999: A distinct cervicovaginal microbiome dysbiosis precedes adverse reproductive outcomes in a primate model of PCOS

Polycystic ovary syndrome (PCOS) is a condition of hyperandrogenism, oligomenorrhea and polycystic ovarian morphology that affects 5-10% of American women and is associated with decreased fertility and adverse pregnancy outcomes. We have shown the effects of chronic hyperandrogenemia and high fat diet (HFD), two features of PCOS on the metabolic and reproductive activity of a primate. While the etiology of PCOS is unclear, there is a potential role for the microbiome in the pathogenesis. Thus, we aimed to investigate if an inflammatory associated microbial dysbiosis occurs in our primate model of PCOS. In our primate model of PCOS, Rhesus macaque females have subcutaneous cholesterol (control) or testosterone (T) implants placed and are fed either a control (14% fat) or high fat diet (36% fat). This design created 4 cohorts: control (C), T alone (T), HFD alone (HFD) and T+HFD (T+HFD) (n=10/group). T and HFD cohorts have decreased fecundity and fertility respectively. All of pregnancies were normal in the C, T, HFD groups, whereas only 57% of pregnancies were normal in the T+HFD group. (Bishop et al 2018). Vaginal and cervical samples were collected during menses, follicular, and luteal phase prior to the fertility trial. DNA was extracted and subjected to 16S pyrosequencing. Microbiome composition was analyzed with standard methods and compared by cohort and examination of inflammation associated genera was compared. We found significantly distinct compositions between cohorts using beta diversity (p=0.001). Furthermore, we found significant increases in taxa associated with adverse pregnancy outcomes, such as Sneathia and Mobiluncus (p < 0.05, Fig.A). When examining the co-occurrence of taxa, we found that Lactobacillus significantly co-occurred with Prevotella and Streptococcus (p<0.05) while Sneathia and Mobiluncus co-occurred together (p <0.001). Additionally, we found functional differences between cohorts using inferred metagenomics (PICRUSt, Fig. C). Our data demonstrate that both diet and androgen treatment result in alterations in the genital tract microbiome characteristic of vaginal dysbiosis associated with inflammation. We speculate that inflammatory associated dysbiosis of the genital tract plays a role in adverse fertility and pregnancy outcomes seen with PCOS.