Fig. 6: The adapted B. animalis subsp. lactis W5S9 rebalances intestinal bile-acid metabolism in NAFLD mice.

a Microbial α-diversity at week 8 (species richness and Shannon index at the species level, WMS data, n = 6 mice per group). Significance is determined by a two-tailed Mann–Whitney test. b PCoA based on Bray-Curtis dissimilarity at the species level (WMS data from n = 6 mice per group). c Pairwise Bray–Curtis dissimilarity between the Control and other groups. Box plots show the median (centre line), 25th and 75th percentiles (box limits) and 1.5× interquartile range (whiskers) for n = 6 mice per group; individual data points are overlaid. Statistical comparisons were evaluated using two-tailed Mann–Whitney tests. d–g OPLS-DA based on the targeted bile-acid metabolic profiles (n = 6 mice per group; OPLS-DA models were fitted using the R package ropls). Score plots (left) and heat maps (right) depict differential metabolites for HFAD vs. Control (d), HFAD-H22B656 (e), HFAD-W5S9 (f), and (g) HFAD-H22B656 vs. HFAD-W5S9. Differential bile acids were defined as those with variable importance in projection (VIP) > 1 and |log₂FC | > 1 for the indicated pairwise comparisons. Top 10 bile acids ranked by VIP score. h Venn diagram of significantly altered bile acids (VIP > 1, |log₂FC | > 1) shared among HFAD vs. Control, HFAD vs. HFAD-H22B656, and HFAD vs. HFAD-W5S9 comparisons. i KEGG pathway enrichment of differential metabolites. Bubble size indicates the richness factor. j Schematic of primary bile-acid biosynthesis (ko00120). Hepatic TCDCA, GCDCA, TCA, and GCA are deconjugated in the gut by BSH (cbh) to regenerate free CDCA and CA (Created in BioRender. Zhang, J. (2026) https://BioRender.com/3qyx3m3). k Concentration of the six key bile acids highlighted in (j). Log₂FC versus HFAD is annotated above each panel. Data are presented as mean ± SEM (n = 6). All shown metabolites met the significance criteria of VIP > 1 and |log₂FC | > 1 compared to the HFAD group. Source data are provided as a Source Data file.