Alkylation, DNA leads to Inadequate DNA repair

General details: The weight of evidence for this KER is strong. It is widely accepted that damaged DNA is subject to repair, and that in the absence of DNA repair, mutations will ensue. Specifically, AGT (Damage Reversal DNA repair: pathway #1 in KE155), also known as O6-methylguanine-DNA methyltransferase (MGMT), reverses alkylation damage by directly transferring alkyl groups from the O6 position of guanine to a cysteine residue on the AGT (or MGMT) molecule, restoring the DNA in a single step. However, transfer of the alky group to AGT results in concomitant inactivation of AGT (Pegg 2011). The mammalian protein is also active on O6-ethylguanine and can remove only one ethyl group from DNA, following which the protein is degraded. Thus, high levels of alkylation damage overwhelm the cellular AGT capacity to remove lesions. In mammalian cells, O4-ethylthymine and O2-ethylthymine are poor substrates for AGT (Fang et al. 2010) and no other DNA repair pathway has been identified that is able to efficiently repair these lesions; consequently, these lesions are extremely persistent in cells. Reviews on this topic have been published (Kaina et al. 2007; Pegg 2011). In the absence of the AGT/MGMT pathway, other DNA repair pathways may be invoked, but the relative efficiency of these pathways is not well understood (further details described below).

The role of nucleotide excision repair (NER; excision repair pathways: #2 in KE155) in alkylation damage repair in mammalian cells remains unclear. Earlier studies using human cell lines suggested that both AGT and NER may be involved in the repair of O6-ethylguanine (Bronstein et al. 1991; Bronstein et al. 1992). Very recently, an alkyltransferase like protein (ATL1) that has homology to AGT has been identified in a range of prokaryotes and lower eukaryotes. This protein has no alkyltransferase activity but can couple O6-alkylguanine damage to NER (Latypov et al. 2012). ATL1 proteins have not yet been identified in mammals.

Some alkyl adducts, such as N7-ethylguanine and N3-ethyladenine, are inherently unstable and may depurinate (i.e., hydrolytic cleavage of the glycosidic bond, which releases adenine or guanine). The resultant abasic sites are normally repaired through error-free pathways although they may occasionally be transformed to DNA strand breaks. In mammals, N-methylpurine DNA glycosylases, such as alkyladenine DNA glycosylase (AAG), have a wide range of substrates including N7-alkylguanine and N3-alkyladenine derivatives (Wyatt et al. 1999). However, there are no specific reports in the literature that the ethylated derivatives are AAG substrates. Glycosylases such as AAG yield abasic sites that are processed as described above. An alternative repair mechanism for repairing minor lesions such as N3-ethylcytosine and N1-ethyladenine is through oxidative dealkylation catalyzed by AlkB and mammalian homologs (Drabløs et al. 2004). This pathway is an error-free damage reversal pathway that releases the oxidized ethyl group as acetaldehyde (Duncan et al. 2002).

A final mechanism through which DNA repair pathways may influence the fate of alkylation damage is through futile cycling of the mismatch repair (MMR; excision repair pathways: #2 in KE155) system at an O6-alkyl G:T mispair. In this scenario, unrepaired O6-alkylguanine is able to mispair with T, and the mispair is recognized by MMR enzymes resulting in the removal of the newly incorporated thymine from the nascent strand opposite the O6-alkyguanine adduct. During DNA repair synthesis, O6-alkylguanine preferentially pairs once again with thymine, reinitiating the repair/synthesis cycle. This iteration of excision and synthesis may produce strand breaks and activate damage signaling pathways (York and Modrich 2006).

If the pathways described above become saturated or do not operate properly, the alkylated DNA will not be repaired and will provide a template for replication of this damaged DNA. This is widely understood and accepted. Many studies have demonstrated that the introduction of plasmids or vectors with alkylated DNA (i.e., unrepaired lesions) into prokaryotic and eukaryotic cells, followed by replication, results in the formation of mutations at the alkylated sites, and that the probability of a mutation occurring at the alkylated site is modified by specific DNA repair genes/pathways (reviewed in Basu and Essigmann 1990; Shrivastav et al. 2010).

Insufficient repair is inferred from the formation and retention of adducts, and the formation of increased numbers of mutations above background (i.e., KE185 - methodologies described therein).

A variety of studies show that alkylated DNA persists for prolonged periods of time post-exposure. For example, persistence of different alkylated nucleotides was shown in livers and brains of C57BL mice exposed to N-methyl-N-nitrosourea, N-ethyl-N-nitrosourea and ethyl methanesulfonate using high-performance liquid chromatography several days post-exposure (Frei et al., 1978). The stability of methyl and ethyl adducts in somatic tissues for various adduct types is summarized in Beranuk, 1990. The in vivo liver half life of methyl adducts ranges from 0-3 days, and liver ethyl adduct half lives can be up to 17 days, indicating poorer repair of oxygen-bound ethyl adducts. This prolonged retention of adducts indicates that there is insufficient repair by AGT or other DNA repair pathways of these adducts.

Studies in both hamsters and rats show persistence of alkylated nucleotides several days post-exposure, indicating lack of DNA repair of some adducts (Scherer et al. 1987; Seiler et al. 1997). For example, 101xC3H mouse hybrid testes exhibited DNA adducts within 1 hour of exposure to ENU (10 or 100 mg/kg by i.p.), but some adducts remained unrepaired six days post-exposure (Sega et al. 1986). O6-ethylguanine adducts were also found in hamster spermatogonia DNA up to four days after exposure to DEN (100 µg/g body weight) (Seiler et al. 1997). O6-ethylguanine adducts were found in spermatogonia 1.5 hours post-exposure to ENU in Syrian Golden hamsters (Seiler et al. 1997). Approximately 30% persisted in spermatogonia four days post-exposure. Moreover, the amount of O6-ethylguanine recovered after a 100 mg ENU/kg exposure was 40% greater than predicted from a linear extrapolation of the amount of O6-ethylguanine recovered after exposure to 10 mg/kg. The data suggest that the high dose exposure to ENU results in depletion of AGT within the testis and permits O6-ethylguanine to persist at higher levels than would be predicted from lower exposure. The relationship between dose and formation of DNA adducts in tubular germ cells is non-linear, indicating relatively rapid repair at low doses that becomes saturated at higher doses (van Zeeland et al. 1990). Thus, with increasing dose, increasing incidence of KE1 (insufficient repair) occurs. This implies that mouse spermatogonia are capable of repairing a major part of the DNA damage at low doses. However, at higher doses the repair process is saturated and mutations begin to occur. Indeed, the dose-response curve for mutations in spermatogonia measured in sperm of exposed males is sub-linear with a clear point of inflection at low sub-chronic doses of ENU (O’Brien et al. 2015).

Finally, both alkyl adducts and mutations increase with increasing doses of alkylating agents in somatic cells and in male germ cells, indicating that DNA repair processes are not operating to remove all of the damage (ability to remove adducts and prevent mutations).

DNA repair is not generally measured directly; thus, insufficient repair is inferred from the retention of adducts or the induction of increases in mutation frequencies post-exposure. In addition, various sizes of alkylation groups (e.g., methyl, ethyl, propyl) can be involved. Although it appears that the larger alkyl adducts tend to be more mutagenic (Beranek, 1990), this is not completely established and there are insufficient data to establish that this is true for germ cells. However, in general, this KER is biologically plausible, broadly accepted for alkyl adducts and has few uncertainties. The direct measurement of insufficient repair can be considered a data gap.