Discovery and taste mechanisms of novel umami peptides from edible grasshopper (Oxya chinensis sinuosa) based on machine learning, molecular docking and dynamics simulation.

This study aimed to reveal umami peptides from grasshopper, and then explored their umami perception mechanism and umami enhancement. Four potential umami peptides (GDRFL, NGWGDY, NVTY and VVNY) were isolated, characterized and screened from grasshopper protein hydrolysate through machine learning and in silico approaches. Subsequently, sensory evaluation and electronic tongue confirmed their umami taste with thresholds of 0.08-0.25 mg/mL, which were lower than that of monosodium glutamate (0.3 mg/mL). The potent umami taste of GDRFL and NGWGDY was mainly driven by their umami amino acid compositions, whereas NVTY and VVNY mainly be attributed to their bitter/hydrophobic amino acid residues (Tyr) at C-terminal. Molecular docking and molecular dynamics simulation further revealed that stable complexes formed between four peptides and umami receptor T1R1/T1R3 via hydrogen bond and hydrophobic interaction, which might be their underlying basis of umami perception. Furthermore, all of them present umami enhancement on MSG through additive effect (R = 0.68-0.86).