Skip to main content
NCBI home page
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 2008 Dec 22;105(52):20689–20694. doi: 10.1073/pnas.0811241106

Replication protein A and proliferating cell nuclear antigen coordinate DNA polymerase selection in 8-oxo-guanine repair

Giovanni Maga a,1, Emmanuele Crespan a, Ursula Wimmer b, Barbara van Loon b, Alessandra Amoroso a, Chiara Mondello a, Cristina Belgiovine a, Elena Ferrari b, Giada Locatelli c, Giuseppe Villani c, Ulrich Hübscher b
PMCID: PMC2634905  PMID: 19104052

Abstract

The adenine misincorporated by replicative DNA polymerases (pols) opposite 7,8-dihydro-8-oxoguanine (8-oxo-G) is removed by a specific glycosylase, leaving the lesion on the DNA. Subsequent incorporation of C opposite 8-oxo-G on the resulting 1-nt gapped DNA is essential for the removal of the 8-oxo-G to prevent G–C to T–A transversion mutations. By using model DNA templates, purified DNA pols β and λ and knockout cell extracts, we show here that the auxiliary proteins replication protein A and proliferating cell nuclear antigen act as molecular switches to activate the DNA pol λ- dependent highly efficient and faithful repair of A:8-oxo-G mismatches in human cells and to repress DNA pol β activity. By using an immortalized human fibroblast cell line that has the potential to induce cancer in mice, we show that the development of a tumoral phenotype in these cells correlated with a differential expression of DNA pols λ and β.

Reactive oxygen species (ROS) are produced during normal cell metabolism and through the action of exogenous agents (1). When ROS react with DNA, the most frequently generated lesion (103 to 104 per cell per day) is 7,8-dihydro-8-oxoguanine (8-oxo-G) (2), whose mutagenic potential in aging, tumor transformation, and neurodegenerative diseases is well established. The presence of 8-oxo-G in the replicating strand can lead to frequent misincorporation of A opposite the lesion by the human replicative DNA polymerases (pols) α, δ, and ε (3). Full repair of 8-oxo-G lesion is guaranteed by 2 different base excision repair (BER) systems: (i) an OGG1-dependent, which targets C:8-oxo-G mispairs, removes the lesion and leaves an intact DNA strand to act as template for the resynthesis step (4); and (ii) a MUTYH-dependent pathway, which targets the A:8-oxo-G base pair and removes the adenine (5–7). Subsequent error-free bypass of the lesion requires a specialized DNA pol that can catalyze the correct incorporation of C opposite 8-oxo-G during the resynthesis step, reconstituting a C:8-oxo-G base pair that could subsequently be repaired by the OGG1-dependent BER. However, the majority of human DNA pols insert an adenine opposite 8-oxo-G on the template strand with high frequencies (10–75% of the time). Thus, the molecular mechanism ensuring correct and efficient repair of A:8-oxo-G mismatches in human cells is currently undetermined.

We have recently shown (8) that the BER enzyme DNA pol λ, which belongs to DNA pol family X (9), is very efficient in performing error-free translesion synthesis past the 2 major oxidative lesions 8-oxo-G (8) and 2-hydroxy-adenine (2-OH-A) (10). Moreover, its fidelity and efficiency is enhanced 2 orders of magnitude by the auxiliary proteins proliferating cell nuclear antigen (PCNA) and replication protein A (RP-A), both for normal and translesion synthesis, resulting in dATP incorporation frequencies opposite 8-oxo-G as low as 10−3. On the other hand, the other major BER enzyme DNA pol β shows a relaxed nucleotide insertion specificity opposite 8-oxo-G, with erroneous (i.e., dATP) incorporation occurring in 20–30% of the cases.

Because of the apparently overlapping roles in BER of DNA pol β and λ, a discriminatory mechanism is required for the repair machinery to properly select DNA pol λ vs. DNA pol β in the MUTYH-dependent BER pathway. The DNA glycosylase MUTYH has been shown to interact with both PCNA and RP-A, suggesting its involvement in replication-coupled BER of A:8-oxo-G mismatches. DNA pol λ also interacts with PCNA and its activity is modulated by RP-A (11–14). In the present work, we show a role of PCNA and RP-A in selecting the most appropriate DNA pol for 8-oxo-G repair. In addition, we investigated the variation of the relative levels of DNA pol λ and β in a model cell line, cen3tel (15). Cen3tel cells acquired the ability to replicate indefinitely, but also showed neoplastic transformation driven by successive stepwise mutations in tumor suppressor genes and oncogenes such as p16INK4a, p14ARF p53, and c-myc, ultimately becoming able to induce tumors when injected in immuno-compromised mice (16). Our results suggest that the misregulated expression of DNA pols β and λ might play an important role in cancer development.

Results

8-Oxo-G Bypass Efficiency and Selectivity of DNA Pol λ Are Independent from Gap Size.

As shown in Fig. 1 A and B and Table S1, DNA pol λ incorporated dCTP better than dATP opposite an 8-oxo-G lesion on DNA substrates with gaps of increasing size (1, 2, and 8 nt), irrespective of the size of the gap. A 1-nt strand displacement event, corresponding to a +2-nt product, was observed on the 1-nt gap substrate (Fig. S1A, lanes 11–14 and 21), whereas no strand displacement was observed on the 2- or 8-nt gapped substrates (Fig. S1 A, lane 22 and B, lanes 11–16). This 1-nt strand displacement depended on both the nucleotide and the enzyme concentration (Fig. S1C, compare lanes 1 and 5 with lanes 9 and 13) and was likely caused by transient “breathing” of the 5′ end, which allowed insertion of 1 additional nucleotide, as already described for DNA pol β (17). Under these conditions, PCNA showed no enhancement of strand displacement (Fig. S1C).

Fig. 1.

Fig. 1.

Open in a new tab

PCNA and RP-A promoted error-free bypass of 8-oxo-G by DNA pol λ and inhibited DNA pol β incorporation opposite an 8-oxoG on gapped DNA templates. Reactions were performed with recombinant DNA pols under the conditions specified in Materials and Methods. (A) Variation of the initial velocities of the reaction catalyzed by 6 nM DNA pol λ on the 1-nt gapped 8-oxo-G template (6 nM) in the presence of dCTP (○) or dATP (▵), as a function of the nucleotide substrate concentration. Values are the mean of 3 independent experiments. Error bars are ± SD. (B) Variation of the initial velocities of the reaction catalyzed by 6 nM DNA pol λ on the 8-nt gapped 8-oxo-G template (6 nM) in the presence of dCTP (○) or dATP (▵), as a function of the nucleotide substrate concentration. Values are the mean of 3 independent experiments. Error bars are ± SD. (C) Effects of increasing amounts of RP-A on the apparent incorporation rates for dATP (light gray bars) or dCTP (dark gray bars) on a 1-nt gapped 8-oxo-G template by DNA pol λ. Values are the mean of 3 independent replicates. Error bars represent ± SD values. (D) RP-A was incubated in the presence of 200 nM 5′-labeled 8-oxo-G or undamaged 1-nt gapped DNA substrates. Samples were run on a native PAGE. The positions of the free probe and the RP-A/DNA complex are shown on the right. (E) DNA pol β was incubated in the presence of the 1-nt gapped 8-oxo-G template and dATP (lanes 1–4 and 9–12) or dCTP (lanes 5–8 and 13–16), in the absence (lanes 1–8) or the presence (lanes 9–16) of PCNA and RP-A. Lane C, control reaction in the absence of nucleotides. (F) Effects of RP-A on the efficiency of primer binding (kcat/Kd) by DNA pols λ and β on control and 8-oxo-G 1-nt gapped DNA templates. The kcat and Kd values were obtained from 3 independent experiments as described in Materials and Methods. Mean values were used to calculate the kcat/Kd ratios.

RP-A Increases the Error-Free Bypass of 8-Oxo-G by DNA Pol λ on a 1-nt Gapped Substrate.

When RP-A was included in the reaction, it reduced the rate of dATP, but not dCTP, incorporation opposite the lesion in a dose-dependent manner (Fig. 1C and Fig. S1 D–F). The selectivity of DNA pol λ for correct dCTP vs. the incorrect dATP increased >100-fold in the presence of RP-A (Table S1). Band-shift experiments (Fig. 1D) performed with either the 8-oxo-G (Fig. 1D, lanes 1–4) or the undamaged (Fig. 1D, lanes 5–8) 1-nt gapped templates clearly demonstrated that RP-A was able to bind both DNA substrates with similar affinities. Comparison of the bound vs. free signals showed that 500 nM RP-A was enough to fully saturate 200 nM of the DNA substrate (Fig. 1D, lanes 4 and 8). This molar excess reflected the change in the affinity of RP-A from 10−9 M for ssDNA to 10−7 M for dsDNA (18).

DNA Pol λ Is More Efficient than DNA Pol β in Error-Free Gap Filling Opposite an 8-Oxo-G Lesion in the Presence of RP-A and PCNA.

As shown in Fig. 1E (lanes 1–8), DNA pol β showed similar incorporation efficiencies for dATP and dCTP opposite the lesion, with a selectivity index of ≈3 (Table S1), whereas, unlike DNA pol λ (8), it showed a greatly reduced efficiency of dCTP incorporation opposite an 8-oxo-G lesion with respect to the undamaged template (Fig. S1G, compare lanes 1–4 with lanes 5–8). Interestingly, PCNA and RP-A inhibited both dCTP and dATP incorporation by DNA pol β opposite the lesion (Fig. 1E, lanes 9–16 and Fig. S1H, lanes 5–8 and 13–16). As a result, in the presence of PCNA and RP-A DNA pol λ displays a 32-fold higher probability than DNA pol β to fill the gap opposite the lesion (Table S1). PCNA reduced dCTP incorporation opposite the lesion by DNA pol β by 30%, whereas RP-A inhibited the same reaction by 70% (Fig. S2A). PCNA or RP-A had no effect on dCTP incorporation opposite an undamaged G (Fig. S2A). DNA pol λ proved to be more efficient in inserting dCTP opposite 8-oxo-G than DNA pol β also in the presence of all 4 dNTPs (Fig. S3). In addition, RP-A selectively inhibited dCTP incorporation by DNA pol β but not by DNA λ opposite 8-oxo-G on a different sequence context (Fig. S4). Thus, RP-A appeared to be the major determinant for the decreased ability of DNA pol β to incorporate dCTP opposite the 8-oxo-G lesion.

RP-A Modulates the DNA Primer Binding Affinity of DNA Pols λ and β, Depending on the Presence of an 8-Oxo-G Lesion on the Template Strand.

We next measured the DNA primer utilization efficiency (kcat/Kd3′OH) of DNA pol λ and β by determining the amount of elongated primer as a function of the gapped DNA substrate concentration, in the presence of a fixed dCTP concentration. The calculated kcat/Kd values are shown in Fig. 1F. In the absence of RP-A, DNA pol β was able to bind a 3′-OH primer end with the same efficiency in the absence (Fig. 1F, lane 1) or in the presence (Fig. 1F, lane 2) of the lesion, whereas DNA pol λ showed 3-fold preference for the damaged template (Fig. 1F, compare lanes 10 and 9). In the presence of 500 nM RP-A and 10 μM dCTP, DNA pol β interacted with the primer in the presence of an 8-oxo-G lesion 4-fold less efficiently than with the undamaged template (Fig. 1F, lanes 5 and 6). When dCTP was used at 0.5 μM in the presence of 500 nM RP-A, the loss of efficiency by DNA pol β for the damaged vs. undamage template was 18-fold (Fig. 1F, lanes 7 and 8). Conversely, when tested under the same conditions (0.5 μM dCTP, 500 nM RP-A), DNA pol λ showed a 2.7-fold preference for the damaged template (Fig. 1F, compare lanes 12 and 11). Under these conditions, DNA pol λ was 50-fold more efficient than DNA pol β in the presence of an 8-oxo-G lesion (Fig. 1F, compare lanes 8 and 12). These data clearly suggest that under low dCTP concentration (and this might be the case in vivo under oxidative stress and other emergency situations) the correct dCTP incorporation is better guaranteed by DNA pol λ in the presence of RP-A.

Physiological Salt Concentrations, PCNA, and RP-A Favor DNA Pol λ Over DNA Pol β on 1-nt Gapped Intermediates Bearing an 8-Oxo-G Lesion.

Increasing salt concentrations (80–150 mM NaCl) reduced the incorporation of dATP and dCTP opposite 8-oxo-G by both DNA pol λ and β (Fig. 2A, lanes 1–8 and 15–22). However, DNA pol λ was significantly more efficient than DNA pol β in dCTP incorporation opposite 8-oxo-G in the presence of 0.1 M NaCl (Fig. 2B). By comparing the respective kcat/Km values (Table S1), DNA pol λ was 23.5-fold more efficient than DNA pol β in dCTP incorporation opposite 8-oxo-G in the presence of 0.1 M NaCl. Overall, dCTP incorporation by DNA pol β was reduced by RP-A by 75%, whereas under the same conditions DNA pol λ was affected <10% (Fig. 2C). At low dCTP concentrations (0.2–1 μM), the activity of DNA pol β was almost completely suppressed, whereas DNA pol λ was able to catalyze efficient error-free incorporation (Fig. S2B, compare lanes 1 and 2 with lanes 5 and 6). Comparison of the relative kcat/Km values, showed that DNA pol λ was 145-fold more efficient than DNA pol β for dCTP incorporation opposite 8-oxo-G in the presence of both RP-A and salt (Table S1). On the control template (i.e., in the absence of the 8-oxo-G lesion) DNA pol β and λ showed similar efficiencies of dCTP incorporation in the presence of both salt and RP-A, suggesting that the observed effect depended on the presence of the lesion (Fig. S2C). In the presence of RP-A and salt, PCNA was able to further reduce the ability to incorporate dCTP opposite 8-oxo-G of DNA pol β (Fig. 2D, lanes 1–6), but not of DNA pol λ (Fig. 2D, lanes 7–12). Quantification of the results showed that PCNA inhibited dCTP incorporation of DNA pol β by 70%, whereas DNA pol λ was not affected (Fig. 2E). Taken together, these data suggest that PCNA and RP-A may favor the recruitment of DNA pol λ over DNA pol β on a 1-nt gap containing an 8-oxo-G lesion. Indeed, we found that either DNA pol λ (Fig. 2F, lane 1) or DNA pol β (Fig. 2F, lane 3) were able to pull down RP-A (visualized with anti-p70 Abs). No unspecific binding of RP-A to the beads (Fig. 2F, lanes 2), nor effect of DNase1 treatment was noted (Fig. 2F, lanes 4–6), ruling out unspecific DNA-mediated interactions. Next, we incubated the 1-nt gapped 8-oxo-G substrate carrying a 5′ biotinylated primer in the presence of 1 μM dCTP, DNA pols β and λ at a 10:1 molar ratio and in the absence or presence of PCNA and RP-A. Bound DNA pols pulled down by streptavidin-coupled agarose beads were visualized by Western blot analysis with specific antibodies. As shown in Fig. 2G, both DNA pols β and λ were recovered in the pellet in the absence of PCNA and RP-A (Fig. 2G, lane 1). Addition of auxiliary proteins, however, caused a marked reduction of bound DNA pol β (Fig. 2G, lane 2) and its correspondent increase in the supernatant (Fig. 2G, lane 4). Association of RP-A to the biotinylated DNA substrate under these conditions did not affect DNA pol λ binding (Fig. S2D). Overall, these data suggested that PCNA and RP-A strongly restrict association of DNA pol β to a 1-nt gapped substrate during dCTP incorporation opposite 8-oxo-G.

Fig. 2.

Fig. 2.

Open in a new tab

PCNA and RP-A favor DNA pol λ gap filling and recruitment over DNA pol β at the 8-oxo-G lesion under physiological salt conditons. Reactions were performed with recombinant DNA pols under the conditions specified in Materials and Methods. (A) DNA pol β (lanes 1–14) or DNA pol λ (lanes 15–28) were incubated in the presence of the 1-nt gapped 8-oxo-G DNA template, 2 μM dATP (lanes 1–4, 9–11, 15–18, and 23–25) or 2 μM dCTP (lanes 5–8, 12–14, 19–22, and 26–28) and in the absence (lanes 1, 5, 15, and 19) or the presence of the indicated amounts of NaCl, either alone (lanes 1–8 and 15–22) or in combination with RP-A (lanes 9–14 and 23–28). Lane C, control reaction in the absence of nucleotides. The structure of the template used is shown at the top. (B) Variation of the apparent reaction velocities for dCTP incorporation opposite 8-oxo-G by DNA pol λ (▵) or DNA pol β (○), as a function of the nucleotide substrate concentration, in the presence of 0.1 M NaCl. Values are the mean of 3 independent replicates. Error bars represent ± SD values. Template and enzyme concentrations were as in A. (C) Effects of increasing amounts of RP-A on the incorporation of dCTP opposite 8-oxo-G by DNA pol β (○) or DNA pol λ (▵). Values are the mean of 3 independent replicates. Error bars represent ± SD values. Reactions conditions were as in A. (D) DNA pol β (lanes 1–6) or DNA pol λ (lanes 7–12) were incubated in the presence of the 1-nt gapped 8-oxo-G DNA template, 10 μM dCTP, RP-A and in the absence (lanes 1 and 7) or presence (lanes 2–6 and 8–12) of increasing amounts of PCNA. Lane C, control reaction in the absence of nucleotides. (E) Effect of increasing amounts of PCNA on the dCTP incorporation opposite 8-oxo-G by DNA pol λ (▵) or DNA pol β (○), in the presence of 0.1 M NaCl, 500 nM RP-A. Values are the mean of 3 independent replicates. Error bars represent ± SD values. Template and enzymes were as in D. (F) Pull-down was performed as outlined in Materials and Methods. Lane 1: DNA pol λ bound to Ni-beads pulls down RP-A; lane 2: Ni-beads incubated with RP-A; lane 3: DNA pol β bound to Ni-beads pulls down RP-A; lanes 4–6: as lanes 1–3, but in the presence of DNase 1 (6 units); lane 7: input RP-A (100 ng); lane 8: input DNA pol λ (100 ng); lane 9: input DNA pol β (100 ng). (G) For pull-down experiments, DNA pols β and λ were incubated at a 10:1 molar ratio, in the presence of the 5′ biotinylated 1-nt gapped 8-oxo-G DNA template, 1 μM dCTP and in the absence (lanes 1 and 3) or presence (lanes 2 and 4) of 200 nM PCNA and 500 nM RP-A. Bound (pellet, P) and unbound (supernatant, S) proteins were detected by immunoblot. One-half of the supernatant was loaded on the gel. The structure of the template is shown on the right. Lane 5, control reaction in the absence of DNA template.

DNA Pol λ Is Responsible for the Majority of Error-Free Gap Filling in the Presence of an 8-Oxo-G Lesion.

Extracts from WT or DNA pol λ knockout (λ−/−) mouse embryonic fibroblasts (MEFs) were compared for their ability to incorporate either dATP or dCTP opposite 8-oxo-G on a 1-nt gapped substrate. Compared with WT MEFs, λ−/− cells showed an increased dATP incorporation (Fig. 3A and Fig. S5A). The dCTP vs. dATP bias of λ−/− MEFs dropped from 2.5 to 0.8. Next, the effect of Aphidicolin, a specific inhibitor of the family B DNA pols α, δ, ε, and ζ, was evaluated on WT, λ−/−, and DNA pol β knockout (β−/−) MEFs, in the absence (Fig. S5 B and C) or the presence (Fig. 3B) of compound Ic (N-9-fluorenylmethoxycarbonyl-aminoalkyl-triphosphate), a specific inhibitor of DNA pol λ (19). The results are quantified in Fig. 3C. In the absence of DNA pol β error-prone bypass of 8-oxo-G was almost exclusively caused by Aphidicolin-sensitive pols, whereas the majority of error-free bypass (75%) depended on DNA pol λ, as indicated by the effects of the inhibitor Ic. In λ−/− cell extracts there was a significant increase in Aphidicolin-resistant error-prone (i.e., dATP) incorporation, likely caused by a more prominent role of DNA pol β. Similar experiments were repeated in the presence of an undamaged 1-nt gapped template (Fig. 3D). No dATP incorporation was observed opposite a normal G on the template (Fig. 3D, lanes 1–6). The dCTP incorporation was substantially Aphidicolin-resistant in all 3 extracts (Fig. 3D, compare lanes 7, 9, and 11 with lanes 8, 10, and 12). Moreover, λ−/− MEFs showed only a marginal decrease in dCTP incorporation activity (Fig. 3D, compare lanes 7 and 8 with lanes 9 and 10), whereas β−/− cells showed a substantially reduced activity (Fig. 3D, compare lanes 7 and 8 with lanes 11 and 12), suggesting that DNA pol β is the major enzyme acting in normal BER. In summary, our results suggest that, in the cellular milieu, DNA pol λ is preferentially used over DNA pol β in gap-filling reactions in the presence of an 8-oxo-G lesion.

Fig. 3.

Fig. 3.

Open in a new tab

The fidelity of 8-oxo-G bypass correlates with the relative levels of DNA pols λ and β in normal and tumor cells. Reactions were performed with cell extracts under the conditions specified in Materials and Methods. (A) WT (lanes 1–6) or λ−/− (lanes 8–12) MEF extracts (2 μg of total proteins) were incubated in the presence of dATP (lanes 1–3 and 7–9) or dCTP (lanes 4–6 and 10–12). Lanes C, control reactions in the absence of extracts. The structure of the template used is shown at the top. (B) β−/− MEF extracts (2 μg of total proteins) were incubated in the presence of the 1-nt gapped 8-oxo-G DNA template, dATP (lanes 1–4 and 9–11), dCTP (lanes 5–8, 12, and 13), and absence (lanes 1, 5, 9, and 12) or presence of compound Ic (lanes 2–4 and 6–8) or Aphidicolin (lanes 10, 11, and 13). (C) Relative activity, expressed as percentage of elongated primer ends, for dATP (light gray bars) or dCTP (dark gray bars) incorporation with the 1-nt gapped 8-oxo-G DNA template (50 nM) by different combinations of MEF extracts (2 μg of total proteins) and Aphidicolin (0.2 μg/μl) or Ic (200 μM). Values are the mean of 3 independent replicates. Error bars represent ± SD values. (D) Two micrograms of total proteins of WT MEFs (lanes 1, 2, 7, and 8), λ−/− MEFs (lanes 3, 4, 9, and 10), or β−/− MEFs (lanes 5, 6, 11, and 12) were incubated in the presence of 10 nM of the 1-nt gapped undamaged template, dATP (lanes 1–6), or dCTP (lanes 7–12), and in the presence (odd lanes) or absence (evens) of Aphidicolin. The structure of the template used is shown at the top. (E) Western blot analysis was performed by using extracts (20 μg of total proteins) from cen3 or cen3tel cells at different PDs, as indicated at the top. (F) Relative incorporation activity, expressed as percentage of elongated primer ends, of extracts (2 μg) from cen3 and cen3tel cells at different PDs, in the presence of 5 μM dATP (white bars), 5 μM dATP + 0.2 μg/μL Aphidicolin (dark gray bars), 5 μM dCTP (light gray bars) and 5 μM dCTP + 0.2 μg/μL Aphidicolin (black bars), on the 1-nt gapped 8-oxo-G DNA template (50 nM).

Relative Levels of DNA Pol λ and β Correlate with the Fidelity of 8-Oxo-G Bypass During Neoplastic Transformation.

The levels of DNA pol β, DNA pol λ, and the BER enzymes AP-endonuclease (APE1) and XRCC1 were evaluated by Western blot analysis in extracts of cen3tel cells at different population doublings (PDs), ranging from nontumorigenic cen3tel cells (PD34) to highly tumorigenic cen3tel cells (PD1012), together with primary cells (cen3 cells, PD22) (Fig. 3E). Quantification of the bands, normalized toward the tubulin marker, is shown in Fig. S5D. DNA pol λ levels raised rapidly, reaching a peak ≈PD190, whereas DNA pol β showed a slower, but steady, increase, reaching a plateau ≈PD600. Thus, at PD190 DNA pol λ levels were ≈3-fold higher than those of DNA pol β, whereas at PD1012 the situation was reversed, with DNA pol β being 2.5-fold higher than DNA pol λ. Next, we analyzed the ability of extracts from cen3 primary fibroblast, cen3tel cells at PD190 and PD1012 to incorporate dATP or dCTP opposite 8-oxo-G, in the absence or presence of Aphidicolin (Fig. S5E and Fig. 3F). As can be seen, the correct dCTP vs. incorrect dATP incorporation ratio opposite 8-oxo-G decreased from 6 at PD190 to <1 at PD1012 and correlated with the correspondent changes in the relative levels of DNA pol λ to DNA pol β from 3:1 to 1:2.5. The levels of another translesion enzyme, DNA pol ι, which were undetectable before PD190, started to raise at later stages (PD >300) of transformation (Fig. 3E). The increase of this highly error-prone enzyme further suggests that deregulation of specialized DNA pols can play a role in tumor development.

Discussion

Earlier investigations suggested that long-patch (LP) BER involving DNA pols δ/ε was the major mechanism for the repair of A:8-oxo-G mismatches (20); however, in vitro reconstituted LP- and short-patch (SP) BER reactions were rather error-prone (21). In this work, we compared the activities of DNA pol λ and DNA pol β, the 2 major SP-BER pols (22–23), and showed that both DNA pol λ and DNA pol β physically interact with RP-A (Fig. 2F), which, together with PCNA, reduced the efficiency of gap-filling by DNA pol β in the presence of an 8-oxo-G lesion (Fig. 1F), by preventing its binding to the 1-nt gapped DNA (Fig. 2G). As a result, PCNA and RP-A ensure that dCTP is incorporated 750-fold better than dATP opposite 8-oxo-G on 1-nt gapped DNA, rendering DNA pol λ 145-fold more efficient than DNA pol β in processing 1-nt gap DNA intermediates containing an 8-oxo-G lesion. RP-A and PCNA increased the thermodynamic barrier for dCTP incorporation opposite 8-oxo-G by DNA pol β from 2 kcal × mol−1 (24) to ≈4 kcal × mol−1, making this event thermodynamically equivalent to a misincorporation (25). Thus, PCNA and RP-A did not provide DNA pols with novel properties, but rather rendered the DNA substrate containing an 8-oxo-G optimal for DNA pol λ but very inefficient for DNA pol β. Two pathways may then act in the cell to remove A:8-oxo-G mispairs: 1 very error-prone and dependent on DNA pols δ/ε and LP-BER and important for immediate, replication-coupled repair of 8-oxo-G (20), and another more accurate one dependent on DNA pol λ and SP BER that might act to “clean up” the newly replicated DNA in the G2 phase before mitosis.

In Fig. 4, we show a model for the MUTYH-dependent efficient and faithful repair of DNA containing A:8-oxo-G mismatches, resulting from error-prone lesion bypass by replicative DNA pols. This model is based on known protein–protein interactions among BER proteins and on the interactions of the DNA pols λ and β with RP-A. The bifunctional DNA glycosylase MUTYH is recruited to the A:8-oxo-G mismatch through interaction with RP-A and PCNA and removes the A. The action of APE1, another PCNA-interacting protein, which restores a 3′-OH moiety, the resulting gapped intermediate, can be recognized by several factors through competition for PCNA binding. In particular, the PCNA-interacting protein XRCC1 binds in vitro both gapped and nicked DNA intermediates with similar affinity. Binding to the gap by XRCC1 could potentially interfere with the filling step by the DNA pol. RP-A might prevent binding of the XRCC1/LigIII complex to the gap by covering it. PCNA and RP-A will also restrict DNA pol β action while favoring error-free gap filling of the 8-oxo-G-containing gap by DNA pol λ. Importantly, DNA pol β will not be inhibited by PCNA and RP-A on a gapped intermediate bearing a normal G, thus ensuring DNA pol β action during OGG1-dependent BER. After gap filling, RP-A and DNA pol λ dissociate, allowing XRCC1/LigIII binding to the resulting nicked intermediate through interaction with PCNA and subsequent ligation. In this way, the error-free processing is ensured at the gap-filling step by the combined action of PCNA and RP-A, so that only C-terminated 3′ ends are presented for ligation to the XRCC1/LigIII complex.

Fig. 4.

Fig. 4.

Open in a new tab

A model for the coordinated action of PCNA, RP-A, and DNA pol λ during A:8-oxo-G mismatch repair. For details see Discussion.

DNA pol λ is highly expressed in germinal tissues, particularly testis (26). Highly proliferating cells, particularly those that store the genetic information to be passed to the offspring, might be advantaged in having a very efficient and faithful mechanism for removing the dangerous promutagenic mispair A:8-oxo-G. Both DNA pol λ and DNA pol β are also overexpressed in many tumor types, along with the error-prone enzyme DNA pol ι (27). High levels of DNA pol β and ι have been shown to correlate with genomic instability, high frequency of mutations, and tumor progression (28–32). We found that DNA pol λ and β levels rise with different kinetics in cen3tel cells. We also found a close correlation between the relative levels of DNA pols λ and β, the fidelity of 8-oxo-G lesion bypass in these cells, and the different stages of neoplastic transformation. In summary, our data indicated that the auxiliary proteins PCNA and RP-A act as molecular switches to activate the DNA pol λ-dependent highly efficient and faithful repair of A:8-oxo-G mismatches in human cells, and that this pathway might be relevant during the development of cancer. In support of this hypothesis, we have previously provided preliminary evidence that targeting DNA pol λ by specific inhibitors can reduce the proliferation of specific types of tumor cells (33).

Materials and Methods

Chemicals.

Deoxynucleotides were purchased from GeneSpin. Labeled γ[32P] ATP was purchased from GE Healthcare. Aphidicolin was purchased from Sigma. All other reagents were of analytical grade and purchased from Merck or Fluka. The 39-, 32-, 31-, and 25-mer oligonucleotides, either unlabeled or 5′-biotinylated, were purchased from MWG. Streptavidine-coupled magnetic beads were from Promega.

DNA Substrates.

All oligonucleotides were purified from polyacrylamide denaturing gels (see SI Text for sequences, purification and template preparation details). Annealing of the 72-mer, either undamaged or containing the 8-oxo G lesion (8-oxo-dG-CE Phosphoramidite; Glen Research), with the 5′-labeled 39-mer primer and the 32-, 31-, or 25-mer terminator oligonucleotides generated the 1-, 2-, and 8-nt gapped templates, respectively.

Cells and Extracts.

Immortalized Polλ+/+, Polλ−/−, Polβ+/+, and Polβ−/− MEFs were grown and lysed according to standard protocols (see SI Text for details).

Telomerase immortalized cen3tel cells were obtained from primary cen3 fibroblasts by infection with an hTERT-containing retrovirus as described (15). Primary and immortalized cells were grown and extracts prepared as described (see SI Text).

Antibodies and Proteins.

Antibodies against Tubulin, APE1, and XRCC1 were purchased from Santa Cruz. Antibodies against DNA pol ι were purchased from Abcam. Recombinant human DNA pols β and λ, RP-A, and PCNA were expressed and purified as described (8). Antibodies against DNA pol β and λ (polyclonal rabbit) and polyclonal chicken antibodies against RP-A p70 were from U.H. and G.M. laboratories.

Pull-Down Assays.

Ni-NTA agarose beads.

Purified recombinant DNA pol λ (3 μg) and purified recombinant DNA pol β (3 μg) were coupled to Ni-NTA beads and used in pull-down experiments with 800 ng of purified recombinant RP-A, as described in SI Text.

Biotinylated substrate.

For details see SI Text.

EMSAs.

Purified recombinant RP-A was incubated 5 min at 37 °C in the presence of 5′-labeled 1-nt gapped substrate, either undamaged or containing the 8-oxo-G lesion, in 10 μL of reaction buffer [50 mM Tris·HCl (pH 7.0), 0.25 mg/ml BSA, 1 mM DTT, and 1 mM Mg2+]. Samples were mixed with nondenaturing gel loading buffer [40% (wt/vol) sucrose, 0.25% bromophenol blue] and subjected to PAGE on a 5% native gel at 4 °C for 2 h at 5 V/cm. Position of the free probe and the protein–DNA complexes on the gel was visualized by laser scanning densitometry.

In Vitro Translesion Assays.

For denaturing gel analysis of DNA synthesis products, the reaction mixtures (10 μL) contained 50 mM Tris·HCl (pH 7.0), 0.25 mg/mL BSA, 1 mM DTT, and 1 mM Mg2+. Concentrations of crude extracts or purified DNA pols λ and β, PCNA, RP-A, dNTPs, and the 5′ 32P-labeled primer/template were as indicated in the figure legends. Reactions were incubated for 5 min at 37 °C (unless otherwise stated) and then stopped by addition of standard denaturing gel loading buffer (95% formamide, 10 mM EDTA, xylene cyanol, and bromophenol blue), heated at 95 °C for 3 min and loaded on a 7 M urea/10% polyacrylamide gel.

Steady-State Kinetic Analysis.

See SI Text for details.

Supplementary Material

Supporting Information
supp_105_52_20689__index.html (652B, html)

Acknowledgments.

This work was supported partly by the Centre National de la Recherche Scientifique and a grant from the Association pour la Recherche sur le Cancer (to G.V.). U.H., B.v.L., E.F., and U.W. are supported by Oncosuisse, Union Bank of Switzerland “im Auftrag eines Kunden,” the Swiss National Science Foundation, and the University of Zürich. G.L. was supported by the French Fondation pour la Recherche Médicale. G.M. was supported by a grant from Vetsuisse Dean FR Althaus and by an Investigator Grant from Associazione Italiana Ricerca sul Cancro. G.V. is a scientist of Institut National de la Santé et de la Recherche Médicale. Work in C.M.'s laboratory is supported by Fondazione Cariplo Grant 2006-0734. E.C. is supported by a Fondazione Italiana per la Ricerca sul Cancro Fellowship.

Footnotes

The authors declare no conflict of interest.

This article contains supporting information online at www.pnas.org/cgi/content/full/0811241106/DCSupplemental.

References

  • 1.Valko M, Izakovic M, Mazur M, Rhodes CJ, Telser J . Role of oxygen radicals in DNA damage and cancer incidence. Mol Cell Biochem. 2004;266:37–56. doi: 10.1023/b:mcbi.0000049134.69131.89. [DOI] [PubMed] [Google Scholar]
  • 2.Collins AR. Oxidative DNA damage, antioxidants, and cancer. BioEssays. 1999;21:238–246. doi: 10.1002/(SICI)1521-1878(199903)21:3<238::AID-BIES8>3.0.CO;2-3. [DOI] [PubMed] [Google Scholar]
  • 3.Avkin S, Livneh Z. Efficiency, specificity, and DNA polymerase dependence of translesion replication across the oxidative DNA lesion 8-oxoguanine in human cells. Mutat Res. 2002;510:81–90. doi: 10.1016/s0027-5107(02)00254-3. [DOI] [PubMed] [Google Scholar]
  • 4.van der Kemp PA, Thomas D, Barbey R, de Oliveira R, Boiteux S. Cloning and expression in Escherichia coli of the OGG1 gene of Saccharomyces cerevisiae, which codes for a DNA glycosylase that excises 7,8-dihydro-8-oxoguanine and 2,6-diamino-4-hydroxy-5-N-methylformamidopyrimidine. Proc Natl Acad Sci USA. 1996;93:5197–5202. doi: 10.1073/pnas.93.11.5197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.McCann JA, Berti PJ. Adenine release is fast in MutY-catalyzed hydrolysis of G:A and 8-Oxo-G:A DNA mismatches. J Biol Chem. 2003;278:29587–29592. doi: 10.1074/jbc.M212474200. [DOI] [PubMed] [Google Scholar]
  • 6.Ohtsubo T, et al. Identification of human MutY homolog (hMUTYH) as a repair enzyme for 2-hydroxyadenine in DNA and detection of multiple forms of hMUTYH located in nuclei and mitochondria. Nucleic Acids Res. 2000;28:1355–1364. doi: 10.1093/nar/28.6.1355. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Russo MT, De Luca G, Degan P, Bignami M. Different DNA repair strategies to combat the threat from 8-oxoguanine. Mutat Res. 2007;614:69–76. doi: 10.1016/j.mrfmmm.2006.03.007. [DOI] [PubMed] [Google Scholar]
  • 8.Maga G, et al. 8-oxo-guanine bypass by human DNA polymerases in the presence of auxiliary proteins. Nature. 2007;447:606–608. doi: 10.1038/nature05843. [DOI] [PubMed] [Google Scholar]
  • 9.Ramadan K, Shevelev I, Hubscher U. The DNA-polymerase-X family: Controllers of DNA quality? Nat Rev Mol Cell Biol. 2004;5:1038–1043. doi: 10.1038/nrm1530. [DOI] [PubMed] [Google Scholar]
  • 10.Crespan E, Hubscher U, Maga G. Error-free bypass of 2-hydroxyadenine by human DNA polymerase λ with proliferating cell nuclear antigen and replication protein A in different sequence contexts. Nucleic Acids Res. 2007;35:5173–5181. doi: 10.1093/nar/gkm568. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Maga G, et al. The human DNA polymerase lambda interacts with PCNA through a domain important for DNA primer binding and the interaction is inhibited by p21/WAF1/CIP1. FASEB J. 2004;18:1743–1745. doi: 10.1096/fj.04-2268fje. [DOI] [PubMed] [Google Scholar]
  • 12.Maga G, et al. DNA elongation by the human DNA polymerase λ polymerase and terminal transferase activities are differentially coordinated by proliferating cell nuclear antigen and replication protein A. J Biol Chem. 2005;280:1971–1981. doi: 10.1074/jbc.M411650200. [DOI] [PubMed] [Google Scholar]
  • 13.Maga G, Shevelev I, Villani G, Spadari S, Hubscher U. Human replication protein A can suppress the intrinsic in vitro mutator phenotype of human DNA polymerase lambda. Nucleic Acids Res. 2006;34:1405–1415. doi: 10.1093/nar/gkl032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Maga G, et al. Human DNA polymerase λ functionally and physically interacts with proliferating cell nuclear antigen in normal and translesion DNA synthesis. J Biol Chem. 2002;277:48434–48440. doi: 10.1074/jbc.M206889200. [DOI] [PubMed] [Google Scholar]
  • 15.Mondello C, et al. Karyotype instability and anchorage-independent growth in telomerase-immortalized fibroblasts from two centenarian individuals. Biochem Biophys Res Commun. 2003;308:914–921. doi: 10.1016/s0006-291x(03)01484-0. [DOI] [PubMed] [Google Scholar]
  • 16.Zongaro S, et al. Stepwise neoplastic transformation of a telomerase-immortalized fibroblast cell line. Cancer Res. 2005;65:11411–11418. doi: 10.1158/0008-5472.CAN-05-1140. [DOI] [PubMed] [Google Scholar]
  • 17.Singhal RK, Wilson SH. Short gap-filling synthesis by DNA polymerase β is processive. J Biol Chem. 1993;268:15906–15911. [PubMed] [Google Scholar]
  • 18.Lao Y, Lee CG, Wold MS. Replication protein A interactions with DNA. 2. Characterization of double-stranded DNA-binding/helix-destabilization activities and the role of the zinc-finger domain in DNA interactions. Biochemistry. 1999;38:3974–3984. doi: 10.1021/bi982371m. [DOI] [PubMed] [Google Scholar]
  • 19.Crespan E, et al. Incorporation of non-nucleoside triphosphate analogues opposite to an abasic site by human DNA polymerases β and λ. Nucleic Acids Res. 2005;33:4117–4127. doi: 10.1093/nar/gki723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Parlanti E, Fortini P, Macpherson P, Laval J, Dogliotti E. Base excision repair of adenine/8-oxoguanine mispairs by an aphidicolin-sensitive DNA polymerase in human cell extracts. Oncogene. 2002;21:5204–5212. doi: 10.1038/sj.onc.1205561. [DOI] [PubMed] [Google Scholar]
  • 21.Hashimoto K, Tominaga Y, Nakabeppu Y, Moriya M. Futile short-patch DNA base excision repair of adenine:8-oxoguanine mispair. Nucleic Acids Res. 2004;32:5928–5934. doi: 10.1093/nar/gkh909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Braithwaite EK, et al. DNA polymerase λ mediates a back-up base excision repair activity in extracts of mouse embryonic fibroblasts. J Biol Chem. 2005;280:18469–18475. doi: 10.1074/jbc.M411864200. [DOI] [PubMed] [Google Scholar]
  • 23.Tano K, et al. Interplay between DNA polymerases β and λ in repair of oxidation DNA damage in chicken DT40 cells. DNA Repair (Amst) 2007;6:869–875. doi: 10.1016/j.dnarep.2007.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Plum GE, Grollman AP, Johnson F, Breslauer KJ. Influence of the oxidatively damaged adduct 8-oxodeoxyguanosine on the conformation, energetics, and thermodynamic stability of a DNA duplex. Biochemistry. 1995;34:16148–16160. doi: 10.1021/bi00049a030. [DOI] [PubMed] [Google Scholar]
  • 25.Florian J, Goodman MF, Warshel A. Computer simulation studies of the fidelity of DNA polymerases. Biopolymers. 2003;68:286–299. doi: 10.1002/bip.10244. [DOI] [PubMed] [Google Scholar]
  • 26.Aoufouchi S, et al. Two novel human and mouse DNA polymerases of the polX family. Nucleic Acids Res. 2000;28:3684–3693. doi: 10.1093/nar/28.18.3684. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Albertella MR, Lau A, O'Connor MJ. The overexpression of specialized DNA polymerases in cancer. DNA Repair (Amst) 2005;4:583–593. doi: 10.1016/j.dnarep.2005.01.005. [DOI] [PubMed] [Google Scholar]
  • 28.Frechet M, Canitrot Y, Cazaux C, Hoffmann JS. DNA polymerase β imbalance increases apoptosis and mutagenesis induced by oxidative stress. FEBS Lett. 2001;505:229–232. doi: 10.1016/s0014-5793(01)02834-4. [DOI] [PubMed] [Google Scholar]
  • 29.Frechet M, et al. Deregulated DNA polymerase β strengthens ionizing radiation-induced nucleotidic and chromosomal instabilities. Oncogene. 2002;21:2320–2327. doi: 10.1038/sj.onc.1205295. [DOI] [PubMed] [Google Scholar]
  • 30.Yang J, Chen Z, Liu Y, Hickey RJ, Malkas LH. Altered DNA polymerase ι expression in breast cancer cells leads to a reduction in DNA replication fidelity and a higher rate of mutagenesis. Cancer Res. 2004;64:5597–5607. doi: 10.1158/0008-5472.CAN-04-0603. [DOI] [PubMed] [Google Scholar]
  • 31.Zhang Y, Yuan F, Wu X, Taylor J-S, Wang Z. Response of human DNA polymerase ι to DNA lesions. Nucleic Acids Res. 2001;29:928–935. doi: 10.1093/nar/29.4.928. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kauffmann A, et al. High expression of DNA repair pathways is associated with metastasis in melanoma patients. Oncogene. 2008;27:565–573. doi: 10.1038/sj.onc.1210700. [DOI] [PubMed] [Google Scholar]
  • 33.Locatelli GA, et al. Diketo hexenoic acid derivatives are novel selective non-nucleoside inhibitors of mammalian terminal deoxynucleotidyl transferases, with potent cytotoxic effect against leukemic cells. Mol Pharmacol. 2005;68:538–550. doi: 10.1124/mol.105.013326. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supporting Information
supp_105_52_20689__index.html (652B, html)
0811241106_0811241106SI.pdf (746.7KB, pdf)

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES