To examine the effect of a prior episode of anoxia on subsequent anoxia-mediated vasorelaxation, norepinephrine-precontracted endothelium-intact rat aortic rings were first exposed to anoxia (95% N2-5% CO2 for 5, 15, or 30 min) then to normoxia (95% O2-5% CO2 for 15 min). These rings were exposed again to anoxia for 30 min. First exposure of rings to anoxia for 30 min resulted in 77 +/- 4% decrease in tone (vasorelaxation), whereas second exposure resulted in only 10 +/- 4% relaxation (n = 11, P < 0.001 vs. relaxation during first exposure). First exposure of rings to anoxia for 5 or 15 min also diminished relaxation to 59 +/- 3 and 19 +/- 8%, respectively, on second exposure to anoxia (both P < 0.01 vs. relaxation during 1st anoxia). Attenuation of vasorelaxation by prior episode of anoxia was not affected by treatment of rings with indomethacin (10(-5) M), the Ca2+ channel blocker felodipine (10(-6) M), the superoxide anion scavenger superoxide dismutase (100 micrograms/ml), or adenosine A1 and A2 blockers (each 10(-6) M). To examine the role of intact functional endothelium in attenuation of vasorelaxation during second anoxic exposure, rings were deendothelialized and treated with the nitric oxide synthase inhibitor NG-monomethyl-L-arginine (L-NMMA; 10(-4) M) or the guanylate cyclase inhibitor methylene blue (MB; 2 x 10(-5) M). In all deendothelialized rings, vasorelaxation during second anoxic exposure was similar to that during first anoxic exposure (100 +/- 0 vs. 98 +/- 3%, P = NS).(ABSTRACT TRUNCATED AT 250 WORDS)