Analysis of Microbial Community Structure and Functional Genes for Volatile Flavor in Stinky Tofu

The distinctive flavor of stinky tofu arises from intricate microbial metabolic networks during traditional fermentation, yet the genetic mechanisms linking microbial community structure to flavor formation remain incompletely resolved. This study employed metagenomic sequencing (Illumina NovaSeq 6000, Q30 > 92%) to generate 7.32 Gb of high-quality data, integrated with functional annotations from KEGG, eggNOG, and CAZy databases, to systematically dissect core microbial taxa and metabolic genes driving flavor biosynthesis. Dominant genera included Pseudomonas (relative abundance: 74.3%), Acinetobacter (14.4%), and Enterobacter (5%), with Pseudomonas putida (12.5%) and Pseudomonas fluorescens (3.2%) orchestrating carbohydrate metabolism (68.22% KEGG pathways) and amino acid degradation via glycoside hydrolases (GHs, 73% of CAZy-annotated enzymes) and dehydrogenases (e.g., 125 lactate dehydrogenase genes). Key flavor compounds, such as diacetyl (379 α-acetolactate synthase genes) and 3-methylbutanoic acid, were synthesized through synergistic pathways. Additionally, Lactococcus and Kluyvera contributed to ester and short-chain fatty acid production via α-keto acid dehydrogenase complexes (55 genes). A total of 410,231 non-redundant genes were identified, annotated to 4690 microbial species, establishing a multi-layered microbial-gene-metabolite regulatory network. This work elucidates the molecular basis of stinky tofu flavor formation and provides a framework for optimizing traditional fermentation processes through targeted microbial engineering.