Relationship of glia to GnRH axonal outgrowth from third ventricular grafts in hpg hosts

Abstract

The homozygous mutant hypogonadal (hpg) mouse lacks a functional gene for the neuropeptide gonadotropin releasing hormone (GnRH). The consequence of this defect is an infantile reproductive tract in adulthood. This condition can be reversed by the implatation of normal fetal preoptic area tissue that contains GnRH neurons. Reversal is always preceded by the outgrowth of GnRH axons into the host target tissue, the median eminence, by a stereotyped pathway. In the current experiments we investigated the cellular nature of the path taken by early emerging GnRH axons focusing on their relationship with astrocytic components and with the specialized ependymal population of this area, the tanycytes. In control tissue glial fibrillary acid protein (GFAP) immunoreactivity was confined to the exterior of cerebral blood vessels and glial limitans. Both GFAP and vimentin, another intermediate filament protein, marked the specialized ependymal cells of this region, the tanycytes. There was a robust reactive astrocytic response to the injury of transplantation in both the donor and host tissue within 5 days of implantation and the reactive astrocytes persisted for 60 days. These cells were GFAP-positive and were present in many areas of the host along the cannula tract and not confined to the area of GnRH axonal outgrowth. Vimentin, another intermediate filament, marked only the specialized ependymal cells of this region, the tanycytes, in both control and grafted tissue. Despite the profound reactive gliosis, GnRH axons were shown to exit the implant as early as 5 days after grafting suggesting that the gliotic process did not constitute a barrier to this phenomenon. At the light microscopic level, double label immunocytochemical studies did not reveal any specific association between GFAP or vimentin-positive cellular processes and these pioneer GnRH fibers. However, since normal GnRH axons had been reported to travel in tanycytic channels through the medial basal hypothalamus we reinvestigated the pattern of early emerging GnRH axons at the ultrastructural level. With this higher resolution, GnRH axons were found adjacent to glial elements along their entire traverse from the graft-host interface, through the host basal hypothalamus to their termination on the hypophysial portal capillaries. At the interface, GnRH-positive axons appeared to exit via glial channels similar to those described in other developing and regenerating systems. In the host, GnRH immunoreactive axonal profiles were surrounded by glial processes though the latter could not be further defined as tanycytic or astroglial. Other, immunonegative, axons were frequently seen in axonal bundles or fascicles and not necessarily in contact with glia. GnRH profiles at the interface or in the basal hypothalamus had a rounded or oval appearance. It was only when such axons reached the portal capillaries that they took on the appearance of more complex growth cones. The dilated portion of the axon was surrounded by tanycytic endfeet and did not make contact with the perivascular space. This same arrangement has been seen previously for normal GnRH axons. Although glial processes seem to provide a permissive substrate for GnRH axonal extension they cannot fully explain the precise targeting of outgrowth since both reactive glia and tanycytic processes are seen in regions where GnRH axons do not exit. Further studies are necessary to determine whether chemotropic factors specific to the region of the median eminence underlie this accurate navigation of the growing axon.