The anti-estrogen toremifen-Fc-1157a or 4-chloro-1,2-diphenyl-1-[4-[2(N,N-dimethylamino)ethoxy]-phenyl]-1- butene is now used for the treatment of breast cancer. This drug is extensively metabolized by cytochrome P450 dependent hepatic mixed function oxidase in man, yielding mainly the N-demethyl-(DMTOR), 4-hydroxy-(4OH-TOR) and deamino-hydroxy-(TOR III) toremifene metabolites. The specific forms of cytochrome P450 involved in these oxidation reactions were examined in 32 human liver microsomal preparations previously characterized with respect to their contents of several known P450 enzymes. Toremifene was demethylated with an apparent Km of 124 microM while it was hydroxylated with an apparent Km of 139 microM. The metabolic rates were 71 +/- 56, 13 +/- 9 and 15 +/- 4 pmol/min/mg microsomal protein, respectively, for DMTOR, 4-OH-TOR and TOR III. The N-demethylation activity was strongly correlated with estradiol 2-hydroxylation (r = 0.75), nifedipine oxidation (r = 0.86), tamoxifen N-demethylation (r = 0.73), testosterone 6 beta-hydroxylation (r = 0.78) and erythromycin N-demethylation (r = 0.84), all these monooxygenase activities known to be supported by CYP3A4 isoform. Furthermore, the CYP3A content of liver microsomal samples, measured by western blot analysis using a monoclonal anti-human CYP3A4 antibody, was strongly correlated with DMTOR formation (r = 0.80). Compounds such as cyclosporin, triacetyl-oleandomycin and testosterone inhibited the N-demethylation of toremifene metabolism at 80, 89 and 56% vs control, respectively, while the formation of TOR III was inhibited at 78, 82 and 73% vs control and the 4-hydroxylation pathway was inhibited no more than about 50% vs control. Prior incubation of microsomes with 100 microM gestodene, known to be a selective mechanism-based inhibitor of CYP3A4 in the presence of NADPH, led to 76 +/- 6 and 76 +/- 5% (N = 5 samples) reductions in the N-demethylation and formation of TOR III, respectively. Polyclonal antibody directed against human CYP3A enzymes inhibited formation of DMTOR and TOR III by 60 and 46%, respectively. The metabolism of toremifene was not activated by alpha-naphthoflavone. Finally, the use of yeasts genetically engineered for expression of human P4501A1, 1A2, 2C9 and 3A4 allowed us to demonstrate that DMTOR and TOR III formations are mediated by P4501A and 3A4 enzymes and by contrast these enzymes are not involved in the 4-hydroxylation pathway.(ABSTRACT TRUNCATED AT 400 WORDS)