2.1.1

Production and Processing of Peptides by Enteroendocrine Cells

EECs have traditionally been classified and named according to the principal hormones they produce as determined by immunostaining ( Table 2.1 ) with each hormone and cell type exhibiting a characteristic distribution along the length of the GI tract. Gastric epithelium, for example, contains a large number of EECs-producing gastrin, somatostatin (SST), ghrelin, or histamine. Small intestine preferentially generates EECs-producing cholecystokinin (CCK), secretin, glucose-dependent insulinotropic polypeptide (GIP), glucagon-like peptides 1 and 2 (GLP-1, GLP-2), peptide YY (PYY), neurotensin (NT), and serotonin (5-HT). In the colon and rectum, EECs have been shown to secrete serotonin, GLP-1, GLP-2, PYY, NT, SST, and insulin-like peptide-5 (INSL5).

Recent molecular techniques examining EEC subpopulations, labeled with fluorescent reporters driven by hormone specific promoters in transgenic mice, have yielded transcriptomic data at odds with the simple EEC classification suggested by the early immunostaining studies that used only one or two antibodies at a time. At the messenger RNA (mRNA) level, there is a high degree of overlap between different EEC types that were originally thought to be distinct, and many single EECs produce mRNA for a number of different gut hormones. Coproduction of several different hormones in the same EECs has been confirmed by immunostaining. It is now thought that intestinal EECs-producing CCK, GIP, secretin, GLP-1, PYY, and 5-HT form a continuum, with individual cells producing a mix of hormones dependent on their position along the GI tract. Within individual cells, there are conflicting views about whether coexpressed hormones are localized in the same or distinct vesicles, but no convincing evidence has yet been presented to show separate mobilization of different hormones from individual cells.

Peptide hormones are biosynthesized as prepropeptides containing N-terminal signal sequences that direct the growing peptide chain into the lumen of the endoplasmic reticulum during translation. Propeptides transit through the Golgi and are packaged into secretory vesicles where they are cleaved by prohormone convertases (PCs) and further posttranslationally modified by, for example, amidation, sulfation, or acylation. The predominant PC identified in most intestinal EECs is PC1/3, which cleaves propeptides at dibasic residues and is likely responsible for the majority of peptide hormone processing in the small intestine and colon. By contrast, PC2 plays a more prominent role in the stomach. EECs are often identified by immunolabeling with antibodies against chromogranins and secretogranins. These large granin proteins are believed to play a functional role in vesicular packaging but are also subject to PC-mediated cleavage, resulting in the generation of smaller peptides that might themselves play signaling roles.

2.1.2

How Do EECs Respond to Nutrition-Related Stimuli?

It has long been recognized that a wide variety of nutritional and nonnutritional signals trigger gut hormone secretion, with some stimuli preferentially linked to the release of certain hormones. Comparisons between plasma gut hormone concentrations following matched nutrient loads administered orally versus intravenously in humans have revealed that most gut hormones are preferentially released after oral nutrient ingestion. Many additional studies have demonstrated that polymeric macronutrients must be digested into monomers (monosaccharides, free fatty acids, and monoacylglycerides or di/tripeptides and amino acids) before they are capable of triggering gut hormone release.

Transcriptomic analysis and single-cell characterization of fluorescently tagged murine EECs have revealed that they express a range of receptors and transporters capable of detecting a wide variety of stimuli. Even at the single cell level, individual EECs produce machinery capable of detecting multiple stimuli. Unlike taste cells in the tongue, therefore, individual EECs seem to be multimodal rather than tuned to respond to single stimuli. There are two major molecular pathways by which EECs detect ingested nutrients—one involving nutrient transporters and a second involving G-protein-coupled receptors (GPCRs).

Enterocytes typically employ ion-coupled transporters to absorb nutrients across the brush border, using inwardly directed gradients for Na ⁺ or H ⁺ ions to drive the uphill absorption of nutrients. A large body of evidence supports the idea that many EECs have hijacked sodium-coupled glucose transporters (SGLT1) on the apical membrane to act as glucose sensors, as the coupled uptake of Na ⁺ ions with glucose molecules generates an inward current capable of triggering electrical activity, leading to Ca ²⁺ entry through voltage-gated Ca ²⁺ channels and activation of vesicular exocytotic pathways. There is some, albeit weaker, evidence that certain amino acids and di/tripeptides might similarly trigger gut hormone release via their Na ⁺ – and H ⁺ -coupled uptake.

Many small molecules are detected by members of the GPCR superfamily, which include receptors specifically responsive to small molecules including long- and short-chain fatty acids, monoacylglycerides, amino acids, bile acids, and bitter tastants. Nutrient and bile acid responsive GPCRs are highly and specifically expressed in EECs within the intestinal epithelium and likely underlie gut hormone responses to ingested fats and protein, as well as bile acids. GPCRs linked to the stimulation of EECs are mostly G _s and G _q coupled, linked, respectively, to the elevation of cytoplasmic cAMP and Ca ²⁺ concentrations. An increasing body of evidence suggests that coincident activation of different signaling pathways in EECs results in synergistic enhancement of gut hormone secretion.

Rather than merely “tasting” the luminal contents, it is increasingly apparent that EECs respond to the local rates of nutrient absorption. In the case of glucose, the rate of SGLT1-mediated glucose uptake by EECs, and hence the degree of glucose-dependent membrane depolarization, will mirror rates of glucose influx by neighboring enterocytes, being determined by the local concentrations of glucose and Na ⁺ ions. Results from perfused intestinal preparations and Ussing chambers have now shown that EEC receptors for long chain fatty acids and bile acids are functionally located on the basolateral rather than the apical surface of EECs, requiring local absorption across the epithelium prior to receptor activation. Linking gut hormone secretion to local nutrient absorption might ensure that the circulating hormonal signal reflects the rate of nutrient entry into the bloodstream, rather than the mass of unabsorbed nutrients in the lumen that do not yet require the activation of a peripheral homeostatic response.

2.1.3

Pathophysiology Affecting Multiple Enteroendocrine Cell Subtypes

In the sections below, we will describe pathologies primarily affecting specific gut hormones, but there are a few conditions that have more generalized effects on the enteroendocrine system. There have been rare case reports of humans born with an almost complete lack of EECs due to mutations in the transcription factor NeuroG3, which is required for cell differentiation down the EEC pathway. Affected neonates presented with severe malabsorptive diarrhea. Rare human cases have also been reported with homozygous loss of PC1/3 due to mutations in the PCSK1 gene, resulting in a variable presentation that can include malabsorptive diarrhea, impaired glucose homeostasis, and obesity, as well as other endocrinopathies attributable to the global deficiency of many active hormones and peptide neurotransmitters in the gut, pancreas, and central nervous system. Secondary EEC deficiency associated with gastrointestinal symptoms has been described in the autoimmune-polyendocrine-candidiasis-ectodermal-dystrophy (APECED) syndrome, associated with a mutation in the AIRE gene.

Neuroendocrine tumors (NETs) of the GI tract can produce a range of unprocessed, partially processed, and fully processed peptide hormones, with the consequence that clinical presentations vary markedly between cases. In most cases, the exact pattern of active peptides produced by an individual tumor is not currently measurable, because of the lack of suitable methodology for the identification and quantification of partially processed peptides.

Some of the most dramatic gut hormone changes in humans have been observed after upper GI surgical procedures such as Roux-en-Y gastric bypass (RYGB) surgery, gastrectomy, or esophagectomy. RYGB and sleeve gastrectomy are performed routinely as a treatment for morbid obesity, but have dramatic metabolic consequences that result in the resolution of the majority of cases of type 2 diabetes. As discussed in some of the sections below, dramatic postprandial elevations of gut hormones such as GLP-1 and PYY in these patients are likely caused by increased nutrient delivery to and absorption in the more distal small intestine, and almost certainly contribute to observed improvements in glucose tolerance and reduced appetite. Similar hormonal changes have been observed in lean subjects, for example, following resection for gastric cancer, and may contribute to some of the symptoms encompassed under the umbrella of “dumping syndrome.”

2.1.4

Details of Specific Gut Hormones With Known Biological and Pathophysiological Roles

In the sections below, we provide details of the major identified gut hormones produced by EECs, focusing particularly on hormones that have known cognate receptors and functional roles. Peptide sequence nomenclature is based on human sequences, as published in the Uniprot/Swissprot database. The list is not exhaustive and does not include the large number of additional signaling peptides produced by non-EECs types in the gut, such as enteric nerves [e.g., vasoactive intestinal peptide (VIP), gastrin-releasing peptide, galanin], Paneth cells (e.g., defensins), enterocytes (e.g., FGF 15/19), immune cells (e.g., interleukins), and as yet unidentified cell types (e.g., guanylin/uroguanylin).

2.2.1

Synthesis

Gastrin is primarily secreted from gastric antral G cells, but has also been identified in the pituitary gland, developing pancreas and sperm. The gastrin gene encodes a 101 amino acid prepropeptide, containing a 21 amino acid N-terminal signal peptide and 80 amino acid progastrin peptide. All subsequent amino acid position nomenclature refers to the position in the preprogastrin peptide. Following cleavage of the signal peptide in the endoplasmic reticulum, progastrin is sulfated at tyrosine 86 and phosphorylated at serine 96. Further processing in the trans-golgi network and secretory vesicles results in the two mature, C-terminal amidated forms—Gastrin34 (G34) and Gastrin17 (G17). Progastrin is cleaved by PC 1/3 (PC1/3) and carboxypeptidase E (CPE) at amino acid positions 58-59 and 92-93 (the latter removing a C-terminal flanking peptide). The resulting 34 amino acid peptide (G34-Gly) is amidated by peptidyl-glycine α-amidating monooxygenase (PAM), with the glycine group acting as an amide donor. G34 is then cleaved by PC2 to G17, with the two forms present in human G cell vesicles at a G34:G17 ratio of 1:9 ( Fig. 2.1 ).

Amino acid sequence of gastrin prepropeptide and secreted peptides, derived from Uniprot/Swissprot database.

Gastrin shares a significant degree of sequence and structural homology with CCK, and G34 and G17 both act through the CCK2 receptor. Whereas G17 and G34 undergo regulated exocytosis, progastrin, and nonamidated forms of G17 and G34 are secreted via the constitutive pathway. They have no known receptor and have previously been regarded as inactive metabolites, although recent evidence suggests that they may play a role in colonic mucosal proliferation and have a complementary role to that of the amidated gastrins.

2.2.2

Secretion

Gastrin secretion is regulated by neuronal-, hormonal-, and nutrient-responsive factors. Gastrin is secreted in response to luminal amino acids detected via apical calcium-sensing receptors (CaSRs), sympathetic and parasympathetic nervous activity, and gastrin-releasing peptide derived from local neurons. Gastrin secretion is inhibited by SST, when the gastric luminal pH is below 3. Chronic use of proton pump inhibitors results in hypergastrinaemia.

2.2.3

Function

Gastrin’s key role in gastric acid secretion has been demonstrated through gastrin infusion and CCK2R antagonist experiments in man, immunoneutralization in dogs, and in gastrin gene knockout in mice. Gastrin acts on enterochromaffin-like (ECL) cells to stimulate histamine secretion, which then acts in a paracrine fashion via H2 receptors on parietal cells to stimulate acid secretion. Interestingly, in gastrin-deficient mice, the coinfusion of G17 and the nonamidated G17-Gly more potently restored gastric acid secretion than G17 alone. In addition to stimulating acute histamine secretion, the gastrin upregulated the expression of histidine decarboxylase, the enzyme responsible for conversion of histidine to histamine, in ECL cells. Although CCK2 receptors are also present on parietal cells, these appear to be only of limited role for gastrin stimulation of parietal cell acid secretion, with the majority of the effect of gastrin on acid secretion arising due to histamine from ECLs.

Gastrin is not essential for the development and maintenance of the gastric mucosa, but gastrin gene knockout mice had reduced numbers of parietal and ECL cells, which could be restored by infusion of gastrin. It therefore appears that gastrin plays a key role in the differentiation and integrity of the gastric mucosa, although the underlying pathways remain subject to ongoing investigation. One pathway of note involves the urokinase plasminogen activator (uPA) family, including uPA and plasminogen activator inhibitors 1 and 2 (PAI1, PAI2), which localize to gastric parietal and ECL cells.

2.2.4

Clinical Aspects

Zollinger-Ellison syndrome, hypergastrinemia secondary to gastrin secreting NETs, is a cause of gastric acid hypersecretion, multiple peptic ulcers, and secretory diarrhea. This is associated with multiple endocrine neoplasia type 1 (MEN1) in up to 20% of cases.

The role of gastrin in gastric mucosal proliferation is of interest in the pathogenesis and treatment of gastric cancer. Specifically, it has been demonstrated that gastrin stimulates the growth of gastric cancer cell lines in vitro by stimulation of CCK2 receptors, and nonendocrine gastric cancer cell lines can secrete gastrin, which may act in an autocrine fashion. Despite this, any link between hypergastrinemia secondary to proton pump inhibitor therapy and an increased prevalence of gastric adenocarcinoma remains controversial. However, there is a more established link between hypergastrinemia and ECL cell NETs of the stomach, evidence arising from potent H2 receptor blockade in rats using loxtidine and transgenic Men1 / Sst knockout mice treated with omeprazole. Gastric carcinoids in man can be associated with hypergastrinemia due to Zollinger-Ellison syndrome (principally in the presence of multiple endocrine neoplasia type 1 [MEN1]) or atrophic gastritis, but not PPI therapy.

2.3.1

Synthesis

Both SST-14 and SST-28 are products of a single 116 amino acid prepropeptide translated from the SST gene. The prepropeptide consists of a 24 amino acid N-terminal signal peptide and a 92 amino acid propeptide, of which the terminal 14 and 28 amino acids correspond to the active SST peptides ( Fig. 2.2 ).

Amino acid sequence of somatostatin prepropeptide and secreted peptides, derived from Uniprot/Swissprot database.

2.3.2

Secretion

Both 14 and 28 amino acid forms of SST are secreted from gastric and intestinal D cells and pancreatic δ cells, with SST-28 predominating in the small intestine, and SST-14 predominating in the rest of the GI tract and pancreas. Gastric D cells differ between the proximal and distal stomach, with oxyntic D cells exhibiting a closed-type morphology and those in the antrum an open-type morphology. Closed-type oxyntic D cells are inhibited by the vagus nerve soon after food ingestion, thereby reducing the tonic inhibitory control by SST of gastrin and histamine secretion that predominates between meals. SST release from the distal antrum is stimulated by nutrient ingestion, reduced gastric pH, CCK, GIP, GLP-1, acetylcholine, VIP, CGRP, and secretin, resulting in a delayed feedback inhibition of gastric secretions that restores acid secretion to basal levels.

2.3.3

Receptors

There are five G-protein-coupled SST receptors, labeled numerically from 1 to 5, with SSTR2 having two isoforms, SSTR2A and SSTR2B. All SSTRs act through pertussis toxin-sensitive pathways (G _i ) to inhibit adenylate cyclase, activate inwardly rectifying potassium channels, and prevent cellular depolarization, calcium influx, and subsequent vesicle exocytosis. SSTRs also activate other downstream pathways that reduce cellular proliferation through the action of protein tyrosine phosphatases on MAPKs. SST-14 and SST-28 bind with equal affinity to SSTRs 1–4, but SST-28 has a 10–30 fold higher affinity for SSTR5 than other SSTRs, whereas SST-14 has reduced affinity at SSTR5.

2.3.4

Function

2.3.5

Clinical Aspects

SST analogues are of considerable utility in the diagnosis and treatment of gastroentero-pancreatic NETs. As many moderately and well-differentiated NETs express receptors to SST, radio-nucleotide labeled SST analogues can be used in the diagnosis and staging of disease and for targeted radiotherapy. Palliative treatment with SST analogues, in the presence of symptomatic NETs, can control hormone-mediated symptoms including diarrhea, tachycardia, and flushing and has recently been shown to delay tumor progression.

Other GI uses of SST analogues are based on limited case series or expert opinion and utilize their counterregulatory and antisecretory effects. The evidence is at present equivocal on the benefits of SST analogues in the prevention of postpancreatectomy cutaneous fistula, or the treatment of enterocutaneous fistula. Long- and short-acting SST analogues have also been used for the management of congenital hyperinsulinemia and reactive hypoglycemia and accelerated intestinal transit after upper GI surgery, with mixed success.

2.4.1

Synthesis

Ghrelin is primarily secreted by X/A-like cells of the stomach, but has also been identified in other tissues including duodenum, pancreas, lymphocytes, and the central nervous system. Following total gastrectomy, circulating total and acyl ghrelin concentrations are undetectable, suggesting that the extra-gastric sources do not contribute significantly to circulating levels. It is encoded by the GHRL gene, located on chromosome 3p25-26. Translation of Ghrl mRNA produces a 117 amino acid preprohormone (preproghrelin), which is cleaved to the active 28 amino acid ghrelin by PC1/3. Ghrelin is modified by the addition of an octanoyl moiety to the hydroxyl group of the serine at position 3 of proghrelin catalyzed by ghrelin O -acyltransferase (GOAT, also membrane bound O -acyltransferase, MBOAT4), although it is unclear whether this step precedes or follows cleavage of proghrelin to ghrelin. Acylation of serine 3 is essential for activity at the GHSR1a receptor, and acyl-ghrelin has historically been regarded as the active, and des-acyl-ghrelin the inactive, form of the peptide, although independent functions have been considered for the latter. The fatty acid chain used for ghrelin acylation appears to derive from the diet.

2.4.2

Secretion

Circulating concentrations of ghrelin are highest in the fasting state, with secretion suppressed by glucose and fat ingestion, and exercise, but less so by protein intake or gastric distension. In vitro experimental evidence exists for direct sensing of fatty acids, glucose, and glutamate by X/A-like cells, for suppression of ghrelin secretion by insulin, leptin, and GLP-1 and for stimulation of ghrelin secretion by glucagon. However, X/A cells are predominantly closed-type EECs making no contact with the gastric lumen, so they are likely regulated primarily by internal factors. Pharmacological experiments in rats demonstrated increased ghrelin secretion in response to muscarinic and beta-adrenergic activity and decreased secretion in response to alpha-adrenergic activity. Plasma ghrelin concentrations increased in healthy humans in response to a cholinergic agonist and were suppressed by a muscarinic antagonist. Vagotomy initially suppressed ghrelin secretion in rats, but seven days postvagotomy plasma ghrelin concentrations were elevated. GLP-1 and PYY, independently and synergistically, suppressed ghrelin secretion in a study of 25 overweight men. Investigation of FACS purified ghrelin secreting cells from mice demonstrated G-protein-coupled actions of α-CGRP, long- and short-chain fatty acids, lactate, SST, GIP, and α-MSH, but interestingly not PYY or GLP-1 on ghrelin secretion.

2.4.3

Function

The majority of experimental studies on ghrelin have focused on the actions of its acyl form on the cognate G-protein-coupled receptor GHSR1a. Des-acyl-ghrelin is presumed to be an inactive metabolite, as it has no identified receptor or major physiological effects.

Ghrelin plays a significant role in appetite regulation, stimulating neuropeptide Y/Agouti-related peptide (NPY/AgRP) neurons within the arcuate nucleus of the hypothalamus, in a counterregulatory fashion to leptin, to stimulate hunger and initiate feeding. Multiple animal and human studies disagree on the relative importance of vagal-mediated ghrelin signaling versus direct central nervous system action of circulating ghrelin. In vivo animal models, and infusion experiments involving human participants who had undergone vagotomy, suggest that an intact vagus nerve is essential for the meal initiation effects of ghrelin. This is consistent with the finding that ghrelin increased the sensitivity of gastric vagal mechanosensory nerves to stretch. However, it has been demonstrated that central nervous system and intraperitoneal administration of ghrelin in rats resulted in similar feeding behavior, and a randomized controlled trial of ghrelin administration in patients after total gastrectomy (i.e., with minimal circulating ghrelin and a truncal vagotomy) improved food intake and reduced body weight loss.

Ghrelin stimulates pituitary growth hormone secretion through its direct actions on GHSR1a. In particular, it seems that ghrelin is a key stimulus of growth hormone-mediated gluconeogenesis in the fasting state. In GOAT or ghrelin knockout mice, prolonged fasting resulted in profound hypoglycaemia, associated with a reduced growth hormone response, which was reversed on infusion of acyl-ghrelin.

Ghrelin also influences glucose homeostasis in a growth hormone-independent fashion. Acyl-ghrelin indirectly inhibited glucose-mediated insulin and glucagon secretion in mouse models by stimulation of SST secretion from pancreatic islet delta cells. Although a population of pancreatic islet ghrelin secreting cells has been described, supporting the possibility that ghrelin might also act in a paracrine fashion within pancreatic islets, the physiological importance of this finding remains controversial. Examination of peripheral insulin sensitivity using an euglycaemic hyperinsulinemic clamp with and without exogenous ghrelin, in hypopituitary patients on stable doses of exogenous growth hormone, demonstrated that ghrelin increased peripheral insulin resistance independent of growth hormone. Both acylated and nonacylated forms of ghrelin have been demonstrated to stimulate fat accumulation in human visceral adipocytes, through enhanced PPARgamma and SREBP1 signaling.

The more direct roles of ghrelin in GI function include stimulation of gastric motility and increasing gastric acid secretion in a vagus and 5-HT-dependent fashion.

2.4.4

Clinical Aspects

The diverse and as yet not completely understood roles of ghrelin in appetite regulation, glucose and energy homeostasis, GI motility, and higher-order cognitive functioning make it a fertile area for ongoing research into disease pathology and pharmacotherapy.

The orexigenic actions of ghrelin, an inverse correlation between fasting ghrelin concentrations and BMI and observations of reduced postprandial suppression in obese humans make it an attractive target for the treatment of obesity. Plasma ghrelin concentrations are reduced in parallel with weight loss after sleeve gastrectomy, but variably increase after Roux-en-Y gastric bypass, suggesting it is not the primary regulator of appetite and body mass in humans. Animal knockout and pharmacological models targeting ghrelin, GOAT, and GHSR1 have resulted in conflicting results regarding metabolic homeostasis and protection from diet-induced obesity, with the most convincing results showing a reduced incidence of diet-induced obesity in Ghsr1 ^−/− mice. One phase I/IIa trial of an antighrelin vaccine was halted due to lack of efficacy.

Conversely, human studies of pharmacological augmentation of the ghrelin axis have proved more fruitful. Partial and total gastrectomy and esophagectomy result in reduced plasma ghrelin concentrations and simultaneous severing of afferent vagal fibers. This is often associated with reduced appetite, weight loss, and impaired quality of life. One small trial of synthetic ghrelin in postgastrectomy patients yielded positive results on food intake. GHSR1a agonists are also in late-phase clinical trials for cancer-related cachexia, with promising early results. It has been proposed that the hyperphagia of Prader-Willi syndrome may be mediated by hyperghrelinemia, although recent evidence suggests the onset of elevated plasma ghrelin concentrations significantly predates hyperphagia and may be unrelated to the phenotype.

The prokinetic effects of ghrelin offer a potential drug target for GI motility disorders. One small study has shown improved gastric emptying following the administration of a GHSR agonist in patients with diabetic gastroparesis, and the drug has entered phase 3 trials.

2.5.1

Synthesis

The discovery of CCK dates back to the suggestion of a hormonal mechanism for gallbladder contraction in 1928, supported by an experiment wherein intestinal mucosal extracts were infused into dogs, cats, and guinea pigs and resulted in gallbladder contraction. The peptide sequence of CCK was first described in 1968, with the C-terminal pentapeptide Gly-Trp-Met-Thr-Phe conserved across all CCK and gastrin peptides.

The CCK gene is located at chromosome 3p22.1. PreproCCK is a 115 amino acid protein, with an N-terminal signal sequence and spacer sequence followed by the bioactive domain. Posttranslational processing involves cleavage of the signal peptide and addition of a C-terminal amide group, followed by cleavage of the 83 amino acid peptides at basic amino acid residues, most likely by PC1/3, although PC2 and PC5 have been implicated in the processing of CCK in rat brain. Addition of a sulfate group to the tyrosine residue seven amino acids from the C-terminus confers activity at the CCK 1 receptor (CCK1R), whereas sulfated and nonsulfated peptides are equally active at the CCK 2 receptor (CCK2R), which also acts as the gastrin receptor.

Measurement of CCK concentrations in plasma is challenging, with acknowledged discrepancies in the sensitivity of immunoassays to different CCK peptides, and no gold standard test. High-pressure liquid chromatography extraction of human intestinal lysates, and of plasma after a meal test, revealed the 58 amino acid peptide to be the most abundant in man, with minor amounts of the 39 and 8 amino acid peptides. These results must, however, be interpreted in light of the finding that the in vivo half-life and hepatic clearance of CCK-8 is markedly faster than that of CCK-58. It is probably reasonable to regard current CCK assays as satisfactory for the examination of relative concentrations of plasma total CCK, but given the variability in the rate of degradation and assay sensitivity between peptides, caution must be exercised when interpreting data attempting to assess the relative concentrations of different length CCK peptides.

It is acknowledged that CCK is widely distributed within the brain, with CCK-8 and CCK-5 being the primary neurotransmitter CCK peptides. In depth discussion of the role of CCK as a central neurotransmitter is beyond the remit of this chapter.

2.5.2

Secretion

CCK is secreted into the circulation from open-type I cells in the duodenal and jejunal mucosa. Plasma concentrations of total CCK rise approximately three- to sevenfold in response to a mixed meal. While it is clear that CCK is secreted in a nutrient specific fashion in man, our knowledge of the underlying receptors is heavily reliant on limited pharmacological experiments and animal and in vitro data. Lipids are the most potent stimulus of CCK secretion, followed by proteins, with only small effects triggered by intraduodenal carbohydrates. Intraduodenal lipid stimulation of CCK secretion is dependent upon medium and long-chain fatty acids acting via FFA1 (GPR40), FFA4 (GPR120), and possibly CD36, with limited effects of short-chain fatty acids on plasma CCK concentration. The mechanisms by which digested proteins stimulate CCK secretion include activation of PEPT1 and the calcium-sensing receptor (CaSR). It appears that carbohydrates play a more limited role in CCK secretion. In one small human study, intraduodenal acidification in the absence of nutrients did not stimulate CCK secretion.

It remains unclear to what extent vagal tone influences CCK secretion, if at all. Studies in vagotomized humans have either failed to take account of altered gastric transit, or found conflicting results, and one limited study in rats suggested that vagal activation could induce CCK secretion, but did not explore this in a physiological fashion.

2.5.3

Function

CCK acts via the G-protein-coupled receptors CCK1R and CCK2R (previously CCKAR and CCKBR). CCK1R has a 500–1000-fold greater affinity for sulfated than nonsulfated CCK peptides, whereas CCK2R will bind both sulfated and nonsulfated CCK, but at a lower affinity than it binds gastrin, which is regarded as the cognate ligand for peripheral CCK2R. Both receptors are primarily G _q coupled and exhibit transcriptional modulation through mitogen-activated protein kinase cascades, and there is some evidence for G _s activity downstream of CCK1R.

CCK acts to promote digestion by delaying gastric emptying, stimulating pancreatic exocrine secretion and gallbladder contraction and stimulating satiety.

The combined physiological actions of CCK and gastrin on CCK2R require the function of both hormones to be considered in tandem. In the stomach, CCK2R is localized in parietal and ECL cells, whereas CCK1R is found in gastric D cells. Gastrin/CCK2R activity stimulates gastric acid secretion indirectly by stimulating histamine release from ECL cells, as well as being a key trophic factor in the maintenance of parietal and ECL cell number and function. Evidence from CCK1R knockout rats suggests that CCK, through CCK1R-mediated stimulation of SST secretion from gastric D cells, also acts in a feedback fashion to reduce gastrin secretion and thus gastric acidity.

Beyond its effects on gastric pH, a randomized crossover study in humans demonstrated that CCK delayed gastric emptying, and also promoted pancreatic secretion of lipase and trypsin in a fashion that was abolished by the CCK1R antagonist loxiglumide. Human gastric smooth muscle demonstrated CCK1R immunoreactivity, although the distribution and activity of CCK receptors in gastric musculature appear to be variable across species.

Studies on rodent pancreatic acinar cells demonstrated direct CCK1R-dependent actions of CCK on enzyme secretion, but there has been considerable controversy about the importance of direct CCK signaling in the human pancreas because in vitro secretion studies of isolated human acinar cells did not demonstrate stimulatory effects of directly applied CCK or CCK1/2R agonists. Until recently, there was therefore a general agreement that the effects of physiological concentrations of CCK on pancreatic exocrine secretion are primarily mediated via CCK1 receptors on vagal afferent fibers, acting via the nucleus tractus solitarius (NTS) and pancreatic vagal efferent fibers. However, a recent study of human pancreatic slice preparations demonstrated direct stimulatory effects of physiological concentrations of CCK8 on acinar cell secretion mediated by CCK1 receptors, and it is therefore likely that direct CCK1R-dependent acinar cell activation contributes to the pancreatic secretory response to CCK in humans as well as rodents.

CCK mediates gallbladder contraction both directly through smooth muscle located CCK1 receptors and indirectly via a vagal circuit. Administration of exogenous CCK caused equal contraction of the gallbladder in vagotomized humans and healthy controls. However, coadministration of atropine or pirenzipine in man abolished the effects of CCK. It seems likely that physiological concentrations of CCK mediate gallbladder contraction via a vagal circuit; however, in the presence of vagotomy, or supraphysiological concentrations, CCK can act directly on the gallbladder.

CCK-mediated sphincter of Oddi relaxation has been described using endoscopic retrograde cholangio-pancreatography (ERCP) in humans, and CCK1 receptors have been identified on the associated smooth muscle, implying at least an endocrine mode of action. It has, however, not been disproven that sphincter of Oddi relaxation is, like gastric and gallbladder smooth muscle, under the control of a vagal circuit mediated by CCK1 receptors on vagal afferent fibers.

The effects of CCK to suppress hunger and stimulate satiety have been well described in humans and animal models. Administration of exogenous CCK in man reduced meal size, and CCK1R blockade with loxiglumide had the opposite effect, as well as reducing satiety during a meal or in response to duodenal lipids. Animal models, including vagotomy, central and peripheral CCK1R blockade, and sham feeding, suggest that both peripheral vagally mediated and central endocrine actions of CCK are responsible for its satiety signaling. No studies have conclusively examined the relative role of central versus peripheral CCK activity in man, although it seems likely that as in animal experiments, CCK1 receptors on the afferent vagus and central CCK1 receptors both act to mediate satiety.

CCK was found to reduce hepatic gluconeogenesis via a vagal circuit in rats. Infusion and antagonist studies in human have, however, failed to demonstrate a significant role for CCK in insulin secretion or glucose metabolism.

2.5.4

Clinical Aspects

The compelling argument that variations in CCK signaling mediate gallstone formation through altered gallbladder contractility has been revisited in recent years. A variant in the CCK1R gene has been linked to reduced CCK1R expression in patients with gallstone disease, and CCK1R knockout or antagonism in mice-induced gallstone formation.

It has been proposed that CCK2 receptors are overexpressed in GI malignancies and may mediate cell proliferation, potentially offering a novel therapeutic target; however, experimental evidence is at present conflicting. The possibility of proton pump inhibitor-mediated hypergastrinemia as a contributing factor to neoplastic progression in Barretťs oesophagus is supported by cell line and organoid studies.

Despite its role in satiety signaling, there is at present no evidence that variations in CCK signaling are a cause of obesity.

2.6.1

Synthesis

In the human, secretin is primarily secreted from S cells in the duodenum and proximal small intestine. The human secretin precursor consists of a signal peptide, a short amino terminal peptide, and the 27 amino acid active peptide, which retains an amide group from the adjacent C-terminal glycine.

The secretin receptor is a G-protein-coupled receptor, which has been isolated from pancreas, stomach, cholangiocytes, bronchi, and brain. It acts primarily via G _s -coupled pathways to elevate intracellular cAMP, with a small effect on IP3 production via G _q pathways.

2.6.2

Secretion

Secretin is released into the circulation in response to delivery of fatty acids or acidic contents to the duodenal and jejunal lumen, although there is a paucity of literature on the cellular mechanisms coupling luminal pH to secretion. Vagal tone appears to play no part in secretin secretion.

2.6.3

Function

Secretin stimulates duodenal Brunner’s gland and pancreatic bicarbonate secretion in response to duodenal acidification. Administration of physiological doses of secretin to humans, with endoscopic duodenal or pancreatic ductal cannulation and sampling, demonstrated a dose-dependent increase in pancreatic bicarbonate secretion. Abolition of the secretin response to a protein meal in dogs using antisecretin serum reduced pancreatic bicarbonate secretion to <20% of control values. Interestingly, infusion of varying doses of secretin and CCK to dogs with gastrostomies and Herrera fistulas demonstrated a synergistic effect of the two hormones. In humans, secretin induces relaxation of the sphincter of Oddi.

Using in vitro cholangiocytes, ex vivo murine bile ducts, and ¹⁴ C-labeled mannitol in humans with T-tube drainage of the bile duct, secretin has been demonstrated to stimulate cholangiocyte bicarbonate secretion and bile flow. Evidence is also accumulating that secretin acts to stimulate proliferation of cholangiocytes, with murine models of cholestasis demonstrating reduced cholangiocyte proliferation in secretin knockout strains, and enhanced bile duct fibrosis on administration of exogenous secretin.

Secretin appears to act in a feedback fashion to inhibit gastric acid secretion. The administration of physiological doses of secretin to dogs and humans reduced pentagastrin and meal-induced gastric acid secretion. Dogs treated with secretin antiserum showed increased gastrin and gastric acid secretion in response to a meal. Based on a rat vagotomy model, it was proposed that secretin-mediated inhibition of gastrin-induced gastric acid secretion is dependent upon vagal pathways. The effect of secretin on gastric acid secretion appears to be potentiated by 5-HT, with the effect of both agents in combination being greater than the sum of their individual actions.

The literature on the effect of secretin on gastrin secretion is contradictory, likely representing the complex interaction of multiple neural and humoral circuits in the control of gastric pH. Early studies involving infusion of secretin into humans and dogs reported reduced plasma basal gastrin concentrations, yet a marked rise was observed in patients with Zollinger-Ellison syndrome. Later studies suggested that even in healthy humans, secretin slightly increases plasma gastrin concentrations.

Human and canine experiments have clearly demonstrated that secretin acts to delay gastric emptying and increase pyloric tone, although information on the exact causative pathways is lacking.

It has recently been argued that secretin, in common with multiple other gut hormones, has a central anorectic role. The administration of secretin to wild type, but not secretin receptor knockout mice, reduced food intake, which could be abolished by pretreatment with a MC4R antagonist. Vagotomy also blocked the anorectic effects of secretin, and the finding of secretin-induced c-Fos activation in POMC neurons of the arcuate nucleus points to a role of secretin on hypothalamic feeding circuits.

Secretin and its receptor are present in the hypothalamus and pituitary and play a role in salt-water homeostasis as a mediator of the angiotensin/aldosterone system, with secretin knockout mice eliciting a defective aldosterone response to a low sodium diet. It has also been reported to be secreted by the posterior pituitary in response to plasma hyperosmolarity, stimulating vasopressin secretion.

2.6.4

Clinical Aspects

Secretin is of use in the assessment of stimulated exocrine pancreatic function, in tandem with either imaging or endoscopy, and to identify gastrinomas.

2.7.1

Secretion

In keeping with its role in the control of gastroduodenal motility, the secretion and action of motilin are interdependent upon meal ingestion, vagal tone, and gastric and duodenal pH. There are very few published studies, however, investigating the stimuli regulating motilin secretion, likely because of the lack of motilin production by standard laboratory rodents. Ingestion or intravenous infusion of lipids was reported to stimulate motilin secretion in humans, whereas oral or intravenous carbohydrate and intravenous amino acids suppressed plasma motilin levels. These results suggest that motilin secretion is regulated more by circulating than luminal nutrient levels. Motilin secretion was also stimulated by acidification, and inhibited by alkalinization, of the duodenal contents. There are conflicting reports on the effect of truncal vagotomy on baseline and stimulated plasma concentration of motilin; however, it seems most likely that the vagus plays a role in the downstream effects of motilin, but not its secretion.

2.7.2

Function

The primary function of motilin seems to be to stimulate GI motility by inducing phase 3 MMC contractions. This is evidenced both by the correlation of increasing plasma motilin concentrations with the onset of phase 3 MMC, and the finding that exogenous motilin, and erythromycin (a motilin receptor agonist) induce phase 3 MMC contractions in humans. This action appears pH- and vagus-dependent—antral or duodenal acidification impaired the stimulatory action of motilin on phase 3 MMCs, an effect that was reduced by vagotomy. Motilin-induced phase 3 MMC contractions were abolished by vagal activity induced by true or sham eating in a shrew model. It is, however, difficult to dissociate the direct efferent effects of the vagus on gastric motility from any vagus-mediated motilin effects in these models. Muscarinic acetyl-choline receptor blockade with atropine abolished erythromycin-induced phase 3 gastric MMCs in humans, suggesting a cholinergic basis for this activity, but this could be mediated by direct myenteric plexus motilin sensing, or a circuit involving the vagus.

Interestingly, it appears that phase 3 MMC contractions and plasma motilin concentrations coincide with times of peak hunger during interdigestive periods, suggesting an orexigenic role for phase 3 MMCs and/or motilin. Erythromycin also induces hunger in healthy volunteers, and the association between hunger, motilin, and phase 3 MMC is dependent upon cholinergic pathways.

2.7.3

Clinical Aspects

The motilin receptor is of interest as a drug target for GI motility disorders including diabetic gastroparesis and critical illness-related malabsorption. Its apparent role in appetite and feeding is also of potential interest in the management of obesity. There is at present no suggestion that disorders of motilin secretion or signaling exist as clinical entities.

2.8.1

Synthesis

The NTS gene, located on chromosome 12, encodes a 170 amino acid prepropeptide, with a 23 amino acid N-terminal signal peptide. Numbering from the first amino acid of the signal peptide, amino acids 24-148 encode “large” Neuromedin N, which contains the five amino acids of Neuromedin N at residues 144-148. NT is sited at amino acids 151-163, and there is a short tail peptide. Posttranslational processing is by PCs at dibasic residues flanking the active peptide sequences.

While originally described as located in NT-secreting N cells in the ileum, recent evidence of pluripotent EECs, showing plasticity of hormone secretion along the GI tract, has demonstrated NT to be colocated and secreted with GLP-1, PYY, and other gut hormones.

2.8.2

Secretion

Luminal fatty acids and alcohols are the most potent secretagogues of NT, with limited roles for saline, glucose, and amino acids in human and canine experiments. Fat-induced NT secretion has been reported to reduce after proximal gastric vagotomy in man, and in healthy volunteers plasma NT concentrations correlated with the volume of oleic acid delivered to the duodenum. Resection of the distal 2/3 of the small intestine in dogs abolished fat-induced NT secretion, implying the majority of NT secreting cells are located in the ileum. Interestingly, studies on isolated perfused rat intestine demonstrated a role for SGLT1 and GLUT2-mediated glucose uptake in NT secretion.

2.8.3

Function

Three NT receptors have been isolated: NTR1 and NTR2 are GPCRs, NTR3 is a single transmembrane receptor with full homology to gp95/sortilin. The exact downstream transduction mechanism of the NT receptors has yet to be adequately described, with conflicting publications suggesting G _i , G _q , and G _s responses in different cellular models, and the possibility that NT acts as an inverse agonist or antagonist at NTR2.

Initial experiments on NT focused on its response to luminal fats and potential as an enterogastrone (i.e., intestinal regulator of gastric acid secretion). An elegant experiment matching plasma NT concentrations during infusion to meal-induced plasma NT concentrations, however, clearly showed no effect of physiological concentrations of NT on peptone-induced gastric acid secretion.

Infusion of supraphysiological concentrations of NT to healthy human volunteers abolished or delayed intestinal MMCs, and in dogs impaired gastroduodenal motility. However, animal models using more physiological concentrations of NT have demonstrated a more nuanced picture, although applying NT to murine interstitial cells of Cajal in vitro did modulate pacemaker electrical activity, acting via NTR1. Rats given NT infusions evidenced delayed intestinal transit in a meal- and vagus-dependent fashion, suggesting that NT may act in synergy with other factors in the control of GI motility. Interestingly, giving NT to humans induced defecation and may point to a differential effect of NT on proximal and distal GI motility.

Exogenous NT at physiological concentrations stimulated pancreatic exocrine secretion in humans, measured by duodenal cannulas, and in dogs potentiated the pancreatic response to duodenal amino acids and hydrochloric acid. Physiological concentrations of NT, and oleic acid-induced NT secretion, had no effect on jejunal fluid secretion/absorption in dogs.

Interestingly, NT may play a role in glucose homeostasis, as it was found to act on isolated rat islets or beta cell cultures to promote insulin secretion at low glucose concentrations and inhibit insulin secretion in high glucose conditions. Infusion experiments in man have, however, demonstrated no effect on glucose homeostasis, in contrast to early rat experiments in which NT-induced hypoinsulinemia and hyperglycemia. . NT was also reported to have an antiapoptotic effect on cultured pancreatic islets, although the underlying mechanisms and physiological relevance are unclear.

2.8.4

Clinical Aspects

The effects of NT on energy homeostasis and central control of appetite are the subject of ongoing study. NT knockout mice were resistant to diet-induced obesity, possibly through impaired intestinal fat absorption. In humans, a longitudinal cohort study showed that healthy volunteers with high fasting plasma pro-NT were at greater risk of developing obesity. In a series of rat experiments, including peripheral administration of NT and NT receptor antagonists, and gastric bypass surgery, circulating NT was demonstrated to reduce food intake, acting via the vagus and directly on the brain, possibly affecting proopiomelanocortin pathways.

2.9.1

Synthesis

The human GIP gene is located on chromosome 17 and encodes a 153 amino acid preprohormone including a 20 amino acid N-terminal sequence peptide. GIP is secreted as a 42 amino acid peptide corresponding to residues 52-93 of the preprohormone. GIP secreting K cells are predominantly located in the duodenum and proximal small intestine, with few GIP cells found in the ileum or colon.

2.9.2

Secretion

GIP secretion is strongly linked to nutrient ingestion, with levels rising ~10-fold within 15 min of food intake. GIP-secreting cells express a variety of nutrient receptors, and GIP release in vivo and in vitro has been demonstrated in response to carbohydrates, amino acids, and fatty acids.

2.9.3

Function

GIP stimulates insulin secretion in the presence of normal and elevated plasma glucose concentrations, but not during periods of hypoglycemia and is a major contributor to the incretin effect in man. The term “incretin effect” describes the finding that insulin rises to a greater extent in response to an oral glucose load than to an intravenous simulation of similar plasma glucose levels and can be fully explained by two intestinally secreted mediators (i.e., GLP-1 and GIP) that potentiate glucose-stimulated insulin secretion. In contrast to GLP-1, GIP also enhances glucagon secretion. Ablation of GIP-producing K-cells or knockout of the GIP receptor (GIPR) had only small effects on glucose tolerance in mice, but the relative importance of GIP in humans cannot be inferred from these results because postprandial glucose profiles reflect a complex interplay between food composition and the relative rates of gastric emptying, digestion/absorption, insulin secretion, and glucose disposal. Studies in humans have estimated that the combined actions of GIP and GLP-1 are responsible for approximately 50% of meal-triggered insulin secretion.

The GIP receptor is a G _s -coupled receptor, acting via elevation of cytoplasmic cyclic AMP concentrations. Its glucose-dependent stimulatory activity on pancreatic β-cells arises because cAMP is a poor stimulus of insulin secretion unless intracellular Ca ²⁺ concentrations are simultaneously elevated by a depolarizing stimulus such as glucose. In vitro studies using rodent islets have also suggested a beta cell protective effect of both GIP and GLP-1. Acting via PKA/CREB/TORC2 and EPAC2/PKB/Akt pathways, it is proposed that beta cell apoptosis is inhibited and cell proliferation promoted. There is, however, limited data supporting an in vivo effect of GIP on beta cell preservation in rat models of type 1 and type 2 diabetes, and proliferative effects observed in rodent islets have not been generally reproducible using human islets.

Conflicting studies have suggested that GIP influences lipid and adipose tissue metabolism. GIP receptor null mice are resistant to the obesogenic effects of a high fat diet, with reduced adipocyte deposition of triglycerides but maintained insulin secretion. This is in contrast to an acute GIP cell depletion mouse model, in which no difference in lipid metabolism was demonstrated. Studies in human volunteers have been equally conflicting, with physiological and supraphysiological increases in GIP concentration resulting in no change to circulating triglyceride concentrations. GIP infusion in man did, however, result in increased adipose tissue blood flow, and fasting plasma GIP concentrations have been shown to correlate inversely with serum low-density lipoprotein concentrations. A recent study demonstrated reduced adipose tissue blood flow, triglyceride and glucose uptake when a GIPR antagonist was infused during a hyperglycemic hyperinsulinemic clamp in healthy human volunteers.

It is also suspected that GIP is a mediator of the daily fluctuations in bone turnover. In vitro studies have demonstrated that GIP stimulates osteoblast and inhibits osteoclast activity, with GIP receptor null mice demonstrating accelerated reduction in bone mineral density and strength associated with increased markers of bone turnover. GIP infusion studies in humans have demonstrated a GIP-related reduction in markers of bone turnover, which appeared to be larger than any secondary effect of increased insulin concentrations, and a GIP receptor missense mutation has been associated with an increased fracture risk in postmenopausal women. While this field is at an early stage, it seems reasonable to conclude that gut-derived signaling, via GIP, is implicated in calcium homeostasis and bone turnover.

Whether GIP plays a role in the central control of appetite remains unclear. GIP receptor null mice were resistant to diet-induced obesity, and even in the absence of leptin, GIP receptor null mice were leaner than ob/ob-only mice, but these phenotypes may be due to increased fat utilization and energy expenditure rather than changes in appetite. In support of this idea, GIP infusions in human volunteers did not reduce appetite or energy intake, in contrast to the clear effects of other GI hormones on eating behavior.

2.9.4

Clinical Aspects

At present, there is little published information on the pathological relevance of GIP. It may play a role in the pathophysiology of type 2 diabetes, as in affected patients, the beta cell response to physiological and pharmacological concentrations of GIP is impaired, although causality remains to be established and beta cell responsiveness to GIP is also reduced indirectly by hyperglycemia. A GIP receptor variant in humans was associated with 2-hour plasma glucose concentrations during an oral glucose tolerance test, but this likely reflects the physiological role of GIP rather than a step on the pathway to type 2 diabetes.

While novel GIP/GLP-1/glucagon multireceptor agonists are in early trials for the treatment of diabetes and obesity, the lack of a potent GIP receptor antagonist has hampered studies into its pathophysiological role in humans. The recent development of GIP _(3-30) , however, promises to shed light in this field.

2.10.1

Synthesis

The human GCG gene is located on chromosome 2 and encodes a 180 amino acid preprohormone, including a 20 amino acid N terminal signaling peptide.

Posttranslational processing of proglucagon in the L-cell is mediated by PC1/3 and generates glucagon-like peptide 1 (GLP-1, amino acids 98-128), glucagon-like peptide 2 (GLP-2, 146-178), glicentin (21-89), and oxyntomodulin (53-89) . Posttranslational processing of proglucagon in the alpha cell is mediated by PC2 and generates glucagon (amino acids 53-81), glicentin-related pancreatic peptide (21-50), and the biologically inactive major proglucagon fragment (92-178) containing the sequences for GLP-1 and GLP-2.

This section will focus on the intestine-derived proglucagon peptides—GLP-1, GLP-2, oxyntomodulin, and glicentin. Pancreatic glucagon secretion is a topic of significant interest, with potential pathological and therapeutic roles in diabetes mellitus and energy balance, but is beyond the scope of this chapter. It remains unclear as to whether pancreatic glucagon (53-81) is additionally produced from the GI tract. Significant assay cross-reactivity between glucagon (53-81), oxyntomodulin, and glicentin make plasma measurements uncertain. While one study detected PC2 in cultured canine ileal L-cells these did not produce significant amounts of glucagon. However, mass spectroscopic analysis of pooled plasma after an oral glucose tolerance test in patients after total pancreatectomy has hinted at the presence of some gut-derived glucagon _(53-81) in man.

2.10.2

Secretion

Proglucagon-producing intestinal L-cells are open-type EECs, located predominantly in the mid to lower small intestine and colon. The proglucagon gene products (GLP-1, GLP-2, oxyntomodulin, glicentin) are cosecreted in response to a variety of possible stimuli. There is good experimental evidence supporting the following substances having a role in L cell secretion: glucose, bile acids, amino acids, and short-, medium-, and long-chain fatty acids. The concentrations of circulating GLP-1, GLP-2, oxyntomodulin, and glicentin are related to the rates of delivery and absorption of nutrients at different regions of the small and large intestines and the corresponding distribution of L-cells.

2.10.3

GLP-1

GLP-1 was originally predicted to be a 37 amino acid peptide, based on the presence of a classic dibasic PC cleavage site at amino acid position 91 on the proglucagon propeptide. Synthetic GLP-1 _(1-37) is, however, biologically inert, and only truncated versions of GLP-1, named GLP-1 _(7-37) and GLP-1 _{(7-36 amide)} , have been isolated from the intestines of multiple species and shown to be metabolically active. The C-terminal glycine of GLP-1 _(7-37) acts as a substrate for amidation to form GLP-1 _{(7-36 amide)} , and while this is the predominant secreted form in humans, significant circulating concentrations of the 7-37 peptide are found in other species. GLP-1 _(7-37) and GLP-1 _{(7-36 amide)} have similar high activity at the GLP-1 receptor (GLP1R).

GLP-1 elicits a range of GI, metabolic, and neurological effects, which can be regarded as facilitating nutrient digestion, absorption, and processing.

The best-studied function of GLP-1 is as an incretin hormone, acting to potentiate glucose-dependent insulin secretion from pancreatic beta cells. To act as an incretin, GLP-1 secretion must be linked to the presence and/or absorption of glucose from the intestine, easily achievable when GLP-1 is released from intestinal L-cells, but less obvious if GLP-1 is derived from intrapancreatic sources, as recently suggested by transgenic mouse studies. These studies however also resulted in impaired Gcg expression levels in the proximal intestine.

The GLP-1 receptor is a G-protein-coupled receptor, first cloned in 1992. GLP1R potentiates insulin secretion through a G _s -coupled pathway, increasing cAMP concentrations and potentiating exocytosis of insulin-containing vesicles in the presence of an elevated cytoplasmic Ca ²⁺ concentration. Ca ²⁺ may also be increased downstream of GLP1R activation through the release of intracellular calcium stores, mediated via phospholipase C- and IP ₃ -mediated pathways. Unlike GIP, GLP-1 also suppresses pancreatic alpha cell glucagon secretion, although the responsible mechanism remains a topic of debate, and likely involves the paracrine action of delta cell SST.

GLP-1 induces the sensation of satiety and suppresses food intake. The mechanism by which GLP-1 exerts this effect remains controversial, but likely includes a combination of stimulation of vagal afferents and direct central sensing of circulating GLP-1 by the brainstem or hypothalamus. Intracerebral infusion of GLP-1, Exendin 9-39 (a GLP-1 receptor antagonist), or both to mice demonstrated that central GLP-1 profoundly inhibits food intake. In rats, an intact vagus appears essential to the satiating effects of intraperitoneal, but not intravenous GLP-1. Human infusion studies, using cognitive tools and functional imaging, have clearly demonstrated that peripherally derived GLP-1 induces satiety. At least some of the central actions of GLP-1 on appetite seem to be mediated through the proopiomelanocortin (POMC)—neuropeptide Y (NPY)/agouti-related peptide (AGRP) neuronal circuit in the hypothalamus, which is also a central site of action of ghrelin and leptin.

Within the GI tract, GLP-1 acts to delay gastric emptying and with PYY contributes to the “ileal brake,” a homeostatic negative feedback loop that matches the rate of nutrient delivery into the duodenum with the rate of intestinal absorption. If nutrients are incompletely absorbed in the proximal gut, their delivery to the distal small intestine stimulates GLP-1 and PYY secretion, thus slowing the transit of further intestinal contents to allow for improved absorption. Evidence in humans for this role arises from experiments combining gastric scintigraphy with infusions of GLP-1 and/or its receptor antagonist Exendin 9-39. Animal models suggest the involvement of both central and peripheral actions of GLP-1, with vagal afferent division and peripheral GLP-1R blockade with Exendin 9-39 reducing the effect of GLP-1, while intracerebral GLP-1 administration also delayed gastric emptying.

In the last decade, GLP-1-based therapies have proven highly efficacious in the treatment of type 2 diabetes. The short plasma half-life of GLP-1 (it is degraded by dipeptidyl peptidase 4/DPP4 within ~2 min) has been circumvented by the modification of the peptide to resist DPP4, or the use of DPP4 inhibitors. The incretin actions of GLP-1, in particular its low risk of hypoglycemia, make it an attractive alternative to insulin therapy. GLP-1 receptor agonist treatment also results in a modest reduction in body weight and has recently been licensed for the treatment of obesity.

Postprandial plasma GLP-1 concentrations are profoundly increased following surgery that accelerates the delivery of nutrients to the small intestine (e.g., Roux-en-Y gastric bypass, sleeve gastrectomy, total or subtotal gastrectomy, esophagectomy, and Whipple’s procedure). Good quality human experimental evidence, including GLP-1R antagonist studies in post-Roux-en-Y gastric bypass patients, supports GLP-1 being a key intermediary in the improved glucose tolerance, reduced energy intake, and reactive hypoglycemia seen after surgery.

Ongoing research suggests that GLP-1 has a protective effect on the myocardium and brain. GLP-1 and its mimetics appear to have a protective effect in myocardial ischemia and improve cardiac contractility in subjects with cardiac failure, in both diabetic and nondiabetic patients independent of plasma insulin concentrations. The proposed protective effect of GLP-1 on the central nervous system is largely based on epidemiological studies associating cerebrovascular and neurodegenerative diseases with type 2 diabetes and in vitro models of multiple sclerosis, amyotrophic lateral sclerosis, and Alzheimer’s and Parkinson’s diseases. While clinical studies of GLP-1 receptor agonists and DPP4 inhibitors in cardiac and neurological disease are in relatively early phases, it seems promising that they may have a therapeutic benefit above and beyond glycemic and weight control.

2.10.4

GLP-2

GLP-2 is cleaved and cosecreted with GLP-1 from intestinal L-cells. It appears to play a role in intestinal mucosal adaptation to injury and response to refeeding, although there are significant gaps in our understanding of its exact mechanism of action. It acts on a specific G _s -coupled receptor, GLP2R.

GLP-2 was initially discovered due to the intestinotrophic actions of some glucagonomas, which presented with intestinal villous hypertrophy. While glp2r ^−/− mice do not appear to develop differently from wild-type controls, experimental models of intestinal resection or injury (chemical, ischemic, and radiation) have shown a key role for GLP-2 in the recovery of the intestinal mucosa. Partial antagonism of the GLP-2 receptor with GLP-2(3-33) in fasted—refed mice prevented the recovery in mucosal mass and architecture seen in control animals.

Localization of the GLP-2 receptor has historically utilized antisera and immunofluorescence techniques; however, the specificity of these methods has been called into question. As yet, it is unclear as to whether GLP-2 has a direct action on enteric mucosal cells, or acts upon enteric neurons.

A GLP-2 mimetic has recently been licensed for the treatment of short bowel syndrome in humans, with evidence supporting a reduction in parenteral nutrition and fluid requirements. There is also interest in the potential role of exogenous GLP-2 in the treatment of osteoporosis, with one randomized controlled trial demonstrating reduced bone turnover in postmenopausal women treated with a GLP-2 analogue.

As with GLP-1, the GLP-2 response to a meal is enhanced after gastric bypass surgery, although there is no current evidence to suggest that it plays any role in the metabolic outcomes of surgery.

2.10.5

Oxyntomodulin

Oxyntomodulin is a 37 amino acid peptide, sharing the N-terminus of glucagon and the C-terminus of glicentin. There is no known specific oxyntomodulin receptor, but it does appear to exert an effect on gastric acid secretion, glucose homeostasis, and energy balance, likely through its action on the GLP-1 and glucagon receptors. It is cosecreted with GLP-1, GLP-2, and glicentin from the intestinal L-cell, with studies on rat intestinal mucosa suggesting approximately 1/3 of glicentin is processed to oxyntomodulin prior to secretion.

Acting via the GLP-1 and glucagon receptors, oxyntomodulin has lower affinity than the cognate ligands and acts in a biased fashion, with full activity on cAMP-mediated pathways but less effect on GRK2/beta-arrestin activity. Animal models have investigated the metabolic role of oxyntomodulin using glucagon receptor knockout mice, Exendin 9-39 and oxyntomodulin modified to have minimal effect at the glucagon receptor. They have demonstrated that oxyntomodulin activity at the glucagon receptor is needed for it to have an effect on weight homeostasis but has some detrimental effects on glucose homeostasis, counteracted by the GLP1R-activity. Oxyntomodulin has also been shown to increase the heart rate of mice in a glucagon receptor-dependent fashion and may act to influence the circadian rhythm in the liver.

The observations that combined GLP-1 and glucagon receptor agonism are superior to GLP-1 receptor action alone for weight loss, and that oxyntomodulin reduced insulin resistance in a mouse model of diet-induced diabetes make oxyntomodulin a promising drug target. This is supported by evidence from a randomized controlled trial of subcutaneous oxyntomodulin in obese humans, which demonstrated a favorable effect on body mass, as a result of reduced energy intake and elevated energy expenditure. Diet-induced obese mice and nonhuman primates treated with a novel dual GLP-1/glucagon agonist demonstrated significant weight loss and reduced fasting insulin and glucose concentrations. Postprandial plasma oxyntomodulin concentrations, as with GLP-1 and GLP-2, are elevated after gastric bypass surgery and may play a role in the metabolic benefits of bariatric surgery.

2.10.6

Glicentin

Glicentin is a 69 amino acid peptide, which contains the entire sequences of oxyntomodulin (and hence glucagon) and glicentin-related pancreatic peptide. It is cleaved from the proglucagon prohormone in the L-cell, with approximately 1/3 of glicentin being further processed to oxyntomodulin. It is unclear whether it is a metabolically inert by-product of proglucagon processing, which is cosecreted from the L-cell, or has a distinct function. There is no known glicentin receptor, and while it has been assayed and found to have an altered plasma concentration in diabetes, obesity and following foregut surgery, these findings mirror the changes in GLP-1, which are appropriate to the condition studied, and so may just indicate L-cell degranulation rather than physiological glicentin function. In the absence of good-quality animal models, or human infusion experiments, it is not possible to comment further on the physiological, pathological, or pharmacological potential of glicentin.

2.11.1

Synthesis

The human PYY gene, located on chromosome 17, encodes a 97 amino acid prepropeptide. Posttranslational processing to PYY _(1-36) is driven by an as yet unconfirmed PC although, given its presence in the L cell, PC1/3 is a prime candidate. PYY _(1-36) acts at all neuropeptide Y family receptors (YR1-5), whereas PYY _(3-36) is a preferential agonist for the Y2 receptor (Y2R). PYY colocalizes, and is cosecreted, with proglucagon gene products from L cells in the intestinal epithelium. The overlap in hormone expression of L cells varies along the GI tract, with proximally situated L cells expressing greater relative concentrations of proglucagon products, with PYY becoming more prevalent in distally located cells in the ileum and colon.

2.11.2

Secretion

L cells are electrically active, open-type endocrine cells, with hormone secretion coupled to elevated intracellular calcium concentration by voltage-gated calcium channels. Diverse nutrient sensing mechanisms, including G-protein-coupled receptors and sodium glucose cotransporters (SGLT1), link the absorption of luminal nutrients to vesicular hormone release. Secretion of PYY is stimulated by carbohydrates, amino acids, lipids (long- and short-chain fatty acids) and bile acids and may also be influenced by neurohumoral circuits involving CCK, GLP-1, VIP, and the vagus nerve.

The proximal to distal increase in PYY expressing cells in the gut may explain the differential secretion profile of GLP-1 and PYY, despite their cosecretion. PYY secretion is more effectively stimulated by ingested lipids than carbohydrates, with carbohydrate-mediated secretion increased by acarbose administration, suggesting that distal delivery of nutrients is necessary for PYY secretion.

Plasma PYY concentrations correlated with energy intake, with greater circulating concentrations found after a 1500 kcal than a 500 kcal meal in healthy volunteers. Bile acids appear to play a role in PYY secretion, as colonic instillation of taurocholate or deoxycholate resulted in a four- to sixfold increase in plasma PYY concentration in healthy volunteers and type 2 diabetic patients. Vagal stimulation in anaesthetized pigs caused PYY secretion and vagotomy in rats suppressed postprandial PYY secretion, yet preservation of the vagus at the time of partial gastrectomy appeared to suppress fasting PYY levels, suggesting a complex and incompletely understood pathway for parasympathetic tone to control PYY secretion. In healthy human volunteers, CCK stimulated PYY secretion in the fasting state.

2.11.3

Function

PYY was initially identified with pancreatic polypeptide (PP) and found to be a potent inhibitor of secretin and CCK-mediated pancreatic exocrine secretion, but has since been found to have multiple roles.

Infusion of PYY into healthy volunteers with gastric and duodenal sampling tubes resulted in a significant reduction in pentagastrin-mediated acid and pepsin secretion, but no change in secretin or CCK-stimulated pancreatic exocrine secretion, although this was measured using a duodenal tube, not direct sampling of the pancreatic duct. Contrasting results in a canine model with pancreatic fistula showed a significant reduction in meal-induced pancreatic exocrine secretion with exogenous PYY infusion. PYY infusion to prairie dogs with gallbladder catheterization resulted in a greater post-CCK relaxation of gallbladder smooth muscle and expansion of storage volume.

PYY mediates delayed gastric emptying and intestinal transit, in a vagus-mediated fashion in animal models. An in depth study of the independent effects of Y1R and Y2R activity in rats suggested that central Y2R activity delays gastric transit, but Y1R activity stimulates gastric emptying, pointing to a balanced role of PYY _(1-36) and PYY _(3-36) .

PYY is a key intermediary of angiotensin 2-mediated salt and water retention in the colon, which acts to stimulate L cell PYY secretion and subsequent Y1R-mediated reduced luminal fluid secretion.

Rodent knockout models have demonstrated that PYY _(3-36) acts through Y2R to inhibit the activity of orexigenic NPY neurons in the hypothalamus and reduce food intake. Exogenous PYY _(3-36) infusion potently reduced food intake in a dose-dependent manner in 16 healthy human volunteers. In a comparison of 12 healthy lean and 12 healthy obese volunteers receiving a placebo-controlled, double-blind, crossover study of PYY _(3-36) , exogenous PYY _(3-36) was found to cause a similar reduction in food intake in both groups, although endogenous PYY production was lower in the obese group, pointing to a picture of PYY deficiency, rather than resistance, in obesity.

The relative importance of central versus peripheral PYY _(3-36) action on eating behavior is a topic of debate. Abdominal vagotomy abolished the effects of exogenous PYY _(3-36) on feeding and c-fos activation in the arcuate nucleus in a rat model. Conflicting evidence has, however, demonstrated that central administration of PYY _(3-36) or a Y2R antagonist influenced eating behavior in rats, PYY _(3-36) reduced food intake in vagotomized rats, and capsaicin denervation of abdominal afferents in rats did not abolish the anorexic effects of PYY _(3-36) .

It remains to be proven whether PYY has a role in glucose homeostasis. Human infusion studies have shown no effect of PYY _(3-36) infusion on insulin secretion or glucose elimination in the fasting or stimulated state. Pyy and npy1r knockout mice did, however, manifest fasting hyperglycemia, although potential confounding effects of diet-induced obesity in this model are difficult to exclude. Interestingly, the profoundly elevated plasma PYY levels seen after bariatric surgery have recently been linked to the improvement of glucose homeostasis through a resting/preserving effect on pancreatic beta-cell mass.

2.11.4

Clinical Aspects

Clinical interest in PYY centers on its potential to reduce energy intake and therefore as a therapy for obesity. Multiple studies have generated conflicting evidence on PYY homeostasis in the obese population, with one well-designed study proposing that PYY secretion is impaired, but its actions unchanged, in the obese population. Examination of PYY secretion before and after weight loss in the obese population showed a marked reduction in fasting PYY concentrations.

As with other distal small intestine secreted peptides, the postprandial PYY response is enhanced after gastric bypass and may contribute to the reduction in energy intake and consequent weight loss seen in this group, with one rodent knockout model showing that PYY is essential for early weight loss after bariatric surgery.

2.12.1

Synthesis

INSL5 is primarily secreted from colonic L cells in the human and mouse and is cosecreted with GLP-1 and PYY. It has also been found in hypothalamic neurons in mice. The INSL5 structure is made up of a 21 amino acid A chain and 24 amino acid B chain, both encoded by the INSL5 gene on human chromosome 1, linked by disulfide bonds. Its cognate receptor is the G-protein-coupled receptor Rxfp4, through which it couples to G _i pathways to reduce cytosolic cAMP concentrations and phosphorylate ERK1/2, p38MAPK, and Akt. Within the GI tract, Rxfp4 has been identified in myenteric and submucosal neurons, as well as in nodose ganglia . The few experimental studies on INSL5 have relied on mouse models due to the difficulty in synthesizing human INSL5, and the fact that INSL5 and Rxfp4 are pseudogenes in the rat and dog.

2.12.2

Secretion

Plasma INSL5 was shown to be suppressed by refeeding following an overnight fast in mice. Calorie restricted mice had a higher plasma concentration of INSL5 than those fed a high fat diet and higher colonic Insl5 mRNA levels. In vitro cultures of human and mouse colonic mucosa secreted INSL5 in response to L-cell cAMP elevation. Plasma INSL5 concentrations in germ free mice were suppressed by a high fat diet, or bacterial colonization, suggesting a role for microbial energy scavenging in INSL5 signaling.

2.12.3

Function

There is a good evidence of an orexigenic role for INSL5 in mice. Exogenous INSL5 dose-dependently increased chow intake in wild-type but not Rxfp4 knockout mice, and immunoneutralization of INSL5 reduced chow intake in wild-type but not Rxfp4 knockout mice. Interestingly, Rxfp4 knockout mice appeared to have reduced preference for a high fat diet. Insl5 knockout mice, however, had no difference in body weight or meal intake compared to wild-type mice, pointing to a nuanced role for the INSL5/Rxfp4 pathway in the control of energy intake.

There is conflicting evidence for the role of INSL5 in glucose homeostasis. Examination of insulinoma cells and a mouse enteroendocrine cell line (GLUTag) revealed Rxfp4 expression and treatment with INSL5 stimulated insulin and GLP-1 secretion. In two studies, Rxfp4 or Insl5 knockout mice had no difference in oral glucose tolerance, although one Insl5 knockout study demonstrated a difference in intraperitoneal glucose tolerance. Another study of Insl5 knockout mice demonstrated impaired glucose tolerance, reduced insulin secretion, and reduced pancreatic islet area and administration of INSL5 to wild-type or diabetic mice improved intraperitoneal glucose tolerance. Without further mechanistic studies examining the interaction between INSL5, islet hormone secretion, and hepatic glucose metabolism, it is not clear what, if any, role INSL5 truly plays in glucose homeostasis.

To date, no infusion or receptor blockade study of INSL5 in man has been published. A single published study examining the role of INSL5 in humans examined the correlation between plasma INSL5, sex hormones, and bariatric surgery, without reaching clear conclusions regarding its metabolic role.

2.12.4

Clinical Aspects

The potential role of INSL5 in energy intake makes it an interesting target for obesity research, although the field needs the development of better tools, including reliable assays for plasma INSL5 detection, to define its exact role and what, if any, pathological or pharmacological role it may play.