Complete sequences of glucagon-like peptide-1 from human and pig small intestine

In the small intestine, proglucagon is processed into the previously characterized peptide "glicentin" (proglucagon (PG) 1-69) and two smaller peptides showing about 50% homology with glucagon: glucagon-like peptide-1 and -2. It was assumed that the sites of post-translational cleavage in the small intestine of the proglucagon precursor were determined by pairs of basic amino acid residues flanking the two peptides. Earlier studies have shown that synthetic glucagon-like peptide-1 (GLP-1) synthesized according to the proposed structure (proglucagon 71-108 or because residue 108 is Gly, 72-107 amide) had no physiological effects, whereas a truncated from of GLP-1, corresponding to proglucagon 78-107 amide, strongly stimulated insulin secretion and depressed glucagon secretion. To determine the amino acid sequence of the naturally occurring peptide we isolated GLP-1 from human small intestine by hydrophobic, gel permeation, and reverse-phase high performance liquid chromatography. By analysis of composition and sequence it was determined that the peptide corresponded to PG 78-107. By mass spectrometry the molecular mass was determined to be 3295, corresponding to PG 78-107 amide. Furthermore, mass spectrometry of the methyl-esterified peptide showed an increase in mass compatible with the presence of alpha-carboxyl amidation. Thus, the gut-derived insulinotrophic hormone GLP-1 is shown to be PG 78-107 amide.     

Naturally occurring products of proglucagon 111-160 in the porcine and human small intestine

Recent studies have revealed that the glucagon gene is expressed in the mammalian intestine. Here it codes for "glicentin" (proglucagon 1-69) and a glucagon-like peptide, proglucagon 78-107, recently isolated from porcine intestine. We studied the fate of the remaining COOH-terminal part of proglucagon (proglucagon 111-160) using radioimmunoassays against proglucagon 111-123 and 126-160. Two peptides were isolated from acid ethanol extracts of porcine ileal mucosa and sequenced: one corresponding to proglucagon 126-158 and one probably corresponding to proglucagon 111-158. By comparing human and porcine proglucagon sequences, Ala117 is replaced by Thr, and Ile138, Ala144, Ile152 and Gln153 are replaced by Val, Thr, Leu, and His. By gel filtration and radioimmunoassay of intestinal extracts it was established that a large part of porcine and virtually all of human proglucagon are processed to release proglucagon 111-123 (designated spacer peptide 2), which, like proglucagon 126-158 must be considered a potential hormonal entity. By isocratic high pressure liquid chromatography human spacer peptide 2 was indistinguishable from synthetic proglucagon 111-122 amide, suggesting that this is the structure of the naturally occurring human peptide.     

Structure, measurement, and secretion of human glucagon-like peptide-2

By using radioimmunoassays toward the cDNA-predicted amino acid sequence of human glucagon-like peptide-2, a peptide was isolated from extracts of human ileum. By mass spectrometry and Edman sequencing, this peptide was identified as human proglucagon 126-158. High-performance liquid chromatography analyses indicated that a similar immunoreactive peptide (iGLP-2) was present in human plasma. Human plasma concentrations of iGLP-2 were elevated 3- to 4-fold at 1 to 2 h after ingestion of 800 to 1200 kcal meals.     

GLP-2 stimulates colonic growth via KGF, released by subepithelial myofibroblasts with GLP-2 receptors

BACKGROUND: Glucagon-like peptide-2 is thought to act as a growth factor for the gut, but the localization of the GLP-2 receptor and mechanism of action on epithelial growth is unclear. METHODS AND RESULTS: We found glucagon-like peptide-2 (GLP-2) receptors mainly on subepithelial myofibroblasts in rat, mouse, marmoset and human small and large intestine by immunohistochemistry and in situ hybridisation. By double labelling we found that these GLP-2 receptor immunoreactive cells also produce smooth muscle actin and keratinocyte growth factor (KGF). By subcutaneous infusion of either GLP-2 alone, GLP-2 plus KGF antibody, KGF antibody alone or saline in mice, we found that KGF antibody abolished the growth promoting effect of GLP-2 in the large intestine, but not in the small intestine. CONCLUSIONS: Our findings suggest that GLP-2 in the gut acts by activating receptors on the subepithelial myofibroblasts, causing the release of growth factors, which in turn stimulate intestinal growth.     

GLP-1 and GLP-2 act in concert to inhibit fasted,but not fed,small bowel motility in the rat

Small bowel motility was studied in rats at increasing (1-20 pmol/kg/min) intravenous doses of either glucagon-like peptide-1 (GLP-1) or glucagon-like peptide-2 (GLP-2) alone, or in combination in the fasted and fed state. There was a dose-dependent inhibitory action of GLP-1 on the migrating myoelectric complex (MMC), where the dose of 5 pmol/kg/min induced an increased MMC cycle length. No effect was seen with GLP-2 alone, but the combination of GLP-1 and GLP-2 induced a more pronounced inhibitory effect, with significant increase of the MMC cycle length from a dose of 2 pmol/kg/min. During fed motility, infusion of GLP-1 resulted in an inhibition of spiking activity compared to control. In contrast, infusion of GLP-2 only numerically increased spiking activity compared to control, while the combination of GLP-1 and GLP-2 resulted in no change compared to control. In summary, this study demonstrates an additive effect of peripheral administration of GLP-1 and GLP-2 on fasted small bowel motility. In the fed state, GLP-1 and GLP-2 seem to display counter-balancing effects on motility of the small intestine.     

Identification and characterization of glucagon-like peptide-1 7-36 amide-binding sites in the rat brain and lung

High-affinity binding sites for glucagon-like peptide-1 7-36 amide (GLP-1 7-36 NH2) were identified in rat brain and lung membranes. Binding of [125I]GLP-1 7-36 NH2 was rapid, reversible, specific, saturable and pH dependent. Specific binding in the central nervous system was particularly high in the hypothalamus and the brain stem. Oxyntomodulin, glucagon-like peptide-1, glucagon-like peptide-2 and glucagon were 100-1000-fold less potent than GLP-1 7-36 NH2 in competition for this binding site.     

Immunoneutralization of endogenous glucagon-like peptide-2 reduces adaptive intestinal growth in diabetic rats

Supraphysiological doses of glucagon-like peptide-2 (GLP-2) have been shown to induce intestinal growth by increasing villus height and crypt depth and by decreasing apoptosis, but a physiological effect of GLP-2 has not yet been demonstrated. Earlier, we found elevated levels of endogenous GLP-2 in untreated streptozotocin diabetic rats associated with marked intestinal growth. In the present study, we investigated the role of endogenous GLP-2 for this adaptive response. We included four groups of six rats: (1) diabetic rats treated with saline, (2) diabetic rats treated with non-specific antibodies, (3) diabetic rats treated with polyclonal GLP-2 antibodies and (4) non-diabetic control rats treated with saline. All animals were treated with once daily intraperitoneal injections for 13 days and killed on day 14. Diabetic rats treated with saline or non-specific antibodies had a significantly (P<0.01) increased area of mucosa (13.00+/-0.64 and 13.37+/-0.60 mm(2), respectively) in the proximal part of the small intestine compared with non-diabetic controls (7.97+/-0.70 mm(2)). In contrast, diabetic rats treated with GLP-2 antibodies had a significantly (P<0.01) smaller increase in area of mucosa in the proximal part of the small intestine (10.84+/-0.44 mm(2)). Antibody treatment had no effect on body weight, blood glucose concentrations and food intake. Thus, blocking of endogenous GLP-2 in a model of adaptive intestinal growth reduces the growth response, providing strong evidence for a physiological growth factor function of GLP-2.     

Characterization of high-affinity receptors for truncated glucagon-like peptide-1 in rat gastric glands

The truncated form of glucagon-like peptide-1 (TGLP-1, or proglucagon 78-108), secreted by the mammalian intestine, has potent pharmacological activities, stimulating insulin release and inhibiting gastric acid secretion. We have characterized high-affinity receptors for this peptide in rat isolated fundic glands. Scatchard analysis of binding studies using mono-125I-TGLP-1(7-36) amide as tracer showed a single class of binding site of Kd (4.4 +/- (SE) .08) x 10(-10) M, with a tissue concentration of 1.0 +/- 0.1 fmol sites/microgram DNA. Whole GLP-1 was approximately 700 times less potent in displacing tracer, while human GLP-2 and pancreatic glucagon produced no significant displacement at concentrations up to 10(-6) M. The data support a physiological role for TGLP-1 in the regulation of gastric acid secretion.     

Crystal structure of the ligand-bound glucagon-like peptide-1 receptor extracellular domain

The glucagon-like peptide-1 receptor (GLP-1R) belongs to Family B1 of the seven-transmembrane G protein-coupled receptors, and its natural agonist ligand is the peptide hormone glucagon-like peptide-1 (GLP-1). GLP-1 is involved in glucose homeostasis, and activation of GLP-1R in the plasma membrane of pancreatic beta-cells potentiates glucose-dependent insulin secretion. The N-terminal extracellular domain (nGLP-1R) is an important ligand binding domain that binds GLP-1 and the homologous peptide Exendin-4 with differential affinity. Exendin-4 has a C-terminal extension of nine amino acid residues known as the "Trp cage", which is absent in GLP-1. The Trp cage was believed to interact with nGLP-1R and thereby explain the superior affinity of Exendin-4. However, the molecular details that govern ligand binding and specificity of nGLP-1R remain undefined. Here we report the crystal structure of human nGLP-1R in complex with the antagonist Exendin-4(9-39) solved by the multiwavelength anomalous dispersion method to 2.2A resolution. The structure reveals that Exendin-4(9-39) is an amphipathic alpha-helix forming both hydrophobic and hydrophilic interactions with nGLP-1R. The Trp cage of Exendin-4 is not involved in binding to nGLP-1R. The hydrophobic binding site of nGLP-1R is defined by discontinuous segments including primarily a well defined alpha-helix in the N terminus of nGLP-1R and a loop between two antiparallel beta-strands. The structure provides for the first time detailed molecular insight into ligand binding of the human GLP-1 receptor, an established target for treatment of type 2 diabetes.     

Truncated glucagon-like peptide-1 interacts with exendin receptors on dispersed acini from guinea pig pancreas. Identification of a mammalian analogue of the reptilian peptide exendin-4.

To find mammalian analogues of exendin-4, a peptide from Helodermatidae venoms that interacts with newly discovered exendin receptors on dispersed acini from guinea pig pancreas, we examined the actions of recent additions to the vasoactive intestinal peptide/secretin/glucagon family of regulatory peptides. In every respect tested, the truncated form of glucagon-like peptide-1, GLP-1(7-36)NH2, mimicked the actions of exendin-4. Like exendin-4, GLP-1(7-36)NH2 caused an increase in acinar cAMP without stimulating amylase release. GLP-1(7-36)NH2-induced increases in cAMP were inhibited progressively by increasing concentrations of the specific exendin-receptor antagonist, exendin(9-39)NH2. In dispersed acini from guinea pig and rat pancreas, concentrations of GLP-1(7-36)NH2 that stimulated increases in cAMP caused potentiation of cholecystokinin-induced amylase release. Binding of 125I-[Y39]exendin-4 or 125I-GLP-1(7-36)NH2 to dispersed acini from guinea pig pancreas was inhibited by adding increasing concentrations of unlabeled exendin-4 or GLP-1(7-36)NH2. We conclude that the mammalian peptide GLP-1(7-36)NH2 interacts with exendin receptors on dispersed acini from guinea pig pancreas. Exendin(9-39)NH2, a competitive antagonist of the actions of GLP-1(7-36)NH2 in pancreatic acini, may be a useful tool for examining the physiological actions of this peptide.     

Glucagon-like peptide-2 acutely increases proximal small intestinal blood flow in TPN-fed neonatal piglets

Glucagon-like peptide-2 (GLP-2) is a gut hormone that is secreted in response to enteral feeding and stimulates small intestinal mucosal growth. We have previously shown that GLP-2 infusion acutely increases portal venous blood flow in TPN-fed piglets. The aim of this study was to localize the vasoactive effect of GLP-2 within the gastrointestinal tissues and other visceral organs in TPN-fed piglets. Tissue blood flow rates were quantified using fluorescent microsphere deposition in anesthetized TPN-fed piglets given intravenous infusion of GLP-2 at either 500 pmol x kg(-1) x h(-1) (low GLP-2, n = 7 pigs) or 2,000 pmol x kg(-1) x h(-1) (high GLP-2, n = 8 pigs) for 2 h. Compared with baseline, the low and the high GLP-2 treatment significantly increased the blood flow rate in the duodenum (+77%) and jejunum (+40% and 80%), respectively, but blood flow to the distal small intestine and colon (-15%) was unchanged or slightly decreased. Baseline mucosal blood flow was five-fold higher than serosal blood flow; however, high GLP-2 treatment increased serosal (+140%) to a larger degree than mucosal blood flow (+73%). The high GLP-2 dose increased pancreatic flow (+34%) but decreased blood flow in the kidneys (-14%) and stomach (-12%), whereas the spleen and brain were unaffected. These findings suggest that the acute GLP-2-mediated stimulation of portal blood flow in TPN-fed piglets occurs principally via increased blood flow through the superior mesenteric artery to the proximal small intestine, a tissue region where the GLP-2R mRNA abundance and trophic GLP-2 effects are greatest.     

Glucagon-like peptide-2 increases sucrase-isomaltase but not caudal-related homeobox protein-2 gene expression

To determine the effect of glucagon-like peptide-2 (GLP-2) on sucrase-isomaltase and caudal-related homeobox protein-2 (Cdx-2) gene expression, male Wistar rats were divided into total parenteral nutrition (TPN)-fed and GLP-2-treated, TPN-fed groups. TPN was given via a jugular line, inserted under anesthesia, for 7 days. The treatment group received 40 microg/day of GLP-2 intravenously with the TPN diet. The small intestine and colon were weighed and measured. Tissue was obtained from the jejunum, terminal ileum, and midcolon. RNA analysis, morphometry, and microdissection were performed. The weight of the small intestine of GLP-2-treated rats was greater than that of TPN-fed rats (P < 0.001). GLP-2 increased the mean metaphase arrests/crypt in both the jejunum and ileum (P < 0.001). Ileal expression of sucrase-isomaltase was increased by 1. 6-fold (P < 0.05). Jejunal expression was increased by a similar amount, although not significantly (P = 0.08). There was no change in Cdx-2 gene expression. Thus GLP-2 can maintain small intestinal morphology and function, but effects on gene expression are not mediated by gross changes in the level of the mRNA for the homeobox protein Cdx-2.     

The importance of the nine-amino acid C-terminal sequence of exendin-4 for binding to the GLP-1 receptor and for biological activity

Exendin-4, a 39-amino acid (AA) peptide, is a long-acting agonist at the glucagon-like peptide-1 (GLP-1) receptor. Consequently, it may be preferable to GLP-1 as a long-term treatment for type 2 diabetes mellitus. Exendin-4 (Ex-4), unlike GLP-1, is not degraded by dipeptidyl peptidase IV (DPP IV), is less susceptible to degradation by neutral endopeptidase, and possesses a nine-AA C-terminal sequence absent from GLP-1. Here we examine the importance of these nine AAs for biological activity of Ex-4, a sequence of truncated Ex-4 analogs, and native GLP-1 and GLP-1 analogs to which all or parts of the C-terminal sequence have been added. We found that removing these AAs from Ex-4 to produce Ex (1-30) reduced the affinity for the GLP-1 receptor (GLP-1R) relative to Ex-4 (IC50: Ex-4, 3.22+/-0.9 nM; Ex (1-30), 32+/-5.8 nM) but made it comparable to that of GLP-1 (IC50: 44.9+/-3.2 nM). The addition of this nine-AA sequence to GLP-1 improved the affinity of both GLP-1 and the DPP IV resistant analog GLP-1 8-glycine for the GLP-1 receptor (IC50: GLP-1 Gly8 [GG], 220+/-23 nM; GLP-1 Gly8 Ex (31-39), 74+/-11 nM). Observations of the cAMP response in an insulinoma cell line show a similar trend for biological activity.     

An histological and immunocytochemical study of the neuroendocrine cells in the intestine of Podarcis hispanica Steindachner, 1870 (Lacertidae)

Summary Numerous endocrine cells can be observed in the gut of the lizard Podarcis hispanica after application of the Grimelius silver nitrate technique. The argyrophilic endocrine cells are usually tall and thin in the small intestine but short, basal, and round in the large intestine. Eleven types of immunoreactive endocrine cells have been identified by immunocytochemical methods. Numerous serotonin-, caerulein/gastrin/cholecystokinin octapeptide-and peptide tyrosine-tyrosine-immunoreactive cells; a moderate number of pancreatic polypeptide-, neurotensin-, somatostatin-, glucagon-like peptide-1-and glucagon-immunoreactive cells, and few cholecystokinin N-terminal-and bombesin-immunoreactive cells were found in the epithelium of the small intestine. Coexistence of glucagon with GLP-1 or PP/PYY has been observed in some cells. In the large intestine a small number of serotonin-, peptide tyrosine-tyrosine-, pancreatic polypeptide-, neurotensin-, somatostatin-and glucagon-like peptide-1-immunoreactive cells were detected. Vasoactive intestinal peptide immunoreactivity was found in nerve fibers of the muscular layer. Substance P-immunoreactive nerve fibers were detected in lamina propria, submucosa and muscular layer. Chromogranin A-immunoreactive cells were observed throughout the intestine, although in lower numbers than argyrophilic cells.     

Metabolism of glucagon-like peptide-2 in pigs: role of dipeptidyl peptidase IV

Little is known about the metabolism of the intestinotropic factor glucagon-like peptide-2 (GLP-2); except that it is a substrate for dipeptidyl peptidase IV (DPP-IV) and that it appears to be eliminated by the kidneys. We, therefore, investigated GLP-2 metabolism in six multicatheterized pigs receiving intravenous GLP-2 infusions (2 pmol/kg/min) before and after administration of valine-pyrrolidide (300 mumol/kg; a well characterized DPP-IV inhibitor). Plasma samples were analyzed by radioimmunoassays allowing determination of intact, biologically active GLP-2 and the DPP-IV metabolite GLP-2 (3-33). During infusion of GLP-2 alone, 30.9+/-1.7% of the infused peptide was degraded to GLP-2 (3-33). After valine-pyrrolidide, there was no significant formation of the metabolite. Significant extraction of intact GLP-2 was observed across the kidneys, the extremities (represented by a leg), and the splanchnic bed, resulting in a metabolic clearance rate (MCR) of 6.80+/-0.47 ml/kg/min and a plasma half-life of 6.8+/-0.8 min. Hepatic extraction was not detected. Valine-pyrrolidide addition did not affect extraction ratios significantly, but decreased (p=0.003) MCR to 4.18+/-0.27 ml/kg/min and increased (p=0.052) plasma half-life to 9.9+/-0.8 min. The metabolite was eliminated with a half-life of 22.1+/-2.6 min and a clearance of 2.07+/-0.11 ml/kg/min. In conclusion, intact GLP-2 is eliminated in the peripheral tissues, the splanchnic bed and the kidneys, but not in the liver, by mechanisms unrelated to DPP-IV. However, DPP-IV is involved in the overall GLP-2 metabolism and seems to be the sole enzyme responsible for N-terminal degradation of GLP-2.     

The release of GLP-1 and ghrelin, but not GIP and CCK, by glucose is dependent upon the length of small intestine exposed

Previous observations suggest that glucagon-like peptide-1 (GLP-1) is released into the bloodstream only when dietary carbohydrate enters the duodenum at rates that exceed the absorptive capacity of the proximal small intestine to contact GLP-1 bearing mucosa in more distal bowel. The aims of this study were to determine the effects of modifying the length of small intestine exposed to glucose on plasma concentrations of GLP-1 and also glucose-dependent insulinotropic peptide (GIP), insulin, cholecystokinin (CCK) and ghrelin, and antropyloric pressures. Glucose was infused at 3.5 kcal/min into the duodenum of eight healthy males (age 18-59 yr) over 60 min on the first day into an isolated 60-cm segment of the proximal small intestine ("short-segment infusion"); on the second day, the same amount of glucose was infused with access to the entire small intestine ("long-segment infusion"). Plasma GLP-1 increased and ghrelin decreased (P < 0.05 for both) during the long-, but not the short-, segment infusion. By contrast, increases in plasma CCK and GIP did not differ between days. The rises in blood glucose and plasma insulin were greater during the long- than during the short-segment infusion (P < 0.05). During the long- but not the short-segment infusion, antral pressure waves (PWs) were suppressed (P < 0.05). Isolated pyloric PWs and basal pyloric pressure were stimulated on both days. In conclusion, the release of GLP-1 and ghrelin, but not CCK and GIP, is dependent upon >60 cm of the intestine being exposed to glucose.     

Increased postprandial responses of GLP-1 and GIP in patients with chronic pancreatitis and steatorrhea following pancreatic enzyme substitution

We aimed to investigate how assimilation of nutrients affects the postprandial responses of glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP) and to evaluate the effect of pancreatic enzyme substitution (PES) on insulin secretion in patients with chronic pancreatitis (CP) and pancreatic exocrine insufficiency (PEI). Eight male patients with CP and PEI were studied. Blood was sampled frequently on two separate days after ingestion of a liquid meal with and without PES, respectively. Eight healthy male subjects served as a control group. beta-Cell responsiveness was estimated as changes in insulin secretion rates in response to changes in postprandial plasma glucose (PG). There was no difference in the PG incremental area under curve (AUC) for patients with and without PES [406 +/- 100 vs. 425 +/- 80 mM.4 h (mean +/- SE), P = 0.8]. The response of total GLP-1 was higher after PES (AUC: 7.8 +/- 1.2 vs. 5.3 +/- 0.6 nM.4 h, P = 0.01), as was the response of total GIP (AUC: 32.7 +/- 7.5 vs. 21.1 +/- 8.3 nM.4 h, P = 0.01). Concurrently, both plasma insulin, plasma C-peptide, and total insulin secretion increased after PES (AUC: 17.7 +/- 4.2 vs. 13.6 +/- 2.9 nM.4 h, P = 0.02; 237 +/- 31.4 vs. 200 +/- 27.4 nM.4 h, P = 0.005; and 595 +/- 82 vs. 497 +/- 80 pmol.kg(-1).4 h, P = 0.01, respectively). beta-Cell responsiveness to glucose was not significantly different on the two study days for patients with CP. These results suggest that the secretion of GLP-1 and GIP is under influence of the digestion and absorption of nutrients in the small intestine and that PES increases insulin secretion.     

Glucagon-Like Peptide-1 Receptor Ligand Interactions: Structural Cross Talk between Ligands and the Extracellular Domain

Activation of the glucagon-like peptide-1 receptor (GLP-1R) in pancreatic β-cells potentiates insulin production and is a current therapeutic target for the treatment of type 2 diabetes mellitus (T2DM). Like other class B G protein-coupled receptors (GPCRs), the GLP-1R contains an N-terminal extracellular ligand binding domain. N-terminal truncations on the peptide agonist generate antagonists capable of binding to the extracellular domain, but not capable of activating full length receptor. The main objective of this study was to use Hydrogen/deuterium exchange (HDX) to identify how the amide hydrogen bonding network of peptide ligands and the extracellular domain of GLP-1R (nGLP-1R) were altered by binding interactions and to then use this platform to validate direct binding events for putative GLP-1R small molecule ligands. The HDX studies presented here for two glucagon-like peptide-1 receptor (GLP-1R) peptide ligands indicates that the antagonist exendin-4_[9-39] is significantly destabilized in the presence of nonionic detergents as compared to the agonist exendin-4. Furthermore, HDX can detect stabilization of exendin-4 and exendin-4_[9-39] hydrogen bonding networks at the N-terminal helix [Val19 to Lys27] upon binding to the N-terminal extracellular domain of GLP-1R (nGLP-1R). In addition we show hydrogen bonding network stabilization on nGLP-1R in response to ligand binding, and validate direct binding events with the extracellular domain of the receptor for putative GLP-1R small molecule ligands.     

Expression and distribution of pituitary adenylate cyclase-activating peptide in human prostate and prostate cancer tissues

The therapeutic potential of the intestinotrophic mediator glucagon-like peptide-2 (1-33) [GLP-2 (1-33)] has increased interest in the pharmacokinetics of the peptide. This study was undertaken to investigate whether the primary degradation product GLP-2 (3-33) interacts with the GLP-2 receptor. Functional (cAMP) and binding in vitro studies were carried out in cells expressing the transfected human GLP-2 receptor. Furthermore, a biologic response of GLP-2 (3-33) was tested in vivo. Mice were allocated to groups treated for 10 days (twice daily) with: (1) 5 microg GLP-2 (1-33), (2) 25 microg GLP-2 (3-33), (3) 5 microg GLP-2 (1-33)+100 microg GLP-2 (3-33), or (4) 5 microg GLP-2 (1-33)+500 microg GLP-2 (3-33). The intestine was investigated for growth changes. GLP-2 (3-33) bound to the GLP-2 receptor with a binding affinity of 7.5% of that of GLP-2 (1-33). cAMP accumulation was stimulated with an efficacy of 15% and a potency more than two orders of magnitude lower than that of GLP-2 (1-33). Increasing doses of GLP-2 (3-33) (10(-7)-10(-5) M) caused a shift to the right in the dose-response curve of GLP-2 (1-33). Treatment of mice with either GLP-2 (1-33) or (3-33) induced significant growth responses in both the small and large intestines, but the response induced by GLP-2 (3-33) was much smaller. Co-administration of 500 microg of GLP-2 (3-33) and 5 microg GLP-2 (1-33) resulted in a growth response that was smaller than that of 5 microg GLP-2 (1-33) alone. Consistent with the observed in vivo activities, our functional studies and binding data indicate that GLP-2 (3-33) acts as a partial agonist with potential competitive antagonistic properties on the GLP-2 receptor.