Influence of DNA repair on nonlinear dose-responses for mutation

Abstract

Recent evidence has challenged the default assumption that all DNA-reactive alkylating agents exhibit a linear dose-response. Emerging evidence suggests that the model alkylating agents methyl- and ethylmethanesulfonate and methylnitrosourea (MNU) and ethylnitrosourea observe a nonlinear dose-response with a no observed genotoxic effect level (NOGEL). Follow-up mechanistic studies are essential to understand the mechanism of cellular tolerance and biological relevance of such NOGELs. MNU is one of the most mutagenic simple alkylators. Therefore, understanding the mechanism of mutation induction, following low-dose MNU treatment, sets precedence for weaker mutagenic alkylating agents. Here, we tested MNU at 10-fold lower concentrations than a previous study and report a NOGEL of 0.0075 µg/ml (72.8nM) in human lymphoblastoid cells, quantified through the hypoxanthine (guanine) phosphoribosyltransferase assay (OECD 476). Mechanistic studies reveal that the NOGEL is dependent upon repair of O(6)-methylguanine (O(6)MeG) by the suicide enzyme O(6)MeG-DNA methyltransferase (MGMT). Inactivation of MGMT sensitizes cells to MNU-induced mutagenesis and shifts the NOGEL to the left on the dose axis.

Fig. 1.

Nonlinear dose-response of MNU in AHH-1 cells quantified through the HPRT assay (solid line). RPE was used to assess toxicity (dotted line). *p < 0.05 using two-sided Dunnett’s test with square root transformed data based on Bartlett’s test for heterogeneity of variance (p > 0.05) and Kolmogorov-Smirnov test for normality (p > 0.05).

Fig. 2.

Frequency of all base substitutions found along the nontranscribed strand of HPRT sequence at increasing concentrations of MNU.

Fig. 2.

Frequency of all base substitutions found along the nontranscribed strand of HPRT sequence at increasing concentrations of MNU.

Fig. 2.

Frequency of all base substitutions found along the nontranscribed strand of HPRT sequence at increasing concentrations of MNU.

Fig. 3.

The proportion of GC → AT changes (bars) increases in concordance with the increase in HPRT MF (line) observed in Figure 1.

Fig. 4.

Linear regression outputs comparing linear to quadratic dose-responses for MNU-induced mutant frequencies in MGMT-active (A) and MGMT-inactive AHH-1 cells (B) at the NOGEL (0.0075 µg/ml) and below. The MGMT-active dose-response was quadratic below the NOGEL (p < 0.05) and displayed a negative gradient for the linear regression (p < 0.05) at −0.0007±0.0006 (A). The MGMT-inactive dose-response was linear (p = 0.15) and had a clearly positive gradient (p < 0.05) (B). Linear model (solid line), observed data (open circles), and quadratic model (dashed line).

Fig. 4.

Linear regression outputs comparing linear to quadratic dose-responses for MNU-induced mutant frequencies in MGMT-active (A) and MGMT-inactive AHH-1 cells (B) at the NOGEL (0.0075 µg/ml) and below. The MGMT-active dose-response was quadratic below the NOGEL (p < 0.05) and displayed a negative gradient for the linear regression (p < 0.05) at −0.0007±0.0006 (A). The MGMT-inactive dose-response was linear (p = 0.15) and had a clearly positive gradient (p < 0.05) (B). Linear model (solid line), observed data (open circles), and quadratic model (dashed line).

Fig. 5.

The fold change in MF in MGMT-active and -inactive AHH-1 cells plotted on a log-linear axis.