Vol.175,No.2JOURNALOFBACTERIOLOGY,Jan.1993,p.428-4370021-9193/93/02042 - PDF document

1 Vol.175,No.2JOURNALOFBACTERIOLOGY,Jan.19
Vol.175,No.2JOURNALOFBACTERIOLOGY,Jan.1993,p.428-4370021-9193/93/020428-10$02.00/0CopyrightC)1993,AmericanSocietyforMicrobiologySuppressionofColElHigh-Copy-NumberMutantsbyMutationsinthepolAGeneofEscherichiacoliYUN-LIANGYANGADBARRYPOLISKY*PrograminMolecular,CellularandDevelopmentalBiology,DepartmentofBiology,IndianaUniversity,Bloomington,Indiana47405Received24July1992/Accepted2November1992WeisolatedthreeEscherichiacolisuppressorstrainsthatreducethecopynumberofamutantColElhigh-copy-numberplasmid.Thesemutationslowerthecopynumberofthemutantplasmidinvivoupto15-fold;thewild-typeplasmidcopynumberisreducedbytwo-tothreefold.Thesuppressorstrainsdonotaffectthecopynumbersofnon-ColEl-typeplasmidstested,suggestingthattheireffectsarespecificforColEl-typeplasmids.TwoofthesuppressorstrainsshowColElallele-specificsuppression;i.e.,certainplasmidcopynumbermutationsaresuppressedmoreefficientlythanothers,suggestingspecificityintheinteractionbetweenthesuppressorgeneproductandplasmidreplicationcomponent(s).AllofthemutationsweregeneticallymappedtothechromosomalpoL4gene,whichencodesDNApolymeraseI.ThesuppressormutationalchangeswereidentifiedbyDNAsequencingandfoundtoaltersinglenucleotidesintheregionencodingtheKlenowfragmentofDNApolymeraseI.TwomutationsmapintheDNA-bindingcleftofthepolymeraseregionandaresuggestedtoaffectspecificinteractionsoftheenzymewiththereplicationprimerRNAencodedbytheplasmid.Thethirdsuppressoraltersaresidueinthe3'-5'exonucleasedomainoftheenzyme.ImplicationsfortheinteractionofDNApolymeraseIwiththeColElprimerRNAarediscussed.TheinitiationofreplicationofColEliscontrolledmainlybyinteractionsbetweentwoplasmid-encodedRNAs:oneisasmallRNAcalledRNAI;anotheristheprimerRNA(RNAII)forleading-strandDNAsynthesis.TheprimerRNAhasseveralunusualstructuralfeatureswhichendowitwithunusualfunctionalproperties(7,44).MoststrikingisitsabilitytoformapersistentRNA-DNAhybridwiththetemplateDNAstrandinthevicinityofthereplicationorigin(12).TheRNA-DNAhybridstructureisasubstrateforthehostenzymeRNaseH,whichcleavestheprimerRNAtogenerateaproperlypositioned3'terminus.ThisterminusissubsequentlyrecognizedbyDNApolymeraseI,whichcat-alyzesformationoftheleadingstrandDNA(11,39).TheabilityoftheprimerRNAtoformapersistenthybridisaconsequenceoftheformationofparticularsecondaryandtertiarystructuraldomainsintheRNA,sincemutationsalteringthesedomainscanreduceorincreasetheefficiencywithwhichhybridformationoccurs(26,27,42).However,thespecificfeaturesofthesedomainswhicharecriticalandhowtheyfunctiontoformthepersistenthybridareun-known.TheabilityoftheprimerRNAtoformapersistenthybridwiththetemplatestrandisthekeystepthatisregulatedinplasmidcopynumbercontrol(28).Thisregulationismedi-atedbythecountertranscriptRNAI,whichiscomplemen-tarytothe5'-terminalregionoftheprimerRNA(20,43).TheinteractionbetweentheseRNAsaltersthefoldingpathwayoftheprimerRNAandpreventstheformationofparticulardomainsrequiredforpersistentRNA-DNAhybridformation(41).AsaconsequenceofinteractionwithRNAI,primerRNAelongationcontinuesbeyondtheoriginregionwithouthybridizing,andtheprimertranscriptisreleasedfromthetemplatewithoutinitiatingleading-strandDNAsynthesis(27,35).RNAIisthereforeatrans-actingnegativeregulatorofreplicationinitiation.Singlebasechangesthat*Correspondingauthor.simultaneouslyalterbothRNAIandprimerRNA,thuspreservingcomplementaritybetweenthem,canneverthelessaffectplasmidcopynumberandplasmidincompatibilityproperties(7,20,45).Theseresultsindicatethattheconfor-mationsofRNAIandprimerRNAareimportanttotheirfunctions(7,41,44,46).WehavepreviouslydescribedagroupofColElhigh-copy-numbermutationsinprimerRNAthatmapoutsideitsregionofcomplementaritywithRNAI(9,36).Thesemuta-tionsincreasetheefficiencywithwhichprimerformsthepersistentRNA-DNAhybridinvitroandincreaseplasmidcopynumberinvivo(36).Eachofthemutationsresponsibleforthephenotypeisasinglebasechangeintheprimer.AtypicalmutationinthisclassiscarriedbyaplasmidcalledpTF128G,whichcontainsaU-to-Gchangeatprimerposi-tion128(9).Wehavealsoshownthatanequilibriumbetweendifferentprimerconformersisalteredbymutationsintheprimer,therebyaffectingtheplasmidcopynumber(8).Wealsoisolatedseveralintragenicsecond-sitesuppressorsofthesecopynumbermutationswhichalterasecondbaseintheprimer(8).Inworkreportedhere,wehaveinvestigatedinteractionsbetweenplasmidandhostreplicationcompo-nentsbysearchingforhostmutationsthatsuppressthehigh-copy-numberphenotypeofthepTF128Gplasmid.WereportonthreesuppressorsthatalterthegeneencodingDNApolymeraseI,ahostfunctionknowntoplayakeyroleinhostDNAreplicationandrepair(19),andinColElreplication(11,12,18)

2 .WeshowthatthesechangesinDNApolymeraseIs
.WeshowthatthesechangesinDNApolymeraseIsuppressdifferentColElmutationsinanallele-specificmanner,suggestingthattheseprimerRNAsinteractdifferentlywithspecificregionsofDNApolymeraseI.MATERIALSANDMETHODSBacterialstrains,growthconditions,andplasmids.AllbacterialstrainsusedarederivativesofEscherichiacoliK-12.DH5[F-endl4hsdRJ7(rK-MK)supE44X-thi-J428 SUPPRESSIONOFColElHIGH-COPY-NUMBERMUTANTS429gyrA96relAl]wasusedastheplasmidhostintheseexper-iments(8).StrainsusedforHfrmappinganddonorstrainsforPlvirtransductionweregiftsfromC.A.Gross,Univer-sityofWisconsin(40).StrainsusedforHfrmappingwereCAG12206,CAG12204,CAG12203,CAG12202,CAG12200,CAG12205,andCAG12201(40).StrainsusedforPlvirtransductionwereCAG18572,CAG18558,CAG18592,CAG18599,CAG18557,CAG18601,CAG18636,CAG18560,CAG18618,CAG18615,CAG18609,CAG18630,CAG18571,CAG18555,andCAG12114(40).ThesestrainsweregrownineitherLuriabroth(LB)(30)or2xYT(38)andsupplementedwiththeappropriateantibioticswiththeindicatedconcen-trationsasneeded:ampicillin,100pg/ml;kanamycin,25pLg/ml;streptomycin,50pg/ml;andchloramphenicol,20pg/ml.DH5wasconvertedtoRecA+byintroductionoftheplasmidpLY10,apCL1921derivative(21)containingawild-typeE.colirecA'gene(10)clonedfrompMZE1,apUC19derivativecarryingrecA+(agiftfromM.Zolan,thisdepartment).PlasmidpLY10,whichconfersstreptomycinresistance,wastransformedintosuppressorstrains.Trans-formantsweregrownovernightwithselection,andtheovernightculturewasstreakedontoLBplatesandtestedforUVlightresistance.PlasmidsusedtodetermineoriginspecificityofthesuppressorstrainswerepTF487(anR6Korigin)(9),pCL1921(apSC101origin)(21),andpJRD215(anRSF1010origin)(4).pCJ105,anF-derivedplasmidconfer-ringchloramphenicolresistance,anditsderivatives(14)weregiftsfromC.M.JoyceatYaleUniversity.Theparentalwild-typeColElplasmidpTF128UandthemutantpTF128Ghavebeendescribedpreviously(8).Hfrmapping.HfrmappingofthesuppressormutationswascarriedoutasdescribedbyMiller(30).EachHfrusedcontainedakanamycinresistancegenemarker20mindown-streamofthetransferorigin.Suppressoranddonorstrainsweregrownovernightat30°CinLBwithappropriateanti-biotics.Dilutionsofconjugatingcellswereincubatedat37°Cfor1htoallowtheexpressionofthekanamycinresistancegeneandthenplatedontoLBplatescontaining100,ugofampicillin,50,ugofstreptomycin,and25,gofkanamycinpermlandincubatedat30°Covernight.Colonieswerereplicaplatedontoplatescontaining100pgofampicillin,3mgofmethicillin,50,ugofstreptomycin,and25,ugofkanamycinperml.Cellscontaininghigh-copy-numberplas-midsformedcoloniesonplatescontaining3mgofmethicillinperml;cellscontainingwild-type-copy-numberplasmidsdidnotformcoloniesontheseplates.PreparationofPlvirtransducingphageandtransduction.PreparationofPlvirtransducingphageandtransductionwereperformedasdescribedbyMiller(30).Donorandrecipientsuppressorstrainsweregrownat30°Cwithappro-priateantibiotics.Alldonorstrainscarryakanamycinresis-tancemarkeratknownchromosomallocations(40).Infecteddonorcellcultureswereincubatedat37°Cuntilcelllysisoccurred,andthetransducingphageswereisolatedbycentrifugation.Infectedrecipientcultureswereincubatedat37°Cfor20minandplatedontoLBplatescontainingappropriateantibiotics.Plateswereincubatedat30°Cover-night.ColoniesresistanttobothkanamycinandampicillinweretestedforColElplasmidcopynumberphenotype.F'mating.F'matingwascarriedoutasdescribedbyMiller(30).DonorandrecipientcellsweregrowninLBat30°Covernightwithappropriateantibiotics.Conjugatingcellswereincubatedat37°Cfor1h.DilutionswereplatedontoLBplateswith100pgofampicillinand20,ugofchloramphenicolperml.Theplateswereincubatedat30°CovernightandreplicaplatedontoLBplateswith100,gofampicillinand20pugofchloramphenicolpermlandontoplateswith400Rgofampicillin,3mgofmethicillin,and20pugofchloramphenicolpermltodetermineantibioticresis-tancephenotypeasdescribedabove.Determinationofmutationposition.ThelocationofthemutationinthepolAgenewasdeterminedafterpolymerasechainreaction(PCR)amplificationsofthechromosomalpolAgeneinto1.7-kband1.3-kbfragments,whichincludedtheentirestructuralgeneandknownpromotersequencesofthepolAgene(15),followedbyDNAsequenceanalysisoftheamplifiedfragments.TheprimersusedforPCRweresynthesizedonanAppliedBiosystemssynthesizerandhadthefollowingdesignationsandsequences:PolAl,CCACATGACGATTI'GCGAGCG;PoLA2,GCACCGGCACCAGCGG;PolA3,GGGCCGTTGAACGTCTJTCGAG;andPolA4,CTAGTGGCCATCACAAAACGTGACAGC.PCRswerecarriedoutwithTaqpolymerase(agiftfromN.R.Pace,thisdepartment)in100RIcontaining50mMKCl,10mMTris-HCl(pH8.3),1.5mMMgCl2,0.01%gelatin,1puMeachprimer,2mMeachdeoxynucleosidetriphosphate(dNTP),and100ngofchromosomalDNA.PCRwascarriedoutfor30cycles

3 ,andreactionproductswerepurifiedbyagaros
,andreactionproductswerepurifiedbyagarosegelelectrophoresisfollowedbyGeneclean(Bio101)treatmenttoremovegenomicDNA,residualoligonucleotides,andadventitiousfragments.DNAsequencingreactionswerecarriedoutwiththeSequenase2.0kit(UnitedStatesBio-chemicalCorporation[USB])accordingtothemanufactur-er'sinstructionswiththefollowingmodifications:14-,ureactionmixturescontaining500to1,000ngofPCRDNAfragments,0.5pmolofoligonucleotideprimer,1RIuof100mMdithiothreitol,2pAof1/10-diluteddNTPmixture(USB),and0.5pAof[a-32P]dCTP(Amersham)weremixedandboiledfor5minandthenfrozenindryice.Sequenase(USB)wasaddedtothereactionmixturesafterthawingatroomtemperature.Reactionswereterminatedimmediatelywithdideoxynucleosidetriphosphates(ddNTPs)(USB).Plasmidcopynumberassay.Bacterialculturescarryingplasmidsweregrownin2xYTmedia.Clearedlysateswerepreparedbyalkalinelysis(9),andplasmidDNAinthelysateswascleavedwithappropriaterestrictionenzymestoproducelinearDNA.DNAswereelectrophoresedin0.8%agarosegelsinTAEbuffer(40mMTris-HCl,18mMEDTA,20mMsodiumacetate[pH7.9]),stainedwithethidiumbromide,destained,andphotographed.DNAamountswerequantitatedbydensitometricanalysis.Isolationofspontaneousbacterialsuppressorstrains.Toselectspontaneoussuppressorstrains,DH5cellscarryingpTF128Gwereinoculatedinto3mlof2xYTwith100,ugofampicillinpermlandincubatedat42°Covernight.ThistemperatureisnonpermissiveforgrowthofDH5carryingpTF128G.Portionsofthisculturewereplatedonto2xYTplatescontaining100,ugofampicillinperml(low-drugplates)andincubatedat30°Covernight.Colonieswerereplicaplatedonto2xYTplatescontaining400,gofampi-cillinand3mgofmethicillinperml(high-drugplates)andincubatedat30°Covernight.Coloniesthatgrewonlow-drugplatesbutnotonhigh-drugplateswereconsideredsuppres-sorcandidatesandsubjectedtoplasmidcopynumberanal-ysis.Nitrosoguanidinemutagenesis.DH5cellscarryingpTF128Gweretreatedwithnitrosoguanidineasdescribedprevi-ously(30).Aftermutagenesis,cellsweregrownat30°Cin2xYTwith100,ugofampicillinpermlovernight.Cellswerespundown,resuspendedinfresh2xYTwith100,ugofampicillinperml,andgrownat42°Cfor8to16h.Suppres-sorcandidateswereisolatedasdescribedabove.VOL.175,1993 430YANGANDPOLISKYPlasmidstabilitydetermination.StrainscarryingColElplasmidsweregrownat30'Cin2xYTmediumintheabsenceofampicillinselection.Cellsweremaintainedinexponentialgrowthphasebyperiodicdilution.Atintervals,sampleswereplatedtodeterminethepercentageofampicil-lin-resistantcellsinthepopulation.Westernblot(immunoblot)analysis.Proteinsampleswereseparatedbyelectrophoresisona6%polyacrylamidegelcontainingsodiumdodecylsulfate(SDS-PAGE),andWest-ernblotting(38)wascarriedoutwithpolyclonalrabbitantibodies(agiftfromC.M.Joyce,YaleUniversity)di-rectedagainstthepurifiedKlenowfragmentofDNApoly-meraseI.Donkeyanti-rabbitimmunoglobulinGconjugatedwithhorseradishperoxidase(AmershamECLkit)togetherwiththehorseradishperoxidasecolordevelopmentreagentwasusedfordetectionofcross-reactingbandsaccordingtotheinstructionsprovidedbythemanufacturer.RESULTSIsolationofextragenicplasmidcopynumbersuppressors.Thedesignofascreenforhostsuppressormutationsofthehigh-copy-numberplasmidpTF128Gwasbasedontheob-servationthatDH5cellscarryingpTF128Ggrownormallyat30'Cin2xYTmediuminthepresenceof100Rgofampicillinpermlbutstopgrowingshortlyafterashiftto420C(Fig.1AandE).Thiseffectoncellgrowthisassociatedwithexten-siveplasmidDNAreplication(datanotshown).Incontrast,DH5cellscarryingwild-typeColElplasmidsgrowexten-sivelyafterashiftto420C(Fig.1AandE);extensivereplicationofplasmidDNAaftertheshiftisnotobserved(datanotshown).Thus,DH5derivativescontainingpTF128Gthatcouldgrowextensivelyat42°Cwereconsid-eredpotentialsuppressorcandidates(Fig.1BandF).Tofurthercharacterizepotentialsuppressormutations,wetookadvantageofasecondconsequenceofthehigh-copy-numberphenotype,viz.,high-levelexpressionoftheplasmid-borneP-lactamaseactivityconferscolonyformingabilityoncellscarryingthemutantplasmid.Wemonitoredtheabilityofcandidateclonestoformcoloniesonplatescontainingbothampicillinandmethicillinasaroughmeasureofintracellularplasmidcopynumber.Thepresenceofmethicillinpotenti-atestheeffectoftheampicillin;higherlevelsofP-lactamasearerequiredtoallowgrowthonplatescontainingbothantibiotics(46).Cellscarryingwild-typeplasmidscannotformcoloniesonplatescontainingbothampicillin(400pug/ml)andmethicillin(3mg/ml)(high-drugplates),whilecellscarryinghigh-copy-numbermutantsareabletoformcolonieswithhighefficiencyonsuchplates(46).SuppressorcandidatescarryingpTF128Gthatwereunabletoformcoloniesonhigh-drugplatesbutcouldformcoloniesonplatescontaining100,ugofampicillinpermlal

4 onewereisolatedforfurtheranalysis.Threec
onewereisolatedforfurtheranalysis.ThreecandidatederivativesofstrainDH5,designatedYY23L,YY112,andYY114,werestudiedfurther.YY23LandYY112werespontaneousmu-tations;YY114wasnitrosoguanidineinduced.ThecopynumberbehaviorofpTF128Gat30°Cinthesethreecandi-datesuppressorstrainsandintheparentalDH5strainisshowninFig.2.ThecopynumberofpTF128Gwasreducedbyafactorof9to15inthesuppressorcandidatesrelativetoitscopynumberinDH5.Allcopynumberdeterminationsweremadebynormalizationtothenon-ColElco-residentplasmidpTF487(8).TheapparentsuppressionofpTF128Gcopynumbercouldresultfromeitheraplasmidorahostmutation.Ifsuppres-sionresultedfromamutationinabacterialgene,thenpTF128Gisolatedfromthecandidatesuppressorcellswouldbeexpectedtodisplayahigh-copy-numberphenotypewhensubsequentlytransformedintofreshDH5cells.Addition-ally,whenpTF128GDNAisolatedfromDH5wastrans-formedintoplasmid-freesuppressorcandidatecells,thecopynumberofpTF128Gisexpectedtobereduced.Usingthesecriteria,weshowedthatYY23L,YY112,andYY114containhostmutations.Effectsofsuppressorstrainsonnon-ColElplasmidorigins.ToassesswhetherthesesuppressorstrainsaffectedcopynumberbehaviorofplasmidsunrelatedtoColEW,plasmidspJRD215(anRSF1010derivative)(4),pTF487(anR6Kderivative)(9),andpCL1920(apSC101derivative)(21)wereseparatelytransformedintoYY23L,YY112,YY114,andDH5.WeobservednodifferenceinplasmidcopynumbersorstabilityoftheseplasmidsinYY23L,YY112,YY114,andDH5(datanotshown).WeconcludethatthesesuppressorsarespecifictoplasmidscontainingtheColElreplicationorigin.EffectofsuppressorstrainsonvariousColE1copynumbermutations.Todeterminewhetherthecopynumbersuppres-sionofpTF128Gwasallelespecific,ColEl-basedplasmidswithdifferenthigh-copy-numbermutationsweretrans-formedintowild-typeandsuppressorstrainstoexaminetheeffectsofsuppressormutations.TwoclassesofcopynumbermutationsinprimerRNAwereexamined.ClassImutationsarelocatedintheprimerregionthatoverlapswithRNAIatprimerpositions19A,53U,and59G(themutantnucleotideisdenotedasdescribedinreference46).ThesechangespotentiallyalterbothprimerandRNAIconformationsalthoughsequencecomplementarityismaintained.Thehigh-copy-numberphenotypecausedbythesemutationscanbeduetodefectiveinhibitionofprimerformationbyRNAIatakineticlevelorduetoalterationsinprimerconformation,eachofwhichreducestheinhibitoryactivityofRNAIonplasmidreplication.ClassIImutationsarelocatedinaparticularprimerdomaincalledstem-loopIV,whichin-cludesprimerpositions128G,124A,131C,132A,and134Aandthe124A/134Adoublemutation(9,36).TheseallelesmapoutsidethecomplementarysequencebetweenRNAIandprimerandreplicatetohighcopynumberdespiteencodingwild-typeRNAI(9,36).PrimerscarryingthesemutationshavealteredconformationscomparedwiththewildtypeandhavebeenshowntoberesistanttoRNAIinhibitioninvitro(9,36).Inaddition,thesemutantprimersformRNA-DNAhybridsatthereplicationoriginmoreefficientlythanwild-typeprimerinvitro(8).CellscarryingvariousColElplasmidswereharvestedinlatelogphase(3x109to4x10cellsperml),andplasmidcopynumbersweredetermined(Fig.3;Table1).SincethesuppressorstrainshadnoeffectonR6K-typeplasmidcopynumber,anR6K-basedplasmid,pTF487,wastransformedintocellstoserveasaninternalcontrol.Toevaluatetheextentofsuppressionat37°C,theamountofwild-typeormutantplasmidDNAwasnormalizedtotheamountofinternalcontrolplasmidpTF4B7,andthesenormalizedval-uesfromwild-typeandsuppressorcellswerecompared.AsshowninTable1,inDH5,variousColElmutantplasmidshad3-to10-fold-increasedcopynumbersrelativetothewild-typeplasmid,dependingontheparticularallele.Ingeneral,classIImutationshadhighercopynumbersthanclassImutationsinDH5.Itisnoteworthythattherelativecopynumberofthewild-typeplasmidwasnotdramaticallylowerinthethreesuppressorstrains;inbothYY23LandYY112,wild-typeplasmidcopynumberswerelessthantwofoldlowerthaninDH5(Table1).Incontrast,suppres-J.BACTERIOL. SUPPRESSIONOFColElHIGH-COPY-NUMBERMUTANTS4314cByso40302010Time(hrs)0ByB300C*YY23U128GMo-YY112/128GxYV114/128G0o.0o0o0jI024681012Time(hrs)D37DC50040030020010000_AIIII024681012Time(hrs)E4200qcBy0246Time(hrs)F420C81012024681012024681012Time(hrs)Time(hrs)FIG.1.Growthbehaviorofwild-typeandsuppressorcellsatdifferenttemperatures.(A,C,andE)Wild-typecellswithpTF128GorpTF128U(wildtype);(B,D,andF)threedifferentsuppressorstrainswithpTF128G.Overnightcellsculturedat30'Cinthepresenceof100pgofampicillinand25,ugofkanamycinpermlwereinoculatedintofresh2xYTmediumcontaining100pgofampicillinand25figofkanamycinandgrownatdifferenttemperatures.ThecelldensitiesoftheseculturesatdifferenttimesafterinoculationweremonitoredwithaKlett-Summersonphotoelectriccolorimeter.sionofdifferentplasmidm

5 utantsinthethreesuppressorstrainsdisplay
utantsinthethreesuppressorstrainsdisplayedallelespecificity,withregardtoboththeplasmidalleleandthesuppressorallele.ScanningverticallyinTable1,itisapparentthatdifferentplasmidallelesweresuppressedtodifferentextentsinaparticularsuppressorstrain.Forexample,pTF128Gwassuppressedabout6-foldinYY23L,whileamutantcarryingthe124Amutationwassuppressedabout14-fold.TherangeofsuppressionamongdifferentclassIImutationswasaboutsevenfoldinYY23LandfourfoldinYY112.ScanninghorizontallyinTable1,itcanbeseenthattheextentofsuppressionofthesameplasmidallelewasdifferentinthevarioussuppressorstrains.Again,pTF124AwasefficientlysuppressedinYY23LbutwassuppressedonlyaboutfourfoldinYY112.Anothermutantplasmid,pTF131C,hadtheoppositebehavior,beingmoreefficientlysuppressedinYY112.ClassImutations(19A,53U,and59G)weresuppressedonlymoderatelybyYY23LandYY112andshowedlessthanatwofoldrangeinsuppressionamongthethreeplasmidalleles(Table1).ThesuppressionpatternexertedbyYY114wasdifferentfromthoseoftheothertwosuppressors.AllclassIImuta-tionsweresuppressedatleast10-foldbyYY114(Table1).Thisextensivesuppressiongreatlynarrowedthecopynum-berrangeexhibitedbyplasmidmutantsinYY114,leadingtotheconclusionthatthissuppressorlackedtheallelicdiscrim-inationshownbytheothertwostrains.YY114alsosup-A=1C37%400-300200-100-csy*AY23/128G_YV112/128GxVY114/128G.DH5/128G*DH5/128U/~~~~~~~~~~~~~~~~~~~~~~i-TSoo400,30022001000VOL.175,19930DH5/128G0*DH5/128U 432YANGANDPOLISKY765TF487ColEl3-FIG.2.Plasmidcopynumberphenotypesinwild-typeandsup-pressorcellsat30'C.Cellswereculturedin2xYTmediumcontain-ing100[tgofampicillinand25ptgofkanamycinperml.Clearedlysateswerepreparedfromlate-log-phaseculturesofeitherwild-typecellscarryingpTF128GorpTF128Uorsuppressorcellscarry-ingpTF128G.EachcellstrainalsocarriedpTF487,anR6Kderiv-ative,asaninternalcontrol.LysateswerelinearizedbyPstIdigestionandelectrophoresedona0.8%agarosegel.Afterbeingstainedwithethidiumbromide,thegelwasphotographed.ThelowerbandineachlaneisthelinearizedColElDNA;theupperbandispTF487DNA.M,1-kb-laddermarkerDNA;lane1,DH5cellswithpTF128Uwild-typeplasmid;lane2,DH5cellswithpTF128Gmutantplasmid;lane3,YY23LwithpTF128G;lane4,YY112withpTF128G;lane5,YY114withpTF128G.pressedthesingleclassImutationtested(53U)ninefold(Table1)-again,aneffectsubstantiallygreaterthanthatobservedwiththeothersuppressors.Geneticmappingofthesuppressormutations.SincethesuppressorstrainswereinitiallyisolatedinarecAmutantbackground,geneticmappingcommencedafterintroductionofpLY10,apSC101-basedplasmidcarryingthewild-typerecA+gene(10).ThelocationofthesuppressormutationinstrainYY23LwasmappedbyusingsevenHfrstrainswithdifferenttransferorigins(40).EachHfrstraincontainedakanamycinresistancemarker20mindownstreamofthetransferorigin(40).WescoredforlinkageofthekanamycinresistancemarkeroftheHfrdonorandincreasedampicillinresistancelevelsduetorestorationofthehighcopynumberofpTF128G,whichwasalsopresentinthesecells.MatingswithHfrstrainsCAG12201andCAG12204(40)wereob-DH5=)0|NIl#l.)CfYY112servedtorestorehigh-levelampicillinresistance.Theseresultspermittedassignmentofthemutationtoapositionbetween82and96minonthechromosome(datanotshown).Tofurtherlocalizethemutation,asetofPlvirlysatespreparedfromstrainswithdefinedtransposoninsertionsinthisregionwereused(40).CellscarryingpTF128GandpLY10weretransducedwiththesephagesandscoredforlinkageofkanamycinresistanceconferredbythetransposonandincreasedampicillinresistance.Thisanalysispermittedassignmentofthesuppressormutationtotheregionbetween87and87.5min(datanotshown).ComplementationofthesuppressorphenotypebyapolAplasmidanddeterminationofmutationposition.SincethepolAgeneencodingDNApolymeraseIislocatedinthe87-to87.5-minvicinity(1)andisknowntobeinvolvedinColE1replicationbothinvivoandinvitro(11,12,18),wedirectlytestedwhetherthesuppressormutationwaslocatedinpolA.AsetofF-derivedplasmidscontainingeitherthewild-typepolAgeneoraportionofthepoLAgene(14)weretrans-formedintosuppressorcells,andthecopynumberofpTF128Gwasdetermined.AsshowninFig.4,amini-Fplasmid,pCJ105,containingthewild-typepoLAgenere-storedthehigh-copy-numberphenotypeofpTF128G.pCJ105,themini-Fplasmidvectorlackinganinsert,hadnoeffectontheplasmidcopynumber.pCJ105derivativescontainingthesequenceofeitheroneofthetwofunctionalmoietiesofDNApolymeraseI,the5'-3'exonucleaseortheKlenowfragment,werealsotransformedintosuppressorstrainsandtestedfortheirabilitytorestorethepTF128Ghighcopynumber.AplasmidcontainingthesequenceencodingtheKlenowfragmentwasabletopartiallyrestorethehigh-copy-numberphenotypeinYY23L,YY112,andYY114(Fig.4).Aplasmidcontainingthesequenceencodi

6 ngthe5'-3'exonucleasewasunabletorestoret
ngthe5'-3'exonucleasewasunabletorestorethisbehavior(Fig.4).Fromtheseresults,wetentativelyconcludedthatthesuppressormutationspresentinYY23L,YY112,andYY114werelocatedintheKlenowfragmentofDNApoly-meraseI.YY23L0TF487ColEIYY114TF487FIG.3.PlasmidcopynumberphenotypesofdifferentColElmutantallelesinwild-typeorsuppressorcellsat37°C.SampleswerepreparedasdescribedforFig.2exceptthatcellswereculturedat370C.ThelocationandnatureofthemutationscarriedbytheColE1derivativesareshownaboveeachlane.ThelowerbandineachlaneisthelinearizedColElDNA;theupperbandispTF487DNA.J.BACTERIOL. SUPPRESSIONOFColElHIGH-COPY-NUMBERMUTANTS433TABLE1.QuantitationoftheextentofsuppressionofdifferentColElcopynumbermutantallelesbythehostsuppressorstrainsaResult"forstrain:PlasmidalleleDH5(ColEl!YY23LYY112YY114TF487)ColE1/TF487DH5ratioColE1/TF487DH5ratioColE1/TF487DH5ratioClassIIWildtype3.0+0.81.6+0.21.91.7±0.21.80.9±0.13.3128G30.7+0.14.9±0.76.34.8±1.26.42.0±0.215.4124A16.6±1.51.2±0.313.84.0±1.64.20.5±0.133.2131C18.0±2.93.8±0.64.72.0±0.39.01.7±0.110.6132A31.4±3.64.2+0.57.51.8±0.117.41.3±0.124.2134A10.8±1.45.3±0.32.01.5±0.17.2ND124A/134A23.0±0.16.3±0.53.73.9±0.15.9NDClassI19A15.2±1.66.3±0.42.44.5±0.53.4ND53U10.8±1.26.5+1.51.73.9±0.12.81.2±0.19.059G9.9±0.77.9±0.21.34.4±0.72.3NDaSampleswerepreparedasforFig.3.ThelinearizedplasmidDNAbandswerequantitatedbydensitometricanalysisafterelectrophoresisinagarosegels."ColE1/TF487,copynumberratioofplasmidColEltopTF487(internalplasmidstandard);DH5ratio,ratiooftheDH5ColE1/TF487valuetothevaluefortheindicatedstrain.Resultsareaveragesfromtwoindependentexperiments,+standarddeviations.ND,notdetermined.YY23LI128GDH5I128UDH5/128G+°++°76543YY1121128GToidentifythemutation(s)presentinthesuppressorstrains,wedirectlysequencedtheentirepolAgenefromthesuppressorsandthewild-typestrainafteramplificationbyPCR.ChromosomalDNAfromtwoseparatecoloniesforeachstrainwasusedforPCRamplification.Theresultsofthesequenceanalysis(Fig.5)showedthatthepolAgeneinYY23LcontainsaC-to-Tsubstitutionatnucleotideposition2540(15)resultinginaPro-to-Leuchangeatposition847ofLY1OthePolIpolypeptide.TheYY112allelecontainsaT-to-Asubstitutionatnucleotideposition2651(15)resultinginaLeu-to-Glnchangeatposition884.TheYY114alleleisaC-to-Tchangeatnucleotideposition1337(15)thatcausesaCO~lESer-to-Phechangeatposition446.ThemutationsinYY23LandYY112maptothepolymerasedomainofthePolIpolypeptide.TheYY114changeislocatedinthe3'-5'exonucleasedomain.Identicalsinglebasesubstitutionsinthepol4geneweredetectedfromthetwoseparatePCRproductsineachcase.YY114/128G00:2T13211GATACCATGCTGGAGTCCTACATTCTCAATDTMLESYLN441F(YY114)T25241GCAGCCATTAACGCGCCAATGCAGGGAACCAANAPMOGT842L(YY23L)FIG.4.ComplementationofsuppressorphenotypewithF-de-rivedplasmids.Wild-typeorsuppressorcellscarryingpTF128GweretransformedwithpCJ105orpCJ105derivativescarryingeitheraportionofortheentirepolAgene.SampleswerepreparedasdescribedforFig.2.pLY10,apSC101derivative,wasusedasaninternalcontrol.Cellswereculturedin2xYTmediumwith100pLgofampicillin,50pLgofstreptomycin,and20figofchloramphenicolperml.ThelowerbandineachlaneisthelinearizedColElDNA;theupperbandispLY10DNA.PolA',pCJ105carryingthewild-typepoU4gene;5'-3'exo,pCJ105carryingthesegmentofthepolAgeneencodingthe5'-3'exonucleasedomainofDNApolymeraseI;Klenow,pCJ105carryingthesegmentofthepolAgeneencodingtheKlenowfragmentundercontrolofthelacUV5promoter.A26411CACGATGAACTGGTATTTGAAGTTCATAAAHDEVFEVHK881Q(YY112)FIG.5.SequencealterationsinthepolAgenesofthesuppressorstrains.ThreeregionsofDNAsequenceofthepoLAgeneareshown,eachshowingthelocationofthesinglebasechangepresentinaparticularsuppressorstrain.Thenumberattheupperleftreferstothenucleotideposition(15).Thebasechangeisshownabovetheline;theaminoacidchangeisindicatedbythearrowhead.ThelowernumberistheaminoacidpositioninDNApolymeraseI.LYb0DolElVOL.175,1993 434YANGANDPOLISKYy97FIG.6.WesternblotanalysisofDNApolymeraseIfromcellsinearlylogphase.Onehundredmicrogramsoftotalcellularproteinwasloadedineachlane.TheantibodywasgeneratedagainstpurifiedKlenowfragment.FourhundrednanogramsofpurifiedKlenowfragment(BoehringerMannheimBiochemicals)wasloadedasapositivecontrolinthelanelabeledKlenow.Arrowhead,positionofDNApolymeraseI.Thenumbersattherightindicatethepositionsofmolecularweightmarkers(inthousands).MMSsensitivity.AcommonphysiologicpropertyofpoIAmutantcellsistheirsensitivitytolowconcentrationsoftheradiomimeticmutagenmethylmethanesulfonate(MMS)(3).poAmutantssuchaspol4laresensitivetoMMSinplateassaysatlevelsaslowas0.04%(6,17).Todeterminewhetherthepol-ColElsuppressorstrainsbehavedsimi-larlytootherpoLAmutan

7 tstrains,wild-typeandsuppressorcellswere
tstrains,wild-typeandsuppressorcellswereplatedontoLBplatescontainingdifferentcon-centrationsofMMSandincubatedatdifferenttemperatures.Weobservednodetectabledifferencebetweenthewild-typeandthethreesuppressorstrainsinsensitivitytoMMSatconcentrationslowerthan0.1%(datanotshown).Thus,thesuppressorstrainshavenodetectableeffectongrowthbehaviororMMSsensitivity,bothofwhicharecommonpropertiesforpoLAmutants(13,17).WeconcludethatthealterationsinDNApolymeraseIcausedbythesemutationsaresubtleandmaybeconfinedtotheireffectsonColElreplication.StabilityofplasmidsandPolIproteininsuppressorstrains.Toexaminethepossibilitythatplasmidinstabilityratherthanplasmidreplicationpersewasresponsibleforloweredplasmidcopynumberinthesuppressorstrains,wemeasuredplasmidstabilityinsuppressorstrainsgrownwithoutselec-tion.WeobservedthatpTF128GandpTF128UwereaboutasstableinthesuppressorstrainsasinDH5(datanotshown).Weconcludethereforethatsuppressionisnotaconsequenceofalteredplasmidstability.AsecondpossibleconsequenceofthepolAmutationsonColElplasmidreplicationcouldbetodecreasethestabilityofPolIprotein,therebyloweringtheamountoffunctionalenzymeinsuppressorcells.AlthoughlargeeffectsonDNApolymeraseIstabilityseemedunlikelygiventhemodesteffectofthesuppressorsonwild-typeplasmidcopynumberbehaviorandMMSsensitivity,itwasimportanttoinvesti-gatethispossibility.Totestthisidea,totalcellularproteinsfromwild-typeandsuppressorcellswereisolatedandelec-trophoresedonSDS-PAGEandthentransferredtoanitro-cellulosemembraneforWesternblotanalysis(38).Apoly-clonalantibodyagainsttheKlenowfragmentofDNApolymeraseIwasusedforthisassay.TheresultsofsuchananalysisshowednodifferenceintheamountofPolIproteininwild-typeorsuppressorcells,fromeitherlogphase(Fig.6)orstationaryphase(datanotshown).Theseresultsareconsistentwiththeideathatsuppressionisnotaconse-quenceofreducedDNApolymeraseIstability.DISCUSSIONWehaveisolatedDNApolymeraseImutantsthatsup-presstheColElhigh-copy-numbermutantpTF128GandotherColEl-typemutants.PreviousselectionsforpolAmutantsutilizedsensitivitytoMMSasascreen(6,13).These"classical"polAmutantsseverelyaffectDNApoly-meraseIactivity,asdeterminedonthebasisoftheireffectsongrowthrate,sensitivitytoMMS,andinabilitytostablymaintainwild-typeColElplasmids(6,13,18).ThisisincontrasttothepolAmutantsdescribedinthisarticle,whichshowallelespecificsuppressionofColElhigh-copy-numbermutantsandhaveminoreffectsonthecopynumberandstabilityofwild-typeColEL.ThesepolAstrainsrepresentanewclassofmutantswhichshouldbeusefulinfuturestructure-functionanalysisofDNApolymeraseI.DespiteextensiveknowledgeaboutthetertiarystructureofDNApolymeraseI,makingspecificassignmentsfortheeffectsofthemutationsdescribedhereonenzymefunctionisdifficult.AsdescribedindetailbyPoleskyetal.(34),theDNApolymeraseIreactionpathwayisquitecomplex.AfterbindingtoDNA,theappropriatedNTPisboundtotheenzyme.Theternarycomplexthenundergoesaconforma-tionalchangepriortophosphodiesterbondformation.Fol-lowingdNTPaddition,theenzymeundergoesasecondconformationalchangepriortoPP1release.Themutationsmayhaveeffectsonanyoneofthesestepsoraltertherate-limitingstepintheoverallreaction.Nonetheless,site-directedmutationalanalysishasbeenvaluableindefiningcertainresiduesandregionsoftheenzymeinvolvedindNTPbindingandsubsequentsteps(33,34).MostofthesemutationsarelocatedinresiduesformingthesurfaceoftheproposedDNA-bindingcleft(Fig.7)(13,16,32-34)andareconsistentwiththeideathatmultipleinteractionsbetweentheenzymeandprimersubstrateoc-cur.Thiscleftcomprisesasix-strandedantiparallel1-sheetformingtheflooranda-helicesmakingupthesides.Thedimensionsofthecleftandmodelbuildingareconsistentwiththeideathattheprimerisnormallyboundinthecleft(16,33,34).Site-directedmutationsinresiduesformingthesurfaceoftheproposedDNA-bindingcleftaffecttheaffinityoftheproteinforDNA,suggestingmultipleclosecontactsbetweenproteinandDNAinthisregion(33,34).ItisreasonabletoassumethatinColElreplicationinitiation,DNApolymeraseIalsohasmultiplecontactswiththeRNaseH-cleavedprimerRNA-DNAhybridinordertoelongatefromthe3'-OHterminusoftheRNAprimer.OfparticularrelevancetothemutationinstrainYY23Ldescribedhere,site-directedmutationsconstructedatresi-dues841and845havebeenshowntocauseanincreaseintheKmfordNTP,suggestingthattheseresiduesmayinter-actwiththeincomingdNTP(34).ThemutationinYY23LisaPro-to-Leusubstitutionatresidue847.Alloftheseresi-duesarelocatedinana-helicaldomaincalledhelixQ,whichispartoftheDNA-bindingcleft(Fig.7)(32).Thisrelativelysimpleassignmentiscomplicatedbytheobservationthatamutationinanothernearbyresidue,849,hasbeenshowntocauselargedecrease

8 sinkcat,indicatingthatthisresiduepartici
sinkcat,indicatingthatthisresidueparticipatesincatalysisofthepolymerasereaction(34).Thus,theregioninwhichresidue847islocatedparticipatesinbothdNTPbindingandcatalyticfunction,anditisnotobvioushowthePro-to-Leuchangeaffectstheseprocesses.However,threecommentsconcerningthischangeandloca-tioncanbemade.First,prolineatposition847istheeighthresidueintheQhelixandisexpectedtodisruptthehelicalconformationbycreatingakink(47).ThiskinkwouldnotbeJ.BACTERIOL. SUPPRESSIONOFColElHIGH-COPY-NUMBERMUTANTSFIG.7.LocationofsuppressoraminoacidsubstitutionsintheKlenowfragment.AmodelofKlenowfragmentderivedfromX-raycrystallography(32)isshownwithcylindersrepresentinga-helicesandarrowsrepresentingP-sheets.Thecirclednumbersindicatethelocationsofthesuppressormutations.expectedintheYY23LQhelix.Despitetheeliminationofthekinkcausedbyproline,themutantenzymedoesnothavedetectablephysiologicaleffects,suggestingthatthekinkdoesnotplayacriticalroleinpolymerasefunction.Indeed,thesecondnoteworthyaspectofthischangeisthatbacteriophageT7DNApolymerase,whichishomologousinpartwithDNApolymeraseIofE.coli,actuallyhasaleucineresidueatthehomologouslocation(5).Thus,leucineatthislocationinE.coliDNApolymeraseIisunlikelytobeincompatiblewithbasicfunctionsoftheenzyme.Thethirdaspectofnoteisthatresidue847isprobablylocatedontheoppositesideofhelixQfromresidues841and845.Thus,itssidechainisunlikelytoextendintotheDNA-bindingcleftanddirectlyparticipateinformationofthecleftsurface.WespeculatethatthischangeinhelixQslightlyaltersthepositionofnearbycriticalresiduessothatcontactswiththeColElprimerRNAareslightlyaltered.Thus,theeffectsofthemutationmaynotbedirecteliminationofacontactbutratheranindirectstereochemicaleffect.Whethertheim-pairedstepisthebindingoftheRNA,bindingofthedNTPs,orachemicalstepisuncleargiventhecloseassociationofthesefunctionsintheactivesite.Themutationinthesecondsuppressorstrain,YY112,changesresidue884fromLeutoGln.ThisresidueislocatedinP-sheet13oftheKlenowfragment(Fig.7)(32).Again,thisdomainispartoftheDNA-bindingcleft,anditcouldpotentiallycontactDNAdirectly(33,34).Site-directedmutationalchangesinthisregionhavebeenmadeatresidues881,882,and883(33,34).Thesidechainofresidue883isthoughttoextendintotheDNA-bindingcleft;amutationatthispositionaffectsDNAbinding(33,34).Residue882hasbeensuggestedtoplayanimportantroleincatalysis(33,34).TheLeu-to-Glnchangeatresidue884altersboththesizeandpolarityofthesidechain.SincethispositionisadjacenttotheresiduewhosesidechainextendsintotheDNA-bindingcleft,aswithresidue847,theeffectsofthechangemaybeindirectandsubtle,affectingpositioningofotherresiduesandreducingtheefficiencyofastepinprimerrecognitionorelongation.Themutationinthethirdsuppressorstrain,YY114,altersresidue446fromSertoGln.ThisresidueislocatedintheDhelix(Fig.7)(32)ofthe3'-5'exonucleasedomainofDNApolymeraseI.Sincetheeditingfunctionoftheexonucleaseactivity(19)isexpectedtoactaftereventsregulatinginitia-tionfrequency,itisdifficulttounderstandhowchangesinthisregioncouldshowplasmidallelespecificity.Infact,weobservedthatthissuppressordidnotshowappreciablediscriminationamongtheColElmutantalleles.WhatisthebasisfortheobtainedColElallelespecificsuppressionbythepolAmutations?IthasbeenshowninvitrothattheDNApolymeraseactivityofthepolA5mutantvarieswidelydependingontheDNAsubstrateusedinthereaction.WhencalfthymusDNAwasused,theactivityofpolA5DNApolymerasewas60-foldgreaterthanwithpoly(dT)1jOpoly(dA)800asthetemplate,comparedwitha2.5-folddifferencewithwild-typeenzyme(29).Recentbio-chemicalanalysisofDNApolymeraseIhasalsoshownthatdifferentsite-directedmutationscanaffecttherateofphos-phodiesterbondformationinvitroandthatthesemutationshavedifferenteffectsondifferentDNAsubstrates(33,34).ConsideringtheconformationaldifferencesintheseDNAsubstrates(2,37),itissuggestedthatthemutationscanaffectthepositioningofthe3'terminusintheDNA-bindingcleft(34).ApplyingtheseideastotheColElwild-typeandmutantprimerRNAsisdifficultintheabsenceofdetailedinformationaboutprimer-DNAconformation.However,mutationsintheColElprimerthatcausehighcopynumberhavebeenshowntoalterprimerconformation,leadingtoenhancedabilitytoformRNA-DNAhybridsinvitro(8).Primerconformationaldifferencesbetweenwild-typeandmutantplasmidsexistpriortoRNaseHcleavage.Theresultsdescribedheresuggestthatthesechangesmaypersist435VOL.175,1993 436YANGANDPOLISKYafterRNaseHcleavageandmaybedifferentiallyrecognizedbyparticularDNApolymeraseImutants.Inthisregard,itisnoteworthythatNaitoetal.(31)isolatedmutationsintheE.colinmhgene,encodingRNaseH,thatsuppressedcertainColElreplicationdefectiveprimermutants(calledcermu

9 -tants)inanallele-specificmanner.Thesewo
-tants)inanallele-specificmanner.TheseworkerssuggestedthatconformationalalterationsinthecerprimerRNAre-sponsibleforthereplicationdefectwerecompensatedbychangesintheconformationofRNaseHthatpermittedittorecognizeand/orprocessthemutantprimer.Theirresultsareconsistentwiththeideathatconformationaldifferencesbetweenwild-typeandmutantprimerRNAscanbediffer-entiallyrecognizedbytheenzymaticcomponentsinteractingwiththem.HowdomutationsinpolAreducethecopynumberofpTF128G?ConsideringthefactthatRNA-DNAhybridefficiencyisincreasedinmutantplasmidsinvitro(8),itmightbethecasethatRNA-DNAhybridformationisratelimitinginwild-typecells.However,mutationsindown-streamenzymaticcomponentssuchasRNaseHandDNApolymeraseImightmakesubsequentstepsratelimiting.ThemutationsinDNApolymeraseIdescribedheremayslightlyalterthebindingand/orpositioningofthemutantRNAprimersothattheefficiencyofrecognitionoftheRNaseH-cleaved3'-OHprimerterminusbyDNApolymeraseIisreduced.PurificationandcharacterizationofthemutantenzymeinDNAreplicationsysteminvitrowithColElDNAastemplatewillhelpresolvetheseissues.AsidefromRNaseHandDNApolymeraseI,whatotherhostproteinsmayplayaroleinColElreplication?AgenecalledpcnBhasbeenshowntoplayaroleinthemaintenanceofplasmidcopynumberofColEl-derivatives(22-24).Mu-tationsinpcnBgenereduceplasmidcopynumber(23).TheproteinpotentiallyencodedbythepcnBgeneshowssimilar-itytotheE.coliccageneproduct,atRNAnucleotidyltransferase(25).InteractionofthepcnBgeneproductwithColElRNAIand/orprimerRNAremainstobeestablished.ACKNOWLEDGMENTSWegratefullyacknowledgeCatherineM.Joyceforhelpfuldis-cussionsandforgiftsofantiserumagainstKlenowfragmentandplasmidpCJ105anditsderivatives.WethankCarolA.GrossforbacterialstrainsandPlvirphage,MiriamE.ZolanforpMZE1,andNormanR.PaceforTaqDNApolymerase.WealsothankJohnP.RichardsonandCarlE.Bauerforhelpfulcommentsonthemanu-script.ThisworkwassupportedbyNIHgrantGM24212.REFERENCES1.Bachmann,B.J.1990.LinkagemapofEscherichiacoliK-12,edition8.Microbiol.Rev.54:130-197.2.Coll,M.,C.A.Frederick,A.H.-J.Wang,andA.Rich.1987.Abifurcatedhydrogen-bondedconformationinthed(A.T)basepairsoftheDNAdodecamerd(CGCAAAl-l-GCG)anditscomplexwithdistamycin.Proc.Natl.Acad.Sci.USA84:8385-8389.3.Cooper,P.K.,andP.C.Hanawalt.1972.RoleofDNApolymeraseIandtherecsysteminexcision-repairinEsch-erichiacoli.Proc.Natl.Acad.Sci.USA69:1156-1160.4.Davison,J.,M.Heusterspreute,N.Chevalier,V.Ha-Thi,andF.Brunel.1987.Vectorswithrestrictionsitebanks.V.pJRD215,awide-host-rangecosmidvectorwithmultiplecloningsites.Gene51:275-280.5.Delarue,M.,0.Poch,N.Tordo,D.Moras,andP.Argos.1990.Anattempttounifythestructureofpolymerase.ProteinEng.3:461-467.6.deLucia,P.,andJ.Cairns.1969.IsolationofanE.colistrainwithamutationaffectingDNApolymerase.Nature(London)224:1164-1166.7.Dooley,T.P.,J.Tamm,andB.Polisky.1985.IsolationandcharacterizationofmutantsaffectingfunctionaldomainsofColE1RNAI.J.Mol.Biol.186:87-96.8.Fitzwater,T.,Y.-L.Yang,X.-Y.Zhang,andB.Polisky.1992.MutationsaffectingRNA-DNAhybridformationoftheColElreplicationprimerRNA;restorationofRNAIsensitivitytoacopy-numbermutantbysecond-sitemutations.J.Mol.Biol.226:997-1008.9.Fitzwater,T.,X.-Y.Zhang,R.Elble,andB.Polisky.1988.ConditionalhighcopynumberColElmutants:resistancetoRNA1inhibitioninvivoandinvitro.EMBOJ.7:3289-3297.10.Horii,T.,T.Ogawa,andH.Ogawa.1980.OrganizationoftherecAgeneofEscherichiacoli.Proc.Natl.Acad.Sci.USA77:313-317.11.Itoh,T.,andJ.Tomizawa.1978.InitiationofreplicationofplasmidColElDNAbyRNApolymerase,ribonucleaseH,andDNApolymeraseI.ColdSpringHarborSymp.Quant.Biol.43:409-418.12.Itoh,T.,andJ.Tomizawa.1980.FormationofanRNAprimerforinitiationofreplicationofColElDNAbyribonucleaseH.Proc.Natl.Acad.Sci.USA77:2450-2454.13.Joyce,C.M.,D.M.Fujii,H.S.Laks,C.M.Hughes,andN.D.F.Grindley.1985.GeneticmappingandDNAsequenceanalysisofmutationsinthepoLAgeneofEscherichiacoli.J.Mol.Biol.186:283-293.14.Joyce,C.M.,andN.D.F.Grindley.1984.Methodfordeter-miningwhetherageneofEscherichiacoliisessential:applica-tiontothepolAgene.J.Bacteriol.158:636-643.15.Joyce,C.M.,W.S.Kelley,andN.D.F.Grindley.1982.NucleotidesequenceoftheEscherichiacolipolAgeneandprimarystructureofDNApolymeraseI.J.Biol.Chem.257:1958-1964.16.Joyce,C.M.,andT.A.Steitz.1987.DNApolymeraseI:fromcrystalstructuretofunctionviagenetics.TrendsBiochem.Sci.12:288-292.17.Kelley,W.S.,andC.M.Joyce.1983.GeneticcharacterizationofearlyambermutationsintheEscherichiacolipoL4geneandpurificationoftheamberpeptides.J.Mol.Biol.164:529-560.18.Kingsbury,D.T.,andD.R.Helinski.1973.Temperature-sensitivemutantsforthereplicationofplasmidsinEscherichiacoli:requirementfordeoxyribon

10 ucleicacidpolymeraseIinthereplicationoft
ucleicacidpolymeraseIinthereplicationoftheplasmidColEl.J.Bacteriol.114:1116-1124.19.Kornberg,A.1980.DNAreplication.W.H.FreemanandCo.,SanFrancisco.20.Lacatena,R.M.,andG.Cesareni.1981.BasepairingofRNAIwithitscomplementarysequenceintheprimerprecursorinhib-itsColElreplication.Nature(London)294:623-626.21.Lerner,C.G.,andM.Inouye.1990.Lowcopynumberplasmidsforregulatedlow-levelexpressionofclonedgenesinEs-cherichiacoliwithblue/whiteinsertscreeningcapability.Nu-cleicAcidsRes.18:4631.22.Liu,J.,andJ.S.Parkinson.1989.GeneticsandsequenceanalysisofthepcnBlocus,anEscherichiacoligeneinvolvedinplasmidcopynumbercontrol.J.Bacteriol.171:1254-1261.23.Lopilato,J.,S.Bortner,andJ.Beckwith.1986.MutationsinanewchromosomalgeneofEscherichiacoliK-12,pcnB,reduceplasmidcopynumberofpBR322anditsderivatives.Mol.Gen.Genet.205:285-290.24.March,J.B.,M.D.Colloms,D.Hart-Davis,I.R.Oliver,andM.Masters.1989.CloningandcharacterizationofanEscherichiacoligene,pcnB,affectingplasmidcopynumber.Mol.Micro-biol.3:903-910.25.Masters,M.,J.B.March,I.R.Oliver,andJ.F.Collins.1990.ApossibleroleforthepcnBgeneproductofEscherichiacoliinmodulatingRNA:RNAinteractions.Mol.Gen.Genet.220:341-344.26.Masukata,H.,andJ.Tomizawa.1984.Effectsofpointmuta-tionsonformationandstructureoftheRNAprimerforColElDNAreplication.Cell36:513-522.27.Masukata,H.,andJ.Tomizawa.1986.Controlofprimerforma-tionforColElplasmidreplication:conformationalchangeoftheprimertranscript.Cell44:125-136.J.BACTERIOL. SUPPRESSIONOFColElHIGH-COPY-NUMBERMUTANTS43728.Masukata,H.,andJ.Tomizawa.1990.Amechanismofforma-tionofapersistenthybridbetweenelongatingRNAandtem-plateDNA.Cell62:331-338.29.Matson,S.W.,F.N.Capaldo-Kimball,andR.A.Bambara.1978.OntheprocessivemechanismofEscherichiacoliDNApolymeraseI,thepolASmutation.J.Biol.Chem.253:7851-7856.30.Miller,J.H.1972.Experimentsinmoleculargenetics.ColdSpringHarborLaboratory,ColdSpringHarbor,N.Y.31.Naito,S.,T.Kitani,T.Ogawa,T.Okazaki,andH.Uchida.1984.Escherichiacolimutantssuppressingreplication-defectivemu-tationsoftheColElplasmid.Proc.Natl.Acad.Sci.USA81:550-554.32.Ollis,D.L.,P.Brick,R.Hamlin,N.G.Xuong,andT.A.Steitz.1985.StructureoflargefragmentofEscherichiacoliDNApolymerasecomplexedwithdTMP.Nature(London)313:762-766.33.Polesky,A.H.,M.E.Dahlberg,S.J.Benkovic,N.D.F.Grindley,andC.M.Joyce.1992.SidechainsinvolvedincatalysisofthepolymerasereactionofDNApolymeraseIfromEscherichiacoli.J.Biol.Chem.267:8417-8428.34.Polesky,A.H.,T.A.Steitz,N.D.F.Grindley,andC.M.Joyce.1990.IdentificationofresiduescriticalforthepolymeraseactivityoftheKlenowfragmentofDNApolymeraseIfromEscherichiacoli.J.Biol.Chem.265:14579-14591.35.Polisky,B.1986.ReplicationcontroloftheColEl-typeplas-mids,p.143-170.InW.ReznikoffandL.Gold(ed.),Maximiz-inggeneexpression.Butterworths,Boston.36.Polisky,B.,X.-Y.Zhang,andT.Fitzwater.1990.MutationsaffectingprimerRNAinteractionwiththereplicationrepressorRNAIinplasmidColEl:potentialRNAfoldingpathwaymutants.EMBOJ.9:295-304.37.Rhodes,D.,andA.Klug.1981.Sequence-dependenthelicalperiodicityofDNA.Nature(London)292:378-380.38.Sambrook,J.,E.F.Fritsch,andT.Maniatis.1989.Molecularcloning:alaboratorymanual,2nded.ColdSpringHarborLaboratory,ColdSpringHarbor,N.Y.39.Seizer,G.,andJ.Tomizawa.1982.SpecificcleavageofthepiSAprimerprecursorbyribonucleaseHattheoriginofDNAreplication.Proc.Natl.Acad.Sci.USA79:7082-7086.40.Singer,M.,T.A.Baker,G.Schnitzler,S.M.Deischel,M.Goel,W.Dove,K.J.Jaacks,A.D.Grossman,J.W.Erickson,andC.A.Gross.1989.AcollectionofstrainscontaininggeneticallylinkedalternatingantibioticresistanceelementsforgeneticmappingofEscherichiacoli.Microbiol.Rev.53:1-24.41.Tamm,J.,andB.Polisky.1985.CharacterizationoftheColElprimer-RNA1complex:analysisofadomainofColElRNA1necessaryforitsinteractionwithprimerRNA.Proc.Natl.Acad.Sci.USA82:2257-2261.42.Tomizawa,J.1985.ControlofColElplasmidreplication:initialinteractionofRNAIandtheprimertranscriptisreversible.Cell40:527-535.43.Tomizawa,J.,andT.Itoh.1981.PlasmidColElincompatibilitydeterminedbyinteractionofRNAIwithprimertranscript.Proc.Natl.Acad.Sci.USA78:6096-6100.44.Tomizawa,J.,andT.Itoh.1982.TheimportanceofRNAsecondarystructureinColElprimerformation.Cell31:575-583.45.Tomizawa,J.,T.Itoh,G.Selzer,andT.Som.1981.InhibitionofColElRNAprimerformationbyaplasmid-specificsmallRNA.Proc.Natl.Acad.Sci.USA78:1421-1425.46.Wong,E.M.,M.A.Muesing,andB.Polisky.1982.Tempera-ture-sensitivecopynumbermutantsofColElarelocatedinanuntranslatedregionoftheplasmidgenome.Proc.Natl.Acad.Sci.USA79:3570-3574.47.Woolfson,D.N.,andD.H.Williams.1990.Theinfluenceofprolineresiduesona-helicalstructure.FEBSLett.277:185-188.VOL.17