Although active oxygen species play important roles in the pathogenesis of various diseases, the molecular mechanism for oxygen toxicity in vascular diseases remains to be elucidated. Since endothelium-derived relaxing factor (EDRF) is inactivated by superoxide radicals in vitro, oxidative stress in and around vascular endothelial cells may affect the circulatory status of animals. To study the role of superoxide radicals and related enzymes, such as superoxide dismutase (SOD), in vascular diseases, we have developed a fusion protein (HB-SOD) consisting of human Cu/Zn-type SOD and a C-terminal basic peptide with high affinity for heparan sulfate on endothelial cells. When injected intravenously, HB-SOD bound to vascular endothelial cells, underwent transcellular transport, and localized within vascular walls by a heparin-inhibitable mechanism. The blood pressure of spontaneously hypertensive rats (SHR) but not normal animals was decreased significantly by HB-SOD. Heparin inhibited the depressor effect of HB-SOD. In contrast, native SOD had no effect on blood pressure of either SHR or normal rats. Neither H2O2-inactivated HB-SOD nor the C-terminal heparin-binding peptide showed such a depressor effect, suggesting that the catalytic function of HB-SOD is responsible for its depressor action. To know the source of superoxide radicals, we determined xanthine oxidase activity in the aorta and uric acid levels in the plasma. Although no appreciable difference in xanthine oxidase activity was found between the two animal groups, uric acid levels were significantly higher in SHR than in normal rats. Oxypurinol, a potent inhibitor of xanthine oxidase, also decreased the blood pressure of SHR but not of normal rats. These findings indicate that superoxide radicals in and around vascular endothelial cells play critical roles in the pathogenesis of hypertension of SHR.