[35S]Guanosine 5'-(gamma-thio)triphosphate autoradiography is a novel technique to detect receptor-dependent activation of G-proteins in brain tissue sections. While an increasing number of reports using this approach are beginning to appear, little effort has been directed to the identification of factors responsible for the heterogeneously distributed [35S]guanosine 5'-(gamma-thio)triphosphate signal in basal conditions. The present study demonstrates that endogenously formed adenosine generates a widespread and prominent adenosine A1 receptor-dependent signal in basal conditions using this technique. Treatment of rat brain tissue sections with the A1-selective antagonist 8-cyclopentyl-1,3-dipropylxanthine dose-dependently (EC50 < 10 nM) suppressed basal [35S]guanosine 5'-(gamma-thio)triphosphate binding in a region-specific manner, an effect fully mimicked by the adenosine-depleting enzyme adenosine deaminase, and less so by the A1 antagonist cirsimarin and by caffeine. That adenosine was continuously formed during the incubation is supported by the constant requirements of adenosine deaminase in order to suppress basal radioligand binding and further by the fact that low micromolar concentrations of adenine nucleotides evoked only adenosine-mimicking and fully 8-cyclopentyl-1,3-dipropylxanthine-sensitive binding responses. In the presence of adenosine deaminase, all responses to adenine nucleotides were abolished, indicating that prior conversion to adenosine was required. Upon stimulation, this technique selectively detected A1 receptor-activated G-proteins, as the non-selective agonists adenosine and 2-chloroadenosine and the A1-selective agonist N6-p-sulfophenyladenosine all evoked only 8-cyclopentyl-1,3-dipropylxanthine-sensitive responses in identical gray matter areas, and also in several white matter areas such as the corpus callosum, anterior commissure, optic tract and cerebellar white matter. Dose-response studies revealed region-specific differences in the magnitude of A1 receptor-stimulated G-protein activation, with the highest response (nine-fold over basal) detectable in the hippocampus. No response to the A2A-selective agonist 2-[(2-aminoethylamino)carbonylethylphenylethylamino]-5'-N-et hylcarboxamidoadenosine or the A3-selective agonist 2-chloro-N6-(3-iodobenzyl)-adenosine-5'-N-methyluronamide was detected in any region. These data reveal that a significant amount of noise inherent to [35S]guanosine 5'-(gamma-thio)triphosphate autoradiography can be eliminated by removal of the adenosine signal, a step likely facilitating detection of responses to other receptors. Furthermore, the data establish [35S]guanosine 5-(gamma-thio)triphosphate autoradiography as a novel and selective approach to directly assess A1 receptor-G-protein coupling in anatomically defined regions of the central nervous system.