Ghrelin: From GH control to feeding behaviour and sleep regulatio

Ghrelin, a 28-amino acid gastric peptide with a n-octanoylation on Ser 3, has recently been identified as an endogenous ligand of the growth hormone secretagogue receptor (GHS-R). In parallel a cDNA encoding a protein which shares sequence similarities with prepromotilin was isolated from a mouse stomach library and its putative product was named prepromotilin-related peptide (ppMTLRP). Mouse and rat ppMTLRP sequences are identical and show 89% identity with human ghrelin. By analogy with promotilin, cleavage of proMTLRP into an 18-amino acid endogenous processed peptide was assumed on the basis of a conserved dibasic motif in position 19-20 of ghrelin sequence. In freely moving animals, both rat and human ghrelin stimulated GH release and human ghrelin 18 was ineffective, whereas in vitro, on superfused pituitaries, the three peptides stimulated GH release to the same extent. Activation of somatostatin (SRIF) release by ether stress did not modify GH response to ghrelin in vivo and ghrelin blunted 25 mM K+-induced-SRIF release from perifused hypothalami in vitro. This finding suggests that ghrelin action on GH secretion is mediated, at least partially, through functional antagonism of somatostatin. Ghrelin is not only effective on GH secretion but also increases gastrointestinal motility and displays orexigenic effects. This multiplicity of effects is in keeping with the pattern of GHS-R expression on hypothalamic Neuropeptide Y, SRIF and GHRH neurons. However, the relationships between endogenous ghrelin and GH secretions and feeding and sleep patterns remained to be investigated. In freely moving adult male rats, repeated administration of ghrelin at 3 to 4-h intervals (one during light-on and two during light-off periods) immediately increased GH release feeding activity and wakefulness while it decreased sleep duration. However, on the total duration of the sampling period (nine hours), only the stimulation of GH secretion and the inhibition of REM sleep were maintained. Endogenous plasma ghrelin levels exhibited pulsatile variations of smaller amplitude and regularity as compared to those of GH. The circulating levels of the two hormones were not strictly correlated, although mean interpeak intervals and pulse frequencies were close. In contrast, ghrelin pulse variations were correlated with food intake episodes in the light-off period and plasma ghrelin concentrations decreased by 26% in the 20 minutes following the end of the food intake periods. A positive correlation between ghrelin levels and active wake was found during the first three hours of the light-off period only. Since ghrelin secretion appeared to be directly related to feeding behaviour, we assessed its plasma levels in anorexia nervosa patients before and after renutrition in comparison with constitutionally thin subjects. Morning fasting plasma ghrelin was doubled in anorexia nervosa patients as compared to age-matched, constitutionally thin women, and it returned to normal values in patients after renutrition. In summary, ghrelin, which was discovered as a GH-secreting factor from the stomach, appears to also be related to the control of feeding behaviour. Whether ghrelin per se or through its indirect effects on feeding (and all attendant metabolic sequelae) acts on sleep remains to be determined.