Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Mar 10.
Published in final edited form as: Mutat Res. 2021 Mar 10;822:111742. doi: 10.1016/j.mrfmmm.2021.111742

DNA damage-signaling, homologous recombination and genetic mutation induced by 5-azacytidine and DNA-protein crosslinks in Escherichia coli

Julie A Klaric 1,a, David J Glass 1, Eli L Perr 1,b, Arianna D Reuven 1,c, Mason J Towne 1, Susan T Lovett 1,*
PMCID: PMC8288033  NIHMSID: NIHMS1713395  PMID: 33743507

Abstract

Covalent linkage between DNA and proteins produces highly toxic lesions and can be caused by commonly used chemotherapeutic agents, by internal and external chemicals and by radiation. In this study, using Escherichia coli, we investigate the consequences of 5-azacytidine (5-azaC), which traps covalent complexes between itself and the Dcm cytosine methyltransferase protein. DNA protein crosslink-dependent effects can be ascertained by effects that arise in wild-type but not in dcm∆ strains. We find that 5-azaC induces the bacterial DNA damage response and stimulates homologous recombination, a component of which is Dcm-dependent. Template-switching at an imperfect inverted repeat (“quasipalindrome”, QP) is strongly enhanced by 5- azaC and this enhancement was entirely Dcm-dependent and independent of double-strand break repair. The SOS response helps ameliorate the mutagenic effect of 5-azaC but this is not a result of SOS-induced DNA polymerases since their induction, especially PolIV, seems to stimulate QP-associated mutagenesis. Cell division regulator SulA was also required for recovery of QP mutants induced by 5-azaC. In the absence of Lon protease, Dcm-dependent QP-mutagenesis is strongly elevated, suggesting it may play a role in DPC tolerance. Deletions at short tandem repeats, which occur likewise by a replication template-switch, are elevated, but only modestly, by 5-azaC. We see evidence for Dcm-dependent and-independent killing by 5- azaC in sensitive mutants, such as recA, recB, and lon; homologous recombination and deletion mutations are also stimulated in part by a Dcm-independent effect of 5-azaC. Whether this occurs by a different protein/DNA crosslink or by an alternative form of DNA damage is unknown.

Keywords: DNA repair, DNA damage response, genome instability, genetic rearrangements, homologous recombination

1. INTRODUCTION

DNA protein crosslinks (DPCs) are a common spontaneous source of DNA damage and can be formed by both nonenzymatic or enzymatic mechanisms. Nonenzymatic DPCs can be induced by radiation and a wide variety of chemicals (Barker et al. 2005). For example, reactive oxygen species and aldehydes, either encountered from the environment or endogenously formed by metabolic processes, can covalently link proteins to DNA. In addition, a number of conserved cellular enzymes are covalently linked to DNA as a reaction intermediate (Ide et al. 2011). Two examples are DNA topoisomerases and DNA cytosine methyltransferases (“CMeTs”). Certain drugs trap these normally short-lived covalent intermediates. The quinolone class of antibiotics and anticancer drugs etoposide and doxorubicin are topoisomerase type IIA poisons for bacterial and eukaryotic enzymes, respectively, and trap the covalent tyrosyl-toposiomerase/phosphoryl-DNA intermediate (also known as the “cleaved complex”). Camptothecin, another chemotherapeutic drug, is a similar poison for eukaryotic type IB topoisomerases (Drlica and Zhao 1997; Nitiss 2009; Pommier et al. 2010). In addition, cytosine methyl transferases (CMeTs) form a covalent complex with C6 of cytosine in their reaction to methylate the C5 of cytosine in DNA or RNA (Figure 1)(Santi et al. 1984; Friedman 1985). The nucleotide analog 5-azacytidine (Figure 1) inhibits the completion of the reaction, trapping the protein covalently attached to the nucleotide base.

Figure 1.

Figure 1.

Open in a new tab

Cytidine-derived structures

DPCs impede both RNA transcription (Nakano et al. 2017) and replication fork progression (Kuo et al. 2007). In humans and other mammals, the inability to repair DPCs accelerates aging and leads to cancer proneness (Craft et al. 1987; Izzotti et al. 1999; Zahn et al. 1999; Wu et al. 2002; Swenberg et al. 2011; Garaycoechea et al. 2012).

In bacteria, DPCs caused by topoisomerase II poisons or 5-azacytidine can lead to double-strand breaks (DSBs). Accordingly, mutants in double-strand break (DSB) repair factors RecA and RecB are sensitive to both to quinolones and 5-azaC (Bhagwat and Roberts 1987; Lal et al. 1988; Butala et al. 2009; Salem et al. 2009; Ide et al. 2011), although there may be other toxic lesions in addition to DSBs. For topoisomerase II poisons, a cytotoxic lesion is some processed form of the trapped “cleaved complex” since killing can be suppressed by inhibitors of RNA and protein synthesis (reviewed nicely in (Drlica and Zhao 1997)). Even in the absence of frank DSB formation, these covalent complexes can block the completion of replication, allowing a second replication fork converging onto the site to produce a broken chromosome (Cox et al. 2000).

In bacteria, 5-azaC is a useful tool to study effects of DPCs in vivo whose targets are C5- cytosine methyl transfereases (CMeT) in restriction/modification systems. Restriction/modification systems are widespread throughout bacteria, and consist of an endonuclease component and a methyltransferase (Roberts et al. 2015). These systems are considered to be primitive innate immune systems, leading to the destruction of invading foreign DNA. A proportion of restriction systems use C5-cytosine methylation (Fig. 1) to provide immunity to restriction endonuclease cleavage (other systems methylate the N6 of adenine or the N4 of cytosine to confer immunity). Because many of these modification enzymes methylate specific DNA recognition sequences and are nonessential in the laboratory environment, we can target DPCs to particular sites to study their consequence in vivo. In addition, in the absence of the methyltransferase the adduct will not form, providing confidence that any effect is due to the DPC alone.

We have recently documented that DPCs in Escherichia coli dramatically stimulate a particular class of mutation. This class of mutation arises by a replication template-switch reaction in an imperfect inverted repeat sequence (or “quasipalindrome”, “QP”)(Ripley 1982; Glickman and Ripley 1984; Yoshiyama and Maki 2003; Dutra and Lovett 2006; Lovett 2017). QPs often constitute mutational hotspots, both in bacteria and in eukaryotes (Hampsey et al. 1988; Greenblatt et al. 1996; Bissler 1998; Viswanathan et al. 2000; Yoshiyama et al. 2001; Yoshiyama and Maki 2003). Using QP mutational reporter strains, we discovered that the DPC-producing agents formaldehyde, ciprofloxacin and 5-azaC are potent mutagens for this class of mutations (Laranjo et al. 2018). For 5-azaC, the mutagenic effect requires E. coli’s endogenous CMeT, Dcm, indicating that it is the trapped DPC that induces mutation at the QP site (Laranjo et al. 2018).

Surprisingly, although the genes required in DSB repair, recA and recB, are necessary to avoid to the lethal effects of DPCs, they are not required for their mutagenicity (Laranjo et al. 2018). This suggests that the mutagenic lesions include something other than DSBs. That the mutagenic lesion may be a block to the DNA replication fork is suggested by the fact that impairment in the DnaB fork helicase greatly elevates QP-associated mutation (QPM) in the absence of DNA damage (Laranjo et al. 2018). To further explore how DPCs lead to genomic instability, we examine here the consequence of DPCs on genetic recombination and DNA damaging signaling. We test the effect of the SOS response, translesion DNA polymerase and the Lon and ClpP proteases on mutagenicity of 5-azaC in E. coli at a QP site. We additionally quantify effects on genomic rearrangements at short tandem repeats, which are caused by a different type of template-switching during DNA replication (Lovett et al. 1993; Lovett and Feschenko 1996). This work finds that DPCs are broadly mutagenic, recombinogenic and inducers of the SOS response. The SOS response is protective in the avoidance of QPM, although unbalanced expression of translesion DNA polymerases, particularly DNA polymerase IV (DinB), is mutagenic. Most of the effects of 5-azaC are due to DPCs formed by the E. coli CMeT, Dcm; however, we also find evidence for Dcm-independent DNA lesions that are lethal or mutagenic in certain sensitive genetic backgrounds.

2. MATERIALS AND METHODS

2.1. Strains and growth conditions:

All of the strains used are derivatives of Escherichia coli K-12 MG1655 (Bachmann 1972; Table 1). Isogenic strains of the indicated genotypes were constructed by P1vir transduction. DNA transformation by electroporation (Dower et al. 1988) was used to introduce plasmids and oligonucleotides for recombineering. The strains were grown at 37 °C in Luria broth (LB, Lennox formulation) medium, consisting of 1% Bacto-tryptone, 0.5% yeast extract, 0.5% sodium chloride and, for plates, 1.5% Bacto-agar. Tetracycline (Tc, 15 µg/mL), chloramphenicol (Cm, 30 µg/mL), ampicillin (Ap, 100 µg/mL) and kanamycin (Km, 45–60 μg/mL) were used for genetic selections. Lac+ reversion mutants were selected on lactose minimal medium (Lac Min XI) containing 56/2 salts (Willetts et al. 1969), 0.2% lactose (Sigma Aldrich, St. Louis, MO, USA), 0.001% thiamine (Sigma Aldrich), and 2% agar (ThermoFisher, Sparks, MD, USA). X-gal (40 µg/mL) and IPTG (0.1 mM) (both from Gold Bio, St. Louis, MO, USA) were included in lactose selection medium as a visual aid for counting colonies. Glucose minimal (Glu Min XI) control plates were made similarly but contained 0.2% glucose in place of lactose. Lactose-X-gal papillation plates were used in disk diffusion assays for mutagenic and recombinogenic effects of DPC agents as previously described (Laranjo et al. 2018; Klaric et al. 2020).

TABLE 1:

E. coli K-12 strains used in this study

Strain number Genotype
MG1655 rph-1 Bachmann 1972
STL1464 lexA::Tn5 Lab collection
STL7180 recA::cat Lab collection
STL7428 sulA::FRT kan Lab collection
STL12071 lexA3 malF::Tn10 kan Lab collection
STL13814 lacZ(C1384G) mphC281::Tn10 Seier et al. 2011
STL14025 lacZ(+11) kan mphC281::Tn10 Seier et al. 2011
STL14095 lacZ(−1G1474) mphC-281::Tn10 Seier et al. 2011
STL14776 lacZ(QP5) xonA::FRT xseA::FRT mphC-281::Tn10 Seier et al. 2011
STL15144 lacZ(QP5) lexA3 malF::Tn10 kan mphC281::Tn10 Laranjo et al. 2018
STL15823 lacZ(QP6) xonA::FRT xseA::FRT mphC-281::Tn10 Laranjo et al. 2018
STL16519 lacZ(QP6) lexA3 malF::Tn10 kan mphC281::Tn10 Laranjo et al. 2018
STL17976 lacZ(QP5) xonA::FRT kan mphC- 281::Tn10 Laranjo et al. 2018
STL17978 lacZ(QP6) xonA::FRT kan mphC- 281::Tn10 Laranjo et al. 2018
STL17980 lacZ(QP5) xseA::FRT kan mphC- 281::Tn10 Laranjo et al. 2018
STL17982 lacZ(QP6) xseA::FRT kan mphC- 281::Tn10 Laranjo et al. 2018
STL18747 lacZ(QP5) dcm::FRT kan mphC281::Tn10 Laranjo et al. 2018
STL18749 lacZ(QP6) dcm::FRT kan mphC281::Tn10 Laranjo et al. 2018
STL19829 lacZ(QP5) recA::cat mphC281::Tn10 Laranjo et al. 2018
STL19831 lacZ(QP6) recA::cat mphC281::Tn10 Laranjo et al. 2018
STL19843 recB::FRT kan Laranjo et al. 2018
STL19845 recO::FRT kan This work
STL20589 lacZ(QP5) mphC-281::Tn10 resiolate of STL17685 Laranjo et al. 2018
STL20590 lacZ(QP6) mphC-281::Tn10 reisolate of STL15654, Seier et al. 2011
STL20956 lacZ(QP5) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 This work
STL20959 lacZ(QP6) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 This work
STL21382 lacZ(QP5) recO::FRT kan mphC- 281::Tn10 This work
STL21384 lacZ(QP6) recO::FRT kan mphC- 281::Tn10 This work
STL21742 dcm::FRT kan This work
STL21845 sulA::FRT This work
STL21944 dcm::FRT kan recA::cat This work
STL21945 lacZ(QP5) dcm::FRT kan recA::cat mphC-281::Tn10 This work
STL21947 lacZ(QP6) dcm::FRT kan recA::cat mphC-281::Tn10 This work
STL22123 lon::FRT kan This work
STL22147 clpP::FRT kan This work
STL22165 lacZ(QP5) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 xonA::FRT xseA::FRT kan This work
STL22169 lacZ(QP5) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 xseA::FRT kan This work
STL22176 lacZ(QP5) clpP::FRT kan mphC- 281::Tn10 This work
STL22178 lacZ(QP6) clpP::FRT kan mphC- 281::Tn10 This work
STL22196 lacZ(QP5) lon::FRT kan mphC- 281::Tn10 This work
STL22198 lacZ(QP6) lon::FRT kan mphC- 281::Tn10 This work
STL22299 lacZ(QP5) recB::FRT kan mphC- 281::Tn10 This work
STL22300 lacZ(QP6) recB::FRT kan mphC- 281::Tn10 This work
STL22644 lacZ(QP6) dcm::FRT recB::FRT kan mphC-281::Tn10 This work
STL22655 lacZ(QP5) dcm::FRT recB::FRT kan mphC-281::Tn10 This work
STL22672 lexA::Tn5 sulA::FRT This work
STL22683 lacZ(QP5) lexA::Tn5 kan sulA::FRT mphC-281::Tn10 This work
STL22684 lacZ(QP6) lexA::Tn5 kan sulA::FRT mphC-281::Tn10 This work
STL22707 lacZ(QP5) sulA::FRT mphC-281::Tn10 This work
STL22709 lacZ(QP6) sulA::FRT mphC-281::Tn10 This work
STL22811 recO::FRT kan dcm::FRT This work
STL22813 lacZ(QP6) dcm::FRT recO::FRT kan mphC-281::Tn10 This work
STL22824 lacZ(QP5) dcm::FRT recO::FRT kan mphC-281::Tn10 This work
STL22826 lacZ(QP6) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 xseA::FRT kan This work
STL22864 lacZ(QP6) dinB::FRT polB::FRT cat umuC::FRT mphC281::Tn10 xonA::FRT xseA::FRT kan This work
STL22869 lacZ(+11) dcm::FRT kan mphC281::Tn10 This work
STL13814 lacZ(C1384G) dcm::FRT mphC281::Tn10 Seier et al. 2011
STL22884 dcm::FRT lon::FRT kan This work
STL22989 sulA::FRT lon::FRT kan This work
STL22991 lacZ(QP5) lon::FRT kan sulA::FRT mphC-281::Tn10 This work
STL22993 lacZ(QP6) lon::FRT kan sulA::FRT mphC-281::Tn10 This work
STL23017 lacZ(QP5) dcm::FRT lon::FRT kan mphC-281::Tn10 This work
STL23019 lacZ(QP6) dcm::FRT lon::FRT kan mphC-281::Tn10 This work
STL23067 lacZ(QP5) dcm::FRT kan sulA::FRT mphC-281::Tn10 This work
STL23097 lacZ(QP5) dcm::FRT kan lon::FRT sulA::FRT mphC-281::Tn10 This work
STL23099 lacZ(QP5) clpP::FRT kan dcm::FRT mphC-281::Tn10 This work
STL23124 lacZ(QP5) clpP::FRT kan dcm::FRT mphC-281::Tn10 This work
Open in a new tab

2.2. Survival assays:

For survival to 5-azacytidine (5-azaC, Sigma Aldrich), MG1655 and isogenic derivative (Table 1) were grown to mid-log phase (OD600 of approximately 0.2– 0.4), then split and treated at the indicated concentrations or untreated for 2 hr prior to serial dilution and plating on LB media. Survival experiments were performed for at least 3 replicates.

2.3. Luciferase assays:

To measure induction of the SOS response, we used luciferase fusions to the recA and dinB promoters as previously described (Goldfless et al. 2006) and strain backgrounds MG1655, STL19845, STL21742, and STL22811. A single colony was used to inoculate 2 mL of LB+Ap media and grown at 37° with agitation for 2 hours until OD600 ≈ 0.5.

The culture was then diluted 1:100 in fresh LB+Ap media and grown again at 37C with agitation for 2 hours. The culture was then diluted 1:100 into fresh LB+Ap media in a Costar 96 Well Assay Plate (Treated Polystyrene, Black Plate, Clear Bottom). The plate was then grown in a BioTek Cytation 1 Plate Reader at 37° shaking for 75 minutes, with OD600 and luminescence readings being taken every 15 minutes. After 75 minutes, 5-azacytidine was added to the appropriate wells to a final concentration of 12.5 ug/mL. Cultures were then grown for 3.5 hours shaking at 37° With OD600 and luminescence readings taken every 15 minutes. Relative luminescence units (RLU) values were calculated by normalizing the luminescence readings (in counts per second) to the OD600 of the cultures.

2.4. Deletion assays:

Deletion of 101 bp tandem repeat in tetA was performed as described using plasmid pSTL57 (Lovett et al. 1994) and strains MG1655 and STL21742. Deletion of 11 bp tandem repeats in lacZ was performed as described (Seier et al. 2011) with strain STL14025. For experiments involving 5-azacytidine (Sigma Aldrich) or formaldehyde (ThermoFisher), mid-log phase cultures were split, and treated with the drug, or not, for 2 hours at the indicated concentrations before serial dilution and plating on appropriate selective media. Data are presented as the fold-change in deletion frequencies of the treated vs. untreated samples.

2.5. Crossover recombination assay:

Recombination between 411 bp of homology in plasmids pSTL330 and pSTL36 was selected by Tc-resistance as described (Lovett et al. 2002) and strain backgrounds MG1655, STL21742 and STL19845. Mid-log phase cultures were split, and for two hours were treated with 5-azaC at the indicated concentrations before serial dilution and plating on appropriate selective media. Both wt and recO mutants were assayed with 4 independent replicates.

2.6. Gene conversion recombination assay:

Gene conversion is selected in this assay by the Lac+ progeny of a strain carrying two mutant alleles of lacZ, one with an internal deletion at the natural lacZ locus (7.83 centisomes) and an internal 500 bp lacZ fragment integrated at attTn7 near glmS (84.28 centisomes). The lacZ recipient allele was constructed by recombineering using a single-strand DNA oligonucleotide and DNA transformation (Sharan et al. 2009). The 5 nucleotides spanning the active site amino acid codon, from nucleotides 1384–1388 of the lacZ ORF, were deleted by transformation of oligonucleotide 5’ ATCACCCGAG TGTGATCATC TGGTCGCTGGG GAATAGGCCA CGGCGCTAAT CACGACGCGC TGTATCGC in a MG1655 strain carrying the pSIM6 recombineering plasmid (Sharan et al. 2009), yielding strain STL20478 after curing the pSIM6 plasmid with growth at 42°. To construct the recombination donor locus, we used a Tn7-based based system to introduce a DNA fragment carrying 250 bp of homology flanking both sides of the lacZ active site deletion into the attTn7 site, near glmS (McKenzie and Craig 2006). The DNA fragment produced by colony PCR with oligonucleotides 5’ GGGGcccggg GCTGATGAAG CAGAACAACT TTAACGCCGT 3’ and 5’ GCttaattaa ACTGTTACCC ATCGCGTGGG CGTATTCGCA 3’ was inserted into temperature-sensitive Tn7 vector pGRG25 (McKenzie and Craig 2006) after cleavage with PacI and XmaI (New England Biolabs). Transformation of this plasmid into strain STL18091 (MG1655 bglG::FRT kan) was selected by Ap-resistance at 30°, after which the plasmid was cured by growth in LB at 42°; integrants of the Tn7 end-flanked fragments were identified by PCR among survivors as described (Laranjo et al. 2017). This allele was transduced into strain STL20478 by cotransduction with bglG::FRT kan. The kan was then removed using FLP plasmid pCP20 (Datsenko and Wanner 2000). The resulting recombination assay strain carrying both loci (attTn7::’lacZ’ lacZas∆) is STL22043. Homologous recombination in this strain generates a lacZ+ locus in its natural site; the donor locus at attTn7 is unchanged. Strains carrying only the donor or only the recipient lacZ locus never yielded Lac+ progeny, nor did recA strains carrying both alleles (unpublished data and Supplemental Figure 1), confirming that lacZ+ isolates arise by homologous recombination between the loci.

For the recombination assays, cultures were grown in LB to midlog phase, then concentrated 5x by centrifugation and resuspension into LB medium containing 50 µg/ml 5-azaC or not. Growth of these split cultures was continued for 2 hours after which they were harvested by centrifugation, washed with 1 × 56/2 salts and serially diluted. Dilutions were plated on Lac Min XI to determine the number of Lac+ recombinants and on Glucose Min XI and LB for total number of CFU. LB plates were incubated at 37° 1 day and minimal plates for 3–4 days.

2.7. Mutation assays.

We used a chromosomal lacZ mutational reporters specific for G to C transversion and for quasipalindrome-associated template-switch mutations, QP5 and QP6, previously described (Seier et al. 2011). The strains used are listed in Table 1, with the genotype including lacZ(QP5) or lacZ(QP6). C to G transversion mutations were measured using STL13814 ((Seier et al. 2011) and a dcm::FRT kan transductant of this strain, STL22870. +1 G frameshift mutations were measured using strain STL14095 carrying the indicated plasmids. 5-azaC effects were determined with mid-log phase split cultures and with treatment for 2 hours prior to dilution and plating.

2.8. SOS polymerase plasmids:

GATEWAY (Thermo Fisher) recombinational cloning technology was used to clone dinB, polB and umuCD’ into the pBAD18 arabinose-controlled expression vector. From MG1655 colonies, the dinB gene was amplified with the following PCR primers 5’ GGGGACAAGT TTGTACAAAA AAGCAGGCTT CGAAGGAGAT AGAACCatgc gtaaaatcattca and 5’ GGGGACCACT TTGTACAAGA AAGCTGGGTC tcataatccc agcacc, the polB gene with 5’ GGGGACAAGT TTGTACAAAA AAGCAGGCTT CGAAGGAGAT AGAACCgtgg cgcaggcagg ttttatc, 5’ ggggACCACT TTGTACAAGA AAGCTGGGTC tcaaaatagc ccaagttgccc. The umuD’C gene was amplified from plasmid MP1 (Addgene) with primers 5’ GGGGACAAGT TTGTACAAAA AAGCAGGCTT CGAAGGAGAT AGAACCatga agagattgca gctcatg, 5’ ggggACCACT TTGTACAAGA AAGCTGGGT Cttatttgac cctcagtaaa tcag and inserted using BP reactions into pDONR201. LR reactions transferred these insert to the pSTL360 plasmid (Dutra and Lovett 2006), a destination plasmid derived from pBAD18 (Guzman et al. 1995).

3. RESULTS

3.1. Recombination pathways and DPC processing:

The Escherichia coli K-12 genome encodes one cytosine methyl transferase, Dcm (“DNA cytosine methylase”). The function of this gene is not known but it is likely a remnant of a restriction/modification system. Dcm methylates the second C in the sequence “CC[A/T]GG”; this is the same specificity as the EcoRII restriction/modification system, encoded on an E. coli antibiotic-resistance conferring-plasmid, derived from a clinical isolate (Bannister and Glover 1968; Yoshimori et al. 1972; Bhagwat et al. 1986; Bhagwat et al. 1990).

E. coli K-12 wild-type cells are not particularly sensitive to killing by 5-azaC after a 2 hour acute exposure and mutants in dcm are only modestly more resistant (Fig. 2). The fraction of potential sites methylated by Dcm is unknown but there is evidence that not every potential Dcm site is methylated (Bhagwat et al. 1990; Ringquist and Smith 1992); for example, EcoRII overexpression confers additional sensitivity to 5-azaC in dcm+ E. coli (Bhagwat et al. 1990; Krasich et al. 2015). With only the endogenous Dcm CMeT, recA and recB mutants, deficient in DSB repair, show enhanced sensitivity (Fig. 2), confirming previous studies (Bhagwat and Roberts 1987; Lal et al. 1988; Butala et al. 2009; Salem et al. 2009; Ide et al. 2011). This sensitivity is partially dependent on Dcm. There is, however, residual sensitivity, suggesting Dcm-independent toxicity of 5-azaC. This residual sensitivity of recA dcm has also been previously documented (Bhagwat and Roberts 1987).

Figure 2.

Figure 2.

Open in a new tab

Fractional survival of E. coli genetic variants at indicated doses of 5-azaC for 2 hours. Data are average of at least 3 experiments; error bars represent standard deviation. A. Recombination mutants B. SOS response mutants C. Protease mutants.

RecA is required for homologous recombination in E. coli, and it participates in two genetically distinct pathways. One of these, involving the RecBCD nuclease, is specific for recombination arising from DSBs, the other, dependent on RecFOR, is specific for recombinational repair of single-strand gaps (reviewed in (Persky and Lovett 2008)). These pathways are defined by the functions that facilitate the loading of RecA, which RecBCD performs on newly resected DNA and RecFOR on ssDNA gaps, the latter displacing bound single-strand DNA binding protein (SSB). The RecA filament bound to ssDNA not only directly facilitates homology search and DNA strand exchange, but is also the inducing signal for the SOS response, a transcriptional response in E. coli to DNA damage (Simmons et al. 2008; Lovett 2010). We show here that RecO mutants exhibit no sensitivity to 5-azaC (Fig. 2A) unlike mutants in RecA or RecBCD.

The SOS response is induced following replication inhibition (Sassanfar and Roberts 1990) after UV irradiation; it and the response to the DNA crosslinkers mitomycin C and cisplatin is promoted by the RecFOR pathway (Thoms and Wackernagel 1987; Whitby and Lloyd 1995; Keller et al. 2001). To ascertain the participation of the RecFOR and RecBCD pathways in processing DPCs, we measured SOS induction in E. coli after 5-azaC treatment, using a luciferase transcriptional reporter fused to the recA promoter. This promoter is regulated as part of the SOS regulon. In these assays, cells are continuously exposed to sublethal concentrations of 5- azaC, beginning in early log phase and monitored through the growth of the culture into stationary phase. Luminescence of the culture is normalized to OD600 of the culture, measured at 15 minute intervals. 5-azaC does indeed induce expression of the recA promoters reaching a maximum about 2 hrs after introduction of the drug. This induction was mostly dependent on dcm, confirming that DPCs are responsible for the induction, and also dependent, as expected, on recA. For both induced and uninduced expression, we saw a growth-dependent increase in expression, peaking at late exponential phase. We do not understand the mechanistic basis of this effect, which suggests that repression of the SOS response may be mitigated as cells approach stationary phase. This increase was not seen in control cultures with the luciferase operon, lacking the promoter regions, so this signal is dependent on the recA promoter. In addition, there was an unexpected, delayed induction of the recA promoter independent of dcm. We saw a modest reduction and delay in recA promoter induction by 5-azaC in a recO mutant, although the effect is obscured by the elevation of expression in recO untreated cultures, possibly due to the accumulation of spontaneous, SOS-inducing DNA lesions in this genetic background. Induction of the recA promoter by 5-azaC was abolished in recB mutants, although the majority of cells were dead by this point, showing that 5-azaC lesions promote SOS induction via double-strand breaks.

3.2. Does 5-azaC induce recombination, as well as the SOS response?

We measured two types of recombination with genetic assays: one selects for crossovers between 411 bp of homology on low copy plasmids (Lovett et al. 2002) and the other detects nonreciprocal gene conversion between an internally deleted lacZ gene and an ectopic internal ‘lacZ’ gene segment sharing 250 bp of homology to each side of the deletion, inserted at a position across the chromosome, near oriC (see Materials and Methods). Crossovers detected in the plasmid assay are primarily dependent on the RecFOR pathway (Lovett et al. 2002); because these plasmids lack Chi sites that protect from degradation by RecBCD (Anderson and Kowalczykowski 1997); there is no detectable contribution of the RecBCD pathway. Using cultures that were split and treated for 2 hr with 5-azaC (Fig. 4), we saw a modest 2- to 3-fold elevation of crossing-over with 5-azaC relative to untreated cultures. This increase was dependent on recO. 5-azaC induced recombination was dependent on dcm as well, with dcm mean fold-induction values of 0.89 ± 0.18, 0.60 ± 0.21, 0.50± 0.39 at 25, 50 and 75 µg/ml 5-azaC, respectively (n=4).

Figure 4.

Figure 4.

Open in a new tab

Crossover recombination induced by 5-azaC, 2 hr treatment at indicated doses, in split cultures of wt or recO mutant strains. Bars represent mean values of the replicates.

Gene conversion between chromosomal lacZ alleles was more strongly stimulated by 5-azaC, with an over 30- fold increase at 50 µg/ml. (Figure 5). Gene conversion detected in this assay, both constitutive and induced, is entirely RecA-dependent, as shown in a Lactose X-gal papillation plate assay (Unpublished results and Supplemental Figure 1) where Lac+ recombinants outgrow as blue papillae on a white background. Gene conversion was reduced 2-fold by recO in quantitative assays, although there was considerable variation among replicates. We saw no strong evidence for Dcm-dependence of this recombination, so there must be a recombinogenic lesion induced by 5-azaC other than Dcm-DPCs. The residual recombination in recO mutants is presumedly dependent on the alternate RecBCD recombination pathway (Persky and Lovett 2008), but we were unable to test recB and recB recO mutants in this assay, due to their strong sensitivity to 5-azaC. In the papillation assay (Supplemental Figure 1), we detected a recombinogenic effect of other DPC-promoting agents, formaldehye and ciprofloxacin, as well as the replication-inhibitor AZT. As with 5-azaC, we saw a modest reduction in recombinants induced by CPX by recO in the papillation assay; recombination induced by AZT appeared to be more strongly recO-dependent (Supplemental Figure 1).

Figure 5.

Figure 5.

Open in a new tab

Gene conversion recombination induced by 50 µg/ml 5-azaC, 2 hr treatment at indicated doses, in split cultures of wt or recO mutant strains. Bars represent mean values of the

3.3. Template-switching at quasipalindromes: effect of recombination functions.

We had previously documented a strong stimulatory effect of 5-azaC on mutagenesis detected by template-switching in a quasipalindromic sequence. In this study, we expanded this analysis to include additional mutants and doses of 5-azaC. We examined, in parallel, the effect of this set of homologous recombination factors on QPM using leading and lagging strand reporter strains, measuring lacZ reversion frequencies with and without treatment with 5-azaC at two doses. For each genetic background, an isogenic dcm derivative was also assayed. There are 8 potential Dcm methylation sites within lacZ. As we had seen previously (Laranjo et al. 2018; Klaric et al. 2020), we observed stimulation of QPM with increasing 5-azaC dose, with a stronger effect with the leading strand template-switch reporter QP5. (Fig. 6) Most, but possibly not quite all, of this induction is lost in dcm mutants. Mutants in recA showed no decrement in mutagenicity; if anything, they showed a slightly higher constitutive and 5-azaC-induced frequencies, at both doses, for both reporters. Mutants in recA had a strong effect of 5-azaC on the lagging strand reporter QP6 where it elevated QPM more than an order of magnitude at the higher dose. Again, as for wt, 5-azaC-induced QPM in recA was largely absent in dcm double mutants. Mutants in recB showed a slightly higher constitutive rate of QPM, but 5-azaC induction could not be tested at the higher dose because of poor survival. Mutants in recO showed a higher constitutive frequency of QPM in untreated cells but looked virtually indistinguishable from wt after 5-azaC treatments. These results suggest that homologous recombination and/or SOS induction promotes higher survival to 5- azaC, with RecA playing a antimutagenic role for QPM. This role may include homologous recombination that removes premutagenic ssDNA gaps, in addition to induction of the SOS response. Elevated constitutive QPM rates for recB and recO mutants would be consistent with the idea that both RecBCD and RecFOR pathways remove spontaneous, premutagenic lesions, with a stronger effect of the latter pathway.

Fig. 6.

Fig. 6.

Open in a new tab

Mutation frequencies with and without 5- azaC treatment in recombination mutants, genotype shown above bracket. Shown are dcm+ and dcm derivatives of wt, recA, recB and recO mutants. QP5 is a leading strand mutational reporter; QP6 a lagging strand reporter. The log scale of the the Y-axis is the same in the two graphs, to aid visual comparison. No treatment frequencies in brown, 1 µg/ml 5-azaC in light

3.4. Template-switching at QPs: effect of SOS response and translesion DNA polymerases:

The SOS response of E. coli consists of a number of functions coordinately induced by DNA damage or replication inhibition (Simmons et al. 2008). The genes for these functions are transcriptionally repressed by LexA, but which are induced by LexA self-cleavage promoted by the accumulation of RecA/ssDNA filaments. By measuring QPM in lexA3 strains, in which the SOS response cannot be induced, we had previously shown that the SOS response is anti-mutagenic for QPM, both constitutively and that induced by DPCs ((Laranjo et al. 2018) and Figure 8.) Mutants in lexA3 are also sensitive to killing by 5-azaC (Figure 2B). Several of the factors induced by the SOS response are the translesion DNA polymerases, Pol II, Pol IV and Pol V (encoded by polB, dinB and umuCD genes, respectively). To determine if these were responsible for the antimutagenic effect of the SOS response, we mutated all three and measured spontaneous QPM rates in a wild-type genetic background and in strains lacking one or both of the 3’ to 5’ exonucleases Exo I and Exo VII. These exonucleases abort QPM, with ExoI playing the larger role (Laranjo et al. 2017); hence QPM rates are much higher in their absence and potentially more sensitive to TLS polymerase effects. We saw no effect in any background of TLS polymerases on mutation frequency, for either the leading or lagging strand QPM reporters (Supplemental Figure 2). The loss of all three TLS polymerases also did not reduce survival to 5-azaC (Supplemental Figure 2B). Therefore, induction of TLS is not responsible for the antimutagenic aspect of the SOS response.

Fig. 8.

Fig. 8.

Open in a new tab

Mutation frequencies with and without 5-azaC treatment in lon and clpP mutants, genotype shown above bracket. Shown are dcm+ and dcm derivatives of lon and clpP mutants, compared to wt assayed in parallel. QP5 is a leading strand mutational reporter. No treatment frequencies in brown, 1 µg/ml 5-azaC in light green, 12.5 µg/ml in dark green. Black bar indicates the median value of

We examined mutants in which the SOS response was constitutively induced, lacking the LexA repressor. Because the cell division inhibitor SulA is induced during the SOS response, lexA∆ strains must carry an additional knockout of sulA to remain viable. With the QP5 reporter, lexA sulA null strains showed diminished 5-azaC-induced QPM relative to wt strains (Figure 7).

Fig. 7.

Fig. 7.

Open in a new tab

Mutation frequencies with and without 5-azaC treatment in lexA mutants, genotype shown above bracket. Shown are dcm+ and dcm derivatives of sulA and lexA∆ sulA compared to WT and lexA3 mutants assayed in parallel. QP5 is a leading strand mutational reporter. No treatment frequencies in brown, 1 µg/ml 5-azaC in light green, 12.5 µg/ml in dark green. Black bar

However, the sulA mutation by itself appeared to depress mutagenesis at the higher 5-azaC concentration (Figure 7). One explanation for this effect is that potential QP mutants require SulA-dependent cell division arrest to recover and form a colony. Using a lactose-papillation assay and the QP5 QPM reporter strain, we confirmed that sulA mutants were clearly less mutable at a wide range of 5-azaC concentrations. (Supplemental Figure 3). The 5-azaC-induced mutagenesis was entirely dependent on dcm (Fig. 7), showing that it is promoted by the DPCs formed by Dcm methyltransferase in the presence of 5-azaC.

The most striking effect of lexA∆, however, was a more than 10-fold elevation of constitutive rates of QPM (in the absence of 5-azaC) relative to wt and sulA strains (Fig. 7). This is the highest constitutive rate seen in our studies. This suggests that the SOS response also can promote QP mutagenesis in some instances.

We wondered whether the unnatural, full-on induction of TLS DNA polymerases was responsible for the high constitutive QPM in lexA∆ strains. We examined strains overexpressing one of the TLS DNA polymerases from an arabinose-regulated pBAD plasmid, with a copy number approximately 20 per cell, in an otherwise wt strain. In the absence of induction, we saw no effect of the TLS polymerases. After 2 hours of induction in arabinose medium, there was a strong mutagenic effect of DNA polymerase IV (DinB) for mutations detected by both QP5 and QP6, 34- and 13-fold, respectively. (Supplemental Figure 4). Pol IV is known to be prone to −1 G frameshift mutations (Kim et al. 1997) and in a lac reporter specific for this mutation, we saw a very large increase in mutation frequency by Pol IV when induced. Induction of Pol II caused about a 3-fold increase in QPM on the leading strand (QP5), and a 2-fold effect on the lagging strand (QP6). Pol V also was modestly mutagenic. This suggests that the TLS polymerases are mutagenic for QPM when their abundance is increased, particularly Pol IV, a potential explanation for the hypermutable constitutive phenotype of lexA∆ sulA strains.

3.5. Template-switching at QPs: effects of proteases:

Because DPCs may be removed by proteolysis (Vaz et al. 2017), we examined two major intracellular, energy-dependent proteases of E. coli for effects on 5-azaC induced QPM. Mutants in the Lon protease were found to be extremely sensitive to killing by 5-azaC (Figure 2C) and this sensitivity was abolished completely by a secondary mutation in the cell division inhibitor, sulA. During recovery from the SOS response, Lon protease destroys SulA, allowing cells to divide again. In the absence of Lon, SOS induction is lethal because cells never recover the ability to divide (Schoemaker et al. 1984). Complete suppression of lon by sulA (Figure 2C) suggests that the failure to turn off the SOS response to resume cell division is predominantly responsible for 5azaC-induced killing in this background. A secondary mutation in Dcm conferred only partial resistance to 5-azaC in lon mutants.

Therefore, some lethal SOS induction derives from Dcm DPCs formed with 5-azaC, but there must be a second SOS-inducing lesion that is Dcm-independent. Mutants in ClpP, a component of both ClpAP and ClpXP protease complexes (Gottesman 1996), did not show any sensitivity to 5-azaC (Figure 2C).

With QPM reporter QP5, mutants in lon showed hypermutability, 20-fold relative to wt, after treatment with the lower dose of 5-azaC (the higher dose could not be tested because of killing) (Figure 10). This elevated mutagenesis was entirely dependent on dcm (Figure 10). Mutants in clpP exhibited normal levels of Dcm-dependent QPM as assayed with reporter QP5.

3.6. Template-switch mediated deletions:

Deletions and expansion of tandem direct repeats occurs by a template-switch mechanism, independent of homologous recombination proteins including recA (reviewed in (Lovett 2017)). With 2 different assays for deletion formation (Lovett et al. 1994; Seier et al. 2011), we tested the effects of DPC-agents 5-azaC and formaldehyde. 5-azaC showed a dose-dependent increase in the deletion frequency of 11 bp tandem repeats within the chromosomal lac locus, with a 4-fold enhancement of deletion at the highest dose of 100 µg/ml (Figure 11a). Note that with a hundred-fold lower dose of 1 µg/ml 5-azaC, we saw a 5-fold enhancement of QPM lac mutagenesis using the QP5 reporter (Fig. 3A), so template-switching that gives rise to short deletions is not as strongly affected by 5-azaC as are QP-templated mutations. The 5-azaC effect was reduced to 2-fold in the dcm mutant (Fig. 11b), although the distributions considerably overlap. This indicates that although some of the deletionogenic effect of 5-azaC may occur through Dcm DPCs, some 5-azaC-induced deletions are independent of Dcm. Formaldehyde had a much stronger effect on deletion in this assay, with an enhancement of over 20-fold in the standard 2-hour treatment (Figure 11a). In a plasmid assay for deletion of 101 bp repeats within the tetA gene, conferring tetracycline resistance, we saw that 2 hr of 5-azaC treatment enhanced deletion frequencies 2-fold, which was reduced to 1.2 fold in the dcm mutant (Figure 11c). Similar treatment of wt type cells with formaldehyde enhanced deletion in this assay 3-fold and it was not appreciably affected by dcm, as expected, since the DPCs induced by FA are nonspecific.

Figure 3.

Figure 3.

Open in a new tab

Induction of the recA promoter as assayed by luciferase expression reporters. Red treated with 12.5 µg/ml 5-azaC; black untreated. 5-azaC was added at time 0 and the culture was followed for 2100 minutes post-treatment. RLU, relative luciferase units is the luminescence divided by the OD600 of the culture.

4. DISCUSSION

We used 5-azacytidine to investigate the genetic consequences of DPCs formed by cytosine methytransferase, Dcm, in E. coli. Using a variety of genetic assays, we measured 5-azaC effects on the DNA damage response, recombination and genetic mutations produced by template switching, at inverted or direct DNA repeats.

4.1. The DNA damage response:

By transcriptional reporter assays for the recA promoter, we show that 5-azacytidine induces the SOS response, a transcriptional response triggered by the formation of RecA filaments on single-strand DNA in vivo. This induction was dependent on Dcm, implicating the Dcm protein-DNA crosslink as the inducing lesion. We saw a modest reduction and delay in a RecO mutant, implicating the RecFOR pathway for SOS induction; however, there was detectable 5-azaC induction independent of RecO that was dependent on RecB.

4.2. Homologous recombination:

A 2-hour treatment with 5-azaC induced homologous recombination about 30-fold, as detected by a gene conversion assay between an internally deleted lacZ gene and a 500 bp homologous lacZ fragment integrated across the chromosome, at attTn7. This increase in recombination was also seen in a dcm mutant, suggesting that lesions other than Dcm-DNA crosslinks can initiate recombination. The identity of these lesions is unknown; we did not find any evidence for recombination induced by guanylurea, a potential 5-azaC byproduct (Towne, unpublished data). There was a small, 2-fold, reduction in 5-azaC-induced recombination in a recO mutant. A plasmid-based assay for crossover recombination, showed a more modest 2–3 fold stimulation by a 2-hr treatment with 5-azaC, which was reduced by recO and dcm. Therefore, we found evidence for homologous recombination induced by 5-azaC, some clearly associated with DPCs and some that may not be.

4.3. Mutability at QP sites/ effects of recombination:

In the absence of RecA, 5-azaC mutagenicity was enhanced at quasipalindromic (QP) sites, dependent on the Dcm methyltransferase. The elevation by recA was particularly striking for the QP6 reporter for lagging-strand template-switching, where loss of recA elevated mutation frequencies, relative to wt, 8- and 68-fold at 1 and 12.5 µg/ml 5-azaC, respectively. This elevation of lagging strand QPM was not due to lack of the SOS response, since lexA3 mutant strains, likewise unable to mount the SOS response, showed no or 3-fold elevation of QP6 rates at 1 and 12.5 µg/ml 5-azaC, respectively, relative to wt type strains. Therefore, the capacity for homologous recombination, negated by recA but not by lexA3, helps avoid QPM by 5-azaC but apparently only, or predominantly, on the lagging strand. We did not see this elevation of QPM in recB mutants, defective in DSB-associated recombination, so this protective recombination is likely to be that associated with single-strand DNA gaps. Gap-filling by recombination should decrease the opportunity for the strand misalignments responsible for template-switching. Other sources of DPCs, including fluoroquinolones and formaldehydes, has previously been shown to induce QPM (Laranjo et al. 2018; Klaric et al. 2020) as do other replication-blocking agents such as AZT (Seier et al. 2012), UV irradiation, mitomycin C and nitrofurazone (Klaric, unpublished results). A commonality among these agents is their ability to arrest DNA replication.

4.4. Mutability at QP sites/ effects of the SOS response and TLS DNA polymerases:

The SOS response does have an antimutagenic effect on 5-azaC -induced leading strand QPM, with 4–6- fold enhancement by lexA3 and 8-fold by recA, relative to wt. Our prior work with the natural QP mutational hotspot in the thyA gene had indicated that the TLS polymerases, under SOS control, had an antimutagenic effect (Dutra and Lovett 2006), a potential explanation for the antimutagenic nature of the SOS response. However, in this work, using QP sites within the lacZ gene, we can find no evidence for such an effect. We saw no effect on mutation rates by introduction of polB dinB umuC knockouts, even when ExoI and ExoVII were absent. Rather, these SOS-induced polymerases, when overexpressed, appeared to be mutagenic for QPM, which is consistent with our finding that lexA∆ strains show an elevated spontaneous level of QPM.

Surprisingly, a sulA mutation, which prevents cell division arrest after DNA damage, reduced the mutagenicity of 5-azaC, as detected on the leading strand QP5 reporter, from 3- to 5-fold. A dramatic reduction in mutability by chronic exposure of cells to 5-azaC was also seen in sulA mutants using lactose X-gal papillation assays. We know of no precedent for sulA effects on mutagenesis, with the exception of a very modest (<2x) effect on stress-induced mutation (Al Mamun et al. 2012). Except for dcm mutants, this is the only strain to exhibit lower frequencies of mutation than wt after 5-azaC treatment. Because we think it unlikely that SulA would play a direct role in potentiating mutagenesis by 5-azaC, we can explain the lower yield of mutants in the sulA strain by proposing that potential mutants fail to survive in the absence of cell division arrest and so are lost in this assay. The loss of sulA did not, however, influence survival of the bulk cell population to 5-azaC, so the vulnerable cells in this case must be a subpopulation.

4.5. Mutability at QP sites/ effects of proteases:

The Lon protease plays a potential role in the avoidance of 5-azaC induced mutations at quasipalindromes; lon mutants were hypermutable by over 50-fold at the lower 5-azaC dose. We documented this elevation using the leading strand QPM reporter, but saw a similar enhancement with the lagging strand reporter (Klaric, unpublished results). One hypothesis is that the persistence of a DPC in the absence of Lon may promote template-switching, whereas prompt removal of the DPC by Lon proteolysis leads to a nonmutagenic outcome. In the bacterium Caulobacter crescentus, Lon has been reported to play a role in degradation of DNA-bound proteins (Zeinert et al. 2018), dependent on Lon’s ability to bind DNA, a property that is shared between C. crescentus and E. coli Lon proteins (Schoemaker and Markovitz 1981; Zehnbauer et al. 1981). E. coli mutants in lon were extremely sensitive to killing by 5-azaC, but this appears to be entirely due to the failure to recover after induction of the SOS response; this sensitivity was suppressed completely by an additional mutation in sulA. Unfortunately, loss of sulA itself depressed QPM and therefore we were unable to test whether the lon hypermutability would be apparent when survival of lon to 5-azaC was enhanced by sulA. Eukaryotic mitochondria employ a Lon ortholog to maintain proteostasis (Bezawork-Geleta et al. 2015) that also binds DNA (Liu et al. 2004) and a role in protection from DPCs would be interesting to investigate. We saw no effect by ClpP. Further analysis will be required to determine which bacterial proteases may aid the removal of DPCs.

4.6. 5-azaC: not just DPCs?

Our data suggest that although Dcm-dependent DPCs largely contribute to genetic effects of 5-azaC, there is some Dcm-independent lesion(s) as well. This is particularly apparent in the survival curves of strains sensitive to killing by 5-azaC (recA, recB, and lon mutants), where dcm improves survival but does not restore it to wildtype levels. In addition, homologous recombination assayed by gene conversion between lacZ alleles is induced strongly by 5-azaC but is largely dcm-independent; 5-azaC-induced deletion of 11 bp in lacZ is partially dcm-independent. Whatever the Dcm-independent lesion is, it must induce the SOS response since sulA entirely suppresses 5-azaC killing of lon mutants, whereas dcm suppresses it only partially. Dcm-independent killing of recA and recB mutants by 5-azaC suggests that the lesion must also lead to DSBs in DNA. 5-azaC can break down into guanylurea (Figure 1), which in DNA mispairs with cytosine, leading to C to G transversion mutations (Jackson-Grusby et al. 1997). In mammalian cells, this C to G transversion correlates with sites of C-methylation, leading to speculation that the CMeT induces the cytosine ring opening that converts it to guanylurea (Jackson-Grusby et al. 1997). However, more recent work suggest that guanylurea deoxribonucleotides can be directly incorporated into DNA during replication (Lamparska et al. 2012). Using a specific reporter for C to G transversion mutations (Cupples and Miller 1989; Seier et al. 2011), we confirmed that 5-azaC does indeed have a strong, 500-fold mutagenic effect on C to G transversion (confirming (Cupples and Miller 1989)), and occurs entirely independent of dcm (Table S1 and (Doiron et al. 1999)). Therefore, guaunylurea lesions may indeed be a potential contributor to dcm-independent effects of 5-azaC, where it may act like a abasic lesion to block DNA synthesis or elicit processing that leads to mutagenesis, template switching or recombination. Assays of guanylurea on induction of gene conversion using our Lac reversion assays, however, were negative (unpublished results). It is also possible that there are additional unknown CMeTs encoded in the E. coli genome that are responsible for the dcm-independent effects.

4.7. Differential effects on template-switching:

Although a template-switch mechanism has been proposed for both QPM at inverted repeats and deletions/expansion at direct repeats (reviewed in (Lovett 2017), we saw a much stronger effect on the former than the latter. We assayed two types of deletions that arise between short tandem repeated sequences, 11 bp in the chromosomal lacZ gene, and 101 bp in the tetA gene on a ColE1 plasmid and both were stimulated modestly, 2–4 fold after 2-hr treatment with 5-azaC. In contrast, comparable treatment yielded stimulation of QPM over 2 orders of magnitude. We do not know the basis for this differential effect but further work may clarify the differences between these template-switching events. Deletion formation detected with these assays was more strongly affected by formaldehyde than by 5-azaC, which may be explained by the ability of formaldehyde to stimulate DNA/DNA crosslinks, in addition to DPCs (Feldman 1973).

Supplementary Material

1

Supplemental Figure 1. Gene conversion recombination induced by AZT or DPCs. Lac X-Gal papillation plates are spread with the lacZ recombination assay strain, either wt, recA or recO mutants. Spotted onto sterile filter disks were 10 µl of water (left) or aqueous solutions of AZT (300 ng/ml), 5-azaC (100 µg/ml), formaldehyde (100 mM) or ciprofloxacin (CPX, 100 µg/ml) (right). A blue ring surrounding the filter disk indicates a recombinogenic effect of the drug. The blue papillae not localized to the disks are spontaneous recombinants arising from the bacterial lawn on the plate.

Supplemental Figure 2. QPM frequencies in strains lacking the TLS DNA polymerases (polB dinB umuC) with and without the DNA exonucleases Exo I and/or Exo VII. Bars indicate the median value of the replicates.

Supplemental Figure 3. QPM induced by 5-azaC in wt and sulA mutants. Lac X-Gal papillation plates are spread with lacZ QP5 assay strains, either wt (left panels) or a sulA mutant (right panels). Spotted onto sterile filter disks were 10 µl of water (“-“, left disks) or aqueous solutions of 5-azaC at the indicated concentrations (“+”, right disks). A blue ring surrounding the filter disk indicates a mutagenic effect of the drug. The blue papillae not localized to the disks are spontaneous mutants arising from the bacterial lawn on the plate.

Supplemental Figure 4. Induced expression of TLS DNA polymerase and effects on mutation frequencies with the indicated lacZ mutational reporter strains. Bars indicate the median value of the replicates.

Table S1. Frequency of C to G transversions with and without 5-azaC, in wild-type and dcm mutants, n= 4

Fig. 9.

Fig. 9.

Open in a new tab

Tandem repeat deletion induced by 5-azaC and FA. Cultures were split and treated for 2 hr with 5-azaC or FA. The mutation frequency in the treated culture relative to that in the untreated is plotted: a value of 1 (indicated by dashed line in A and C) means there is no effect. Bars indicate mean values. A. Deletion of 11 bp repeats in chromosomal lacZ. B. Effect of dcm on 11 bp lacZ deletion. C. Deletion of 101 bp repeats on plasmid-encoded tetA and effects of dcm.

Highlights:

  • 5-azacytidine is broadly mutagenic and recombinogenic

  • In E. coli, 5-azaC promotes genetic instability through Dcm methyltransferase.

  • There are other unknown recombinogenic and mutagenic lesions induced by 5-azaC besides Dcm/DNA crosslinks

  • 5-azaC induces the SOS response, which protects cells from killing and genetic instability

5. Acknowledgments:

We thank Hirotada Mori and the National Genetics Institute of Japan for the collection of E. coli knockout mutants. Funding: This work was supported by the National Institutes of Health, grants R01 GM51753 and P01 GM105473, T32 GM007122 and National Science Foundation REU grant DBI-1359172.

Abbreviations:

DPCs

DNA protein crosslinks

CMeT

cytosine methyltransferase

5-azaC

5- azacytidine

DSBs

double strand breaks

ssDNA

single-strand DNA

QP

quasipalindrome

QPM

quasipalindrome-associated mutation

Bibliography

  1. Al Mamun AA, Lombardo MJ, Shee C, Lisewski AM, Gonzalez C et al. , 2012. Identity and function of a large gene network underlying mutagenic repair of DNA breaks. Science 338: 1344–1348. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Anderson DG, and Kowalczykowski SC, 1997. The recombination hot spot chi is a regulatory element that switches the polarity of DNA degradation by the RecBCD enzyme. Genes & development 11: 571–581. [DOI] [PubMed] [Google Scholar]
  3. Bannister D, and Glover SW, 1968. Restriction and modification of bacteriophages by R+ strains of Escherichia coli K12. Biochem Biophys Res Commun 30: 735–738. [DOI] [PubMed] [Google Scholar]
  4. Barker S, Weinfeld M and Murray D, 2005. DNA-protein crosslinks: their induction, repair, and biological consequences. Mutat Res 589: 111–135. [DOI] [PubMed] [Google Scholar]
  5. Bezawork-Geleta A, Brodie EJ, Dougan DA and Truscott KN, 2015. LON is the master protease that protects against protein aggregation in human mitochondria through direct degradation of misfolded proteins. Sci Rep 5: 17397. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bhagwat AS, Johnson B, Weule K and Roberts RJ, 1990. Primary sequence of the EcoRII endonuclease and properties of its fusions with beta-galactosidase. J Biol Chem 265: 767–773. [PubMed] [Google Scholar]
  7. Bhagwat AS, and Roberts RJ, 1987. Genetic analysis of the 5-azacytidine sensitivity of Escherichia coli K-12. Journal of Bacteriology 169: 1537–1546. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bhagwat AS, Sohail A and Roberts RJ, 1986. Cloning and characterization of the dcm locus of Escherichia coli K-12. J Bacteriol 166: 751–755. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bissler JJ, 1998. DNA inverted repeats and human disease. Front Biosci 3: d408–418. [DOI] [PubMed] [Google Scholar]
  10. Butala M, Busby SJW and Lee DJ, 2009. DNA sampling: a method for probing protein binding at specific loci on bacterial chromosomes. Nucleic Acids Res 37: e37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Cox MM, Goodman MF, Kreuzer KN, Sherratt DJ, Sandler SJ et al. , 2000. The importance of repairing stalled replication forks. Nature 404: 37–41. [DOI] [PubMed] [Google Scholar]
  12. Craft TR, Bermudez E and Skopek TR, 1987. Formaldehyde mutagenesis and formation of DNA-protein crosslinks in human lymphoblasts in vitro. Mutat Res 176: 147–155. [DOI] [PubMed] [Google Scholar]
  13. Cupples CG, and Miller JH, 1989. A set of lacZ mutations in Escherichia coli that allow rapid detection of each of the six base substitutions. Proc Natl Acad Sci U S A 86: 5345–5349. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Datsenko KA, and Wanner BL, 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc Natl Acad Sci U S A 97: 6640–6645. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Doiron KM, Lavigne-Nicolas J and Cupples CG, 1999. Effect of interaction between 5- azacytidine and DNA (cytosine-5) methyltransferase on C-to-G and C-to-T mutations in Escherichia coli. Mutat Res 429: 37–44. [DOI] [PubMed] [Google Scholar]
  16. Dower W, Miller J and Ragsdale C, 1988. High efficiency transformation of E. Coli by high voltage electroporation. Nucleic Acids Research 16: 6127–6145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Drlica K, and Zhao X, 1997. DNA gyrase, topoisomerase IV, and the 4-quinolones. Microbiol Mol Biol Rev 61: 377–392. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Dutra BE, and Lovett ST, 2006. Cis and trans-acting effects on a mutational hotspot involving a replication template switch. J Mol Biol 356: 300–311. [DOI] [PubMed] [Google Scholar]
  19. Feldman MY, 1973. Reactions of nucleic acids and nucleoproteins with formaldehyde. Prog Nucleic Acid Res Mol Biol 13: 1–49. [DOI] [PubMed] [Google Scholar]
  20. Friedman S, 1985. The irreversible binding of azacytosine-containing DNA fragments to bacterial DNA(cytosine-5)methyltransferases. J Biol Chem 260: 5698–5705. [PubMed] [Google Scholar]
  21. Garaycoechea JI, Crossan GP, Langevin F, Daly M, Arends MJ et al. , 2012. Genotoxic consequences of endogenous aldehydes on mouse haematopoietic stem cell function. Nature 489: 571–575. [DOI] [PubMed] [Google Scholar]
  22. Glickman BW, and Ripley LS, 1984. Structural intermediates of deletion mutagenesis: a role for palindromic DNA. Proc Natl Acad Sci U S A 81: 512–516. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Goldfless SJ, Morag AS, Belisle KA, Sutera VA and Lovett ST, 2006. DNA repeat rearrangements mediated by DnaK-dependent replication fork repair. Molecular Cell 21: 595–604. [DOI] [PubMed] [Google Scholar]
  24. Gottesman S, 1996. Proteases and their targets in Escherichia coli. Annu. Rev. Genet 30: 465–506. [DOI] [PubMed] [Google Scholar]
  25. Greenblatt MS, Grollman AP and Harris CC, 1996. Deletions and insertions in the p53 tumor suppressor gene in human cancers: confirmation of the DNA polymerase slippage/misalignment model. Cancer Res 56: 2130–2136. [PubMed] [Google Scholar]
  26. Guzman L, Belin D, Carson M and Beckwith J, 1995. Tight regulation, modulation, and high-level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177: 4121–4130. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Hampsey DM, Ernst JF, Stewart JW and Sherman F, 1988. Multiple base-pair mutations in yeast. J Mol Biol 201: 471–486. [DOI] [PubMed] [Google Scholar]
  28. Ide H, Shoulkamy MI, Nakano T, Miyamoto-Matsubara M and Salem AM, 2011. Repair and biochemical effects of DNA-protein crosslinks. Mutat Res 711: 113–122. [DOI] [PubMed] [Google Scholar]
  29. Izzotti A, Cartiglia C, Taningher M, De Flora S and Balansky R, 1999. Age-related increases of 8-hydroxy-2’-deoxyguanosine and DNA-protein crosslinks in mouse organs. Mutat Res 446: 215–223. [DOI] [PubMed] [Google Scholar]
  30. Jackson-Grusby L, Laird PW, Magge SN, Moeller BJ and Jaenisch R, 1997. Mutagenicity of 5-aza-2’-deoxycytidine is mediated by the mammalian DNA methyltransferase. Proc Natl Acad Sci U S A 94: 4681–4685. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Keller KL, Overbeck-Carrick TL and Beck DJ, 2001. Survival and induction of SOS in Escherichia coli treated with cisplatin, UV-irradiation, or mitomycin C are dependent on the function of the RecBC and RecFOR pathways of homologous recombination. Mutat Res 486: 21–29. [DOI] [PubMed] [Google Scholar]
  32. Kim SR, Maenhaut-Michel G, Yamada M, Yamamoto Y, Matsui K et al. , 1997. Multiple pathways for SOS-induced mutagenesis in Escherichia coli: an overexpression of dinB/dinP results in strongly enhancing mutagenesis in the absence of any exogenous treatment to damage DNA. Proc Natl Acad Sci U S A 94: 13792–13797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Klaric JA, Perr EL and Lovett ST, 2020. Identifying Small Molecules That Promote Quasipalindrome-Associated Template-Switch Mutations in Escherichia coli. G3 (Bethesda). [DOI] [PMC free article] [PubMed]
  34. Krasich R, Wu SY, Kuo HK and Kreuzer KN, 2015. Functions that protect Escherichia coli from DNA-protein crosslinks. DNA Repair (Amst) 28: 48–59. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kuo HK, Griffith JD and Kreuzer KN, 2007. 5-Azacytidine induced methyltransferase- DNA adducts block DNA replication in vivo. Cancer research 67: 8248–8254. [DOI] [PubMed] [Google Scholar]
  36. Lal D, Som S and Friedman S, 1988. Survival and mutagenic effects of 5-azacytidine in Escherichia coli. Mutat Res 193: 229–236. [DOI] [PubMed] [Google Scholar]
  37. Lamparska K, Clark J, Babilonia G, Bedell V, Yip W et al. , 2012. 2’-Deoxyriboguanylurea, the primary breakdown product of 5-aza-2’-deoxyribocytidine, is a mutagen, an epimutagen, an inhibitor of DNA methyltransferases and an inducer of 5-azacytidine-type fragile sites. Nucleic Acids Res 40: 9788–9801. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Laranjo LT, Gross SJ, Zeiger DM and Lovett ST, 2017. SSB recruitment of Exonuclease I aborts template-switching in Escherichia coli. DNA Repair (Amst) 57: 12–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Laranjo LT, Klaric JA, Pearlman LR and Lovett ST, 2018. Stimulation of Replication Template-Switching by DNA-Protein Crosslinks. Genes (Basel) 10. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Liu T, Lu B, Lee I, Ondrovicova G, Kutejova E et al. , 2004. DNA and RNA binding by the mitochondrial lon protease is regulated by nucleotide and protein substrate. J Biol Chem 279: 13902–13910. [DOI] [PubMed] [Google Scholar]
  41. Lovett ST (Editor), 2010. The DNA damage response. American Society of Microbiology Press, Washington, DC. [Google Scholar]
  42. Lovett ST, 2017. Template-switching during replication fork repair in bacteria. DNA Repair (Amst) 56: 118–128. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Lovett ST, Drapkin PT, Sutera VA Jr. and Gluckman-Peskind TJ, 1993. A sister-strand exchange mechanism for recA-independent deletion of repeated DNA sequences in Escherichia coli. Genetics 135: 631–642. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Lovett ST, and Feschenko VV, 1996. Stabilization of diverged tandem repeats by mismatch repair: evidence for deletion formation via a misaligned replication intermediate. Proc Natl Acad Sci U S A 93: 7120–7124. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lovett ST, Gluckman TJ, Simon PJ, Sutera VA Jr. and Drapkin PT, 1994. Recombination between repeats in Escherichia coli by a recA-independent, proximity-sensitive mechanism. Mol Gen Genet 245: 294–300. [DOI] [PubMed] [Google Scholar]
  46. Lovett ST, Hurley RL, Sutera VA Jr., Aubuchon RH and Lebedeva MA, 2002. Crossing over between regions of limited homology in Escherichia coli. RecA-dependent and RecA- independent pathways. Genetics 160: 851–859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. McKenzie GJ, and Craig NL, 2006. Fast, easy and efficient: site-specific insertion of transgenes into enterobacterial chromosomes using Tn7 without need for selection of the insertion event. BMC Microbiol 6: 39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Nakano T, Xu X, Salem AMH, Shoulkamy MI and Ide H, 2017. Radiation-induced DNA-protein cross-links: Mechanisms and biological significance. Free Radic Biol Med 107: 136–145. [DOI] [PubMed] [Google Scholar]
  49. Nitiss JL, 2009. Targeting DNA topoisomerase II in cancer chemotherapy. Nat Rev Cancer 9: 338–350. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Persky NS, and Lovett ST, 2008. Mechanisms of recombination: lessons from E. coli. Crit Rev Biochem Mol Biol 43: 347–370. [DOI] [PubMed] [Google Scholar]
  51. Pommier Y, Leo E, Zhang H and Marchand C, 2010. DNA topoisomerases and their poisoning by anticancer and antibacterial drugs. Chem Biol 17: 421–433. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Ringquist S, and Smith CL, 1992. The Escherichia coli chromosome contains specific, unmethylated dam and dcm sites. Proc Natl Acad Sci U S A 89: 4539–4543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Ripley L, 1982. Model for the participation of quasi-palindromic DNA sequences in frameshift mutation. Proceedings of the National Academy of Sciences of the United States of America 79: 4128–4132. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Roberts RJ, Vincze T, Posfai J and Macelis D, 2015. REBASE--a database for DNA restriction and modification: enzymes, genes and genomes. Nucleic Acids Res 43: D298–299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Salem AM, Nakano T, Takuwa M, Matoba N, Tsuboi T et al. , 2009. Genetic analysis of repair and damage tolerance mechanisms for DNA-protein cross-links in Escherichia coli. J Bacteriol 191: 5657–5668. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Santi DV, Norment A and Garrett CE, 1984. Covalent bond formation between a DNA-cytosine methyltransferase and DNA containing 5-azacytosine. Proc Natl Acad Sci U S A 81: 6993–6997. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Sassanfar M, and Roberts JW, 1990. Nature of the SOS-inducing signal in Escherichia coli. The involvement of DNA replication. J Mol Biol 212: 79–96. [DOI] [PubMed] [Google Scholar]
  58. Schoemaker JM, Gayda RC and Markovitz A, 1984. Regulation of cell division in Escherichia coli: SOS induction and cellular location of the sulA protein, a key to lon-associated filamentation and death. J Bacteriol 158: 551–561. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Schoemaker JM, and Markovitz A, 1981. Identification of the gene lon (capR) product as a 94-kilodalton polypeptide by cloning and deletion analysis. J Bacteriol 147: 46–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Seier T, Padgett DR, Zilberberg G, Sutera VA Jr., Toha N et al. , 2011. Insights into mutagenesis using Escherichia coli chromosomal lacZ strains that enable detection of a wide spectrum of mutational events. Genetics 188: 247–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Seier T, Zilberberg G, Zeiger DM and Lovett ST, 2012. Azidothymidine and other chain terminators are mutagenic for template-switch-generated genetic mutations. Proc Natl Acad Sci U S A 109: 6171–6174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Sharan SK, Thomason LC, Kuznetsov SG and Court DL, 2009. Recombineering: a homologous recombination-based method of genetic engineering. Nat Protoc 4: 206–223. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Simmons LA, Foti JJ, Cohen SE and Walker GC, 2008. The SOS Regulatory Network. EcoSal Plus 3. [DOI] [PubMed] [Google Scholar]
  64. Swenberg JA, Lu K, Moeller BC, Gao L, Upton PB et al. , 2011. Endogenous versus exogenous DNA adducts: their role in carcinogenesis, epidemiology, and risk assessment. Toxicol Sci 120 Suppl 1: S130–145. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Thoms B, and Wackernagel W, 1987. Regulatory role of recF in the SOS response of Escherichia coli: impaired induction of SOS genes by UV irradiation and nalidixic acid in a recF mutant. J Bacteriol 169: 1731–1736. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Vaz B, Popovic M and Ramadan K, 2017. DNA-Protein Crosslink Proteolysis Repair. Trends Biochem Sci 42: 483–495. [DOI] [PubMed] [Google Scholar]
  67. Viswanathan M, Lacirignola JJ, Hurley RL and Lovett ST, 2000. A novel mutational hotspot in a natural quasipalindrome in Escherichia coli. J Mol Biol 302: 553–564. [DOI] [PubMed] [Google Scholar]
  68. Whitby MC, and Lloyd RG, 1995. Altered SOS induction associated with mutations in recF, recO and recR. Mol Gen Genet 246: 174–179. [DOI] [PubMed] [Google Scholar]
  69. Wu FY, Lee YJ, Chen DR and Kuo HW, 2002. Association of DNA-protein crosslinks and breast cancer. Mutat Res 501: 69–78. [DOI] [PubMed] [Google Scholar]
  70. Yoshimori R, Roulland-Dussoix D and Boyer HW, 1972. R factor-controlled restriction and modification of deoxyribonucleic acid: restriction mutants. J Bacteriol 112: 1275–1279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. Yoshiyama K, Higuchi K, Matsumura H and Maki H, 2001. Directionality of DNA replication fork movement strongly affects the generation of spontaneous mutations in Escherichia coli. Journal of Molecular Biology 307: 1195–1206. [DOI] [PubMed] [Google Scholar]
  72. Yoshiyama K, and Maki H, 2003. Spontaneous hotspot mutations resistant to mismatch correction in Escherichia coli: transcription-dependent mutagenesis involving template-switching mechanisms. J Mol Biol 327: 7–18. [DOI] [PubMed] [Google Scholar]
  73. Zahn RK, Zahn-Daimler G, Ax S, Hosokawa M and Takeda T, 1999. Assessment of DNA-protein crosslinks in the course of aging in two mouse strains by use of a modified alkaline filter elution applied to whole tissue samples. Mech Ageing Dev 108: 99–112. [DOI] [PubMed] [Google Scholar]
  74. Zehnbauer BA, Foley EC, Henderson GW and Markowitz A, 1981. Identification and purification of the Lon+ (capR+) gene product, a DNA-binding protein. Proc. Natl. Acad. Sci. USA 78: 2043–2047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Zeinert RD, Liu J, Yang Q, Du Y, Haynes CM et al. , 2018. A legacy role for DNA binding of Lon protects against genotoxic stress. bioRxiv: 10.1101/317677. [DOI]

Associated Data

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

Supplementary Materials

1

Supplemental Figure 1. Gene conversion recombination induced by AZT or DPCs. Lac X-Gal papillation plates are spread with the lacZ recombination assay strain, either wt, recA or recO mutants. Spotted onto sterile filter disks were 10 µl of water (left) or aqueous solutions of AZT (300 ng/ml), 5-azaC (100 µg/ml), formaldehyde (100 mM) or ciprofloxacin (CPX, 100 µg/ml) (right). A blue ring surrounding the filter disk indicates a recombinogenic effect of the drug. The blue papillae not localized to the disks are spontaneous recombinants arising from the bacterial lawn on the plate.

Supplemental Figure 2. QPM frequencies in strains lacking the TLS DNA polymerases (polB dinB umuC) with and without the DNA exonucleases Exo I and/or Exo VII. Bars indicate the median value of the replicates.

Supplemental Figure 3. QPM induced by 5-azaC in wt and sulA mutants. Lac X-Gal papillation plates are spread with lacZ QP5 assay strains, either wt (left panels) or a sulA mutant (right panels). Spotted onto sterile filter disks were 10 µl of water (“-“, left disks) or aqueous solutions of 5-azaC at the indicated concentrations (“+”, right disks). A blue ring surrounding the filter disk indicates a mutagenic effect of the drug. The blue papillae not localized to the disks are spontaneous mutants arising from the bacterial lawn on the plate.

Supplemental Figure 4. Induced expression of TLS DNA polymerase and effects on mutation frequencies with the indicated lacZ mutational reporter strains. Bars indicate the median value of the replicates.

Table S1. Frequency of C to G transversions with and without 5-azaC, in wild-type and dcm mutants, n= 4

NIHMS1713395-supplement-1.pdf (528.7KB, pdf)

RESOURCES