Quinone drugs are used extensively as anti-neoplastic agents. The mechanism of their actions and the reasons for their unfavorable side effects are not well understood. Mitomycin C (MC) is an N-heterocyclic quinone with chemotherapeutic action against solid tumors. Previous research has led to the development of a model for drug activation involving NADPH reduction of the drug via microsomal mixed-function oxidases. We tested the possibility that NADPH is provided from the hexose monophosphate shunt (HMPS). The MC did indeed increase HMPS activity aerobically, while not affecting Kreb's cycle activity. Anaerobic stimulation of the shunt is also predicted by the model. However, under hypoxic conditions no HMPS or Kreb's activity was observed in MC-treated or untreated samples. Other investigators have documented the involvement of reactive oxygen species in microsomal systems in vitro. The oxygen requirement for MC stimulation of HMPS suggests oxygen radical involvement. We carried out experiments using [14C]-formate as a scavenger for hydrogen peroxide. There was no apparent change in H2O2 production when MC was added. Catalase is known to be involved in peroxide metabolism in vivo; however, addition of the catalase inhibitor sodium azide did not alter endogenous or MC-stimulated shunt activity. The microsomal inhibitor SKF-525A (10(-3) M) prevented MC stimulation of the HMPS, which is consistent with the model implicating microsomal enzymes in MC metabolism. Overall, we have shown the oxygen dependence of endogenous and MC-stimulated shunt activity, and the results provide evidence for MC activation of oxidative metabolism by a mechanism which involves microsomes.