| The bitterness in soy protein hydrolysates is a major obstacle to acceptance of soy products by consumers. The primary objective of this research was to evaluate the bitterness of 4 different molecular weight peptide fractions obtained from Protex 7L-treated soy hydrolysate. First, a theoretical analysis was performed to predict hydrolytic cleavage points of 3 different bitter- and non-bitter-producing proteases on soy protein at 4% degree of hydrolysis (DH) and hypothesize how peptide size influences bitterness. Protex 7L-treated soy hydrolysate was fractionated by gel filtration, desalinated by ultra-filtration, freeze-dried, and re-diluted to 5% w/v in Milli-Q water for sensory evaluation. Molecular weight of the fractions was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Panelists' sensitivity to Multifect Neutral (MN)-treated soy hydrolysate was compared to caffeine, quinine, leucine, and phenylalanine. Panelists' perceptions of the bitterness of MN-treated soy hydrolysate were compared to the bitterness of leucine and phenylalanine free amino acids. Panelists were screened for bitterness sensitivity and 15 highly sensitive panelists were selected to evaluate the Protex 7L-treated soy hydrolysate fractions, MN-treated soy hydrolysate, caffeine, quinine, leucine, and phenylalanine by using a modified triangle test. Comparison of panelist sensitivity was evaluated by Cohen's kappa coefficient and Fisher's exact test (p ≤ 0.05). Comparison of bitterness was analyzed by Cohen's kappa coefficient and McNemar's test (p ≤ 0.05).;A fraction consisting of low molecular weight peptides (estimated 1-5 kDa) was identified as bitter (p = 0.009) as well as the unfractionated hydrolysate (p = 0.088). The remaining 3 fractions (2 larger MW and one < 1 kDa) were not bitter. Panelists selected for bitterness sensitivity had an average threshold of 0.98 mM for caffeine, 8.9 muM for quinine, 5.3 mM for leucine, 3.4 mM for phenylalanine, and 5.2 g/100 mL for MN-treated soy hydrolysate. The kappa coefficient showed poor agreement between panelist sensitivity to soy hydrolysate in relation to caffeine, quinine, and phenylalanine, and fair agreement in relation to leucine. Fisher's exact test showed a non-significant p value for panelist sensitivity to soy hydrolysate in relation to caffeine, quinine, leucine, and phenylalanine, indicating that panelist sensitivity to soy hydrolysate was independent of panelist sensitivity to caffeine, quinine, leucine, and phenylalanine. While all panelists identified caffeine and quinine as bitter, 41%, 62%, and 86% identified leucine, phenylalanine, and soy hydrolysate as bitter, respectively. In both bitter perception relationships to MN-treated soy hydrolysate, leucine and phenylalanine showed poor agreement by the kappa coefficient and significance by McNemar's test, indicating that the bitterness in leucine and phenylalanine is different than the bitterness in soy hydrolysate. This suggests that free hydrophobic amino acids such as leucine and phenylalanine are not responsible for the bitterness of soy protein hydrolysate. In bitterness sensory studies of soy hydrolysate, neither leucine nor phenylalanine is recommended for panelist training. Although caffeine and quinine are recognized standards for bitterness, small peptides may be a better standard to use in bitterness training for soy hydrolysates.;Several factors are likely to be the cause for bitterness in protein hydrolysates. The hydrophobicity, primary sequence, spatial structure, molecular weight, and bulkiness of peptides tend to be inter-related and a combination of these factors is most likely responsible for bitterness. However, these bitterness models do not explain why certain proteases do not produce bitter hydrolysates. Further research on the bitterness of protein hydrolysates coupled with valid sensory analysis is still needed. |
