The Ongoing Saga of Obestatin: Is It a Hormone?

Abstract

Posttranslational processing of preproghrelin encoded by the ghrelin (GHRL) gene is predicted to yield two peptides, ghrelin and obestatin, which have been proposed to have opposite actions on feeding behavior (1, 2). Although ghrelin increases food intake (1), obestatin has been shown to suppress it (2). This commentary focuses on the recent controversy about the role of obestatin as a hormone. The GHRL gene is predominantly expressed in the stomach in X/A-like cells (specialized oxyntic gland cells) but also at low levels in the duodenum, and the rest of the bowel, pancreas, pituitary, hypothalamus, kidney, and placenta (1, 3). The GHRL gene encodes a polypeptide containing 117 residues, called preproghrelin, which undergoes stepwise processing, including proteolysis and acylation, to form ghrelin (Fig. 1) (1, 4, 5). Based on bioinformatics, Zhang et al. (2) predicted that preproghrelin can also undergo additional processing at two potential proteolytic sites to form obestatin (Fig. 1). Purification of the immunoreactive obestatin from the rat stomach extract revealed the sequence to be consistent with the obestatin sequence deduced from rat preproghrelin cDNA. The investigators reported that synthetic obestatin, when injected ip (12.5–1000 nmol/kg) or intracerebroventricularly (8 nmol/kg), acutely suppressed food intake over 5 h in adult male mice in a dose-dependent fashion. Obestatin reduced gastric emptying and inhibited jejunal contractions acutely. In adult male rats, obestatin suppressed body weight gain over a 7-d period. The hormone was reported to bind to G protein-coupled receptor-39 (GPR39) (2). As often happens in highly competitive areas of science, many original findings have been challenged. First, three groups were unable to confirm the binding of obestatin to GPR39 (6–8). Second, the original investigators themselves could not reproduce the binding of obestatin to tissue homogenates or recombinant GPR39 (9). Thus, the receptor involved in action of obestatin remains unknown. Third, obestatin lacks specific binding at the blood-brain barrier and undergoes rapid degradation (10). Furthermore, many other laboratories failed to observe any significant effects of obestatin on food intake in animal models. In doses ranging from 12–1000 nmol/kg body weight ip or approximately 1.2–12 nmol/kg intracerebroventricularly, obestatin did not reduce food intake in adult mice or rats over 1–24 h or over 7–9 d (6, 11–17). In defense of their original findings, Zhang et al. (9) reported that obestatin, 750 nmol/kg ip, reduced food intake in mice when animals had access to food at precisely 15 min after injection. Recently, Lagaud et al. (18) also observed a mild, but significant reduction in food intake with 10–300 nmol/kg obestatin given ip to mice and rats; however, higher doses were ineffective. These authors also reported reduction in food intake and weight gain over a 7-d period with three times daily obestatin injections in mice (18). Nonetheless, in the hands of most of the investigators, obestatin has shown no effects on body weight. Obestatin does not affect secretion of any pituitary hormone or corticosterone in rat (17, 19); other investigators (20) have found little or no effects of obestatin on gastric or jejunal motility. Whether obestatin reduces thirst (16), promotes sleep (21), or affects memory and anxiety (22) remains to be confirmed. Serum obestatin levels did not change upon fasting or feeding in rats (2); but in mice, a slight reduction was seen upon fasting (14). In humans also no postprandial changes in serum obestatin levels have been noted (23). For a hormone proposed to affect food intake acutely, this lack of change in circulating concentrations postprandially is surprising. The intriguing feature of the preproghrelin protein is its proteolytic cleavage, such that during one processing step ghrelin is released, and with another, obestatin is released. Studies of animals lacking various prohormone convertases (PCs) suggest that PC1 is involved in processing of proghrelin to ghrelin as well as obestatin, although direct processing of proghrelin to obestatin was not evaluated (4). A recent study suggests that PC7 and furin may also be involved in ghrelin processing (5). These proteases recognize di-basic residues K-R or R-R and cleave the peptide bond carboxy terminal to the dibasic pair (24). They may also recognize monobasic residue cleavage sites, but the rules of engagement may be different or inefficient. Although the release of ghrelin after the initial removal of signal peptide involves a single cleavage, the release of obestatin requires cleavage at both the aminoand carboxy-terminus. The efficiency with which these two cleavages might occur is not clear because the proposed cleavage sites lack the more efficient dibasic Abbreviations: GHRL, Ghrelin (gene); GPR39, G protein-coupled receptor-39; PC, prohormone convertase.