JOURNAL OF BACTERIOLOGY, Oct. 2010, p. 4954–4962 Vol. 192, No. 190021-9193/10/$12.00 doi:10.1128/JB.00667-10Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Repair of DNA Double-Strand Breaks following UV Damage in ThreeSulfolobus solfataricus Strains�†

Michael L. Rolfsmeier, Marian F. Laughery, and Cynthia A. Haseltine*School of Molecular Biosciences, Washington State University, Pullman, Washington 99164

Received 10 June 2010/Accepted 25 July 2010

DNA damage repair mechanisms have been most thoroughly explored in the eubacterial and eukaryoticbranches of life. The methods by which members of the archaeal branch repair DNA are significantly less wellunderstood but have been gaining increasing attention. In particular, the approaches employed by hyperther-mophilic archaea have been a general source of interest, since these organisms thrive under conditions thatlikely lead to constant chromosomal damage. In this work we have characterized the responses of threeSulfolobus solfataricus strains to UV-C irradiation, which often results in double-strand break formation. Weexamined S. solfataricus strain P2 obtained from two different sources and S. solfataricus strain 98/2, a popularstrain for site-directed mutation by homologous recombination. Cellular recovery, as determined by survivalcurves and the ability to return to growth after irradiation, was found to be strain specific and differeddepending on the dose applied. Chromosomal damage was directly visualized using pulsed-field gel electro-phoresis and demonstrated repair rate variations among the strains following UV-C irradiation-induceddouble-strand breaks. Several genes involved in double-strand break repair were found to be significantlyupregulated after UV-C irradiation. Transcript abundance levels and temporal expression patterns for double-strand break repair genes were also distinct for each strain, indicating that these Sulfolobus solfataricus strainshave differential responses to UV-C-induced DNA double-strand break damage.

Cells have evolved molecular mechanisms to meet the chal-lenge of maintaining genomic integrity by rapidly respondingto environmental stresses that can damage proteins and DNA.One of the most common forms of damage is caused by UVlight (UV) exposure. High-energy short-wavelength UV-Clight is absorbed directly by DNA and induces both cyclobu-tane pyrimidine dimers between adjacent thymidine or cyto-sine residues as well as pyrimidine-pyrimidone photoproductsbetween adjacent pyrimidine residues. Mechanisms for repairof these lesions appear to be present in all organisms and arethought to occur through either light-independent nucleotideexcision repair (NER) or light-dependent photoreactivationusing photolyases (for reviews, see references 33 and 34).UV-C irradiation also causes the production of reactive oxygenspecies, which can result in DNA double-strand breaks (DSBs)(6, 44). Our primary understanding for repair of DSBs hascome from studies focused primarily on bacteria and eu-karyotes. In Escherichia coli, these breaks are repaired throughthe action of the RecA protein, which assists in recombina-tional repair of single-strand regions produced through repli-cation fork arrest at UV lesions and in DSB repair by extendedsynthesis-dependent strand annealing (SDSA) and homolo-gous recombination (9, 19). Eukaryotes employ nonhomolo-gous end joining as well as DSB repair by SDSA and homol-ogous recombination mechanisms to repair these breaks (forrecent reviews, see references 20 and 21).

The archaeal domain is distinct from the bacterial and eu-karyotic branches of life and includes organisms that inhabitextreme environmental habitats that might be expected to ac-celerate spontaneous DNA damage. Interest in these microbeshas grown steadily since their identification, as species livingunder such conditions are likely to have robust mechanisms tocontend with continual genomic insults. In particular, DNAdamage repair and genomic stability have been the focus ofseveral recent studies in the archaea. Two genome-wide effortswith Halobacterium strain NRC-1 have focused on transcrip-tion following UV irradiation and demonstrated that at highdoses, photoreactivation is a major mechanism for repair,while at lower doses, homologous recombination is involved incorrection of DNA damage (4, 24). Sulfolobus spp. occupyterrestrial hot springs with temperatures ranging from 70 to85°°C, an environment that is likely to lead to DNA damage bygeneration of reactive oxygen species, hydrolytic deaminationof nucleotide bases, and UV exposure. Studies of Sulfolobusacidocaldarius have shown that cellular sensitivity to UV irra-diation and spontaneous mutation rates are similar to those ofE. coli, indicating efficient DNA damage repair in these hyper-thermophiles (15, 46). Additionally, photoreactivation-basedrepair and genetic marker exchange were apparent during ex-posure to visible light, suggesting that DNA lesions and DSBsstimulate exchange (11, 36). Genome-wide transcriptional ex-amination of the UV damage response in Sulfolobus solfatari-cus strains PH1 and P2 and in S. acidocaldarius showed anapparent increase of no more than 2-fold in expression ofDNA repair genes, suggesting a potential lack of repair path-way inducibility (11, 12).

S. solfataricus has emerged as a popular model system forhyperthermophilic archaea, since it is relatively easy to grow inthe laboratory in liquid medium as well as on plates. S. solfa-

* Corresponding author. Mailing address: Washington State Univer-sity, School of Molecular Biosciences, 137 Biotechnology Life Sci-ences, Pullman, WA 99164. Phone: (509) 335-6148. Fax: (509) 335-1907. E-mail: chaseltine@@wsu.edu.

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

� Published ahead of print on 30 July 2010.

4954

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

taricus strain 98/2 has become especially useful, since a geneticselection system utilizing lactose and the lacS gene has beendeveloped and directed gene replacement through homolo-gous recombination has been demonstrated (35, 47). Curi-ously, while equivalent lacS mutant backgrounds exist in otherstrains (specifically S. solfataricus strains PH1 and P2), effortsat in vivo-directed gene replacement have thus far been suc-cessful only in strain 98/2 (8, 10, 16, 18, 23, 45). The apparentinability of some S. solfataricus strains to perform homologousrecombination-based gene replacement suggests a fundamen-tal difference in the repair mechanisms employed from strainto strain and prompted our comparison of cellular DNA dam-age repair. We specifically examined three S. solfataricusstrains, strain P2 obtained from two different sources andstrain 98/2. Here we describe the differential ability of thestrains to survive UV-C irradiation and the rates at which theyreturn to growth. Additionally, we describe the rate of chro-mosome repair following UV-C-induced DSB formation andthe transcriptional response of DNA damage repair genes.

MATERIALS AND METHODS

Cell cultivation and UV exposure. For all three S. solfataricus strains, the cellswere grown to an optical density at 540 nm of 0.2 to 0.4 in basal salts medium asdescribed previously (30) containing 0.1% (wt/vol) sucrose and 0.1% (wt/vol)tryptone (ST medium) or 0.2% (wt/vol) glucose and 0.2% (wt/vol) tryptone (GTmedium) prior to UV exposure. The cells were exposed to 0, 100, 200, or 300mJ/cm2 of UV-C (UVC-508; Ultra-Lum Inc., Claremont, CA). Irradiated cellswere then added to prewarmed ST medium and cultivated in the dark at 80°°Cwith shaking. For viable-cell determination, irradiated and control samples weregrown in GT medium, diluted using the same medium, and plated in the dark on0.8% (wt/vol) Gelrite (Kelco) GT plates at a pH of 3.0. The plates were incu-bated at 80°°C in a humid chamber for approximately 5 days, and colonies werecounted. Growth rates were determined by spectrophotometric analysis at 540nm of cultures grown in liquid, and generation times were calculated using Prizm4.0 software using a minimum of seven independent cultures. Representativereturns to growth curves were determined by analysis of three separate growthcycles.

Pulsed-field gel electrophoresis. The cells were grown to early log phase(optical density at 540 nm of 0.2 to 0.4) and exposed to a UV dose of 100 mJ/cm2

and then returned to growth. At the indicated times, 4 �� 109 cells wereremoved and collected by centrifugation at 4°°C. The cells were resuspended in0.5 volume of 1�� medium salts, warmed to 50°°C, and added to 0.5 volume ofmelted 2% agarose (InCert). Plugs were cast using a mold (Bio-Rad) and solid-ified at 4°°C. Solidified plugs were treated in 10 mM Tris-Cl (pH 8.0), 1 mMEDTA plus 1% N-lauroyl sarcosine sodium salt and 1 mg/ml proteinase K(Bioline) overnight at 50°°C. This was followed by a second overnight incubationin proteinase K at 50°°C. The plugs were then washed 3 times for 30 min eachtime in ET buffer (20 mM Tris-Cl [pH 8.0], 50 mM EDTA) and then washed 3times for 30 min each in 0.1�� ET buffer. After the agarose plugs were washedwith TE buffer (10 mM Tris-Cl [pH 8.0], 1 mM EDTA) two times for 10 min eachtime, they were then subjected to digestion by SfiI (NEB) for 4 h using themanufacturer’s recommended conditions in a total volume of 250 ��l. DNA wasanalyzed using a Bio-Rad CHEF-DR III system with a buffer temperature of14°°C. The gels were made of 1% agarose, 0.5�� TBE (9 mM Tris, 9 mM boricacid, 0.2 mM EDTA) and were run for 24 h at 5.5 V/cm using a 120o includedangle with switch times of 60 to 120 s. The gels were subsequently stained withethidium bromide and photographed with a Geneflash gel documentation system(Syngene). Chromosome repair was measured using the GeneTools (Syngene)quantification software. The zero-hour time point (untreated sample) was des-ignated 100% repair. The values for 24-h time points were normalized to thevalues for untreated samples.

PCRs. Standard PCRs were performed in 1�� ThermoPol buffer (New En-gland BioLabs [NEB]) with 200 pmol of the appropriate primer (see Table S1 inthe supplemental material), 200 ��M dinucleotide triphosphates (NEB), and 2.5U Taq (NEB). PCRs for gene sequencing were performed in 1�� ExTaq buffer(Takara Bio) with 200 pmol of the appropriate primer, 200 ��M dinucleotidetriphosphates (NEB), and 1.25 U ExTaq (Takara Bio).

Gene sequencing and sequence analysis. Gene sequencing was performed byAmplicon Express (Pullman, WA). DNA sequences were converted to proteinsequence using the Translate�� program (SeqWeb GCG Wisconsin SequenceAnalysis Package). Protein alignments were produced using the Pretty program(SeqWeb GCG Wisconsin Sequence Analysis Package), and differing residueswere highlighted by hand. Gene sequences were found to match genome infor-mation available at the NCBI with the exception of the S. solfataricus strain P2-Aral2 sequence, which has been deposited in GenBank under accession numberHM462249. BLAST analysis for the novel insertion in the strain 98/2 rad54 genewas performed using the 327-bp insertion sequence as a query with nucleotideBLAST and the Sulfolobus genome information available at the NCBI website.

RNA isolation and cDNA preparation. Samples in 6- to 10-ml volumes wereremoved from cultures grown as described above at the times indicated. Cellswere collected by centrifugation at 10,500 �� g. RNA was isolated using theRiboPure bacteria kit (Ambion) and the manufacturer’s protocol. Contaminat-ing DNA was removed from RNA samples using the DNA-free kit (Ambion)following the manufacturer’s protocol. Resulting nucleic acid samples werequantified at 260 nm using a Coulter Beckman DU-800 spectrophotometer.cDNA was prepared from DNA-free RNA using the RevertAid first-strandcDNA synthesis kit (Fermentas) with the random hexamer primer and by usingthe manufacturer’s recommendations.

qRT-PCR. Quantitative reverse transcriptase PCR (qRT-PCR) was performedwith 10 ng template cDNA, Fast SYBR green master mix (Applied Biosystems),and 180 pmol of each primer (see Table S1 in the supplemental material) in a20-��l reaction volume. qRT-PCR was performed using an Applied Biosystems7500 Fast Real-Time PCR system using the Fast, ����CT, SYBR green settings.cDNA template was validated over 4 orders of magnitude, following the manu-facturer’s protocol. Transcript levels were normalized to the S. solfataricus 23SrRNA gene (annotated as Ssor04 at http://www-archbac.u-psud.fr/projects/sulfolobus/). Normalized gene expression was calculated using the comparativethreshold cycle (CT) method, also known as the ����CT method (22, 37).����CT calculations were performed using the equations provided in the userbulletin for the ABI Prism 7700 sequence detection system (2a).

RESULTS

Cellular response following UV-C radiation exposure wasvariable and dependent on the source of the strain. We exam-ined three separate sources of S. solfataricus strains for cellularresistance to UV-C irradiation (see Table S1 in the supple-mental material). Strain P2-A was purchased directly from theAmerican Type Culture Collection (ATCC) and is designatedS. solfataricus P2 as deposited by Wolfram Zillig into the Deut-sche Sammlung von Mikroorganismen und ZellkulturenGmbH (DSMZ). Strain P2-B was acquired from YvanZivanovic (Universite Paris-Sud) and was used in the S. solfa-taricus P2 genome sequencing project (40). Strain 98/2 wasprovided by Paul Blum (University of Nebraska––Lincoln) andhas been the subject of recent genomic sequencing efforts(GenBank accession no. ACUK00000000 and CP001800.1).We found that the strains showed differing responses to UVdamage that were dependent on the dose applied. Cellularsurvival was determined by colony formation on solid mediumand is shown in Fig. 1. At the lowest UV dose of 100 mJ/cm2,all three strains demonstrated resistance, with a survival rate ashigh as 23% for strain 98/2 and survival rates of 11% and 13%for strains P2-A and P2-B, respectively. At a UV dose of 200mJ/cm2, the P2-A and 98/2 strains showed nearly identicalsensitivities, with cellular survival between 1.7 and 2.2%. TheP2-B strain was more sensitive to UV damage at this higherdose, displaying 5- to 7-fold-lower survival than the other twostrains. At the highest UV dose of 300 mJ/cm2, strain 98/2 wasthe most sensitive strain, exhibiting only 0.004% survival com-pared to 0.18% and 0.08% survival for strains P2-A and P2-B,respectively.

To better assess the ability of S. solfataricus to recover from

VOL. 192, 2010 SULFOLOBUS UV-INDUCED DOUBLE-STRAND BREAK REPAIR 4955

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

UV-C damage, we examined the return of the cells to growthin liquid medium. Each strain was exposed to various doses ofUV light, and growth was monitored by measuring the opticaldensity (Fig. 2). In the absence of UV damage, the strainsshowed slightly different generation times in rich medium.Strain 98/2 grew the most rapidly, with a doubling time ofapproximately 5.9 h �� 0.7 h. The next fastest growth rate wasseen with strain P2-A, with a doubling time of approximately7.5 �� 0.6 h. Strain P2-B grew more slowly than either strain98/2 or P2-A and had a doubling time of approximately 8.3 ��1.3 h. After exposure to 100 mJ/cm2 UV-C irradiation, all threestrains began recovery in the first 4 to 5 h, although at a slightlydiminished growth rate. Strain 98/2 recovered the most quicklyfrom a low UV-C dose, within 2 h following exposure to 100mJ/cm2, which was not unexpected based on observed highersurvival rates on solid medium as shown in Fig. 1. Both P2strains restarted growth approximately 4 to 6 h earlier thanstrain 98/2 following exposure to 200 mJ/cm2. After applicationof the highest UV-C dose of 300 mJ/cm2, both P2 strains (P2-Aand P2-B) demonstrated recovery as measured by growth by30 h after exposure. Strain 98/2, however, did not recover fromthis exposure within the first 100 h. Although the two higherdoses of UV-C irradiation (200 and 300 mJ/cm2) requiredsignificantly longer time periods for restarted growth to bedetected than the 100-mJ/cm2 dose, the growth rates achievedupon regrowth were similar to those observed in the absence ofUV-C exposure. Taken together with the survival curve data,these results suggest that there may be a differential responseto UV damage in the three strains.

Chromosomal repair rates differed between the strains un-der identical conditions. It has been previously reported thatUV treatment of S. solfataricus strain PH1 results in double-strand break formation, with breaks likely arising from thepresence of lesions that remain unrepaired as the cells progressthrough replication (11). To evaluate the ability of each strainto repair damage, we used pulsed-field gel electrophoresis(PFGE) to directly visualize the state of the chromosome forboth P-2 strains and strain 98/2 (Fig. 3). To date, all publishedSulfolobus genomes have been reported to be circular andconsist of close to 3,000 genes (7, 17, 40). S. solfataricus strainP2-B was the subject of the first Sulfolobus genome sequencing

project and has a circular genome of 2.99 Mbp with 3,287 openreading frames and a GC content of 35.8% (40). The other S.solfataricus P2 strain, P2-A, is annotated as the same strain asP2-B at the ATCC and is expected to have a genome of similarcomposition. While not yet published, publicly available se-quence information for strain 98/2 has recently been found tohave a 2.68-Mbp genome of 35.4% GC with 3,055 predictedopen reading frames (GenBank accession no. ACUK00000000and CP001800.1).

To better visualize chromosome repair by PFGE, genomicDNA was subjected to restriction endonuclease digestion usingSfiI. This enzyme was anticipated to cut rarely in the S. solfa-taricus genome due to a 13-bp recognition sequence that re-quires a 4-bp GC sequence at each end. We chose the 100-mJ/cm2 irradiation dose to assess chromosome damage and repair,since cellular survival was found to be at least 10% for each ofthe strains under these conditions (as shown in Fig. 1). In theabsence of UV-C irradiation, indicated as the time zero lane

FIG. 1. Representative survival curves for three S. solfataricusstrains following exposure to UV-C radiation. Survival of strains P2-A,P2-B, and 98/2 is shown. Values are the means �� standard deviations(error bars) for triplicate experiments (plated in duplicate). Many ofthe error bars for survival data were smaller than the symbol and aretherefore not visible in the figure.

FIG. 2. Representative growth curves showing the effect of irradi-ation on return to growth. Cells were treated with various doses ofUV-C irradiation and then returned to growth in the dark. Culturegrowth was measured by optical density at 540 nm (OD540). StrainsP2-A (A), P2-B (B), and 98/2 (C) were examined. The UV doses were0, 100, 200, and 300 mJ/cm2.

4956 ROLFSMEIER ET AL. J. BACTERIOL.

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

for each gel, from the digested chromosome, two distinct bandsfor strain P2-A were observed (Fig. 3A). Four distinct digestfragments were apparent for strain P2-B, while three frag-ments were seen for strain 98/2 (Fig. 3B and C, respectively).The different digest patterns suggest that there is some se-quence disparity between the two P2 strains that results in thepresence of additional or differentially located restriction sites.Within 1 h of UV-C irradiation, double-strand break forma-

tion was apparent for each of the strains and manifested aslower-molecular-weight smears seen toward the bottom of thegels that is concurrent with the loss of the digest fragmentsobserved before irradiation. The state of the chromosome wasassessed over a period of 24 h following the return to growth,and during this time, the low-molecular-weight DNA smearsresolved into the digest fragments that are indicative of anintact chromosome. Among the three strains, strain P2-B ap-peared to have the most accumulated breaks as indicated bythe intensity of the lower-molecular-weight smears (Fig. 3B).This strain demonstrated an approximate 50% chromosomalrepair rate by 14 h after returning to growth, while the othertwo strains achieved approximately 50% repair several hourssooner; for strain P2-A, this occurred between 7 and 9 h, andfor strain 98/2, this level of repair was apparent at 9 h (Fig.3D). Overall, strain 98/2 demonstrated the slowest averagerecovery from UV-C-induced double-strand breaks as shownby quantification of restriction fragments at the 2-, 3-, 5-, and7-h time points compared to the P2 strains (Fig. 3D).

Genes implicated in DSB repair were transcriptionally up-regulated in response to UV-C irradiation. Several genes havebeen identified in Sulfolobus species that are expected to beinvolved in DSB repair in vivo. Biochemical studies have elu-cidated the activities of the RadA protein (a homologue ofeukaryotic Rad51 and bacterial RecA proteins) from both S.solfataricus and Sulfolobus tokodaii and the Rad54 proteinfrom S. solfataricus and suggest that these proteins are involvedin the repair of DNA DSBs through homologous recombina-tion (3, 14, 31, 38, 39, 42, 48). The Mre11 protein from Sul-folobus acidocaldarius has been found to be associated withDNA following damage and directly interacts with Rad50, im-plying a role in break repair for both of these proteins (28).RadA paralogues identified in S. solfataricus and S. tokodaiiwere initially implicated in DSB repair by virtue of their ho-mology to the RadA protein, but subsequent efforts indicatethat at least some of these proteins are directly involved in therepair process (1, 25, 41, 42). These paralogue proteins havepreviously been described as members of the newly designatedarchaeal RadC (aRadC) family (13), but no gene names havebeen assigned to the individual open reading frames (ORFs)apart from the corresponding P2 genome project number inthe case of the three paralogues found in S. solfataricus (1, 25).Since our study includes examination of the paralogue genes inmore than one S. solfataricus strain and use of a single ORFdesignation relating to just one genome project could be con-fusing, for simplicity here we have named the genes encodingthe paralogues ral genes (for radA-like), where the publishedP2 ORF Sso2452 is ral1, ORF Sso0777 is ral2, and ORFSso1861 is ral3. The corresponding genes in strains P2-A and98/2 are named in the same manner.

To examine the possibility that the expression of these genescould be involved in the DSB repair observed by PFGE, wefirst verified the presence of each gene in the three S. solfa-taricus strains. PCR primers were designed based on P2 nucle-otide sequences from GenBank (accession no. AE006641) andused to amplify the putative DSB repair-associated genes fromall three strains (Fig. 4). We found that each strain carries theradA gene, all three ral genes (ral1, ral2, and ral3) and rad50,mre11, and rad54. Every gene was of the expected size based onthe published S. solfataricus P2 genome, with the exception of

FIG. 3. Repair of S. solfataricus chromosomal DNA followingUV-C irradiation at 100 mJ/cm2 as visualized by pulsed-field gel elec-trophoresis. Representative gels are shown for each of the three strains(strain P2-A [A], P2-B [B], and 98/2 [C]) where each lane containsDNA extracted from 5 �� 108 cells as visualized by staining withethidium bromide. Chromosomal DNA was digested with SfiI, and therestriction fragments are indicated by the arrows, while fragmentsresulting from DSB formation are labeled broken. The length of timeafter irradiation is shown in hours. The rate of repair for each strainwas determined by quantitation of the largest digest fragment and isshown as the average of four to seven replicate experiments in panel D.

VOL. 192, 2010 SULFOLOBUS UV-INDUCED DOUBLE-STRAND BREAK REPAIR 4957

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

the rad54 gene (Fig. 4A). rad54 PCR amplification productswere of different sizes in each of the three strains, with thelargest PCR amplicon found in strain P2-B and the smallest instrain P2-A. Sequencing of the rad54 amplicons showed thepresence of insertions that resulted in larger DNA productsfollowing PCR amplification (Fig. 4B). The rad54 gene fromstrain P2-A appears to be the intact wild-type sequence and is2,721 bp in length. The insertion in the rad54 gene from strainP2-B is identical to that reported in the published S. solfatari-cus P2 genome sequence (14, 40) and is the 966-bp insertionelement ISC1173. The insertion in the rad54 gene of strain 98/2is smaller than that in strain P2-B and in a different location.This insertion is 327 bp in length and when used as a query inBLAST analysis, is more than 90% identical to 7 sequences ofidentical size found across the genome of strain 98/2. Each ofthese 7 additional sequences is located in a probable intergenic

region, suggesting that the insertion is likely noncoding andcould be derived from a insertion element. The insertions inthe rad54 genes of strains P2-B and 98/2 result in prematurestop codons that are likely to preclude production of full-length protein in vivo.

Additional sequence analysis revealed that radA, ral3, rad50,and mre11 are identical in all three strains examined, but thereare nucleotide differences in the ral1 and ral2 genes. For com-parative purposes, we determined gene sequence identity usingthe P2-B sequence as a reference, as it currently serves as thesequenced type strain for S. solfataricus. Nucleotide sequencesfor ral1 were 100% identical in strains P2-A and 98/2, but theP2-B sequence differed at 97 specific residues and was only87.7% identical; at the protein sequence level, the Ral1 pro-teins of strains P2-A and 98/2 were 100% identical to eachother but only 95% identical to the Ral1 protein of strain P2-B(see Fig. S1 in the supplemental material). ral2 gene sequencesfor strains P2-A and 98/2 were 99.4% identical to each other,differing at just 3 nucleotides. Following gene translation, theresulting protein sequences were 99% identical to each other.In comparison with ral2 sequence from strain P2-B, ral2 instrain P2-A was 89.2% identical and in strain 98/2 was 88.6%identical. Translation of the ral2 gene from strain P2-A re-sulted in a sequence that is 93% identical to that of strain P2-B,while the Ral2 protein sequence from strain 98/2 was only 91%identical to that from strain P2-B (see Fig. S2 in the supple-mental material).

Using quantitative real-time reverse transcriptase PCR andthe Applied Biosystems 7500 Fast Real-Time PCR detectionsystem, we determined the levels of expression for radA, ral1,ral2, ral3, rad50, mre11, and rad54 in all three S. solfataricusstrains following exposure to 100 mJ/cm2 of UV-C irradiation(Fig. 5). We found that expression for this subset of genes afterDNA damage was different for each strain both over time andin intensity. Strain P2-A showed the earliest transcriptionalresponse, with the highest level of expression of radA, ral1,ral2, ral3, and rad54 occurring 30 min after exposure (Fig. 5).We observed an increase of more than 9-fold in the expressionof radA in strain P2-A at the 30-min time point, with expressiondecreasing over subsequent time points, followed by a mildincrease in expression of more than 2-fold 120 min after UV-Cexposure (Fig. 5A). Similar increases in transcription andequivalent timing were observed for each of the ral transcriptsas well as the rad54 transcript, with the highest level of expres-sion occurring at 30 min, which then tapered off over theremainder of the time course. Strain P2-B showed comparableexpression level increases for each of the genes, but the timingof increased expression was different from that of strain P2-A(Fig. 5C). Levels of transcript production were found to beginto increase 45 min after exposure, with peak expression for theradA, ral, and rad54 transcripts occurring 90 min after damage.The radA paralogues and rad54 were also induced for expres-sion, with upregulation ranging between 6- and 10-fold for theral genes and more than 4-fold induction for rad54. Expressionlevels for rad50 and mre11 transcripts were much higher thanthose observed for the radA, ral, and rad54 genes for strainsP2-A and P2-B (Fig. 5B and D). More than 20-fold inductionfor rad50 and mre11 was seen in both cases but at differingtimes. Significant induction of transcription from these genesoccurred 120 and 90 min after damage for strains P2-A and

FIG. 4. PCR amplification of genes likely to be involved in double-strand break repair. (A) The genes encoding RadA, Ral1, Ral2, Ral3,Rad50, Mre11, and Rad54 were amplified from genomic DNA isolatedfrom strains P2-A, P2-B, and 98/2 as indicated. The products wereelectrophoresed using gels of 0.8% agarose and 1�� TBE and visu-alized by ethidium bromide staining. (B) Schematic representation ofthe insertions present in the rad54 gene in strains P2-B and 98/2.Uninterrupted rad54 from strain P2-A is shown at the top. The inser-tion in rad54 from strain P2-B begins at nucleotide 2360 and is 966 bpin length, while the rad54 insertion in strain 98/2 begins at nucleotide1760 and is 327 bp in length.

4958 ROLFSMEIER ET AL. J. BACTERIOL.

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

P2-B, respectively, with lower levels of transcript productionduring the remainder of the time course.

Strain 98/2 showed the weakest transcriptional response forall the genes examined relative to the 23S rRNA gene, withlevels of radA transcript never exceeding a 3-fold increase atany point over the time course (Fig. 5E). The ral and rad54genes from strain 98/2 were similarly not highly induced fortranscription over the period examined following damage. Forrad50 and mre11, strain 98/2 again showed a reduced transcrip-

tional response to damage, with induction of these genes at amuch lower level than that seen for the P2 strains, increasing toa maximum of only 10-fold (Fig. 5F). The overall reducedtranscriptional response of strain 98/2 suggested that strain98/2 reacts to UV-C DNA damage by upregulating DSB repairgenes at a low level over an extended period of time or that thisstrain utilizes an alternate pathway for repair. An alternatepathway could involve nucleotide excision repair, and recently,nucleotide excision-related proteins have been identified and

FIG. 5. Quantitative real-time PCR detection of transcripts from likely DSB repair genes following UV-C irradiation. Cells from each of thethree strains were exposed to 100 mJ/cm2 of UV-C irradiation, and RNA samples were isolated at the times indicated. Normalized expression levelsfor radA, ral1, ral2, ral3, and rad54 transcripts from strains P2-A, P2-B, and 98/2 are shown in panels A, C, and E, respectively. Normalizedtranscript expression levels for rad50 and mre11 are represented for strains P2-A, P2-B, and 98/2 in panels B, D, and F, respectively. Experimentalresults are shown as the averages plus standard deviations (error bars) from a minimum of three replicates.

VOL. 192, 2010 SULFOLOBUS UV-INDUCED DOUBLE-STRAND BREAK REPAIR 4959

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

characterized for S. solfataricus. Specifically, the SulfolobusXPB-Bax1 protein complex is likely equivalent to eukaryoticXPG and has been shown to function as a helicase-nucleaseresponsible for unwinding and cleaving DNA substrates thatmimic those likely to be abundant during nucleotide excisionrepair processes (29, 32, 43). We therefore examined geneexpression levels for bax1 and xpd2 over a similar time coursefor each of the three strains (Fig. 6). Strain P2-A showed theearliest transcriptional induction for these genes, with expres-sion levels reaching more than 6.5-fold 30 min after damage,while strain P2-B transcription levels reached a maximum ofmore than 5-fold after 45 min (Fig. 6A and B). Strain 98/2 didnot demonstrate significant induction of either bax1 or xpd2following UV-C irradiation but instead displayed a low-level

(no more than 2-fold) persistent increase in transcript produc-tion over the time course (Fig. 6C).

DISCUSSION

This work demonstrates that three S. solfataricus strainshave marked physiological differences in their response toDNA damage induced by exposure to UV-C irradiation. Oursurvival curves indicated that strain 98/2 was the most vulner-able to high UV-C doses, but it had slightly more resistance tolower doses than strain P2-A or P2-B did. To our surprise, twoP2 strains from two different sources (one was purchased fromATCC, and another was obtained directly from the sequencingproject) were strikingly different. Survival varied followingUV-C exposure and was dependent on dose, with strain P2-Bdisplaying more overall sensitivity to damage than strain P2-A.The differences between these two strains were again apparentin the rate at which chromosomal DSB repair occurs. Whilestrain P2-B shows early immediate repair, this is followed by aprolonged delay, and 50% genomic repair is not achieved untilseveral hours after the same point is reached in strain P2-A.Sequence variations were also found between the two P2strains; several of the genes we examined were identical, butnucleotide disparities were directly apparent in the ral1 andral2 genes and the P2-B rad54 gene is interrupted by an inser-tion sequence. Variability in sequence was also indirectly indi-cated through the differential SfiI genome digestion patterns.Although we cannot, in the work described here, ascertain theextent of sequence variation between the two strains identifiedas P2, the data do suggest that the strains are not interchange-able. Additional sequencing efforts will be necessary to furtherdelineate the differences between the two strains.

Examination of the cellular transcriptional response toDSBs caused by UV-C damage indicates that the timing andintensity of gene expression are dependent on the strain. Strain98/2 responds to damage with a somewhat consistent low-levelgene expression pattern that begins within the first 30 min. Incontrast, strains P2-A and P2-B have a more robust responseand strongly upregulate radA, ral1, ral2, ral3, rad54, rad50, andmre11. We were again able to discern differences in the ele-vated transcriptional responses of the two P2 strains. StrainP2-A responds to damage more quickly than strain P2-B doesand also modulates gene expression more tightly, exhibiting aburst of expression between 30 and 45 min after damage. Incontrast, strain P2-B initiates transcriptional upregulationlater, 45 min after damage, but persists in increased transcriptproduction that peaks 90 min after damage. Overall, we foundthat despite the nuances in timing and duration, expression ofDSB repair genes was clearly temporal in nature, and theresponse time was rapid for the P2 strains. Over the timecourse examined, strain 98/2 never induced expression of theDSB repair genes to the level seen for the P2 strains andappeared to respond to UV-C DNA break damage by mod-estly upregulating these genes but maintaining this level ofexpression for a prolonged time.

Our transcriptional studies have demonstrated for the firsttime that there is a measurable regulatory response to DSBdamage arising from UV-C irradiation in S. solfataricus thatinvolves multiple genes. Prior to this work, the only report ofincreased expression of DSB-related genes following UV irra-

FIG. 6. Quantitative real-time PCR detection of transcripts frombax1 and xpd2 following UV-C irradiation. Cells from each of the threestrains were exposed to 100 mJ/cm2 of UV-C irradiation, and RNAsamples were isolated at the times indicated. Normalized expressionlevels for bax1 and xpd2 transcripts from strains P2-A, P2-B, and 98/2are shown in panels A, B, and C, respectively.

4960 ROLFSMEIER ET AL. J. BACTERIOL.

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

diation showed moderate transcriptional elevation (between 2-and 3-fold) for rad50 and mre11 in S. solfataricus strain PH1(11). Each of the strains we examined in this work increaseexpression of genes expected to be involved in DSB repair tovarious levels in response to damage incurred by UV-C. Theinduction of rad54 and the ral genes observed here is the firstindication of an in vivo role for Rad54 and Ral proteins inDNA damage repair in Sulfolobus. Additionally, we found thatthe rad54 gene is not essential for S. solfataricus, as the pres-ence of insertions in the coding sequence were well toleratedby the cell. There is no apparent correlation between the stateof the rad54 gene and the ability of cells to recover fromUV-C-induced DNA damage; further understanding of therole Rad54 protein plays in vivo for DSB repair will best beachieved through isogenic mutant strain construction.

Overall, the cellular response to UV-C damage was clearlystrain dependent, and the window of gene expression was tran-sient. This was especially apparent for strain P2-A. The largesttranscriptional increase occurs 30 min after exposure in strainP2-A, but transcript levels were significantly lower 15 minearlier or later in the time course. Transcript stability in S.solfataricus has been shown to be variable, with some RNAshaving very short half-lives and some remaining stable for aslong as 2 h (2, 5). It is possible that expression of DSB repairgenes is even more highly upregulated than we were able todetect, as our time points were a full 15 min apart and RNAturnover could be rapid. Prior to this work, no significantincrease in DSB repair-related gene expression in response toUV damage has been reported for S. solfataricus. Genome-wide microarray experiments for strains PH1 and strain P2(obtained from the DSMZ) indicated only mild increases (upto 2-fold) in transcription for DSB repair genes (11, 12). Thisdisparity could be the result of differential sensitivity betweenmicroarray data and quantitative real-time PCR or could arisefrom simple strain differences. Strain PH1 and the DSMZ-derived P2 strain might be more like strain 98/2 in their re-sponse to UV damage and induce expression of DSB repairgenes at a low level that persists over a period of time.

We have identified a fundamental difference in the physio-logical response to UV-C damage in the P2 strains and instrain 98/2 that is most pronounced at the transcript inductionlevel. The experimentally useful ability of strain 98/2 to permitsite-directed gene replacement through homologous recombi-nation could be the result of the observed mild but persistentupregulation of DSB repair protein transcripts following theapplication of a DNA damaging agent, in this case UV-Cirradiation. DNA uptake by S. solfataricus is facilitated byelectroporation, a transformation method that has been re-ported to cause cellular damage by inducing DNA breaks (26,27). Assuming similar DSB repair protein stability in vivo inSulfolobus, it may be that the ease with which strain 98/2performs directed gene replacement using exogenous DNA isthe result of persistent DSB repair pathway induction thatwould give cells an extended opportunity to accomplish homol-ogous recombination. A shorter length of time of expression ofDSB repair genes in the P2 strains may result in a moreabbreviated time frame for efficient homologous recombina-tion and could explain the failure of site-directed gene replace-ment efforts in these strains.

ACKNOWLEDGMENTS

This work was supported by grants from the American Cancer So-ciety IRG-77003-26 and from the National Science Foundation MCB-0951125.

We thank Yvan Zivanovic and Paul Blum for their generous gifts ofS. solfataricus strains P2 and 98/2, respectively. Both Andy Galbraithand Bill Graham critically read the manuscript. Additionally, we thankAndy Galbraith for technical assistance.

REFERENCES

1. Abella, M., S. Rodriguez, S. Paytubi, S. Campoy, M. F. White, and J. Barbe.2007. The Sulfolobus solfataricus radA paralogue sso0777 is DNA damageinducible and positively regulated by the Sta1 protein. Nucleic Acids Res.35:6788––6797.

2. Andersson, A. F., M. Lundgren, S. Eriksson, M. Rosenlund, R. Bernander,and P. Nilsson. 2006. Global analysis of mRNA stability in the archaeonSulfolobus. Genome Biol. 7:R99.

2a.Applied Biosystems. 2001. User bulletin #2 for the ABI Prism 7700 sequencedetection system. Applied Biosystems, Foster City, CA.

3. Ariza, A., D. J. Richard, M. F. White, and C. S. Bond. 2005. Conformationalflexibility revealed by the crystal structure of a crenarchaeal RadA. NucleicAcids Res. 33:1465––1473.

4. Baliga, N. S., S. J. Bjork, R. Bonneau, M. Pan, C. Iloanusi, M. C. Kotte-mann, L. Hood, and J. DiRuggiero. 2004. Systems level insights into thestress response to UV radiation in the halophilic archaeon HalobacteriumNRC-1. Genome Res. 14:1025––1035.

5. Bini, E., V. Dikshit, K. Dirksen, M. Drozda, and P. Blum. 2002. Stability ofmRNA in the hyperthermophilic archaeon Sulfolobus solfataricus. RNA8:1129––1136.

6. Cadet, J., E. Sage, and T. Douki. 2005. Ultraviolet radiation-mediated dam-age to cellular DNA. Mutat. Res. 571:3––17.

7. Chen, L., K. Brugger, M. Skovgaard, P. Redder, Q. She, E. Torarinsson, B.Greve, M. Awayez, A. Zibat, H. P. Klenk, and R. A. Garrett. 2005. Thegenome of Sulfolobus acidocaldarius, a model organism of the Crenarcha-eota. J. Bacteriol. 187:4992––4999.

8. Cooper, C. R., A. J. Daugherty, S. Tachdjian, P. H. Blum, and R. M. Kelly.2009. Role of vapBC toxin-antitoxin loci in the thermal stress response ofSulfolobus solfataricus. Biochem. Soc. Trans. 37:123––126.

9. Courcelle, J., J. R. Donaldson, K. H. Chow, and C. T. Courcelle. 2003. DNAdamage-induced replication fork regression and processing in Escherichiacoli. Science 299:1064––1067.

10. Friest, J. A., Y. Maezato, S. Broussy, P. Blum, and D. B. Berkowitz. 2010.Use of a robust dehydrogenase from an archael hyperthermophile in asym-metric catalysis-dynamic reductive kinetic resolution entry into (S)-profens.J. Am. Chem. Soc. 132:5930––5931.

11. Frols, S., P. M. Gordon, M. A. Panlilio, I. G. Duggin, S. D. Bell, C. W.Sensen, and C. Schleper. 2007. Response of the hyperthermophilic archaeonSulfolobus solfataricus to UV damage. J. Bacteriol. 189:8708––8718.

12. Gotz, D., S. Paytubi, S. Munro, M. Lundgren, R. Bernander, and M. F.White. 2007. Responses of hyperthermophilic crenarchaea to UV irradia-tion. Genome Biol. 8:R220.

13. Haldenby, S., M. F. White, and T. Allers. 2009. RecA family proteins inarchaea: RadA and its cousins. Biochem. Soc. Trans. 37:102––107.

14. Haseltine, C. A., and S. C. Kowalczykowski. 2009. An archaeal Rad54 pro-tein remodels DNA and stimulates DNA strand exchange by RadA. NucleicAcids Res. 37:2757––2770.

15. Jacobs, K. L., and D. W. Grogan. 1997. Rates of spontaneous mutation in anarchaeon from geothermal environments. J. Bacteriol. 179:3298––3303.

16. Jonuscheit, M., E. Martusewitsch, K. M. Stedman, and C. Schleper. 2003. Areporter gene system for the hyperthermophilic archaeon Sulfolobus solfa-taricus based on a selectable and integrative shuttle vector. Mol. Microbiol.48:1241––1252.

17. Kawarabayasi, Y., Y. Hino, H. Horikawa, K. Jin-no, M. Takahashi, M.Sekine, S. Baba, A. Ankai, H. Kosugi, A. Hosoyama, S. Fukui, Y. Nagai, K.Nishijima, R. Otsuka, H. Nakazawa, M. Takamiya, Y. Kato, T. Yoshizawa, T.Tanaka, Y. Kudoh, J. Yamazaki, N. Kushida, A. Oguchi, K. Aoki, S. Masuda,M. Yanagii, M. Nishimura, A. Yamagishi, T. Oshima, and H. Kikuchi. 2001.Complete genome sequence of an aerobic thermoacidophilic crenarchaeon,Sulfolobus tokodaii strain 7. DNA Res. 8:123––140.

18. Korencic, D., I. Ahel, J. Schelert, M. Sacher, B. Ruan, C. Stathopoulos, P.Blum, M. Ibba, and D. Soll. 2004. A freestanding proofreading domain isrequired for protein synthesis quality control in Archaea. Proc. Natl. Acad.Sci. U. S. A. 101:10260––10265.

19. Kuzminov, A. 2001. DNA replication meets genetic exchange: chromosomaldamage and its repair by homologous recombination. Proc. Natl. Acad. Sci.U. S. A. 98:8461––8468.

20. Lieber, M. R. 2010. The mechanism of double-strand DNA break repair bythe nonhomologous DNA end-joining pathway. Annu. Rev. Biochem. 79:181––211.

VOL. 192, 2010 SULFOLOBUS UV-INDUCED DOUBLE-STRAND BREAK REPAIR 4961

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from

21. Lisby, M., and R. Rothstein. 2009. Choreography of recombination proteinsduring the DNA damage response. DNA Repair (Amsterdam) 8:1068––1076.

22. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expressiondata using real-time quantitative PCR and the 2(��Delta Delta C(T))method. Methods 25:402––408.

23. Mandal, D., C. Kohrer, D. Su, S. P. Russell, K. Krivos, C. M. Castleberry,P. Blum, P. A. Limbach, D. Soll, and U. L. RajBhandary. 2010. Agmatidine,a modified cytidine in the anticodon of archaeal tRNA(Ile), base pairs withadenosine but not with guanosine. Proc. Natl. Acad. Sci. U. S. A. 107:2872––2877.

24. McCready, S., J. A. Muller, I. Boubriak, B. R. Berquist, W. L. Ng, and S.DasSarma. 2005. UV irradiation induces homologous recombination genesin the model archaeon, Halobacterium sp. NRC-1. Saline Systems 1:3.

25. McRobbie, A. M., L. G. Carter, M. Kerou, H. Liu, S. A. McMahon, K. A.Johnson, M. Oke, J. H. Naismith, and M. F. White. 2009. Structural andfunctional characterisation of a conserved archaeal RadA paralog with an-tirecombinase activity. J. Mol. Biol. 389:661––673.

26. Meaking, W. S., J. Edgerton, C. W. Wharton, and R. A. Meldrum. 1995.Electroporation-induced damage in mammalian cell DNA. Biochim. Bio-phys. Acta 1264:357––362.

27. Meldrum, R. A., M. Bowl, S. B. Ong, and S. Richardson. 1999. Optimisationof electroporation for biochemical experiments in live cells. Biochem. Bio-phys. Res. Commun. 256:235––239.

28. Quaiser, A., F. Constantinesco, M. F. White, P. Forterre, and C. Elie. 2008.The Mre11 protein interacts with both Rad50 and the HerA bipolar helicaseand is recruited to DNA following gamma irradiation in the archaeon Sul-folobus acidocaldarius. BMC Mol. Biol. 9:25.

29. Richards, J. D., L. Cubeddu, J. Roberts, H. Liu, and M. F. White. 2008. Thearchaeal XPB protein is a ssDNA-dependent ATPase with a novel partner.J. Mol. Biol. 376:634––644.

30. Rolfsmeier, M., C. Haseltine, E. Bini, A. Clark, and P. Blum. 1998. Molec-ular characterization of the alpha-glucosidase gene (malA) from the hyper-thermophilic archaeon Sulfolobus solfataricus. J. Bacteriol. 180:1287––1295.

31. Rolfsmeier, M. L., and C. A. Haseltine. 2010. The single-stranded DNAbinding protein of Sulfolobus solfataricus acts in the presynaptic step ofhomologous recombination. J. Mol. Biol. 397:31––45.

32. Rouillon, C., and M. F. White. 2010. The XBP-Bax1 helicase-nuclease com-plex unwinds and cleaves DNA: implications for eukaryal and archaealnucleotide excision repair. J. Biol. Chem. 285:11013––11022.

33. Sancar, A. 1996. DNA excision repair. Annu. Rev. Biochem. 65:43––81.34. Sancar, A. 2003. Structure and function of DNA photolyase and crypto-

chrome blue-light photoreceptors. Chem. Rev. 103:2203––2237.35. Schelert, J., M. Drozda, V. Dixit, A. Dillman, and P. Blum. 2006. Regulation

of mercury resistance in the crenarchaeote Sulfolobus solfataricus. J. Bacte-riol. 188:7141––7150.

36. Schmidt, K. J., K. E. Beck, and D. W. Grogan. 1999. UV stimulation ofchromosomal marker exchange in Sulfolobus acidocaldarius: implications forDNA repair, conjugation and homologous recombination at extremely hightemperatures. Genetics 152:1407––1415.

37. Schmittgen, T. D., and K. J. Livak. 2008. Analyzing real-time PCR data bythe comparative C(T) method. Nat. Protoc. 3:1101––1108.

38. Seitz, E. M., J. P. Brockman, S. J. Sandler, A. J. Clark, and S. C. Kowalc-zykowski. 1998. RadA protein is an archaeal RecA protein homolog thatcatalyzes DNA strand exchange. Genes Dev. 12:1248––1253.

39. Seitz, E. M., and S. C. Kowalczykowski. 2000. The DNA binding and pairingpreferences of the archaeal RadA protein demonstrate a universal charac-teristic of DNA strand exchange proteins. Mol. Microbiol. 37:555––560.

40. She, Q., R. K. Singh, F. Confalonieri, Y. Zivanovic, G. Allard, M. J. Awayez,C. C. Chan-Weiher, I. G. Clausen, B. A. Curtis, A. De Moors, G. Erauso, C.Fletcher, P. M. Gordon, I. Heikamp-de Jong, A. C. Jeffries, C. J. Kozera, N.Medina, X. Peng, H. P. Thi-Ngoc, P. Redder, M. E. Schenk, C. Theriault, N.Tolstrup, R. L. Charlebois, W. F. Doolittle, M. Duguet, T. Gaasterland, R. A.Garrett, M. A. Ragan, C. W. Sensen, and J. Van der Oost. 2001. Thecomplete genome of the crenarchaeon Sulfolobus solfataricus P2. Proc. Natl.Acad. Sci. U. S. A. 98:7835––7840.

41. Sheng, D., M. Li, J. Jiao, J. Ni, and Y. Shen. 2008. Co-expression with RadAand the characterization of stRad55B, a RadA paralog from the hyperther-mophilic crenarchaea Sulfolobus tokodaii. Sci. China C Life Sci. 51:60––65.

42. Sheng, D., S. Zhu, T. Wei, J. Ni, and Y. Shen. 2008. The in vitro activity ofa Rad55 homologue from Sulfolobus tokodaii, a candidate mediator inRadA-catalyzed homologous recombination. Extremophiles 12:147––157.

43. White, M. F. 2009. Structure, function and evolution of the XPD family ofiron-sulfur-containing 5'�333'� DNA helicases. Biochem. Soc. Trans. 37:547––551.

44. Wiseman, H., and B. Halliwell. 1996. Damage to DNA by reactive oxygenand nitrogen species: role in inflammatory disease and progression to cancer.Biochem. J. 313:17––29.

45. Wong, J. H., J. A. Brown, Z. Suo, P. Blum, T. Nohmi, and H. Ling. 2010.Structural insight into dynamic bypass of the major cisplatin-DNA adduct byY-family polymerase Dpo4. EMBO J. 29:2059––2069.

46. Wood, E. R., F. Ghane, and D. W. Grogan. 1997. Genetic responses of thethermophilic archaeon Sulfolobus acidocaldarius to short-wavelength UVlight. J. Bacteriol. 179:5693––5698.

47. Worthington, P., V. Hoang, F. Perez-Pomares, and P. Blum. 2003. Targeteddisruption of the alpha-amylase gene in the hyperthermophilic archaeonSulfolobus solfataricus. J. Bacteriol. 185:482––488.

48. Yang, S., X. Yu, E. M. Seitz, S. C. Kowalczykowski, and E. H. Egelman. 2001.Archaeal RadA protein binds DNA as both helical filaments and octamericrings. J. Mol. Biol. 314:1077––1085.

4962 ROLFSMEIER ET AL. J. BACTERIOL.

on May 9, 2016 by P

EN

N S

TA

TE

UN

IVhttp://jb.asm

.org/D

ownloaded from