Identification of Novel Umami Peptides from Yeast Protein through Enzymatic, Sensory, and In Silico Approaches

This study aimed to rapidly develop novel umami peptides using yeast protein as an alternative protein source. Yeast protein hydrolysates exhibiting pronounced umami intensity were produced using flavorzyme under optimum conditions determined via a sensory-guided response surface methodology. Six out of 2138 peptides predicted to possess umami taste by composite machine learning and assessed as nontoxic, nonallergenic, water-soluble, and stable using integrated bioinformatics were screened as potential umami peptides. Sensory evaluation results revealed these peptides exhibited multiple taste attributes (detection threshold: 0.37 ± 0.10–1.1 ± 0.30 mmol/L), including umami. In light of the molecular docking outcomes, it is inferred that hydrogen bond, hydrophobic, and electrostatic interactions enhanced the theoretically stable binding of peptides to T1R1/T1R3, with their contributions gradually diminishing. Hydrophilic amino acids within T1R1/T1R3, especially Ser, may play a particularly pivotal role in binding with umami peptides. Future research will involve establishing heterologous cell models expressing T1R1 and T1R3 to delve into the cellular physiology of umami peptides. Peptide sequences (FADL, LPDP, and LDIGGDF) also had synergistic saltiness-enhancing effects; to overcome the limitation of not investigating the saltiness enhancement mechanism, comprehensive experiments at the molecular and cellular levels will also be conducted. This study offers a rapid umami peptide development framework and lays the groundwork for exploring yeast protein taste compounds.