The seminal chemical-biological role of octanoy-lation as the initial immediate critical chemical step in determining the bioactive receptor conformation of the ghrelin peptide for the GHS type 1a receptor is notable and intriguing. We are currently working with Shaoxing Wu to study the effect of octanoylation on the ghrelin conformation determined from NMR analysis in conjunction with ghrelin conformation as determined from theoretical molecular modelling. The long-term objective is to evaluate changes in ghrelin conformation following substitution of shorter and longer car-boxylic acids of the octanoyl group. It is hypothesised that at certain anatomical peripheral tissue sites, carboxylic acids other than octanoic acid are covalently linked to Ser3 of ghrelin, which produce ghrelin molecules with different conformations that act on the putative ghrelin receptor subtypes. Novel in this hypothesis is that ghrelin molecular subtypes are generated without changing the open reading frame of the gene for the ghrelin peptide. If this occurs, the effects on the physiology of the ghrelin system would be major. If this hypothesis is incorrect, it could be because the special octanoylation enzyme complex is so specific in action that it only links octanoic carboxylic acid to Ser3 of the full-length ghrelin peptide. So far this key important specific enzyme complex has neither been directly demonstrated to exist nor its regulation characterised. Unknown is whether the specificity of this enzyme may be only for the ghrelin peptide and only selective for the Ser3 amino acid residue, while the esterification reaction may be more permissive. Envisioned is the possibility that carboxylic acids of different chain lengths may be substituted at Ser3, which are tissue, metabolic and nutrition dependent.
Several findings have led to this possible broader, less restrictive 'carboxylic acid hypothesis'. These include the following points. (1) Octanoylation as the critical immediate determinant of the bioactive chemical conformation of the ghrelin peptide raises the possibility that different carboxylic acids added at Ser3 may occur and may induce different conformations of the ghrelin molecule that bind to select ghrelin subtype receptors. (2) Octanoylation of the ghrelin peptide has been directly established and validated chemically only for ghrelin synthesised in the stomach but not at other anatomical tissue and organ sites, where only mRNA or in situ hybridisation evidence of the ghrelin peptide exists and/or that immunohistochemistry evidence specifically detects full-length bioactive octanoylated ghrelin. Thus, at peripheral anatomical sites other than the stomach, the ghrelin peptide may be desoctanoyl ghrelin or may have a different carboxylic acid covalently linked to the Ser3 residue. (3) Putative ghrelin receptor subtypes have been proposed as a result of different biological profiles of select GHRPs. (4) Binding affinities for the ghrelin GHS receptor vary from high to low as a function of the carbon atom chain length of the carboxylic acid substituted for the octanoyl group of the ghrelin molecule. This may suggest the spectrum and number of ghrelin receptor types could be considerable since the receptor binding affinities were so variable for the type 1a ghrelin receptor when different carboxylic acids were substituted for the octanoyl group on the Ser3 residue of the full-length ghrelin peptide. It is possible to envision that these variable binding affinities may indicate different subtypes of ghrelin. The ghrelin molecule subtypes and ghrelin receptor subtypes may both exist in the putative peripheral ghrelin sys tem as well as the CNS system, but they may be regulated by the metabolic, nutritional and hormonal status. Additionally, another molecular form of ghrelin itself has been identified in the stomach, i.e., des[Gln14]-ghrelin in a 27 amino acid Ser3 octanoylated peptide, which is the result of alternative splicing of the ghrelin gene. It has the same bioactivity of the primary 28 amino acid octanoylated ghrelin.
Besides the GHRP/ghrelin GHS type 1a receptor, the Merck group cloned another related GHS receptor designated type 1b in 1997. The type 1b receptor is a truncated version of the type 1a GHS receptor because only TM-1 through TM-5 domains are encoded. Its function is still unknown. Neither the GHRPs nor ghrelin bind to this receptor and the type 1a and 1b receptors are localised to separate chromosomes. When selective, sensitive hybridisation probes for the type 1a and 1b receptors were utilised, the mRNA distribution in normal human tissues demonstrates the truncated type 1b receptor is widely distributed while the type 1a GHRP/ghrelin active receptor is much more restricted, i.e., predominantly in the pituitary gland but also in the thyroid gland, pancreas, spleen, myocardium and adrenal gland. In contrast to the distribution of the active type 1a receptor, the expression of the mRNA distribution of the ghrelin peptide is widespread in human tissues, which suggests that ghrelin may be acting on selected receptor subtypes in peripheral tissues . These results indicate the necessity of utilising subtype-specific probes in order to distinguish the widespread but inactive truncated receptor from the active GHS receptor, as well as the need to attempt to identify additional ghrelin receptor subtypes.
Tomassetto et al. identified the motilin-related peptide from the stomach in a separate study via a recombinant cDNA approach, in contrast to the GHS-type 1a receptor approach utilised by Kojima et al., which identified octanoylated ghrelin . Significantly, the motilin-related peptide is chemically but not biologically identical to desoctanoyl ghrelin, particularly since octanoylation of the 28 amino acid ghrelin peptide is essential for binding to the type 1a ghrelin receptor and its enhancement of GH release and food intake. Additionally, on the basis of chemical similarities between octanoylated ghrelin, motilin-related peptide and the duodenal hormone motilin, an important regulator of gastrointestinal motility, it is reasonable to project that octanoylated ghrelin also may have a motilin-like effect on gastrointestinal motility. Most informative and meaningful are the recent studies and detailed discussion of Inui et al. in support of the unique and potent physiological gastro-prokinetic effect of ghrelin and its relationship to the increase of food intake by ghrelin . A special aspect is the functional interde-pendency of the stomach and hypothalamus for the integration of feeding and associated auto-nomic, neuroendocrine and gastrointestinal functions emphasised by Inui. It becomes apparent that the route of administration or the anatomical origin of endogenous secretion of ghrelin is a basic issue that must be considered. Peripheral versus central intracerebroventricular (ICV) administration may produce differing effects on gastric acid secretion, motility or food intake. Inui et al. and others (vide infra) bring out the potential physiological functional significance of only a very small amount of 'ghrelin' synthesised in the hypothalamus, particularly the arcuate nucleus [20,21]. A basic point still in need of chemical validation is whether bioactive octanoylated ghrelin is produced in the hypothalamus because desoc-tanoyl ghrelin is non-functional in regard to these effects. Obviously, these are convoluted biological regulatory issues, suggesting that a coordinated approach will be required to resolve these complex issues, i.e., whether the bioactive ghrelin molecule is synthesised in the hypothalamus.
NMR studies of ghrelin and five of its truncated analogues in solution detail that all of them behave as random coils . NMR comparison of the N-terminal octanoylated C-terminal amidated pentapeptide ghrelin fragment, active in vitro, with the N-terminal portion of the full-length ghrelin further indicates that they are similar. An exception is the presence of two additional nuclear Overhauser effects (NOE) between the Phe4 NH proton and the protons of the beta-methylene of the Ser3 residue of this pentapeptide fragment. Circular dichroism spectroscopy of ghrelin pentapeptides in water also indicates a conforma tion as random coils. However, molecular modelling of GHRP-6 and ghrelin with the incorporation of these NMR results did not account for nor agree with the binding of these peptides to the ghrelin GHS type 1a receptor.
In 1998, Scott Feighner and his Merck collaborators published mutation studies on the human GHS 7 transmembrane (TM) domain G-protein coupled receptor. They found that mutation of amino acids on TM 2, 3 and 5, 6 both affect and activate the GHS receptor . This group developed a three-dimensional docking model of the Merck benzolactam and spiropiperidine non-pep-tide GHSs as well as GHRP-6. Mutating glutamic acid to glutamine at position 124 in TM 3 resulted in a non-functional receptor for each of the three different dissimilar chemical types of GHRP/GHSs. Since each GHRP has an essential positive-charged atom at the N-terminus, the non-functional receptor with the glutamic acid to glutamine mutation was explained by eliminating the counter ion interaction between these three GHSs and the receptor. Furthermore, the TM 2, 5 and 6 mutations induced different effects on the binding and activation of these three chemically different GHRP/GHSs. This led to the speculation by this group that these three GHSs probably bind to the same receptor site in different molecular orientations.
Was this article helpful?