Vol. 176, No. 24JOURNAL OF BAcTERIOLOGY, Dec. 1994, p. 7439-74460021-9193/94/$04.00+0Copyright © 1994, American Society for Microbiology

Novel Ionizing Radiation-Sensitive Mutantsof Deinococcus radiodurans

KUMARASWAMY S. UDUPA, PEGGY A. O'CAIN, VALERIE MATTIMORE, AND JOHN R. BATTISTA*Department of Microbiology, Louisiana State University, Baton Rouge, Louisiana 70803

Received 11 July 1994/Accepted 3 October 1994

Two new loci, irrB and irrI, have been identified in Deinococcus radiodurans. Inactivation of either locusresults in a partial loss of resistance to ionizing radiation. The magnitude of this loss is locus specific anddifferentially affected by inactivation of the uvrA gene product. An irrB uvrA double mutant is more sensitiveto ionizing radiation than is an irrB mutant. In contrast, the irrI uvrA double mutant and the irrI mutant areequally sensitive to ionizing radiation. The irrB and irrI mutations also reduce D. radiodurans resistance to UVradiation, this effect being most pronounced in uvrA+ backgrounds. Subclones of each gene have been isolated,and the loci have been mapped relative to each other. The irrB and irrI genes are separated by approximately20 kb of intervening sequence that encodes the uvrA and pol genes.

Members of the family Deinococcaceae are distinguished bytheir extraordinary resistance to the lethal and mutageniceffects of many DNA-damaging agents, including mitomycin,UV, and ionizing radiation (16, 24). For example, an exponen-tial-phase culture of Deinococcus radiodurans R1 can with-stand 5,000 Gy of y radiation without loss of viability (20). Thecapacity to survive such massive insults to their genetic integ-rity suggests that the Deinococcaceae have evolved distinctivemechanisms ofDNA damage tolerance, and available evidenceargues that efficient DNA repair is an integral part of thistolerance. The enzymology of the deinococci's DNA repairprocesses is poorly understood, however.

Studies of D. radiodurans R1 have shown that DNA damageresistance is, in part, mediated by the activities of two endo-nucleases, designated a and 1 (6, 19). Endonuclease a assistsin the repair of a broad range of DNA damage, and thisenzyme appears to be functionally analogous to the ABCexcinuclease of Escherichia coli (19). Until very recently, it wasbelieved that endonuclease a was encoded by two genes, mtcAand mtcB, but Minton has reported (12, 13) that the mtcABregion ofD. radiodurans is a single gene that encodes a proteinthat is homologous with the UvrA protein of E. coli. Muta-tional inactivation of the uvrA locus of D. radiodurans renderscells sensitive to the lethal effects of mitomycin and hypermut-able to simple alkylating agents such as MNNG (N-methyl-N'-nitro-N-nitrosoguanidine). Experimental evidence indicatesthat endonuclease a is the product of the D. radiodurans uvsC,uvsD, and uvsE genes. This repair protein has a substratespecificity narrower than that of endonuclease a, however, andmay be a pyrimidine dimer-specific endonuclease. Cell extractsof D. radiodurans that exhibit endonuclease 13 activity havebeen shown to catalyze the initial incision necessary for theremoval of pyrimidine dimers (7). The two endonucleases haveoverlapping activities, and both proteins must be inactivated toprevent the excision of UV-induced DNA damage (6). uvrAuvs double mutants are sensitive to UV light, whereas strainsthat carry only a uvrA or uvs mutation exhibit wild-type levelsof UV resistance. Specific processes that mediate the repair ofionizing radiation-induced DNA damage in the deinococci

* Corresponding author. Mailing address: Department of Microbi-ology, 508 Life Sciences Building, Louisiana State University, BatonRouge, LA 70803. Phone: (504) 388-2810. Fax: (504) 388-2597.

have not been described. Endonucleases at and 1 do notappear to play an essential role in ionizing radiation resistance,since uvrA uvs double mutants are nearly as resistant to -yradiation as D. radiodurans R1 (6). Mutational inactivation ofthe pol or rec loci of D. radiodurans R1 results in strains thatare ionizing radiation sensitive (IRS) (8, 9, 14, 17), but thesestrains are also sensitive to mitomycin and UV light, indicatingthat the pol and rec gene products are involved in repairpathways common to all three types of DNA damage. The poland rec gene products of D. radiodurans have recently beenshown to be homologs of E. coli DNA polymerase I (9) andRecA protein (8), respectively. Other defects associated withIRS strains of D. radiodurans have not been characterized.As a first step toward developing a better understanding of

the deinococci's remarkable radioresistance, we have screened45,000 MNNG-mutagenized colonies of D. radiodurans 302 forsensitivity to ionizing radiation. This screen led to the isolationof 49 IRS strains. We report here the initial characterization oftwo IRS strains that have been designated IRS18 and IRS41.Each strain carries a unique and previously undescribed defectthat severely impairs the ability of D. radiodurans to surviveexposure to ionizing radiation.

MATERMILS AND METHODS

Bacterial strains and plasmids. Bacterial strains and plas-mids used in this study are listed in Table 1. All D. radioduransstrains were grown at 30°C in TGY broth (0.5% tryptone, 0.3%yeast extract, 0.1% glucose) or on TGY agar (1.5% agar). E.coli strains were grown in Luria-Bertani broth or on Luria-Bertani plates at 37°C (23). Plasmids were routinely propa-gated in E. coli DH5a MCR.MNNG mutagenesis. Exponentially growing cultures of D.

radiodurans 302 were treated with 20 ,ug ofMNNG per ml andincubated for 2 h at 30°C with shaking (25). A 100-,ul aliquot ofmutagenized cells was diluted in 25 ml of TGY and incubatedfor an additional 18 h. The resulting culture was diluted andplated at 200 to 300 CFU per plate. The mutagenized popu-lation was screened for ionizing radiation sensitivity by patch-ing individual colonies onto TGY plates and exposing the platesto 5,000 Gy of y radiation (60Co source, 20 Gy/min, 22°C).Colonies of untreated D. radiodurans 302 patched onto eachplate served as a control. Putative IRS mutants were identifiedby comparing the growth of mutagenized colonies with that of

7439

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7440 UDUPA ET AL.

TABLE 1. Bacterial strains and plasmids

Bacterial strain or plasmid Relevant description Source or reference

D. radioduransR1 ATCC 13939 2302 As R1 but uvrAl (mtcA)a 18262 As R1 but uvrA2 (mtcB)a 18UV17 As R1 butpol 14IRS18 As 302 but irrBl This studyIRS181 As IRS18 but uvrA' This studyIRS41 As 302 but irrll This studyIRS411 As IRS41 but uvrA' This studyLS18 As R1 but streptomycin resistant This study

E. coliDH5ct-MCR F mcrA A(mrr-hsdRMS-mcrBC) 480dlacZAM15 AlacX74 end41 recAl deoR A(ara-leu) Life Technologies, Inc.,

7697 araD139 galU galK nupG rpsL Gaithersburg, Md.NM554 recA13 araD139 A(ara-leu)7697 AiacAl7 galU galK hsdR514 mcrA mcrB rpsL 21

pUE58 pAT153::uvrA+ (mtcA+) irrB+ 5.6-kb subclone of D. radiodurans genomic DNA 1carrying a portion of the wild-type uvrA gene that restores uvrAl (mtcA) strains tomitomycin resistance and the irrB gene

pPG11 pBluescript SK+ derivative with a 5.2-kb insert of D. radiodurans genomic DNA 9carrying a 1.98-kb fragment of the pol gene

pPG12 pBC SK+ derivative with a 1.2-kb insert of D. radiodurans genomic DNA carrying a 90.9-kb fragment of the pol locus

pKU1 pWEl5::uvrA+ (mtcAB+) irrB+ 40-kb cosmid clone from D. radiOdurans R1 genomic This studyDNA carrying the uvrA gene and the irrB gene

pPOl pWE15::uvrA+ (mtcB+) pol+ irrI+ 40-kb cosmid clone from D. radiodurans R1 genomic This studyDNA carrying a portion of the wild-type uvrA gene that restores uvrA2 (mtcB) strainsto mitomycin resistance, the pol gene, and the irrI gene

pPOll pACYC184::pol+ irrI+ 14-kb subclone of pPO1 This studypPO12 pACYC184::irrI+ 5-kb subclone of pPO1 This study

a The mtcAB region of the D. radiodurans chromosome has recently been sequenced by Minton (12, 13) and shown to contain a single gene whose product ishomologous with the UvrA protein of E. coli. In light of this new information, Minton has recommended that the mtcA mutation of D. radiodurans 302 and the mtcBmutation of D. radiodurans 262 be designated uvrAl and uvrA2, respectively.

strain 302 48 h after irradiation. A total of 45,000 mutagenizedcolonies from 11 separate MNNG-treated cultures werescreened. Putative mutants were streaked to isolation, and theIRS phenotype was confirmed by repatching and irradiatingfive isolated colonies from each isolation streak.Transformation in liquid culture. Calcium chloride from a 1

M stock solution was added to D. radiodurans cultures inexponential growth phase until a final concentration of 30 mMwas achieved. This mixture was incubated at 30°C for 80 min(26). Either 1 ,ug of plasmid DNA or 10 ,ug of chromosomalDNA was added to 1 ml of TGY containing 2 X 107 cells, andthe mixture was incubated on ice for 30 min. The transforma-tion mixture was diluted 10-fold with TGY broth and incu-bated for another 18 h at 30°C. This transformation protocolwas used to identify uvrA+ cells. Transformants were selectedon TGY plates containing 60 ng of mitomycin per ml (18).Dot transformation. Calcium chloride was added to D.

radiodurans cultures in exponential growth phase until a finalconcentration of 30 mM was achieved. This mixture wasincubated at 30°C for 80 min. A 100-,ul aliquot of cells wasspread onto a TGY plate and incubated for 4 to 6 h at 30°C.Transformations were performed by dotting either chromo-somal or plasmid DNA onto the plate. Twenty-four hourslater, the bacterial lawn was replica plated and selectivepressure was applied. To select mitomycin-resistant transfor-mants, lawns were replica plated onto TGY agar containing 60ng of mitomycin per ml. To select for ionizing radiation-resistant transformants, lawns were plated on TGY agar and yirradiated at 6,000 Gy. Three days after irradiation, the plateswere scored for restoration of ionizing radiation resistancewithin the area where the DNA had been dotted.

Chromosomal DNA isolation. TGY broth (200 ml) wasinoculated with a 2-ml overnight culture (2 X 10' CFU/ml) ofD. radiodurans. After 48 h, the 200-ml cultures were harvestedby centrifugation at 4°C at 3,000 X g for 15 min. Pellets wereresuspended in 20 ml of 95% ethanol and held at roomtemperature for 10 min to remove the outer membrane of D.radiodurans. The ethanol-stripped cells were collected by cen-trifugation at 4°C at 3,000 X g for 15 min, and the resultingpellet was resuspended in 9 ml of TE buffer (10 mM Tris-HCl,0.1 mM EDTA; pH 8.0). Two milligrams of lysozyme (SigmaChemical, St. Louis, Mo.) was added to stripped cells, and thismixture was incubated at 37°C for 30 min. A 0.5-ml volume of10% sodium dodecyl sulfate and 50 ,ul of proteinase K (SigmaChemical) (20 mg/ml) were added to lysozyme-treated cells,and the mixture was incubated for 12 h at 56°C. Lysed cellswere transferred to a centrifuge tube and extracted once withan equal volume of phenol-chloroform (1:1) and twice withequal volumes of chloroform-isoamyl alcohol (24:1). The DNAwas precipitated by adding 1 ml of 3 M sodium acetate (pH7.0) and 20 ml of ice-cold 100% ethanol to the extractedmaterial. The DNA was spooled out with a curved glass rodand washed twice with 70% ethanol. The DNA was air dried,dissolved in 5 ml of TE buffer (pH 8.0), and stored at 4°C.

Construction of pWE15 cosmid library. D. radiodurans R1chromosomal DNA was partially digested with Sau3AI (NewEngland Biolabs, Beverly, Mass.), and the restriction frag-ments were size selected on a linear sodium chloride gradientranging from 1.25 to 5.0 M NaCl. The fractions containingfragments of approximately 40 kb were ligated to a uniqueBamHI site of the cosmid vector pWE15 (Stratagene CloningSystems, La Jolla, Calif.). The recombinants were packaged

J. BACTERIOL.

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

IONIZING RADIATION-SENSITIVE D. RADIODURANS MUTANTS 7441

B

FIG. 1. (A) Pattern of growth on a D. radiodurans R1 spread plate dot transformed with 200 ng of chromosomal DNA from LS18, astreptomycin-resistant isolate of D. radiodurans R1. Transformants were selected on TGY agar containing 50 jig of streptomycin per ml. (B)Pattern of growth on an IRS18 plate dot transformed with 500 ng of chromosomal DNA from strains IRS41 to IRS47. Following replica plating,the IRS18 lawn was exposed to 6,000 Gy of y radiation. Areas of growth indicate successful transformation to ionizing radiation resistance.

with Gigapak-XL in vitro packaging extract (Stratagene Clon-ing Systems) and transfected into E. coli NM554. Successfullytransfected cells were selected on Luria-Bertani plates contain-ing 50 jig of ampicillin per ml. The titers obtained fromampicillin selection indicated that there were at least 66,000cosmid clones per jig of size-fractionated DNA. Restrictiondigestion using NotI and EcoRI on randomly selected cosmidclones revealed that the insert sizes ranged from 38 to 42 kb.Cosmid DNA was isolated by an alkaline lysis protocol (3).

Survival curves. Only D. radiodurans cultures in exponentialgrowth phase were evaluated for their ability to survive UV orionizing radiation. To measure ionizing radiation resistance,cultures were divided into 1-ml aliquots, placed in Eppendorftubes, and exposed to a 60Co source with a dose rate of 20Gy/min. The sample chamber of the irradiator used is acylinder that is 150 mm in diameter and 50 cm long. The 6Cosource surrounds the sample chamber during irradiation. Foreach irradiation, samples were placed at the positions alongthe length of the sample chamber where optimal dose ratescould be achieved. The sample chamber was held at 22°Cduring irradiation. Cells were removed from the source afterthe desired accumulated dose was achieved. Irradiated cellswere diluted, plated in triplicate on TGY agar plates, andincubated for 3 days at 30°C before survivors were scored.

Survival following exposure to UV light was evaluated byplating appropriate dilutions of D. radiodurans cultures onTGY agar and irradiating the surface of each plate with agermicidal lamp.

Subcloning the irrI locus. The cosmid pPO1 was restrictedwith either EcoRI or NcoI. The resulting fragments wereligated into compatible sites on pACYC184. Subclones carry-ing the irrI+ locus were identified by their ability to restore irrImutants of D. radiodurans to ionizing radiation resistancefollowing dot transformation.

RESULTS

D. radiodurans and dot transformation. Dot or in situtransformation is a technique that has been used to facilitate

genetic manipulation of Acinetobacter calcoaceticus (11) andthe cyanobacterium Synechocystis sp. strain 6803 (5). In thisprocedure, transforming DNA is simply dotted onto a freshlyspread plate. Once a lawn has grown, it is replica plated andtransformants are identified by applying appropriate selectivepressure. We have determined that D. radiodurans R1 can alsobe transformed by this method. Approximately 2 X 107 R1cells obtained from an exponential-phase culture were spreadonto TGY agar and dotted with 200 ng of chromosomal DNAobtained from LS18, a streptomycin-resistant isolate of D.radiodurans R1. Once plated cells had grown into a lawn, thelawn was replica plated onto TGY agar containing 50 ,ug ofstreptomycin per ml. Following a 2-day incubation, the onlycolonies that appeared were those successfully transformed tostreptomycin resistance within the area where the transformingDNA was dotted (Fig. 1A).IRS strains ofD. radiodurans. A large yield of mutants of D.

radiodurans can be generated by treating D. radiodurans 302with the alkylating agent MNNG (25). Strain 302 is approxi-mately 50 times more mutable than the wild-type R1 strainfollowing MNNG treatment because it is a uvrA mutant (i.e., itcarries a defect in endonuclease cx). This defect preventseffective repair of base damage caused by the alkylating agent.We have screened 45,000 MNNG-mutagenized colonies ofstrain 302 for sensitivity to 5,000 Gy of ry radiation and haveidentified 49 IRS strains. To our knowledge, this is the onlylarge-scale screen for IRS mutants of D. radiodurans that hasbeen reported. The mutant strains recovered were designatedIRSlito IRS49.IRS18 and IRS41. During preliminary characterization of

the IRS strains, it was noted that IRS18 and IRS41 exhibited<1% survival relative to their parent strain following admin-istration of 6,000 Gy of y radiation. The mutations responsiblefor the sensitivity of IRS18 and IRS41 to ionizing radiationdiffer from each other and are unique within the collection ofIRS strains isolated. This was determined by testing whetherchromosomal DNA isolated from other IRS strains couldrestore IRS18 or IRS41 to wild-type levels of radioresistance ina dot transformation protocol similar to that described above.

VOL. 176, 1994

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7442 UDUPA ET AL.

100

z00

z5U

C,

10-1

10-2

10-3

10-4

100

z0

en

0

z

CO)

0 1200 2400 3600 4800 6000 7200 8400

y RADLATION (Gy)FIG. 2. Representative survival curves for IRS18 (triangles) and

IRS181 (squares) following exposure to My radiation. Survival of strain302 (circles) is also shown. Values are means ± standard deviations ofduplicate experiments.

Freshly spread cultures of either IRS18 or IRS41 were dottedin separate locations with 2.0 pug of chromosomal DNA fromthe other IRS strains. The resulting lawn was replica platedonto a TGY plate and -y irradiated at 6,000 Gy to select forcells transformed to ionizing radiation resistance. The restora-tion of radioresistance, as evidenced by heavy growth withinthe area of the dotted DNA (Fig. 1B), indicates that themutation found in the IRS strain undergoing transformation(the recipient strain) has been replaced with a wild-typesequence in a recombinational event. The transforming DNA(the donor strain's chromosomal DNA) is the source of thatwild-type sequence. In other words, the mutations that madethe donor and recipient strains IRS are different. When IRS18served as the recipient, chromosomal DNA from all other IRSstrains restored wild-type levels of radioresistance to thisstrain. Similarly, radioresistance was restored to IRS41 upontransformation with chromosomal DNA from all other IRSstrains. Chromosomal DNA from IRS18 could not restoreIRS18 to radioresistance, nor could IRS41 DNA restoreIRS41.

Since both strains take up and stably incorporate exog-enously added chromosomal DNA during natural transforma-tion, it is assumed that they are recombination proficient (i.e.,rec'). The efficiency of transformation was determined to beabout 0.1% for IRS18 and IRS41, identical to that obtainedwhen strain 302 was transformed. This was determined byadding 1 pRg of LS18 chromosomal DNA to 2 X 107 exponen-tial-phase cells of each strain in 1 ml of liquid medium andcalculating the efficiency with which LS18's streptomycin resis-tance marker was transferred to the strain.

Survival of IRS18 and IRS41 following exposure to 'y

radiation. Representative survival curves for IRS18 and IRS41following exposure to -y radiation are depicted in Fig. 2 and 3,respectively. Both strains were substantially more sensitive toionizing radiation than strain 302. For example, approximately10% of the IRS18 culture (Fig. 2) and 1% of the IRS41 culture

10-1

10-2

10-3

10-4 -0 1200 2400 3600 4800 6000 7200

yRADIATION (Gy)FIG. 3. Representative survival curves for IRS41 (triangles) and

IRS411 (squares) following exposure to y radiation. Survival of strain302 (circles) is also shown. Values are means ± standard deviations ofduplicate experiments.

(Fig. 3) survive exposure to 3,600 Gy of y radiation, whereas86% of the 302 culture survives (Fig. 2 and 3). The shouldercharacteristic of wild-type strains of Deinococcus was missingfrom the survival curves of IRS18 and IRS41. At radiationdoses of between 2,000 and 8,400 Gy, IRS41 was approxi-mately 10-fold more sensitive than IRS18 at a given dose.The influence of a uvrA mutation on the IRS phenotype of

IRS18 and IRS41. Since IRS18 and IRS41 are derived fromstrain 302, they are sensitive to mitomycin as well as ionizingradiation. To determine whether the lack of the uvrA geneproduct contributed to the IRS phenotype observed, IRS18and IRS41 were transformed to mitomycin resistance and thenscreened for ionizing radiation resistance. Transformation tomitomycin resistance was accomplished by using plasmidpUE58 (1). This plasmid contains a 5.6-kb subclone thatencodes a portion of the wild-type uvrA locus that restoresmitomycin resistance to derivatives of strain 302 but not tostrain 262. This plasmid cannot replicate in D. radiodurans, andrestoration of mitomycin resistance requires homologous re-combination between plasmid and chromosomal DNA (1).pUE58 was linearized and added to competent liquid culturesof IRS18 and IRS41. Linearization was necessary to ensurethat only double crossover events would give rise to themitomycin-resistant transformants. Successful transformantswere selected by plating transformed cultures on TGY agarcontaining 60 ng of mitomycin per ml. Resistant colonies werepatched onto fresh TGY plates and screened for resistance to6,000 Gy of y radiation. The results of this screen are summa-rized in Table 2. It was possible to obtain mitomycin-resistanttransformants that remained IRS from both IRS18 and IRS41,suggesting that the MNNG-induced mutations introduced intothese strains were solely responsible for the IRS phenotypeobserved. Mitomycin-resistant derivatives of IRS18 and IRS41were designated IRS181 and IRS411, respectively. Since thebiochemical nature of the defects in the IRS strains is un-known, wild-type loci identified by mutational inactivation in

J. BAcriERIOL.

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

IONIZING RADIATION-SENSITIVE D. RADIODURANS MUTANTS 7443

TABLE 2. Transformation of IRS18 and IRS41 to mitomycinresistance with plasmid pUE58

No. of coloniesa CotransformationStrain Mtcr Mtcr 1W Mtcr IRS frequency (%)

IRS18 (uvrAl irrBl) 751 166 585 22IRS41 (uvrAl irrIl) 499 0 499 0

a Abbreviations: Mtcr, mitomycin resistant; IRE, ionizing radiation resistant;IRS, ionizing radiation sensitive.

these strains were designated irr (for ionizing radiation resis-tance). Separate irr genes were identified by a letter code. Thewild-type loci inactivated in IRS181 and IRS411 were desig-nated irrB and irrI, respectively. The mutant alleles found inthese strains are referred to as irrBl and irrIl.IRS181 and IRS411 remained IRS, and representative sur-

vival curves are depicted in Fig. 2 and 3. IRS18 (Fig. 2) wasclearly more sensitive to ionizing radiation than IRS181 (Fig.2), its uvrA+ isogenote. When exposed to doses of ionizingradiation of less than 3,000 Gy, IRS18 and IRS181 behavedsimilarly, but at doses in excess of 3,000 Gy the presence of thewild-type uvrA gene product enhanced survival. IRS41 andIRS411 exhibited virtually identical survival curves (Fig. 3),showing no evidence that the uvrA gene product influencedionizing radiation resistance in an irrI background.Mapping the irrB and irrI loci. As illustrated in Table 2, only

80% of the mitomycin-resistant transformants derived fromIRS18 were sensitive to ionizing radiation, whereas all mito-mycin-resistant transformants derived from IRS41 were IRS.This indicated that portions of the uvrA+ and irrB+ genes werelocated within the 5.6-kb fragment cloned into pUE58 and thatthe ionizing radiation-resistant recombinants were the result ofcotransformation of these loci. Since only IRS18 was trans-formed to ionizing radiation resistance with pUE58, the irrIlocus was assumed to be outside the pUE58 fragment. In anattempt to map the positions of irrB and irrI relative to eachother, a pWE15 cosmid library was screened by dot transfor-mation for clones capable of restoring ionizing radiationresistance to IRS181 and IRS411. The first cosmid found,pKU1, restored IRS181, but not IRS411, to ionizing radiationresistance (Table 3). This cosmid failed to restore ionizingradiation resistance to any of the other 48 IRS strains (data notshown) or to the pol mutant UV17. pKU1 also restoredmitomycin resistance to the uvrA strains, 302 and 262 (Table4), indicating that the irrB and uvrA loci were located within

TABLE 3. Restoration of ionizing radiation resistance in selectedstrains of D. radiodurans following transformation with subclones

of the uvrA, pol, irrB, and irrI genes0Result for plasmidb:

StrainpP01 pKU1 pUE58 pPG11 pPG12 pPO11 pPO12

IRS18 (uvrAl - + +irrBl)

IRS181 - + + - - - -(irrBl)

IRS41 (uvrAl + - - - - + +irrIl)

IRS411 (irrIl) + - - - - + +UV17(pol) + - - + + + -

a See Table 1 for a detailed description of the plasmids used in this analysis.b +, restoration of IRS strain to a radioresistant phenotype following trans-

formation with the donor DNA; -, no restoration.

TABLE 4. Restoration of mitomycin resistance in selected strainsof D. radiodurans following transformation with subclones of the

uvrA, pol, irrB, and irrI genesaResult for plasmidb:

StrainpP01 pKU1 pUE58 pPG11 pPG12 pPO11 pPO12

IRS18 (uvrAl - + + - - - -irrBl)

IRS41 (uvrAl - + +irrIl)

302 (uvrAl) - + +262 (uvrA2) + + - +UV17 (pol) + - - + + + -

a See Table 1 for a detailed description of the plasmids used in this analysis.b +, restoration of mitomycin-sensitive strain to mitomycin-resistant pheno-

type following transformation with the donor DNA; -, no restoration.

this cosmid's 40-kb cloned insert and that the irrI and pol lociwere found outside this region of the chromosome.A second cosmid, pPO1, that restored IRS411, UV17, and

IRS34 to ionizing radiation resistance was identified (Table 3).It did not restore radioresistance to any other IRS strain (datanot shown). Although the mutation responsible for IRS34 hasnot yet been characterized, the gene inactivated does notappear to be related to the irrI or pol genes (29). pPO1 alsotransformed strain 262 to mitomycin resistance, but, in contrastto pKU1, it did not restore mitomycin resistance to strain 302,indicating that the DNA sequences of the inserts found onpKU1 and pPO1 partially overlap and that a portion of theuvrA gene is within the area of the overlap.To obtain a higher-resolution map of the area of the D.

radiodurans chromosome defined by the cosmids pKU1 andpPO1, plasmids carrying subclones of the wild-type uvrA, pol,irrB, and irrI loci were used as donor DNAs in a series of dottransformation experiments that attempted to restore eitherionizing radiation resistance or mitomycin resistance to strainscarrying defects at one of these loci. The results from these dottransformations are summarized in Tables 3 and 4.As indicated earlier, the 5.6-kb insert of pUE58 (1) includes

a portion of the uvrA and irrB genes. Wild-type resistance tomitomycin and ionizing radiation was restored to strains 302and IRS181, respectively, when transformed with this plasmid.pPG11 (9) carries a 5.2-kb insert that includes 1,980 bp of the3' end of the pol gene and 1.3 kb of downstream sequence.pPG12's 1.2-kb insert encodes the final 980 bp of the pol geneand 220 bp of downstream sequence. pPG11 and pPG12restored strain UV17 to ionizing radiation resistance. pPG11also transformed strain 262 to mitomycin resistance. pPG12did not, however, restore mitomycin resistance to strain 262.The uvrA locus is apparently located in the region downstreamfrom the pol gene, placing the pol gene near the overlapbetween pKU1 and pPO1. The subclones located withinpPO11 and pPO12 were derived from the 40-kb insert presentin pPO1. pPO11 has a 14-kb insert that restores ionizingradiation resistance to the irrI and pol strains. pPO12's 5.0-kbinsert transformed only the irrI strains to radioresistance.The observations summarized in Tables 3 and 4 allowed the

uvrA,pol, irrB, and irrI loci to be ordered relative to each other.The irrB and irrI genes are separated by approximately 20 kb ofintervening sequence containing the uvrA and pol loci. Thefour loci are arranged as follows: irrB-uvrA-pol-irrI.

Survival of the irrB and irrI mutants following exposure toUV radiation. As illustrated in Fig. 4 and 5, the irrB and irrIstrains of D. radiodurans exhibited sensitivity to UV lightcompared with strain 302, but the degree of sensitivity ob-

VOL. 176, 1994

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7444 UDUPA ET AL.

100 0 loss of viability. Survival of IRS181 cultures decreased rapidlyat UV doses in excess of 300 J/m2.

This enhanced lethality is observed only when the uvrA+strain is exposed to UV light. There is no evidence of enhancedlethality when uvrA + irrB or uvrA + irrI strains are exposed to

10-1 - ionizing radiation (Fig. 2 and 3), and IRS181 and IRS411 areZ as resistant to mitomycin as D. radiodurans R1 (data not01= \^shown). This suggests that the repair pathways of D. radio-iU Y\ durans for UV-, mitomycin-, and y-radiation-induced DNA

damage differ significantly from each other even though repair102 of all three types of damage seems to use the uvrA genez5 product.

CO ~~~~~~~~~~~~~~DISCUSSION103-s

Cultures of D. radiodurans display an extraordinary ability towithstand the lethal effects of ionizing radiation (16). Thetypical y radiation survival curve during exponential growth forthis species exhibits a shoulder of resistance to 5,000 Gy in

04 which there is no loss of viability (20). At doses above 5,000 Gy,0 100 200 300 400 500 600 damage apparently begins to overwhelm the cell's defenses and

there is an exponential loss of viability. A dose of 6,000 Gy is,UVDOSE(J/m)

on average, required to inactivate a single CFU of D. radio-UV DOSE (J/m2) durans R1. In terms of DNA damage, this dose of radiation will

FIG. 4. Representative survival curves for IRS18 (triangles) and induce approximately 200 double-strand breaks, over 3,000IRS181 (squares) following exposure to UV radiation. Survival of single-strand breaks, and >1,000 sites of base damage per D.strain 302 (circles) is also shown. Values are means ± standard radiodurans genome (24).deviations of duplicate experiments. All available evidence indicates that the ability of D. radio-

durans to tolerate ionizing radiation is a function of the abilityof this species to repair the DNA damage generated by

served depended on whether or not the strain was uvrA+. ionizing radiation, but relatively little is known about the DNAIRS18 and IRS41 were more resistant to UV exposure than repair systems of this organism. To date, only two deinococcaltheir uvrA' isogenotes, IRS181 and IRS411. Evidence of this proteins have been directly associated with ionizing radiationenhanced lethality was apparent at all doses examined when resistance: (i) the rec gene product, a homolog of E. coli RecAIRS411 was exposed to UV light, whereas IRS181 was found to protein (8), and (ii) the po gene product, a homolog of E. colitolerate exposure to UV doses of <300 J/m2 without significant DNA polymerase I (9). Given the complexity of DNA repair in

other species, it is reasonable to expect that the deinococci relyon the activity of other, as yet unidentified proteins to survive

100 * the cellular damage caused by -y radiation.We have initiated a search for proteins involved in D.

radiodurans's repair of ionizing radiation-induced DNA dam-age. MNNG-mutagenized populations of D. radiodurans 302were screened for IRS mutants with the assumption that many

10-1- of these mutants carry defects in DNA repair proteins. Az preliminary characterization of two of the IRS mutants iso-0 \ lated in this screen has been presented here.0 iIRS18 and IRS41 (Table 1) are derivatives of strain 302 that,

in addition to being sensitive to mitomycin, are sensitive to(D 10-2 ionizing radiation (Fig. 2 and 3). Each strain carries a novelz \ mutation that is responsible for the IRS phenotype. The locus> mutationally inactivated in IRS18 has been designated irrB,

and that inactivated in IRS41 has been designated irrL. Theto 4 1 \ mapping studies conducted have shown that the irrB and irrl

10-3 genes are separated by approximately 20 kb of interveningchromosomal DNA sequence and that the uvrA and po genesare located in this region.IRS18 (irrB1 uvrAl) cultures are substantially more sensitive

014 to ionizing radiation than are cultures of strain 302 (uvrAl)(Fig. 2), suggesting that the irrB gene product plays a signifi-

0 100 200 300 400 500 600 cant role in the ionizing radiation resistance of D. radiodurans.This enhanced sensitivity is all but eliminated, however, when

UV DOSE (J/m2) IRS18 is made uvrA+ and endonuclease a activity is restored.FIG. 5. Representative survival curves for IRS41 (triangles) and IRS181 (irrB1 uvrA+) is only slightly more sensitive than strain

IRS411 (squares) following exposure to UV radiation. Survival of 302 to ionizing radiation (Fig. 2). It appears that the simulta-strain 302 (circles) is also shown. Values are means ± standard neous inactivation of endonuclease x and the irrB gene productdeviations of duplicate experiments. renders IRS18 sensitive to the lethal effects of ionizing radia-

J. BAcrERIOL.

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

IONIZING RADIATION-SENSITIVE D. RADIODURANS MUTANTS 7445

tion. Endonuclease a has not previously been associated withD. radiodurans's tolerance of ionizing radiation-induced cellu-lar damage, but as indicated in Fig. 2, this enzyme appears toplay a role in the organism's survival following exposure to yradiation, especially at doses in excess of 3,600 Gy. This role isapparent only in an irrB background, suggesting that the irrBgene product and endonuclease a perform related or redun-dant functions. This situation is analogous to that observed forUV resistance in D. radiodurans, in which the activities ofendonuclease a and endonuclease 13 overlap and inactivationof both proteins is necessary to generate a UV-sensitivemutant strain.The irrB gene product also affects survival following UV

irradiation of D. radiodurans but to a lesser extent than it doessurvival following exposure to ionizing radiation (Fig. 4). Inaddition, the extent of this protein's influence is modified bythe presence of endonuclease a. IRS18 does not exhibit strain302's characteristic shoulder of UV resistance. Instead, there isa gradual loss of viability with increasing UV doses. IRS181'ssurvival curve parallels that of IRS18 until the accumulateddose exceeds 400 J/m2. At this point, IRS181's loss of viabilityis accelerated relative to that of IRS18, suggesting that the irrBgene product modifies or ameliorates a potentially lethalactivity associated with the action of endonuclease a onUV-damaged DNA. These observations are difficult to recon-cile with the apparent cooperative nature of IrrB and endonu-clease a activity following -y irradiation.The irr mutants, IRS41 and IRS411, are extremely sensitive

to -y radiation, and the survival curves obtained for the twostrains are virtually superimposable (Fig. 3). This indicates thatthe functions of the irrI gene product and endonuclease a donot overlap and that inactivation of the irrl gene is sufficient torender D. radiodurans IRS.IRS41 was only slightly more sensitive to UV than strain 302,

its irrI+ parent (Fig. 5). In contrast, IRS411 was extremelysensitive to UV light, exhibiting a dramatic reduction insurvival following UV doses that are sublethal to IRS41 andstrain 302. This result suggests that the irrI gene productregulates the activity of either endonuclease a or an enzymaticactivity that arises subsequent to the action of endonuclease a.

In many respects, the effect of the irrI mutation in a uvrA+background following UV irradiation is similar to the effectsobserved when wild-type D. radiodurans is treated with chlor-amphenicol prior to UV irradiation. Immediately followingUV irradiation, there is evidence of extensive DNA degrada-tion, presumably resulting from the enzymatic removal ofDNA damage (15, 27). The degradative process appears to beregulated by an inducible protein, since the administration ofchloramphenicol prior to irradiation allows DNA degradationto proceed unchecked, with lethal consequences. In uvrAmutants, the lethal effect of chloramphenicol addition is notobserved, indicating that chloramphenicol stops the synthesisof a protein that is down-regulating the process of DNAdegradation (10). Inactivation of endonuclease a also preventsthis lethal degradation, presumably because this enzyme eithercatalyzes the degradation or initiates the degradative process(10). Endonuclease a could, by incising the DNA at the site ofdamage, trigger the action of an exonuclease that degradesDNA until it is inactivated by a second protein. Conceptually,this type of interaction is analogous to that observed when thegam protein of phage lambda inactivates the RecBCD exonu-clease of E. coli (22). The similarities between the effects ofchloramphenicol and the effects of the irrl mutation on UVresistance suggest that the irr gene product is a regulatoryprotein and that it may be the protein inhibited by chloram-phenicol pretreatment.

If the irrI gene product is a regulatory protein involved in thecontrol of DNA degradation following the action of endonu-clease ao, then why is there no difference in the survival ofIRS41 and IRS411 following y irradiation? DNA degradationdoes occur after y irradiation (4, 28), and wild-type cells aremore radiosensitive if pretreated with chloramphenicol. Inaddition, observations presented here suggest that endonucle-ase ao acts on ionizing radiation-damaged DNA. Since the datapresented indicate that endonuclease a('s role in ionizingradiation resistance is redundant, the irrl gene product couldeffect an enzymatic activity that arises after the action ofendonuclease ax as well as other proteins that repair ionizingradiation-induced DNA damage. If, as postulated, the irrprotein inhibits an exonuclease activity, then there is no reasonto assume that this exonuclease functions on substrates gener-ated solely by endonuclease a.

Clearly, D. radiodurans has evolved a complex strategy forcoping with DNA damage, and the irrB and irrI gene productsare critical components in that overall strategy. Our presentunderstanding of the physiology of D. radiodurans is toorudimentary, however, to allow us to define how the irrB andirrI gene products contribute to this organism's extreme ra-dioresistance without further investigation.

ACKNOWLEDGMENTS

We are grateful to Kenneth W. Minton for providing several of thestrains and plasmids used in this work and for sharing unpublishedresults. We also thank Mary Hollis Quinn for helpful discussions.

This work was supported by Louisiana Education Quality SupportFund grant LEQSF (1992-94)-RD-A-03 from the Louisiana Board ofRegents.

REFERENCES1. Al-Bakri, G., M. Mackay, P. Whittaker, and B. E. B. Moseley.

1985. Cloning of the DNA repair genes mtcA, mtcB, uvsD, uvsE,and the leuB gene from Deinococcus radiodurans. Gene 33:305-311.

2. Anderson, A. W., H. C. Nordon, R. F. Cain, G. Parrish, and D.Duggan. 1956. Studies on a radio-resistant micrococcus. I. Isola-tion, morphology, cultural characteristics and resistance to gammaradiation. Food Technol. 10:575-578.

3. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extractionprocedure for screening recombinant plasmid DNA. NucleicAcids Res. 7:1513.

4. Dean, C. J., J. G. Little, and R. W. Serianni. 1970. The control ofpost-irradiation DNA breakdown in Micrococcus radiodurans.Biochem. Biophys. Res. Commun. 39:126-134.

5. Dzelzkalns, V. A., and L. Bogorad. 1988. Molecular analysis of amutant defective in photosynthetic oxygen evolution and isolationof a complementing clone by a novel screening procedure. EMBOJ. 7:333-338.

6. Evans, D. M., and B. E. B. Moseley. 1983. Roles of uvsC, uvsD,uvsE, and mtcA genes in the two pyrimidine dimer excision repairpathways of Deinococcus radiodurans. J. Bacteriol. 156:576-583.

7. Evans, D. M., and B. E. B. Moseley. 1985. Identification and initialcharacterization of a pyrimidine dimer UV endonuclease (UVendonuclease 1) from Deinococcus radiodurans: a DNA-repairenzyme that requires manganese ions. Mutat. Res. 145:119-128.

8. Gutman, P. D., J. D. Carroll, C. I. Masters, and K. W. Minton.1994. Sequencing, targeted mutagenesis and expression of a recAgene required for extreme radioresistance of Deinococcus radio-durans. Gene 141:31-37.

9. Gutman, P. D., P. Fuchs, L. Ouyang, and K. W. Minton. 1993.Identification, sequencing, and targeted mutagenesis of a DNApolymerase gene required for the extreme radioresistance ofDeinococcus radiodurans. J. Bacteriol. 175:3581-3590.

10. Hanson, M. T. 1982. Rescue of mitomycin C- or psoralen-inactivated Micrococcus radiodurans by additional exposure toradiation or alkylating agents. J. Bacteriol. 152:976-982.

11. Harnett, C., E. L. Neidle, K. Ngai, and L. N. Ornston. 1990. DNA

VOL. 176, 1994

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

7446 UDUPA ET AL.

sequences of genes encodingAcinetobacter calcoaceticus protocat-echuate 3,4-dioxygenase: evidence indicating shuffling of genesand of DNA sequences within genes during their evolutionarydivergence. J. Bacteriol. 172:956-966.

12. Minton, K. W. 1994. DNA repair in the extremely radioresistantbacterium Deinococcus radiodurans. Mol. Microbiol. 13:9-15.

13. Minton, K. W. (Uniformed Services University of the HealthSciences). 1994. Personal communication.

14. Moseley, B. E. B. 1967. The isolation and some properties ofradiation sensitive mutants of Micrococcus radiodurans. J. Gen.Microbiol. 49:293-300.

15. Moseley, B. E. B. 1967. The repair of DNA in Micrococcusradiodurans following ultraviolet irradiation. J. Gen. Microbiol.48:4-24.

16. Moseley, B. E. B. 1983. Photobiology and radiobiology of Micro-coccus (Deinococcus) radiodurans. Photochem. Photobiol. Rev.7:223-275.

17. Moseley, B. E. B., and H. J. R. Copland. 1975. Isolation andproperties of a recombination-defective mutant of Micrococcusradiodurans. J. Bacteriol. 121:422-428.

18. Moseley, B. E. B., and H. J. R. Copland. 1978. Four mutants ofMicrococcus radiodurans defective in the ability to repair DNAdamaged by mitomycin C, two of which have wild-type resistanceto ultraviolet radiation. Mol. Gen. Genet. 160:331-337.

19. Moseley, B. E. B., and D. M. Evans. 1983. Isolation and propertiesof strains of Micrococcus (Deinococcus) radiodurans unable to exciseultraviolet light-induced pyrimidine dimers from DNA: evidenceof two excision pathways. J. Gen. Microbiol. 129:2437-2445.

20. Moseley, B. E. B., and A. Mattingly. 1971. Repair of irradiatedtransforming deoxyribonucleic acid in wild-type and a radiation-sensitive mutant ofMicrococcus radiodurans. J. Bacteriol. 105:976-983.

21. Raleigh, E. A., N. E. Murray, H. Revel, R. M. Blumenthal, D.Westaway, A. D. Reith, P. W. J. Rigby, J. Elhai, and D. Hanahan.1988. McrA and McrB restriction phenotypes of some E. colistrains and implications for gene cloning. Nucleic Acids Res.16:1563-1575.

22. Sakai, Y., E. Karu, S. Linn, and H. Echols. 1973. Purification andproperties of the gamma-protein specified by bacteriophage lamb-da: an inhibitor of the host RecBC recombination enzyme. Proc.Natl. Acad. Sci. USA 70:2215-2219.

23. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecularcloning: a laboratory manual, 2nd ed., p. 1.25-1.28. Cold SpringHarbor Laboratory Press, Cold Spring Harbor, N.Y.

24. Smith, M. D., C. L. Masters, and B. E. B. Moseley. 1992. Molecularbiology of radiation resistant bacteria, p. 258-280. In R. A.Herbert and R. J. Sharp (ed.), Molecular biology and biotechnol-ogy of extremophiles. Chapman & Hall, New York.

25. Tempest, P. R., and B. E. B. Moseley. 1980. Defective excisionrepair in a mutant of Micrococcus radiodurans hypermutable bysome monofunctional alkylating agents. Mol. Gen. Genet. 179:191-199.

26. Tigari, S., and B. E. B. Moseley. 1980. Transformation in Micro-coccus radiodurans: measurement of various parameters and evi-dence for multiple independently segregating genomes per cell. J.Gen. Microbiol. 119:287-296.

27. Varghese, A. J., and R. S. Day. 1970. Excision of cytosine-thymineadduct from the DNA of ultraviolet-irradiated Micrococcus radio-durans. Photochem. Photobiol. 11:511-517.

28. Vukovic-Nagy, B., B. W. Fox, and M. Fox. 1974. The release ofDNA fragments after X-irradiation of Micrococcus radiodurans.Int. J. Radiat. Biol. 25:329-337.

29. Udupa, K., and J. R. Battista. 1994. Unpublished data.

J. BACTERIOL.

on Septem

ber 2, 2020 by guesthttp://jb.asm

.org/D

ownloaded from