The Gastrointestinal Tract and Control of Food Intake

62.1.2.1

CCK

The action of exogenous and endogenous CCK on food intake is quite well understood and there is now overwhelming evidence that CCK plays a physiological role in the regulation of food intake, principally via an effect on meal size and duration. CCK is secreted from I cells located principally in the duodenal mucosa. However CCK-containing EECs are found along the length of the small intestine. Exogenous injection of CCK decreases meal size in many species, including rodents, nonhuman primates, and humans. The ability of CCK to decrease food intake is mediated by an action of CCK at the CCK type 1 (CCK1R) as demonstrated by the much decreased potency of gastrin to decrease food intake, the effectiveness of specific and potent CCK1R antagonists, and by the lack of response to exogenous CCK in transgenic mice with a deletion of the CCK1R. Moreover, the ability of fats and protein in the intestinal lumen, which release CCK from I cells, to decrease food intake is also abolished by CCK1R antagonist, showing a role for endogenous CCK in intestinal feedback control of food intake. Administration of CCK not only has effects to decrease meal size and inhibit ongoing feeding but will also induce the sequence of behavioral changes associated with satiation, including a decrease in activity and in rodents, the onset of grooming behaviors. This supports the idea of CCK as being a physiological satiety factor. In humans, administration of CCK can induce the feeling of satiation and decrease meal size. Administration of CCK1R antagonists in humans increases hunger scores, meal size, and caloric intake.

The use of naturally occurring genetic mutations and the generation of transgenic animals have also helped define a physiological role of CCK in the control of food intake. Experiments in mice with a CCK1R deletion have shown that these mice ingest larger and longer meals. However, there is little change in overall daily food consumption or change in body weight in these mice, even in the long term, although a potential confounder in interpretation is that CCK1R-null mice were generated on an obese-resistant mouse strain. Similarly, CCK-null mice have normal food intake, body weight, and adiposity, although individual meals are larger. A spontaneously occurring 64 kb genetic deletion in CCKA receptors in a strain of rats results in a different phenotype; Otsuka Long Evans Tokushima Fatty (OLETF) rats were originally described as developing obesity and diabetes on regular laboratory chow diet and were selectively bred to provide a consistent phenotype of obesity, hyperglycemia, and late-onset insulin resistance. These rats are insensitive to the ability of exogenous CCK to inhibit food intake. Moreover, these rats eat longer and larger meals, and unlike CCK1R-knockout mice, do become hyperphagic, and gain body weight even when ingesting regular low fat laboratory chow. When OLETF rats were pair-fed to their lean counterparts, they remained lean, suggesting that CCK and the CCK1R may indeed play a role in the long-term regulation of body weight. Loss of CCK1R expression in hypothalamus in rats drives overexpression of neuropeptide Y (NPY) in dorsal medial hypothalamus (DMH). In support of a role for central CCK1R are data from experiments where direct injection of CCK into the hypothalamus decreases food intake in rats.

Plasma levels of CCK rise after a meal or following perfusion of either lipid or protein-rich solutions into the proximal small intestine in humans and experimental animals. However, although CCK may have humoral actions to regulate the gallbladder and pancreas, for example, its action to decrease food intake and stop eating are mediated via extrinsic, vagal afferent pathway. In rats, either systemic capsaicin pretreatment or perivagal capsaicin treatment to functionally ablate vagal afferents blunts the effect of exogenous and endogenous CCK to inhibit feeding. Vagal afferents innervating the abdominal organs express CCK1Rs and CCK induces action potentials in vagal afferent neurons both in vivo and in vitro. Similarly, activation of vagal afferent neurons by lipid is abolished by CCK1R antagonist, suggesting that endogenous CCK also acts to increase vagal afferent firing. Consistent with these observations, CCK activates second-order neurons in brainstem, the initial central nervous system site of integration of signals arising from the GI tract.

62.1.2.2

PYY

Peptide YY (so denoted because of the tyrosine residue at both terminals) (PYY) is secreted from EECs located in the proximal and distal small intestine, and the proximal colon; PYY is released as a 39 amino acid peptide in response to either lipids or carbohydrate in the intestinal lumen. PYY is also released in response to short-chain fatty acids generated by the gut microflora in the colon. The intact PYY molecule is rapidly broken down to produce PYY3-36, the form that inhibits food intake. In rodents, administration of PYY3-36 decreases food intake, and in humans, peripheral infusion of PYY3-36 decreases hunger and 24 h caloric intake. Initial reports on the effects of PYY3-36 to inhibit food intake were difficult to reproduce in some laboratories; however, it seems that the expression of its effects are sensitive to environmental stressors and it is now generally agreed that exogenous administration of PYY3-36 inhibits food intake. Thus, the acute anorectic effects of PYY3-36 have been reproduced in several species including rodents, nonhuman primates, and humans. Like CCK, the predominant effect of PYY3-36 is to decrease the size and duration of meals.

Receptors for PYY belong to the NPY family of receptors; PYY3-36 is a potent agonist for the Y2 receptor subtype and inhibits food intake via the Y2 receptor subtype. PYY3-36 is without effect on food intake in the Y2 receptor null mouse, and administration of Y2 receptor antagonists blocks the action of PYY3-36. The precise mechanism and location of action of PYY3-36 to influence eating behavior are not fully elucidated. PYY3-36 can cross the blood-brain barrier and will, therefore, have access to central Y2 receptor. There is evidence to suggest that PYY3-36 acts on Y2 receptors in the arcuate nucleus of the hypothalamus, a region of the brain crucial in the regulation of food intake and body weight. PYY3-36 directly administered into the arcuate nucleus inhibited food intake and also activates arcuate neurons, as measured by expression of the early gene c-fos. It is thought that the mode of action is on proopiomelanocortin (POMC)-expressing neurons to reduce NPY expression and release, which will in turn increase activity in POMC neurons and reduce appetite.

However, similar to CCK, there are Y2 receptors expressed by vagal afferent neurons innervating the GI tract, and it is possible that this pool of receptor plays a role in mediating the effects of PYY to inhibit food intake. At least two separate laboratories have shown that the anorectic effect of PYY3-36 is absent in vagotomized rats ; however, neither vagotomy nor capsaicin treatment attenuated the effects in mice.

Although there is considerable evidence that exogenous PYY can inhibit food intake, the physiological role of PYY has been harder to establish. Recently, the use of genetically manipulated mice has shed some light on the potential role of PYY. PYY ^−/− mice are hyperphagic and have an increased body weight and adiposity, even when ingesting a regular laboratory chow diet ; moreover, the phenotype of this knockout mouse can be rescued by administration of exogenous PYY3-36. However, other studies have found only minor changes in food intake and body weight in female PYY-null mice, and yet another study has found no change at all. Overexpression of PYY was without effect on either food intake or body weight in mice. Mice deficient in the Y2 receptor demonstrate a small but significant increase in body weight even when ingesting chow, an effect that was exacerbated when these mice were allowed to ingest a high-fat diet. Thus, the physiological role of PYY to inhibit food intake is not fully established at this time. However, there is considerable interest in the role of PYY released from the gut in mediating the beneficial effects of bariatric surgery on obesity; this will be considered later in this chapter.

62.1.2.3

GLP-1

Considerable attention has recently focused on the role of the incretin hormone, GLP-1, in the regulation of food intake and body weight. GLP-1 is expressed in EEC found throughout the small intestine and the proximal large intestine; in the small intestine, there is an increase in density from the proximal to the distal gut. GLP-1 is released by lumenal nutrients, such as glucose and long-chain triglyceride, and also by other factors including bile acids and short-chain fatty acids. Release of GLP-1 is more complex than some other gut hormones as it is directly regulated by contact of EEC with nutrients and also indirectly via release of GIP, another incretin hormone, from the proximal gut. The relative contribution of these two mechanisms is unclear and may depend on the size of the meal and thus nutrient spread to the distal small intestine.

Exogenous administration of GLP-1 decreases food intake in several species including humans, rats, and mice. The effect of GLP-1 is inhibited by a GLP-1 receptor antagonist and is absent in GLP1-R-null mice. However, the pathway by which GLP-1 mediates effects on food intake is unclear. GLP1-Rs are expressed on vagal afferent neurons, and there is compelling evidence to suggest that peripheral GLP-1, possibly including that released from EEC, acts on this pathway to inhibit food intake. There are numerous pieces of evidence to support this mechanism of action, including reduced effectiveness of exogenous GLP-1 or low doses of the long-acting agonist excendin-4 after capsaicin treatment or subdiaphragmatic afferent vagotomy in rodents. In addition, in humans with vagotomy in combination with pyloroplasty, intravenous GLP-1 has no effect on meal size. Perhaps the most compelling evidence for a role of GLP-1 Rs expressed on vagal afferents comes from experiments in rats where GLP-1Rs in vagal afferent neurons is knocked down using a lentiviral vector. After 3 weeks, the investigators reported an approximate 50% decrease in RNA and protein expression of the receptor, which was accompanied by an increase in meal size and duration in the dark phase, and a complete abolition of the effect of exogenous GLP-1 administration on food intake. Although GLP-1Rs are expressed in a number of areas of the CNS involved in the control of food intake, including the hypothalamus, limbic regions, and brainstem, there is little evidence to suggest that GLP-1 released from EEC influences these populations of receptors in the brain.

However, whether peripheral GLP-1 plays a role in long-term regulation of food intake or body weight is unclear. GLP-1R KO mice have overall normal daily food intake and maintain body weight. Therefore, the role of GLP-1 in long-term regulation of food intake unclear, but similar to CCK may be involved in short-term regulation of meal size and duration. Unlike CCK, long-acting agonists of GLP-1R do reduce body weight in models of increased body weight such as diet-induced obesity in rodents or obese humans, or in type 2 diabetes. Moreover, RYGB and VSG in type 2 diabetics and in experimental animals are accompanied by increase in plasma levels of GLP-1 and have been hypothesized to play a role in the beneficial effects of bariatric surgery on appetite, food intake, and glucose homeostasis.

62.1.2.4

Leptin

Although leptin is largely considered to be an adipocytes hormone, leptin is also found in endocrine cells within the gastric mucosa. As such, it can be considered to be anorexigenic gut hormone. However, the extent to which gastric leptin contributes to overall control of food intake or body weight is unclear. Exogenous administration of leptin inhibits food intake in humans and experimental animals; this effect is likely mediated at least in part via a direct action of leptin on neurons in the hypothalamus. However, vagal afferent neurons express ObR and leptin activate these neurons, as determined in electrophysiological studies and also in cell culture. Deletion of leptin receptors in vagal afferent neurons induces a mild-obese phenotype, changes meal patterns, and markedly reduces the ability of CCK to inhibit food intake. Taken together, the data support a role for gastric leptin and vagal afferent neurons in eating behavior and body weight.

62.1.3.1

Ghrelin

Until 1999, all identified gut hormones produced an inhibition of food intake upon exogenous administration and led to the concept of intestinal feedback inhibition in response to ingested nutrients. However, in 1999, ghrelin was identified in EEC in the gastric mucosa. Ghrelin is an endogenous ligand for the growth hormone secretagogue receptors (GHSR). Ghrelin is secreted from EEC the gastric mucosa; these cells are very numerous, accounting for around 20% of the total EEC cell population in the gastric mucosa. It is interesting to note that these are of the closed type suggesting that they may not be directly regulated by luminal stimuli (in contrast to the majority of EEC found in the small intestine). Unlike other gut hormones that are released by the presence of food and initiate intestinal feedback inhibition of gut function and food intake, fasting elevates circulating levels of ghrelin and feeding decreases plasma levels of ghrelin. Indeed, plasma levels of ghrelin are highest right before the onset of a meal in humans and in rodents. Moreover, the postprandial decrease in plasma ghrelin does not seem to rely on exposure of the gastric or intestinal mucosa to nutrients, but rather is likely to be neurally mediated. Consistent with these observations, exogenous administration of ghrelin stimulates food intake by inducing meal initiation and also by increasing the number of meals, rather than any change in meal size or duration. The majority of the neural regulation of the oral and cephalic response to a meal, including stimulation of peptides from EECs the gastric mucosa, is mediated by the parasympathetic subdivision of the autonomic nervous system; however, it is interesting to note that the release of ghrelin seems to be regulated by the sympathetic innervation.

Ghrelin is a 28 amino acid peptide that undergoes considerable posttranslational modification. There is n-octanoylation (i.e., acylation) of the serine residue in position 3, but both the acylated and desacylated form of ghrelin seem to occur naturally. However, the biologically active form of ghrelin to inhibit food intake is only the acelyated form. In addition, there is evidence that other long-chain fatty acids can substitute for the octanoic acid. It is of interest to note that ghrelin is synthesized and stored not only in EECs in the gastric mucosa but also in neurons in the central nervous system, including those involved in the regulation of food intake and appetite in the brainstem and hypothalamus.

Ghrelin can have direct effects on neurons within the hypothalamus to initiate meals. Ghrelin receptors are expressed by NPY/AGRP neurons in the arcuate nucleus of the hypothalamus; injection of ghrelin into this region results in the onset of a meal, supporting this as a site of action of endogenous ghrelin. Ghrelin receptors (GHSR1A) are also expressed by vagal afferents, although the role of this pool of receptors in mediating the effects of ghrelin is highly controversial. Administration of ghrelin inhibits CCK signaling in vagal afferent neurons and subdiaphragmatic vagotomy or capsaicin treatment to ablate afferent neurons inhibits the influence of exogenous ghrelin on feeding. However, subdiaphragmatic vagal deafferentation had no effect on the ability of ghrelin to stimulate feeding. These data suggest that ghrelin activation of vagal afferent neurons may prevent meal termination, while central ghrelin receptors may initiate food intake.

It has been reported, however, that ghrelin ^−/− mice are of normal body weight and adiposity, and there is no alteration in food intake compared to wild-type mice. However, ghrelin-null mice are resistant to the obesigenic effects of a high-fat diet. Similarly, GHSR1a KO have normal food intake and show no overall change in body weight on chow but, like ghrelin-null mice, are protected from diet-induced obesity, in part due to a decrease in food intake. The location of the receptor mediating these effects is not clear, nor whether gastric ghrelin is playing a role in these phenotypic changes.

62.1.3.2

Endocannabinoids and Lipid Mediators

One of the newer class of mediators generated in the GI tract that alters food intake are lipid mediators, including the endocannabinoids, anandamide, and also oleoethanolamide (OEA). N -acylethanolamines are a group of fatty acids derivatives. Many of these act as agonists at cannabinoid receptors, although there is evidence that these lipid mediators also activate other receptors and ion channels.

There is evidence that lipid mediators can act both within the brain and the GI tract to alter food intake. It is interesting to note that in contrast to other anorexigenic gut mediators such as CCK and PYY that decrease food intake by decreasing meal size, endocannabinoids act via prolonging intermeal interval. Perhaps the best studied of this class of anorexigenic mediators is OEA. OEA is an endogenous fatty acid ethanolamine that is produced in the intestinal mucosa. Levels of synthesis are regulated by feeding, with increased levels in the fed state and a decreased level during a period of fasting. Exogenous administration of OEA inhibits feeding in rodents and also can decrease body weight following long-term administration, at least in rodents.

The evidence suggests that OEA acts via a vagal afferent pathway to inhibit food intake; the decrease in food intake in response to exogenous OEA is abolished by perivagal capsaicin treatment, and exogenous administration induces an increase in neuronal activity as determined by immunochemical localization of fos protein in the nucleus of the solitary tract (NTS), the site of termination of vagal afferent neurons. Moreover, central administration of OEA has no effect on food intake. In short-term culture, OEA can activate vagal afferents and this seems to be mediated via a TRPV1 pathway since it is not present in nodose neurons from TRPV1-null mice and vagal afferent neuronal responses to OEA were temperature-sensitive and blocked by a TRPV1 channel blocker. However, OEA can activate a number of different receptors in addition to TRPV1, including CB1Rs, PPAR, and GPR119, all of which are expressed in the gut and have been shown to be involved in control of feeding.

It is interesting to note that is has been recognized that all effects of intestinal lipid are not mediated by CCK1Rs; infusion of lipid into the small intestine or ingestion of lipid inhibits food intake that is markedly reduced but not abolished be treatment with specific CCK1R antagonists or also in CCK1R-null mice. It is possible that other ligands and receptors, such as lipid mediators and PPAR agonists, may be involved in the pathway by which lipid inhibits food intake. For example, in either CD36 or PPAR-null mice, there is no inhibition of food intake in response to fat. This requires further study but illustrates the complicated nature and possible redundancy of the pathways and mediators involved in intestinal feedback regulation of food intake.