Merbarone (MB), a nonsedating derivative of thiobarbituric acid, wa recently found to induce profound hypouricemia. When incubated with xanthine oxidase (XO) and hypoxanthine in vitro, MB is both an inhibitor of XO and degraded by the XO-hypoxanthine interaction. Compared with allopurinol (K(i) = 0.025 muM), MB is a very weak inhibitor of XO (K(i) = 51 +/- 8 muM). MB interacts with XO in the presence of hypoxanthine to yield three chromatographically separate products. One of these products has been identified by HPLC retention time and spectral characteristics as 2-oxo-2-desthiomerbarone (2-oxo-MB). The other two products are thought to be S-oxide intermediates in the oxidative desulfuration of this drug. Formation of these products was blocked by catalase, suggesting that the conversion was dependent on reactive oxygen species (especially H2O2) generated by the hypoxanthine-XO system. This suggestion was confirmed by incubating MB with H2O2. In vitro studies with rat liver microsomes have documented the formation of 2-oxo-MB and 4'-OH-MB (4'-OH-MB), the latter being identified by the characteristic HPLC retention time of its acetylated derivative. The formation of 4'-OH-MB has many characteristics of a cytochrome P-450-dependent monooxygenase reaction (NADPH requirement and SKF 525-A inhibition); formation of 2-oxo-MB occurs by a different mechanism that is, as yet, uncharacterized. Incubation of kidney microsomes with MB generated 2-oxo-desthiomerbarone but no detectable 4'-OH-MB. These results indicate that MB is a weak inhibitor of XO, and that reactive oxygen species generated by cytochrome P-450 and or by flavin-containing enzymes such as XO play a key role in the oxidative desulfuration to 2-oxo-MB. The hydroxylation of MB to 4'-OH-MB by liver microsomes is, however, mediated by NADPH and cytochrome P-450-dependent metabolism only. The pertinence of these in vitro observations to the metabolism, antitumor, nephrotoxicity, and hypouricemic effects of MB is discussed.