IGF-I variants which bind poorly to IGF-binding proteins show more potent and prolonged hypoglycaemic action than native IGF-I in pigs and marmoset monkeys.

The relative acute hypoglycaemic potencies of IGF-I and several variants of IGF-I which bind poorly to the IGF-I binding proteins (IGFBPs) have been examined in marmosets (Callithrix jacchus) and the pig. In the marmoset study, IGF-I and des(1-3)IGF-I were compared in anaesthetised and conscious animals in a range of bolus doses from 42 to 270 micrograms/kg body weight. In the pig study, IGF-I was compared with four variants, des(1-3)IGF-I long-IGF-I, R3IGF-I and long-R3IGF-I (LR3IGF-I), which show reduced affinity for the IGFBPs as well as with insulin. Doses in the pig were 20 and 50 micrograms/kg body weight for the IGFs and 3 micrograms/kg for insulin. In each study serial blood samples were taken from 30 min before to 4 h after the bolus injection. Plasma glucose levels were decreased in a dose-responsive manner with the pig more sensitive than either the conscious or anaesthetised marmoset (maximum lowering 4.8, 3.7 and 2.5 mmol/l respectively). The IGF variants were consistently 2- to 3-fold more potent than IGF-I in each animal for lowering of plasma glucose to the nadir, with the potency reflecting the relative affinities for binding to the IGFBPs and the IGF-I receptors. Thus, hypoglycaemic potency was in the order IGF-I < long-IGF-I < R3IGF-I approximately LR3IGF-I < des (1-3)IGF-I. Notably the variants suppressed plasma glucose levels over a much longer period than did IGF-I, the cumulative suppression over four hours showing an approximately 4- to 8-fold increase in the extent of hypoglycaemia. The prolonged suppression was not simply proportional to the hypoglycaemic nadir; at doses equipotent for glucose lowering, the cumulative hypoglycaemic effect for the variants in either species was about 2-fold that for IGF-I. The differential effect of the variants in the marmoset could not be accounted for by correlated changes in plasma insulin, IGF-I or IGFBP levels in plasma. Indirect effects via inhibition of glucagon, or direct effects via hepatic insulin receptors are postulated to account for the results. There was a dose-related reduction in plasma amino acids in the pig but, unlike the case for plasma glucose, only one analogue, LR3IGF-I was more potent than IGF-I. The response to LR3IGF-I was accentuated at the high dosage but on the basis of the other variants tested this effect could not be ascribed to either of the incorporated molecular variations. Despite their more rapid clearance from the circulation, variants of IGF-I which show lower affinity for binding to IGFBPs show proportionately superior potency for sustained hypoglycaemic action. Since our data were obtained in animal models of accepted relevance to humans these results point to the possible superior efficacy of the variants, especially des(1-3)IGF-I, over IGF-I for use as an adjunct to insulin treatment of hyperglycaemic conditions.