Revealing metabolite profiles in soy sauce and exploring their correlation with umami taste using UPLC-Orbitrap-MS/MS and GC-Tof-MS derivatization.

Soy sauce has a rich base of non-volatile substances, but existing studies are insufficient. This study analyzed the metabolites of 19 Chinese commercial soy sauces by UPLC-Orbitrap-MS/MS and GC-Tof-MS derivatization, and detected 674 and 230 kinds of substances, respectively, that could be grouped into 12 different classes of compounds, such as peptides, amino acid derivatives, organic acids, sugars, sugar alcohols, amino acids and so on. For the first time, 215 dipeptides and 91 amino acid derivatives in soy sauce were analyzed in detail and systematically from the perspective of composition and amino acid structure. The flavor profile of soy sauce was obtained by electronic tongue analysis, and orthogonal projections to latent structures (OPLS), random forest (RF), correlation were used to screen potential compounds associated with umami. The intersection of the three methods yielded 9 substances, including 4 reported umami-taste compounds, i.e., Glu, Fru-Glu, Inosine 5'prime-monophosphate (IMP) and Arg-Ser, as well as 5 others that may potentially contribute to umami or be associated with umami-taste producing microorganisms, including His-Asn and Homoserine lactone. This study will advance the understanding of soy sauce metabolites, and provide an in-depth reference for dipeptides and amino acid derivatives in soy sauce.