The endothelium-derived relaxing factor is rapidly inactivated by superoxide radicals, and atherosclerotic vessels generate excess radical species. We tested the hypothesis that an imbalance between intrinsic superoxide dismutase (SOD) activity and the generation of superoxide radicals in atherosclerotic arteries may result in augmented inactivation of endothelium-derived relaxing factor. Vascular SOD was increased in normal and cholesterol-fed (1% cholesterol for 4 months) rabbits approximately twofold by treatment with polyethylene-glycolated SOD (PEG-SOD; 41,000 units/kg/day i.m.) for 1 week. Aortic rings from these animals and nontreated control and atherosclerotic rabbits subsequently were studied in organ chambers. Endothelium-dependent relaxations to acetylcholine and the calcium ionophore A23187 were improved by PEG-SOD in atherosclerotic but not in normal rabbits. PEG-SOD pretreatment did not alter endothelium-independent relaxations to nitroprusside. Thus, treatment with PEG-SOD can partially restore impaired endothelium-dependent relaxation of atherosclerotic arteries. We conclude that generation of oxygen-derived radicals likely contributes to endothelial dysfunction of atherosclerotic arteries.

Endothelium-derived relaxing factor (EDRF) is rapidly inactivated by radicals. Endothelial cells possess several antioxidant defense mechanisms. It is not clear which intrinsic antioxidant defense systems are important to preserve the release of biologically active EDRF. We impaired antioxidant defense in normal vascular tissue by inhibiting catalase activity with 3-amino-1,2,4-triazole (AT), superoxide dismutase with diethyldithiocarbamate (DETC), and by reducing glutathione content via inhibiting glutathione synthesis with L-buthionine-(S,R)-sulfoximine (BSO). Pretreatment of rabbit aorta in vitro with DETC markedly reduced endothelium-dependent relaxation in response to acetylcholine and calcium ionophore A23187 and, to a lesser extent, reduced endothelium-independent relaxation in response to nitroprusside. Pretreatment of cultured bovine aortic endothelial cells (BAEC) with DETC did not alter release of nitrogen oxides (measured by chemiluminescence), but, the effluent of pretreated cells showed marked depression in vasodilator activity (measured by bioassay). Pretreatment of rabbit aorta in vitro with AT did not alter endothelium-dependent and -independent relaxations. Pretreatment of BAEC with BSO did not alter the release of nitrogen oxides or the vasodilator activity. These results suggest that endothelial superoxide dismutase activity, but not catalase or glutathione, is necessary for the release of biologically active EDRF. An imbalance of the intrinsic superoxide dismutase and the production of superoxide anions may therefore predispose to impaired endothelium-dependent relaxations and alter vascular reactivity.

No abstract

Copper- and zinc-containing superoxide dismutase (CuZnSOD), manganese-containing superoxide dismutase (MnSOD), catalase (CAT), glutathione peroxidase (GPX, both Se-dependent and Se-independent), and glutathione reductase (GR) were measured in normal, nitrosoguanidine-transformed and SV40-transformed mouse liver cells in culture, as well as in mouse liver homogenates. Enzyme activities were compared on the basis of 3 different endpoints: per mg protein, per mg DNA, and per 10(6) cells. Except for GR, activity of all the measured anti-oxidant enzymes was much higher in vivo than in vitro. All of the anti-oxidant enzyme activities were lower in general in the 2 transformed cell lines than in the in vitro normal cell line, except Cu-ZnSOD, which showed little change. However, MnSOD was the only enzyme which showed lowered activity in both transformed cell lines, no matter what endpoint was used. This finding is in agreement with previous work showing lowered MnSOD activity in tumor cells.

A simple and sensitive method for the simultaneous visualization of glutathione peroxidase and catalase on polyacrylamide gels is described. The procedure included: (1) running samples on a 7.5% polyacrylamide gel, (2) soaking the gel in a certain concentration of reduced glutathione (0.25-2.0 mM), (3) soaking the gel in GSH plus H2O2 or cumene hydroperoxide, (4) finally staining with a 1% ferric chloride 1% potassium ferricyanide solution. The best concentration of glutathione for simultaneous visualization of glutathione peroxidase in mouse liver homogenates and also it is specific for glutathione peroxidase since other peroxidases such as lactoperoxidase, horseradish peroxidase and glutathione S-transferase cannot be visualized. Using this method, it was found that unlike catalase, glutathione peroxidase is heat resistant (68 degrees C, 1 min), but sensitive to 10 mM sodium iodoacetate.