The combined impact of testosterone and Western-style diet on endometriosis severity and progression in rhesus macaques†

Abstract

Polycystic ovary syndrome (PCOS) is associated with irregular menstrual cycles, hyperandrogenemia, and obesity. It is currently accepted that women with PCOS are also at risk for endometriosis, but the effect of androgen and obesity on endometriosis has been underexplored. The goal of this study was to determine how testosterone (T) and an obesogenic diet impact the progression of endometriosis in a nonhuman primate (NHP) model. Female rhesus macaques were treated with T (serum levels approximately 1.35 ng/ml), Western-style diet (WSD; 36% of calories from fat compared to 16% in standard monkey chow) or the combination (T + WSD) at the time of menarche as part of a longitudinal study for ~7 years. Severity of endometriosis was determined based on American Society for Reproductive Medicine (ASRM) revised criteria, and staged 1-4. Stages 1 and 2 were associated with extent of abdominal adhesions, while stages 3 and 4 were associated with presence of chocolate cysts. The combined treatment of T + WSD resulted in earlier onset of endometriosis and more severe types associated with large chocolate cysts compared to all other treatments. There was a strong correlation between glucose clearance, homeostatic model assessment for insulin resistance (HOMA-IR), and total percentage of body fat with presence of cysts, indicating possible indirect contribution of hyperandrogenemia via metabolic dysfunction. An RNA-seq analysis of omental adipose tissue revealed significant impacts on a number of inflammatory signaling pathways. The interactions between obesity, hyperandrogenemia, and abdominal inflammation deserve additional investigation in NHP model species.

Keywords: endometriosis; metabolism; obesity; polycystic ovary syndrome; testosterone.

Figure 1

Rhesus late-stage endometriotic lesions from three separate T + WSD-treated females (Animal 1, Animal 2, Animal 3) show glandular (gl)-like and stromal (str)-like characteristics and expected expression of proteins associated with endometrial function. Expression of androgen receptor (AR, A–C), estrogen receptor α (ESR1, D–F), progesterone receptor (PGR, G–I), and marker of proliferation ki-67 (MKI67/Ki67, J–L) in endometriotic lesions. All tissues were collected at necropsy for late-stage disease (stage 4). Positive protein expression is denoted by brown staining, while blue counter-stained nuclei are negative for these proteins. All images were visualized at 20X magnification with the same brightness. Inset image to left in panel A shows hematoxylin and eosin (H&E) staining of glandular-like morphology. Inset images in right of panel A, and left in panels D, G, and J show negative IgG controls. Inset images in left of panels B, E, H and K show increased magnification images of gl-like structures. Inset images in left of panels C, F, I and L show increased magnification images of str-like cells.

Figure 2

Linear correlation between time to onset of diagnosis with stage 3 disease and BMI at time of diagnosis. BMI measurements of individual females at time of stage 3 endometriosis diagnosis were compared to the time of diagnosis during the study (in years). Females in treatment groups C (n = 4), T (n = 5), WSD (n = 6), and T + WSD (n = 9) are represented by individual symbols with colors by group (figure legend). Dashed line represents linear fit and Pearsons bias-adjusted correlation value, as well as significance, is depicted in graph. See Results section for further details.