Balance between bile acid conjugation and hydrolysis activity can alter outcomes of gut inflammation

Abstract

Conjugated bile acids (BAs) are multi-functional detergents in the gastrointestinal (GI) tract produced by the liver enzyme bile acid-CoA:amino acid N-acyltransferase (BAAT) and by the microbiome from the acyltransferase activity of bile salt hydrolase (BSH). Humans with inflammatory bowel disease (IBD) have an enrichment in both host and microbially conjugated BAs (MCBAs), but their impacts on GI inflammation are not well understood. We investigated the role of host-conjugated BAs in a mouse model of colitis using a BAAT knockout background. Baat^-/- KO mice have severe phenotypes in the colitis model that were rescued by supplementation with taurocholate (TCA). Gene expression and histology showed that this rescue was due to an improved epithelial barrier integrity and goblet cell function. However, metabolomics also showed that TCA supplementation resulted in extensive metabolism to secondary BAs. We therefore investigated the BSH activity of diverse gut bacteria on a panel of conjugated BAs and found broad hydrolytic capacity depending on the bacterium and the amino acid conjugate. The complexity of this microbial BA hydrolysis led to the exploration of bsh genes in metagenomic data from human IBD patients. Certain bsh sequences were enriched in people with Crohn's disease particularly that from Ruminococcus gnavus. This study shows that both host and microbially conjugated BAs may provide benefits to those with IBD, but this is dictated by a delicate balance between BA conjugation/deconjugation based on the bsh genes present.

Fig. 1. Under DSS induced colitis wild-type and ko mice comparison.

a Experimental animal model in the Baat^−/− KO mice versus wild-type mice under DSS treatment for 7 days (n = 4–5). b Weight changes during the experiment. c Daily DAI scores. d Example images of the colons from each group of mice. e The colon lengths of each animal. f Representative photomicrographs of H&E-stained (original magnification, ×100, scale bar = 100 μm). g Histology scores with representative H&E staining. Data in (b, c) expressed as the mean ± SEM; n = 4–5 animals per group; all boxplots are the interquartile range, center line is the median, with the whiskers denoting minima and maxima; statistics analyzed by one-way ANOVA with Tukey’s post hoc, and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 2. Baat^−/− KO mice supplemented with 0.3% taurine-conjugated cholic acid under DSS-induced colitis.

a Experimental animal design using Baat^−/− KO mice fed with 0.3% TCA under DSS-induced colitis (n = 6-8). (b) Weight changes during the experiment (Data expressed as the mean ± SEM). c Example images of the colons from each group of mice. d The colon lengths of each animal. e Spleen index (spleen weight: body weight ratio). f DAI scores on day 13. g Representative photomicrographs of alcian blue-periodic acid-Schiff (AB-PAS)-stained colon sections (original magnification, ×100, scale bar = 100 μm). h Goblet cell counts from each group of mice. i Representative photomicrographs of MUC-2 staining colon sections (scale bar = 50 μm). j Quantification of MUC2 staining. (k–n) mRNA expression of MUC-2, ZO-1, FXR and TGR5 in colon. All boxplots are the interquartile range, center line is the median, with the whiskers denoting minima and maxima; n = 6–8 animals per group. Weight changes, colon lengths, spleen index, DAI scores, AB-PAS staining, goblet cell counts, and mRNA expression data were tested by one-way ANOVA with Tukey’s post hoc. MUC-2 staining data analyzed by Kruskal-Wallis test followed by Dunn’s test. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 3. DSS-induced ulcerative colitis in Baat^−/− KO mice supplemented with 0.3% taurine-conjugated cholic acid modifies microbiome, metabolome and bile acid profiles.

a Microbiome alpha diversity determined by the Shannon index; n = 4-5 animals per group (b) The principal coordinate analysis (PCoA) based on Bray-Curtis dissimilarity shows the microbiome β-diversity; n = 4–5 animals per group. c Relative abundance of Lachnospiraceae. d–g The principal coordinate analysis (PCoA) based on Bray-Curtis dissimilarity of liver, serum, cecum, and feces. h Average relative abundance of TCA, GCA, CA, DCA, oxoCA, CA-MCY, CA-MCYO, CA-MCYO₂, and MCBAs. Heatmap values (Z-score by rows) represent mean abundance of each BA in each group. Grey color indicates values that were not detected. The number of mice in the microbiome and metabolome results was n = 4–5 mice per group. All boxplots are the interquartile range, center line is the median, with the whiskers denoting minima and maxima. Microbiome data analyzed by one-way ANOVA with Tukey’s post hoc, and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Metabolites data analyzed by Kruskal-Wallis followed by Dunn’s post hoc, # indicated the significant differences between HC and TCA groups (p < 0.05), × indicated the significant differences between HC and DSS groups (p < 0.05), and * indicated the significant differences between DSS and DSS-TCA groups (p < 0.05).

Fig. 4. Conjugated cholic acids hydrolysis across bacterial strains and wild-type mice supplemented 50 mg/kg TCA or GluCA via peanut butter pellets self-administrated method under DSS-induced colitis.

a The abundances of conjugated cholic acids are represented as colors corresponding to log10­transformed peak areas (area under the curve, AUC). Values represent the mean from n = 3 replicates of log10 areas changes compare to medium after 48 h in vitro culture. Phylogenetic tree of 17 Bacteroidetes strains using GTDB-tk based on whole bacteria genome. b Experimental animal design in wild-type C57BL/6 J mice fed with a final dose of 50 mg/kg TCA or GluCA peanut butter pellets under DSS-induced colitis (n = 10 each group). c–e Relative abundance of CA, DCA and oxocholic acid abundance in feces. Fecal metabolites data was analyzed by Kruskal-Wallis followed by Dunn’s post hoc, and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. All boxplots are the interquartile range, center line is the median, with the whiskers denoting minima and maxima.

Fig. 5. Characterizations of BSHs in IBD cohorts.

a, b The alpha diversity of BSHs in IBD patients and healthy controls, which was measured using the Shannon index and Observed BSHs. c The β-diversity of BSHs was measured by principal coordinate analysis (PCoA) based on Bray-Curtis dissimilarity. d The cumulative relative abundance of BSHs in IBD patients and healthy controls. e–g Taxonomic characterization of BSHs at phylum, family, and genus levels. h–k Examples of significant differing BSHs in IBD compared to healthy controls. Data was analyzed by two-sided Wilcox rank-sum test followed by BH correction, and *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, sample sizes available in methods. All boxplots are the interquartile range, center line is the median, with the whiskers denoting minima and maxima.