Review Article Understanding DNA Repair in Hyperthermophilic Archaea: Persistent … · 2019. 7. 31. · DennisW.Grogan Department of Biological Sciences, University of Cincinnati,

Review ArticleUnderstanding DNA Repair in Hyperthermophilic ArchaeaPersistent Gaps and Other Reasons to Focus on the Fork

Dennis W Grogan

Department of Biological Sciences University of Cincinnati 614 Rieveschl Hall Clifton Court Cincinnati OH 45221-0006 USA

Correspondence should be addressed to Dennis W Grogan grogandwucmailucedu

Received 26 February 2015 Accepted 21 May 2015

Academic Editor Yoshizumi Ishino

Copyright copy 2015 Dennis W Grogan This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Although hyperthermophilic archaea arguably have a great need for efficient DNA repair they lack members of several DNA repairprotein families broadly conserved among bacteria and eukaryotes Conversely the putative DNA repair genes that do occur inthese archaea often do not generate the expected phenotype when deletedThe prospect that hyperthermophilic archaea have someunique strategies for coping with DNA damage and replication errors has intellectual and technological appeal but resolving thisquestion will require alternative copingmechanisms to be proposed and tested experimentallyThis review evaluates a combinationof four enigmatic properties that distinguishes the hyperthermophilic archaea from all other organisms DNApolymerase stalling atdU apparent lack of conventional NER lack of MutSL homologs and apparent essentiality of homologous recombination proteinsHypothetical damage-coping strategies that could explain this set of properties may provide new starting points for efforts to definehow archaea differ from conventional models of DNA repair and replication fidelity

1 Genome Integrity and Archaea

The importance of maintaining the integrity of cellular gen-omes and the diversity of processes that threaten it canbe seen in the network of sophisticated mechanisms thatcope with DNA damage and replication errors in even thesimplest microorganisms Consistent with their deep inte-gration into cellular biology damage-coping strategies showboth conservation and divergence eukaryotes and bacteriaemploy similar sets of basic coping strategies for example butthese differ in their mechanistic details The similarities mayreflect the common biological threats posed by DNA lesionsand mutations whereas the differences point to the deepevolutionary divergence between these two fundamentallydifferent types of cells [1]

There are legitimate reasons to expect that archaeaespecially the hyperthermophiles differ from both bacteriaand eukaryotes with respect tomechanisms of genomemain-tenance As predicted by its distinct evolutionary history[2] the third domain of life shares certain fundamentalmolecular features only with eukaryotic cells others onlywith bacterial cells and has yet other features that areuniquely archaeal With respect to genome maintenance

archaea employ ldquoeukaryoticrdquo DNA replication proteins topropagate small circular chromosomes in the context ofa small ldquoprokaryoticrdquo cell [3ndash7] In addition many groupsof archaea are adapted to environmental extremes whichimpose molecular constraints not accommodated by themicroorganisms used historically as models of molecu-lar biology To the extent that archaea may have geneticmechanisms not represented in bacteria or eukaryotes theyrepresent a potential source of new information concerningearly cellular evolution as well as processes that may be usedfor biotechnology especially if these processes occur in thearchaea that are adapted to extreme conditions

Molecular mechanisms not known from other systemsare by nature however difficult to define and confirmingsuch mechanisms will require genetics including the con-struction and analysis of mutant strains Because hyper-thermophilic archaea (HA) are challenging to manipulateas living cells their genetic processes have been addressedoverwhelmingly via biochemical and structural analysesof purified proteins with little input from the comple-mentary and essential perspective of functional geneticsParaphrasing a sentiment once expressed by the authorrsquos

Hindawi Publishing CorporationArchaeaVolume 2015 Article ID 942605 12 pageshttpdxdoiorg1011552015942605

2 Archaea

doctoral research advisor ldquoWhen you do biochemistryyou tell the organism what you think is important whenyou do genetics the organism tells you what it thinks isimportantrdquo

2 The Functional-Genetic Era forHyperthermophilic Archaea

For HA basic genetic analyses have become feasible only inthe past decade and one of the first of these studies targetedthe reverse DNA gyrase of Thermococcus kodakaraensis[8] Reverse gyrases (Rgy) are single-strand-nicking DNAtopoisomerases that introduce positive supercoils into DNAand appear only in extreme- or hyperthermophilic archaeaand bacteria [9] Although its natural distribution suggestedto many that reverse gyrase activity is essential for life at hightemperature Atomi et al [8] successfully isolated a deletionmutant demonstrating that Rgy is not essential for Thermo-coccus viability The Thermococcus mutant nevertheless grewat a reduced rate and with a decreased maximal temperatureThus whereas Rgy appears not to be absolutely requiredfor cellular function inThermococcus the mutant phenotypeargued that it must play a role of some importanceThe latterconclusion has been reinforced by failure of later efforts todelete the corresponding genes of other hyperthermophilicarchaea [10]

In the ensuing decade progress in the genetic analysisof DNA metabolic enzymes of HA has been substantial buthard-won As in the Rgy study the archaeal mutants aretypically constructed by replacing the target gene with aselected marker and therefore represent unambiguous lossof gene function However the resulting genotype oftendoes not match that predicted by the bacterial or eukaryoticcounterpart On one hand this reinforces the evidenceof deep divergence separating archaea from bacteria andeukaryotes and the incentive for their analysis but on theother hand it intensifies the need to formulate and test thenext generation of hypotheses This review and commen-tary selectively interprets functional genetic data relatingto genome stability in archaea with an emphasis on thehyperthermophiles It proposes that certain patterns of resultsdo not fit well with the schemes of genome-maintenancemechanisms identified in eukaryotic or bacterial modelsand suggests ways in which atypical or primordial damage-coping strategies could in principle support fundamen-tally different strategies of genome maintenance in theHA

3 Mutant Phenotypes

Table 1 lists functional studies of HA genes that have beenimplicated in DNA repair or related processes (usually bysequence similarity) grouped according to the phenotypicoutcome that was observed Mutants placed in Group Iseem to be relatively straightforward to interpret since theirphenotypes indicate a role of the deleted gene in DNArepair or mutagenesis roughly in line with that predictedfrom its sequence Examples include DNA photolyases

at least one TLS polymerase Holliday-junction resolvasesa 31015840-flap endonuclease a recombinase homolog and twoputative uracil DNA glycosylases (UDG) In most cases thedocumented phenotype is increased sensitivity to one ormore DNA-damaging agents although the UDG mutantsshowed impaired growth [10] and the Y-family polymerasemutant showed an altered spectrumof spontaneousmutation[11]

Mutants in Group II identify genes with minimal impactonDNA repair capacity as evidenced by little or no detectableincrease in sensitivity to radiation or DNA-damaging chem-icals In some cases notably HJ resolvases Hjc and Hjesingle mutants exhibited only marginal phenotypes but thedouble mutant was apparently nonviable [12] The lattercase thus suggests that some cellular function of these twoenzymes is important and is fulfilled by two at least partiallyredundant enzymes An important feature of the Group IIcases (Table 1) is that most of them encode homologs ofeukaryotic NER proteins which were expected to participatein generalized repair of bulky helix-distorting DNA lesionsAlthough redundancy of function could in principle accountfor this result the corresponding eukaryotic genes typicallydo not exhibit such redundancy with respect to DNA repair[13]

Group III of Table 1 represents genes which could notbe deleted and thus may encode an essential functionAlthough the genetic tools for HA remain limited insome cases lethality could be further supported by specificgenetic tests Some of the proteins are predicted to playimportant roles in basic replication and thus their appear-ance in this group is fully plausible examples include thestructure-specific endonuclease Fen1 which is implicatedin maturation of Okazaki fragments Most of the genes inthis category are associated specifically with homologousrecombination which is however not an essential process innormal cells of bacteria unicellular eukaryotes ormesophilicarchaea

As noted in several of these studies the number ofputative DNA-repair genes that fall into the latter two cate-gories (Groups II and III) adds to the accumulating evidencethat HA do not follow the bacterial or eukaryotic modelsof DNA repair in any strict sense The available evidence(discussed below) nevertheless argues that they do replicatetheir genomes accurately and cope effectively with DNAdamage This poses the interesting but challenging opportu-nity of identifying precisely how the genome-maintenancemechanisms of HA differ from those defined historically inbacteria and eukaryotic cells This review and commentaryhighlights four enigmata of genome-stability mechanismsin HA some of which have persisted for two decades (i)the functional role of archaeal DNA polymerase stalling atdU (ii) the apparent lack of repair-specific functions for theHA homologs of eukaryotic NER proteins (iii) the specificabsence of MutSL homologs in HA and (iv) the essentialityof HR proteins in HA Although the interpretations providedhere are broad and speculative attention to unresolvedquestions such as these represents a logical prerequisite touncovering genetic processes that distinguish archaea andespecially HA from other forms of life

Archaea 3

Table 1 Gene-deletion studies in hyperthermophilic archaea

Gene ORF ID Organism Predicted function Phenotype ReferenceGroup I (relevant phenotype)

udg2 SiRe 0084 S islandicus Uracil DNA glycosylase Impaired growth [10]udg4 SiRe 1884 S islandicus Uracil DNA glycosylase Impaired growth [10]hefxpf TK1021 Tc kodakarensis Structure-specific endonuclease General damage sensitivity [34]hjmhel308 TK1332 Tc kodakarensis DNA helicase Sensitive to mitomycin C [34]phr Saci 1227 S acidocaldarius DNA photolyase Loss of photoreactivation [59]dbh Saci 0554 S acidocaldarius Y-family DNA polymerase Increased GC to TA [11]radC1 SiRe 0240 S islandicus RadA paralog Increased damage sensitivity [60]phrB SiRe 0261 S islandicus DNA photolyase Photorepair deficiency [10]

Group I1015840 (possible phenotype)xpf SiRe 1280 S islandicus 31015840-flap endonuclease Slow DNA replication [10]dpo4 S solfataricus Y-family DNA polymerase Possible sensitivity to cisplatin [61]xpb TK0928 Tc kodakarensis NER-related helicase Marginal damage sensitivity [36]xpd TK0784 Tc kodakarensis NER-related helicase Marginal damage sensitivity [36]

Group II (no observed impact)xpd SiRe 1685 S islandicus NER-related helicase None observed [10]xpb1 SiRe 1128 S islandicus NER-related helicase None observed [10]xpb2 SiRe 1526 S islandicus NER-related helicase None observed [10]bax1 SiRe 1524 S islandicus Nuclease None observed [10]uvde Saci 1096 S acidocaldarius Putative UV endonuclease No effect on UV survival [59]hjc TK1175 Tc kodakarensis Holliday-junction resolvase None observed [36]hjc SiRe 1431 S islandicus Holliday-junction resolvase None observed [10]hje SiRe 930 S islandicus Holliday-junction resolvase None observed [10]

Group II1015840 (synthetic lethality)hje + hjc SiRe 930 amp 1431 S islandicus Holliday-junction resolvases Double mutant is lethal [62]

Group III (presumed lethal)radA TK1899 Tc kodakarensis Recombinase Lethal (not recovered) [36]rad50 TK2211 Tc kodakarensis DSB end-processing Lethal (not recovered) [36]mre11 TK2212 Tc kodakarensis DSB end-processing Lethal (not recovered) [36]herA TK2213 Tc kodakarensis DSB end-processing Lethal (not recovered) [36]nurA TK2210 Tc kodakarensis DSB end-processing Lethal (not recovered) [36]xpgfen1 TK1281 Tc kodakarensis 51015840-flap endonuclease Lethal (not recovered) [36]radA SiRe 1747 S islandicus Recombinase Lethal [10]herA SiRe 0064 S islandicus Bipolar helicase Confirmed lethal [62 63]xpgfen1 SiRe 1830 S islandicus 51015840-flap endonuclease Lethal [10]nurA SiRe 0061 S islandicus DSB end-processing Confirmed lethal [10 62]rad50 SiRe 0062 S islandicus DSB end-processing Confirmed lethal [10 62]mre11 SiRe 0063 S islandicus DSB end-processing Confirmed lethal [10 62]hjm SiRe 0250 S islandicus DSB end-processing Lethal [62 64]

4 The Enigma of DNA PolymerasesBlocked by Template dU

Although early characterization of HA DNA polymerasesshowed that they have certain advantages for PCR relative toTaq DNA polymerase it was also found that HA polymerasesbind tightly to any dU they encounter in the templatestrand [14] which impedes PCR Driven by the practicalincentives to use HA DNA polymerases in PCR variousbiochemical techniques have been developed to overcomethis problem In addition to enabling PCR by HA DNA

polymerases these techniques illustrate the potential biolog-ical challenges posed by this uniquely archaeal property ofthese polymerases

One technique which promotes PCR by archaeal poly-merases involves supplementing the reaction mixture witha thermostable uracil DNA glycosylase (UDG) [15] BecauseUDG destroys the dU-containing strand as a usable templatein any subsequent cycle its beneficial impact implies thatdU binding does not merely slow the HA polymerase butsomehow removes a significant fraction of it from contribut-ing to DNA synthesis This suggests that the polymerasedU

4 Archaea

(a)

UU

U

dU inlagging-strandtemplate

dU inleading-strand

template

(substrate for BER)

U

U

(subject to breakage)

Fork regression

Polymerase uncoupling


(b)

(c)

Figure 1 Possible fates of a replicative polymerase stalled at template dU (a) Archaeal polymerases stall with the nascent 31015840 end 4 nt aheadof the template dU [23] (b) If the replication fork responds with immediate fork reversal the dU is restored to its original context and BERis required for resumption of fork progress (c) If tight coupling is not preserved the Mcm helicase and nonstalled polymerase (not drawn)would continue creating ssDNA region on the stalled template vulnerable to structure- or single-strand-specific nucleases Arrowheads onDNA strands represent 31015840 ends

complex is both unproductive and stable and raises questionsabout what impact such properties would have in vivo

Another technique involves supplementing the reactionmixture with a thermostable dUTPase [16] This enzymeremoves dUTP from the cellular pool of dNTPs thus pre-venting its incorporation into nascent DNA in vivo Theimpact of this technique suggests that much of the templatedU that the polymerases of HA encounter during PCRresults from cytosine deaminated before not after it wasincorporated into DNA This is consistent with in vitrostudies demonstrating limited discrimination against dUTPby DNA polymerases of HA [17] and raises the biologicallyrelevant question as to how well HA purge dUTP from theirnucleotide pools during chromosome replication A thirdapproach to improved PCR by these polymerases has beento mutate the specific dU-binding pocket which mediatesthe stalling This has yielded fully active and thermostablearchaealDNApolymeraseswhich are nevertheless insensitiveto the occurrence of dU in the template strand [18 19]

The only biological rationale offered consistently in theliterature for the specific stalling of the replicative poly-merases of HA at template dU is to avoid GC to AT transi-tion mutations The logic is admittedly simple and plausible(i) hydrolytic deamination of dC occurs spontaneously inDNA and is greatly accelerated by high temperature and (ii)most DNA polymerases (including those of HA) insert dAopposite the resulting dU thus creating a transitionmutationAlthough avoiding such mutation is clearly a benefit themagnitude of this benefit has not been weighed against therisks and costs which polymerase stalling imposes on thearchaeal cell nor against the benefits of conventional repairnor has it been evaluated critically in the wider context ofDNA metabolism

For example why does the property of polymerasestalling at dU correlate with phylogenetic domain not withgrowth temperature All the replicative DNA polymerases of

archaea that have been examined but none of the eukaryoticor bacterial enzymes stall at dU [20] This natural distri-bution thus seems to undermine the argument that HAneed polymerase stalling specifically to cope with a highrate of spontaneous cytosine deamination in DNA as hyper-thermophilic bacteria should have a similar need whereasmesophilic archaea should not It also seems striking thatD-family DNA polymerases which replicate genomes andoccur only in archaea are also inhibited by dU but througha different mechanism [21] Moreover there seems to be (atleast superficially) no evidence that HA are any less efficientin the conventional base-excision repair (BER) of dU thanthermophilic bacteria are Corresponding N-glycosylasessupported by AP lyases AP endonucleases and related BERenzymes are well represented in HA [22]

Other more mechanistic questions concern the fact thatthe stalled DNA polymerase precludes BER by sequesteringthe dU within the binding pocket of the polymerase [23]Repair thus requires some formof replication-fork reversal ordisassemblyHow this processmightwork remains difficult totest The apparently most favorable and commonly proposedoutcome is reversal of the fork allowing BER of the dU inits original context (Figure 1) Recently an archaeal DNAprimase has been found to bypass dU and other damagedbases in vitro [24] and this may allow bypass with mini-mal reversal or rearrangement of the fork Other plausiblescenarios can be proposed which have not been discussedin the literature these involve intermediate structures thatare vulnerable to single-strand or structure-specific nucle-ases leading potentially to breakage (or ldquocollapserdquo) of thereplication fork (Figure 1) As argued below fork reversaldisassembly or even breakage may be a routine processin HA chromosome replication and may underlie uniquelyarchaeal strategies to repairDNA lesions Even in this contexthowever the antimutagenesis argument fails to identify thenet advantage for archaea to invest deeply in a radical

Archaea 5

alternative to the relatively simple low-risk strategy of BERwhich these archaea also employ

More evidence of deeper complexity in the dU-stallingphenomenon comes from recent gene-deletion studiesGenome annotations depict S islandicus as having ensuredan adequate level of dU repair as it encodes no fewer thanfour putative UDG proteins However efforts to delete eachof the genes individually were successful in only two casesand the successful constructs showed impaired growth underlaboratory conditions [10] Since neither lethality nor growthimpairment is a characteristic of UDG mutants even ofsimilarly thermophilic bacteria [25] these results argue thatthe encoded Sulfolobus proteins have roles beyond the basicredundant BER of dU predicted by their annotation as fourUDGs

5 Do Archaea Need dU for a Specific Aspect ofNormal Chromosome Replication

As noted above deamination of dC is not the only sourceof dU it can also be inserted opposite dA via dUTPincorporation during polymerizationThe dU formed by thisroute is not intrinsically mutagenic and its abundance wouldbe determined by the discriminative accuracy of the DNApolymerase against dUTP [17] and the level of dUTP in thecellular pools To the authorrsquos knowledge neither dUTP levelsnor rates of dUTP incorporation have been measured inarchaeal cells

Removal of dU from archaeal DNA is expected to occurat a steady rate dictated largely by intracellular levels of UDGIn contrast dU incorporation occurs only at sites of DNAsynthesis dUwould thus bemost abundant immediately afterpassage of the replication fork and would decay steadily overtime thereafter The contrasting kinetics of incorporationversus removal could in principle make dU a ldquobiochemicalsignature of youthrdquo marking strands that remain uncorrectedby BER because they have been synthesized recently It istempting to question whether this or some other molecularldquotime stamprdquo could control important features of genomereplication in archaea E coli uses methylation of the tetranu-cleotide GATC in an analogous way to distinguish old andnew strands after the replication fork has passed and thissignal is necessary for the accuracy of postreplicative MMR[26ndash28] In HA dU could in principle play an analogousrole in a uniquely archaeal error-correction system or itcould serve to limit the extent or timing of DNA synthesis inrecently replicated regions Alternatively optimal replicationin HA may simply require periodic fork reversal in theabsence of replication-blocking lesions for which dU in thetemplate strand may provide a specific trigger

Whatever cellular processes dU abundance may con-trol the ldquobiochemical markerrdquo hypothesis predicts that aparticular balance of the relevant biochemical activities(dUTPase UDG and polymerase stalling) may be importantfor normal cellular function Thus artificial manipulation ofthese activities may perturb normal replication or growthwhich may explain the unexpected results of UDG genedeletion in S islandicus [10] It may be feasible to test thismore directly bymanipulating intracellular levels of dUTPase

and of individual UDGs through cloning in appropriateexpression vectors [29] It also may be possible to expressphysiological levels of the recently engineered dU-insensitiveDNA polymerases [18 19] in HA cells using similar vectors

6 The Enigma of Unemployed EukaryoticNER Proteins

Historically the molecular dissection of DNA repair path-ways including that of nucleotide excision repair (NER)depended heavily on the nonessential nature of these func-tions in microorganisms which allowed completely repair-defective mutants to be isolated and characterized NERis one of the most generalized systems of DNA repairits substrates represent a broad spectrum of lesions thatare bulky helix-distorting or both Despite having broadlysimilar strategies bacterial and eukaryotic NER aremediatedby proteins that share little homology Furthermore majorgroups of eukaryotes diverge with respect to the damagerecognition proteins which initiate NER yeast has one set ofdamage-recognition proteins plant cells have a different setandmammalian cells have several additional proteins [13 30]In addition to being diverse the damage-recognition proteinsare also the only NER-specific proteins of eukaryotes asall ldquodownstreamrdquo NER proteins play other roles in the cellThe overall pattern thus suggests an evolutionary scenarioin which NER arose rather recently and multiple times byrecruitment of suitable proteins to perform the damage-recognition step

Superficially genome annotations of mesophilic archaeasuggest a redundancy ofNER systems inwhich an incompleteset of eukaryotic NER genes commonly annotated as ldquoXPrdquo(xeroderma pigmentosum) homologs is superimposed on acomplete set of bacterial genes represented by homologsof E coli uvrABC None of the archaeal genomes encodesa clear homolog of an eukaryotic NER damage-recognitionprotein and the available experimental evidence suggests thatonly the Uvr proteins mediate NER in these archaea Forexample the damaged oligonucleotide excised by the moder-ately thermophilic Methanobacterium thermoautotrophicum(ie Methanothermobacter thermoautotrophicus) strain ΔH(which has a complete set of uvr genes) has the lengthcharacteristic of bacterialNER not eukaryoticNER [31] Sim-ilarly in the mesophilic halophile Halobacterium salinaruminactivation of any uvr homolog is sufficient to render thecells UV-sensitive [32] Thus the redundancy of NER inmesophilic archaea suggested by genome analysis is moreapparent than real as these archaea seem to rely functionallyon the simpler NER system they share with bacteria [33]

This conclusion raises questions for the HA which uni-formly lack the UvrABC homologs found in the mesophilicarchaea Three possibilities can be considered (i) in HAlack of the UvrABC system is compensated by recruitmentof unidentified HA-specific damage-recognition proteinswhich in combination with the archaeal XP homologsreconstruct an eukaryotic-style NER (ii) HA use fundamen-tally different ldquoalternative excisionrdquo pathways [1] to repairduplexDNAbetween rounds of replication or (iii)HA simply

6 Archaea

X

Lesion inleading-strand

template

Xand

Nicked or gapped chromosome plus ds end with lesion

Hef Xpf

(a)

X

lagging-strandtemplate

Lesion in

Fen1Xpg Xor


(b)

Figure 2 Fork breakage ldquocollapserdquo in response to replication-blocking lesions Adducts or other large helix-distorting lesions (symbolizedby X) block a polymerase molecule inducing a certain degree of uncoupling and ssDNA exposure at the base of the fork Cleavage by anappropriate single-strand-specific or structure-specific endonuclease can for each case generate dsDNA end that retains the lesion The topstrand in each DNA duplex is oriented with the 51015840 end on left free 31015840 ends are represented by arrowheads whereas broken lines indicate therest of the chromosome Carets (and) locate the single cut that would bemade by the indicated nuclease For the cleavage preferences of archaealHef Xpf and Fen1Xpg see [36 65 66]

forgo preemptive repair and deal with bulky helix-distortingDNA lesions after they stall the replication fork

The accumulating experimental evidence argues increas-ingly against (i) as deletions of most archaeal XP homologshave negligible effects on the survival of UV and otherrelevant DNA damage (Table 1) While it could be arguedthat the genes which showed no apparent role in repairduplicate the function of another gene such redundancyis not seen typically in bacteria and eukaryotes It shouldalso be noted that HA can remove such lesions and donot merely bypass them as UV photoproducts are steadilylost from the genomic DNA of UV-irradiated cultures uponresumption of incubation [34 35] In addition the removal ofUV photoproducts is not affected by the transcription of thedamaged DNA which argues further that HA do not employthe archaeal RNA polymerase extensively as a DNA damagedetector [34 35]

7 Do HA Use DNA ReplicationItself to Target Bulky Helix-DistortingLesions for Removal

Other experimental evidence can be interpreted as support-ing (iii) above Among various HA deletion mutants forexample one of the few that fits the phenotype expected ofa generalized excision repair deficiency is the Δhef mutant ofThermococcus kodakarensis [36] The encoded Hef protein isrelated to eukaryotic FANCM proteins and like them bindsto branchedDNAs thatmimic stalled replication forks whereit cleaves the duplex near the joint [36]This specificity seemstailored to the expected structures of archaeal replicationforks stalled by unrepaired bulky adducts particularly if thereplicative helicase archaeal Mcm mimics the eukaryoticMcm in its ability to progress past most DNA lesions leavingthem to stall the polymerase [37] If the leading-strandpolymerase encounters the lesion its blockage would createa partially single-stranded region near the junction of thefork (Figure 2) Cleavage at the base of the fork by Hef ora 31015840-flap endonuclease such as archaeal Xpf or cleavage ofthe large single-stranded region by ssDNA endonucleasewould break the fork in a way that leaves the lesion on

the resulting end Conversely a lesion encountered on thelagging-strand template would yield a similar result if theY-structure were cleaved by a 51015840-flap endonuclease suchas Fen1Xpg (Figure 2) The logic of the proposed strandspecificity is that in both cases it positions the lesion forremoval by ds end-processing which is central to replicationfork reassembly via HR (discussed in more detail below) Incontrast models of replication-fork ldquocollapserdquo and reassem-bly in bacteria have historically proposed steps that leavethe lesion in a continuous (albeit potentially gap-containing)DNA this feature may reflect the rationale that providingadditional chances for conventional NER represents the beststrategy for successful replication [36 38ndash40]

Although inducing fork disassembly and breakage inorder to expose and remove DNA lesions seems drastic(as argued above for dU) the lesions in question are bydefinition intrinsically incompatible with replicative poly-merases and (unlike dU)may have no alternative pathways ofremoval in HA A strategy like this may represent one of thesimplest and earliest responses to DNA damage and its basicfeasibility is demonstrated by the relatively robust growthof bacterial and eukaryotic NER mutants under laboratorycultivation In these cells the low level of bulky lesionsthat form endogenously (eg proteinDNA cross-links) isusually handled at the replication fork by HR-dependentprocesses such as fork reassembly initiated by fork breakage(discussed below) template-switching or lesion-skippingprocesses [41]

8 The Enigma of Accurate Replicationwithout MutSL Homologs

Researchers analyzing the first Pyrobaculum genomesequences noted that these HA genomes did not encode anyhomologs of the E coli MutS or MutL proteins [42] Thesetwo proteins cooperate to remove replication errors fromdsDNA soon after the bacterial replication fork has passed(sometimes termed the ldquospellcheckerrdquo function) they alsotrigger the abortion of HR between sequences that differat multiple sites (termed the ldquoantirecombinationrdquo function)[26 27] Among archaea the absence of MutSL homologs

Archaea 7

applies specifically to hyperthermophiles and thus parallelsthe absence of Uvr proteins Also similar to Uvr absenceof MutSL homologs in HA does not correlate with anyobvious defect in maintaining genome integrity althoughit must be noted that few HA have been analyzed in thisrespect Sulfolobus spp have been found to replicate theirchromosomes accurately although in some species this maybe masked by high levels of transposable element activity[43ndash45] The error rate per genome replication is lower in Sacidocaldarius than in nearly all other microbial genomessimilarly analyzed interestingly however single-nucleotideframeshifts in short mononucleotide runs are prominentin the spectrum which is a qualitative characteristic ofMMR-deficient cells [44]

These results demonstrate that Sulfolobus spp imitatethe results of conventional MMR quantitatively but perhapsnot qualitatively The absence of any MutSL pair argues thatSulfolobus spp and other HA either (i) lack any form ofpostreplicational mismatch repair or (ii) employ some alter-native to the conventional (ie MutSL-dependent) strategyIn order to explain the low error rate hypothesis (i) wouldrequire HA to have a much higher accuracy of polymer-ization in vivo than that exhibited by known replicativeDNA polymerases in vitro Similarly (ii) would require somesystem that recognizes mismatches after the duplex leaves thepolymerase and removes the newly synthesized strand butdoes not involve a MutS or MutL protein

Explaining the accurate chromosome replication in Sul-folobus spp must also account for their multiple translesionsynthesis (TLS) DNA polymerases which are nonproces-sive but error-prone The Sulfolobus Y-family polymerasesDpo4 (S solfataricus) and Dbh (S acidocaldarius) havebeen cocrystallized with many DNA substrates providingunprecedented resolution of biochemical and structural anal-ysis of TLS in vitro [46] Both polymerases are highly error-prone when replicating intact template in vitro with Dbhmaking 1-nt deletions at frequencies over 50 oppositecommon sequence motifs such as 51015840-GYY- [47] Howeverthe genetic consequences of inactivating Dbh indicate thatthis polymerase plays at least one antimutagenic role in vivonamely the suppression of spontaneous GC to TA transver-sions [11] The hypothesis that this suppression reflectsaccurate bypass of oxidized guanine (oxoG) is supported byanalysis of individual TLS events past this damaged base(Sakofsky amp Grogan unpublished) However the most com-mon spontaneousmutations observed in the S acidocaldariuschromosome are not affected by inactivating Dbh (Sakofskyamp Grogan unpublished) Given the extreme inaccuracy ofthis DNA polymerase in vitro these results suggest that Dbhis normally excluded from the Sulfolobus replication fork byan unidentified yet apparently effective mechanism

9 Mismatch Processing in Sulfolobus

Processing ofmismatchedDNAhas been detected geneticallyin S acidocaldarius but in the context of genetic recombi-nation not that of replication accuracy The evidence comesfrom transforming auxotrophic mutants with linear DNAsthat contain several closely spaced phenotypically silent

base pair substitutions (BPSs) as genetic markers [48] Thecomplex pattern of markers recovered after transformationwith ssDNA (Figure 3(a)) indicated that a short-patch modeof gene conversion (nonreciprocal HR) representing local-ized strand removal and resynthesis had created multiplediscontinuous tracts during recombination Even greatercomplexity was observed in cells transformed with dsDNAthat was itself a preformed heteroduplex in which bothstrandswere distinguished fromeach other aswell as from therecipient chromosome (Figure 3(b))The patterns from thesetriallelic crosses suggested that the preformed heteroduplexwas processed via similar localized strand removal and resyn-thesis before synapsis with the recipient chromosome [48]

10 Could Fork Regression Target ReplicationErrors for Removal

Transient reversal of the replication fork to form a four-armstructure commonly termed a ldquochicken footrdquo is thought to berelatively frequent inDNAreplication and is proposed to havefunctional roles such as providing error-free alternatives toTLS when the leading-strand template has an unrepairedlesion [49 50] One property of regressed forks that hasnot received much attention is the distinctive nature of thefourth (ie extruded) armwhich (i) represents the onlyDNAend of the structure (albeit not necessarily a blunt one) and(ii) includes only the newly synthesized daughter strands(Figure 4) Thus a ldquochicken footrdquo contains in the config-uration of its strands the information needed to find andremove a recent replication error (Figure 4) Although thishypothetical ldquospellcheckerrdquo would remove a correct strandalong with the replication error this price may be consideredsmall in light of the potential simplicity of the mechanism(which does not need to track strand discontinuities) and thebenefits of the resulting replication accuracy However thishypothetical mode of strand discrimination would imposesome unusual and perhaps unrealistic requirements Thereplication fork should presumably reverse only when areplication error has left either of the DNA polymerases forexample also destruction of the fourth armwould need to beprevented in forks reversed in the course of other processessuch as error-free lesion bypass [50]

11 The Enigma of Essential Recombination

Recombination proteins play diverse roles across biologyHistorically analysis of homologous recombination (HR)emphasized genetic diversification that is the reassortmentof alleles in eukaryotic meiosis and the acquisition of newtraits in bacteria following chromosomal DNA transferIncreasingly however HR is considered to support successfulgenome replication by providing repair of double-strandbreaks (DSBs) In particular the central process invasion ofa DNA duplex by a single-stranded 31015840 end with a matchingsequence provides a route to reassembling broken replicationforks [37 38]

HR in bacteria can be eliminated nearly quantitativelyby inactivating the homolog of the E coli recA gene or ineukaryotes by inactivating the Rad52 protein which loads

8 Archaea

ssDNA

Recipient chromosome

Transformant chromosome

Heteroduplexformation

Localizedstrand removal

Resynthesis

+

(a)



Preformedheteroduplex

+

+

Heteroduplex formation

Partial resynthesis

Partial resynthesis

Localized strand removal


(b)

Figure 3 Patchy recombination in Sulfolobus Cells transformed with linear DNAs marked at multiple sites (or mated with multiply markeddonor cells) produce recombinants that indicate erratic localized strand loss from heteroduplex intermediates [48 67] (a) Markers areacquired from ssDNA as multiple short tracts some of which consist of a single marker (b) A preformed heteroduplex containing twodistinct donor alleles at each marked position generates similar short patches representing all three possible alleles Semicircular symbolsdepict ldquosilentrdquo genetic markers (synonymous mutations) and the colors (black white and gray) depict different alleles Different colorsopposite to each other indicate a mismatch that is eventually resolved by strand removal and resynthesis

(forkregression)

(ds enddegraded)

Replicationerror

Figure 4 Repositioning of replication errors by fork reversal Ifa replication fork reverses shortly after making a replication errorthe resulting mismatch would be localized to the short (extruded)arm In principle this could provide a basis for removing replicationerrors without involving a conventional (MutSL-dependent) MMRsystem The white semicircle represents a polymerase error heavylines depict parental (template) strands

the eukaryotic equivalent of RecA (Rad51) onto ssDNAArchaea have their own recombinases designated ldquoRadArdquorelated to both the bacterial and eukaryotic proteins [51]All these proteins mediate a central step in HR in whichssDNA finds its complement in dsDNA and pairs with itIn addition to abolishing the processes of crossing-over andgene conversion between two homologous duplex DNAsrecA or rad52 mutations make cells dramatically sensitive toDNAdamage and this sensitivity is exacerbated ifNER is also

disabledThis supports the argument that the immediate andperhaps the original function of HR is to enable replicationforks blocked by unrepaired lesions to disassemble reformand resume DNA replication [39 52]

The remaining proteins required for HR prepare theDNA end(s) for strand invasion or resolve the structures thatresult In bacteria homologs of the E coli RecBCD proteinsprepare the ds end for invasion of the intact duplex TheRecBCD complex is a powerful helicaseexonuclease thatinitially unwinds and degrades both strands of dsDNA endbut eventually converts to a form that leaves 31015840 extensionswhich are needed for strand invasion [37 38] Anothercomplex in bacteria consisting of the SbcCD proteins canalso perform this function and eukaryotic cells have corre-sponding ldquoMRXrdquo complexes composed of related eukaryoticproteinsMre11 and Rad50 (the third protein X differs amongdifferent eukaryotes) Archaea have their own version ofMRX composed of archaeal Mre11 and Rad50 homologsand a uniquely archaeal complex of bipolar helicase andendoexonuclease HerA-NurA [53]

Among archaea genetic dissection of HR functions hasmade the most progress in the extreme halophiles The radArad50 andmre11 genes have been deleted individually and incombinations from Halobacterium salinarum and Haloferaxvolcanii for example [54ndash56] Most of the mutants showed

Archaea 9

XX

X

Fork breakage products

Resection of both strandsformation of a 3998400 extension

Gaps filled++

(a)

Strand invasion D-loop formation

D-loop extensionopposite-strand priming

Holliday-junctionresolution gap filling

(b)

Figure 5 Regeneration of broken replication forks by HR functions (a) The products of archaeal replication fork breakage proposed inFigure 2 are expected to retain various lesions gaps and overhangs possible structures are illustrated to the left of each bracketThese variousstructuresmust be converted to a ldquocleanrdquo 31015840 extension on the ds end and an intact continuous duplex with which it can recombine (b) Proteinsof double-strand-break repair (homologous recombination) promote subsequent reassembly of the replication fork The steps depicted hereare those commonly proposed for bacteria and eukaryotes [37 38 52] As in previous figures broken lines indicate the remainder of thechromosome and arrowheads mark 31015840 ends

a modest increase in radiation sensitivity but in HaloferaxradA could not be deleted from a mutant lacking all fourof the normal replication origins [57] In contrast to thehalophiles HA have not permitted deletion of the radA genedespite multiple attempts by multiple groups Furthermorefour other proteins of HA share this property of apparentessentiality the archaeal homologs of Mre11 and Rad50 aswell as the HerA and NurA proteins All of these proteins areimplicated in the central steps of HR that is the processingof dsDNA and for RadA loading and subsequent strandinvasion and to the authorrsquos knowledge they have not beenidentified with any other nonrecombination function Themre11 rad50 herA and nurA genes are typically clustered inHA genomes and the proteins have been observed to formfoci in 120574-irradiated cells [58] The apparent essentiality ofthese proteins therefore seems to coincide with the othergenetic features discussed above that distinguish HA from allother cellular organisms including mesophilic archaea

12 Do HA Need HR to Survive Fork-CenteredDNA Repair

In principle the essentiality of RadA Mre11 Rad50 HerAand NurA specifically in HA could reflect biochemicalfunctions unique to the HA versions of these proteinsAlthough this possibility remains intriguing it seems moreparsimonious to ask whether the broadly conserved functionof these proteins (specifically HR between a dsDNA end anda corresponding duplex) is not simply in greater demand orplays a more critical role in HA than in other organisms IfHA indeed lack a dedicated preventive-maintenance systemlike NER to repair bulky helix-distorting lesions beforereplication they should experience frequent lesion-inducedfork stalling Bacterial and eukaryotic cells experience sucharrest and HR proteins are required for many of theirresponses to this challenge [39 52] In HA therefore (i) theability to remove UV lesions from genomic DNA in theabsence of conventional NER [34] (ii) the importance ofHefXpf activity for full DNA repair capacity [36] and (iii)

the essentiality of basic HR functions for viability (Table 1)combine to suggest the possibility that breakage and HR-mediated reassembly of replication forks blocked at DNAlesions may serve as the major pathway for removing suchlesions

Most schemes of fork ldquocollapserdquo (breakage) and reassem-bly that have been proposed for bacteria or eukaryotes eitherrestore the context of the original lesion so that it can berepaired by conventional NER or leave it behind in anassociated daughter-strand gap [39 52] In the context of HAit would seem to benefit the cell more if the endonucleolyticcleavage that breaks the replication fork would leave thelesion near the ds end (Figure 4) The proposed advantageof this feature is that the unwinding and bidirectional nucle-olytic activities typical of DSB end-processing complexesseem well suited to remove a broad spectrum of DNAlesions including clustered and opposed ones with minimalrequirements for damage recognition (Figure 5) Similarly byanalogy to the RecBCD complex once a bidirectional strandremoval has discarded the lesion limiting or inhibiting the 31015840-specific activity would in principle be sufficient to generatethe 31015840 tail required to initiate fork reconstruction

Regardless of itsmolecular details the basic principle thatall replication-blocking DNA lesions may be funneled downa pathway culminating in fork breakage can be expected tomake successful replication of the chromosomedependent onHR to a degree not represented in any other major group oforganisms Although NER mutants of bacteria or eukaryotesmanage to reproduce under laboratory conditions on a basissimilar to that proposed here for wild-type HA the envisagedarchaeal strategy seems unlikely to succeed in nature withoutmolecular adaptations that make the requisite fork-breakageand reassembly processes unusually efficient and reliableGenes or biochemical functions found to be unexpectedlyimportant for normal chromosomal replication in HA mayinclude those that embody such adaptations

Finally whether or not any of the processes proposed hereultimately prove to operate in HA the prospect remains thatthese organisms may integrate repair recombination andreplication of DNA much more tightly than any other type

10 Archaea

of cell that has been examined previously Such integrationis expected to make functional studies challenging andanticipates a need for genetic techniques such as constructionof single-gene knockout libraries and conditional mutantsepistasis testing and analysis of mutants with biophysicalmethods and the full spectrum of molecular biology tech-niques Although activities and molecular interactions ofindividual enzymes will continue to be identified in HA thathave a plausible connection with DNA repair and replicationfidelity the most meaningful test of progress in this area willbe whether the proposed contribution to genome mainte-nance can be confirmed in living archaeal cells

Conflict of Interests

The author declares that no conflict of interests exists regard-ing the publication of this paper

Acknowledgment

The author thanks C Sakofsky for helpful discussion regard-ing the paper

References

[1] E C Friedberg G C Walker W Siede R D Wood R ASchultz and T EllenbergerDNA Repair andMutagenesis ASMPress (American Society for Microbiology) Washington DCUSA 2006

[2] C R Woese O Kandler and M L Wheelis ldquoTowards anatural system of organisms proposal for the domains ArchaeaBacteria and Eucaryardquo Proceedings of the National Academy ofSciences of the United States of America vol 87 no 12 pp 4576ndash4579 1990

[3] E R Barry and S D Bell ldquoDNA replication in the archaeardquoMicrobiology and Molecular Biology Reviews vol 70 no 4 pp876ndash887 2006

[4] S Ishino LM Kelman Z Kelman and Y Ishino ldquoThe archaealDNA replication machinery past present and futurerdquo Genesand Genetic Systems vol 88 no 6 pp 315ndash319 2014

[5] L M Kelman and Z Kelman ldquoArchaeal DNA replicationrdquoAnnual Review of Genetics vol 48 no 1 pp 71ndash97 2014

[6] M OrsquoDonnell L Langston and B Stillman ldquoPrinciples andconcepts of DNA replication in bacteria archaea and eukaryardquoCold Spring Harbor Perspectives in Biology vol 5 no 7 2013

[7] F Sarmiento F Long I Cann and W B Whitman ldquoDiversityof the DNA replication system in the archaea domainrdquoArchaeavol 2014 Article ID 675946 15 pages 2014

[8] H Atomi R Matsumi and T Imanaka ldquoReverse gyrase is not aprerequisite for hyperthermophilic liferdquo Journal of Bacteriologyvol 186 no 14 pp 4829ndash4833 2004

[9] P Forterre ldquoA hot story from comparative genomics reversegyrase is the only hyperthermophile-specific proteinrdquo Trends inGenetics vol 18 no 5 pp 236ndash237 2002

[10] C Zhang B Tian S Li et al ldquoGenetic manipulation in Sul-folobus islandicus and functional analysis of DNA repair genesrdquo Biochemical Society Transactions vol 41 no 1 pp 405ndash4102013

[11] C J Sakofsky P L Foster and D W Grogan ldquoRoles of theY-family DNA polymerase Dbh in accurate replication of the

Sulfolobus genome at high temperaturerdquoDNARepair vol 11 no4 pp 391ndash400 2012

[12] Q Huang Y Li C Zeng et al ldquoGenetic analysis of the Holl-iday junction resolvases Hje and Hjc in Sulfolobus islandicusrdquoExtremophiles vol 19 no 2 pp 505ndash514 2015

[13] O D Scharer ldquoNucleotide excision repair in eukaryotesrdquo ColdSpring Harbor Perspectives in Biology vol 5 no 10 Article IDa012609 2013

[14] R S Lasken D M Schuster and A Rashtchian ldquoArchaebacte-rial DNApolymerases tightly bind uracil-containingDNArdquoTheJournal of Biological Chemistry vol 271 no 30 pp 17692ndash176961996

[15] X-P Liu and J-H Liu ldquoCharacterization of family IV UDGfrom Aeropyrum pernix and its application in hot-start PCR byfamily B DNA polymeraserdquo PLoS ONE vol 6 no 11 Article IDe27248 2011

[16] H H Hogrefe C J Hansen B R Scott and K B NielsonldquoArchaeal dUTPase enhances PCR amplifications with archaealDNA polymerases by preventing dUTP incorporationrdquo Pro-ceedings of the National Academy of Sciences of the United Statesof America vol 99 no 2 pp 596ndash601 2002

[17] P Gruz M Shimizu F M Pisani M de Felice Y Kanke andT Nohmi ldquoProcessing of DNA lesions by archaeal DNA poly-merases from Sulfolobus solfataricusrdquo Nucleic Acids Researchvol 31 no 14 pp 4024ndash4030 2003

[18] E Gaidamaviciute D Tauraite J Gagilas and A LagunaviciusldquoSite-directed chemical modification of archaealThermococcuslitoralis Sh1BDNApolymerase acquired ability to read throughtemplate-strand uracilsrdquo Biochimica et Biophysica Acta vol1804 no 6 pp 1385ndash1393 2010

[19] S K Jozwiakowski B J Keith L Gilroy A J Doherty and B AConnolly ldquoAn archaeal family-B DNA polymerase variant ableto replicate past DNA damage occurrence of replicative andtranslesion synthesis polymerases within the B familyrdquo NucleicAcids Research vol 42 no 15 pp 9949ndash9963 2014

[20] J Wardle P M J Burgers I K O Cann et al ldquoUracilrecognition by replicative DNA polymerases is limited to thearchaea not occurringwith bacteria and eukaryardquoNucleic AcidsResearch vol 36 no 3 pp 705ndash711 2008

[21] T T Richardson L Gilroy Y Ishino B A Connolly andG Henneke ldquoNovel inhibition of archaeal family-D DNApolymerase by uracilrdquo Nucleic Acids Research vol 41 no 7 pp4207ndash4218 2013

[22] S Grasso and G Tell ldquoBase excision repair in Archaea back tothe future inDNA repairrdquoDNARepair vol 21 pp 148ndash157 2014

[23] M J Fogg L H Pearl and B A Connolly ldquoStructural basisfor uracil recognition by archaeal family B DNA polymerasesrdquoNature Structural Biology vol 9 no 12 pp 922ndash927 2002

[24] S K Jozwiakowski F Borazjani Gholami and A J DohertyldquoArchaeal replicative primases can perform translesion DNAsynthesisrdquo Proceedings of the National Academy of Sciences of theUnited States of America vol 112 no 7 pp E633ndashE638 2015

[25] T Sakai S-I Tokishita K Mochizuki A Motomiya HYamagata andTOhta ldquoMutagenesis of uracil-DNAglycosylasedeficient mutants of the extremely thermophilic eubacteriumThermus thermophilusrdquo DNA Repair vol 7 no 4 pp 663ndash6692008

[26] S Acharya P L Foster P Brooks and R Fishel ldquoThe coor-dinated functions of the E coli MutS and MutL proteins inmismatch repairrdquo Molecular Cell vol 12 no 1 pp 233ndash2462003

Archaea 11

[27] GM Li ldquoMechanisms and functions of DNAmismatch repairrdquoCell Research vol 18 no 1 pp 85ndash98 2008

[28] J A Surtees J L Argueso and E Alani ldquoMismatch repair pro-teins key regulators of genetic recombinationrdquo Cytogenetic andGenome Research vol 107 no 3-4 pp 146ndash159 2004

[29] S Berkner A Wlodkowski S-V Albers and G Lipps ldquoInduci-ble and constitutive promoters for genetic systems in Sulfolobusacidocaldariusrdquo Extremophiles vol 14 no 3 pp 249ndash259 2010

[30] V Ly A Hatherell E Kim A Chan M F Belmonte and D FSchroeder ldquoInteractions betweenArabidopsisDNArepair genesUVH6 DDB1A and DDB2 during abiotic stress tolerance andfloral developmentrdquo Plant Science vol 213 pp 88ndash97 2013

[31] M Ogrunc D F Becker S W Ragsdale and A Sancar ldquoNu-cleotide excision repair in the third kingdomrdquo Journal ofBacteriology vol 180 no 21 pp 5796ndash5798 1998

[32] D J Crowley I Boubriak B R Berquist et al ldquoThe uvrAuvrB and uvrC genes are required for repair of ultraviolet lightinduced DNA photoproducts in Halobacterium sp NRC-1rdquoSaline Systems vol 2 article 11 2006

[33] M F White ldquoDNA repairrdquo in Archaea Evolution Physiologyand Molecular Biology R A Garrett and H P Klenk Eds pp171ndash182 Blackwell Publishing Malden Mass USA 2007

[34] R Dorazi D Gotz S Munro R Bernander and M F WhiteldquoEqual rates of repair of DNAphotoproducts in transcribed andnon-transcribed strands in Sulfolobus solfataricusrdquo MolecularMicrobiology vol 63 no 2 pp 521ndash529 2007

[35] V Romano A Napoli V Salerno A Valenti M Rossi and MCiaramella ldquoLack of strand-specific repair of UV-inducedDNAlesions in three genes of the archaeon Sulfolobus solfataricusrdquoJournal of Molecular Biology vol 365 no 4 pp 921ndash929 2007

[36] R Fujikane S Ishino Y Ishino and P Forterre ldquoGenetic analy-sis of DNA repair in the hyperthermophilic archaeonThermo-coccus kodakaraensisrdquo Genes and Genetic Systems vol 85 no4 pp 243ndash257 2010

[37] S L Forsburg ldquoThe MCM helicase linking checkpoints to thereplication forkrdquo Biochemical Society Transactions vol 36 no1 pp 114ndash119 2008

[38] M M Cox ldquoRecombinational DNA repair of damaged repli-cation forks in Escherichia coli questionsrdquo Annual Review ofGenetics vol 35 pp 53ndash82 2001

[39] M M Cox M F Goodman K N Kreuzer D J Sherratt S JSandler and K J Marians ldquoThe importance of repairing stalledreplication forksrdquo Nature vol 404 no 6773 pp 37ndash41 2000

[40] M S Dillingham M Spies and S C KowalczykowskildquoRecBCD enzyme is a bipolar DNA helicaserdquo Nature vol 423no 6942 pp 893ndash897 2003

[41] J T P Yeeles J Poli K J Marians and P Pasero ldquoRescuingstalled or damaged replication forksrdquo Cold Spring HarborPerspectives in Biology vol 5 no 5 Article ID a012815 2013

[42] S T Fitz-Gibbon H Ladner U-J Kim K O Stetter M ISimon and J H Miller ldquoGenome sequence of the hyperther-mophilic crenarchaeon Pyrobaculum aerophilumrdquo Proceedingsof the National Academy of Sciences of the United States ofAmerica vol 99 no 2 pp 984ndash989 2002

[43] S Berkner and G Lipps ldquoMutation and reversion frequenciesof different Sulfolobus species and strainsrdquo Extremophiles vol12 no 2 pp 263ndash270 2008

[44] D W Grogan G T Carver and J W Drake ldquoGenetic fidelityunder harsh conditions analysis of spontaneous mutationin the thermoacidophilic archaeon Sulfolobus acidocaldariusrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 98 no 14 pp 7928ndash7933 2001

[45] EMartusewitsch CW Sensen and C Schleper ldquoHigh sponta-neous mutation rate in the hyperthermophilic archaeon Sulfol-obus solfataricus is mediated by transposable elementsrdquo Journalof Bacteriology vol 182 no 9 pp 2574ndash2581 2000

[46] J D Pata ldquoStructural diversity of the Y-family DNA poly-merasesrdquo Biochimica et Biophysica ActamdashProteins and Pro-teomics vol 1804 no 5 pp 1124ndash1135 2010

[47] O PotapovaND F Grindley andCM Joyce ldquoThemutationalspecificity of the Dbh lesion bypass polymerase and its implica-tionsrdquo The Journal of Biological Chemistry vol 277 no 31 pp28157ndash28166 2002

[48] D Mao and D W Grogan ldquoHeteroduplex formation mis-match resolution and genetic sectoring during homologousrecombination in the hyperthermophilic archaeon Sulfolobusacidocaldariusrdquo Frontiers in Microbiology vol 3 article 1922012

[49] J Atkinson and P McGlynn ldquoReplication fork reversal and themaintenance of genome stabilityrdquo Nucleic Acids Research vol37 no 11 pp 3475ndash3492 2009

[50] J E Sale ldquoCompetition collaboration and coordinationmdashdetermining how cells bypass DNA damagerdquo Journal of CellScience vol 125 no 7 pp 1633ndash1643 2012

[51] E M Seitz J P Brockman S J Sandler A J Clark and S CKowalczykowski ldquoRadA protein is an archaeal RecA proteinhomolog that catalyzes DNA strand exchangerdquo Genes andDevelopment vol 12 no 9 pp 1248ndash1253 1998

[52] A Kuzminov ldquoDNA replication meets genetic exchange chro-mosomal damage and its repair by homologous recombinationrdquoProceedings of the National Academy of Sciences of the UnitedStates of America vol 98 no 15 pp 8461ndash8468 2001

[53] N J Rzechorzek J K Blackwood S M Bray J D MamanL Pellegrini and N P Robinson ldquoStructure of the hexamericHerA ATPase reveals a mechanism of translocation-coupledDNA-end processing in archaeardquo Nature Communications vol5 article 5506 2014

[54] S Delmas I G Duggin and T Allers ldquoDNA damage inducesnucleoid compaction via the Mre11-Rad50 complex in thearchaeon Haloferax volcaniirdquo Molecular Microbiology vol 87no 1 pp 168ndash179 2013

[55] A Kish and J DiRuggiero ldquoRad50 is not essential for theMre11-dependent repair of DNA double-strand breaks in Halobac-terium sp strain NRC-1rdquo Journal of Bacteriology vol 190 no15 pp 5210ndash5216 2008

[56] W G Woods and M L Dyall-Smith ldquoConstruction and anal-ysis of a recombination-deficient (radA) mutant of Haloferaxvolcaniirdquo Molecular Microbiology vol 23 no 4 pp 791ndash7971997

[57] M Hawkins S Malla M J Blythe C A Nieduszynski and TAllers ldquoAccelerated growth in the absence of DNA replicationoriginsrdquo Nature vol 503 no 7477 pp 544ndash547 2013

[58] A Quaiser F Constantinesco M F White P Forterre andC Elie ldquoThe Mre11 protein interacts with both Rad50 andthe HerA bipolar helicase and is recruited to DNA followinggamma irradiation in the archaeon Sulfolobus acidocaldariusrdquoBMCMolecular Biology vol 9 article 25 2008

[59] C J Sakofsky L A Runck and D W Grogan ldquoSulfolobusmutants generated via PCR products which lack putativeenzymes of UV photoproduct repairrdquoArchaea vol 2011 ArticleID 864015 12 pages 2011

[60] P-J Liang W-Y Han Q-H Huang et al ldquoKnockouts of RecA-like proteins RadC1 and RadC2 have distinct responses to DNA

12 Archaea

damage agents in Sulfolobus islandicusrdquo Journal of Genetics andGenomics vol 40 no 10 pp 533ndash542 2013

[61] J H Y Wong J A Brown Z Suo P Blum T Nohmi andH Ling ldquoStructural insight into dynamic bypass of the majorcisplatin-DNA adduct by Y-family polymerase Dpo4rdquo TheEMBO Journal vol 29 no 12 pp 2059ndash2069 2010

[62] Q Huang L Liu J Liu J Ni Q She and Y Shen ldquoEfficient51015840-31015840 DNA end resection by HerA and NurA is essential forcell viability in the crenarchaeon Sulfolobus islandicusrdquo BMCMolecular Biology vol 16 article 2 2015

[63] T Zheng Q Huang C Zhang J Ni Q She and Y ShenldquoDevelopment of a simvastatin selection marker for a hyper-thermophilic acidophile Sulfolobus islandicusrdquo Applied andEnvironmental Microbiology vol 78 no 2 pp 568ndash574 2012

[64] Y Hong M Chu Y Li et al ldquoDissection of the functionaldomains of an archaealHolliday junctionhelicaserdquoDNARepairvol 11 no 2 pp 102ndash111 2012

[65] J A Roberts S D Bell and M F White ldquoAn archaeal XPFrepair endonuclease dependent on a heterotrimeric PCNArdquoMolecular Microbiology vol 48 no 2 pp 361ndash371 2003

[66] M W Kaiser N Lyamicheva W Ma et al ldquoA comparisonof eubacterial and archaeal structure-specific 51015840-exonucleasesrdquoThe Journal of Biological Chemistry vol 274 no 30 pp 21387ndash21394 1999

[67] J Rockwood D Mao and DW Grogan ldquoHomologous recom-bination in the archaeon Sulfolobus acidocaldarius effects ofDNA substrates and mechanistic implicationsrdquo Microbiologyvol 159 no 9 pp 1888ndash1899 2013

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology

2 Archaea

doctoral research advisor ldquoWhen you do biochemistryyou tell the organism what you think is important whenyou do genetics the organism tells you what it thinks isimportantrdquo

2 The Functional-Genetic Era forHyperthermophilic Archaea

For HA basic genetic analyses have become feasible only inthe past decade and one of the first of these studies targetedthe reverse DNA gyrase of Thermococcus kodakaraensis[8] Reverse gyrases (Rgy) are single-strand-nicking DNAtopoisomerases that introduce positive supercoils into DNAand appear only in extreme- or hyperthermophilic archaeaand bacteria [9] Although its natural distribution suggestedto many that reverse gyrase activity is essential for life at hightemperature Atomi et al [8] successfully isolated a deletionmutant demonstrating that Rgy is not essential for Thermo-coccus viability The Thermococcus mutant nevertheless grewat a reduced rate and with a decreased maximal temperatureThus whereas Rgy appears not to be absolutely requiredfor cellular function inThermococcus the mutant phenotypeargued that it must play a role of some importanceThe latterconclusion has been reinforced by failure of later efforts todelete the corresponding genes of other hyperthermophilicarchaea [10]

In the ensuing decade progress in the genetic analysisof DNA metabolic enzymes of HA has been substantial buthard-won As in the Rgy study the archaeal mutants aretypically constructed by replacing the target gene with aselected marker and therefore represent unambiguous lossof gene function However the resulting genotype oftendoes not match that predicted by the bacterial or eukaryoticcounterpart On one hand this reinforces the evidenceof deep divergence separating archaea from bacteria andeukaryotes and the incentive for their analysis but on theother hand it intensifies the need to formulate and test thenext generation of hypotheses This review and commen-tary selectively interprets functional genetic data relatingto genome stability in archaea with an emphasis on thehyperthermophiles It proposes that certain patterns of resultsdo not fit well with the schemes of genome-maintenancemechanisms identified in eukaryotic or bacterial modelsand suggests ways in which atypical or primordial damage-coping strategies could in principle support fundamen-tally different strategies of genome maintenance in theHA

3 Mutant Phenotypes

Table 1 lists functional studies of HA genes that have beenimplicated in DNA repair or related processes (usually bysequence similarity) grouped according to the phenotypicoutcome that was observed Mutants placed in Group Iseem to be relatively straightforward to interpret since theirphenotypes indicate a role of the deleted gene in DNArepair or mutagenesis roughly in line with that predictedfrom its sequence Examples include DNA photolyases

at least one TLS polymerase Holliday-junction resolvasesa 31015840-flap endonuclease a recombinase homolog and twoputative uracil DNA glycosylases (UDG) In most cases thedocumented phenotype is increased sensitivity to one ormore DNA-damaging agents although the UDG mutantsshowed impaired growth [10] and the Y-family polymerasemutant showed an altered spectrumof spontaneousmutation[11]

Mutants in Group II identify genes with minimal impactonDNA repair capacity as evidenced by little or no detectableincrease in sensitivity to radiation or DNA-damaging chem-icals In some cases notably HJ resolvases Hjc and Hjesingle mutants exhibited only marginal phenotypes but thedouble mutant was apparently nonviable [12] The lattercase thus suggests that some cellular function of these twoenzymes is important and is fulfilled by two at least partiallyredundant enzymes An important feature of the Group IIcases (Table 1) is that most of them encode homologs ofeukaryotic NER proteins which were expected to participatein generalized repair of bulky helix-distorting DNA lesionsAlthough redundancy of function could in principle accountfor this result the corresponding eukaryotic genes typicallydo not exhibit such redundancy with respect to DNA repair[13]

Group III of Table 1 represents genes which could notbe deleted and thus may encode an essential functionAlthough the genetic tools for HA remain limited insome cases lethality could be further supported by specificgenetic tests Some of the proteins are predicted to playimportant roles in basic replication and thus their appear-ance in this group is fully plausible examples include thestructure-specific endonuclease Fen1 which is implicatedin maturation of Okazaki fragments Most of the genes inthis category are associated specifically with homologousrecombination which is however not an essential process innormal cells of bacteria unicellular eukaryotes ormesophilicarchaea

As noted in several of these studies the number ofputative DNA-repair genes that fall into the latter two cate-gories (Groups II and III) adds to the accumulating evidencethat HA do not follow the bacterial or eukaryotic modelsof DNA repair in any strict sense The available evidence(discussed below) nevertheless argues that they do replicatetheir genomes accurately and cope effectively with DNAdamage This poses the interesting but challenging opportu-nity of identifying precisely how the genome-maintenancemechanisms of HA differ from those defined historically inbacteria and eukaryotic cells This review and commentaryhighlights four enigmata of genome-stability mechanismsin HA some of which have persisted for two decades (i)the functional role of archaeal DNA polymerase stalling atdU (ii) the apparent lack of repair-specific functions for theHA homologs of eukaryotic NER proteins (iii) the specificabsence of MutSL homologs in HA and (iv) the essentialityof HR proteins in HA Although the interpretations providedhere are broad and speculative attention to unresolvedquestions such as these represents a logical prerequisite touncovering genetic processes that distinguish archaea andespecially HA from other forms of life

Archaea 3












4 Archaea

(a)

UU

U


dU inleading-strand

template

(substrate for BER)

U

U


Fork regression



(b)

(c)








Archaea 5












6 Archaea

X


template

Xand


Hef Xpf

(a)

X


Lesion in

Fen1Xpg Xor


(b)










Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Archaea 3












4 Archaea

(a)

UU

U


dU inleading-strand

template

(substrate for BER)

U

U


Fork regression



(b)

(c)








Archaea 5












6 Archaea

X


template

Xand


Hef Xpf

(a)

X


Lesion in

Fen1Xpg Xor


(b)










Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

4 Archaea

(a)

UU

U


dU inleading-strand

template

(substrate for BER)

U

U


Fork regression



(b)

(c)








Archaea 5












6 Archaea

X


template

Xand


Hef Xpf

(a)

X


Lesion in

Fen1Xpg Xor


(b)










Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Archaea 5












6 Archaea

X


template

Xand


Hef Xpf

(a)

X


Lesion in

Fen1Xpg Xor


(b)










Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

6 Archaea

X


template

Xand


Hef Xpf

(a)

X


Lesion in

Fen1Xpg Xor


(b)










Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Archaea 7












8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

8 Archaea

ssDNA





Resynthesis

+

(a)




+

+


Partial resynthesis

Partial resynthesis



(b)


(forkregression)

(ds enddegraded)

Replicationerror






Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Archaea 9

XX

X



Gaps filled++

(a)




(b)









10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

10 Archaea




Acknowledgment


References




























Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Archaea 11



































12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

12 Archaea
















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology








Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Documents

Review Article Understanding DNA Repair in Hyperthermophilic Archaea: Persistent … · 2019. 7. 31. · DennisW.Grogan Department of Biological Sciences, University of Cincinnati,