Expression and Mutational Analysis of DinB-Like Protein DR0053 in Deinococcus radiodurans

In order to understand the mechanism governing radiation resistance in Deinococcus radiodurans, current efforts are aimed at identifying potential candidates from a large repertoire of unique Deinococcal genes and protein families. DR0053 belongs to the DinB/YfiT protein family, which is an over-represented protein family in D. radiodurans. We observed that dr0053 transcript levels were highly induced in response to gamma radiation (γ-radiation) and mitomycin C (MMC) exposure depending on PprI, RecA and the DrtR/S two-component signal transduction system. Protein profiles demonstrated that DR0053 is a highly induced protein in cultures exposed to 10 kGy γ-radiation. We were able to determine the transcriptional start site of dr0053, which was induced upon irradiation, and to assign the 133-bp promoter region of dr0053 as essential for radiation responsiveness through primer extension and promoter deletion analyses. A dr0053 mutant strain displayed sensitivity to γ-radiation and MMC exposure, but not hydrogen peroxide, suggesting that DR0053 helps cells recover from DNA damage. Bioinformatic analyses revealed that DR0053 is similar to the Bacillus subtilis protein YjoA, which is a substrate of bacterial protein-tyrosine kinases. Taken together, the DNA damage-inducible (din) gene dr0053 may be regulated at the transcriptional and post-translational levels.


Introduction
Deinococcus radiodurans (D. radiodurans) is a strain of polyextremophilic bacteria that is capable of withstanding up to 15kGy gamma radiation (γ-radiation) [1], several weeks of desiccation [2], 500 J/m 2 UV-C radiation, [3] and various DNA damaging chemicals such as mitomycin C (MMC) [4], with almost no loss of viability. Furthermore, these bacteria exhibit unmatched resistance to oxidative stress after stimulation with a variety of different stresses [5]. Therefore, the mechanisms that underlie the extreme tolerance to multiple stresses in this Canada). The source strength was approximately 215 kCi at a dose rate of 15 kGy/h. Following irradiation, the Deinococcus cells were harvested for the subsequent analyses.

Construction of mutant strains
The D. radiodurans mutant strains were constructed using the double cross-over recombination method as previously described [25]. For the construction of the recA disruption mutant (recA::cm), the 1.4-kb fragment containing recA was PCR-amplified from D. radiodurans genomic DNA using the sequence-specific primer set recA-1F and recA-1R (S1 Table) and cloned into the pGEM-T Easy vector (Promega, Madison, WI). The resulting plasmid was digested with NruI and ligated with the chloramphenicol resistance cassette (916 bp) obtained from the pKatCAT plasmid [26]. The whole ligated product (2,317 bp) was PCR-amplified using the same primer set used for recA amplification and transformed into D. radiodurans cells as previously described [27]. The chloramphenicol-resistant transformants were grown for several generations in TGY supplemented with chloramphenicol. The recA disruption was confirmed by diagnostic PCR using the specific primers recA-2F and recA-2R and nucleotide sequencing (S1 Fig. and S1 Table). Using the primer sets dr0055-1F/dr0055-1R and dr0055-2F/dr0055-2R (S1 Table), approximately 1 kb of the upstream and downstream dr0055 regions, respectively, were PCR-amplified to construct the D. radiodurans dr0055 deletion mutant (Δdr0055). The upstream and downstream regions were cloned into the XhoI/EcoRV and XbaI/PstI sites of pKa-tAPH3 [28], respectively. The resultant recombinant plasmid was transformed into D. radiodurans cells, and the transformants were screened on TGY-kanamycin agar plates. Since the sizes of the kanamycin-resistant cassette (aph) and dr0055 gene are very similar, the PCR product obtained from the mutant and wild-type was restriction-digested with NdeI to confirm the replacement of the dr0055 gene with aph that has the restriction site for NdeI (S1 Fig.). The D. radiodurans dr0053 deletion mutant (Δdr0053) was constructed as the dr0055 mutant with some modifications. Briefly, the primer sets dr0053-1F/dr0053-1R and dr0053-2F/dr0053-2R (S1 Table) were used to amplify the upstream (1,852 bp) and downstream (1,856 bp) dr0053 regions. The dr0053-1R primer is complementary to dr0053-2F; hence, in a second PCR reaction with the dr0053-1F and dr0053-2R primers, the two separate PCR products were combined, generating a fragment lacking the dr0053 ORF. The final PCR product was cloned into the pGEM T-easy vector (Promega). It was subsequently digested with SmaI and then ligated to a 1-kb HincII fragment harboring the aph cassette from pKatAPH3 [28]. This resultant plasmid was transformed into D. radiodurans cells, and the transformants were screened on TGY-kanamycin agar plates. Gene replacement was confirmed by diagnostic PCR using the primers dr0053-3F and dr0053-3R (S1 Fig., S1 Table), which bind outside the mutant cassette on the genomic D. radiodurans DNA.

Quantitative real-time PCR (qRT-PCR)
A 5-ml culture grown to log phase (OD 600 % 1.0) was irradiated or incubated in the presence of MMC (5 μg/ml) or H 2 O 2 (60 mM) for 1 h. After stress exposure, the cells were collected by centrifugation, re-suspended in 1 ml of RiboEX reagent (GeneAll Biotechnology, Korea) and lysed in a Precellys 24 bead beater (Bertin Technologies, France) using 0.1 mm-diameter Zirconia/Silica beads (Biospec Products, USA). The total RNA was purified using the RNeasy Mini kit (Qiagen, Germany) and RNase-free DNase (Qiagen) according to the manufacturer's instructions. For real-time PCR analysis, cDNA was synthesized from 1 μg of total RNA using the PrimeScript first-strand cDNA Synthesis Kit (Takara Bio Inc., Japan) as recommended by the manufacturer's instructions. Real-time qPCR amplification was performed with SYBR Premix Ex Taq (Takara) on an Eco™ Real-Time PCR System (Illumina, USA). The PCR reactions were performed as follows: one cycle of 95°C for 5 m followed by 40 cycles of 95°C for 10 s and 60°C for 30 s. The housekeeping gene dr1343, which encodes glyceraldehyde-3-phosphate dehydrogenase, was chosen as the internal control, because its expression level remains unaffected by ionizing radiation [29]. The primers used in for the qRT-PCR assay are summarized in S1 Table. The D. radiodurans lexA1 (XLK1) and lexA2 (XL2K1) mutant strains were provided by Dr. K. Satoh of the Japan Atomic Energy Agency, and the pprI, drtR, and drtS mutant strains were constructed previously and stored in our laboratory [30,31].

Primer extension assay
The total RNA from the 25 ml of culture was isolated as described in the earlier section (qRT-PCR). The oligonucleotide primer P dr0053 (S1 Table) was end-labeled with 80 μCi of [γ-32 P]-dATP (Amersham Pharmacia Biotech, UK) and 10 units of T4 polynucleotide kinase (Life Technologies, USA) for 30 m at 37°C. The labeling mixture was heated at 70°C for 10 m and purified using MicroSpin G-25 columns (Amersham Pharmacia Biotech). The [γ-32 P] end-labeled primer (0.5 pmoles) was resuspended in 4 μl of 5× hybridization buffer (1.25 M KCl, 10 mM Tris) and 30 μg of D. radiodurans RNA. For hybridization, the mixture was heated to 60°C for 3 m and then cooled to room temperature for 1 h. Subsequently, 5 μl of the reaction solution containing 5 μg of actinomycin D, 700 μM dNTPs, 10 mM MgCl 2 , 5 mM DTT, 20 mM Tris (pH 8.7), 30 units of RNasin (Promega, USA), and 150 units of Superscript Reverse Transcriptase (Life Technologies) was added. The mixture was incubated at 42°C for 70 m and treated with 100 units of RNase T1 (Roche, Switzerland) and 2 μl of 0.5 M EDTA at 37°C for 15 m. The sample was precipitated with ethanol after the addition of 1.4 μl of 5 M NaCl and 2.5 volumes of absolute ethanol and then washed with 75% ethanol. Each sample was resuspended in 6 μl of formamide dye and 4 μl of TE buffer (pH 8.0) and denatured at 90°C for 3 m. To map the first nucleotide of the reaction products, aliquots of each of the reactions were subjected to electrophoresis on 6% polyacrylamide-8 M urea gels alongside sequencing reactions initiated with the same primers that were used for the primer extension reactions. The gels were subsequently dried under a vacuum on filter paper, and the extension products were analyzed on a Fuji Bio-Imaging Analyzer BAS-2500 (Fujifilm, Japan). The primer extension products were run in parallel to the corresponding reactions to map the start site of the dr0053 transcript.
Two-dimensional (2-D) gel electrophoresis D. radiodurans cultures grown to log phase (OD 600 % 1.0) were exposed to 10 kGy of γ-radiation and allowed to recover for 1 h. Subsequently, the cultures were washed with phosphate buffered saline, resuspended in lysis buffer (8 M urea, 2 M thiourea, 2% CHAPS, 1% Pharmalyte, 1% dithiothreitol, and protease inhibitor) and incubated at room temperature for 5 m. The cellular debris was removed by centrifugation (10,000×g/3 m/4°C), and the clear supernatant was precipitated with 100% acetone for 3 h at −20°C. The protein pellet was resuspended in solubilization buffer [9 M urea, 2% β-mercaptoethanol, 2% pharmalyte (pH 4-7), and 8 mM PMSF] and resolved by isoelectric focusing in the first dimension (4-7 IPG strips, Amersham Pharmacia Biotech) using a MultiphorTM II electrophoresis system (GE Healthcare, USA) followed by 12% SDS-PAGE (SE600 Holliston, USA). The protein profile of three biological replicates was analyzed using the PDQuest™ 2-D analysis software (Bio-Rad, USA). The protein spots that differed in intensity by >2-fold in intensity between the control and irradiated samples and passed the t-test control (p<0.05) were identified by peptide mass fingerprinting with the search programs MS-FIT (UCSF Mass Spectrometry Facility, http://prospector.usuf.edu) and Mascot (Matrix Science, London, UK, http://www.matrixscience.com) and the NCBInr and Swiss-Prot databases.

Survival analysis
Cells grown to log phase (OD 600 % 1.0) were adjusted to *10 7 CFU/ml in TGY medium and then exposed to different types of DNA-damaging agents. The cells were treated as described below, serially diluted in 0.85% NaCl, spotted on TGY plates, and incubated at 30°C for two days prior to the enumeration of the colonies. For γ-radiation treatment, the cells were irradiated with different doses of γ-radiation at room temperature using a cobalt-60 γ-ray irradiator. The resistance to MMC was determined by exposing D. radiodurans cultures to 10 μg/ml MMC in TGY broth. The treated cultures were incubated for 1 or 2 h in the presence of MMC with shaking at 30°C, and the cells were then harvested to determine their viability. For H 2 O 2 stress treatment, the wild-type and mutant strains were incubated with H 2 O 2 (0, 40 and 60 mM) with shaking at 30°C for 1 h, and the cell survival fraction was calculated. For UVstress, the cells were serially diluted in TGY broth and spread on TGY agar plates. After the culture was soaked onto the plates, the plates were exposed to UVC light in a UVC ultraviolet crosslinker (CX-2000, UVP, USA) at 20 J/m 2 /s for different time intervals.

Purification of DR0053
The genomic DNA from D. radiodurans was isolated using the Wizard Genomic DNA purification kit (Promega), according to the manufacturer's instructions. PCR was performed with the primers (dr0053-4F and dr0053-4R) listed in S1 Table. The PCR product was purified, digested, and ligated into pET28a. The plasmid was then introduced into E. coli BL21(DE3) cells by transformation. The DR0053 6XHis protein was purified by Ni-NTA agarose (Promega) affinity chromatography, according to the manufacturers' instructions. In brief, E. coli BL21(DE3) cells carrying the pET28a-DR0053 vector were cultured with shaking at 37°C in LB medium (1% tryptone, 0.5% yeast extract, and 0.5% NaCl) containing 50 μg/ml kanamycin to an OD 600 of 0.8. Protein expression in the cells was then induced by incubating the cells with 0.5 mM isopropyl β-D-1 thiogalactopyranoside (IPTG) for 4 h. The cells were lysed by incubating the cells in 50 mM Tris-HCl (pH 8.0) containing 4 mg/ml lysozyme for 30 m at 37°C. The cellular extracts were then clarified by centrifugation for 30 m at 3800×g, incubated for 2 h with Ni-Agarose beads (GE Healthcare), and applied to a poly-prep column (Bio-Rad). The flowthrough was reloaded onto the column and then washed four times with 10 ml of washing buffer (1.5 M NaCl, 25 mM imidazole, and 20 mM Tris, pH 7.5). The bound protein was eluted with 4 ml of binding buffer containing 500 mM imidazole.

DNase activity assay
The DNase activity of DR0053 was assessed by the DNase agar plate and DNA degradation methods according to previous studies [35]. The purified DR0053 6XHis protein (10 μg) or DNase (Qiagen; 1 μg) in PBS (phosphate buffered saline) with 1 mM MgCl 2 was spotted onto the DNase test agar plates (BD Biosciences, USA). After incubating the plate at 30°C for 5 m, the plate was flooded with 10 ml of 2 N HCl. The excess acid was removed with a vacuum pipette, and the clear zones around DR0053 were compared with DNase as the positive control. The PCR product of the Deinococcal dr0023 and dr0024 (3 kb) regions was used to test the direct nuclease activity of the DR0053 protein. The purified PCR product in distilled water with 1 mM MgCl 2 was incubated with the indicated concentration of DR0053 or DNase for 30 m and the DNA was visualized on a 1% agarose gel.

dr0053 expression is highly induced upon γ-radiation and MMC exposure
The din genes in Gram-positive bacteria are induced in response to environmental stressors such as chemical mutagens, radiation, and oxidative stress. Additionally, γ-radiation-induced dr0053 expression has been previously studied by microarray and quantitative real-time PCR (qRT-PCR) [9,21]. To confirm and examine the pattern of dr0053 expression under DNAdamage conditions, we performed qRT-PCR using total RNA isolated from wild-type D. radiodurans R1 that had been exposed to different doses (0, 5, 10 or 15 kGy) of γ-radiation, MMC (5 μg/ml), or H 2 O 2 (60 mM). dr0053 transcript levels remained relatively unchanged immediately after radiation exposure (data not shown). However, there was a marked increase in its expression after 1 h of post-irradiation recovery (PIR) (Fig. 1A). After exposure to 5 kGy, dr0053 levels were increased by approximately 70-fold compared with the non-irradiated counterpart. These levels were almost doubled after 10 kGy exposure and also remained similar after 15 kGy exposure. dr0053 induction during PIR is consistent with the previous result that dr0053 expression peaked 1.5 h after a 15 kGy dose of γ-radiation [21]. dr0053 expression was also highly upregulated by treatment with another DNA damaging agent (MMC) but not with H 2 O 2 , suggesting that DR0053 is directly or indirectly responsive to DNA damage but not to oxidative stress ( Fig. 1B and 1C).

PprI, RecA, and DrtR are involved in dr0053 activation
The SOS response is a transcriptional circuit that is activated upon DNA damage. The RecA and LexA proteins play key roles in the regulation of the SOS response. RecA activates the SOS response, whereas LexA, a transcriptional repressor, negatively regulates SOS induction [36]. In B. subtilis, dinB is a canonical SOS gene; its expression is not activated by DNA-damaging agents in the absence of recA, and its promoter has a LexA binding site [37]. Although the error-prone SOS repair system is not observed in D. radiodurans due to the absence of dinP and umuC, D. radiodurans encodes RecA (DR2340) and the two LexA homologues LexA1 (DRA0344) and LexA2 (DRA0074) [5]. We investigated the effects of RecA and LexA on dr0053 expression using real-time PCR analysis. The marked activation of dr0053 expression by γ-radiation, which was observed in wild-type strains, was drastically compromised in the recA mutant strain (Fig. 2A). The lack of lexA1 or lexA2 did not affect dr0053 expression ( Fig. 2A), thus indicating that LexA is not involved in dr0053 regulation. The PprI protein, a Deinococcus-specific regulator, stimulates recA transcription and translation following exposure to γ-radiation [19,29]. Similar to the recA mutant strain, dr0053 expression was not fully activated in the pprI mutant strain ( Fig. 2A). This finding suggests that PprI can affect dr0053 expression via its regulatory function on RecA. DrtR (DR2415) and DrtS (DR2416) are a histidine kinase (HK) and a response regulator (RR) of a novel two-component signal transduction system (TCS), respectively, which responds to DNA damage and plays a role in the resistance of D. radiodurans to DNA damaging agents [31]. When dr0053 expression was examined in the drtR and drtS mutant strains, dr0053 induction was very poor compared with the wild-type strain (Fig. 2B). Because PprI functions as a general switch to activate DNA repair and various defense pathways [20], we also examined drtR expression in the pprI mutant strain to investigate the effect of PprI on drtR. γ-Radiation-dependent activation of drtR expression was observed in the wild-type and recA mutant strains, but the activation disappeared in the pprI mutant strain (Fig. 2C), indicating that PprI is involved in drtR regulation. Additionally, recA expression level were not different between the wild-type and drtR mutant strains before and after γ-radiation (data not shown). Because RecA and DrtR are necessary for dr0053 expression and are governed by PprI, PprI appears to be the primary regulator involved in dr0053 activation by γ-radiation. Expression profiles of dr0053 and drtR in mutant strains. Bacterial cultures were exposed to γradiation and allowed to recover for 1 h. After total RNA isolation, qRT-PCR analysis was performed to determine dr0053 (A and B) and drtR transcript levels (C). The fold increase in each strain was obtained by dividing the mRNA levels from the irradiated cells by that from the non-irradiated cells. Error bars indicate the standard deviations from three independent experiments performed in duplicate.
doi:10.1371/journal.pone.0118275.g002 dr0053 promoter analysis dr0053 transcript levels were highly induced upon γ-radiation (Fig. 1A); therefore, it must possess a strong radiation-inducible promoter. First, we performed a primer extension assay to determine the transcriptional start site of the dr0053 gene. In total, two primer extension products named P dr0053-1 and P dr0053-2 were detected from the total RNA of 10 kGy-irradiated cells. Their corresponding transcription start sites were mapped to positions-20 and-10 The wild-type strains harboring pRADZ3 (R1-pRADZ3) and the dr0053 mutant strains harboring pB6 (Δdr0053-pB6) were exposed to γ-radiation and were allowed to recover for 1 h. After total RNA isolation, qRT-PCR analysis was performed to determine dr0053 transcript levels. Relative expression values were determined by defining the mRNA levels from non-treated R1 cultures as 1. Error bars indicate the standard deviations from three independent experiments conducted in duplicate. nucleotides upstream of the dr0053 translational start site (Fig. 3A). Next, we performed a promoter deletion analysis to identify the regulatory regions involved in γ-radiation responsiveness. As illustrated in Fig. 3B, various promoter fragments of dr0053 were cloned into the lacZ reporter plasmid pRADZ3 [32]. D. radiodurans cells harboring pB1 and pB2 plasmids demonstrated approximately three-fold activation in response to 10 kGy radiation (Fig. 3C). The longer product P dr0053-1 was induced only after irradiation, whereas the shorter product P dr0053-2 was expressed constitutively in both irradiated and non-irradiated samples (Fig. 3A). To assess the contribution of P dr0053-2 to dr0053 expression, we measured β-galactosidase activity from R1 cells harboring the pB3 plasmid that had only P dr0053-1 . Deletion of the putative -10 region and the transcription start site of P dr0053-2 , which spans the region from -8 to -19 relative to the start codon of dr0053, increased dr0053 expression under both irradiated and non-irradiated conditions, but the γ-radiation-dependent activation of dr0053 was still observed (Fig. 3C). These results demonstrate that dr0053 expression is driven by P dr0053-1 and not by P dr0053-2 . Thus, the shorter product may have been generated by pre-maturation and/or degradation of the longer transcript.
The shortest promoter (pB4) displayed the highest expression of the constructed plasmids, but the β-galactosidase activity remained almost constant under both irradiated and non-irradiated conditions (Fig. 3C). This result indicates that the 133-base-pair (bp) region (from -133 to -266 nucleotides relative to the start codon) between the pB3 and pB4 promoter fragments may provide a binding site for a transcriptional repressor, thereby regulating dr0053 expression in response to γ-radiation. As the LexA repressor is not involved in dr0053 expression ( Fig. 2A), we examined the proximal region of dr0053 to search for putative repressor proteins. The dr0055 ORF, which encodes a repressor protein, is located approximately 441 bp upstream of dr0053. To assess whether DR0055 is a repressor of the dr0053 gene, a dr0055 deletion mutant was constructed, and dr0053 expression was examined in this construct using qRT-PCR. There was marginal change in dr0053 expression in dr0055 deletion mutant (data not shown), indicating that DR0055 is not a repressor of dr0053. Therefore, further research is necessary to determine how dr0053 is repressed under non-irradiated conditions. When comparing pB2 with pB3, the absence of a 12-bp region increased dr0053 expression (Fig. 3C). The dr0053 promoter activity was also reduced in pB5 by extension of the 12-bp region compared to pB4 (Fig. 3C). These results demonstrate that the 12-bp region, which is adjacent to the start codon, plays a role in the repression of dr0053 regardless of γ-radiation.
The fold increase in activity from the cloned dr0053 promoters was much lower than its expression from the native genomic site (Figs. 1 and 3). Because the 12-bp region of P dr0053-1 (positions -8 to -19 relative to the start codon) affected dr0053 expression (Fig. 3C), sequences around the start of the DR0053 coding region might influence dr0053 activation in response to γ-radiation. To investigate this possibility, we amplified the DNA fragment, which covers almost the entire DR0054 and DR0053 ORFs, and cloned the 849-bp region into pRADZ3 (Fig. 3B). The pB6 plasmid was introduced into the dr0053 mutant strains (Δdr0053-pB6), and the fold increase of dr0053 was measured by qRT-PCR. dr0053 mRNA levels increased approximately four-fold in Δdr0053-pB6 compared with R1 cells before γ-radiation (Fig. 3D). Following γ-radiation, we observed a 4.5-fold increase in dr0053 mRNA levels in Δdr0053-pB6, whereas dr0053 levels increased more than 200-fold in R1 cells (Fig. 3D). This result shows that the low activation of the cloned dr0053 promoters is not related to a lack of specific sequences surrounding the translational start codon. Because the pRADZ3 plasmid used in this study has copy number similar to that of the chromosome, at 7 to 10 copies per cell [32], this difference in activity may be attributed to the difference in the DNA context surrounding the dr0053 promoter region in the genome and the plasmid. It is known that DNA supercoiling generated at a local level by transcription can influence nearby events in the same DNA molecule such as promoter activity [38].

DR0053 is produced at high levels during post-irradiation recovery
When cells encounter any stress, there are often a multitude of changes that occur at the transcriptional level. However, those proteins that are highly critical for stress recovery are preferentially translated into protein. In an effort to determine whether DR0053 is truly induced at the protein level and to identify other DinB/YfiT family of proteins, two-dimensional electrophoresis was performed, and the protein profile of D. radiodurans R1 cells exposed to 10 kGy of γ-radiation was compared with that of unirradiated cells after 1 h of PIR. Of the 13 spots that displayed significant changes, eight were up-regulated, and five were down-regulated in response to γ-radiation ( Fig. 4 and Table 1). All of these protein spots were identified by MALDI-TOF. DR0053 was among the top induced proteins observed in this 2D protein profile (Fig. 4). In addition, other up-regulated protein spots were identified as DdrA [10], DdrD [9], the single strand binding protein SSB [39], the pleiotropic protein promoting DNA repair PprA [15], and the tellurium resistance protein TerB [40], all of which have been demonstrated to be induced in response to γ-radiation and have been implicated in radiation resistance. However, we were unable to identify any other member of the DinB/YfiT family of proteins in our protein profile.

Gene disruption of dr0053 increases susceptibility to DNA damage
Since dr0053 was highly induced upon γ-radiation and MMC exposure (Figs. 1 and 4), it may play an important role in survival under DNA-damaging conditions. First, to assess DR0053 function, a D. radiodurans dr0053 deletion mutant was constructed. The dr0053 gene was replaced with a kanamycin-resistant cassette (aph) under a constitutively expressed Deinococcal Protein profiles of non-irradiated and irradiated D. radiodurans. D. radiodurans R1 cultures were exposed to 10 kGy of γ-radiation and allowed to recover for 1 h. Whole-cell proteins were extracted and resolved by isoelectric focusing in the first dimension using 4-7 IPG strips and by 12% SDS-PAGE in the second dimension. The protein spots (marked in circles) that displayed significant differences in expression were selected for further analysis. catalase (kat) promoter. The deletion of dr0053 did not confer any changes in the growth rate compared with the wild-type strain R1 (data not shown). The dr0053 mutant strain and D. radiodurans R1 were also evaluated for their ability to withstand γ-radiation, MMC, H 2 O 2 , and UVC. Under conditions of 15 kGy γ-radiation, the deletion mutant strain was more sensitive than the wild-type. Similarly, the mutant cells were more sensitive to MMC exposure (at 10 μg/ml) than the wild-type R1 cells. However, no significant differences in survival were observed between the mutant and the wild-type strains in response to H 2 O 2 and UVC stresses (Fig. 5). The lack of response of the dr0053 mutant to H 2 O 2 is consistent with the unchanged expression of dr0053 after H 2 O 2 treatment (Fig. 1). This finding suggests that DR0053 may play a role in the survival of D. radiodurans under the DNA damaging conditions generated during γ-radiation and MMC treatment.
DR0053 is a homologue of the B. subtilis YjoA protein In D. radiodurans, it is speculated that DinB-like family proteins are metal-dependent hydrolases because they have three conserved histidine residues, which indicate metal-binding properties [23,24]. Thus, DinB-like proteins are predicted to function as nucleases involved in the cleaning up of DNA damaged products, which are formed immediately after exposure to γradiation [23,24]. To investigate whether DR0053 has DNase activity, cell lysates of wild-type and dr0053 mutant cells were spotted onto DNase test agar according to a previous study [35]. DNase activity was found to be similar in both wild-type and dr0053 cell lysates, even after exposure to 10 kGy of γ-radiation (data not shown). To examine this DNase activity in vitro, we purified the DR0053 protein and compared its activity with that of commercially available DNase I using the same method. A distinct zone of clearance was observed only in the area surrounding commercial DNase I (Fig. 6A). To measure its DNase activity directly, DR0053 was incubated with PCR products amplified from the genomic DNA of R1 cells. However, no detectable DNase activity was seen even when using unusually high concentrations of DR0053 Table 1. List of differentially regulated proteins in R1 cells exposed to 10 kGy of γ-radiation. Regulation of DR0053 in D. radiodurans  ( Fig. 6B). Interestingly, the first and second histidine residues, which are conserved in metaldependent hydrolases, are substituted with glutamic acid in DR0053 [23].
To gain a better understanding of the structural/functional determinants of the DR0053 protein, a bioinformatic analysis was performed. The pfam domain search demonstrated that DR0053 belongs to the DinB protein family (pfam05163). This family consists of seven SYSTERS (SYSTEmatic Re-Searching) protein families [41]. Among them, DR0053 belongs to the family of cluster O154983 with six other proteins (Fig. 7A). The relationship between DR0053 and the other proteins in this cluster was investigated through neighbor-joining methods of MEGA 5.0 program with 1,000 bootstrap replicates [42]. A phylogenetic tree indicated that DR0053 is located in a distinct phylogenetic branch with YjoA of B. subtilis (Fig. 7A). It suggests that DR0053 function might be more similar to YjoA than any other proteins. A structure prediction analysis using PHYRE and HHpred algorithms [43,44] also identified DR0053 as having high structural similarity to the DinB-like protein (YjoA) of B. subtilis (HHpred; 100% probability, E-value = 2e-27) (Fig. 7B). These show that DR0053 is a homologue of the B. subtilis YjoA protein. . The sequences were aligned in a ClustalW2 algorithm, and a consensus tree was generated using a 1,000× repeated bootstrapping process with the MEGA 5 program [42]. The numbers on the nodes are the frequency (in percentages) in which a cluster appears in a bootstrap test. (B) Sequence alignment of DR0053 with YjoA (PDB code 3dka) using the HHpred program [44]. Sequences 'ss_pred' and 'ss_dssp' denote the PSI-PRED secondary structure prediction and the secondary structure assigned by DSSP (Dictionary of Secondary Structure of Proteins) program. Upper and lower case amino acids in the consensus sequences indicate high (!60%) and moderate (!40%) conservation, respectively. Red, blue, and black letters represent charged, polar, and hydrophobic residues, respectively. Symbols indicating the quality of the column-column match: '|' very good, '+' good, 'Á' neutral, and '−' bad. doi:10.1371/journal.pone.0118275.g007

Discussion
The polyextremophilic D. radiodurans encodes 13 DinB/YfiT homologues, which is the highest number identified in any bacteria to date [23,24]. Such particular expansions of certain protein families are favored during evolution to aid in organism survival [23]. Considering the extreme multiple stress tolerance of this bacterium, genes belonging to such protein families may hold essential information to help elucidate its resistance mechanisms. The prototype of Deinococcal DinB family proteins is B. subtilis DinB whose expression is controlled by the SOS system [23,37]. In D. radiodurans, RecA positively regulated the expression of dr0053, one of the DinB family proteins, as it does in B. subtilis; however, LexA was not involved in this regulation ( Fig. 2A). This result is consistent with the finding that neither of the Deinococcal LexA homologues (LexA1 and LexA2) repressed recA expression although they are cleavable by RecA [33,45]. In addition to its participation in the SOS response, B. subtilis RecA is also responsible for DNA damage-dependent alterations in gene expression for nearly 600 genes, most of which are not repressed directly by LexA [36]. These observations indicate the presence of another transcriptional repressor, substituting for LexA, which suppresses dr0053 expression under non-irradiated conditions. Deletion analysis of the dr0053 promoter demonstrates the possibility that unidentified repressor binding sites are present in the 133-bp region upstream of the transcriptional start site of P dr0053-1 (Fig. 3). Some regulators are involved in the repression of radiation-inducible genes under non-irradiated conditions. The deletion of pprM (DR0907), which encodes a modulator of the PprI-dependent DNA damage response, and recX (DR1310) results in constitutive production of PprA and RecA, respectively, regardless of γ-radiation treatment [46,47].
The TCS, which is composed of an HK and an RR, is one of the most ubiquitous means by which bacteria sense, respond, and adapt to environmental changes. The HK perceives the environmental signal and transduces the signal to its cognate RR which, in turn, activates the specific response to adapt the cell to its new surroundings [48]. Until now, three RRs, DrRRA (DR2418), RadR (DRB0091), and DrtR, have been shown to be necessary for radiation resistance in D. radiodurans [31,49,50]. Deletion of drRRA downregulates the transcriptional levels of numerous genes related to stress response and DNA repair, such as kat, sod, recA and pprA [49]. Microarray analysis demonstrated that the drRRA mutation slightly reduced dr0053 expression under both normal and irradiation stress conditions [49]. Taking the effect of RecA on dr0053 expression into consideration ( Fig. 2A), DrRRA is likely to have a positive effect on dr0053 expression via RecA. Here, we also observed that DrtR is necessary for dr0053 activation by γ-radiation (Fig. 2B), although DrtR was not related to RecA (Fig. 2C). The involvement of two RRs in dr0053 regulation implies that the function of DR0053 is intimately connected to environmental changes.
The dr0053 gene was induced in response to γ-radiation in the presence of RecA, DrtR (Fig. 2) and DrRRA [49], and expression of these regulators decreased in the pprI mutant strains [19,49]. Therefore, PprI appears to be the primary regulator of dr0053 and might exert its effect on dr0053 expression via RecA, DrtR, and/or DrRRA. The dr0053 gene is found among PprI-dependent genes [20]. However, we observed a three-and six-fold induction in dr0053 expression in pprI mutant strains in response to 1 and 2 kGy of γ-radiation, respectively ( Fig. 2A). The double mutant ΔdrRRAΔpprI was more sensitive to γ-radiation than either the ΔdrRRA or ΔpprI single mutant [51]. In addition, RecA overexpression could partially restore the radioresistance of a pprI mutant strain [19]. These findings suggest that PprI and its downstream regulators have non-overlapping routes in addition to a common pathway for the regulation of target genes. Taken together, dr0053 expression is likely to be finely tuned by a multi-layered regulatory scheme in which PprI, RecA, DrRRA, and/or DrtR act together with unidentified regulatory proteins.
DR0053 has a critical residue found in metal-dependent catalytic enzymes, such as hydrolases and nucleases. Thus, it was previously proposed to encode a nuclease that helps clean up damaged DNA resulting from acute DNA damaging stresses [23,24]. However, when we examined its nuclease activity using the DNase test agar plate and Deinococcal DNA degradation assays according to a previous study [35], no DNase activity was detected under our experimental conditions (Fig. 6). Instead, we found out that DR0053 has a structural conformation that resembles the YjoA protein of B. subtilis (Fig. 7). It was recently discovered that YjoA is one of the substrates of the bacterial protein tyrosine-kinase (BY-kinase) PtkA in B. subtilis [52]. It should be noted that 9 of 36 (25%) identical amino acid residues between YjoA (153 aa in length) and DR0053 (158 aa in length) are concentrated in the C-terminal tail region (17 aa in length) of each protein. The phosphorylation site of YjoA is Y150 [45]. Thus, the conservation of this segment implies that DR0053 can be a substrate of a homologue of PtkA in D. radiodurans, and the Y155 of DR0053 serves as a putative phosphorylation site. Protein phosphorylation is a widespread post-translational modification that plays a key role in the regulation of cellular functions [53]. The eukaryotic-type serine/threonine protein kinase (eSTPK) DR2518, whose synthesis and phosphorylation are induced by γ-radiation, has been characterized for its role in bacterial responses to DNA damage [54]. Recently, PprA was identified as a substrate for this protein kinase [55], suggesting that protein phosphorylation, which can change an enzyme's activity level, cellular localization, or interaction with partners of the target protein [53], is a defense mechanism in D. radiodurans.
DNA repair genes, which are under the control of PprI in response to γ-radiation, exhibit phase-dependent expression during PIR: recA is significantly increased at the early and middle phases and return to normal levels in the late phase [20]. In the presence of PprI, interestingly, dr0053 shows a recA-like expression pattern during PIR [20]. In addition, DR0053 was highly produced after γ-radiation (Fig. 4), and the dr0053 mutant displayed increased sensitivity to γradiation and MMC exposure (Fig. 5). Taken together, these observations suggest that DR0053 is necessary to cope with the stress generated from these damaging agents. DR0053 is a probable substrate protein for BY-kinase of D. radiodurans (Fig. 7). BY-kinases, which phosphorylate tyrosine residues on their substrate proteins, are involved in several cellular processes, including the heat shock response, DNA replication, and the cell cycle. However, they have been best characterized for their involvement in the production of exopolysaccharide (EPS) [52,56]. A sequence homology search using the PSI-BLAST tool revealed that DRA0033, denoted "ExoP-related protein", is homologous to PtkA and is located within the gene cluster involved in EPS biosynthesis in D. radiodurans. This is consistent with the fact that most experimentally validated BY-kinases are encoded by genes located in large operons involved in EPS biosynthesis and export [53]. Although further research is warranted to identify a link between DR0053 and DRA0033, the location of DRA0033 provides a clue to the role of DR0053.
Supporting Information S1 Table. Primers used in this study. (DOC) S1 Fig. Verification of gene disruption by PCR. PCR fragments were amplified from genomic DNA of the wild-type R1 (lane 1) and its isogenic mutant strains (lane 2) using the primers listed in S1 Table. PCR products of the recA and dr0053 mutant strains were larger than those of R1 due to the antibiotic marker insertion. The PCR product obtained from the dr0055 mutant strain was digested with NdeI. M denotes the DNA size markers. (TIF)