2,6-Dithiopurine (DTP) has been proposed as a possible chemopreventive agent because of its ability to react with electrophiles. Acrolein, an electrophilic metabolite of cyclophosphamide (CP) involved in the toxicities of this anticancer drug, can be scavenged by DTP. The present study examined the effect of DTP treatment on CP-mediated bladder and lung toxicity in male ICR mice. Mice fed a diet containing 4% DTP that were treated intraperitoneally (i.p.) with 350 mg/kg CP showed no significant bladder damage (measured as bladder blood content at 48 h) with respect to the group fed a control diet. DTP (50 and 100 mg/kg), given i.p. 0.5 and 7 h after the initial injection of CP, also prevented the bladder damage when compared with the group receiving CP alone. Surprisingly, although neither parenteral CP nor DTP alone caused any mortality at these doses, the combined treatment resulted in 67% mortality within 3 days. At 24 h after CP + DTP, blood urea nitrogen was elevated 6-fold and urine volumes decreased by 70%. Histopathological analyses revealed a diffuse myocardial degeneration and necrosis, severe granular degeneration in the liver, abundant cellularity and infiltrates in interalveolar spaces in the lung and swollen nephron epithelial cells with some necrosis. All mice survived treatment when the dose of CP was lowered to 250 and 25-75 mg/kg DTP was given i.p. 0.5 and 7 h after CP. These DTP regimens reduced the degree of CP-induced lung toxicity, measured by [3H]thymidine incorporation into lung DNA 7 days after CP, in a dose-dependent manner. DTP (75 mg/kg) also reduced CP-induced lung fibrosis estimated by lung hydroxyproline content 28 days after CP. Analyses of urine from mice given CP + DTP revealed large amounts of the metabolic product dithiouric acid, smaller amounts of the parent DTP and several smaller peaks. The major unique metabolite peak was collected and analyzed by mass spectrometry, but did not correspond to either acrolein-DTP or acrolein-dithiouric acid. Thus, either very small amounts of an acrolein adduct are generated, the adduct is broken down to an unidentified product, or the ability of DTP to prevent CP-induced lung and bladder damage is related to some other mechanism. The possibility that mercapturic acid metabolites of acrolein released the parent electrophile in the urine was not supported by the finding that probenecid did not prevent CP-induced bladder toxicity.