We examined and compared the actions of IGF-I and -II on the release of insulin from isolated, intact rat islets of Langerhans within a perifusion system. Islets were isolated from adult male rats by collagenase digestion and Ficoll gradient separation, and were maintained in tissue culture for 48 h before perifusion. Following an equilibration period, islets were perifused with medium containing 2·7 m m glucose from 0 to 30 min, and 2·7 11·1 or 16·7 m m glucose from 30 to 90 min. All chambers then received medium with 2·7 m m glucose from 90 to 120 min. Various doses (6·7–53 n m ) of IGF-I, des(1–3) IGF-I or IGF-II were given either as a pulse between 30 and 35 min, or continuously from 30 to 90 min. Insulin was measured in effluent medium by RIA. When 11·1 m m glucose was administered after 30 min an immediate increase in insulin release occurred, from a baseline of 1–3 pmol/fraction to approximately 7 pmol/fraction. The elevated rate of release was maintained until 90 min, and fell when the glucose concentration was lowered. Glucose at 16·7 m m was a less effective insulin secretogogue than was 11·1 m m . When islets received a pulse infusion of IGF-I (13·3 n m ) at 30 min in the presence of 11·1 m m glucose, a statistically significant increase ( P <0·005) in insulin release occurred, of approximately 10 pmol/fraction in excess of that seen with glucose alone. The IGF-I-stimulated insulin release was still higher than controls at 115 min. When the concentration of IGF-I was altered between 6·7 n m and 53 n m , maximum insulin release was achieved with 13·3 n m IGF I, both lower and higher concentrations being less effective. A significant inhibition of insulin release occurred with 53 n m IGF-I compared with glucose alone. IGF-II (13·3 n m ) did not significantly increase insulin release, while 53 n m IGF-II significantly inhibited release of insulin relative to controls. Des(1–3) IGF-I (13·3 n m ), which has a reduced binding affinity for IGF-binding proteins (IGFBPs), administered with 11·1 m m glucose caused an immediate increase in insulin release, which fell to control values within 30 min. Western ligand blot analysis identified four IGFBP species in perifused islets, of 46 kDa, 35 kDa, 28 kDa and 19 kDa respectively, of which the 28 kDa species was identified immunologically as IGFBP-1. When IGF-I was administered continuously from 30 to 90 min it inhibited glucose-stimulated insulin release at all concentrations used. The results suggest that under perifusion conditions, IGF-I can act both as a potent insulin secretogogue, augmenting the actions of glucose, and as an inhibitor of insulin release, depending on concentration and kinetics of administration.

Journal of Endocrinology (1997) 153, 15–25