Immobilized insulin for high capacity affinity chromatography of insulin receptors.

Insulin receptors can be purified by affinity chromatography on immobilized insulin, but published methods all suffer from a rather low capacity of the affinity columns. By using insulin that has been protected in positions A1 and B29, we have been able to couple the insulin selectively through the B1 amino group to divinyl sulfone-activated agarose. The N terminus of the B-chain is the most innocuous site as far as receptor-insulin interaction is concerned, and this strategy allowed us to make affinity columns with capacities of several milligrams of receptor/ml of resin. The receptor used was the soluble ectodomain of the human insulin receptor, produced in transfected baby hamster kidney cells. The column preparation and the elution conditions are described in detail, as the efficacy of the purification depends strongly on both. The purity of the eluted receptors was so high that quantitative amino acid analysis fitted with theory. The molar absorption coefficient at 278.5 nm was 296,000 M-1 cm-1. Finally, it could be unequivocally established that the soluble receptor binds two molecules of insulin with equal affinity.

N-(9-Xanthenyl)amino acid N-carboxyanhydrides in the solid-phase synthesis of a Lys-Lys-linked cyclic leucine-enkephalin.

The side-chain-protected heptapeptide Phe-Leu-Lys(F3Ac)-Lys(F3Ac)-Tyr(Cac)-Gly-Gly is assembled by stepwise chain elongation on a polystyrene resin, using inter alia N-(9-xanthenyl)-amino acid N-carboxyanhydrides. After HBr cleavage from the resin, cyclization via carbodiimide/N-hydroxysuccinimide is carried out in a yield of 24%. Cleavage of the 9-carbazolylcarbonyl group by hydrazinolysis, and alkaline hydrolysis of the trifluoroacetyl groups, affords cyclo (-Tyr-Gly-Gly-Phe-Leu-Lys-Lys-) in an over-all yield of 12%. The structure of the peptides is confirmed by mass spectrometry.

The preparation of porcine (Tyr-A14-2H) insulin and its characterization by mass spectrometry.

The title compound was prepared by catalytic deiodination of (Tyr-A14-3-I) insulin with deuterium gas. Under special conditions (large excess of PdO in CH3OD/D2O at pH 9.2) developed to minimize problems due to poisoning of the catalyst, deuterated insulin was produced in yields of about 35% after purification by reversed phase HPLC. For analysis, the deuterated insulin was digested with V8 protease and was shown by mass spectrometry to have incorporated deuterium to an extent of 96.6 atom %, exclusively in a pentapeptide containing Tyr A14. The title compound should prove useful to those workers studying insulin by mass spectrometry, and use of the method with tritium gas in place of deuterium should permit the preparation of a specifically labelled radioactive insulin analogue which behaves identically to the natural hormone.

Soluble, fatty acid acylated insulins bind to albumin and show protracted action in pigs.

We have synthesized insulins acylated by fatty acids in the epsilon-amino group of LysB29. Soluble preparations can be made in the usual concentration of 600 nmol/ml (100 IU/ml) at neutral pH. The time for 50% disappearance after subcutaneous injection of the corresponding TyrA14(125I)-labelled insulins in pigs correlated with the affinity for binding to albumin (r = 0.97), suggesting that the mechanism of prolonged disappearance is binding to albumin in subcutis. Most protracted was LysB29-tetradecanoyl des-(B30) insulin. The time for 50% disappearance was 14.3 +/- 2.2 h, significantly longer than that of Neutral Protamine Hagedorn (NPH) insulin, 10.5 +/- 4.3 h (p < 0.001), and with less inter-pig variation (p < 0.001). Intravenous bolus injections of LysB29-tetradecanoyl des-(B30) human insulin showed a protracted blood glucose lowering effect compared to that of human insulin. The relative affinity of LysB29-tetradecanoyl des-(B30) insulin to the insulin receptor is 46%. In a 24-h glucose clamp study in pigs the total glucose consumptions for LysB29-tetradecanoyl des-(B30) insulin and NPH were not significantly different (p = 0.88), whereas the times when 50% of the total glucose had been infused were significantly different, 7.9 +/- 1.0 h and 6.2 +/- 1.3 h, respectively (p < 0.04). The glucose disposal curve caused by LysB29-tetradecanoyl des-(B30) insulin was more steady than that caused by NPH, without the pronounced peak at 3 h. Unlike the crystalline insulins, the soluble LysB29-tetradecanoyl des-(B30) insulin does not elicit invasion of macrophages at the site of injection. Thus, LysB29-tetradecanoyl des-(B30) insulin might be suitable for providing basal insulin in the treatment of diabetes mellitus.