Alkylation, DNA leads to Increase, Mutations

Alkylating agents can cause a variety of adducts and DNA damage (e.g., alkali labile sites, DNA strand breaks, etc.) that are potentially mutagenic and clastogenic. This KER focuses on the probability that an alkyl DNA adduct will lead to a mutation.

Not all adducts are equally mutagenic. Very generally, chemicals that preferentially cause O-alkylation in DNA induce DNA sequence changes, whereas chemicals that cause N-alkylation of DNA are more efficient inducers of structural chromosomal aberrations (reviewed in Beranek 1990). Indeed, a review of the biological significance of N7 alkyl-guanine adducts concluded that these adducts simply be used to confirm exposure to target tissue (Boysen et al., 2009), because the vast majority of studies shows that these adducts do not cause mispairing. A variety of work has demonstrated that N7-alkylguanine adducts can be bypassed essentially error free (e.g., Philippin et al., 2014; Shrivastav et al., 2010). Moreover, alkylation can involve modification with different sizes of alkylation groups (e.g., methyl, ethyl, propyl). Although response to these is qualitatively similar with respect to the key events, in general, larger alkylating groups tend to be more mutagenic (Beranek, 1990). It is widely known that chemicals that preferentially cause O-alkylation in DNA induce mutations. ENU (N-ethyl-N-nitrosourea) is a prototypical O-alkylating agent and the most studied male germ cell mutagen.

Alkylating agents are prototypical somatic and male germ cell mutagens.

Evidence in somatic cells It is well established that transfection of cells with alkylated DNA leads to mutation at the sites of alkyl damage. The design of these experiments requires waiting for cellular replication in order to produce the mutation, confirming the temporal concordance of DNA alkylation and subsequent mutation. A summary of the empirical data to support this is reviewed in Shrivastav et al. (2010).

Various studies have examined the dose-response of DNA adduct levels and mutations. These studies demonstrate that alkyl adducts can be seen at lower doses in the absence of increased mutations both in vitro and in vivo, or demonstrate equal or increased incidence of adducts relative to mutations at the same doses. For example, following exposure of AHH-1 cells to increasing concentrations of MMS, a linear increase in alkylated DNA is measured with significant increases occurring in adduct levels at 0.25 µg/ml (Swenberg et al. 2008). Significant increases in mutations in the HPRT gene in the same cells are not measurable until 1.25 µg MMS/ml (Swenberg et al. 2008) (Figure 1). In vivo, time-series analysis of λlacZ transgenic mice exposed to a single dose of either ENU or DEN, demonstrate that global and lacZ-specific O6-EtGua adducts occur within hours of exposure in the liver, with the bulk of adducts removed by three days post-exposure (Mientjes et al. 1996 and 1998). In contrast, mutant frequency does not begin to significantly increase until three days post-exposure, demonstrating temporal concordance of adduct and mutation formation (see Figure 1 in Mientjes et al. 1998). Levels of O6-EtGua adducts are also consistently higher than the induced mutation frequency per nucleotide. In the bone marrow, DEN is not metabolized and thus is unable to create O6-EtGua adducts. The finding of lack of O6-alkyl adducts in bone marrow is consistent with the lack of an increase in mutations observed in this tissue (Mientjes et al. 1998). This is in contrast to ENU exposure (a direct acting mutagen that does not require metabolic activation), where both adducts and mutations increase in a concordant fashion in the bone marrow.

This pattern of adduct incidence versus mutation incidence is consistent for somatic tissues in rodents in vivo for other types of adducts (we were not able to find suitable dose-response studies to compare oxygen-alkyl adducts to mutation frequencies in vivo). For example, MutaMouse males were exposed to increasing doses of the polycyclic aromatic hydrocarbon dibenzo[a,l]pyrene (forms mutagenic bulky DNA adducts) for 28 days (followed by a 3 day break for mutation fixation) following OECD protocol TG488 (Malik et al. 2013). Significant increases in hepatic DNA adducts were found at 25 mg/kg, but increases in lacZ mutant frequency in liver did not occur until 50 mkg/kg. Bulky DNA adducts in both the livers and lungs of MutaMouse males exhibit an order of magnitude greater incidence per nucleotide than mutations in the lacZ gene using a similar experimental design following exposure to benzo[a]pyrene (Labib et al. 2012; Malik et al. 2012). Lower tissue adduct burden is correlated with lower tissue-specific gene mutation frequencies in Big Blue mice exposed to benzo[a]pyrene (Skopek et al. 1996). Thus, adducts in DNA occurs at lower doses than mutations and are correlated with mutation burden in somatic tissues for different types of DNA adducts.

Evidence in germ cells: No study has compared dose-response for adduct formation and mutation in a single experiment on germ cells. However, it is possible to look across experiments. It is important to note that adducts are measured immediately following exposure because they are relatively quickly repaired. However, analysis of lacZ mutation requires collection of mature sperm from the caudal epididymis. Thus, sperm is collected 42 or 49 days post-exposure (OECD TG488). This is because spermatogonia can not be sampled directly for these purposes. Therefore, comparison of adducts to mutations in pre-meiotic male germ cells requires sampling at different time points for these endpoints (early for adducts, much later for mutations), which is consistent with the expected temporal order of events, with adducts occurring before mutations.

Dose-response for alkyl adduct levels has been very well characterized in mouse testicular DNA for ENU, EMS and DES (van Zeeland et al. 1990). A summary of dose-response data for mouse exposure to ENU and mutation analysis using the transgenic rodent mutation assay in sperm is given in the attached Table I and Figure 2. These studies involve acute injections or oral gavage studies only. Alkyl adducts are evident in gonadal tissues within 2 hours of exposure (van Zeeland et al. 1990) and are fairly efficiently removed within days of exposure in the absence of continued exposure. For this analysis, transgene mutant frequencies were converted to per nucleotide mutation frequency by dividing mutant frequency by the length of the lacZ gene (3096 bp). The data demonstrate that alkyl adduct incidence at low doses is much greater per nucleotide than transgene mutations in the lacZ locus. Adducts are observed to increase substantially at the lowest exposure dose (10 mg/kg), whereas mutation increases in lacZ are marginal at 25 mg/kg. Alkyl adducts in mouse testes following ENU exposure were in the range of approximately 40 in 10E7 nucleotides for 80 mg/kg ENU exposure (van Zeeland et al. 1990). Conversion of the data in O’Brien et al. (2015) to mutations per nucleotide (by dividing mutant frequency by the length of the lacZ locus, which is 3096 bp) produces an estimated induced mutation frequency in spermatogonia of approximately 4 mutations per 10E7 nucleotides for the highest dose (100 mg/kg ENU), an increase of approximately 2 in 10E7 above controls. This suggests that the incidence of adducts is an order of magnitude greater than incidence of mutations. Similarly, exposure to a single dose of 250 mg/kg EMS leads to an over 10-fold increase in the number of alkyl adducts (although the majority are on nitrogen atoms, with only a small proportion on oxygen) (van Zeeland et al. 1990), but only a marginal 2-fold increase in lacZ mutation frequency [van Delft et al. 1997]. Indeed, Van Zeeland et al. (1990) estimate that approximately 10 O6-ethylguanine adducts are required in the gene-coding region to generate a mutation.

Analysis of germ cell mutation during a sub-chronic exposure was carried out by O’Brien et al. (2015). The lower doses used in that study revealed that significant increases in mutations occurred only after 28 days of exposure to the highest dose of 5 mg/kg ENU (cumulative dose of 140 mg/kg), again supporting that higher adduct burdens are required to lead to mutations.

In general, many studies in different mouse strains have used similar experimental designs to conclusively demonstrate that exposure to a variety of alkylating agents causes mutations in spermatogonia (Brooks and Dean 1997; Douglas et al. 1995; Katoh et al. 1997; Katoh et al. 1994; Liegibel and Schmezer 1997; Mattison et al. 1997; O'Brien et al. 2014; O’Brien et al. 2015; Renault et al. 1997; Suzuki et al. 1997; Swayne et al. 2012; Tinwell et al. 1997; van Delft et al. 1997). These studies have been done using a single dose and thus do not enable further comparison of the concordance of dose-response. We also note that ENU exposure of pre-meiotic male germ cells in fish (transgenic medaka) also causes significant increases in mutations observed in spermatozoa (Norris and Winn 2010), supporting the effects of alkylating agents on mutations in male pre-meiotic germs across taxa using similar experimental designs.

As described above, not all alkyl adducts are mutagenic. The proportion of oxygen-alkylation and the type of mutation (with ethylation > methylation) will govern mutagenicity, but there are few empirical data to support this. There are no inconsistencies or uncertainties for ENU or iPMS; other alkylating agents (EMS, MMS) have yielded some discrepancies in the transgenic rodent mutation assay. However, the experimental protocols applied were sub-standard (the OECD TG for this analysis was revised and published in 2013). Overall, more work is needed on alkylating agents other than ENU to fill important data gaps.