Recent in situ hybridization studies have provided information on the specific distribution of GHR gene expression. In most species so far investigated, the main neural sites of expression of GHR in the CNS of adult animals are in the hypothalamus and hippocampus with the majority of studies carried out in the rat (Fig. 1). Within the hypothalamus, there is increasing evidence that GHRs are involved in a short loop feedback regulating GH secretion. Initial studies from Burton et al. (18) localized GHR expression to periventricular nucleus (PeN) somatostatin (SRIF) neurons, consistent with a feedback loop increasing SRIF expression and release in the face of high GH expression. GHR transcripts were also found in the arcuate nucleus (ARC) consistent with the idea that GH feedback inhibits growth hormone releasing hormone (GHRH) expression. Although ARC GHR expression is present in a small number of GHRH-containing neurons (19), the majority of ARC GHR-expressing cells also express neuropeptide-Y (NPY) (20). Thus, GH feedback inhibition of GHRH may be indirect, perhaps via changes in activity of NPY cells (see below). Other hypothalamic structures, such as paraventricular nucleus (PVN) also express GHRs, but their function here is less obviously related to the direct neuroendocrine control of the GH axis. Similar considerations apply to extrahypothalamic sites of GHR expression, prominent among which is the hippocampus (Fig. 1). The function of GHR expression in this region is not so clearly understood, but is consistent with a role for GH in consolidation of memory (see below).
Although in situ hybridization is a powerful technique, it is important to recall its limitations. It describes mRNA located in the site of transcription—there is no guarantee that the RNA will always be translated into functional protein, when it will be translated, or where such GHR protein itself is finally located. As a plasma membrane receptor, GHR may be transported considerable distances from the sites of synthesis either to dendrites or to axon terminals. Conversely, its final target may not be the cell surface; some GHR protein may be targeted intracellularly or even to the nuclei of the cells in which it is expressed, perhaps to signal there (21). For these reasons it will be interesting to determine the intracellular location of GHR proteins in specific regions of the CNS at a higher resolution and to evaluate the physiological significance of the observation that, in addition to neuronal cells, some GHRs may be expressed in glial, microvascular, or other nonneural elements in the CNS (22).
A second important problem is the heterogeneity of GHR transcripts in these tissues. Almost all in situ hybridization studies have been carried out with probes directed against exons for the coding region of GHRs. However, it has become clear that there are a variety of upstream 5' untranslated regions (UTRs) contributing alternative first exons to both GHR and GHBP transcripts in all species that have been examined. In the liver, these 5' UTRs are regulated and are much more closely correlated with changes in GHR or GHBP protein than probes against the common coding exons that do not distinguish these 5' UTRs (23,24). Little is yet known about the 5' UTRs in brain GHR transcripts other than the fact that the liver-specific rat GHR1 exon (V2 in the terminology of Domene et al.) is not detectable in brain (25), whereas the 5' UTR identified as V4 by Domene et al. (25) is prominently expressed in brain tissue. Mapping the relative distributions of transcripts containing these alternate 5' UTRs in the CNS may give further insights into the regulation of GHR expression in the brain because they may underlie the differential effects of peripheral steroid hormones on brain vs hepatic GHR expression, or the opposite changes in GHR expression in brain and liver seen from fetal life to adulthood.
Some information has recently been reported for the genomic organization of these upstream UTRs in the human (26). Zou et al. (26) have found that V1,V4,V7, and V8 UTRs are all located in close proximity, upstream of the hGHR gene and that there may be other UTRs in addition. Practical benefits could flow from further studies of these 5' UTRs in humans. For instance, specific analyses for mutations in the coding exons of the human GHR have identified possible causes of partial GH insensitivity in some short individuals (27). A similar approach, but one that looks for mutations in the upstream UTR exons, may prove equally fruitful to identify deletions or mutations in individual 5' UTRs that could theoretically give rise to tissue-specific GH insensitivity. There is very little information on the genomic arrangements of the 5' UTRs in other species to date.
Was this article helpful?