Cross Talk between Stimulatory and Inhibitory Guanosine 5’-Triphosphate Binding Proteins: Role in Activation and Desensitization of the Adenylate Cyclase Response to Vasopressin_t

The inhibitory GTP-binding protein (G,) is known to mediate the effects of a number of hormones that act through specific receptors to inhibit adenylate cyclase. In this study we examined the mechanism whereby Gi modulates the response of adenylate cyclase to a stimulatory hormone and its role in desensitization. In membranes prepared from the cultured renal epithelial cell line LLCPKi, adenylate cyclase activity was stimulated 16-fold by 1–2 pM lysine vasopressin. Addition of GTP (1–100 pM) resulted in stimulation of basal activity but inhibition of hormone-stimulated activity (~40% inhibition at 100 AM GTP). This constrasts with the usual effect of GTP to support or augment activation by stimulatory receptors. The inhibitory effect was abolished by pertussis toxin, which had little effect on basal activity in the absence or presence of added GTP or on vasopressin-stimulated activity in the absence of added GTP. GTP-mediated inhibition was vasopressin concentration dependent. At concentrations of vasopressin below the kij2for enzyme activation (approximately 0.6 nM), GTP was stimulatory, and at higher concentrations, GTP was inhibitory. The inhibitory effect of GTP was also observed for a V₂-receptor agonist and was not abolished by a V_rreceptor antagonist, indicating that a distinct V1 receptor did not mediate inhibition of adenylate cyclase. Using the known subunit structure of adenylate cyclase, we developed the minimal mechanism that would incorporate a modulatory role for Gi in determining net activation of adenylate cyclase by a stimulatory hormone. The predicted enzyme activities for basal and maximal hormone stimulation in the presence and absence of GTP were generated, and model parameters were chosen to match the experimental observations. Direct interaction of the receptor with Gj was not required, but a direct inhibitory interaction of the a, subunit with G; was necessary for the model to yield the experimentally observed effects of GTP on basal and hormone-stimulated enzyme activity. Elimination of input in the model without changing any other model parameters mimicked the experimentally observed effects of pertussis toxin. Despite the lack of a requirement for direct interaction of the receptor with Gj, the model predicted the inhibitory effect of GTP to be hormone concentration dependent. We also observed that forskolin-stimulated activity was inhibited by GTP, an effect that could be reversed by pertussis toxin, further supporting that interaction of the vasopressin receptor with Gi was not required for GTP-mediated inhibition. The effects of pertussis toxin and GTP in desensitized cells observed experimentally were compared with the pattern predicted by the model. Pertussis toxin pretreatment did not affect homologous desensitization, excluding a role for Gj in homologous desensitization. The results obtained for desensitized cells were predicted by postulating a defect in hormone-stimulated guanyl nucleotide exchange on G_s in the model. These results allow us to propose a comprehensive mechanism whereby cross talk between Gi and G_sdetermines net adenylate cyclase activation by a stimulatory hormone. The modulatory role of Gj does not require its direct interaction with the receptor but does require a direct inhibitory interaction with the catalytic unit. The results also exclude a direct role for in homologous desensitization.