Androgens and the ovary
Between 10-15% of women suffer from polycystic ovary syndrome (PCOS), making it the most common cause of female infertility. Clinical features of PCOS include high circulating levels of ovarian androgens (T and A4), anovulation and obesity. The aetiology of this reproductive endocrinopathy is likely to be multifactorial, through the interplay of genetics, epigenetics and environmental factors. Primate research into sexual behaviour development noted that fetally androgenised monkeys developed symptoms like those of PCOS. There are now multiple animal models of PCOS using primates, sheep, rats and transgenic mice. The investigations described in this thesis use rodent models to examine the role of androgens in the pathogenesis of female infertility. An attempt to generate a granulosa cell specific androgen receptor knockout mouse model will first be described, followed by several studies into the developmental programming of female Wistar rat infertility and metabolism by steroid hormones. Initial investigations showed that testosterone proprionate (TP) administered to female rats during different windows of fetal and neonatal life alters the reproductive and metabolic axes of the adult animals. Fetal plus neonatal TP exposure led to complete ovarian dysgenesis, while postnatal exposure produced a PCOS-like phenotype. Animals which received TP postnatally were heavier and had an increased proportion of primordial follicles in their ovaries by postnatal day (pnd) 90 of life. Evaluation of this PCOS model showed that neonatally androgenised rats had ovarian follicles with larger antra and a greater ovarian stromal compartment. In addition, these animals were heavier when compared to controls. However, unlike human studies, neonatally androgenised rats showed no differences in circulating gonadotrophin or ovarian androgen levels. Nor did they show any programming effect of neonatal TP upon the theca interna by pnd 90. Further investigations to narrow the windows and dose of TP required to produce a PCOS phenotype showed that TP administered in an early window of neonatal life, between postnatal days (pnd) 1-6 not only led to anovulation, but potentially reprogrammed the hypothalamic-pituitary axis, as there was minimal gonadotrophin response to reduced ovarian negative feedback (inhibin B and estradiol) in these rats. Neonatal TP also affected the rat metabolic axis with adult animals becoming heavier after weaning without any change in food intake. Animals developed mesenteric and retroperitoneal obesity along with insulin resistance (IR). Increased hepatic glucocorticoid turnover and altered adipokine expression were also noted in neonatally androgenised females, possibly contributing to the pathogenesis of obesity. No effect of TP dose upon the severity of infertility or metabolic abnormalities in adult animals was observed. To delineate which features of the rat PCOS model resulted from androgenic, estrogenic or corticosteroid action, a final study used administration of different steroids during the early window of postnatal life: TP, estradiol valerate (EV), dihydrotestosterone (DHT), dehydroepiandrosterone (DHEA) and dexamethasone (DEX). The anovulatory PCO-like phenotype observed with TP was also seen in animals which received EV, but not those which received DHT, DHEA or DEX. TP and EV treatment also resulted in a reduction of ovarian follicle numbers and activated follicle proportions, with an increase in primordial follicle proportions. Although glucose tolerant, animals treated with TP and EV were highly IR. Unlike dexamethasone, DHT and DHEA also produced IR in adult animals, to a lesser extent than TP and EV. Taken collectively, the results described in this thesis demonstrate that the PCOS-like phenotype observed in the neonatally androgenised female rat is likely to be due to the estrogenic actions of testosterone, potentially through as yet unknown epigenetic mechanisms. The programming of the metabolic components described may additionally be due to the actions of androgens. Furthermore, these studies demonstrate a novel estrogenic effect of neonatal steroids upon primordial follicle populations and show that the neonatally androgenised rat may be a rational PCOS model in a poly-ovulatory species.
The following license files are associated with this item: