Search

BIOSYNTHETIC studies with human islet cell tumours1,2, rat2,3 and codfish4 islets, and foetal calf pancreas5 have shown that insulin is formed from proinsulin, a single-chain polypeptide precursor. This protein has subsequently been purified from commercial preparations of crystalline insulin6, and the amino-acid sequences of porcine7 and bovine8 proinsulin have been determined. The molecule begins at the N-terminus with the B-chain sequence of insulin, terminates with the insulin A-chain and bears a connecting segment of thirty (bovine) or thirty-three (porcine) amino-acids linking the two chains. Although 30 per cent of the amino-acid sequences of these connecting segments differ, identical pairs of basic amino-acids occur at each end in both species. Limited tryptic digestion in each case liberates desalanyl–insulin, alanyl–arginine, free arginine and the remainder of the connecting segment as an intact peptide6,7.

INSULIN is a well defined globular protein made up of two small peptide chains linked together through disulphide bonds1,2. It contains an unusually high proportion of hydrophobic residues and thus tends to readily self-associate3,4. In many species the hormone forms dimers in solution and, in the presence of Zn2,4ions, hexamers4,5. Recently, Hodgkin and coworkers6,7 have determined the crystal structure of the rhombohedral form of porcine Zn-insulin at 1.9 A resolution6,7. Two molecules of insulin occur in the asymmetric unit, related through an approximate twofold axis to form a nearly symmetrical dimer. Three such dimers in turn associate through coordination of histidine residues B10 with two zinc ions to form a nearly spherical hexamer. In spite of the availability of the amino acid sequences of a wide variety of vertebrate insulins2,8, however, comparative studies on the detailed structure and modes of association of vertebrate insulins have been lacking.

PROINSULIN is a single-chain polypeptide in which the A and B chains of insulin are linked together by the connecting peptide1. The terminals of the connecting peptides are occupied by the basic dipeptides, Arg–Arg and Lys–Arg, respectively, and proinsulin C-peptide is defined as connecting peptide minus the basic dipeptides. The immunological reactivity of synthetic [59-formyllysine]-bovine proinsulin 31–60, which comprises the entire sequence of the connecting peptide segment of bovine proinsulin, is indistinguishable from that of natural bovine proinsulin when compared on an equimolar basis2. The immunoassay system involved a bovine proinsulin antiserum3, which was purified by the removal of the antibodies directed towards the insulin region of the molecule. We have now evaluated the structural features of bovine connecting peptide segment which determine its reactivity with bovine proinsulin antisera. For this purpose, homogeneous synthetic peptides of defined structures serve as excellent substrates. We synthesised twelve peptides related to bovine proinsulin C-peptide which has the following amino acid sequence

We have previously identified a structurally abnormal insulin in the serum and pancreas of a middle-aged man with diabetes mellitus which arose from a leucine for phenylalanine substitution at position 24 or 25 of the insulin B chain; further analysis of the patient's leukocyte DNA showed that one of the patient's insulin alleles had undergone mutation resulting in loss of an MboII restriction site normally present in the human insulin gene. Two additional and unrelated patients with the same clinical syndrome have now been identified (ref. 4 and unpublished results). All of these patients showed hyperglycaemia typical of diabetes and with marked hyperinsulinaemia typical of insulin resistance, but all three show normal tolerance to exogenously administered insulin. As the opportunity of examining pancreatic tissue from patients suspected of secreting insulin variants is rare, we have developed a method combining HPLC and radioimmunoassay to identify insulin variants isolated from human sera. By this method we have shown that all three patients noted above secrete structurally variant and chemically distinct insulins. In correction of our original assignment, one is identified as [LeuB25]insulin.

Immunoprecipitation and tryptic peptide analysis of newly synthesized proteins from rat islets have identified an 18,000 molecular weight (MW) protein as proglucagon. Conversion of this precursor was kinetically similar to the conversion of proinsulin and resulted in the formation of both pancreatic glucagon and a 10,000-MW protein lacking this hormonal sequence.