Genotoxicity of trophosphamide in mouse germ cells: assessment of micronuclei in spermatids and chromosome aberrations in one-cell zygotes

Abstract

The genotoxicity of trophosphamide (TP) in mouse germ cells was assessed by the cytogenetic analysis of micronuclei in spermatids and chromosome aberrations in one-cell zygotes and compared with the genotoxicity in somatic cells evaluated by the micronucleus reticulocyte assay. Single acute doses of 50, 75, 100 and 150 mg/kg were studied after i.p. injection. TP was only weakly mutagenic for preleptotene spermatocytes-differentiating spermatogonia, but clear-cut cytotoxic effects were demonstrated after treatment of these cells by a dose-dependent reduction of the ratio between Golgi and cap phase spermatids. Effects induced in post-meiotic stages were estimated, after mating the treated males with untreated superovulated females, by the frequencies of zygotes with chromosome aberrations: a peak of genetic damage was detected in late spermatids, with as many as 55% zygotes with aberrations, but spermatozoa and early spermatids were also clearly affected. When compared with matched solvent-injected controls, the lowest effective dose in spermatozoa and late spermatids was 100 mg/kg, although the 3- to 4-fold increases detected at 50 mg/kg were also statistically significant when compared with a pool of laboratory controls. In peripheral blood reticulocytes, the micronucleus frequencies were increased by 3-20 times the respective baseline values in the individual animals. A marked cytotoxic effect on bone marrow cells was revealed by the reduction of the proportion of early reticulocyte stages, which dropped to 20% of the control value at 150 mg/kg. Both genotoxic and cytotoxic effects were higher in bone marrow than in germ cells of the same animals, pointing to a generalized higher susceptibility of somatic cells to TP, possibly related to chemical distribution and target organ accessibility. The accurate description of stage- and dose-effect relationships in germ cells of experimental models is crucial for genetic risk assessment after chemical exposure. The approaches applied in this study may contribute to this goal.