We have studied the modulation of cyclic AMP (cAMP) accumulation by the human immunodeficiency virus type 1 (HIV 1) protein Tat in microglia and astrocyte cultures obtained from neonatal rat brain. Pretreatment of microglia with recombinant Tat resulted in a dose- and time-dependent decrease of cAMP accumulation induced by subsequent exposure to isoproterenol (1 microM). The inhibitory action of 100 ng/mL Tat approached 50% after 4 h of preincubation and reached a maximum of 70% after 24 h. The Tat-induced time- and dose-dependent decrease of cAMP accumulation was observed also when microglial cultures were stimulated with the adenylyl cyclase activator forskolin (100 microM). In both cases, Tat inhibitory action was 70% reverted by a specific monoclonal anti-Tat antibody, but was not prevented either by the phosphodiesterase inhibitor 3-isobutyl-1-methyl-xantine (100 microM) or by a 16-h pretreatment of microglial cultures with the Gi protein inhibitor pertussis toxin (10 ng/mL). All these results suggested that the viral protein acts at a step of the cAMP transduction pathway other than receptors, G proteins and phosphodiesterases. The target of Tat appeared to be adenylyl cyclase, whose activity was markedly reduced (up to 60%) in membranes prepared from Tat-treated microglial cells, both in basal conditions and after stimulation with isoproterenol and forskolin. The inability of the competitive inhibitor of nitric oxide synthase N(G)-monometyl- L-arginine (20 and 200 microM) to revert Tat action on forskolin-induced cAMP accumulation, and of two potent nitric oxide donors, PAPA and DETA (0.1-2 m M), to alter forskolin-induced cAMP accumulation, excluded an involvement of nitric oxide in Tat-induced adenylyl cyclase inhibition. On the contrary, two inhibitors of nuclear factor kappaB activation, N-tosyl-( L)-phenylalanine chloromethyl ketone (10 microM) and SN50 (25 microM), markedly prevented the reduction of forskolin-evoked cAMP accumulation by Tat, suggesting a possible role for this nuclear transcriptional factor in the regulation of adenylyl cyclase by Tat in microglia. This assumption was strengthened by the ability of lipopolysaccharide (100 ng/mL, 4 h) to mimic the inhibitory effect of the viral protein. Conversely, astrocyte cAMP accumulation was unaffected by the viral protein, as tested at various concentrations and time points. Finally, Tat inhibition of microglial adenylyl cyclase was not due to non-specific cytotoxicity. As cAMP has been reported to exert a neuroprotective role in several in vivo and in vitro models of brain pathologies, and microglia is believed to mediate Tat-induced neurotoxicity, these results suggest that the ability of Tat to inhibit cAMP synthesis in microglia may contribute to neuronal degeneration and cell death associated with HIV infection.