Deinococcus radiodurans, an extremely radioresistant bacterium has a multipartite genome system and ploidy. Mechanisms underlying such types of bacterial genome maintenance and its role in extraordinary radioresistance are not known in this bacterium. Chromosome I (Chr I), chromosome II (Chr II) and megaplasmid (Mp) encode its own set of genome partitioning proteins. Here, we have characterized P-loop ATPases of Chr II (ParA2) and Mp (ParA3) and their roles in the maintenance of genome copies and extraordinary radioresistance. Purified ParA2 and ParA3 showed nearly similar polymerization kinetics and interaction patterns with DNA. Electron microscopic examination of purified proteins incubated with DNA showed polymerization on nicked circular dsDNA. ParA2 and ParA3 showed both homotypic and heterotypic interactions to each other, but not with ParA1 (ParA of Chr I). Similarly, ParA2 and ParA3 interacted with ParB2 and ParB3 but not with ParB1 in vivo. ParB2 and ParB3 interaction with cis-elements located upstream to the corresponding parAB operon was found to be sequence-specific. Unlike single mutant of parA2 and parA3, their double mutant (ΔparA2ΔParA3) affected copy number of cognate genome elements and resistance to γ-radiation as well as hydrogen peroxide in this bacterium. These results suggested that ParA2 and ParA3 are DNA-binding ATPases producing higher order polymers on DNA and are functionally redundant in the maintenance of secondary genome elements in D. radiodurans. The findings also suggest the involvement of secondary genome elements such as Chr II and Mp in the extraordinary radioresistance of D. radiodurans.
Until recently, the bacterial genomes were synonyms of a single circular chromosome and extrachromosomal plasmids. Now we know that there are many bacteria that harbor multipartite genome system. The numbers of copies of these genome elements, including primary chromosomes may vary from one to several copies per cell . In general, the primary chromosome is larger and tends to have significantly more conserved housekeeping genes that encode for core cellular functions and contribute to greater conservation of the contents. On the other hand, the secondary genome elements show a greater variability and encode accessory functions associated with adaptation and survival in different niches and largely contribute to stress tolerance [2–4]. The secondary chromosomes are normally smaller than primary chromosome  and believed to have originated by the mechanisms like the split of a primary chromosome, chromosome duplication or acquisition of a large plasmid with essential genes. The faithful inheritance of multipartite genome system and the maintenance of ploidy are not fully understood in bacteria. The genome segregation in bacteria harboring single circular chromosome and low copy plasmids occurs largely by the involvement of the tripartite genome segregation (TGS) system. The TGS consists of an origin-proximal cis-acting (centromere-like) DNA sequences, the centromere binding adaptor proteins like ParB or ParB homologs and the P-loop Walker ATPases like ParA or Par A like motor proteins [6,7].
Deinococcus radiodurans, a radioresistant bacterium, is characterized with an efficient DNA double-strand break repair and the strong oxidative stress tolerance mechanisms [8–10]. The cytogenetic features like multipartite genome system with two chromosomes (2 648 638 bp and 412 348 bp) and plasmids (177 466 bp and 45 704 bp) and their ploidy are equally interesting features in this bacterium . Molecular mechanisms underlying the evolution and maintenance of multipartite genome system and its faithful inheritance into daughter cells, maintenance of ploidy and its functional significance in extreme phenotypes of this bacterium are not clearly understood. The chromosome I (Chr I) and chromosome II (Chr II) contain one putative parAB operon each, while megaplasmid (Mp) contains two putative parAB operons . Previously, the partitioning system of Chr I including the centromeric sequences (segS) of Chr I and ‘Par’ proteins (ParA1 and ParB1) has been characterized . The partitioning mechanisms of Chr II and Mp (together referred to as secondary genome elements unless specified) are not known and would be worth studying. However, it was shown that ATPase encoded from DR_A0001 in Chr II (ParA2) functions differently in the absence of its cognate ParB2 .
Here, we have characterized ParA2 and P-loop ATPase encoded by DR_B0001 (hereafter referred as ParA3) on Mp and established their roles in the cognate genome copy maintenance and in resistance to different abiotic stresses in D. radiodurans. We demonstrated that ParA2 and ParA3 have nearly similar functions in vitro and can complement each other's phenotype in vivo. They produced higher order polymers on DNA in the presence of ATP, and ATP hydrolysis was not stimulated by non-specific DNA, but led to the conformational change that was very similar to the one observed in the presence of ADP. These proteins did interact with each other, but not with ParA1. Also, ParA2 and ParA3 interacted with ParB2 (ParB encoded on Chr II) and ParB3 (ParB encoded on Mp), but not with ParB1 in vivo. ParB2 and ParB3 showed sequence-specific interaction with cis-elements located upstream to the parAB operons of the secondary genome elements. Unlike single mutants, the double mutant of parA2 and parA3 (ΔparA2ΔparA3) showed a reduction in the copy number of secondary genome elements and was found sensitive to γ-radiation as well as to hydrogen peroxide (H2O2). In comparison with single mutant, the double mutant showed an increase in cell size and septum trapped nucleoid phenotype under microscopic observation. These results together suggested that nearly similar in vitro characteristics of ParA2 and ParA3 might have allowed them to complement each other's role in vivo. Furthermore, ploidy in secondary genome elements seems to play important roles in the radioresistance and oxidative stress tolerance in this bacterium.
Bacterial strains, plasmids and materials
Deinococcus radiodurans R1 (ATCC13939) was a kind gift from Professor J. Ortner, Germany . It was grown in TGY (tryptone [1%], glucose [0.1%] and yeast extract [0.5%]) medium at 32°C. E. coli strain NOVABLUE was used for cloning and maintenance of all the plasmids. E. coli strain BTH101 (cyaA−) (hereafter referred to as BTH101) was used for the co-expression of different ParA proteins on BACTH (bacterial two-hybrid system) plasmids to monitor protein–protein interaction in E. coli. E. coli strain BL21 (DE3) pLysS was used for the expression of recombinant proteins. E. coli cells harboring pUT18, pKNT25 and pET28a (+) and its derivatives were maintained in the presence of required antibiotics. Shuttle expression vector pVHS559  and their derivatives were maintained in the presence of spectinomycin (70 µg/ml) in E. coli and D. radiodurans, whereas pRadgro  and their derivatives were maintained in the presence of ampicillin (100 µg/ml) in E. coli and chloramphenicol (5 µg/ml) in D. radiodurans. Single mutants of parA2 (ΔparA2::nptII) and parA3 (ΔparA3::aadA) as well as their double mutant (ΔparA2::nptII–ΔparA3::aadA) from D. radiodurans used in this study, has been described recently . Standard protocols for all recombinant techniques were used as described in ref. . All the bacterial strains and plasmids used in this study have been listed in Supplementary Table S2. Antibodies against T18 (SC-33620) and T25 (SC-13582) domains of CyaA of Bordetella pertussis, respectively, were procured commercially (Santa Cruz Biotechnology, Inc). Antibody against polyhistidine tag was purchased from Sigma Chemicals Company, U.S.A. Molecular biology-grade chemicals and enzymes were procured from Sigma Chemicals Company, U.S.A., Roche Biochemicals, Mannheim, Germany, New England Biolabs, U.S.A. and Merk India Pvt. Ltd., India. Radiolabeled nucleotides were obtained from the Board of Radiation and Isotope Technology, Department of Atomic Energy (DAE), India (BRIT, India).
Bioinformatics analysis and molecular modeling
Multiple sequence alignment and functional motif search in ParA1 (DR_0013), ParA2 (DR_A0001) and ParA3 (DR_B0001) proteins were carried out using standard online bioinformatics tools as described earlier [13,18]. In brief, the amino acid sequences of ParA1, ParA2 and ParA3 proteins were subjected to a PSI-BLAST search with the SWISSPROT database. After five iterations, the sequences obtained were aligned by CLUSTALW and the conserved deviant of the Walker box ATP-binding motif and DNA-binding motif were searched. Homologous sequences were aligned by T-COFFEE and the conserved motifs were marked. ParA1, ParA2 and ParA3 proteins were modeled by I-TASSER server (http://zhanglab.ccmb.med.umich.edu/I-TASSER/) . The models were validated by Swiss model workspace encompassing the package of Anolea, DFire, QMEAN, Gromos, DSSP, Promotif and ProCheck (http://swissmodel.expasy.org/workspace/). The template used for the modeling of protein structure was derived from the Soj structure of Thermus thermophilus (PDB ID: 2BEK) . The modeled structure of both ParA2 and ParA3 were superimposed with Soj structure (PDB ID: 2BEK) of T. thermophilus as well as with deinococcal ParA1 model structure using Pymol software.
Construction of recombinant plasmids
Details of the primers used in the construction of recombinant plasmids and plasmids used in this study are given in Supplementary Tables S1 and S2, respectively. A suicidal plasmid pNOSOUT conferring spectinomycin resistance in D. radiodurans was constructed for creating gene knockout in this bacterium and described in ref. . Genomic DNA of D. radiodurans R1 was prepared as reported previously  and coding sequences of ParA2 (DR_A0001), ParA3 (DR_B0001), ParB2 (DR_A0002) and ParB3 (DR_B0002) were PCR amplified from genomic DNA using gene-specific primers as given in Supplementary Table S1. PCR products were ligated at the Nde I and Xho I sites in pET28a (+) to yield pETA2, pETA3, pETB2 and pETB3, respectively. These plasmids were sequenced for the presence of cloned genes and further used for protein expression and purification or generation of polyhistidine-tagged translational fusion for in vivo interaction study.
For protein–protein interaction studies in E. coli, the coding sequences of ParA1 (DR_0013), ParA2 and ParA3 were PCR amplified using gene-specific primers as given in Supplementary Table S1 and ligated in BACTH plasmids, namely pKNT25 and pUT18 as given in Supplementary Table S2. For in vivo interaction of ParA and ParB proteins in D. radiodurans, the translational fusion of ParA2, ParA3, ParB1, ParB2 and ParB3 were generated with polyhistidine tag (from their pET28a (+) variants) in pRADgro vector using PETHisF and PETHisR primers as given in Supplementary Table S1, and recombinant plasmids obtained were named as pRADhisA2, pRADhisA3, pRADhisB1, pRADhisB2 and pRADhisB3 respectively. Similarly, T18 tag fusions of different parAs were PCR amplified from their pUT18 variants and were cloned in pVHS559 using BTHF (PV) and BTHR (PV) primers as given in (Supplementary Table S1), and resultant plasmids were named as pV18A1, pV18A2 and pV18A3, respectively (Supplementary Table S2). D. radiodurans cells harboring pVHS559 derivatives as well as pRADgro derivatives in different combinations were induced with 5 mM isopropyl-β-d-thiogalactopyranoside (IPTG) for the expression of recombinant proteins from pVHS559 derivatives while pRADgro variants express recombinant proteins constitutively. The expression of each fusion protein in D. radiodurans was ascertained by immunoblotting using antibodies against T18 tag of polyhistidine tag, respectively (Supplementary Figure S1C–E). Recombinant plasmids pKNTA1, pKNTA2 and pKNTA3 used in this study were constructed previously .
Generation of insertional deletion mutants of parA2 and parA3
The single (ΔparA2 or ΔparA3) and double mutants (ΔparA2ΔparA3) of parA2 and parA3 genes were generated as described recently . In brief, ∼1 kb upstream and downstream region from mid of parA2 and parA3 ORFs were PCR amplified and cloned in pNOKOUT (KanR) and pNOSOUT (SpecR) to yield pNOKA2UD and pNOSA3UD, respectively. The upstream fragments were cloned at Kpn I–Apa I and downstream at Bam HI–Xba I sites. These constructs were linearized by Xmn I and transformed into D. radiodurans separately as well as together and transformants were grown for several generations under required, selection pressure until the homozygous insertion and the replacements of middle portion parA2 with nptII cassette and parA3 with aadA cassette were achieved in the genome of D. radiodurans. This was ascertained by PCR amplification using parA2 and parA3 gene-specific as well as antibiotic cassettes (nptII and aadA) specific primers in different combination.
Purification of recombinant proteins
The recombinant ParA2, ParA3, ParB2 and ParB3 were purified from E. coli BL21 (DE3) pLysS expressing these proteins on pETA2, pETA3, pETB2 and pETB3, respectively, as described recently . In brief, mid-logarithmic phase cells of E. coli BL21 (DE3) pLysS-expressing recombinant proteins were induced with 0.5 mM IPTG allowed to grow at 37°C for 3 h and kept overnight at 18°C. The cells were pelleted and stored at −70°C. The cell pellet was thawed and suspended in buffer A (20 mM Tris–HCl [pH 7.6], 300 mM NaCl) containing 10 mM imidazole, 0.5 mg/ml lysozyme, 1 mM PMSF, 1 mM MgCl2, 0.05% NP-40, 0.05% Triton X-100, protease inhibitor cocktail and 10% glycerol) and incubated at 37°C for 30 min. The mixture was sonicated for 10 min at 10 s pulses with intermittent cooling for 15 s at 25% amplitude. The cell-free extract obtained after centrifugation at 11,000 rpm for 30 min at 4°C was loaded onto a pre-equilibrated Ni-NTA column. The column was thoroughly washed with buffer A containing 50 mM imidazole and recombinant protein was eluted with buffer A containing 200, 250 and 300 mM imidazole. Fractions were analyzed on SDS–PAGE and those containing nearly pure proteins were pooled and protein was further purified using anion exchange column chromatography. Different fractions containing pure protein were pooled and concentrated using 10 kDa cutoff spin columns. The protein solution was centrifuged at 16 000 rpm for 30 min and the supernatant containing mostly soluble protein was dialyzed in a buffer containing 20 mM Tris–HCl (pH 7.6), 200 mM NaCl, 50% glycerol and 1 mM PMSF and stored at −20°C (Supplementary Figure S2). Protein concentration was determined by taking OD at 280 nm in NanoDrop (Synergy H1, Hybrid Multi-Mode Reader Biotek) using mass extinction coefficient of both the proteins. The refolding of purified ParA2 and ParA3 proteins was confirmed by recording Circular Dichroism spectroscopy in phosphate buffer using JASCO, J815, Japan as described previously  (Supplementary Figure S2).
Electrophoretic mobility shift assay
The DNA-binding activity of ParA proteins was assayed by electrophoretic mobility shift assay (EMSA) as described in ref. . In brief, different concentrations of proteins (0–2.5 µM) were incubated with 100 fmol of 3 kb linear dsDNA (EcoRI linearized pBluescript II SK+) in a total volume of 30 µl containing DNA-binding buffer B (50 mM Tris–Cl (pH 8.5), 75 mM NaCl, 5 mM MgSO4 and 0.5 mM 1,4-dithiothreitol) for 10 min at 25°C in the absence and presence of 1 mM ATP, ADP or ATP-γ-S. The reaction mixture was mixed with 10% glycerol and loaded in 0.8% agarose gel. Agarose gel electrophoresis was performed in 0.5× TBE buffer at 50 mV at 8°C and gels were stained with ethidium bromide. Data were documented and analyzed for a shift in mobility with respect to the free DNA probe. The mobility retardation of the nucleoprotein complex (NPC) for each concentration has been calculated as difference in distance (cm) travel at each concentration with respect to total migration of DNA probe. It has been further plotted with respect to the different concentration of ParA2 and ParA3 as mean ± SD.
Similarly, the ParB2 and ParB3 interaction with cis-elements (cisII and cisMP) as well as non-specific DNA was studied by EMSA as described previously . For that, the ∼400 bp fragment containing 10 direct repeats (cisII) or 8 direct repeats (cisMP) of 17 mer or 16 mer (Supplementary Figure S3) located upstream to parA2B2 in chromosome II and parA3B3 in Mp, respectively, were PCR amplified using sequence-specific primers (see Supplementary Table S1). The PCR products were gel purified. DNA substrates were labeled with [γ32P] ATP using T4 polynucleotide kinase. Approximately, 30 nM labeled substrate was incubated with different concentrations of purified recombinant ParBs in a reaction buffer containing 50 mM Tris–HCl (pH 8.0), 75 mM KCl, 5 mM MgSO4 and 0.1 mM DTT at 37°C for 15 min. For the competition assay, a saturating concentration of ParBs was incubated with cis sequences before the different concentration of 400-bp non-specific competitor DNA (mid of ftsZ gene; Supplementary Table S1) was added and further incubated as per experimental requirements. A 10-fold higher concentration of cold cis sequences was also used in a competition assay for respective ParB and cis interaction. Mixtures were separated on 6% native PAGE gels, the gels were dried and autoradiograms were developed on X-ray films. The band intensity of bound and unbound fraction was determined by using ImageJ 2.0 software. The fraction of DNA probe bound to the protein was plotted as a function of the protein concentration by using GraphPad Prism 5. The Kd value for the curve fitting of individual plots was determined as described before .
Fluorescence anisotropy was measured as described in ref. . In brief, an equimolar concentration of 5′ fluorescein-labelled oligonucleotide Phi-W (5′ fluorescein-CGTTCTTATTACCCTTCTGAATGTCACGCTGATTATTTTGACTTTGAGCGTATCG-3′) was annealed to its complementary unlabeled oligonucleotide Phi-C to create fluorescein labeled double-stranded DNA . A 50 µl of the reaction mixture containing the different concentrations of protein (0.5–2.0 µM) was incubated with 20 nM 5′ fluorescein labeled double-stranded DNA (55 mer) in DNA-binding buffer B in the absence and presence of 1 mM ATP at 25°C for 10 min. Fluorescence signals were recorded at an excitation of 480 nm and emission at 520 nm at 25°C using FLS 980 spectrofluorimeter (Edinburgh Instruments). The data were analyzed and plotted with the curves fitting using GraphPad Prism 5.
The sedimentation analysis of ParA proteins was performed under different conditions as described in ref. [14,25]. In brief, the recombinant ParA2 and ParA3 proteins were centrifuged at 22 000×g for 15 min at 4°C to remove any aggregate. The 2 µM proteins were incubated with 0.5 pmol linear 3 kb dsDNA and 1 mM of ATP, ADP or ATP-γ-S in 30 µl for 10 min at room temperature. Similarly, a titration of DNA concentration (0–1.5 pmol) was done with both proteins in the absence and presence of 1 mM ATP or ADP only. Proteins incubated without DNA was used as a negative control. The mixtures were centrifuged at 22 000×g for 30 min at 25°C. The supernatants were removed carefully, and the pellet was resuspended in 30 µl of buffer B and mixed with 30 µl of 2× SDS-loading buffer. The mixture was heated at 95°C for 10 min, centrifuged and components were analyzed on 12% SDS–PAGE gels. Protein gel was stained with Coomassie Brilliant Blue R stain and protein band intensity was measured densitometrically by using ImageJ 2.0 software, and data were plotted as the ratio of pellet to supernatant using GraphPad Prism5.
Dynamic light scattering
Dynamic light scattering (DLS) was measured using a Horiba Scientific Nanopartica SZ-100 instrument as described previously [12,26]. In brief, all the solutions used in this study were passed through a 0.2 µm filter and proteins were centrifuged at 22 000×g for 30 min at 4°C. 2 µM proteins were incubated with 0.1 pmol 3 kb linear dsDNA in the absence and presence of 1 mM ATP or ADP. Light scattering at 90° was measured at 25°C for 30 min at a regular interval of 30 s. The data obtained as kilo counts per second (KCPS) were analyzed using in-built software (SZ-100) and plotted. The curve was smoothened using GraphPad Prism 5.
ATPase activity measurement using TLC
ATPase activity was measured as the release of [32P]-αADPs from [32P]-αATPs by TLC as described earlier [23,25]. In brief, 2 µM proteins were mixed with 30 nM [32P]-α ATP and 0.1 pmol dsDNA in different combinations in a total volume of 30 µl containing buffer B containing 2 mM Mg2+ and incubated at 37°C for 0–40 min. The reaction was stopped with 10 mM EDTA solution and 1 µl of the reaction mixture was spotted on PEI-Cellulose F+ TLC sheet. Spots were air-dried, and components were separated on a solid support in a buffer system containing 0.75 M KH2PO4/H3PO4 (pH 3.5) and autoradiogram was developed. Spot intensities of the samples were determined densitometrically using ImageJ 2.0 software, % ADP to ATP ratios were calculated and plotted using the GraphPad Prism 5 software.
ATP binding and hydrolysis by measurement of intrinsic tryptophan fluorescence
Since both ParA2 and ParA3 proteins have 2 Trp residues, the nucleotides (ADP/ATP/ATP-γ-S) binding and hydrolysis by ParA2 and ParA3 in the absence and presence of dsDNA was measured as a function of intrinsic tryptophan fluorescence of these proteins. In brief, 2 µM of ParA2 or ParA3 were incubated in buffer B containing 0.5 mM MgSO4 for 30 min in the absence and presence of 0.1 mM of ADP, ATP or ATP-γ-S and 0.1 pmol dsDNA in 30 µl reaction volume. The emission spectra of tryptophan were obtained by excitation at 295 nm and spectral scanning of emission from 315 to 401 nm at an interval of 2 nm using FLS 980 spectrofluorimeter (Edinburgh Instruments). Furthermore, 2 µM of ParA2 or ParA3 was preincubated in buffer B for 2 min, and the emission spectra were acquired at 0, 10, 15, 20 and 30 min after the addition of 0.1 mM ATP. Spectra were corrected for background and Raman scattering by subtracting buffer spectra. The obtained spectra for each time points were compared with spectra for each protein incubated with 1 mM ADP for 30 min. The data were analyzed and plotted using GraphPad Prism 5.
Protein–protein interaction studies
Protein–protein interactions were monitored using a BACTH as described in ref. [27,28]. In brief, BTH101 was co-transformed with different plasmids like pUT18A1, pUT18A2 and pUT18A3, and pKNTA1, pKNTA2 and pKNTA3 in different combinations. Empty vectors were transformed in different combinations and used as negative controls while pUTEFA and pKNTEFZ were used as positive control. Recombinant cells expressing these proteins in different combinations were spotted on LB agar plate containing 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) (40 µg/ml), IPTG (0.5 mM) and antibiotics as required. Plates were incubated at 30°C overnight and the appearance of white–blue color colonies was recorded. In parallel, the levels of β-galactosidase activity were measured from the same liquid cultures grown overnight with 0.5 mM IPTG as described earlier [27,29]. The β-galactosidase activity was calculated in Miller units as described in ref.  and plotted with the standard deviation in GraphPad Prism 5.
Interaction of proteins in surrogate E. coli was monitored by co-immunoprecipitation (co-IP). For that, the total proteins of the recombinant E. coli BTH101 cells expressing ParA1, ParA2 and ParA3 proteins of D. radiodurans on BACTH plasmids (Supplementary Table S2) in different combinations were immunoprecipitated using polyclonal antibodies against T25 tag as described in ref. . The immunoprecipitates were separated on SDS–PAGE, blotted and detected using monoclonal antibodies against T18 tag. Signals were detected using anti-mouse secondary antibodies conjugated with alkaline phosphatase using BCIP/NBT substrates (Roche Biochemical, Mannheim).
Interaction of different ParAs with ParBs was monitored in D. radiodurans by using co-IP. For that, the cell-free extracts of D. radiodurans expressing ParAs as on pV18A1, pV18A2 and pV18A3 (Supplementary Figure S1) in different combinations with ParA2 from pRADhisA2 and ParA3 from pRADhisA3 (Supplementary Figure S1) were prepared and immunoprecipitated using polyhistidine antibodies as described earlier [22,29]. Similarly, the cell-free extracts of D. radiodurans expressing different ParAs (fused to T18 tag as described above) in different combinations with different deinococcal ParBs (fused to polyhistidine tag as expressed from pRADhisB1, pRADhisB2 and pRADhisB3 were prepared and immunoprecipitated using polyhistidine antibodies as described earlier [22,29]. Immunoprecipitate was purified using Protein G Immunoprecipitation Kit (Cat. No. IP50, Sigma–Aldrich, Inc.). The immunoprecipitates were separated on SDS–PAGE, blotted onto PVDF membrane and hybridized using monoclonal antibodies against T18 tag. The hybridization signals were detected using anti-mouse secondary antibodies conjugated with alkaline phosphatase using BCIP/NBT substrates (Roche Biochemical, Mannheim) as described above.
Cell survival studies
Deinococcus radiodurans R1 and its parA mutants were treated with 6 kGy γ-radiation as well as different doses of hydrogen peroxide as described in ref. . In brief, the bacteria were grown in TGY medium with appropriate antibiotics at 32°C were washed and suspended in sterile PBS and treated with 6 kGy γ-radiation at dose rate 1.81 kGy/h (Gamma Cell 5000, 60Co, Board of Radiation and Isotope Technology, DAE, India). Gamma irradiated cells and respective controls maintained under identical conditions (SHAM) controls were washed in PBS and suspended in the fresh TGY medium. These cells were grown in TGY medium in 48-well microtiter plates in replicates at 32°C for 42 h. Optical density at 600 nm (OD600) was measured online in the Synergy H1 Hybrid multi-mode microplate reader. The growth rate was calculated from growth curve using formula , where Nt is OD600 at time t, N0 is OD600 at the start of the growth curve, r is growth rate and t is time passed) and plotted for each sample type.
For hydrogen peroxide treatment, the exponentially growing cells were exposed to different concentration of H2O2 for 30 min. The serial dilutions were made and plated on TGY agar medium containing antibiotics as required . The colony-forming units were recorded after 48 h of incubation at 32°C. The surviving fractions were expressed as the percentage of colony-forming units obtained after treatment with respect to untreated cells. We have also calculated D10 values from the survival curve for each sample and plotted.
Genome copy number determination using quantitative real-time PCR
The single and double mutant cells of similar OD at 600 nm were harvested by centrifugation and their cell number was determined using a Neubauer cell counter. The cells were washed with 70% ethanol solution and lysed in a lysis solution containing 10 mM Tris–HCl (pH 7.6), 1 mM EDTA and 4 mg/ml lysozyme at 37°C. The lysed cells were centrifuged (10 000 rpm, 5 min) to remove cell debris. The lysis efficiency was verified by plating of lysed supernatant on TYG agar plates. The integrity of genomic DNA was confirmed by agarose gel electrophoresis. The serial dilutions of cytoplasmic extract were made and 0.1 ml of it was used for further analysis of genomic copy number using quantitative real-time PCR as described in ref. . In brief, a fragment of ∼300 bps was amplified using standard PCRs with isolated genomic DNA from D. radiodurans as a template. The PCR product was gel purified and the amount of DNA was quantified by nanodrop and the concentrations of DNA molecules were calculated using the molecular mass computed with ‘oligo calc’ (www.basic.northwestern.edu/biotools). A dilution series was generated for each standard fragment and used for qPCR analysis with the dilution series. Two different genes per replicon with similar PCR efficiency were selected in D. radiodurans, namely ftsE and ftsZ for chromosome I, pprA and Dr_A0155 for chromosome II, Dr_B003 and Dr_B0104 for Mp and Dr_C001 and Dr_C018 for small plasmid (Sp) (Supplementary Table S1). The PCR efficiency of each gene was ascertained and was found to be >96% for each (data not shown). The qPCR was carried out by following the minimum information for publication of quantitative real-time PCR experiments (MIQE) guidelines using Roche Light cycler  and the cycle threshold (Ct) values were determined. Three independent biologic replicates were used for each sample. The replicon copy number is quantified by comparing the results with a dilution series of a PCR product of known concentration that is used as a standard (Supplementary Figure S4). The copy number of each replicon by both genes per cell was calculated using the cell number present at the time of cell lysis. An average of copy number reflected from two genes per replicon was represented with appropriate bio-statistical analysis.
Fluorescence microscopy of D. radiodurans and its mutants was carried out as described previously , using an Olympus IX83 inverted fluorescence microscope equipped with an Olympus DP80 CCD monochrome camera. In brief, the cells were grown until the exponential phase, harvested and washed with PBS. Cells were resuspended in PBS and stained with DAPI (4′,6-diamidino-2-phenylindole, dihydrochloride) (0.2 µg/ml) for nucleoid and Nile red (1 µg/ml) for the membrane and washed two times with PBS. These cells were mounted on glass slides coated with 0.8% agarose and imaged for DAPI and Nile red signals using DAPI and TRITC (tetramethylrhodamine isothiocyanate) channels under a fluorescence microscope, respectively. Images were aligned using an in-built software, cellSens. Each image has been presented in isolated as well as merged channels. The brightness and contrast of all images were adjusted using Adobe Photoshop 7.0. Nearly 500 cells from both wild type and mutants were examined for cell area determination using counting and measure tool of cellSens. Furthermore, scatter plot of cell area (µm2) vs sample type was plotted using GraphPad Prism 5. We performed a line scan analysis of many cells from each sample type through the cellSens software by following its manual. In line scan analysis, we scanned fluorescence intensity of DAPI and Nile red signals across a line in a cell to find the relative position of nucleoid and membrane (or septum). The percentage of cells showing nucleoids located between septum (septum trapped nucleoids) and defect on tetrads separation was calculated in each sample type and plotted using GraphPad Prism 5 software. The experiments were repeated to ensure the reproducibility and significance of these data.
The DNA-binding activity of ParAs and their polymerization on DNA was monitored by TEM on an electron microscope (Model JEOL2000FX, Japan) using previously described protocols [20,25]. In brief, 100 ng nicked circular ϕX174 RF II dsDNA was mixed with 1.5 µM ParA2 or ParA3 alone or with 1 mM ADP or ATP in buffer B containing 50 mM Tris–Cl (pH 8.5), 75 mM NaCl, 5 mM MgSO4 and 0.5 mM DTT in different combinations. Protein without DNA and high-energy phosphates were used as a control. This mixture was incubated at 37°C for 10 min before application to UV-activated carbon-coated 200 mesh copper grids. This mixture was placed on the charged side of the grid for 2 min and then washed in stage II distilled water. The grids were negatively stained with 10 µl of 2% (w/v) uranyl acetate followed by a washing in stage II distilled water. These were further blotted to dry and incubated for 1 h under vacuum before imaging. The grids were observed under JEOL 2000FX, Japan electron microscope at 100 kV and 50 000–200 000× magnification. Digital images were collected on a CCD camera as described previously .
To ensure statistical significance of data, we have performed different statistical analysis like ‘Student's t-test’ or ANOVA as required and mentioned in figure legends. Significance value (P value) obtained at 95% confidence intervals are shown as (*) for <0.05, (**) for <0.01 and (***) for <0.001.
ParAs encoded on secondary genome elements show higher sequence similarity to each other
It has previously been shown that secondary genome ParAs of D. radiodurans are different from chromosome I ParA (ParA1) of D. radiodurans . When the amino acid sequences of ParA2 and ParA3 (hereafter referred as secondary genome ParAs) were aligned with chromosomal ParAs using CLUSTALW program (Supplementary Figure S5A), ParA1 contains ∼48 amino acids extra at its N-terminus (Supplementary Figure S5A) while rest of the region of all the ParAs is conserved. The ParAs of secondary genome elements contain Walker A, Walker A′ and Walker B motifs like other known ParAs (Supplementary Figure S5A). ParA2 and ParA3 show higher sequence similarity to each other than to ParA1. For instance, the amino acid sequence of ParA1 is 25% identical with ParA2 and 23% identical with ParA3 while ParA2 and ParA3 show 42% identity with each other. Homology modeling results showed that both ParA2 and ParA3 are structurally closet to Soj protein (PDB ID; 2BEK) of T. thermophilus. The structures built on the template of Soj for both the proteins have aligned perfectly to each other as well as to the Soj of T. thermophilus (Supplementary Figure S5B,C). However, the ∼48 amino acids at N-terminal in ParA1 was extra and hanged around the remaining parts of the 3D modeled structure, which were nearly superimposable with secondary ParAs (Supplementary Figure S5D). These results suggested that ParA2 and ParA3 proteins of D. radiodurans are very similar to each other and seem to be different from ParA1, at least in silico.
ParA2 and ParA3 are characterized as the DNA-binding ATPases
The DNA-binding activity of ParA2 and ParA3 was monitored by EMSA and fluorescence anisotropy. Both the proteins showed a nearly similar binding pattern with non-specific dsDNA and the effect of ATP and ADP on DNA-binding activity was also same (Figure 1). For instance, the size of the NPC increased progressively (reflected as slower mobility) with the increase in protein concentration, which did not change in the presence of ADP. ParA3 and ParA2 binding to DNA was significantly stimulated in the presence of ATP and ATP-γ-S as compared with protein controls (Figure 1). Fluorescence anisotropy results further supported that both the proteins interact with dsDNA in almost similar fashion (Supplementary Figure S6). In the presence of ATP, however, these proteins showed a significant increase in anisotropy. Fluorescence anisotropy has been used previously in determining the nature of DNA–protein interaction .
Effect of nucleotides on DNA–protein interaction activity of ParA2 and ParA3.
Since ATP has made a significant effect on ParA2 and ParA3 interaction with DNA, and presumably, the assembly of these proteins on dsDNA, the metabolic fate of ATP by these proteins was monitored. Both these proteins could hydrolyze ATP into ADP and Pi and this activity was not enhanced in the presence of dsDNA (Figure 2). Previously, dsDNA stimulation of ATP hydrolysis of ParA1 was shown in the presence of centromere-ParB1 complex . A possibility of ATPase activity stimulation in the presence of centromere-ParB NPC cannot be ruled out. The results suggested that ParA2 and ParA3 proteins of D. radiodurans coat the DNA in the presence of ATP forming higher order structures independently of ATP hydrolysis, at least in vitro. These results together suggested that both the ParAs of secondary genome elements are DNA-binding ATPase.
Effect of dsDNA on ATPase activity of ParA2 and ParA3.
ParB2 and ParB3 showed specific interaction with the cis-element located upstream to parAB operons of the secondary genome in D. radiodurans
The recombinant ParB2 and ParB3 proteins were purified to homogeneity from E. coli (Supplementary Figure S2D). The cis-elements containing multiple direct repeats located upstream to parAB operons which are having the signature of putative centromeres were identified bioinformatically (Supplementary Figure S3) The cis-elements in chromosome II (cisII) and Mp (cisMP) were PCR amplified and interaction of the recombinant ParB2 and ParB3 was checked in vitro by EMSA. ParB2 and ParB3 showed sequence-specific interaction with cisII and cisMP sequences (Figure 3) with a Kd value of 0.41 ± 0.007 µM and 0.60 ± 0.04 µM, respectively. The ParB bound to respective cis-elements remained unaffected even in the presence of a 100-fold higher molar concentration of non-specific DNA while titrated out with 10-fold less molar concentration of specific DNA as compared with non-specific DNA (Figure 3). ParB2 or ParB3 also showed specific binding to non-cognate cis-elements like cisMP and cisII albeit with lower affinities. For instance, the Kd of ParB2 for cisMP (1.16 ± 0.02 µM) was nearly 3-fold higher than its Kd for cisII (0.41 ± 0.007 µM). Similarly, Kd of ParB3 for cisII (1.47 ± 0.03 µM) was nearly 2.5 times higher than its Kd for cisMP (0.60 ± 0.04 µM) (Figure 3). Surprisingly, both ParBs did not interact with non-specific DNA even at very high concentration (Supplementary Figure S7) indicating the motif specificity of these proteins rather than nucleotide sequence per se. These results clearly suggested that ParB2 and ParB3 bind to cis-elements present upstream to parAB operons with a strong possibility of cross-talk between segregation systems of secondary genome elements in D. radiodurans. Whether these motifs as a whole or a few repeats function as centromeres are not known and will be investigated independently.
Secondary genome encoded ParB–protein interaction with cis-elements located upstream to parAB operon in D. radiodurans.
Secondary genome ParAs produced higher order complexes on DNA, which increased further in the presence of ATP
The possibility of ParA2 and ParA3 proteins forming higher order NPC on DNA as a function of ATP was further investigated by sedimentation assay and DLS. For sedimentation analysis, the purified ParA2 and ParA3 were incubated with ATP, ADP as well as ATP-γ-S in the presence and absence of dsDNA. The amount of protein present in the pellet and supernatant was analyzed on SDS–PAGE and quantified. The amount of protein in the pellet had increased in the presence of DNA as compared with protein control, which significantly increased further in the presence of ATP as compared with ATP and DNA controls (Figure 4A,B). Interestingly, ADP did not increase the sedimentation of these proteins when compared with protein and DNA controls. To know if the ATP effect on producing bigger pellet was affected by the ATPase activity, the sedimentation of both the proteins was checked in the presence of non-hydrolyzable ATP (ATP γ-S), and the results were nearly similar to that of ATP (Supplementary Figure S8). A similar observation was made previously for ParA2 protein of Vibrio cholerae .
Effect of high-energy nucleotides on ParAs interaction with DNA.
Furthermore, the increase in the size of the NPC was also measured by DLS (Figure 4C,D). The results fully corroborated the sedimentation assay. For instance, the intensity of light scattering with both ParA2 and ParA3 and dsDNA controls was constant in the range of 900–1000 KCPS, irrespective of the presence of ATP. However, proteins in the presence of dsDNA showed a rapid increase in scattering with the KCPS increasing to more than 3000 in 5 min, which increased further with ATP (KCPS ∼6000 in 30 min) but no effect was seen with ADP (KCPS ∼3000 like DNA + protein) (Figure 4C). Since the increase in the intensity of light scattering in the presence of ATP was observed at a ratio of protein to DNA that had reached to saturation, this effect of ATP seems to be due to an increase in deposition of ParA–ATP complex over DNA. These results together suggested that ParA2 and ParA3 could bind to dsDNA, and the presence of ATP not ADP induced the interaction of ParAs with DNA which resulted in an increase in the size of NPC. Notably, both the proteins showed a similar pattern of interaction with DNA with respect to nucleotides at least in vitro.
Unlike ADP, ATP stimulated ParA2 and ParA3 binding on DNA
DNA–protein interaction in the presence of ATP and ADP was imaged by TEM. These proteins appeared as oligomers both in the presence of ATP and ADP. However, they showed nucleation on dsDNA and the density of nucleation had increased further in the presence of ATP but not with ADP (Figure 5). These results concurred the findings from sedimentation assay and DLS. Furthermore, the incubation of DNA with the protein produced the patterns of alternate dark and light spots on DNA, which further increased in the presence of ATP. In conclusion, both ParA2 and ParA3 proteins showed nearly similar results in TEM and ATP but not ADP has affected these proteins binding to dsDNA.
ATP to ADP conversion by ATPase activity of ParAs leads to a conformational change
Conformational change in the protein was monitored as the change in intrinsic fluorescence of tryptophan. Both ParA2 and ParA3 showed excitation maxima (λExt = 295) at 295 nm and emission maxima (λEm = 327) at 327 nm in the aqueous solution. Therefore, the relative fluorescence of Trp in these proteins was measured at 327 nm in the presence of different nucleotides (ATP, ADP, ATP-γ-S) and dsDNA (Figure 6A,C). Results showed a significant increase in Trp fluorescence in the presence of ATP and ADP but not with ATP-γ-S. This indicated that ParAs hydrolyze ATP into ADP and it is ADP binding with protein that led to an increase in intrinsic fluorescence of the protein perhaps through a conformational change. Such an increase in intrinsic fluorescence has been implicated in the movement of Trp residues in the hydrophobic micro-environment due to a conformational change in the proteins . Interestingly, the increase in fluorescence was higher when incubated with ADP alone, which decreased significantly in the presence of DNA. ParA incubated with ATP and DNA separately showed an increase in intrinsic fluorescence, which did not change further when both DNA and ATP were present together. This indicated that ParA protein interaction with DNA and ATP creates conformational change, perhaps required for protein polymerization on DNA. Since the presence of ADP did not increase polymerization, but showed an increase in Trp fluorescence, it indicated the possibility of a conformation change by ADP, and the ATP effects seem to be by the conversion of ATP to ADP by the ATPase activity of these proteins.
Electron microscopic imaging of ParA–protein interactions with nicked circular dsDNA.
Nucleotides effects on tryptophan fluorescence of secondary genome ParAs.
To understand if the hydrolysis of ATP to ADP by ParAs can affect protein conformation, time course kinetics of intrinsic fluorescence change due to ATP hydrolysis was monitored. We noticed that the fluorescence of ParA increases with incubation time and that reaches close to ADP control (Figure 6B,D). On the other hand, the change in the intrinsic fluorescence of ParA was not observed in the presence of ATP-γ-S (Figure 6A,C). This clearly suggests that ParA binding with ATP alone is possibly not causing a structural change in the protein rather it is the conversion of ATP to ADP. It has previously been shown that in Type I mechanism of plasmid segregation, the ParA binding to ATP and hydrolysis favors assembly and disassembly of the segregation protein complex with DNA, while ParA–ADP complex is an antagonist to this DNA–protein interaction and exist as a monomer . Therefore, the possibility of ParA–ATP and ParA–ADP ratios determining the polymerization and depolymerization dynamics of secondary ParAs cannot be ruled out and would be worth studying independently.
Para2 and ParA3 showed homotypic and heterotypic interactions
The ParA interaction was studied using BACTH  and co-IP in E. coli, co-expressing these proteins in different combinations as well as in D. radiodurans as described in the experimental procedures. Expression of these chimeras was confirmed by immunoblotting (Supplementary Figure S1). The E. coli cells co-expressing ParA1, ParA2 and ParA3 on BACTH plasmids in different combinations were screened for CyaA regulated β-galactosidase expression in E. coli BTH101 an E. coli host lacking active CyaA (see methods; ). Different ParA proteins showed homotypic interactions as indicated from the blue color colonies in spot assay and β-galactosidase activity in liquid culture (Figure 7A). In addition, ParA2 and ParA3 interacted with each other while none of them showed interaction with ParA1. These results were confirmed by co-IP in E. coli cells co-expressing these deinococcal ParAs tagged with T18 or T25 domains of CyaA in different combinations. For instance, when immunoprecipitation was carried out using T25 antibodies and the presence of interacting partner(s) tagged with T18 was detected by using T18 antibodies, all ParAs were coimmunoprecipitated with T18-fused species that included ParA2 and ParA3 but not ParA1 (Figure 7B–D).
Protein–protein interaction of ParA proteins.
In D. radiodurans, the cell-free extract of the cells co-expressing ParA1, ParA2 and ParA3 tagged with T18, and ParA2 or ParA3 with polyhistidine tag on plasmids was subjected to co-IP using anti-polyhistidine antibodies and interacting partner(s) if any are detected using T18 antibodies (Figure 8A). Nearly similar results were obtained as that of BTH and co-IP analysis in heterologous host E. coli. These results provided evidence that all the ParAs interacted homotypically while secondary genome ParAs can cross-talk to each other, but not with their homolog in the primary chromosome (Figure 8B). Thus, there seem to be structural and functional similarities among secondary genome's ParAs particularly ParA2 and ParA3 and a possibility of functional complementation of them for each other in vivo cannot be ruled out.
In vivo interactions of secondary genome encoded ParA proteins in D. radiodurans.
ParAs interact to ParBs in vivo
The interaction of ParA1, ParA2 and ParA3 with ParB1, ParB2 and ParB3 in this bacterium was checked in different combinations using co-IP assay. For that, the cell-free extract from the cells co-expressing all ParAs tagged with T18, and all ParBs with polyhistidine tag, on plasmids was subjected to co-IP using anti-polyhistidine antibodies and interacting partner(s) if any would be detected using T18 antibodies. Results showed in vivo interaction of all the ParAs with their cognate ParBs. However, the secondary genome ParAs also interacted with non-cognate secondary genome ParBs (Figure 9A–C). Interestingly, ParA1 showed interaction with ParB1 only while none of the secondary genome ParAs interacted with ParB1. These results suggested a possible functional redundancy in the segregation process of secondary genome replicons.
In vivo interactions of ParA and ParB proteins in D. radiodurans.
ΔParA2ΔParA3 mutant showed a reduction in the copy number of secondary genome elements
Since ParA plays a crucial role in genome segregation, and secondary genome encoded ParA2 and ParA3 show nearly similar functions in vitro, the possibility of these ParAs affecting genome maintenance in D. radiodurans was tested. Both single (ΔparA2 and ΔparA3) and double (ΔparA2ΔparA3) mutants of parA2 and parA3 were generated in D. radiodurans (Figure 10A) . The genome copy number of these mutants was compared with wild-type cells. The copy numbers of Chr I did not change in any of the mutants while the Chr II and Mp copy numbers were reduced in double mutant but not in the single mutants (Figure 10B). This suggested that secondary genome ParA2/ParA3 deletion that may have affected their segregation in dividing cells has also affected replication of secondary genome elements by a yet unknown mechanism(s).
Roles of secondary genome ParAs on genome copy number and extremotolerant phenotypes of D. radiodurans.
D. radiodurans cells lacking both ParA2 and ParA3 showed sensitivity to γ-radiation and H2O2
The effect of parA2 and parA3 deletion on growth response of D. radiodurans was monitored under normal and DNA damaging conditions. The double mutant showed a relatively slow growth under normal conditions as well as higher sensitivity to γ-radiation and H2O2 as compared with single mutant and wild type (Figure 10C–F). Thus, the double mutant that had reduced copy number of Chr II and Mp was also found to be more sensitive to γ-radiation and H2O2 as compared with their single mutant and wild type. This clearly indicated the role of secondary genome elements in normal growth and DNA damage tolerance in D. radiodurans. The single mutants have not shown a significant effect on growth, resistance to γ-radiation and hydrogen peroxide, and change in copy number of secondary genome elements. This further suggested a strong possibility of ParA2 and ParA3 complementing each other's role in these functions in vivo.
ΔParA2ΔParA3 double mutant showed a different morphology
Cell morphology and nucleoid phenotype of wild type, ΔparA2, ΔparA3 single mutants as well as ΔparA2ΔparA3 double mutant were monitored under the fluorescence microscope. Line scan analysis was carried out for a large number of cells in the region of interest (ROI). Figure 11 showed the line scan analysis of one ROI as the representative data for each sample. We observe that double mutant has a higher percentage of cells containing their nucleoid trapped between the septum and showing defects in the separation of the tetrads colony that usually happen to wild-type cells during normal cell division (Figure 11A,B). The average cell area of double mutant (∼8 µm2) was significantly higher than the cell areas of wild type and single mutants (∼4 µm2) (Figure 11A,C). A nearly similar phenotype of nucleoid trapped between septum has been reported in Noc null mutant of Staphylococcus aureus . Quantitatively, ∼25% cells were showing septum trapped nucleoid in double mutant as compared with less than 4% in single mutants and wild type. These finding together suggested that both ParA2 and ParA3 proteins regulate DNA translocation during cell division and they could complement the function of each other in vivo.
Microscopic studies of parAs mutants in D. radiodurans.
Genome duplication followed by its accurate segregation is pre-requisite for productive cell division in all organisms. Interdependent regulation of these processes has not been discussed in greater details. However, the involvement of TGS in the segregation of the bacterial genome has been studied to a greater extent in bacteria harboring single circular chromosome and low copy plasmids [6,7]. The genome of MGH bacteria also encodes TGS. Very limited studies have been carried out on the mechanisms of genome segregation in MGH bacteria. For instance, in V. cholerae and Burkholderia cenocepacia, the TGS of primary and secondary chromosomes have been shown to function independently [35–38]. In V. cholerae, Chr I segregation shares maximum similarities with the segregation of pB171 . In D. radiodurans, the TGS of the primary chromosome has been characterized previously .
Here, we have brought forth functional characterization of P-loop Walker ATPases encoded on secondary genome elements. We demonstrated that both ParA2 and ParA3 have higher sequence similarity at amino acid levels, show nearly similar biochemical and biophysical characteristics in vitro and could compensate for the loss of each other in vivo. We found that secondary genome ParAs and or ParBs interact to self as well as to each other, but not with primary genome ParA and/or ParB in D. radiodurans. A double mutant lacking both ParA2 and ParA3 showed slower growth and higher sensitivity to DNA damaging agents. These phenotypes could be implicated due to a reduction in copy number of secondary genome elements. Interestingly, the phenotype loss due to secondary genome ParA deletion was not compensated by the presence of primary chromosomal ParA, indicating a strong possibility of independent segregation of primary chromosome and secondary genome elements.
Earlier, the roles of ParA and ParB in multiple processes like chromosome replication, segregation and cell division have been reported in different bacteria [40–47]. The roles of ParA and ParB in the normal growth of different bacteria have been found to be different. For instance, ParA and ParB encoded on the chromosome in Caulobacter crescentus has been shown to involve in cell cycle progression and cell division, and their null mutants are lethal . C. crescentus follows asymmetric genome segregation where one copy of duplicated oriC gets traversed to opposite poles due to retraction of ParA filament upon depolymerization after it encounters to ParB-centromere complex [6,49]. On the other hand, although the loss of parAB in Pseudomonas putida, Pseudomonas aeruginosa, Streptomyces coelicolor and Bacillus subtilis has caused segregation defect, the parAB mutants do not show lethality in these bacteria [50–53]. In V. cholerae, parAB of Chr I is indispensable for normal growth . Here, we found that the double mutant of parA2 and parA3, in D. radiodurans produces phenotypes like reduced copy number of secondary genome elements, growth retardation albeit low under normal conditions, and higher sensitivity to γ-radiation and H2O2. Interestingly, we observed that the secondary genome ParA deletion does not affect Chr I copy number indicating that both primary chromosome and secondary genome elements perhaps segregate independently. Interestingly, we found that ΔparA2ΔparA3 double mutant showed a significantly increased cell size and septum trapped nucleoid phenotype in D. radiodurans. However, the loss of ParAs which presumably have arrested, DNA segregation affecting replication and thus copy number is intriguing and offers a possibility of an interdependent regulation of segregation and replication in this bacterium.
In summary, we show that the genome partitioning P-loop ATPases ParA2 and ParA3 encoded on secondary genome elements are biochemically and biophysically similar. The ParA proteins interact to its cognate ParB but not with other ParBs. ParAs of the secondary genome showed both homotypic and heterotypic interaction amongst themselves but not with Par proteins of the primary chromosome. The ParB2 and ParB3 proteins show specific interaction with both cisII and cisMP elements albeit with different affinity in D. radiodurans (Figure 3). Nearly no effect of single deletion of either ParA2 or ParA3 while a profound effect of the double mutant on copy number and response to γ-radiation and H2O2 was observed, suggesting a strong possibility of these ParAs complementing the function of each other in vivo and seems to be having no role in the maintenance of primary chromosome. While the real-time demonstration on how genome segregation arrest can affect DNA replication would be worth pursuing independently, the available data provide a strong evidence that (i) all the secondary genome ParAs interact with all the ParBs encoded on secondary genome, (ii) arrest of genome segregation (at least secondary genome) affects genome duplication indicating an interdependent regulation of these processes, and (iii) the copy number reduction in ChrII and Mp affects the wild-type response to γ-radiation and oxidative stress in D. radiodurans.
bacterial two-hybrid system
Department of Atomic Energy
dynamic light scattering
electrophoretic mobility shift assay
region of interest
tripartite genome segregation
G.K.M. conducted experiments, analyzed data and wrote the paper. S.K. conducted experiments, analyzed data and wrote the manuscript. N.N.K. and R.T. conducted TEM imaging. H.S.M. is a Principal Investigator, conceived idea, analyzed results, discussed, wrote the paper and communicated for publication.
This work was supported by institutional funds only.
The authors are thankful to Ms Shruti Mishra and Dr Anubrata Das for their technical comments on the manuscript. The authors also thank Dr Debes Ray for his help in DLS experiments, Dr Rajani Kant Chittela and Dr Alka Gupta for their help in electron microscopic studies. G.K.M. is grateful to the DAE, Government of India, for research fellowships.
The Authors declare that there are no competing interests associated with the manuscript.
A part of this work was presented in FEMS-2017 congress held at Valencia, Spain and rewarded as the ‘Early Career Scientist Award’.