DNA Polymerase II Supports the Replicative Bypass of N²-Alkyl-2′-deoxyguanosine Lesions in Escherichia coli Cells

Abstract

Alkylation represents a main form of DNA damage. The N² position of guanine is frequently alkylated in DNA. The SOS-induced polymerases have been shown to be capable of bypassing various DNA damage products in Escherichia coli. Herein, we explored the influences of four N²-alkyl-dG lesions (alkyl = ethyl, n-butyl, isobutyl, or sec-butyl) on DNA replication in AB1157 E. coli cells and the corresponding strains with polymerases (Pol) II, IV, and V being individually or simultaneously knocked out. We found that N²-Et-dG is slightly less blocking to DNA replication than the N²-Bu-dG lesions, which display very similar replication bypass efficiencies. Additionally, Pol II and, to a lesser degree, Pol IV and Pol V are required for the efficient bypass of the N²-alkyl-dG adducts, and none of these lesions was mutagenic. Together, our results support that the efficient replication across small N²-alkyl-dG DNA adducts in E. coli depends mainly on Pol II.

Graphical Abstract

Cells are continuously exposed to alkylating agents that can damage DNA, and the N² position of guanine is a common site of alkylation.^1,2 For instance, the N² of guanine is susceptible to reaction with formaldehyde and acetaldehyde, which can be produced endogenously or present in external sources, including diesel exhaust, cigarette smoke, etc.^3,4 In addition, methylglyoxal, an air pollutant and a byproduct of endogenous glycolysis, can induce the stable N²-(1-carboxyethyl)-2′-deoxyguanosine (N²-CE-dG) in DNA.⁵

DNA adducts, if left unrepaired, can block DNA replication and induce nucleobase substitutions in that process; if mutations occur in oncogenes or tumor suppressor genes, these adducts may contribute to carcinogenesis.^5–7 To ameliorate the genotoxic effects of DNA lesions, cells are equipped with various DNA repair machineries to remove these lesions. Cells are also evolved with DNA damage tolerance pathways to cope with unrepaired lesions, where translesion synthesis (TLS) is one of these pathways for overcoming replication blockage conferred by DNA lesions.^1,8,9 In Escherichia coli, three TLS polymerases (Pol II, Pol IV, and Pol V) can be induced under SOS conditions; Pol II and Pol IV participate in mainly error-free TLS of specific DNA lesions, whereas Pol V can bypass a wide range of DNA lesions in a more error-prone manner.¹⁰

Several studies have been conducted to examine the effects of N²-alkyl-dG adducts on DNA replication in cells. Yuan et al.¹¹ showed that the R diastereomer of N²-CE-dG blocks DNA replication more strongly than the S diastereomer in E. coli, and replicative bypass of these lesions is largely error-free, where Pol IV is required for the accurate and efficient bypass of these lesions. In addition, faithful replication of N²-CE-dG and a number of simple N²-alkyl-dG lesions (alkyl = Me, Et, nPr, and nBu) in mammalian cells requires polymerases κ and ι, where Pol κ is the mammalian ortholog of E. coli Pol IV.^11,12 On the other hand, Shrivastav et al.¹³ found that the replication across N²-Me-dG and N²-Et-dG is accurate and efficient in E. coli cells; depletion of Pol IV, however, does not perturb the efficiency or fidelity of replication across these lesions. It has not yet been investigated systematically how simple N²-alkyl-dG lesions influence DNA replication in E. coli cells or how the three SOS-induced DNA polymerases modulate the replication across these lesions. To answer these questions, we examined the efficiencies and fidelities of replication across four N²-alkyl-dG lesions with different sizes and various structures of alkyl groups (N²-Et-dG, N²-nBu-dG, N²-iBu-dG, and N²-sBu-dG, Figure 1) in E. coli cells and the functions of the SOS-induced DNA polymerases in supporting their replication bypass.

Figure 1.

N²-Alkyl-dG lesions investigated in this study.

We utilized a modified competitive replication and adduct bypass (CRAB) assay¹¹ to explore how replication efficiency and fidelity of a single-stranded M13 plasmid are influenced by the presence of site-specifically inserted N²-alkyl-dG lesions (Figure 2 and Figures S1–S5). In brief, lesion-containing single-stranded M13 plasmids were mixed individually with a lesion-free competitor plasmid at a specific molar ratio and allowed to replicate in SOS-induced AB1157 E. coli cells that are proficient in TLS or with one or more SOS-induced DNA polymerases being genetically depleted. In this respect, the competitor genome, which harbors three more nucleotides but is lesion-free, acts as an internal standard for measuring the replication efficiency across the damage site. After progeny genome isolation, PCR amplification, and restriction digestion, the released oligodeoxyribonucleotides (ODNs) were analyzed by LC-MS/MS and native PAGE to identify and quantify the replication products.

Figure 2.

Restriction digestion and postlabeling method for determining the cytotoxic and mutagenic properties of N²-alkyl-dG lesions in wild-type AB1157 E. coli and the isogenic strains deficient in one or more SOS-induced DNA polymerases. (A) Restriction digestion and selective radiolabeling of the original lesion-containing strand or its complementary strand. (B,C) Representative gel images showing the BbsI/MlucI-produced restriction fragments of interest from the PCR products of progeny genomes of the indicated lesion- or control dG-containing plasmids isolated from SOS-induced wild-type (WT) and Pol II-deficient AB1157 E. coli cells. (D) Bypass efficiencies of N²-alkyl-dG lesions in SOS-induced wild-type AB1157 cells and isogenic polymerase-deficient cells. (E) Bypass efficiencies of N²-alkyl-dG lesions in wild-type and Pol II-deficient AB1157 E. coli cells without SOS induction. The bypass efficiency was calculated by using the following equation: bypass efficiency (%) = (10mer lesion signal/13mer competitor signal)/(10mer control signal/13mer competitor signal) × 100%. The data represent the mean ± SEM of results acquired from three independent replication experiments. *, 0.01 ≤ P < 0.05; **, 0.001 ≤ P < 0.01; ***, P < 0.001. The P values in (D) were calculated using a two-tailed, unpaired t test, and the values refer to the comparisons between SOS-induced WT and the isogenic TLS polymerase-deficient cells (listed above the columns for the polymerase-deficient cells).

The results from native PAGE analyses of the ensuing radiolabeled fragments revealed that no mutagenic products were induced by any of the N²-alkyl-dG lesions, and a similar finding was made from LC-MS/MS analysis of the corresponding non-radiolabeled restriction fragments (Figure 2B,C and Figures S3–S5). We also determined the bypass efficiencies of the N²-alkyl-dG lesions by measuring the ratio of intensity of the 10mer band arising from the lesion-containing genome over that of the 13mer band emanating from the lesion-free competitor genome and further normalizing the ratio to that observed for the control dG-containing genome. The results showed that all four N²-alkyl-dG lesions could impede DNA replication in SOS-induced E. coli cells, with the bypass efficiencies for N²-Et-dG, N²-nBu-dG, N²-iBu-dG, and N²-sBu-dG being 36.5, 31.5, 27.1, and 28.5%, respectively. Hence, the three N²-Bu-dG lesions display blockage effects on DNA replication slightly stronger than that with N²-Et-dG; the replication efficiencies across the three N²-Bu-dG lesions are, nonetheless, not appreciably affected by the structures of the butyl groups (nBu, iBu, and sBu) (Figure 2D).

It is worth noting that the bypass efficiency for N²-Et-dG observed in the present study is lower than what was observed previously by Shrivastav et al.,¹³ which could be attributed to different flanking sequences of the lesion employed in the two studies. In this vein, sequence contexts surrounding DNA damage sites are known to modulate the efficiencies and fidelities of DNA replication across these sites.^14,15 The bypass efficiencies for N²-Et-dG and N²-nBu-dG obtained from this study were also lower than what Wu et al.¹² reported recently for the same lesions in HEK293T human embryonic kidney cells, which could be due to differences in replication machineries in E. coli and human cells.

We next examined the roles of Pol II, Pol IV, and Pol V in bypassing the N²-alkyl-dG adducts by performing the replication experiments using the SOS-induced isogenic E. coli strains where these polymerases were individually or concurrently ablated. Except that depletion of Pol IV alone did not give rise to significant drops in bypass efficiencies for N²-Et-dG or N²-nBu-dG, individual depletion of each of the three SOS-induced DNA polymerases led to substantial attenuations in bypass efficiencies for all four N²-alkyl-dG lesions, with the most pronounced decreases being observed for the Pol II-deficient background (Figure 2D). Additional drops in bypass efficiencies were observed for the three N²-Bu-dG lesions in Pol IV and Pol V double knockout background compared to those with depletion of either polymerase alone (Figure 2D).

For comparison, we also examined how these lesions modulate the replicative bypass of these lesions in wild-type and Pol II-depleted AB1157 cells without SOS induction. Our results showed that SOS induction led to augmented bypass efficiencies for all four lesions in AB1157 cells (Figure 2D,E). Depletion of Pol II, however, does not alter appreciably the bypass efficiencies for any of the four N²-alkyl-dG lesions in AB1157 cells without SOS induction (Figure 2E). It was shown previously that SOS induction could give rise to a 7-fold elevation in expression level of Pol II.¹⁶ Thus, the lack of effect of Pol II deletion on the bypass efficiencies of these lesions in uninduced E. coli cells could be due to the relatively low level of expression of Pol II in wild-type AB1157 cells without SOS induction.

Exposure to endogenous and exogenous genotoxic agents can lead to the formation of various DNA lesions, many of which block replicative DNA polymerases and require TLS polymerases for their replicative bypass.¹⁷ A large body of literature revealed the roles of B- and Y-family polymerases in modulating the cytotoxic and mutagenic properties of various DNA lesions.^18,19 In this vein, Pol IV and Pol V were found to participate in error-free TLS and induce a −1 frameshift mutation at the site of an N²-dG adduct of benzo[a]pyrene; Pol II, however, is involved in bypassing the bulky N²-dG adduct of acetylaminofluorene and elicits a −2 frameshift mutation at the lesion site.^20,21 In addition, previous in vitro biochemical studies showed that purified Pol IV preferentially inserts a dCMP opposite N²-Et-dG, N²-iBu-dG, and N²-CE-dG in template DNA,^11,22 though the kinetic parameters for nucleotide insertions opposite simple N²-alkyl-dG lesions were not measured. It will be important to determine, in the future, the efficiencies and fidelities of Pol II- and Pol IV-catalyzed nucleotide incorporation opposite the N²-alkyl-dG lesions.

The major finding from the present study is about the contributions of the three SOS-induced DNA polymerases, namely, Pol II, Pol IV, and Pol V, in bypassing four minor-groove N²-alkyl-dG adducts (alkyl = Et, nBu, iBu, and sBu) in E. coli cells. Similar to what were observed for N²-furfuryl-dG and N²-tetrahydrofuran-2-yl-methyl-dG lesions,¹³ we found that the smaller N²-Et-dG and N²-Bu-dGs were not mutagenic in AB1157 cells or any of the isogenic polymerase-deficient strains tested. Our results support that all three TLS polymerases in E. coli are involved in the error-free TLS of these four lesions. A small alkyl group adducted to the N² of guanine does not strongly impair the base-pairing property of the nucleobase, whereas those adducts with large alkyl groups can induce DNA frameshifts or single-base substitutions during DNA replication.²⁰ Meanwhile the loss of Pol II results in the most marked diminutions in bypass efficiencies in SOS-induced AB1157 cells, underscoring the major role of this polymerase in overcoming the replication blockage imposed by these N²-alkyl-dG lesions.

It is worth noting that Pol IV was found to be the major DNA polymerase responsible for bypassing N²-CE-dG lesions in vivo,¹¹ whereas Pol II was the main polymerase involved in bypassing the N²-dG lesions with small alkyl groups. The exact reason for these differences remains unclear and awaits further investigation. In this vein, the X-ray crystal structure of a ternary complex of E. coli Pol II, duplex DNA, and an incoming dCTP showed that the active site of the polymerase facing the minor-groove N² position of template dG is quite spacious,²³ which should be able to accommodate alkyl groups adducted to the N² of dG (Figure S6). This may explain Pol II’s role in supporting the accurate and efficient bypass of N²-alkyl-dG lesions. In this context, it is of note that minor-groove O²-alkyl-dT lesions are highly mutagenic in E. coli cells and their efficient bypass mainly requires Pol V.²⁴

These previously published results, together with the observations made from the present study, indicate that all three TLS polymerases can participate in bypassing minor-groove lesions; Pol V tends to transverse, in an error-prone manner, those lesions with the hydrogen bonding properties of the nucleobases being disrupted (e.g. the O²-alkyl-dT lesions), whereas Pol II and Pol IV tend to bypass accurately minor-groove N²-alkyl-dG lesions. Moreover, DNA Pol II, which is B-family DNA polymerase, was also found to participate in the error-free bypass of a major-groove N⁶-benz[a]anthracene adenine adduct.²⁰ Along this line, several B-family polymerases were shown to possess a conserved motif that scans the DNA minor groove for lesions and misincorporations.²⁵ Therefore, Pol II is capable of bypassing accurately both major- and minor-groove alkylated DNA lesions. Together, the results from previous studies and the current work reveal that nuances of TLS can be modulated by the subtle differences in chemical structures of DNA lesions.