Endothelium-dependent relaxation to hydrogen peroxide in canine basilar artery: a potential new cerebral dilator mechanism.

In prostglandin F2alpha(PGF2alpha)-precontracted isolated canine basilar arterial rings, hydrogen peroxide (H2O2) produced endothelium-dependent relaxations at concentrations of from 4.4 x 10(-7) - approximately 4.4 x 10(-5) M. Removal of extracellular Ca2+ ([Ca2+]0) attenuated the relaxant effects of H2O2. Complete inhibition of H2O2 relaxant action was obtained after buffering intracellular Ca2+ ([Ca2+]i), in the endothelial cells, with 10 microM 1,2-bis (2-aminophenoxy) ethane-N,N,N',N'-tetraacetic acid (BAPTA-AM). The H2O2-induced relaxations could be abolished completely by 1200 u/ml catalase and was suppressed significantly by 0.5 microM atropine, 150 microM NG-monomethyl-arginine (L-NMMA), 50 microM NG-nitro-L-arginine methyl ester (L-NAME), 1 microM Fe2+, or 5 microM methylene blue. These inhibitory effects of L-NMMA, L-NAME, or atropine could be reversed partly by 50 microM L-arginine. The Fe2+ inhibition of H2O2-stimulated relaxation was reduced significantly by either 1 mM deferoxamine (a Fe2+ chelator) or 100 microM dimethyl sulfoxide (DMSO, a *OH scavenger). Such relaxant effects of H2O2 were enhanced, significantly, by an acetylcholinesterase antagonist, neostigmine. A variety of pharmacological antagonists (of diverse vasodilator agents) could not inhibit the relaxant action of H2O2. Our observations suggest that at suitable pathophysiological concentrations, H2O2 could induce release of an endothelium-derived relaxing factor (EDRF), probably nitric oxide (NO), from endothelial cells of the canine cerebral artery. The H2O2 relaxant effects are clearly Ca2+-dependent, require formation of cyclic guanosine monophosphate (cGMP), and may be associated with release of endogenous acetylcholine (ACh).