Pharmacological interference with phospholipase A2 activity reveals mechanistic differences between glucose and glyceraldehyde induced insulin release: implication for coupling of glucose metabolism to phospholipase A2 activity.

Prostaglandin E2 levels in isolated rat islets were increased from 64 +/- 11 pg/30 islets when incubated in medium containing 2 mM glucose to 115 +/- 9 pg/30 islets in medium containing 20 mM glucose. In contrast, glyceraldehyde (10 mM) reduced prostaglandin E2 levels to 29 +/- 6 pg/30 islets. Inhibition of glucose metabolism by mannoheptulose (10 mM) abolished the stimulatory effect of glucose on prostaglandin E2 levels and inhibited glucose-induced insulin release. The cyclooxygenase inhibitor, flurbiprofen (20 microM), did not affect insulin release caused by glucose or glyceraldehyde. In the presence of 1 mg/ml bovine serum albumin, insulin secretion induced by 20 mM glucose (6.9 +/- 1.1% of islet insulin content) was reduced by the lipoxygenase inhibitor BW755 C (20 microM) to 3.1 +/- 0.6%, and by the phospholipase A2 inhibitor, p-bromophenacyl bromide (10 microM), to 2.1 +/- 0.8%. In the absence of bovine serum albumin the inhibitory action of BW755 C and p-bromophenacyl bromide on glucose-induced insulin release was significantly more pronounced. These drugs whether in the presence or absence of bovine serum albumin, did not affect glyceraldehyde-stimulated insulin secretion. Glyceraldehyde (10 mM), potentiated glucose-induced insulin release in the presence of 2-8 mM glucose, but not for 10-20 mM glucose. Although the phospholipase A2 activator, melittin, initiated insulin release in the presence of 2 mM glucose and enhanced 10 mM glyceraldehyde-stimulated insulin secretion it had no effect on 20 mM glucose-induced insulin release. These two stimulatory effects of melittin on insulin release were totally abolished by p-bromophenacyl bromide.(ABSTRACT TRUNCATED AT 250 WORDS)