6α-hydroxylated bile acids mediate TGR5 signalling to improve glucose metabolism upon dietary fiber supplementation in mice

Abstract

Objective: Dietary fibres are essential for maintaining microbial diversity and the gut microbiota can modulate host physiology by metabolising the fibres. Here, we investigated whether the soluble dietary fibre oligofructose improves host metabolism by modulating bacterial transformation of secondary bile acids in mice fed western-style diet.

Design: To assess the impact of dietary fibre supplementation on bile acid transformation by gut bacteria, we fed conventional wild-type and TGR5 knockout mice western-style diet enriched or not with cellulose or oligofructose. In addition, we used germ-free mice and in vitro cultures to evaluate the activity of bacteria to transform bile acids in the caecal content of mice fed with western-style diet enriched with oligofructose. Finally, we treated wild-type and TGR5 knockout mice orally with hyodeoxycholic acid to assess its antidiabetic effects.

Results: We show that oligofructose sustains the production of 6α-hydroxylated bile acids from primary bile acids by gut bacteria when fed western-style diet. Mechanistically, we demonstrated that the effects of oligofructose on 6α-hydroxylated bile acids were microbiota dependent and specifically required functional TGR5 signalling to reduce body weight gain and improve glucose metabolism. Furthermore, we show that the 6α-hydroxylated bile acid hyodeoxycholic acid stimulates TGR5 signalling, in vitro and in vivo, and increases GLP-1R activity to improve host glucose metabolism.

Conclusion: Modulation of the gut microbiota with oligofructose enriches bacteria involved in 6α-hydroxylated bile acid production and leads to TGR5-GLP1R axis activation to improve body weight and metabolism under western-style diet feeding in mice.

Keywords: bile acid metabolism; diabetes mellitus; dietary fibre; glucagen-like peptides.

Figure 1

Oligofructose (OFS) supplementation increases caecal 6α-hydroxylated bile acid levels. (A) Bile acid profile in the caecum. (B) Total and secondary bile acid levels in the caecum. (C) Caecal levels of tauro-beta muricholic acid (TβMCA), beta muricholic acid (βMCA), omega muricholic acid (ωMCA), hyocholic acid (HCA) and hyodeoxycholic acid (HDCA) as well as the ratio of conjugated and unconjugated ωMCA, HDCA and HCA over conjugated and unconjugated βMCA levels. (D) Oral glucose tolerance test (OGTT) and its corresponding area under the curve (AUC). (E) Insulin levels, at fasting and 15 min during OGTT. (F) Colonic Gcg gene expression and L-cell density. Data are presented as mean±SEM of n= 9−10 per group. Kruskal-Wallis test with multiple comparison test using the original false discovery rate method of Benjamini and Hocheberg was performed for panels B and C. One-way analysis of variance was performed followed with post-hoc Tukey’s test for panels D to F. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001 (see also online supplemental figures 1 and 2). DCA, deoxycholic acid; LCA, lithocholic acid; WSD, western style diet.

Figure 2

Altered gut microbiota profile by oligofructose (OFS) supplementation correlates significantly with 6α-hydroxylated bile acid levels in the caecum. (A) Relative abundance of bacterial phyla. (B) Relative abundance of selected bacterial families and genera with a false discovery rate of p<0.05. (C) Spearman correlations between the relative abundance of bacterial genera and bile acid profile in the caecum as well as with body weight gain and white adipose tissue weights. Data are presented as mean±SEM. n=9–10 per group. Kruskal-Wallis test was performed followed by a multiple comparison test using the original false discovery rate (FDR) method of Benjamini and Hocheberg. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001. For panel C: a: p<0.05, b: p<0.01, c: p<0.001 and d: p<0.0001 (see also online supplemental figures 3 and 4). HDCA, hyodeoxycholic acid; WSD, western style diet; βMCA, beta muricholic acid; ωMCA, omega muricholic acid; TβMCA, tauro-beta muricholic acid,

Figure 3

Oligofructose (OFS) supplementation increased the capacity of gut bacteria to produce 6α-hydroxylated bile acids in vivo. (A and B) Body weight gain (BWG) and oral glucose tolerance test (OGTT), respectively of wild-type (WT) germ-free mice colonised with the caecal content of western style diet (ConvD WSD) or western style diet enriched with oligofructose fed mice (ConvD WSD-OFS). (C) Fasting glucose levels before OGTT. (D) Active GLP-1 levels in the portal vein. (E) Caecal bile acid levels. Data are presented as mean±SEM. Two-way analysis of variance with Bonferroni’s multiple comparison test was performed for panels A and B and Mann-Whitney non-parametric test was used when two groups were compared (panels C–E). n=6–8 per group. *p<0.05, **p<0.01 and ***p<0.001 (see also online supplemental figure 5). HCA, hyocholic acid; HDCA, hyodeoxycholic acid; βMCA, beta muricholic acid; ωMCA, omega muricholic acid; TβMCA, tauro-beta muricholic acid.

Figure 4

The beneficial metabolic effects of oligofructose (OFS) supplementation are dependent on TGR5 pathway activity. (A and B) Body weight gain (BWG) (n=10–12 per group) and oral glucose tolerance test (OGTT) (n=7–11 per group) of wild-type (WT) and TGR5 KO mice. (C) Fasting glucose levels (n=10–12 per group). (D) Fasting insulin levels (n=7–11 per group). (E) Caecal bile acid levels (n=7–10 per group). (F) Colonic gene expression of preproglucagon (Gcg), neurogenin 3 (Ngn3) and proconvertase 1 (Pcsk1) (n=6–8 per group). (G) Colonic L-cell density (n=5–8 per group) and active GLP-1 levels in the portal vein (10–12 per group). Data are presented as mean±SEM. Mixed-effects model test was used for BWG analysis (****p<0.0001: WT WSD-OFS vs WT WSD, ####p<0.0001: WT WSD-OFS vs TGR5KO WSD, $$$$p<0.0001: WT WSD-OFS vs TGR5KO WSD-OFS). Two-way analysis of variance with Bonferroni’s multiple comparison test was performed for panels B to G. D: diet, G: genotype, D&G: diet×genotype. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001 (see also online supplemental figure 6). HCA, hyocholic acid; HDCA, hyodeoxycholic acid; βMCA, beta muricholic acid; ωMCA, omega muricholic acid; TβMCA, tauro-beta muricholic acid; WSD, western style diet.

Figure 5

Hyodeoxycholic acid (HDCA) treatment improves host glucose metabolism in a TGR5-dependent mechanism. (A) Body weight gain (BWG) of wild-type (WT) mice treated orally with vehicle or 50 mg/kg of HDCA. (B) Fasting glucose and insulin levels. (C) Oral glucose tolerance test (OGTT) of WT mice treated intraperitoneally with vehicle or 5 µg of Exendin 9–39 20 min before OGTT. (D) Insulin levels during OGTT. (E) BWG of TGR5 KO mice treated orally with vehicle or HDCA (50 mg/kg of body weight). (F) Fasting glucose and insulin levels. (G) OGTT of TGR5 KO mice. (H) Insulin levels during OGTT. Data are presented as mean±SEM. Two-wayanalysis of variance with Bonferroni’s multiple comparison test was performed for BWG analysis and insulin secretion test. Mann-Whitney non-parametric test was used when two groups were compared. n=6–10 per condition for WT group and n=5 per condition for TGR5 KO group. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001 (see also online supplemental figure 7).

Figure 6

Exendin-4 supplementation improves the metabolic defect of TGR5 KO fed western-style diet (WSD) enriched with oligofructose (OFS). (A and B) Body weight gain (BWG) and oral glucose tolerance test (OGTT) of TGR5 KO mice fed with WSD-OFS and receiving a daily dose of 0.9% of NaCl as a vehicle or exendin-4 (2 nmol/kg/day) using Alzet minipumps. (C) Fasting glucose and insulin levels. (D) Insulin levels during OGTT. (E) Body composition at sacrifice. Data are presented as mean±SEM. Two-way analysis of variance with Bonferroni’s multiple comparison test was performed for BWG analysis and Mann-Whitney non-parametric test was used when two groups were compared. n=5–6 per group. *p<0.05, **p<0.01, and ***p<0.001.