Peptide YY Inhibits the Insulinotropic Action of Gastric Inhibitory Polypeptide

Transcription

1 GASTROENTEROLOGY 1989;96:69-4 Peptide YY Inhibits the Insulinotropic Action of Gastric Inhibitory Polypeptide YAN-SHI GUO, POMILA SINGH, EDWIN DRA VIAM, GEORGE H. GREELEY, Jr., and JAMES C. THOMPSON Department of Surgery, The University of Texas Medical Branch, Galveston, Texas Peptide YY (PYY) is released from the gut after ingestion of fat or after a meal. The purpose of this investigation was to examine the effect of PYY on gastric inhibitory polypeptide (GIP)-stimulated insulin release in conscious dogs with gastric and duodenal fistulas. In control experiments, 6 dogs received GIP (4 pmollkg, i.v., for 1 h) and glucose (.6 glkg, i.v., for 1 h); the integrated insulin response over a l-h period was 142 ± 32.7 ng-6 mini ml. The plasma GIP levels achieved by this procedure were similar to those observed by intraduodenal infusion of Lipomul (2 mllmin), suggesting that the dose of GIP used was within the physiologic range. Intravenous infusion of three different doses of PYY (1, 2, or 4 pmollkg. h) caused a significant inhibition of insulin release stimulated by GIP + glucose; the integrated insulin response was reduced to 15, 88, and 79 ng-6 minlml, respectively. On the other hand, PYY (4 pmoll kg. h) had no effect on insulin secretion induced by intravenous glucose (.6 glkg. h) alone. These results indicate that PYY specifically inhibits the insulinotropic action of GIP and that PYY may play a negative-feedback regulatory role in the enteroinsular axis. Considerable evidence has been accumulated to support the theory that gastrointestinal peptides playa significant role in the release of insulin after a meal. This role has been called the incretin effect (1,2), and the functional relationship between the gut and pancreatic islets has been referred to as the enteroinsular axis (1,2). Gastric inhibitory polypeptide (GIP) is presently the strongest incretin candidate (1-3). Although the positive effect of the enteric hormones on insulin release has b e well ~ n documented, the negative effect has not been. Peptide YY (PYY), a 36-amino acid polypeptide, was recently isolated from porcine duodenal extract (4). Peptide YY is found in high concentrations in the endocrine cells of the small intestine and colon of humans, dogs, and rats (5-9). The pancreas was recently identified, by means of a clonal cdna probe, as the major site of PYY mrna synthesis (1). Plasma PYY levels have been reported to increase after a meal and after intraduodenal infusion of oleic acid (8,11,12). Peptide YY has a broad spectrum of biologic actions. We and others have shown that PYY inhibits (a) stimulated gastric acid secretion (12-14), (b) pancreatic exocrine secretions (4,12,15), (c) 2-deoxy-glucose- and neuropeptide-stimulated insulin release (16), (d) gastric and intestinal motility (17-19), and (e) intestinal blood flow (17). The present study was undertaken to examine the inhibitory action of PYY on GIP-stimulated insulin release in conscious dogs. Materials and Methods Six dogs (18-22 kg) of either sex were prepared with gastric and duodenal cannulas and were allowed to recover for 3 wk before the following experiments were done. The dogs were fasted for 18 h before the initiation of an experiment, with free access to tap water. During each experiment, the gastric and duodenal cannulas were opened and the stomach and duodenum of each dog were rinsed gently with warm water. Two Teflon catheters were then placed in the veins of both hind limbs of the dogs, one for infusion of drugs and one for collection of blood samples. Experimental Design All experiments were done in random order and no more than two experiments were done per week. Experiment 1. On separate days the effect of each insulinotropic agent, either alone (control) or in combination with a single dose of PYY, was studied. After collec- Abbreviation used in this paper: PYY, peptide YY by the American Gastroenterological Association /89/$3.5

2 March 1989 PEPTIDE YY INHIBITS GIP INSULlNOTROPIC ACTION 691 tion of two basal blood samples (15-min interval), porcine GIP (Bachem, Torrance, Calif.; 4 pmollkg. h) and glucose (Sigma Chemical Co., St. Louis, Mo.;.6 glkg. h) were intravenously infused, with or without different doses of porcine PYY (Peninsula Laboratories, Belmont, Calif.; 1, 2, or 4 pmollkg. h), for 1 h. Experiment 2. Glucose (.6 g/kg h) was given alone, in the presence or absence of PYY (4 pmollkg. h). Experiment 3. To determine the physiologic range of plasma GIP levels, dogs were given Lipomul (The UpJohn Company, Kalamazoo, Mich.) intraduodenally by means of a constant infusion pump (Harvard Apparatus, South Natick, Mass.) at the rate of 2 ml/min. Plasma GIP levels were measured by a specific radioimmunoassay as given below. Collection of Blood and Laboratory Analysis Blood samples were collected every 15 min into chilled glass tubes containing 15 U/ml of sodium heparin (Organon Inc., West Orange, N.J.) and 1 U/ml of Aprotinin (Sigma). Plasma was separated and stored at -2 C for measurement of insulin or GIP. Plasma insulin levels were determined by a double-antibody radioimmunoassay using an antibody generated in guinea pigs and a porcine standard as described previously (2). The sensitivity of insulin radioimmunoassay was 4 pg/tube, and the IDso value of the insulin antisera for insulin was.5 ng/tube. The insulin antisera did not cross-react with somatostatin, glucagon, pancreatic polypeptide, gastrin, or cholecystokinin. The intraassay and interassay coefficients of variation were 5 ± 2% and 1 ± 3%, respectively. Plasma GIP levels were determined using a double-antibody radioimmunoassay procedure developed by Kuzio and colleagues (21) as previously described (13). Blood samples were also collected in sodium fluoride and potassium oxalate tubes, and glucose concentrations were measured by a Beckman Glucose Analyzer (Beckman Instruments, Fullerton, Calif.). Analysis of Data Insulin and GIP levels (in picograms per milliliter) or glucose concentrations (change in milligrams per deciliter) are expressed as mean ± SEM. The integrated release (from to 6 min) of insulin was calculated from the areas beneath the curve and above the mean basal line as described previously (22) and expressed as nanogramminute per milliliter. The signed-rank test was used to compare the effect of GIP or GIP + PYY on the integrated insulin release (23). Differences with a p value of <.5 were considered statistically significant. Results Effect of Peptide YY On Gastric Inhibitory Polypeptide-Stimulated Insulin Release Intravenous infusion of GIP (4 pmollkg. h) in combination with glucose (.6 g/kg. h) sharply raised plasma insulin levels in dogs. The release of insulin, in response to a continuous infusion of GIP + glucose, was biphasic with a peak increase (3878 ± 936 pg/ml) occurring at 15 min, followed by a phase of sustained secretion (Figure 1A). The 1-h integrated insulin response in control animals (GIP + glucose treated) was 142 ± 32.7 ng-6 minlml - E : G:i > ~ c 2 ::I I/).= <"a E 1 I/) <"a a: A GIP 4 pmollkg-h Glucose.6 g /kg-h PYY 1-4 pmllkg-h N = E... c 8 r-'- 'E 12 * CO I 1 C * *t -.E :; 8 I/)... N ~.= g 8 " Q) >- >- >- >- >- >- a.. a.. a.. -<"a ~, Q) -c a.. a.. a.. a.. a a a Figure 1. A. Incremental plasma insulin level in response to i.v. injection of GIP + glucose alone (e) or in combination with different doses of PYY [1 (), 2 (D.), and 4 () pmol/kg. hj. B. Integrated insulin response to GIP + glucose with or without different doses of PYY (1, 2, and 4 pmollkg. h) and to glucose alone (*p <.5 vs. GIP + glucose alone, tp <.5 vs. GIP + glucose + 1 pmollkg. h of PYY). Blood samples (including basal samples) were collected at 15-min intervals.

3 692 GUO ET AL. GASTROENTEROLOGY Vol. 96, No.3 12 Plasma 8 insulin level (pg/mi) 4 N = 6 o I I B1 B Figure 2. Comparison of the changes in the incremental plasma insulin in dogs in response to glucose (.6 g/kg. h) alone (e) or in combination with PYY (4 pmollkg. h, 6). B 1, B 2, and first and second blood samples were collected at -15,, 5, and 15 min, respectively; other samples were collected at 15-min intervals. (Figure lb). Concurrent intravenous administration of PYY (1, 2, and 4 pmol/kg. h), from to 6 min, caused a significant inhibition of GIP + glucose-stimulated insulin release (Figure 1). The l-h integrated insulin response in PYY-treated dogs was 15 ± 24 ng-6 min/ml (1 pmol/kg. h of PYY), 88 ± 16 ng-6 min/ml (2 pmol/kg. h of PYY), and 79 ± 2 ng-6 min/ml (4 pmol/kg. h of PYY), respectively, all of which were significantly different from the insulin response to GIP + glucose alone (p <.5, Figure lb). The inhibition achieved with 4 pmol/kg. h of PYY was also significantly greater than that observed with 1 pmollkg. h of PYY (p <.5, Figure lb). Effect of Peptide YY on Glucose-Induced Insulin Release Intravenous injection of glucose (.6 g/kg. h) alone also stimulated insulin release (Figure 2). Simultaneous infusion of PYY (4 pmol/kg. h) failed to inhibit glucose-induced insulin secretion (Figure 2). Effect of Peptide YY on Blood Glucose Levels Five minutes after intravenous infusion of glucose (.6 g/kg. h), blood glucose increased 4-1 mg/dl (Table 1), and 15 min later blood glucose levels rose mg/dl. Peptide YY had no significant effect on blood glucose levels achieved either after intravenous administration of glucose alone or in combination with GIP (Table 1). Plasma Gastric Inhibitory Polypeptide Levels After Infusion of Gastric Inhibitory Polypeptide or Lipomul Intravenous infusion of GIP (4 pmol/kg. h) rapidly increased the plasma GIP levels (Figure 3). The peak value of plasma GIP at 6 min was 4493 ± 34 pg/ml. After discontinuing GIP administration, the plasma GIP levels markedly dropped (Figure 3). Intraduodenal infusion with Lipomul (2 ml/min) caused a gradual increase in the levels of plasma GIP. The peak value achieved at 6 min was 433 ± 153 pg/ml, which was similar to the peak level achieved on Lv. infusion of GIP (4 pmol/kg. h) (Figure 3). On cessation of Lipomul infusion, the plasma GIP levels remained high and did not return to basal levels for the period of the present study (Figure 3). Discussion Gastric inhibitory polypeptide is an important, and probably the leading, incretin candidate (1-3). Gastric inhibitory polypeptide is released after ingestion of glucose, galactose, sucrose, fat, and certain amino acids (1). Under normoglycemic conditions, GIP does not cause insulin release, but in the presence of higher than normal glucose concentrations (+ 2 mg/dl above basal), GIP causes significant insulin release, suggesting that the insulinotropic action of GIP is glucose-dependent (1-3). Table 1. Effect of Peptide YY on Blood Glucose Levels Provoked by Intravenous Infusion of Glucose and Gastric Inhibitory Polypeptide Agent G 4 ± 1 19 ± 4 38 ± 4 43 ± 7 48 ± 1 16 ± 1 2 ± 2 G + PYY 4 pmol/kg. h 1 ± 7 32 ± ± 9 48 ± ± 8 11 ± 1 2 ± 2 G + GIP 21 ± 4 32 ± 6 36 ± 5 28 ± 6 13 ± 4 8 ± 5 G + GIP + PYY 2 pmol/kg. h 32 ± 3 4 ± 9 35 ± 9 3 ± 11 7±4 4 ± 3 G + GIP + PYY 4 pmol/kg. h 28 ± 8 37 ± ± 9 51 ± ± 6 9 ± 5 G, glucose; GIP, gastric inhibitory polypeptide; PYY, peptide YY. All values (mean ± SEM) are expressed as change in milligrams per deciliter, n = 6.

4 March 1989 PEPTIDE YY INHIBITS GIP INSULINOTROPIC ACTION 693 GIP 4 pmol/kg-h Lipomul Glucose.6 g/kg-h E Ol a. A -a> > 4 4 ~ Cl. CJ ctl E 2 2 en ctl CL Figure 3. Comparison of the plasma GIP levels in response to intraduodenal infusion of Lipomul (2 mllmin, A) and intravenous administration of GIP (4 pmollkg. h) + glucose (.6 g/kg. h, B). Blood samples were collected at 15-min intervals. In the current study, intravenous infusion of GIP and glucose resulted in a marked elevation of plasma GIP and insulin levels, as expected. The peak level of GIP achieved on Lv. infusion of GIP was similar to that observed by intraduodenal infusion of Lipomul (Figure 3), suggesting that the dose of GIP used in the present study is within the physiologic range. Simultaneous infusion of PYY at dosages of 1, 2, and 4 pmollkg. h, which have also been reported to achieve physiologic levels (12), caused a significant inhibition of GIP-stimulated insulin release in a dose-related fashion. As PYY (4 pmollkg. h) failed to inhibit insulin release in response to i.v. glucose alone, PYY may specifically inhibit the incretin action of GIP. The inhibitory mechanism of PYY action on GIPelicited insulin release is unknown. In a preliminary study, we have observed that pancreatic islets are positive for the presence of specific binding sites for PYY (unpublished data), suggesting that PYY may directly act on islet cells. Another possibility is that PYY increases hepatic extraction of insulin (i.e., facilitates the insulin degradation rate). This possibility remains to be examined through determining the C-peptide levels (1). Besides the current findings, we have recently observed that PYY also depresses 2-deoxy-glucose-, gastrin-releasing peptide-, tetragastrin-, and cholecystokinin-8-stimulated insulin release, but has no effect on vasoactive intestinal peptide- and bethanecol-evoked insulin secretion (16,24). The selective effect of PYY on insulin release provoked by some insulinotropic agents suggests that the inhibitory mechanism of PYY action may be exerted at specific and possibly multiple sites. Ten years ago, Pederson and colleagues (25) found that, although serum GIP levels achieved by infusion of porcine GIP were comparable to those resulting from ingestion of Lipomul, the serum insulin response during GIP infusion was greater than that with oral Lipomul. Additionally, serum insulin response to Lv. glucose plus oral Lipomul was significantly less than the insulin response to oral glucose alone, even though the serum GIP response to triglycerides was considerably greater than that to oral glucose. These experiments indicate that GIP released in response to fat, even at high blood glucose levels, always has a weak ability to stimulate insulin release. This phenomenon has not been explained satisfactorily. Endogenous PYY, at least in part, may be responsible for the reduced response, as fat is a potent stimulator of PYY release as well (11,12). The PYY release in response to fat may inhibit the insulinotropic effect of GIP, as shown in the present study. Fat, on a weight and molar basis, has been shown to be more potent than glucose in releasing GIP (26). Thus, release of GIP could pose a potential teleologic danger to the organism, as it might produce severe hypoglycemia if its release resulted in insulin release. To further this speculation, one might assume that the organism might require protection against the insulinotropic effect of GIP after an oral meal containing fat. One protective mechanism would be that the insulinotropic action of GIP is glucosedependent. Another may be that PYY release, stimulated by fat, may depress the insulinotropic action of GIP, as shown in the present study. In summary, our results indicate that PYY may be a negative feedback factor in the enteroinsular axis. This factor may have an important physiologic significance because it can avoid an exaggerated insulin response to fat that could otherwise cause a dangerous level of hypoglycemia.

5 694 GUO ET A1. GASTROENTEROLOGY Vol. 96, No.3 References 1. Creutzfeldt W, Ebert R. The enteroinsular axis. In: Go VLW, Gardner JD, Brooks FP, Lebenthal E, DeMagno EP, Scheele GA, eds. The exocrine pancreas: biology, pathology and disease. New York: Raven, 1986: Greeley GH Jr. Entero-insulinar axis: incretin candidates. In: Thompson JC, Greeley GH Jr, Rayford PL, Townsend CM Jr, eds. Gastrointestinal endocrinology. New York: McGraw Hill, 1986: Khalil T, Alinder G, Rayford PRo Gastric inhibitory peptide. In: Thompson JC, Greeley GH Jr, Rayford PL, Townsend CM Jr, eds. Gastrointestinal endocrinology. New York: McGraw Hill, 1986: Tatemoto K. Isolation and characterization of peptide YY (PYYJ, a candidate gut hormone that inhibits pancreatic exocrine secretion. Proc Nat! Acad Sci USA 1982;79: El-Salhy M, Grimelius L, Wilander E, et al. Immunocytochemical identification of polypeptide YY (PYY) cells in the human gastrointestinal tract. Histochemistry 1983;77: El-Salhy M, Wilander E, Juntti-Berggren L, Grimelius 1. The distribution and ontogeny of polypeptide YY (PYY)- and pancreatic polypeptide (PP}-immunoreactive cells in the gastrointestinal tract of rat. Histochemistry 1983;78: Leduque P, Paulin C, Dubois PM. Immunocytochemical evidence for a substance related to the bovine pancreatic polypeptide-peptide YY group of peptides in the human fetal gastrointestinal tract. Regul Pept 1983;6: Taylor 11. Distribution and release of peptide YY in dog measured by specific radioimmunoassay. Gastroenterology 1985;88: O'Donohue TL, Chronwall BM, Pruss RM, et al. Neuropeptide Y and peptide YY neuronal and endocrine systems. Peptides 1985;6: Leiter AB, Toder A, Taylor IL, Huntress F, Mandel G, Goodman RH. Identification of the pancreas as the major site of peptide YY mrna synthesis using a cloned cdna probe (abstr). Gastroenterology 1987;92: Aponte GW, Fink AS, Meyer JH, Tatemoto K, Taylor IL. Regional distribution and release of peptide YY with fatty acids of different chain length. Am J PhysioI1985;249:G Pappas TN, Debas HT, Goto Y, Taylor 11. Peptide YY inhibits meal-stimulated pancreatic and gastric secretion. Am J PhysioI1985;248:Gl Guo Y-S, Singh P, Gomez G, Greeley GH Jr, Thompson JC. Effect of peptide YY on cephalic, gastric and intestinal phases of gastric acid secretion and on the release of gastrointestinal hormones. Gastroenterology 1987;92: Guo Y-S, Fujimura M, Lluis F, Tsong Y, Greeley GH Jr, Thompson JC. Inhibitory action of peptide YY on gastric acid secretion. Am J PhysioI1987;253:G Lluis F, Fujimura M, Guo Y-S, Gomez G, Greeley GH Jr, Thompson Je. Peptide YY is an important regulation of pancreatic exocrine secretion. Surg Forum 1985;36: Guo Y-S, Singh P, DeBouno JF, Thompson JC. Effect of peptide YY on insulin release stimulated by 2-deoxy-glucose and neuropeptides in dogs. Pancreas 1988;3: Lundberg JM, Tatemoto K, Terenius L, et al. Localization of peptide YY (PYY) in gastrointestinal endocrine cells and effects on intestinal blood flow and motility. Proc Natl Acad Sci USA 1982;79: Suzuki T, Nakaya M, Itoh Z, Tatemoto K, Mutt V. Inhibition of interdigestive contractile activity in the stomach by peptide YY in Heidenhain pouch dogs. Gastroenterology 1983;85: Al-Saffar A, Hellstrom PM, Nylander G. Correlation between peptide YY-induced myoelectric activity and transit of smallintestine contents in rats. Scand J GastroenteroI1985;2: Greeley GH Jr, Thompson Je. Insulinotropic and gastrinreleasing action of gastrin-releasing peptide (GRP). Regul Pept 1984;8: Kuzio M, Dryburgh JR, Malloy KM, Brown JC. Radioimmunoassay for gastric inhibitory polypeptide. Gastroenterology 1974;66: Thompson JC, Reeder DD, Bunchman HH, Becker HD, Brandt EN. Effect of secretin on circulating gastrin. Ann Surg 1972; 176: Elashoff JD. Down with multiple t-tests. Gastroenterology 1981;8: Guo Y-S, Gomez G, Thompson JC, Singh P. Feedback regulation by peptide YY of enteroinsular axis (abstr). Gastroenterology 1988;94:A Pederson PA, Schubert HE, Brown JC. Gastric inhibitory polypeptide. Its physiologic release and insulinotropic action in the dog. Diabetes 1975;24: Krarup T, Holst n, Larsen K1. Responses and molecular heterogeneity of IR-GIP after intraduodenal glucose and fat. Am J PhysioI1985;249:E Received February 8, Accepted October 11, Address requests for reprints to: Dr. Pomila Singh, Associate Professor, The University of Texas Medical Branch, Department of Surgery, 6.22 Old John Sealy, Route E32, Galveston, Texas This work was supported by grants from the National Institute of Health (ROl DK 15241, POl DK 3568, and CA 38651). The authors thank Kay Smith for assistance in the preparation of this manuscript, and Robert Washington, Perry Orchard, and Azar Owlia for technical assistance.