Analysis of protein in soy flour and in tapioca based snack chips

Two experiments were conducted: El to evaluate a rapid analysis method for the proteins in PDI 20 and PDI 85 defatted soy flours (SF) and E2 to investigate the interactions of soy proteins during production of tapioca based snack chips. In El total proteins were extracted from each SF by a 0.03M Tris buffer (pH 8.0) with and without lOMm 2-mercaptoethanol (ME) and were separated into 2S, 7S and IIS fractions. Total protein samples and each fraction were analyzed by high performance liquid chromatography (HPLC) on a TSK 3000-SW size exclusion column. E2 consisted of three parts. In the first part, soy proteins from each production stage: flour, batter, gel and cooked (baked), of tapioca based snack chips containing each SF were analyzed by HPLC. A second part consisted of measurement of percent protein extractable from each stage by buffer alone or with 0.5M NaCl, 6M urea, 0.2M ME, or 20% propylene glycol (PG) to determine specific protein interaction forces. In the last part, the percent moisture and protein of each chip production stage containing each SF and of baked and fried chips were measured, and the hardness and expansion during cooking of chips were determined. Two replications were completed for El and for each part of E2.

A total of 11 peaks were found in the chromatograms of the total soy proteins and among the chromatograms of the fractions of each SF in El. Time for HPLC analysis was less than 45 min. Tentative identity (ID) of the peaks in the 2S chromatogram included a small amount of beta-conglycinin, lipoxygenase and monomers and aggregates of the two trypsin inhibitors (TI). One 2S peak eluting at 23.9 min was positively identified as an aggregate of the Kunitz TI. Major peaks in the 7S chromatogram were identified as an aggregate of beta-conglycinin with a molecular weight (MW) of over 500,000 Daltons and glycinin. The major peak in the 11S chromatogram was glycinin with a MW of 325,000 Daltons. The chromatogram of PDI 85 SF had better peak resolution and sharper peaks than the chromatogram of PDI 20 SF.

Heating to produce the gel stage of the chip resulted in lower (P<0.05) levels of extractable protein and a decrease ((P<0.05) in the combined peak areas of the beta-conglycinin protomer and aggregate and the glycinin in the chromatogram. Baking did not further reduce protein solubility. Data from the protein interaction study showed that disulfide bonds, hydrogen bonds and hydrophobic interactions were involved in the heat denaturation and reduced solubility of the proteins from the SF of different PDIs and at the gel and cooked stages of the chips. Baked and fried chips from either SF contained, respectively, 17.5% and 12.2% protein and 5.0 and 3.8% moisture. Baked chips containing PDI 20 SF were the hardest and had the least expansion during cooking while fried chips from PDI 85 SF were the easiest fractured with next to the least expansion during cooking.