Abstract
Polycystic ovary syndrome (PCOS) affects approximately 20% of reproductive-aged women and is characterised by elevated androgens, menstrual irregularity, and polycystic ovary morphology. Although the etiology of PCOS is unclear, exposure to prenatal androgen excess (PNA) is implicated in programming reproductive dysfunction associated with the disorder. It is hypothesized that PNA directly disrupts steroid hormone feedback to key hypothalamic neurons that regulate gonadotropin releasing hormone (GnRH) neuron activity, leading to dysregulation of the hypothalamic-pituitary-ovarian (HPO) axis controlling reproduction. Although both pharmacological blockade and genetic deletion of androgen receptors (AR) can ameliorate PCOS-like features, the key target through which androgen excess exerts its effects remains unclear. Preclinical studies using neuron-specific AR knockout models implicate the brain as a central mediator, but whether it is the primary site driving PNA-induced reproductive dysfunction is unknown.
This project employed a reductionist approach to determine whether androgen signaling in the forebrain is required for the development of PCOS-like traits in PNA mice. Neuron specific AR knockout (NeurARKO) mice were generated by crossing ARfl/fl mice with CamKllĪ±-Cre+/- mice, and pregnant dams were treated with dihydrotestosterone or vehicle during late gestation. Deletion of AR during embryonic development was confirmed using immunohistochemistry. We hypothesised that PNA NeurARKO mice would be protected from developing PNA-induced reproductive and neuroendocrine PCOS-like features.
As expected, PNA wildtype mice exhibited delayed pubertal onset, disrupted reproductive cycles, elevated androgen levels, and subtle alterations in ovarian morphology, consistent with prior characterisations. Contrary to our hypothesis, these reproductive deficits persisted in PNA NeurARKO mice. In addition, PNA-induced downregulation of hypothalamic progesterone receptor expression was restored in PNA NeurARKO mice, suggesting that rescuing the sensitivity of GnRH neuronal afferents to progesterone feedback is insufficient to restore reproductive function. Analysis of the GABA-to-GnRH circuit revealed that PNA NeurARKO mice exhibited elevated GABAergic innervation of GnRH neurons, which was surprisingly undetected in PNA WT mice. Together, our findings indicate that AR-expressing forebrain neurons do not solely mediate PNA programming of PCOS-like features. Further investigation of androgen-sensitive targets beyond the hypothalamus are required to advance the development of mechanism-based therapies for women with PCOS.