Available evidence for oxidative stress after angioplasty is indirect or ambiguous. We sought to characterize the pattern, time course, and possible sources of free radical generation early after arterial balloon injury. Ex vivo injury performed in arterial rings in buffer with lucigenin yielded a massive oxygen-dependent peak of luminescence that decayed exponentially and was proportional to the degree of injury. Signals for injured vs. control arteries were 207. 1 +/- 17.9 (n = 13) vs 4.1 +/- 0.7 (n = 22) cpm x 10(3)/mg/min (p <. 001). Data obtained with 0.25 mmol/l lucigenin were validated with 0. 005-0.05 mmol/l lucigenin or the novel superoxide-sensitive probe coelenterazine (5 micromol/l). Gentle removal of endothelium prior to injury scarcely affected the amount of luminescence. Lucigenin signals were amplified 5- to 20-fold by exogenous NAD(P)H, and were >85% inhibited by diphenyliodonium (DPI, a flavoenzyme inhibitor). Antagonists of several other potential free radical sources, including xanthine oxidase, nitric oxide synthase, and mitochondrial electron transport, were without effect. Overdistension of intact rabbit iliac arteries in vivo (n = 7) induced 72% fall in intracellular reduced glutathione and 68% increase in oxidized glutathione, so that GSH/GSSG ratio changed from 7.93 +/- 2.14 to 0. 81 +/- 0.16 (p <.005). There was also 28.7% loss of the glutathione pool. Further studies were performed with electron paramagnetic resonance spectroscopy. Rabbit aortas submitted to ex vivo overdistension in the presence of the spin trap DEPMPO (5-diethoxy-phosphoryl-5-methyl-1-pyrroline-N-oxide, 100 mmol/l, n = 5) showed formation of radical adduct spectra, abolished by DPI or superoxide dismutase. Computer simulation indicated a mixture of hydroxyl and carbon-centered radical adducts, likely due to decay of superoxide adduct. Electrical mobility shift assays for NF-kappaB activation were performed in nuclear protein extracts from intact or previously injured rabbit aortas. Balloon injury induced early NF-kappaB activation, which was decreased by DPI. In conclusion, our data show unambiguously that arterial injury induces an immediate profound vascular oxidative stress. Such redox imbalance is likely accounted for by activation of vessel wall NAD(P)H oxidoreductase(s), generating radical species potentially involved in tissue repair.