Clastogenic and aneugenic effects of tamoxifen and some of its analogues in hepatocytes from dosed rats and in human lymphoblastoid cells transfected with human P450 cDNAs (MCL-5 cells).

Tamoxifen and its analogues 4-hydroxytamoxifen, toremifene, 4-hydroxytoremifene, clomifene and droloxifene were tested for clastogenic effects in a human lymphoblastoid cell line (MCL-5) expressing elevated native CYP1A1 and containing transfected CYP1A2, CYP2A6, CYP2E1 and CYP3A4 and epoxide hydrolase and in a cell line containing only the viral vector (Ho1). MCL-5 or Ho1 cells were incubated with 4-hydroxytamoxifen, 4-hydroxytoremifene, clomifene or droloxifene and the incidence of micronuclei estimated. With MCL-5 cells there was an increase in micronuclei with 4-hydroxytamoxifen, 4-hydroxytoremifene and clomifene but not with droloxifene. With Ho1 cells only 4-hydroxytamoxifen and 4-hydroxytoremifene caused an increase in micronuclei. MCL-5 cells were incubated with tamoxifen, 4-hydroxytamoxifen, toremifene, droloxifene, clomifene or diethylstilbestrol (0.25-10 microg/ml) for 48 h and subjected to 3 h treatment with vinblastine (0.25 microg/ml) to arrest cells in metaphase. The incidence of cells with chromosomal numerical aberrations (aneuploidy) was increased in cells treated with tamoxifen, 4-hydroxytamoxifen, toremifene, clomifene and diethylstilbestrol but not droloxifene. The frequency of cells with structural abnormalities (excluding gaps) was increased in cells treated with tamoxifen and toremifene but not 4-hydroxytamoxifen, clomifene, droloxifene or diethylstilbestrol. The clastogenic activities of tamoxifen (35 mg/kg), toremifene (36.3 mg/kg), droloxifene (35.2 mg/kg) and diethylstilbestrol (25 mg/kg) were compared in groups of four female Wistar rats. Each chemical was dissolved in glycerol formal, administered as a single dose by gavage and hepatocytes isolated by collagenase perfusion 24 h later. The cells were cultured in the presence of epidermal growth factor (40 ng/ml) for 48 h, colchicine (10 microg/ml) being added for the final 3 h of incubation. At least 100 chromosomal spreads were examined from each animal for the presence of numerical and structural abnormalities. The incidences of aneuploidy following treatment were: tamoxifen 81%, toremifene 46%, droloxifene 9.6%, diethylstilbestrol 45.7%, vehicle control 5.3%. The incidences of chromosomal structural abnormalities excluding gaps were: tamoxifen 4.3%, toremifene 0.8%, droloxifene 0.5%, diethylstilbestrol 0.8%, control 0.5%. The incidence of chromosomal structural aberrations excluding gaps in the treated animals was not statistically significantly different from controls except in the tamoxifen-treated group. Tamoxifen (35 mg/kg per os) and toremifene (36.3 mg/kg per os) were dosed to rats for 4 weeks and chromosomal spreads made from hepatocytes. The incidences of aneuploidy were: tamoxifen 94%, toremifene 57%, control 6.5%. The incidences of chromosomal aberrations excluding gaps were: tamoxifen 12%, toremifene 1%, control 0.5%. The incidence of tamoxifen-induced chromosomal structural abnormalities was significantly elevated compared with control levels. The results demonstrate that tamoxifen and toremifene are the only two drugs tested in the study that cause chromosomal structural and numerical aberrations in vitro and tamoxifen is the only drug that induces both these effects in rat liver cells stimulated to divide in culture following oral dosing. Since chromosomal mutations require cell division for their manifestation and tamoxifen is the only compound of those tested that causes hyperplasia in the rat liver, chromosomal aberrations and aneuploidy in the rat liver would only be expected to occur following treatment with tamoxifen alone, although aneuploidy could be induced by toremifene in conjunction with a promoter such as phenobarbitone.