Regulation of adenylyl cyclase activity by guanine nucleoside tri- and diphosphates as well as by stimulatory and inhibitory receptors was studied in canine cardiac sarcolemmal membranes. Guanosine triphosphate (GTP) increased adenylyl cyclase activity by a maximum of 80%, with an EC50 value of 0.7 mumol/l. The addition of the beta-adrenoceptor agonist, isoprenaline (100 mumol/l), caused a further, about 100%, increase in GTP-stimulated activity. The nucleoside diphosphate (GDP) also activated cardiac adenylyl cyclase, but in a biphasic manner. At low concentrations (EC50 0.12 mumol/ l). GDP increased enzyme activity by about 80%, followed by a plateau at 0.5-2 mumol/l and a second increase to a maximum of 60% with an EC50 value of 14 mumol/l. The stable GDP analog, guanosine 5'-O-(2-thio)diphosphate (GDP beta S), also increased cardiac adenylyl cyclase activity, but in a monophasic manner, by a maximum of 150%, with an EC50 of 0.4 mumol/l. Addition of uracil diphosphate (UDP) (3 mmol/l), which completely inhibited transphosphorylation of GDP to GTP, did not reduce adenylyl cyclase stimulation by low concentrations of GDP, whereas enzyme stimulation by high GDP concentrations was almost completely attenuated. Furthermore, pretreatment of the membranes with cholera toxin led to an increased stimulation of adenylyl cyclase activity by high concentrations of GDP. These findings suggest that the second phase of adenylyl cyclase stimulation by GDP is due to transphosphorylation of GDP to GTP, associated with activation of Gs proteins, and that stimulation by GDP itself (first phase) and endogenously formed GTP (second phase) is additive. However, in contrast to exogenously added GTP, beta-adrenoceptor activation did not enhance GDP-stimulated adenylyl cyclase activity. Furthermore, in the presence of 1 mumol/l GDP, the addition of GTP did not cause any further increase in enzyme activity. On the other hand, the muscarinic acetylcholine receptor agonist carbachol inhibited both GTP- and GDP-activated adenylyl cyclase. The inhibition of GDP-stimulated activity was lost when formation of GTP from GDP was blocked. The contrasting effects of endogenously formed GTP and exogenous GTP suggest that the formation of GTP from GDP is closely linked to the activation site of adenylyl cyclase, i.e. the stimulatory Gs protein. This receptor-independent activation can apparently bypass beta-adrenoceptor-dependent activation of cardiac adenylyl cyclase.