Prediabetes and type 2 diabetes but not obesity are associated with alterations in bile acid related gut microbe-microbe and gut microbe-host community metabolism

Abstract

The interplay between bile acids (BAs) and metabolic diseases has gained importance in recent years, with a variety of studies investigating their relationship with diverging results. Therefore, in the present study we performed a detailed analysis of BA metabolism in 492 subjects with different metabolic phenotypes. Besides microbiomics and metabolomics this investigation included in silico analysis of community metabolism to examine metabolic interchange between different microbes as well as microbes and the human host. Our findings revealed distinct changes in the BA profiles of patients with diabetes and prediabetes, whereas obesity alone had no influence on circulating BAs. Impaired glycemic control led to increased circulating BAs, a shift toward more secondary BAs, and an increase in the ratio of glycine to taurine-conjugated BAs. Additional analyses revealed that the ratio of glycine to taurine conjugation demonstrated variations between the single BAs, cholic acid (CA), chenodeoxycholic acid (CDCA) and deoxycholic acid (DCA), regardless of the metabolic status, with CA having a higher fraction of taurine conjugation. Furthermore, we found that microbiome alterations are associated with BAs, independent of diabetes or obesity. Analysis of microbial community metabolism revealed differential relative pathway abundance in relation to diabetes, particularly those related to membrane and polyamine synthesis. Increased bacterial cross-feeding of polyamines, galactose, and D-arabinose also coincided with an increase in BA. Notably, our serum metabolome analysis mirrored several of the previously in silico predicted exchanged metabolites, especially amino acid metabolism. Therefore, targeting BA metabolism may be a future approach for the treatment of metabolic diseases, especially prediabetes and type 2 diabetes.

Keywords: Bile acids; gut community metabolism; gut microbiome; type 2 diabetes.

Figure 1.

Total bile acid concentration in the FoCus cohort (a) Comparison of underweight, normal weight and obese patients (b) Comparison of prediabetic, diabetic and healthy participants. c) Total bile acid concentration in normal weight, obese patients with and without prediabetes (d) Comparison of single bile acids in normal weight and obese patients without pre/diabetes. (e-h) Generalized linear regression evaluating the impact of gender, age and (e)metabolic health groups/f)T2D status/g)bmi/h)obesity with and without preT2D on total bile acids as the dependent variable.

Figure 2.

Conjugation of the bile acids (a) Percentages of unconjugated, glycine and taurine- conjugated bile acids in the normal weight group. (b) Ratios of glycine- to taurine- conjugated bile acids in the obesity groups (c) Glycine to taurine ratio of the single bile acids in the diabetes groups. (d-f) Generalized linear regression evaluating the impact of BMI, gender, age and diabetes status for (d) ca/ (e)cdca/ (f) dca on the ratio of glycine to taurine conjugation as the dependent variable. (g) Generalized linear regression evaluating the impact of metabolic status, age and gender on the ratio of glycine to taurine conjugation as the dependent variable.

Figure 3.

Association of bile acid traits with biological sex and metabolic biomarkers (a) Correlation matrix of bile acid weighted score (BWS), total bile acids and single bile acids with metabolic serum biomarkers. Only significant correlations (fdr >0.05) are displayed. Numbers in boxes refer to correlation coefficients, with darker coloring indicating stronger correlation. List of fdr-corrected p-values can be found in supplement S4–1. (b) total BA levels in males and females (c) Total BA levels in normal, prediabetic and diabetic probands by sex.

Figure 4.

Association of chronic diseases at the baseline (a) bile acid weighted score (BWS) (b) deoxycholic acid (DCA) (c) Total bile acids significance was tested using a logistic regression. Bile acids (BA), coronary heart disease (CHD).

Figure 5.

Pathway abundance of Spermidine synthesis in BWS quartiles.

Figure 6.

Enriched kegg-pathways a) in the serum and b) in the urine of patients with high circulating total bile acids.

Figure 7.

Total bile acid concentration in intervention cohorts. Longitudinal changes in nine BAs during surgical and non-surgical intervention. Associations were tested using paired Wilcoxon and Kruskal-Wallis tests with fdr correction.