- Top of page
- Distally Arising GI Signals
- Ileal Signals and Satiation
- Apo A-IV
WOODS, STEPHEN C. Dietary synergies in appetite control: distal gastrointestinal tract.
Version of Record online: 6 SEP 2012
2006 North American Association for the Study of Obesity (NAASO)
Special Issue: Dietary Synergies in Appetite Control
Volume 14, Issue S7, pages 171S–178S, July 2006
Woods, S. C. (2006), Dietary Synergies in Appetite Control: Distal Gastrointestinal Tract. Obesity, 14: 171S–178S. doi: 10.1038/oby.2006.301
WOODS, STEPHEN C. Dietary synergies in appetite control: distal gastrointestinal tract.
Myriad signals interact to influence how many calories are consumed in individual meals. Some of these signals begin before any food is ingested, and others are generated as food interacts with receptors from the mouth to the intestinal wall. This article reviews the influence of signals arising from the more distal regions of the small intestine, focusing on glucagon-like peptide-1 (GLP-1),1 peptide tyrosine-tyrosine (PYY), apolipoprotein A-IV (apo A-IV), and enterostatin. When administered to animals and humans, there is strong evidence that meal size can be reduced by each of these signals. An important point is that even though these intestinal signals can be secreted in response to food stimulating local receptors in the distal gut, it is more likely that neural and humoral reflexes normally stimulate their release during meals.
This is a review of signals secreted from the distal gastrointestinal (GI) tract that modulate food intake, providing a complement to the previous discussion by G.P. Smith on signals from the proximal gut. I begin with several general points relevant to consideration of the control of eating and meal size. Myriad signals influence ingestion as indicated in Figure 1.
The sight, smell, and taste of foods influence the initiation of eating and the selection of specific foods as well as contributing to meal size. The influence of these factors is malleable, with flavor aversions and preferences developing as a result of experience with specific foods. As a general oversimplification, tastes that are not rejected outright contribute a positive effect to food intake, driving further consumption during an ongoing meal. As their potency wanes and/or as inhibitory signals generated by the food interacting with the GI tract accumulate, continued eating is inhibited.
As ingested food interacts with the lining of the stomach and intestines, numerous signals are generated, many of which contribute to satiation or the perception of fullness. Included are several gut peptides that are secreted in response to specific macronutrients and gastric distention (1, 2). The generally accepted model is that GI signals that limit meal size, collectively known as satiety signals, are secreted in proportion to ingested calories and accumulate during a meal, ultimately providing an integrated signal reflecting the quantity and quality of ingested food and causing eating to stop.
Figure 1 summarizes the array of signals generated by the GI tract that converge on the brain to influence food intake. As indicated, in addition to signals that elicit satiety, other signals, in particular several hormones, are secreted in proportion to body fat and interact with satiety signals to influence food intake, and these have been reviewed elsewhere (3, 4). Most noteworthy are leptin and insulin, and these are critically important for maintaining body weight (body fat, actually) relatively constant over long intervals, for when the insulin or leptin signal to the brain is altered on a chronic basis, body weight changes.
Adiposity signals are thought to act by changing the sensitivity of the brain to the action of satiety signals (5, 6, 7, 8, 9, 10, 11, 12, 13). For example, if one loses weight, the levels of insulin and leptin reaching the brain decrease, and this results in reduced sensitivity to cholecystokinin (CCK) and other satiety signals. The consequence is that individuals eat larger average meals until weight is restored. Likewise, the response to weight gain is increased adiposity signals, increased sensitivity to satiety signals, and consumption of smaller meals. Maintenance of body weight is sufficiently important that if constraints are imposed on meal size, individuals readily adopt alternative meal patterns to obtain the requisite number of calories each day (14, 15, 16, 17, 18, 19, 20, 21). The key point is that what is regulated physiologically is the amount of food eaten once a meal begins rather than the meal pattern per se and this, in turn, is related to the maintenance of a relatively stable body weight.
To perform a routine act such as eating a meal, an individual faces several challenges or goals that must be satisfied (Figure 2). The first is to find and obtain adequate ingestible energy; this foraging behavior is not covered in this review. Once food is secured and eating begins, a second major challenge is to take in sufficient energy, but not excessive energy, within the course of a single meal. This is facilitated, in part, by making anticipatory responses once a meal, and particularly a large meal, is imminent, because only when the body is adequately prepared to cope with the pending caloric load can a large number of calories be safely consumed; i.e., directing the consumption of adequate energy without overconsuming is a major challenge with which the hungry brain has to grapple (16, 19, 20, 21, 22). After a meal has ended, an important feature of the post-ingestive state is resting and allowing the absorptive and digestive process to proceed without interference from other activities.
The brain recruits a number of important responses to meet these challenges. One key strategy for consideration in this discussion is the generation of neurally elicited cephalic responses such as insulin secretion and various digestive peptides to better cope with the influx of ingested nutrients into the blood. Another is to monitor accurately and continuously how the meal is proceeding to truncate the meal at the appropriate point with regard to ingested calories; i.e., the rate at which calories and, perhaps, even the mix of macronutrients is entering the GI tract is a major determinant of when satiation will occur.
Cephalic responses are critically important in the timeline of a meal as depicted in Figure 3. They are elicited by cues that herald an influx of new calories into the body, and they are initiated before actual ingestion begins and probably continue so long as food is being tasted (19, 23, 24, 25, 26, 27, 28, 29). Most, if not all, are neurally elicited GI secretions that make the digestive and absorptive processes more efficient, enabling, for example, the more rapid removal of glucose from the blood during meals. Once the meal begins, other signals arise as food enters the stomach and subsequently the duodenum. These within-meal signals, including gastric distention and satiety peptides such as CCK, overlap temporally with cephalic responses, and both provide important feedback to the brain. As satiety signals accumulate, their collective impact causes eating to stop and the meal to be ended. Key unanswered questions concern which satiety signal, or mix of satiety signals, is most potent, and whether satiety signals could be mimicked by drugs or enhanced by specific foodstuffs to reduce food intake therapeutically.
There is a logical conundrum when considering satiety signals secreted from the distal gut (jejunum and ileum) because ingested food might not move that far along the GI tract before eating stops; i.e., if the release of critical ileal peptides is stimulated by partially digested chyme passing through the intestinal lumen, distal GI hormones could at best be secreted late within a meal. Therefore, if distal peptides are important for ending a meal, it is more likely that other more proximally arising meal-related signals stimulate the distal gut, causing it to release its secretions within the time frame of a single meal. As an example and as discussed above, the taste of food in the mouth continues throughout the entire eating episode, from the first bite to the final swallow, such that a taste bud-to-brain-to-distal GI tract neural reflex might trigger the release of satiety signals from the distal intestine. However, the physical qualities of taste probably do not vary considerably over the course of a meal (30), such that taste per se is unlikely to have a major impact on terminating most meals. The relative impotence of taste to terminate eating is also demonstrated by the lack of satiation during sham eating (31, 32, 33). There is evidence, however, for a reduction in the hedonic pleasure elicited by tastants over the course of a meal, a phenomenon called sensory-specific satiety that could contribute to a cessation of eating, especially when only one food is available to eat (34).
Because different signals that individually are able to reduce meal size become active at different times during a meal, an interesting and important question concerns whether each is necessary or whether there is some integrated signal based on additive or synergistic interactions of the mix. Alternatively, it may be the case that the individual signals are redundant and that only one is necessary or sufficient during any given meal. The sham feeding paradigm makes it clear that neither taste alone nor the act of eating and passing food into the stomach is sufficient to terminate a meal, whereas putting food directly into the duodenum is sufficient when an animal is tasting and swallowing food (35, 36). Although this could be taken as strong evidence that signals such as CCK that arise mainly from the duodenum are all that is required to cause satiation during normal meals, it is just as feasible that food in the duodenum elicits neural, hormonal, or paracrine reflexes that, in turn, cause the release of ileal or pancreatic signals as depicted in Figure 3 and that the combination of all of the GI signals is what is important during normal meals.
The important point is that when a meal is being eaten, especially a mixed meal containing different foodstuffs and macronutrients, many and, perhaps, dozens of signals provide continuous information to the brain. Most of this information arrives at the hindbrain where a decision to stop eating is made. This latter conclusion is based on the ability of CCK to reduce meal size in animals whose forebrain is no longer attached to the hindbrain (37). That said, under normal conditions, the forebrain is critically important because factors such as relative adiposity, the social situation, or a particularly appealing dessert can easily influence the amount eaten.
For the brain to make an informed decision as to when to stop eating, it needs to use the most reliable information, and it needs this information as early within a meal as possible. In most experiments with animals and humans in which satiation or satiety has been assessed, the subjects consume a familiar food, often consisting of a single foodstuff. The familiarity aspect is important for two reasons. The first is that novel foods may not be eaten in sufficient quantity to provide an unambiguous assessment of how many calories would otherwise be consumed, and the second is that previous experience with the food implies that the individual could well have learned important associations between specific signals such as CCK and the number of calories ultimately entering the circulation; i.e., the nature of most experiments, especially those using rats or mice, dictates that subjects know precisely what each meal-generated signal means in a caloric or macronutrient sense.
There are implications of this premise for strategies to consider meal size as a therapeutic target. If a novel food were designed that elicited a disproportionately large secretion of one or another satiety signal, or if a pharmaceutical agent were able to mimic a particular satiety signal, it is probably the case that these treatments would be efficacious in reducing meal size, at least for a while. Evidence suggests, however, that over the span of several meals using the same intervention, the brain would learn to stop paying attention to that particular satiety signal because it would no longer be a reliable indicator of how many calories are consumed; i.e., the brain could simply shift its focus to one or another of the other signals that remain good harbingers of the calories that are being consumed. Considerable evidence indicates that the efficacy of CCK at reducing meal size can be manipulated by dissociating its levels from calories consumed (38, 39). Likewise, the ability of gastric distention to serve as a caloric signal can also be changed, as demonstrated by experiments in which non-caloric bulk is added to food (40). In the first few days of such a manipulation, animals evidently pay attention to stomach volume cues and stop eating when a particular volume has been achieved, but over the course of several days, the animals increase the volume consumed within each meal such that caloric content returns to baseline levels. This can be interpreted to indicate that the animals have learned to disregard stomach distention and pay more attention to other cues in determining when to end a meal. As a final point, there is also evidence that even if a particular satiation-evoking manipulation were to remain efficacious over time, individuals would adapt to eating smaller meals by consuming more frequent meals (17).
Independent of whether chyme reaches receptors lining the lumen of the ileum or not during a meal, there is no doubt that ileal peptides are secreted during the course of a meal that interact with local receptors as well as entering the blood. As discussed above, the actual mechanism may well rely on neural or hormonal stimulation of enteroendocrine cells lining the ileum. The two best known ileal compounds that reduce food intake when exogenously administered are GLP-1 and PYY. Both have been reported to be efficacious in both animals and humans. GLP-1 analogs are, in fact, being prescribed to treat type 2 diabetes, and off-label use suggests that they reduce food intake as well (41, 42). GLP-1's best known action is as an incretin, stimulating insulin secretion during meals and, thereby, improving glucose tolerance (43, 44, 45).
GLP-1 is made by post-translational modification of the precursor molecule proglucagon, which is synthesized by enteroendocrine L cells in the ileum and colon (46). GLP-1 secretion is elicited by both nutrient and neurohumoral stimulation and GLP-1 circulates as both GLP-1 (7–37) and GLP-1 (7–36) (47). Within 2 minutes or so, most plasma GLP-1 is degraded by the enzyme dipeptidyl-peptidase IV (DPP-IV) (47). The GLP-1 receptor (GLP-1r) is expressed in the gut and endocrine pancreas and in the central nervous system (48). The release of GLP-1 from the distal gut results from both direct and indirect mechanisms (49). Although GLP-1 is rapidly secreted from the ileum during a meal, the process does not require actual nutrient contact with the endocrine L cells (50). Rather, secretion is thought to result from neural or humoral signals arising from the proximal intestine. Because GLP-1 inhibits GI motility, reduces GI secretions, and attenuates gastric emptying, it has been touted as a major component of the ileal brake, an inhibitory feedback mechanism by which the distal gut regulates the overall transit of nutrients along the GI tract (45, 51).
Exogenous GLP-1 reduces food intake in animals and humans (52, 53, 54, 55, 56, 57). The anorectic actions of GLP-1 are probably mediated through both peripheral and central mechanisms, and a population of neurons that synthesize GLP-1 is located in the brainstem and projects to hypothalamic and brainstem areas important in the control of energy homeostasis (58, 59). In rats, central administration of GLP-1 dose-dependently reduces food intake (52, 53, 54, 55, 56, 57), an effect that is reversed by coadministration of the GLP-1r antagonist exendin (9–39) (52). Centrally administered GLP-1 reduces food intake through at least two mechanisms. GLP-1rs in the hypothalamus are thought to reduce intake by the normal controls of caloric homeostasis (60, 61, 62). GLP-1rs in the amygdala, on the other hand, are thought to reduce food intake by eliciting symptoms of malaise (60, 61, 62) and are presumably responsible for the conditioned taste aversions caused by GLP-1 (56, 57, 63).
Within the brain, GLP-1 mediates both the endocrine and the behavioral responses to stress in rats (64), implying that the GLP-1 signal has complex actions, only some of which are directly relevant to the control of caloric homeostasis. Peripherally administered GLP-1 elicits satiety in healthy (65), obese (66), and diabetic humans (67, 68). Because the half-life of active GLP-1 is <2 minutes, any direct effects are likely transient such that the reduction of food intake is more likely a result of GLP-1's inhibitory effects on GI transit and/or reduced gastric emptying (69). However, peripherally administered GLP-1 does cross the blood-brain barrier (70), such that its role and relative contribution within the central nervous system on food intake are not conclusive.
Because it both reduces food intake and stimulates insulin secretion, GLP-1 is a logical candidate as a therapeutic agent in the treatment of obesity and type 2 diabetes (71, 72). However, as discussed above, a major problem in the potential use of GLP-1 as a treatment is its rapid degradation by DPP-IV. Hence, treatment strategies based on GLP-1 will likely use compounds that are not so rapidly destroyed, such as exendin-4, a 39-amino acid peptide originally isolated from the venom of the Gila monster salivary gland that shares 53% homology with GLP-1 (7–36) (73, 74). Exendin-4 (exenatide) is a potent agonist for the GLP-1r and has a significantly greater one-half-life than endogenous GLP-1 (75). In animals and humans, exendin-4 reduces gastric emptying (76), lowers fasting plasma glucose (76, 77), and reduces food intake (76, 78, 79), supporting its potential use as a possible treatment for obesity and diabetes. An alternative strategy may be to use compounds that compromise DPP-IV, thus prolonging the one-half-life of endogenous GLP-1.
PYY is a member of the pancreatic polypeptide family, a group of peptides that includes pancreatic polypeptide and neuropeptide Y (NPY). Like proglucagon-derived peptides, PYY is synthesized and secreted by L cells in the distal ileum and colon (80), and most L cells that secrete GLP-1 also secrete PYY (81). PYY is secreted as PYY (1–36) and is degraded to PYY (3–36) by DPP-IV (82, 83). Receptors that mediate the effects of PYY belong to the NPY receptor family and include Y1, Y2, Y4, and Y5 (84). PYY (1–36) is an agonist for the Y1 and Y2 receptors (82, 83) and is a potent orexigen within the brain (85). Once PYY (3–36) is formed, it is a highly selective agonist activity for the Y2 receptor and has been reported to reduce food intake (86, 87, 88). The secretion of PYY from the gut is proportional to the caloric density of the ingested nutrients (89), with lipids and carbohydrates being the primary nutrients in the stimulation of PYY release (90, 91).
Like GLP-1, PYY has been implicated as a major component of the ileal brake (92). Its secretion can be stimulated by the presence of nutrients, particularly lipids, within the ileum itself (93) or else before direct nutrient contact because of neurohumoral signals originating from the proximal gut (94). Pharmacological experiments have yielded conflicting results regarding PYY as a satiety signal. Several reports indicate that PYY is an orexigenic peptide (85, 95, 96, 97, 98, 99), with feeding stimulatory properties superior to those of NPY (for review, see (96)). This is particularly the case when PYY is administered directly into the cerebral ventricles. In contrast, it was recently reported that peripheral administration of PYY (3–36) reduces food intake in rodents, non-human primates, and humans (86, 87, 88). These findings are controversial, especially for rodents (100, 101), and there are no clear conclusions other than that PYY stimulates food intake when administered into the brain.
It has been hypothesized that PYY influences food intake through its interaction with Y2 receptors in the hypothalamic arcuate nucleus (ARC) (86, 87, 88). The ARC is a major conduit for feeding-related signals (3, 4, 102, 103), and it is thought that circulating PYY gains access to this area of the brain because it freely crosses the blood-brain barrier (104).
Apo A-IV is a large peptide synthesized by intestinal cells during the packaging of digested lipids into chylomicrons that subsequently enter the blood by the lymphatic system (105, 106, 107). Apo A-IV is also synthesized in the ARC (108). Systemic or central administration of apo A-IV reduces food intake and body weight of rats (105, 106, 107, 109), and administration of apo A-IV antibodies has the opposite effect (110). Because both intestinal and hypothalamic apo A-IV are regulated by absorption of lipid but not carbohydrate (111), this peptide may be an important link between short- and long-term regulators of body fat (105, 106, 107).
Another digestion-related peptide, enterostatin, is also closely tied to lipid digestion. When ingested fat enters the intestine, the exocrine pancreas is stimulated to secrete lipase and colipase to aid in the digestive process; enterostatin is a cleavage by-product of the formation of colipase from procolipase. Administration of exogenous enterostatin either systemically (112, 113) or directly into the brain (114) reduces food intake, and, when rats are given a choice of foods to eat, the reduction is specific for fats; that is, enterostatin does not decrease the intake of carbohydrates or proteins (115). Therefore, two peptides that are secreted from the intestines during the digestion and absorption of lipids, apo A-IV and enterostatin, act as signals that decrease food intake, and at least one of them selectively reduces the intake of fat. Macronutrient specificity has not been assessed with apo A-IV.
Nonstandard abbreviations: GLP-1, glucagon-like peptide-1; PYY, peptide tyrosine-tyrosine; apo A-IV, apolipoprotein A-IV; GI, gastrointestinal; CCK, cholecystokinin; DPP-IV, dipeptidyl-peptidase IV; GLP-1r, GLP-1 receptor; NPY, neuropeptide Y; ARC, arcuate nucleus; PVN, paraventricular nucleus.