Answers are grounded in this paper and cite their sources.
Fermentation is driven by dynamic interactions between substrates and microbial communities. However, the complexity of natural consortia limits their predictive control. To address this problem, we developed a gnotobiotic kimchi model via inoculation with defined lactic acid bacterial (LAB) consortia and individual strains under controlled temperatures. Consortium fermentations exhibited temperature-dependent growth and acidification, with stable proliferation under 6 °C and 10 °C, whereas 15 °C induced a rapid shift in community structure. Phase normalisation revealed convergence of microbial and metabolite profiles along a shared successional pattern. Machine learning identified a nine-metabolite signature (lactate, sucrose, fructose, glycine, glucose, succinate, threonine, choline, and glutamate) that accurately classified the fermentation phases with robust performance across internal and external datasets. Network analyses highlighted Leuconostoc mesenteroides and Lactococcus lactis as keystone metabolic species, whereas mono-association fermentation uncovered strain-specific metabolic endpoints and strategies. These findings established a predictive framework for LAB-driven kimchi fermentation. • Gnotobiotic kimchi models reveal temperature-dependent LAB growth and acidification. • Phase-resolved analysis uncovered robust LAB-driven fermentation dynamics. • Machine learning identified nine key metabolites for accurate phase classification. • Leuconostoc mesenteroides and Lactococcus lactis as metabolic keystone species. • Mono-association fermentations defined strain-specific metabolic endpoints and strategies.
No takes yet. Share an insight, caveat, or question.
Kim et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: