Catecholamines inhibit adenylate cyclase in pancreatic B-cells, but the importance of the resulting fall in cAMP concentration for the decrease in insulin release remains controversial. Adrenaline caused a dose-dependent inhibition (EC50 = 5.7 nM) of insulin release by mouse islets incubated in a medium containing 15 mM glucose. Supplementation of the medium with 500 microM dibutyryl-cAMP or 1 microM forskolin potentiated the effect of glucose on release and attenuated the inhibition by 1 and 10 nM adrenaline; the EC50 value was increased 2-fold. The inhibitory action of 100 nM or 1 microM adrenaline was, however, not affected. This apparent change in adrenaline potency was not simply due to the larger rate of release since it was not observed when the effect of glucose was potentiated by cytochalasin-B. However, when the same rate of insulin release as that produced by 15 mM glucose alone was achieved by combining 10 mM glucose and 250 microM dibutyryl-cAMP, the inhibitory potency of adrenaline was unaffected. Intracellular microelectrodes were used to determine whether the changes in B-cell membrane potential brought about by adrenaline are mediated by a fall in cAMP levels. Addition of dibutyryl-cAMP or forskolin to a medium containing 10 or 15 mM glucose increased the Ca(2+)-dependent electrical activity triggered by the sugar. However, this did not prevent adrenaline from hyperpolarizing the membrane transiently and causing a steady-state decrease in the intensity of the electrical activity.(ABSTRACT TRUNCATED AT 250 WORDS)