The induction of unscheduled DNA synthesis (UDS) in rat primary hepatocytes is used as an in vitro screen for genotoxicity. The purpose of this study was to develop criteria for statistical evaluation of autoradiographic grain counts from the UDS assay. Based on our historical solvent control population (N = 17), we first established a predictive two-sided symmetric tolerance interval providing 95% confidence for coverage of 99% of the distribution of the mean net nuclear grain count for a new solvent control observation. The tolerance interval was determined to run from -2.82 to 4.03 net nuclear grains. When the concurrent solvent control (C) proves admissible (-2.82 less than C less than 4.03), the mean net nuclear grain counts from treatment groups are compared to the concurrent solvent control value. If the difference exceeds 3.09 at two successive concentrations, the test compound is considered positive. We have tested eight carcinogens (MNNG, AFB1, 4DAB, 4NQO, DMBA, 2AAF, DEN and 2AA) utilizing these criteria. All eight were positive. Four non-carcinogens (2AB, pyrene, anthracene and alpha NA) were negative. Each of these chemicals was tested twice using our standard protocol, which employs a preliminary range-finding cytotoxicity assay. We conclude that these criteria provide adequate sensitivity and specificity to be applicable for the evaluation of the results of routine UDS assays with compounds of unknown genotoxic potential.
Read full abstract