Endocrine Abstracts

Androgen excess is a hallmark feature of polycystic ovary syndrome (PCOS), the most common form of anovulatory infertility. Clinical and preclinical evidence links developmental exposure to hyperandrogenism with programming the reproductive traits of PCOS. While the critical androgen targets remain to be determined, GABAergic neurons in the brain are postulated to be involved. GABA levels are elevated in the cerebrospinal fluid of PCOS patients, and both GABA transmission and innervation to gonadotrophin relseasing hormone (GnRH) neurons are elevated in prenatally androgenised models of PCOS. Here, we tested the hypothesis that androgen signalling in GABAergic neurons is critical in PCOS pathogenesis in a well-characterised, hyperandrogenic mouse model of PCOS. To generate mice with GABA neuron-specific knockout of androgen receptors (referred to as GABARKO), vesicular GABA transporter (VGAT)-ires-Cre (VGAT-Cre^± ) mice were crossed with androgen receptor flox (AR^fl/fl ) mice. Experimental animals were produced by crossing heterozygous VGAT-Cre^± ; AR^fl/wt females with homozygous AR^fl/Y males. To generate experimental animals with fluorescent GnRH neurons, VGAT-Cre^± ;AR^fl/wt females were bred with AR^fl/Y ;GnRH-GFP^+/+ males. VGAT-Cre^-/- ; AR^fl/fl females were used as controls (referred to as wildtype). Prenatally androgenized (PNA) PCOS-like mice and vehicle controls (VEH) were generated by injecting pregnant dams s.c. with 100 μl of either dihydrotestosterone (250 μg) dissolved in sesame oil or an oil vehicle alone on gestational days 16, 17 and 18. Female offspring were phenotyped for PCOS-like reproductive features and fixed brain sections were subsequently assessed for elevated GABAergic input to GnRH neurons using immunofluorescence and confocal microscopy. GABARKO did not impact the timing of pubertal onset, estrous cyclicity or ovarian morphology in VEH mice. As expected, wildtype PNA mice exhibited delayed pubertal onset and complete acyclicity compared to VEH mice. The majority of GABARKO PNA mice remained acyclic and ovarian morphology was unaffected across groups. However, while PNA predictably increased the density of putative GABAergic synapses to gonadotropin-releasing hormone (GnRH) neurons in adult WT mice, this PNA-induced plasticity in GABAergic innervation to the GnRH neurons was absent in GABARKO mice. Together, these findings suggest that while direct androgen signalling in GABA neurons is largely not required for the development of the reproductive PCOS-like traits in androgenised models of PCOS, developmental programming of GnRH neuron innervation is dependent upon androgen signalling in GABA neurons.