Gut Peptide Signaling in the Controls of Food Intake
Department of Psychiatry and Behavioral Sciences, Ross 618, 720 Rutland Avenue, Johns Hopkins University School of Medicine, Baltimore, MD 21286. E-mail: firstname.lastname@example.org
During a meal and after a meal, ingested nutrients alter the release of a variety of gut peptides that have the potential to modulate food intake. Such feedback peptide signaling can be conceptualized as having three outcomes: meal termination, inhibitory modulation of intake in subsequent meals, and orexigenic modulation. Cholecystokinin, pancreatic glucagons, and amylin are examples of peptides involved in meal termination. They are released rapidly with the onset of feeding and have short durations of action. Peptide YY(3-36) and glucagon-like peptide 1 are peptides for which longer-term feeding inhibitory actions have been proposed. They are released from the distal intestine and have longer durations of actions. Ghrelin is a gastric peptide that stimulates food intake after its exogenous administration. Plasma ghrelin levels fall with feeding and rise with food deprivation. All of these gut peptides have vagal or dorsal hindbrain mediation. Their potential as targets for the development of anti-obesity treatments is under study.
Ingested nutrients come into contact with multiple sites in the gastrointestinal tract that have the potential to alter peptide and neural signaling. Such signaling can serve as feedback mediators influencing current or subsequent food intake. Recent work has characterized the ability of multiple gut peptides to affect eating and, consistent with their different patterns of release around meals, three separate roles for these peptides in overall eating control have been suggested. Most gut peptides with roles in food intake are inhibitory. These can be divided into those that are involved in meal termination and those that inhibit the amount of food consumed across multiple meals. The exception is ghrelin, a peptide with a proposed role in stimulating food intake. In the following sections, the actions of peptides in each of these categories will be discussed.
A number of peptides are released into circulation at the onset of eating or shortly thereafter. A few of these have been shown to play feedback roles in the control of the size of the meal that stimulates their release. Among these are the intestinal peptide, cholecystokinin (CCK),1 and the pancreatic peptides, glucagon and amylin. These peptides share a number of properties including relative brief durations of action, the ability to produce reductions in meal size after their exogenous administration, documented roles for the endogenous peptide in limiting meal size, and signal transduction involving either the vagus nerve and/or the dorsal hindbrain.
CCK was the first gut peptide shown to play a role in the control of eating (1). Peripheral CCK inhibits food intake and induces a behavioral satiety sequence (2). Consistent with its rapid release from the proximal intestine and short duration of effective bioactivity, the actions of CCK are specific to terminating an individual meal. Repeated meal contingent CCK administration results in reliable decreases in meal size but no overall decrease in food intake (3). The actions of glucagon and amylin are similar to those of CCK. Eating rapidly elicits an increase in pancreatic glucagon secretion, and meal-released glucagon is rapidly cleared from the circulation by the liver (4). Hepatic–portal infusions of glucagon at meal onset elicit dose-related reductions in meal size (LaSauter and Geary, 1991). Amylin also inhibits feeding in a dose-dependent and behaviorally specific manner after peripheral administration (5). Amylin is obligatorily cosecreted with insulin by pancreatic β cells (6). Thus, amylin levels rise rapidly with meal onset and remain elevated for a significant period of time during and after meals. Similar to CCK and glucagon, meal-contingent amylin administration reduces meal size and meal duration without affecting intake in subsequent meals (7).
CCK's feeding inhibitory actions are mediated by the afferent vagus. Afferent vagal fibers innervating the stomach and proximal intestine contain CCK receptors, and CCK activates fibers that are sensitive to the volume and/or chemical composition of gastrointestinal contents, providing feedback to the brain on the overall nutrient content of a meal (8). Disruption of vagal afferent signaling from the upper gastrointestinal tract blocks the ability of CCK to inhibit eating (9). Consistent with a hepatic portal site of action, glucagon-induced feeding inhibition is prevented by transecting vagal afferent fibers that innervate the liver (10).
In contrast, the satiety actions of amylin are not vagally dependent. Amylin's feeding inhibitory actions are mediated by the area postrema (AP), a circumventricular organ with a porous blood–brain barrier that lies adjacent to the nucleus of the solitary tract (NTS) the site of vagal afferent terminations. The AP contains amylin receptors, AP lesions block the ability of amylin to inhibit eating, and amylin administration activates cells in both the AP and underlying NTS (11). NTS lesions block the ability of both CCK and glucagon to inhibit food intake (12, 13). This similar hindbrain mediation suggests that the three peptides may be interacting with a common substrate at this level of the neural axis.
The inhibition of eating within meals has been shown to be an important action of endogenous CCK, glucagon, and amylin. Blockade of CCK signaling by specific antagonists or genetic defects resulting in the absence of CCK1 receptors (14, 15) results in increases in meal size. A role for endogenous glucagon in the control of meal size is supported by data showing the ability of hepatic portal infusions of glucagon antibodies to increase meal size (16). A feeding role for endogenous amylin in eating control is supported by experiments showing increases in eating in response to administration of amylin antagonists (17).
CCK analogs with increased duration of action potently inhibit food intake beyond individual meals and can do so over multiple days (18). CCK compounds with oral bioavailability have been developed, and their ability to affect body weight has received preliminary assessments. Similar amylin or glucagon compounds have yet to be developed.
Gut Peptides with Longer-term Feeding Inhibitory Actions
Gut peptides secreted from the distal intestine have been suggested to play longer-term roles in energy balance. Among these are glucagon like peptide-1 (GLP-1) and peptide YY(3-36) [PYY(3-36)]. Both of these peptides are secreted from the L cells in the distal intestine and have patterns of release that occur with a longer latency and duration than those of peptides involved in meal termination. PYY(3-36), an alternate form of PYY, is released in response to the intraintestinal presence of nutrient digestive products. With ingestion, plasma levels begin to increase within 15 minutes, peak at ∼90 minutes, and can remain elevated for as long as 6 hours (19). GLP-1 levels also rise during a meal, peaking at a time after meal termination and remaining elevated between meals (20).
Although there has been some controversy over the reproducibility of the findings in rodents (21), PYY(3-36) has been shown to inhibit food intake in a number of species, including non-human primates and humans (22, 23). GLP-1 actions in food intake have been more difficult to assess. GLP-1 is rapidly metabolized, and bolus GLP-1 administration has often proven to be ineffective on inhibiting food intake (24). However, GLP-1 infusions before and during meal periods have been shown to reduce amounts consumed (25). The availability of a peptide GLP-1 agonist with a longer duration of action has allowed a more complete assessment of the effects of increased GLP-1 signaling on food intake. Peripheral bolus administration of such a compound potently suppresses food intake and does so for extended time intervals (26).
Originally, peripheral PYY(3-36) was proposed to affect feeding by directly inhibiting hypothalamic arcuate nucleus neuropeptide Y (NPY)-containing neurons (22). NPY is a peptide that increases eating when administered into the brain, and PYY(3-36) has been shown to have relative specificity for the presynaptic inhibitory Y2 NPY receptor subtype. Consistent with this proposal, peripheral PYY(3-36) was shown to decrease arcuate NPY mRNA expression. However, recent work supports the view that, rather than directly accessing brain sites, both PYY(3-36) and GLP-1 may inhibit feeding through alterations in vagal signaling (27). Surgical vagotomy blocks the ability of peripheral administration of both peptides to inhibit eating.
It is not yet clear whether feeding inhibition is a physiological action of meal-induced PYY(3-36) or GLP-1. Infusions that reduce food intake produce plasma levels greatly in excess of those found after a meal (28). There is a recent suggestion that combining PYY(3-36) and GLP-1 may allow feeding inhibitions at plasma levels of each closer to those found after a meal (29). The pattern of release of PYY(3-36) and its effects on hypothalamic gene expression suggest that it is not involved in meal termination but exerts its influences on food intake over longer periods. Consistent with this, peripheral PYY(3-36) has been reported to prolong the latency to the first meal after administration and to have a prolonged effect on meal size through a daily 6-hour feeding period (23). GLP-1 analogs have been shown in preclinical studies to exert lasting effects on food intake, suggesting a potential target for drug development (26). Whether a Y2 PYY(3-36) analog could have similar potential has yet to be shown.
Orexigenic Gut Signaling
Ghrelin, a recently identified peptide, is synthesized primarily in gastric epithelial tissue. Eating affects plasma ghrelin levels but in a manner opposite to the effects on peptides that inhibit eating. Ghrelin levels are high immediately before meals and fall in response to nutrient ingestion (30). Ghrelin administration has been shown to increase food intake in rodents and in a buffet test meal situation in human subjects (31, 32). Repeated ghrelin administration can result in increases in body weight, and these increases seem to be mediated through alterations in both food intake and energy expenditure (33). Consistent with its pattern of release, ghrelin seems to be involved in stimulating meal initiation. Ghrelin administration reduces the latency to begin eating and stimulates additional meals but does not alter meal size (34). A physiological role for ghrelin in food intake is supported by recent data showing the ability of ghrelin antagonists to reduce feeding (35).
Ghrelin also increases eating when administered directly into the brain, suggesting a central site of action for peripheral ghrelin. The feeding stimulatory actions of both peripheral and central ghrelin seem to involve arcuate nucleus NPY neurons. Arcuate NPY-containing neurons express growth hormone secretagogue receptors, and both peripheral and central ghrelin administrations activate arcuate NPY neurons and increase NPY mRNA expression (36, 37). However, how peripheral ghrelin activates arcuate NPY neurons is not clear. Unlike the AP, the arcuate nucleus does not have fenestrated capillaries, and although ghrelin transport across the blood–brain barrier has been shown, the major direction of that transport seems to be from the brain to the periphery (38). Ghrelin neurons have been identified in the periventricular hypothalamus, raising the possibility that there are separate central and peripheral ghrelin actions (39). Vagal mediation of the feeding actions of peripheral ghrelin has also been suggested. Ghrelin reduces activity in gastric vagal afferents, and vagotomy has been reported to block the ability of peripheral but not central ghrelin to stimulate food intake (40).
As well as fluctuating with meals, ghrelin levels are elevated by acute or chronic food deprivation or by weight loss, and ghrelin levels fall with weight gain (41). Low ghrelin levels in obesity suggest that a ghrelin antagonist may not be an ideal tool for weight loss, but the fact that levels are elevated with weight loss suggests that such a compound could be a useful aid for preventing weight gain after dieting.
Gut peptide signaling modulates eating, either stimulating intake or providing feedback inhibition within individual or across multiple meals. Vagal afferent neurons and/or hindbrain signaling seem to mediate the feeding actions of many of these peptides. These signaling pathways are potential targets for anti-obesity drug development.
Nonstandard abbreviations: CCK, cholecystokinin; AP, area postrema; NTS, the nucleus of the solitary tract; GLP-1, glucagon like peptide-1; PYY(3-36), peptide YY(3-36); NPY, neuropeptide Y.