1. We have recently provided evidence that [35S]-adenosine 5'-O-[3-thiotriphosphate] ([35S]-ATP gamma S) can label the human bladder recombinant P2X1 purinoceptor (human P2X1 purinoceptor). In this study we have characterized the binding of [35S]-ATP gamma S to a second P2X purinoceptor subtype, the rat PC12 phaeochromocytoma cell recombinant P2X2 purinoceptor (rat P2X2 purinoceptor), and compared its binding properties with those of both endogenous and recombinant P2X1 purinoceptors. 2. Infection of CHO-K1 cells with the rat P2X2 purinoceptor using Semliki forest virus (SFV) resulted in the expression of high affinity (pKd = 9.3; Bmax = 18.1 pmol mg-1 protein) binding sites for [35S]-ATP gamma S but not for [3H]-alpha, beta-methylene ATP ([3H]-alpha beta meATP). Since functional P2X purinoceptors could be detected electrophysiologically in these cells, but not in non-infected or CHO-K1 cells infected with SFV containing the LacZ gene, these results suggest that the rat P2X2 purinoceptor can be labelled using [35S]-ATP gamma S. 3. The binding characteristics of the rat P2X2 purinoceptor were compared with those of the human P2X1 purinoceptor, which was also expressed in the CHO-K1 cells using SFV. A major difference between the two recombinant P2X purinoceptor types was in the binding characteristics of alpha, beta-methylene ATP (alpha beta meATP). Thus, in the absence of divalent cations, alpha beta meATP possessed low affinity for both the human P2X1 purinoceptor (pIC50 = 7.2) and rat P2X2 purinoceptor (pIC50 = 7.1) labelled using [35S]-ATP gamma S. However, when the recombinant P2X purinoceptors were labelled with [3H]-alpha beta meATP in the presence of 4 mM CaCl2, the affinity of alpha beta meATP for the human P2X1 purinoceptor increased (pIC50 for alpha beta meATP = 8.2), while the affinity of the rat P2X2 purinoceptor for alpha beta meATP did not change (pIC50 for alpha beta meATP = 6.8). 4. Affinity estimates of 15 other nucleotide analogues for the [35S]-ATP gamma S binding sites on the two recombinant P2X purinoceptor subtypes were surprisingly similar (less than 5 fold difference), the only exception being 2'-deoxy ATP which possessed 8 fold higher affinity for rat P2X2 than for human P2X1 purinoceptors. In contrast dextran sulphate and the P2 purinoceptor antagonists, pyridoxalphosphate-6-azophenyl-2', 4'-disulphonic acid and 4,4'-diisothiocyanatostilbene-2,2' disulphonic acid, possessed 7 to 33 fold higher affinity for the human P2X1 than for the rat P2X2 purinoceptor. These data provide a correlation coefficient (r) of 0.894. 5. There was some evidence for species differences in the P2X1 purinoceptor. Thus, most nucleotides possessed slightly greater (up to 9-10 fold), while the P2 purinoceptor antagonists possessed slightly lower (up to 7-16 fold), affinity for the endogenous rat vas deferens and rat bladder P2X1 purinoceptors than for the human recombinant P2X1 purinoceptor. These differences were reflected in a slightly lower correlation coefficient, when comparing across species between the human recombinant P2X1 purinoceptor and the endogenous P2X1 purinoceptors labelled in either the rat deferens (r = 0.915) or the rat bladder (r = 0.932), than when comparing within species between the endogenous rat vas deferens and rat bladder P2X1 purinoceptors (r = 0.995). 6. In summary, [35S]-ATP gamma S can be used to label the recombinant P2X1 and P2X2 purinoceptors. Despite the marked differences reported between these two forms of P2X purinoceptor in functional studies, the differences in binding studies were more limited. However, a number of antagonists could discriminate between the P2X purinoceptor subtypes in the binding studies raising expectations that selective antagonists for these receptors can be developed.