The role of endothelium in angiotensin II-induced contractions of the rabbit aorta and the mechanism involved were investigated. Destruction of the endothelium significantly shifted the concentration-response curve for angiotensin II to the left in a non-parallel manner and enhanced the maximal response. The EC(50) and E(max) values obtained from the rings with and without functional endothelium were 2.44 +/- 0.13 x 10(-9) M, 4.50 +/- 0.45 g and 1.21 +/- 0.14 x 10(-9) M (n = 8, P < 0.05), 5.73 +/- 0.55 g (n = 8, P < 0.05), respectively. Indomethacin (10(-5) M) did not significantly alter the concentration-dependent response to angiotensin II in the presence of endothelium. Three inhibitors of nitric oxide synthase (N-G-monomethyl-L-arginine; N-G-nitro-L-arginine, and N-G-nitro-L-arginine methyl ester) at 10(-4) M caused a similar endothelium-dependent potentiation of angiotensin II-induced contractions in the aortic rings with, but not in those without endothelium. These effects were reversed by L-arginine (3 X 10(-3) M) but not by D-arginine (3 x 10(-3) M). Angiotensin II in a concentration range of 10(-16) to 10(-6) M did not relax the endothelium-intact rings precontracted with phenylephrine (2 x 10(-7) M). In the presence of endothelium, the angiotensin II subtype 2 receptor antagonist, 1-[(4-amino-3-methylphenyl]-5-(diphenylacetyl H-imidazol, [4,5-C]pyridine-6-carboxylic acid (PD 123177), caused relaxation of the rings precontracted with phenylephrine nor alteration of the concentration-response curve for angiotensin II. These results provide evidence for endothelium-dependent inhibition of angiotensin II-induced contractions in the rabbit aorta. They support the assumption that a basal or stimulated release of endothelium-derived relaxing factor (NO) counteracts the contractions induced by angiotensin II in this preparation.