Vascular free radical release. Ex vivo and in vivo evidence for a flow-dependent endothelial mechanism.

Mechanisms underlying production of vascular free radicals are unclear. We hypothesized that changes in blood flow might serve as a physiological stimulus for endothelial free radical release. Intact isolated aortas from 45 rabbits were perfused with the spin trap alpha-phenyl-N-tert-butylnitrone (PBN, 20 mmol/L) and formed radical adducts detected by electron paramagnetic resonance spectroscopy (EPR). Sequential perfusion at 2, 7.5, and 12 mL/min changed cumulative vascular PBN radical adduct yields, respectively, from 3.2 +/- 0.9 to 4.1 +/- 0.7 (P < .05) and 7.0 +/- 1.5 (P < .005) pmol/mg with endothelium and from 3.6 +/- 1.6 to 3.8 +/- 1.4 and 2.2 +/- 0.8 pmol/mg without endothelium (P = NS). In endothelialized aortas, superoxide dismutase (SOD) completely blocked flow-induced free radical production, whereas inactivated SOD, indomethacin, and the nitric oxide synthetase antagonist nitro-L-arginine methyl ester (L-NAME) had no effect; relaxations to acetylcholine remained unchanged with higher flows. To assess the role of flow on in vivo radical production, femoral arterial plasma levels of the ascorbyl radical, a stable ascorbate oxidation product, were measured by direct EPR in 56 other rabbits. Ascorbyl levels were assessed at baseline (30.2 +/- 0.7 nmol/L) and at peak-induced iliac flow changes. Flow increases from 25% to 100% due to saline injections through an extracorporeal aortic loop induced significant dose-dependent increases in ascorbyl levels (n = 5). In addition, after papaverine bolus injections, flow increased by 114 +/- 8% versus baseline, and ascorbyl levels increased by 5.4 +/- 0.7 nmol/L (n = 31, P < .001); similar results occurred with adenosine, isoproterenol, or hyperemia after 30-second occlusions (P < .05, n = 4 or 5 in each group). Active SOD completely blocked papaverine-induced ascorbyl radical increase, despite preserved flow response (delta ascorbyl = 0.02 +/- 1.6 nmol/L, P = NS); inactivated SOD, catalase, indomethacin, and L-NAME had no effect. Blood flow decreases of 65% to 100% due to phenylephrine or 60-second balloon occlusions were accompanied by an average decrease of 4.4 nmol/L (P < .05) in ascorbyl levels. No change in ascorbyl signal was observed when rabbit blood alone was submitted to in vitro flow increases through a tubing circuit. Thus, increases in blood flow trigger vascular free radical generation; such a response seems to involve endothelium-derived superoxide radicals unrelated to cyclooxygenase or nitric oxide synthetase activities. This mechanism may contribute to explain vascular free radical generation in physiological or pathological circumstances.