Abstract
Polycystic ovary syndrome (PCOS) is the most common endocrinopathy affecting 10% of reproductive-aged women. The clinical hallmarks of PCOS include hyperandrogenism, oligo-anovulation, and polycystic ovarian morphology. Females with PCOS also experience significant metabolic and psychosocial comorbidities. Experimental PCOS can be modelled by prenatal androgen exposure (PNA), a suspected aetiology, allowing investigation of the unknown underpinning mechanism. Adult PNA mice recapitulate the cardinal features of PCOS, and exhibit aberration of the hypothalamus-pituitary-gonadal axis coordinating reproduction.
Both adult PCOS patients and PNA female mice treated with an androgen receptor (AR) antagonist demonstrate restoration of reproductive function, indicating AR as a clinical target. Although the tissue mediating hyperandrogenism remains undefined, evidence from alternative models of PCOS suggest the brain has a critical role. Moreover, amelioration of PCOS-like reproductive traits, modelled by peripubertal chronic dihydrotestosterone exposure, was reported in female mice with a cell-specific deletion of androgen receptor from neurons (NeurARKO). It remains unknown whether central androgen signalling is also required for the development of PNA-induced PCOS-like features.
This study aimed to investigate whether NeurARKO is protective for the development of PCOS-like phenotypes following prenatal androgen exposure. As neuronal androgen signalling has been shown to mediate the effects of androgen excess, we hypothesised that an amelioration of PNA-driven PCOS-like traits would be observed in NeurARKO mice.
To generate female NeurARKO offspring, androgen receptor floxed mice were mated with Camkll-Cre mice. During late gestation, pregnant dams were injected with either DHT to programme PCOS-like features, or oil as a control. Neuron-specific loss of AR expression was confirmed by determining the period of CamkllCre reported expression and immunohistochemistry for AR in adult brain sections. Reproductive features of PNA NeurARKO females were then characterised to determine the extent of restoration. Pubertal onset was measured by the day of vaginal opening and day of first estrus. Reproductive cycling was determined by recording vaginal cytology during adulthood. Moreover, ovarian histology was assessed following perfusion to determine the frequency of corpus lutea, preovulatory follicles, and to assess the follicle wall morphology.
Reported expression of CamkllCre indicates Cre mediated excision of AR occurs prior to prenatal androgenisation. The NeurARKO genotype was found to be protective of the PNA-induced delay in pubertal onset in a small cohort of female mice, suggesting neuronal androgen signalling mediates early androgen excess programming of delayed pubertal onset. No restoration of estrous cyclicity was observed in adult PNA NeurARKO mice despite the presence of corpora lutea in ovarian sections. Thus, our findings suggest that the brain may not be the sole mediator of PNA programming PCOS-like features manifesting in adulthood. Further elucidating the central role of androgen signalling underpinning PCOS pathogenesis is essential for the development of mechanism-based therapeutics.