Epoxides metabolically produced from some known carcinogens and from some clinically used drugs. I. Differences in mutagenicity.

The epoxide metabolites of two clinically used drugs and an experimental psychotropic agent, carbamazepine 10,11-oxide, cyproheptadine 10,11-oxide and cyclobenzaprine 10,11-oxide, were fully devoid of any mutagenic activity under conditions where K-region-epoxide metabolites of some known carcinogens, such as benzo(a)pyrene, proved to be potent frameshift mutational agents for Salmonella typhimurium TA 1537 and TA 1538. All epoxides tested were non-mutagenic for TA 1535, designed to detect substitution mutations. The 10,11-epoxides of the three drugs, carbamazepine, cyproheptadine and cyclobenzaprine, were not cytotoxic to any of the bacterial tester strains used, precluding that mutagenicity might have been overshadowed by cytotoxicity. When the mutagen, precursor, benzo(a)pyrene, was incubated together with TA 1537 and a mammalian microsomal preparation in the presence of a system generating the co-factor necessary for mono-oxygenase activity, activation to mutagenic species was observed which was dramatically increased in the presence of a potent epoxide hydratase inhibitor, 1,1,1-trichloropropene 2,3-oxide, suggesting epoxide(s) as the (or one of the) mutagenically active species metabolically produced in situ. None of these effects was observed with the three medical drugs. Moreover, the observation that the alkene oxide 4-phenylstyrene 7,8-oxide is mutagenic to the two strains TA 1537 and TA 1538 but the K-region arene oxide derived from 7,12-dimethylbenz(a)anthracene is inactive for the latter strain indicates that epoxidation of an aromatic double bond of a polycyclic hydrocarbon is neither a necessary nor a satisfying condition for frameshift mutagenesis to occur.