Murine leukemia L1210 cells grown for 5-7 d in the presence of 1% serum without added selenium [Se(-) cells] expressed < 5% of the glutathione peroxidase (GPX) activity of selenium-supplemented controls [Se(+) cells]. Clonogenic survival assays indicated that t-butyl hydroperoxide (t-BuOOH) is much more toxic to Se(-) cells (LC50 approximately 10 muM) than to Se(+) or selenium-repleted [Se(-/+)] cells (LC50 approximately 250 muM). Hypersensitivity of Se(-) cells to t-BuOOH was partially reversed by treating them with Ebselen, a selenoperoxidase mimetic; thus, selenoperoxidase insufficiency was probably the most serious defect of Se deprivation. Cytotoxicity of t-BuOOH was inhibited by desferrioxamine and by alpha-tocopherol, indicating that redox iron and free radical intermediates are involved. Elevated sensitivity of Se(-) cells to t-BuOOH was accompanied by an increased susceptibility to free radical lipid peroxidation, which became even more pronounced in cells that had been grown in arachidonate (20:4, n - 6) supplemented media. That glutathione (GSH) is required for cytoprotection was established by showing that Se(+) cells are less resistant to t-BuOOH after exposure to buthionine sulfoximine (BSO), an inhibitor of GSH synthesis, or 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU), an inhibitor of glutathione reductase. Coupled enzymatic assays indicated that Se(+) or Se(-/+) cells metabolize t-BuOOH 20-25 times more rapidly than Se(-), consistent with the measured difference in GPX activities of these cells. Correspondingly, when challenged with t-BuOOH, Se(+) cells showed an initial loss of GSH and elevation of GSSG that exceeded that of Se(-) cells. It was further shown that like Se(-) cells, BSO- or BCNU-treated Se(+) cells metabolize t-BuOOH more slowly than nontreated controls. These results clearly indicate that selenoperoxidase action in the glutathione cycle is a vital element in cellular defense against toxic hydroperoxides.