Evaluation of four methods for scoring cytoplasmic grains in the in vitro unscheduled DNA synthesis (UDS) assay

The in vitro unscheduled DNA synthesis (UDS) assay measures DNA repair (incorporation of [3H]thymidine) following in vitro treatment of rat primary hepatocytes. The autoradiographic method was used to detect UDS by counting developed silver grains in the photographic emulsion overlaying nuclei and cytoplasmic areas of the hepatocytes. In this communication we report results using 4 scoring methods: (1) the 2 most heavily labeled cytoplasmic areas adjacent to the nucleus (our standard method), (2) the cytoplasmic area left of the nucleus, (3) the cytoplasmic areas left and right of the nucleus, and (4) 2 cytoplasmic areas whose positions were selected at random. Rat primary hepatocyte cultures treated with a medium control, a solvent control (dimethyl sulfoxide) and 5 known genotoxic chemicals (2-acetylaminofluorene, dimethylnitrosamine, diethylnitrosamine, methyl methanesulfonate and ethyl methanesulfonate) were scored using these 4 methods. The average or maximum cytoplasmic grain count was subtracted from the nuclear grain count to yield net grains/nucleus (NG). In general, NG counts for Methods 2, 3 and 4 were similar, although shifted about 3-10 grains higher than Method 1 for controls and most treated groups. Methods 2, 3 and 4 showed more experiment-to-experiment variability in sensitivity for detecting statistically significant increases in treated groups than did our standard method. Thus, the alternative methods afforded no consistent improvements in sensitivity or reduction of variability for this assay. Subtraction of the average or the highest cytoplasmic count had virtually no effect on the sensitivity of the assay, but simply requires an appropriate adjustment of the criteria for a positive response.