Elucidating the functional role of Mycobacterium smegmatis recX in stress response

The RecX protein has attracted considerable interest because the recX mutants exhibit multiple phenotypes associated with RecA functions. To further our understanding of the functional relationship between recA and recX, the effect of different stress treatments on their expression profiles, cell yield and viability were investigated. A significant correlation was found between the expression of Mycobacterium smegmatis recA and recX genes at different stages of growth, and in response to different stress treatments albeit recX exhibiting lower transcript and protein abundance at the mid-log and stationary phases of the bacterial growth cycle. To ascertain their roles in vivo, a targeted deletion of the recX and recArecX was performed in M. smegmatis. The growth kinetics of these mutant strains and their sensitivity patterns to different stress treatments were assessed relative to the wild-type strain. The deletion of recA affected normal cell growth and survival, while recX deletion showed no significant effect. Interestingly, deletion of both recX and recA genes results in a phenotype that is intermediate between the phenotypes of the ΔrecA mutant and the wild-type strain. Collectively, these results reveal a previously unrecognized role for M. smegmatis recX and support the notion that it may regulate a subset of the yet unknown genes involved in normal cell growth and DNA-damage repair.

Bacterial strains, plasmids and growth conditions. Bacterial strains and plasmids used in this study are described in Table 1. The M. smegmatis mc 2 155 ΔrecX and ΔrecAΔrecX strains were generated as described 28 . The pJV53 vector (encoding mycophage derived recombinases under acetamide-inducible promoter was transformed into M. smegmatis mc 2 155 28 . The ΔrecAΔrecX and ΔrecX mutants were derived from M. smegmatis mc 2 155 strain. These strains were grown in liquid 7H9 Middlebrook broth (Difco Laboratories, Detroit, MI) supplemented with 0.05% Tween 80 and 10% albumin-dextrose-catalase (ADC) (henceforth referred to as 7H9 broth) with continuous agitation (120 rpm) at 37 °C for 2 days. M. smegmatis mc 2 155 was grown on 7H10 agar (4.7 g/L) plates supplemented with 10% ADC. The starter cultures were prepared by inoculating a single colony of M. smegmatis mc 2 155 into 15 ml of 7H9 broth and grown at 37 °C in a shaking incubator (120 rpm), until the cultures reached an OD 600 of 1. Then 1.5 ml of the cell suspension was inoculated into 30 ml of 7H9 Middlebrook medium to give an OD 600 of 0.05. The antibiotics (kanamycin and hygromycin) were added to a final concentration of 25 µg/ml and 100 µg/ml respectively.
The M. smegmatis ΔrecA mutant was a kind gift from Elaine O. Davis, National Institute for Medical Research, London. The characteristics of plasmids (pJV53, pVV16 and p0004S) have been described elsewhere 29 . The bacterial cultures were exposed to a UV lamp (UVP Inc. San Gabriel, CA) at 254 nm for different time intervals ranging from 5 to 20 sec. The dosage was measured using a radiometer (UVItec Ltd, Cambridge, UK).The E. coli DH5α strain (Novagen, Madison, WI) was used in the recombinant DNA experiments. The cultures were incubated at 37 °C with shaking at 120 rpm, and OD 600 values were recorded every 3 h, until the cultures reached an OD 600 of 2.5.
Cloning of recA and recX genes. M. smegmatis recA (MSMEG_2723) and recX (MSMEG_2724) ORFs were obtained from the SmegmaList website (http://svitsrv8.epfl.ch/mycobrowser/smegmalist.html). The coding sequences of recA and recX were amplified by PCR using gene-specific forward and reverse primers ( Table 2). The pVV16-vector was digested with NdeI and HindIII. The PCR products were gel purified, and ligated separately into the linear plasmid pVV16 DNA at www.nature.com/scientificreports www.nature.com/scientificreports/ NdeI/HindIII site. The resulting plasmids contained either full-length M. smegmatis recA or recX genes under the control of the Hsp60 constitutive promoter.
Complementation assay. The complementation assay was performed using M. smegmatis mc 2 155 ΔrecA, ΔrecAΔrecX and ΔrecX mutant strains. The plasmid complemented strains bearing wild-type copies of M. smegmatis recA or recX ORFs were created by electroporation. These were grown in liquid 7H9 broth to mid-log phase, until cultures reached an OD 600 of 0.6. The cultures were centrifuged and pellets were resuspended to an OD 600 of 0.8 in sterile Milli-Q water. An aliquot from each culture was serially diluted and spotted on 7H10 agar plate containing 100 µg/ml of hygromycin B. The cells were exposed to 10 J/m 2 of 254 nm UV light and incubated at 37 °C in the dark for 3 days. The images were acquired in the digitization mode using a Quant LAS 4000 imaging system (GE Healthcare, Piscataway, NJ).
Construction of recX and recA-recX gene replacement mutants in M. smegmatis. The DNA manipulations were carried out using standard protocols 30 . Plasmid DNA and DNA fragments were isolated from the gels using Qiagen miniprep kit (Qiagen India, New Delhi) according to the manufacturer's instructions. The genomic DNA of M. smegmatis was PCR-amplified as described 31 . The electrocompetent M. smegmatis mc 2 155 cells were prepared and stored as described 32 . The construction of allelic exchange substrate (AES) and its subsequent delivery to the recombineering strain of M. smegmatis mc 2 155 was carried out as described 28 . Table 3 shows the sequences of upstream and downstream primers corresponding to the M. smegmatis recA-recX locus. For recX gene deletion, the upstream and downstream flanking regions of M. smegmatis mc 2 155 recX were PCR-amplified from the genomic DNA using specific primers that contained the restriction sites in the counter-selectable vector p0004S. The upstream and downstream flank fragments were digested with PflMI and directionally cloned on either side of the Hyg resistance gene to generate the recX-AES. To generate allelic exchange substrates, the upstream and downstream flanking sequences of recA-recX locus were PCR-amplified using genomic DNA as the template and specific primers. The amplified products were digested with PflMI and directionally cloned on either side of the Hyg resistance gene. The resulting constructs were independently transformed into E. coli DH5α and plated on LB agar containing 100 μg/ml of hygromycin. The positive clones were identified by restriction analysis of the isolated plasmid using EcoRV to generate the linear blunt-ended recombineering-proficient (containing upstream region-hyg r -downstream region) DNA fragments. preparation of M. smegmatis mc 2 155 recombineering strain and construction of ΔrecX and ΔrecAΔrecX strains. The M. smegmatis mc 2 155 derivative harbouring the pJV53 vector was used in recombineering 28 . The AES constructs generated for disruption of recX and recArecX were linearised using EcoRV and then isolated by electroelution. The linearized AES constructs (100 ng each), were independently electroporated into the M. smegmatis mc 2 155 recombineering strain. The transformed cells were cultured with shaking at 37 °C for 4 h. The transformants were plated on 7H10 agar plates, supplemented with 100 μg/ml hygromycin B and 25 µg/ml kanamycin. The transformants were selected and their genomic DNA was isolated and analyzed by PCR to ascertain recombination in the M. smegmatis mc 2 155 genome at the allelic site.

Analysis of M. smegmatis recA and recX gene expression under hypoxic conditions. Wayne and
Hayes dormancy culture system was used to study the hypoxic response of M. smegmatis recA and recX expression as described 33,34 . Briefly, M. smegmatis mc 2 155 strain was grown as described above until the culture reached an OD 600 of 0.6. An aliquot of the culture was diluted in Middlebrook 7H9 broth to an OD 600 of 0.05. Equal volumes (5 ml) of diluted culture were placed in each of 100 ml screw-cap flasks having a headspace ratio of ~ 0.5 (35 ml of head space air vol. to 65 ml of liquid medium) containing a stir bar. The cultures were gently shaken (120 rpm) at 37 °C. To ascertain hypoxic culture conditions, we used a standard color assay (viz., methylene blue reduction) 35 . Methylene blue (1.5 μg/ml) was added to cultures, and reduction was monitored visually (by decolorization). To avoid aeration of cultures during sampling, separate flasks were used for each time point in these experiments. A small portion of the culture from each flask was used for Western blot analysis, while the rest was used for total RNA extraction and RT-qPCR as described below.
DNA damage-sensitivity assays. The M. smegmatis mc 2 155 wild-type, ΔrecA, ΔrecX and ΔrecAΔrecX strains were grown in 7H9 broth to mid-log phase to a cell density of 0.6. After 3 h of incubation with different concentrations of ciprofloxacin, MMS or hydrogen peroxide, the cells were collected by centrifugation and resuspended in 1 ml sterile Milli-Q water at a cell density of 0. 8 Table 3. The sequences of oligonucleotides used for the construction of AES. (RP and FP stand for reverse and forward primers respectively). Note: the sequences shown in bold are the recognition sites of PflMI and those underlined are for EcoRV. Primer pairs 1, 2 and 3, 4 were used for the amplification of upstream region ofrecA and recX genes, respectively. Primers 5 and 6 were used to amplify downstream region of recX gene.
www.nature.com/scientificreports www.nature.com/scientificreports/ The cDNA synthesis was performed using the following parameters: 25 °C for 10 min, 42 °C for 30 min followed by termination at 85 °C for 5 min. The RT-qPCRs (in 20 µl) were carried out with 2 ng of cDNA template, 10 µl of SYBR green (2X), 0.5 µl of both forward and reverse primers (0.25 µM each) in the BioRad iQ5 multicolor real time PCR detection system. The cycling program comprised initial denaturation at 95 °C for 10 min, denaturation at 95 °C for 15 sec, primer annealing at 60 °C for 30 sec and extension at 72 °C for 30 sec. Following amplification, a melting curve analysis was performed to verify the authenticity of the amplified product. Similarly, the levels of MSMEG_2725, MSMEG_2726 and MSMEG_2720 ORFs in M. smegmatis mc 2 155 wild-type, ΔrecA ΔrecX and ΔrecX mutant cells were determined by RT-qPCR.
Western blot analysis. The whole cell lysates of M. smegmatis mc 2 155 cells, grown under aerobic or hypoxic conditions, were prepared as described above (see under RNA isolation and RT-qPCR). The antibody against GroEL was purchased from Sigma-Aldrich (St. Louis, MO). The polyclonal antibodies against M. tuberculosis RecA and RecX were raised in rabbits and characterized as described 14 . The whole cell lysate (30 µg of protein) was boiled in SDS/PAGE buffer for 10 min, and the proteins were separated by SDS-PAGE using 10% (wt/vol) polyacrylamide gels as described 36 . After gel electrophoresis, the proteins were transferred onto a polyvinylidene difluoride membrane. The membranes were incubated with either anti-RecA, anti-RecX or anti-GroEL antibodies at a dilution of 1:15000, 1:1000 and 1:80000 respectively as described 37 . The membranes were blocked with 3% BSA in phosphate buffered saline (137 mM NaCl, 2.7 mM KCl, 10 mM Na 2 HPO 4 and 1.8 mM KH 2 PO 4 ) and washed with TBST (50 mM Tris-HCl, pH 8, 150 mM NaCl and 0.1% Tween 20). Subsequently, the membranes were incubated with peroxidase-conjugated (Sigma-Aldrich, St. Louis, MO) secondary antibody. The blots were developed using luminol and hydrogen peroxide. The digital images of Western blot bands were acquired using the Quant LAS-4000 chemidocumentation system (GE Healthcare Life Sciences, Pittsburgh, PA), and quantified using the ImageJ software. statistics and data presentation. Statistical significance was determined using the unpaired t-test. P-value (p) less than 0.05 indicate significant difference and p > 0.05 is not significant (ns). The symbols *, ** and *** correspond to p < 0.05, p < 0.01 and p < 0.001 respectively.

Results and Discussion
Hypoxia causes an increase in the expression of M. smegmatis recA and recX genes. During the course of infection and proliferation, M. tuberculosis is exposed to physiological stress conditions such as hypoxia and acidic pH stress, which could compromise its survival [38][39][40] . For instance, acute hypoxic stress arrests DNA replication and triggers robust DNA damage [41][42][43] . An increase in the expression of genes of the SOS pathway, including umuDC, polB, recN, sulA, uvrA, uvrB, and uvrD, begin upon DNA damage [44][45][46] . To this end, hypoxia-induced gene expression profiling in M. smegmatis and M. tuberculosis during latent infection has provided valuable insights on the nature of latent tubercle bacilli and its treatment 47 .
Previous studies in M. tuberculosis 48 , Thiobacillus ferrooxidans 49 and Streptomyces lividans 50 have demonstrated that recA is co-transcribed with recX. Furthermore, an overexpression of RecA (a positive regulator of SOS response) was found to be toxic in the absence of RecX in certain bacterial species including Pseudomonas aeruginosa 51 , S. lividans 50 , M. smegmatis 52 , and Xanthomonas oryzae 53 . These findings are consistent with the notion that RecX acts as an anti-recombinase to quell inappropriate recombination events promoted by RecA 14 .
In contrast, an overexpression of RecA in the E. coli ΔrecX strain is not deleterious and the deletion of recX does not affect recA expression 54   www.nature.com/scientificreports www.nature.com/scientificreports/ normal growth conditions because of the existence of a transcription termination site between the recA and recX coding sequences. However, a recA-recX mRNA transcript resulting from transcriptional read-through accumulates in the range of 5-10% of the total recA mRNA content 54 . The recX transcript was hardly detectable during the vegetative growth of E. coli, but robust expression of recX occurs following treatment with DNA damaging agents 15,55 .
The genomes of both M. smegmatis and M. tuberculosis contain a single copy each of recA and recX in a single operon. To investigate the expression and regulation of these genes under aerobic and hypoxic growth in M. smegmatis, a commonly used surrogate model for M. tuberculosis, the relative abundance of mRNA at various growth phases was measured by a RT-qPCR assay and normalized to that of the constitutively expressed 16S rRNA gene. The cycle threshold (Ct) values for M. smegmatis recA and recX mRNA were normalized against the Ct value of the16S rRNA gene. In contrast to E. coli 15,55 , significant amount of recA mRNA transcripts were seen at early-log phase under both aerobic and hypoxic culture conditions (Fig. 1A). Interestingly, hypoxic conditions enhanced the accumulation of recA mRNA by 1.5-fold at mid-log phase, and decreased thereafter as the cells entered the stationary phase. A similar pattern was observed with recX mRNA abundance under both aerobic and hypoxic conditions, albeit at reduced levels (Fig. 1B). These observations support the idea that the two genes are co-transcribed and coordinately regulated under aerobic and hypoxic growth conditions. Intriguingly, no difference was seen in the amounts of mRNA under aerobic and hypoxic growth conditions at the early-log phase. The transcriptional profiles of M. smegmatis growth-phase-specific marker genes (sigH2 and sigH7) 56 was used to ascertain the times of the early-log, the mid-log and the stationary phases (Fig. 1C). In multi-drug resistant clinical strains of M. tuberculosis, the levels of both recA and recX mRNA were found to be higher than in the drug-susceptible strains and the levels of recX mRNA increased concomitantly with a rise in recA mRNA 57 .
However, it should be noted that the mRNA levels cannot be used as surrogates for the corresponding protein levels. Therefore, the abundance of RecA and RecX proteins in M. smegmatis was measured under aerobic and hypoxic conditions by a Western blot assay using anti-RecA and anti-RecX antibodies. GroEL was probed as a protein-loading control. The densitometric analysis of immunoblots revealed that cells accumulated higher levels of RecA under hypoxic growth conditions compared to aerobic conditions ( Fig. 2A,B). A comparative analysis showed that the cells consistently accumulated 2-fold higher levels of RecA at mid-log phases relative to the early-log phase under aerobic growth conditions, which decreased at the stationary phase to levels seen at the early-log phase ( Fig. 2A,B). A similar pattern was observed for RecX, albeit the levels were less pronounced than that of RecA ( Fig. 2A,C). Although the mRNA and protein expression patterns were comparable, important differences were noted. First, the RecX protein was hardly detectable at early-log phase during hypoxic conditions. Second, the abundance of RecA mRNA and protein were higher compared to RecX.
Although, these results suggest that M. smegmatis recA and recX genes are induced under hypoxic conditions, there is no correlation between the amount of recX mRNA and corresponding protein in the early-log phase. Both M. tuberculosis and M. smegmatis have been shown to stabilize their mRNA transcripts under growth-inhibiting conditions [58][59][60] . Among possible mechanisms, tRNA reprogramming and selective codon-biased translation have been shown to play a role in mycobacteria in response to hypoxia 61 . One possible explanation for the lack of correlation between the recX mRNA versus protein level in the early log phase could be due translational repression of recX mRNA.

Construction of ΔrecX and
ΔrecA ΔrecX mutant strains. The difference in the RecA and RecX protein levels in M. smegmatis indicate possible regulation of gene expression at the level of transcription. In many bacteria studied so far, the recX gene is located on the same coding strand downstream of recA and the two genes are co-transcribed 48,50,53,54,62,63 . However, in the lexA-recA-recX locus of Xanthomonas pathovars, each gene is expressed from its own promoter 64 . Furthermore, in D. radiodurans 65,66 , B. subtilis 67 and N. gonorrhoeae 26 , the recA and recX genes are separated by several hundred kb in length and transcribed from their own promoters.
In a number of mycobacterial species, as well as in a few other bacteria, the 5′ region of the recX coding sequence overlaps the 3′ region of the recA gene. The overlap is 32 bp long in M. smegmatis 48  , the S. lividans recX mutants displayed decreased resistance to UV damage 42 and the N. gonorrhoeae recX mutant showed a small decrease in its ability to survive DNA damage that was caused by double-strand breaks 26 . However, the impact of deletion of both recA and recX genes on the growth characteristics and DNA damage repair has not been investigated in any organism. Moreover, a full understanding of the in vivo role of recX in mycobacteria is not fully understood.
Previous studies have demonstrated that the M. smegmatis ΔrecA strain exhibited sensitivity to UV-induced DNA damage and fails to promote HR 52 . We combined the recA mutation with the deletion of recX to investigate the effects of the double mutations. For this purpose, M. smegmatis mc 2 155 recX single and recArecX double mutant strains were generated. As a control, the effect of recA deletion in M. smegmatis was re-evaluated for direct comparison. Using the recombineering method, which is based on a protocol developed for M. tuberculosis and M. bovis BCG 28 , the M. smegmatis mc 2 155 ΔrecA and ΔrecAΔrecX mutants were generated using 3.279 kb and 3.302 kb linear ΔrecA::hyg and ΔrecAΔrecX::hyg AES constructs, respectively. Approximately 100 ng of linear AES DNA fragments were generated by restriction digestion of the respective plasmids with EcoRV (Fig. 3A). These were transformed into competent M. smegmatis mc 2 155:pJV53 recombineering cells 70  www.nature.com/scientificreports www.nature.com/scientificreports/ mutant strains were screened using PCR to determine if the recX and recArecX genes were deleted from the chromosome (data not shown). The correct mutants were grown in 7H10 Middlebrook agar medium for 5-8 generations to allow the loss of pJV53.

Genotypic and phenotypic analysis of M. smegmatis ΔrecX and ΔrecAΔrecX mutants.
After screening by PCR, 2 each of the ΔrecX and ΔrecAΔrecX knockout mutants of M. smegmatis mc 2 155 were obtained. The ΔrecX and ΔrecAΔrecX mutants were characterized by restriction enzyme mapping and Southern blot hybridization (Fig. 3). Upon hybridization with appropriate radiolabeled probes, a 4.1 kb fragment was seen in the case of wild-type M. smegmatis mc 2 155 cells. The predicted 3.14 kb and 2.09 kb fragments were observed in the ΔrecX and ΔrecAΔrecX mutants respectively (Fig. 3B,C). Both these bands are smaller than 4.1 kb, indicating that the recX and recArecX genes have been deleted by allelic replacement. The frequency of allelic exchange with respect to recX and recArecX was in the range of 70% and 80% respectively.
One caveat of this analysis is that the observed effects of ΔrecX and ΔrecAΔrecX mutations could result from the polar effects of hygromycin-resistance gene insertion. In general, genetic alterations around the recA-recX locus could lead to polar effects on the expression of genes downstream of recA-recX locus. Two independent experiments were carried out to investigate the potential polar effects. First, the ΔrecX and ΔrecAΔrecX mutants were complemented with the functional copies of the recA and recX genes. The transformants were evaluated for their ability to grow in a standard culture medium and protect mutant cells against UV irradiation. Ten-fold serial dilutions were spotted on 7H10 agar plates and analyzed. As shown in Fig. 4A,B, the wild-type recA partially complemented the ΔrecA and ΔrecAΔrecX mutant strains, as deduced from their ability to support growth and provide protection against UV irradiation. However, wild-type recX failed to rescue the phenotype of ΔrecAΔrecX mutant strains observed upon DNA damage induction with UV irradiation. Since, deletion of recX www.nature.com/scientificreports www.nature.com/scientificreports/ had no discernible effect on its growth under normal and DNA damaging conditions, ΔrecX and its corresponding complemented strain showed similar growth as that of wild-type.
Second, we assessed the expression of two M. smegmatis mc 2 155 genes, gluD (MSMEG_2725) and glnH (MSMEG_2726), located downstream of the recA recX locus in the ΔrecA ΔrecX and ΔrecX mutant strains in comparison with the wild-type strain. The M. smegmatis hyd2 (MSMEG_2720) gene located upstream of the recA-recX locus was used as a positive control. The total RNA was prepared from the wild-type, ΔrecA ΔrecX and ΔrecX mutant strains. The relative abundances of the gene transcripts were determined by quantitative RT-qPCR and normalized to that of the constitutively expressed chromosomal 16S rRNA gene. The expression levels were measured from cells grown in the late exponential growth phase. In all cases, the gene expression profiles of both the upstream and downstream ORFs were similar for the wild-type and mutant strains (Fig. 4C). Taken together, these results exclude the possibility that insertion of the hygromycin-resistance gene causes polar effects.  www.nature.com/scientificreports www.nature.com/scientificreports/  www.nature.com/scientificreports www.nature.com/scientificreports/ increased the length of the lag phase and concomitantly reduced the exponential growth phase, suggesting that recA plays a role in the growth and viability of M. smegmatis. Furthermore, these results are consistent with the known function of RecA in the rescue of stalled or collapsed replication forks [71][72][73] . Because the recA and recX genes form a single operon, the impact of their deletion on M. smegmatis growth was investigated. The ΔrecA ΔrecX knockout strain exhibited a growth phenotype that was intermediate between the ΔrecA mutant and the wild-type strain; however the growth at late log and stationary phases was similar to that of ΔrecA and wild-type strains.
Deletion of recA, but not recX, renders M. smegmatis more susceptible to DNA damaging agents. Similar to other eubacteria, the mycobacterial RecA protein plays a crucial role in regulating the SOS response upon DNA damage [74][75][76][77] . To test whether the absence of recArecX and recX affect the ability of M. smegmatis to effectively repair damaged DNA, the mutants were exposed to a range of agents with varied mechanisms of DNA damage. In these experiments, stress was induced by exposing the ΔrecA, ΔrecX and ΔrecA ΔrecX mutant strains to UV-irradiation, MMS, H 2 O 2 or ciprofloxacin during the exponential phase (OD 600 of 0.6) of the bacterial growth. After 3 h incubation, the cells were washed and their viability was assayed by spotting ten-fold serial dilutions of the cultures on Middlebrook 7H10 agar plates containing hygromycin (100 µg/ml). The most appropriate concentration/dose of the DNA-damaging agents was determined after evaluating the differences in cell viability between the strains using plate assays. In the absence of DNA-damaging agents, all the strains exhibited similar levels of cell viability (Fig. 6). In contrast, in the presence of DNA-damaging agents, ΔrecA strain showed a pronounced growth defect accompanied with 100-fold decrease in plating efficiency relative to the wild-type strain. This phenotype is consistent with the available literature. In contrast, the lethal effect of UV, MMS or ciprofloxacin, but not H 2 O 2 , was partially blocked by deletion of the recX gene in the ΔrecA strain. Although the basis for lack of H 2 O 2 effect is not clear, the concentration used was probably not high enough to affect growth. Interestingly, the ΔrecX strain exhibited a slightly more resistant phenotype against all four DNA-damaging agents than the ΔrecA ΔrecX strain, indicating a possible gain-of-function due to loss of recX gene. Given these results, it is apparent that recA plays an active role in the SOS response and that the sensitivity of DNA damaging agents is slightly suppressed in ΔrecX strain.

Survival after treatment with different concentrations of DNA-damaging agents.
Mycobacteria experience a number of adverse stress conditions, such as oxidative, nutritional and drug-induced stresses [78][79][80][81][82] . To further explore cell viability differences, experiments were carried out in which the cell viability was measured by colony forming units (CFUs). First, the viability of M. smegmatis mc 2 155 ΔrecA, ΔrecX and ΔrecAΔrecX mutants were tested in comparison with that of the wild-type by challenging them with increasing concentrations of MMS. The strains were grown in the presence of indicated concentrations of MMS, until the cultures reached an OD 600 of 0.6. The cell viability was assayed by plating predetermined number of cells on 7H10 agar plates. The cells were judged alive if they were able to divide and form colonies. The mutants, unlike the wild-type strain, showed increased sensitivity to MMS in a concentration-dependent manner. The quantitative analysis indicated that the ΔrecA mutant exhibited relatively higher sensitivity to MMS, which increased with increasing concentrations of MMS (Fig. 7A). In the case of ΔrecX, the cells were ~2-fold less sensitive to MMS compared to the ΔrecA cells. On the other hand, the ΔrecA ΔrecX mutant showed more sensitivity to MMS in contrast to the ΔrecX mutant. The sensitivity of the ΔrecX mutant to MMS indicate that recX may have yet unidentified targets in addition to RecA. This assumption needs further investigation.
Next, the sensitivities of the M. smegmatis mc 2 155 ΔrecA, ΔrecX and ΔrecA ΔrecX mutant strains relative to the wild-type strain were determined by exposing them to increasing concentrations of H 2 O 2 . The results show that the ΔrecA mutant was less sensitive to this treatment (Fig. 7B). The deletion of recX alone, or deletion of the recA-recX locus, had only a minor effect on the viability and proliferation of mutant cells in response to H 2 O 2 treatment. Notably, the ΔrecA and ΔrecAΔrecX mutant strains exhibited a sensitive growth phenotype www.nature.com/scientificreports www.nature.com/scientificreports/ to UV irradiation and ciprofloxacin that is significantly much more severe than to MMS or H 2 O 2 (Fig. 7C,D). Furthermore, ΔrecA strain displayed greater sensitivity to UV irradiation and ciprofloxacin than the ΔrecA ΔrecX double mutant strain. the expression of recA and recX genes are induced by DNA damage. The regulatory elements upstream of the recA gene are not conserved in all mycobacterial species. For example, the M. tuberculosis recA gene is transcribed from two promoters: both are DNA damage-inducible, albeit through different mechanisms. The promoter P1 of M. tuberculosis recA (proximal to the start codon) can be induced following DNA damage independent of the LexA and RecA proteins. In contrast, the promoter P2 (located away from the recA start codon) is regulated by LexA, which is functionally analogous to the E. coli recA promoter [83][84][85] . Interestingly, the mechanism of DNA damage induction in M. tuberculosis is not fully conserved in M. smegmatis 46,75 . These findings emphasize the need to understand the expression and regulation of the M. smegmatis recA and recX genes and their roles in response to DNA-damaging agents.
As described above, hypoxia caused an increase in the expression of M. smegmatis recA and recX genes (Fig. 1). To further corroborate the notion that the expression of M. smegmatis recA and recX genes is damage-inducible, www.nature.com/scientificreports www.nature.com/scientificreports/ the kinetics of induction with and without DNA damage was determined using RT-qPCR. Total RNA samples were prepared from M. smegmatis cells harvested at 0, 3, 6, 9 and 12 h after exposure to UV light as well as from untreated cultures and analysed as described in the Methods section. The results revealed no significant differences between the recA and recX transcript levels in untreated cells. In contrast, a marked increase in the recA and recX mRNA transcripts was observed 3 h after exposure to UV radiation, and then decreased slightly thereafter. A comparison of the RT-qPCR data indicates important differences between the recA and recX transcript levels. Exposure to UV radiation led to a ~8-fold increase in recA mRNA over control, while recX by ~3-fold (Fig. 8A,B). Thus, although recA and recX mRNA exist in the same transcriptional unit, these results support the idea that the production and/or stability of recX mRNA transcript is subject to an additional posttranscriptional regulatory mechanism.
These results led us to perform additional experiments to determine the kinetics of accumulation of RecA and RecX proteins in M. smegmatis cells with or without DNA damage. Western blotting assays were performed on whole cell lysates using polyclonal antibodies raised against RecA and RecX (Fig. 8C). The quantification of Western blots showed that the cells contained significant quantities of both RecA and RecX proteins in uninduced cells (Fig. 8D,E). By comparison, a 2-fold increase in the abundance of both RecA and RecX (over control) was seen in cells 3 h after exposure to UV radiation and decreased thereafter. Importantly, the induction of RecA and RecX proteins exhibited a pattern reminiscent of that seen in cells under hypoxic conditions (Fig. 2), although the mechanisms by which they damage DNA are likely to be different.

Concluding remarks.
Numerous studies have demonstrated that recA and recX perform a broad range of functions related to DNA repair and recombination 7,10 . To our knowledge, the stimuli that activate the expression of M. smegmatis recA and recX genes have not been identified. In this study, the expression levels of these genes were assessed in cells at aerobic growth and in response to various stress conditions. In M. smegmatis, similar to many bacterial species, recA and recX belong to the same operon, and the recX gene is located immediately downstream of the recA gene, and share overlapping coding regions 86 . It was found that DNA-damaging agents induced the expression of both genes to different extents; however, the expression ratios follow a similar pattern. Interestingly, the levels of both RecA and RecX remain high under stress conditions compared to aerobic growth conditions.
Several studies have demonstrated alternative roles for recX in recombinational DNA repair promoted by RecA. While RecX protein physically interacts with RecA, and functions as a potent inhibitor of all known functions of the latter in many bacterial species 14,46,48 , it potentiates homologous recombination in N. gonorrhoeae and B. subtilis 25,26 . To gain insights into the role of recX in stress response in mycobacteria, knockout mutants of M. smegmatis recX and recArecX were constructed. Interestingly, the deletion of recX in M. smegmatis resulted in a slightly more resistant phenotype against all four DNA-damaging agents, indicating a possible gain-of-function due to loss of recX gene. The molecular basis of this effect is not clear, which seems worthwhile to explore in future studies. Whilst our analysis was mainly focused on the genetic interaction between recA and recX in (C) UV radiation and (D) ciprofloxacin. The error bars represent the standard error of the mean calculated from 3 independent replicates. Significant differences are indicated by *p < 0.05; **p < 0.01 and ***p < 0.001. ns, not significant.
www.nature.com/scientificreports www.nature.com/scientificreports/ recombinational DNA repair pathway, interestingly, recX seems to play an important role in bacterial growth. In summary, these findings are consistent with the idea that M. smegmatis recX plays an important role in DNA repair/recombination processes under adverse conditions, perhaps by regulating the detrimental effects of a subset of the yet unknown genes.