It has remained unclear so far why the glucagon secretion has the inverse glucose dependence of insulin secretion. The main question is whether this pattern is due to alpha-cell intrinsic mechanisms or to interactions within the pancreatic islet.
Insulin and glucagon were determined by ELISA from the fractionated efflux of batch-perifused NMRI mouse islets. Alpha-cells were isolated from these islets by incubation with alloxan and culturing the surviving cells for 24h. In such cells the cytosolic Ca2+ concentration ([Ca2+]i) was measured by microfluorometry.
In the presence of 1 mM glucose 20 mM arginine had a moderate stimulatory effect on glucagon secretion but not on insulin secretion. Adding either 30µM gliclazide or 500µM tolbutamide or 15 mM KCl increased insulin secretion before it diminished glucagon secretion. In the combined presence of 20 mM arginine and 30µM gliclazide (1 mM glucose) the a2-adrenoceptor agonist clonidine suppressed insulin secretion and markedly increased glucagon secretion. In a situation of low insulin secretion clonidine did not affect glucagon secretion. In contrast to single beta cells, single non-beta cells (survivors of alloxan treatment) did not respond to 500µM tolbutamide by increasing [Ca2+]i, but did so in response to 20 mM arginine. Of these 50% responded to 1 mM glutamate and are thus alpha-cells.
The paradoxical decrease of glucagon secretion in response to depolarizing non-nutrients depends on paracrine insulin release. The inability of KATP channel closure to increase [Ca2+]i in alpha cells suggests that nutrient recognition differs from that in beta cells.
