Role of Bacillus subtilis DNA Glycosylase MutM in Counteracting Oxidatively Induced DNA Damage and in Stationary-Phase-Associated Mutagenesis

The bacterial strains and plasmids used in this study are listed in Table 1. B. subtilis YB955 is a prophage-“cured” strain that contains the hisC952, metB5, and leuC427 auxotrophic mutations (15–17). B. subtilis strains were maintained on tryptic blood agar base (TBAB) (Acumedia Manufacturers, Inc., Lansing, MI). Liquid cultures of B. subtilis strains were grown in Penassay broth (PAB) (antibiotic A3 medium; Difco Laboratories, Sparks, MD). E. coli cultures were grown in Luria-Bertani (LB) medium. When required, neomycin (Neo; 10 μg ml⁻¹), tetracycline (Tet; 10 μg ml⁻¹), spectinomycin (Sp; 100 μg ml⁻¹), kanamycin (Kan; 10 μg ml⁻¹), ampicillin (Amp; 100 μg ml⁻¹), chloramphenicol (Cm; 5 μg ml⁻¹), erythromycin (Ery; 1 μg ml⁻¹), rifampin (Rif; 10 μg ml⁻¹), or isopropyl-β-d-thiogalactopyranoside (IPTG; 1 mM) was added to media. Hydrogen peroxide (H₂O₂) and 1,1′-dimethyl-4,4′-bipyridinium dichloride (Paraquat [PQ]) were obtained from Sigma-Aldrich (St. Louis, MO).

To obtain a mutM mutant strain in the genetic background YB955, chromosomal DNA of strain B. subtilis PERM599 (PS832 ΔmutM::tet) was isolated and used to transform competent cells of strain B. subtilis YB955, generating the strain B. subtilis PERM571 (YB955 ΔmutM::tet).

For complementation of the mutM mutation, a copy of this gene was placed ectopically at the amyE locus under the control of the IPTG-inducible hyperspank promoter (Phs). To this end, the open reading frame of mutM was amplified by PCR using genomic DNA of B. subtilis YB955 as a template and Vent DNA polymerase (New England BioLabs, Beverly, MA). Oligonucleotide primers 5′-AAGTCGACGAGATAGGAAGTGATG-3′ (forward) and 5′-ATGCATGCGGGAAAGTGAAAAATC-3′ (reverse) containing SalI and SphI sites (underlined), respectively, were used in the PCR. The PCR product was ligated into the pCR-Blunt-II-TOPO vector (Invitrogen Life Technologies, Carlsbad, CA), generating plasmid pPERM735 (Table 1). The mutM gene was released from this plasmid and inserted into the integrative vector pdr-111-amyE-Phyperspank (a gift from David Rudner). The resulting construct, pPERM792, was introduced by transformation into the B. subtilis strains PERM571 and PERM573 to generate the strains B. subtilis PERM1199 and PERM794.

To obtain a mutM mutY double mutant, the strain PERM571 (YB955 ΔmutM::tet) was transformed with genomic DNA of PERM704 (YB955 mutY::sp) (18), generating the strain B. subtilis PERM572 (YB955 ΔmutM::tet mutY::sp) (Table 1). The double homologous-recombination event resulting in inactivation of the gene of interest was confirmed by PCR with specific oligonucleotide primers (data not shown).

Construction of a transcriptional fusion between mutM and the lacZ gene was performed in the integrative plasmid pMUTIN4 (19). To this end, a 405-bp fragment, extending from 264 bp upstream to 141 bp downstream of the mutM open reading frame (ORF) start codon was amplified using Vent DNA polymerase (New England BioLabs, Beverly, MA) and oligonucleotide primers 5′-CGCGAATTCCGATTCAAGGAAGCGCCG-3′ (forward) and GCCGGATCCTCGCGCAAATTCCTCCGG-3′ (reverse) with EcoRI/BamHI restriction sites (underlined), respectively. The PCR product was ligated into pCR-Blunt-II-TOPO (Invitrogen Life Technologies, Carlsbad, CA), generating plasmid PERM617. The EcoRI/BamHI fragment was ligated into pMUTIN4, which had previously been digested with the same restriction enzymes. The resulting construct containing the mutM-lacZ fusion was designated pPERM657 and was introduced by transformation into competent cells of strain B. subtilis YB955 to generate strain B. subtilis PERM659 (Table 1).

An in-frame translational fusion between mutM and the FLAG epitope was constructed in the vector pMUTIN-FLAG (20). To this end, a DNA fragment from 15 bp upstream (including the Shine-Dalgarno sequence) of the translational start codon to the last codon of the mutM ORF was amplified by PCR by utilizing Vent DNA polymerase (New England BioLabs) and the oligonucleotide primers 5′-GGAAGCTTCAGAGATAGGAAGTCATGGAT-3′ (forward) and 5′-GGGGTACCGTTTTTTGTCTGGCACTTTCG-3′ (reverse), which inserted HindIII and KpnI sites (underlined) into the cloned DNA. The PCR-amplified DNA fragment (849 bp) was first ligated into pCR-Blunt-II-TOPO (Invitrogen, Carlsbad, CA) and then replicated in E. coli XL10-GOLD Kan^r (Stratagene, Cedar Creek, TX). The resulting construct (PERM698) was treated with HindIII and KpnI, the 849-bp mutM insert was ligated into HindIII/KpnI-treated pMUTIN-FLAG, and the ligation products were introduced by transformation into competent cells of E. coli XL10-GOLD Kan^r (Stratagene, Cedar Creek, TX). This strategy generated plasmid pPERM748, which was used to transform B. subtilis YB955, generating strain B. subtilis PERM796 (Table 1). The crossover events leading to insertion of the mutM-lacZ and mutM-FLAG fusions into the corresponding loci were confirmed by PCR with specific oligonucleotide primers (data not shown).

B. subtilis strain PERM659, containing a transcriptional mutM-lacZ fusion, was propagated in liquid A3 medium. Aliquots of 1 ml were collected from cultures at exponential growth phase, stationary phase, or sporulation. Cells were washed with 0.1 M Tris-HCl (pH 7.5), pelleted by centrifugation, and stored at 20°C until determination of β-galactosidase activity (21). Briefly, washed-cell samples were first disrupted with lysozyme and subjected to centrifugation; the β-galactosidase activity present in the supernatant was then determined as previously described, using ortho-nitrophenyl-β-d-galactopyranoside (ONPG) as a substrate.

Total RNA from exponentially growing or stationary-phase B. subtilis YB955 cells grown in A3 medium was isolated by using TRI Reagent (Molecular Research Center, Inc., Cincinnati, OH). Reverse transcription-PCRs (RT-PCRs) were performed with the RNA samples and a Master AMP RT-PCR kit (Epicentre Technologies, Madison, WI) according to the manufacturer's instructions. The primers used for RT-PCR were 5′-GGTCTTCATCAAGCCGAGGAG-3′ (forward) and 5′-ATTTCAGCGTGAAGGGTTTTG-3′ (reverse), which generated a 377-bp RT-PCR product extending from 249 bp downstream of the start codon of mutM to 626 bp downstream of this point. As a control, in each experiment, the absence of chromosomal DNA in the RNA samples was assessed by PCRs carried out with Vent DNA polymerase (New England BioLabs) and the set of primers described above. The size of the RT-PCR product was determined by utilizing the 1-kb-Plus DNA ladder (Life Technologies, Rockville, MD) during agarose gel electrophoresis.

B. subtilis strain YB955 was cultivated with shaking in liquid antibiotic A3 medium at 37°C. Aliquots of 1.5 ml were collected from cultures during the exponential, transition, or stationary phase of growth. Cells were collected by centrifugation (16,000 × g; 1 min), washed twice with 25 mM Tris-HCl (pH 7.5) buffer, and stored at 20°C. Bacterial pellets were resuspended in 0.3 ml of the same buffer supplemented with a protease inhibitor cocktail (Roche, Mannheim, Germany) and were disrupted by sonication with a VCX130-PB Vibra-Cell apparatus (Sonics and Materials Inc., Newton, CT). The cell lysate was subjected to centrifugation to eliminate undisrupted cells and cell debris. The supernatant was separated, and its protein concentration was determined with a Coomassie (Bradford) protein assay kit (Pierce, Rockford, IL). Protein aliquots (100 μg) were separated in SDS-12% polyacrylamide gels and then electrotransferred to polyvinylidene difluoride (PVDF) membranes. Western blot analyses were performed with a FLAG monoclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA) diluted 10,000-fold and then processed with an enhanced-chemiluminescence (ECL) Western blotting system (Amersham Pharmacia, Little Chalfont, Buckinghamshire, United Kingdom).

Essentially, cultures were grown in flasks containing antibiotic A3 medium with aeration (250 rpm) at 37°C until 90 min after the cessation of exponential growth (designated T₉₀ [90 min after the time point in the culture when the slopes of the logarithmic and stationary phases of growth intercepted]). Growth was monitored with a spectrophotometer measuring the optical density at 600 nm (OD₆₀₀). The stationary-phase mutagenesis assays were performed as previously described (22, 23) using solid Spizizen minimal medium (SMM; 1× Spizizen salts supplemented with 0.5% glucose and either 50 μg or 200 ng of the required amino acid per ml and 50 μg each of isoleucine and glutamic acid per ml). The concentration of the amino acid used depended on the reversion tested. For instance, in selecting His⁺ revertants, 50 μg (each) of methionine and leucine ml⁻¹ was added to the medium and 200 ng of histidine ml⁻¹ was added. Isoleucine and glutamic acid were added as described previously (16) in order to protect the viability of the cells. The number of revertants was scored daily. The initial number of bacteria plated for each experiment was estimated by serial dilution of the bacterial cultures and then plating of the cells on LB medium. The experiments were repeated at least three times.

Frequencies of spontaneous mutation to rifampin resistance in growing cells were determined as previously described (13). Essentially, the appropriate strains were grown for 12 h at 37°C in antibiotic A3 medium with proper antibiotics. Mutation frequencies were determined by plating aliquots on six LB plates containing 10 μg ml⁻¹ rifampin, and the rifampin-resistant (Rif^r) colonies were counted after 1 day of incubation at 37°C. The number of cells used to calculate the frequency of mutation to Rif^r was determined by plating aliquots of appropriate dilutions on LB plates without rifampin and incubating the plates for 24 to 48 h at 37°C. These experiments were repeated at least three times (24).

B. subtilis strains YB955, PERM571, and PERM1199 were grown in LB medium with aeration (250 rpm) at 37°C. Growth was monitored with a spectrophotometer measuring the optical density at 600 nm (OD₆₀₀). Before cessation of exponential growth (OD₆₀₀, ∼0.6), the cells were collected by centrifugation (6,500 × g, 5 min), washed twice with phosphate-buffered saline (PBS; pH 7.2), and resuspended in the same buffer. Cell aliquots of equal volumes were treated with different final concentrations of H₂O₂ or PQ and incubated for 30 min at 37°C with shaking. The total viable-cell numbers in each culture were determined by spotting serial dilutions of the cultures on LB agar plates. The number of colonies was counted after 24 h of incubation at 37°C.

For determination of mutation frequencies and oxidative-stress sensitivity, differences were calculated by performing one-way analysis of variance (ANOVA) followed by a Tukey's post hoc analysis. Significance was set at a P of <0.05.

ROS-promoted DNA lesions, including 8-oxo-G, may potentially generate mutagenesis and cell death (25, 26). The 8-oxo-G lesion is processed through the BER pathway with the specific participation of MutM, which eliminates this oxidized base from DNA (5, 27, 28). Thus, we analyzed whether MutM conferred protection to growing B. subtilis cells from the cytotoxic effects of oxidative stress. To this end, growing cells of a mutM knockout strain and its MutM-proficient parental strain were treated with increasing amounts of H₂O₂ or PQ. Results showed that disruption of mutM significantly sensitized exponentially growing cells of B. subtilis to these oxidizing agents (Fig. 1A and B). This result reveals a role for MutM in conferring to B. subtilis protection from the lethal effects of H₂O₂ and PQ. In support of this notion, the susceptibility to H₂O₂ and PQ of the mutM strain was reestablished to the level of the parental strain, YB955, following expression of the wild-type mutM gene from the IPTG-inducible Phs promoter (Fig. 1A and B). Taken together, these results strongly suggest that MutM plays a significant role in preventing the cytotoxic effects of 8-oxo-G and possibly of other related lesions, including the opened ring derivative formamidopyrimidine (FaPy) (29–32), thus contributing to B. subtilis survival. However, in addition to inducing the formation of oxidized bases, H₂O₂ and PQ may promote other types of DNA lesions, including 8-OxoG·A mispairs and apurinic/apyrimidinic (AP) sites, as well as single- and/or double-strand DNA breaks (33). Therefore, in addition to MutM, other repair proteins, including MutY, Nth, and the AP-endonucleases Nfo and ExoA, most probably contribute to protecting B. subtilis from the genotoxic effects of H₂O₂ (33).

Of note, the absence of MutM also decreased the H₂O₂ resistance of E. coli cells, and such effect was associated with an increased amount of 8-oxo-G lesions in the genome of this microorganism (34). However, the protective role conferred by MutM against H₂O₂-promoted DNA damage has also been described to occur in other bacteria, including Pseudomonas aeruginosa and Mycobacterium smegmatis. Thus, cells of these strains lacking MutM were significantly more susceptible to H₂O₂ treatment than their MutM-proficient counterparts (35, 36).

Due to an anticipated role of the MutM protein in preventing the mutagenic and cytotoxic effects of 8-oxo-G (13, 37), the mutation frequency to a Rif^r phenotype was determined in growing cells of the MutM-deficient and parental YB955 strains. The results revealed that the loss of MutM increased the spontaneous mutation frequency to Rif^r 5-fold in comparison with that of an isogenic strain that produced a functional MutM protein (Fig. 1C). From these results, we propose that MutM prevents the spontaneous mutagenic events promoted by oxidative stress in growing B. subtilis cells. Two observations support this contention: the levels of mutation to Rif^r calculated in the MutM-deficient strain were restored to the levels of the parental strain following expression of mutM from the IPTG-inducible Phs promoter (Fig. 1C), and the overexpression of mutM induced a significant decrease in the frequency of mutation to Rif^r of a hypermutagenic strain that was deficient for MutM, MutY, and MutT (Fig. 1D).

In agreement with a previous report (37), our results revealed that cells of the B. subtilis strain YB955 lacking MutM showed a slight but statistically significant increase in their frequency of spontaneous mutation to Rif^r relative to that of the MutM-proficient parental strain. It must be pointed out that MutM-deficient strains of E. coli and Pseudomonas putida also presented a mutagenic Rif^r phenotype; however, in these bacteria, as well as in B. subtilis (13), the single MutY deficiency conferred a stronger mutagenic effect than that observed in the strains lacking MutM (5, 36, 38, 39). These results suggest the existence of alternative repair pathways that compensate for the absence of MutM; in agreement with this notion, the genomes of the three microorganisms discussed above contain the gene for Nth, a DNA glycosylase capable of processing 8-oxo-G and AP sites (40–42). In the case of B. subtilis, it was recently shown that the genetic inactivation of Nth not only increases this bacterium's spontaneous Rif^r mutation frequency but also sensitizes it to the ROS promoter agent H₂O₂ (33).

We next investigated the role played by MutM in the stationary-phase-associated mutagenesis (SPM) of B. subtilis. These experiments were performed in strain B. subtilis YB955, which is auxotrophic for three amino acids due to the chromosomal mutations hisC952 (amber), metB5 (ochre), and leuC427 (missense). This strain has been validated and widely used as a model system to understand how mutations are generated in amino-acid-starved cells (18, 23, 43). Analysis of frequencies of reversion to his, met, and leu in cell cultures that were starved for each of these amino acids revealed that MutM contributes to mutagenesis in starved B. subtilis cells. As shown in Fig. 2, the MutM-deficient strain significantly increased the frequency of his, met, and leu reversions in reference to those generated by parental strain YB955. These results strongly suggest that unrepaired 8-oxo-G lesions that accumulate in the MutM-deficient strain promote stationary-phase-associated mutagenesis in B. subtilis. In a marked contrast with our results, the single absence of MutM did not promote mutagenesis in starved cells of E. coli and P. putida (11, 12). However, in E. coli and P. putida, the lack of MutY did induce a significant increase in the production of stationary-phase-associated mutations (11, 44), suggesting that accumulation of nonprocessed 8-oxo-G lesions contribute to stationary-phase-associated mutagenesis in these strains. In support of this notion, when the mutM mutation was combined with a deficiency in MutY, the mutation frequency was further enhanced in starved E. coli cells (11). Thus, despite the fact that the lack of MutY also contributes to SPM in B. subtilis (18), our results clearly indicate that the sole disruption of mutM also favored this type of mutagenesis in this microorganism. In support of this contention, an ectopic copy of mutM expressed from the IPTG-inducible Phs promoter diminished the numbers of His⁺, Met⁺, and Leu⁺ revertants relative to those produced by parental strain YB955 (Fig. 2). Moreover, we corroborated the finding that the genetic defect in mutM did not affect the survival of B. subtilis cells starved for his, met, and leu during the 10 days that the SPM experiments lasted (see Fig. S1 in the supplemental material).

As described above, MutM confers protection to growing B. subtilis cells from the toxic effects of H₂O₂, and its deficiency promotes adaptive mutagenesis in nutritionally stressed cells. These results suggest that mutM may be expressed in the exponential and stationary phases of growth of this microorganism. To explore this notion, we analyzed the temporal pattern of expression of mutM and determined the levels of its encoded product during the life cycle of B. subtilis. The levels of transcription were determined by employing B. subtilis strain PERM659, which harbors a genomic copy of a transcriptional mutM-lacZ fusion (Table 1). The results showed that this strain expressed barely similar levels of β-galactosidase during the exponential transition (from exponential to stationary phase) and the first hours of stationary phases of growth (Fig. 3A). However, the expression levels of the reporter lacZ gene commenced to diminish during the late stationary phase of growth. Results from an RT-PCR experiment performed with RNA samples collected during exponential growth as well as during the transition and stationary phases of growth confirmed the presence of mutM mRNAs during the three developmental phases analyzed (Fig. 3B). In agreement with this result, we also detected a MutM-FLAG protein in actively growing cells of a B. subtilis strain and in cells in the stationary phase of growth harboring an in-frame translational mutM-FLAG fusion (Fig. 3C). Based on these and previous results (45, 46), it is feasible to propose that B. subtilis expresses mutM during its entire life cycle to contend with the genotoxic and cytotoxic effects of ROS. However, despite the role displayed by mutM in protecting B. subtilis from oxidatively induced DNA damage, we did not find evidence that this gene is part of the gene circuitries that respond to distinct types of stressful conditions, including DNA damage and oxidative or osmotic stress (47–53). This conclusion was deduced from experiments showing that H₂O₂ (0.1%), mitomycin C (0.5 μg ml⁻¹), and NaCl (4%) did not turn on the transcription of a mutM-lacZ fusion inserted into the genome of strain B. subtilis YB955 (Fig. S2). Therefore, in conjunction with previous reports (47, 54–56), it is feasible to conclude that expression of mutM is not under the control of the master regulator RecA/DinR, PerR, or σ^B.

In contrast, in E. coli, mutM is under the negative transcriptional control of the Fur, Fnr, and ArcA regulators; thus, the mRNA levels of mutM are enhanced in this bacterium by ROS-producing chemicals, including Paraquat. It is noteworthy that the levels of expression of mutY are repressed under the stressful conditions that activate mutM (14, 57, 58).

Thus, the ability of B. subtilis to keep active the synthesis of MutM during the logarithmic and stationary phases is in agreement with our results that demonstrated antimutagenic roles of this repair protein in both stages of growth (Fig. 1 and 2). Moreover, the presence of MutM in the stationary phase of B. subtilis but its apparent absence in E. coli (57) may explain why the single disruption of mutM did not promote mutagenesis in starved cells of E. coli unless combined with a mutation in MutY (11). Alternatively, the existence in E. coli of repair proteins that process 8-oxo-G lesions, including Nth and Nei (11, 59), may suppress mutagenesis in starved E. coli cells deficient for MutM.

Our analysis of his, met, and leu reversions in nutritionally stressed B. subtilis cells showed that deficiencies in mutM significantly increased the mutagenesis levels in the three alleles tested. However, the mechanisms involved in generating such reversions may be different; thus, for the his and met alleles, ROS-promoted synthesis of 8-oxo-G may be responsible for these reversions. In support of this contention, genetic inactivation of mutY in the MutM-deficient strain dramatically increased the production of His and Met revertants in the resulting mutM mutY mutant (Fig. 4). In contrast, the levels of reversion of the leu allele in the mutM mutY strain were reduced compared to those observed in the mutM and parental YB955 strains (Fig. 4C). This result suggest that MutY promotes reversions in the leuC allele; in support of this notion, the levels of Leu⁺ revertants were almost completely ablated in the MutY-deficient strain (Fig. 4C). Furthermore, a previous study demonstrated that processing of accumulated G·A mismatches in starved B. subtilis by MutY is involved in generating stationary-phase-associated Leu⁺ revertants (18).

As shown in this and previous reports (13, 18), DNA repair proteins that process ROS-induced DNA damage play prominent roles in modulating mutagenesis in starved bacterial cells. Nevertheless, current reports have shown that in B. subtilis, this type of mutation is also dependent on Mfd, a protein that couples transcription with the DNA repair machinery (43, 60). It was recently found that production of Leu⁺ prototrophs in MutY-deficient B. subtilis cells of strain YB955 are fully dependent on a functional Mfd protein (M. Gómez-Marroquín, E. A. Robleto, and M. Pedraza-Reyes, unpublished results). Therefore, we are currently investigating how Mfd coordinates the activities of repair proteins of the GO system to generate mutations that occur in nutritionally stressed B. subtilis cells.