Development of a standard protocol for in vitro cytogenetic testing with Chinese hamster ovary cells: comparison of results for 22 compounds in two laboratories.

A major problem of cytogenetics testing in mammalian cells is lack of agreement of results among laboratories. Our objective was to develop a sensitive in vitro test protocol that was applicable to large-scale chemical screening and yielded comparable results in two laboratories. We used sister chromatid exchange (SCE) and chromosome aberration (CAb) tests in Chinese hamster ovary (CHO) cells. The initial protocol used standard cell densities, medium, batch of rat liver S9 for metabolic activation; positive, negative, and solvent controls; staining and scoring techniques; and fixation times. Treatment without S9 was for 8-12 hr (CAb) or 26 hr (SCE), and with S9 for 2 hr in serum-free medium. Bromodeoxyuridine (BrdUrd) (10 microM) was added to SCE cultures only, 2 hr after addition of the test chemical. Doses were based on the 50% toxicity level in a preliminary test of cell survival 24 hr after treatment. One hundred cells (CAb) or 50 cells (SCE) were scored from each control and from five dose levels. Five clastogens were tested in the first two-laboratory comparison: mitomycin-C, triethylenemelamine, N-methyl-N'-nitro-N-nitrosoguanidine, cyclophosphamide, and benzo(alpha)pyrene. There was quite good agreement between laboratories. Seventeen compounds were then tested "blind" in the two laboratories. As testing proceeded, some discrepancies occurred between the laboratories, and the protocol was modified in attempts to improve the resolution of marginal responses and make dose selection more consistent. The preliminary test for cell survival was omitted. A 10(5) dose range in a half-log series was tested, and cells were scored at the highest dose at which sufficient mitotic cells were obtained, and at the next two lower doses. By delaying fixation times, SCE and CAb were scored at doses that inhibited cell cycle progression. This protocol gave comparable results in the two laboratories in many cases and by testing up to a maximum dose, limited by solubility and/or toxicity, should detect a high proportion of clastogens and SCE inducers.