In comparison to the embryo, very little work has been carried out on the expression and role of Hox genes in the adult animal. An expression profile of all 39 vertebrate Hox genes on a select panel of adult human tissues reveals that in fact these genes are widely expressed throughout the adult human and a colinear pattern of expression is displayed similar to that of the developing embryo. Of particular interest is the abundance of Hox genes that are expressed within the adult adrenal gland. Adrenal cortical cells are continuously renewed to sustain production of zonal steroids. Cell proliferation occurs at the periphery of the cortex and cells are then displaced centripetally, phenotypically switching as they migrate through the gland before undergoing apoptosis at the zona reticularis/medullary boundary. It is still unclear which mechanisms cause the cells to differentiate as they cross the zonal boundaries and we hypothesise that Hox genes may be involved in the phenotypic switching of the adrenocortical cells. In situ hybridisation experiments were carried out on adult rat adrenal gland sections and Hox gene expression was localized within the zonal borders, coinciding with the localization of cells that undergo phenotypic differentiation, and thus supporting our hypothesis that Hox genes may be involved in the phenotypic switching of the adrenocortical cells. As in the developing embryo, the genes display colinear expression with the 3' Hox genes being expressed within the outer gland and the 5' genes within the inner zones.

Immunological and cellular stress signals trigger the release of corticotropin-releasing hormone (CRH) from the spleen, thymus and inflamed tissue. In vivo and in vitro studies generally suggest that peripheral, immune CRH has pro-inflammatory effects and acts in a paracrine manner by binding to CRH-R1 and CRH-R2 receptors on neighboring immune cells. However, it now seems likely that some of the suggested pro-inflammatory actions of CRH may be attributed to novel CRH-like peptides or to the related peptide, urocortin, which is also present in immune cells and has especially high affinity for CRH-R2 receptors.

The expression and activities of corticotrophin-releasing factor (CRF), urocortin (UCN), the CRF-binding protein (CRF-BP) and CRF receptors in rat brain have been well documented; however, information regarding their peripheral distributions remains incomplete. Given the multiple immunomodulatory effects of peripherally administered CRF and UCN and the high levels of CRF receptor type 2 (CRF-R2) mRNA and protein expressed in the heart, the lymphoid organs and heart have become targets for some of the latest CRF-related research. Here we demonstrate the presence of UCN mRNA in both the rat spleen and human Jurkat T-lymphoma cells using 3'-RACE (rapid amplication of cDNA ends) PCR. Following on from these initial results, we used semi-quantitative RT-PCR to carry out a comprehensive study assessing the relative amounts of CRF, UCN, CRF-R1, CRF-R2 and CRF-BP mRNAs in the brain, thymus, spleen and heart of normal, untreated rats. The rank orders of mRNA abundance in each of the tissue types were as follows: for CRF, brain>>thymus=spleen=heart; for UCN, heart>/=brain>thymus>spleen; for CRFR1, brain>>thymus>spleen (absent in heart); for CRF-R2, brain=heart>thymus>spleen; and CRF-BP was only detectable in the brain. We have provided evidence for the existence of CRF, UCN, CRF-R1 and CRF-R2 expression in resting immune cells, with UCN expression being particularly predominant in the rat thymus and human Jurkat cells. Additionally, the high levels of UCN mRNA detected in heart corresponded to the high expression of CRF-R2 mRNA, suggesting an important role for UCN/CRF-R2 coupling in this tissue.

To purify novel ligands for the corticotrophin-releasing factor binding protein (CRF-BP) from ovine brain, whole brain was homogenised in methanol and the supernatant extracted on Sep-pak C18 cartridges followed by a preliminary HPLC step. Three peaks of ovine CRF-BP ligand activity were detected in the HPLC fractions, the first two of which were also detected by a specific corticotrophin-releasing factor two-site immunoradiometric assay, the third peak being detected by a human CRF-BP ligand assay, which will not detect ovine CRF. Human CRF-BP ligand-containing fractions were further purified by affinity chromatography on a human recombinant CRF-BP column with two additional HPLC steps. The human CRF-BP ligand was found to: (a) possess a molecular mass of 4707 Daltons, (b) have an N-terminal amino acid sequence (5 residues) identical to rat urocortin, (c) be detected by a specific urocortin radioimmunoassay, (d) have high affinity for both the human and ovine CRF-BPs and (e) be present in many regions of the ovine brain. Additionally, a 300 bp cDNA fragment sharing 83% homology with the rat urocortin gene was cloned from ovine brain, the product of which was predicted to have an identical amino acid sequence to that of rat urocortin. These pieces of information confirmed the identity of the human CRF-BP ligand as an ovine urocortin. The specially developed CRF-BP ligand assays showed that the rank orders of affinity of the CRF family members for human CRF-BP were: carp urotensin-1>>human CRF=rat/ovine urocortin>human urocortin>>frog sauvagine>>ovine CRF, and those for the ovine CRF-BP were: carp urotensin-1> human CRF=rat/ovine urocortin>human urocortin> frog sauvagine>>ovine CRF. This study describes a successful technique for the purification and detection of peptide ligands for the CRF-BP. We conclude that urocortin is the principal ligand for the CRF-BP in ovine brain and we could find no evidence for a centrally located mammalian sauvagine-like peptide.

In view of the reported inflammatory effects of corticotrophin-releasing factor (CRF) and the associated regulatory elements in the gene of its binding protein (BP), we postulate that both BP as well as novel BP ligands other than CRF may be involved in inflammatory disease. We have investigated BP in the blood of patients with arthritis and septicaemia, and have attempted to identify CRF and other BP-ligands in synovial fluid. The BP was found to be significantly elevated in the blood of patients with rheumatoid arthritis and septicaemia. There was less BP-ligand and CRF in synovial fluid from patients with rheumatoid arthritis than from those with osteo- or psoriatic arthritis. There was at least 10-fold more BP-ligand than CRF in the fluid of all three groups of patients. A small amount of immunoreactive human (h)CRF, eluting in the expected position of CRF-41, was detected after high-pressure liquid chromatography of arthritic synovial fluid; however, the bulk of material with BP-ligand binding activity eluted earlier, suggesting that synovial fluid contained novel peptides that interacted with the BP. These results would suggest that the BP and its ligands could play an endocrine immunomodulatory role in inflammatory disease.

As the association of corticotrophin-releasing factor (CRF) with its binding protein (BP) to form a dimer complex (CRF2/BP2) appears to be dependent on the nature of the ligand we have compared the circular dichroism difference spectra after association of the BP with ovine (o) CRF, human (h) CRF and the alpha-helical CRF (9-41) antagonist. All three ligands caused a negative change in molar ellipticity above 210 nm, with oCRF having the least and hCRF the greatest effect. Below 210 nm there was a marked divergence of difference spectra, with the reaction with the natural peptides, hCRF and oCRF, resulting in a positive change in ellipticity, whilst that with the antagonist produced a negative change. In view of the BP spectrum indicating predominantly beta-sheet and the peptides showing mainly alpha-helix these results were interpreted as the changes above 210 nm being due to dimerization and below 210 nm to a change in the conformation of ligand on binding. The opposite change in alpha-helicity of the antagonist observed on binding compared with the two natural CRF peptides could have fundamental pharmacological implications.