Revealing the Characteristics and Correlations Among Microbial Communities, Functional Genes, and Vital Metabolites Through Metagenomics in Henan Mung Bean Sour

Abstract

Henan mung bean sour (HMBS) is the raw material for mung bean sour noodles (MBSNs), a traditional fermented food. To investigate the characteristic flavor compounds, we have detected the content of free amino acids (FAAs) and key metabolites including organic acids, sugars, and alcohols. The results revealed that the content associated with umami, sweetness, and bitterness (TVA > 1) showed significant differences. Metagenomic analysis indicated that Lactobacillus delbrueckii was the dominant and characteristic species in WJ and LY15, whereas Bifidobacterium mongoliense, Lactiplantibacillus plantarum, and Acetobacter indonesiensis were the dominant species in GY. The abundance of functional genes related to carbohydrate and amino acid metabolism was higher in WJ and LY15. There was a strong correlation between dominant genera and vital metabolites (r |>| 0.7). This study provides a theoretical foundation for the development of HMBS.

Keywords: Henan mung bean sour; dominant genera; functional genes; metabolites; metagenomics.

Figure 1

(A) Radar chart of sensory evaluation based on K-means clustering; (B) comparison of pH and total sensory scores of different samples. TVA, taste activity value. VIP, variable importance in the projection. Significant differences are denoted by different lowercase letters and uppercase letters (p < 0.05).

Figure 2

(A) Free amino acids profiles were subjected to partial least squares analysis (PLS-DA) R2X = 0.993, R2Y = 0.999, Q2 = 0.993. (B) Changes in critical metabolites in a waterfall plot in different samples. The content changes of vital metabolites in different samples (C,D). Data with different lowercases in the same index were significantly different.

Figure 3

Differences in microbial community composition across samples from different regions. Variations in alpha diversity indices among samples are shown, including the Sobs index (A), Shannon index (B), and Simpson index (C). Venn diagrams depict phylum level (D), genus level (E), and species level (F), respectively. Relative abundance of microbial community at the genus (G) and species (H) level.

Figure 4

Beta diversity analysis and identification of different flora. PCoA score plots from all samples at the phylum level (A), genus level (B), and species level (C). Cladogram of LEfSe analysis based on shotgun metagenomics at different taxonomic levels (D). LEfSe differential analysis (LDA > 4) (E).

Figure 5

Functional gene annotation plots. Circos sample-function relationship plots are commonly used to show the distribution of functions present in different microbial samples. The bar graphs show the differences in the mean relative abundance of the same function between different groups and are annotated as to whether the differences are significant (p-value values and asterisks indicate significant differences). (A) COG annotated Circos plot. (B) COG annotated bar graph plot. (C) KEGG annotated Circos plot. (D) KEGG annotated bar graph plot. (E) CAZy annotated Circos plot. (F) CAZy annotated bar graph. * 0.01 < p ≤ 0.05.

Figure 6

Differences in carbohydrate metabolism pathways across different samples (A) and differences in amino acid metabolism pathways (B). (C,D) represent the correlation between dominant bacterial genera and 17 free amino acids, as well as between even key metabolites, respectively. * means significant differences among samples (p < 0.05).