DNA Double-Strand Break Repair at �15°C

Markus Dieser, John R. Battista, Brent C. Christner

Louisiana State University, Department of Biological Sciences, Baton Rouge, Louisiana, USA

The survival of microorganisms in ancient glacial ice and permafrost has been ascribed to their ability to persist in a dormant,metabolically inert state. An alternative possibility, supported by experimental data, is that microorganisms in frozen matricesare able to sustain a level of metabolic function that is sufficient for cellular repair and maintenance. To examine this experimen-tally, frozen populations of Psychrobacter arcticus 273-4 were exposed to ionizing radiation (IR) to simulate the damage in-curred from natural background IR sources in the permafrost environment from over �225 kiloyears (ky). High-molecular-weight DNA was fragmented by exposure to 450 Gy of IR, which introduced an average of 16 double-strand breaks (DSBs) perchromosome. During incubation at �15°C for 505 days, P. arcticus repaired DNA DSBs in the absence of net growth. Based onthe time frame for the assembly of genomic fragments by P. arcticus, the rate of DNA DSB repair was estimated at 7 to 10 DSBsyear�1 under the conditions tested. Our results provide direct evidence for the repair of DNA lesions, extending the range ofcomplex biochemical reactions known to occur in bacteria at frozen temperatures. Provided that sufficient energy and nutrientsources are available, a functional DNA repair mechanism would allow cells to maintain genome integrity and augment micro-bial survival in icy terrestrial or extraterrestrial environments.

The results from decades of studying the stability of DNA in bac-teria may be summarized in two uncomplicated axioms. First,

genomic DNA is not inert; the genetic material will degrade withtime, and damage will accumulate unless that degradation is repaired.And second, if this degradation is not reversed, the cell will die. Therepair processes that cope with DNA damage are integral to life, and itis assumed that all species express means of maintaining genetic in-tegrity.

The importance of DNA repair to cell viability motivated ex-tensive study of the biochemistry and molecular biology of theserepair processes in model organisms such as Escherichia coli (1).Although the advantages of using this species for such studies arewell documented, there is a question as to whether detailed knowl-edge of DNA repair in E. coli as observed under standard labora-tory conditions adequately represents the full spectrum of bacte-rial responses to DNA damage. For instance, DNA repair is anenergy-intensive activity (2, 3), but microbial species in naturerarely experience optimal conditions for growth or have access tonutrient excess. Without a generous supply of energy, cells mayneed to generate a more measured response, rationing availableresources in a manner that prioritizes the most important tasks forsurvival. Moreover, viability in the absence of cellular reproduc-tion is sustained only by cells that do not accumulate a level ofdamage (e.g., to DNA) that exceeds a threshold beyond whicheffective repair is no longer possible.

Genomic DNA and viable bacteria are preserved in ancient iceand permafrost for hundreds of thousands to millions of years(4–8). The presence of a solute-rich liquid phase within the icematrix has been argued to be a habitat suitable for microorgan-isms (9–11). Furthermore, experimental data support the hypoth-esis that certain bacteria and fungi are metabolically active underfrozen conditions (12), and there is currently no evidence for aminimum temperature threshold for metabolism (13). Amato etal. (14) argued that the low rates of DNA synthesis observed inbacteria at �15°C would be sufficient to offset DNA damage in-curred by natural background ionizing radiation (IR) over geo-logical timescales. Additionally, Johnson et al. (8) found microbial

DNA preserved in permafrost samples as old as 600 kiloyears (ky)and speculated that active DNA repair by certain bacterial phylawas the most likely explanation for the DNA integrity observed.

Information on the metabolic capabilities of microorganismsat very low temperatures has provided a new perspective on mi-crobial survival strategies in the Earth’s cryosphere (12, 15, 16);however, specific knowledge about their physiology under frozenconditions is still very limited. More precisely, while it is clear thatDNA damage and repair are highly relevant to microbial survival,it is not known if microorganisms have the capacity to repair DNAdamage under the conditions existing in an ice matrix (i.e., highionic strength and low water activity and temperature). Here wereport on experiments in which populations of Psychrobacterarcticus 273-4 were exposed to IR, and subsequently, the cells re-paired lesions in their genomic DNA at �15°C. We discuss thebroader relevance of how a functional pathway of double-strandbreak (DSB) repair would be superior to dormancy in alleviatingthe problem of incurring formidable damage to DNA over ex-tended time frames and contribute to microbial persistence inglacial ice and permafrost environments.

MATERIALS AND METHODSBacterial strain and culture conditions. Pure cultures of Psychrobacterarcticus 273-4 (DSM 17307; equivalent to VKM B-2377) were grown aer-obically with shaking (200 rpm) at 22°C in R2A broth (17). Cells from themid-exponential phase were harvested by centrifugation at 5,000 � g,washed twice in phosphate-buffered saline (PBS; pH 7.2), and suspendedin either PBS or 0.8% R2A broth. The viable cell concentration in eachsample was determined by standard dilution plating.

Cells for the DNA repair experiments were suspended in an ice-cold

Received 22 August 2013 Accepted 25 September 2013

Published ahead of print 27 September 2013

Address correspondence to Brent C. Christner, xner@lsu.edu.

Copyright © 2013, American Society for Microbiology. All Rights Reserved.

doi:10.1128/AEM.02845-13

7662 aem.asm.org Applied and Environmental Microbiology p. 7662–7668 December 2013 Volume 79 Number 24

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

R2A medium that was diluted to 0.8% of the standard recipe (i.e., a finalconcentration of �10 mg C liter�1). Triplicate samples were prepared foreach time point and assay. One- or two-milliliter samples of the cell sus-pensions were frozen by transferring the tubes to a �80°C freezer andstoring the samples for at least 16 h. After exposure to IR, the samples weretransferred to a �15 � 1°C incubator (Thermo Scientific Revco ULTmodel 350-3-A32 freezer) for up to 505 days. At each experimental timepoint, the viable cell count was determined using standard dilution plat-ing; dilutions were performed in PBS, and the number of CFU was deter-mined in triplicate following incubation at 22°C for 3 days on R2A me-dium solidified with 1.5% agar.

Exposure to ionizing radiation. To establish the sensitivity of P. arcti-cus to IR, cells suspended in PBS were exposed to IR at a rate of 5.0 � 0.14Gy min�1 for 30, 60, and 90 min at 22°C, using a 60Co irradiator (model484; J. L. Sheppard and Associates). The dose rate was directly measuredby Fricke dosimetry (18). Survival was determined at each time point bytriplicate serial dilution plating of the samples on agar-solidified R2Amedium. The surviving fraction at each dose was calculated by dividingthe average number of CFU formed postirradiation by the number ofCFU in the control. The dose response of P. arcticus to IR was expressed asthe D37 value (i.e., the dose at which 37% of the cells survived) and wascalculated from the slope of the survival curve (Fig. 1).

Individual samples (n � 48) of the frozen cell suspensions were placedin a glass beaker, overlain with dry ice, and transported from the �80°Cfreezer to the 60Co irradiator in an insulated cooler containing ice packschilled to �80°C. In the irradiation chamber, the beaker was surroundedby the ice packs to keep the samples at low temperature during irradiation.Cell populations of P. arcticus were exposed to 450 Gy of IR, and allsamples remained frozen during the 90-min exposure. Identical samplesthat were not irradiated served as controls. After irradiation, the tubeswere sorted and boxed within a �20°C walk-in freezer and were trans-ferred to a �15 � 1°C incubator at �0.5 h postirradiation.

DNA extraction and PFGE. At each experimental time point, three sam-ples were removed from the freezer, melted at 22°C, and immediately pre-pared for DNA extraction. The cells were harvested by centrifugation, sus-pended in a prewarmed buffer (50°C; 10 mM Tris, 50 mM EDTA, 20 mMNaCl; pH 7.2), and embedded in 1.6% low-melting-point agarose (50°C; 1%final concentration) within 15 min of removal from the freezer. To enhancepolymerization, the plugs were incubated at 4°C for 10 min. Cell lysis wasperformed by placing the low-melting-point agarose plugs in a 0.5 M EDTA

(pH 8.0) solution containing 1 mg ml�1 proteinase K, 1% (wt/vol) lauroylsarcosine, and 1% (wt/vol) sodium dodecyl sulfate and incubating them for24 h at 50°C. After lysis, the plugs were washed twice in 5 ml of deionizedwater and four times in TE buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, pH8.0) for 30 min with agitation. A 2-mm slice was cut from each plug, and thechromosomal DNA was digested with 25 units of NotI in 1� RE buffer (60mM Tris-HCl, pH 7.9, 1.5 M NaCl, 60 mM MgCl2, 10 mM dithiothreitol)supplemented with bovine serum albumin (BSA; final concentration, 0.1 mgml�1) (all from Promega) for 48 h at 50°C. After restriction enzyme digestion,the plugs were washed twice in 5 ml of TE buffer as described above. A DNAstandard (Saccharomyces cerevisiae chromosomal DNA; Lonza) was includedto facilitate sizing of the fragments and comparisons of migration patternsbetween individual gels. High-molecular-weight DNAs were separated bypulsed-field gel electrophoresis (PFGE) on a 1% agarose gel containing 0.5�TBE (45 mM Tris-borate, 1 mM EDTA, pH 8.0) at 6 V cm�1, with a 120°angle and a ramp switch time from 50 to 90 s over 26 h at 12°C. The separatedDNAs were visualized by staining with ethidium bromide. The gel was digi-tally photographed, and electropherograms were obtained and analyzed withthe Bio-Rad Quantity One (version 4.2) software package. Background wassubtracted using the software’s rolling disk method, and the width and heightof the automatically detected peaks were adjusted manually (Quantity Oneuser guide for version 4, Bio-Rad). The shape of each band conformed to aGaussian model, which was used to mathematically describe the PFGE pro-files. The data from each electropherogram were normalized to the highestpixel intensity (i.e., expressed as 100% intensity), allowing individual peaks tobe distinguished and compared more easily.

The frequency at which DNA DSBs were generated in the 2.65-Mbgenome of P. arcticus (19) was estimated based on a rate of 1 DSB per 10Gy per 5 � 109 Da of double-stranded DNA (dsDNA) (20). In the case ofP. arcticus, 29 Gy would generate one DSB per chromosome, and 450 Gy(i.e., the experimental dosage) would introduce 16 DSBs. Similarly, theaverage number of DSBs that would be expected in each NotI restrictionproduct was calculated (Table 1). The kinetics of DNA repair was deter-mined using the time-dependent appearance of intact chromosomal frag-ments in the electropherograms. Using this approach, the average cumu-lative number of DSBs repaired from triplicate samples was calculated ateach time point. For example, if an average of 16 DSBs must be repaired toassemble the 2.65-Mb P. arcticus chromosome, then the repair of 6 DSBswould be required for a DNA fragment of 0.997 Mb (Table 1, data forband 7). The rate of DSB repair was determined from the slope of a fittedlinear regression model. To represent the variability observed betweensample replicates, the minimum and maximum numbers of DNA DSBsrepaired by 505 days were used to calculate low and high rate estimates forthe number of DNA DSBs repaired.

RESULTS

The survival of P. arcticus after exposure to 150, 300, and 450 Gy ofIR followed an exponential-decay function (Fig. 1). From these

TABLE 1 Predicted sizes of NotI restriction fragments of thePsychrobacter arcticus chromosome and estimated average numbers ofDNA DSBs introduced into each region after exposure to 450 Gy ofionizing radiation

BandFragmentsize (kb)

Avg no. ofDNA DSBs

1 33 �12 114 �13 229 14 305 25 375 26 598 47 997 6Genome 2,650 16

FIG 1 Survival curve of Psychrobacter arcticus as a function of ionizing radia-tion (IR) dose on a logarithmic scale. Data points represent means and stan-dard deviations for triplicate plate counts. The dotted vertical line representsthe D37 value, and the dashed line represents average numbers of double-strand breaks (DSBs) introduced into the genome of P. arcticus. The upper xaxis shows the predicted time frame in permafrost to receive the IR dosesapplied in this study, based on the measurements of McKay (2 mGy year�1)(37).

DNA Repair at �15°C

December 2013 Volume 79 Number 24 aem.asm.org 7663

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

data, the D37 of P. arcticus was determined to be 253 Gy, whichrepresents the dosage at which each cell, on average, experienced alethal event. The D37 dosage corresponded to the formation of anaverage of 9 DSBs, and based on the linear accumulation of DNADSBs (i.e., 1 DSB per 10 Gy per 5 � 109 Da of dsDNA [20]), 16DSBs, on average, were introduced into the chromosome of P.arcticus after exposure to 450 Gy of IR (Fig. 1).

A 9-fold reduction in culturability of P. arcticus cells occurredas a result of freezing in low-nutrient medium (0.8% R2A me-

dium) at �80°C and thawing at 22°C (Fig. 2B). The additive effectof exposure to 450 Gy reduced the P. arcticus population an addi-tional 12-fold (Fig. 2B). Long-term incubation at �15°C showedno measureable difference in culturability of irradiated and con-trol samples (unpaired two-tailed t test; P � 0.369). Cell survivalover 505 days at �15°C was biphasic, with an initial 80-day expo-nential decay in which approximately 80% of the population lostculturability, followed by a plateau in the residual fraction (Fig.2A). The data from the irradiated and control samples between 80and 505 days were not statistically different (one-way analysis ofvariance [ANOVA]; P � 0.161), indicating that a subpopulationof the cells (�20%) were more resistant to the effects of storage at�15°C.

Pulsed-field gel electrophoresis was used to visualize high-mo-lecular-weight DNA molecules isolated from populations of P.arcticus (Fig. 3). Digestion of the chromosomal DNA with NotIfacilitated the movement of large-molecular-weight DNA into thegel and provided a means of monitoring the reassembly of chro-mosomal DNA following the introduction of IR-induced DSBs.Restriction of the intact chromosome with NotI generated sevenfragments (Fig. 3 and 4A) that corresponded in size to predictionsbased on the genome sequence. The lowest-molecular-weightband (33 kb) was difficult to resolve under the electrophoreticconditions used. In addition to the seven expected genome frag-ments, three additional bands (700, 1,600, and 2,200 kb) (Fig. 3)were regularly observed that were consistent with incompletelydigested products (e.g., bands 4 and 5 are neighboring fragmentson the chromosome that together form an �700-kb molecule).Exposure to 450 Gy of IR led to a drastic change in the pattern ofP. arcticus genome fragments as observed by PFGE, reducing theresolvable chromosomal DNA to a population of molecules be-tween 30 and 200 kb (Fig. 3 and 4A). This was consistent withdose-rate calculations that predicted the introduction of 16 DSBs

FIG 2 Effect of freezing on survival of control (�) and irradiated (�) (450 Gy)cell suspensions of Psychrobacter arcticus incubated at �15°C for 505 days. Thedata are shown as fractions of culturable cells (A) and numbers of CFU ml�1

(B), with the square (�) representing the CFU ml�1 prior to freezing andirradiation. Error bars represent standard deviations from the means for trip-licate measurements.

FIG 3 Pulsed-field gel electrophoresis of NotI-restricted genomic DNA from Psychrobacter arcticus after exposure to 450 Gy of IR and subsequent incubation at�15°C. Fragment sizes in kilobases (kb) and number designations of the NotI-digested bands (1 to 7) are given on the left. The question marks indicateincompletely digested DNA fragments. No IR, chromosomal DNAs from frozen samples that were not exposed to IR. Samples irradiated with 450 Gy andexamined after 0, 30, 80, 155, 245, 310, 400, and 505 days of frozen incubation are indicated. Duplicate samples are shown for each time point. Each picture is anegative image of an ethidium bromide-stained gel.

Dieser et al.

7664 aem.asm.org Applied and Environmental Microbiology

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

into the 2.65-Mb chromosome of P. arcticus, producing an aver-age fragment size of �170 kb. Based on the size of each NotIfragment, the average numbers of DSBs introduced after exposureto 450 Gy ranged from less than one to six DSBs (Table 1).

During the first 155 days, the DNA remained fragmented, mi-grating as a low-molecular-weight smear (Fig. 3). Bands 3 to 5,

corresponding to the smaller NotI chromosomal fragments (229,305, and 375 kb), appeared in some PFGE profiles at 245 days, andby 310 days, these fragments were clearly resolved in all replicates(Fig. 3 and 4A). Although bands 1 and 2 (33 and 114 kb) could bevisualized by PFGE, they were masked in the electropherogramsdue to the weak contrast between the fragments and the large pool

FIG 4 (A) Pulsed-field gel electropherograms from a Psychrobacter arcticus population after exposure to 450 Gy of IR and long-term incubation at �15°C. Eachtime point consists of 3 plots derived from independent samples. Labels 1 to 7 indicate the resolvable NotI digestion fragments and coincide with the banddesignations in Table 1 and Fig. 3. (B) DNA DSB repair kinetics. Solid line, regression line for DSBs repaired, with standard deviations of the data (error bars);dotted lines, range of DSBs repaired, based on the minimum and maximum numbers of NotI bands observed after 505 days. MW, molecular weight.

DNA Repair at �15°C

December 2013 Volume 79 Number 24 aem.asm.org 7665

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

of low-molecular-weight DNA. Since NotI fragments 1 to 5 (33 to375 kb) required an average of two or fewer DSB repair events forreassembly (Table 1), these were the first genome segments ob-served, as would be expected if active DNA repair was occurring.The two highest-molecular-weight bands (598 and 997 kb) werenot observed before 505 days (Fig. 3 and 4A).

The electropherograms generated from the PFGE images wereused to construct Gaussian models (Fig. 4A) that characterizedeach band and aided in describing the kinetics of DNA DSB repair(Fig. 4B). After 155 days of incubation at �15°C, measureablerepair and reassembly of damaged chromosomes had occurred.Using the cumulative number of DNA DSBs repaired in triplicatesamples from each time point, averages of 3, 6, and 13 DSBs wererepaired after 245, 310, and 505 days, respectively (Fig. 4B). Fromthe slope of the linear regression, a repair rate of 8 DSBs year�1 wascalculated. Also shown in Fig. 4B are rates of DSB repair based onthe minimum and maximum numbers of DNA DSBs repaired ineach replicate after 505 days (dotted lines). Either six or seven NotIfragments were observed after 505 days (Fig. 4A), correspondingto the repair of 12 and 16 DSBs, respectively. Based on these data,we infer that P. arcticus repaired between 7 and 10 DSBs year�1

under the conditions tested.

DISCUSSION

P. arcticus was isolated from a Siberian permafrost horizon thathad been frozen for 20 to 30 ky at temperatures of �10 to �12°C(21). Subsequently, this species has provided valuable insight intothe growth (22), genomic composition (19), and gene expression(23) of permafrost bacteria, as well as the effect of low water ac-tivity (24) on these bacteria. P. arcticus has one of the coldestreported growth temperatures (�10°C) (25), which is eclipsedonly by those of Psychromonas ingrahamii 37 (�12°C) (26), Col-wellia psychrerythraea strain 34H (�12°C) (27), and Planococcushalocryophilus Or1 (�15°C) (28). The genome of P. arcticus en-codes a complete complement of proteins involved in canonicalbacterial DNA repair, including base excision repair, nucleotideexcision repair, mismatch repair, and homologous recombination(HR) (19). Given the opportunity to function, these proteinsshould be capable of efficiently dealing with DNA damage as itappears. Survival in ancient permafrost and the capacity to growand metabolize at subzero temperatures (14, 24, 25) were thus theprimary rationales for using P. arcticus as a model for examining ifbacteria can repair their DNA under conditions physicochemi-cally similar to those found in situ, i.e., low temperature and or-ganic carbon concentrations (total of �10 mg C liter�1) similar tovalues reported for permafrost (29, 30).

Exposure to 450 Gy of IR introduced an average of 16 DSBsinto the P. arcticus genome, as well as other forms of DNA damagenot detectable by PFGE (e.g., single-strand lesions and damage tonitrogenous bases), and led to a 12-fold reduction in the numberof CFU ml�1. This dose was empirically determined to be theminimum exposure that created sufficient DSB damage to elimi-nate the resolvable NotI fragments (Fig. 3). The data from Fig. 3and 4 imply that reassembly of damaged chromosomal DNA didnot proceed immediately after exposure to IR, with a linear rate ofDSB repair detected only after 155 days of incubation at �15°C(Fig. 4B). The time frame for the initiation of genome assemblycoincided with a period in which the number of CFU stabilizedand remained statistically unchanged for �350 days (Fig. 2). Al-though the repair of DSBs had no appreciable effect on cell cul-

turability (Fig. 2), the sublethally damaged cells were assayed byplating on culture medium, which would have provided the oppor-tunity for them to repair their DNA prior to initiating colony forma-tion. This result brings a new perspective to the observations of Alurand Grecz (31) and Grecz et al. (32), who incubated cell suspensionsof E. coli for up to 12 months at �20°C and demonstrated that freez-ing caused single- and double-strand breaks in the genome whichwere associated with decreased survival. Unexpectedly, it was ob-served that the extent of DSBs began to decrease after 4 months post-freezing, and the authors speculated that “random reassociation andaggregation of the initial DNA fragments” was the most likely expla-nation for the results (32). Whether or not the DNA repair activity weobserved is a property exclusive to cold-adapted species or is morebroadly distributed in the microbial world represents fertile territoryfor further study.

The major, ubiquitous pathway of DNA DSB repair in bacteriais homologous recombination (HR) (33). Some species (e.g., Ba-cillus subtilis) also possess a rudimentary analog to the eukaryoticnonhomologous end-joining pathway (NHEJ) (34), and a modi-fied version of NHEJ, named alternative end joining (A-EJ), existsin E. coli (35). Recently, genome condensation under stress con-ditions was suggested as an additional DSB repair mechanism thatoperates in E. coli (36). Of these DNA DSB repair mechanisms,only genes involved in HR pathways have been identified in the P.arcticus genome (19). The P. arcticus inoculum used in our exper-iments was harvested from the exponential phase of growth, anddespite the removal of nutrients, DNA repair enzymes and paren-tal scaffolds for HR could have existed.

Amato et al. (14) examined DNA synthesis in frozen samples ofP. arcticus incubated with various nutrient amendments and mea-sured rates at �15°C that ranged from 20 to 1,625 bp cell�1 day�1.Although their study did not specifically examine DNA repair, itwas concluded that if suitable nutrients and redox couples werepresent, such a rate of DNA metabolism would be adequate foroffsetting the DNA damage incurred from natural and cosmicsources of IR in terrestrial and extraterrestrial permafrost. Basedon IR dose values in Siberian permafrost (�2 mGy year�1) (37),each cell in a dormant P. arcticus population would incur an av-erage of one DSB every 14,500 years, and the D37 would be reachedin 126,500 years (Fig. 1). As an extreme comparison, the time toreach the D37 would be as few as 305 years on the surface of Mars,which receives solar energetic protons and high-energy galacticcosmic rays (as well as germicidal short-wavelength UV radiation)at a rate estimated to be 830 mGy year�1 (38). In this end-memberscenario, IR would induce an average of one DSB per cell every 35years on the Martian surface, a frequency that is 280-fold lowerthan the rate of DSB repair measured in this study (Fig. 4B). Ourexperiments delivered a dose of IR in 90 min that would be equiv-alent to the cumulative dosage P. arcticus would be exposed toover 225 ky in permafrost (Fig. 1). As such, a more restrainedresponse to DNA damage may be ample for cellular maintenanceunder IR conditions that more closely resemble those found innature. It is also important to note that our estimates of DSBsinduced by IR were based on experiments with liquid cell suspen-sions (20) and therefore may not be directly applicable to thoseunder frozen conditions. However, the sizes of genome fragmentsobserved after exposure to 450 Gy of IR (Fig. 3) agreed well withpredicted values (Table 1). If our data and related calculations arean approximation of P. arcticus’s physiological potential in per-mafrost, a rate of 7 to 10 DSBs repaired year�1 at �15°C would be

Dieser et al.

7666 aem.asm.org Applied and Environmental Microbiology

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

more than sufficient to offset DSB damage resulting from IR interrestrial permafrost or ice on the surface of Mars. Although thisstudy focused on a cold-adapted bacterium from the cryosphere,our results may have broader implications for understanding mi-crobial longevity in the deep biosphere (e.g., see references 39 and40), where microbes persist on scant energy sources and the turn-over of microbial biomass is estimated to be hundreds to severalthousand years (41, 42).

In his review of the endogenous hydrolytic and oxidative pro-cesses that decompose DNA, Lindahl (43) described how genomicDNA would progressively accumulate damage when “deprived ofthe repair mechanisms provided in living cells.” DNA is inherentlyunstable, exhibiting spontaneous base loss, deamination, andstrand breaks. Although the rates of these processes are extremelylow under physiological conditions, they will eventually degradethe DNA to a point where information storage is no longer possi-ble. Lindahl concluded that under the best conditions, genomicDNA will degrade to short nonfunctional fragments within tens ofthousands of years, and the recovery of low-molecular-weightDNA fragments from long-dead animals and plants supports thisprediction (e.g., see reference 44 and references within). However,the rates of DNA decomposition described by Lindahl (43) are notrelevant in considering the longevity of genomic DNA if a cellexpresses DNA repair processes. We suggest that as long as cellsare capable of repair, the long-term effects of spontaneous decom-position or natural radioactive decay on their genomic DNA canbe delayed and cell survival extended.

Conclusions. This study demonstrated that P. arcticus has thecapacity to overcome damage induced to its genome by IR at a rate�100,000 times faster than DNA DSB lesions would occur in itsnative permafrost environment. This has expanded our under-standing of bacterial physiology at low temperature and providesdirect evidence of a mechanism that could increase the longevityof microbes in frozen matrices. Considering the instability ofDNA, the capacity to conduct metabolic activity under icy condi-tions is a long-term survival strategy that is superior to dormancy.The ability to synthesize (14) and repair (Fig. 4B) DNA at temper-atures relevant to Siberian permafrost and the northern polar re-gion of Mars (45) suggests that long-term survival of P. arcticusunder such conditions may be limited only by the water activityand availability of suitable redox couples and nutrients. In conclu-sion, our investigation of DNA repair in a permafrost bacteriumhas revealed an important factor pertinent to theoretical predic-tions of microbial longevity while frozen (37, 38) and is clearlyrelevant to discussions concerning the survival of microbial life inicy extraterrestrial environments.

ACKNOWLEDGMENTS

We thank Gregg S. Pettis for assistance with PFGE.This work was supported by grants to B.C.C. from the National

Aeronautics and Space Administration (grants NNX10AR92G andNNX10AN07A) and the Louisiana Board of Regents.

REFERENCES1. Persky NS, Lovett ST. 2008. Mechanisms of recombination: lessons from

E. coli. Crit. Rev. Biochem. Mol. Biol. 43:347–370.2. Kowalczykowski SC. 1991. Biochemistry of genetic recombination: ener-

getics and mechanism of DNA strand exchange. Annu. Rev. Biophys. Bio-phys. Chem. 20:539 –575.

3. Kowalczykowski SC, Dixon DA, Eggleston AAK, Lauder SD, RehrauerWM. 1994. Biochemistry of homologous recombination in Escherichiacoli. Microbiol. Rev. 58:401– 465.

4. Christner BC, Mosley-Thompson E, Thompson LG, Reeve JN. 2003.Bacterial recovery from ancient ice. Environ. Microbiol. 5:433– 436.

5. Christner BC, Royston-Bishop G, Foreman CM, Arnold BR, Tranter M,Welch KA, Lyons WB, Tsapin AI, Studinger M, Priscu JC. 2006.Limnological conditions in subglacial Lake Vostok, Antarctica. Limnol.Oceanogr. 51:2485–2501.

6. Bidle KD, Lee SH, Marchant DR, Falkowski PG. 2007. Fossil genes andmicrobes in the oldest ice on Earth. Proc. Natl. Acad. Sci. U. S. A. 104:13455–13460.

7. Gilichinsky DA, Wilson GS, Friedmann EI, Mckay CP, Sletten RS,Rivkina EM, Vishnivetskaya TA, Erokhina LG, Ivanushkina NE, Koch-kina GA, Shcherbakova VA, Soina VS, Spirina EV, Vorobyova EA,Fyodorov-Davydov DG, Hallet B, Ozerskaya SM, Sorokovikov VA,Laurinavichyus KS, Shatilovich A, Chanton JP, Ostroumov VE, TiedjeJM. 2007. Microbial populations in Antarctic permafrost: biodiversity,state, age, and implication for astrobiology. Astrobiology 7:275–311.

8. Johnson SS, Hebsgaard MB, Christensen TR, Mastepanov M, NielsenR, Munch K, Brand T, Gilbert MT, Zuber MT, Bunce M, Rønn R,Gilichinsky D, Froese D, Willerslev E. 2007. Ancient bacteria showevidence of DNA repair. Proc. Natl. Acad. Sci. U. S. A. 104:14401–14405.

9. Price PB. 2000. A habitat for psychrophiles in deep Antarctic ice. Proc.Natl. Acad. Sci. U. S. A. 97:1247–1251.

10. Barletta RE, Priscu JC, Mader HM, Jones WL, Roe CH. 2012. Chemicalanalysis of ice vein microenvironments. II. Analysis of glacial samplesfrom Greenland and Antarctica. J. Glaciol. 58:1109 –1118.

11. Dani KGS, Mader HM, Wahman JL, Wolff EW. 2012. Modelling theliquid-water vein system within polar ice sheets as a potential microbialhabitat. Earth Planet. Sci. Lett. 333–334:238 –249.

12. Doyle S, Dieser M, Broemsen E, Christner B. 2012. General charac-teristics of cold-adapted microorganisms, p 103–125. In Whyte L,Miller RV (ed), Polar microbiology: life in a deep freeze. ASM Press,Washington, DC.

13. Price PB, Sowers T. 2004. Temperature dependence of metabolic rates formicrobial growth, maintenance, and survival. Proc. Natl. Acad. Sci.U. S. A. 101:4631– 4636.

14. Amato P, Doyle SM, Battista JR, Christner BC. 2010. Implications ofsubzero metabolic activity on long-term microbial survival in terrestrialand extraterrestrial permafrost. Astrobiology 10:789 –798.

15. de Lorenzo V. 2011. Genes that move the window of viability of life:lessons from bacteria thriving at cold extremes. Bioessays 33:38 – 42.

16. Verde C, di Prisco G, Giordano D, Russo R, Anderson D, Cowan D.2012. Antarctic psychrophiles: models for understanding the molecularbasis of survival at low temperature and response to climate change. Bio-diversity 13:249 –256.

17. Reasoner DJ, Geldreich EE. 1985. A new medium for the enumerationand subculture of bacteria from potable water. Appl. Environ. Microbiol.49:1–7.

18. Fricke H, Hart EJ. 1966. Chemical dosimetry, p 167–239. In Attix FH,Roesch WC (ed), Radiation dosimetry. Academic Press, New York, NY.

19. Ayala-del-Río H, Chain PS, Grzymski JJ, Ponder MA, Ivanova N,Bergholz P, Di Bartolo G, Hauser L, Land M, Bakermans C, RodriguesD, Klappenbach J, Zarka D, Larimer F, Richardson P, Murray A,Thomashow M, Tiedje JM. 2010. The genome sequence of Psychrobacterarcticus 273-4, a psychroactive Siberian permafrost bacterium, revealsmechanisms for adaptation to low-temperature growth. Appl. Environ.Microbiol. 76:2304 –2312.

20. Burrell AD, Dean CJ. 1975. Repair of double-strand breaks in Micrococ-cus radiodurans. Basic Life Sci. 5B:507–512.

21. Vishnivetskaya T, Kathariou S, McGrath J, Gilichinsky D, Tiedje JM.2000. Low temperature recovery strategies for the isolation of bacteriafrom ancient permafrost sediments. Extremophiles 4:165–173.

22. Bakermans C, Tsapin AI, Souza-Egipsy V, Gilichinsky DA, NealsonKH. 2003. Reproduction and metabolism at �10°C of bacteria isolatedfrom Siberian permafrost. Environ. Microbiol. 5:321–326.

23. Bergholz PW, Bakermans C, Tiedje JM. 2009. Psychrobacter arcticus273-4 uses resource efficiency and molecular motion adaptations for sub-zero temperature growth. J. Bacteriol. 191:2340 –2352.

24. Ponder MA, Thomashow MF, Tiedje JM. 2008. Metabolic activity ofSiberian permafrost isolates, Psychrobacter arcticus and Exiguobacteriumsibiricum, at low water activities. Extremophiles 12:481– 490.

25. Bakermans C, Ayala-del-Río HL, Ponder MA, Vishnivetskaya T,Gilichinsky D, Thomashow MF, Tiedje JM. 2006. Psychrobacter cryoha-

DNA Repair at �15°C

December 2013 Volume 79 Number 24 aem.asm.org 7667

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from

lolentis sp. nov. and Psychrobacter arcticus sp. nov., isolated from Siberianpermafrost. Int. J. Syst. Evol. Microbiol. 56:1285–1291.

26. Auman AJ, Breezee JL, Gosink JJ, Kaempfer P, Staley JT. 2006. Psy-chromonas ingrahamii sp. nov., a novel gas vacuolate, psychrophilic bac-terium isolated from Arctic polar sea ice. Int. J. Syst. Evol. Microbiol.56:1001–1007.

27. Wells LE, Deming JW. 2006. Characterization of a cold-active bacterio-phage on two psychrophilic marine hosts. Aquat. Microb. Ecol. 45:15–29.

28. Mykytczuk NCS, Foote SJ, Omelon CR, Southam G, Greer CW, WhyteLG. 2013. Bacterial growth at �15°C; molecular insights from the perma-frost bacterium Planococcus halocryophilus Or1. ISME J. 7:1211–1226.

29. Rivkina E, Laurinavichius K, McGrath J, Tiedje J, Shcherbakova V,Gilichinsky. 2004. Microbial life in permafrost. Adv. Space Res. 33:1215–1221.

30. Douglas TA, Fortier D, Shur YL, Kanevskiy MZ, Guo L, Cai Y, BrayMT. 2011. Biogeochemical and geocryological characteristics of wedgeand thermokarst-cave ice in the CRREL permafrost tunnel, Alaska. Per-mafrost Periglac. Processes 22:120 –128. doi:10.1002/ppp.709.

31. Alur MD, Grecz N. 1975. Mechanism of injury of Escherichia coli byfreezing and thawing. Biochem. Biophys. Res. Commun. 62:308 –312.

32. Grecz N, Hammer TL, Robnett CJ, Long MD. 1980. Freeze-thaw injury:evidence for double strand breaks in Escherichia coli DNA. Biochem. Bio-phys. Res. Commun. 93:1110 –1113.

33. Wigley DB. 2013. Bacterial DNA repair: recent insights into the mecha-nism of RecBCD, AddAB and AdnAB. Nat. Rev. Microbiol. 11:9 –13.

34. Alonso JC, Cardenas PP, Sanchez H, Hejna J, Suzuki Y, Takeyasu K.2013. Early steps of double-strand break repair in Bacillus subtilis. DNARepair 12:162–176.

35. Chayot R, Montagne B, Mazel D, Ricchetti M. 2010. An end-joiningrepair mechanism in Escherichia coli. Proc. Natl. Acad. Sci. U. S. A. 107:2141–2146.

36. Shechter N, Zaltzman L, Weiner A, Brumfeld V, Shimoni E, Fridmann-Sirkis Y, Minsky A. 2013. Stress-induced condensation of bacterial ge-nomes results in re-pairing of sister chromosomes: implications for dou-ble-strand DNA break repair. J. Biol. Chem. 288:25659 –25667. doi:10.1074/jbc.M113.473025.

37. McKay CP. 2001. The deep biosphere: lessons for planetary exploration, p315–327. In Frederickson JK, Fletcher M (ed), Subsurface microbiologyand biogeochemistry. Wiley-Liss, New York, NY.

38. Dartnell LR, Desorgher Ward LM, Coates AJ. 2007. Modelling thesurface and subsurface Martian radiation environment: implications forastrobiology. Geophys. Res. Lett. 34:L02207.

39. Morono Y, Terada T, Nishizawa M, Ito M, Hillion F, Takahata N,Sanoe Y, Inagaki F. 2011. Carbon and nitrogen assimilation in deepsubseafloor microbial cells. Proc. Natl. Acad. Sci. U. S. A. 108:18295–18300.

40. Roussel EG, Cambon Bonavita MA, Querellou J, Cragg BA, Webster G,Prieur D, Parkes RJ. 2008. Extending the sub-sea-floor biosphere. Science320:1046. doi:10.1126/science.1154545.

41. Biddle JF, Lipp JS, Lever MA, Lloyd KG, Sorensen KB, Anderson R,Fredricks HF, Elvert M, Kelly TJ, Schrag DP, Sogin ML, Brenchley JE,Teske A, House CH, Hinrichs KU. 2006. Heterotrophic Archaea domi-nate sedimentary subsurface ecosystems off Peru. Proc. Natl. Acad. Sci.U. S. A. 103:3846 –3851.

42. Jørgensen BB. 2011. Deep subseafloor microbial cells on physiologicalstandby. Proc. Natl. Acad. Sci. U. S. A. 108:18193–18194.

43. Lindahl T. 1993. Instability and decay of the primary structure of DNA.Nature 362:709 –715.

44. Willerslev E, Cooper A. 2005. Ancient DNA. Proc. Biol. Sci. 272:3–16.45. Jakosky BM, Nealson KH, Bakermans C, Ley RE, Mellon MT. 2003.

Subfreezing activity of microorganisms and the potential habitability ofMars’ polar regions. Astrobiology 3:343–350.

Dieser et al.

7668 aem.asm.org Applied and Environmental Microbiology

on February 6, 2020 by guest

http://aem.asm

.org/D

ownloaded from