Sci. Aging Knowl. Environ., 2 March 2005
Vol. 2005, Issue 9, p. re3
[DOI: 10.1126/sageke.2005.9.re3]


Oxidative Mutagenesis, Mismatch Repair, and Aging

Amy M. Skinner, and Mitchell S. Turker

Amy M. Skinner and Mitchell S. Turker are at the Center for Research on Occupational and Environmental Toxicology, Oregon Health & Science University, Portland, OR 97239, USA. Amy M. Skinner is also in the Department of Environmental and Molecular Toxicology at Oregon State University, Corvallis, OR 97331, USA. Mitchell S. Turker is also in the Department of Molecular and Medical Genetics, Oregon Health & Science University, Portland, OR 97239, USA. E-mail: turkerm{at} (M.S.T.)

Key Words: oxidative stress • oxidative mutagenesis • reactive oxygen species • mismatch repair • cancer

Abstract: A PubMed search for the term "oxidative stress" yields over 29,000 articles published on the subject over the past 10 years; more than 2000 of these articles also include the term "aging" in their title or abstract. Many theories of aging predict causal roles for oxidative stress in the myriad of pathological changes that occur as a function of age, including an increasing propensity to develop cancer. A possible link between aging and cancer is the induction and accumulation of somatic mutations caused by oxidative stress. This Review focuses on small mutational events that are induced by oxidative stress and the role of mismatch repair (MMR) in preventing their formation. It also discusses a possible inhibitory effect of oxidative stress on MMR. We speculate that a synergistic interaction between oxidative damage to DNA and reduced MMR levels will, in part, account for an accumulation of small mutational events, and hence cancer, with aging.

Introduction Back to Top

Oxidative stress is believed to increase with aging and to play roles in age-related pathological processes such as neurodegeneration, atherosclerosis, and cancer (1, 2). We have speculated that an accumulation of mutations from oxidative damage plays a role in the age-related increase in cancer incidence (3). There are a number of ubiquitous agents found both environmentally and endogenously that contribute to cellular levels of reactive oxygen species (ROS). Examples of environmental oxidizers include ionizing and solar radiation (4), automobile exhaust, cigarette smoke (5), other forms of air pollution, and food contamination (6). Endogenous oxidizing sources include byproducts of inefficient oxidative phosphorylation reactions (see Nicholls Perspective and Kristal Perspective) and bacteriocidal and tumoricidal respiratory bursts produced by leukocytes (7). Not surprisingly, the cell has a variety of defense mechanisms (for example, see Sampayo Perspective) to keep ROS from wreaking intracellular havoc; however, the balance between ROS and antioxidant defense can be dysregulated if the amount of ROS exceeds available antioxidant defenses. Once this occurs, the initial damage may perpetuate further damage, leading to a continuous cycle of oxidative stress (4).

ROS cause a myriad of DNA lesions [for reviews, see (4, 8-10)]. Therefore, it is difficult to simply treat cells with oxidants, identify mutations, and then attribute these mutations to specific forms of oxidative damage. Instead, many studies have focused on specific forms of oxidative DNA damage and the types of mutations that they produce. Fig. 1 lists a subset of oxidative base modifications and the types of base-pair mutations that they are known to cause. As discussed below, oxidative stress also appears to produce frameshift mutations (11).

View larger version (9K):
[in this window]
[in a new window]
Fig. 1. Specific oxidative DNA lesions and the mutations they produce.

In this Review, we highlight studies from a number of laboratories that confirm and extend the types of mutations caused by oxidative damage. In many cases, investigators have turned to cells with compromised repair or salvage pathways to demonstrate mutagenic effects. We review some of these studies and then focus on mismatch repair (MMR; see Shcherbakova Review) because of data showing that oxidative stress can impair its function. Oxidative stress increases with age (12), suggesting that some age-related mutations could be caused by nonheritable decreases in MMR. We therefore conclude by highlighting several aging-related studies to look for evidence that oxidative stress plays a role in an age-associated accumulation of small mutational events. Even the relatively small numbers of studies we cite will illustrate that a diversity of systems and approaches can lead to a diversity of results. Nonetheless, some common themes emerge. This Review is intended to be illustrative rather than comprehensive, and we therefore apologize to the many authors whose work is not cited here.

Large mutations such as multilocus deletions (13, 14) and recombination between regions of homologous chromosomes (15) are also induced by oxidative stress, but because of the focus of this Review on an interaction between oxidative mutagenesis and MMR, these large events are not discussed. Likewise, we do not address the issue of mitochondrial mutation and oxidative stress, although it is noted that age-related deletion mutations in the mitochondrial genome may account for a significant fraction of intracellular oxidative stress (16).

Base-Pair Substitutions Induced by Oxidative Damage in MMR-Proficient Cells Back to Top

A number of approaches have been used to define the types of mutations induced by oxidative stress. The universality of repair and scavenger systems in prokaryotic and eukaryotic systems allows results from these systems to be readily compared. On the other hand, the presence of these often-redundant systems within each cell makes it difficult to detect mutations induced by oxidative stress. An additional caveat is that there is no single accepted method of producing oxidative damage, making comparisons between studies somewhat tenuous. Nonetheless, many studies have taken approaches allowing relatively unambiguous results showing the induction of base-pair changes by oxidative mutagenesis; a subset is highlighted below.

Mutations induced by cellular exposure to an oxidative stress

The most straightforward conceptual approach to show that oxidative stress can induce mutations is to treat DNA repair-proficient cells with an oxidizing agent and then demonstrate an increase in the mutant frequency and/or alteration of the mutational spectrum. For example, primary human T lymphocytes exposed to hydrogen peroxide for 1 hour exhibited a dose-dependent increase in the frequency of mutations in Hprt (a useful marker for this type of experiment, because loss-of-function mutations at this locus confer resistance to a toxic compound) both 5 and 8 days after treatment (17). The mutant frequency in cells treated with 2.7 mM H2O2 increased from threefold at day 5 to fivefold at day 8, indicating that mutations in these cells continue to arise after treatment. In both the treated and nontreated cells, the most commonly observed base-pair substitution was G:C->A:T. However, this base-pair substitution was found 2.5 times more frequently in cells exposed to hydrogen peroxide than in unexposed cells, suggesting the induction of this base-pair substitution by ROS. This result led to the speculation that the spectrum of mutations found under routine cell growth conditions is due to endogenous oxidative stress.

The conditions in which cultured mammalian cells are routinely grown do not represent the conditions that are found in the body. One major difference is oxygen tension. A recent comparison of mutations within a transgenic lacZ gene in mouse embryo fibroblasts (MEFs) grown in 20% oxygen or 3% oxygen revealed far lower levels of mutations for the cells grown in 3% oxygen (see Hornsby Perspective). This was particularly the case for G:C->T:A base-pair substitutions and, to a lesser extent, for G:C->A:T substitutions (18). These results therefore suggest that the oxidative stress that occurs when MEFs are grown in 20% oxygen is mutagenic. Comparison of this study with that presented above also shows that different cell types, approaches, and mutational targets can lead to both similar (G:C->A:T substitutions) and different (G:C->T:A substitutions) observations for the types of mutations induced by oxidative stress.

Mutations induced by treatment of DNA with oxidants

Isolated DNA can be directly treated with an oxidizing agent and then introduced into cells to determine the mutagenic effect of treatment. This approach provided a novel mutational signature for oxidative stress in a series of studies conducted in the laboratory of Lawrence Loeb. In an initial experiment (19), M13mp2 single-stranded DNA (a bacteriophage vector that also contained a lacZ insert) was incubated with copper ions to cause oxidative stress and transformed into SOS-induced Escherichia coli. (The SOS system of E. coli is activated by forms of DNA damage that block DNA replication; this system allows damaged DNA to be replicated by low-fidelity DNA polymerases.) The use of single-stranded DNA allowed for unambiguous identification of the mutation that occurred. Ninety percent of the mutations identified were single base-pair substitutions, with C->T and G->T events representing 43 and 33% of the induced base-pair substitutions, respectively. Surprisingly, a tandem base-pair substitution associated with ultraviolet (UV) mutagenesis, CC->TT (20), was found in 14% of the sequenced samples.

The above experiment was interpreted as evidence for the induction of tandem CC->TT mutations by oxidative stress. To test this possibility, a single-stranded bacteriophage containing lacZ{alpha}, a portion of the lacZ gene with an inactivating base-pair substitution, was designed to detect single C->T or tandem CC->TT base-pair substitutions that would restore gene activity (21). This reversion construct (that is, designed to detect base-pair substitutions that revert the nonfunctional construct to one that is functional) was treated with iron or copper in combination with hydrogen peroxide and transformed into SOS-induced E. coli. UV exposure was used as a control. Whereas most induced reversion mutations were C->T base-pair substitutions, tandem CC->TT mutations represented nearly 3% of the total mutations induced by oxidative stress and 7% of the mutations induced by UV. The tandem mutations were not observed in untreated samples. Moreover, the hydroxyl scavenger mannitol was found to reduce the frequency of ROS-induced tandem mutations but not those induced by UV. These results provide further evidence that oxidative stress can induce tandem CC->TT mutations and suggest further that a separate mutagenic pathway is followed after UV exposure to form these tandem mutations.

In a third study by the Loeb group (22), the induction of the tandem CC->TT mutations by oxidative damage was again demonstrated in E. coli, this time by treating a double-stranded plasmid with hydrogen peroxide, nickel, and the tripeptide glycine-glycine-histidine or hydrogen peroxide plus copper (the metal-catalyzed Fenton reaction). This study used a novel assay based on the polymerase chain reaction to detect the tandem mutations at a CC target site in the rat pol{beta} gene located on the plasmid. However, this assay could not detect the induction of CC->TT mutations at the endogenous pol{beta} locus in mammalian cells exposed to hydrogen peroxide. As discussed below, we have now been able to make such a demonstration in mammalian cells deficient in MMR.

Mutations induced by oxidatively modified nucleotides

Another approach to demonstrate a role for oxidative stress in mutation induction is to transfect oxidatively modified nucleotides directly into cells. Inoue et al. (23) transfected 8-hydroxydeoxyguanosine triphosphate (8-OH-dGTP) or 2-hydroxydeoxyadenosine triphosphate (2-OH-dATP) into E. coli and selected for lacI mutants. The mutation frequencies in cells transfected with 8-OH-dGTP or 2-OH-dATP were increased 2.4-fold and 2.5-fold, respectively, above the mutation frequency in nontransfected cells. No increase in mutant frequency was observed when unmodified nucleotides (dGTP or dATP) were tested, ruling out the possibility that the induced mutations were caused by a nucleotide pool imbalance. A:T->C:G transversions were found in 69% of mutant clones induced by 8-OH-dGTP. These mutations are the result of the favorable pairing between adenine and 8-OH-dG. When 2-OH-dATP was transfected into cells, 43% of the identified mutations were G:C->T:A transversion mutations. These mutations are the result of the favorable pairing between guanine and 2-OH-dA. In total, this approach reveals that oxidatively damaged nucleotides in the nucleotide DNA precursor pool can play mutagenic roles.

Mutations in cells deficient in free radical scavenging

Free radical scavenger proteins neutralize free radicals before they can cause damage to DNA or other macromolecules. If free radical scavengers play a role in reducing the mutagenic effects of oxidative stress, the elimination of these scavengers should increase mutagenesis in deficient cells, particularly when these cells are exposed to oxidizing agents. In one study (24), the cytotoxic and mutagenic responses in both untreated and hydrogen peroxide-treated E. coli strains defective in (i) catalase activity (katG katE), which can detoxify hydrogen peroxide, or (ii) superoxide dismutase (SOD) activity (sodA sodB), which can scavenge superoxide radicals, were examined. Under normal growth conditions, the sodA sodB mutant frequency was nearly fivefold higher than the wild-type mutant frequency, and the katG katE mutant frequency was nearly threefold higher than that found in wild-type cells.

Despite the increased mutation frequency, the spontaneous mutational spectrum for the katG katE deficient cells showed no shift when compared with that of wild-type E. coli. However, when these cells were exposed to hydrogen peroxide, over 43% of all induced mutations in the lacI gene were G:C->T:A transversion mutations as compared with 17% in the untreated cells (25). Likewise, there was no substantial increase in the percentage of spontaneous G:C->T:A transversion mutations, as compared with wild-type cells, in the sodA sodB defective cells when mutations in a second mutational target, the supF gene, were monitored. However, when the concentration of iron, which participates in the Fenton reaction, was increased in these cells via the inclusion of a mutation in the gene encoding a repressor of iron uptake (fur), almost half of all mutations (49%) were G:C->T:A transversion mutations (26). Both sets of experiments implicate 8-OH-dGTP in the induction of mutations and confirm a role for the free radical scavengers in preventing oxidative mutagenesis. These enzymes become particularly important when intracellular concentrations of radicals are increased above those normally present within the cells.

Mutations in cells deficient in repair of oxidative lesions

A number of DNA repair pathways have been implicated in the prevention of oxidative mutagenesis via direct repair of oxidative damage. There are several base excision repair enzymes that are involved in helping the cell cope with 8-OH-dGTP, which readily binds to both cytosine and adenine. When bound to cytosine, the FPG (formamidopyrimidine-DNA glycosylase) or OGG1 (8-oxoG DNA N-glycosylase/AP lyase) enzymes (in prokaryotes and eukaryotes, respectively) function to remove the 8-OH-dGTP adduct. If this adduct is bound to adenine, the MutY or MYH1 enzymes (in prokaryotes and eukaryotes, respectively) function to remove the adenine base.

In one set of experiments, untreated and gamma-irradiated M13 DNA, with lacZ{alpha} inserted as a reporter gene, were transformed into wild-type and fpg deficient E. coli (in which the single-stranded DNA is converted into double-stranded DNA), followed by measurement of mutant frequencies and mutational spectra (27). Gamma radiation causes significant amounts of oxidative damage (10). The spontaneous mutant frequencies were the same for the wild-type and fpg deficient cells, as were the mutational spectra, with G:C->A:T and G:C->T:A found in approximately half and one-quarter of mutant cells, respectively. However, examination of the strand distribution of G:C->T:A mutations revealed that there was a 3.5-fold higher level for the (–) strand (the replication template) versus the (+) strand in the fpg deficient cells when compared with the wild-type cells, as would be expected if 8-OH-dGTP were playing a causal role. This ratio increased to 7.0-fold when the frequency of mutations in irradiated DNA transformed into the fpg deficient cells was compared with the mutation frequency in identically treated DNA transformed into wild-type cells. These data provide evidence that the FPG protein plays a role in preventing G:C->T:A mutations caused by 8-OH-dGTP. ogg1 deficient yeast cells also exhibit elevated levels of G:C->T:A mutations (28).

Following up on the above experiments with the fpg deficient E. coli cells, the same group examined spontaneous and gamma radiation-induced mutations in cells that were deficient for both fpg and mutY (29). In these cells, 66% of spontaneous mutations and 87% of induced mutations were G:C->T:A transversions. These levels represent 10- and 20-fold increases, respectively, when compared with spontaneous and induced levels in wild-type cells. G:C->T:A mutations were the predominant mutation within the K-ras proto-oncogene of lung tumors developing in mice with an equivalent genotype; that is, null for both Myh and Ogg1 (30).

Oxidative Mutagenesis and MMR Back to Top

The set of experiments discussed immediately above demonstrates the importance of base excision repair in averting oxidative mutagenesis. Other studies have suggested that MMR proteins can also interact with oxidative lesions (31-33), but the ways in which they do so have not been resolved. Regardless of how such an interaction might take place, a number of experimental approaches have strongly suggested that MMR plays a role in preventing oxidative mutagenesis, and others have suggested that oxidative stress can induce frameshift mutations in repeated sequences, which are a signature event associated with MMR deficiency. These experiments and a possible molecular explanation are provided below, as are experiments showing a role of MMR in preventing oxidative mutagenesis.

Oxidative stress induces frameshift mutations characteristic of MMR-deficient cells

Jackson et al. (34) cloned several out-of-frame microsatellite sequences into plasmids containing the {beta}-lactamase gene and transformed E. coli for the purpose of determining the effect of oxidative stress on microsatellite stability. Microsatellite sequences contain runs of mono- or dinucleotides; high levels of frameshifts within these sequences are a hallmark of MMR deficiency (35). The spontaneous frameshift frequency for both dinucleotide [(CA)11 and (CA)12] and mononucleotide [(A)10 and (G)10] sequences in a wild-type repair-proficient strain of E. coli was found to be approximately 10–5. As expected, when an MMR-deficient strain (mutH) was used, a large (340-fold) increase in frameshift mutations was observed. When plasmid DNA was treated with 1 or 50 µM hydrogen peroxide in the presence or absence of metals and then transformed into repair-proficient cells, frameshift mutations restoring gene expression were observed at levels as high as 10-fold above those observed when untreated DNA was added to the same cells. This was the first demonstration that oxidative stress could induce mutations characteristic of MMR deficiency. These results could be reproduced when the cells were treated with the hydrogen peroxide and metals instead of pretreatment of the plasmid DNA (36).

Subsequent studies have shown that oxidative stress can also induce frameshift mutations in mammalian cells. In one study, a plasmid containing the gene encoding green fluorescent protein (GFP) was modified to contain a CA microsatellite with 13 repeats. Loss of one repeat unit creates an in-frame coding region allowing expression of GFP. This plasmid was transfected into a corrected version (a version with restored MMR function) of the MMR-deficient HCT116 human colon cancer cell line, and the percentage of cells expressing GFP was measured by flow cytometry after exposure of the cells to hydrogen peroxide (37). Among several procedures tested, repeated exposure to hydrogen peroxide had the strongest effect, increasing the frequency of expressing cells as high as 16-fold when compared with control cells.

In a separate study (38), microsatellite mutation frequencies were examined in a human lymphoblastoid cell line. These cells were transfected with a vector containing the herpes simplex virus thymidine kinase (tk) gene with an in-frame (TTCC/AAGG)9 microsatellite sequence. Because a variety of mutational events could inactivate tk (for example, base-pair substitutions, gross rearrangements, and frameshifts), the relative percentage of these events could be compared. Frameshift mutations were most common in this assay in both untreated (62%) and hydrogen peroxide-treated cells (60%). Nonetheless, a fivefold increase in the mutant frequency after exposure was consistent with the induction of frameshift mutations by oxidative stress.

Oxidative stress depletes MMR

The above experiments show that oxidative stress can cause mutations characteristic of MMR deficiency, albeit at lower frequencies. One possible explanation is that oxidative stress directly reduces MMR functionality. To address this possibility, Chang et al. (39) examined MMR activity in cell extracts from human erythroleukemia (HEL) cells exposed to hydrogen peroxide for 24 hours. The extracts were used to measure the detection and repair of heteroduplex DNA containing either a G/T mismatch or a two-base-pair loop [insertion/deletion loop (IDL)]. The DNA preparations were then introduced into a mutS strain of E. coli (to prevent further repair), and the repair that had occurred in the HEL cell extracts was visualized by the presence of galactosidase activity, which makes the bacterial colonies blue. Cellular exposure to hydrogen peroxide was found to reduce both mismatch and IDL repair. To ensure that the reduction in repair was caused by loss of MMR activity, the extracts from the hydrogen peroxide-treated cells were complemented with recombinant human MutS{alpha} (hMutS{alpha}) or hMutL{alpha} protein complexes, or a mixture of both protein complexes. Recombinant hMutS{beta} protein was also used for restoration of IDL repair. MutS and MutL dimers combine to create functional MMR complexes (35). In most cases, significantly increased levels of repair were observed when purified MutL or MutS protein complexes were added to the extracts from hydrogen peroxide-treated cells, and for both assays MMR levels were restored to near-normal levels when both a MutS and a MutL complex were added to the cell extracts. Additional work showed that intracellular protein concentrations of hMSH6 (part of the MutS{alpha} complex) and hPMS2 (part of the MutL{alpha} complex) were considerably decreased after hydrogen peroxide exposure, but the concentrations of other MMR proteins were unchanged. This observation shows specificity for the effect of oxidative stress on MMR function.

The MMR pathway might be particularly sensitive to environmental impairment because it is also sensitive to hypoxia (40) and metals (41). The above study was the first to suggest that oxidative stress decreases MMR levels by specifically targeting several of the MMR proteins. MMR-deficient cells have spontaneous mutator phenotypes because they can't repair base-pair mismatches that are a consequence of DNA replication (35). Therefore, environmental or endogenous perturbation of MMR proteins suggests a second pathway for mutations induced by oxidative stress, with the first being direct DNA damage.

Oxidative mutagenesis in MMR-deficient cells

The possibility that oxidative stress is simultaneously decreasing MMR and causing oxidative DNA lesions raises the possibility of a synergistic interaction. If so, it is important to define the types of mutations that are induced by oxidative stress in MMR-deficient cells. To address this, our laboratory has conducted several studies to identify mutations caused by oxidative stress in MMR null mouse cells.

In one study, we measured the frequency of mutations in Aprt (a useful mutational target because it is both selectable and dispensable in mammalian cells) in mouse cells deficient for the MMR protein MLH1. We found that 35% of mutant Aprt alleles from hydrogen peroxide-exposed Mlh1 null cells had two well-separated base-pair substitutions, as compared with 9% of mutant alleles from untreated Mlh1 null cells (42). Only one base-pair substitution is required to eliminate Aprt expression (43, 44), ruling out the possibility that both mutations were required for the selected phenotype. This result suggests that the induction of alleles with multiple mutations can be a signature for oxidative mutagenesis in MMR null cells. Interestingly, multiple mutations within the TP53 gene in human cancers have been reported (45, 46). Cancers have high levels of oxidative stress (47).

The most common spontaneous base-pair substitution within mouse Aprt in Pms2 null cells is A:T->G:C (60% of all base-pair substitutions detected) (42, 48), a base-pair substitution that is relatively rare in DNA repair-proficient cells. To determine whether this observation could be explained by an increased rate of A:T->G:C mutations relative to other base-pair substitutions, we developed a reversion assay to examine A:T->G:C, G:C->A:T, and G:C->T:A substitutions within Aprt in the Pms2 null cells (49). The results demonstrated an approximately 10- to 20-fold increase in the spontaneous rate of A:T->G:C base-pair substitutions relative to the rates for other base-pair substitution events. The A:T->G:C mutation rate fell fourfold when the cells were grown in the presence of an antioxidant mixture, thereby demonstrating that (i) oxidative stress plays an important role in the formation of this mutation, and (ii) PMS2 protein plays a role in blocking its formation. The antioxidant mix did not decrease the rate of G:C->T:A mutations, most likely because the OGG1 and MYH1 proteins are actively suppressing these mutations even in MMR null cells.

The reversion construct we used to detect G:C->A:T mutations had the added advantage of allowing us to detect tandem GG:CC->AA:TT events within the same codon. This feature allowed us to compare the relative frequencies of each mutational event in a single assay (50). Spontaneous reversion mutations were relatively rare (~0.6–6) in a wild-type kidney cell line, and all were single C->T base-pair substitutions. As expected, spontaneous mutations were far more common in the Pms2 null cells (37 x 10–6). Whereas most were single C->T events (88%), the remainder (12%) were tandem CC->TT events (50). These tandem events do not occur spontaneously in DNA repair-proficient cells (20). Tandem CC->TT events were induced by UV in the repair-proficient cells (16%) at frequencies consistent with previous studies (44). Surprisingly, most of the mutations (64%) induced by UV in the Pms2 null cells were tandem events, suggesting a role for Pms2 in suppressing these mutations when caused by UV damage. Having confirmed the ability of this assay to detect the induction of tandem CC->TT mutations, we next asked if we could detect the induction of tandem mutations by oxidative stress as shown in E. coli by Loeb and colleagues (see above). Although we could not do so when the DNA repair-proficient cells were exposed to a mixture of hydrogen peroxide, copper, and iron, almost all (94%) revertants from the Pms2 null cells exposed to the same mixture exhibited tandem CC->TT mutations. This observation confirms that tandem CC->TT mutations can indeed be a signature for oxidative mutagenesis, but that MMR deficiency may be a prerequisite.

Overexpression of hMTH1 in MMR null cells can suppress mutagenesis

As discussed above, contamination of the DNA precursor pool with oxidatively damaged nucleotides can cause mutations (23). MTH1 protein hydrolyzes 8-oxo-dG and 2-oxo-dA, which provides a means of "sanitizing" these oxidized bases from the nucleotide pool. It is a homolog of E. coli MutT. In a study conducted by Russo et al. (51), hMTH1 cDNA was transfected into mouse embryo fibroblasts (MEFs) derived from mice null for Msh2. Two hMTH1-expressing cell clones were isolated, characterized, and shown to have 10- and 50-fold increased levels of MTH1 activity, respectively, relative to nontransfected cells. These expression levels led to 2- and 17-fold decreases, respectively, in the spontaneous Hprt mutation rate. The mutation rate in the highly expressing clone was near that observed for wild-type MEFs, strongly suggesting that MMR plays an important role in preventing mutations caused by oxidized nucleotides (that is, eliminating these damaged nucleotides eliminates the need for MMR).

The most common mutation at the Hprt locus in the nontransfected MMR null cell lines was a frameshift within a run of guanine nucleotides (36%). The level of frameshift mutations dropped 28-fold in the highly expressing clone relative to the untransfected cells, suggesting that 8-oxo-dG plays a role in generating frameshift mutations. As we found in the Pms2 null cells (42), A:T->G:C mutations were the most common base-pair substitution event in the nontransfected Msh2 null cells (38% of all base-pair substitutions). The frequency of these mutations was reduced 44-fold in the highly expressing hMTH1 clone suggesting that 2-oxo-dA is largely responsible for their formation in the MMR null cells, and by extension that MMR is responsible for preventing the mutagenic effect of incorporated 2-oxo-dA.

Do Mutational Spectra Suggest an Interaction Between Aging, MMR, and Oxidative Stress? Back to Top

One of the few areas of aging-related research and theory in which there is near-universal agreement is that oxidative stress increases with aging (12). Based on the above presentations, we speculate that there is a synergistic interaction in which oxidative stress both causes DNA damage and compromises MMR. This interaction would lead to increased rates of mutations with aging that should have characteristics of mutations caused by oxidative damage and/or MMR deficiency. For example, evidence for microsatellite instability has been detected in human peripheral blood cells as a function of age (52, 53).

Almost all work examining the spectrum of mutations that accumulate with aging has been performed in the mouse, and most studies designed to examine small mutational events have examined mutations in prokaryotic transgenes. These transgenes allow a broad spectrum of base-pair substitutions and in some cases frameshift events to be measured. A recent study using the transgenic MutaMouse (which carries multiple copies of lacZ) to examine mutations in different regions of the digestive tract in young (2 months) and old (23 months) mice (54) found that the frequency of multiple base-pair substitutions increased with aging in cells of the distal part of the intestine and the colon (8 of 92 mutant genes in old mice versus 1 of 83 mutant genes in young mice). Moreover, nearly 30% of these age-related base-pair substitutions were A:T->G:C transversions. As mentioned above, we have reported both multiple mutations (42) and A:T->G:C transversions (49) to be potential signatures of oxidative mutagenesis in MMR null cells.

We also reported tandem mutations as a signature of oxidative mutagenesisis in MMR null cells (50). An analysis of several thousand mutations in the transgenic lacI gene of Big Blue mice of different ages revealed an age-related increase in tandem mutations, particularly in the kidney, liver, and adipose tissue (the colon and intestine were not examined) (55). However, most of these tandem mutations were GG->TT events, not CC->TT as we observed in the MMR null cells exposed to oxidative stressors. Our assay would not have detected the GG->TT events.

A recent comparison of base-pair substitutions within lacZ for five tissues (the brain, heart, liver, spleen, and small intestine) of young (3.5 months) and aged (32 months) transgenic mice revealed remarkably similar mutational spectra for all tissues in the young cohort, with the predominant substitution identified as G:C->A:T transitions (56). Such mutations are generally believed to be caused by spontaneous cytosine deamination. This result shows that developmentally related mutations are similar in a wide range of tissues. The G:C->A:T transitions increased significantly in the brains and hearts of aged versus young mice, with almost all of these mutational events occurring at CpG dinucleotides, the target for DNA methylation. Because lacZ is heavily methylated in the transgenic mice, this result suggests strongly that deamination of methyl cytosine is the predominant age-related base-pair substitution in these nonproliferative tissues. In contrast, the age-related increases in G:C->A:T transitions in the liver and small intestine were mostly at non-CpG sites, suggesting another cause such as oxidative mutagenesis. Further evidence for oxidative mutagenesis was seen in the small intestine, the most proliferative tissue examined, where significant increases were observed in G:C->T:A mutations and one-base-pair deletions at repeat sequences. As discussed above, both mutational events can be signatures for oxidative mutagenesis under certain conditions. However, it does not appear that G:C->T:A mutations are preferentially enhanced in MMR null cells (49), suggesting that perhaps another repair pathway or a salvage pathway can also be inhibited during the aging process in response to oxidative stress or other age-related factors. Tandem mutations were not reported for this study.

Conclusions Back to Top

Oxidative stress clearly has mutagenic consequences, but these consequences are often difficult to measure in DNA repair-proficient cells because of the large number of protective systems (such as free radical salvage and DNA repair) that cells have at their disposal. Oxidative stress also appears to inhibit MMR. As suggested by studies presented here, this combination could lead to an age-related accumulation of mutations that are characteristic of oxidative mutagenesis in MMR-deficient cells. By extension, this could explain an age-related increase in cancers in tissues that are chronically exposed to oxidative stress, particularly in tissues that have large numbers of cycling cells. Admittedly, these possibilities are quite speculative, as are essentially all hypotheses that try to tie together mutation, aging, and cancer (3). However, they at least raise some questions that can be answered experimentally with work designed to determine the effect of oxidative stress on MMR and perhaps other DNA repair pathways, and the types of mutations caused by oxidative stress in cells in which these pathways are impaired. Finally, this is just a piece of the puzzle. We did not discuss large nuclear mutations or mitochondrial mutations induced by oxidative stress. The latter mutations may also contribute to oxidative stress, setting up conditions of a feedback cycle between oxidative stress and oxidative mutagenesis (3).

March 2, 2005
  1. Y. H. Wei, H. C. Lee, Oxidative stress, mitochondrial DNA mutation, and impairment of antioxidant enzymes in aging. Exp. Biol. Med. (Maywood) 227, 671-682 (2002).[Abstract/Free Full Text]
  2. J. E. Klaunig, L. M. Kamendulis, The role of oxidative stress in carcinogenesis. Annu Rev Pharmacol. Toxicol. 44, 239-267 (2004).[CrossRef][Medline]
  3. M. S. Turker, Somatic cell mutations: Can they provide a link between aging and cancer? Mech. Ageing Dev. 117, 1-19 (2000).[CrossRef][Medline]
  4. J. Cadet, T. Delatour, T. Douki, D. Gasparutto, J. P. Pouget, J. L. Ravanat, S. Sauvaigo, Hydroxyl radicals and DNA base damage. Mutat. Res. 424, 9-21 (1999).[CrossRef][Medline]
  5. R. W. Gracy, J. M. Talent, Y. Kong, C. C. Conrad, Reactive oxygen species: The unavoidable environmental insult? Mutat. Res. 428, 17-22 (1999).[Medline]
  6. S. Loft, H. E. Poulsen, Cancer risk and oxidative DNA damage in man. J. Mol. Med. 74, 297-312 (1996).[CrossRef][Medline]
  7. C. B. Klein, K. Frenkel, M. Costa, The role of oxidative processes in metal carcinogenesis. Chem. Res. Toxicol. 4, 592-604 (1991).[CrossRef][Medline]
  8. S. Bjelland, E. Seeberg, Mutagenicity, toxicity and repair of DNA base damage induced by oxidation. Mutat. Res. 531, 37-80 (2003).[CrossRef][Medline]
  9. J. Termini, Hydroperoxide-induced DNA damage and mutations. Mutat. Res. 450, 107-124 (2000).[Medline]
  10. J. Cadet, T. Douki, D. Gasparutto, J. L. Ravanat, Oxidative damage to DNA: Formation, measurement and biochemical features. Mutat. Res. 531, 5-23 (2003).[CrossRef][Medline]
  11. A. L. Jackson, L. A. Loeb, On the origin of multiple mutations in human cancers. Semin. Cancer Biol. 8, 421-429 (1998).[CrossRef][Medline]
  12. G. M. Martin, S. N. Austad, T. E. Johnson, Genetic analysis of ageing: Role of oxidative damage and environmental stresses. Nat. Genet. 13, 25-34 (1996).[CrossRef][Medline]
  13. T. K. Hei, S. X. Liu, C. Waldren, Mutagenicity of arsenic in mammalian cells: Role of reactive oxygen species. Proc. Natl. Acad. Sci. U.S.A. 95, 8103-8107 (1998).[Abstract/Free Full Text]
  14. M. Filipic, T. K. Hei, Mutagenicity of cadmium in mammalian cells: Implication of oxidative DNA damage. Mutat. Res. 546, 81-91 (2004).[Medline]
  15. M. S. Turker, B. M. Gage, J. A. Rose, D. Elroy, O. N. Ponomareva, P. J. Stambrook, J. A. Tischfield, A novel signature mutation for oxidative damage resembles a mutational pattern found commonly in human cancers. Cancer Res. 59, 1837-1839 (1999).[Abstract/Free Full Text]
  16. K. K. Singh, Mitochondrial dysfunction is a common phenotype in aging and cancer. Ann. N Y Acad. Sci. 1019, 260-264 (2004).[CrossRef][Medline]
  17. S. Diaz-Llera, A. Podlutsky, A. M. Osterholm, S. M. Hou, B. Lambert, Hydrogen peroxide induced mutations at the HPRT locus in primary human T-lymphocytes. Mutat. Res. 469, 51-61 (2000).[Medline]
  18. R. A. Busuttil, M. Rubio, M. E. Dolle, J. Campisi, J. Vijg, Oxygen accelerates the accumulation of mutations during the senescence and immortalization of murine cells in culture. Aging Cell 2, 287-294 (2003).[CrossRef][Medline]
  19. L. K. Tkeshelashvili, T. McBride, K. Spence, L. A. Loeb, Mutation spectrum of copper-induced DNA damage. J. Biol. Chem. 266, 6401-6406 (1991).[Abstract/Free Full Text]
  20. F. Hutchinson, Induction of tandem-base change mutations. Mutat. Res. 309, 11-15 (1994).[Medline]
  21. T. M. Reid, L. A. Loeb, Tandem double CC->TT mutations are produced by reactive oxygen species. Proc. Natl. Acad. Sci. U.S.A. 90, 3904-3907 (1993).[Abstract/Free Full Text]
  22. T. G. Newcomb, K. J. Allen, L. Tkeshelashvili, L. A. Loeb, Detection of tandem CC->TT mutations induced by oxygen radicals using mutation-specific PCR. Mutat. Res. 427, 21-30 (1999).[Medline]
  23. M. Inoue, H. Kamiya, K. Fujikawa, Y. Ootsuyama, N. Murata-Kamiya, T. Osaki, K. Yasumoto, H. Kasai, Induction of chromosomal gene mutations in Escherichia coli by direct incorporation of oxidatively damaged nucleotides. New evaluation method for mutagenesis by damaged DNA precursors in vivo. J. Biol. Chem. 273, 11069-11074 (1998).[Abstract/Free Full Text]
  24. J. Alhama, J. Ruiz-Laguna, A. Rodriguez-Ariza, F. Toribio, J. Lopez-Barea, C. Pueyo, Formation of 8-oxoguanine in cellular DNA of Escherichia coli strains defective in different antioxidant defences. Mutagenesis 13, 589-594 (1998).[Abstract/Free Full Text]
  25. J. Ruiz-Laguna, C. Pueyo, Hydrogen peroxide and coffee induce G:C->T:A transversions in the lacI gene of catalase-defective Escherichia coli. Mutagenesis 14, 95-102 (1999).[Abstract/Free Full Text]
  26. T. Nunoshiba, F. Obata, A. C. Boss, S. Oikawa, T. Mori, S. Kawanishi, K. Yamamoto, Role of iron and superoxide for generation of hydroxyl radical, oxidative DNA lesions, and mutagenesis in Escherichia coli. J. Biol. Chem. 274, 34832-34837 (1999).[Abstract/Free Full Text]
  27. G. K. Kuipers, H. A. Poldervaart, B. J. Slotman, M. V. Lafleur, The influence of formamidopyrimidine-DNA glycosylase on the spontaneous and gamma-radiation-induced mutation spectrum of the lacZ alpha gene. Mutat. Res. 435, 141-150 (1999).[Medline]
  28. D. Thomas, A. D. Scot, R. Barbey, M. Padula, S. Boiteux, Inactivation of OGG1 increases the incidence of G.C->T.A transversions in Saccharomyces cerevisiae: Evidence for endogenous oxidative damage to DNA in eukaryotic cells. Mol. Gen. Genet. 254, 171-178 (1997).[CrossRef][Medline]
  29. G. K. Kuipers, B. J. Slotman, H. A. Poldervaart, C. A. Reitsma-Wijker, M. V. Lafleur, The influence of combined Fpg- and MutY-deficiency on the spontaneous and gamma-radiation-induced mutation spectrum in the lacZalpha gene of m13mp10. Mutat. Res. 461, 189-195 (2000).[Medline]
  30. Y. Xie, H. Yang, C. Cunanan, K. Okamoto, D. Shibata, J. Pan, D. E. Barnes, T. Lindahl, M. McIlhatton, R. Fishel, J. H. Miller, Deficiencies in mouse Myh and Ogg1 result in tumor predisposition and G to T mutations in codon 12 of the K-ras oncogene in lung tumors. Cancer Res. 64, 3096-3102 (2004).[Abstract/Free Full Text]
  31. A. Mazurek, M. Berardini, R. Fishel, Activation of human MutS homologs by 8-oxo-guanine DNA damage. J. Biol. Chem. 277, 8260-8266 (2002).[Abstract/Free Full Text]
  32. J. Wyrzykowski, M. R. Volkert, The Escherichia coli methyl-directed mismatch repair system repairs base pairs containing oxidative lesions. J. Bacteriol. 185, 1701-1704 (2003).[Abstract/Free Full Text]
  33. C. Colussi, E. Parlanti, P. Degan, G. Aquilina, D. Barnes, P. Macpherson, P. Karran, M. Crescenzi, E. Dogliotti, M. Bignami, The mammalian mismatch repair pathway removes DNA 8-oxodGMP incorporated from the oxidized dNTP pool. Curr. Biol. 12, 912-918 (2002).[CrossRef][Medline]
  34. A. L. Jackson, R. Chen, L. A. Loeb, Induction of microsatellite instability by oxidative DNA damage. Proc. Natl. Acad. Sci. U.S.A. 95, 12468-12473 (1998).[Abstract/Free Full Text]
  35. A. B. Buermeyer, S. M. Deschenes, S. M. Baker, R. M. Liskay, Mammalian DNA mismatch repair. Annu. Rev. Genet. 33, 533-564 (1999).[CrossRef][Medline]
  36. A. L. Jackson, L. A. Loeb, Microsatellite instability induced by hydrogen peroxide in Escherichia coli. Mutat. Res. 447, 187-198 (2000).[Medline]
  37. C. Gasche, C. L. Chang, J. Rhees, A. Goel, C. R. Boland, Oxidative stress increases frameshift mutations in human colorectal cancer cells. Cancer Res. 61, 7444-7448 (2001).[Abstract/Free Full Text]
  38. M. L. Maneval, K. A. Eckert, Effects of oxidative and alkylating damage on microsatellite instability in nontumorigenic human cells. Mutat. Res. 546, 29-38 (2004).[Medline]
  39. C. L. Chang, G. Marra, D. P. Chauhan, H. T. Ha, D. K. Chang, L. Ricciardiello, A. Randolph, J. M. Carethers, C. R. Boland, Oxidative stress inactivates the human DNA mismatch repair system. Am. J. Physiol. Cell Physiol. 283, C148-C154 (2002).[Abstract/Free Full Text]
  40. V. T. Mihaylova, R. S. Bindra, J. Yuan, D. Campisi, L. Narayanan, R. Jensen, F. Giordano, R. S. Johnson, S. Rockwell, P. M. Glazer, Decreased expression of the DNA mismatch repair gene Mlh1 under hypoxic stress in mammalian cells. Mol. Cell Biol. 23, 3265-3273 (2003).[Abstract/Free Full Text]
  41. Y. H. Jin, A. B. Clark, R. J. Slebos, H. Al-Refai, J. A. Taylor, T. A. Kunkel, M. A. Resnick, D. A. Gordenin, Cadmium is a mutagen that acts by inhibiting mismatch repair. Nat. Genet. 34, 326-329 (2003).[CrossRef][Medline]
  42. C. Y. Shin, I. Mellon, M. S. Turker, Multiple mutations are common at mouse Aprt in genotoxin-exposed mismatch repair deficient cells. Oncogene 21, 1768-1776 (2002).[CrossRef][Medline]
  43. C. Y. Shin, O. N. Ponomareva, L. Connolly, M. S. Turker, A mouse kidney cell line with a G:C->C:G transversion mutator phenotype. Mutat Res. 503, 69-76 (2002).[Medline]
  44. N. H. Khattar, M. S. Turker, A role for certain mouse Aprt sequences in resistance to toxic adenine analogs. Somat. Cell Mol. Genet. 23, 51-61 (1997).[Medline]
  45. B. S. Strauss, Silent and multiple mutations in p53 and the question of the hypermutability of tumors. Carcinogenesis 18, 1445-1452 (1997).[Abstract/Free Full Text]
  46. B. S. Strauss, Role in tumorigenesis of silent mutations in the TP53 gene. Mutat. Res. 457, 93-104 (2000).[Medline]
  47. S. Toyokuni, K. Okamoto, J. Yodoi, H. Hiai, Persistent oxidative stress in cancer. FEBS Lett. 358, 1-3 (1995).[CrossRef][Medline]
  48. C. Shao, M. Yin, L. Deng, P. J. Stambrook, T. Doetschman, J. A. Tischfield, Loss of heterozygosity and point mutation at Aprt locus in T cells and fibroblasts of Pms2(–/–) mice. Oncogene 21, 2840-2845 (2002).[CrossRef][Medline]
  49. C. Y. Shin, M. S. Turker, A:T->G:C base pair substitutions occur at a higher rate than other substitution events in Pms2 deficient mouse cells. DNA Repair (Amsterdam) 1, 995-1001 (2002).[CrossRef]
  50. C. Y. Shin-Darlak, A. M. Skinner, M. S. Turker, A role for Pms2 in the prevention of tandem CC->TT substitutions induced by ultraviolet radiation and oxidative stress. DNA Repair (Amsterdam) 4, 51-57 (2005).
  51. M. T. Russo, M. F. Blasi, F. Chiera, P. Fortini, P. Degan, P. Macpherson, M. Furuichi, Y. Nakabeppu, P. Karran, G. Aquilina, M. Bignami, The oxidized deoxynucleoside triphosphate pool is a significant contributor to genetic instability in mismatch repair-deficient cells. Mol. Cell Biol. 24, 465-474 (2004).[Abstract/Free Full Text]
  52. S. Krichevsky, G. Pawelec, A. Gural, R. B. Effros, A. Globerson, D. B. Yehuda, A. B. Yehuda, Age related microsatellite instability in T cells from healthy individuals. Exp. Gerontol. 39, 507-515 (2004).[CrossRef][Medline]
  53. A. Ben Yehuda, A. Globerson, S. Krichevsky, H. Bar On, M. Kidron, Y. Friedlander, G. Friedman, D. Ben Yehuda, Ageing and the mismatch repair system. Mech. Ageing Dev. 121, 173-179 (2000).[CrossRef][Medline]
  54. T. Ono, H. Ikehata, V. P. Pithani, Y. Uehara, Y. Chen, Y. Kinouchi, T. Shimosegawa, Y. Hosoi, Spontaneous mutations in digestive tract of old mice show tissue-specific patterns of genomic instability. Cancer Res. 64, 6919-6923 (2004).[Abstract/Free Full Text]
  55. K. A. Hill, J. Wang, K. D. Farwell, S. S. Sommer, Spontaneous tandem-base mutations (TBM) show dramatic tissue, age, pattern and spectrum specificity. Mutat. Res. 534, 173-186 (2003).[Medline]
  56. M. E. Dolle, W. K. Snyder, D. B. Dunson, J. Vijg, Mutational fingerprints of aging. Nucleic Acids Res. 30, 545-549 (2002).[Abstract/Free Full Text]
  57. M. Moriya, Single-stranded shuttle phagemid for mutagenesis studies in mammalian cells: 8-oxoguanine in DNA induces targeted G.C->T.A transversions in simian kidney cells. Proc. Natl. Acad. Sci. U.S.A. 90, 1122-1126 (1993).[Abstract/Free Full Text]
  58. P. T. Henderson, J. C. Delaney, F. Gu, S. R. Tannenbaum, J. M. Essigmann, Oxidation of 7,8-dihydro-8-oxoguanine affords lesions that are potent sources of replication errors in vivo. Biochemistry 41, 914-921 (2002).[CrossRef][Medline]
  59. C. J. Wiederholt, M. M. Greenberg, Fapy.dG instructs Klenow exo(-) to misincorporate deoxyadenosine. J. Am. Chem. Soc. 124, 7278-7279 (2002).[CrossRef][Medline]
  60. H. Kamiya, T. Ueda, T. Ohgi, A. Matsukage, H. Kasai, Misincorporation of dAMP opposite 2-hydroxyadenine, an oxidative form of adenine. Nucleic Acids Res. 23, 761-766 (1995).[Abstract/Free Full Text]
  61. A. Guy, A. M. Duplaa, J. Ulrich, R. Teoule, Incorporation by chemical synthesis and characterization of deoxyribosylformylamine into DNA. Nucleic Acids Res. 19, 5815-5820 (1991).[Abstract/Free Full Text]
  62. J. M. McNulty, B. Jerkovic, P. H. Bolton, A. K. Basu, Replication inhibition and miscoding properties of DNA templates containing a site-specific cis-thymine glycol or urea residue. Chem. Res. Toxicol. 11, 666-673 (1998).[CrossRef][Medline]
  63. A. A. Purmal, G. W. Lampman, J. P. Bond, Z. Hatahet, S. S. Wallace, Enzymatic processing of uracil glycol, a major oxidative product of DNA cytosine. J. Biol. Chem. 273, 10026-10035 (1998).[Abstract/Free Full Text]
  64. D. A. Kreutzer, J. M. Essigmann, Oxidized, deaminated cytosines are a source of C->T transitions in vivo. Proc. Natl. Acad. Sci. U.S.A. 95, 3578-3582 (1998). [Abstract/Free Full Text]
  65. The authors' work on oxidative mutagenesis in MMR null cells was supported by NIH grants CA56383 and CA76528 and by National Institute of Environmental Health Sciences grant 1P42 ES10338. A.S. is supported by training grant T32 ES07060 and is a contributing editor-assistant for SAGE KE. We thank M. Pieretti for helpful comments on this manuscript.
Citation: A. M. Skinner, M. S. Turker, Oxidative Mutagenesis, Mismatch Repair, and Aging. Sci. Aging Knowl. Environ. 2005 (9), re3 (2005).

Science of Aging Knowledge Environment. ISSN 1539-6150