THE JOURNAL OF BIOLOGICAL CHEMISTRY
`
`Vol. 268, No. 26, Issue of September 15, pp. 19650-19655,1993
`Printed in U.S.A.
`
`Exendin-4 Is a High Potency Agonist and Truncated Exendin-(9-39)-
`amide an Antagonist at the Glucagon-like Peptide 1-(7-36)-amide
`Receptor of Insulin-secreting ,&Cells*
`
`(Received for publication, March 5, 1993, and in revised form, May 7, 1993)
`
`Riidiger Goke, Hans-Christoph Fehmann, Thomas LinnS, Harald Schmidt, Michael Krause9,
`John EngT, and Burkhard GokeII
`From the Departments of Internal Medicine, Philipps University of Marburg, 3550 Marburg, the $Justus-Liebig University of
`Giessen, 6300 Giessen, the §Microchemistry Unit of the Institute of Molecular Biology and Cancer Research, 3550 Marburg,
`Germany and the Wolomon A. Berson Research Laboratory, Veterans Affairs Medical Center, Bronx, New York 10468
`
`Exendin-4 purified from Heloderma suspecturn
`venom shows structural relationship to the important
`incretin hormone
`glucagon-like peptide 1-(7-36)-
`amide (GLP-1). We demonstrate that exendin-4 and
`truncated exendin-(9-39)-amide specifically interact
`with the GLP-1 receptor on insulinoma-derived cells
`and on lung membranes.
`Exendin-4 displaced "'I-
`GLP- 1, and unlabeled GLP- 1 displaced lZ6I-exendin-4
`from the binding
`site at rat insulinoma-derived
`RINmSF cells. Exendin-4 had,
`like GLP-1, a pro-
`nounced effect on intracellular CAMP generation,
`which was reduced by exendin-(9-39)-amide. When
`combined, GLP-1 and exendin-4 showed additive ac-
`tion on CAMP. They each competed
`with the radio-
`labeled version of the other peptide in
`cross-linking
`experiments. The apparent molecular mass of the re-
`spective ligand-binding protein complex was 63,000
`Da. Exendin-(9-39)-amide abolished the cross-linking
`of both peptides. Exendin-4, like GLP-1, stimulated
`dose dependently the glucose-induced insulin
`wcre-
`tion in isolated rat
`islets, and, in mouse insulinoma
`@TC-1 cells, both peptides stimulated the proinsulin
`gene expression at the level of transcription. Exendin-
`(9-39)-amide reduced these
`effects. In conclusion,
`exendin-4 is an agonist and exendin-(9-39)-amide is
`a specific GLP- 1 receptor antagonist.
`
`glucagon show some sequence homology to GLP-1,' these
`peptides exhibit no interactions with the recently discovered
`(61, biochemically characterized (6-8), and cloned (9) GLP-1
`receptor. We have previously studied the insulin stimulatory
`action of the reptile peptide helodermin, which was isolated
`from Gila monster venom (lo), because of the striking struc-
`tural homology of helodermin and GLP-1. Since helodermin
`induced only a weak stimulatory effect on insulin secretion
`(lo), the question of whether this peptide interacts with
`endocrine pancreatic islet GLP-1 receptors was not further
`addressed.
`The venom of the lizard Gila monster contains several
`bioactive peptides. Helodermin was isolated from Heloderma
`suspecturn venom and the helospectins I and I1 from Helod-
`erma horridurn (11,12). In recent studies, two new biologically
`active peptides, exendin-3 (13) and -4 (14) have been identi-
`fied from Helodermatidae venoms (15, 16). In our search for
`naturally occurring GLP-1 receptor agonists/antagonists, we
`became interested in the exendins, because exendin-3 and -4
`have been shown to enhance insulin secretion when infused
`in dogs (17). Most interestingly, the exendins share a 53%
`sequence homology with GLP-l-(7-36)-amide.
`Consequently, we investigated whether the putative biolog-
`ical action of exendin-4 at the endocrine pancreas is mediated
`by GLP-l-(7-36)-amide receptors. For this, we utilized as P-
`cell models the well characterized rat insulinoma-derived
`RINm5F cell line (18) in which we previously discovered GLP-
`1-(7-36)-amide receptors (6). Studies with mouse insulinoma-
`derived PTC-1 cells have been included to test for effects on
`proinsulin gene regulation. In order to complete those inves-
`tigations of insulinotropic effects, we looked at insulin secre-
`tion studies with normal islets. It was also intriguing to find
`out whether truncated exendin-(9-39)-amide behaves as a
`GLP-1 receptor antagonist at the @-cell as was suggested for
`acinar cells of the guinea pig exocrine pancreas (16).
`Rat lung membranes also possess GLP-l-(7-36)-amide
`receptors with biochemical properties that are somewhat dif-
`ferent from the P-cell receptors on RINm5F cells (19-21). We
`found these receptors localized to mucus glands in the trachea
`and on smooth muscle of the pulmonary artery (22) and
`believe that they contribute to the non-adrenergic non-cholin-
`ergic peptidergic regulation of lung function. It was therefore
`of additional interest to study the interaction of exendin-4
`and GLP-l-(7-36)-amide with the GLP-1 lung receptor.
`
`The abbreviations used are: GLP-1, glucagon-like peptide 1; KRB,
`Krebs-Ringer buffer.
`
`~~
`
`Intestinal peptide hormones, released into the circulation
`after a meal, are important augmentors of the postprandial
`insulin release stimulated by absorbed nutrients (incretin
`effect) (1, 2). Glucagon-like peptide 1-(7-37)/(7-36)-amide is
`now identified as probably the most important hormonal
`mediator in this enteroinsular axis and in glucose homeostasis
`(3, 4). This peptide is derived from the intestinal posttrans-
`lational proglucagon processing and is a member of an ex-
`tended family of bioactive peptides, the glucagon-secretin-
`vasoactive intestinal polypeptide family, all of which are
`closely related one to the other in their amino acid sequences
`(3, 5). Although mammalian peptides such as secretin and
`* This work was supported by Deutsche Forschungsgemeinschaft
`Grants Go 492/2-1 (to R. G.) and Go 417/3-1 (to B. G.) The costs of
`publication of this article were defrayed in part by the payment of
`page charges. This article must therefore be hereby marked "adver-
`tisement" in accordance with 18 U.S.C. Section 1734 solely to indicate
`this fact.
`11 To whom correspondence should be addressed: Laboratory of
`Molecular Endocrinology, Division of Gastroenterology and Metab-
`olism, Dept. of Internal Medicine, Philipps University of Marburg,
`Baldinger Str., W-3550 Marburg, Germany. Fax: 49-6421-285648.
`
`19650
`
`SANOFI-AVENTIS Exhibit 1018 - Page 19650
`
`IPR for Patent No. 8,951,962
`
`

`
`TABLE I
`K d values of binding experiments with RINm5F cells and
`rat lung membranes
`Kd
`RINm5F cell experiments
`Inhibition of '"I-GLP-1 (7-36) amide binding by
`3.45 f 0.7 X 10"' M
`GLP-1 (7-36) amide
`1.36 f 0.4 X 10"' M
`Exendin-4
`2.99 f 0.7 X IO-' M
`Exendin (9-39) amide
`Inhibition of 12sI-exendin-4 binding by
`4.27 f 0.7 X io-' M
`GLP-1 (7-36) amide
`2.73 f 0.3 X 10"' M
`Exendin-4
`3.99 -+ 1.2 x lo-' M
`Exendin (9-39) amide
`Kd
`Lung membranes
`Inhibition of 1251-GLP-1 (7-36) amide binding by
`3.93 f 0.2 X 1o"o M
`GLP-1 (7-36) amide
`1.34 f 0.2 X 10"' M
`Exendin-4
`5.25 f 0.3 X IO-' M
`Exendin (9-39) amide
`Inhibition of '261-exendin-4 binding by
`7.65 f 1.4 X 10"' M
`GLP-1 (7-36) amide
`2.55 f 0.3 X 10"' M
`Exendin-4
`6.88 f 1.6 X lo-' M
`Exendin (9-39) amide
`
`100
`
`B
`
`6c
`
`20
`
`O
`
`-11 -10 -8
`
`-s
`
`-7 -8
`
`100
`
`2
`4 80
`a
`P 60
`B
`i
`
`40
`
`I
`
`6c
`
`20
`
`0
`
`100
`
`a 2 80
`a
`B U 60
`-
`Y i 40
`
`6c 20
`
`-8 -8 -7
`-11 -10
`-6
`Peptide log (M)
`
`\
`
`Exendins on GLP-1 Receptor of Islet P-Cells
`19651
`(Munchen, Germany);
`NaIz6I from Amersham-Buchler (Braun-
`schweig, Germany); human serum albumin from Behring (Marburg,
`Germany); bacitracin from Serva (Heidelberg, Germany).
`RINm5F (23) and flC-1 (28) cells were grown in plastic culture
`bottles under conditions as previously described. Before experimen-
`tation, cells were detached from the bottles using phosphate-buffered
`saline (NaC1, 136 mM; KC1, 2.7 mM; NazHP04, 8.1 mM; KHzPO,, 1.5
`mM, pH 7.3) containing 0.7 mM EDTA.
`Lung membranes were prepared as described in detail previously
`(8). Rat lungs were homogenized, and the homogenate was layered
`over a 41% (w/v) sucrose solution and centrifuged at 4 "C for 60 min
`at 100,000 X g. The band at the interface of the layers represented
`the membranes and was collected, diluted, and centrifuged at 4 "C for
`30 min at 40,000 X g. The pellet was resuspended in a modified Krebs-
`Ringer buffer (KRB) (Tris-HC1,2.5 mM; NaC1,120 mM; MgSO4,1.2
`mM; KCl, 5 mM; CH3COONa, 15 mM, pH 7.4) containing 1% (w/v)
`human serum albumin, 0.1% (w/v) bacitracin, and 1 mM EDTA
`frozen in liquid nitrogen and stored at -80 'C. Protein concentration
`was determined as described by Bradford (24).
`12'I-GLP-1-(7-36)-amide was prepared as described previously (6).
`Iodination of exendin-4 was carried out using a modification of this
`method. Briefly, exendin-4 (5 pg/5 pl of 0.1% (v/v) trifluoroacetic
`acid) was diluted with 95 pl of sodium phosphate buffer (0.2 mM, pH
`7.5) and was added to a microcentrifuge tube containing a film of
`iodogen. The reaction was started by addition of 1 mCi Na'"1
`and
`proceeded for 3 min at 20 "C. The crude tracer was then injected onto
`a C18 Nova-pak reverse-phase high performance liquid chromatogra-
`phy column (0.46 X 15 cm; Waters Associates, Milford, MA) equili-
`brated at a flow rate of 1.5 ml/min 0.1% trifluoro acetic acid. The
`concentration of acetonitrile in the eluting solvent was raised to 70%
`(v/v) over 60 min. The peak of radiolabeled tracer was stored at
`-20 'C. Under the conditions of chromatography, labeled peptide was
`separated from unlabeled peptide, and the specific activity of the
`label was estimated to be approximately 65 TBq/mmol.
`Binding assays were performed as follows. RINm5F cells were
`detached from the culture bottles, were centrifuged (100 X g, 5 min),
`and the pellet was resuspended in KRB. Cells (-lo6) or lung mem-
`branes (1 mg of protein/ml) were incubated in KRB at 37 "C with
`radiolabeled peptide in the absence and presence of a range of
`concentrations of unlabeled peptide. After 30 min, the reaction was
`stopped by centrifugation of sample aliquots through an oil layer (6).
`Cell surface-associated radioactivity in the pellet was determined
`using a y-counter. Specific binding was calculated as the difference
`between the amount of 1261-GLP-1-(7-36)-amide or lZ6I-exendin-4
`bound in the absence (total binding) and presence (nonspecific bind-
`ing) of 1 p~ unlabeled GLP-1 or exendin-4, respectively.
`Cross-linking experiments were carried out as described previously
`(7, 20). Briefly, plasma membranes (1 mg of protein/ml) were incu-
`bated with '261-GLP-1-(7-36)-amide or 12SI-exendin-4 (-8 X 10' c.p.m.
`each) in 50 mM HEPES buffer pH 7.5 containing 0.02% (w/v) human
`serum albumin for 30 min at 37 "C in the presence and absence of
`unlabeled peptides (see Fig. 2). After centrifugation for 5 min at
`10,000 X g and 4 "C, the pellet was resuspended in ice-cold 10 mM
`HEPES, 300 mM mannitol buffer (pH 9.0). Disuccinimidyl suberate
`dissolved in dimethylsulfoxide was added to give a final concentration
`of 0.1 mM. After an incubation of 10 min at 0 'C, the reaction was
`terminated by addition of ammonium acetate (final concentration, 10
`mM), the membranes were centrifuged, and the pellets were resus-
`pended in HEPES-mannitol buffer (pH 7.5). An incubation for 30
`min at 25 "C was carried out in order to dissociate noncovalently
`bound tracer. After centrifugation, pellets were resuspended in 10
`mM sodium phosphate buffer (pH 7.5) containing 2% (w/v) SDS.
`Samples were boiled for 5 min in the presence of 5% (v/v) mercap-
`toethanol. Electrophoresis was carried out as described by Laemmli
`(25). After drying, gels were exposed to Kodak type AR film for up
`to 2 weeks at -80 "C using a light-intensifying screen.
`CAMP concentrations were determined as previously described (6).
`Briefly, cells (-lo6) in 500 pl of buffer (NaCl, 113 mM; KCl, 4.7 mM;
`KHzPO4, 1.2 mM; HEPES, 10 mM; CaClZ, 2.5 mM; MgSO,,
`1.2 mM,
`pH 7.4) containing 0.1% (w/v) human serum albumin were incubated
`for 3 min at 37 "C in the presence and absence of peptide. The
`reaction was stopped by addition of ice-cold trichloroacetic acid (12%)
`(w/v) and sonification of the sample. CAMP concentrations were
`determined using a kit from Du Pont-New England Nuclear according
`to the manufacturer's instructions.
`Rat pancreatic islets were isolated aseptically by collagenase and
`discontinuous Ficoll density gradient centrifugation at 800 X g for 10
`min (26). Islets were maintained overnight at 37 "C, pH 7.4, in a gas
`
`-6
`
`0
`
`-11 -10 -8 -8 -7
`Peptide log (M)
`Peptide log (M)
`FIG. 1. Binding experiments were performed for 30 min at
`37 "C. The tracer concentrations utilized amounted to 150 PM. In
`RIN cell experiments, 25% of total radioactivity was specifically
`bound (in lung membrane experiments, 8%). For the respective Kd
`values, see Table I. A, binding of '2'-I-GLP-1-(7-36)-amide to
`RINm5F cells. Labeled GLP-1 was concentration-dependently dis-
`placed by unlabeled GLP-l-(7-36)-amide, exendin-4, and exendin-
`(9-39)-amide. Data show means of six experiments. B, binding of
`I-exendin-4 to RINm5F cells. The order of potency of displacement
`by the unlabeled peptides was exendin-4 > GLP-1= exendin-(9-39)-
`amide. Shown are means of six experiments. C, inhibition of "'-1-
`GLP-l-(7-36)-amide binding to lung membranes by exendin-4, GLP-
`1, and exendin-(9-39)-amide (means of six experiments). D, inhibi-
`tion of 125-I-exendin-4 binding to lung membranes by exendin-4,
`GLP-1, and exendin-(9-39)-amide (means of six experiments).
`
`MATERIALS AND METHODS
`Exendin-4, exendin-(9-39)-amide, and [Y3']exendin-4 were syn-
`thesized as has been detailed before (16). Briefly, the peptides were
`produced on solid-phase support (PAL resin) utilizing activated N-
`(9-fluorenyl)methoxycarbonyl amino acids on a Milligen 9050 peptide
`synthesizer (Milligen, Burlington, MA) and were purified by prepar-
`ative high pressure liquid chromatography.
`All other peptides were purchased from Bachem (Heidelberg, Ger-
`many). Iodogen and disuccinimidyl suberate was from Pierce
`
`SANOFI-AVENTIS Exhibit 1018 - Page 19651
`
`IPR for Patent No. 8,951,962
`
`

`
`19652
`
`Exendins on GLP-1 Receptor of Islet @-Cells
`
`94
`
`67
`
`43
`
`67
`
`43
`
`l2%-GLP-1
`I2%-Exendin-4
`FIG. 2. Autoradiographs of 1261-GLP-1-(7-36)-amide (left panel) and l2'1-exendin-4 (right panel) cross-linked to receptors
`solubilized from RINm5F cell membranes. RINm5F cell membranes were incubated for 30 min at 37 "C with T-GLP-1 (kft) or lz5I-
`exendin-4 (right) in the absence (first l a n e ) or presence of unlabeled peptides. The membranes were then pelleted, washed, and cross-linked
`(0.1 mM disuccinimidyl suberate) for 10 min at 4 "C followed by SDS-electrophoresis. The autoradiograph presented is representative of three
`separate experiments. The mobilities of phosphorylase b (molecular mass, 94 kDa), human serum albumin (67 kDa), and ovalbumin (43 kDa)
`are indicated by the numbers at the left of the respective panek.
`
`phase of 5% CO:, in air in plastic tissue culture dishes, each containing
`5 ml of RPMI 1640 (Flow Lab, Irivine, UK) supplemented with 11
`mM glucose and 10% fetal calf serum. Insulin secretion was measured
`in static incubations. Batches of 10 islets were incubated for 45 min
`in a gently shaking water bath (37 "C) in tubes containing 500 pl of
`KRB buffer solution (pH 7.4) supplemented with 2 mg/ml bovine
`serum albumin (Sigma), 10 mM HEPES, 10 mM glucose, and varying
`concentrations of GLP-l-(7-36)-amide and exendin-4. Incubation
`was preceded by a 30-min preincubation period in glucose-free me-
`dium. Insulin was analyzed by radioimmunoassay utilizing rat insulin
`as standard (Novo, Mainz, Germany). The assay sensitivity was 5
`microunits/ml, the interassay variance was 11.8%, and the intraassay
`variance amounted to 9.7%. Mean protein content of 10 rat islets was
`4.95 ? 0.84 pg.
`The effect of exendin-4 on proinsulin biosynthesis was studied in
`flC-1 cells, as has been detailed before (27). Briefly, cells were
`transfected with the rat insulin I promotor-indicator plasmid using
`the diethylaminoethyl-dextran method. The plasmid contained 410
`base pairs of the rat insulin I gene promotor fused to the transcrip-
`tional reporter gene encoding the bacterial enzyme chloramphenicol
`acetyltransferase. The plasmid was constructed by linking the PuuII/
`RsaI fragment (-140 to +65) of the rat insulin I gene into the vector
`PO-CAT. Chloramphenicol acetyltransferase activities were deter-
`mined according to routine procedures (27). Chloramphenicol acetyl-
`transferase activities were determined in 55 pl of cell lysates, and
`incubation with the assay mixture (27) was carried out for 120 min.
`Treatment of cells with exendin-4, GLP-1, or exendin-(9-39)-amide
`was started immediately after transfections and continued for 24 h
`according to (27).
`Scatchard analysis was done using a version of the program LI-
`GAND modified for the IBM PC (G. A. McPherson, Elsevier-Biosoft,
`Cambridge, UK). Statistical analysis was performed using Student's
`t test for paired data. Data are expressed as means f S.E. p values of
`less than 0.05 were considered to be significant.
`
`RESULTS
`Binding Experiments-See Table I. Binding of '251-GLP-1-
`(7-36)-amide to RINm5F cells was inhibited concentration-
`dependently by unlabeled GLP-l-(7-36)-amide, exendin-4,
`and exendin-(9-39)-(Fig. L4). Binding of lZ5I-exendin-4 to
`RINm5F cells was also inhibited concentration-dependently
`by the peptides (Fig. 1B); the order of potency was exendin-
`4 > GLP-l-(7-36)-amide = exendin (9-39).
`Using lz5I-GLP-l-(7-36)-amide in binding studies with lung
`membranes, the order of potency of peptides displacing the
`tracer from its binding sites was exendin-4 > GLP-1-(7-36)-
`amide > exendin-(9-39)-(Fig. IC). In experiments with lZ5I-
`exendin-4, unlabeled exendin-4 was most potent in displacing
`the tracer followed by GLP-l-(7-36)-amide and exendin-(9-
`39)-(Fig. 1D).
`Vasoactive intestinal peptide and peptide histidine isoleu-
`cine in concentrations up to 1 phd were not able to displace
`'"I-exendin-4 from its binding sites on lung membranes (data
`not shown).
`Cross-linking Experiments-Incubation of RINm5F plasma
`membranes with 1251-GLP-1-(7-36)-amide followed by chem-
`ical cross-linking and solubilization identified a single ligand-
`binding protein complex with an apparent molecular mass of
`63,000 Da (Fig. 2, left). This band was not detectable when
`the incubations were carried out in the presence of 1 ~ L M
`unlabeled GLP-l-(7-36)-amide. Furthermore, addition of un-
`or exendin-(9-39)-amide abolished
`labeled exendin-4 (1 p ~ )
`
`the specific labeling of the binding site. The same was true
`for experiments with lz5I-exendin-4 (Fig. 2, right). The ligand-
`
`SANOFI-AVENTIS Exhibit 1018 - Page 19652
`
`IPR for Patent No. 8,951,962
`
`

`
`19653
`
`,
`-7 [
`
`I ~ u l
`
`
`
`-9
`
`-8
`
`
`
` -12 -11 -10
`
`
`
`
`
`
`
`
`
`
`
`Exendins on
`
`220
`
`1
`
`10 GLP-l(7-36)
`0 Exendin-4
`+ Exendin(9-39)
`V GLP-l(7-36)
`
`!
`
`GLP-1 Receptor of Islet P-Cells
`90 .
`
`0 Ex 4
`
`0 E X 4 +
`100 nu Ex 9-39
`
`=
`"
`S s
`
`0
`
`R VI o al
`
`VI
`
`8o
`
`70 -
`-
`60
`-
`50
`
`30 40
`
`.c -
`- : 20 -
`- -
`S .- - 10 -
`=
`
`/!/I\{
`0 - €&-€
`
`-
`
`-
`e
`5
`
`c
`
`-10
`
`-9
`
`-8
`
`-7
`
`-6
`
`Peptide log [ M I
`FIG. 3. Effects of GLP-l-(7-36)-amide, exendin-4, exen-
`din-(9-39)-amide, and combinations of GLP-l/exendin-4
`with truncated exendin-(9-39)-amide on the intracellular
`generation of cAMP in RINm5F cells (means 2 S.E. of six
`experiments). At concentrations of 1 and 10 nM, exendin-4 was
`more effective than GLP-1 ( p < 0.05).
`
`0
`
`LC
`
`m
`0 ) -
`2
`.- 2
`v) - c
`
`3
`
`0 OW-1
`m GW-1 +
`0.1 nu Ex 9-39
`A OW-1 +
`100 nu EX 0-30
`
`90
`
`80 -
`
`70 -
`-
`60
`50 -
`
`4 0 -
`30 -
`-
`20
`l a -
`
`0 -
`
`-
`0 -
`
`220 I
`
`200 -
`
`-
`
`n 0
`
`Y-
`
`180 -
`
`160 -
`
`CW-1(7-36)
`Exendin-4
`Q GLF"l(7-36)
`+ Exendin 4
`
`I
`
`I
`I
`I
`-8
`-12
`-11
`-10
`-9
`-7
`FIG. 5. Effect of exendin-4 (Ex 4 ) (upper panel) and GLP-
`1-(7-36)-amide (lower panel) in the absence or presence of
`truncated exendin-(9-39)-amide (Ex 9-39) on glucose-in-
`duced (10 mM) insulin secretion from isolated rat islets (means
`f S.E. of six experiments).
`
`t l o g u l
`
`140
`
`8?
`a
`2
`9 120
`-
`
`-
`
`tn
`
`l o o 80
`
`0.1 nM
`FIG. 4. The combined effect of GLP-1-(7-36)-amide plus
`exendin-4 (0.1 and 1 nM and 1 M) on intracellular cAMP
`generation in RINm5F cells. Given are means f S.E. of six
`experiments.
`
`protein complex identified in experiments with '251-exendin-
`4 showed the same apparent molecular mass as has been
`found with lZ5I-GLP-1-(7-36)-amide, and its
`labeling was
`antagonized by an excess of unlabeled GLP-1 or exendin-4.
`Cross-linking experiments with lung membranes had an
`identical outcome (data not shown). However, the apparent
`molecular mass of the binding sites in lung for exendin-4 and
`GLP-l-(7-36)-amide was 55,000 Da. This difference in mo-
`lecular masses between the GLP-1 receptor of RINm5F cells
`and lung has been described before (21).
`Determination of CAMP Concentrations-Exendin-4
`seemed slightly more efficient in stimulating cAMP produc-
`tion in RINm5F cells than GLP-l-(7-36)-amide (Fig. 3). The
`action of both, exendin-4 and GLP-l-(7-36)-amide, was in-
`
`hibited by exendin-(9-39)-amide (1 p ~ ) . Truncated exendin-
`
`(9-39)-amide alone did not alter the cAMP levels (Fig. 3). If
`GLP-l-(7-36)-amide and exendin-(9-39)-amide were used in
`
`equimolar concentrations (1 p ~ ) , exendin-(9-39)-amide could
`not block the effect of GLP-l-(7-36)-amide on cAMP gener-
`ation (Fig. 3). The combination of GLP-l-(7-36)-amide and
`exendin-4 showed an additive effect on cAMP levels when
`compared with the effect of each peptide alone (Fig. 4).
`Insulin Release Stimulated by GLP-l-(7-36)-Amide and Ex-
`endin-4"Insulin secretion in the presence of 10 mM glucose
`and in the absence of GLP-1-(7-36)-amide or exendin-4 (con-
`trols) was 154 +- 31 microunits/pg of protein/h. (Insulin
`secretion at 2.8 mM glucose was 38 k 18 units/g of protein/
`h.) In normal islets, both peptides, exendin-4 and GLP-1-(7-
`36)-amide, dose dependently enhanced the glucose-induced
`insulin release (Fig. 5 ) . The stimulatory effect of exendin-4
`became significant (p < 0.05) at a concentration of 10"'
`M
`and was strongly inhibited by
`M exendin-(9-39)-amide
`over the whole range of exendin-4 concentrations tested (Fig.
`5, upper panel).
`Fig. 5 (lower panel) demonstrates
`that exendin-(9-39)-
`amide dose dependently reduced the insulin stimulatory effect
`of GLP-l-(7-36)-amide.
`Combinations of submaximal concentrations of GLP-1 and
`exendin-4 (10"' M, respectively) showed, as in the cAMP
`experiments (Fig. 4), an additive stimulatory effect on insulin
`secretion (data not shown). At low glucose levels (2.8 mM
`glucose), GLP-1 and exendin-4 left the insulin release unal-
`tered.
`Effect on Proinsulin Gene Promotor Activity-PTC-1 cells
`
`SANOFI-AVENTIS Exhibit 1018 - Page 19653
`
`IPR for Patent No. 8,951,962
`
`

`
`19654
`
`200
`
`Exendin 4
`GLP 1
`
`T
`
`Exendins on GLP-1 Receptor of Islet &Cells
`
`DISCUSSION
`Gut endocrinology faces the dilemma that the classic ap-
`proach of endocrinology to evaluate the significance of an
`endocrine organ or system simply by its removal is not pos-
`sible. The GLP-1 generating and releasing L-cells belong to
`the diffuse endocrine system of the gut (28). Although it was
`possible to collect a wealth of data concerning the GLP-1
`action at its target cells (3-5,29), the assumption of a possible
`importance of this gut hormone within the enteroinsular axis
`is mainly based on indirect evidence. A direct proof would be
`possible when a selective blockade or induction of the GLP-1
`action could be facilitated. Our data now offer the possibility
`to employ exendin-(9-39)-amide as a potent antagonist and
`exendin-4 as a efficient agonist for such studies.
`Exendin-4 is a 39-amino acid peptide that has been purified
`from H. suspecturn venom (14). It shares 53% structure ho-
`mology with GLP-1-(7-36)-amide. Furthermore, a previous
`report in this journal presented the concept that GLP-1 binds
`at exendin receptors on pancreatic acinar cells (16). There-
`fore, we studied whether the exendins and GLP-1 share the
`same receptor on endocrine pancreatic P-cells. It is probably
`not more than a semantic problem whether such a putative
`common receptor is primarily addressed as GLP-1 or, instead,
`as an exendin receptor. However, GLP-1 is, at least for the
`moment, the single candidate as the endogenous ligand for
`this receptor. The association of exendin and GLP-1 is remi-
`niscent of other examples of non-mammalian peptides that
`have mammalian homologues. For example, the amphibian
`peptides bombesin and cerulein have been matched with the
`mammalian gastrin-releasing peptide and cholecystokinin,
`respectively. As for these other examples, the shared biological
`properties of GLP-1 and exendin-4 are probably based upon
`their homologues’ primary structures.
`Our data indicate that exendin-4 and GLP-l-(7-36)-amide
`compete for the same binding site on RINm5F cells. Exendin-
`4 even binds with higher affinity to RINm5F cells than GLP-
`1-(7-36)-amide. Binding of both peptides was inhibited by
`exendin (9-39). Cross-linking of the binding proteins with
`radioactive labeled exendin-4 and GLP-l-(7-36)-amide
`re-
`vealed the same molecular mass for both ligand-protein com-
`plexes. Receptor binding of exendin-4 and GLP-1-(7-36)-
`amide appears to be associated with stimulation of CAMP
`production and potentiation of the glucose-induced insulin
`release. Truncated exendin-(9-39)-amide antagonized each of
`these biological effects of both peptides. Furthermore, exen-
`din-4, like GLP-1, stimulated the proinsulin gene promotor
`activity, whereas exendin-(9-39)-amide reduced this action.
`However, after repeated and prolonged administration over
`24 h, truncated exendin-(9-39)-amide seems to exhibit a
`partial agonistic activity on proinsulin promotor activation.
`In this context, the use of [Y3’]exendin-4 deserves also a
`further comment. In preliminary experiments, we and others
`(13) have found that the addition of tyrosine at the position
`39 does not significantly alter the effect of exendin-4 on
`intracellular CAMP. On the other hand, it appears that the
`10-fold increases in the Kd shown for GLP-1 and exendin-4
`in RIN cells when radiolabeled exendin-4 was used as tracer
`(as compared with when labeled GLP-1 was utilized as tracer)
`could indicate some complexity in the use of the labeled
`analogue in binding studies, which suggests a careful inter-
`pretation of the respective Kd values.
`Previously, we identified GLP-l-(7-36)-amide receptors on
`rat lung membranes. Our data now show that exendin-4 and
`GLP-l-(7-36)-amide, similar to the results of the experiments
`with RINm5F cells, compete for the same binding site on lung
`membranes and that exendin-4 binds with higher affinity.
`
`7
`
`8
`
`9
`
`7
`
`8
`
`9
`
`
`
`T
`
`T
`
`21
`0 c
`
`100
`
`~
`
`~~
`
`-
`EX-4
`Ex-4
`GLP-1
`GLP-1
`+
`+
`+
`-
`-
`Ex-(8-38)
`FIG. 6. Upper panel, induction of proinsulin gene promotor-driven
`chloramphenicol acetyltransferase activity by exendin-4 (hatched
`bars on the left) and GLP-l-(7-36)-amide (crossed burs on the right).
`Whereas exendin-4 exerted a significant stimulation ( p < 0.05) over
`the whole range of tested concentrations, the GLP-1-induced effect
`was not significant at 1 nM. Lower panel, truncated exendin-(9-39)-
`amide (1 pM) reduced the GLP-1-induced (10 nM) or exendin-4-
`induced (10 nM) increase of the promotor activity. Given are means
`k S.E. of four experiments.
`
`were transiently transfected with a plasmid containing 410
`base pairs of the rat insulin I gene promotor fused to the
`transcriptional reporter gene encoding the bacterial enzyme
`chloramphenicol acetyltransferase. After transfection, cells
`were repeatedly treated for 6,10, and 24 h, and both peptides,
`GLP-1-(7-36)-amide and exendin-4, induced chlorampheni-
`col acetyltransferase activities in a dose-dependent manner
`(Fig. 6, upper panel). Tested in the same concentration range,
`exendin-4 grossly appeared to be equipotent to GLP-1. The
`stimulatory effect of both peptides on proinsulin promotor
`activity was maximal at 10 nM. The lower panel shows that
`the additional introduction
`of truncated exendin-(9-39)-
`amide (1 PM) in combination with GLP-1 or exendin-4 inhib-
`ited but not fully abolished both the GLP-1-induced (10 nM)
`and exendin-4-induced (10 nM) increase of the proinsulin
`promotor activity. It seems that the repeated and prolonged
`(24 h) administration of truncated exendin-(9-39)-amide ex-
`hibited some partial agonistic activity against proinsulin pro-
`motor activation.
`
`SANOFI-AVENTIS Exhibit 1018 - Page 19654
`
`IPR for Patent No. 8,951,962
`
`

`
`Exendins on GLP-1 Receptor of Islet P-Cells
`19655
`3. Goke, R., Fehmann, H. C., and Goke, B. (1991) Eur. J. Clin. Invest. 2 1 ,
`Vasoactive intestinal peptide and peptide histidine isoleucine
`135-144
`are able to displace 1251-GLP-1-(7-36)-amide from its binding
`4. Fehmann, H. C., Goke, R., and Goke, B. (1992) Mol. Cell. Endocrinol. 8 5 ,
`c39-c44
`sites to rat lung membranes (8) but do not compete with "'1-
`5. Fehmann, H. C., and Habener, J. F. (1992) Trends Endocrinol. Metab. 3 ,
`exendin-4 for binding to lung membranes. This may be a
`158-163
`6. Goke, R., and Conlon, J. M. (1988) J. Endocrid 116,357-362
`further reflection of the high affinity binding of exendin-4 to
`7. Goke, R., Cole, T., and Conlon, J. M. (1989) J. Mol. Endocrinol. 2 , 93-98
`8. Goke, R., Oltmer, B., Shelk, S. P., and Goke, B. (1992) FEBS Lett. 3 0 0 ,
`GLP-1-(7-36)-amide receptors in this tissue.
`232-236
`Our data now raise the possibility of defining the exact
`9. Thorens, B. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,8641-8645
`10. Fehmann, H. C., Goke, R., Goke, B., Eissele, R., and Arnold, R. (1991) Int.
`physiologic role for GLP-1 utilizing a specific GLP-l-antag-
`J. Pancreatol. 8,289-303
`onistic compound (exendin-(9-39)-amide). It now seems fea-
`11. Parker, D. S., Raufman, J. P., ODonohue, T. L., Bledsoe, M., Yoshida, H.,
`and Pisano, J. J. (1984) J. Bml. Chem. 259,11751-11755
`sible to quantify GLP-1's contribution to the incretin effect.
`12. Vandermeers, A., Vandermeers-Plret, M. C., Robberecht, P., Waelbrock,
`M., Dehaye, J. P., Winand, J., and Christophe, J. (1984) FEBS Lett.
`Recent clinical studies have focused mainly on a possible
`1 an 3 7 7 - 3 7 ~
`"V) 1 I " 1 I"
`therapeutic significance of GLP-1 in the treatment of non-
`13. Eng, J., Andrews, P. C., Kleinman, W. A., Singh, L., and Raufman, J.-P.
`(1990) J. Biol. Chem. 265,20259-20262
`insulin-dependent diabetes mellitus (type I1 diabetes) (30,
`14. En , J , Kleinman, W. A., Singh, L., Singh, G., and Raufman, J.-P. (1992)
`31). So, future studies need to investigate whether exendin-4
`Bwl. Chem. 267,7402-7405
`15. Raufman, J.-P., Singh, L., and Eng, J. (1991) J. Biol. Chem. 2 6 6 , 2897-
`could be an even more potent insulin-releasing therapeutic
`van')
`16. Raufman, J.-P., Singh, L., Singh, G., and Eng, J. (1992) J. Biol. Chem.
`alternative in diabetes therapy. The latter idea is supported
`267,21432-21437
`18. W a h i m , C. B.,'and Pozzan, f. i19&) J. h o l . Chem. 259,2262-2267
`by a preliminary report, that exendin-4 evoked a greater
`17. En J and Eng C. (1992) Re ul Pe t. 40 142 (abstr.)
`insulin secretory response than GLP-1 when infused in dogs
`19. Richter, G., Goke, R., Goke, B., and Arnold, R. (1990) FEBS Lett. 2 6 7 ,
`7 ~ a n
`(17).
`20. RilEc; G., Goke, R., Goke, B., Schmidt, H., and Arnold, R. (1991) FEBS
`Taken together, our study shows that exendin-4 is a super
`Lett. 280,247-250
`Goke, R., Kolligs, F.? Richter, G., Lankat-Buttgereit, B., and Goke, B.
`21.
`agonist and exendin-(9-39)-amide a potent antagonist at the
`(1993) Am. J. Phys~oL 2 6 4 , L146-Ll52
`Richter, G., Feddersen, 0.. Wagner, U., Barth, P., Goke, R., Goke, B. (1993)
`GLP-1-(7-36)-amide receptor on RINm5F cells and rat lung
`22.
`Praz, G. A., &alb&, 8. A,,, Wollheim, C. B., Blondel, B., Strauss, A. J., and
`Am. J. Ph swl in ress
`membranes. Both peptides are useful tools to further inves-
`23.
`tigate the physiological role of GLP-l-(7-36)-amide. Further-
`Renold A. E. (1983) Bmchem. J. 210,345-352
`24.
`(1976) A d . Bwchem. 72,248-254
`