Evidence for hormesis in mutagenicity dose-response relationships.

This study assessed the occurrence of hormetic dose responses from three previously published data sets [1-3] with 825 chemicals in three Ames assay tester strains (i.e., TA97, TA98, TA100) with and without the S9 fraction, using a five dose protocol and semi-log dose spacing. Ninety-five (95) (11.5%) chemicals satisfied the multiple a priori entry criteria, with a total of 107 assays. Of the assays satisfying the entry criteria, 61 involved TA100, a strain that detects base-pair substitution mutations. 29.5% (18/61) satisfied the statistical evaluative criteria for hormesis, exceeding that predicted by chance by 4.0-fold (p<0.001). The remaining 46 assays involved TA97 and TA98, strains that detect frameshift mutations. Of these 46 assays, the overall responses for the lowest two doses closely approximated the control response (e.g., 101.77% of the control for TA98; 99.20% for TA97). Only 2.2% (1/46) of the assays satisfied the evaluative criteria for hormesis. In conclusion, these data support a hormetic model for TA100, whereas the responses for TA97 and TA98 are consistent with a threshold dose-response model.