Anatomical and functional characterization of hypothalamic insulin signaling in the metabolic control of reproductive function in mice

Central insulin signaling plays a key role in regulating the neuroendocrine reproductive axis. Mice exhibiting brain-specific insulin receptor gene (Insr) deletion (NIRKO mice) or diabetic insulin insufficiency display hypothalamic hypogonadism in association with absent or decreased insulin signaling. However, the specific neuronal target cells mediating insulin’s central effects on the reproductive axis remain largely unidentified. Using a combination of techniques, including immunohistochemistry to identify insulin responsive neurons as well as cell type-specific insulin receptor (InsR) knockout (KO) experiments, I investigated whether insulin’s central effects on the neuroendocrine reproductive axis are mediated directly via gonadotropin-releasing hormone (GnRH neurons), which are the final central drivers of reproductive function, or indirectly via putative insulin-sensitive neuronal populations implicated in the metabolic control of fertility. My first aim was to determine whether GnRH neurons are direct targets of insulin signaling. Encouragingly, I detected Insr mRNA and InsR protein in GnRH-immunolabeled neurons. However, I found no evidence of insulin-induced phosphorylated AKT (pAKT) or phosphorylated ERK1/2 (pERK1/2) in GnRH neurons. These results demonstrate GnRH neurons can directly respond to insulin; however, there was no clear evidence of insulin-induced signaling in GnRH neurons. This suggests insulin’s central effects on the reproductive axis are likely mediated by GnRH afferent neurons. Therefore, I next investigated whether insulin’s central effects on fertility are mediated via kisspeptin neurons, which are the most potent known activators of GnRH neurons. I used the Cre-lox system to generate kisspeptin-specific InsR knockout (KIRKO) mice and then assessed multiple reproductive parameters. Disappointingly, no significant differences in puberty onset, estrous cyclicity, or fertility were observed in the female or male KIRKO mice in comparison to their control littermates. The results from this experiment demonstrate InsR signaling via kisspeptin cells is not critically involved in mediating insulin’s central effects on the neuroendocrine reproductive axis. I then broadened my approach and next investigated whether insulin signaling via either gamma-aminobutyric acid (GABA) or glutamate-expressing neurons plays a critical role in the metabolic regulation of reproduction by generating GABA-specific and glutamate-specific InsR knockout mice, again using the Cre-lox system. Surprisingly, given the widespread expression of GABAergic and glutamatergic neurons, and hence the significant reduction in hypothalamic Insr mRNA expression in both GABA-specific and glutamate-specific InsR knockout mice, no reproductive abnormalities were found. However, female GABA-specific InsR knockout mice exhibited a significant metabolic phenotype, demonstrating insulin signaling via GABAergic neurons plays a critical role in the regulation of energy homeostasis. Nevertheless, the neuronal mediators of insulin’s central effects on fertility remained unidentified. My next aim was to test whether insulin signaling in neuronal cells, versus neuroglia cells, is indeed critical for normal metabolic control of reproductive function. The Cre-lox system was utilized once more to generate forebrain neuron-specific InsR knockout mice. Consistent with the general belief that neuronal insulin signaling plays a role in modulating metabolic function, female and male knockout mice exhibited a significant metabolic phenotype. However, in the absence of insulin signaling via forebrain neurons, male and female mice exhibited normal reproductive maturation, estrous cyclicity, and fertility. Collectively, the findings from these InsR KO studies challenge the mainstream view that neuronal insulin signaling is a critical regulator of reproductive function. The final aim of this thesis was to investigate whether impaired insulin signaling in the brain plays an etiological role in the pathogenesis of polycystic ovary syndrome (PCOS), which is a leading cause of infertility in women. A pre-natal androgen (PNA)-treated mouse model of PCOS previously shown to exhibit the hallmark features of PCOS was used to determine whether impaired central insulin signaling is evident in PCOS. However, under fasting conditions as well as in response to insulin treatment, no impairments in hypothalamic pAKT activation were observed between the PNA-treated and control mice, suggesting impaired hypothalamic insulin signaling is not central to the pathogenesis of PCOS. In summary, this thesis challenges the widely held view that neuronal insulin signaling plays a critical role in the metabolic regulation of fertility