Exposure to a Healthy Gut Microbiome Protects Against Reproductive and Metabolic Dysregulation in a PCOS Mouse Model

Abstract

Polycystic ovary syndrome (PCOS) is a common endocrine disorder affecting ∼10% to 15% of reproductive-aged women worldwide. Diagnosis requires two of the following: hyperandrogenism, oligo-ovulation or anovulation, and polycystic ovaries. In addition to reproductive dysfunction, many women with PCOS display metabolic abnormalities associated with hyperandrogenism. Recent studies have reported that the gut microbiome is altered in women with PCOS and rodent models of the disorder. However, it is unknown whether the gut microbiome plays a causal role in the development and pathology of PCOS. Given its potential role, we hypothesized that exposure to a healthy gut microbiome would protect against development of PCOS. A cohousing study was performed using a letrozole-induced PCOS mouse model that recapitulates many reproductive and metabolic characteristics of PCOS. Because mice are coprophagic, cohousing results in repeated, noninvasive inoculation of gut microbes in cohoused mice via the fecal-oral route. In contrast to letrozole-treated mice housed together, letrozole mice cohoused with placebo mice showed significant improvement in both reproductive and metabolic PCOS phenotypes. Using 16S rRNA gene sequencing, we also observed that the overall composition of the gut microbiome and the relative abundance of Coprobacillus and Lactobacillus differed in letrozole-treated mice cohoused with placebo mice compared with letrozole mice housed together. These results suggest that dysbiosis of the gut microbiome may play a causal role in PCOS and that modulation of the gut microbiome may be a potential treatment option for PCOS.

Copyright © 2019 Endocrine Society.

Figures

Figure 1.

Cohousing letrozole mice with placebo mice protected against development of the PCOS metabolic phenotype. Design of cohousing study with pubertal female mice housed two per cage in three different housing arrangements that resulted in four groups of mice (n = 8 per group): P, LET, Pch, and LETch (A). Letrozole treatment resulted in metabolic dysfunction compared with placebo including (B–F) increased weight, abdominal adiposity, FBG, and insulin levels and insulin resistance. (B–F) Compared with LET mice, LETch mice showed a decrease in body weight, a decrease in abdominal adiposity, a decrease in FBG and insulin levels, and restored insulin sensitivity. Graph error bars represent SEM. Different letters or an asterisk symbol were used to indicate significant differences in a one-way ANOVA or repeated-measures two-way ANOVA followed by post hoc comparisons with the Tukey-Kramer honestly significant difference test (P < 0.05).

Figure 2.

Letrozole mice cohoused with placebo mice did not become hyperandrogenemic or acyclic. The cohousing study included four groups of mice (n = 8 per group): P, LET, Pch, and LETch. Letrozole treatment resulted in increased (A) testosterone and (B) LH levels compared with placebo. LETch mice displayed a decrease in (A) testosterone and (B) LH and (C) a restoration of estrous cyclicity compared LET mice stuck in diestrus. Stages of the estrous cycle are indicated as diestrus (D), metestrus (M), estrus (E), and proestrus (P). Graphs illustrating the estrous cycle stages of representative mice from the four groups are shown. Graph error bars represent SEM. Different letters were used to indicate significant differences in a one-way ANOVA followed by post hoc comparisons with the Tukey-Kramer honestly significant difference test (P < 0.05).

Figure 3.

Cohousing letrozole mice with placebo mice improved the ovarian phenotype. The cohousing study included four groups of mice (n = 8 per group): P, LET, Pch, and LETch. (A) Letrozole treatment resulted in a lack of corpora lutea, cystlike follicles, and hemorrhagic cysts in the ovaries compared with placebo mice. (A) Unlike LET mice, LETch mice lacked polycystic ovaries, and their ovaries contained corpora lutea (CL) which is evidence of ovulation. Scale bars represent 250 µm. (B–E) Letrozole treatment also resulted in increased ovarian weight and increased mRNA expression of several ovarian genes important in ovarian follicular development and steroidogenesis. (B) Ovarian weight was lower in LETch mice compared with LET mice. Fshr and Cyp19 mRNA levels were similar between LET and LETch mice, whereas Cyp17 was lower in LETch mice compared with LET mice. Graph error bars represent standard error of the mean. Different letters were used to indicate significant differences in a one-way ANOVA followed by post hoc comparisons with the Tukey-Kramer honestly significant difference test (P < 0.05).

Figure 4.

Gut microbiome was similar in all cohousing treatment groups before treatment. The cohousing study included four groups of mice: P, LET, Pch, and LETch (n = 8 per group with the exception of n = 7 for P time 4 and n = 6 for Pch and LETch time 5). No significant differences in (A) gut microbial richness (α-diversity, Faith PD) or (B) community composition (β-diversity, weighted UniFrac) were observed among cohousing treatment groups before treatment (week 0; n = 8 per group). One-way ANOVA was used to compare α-diversity among the groups, and analysis of similarity (ANOSIM) test was used to compare β-diversity among the groups.

Figure 5.

Cohousing letrozole mice with placebo mice did not restore α-diversity of the gut microbiome. The cohousing study included four groups of mice: (A) P, (B) LET, (C) Pch, and (D) LETch (n = 8 per group with the exception of n = 7 for P time 4 and n = 6 for Pch and LETch time 5). α-Diversity as approximated by Faith PD ranked estimate was graphed over time for the four groups. Results of linear regression model (LM) and P value are in the box insets, and the gray shaded area indicates the 95% CI for the line of best fit. P values for the linear mixed effects model (LME) were obtained by the likelihood ratio test of the full model, with the effect in question (time) against the model without the effect in question, and are in the box insets.

Figure 6.

Cohousing letrozole mice with placebo mice influenced the overall composition of the gut bacterial community over time. The cohousing study included four groups of mice: P, LET, Pch, and LETch (n = 8 per group with the exception of n = 7 for P time 4 and n = 6 for Pch and LETch time 5). (A) Unconstrained PCoA of weighted UniFrac distances demonstrated changes in the microbial composition (β-diversity) among samples collected after treatment. Permutational ANOVA of the weighted UniFrac distances indicated that cohousing had a strong influence on the gut microbial community (P = 0.001). (B) Constrained CAP of weighted UniFrac distances further illustrated the relationship between β-diversity and posttreatment with a significant effect of constraining the data based on the cohousing treatment group (P = 0.001). (C–H) Samples from the different groups were then compared at each time point. Permutational ANOVA of the weighted UniFrac distances was done for each time point.

Figure 7.

Specific bacterial genera were associated with improvement of the PCOS phenotype during cohousing. The cohousing study included four groups of mice: P, LET, Pch, and LETch (n = 8 per group with the exception of n = 7 for P time 4 and n = 6 for Pch and LETch time 5). Results from the DESeq2 differential abundance analysis were expressed as log2 fold change for (A) the comparison of P and LET mice and (B) the comparison of LETch and LET mice. Positive log2 fold changes represent bacterial genera increased in (A) P mice relative to LET mice or (B) LETch relative to LET mice, whereas negative changes represent bacterial general increased in (A) LET relative to P mice or (B) LET relative to LETch mice. *P < 0.05; **P < 0.01; ***P < 0.001.