Recent studies suggest that the endogenous opioid peptide, dynorphin, is an important mediator of progesterone negative feedback on GnRH pulse frequency in the ewe. These experiments tested this hypothesis by examining the effects of progesterone on dynorphin A concentrations in cerebrospinal fluid (CSF) collected from the third ventricle and expression of preprodynorphin (PPD) mRNA in hypothalamic nuclei. CSF was collected every 10 min for 5 h in three groups of ewes: 1) ovary-intact ewes during the luteal phase (d 6-7 of estrous cycle); 2) ewes 6-7 d after ovariectomy (OVX); and 3) OVX ewes treated for 6-7 d with implants that produced luteal-phase progesterone levels. Diencephalic tissue from these ewes was then collected and processed for in situ hybridization using an ovine cDNA probe against PPD. Progesterone treatment increased dynorphin A concentrations in CSF over that observed in untreated OVX ewes; CSF dynorphin A concentrations in ovary-intact ewes were midway between the other groups. OVX significantly decreased the number of PPD mRNA-expressing cells in the preoptic area (POA), anterior hypothalamic area (AHA), and arcuate nucleus (ARC), with no change seen in any other PPD-expressing nuclei. Progesterone treatment of OVX ewes restored PPD expression in the POA and AHA to levels seen in luteal-phase animals but had no effect on PPD expression in the ARC. These results are consistent with the hypothesis that progesterone acts via dynorphin neurons to inhibit pulsatile GnRH secretion and point to dynorphin neurons in the POA, AHA, and ARC as potential mediators of this action during the luteal phase.

An increase in the response of GnRH neurons to estrogen negative feedback is responsible for seasonal anestrus in the ewe, but the underlying neural mechanisms remain largely unknown. Neural plasticity may play an important role because the density of synaptic input to GnRH neurons changes with seasons. Moreover, the transition from breeding to anestrous season requires thyroid hormones, which are also required for neuronal development. In the first experiment, we examined whether the decrease in synapses on GnRH neurons is critical for the transition to anestrus by comparing synaptic input in thyroidectomized and thyroid-intact controls, using electron microscopic analysis. Thyroidectomized ewes remained in the breeding season, but the number of synaptic contacts on their GnRH cells was not different from those in thyroid-intact ewes that were anestrus. The next experiment tested whether there was a seasonal change in morphology of the A15 dopaminergic neurons that mediate estrogen negative feedback during anestrus by analyzing synapsin-positive close contacts onto A15 neurons with confocal microscopy. There was a 2-fold increase in these close contacts onto dendrites of A15 neurons in anestrus and a corresponding increase in the length of A15 dendrites at this time of year. The increase in dendritic length was blocked by thyroidectomy, but this procedure did not significantly affect synaptic input to A15 neurons. These results provide initial evidence that the seasonal change in synapses on GnRH neurons is not sufficient for the transition into anestrus but that plasticity of the A15 dopaminergic neurons mediating estrogen negative feedback may contribute to this seasonal alteration.