A new technique, high-performance liquid chromatography with reductive mode electrochemical detection on a mercury drop (HPLC-EC), has been used for analyzing lipid hydroperoxide (LOOH) formation in photooxidatively stressed L1210 leukemia cells. Highly specific and sensitive for peroxides (detection limits < 0.5 pmol for cholesterol hydroperoxides and < 50 pmol for phospholipid hydroperoxides), this approach allows different classes of LOOH to be separated and determined in minimally damaged cells. L1210 cells in serum-containing growth medium were irradiated in the presence of merocyanine 540 (MC540), a lipophilic photosensitizing dye. Lipid extracts from cells exposed to a light fluence of 0.11 J/cm2 (which reduced clonally assessed survival by 30%) showed 12-15 well-defined peaks in HPLC-EC. None of these peaks was observed when cells were irradiated without MC540 or when dye/light-treated samples were reduced with triphenylphosphine prior to analysis. Three peaks of relatively low retention time (< 12 min) were assigned to the following species by virtue of comigration with authentic standards: 3 beta-hydroxy-5 alpha-cholest-6-ene-5-hydroperoxide (5 alpha-OOH), 3 beta-hydroxycholest-4-ene-6 beta-hydroperoxide (6 beta-OOH), and 3 beta-hydroxycholest-5-ene-7 alpha/7 beta-hydroperoxide (7 alpha/7 beta-OOH). Formation of 5 alpha-OOH and 6 beta-OOH (single oxygen adducts) was confirmed by subjecting [14C]cholesterol-labeled cells to relatively high levels of photooxidation and analyzing extracted lipids by HPLC with radiochemical detection. Material represented in a major peak at 18-22 min on HPLC-EC was isolated in relatively large amounts by semipreparative HPLC and shown to contain phospholipid hydroperoxides (predominantly phosphatidylcholine species, PCOOH) according to the following criteria: (i) decay of 18-22 min peak during Ca2+/phospholipase A2 treatment, with reciprocal appearance of fatty acid hydroperoxides; (ii) reduction of peroxide during treatment with reduced glutathione and phospholipid hydroperoxide glutathione peroxidase, but not glutathione peroxidase; and (iii) comigration with PCOOH standards in thin-layer chromatography. HPLC-EC analysis revealed quantifiable amounts of PCOOH and ChOOH at a light fluence that clonally inactivated < 10% of the cells, which allows for the possibility that photoperoxidative damage plays a causal role in cell killing.