SiteFinding-PCR

doi:10.1093/nar/gni124

SiteFinding-PCR:asimpleandefficientPCRmethodforchromosomewalking

GuihongTan,YinGao,MiaoShi,XinyueZhang,ShanpingHe,ZhangliangChenandChengcaiAn*

TheNationalLaboratoryofProteinEngineeringandPlantGeneticEngineering,CollegeofLifeSciences,PekingUniversity,Beijing100871,China

ReceivedApril7,2005;RevisedJune27,2005;AcceptedJuly18,2005

ABSTRACT

Inthispaper,wepresentanovelPCRmethod,termedSiteFinding-PCR,forgeneorchromosomewalking.ThePCRwasprimedbyaSiteFinderatalowtempera-ture,andthenthetargetmoleculeswereamplifiedexponentiallywithgene-specificandSiteFinderprimers,andscreenedoutbyanothergene-specificprimerandavectorprimer.However,non-targetmole-culescouldnotbeamplifiedexponentiallyowingtothesuppressioneffectofstem–loopstructureandcouldnotbescreenedout.Thissimplemethodprovedtobeefficient,reliable,inexpensiveandtime-saving,andmaybesuitableforthemoleculesforwhichgene-specificprimersareavailable.Moreimportantly,largeDNAfragmentscanbeobtainedeasilyusingthismethod.TodemonstratethefeasibilityandefficiencyofSiteFinding-PCR,weemployedthismethodtodochromosomewalkingandobtained16positiveresultsfrom17samples.INTRODUCTION

SeveralPCRmethodshavebeendevelopedforisolatinganunknownsegmentadjacenttoaknownDNAsequence,includ-inginversePCR(1–6),ligation-mediatedPCR(7–18)andrandomlyprimedPCR(19–24).Eachofthesemethodshasparticularadvantages.Forexample,inversePCRhashighspecificity.TAIL-PCRisespeciallysuitableforlarge-scalemanipulationbecauseitcanbeeasilymanipulatedandcanbeautomated(19).However,mostofthesemethodsrequirecomplicatedmanipulations,includingrestrictioncleavage,Southernblotanalysis,ligationortailingbeforePCRampli-fication.Withsomeofthesemethods,arbitraryprimingpro-ducesamplificationofnon-targetmolecules,whichconstitutethebulkofthefinalproduct(18).Furthermore,mostproductsofthemarelimitedtoalengthof<1kb.

Here,wereportasimpleandefficientPCRmethod,i.e.SiteFinding-PCR,forgeneorchromosomewalking.TheprincipleandtheprocedureofSiteFinding-PCRareoutlinedinFigure1,andthedetailedthermalcyclersettingsarelistedinTable1.Weverifiedthefeasibilityofourprotocolintwoproofsofprinciplestudies:(i)weamplifiedandclonedthesequencecontiguoustotheknownsequenceofanovelcyanophage,and(ii)weclonedandsequencedtheinsertionsitesofagrobacteriumT-DNA(25–27)insertedintotheArabidopsisgenome.

MATERIALSANDMETHODSTemplateDNAandoligonucleotides

CyanophageP4wasisolatedfromKunmingLakeattheSummerPalaceinBeijinganditsgenomicDNAwasextractedfromthelysateofacyanobacteriainaccordancewiththeM13phageDNApreparationprotocoldescribedbySambrookandRussell(28).ThegenomicDNAsofArabidopsismutantswereextractedaccordingtothemethoddescribedbyLiuetal(19).TheoligonucleotidesandthecorrespondingprimersoftheSiteFindersareshowninFigure2A.Thegene-specificprimersemployedareshowninFigure2BandC.SiteFinding

ThePCRmixtureincluded2mlof10·longTaqDNApoly-merasebuffer,2mlofmixeddNTPsolution(2.5mMeachofdATP,dTTP,dCTPanddGTP),0.5UoflongTaqDNApolymerase(BeijingTianWeiTimesTechnologyCo.Ltd,China),10pmolofSiteFinderand10–200ngoftemplateDNA.Thefinalvolumewasbroughtto20mlwithMilli-Qwater,andthenasinglecyclePCRcyclewasrun(Table1).NestedPCR

FortheprimaryroundofPCR,5mlofprimermixture(50pmolofSFP1,10pmolofGSP1and1·TaqDNApolymerasebuffer)wasaddedtoPCRtubesorthewellsofaPCR

*Towhomcorrespondenceshouldbeaddressed.Tel:+861062752405;Fax:+861062751526;Email:chcaian@pku.edu.cnÓTheAuthor2005.PublishedbyOxfordUniversityPress.Allrightsreserved.

Theonlineversionofthisarticlehasbeenpublishedunderanopenaccessmodel.Usersareentitledtouse,reproduce,disseminate,ordisplaytheopenaccessversionofthisarticlefornon-commercialpurposesprovidedthat:theoriginalauthorshipisproperlyandfullyattributed;theJournalandOxfordUniversityPressareattributedastheoriginalplaceofpublicationwiththecorrectcitationdetailsgiven;ifanarticleissubsequentlyreproducedordisseminatednotinitsentiretybutonlyinpartorasaderivativeworkthismustbeclearlyindicated.Forcommercialre-use,pleasecontactjournals.permissions@oupjournals.org

e122NucleicAcidsResearch,2005,Vol.33,No.13PAGE2

OF

7

Figure1.SchematicoutlineofSiteFinding-PCRmethodforchromosomewalking.Knownandunknownsequencesaredepictedwiththickandthinlines,respectively.BluesegmentsshowtheexpectedSiteFindertargets.Gene-specificprimers(GSPs)1–3canannealwithknownsequences(whitearrows).(1)Originalgenomicdouble-strandtemplates,showingtargetmoleculeandnon-targetmolecules.(2)SiteFindingreaction:afterlowtemperatureprimingbyaSiteFinder,onestrandofthetargetgenewasreplacedbylongTaqDNApolymerase,whichgenerateddouble-strandedtargetmoleculesofdifferentlengths.(3)NestedPCR:thetargetDNAwasexponentiallyamplifiedbynestedPCRwithGSPsandSiteFinderprimers(SFPs)1and2,whilenon-targetgeneamplificationwassuppressedbythestem–loopstructureoftheDNA.(4)Cloningtargetmolecules:afterbeingcleavedwithNotI,thePCRproducts(generatedbyGSP2andSFP2)werepurifiedbyagarosegelelectrophoresis,andthenthepurifiedDNAwasclonedintoapBluescriptSK(+)vectorlinearizedbyNotIandEcoRV.(5)Screeningandsequencing:thecloneswerescreenedbycolony-PCRwiththethirdgene-specificprimer(GSP3)andavectorprimer(M13reverseprimerorT3primer),andthetargetmoleculeswerescreenedoutandsequencedsubsequently.

plate(25mlfinalvolume)onice,andthenthePCRwasrunfor30cycles(Table1).Forthesecondaryreaction,1mloftheprimaryPCRproductswasdilutedinto100–1000mlMilli-Qwater,andthen1mlofthedilutedproductswascombinedwith49mlofthesecondaryPCRmixture,whichcontained1·longTaqDNApolymerasebuffer,25mMdNTPs,0.8UoflongTaqpolymerase,0.2mMeachofinternalspecificprimer(GSP2)andinternalSiteFinderprimer(SFP2),andthenthePCRwas

PAGE3

OF

7NucleicAcidsResearch,2005,Vol.33,No.13e122

runfor30cycles(Table1).Meanwhile,anotherPCRwasruninwhichallofthecomponentsandtheirconcentrationswerethesameasdescribedaboveexceptthatGSP3wasusedinsteadofGSP2.Finally,theproductswereseparatedon1.2%agarosegels.Specificitywasjudgedbythedifferenceinproductsize,whichwasconsistentwiththedistancebetweenthepositionsofGSP2andGSP3.Cloningthetargetmolecules

(i)ThePCRproductsweredirectlycleavedasfollows:43mlofPCRproductsweremixedwith5mlof10·enzymebufferand2ml(20U)ofNotI(NewEnglandBiolabs,USA),andthenincubatedovernightat37󰀁C.(ii)ThedigestedPCRproductswereseparatedbyelectrophoresison1.2%agar-osegelsandpurifiedwithWizardÒSVGelandPCRClean-UpSystem(Promega,USA).(iii)ThepurifiedDNAwascloned

Table1.CyclingconditionsusedforSiteFinding-PCRonPTC-200PeltierTermalCyclerReactionSiteFindingPrimarySecondary

Cycle(s)113011301

Thermalcondition92󰀁C25󰀁C68󰀁C94󰀁C95󰀁C72󰀁C94󰀁C95󰀁C72󰀁C

(2min),95󰀁C(1min),(1min),rampto

over3min,68󰀁C(10min)(1min)

(10s),68󰀁C(6min)(5min)(1min)

(10s),68󰀁C(6min)(5min)

intothepBluescriptSK(+)vectorbetweentherestrictionsitesofNotIandEcoRV.Screeningandsequencing

Severalcloneswereselectedatrandomandtransferredindi-viduallyintoPCRtubes,eachofwhichcontained15mlofthefollowingPCRmixture:1·TaqDNApolymerasebuffer,25mMdNTPs,0.5UofTaqDNApolymerase(BeijingTianWeiTimestechnologyCo.Ltd),0.2mMoftheM13reverseprimer(50-CAGGAAACAGCTATGAC-30)orT3sequencingprimer(50-AATTAACCCTCACTAAAGGG-30),andthecorrespondinggene-specificprimer3(GSP3).Theclonesweresubjectedto30cyclesofPCR,andthereac-tionparameterswereasfollows:denaturationat95󰀁Cfor10s,annealingat56󰀁Cfor30sandextensionat72󰀁Cfor2.5min,plusaninitialdenaturationstepof1minat94󰀁Candafinalextensionof5minat72󰀁C.ThePCRproductswereexaminedona1.2%agarosegel,andthepositivecloneswereselectedforsequencing(DNAsequencingwascarriedoutbyDalianTaKaRaBiotechnologyCo.Ltd,China).DNAsequenceanal-yseswerecarriedoutusingtheBLASTprogram(http://ncbi.nlm.nih.gov).RESULTS

IdentificationofanunknownregiononcyanophageDNAbySiteFinding-PCR

TheCyanophageP4genomicDNAwasextractedfromthelysateofacyanobacteria,anda769bpfragmentwascloned

Figure2.(A)SequencesoftwoSiteFindersandtheirprimers(SFP1andSFP2).SiteFinder-1and2differedonlyattheir30ends,andcontainedararerestrictionenzymesiteforNotI,whichfacilitatescloningwithcommonlyusedvectors,suchaspBluescriptSK(+).Thefourspecificnucleotidesunderlinedwithbluebaratthe30endsoftheSiteFinders,withthehelpofNNNNNN,wereusedtoannealwiththecomplimentarysitesongenomicDNAsatlowtemperatureandinitiateSiteFinding-PCR.SFP1andSFP2wereusedintheprimaryandsecondaryreactions,respectively.(B)Threegene-specificprimers(GSPs)forCyanophageP4(P4-1,P4-2andP4-3)areindicatedbyblackarrows.ThedistancefromP4-2toP4-3was31bp.(C)ThreeGSPsforArabidopsisT-DNAinsertionmutants(DL1,DL2andDL3)designedbasedontheT-DNAsequenceofpSki015areindicatedbyblackarrows.ThedistancefromDL2toDL3was59bp.

e122NucleicAcidsResearch,2005,Vol.33,No.13PAGE4

OF

7

intothepBluescriptSK(+)vectorandsequenced(M.Shietal.,unpublisheddata).Basedontheknownsequence,threepri-mers,P4-1(GSP1),P4-2(GSP2)andP4-3(GSP3),weredesigned(Figure2B).AftertworoundsofPCR,theproductswereseparatedona1.2%agarosegelandexamined.ADNAfragmentof󰀃4.5kbwasrecoveredandclonedintothepBlue-scriptSK(+)vector.EightcloneswerescreenedwiththeM13reverseprimerandP4-3,amongwhichfivepositivecloneswereobtained.Oneclonewitha󰀃4.5kbCyanophageP4DNAfragmentwasselectedandsequenced.Thesequencingresults(SupplementaryMaterial1)showedthattheclonedproductwas4617bpinlength(Figure3B,I)andcontainedanothertwoGCCTsites,oneofwhichwas293bpawayfromtheP4-2locus,andtheotherwas707bpawayfromtheP4-2locus(Figure3B,I).Thesequencewassubsequentlycon-firmedbyCyanophageP4genomesequencing(M.Shietal.,unpublisheddata).Toevaluatethefeasibilityofthisapproachforcomplicatedgenomes,DNAmixtures(0.1ngofCyanophageP4DNAplus300ngofricegenomeDNA)wereutilizedasoriginaltemplates,andweobtainedthesameresultsasthoseusingtheP4CyanophagegenomeDNAastheoriginaltemplates.

IdentificationoftheT-DNAinsertionsitesofArabidopsismutants

Toverifythefeasibilityofthisapproachforacomplicatedgenome,weusedittoidentifyanArabidopsismutantsiteinsertedbyT-DNA(25–27)ofthebinaryvectorpSki015(29,30),whichhadformerlybeenidentifiedbyTAIL-PCR(SupplementaryMaterial2).ThreeGSPs(DL1,DL2andDL3),designedfromknownregionsadjacenttotheleftborderoftheT-DNAinpSki015(Figure2C),wereusedtogetherwithtwoSFPs(SFP1andSFP2).AftertworoundsofPCR,weobtainedtwospecificDNAbandsof󰀃2.2kband󰀃0.7kb(Figure3A,lane3).TheDNAwasrecoveredfromthe󰀃2.2kbband,andthenclonedandsequenced(Figure3B,II).Thesequencingresultsshowedthattheinsertionsitewasconsis-tentwiththatobtainedbyTAIL-PCR(SupplementaryMate-rial2).Tofurtherdemonstratethefeasibility,efficiencyand

universalityoftheSiteFinding-PCRmethod,thegenomicDNAsfrom15otherArabidopsismutantswereusedastem-platestoidentifyArabidopsisgenomicsequencesflankingtheT-DNAleftborders.Weobtainedspecificbandsfrom14mutants(Figure4AandB),amongwhich9samples(samples1–3and8–13)hadformerlybeenidentifiedbyTAIL-PCR(SupplementaryMaterials3and4).TheotherscouldnotbeidentifiedsuccessfullybyTAIL-PCR.Here,only1outofthe15samplesdidnotshowspecificamplification(datanotshown).Weclonedandsequenced28specificbandsindicatedwithwhitearrowsinFigure4AandB.Sequenceanalyses(SupplementaryMaterial4)revealedthatwehadidentifiedthesitesinsertedbyT-DNAfor14samples,amongwhichSamples3and5hadtwoinsertionsites(Figure4AandC).DISCUSSION

PrincipleofSiteFinding-PCR

ItiswellknownthatthecontroloffalseprimingisveryimportantforPCR.However,atthe30endofaprimer,theremustbeseveralnucleotidesthatcomplementthetem-plateaccurately,eveninthecaseoffalsepriming.WeutilizedthischaracteristictoinitializeSiteFinding-PCRinourproto-col.Theoligonucleotide50-NNNNNNGCCT-30atthe30endofSiteFinder-1or50-NNNNNNGCGC-30atthe30endofSiteFinder-2wasutilizedtofindtheGCCTorGCGCsitesonthetargetmoleculesandnon-targetmolecules.SiteFinding-PCRamplificationcouldthenbeprimedatlowtemperature.Innon-targetmoleculeamplification,thePCRproductscon-taineddouble-strandedSiteFindersequencesatbothends,andtheendsoftheindividualDNAstrandsformedastem–loopstructurefollowingeverycycleduetothepresenceofinvertedterminalrepeats.Stem–loopstructuresaremoresta-blethantheprimer-templatehybrid,andthereforesuppressexponentialamplification(14,31,32).However,inthecaseoftargetmoleculeamplification,adistalgene-specificprimerextendsaDNAstrandthroughtheSiteFinderaftertheSiteFindingreaction,andthetargetmoleculecontainstheSiteFindersequenceonlyatoneend.Asaresult,the

Figure3.ChromosomewalkingfortheCyanophageP4andanArabidopsismutantusingtheSiteFinding-PCRmethod.SiteFinder-1wasusedtoinitializetheSiteFindingreaction.(A)ProductsofthesecondaryroundofPCR(lanes1–2forP4Cyanophageand3–4forArabidopsismutant).Lanes1–4containedPCRproductsobtainedwithprimercouplesSFP2/P4-2,SFP2/P4-3,SFP2/DL2andSFP2/DL3,respectively.(B)ClonedandsequencedPCRproducts.Lanes1and2bothshowedthreespecificproducts(A),andthelargestoneinlane1wasclonedandsequencedasindicatedin(I).TherewereanothertwoGCCTsitesonthe4617bpfragmentoftheCyanophageP4sequence,whichareindicatedwiththeblackarrowheadsin(I)(firstsiteandsecondsite).Thesefindingswereconsistentwiththegelelectrophoresisresults(whitearrowsinlane1).Lanes3and4bothshowedtwospecificproducts(A),andthelargeroneinlane3wasclonedandsequencedasindicatedin(II).TherewasanotherGCCTsiteonthe2270bpfragmentoftheArabidopsisDNAasindicatedwiththeblackarrowheadsin(II)(firstsite),whichwasalsoconsistentwiththegelelectrophoresis(whitearrowinlane3).

PAGE5

OF

7NucleicAcidsResearch,2005,Vol.33,No.13e122

Figure4.IdentificationofT-DNAinsertionsitesof14ArabidopsismutantswithSiteFinding-PCR.LanesIIandIII:theproductsgeneratedbySFP2/DL2andSFP2/DL3,respectively.SamplesarelabeledasS1–S14.(A)TheproductsofS1–S7,initiallyprimedbySiteFinder-1.WhitearrowsinlaneIIindicatedtheclonedproducts.Digitalsinboxedareasshowdifferentinsertionsites.AsforS3,1-1and1-2showtwodifferentproductsofthe1stinsertionsite;2-1,2-2and2-3inS3refertothreeproductsofthe2ndinsertionsite.(B)Theproductsofsamples8–14,initiallyprimedbySiteFinder-2.WhitearrowsinlaneIIindicatedtheclonedproducts.(C)Comparisonofthe30endoftheSiteFinderwiththelargestorlargerfragment.BoldblacklinesegmentsandgreylinesegmentsrepresenttheT-DNAandArabidopsisDNAs,respectively.ThearrowsrepresentSFP2.Thesizesofthelargestorlargerspecificfragmentoftheinsertedsiteareindicatedontheright,andthesmalleronesaremarkedinthemiddlewithbluebars.NNNNNNwith0–3mismatchnt(courier)helpedthe30ends(underlinedwithbluebars)ofSiteFinder-1andSiteFinder-2toannealaccuratelywith30-CGGA-50(GCCTsite)and30-CGCG-50(GCGCsite),respectively.

stem–loopstructuredoesnotform,andthusthePCRampli-ficationcanproceedsmoothly,usingalongdistancethermo-stableDNApolymerase(longTaqDNApolymerase)withatwo-temperaturecyclingprotocol(Table1).

Twokeypointspertainingtothismethodmaybenoted.First,the4–6ntoligonucleotideatthe30endoftheSiteFinderinitializesthereactionatlowtemperature;second,thestem–loopstructuresuppressesexponentialamplificationeffi-ciently.Inoperation,weutilizedtwo4ntoligonucleotides,GCCTandGCGC,toprimethePCRat25󰀁C.Infact,any4,5or6ntoligonucleotidescouldbeused,whilethelengthandconstitutionoftheoligonucleotideswouldinfluencetheresults.Importantly,theannealingtemperatureintheSiteFindingreactionshouldbeadjustedappropriatelyfor

e122NucleicAcidsResearch,2005,Vol.33,No.13thespecificoligonucleotidesused.Forexample,wecouldusea6ntoligonucleotideatahighertemperaturetoobtainlongertargetmoleculesanda4ntoligonucleotideatalowertem-peraturetoobtainshortertargetmolecules.Inthispaper,weobtainedmorespecificDNAbandsusingGCCT(Figure4A)thanusingGCGC(Figure4B),becausethelatterexhibitedaloweremergencefrequencyintheArabidopsisgenome(18).TohelptheSiteFinderannealwithcomplementarysitesongenomicDNA,oligonucleotideswithanappropriatenumberofrandomnucleotidesorinosinestriphosphateadjacenttothefixed4–6ntcouldbesynthesized.Inourexperiment,sequenc-ingresultsdemonstratedthatthe30endsofSiteFinder-1andSiteFinder-2accuratelyannealedwith30-CGGA-50(GCCTsite)and30-CGCG-50(GCGCsite),respectively,underthehelpofNNNNNNwith0–3mismatchnucleotides(courierinFigure4C)owingtolowstringencySiteFindingpriming.FurtheranalysesshowedthatthemismatchmightbecreatedduringPCRcycles,becausetherewere󰀃0.1%errorratesoccurredbyextendingalignmentswithArabidopsisgenomesequence(SupplementaryMaterial4).

SiteFinderscontainararerestrictionenzymesiteforNotI,andPCRamplificationwasperformedusingalongTaqDNApolymerase,whichcouldonlyproducethefragmentswithbluntends.Therefore,afterdigestingtheproducts(generatedbySFP2andGSP2)withNotI,weobtainedtargetmoleculeswithonestaggeredendandonebluntend(Figure1,panel4),whichfacilitatedthecloningoftargetmoleculesintopBlue-scriptSK(+)linearizedbyNotIandEcoRV(Figure1,panel4).Incontrast,non-targetmoleculeshadeithertwoadhesiveter-minalsatbothendsorastem–loopstructurewithoneadhesiveterminal(Figure1,panel3),whichpreventedthemfrombeingcloned.Furthermore,aninternalspecificprimer(GSP3)wasusedtoscreentheclones,whichensuredthatonlythoseclonespossessingthespecificproductswouldbescreenedoutbecauseonlythetargetmoleculespossessedthecomplemen-tarysiteforGSP3(Figure1,panel5).AdvantagesofSiteFinding-PCR

Ingeneral,therearethreekindsofPCRmethodsavailableforchromosomewalking:inversePCR(1–6),LM–PCR(7–18)andRP–PCR(19–24).InversePCRandLM–PCRaredepen-dentupontheuseofrestrictionendonucleasesbeforePCRandrequireadditionalsteps,whichmaydecreasetherecoveryefficiencyoftheproduct,increasethechanceofcontamina-tion,andaretime-consuming.Inaddition,itwouldbeexpen-sivebecausemanyrestrictionendonucleaseshadtobeusedduetounpredictablerestrictionendonucleasesitesinunknownregionsofthetargetmolecule.RP–PCR,e.g.TAIL–PCR,isapopularmethodforchromosomewalking,especiallyfortheidentificationofT-DNAortransposoninsertions,butitsamplifiedproductsareusuallysmall,andtheoptimumcon-ditionshavetobeestablishedempirically.

ComparedwithotherPCRmethodsforchromosomewalk-ing,theSiteFinding-PCRmethodhasseveraladvantages.(i)Simplicity:theSiteFinding-PCRdoesnotrequireadditionalmanipulations(suchasSouthernblotanalysis,restrictiondigestion,ligationortailing)beforePCR,andtheproducts’specificitycanbeconfirmedbyasimpleagarosegelelectro-phoresisanalysis.(ii)Specificity:becauseofthestem–loopstructuresuppressioneffect,co-amplificationofnon-specific

PAGE6

OF

7

productsisveryweak.Furthermore,anynon-specificproductwouldnotbeclonedandscreenedout(Figure1,panels4and5).(iii)Sensitivity:weused10ngofArabidopsisgenomicDNAastheoriginaltemplateandobtainedgoodresults.(iv)Efficiency:weobtained16positiveresultsfrom17samples,andonlyonereactiondidnotyieldspecificproducts,whichmaybeduetothelackofcorrespondingsitesadjacenttotheknownsequence.However,anotherSiteFindercouldbeusedtocompletethewalking,i.e.SiteFinding-PCRmaybepoten-tiallysuitableforanyDNAonwhichgene-specificprimerscanbedesigned.(v)Longspecificproduct:weobtained17specificproductsthatwere>1kb(SupplementaryMaterial4),2ofwhichwere>4.5kb(Figures3AandFigure4AandB).(vi)Lowcostandtime-saving:theenzymesrequiredfortheprocedureareonlylongTaqDNApolymerase,NotIandT4DNAligase;thePCRamplificationcanbecompletedconsecutivelywithin6–8h.

ThepotentialapplicationsofSiteFinding-PCR

TheadvantagesoftheSiteFinding-PCRmethodmakeitusefulforthefollowingapplications:(i)rapidlyobtainingafullgenesequenceorfindingpromotersandregulatoryelementsfromclonedcDNAfragments,(ii)determiningtheexon–intronboundarieswithingenes,(iii)walkingupstreamordownstreamfromsequence-taggedsitesinknownorunknowngenomesforgenefunctionstudies,(iv)obtainingnon-conservedregionsofgenesinuncharacterizedspecies,accord-ingtotheconservedsequencesofreportedgenesand(v)fillinginthegapsingenomesequencing.SUPPLEMENTARYMATERIAL

SupplementaryMaterialisavailableatNAROnline.ACKNOWLEDGEMENTS

WearegratefultoDrXinLiandDrYanxiaZhangforprovidingArabidopsismaterials,andourcolleaguesYeWang,ShuangliMi,WanqiangQianandJiayuGufortheirhelpfulsuggestionsduringthepreparationofthismanuscript.ThisworkwassupportedbygrantsfromtheNationalKeyBasicScience‘973’Program(No.G20000016204),theState‘863’HighTechnologyR&DProgram(No.2004AA214072andNO.2002AA206631)andtheMinistryofEducationDoctor’sFoundationProgram(No.20030001083)oftheChineseGovernment.FundingtopaytheOpenAccesspublicationchargesforthisarticlewasprovidedbyPekingUniversity,P.R.China.

Conflictofintereststatement.Nonedeclared.REFERENCES

1.Triglia,T.,Peterson,M.G.andKemp,D.J.(1988)AprocedureforinvitroamplificationofDNAsegmentsthatlieoutsidetheboundariesofknownsequences.NucleicAcidsRes.,16,8186.

2.Ochman,H.,Gerber,A.S.andHartl,D.L.(1988)Geneticapplicationsofaninversepolymerasechainreaction.Genetics,120,621–623.

3.Silverman,G.,A.,Ye,R.D.,Pollock,K.M.andSadlerJ.E.,Korsmeyer,S.J.(1989)Useofyeastartificialchromosomeclonesformappingandwalkingwithinhumanchromosomesegment18q21.3.Proc.NatlAcad.Sci.USA,86,7485–7489.

PAGE7

OF

74.Huckaby,C.S.,Kouri,R.E.,Lane,M.J.,Peshick,S.M.,Carroll,W.T.,Henderson,S.M.,Faldasz,B.D.,Waterbury,P.G.andVournakis,J.N.(1991)Anefficienttechniqueforobtainingsequencesflankinginsertedretroviruses.Genet.Anal.Tech.Appl.,8,,151–158.

5.Arveiler,B.andPorteous,D.J.(1991)AmplificationofendfragmentsofYACrecombinantsbyinversepolymerasechainreaction.Technique,3,24–28.

6.Keim,M.,Williams,R.S.andHarwood,A.J.(2004)AninversePCRtechniquetorapidlyisolatetheflankingDNAofdictyosteliuminsertionmutants.Mol.Biotechnol.,26,221–224.

7.Riley,J.,Butler,R.,Ogilvie,D.,Finniear,R.,Jenner,D.,Powell,S.,

Anand,R.,Smith,J.C.andMarkham,A.F.(1990)Anovel,rapidmethodfortheisolationofterminalsequencesfromyeastartificialchromosome(YAC)clones.NucleicAcidsRes.,18,2887–2890.

8.Rosenthal,A.andJones,D.S.C.(1990)Genomicwalkingandsequencingbyoligo-cassettemediatedpolymerasechainreaction.NucleicAcidsRes.,18,3095–3096.

9.Rosenthal,A.,MacKinnon,R.N.andJones,D.S.C.(1991)PCRwalingfrommicrodissectioncloneM54identifiesthreeexonsfromthehumangenefortheneuralcelladhesionmoleculeL1(CAM-L1).NucleicAcidsRes.,19,5395–5401.

10.Espelund,M.andJakobsen,K.S.(1992)Cloninganddirectsequencingof

plantpromotersusingprimeradaptermediatedPCRonDNAcoupledtoamagneticsolidphase.BioTechniques,13,74–81.

11.Jones,D.H.andWinistorfer,S.C.(1992)Sequencespecificgenerationofa

DNApanhandlepermitsPCRamplificationofunknownflankingDNA.NucleicAcidsRes.,20,595–600.

12.Jones,D.H.andWinistorfer,S.C.(1993)Genomewalkingwith2-to4-kb

stepsusingpanhandlePCR.PCRMethodsAppl.,2,197–203.

13.Warshawsky,D.andMiller,L.(1994)Arapidgenomicwalkingtechnique

basedonligation-mediatedPCRandmagneticseparationtechnology.BioTechniques,16,792–798.

14.Siebert,P.D.,Chenchik,A.,Kellogg,D.E.,Lukyanov,K.A.and

Lukyanov,S.A.(1995)AnimprovedPCRmethodforwalkinginunclonedgenomicDNA.NucleicAcidsRes.,23,1087–1088.

15.Sterky,F.,Holmberg,A.,Alexandersson,G.,Lundeberg,J.andUhlen,M.

(1998)Directsequenceingofbacterialartificialchromosone(BACs)andprokaryoticgenomesbybiotin-capturePCR.J.Biotechnol.,246,810–813.

16.Dai,S.M.,Chen,H.H.,Chang,C.,Riggs,A.D.andFlanagan,S.D.(2000)

Ligation-mediatedPCRforquantitativeinvivofootprinting.Nat.Biotechnol.,18,1108–1111.

17.Zhang,Z.andGurr,S.J.(2002)A‘stepdown’PCR-basedtechniquefor

walkingintoandthesubsequentdirect-sequenceanalysisofflankinggenomicDNA.MethodsMol.Biol.,192,343–350.

18.Yuanxin,Y.,Chengcai,A.,Li,L.,Jiayu,G.,Guihong,T.andZhangliang,C.

(2003)T-linker-specificligationPCR(T-linkerPCR):anadvancedPCR

NucleicAcidsResearch,2005,Vol.33,No.13e122

techniqueforchromosomewalkingorforisolationoftaggedDNAends.NucleicAcidsRes.,31,e68.

19.Liu,Y.G.,Mitsukawa,N.,Oosumi,T.andWhittier,R.F.(1995)EfficientisolationandmappingofArabidopsisthalianaT-DNAinsertjunctionsbuthermalasymmetricinterlacedPCR.PlantJ.,8,457–463.

20.Liu,Y.G.andWhittier,R.F.(1995)ThermalasymmetricinterlacedPCR:automatableamplificationandsequencingofinsertendfragmentsfromP1andYACclonesforchromosomewalking.Genomics,25,674–681.21.Terauchi,R.andKahl,G.(2000)RapidisolationofpromotersequencesbyTAIL-PCR:the5-flankingregionsofPalandPgigenesfromyams(Dioscorea).Mol.Gen.Genet.,263,554–560.

22.

Weigel,D.,Ahn,J.H.,Blazquez,M.A.,Borevitz,J.O.,Christensen,S.K.,Fankhauser,C.,Ferrandiz,C.,Kardailsky,I.,Mlancharuvil,E.J.,Neff,M.M.etal.(2000)ActivationtagginginArabidopsis.PlantPhysiol.,122,1003–1013.

23.Singer,T.andBurke,E.(2003)High-throughputTAIL-PCRasatooltoidentifyDNAflankinginsertions.MethodsMol.Biol.,236,241–272.

24.

Antal,Z.,Rascle,C.,Fevre,M.andBruel,C.(2004)SingleoligonucleotidenestedPCR:arapidmethodfortheisolationofgenesandtheirflankingregionsfromexpressedsequencetags.Curr.Genet.,46,240–246.

25.Chang,S.S.,Park,S.K.,Kim,B.C.,Kang,b.J.,Kim,D.U.andNam,H.G.(1994)StablegenetictransformationofArabidopsisthalianabyAgrobacteriuminoculationinplanta.PlantJ.,5,551–558.

26.Gelvin,S.B.(2003)Agrobacterium-mediatedplanttransformation:thebiologybehindthe‘gene-jockeying’tool.MicrobiolMol.Biol.Rev.,67,16–37,tableofcontents.

27.Tzfira,T.,Li,J.,Lacroix,B.andCitovsky,V.(2004)AgrobacteriumT-DNAintegration:moleculesandmodels.TrendsGenet.,20,375–383.28.Sambrook,J.andRussell,D.W.(2001)MolecularCloning:ALaboratoryManual.3rdedn.ColdSpringHarborLaboratoryPress,ColdSpringHarbor,NY.

29.Memelink,J.(2003)T-DNAactivationtagging.MethodsMol.Biol.,236,345–362.

30.

Weigel,D.,Ahn,J.H.,Bla

´zquez,M.A.,Borevitz,J.O.,Christensen,S.K.,Fankhauser,C.,Ferra

´ndiz,C.,Kardailsky,I.,Malancharuvil,E.J.,Neff,M.M.etal.(2000)ActivationtagginginArabidopsis.PlantPhysiol.,122,1003–1013.

31.

Matz,M.,Shagin,D.,Bogdanova,E.,Britanova,O.,Lukyanov,S.,Diatchenko,L.andChenchik,A.(1999)AmplificationofcDNAendsbasedontemplate-switchingeffectandstep-outPCR.NucleicAcidsRes.,27,1558–1560.

32.

Chenchik,A.,Diachenko,L.,Moqadam,F.,Tarabykin,V.,Lukyanov,S.andSiebert,P.D.(1996)Full-lengthcDNAcloninganddeterminationofmRNA50and30endsbyamplificationofadaptor-ligatedcDNA.BioTechniques,21,526–534.