Effects of estradiol on gonadotropin-releasing hormone neurons

Abstract

The gonadotropin-releasing hormone (GnRH) neurons, located within the hypothalamus of the mammalian brain, have been established as the key output cells mediating all influences that control fertility, conveyed by the central nervous system. Gonadotropins (luteinizing hormone; LH and follicle stimulating hormone; FSH), secreted by the pituitary in response to hypothalamic signals from GnRH neurons, stimulate the release of gonadal steroids from the ovaries. The gonadal steroid, estradiol (E2), functions as a diverse and indispensible hormone capable of modulating multiple facets of neuronal functioning. Circulating E2 provides neuroendocrine feedback to the hypothalamus and pituitary, through negative and positive feedback mechanisms depending on the hormonal milieu across the ovulatory cycle. E2 exerts a negative feedback response on GnRH neurons to keep basal gonadotropin secretion low and switches over to positive feedback to generate the pre-ovulatory gonadotropin surge, required for ovulation. Although GnRH neurons express estrogen receptor β (ERβ), the other receptor subtype, ERα, has been described as the main receptor mediating positive feedback. However, a role for both ERα and ERβ in estrogen negative feedback remains likely. The mechanisms underlying negative feedback actions of estrogen upon the GnRH neuronal network are unclear and likely to require different mechanisms of E2 action including direct actions through ERβ-expressing GnRH neurons or indirectly through ERα-expressing afferent inputs to GnRH neurons. E2 is known to act through the non-classical pathway to rapidly activate intracellular signaling cascades to alter cellular function and it is believed that these rapid non-classical E2 actions are likely contributors involved in mediating estrogen negative feedback in mice. The purpose of this study was to investigate the mechanisms mediating estrogen negative feedback on the GnRH neuronal system.

The specific intracellular signaling pathways mediating rapid non-classical E2 actions were first addressed employing both in vivo and in vitro acute brain slice approaches involving dual-label fluorescence and peroxidase-based immunohistochemistry in a negative feedback model. This model involved ovariectomizing mice and testing the ability of the acute E2 treatment to rapidly induce the phosphorylation of the cyclic-AMP response element binding protein (CREB) and the upstream molecule, extracellular signal-regulated kinase 1 and 2 (ERK1/2). The administration of E2 to adult female mice in vivo resulted in the rapid activation of ERK1/2 in GnRH neurons in a time-dependent manner and in vitro studies using pharmacological antagonists showed that ERK1/2 was essential for E2-induced CREB phosphorylation in GnRH neurons. Upstream to this, protein kinase A (PKA) and calcium/calmodulin dependent protein kinase type II (CaMKII), but not protein kinase C were found to be necessary for E2-induced phosphorylation of ERK1/2 and CREB. Interestingly, this rapid E2 signaling cascade in GnRH neurons was found to require both direct and indirect E2 actions. E2 failed to phosphorylate ERK1/2 and CREB in GnRH neuron- specific ERβ knockout mice in vivo. Equally, however, a cocktail of tetrodotoxin and γ-aminobutyric acid (GABA)/glutamate receptor antagonists also blocked E2-induced ERK1/2 phosphorylation in GnRH neurons in vitro. The results from this study also show that the non-classical effects of E2 are dependent upon both direct ERβ mechanisms as well as indirect actions mediated by afferent inputs most probably through other ERα-expressing neurons projecting to the GnRH neurons.

The next part of the study examined the involvement of receptors, ERα and ERβ, in mediating negative feedback in the adult brain using conditional tamoxifen-inducible CaMKIIα promoter driven neuron-specific ERα and ERβ knockout mouse lines. Mice devoid of ERα from all CaMKIIα-expressing forebrain neurons in adulthood exhibited disrupted estrous cycles while the elimination of ERβ from CaMKIIα-expressing neurons displayed normal estrous cyclicity. Measurements of plasma LH levels by radioimmunoassay prior to ovariectomy, 2 weeks post-ovariectomy and 3 h after a single E2 injection provide an indication of estrogen’s negative feedback influence on GnRH neurons. E2 treatment failed to suppress LH levels in the tamoxifen inducible ERα knockouts but was effective in the inducible ERβ knockouts, suggesting that CaMKIIα-expressing ERα neurons but not CaMKIIα-expressing ERβ neurons play an essential role in exerting estrogen negative feedback actions. The effects of non-classical E2 signaling and the ability of E2 to regulate estrogen negative feedback actions in aging were examined to further investigate the functional implications of negative feedback. The ability of E2 to induce ERK1/2 phosphorylation was absent in both middle-aged GnRH neuron-specific ERβ knockout mice and their age-matched wild-type littermates. Although the middle-aged mice exhibited normal estrous cyclicity, the ability to exert negative feedback was dependent on ERβ on GnRH neurons during aging.

Together, the results from these studies suggest a new estrogen pathway in GnRH neurons that has the potential to mediate estrogen negative feedback and suggest that ERα is the critical receptor subtype controlling negative feedback. These observations provide more insight into E2 actions on the GnRH neuronal network, crucial to furthering our understanding of estrogen’s actions on GnRH neurons.