Effect of a recD Mutation on DNA Damage Resistance and Transformation in Deinococcus radiodurans

The bacterium Deinococcus radiodurans is resistant to extremelyhigh levels of DNA-damaging agents such as UV light, ionizingradiation, and chemicals such as hydrogen peroxide and mitomycinC. The organism is able to repair large numbers of double-strandbreaks caused by ionizing radiation, in spite of the lack ofthe RecBCD enzyme, which is essential for double-strand DNAbreak repair in Escherichia coli and many other bacteria. TheD. radiodurans genome sequence indicates that the organism lacksrecB and recC genes, but there is a gene encoding a proteinwith significant similarity to the RecD protein of E. coli andother bacteria. We have generated D. radiodurans strains witha disruption or deletion of the recD gene. The recD mutantsare more sensitive than wild-type cells to irradiation withgamma rays and UV light and to treatment with hydrogen peroxide,but they are not sensitive to treatment with mitomycin C andmethyl methanesulfonate. The recD mutants also show greaterefficiency of transformation by exogenous homologous DNA. Theseresults are the first indication that the D. radiodurans RecDprotein has a role in DNA damage repair and/or homologous recombinationin the organism.

INTRODUCTION

Top

INTRODUCTION

The bacterium Deinococcus radiodurans is one of the most resistantto ionizing radiation (IR) and other DNA-damaging agents ofall known organisms (2, 6, 39). D. radiodurans DNA is extensivelyfragmented after treatment with IR, but the broken chromosomescan be reassembled, and the organism survives except at veryhigh radiation doses (2, 6, 49). D. radiodurans is also resistantto UV light and a variety of chemicals that damage DNA. Theability to survive extensive DNA damage is believed to be aconsequence of resistance to desiccation, a condition that alsoleads to oxidative damage to DNA (35).

Several factors are believed to contribute to the extraordinaryDNA damage resistance of D. radiodurans. The organism has multiplegenome copies (21, 22), which presumably facilitates DNA repairvia homologous recombination (49). In addition, its condensednucleoid structure may keep broken DNA ends aligned for facilerepair (31, 51). D. radiodurans also accumulates Mn²⁺ to highintracellular concentrations and has a low concentration ofinternal Fe²⁺. Manganese ion is thought to act as an antioxidantto prevent oxidative damage to cellular components (7, 8, 17).

Much recent research has been directed at identifying and studyingthe DNA repair enzymes and pathways in D. radiodurans. Repairof double-strand breaks such as those which arise from IR orduring replication of a damaged DNA template generally occursvia homologous recombination in bacteria (26-28, 38). The brokenDNA ends are first attacked by a helicase/nuclease enzyme (RecBCDin Escherichia coli and AddAB in Bacillus subtilis) (5). Thenuclease and helicase activities of the RecBCD enzyme processthe broken DNA ends, and the RecBCD enzyme loads the RecA proteinonto the resulting 3'-terminated single-stranded DNA strand.The RecA single-stranded DNA filament then binds to a homologousDNA duplex and initiates DNA strand exchange and recombination,leading ultimately to the repair of the DNA break.

Repair of IR-induced DNA damage in D. radiodurans is RecA dependent,since recA mutants are highly sensitive to IR and other DNA-damagingagents (9, 19). Interestingly, there are no identifiable homologuesof genes that would encode either the RecB or RecC proteinsin the sequenced D. radiodurans genome, nor are there genesrelated to addAB (48). However, the organism has a gene annotatedas recD, which encodes a protein with amino acid sequence similarityto the RecD protein from E. coli and other bacteria. The functionof the D. radiodurans RecD protein is unknown, since RecD proteinsin other organisms have no known function except as part ofthe RecBCD enzyme.

RecD proteins share amino acid sequence motifs that are foundin the enzymes grouped as superfamily I helicases (18). Genesthat encode proteins closely related to RecD from E. coli havebeen identified in a number of bacterial genomes and are groupedin a single group of orthologues in the Clusters of OrthologousGroups database (accession number COG0507) (see http://www.ncbi.nlm.nih.gov/COG/)(45). Further analysis of these RecD protein sequences showsthat they can be divided into two subfamilies based on the aminoacid sequences N terminal to the conserved helicase motifs (42,47). One subfamily (referred to as RecD1 in reference 42) includesRecD subunits of RecBCD enzymes. The second subfamily (RecD2in reference 42) includes the D. radiodurans RecD protein andclosely related proteins from a number of other bacteria includingB. subtilis, Lactococcus lactis, Streptococcus pyogenes, andChlamydia species (42, 47). Most of these bacteria lack a RecBCDenzyme and instead have the AddAB enzyme. D. radiodurans isunusual in that it has a RecD-like protein but neither RecBCDnor AddAB.

In previous work, we purified the D. radiodurans RecD proteinand found that it is a DNA helicase capable of efficient unwindingof short double-stranded DNA (dsDNA) substrates (47). In thiswork, we have generated recD mutations in the D. radioduranschromosome by insertional mutagenesis. The recD mutants aresensitive to irradiation with gamma rays and UV light and totreatment with hydrogen peroxide, but they are resistant tomitomycin C (MMC) and to methyl methanesulfonate (MMS). A recDmutant is also more transformable than the wild type in assaysinvolving the uptake of DNA from the medium and incorporationinto the chromosome by homologous recombination. These resultsindicate that RecD has a role in the extraordinary DNA damageresistance of D. radiodurans, and they provide the first indicationof a role for any member of this RecD protein family in DNAmetabolism.

MATERIALS AND METHODS

Top

MATERIALS AND METHODS

Bacterial strains and culture conditions. Wild-type D. radiodurans strains R1 and BAA-816 were obtainedfrom the ATCC, Manassas, VA, and from Michael Daly, UniformedServices University of the Health Sciences, respectively. D.radiodurans strain LS18 (streptomycin resistant) was a giftfrom John Battista, Louisiana State University. E. coli strainDH5 ${alpha}$ was used for routine cloning procedures. TGY medium contains0.5% tryptone, 0.3% yeast extract, and 0.1% dextrose. LB mediumcontains 1% tryptone, 0.5% yeast extract, and 1% NaCl. Antibioticswere added to D. radiodurans cultures as follows: 8 µg/mlkanamycin, 3 µg/ml chloramphenicol, and 50 µg/mlstreptomycin unless stated otherwise. Antibiotics were addedto E. coli cultures as follows: 30 µg/ml kanamycin and100 µg/ml ampicillin.

PCRs. PCR mixtures (50 µl) contained 100 ng genomic DNA templateor 0.1 ng plasmid DNA template, 0.4 µM primers, and 200µM deoxynucleoside triphosphates with Pfu DNA polymerase(Stratagene Corp.). Other conditions were as recommended bythe DNA polymerase supplier. Primer sequences are listed inTable S1 in the supplemental material. D. radiodurans chromosomalDNA was isolated essentially as described previously (13).

Southern blotting. Southern blotting was performed using genomic DNA isolated fromwild-type D. radiodurans or recD mutants as described previously(13). The DNA was digested with appropriate restriction endonucleases(see figure legends) and subjected to agarose gel electrophoresis.DNA was transferred from the gel to uncharged nitrocellulosemembranes (Optitran membrane; Schleicher & Schuell) by capillarytransfer and fixed by baking at 120°C for 30 min. Digoxigenin-labeledprobes were created with the PCR DIG Probe Synthesis kit (Roche).Hybridization was done at 42°C with shaking for 60 min afterblocking with salmon sperm DNA, followed by several washes at60°C. Detection was done using anti-digoxigenin-AP Fab fragments(Roche) and ECF substrate (catalog no. RPN5785; Amersham). Southernblots were visualized using a Storm PhosphorImager (GE Healthcare).

Creation of D. radiodurans recD mutants. Disruption of the recD gene was accomplished by insertionalmutagenesis (16). The kanamycin resistance gene and promoterregion (kan gene) were PCR amplified from the plasmid pCR-Blunt(Invitrogen Corp.) using primers that introduced SacII and PshAIsites. The D. radiodurans recD gene in plasmid pDr-RecD.ptz(47) and the kan PCR product were digested with PshAI and SacIIand ligated to create plasmid pRecDKanDis. Primers complementaryto the ends of the recD sequence (DRecD1 and DRecD2) (Fig. 1Aand see Table S1 in the supplemental material) were used toPCR amplify the disrupted recD gene (recD::kan gene) using pRecDKanDisas the template. The PCR product was used to transform D. radioduransstrain R1 to kanamycin resistance as described previously (34)(recD::kan mutant 1).

View larger version (50K):
[in this window]
[in a new window]

FIG. 1. recD::kan mutant construction and verification. (A) Map of the recD gene region in the wild-type D. radiodurans chromosome (top) and after disruption by insertion of the kanamycin resistance (kan) gene (bottom). The kan gene replaced a 48-bp PshAI-SacII fragment within the recD gene. The recD gene is transcribed left to right, while kan is transcribed right to left, in this diagram. (B) PCRs were done with primers Prom1 and DRecD2 (lanes 1 and 2) and Prom1 and KanFor (lanes 3 and 4) using genomic DNA isolated from either wild-type (wt) or recD::kan mutant (mut) cells. Primer sequences are given in Table S1 in the supplemental material, and their annealing locations are indicated in A. (C) Southern blots. Genomic DNA isolated from the recD::kan mutant was digested with EagI (lane 1), EcoRI (lane 2), ApaI (lane 3), and NcoI/PstI (lane 4) and displayed on a 0.9% agarose gel. The DNA was transferred onto a nitrocellulose membrane, hybridized to a kan gene probe (left-hand panel), stripped, and rehybridized to an undisrupted recD gene probe (right-hand panel). Molecular weights were determined from the signal due to nonspecific hybridization of the digoxigenin-labeled probe to molecular weight markers run in an adjacent lane (see Fig. S1 in the supplemental material). The expected band sizes are 2.3 and 2.5 kb (lane 1), 4.7 kb (lane 2), 4.4 kb (lane 3), and 6.1 and 2.7 kb (lane 4).

We generated several versions of the recD::kan mutation to ensurethat the phenotypic effects were due to the recD disruptionand no other mutation. First, we transferred kanamycin resistanceto fresh wild-type cells by transforming D. radiodurans strainBAA-816 with genomic DNA from recD::kan mutant 1 to make recD::kanmutant 2. We then PCR amplified the recD::kan gene using genomicDNA from recD::kan mutant 2 and again transferred it to freshBAA-816 wild-type cells (recD::kan mutant 3). The results shownin Fig. 2 were obtained with recD::kan mutant 1, while the resultsin Fig. 3 to 6 were obtained using recD::kan mutant 3. The structureof each recD::kan disruption mutant was confirmed by PCR andby Southern blotting (see Results).

View larger version (8K):
[in this window]
[in a new window]

FIG. 2. Growth of wild-type and recD::kan mutant D. radiodurans strains. Cells were grown overnight (with kanamycin for the recD::kan mutant), diluted 1:200 into fresh TGY medium with no antibiotic, and shaken at 30°C. Samples were removed at the indicated times after dilution, and the OD₆₅₀ was measured with a Perkin-Elmer Lambda Bio spectrophotometer in a 1-cm-path-length cuvette. Closed circles, wild-type strain R1; inverted open triangles, recD::kan mutant.

View larger version (8K):
[in this window]
[in a new window]

FIG. 3. Sensitivity of the recD::kan mutant to gamma and UV irradiation and to hydrogen peroxide. Closed circles, wild-type strain BAA-816; open inverted triangles, recD mutant. (A) Gamma irradiation was from a ⁶⁰Co source at 100 Gy/min on ice. (B) Cells were spread onto TGY plates and irradiated with UV light from a germicidal lamp at 90 J/m²/min. (C) Cells were treated with hydrogen peroxide (0.03%; 8.8 mM) for the indicated times and then spread onto TGY plates. The results for gamma and UV irradiation (A and B) are from two independent experiments (n = 6), while the results for hydrogen peroxide are from a single experiment (n = 3).

View larger version (11K):
[in this window]
[in a new window]

FIG. 6. Test for complementation of the recD::kan disruption mutation by a plasmid-borne recD gene. Sensitivity to irradiation with UV light was measured as described in the legend of Fig. 3 and Materials and Methods. Strains containing the plasmid were grown in the presence of chloramphenicol (3 µg/ml) but were grown without antibiotic after UV irradiation. Closed circles, wild type; open diamonds, wild-type cells transformed with plasmid pRAD1-recD; open inverted triangles, recD::kan mutant; closed squares, recD::kan transformed with pRAD1-recD. The results are representative data (n = 3) from four independent experiments.

The recD gene was also deleted completely and replaced by achloramphenicol resistance gene (cam gene). The cam gene inplasmid p11530 (29) was PCR amplified using Taq DNA polymerase(Fermentas) and ligated into the T/A cloning vector pCR2.1 (InvitrogenCorp.) to make pCR2.1-cam. A 558-bp D. radiodurans chromosomalDNA fragment upstream of the recD gene was amplified by PCRusing primers that introduced HindIII and BamHI recognitionsites at the upstream and downstream ends, respectively. A 662-bpD. radiodurans chromosomal DNA fragment downstream of the recDgene was amplified by PCR using primers that introduced XhoIand NsiI recognition sites at the upstream and downstream ends,respectively. These two PCR products were digested with HindIIIand BamHI, or with XhoI and NsiI, and ligated successively intothe pCR2.1-cam plasmid construct cut with the same pairs ofrestriction endonucleases. This D. radiodurans DNA fragmentin pCR2.1 (2,501 bp), with the cam gene replacing the recD gene,was PCR amplified, and the product was used to transform recD::kanmutant 3 (see above) to chloramphenicol resistance, generatingthe ${Delta}$ recD::cam mutant. The complete deletion of the previouslydisrupted recD gene was confirmed by the inability of the chloramphenicol-resistanttransformants to grow in the presence of kanamycin and by Southernblotting (see Results). The recD coding region is completelydeleted in this mutant, along with 104 bp upstream of the ATGstart codon and 79 bp downstream of the TAA stop codon.

DNA damage sensitivity assays. Cells for each assay were grown in TGY broth with appropriateantibiotics to mid-log phase (optical density at 600 nm [OD₆₀₀]of 0.5, corresponding to 2 x 10⁸ cells/ml). Gamma and UV irradiationand MMC sensitivity experiments were done with recD::kan mutants2 and 3 (see above), with strain BAA-816 as the correspondingwild-type strain. Sensitivity to MMS and hydrogen peroxide wastested only with recD::kan mutant 3.

Sensitivity to gamma irradiation was determined by using a NeutronProducts model 200324 ⁶⁰Co source. Cells were irradiated in1.0-ml portions at about 60 Gy/min in microcentrifuge tubeson ice. Irradiated and mock-irradiated cells were serially dilutedin triplicate in TGY medium and spread onto TGY plates withoutantibiotics. The plates were incubated at 30°C, and colonieswere counted after 2 to 3 days.

For UV sensitivity experiments, cells were serially dilutedin triplicate in TGY medium and spread onto TGY plates. Afterthe medium was absorbed, the plates were opened and exposedto UV-B light from a 15-W germicidal lamp (FG15T8; Fisher) ata rate of 90 J/m²/min (measured with a UVX Radiometer; UVP,Upland, CA). The plates were incubated at 30°C, and colonieswere counted after 2 to 3 days.

Sensitivity to MMC, MMS, or H₂O₂ was measured by incubatingcells in liquid TGY medium containing either 10 µg/mlMMC, 30 mM MMS, or 0.03% H₂O₂ or with no addition, with shakingat 30°C. At selected times, aliquots from the culture wereserially diluted in TGY medium and spread onto TGY plates. Theplates were incubated at 30°C, and colonies were countedafter 2 to 3 days.

Transformation experiments. Transformation experiments were done according to two differentprotocols. First, D. radiodurans strains were grown in 3.0 mlTGY broth to mid-log phase (OD₆₀₀ of 0.5), treated with CaCl₂,and transformed essentially as described previously (12). Competentcells (2 x 10⁸ CFU/ml) were mixed with transforming DNA fromstrain LS18 (5 µg; 10 µg/ml in the transformationmixture), serially diluted, and plated onto TGY plates to determinetotal cell numbers and onto TGY plates with streptomycin (5µg/ml) to assess the number of transformants. Plates wereincubated at 30°C and scored after 2 to 3 days. Transformationefficiency was calculated as transformants/µg DNA/totalviable cells.

Alternatively, cells were grown to mid-log phase (OD₆₀₀ of 0.5),harvested, resuspended at 6 x 10⁹ CFU/ml in 100 µl ofTGY medium supplemented with 25 mM CaCl₂ and 10% glycerol, andfrozen at –80°C, as described previously (34). Foreach experiment, cells were thawed and mixed with transformingDNA (2 µg; 20 µg/ml in the transformation mixture),and transformation was continued as described previously (34).Transformed cells were spread onto plates containing streptomycin(5 or 50 µg/ml) and scored as described above.

Complementation experiments. The D. radiodurans recD gene, along with 948 bp of DNA upstreamof the RecD start codon, was PCR amplified from genomic wild-typeDNA and ligated into BamHI-cleaved pRAD1 (36) to make pRAD1-recD.The recD gene and upstream DNA inserted into this plasmid weresequenced, and no differences from the sequence in GenBank (accessionnumber AE000513) were found. The pRAD1-recD plasmid was transformedinto D. radiodurans cells (wild-type or recD::kan mutant strains)as described above, and transformants were selected for theirability to grow on chloramphenicol and kanamycin. Sensitivityof the recD::kan (pRAD1-recD) cells to UV irradiation was testedas described above.

RESULTS

Top

RESULTS

Verification of recD::kan gene disruption. The presence of the desired recD gene disruption in the recD::kanmutant, and the absence of undisrupted wild-type recD genes,was verified by PCR and Southern blotting. Chromosomal DNA wasisolated from wild-type and kanamycin-resistant colonies, andPCRs were done using primers Prom1 and DRecD2, which shouldanneal to both the wild-type and recD::kan mutant strains (Fig.1A). The wild-type DNA gave the expected 2.2-kb fragment, whilethe recD::kan mutant gave the expected 3.2-kb fragment due tothe insertion of the kan gene (Fig. 1B). The mutant gave theexpected 1.6-kb fragment with primers Prom1 and KanFor, a primerspecific for the kan gene, while no fragment was made from wild-typeDNA with these primers (Fig. 1B).

The structure of the recD::kan mutant was confirmed by Southernblotting. Genomic DNA from the recD::kan mutant was digestedwith four different restriction endonucleases and analyzed witha kan gene probe. DNA fragments of the expected size for thedesired recD gene disruption were detected for each enzyme used(Fig. 1C). The same membrane was stripped and reprobed witha recD gene probe. The fragments detected were as expected forthe disrupted recD gene (and the same as those detected withthe kan gene probe), with no detectable undisrupted, wild-typerecD gene (Fig. 1C).

Growth of the recD::kan mutant. The recD::kan mutant strain grew at the same rate as the wildtype and reached essentially the same final cell density (Fig.2). Retention of the disrupted gene in the mutant cells, whichwere grown without kanamycin, was verified by PCR using genomicDNA isolated from the recD::kan cells at the end of the experiment(data not shown).

DNA damage sensitivity of the recD::kan mutant. The recD::kan mutant cells are more sensitive to UV and gammairradiation than the wild type (Fig. 3A and B). The mutant exhibitsa "shoulder" of resistance to moderate doses that is characteristicof wild-type D. radiodurans, but the mutant is more sensitiveto higher doses of these types of radiation than the wild type.The recD::kan mutant is also more sensitive to treatment withhydrogen peroxide than the wild type (Fig. 3C). These resultsare the first indication that any member of the RecD2 proteinsubfamily (see the introduction) (42) is involved in DNA metabolismin any bacterium in which they have been identified. In contrastto the DNA-damaging agents shown in Fig. 3, recD::kan mutantcells are no more sensitive to MMC and to MMS than are wildtype cells (Fig. 4).

View larger version (10K):
[in this window]
[in a new window]

FIG. 4. Sensitivity of the recD::kan mutant to MMC and MMS. Closed circles, wild-type strain BAA-816; open inverted triangles, recD mutant. Cells were treated with MMC (10 µg/ml) (A) or with 30 mM MMS (B) for the indicated times and then spread onto plates lacking either drug. The results for MMC are from two independent experiments (n = 6), and the results for MMS are from one experiment (n = 3).

UV sensitivity of the ${Delta}$ recD::cam deletion mutant. We attempted to generate a complete deletion of the recD genestarting with wild-type cells but were unsuccessful for unknownreasons. However, we were able to replace the disrupted recD::kangene with a cam gene and, in the process, delete the entirerecD gene sequence (see Materials and Methods). The structureof the ${Delta}$ recD::cam deletion mutation was verified by Southernblotting. A cam gene probe hybridized with chromosomal DNA fragmentsof the expected size from the deletion mutant, while no fragmentswere detected with either a kan gene or a recD gene probe (seeFig. S1 in the supplemental material) (data not shown). The ${Delta}$ recD::cam deletion mutant is as sensitive to gamma and UV irradiationas the recD::kan disruption mutant (Fig. 5).

View larger version (15K):
[in this window]
[in a new window]

FIG. 5. Sensitivity of the ${Delta}$ recD::cam deletion mutant to gamma and UV irradiation. Irradiation with (A) gamma rays and (B) UV light was done as described in Materials and Methods and the legend of Fig. 3. Closed circles, wild-type strain BAA-816; closed squares, ${Delta}$ recD::cam deletion mutant; open inverted triangles, recD::kan disruption mutant. Results for gamma irradiation are from a single experiment (n = 3), while those for UV irradiation are from two independent experiments (n = 6).

Complementation of the recD::kan mutation. We next tested whether the UV sensitivity of the recD::kan mutantstrain could be complemented by a plasmid-borne copy of thewild-type recD gene. We inserted the entire recD coding region,as well as a large portion of genomic DNA upstream of recD,into plasmid pRAD1, which replicates at low copy number in D.radiodurans cells. The recD gene in this plasmid should be expressedfrom its natural promoter, which we presume is present in theupstream DNA. The pRAD1-recD plasmid partially complements theUV sensitivity of the recD::kan mutant (Fig. 6).

We are uncertain why there is reproducibly partial complementationof the recD::kan mutation by plasmid pRAD1-recD. We consideredthe possibility that the observed phenotype of the recD::kanmutant could be the combined effect of the recD disruption mutationas well as an additional mutation in some gene other than recD.However, we believe that this is very unlikely for the followingreasons: (i) the PCR analysis and Southern blots (Fig. 1) indicatethat the kan gene insertion is in the recD gene as expected,with no secondary insertion that could contribute to the observedphenotype, and (ii) as described in Materials and Methods, wemade the recD::kan disruption mutation three times. Mutants2 and 3 were equally sensitive to UV and IR, and both were resistantto MMC (data not shown). All three mutant strains showed similarenhanced transformability (see below). These observations makeit extremely unlikely that the recD mutant phenotypes couldbe due to a second mutation elsewhere in the genome, as it isvery unlikely that the second mutation would be transferredalong with recD::kan to naïve recipient cells.

It is conceivable that the insertion into the recD gene couldhave a polar effect on neighboring genes. However, an examinationof the D. radiodurans genome sequence indicated that there areno other genes in the vicinity of recD that are likely to beinvolved in DNA metabolism. The nearest annotated open readingframe in the genome downstream of recD begins 229 bp after thepredicted recD stop codon. This open reading frame, and theone that follows, would encode proteins of unknown functionand with no significant similarity to other proteins in theGenBank database. It is thus very unlikely that the observedphenotype arises from an effect on expression of these putativeproteins. The nearest gene upstream of recD is predicted tobe transcribed in the opposite direction from recD and thuswould be unlikely to feel a polar effect of recD gene disruption.

Transformation efficiency of the recD::kan mutant. Finally, we tested the ability of recD::kan mutant cells totake up and be transformed by exogenous DNA. D. radioduransis naturally competent to take up DNA from its environment andintegrate the transforming DNA into its chromosome via homologousrecombination (41). The transformation efficiency of the recD::kanmutant was greater than that of the wild type with transformingDNA from D. radiodurans strain LS18, which carries streptomycinresistance. Using a protocol in which wild-type D. radioduransis transformed relatively efficiently (12), the recD::kan mutantwas transformed with a three- to sevenfold-greater efficiencythan the wild type (Table 1). The difference in transformationefficiency between the mutant and wild type was even greater(30- to 100-fold) when cells were treated as described previously(34) (Table 1).

View this table:
[in this window]
[in a new window]

TABLE 1. Transformation efficiency of the recD::kan mutant^a

DISCUSSION

Top

DISCUSSION

The RecD protein is well known as a subunit of the RecBCD enzymeand has been studied extensively in that enzyme from E. coli.There has been much less work on the proteins in the RecD2 subfamilythat includes homologues from D. radiodurans, B. subtilis, Chlamydia,and other bacteria (see the introduction). The results reportedin this paper indicate that RecD serves some function in DNAdamage repair and/or recombination, at least in D. radiodurans.A study similar to ours, in which a recD disruption mutationwas constructed in the D. radiodurans chromosome, has been reportedrecently (50). Those workers found, as we did, that the recDmutant is sensitive to treatment with hydrogen peroxide. However,those authors reported that the recD cells are less sensitiveto UV than we have found and that they are not sensitive toIR. The reason for this difference between the two sets of resultsis not clear at this time.

Role of RecD in DNA repair in D. radiodurans. It is not yet possible to specify a precise function for RecDin DNA damage resistance in D. radiodurans. The DNA-damagingagents used in this work give rise to a variety of forms ofdamage, and the phenotype of the recD mutant is somewhat differentfrom those of other D. radiodurans mutants studied previously,as elaborated below. Thus, much further work is required beforeRecD can be placed in a specific DNA repair pathway.

A significant effect of gamma irradiation is the productionof dsDNA breaks, and a mechanism by which D. radiodurans canrepair thousands of double-strand breaks was proposed recently(49). Double-strand ends are thought to be resected by an unknownnuclease (10, 14, 30), and the resulting single-stranded 3'ends serve to prime DNA synthesis either by annealing to homologoussingle strands or by invading homologous double-stranded regions.D. radiodurans cells with mutations in the recA and polA genesare highly sensitive to IR (9, 19, 20), indicating essentialroles for the RecA protein and DNA polymerase I in the repairprocess. However, the additional enzymes thought to be requiredfor the proposed repair pathway, including helicases that mayact at several steps (49), are not yet known.

The sensitivity of the recD mutants to gamma irradiation mayindicate a role for RecD in the repair of double-strand breaks.However, the recD mutants are much less sensitive to IR thanare the recA and polA mutants. It is possible that RecD doesact in double-strand-break repair but that another helicase(s)overlaps in function with RecD and can partially substitutefor RecD and thus lessen the effect of the recD mutation. Asecond possibility is that there is more than one pathway forthe repair of DNA damage from IR in D. radiodurans, only oneof which is affected by the recD mutation.

The fact that the recD mutant is sensitive to IR and hydrogenperoxide but not to MMC suggests that RecD might be involvedin resistance to oxidation damage rather than in the repairof double-strand breaks caused by IR. Repair by nucleotide excisionrepair of closely spaced single-strand adducts and interstrandcross-links produced by MMC gives rise to double-strand breaksin E. coli (11) and D. radiodurans (25). Consistent with this,D. radiodurans uvrA and uvrD mutants and recA and polA mutantsare quite sensitive to MMC as well as to IR. The fact that therecD mutant is not sensitive to MMC thus suggests that it maynot act in double-strand-break repair. Both IR and hydrogenperoxide lead to the formation of single-strand breaks and avariety of oxidative products from the DNA bases (15, 23). Double-strandbreaks are produced at much lower levels by hydrogen peroxidethan by IR, since peroxide-induced DNA damage at adjacent positionsin opposite strands is unlikely (15). (The specific effect ofH₂O₂ on the DNA of D. radiodurans, at the concentrations usedin our experiments, has not been examined.)

Role for RecD in transformation. The elevated level of transformation in the recD::kan mutantis similar to the effect of mutations in genes encoding severaldifferent helicases in other organisms. Interestingly, E. colirecD mutants exhibit three- to sixfold-greater levels of recombinationthan wild-type cells in some assays (4), and E. coli recD mutantsare more readily transformable than the wild type with lineardsDNA (43). The loss of the RecD subunit in E. coli disablesthe nuclease activity of RecBCD (4), and therefore, transformingDNA is not degraded in the cell. Further work will be neededto see whether RecD is involved in DNA degradation in D. radiodurans.The D. radiodurans recD mutants differ from E. coli recD mutantsin that the latter mutants are not sensitive to DNA-damagingagents such as UV light (4, 44).

Several other helicases, including UvrD, its eukaryotic homologueSrs2, and several RecQ homologues, also suppress the levelsof homologous recombination. UvrD and Srs2 can displace theRecA protein from DNA to suppress recombination (1, 3, 40, 46),while the RecQ homologues unwind recombination intermediatesand prevent their resolution to recombinant products (24, 32,33). Interestingly, a D. radiodurans uvrD mutant shows lowerlevels of recombination than does the wild type (37). It couldbe that RecD serves the antirecombinogenic function in D. radioduransthat is served by the UvrD helicase in E. coli.

ACKNOWLEDGMENTS

We thank John Battista and Michael Daly for gifts of materialsand for helpful discussions, Alex Holland and Mary Lidstromfor the gift of E. coli-D. radiodurans shuttle plasmids, AliaWeaver of the University of Maryland gamma irradiation facilityfor assistance with the ⁶⁰Co source and gamma irradiation experiments,and Wei Li for assistance with Southern blotting.

This work was supported by funds from the NIH (grant no. GM-39777).

FOOTNOTES

* Corresponding author. Mailing address: Department of Chemistry and Biochemistry, University of Maryland, College Park, MD 20742. Phone: (301) 405-1821. Fax: (301) 314-9121. E-mail: djulin{at}umd.edu

Published ahead of print on 11 May 2007.

Supplemental material for this article may be found at http://jb.asm.org/.

REFERENCES

Top