Pharmacological approach to understanding the control of insulin secretion in human islets.

Aims To understand better the control of insulin secretion by human β cells and to identify similarities to and differences from rodent models. Methods Dynamic insulin secretion was measured in perifused human islets treated with pharmacological agents of known modes of action. Results Glucokinase activation ( Ro28-1675) lowered the glucose threshold for stimulation of insulin secretion to 1 mmol/ L ( G1), augmented the response to G3-G5 but not to G8-G15, whereas tolbutamide remained active in G20, which indicates that not all K_ATP channels were closed by high glucose concentrations. An almost 2-fold greater response to G15 than to supramaximal tolbutamide in G3 or to KCl+diazoxide in G15 vs G3 quantified the contribution of metabolic amplification to insulin secretion. Both disruption (latrunculin- B) and stabilization (jasplakinolide) of microfilaments augmented insulin secretion without affecting metabolic amplification. Tolbutamide-induced insulin secretion was consistently greater in G10 than G3, with a threshold at 1 and maximum at 10 µmol/ L tolbutamide in G10, vs 10 and 25 µmol/ L in G3. Sulphonylurea effects were thus clearly glucose-dependent. Insulin secretion was also increased by inhibiting K channels other than K_ATP channels: Kv or BK channels (tetraethylammonium), TASK-1 channels ( ML-365) and SK4 channels ( TRAM-34). Opening K_ATP channels with diazoxide inhibited glucose-induced insulin secretion with half maximum inhibitory concentrations of 9.6 and 24 µmol/ L at G7 and G15. Blockade of L-type Ca channels (nimodipine) abolished insulin secretion, whereas a blocker of T-type Ca channels ( NNC-55-0396) was ineffective at specific concentrations. Blockade of Na channels (tetrodotoxin) did not affect glucose-induced insulin secretion. Conclusions In addition to sharing a K_ATP channel-dependent triggering pathway and a metabolic amplifying pathway, human and rodent β cells were found to display more similarities than differences in the control of insulin secretion. [ABSTRACT FROM AUTHOR]