Pathway and Production Differences in Branched-Chain Hydroxy Acids as Bioactive Metabolites in Limosilactobacillus fermentum, Ligilactobacillus salivarius, and Latilactobacillus sakei

Abstract

Branched-chain hydroxy acids (BCHAs) as bioactive metabolites of Lactobacillaceae include 2-hydroxy isovaleric acid (HIVA), 2-hydroxy isocaproic acid (HICA), and 2-hydroxy-3-methyl isovaleric acid (HMVA). Combining targeted and untargeted metabolomics, this study elucidates differences in extracellular BCHA production in Limosilactobacillus fermentum, Ligilactobacillus salivarius, and Latilactobacillus sakei alongside comparing comprehensive metabolic changes. Through targeted metabolomics, BCHA production among 38 strains exhibited strain specificity, except for L. sakei, which showed significantly lower BCHA production. Explaining the lower production in L. sakei, which lacks the branched-chain amino acid (BCAA)-utilizing pathway, comparison of BCHA production by precursor reaction revealed that the pathway utilizing BCAAs is more dominant than the pathway utilizing pyruvate. Expanding upon the targeted approach, untargeted metabolomics revealed the effects of the reaction compound on other metabolic pathways besides BCHAs. Metabolism alterations induced by BCAA reactions varied among species. Significant differences were observed in glycine, serine, and threonine metabolism, pyruvate metabolism, butanoate metabolism, and galactose metabolism (p < 0.05). These results emphasize the importance of the synergy between fermentation strains and substrates in influencing nutritional components of fermented foods. By uncovering novel aspects of BCAA metabolism pathways, this study could inform the selection of fermentation strains and support the targeted production of BCHAs.

Keywords: 2-hydroxy isocaproic acid (HICA); 2-hydroxy isovaleric acid (HIVA); 2-hydroxy-3-methyl valeric acid (HMVA); Lactobacillaceae; bioactive metabolites; metabolomics; probiotics.

Figure 1

Branched-chain hydroxy acid (BCHA) production in MRS cultivation. Different letters in the bar graph indicate a significant difference determined by ANOVA followed by Duncan’s multiple-range test (p < 0.05).

Figure 2

Branched-chain hydroxy acid (BCHA) production in the reaction mixture experiments. (A) Limosilactobacillus fermentum EFEL 6804, (B) Ligilactobacillus salivarius KGMB 2057, and (C) Latilactobacillus sakei KACC 12414. Different letters in the bar graph indicate a significant difference determined by ANOVA followed by Duncan’s multiple-range test (p < 0.05).

Figure 3

Pathway difference in the branched-chain hydroxy acids (BCHAs) in Limosilactobacillus fermentum, Ligilactobacillus salivarius, and Latilactobacillus sakei.

Figure 4

Untargeted metabolomics heatmap of the reaction mixture experiment. The relative metabolite levels were normalized to non-reaction control groups specific to each strain. All metabolites were significantly distinguished in each group by one-way analysis of variance (p < 0.05).

Figure 5

(A) Principal component analysis of the reaction mixture experiment analyzed by UHPLC–Orbitrap–MS/MS in positive ion mode. (B) Significantly altered metabolism pathways under BCHA precursor reactions. LFE 6804, Limosilactobacillus fermentum EFEL 6804; LSL 2057, Ligilactobacillus salivarius KGMB 2057; LSK 12414, Latilactobacillus sakei KACC 12414.