Dantrolene-induced inhibition of insulin release. A mechanism independent of effects on calcium fluxes

Production of phosphatidylethanol by phospholipase D phosphatidyl transferase in intact or dispersed pancreatic islets: evidence for the in situ metabolism of phosphatidylethanol

To determine if phospholipase D is present in intact adult islets, we took advantage of the fact that, in the presence of ethanol, this enzyme generates phosphatidylethanol via transphosphatidylation. Extracts of cells prelabeled with [14C]arachidonate, [14C]myristate, or [14C]stearate were analyzed via three TLC systems; the identify of phosphatidylethanol was further confirmed via incorporation of [14C]ethanol into the same phospholipid bands. The phorbol ester 12-O-tetradecanoylphorbol-13-acetate stimulated phosphatidylethanol (to 603% of basal by 60 min) both in intact adult islets and in dispersed neonatal islet cells. A nonphorbol activator of protein kinase C (mezerein) also stimulated phospholipase D, whereas a phorbol which does not activate protein kinase C (4 alpha-phorbol-12,13-didecanoate) was virtually inactive. The effects of the active phorbol ester or of mezerein were reduced by the protein kinase C inhibitor H-7 and were virtually eliminated by prior down-regulation of that enzyme. In addition, a calcium-selective ionophore (ionomycin) or fluoroaluminate also activated the islet phospholipase D. When accumulation of phosphatidylethanol (labeled with any of three fatty acids) was induced by a preincubation in the presence of ethanol plus agonist, which then were removed, phosphatidylethanol declined by 34-47% over a subsequent 60-min incubation. Thus, while phosphatidylethanol is relatively stable metabolically, it is detectably degraded (a variable overlooked in previous studies). In the absence of ethanol, stimulated islet cells generated phosphatidic acid, although such hydrolysis was less evident than transphosphatidylation. Ethanol provision distinguished phosphatidate formed via phospholipase D (inhibition, via phosphatidylethanol formation) from that due predominantly to phospholipase C (phosphatidate not inhibited). In view of our recent findings that phosphatidic acid (or exogenous phospholipase D) has potent insulinotropic effects, this pathway could play a role in stimulus-secretion coupling; conversely, stimulation of transphosphatidylation at the expense of hydrolysis could contribute to the inhibition of secretion caused by ethanol.

Stimulation of insulin release by phospholipase D. A potential role for endogenous phosphatidic acid in pancreatic islet function

Although exogenous phosphatidic acid (PA) has been shown to promote insulin release, the effects of endogenous PA on endocrine function are largely unexplored. In order to generate PA in situ, intact adult-rat islets were treated with exogenous phospholipases of the D type (PLD), and their effects on phospholipid metabolism and on insulin release were studied in parallel. Chromatographically purified PLD from Streptomyces chromofuscus stimulated the accumulation of PA in [14C]arachidonate- or [14C]myristate-prelabelled islets, and also promoted insulin secretion over an identical concentration range. During 30 min incubations, insulin release correlated closely with the accumulation of [14C]arachidonate-labelled PA (r2 = 0.98; P less than 0.01) or [14C]myristate-labelled PA (r2 = 0.97; P less than 0.01). Similar effects were seen both in freshly isolated and in overnight-cultured intact islets. In contrast, PLDs (from cabbage or peanut) which do not support phospholipid hydrolysis at the pH of the extracellular medium also did not promote insulin release. The effects on secretion of the active PLD preparation were inhibited by modest cooling (to 30 degrees C); dantrolene or Co2+ also inhibited PLD-induced secretion without decreasing PLD-induced PA formation. Additionally, the removal of PLD left the subsequent islet responsiveness to glucose intact, further supporting an exocytotic non-toxic mechanism. PLD-induced insulin release did not appear to require influx of extracellular Ca2+, nor could the activation of protein kinase C clearly be implicated. During incubations of 30 min, PLD selectively generated PA; however, more prolonged incubations (60 min) also led to production of some diacyglycerol and free arachidonic acid concomitant with progressive insulin release. These data suggest that PLD activation has both rapid and direct effects (via PA) and more delayed, secondary, effects (via other effects of PA or the generation of other lipid signals). Taken in conjunction with our demonstration that pancreatic islets contain an endogenous PLD which generates PA [Dunlop & Metz (1989) Biochem. Biophys. Res. Commun. 163, 922-928], these studies provide evidence suggesting that PLD activation (and possibly other pathways leading to PA formation) could play a role in stimulus-secretion coupling in pancreatic islets.