EFFECTS OF THE PHORBOL ESTER PHORBOL 12-MYRISTATE 13-ACETATE (PMA) ON ISLET-CELL RESPONSIVENESS

Collagenase-isolated rat islets were labelled for 2 h in myo-[2-H-3]inositol solution supplemented with 2.75 mM-glucose. The phorbol ester phorbol 12-myristate 13-acetate (PMA; 0.1 or 1-mu-M) was also present in some experiments. After labelling, islets were washed and then perifused in 2.75 mM-glucose to establish basal [H-3]inositol-efflux and insulin-secretory rates. Subsequently, the responses of these islets to stimulation with various agonists were assessed. Inositol phosphate accumulation was measured at the termination of the perifusion. In separate experiments, the cellular location of protein kinase C (PKC) after PMA pretreatment was measured by quantitative immunoblotting of membrane and cytosolic fractions. The following observations were made. (1) Labelling in 0.1-1 /mu-M-PMA had no deleterious effect on total [H-3]inositol incorporation during the 2 h labelling period. However, islets labelled for 2 h in 1-mu-M-PMA were unable to respond, in terms of increases in insulin release, to a 1-mu-M-PMA stimulus during the subsequent perifusion. (2) As compared with the responses of control islets labelled in 2.75 mM-glucose alone, islets labelled in the additional presence of 1-mu-M-PMA displayed a significant impairment in phosphoinositide (PI) hydrolysis, but an enhancement of both first- and second-phase insulin secretion, in response to subsequent 20 mM-glucose stimulation. (3) Decreasing extracellular Ca2+ level to 0.1 mM and including the Ca2+-channel antagonist nitrendipine (0.5-mu-M) along with 1-mu-M-PMA during the [H-3]inositol-labelling period did not alter the response of the islets to the subsequent addition of 20 mM-glucose. Glucose-induced PI hydrolysis was still inhibited and 20 mM-glucose-induced insulin release was still enhanced. (4) A markedly amplified and sustained insulin-secretory response to 200-mu-M-tolbutamide in the presence of 2.75 mM-glucose was also obtained from 1-mu-M-PMA-pretreated islets. This contrasts sharply with the small and transient response to tolbutamide noted in control islets. (5) When present only during the perifusion phase of the experiments, nitrendipine (0.5-mu-M) abolished the amplified insulin-secretory responses to both 20 mM-glucose and 200-mu-M-tolbutamide noted in PMA-pretreated islets. (6) Prior labelling in 1-mu-M-PMA dramatically amplified the insulinotropic effect of 25 mM-K+ or 5-mu-M-A23187 stimulation. The amplified insulin-secretory response to K+, but not to A23187, was abolished by inclusion of nitrendipine during the perifusion. (7) Labelling in 1-mu-M-PMA also decreased PI hydrolysis in response to cholecystokinin (CCK-8S) stimulation. (8) A 2 h pre-exposure to 1-mu-M-PMA resulted in the persistent translocation of immunoreactive PKC to the membrane fraction. These results support the following conclusions. PMA pretreatment desensitizes the islet, in terms of agonist-induced increases in PI hydrolysis, to glucose, CCK-8S or tolbutamide. In spite of this inhibitory effect on PI hydrolysis, prior exposure to PMA dramatically amplified the insulin-secretory response to subsequent agonist stimulation. It is suggested that the chronic activation of PKC by PMA induces feedback inhibition of PI hydrolysis. However, the sensitivity of PKC activation to Ca2+, a result of persistent PKC membrane translocation, is dramatically increased. This accounts for the amplified insulin-secretory response despite diminished PI hydrolysis. These findings emphasize the importance of PKC activation in glucose-induced insulin secretion.