Gut hormones in the control of appetite


Corresponding author S. R. Bloom, Department of Metabolic Medicine, Imperial College Faculty of Medicine, Hammersmith Campus, Du Cane Road, London W12 ONN, UK. Email:


Obesity is one of the greatest threats to the health of the developed world. In order to design effective drugs to treat the alarming increase in obesity, it is essential to understand the physiology of normal appetite control and the pathophysiology of obesity. The hypothalamus interprets and integrates neural and humoral inputs to provide a coordinated feeding and energy expenditure response. Recent evidence suggests that certain gut hormones – ghrelin, polypeptide YY, pancreatic polypeptide, glucagon-like-peptide 1 and oxyntomodulin – have a physiological role in governing satiety via the hypothalamus. Gut hormone appetite-regulatory systems represent a potential target for the design of antiobesity drugs.

The obesity epidemic

Obesity is one of the greatest threats to the health of the developed world. Levels of obesity, clinically defined as a body mass index of 30 kg m−2 or more, are increasing at an alarming rate. The prevalence of adult obesity in England and Wales increased from 6% of men and 8% of women in 1980 to 17% of men and 21% of women in 1998 (Bourn, 2001). Similar increases are evident throughout the developed world and are beginning to be mirrored in the developing world. Such an epidemic has more than mere aesthetic consequences. Obesity kills. Obese individuals show a marked increase in mortality, attributable to increased risk of type 2 diabetes mellitus, cardiovascular disease and certain forms of cancer (Visscher & Seidell, 2001; Bray, 2003). In the UK, obesity is now responsible for more than 1000 premature deaths per week, and swallows at least half a billion pounds of NHS funds per year (Bourn, 2001).

Even modest weight loss can significantly reduce the morbidity and mortality associated with diabetes and cardiovascular conditions. However, public and private health advice to reduce food intake and increase exercise has not prevented the obesity epidemic from accelerating. Gastrointestinal surgery is the only treatment shown to achieve long-term weight loss (Frandsen et al. 1998), but practical constraints and the significant associated morbidity limit its use to extreme cases. The search is therefore underway throughout industry and academia for possible pharmacological treatments (Clapham et al. 2001; Halford et al. 2004).

In order to design effective drugs to treat obesity it is essential to understand the physiology of normal appetite control and the pathophysiology of obesity. The last decade has seen great advances in our understanding of the central and peripheral mechanisms involved in the regulation of energy balance. This review will examine these mechanisms and discuss possible targets for pharmaceutical intervention.

The regulation of body weight

The regulation of body weight is homeostatic. On an individual, long-term basis, energy balance is remarkably precise, particularly when considering the large day-to-day variation in food intake and energy expenditure. However, there does appear to be a bias in this system to favour the conservation and storage of energy. While this bias makes sense from an evolutionary perspective – it being safer to carry unneeded energy stores than to lack required energy in a time of need – it also explains the rise in obesity seen in most modern environments which promote low levels of physical activity and possess a surfeit of calorie-dense foodstuffs (Wilding, 2002).

Central mechanisms The hypothalamus is the key central nervous system (CNS) region involved in appetite regulation, though other brain regions, including the nucleus tractus solitarius and the area postrema, also play a role. Early lesioning and stimulation experiments led to a ‘dual centre’ hypothesis which proposed that the ventromedial hypothalamic nucleus was a ‘satiety centre’ and the lateral hypothalamus a ‘feeding centre’ (Vettor et al. 2002). Over time this simple picture has changed into a more sophisticated model in which a number of discrete neuronal pathways within specific hypothalamic nuclei have been shown to form a more complex, integrated neural network. Numerous neurotransmitters and modulators are thought to be involved in the hypothalamic regulation of appetite (Kalra et al. 1999).

The hypothalamus interprets and integrates a number of neural and humoral inputs to coordinate feeding and energy expenditure in response to conditions of altered energy balance. Long-term signals communicating information regarding the body's energy stores, endocrine status and general health are predominantly humoral. Short-term signals, including gut hormones and neural signals from higher brain centres and the gut, regulate meal initiation and termination. Both short- and long-term signals can also affect energy expenditure via sympathetic nervous outflow to brown adipose tissue and by effecting the secretion of various pituitary hormones (Schwartz et al. 1999).

The hypothalamic arcuate nucleus (ARC), which is known as the infundibular nucleus in man, appears to play a crucial role in receiving and integrating such signals (see Fig. 1). The ARC is situated at the base of the hypothalamus and is incompletely isolated from the general circulation by the blood–brain barrier, allowing direct access of circulating factors to ARC neurones. There are two main populations of ARC neurones involved in the regulation of food intake: appetite-inhibiting pro-opiomelanocortin (POMC) neurones, and appetite-stimulating neuropeptide Y (NPY) and agouti-related peptide (AgRP) coexpressing neurones (Williams et al. 2001). POMC is a precursor molecule that gives rise to the anorectic peptide α-melanocyte-stimulating hormone (α-MSH), an agonist at the melanocortin 3 (MC3) and melanocortin 4 (MC4) receptors (Ellacott & Cone, 2004). Central administration of α-MSH inhibits feeding (Rossi et al. 1998). Transgenic animals and humans lacking the POMC gene or with MC4 receptor mutations are hyperphagic and obese (Yang & Harmon, 2003). AgRP is the endogenous antagonist of the MC3 and MC4 receptors, and central injection of AgRP dramatically increases food intake (Rossi et al. 1998). Neuropeptide Y is also a potent orexigenic agent, though its effects are shorter lived than AgRP. It stimulates feeding, possibly via the Y1 and Y5 receptors, though it also has effects on other endocrine systems (Gehlert, 1999).

Figure 1.

Circulating hormones influencing energy homeostasis via the arcuate nucleus
Continuous lines indicate stimulatory effects and dashed lines indicate inhibitory effects. For explanation of abbreviations, see text. AgRP, agouti-related peptide; GLP-1, glucagon-like-peptide 1; αMSU, alpha-melanocyte-stimulating hormone; NPY, neuropeptide Y, Oxm, oxyntomodulin; POMC, pro-opiomelanocortin; PP, pancreatic polypeptide; PYY, peptide YY.

Both the POMC and NPY/AgRP neuronal populations project to other hypothalamic nuclei, in particular the paraventricular nucleus (PVN), which is known to be critical in the regulation of food intake and energy expenditure. The PVN also assimilates inputs from other hypothalamic nuclei, including the lateral hypothalamic area, the ventromedial nucleus and the dorsomedial nucleus.

Adipostat factors

Leptin In 1953 Kennedy proposed that body weight was maintained by the regulation of body fat content (Kennedy, 1953). His adipostat mechanism anticipated the presence of an unknown circulating factor that provides the hypothalamus with information on the extent of body fat stores. Though several adipostat factors were proposed in the intervening years, the discovery of leptin in 1994 revolutionized the field (Zhang et al. 1994). In accordance with its putative adipostat role, leptin is expressed and secreted by adipocytes in white adipose tissue and circulates in plasma at concentrations proportional to fat mass, with a relatively long half-life. Peripheral or CNS administration of leptin to rodents reduces food intake and body weight and increases energy expenditure (Friedman & Halaas, 1998). Much lower doses are required with CNS administration, and peripheral leptin administration activates hypothalamic neurones expressing the leptin receptor, suggesting that these effects are mediated via the hypothalamus. Endogenous, peripherally secreted leptin enters the CNS either by active uptake or simple diffusion in areas outside the blood–brain barrier. Leptin directly inhibits orexigenic ARC NPY/AgRP neurones and stimulates anorectic ARC POMC neurones (Sahu, 2003). Leptin therefore acts as the afferent limb of a body fat regulation feedback loop.

The hyperphagic and obese ob/ob mouse lacks functional leptin (Zhang et al. 1994). Children with congenital leptin deficiency have a voracious appetite and also become obese. Administration of exogenous leptin ameliorates these abnormalities in mice and men (Pelleymounter et al. 1995; Farooqi et al. 2002). However, the vast majority of obese humans have normally functioning ob genes and high plasma leptin levels reflecting their high fat mass, suggesting leptin resistance in obese individuals (Considine et al. 1996). The mechanism for this is unknown, though it may involve reduced passage and capillary transport of leptin into the CNS, reduced leptin receptor expression and/or suppressed intracellular signalling. These factors may be responsible for the limited efficacy of leptin as an antiobesity drug in human trials to date (Mantzoros & Flier, 2000). While the absence of circulating leptin, communicating low or nonexistent body fat stores, has profound effects on appetite, body weight and fertility, raised leptin levels have much less dramatic results. Leptin may therefore play an important role during periods of starvation, but be less significant when food is freely available.

Insulin Insulin may play a similar role to leptin as an adiposity signal. Although insulin is not released from adipocytes themselves, basal circulating levels of insulin correlate with body adiposity level (Polonsky et al. 1988; Woods & Seeley, 1998) and disruptions of insulin sensitivity are associated with both obesity and diabetes (Kahn & Flier, 2000). Besides its well-characterized peripheral effects, insulin is thought to inhibit NPY/AgRP neurones in the ARC (Kalra et al. 1999). Intracerebroventricular (i.c.v.) administration of insulin reduced food intake and body weight in rodents (Air et al. 2002) and CNS administration of antibodies specific to insulin increased food intake and body weight (McGowan et al. 1992). Insulin-deficient animals are hyperphagic. Though this effect may be secondary to the absence of the peripheral effects of insulin, it is interesting to note that CNS insulin replacement to type 1 diabetic rats alleviates their hyperphagia, suggesting that insulin may have a direct anorectic effect (Sipols et al. 1995).

Gut hormones

The increase in hunger before a meal and the reduction following it cannot be explained by the small changes in plasma leptin. Similarly, the acute postprandial rise in insulin does not appear to reduce appetite directly (Chapman et al. 1998). There is evidence that the hypothalamus may sense nutrients and adjust food intake accordingly (Levin et al. 1999; Obici et al. 2002), but mimicking the postprandial rise in circulating nutrients by exogenous infusion has little effect on appetite (Opara et al. 1996; Stratton & Elia, 1999).

Postprandial satiety might be explained by a gut sensor system, signalling from the gut to the appetite centres in the brain. Both neural and endocrine afferent signalling appear to be involved. The presence of food in the stomach stimulates both mechanical and chemical receptors to transmit signals to the brainstem via vagal afferent fibres (Fraser et al. 1995; Willing & Berthoud, 1997; Mathis et al. 1998). Gastric loading volume-dependently decreases short-term food intake, independent of nutrient content. This suggests that stomach distension is more important than content in the neural suppression of food intake (Phillips & Powley, 1996). Nutrient content plays a larger role in the small intestine, stimulating vagal pathways that may contribute to satiety (Schwartz, 2000).

Such neural signals are integrated with endocrine signals from the gastrointestinal tract. Gut hormones have an important physiological role in postprandial satiety. The gut contains a diffuse population of endocrine cells that release several circulating hormones in response to changes in luminal nutriment content. Recent evidence has shown that gut hormones administered at physiological concentrations can influence appetite in rodent models and humans (Wren et al. 2001a; Batterham et al. 2002, 2003b; Cohen et al. 2003).

Cholecystokinin The role of cholecystokinin (CCK) in pancreatic secretion and gallbladder contraction was already well established when it was demonstrated that CCK also reduced meal size in a dose-dependent manner in rats (Gibbs et al. 1973). CCK was the first gut hormone implicated in the control of appetite. The anorectic effects of CCK have since been confirmed in rodent and human studies (Kissileff et al. 1981; Muurahainenn et al. 1988; Moran & Schwartz, 1994). The presence of digestive products in the intestinal lumen stimulates the release of CCK from the proximal small intestine. The anorectic effects of CCK appear to be mediated by the CCKA receptor via the vagal nerve, and are abolished by vagotomy (Smith et al. 1981). Preprandial administration of CCKA antagonists increases meal size in humans and other species (Hewson et al. 1988; Reidelberger & O'Rourke, 1989; Beglinger et al. 2001). Otsuka Long Evans Tokushima Fatty rats, which lack CCKA receptors, are obese and fail to reduce their appetite in response to CCK (Moran, 2000).

However, though the CCKA receptor knockout mouse is also insensitive to the anorectic action of CCK, it has normal food intake and body weight. In addition, tolerance rapidly develops to continuously infused CCK (Crawley & Beinfeld, 1983). Though administering CCK intermittently to rats preserves its anorectic effect, there is little change in daily food intake or body weight because of compensatory eating between injections (West et al. 1984). Therefore, despite its anorectic pedigree, CCK may prove difficult to exploit in the design of antiobesity agents.

Ghrelin Ghrelin is the only peripherally active appetite-stimulating hormone so far discovered. Ghrelin potently stimulates food intake and growth hormone secretion following peripheral administration in man and rats (Tschop et al. 2000; Wren et al. 2000, 2001a). Plasma ghrelin levels are inversely correlated with body weight and rise following weight loss in humans (Cummings et al. 2002). The major source of circulating ghrelin is the stomach, though ghrelin mRNA and immunoreactivity are also found in other regions of the gastrointestinal tract (Date et al. 2000). Ghrelin is composed of 28 amino acids with an acyl sidechain attached to the serine residue at position 3. This acyl group is crucial to ghrelin's orexigenic and growth hormone-releasing actions, which are mediated through the growth hormone secretagogue receptor (GHS-R; Kojima et al. 1999). The GHS-R is highly expressed in the hypothalamus, including the ARC, but also found in the brainstem, pituitary, gastrointestinal tract, adipose tissue and other peripheral tissues (Petersenn, 2002). It has been suggested that des-acylated ghrelin has other biological functions mediated by separate GHS-R subtypes (Baldanzi et al. 2002).

Circulating ghrelin concentrations rise during fasting and fall rapidly after a meal. Ghrelin may therefore be involved in meal initiation (Cummings et al. 2001), though recent work has shown that circulating ghrelin levels do not predict intermeal interval in humans (Callahan et al. 2004). Though calorie intake appears to be the primary regulator of plasma ghrelin levels (Tschop et al. 2000), the exact mechanisms mediating ghrelin release are unknown. There is some suggestion that glucose and/or insulin suppress ghrelin release (Yoshihara et al. 2002), though another study has shown that physiological levels of either appear to have little effect on plasma ghrelin concentrations (Schaller et al. 2003). Circulating ghrelin levels are lower in obese individuals, perhaps reflecting a feedback mechanism to reduce appetite (Tschop et al. 2001).

The orexigenic effects of peripheral ghrelin are mediated via the CNS. Peripheral administration of ghrelin activates neurones in the ARC and the PVN (Ruter et al. 2003) and i.c.v. administration of ghrelin increases appetite (Wren et al. 2000). Infusing anti-ghrelin antibodies into the rat brain inhibits fasting-induced feeding. The orexigenic effects of ghrelin may be mediated via NPY and AgRP. Central injection of ghrelin activates NPY/AgRP neurones and increases hypothalamic NPY mRNA expression. NPY and AgRP antagonists and antibodies block ghrelin-induced increases in feeding (Nakazato et al. 2001). In addition, in NPY and AgRP double knockout mice, the orexigenic action of ghrelin is abolished (Chen et al. 2004). Recently, ghrelin has been found to be expressed in a previously uncharacterized neuronal population adjacent to the third ventricle (Cowley et al. 2003). These hypothalamic ghrelin neurones may be involved in another hypothalamic appetite circuit, though the relationship between central and peripheral ghrelin signalling is currently unknown. It is interesting to note that the patterns of neuronal activation following peripheral and central ghrelin administration differ (Lawrence et al. 2002; Ruter et al. 2003).

Chronic peripheral ghrelin administration causes hyperphagia and obesity in rats (Tschop et al. 2000; Wren et al. 2001b). The ghrelin system therefore offers a potential target for long-term antiobesity therapy. Though there is little change in body weight or food intake in ghrelin or GHS-R knockout animal models, this may be due to compensatory developmental changes in other appetite regulatory systems (Sun et al. 2003, 2004). Further work is required to investigate whether ghrelin antagonists might be viable antiobesity drugs.

Pancreatic polypeptide and peptide YY Pancreatic polypeptide (PP) and peptide YY (PYY) are anorectic gut hormones belonging to the PP fold peptide family, of which NPY is also a member. All three peptides are 36 amino acids long, with primary structures incorporating a number of tyrosine residues. The PP fold is a tertiary structural motif common to all members and important in the binding and activation of the six G protein-coupled receptor subtypes – Y1, Y2, Y3, Y4, Y5 and Y6 –that mediate the overlapping physiological actions of the family (Keire et al. 2000).

PP binds with greatest affinity to Y4 and Y5 receptors. It is found predominantly in the endocrine pancreas, and its release is stimulated by food intake. Postprandial plasma levels are proportional to caloric intake, though other factors, including gastric distension, vagal tone, blood glucose levels and other gastrointestinal hormones, also influence its release. PP inhibits pancreatic exocrine secretion and gallbladder contraction and modulates gastric acid secretion and gastrointestinal motility.

Peripheral administration of PP suppresses food intake and gastric emptying in rodents (Asakawa et al. 2003). In contrast, central administration of PP increases food intake and gastric emptying. It has been suggested that the effect of PP on food intake may therefore be secondary to the effects on gastric emptying (Katsuura et al. 2002). However, intravenous infusion of PP has also been shown to reduce food intake in humans without affecting gastric emptying (Batterham et al. 2003b). The inimical effects of central and peripheral PP may represent the activation of different receptor subtypes. Peripheral PP administration activates neurones in the area postrema, an area where the Y4 receptor is highly expressed (Whitcomb et al. 1997). Central PP may mediate its orexigenic effects through the Y5 receptor, as Y5 knockout mice show a reduced PP-induced increase in food intake (Kanatani et al. 2000). However, Y5 receptor antisense oligonucleotides do not affect this response, suggesting that another receptor may be involved (Flynn et al. 1999).

Peripheral PP increases hypothalamic NPY expression and also increases oxygen consumption and stimulates the sympathetic nervous system, suggesting that it may increase energy expenditure (Asakawa et al. 2003). Chronic administration of PP to ob/ob mice lowers body weight gain, and transgenic mice over-expressing peripheral PP have lower food intake and body weight than controls (Ueno et al. 1999). PP may therefore represent a potential long-term obesity therapy.

PYY is produced by gut endocrine cells and is released into the circulation postprandially (Adrian et al. 1985). Peripheral administration of full-length PYY increases fluid and electrolyte absorption from the ileum and inhibits gut motility, meal-induced pancreatic and gastric secretion and gallbladder and gastric emptying. It also acts as a vasoconstrictor. These effects are thought to be mediated both directly in the periphery and centrally via the area postrema and the nucleus of the solitary tract.

Full-length PYY binds with high affinity to all Y receptors. However, the majority of PYY found stored in cells and present in the circulation is in the form PYY3-36, a truncated 34 amino acid peptide generated by the cleavage of two amino acids from the N terminus. PYY3-36 has high affinity for the Y2, and a lesser affinity for Y1 and Y5 receptors. Peripheral administration of PYY3-36 at doses mimicking postprandial levels activates neurones in the ARC and markedly inhibits food intake in rodents and man. Direct intra-ARC injection also reduces food intake in rats. Food intake and body weight are reduced in rodents chronically treated with peripheral PYY3-36 (Batterham et al. 2002).

However, i.c.v. injection of either PYY or PYY3-36 increases food intake in rats. The actions of peripheral PYY3-36 on satiety appear to be mediated by a direct action on the arcuate Y2 receptor, a presynaptic inhibitory receptor of NPY neurones. This inhibition reduces ARC NPY expression and release, and also results in increased POMC neurone activity. The anorectic effect of peripheral PYY3-36 is abolished in Y2 knockout mice and reduced by a selective Y2 agonist. Though PYY does not have a high affinity for the Y1 and Y5 receptors, they may mediate its central orexigenic action, since PYY3-36-stimulated food intake is reduced in Y1 knockout mice and Y5 knockout mice (Kanatani et al. 2000).

Glucagon-like peptide 1 and oxyntomodulin Glucagon-like peptide-1(7–36) amide (GLP-1) and oxyntomodulin (Oxm) are produced by posttranslational processing of the preproglucagon gene in the CNS and the intestine and colon. Both peptides are released into the circulation in response to nutrient intake and appear to function as satiety signals.

Physiological actions of GLP-1 include stimulation of insulin release, suppression of gastric acid secretion and slowing of gastric emptying. GLP-1 receptors are found in the brainstem, ARC and PVN, in addition to various peripheral tissues. Central administration of GLP-1 to rats inhibits food intake and activates ARC and PVN neurones (Turton et al. 1996). Peripheral administration to rats and humans also inhibits food intake, and activates neurones in the brainstem of rats (Verdich et al. 2001a; Yamamoto et al. 2003).

Like GLP-1, Oxm also inhibits gastric acid secretion and reduces food intake when administered centrally or peripherally (Dakin et al. 2002, 2004). As with GLP-1, these anorectic effects can be blocked by the GLP-1 receptor antagonist, exendin(9–39) (Tang-Christensen et al. 2001), suggesting that Oxm also signals via the GLP-1 receptor. However, Oxm and GLP-1 are equally efficacious at reducing food intake, yet the affinity of Oxm for the GLP-1 receptor is much lower than that of GLP-1 (Fehmann et al. 1994). It is possible that another as yet unknown Oxm receptor exists. Such a receptor must have similarities to the GLP-1 receptor, since it is antagonized by exendin(9–39). Peripheral administration of Oxm activates neurones in the ARC, and intra-ARC administration of exendin(9–39) blocks the anorectic actions of peripheral Oxm but not GLP-1, suggesting that Oxm and GLP-1 may reduce food intake by different pathways (Dakin et al. 2004). Peripheral Oxm may reduce food intake via the ARC, while GLP-1 may act on appetite through the area postrema in the brainstem (Yamamoto et al. 2003).

Chronic central administration of GLP-1 and central or peripheral administration of Oxm cause a reduction in weight gain (Meeran et al. 1999; Dakin et al. 2002, 2004). A reduction in food intake appears to be responsible in GLP-1-treated animals; however, Oxm-treated animals lost more weight than pair-fed controls, suggesting that Oxm may also increase energy expenditure.

While intravenous infusion of GLP-1 in man has a significant effect on appetite at physiological concentrations, only pathophysiological levels of circulating Oxm have been demonstrated to reduce calorie intake (Verdich et al. 2001a; Cohen et al. 2003). Further work is required to elucidate the physiological significance of Oxm in human appetite. Interestingly, Oxm infusion reduces plasma ghrelin levels in humans, which might be responsible for the observed reduction in appetite (Cohen et al. 2003).

Current treatments

Efforts to develop pharmacological treatments for obesity have increased tremendously in the last decade, spurred by the increase in obesity levels and the subsequent recognition of obesity as a chronic medical disease (World Health Organization, 1998). However, there are currently only two drugs licensed in the UK for the treatment of obesity. Orlistat interferes with the actions of gastric and pancreatic lipases, reducing fat absorption from the gastrointestinal tract. This results in a reduced calorie intake, but is associated with significant gastrointestinal side-effects if patients do not comply with a strict low-fat diet. Sibutramine is a re-uptake inhibitor of noradrenaline (NA) and serotonin (5-HT), hence enhancing central NA and 5-HT signalling. Sibutramine promotes a feeling of satiety and therefore reduces food intake. However, because sibutramine has a number of side-effects, including slightly raising pulse rate and blood pressure, it is not suitable for many obese patients who have comorbidities such as high blood pressure.

Both orlistat and sibutramine result in modest weight loss in clinical trials (Padwal et al. 2003). At present the long-term safety or efficacy of either is not clear, and there are some mild concerns regarding cardiovascular side-effects with sibutramine and vitamin deficiency with orlistat. Sibutramine and orlistat are not approved for treatment periods longer than 1 or 2 years, respectively.

Future treatments

More efficacious drugs are required to combat the obesity epidemic. Though the hypothalamic appetite circuits offer a plethora of potential pharmacological targets, few of the drugs targeting them make it to human trials. In addition to the problem of gaining access to the CNS, many of the neuropeptides involved in hypothalamic appetite control also signal in extrahypothalamic regions, where they may be involved in systems unrelated to appetite. For example, drugs targeting the NPY or MC4 receptor systems will have actions throughout the brain, since only a small subset of these receptors are involved in appetite control. The side-effects of sibutramine may be due to it acting as a sympathomimetic in central and peripheral signalling circuits.

In contrast to the current chemical agents which affect systems relatively indiscriminately, the use of gut hormones as therapeutic agents would have the advantage of targeting only the relevant appetite control systems. It seems likely that permanent weight loss will require continuous treatment throughout the life of a patient, or at least for extended periods of time. Since gut hormones are released daily, they have a low expectation of side-effects, and tachyphylaxis is also less likely.

Demonstrating that gut hormones can reduce appetite in normal-weight volunteers does not guarantee the same results in the obese. It is possible that gut hormone resistance contributes to the problem of obesity, as leptin resistance is believed to. However, exogenous administration of GLP 1 or PYY3-36 reduces appetite to a similar degree in obese volunteers and normal-weight controls (Verdich et al. 2001a; Batterham et al. 2003a). In addition, there is evidence to suggest that changes in the release of circulating gut hormones, rather than resistance, may contribute towards the pathophysiology of obesity. Obese humans have lower circulating PYY and GLP-1 than normal-weight controls (Verdich et al. 2001b; Batterham et al. 2003a) and food intake appears not to suppress circulating ghrelin levels in obese individuals (English et al. 2002). It has also been reported that PP levels are lower in the obese (Lassmann et al. 1980), though other studies have not confirmed these results (Pieramico et al. 1990; Wisen et al. 1992). If such changes in gut hormone profile are even partly responsible for obesity, then correcting them with exogenous hormones or antagonists would be a particularly appropriate therapy.

Gut hormones are peptides which have a relatively short circulating half-life and have to be given parenterally. While chronic parental administration of gut hormones thus requires a far from ideal mode of delivery (e.g. subcutaneous injection), it is possible that novel administration routes (e.g. intranasally), or the development of small molecule mimetics or breakdown-resistant analogues with similar bioavailability and brain penetration might lead to a safe, long-term and effective treatment for obesity.

Interestingly, the reduction in appetite and body weight associated with bariatric surgery may be partly due to changes in the release of gut hormones. There is evidence from rodent models of gastrointestinal bypass surgery that a transferable humoral factor is responsible for the associated reduction in appetite (Atkinson & Brent, 1982). Gastointestinal bypass surgery elevates circulating Oxm and PYY (Sarson et al. 1981; Naslund et al. 1997). Ghrelin levels are lower following gastric surgery and do not increase as expected during the associated weight loss (Cummings et al. 2002; Stoeckli et al. 2004).

Further work is required to fully understand the multiple signals regulating appetite and body weight. However, gut hormone signalling systems are shaping up as potentially powerful weapons in the fight against obesity.



The authors wish to express their thanks to Dr W. Dhillo and Dr S. Stanley for assisting with the preparation of this manuscript.