Non-linear dose–response of DNA-reactive genotoxins: Recommendations for data analysis

Until recently, there has only been a limited amount of data available on the kinetics of mutation induction in the low dose region of exposure. In our publication Doak et al. [S.H. Doak, G.J. Jenkins, G.E. Johnson, E. Quick, E.M. Parry, J.M. Parry, Mechanistic influences for mutation induction curves after exposure to DNA-reactive carcinogens, Cancer Res. 67 (2007) 3904–3911] we showed that the two alkylating agents methyl-methanesulfonate (MMS) and ethyl-methanesulfonate (EMS) possess non-linear dose–response curves with no observed effect levels (NOEL) for mutation or chromosomal damage in vitro. These experiments were carried out in the AHH-1 human lymphoblastoid cell line, using the hypoxanthine phosphoribosyl transferase (HPRT) assay and the cytokinesis-block micronucleus (CBMN) assay, respectively. We have now carried out more advanced statistical analyses to define threshold values, which is critical as it has a dramatic impact on hazard and risk assessment. To do this, we re-analysed the data to see if the linear model or a more complex model (hockey stick or quadratic) gave a significant better fit of the data. For both EMS and MMS cytokinesis-block micronucleus data sets, the hockey stick model gave the most significant fit. The same was true for EMS, MMS and surprisingly ethylnitrosourea (ENU) in the HPRT assay in human AHH-1 cells. However, methylnitrosourea (MNU) was linear in both assays. These further analyses have shown that EMS and MMS have clear thresholds for both gene mutation and chromosome damage, as does ENU for gene mutation in AHH-1 cells. MNU was linear for gene and chromosome mutation and so was ENU for chromosome mutations at the concentrations tested. These findings correlate closely with those in vivo findings of Gocke et al. [E. Gocke, L. Müller, In vivo studies in the mouse to define a threshold for the genotoxicity of EMS and ENU, Mutat. Res. (this issue)] and together these data show a true threshold for EMS both in vitro and in vivo. In this report, we will discuss the approaches that were taken to investigate potential threshold dose–response curves for DNA-reactive genotoxic compounds, with recommendations for further studies.