Selenoperoxidase-dependent glutathione cycle activity in peroxide-challenged leukemia cells.

Murine leukemia L1210 cells rendered deficient in glutathione peroxidase (GPX) and phospholipid hydroperoxide glutathione peroxidase (PHGPX) by Se deprivation (L.Se(-) cells) were found to be more sensitive to tert-butyl hydroperoxide (t-BuOOH) cytotoxicity than Se-replete controls (L.Se(+) cells). Human K562 cells, which express PHGPX, but not GPX, were also more sensitive to t-BuOOH in the Se-deficient (K.Se(-)) than Se-satisfied (K.Se(+)) condition. In examining the metabolic basis for selenoperoxidase-dependent resistance, we found that glucose-replete Se(-) cells reduce t-BuOOH to t-butanol far more slowly than Se(+) cells, the ratio of the first-order rate constants approximating that of the GPX activities (L1210 cells) or PHGPX activities (K562 cells). Monitoring peroxide-induced changes in GSH and GSSG gave consistent results; e.g., glucose-depleted L.Se(+) cells exhibited a first order loss of GSH that was substantially faster than that of glucose-depleted L.Se(-) cells. Under the conditions used, peroxide-induced conversion of GSH to GSSG could be stoichiometrically reversed by resupplying D-glucose, indicating that no significant lysis or GSSG efflux and/or interchange had taken place. The apparent first-order rate constant for GSH decay increased progressively for L1210 cells expressing a range of GPX activities from approximately 5% to 100%, demonstrating that peroxide detoxification is strictly dependent on enzyme content. The initial rate of 14CO2 release from D-[1-14C]glucose supplied in the medium was much greater for L.Se(+) or K.Se(+) cells than for their respective Se(-) counterparts, consistent with greater hexose monophosphate shunt activity in the former. These results highlight the importance of selenoperoxidase action in the glutathione cycle as a means by which tumor cells cope with hydroperoxide stress.