A critique of methods to measure cytotoxicity in mammalian cell genotoxicity assays

Here we show the results of comparing cell viability, cytotoxicity, and apoptosis assays for measuring the time- and dose-dependent toxic effects of tamoxifen on HepG2 cells. The quantitation of adenosine 5'-triphosphate (ATP), 5-(3-carboxymethoxyphenyl)-2-(4,5- dimethylthiazolyl)-3-(4-sulfophenyl) tetrazolium, inner salt (MTS) tetrazolium reduction, and resazurin reduction methods used to estimate the number of viable cells all showed a similar trend of decreased cell viability after longer periods of tamoxifen exposure to HepG2 cells. The release of lactate dehydrogenase (LDH) as a marker for cells with a compromised membrane and the increase in caspase-3/7 activity as a marker for apoptosis were both shown to increase using the same tamoxifen exposure conditions that caused a decrease in HepG2 cell viability. The longer the duration of exposure of tamoxifen, the lower the concentration required to kill or induce apoptosis in HepG2 cells. In contrast, there was no change in LDH release from HL-60 cells using conditions of vinblastine treatment that caused an increase in caspase activity and a decrease in ATP content, suggesting a difference in the mechanism of cell death between the two model systems. Both the density of parent stock cultures used as a source of cells to prepare assay plates and the density of cells per well in the assay plates were demonstrated to be factors than can influence the apparent potency of a toxin in viability, toxicity, and apoptosis assays. These results illustrate the importance of understanding the kinetics and mechanism of cell death of each in vitro model system as prerequisites for choosing the most appropriate assay method.

International guidelines for cytotoxicity limits for the in vitro chromosomal aberration assay require reductions in cell growth of greater than 50%. This sets no upper limit on toxicity and there is concern about the number of false or irrelevant results obtained in the aberration assay, i.e., positive results at toxic dose levels only, with no evidence for primary DNA damaging ability and with negative results in the other genotoxicity tests. We have previously proposed that no truly genotoxic compound would be missed if the toxicity of the highest dose did not exceed 50%. Cell growth measured by cell counts as a percentage of controls can underestimate toxicity. For example, if we seed half a million cells per culture, and the controls double to 1 million during the experiment, a culture that truly has no growth will still have a cell count 50% of the control. Measurement of population doublings (PDs) more accurately assesses cell growth. To assess the use of PD in dose selection, we examined previous data from this lab and data from new experiments with "true," primary DNA damaging clastogens, and with clastogens, including drugs, thought to act indirectly, through cytotoxicity-associated mechanisms. We compared aberration results where the highest doses scored were based on 50% reductions in final cell counts with results obtained when the highest doses were based on PD. The PD method allows detection of true clastogens, including those that are active in a range with some toxicity, and reduces the number of toxicity-related "false"-positive results. Copyright 2004 Wiley-Liss, Inc.