Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Release and metabolism of gastric inhibitory polypeptide Kieffer, Timothy James 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-894199.pdf [ 4.54MB ]
JSON: 831-1.0088076.json
JSON-LD: 831-1.0088076-ld.json
RDF/XML (Pretty): 831-1.0088076-rdf.xml
RDF/JSON: 831-1.0088076-rdf.json
Turtle: 831-1.0088076-turtle.txt
N-Triples: 831-1.0088076-rdf-ntriples.txt
Original Record: 831-1.0088076-source.json
Full Text

Full Text

RELEASE AND METABOLISM OF GASTRIC INHIBITORYPOLYPEPTIDEbyTIMOTHY JAMES KIEFFERB.Sc., The University of British Columbia, 1989A THESIS SUBMrrIED iN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFDOCTOR OF PHILOSOPHYinTHE FACULTY OF GRADUATE STUDIESDepartment of PhysiologyWe accept this thesis as conformingto the required standardTHE UNWERS1TY OF BRITISH COLUMBIAApril 1994© Timothy James Kieffer, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.(Signature)_________________Department of_______________The University of British ColumbiaVancouver, CanadaDate t’tDE-6 (2/88)11ABSTRACTThis thesis reports methodology that was developed to isolate and enrich intestinalendocrine cell preparations by centrifugal elutriation and short-term culture to enable thestudy of the regulation of IRGIP secretion at the cellular level. Adherent canine epithelialcell cultures contained —10% IRGIP cells. The release of IRGIP from these cells wassignificantly increased by incubation with depolarizing concentrations of K, glucose,somatostatin immunoneutralizing antibody, or GRP, in addition to pharmacologicalelevations of intracellular cAMP or Ca2. It is concluded that this method provides ameans of further investigating the cellular mechanisms controlling GIP release.Enteroendocrine tumor cell lines were also investigated as an alternate source ofGIP cells. A cell line derived from intestinal tumors of transgenic mice (STC- 1) was sub-cloned to produce a cell line with —30% IRGIP (STC 6-14). HPLC of STC 6-14 extractsindicated that the tumor cell derived IRGIP eluted with synthetic porcine GTP 1-42.Release of IRGIP from STC 6-14 cells; increased in a concentration dependent fashion inresponse to glucose, was augmented by the addition of somatostatin neutralizingantibody, and attenuated by exogenous somatostatin. Immunoreactive somatostatin(IRSS) release was significantly increased by adding GIP to the incubation medium. It isconcluded that this cell line represents a means of rapidly obtaining large numbers of GIPcells, and thus should be useful to investigate stimulus-secretion coupling in the GIP cell.GIP secreted by STC 6-14 cells was metabolized by a serum constituent tobiologically inactive GIP 3-42. ‘251-GIP was purified by HPLC and used as a highlysensitive means to further investigate the degradation of GIP by serum. The removal ofthe N-terminal dipeptide by serum could be blocked by diprotin A, a competitive inhibitorof dipeptidyl peptidase IV (DPP IV). No GIP 3-42 was produced by incubation of GIPwith serum from rats specifically lacking DPP IV. Infusion of‘251-GIP into rats, followedby HPLC analysis, indicated that 50% was metabolized to‘251-GIP 3-42 by —1.5 mm. It isconcluded that DPP IV is a primary degradative and inactivating enzyme of GIP.111TABLE OF CONTENTSABSTRACT iiTABLE OF CONTENTSLIST OF FIGURES viiiACKNOWLEDGEMENTS xiiiCHAPTER 1iNTRODUCTION1.1 OVERVIEW 11.2 DISCOVERY OF GIP AS AN ENTEROGASTRONE1 .2. 1 Enterogastrone concept 21 .2.2 Evidence for existence of OW 21.2.3 IsolationofGlP 31 .2.4 Glucagon superfamily 41.3 ENTEROGASTRONE ACTIONS OF GIP1 .3. 1 Effect of GIP on gastric secretion 51 .3.2 Mechanism of gastric action of GIP 51 .3 .3 Interaction with other enterogastrones 71.4 DISCOVERY OF GIP AS AN INCRETIN1 .4. 1 Incretin concept 81 .4.2 Enteroinsular axis 91 .4,3 Evidence for GIP as an incretin 101.5 INSULINOTROPIC ACTIONS OF GIP1 .5. 1 Effect of OP on insulin secretion 111 .5 .2 Other potential incretins 121 . 5 . 3 Interactions between insulin secretagogues 131 .5.4 Relative contribution of GIP vs tGLP-I 14iv1 .5.5 Participation of GIP and tGLP-I in an enteroendocrine loop 151 .5.6 Mechanism of action of GIP on the B-cell 161 .5 .7 Nature of glucose dependency 181.6 OTHERACTIONSOFG1P 191.7 GJPGENE1 .7. 1 Gene structure and post-translational processing 201 .7.2 Gene regulation 221.8 GIP CELL DISTRIBUTION AND CHARACThRIZATION 221.9 MEASUREMENT OF GIP1 .9. 1 Development of GIP radioimmunoassay 241 .9.2 Physiological levels of GIP 241 .9.3 Immunoreactive forms of GIP 251.10 GIP SECRETION1. 10. 1 Response to a mixed meal 271 . 10.2 Response to carbohydrate 281.10.3 Response tofat 321 . 10.4 Response to protein 341. 10.5 ANS control of GIP release 361 . 10.6 Other modulators of GIP release 381. 10.7 Feedback inhibition of GIP release 391.11 GIP RECEPTORS1. 11 . 1 Biologically active site(s) 421 . 11 .2 OW binding studies 441.12 OP METABOLISM 471.13 GIP PATHOPHYSIOLOGY 491.14 THESIS INVESTIGATION 52VCHAPTER 2GASTRIC INHIBITORY POLYPEPTIDE RELEASE FROM ISOLATEDCANINE AND PORCINE ENDOCRINE CELLS2.1 INTRODUCTION 542.2 MATERIALS and METHODS2.2. 1 Animal tissues 542.2.2 Isolation of mucosal cells 552.2.3 Enrichment of mucosal endocrine cells 552.2.4 Endocrine enriched cell culture 562.2.5 Immuocytochemistry 572.2.6 Release experiments 582.2.7 Secretagogues 592.2.8 Radioimmunoassay for GIP 602.2.9 Radioimmunoassay for SS 632.2. 10 Expression of results 642.3 RESULTS2.3. 1 Mucosa digestion 652.3.2 Endocrine cell enrichment and culture 652. 3 .3 ICC of intact mucosa and cultured cells 672.3 .4 IRGIP secretion in response to K 672. 3 .5 IRGIP secretion in response to glucose 682.3 .6 IRGIP secretion in response to A23 187 682.3 .7 IRGIP secretion in response to GRP 692.3.8 IRGIP secretion in response to forskolin 692.4 DISCUSSION 83viCHAPTER 3GASTRIC INHIBITORY POLYPEPTIDE RELEASE FROM A TUMOR-DERIVED CELL LINE3.1 INTRODUCTION 893.2 MATERIALS and METHODS3.2. 1 Culture and production of STC 6-14 cell line 903 .2.2 Release experiments 913.2.3 Secretagogues 923.2.4 Peptide quantification 933.2.5 Immunocytochemistry 933.2.6 HPLC 933 .2.7 Expression of results 943.3 RESULTS3.3. 1 Characterization of STC 6-14 cells 943 .3 .2 IRGIP secretion in response to glucose 953 . 3 .3 IRSS secretion in response to glucose and GIP 963.4 DISCUSSION 103CHAPTER 4GASTRIC INHIBITORY POLYPEPTIDE METABOLISM BY DIPEPTIDYLPEPTifiASE IV4.1 INTRODUCTION 1074.2 MATERIALS and METHODS4.2. 1 Incubation of GIP with serum 1084.2.2 Dipeptidyl peptidase IV-negative rats 1094.2.3 Dipeptidyl peptidase IV assay 1094.2.4 HPLC purification of‘251-GIP 110vii4.2.5 Incubation of ‘I-GW with serum .1114.2.6 In vivo experiments with 125-GIP 1124.2.7 Dipeptidyl peptidase IV in Zucker rats and NIDDM subjects 1134.3 RESULTS4.3. 1 Incubation of GIP with serum 1134.3 .2 Dipeptidyl peptidase IV assay 1144.3 .3 Label specific activity 1144.3.4 Incubation of purified ‘I-GIP with Wistar rat serum 1154.3.5 Incubation of purified ‘I-GIP with DPP IV-negative rat serum 1164.3.6 Incubation of purified ‘I-GIP with CEP 1174.3 .7 The analysis of‘25I-GIP degradation in vivo 1174.4 DISCUSSION 144CHAPTER 5SUMMARYSUMMARY AND FUTURE DIRECTIONS 151REFERENCES 160vi”LIST OF FIGURES1 Amino acid sequences of porcine, human, bovine, and rodent gasthcinhibitory polypeptide 42 Sequence of a cDNA encoding the human GIP precursor 213 A) IRGIP and IRSS content from the canine single cell suspension priorto elutriation and the two fractions resulting from elutriation.B) IRGIP content from isolated porcine cell fractions 714 Canine jejunum immunostained for GIP using the peroxidase method 725 Isolated canine epithelial cells after 40 h culture, immunostained for GIP. . . . 736 Isolated canine epithelial cells after 40 h culture, immunostained forsomatostatin 747 Canine jejunum stained for neutral and acid mucins using a combinedalcian blue-periodic-acid Schiff technique 758 Isolated canine epithelial cells after 40 h culture, stained for neutral andacid mucins using a combined alcian blue-periodic-acid Schiff technique. . . . 769 Effect of K concentrations of 10 to 55 mM on basal IRGIP release (5 mMK+) from isolated canine and porcine endocrine cells in the presence of 5mM glucose 7710 Concentration-response relationship between glucose (5 to 20 mM) andLRGIP secretion from isolated canine epithelial cells 7811 Concentration-response relationship between glucose (5 to 20 mM) andIRGIP secretion from isolated porcine epithelial cells 7912 Effect of 1, 5, and 10 IIM A23 187 on basal release of IRGIP in thepresence of 5 mM glucose 80ix13 Effect of graded concentrations of GRP (0 to 100 nM) on IRGIP releasefrom cultured canine epithelial cells 8114 Effect of 0.1 to 100 .tM forskolin in the presence of 5 mM glucose onIRGIP release from isolated canine epithelial cells 8215 Cultured STC 6-14 cells immunostained for GIP 9716 HPLC elution profile of natural porcine GIP and synthetic porcine GIPfrom a C18 column with an acetonitrile gradient of 28-33% over 10 mm 9817 IRGIP content of HPLC fractions:A) Overlay of elution profiles for natural porcine GIP and syntheticporcine GW.B) Overlay of elution profiles for STC 6-14 cell extract and culturemedium 9918 Effect of 250 anti-somatostatin antibody (SOMA-lO) or 10 nMsomatostatin on the dose-response relationship between glucoseconcentrations (5 to 20 mM) and IRGIP secretion from STC 6-14 cells . . . . 10019 Concentration-response relationship between glucose concentrations(5 to 20 mM) and IRSS secretion from STC 6-14 cells 10120 Effect of graded concentrations of porcine GTP (0-100 nM) on IRSSrelease from STC 6-14 cells in the presence of 5 mM glucose 10221 Determination of relative DPP IV levels in Fischer DPP TV-positive) andDPP TV-negative rats from rate of production of p-nitroaniline.A) Standard curve of absorbance for increasing concentrations of pnitroaniline (mM)B) Plot of p-nitroaniline concentration vs. time (mm) deterimined fromabsorbance values converted to mM p-nitroaniline from curve A 119x22 Determination of the relative levels of DPP IV in NIDDM and normalsubjects from the production of p-nitroaniline after 45 mm 12023 Profile of synthetic porcine GIP iodination mixture eluted on Sephadex G15 12124 A) % bound for a GIP standard curve using increasing GIP concentrations,and a self displacement curve using increasing amounts of ‘I-GIP.B) Plot of mass of GIP standard vs.1I-GIP radioactivity as deteriminedfrom A) for a number of points 12225 A) HPLC elution profile of ‘I-GIP at 32% acetonitrile for 10 mm,followed by a gradient to 38% from 10 to 20 mm.B) HPLC elution profile of peak 2 after 25 h incubation at 37°C 12326 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistarrat serum (10%) at 37°C.A) duration = 10 mm. B) duration = lh 12427 Both figures are HPLC elution profiles of‘251-GJP incubated with Wistar ratserum (10%) at 37°C. A) duration= ih. B) duration = 12528 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistar ratserum (10%) at 37°C. A) duration= Iih. B) duration = 24Ji 12629 Both figures are HPLC elution profiles of 125-GIP incubated with Wistar ratserum (10%) at 4°C. A) duration = 4..h. B) duration = 12730 Both figures are HPLC elution profiles of 125-GIP incubated with Wistar ratserum (10%) at 4°C. A) duration =.19J B) duration = 24Ji 12831 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistarrat serum (10%) and diprotin A (0.1 mM) at 37°C. A) duration =B) duration = 6h 129xi32 Both figures are HPLC elution profiles of‘251-GJP incubated with Wistarrat serum (10%) and diprotin A (0.1 mM) at 37°C. A) duration= jh.B) duration = 13033 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistarrat serum (10%) and aprotinin (2%) at 37°C. A) duration= i.h.B) duration= 13134 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistarrat serum (10%) and aprotinin (2%) at 37°C. A) duration= jjj1.B) duration 2j1 13235 Both figures are HPLC elution profiles of 125-GIP incubated with Wistarrat serum (10%) and bacitracin (50 U/mi) at 37°C. A) duration= 2.h.B) duration = 19.5 h 13336 Both figures are HPLC elution profiles of‘251-GIP incubated with Wistarrat serum (10%) containing diprotin A (0.1 mM), aprotinin (2%) andbacitracin (50 U/mi), at 37°C. A) duration = 3.25 h. B) duration= 2h.... 13437 Both figures are HPLC elution profiles of 125-GIP incubated with DPP IV-negative rat serum (10%) at 37°C. A) duration= 2h. B) duration = lb.... 13538 Both figures are HPLC elution profiles of 125-GJP incubated with DPP IV-negative rat serum (10%) and aprotinin (2%) at 37°C. A) duration= 4...h.B)duration=.j2 13639 Both figures are HPLC elution profiles of 125-GIP incubated with DPP IV-negative rat serum (10%) and aprotinin (2%) at 37°C. A) duration= lib.B)duration=27h 13740 Both figures are HPLC elution profiles of‘251-GIP incubated with charcoalextracted human plasma (10%) used in the GIP MA, at 37°C.A) duration=j. B) duration= 138xii41 Both figures are HPLC elution profiles of 125-GIP incubated for 1 h withWistar rat serum (neat) containing diprotin A (0.1 mM), aprotinin (2%),purified by SepPak, lyophilized, and reconstiuted.A) incubation on ice. B) incubation at 37°C 13942 Both figures are HPLC elution profiles of 125-GIP infused into a Wistarrat, collected in 2 ml serum with diprotin A (0.1 mM) and aprotinin (2%),purified by SepPak, lyophilized and reconstituted.A) duration = 2 mm. B) duration =5 mm 14043 Both figures are HPLC elution profiles of‘25101P infused into a Wistarrat, collected in 2 ml serum with diprotin A (0.1 mM) and aprotinin (2%),purified by SepPak, lyophilized and reconstituted.A) duration = 10 mm. B) duration =20 mm 14144 Both figures are HPLC elution profiles of‘251-GTP infused into a DPP IV-negative rat, collected in 2 ml serum with diprotin A (0.1 mM) andaprotinin (2%), purified by SepPak, lyophilized and reconstituted.A) duration = 2 mm. B) duration = 10 mm 14245 Graphs summarizing the conversion rate of‘251-GIP 1-42 to‘251-GIP 3-42both in vitro and in vivo calculated from peak areas of Figures 5 to 22,plotted as % GIP 3-42 (% of GIP 1-42 + 3-42) vs. time.A) Relative %‘251-GIP 3-42 production in Wistar rat serum at 37°C(normal serum), + diprotin A (0.1 mM), + bacitracin (50 U/mI), or @ 4°C.B) Rate of formation of 125-GIP 3-42 from‘25101P 1-42 infused in aWistar rat 143xli’ACKNOWLEDGEMENTSI was introduced to the U.B.C. Physiology Department in 1988 by Dr. JohnBrown, and was fortunate to obtain a summer studentship position working for him. Ithank Dr. Brown for cultivating my interest in research and aiding in my development asa scientist. Other members of the then MRC Regulatory Peptide Group were also equallyimportant in my training. Dr. Kenny Kwok emphasized the need for ‘controls’ and hispharmacology background was helpful. Dr. Alison Buchan guided me throughimmunocytochemical techniques and the isolation of gut endocrine cells. Dr. ChrisMcIntosh was an endless source of answers, and directed my work with the HPLCsystem. Finally, my supervisor, Dr. Ray Pederson, continually guided my research andsupported me throughout my degree. He also provided annual ‘retreats’ to Mayne Island,which I will always remember. Throughout my degree, I received a great deal oftechnical support, particularly from Haidee Barker, Dr. Zhongxian Huang, NarinderDhatt, Leslie Checknita, Marie Langton, John Sanker and Joe Tay. Fellow graduatestudents were also a large part of my experience, especially Glenn Morrow who providedconstant feedback on experiments, and the odd ‘one-on-one’ challenge. Finally, I wouldlike to thank my family and Stephanie for continual support and encouragement.1CHAPTER 1INTRODUCTION1.1 OVERVIEWIn 1902, Bayliss and Starling noted that water and bicarbonate secretion from thepancreas increased when weak acid was introduced into the duodenum, but not whenadministered intravenously. Furthermore, the intravenous injection of an acid extract ofupper intestinal mucosa resulted in copious secretion from the pancreas. Thus, thepresence of “secretin” was proposed. Bayliss and Starling also searched for a generalname that would convey the meaning of “chemical messenger”. It was W. B. Hardy thatproposed the name “hormone” derived from the Greek word for “I arouse to activity”, andalthough it did not suggest the property of messenger, it was adopted. The word firstappeared in print in Starling’s Croonian Lecture of 1905.The original concept- that a chemical substance called a hormone, liberated byone kind of cell, and carried by the blood stream to act on a distant target cell -represented a major advance in physiological thinking. Hormones are principal players inthe maintenance of physiological homeostasis, as perturbation of this equilibrium leads totheir appropriate corrective secretion. These responses are monitored and feedback loopsexist to inhibit further stimulation of hormone release in a regulatory fashion. To beeffective, this feedback must be accompanied by the clearance of previously secretedhormone from the circulation through enzymatic inactivation and/or removal of thesubstance by the liver or kidneys.Investigation of a hormone first requires isolation and identification of the activeprinciple in order that the biological response be reproduced by administration ofexogenous hormone. The demonstration of an endocrine-mediated biological responsethen allows examination of other components of the system. Important physiologicalquestions are left outstanding without knowledge of stimulus-secretion coupling,2circulating hormone levels, and duration of action and metabolism in plasma. The goalof these investigations was to further understand the secretion and metabolism of thehormone GIP (Gastric Inhibitory Polypeptide).1.2 DISCOVERY OF GIP AS AN ENTEROGASTRONE12.1 ENTEROGASTRONE CONCEPTIt was first reported in 1886 by Ewald and Boas that the addition of olive oil to atest meal produced an inhibition of gastric secretion in human subjects. In 1910, Pavlovdemonstrated that fat added to a meal fed to dogs inhibited the secretion of acid andpepsin. In 1926, Farrell and Ivy made the observation that the acid response to a mealcould be inhibited by injection of extracts of the duodenal mucosa and suggested that ahormonal mechanism must be involved. Kosaka and Lim (1930) proposed the term“enterogastrone” to describe a putative hormone which was secreted in response to fat orits digestive products in the intestinal lumen and inhibited gastric acid secretion.12.2 EVIDENCE FOR EXISTENCE OF GIPTwo candidate intestinal hormones apparently satisfying the requirements forenterogastrone were secretin (Bayliss and Starling, 1902) which was isolated for itsability to stimulate pancreatic secretion, and cholecystokinin (CCK; Ivy and Oldberg,1927), a potent stimulator of gall bladder contraction. Impure preparations of both werefound to have acid inhibitory activity in the vagally denervated (Heidenhain) caninegastric pouch (Gillespie and Grossman, 1964). However, in apparent conflict with thedescribed acid inhibitory effects of CCK, Magee and Nakamura (1966) observed thatCCK preparations were capable of stimulating acid secretion under fasting conditionswhen administered intravenously. Brown and Pederson (1970a) tested the hypothesisthat the gastric effects resulted from the actions of factors other than CCK present in theimpure preparations. The acid secretory effects of two different preparations of CCK3designated 10% and 40% pure on the basis of gallbladder stimulating potency werecompared in dogs prepared with Heidenhain pouches. The 40% pure preparationproduced a greater stimulatory effect on acid secretion than the 10% pure preparation atdoses that yielded comparable gallbladder activity. It was proposed that either a gastricstimulant had been concentrated or an inhibitor of acid secretion had been removedduring the purification procedure. Support for the latter was provided in 1971 (Pederson)when it was demonstrated that the 10% pure preparation of CCK was a more potentinhibitor of pentagastrin stimulated acid secretion than the 40% pure preparation.12.3 ISOLATION OF GIPThe canine Heidenhain pouch was used as a bioassay model for the isolation ofthe putative inhibitor of acid secretion. Brown et al. (1969, 1970) purified the activesubstance from extracts of hog duodeno-jejunal mucosa to a degree of homogeneitysuitable for amino acid analysis to be performed, and named it gastric inhibitorypolypeptide, or GIP (Brown, 1971; Brown and Dryburgh, 1971). Pure GIP was shown tobe a potent inhibitor of gastric acid and pepsin secretion in the dog (Pederson and Brown,1972). The amino acid sequence of the porcine peptide, as determined by Brown andDryburgh (1971) and later corrected by Jörnvall eta!. (1981) is shown in Figure 1, alongwith that of human, bovine, and rat GIP. The high degree of conservation of the peptideamong species suggests that GIP has an important physiological role.41 2 3 4 5 6 7 8 9 10 11 12 13 14PORCINE: Tyr—Aia—Giu—Gly—Thr—Phe—Ile—Ser—Asp—Tyr—Ser—Ile—Ala-MetHUMAN:BOVINE:RODENT:15 16 17 18 19 20 21 22 23 24 25 26 27 28PORCINE: Asp-Lys-Iie-Arg-Gln-Gln-Asp-Phe-Val-Asn-Trp-Leu-Leu-AiaHUMAN: -His-BOVINE:RODENT:29 30 31 32 33 34 35 36 37 38 39 40 41 42PORCINE: Gin-Lys-Giy-Lys-Lys-Ser-Asp-Trp-Lys-His-Asn-Ile-Thr-GinHUMAN: -AsnBOVINE: -lie-RODENT: -Asn- -LeuFIGURE 1: Amino acid sequences of porcine, human, bovine, and rodent gastricinhibitory polypeptide (GIP). From Jömvall et a!., 1981 (porcine); Moody et at., 1984(human); Carlquist et at., 1984 (bovine) and Sharma et at., 1992 (rat).12.4 GLUCAGON SUPERFAMILYThe sequence of GIP, as well as the structure of its precursor peptide (Takeda etat., 1987) and the gene encoding it (Inagaki et at., 1989) indicate that GIP belongs to theglucagon superfamily of peptides, which share sequence homologies. This familyincludes glucagon, glicentin, the glucagon-like peptides (GLP-I and GLP-II), secretin,vasoactive intestinal peptide (VIP), peptide histidine-methionine (PHM), peptidehistidine-isoleucine (PHI), peptide histidine-valine (PHV), growth hormone releasingfactor (GRF), pituitary adenylate cyclase activating peptides (PACAP-38 and PACAP27), helodermin, helospectins (I and II), extendins (3 and 4), and somatoliberin (Arimura,1992; Eng et at., 1990; Parker et a!., 1984; Raufman et at., 1991; Rai et at., 1993;Robberecht et a!., 1991). It has been suggested that these hormones probably arose froma common ancestral gene (Bell, 1986; Campbell and Scanes, 1992).51.3 ENTEROGASTRONE ACTIONS OF GIP13.1 EFFECT OF GIP ON GASTRIC SECRETIONDogs surgically prepared with Heidenhain pouches of the body of the stomachwere used to determine the acid inhibitory action of GJP and served as a bioassay for itsisolation. Using this vagally and sympathetically denervated gastric pouch preparation,GIP dose-dependently inhibited pentagastrin-stimulated acid secretion, with the highestdose able to produce —80% inhibition (Pederson and Brown, 1972). However, in thevagally innervated gastric remnant, the same authors observed only —40% inhibition ofacid secretion induced by insulin hypoglycemia. Subsequent studies in man (Maxwell etal., 1980) and rat (El-Munshid et al., 1980) also found GIP to be a weak inhibitor ofpentagastrin-stimulated acid secretion when innervation was intact, and it was speculatedthat GIP may not be physiologically important as an enterogastrone.In a comparative study in the dog stomach, Soon-Shiong et al. (1979) reproducedthe profound inhibition of pentagastrin-stimulated acid secretion observed earlier in theHeidenhain pouch but also found Gil’ had only weak effects on the innervated gastricremnant. Yamagishi and Debas (1980) noted that if oleic acid was introduced into theduodenum, a GIP infusion was able to completely inhibit acid secretion in the innervateddog stomach in response to a meal. The possible involvement of a cholinergicmechanism antagonistic to the action of GIP on the stomach was suggested by theobservation that the acid inhibitory effect of this hormone in the denervated gasthc pouchof the dog could be blocked by the intravenous infusion of urecholine (Soon-Shiong etat., 1979). It therefore seemed possible that GIP exerted its inhibitory effect on theparietal cell indirectly, via the release of an inhibitor under cholinergic control.1.3.2 MECHANISM OF GASTRICACTIONOF GIPIn 1981, McIntosh et al. suggested that gastric somatostatin might mediate theacid inhibitory action of GJP. In testing this hypothesis, they found a strong stimulation6of immunoreactive somatostatin (IRSS) release from the perfused stomach by GIP. Inaddition, both acetyicholine and vagal stimulation potently inhibited thesomatostatinotropic activity of GIP. These data accounted for the observation of weakacid inhibitory activity of GIP in the innervated stomach. McIntosh et al. (1981)proposed a model whereby GIP released by the presence of fat in the duodenumstimulated somatostatin secretion upon reaching the stomach. However, thecholinergically or neurally mediated inhibition of GIP-stimulated IRSS secretion in therat was only partially blocked by atropine (McIntosh et al., 1979). Naloxone partiallyreversed the vagally induced inhibition of GIP-stimulated IRSS release, suggesting thatopioid peptides were released and also inhibited IRSS release during the period of vagalstimulation (McIntosh et a!., 1983). Subsequent studies have demonstrated thatnoncholinergic transmitters, including the opioid peptides (enkephalins, dynorphins) andthe tachykinins (substance P, neurokinin A), also inhibit somatostatin secretion andreverse the actions of GIP on the stomach (Kwok et a!., 1988a; McIntosh et al., 1989;1990).In order for GIP to act as a physiological inhibitor of acid secretion, a mechanismcapable of antagonizing the parasympathetic influence would be required. McIntosh eta!. (1981) suggested that in addition to stimulating GIP secretion, fat in the duodenummay activate sympathetic fibres which provide such control. Evidence for this hypothesiswas supplied by the observation that catecholamines stimulate gastric IRSS release in theperfused stomach (McIntosh etal., 1981; Goto etal., 1981; Koop etal., 1981). McIntoshet a!. (1981) also observed a marked increase in IRSS with stimulation of preganglionicsympathetic nerves together with simultaneous infusion of atropine. Alternatively,Brown et a!. (1989) suggested that the inhibitory effect of GIP may be most profoundwhen the parasympathetic activity of the stomach is minimal, such as the interdigestiveperiod, when OP levels have been shown to remain elevated (Jorde et a!., 1980; Salera et7a!., 1983). While the mechanism of action is still not completely understood, GIPappears to fulfill the requirements of an enterogastrone.13.3 INTERACTION WITH OTHER ENTEROGASTRONESSequencing of the preproglucagon gene by two separate groups (Bell et al., 1983;Lopez et at., 1983) revealed that the precursor peptide contained, in addition to glucagonand glicentin, the sequence of two glucagon-like peptides (GLP-I and GLP-II). Mojsovet at. (1986) showed that post-translational processing of the preprohormone yielded aC-terminal 30 residue form of GLP-I, with amino acids 7-36-NH2 (tGLP-I). As amember of the glucagon superfamily (section 1.2.4), tGLP-I has a high degree ofhomology with GIP, sharing 11 amino acids. Like GIP, tGLP-I has also been shown toinhibit pentagastrin and meal stimulated acid secretion in man (Schjoldager et a!., 1989;O’Halloran et a!., 1990; Wettergren et al., 1993). In a seeming paradox, tGLP-Istimulated H production in an enriched preparation of isolated rat parietal cells(Schmidler et al., 1991). It was suggested that in vivo, direct stimulation by tGLP-I of theparietal cells might be counterbalanced by indirect inhibitory mechanisms that areexcluded in the in vitro cell system. Like GIP, tGLP-I may act indirectly via somatostatinrelease, as tGLP-I has been reported to stimulate somatostatin release in the isolatedperfused rat stomach (Eissele et at., 1990). It is quite likely, therefore, that GIP fulfillsthe enterogastrone role by acting in concert with other peptides, such as CCK, secretin,and tGLP-I, in addition to nervous inhibitory mechanisms that may be initiated during thedigestive and absorptive process in the upper small bowel.Nauck et al. (1992) recently questioned whether GIP and tGLP-I were in factphysiological enterogastrones in humans. These investigators examined the effects ofintravenous human GIP and/or tGLP-I on pentagasthn-stimulated gastric volume andacid output. Pentagastrin significantly stimulated acid output, but neither GIP nor tGLPI, either alone or in combination, reduced pentagastrin-stimulated gasthc acid secretion.8It was therefore concluded that these hormones were not likely enterogastrones in man.However, these authors did not acknowledge the data from the previous decade whichclearly indicated that the enterogastrone effects of these hormones are likely mediated bygastric somatostatin, which is also under neural control. Thus, in the absence ofintraduodenal stimuli (i.e. not a physiologically relevant design), there may be nosympathetic activity to antagonize the parasympathetic inhibition of somatostatin release,and thus no reduction in gastric acid secretion (see section 1.3.2). Thus, to date, the roleof GIP as an enterogastrone in man is unresolved.1.4 DISCOVERY OF GIP AS AN INCRETIN1.4.1 INCRETIN CONCEPTIn 1902, Bayliss and Starling speculated that signals arising in the gut afteringestion of nutrients may elicit endocrine responses and affect the disposal ofcarbohydrates, as well as stimulating pancreatic bicarbonate secretion. This came as aresult of their discovery of secretin as a regulator of pancreatic secretion. Moore et al.(1906) postulated that the duodenum produced a “chemical excitant” for pancreaticsecretion, the absence of which caused diabetes. Their attempted treatment of diabetesmellitus by injection of gut extracts was unsuccessful. Zunz and Labarre (1929) andLabarre and Still (1930) prepared an intestinal extract free of secretin activity which wasable to produce hypoglycemia in dogs. Thus Labarre (1932) introduced the term“incretin” to describe humoral activity of the gut that might enhance the endocrinesecretion of the pancreas.Interest in the search for the active principle in the duodenum waned with theisolation of insulin, allowing the effective control of blood glucose in diabetes mellitus(Banting et a!., 1922). Also, studies in 1940 by Loew et a!. questioned the existence ofgastrointestinal factors with insulin-releasing potency. It was not until the developmentof a reliable radioimmunoassay for insulin in the 1960’s by Berson and Yalow, allowing9measurement of circulating levels of this hormone, that interest was renewed. It becameapparent that the direct effect of nutrients and their metabolites on the islets could notadequately account for the amount of insulin released. Absorption of nutrients from thegut seemed to be accompanied by the release of additional factors. The insulin responseto intravenous glucose was much smaller than to either oral glucose (Elrick et al., 1964)or intrajejunal infusion (Mcintyre et a!., 1964), even though there was a greater increasein blood glucose levels with intravenous administration. It was concluded from theseobservations that a humoral substance was released from the jejunum during glucoseabsorption which acted to stimulate insulin release. Perly and Kipnis (1967) estimatedthat as much as half of the insulin secreted following an oral glucose load was a result ofgastrointestinal factors.1.4.2 ENTEROINSULAR AXISIn 1969, Unger and Eisentraut named the connection between the gut and thepancreatic islets described above the “enteroinsular axis”. While Labarre had noted anincretin or hormonal effect from the gut on insulin release (1932), he had provided earlierevidence that insulin could be released by vagal stimulation (Labarre, 1927). It becameapparent that the gastrointestinal neural system might participate in the enteroinsular axisnot only by direct stimulation of the islets, but also by regulating the release of an incretininto the blood. As it was clear that the pancreas played a major role in maintainingcirculating glucose levels, it also seemed logical that nutrients, like glucose, were able toact directly on islets to control hormone release. Creutzfeldt (1979) therefore suggestedthat the enteroinsular axis encompass nutrient, neural and hormonal signals from the gutto the islet cells secreting insulin, glucagon, somatostatin, or pancreatic polypeptide.Furthermore, Creutzfeldt (1979) defined the criteria for fulfillment of the hormonal orincretin part of the enteroinsular axis as: 1) it must be released by nutrients, particularly10carbohydrates, and 2) at physiological levels, must stimulate insulin secretion in thepresence of elevated blood glucose levels.1.4.3 EVIDENCE FOR GIP AS AN INCRETINIn 1966, Dupré and Beck demonstrated that a crude preparation of CCK possessedinsulinotropic activity. In 1972, Rabinovitch and Dupré found that this insulinotropicaction could be removed by further purification of the CCK. This resembled the loss ofthe acid inhibitory activity previously reported by Brown and Pederson (1970a) in thepurification of GIP from CCK, and led Dupré to the hypothesis that Gil’ may haveinsulin-releasing capabilities. By 1973, Dupré et al. demonstrated that a purifiedpreparation of GIP infused intravenously in humans in concert with glucose, resulted inthe stimulation of insulin release and an improvement in glucose tolerance. The insulinresponse was sustained for the duration of the GIP infusion and was not observed in theeuglycemic state.While no specific GIP antagonists have been developed to further establish thephysiological contribution of GIP to the postprandial insulin response, antisera to GIPhave been produced which may suppress its biological activity. One of the first attemptsat GIP immunoneutralization study was by Lauritsen et al. (1981). Insulin release wasmeasured in rats given an oral glucose tolerance test (OGTT) and compared to an OGTTplus infusion of GLP antisera via the jugular vein. They found that the insulin releaserelative to the glycemic stimulus (the insulinogenic index) was significantly depressed byanti-Gil’ serum and was comparable to that observed in rats which received intravenousglucose. The GIP antiserum completely suppressed the incretin effect of the oral glucose,suggesting GIP to be the only incretin. However, the GIP antiserum used in these studieswas prepared using a 40% pure GIP preparation which was derived from porcine CCKextracts. It is therefore probable that the antisera preparation also bound other potentialincretin factors contained in the impure gut extract.11Ebert and coworkers (1979a) tested the effect of intravenous GIP antisera raisedto a much purer form of Gil? on intestinal stimuli for IRGIP release in rats. Duringintraduodenal acid administration along with intravenous glucose infusion, GIPantibodies strongly inhibited the initial increase in insulin levels, although after 20-30mm, circulating insulin levels were the same as in control rats. Similarly, in 1982, Ebertand Creutzfeldt found the incretin effect of GIP was strongest immediately after anintraduodenal glucose load, and the GIP antisera did not completely block the incretineffect. It was therefore concluded from these studies that GIP was not the exclusiveincretin and that additional gut factors with insulinotropic activity existed. Since thenervous system also influences glucose-induced insulin secretion following an oral orintraduodenal glucose load (Porte et al., 1973), the preserved incretin effect after GIPantiserum injection could be related to the neural part of the enteroinsular axis. Tocircumvent potential neural effects, Ebert et a!. (1983) examined the insulinotropicpotency of differently prepared gut extracts. They found that intravenous rat gut extractsexerted insulinotropic activity even after removal of GIP by immunoadsorption.Approximately 50% of the incretin activity of the gut extracts remained suggesting theexistence of other hormonal gut factors with insulinotropic activity.1.5 INSULINOTROPIC ACTIONS OF GIP15.1 EFFECT OF GIP ON INSULIN SECRETIONThe observation by Dupré et a!. (1973) that GIP was not insulinotropic undereuglycemic conditions suggested glucose played an important role in regulating thisaction of GIP. It was also established that the insulinotropic action of GIP was glucose-concentration dependent in vivo in dog (Pederson et a!., 1975b), man (Elahi et a!., 1979)and in the perfused rat pancreas (Pederson and Brown, 1976) where a glucose thresholdof approximately 5.5 mM for the insulinotropic action of GIP was observed. Thisprompted Brown and Pederson (1976a) to suggest that GIP be given the alternate12designation Glucose-dependent Insulinotropic Polypeptide. Thus while ingestion of fathas been shown to be a potent stimulus for IRGIP release (see section 1.10.3), no increasein insulin was observed unless intravenous glucose was administered as well (Brown,1974; Cleator and Gourlay, 1975; Ross and Dupré, 1978). There have even been reportsthat GIP may attenuate the stimulation of insulin secretion under euglycemic conditions(Opara and Go, 1993). These observations demonstrate that the glucose-dependency ofGIP-stimulated insulin secretion provides an important safeguard against hypoglycemiaby preventing the inappropriate stimulation of insulin release during a high fat, lowcarbohydrate meal.Not only was it shown that the insulinotropic action of GIP depended on thepresence of a threshold glucose concentration, but it also appeared that glucose couldpotentiate the action of GIP on the B-cell. At a fixed GIP concentration, increasedglucose concentrations stimulated insulin secretion in more than an additive manner(Pederson and Brown, 1976). In the perfused rat pancreas, the maximum potentiatingaction of glucose on GIP-stimulated insulin release was observed at approximately 16mM (Brown, 1982). In the presence of 17.8 mM glucose, when GIP was delivered to therat pancreas in a linear gradient from 0 to 200 pM, the insulinotropic effect of GIP wasinitiated at concentrations as low as 70 pM (Pederson et a!., 1982), concentrations withinthe physiological range (see section 1.9.2). It was also noted that the effect of GIP oninsulin release from the perfused rat pancreas was dose-dependent (Pederson and Brown,1976).15.2 OTHER POTENTIAL INCRETINSSeveral other gastrointestinal hormones have been investigated as potentialincretins involved in the enteroinsular axis, including gastrin, secretin, CCK, and tGLP-I(Creutzfeldt and Ebert, 1988; Brown, 1988). Kreymann et a!. (1987) showed that tGLP-Iwas present in the small intestine of man, and observed a sustained increase in plasma13levels following an oral glucose load or ingestion of a mixed meal. At postprandialconcentrations, Mojsov et al. (1987) showed potent insulinotropic actions of tGLP-I inthe perfused rat pancreas. In humans, tGLP-I infusion at physiological concentrations inthe presence of an intravenous glucose load significantly enhanced insulin release andsignificantly reduced peak plasma glucose concentrations. Furthermore, like GIP, theeffectiveness of tGLP-I has been shown to be glucose-dependent (Weir et al., 1989). Ittherefore appears that tGLP-I also fulfills the criteria of an incretin.15.3 INTERACTIONS BETWEEN INSULIN SECRETAGOGUESRasmussen and his colleagues have suggested that interactions between variousinsulin secretagogues occur at the level of signal transduction with the B-cell (Zawalichand Rasmussen, 1990; Rasmussen et al., 1990). Thus, agonists that act via distinctintracellular pathways tend to potentiate the action of each other on the B-cell, while thosethat act via the same intracellular pathway tend to have additive stimulatory effects.Evidence for this hypothesis is the observation of potentiating interactions between GIPand acetyicholine (Verchere et a!., 1991) as well as CCK (Sandberg et a!., 1988) in theperfused rat pancreas. GIP appears to act by stimulating cAMP production in the B-cell(see section 1.5.6), while acetyicholine and CCK act via the phosphoinositide pathwayand mobilization of intracellular calcium (Prentki and Matschinsky, 1987). On the otherhand, activators of cAMP, such as tGLP-I tend to exert additive effects with GIP oninsulin release (Fehmann et al., 1989).In addition to interacting at the B-cell with other hormones and neurotransmitters,GIP stimulated insulin release is subject to modulation by nutrients such as arginine. In197 8, Pederson and Brown demonstrated that in the presence of a glucose concentrationbelow the threshold for GIP-stimulated insulin release (2.7 mM), arginine (5 or 10 mM)was able to potentiate the insulinotropic action of OW. At higher concentrations (20mM) arginine attenuated the insulinotropic effect of GIP, suggesting that GIP and14arginine acted on the B-cell via a similar mechanism (Pederson and Brown, 1978).Support for this hypothesis has come from the observation of similar interactions ofarginine and GIP on insulin release in vivo in man (Elahi et al., 1982).15.4 RELATIVE CONTRIBUTION OF GIP vs tGLP-IThere have been numerous studies aimed at determining the relative importanceof GIP and tGLP-I in the enteroinsular axis. While some studies indicated both peptideswere equally effective (Schmid et al., 1990; Kieffer et a!., 1993), other dose-responseanalyses of the insulinotropic activity of GIP and tGLP-I indicated that lower plasmaconcentrations of tGLP-I are necessary to augment insulin secretion, particularly atelevated glucose concentrations (D’Alessio et al., 1989; Holst et aL, 1987; Mojsov et a!.,1987; Shima et a!., 1988; Weir et a!., 1989; Nauck et a!., 1989; Krarup et al., 1987c;Kreyman et a!., 1987). On the other hand, while tGLP-I appears more potent than GIP,the rise in IRGIP following oral glucose is greater in magnitude than the increment inplasma immunoreactive tGLP-I after a similar load (Nauck et a!., 1986; Salera et al.,1983; Kreyman et aL, 1987; ørskov and Hoist, 1987; Ørskov et al., 1991; Nauck et a!.,1993). Nauck et al. (1993) examined the incretin effect of Gil’ and/or tGLP-I infused atdoses chosen to produce plasma levels roughly comparable to values measured after oralglucose in human subjects (--450 pM and —50-60 pM respectively). It was observed thatthe doses used were sufficient to augment the B-cell response to intravenous glucose, tovalues not significantly different from those after oral glucose. It was also noted that GIPmade a major contribution to the incretin effect after oral glucose, while tGLP-I appearedto mediate a smaller proportion. Nauck et al. (1993) concluded that GIP and tGLP-Itogether were sufficient to explain the full incretin effect after oral glucose in normalindividuals. Fehmann et a!. (1989) and Suzuki et a!. (1992) noted synergistic stimulatoryincretin effects of tGLP-I and GIP on the rat pancreas and concluded that the hormonesact “in concert” to guarantee adequate insulin responses to a meal. Kreymann et a!.15(1987), found that when human volunteers consumed a test breakfast, plasma IRGIPlevels rose more rapidly than tGLP-I, implying that GIP mediated the early insulinresponse to a meal while tGLP-I was more important in the late response.15.5 PARTICIPATION OF GJP AND tGLP-I IN AN ENTEROENDOCRINE LOOPThe observed differences in IRGIP and tGLP-I secretion rates have beeninterpreted differently by different investigators. The two insulinotropic peptides GIPand tGLP-I have distinct patterns of distribution in the intestine. GIP cells are locatedmainly in the upper small intestinal mucosa, found in highest concentrations in theduodenum and, to a much lesser extent, in the ileum and colon (see section 1.8). tGLP-Iproducing L-cells, on the other hand, are located predominantly in the ileum and colon,with 10- to 40-fold lower concentrations of proglucagon-derived peptides in theduodenum (Eissele et al., 1992). As nutrients rarely reach the ileum before postprandialinsulin responses are observed, the role of tGLP-I as a physiological incretin acting fromthis location has been questioned. In 1991, Roberge and Brubaker made the interestingobservation that intestinal proglucagon-derived peptide secretion was stimulated equallyby fat in the duodenum as compared to fat in the ileum. It was concluded that duodenalfat either stimulates the enteric nervous system or the secretion of a factor that, in turn,stimulates the release of intestinal proglucagon-derived peptides, including tGLP-I.Brubaker (1991) also reported on the regulation of intestinal proglucagon-derived peptidesecretion by intestinal regulatory peptides using a fetal rat intestinal culture model. It wasobserved that GIP stimulated intestinal proglucagon-derived peptide secretion atphysiological concentrations and in a dose-dependent fashion. This prompted the groupto hypothesize an ‘enteroendocrine loop’, whereby GIP might stimulate the release tGLP-Ibefore arrival of nutrients in the ileum, allowing tGLP-I to act as an incretin.Recently, Roberge and Brubaker (1993) investigated this hypothesis. They found,as did Kreymann et al. (1987), that the rise in plasma IRGIP levels in response to16duodenal nutrients occurred slightly before intestinal proglucagon-derived peptide levels,suggesting a relationship between the two peptides. Intravenous infusion of GIP at a dosesufficient to yield concentrations similar to those observed after duodenal fatadministration induced a 2-fold increase in plasma levels of intestinal proglucagonderived peptides that were independent of glycemic levels. No increment in intestinalproglucagon-derived peptides was found in response to infusion of CCK. It wastherefore concluded that this enteroendocrine loop between the duodenal peptide GIP andthe ileal proglucagon-derived peptides may account for some of the early rises insecretion of tGLP-I observed in response to nutrient ingestion. Nauck et al. (1993)however questioned this hypothesis when intravenous GIP infusions failed to reproducean increase in tGLP-I levels mimicking levels attained after oral glucose in humansubjects.1.5.6 MECHANISM OF ACTION OF GIP ON THEJ3-CELLThe mechanism by which GD? stimulates insulin release from the B-cell is notfully understood. This has largely been due to the lack of homogenous preparations of B-cells, which have only recently become available. Using a hamster pancreatic B-cell line(In III), Amiranoff et al. (1984) showed that GIP produced a concentration-dependentincrease in cAMP content in the cells which paralleled insulin release. The idea that GIPacted via stimulation of B-cell G-protein coupled adenylate cyclase was latersubstantiated by studies with cultured rat islets (Siegle and Creutzfeldt, 1985), the ratinsulinoma cell line RINm5F (Gallwitz et a!., 1993), and human insulinoma tissue invitro (Maletti et a!., 1987). The closely related peptide tGLP-I was also shown toincrease cAMP levels in a rat islet cell line (Drucker et al., 1987; Göke and Conlon,1988; Gallwitz et a!., 1993). An alternative pathway for the augmentation of insulinsecretion by GD? was suggested by Lardinois et a!. (1990). Using neonatal rat islet cellcultures, they observed that GIP-stimulated insulin release could be suppressed by17inhibitors of the membrane associated enzyme phospholipase A2 and intracellularlipoxygenase and cyclooxygenase. It was concluded that GIP exerted its influence in partby modulating membrane associated phospholipase A2 activity, and that the formation ofintracellular lipoxygenase products appeared to be a pivotal step in the insulinotropicaction of GIP.More recently, Wahi et a!. (1992) investigated the potential for a role of ionicfluxes in pancreatic beta cells as a target for the action of GIP. Using mouse pancreaticislets, it was observed that the amplification of insulin release by GIP did not occur in theabsence of Ca2 in the extracellular medium, and did not involve changes in the cellularcontent of inositol triphosphate. It was therefore concluded that the amplifying effect ofGIP on insulin release was due to an effect on Ca2 uptake, and subsequent increase inelectrical activity. It was also pointed out that an additional contribution of the other isletcell types (c and 6) should also be taken into account, although the mouse islet containsmore than 80% B-cells. A potential role for glucagon is suggested by the fact that GIPhas been shown to increase glucagon release (Szecowka et al., 1982b), and theobservation by Pipeleers et al. (1985) that glucose-induced insulin release from isolatedB-cells depended on the cAMP levels, which were markedly increased after addition of(Bu)2cAMP, glucagon, or pancreatic tx-cells.Using a hamster B-cell line (HIT T15) Lu et al. (1993) found that both tGLP-Iand GIP-stimulated increases in cAMP resulted in increased extracellular Ca2 influxthrough voltage-dependent Ca2 channels. Neither peptide altered phosphoinositidemetabolism, further underlining the fact that the mobilization of intracellular Ca2 fromendoplasmic reticulum is not involved in the GIP and tGLP-I signal transductionpathways. Yada et a!. (1993) also observed that tGLP-I potentiated rises in cAMP andresultant increases in the cytosolic free Ca2 concentration in size selected rat pancreaticB-cells. It was determined that the Ca2 influx resulted from an enhanced activity of Ltype Ca2 channels in the B-cell plasma membrane.1815.7 NATURE OF GLUCOSE-DEPENDENCYThe glucose-dependency of the insulinotropic action of GIP has been well-documented in rat, dog, man (Dupré et a!., 1973; Pederson et a!., 1975b; Elahi et al.,1979; Pederson and Brown, 1976; Zawalich et a!., 1993, Siegel et al., 1992), and in 13-celllines (Kieffer et a!., 1993; Lu et a!., 1993). Other insulin secretagogues such as CCK(Verspohi and Ammon, 1987; Zawalich et a!., 1987), acetylcholine (Garcia et a!., 1988;Verchere et a!., 1991) and tGLP-I (Weir et a!., 1989; Komatsu et a!., 1989) have alsoexhibited glucose-dependency. The mechanism by which GIP and other insulinotropicagents are glucose-dependent is not fully understood. D-glyceraldehyde, an intermediateof the glycolytic pathway, was able to potentiate the insulinotropic action of GIP in theabsence of glucose from the perfused rat pancreas (Brown, eta!. 1981). Mannoheptulose,which blocks glycolysis, abolished GIP-stimulated insulin secretion in the samepreparation (MUller et a!., 1982). These studies strongly suggest that glucose metabolismis a prerequisite for GIP-stimulated insulin release to occur.Rasmussen et at. (1990) proposed that GIP-induced increases in cAMP contentleading to the conversion of latent to operative Ca2channels are ineffective in producingsufficient Ca2 influx for insulin release unless the B-cell is partially depolarized by theactions of glucose. Postprandial changes in plasma glucose concentrations are thussuggested to primarily serve as conditional modifiers of insulin secretion. Calmodulin,activated by glucose-induced rises in intracellular Ca2, has been shown to stimulateadenylate cyclase activity in rat islets (Sharp et at., 1980) and was proposed by Holz andHabener (1992) to be required for tGLP-I stimulated insulin release. Although this mayalso explain the glucose-dependency of GIP,Ca2-calmodulin does not always appear toactivate adenylate cyclase activity in islets (Thams et a!., 1982). Furthermore, transgenicmice with elevated levels of B-cell calmodulin develop severe diabetes, even though theirpancreatic B-cells contain reserve levels of insulin (Epstein et a!., 1992).19Another possible mechanism by which glucose may sensitize the B-cell to furtherstimulation by insulinotropic agents is through the expression of functional hormonereceptors. The transport of glucose, and/or its metabolism resulting in the production ofvarious intracellular messengers could be involved in the expression or allostericactivation of receptors for insulinotropic hormones such as GIP and tGLP-L Thus in theabsence of glucose, a reduction in the expression of functional hormone receptors wouldbe expected and it would therefore be predicted that GIP or tGLP-I would not alter cAMPlevels; an observation that was recently made using a hamster B-cell line (Lu et al., 1993).If in fact glucose transport is necessary for ‘correct GIP receptor expression, then in caseswhere glucose transport is altered, a corresponding alteration in the insulinotropic actionof GIP would be expected. Reduced expression of the B-cell high Km glucosetransporters (GLUT2) has been demonstrated in the non-insulin-dependent diabetesmellitus model, the obese Zucker rat (Johnson et al., 1990; Orci et al., 1990). Thisobservation has also been observed in transformed B-cells, combined with the presence ofa glucose transporter isoform with a lower Km for glucose (GLUT1; Brant et al., 1992)and alterations in the expression of various hexokinases (Visher et a!., 1987). In both theobese Zucker rat and the transformed B-cell line, BTC3, a significant reduction in theglucose threshold for GIP-stimulated insulin release has been observed (Chan et al.,1984; Kieffer et a!., 1993). Examination of the hypothesis that altered GIP receptorexpression may contribute to these observations may soon be possible with the recentcloning of the GIP receptor (Usdin et al., 1993).1.6 OTHER ACTIONS OF GIPIn addition to enterogastrone and insulinotropic actions, GIP has been suggestedto have other functions, although whether they are physiological or not is uncertain. Thislist includes actions on gastrointestinal motility, mesenteric blood flow, intestinalsecretion, and release of anterior pituitary hormones (Brown et a!., 1989). The recent20discovery of the GIP receptor in tissues outside the gut (brain; see section 1.11.2)suggests that there may be as yet undiscovered actions of this hormone. Potentialanabolic actions of GJP are better documented. In the liver, GIP has been demonstratedto inhibit glucagon-induced lipolysis and diminish glucagon-stimulated hepatic glucoseproduction (Hartmann et at., 1986). GIP may also increase glucose uptake in adipocytesand other peripheral tissues such as muscle (Hauner et a!., 1988). These insulin-likeactions are compatible with the insulinotropic effect of GIP in promoting glucoseutilization. In adipose tissue, GIP has been shown to increase lipoprotein lipase activity,(Eckel et at., 1979), inhibit glucagon-induced lipolysis (Dupré et at., 1976), andpotentiate the insulin-stimulated incorporation of fatty acids into triglycerides (Beck andMax, 1983). GIP has also recently been shown to stimulate fatty acid synthesis inadipose tissue (Oben et al., 1991). A role for GIP in the control of fat metabolism is notsurprising considering the potent stimulatory effect of fat on IRGIP release, which in theabsence of glucose, is not insulinotropic (Beck, 1989).1.7 GIP GENE1.7.1 GENE STR UCTURE AND POST-TRANSLATIONAL PROCESSINGThe sequence of the human GIP gene was reported by Takeda et at. in 1987. ThecDNA sequence coding the human GIP precursor (Figure 2) indicates that humanpreproGlP is a protein of 153 amino acids, with a predicted molecular weight of 17,107daltons. The sequence indicates that proteolytic processing at single arginine residues ateither end of the GIP sequence would yield GIP 1-42 (Figure 2). The structuralorganization of human preproGlP was reported by this group to be: putative signalpeptide (21 amino acids), NH2-terminal peptide (30 amino acids), GIP (42 amino acids)and COOH-terminal peptide, (60 amino acids). Subsequently, Inagaki et al. (1989)revealed that the human GIP gene consists of six exons separated by five introns, withexons 3 and 4 encoding mature human GIP. Recently, rat preproGiP was reported to be21144 amino acids in length and comprised of the GIP peptide itself, N- and C-terminalflanking peptides of 22 and 59 amino acids respectively and a typical hydrophobic signalpeptide (Sharma et a!., 1992).—90 —80 —70 —60 —505’ - AGGCTCAGAAGGTCCAGAAATCAGGGGAAGGAGACCCCTATCTGTCCTTCTTCTGGAA—40 —30 —20 —10 1 10 20GAGCTGGAAAGGAAGTCTGCTCAGGAAATAACCTTGGAAGATGGTGGCCACGAAGACCTTMetVa lAlaThrLysThrPh30 40 50 60 70 80TGCTCTGCTGCTGCTGTCCCTGTTCCTGGCAGTGGGACTAGGAGAGAAGAAAGAGGGTCAeA1aLeuLeuLeuLeuSerLeuPhLuA1aVa1G1yLeuG1yG1uLysLysG1uG1yHi90 100 110 120 130 140CTTCAGCGCTCTCCCCTCCCTGCCTGTTGGATCTCATGCTAAGGTGAGCAGCCCTCAACCsPheSerAlaLeuProSerLeuProValGlySerHisAlaLysValSerSerProGlnPr150 160 170 180 190 200TCGAGGCCCCAGGACGCGGAAGGGACTTTCATCAGTGACTACAGTATTGCCATGGACAAoArgGlyP roArgTyrAlaGluGlyThrPhelleSerAspTyrSerfleAlaNetAspLy210 220 230 240 250 260GATTCACCAACAAGACTTTGTGAACTGGCTGCTGGCCCAAAAGGGGAAGAAGATGACTG$ I1eHisG1nG1nAspPheVa1AsnTrpLeuIeuA1aG1nLysG1yLysLysAsnAspTr270 280 290 300 310 320GAAACACAACATCACCCAGAGGGAGGCTCGGGCGCTGGAGCTGGCCAGTCAAGCTAATAGPlaysHisAsnhleThrGlnArgGluAlaArgAlaLeuGluLeuAlaSe rGlnAlaAsnAr330 340 350 360 370 380GAAGGAGGAGGAGGCAGTGGAGCCACAGAGCTCCCCAGCCAAGAACCCCAGCGATGAAGAgLysGiuGluGluAlaValGiuP roGlriSerSerP roAlaLysAsnProSerAspGluAs390 400 410 420 430 440TTTGCTGCGGGACTTGCTGATTCAAGAGCTGTTGGCCTGCTTGCTGGATCAGACAAACCTpLeuLeuArgAspLeuLeul leGlnGluLeuLeuAlaCysLeuLeuAspGlnThrAsnLe450 460 470 480 490 500CTGCAGGCTCAGGTCTCGGTGACTCTGACCACACCCAGCTCAGGACTCGATTCTGCCCTTuCysArgLeuArgSerArg510 520 530 540 550 560CACTTAGCACCTGCCTCAGCCCCACTCCAGAATAGCCAAGAGCCCCCTGT570 580 590 600 610TTATGCTAAGTCGAGCCCATTGTGAAATTTATTAAJATGACTACTGAGCACT- 3’FIGURE 2: Sequence of a cDNA encoding the human GIP precursor. The numbering ofnucleotides starts from the first adenine of the initiating methionine to the last nucleotidejust before the poly(A) tract. Residues in the 5’ untranslated region have negativenumbers. The deduced amino acid sequence of preproGlP is indicated. The regions ofthe putative signal sequence and mature GIP are underlined and in bold, respectively.(From Takeda et a!., 1987)221.7.2 GENE REGULATIONThe mechanism of GIP gene regulation is not completely understood. Thepromoter region of the human GIP gene contains potential binding sites for multipletranscription factors, including Spi, AP-1 and AP-2, but the roles of these sites areunknown (Inagaki et at., 1989). More recently, this same group investigated whether GIPgene expression is regulated by cAMP and glucose. Two cAMP response element (CRE)binding protein (CRE-BP) binding sites were identified in the promoter region (Someyaet at., 1993). CRE is an inducible enhancer of genes which can be transcribed inresponse to increased cAMP levels (Comb et at., 1986). CRE-BP1 can dimerize with thec-jun protooncogene product (c-Jun) (Macgregor et at., 1990), and interestingly, c-Junrepresses the cAMP-induced activity of the insulin promotor (Inagaki et at., 1992). Apossible mechanism of this c-Jun action is by its ability to inhibit the binding of someCRE-BP’s, such as CRE-BP1, to the CREs of the insulin gene by formation of aheteroclimer. These results, and another observation that the level of c-Jun is dramaticallyincreased by glucose deprivation in hamster insulinoma (HIT T15) cells (Inagaki et al.,1992), suggest that glucose may regulate expression of the human insulin gene throughCREs and c-Jun. Mutation analysis showed that the two CREs are required for basalpromoter activity. Interestingly, the GIP promoter activity was repressed by c-Jun likethe human insulin promoter, possibly through the CREs.1.8 Gil’ CELL DISTRIBUTION AND CHARACTERIZATIONProof of the existence of endocrine cells in the gastrointestinal mucosa came withthe use of electron microscopy by Solcia et a!. (1967). Several morphologically distinctcell types were soon identified and correlated with the production of members of theexpanding family of gut endocrine peptides. Immunoreactive GIP (IRGIP) cells havebeen located in the upper small intestine of ruminants (Bunnett and Harrison, 1986), man,pig, dog (Buffa et al., 1975) and rats (Buchan et at., 1982). In the gastrointestinal tract of23dog and man, IRGIP is present in cells predominantly in the mid-zone of the glands in theduodenum, and to a lesser extent in the jejunum (Polak et al., 1973). Other studies havefound a few IRGIP cells as far as the terminal ileum in rat (Buchan et al., 1982) and man(Fern et aL, 1983). IRGIP cells are part of the amine precursor uptake decarboxylation(APUD) series (Pearse, 1968). Ultrastructural studies of human IRGIP cells indicated acharacteristic appearance of the K-cell; intracellular secretory granules having a smallelectron-dense core surrounded by a concentric electron-lucent halo (Buchan, 1978). Inthe dog, however, IRGIP cells identified in the duodenum contained uniformly electron-dense secretory granules consistent with the cell type recognized as the I-cell of theendocrine cell classification (Usellini et al., 1984). Thus some species specificity existsregarding the ultrastructure of the GIP cell.Some immunocytochemical studies indicated that GIP was co-localized withglucagon in the pancreatic a-cells in mammals (Smith et at., 1977; Alumets et at., 1978;Ahrén et a!., 1981). Other studies indicated, however, that the staining of the pancreatica-cells could be blocked by preincubation of the antisera with glicentin (Larson andMoody, 1980) or glucagon (Buchan et al., 1978), suggesting that these results were dueto cross-reactivity of the GIP antisera used with glucagon or the glucagon-like peptidesand their precursors. Furthermore, studies using RNA blot analysis have only detectedthe presence of human preproGlP mRNA in human intestine, and not in the pancreas(Takeda et at., 1987; Inagaki et at., 1989; Sharma et a!., 1992). It is interesting to notethat IRGIP has been localized to a distinct cell type in the islet organ of elasmobranchs,although its function there is not known (Hazeiwood, 1989). It appears that GIP cells arenot found phylogenetically in the gut mucosa until a distinct islet organ evolves uponwhich the peptide may act, i.e. the hagfish, suggesting the possibility of a functionalenteropancreatic axis at this evolutionary level (Falkmer et a!., 1980).241.9 MEASUREMENT OF GIP1.9.1 DEVELOPMENT OF GIP RADJOIMMUNOASSAYWith the purification of GIP came the possibility of development of techniquesfor measuring endogenous hormone secretion. This was a prerequisite for theestablishment of GIP as a hormone, in order to demonstrate the polypeptide in blood andtissues and release from tissues by physiological mechanisms. The development of aradioimmunoassay (RIA) for GIP was first achieved in 1974 by Kuzio et al. Antiserawere produced in guinea pigs by subcutaneous immunization with GIP emulsified withFreund’s adjuvant, followed by boosting with GIP conjugated to bovine serum albuminusing the carbodiimide method. Antisera were shown to have no measurable crossreactivity with natural secretin, synthetic glucagon, synthetic human gastrin, pure porcineCCK, pure porcine motilin, and pure porcine vasoactive intestinal peptide. GIP waslabeled with 125J by a minor modification of the chloramine-T technique, and purified ona Sephadex G-15 column. The RIA developed had a sensitivity range of 25 to 250 pg,and Kuzio et al. (1974) reported a mean fasting serum concentration of 237 ± 14 pg/miand >1200 pg/mi after a meal.1.9.2 PHYSIOLOGICAL LEVELS OF GIPSubsequent to the development of the first GIP RIA, other assays followed(Moody and Lauritsen, 1977; Morgan et al., 1978; McLoughlin et al., 1979; Ebert et al.,1979b; Sarson et al., 1980; Burhol et al., 1980a; Wolfe and McGuigan, 1982; Jorde et al.,1983; Sheu et al., 1987; Wishart et al., 1992; Moody et a!., 1992). There appeared to bea general agreement on the pattern of IRGIP release after ingestion of nutrients, whereasthe absolute values measured differed widely, ranging from 9 to 80 pmol/l in the fastingstate, and from 35 to 700 pmol/l after meals. In 1983, Jorde et al. attempted to resolvethis disparity by measuring fasting and postprandial IRGIP values in man with sevendifferent antisera, including two of their own. Even under the same assay conditions,25mean fasting IRGIP levels ranged from 12 to 92 pmol/l, and the mean postprandial IRGIPvalues ranged from 35 to 235 pmol/l. It was concluded that this variability was a result ofdifferent cross reactivities of the antisera with human GIP. Porcine GIP (two amino acidsdifferent) of varying purities had been used as standards and to raise antisera for RIAsused to measure IRGIP in man.Amland et a!. (1984) measured fasting and postprandial IRGIP values in pigs,rats, dogs and man with five different antisera and porcine GIP standards and tracers.The mean IRGIP values in rats, dogs, and man varied considerably, depending on theantiserum used, whereas all the antisera recorded fairly similar IRGIP values in pigs.These findings again demonstrated the species differences in immunological properties ofGIP, and thus the importance of using GIP from the appropriate species as standards.The sequencing of human GIP and its successful production by both chemical synthesis(Yajima et a!., 1985; Fujii et al., 1986) and recombinant DNA techniques (Chow et a!.,1990) should have allowed the proper evaluation of IRGJP concentrations in man. Oddlyenough, only two groups at present appear to have used human GIP in assays employedfor determining plasma IRGIP levels in man. Kreymann et al. (1987) used synthetichuman GIP as standards for the measurement of IRGIP in human subjects in response tooral glucose and a test meal. Basal values were 47± 16 (mean ± SEM) and peaked at 174± 24, 30 mm after receiving oral glucose. The material used for tracer production was notstated. Using synthetic human GIP for tracer and standards, Nauck et a!. (1992)measured basal IRGIP levels in human subjects as 69 ± 18 pM, and peak IRGIPconcentrations after oral glucose of 340 ± 39 pM. It is not surprising that these peaklevels are higher than others previously reported in response to oral glucose.1.9.3 IMMUNOREACTIVE FORMS OF GIPIn 1975, Brown et a!. reported that the antisera used in the RIA developed byKuzio et a!. (1974) identified more than one molecular form of IRGIP in plasma and gut26tissue. In fact, all seven of the antisera tested by Jorde et al. (1983) recognized threemolecular forms of IRGIP to variable degrees, likely contributing to the markeddiscrepancies in reported IRGIP levels. When subjected to gel filtration, one highmolecular weight component eluted in the void volume of the column, while the twoother components eluted at positions corresponding to molecular weights ofapproximately 8000 and 5000 daltons (Brown et a!., 1975). The 5000 IRGIP form is themajor component of GIP in both blood and tissues and has been used for theestablishment of RIAs (see Figure 1 for sequence). After gel filtration of extracts ofporcine and human small intestine, a fourth IRGTP form was found eluting between 8000IRGIP and void volume (Krarup and Holst, 1984). This fourth form has never beendetected in plasma. Krarup and Hoist (1984) concluded that since all five antisera testedreacted with this IRGIP form it may represent a prohormone to 5000 IRGIP.No void volume IRGIP (V0-IRGIP) has been detectable after gel filtration ofextracts of gut mucosa (Bacarese-Hamilton et al., 1984; Krarup and Hoist, 1984). Thiscomponent could be removed from plasma by alcohol extraction, indicating itsassociation with high molecular weight constituents of plasma (Krarup, 1988). Thissuggestion had been made by Dryburgh (1977), when it was noted that pre-treatment ofserum by boiling or by addition of 6 M urea diminishedV0-IRGJP and increased 5000IRGIP. In addition,V0-IRGIP does not appear to contribute to the increase in IRGIP inresponse to glucose, fat, or a meal (Krarup et a!., 1985; Krarup et a!., 1987b). Krarup(1988) concluded that V0-IRGIP may represent large molecular weight proteinsinterfering in the assays.While 5000 IRGIP was shown to increase in plasma 60 mm after intraduodenalglucose or fat had been given in man, 8000 IRGIP showed small and inconsistent changesafter both stimuli (Krarup et al., 1985; Krarup et a!., 1987b). It was concluded by Krarup(1988) that 8000 IRGIP contributed little to the increase in IRGIP after glucose and fat,and only modestly to that observed after a meal. A partially purified 8000 IRGIP from27porcine gut did not stimulate insulin release even when used at pharmacological doses inan isolated porcine pancreas system (Krarup et al., 1987a). Otte et al. (1984) found thatthe entire sequence of GIP was not contained in an 8000 IRGIP form extracted from piggut and concluded that it was not a proform of GIP. Since 8000 IRGIP cross reacts withmost GIP antisera (Jorde et a!., 1983; Krarup, 1988) there must be a degree of homologybetween this form and 5000 IRGIP. It is possible, therefore, that 8000 IRGIP is derivedfrom a GIP precursor.1.10 GIP SECRETION1.10.1 RESPONSE TO A MIXED MEALIt is generally agreed that IRGIP levels increase approximately six times abovebasal in response to ingestion of a mixed meal, although absolute values varysignificantly (Kuzio et al., 1974; Morgan et a!., 1978; Jorde et a!., 1980; Jorde et al.,1983; Amland et a!., 1984). Plasma levels of IRGIP remained elevated for up to sixhours after a meal, and thus during a 24 h period, with three meals ingested, IRGIP wassignificantly high all day (Jorde et a!., 1980; Salera et al., 1983; Jones et al., 1985). Becket a!. (1984) observed that ingestion of a meal with a higher caloric level yielded asignificantly greater IRGIP response, even though blood glucose levels were notsignificantly different. It was concluded that this was an adaptation to the greater amountof food ingested, and that the GIP producing cells have a rapid mechanism of adapting tothe ingested caloric load. Service et a!. (1983) also observed a highly significantassociation between meal size and post-prandial plasma IRGIP responses.Oektedalen et al. (1983), compared the IRGIP response to a meal in previouslyfasted and fed individuals. It was observed that the IRGIP response to oral glucose of atest meal was augmented after food deprivation. It was suggested that the increasedcatecholamine levels associated with fasting may have caused the increased IRGIPrelease, as the adrenergic nervous system has been shown to modify the IRGIP release28(see section 1.10.5). Hampton et a!. (1983) found the IRGIP response to a meal in ratsfed a high fat diet for 4 days prior was greater than controls. Deschodt-Lanckman et al.(1971) showed that pancreatic lipase activity in the rat could be stimulated by a high fatdiet. This increased lipase activity might result in an increase in the rate of absorption offat, and thus a greater IRGIP response (section 1.10.3). Morgan et a!. (1988), however,only noted a small but significant increase in IRGIP release by human subjects previouslyon a high fat diet for 35 days, suggesting that species differences are likely to exist.1.10.2 RESPONSE TO CARBOHYDRATESEvidence supporting the role of GIP as an incretin came with the first observationof its release in response to oral glucose, followed almost immediately by an increase inplasma insulin levels (Cataland et al., 1974). A direct relationship between the glucoseload ingested and the release of IRGIP was observed in dog (Pederson et al., 1975b) andman (Falko et al., 1980). The fact that luminal stimulation by glucose was necessary forIRGIP release was proven when no changes in circulating levels of this hormone wereobserved when glucose was administered intravenously (Cataland et a!., 1974; Pedersonet a!., 1975b; Cleator and Gourlay, 1975; Andersen et a!., 1978).The digestion of complex carbohydrates occurs in the gut through the action ofpancreatic enzymes and brush border hydrolases such as sucrase and maltase. Finalproducts of digestion, D-glucose, D-galactose, and D-fructose, can then be absorbed byenterocytes lining the upper third of the intestinal villi. The absorptive process occurs inthe duodenum and jejunum and is generally complete before the chyme reaches theileum. The pyranoses are absorbed by a two-stage process. Glucose and galactose enterepithelial cells by an energy requiring Na-dependent transporter, SGLT1 (Wright, 1993).In contrast, fructose is absorbed down its concentration gradient by the facilitativeglucose transporter, GLUT5 (Burant et a!., 1992). The second step is the downhilltransport of sugar out of the enterocyte across the basolateral membrane into the blood,29by the facilitated sugar transporter GLUT2 (Thorens, 1993). During some point in thisprocess, endocrine cells release GIP; the actual step that triggers this secretion is howevernot understood.In an attempt to determine the carbohydrate specificity for GIP release, Morgan(1979) studied the IRGIP responses to the sugars glucose, galactose, sucrose, and fructosegiven as an oral carbohydrate tolerance test in human subjects, or perfused through the ratsmall intestine. In both models, IRGIP was released by glucose, galactose and sucrose,but not fructose. It was also noted that the response to sucrose was significantly delayedrelative to the rise in response to glucose and galactose. The results indicated that IRGIPsecretion was dependent on the active transport of monosaccharides. The delay in IRGIPresponse to sucrose was explained by the fact that sucrose must be hydrolysed intofructose and glucose by a brush border hydrolase prior to absorption. Similar resultswere reported by Sirinek et al. (1979; 1983) from studies of intraduodenal infusion ofcarbohydrates in dogs. Sirinek et a!. (1983) suggested, on the basis of their data, that thestructural integrity of the glucose molecule from the C-i to C-4 carbon atom, a freealdehyde group on the C-i carbon atom and a cyclic structure are all necessary for boththe active transport of glucose and the release of GIP.Further evidence for the necessity of active sugar transport of monosaccharidesfor IRGIP release has been supplied by studies in which transport is blocked. Morgan(1979) observed that reduction of the post-prandial rise in blood glucose by guar (a gumproduced from the seeds of the Indian Cluster bean, Cyonopsis tetraagonoloba) resultedin a significant reduction in IRGIP release. Sykes et al. (1980) blocked the Nadependent carrier protein with phloridzin, a 8-glucoside that binds but is not transported(Hopfer, 1987) and found that glucose-stimulated IRGIP release in rats was abolished.This group also provided further information on the minimal structural requirements forthe release of IRGIP in response to carbohydrates. Of the numerous monosaccharidestested, only those with the structural requirements for active transport by the Na30dependent hexose pathway (Hopfer, 1987) were effective as stimulants of IRGIP release.Furthermore, their observations that x-methylglucoside and 3-O-methylglucosestimulated significant IRGIP release indicated that transport out of the cell, ormetabolism, was not necessary for IRGIP release, respectively. While similar resultswere obtained by Flatt et al. (1989) with mice, others have observed no IRGIP responseto 3-O-methylglucose in rats (Fushiki et a!., 1992).While the majority of data seem to suggest that the Na-dependent glucosetransporter was involved in the mechanism by which GIP cells ‘sense’ glucose, theevidence was not definitive. Unlike Morgan (1979) and Sykes et al. (1980), Flatt et al.(1989) observed significant IRGIP release by fructose. In addition, Flatt et a!. (1989)demonstrated a delayed but large IRGIP response by 2-deoxyglucose, which is nottransported by the glucose or fructose luminal transport system (Hopfer, 1987). Sykes eta!. (1980), however, found no significant response to luminal perfusion of 2-deoxyglucose in the rat. Whether these observations indicate species differences or amechanism other than the Na-dependent glucose transporter for GIP release is not clear.In hopes of better characterizing the mechanism by which GIP cells ‘sense’glucose, Fushiki et a!. (1992) studied IRGIP release into the portal vein of rats inresponse to duodenal infusion of glucose in the presence of inhibitors of some putativeglucose sensors and carriers in the intestinal lumen. Gymnemic acid, the active principleof Gymnema sylvestre leaves, is a glucuronide of triterpene which can inhibit Nadependent active glucose transport in the small intestine (Yoshioka, 1986). Fushiki et a!.(1992) found that like phloridzin, gymnemic acid markedly suppressed the increase inportal IRGIP in response to intraduodenal glucose infusion. In the presence ofcytochalasin B, a competitive inhibitor of facilitative glucose transporters, there was nosignificant change in the portal IRGIP concentration after glucose administration,suggesting that a facilitative glucose transporter was not directly involved in the IRGIPrelease by the small intestinal endocrine cells (Fushiki et a!., 1992).31Fushilci et at. (1992) investigated the possibility that IRGIP release by glucoseoccurred by a mechanism similar to that by which glucose acts to release insulin from thepancreatic B-cell. Glucokinase has been demonstrated to be a key enzyme for glucoserecognition by B-cells as well as liver cells (Matschinsky, 1990). Infusion ofmannoheptulose, an inhibitor of glucokinase (Matschinsky, 1990), did not affect theportal vein plasma IRGIP concentration (Fushiki et a!., 1992). Glybenclamide, a potentinsulin secretagogue acting via closure of the ATP-sensitive K channels (Trube et a!.,1986), also had no effect on portal vein plasma IRGIP concentration. These findingssuggested that the mechanism underlying GIP release in response to luminal glucose inthe small intestine was different from that underlying insulin release in response toplasma glucose in the pancreas.A final putative mechanism for glucose ‘sensing’ by GIP cells was mediation by aneural pathway (Fushiki et at., 1992). In 1978, Mei found a glucoreceptor existed in theduodenum and proximal jejunum, which transmitted via vagal neurons. During recordingfrom nodose ganglia by means of extracellular glass microelectrodes, Mei (1978) notedthat glucose and galactose preferentially stimulated the sensor, but fructose did not.Furthermore, this duodenal glucose sensor could be blocked by 1% procaine infusion.Fushiki et a!. found that infusion of 10 g/l procaine and 10 g/l lidocaine in saline did notaffect the glucose-induced portal vein plasma IRGIP concentration, suggesting that theprocaine-sensitive vagal glucoreceptor was not involved in IRGIP release.Overall, the present data suggest that glucose-mediated GIP release does not occuras a result of the direct actions of the duodenal procaine-sensitive vagal glucose sensor, acytochalasin B sensitive facilitative glucose transporter, a mannoheptulose sensitiveglucokinase, or the ATP-sensitive K channels. It appears that neither glucosemetabolism, nor glucose transport out of the cell are necessary for GIP release, althoughthis requires further verification. It seems glucose transport via a phloridzin sensitiveNa-dependent transporter like SGLT1 is involved in glucose recognition leading to GIP32release but some data indicate a specificity different from that known for SGLT1.Furthermore, it is not known whether the GIP cells directly ‘sense’ glucose, or whethermessages are transmitted by neighbouring glucose-transporting mucosal cells. Sykes eta!. (1980) postulated that tight junctions between the two types of cells allowing changesin the flux of sodium ions generated in the mucosal cells by interaction between thestimulatory monosaccharides and their transport proteins might offer a possibleexplanation for their ability to stimulate release of IRGIP.1.10.3 RESPONSE TO FATSTriglycerides form the primary lipids of a normal diet. Their digestion occursprimarily in the duodenum and upper jejunum, where pancreatic lipases act on fatemulsified by bile. In the intestinal lumen, triglycerides are converted to free fatty acidsand monoglycerides, form micelles, and then diffuse among the microvilli that form thebrush border. The high lipid solubility of the fatty acids and monoglycerides promotestheir diffusion into the enterocytes. The subsequent fate of the fatty acids depends ontheir size. Fatty acids containing less than 10-12 carbon atoms pass from the mucosalcells directly into the portal blood, where they are transported as free (unesterified) fattyacids. Larger fatty acids are reesterified with monoglycerides to re-form triglycerides,which accumulate in chylomicra and are ejected by exocytosis Out of the cell, into thelymph. In order for GIP to fulfill the role of an enterogastrone, it is necessary to establishits release by some stage of this fat digestion process.The first report of IRGIP release in response to fat was made in 1974 by Brownshortly after the development of a GJP radioimmunoassay (Kuzio et al., 1974). Brownobserved a peak in serum IRGIP levels in human volunteers approximately 2 h afteringestion of 100 ml corn oil suspension (Lipomul). A similar response was noted byFailco et at. (1975), and Cleator and Gourlay (1975), and these investigators also reportedno increase in serum glucose or insulin concentrations. These studies showed quite33conclusively that endogenously released IRGIP was not insulinotropic in the absence ofhyperglycemia. Pederson et a!. (1975b) studied the IRGIP response to graded oral dosesof fat and glucose in dogs, and found triglycerides produced a greater and moreprolonged IRGIP response (5 h) than the glucose loads (2.5 h). It was suggested that therate of gastric emptying might be a contributing factor to these observations. Pederson eta!. (1975b) pointed out that the release of IRGIP by fat and the nature of this response fitwell with an enterogastrone-like action, whereas the rapid response to oral glucose wasmore relevant to the potentiation of a rapidly rising insulin response.In 1981, Ross and Shaffer found that the process of hydrolysis was necessary forGIP release. This came from the observation that there was no IRGIP release frompatients with cystic fibrosis and defective fat lypolysis following ingestion of corn oil, butthat it returned to levels similar to controls with coingestion of pancreatic enzymes. Asimilar conclusion was made by Ohneda et al. (1983) when it was observed that anIRGIP response to fat in pancreatectomized dogs occurred only during coingestion ofpancreatic enzymes. Imamura et at. (1988) found a reduced IRGIP response to ingestionof a fat-enriched meal in patients with disturbed fat metabolism resulting from externalbiliary drainage. Interestingly duodenal infusion of bile alone has been demonstrated toincrease IRGIP release (Burhol et a!., 1980b). Thus emptying of the gallbladder inresponse to the presence of fat in the duodenum may enhance the release of IRGIP inresponse to fat hydrolysis.The suggestion that fatty acids were the product of triglyceride hydrolysisresponsible for IRGIP release came from observations that monoglycerides had no effecton IRGIP levels (Ross and Shaffer, 1981; Ohneda eta!., 1984; Williams eta!., 1981). In1976, O’Dorisio et a!. reported only a modest IRGIP response to an intraduodenalinfusion of medium chain triglycerides in dogs, when compared to the IRGIP releasepreviously obtained from long chain fatty acids (Brown, 1974; Falko et a!., 1975). Otherstudies on the effects of fatty acid chain length on IRGIP release have since confirmed34that long chain fatty acids exert strong stimulation, while medium or short chain fattyacids do not (Ross and Shaffer, 1981; Ohneda et a!., 1984 Kwasowski et al., 1985). Inaddition to fatty acid chain length, it also appears that the degree of saturation mightaffect their ability to stimulate IRGIP release. Lardinois et a!. (1988) compared theIRGIP response to saturated, monosaturated and polyunsaturated fats in man, and foundthe latter yielded the smallest IRGIP secretion.As short and medium chain fatty acids are absorbed faster than long chain, GIPrelease is not related only to the rate of cellular uptake of fatty acids. Kwasowski et a!.(1985) suggested that the stimulus for fat-induced GIP release might be generated duringthe intracellular handling and metabolism of fatty acids. Short- and medium-chain fattyacids are transferred across the intestinal epithelium without esterification. However,long-chain fatty acids are conveyed to the smooth endoplasmic reticulum for esterficationprior to incorportation into chylomicrons and eventual exocytosis into the intercellularcompartment. The GIP-releasing action of fatty acids may be coupled therefore to theextent of esterification, an energy-consuming metabolic process confined to long-chainfatty acids. Previous studies on the mechanism of carbohydrate-stimulated GIP releasehave shown that only actively transported sugars stimulate IRGIP secretion (see section1.10.2). Thus there may be a common link between metabolic and secretory eventsresponsible for nutrient-regulation of GIl? release from the intestinal GIP cells.1.10.4 GIP RESPONSE TO PROTEINSThe majority of protein digestion occurs in the duodenum and jejunum, wherepancreatic proteases such as trypsin, chymotrypsin, and carboxypeptidase act to convertdietary protein to small peptides. The brush border of the upper small intestine contains anumber of peptidases, including aminopeptidases, dipeptidases and dipeptidylaminopeptidases, which further reduce the peptides to oligopeptides and amino acids.The small peptides and amino acids are then transported across the brush border plasma35membrane into the epithelial cells. Transport of dipeptides and tripeptides across thebrush border plasma membrane is a secondary active transport process powered by theelectrochemical potential difference of Na across the membrane. Normally amino acidsare transported into the enterocyte by way of certain specific amino acid transportsystems, some dependent on Na+ and others not. Once in the enterocyte, transportsystems of the basolateral membrane then transport amino acids out of the cell, againwith some dependent on Na, and others not.Brown (1974) reported that ingestion of a meat extract did not result in asignificant release of IRGIP. A similar result was observed by Cleator and Gourlay(1975) with either beef steak or a meat extract, while a small but significant response toleucine was noted. Thomas et al. (1976) were able to demonstrate that intraduodenaladministration of a mixture of amino acids elevated serum IRGJP concentrations (peakingat 30 mm) in association with an increased insulin response. More specifically, theyshowed that a mixture containing arginine, histidine, isoleucine, lysine and threoninecaused a marked rise in integrated IRGIP and insulin secretion, while a combination ofmethionine, phenylalanine, tryptophan and valine had only a minimal effect on release ofthese two hormones. It was suggested by Thomas et a!. (1978) that the higherconcentration of amino acids as compared to the previous studies was responsible for theobserved IRGIP response, as concentrations of 10 or 20 mM in the small intestinedemonstrated no effect. O’Dorisio et at. (1976) demonstrated a significant IRGIPresponse with a mixture of amino acids totaling —38 mM.In 1982, Schulz et al. examined the IRGIP response to amino acids in rats treatedwith a corticosteroid or alloxan. These compounds are thought to increase Na KATPase activity, and thus the Na-dependent active absorption process (Schulz et a!.,1982). A mixture of amino acids, including those shown by Thomas et at. (1976) tostimulate IRGIP release, were also found to be stimulatory by Schulz et a!. (1982), and,furthermore, both groups of treated rats showed significantly higher serum IRGIP levels.36These results were attributed to an increased absorption of the amino acids owing to anincreased Na K ATPase activity in the intestinal mucosa of corticosteroid- and alloxantreated rats. These data therefore suggested that as with glucose or fat, the release of GIPafter amino acid stimulation seemed to depend more on an active absorption than on theirpresence within the intestinal lumen.In 1982, Wolfe and McGuigan measured the IRGIP response following a peptonemeal in the dog. The increase in IRGIP release in response to the peptone meal in thisstudy was maximal in peripheral venous serum at 15 mm. This is similar to the responseto glucose feeding and contrasts with the more prolonged response to fat ingestion.IRGIP release has not been demonstrated previously after feeding or infusion of a proteinof peptone meal that is free of carbohydrate or fat. It was suggested that these differencesmight result from the sensitivity of the GIP radioimmunoassay used in this study whichmeasured basal and peak IRGIP levels at 34 ± 10 and 323 ± 95 pg/ml, respectively;values lower than those previously reported. No significant increase in insulin was noted,consistent with the glucose-dependent action of GIP. Wolfe and McGuigan (1982)suggested that their demonstration of GIP release after a peptone meal accompanyingpeptone-stimulated gastrin release and gastric acid secretion was supporting of theenterogastrone role of GIP.1.10.5 A UTONOMIC NERVOUS SYSTEM CONTROL OF GIP RELEASEConflicting data, both in human and animal studies, exist regarding theinvolvement of cholinergic and adrenergic factors in GIP secretion. Vagotomy has beendemonstrated either to enhance (Thomford et al., 1974; Imamura et al., 1984; Yoshiya etal., 1985), reduce (Lauritsen et al., 1982; Imamura et al., 1984), or to have no effect(Gayle and Ludewig, 1978; Yoshiya et al., 1985) on the IRGIP response to nutrientingestion. Sham feeding, which produced cephalic-vagal stimulation, had no effect onIRGIP release (Taylor and Feldman, 1982), as did direct vagal stimulation (Berthoud et37al., 1982). Studies with atropine have demonstrated either reduced (Baumert et al., 1978;Larrimer et al., 1978) or no effect (Sirinek eta!., 1981; Nelson eta!., 1986) on the IRGIPresponse to nutrient ingestion. Finally, acetylcholine decreased the IRGIP response tointragastric glucose administration (Williams and Beisbroeck, 1980).As with parasympathetic stimulation, controversy also exists on the effect of thesympathetic nervous system on GIP release. Epinephrine has been shown to inhibit(Williams and Biesbroeck, 1980) or have no effect (Sirinek et al., 1977b) on glucose-stimulated IRGIP release. Frier et a!. (1984) examined the influence of adrenergicdenervation on the IRGIP response to feeding in man. There was no significantdifference in meal-stimulated IRGIP release in sympathectomized subjects (with acomplete transection of the spinal cord above the sympathetic outflow) compared tocontrols. Blockade of cz-adrenorecetors has been shown to augment (Sirenek et a!.,1977b) or have no effect on nutrient stimulated IRGIP release (Flaten, 1981), while 13-adrenoreceptor antagonists seem to inhibit nutrient stimulated IRGIP release (Sirinek etal., 1977b; Flaten, 1981). Stimulation of x-adrenoreceptors has been shown to reduce theIRGIP response to oral glucose (Salera et al., 1982a), while B-adrenoreceptor stimulationhas been shown to increase basal and glucose stimulated IRGIP release (Flaten et a!.,1982; Kogire et a!., 1990), or have no effect (Salera et a!., 1982a).The multiple effects of the autonomic nervous system on the gastrointestinal tractprobably conthbute to the difficulty of establishing the role of its components on IRGIPrelease. Alterations in gastric emptying, intestinal secretion and absorption of fluids,motility, and blood flow by the autonomic nervous system are all likely to influenceIRGTP release. It is therefore difficult to make inferences on the direct action of thesympathetic or parasympathetic nervous system on GIP cells.381.10.6 OTHER MODULATORS OF GIP RELEASESomatostatin appears to be an effective inhibitor of GIP release. Pederson et al.(1975a) demonstrated that a rapid intravenous injection of somatostatin in dogs delayedthe increase in serum concentrations of insulin, JRGIP and glucose following oralglucose. During a sustained somatostatin infusion, the IRGIP response to oral glucosewas suppressed until cessation of the infusion, even though serum glucose levels wereelevated to the same levels as controls. This suggested that inhibition of glucose-stimulated IRGIP release was not a result of the inhibition of glucose absorption. Thishypothesis was confirmed by Creutzfeldt and Ebert (1977), who demonstrated that IRGIPand insulin responses to a test meal were completely suppressed during somatostatininfusion, whereas blood glucose levels were not different from controls. Kraenzlin et a!.(1985b) demonstrated that a long acting somatostatin-analogue, SMS 201 995, alsoeffectively suppressed the postprandial release of IRGIP.Intravenous gastrin has been demonstrated to increase the IRGIP response tointraduodenal glucose in dogs (Sirinek et at., 1977a). This augmentation could not havebeen caused by the stimulation of gastric H secretion, since acidification of the duodenalglucose did not augment the IRGIP release. Jorde et a!. (1981) further examined thisputative effect of gastrin by examining the IRGIP response to a mixed liquid test meal ina group of achiorhydric patients with high serum gastrin levels compared to a normalgroup of human subjects. A significantly higher IRGIP response was observed in theachlorhydric group than in the control group. Studies on patients after total anteroduodeno-pancreatectomy that have undetectable levels of serum gastrin show asignificant IRGIP response to a test meal, indicating that gastrin is not a prerequisite forGIP release (Creutzfeldt et al., 1976). However, these patients have a very rapid gastricemptying, which may explain their elevated serum IRGIP levels.Evidence for a relationship between GIP and gastrin has also been supplied byMorgan et a!. (1985). These investigators observed a highly significant negative39correlation between circulating gastrin and IRGIP levels after control and guar gummeals. Addition of guar gum to an oral glucose load reduced the postprandial secretionof both IRGIP and insulin. Studies in dogs by Wolfe et al. (1983) using anti-GIPantibodies have shown that IRGIP can function as a physiological inhibitor of gastric acidsecretion through its effect on gastrin release. Morgan et al. (1985) therefore suggestedthat the guar gum attenuated IRGIP response to the protein meal, leading to an attenuatedsomatostatin response and as a result unrestrained gastrin secretion. Exogenous GIPinfusions have been shown to inhibit gastrin release (Villar et al., 1976; Arnold et al.,1978), the effect seemingly mediated via an increase in gastric somatostatin secretion(McIntosh et al., 1979).In addition to gastrin, other compounds have also been demonstrated to alterIRGIP release. Increases in IRGIP levels have been noted in response to intravenousadministration of the neuropeptide gastrmn-releasing peptide (McDonald et al., 1981;Greely et al., 1986a; 1986b), while the neuropeptide calcitonin gene-related peptideinhibited basal levels of IRGIP in man (Kraenzlin et a!., 1985a). A reduction inpostprandial IRGIP release has also been noted in response to intravenous morphine(Champion et a!., 1982) and oral administration of the prostaglandins enprostil orrioprostil (Nicholl et a!., 1986; Demol and Wingender, 1989; Schwartz and Saito, 1989).It is not clear whether these actions result from direct interactions with GIP cells, orindirectly by slowing gastric emptying, reduced intestinal nutrient absorption, or by theaction of another mediating peptide or neurotransmitter.1.10.7 FEEDBACK INHIBITION OF GIP RELEASEIn 1975, Brown et a!. proposed a feedback inhibitory control mechanism of GIPrelease involving insulin. This hypothesis was based on their observation that anintravenous bolus injection of insulin during ingestion of fat significantly reduced thetriglyceride-induced IRGIP increment. Support came from studies demonstrating40attenuated IRGIP responses to oral fat ingestion during simultaneous infusion of glucose,and, thereby, stimulation of endogenous insulin release (Cleator and Gourlay, 1975;Crockett et al., 1976; Ross and Dupr, 1978; Ebert et at., 1979a). In 1978 Sirinek et at.suggested that insulin was also capable of attenuating the GIP response to oral glucose.However, Andersen et at. (1978) noted that during a euglycemic clamp with a continuousinsulin infusion resulting in hyperinsulinemia, a significant IRGIP response to oralglucose was noted. Andersen et at. concluded that glucose-stimulated GIP release wasnot inhibited in the presence of marked hyperinsulinemia. This finding was supported bythe observations that hyperinsulinemia induced by exogenous insulin (Service et at.,1978), or endogenous insulin (Collier et at., 1984) had no effect on carbohydrate-stimulated IRGIP secretion in normal subjects. Furthermore, in juvenile diabeticswithout insulin reserve, fat-, but not glucose-induced IRGIP secretion was reduced byexogenous insulin infusion (Creutzfeldt et at., 1980). In a recent investigation with rats,however, Bryer-Ash et at. (1994) observed that the IRGIP response to oral glucose wassuppressed by hyperinsulinemia, and this suppression was attenuated whenhyperinsulinemia was accompanied by hyperglycemia. It is therefore possible thatspecies differences may exist in the insulin feedback inhibition of GIP release by oralglucose.Verdonk et a!. (1980), using the glucose insulin clamp technique, questioned theeffectiveness of insulin as an inhibitor of fat-stimulated IRGIP secretion. These authorsused hypoglycemic, euglycemic, and hyperglycemic clamp conditions and found thatinsulin, under euglycemic conditions, did not exert an inhibitory effect on fat-stimulatedIRGIP levels. Since in their study, the hypoglycemic clamp increased and thehyperglycemic clamp decreased the IRGIP response to oral fat, they suggested an effectof glycemia itself on IRGIP secretion in the presence of hyperinsulinemia. However,Stöckmann et a!. (1984) postulated that the 2 h hyperinsulinemic period precedingingestion of fat in the study by Verdonk et a!. may have prevented the demonstration of41the insulin effect on fat-induced IRGIP release. Stöckmann et a!. (1984) found that theresponse of IRGIP to oral fat was inhibited by 63% if insulin infusion was started at thetime of fat ingestion, whereas no inhibition was seen if a 2 h hyperinsulinemic periodproceeded the fat load. It was therefore concluded that insulin does inhibit fat-inducedIRGIP secretion in normal man, but prior hyperinsulinemia masks this insulin effect,probably by decreasing the sensitivity of the GIP cells to insulin.Recently, Takahashi et al. (1991) compared the effect of insulin and glucose onfat-induced GIP and tGLP-I release in humans. The response of both hormones to fatingestion was measured during Continuous glucose infusion and during ahyperinsulinemic euglycemic glucose clamp. The release of GIP and tGLP-I wassuppressed in the hyperglycemic, hyperinsulinemic state. However, while GIP was alsosuppressed in the normoglycemic hyperinsulinemic state, the release of tGLP-I was not.Takahashi et a!. (1991) thus concluded, as previously reported, that insulin inhibited fat-induced GIP, but the secretion of tGLP-I was more likely inhibited by a direct action ofglucose.Dryburgh et al. (1980a; 1980b) investigated the possibility that C-peptide mightinhibit the release of IRGIP. At a dose three-times that found in the fed rat, C-peptidetotally abolished the IRGIP response of the perfused rat intestine to fat stimulation. Itwas also demonstrated that C-peptide released after stimulation of the pancreas byglucose and tolbutamide administration given intravenously in combination with insulinantiserum significantly inhibited fat-stimulated IRGIP release. However, this mechanismcould not account for the reduced IRGIP response to fat during concomitant insulininfusion observed by Stöckmann et a!. (1984) under their experimental conditions. Cpeptide levels were not elevated, and were in fact decreased by about 50% by insulininfusion, consistent with the suggested feedback inhibition of insulin secretion by insulin(Service et a!., 1978; Elahi et a!., 1982).421.11 GIP RECEPTORS1.11.1 BIOLOGICALLY ACTIVE SITE(S)It is possible, that the enterogastric and insulinotropic actions of GIP are mediatedvia two different receptor types, which interact with different regions of the polypeptide.Evidence for this hypothesis emerges from studies testing the biological activity ofvarious regions of the GIP molecule. The cleavage of GIP by trypsin was first usedduring amino acid sequence determination of the polypeptide (Brown et a!., 1970). Anunpurified mixture of tryptic fragments, including the major products 1-16, 19-30, and34-42, had no enterogastrone activity, as tested in the denervated pouch of the fundus ofthe stomach in dogs. Brown and Pederson (1970b) reported that a purified cyanogenbromide produced C-terminal fragment with amino acids 15-42 was sufficient to inhibitacid secretion in the same model. Moroder et a!. (1978) tested synthetic porcine GIP 1-38 and found it to be a “poort’ inhibitor of gastric acid secretion. However, it was laterfound that there was an extra glutamine residue in this sequence at position 30, whichperhaps conthbuted to that observation. Pederson et al. (1990) observed only a weaksomatostatinotropic effect in a combined isolated perfused pancreas and stomachpreparation of the rat with synthetic porcine GIP 1-30. Furthermore, Rossowski et a!.(1992) found that unlike the parent peptide, porcine GIP 1-30 was not able to inhibitpentagastrin-stimulated gastric acid secretion in rats. These observations, whencombined with that made by Brown and Pederson (1970b) indicate that the C-terminalportion of the polypeptide is likely the site of the acid inhibitory action of GIP.The first GIP fragment tested for insulinotropic activity was the purified cyanogenbromide fragment 15-42 produced by Brown and Pederson (1970b). In 1976, theseinvestigators tested this C-terminal peptide in the perfused rat pancreas and found that itwas approximately 40% as potent as an insulinotropic agent as intact porcine GIP(1976b). Using the same bioassay, Maletti et al. (1986) established that an HPLCpurified fragment of bovine GIP resulting from enterokinase digestion (amino acids 17-4342) was approximately 32% as insulinotropic as native bovine GIP. The N-terminalsynthetic fragments bovine GIP 1-39 (Sandberg et at., 1986), porcine GIP 1-38 (Moroderet al., 1978), human GIP 1-31 (Carquist, 1987) and porcine GIP 1-30 (Pederson et al.,1990; Gallwitz et a!., 1993) were shown to be equipotent with native GIP. In 1986,Schmidt et a!. found that HPLC purified porcine GIP 19-42 derived by proteolyticcleavage of the natural peptide was not significantly insulinotropic in isolated ratpancreatic islets. This result, when considered with those by Maletti et a!. (1986), andPederson et al. (1990), suggests that the GIP region of 19-30 does not contain all thenecessary amino acids for its insulinotropic activity, and 17-30 would be predicted toretain approximately 32% insulinotropic action. Blundell et al. (1976) pointed out thathydrophobic amino acid regions of the glucagon superfamily are probably important forreceptor binding. The amino acid sequences 6-14 and 19-27 constitute the hydrophobicregions in GIP. It is also evident that the region 19-30 has so far been shown to becompletely conserved in all species (Figure 1).Interestingly, in 1981, Brown et at. reported that the purest preparations of GIPexisting at that time, contained a minor peptide component contributing approximately5% to the total peptide content as shown by isotachophoresis and HPLC. Concomitantly,Jörnvall et at. confirmed the heterogeneity in the porcine OP preparation, and sequenceanalysis of the minor component suggested that it corresponded to GIP 3-42 (lackingresidues 1 and 2). Jörnvall et at. (1981) reported 20% contamination by this fragment inGIP preparations, and suggested that it was formed by secondary processing ordegradation, through susceptibility to attack by aminopeptidase, elastase, dipeptidylaminopeptidase or related enzymes in the intestine. Again in 1981, Brown et al.compared the insulinotropic and somatostatinotropic actions of these two components inisolated perfused rat pancreas and stomach preparations. GIP 3-42 was found to lacksignificant actions on both IRSS and insulin release when compared with intact GIP 1-42.The lack of insulinotropic activity of OP 3-42 was confirmed by Moody et a!. (1981)44and Schmidt et a!. (1986). In 1987, Schmidt et a!. reported that this shorter form of OWcomprised 32% of available natural GIP, and also found 4% CCK to be present.Furthermore, they noted that GIP 3-42 did not exhibit antagonistic activity to GIP 1-42,even at a 10-fold molar excess. Cariquist et a!. (1984) used Staphylococcus aureus V8 tocleave the N-terminal tripeptide from bovine GIP, resulting in GIP 4-42, and tested itsinsulinotropic activity in the perfused rat pancreas (Maletti et a!., 1986). In 3 Out of 6 ratstested, a “small” stimulation of insulin release as compared to GIP 1-42 was noted(—10%). It is not clear why the biological activity of inactive GIP 3-42 appears to beregained with the further removal of N-terminal amino acids.1.11.2 GIP BINDING STUDIESAn important step in confirming the actions of a peptide hormone on target tissuesis the demonstration of specific binding sites. Attempts at studying the receptors thatmediate the action of GIP on normal tissues have proven very difficult. Brown et a!.(1989) suggested two reasons to explain the unsuccessful attempts. Firstly, it was notedthat iodination of GIP resulted in a heterogenous population of iodinated peptides,potentially altering the receptor binding and biological activity. Secondly, methods usedto isolate B-cells or islets appeared to critically influence the GIP receptor. Indirectevidence for this was the observation that isolated pancreatic islets or islet cells onlyresponded to pharmacological concentrations of GIP, unlike responses to other stimuli(Schafer and Schatz, 1979; Schauder et a!., 1975; Schauder et at., 1976). This impliedthat the GIP receptor was particularly susceptible to enzymatic damage during cellisolation. Verchere (1991) was able to demonstrate binding sites for GIP on cultured ratislets using HPLC purified, biologically active‘25I-GIP. Significant displacement of theradioligand by GIP was observed at concentrations as low as 1 nM.Tumor cells have been recognized as a potential source for large numbers of OPreceptors. Malleti et a!. (1984) and Couvineau et a!. (1984) were able to demonstrate45high affinity binding sites in membrane preparations from hamster B-cell tumors. Malettiet al. (1984) ensured that the label was homogenous by HPLC purification, anddemonstrated biological activity in the perfused rat pancreas. Amiranoff et al. (1984;1985) provided evidence for specific GIP binding sites in the pancreatic tumor cell line InIII. Furthermore, GIP binding was correlated with an increase in cAMP levels andinsulin release. Binding of‘251-GIP in these studies was found to be saturable and couldnot be displaced by peptides structurally related to GIP. Both high affinity (Amiranoff etal.: KD = 7 nM; 3000 binding sites/cell, Maletti et a!.: KD = 2 nM; 219 binding sites/cell)and low affinity (Amiranoff et al. :KD = 800 nM; 150,000 binding sites/cell, Maletti et a!.:KD = 39; 1250 binding sites/cell) binding sites for GIP were identified. By using a cross-linker to prevent dissociation of 125GJP bound to In III membranes, Amiranoff et al.(1986) identified a 59-kDa membrane protein that specifically bound GIP. FunctionalGIP receptors have also been demonstrated in human insulinoma plasma membranes(Maletti et al., 1987), again with both high affinity (KD = 0.2 nM) and low affinity (KD =8.4 nM) binding sites, and in the mouse derived B-cell line, BTC3 (Kieffer et al., 1993).These receptors belong to a distinct family of G-protein coupled receptors referred to asthe secretin-VIP receptor family (Christophe eta!., 1986; Rosselin, 1986).Binding sites for GIP in vivo were reported by Whitcomb et al. (1984) using 125J..GIP in rats. Specific and displaceable binding of GIP to its receptors was demonstratedin the pancreas, glandular portion of the stomach, throughout the intestine and variousmuscle groups. No OP receptors could be demonstrated in the liver, adrenal gland,spleen, kidney, submandibular gland, testis, epididymis, prostate or seminal vesicles.Recently, using a molecular approach, Usdin et a!. (1993) cloned a novel receptor, which,when expressed in a cell line demonstrated activation only by GIP. Northern blots,reverse-transcription PCR and in situ hybridization demonstrated the receptor in tissuesknown to respond to OP (pancreas, gut, adipose tissue) and in novel locations such as theheart, brain, and inner layers of the adrenal cortex, where physiological effects of OP46have not been described. No receptor mRNA was found in kidney, spleen or liver. Asneither GIP nor its effects have been described in the central nervous system, except athigh concentrations (Ottlecz et al., 1985) these findings might suggest the presence of anovel, homologous peptide in the brain. Interestingly, the related peptide tGLP-I hasbeen observed and its release demonstrated in the rat brain (Shimazu et al., 1987;Kreymann et a!., 1989) where it might act as a neurotransmitter.As the results of biological assays with GIP fragments do not demonstrateinteractions with the receptor as either an agonist or antagonist, the availability of specificreceptor binding assays is important. Some fragments of GIP shown to be biologicallyactive (GIP 1-31; 1-30; 17-42) were shown to competitively inhibit the binding of 125j..GIP (Maletti et a!., 1987; Kieffer et al., 1993; Maletti et a!., 1986; Gallwitz et a!., 1993).On the other hand, fragments not expected to yield insulinotropic activity, such as GIP19-30 (see section 1.11.1), did not displace binding of the tracer to insulinomamembranes (Maletti et al., 1986). Likewise, the GIP fragment 1-27 was unable todisplace OW binding to RINm5F cells (Gallwitz et al., 1993). Interestingly, the GIPfragment 4-42, which demonstrated only partial insulinotropic action, was capable ofcompeting equally with GIP 1-42 for tracer binding (Maletti et a!., 1986). In this regard,GIP 4-42 may act as a partial antagonist.In 1991, Fehmann and Habener showed that the tGLP-I receptor in the glucose-responsive B-cell line HIT-T15 could rapidly and reversibly desensitize in response tosupraphysiological concentrations of tGLP-I. Preperifusion with GIP had no effect on thetGLP-I response indicating the tGLP-I receptor on the HIT-T 15 cells was distinct fromthat of OW. Gallwitz et a!. (1993) arrived at the same conclusion with the RINm5F cellline when it was observed that‘25I-GIP binding was not displaced by tGLP-I, and 125ItGLP-I binding was not altered by GIP. Fehman and Habener (1991a) also suggested thatthe OW receptor could undergo homologous desensitization from their observation thatprior exposure of HIT-T15 cells to GIP (100 nM) also reduced the insulin secretion47during stimulation with 10 nM GIP. This hypothesis was also supported by thedemonstration that high concentrations of GJP inhibit both insulin secretion from theperfused pancreas (Szecowka et al., 1982a) and cAMP generation in human (Malleti etal., 1987) and rat (Gallwitz et al., 1993) insulinoma cells, possibly by homologousdesensitization. Furthermore, in rats subjected to total parenteral nutrition for a period of6 days, there was a 30% increase in insulin release from the perfused pancreas inresponse to GIP (Pederson et al., 1985). It was concluded that subjecting the pancreas tochronic low concentrations of GIP could induce an increase in the sensitivity or numberof GIP receptors.1.12 GIP METABOLISMIn 1975, Brown et al. determined the half-life of porcine IRGIP in the plasma ofnormal subjects to be approximately 21 mm. Identical values were obtained usingporcine GJP in man by Elahi et a!. (20.4 ± 2.37 mm; 1979) and Sarson et al. (20.3 ± 1.2mm; 1982), and the metabolic clearance rate was calculated as 2.6 ± 0.1 ml/kg•min(Sarson et a!., 1982). While even higher values have been obtained for the half-life oftGLP-I in rat (39.5 ± 15.5 mm; Oshima et a!., 1988) and man (45.9 ± 8.8 mm; Oshima etal., 1991), other related peptides are cleared from the plasma much quicker, with half-lives of 5 mm or less (secretin: 2.5 mm, Häcki et a!., 1977; vasoactive intestinal peptide:1 mm, Modlin et a!., 1978; glucagon: 5.5 mm, Jaspan and Rubenstein, 1977). It thereforeappears that exogenously administered GIP is cleared relatively slowly from humanplasma. Wolfe and McGuigan (1982) measured the half-time of disappearance forporcine IRGIP in the dog at 7.6 ± 1.5 mm, thus indicating that some species differencesmight exist.In 1979, Elahi et al. reported that the prevailing state of glycemia could influencethe metabolic half-life of GIP. Using porcine GIP in human subjects, Elahi et a!. (1979)found that under euglycemic conditions, the half-life was 26.3 ± 4.57 mm, during mild48hyperglycemia (8 mM), 23.2 ± 3.16 mm, and under moderate hyperglycemia (13 mM)dropped to 13.1 ± 2.59 mm. As the moderate hyperglycemia was also associated withhyperinsulinemia, Elahi and associates (1979) were unable to conclude which factor mayhave been responsible for the shortened IRGIP half-life. In order to differentiate betweenthese two factors, Andersen et at. (1980) examined the half-life of exogenous porcineGIP in fasting dogs during hyperinsulinemia and hyperglycemia, alone and incombination. The half-life was significantly reduced by hyperglycemia alone or incombination, but there was no change with hyperinsulinemia alone, suggesting thathyperglycemia, but not hyperinsulinemia could significantly enhance the metabolicclearance rate of IRGIP. Recently, Sheu et at. (1987) also noted a greater metabolicclearance rate of porcine GIP in fed versus fasted rats. However, not all investigatorshave reported similar observations. Nauck et at. (1989) observed no difference in thehalf-life (18 mm) or metabolic clearance rate (‘—6 mlIkgmin) of human GIP infused innormal subjects during basal (5 mM) or hyperglycemic (8 mM) clamps.All pancreatic and gastrointestinal hormones must traverse the liver prior toreaching the general circulation. The liver is major clearance site for some hormones,such as insulin, of which approximately 50% of that secreted is removed with eachtranshepatic circulation (Ishida et at., 1983; Stoll et al., 1970). On the other hand, thereappears to be no hepatic extraction of endogenous or exogenous Gil’ (Hanks et at., 1984;Chap et al., 1987). The kidney also inactivates or clears several polypeptide hormones,such as insulin, glucagon and gastrin from the circulation (Chamberlain and Stimmier,1967; Sherwin et al., 1976; Davidson et al., 1973). O’Dorisio et at. (1977) found a renalarterial-venous IRGIP difference of 39% in dogs during intraduodenal perfusion ofglucose, and found that uremic patients had higher than normal fasting and stimulatedIRGIP concentrations. It therefore appears that the kidney is the major site of GIPclearance.49An important factor that must be considered when examining the metabolism of ahormone is the possibility that hormone measurements associated with immunoreactivitymay not necessarily correspond to the biological activity of that hormone. There arenumerous proteases in plasma that could potentially render a hormone biologicallyinactive, yet still immunoreactive in an RIA. Thus, while measurements of the half-lifeof IRGIP levels in plasma are useful, they must also be combined with studies examiningthe duration of biological activity of circulating GIP. The importance of this issue hasrecently been underscored by the observation that dipeptidyl peptidase IV hydrolyzes GIPin human serum to the biologically inactive fragment GIP 3-42 and thus may play animportant role in the metabolism of GIP (Mentlein et a!., 1993). Present GIP antiseramay be unable to distinguish this truncated N-terminal form, suggesting that GIP 3-42likely contributes to IRGIP measurements in the circulation. Further discussion of thisissue appears in Chapter 4 of this thesis.1.13 GIP PATHOPHYSIOLOGYThe possibility exists that GIP may be involved in the etiology ofpathophysiological states. In particular, in light of the fact that GIP is a major incretincandidate of the enteroinsular axis, the role of GIP in diseases associated with insulindeficiency and excess have been extensively investigated. The results of these studies,however, leave no clear answer to the function of GIP in these states. For instance, innon-insulin dependent diabetes mellitus (NIDDM), circulating IRGIP levels have beenreported to be increased (Brown et al., 1975; Mazzaferri et a!., 1985; Elahi et at., 1984;Cox et al., 1981; Lardinois et at., 1985; Jones et a!., 1989b; Ross et at., 1977), normal(Service et at., 1984; Levitt et at., 1980; Alam et al., 1992), or decreased (Alam andBuchanan, 1980; Service et a!., 1984; Groop, 1989; Nauck et al., 1993) following the oraladministration of nutrients. Similarly, in obese subjects, some studies have demonstratedelevated fasting levels of IRGIP (Elahi et a!., 1984; Mazzaferri et al., 1985; Salera et al.,501982a), and others exaggerated (Creutzfeldt et a!., 1978; Elahi et a!., 1984; Mazzaferri eta!., 1985; Ebert et al., 1979a; Jones et al., 1989a; Fukase et a!., 1993), normal (Lauritsenet al., 1980; Jorde et al., 1983; Amland et al., 1984; Ebert and Creutzfeldt, 1989) orblunted (Service et al., 1984; Groop, 1989) levels after oral nutrient ingestion.Among other defects in insulin secretion in NIDDM patients, there is a reduced orabsent incretin effect: insulin release is no longer stimulated more by oral as compared to“isoglycemic” intravenous glucose (Perley and Kipnis, 1967; Nauck et a!., 1986; Tronieret a!., 1985). A reduced incretin effect could be caused by impaired secretion of relevantincretin hormones or by B-cell insensitivity to their insulinotropic action. While the dataon GIP secretion are not conclusive, reports of the insulinotropic action of GIP inNIDDM patients seem to be in agreement. In streptozotocin diabetic rats, theinsulinotropic effectiveness of GIP was significantly reduced (Suzuki et a!., 1990). Inhuman diabetic subjects, there appears to be a significantly lower insulin response tointravenous GIP when compared to normal subjects (Jones et al., 1987; Krarup et a!.,1987c; Nauck et a!., 1993; Meneilly et a!., 1993). Meneilly et a!. (1993) found that thesulfonylurea glyburide significantly increased the B-cell response to GIP infusion indiabetic subjects, but the observed insulin release was still approximately 10-fold lessthan that of controls. Interestingly, Nauck et al. (1993) found no significant difference inthe insulin response to tGLP-I in diabetic vs. normal subjects. Holz et a!. (1993)postulated that since tGLP-I appeared able to compensate for a defect in the glucosesignaling pathway that regulates insulin secretion from B-cells, making them “glucosecompetent”, tGLP-I might be useful for the treatment of NIDDM.The presence of a glucose threshold for the insulinotropic action of GIP isteleologically necessary to prevent inappropriate insulin release and hypoglycemia. Sinceuntreated patients with NIDDM have plasma glucose levels above this threshold even inthe fasting state, Jones et a!. (1989) investigated whether GIP was insulinotropic duringfasting in these individuals. It was found that the glucose threshold for the insulinotropic51action of GIP was not altered in the NIDDM patients, and thus there was a significantinsulin response to intravenous GIP even during fasting. The resulting secretion ofinsulin was, however, insufficient to decrease plasma glucose levels. In normal subjects,persistently elevated GIP levels only have a short lived effect on insulin release, asplasma glucose levels are only above the necessary threshold for a relatively short timefollowing meals. Jones et al. (1989b) reasoned that in untreated NIDDM subjectsexhibiting an insulinotropic effect to GIP in the hyperglycemic fasting state, persistentlyelevated GIP would be likely to exert an effect on insulin secretion throughout the day.Obesity is often associated with glucose intolerance and hyperinsulinemia in boththe fed and fasted state (Perley and Kipnis, 1967; Creutzfeldt et a!., 1978; Chan et al.,1984). Unlike diabetics, however, B-cell sensitivity to GIP appears to be unaltered inobese humans. Intravenous infusion of porcine GIP during a hyperglycemic clampproduced a similar insulin response in both lean and obese volunteers (Amland et al.,1985). Elahi et a!. (1984) calculated B-cell sensitivity to GIP from the insulin and IRGIPresponses to oral glucose and found it normal in obese individuals. The GIP sensitivityof B-cells from the hyperinsulinemic obese Zucker rat has also been measured. Chan eta!. (1984) found that the insulin response to GIP was enhanced in the pancreas of obeseanimals and additionally, that the glucose threshold for the insulinotropic action of GIPwas well below fasting levels in these rats. Thus, as with NIDDM, there may beinappropriate GIP insulinotropic activity occurring associated with obesity.How might the insulinotropic activity of GIP become uncontrolled? A defect inthe feedback control of GIP secretion by insulin was proposed as the explanation for theexaggerated [RGIP release after food stimulation in obesity (Brown and Otte, 1978) andin NLDDM (Crockett et a!., 1976; Ross et al., 1977). Hampton et a!. (1983) studied theeffects of pre-treatment with a high fat diet on the IRGIP and insulin responses to oral fatand glucose in rats. The pre-treatment resulted in increased IRGIP secretion in responseto oral fat and abolition of the feedback inhibition of exogenous insulin on fat-stimulated52IRGIP release. It was also noted that some degree of insulin resistance was established.Willms et al. (1978), found that 5 days of caloric restriction abolished the exaggeratedIRGIP response to fat or a mixed meal in obese subjects, suggesting that a previouslyhigh calorie intake might have been responsible for the excessive IRGIP production.Creutzfeldt et al. (1978) also reported an exaggerated IRGIP response in obese subjectsgiven a high calorie mixed meal. This was not, however, observed in subjects after 5days of dietary restriction. Hampton et al. (1983) suggested that this might be due to adecreased responsiveness of the GIP secreting cells to insulin due to a reduction in insulinreceptor numbers. Investigation of insulin receptors on human monocytes has shown thatinsulin receptor number can be altered by dietary changes (Pedersen et al., 1980) and thatthese changes can occur rapidly - within 24 h (Schluter et a!., 1980).1.14 THESIS INVESTIGATIONIn summary, the hormone GLP, has been extensively studied as a potentialenterogastrone and an incretin involved in the enteroinsular axis. In addition, evidencehas been presented that suggests GIP has other roles, such as direct actions on fatmetabolism. These activities have been substantiated by the location of specific GIPreceptors on the respective target tissues. Studies have revealed specific regions of themolecule to be important for receptor interactions, and activation of intracellularmessenger systems have been associated with GIP binding. The endocrine cells thatrelease GIP have been identified, and the gene coding for the prepropeptide sequenced.Numerous GIP assays have been established allowing for the measurement of GIPresponses to the ingestion of nutrients or the intravenous administration of varioussecretagogues, and have enabled an estimation of the half-life of GIP immunoreactivity inplasma.Observations from circulating IRGIP profiles do not allow direct assessment ofthe actions of GIP secretagogues and inhibitors because of the potential interactions53among paracrine, endocrine, neural and luminal influences at the level of the GIP cell.These confounding factors can only be overcome by studying GIP release at the cellularlevel. Such methods allow the investigation of many components potentially influencingGIP release, individually and in combination. Studies at the cellular level also allowinvestigation of the intracellular mechanisms by which agents act on the GIP cell. Oneobjective of these thesis investigations was therefore to develop methods that wouldenable the study of GIP secretion at the cellular level in a controlled environment.Observations of circulating IRGIP levels have also been used to evaluate theimmunoreactive half-life of the hormone. It is not known, however, how accurately thesemeasurements reflect the time of biological activity of circulating GIP. There is evidenceto suggest that peptidases may convert GIP to a biologically inactive fragment, that isimmunologically indistinguishable from the native form. Reports of circulating IRGIPlevels may therefore have little meaning in terms of the actual physiological potency ofplasma GJP. A second aim of the thesis investigations was consequently to carefullyexamine the metabolic processing of GIP and to determine a physiological time frameduring which this process occurs. A better understanding of GIP secretion andmetabolism is critically important to resolving the potential role this hormone may havein both normal and pathophysiological situations.54CHAPTER 2GIP RELEASE FROM ISOLATED CANINE AND PORCINE ENDOCRINE CELLS2.1 INTRODUCTIONA major factor impeding study of isolated GIP cells is the diffuse nature of thedistribution of endocrine cells throughout the small intestine. Separation techniques toproduce enriched preparations of gastric endocrine cells were developed by Soil andcolleagues in 1984. These techniques were used to study the release of somatostatin andhave been modified to permit the study of the local regulation of peptide secretion frompurified gastrin (Giraud et a!., 1987; Campos et a!., 1990), cholecystokinin (Barber et al.,1986b; Koop and Buchan, 1992), neurotensin (Barber et a!., 1986a), peptide YY (Aponteet a!., 1988), motilin (Poitras et al., 1993) and enteroglucagon cells (Buchan et al., 1987).The aim of these studies was to modify these techniques to allow for the isolation,enrichment, and culture of intestinal endocrine cells in order to study the local regulationof GIP secretion.2.2 MATERIALS AND METHODS22.1 ANIMAL TISSUESInitial studies were performed with adult mongrel dogs (n = 10), but due tounavailability of dogs, the model was later adapted for use with pigs (n = 10). Fastedmale and female dogs or pigs were sedated with phentanol triperidol (0.1 mi/kg)administered with atropine (0.05 mg/kg), anesthetized with intravenous sodiumpentobarbital (30 mg/kg) and prepared for abdominal surgery. The animals were thenbled from the abdominal vena cava, and the upper small bowel was removed. Theintestine was immediately cut open and washed in ice-cold Hanks’ balanced-salt solution(HBSS; Gibco Laboratories, Burlington, Ont.) —pH 7.4 containing 0.1% bovine serumalbumin (BSA fraction V; Sigma Chemical, St. Louis, MO).5522.2 ISOLATION OF MUCOSAL CELLSThe canine or porcine mucosa was blunt dissected from the intestine, weighed andchopped finely with scalpels. Aliquots (10 g) of canine mucosa were initially digested in50 ml basal medium Eagle (BME; Gibco) —pH 7.4, with a mixture of enzymes consistingof 75 U/mi type I coliagenase (Sigma), 75 U/mi type XI collagenase (Sigma), 0.9 U/mitype IX protease (Sigma) and 1 U/mi trypsin (Worthington Biochemical Corp, FreeholdNJ) for 1 h in a shaking water bath at 37°C. The total volume was then doubled withHBSS-BSA and allowed to settle —10 mm. The supemate containing detached cells wasdiscarded, and the remaining tissue was further digested in the enzyme mixture for two45 mm periods, with each step followed by the addition of 300 1.11 of 0.5 Methylenediaminetettraacetic acid (EDTA; BDH, Toronto, Ont.) for 15 mm. A similarprocedure was followed for the porcine mucosa, but as it was much thinner, thedigestions were performed only with 75 U/mi type I collagenase (Sigma). The cellsuspension resulting from digest 3 or 4 of the canine and porcine mucosa was filteredthrough Nitex mesh (200 Lm, B & SH Thompson, Scarborough, Ont.) and washed andcentrifuged at 200 x g (Baxter Biofuge) twice with HBSS-BSA supplemented with 0.01%dithiothreitol (DTT; Sigma) and 0.001% deoxyribonuclease (DNAse; Sigma). The cellswere then filtered a second time through fine Nitex mesh (62 jim, B & SH Thompson),counted, and diluted in HBSS-BSA-DTT-DNAse to 6 x 106 cells/mi for elutriation.22.3 ENRiCHMENT OF MUCOSAL ENDOCRINE CELLSCounter-flow centrifugal elutriation of cells was first reported by Lindahl in 1948.This procedure separates cells on the basis of their sedimentation coefficients. The cellsuspension is pumped into a rotating chamber, and cells are held where theirsedimentation rate is balanced by the flow of fluid through the separation chamber.Different fractions of homogeneous cells are then ‘eluted’ by either increasing the flowrate through the chamber or decreasing the centrifugal speed. The appropriate flow rates56and centrifugation speeds were determined empirically. Batches consisting of 1.5 x 108canine or porcine dispersed cells were introduced into the Beckman elutriator (model J2-21M/E; Beckman, Los Angeles, CA) via a pump (Cole Parmer) connected to a sterilesource of HBSS-BSA. The enzyme dispersed canine cells were loaded into the elutriatorchamber at a rotor speed of 2500 rpm with a flow rate of 25 mi/mm and washed for 2 mmA 100 ml fraction (Fl) was collected after changing the flow rate to 30 mI/mm. A second100 ml fraction (F2) was obtained at a rotor speed of 2100 rpm and a flow rate of 55ml/min. The elutriation procedure for the dispersed porcine mucosal cells was slightlydifferent. Cells were loaded at a rotor speed of 3000 rpm and a flow rate of 25 mi/mm.The Fl and F2 fractions were eluted at 2800 rpm, 30 mI/mm and 2200 rpm, 55 mi/mmrespectively.Samples of cells (5 x 106, in triplicate) from the pre-eluted cells, Fl and F2 wereextracted in 1 ml of 2 M acetic acid in 1.5 ml Eppendorf microcentrifuge tubes. Thetubes were boiled for 10 mm, centrifuged for 5 mm at 7000 x g (Baxter Biofuge) at 4°Cand the supernate stored at -20°C for subsequent radioimmunoassay (RIA). Prior tobeing assayed, these extracts were neutralized with 10 M NaOH. Results indicated thatthe majority of canine and porcine IRGIP cells eluted in F2, and consequently these cellswere cultured for immunocytochemical and hormone release studies.22.4 ENDOCRINE ENRICHED CELL CULTURECanine and porcine cells from F2 were concentrated by centrifuging at 200 x g for10 mm, and then resuspended in sterile culture medium. The canine culture mediumconsisted of 47.5% Dulbecco’s modified Eagle medium (DMEM, Terry Fox Laboratory,Vancouver, BC) and 47.5% Ham’s F-12K (Terry Fox Lab), containing 5.5 mM glucose,supplemented with 5% fetal calf serum (FCS, Gibco), 2 ng/ml nerve growth factor(Collaborative Research, Bedford, MA), 8 pg/ml insulin (Sigma), 1 rig/mI hydrocortisone(Sigma), 50 pg/ml gentamycin sulfate (Sigma), 0.25 jig/mI amphotericin B (Fungizone,57Gibco), 50 U/mi penicillin (Gibco), 50 ag/ml streptomycin (Gibco), and 20 iM cytosineB-D-arabinofuranoside (Gibco). The porcine culture medium was identical with theexception of 10% heat inactivated porcine serum substituted for the 5% FCS. Plates forculturing the cells (12-well Costar, Cambridge, MA) were coated with collagen topromote cell adherence. To prepare the collagen coated plates, rat tail collagen wasdissolved in 0.0 17 M acetic acid (3.5 mg collagenlml) and then added to each well andleft for 45 mm. The solution was then aspirated, and the plates allowed to air dry prior tostorage at 4°C. Cells from F2 were seeded at 5 x 106.m11.we111 in the collagen coatedculture plates and maintained at 37°C in a humidified atmosphere of 5% CO2 in air for 40h.22.5 IMMUNOCYTOCHEMISTRYBoth cultured cells and intact intestine were prepared for immunocytochemicalanalysis. Sections of intestine (1 cm2) were fixed in Bouin’s solution (75 ml saturatedpicric acid, 25 ml 40% formaldehyde, 5 ml acetic acid) for 1 h, and then washed andstored in 70% ethanol prior to processing. The tissue was dehydrated through a gradedseries of alcohol and xylene using a histomatic tissue processor (model 166; Fisher). Thesamples were then embedded in paraffin wax using a histo-center IT-N (Thermolyne).Sections 5 I.Im thick were cut using a microtome (1 130/Biocut; Reichert-Jung) and driedon gelatin coated slides. Prior to staining, wax was removed by soaking the sections for10 mm in xylene followed by clearing in petroleum ether for 2 mm. Cultured cells werewashed 2 x in phosphate-buffered saline (8.0 g/l NaCI, 0.2 g/l KH2PO4,1.15 g/lNa2HPO4)plus 0.1% sodium azide (NaN3; PBS-azide). The cells were then fixed inBouin’s solution for 10 mm and then washed 3 x and stored in PBS-azide.Tissue sections and cultured cells were immunostained with primary antibodiesfor all hormones listed in Table 1 using the avidin-biotin peroxidase method. To preventnonspecific staining caused by the presence of endogenous peroxidase activity, samples58were immersed in PBS-azide containing 0.3% H20(BDH Inc. Toronto, Ont.) for 30 mmat room temperature and then washed in PBS-azide. The fixed cells and 5 jim sections ofjejunum were incubated with the antibodies diluted in PBS-azide containing 10% swineserum as listed in Table 1 for a period of 48 h at 4°C. After washing 3 x in PBS-azide,the biotinylated secondary antibodies (Table 1) were applied and incubated for 1 h atroom temperature. The secondary antibodies were then localized with a premixed (10mm) avidin-biotin peroxidase solution (Vectastain ABC Kit; Vector Laboratories,Burlingame CA) for 1 h. The resulting complex was identified with pH 7.6 tris buffersolution of diaminobenzidine (BDH; 125 p.g/100 ml) containing 40 mg NH4C1, 200 mgD-glucose, and 0.3 mg glucose oxidase (Sigma). Staining was monitored with a ZeissAxiophot microscope, and the reaction was terminated after the appearance of positivecells (15-45 mm) by washing in PBS-azide. In order to quantify staining, a minimum of10 groups of approximately 100 cells were observed for positive cells, which were thenexpressed as a % of the total number counted.The cultures and intact mucosa were also stained for mucin containing cells usinga combined alcian blue-periodic-acid Schiff technique. Fixed cells were treated with 1%alcian blue (Fisher, Fair Lawn, NJ) in 3% acetic acid for 5 mm and then washed indistilled water. The cells were treated with 1% aqueous periodic acid solution for 5 mmand again washed in distilled water. Finally, they were washed in Schiff’s reagent (0.5%basic fuchsin, Fisher; 0.5% sodium metabisulphite, Fisher; 0.5% hydrochloric acid, and0.5% decolorizing charcoal, Fisher) for 15 mm followed by 10 mm of running water.Neutral mucins stained red and acid mucins blue (pink and purple respectively in figures).2.2.6 RELEASE EXPERIMENTSOn the experimental day, the culture medium was removed by pipet and theporcine or canine cells washed with 1 ml release medium, consisting of DMEM (Sigma)supplemented with 25 mM NaHCO3,25 mM HEPES, 5 mM glucose, 1% FCS, and 2%59aprotinin (Trasylol, 500 kallikrein-inhibiting U/mi, Miles Pharmaceuticals, Rexdale,Ont.). This was followed by addition of 1 ml release medium containing the appropriateglucose concentration to each well. Stock solutions of other secretagogues were added as25 .tl of a 40 x concentration to 975 of release medium in 5 mM glucose, in triplicate.Cells were incubated for 2 h, after which the release medium was removed and stored in1.5 ml Eppendorf microcentrifuge tubes on ice. The release medium was centrifuged for5 mm at 7000 x g (Baxter Biofuge) at 4°C and the supernate stored at -20°C forsubsequent GIP and somatostatin RIA. Adherent cells in control wells (2 wells per plate)were extracted in 2 M acetic acid for determination of total cell peptide content per well.After centrifugation to remove particulate matter, the supernate was stored at -20°C.Prior to being assayed, these extracts were neutralized with 10 M NaOH. Peptidesecretory values from each well were expressed as percentage of total cell peptide content(%TCC) as measured from extracts.2.2.7 SECRETAGOGUESThe somatostatin antibody (SOMA 10; Regulatory Peptide Group, Vancouver,BC, Canada) was diluted in release medium to 10 jig/mI/well. Porcine gastrin releasingpeptide (GRP, Penninsula Laboratories mc, Belmont CA) was diluted to a concentrationof 4 x 10-6 M with release medium and kept on ice. This solution was serially diluted toform other stock solutions of 4 x iOfl, 4 x 10-8, and 4 x 10 M. Forskolin (Sigma) wasprepared as a stock solution of 4 x iO M in ethanol and serially diluted in DMEMrelease medium to concentrations of 4 x 10, 4 x iO, and 4 x 10-6 M. A23 187 (Sigma)was prepared as a stock solution of 4 x 10 M in dimethylsulfoxide (DMSO) and seriallydiluted in DMEM release medium to concentrations of 2 x i0 and 4 x lO M. Theeffect of the solvents ethanol and DMSO on IRGIP or IRSS release was tested by addingidentically diluted solvents to medium in corresponding control wells. Peptide values60resulting in these controls were subtracted from the appropriate values obtained withforskolin or A23 187.22.8 RADJOIMMUNOASSAY FOR GIPThe RIA used to measure immunoreactive GIP (IRGIP) was originally described byKuzio et a!. (1974) and modified by Morgan et a!. (1978). The‘251-GIP used for the RIAwas prepared by a slight modification of the chloramine-T method as was performed byGreenwood et al. (1963). The column for label purification was prepared prior to theiodination with Sephadex G-15 (Pharmacia, Uppsala, Sweden). The G-15 was first swollenin 0.2 M acetic acid with the ‘fines’ repeatedly decanted, and then degassed under watervacuum for —2 h. A 1:1 ‘slurry’ of gel and acetic acid was then poured into a 0.5 x 20 cmplastic pipet with a glass wool plug and allowed to settle at an eluent (0.2 M acetic acid plus5.0% BSA; RIA Grade) flow rate of -250 for 4-5 h. The column buffer wassupplemented with 2% aprotinin for -1 h before the iodination.The iodination reaction was performed in a siliconized test tube containing 5 p.g ofporcine GIP (Regulatory Peptide Group) dissolved in 100 p.1 40 mM pH 7.5 phosphate buffer(10 x stock prepared by titrating 0.4 M Na2HPO4with 0.4 M NaH2PO4to pH 7.5). In a fumehood behind lead, 0.5 mCi Na125 (NEN Research Products, Wilmington, DE) and 10 p.1chioramine-T (Fisher, Fair Lawn, NJ; 4 mg/ml phosphate buffer) were added while thereaction vessel was agitated. The reaction was stopped after 15 sec by the addition of 20 p.1sodium metabisulfite (Fisher; 12.6 mg/mi phosphate buffer). Separation of the free iodinefrom the 125-labeled hormone was achieved using gel filtration on the Sephadex G-15column previously poured. Fractions of 400 p.1 were collected at a flow rate of 250 p.1/mmand 10 p.1 aliquots were counted (LKB Wallace 1277 Gammamaster). A typical columnprofile obtained with these counts is shown in Figure 23.Fractions comprising the‘251-GIP peak were assayed for ‘damage’- or the ability ofthe label to bind to charcoal. The 10 jii fractions were diluted to —5000 cpm/100 p.1 of assay61buffer (pH 6.5, 40 mM phosphate buffer containing 5% CEP and 2% aprotinin). Then, 100p.1 aliquots of the diluted fractions were added in triplicate to tubes containing 900 p.1 assaybuffer. Dextran-coated charcoal was prepared by dissolving 2.5 mg/ml dextran (T-70,Pharmacia) in 40 mM phosphate buffer (pH 6.5), adding 2.5 mg/mi activated charcoal (NoritDecolorizing Carbon), 0.75% aprotinin, plus 5% CEP, and stirring for 2h at 4°C. Each tubeof diluted label received 200 p.1 of the dextran-coated charcoal and was vortexed andcentrifuged (30 mm, 1000 x g). Both the supematant and the charcoal pellet were countedand the percent of total radioactivity bound (%B) to charcoal (i.e. 125-GIP) was calculated.The fraction from the‘251-GIP peak with the highest radioactivity and the first 1-2 fractionson the descending portion of the peak usually had the greatest adsorption to charcoal (usually>95%). These fractions were also previously shown to have the greatest immunoreactivity inthe RIA, possibly due to decreased substitutions of iodine into the GIP molecule (Kuzio etal., 1974). Therefore, these fractions were pooled and diluted to 2.5 x 106 cpmIlOO p.1 in a1:1 mixture of column buffer and acid ethanol (750 mill 95% ethanol plus 15 mi/lconcentrated HC1), and stored at -20°C until used in the RIA.The rabbit anti-porcine GIP serum used for the RIA (RK343F also referred to asLMR34) was generously supplied by Dr LM Morgan (University of Surrey, UK). Negligiblecross-reactivity of this antibody was demonstrated with cholecystokinin, insulin, pancreaticpolypeptide, glucagon, secretin, and vasoactive intestinal peptide (Morgan et al., 1978). Thisantisera was stored as 100 p.1 aliquots of a 1:40 dilution in assay buffer at -20°C, and dilutedto 1:6000 for use in the RIA (final dilution in RIA of 1:3 x 10).Stock standards for the RIA were prepared from 100 p.g porcine GIP (RegulatoryPeptide Group) dissolved in 0.2 M acetic acid plus 0.5% BSA (Pentex) and 2% aprotinin to aconcentration of 100 p.g/ml. Aliquots of 1 p.g (10 p.1) were then lyophilized and stored at-20°C. When needed, an aliquot was dissolved in 100 p.1 acetic acid (100 mM) and 100 p.1assay buffer to achieve a concentration of 5 p.g/ml. This was then further diluted to 20 ng/ml62by diluting 40 p1 up to 10 ml in assay buffer, and then serially diluted in assay buffer toprovide a range of standards down to 78 pg/ml.On the first day of the assay, 200 p1 assay buffer was added to all sample tubesfollowed by either 100 p1 of porcine GIP standard (7.8 - 2000 pgIlOO p1) or 100 p1 ofunknown sample, and 100 p1 of antibody (LMR34). For the non-specific binding (NSB) andzero-binding (Bo) tubes, 100 p1 of buffer was used to replace the antibody or standard,respectively. All tubes were vortexed and incubated 24 h at 4°C. On day two, 100 p1 ofporcine GIP tracer (—5000 cpmIlOO pi) was added to each tube and vortexed. On day three,500 pi of polyethylene glycol (250 g/l H2O PEG 8000; Fisher) was added to all but the totalcounts and the tubes were vortexed and centrifuged at 1000 x g for 45 mm, after which thesupematant was decanted and the pellets were allowed to dry overnight. The specificradioactivity in the bound fraction (B) was then determined by counting the pellets for 2 mmon a gamma counter (LKB Wallace 1277 Gammamaster) and subtracting the NSB. Samplevalues were computed from a spline - smooth plot of B expressed as a percentage of the Boagainst concentration using an IBM 386 microprocessor with RiaCalc (Pharmacia).Controls for the GIP RIA were prepared by dissolving 1 pg porcine GIP(Regulatory Peptide Group) in 500 ml of 40 mM phosphate buffer (pH 6.5) containing0.5% BSA (Pentex) and 2% aprotinin (200 pg/100 jil). Aliquots of 1 ml were stored inEppendorf tubes at -20°C until use in the RIA. Inter- and intra-assay variation calculatedfrom these controls was 10 ± 2% and 5 ± 1%, respectively. Control studies were alsoperformed to examine the effect of incubation of GIl’ at 37°C in release medium, withand without isolated cells. No significant difference was noted in standard curvesprepared from porcine GIP standards under normal assay conditions or after 2 hincubation with cells.6322.9 RADIOIMMUNOASSAY FOR SOMATOSTATINThe RIA used to measure immunoreactive somatostatin (IRSS) was as describedby McIntosh et a!. (1978; 1987). Synthetic Tyr’-somatostatin was iodinated by thechloramine-T method, and the labeled peptide was separated from free 125J and damaged’material by adding 1 ml hormone free plasma followed by 20 mg microfine silica (QUSOG32; Philadelphia Quartz Co., USA). After centrifugation the pellet was washed twicewith 1 ml distilled water, and the label eluted with 1 ml acetic acid/acetone/water(0.1/3.9/4). The label was then diluted in 0.1 M acetic acid containing 0.5% BSA(Pentex, Miles Labs) to 1 x 106 cpm/10 p.! and lyophilized and stored at -20°C. On theday of the RIA, an aliquot of label was dissolved in 1 ml 2 mM ammonium acetate buffer(pH 4.6) and applied to a column (0.9 x 10 cm) of CM-cellulose (CM 52; Whatman)equilibrated in the same buffer. The column was then washed with —50 ml bufferpumped at a flow rate of 1 mI/mm and the label was eluted with 200 mM ammoniumacetate (pH 4.6). The peak fraction (2 mm/fraction) was then neutralized with 2 MNaOH, and diluted in assay buffer (50 mM barbital buffer pH 7.4 containing 0.01%merthiolate, 1% aprotinin, and 0.5% BSA) to -P3500 cpm/100 somatostatin monoclonal antibody (SOMA 03; Regulatory Peptide Group) wasused in the assay, prepared from a stock of 1:100 (in 0.9% NaC1 containing 0.1% NaN3and 0.5% BSA) stored at 7°C. Antibody was diluted to 1:1 x 106 in assay buffer the dayof the RIA for a final concentration of 1:4 x 106. This antibody was shown not to cross-react with GIP, gastrin or motilin and is directed towards the central region of thesomatostatin- 14 molecule, recognizing somatostatin- 14 and -28 equally (McIntosh et aL,1987).Assay standards were prepared by dissolving synthetic cyclic somatostatin-14(Penninsula) in 100 mM acetic acid containing 0.5% BSA (Pentex) to obtain aconcentration of 100 p.g/ml. Aliquots of 5 p.g were then lyophilized and stored at -20°C.On the day of the assay, an aliquot was reconstituted in 100 p.! cold distilled water64followed by 400 assay buffer to obtain a concentration of 10 p.g/ml. This was thenserially diluted (1:10, 1:10, 1:10, 1:20) to a concentration of 500 pglml, and thensequentially 1:2 to 3.9 pg/mi for a total of 8 standards for the RIA.The same general procedure was used for the RIA as in the GIP assay, with thefollowing differences. All the assay constituents were added to the tubes the first day inthe following order, 100 p.! standard or sample, 100-300 p.1 assay buffer, 100 .tl antibodyand 100 p.! label for a total volume of 400 p.1/tube. The assays were then vortexed andincubated for —72 h at 4°C. The bound peptide was then separated from the unbound byadding 1 ml dextran-coated charcoal (plus CEP) to each tube (except total counts),allowing the assays to sit —15 mm, centrifuging at 1000 x g for 30 mm, and thendecanting the supematant. After drying, the pellet was counted for 3 mm on a gammacounter, and sample values were computed from a spline - smooth plot of standard Bexpressed as a percentage of the Bo against concentration.Samples containing known amounts of somatostatin for controls could not be keptfor long periods of time because of loss of immunoreactivity. Interassay variation wastherefore computed from standards of 10 standard curves at 11 ± 2%. Intraas sayvariation as computed from standards (20 duplicates) placed randomly throughout theassay was 5 ± 1%.2.2.10 EXPRESSION OF RESULTSRelease data were calculated as mean ± SEM for a percentage of the total cellcontent (%TCC) to reduce variation between animals. Statistical significance wasdetermined using the StudenCs t test and was set at the 5% level. Support for using thismethod as opposed to an analysis of variance test followed by one of the a posteriori testswas provided by Sinclair (1988).652.3 RESULTS2.3.1 MUCOSA DIGESTIONThe digestion protocol described for both canine and porcine mucosal cells wasempirically determined to give the greatest yield of IRGIP cells. Other methodsattempted, such as tissue homogenization, were not successful. The duration of timebetween bleeding the animal and retrieving the tissue into cold Hank’s (period of tissueanoxia) was determined to be a critical factor for the cell survivability rate. Porcinebowel supplied by a slaughter house which was anoxic for 5-10 mm usually did not yieldsignificant numbers of viable endocrine cells. In addition, these animals were usually notfasted prior to slaughter, and the presence of pancreatic and intestinal enzymes in theintestinal lumen undoubtediy contributed to tissue degradation.While the enzymatic digestion used in dispersing the canine mucosal cellsundoubtedly resulted in cell damage, collagenase on its own was not sufficient. With justtype I collagenase, 8-10 digests were required to disperse the mucosa and this repetitiveprocess led to excessive cell death. In contrast, the porcine mucosa was much thinner anddigestion with type I collagenase was found sufficient to disperse the cells after 3-4digests. The enzyme mixture used for canine mucosa completely digested the porcinetissue in 1-2 digests, but the cells did not recover from this process. In both the canineand porcine experiments, extracts of cells were kept from each digestion for assay ofIRGIP content. It was usually noted that the later digests (3 and 4) had the greatestconcentration of IRGIP. Early digests were observed to contain greater numbers of redblood cells, cellular debris, and non-viable cells. Viability of cells from later digests was—98%, as assessed by trypan blue exclusion.2.3.2 ENDOCRINE CELL ENRICHMENT AND CULTUREThe elutriation procedure yielded two main fractions, designated Fl and F2. TheFl fraction contained most of the bacteria, red blood cells, and debris that was present in66the digested cells. The F2 fractions for both porcine and canine cells were collectedunder similar elutriation conditions, and made up a population of cells of similar size.The IRGIP and IRSS content of 5 x 106 canine cells from the eluted fractions wascompared to that of the same number of cells from a pre-elutriation (PE) sample (Figure3A). While there was no significant difference between the PE and Fl fractions, the F2fractions contained approximately 3 times the IRGIP and IRSS concentrations as the PEfractions (p<O.005). The canine F2 fraction (5 x 106 cells) contained 38 ± 6 ng IRGTPand 7.0 ± 0.6 ng IRSS. A similar enrichment of IRGIP was observed in the elutriatedporcine cells (Figure 3B), with the F2 fraction (15.4 ± 2.5 ng) containing —3 times the PEfraction (5.0 ± 0.8 ng). Although IRSS was not consistently measured in the porcineextracts, preliminary results suggest that while there was enrichment, considerably lessIRSS was present when compared to the same number of eluted canine cells (Fl: 52 pg,F2:321 pg;n=2).The culture conditions were found to be an important factor in the success of thepeptide secretion experiments. Collagen coated plates provided a suitable substrate forendocrine cells to attach to. These cells did not adhere to non-coated plates, and othersubstrates, such as poly-l-lysine, laminin, and fibronectin produced various degrees ofsuccess. As only 20 - 25% of the cells in each well adhered, the maximum number ofcells deemed possible to culture (5 x 106) were added to each well. This was important inorder to attain sufficient numbers of IRGIP cells to produce measurable levels of IRGIP.Attempts at culturing the dispersed and enriched cells in large flasks and then transferringto Eppendorfs for release experiments were also not successful. After 40 h culture, 98%of adherent cells were viable as assessed by trypan blue exclusion. The viabilitydecreased rapidly as cells were cultured longer. Extracts of canine cells contained 11.5 ±2.5 ng IRGIP/well and 1.4 ± 0.2 ng IRSS /well, while porcine cells contained 6.6 ± 0.5 ngIRGIP/well.6723.3 IMMUNOCYTOCHEMISTRY OF INTACT MUCOSA AND CULTURED CELLSFigure 4 shows canine jejunum immunostained for IRGIP using the peroxidasemethod. The IRGIP cells, shown by arrows are sparsely distributed, and account for <0.1% of the cells of the intact mucosa. A similar observation was made for IRSSdistribution (data not shown). Figures 5 and 6 show the isolated canine mucosal cellsafter enrichment and 40 h culture, immunostained using the peroxidase method for GIPand SS, respectively. At the time of plating, the F2 fraction contained mainly single,spherical cells, but after culture, endocrine cells appeared to flatten and adopt a moreovoid or triangular cell appearance (see Figure 5 & 6). During the culture period,adhered clusters of up to 200 cells formed, including the endocrine cells, and these cellswere observed to form connections with neighbouring cells. IRGIP cells accounted forapproximately 10% of the adherent canine cell population, while IRSS cells accountedfor 5% on average. A similar proportion of cultured porcine cells stained positive forIRGJP (8%), while IRSS cells added up to —1% of the adherent population (data notshown). A large portion of cells in the intact mucosa were mucin containing (Figure 7),and accounted for —80% of the adherent canine (Figure 8) and porcine cell population.Less than 1% of the canine or porcine cultured cells stained for motilin, secretin,gastrin/CCK, glicentin, or neurotensin (data not shown).2.3.4 IRGIP SECRETION IN RESPONSE TO KThe effect of depolarization of the isolated canine and porcine endocrine cells wasinvestigated using K. Figure 9 shows the IRGIP secretory response to increasingconcentrations of K from 10 to 55 mM over a 2 h period from canine and porcine cells.Basal release of IRGIP in 5 mM glucose, 5 mM K was 2.7 ± 0.4 %TCC (310 ± 45pg/ml) for canine cells and 4.6 ± 0.4 %TCC (304 ± 23 pg/ml) for porcine cells.Potassium concentrations ranging from 20 to 55 mM significantly stimulated IRGIPrelease from canine cells in a concentration-dependent fashion when compared to basal68(p<O.O5). A similar response was observed in the porcine cells, however, a significantincrease in IRGIP release was also observed at 10 m’vI K (p<O.O5). Stimulation at 55mM K was 18.8 ± 2.1 %TCC (2.2 ± 0.2 ng/weli), or approximately 7 times basal release(p<O.OO1)for canine cells and approximately 4 times basal release (p<O.OOl) for porcinecells (19.9 ± 2.0 %TCC or 1.3 ± 0.1 ng/well). Basal release of IRSS from canine cellswas 2.0 ± 0.2 %TCC (28 ±3 pg/mi) and increased significantly by 55 mM K to 4.1 ± 0.3%TCC (57 ±4 pg/well; p<O.O5, n =3, data not shown).23.5 IRGIP SECRETION INRESPONSE TO GLUCOSEFigures 10 and 11 show the effects of graded glucose concentrations from 5 to 20mM on IRGIP secretion from canine and porcine cells, respectively. In the canine cells,the release of IRGIP was enhanced significantly (p<O.O5) over basal by glucose levels of15 and 20 mM, while the addition of 10 mM glucose had no significant effect. Incontrast, IRGIP release from porcine cells was significant at glucose concentrations 10mM (p.cz0.O5). With the addition of 10 p.g/well somatostatin antibody SOMA 10 to thecanine cells, IRGIP release in the presence of 5 mM glucose significantly increased to 6.1± 0.7 %TCC (p<O.O5), comparable to that induced by 20 mM glucose alone (Figure 10).Glucose had a minimal effect on IRSS release from canine cells, yielding a significantincrease only at 20 mM glucose (from 2.0 ± 0.2 %TCC to 2.8 ± 0.3 %TCC; p<O.05, n =4, data not shown).2.3.6 IRGIP SECRETION IN RESPONSE TO A23187The involvement of Ca2 in the release of IRGIP was investigated using the Ca2ionophore A23 187. Figure 12 shows the effect of A23 187 at concentrations of 1, 5, and10 .tM on IRGIP release from canine epithelial cells in the presence of 5 mM glucose.While the addition of 5 p.M A23 187 produced only a small but significant (p<O.05)increase over basal IRGIP release, the addition of 10 p.M A23187 resulted in an IRGIP69output approximately 4.1 times basal (to 11.2 ± 1.8 %TCC; p<O.Ol). IRSS release wasincreased significantly by 10 pM A23 187, from 2.0 ± 0.2 %TCC to 3.9 ± 0.6 %TCC(p<O.O5, n =4, data not shown).2.3 .7 JRGIP SECRETION INRESPONSE TO GRPTo investigate the potential for peptidergic control of IRGIP secretion, the effectof graded concentrations of porcine GRP on IRGIP release from canine epithelial cells inthe presence of 5 mM glucose was examined (Figure 13). GRP significantly (p<O.O5)stimulated IRGIP release in a concentration-dependent fashion with significant releaseoccurring at a concentrations of 1 nM and greater. At 100 nM GRP, IRGIP release was7.2 ± 1.0 %TCC, approximately 2.7 times basal (p<O.Ol). This concentration of GRPalso significantly increased IRSS release from 2.0 ± 0.2 %TCC to 2.9 ± 0.4 %TCC(p<O.O5, n =4, data not shown).2.3.8 IRGIP SECRETION IN RESPONSE TO FORSKOLINA possible role for adenylate cyclase in IRGIP secretion was investigated usingforskolin. Figure 14 shows IRGIP release from canine epithelial cells in response toforskolin concentrations ranging from 0.1 to 100 .tM. Forskolin at a concentration of 1JiM significantly (p<O.O5) increased IRGIP release over basal. The maximum effectobserved was achieved with the highest forskolin concentration tested (100 JiM) whereIRGIP release was approximately 4.9 times basal (to 13.1 ± 1.8 %TCC; p<O.Ol).Concentrations of forskolin from 1-100 JiM also significantly increased IRSS release,with 100 JiM yielding 3.3 ± 0.5 %TCC compared to 2.0 ± 0.2 %TCC in the absence offorskolin (p<O.05, n =4, data not shown).70Peptide Dilution Source SpeciesGastrin/CCK 1:100 Walsh MouseGlucagon 1:100 Gregor MouseSecretin 1:500 Polak RabbitSomatostatin 1:1000 RPG MouseInsulin 1:1000 RPG Guinea PigGlicentin 1:1000 Moody RabbitNeurotensin 1:2000 RPG MousePP 1:2000 Buchanan RabbitGIP 1:5000 RPG MouseMotilin 1:5000 RPG MouseRabbit IgG 1:300 Vector GoatMouse IgG 1:300 Vector GoatGuinea Pig IgG 1:300 Vector GoatTABLE 1: Antisera used in this thesis.CCK- cholecystokinin; GIP-gastric inhibitory polypeptide; PP- pancreatic polypeptideRPG- Regulatory Peptide Group; Dr. 3. Walsh- Cure, UCLA; Dr. J. Polak -Hammersmith Hospital, RPMS; Dr. A. Moody- Novo Nordisk, Cophenhagen; Dr. K.Buchanan - University of Belfast; Vector Laboratories Inc.- Burlingame CA.71I IFRACTIONFIGURE 3: A) IRGIP and IRSS content (ng/5 x 106 cells) from the canine single cellsuspension prior to elutriation (PE) and the two fractions resulting from eluthation (Fl &F2). For IRGIP n = 9, for IRSS n = 4. B) IRGIP content from isolated porcine cellfractions. n = 10. * represents significance to a level of p<O.05 when compared to PE.Q IRGIPIRSS*iABT TTI50-. 20-1o.0.20.In 10-c5.0.I IPE Fl F2FRACTIONPE F2c-,/i,tfç. \p.L’;:‘31- I,.— ,:v..i\) ‘:72FIGURE 4: Canine jejunum immunostained for GIP using the peroxidase method.Magnification is 50 x. CM = circular muscle, MM = muscularis mucosae, SM =submucosa, V = villus./ ,f73FIGURE 5: Isolated canine epithelial cells after 40 h culture, immunostained for GIP.Magnification is 400 x. Arrow indicates a positive cell.74,_FIGURE 6: Isolated canine epithelial cells after 40 h culture, immunostained forsomatostatin. Magnification is 200 x. Arrow indicates a positive cell.r75FIGURE 7: Canine jej unum stained for neutral (red) and acid (blue) mucins usinga combined alcian blue-periodic-acid Schiff technique. Magnification is 50 x.76• 4%•LFLL :/V41liir•\. •.-• ••FIGURE 8: Isolated canine epithelial cells after 40 h culture, stained for neutral (red)and acid (blue) mucins using a combined alcian blue-periodic-acid Schiff technique.Magnification is 400 x.7725.CANINE20. PORCINE15. *10.K (mM)FIGURE 9: Effect of K concentrations of 10 to 55 mM on basal IRGIP release (5 mMK+) from isolated canine and porcine endocrine cells in the presence of 5 mM glucose. Inthis and subsequent figures, levels are expressed as mean ± SEM %TCC. IRGIPresponses were compared to basal release (5 mM glucose, 5 ni?vI K+).* p<0.05, canine n =5, porcine n = 8.**I*05I I35 55788-0—r — —r — —r- — —r- — —r -5 5+cS 10 15 20GLUCOSE (mM)FIGURE 10: Concentration-response relationship between glucose (5 to 20 mM) andIRGIP secretion from isolated canine epithelial cells. 5+czS indicates 5 mM glucose with10 jig/ml anti-somatostatin monoclonal antibody (n = 4). IRGIP responses werecompared to basal. *p<005, n 6.79FIGURE 11: Concentration-response relationship between glucose (5 to 20 mM) andIRGIP secretion from isolated porcine epithelial cells. IRGIP responses were comparedto basal. *p<O.05, n =7.15***010 15GLUCOSE (mM)208015.**5.________ __—__A23187 (jiM)FIGURE 12: Effect of 1, 5, and 10 jiM A23187 on basal release of IRGIP in thepresence of 5 mM glucose. IRGIP responses were compared to basal. *p<0.05, n = 10.8110.*8 J*c.)*4. -I2-0-— 1 —0 0.1 1 10 100GRP (nM)FIGURE 13: Effect of graded concentrations of GRP (0 to 100 nM) on IRGIP releasefrom cultured canine epithelial cells. IRGIP responses were compared to basal. *p<O.05,n =4.8215.1** _I___:r_0 -_____________— —r —0 0.1 1 10 100FORSKOLIN (j.tM)FIGURE 14: Effect of 0.1 to 100 pM forskolin in the presence of 5 mM glucose onIRGIP release from isolated canine epithelial cells. IRGIP responses were compared tobasal. *p<O.05, n =4.832.4 DISCUSSIONAlthough it is known that IRGIP is released into the circulation in response to ameal, the mechanisms by which the components of a meal in the lumen of the smallintestine act to cause GIP secretion are unknown. A greater understanding of thesemechanisms will only become available with in vitro experiments to explore the directeffect of various substances on the GIP cell. Studying the release of IRGIP at the cellularlevel requires enrichment of the otherwise diffusely located endocrine cells or the use oftumor-derived cell lines which express IRGIP. The objective of these experiments was todevelop a model for the isolation, enrichment and culture of canine and porcine intestinalendocrine cells.Dispersion of canine and porcine duodenal and jejunal mucosa by sequentialenzyme digestion progressively increased the IRGIP content of the resulting cellsuspension. This cell suspension was further enriched for endocrine cells by elutriationfollowed by culture. The culture technique has often been found useful after counterfiowelutriation to facilitate further enrichment in the studied cell content by providingselective attachment of the endocrine cells to the culture dishes (Aponte et al., 1988;Barber et a!., 1986; Buchan et a!., 1987; Giraud et a!., 1987; Koop and Buchan, 1992;Soil et a!., 1984). The overall enrichment of cultured endocrine cells was as much as100-fold, with IRGIP and IRSS cells accounting for 10% and 5% respectively of theadherent canine cells, plus 8% and 1% respectively of the attached porcine cellpopulation. The majority of the remaining cells were mucin and goblet cells, stainingpositive for neutral and/or acidic mucins.The fact that 1 to 2 days was required for cells to adhere and adopt a more normalappearance indicates that these cells likely sustained damage as a result of the chemicaland mechanical forces of the digestion procedure. Using similar methods, Barber et al.(1986) documented the beneficial effects of short term culture of isolated canine ilealmucosal cells. Freshly isolated neurotensin containing cells incubated at 4°C for 60 mm84released —12% of their neurotensin content, while after 48 h culture, basal release was—1%. The culture period was thus important for the isolated cells to functionallystabilize. In the present investigations, basal release of IRGIP was 2.7 ± 0.4 %TCC(canine) and 4.6 ± 0.35 %TCC (porcine), values similar to those previously reported fromother comparably studied isolated cells.Membrane potential difference plays an important role in the control of manycellular processes, including endocrine secretion (Barber et a!., 1986; Matthews andSakamoto, 1975; Taraskevich and Douglas, 1984). To assess the ability of the cells torelease peptide in response to a depolarizing stimulus, cells were exposed to a series ofK concentrations. Basal IRGIP release in the presence of 5 mM K was significantlyincreased by K concentrations of 20 mM and greater for canine cells and from 10 mMfor porcine cells. Release in response to 55 mM K was approximately 7 times and 4times that of the basal canine and porcine cell release, respectively. This sameconcentration of K doubled IRSS release compared to basal in the canine cells. Klikely produces these effects by depolarizing the cells, thus opening Ca2 channels andincreasing the intracellular Ca2 concentration, leading to peptide release. IRGP releaseby K indicated these cells were electhcally active and that depolarization resulted inexocytosis.Glucose is thought to be one of the most important physiological stimuli for GIPsecretion in vivo. The isolated porcine and canine endocrine cells responded to gradedglucose by releasing IRGIP in a concentration-dependent fashion. There was, however,no significant increase in IRGIP release from the canine endocrine cells until aconcentration of 15 mM glucose was reached in the release medium. Normalphysiological diets result in luminal concentrations of glucose in dogs of < 20 mM(Ferraris et al., 1990). Therefore, one would expect canine GIP cells to be sensitive tolower concentrations of glucose then was observed in this study. A probable explanationfor the apparent lack of sensitivity of the isolated canine endocrine cells to glucose is that85the GIP cells may have been tonically inhibited. Unlike the intact intestinal mucosa, theisolated epithelial cells in this model are in a non-vascular system. Metabolic productsand other secreted peptides that would normally be removed by the circulation remain inthe medium. Immunocytochemistry of the cultured canine intestinal mucosal cellsrevealed that cultures contained 5% IRSS cells, and IRSS levels during basal conditionswere 28 ± 3 pg/mi. IRSS release was increased modestly by 20 mM glucose to 39 ± 1pg/mi. Although IRSS cells may not appear adjacent to IRGIP cells in the intact gut(Buchan et at., 1982), somatostatin has been previously demonstrated to inhibit IRGIPrelease in man (Salera et al., 1982b), dog (Pederson et a!., 1975a) and rat (Ho et at.,1987). It is possible, therefore, that endogenous somatostatin present in the mediuminhibited IRGIP release. Support for this is the observation that the addition ofsomatostatin immunoneutralizing antibody (Seal et al., 1987) to basal mediumsignificantly increased IRGIP release. The significant IRGIP release from porcineendocrine cells at a lower glucose concentration (10 mM) might reflect the fewer IRSScells present, and resultant lower levels of endogenous IRS S.Considering that glucose transport is a requirement for GIP release (see section1.10.2), the sensitivity of the isolated cells to glucose may also be a reflection of theactivity of luminal glucose transporters. It is possible that these transporters weredamaged by the digestive process, and were not functioning normally after 40 h culture.Alternatively, these transporters which exist in a glucose-free environment (during theinterdigestive period), may be down regulated by 40 h culture in 5.5 mM glucose. It isalso possible that the GIP cells may not possess luminal glucose transporters and thusrequire messages from neighbouring ‘glucose-sensitive’ cells in order to release GIP inresponse to luminal glucose. The degree of IRGIP response to glucose may therefore berelated to the degree of reaggregation of the cells upon the seeding on collagen coatedplates in order that the appropriate connections are made. Thus other factors could86potentially contribute to the weak IRGIP response to glucose observed from the isolatedcells.The Ca2 ionophore A23187 increases cytosolic Ca2 concentrations.Concentrations of 5 .tM and greater were able to significantly increase IRGIP releasefrom the cultured canine epithelial cells . At 10 .tM A23187, release of IRGIP wasapproximately 4.1 times that of basal. These data suggest that signal transductionmechanisms in the IRGIP cell involve an increase in intracellular Ca2 concentration.IRSS levels were also increased to twice that of basal by 10 tM A23 187. Somatostatinsecretion from isolated canine fundic mucosal cells (Chan and Soil, 1988) or humanantral cells (Buchan et a!., 1990) has been reported to be stimulated by treatment withA23 187. In contrast, cultured fetal rat intestinal cells did not respond to modulation ofCa2 fluxes (Brubaker et at., 1990). It is not known if the action of A23 187 on IRSSsecretion observed in the present study was direct, or whether IRGIP release in responseto A23 187 stimulated IRSS release. GIP has been shown previously to stimulate gastricIRSS release (McIntosh eta!., 1981).Gasthn releasing peptide (GRP) has been located in neurons and nerve fibres ofcanine duodenum and jejunum (Vigna et at., 1987). In rats and dogs, GRP infusion hasbeen demonstrated to cause IRGIP release (Greely et at., 1986a, 1986b; McDonald et at.,1981) although it was not known whether this was a direct effect on Gil? cells. GRP isthought to act through 1P3 to release Ca2 from intracellular stores (Gallacher et at.,1990). It was therefore of interest to determine the effects of GRP on IRGIP secretion inthe isolated epithelial cell model. Concentrations of GRP from 1 to 100 nM produced asignificant increase in IRGIP release over basal in a concentration-dependent fashion.This provides evidence that GRP may be involved in vivo in the control of GIP release,possibly by an IP3/Ca2mediated pathway. It is, however, notable that other isolatedendocrine cell preparations have yielded significant responses to GRP at concentrationsas low as 0.01 fM (Campos et a!., 1990). It is possible, therefore, that IRGIP release in87response to GRP observed in the present study was mediated by a different subtype of theGRP receptor, or via a related receptor, such as that for neuromedin B. The addition of100 nM GRP also resulted in a modest increase in IRSS release (to 2.9 ± 0.4 %TCC).GRP infusion in dogs has previously been demonstrated to increase plasma IRSS(Schusdziarra et a!., 1980).To determine if activation of cAMP-dependent intracellular pathways wouldresult in IRGIP secretion from the cultured canine epithelial cells, forskolin, an activatorof adenylate cyciase was tested. Forskolin at concentrations of 1 to 100 iM producedsignificant increases in IRGIP release when compared to basal. The effect of 100 p.Mforskolin was disproportionally higher than at lower concentrations tested, and may bereflective of the lack of specificity of forskolin for adenylate cyclase at this dose. Highconcentrations of forskolin may also affect a variety of ion channel functions,independent of cAMP systems (Leidenheimer et at., 1990; Ticku and Mehta, 1990;Watanabe et a!., 1987). With these limitations in mind, it appears that canine GIP cellsare responsive to activation of adenylate cyclase. Whether the effects on stimulation ofCa2 and cAMP systems yield secretory responses that are additive or potentiating is asubject for further investigation. Forskolin concentrations of 1 to 100 p.M alsosignificantly increased the release of IRSS (p<O.O5). Somatostatin release in response todibutyryl cAMP or forskolin has been reported in isolated canine fundic somatostatincells (Soil et a!., 1984) in human antral somatostatin cell cultures (Buchan et at., 1990),and in fetal rat intestinal cells in culture (Brubaker et al., 1990).In conclusion, a method has been established to enrich duodenal/jejunal endocrinecells, yielding large adhered cell clusters in culture of approximately 10% IRGIP in thecanine and 8% IRGIP in the porcine preparation. Both depolarization by K and nutrientstimulation with glucose yielded an increase in IRGIP release from the cultured canineand porcine cells. In addition, pharmacological investigation of potential signaltransduction systems involved in stimulus-secretion coupling in canine GIP cells88suggested the participation of both cAMP and Ca2 - dependent pathways. Evidence wasalso provided for a direct role of GRP on receptor-dependent stimulation of IRGIPrelease. The use of a somatostatin immunoneutralizing antibody indicated that basalIRGIP release was tonically inhibited by exogenous IRS S. The fewer IRSS cells in theisolated porcine cells may, therefore, have contributed to the greater basal release andsensitivity to lower glucose or K concentrations. IRSS secretion was significantlystimulated by high levels of all the above agents used to examine IRGIP release, but it isunclear if these are direct actions or a result of endogenously released peptides such asGIP. This model may prove useful for further elucidating the cellular mechanismscontrolling the release of the intestinal endocrine hormones OW and somatostatin.Furthermore, it may be possible to adapt these methods to isolate epithelial cells from therat, where models for disease states potentially characterized by abnormal GIP release,such as obesity, exist. Such studies at the cellular level may be the only way to resolvepossible aberrant OW release, and the mechanisms responsible.89CHAPTER 3GIP RELEASE FROM A TUMOR-DERIVED CELL LINE (STC 6-14)3.1 INTRODUCTIONThe rationale for studying GIP release at the cellular level was presented inchapter 2. By using cell preparations, the culture environment can be controlled andprogressive changes in intracellular and intercellular events can be directly monitored(Hassall et a!., 1989). A major factor impeding such studies is the diffuse disthbution ofthe entero-endocrine cells (including IRGIP cells) throughout the small intestine. Onestrategy employed has been the development of methods for the isolation enrichment andculture of gut endocrine cells, and the adaptation of these methods to specifically examineGIP release (chapter 2). While this model has proven useful for studying the mechanismscontrolling IRGIP release, endocrine cell yield remains a critical problem.Recently, hormone secreting cell lines have been obtained from neuroendocrinetumors of transgenic mice expressing oncogenes in neuroendocrine cells. The generalstrategy involves the construction of a hybrid gene composed of the regulatory regionfrom a hormone gene linked to the coding region of an oncogene. This hybrid gene isthen transferred into a mouse germ line via microinjection of fertilized eggs and theresulting transgenic mice are able to express the oncogene in the cell types that normallyexpress the hormone regulatory sequences (Hanahan, 1988). Recently, two lineages oftransgenic mice were crossed to produce double transgenics. One of these lines carried ahybrid gene construct linking the rat insulin promoter, which drives expression inpancreatic B-cells, to the SV4O early region encoding the potent oncogene large Tag(RIP lTag2; Hanahan, 1985). The other line carried a polyoma small T antigen genelinked to the rat insulin promoter (RIP2PyST1; Seth et a!., 1991). Progeny of these micefollowed two distinct tumorigenesis pathways, either resulting in the development ofencapsulated B-cell tumors of the pancreas, or unexpected neuroendocrine cell tumors of90the small intestine (Grant et a!., 1991). These tumors were excised, minced and injectedas small clumps into mice to yield cell lines (Grant et a!., 1991). The tumor cells did notstain with periodic acid-Schiff/alcian blue, indicating a lack of mucus secretion, butretained strong positivity for the Grimelius’ silver method and chromogranin A, revealingtheir endocrine nature (Rindi et al., 1990). It appeared that a well differentiatedendocrine cell producing secretin was the original target of transgene expression in thegut of RIP1Tag2IRIP2PyST1 mice, as T-antigen-positive mucosal neuroendocrine cellsonly expressed secretin (and not the other tested hormones). A high proportion of cellsstained for secretin, and thus the name ecretin producing lumor Lells (or -Iype Lell;abbreviated STC) was given to the cell line. Other cells stained for the different areas ofthe proglucagon molecule (GLPI, GLPII, glicentin and glucagon) and scattered rare cellsreacted to neurotenisn and pancreatic polypeptide antisera, while no immunoreactivitywas detected for insulin in the STC-1 cells (Rindi et al., 1990). Experiments reported inthis chapter explored the use of this tumor cell line to study GIP release at the cellularlevel.3.2 MATERIALS AND METHODS32.1 CULTURE AND PRODUCTION OF STC 6-14 CELL LINEThe STC-1 cell line was kindly supplied by Dr. D. J. Drucker (Banting and BestResearch Group, Department of Medicine, Toronto Ont). Cells were cultured inDulbecco’s modified Eagle’s medium (DMEM; Terry Fox Laboratory, Vancouver, BC)supplemented with 25 mM glucose, 2.5% fetal calf serum (FCS; Sigma Chemical, St.Louis, MO), 12.5% horse serum (HS; Gibco Laboratories, Burlington, Ont.), 1-glutamine(Sigma, 0.15 mg/mi), penicillin (Gibco, 50 units/mi), and streptomycin (Gibco, 50 p.g/ml)at 37°C in 5% CO2 in air in 250 ml Falcon tissue culture flasks (Becton DickinsonCanada Inc., Toronto, Ont.). The cells were subcultured as required or whenapproximately 80% confluency was reached, by harvesting with trypsin-EDTA (Sigma).91Some cells were cultured on collagen coated 12-well plates until —80% confluency wasachieved and fixed in Bouins for immunocytochemical analysis, as described in section2.2.5. STC-1 cells were also routinely frozen in media containing DMSO and stored inliquid nitrogen.Data from immunocytochemical studies indicated that a small number of STC- 1cells were immunoreactive for GIP. In an attempt to increase the proportion of cellsexpressing GIP, the STC- 1 cells were sub-cloned. Cells were plated at an average of 1cell/well in 96 well Falcon culture plates (10 plates; Becton Dickinson Canada Inc.) andcultured as above to produce clones. Supematant from each well containing clones wasstored at -20°C for later IRGIP determination using the GIP RIA described in section2.2.8. Cells from RIA-positive wells were transferred to progressively larger wells toincrease cell numbers, and also immunostained to verify the presence of IRGIPcontaining cells (see section 2.2.5 for methods). Sub-clones were produced with varyingproportions of IRGIP expressing cells. However, those clones with a majority of IRGIPcells appeared slow growing and difficult to culture. A stable clone, STC 6-14, wastherefore chosen as the cell line to use for IRGIP release studies. The STC 6-14 cellswere transferred to culture flasks and maintained as above to establish sufficient cellnumbers for experiments to be conducted. Cells were routinely frozen in mediacontaining DMSO and stored in liquid nitrogen. The cells were subcultured as requiredor when approximately 80% were confluent, by harvesting with trypsin-EDTA. Cellsfrom passages 26-39 were used in this study.32.2 RELEASE EXPERIMENTSFor IRGIP secretory studies, 5 x iO cells in 1 ml of culture medium were platedin each well of collagen coated Costar 12-well plates and cultured for 4 days. On theexperimental day, the medium was removed by pipet, and changed to 5 mM glucoseDMEM (prepared as described above) and incubated 4-5 h. The culture medium was92then removed by pipet and the cells washed with 1 ml release medium consisting ofDMEM supplemented with 5 mM glucose, 1% FCS, and 2% aprotinin (Trasylol, 200kallikrein-inhibiting U/ml, Miles Pharmaceuticals, Rexdale, Ont.). This was followed byaddition of 1 ml of release medium containing the appropriate stimulus to each well.Each stimulus was tested in triplicate, for n = 1. The cells were incubated for 2 h, afterwhich the release medium was removed and stored in 1.5 ml Eppendorf microcentrifugetubes on ice. Samples were centrifuged for 5 mm at 7000 x g (Baxter Biofuge) at 4°Cand the supematant stored at 200C for subsequent peptide determination. Adherent cellsin control wells (2 wells per plate) were extracted in 2 N acetic acid for determination oftotal cell peptide content per well. After boiling 10 mm and centrifugation to removeparticulate matter, the supernatant was stored at -20°C for subsequent IRGIP orimmunoreactive SS (IRSS) assay. IRGIP and IRSS secretory values from each well wereexpressed as percentage of total cell content (%TCC) of IRGIP or IRSS as measured fromextracts. Control plates that were not used for the secretion experiments were fixed withBouins for 10 mm and then stored in phosphate buffered saline (PBS) to be subsequentlyexamined for peptides by immunocytochemistry.32.3 SECRETAGOGUESPorcine GIP (pGIP; Regulatory Peptide Group, Vancouver, BC) was diluted to aconcentration of 4 .tM with glucose-free DMEM release medium and kept on ice. Thissolution was serially diluted to 400, 40, and 4 nM. Final concentrations of pGJP per wellwere obtained by adding 25 jil of the different pGIP solutions to 1 ml medium in eachwell. An appropriate volume of a solution of 400 nM SS (SS-14; Peninsula, BelmontCA) was added to release media to give a final concentration of 10 nWwell before themedia were added to the appropriate wells. The SS antibody (SOMA-lO; RegulatoryPeptide Group) was diluted in release medium to 250 p.glmllwell. Synthetic porcine GIPused for HPLC standard was obtained from Peninsula.9332.4 PEPTIDE QUANTIFICATIONPeptides (IRGIP & IRSS) were quantified by RIA as described in sections 2.2.8and 2.2.9. Acetic acid extracts were adjusted to pH —7 with 10 M NaOH prior toassaying.32.5 IMMUNOCYTOCHEMISTRYThe STC 6-14 cells were immunostained with primary antibodies for allhormones listed in Table 1 using the avidin-biotin peroxidase method as previouslydescribed in section HIGH-PERFORMANCE LIQUID CHROMATOGRAPHY (HPLC)Reverse-phase HPLC was used to characterize the IRGIP extracted from STC 6-14cells, and in the culture medium. A pellet of 1.5 x iO cells was extracted with 5 ml 2 Nacetic acid, followed by boiling 10 mm and centrifugation (7000 x g, 10 mm). Thesupernatant was then dried using a SpeedVac (Savant Instruments Inc., Farmingdale, NY)and stored at -20°C. Medium that had cultured STC 6-14 cells for 4 days was also processedfor HPLC analysis. Samples of 20 ml were adjusted to pH —2 with acetic acid. SepPakcartridges (C18; Waters) were prepared by priming with 5 ml acetonitrile plus 0.1%thfluoroacetic acid (CH3CN + 0.1% TFA), followed by 10 ml H20 (plus 0.1% TFA). Themedia samples were then applied to the cartridge and washed with 50 ml H20 plus TFA,followed by 20 ml 10% CH3N plus TFA. It was then eluted with 2 ml 60% CH3N plusTFA, and the products were dried by Speed Vac and stored at -20°C. An aliquot of purifiedculture medium or the cell extract (10 .tg/jil) was dissolved in distilled filtered water and of this solution was loaded onto a tBondpak C18 column (4.6 x 250 mm; Waters). TheHPLC equipment (Waters) consisted of a model 512 WISP autoinjector, two model 510pumps, a model 441 absorbance detector, and a NEC computer and printer for signalprocessing using the Waters Maxima 820 program. The loaded material was eluted using a94gradient that had previously been empirically determined to resolve the components ofnatural porcine GIP. Samples were loaded at 28% CH3N plus 0.1% TFA, and after 1 mm,were eluted by a CH3N gradient (28- 33% in water) containing 0.1 % TFA, run over 10mm at a flow rate of 1 mI/mm. The column was then maintained a further 3 mm at 33%CH3N followed by a 4 mm gradient up to 70% CH3N. After 2 mm at 70% CH3N, it wasthen re-equilibrated to 28% CH3N over 2 mm. The run was complete after a 10 mm periodof equilibration, and was followed by 100 p.1 H20 injected and subjected to the same gradientto ensure no residual GIP remained on the column and was collected in the next sample run.Fractions were collected every 0.5 mm and were dried using a SpeedVac prior to being storedat -20°C. Samples were then dissolved in 500 p.1 assay buffer prior to being assayed forIRGIP.32.7 EXPRESSION OF RESULTSRelease data are presented as mean ± SEM for a percentage of the total cellcontent (%TCC). Statistical significance was determined using the Student’s t test andwas set at the 5% level.3.3 RESULTS3.3.1 CHARACTERIZATION OF STC 6-14 CELLSSub-cloning of the STC- 1 cell line resulted in a stable cell line, designated STC 6-14, that consisted of approximately 30% IRGIP cells (Figure 15). Roughly the sameproportion of cells exhibited SS immunoreactivity, and <2% immunostained for eitherglucagon, gastrin, secretin, glicentin, pancreatic polypeptide, or neurotensin. Noimmunoreactivity for insulin was observed. In order to examine IRGIP expressed in STC6-14 cells, HPLC analysis of cell extracts was employed. The narrow gradient waschosen to resolve the components of natural porcine GIP in a fashion similar to thatwhich had been previously reported (see section 1.1 1.1). Three major components of95natural porcine GIP were resolved, with retention times of 11.37, 12.47 and 13.12 mmrespectively, while synthetic porcine GIP eluted as a mono-component at 13.01 mm(Figure 16). For following runs, fractions were collected and assayed for IRGIP, assamples were below the absorbance detection limits of the HPLC system. Analysis of anatural porcine Gil? standard in this fashion yielded 3 IRGIP peaks (a, b, c), the second ofwhich (b) eluted in the same fraction as synthetic porcine GIP (Figure 17A). On aseparate elution, RIA analysis of collected fractions revealed 1 major IRGIP peak for theSTC 6-14 cell extract, but 3 major IRGIP peaks (i, iii) for the STC 6-14 culturemedium (Figure 17B). While the IRGIP in the cell extract was present in the samefractions as synthetic porcine GIP 1-42, the 3 IRGIP peaks from cell medium were indifferent fractions. The largest IRGIP fraction of the cell medium (peak i) eluted at thesame fraction as peak a of natural porcine G]P.33.2 IRGIP SECRETION INRESPONSE TO GLUCOSEThe effects of SS and SS monoclonal antibody (SOMA-lO) on glucose stimulatedIRGIP release are shown in Figure 18. After 4 days culture, each well of cells on averagecontained 33.3 ± 1.4 ng of IRGIP, and basal release in the presence of 5 mM glucose was733 ± 58 pgml-’2h’ (2.20 ± 0.17 %TCC). Incubation of the cells at glucoseconcentrations of 10 mM and higher resulted in significantly greater IRGIP release thanat 5 mM glucose, to a maximum as tested of 4.20 ± 0.42 %TCC at 20 mM. In thepresence of 10 nM SS however, glucose had no significant effect on IRGIP release. Theaddition of SOMA-lO to the graded glucose concentrations resulted in significantlygreater IRGIP responses at all but 20 mM glucose. At 5 mM glucose, IRGIP release wasincreased by the addition of SOMA-lO to a level comparable to that obtained for 15 mMglucose alone.3.3.3 IRSS SECRETION IN RESPONSE TO GLUCOSE AND GIPThe release of IRSS from STC 6-14 cells in response to 4 concentrations ofglucose is shown if Figure 19. The average content of IRSS in each well after 4 daysculture was 18.4 ± 1.5 ng. Basal release of IRSS was 2.05 ± 0.19 %TCC, and increasedsignificantly in response to 15 and 20 mM glucose. The addition of natural porcine GIPat concentrations of 1 nM and greater to release medium containing 5 mM glucose alsosignificantly increased IRSS release in a concentration dependent fashion (Figure 20). Atthe highest pGIP concentration tested (100 nM) IRSS release was doubled (4.10 ± 0.44%TCC).96• L— - •,.“97FIGURE 15: Cultured STC 6-14 cells immunostained for GIP (black), x 400.I1F. a•4 ,..— .-Ii)c(- •b•. —•..IJ98-5.6-5.7LI).nGIPsGIp-5.9FIGURE 16: HPLC elution profile of natural porcine GIP (5 .tg) and synthetic porcineGIP (3 tg) from a C18 column with an acetonitrile gradient of 28-33% over 10 mm, at aflow rate of 1 mi/mm.TIME (mm)2099AB1301201O0b0 10 20 30FRACTION40 50FIGURE 17: Immunoreactive GIP (IRGIP) content of HPLC fractions eluted from a C18column with an acetonitrile gradient of 28-33% over 10 mm. A) Overlay of elutionprofiles for natural porcine GIP and synthetic porcine GIP. B) Overlay of elution profilesfor STC 6-14 cell extract and culture medium.4032aNATURAL pGIP—0---— SYNTHETIC pGIPC40 50010 30FRACTIONISTC 6-14 MEDIA—0—— STC 6-14 EXTRACT11111c.)10076__543210FIGURE 18: Effect of 250 pgfm1 anti-somatostatin antibody (SOMA-lO) or 10 nMsomatostatin (SS) on the dose-response relationship between glucose concentrations (5 to20 mM) and IRGIP secretion from STC 6-14 cells. For glucose alone, n = 10; for glucose+ anti-SS, n = 7; for glucose + 10 nM SS, n = 4. In this and subsequent figures, IRGIPlevels are expressed as mean ± SEM percent total cell content (TCC). IRGIP responseswere compared to 5 mM glucose alone. * p< 0.05.GLUCOSE (mM)1014 **_3__Ci)Ci)1-0.— I — I — I -5 10 15 20GLUCOSE (mM)FIGURE 19: Concentration-response relationship between glucose concentrations (5 to20 mM) and immunoreactive somatostatin (IRSS) secretion from STC 6-14 cells. n = 9.IRSS responses were compared to 5 mM glucose. * p< 0.05.c-)Ci)102***S432100 0.1 1GIP (nM)FIGURE 20: Effect of graded concentrations of porcine GIP (0-100 nM) onimmunoreactive somatostatin (IRSS) release from STC 6-14 cells in the presence of 5mM glucose. n =7. IRSS responses were compared to 5 mM glucose in the absence ofGIP. * p< 0.05.10 1001033.4 DISCUSSIONCell lines producing gut peptides are able to proliferate in culture and can providelarge numbers of viable tumor cells free of contamination by other cell types. They aretherefore valuable models for studying the synthesis, storage and release of gut peptidesin definable environments. The STC-1 cell line was derived from an intestinal endocrinetumor that developed in mice carrying transgenes consisting of the rat insulin promoterlinked to the potent viral oncogene SV4O T antigen and to the polyoma virus small Tantigen (Rindi et a!., 1990). This cell line has been used previously to study the releaseof cholecystokinin (CCK) and secretin (Chang et al., 1992; Mangel et al., 1993). In thesestudies, it was concluded that the STC-1 cells may serve as a useful model system toinvestigate the secretory mechanisms for these peptides. In the present study, theseprimarily ecretin-secreting tumor ells were sub-cloned to enrich the population ofIRGIP secreting cells, resulting in the STC 6-14 clone.Immunohistochemical analysis of the STC 6-14 cells revealed that IRGIP andIRSS were the main endocrine cell types comprising the cell line, and that small numbersof glucagon, gastrin, secretin, glicentin, pancreatic polypeptide, and neurotensin werealso present. The fact that the STC cell lines secrete a number of peptides must beconsidered. Heterogeneity of hormone expression might be related to the knownplasticity of endocrine cells when transformed (Tischler, 1983). Rindi et a!. (1990) notedthat the spectrum of hormones expressed apparently increased with tumor progression,suggesting that the proliferating intestinal neuroendocrine cell is able to switch tomultiple alternative differentiated states. This quality has been ascribed to stem cells ofthe intestinal epithelium (Cheng and Leblond, 1974a; 1974b; Pictet et a!., 1976). Rindi etal. (1990) also noted it was surprising that in RIP1Tag2/RIP2PyST1 the regulatorysequences of rat insulin II gene induced transformation in the secretin cell, an epithelialendocrine cell anatomically and functionally quite distinct from the insulin producing B-cell. This might be explained by the fact that the pancreas and upper gut have a common104embryologic origin (Pictet and Rutter, 1972; Fallcmer et at., 1984). Whether more thanone peptide is secreted from one cell was not determined, but is the subject of furtherinvestigation.CCK and secretin released from the STC-1 cell line have been previously shownto have the same retention times on HPLC as the naturally occurring peptides (Chang eta!., 1992). HPLC was also employed in the current investigations to characterize GIPexpressed by the STC 6-14 cells. The gradient derived for the elution of GIP by HPLCwas capable of resolving three main components of natural porcine GIP, similar to thatwhich had been previously reported (Brown et al., 1981; Jömvall eta!., 1981; Schmidt eta!., 1987). The main component had a retention time comparable to that of syntheticporcine GIP, and was therefore assigned GIP 1-42. The other two peaks were identifiedby comparisons to previous HPLC profiles of natural porcine OW (Brown et a!., 1981;Jörnvall et al., 1981; Schmidt et a!., 1987). These reports identified the two precedingpeaks as CCK and GIP 3-42, respectively. As the content of GIP in the STC 6-14 culturemedium and extracts loaded on the HPLC (using the same gradient conditions) wasbelow the absorbance detection limit, eluted fractions of subsequent runs were collected,examined for GIP immunoreactivity, and compared to identically treated GIP standardsrun the same day. Three main IRGIP components were resolved from natural porcineGIP. The largest component (b) was found in the same fraction as synthetic porcine GIPand was assigned GIP 1-42. Peak a of the natural GIP preparation eluted prior to peak band was hypothesized to be GIP 3-42 based on the previous reported elution profiles ofthis OW preparation. The difference in elution profiles between the absorbance (Figure16) and the GIP immunoreactivity (Figure 17A) may have occurred as a result ofreplacing the C18 column between these runs. IRGIP in the eluted STC 6-14 extract wasfound in the same fractions as synthetic porcine GIP and was assigned GIP 1-42.However, IRGIP in eluted STC 6-14 culture medium was found in three fractions, thelargest of which (peak i) eluted in the same fractions as peak a of the natural GIP,105believed to be GIP 3-42. The other IRGIP peaks of the STC 6-14 culture medium couldrepresent putative forms of GIP that were previously suggested to exist (see section1.9.3), or could be other 1RGIP fragments. Thus while 1RGIP contained in the STC 6-14cells does appear to have identical HPLC characteristics to synthetic GIP, IRGIP purifiedfrom the cell culture medium appears to have been degraded, possibly largely to GIP 3-42. This phenomenon is further investigated in chapter 4 of this thesis.Glucose is thought to be one of the main physiological regulators of GIP secretionin vivo. Concentrations of glucose from 10 to 20 mM significantly increased IRGIPrelease from STC 6-14 cells, however, there was only an approximately 2-fold increase inIRGIP release at 20 mM when compared to basal. This increase in IRGIP secretion issignificantly smaller than that noted in the isolated canine and porcine endocrine cells(Figures 10 & 11). It is possible that these tumor cells have glucose regulatorymechanisms that differ from those of normal endocrine cells. Such a case has previouslybeen observed with other tumor-derived endocrine cell lines (Brant et at., 1992;Nagamatsu and Steiner, 1992; Visher et al., 1987). In view of the high proportion ofsomatostatin-secreting cells present in the STC 6-14 cell cultures, it is likely thatendogenous somatostatin contributes to the blunted IRGIP response observed followingaddition of glucose. Studies of glucose-stimulated IRGIP release in the presence of thesomatostatin monoclonal antibody tend to support this hypothesis. This antibodysignificantly increased glucose-stimulated IRGIP release at glucose concentrations lessthan 20 mM, presumably by immunoneutralizing IRSS. Furthermore, the addition ofexogenous SS was able to completely suppress glucose-stimulated IRGIP release, at allglucose concentrations tested. SS has been previously demonstrated to inhibit IRGIPrelease in man (Salera et a!., 1982), dog (Pederson et a!., 1975) and rat (Ho et al., 1987).The cellular mechanisms responsible for the inhibitory effect of SS on IRGIP secretionare presently not understood, probably because of the prior lack of suitable in vitromodels.106Concomitant determination of IRSS revealed that glucose concentrations of 15and 20 mM significantly increased IRSS release when compared to basal. Previous invivo studies have also demonstrated the release of gut IRSS in response to luminalglucose (Schusdziarra et at., 1978). It is not clear in the present case whether thissecretion is a direct result of glucose, or whether glucose-stimulated IRGIP is acting as asecretagogue of IRSS. In order to examine if IRGIP could stimulate release of IRSS fromthe STC 6-14 cells, pGIP was added to the basal release medium. At concentrations of 1nM and greater, pGIP significantly increased IRSS release. Verification thatendogenously released IRGIP stimulates IRSS release requires the development ofantibodies capable of neutralizing the action of IRGIP. GIP has previously beendemonstrated to stimulate secretion of IRSS from the perfused rat stomach (McIntosh etal., 1981).In conclusion, the tumor-derived STC 6-14 cell line represents a readily availablesource of large numbers of IRGIP and IRSS cells for intensive study of cellularmechanisms of hormone release. IRGIP expressed by these cells has the same HPLCretention time as porcine GIP 1-42. Glucose-stimulated IRGIP release is augmented bythe addition of a somatostatin antibody, possibly by immunoneutralization of endogenousIRSS. The addition of exogenous SS suppressed glucose-stimulated IRGIP release.IRSS release from STC 6-14 cells is enhanced by the addition of glucose, but it is unclearif this increased release is secondary to glucose-stimulated IRGIP release. The additionof exogenous pGIP stimulated IRSS release in a concentration dependent fashion. TheSTC 6-14 cell line may thus be a useful model to further study the interactions andrelease of GIP and SS.107CHAPTER 4METABOLISM OF GIP4.1 INTRODUCTIONStudies in chapter 3 indicated that while IRGIP extracted from STC 6-14 cellseluted with the same retention time on HPLC as native porcine GIP, GIP secreted fromthese cells had a significantly different HPLC profile. Furthermore, it appeared that themajor constituent of IRGIP in the STC 6-14 culture medium eluted identically with GIP3-42 that was a component of preparations of porcine GIP produced by the method ofBrown et at. (1970). This observation suggested that the conversion of GIP 1-42 to GIP3-42 might be occurring in the medium. As this culture medium contained a total of 15%serum, it was also hypothesized that the N-terminal Tyr-Ala of GIP could be the substratefor an enzyme present in serum. The physiological implication of this process wasevident from the observation that GIP 3-42 separated from the natural preparations byHPLC was biologically inactive (see section 1.11.1).Jömvall et a!. (1981) suggested that the GIP fragment which constituted >20% ofpurified GIP was formed by secondary processing or degradation, through susceptibilityto attack by aminopeptidase, elastase, dipeptidyl aminopeptidase or related enzymes inthe intestine. As GIP belongs to a superfamily of peptides sharing sequence homology(see section 1.2.4), it could be predicted that such an enzyme would act on otherhormones with similar N-terminal amino acid sequences. In 1986, Frohman et al.observed that growth hormone-releasing hormone (GRF 1-44) was cleaved to 3-44 by anenzyme in plasma. Furthermore, this enzyme product, like the GIP fragment, wasbiologically inactive. As a member of the glucagon superfamily, N-terminal amino acids(1 and 2) of GRF are identical to GIP (Tyr-Ala). In 1989, Frohman et at. identified theplasma enzyme responsible as dipeptidyl peptidase IV (DPP IV). They clearly108demonstrated that this enzyme was present in human plasma and quickly inactivated GRFin vivo.The objective of studies in this chapter was to determine if DPP IV was alsoresponsible for the putative production of GIP 3-42 found in the STC 6-14 growthmedium, and more importantly, whether this enzyme might play an important role in GIPmetabolism in vivo. Furthermore, in light of a potential role for OP in some diseasestates (see section 1.13), it was also of interest to determine if levels of DPP IV werealtered in the obese Zucker rat, and NTDDM subjects. It was hypothesized that elevationsof plasma DPP IV leading to rapid incretin degradation, might manifest as hyperglycemiadue to a reduction in the biological potency of the enteroinsular axis.4.2 MATERIALS AND METHODS4.2.1 INCUBATION OF GIP WITH SERUMIn order to test the hypothesis that the serum component of the STC 6-14 culturemedium was responsible for the degradation of GIP, synthetic porcine OW (5 j.tglml;Peninsula) was incubated at 37°C in 40 mM phosphate buffer (pH 7.0), with and withoutserum (12.5% HS, 2.5% FCS). At time intervals of 1, 10, 30 mm; and 1, 3, 6, and 24 h,100 jil aliquots were transferred to Eppendorf tubes, immediately flash frozen bysubmersion in ethanol and dry ice and then stored at -70°C. Sample volumes of 50 p1were then analyzed by reverse-phase HPLC as previously described in section 3.2.6.These studies indicated a time-dependent breakdown of GIP 1-42 to the product believedto be GIP 3-42, in the presence of serum. Similar experiments were performed with andwithout the addition of 0.1 mM diprotin A (Sigma). Diprotin A is a bacterial tripeptide(Ile-Pro-ile) that has been demonstrated to competitively inhibit DPP IV (Umezawa et al.,1984) and was shown previously by Frohman et al. (1989) to greatly reduce theconversion of GRF 1-44 to GRF 3-44.10942.2 DIPEPTIDYL PEPTIDASE IV-NEGAT1VE RATSRecently, studies have demonstrated that Fischer-344 rats from the JapaneseCharles River Inc. (DPP TV-negative) specifically lack DPP IV, whereas Fischer-344 ratsfrom sources in the United States (DPP TV-positive) possess normal DPP IV activity(Tiruppathi et a!., 1990; Watanabe et al., 1987). It was felt that these animals wouldserve as ideal models with the DPP TV-negatives serving as controls, to investigate apossible role for DPP IV in the metabolism of GIP in vivo. Two breeding pairs of theserats were generously supplied by Dr. F. H. Liebach (Medical College of Georgia,Augusta, Georgia) and a colony was established at the department of Physiology at UBC.42.3 DIPEPTIDYL PEPTIDASE IV ASSAYIn order to verify the lack of DPP TV in the DPP TV-negative rats, a DPP TV assaywas established, using a slightly modified method from that which was previouslydescribed (Matumura, 1985). Blood was collected from the tall of five DPP TV-negativeand five DPP TV-positive rats, and the serum separated. The substrate for the enzymewas a solution of 1.4 mM Gly-Pro-p-nitroanilide (Sigma) in 114 mM Tris buffer pH 8.0(700 mg Tris HC1 + 844 mg Tris Base to 100 ml distilled H20). Test tubes containing 1ml of the Tris buffer were incubated at 37°C, followed by the addition of 100 il testserum. The absorbance at 410 nm was then recorded immediately and at 5 mm intervals,using a U.V. spectrophotometer (Canlab Pye Unicam SP8-100) to monitor the appearanceof the yellow product of enzyme activity, p-nitroaniline. A standard curve was preparedby reading the absorbance of p-nitroaniline (Sigma) solutions of 0.005 to 1 mM in Trisbuffer (Figure 21A). Absorbance of serum samples with substrate (time = 0) wassubtracted from all further timed readings. The resulting absorbance values were thendetermined from the standard curve, and the resulting conversion rate to p-nitroanilinewas plotted.1104.2.4 HPLC PURIFICATION OF‘251-GIPIn order to perform in vivo studies to confirm the action of DPP IV on GIP, it wasdesirable to develop a highly sensitive assay system to monitor the conversion of GIP 1-42 to GJP 3-42. Since GIP 1-42 can be separated from GIP 3-42 by HPLC (see section3.3.1), it was hypothesized that HPLC could also resolve 125-GIP 1-42 from‘251-GIP 3-42. The use of radioactive tracer would enable the use of physiological levels of GIP invivo, and thus allow for a highly sensitive, accurate estimation of the conversion of GIP1-42 to inactive GIP 3-42. Tracer was therefore prepared, using the methods outlined insection 2.2.8, with the exception that synthetic porcine GIP (Peninsula) was used insteadof the natural purified form (known to contain a high proportion of GIP 3-42; see section1.11.1). Immediately after iodination, the peak fraction of 125-GIP was purified on anHPLC system that was used exclusively for purifying radioiodinated peptides.Water (distilled and treated through a Waters MilliQ H20 filtration system) andacetonitrile (CH3CN) were prepared by adding 0.1% thfluoroacetic acid (TFA) to each,and degassing and filtering through either 0.22 I.Lm (H20) or 0.45 jim (CH3CN) filters(Waters). These solvents were delivered to a j.tBondapak C18 column (Waters) using two1 lOB Solvent Delivery Module pumps (Beckman Instruments Inc, San Ramon, CA)controlled by a programmable 421A Controller (Beckman). The‘251-GIP was injectedusing a 100 jil needle syringe (Hamilton Co., Reno, NV) into a 210A Injector Port(Beckman) with 100 jil capacity. The tracer was loaded at a CH3N concentration of32%, and eluted by increasing the CH3N content to 38% over 10 mm, and maintainedfor a further 5 mm. The column was then washed by increasing the CH3Nconcentration to 60% over 5 mm, and re-equilibrated at 32% for 10 mm prior to injectionof another sample. This program had previously been empirically determined byVerchere (1991) to provide separation of the tracer components. Radioactivity of thecolumn eluant was measured with a 170 Radioisotope Detector (Beckman) and chartedon a Recordall Series 5000 recorder (Fisher). Eluant fractions were collected every 0.5111mm into 13 x 10 mm siliconized glass test tubes containing 10 p.1 BSA (5%; RIA Grade).Aliquots (10 p.1) of these fractions were counted to ensure recovery and separation, andpeak fractions were pooled, frozen at -70°C, lyophiized, and stored at -20°C.The specific activity of the‘251-GIP was estimated by two methods. First, thetotal radioactivity (cpm) eluted from the G- 15 column for the first peak (‘251-GIP) wasdetermined as the sum of the radioactivity in each fraction of the peak (= x p.Ci).Assuming that all 5 p.g GIP was iodinated, the specific activity (S.A.) was then estimatedas: S.A. (p.CiIp.g) = x p.Ci/5 jig. Specific activity of the purified tracer was estimatedfrom self-displacement curves, using increasing concentrations of‘251-GIP in a GIP RIA.A displacement curve was formed beginning with --5000 cpm, and then doublingamounts up to 160,000 cpm. A standard curve was also prepared similar to the normalGIP RIA (see section 2.2.8), with the exception that the label was added at the same timeas the standards. In this fashion, both displacement curves allowed equal time for labeledand unlabeled peptide to compete for binding to the antibody. The ratio of bound to totaltracer (%B) was then calculated for each concentration of ‘I-GIP and the results plottedas %B vs. total counts added, on the same axes as the standard curve. Assuming that theantiserum used bound‘251-GIP and unlabeled GIP equally well, both the GIP standardconcentration and‘251-GIT’ radioactivity were determined for several values of %B in themid-range of the curve and a plot of GIP mass (pg) vs radioactivity (cpm) was produced.The S.A. of the label was then estimated as the inverse slope of this plot.4.2.5 INCUBATION OF‘251-GIP WITH SERUMThe hypothesis that‘251-GIP could be metabolized by serum and the componentsresolved by HPLC was tested. Purified label (peak 2) was diluted to 500,000 cpm/50 p.1in 40 mM phosphate buffer (pH 7.0) and 50 p.1 aliquots were added to siliconized testtubes containing 400 p.1 of the same buffer. After gentle mixing, 50 p.1 of rat serum wasadded to each tube, and incubated at 37°C. Some samples were also incubated at 4°C to112examine the effect of temperature. Similar serum/label mixtures were prepared withbuffer containing either 0.1 mM diprotin A, 2% aprotinin, 50 U/mi bacitracin (Upjohn),or all three combined as a ‘cocktail’. After incubation times ranging from 10 mm to 48 h,100 aliquots were removed and analyzed by HPLC for breakdown of1I-GP 1-42.42.6 IN VIVO EXPERIMENTS WITH 125-GIPExperiments were next carried out to measure the degradation of‘251-GTP 1-42 invivo. Age matched Fischer 344 rats and DPP TV-negative rats (—350 g) were fastedovernight and anesthetized by intraperitoneal injection of sodium pentobarbital (60mg/kg). The right jugular vein was cannulated with PE 90 tubing filled with heparinizedsaline and attached to a 3-way stopcock. The abdomen was then exposed by a midlineincision and the abdominal aorta cannulated with heparinized saline-filled PE16O tubingattached to a 3-way stopcock. At time 0, —5 x 106 cpm purified GIP label (peak 2; —22.5ng GIP) in a volume of 100 I.Ll 40 mM phosphate buffer (pH 7.0) was injected into thejugular vein, followed by 200-300 j.ti of heparinized saline. At intervals of 2, 5, 10, and20 mm, 2 ml blood samples were collected from the dorsal aorta and transferredimmediately to Eppendorf tubes on ice containing diprotin A and aprotinin (finalconcentration 0.1 mM and 2%, respectively). Samples were then centrifuged at 7,000 x gfor 5 mm at 4°C and the serum removed and stored on ice.A C18 SepPak cartridge (Waters) was used to purify the‘251-GIP from the serum.The SepPak was first primed with 5 ml CH3N (plus 0.1% TFA), followed by 10 ml H20(plus 0.1% TFA). The serum sample was then applied to the cartridge, and washed with10 ml H20 plus TFA followed by 10 ml of 20% CH3N plus TFA. The label was theneluted into a siliconized test tube containing 10 BSA (5%; RIA Grade) with 2 ml 40%CH3N plus TFA, lyophilized and stored at -20°C. Each step of the procedure wasmonitored with a portable gamma detector to ensure maximal recovery of the tracer.113Lyophilized samples were then dissolved in 100 jil of distilled H20 and analyzed byHPLC as described in section 4.2.4.Control experiments were performed in order to ensure that during the samplecollection procedure and sample treatment steps, no further degradation of the labeloccurred. Approximately 6 x iO cpm was divided into two aliquots of 1 ml serumcontaining diprotin A (0.1 mM) and aprotinin (2%). One aliquot was maintained on icefor 1 h, while the other was incubated at 37°C for 1 h. Both samples were then purifiedusing the SepPak, and lyophilized as the rest of the samples. These controls were thenanalyzed by HPLC for any degradation of the ‘I-GIP.4.2.7 DIPEPTIDYL PEPTIDASE IV IN ZUCKER RATS AND NIDDM PATIENTSSerum samples were collected from fed Wistar rats (4) plus fed lean and fatZucker rats (4 of each) and stored at -20°C. In collaboration with Dr. G. S. Meneilly(Department of Medicine, UBC) samples were also collected from 12 fasted non-obeseNTDDM patients receiving treatment with oral hypoglycemic agents, and 12 fasted agematched controls with normal glucose tolerance, and stored at -20°C. All samples werethen assayed for DPP IV by the method described in section RESULTS4.3.1 INCUBATION OF GIP WITH SERUMIncubation of GIP with serum yielded a time-dependent degradation, resulting inproducts that eluted by HPLC identically to that observed for STC 6-14 cell culturemedium (Figure 17 chapter 3). This process did not occur in the absence of serum, andwas considerably reduced by the addition of 0.1 mM diprotin A (data not shown).1144.3.2 DIPEPTIDYL PEPTIDASEIVASSAYThe assay used to measure serum levels of DPP IV relied on the ability of thisenzyme to convert Gly-Pro-p-nitroanilide to the colored product, p-nitroaniline. Thedevelopment of the product was monitored with a spectrophotometer at 410 nm and wasconverted to mM p-nitroaniline produced using a standard curve (Figure 21A). Under theassay conditions used, serum from Fischer rats was able to produce p-nitroaniline at a rateof —1.66 I.LM/min, while production with serum from DPP TV-negative rats occurred at amuch slower rate of —0.17 p.M/mm (Figure 21B). There was no significant difference inDPP IV activity in the same serum volume from Fischer, Wistar, lean Zucker, or fatZucker rats (data not shown). Similarly, there was no significant difference in the DPPIV activity in serum from NIDDM patients tested when compared to controls.Absorbance values at 45 mm were corrected for the absorbance of the serum + substrateat time 0, and then converted to mM p-nitroaniline produced using the standard curve.The average concentration of p-nitroaniline after 45 mm for control subjects was 0.134 ±0.004 mM (—2.97 p.M/mm), and for diabetic subjects was 0.128 ± 0.004 mM (—2.84.tM/min).4.3.3 LABEL SPECIFIC ACTIVITYThe specific activity of the 125-GIP eluted from the G-15 column duringiodination (Figure 23) was estimated at —35 p.Ci/p.g. This material was then purified byreverse-phase HPLC, which revealed a heterogeneous population of iodinated peptides,consisting of 4 major components, the largest being peak 2 (Figure 25A). The specificactivity of this fraction was estimated from self displacement curves to be —100 p.Ci/p.g(Figure 24), representing a significant enrichment compared to the unpurified tracer.1154.3.4 INCUBATION OF PURIFIED 125I-GIP WITH WISTAR RAT SERUMThe stability of HPLC purified ‘I-GW 1-42 was investigated by incubation at37°C for up to 24 h in the absence of serum. At intervals of --6 h, labeled GIP wasremoved and analyzed by HPLC. Even after 24 h, a single component was observed(Figure 25B). This 125-GJP 1-42 eluted slightly later (—21-22 mm) then when collectedas peak 2 (—19 mm; Figure 25A), perhaps attributable to the use of a new C18 column. Asingle component was also observed after 10 mm incubation with 10% Wistar rat serum(Figure 26A), but after 1 h, 125-GIP had resolved into two distinct major components.One component eluted in the original position (—21-22 mm) and thus consisted of 1251GIP 1-42 (peak 1), while the new component eluted earlier (19-20 mm). This separationis similar to that observed in the identification of GIP 3-42 in natural GJP preparations(see Figure 16) and therefore this peak 2 was assigned‘251-GIP 3-42. With time, asaliquots of incubated‘251-GIP and serum were analyzed by HPLC, a progression from125-GIP 1-42 to putative‘251-GLP 3-42 was observed (Figures 27 & 28). Furthermore,other‘251-GIP fragments with different elution times appeared at a significantly slowerrate, the majority eluting at —3 mm (peak 3; Figures 27 & 28).The effect of temperature on the metabolism of‘251-GIP 1-42 was investigated byperforming the same experiments at 4°C. Both peaks 1 & 2 could be observed, indicatingthat the conversion of 125-GIP 1-42 to putative‘2I-GIP 3-42 was still occurring (Figures29 & 30). However, the rate of this reaction (as determined by the changes of the areas ofthe two peaks with time) was considerably slower. Interestingly, at this temperature, anabsence of other peaks including peak 3 was noted, indicating no further degradation ofthe tracer had occurred (Figures 29 & 30)The effect of enzyme inhibitors on the degradation of‘251-GIP was alsoinvestigated. Diprotin A, a competitive inhibitor of DPP IV, was able to significantlyreduce the degradation of‘251-GIP 1-42 to putativer2sIGIP 3-42. This was evident fromthe significantly reduced rate of appearance of peak 2 and reduction in peak 1 when 0.1116mM diprotin A was added to the serum and‘251-GTP (Figures 31 & 32). However, thisinhibitor did not alter the degradation of 125-Gll’ to other fragments, with considerableradioactivity eluting in peak 3 as incubation time progressed (Figures 31 & 32). Incontrast, the protease inhibitor aprotinin had no significant effect on the conversion of125IGIp 1-42 to putative‘251-GIP 3-42, as evident from the reduction of peak 1 andproduction of peak 2 with time by Wistar rat serum in the presence of 2% aprotinin(Figures 33 & 34). However, this inhibitor did reduce the rate of production of peak 3(Figures 33 & 34). The antibiotic bacitracin (50 U/mi) significantly reduced theconversion of‘251-GJP 1-42 to putative‘251-GJP 3-42 by serum (peak 1 to peak 2; Figure35), but not to the same degree as diprotin A. Bacitracin, however, greatly reduced thefurther fragmentation of‘251-GIP as indicated by the near absence of other peaks,including peak 3 (Figure 35). When used in combination, these inhibitors effectivelyreduced the degradation of‘251-GIP (Figure 36). After 20 h incubation of‘251-GIP withWistar rat serum and the inhibitor ‘cocktail’ at 37°C, —35% existed as putative 12I-GIP 3-42 (peak 2) and the remainder as ‘I-GIP 1-42 (peak 1).4.3.5 INCUBATION OF PURIFIED 125-GIP WiTH DPP [V-NEGATIVE RAT SERUMThe effect of DPP TV-negative serum on the degradation of‘25GW as analyzed byHPLC is shown in Figure 37. For a period of up to 5 h, no peak 2 could be resolved,suggesting that no‘251-GIP 3-42 was produced. However, as with Wistar rat serum, otherpeaks representing other1I-GIP fragments were observed, including peak 3 (Figure 37).The addition of 2% aprotinin significantly reduced the rate of production of other l2SjGIP fragments by serum from DPP P1-negative rats, as evident from the reduction in thesize of other peaks, including peak 3 (Figures 38 & 39).1174.3.6 INCUBATION OF PURIFIED 125I-GIP WiTH CEPThe buffer used for the GIP RIA contains 5% charcoal extracted human plasma(CEP; see section 2.2.8). In order to determine if this plasma contains enzymes capableof degrading GIP, CEP was diluted and incubated at 37°C with‘251-GIP and wasanalyzed by HPLC. After 3 h incubation, —48%‘251-GIP was found in peak 2, theputative‘251-GIP 3-42 (Figure 40). Furthermore, by 7 h, further significant degradationto other‘251-GIP fragments had occurred, including a significant proportion in peak 3(Figure 40). These observations would suggest that the addition of this plasma to the GIPassay buffer would promote the degradation of the GIP tracer, and possibly the standardsand samples.4.3 .7 THE ANALYSIS OF‘251-GIP DEGRADATION IN VIVOMethods were developed to investigate the degradation of GIP in vivo. It wasexpected that 125-GIP 1-42 would be rapidly degraded to‘251-GIP 3-42 in the circulation,and rapid delivery of the tracer and expeditious blood sampling were deemed necessary.Label was therefore delivered by injection through a jugular vein cannula and plasmasamples were collected directly from the aorta. It was also imperative that any furtherdegradation subsequent to sample collection be prevented. Based on in vitro results itwas determined that in order to keep degradation to a minimum, samples should becollected into tubes in ice, containing diprotin A and aprotinin (final concentrations 0.1mM and 2%, respectively). The collected samples were then purified by SepPak asdescribed earlier (section 4.2.6). Recovery of tracer from plasma collection tolyophilization was approximately 80%. In order to verify that there was minimal labeldegradation through all steps of the procedure, 125-GIP was added to serum containingthe inhibitors, and was treated identically to the experimental samples. However, prior tobeing purified by SepPak, one aliquot was stored on ice for 1 h, and another wasincubated at 37°C for 1 h. The results of the HPLC analysis are shown in Figure 41.118Neither condition resulted in measurable degradation of the label, indicating thatexperimental samples were treated appropriately.The degradation of label occurred very quickly in vivo. By 10 mm, only --9% ofthe‘251-GIP was represented by peak 1(125-GIP 1-42; Figure 42 & 43). The remainderof tracer appeared to be‘251-GIP 3-42 (peak 2) with no other degradation productsdiscernible. In contrast, there was no degradation of 125G1P infused into DPP IV-negative rats by 10 mm (Figure 44). In order to summarize some of the data on 125-GIPbreakdown in serum, the relative proportions of putative‘251-GIP 3-42 and 125-GIP 1-42were computed by measuring the area of peak 1 and peak 2. By repeating this for varioustime intervals, plots of the rate of conversion to putative 125-GIP 3-42 were produced(Figure 45). Figure 45A indicates that the time required to yield 50% putative‘251-GIP3-42 was —2.2 h in 10% normal Wistar rat serum. This time was significantly reduced bythe addition of bacitracin to --5.6 h, and to 16.1 h by storage at 4°C. The addition ofdiprotin A lengthened the time required to acquire 50% putative‘251-GIP 3-42 to —29 h.Similar computations were done with the in vivo data from one Wistar rat. Only —1.5 mmwas required for 50% of the‘251-GTP to be converted to putative‘251-GIP 3-42, a rateapproximately 88-times faster than was observed with the 10% rat serum in vitro. Nodegradation of 125-GIP 1-42 to putative‘251-GIP 3-42 was observed in DPP TV-negativerats, in vitro or in vivo.zC(I)zzCz119p-NITRO ANILINE (mM)5----0----A1. 21: Determination of relative DPP IV levels in Fischer (DPP IV-positive) andDPP TV-negative rats from rate of production of p-nitroaniline. A) Standard curve ofabsorbance for increasing concentrations of p-nitroaniline (mM); n = 10. B) Plot of pnitroaniline concentration vs. time (mm) determined from absorbance values converted tomM p-nitroaniline from curve A. Control = substrate in the absence of serum; n = 8.FISCHERDPP IV NEG.CONTROL20 30TIME (mm)1204-z-zC-0-— I—CONTROLS DIABETICSFIGURE 22: Determination of the relative levels of DPP IV in NDDM (diabetic) andnormal (control) subjects from the production of p-nitroaniline after 45 mm. Averageconcentrations of p-nitroaniline were 0.134 ± 0.004 mM (--2.97 p.M/mm) for controls (n= 12), and 0.128 ± 0.004 mM (—2.84 p.M/mm) for diabetics (n = 12). These values arenot significantly different.1FIGURE 23: Profile of synthetic porcine GIP iodination mixture eluted on Sephadex G15. The iodination mixture was applied to the column immediately following iodination,and fractions were collected every 1.5 mm.‘251-radioactivity was measured from 10 i.tlaliquots, to produce the profile. Free Na125 eluted after‘251-GIP. *denotes samplesused.121S432 Na125010 20FRACTION30122A50-40-30- o20-10. I I I0.( 0.1 1 10 100Mass GIP Standard (pg x 102) (....)125i Radioactivity (cpm x 1O) (-.-)B0.5-0.4-oZ0.3-0.2-0.1- 00- I I I I0 2 4 6 8 10125! Radioactivity (cpm x 10)FIGURE 24: A) % bound (%B) for a GIP standard curve using increasing GIPconcentrations (empty circles), and a self displacement curve using increasing amounts of‘251G1P (filled circles). B) Plot of mass of GIP standard vs.‘251-GIP radioactivity asdetermined from A for a number of points. The inverse of the slope of this line was usedto calculate the specific activity of HPLC purified ‘I-G1P (peak 2; —100 .tCi/tg).EV3530251234AB1510 1500 5TIME (mm)30 35EV5-4-3-2-1-0 I I I I I I I0 5 10 15 20 25 30 35TIME (mm)FIGURE 25: A) HPLC elution profile of‘251-GIP (—10 cpm) at 32% acetonitrile for 10mm, followed by a 10 mm gradient up to 38%. After a further 5 mm at 38%, the columnwas washed by a 5 mm gradient up to 60%, and then at 30 mm was returned to 32% over5 mm. Major peaks numbered 1-4 are discussed in the text. B) HPLC elution profile ofpeak 2(‘251-GIP in subsequent figures) after 24 h incubation at 37°C (—5 x i0 cpm).124A10-8- 10O1I52IO53IO3I5TIME (mm)B10-8--1 10I:IIITIME (mm)FIGURE 26: Both figures are HPLC elution profiles of‘251-GIP (—10 cpm) incubatedwith Wistar rat serum (10%) at 37°C. A) duration = 10 mm. The single peakcorresponds to‘251-GIP 1-42. B) duration= j. Peak 1 is‘251-GIP 1-42, peak 2corresponds to‘251-GIP 3-42. In this and subsequent figures, the acetonitrile gradient isidentical to that described for Figure 2.125A10-8-2..__10— I I I I I) 5 10 15 20 25 30 35TIME (mm)B10.8-E. 30JIITIME (mm)FIGURE 27: Both figures are HPLC elution profiles of‘251-GIP (—1O cpm) incubatedwith Wistar rat serum (10%) at 37°C. A) duration = 111. B) duration = 6h. In bothfigures, peak 1 is I-GIP 1-42, peak 2 corresponds to‘251-GIP 3-42, and peak 3 consistsof unidentified‘251-GIP fragments.126A10-8-6-S.?01’52’O2’53’O3’5TIME (mm)B10-8--5 4..Ca1’01’52’02’53’03’5TIME (mm)FIGURE 28: Both figures are HPLC elution profiles of 125-GIP (—10 cpm) incubatedwith Wistar rat serum (10%) at 37°C. A) duration = .11h. B) duration = 24h. In bothfigures, all peaks consist of unidentified 125-GIP fragments.127A10-8-6-1°O1’52’O2’53’O3’5TIME (mm)B10-8--TIME (mm)FIGURE 29: Both figures are HPLC elution profiles of‘251-GIP (—10 cpm) incubatedwith Wistar rat serum (10%) at 4°C. A) duration = 4h. B) duration= J1. In bothfigures, peak 1 is‘251-Gll? 1-42, and peak 2 corresponds to‘251-GIP 3-42.128A10-8--E40111111TIME (mm)B10-8-0I;IIITIME (mm)FIGURE 30: Both figures are HPLC elution profiles of‘251-GIP (—‘1O cpm) incubatedwith Wistar rat serum (10%) at 4°C. A) duration = 19h. B) duration = 24h. In bothfigures, peak 1 is‘251-GJP 1-42, and peak 2 corresponds to ‘I-GW 3-42.129A10.8-6-ITIME (mm)B10-8-6-ITIME (mm)FIGURE 31: Both figures are HPLC elution profiles of 125-GIP (-10 cpm) incubatedwith Wistar rat serum (10%) and diprotin A (0.1 mM) at 37°C. A) duration= 2Ji. B)duration = j1. In both figures, peak 1 is‘251-GIP 1-42, peak 2 corresponds to‘251-GIP3-42, and peak 3 consists of unidentified‘25I-GW fragments.130A10-8-6.E1’01’52’02’53’03’5TIME (mm)B10-8- 36-2- 2vJ0- I I I I I0 5 10 15 20 25 30 35TIME (mm)FIGURE 32: Both figures are HPLC elution profiles of‘251-GIP (—10 cpm) incubatedwith Wistar rat serum (10%) and diprotin A (0.1 mM) at 37°C. A) duration = i2Ji. B)duration = ..h. In both figures, peak 1 is‘251-GIP 1-42, peak 2 corresponds to‘251-GIP3-42, and peak 3 and others consist of unidentified‘251-GIP fragments.131A10-8-6-o.I’IIITIME (mm)B10-8-6-I 0E1I0 115 2’O 5 310 315TIME (mm)FIGURE 33: Both figures are HPLC elution profiles of‘251-GIP (‘—10 cpm) incubatedwith Wistar rat serum (10%) and aprotinin (2%) at 37°C. A) duration = iji. B) duration= i.h. In both figures, peak 1 is 125-GIP 1-42, peak 2 corresponds to‘251-GIP 3-42, andpeak 3 consists of unidentified 125-GIP fragments.132A10-8-1111110 ITIME (mm)B10-8-:_________________________________TIME (mm)FIGURE 34: Both figures are HPLC elution profiles of 125-GIP (—1O cpm) incubatedwith Wistar rat serum (10%) and aprotiniri (2%) at 37°C. A) duration = iLh. B) duration= 5j1. In both figures, peak 1 is ‘I-GW 1-42, peak 2 corresponds to 115-GIP 3-42, andpeak 3 and others Consist of unidentified 125-GIP fragments.133A10-8-6-E.ohol’52’02’53’o3’5TIME (mm)B10.8-6-4.:HOlI5I:233TIME (mm)FIGURE 35: Both figures are HPLC elution profiles of‘251-GIP (—1O cpm) incubatedwith Wistar rat serum (10%) and bacitracin (50 U/mi) at 37°C. A) duration = 2h. B)duration = 19.5 h. In both figures, peak 1 is‘251-GIP 1-42, peak 2 corresponds to 1251..GIP 3-42, and peak 3 consists of unidentified ‘I-GW fragments.134A10-8-16-TIME (mm)B10-8-6-ITIME (mm)FIGURE 36: Both figures are HPLC elution profiles of 125-GIP (-10 cpm) incubatedwith Wistar rat serum (10%) containing diprotin A (0.1 mM), aprotinin (2%) andbacitracin (50 U/mi), at 37°C. A) duration = 3.25 h. B) duration = 2Ji. In both figures,peak 1 is‘251-GIP 1-42, and peak 2 corresponds to‘251-GIP 3-42.135A10-8-4.2-30.0 5 10 15 20 25 30 35TIME (mm)B10-8-6-E,.110 1’S 210 2’S 310ITIME (mm)FIGURE 37: Both flgures are HPLC elution profiles of‘251-GIP (--10 cpm) incubatedwith DPP IV-negative rat serum (10%) at 37°C. A) duration 2..h. B) duration ih. Inboth figures, peak 1 is 125-GIP 1-42, and peak 3 consists of unidentified 125-GIPfragments.136A10-8-a6-E..2-0- I I 1 I0 5 10 15 20 25 30 35TIME (mm)B10-8-6-ITIME (mm)FIGURE 38: Both figures are HPLC elution profiles of‘251-GIP (—1O cpm) incubatedwith DPP TV-negative rat serum (10%) and aprotinin (2%) at 37°C. A) duration= 4..h.B) duration &h. In both figures, peak 1 is‘251-GIP 1-42, and peak 3 consists ofunidentified‘251-GIP fragments.137A10-8-6-01I0 115 210 215 3’O315TIME (mm)B10.8.6. 3E-Ca1101’52’02’53’03’5TIME (mm)FIGURE 39: Both figures are HPLC elution profiles of‘251-GTP (—10 cpm) incubatedwith DPP TV-negative rat serum (10%) and aprotinin (2%) at 37°C. A) duration = iih.B) duration = 27 h. In both figures, peak 1 is 125-GIP 1-42, and peak 3 and othersconsist of unidentified‘251-GIP fragments.138A10-8-6-E.‘‘‘‘‘I’TIME (mm)B10-8--4,.V0EiI0 1’S 2’: 25 310ITIME (mm)FIGURE 40: Both figures are HPLC elution profiles of‘251-GIP (--10 cpm) incubatedwith charcoal extracted human plasma (10%) used in the GIP RIA, at 37°C. A) duration= iii. B) duration = Lh. In both figures, peak 1 is‘251-GIP 1-42, peak 2 corresponds to‘251-GIP 3-42, and peak 3 consists of unidentified‘251-GJP fragments.;- 5.0.E2.5.EV7.5.139AB1I I I I5 10 15 20TIME (mm)I I I25 30 3510.0C7.5-5.0-2.5-0.0- I0 5 10 15TIMEFIGURE 41: Both figures are HPLC elution profiles of‘251-GIP (-40 cpm) incubatedfor 1 h with Wistar rat serum (neat) containing diprotin A (0.1 mM), aprotinin (2%),purified by SepPak, lyophilized, and reconstituted as described in section 4.2.6. A)incubation on ice. B) incubation at 37°C. In both figures, peak 1 is 125-GIP 1-42.20 25 30 35(mm)140A10-8-6-.O1’52:2’53’O3’5TIME (mm)B7.5.;-‘ 5.0-115 2’SI ITIME (mm)FIGURE 42: Both figures are HPLC elution profiles of‘251-GIP infused into a Wistarrat, collected in 2 ml serum with diprotin A (0.1 mM) and aprotinin (2%), purified bySepPak, lyophilized and reconstituted, as described in section 4.2.6. A) duration = 2 mm.B) duration = 5 mm. In both figures, —8 x iO cpm was loaded, peak 1 is 125-GIP 1-42and peak 2 corresponds to‘251-GIP 3-42.141A2.5-2.0-TIME (mm)B2.5.2.0.ITIME (mm)FIGURE 43: Both figures are HPLC elution profiles of‘251-GIP infused into a Wistar rat,collected in 2 ml serum with diprotin A (0.1 mM) and aprotinin (2%), purified by SepPak,lyophilized and reconstituted, as described in section 4.2.6. A) duration = 10 mm. B)duration = 20 mm. In both figures, ‘—4 x 10 cpm was loaded, peak 1 is“51-GIP 1-42, peak2 corresponds to‘251-GJP 3-42, and peak 3 consists of unidentified 125-GIP fragments.EE1421I I I25 30 35A10-8-6-4-2-0- I I I0 5 10 15 20TIME (mm)B2. 44: Both figures are HPLC elution profiles of‘251-GIP infused into a DPP IV-negative rat, collected in 2 ml serum with diprotin A (0.1 mM) and aprotinin (2%), purifiedby SepPak, lyophilized and reconstituted, as described in section 4.2.6. A) duration= 2mm, —10 cpm loaded. B) duration = 10 mm, --.5 x iO cpm loaded. In both figures, peak 1is 125I.G1P 1-42 and peak 3 consists of unidentified ‘I-GIP fragments.10 5 10 15 20 25 30TIME (mm)143A100- Normal Serumyt0 + Diprotin A-075 - ----0---- + Bacitracin--A @ 4°C150-__02 5 — /—.4/,___I_I,.0 5 10 15 20TIME (h)B100.7550.25-0. I I I Io 2 4 6 8 10TIME (mm)FIGURE 45: Graphs summarizing the conversion rate of‘251-GIP 1-42 to‘251-GIP 3-42both in vitro and in vivo calculated from peak areas of Figures 5 to 22, plotted as % GIP3-42 (% of GIP 1-42 + 3-42) vs. time. A) Relative %‘251-GIP 3-42 production in Wistarrat serum at 37°C (normal serum), + diprotin A (0.1 mM), + bacitracin (50 U/ml), and @4°C. B) Rate of formation of‘251-GIP 3-42 from 125-GIP 1-42 infused in a Wistar rat.1444.4 DISCUSSIONThe enzyme DPP IV preferentially cleaves peptides and proteins having either XPro, X-Hyp or X-Ala at their N-termini (Walter et a!., 1980), and to a lesser extent X-Ser,X-Thr, and X-Val (Martin et a!., 1993). DPP IV was first identified in rat liver (HopsuHavu and Glenner, 1966), and has since been found ubiquitously distributed throughoutmammalian tissues, with highest activity in the kidney and the intestinal brush-bordermembrane, as well as in insects, bacteria, and yeast (see Yaron and Naider, 1993 forreview). The primary structure of rat liver DPP IV was deduced from its cDNA, with acalculated mass for the 767-residue polypeptide of 88,107 daltons (Ogata eta!., 1989). Inkeeping with the widespread occurrence of DPP IV, evidence suggests that the enzyme isinvolved in a number of biochemical processes. This list includes; renal transport andintestinal digestion of proline-containing peptides, immunological activation ofimmunocompetent cells, and fibronectin-mediated adhesion (Yaron and Naider, 1993).DPP IV also appears to have a role in the inactivation of biologically relevant peptides(Ahmad et a!., 1992; Frohman et a!., 1989; Mentlein et a!., 1993), and it is theinvolvement of DPP IV in the metabolism of GIP that was the focus of research discussedin this chapter.The observation made in chapter 3 that IRGIP in STC 6-14 culture medium eluteddifferently than IRGIP extracted from the cells, suggested the possibility of GIPdegradation. These observations were reproduced by incubating GIP in serum, yielding atime-dependent breakdown of GIP 1-42. The retention time of the major metabolicproduct, eluting just prior to GIP 1-42 standards, suggested it was GIP 3-42. GIP 3-42has been sequenced, and shown to be a significant component of natural GIPpreparations, eluting just prior to GIP 1-42 on reverse-phase HPLC (JUrnval eta!., 1981,Schmidt et a!., 1987). Further evidence that this component of GIP degradation by serumwas GIP 3-42 was supplied by the observation that the inclusion of diprotin A, acompetitive inhibitor of DPP IV, significantly reduced the production of the peak145believed to be GIP 3-42. The enzyme DPP IV, had previously been shown to beresponsible for removal of the same two N-terminal residues from growth hormonereleasing factor (GRF) in human serum (Frohman et a!., 1986; 1989).Subsequent to these studies, a report confirming the degradation of GIP by DPP1V in serum was published (Mentlein et a!., 1993). These authors found that DPP IVpurified from human placenta hydrolyzed the N-terminal dipeptide from GIP, GRF,tGLP-I, and PHM. Km values of 4- 34 jiM and Vmax values of 0.6 - 3.8 j.tmolmin-’mg4protein were determined for the enzymatic degradation of the 3 peptides by the purifiedpeptidase. The same fragments were identified when GIP or tGLP-I were incubated withhuman serum, and analyzed by HPLC. Furthermore, as was observed in the presentstudy, this reaction could be inhibited by 0.1 mM diprotin A. Mentlein et al. (1993)concluded that DPP IV initiates the metabolism of GIP in serum.Confirmation of a role for DPP IV in the metabolism of GIP required in vivoevidence. Frohman et a!. (1986) proposed a role for DPP IV in GRF metabolism, bydemonstrating degradation of GRF 1-44 to GRF 3-44 in vivo. These observations wereaccomplished by injecting subjects with GRF and analyzing serum by HPLC. Therewere, however, several disadvantages to this system, one being the requirement ofinfusing supraphysiological doses of hormone. Subjects were injected with GRF at 1jig/kg, and even at this dose, plasma concentrations were well beneath the concentrationsrequired for UV detection after HPLC, thus necessitating the collection of eluted fractionsand analysis by rad.ioimmunoassay. Furthermore, the interpretation of these experimentswere compounded by the presence of endogenous hormone. A primary objective of thepresent investigations was therefore to develop a highly sensitive assay to detect theconversion of GIP 1-42 to GIP 3-42 in vivo, to support the in vitro results.The HPLC profile obtained with the‘251-GIP in the present study, is very similarto that previously reported by others (Maletti et a!., 1986; Verchere, 1991). Maletti et al.(1986) found that tracer from the two largest peaks was able to stimulate insulin release146from the isolated perfused rat pancreas similarly to natural GIP. These authors also hadpreliminary evidence from enzyme cleavage that both of these iodinated fractions wereiodinated only at the tyrosine in position 10, and not at the tyrosine in position 1.Verchere (1991) investigated the iodination state of the four major peaks of 125-GIPobtained after HPLC. Peaks 1-4 contained predominantly monoiodinated tyrosines(MIT), with peak 2 being the most pure, consisting of almost 100% MIT. Furthermore,in a binding assay using tumor derived B-cells (BTC3), peak 2 had the greatest specificbinding (Verchere, 1991). In combination, these investigations suggested that peak 2consisted of GIP exclusively monoiodinated at the tyrosine in position 10. This peak wastherefore deemed ideal for the present studies, as any iodination at tyrosine in position 1could potentially alter the ability of DPP IV to cleave GIP between residues 2 and 3.Furthermore, this peak appeared to retain biological activity (Maletti et al., 1986;Verchere, 1991), and made up the greatest proportion of the iodinated product, allowingfor maximum yields. All subsequent experiments were therefore performed with peak 2of the HPLC separated ‘I-GIP.In order to confirm that the purified 1251..Gu) would still serve as a substrate forDPP IV, and that the products of enzyme degradation could be readily identified andquantified, experiments involving the incubation of GJP with serum followed by HPLCwere repeated with the tracer. These in vitro experiments clearly demonstrated the timedependent metabolism of GIP into products that could be resolved. As had beenobserved with the metabolism of natural GIP in the present experiment and by Mentleinet a!. (1993), there appeared to be one primary product which eluted just prior to‘251-GIP1-42, that was considered ‘I-GIP 3-42. The time required for conversion of 50% of the‘I-GP to‘251-GJP 3-42 under the assay conditions was estimated at 2.2 h. Other peaksappeared indicating further degradation of the tracer, but these reactions occurred at amuch slower rate. Further evidence that the primary product was 125-GIP 3-42 wasprovided by the observation that the inclusion of diprotin A in the reaction significantly147reduced the rate of production of this peak (50% 125-GIP 3-42 @ —29 h). Furthermore,rats specifically lacking DPP IV, served as suitable controls, and serum from theseanimals was unable to metabolize‘251-GIP 1-42 to‘251-GIP 3-42.This highly sensitive model also allowed for further characterization of themetabolism of 125-GIP 1-42. As expected, the inhibitor aprotinin had no effect on theproduction of‘251-GLP 3-42. Aprotinin did however significantly reduce subsequentmetabolism to other unidentified products, likely by inhibiting the action of otherenzymes such as trypsin. Buckley and Lundquist (1992) reported that bacitracin was aneffective inhibitor of enzymatic cleavage of tGLP-I, and therefore this inhibitor was alsoinvestigated. Bacitracin significantly reduced the degradation of 125-GIP 1-42 to 1251..GIP 3-42, but not to the degree that was observed with diprotin A (50% 125-GIP 3-42 @—5.6 h). However, while diprotin A did not appear to significantly affect subsequentlabel degradation, bacitracin was highly effective. Even after 19.5 h incubation at 37°C,very small amounts of other products were identifiable (other than‘251-GIP 3-42). Incombination, diprotin A, aprotinin, and bacitracin provide highly effective protectionagainst‘251-GJP degradation.Interestingly, incubation with serum at 4°C resulted in significant degradation of‘251.GW to‘251-GIP 3-42, although the rate was much slower than the same reaction at37°C (50%‘251-GIP 3-42 @ —16.1 h). Other enzymes capable of metabolizing GIPappeared to be completely ineffective at this temperature, however, as indicated by thelack of other significant peaks (products) even after 24 h incubation. Thus in addition toincluding the enzyme inhibitor ‘cocktail’, these results would suggest that storage at 4°Cis an effective method to prevent degradation of GIP. While this study clearly supportsthe current practice of performing GIP assays at 4°C, the use of charcoal extracted plasma(CEP), a major assay constituent, requires further consideration. CEP was clearly able todegrade the tracer into multiple products, as indicated by the multiple peaks obtained byHPLC. This degradation of‘251-GIP, could very likely also extend to sample148degradation, which could affect the binding of the products by the antiserum. Thesestudies, therefore, clearly do not support the use of plasma in the GIP radioimmunoassay.This model thus appeared to fulfill the requirements of a suitable assay for themetabolism of GJP, and therefore in vivo experiments were performed. Approximately 5x 106 cpm‘251-GIP with a specific activity of —100 tCi/p.g (equivalent to 22.5 ng/GIP),was injected in each rat. Thus for a plasma volume of —10 ml, the GIP concentrationwould be —2 ng/ml, a value within the physiological range. This method thus allows theuse of concentrations at least 10-fold lower then those used by Frohman et a!. (1986) inthe investigation of in vivo metabolism of GRF. As expected, the rate for‘251-GIPdegradation was much more rapid in vivo than was observed in vitro with diluted serum.Only —1.5 mm was required for half of the‘251-GIP 1-42 to be converted to‘251-GIP 3-42. This represents an 88-fold reduction compared to the time observed in vitro forequivalent degradation. This result is comparable to that obtained by Frohman et al.(1986) who found that 1 mm after injection of 1 pjg/kg GRF 1-44, 43 ±7 % of the totalimmunoreactive GRH was GRH 3-44. This short half-life of DPP IV substrates is likelya reflection of the fact that DPP IV is found in high concentrations as an ectoenzyme ofthe plasma membrane of numerous cell types, in addition to plasma (Yaron and Naider,1993).There are many implications of these results that should be considered. The datapresented suggest that DPP IV is a primary enzyme involved in the degradation of GIP invivo. The product of this reaction is GIP 3-42, a polypeptide that has been previouslydemonstrated to lack somatostatinotropic and insulinotropic activity (Brown eta!., 1981;Moody et a!., 1981; Schmidt et al., 1986; 1987). There have been no assays reported todate using antisera that can distinguish between GIP 1-42 and GIP 3-42. Given thisinformation, the significance of sustained elevation of circulating GIP levels must be reevaluated. For example, reports of GIP half-lives of —20 mm (see section 1.12) andcirculating levels elevated for hours (see section 1.10.1) must be considered in light of the149time course of conversion of GIP 1-42 to GIP 3-42. Thus the insulinotropic activityappears regulated in such a fashion that secreted GIP only acts on the pancreas for onecirculation and is then inactivated to prevent overproduction of insulin andhypoglycemia.While GIl’ 3-42 is not an antagonist of the insulinotropic activity of GIl’ 1-42,other smaller fragments such as GIP 4-42 could act as antagonists, or may even beinsulinotropic (see section 1.11.2). It remains possible, therefore, that the furtherenzymatic degradation of GIP that was observed to occur, albeit at a comparatively slowrate, might yield GIP fragments smaller than 3-42 with biological activity, or the abilityto act as antagonists. It is also possible that the addition of 125J to the tyrosine at position10 of the GIP polypeptide to produce the tracer used for metabolism studies hinderedenzymatic cleavage near this residue. These queries can be addressed by the separation,identification, and determination of biological activities of further serum degradationproducts of GIP.Over activity of DPP IV could result in increased metabolism of the insulinotropichormones Gil’ and tGLP-I, and thus a reduced insulin response to the secretion of thesehormones. This might then result in insufficient insulin secretion in response to a glucoseload, yielding diabetic-type symptoms. Interestingly, urinary concentrations of DPP IVhave been shown to be elevated in NIDDM patients (Nagata et a!., 1988; Takasawa et a!.,1990; Nukada et a!., 1992). As this elevation preceded the onset of microalbuminuriathat is associated with nephropathy in diabetics, it was suggested that the alteration of theproximal tubules occurs prior to pathological changes of the glomeruli and brings aboutenzyme leakage from the brush border. If this enzyme dissociation is a response to theelevated glucose levels, then it is possible that maintained high glucose could also resultin DPP IV dissociation from other tissues, such as endothelium, and lead to elevatedlevels in the plasma. However, neither in the obese Zucker rat, nor in the 12 NIDDMsubjects tested, were DPP IV activity rates significantly different from controls. The150possibility of altered DPP IV activity in subjects with diabetes requires furtherinvestigation.The mechanisms that regulate serum levels of DPP IV are not currentlyunderstood. In light of the metabolic role of this enzyme for degradation of incretins, it ispossible that DPP IV, like incretins, is influenced by nutrient absorption. Suzuki et al.(1993) recently reported that rat intestinal DPP IV levels could be modified by diet. After7 days on a high proline (gelatin) diet, DPP IV activity and mRNA levels in brush-bordermembranes were three- to six-fold greater. No analysis of serum concentrations weremade in this study. Mentlein et al. (1993) measured serum DPP IV levels in subjects preand postprandial and found no significant difference with an n = 3. Clearly, regulation ofserum DPP IV levels requires further investigation.In conclusion, separation of the cleavage products of GIP in vitro and in vivo,combined with the influence of enzyme inhibitors and serum from DPP IV-negative ratsindicates that dipeptidyl peptidase IV is a principal enzyme resulting in the degradationand biological inactivation of GIP. Preliminary data indicate no significant difference inDPP IV activity for either obese Zucker rats or NIDDM subjects when compared tocontrols. The ubiquitous nature of this enzyme, including epithelial cells of the intestine,might explain the isolation of considerable quantities of GIP 3-42 from natural GIPpreparations. This and smaller enzyme fragments of OW undoubtedly contribute to theoverall immunoreactivity measured by GIP radioimmunoassays currently in use.Furthermore, the inclusion of plasma in assay buffer, results in significant degradation ofas well as perhaps GIP in standards and unknowns. This can be prevented byincluding diprotin A, aprotinin, and bacitracin in the assay buffer, in addition tomaintaining all samples at 4°C.151CHAPTER 5SUMMARY AND FUTURE DIRECTIONSThe role of gastric inhibitory polypeptide as a hormone has been well established.The polypeptide has been isolated, purified, and chemically identified, and its cells oforigin located. Secretagogues for GIP release in vivo have been identified. The genecoding GIP has been isolated, sequenced, and the regulation of its transcriptioninvestigated. Induction of endogenous hormone release is accompanied by insulinrelease, and GIP injection in amounts that result in physiologically attainable levelsmimic the biological actions. A GIP receptor has been cloned, sequenced, and located ontarget tissues, and the signal transduction system has been studied. Finally, putative rolesfor GIP in pathophysiological conditions, such as diabetes mellitus and obesity have beensuggested, although data are equivocal. What is least well understood about Gil’, are themechanisms controlling its release, and its duration of biological activity i.e. metabolism.The objectives of investigations reported in this thesis were, therefore, firstly to developmethods to investigate GIP release at the cellular level, and secondly, to develop asensitive assay to monitor the degradation of GIP in plasma.A major difficulty in studying the release of GIP at the cellular level is the diffusedistribution of the endocrine cells of origin in the upper small intestine. Methodologywas therefore developed involving preferential enrichment of intestinal endocrine cellpreparations by centrifugal elutriation and short-term tissue culture in order to enable thestudy of the local regulation of IRGIP secretion from isolated endocrine cells. Canineintestinal duodenal and jejunal epithelial cell preparations enriched for endocrine cellsreleased IRGIP in response to depolarization, as well as the principal GIP secretagogueidentified in studies in vivo, glucose. It was demonstrated indirectly that glucosestimulated IRGIP release was inhibited by somatostatin, which is also likely a mediatorof GIP release in vivo. Increasing intracellular levels of cAMP or Ca2 by152pharmacological activators resulted in significant secretion of IRGIP, indicating thatthese intracellular compounds are likely involved as messengers in the signal transductionsystem regulating GIP release. Finally, receptor-dependent activation of intracellularCa2+ messenger systems by the neuropeptide GRP yielded concentration-dependentincreases in IRGLP output. This provides one of possibly many examples of neuropeptidecontrol mechanisms for GIP release. While most of these observations were made withisolated canine epithelial cells, preliminary data suggests that porcine tissue is suitable forthe continuation of these studies.These studies represent the first data on GIP release at the cellular level. Other invitro models, such as mucosal sections or isolated bowel preparations, have for the mostpart proven unsuccessful for studying GIP release, due to the problems associated withischemia in intestinal tissues, and the diffuse distribution of the GIP cells. With theability to culture the isolated cells, comes the ability to readily control their environment -a key factor in the investigation of stimulus-secretion coupling. Understanding themechanism by which a compound acts on the GIP cell, whether it be luminal, endocrine,paracrine, or neuronal, is very difficult to establish from in vivo studies. The methodsused in these investigations resulting in cultured GIP cells, allowed for observations ofthe effects of luminal (glucose), endocrine/paracrine (somatostatin) and neuronal (GRP)stimuli on GIP release.There are many unanswered questions which remain concerning GIP release,which may be addressed using this model. For example, it might now be possible toinvestigate the influence of insulin on GIP release to resolve the potential role of insulinas a feedback-inhibitor of GIP secretion. The mechanisms by which nutrients, such asglucose and fat, act to cause GIP release may also be investigated. It is not known whatintracellular processes in the metabolism of these nuthents result in GIP expression.Techniques are available in this laboratory to directly examine Ca2 flux in individualcells in response to secretagogues. This could facilitate answering these remaining153questions, and corroborate data presented in this thesis suggesting that changes inintracellular Ca2 are important in the regulation of GIP release.It also remains to be determined whether nuthents or their digestion products actdirectly on the GIP cell, or indirectly on neighbouring cells to stimulate GIP release. Ifthe spatial organization of cells is important in allowing signal transduction from nutrientsensitive cells to GIP cells, then it can be speculated that the reaggregation of dispersedcells during the culture process might influence the GIP response to certain stimuli. Withthe identification of the glucose transporter responsible for the activation of OW release,and its localization on specific cell types, it will also be important to examine itsdistribution in the cultured cells. Specifically, it is of interest whether the cells attach andaggregate on the collagen coated wells in similar patterns that occur in vivo. To this end,a small chamber was constructed to enable culture of isolated cells on a microscope stagefrom which time-lapse photographs may be taken to observe the process of cellaggregation, growth of cellular projections, formation of cell to cell connections etc.which might facilitate these studies.One negative aspect of the static release studies described in this thesis is thatmetabolites, and other secretory products of cells in the cultures may influence the releaseof GJP. Evidence for this was supplied by the ability of somatostatin immunoneutralizingantibody to increase IRGIP release. It would therefore be desirable to examine therelease of GIP from cells in a perifusion apparatus, where fresh medium is continuallyflowing over the cells and thus removing any secreted compounds. Such a system wouldalso allow investigation of the dynamics of the GIP response to various secretagogues.For example, the IRGIP response may be biphasic, or exhibit desensitization to prolongedstimuli. Finally, a molecular approach would also be beneficial, allowing determinationof the action of secretagogues on GIP mRNA levels. It would then be possible to addresshow much of the observed release of GIP is from GIP stores, and how much is fromnewly synthesized hormone.154Tumor cell lines were also investigated as an alternate source of GIP cells forstudy. A cell line derived from intestinal tumors of transgenic mice (STC- 1) was sub-cloned to produce a stable cell line with approximately 30% IRGIP and 30% IRSS cells.This new clone (STC 6-14) was used to study the release of IRGIP in response to glucoseand interactions with somatostatin. HPLC of extracts of STC 6-14 cells indicated that thetumor cell derived IRGIP had the same retention time as natural porcine GIP 1-42.Release of IRGIP from STC 6-14 cells; increased in a concentration dependent fashion byglucose, was attenuated by the addition of somatostatin, and was augmented by additionof a somatostatin neutralizing antibody, presumably by immunoneutralizingendogenously released somatostatin. The addition of exogenous porcine GIP in thepresence of 5 mM glucose produced a concentration dependent increase in IRSS release.These represent the first reported data on GIP release from a tumor cell line. The STC 6-14 cells may be useful to further investigate the cellular mechanisms controlling therelease of GIP and somatostatin, in addition to their interactions.This cell line provides a readily available, inexpensive supply of GIP cells.Studies on the cellular regulation of GIP secretion should be done concomitantly withstudies on primary cells. Caution must be used when interpreting data from transformedcells, which might be regulated in a significantly different fashion than their naturalcounterpart. For example, in most malignant epithelial cells, there is an impairment ofthe metabolism of glucose with high rates of glucose consumption, aerobic glycolysis andan unusual accumulation of glycogen as compared with the normal tissue (see section 3.4and Zweibaum et a!., 1991 for review). Such cellular alterations might therefore result inaberrant hormone regulation and release mechanisms. A multidisciplinary approachincluding parallel studies with primary cells is therefore desirable.Most intestinal receptors were initially identified in isolated intestinal epithelialcell preparations from various species (Laburthe and Amiranoff, 1989). Cell lines offerseveral advantages over other tissue preparations for studying neurohormonal receptors155(Zweibaum et a!., 1991). Studies of the GIP receptor on pancreatic tumor derived 13-cellsare reviewed in section 1.11.2. In light of the fact that the GIP receptor on somatostatincells may be different then that on 13-cells (see section 1.11.1), the somatostatin secretingSTC 6-14 cell line may be a useful model to further investigate this hypothesis,particularly in view of the recent cloning of the GIP receptor (Usdin et a!., 1993).Previous studies have used somatostatin-secreting tumor cell lines to study the bindingcharacteristics of tGLP-I (Fehmann and Habener, 1991b; Gros et a!., 1993), but no suchreports on GIP binding have been forthcoming to date.All of the studies suggested for the primary GIP cells should also be performed onthe STC 6-14 cells. However, more importantly, further experiments should first beaimed at investigating characteristics of the tumor cells. In particular, it is important todetermine if the intracellular organization of these cells, as studied by electronmicroscopy, is characteristic of normal endocrine cells, e.g. secretory granule size andarrangement. Specifically, as this population of cells secretes a multitude of peptides (seesection 3.3.1), it is of interest to determine if there are cells expressing more than onehormone. These studies are currently in progress at the Department of Physiology, UBC.In addition, the STC 6-14 cell line provides a source of murine GIP for sequencing andassessment of the biological activity.Studies described in chapter 4 implicate DPP IV as the main degradation andinactivating enzyme for GIP in the circulation. This hypothesis is based on theobservations of the degradation of GIP 1-42 to putatitive GIP 3-42. Evidence that theDPP IV product is GIP 3-42 is supplied by its elution position just prior to GIP 1-42, andthe finding that this inactive peptide was absent or reduced by including diprotin A, aspecific inhibitor of DPP IV or by incubation with serum from rats DPP IV dificient.Furthermore, a recent report by Mentlein et a!. (1993) confirmed that GIP 1-42 isdegraded to GIP 3-42 in vitro by purified DPP IV or serum, by sequencing the reactionproducts. In order to confirm that GIP 3-42 is the main product of experiments156performed in this thesis, sequencing analysis should also be performed. In order to’ avoidsequencing radioactive peptide,‘271-GIP could be prepared and purified similar to thatwhich was performed for 125-GIP. After incubation with serum, the products could beseparated using the same HPLC protocols, collected and then sequenced.Studies performed in vivo suggest that 50% of a bolus injection of GIP can beinactivated to GIP 3-42 in as little as 1.5 mm. Rapid hormone inactivation in conjunctionwith the glucose-dependency of GIP, may act to ensure the prevention ofhyperinsulinemia and subsequent hypoglycemia. Since GIP 3-42 is biologically inactive,yet still immunoreactive, reports of circulating IRGIP levels in health and disease must bere-considered. Information regarding the circulating levels of biologically inactive GIPfragments can be misleading. Therefore, it is desirable to obtain N-terminal directedantibodies to establish an RIA, that can be used to distinguish between GIP 1-42 and GJP3-42. Such an assay would provide a clearer picture of circulating levels of biologicallyactive GIP secreted.As discussed in section 4.4, there is very little currently known about themechanisms that control the levels of the enzyme DPP IV. If diet can alter intestinalbrush border levels of this enzyme, this might lead to modulation of serumconcentrations. Alterations in the levels of DPP IV could potentially have profoundeffects on the biological actions of GIP and its other substrates eg. tGLP-I. Elevated DPPIV levels would be expected to reduce the half-life of biologically active GIP, and thusreduce the incretin effect, potentially leading to hyperglycemia. Initial studiesinvestigating the DPP IV activity in serum from obese Zucker rats and NIDDM subjectsrevealed no abnormalities. However, it does remain possible that other factors, such asdiet or medical treatment, might alter DPP IV activity. Furthermore, serum samples werecollected from NTDDM patients in the fasting state, during which GIP is normally neithersecreted nor insulinotropic. In order to test the hypothesis that DPP IV serum activity isaltered in NJDDM subjects, it is important, therefore, to also examine the enzyme activity157during the post-absorptive state, when GIP is insulinotropic. In addition, due to theubiquitous distribution of DPP IV, with many potential sites for the metabolism of GIPother than in serum, in vivo GIP degradation studies should also be performed withNIDDM subjects. As diabetes is a multifactorial disorder, with many potentialcontributing abnormalities, further investigation into a potential role for DPP IV iswarranted.The DPP TV-negative rats provide a model for the investigation of theconsequences of a deficiency of this enzyme. The half-life for biological activity of GIPin these animals is considerably longer then controls, limited only by degradation byother enzymes and extraction by organs such as the kidney. It might be conceived thatthe greater duration of insulinotropic activity of GIP (and tGLP-I) could potentially resultin over production of insulin and hypoglycemia. However, as glucose is the primarysource of cell metabolic energy, it might be expected that these animals adaptphysiologically to ensure an appropriate insulin response to ingestion of a meal.Preliminary investigations indicate that these animals do in fact have a normal insulinresponse to an oral glucose tolerance test, resulting in the appropriate maintenance ofblood glucose levels. One protective measure that could explain this observation is theglucose-dependent nature of GIP. As soon as elevated plasma glucose levels (followingan oral glucose challenge) return to basal (below threshold for GIP action), any non-metabolized GIP 1-42 will not be insulinotropic, thus preventing hypoglycemia.Alternatively, or additionally, desensitization of B-cells to GIP could result from elevatedlevels of GIP 1-42. In light of the fact that elevated GIP or tGLP-I levels have beendemonstrated to cause B-cell desensitization to these hormones (see section 1.11.2), it isof interest to determine if similar phenomena are also occurring in the DPP TV-negativerats. Preliminary studies suggest that the pancreas from these animals is in factdesensitized to GIP, as evident from a reduced insulin response from the perfused158pancreas to GIP in DPP TV-negative rats when compared to controls (Pederson andKieffer, unpublished observations).Desensitization of the DPP TV-negative rat B-cells to the incretins GIP and tGLP-Imay occur as a result of down-regulation of the receptors to these incretins. Thishypothesis could be tested by examining the binding of these peptides to B-cells fromDPP IV-negative and control rats. Methods for rat B-cell isolation and examination ofGIP binding have been previously described (Verchere et a!., 1991). It would also beinteresting to examine the effect of total parenteral nutrition (TPN) on the enteroinsularaxis in these rats. In previous studies with normal rats, Pederson et al. (1985) showedthat while the IRGIP response to oral glucose was normal following TPN, the isolatedperfused pancreas showed a 30% increase in the insulin release in response to GIP afterTPN. It was hypothesized that the increase in B-cell sensitivity to GIP may be causallyconnected to the exposure of the pancreas to chronically low levels of GIP during TPN.Perhaps this procedure could reverse the B-cell desensitization to GIP observed in theDPP IV-negative rats.Evidence provided in this thesis that implicates DPP IV as a principal inactivatingenzyme of GIP suggests a strategy for the development of GIP analogs with a longer half-life in the circulation. DPP IV has an absolute requirement for the L configuration of theamino acid residue, both in the penultimate and the N-terminal position (Yaron andNaider, 1993). Thus substitution of D-amino acids for the L-amino acids His-Ala at theN-terminus of tGLP-I has been reported to result in resistance to DPP IV cleavage(Buckley and Lundquist, 1992). It is, therefore, likely that substitution of the GIP N-terminal amino acids Tyr-Ala with corresponding D configurations might make theresulting analogue resistant to degradation by DPP IV, with an attendant increase ininsulinotropic potency. Such a compound could potentially have therapeutic actions insubjects with NIDDM who have insufficient insulin release to regulate glucosemetabolism. This concept is supported by recent observations that tGLP-I (which may be159more insulinotropic then native GIP; see section 1.5.4) might have antidiabetogeniceffects in subjects with NIDDM (Gutniak et al., 1992; Holz et al., 1993; Nathan et al.,1992; Nauk et a!., 1993).In conclusion, GIP is a hormone with important anabolic functions, of which themost extensively studied is undoubtedly its insulinotropic action. The role of GIP as anincretin in the enteroinsular axis has been well established, and potential links withobesity and NTDDM have been made. The experiments outlined in this thesis have beendirected towards gaining a better understanding of the cellular mechanisms controllingGIP release and biological inactivation of this hormone in vivo. Data have beenpresented which implicate both the cAMP and Ca2 intracellular messenger systems inthe release of GIP, and the enzyme dipeptidyl peptidase IV in the rapid inactivation ofGIP in vivo. The models developed to examine the release and metabolism of GIP thatare presented in this thesis should enable further future investigation of these aspects ofGIP physiology in both normal and pathophysiological states.160REFERENCESAhmad S, Wang L, Ward PE (1992) Dipeptidyl(amino)peptidase TV and aminopeptidaseM metabolize circulating substance P in vivo. J. Pharmacol. Exp. Therap. 260:1257-1261.Ahrén B, Häkanson R, Lundquist I, Sjolund K, Sundler F (1981) GIP-likeimmunoreactivity in glucagon cells. Interactions between GIP and glucagon oninsulin release. Acta Physiol. Scand. 112:233-242.Alam MJ, Buchanan KD (1980) Gastric inhibitory polypeptide (GIP) responses inprimary diabetes. [Suppi. 1] Regul. Pept. p 52 (Abstract).Alam MJ, Buchanan KD (1993) Conflicting gastric inhibitory polypeptide data: possiblecauses. Diabetes Res. Clin. Prac. 19:93-101.Alam MJ, Kerr JI, Cormican K, Buchanan KD (1992) Gastric inhibitory polypeptide(GIP) response in diabetes using a highly specific antiserum. Diabetic. Med. 9: 542-545.Aluments JR, Häkanson R, O’Dorisio T, Sjolund K, Sundler F (1978) Is GIP a glucagoncell constituent? Histochem. 58:253-256.Amaland PF, Jorde R, Revhaug A, Myhre ESP, Burhol P0, Giercksky KE (1984) Fastingand postprandial GIP values in pigs, rats, dogs, and man measured with five differentGIP antisera. Scand. J. Gastroenterol. 19:1095-1098.Amiranoff B, Couvineau A, Vauclin-Jacques N, Laburthe M (1986) Gastric inhibitorypolypeptide receptor in hamster pancreatic beta cells: direct cross-linking,solubilization and characterization as a glycoprotein. Eur. J. Biochem. 159:353-358.Amiranoff B, Vauclin-Jacques N, Laburthe M (1984) Functional GIP receptors in ahamster pancreatic beta-cell line ml 11: specific binding and biological effects.Biochem. Biophys. Res. Commun. 123:671-676.Amiranoff B, Vauclin-Jacques N, Laburthe M (1985) Interaction of gastric inhibitorypolypeptide (GIP) with the insulin-secreting pancreatic beta-cell line In 111:characteristics of GIP binding sites. Life Sci. 36:807-8 13.Andersen DK, Elahi D, Brown JC, Tobin JD, Andrs R (1978) Oral glucoseaugmentation of insulin secretion: interactions of gastric inhibitory polypeptide withambient glucose and insulin levels. J. Clin. Invest. 62:152-161.Andersen DK, Putnam WS, Hanks JB, Wise JE, Lebovitz HE, Jones RS (1980)Enhancement of the metabolic clearance of gastric inhibitory polypeptide (GIP) byhyperglycemia. Regul. Pept. [Suppl.J 1:S3.Aponte GW, Taylor IL, Soil AH (1988) Primary culture of PYY cells from canine colon.Am. J. Physiol. 254:G829-G836.Arimura A (1992) Receptors for pituitary adenylate cyclase-activating polypeptide:comparisons with vasoactive intestinal peptide receptors. Trends Endocrinol. Metab.3:288-294.161Arnold R, Ebert R, Creutzfeldt W, Becker HD, Borger H (1978) Inhibition of gastric acidsecretion by gastric inhibitory polypeptide (GIP) in man. Scand. 3. Gastroenterol.[Suppi. 49] 13:11.Bacarese-Hamilton AJ, Adrian TE, Bloom SR (1984) Human and porcineimmunoreactive gastric inhibitory polypeptides (IR-GIP) are not identical. FEBS Lett.168:125-128.Banting FG, Best CH, Collip JB, Campbell WR, Fletcher AA (1922) Pancreatic extractsin the treatment of diabetes mellitus. Can. Med. Assoc. J. 12:141.Barber DL, Buchan AMJ, Walsh JH, Soil AH (1986a) Isolated canine ileal mucosal cellsin short-term culture: a model for the study of neurotensin release. Am. J. Physiol.250:G374-G384.Barber DL, Walsh IH, Soll AH (1986b) Release and characterization of cholecystokininfrom isolated canine jejunal cells. Gastroenterol. 9 1:627-636.Baumert Jr JE, Cataland S, Tetirick Jr CE, Pace G, Mazzaferri EL (1978) Effect ofatropine on meal-stimulated gastrin and gastric inhibitory polypeptide (GIP) release.3. Clin. Endocrinol. Metab. 46:473-476.Bayliss WM, Starling EH (1902) The mechanism of pancreatic secretion. 3. Physiol.(London) 28:235-334.Beck B (1989) Gastric inhibitory polypeptide: a gut hormone with anabolic functions. 3.Mol. Endocrinol. 2:169-174.Beck B, Max JP (1983) Gastric inhibitory polypeptide enhancement of the insulin effecton fatty acid incorporation into adipose tissue in the rat. Regul. Pept. 7:3-8.Beck B, Villaume C, Chayviaile JA, Gariot P. Ulmer M, Desalme A, Debry G (1984)Influence of caloric intake on gastric inhibitory polypeptide, VIP and gastrin releasein man. Peptides 5:403-406.Bell GI (1986) The glucagon superfamily: precursor structure and gene organization.Peptides 7:27-36.Bell GI, Santerre RF, Mullenbach GT (1983) Hamster preproglucagon contains thesequence of glucagon and two related peptides. Nature 302:716-718.Berson SA, Yallow RS (1959) Immunoassay of endogenous plasma insulin in humansubjects by immunologic methods. Nature 184:1648-1649.Berthoud HR, Trimble ER, Moody AJ (1982) Lack of gastric inhibitory polypeptide(GIP) response to vagal stimulation in the rat. Peptides 3:907-912.Blundell TL, Dockerill 5, Sasaki K, Tickle U, Wood SP (1976) The relation of structureto storage and receptor binding of glucagon. Metab. [Suppl. 1] 25:1331-1336.Bodanszky M, Bodanszky A (1986) Conformation of peptides of the secretin-VIPglucagon family in solution. Peptides [Suppl. 1] 7:43-48.162Brant AM, MeCoid S, Thomas HM, Baldwin SA, Davies A, Parker JC, Gibbs EM, GouldGW (1992) Analysis of the glucose transporter content of islet cell lines: implicationsfor glucose-stimulated insulin release. Cellular. Sig. 4:641-650.Brown JC (1971) A gastric inhibitory polypeptide. I. The amino acid composition and thetryptic peptides. Can. J. Biochem. 49:255-261.Brown JC (1974) Gastric inhibitory polypeptide (GIP). In: Endocrinology 1973. S.Taylor, ed., Heinemann, London, pp. 276-284.Brown JC (1982) Gastric inhibitory polypeptide. In: Monographs on Endocrinology.Springer-Verlag, Berlin, Germany.Brown JC (1988) Enteroinsular axis. In: Baillières Clinical Endocrinology andMetabolism 2(2) :359-373.Brown JC, Buchan AMJ, McIntosh CHS, Pederson RA (1989) Gastric inhibitorypolypeptide. In: Handbook of physiology, the gastrointestinal system. Eds. S.G.Schultz, G.M. Makhlouf, B.B. Rauner. Am. Phsyiol. Soc. Bethesda MD, pp. 403-430.Brown JC, Dahl M, Kwauk 5, McIntosh CHS, Muller M, Otte SC, Pederson RA (1981)Actions of GTP. Peptides 2 (suppi. 2) 24 1-245.Brown JC, Dryburgh JR (1971) A gasthc inhibitory polypeptide II. The complete aminoacid sequence. Can. J. Biochem. 49:867-872.Brown JC, Dryburgh JR, Ross SA, Dupré J (1975) Identification and actions of gastricinhibitory polypeptide. Recent Prog. Horm. Res. 31:487-532.Brown JC, Mutt V. Pederson RA (1970) Further purification of a polypeptidedemonstrating enterogastrone activity. J. Physiol. 209:57-64.Brown JC, Otte SC (1978) Gastrointestinal hormones and control of insulin secretion.Diabetes 27:782-789.Brown JC, Pederson, RA (1970a) A multiparameter study on the action of preparationscontaining choleocystokinin-pancreozymin. Scand. J. Gastroenterol. 5:537-541.Brown JC, Pederson, RA (1970b) Cleavage of a gastric inhibitory polypeptide withcyanogen bromide and the physiological action of the C-terminal fragment. Proc.Physiol. Soc. June 52P (Abstract).Brown JC, Pederson RA (1976a) GI hormones and insulin secretion. Endocrinology.Proceedings of the Vth International Congress of Endocrinology 2:568-570.Brown JC, Pederson RA (1976b) The insulinotropic action of gastric inhibitorypolypeptide in the perfused isolated rat pancreas. Endocrinology 99:780-785.Brown JC, Pederson, RA, Jorpes E, Mutt V (1969) Preparation of highly activeenterogastrone. Can. 3. Physiol. Pharmacol. 47:133-114.Brubaker PL (1991) Regulation of intestinal proglucagon-derived peptide secretion byintestinal regulatory peptides. Endocrinology 128:3175-3182.163Brubaker PL, Drucker DJ, Greenberg GR (1990) Synthesis and secretion of somatostatin28 and -14 by fetal rat intestinal cells in culture. Am. J. Physiol. 258:G974-G981.Bryer-Ash M, Cheung A, Pederson RA (1994) Feedback regulation of glucose-depedentinsulinotropic polypeptide (GIP) secretion by insulin in conscious rats. Regul. Pept.(in press).Buchan AMJ, Ingman-Baker J, Levy 3, Brown JC (1982) A comparison of the ability ofserum and monoclonal antibodies to gastric inhibitory polypeptide to detectimmunoreactive cells in the gastroenteropancreatic system of mammals and reptiles.Histochem. 76:341-349.Buchan AMJ, Barber DL, Gregor M, Soll AH (1987) Morphologic and physiologicstudies of canine ileal enteroglucagon-containing cells in short-term culture.Gastroenterol. 93:791-800.Buchan AMJ, Curtis SB, Meloche RM (1990) Release of somatostatin immunoreactivityfrom human antral D cells in culture. Gastroenterol. 99:690-696.Buchan AMJ, Polak JM, Capella C, Solcia E, Pearse AGE (1978) Electronimmunocytochemical evidence for the K cell localisation of gastric inhibitorypolypeptide (GIP) in man. Histochem. 56:37-44.Buckley DI, Lundquist P (1992) Analysis of the degradation of insulinotropin [GLP-1(7-37)j in human plasma and production of degradation resistant analogs. Regul. Pept.40:117 (Abstract).Buffa B, Polak TM, Pearse AGE, Solcia E, Grimelius L, Capella C (1975) Identificationof the intestinal cell storing gastric inhibitory polypeptide. Histochem. 43:249-255.Bunnett NW, Harrison FA (1986) Immunocytochemical localization of gastric inhibitorypolypeptide and glucagon in the alimentary tract of ruminants. Quart. 3. Exp. Physiol.7 1:433-441.Burant CF, Takeda 3, Brot-Laroche E, Bell GI, Davidson NO (1992) Fructose transporterin human spermatozoa and small intestine is GLUT5. J. Biol. Chem. 267:14523-14526.Burhol PG, Jorde R, Waldun HL (1980a) Radio-immunoassay of plasma gastricinhibitory polypeptide (GIP), release of GIP after a test meal and duodenal infusion ofbile and immunoreactive plasma GIP components in man. Digest. 20:336-345.Burhol PG, Lygren I, Waldum HL, Jorde R (1980b) The effect of duodenal infusion ofbile on plasma VIP, GIP, and secretin and on duodenal bicarbonate secretion. Scand.J. Gastroenterol. 15:1007-1011.Campbell RM, Scanes CG (1992) Evolution of the growth hormone-releasing factor(GRF) family of peptides. Growth Reg. 2:175-191.Campos RV, Buchan AMJ, Meloche RM, Pederson RA, Kwok YN, Coy DH (1990)Gastrin secretion from human antral G cells in culture. Gastroenterol. 99:36-44.164Cariquist M (1987) Solid phase synthesis of a 31-residue fragment of human glucose-dependent insulinotropic polypeptide (GIP) by the continuous flow polyamidemethod. Acta Chem. Scand. 41:494-498.Cariquist M, Maletti M, Jörnvall H, Mutt V (1984) A novel form of gastric inhibitorypolypeptide (GIP) isolated from bovine intestine using a radioreceptor assay.Fragmentation with staphylococcal protease results in GIP 1-3 and GIP 4-42,fragmentation with enterokinase in GIP 1-16 and GJP 17-42. Eur. J. Biochem.145:573-577.Cataland 5, Crockett SE, Brown JC, Mazzaferri EL (1974) Gastric inhibitory polypeptidestimulation by oral glucose in man. I. Clin. Endocrinol. Metab. 39:223-228.Chamberlain MJ, Stimmler L (1967) The renal handling of insulin. J. Clin. Invest.49:791-797.Champion MC. Sullivan SN, Bloom SR, Adrian TE, Christofides ND (1982) The effectsof naloxone and morphine on postprandial gastrointestinal hormone secretion. Am. 3.Gastroenterol. 77:617-620.Chan CB, Pederson RA, Buchan AMJ, Tubesing KB, Brown JC (1984) Gastric inhibitorypolypeptide (GIP) and insulin release in the obese zucker rat. Diabetes 33:536-542.Chan CB, Soil AH (1988) Pertussis toxin-sensitive cholinergic inhibition of somatostatinrelease form canine D-cells. Am. J. Physiol. 255:G424-G428.Chang CL, Chang TM, Chey WY (1992) Characterization of the release ofcholecystokinin and secretin from a murine intestinal cell line. Regul. Peptides 40:122(Abstract).Chap Z, O’Dorisio TM, Cataland 5, Field JB (1987) Absence of hepatic extraction ofgastric inhibitory polypeptide in conscious dogs. Dig. Dis. Sci. 32:280-284.Cheng H, Leblond CP (1974a) Origin, differentiation and renewal of the four epithelialcell types in the mouse small intestine: III enteroendocrine-cells. Am. J. Anat.141:503-520.Cheng H, Leblond CP (1974b) Origin, differentiation and renewal of the four epithelialcell types in the mouse small intestine: V unitarian theory of the origin of the fourepithelial cell types. Am. 3. Anat. 141:537-562.Chow BC, Morrow GW, Ho M, Pederson RA, McIntosh CHS, Brown JC, MacGillivrayRTA (1990) Expression of recombinant human glucose-dependent insulinotropicpolypeptide in Esherichia coli by sequence-specific proteolysis of a protein A fusionprotein. Peptides 11:1069-1074.Christophe 3, Svoboda M, Lambert M, Waelbroeck M, Winand 3, Dehaye JP,Vandermeers-Piret MC, Vandermeers A, Robberecht P (1986) Effector mechanismsof peptides of the VIP family. Pept. [Suppl. 1] 7:101-107.Cleator IG, Gourlay RH (1975) Release of immunoreactive gastric inhibitory polypeptide(IR-GIP) by oral ingestion of food substances. Am. J. Surg. 130: 128-135.165Collier G, McLean A, ODea K (1984) The effect of co-ingestion of fat on the metabolicresponse to slowly and rapidly absorbed carbohydrates. Diabetologia 26:50-54.Comb M, Birnberg NC, Seasholtz HE, Goodman HM (1986) A cyclic AMP- and phorbolester-inducible DNA element. Nature 323:353-356.Couvineau AB, Amiranoff N, Vauclin-Jacques, Laburthe M (1984) The GIP receptor onpancreatic beta cell tumor: molecular identification by covalent cross-linking.Biochem. Biophys. Res. Commun. 122:283-288.Cox JS, ODorisio TM, Cataland S, Crockett SE (1981) Gastric inhibitory polypeptidehypersecretion in diabetes mellitus: effect of sulfonylurea treatment. J. Clin.Endocrinol. Metab. 52: 1002-1005.Creutzfeldt W (1979) The incretin concept today. Diabetologia 16:75-85.Creutzfeldt W, Ebert R (1977) Release of gastric inhibitory polypeptide (GIl?) to a testmeal under normal and pathophysiological conditions in man. In: Bajaj JS (ed)Diabetes. Excerpta Medica, Amsterdam, pp 63-75.Creutzfeldt W, Ebert R (1988) The incretin concept. In: Advances in Metabolic DisordersAcademic Press 11:333-367.Creutzfeldt W, Ebert R, Arnold R, Frerichs H, Brown JC (1976) Gastric inhibitorypolypeptide (GIP), gastrin and insulin: response to a test meal in coeliac disease andafter duodeno-pancreatectomy. Diabetologia 12:279-286.Creutzfeldt W, Ebert R, Willms B, Frerichs H, Brown JC (1978) Gastric inhibitorypolypeptide (GIP) and insulin in obesity: increased response to stimulation anddefective feedback control serum levels. Diabetologia 14:15-24.Creutzfeldt W, Talaulicar M, Ebert R, Willms B (1980) Inhibition of gastric inhibitorypolypeptide (GIP) release by insulin and glucose in juvenille diabetes. Diabetes29:140-145.Crockett SE, Cataland S, Falko JM, Mazzaferri EL (1976) The insulinotropic effect ofendogenous gastric inhibitory polypeptide in normal subjects. J. Clin. Endoncrinol.Metab. 42:1098-1103.D’Alessio DA, Fujimoto WY, Ensinck JW (1989) Effects of glucagon-like peptide 1 (7-36) on release of insulin, glucagon, and somatostatin by rat pancreatic islet cellmonolayer cultures. Diabetes 38:1534-1538.Davidson WD, Springberg PD, Falkinburg NR (1973) Renal extraction and excretion ofendogenous gastrin in the dog. Gastroenterol. 64:955-961.Demol P, Wingender W (1989) Effects of rioprostil on the postprandial glucose, GIP,insulin and gastrin levels in volunteers. Scand. J. Gastroenterol. [Suppl. 164] 24:112-117.Deschodt-Lanckman M, Robberecht P. Camus J, Christophe J (1971) Short-termadaptation of pancreatic hydrolases to nutritional and physiological stimuli in adultrats. Biochimie 53:789-796.166Drucker DJ, Philippe J, Mojsov S, Chick WL, Habener iF (1987) Glucagon-like peptide Istimulates insulin gene expression and increases cyclic AMP levels in a rat islet cellline. Proc. Natl. Acad. Sci. USA, 84: 3434-3438.Dryburgh JR (1977) Immunological techniques in the investigation of the physiologicalfunctions of gastric inhibitory polypeptide and motiin. Vancouver, Canada: Univ. ofB.C. PhD dissertation.Dryburgh JR. Hampton SM, Marks V (1980a) Endocrine pancreatic control of the releaseof gastric inhibitory polypeptide: a possible physiological role for C-peptide.Diabetologia 19:397-401.Dryburgh IR, Hampton SM, Morgan LM, Marks V (1980b) C-peptide-induced inhibitionof GIP release. Regul. Pept. [Suppi.] 1:28.Dupré J, Greenidge N, McDonald TJ, Ross SA, Rubinstein D (1976) Inhibition of actionof glucagon in adipocytes by gastric inhibitory polypeptide. Metabol. 25:1197-1199.Dupré J, Ross SA, Watson D, Brown JC (1973) Stimulation of insulin secretion bygastric inhibitory polypeptide in man. 3. Clin. Endocrinol. Metab. 37:826-828.Dupré J, Beck JC (1966) Stimulation of release of insulin by an extract of intestinalmucosa. Diabetes 15:555-559.Ebert R, Creutzfeldt W (1982) Influence of gastric inhibitory polypeptide antiserum onglucose-induced insulin secretion in rats. Endocrinology 111:1601-1606.Ebert R, Frerichs H, Creutzfeldt W (1979a) Impaired feedback control of fat inducedgastric inhibitory polypeptide secretion by insulin in obesity and glucose intolerance.Eur. 3. Clin. Invest. 9:129-135.Ebert R, Illmer K, Creutzfeldt W (1979b) Release of gastric inhibitory polypeptide (GIP)by intraduodenal acidification in rats and humans and abolishment of the incretineffect by OW-antiserum in rats. Gastroenterol. 76:515-523.Ebert R, Unger H, Creutzfeldt W (1983) Preservation of incretin activity after removal ofgastric inhibitory polypeptide (GIP) from rat gut extracts by immunoadsorption.Diabetologia 24:449-454.Eckel RH, Fujimoto WY, Brunzell IL) (1979) Gasthc inhibitory polypeptide enhanceslipoprotein lipase activity in cultured preadipocytes. Diabetes 28:1141-1142.Elahi D, Andersen DK, Brown JC, Debas HT, Hershcopf RJ, Raizes GS, Tobin JD,Andres R (1979) Pancreatic a- and B-cell responses to GIP infusion in normal man.Am. 3. Physiol. 237:E185-E191.Elahi D, Andersen DK, Muller DC, Tobin JD, Brown C JC, Andres R (1984) The entericenhancement of glucose-stimulated insulin release. The role of GIP in aging, obesityand noninsulin-dependent diabetes mellitus. Diabetes 33:950-957.Elahi D, Raizes OS, Andres R, Hershcopf RJ, Muller DC, Tobin JO, Andersen DK(1982) Interaction of arginine and gastric inhibitory polypeptide on insulin release inman. Am 3. of Physiol. 242:E343-E35 1.167El-Munshid HA, Häkanson R, Lindberg G, Sundler F (1980) Effects of variousgastrointestinal peptides on parietal cells and endocrine cells in the oxyntic mucosa ofrat stomach. J. Physiol. (London) 305:249-265Elrick ML, Stimmler L, Hlad CJ, Arai Y (1964) Plasma insulin response to oral andintravenous glucose administration. J. Clin. Endocrinol. 24: 1076-1082.Eissele R, Göke R, Willemer S, Harthus H-P, Vermeer H, Arnold R, Göke B (1992)Glucagon-like peptide- 1 cells in the gastrointestinal tract and pancreas of rat, pig andman. Eur. J. Clin. Invest. 22:283-29 1.Eissele R, Koop H, Arnold R (1990) Effect of glucagon-like peptide-1 on gastricsomatostatin and gastrin secretion in the rat. Scand. J. Gastroenterol. 25 :449-454.Eng J, Andrews PC, Kleinman WA, Singh L, Raufman JP (1990) Purification andstructure of extendin-3, a new pancreatic secretagogue isolated from Helodermahorridum venom. J. Biol. Chem. 265:20259-20262.Epstein PN, Ribar TJ, Decker GL, Yaney G, Means AR (1992) Elevated B-cellcalmodulin produces a unique insulin secretory defect in transgenic mice.Endocrinology 130:1387-1393.Ewald CA, Boas J (1886) Beitrage zur Physiologie und Pathologie der Verdauung II.Virchows Arch [Path Anat] 104:271-305.Falkmer 5, Ebert R, Creutzfeldt W (1980) The entero-insular axis In: Frontiers ofHormone Research, vol 7. Creutzfeldt W, Ed. Basel: S Karger, pp 1-6.Fallcmer S, El-Salhy M, Titlbach M (1984) Evolution of the neuroendocrine system invertebrates. A review with particular reference to the phylogeny and postnatalmaturation of the islet parenchyma. In: Evolution and Tumor Pathology of theNeuroendocrine System, Falkmer S, Hakanson R, Sundler F, eds. Amsterdam:Elsevier Science Publication, pp 59-87.Fallcmer S, Ostberg Y (1985) Comparative morphology of pancreatic islets in animals. In:The diabetic pancreas, BW Vollc and ER Arquilla, chapt. 2. New york: Elsevier/NorthHolland.Failco JM, Crockett SE, Cataland 5, Mazzaferri EL (1975) Gastric inhibitory polypeptide(GIP) stimulated by fat ingestion in man. J. Clin. Endocrinol. Metab. 41:260-265.Falko 3M, Crockett SE, Cataland 5, ODorisio TM, Kramer W, Mazzaferri EL (1980) Theeffect of increasing doses of ingested glucose on insulin and gastric inhibitorypolypeptide (GIP) concentrations in man. Clin. Endocrinol. 13:587-593.Farrell JI, Ivy AC (1926) Studies on the motility of the transplanted gastric pouch. Am. J.of Physiol. 46:227-228.Fehmann HC, Göke B, Göke R, Truatmann ME, Arnold R (1989) Synergistic stimulatoryeffect of glucagon-like peptide- 1 (7-36) amide and glucose-dependent insulin-releasing polypeptide on the endocrine rat pancreas. Fed. Eur. Biochem. Soc.252:109-112.168Fehmann HC, Habener JF (1991a) Homologous desensitization of the insulinotropicglucagon-like peptide-I(7-37) receptor on insulinoma (HIT-T15) cells. Endocrinology128:2880-2888.Fehmann HC, Habener IF (1991b) Functional receptors for the insulinotropic hormoneglucagon-like peptide-I(7-37) on a somatostatin secreting cell line. Fed. Eur.Biochem. Soc. 279:335-340.Ferraris RP, Yasharpour S, Lloyd KCK, Mirzayan R, Diamond 3M (1990) Luminalglucose concentrations in the gut under normal conditions. Am. 3. Physiol. 22:G822-G837.Fern GL, Adrian TE, Ghatei MA, O’Shaugnessy DJ, Probert L, Lee YC, Buchan AMJ,Polak JM, Bloom SR (1983) Tissue localization and relative distribution of regulatorypeptides in separated layers from the human bowel. Gastroenterol. 84:777-786.Flaten 0 (1981) The effect of adrenoceptor blockade on the release of gastric inhibitorypolypeptide after intraduodenal glucose in humans. Scand. 3. Gastroenterol. 16:641-645.Flaten 0, Sand T, Myren 3 (1982) Beta-adrenergic stimulation and blockade of therelease of gastric inhibitory polypeptide and insulin in man. Scand. 3. Gastroenterol.17:283-288.Flatt PR, Kwasowski P, Bailey CJ (1989) Stimulation of gastric inhibitory polypeptiderelease in ob/ob mice by oral administration of sugars and their analogues. 3. Nutr.119:1300-1303.Frier BM, Corrall 3M, Adrian TE, Bloom SR, Strachan RK, Heading RC (1984) Theinfluence of adrenergic denervation on the response to feeding of the gastropancreaticsystem in man. Clin. Endocrinol. 21:639-647.Frohman LA, Downs TR, Heimer EP, Felix AM (1989) Dipeptidylpeptidase IV andtrypsin-like enzymatic degradation of human growth hormone-releasing hormone inplasma. 3. Clin. Invest. 83:1533-1540.Frohman LA, Downs TR, Williams TC, Heimer EP, Pan YCE, Felix AM (1986) Rapidenzymatic degradation of growth hormone-releasing hormone by plasma in vitro andin vivo to a biologically inactive product cleaved at the NH2 terminus. 3. Clin. Invest.78:906-913.Fujii N, Sakurai M, Akaji K, Nomizu M, Yajima H, Mizuta K, Aono M, Moriga M,Inoue K, Hosotani R, Tobe T (1986) Studies on peptides. CXXXIX. Solutionsynthesis of a 42-residue peptide corresponding to the entire amino acid sequence ofhuman glucose-dependent insulinotropic polypeptide (GIP). Chem. Pharm. Bull.34:2397-2410.Fukase N, Igarashi M, Takahashi H, Manaka H, Yamatani K, Daimon M, Tominaga M,Sasaki H (1993) Hypersecretion of truncated glucagon-like peptide-l and gastricinhibitory polypeptide in obese patients. Diabet. Med. 10:44-49.Fushiki T, Kojima A, Imoto T, Inoue K, Sugimoto E (1992) An extract of Gymnemasylvestre leaves and purified gymnemic acid inhibits glucose-stimulated gastricinhibitory peptide secretion in rats. 3. Nutr. 122:2367-2373.169Gallacher DV, Hanley MR, Petersen OH, Roberts ML, Squire-Pollard LG, Yule DI(1990) Substance P and bombesin elevate cytosolic Ca2 by different molecularmechanisms in a rat pancreatic acinar cell line. J. Physiol. 426:193-207Gallwitz B, Witt M, Fölsch UR, Creutzfeldt W, Schmidt WE (1993) Binding specificityand signal transduction of receptors for glucagon-like peptide-1(7-36)amide andgastric inhibitory polypeptide on RINm5F insulinoma cells. J. Mol. Endocrinol.10:259-268.Garcia MC, Hermans MP, Henquin IC (1988) Glucose-, calcium- and concentration-dependence of acetyicholine stimulation of insulin release and ionic fluxes in mouseislets. Biochem. J. 254:2 1 1-2 18.Gayle RG, Ludewig RM (1978) Effect of truncal, selective and highly selectivevagotomy on fat-induced gastric inhibitory polypeptide release. Surg. Forum 29:397-402.Gillespie IE, Grossman MI (1964) Inhibitory effect of secretion and cholecystokinin onHeindenhain pouch responses to gastrin extract and histamine. Gut 5:342-345.Giraud AS, Soll AH, Cuttita F, Walsh JH (1987) Bombesin stimulation of gastrin releasefrom canine gastrin cells in primary culture. Am. J. Physiol. 252:G413-G420.Göke R, Conlon JM (1988) Receptor for glucagon-like peptide-1(7-36)amide on ratinsulinoma-derived cells. 3. Endocrinol. 166:357-362.Goto Y, Berelowitz M, Frohman LA (1981) Effect of catecholamines on somatostatinsecretion by isolated perfused rat stomach. Am. J. Physiol. 240:E274-E278.Grant SGN, Seidman I, Hanahan D, Bautch V (1991) Early invasiveness characterizesmetastatic carcinoid tumors in transgenic mice. Cancer Res. 5 1:4917-4923.Greely Jr GH, Spannagel A, Hill FLC, Thompson JC (1986a) Comparison of the actionsof bombesin, gastrin-releasing peptide-27, neuromedin B, and gastrin-releasingpeptide-lO in causing release of gastrin and gastric inhibitory peptide in rats. Proc.Soc. Exp. Biol. Med. 183:136-139.Greely Jr GH, Spannagel A, Trowbridge I, Thompson JC (1986b) Effect of bombesin andgastrin-releasing peptide on the release of gastric inhibitory polypeptide and insulin inrats. Proc. Soc. Exp. Biol. Med. 182:540-542.Greenwood FC, Hunter WM, Glover IS (1963) The preparation of j13 -labelled humangrowth hormone of high specific radioactivity. Biochem. 1. 89:114-123.Groop PH (1989) The influence of bodyweight, age and glucose tolerance on therelationship between GIP secretion and beta-cell function in man. Scand. J. Clin. Lab.Invest. 49:367-379.Gros L, Thorens B, Bataille D, Kervran A (1993) Glucagon-lilce peptide-1-(7-36)amide,oxyntomodulin, and glucagon interact with a common receptor in a somatostatinsecreting cell line. Endocrinology 133:631-638.170Gutniak M, Ørskov C, Hoist JJ, Ahrén B, Efendic S (1992) Antidiabetogenic effect ofglucagon-like peptide-1 (7-36)amide in normal subjects and patients with diabetesmellitus. N. Engi. J. Med. 326:1316-1322.Hicki WH, Bloom SR, Mitznegg P, Domschke W, Domschke S, Belohiavek D, DemlingL, WUnsch E (1977) Plasma secretin and pancreatic bicarbonate response toexogenous secretin in man. Gut 18:191-195.Hanks JB, Andersen DK, Wise JE, Putnam WS, Meyers WC, Jones RS (1984) Thehepatic extraction of gastric inhibitory polypeptide and insulin. Endocrinology115:1011-1018.Hampton SM, Kwasowski P, Tan K, Morgan LM, Marks V (1983) Effect of pretreatmentwith a high fat diet on the gastric inhibitory polypeptide and insulin responses to oraltriolein and glucose in rats. Diabetologia 24:278-28 1.Hanahan D (1984) Heritable formation of pancreatic B cell tumors in transgenic miceexpressing recombinant insulin/simian virus 40 oncogenes. Nature (Lond.) 315:115-122.Hanahan D (1988) Dissecting multistep tumorigenesis in transgenic mice. Annu. Rev.Genet. 22:479-519.Hanks JB, Andersen DK, Wise JE, Putnam WS, Meyers WC, Jones RS (1984) Thehepatic extraction of gastric inhibitory polypeptide and insulin. Endocrinology115:1011-1018.Hartmann E, Ebert R, Creutzfeldt W (1986) Insulin-dependent inhibition of hepaticglycogenolysis by gastric inhibitory polypeptide (GIP) in perfused rat liver.Diabetologia 29:112-114.Hassall CJS, Allen TGJ, Pittam BS, Burnstock G (1989) The use of cell and tissue culturetechniques in the study of regulatory peptides. EXS 56:113-136.Hauner H, Glatting G, Kaminska D, Pfeiffer EF (1988) Effects of gastric inhibitorypolypepetide on glucose and lipid metabolism of isolated rat adipocytes. Ann. Nutr.Metab. 32:282-288.Hazelwood RL (1989) Evolutionary, phylogenetic, and comparative considerations. In:The endocrine pancreas. Prentice Hall. chpt. 10.Ho LT, Pu HF, Sheu WJ, Wang WC, Wang PSG (1987) Inhibition of somatostatin ofglucose-induced release of gastric inhibitory polypeptide in rats. Chin. J. Physiol.30:45-53.Holst JJ, Ørskov C, Vagn Nielson 0, Schwartz T (1987) Truncated glucagon-like peptide1, an insulin-releasing honuone from the distal gut. FEBS Lett. 211:169-174.Holz GG, Habener IF (1992) Signal transduction crosstalk in the endocrine system:pancreatic B-cells and the glucose competence concept. TIBS 17:388-393.Holz GG, Kühtreiber WM, Habener iF (1993) Pancreatic beta-cells are rendered glucose-competent by the insulinotropic hormone glucagon-like peptide- 1(7-37). Nature361:362-365.171Hopfer U (1987) Membrane transport mechanisms for hexoses and amino acids in thesmall intestine. In: Physiology of the gastrointestinal tract. 2nd ed. LR Johnson (ed)Raven Press, New York, Chpt 55.Hopsu-Havu VK, Glenner GG (1966) A new dipeptide naphthylamidase hydrolyzingglycyl-prolyl-13-naphthylamide. Histochemie 7:197-201.Imamura M, Kameyama J, Naito H, Sato T, Ohneda A (1984) Influence of vagotomyupon GIP release in patients with peptic ulcer. Tohuku J. Exp. Med. 143:335-344.Imamura M, Takahashi M, Sasaki I, Yamauchi H, Sato T (1988) Effects of the pathwayof bile flow on the digestion of fat and release of gastrointestinal hormones. Am. J.Gastroenterol. 83:386-392.Inagaki N, Maekawa T, Sudo T, Ishii 5, Seino Y, Imura H (1992) C-Jun represses thehuman insulin promoter activity that depends on multiple cAMP response elements.Proc. Nati. Acad. Sci. USA 89: 1045-1049.Inagaki N, Seino Y, Takeda J, Yano H, Yamada Y, Bell GI, Eddy RL, Fukushima Y,Byers MG, Shows TB, Imura H (1989) Gastric inhibitory polypeptide: structure andchromsomal location of the human gene. Mol. Endocrinol. 3:1014-1021.Ishida T, Chap Z, Chou J, Lewis R, Hartley C, Entman M, Field J (1983) Differentialeffects of oral, peripheral intravenous and intraportal glucose on hepatic glucoseuptake and insulin and glucagon extraction in conscious dogs. J. Clin. Invest. 72:590-601.Ivy AC, Oldberg E (1927) Contraction and evacuation of the gall-bladder caused byhighly purified “secretin” preparation. Proc. Soc. Exp. Biol. Med. 25:113-115.Jaspan JB, Rubenstein AH (1977) Circulating glucagon plasma profiles and metabolismin health and disease. Diabetes 26:887-904,Johnson JH, Ogawa A, Chen L, Orci L, Newgard CB, Alam T, Unger RH (1990) Underexpression of B cell high Km glucose transporters in noninsulin-dependent diabetes.Science 250:546-549.Jones JR. Owens DR, Luzio SD, Hayes TM (1989a) Obesity is associated with increasedpostprandial GIP levels which are not reduced by dietary restriction and weight loss.Diabet. Metab. 15:11-22.Jones IR, Owens DR, Luzio SD, Hayes TM (1989b) Glucose dependent insulinotropicpolypeptide (GIP) infused intravenously is insulinotropic in the fasting state in type 2(non-insulin dependent) diabetes mellitus. Horm. Metabol. Res. 21:23-26.Jones IR, Owens DR, Moody AJ, Luzio SD, Morris T, Hayes TM (1987) The effects ofglucose-dependent insulinotropic polypeptide infused at physiological concentrationsin normal subjects and type 2 (non-insulin-dependent) diabetic patients on glucosetolerance and B-cell secretion. Diabetologia 30:707-711.Jones IR, Owens DR, Vora J, Luzio SD, Hayes TM (1989c) A supplementary infusion ofglucose-dependent inuslinotropic polypeptide (GIP) with a meal does notsignificantly improve the B cell response or glucose tolerance in type 2 diabetesmellitus. Diabetes Res. Clin. Prac. 7: 263-9.172Jones IR, Owens DR, Sarsons DL, Bloom SR (1985) Day profiles of glucose dependentinsulinotropic polypeptide (GIP) in normal subjects. Horm. Metabol. Res. 17:660-662.Jorde R, Burhol PG, Schulz TB (1983) Fasting and postprandial plasma GIP values inman measured with seven different antisera. Reg. Pept. 7:87-94.Jorde R, Burhol PG, Waldum HL (1981) The effect of a test meal on plasma gastricinhibitory polypeptide and serum gastrin in patients with achiorhydria andhypergastrinemia, and in patients with normal gastric H secretion. Digest. 21:304-309.Jorde R, Burhol PG, Waldum HL, Schulz TB, Lygren I, Florholmen J (1980) Diurnalvariation of plasma gastric inhibitory polypeptide in man. Scand. J. Gastroenterol.15:617-619.Jörnvall H, Carlquist M, Kwauk S, Otte SC, McIntosh CHS, Brown JC, Mutt V (1981)Amino acid sequence and heterogeneity of gastric inhibitory polypeptide (GIP).FEBS Lett. 123:205-210.Kieffer TJ, Verchere CB, Fell CD, Huang Z, Brown JC, Pederson RA (1993) Glucose-dependent insulinotropic polypeptide (GJP)-stimulated insulin release from a tumor-derived 13-cell line (13TC3). Can. J. Physiol. Pharmacol. 7 1:917-922.Kogire M, Izukura M, Gomez G, Uchida T, Greely Jr G, Thompson JC (1990)Terbutaline, a132-adrenoreceptor agonist, inhibits gastric acid secretion and stimulatesrelease of peptide YY and gastric inhibitory polypeptide in dogs. Dig. Dis. Sci.35:453-457.Koop H, Behrens I, McIntosh CHS, Pederson RA, Arnold R, Creutzfeldt W (1981)Adrenergic modulation of gastric somatostatin release in rats. FEBS Lett. 118:248-250.Koop I, Buchan AMJ (1992) Cholecystokinin release from isolated canine epithelial cellsin short-term culture. Gastroenterol. 102:28-34.Komatsu R, Matsuyama T, Namba M, Watanabe N, Itoh H, Kono N, Tarui 5 (1989)Glucagonostatic and insulinotropic action of glucagonlike peptide 1 -(7-36)-amide.Diabetes 38:902-905.Kosaka T, Lim RKS (1930) Demonstration of the humoral agent in fat inhibition ofgastric secretion. Proc. Soc. Exp. Biol. Med. 27:890-891.Kraenzlin ME, Ch’ng JLC, Mulderry PK, Ghatei MA, Bloom SR (1985a) Infusion of anovel peptide, calcitonin gene-related peptide (CGRP) in man. Pharmacokinetics andeffects on gastric acid secretion and on gastrointestinal hormones. Regul. Pept.10:189-197.Kraenziin ME, Wood SM, Neufeld M, Adrian TE, Bloom SR (1985b) Effect of longacting somatostatin-analogue, SMS 201995, on gut hormone secretion in normalsubjects. Experimentia (Birkhäuser Verlag) 41:738740.173Krarup T (1988) Immunoreactive gastric inhibitory polypeptide. Endocrine Rev. 9:122-134.Krarup T, Hoist JJ (1984) The heterogeneity of gastric inhibitory polypeptide in porcineand human gastrointestinal mucosa evaluated with five different antisera. Reg. Pept.9:35-46.Krarup T, Hoist JJ, Knuhtsen S, Moody AJ, Nielsen OV (1987a) Effect of porcine 8000dalton immunoreactive gastric inhibitory polypeptide on the isolated perfused pigpancreas. Diabetes 36:7A (Abstract).Krarup T, Hoist JJ, Larsen KL (1985) Responses and molecular heterogeneity of IR-GIPafter intraduodenal glucose and fat. Am. 3. Physiol. 249:E195-E200.Krarup T, Hoist JJ, Madsbad S (1987b) Heterogeneity of immunoreactive gastricinhibitory polypeptide in the plasma of newly diagnosed type 1 (insulin-dependent)diabetics. Acta Endocrinol. (Copenh.) 114:74-78.Krarup T, Saurbrey N, Moody AJ, KUhl C, Madsbad S (1987c) Effects of porcine gastricinhibitory polypeptide on beta-cell function in type I and II diabetes meilitus.Metabolism 36:677-682.Kreymann B, Ghatei MA, Burnet P, Williams G, Kanse S, Diani AR, Bloom SR (1989)Characterization of glucagon-like peptide-1-(7-36)amide in the hypothalamus. BrainRes. 502:325-331.Kreymann B, Ghatei MA, Williams G, Bloom SR (1987) Glucagon-like peptide 1 7-36: aphysiological incretin in man. Lancet 2:1300-1304.Kuzio M, Dryburgh JR, Malloy KM, Brown JC (1974) Radioimmunoassay for gastricinhibitory polypeptide. Gastroenterol. 66:357-364.Kwasowski P, Flatt PR, Bailey CT, Marks V (1985) Effects of fatty acid chain length andsaturation on gastric inhibitory polypeptide release in obese hyperglycaemic (ob/ob)mice. Biosc. Rep. 5:701-705.Kwok YN, McIntosh CHS, Sy H, Brown JC (1988a) Inhibitory action of tachykinins andneurokinins on somatostatin release from the isolated perfused stomach. J. Pharmacol.Exp. Therap. 246:726-731.Kwok YN, Verchere CB, McIntosh CHS, Brown JC (1988b) Effect of galanin onendocrine secretions from the isolated perfused rat stomach and pancreas. Eur. J.Pharmacol. 145:49-54.Labarre J (1927) Sur l’augmentation de la teneur en insuline du sang veineuxpancreatique apres excitation du nerf vague. Comptes Rendus des Seances de laSociet de Biologie et de ses Filiales (Paris) 96:193-196.Labarre J (1932) Stir les possibilities d’un traitement du diabete par Pincrétine. Bull.Acad. Roy. Soc. Med. Belg. 12:620-634.Labarre J, Still EV (1930) Studies on the physiology of secretin. Am. J. Physioi. 91:649-653.174Lardinois CK, Mazzaferri EL, Stanch GH, Hollenbeck CB, Reaven GM (1985) Increasedgastric inhibitory polypeptide is not reduced in patients with noninsulin-dependentdiabetes mellitus treated with intense insulin therapy. J. Clin. Endocrinol. Metab.61: 1089-1092.Lardinois CK, Richeson III RB, Stanch GH, Mazzu D, Mazzaferri EL (1990) Gastricinhibitory polypeptide mechanisms of augmenting insulin secretion. Life Sci.47:1015-1022.Lardinois CK, Stanch GH, Mazzaferri EL, DeLett A (1988) Polyunsaturated fatty acidsaugment insulin secretion. J. Am. College Nutr. 6:507-5 15.Larrimer JN, Mazzaferri EL, Cataland S, Mekhjian HS (1978) Effect of atropine onglucose-stimulated gastric inhibitory polypeptide. Diabetes 27:638-642.Larrson LI, Moody AJ (1980) Glicentin and gastric inhibitory polypeptideimmunoreactivity in endocrine cells of the gut and pancreas. J. Histochem. Cytochem.28:925-933.Lauritsen KB, Christensen KC, Stokhoim KH (1980) Gastric inhibitory polypeptide(GIP) release and incretin effect after oral glucose in obesity and after jejunoilealbypass. Scand. J. Gastroenterol. 15:489-492.Lauritsen K, Hoist 33, Moody AJ (1981) Depression of insulin release by anti-GIP serumafter oral glucose in rats. Scand. J. Gastroenterol. 16:417.Lauritsen KB, Moesgaard F, Nielson ML, Jensen SL (1982) Release of gastric inhibitorypolypeptide and insulin in response to intrajejunal glucose in duodenal ulcer patientsbefore and after truncal vagotomy. Scand. J. Gastroenterol. 17:507-511.Leidenheimer NJ, Browning MD, Dunwiddie TV, Hahner LD, Harris RA (1990)Phosphorylation-independent effects of second messenger system modulators on yaminobutyric acid receptor complex function. Mol. Pharmacol. 38:823-828.Levitt NS, Vinik AT, Child PT (1980) Glucose-dependent insulin-releasing peptide innoninsulin-dependent maturity-onset diabetes: effects of autonomic neuropathy. 3.Clin. Endocrinol. Metab. 51:254-258.Lindahl PE (1948) Principle of a counter-streaming centrifuge for the separation ofparticles of different sizes. Nature 161:648-649.Loew ER, Gray JS, Ivy AC (1940) Is a duodenal hormone involved in carbohydratemetabolism? Am. J. Physiol. 129:659-663.Lopez LC, Frazier ML, Su CJ, Human A, Sanders GF (1983) Mammalian pancreaticpreproglucagon contains three glucagon-related peptides. Proc. Nati. Acad. Sc.(U.S.A.) 80:5485-5489.Lu M, Wheeler MB, Leng XH, Boyd Ill AE (1993) The role of free cytosolic calciumlevel in B-cell signal transduction by gastric inhibitory polypeptide and glucagon-likepeptide 1(7-37). Endocrinology 132:94-100.Macgregor PF, Abate C, Curran T (1990) Direct cloning of leucine zipper proteins: Junbinds cooperatively to the CRE with CRE-BP1. Oncogene 5:45 1-458.175Magee DF, Nakamura M (1966) Action of pancreozymin preparations on gastricsecretion. Nature 212: 1487-1488.Maletti M, Altman 33. Hoa DHB, Carlquist M, Rosselin G (1987) Evidence of functionalgastric inhibitory polypeptide (GIP) receptors in human insulinoma. Binding ofsynthetic human GIP (1-31) and activation of adenylate cyclase. Diabetes, 36: 1336-1340.Maletti M, Cariquist M, Portha B, Kergoat M, Mutt V, Rosselin G (1986) Structuralrequirements for gastric inhibitory polypeptide (GIP) receptor binding and stimulationof insulin release. Peptides [Suppi 1] 7:75-78.Maletti M, Portha B, Cariquist M, Kergoat M, Laburthe M, Marie JC, Rosselin G (1984)Evidence for and characterization of specific high affinity binding sites for gastricinhibitory polypeptide in pancreatic B-cells. Endocrinology 115:1324-1331.Mangel AW, Snow ND, Misukonis MA, Basavappa S, Middleton JP, Fitz JO, Liddle RA(1993) Calcium-dependent regulation of cholecystokinin secretion and potassiumcurrents in STC-1 cells. Am. 3. Physiol. 264:G1031-G1036.Martin RA, Cleary DL, Guido DM, Zurcher-Neely HA, Kubiak TM (1993) Dipeptidylpeptidase IV (DPP-IV) from pig kidney cleaves analogs of bovine growth hormone-releasing factor (bGRF) modified at position 2 with Ser, Thr, or Val. Extended DPPIV substrate specificity? Biochim. Biophys. Acta 1164:252-260.Matschinsky FM (1990) Glucokinase as glucose sensor and metabolic signal generator inpancreatic B-cells and hepatocytes. Diabetes 39:647-652.Matthews EK, Sakamoto Y (1975) Electrical characteristics of pancreatic islet cells. J.Physiol. Lond. 246:421-437.Matumura C (1985) Dipeptidyl aminopeptidase IV (DAP IV) in urine. 3. Jpn. Pediatr.Soc. 33:287-291.Maxwell V, Shulkes A, Brown JC, Solomon TB, Walsh JH, Grossman MI (1980) Effectsof gastric inhibitory polypeptide on pentagastrin-stimulated acid secretion in man.Dig. Dis. Sci. 25:113-116.Mazzaferri EL, Starich GH, Lardinois CK, Bowen GD (1985) Gastric inhibitorypolypeptide responses to nutrients in caucasians and american indians with obesityand noninsulin-dependent diabetes mellitus. J. Clin. Endocrinol. Metab. 61:313-321.McDonald TJ, Ghatei MA, Bloom SR. Track NS, Radziuk 3, Dupré 3, Mutt V (1981) Aqualitative comparison of canine plasma gastroenteropancreatic hormone responses tobombesin and the porcine gastrin-releasing peptide (GRP). Regul. Pept. 2:293-304.McIntosh C, Arnold R, Boothe E, Becker H, Köbberling 3, Creutzfeldt W (1978)Gastrointestinal somatostatin: extraction and radioimmunoassay in different species.Gut 19:655-663.McIntosh CHS, Bakich V, Brown JC, Kwok YN, Nishimura E, Pederson RA (1989)Somatostatin and gastrin secretion from the isolated perfused rat stomach: neural andhormonal regulation and the effect of the ulcerogen cysteamine. In: Szabo S (ed)176Ulcer disease: new aspects of pathogenesis and pharmacology. CRC Press, BocaRaton pp. 217-228.McIntosh CHS, ha X, Kwok YN (1990) Characterization of the opioid receptor typemediating inhibition of rat gastric somatostatin release. Am. J. Physiol. 259:0922-G927.McIntosh CHS, Kwok YN, Mordhorst T, Nishimura E, Pederson RA, Brown JC (1983)Enkephalinergic control of somatostatin secretion from the perfused rat stomach. Can.J. Physiol. Pharmacol. 61:657-663.McIntosh CHS, Kwok YN, Tang C, Brown JC (1987) The use of monoclonal antibodiesin RIA for gastrointestinal hormones. I. Clin. Immunoassay 10:79-84.McIntosh CHS, Pederson RA, Koop H, Brown JC (1979) Inhibition of GIP stimulatedsomatostatin-like immunoreactivity (SLI) by acetyicholine and vagal stimulation. In:Miyoshi A, ed. Gut Peptides: secretion, function and clinical aspects. Tokyo:Kodansha 100-104.McIntosh CHS, Pederson RA, Koop H, Brown JC (1981) Gasthc inhibitory polypeptidestimulated secretion of somatostatin-like immunoreactivity from the stomach:inhibition by acetyicholine or vagal stimulation. Can. J. Physiol. Pharmacol. 59:468-472.McIntyre N, Holdsworth CD, Turner DS (1964) New interpretation of oral glucosetolerance. Lancet 11:20-2 1.McLoughlin JC, Buchanan KD, Alam MJ (1979) A glycoside-hydrolase inhibitor intreatment of dumping syndrome. Lancet 2:603-605.Mei N (1978) Vagal glucoreceptors in the small intestine of the cat. 3. Physiol. 282:485-506.Meneilly GS, Bryer-Ash M, Elahi D (1993) The effect of glyburide on B-cell sensitivityto glucose-dependent insulinotropic polypeptide. Diabetes Care 16:110-114.Mentlein R, Gallwitz B, Schmidt WE (1993) Dipeptidyl-peptidase IV hydrolyses gastricinhibitory polypeptide, glucagon-like peptide- 1(7 -36)amide, peptide histidinemethionine and is responsible for their degradation in human serum. Eur. J. Biochem.214:829-835.Modlin IM, Mitchell SJ, Bloom SR (1978) The systemic release and pharmacokinetics ofVIP. In: Bloom SR, ed. Gut hormones. Edinburgh: Churchill Livingstone, 470-474.Mojsov S, Heinrich G, Wilson IB, Ravazzola M, Orci L, Habener JF (1986)Preproglucagon gene expression in pancreas and intestine diversifies at the level ofpost-translational processing. J. Biol. Chem. 261:11880-11889.Mojsov S, Weir GC, Habener iF (1987) Insulinotropin: glucagon-like peptide (7-37) coencoded in the glucagon gene is a potent stimulator of insulin release in the perfusedrat pancreas. I. Clin. Invest. 79:616-619.Moody AJ, Damm JØrgensen K, Thim L (1981) Stucture-function relationships in porcineGIP. Diabetologia 21:306 (Abstract).177Moody AJ, Krarup T, Larsen UD (1992) The reactivity of anti-porcine gastric inhibitorypolypeptide (GIP) rabbit serum R65 with synthetic human and porcine GIP. Scand. J.Clin. Invest. 52:103-106.Moody AJ, Lauritsen KB (1977) A sensitive radio-immunoassay for plasma GIP. In:Speranza V, Basso N, Lezoche E (eds) Gastrointestinal Hormones and Pathology ofthe Digestive System Arti Grafiche Tris. Rome, p 148 (Abstract).Moody AJ, Thim L, Valverde I (1984) The isolation and sequencing of human gastricinhibitory polypeptide (GIP) FEBS Lett. 172:142-148.Moore B, Edie ES, Abram JM (1906) On the treatment of diabetes mellitus by acidextract of duodenal mucous membrane. Biochem. 1. 1:28-38.Morgan LM (1979) Immunoassayable gastric inhibitory polypeptide: investigations intoits role in carbohydrate metabolism. Annals Clin. Biochem. 16:6-14.Morgan LM, Hampton SM, Tredger JA, Cramb R, Marks V (1988) Modification ofgastric inhibitory polypeptide (GIP) secretion in man by a high-fat diet. Br. J. Nutr.59:373-380.Morgan LM, Morris BA, Marks V (1978) Radioimmunoassay of gastric inhibitorypolypeptide. Ann. Clin. Biochem. 15:172-177.Morgan LM, Tredger IA, Madden A, Kwasowski P, Marks V (1985) The effect of guargum on carbohydrate, fat and protein stimulated gut hormone secretion. Br. J. Nutr.53:467-475.Moroder L, Hallet A, Thamm P, Wilschowitz L, Brown JC, Wünsch E (1978) Studies ongastric inhibitory polypeptide: synthesis of the actatriacontapeptide GIP1-38 with fullinsulinotropic activity. Scand. I. Gastroent. [Suppi. 49] 13:129 (Abstract).Muller MK, Pederson RA, Brown JC (1982) Interaction of gastric inhibitory polypeptideand arginine in the perfused pancreas of rats treated with triiodothyronine. Can. 3.Physiol. Pharmacol. 60:297-301.Nagamatsu 5, Steiner D (1992) Altered glucose regulation of insulin biosynthesis ininsulinoma cells: mouse BTC3 cells secrete insulin-related peptides predominantly viaa constitutive pathway. Endocrinology 130:748-754.Nagata T, Umeda Y, Urakami M, Inada M (1988) Urinary dipeptidyl aminopeptidase IV(DAP IV) activity in patients with impaired glucose tolerance and diabetes mellitus. J.Jpn. Diabetic Soc. 31:169-171.Nathan DM, Schreiber E, Fogel H, Mojsov S, Habener JF (1992) Insulinotropic action ofglucagonlike peptide-I-(7-37) in diabetic and nondiabetic subjects. Diabetes Care15:270-276.Nauck MA, Bartels E, ørskov C, Ebert R, Creutzfeldt W (1992) Lack of effect ofsynthetic human gastric inhibitory polypeptide and glucagon-like peptide 1 [7-36amide] infused at near-physiological concentrations on pentagastrmn-stimulated gastricacid secretion in normal human subjects. Digestion 52:214-221.a178Nauck MA, Heimesaat MM, Ørskov C, Hoist JJ, Ebert R, Creutzfeldt W (1993)Preserved incretin activity of glucagon-like peptide 1 [7-36 amide] but not ofsynthetic human gastric inhibitory polypeptide in patients with type 2 diabetesmellitus. J. Clin. Invest. 91:301-307.Nauck MA, Homberger E, Siegel EG, Allen RC, Eaton RP, Ebert R, Creutzfeldt W(1986) Incretin effects of increasing glucose loads in man calculated from venousinsulin and C-peptide responses. J. Clin. Endocrinol. Metab. 63:492-498.Nauck MA, Schmidt WE, Ebert R, Strietzel J, Cantor P, Hoffmann G, Creutzfeldt W(1989) Insulinotropic properties of synthetic human gastric inhibitory polypeptide inman: interactions with glucose, phenylalanine, and cholecystokinin-8. J. Clin.Endocrinol. Metab. 69:654-662.Nelson RL, Go VLW, McCullough AJ, Ilstrup DM, Servic FJ (1986) Lack of a directeffect of the autonomic nervous system on glucose-stimulated gastric inhibitorypolypeptide (GIP) secretion in man. Dig. Dis. Sci. 3 1:929-935.Nicholl C, Sevelius H, Bloom SR (1986) Enprostil markedly reduces the postprandialrises in insulin and glucose-dependent insulinotropic peptide. Diabet. Med. 3:358a(Abstract).Nukada 0, Kobayashi M, Moriwake T, Kanzaki 5, Himei H, Yoda T, Seino Y (1992)Urinary glycylprolyl dipeptidyl aminopeptidase (GP-DAP) in insulin-dependentdiabetic patients. Acta Paediatr. 81:907-911.Oben J, Morgan L, Fletcher J, Marks V (1991) Effect of the enter-pancreatic hormones,gastric inhibitory polypeptide and glucagon-like polypeptide-1(7-36) amide, on fattyacid synthesis in explants of rat adipose tissue. J. Endocrinol. 130:267-272.O’Dorisio TM, Cataland 5, Stevenson M, Mazzaferri EL (1976) Gastric inhibitorypolypeptide (GIP). Intestinal distribution and stimulation by amino acids andmedium-chain triglycerides. Am. J. Dig. Dis. 21:761-765.ODoriso TM, Sirinek KR, Mazzaferri EL, Cataland S (1977) Renal effects on serumgastric inhibitory polypeptide (GIP). Metab. 26:651-656.Oektedalen 0, Opstad PK, Jorde R (1983) Increased plasma response of gastric inhibitorypolypeptide to oral glucose and a liquid meal after prolonged starvation in healthyman. Digest. 26:114-123.Ogata 5, Misumi Y, Ikehara Y (1989) Primary structure of rat liver dipeptidyl peptidaseIV deduced from its cDNA and identification of the NH2-terminal signal sequence asthe membrane-anchoring domain. J. Biol. Chem. 264:3596-3601.OHalloran DJ, Nilcou GC, Kreymann B, Ghatei MA, Bloom SR (1990) Glucagon-likepeptide-1 (7-36)-NH2:a physiological inhibitor of gastric acid secretion in man. J.Endocrinol. 126:169-173.Ohneda A, Kobayashi T, Nihei J (1983) Response of gastric inhibitory polypeptide (GIP)to oral administration of nutrients in normal and pancreatectomized dogs. Reg. Pept.5:263-272.179Ohneda A, Kobayashi T, Nihei J (1984) Response of gastric inhibitory polypeptide to fatingestion in normal dogs. Reg. Pept. 8:123-130.Opara EC, Go VLW (1993) Dual effects of gastric inhibitory polypeptide on insulinsecretion. Pancreas 8:39-43.Orci L, Ravazzola M, Baetens D, Inman L, Amherdt M, Peterson RG, Newgard CB,Johnson IH, Unger RH (1990) Evidence that down-regulation of B-cell glucosetransporters in non-insulin-dependent diabetes may be the cause of diabetichyperglycemia. Proc. Nail. Acad. Sci. USA 87:9953-9957.ørskov C, Holst JJ (1987) Radio-immunoassays for glucagon-like peptides 1 and 2(GLP-1 and GLP-2). Scand. J. Clin. Lab. Invest. 47:165-174.ørskov C, Jeppesen JU, Madsbad S, Hoist JJ (1991) Progiucagon products in plasma ofnoninsulin-dependent diabetics and nondiabetic controls in the fasting state and afteroral glucose and intravenous arginine. J. Clin. Invest. 87:415-423.Oshima I, Hirota M, Ohboshi C, Shima K (1988) Comparison of half-disappearancetimes, distribution volumes and metabolic clearance rates of exogenous glucagon-likepeptide 1 and glucagon in rats. Regul. Pept. 21:85-93.Oshima I, Yamamoto C, Shima K (1991) Half-disappearance time of endogenous GLP-1immunoreactivity in man. Horm. Metab. Res. 23:240-242.Otte SC, Mutt V. McIntosh CHS, Brown JC (1984) Purification and amino acidcompostion of an 8,000 dalton immunoreactive form of GIP. Dig. Dis. Sci. [Suppl.]29:63S (Abstract).Ottlecz A, Samson WK, McCann SM (1985) The effects of gastric inhibitory polypeptide(GIP) on the release of anterior pituitary hormones. Peptides 6:115-119.Parker DS, Raufman JP, ODonohue TL, Bledsoe M, Yoshida H, Pisano JJ (1984) Aminoacid sequences of helospectins, new members of the glucagon superfamily, found ingila monster venom. J. Biol. Chem. 259:11751-11755.Pavlov IV (1910) The work of the digestive glands. Translated by Thompson WH.Griffin, London.Pearse AGE (1968) Common cytochemical and ultrastructural characteristics of cellsproducing polypeptide hormones (the APUD series) and their relevance to thyroidand ukimobranchial C cells and calcitonin. Proc. R. Soc. Lond. Biol. Sci. 170:71-80.Pedersen 0, Hjollund E, Lindskov HO, Schwartz Sorensen N (1980) Increased insulinreceptors on monocytes from insulin dependent diabetics after high starch high fibrediets. Diabetologia 19:306 (Abstract).Pederson RA (1971) The isolation and physiological actions of gastric inhibitorypolypeptide. Vancouver, Canada: Univ. of B.C. PhD dissertation.Pederson RA, Brown JC (1972) Inhibition of histamine-, pentagastrin-, and insulin-stimulated canine gastric secretion by pure ‘gastric inhibitory polypeptide’.Gastroenterol. 62:393-400.180Pederson RA, Brown IC (1976) The insulinotropic action of gastric inhibitorypolypeptide in the perfused rat pancreas. Endocrinology 99:780-785.Pederson RA, Brown IC (1978) Interactions of gastric inhibitory polypeptide, glucoseand arginine on insulin and glucagon secretion from the perfused rat pancreas.Endocrinology 102:610-615.Pederson RA, Buchan AMJ, Zahedi-Asi S, Chan CB, Brown IC (1982) Effect of jejunoileal bypass in the rat on the entero-insular axis. Regul. Pept. 5:53-63.Pederson RA, Dryburgh JR, Brown JC (1975a) The effect of somatostatin on release andinsulinotropic action of gastric inhibitory polypeptide. Can. J. Physiol. Pharmacol.53:1200-1205.Pederson RA, Innis SM, Buchan AMJ, Chan CB, Brown IC (1985) The effect of totalparenteral nutrition (TPN) on the enteroinsular axis in the rat. Regul. Pept. 10:199-206.Pederson RA, Mochizuki T, Yanaihara C, Yanaihara N, Brown JC (1990) Reducedsomatostatinotropic effect of a GIP fragment (pGIP[l..30]-NH2)with insulinotropicactivity. Digestion 46 [Suppi. 1] 86 (Abstract)Pederson RA, Schubert HE, Brown IC (1975b) Gastric inhibitory polypeptide. Itsphysiologic release and insulinotropic action in the dog. Diabetes 24:1050-1056.Perley MI, Kipnis DM (1967) Plasma insulin responses to oral and intravenous glucose:studies in normal and diabetic subjects. I. Clin. Invest. 46:1954-1962.Pictet RL, Rall LB, Phelps P, Rutter WJ (1976) The neural crest and the origin of theinsulin-producing and other gastro-intestinal hormone-producing cells. Science191:191-192.Pictet RL, Rutter WI (1972) Development of the embryonic endocrine pancreas. In:Steiner DF, Frenkel M, eds. Handbook of Physiology, Sec. 7, Am. Physiol. Soc.,Washington pp 25-66.Pipeleers DG, Schuit FC, Veld PA, Maes E, Hooghe-Peters EL, Van de Winkel M, GeptsW (1985) Interplay of nutrients and hormones in the regulation of insulin release.Endocrinology 117:824-833.Poitras P, Dumont A, Cuber IC, Trudel L (1993) Cholinergic regulation of motilinrelease from isolated canine intestinal cells. Pept. 14:207-2 13.Polak 3M, Bloom SR, Kuzio M, Brown IC, Pearse AGE (1973) Cellular localization ofgastric inhibitory polypeptide in the duodenum and jejunum. Gut 14:284-288.Porte D Ir, Girardier L, Seydoux J, Woliheim CB, Kanazawa Y, Postemak 3 (1973)Neural regulation of insulin secretion in the dog. 3. Clin. Invest. 52:210-214.Prentki M, Matschinsky FM (1987) Ca2, cAMP, and phospholipid-derived messengersin coupling mechanisms of insulin secretion. Physiol. Rev. 4:1185-1248.181Rabinovitch A, Dupré J (1972) Insulinotropic and glucagonotropic activities of crudepreparation of cholecystokinin-pancreozymin. Clin. Res. 20:945 (Abstract).Rai A, Singh G, Raffaniello R, Eng J, Raufman JP (1993) Actions of Helodermatidaevenom peptides and mammalian glucagon-like peptides on gastric chief cells. Am. J.Physiol. 265:G1 18-G 125.Raufman JP, Singh L, Eng J (1991) Extendin-3, a novel peptide from Helodermahorridum venom, interacts with vasoactive intestinal peptide receptors and a newlydescribed receptor on dispersed acini from guinea pig pancreas. J. Biol. Chem.266:2897-2902.Rasmussen H, Zawalich KC, Ganesan S, Calle R, Zawalich WS (1990) Physiology andpathophysiology of insulin secretion. Diabetes Care 13:655-666.Rindi G, Grant SGN, Yiangou Y, Ghatei MA, Bloom SR, Bautch VL, Solcia E, Polak JM(1990) Development of neuroendocrine tumors in the gastrointestinal tract oftransgenic mice. Am. 3. Pathol. 136:1349-1363.Robberecht P, Gourlet P. Cauvin A, Buscail L, De Neef P. Arimura A, Christophe J(1991) PACAP and VIP receptors in rat liver membranes. Am. J. Physiol. 260:G97-G102.Roberge JN, Brubaker PL (1991) Secretion of proglucagon-derived peptides in responseto intestinal luminal nutrients. Endocrinology 128:3169-3174.Roberge JN, Brubaker PL (1993) Regulation of intestinal proglucagon-derived peptidesecretion by glucose-dependent insulinotropic peptide in a novel enteroendocrineloop. Endocrinology 133:233-240.Ross SA, Brown JC, Dupré 3 (1977) Hypersecretion of gastric inhibitory polypeptidefollowing oral glucose in diabetes mellitus. Diabetes 26:525-529.Ross SA, Dupré J (1978) Effects of ingestion of triglyceride or galactose on secretion ofgastric inhibitory polypeptide and on responses to intravenous glucose in normal anddiabetic subjects. Diabetes 27:327-333.Ross SA, Shaffer EA (1981) The importance of triglyceride hydrolysis for the release ofgastric inhibitory polypeptide. Gastroenterol. 80:108-111.Rosselin G (1986) The receptors of the VIP family peptides (VIP, secretin, GRF, PHI,PHM, GIP, glucagon and oxyntomodulin). Specificities and identity. Pept. [Suppl. 1]7:89-100.Rossowski WJ, Zacharia S. Mungan Z, Ozmen V, Ertan A, Baylor LM, Jiang NY, CoyDH (1992) Reduced gastric acid inhibitory effect of a pGIP(1-30)NH2fragment withpotent pancreatic amylase inhibitory activity. Regul. Pept. 39:9-17.Salera M, Ebert R, Giacomoni P. Pironi L, Venturi S, Corinaldesi R, Miglioli M, BarbaraL (1982a) Adrenergic modulation of gastric inhibitory polypeptide secretion in man.Dig. Dis. Sci. 27:794-800.182Salera M, Giacomoni P. Pironi L, Ustra C, Capelli M, Giorgi A, Miglioli M, Barbara L(1983) Circadian rhythm of gastric inhibitory polypeptide (GIP) in man. Metabolism32:21-24.Salera M, Pironi L, Giacomoni P, Venturi S, Capelli M, Miglioli M, Barbara L (1982b)Effect of somatostatin on fasting and glucose-stimulated gastric inhibitorypolypeptide release in man. Digest. 24:126-132.Sandberg E, Ahrén B, Tendler D, Carlquist M, Efendic S (1986) Potentiation of glucose-induced insulin secretion in the perfused rat pancreas by porcine GIP (gastricinhibitory polypeptide), bovine GIP, and bovine GIP (1-39). Acta. Physiol. Scand.127:323-326.Sandberg E, Ahrén B, Tendler D, Efendic S (1988) Cholecystokinin-33 potentiates andvasoactive intestinal peptide inhibits gastric inhibitory polypeptide-induced insulinsecretion in the perfused rat pancreas. Acta Endocrinol. (Copenhagen) 117:545-551.Sarson DL, Bryant MG, Bloom SR (1980) A radioimmunoassay of gastric inhibitorypolypeptide in human plasma. J. Endocrinol. 85:487-496.Sarson DL, Hayter RC, Bloom SR (1982) The pharmacokinetics of porcine glucose-dependent insulinotropic polypeptide (GJP) in man. Eur. J. Clin. Invest. 12:457-461.Schafer R, Schatz H (1979) Stimulation of (pro-)insulin biosynthesis and release bygastric inhibitory poypeptide isolated islets of rat pancreas. Acta Endocrinol.(Copenhagen) 91:493-500.Schauder P, Brown JC, Frerichs H, Creutzfeldt W (1975) Gastric inhibitory polypeptide:effect on glucose-induced insulin release from pancreatic islets in vitro. Diabetologia11:483-484.Schauder P, McIntosh C, Arends J, Arnold R, Frerichs H, Creutzfeldt W (1976)Somatostatin and insulin release from isolated rat pancreatic islets stimulated byglucose. FEBS Lett. 68:225-227.Schjoldager BTG, Mortensen PE, Cliristiansen J, ørskov C, Holst JJ (1989) GLP-1(glucagon-like peptide 1) and truncated GLP-1, fragments of human proglucagon,inhibit gastric acid secretion in humans. Dig. Dis. Sci. 34:703-708.Schluter KJ, Petersen KG, Kerp L (1980) Rapid changes in insulin binding affinity ofmonocytes in male and female volunteers after breakfast. Diabetologia 19:3 13(Abstract).Schmid R, Schusdziarra V, Aulehner R, Weigert N, Classen M (1990) Comparison ofGLP- 1 (7-36amide) and GIP on release of somatostatin-like immunoreactivity andinsulin from the isolated rat pancreas. Z. Gastroenterol. 28:280-284.Schmidt WE, Siegel EG, Gallwitz B, Kümmel H, Ebert R, Creutzfeldt W (1986)Characterization of the insulinotropic activity of fragments derived from gastricinhibitory polypeptide. Diabetologia 29:591A (Abstract).Schmidt WE, Siegel EG, Kümmel H, Gallwitz B, Creutzfeldt W (1987) Commerciallyavailable preparations of porcine glucose-dependent insulinotropic polypeptide (GIP)183contain a biologically inactive GIP-fragment and cholecystokinin-3 31-39.Endocrinology 120:835-837.Schmidtler J, Schepp W, Janczewska I, Weigert N, FUrlinger C, Schusdziarra V. ClassenM (1991) GLP-1-(7-36) amide, -(1-37), and -(1-36) amide: potent cAMP-dependentstimuli of rat parietal cell function. Am 3. Physiol. 260:G940-G950.Schulz TB, Jorde R, Burhol PG, Christensen 0 (1982) Augmented release of gastricinhibitory polypeptide into the portal vein in response to intraduodenal glucose andamino acids in anesthetized rats treated with methyiprednisolone of alloxan. Scand. J.Gastroenterol. 17:357-362.Schusdziarra V, Harris V, Conlon JM, Arimura A, Unger RH (1978) Pancreatic andgastric somatostatin release in response to intragastric and intraduodenal nutrients andHCL in the dog. 3. Clin. Invest. 62:509-5 18.Schusdziarra V, Rouiller D, Harris V. Pfeiffer EF, Unger RH (1980) Effect of bombesinupon plasma somatostatin-lilce immunoreactivity, insulin and glucagon in normal andchemically sympathectomized dogs. Regul. Pept. 1:89-96.Schwartz KE, Saito T (1989) Suppresion of alimentary lipemia in man by a prostaglandinanalogue (enprostil). Atherosci. 79:23 1-236.Seal AM, Meloche RM, Liu YQE, Buchan AMJ, Brown JC (1987) Effects of monoclonalantibodies to somatostatin on somatostatin-induced and intestinal fat-inducedinhibition of gastric acid secretion in the rat. Gastroenterol. 92:1187-1192.Service Fl, Hall LD, Westland RE, O’Brien PC, Go VLW, Matmond MW, Rizza RA(1983) Effects of size, time of day and sequence of meal ingestion on carbohydratetolerance in normal subjects. Diabetologia 25:316-321.Service FJ, Nelson RL, Rubenstein AH, Go VLW (1978) Direct effect of insulin onsecretion on insulin, glucagon, gastric inhibitory polypeptide and gastrin duringmaintenance of normoglycemia. J. Clin. Endocrinol. Metab. 47:488-493.Service FJ, Rizza RA, Westland RE, Hall LD, Gerich JE, Go VLW (1984) Gastricinhibitory polypeptide in obesity and diabetes. J. Clin. Endocrinol. Metab. 58:1133-1140.Sharma SK, Austin C, Howard A, Lo G, Nicholl CG, Legon S (1992) Characterization ofrat gastric inhibitory peptide eDNA. J. Mo!. Endocrinol. 9:265-272.Sharp GWG, Wiedenkeller DE, Kaelin D, Siegel EG, Wollheim CB (1980) Stimulationof adenylate cyclase by Ca2 and calmodulin in rat islets of langerhans. Diabetes29:74-77.Sherwin RS, Basti C, Finkeistein FO (1976) Influence of uremia and hemodialysis on theturnover and metabolic effects of glucagon. 3. Clin. Invest. 57 :722-731.Sheu WJ, Pu HF, Wang SW, Ho LT, Wang PSG (1987) Metabolic clearance rate andsecretion rate of gastric inhibitory polypeptide in the rat. Chin. J. Physiol. 30:25-33.184Shima K, Hirots M, Ohboshi C (1988) Effect of glucagon-like peptide-1 on insulinsecretion. Regul. Pept. 22:245-252.Shimazu I, Hirota M, Ohboshi C, Shima K (1987) Identification and localization ofglucagon-like peptide-1 and its receptor in rat brain. Endocrinology 121:1076-1082.Siegel EG, Creutzfeldt W (1985) Stimulation of insulin release in isolated islets by GIP inphysiological concentrations and its relation to islet cyclic AMP content.Diabetologia, 28: 857-861.Siegel EG, Schulze A, Schmidt WE, Creutzfeldt W (1992) Comparison of the effect ofGIP and GLP- 1 (7-36 amide) on insulin release from rat pancreatic islets. Eur. J. Clin.Invest. 22:154-157.Sinclair JD (1988) Multiple t tests are appropriate in science. TIPS 9:12-13.Sirinek KR, Cataland S, ODorisio TM, Mazzaferri EL, Crockett S. Pace WG (1977a)Augmented gastric inhibitory polypeptide response to intraduodenal glucose byexogenous gastrin and cholecystokinin. Surg. 82:438-442.Sirinek KR, Cataland S, O’Dorisio TM, Mazzaferri EL, Crockett S, Pace WG (1977b)Sympathetic neuroendocrine regulation of gastric inhibitory polypeptide response toglucose. Surg. Forum 28:377-379.Sirinek KR, Levine BA, O’Dorisio TM, Cataland S (1981) Parasympatheticneuroendocrine regulation of GIP response to glucose. J. Surg. Res. 30:343-347.Sirinek KR, Levine BA, O’Dorisio TM, Cataland S (1983) Gastric inhibitory polypeptide(GIP) release by actively transported, structurally similar carbohydrates. Proc. Soc.Exp. Biol. Med. 173:379-385.Sirinek KR, O’Dorisio TM, Levine BA, Cataland S (1979) Selective release of gastricinhibitory polypeptide by some actively transported, structurally similarcarbohydrates. Surg. Forum 30:358-360.Sirinek KR, Pace WG, Crocket SE, O’Dorisio TM, Mazzaferri EL (1978) Insulin-inducedattenuation of glucose-stimulated gastric inhibitory polypeptide secretion. Am. J.Surg. 135:151-155.Smith PH, Merchant FW, Johnson DG, Fujimoto WY, Williams RH (1977)Immunocytochemical localization of gastric inhibitory polypeptide-like materialwithin A-cells of the endocrine pancreas. Am. J. Anat. 149:585-590.Soll AH, Yamada T, Park J, Thomas LP (1984) Release of somatostatin-likeimmunoreactivity from canine fundic mucosal cells in primary culture. Am. 3.Physiol. 247:G558-G566.Someya Y, Inagaki N, Maekawa T, Seino Y, Ishii S (1993) Two 3’,5’-cyclic-adenosinemonophosphate response elements in the promoter region of the human gastricinhibitory polypeptide gene. Fed. Eur. Biochem. Soc. 3 17:67-73.Soon-Shiong P, Debas HT, Brown JC (1979) The evaluation of gastric inhibitorypolypeptide (GIP) as the enterogastrone. 3. Surg. Res. 26:681-686.185Starling EH (1905) The Croonian lectures on: the chemical correlation of the functions ofthe body. The Lancet 2:339-341.Stockmann F, Ebert R, Creutzfeldt W (1984) Preceding hyperinsulinemia preventsdemonstration of insulin effect on fat-induced gastric inhibitory polypeptide (GIP).Diabetes 33:580-585.Stoll RW, Touber IL, Menahan LA, Williams RH (1970) Clearance of porcine insulin,proinsulin, and connecting peptide by the isolated rat liver. Proc. Soc. Exp. Biol.Med. 133:894-896.Suzuki S, Kawai K, Ohashi S, Watanabe Y, Yamashita K (1992) Interaction of glucagonlike peptide-1(7-36)amide and gastric inhibitory polypeptide or cholecystokinin oninsulin and glucagon secretion from the isolated perfused rat pancreas. Metabol.41:359-363.Suzuki S, Kawai K, Ohashi S. Mukai H, Murayama Y, Yamashita K (1990) Reducedinsulinotropic effects of glucagonlike peptide I-(7-36)-amide and gastric inhibitorypolypeptide in isolated perfused diabetic rat pancreas. Diabetes 39:1320-1325.Suzuki Y, Erickson RH, Sedimayer A, Chang SK, Ikehara Y, Kim YS (1993) Dietaryregulation of rat intestinal angiotensin-converting enzyme and dipeptidyl peptidaseIV. Am. J. Physiol. 264:G1 153-G1 159.Sykes S, Morgan LM, English J, Marks V (1980) Evidence for preferential stimulation ofgastric inhibitory polypeptide secretion in the rat by actively transportedcarbohydrates and their analogues. J. Endocrinol. 85:201-207.Szecowka 3, Grill V, Sandberg E, Efendic S (1982a) Effect of Gil’ on the secretion ofinsulin and somatostatin and the accumulation of cyclic AMP in vitro in the rat. ActaEndocrinol. (Copenh.) 99:416-421.Szechowka J, Lins PE, Efendic S (1982b) Effects of cholecystokinin, gastric inhibitorypolypeptide, and secretin on insulin and glucagon secretion in rats. Endocrinology110:1268-1272.Takahashi H, Manaka H, Suda K, Fukase N, Sekikawa A, Eguchi H, Tominaga M, SasakiH (1991) Hyperglycaemia but not hyperinsulinaemia prevents the secretion ofglucagon-like peptide-1 (7-36 amide) stimulated by fat ingestion. Scand. 3. Clin. Lab.Invest. 51:499-507.Takasawa M, Kasahara S, Igarashi K (1990) Urinary glycylprolyldipeptidylaminopeptidase acitivity in diabetes mellitus. 3. Jpn. Diabetic Soc. 33:287-291.Takeda 3, Seino Y,Tanaka KI, Fukumoto H, Kayano T, Takahashi H, Mitani T, KuronoM, Suzuki T, Tobe T (1987) Sequence of an intestinal cDNA encoding human gastricinhibitory polypeptide precursor. Proc. Nat!. Acad. Sci. USA 84:7005-7008.Taraskevich PS, Douglas WW (1984) Electrical activity in adenohypophyseal cells andeffects of hypophysiotropic substances. Federation Proc. 43:2373-2378.186Taylor IL, Feldman M (1982) Effect of cephalic-vagal stimulation on insulin, gastricinhibitory polypeptide, and pancreatic polypeptide release in humans. J. Clin.Endocrinol. Metab. 55:1114-1117.Thams P. Capito K, Hedeskov CJ (1982) Different effects ofCa2-calmodulin onadenylate cyclase activity in mouse and rat pancreatic islets. Biochem. 3. 206:97-102.Thomas FB, Mazzaferri EL, Crockett SE, Mekhjian HS, Gruemer ND, Cataland S (1976)Stimulation of secretion of gastric inhibitory polypeptide and insulin by intraduodenalamino acid perfusion. Gastroenterol. 70:523-527.Thomas FB, Sinar D, Mazzaferri EL, Cataland 5, Mekhjian HS, Caidwell JH, Fromkes 33(1978) Selective release of gastric inhibitory polypeptide by intraduodenal amino acidperfusion in man. Gastroenterol. 74:1261-1265.Thomford NR, Sirinek KR, Crockett SE, Mazzaferri EL, Crockett 5, Pace WG (1974)Gastric inhibitory polypeptide response to oral glucose after vagotomy andpyloroplasty. Arch. Surg. 109:177-182.Thorens B (1993) Facilitated glucose transporters in epithelial cells. Annu. Rev. Physiol.55:591-608.Ticku MK, Mehta AK (1990) y-aminobutyric acid receptor desensitization in mice spinalcord cultured neurons: lack of involvement of protein kinases A and C. Molec.Pharm. 38:719-724.Tiruppathi C, Miyamoto Y, Ganapathy V, Roesel RA, Whitford G, Leibach FH (1990)Hydrolysis and transport of proline-containing peptides in renal brush-bordermembrane vesicles from dipeptidyl peptidase TV-positive and dipeptidyl peptidaseTV-negative rat strains. 3. Biol. Chem. 265:1476-1483.Tischler AS (1983) Gut peptide production by tumor cell lines in culture. Scand. 3.Gastroenterol. [Suppl.] 18:33-43.Tronier B, Deigard A, Andersen T, Madsbad 5 (1985) Absence of incretin effect in obesetype II and diminished effect in lean type IT and obese subjects. Diabetes Res. Clin.Pract. [Suppl. 1] 1:S568 (Abstract).Trube G, Rorsman P, Ohno-Shosaku T (1986) Opposite effects of tolbutamide anddiazoxide on the ATP-dependent K channel in mouse pancreatic B-cells. PfluegersArch. 407:493-499.Umezawa H, Aoyagi T, Ogawa K, Naganawa H, Hamada M, Takeuchi T (1984)Diprotins A and B, inhibitors of dipeptidyl aminopeptidase IV, produced by bacteria.3. Antibiotics XXXVII:422-425.Unger RH, Eisentraut AM (1969) Entero-insular axis Arch. Intern. Med. 123:261-266.Usdin TB, Mezey E, Button DC, Brownstein MJ, Bonner TI (1993) Gastric inhibitorypolypeptide receptor, a member of the secretin-vasoactive intestinal peptide receptorfamily, is widely distributed in periperal organs and the brain. Endocrinology133:2861-2870.187Usellini L, Capella C, Solcia E, Buchan AMJ, Brown IC (1984) Ultrastructurallocalization of gastric inhibitory polypeptide (Gil’) in a well characterized endocrinecell of canine duodenal mucosa. Histochem. 80:85-89.Verchere CB (1991) The role of gastric inhibitory polypeptide in the regulation ofpancreatic endocrine secretion. Vancouver, Canada: Univ. of B.C. PhD dissertation.Verchere CB, Kwok YN, Brown JC (1991) Modulation of acetyicholine-stimulatedinsulin release by glucose and gastric inhibitory polypeptide. Pharmacol. 42:273-282.Verdonk CA, Rizza RA, Nelson LE, Go VLW, Gerich JE, Service FJ (1980) Interactionof fat-stimulated gastric inhibitory polypeptide on pancreatic alpha and beta cellfunction. J. Clin. Invest. 65:1119-1125.Verspohl EJ, Ammon HPT (1987) Cholecystokinin (CCK8) regulates glucagon, insulinand somatostatin secretion from isolated islets: interaction with glucose. PfugersArchives 410:284-287.Vigna SR, Mantyh CR, Giraud AS, Soll AH, Walsh JH, Mantyh PW (1987) Localizationof specific binding sites for bombesin in the canine gastrointestinal tract.Gastroenterol. 93:1287-95.Villar HV, Fender HR, Rayford PL, Bloom SR, Ramus NI, Thompson IC (1976).Suppression of gastrin release and gastric secretion by gastric inhibitory polypeptide(GIP) and vasoactive intestinal polypeptide (VIP). Annals Surg. 184:97-102.Visher U, Blondel B, Wollheim CB, Hoppner W, Seitz HI, lynedjian PB (1987)Hexokinase isoenzymes in RIN-m5F insulinoma cells. Biochem. 1. 241: 249-255.Wahi MA, Plehn RI, Landsbeck EA, Verspohi EJ, Ammon HPT (1992) Are ionic fluxesof pancreatic beta cells a target for gastric inhibitory polypeptide? Mol. Cell.Endocrinol. 90:117-123.Walter R, Simmons WH, Yoshimoto T (1980) Proline specific endo- and exopeptidases.Mol. Cell. Biochem. 30:111-126.Watanabe K, Gola M (1987) Forskolin interaction with voltage-dependent K channels inHelix is not mediated by cyclic nucleotides. Neurosci. Lett. 78:211-216.Watanabe Y, Kojima T, Fujimoto Y (1987) Deficiency of membrane-bounddipeptidylaminopeptidase IV in a certain rat strain. Experientia 43:400-401.Weir GC, Mosjov S. Hendrick GC, Habener IF (1989) Glucagonlike peptide 1 (7-37)actions on endocrine pancreas. Diabetes 38:338-342.Wettergren A, Schjoldager B, Mortensen PE, Myhre J, Christiansen I, Holst II (1993)Truncated GLP-1 (proglucagon 78-107-amide) inhibits gastric and pancreaticfunctions in man. Dig. Dis. Sci. 38:665-673.Whitcomb DC, O’Dorisio TM, Cataland S, Nishikawara MT (1984) In vivo binding ofinsulin to target tissues using a new in vivo radioreceptor assay. Dig. Dis. Sci. 29:95S(Abstract).188Williams RH, Biesbroek TB (1980) Gastric inhibitory polypeptide and insulin secretionafter infusion of acetylcholine, catecholamines, and gut hormones. Proc. Soc. Exp.Biol. Med. 163:39-47.Williams RH, May 3M, Biesbroek JB (1981) Determinants of gastric inhibitorypolypeptide and insulin secretion. Metab. 30:36-40.Wilims B, Ebert R, Creutzfeldt W (1978) Gastric inhibitory polypeptide (GIP) andinsulin in obesity: II. Reversal of increased responses to stimulation by starvation orfood restriction. Diabetologia 14:379-387.Wishart J, Morris HA, Horowitz M (1992) Radioimmunoassay of gastric inhibitorypolypeptide in plasma. Clin. Chem. 38:2156-2157.Wolfe MM, Hocking MP, Maico DG, McGuigan TB (1983) Effects of antibodies togastric inhibitory peptide oa gastric acid secretion and gastrin release in the dog.Gastroenterol. 84:941-948.Wolfe MM, McGuigan JR (1982) Release of gastric inhibitory peptide following apeptone meal in the dog. Gastroenterol. 83:864-872.Wright E (1993) The intestinal Na/glucose transporter. Annu. Rev. Physiol. 55:575-589.Yada T, Itoh K, Nakata M (1993) Glucagon-like peptide-1-(7-36)amide and a rise incyclic adenosine 3’,5’-monophosphate increase cytosolic free Ca2 in rat pancreatic B-cells by enhancing Ca2 channel activity. Endocrinology 133:1685-1692.Yajima H, Fujii N, Akaji K, Sakurai M, Nomizu M, Misuta K, Aona M, Moriga M, InoueK, Histotani R, Tobe T (1985) Synthesis of a 42-residue peptide corresponding to theentire amino acid sequence of human GIP. Chem. Pharm. Bull. (Tokyo) 33:3578-3580.Yamagishi T, Debas HT (1980) Gastric inhibitory polypeptide (GIP) is not the primarymediator of the enterogastrone action of fat in the dog. Gastroenterol. 78:931-936.Yaron A, Naider F (1993) Proline-dependent structural and biological properties ofpeptides and proteins. Crit. Rev. Biochem. Mo!. Biol. 28:31-81.Yoshioka S (1986) Inhibitory effects of gymnemic acid and an extract from the leaves ofZizyphus jujuba on glucose absorption in the rat small intestine. J. Yonago Med.Assoc. 37:142-154.Yoshiya K, Yamamura T, Ishikawa Y, Utsunomiya 3, Takemura 3, Takeda 3, Seinao Y,Imura H (1985) Effect of truncal vagotomy on GLP release induced by intraduodenalglucose or fat in dogs. Digest. 31:41-46.Zawalich WS, Rasmussen H (1990) Control of insulin secretion: a model involving Ca2,cAMP and diacyglycerol. Mol. Cell. Endocrinol. 70:119-137.Zawalich WS, Takuwa N, Takuwa Y, Diaz VA, Rasmussen H (1987) Interactions ofcholecystokinin and glucose in rat pancreatic islets. Diabetes 36:426-433.68Tççz-€zddA1wsApuiisiwons9qj(ojois(qjjojooqpui:uIuo!lDunJputuoiirniuaijjipii’Iu!1s’ujosipnisusuqu°painnoJo(1661)apluAno’]‘1ssi9‘jqunqw‘vmnqpvv-OVTEOSAqJiuiqais1JDudnpuiuii,uixsuoraxssjauusuowpi:s11Du1dnpuJ1u!uouisijpsnbiojoisAqdSUO!11I1AsppnijsuoT1nquwo(6z6T)fuiqrT‘zunE8-LLVVicIjjn’>jSSuATSuodsJJpguoi-pddmuonjJounuui(66T)HUSSflWSJ‘D)IP!T’”Z‘SM


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items