Gut Steroids and Microbiota: Effect of Gonadectomy and Sex

Abstract

Sex steroids, derived mainly from gonads, can shape microbiota composition; however, the impact of gonadectomy and sex on steroid production in the gut (i.e., gut steroids), and its interaction with microbiota composition, needs to be clarified. In this study, steroid environment and gut steroidogenesis were analysed by liquid chromatography tandem mass spectrometry and expression analyses. Gut microbiota composition as branched- and short-chain fatty acids were determined by 16S rRNA gene sequence analysis and gas chromatography flame ionisation detection, respectively. Here, we first demonstrated that levels of pregnenolone (PREG), progesterone (PROG), and isoallopregnanolone (ISOALLO) were higher in the female rat colon, whereas the level of testosterone (T) was higher in males. Sexual dimorphism on gut steroidogenesis is also reported after gonadectomy. Sex, and more significantly, gonadectomy, affects microbiota composition. We noted that a number of taxa and inferred metabolic pathways were associated with gut steroids, such as positive associations between Blautia with T, dihydroprogesterone (DHP), and allopregnanolone (ALLO), whereas negative associations were noted between Roseburia and T, ALLO, PREG, ISOALLO, DHP, and PROG. In conclusion, this study highlights the novel sex-specific association between microbiota and gut steroids with possible relevance for the gut-brain axis.

Keywords: branched- and short-chain fatty acids; gastrointestinal tract; gut microbiota; mucosa; pregnenolone; sex dimorphism; sex steroids; stool.

Figure 1

Steroid levels assessed by LC-MS/MS in the colons of male or female gonadectomized (GDX) rats compared with a sham-operated control. The panels represent the levels of (a) pregnenolone (PREG); (b) pregnenolone sulfate (PREG sulfate); (c) progesterone (PROG); (d) dihydroprogesterone (DHP); (e) allopregnanolone (ALLO); (f) isoallopregnanolone (ISOALLO); (g) testosterone (T); (h) dihydrotestosterone (DHT); (i) 5α-androstane-3α, 17β-diol (3α-diol); (j) 17β-estradiol (17β-E). Data are expressed as pg/mg and are the mean ± SEM. n = 6 animals for each experimental group. u.d.l. = under detection limit. Limit of detection for ALLO is 0.1 pg/mg, for T is 0.02 pg/mg and for DHT is 0.05 pg/mg. The two-way ANOVA was used for statistical analysis. * p < 0.05, ** p < 0.01, *** p < 0.001. The multiple comparison Tukey post-hoc test was used. ° p < 0.05, °° p < 0.01, °°° p < 0.001 vs. male control group; ### p < 0.001 vs. male GDX group; § p < 0.05, §§ p < 0.01, §§§ p < 0.001 vs. female control group.

Figure 2

Expression of enzymes detected in rat colons. Data were obtained by real time PCR in panels (a) cholesterol side-chain cleavage enzyme (P450scc); (b) sterol regulatory element-binding protein 2 (SREBP2); (c) 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMG-CoA R); (d) low-density lipoprotein receptor (LDL R); (e) steroid sulfatase (STS); (f) hydroxysteroid sulfotransferase (SULT2B1); (g,h) 3β-hydroxysteroid dehydrogenase (3β-HSD); (i) 5α-reductase type II (5α-R II); (j) 5α-reductase type I (5α-R I); (k) 3α-hydroxysteroid-oxidoreductase (3α-HSOR), and by Western blot analysis in panel (l) aromatase. The columns represent the mean ± SEM after normalization with 36B4 rRNA in colon of adult rats. n = 6 animals for each experimental group. The unpaired Student’s t-test was used for statistical analysis. * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 3

Microbial diversity. (a) Alpha diversity (richness and phylogenetic diversity) was increased following gonadectomy, but it did not differ significantly between sexes or substrates. (b) Beta diversity (CCA of sample composition), showing the significant influence of both experimental group and sex, as illustrated by the strong separation of labels (centroids) for F and M controls, which is resolved by GDX. The substrate had a non-significant effect, as illustrated by the co-location of labels near the origin. (c) The same CCA showed taxa as points closest to samples where they were most abundant (e.g., Gammaproteobacteria tended to be more abundant in control-F). Crossed circles indicate location of group centroids. Paired samples are joined by a line. Asterisk (*) denotes significant difference at FDR < 0.05, whereas a horizontal bracket indicates that the difference is between the GDX and control overall. Abbreviations: F: female; M: male; GDX: gonadectomy; CCA: constrained correspondence analysis.

Figure 4

Differential effects of sex and substrate on the abundances of individual microbial genera. (a) Taxa which were differentially abundant (FDR < 0.05) between mucosal and stool samples for different groups across the experimental design. (b) Taxa which were differentially abundant (FDR < 0.05) between female and male samples for different groups across the experimental design.

Figure 5

Microbial composition and taxa significantly affected gonadectomy: (a): Phylum level; (b): Family level; (c): Genus level; (d): Summary of genera which differ significantly across the study (FDR < 0.05), whether in both females (F) and males (M), F only, or M only. Some genera (e.g., Bacteroides, Alloprevotella) appear higher in both the control and GDX rats as there are multiple distinct members (e.g., species) within that genus, which show different responses to sex and GDX. (a–c): all taxa > 3.5% abundance, in 10% of samples (remainder grouped to “etc.”). “*” indicates taxa differentially abundant between experimental groups (GDX, control); “†” indicates taxa differentially abundant between substrates (mucosa, stool); “§” indicates differentially abundant taxa between sexes (male, female).

Figure 6

Faecal short-chain fatty acid and branched chain fatty acid levels in all groups.

Figure 7

Microbiome features associated with control are positively correlated (green) with an abundance of multiple steroids. This also extends to predicted metabolism, which only shows positive correlations with steroid levels; many of these steroids are also significantly more abundant in the control rats (*, Figure 2). This unified pattern suggests that the control gut microbiome metabolizes a range of host steroids and is negatively affected when these substrates are depleted by gonadectomy. Disruption via GDX favours an alternate cohort of bacteria, but it does not foster new predicted metabolic activities. Heatmap: colours reflect positive (green) or negative (purple) Spearman’s ρ correlations between microbial features (ASVs, predicted ECPs and biological parameters: steroids, fatty acids, Shannon’s alpha diversity). Top row (GDX effect): summarizes associations relevant to experimental design; bottom row (non-spec. effect): summarizes incidental associations which do not cluster with the main experimental effect. Right-hand margin: unit-scaled abundances, significant differences if applicable (*), and categories for biological parameters in this study (steroid, fatty acids, alpha diversity). Bottom margin: scaled abundances, and microbial taxonomy (ASVs) or pathway types (ECPs) of features. All microbial features (ASV, EC) shown here were found to differ significantly between control and GDX (FDR < 0.05).

Figure 8

Schematic representation of gut steroidogenesis. Framed steroids reported have been assessed by LC-MS/MS. Pink frame: higher levels in intact female; blue frame: higher levels in intact male; grey frame: no sex difference. Arrows represent the effect of gonadectomy in female (pink) and in male (blue).