Identification and distribution of neuroendocrine gonadotropin-releasing hormone neurons in the ewe

The final common pathway controlling reproductive function in vertebrates is the GnRH neuron and its projection to the median eminence (ME), site of peptide release into the pituitary portal system. GnRH neurons are widely distributed; therefore we sought to test the hypothesis that those projecting to the ME are located in specific regions. We used as a model the sheep, a species in which a great deal of information regarding the physiology of GnRH secretion is known. To identify cells projecting to the ME (i.e., neuroendocrine neurons), ewes (n = 10) received injections into the ME of neuronal tract-tracing compounds: cholera toxin-beta subunit (CT-beta) or one of two fluorescent compounds (rhodamine isothiocyanate or fluorescein-conjugated dextran). Forty-eight h later, animals were perfused intracranially and their brains were processed for immunocytochemical localization of GnRH and CT-beta using a dual-immunofluorescent procedure or by single-label immunofluorescent visualization of GnRH combined with direct visualization of fluorescent tracers. Small, well-circumscribed injections into the ME were made successfully in 6 of 10 animals, and these overlapped the location of GnRH terminals and fibers. Neuroendocrine GnRH neurons (those GnRH neurons containing retrogradely transported tracer) were identified throughout their previously reported range: within the diagonal band of the Broca/medial septal region, medial preoptic area (MPOA), anterior hypothalamic area, and medial basal hypothalamus. Although the absolute number of neuroendocrine GnRH neurons varied by region, the percentage of the total GnRH population within each of these areas that was retrogradely labeled did not differ (p > 0.05). Injections placed unilaterally within the ME labeled a similar proportion of GnRH cells both ipsilateral and contralateral to the injection site in all areas except the MPOA, where ipsilaterally labeled cells were approximately twice as numerous as those labeled contralaterally. Injections that missed the ME and were placed either into the third ventricle or into the arcuate nucleus labeled only 0.5% and 4-11% of GnRH neurons, respectively. These results do not support the hypothesis that in the ewe, GnRH neurons projecting to the ME are localized to specific regions. Thus, we postulate that GnRH release into the hypophyseal portal system reflects the output of GnRH neurons located in multiple areas.