Microbiome analysis and random forest algorithm-aided identification of the diacetyl-producing microorganisms in the stacking fermentation stage of Maotai-flavor liquor production

ABSTRACT Diacetyl imparts a characteristic buttery aroma to foods and beverages. However, when present at high levels in Maotai-flavor liquor, it generates an undesirable buttery off-odor. The aim of this study was to identify the diacetyl-producing microorganisms in Maotai-flavor liquor with an ultimate aim to develop methods that could reduce diacetyl production. To explore the microbial origin of diacetyl in Maotai-flavor liquor, the off-odor intensity and diacetyl content of stacking fermented grains, which are used in the production of Maotai-flavor liquor, were evaluated. Then, 16S rRNA V3-V4 high-throughput sequencing and the random forest algorithm were combined to analyze the potential diacetyl producers. The featured selection results indicated that Lactobacillus was the major diacetyl contributor in the fermented grains, followed by Staphylococcus, Weissella, Pediococcus, and Klebsiella. To verify this, 123 lactic acid bacteria (LAB) isolates were selected from the high diacetyl off-odor group samples, and the genus Lactobacillus accounted for more than 90% of the identified isolates. The results of our study showed that L. plantarum, L. pentosus, and L. fermentum were the major diacetyl-producing microorganisms (>90 mg/L). Moreover, the fermentation characteristics of the high-yield strain, L. plantarum MTL-09, showed that temperature and pH had strong effects on diacetyl production, which may offer a strategy for inhibiting the off-odor by controlling the pH or temperature during fermentation. The methods may also be useful for identification of key microorganisms for fermented foods and alcoholic beverages.