The aim of this study was to characterise the receptor(s) mediating relaxations to adenosine and its analogues in the hamster isolated aorta. Adenosine relaxed the aorta but there was no significant difference between pIC20 values in the absence and presence of 8-sulphophenyltheophylline (8-SPT, 50 microM), although there was a small right-shift (approximately threefold) of the lower portion of the curve in the presence of 8-SPT. However, in the presence of the adenosine uptake inhibitor nitrobenzylthioinosine (NBTI, 1 microM), curves to adenosine were left-shifted by approximately 100-fold and an apparent pK(B) for 8-SPT of 5.79+/-0.05 was obtained. Likewise, 5'-N-ethylcarboxamidoadenosine (NECA) relaxed the aorta but curves were biphasic. The first phase of the curve was blocked by 8-SPT (10-100 microM, pA2 = 5.75+/-0.14) and the A2A-selective antagonist 4-(2-[7-amino-2-(2-furyl) [1,2,4]-triazolo[2,3-a][1,3,5]triazin-5-ylaminolethyl) phenol (ZM 241385, 3 nM-1 microM, pK(B)=9.17+/-0.10). Similarly, the A2A-selective agonist 2-[p)-(2-carbonylethyl)-phenylethylamino]-5'-N-ethylcarboxam idoadenosine (CGS 21680) relaxed the tissues but curves were biphasic and the first phase was again blocked by ZM 241385 (10 nM, apparent pK(B)=9.06+/-0.34). In contrast, relaxations to N6-R-phenylisopropyladenosine (R-PIA), N6-cyclopentyladenosine (CPA), 2-chloroadenosine (2-CADO) and N6-(3-iodobenzyl)-adenosine-5'-N-methyluronamide (IB-MECA) were not blocked by 8-SPT (50 microM). Responses to IB-MECA were also not blocked by the A3 receptor antagonist 3-ethyl-5-benzyl-2-methyl-6-phenyl-4-phenylethynyl-1,4-(+/-)-dihyd ropyridine-3,5-dicarboxylate (MRS 1191, 30 microM). The asymptote of the first phase of curves to NECA was markedly reduced (and in some preparations the first phase was completely abolished) both in the presence of N(G)-nitro-L-arginine methyl ester (L-NAME, 0.1 mM), and in the absence of endothelium. Likewise, the first phase of curves to CGS 21680 was abolished both in the presence of L-NAME (0.1 mM) and in the absence of endothelium. In contrast, there were only relatively small shifts to the right of curves to adenosine and the other analogues in the presence of L-NAME or the absence of endothelium (between three- and fivefold). The data suggest the presence of A2A receptors which are located on the endothelium and mediate release of nitric oxide. These receptors are activated by NECA, CGS 21680 and adenosine (in the presence of uptake blockade). The resistance to blockade of relaxations to adenosine (in the absence of uptake inhibitor), CPA, R-PIA, 2-CADO, IB-MECA and high concentrations of NECA and CGS 21680 by 8-SPT or ZM 241385 suggests the presence of an additional mechanism(s). Data obtained with adenosine in the absence and presence of NBTI suggest that the endogenous ligand may cause relaxation via an intracellular mechanism.