The structure of human preproglucagon, as deduced from nucleotide sequencing of the preproglucagon gene, contains two glucagon-like peptides (GLP-1 and GLP-2) in the portion C-terminal to glucagon. A rabbit antiserum was raised against synthetic GLP-1-(1-19) which had 20% cross-reactivity with synthetic GLP-1 and des-Gly37-GLP-1 amide, two possible forms of the GLP-1 whole molecule, but no significant cross-reactivity with glucagon or other pancreatic peptides. Immunocytochemistry revealed that the distribution of GLP-1-(1-19) immunoreactivity followed that of glucagon-like immunoreactivity in the normal human pancreas and in two human glucagon-secreting pancreatic tumors. Chromatography of human pancreas extracts on Sephadex G-50 gave peaks of cross-reactivity at Kav values of 0.06-0.16, 0.34-0.39, 0.54-0.58 (the elution position of synthetic GLP-1), and 0.64-0.70. The concentration of immunoreactivity in the Kav 0.54-0.58 peak measured by RIA using GLP-1 or des-Gly37-GLP-1 amide as standard was 94 +/- 7 pmol/g (mean +/- SEM), while the total pancreatic glucagon content was 4.8 +/- 0.8 nmol/g. One extract of a human glucagon-secreting pancreatic tumor contained a prominent peak of GLP-1-(1-19) peptide cross-reactivity with properties identical to those of GLP-1 or des-Gly37-GLP-1 amide on gel filtration and reverse phase high pressure liquid chromatography, but another tumor contained a preponderance of cross-reactive forms of greater molecular size. Pretreatment plasma from three patients with radiological and biochemical evidence of glucagon-secreting tumors contained a peak of cross-reactivity with the chromatographic properties of intact GLP-1. The low concentrations of intact GLP-1 in normal pancreas compared with pancreatic glucagon concentrations suggest that the majority of the proglucagon is cleaved in a manner that does not produce GLP-1, as defined by its delimiting pairs of basic amino acid residues.

Molecular forms of the glucagon‐like peptides (GLP) encoded by the human preproglucagon gene were analysed by chromatography combined with specific radioimmunoassays to the synthetic peptides. Whereas extracts of human pancreas and a glucagonoma contained a large proglucagon cleavage product possessing both GLP‐1 and GLP‐2 immunoreactivities, extracts of human intestine contained products corresponding to free GLP‐1 and a small amount ofchromatographically distinct GLP‐2 immunoreactivity. It is concluded that post‐translational processing of proglucagon differs in pancreas and intestine, so that the C‐terminal portion of the molecule is cleaved to liberate free GLP‐1 in the intestine. Further processing or degradation results in loss especially of GLP‐2 immunoreactivity.

Bombesin-like immunoreactivity has been measured in pancreatic tissues of man (12.4 +/- 1.2 pmol/g), pig (15.8 +/- 3.2), calf (4.3 +/- 0.9), rat (8.5 +/- 1.2) and guinea-pig (2.8 +/- 0.6) by a specific radioimmunoassay. Gel filtration of the pancreatic extracts revealed 2 major immunoreactive peaks: the earlier peak was eluted in the position of porcine gastrin-releasing peptide, and the later peak was eluted just after the amphibian bombesin standard. Immunocytochemistry demonstrated the presence of bombesin-like immunoreactivity in nerves in the rat pancreas, particularly in the exocrine pancreas, and occasionally in the peri-insular spaces. Isolated rat pancreatic islets were found to contain small quantities of bombesin-like immunoreactivity (0.037 +/- 0.003 fmol/islet) suggesting that mammalian bombesin-like peptides may be involved in the regulation of endocrine as well as exocrine pancreatic secretion.

Glucagon-like peptide-1 does not have specific, high-affinity receptors on rat liver membranes, does not displace glucagon from glucagon receptors on these membranes and does not stimulate the production of cyclic AMP by isolated rat hepatocytes. In the presence of glucagon, high concentrations of glucagon-like peptide-1 do not significantly alter the production of cyclic AMP. Thus, glucagon-like peptide-1 appears unlikely to have a direct action on hepatic carbohydrate metabolism.

Summary. Although glucagon-like peptide-1 has the appearance of a glucagon-homologue that may be co-secreted with glucagon, synthetic glucagon-like peptide-l-(1-37) does not significantly affect plasma glucose and insulin concentrations when administered at high doses (100 and 400 ~tg) to cortisone-pretreated rabbits. This synthetic preparation thus lacks the primary metabolic effect of glucagon at the doses tested. An intra-or extra-pancreatic role of glucagon-like peptide-1 has yet to be discovered. Key words:Glucagon-like peptide-1, glucose, insulin, glucagon, bioassay.The nucleotide sequences of cDNA derived from hamster [1] and ox [2] glucagon mRNA and the human preproglucagon gene [3] show that pre-proglucagon contains two further glucagon-like peptides (GLP-1 and -2) C-terminal to the glucagon sequence. The amino-acid sequence of GLP-1 is completely conserved between the three mammalian species studied and shows a high degree of homology with similar deduced amino-acid sequences in angler-fish pre-proglucagons [4]. It is defined by delimiting pairs of basic amino-acid residues, that form potential cleavage points. GLP-1 thus has the appearance of a highly conserved biologically active peptide that may be co-secreted with glucagon, and it is possible that it has a modulating role in carbohydrate metabolism. In the first instance we have tested synthetic GLP-1 for its effect on plasma glucose and insulin concentrations in rabbits. Material and methodsSynthetic GLP-l-(1-37) was synthesized by a solid phase method [5] (Bachem, Torrance, CA, USA). It was tested for its effect on plasma glucose and insulin by a modification of the twin cross-over bioassay for glucagon [6], using crystalline glucagon (Novo Industri, Copenhagen, Denmark) as a standard. Twelve rabbits were injected (25 mg subcutaneously) on day 0 with cortisone acetate injection (British Pharmacopoeia; Cortistab, The Boots Company, Nottingham, UK). The assay was performed on days 2 and 3, the rabbits being deprived of food for 18 h before each part of the assay, but with free access to water. The peptides were dissolved immediately before each experiment in 1.6% (vol/vol) aqueous glycerine containing 0.2% (wt/vol) phenol, adjusted to pH 3 with HC1.The rabbits were randomized into four groups of three each, and they received a subcutaneous injection of 1 ml of diluent, as a control, followed after 60 min by either GLP-1 (400 gg or 100 gg) or glucagon (24 .ag or 6 ~xg) in the same volume. Blood samples (0.3 ml) were taken from a marginal ear vein at times -40, 0, 20 and 60 rain in relation to both the diluent and peptide injections. The next day, the experiment was repeated with a twin cross-over, so that rabbits that had received the higher dose of glucagon now received the lower dose of GLP-1. The blood samples were collected into fluoride-oxalate centrifuge tubes, containing dried aprotinin (200 Kallikrein Inhibitor Units; Trasylol, Bayer, Wuppertal, FRG), and centrifuged immediately at 1600 g for 10 min. The plasma was frozen on solid CO2 and st...