Abstract
Gonadotropin-releasing hormone (GnRH) neurons are the central regulators of reproductive function in all mammals. The cell bodies of GnRH neurons are unique in that they are not confined to a discrete nucleus, but rather exist as a scattered continuum of cells throughout the basal forebrain. In rodent species, most of these GnRH neurons reside within three anatomical divisions of the brain: the medial septum (MS), the rostral preoptic area (rPOA) and the anterior hypothalamic area (AHA). Typically, these neurons are thought of as one large homogenous group, serving similar functions, however, recent anatomical and functional evidence suggests that this is not the case. One way to distinguish subsets of neurons is by their patterns of projection throughout the brain. Until now, the projection patterns of GnRH neurons have only ever been assessed for the population as a whole, due to the lack of ability to distinguish subdivisions from one another.
The recent development of novel transgenic tools has enabled us to visualize GnRH neurons and their projections in their anatomical subdivisions of the MS, rPOA or AHA for the first time. An adenovirus containing a transgene for enhanced, farnesylated green fluorescent protein (Ad-iZ/EGFPf) was injected intracranially at stereotaxic coordinates for the MS (n=4), rPOA (n=6) or AHA (n=4) into female transgenic GnRH-Cre mice. Using this approach, Ad-iZ/EGFPf was specifically targeted to GnRH neurons in each region of interest, in order to “fill” these cells to the most distal ends of their projections. The first aim of this project was to assess how accurately GnRH neurons in MS, rPOA and AHA could be specifically targeted. Injections filled between 5 and 20 cells in most animals with accurate injection sites. In animals that received MS injections, more GnRH neurons in the MS were filled than in the rPOA (P < 0.05) and AHA (P < 0.05). Similarly, animals that received rPOA injections had more filled cells in the rPOA compared with the MS (P < 0.0001) and AHA (P < 0.001). In animals injected in the AHA there was no significant difference in the number of filled cells in the AHA compared with the MS and rPOA. In wild-type controls (n=3) and animals that received off-target injections (n=3), no filled GnRH neurons or projections were present.
In the second aim of this project the distribution of projections from Ad-iZ/EGFPf filled GnRH neurons residing in the MS, rPOA and AHA were mapped. Across all animals, GnRH neuron fibre projections that were positive for GFP were found in 120 different regions of the brain, including nuclei, subnuclei and white matter tracts. These regions were found across several major brain divisions, in the hypothalamus, septum, thalamus, cerebral cortex, pallidum, striatum, amygdala, hippocampus and midbrain. The broad distribution of GnRH neuron projections highlights the diverse functions that these neurons are potentially influencing within the brain, as well as the power of the viral cell-filling approach used to visualize the full extent of these neurons. In the third aim, the projection patterns from GnRH neurons in the MS, rPOA and AHA were compared. Remarkably, 60% of the brain regions that contained fibre projections only did so from one or a combination of any two subpopulations of GnRH neurons, indicating that the projection patterns of these subdivisions is not homogenous. Notably, fibre projections in the vomeronasal amygdala originated exclusively from GnRH neurons in the AHA. Areas involved with olfactory processing likewise only received projections from MS and AHA GnRH neurons. Surprisingly, the largest division of GnRH neurons, the rPOA, had the most confined pattern of projection, but projected robustly to the median eminence suggesting a primarily hypophysiotropic role.
For the first time, it has been possible to interrogate the projection patterns of anatomical subdivisions of GnRH neurons, which has revealed that they are far from homogenous. Overall, these results provide strong support for the existence of GnRH neuron subpopulations, highlighting that these neurons should be treated as similar but separate entities. Identifying GnRH neural subpopulations and delineating their respective roles could have wide applications, from increasing reproductive success in livestock, to teasing apart the ongoing mysteries surrounding infertility in humans.