Purification and Characterization of Insulin, Glucagon, and Two ...

0013-7227/98/$03.00/0 Endocrinology Copyright © 1998 by The Endocrine Society

Vol. 139, No. 8 Printed in U.S.A.

Purification and Characterization of Insulin, Glucagon, and Two Glucagon-Like Peptides with Insulin-Releasing Activity from the Pancreas of the Toad, Bufo marinus* J. MICHAEL CONLON, YASSER H. A. ABDEL-WAHAB, FINBARR P. M. O’HARTE, PER F. NIELSEN, AND JONATHAN WHITTAKER Regulatory Peptide Center (J.M.C.), Department of Biomedical Sciences, Creighton University School of Medicine, Omaha, Nebraska 68178-0405; School of Biomedical Sciences (Y.H.A.A.-W., F.P.M.O.), University of Ulster at Coleraine, Coleraine BT52 1SA, Northern Ireland; Novo Nordisk A/S, Health Care Discovery (P.F.N.), 2880 Bagsvaerd, Denmark; and Hagedorn Research Institute (J.W.), 2820 Gentofte, Denmark ABSTRACT Insulin and four peptides derived from the posttranslational processing of proglucagon have been isolated in pure form from the pancreas of the cane toad, Bufo marinus. Although Bufo insulin contains 9 amino acid substitutions, compared with human insulin, all those residues that are considered to be involved in receptor-binding and in dimer and hexamer formation have been conserved. Bufo insulin was, however, more potent (4-fold) than human insulin in inhibiting the binding of [125I-Tyr-A14] insulin to the soluble fulllength recombinant human insulin receptor, which is probably a consequence of the substitution (Thr 3 His) at position A-8. Bufo glucagon was isolated in two molecular forms: glucagon-29 shows only one amino acid substitution (Thr29 3 Ser), compared with human glucagon; and glucagon-36 comprises glucagon-29, extended from its

T

HE STRUCTURE and expression of the glucagon gene in vertebrates is complex and only partially understood. Nucleotide sequence analysis of cDNAs encoding preproglucagons from several mammalian species have shown that the glucagon is cosynthesized with two structurally related peptides, termed glucagon-like peptide (GLP)-1 and GLP-2, by a single mRNA that is identical in all tissues from that species (1). In contrast, nucleotide sequence analysis of cloned DNA complementary to preproglucagon mRNA from the pancreatic islets of the anglerfish, Lophius americanus (2) and from the chicken pancreas (3) has shown that the preproglucagons from these species do not contain a region corresponding to mammalian GLP-2. It seemed paradoxical, therefore, that both GLP-1 and GLP-2 were isolated, together with glucagon, from an extract of the pancreas of the amphibians, Rana catesbeiana (American bullfrog) (4) and Amphiuma tridactylum (three-toed amphiuma) (5), because amphibians are considered to be phylogenetically more ancient than birds. The situation has been clarified somewhat by the demonstration that cDNAs encoding preproglucaReceived January 29, 1998. Address all correspondence and requests for reprints to: Dr. J. M. Conlon, Department of Biomedical Sciences, Creighton University School of Medicine, Omaha, Nebraska 68178-0405. E-mail: jmconlon@ creighton.edu. * This work was supported by the National Science Foundation (IBN-9418819).

C-terminus by Lys-Arg-Ser-Gly-Gly-Met-Ser. The human proglucagon gene contains one copy of glucagon-like peptide (GLP)-1, a potent insulin secretogogue, and one copy of GLP-2 that is devoid of insulinreleasing activity. In contrast, two proglucagon-derived peptides with 32- and 37-amino acid residues (GLP-32 and GLP-37), displaying greater structural similarity to human GLP-1 than to GLP-2, were isolated from Bufo pancreas. Both peptides produced concentrationdependent increases in insulin release from glucose-responsive rat insulinoma-derived BRIN-BD11 cells. The threshold concentrations producing a significant (P , 0.001) effect were 10-8 M (GLP-32) and 10-9 M (GLP-37), and the maximum increase in the rate of insulin release produced by 10-6 M concentrations of both peptides was approximately 5-fold. (Endocrinology 139: 3442–3448, 1998)

gons isolated from the chicken and trout intestines contain the GLP-2 sequence (6). It was proposed that an alternative RNA splicing mechanism generates a preproglucagon mRNA in the pancreas that lacks the region encoding the GLP-2 sequence, whereas the intestinal precursor encodes both GLP-1 and GLP-2. In mammals, the truncated form of GLP-1 [GLP1(7–36)amide] is the most potent insulinotropic peptide yet discovered (7), and it has been shown that cultured amphibian pancreatic islets also respond to the peptide, with increased release of insulin (8). In contrast, GLP-2 is devoid of significant insulinotropic activity but may play a physiologically important role as an intestinal trophic factor (9). It has recently been shown that a single cloned cDNA encoding preproglucagon from the clawed toad Xenopus laevis, a tetraploid species, contains the sequence of three peptides with structural similarity to GLP-1, in addition to sequences corresponding to mammalian glucagon and GLP-2 (10). Synthetic replicates of the toad GLP-1 peptides stimulated insulin-release from the rat pancreas. At this time, the pathway of posttranslational processing of Xenopus preproglucagon is not known, so that it is unclear which, if any, of the GLP-1 peptides are produced in vivo. The present study describes the purification and structural characterization of insulin and four peptides derived from proglucagon from an extract of the pancreas of a second toad, Bufo marinus, a diploid species. The insulinotropic activity of

3442

Bufo PANCREATIC HORMONES

the two endogenous peptides with structural similarity to mammalian GLP-1 peptides was investigated using a glucose-responsive rat insulinoma-derived cell line BRINBD11 (11). Materials and Methods Tissue extraction Adult toads (Bufo marinus) of both sexes were obtained from a commercial source and killed by pithing. The pancreas was collected from 130 specimens and immediately frozen on dry ice. The tissue (26.7 g, wet wt) was homogenized with ethanol/0.7 m HCl (3:1, vol/vol; 300 ml) using a Waring blender and was stirred for 2 h at 0 C, as previously described (12) After centrifugation (1600 3 g; 30 min; 4 C), ethanol was removed from the supernatant under reduced pressure. Peptide material was isolated from the extract using Sep-Pak C18 cartridges (Waters Associates, Milford, MA), as previously described (12). Bound material was eluted with acetonitrile/water/trifluoroacetic acid (70.0:29.9:0.1) and lyophilized.

RIA Insulin-like immunoreactivity was measured using an antiserum raised against pig insulin (13). Glucagon-like immunoreactivity was measured with an antiserum directed against a site in the COOHterminal region of porcine glucagon (14).

Peptide purification The pancreatic extract, after partial purification on Sep-Pak cartridges, was redissolved in 1 m acetic acid (2 ml) and was chromatographed on a 1.6 3 90-cm column of Sephadex G-25 (Pharmacia Biotech, Uppsala, Sweden), equilibrated with 1 m acetic acid. The column was eluted at a flow rate of 24 ml/h, and fractions (2.0 ml) were collected. Absorbance was measured at 280 nm. The concentrations of insulin-like and glucagon-like immunoreactivities in the fractions were determined at a dilution of 1:30. Immunoreactive fractions were pooled (total vol 5 16 ml) and pumped onto a 1 3 25-cm Vydac 218TP510 C18 reversedphase HPLC column (Separations Group, Hesperia, CA), equilibrated with 0.1% trifluoroacetic acid/water at a flow rate of 2 ml/min. The

FIG. 1. Reversed-phase HPLC on a semipreparative Vydac C18 column of an extract of the pancreas of Bufo marinus after partial purification by gel permeation chromatography. Peak 1 contained insulin and glucagon-36, peak 2 contained glucagon-29, peak 3 contained GLP-32, and peak 4 contained GLP-37. The dashed line shows the concentration of acetonitrile in the eluting solvent and the arrows show where peak collection began and ended. ABS280 is the absorbance at 280 nm.

3443

concentration of acetonitrile in the eluting solvent was raised to 21% over 10 min and to 49% over 60 min using linear gradients. Absorbance was measured at 214 and 280 nm, and fractions (1 min) were collected. Bufo insulin and glucagon-36 (designated: peak 1, in Fig. 1), glucagon-29 (peak 2), GLP-32 (peak 3), and GLP-37 (peak 4) were purified to near-homogeneity, as assessed by a symmetrical peak shape, by successive chromatographies on a 0.46 3 25-cm Vydac 214TP54 C4 column and a 0.46 3 25-cm Vydac 219TP54 phenyl column using the elution conditions summarized in Fig. 2.

Structural analysis Purified Bufo insulin (approximately 2 nmol) was incubated for 3 h at room temperature with dithiothreitol (2 mg) in 0.1 m Tris-HCl-6 m guanidine hydrochloride buffer, pH 7.5 (0.4 ml), under an atmosphere of argon. Cysteine residues were derivatized by addition of 4-vinylpyridine (3 ml), and the pyridylethylated A- and B-chains of insulin were separated on a 0.46 3 25-cm Vydac C4 column, under the conditions used for the purification of intact insulin (Fig. 2A). Bufo glucagon-36 (2 nmol) was incubated with 1-Tosylamide-2-phenylethylchloromethyl ketone-treated trypsin from bovine pancreas (10,000 U/mg, solid) (2 mg) for 3 h at 37 C in 0.2 m ammonium bicarbonate solution, pH 7.8 (200 ml). Tryptic fragments were isolated by chromatography on a 0.46 3 25-cm Vydac 218TP54 (C18) column equilibrated with 0.1% trifluoroacetic acid/water at a flow rate of 1.5 ml/min. The concentration of acetonitrile in the eluting solvent was raised to 42% over 60 min using a linear gradient. Amino acid compositions were determined by precolumn derivatization with phenylisothiocyanate, using an Applied Biosystems model 420A derivatizer (Foster City, CA), followed by separation of the phenylthiocarbamyl amino acids by reversed-phase HPLC. Hydrolysis in 5.7 m hydrochloric acid (24 h at 110 C), of approximately 500 pmol of peptide, was carried out. The primary structures of the peptides were determined by automated Edman degradation, using an Applied Biosystems model 471A sequenator, modified for on-line detection of phenylthiohydantoin amino acids under gradient elution conditions. Standard operating procedures were used (Applied Biosystem model 471A Protein Sequencer User’s Manual), and the detection limit was 1 pmol. Mass spectrometry was performed on a Voyager RP MALDI-TOF instrument (Perspective Biosystems Inc., Framingham, MA) equipped with a nitrogen laser (337 nm). The instrument was operated in linear

3444

Endo • 1998 Vol 139 • No 8


FIG. 2. Purification by reversed-phase HPLC on an analytical Vydac C4 column of Bufo marinus insulin and glucagon-36 (A), glucagon-29 (B), GLP-32 (C), and and GLP-37 (D). The column was eluted at a flow rate of 1.5 ml/min. mode with delayed extraction, and the accelerating voltage in the ion source was 25 kV. The accuracy of the mass determinations was within 0.1%.

Insulin binding studies. Competitive binding studies were carried out using the soluble fulllength recombinant human insulin receptor, expressed in 293EBNA cells (an adenovirus transformed human kidney cell line expressing EBV nuclear antigen) (15). Porcine insulin binds to two population of binding sites in the full-length receptor, with Kd values of 2.8 pm and 0.51 nm. The abilities of Bufo insulin (purity . 98%) and human insulin to inhibit the binding of [3-[125I]iodotyrosine-A14] human insulin (specific radioactivity, 74 TBq/mmol; Novo Nordisk, Bagsvaerd, Denmark) to the soluble form of the insulin receptor were determined using a procedure previously described in detail (15, 16). All determinations were performed in quadruplicate.

Insulin-releasing activity BRIN-BD11 cells (11) were cultured in RPMI-1640 tissue culture medium containing 10% (vol/vol) FCS and 11.1 mm glucose. Cells were maintained in sterile tissue culture flasks at 37 C in an atmosphere of 5% CO2-95% air. For measurement of insulin-release from cell monolayers, the cells were seeded into 24-multiwell plates (Nunc, Roskilde, Denmark) at a density of 2.5 3 105 per well and allowed to attach during overnight culture. The culture medium was replaced by 1.0 ml of buffer

(composition in mm: NaCl 115, KCl 4.7, CaCl2.2H2O 1.28, KH2PO4 1.2, MgSO47H2O 1.2, NaHCO3 10; pH 7.4) containing BSA (1 g/liter) and glucose (1.1 mm), and the cells were incubated for 40 min at 37 C. Test incubations were performed using the same buffer supplemented with 5.6 mm glucose and peptides at appropriate concentrations. After 20 min of incubation, the supernatants were removed from each well, and insulin-like immunoreactivity was measured by RIA. Six independent experiments were performed. Data are expressed as mean 6 sem. Effects on insulin release were analyzed by Student’s paired t test, and differences were considered to be significant at P , 0.05.

Results Peptide purification

The insulin-like immunoreactivity and the glucagon-like immunoreactivity in the extract of Bufo pancreas, after partial purification on Sep-Pak cartridges, were eluted from a Sephadex G-25 gel permeation column in the same fractions as a broad peak, with Kav between 0.2 and 0.4. These fractions were pooled and injected onto a semipreparative Vydac C18 reversed-phase HPLC column (Fig. 1). The major uv-absorbing peak in the chromatogram, designated: peak 1, contained insulin-like immunoreactivity, and peak 2 contained glucagon-like immunoreactivity. The insulin-containing peak was


3445

rechromatographed on an analytical Vydac C4 column (Fig. 2A) and was resolved into two major components. The peak designated I contained insulin, and peak G36 was subsequently shown to contain glucagon extended from its Cterminus by 7 amino acid residues. The glucagon-containing peak was eluted from an analytical Vydac C4 column as the distinct, well-resolved peak, designated G29 (Fig. 2B). Chromatographic analysis of the major peaks, designated 3 and 4 in Fig. 1 on an analytical Vydac C4 column, is shown in Fig. 2, C and D. Peak 3 material, subsequently shown to contain a GLP with 32 amino acid residues (GLP-32), and peak 4 material, subsequently shown to contain a GLP peptide with 37 amino acid residues (GLP-37), were eluted as wellresolved major peaks. The Bufo peptides were purified to near homogeneity, as assessed by symmetrical peak shape, by a final chromatography on analytical Vydac phenyl column. The final yields of pure peptides (determined by amino acid analysis) were: insulin, 21 nmol; glucagon-29, 33 nmol; glucagon-36, 17 nmol; GLP-32, 91 nmol; and GLP-37, 44 nmol. Structural characterization

The primary structures of the pyridylethylated A-chains and B-chains of Bufo insulin and of glucagon-29, glucagon-36, GLP32, and GLP-37 were determined by automated Edman degradation, and the results are shown in Fig. 3. In all cases, the results of amino acid composition analysis of the peptides were consistent with the results of sequence analysis, demonstrating that the full sequences of the peptides had been obtained. The data indicated that the Bufo peptides were more than 98% pure. The amino acid sequence at the COOH-terminus of glucagon36 was established by Edman degradation of the products of digestion with trypsin. Four peptide fragments were identified, corresponding to residues (1–12), (13–17), (19–30), and (32–36). The primary structure of the C-terminal fragment was determined as Ser-Gly-Gly-Met-Ser. The primary structures of the proglucagon-derived peptides were confirmed by mass spectrometry. The observed molecular mass of glucagon-29 was 3468, compared with a calculated mass of 3467; glucagon-36, observed mass 5 4173, calculated mass 5 4170; GLP-32, observed mass 5 3670, calculated mass 5 3668; GLP-37, observed mass 5 4306, calculated mass 5 4303. The mass spectrum of glucagon-36 also gave a prominent signal at 4086, suggesting that the peptide was contaminated with the [des-Ser36] fragment. Receptor-binding properties of Bufo insulin

The abilities of Bufo insulin and human insulin to inhibit binding of 125I-labeled human insulin to the soluble fulllength human insulin receptor are compared in Fig. 4. The concentration of Bufo insulin producing a 50% inhibition of binding was 11 pm (range, 9 –12 pm). The corresponding value for human insulin, in incubations carried out at the same time and under identical conditions, was 42 pm (range, 35– 46 pm). Insulin-releasing activity

In the absence of peptide secretogogues, the rate of insulin-release from BRIN-BD11 cells incubated in

FIG. 3. Amino acid sequences of the A-chain and B-chain of insulin, glucagon-29, glucagon-36, GLP-32, and GLP-37 isolated from the pancreas of Bufo marinus.

medium containing 5.6 mm glucose was 1.20 6 0.10 ng insulin/106 cellsz20 min. The rate of insulin-release was significantly (P , 0.05) greater from cells incubated in medium containing 16.7 mm glucose (2.04 6 0.13 ng insulin/106 cellsz20 min). As shown in Fig. 5, both GLP-32 and GLP-37 produced a concentration-dependent increase in the rate of insulin-release. The minimum concentration producing a significant (P , 0.001) increase in secretion was 10-9 m for GLP-37 and 10-8 m for GLP-32. The maximum increase in rate of insulin-release produced by the highest dose of peptide tested (10-6 m) was the same for both peptides (approximately 5-fold). Discussion

The primary structure of B. marinus insulin is compared with that of human insulin and other known amphibian insulins [the anurans, clawed toad Xenopus laevis (17) and bullfrog, Rana catesbeiana (18); the urodeles, three-toed amphiuma Amphiuma tridactylum (12) and lesser siren Siren intermedia (19); the caecilian Typhlonectes natans (20)] in Fig. 6. Traditionally, the receptor-binding region of human insulin is considered to involve contributions from amino acid residues at positions A1-A3, A5, A19, A21, and B22B24 (21), although more recent data (22) has suggested that

3446


Endo • 1998 Vol 139 • No 8

FIG. 4. Comparison of the abilities of Bufo and human insulins to inhibit the binding of [125I-Tyr-A14] human insulin to the soluble recombinant human insulin receptor. Data are presented as the B/Bo ratio, where B is 125I-labeled insulin specifically bound at the indicated insulin concentration and B0 is the binding in the absence of added insulin. Each point represents mean 6 SEM of four independent experiments, with Bufo (Œ—Œ) and human (f—f) insulins.

the role of residue A21 may be relatively minor. All these residues have been conserved in Bufo insulin. Similarly, those residues in human insulin involved in dimer formation (B12, B16, B20, B24, B26, and B28) and hexamer formation (B6, B10, B14, B17, B18, A13, A14) (20) have also been fully conserved in Bufo insulin. Bufo insulin shares with insulins from Xenopus (17) and the amphiuma (12) the presence of a histidine residue at position A8. This residue, by forming stabilizing structural motifs in the insulin molecule that are of critical importance for receptor recognition, is probably responsible for the observed increase in binding affinity of these insulins for mammalian insulin receptors (23). The pathway of posttranslational processing of preproglucagon in mammals is relatively well understood. In the pig pancreas, for example, the precursor is processed predominantly to an N-terminal flanking peptide (termed glicentin-related pancreatic peptide), glucagon, an intervening hexapeptide sequence and an unprocessed peptide containing both the GLP-1 and GLP-2 sequences. In the pig intestine, the molecular form of GLP-1 with insulinotropic activity [GLP-1(7–37) and/or GLP-1(7–36)amide] is generated by the combined action of a prohormone convertase, cleaving at dibasic residue sites, and a protease cleaving at a single arginyl residue to remove an N-terminal leader sequence (24). The present data provide further support for the conclusion (5, 25) that the primary structure of glucagon has been very strongly conserved among tetrapods. As shown in Fig. 6, Bufo glucagon is identical to glucagon from the bullfrog R. catesbeiana (4) and shows only one amino acid substitution

FIG. 5. Effects of Bufo GLP-32 (A) and Bufo GLP-37 (B) on insulin release from BRIN-BD11 rat insulinoma-derived cells, incubated at 5.6 mM glucose concentration. After a 40-min preincubation, cells were incubated with peptides at the concentrations (conc) shown for 20 min. Values are means 6 SEM for six independent experiments. *, P , 0.001, compared with basal (5.6 mM glucose only) release.

(Thr29 3 Ser), compared with glucagons from the human and from Xenopus (10). The C-terminally extended form of Bufo glucagon contains only one substitution (Ile35 3 Met), compared with the corresponding peptide isolated from bullfrog pancreas (4). A C-terminally extended form of glucagon with 37 amino acid residues, often referred to as oxyntomodulin, is present in variable amounts in the intestines and pancreata of mammals (26). In the pancreas of B. marinus, as in the pancreata of the bullfrog (4) and the amphiuma (5), GLP-1 is stored in the mature, fully processed form that corresponds to pig intestinal GLP-1(7–37). An N-terminally extended form of GLP-1 was probably not present in the extract of Bufo pancreas in major abundance, but its presence as a minor component remains a possibility. The most unexpected feature of the present study is the fact that two peptides with structural similarity to human GLP-1(7–37) were isolated from Bufo pancreas, whereas a peptide with structural similarity to GLP-2 was purified from bullfrog and amphiuma pancreas, in addition to a single GLP-1-related peptide. As shown in Fig. 7, Bufo GLP-32 shows close structural similarity with bullfrog GLP-1 (three amino acid substitutions) (4), whereas


3447

FIG. 7. A comparison of the primary structures of glucagon and the GLPs isolated from Bufo marinus pancreas with the corresponding peptides from other amphibian species and with human GLP-1 and GLP-2. (2) denotes residue identity. FIG. 6. A comparison of the primary structure of insulins from Bufo marinus with insulins from other amphibian species and with human insulin. (2) denotes residue identity.

Bufo GLP-37 resembles most strongly the predicted amino acid sequence of GLP-1A, identified in Xenopus proglucagon (10). A structure-activity study, in which each amino acid in human GLP-1(7–36)amide was replaced by alanine, has identified residues 7, 10, 12, 13, 15, 28, and 29 as being important for high affinity binding by cells expressing the rat pancreatic GLP-1 receptor (27, 28). Bufo GLP-32 contains one substitution (Ile 3 Val at the position corresponding to residue 29) and GLP-37 contains one substitution (Phe3 Tyr at the position corresponding to residue 12) among these critical residues. The fact that GLP-32 is significantly less potent than GLP-37 suggests that the substitution at position 29 may have a deleterious effect on the insulinotropic action of the peptides. Synthetic replicates of three peptides with structural similarity to GLP-1, identified in the proglucagon sequence of Xenopus, produced an 8- to 10-fold greater increase in insulin release from the perfused rat pancreas, compared with the release produced by 16 mm glucose alone (10). A peptide with appreciable structural similarity to either a mammalian GLP-2 or to either of the three known amphibian GLP-2 peptides (Fig. 7) was not identified in the extract of Bufo pancreas. In the absence of nucleotide sequence data of cDNAs or genomic fragments, it is unclear whether Bufo GLP-32 and GLP-37 are the products of posttranslational of two separate preproglucagons, both of which contain an identical glucagon sequence but lack a GLP-2 sequence, or are derived from a single preproglucagon containing a single copy of glucagon, GLP-32, and GLP-37. The fact that glucagon was isolated in appreciably lower yield than GLP-32 does not favor the former hypothesis.

Acknowledgments The authors thank Drs. E. Burcher and F. Warner, University of New South Wales, Sydney, Australia, for a gift of Bufo pancreas; and Professor P. F. Flatt, University of Ulster, Northern Ireland, for use of BRIN-BD11 cells.

References 1. Novak U, Wilkes A, Buell G, McEwen S 1987 Identical mRNA for proglucagon in pancreas and gut. Eur J Biochem 164:553–558 2. Lund PK, Goodman RH, Dee PC, Habener JF 1982 Pancreatic preproglucagon cDNA contains two glucagon-related coding sequences arranged in tandem. Proc Natl Acad Sci USA 79:345–349 3. Hasegawa S, Terazono K, Nata K, Takada T, Yamamoto H, Okamoto H 1990 Nucleotide sequence determination of chicken glucagon precursor cDNA. Chicken preproglucagon does not contain glucagon-like peptide II. FEBS Lett 264:117–120 4. Pollock HG, Hamilton JW, Rouse JB, Ebner KE, Rawitch AB 1988 Isolation of peptide hormones from the pancreas of the bullfrog (Rana catesbeiana). J Biol Chem 263:9746 –9751 5. Cavanaugh ES, Nielsen PF, Conlon JM 1996 Isolation and structural characterization of proglucagon-derived peptides, pancreatic polypeptide and somatostatin from the urodele, Amphiuma tridactylum. Gen Comp Endocrinol 101:12–20 6. Irwin DM, Wong J 1995 Trout and chicken proglucagon: alternative splicing generates mRNA transcripts encoding glucagon-like peptide 2. Mol Endocrinol 9:267–277 7. Mojsov S, Weir G, Habener JF 1987 Insulinotropin: glucagon-like peptide 1 (7–37) co-encoded in the glucagon gene is a potent stimulator of insulin release in the perfused rat pancreas. J Clin Invest 79:616 – 619 8. Mommsen TP, Moon TW 1994 Glucagon-like peptides: structure-function relationships and evolution. In: Davey KG, Peter RE, Tobe S (eds) Perspectives in Comparative Endocrinology. National Science Research Council, Ottawa, pp 493– 498 9. Drucker DJ, Ehrlich P, Asa SL, Brubaker PL 1996 Induction of intestinal epithelial proliferation by glucagon-like peptide 2. Proc Natl Acad Sci USA 93:7911–7916 10. Irwin DM, Satkunarajah M, Wen Y, Brubaker PL, Pederson RA, Wheeler MB 1997 The Xenopus proglucagon gene encodes novel GLP-1-like peptides with insulinotropic properties. Proc Natl Acad Sci USA 94:7915–7920 11. McClenaghan NH, Barnett CR, Ah-Sing E, Abdel-Wahab YHA, O’Harte FPM, Yoon TW, Swanson-Flatt SK, Flatt PR 1996 Characterization of a novel glucose-responsive insulin-secreting cell line BRIN-BD11, produced by electrofusion. Diabetes 45:1132–1140 12. Conlon JM, Cavanaugh ES, Mynarcik DC, Whittaker J 1996 Characterization

3448

13. 14. 15. 16.

17. 18. 19. 20.


of an insulin from the amphiuma (Amphibia: Urodela) with an N-terminally extended A-chain and high receptor-binding affinity. Biochem J 313:283–287 Flatt PR, Bailey CJ 1981 Abnormal plasma glucose and insulin responses in heterozygous (ob/1) mice. Diabetologia 20:573–577 Conlon JM, Thim L 1985 Primary structure of glucagon from an elasmobranchian fish, Torpedo marmorata. Gen Comp Endocrinol 60:398 – 405 Williams PF, Mynarcik DC, Yu GQ, Whittaker J 1995 Mapping of an NH2terminal ligand binding site of the insulin receptor by alanine scanning mutagenesis. J Biol Chem 270:3012–3016 Mynarcik DC, Williams PF, Schaffer L, Yu GQ, Whittaker J 1997 Analog binding properties of insulin receptor mutants. Identification of amino acids interacting with the COOH terminus of the B-chain of the insulin molecule. J Biol Chem 272:2077–2081 Shulinder AR, Bennett C, Robinson EA, Roth J 1989 Isolation and characterization of two different insulins from an amphibian, Xenopus laevis. Endocrinology 125:469 – 477 Conlon JM, Yano K, Chartrel N, Vaudry H, Storey KB, Freeze tolerance in the wood frog Rana sylvatica is associated with unusual structural features in insulin but not in glucagon. J Mol Endocrinol, in press Conlon JM, Trauth SE, Sever DM 1997 Purification and structural characterization of insulin from the lesser siren, Siren intermedia (Amphibia: Caudata). Gen Comp Endocrinol 106:295–300 Conlon JM, Hilscher-Conklin C, Boyd SK 1995 Purification and structural characterization of insulin from a caecilian, Typhlonectes natans (Amphibia: Gymnophiona). Peptides 16:1385–1388

Endo • 1998 Vol 139 • No 8

21. Baker EN, Blundell TL, Cutfield JF, Cutfield SM, Dodson EJ, Dodson GG, Hodgkin DMC, Hubbard RE, Isacs NW, Reynolds CD, Sakabe K, Sakabe N, Vijayan NM 1988 The structure of the 2 Zn pig insulin crystal at 1.5 A resolution. Philos Trans R Soc Lond Biol 319:369 – 456 22. Kristensen C, Kjeldsen T, Wiberg FC, Schaffer L, Hach M, Havelund S, Bass J, Steiner DF, Andersen AS 1997 Alanine scanning mutagenesis of insulin. J Biol Chem 272:12978 –12983 23. Muggeo M, Ginsberg BH, Roth J, Neville DM, De Meyts P, Kahn CR 1979 The insulin receptor in vertebrates is functionally more conserved during evolution than insulin itself. Endocrinology 104:1393–1402 24. Orskov C, Holst JJ, Knuhtsen S, Baldissera FGA, Poulsen SS, Vagn Nielsen O 1986 Glucagon-like peptides GLP-1 and GLP-2, predicted products of the glucagon gene, are secreted separately from pig small intestine but not pancreas. Endocrinology 119:1467–1475 25. Plisetskaya EM, Mommsen TP 1996 Glucagon and glucagon-like peptides in fishes. Int Rev Cytol 168:187–257 26. Kervan A, Blanche P, Bataille D 1987 Distribution of oxyntomodulin and glucagon in the gastrointestinal tract and the plasma of the rat. Endocrinology 121:704 –713 27. Adelhorst K, Hedegaard BB, Knudsen LB, Kirk O 1994 Structure-activity studies of glucagon-like peptide-1. J Biol Chem 269:6275– 6278 28. Gallwitz B, Witt M, Paetzold G, Morys-Wortmann C, Zimmermann B, Eckart K, Fo¨lsch UR, Schmidt WE 1994 Structure/activity characterization of glucagon-like peptide-1. Eur J Biochem 225:1151–1156