Abstract
Mammalian fertility is governed by a neural circuit that ultimately controls the activity of gonadotropin-releasing hormone (GnRH) neurons. Activation of GnRH neurons results in the pituitary release of luteinising hormone (LH) and follicle-stimulating hormone. Prior to ovulation, there is a surge in the release of LH, due to the increased activity of afferent kisspeptin neurons in the rostral periventricular area of the third ventricle (RP3V). These kisspeptin neurons are under the control of circadian inputs from the suprachiasmatic nucleus (SCN) and ovarian steroid hormone feedback. One key circadian input to kisspeptin neurons is the vasopressin (AVP) neurons from the SCN. Recent evidence has shown that AVP increases RP3V kisspeptin neuron activity, in relation to the level of ovarian steroids. As such, it is hypothesised that SCN AVP neurons are a key input onto RP3V kisspeptin neurons and their activation underpins the downstream activation of kisspeptin neurons and the resulting preovulatory LH surge.
Using genetically-mediated anterograde viral tract tracing, the projection from SCN AVP neurons was traced in the Avp-cre mouse model. Avp-cre projections to the RP3V were found in the RP3V and were in close appositions to kisspeptin neurons present. There was a significant correlation between the amount of viral-expression in Avp-cre neurons in the SCN to the density of projections seen in the RP3V; there was no correlation between other AVP- expressing regions and the RP3V. This indicates that the Avp-cre projection to the RP3V is specifically from the SCN, rather than from other areas if the brain.
To determine whether the Avp-cre projections to RP3V kisspeptin neurons were communicating by GABAergic fast-synaptic transmission, a viral vector expressing a channelrhodopsin (ChR2) variant was injected into the SCN of female Avp-cre:Kiss1-hrGFP mice. Using short pulses of blue light, the ChR2-expressing projections were optogenetically stimulated while whole-cell electrophysiological recordings were made from identified kisspeptin neurons in the RP3V. Despite the Avp-cre appositions to RP3V kisspeptin neurons, very few received inhibitory postsynaptic currents from ChR2-expressing Avp-cre projections in response to optogenetic stimulation. There were no differences to this input across the oestrous cycle.
To determine the contribution of AVPergic to RP3V kisspeptin neurons, high-frequency light stimulation (HFLS) of Avp-cre projections was carried out whilst on-cell loose-patch recordings were made from identified RP3V kisspeptin neurons. There was a significant increase in kisspeptin firing rate following HLFS solely in proestrus, which was not seen in other oestrous cycle stages. This response was able to be blocked by the V1R antagonist Manning Compound, indicating it was mediated by AVP release. Surprisingly, a second response was found, solely in oestrus. Rather than a delayed, AVP-mediated excitation of kisspeptin neurons, an immediate inhibition in kisspeptin firing was noted. This response was not seen in other oestrous cycle stages. To determine the potential mechanism of this inhibition, the GABAAR antagonist, gabazine, was applied throughout HFLS. This inhibited the HFLS- induced decrease during oestrus, suggesting it was mediated by GABA release. Both the excitatory and inhibitory responses seen across the oestrous cycle correlate to time points where the LH surge should and should not be initiated respectively. This points towards a mechanism by which SCN AVP neurons can either trigger or inhibit the onset of the LH surge.
This work was followed by an examination of the SCN-to-RP3V circuit in a mouse model of polycystic ovary syndrome (PCOS), which does not naturally mount an LH surge. In the prenatally androgen-treated (PNA) mouse model, the Avp-cre projection to the RP3V was reduced; as was the proportion of close appositions made with kisspeptin neurons. Despite this, kisspeptin neurons from the PNA mouse were able to respond to AVP application.
Overall, this work has confirmed a projection from SCN AVP neurons to RP3V kisspeptin neurons, that comprises an indirect pathway leading to GnRH neuron activation. Activation of this circuit shows oestrous cycle-dependent changes that are consistent with increases in neural activity at the time of the LH surge. Finally, as this circuitry is critical for the timing of the LH surge, its reduction in a clinically relevant mouse model, may be an underlying cause of ovulatory dysfunction seen in PCOS.