photographs are displayed in Figure 3-3. There was no apparent differences in electrophoresis patterns related to molecular weight (MW) subunits. All proteinate with or without reducing agent, contained subunits with the highest MW=67-68 kD and the lowest MW=13-14 kD.

Electrophoretic patterns of undenatured samples (not shown in figure) revealed the differences between protein polymers of AP and MAP or CAP samples. AP contained polymers that were not clearly separated on the gel, but had bands of different polymers. While MAP and CAP did not show any band at aqll, due to polymer size too large to enter the stacking gel. This could be the result of Mg- or Ca-bridging on the protein polymer.

Addition of 2-ME (Figure 3-3) gave a different MW pattern compared to the one without 2-ME (Figure 3-3). Electrophoretic patters of samples without the reducing agent (Figure 3-3) showed the high MW subunits (>100 kD) which could not enter the stacking gel (4% acrylamide) which apparently dissociated into smaller subunits after 2-ME treatment, as it disappeared from wells on the stacking gel. the SDS-PAGE of unreduced samples resulted in the densest bands at subunits having MW=53 kD, which after treatment with reducing agent the denest bands were at subunits having MW-47 kD, and MW=24-26 kD.

Molecular weight patterns of undenatured proteinate samples indicated that forces, besides disulfide bonds, held the subunits together to form large protein polymers. These forces could be hydrophobic interactions, hydrogen bonds, electrostatic interactions, and/or salt bridges. The similarity in MW pattern of subunits with AP, MAP, and CAP samples, regardless of temperature and pH treatment combinations during extraction, indicated similarity in the type of protein subunits that had been coagulated. The difference in resolution of protein polymers from undenatured samples indicated there might some bridging effects of magnesium or calcium ion that were absent from acid coagulation.

Data indicate that further exploration of yellow pea extracted proteinate is warranted. The coagulant used did not significantly influence yield or nitrogen content. Though there were no differences in the type of protein extracted, salt-coagulated proteinate contained calcium or magnesium, at 3.04% (w/w). Adding value to the protein produced, particularly if dietary calcium and magnesium may have a protective effect against certain diseases (Berner et al, 1990; Anonymous, 1991a). Amino acid composition on all three proteinate showed similar pattern, which had ratio around 60:40 for hydrophilic to hydrophobic ratio merits further pursuit as a food source. The information obtained by this study will facilitate the background of thought for developing viable high-vegetable-protein foods with high calcium content.

REFERENCES

Aguilera, J.M. and Garcia, H.D. Protein extraction from lupin seeds: a mathematical model. Int. J. Food Sci. and Tech. 1989, 24, 17-27.

Anonymous. Calcium and magnesium needed for healthy blood pressure. Food Processing. 1991a. 52, 150.

AOAC. Official Methods of Analysis. Association of Official Analytical Chemist. 15th. Ed. herlich, K. 1990, vol. 1, No. 960, 52, p. 342. Arlington, VA.1984.

Berner, L.A., McBean, L.D. and Lofgren, P.A. Calcium and chronic disease prevention: Challanges to the food industry. Food Tech. 1990. March, 44, 50, 57 -60, 68, 70.

Brooks, J.R. and Morr, M.V. Phosphorus and phytate content of soybean protein components. J. Agric. Food Chem. 1984, 32, 672-674.

Carnes, J. Personal communication. International Seeds, Inc. Halsey, OR. 1988

Cochran, W.G. and Cox, G.M. Some methods for the study of response surfaces, Ch.8A. p.335. In "Experimental Design" 2nd ed. 5th printing. John Wiley and Sons, Inc. NY., 1957, p. 335.

Cumming, D.B., Stanley, D.W. and DeMan, J.M. 1973. Fate of water soluble soy protein during thermoplastic extrusion. J. Food Sci. 1973, 38, 320-323.

Davis, K.R. Effect of processing on composition and Tetrahymena relative nutritive value of green and yellow peas, lentils, and white pea beans. Cereal Chem. 1981, 58, 454-459.

England, D.C., Chitko, C., Dickson, R. and Cheeke, P.R. Utilization of yellow peas in swine grower-finisher diets. Oregon State Univ., Agric. Exper. Stat. Special Rep. 1986, (771), 1-7.

Gebre-Egziabher, A. and Sumner, A.K. Preparation of high protein curd from field peas. J. Food Sci. 1983, 48, 375-377,378.

Hsu, D.L., Leung, H.K., Morad, M.M., Finney, P.L. and Leung, C.T. Effect of germination on electrophoretic, functional and bread-baking properties of yellow pea, lentil, and faba bean protein isolates. Cereal Chem. 1982, 58, 344-350.

Kantha, S.S., Hettiarachchy, N.S., and Erdman Jr., J.W. Nutrient, antinutrient contents, and solubility profiles of nitrogen, phytic acid, and selected minerals in winged bean flour. Cereal Chem. 1986, 63, 9-13.

Kim, S.Y., Park, P.S.W. and Rhee, K.C. Functional properties of proteolytic enzyme modified soy protein isolate. J. Agric Food Chem. 1990, 38, 651-656.

Kinsella, J.E. 1979. Functional properties of soy proteins. J Am. Oil Chem. Soc. 1979, 56, 242-258.

Kroll, R.D. Effect of pH on the binding of calsium ions by soybean proteins. Cereal Chem. 1984<, 61(6), 490-495.

Laemmli, U.K. Cleavage of structural proteins during the assembly of the head of bacteriophage T-4. Nature. 1970, 227, 680-685.

Leterme, P., Monmart, T. and Baudart, E. Amino acid composition of pea (Pisum sativum) proteins and protein profile of pea flour. J. Sci. Food Agric. 1990, 53, 107-110.

Lu, J.Y., Carter, E. and Chung, R.A. Use of calcium salts for soybean curd preparation. J. Food Sci. 1980, 45, 32-34.

McCurdy, S.M. and Knipfel, J.E. Investigation of faba bean protein recovery and application to pilot scale processing. J. Food Sci. 1990 55, 1093-1094, 1101.

Naczk, M., Rubin, L.J. and Shahidi, F. Functional properties and phytate content of pea protein preparations. J. Food Sci. 1986, 51, 1245-1247.

Okezie, B.O. and Bello, A.B. Physicochemical and functional properties of winged bean flour and isolate compared to with soy isolate. J. Food Sci. 1988 , 53(2), 450-454.

Rhee, K.C., Cater, C.M. and Mathew, K.F. Simultaneous recovery of protein and oil from raw peanuts in an aquous system J. Food Sci. 1972, 37, 90-93.

Savage, T.F., Nakaue, H.S., Holmes, Z.A. and Taylor, T.M. 1986. Feeding value of yellow peas (Pisum sativum L. Variety Miranda) in market turkeys and sensory evaluation of carcasses. Poultry Sci. 1986, 65, 1383-1390.

Shen, C.F., De Man, L., Buzzel, R.I. and De Man, J.M. Yield and quality of tofu as affected by soybean and soymilk characteristics: Glucono-delta-lactone coagulant. J. Food Sci. 1991, 56, 109-112.

Sumner, A.K., Nielsen, M.A. and Youngs, C.G. Production and evaluation of pea protein isolate. J. Food Sci. 1981, 46, 364-366, 372. Thompson, L.U. Preparation and evaluation of mung bean protein isolates. J. Food Sci. 1977, 42, 202-206.

Thompson, L.U. Reduction of phytic acid concentration in protein isolates by acylation techniques. J. Am. Oil. Soc.. 1987, 64, 1712-1717.

Wang, C.H. and Damodaran, S. Thermal gelation of globular proteins: Weight-average molecular weight dependence of gel strength. J. Agric. Food Chem. 1990, 38(5), 1157-1164.

Wang, H.L. and Cavins, J.F. Yield and amino acid composition of fractions obtained during tofu production. Cereal Chem. 1989, 66, 359-361.

Wang, H.L., Swain, E.W. and Kwolek, W.F. Effect of soybean varieties on the yield and quality of tofu. Cereal Chem. 1983, 60, 245-248.