The dynamics and mechanisms of extracellular release of hydrogen peroxide (H2O2) from bovine pulmonary artery endothelial cells (EC) subjected to anoxia, hypoxia, and hypoxia followed by reoxygenation were examined using various inhibitors of enzymatic systems in intact cells and by direct measurement of H2O2 production from isolated EC plasma membranes. Extracellular H2O2 was measured with a fluorometric assay. EC exposed to hypoxia (3% O-2) and anoxia (0% O-2) released less H2O2 (29.6 +/- 1.3% and 4.2 +/- 0.7%, respectively) compared with EC exposed to normoxia (20% O-2). The extracellular release of H2O2 from EC previously exposed to hypoxia for 24 h increased immediately after reoxygenation (20% O-2) to 272 +/- 48%, as compared with EC exposed continuously to normoxia (100% release), Inhibition of xanthine oxidase (XO) by allopurinol did not reduce the release of H2O2 from cells exposed to normoxia or hypoxia followed by reoxygenation. Furthermore, inhibitors of cyclooxygenase (indomethacin), phospholipase A(2) (quinacrine and chlorpromazine), nitric oxide synthase (L-arginine analogs), the mitochondrial electron transport chain (rotenone and cyanide), and cytochrome P-450 (methoxypsoralen) had no or minimal effect on this release. On the other hand, inhibitors of protein kinase C (calphostin and staurosporine) and NADPH oxidase (diphenyliodonium) reduced the release of H2O2 from EC in a dose-dependent manner in both exposure groups. In separate experiments, plasma membranes isolated from EC were found to produce H2O2 in the presence of NADH or NADPH as electron donors, This was inhibited by diphenyliodonium but not by allopurinol. These results suggest that the baseline production of H2O2 by EC and the increase in H2O2 released from EC in response to hypoxia followed by reoxygenation is not due to XO activity, but rather to a membrane-bound enzyme similar to NADPH oxidase. This effect may be mediated through protein kinase C activity.