METHODOLOGY Open Access An improved allelespecific PCR - PDF document

METHODOLOGYOpenAccess Animprovedallele-specificPCRprimerdesign methodforSNPmarkeranalysisandits application JingLiu 1 † ,ShunmouHuang † ,MeiyuSun 1 ,ShengyiLiu 1 ,YumeiLiu 2 ,WanxingWang 2 ,XiurongZhang 1 , HanzhongWang 1 andWeiHua 1* Abstract Background: AlthoughSingleNucleotidePolymorphism(SNP)markerisaninvaluabletoolforpositionalcloning, associationstudyandevolutionaryanalysis,lowSNPdetectionefficiencybyAllele-SpecificPCR(AS-PCR)stillrestricts itsapplicationasmolecularmarkerlikeothermarkerssuchasSimpleSequenceRepeat(SSR).Toovercomethis problem,primerswithasinglenucleotideartificialmismatchintroducedwithinthethreebasesclosesttothe3 ’ end (SNPsite)havebeenusedinAS-PCR.However,foroneSNPsite,ninepossiblemismatchescanbegenerated amongthethreebasesandhowtoselecttherightonetoincreaseprimerspecificityisstillachallenge. Results: Inthisstudy,differentfromthepreviousreportswhichusedalimitedquantityofprimersrandomly (severalordozenpairs),wesystematicallyinvestigatedtheeffectsofmismatchbasepairs,mismatchsitesandSNP typesonprimerspecificitywith2071primerpairs,whichweredesignedbasedonSNPsfrom Brassicaoleracea 01-88and02-12.Accordingtothestatisticalresults,we(1)foundthattheprimersdesignedwithSNP(A/T),in whichthemismatch(CA)inthe3 rd ’ end,hadthehighestallele-specificity(81.9%).This informationcouldbeusedwhendesigningprimersfromalargequantityofSNPsites;(2)performedtheprimer designprinciplewhichformstheoneandonlybestprimerforeverySNPtype.Thisisneverreportedinprevious studies.Additionally,wefurtheridentifieditsavailabilityinrapeseed( BrassicanapusL. )andsesame( Sesamum indicum ).Highpolymorphismpercent(75%)ofthedesignedprimersindicateditisageneralmethodandcanbe appliedinotherspecies. Conclusion: ThemethodprovidedinthisstudycangenerateprimersmoreeffectivelyforeverySNPsitecompared tootherAS-PCRprimerdesignmethods.Thehighallele-specificefficiencyoftheSNPprimerallowsthefeasibility forlow-tomoderate-throughputSNPanalysesandismuchsuitableforgenemapping,map-basedcloning,and marker-assistedselectionincrops. Keywords: SNP,AS-PCR,Mismatch,Polymorphism,Destabilization Introduction SingleNucleotidePolymorphisms(SNPs)aresinglebase differencesbetweenDNAofdifferentindividuals.Once discovered,SNPscanbeconvertedintogeneticmarkers thatcanbeassayed[1,2].Asthemostabundantandsta- bileformofgeneticvariationinmostorganism genomes,SNPsaremoresuitableforgenotypingmar- kerscomparedtotheconventionalmarkerssuchas AFLP(Amplifiedfragmentlengthpolymorphism)and SSR(SimpleSequenceRepeat).Withthedevelopmentof bio-technology,SNPsarebecomingfavoredgeneticmar- kersthatareusedinmarker-assistedbreeding[3],map- basedcloning[4],studyofevolutionaryconservations betweendifferentspecies[5,6],andthedetectionofrisk- associatedalleleslinkedtohumandiseases[7]. *Correspondence: huawei@oilcrops.cn † Equalcontributors 1 KeyLaboratoryofBiologyandGeneticImprovementofOilCrops,Ministry ofAgriculture,OilCropsResearchInstituteoftheChineseAcademyof AgriculturalSciences,Wuhan430062,People ’ sRepublicofChina Fulllistofauthorinformationisavailableattheendofthearticle PLANT METHODS ©2012Liuetal.;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycited. Liu etal.PlantMethods 2012, 8 :34 http://www.plantmethods.com/content/8/1/34 Recently,massiveparallelsequencingplatformssuchasGSFLX(Roche),Solexa(Illumina)andSOLID(Ap-pliedBiosstems)havesignificantlyreducedthecostofhighthroughoutsequencing[8].Alargenumberofgen-omesandtranscriptomeshavebeenrapidlysequencedusingthesenewplatformstoidentifynovelSNPsinmaize[9],rapeseed[10]andhuman[11]etc.Alargevarietyoftechniquesforhigh-throughputSNPgenotyp-inghavealsobeendevelopedusingTaqman[12],Ampli-fluor[13],genomere-sequencing[14,15],andSNParrays[16,17].Thesetechniquesareexpensiveandre-quirespecializedequipments,whichcostmorestandardprimersandarenotpracticalforassayinglow-tomoderate-throughputSNPs.Hence,thereisaneedforsimpleandaccurategenotypingassaysthatcanbeimplementedinlaboratorieslackingaccesstosophisti-catedequipment.TraditionalSNPgenotypingmethodssuchasCAPs(TheCleavedAmplifiedPolymorphicSequence),dCAPs(derivedCAPS),andAS-PCR(Allele-specificPCR)arewidelyusedforlow-throughputapplicationsinplantre-search.Inapplication,CAPSanddCAPSarerestrictedbyendnucleasesitesthatcouldbeinefficientandnotcost-effective[18-20].AS-PCRisbasedontheextensionofprimeronlywhenits3endisaperfectlycomplemen-tedtothetemplate[21].Inprinciple,SNPscanbedetectedusingallele-specificPCRprimersbasedontheterminalnucleotideofaprimerthatcorrespondstoaspecificSNPsite.However,reliablediscriminationbe-tweentheallelesisnotsufficienttoachieveusingthisdescribedmethod.Toovercomethisproblem,allele-specificprimerswithanadditionalbasepairchangewithinthethreebasesclosesttotheSNPsitebetweenalleleshavebeenused[21,22].EachspecificSNPsiteinanallelecangenerateatleast18possibleprimerswithonemismatchbase[23].TheSNAPERprogramgener-atesalistofupto16possibleprimersperSNPsiteforeachallele[23].Therefore,choosingadditionalmis-matchestoincreaseprimerspecificityhasbeenachal-lengeforAS-PCR[23].SomestudieshaveproposedcriteriafordesigningAS-PCRprimers.Hayashietal(2004)proposedthatbasepairmismatchescreatedthroughT-GorC-Atransversionsatthirdbasefrom3endcouldincreasetheallele-specificity[24].Hirotsu(2010)identifiedA-TtransversionandA-Gtransitionwereusefulbasepairmismatchesforimprovementofallele-specificamplification[25].TheWASPtoolcouldalsobeusedtointroducemismatchesatthepenultimatetotheterminal)baseoftheprimer[26,27].How-ever,moststudiesusedonlyalimitedquantityofpri-mers,whichmighthavesomeinfluencesonefficiencyofSNPprimerspecificity.Inthisstudy,over2000primerpairs,whichweredesignedbasedonSNPsbetweenB.oleracealines01-88and02-12,wereusedtoanalyzetheeffectsofdifferentSNPtypes,mismatchbasesandsiteswithinthethreebasesclosesttothe3endonprimerspecificity.Basedontheseresults,weadvancedtheSNPprimerdesignprinciple.ComparedtotraditionalSNPgenotypingmethods,ourmethodcouldprovideacost-effectiveal-ternativeforhighefficientspecificprimersandwouldgreatlyfacilitateplantresearch.SNPanalysisofB.oleracea01-88and02-12genomeTheassemblyofgenomesequencesofB.oleracea02-12hasbeenaccomplished(unpublished).ToidentifySNPsbetweenB.oleracealines01-88and02-12,gen-omeDNAofline01-88wasre-sequencedandatotalof119millionreadswereobtained.Togethigh-qualitySNPs,thesequencedatawassubjectedtostringentfil-tering:Thereadsfromline01-88werecomparedtothesequencesofline02-12usingBLASTN.Sitescontaining Table1PutativeSNPsidentifiedbetweenB.oleraceagenomesof01-88and02-12SNP(01-88/02-12)No.ofeverySNPtypeCausedbyTotalNo.oftransitionandtransversionA/G201247transition806387C/T202316G/A202801transitionT/C200023A/C79813transversion615726G/T79800C/A80029transversionT/G79782A/T99848transversionT/A99743G/C48372transversionC/G48339etal.PlantMethodsPage2of9http://www.plantmethods.com/content/8/1/34 tri-allelicorhighdegreeofpolymorphismwereomitted. Thesequencescontainingover8readsmappedto uniquesitesinthegenomesequencewereextracted fromline01-88.Pairwisealignmentwasusedtoevalu- atetheSNPsbetweengenomesequencesof01-88and 02-12.Thealignmentresultrevealedatotalof1,422,113 SNPsexistedbetween B.oleracea lines01-88and02-12 (anaverageofoneSNPinevery360bpfragment) (Table1).AnalysistotheSNPsshowedoverhalf(56.7%, 806387/1422113)ofthenucleotidechangesweretransi- tions(A-GorC-T).Transversions(A-T,A-C,C-Gand G-T)accountedfor43.3%(615726/1422113)ofthe detectedSNPs(Table1). TofurtheridentifytheputativeSNPsandestimatethe proportionoffalsepositives,96SNPsitesderivedfrom only8-readsequenceswerechosenrandomly.Primers weredesignedaccordingtothegenomesequencesnear theseSNPsitesandallampliconscouldgenerateabout 500bpfragmentsinwhichcontainingthecorresponding SNPs.Sangersequencingresultsshowed93SNPswere identicalwiththeputativeSNPsand3SNPswereun- predictedorundetectable.Thisindicatedaveryhigh proportionofSNPsreallyexistedbetweenlines01-88 and02-12. DesignandefficiencydetectionofSNPprimers Primerdesignstrategyforallele-specificPCRwasillu- stratedinFigure1.PrimerP1,whichwasdesignedbased on02-12genomesequence,formedamismatch(TG)in 3 ’ endwiththeDNAsequenceof01-88.Whileinmost cases,itstillcouldamplifythebandwith01-88effi- ciently.Afterintroducinganothermismatch(GA)inthe 2 nd siteclosesttothe3 ’ endinprimerP2,itcouldonly amplifythebandwith02-12(Figure1a).Basedonthe methoddescribedbyCha etal (1992),mismatchsites andmismatchbasesclosesttotheSNPsitewereran- domlychosen[21].Therefore,ninepossiblemismatches couldbegeneratedamongthethreebasesclosestto 3 ’ endofprimer(Figure1b). Inthisstudy,1686 B.oleracea SNPsincluding12kinds ofSNPswerechosenforprimerdesign(Additionalfile1). Amongthem,someofthoseSNPscouldformthesame 3 ’ endmismatchesin01-88withallele-specificprimers. Therefore,allSNPscouldbeclassifiedintoeightkindsin- cludingA/GandC/T,T/CandG/A,A/CandG/T,T/G andC/A,A/T,T/A,C/G,G/C(Table2)basedonmis- matchtypes(forexample,forSNPsA/GandC/T,the3 ’ endmismatchoftheallele-specificprimerdesignedfor 02-12isTGin01-88).Moreover,forconvenienceofdata analysis,wefurthercompressedintofourSNPtypesin- cludingA/G(T/C),A/C(T/G),A/T(T/A)andG/C(C/G) basedontheirdestabilizationeffectsinmismatchbase pairs(Table2)[28,29]. Inthisstudy,alltheprimerswereusedtoamplifythe genomeDNAoflines01-88and02-12,andPCRproducts weredetectedon2.5%agarosegelbyelectrophoresis.The Figure1 SchematicrepresentationoftheAS-PCRprimerdesign.a ,PrimerP1formsaperfectmatchwithallelefrom02-12,butamismatch basepairatthe3 ’ endwiththeDNAsequenceofallelefrom01-88.Itcouldamplifythebandinbothoftwolines01-88and02-12.PrimerP2 formstwomismatchbasepairswithallelefrom01-88atthe3 ’ endandinthe3 rd nucleotidefromthe3 ’ end,whileamismatchbasepairinthe 3 rd nucleotidewithallelefrom02-12.Itamplifiedthebandonlyin02-12. b ,SchematicrepresentationofdifferentmismatchesduringtheSNP primerdesign. Liu etal.PlantMethods 2012, 8 :34 Page3of9 http://www.plantmethods.com/content/8/1/34 resultsshowedthatamongthe1686primerpairs,490pairs(29.1%)displayedpolymorphism(Table3).WhenclassifiedbySNPtypes,thepercentsofpolymorphismofA/G(C/T),A/C(G/T),A/T(T/A),C/G(G/C)were26.4%(111/420),31.3%(131/418),34%(148/435)and37.5%(155/413)respectively.Whenclassifiedbymismatchsites,polymorphismpercentsofthe2and4baseloca-tionclosesttotheSNPsiteswere25.1%(139/554),32.9%(189/575)and29.1%(162/557),respectively.IfSNPtypeandmismatchsiteinprimerswerebothconsidered,thepolymorphismpercentofSNPA/T(T/A)whichhadmis-matchesinthe3sitewas45.9%(72/157),followedbyC/G(G/C)inthe2sitewith43.3%(58/134)and3sitewith39.4%(56/142).ThepolymorphismpercentofSNPA/G(C/T)was26.4%(111/420),whichalsohasthelowestpolymorphismsinallmismatchsites.Moreover,thepolymorphismpercentsofSNPprimersclassifiedbythetypesofmismatchbasesinthe2and4mismatchsiteswereanalyzed(Table4).Resultsshowedthatpolymorphismpercentsof8mismatchtypeswerecloselyequivalentinthe2sites(22.5%-26.4%).Highpolymorphismpercentsappearedintwomismatchbasepairsinthe3baselocation(CA,46.8%,37/79;TG,38.6%,27/70).Atthe4baseawayfromtheSNPsite,thepolymorphismpercentsofprimerswithmismatchbasepairsGA,TC,TTandCCwereover30%(31.7%-33.3%).BasedontheresultsinTable3andTable4,wein-ferredthatprimers,whichhadCAmismatchinthe3baseclosesttoSNPsiteA/T(T/A),hadthehighestpolymorphismbetweenlines01-88and02-12.Toiden-tifythisconjecture,385primersdesignedfromSNPA/T(T/A),whichcontainedGorTbaseinthe3nucleo-tide,werefurtherchosenforpolymorphismanalysis(Additionalfile2).Thedetectionresultsshowedthat295SNPprimerswerepolymorphic(295/385,76.6%)be-tweenlines01-88and02-12.WhentheseprimerswereclassifiedbasedontheSNPs(A/TorT/A)in3end,thespecificitypercentofSNP(A/T)(158/193,81.9%)washigherthanthatofSNP(T/A)(137/192,71.4%)(Figure2aandFigure2b).ApplicationofthedesignmethodBesidesthehighestpolymorphicprimersmentionedabove,thehighpolymorphicprimerscouldbefoundineverySNPtypebasedonourresults.Theprincipleofpri-merdesignisdescribedasfollowed:firstly,foreverykindofSNP,themismatchsite(the2and4siteclosesttotheSNPsite)ischosenaccordingtotheresultofpoly-morphismpercentinTable3.FortheprimersofthreeSNPtypesA/T(T/A),A/C(G/T)andA/G(C/T),thehighestpolymorphismpercentsare45.9%,37.4%and30.7%,respectivelyandthemismatchesinprimersarealllocatedinthe3site.WhilefortheSNPtypeC/G(G/C),themismatchesinthe2siteclosestto3endofprimersshowhighestpolymorphismpercent(43.3%).Forthebase Table2DestabilizationstrengthofeightcombinationsofmismatchnucleotidepairingSNPtypesmismatchtypesatendofprimerinDestabilizationstrengthofmismatchtype(01-88/02-12)A/GTGStrongC/TGTG/AACT/CCAA/CTCWeakG/TCTC/AAGT/GGAA/TTTStrongandMediumT/AAAG/CCCStrongandMediumC/GGG Table3EffectofmismatchsitesandSNPtypesonthespecificityofallele-specificPCRsitesclosesttothe3ofprimersPolymorphismpercentofSNPPolymorphismpercentofprimersforeverySNPtypeA/G,C/TA/C,G/TA/T,T/AC/G,G/CThe2base25.1%23.0%26.8%26.6%43.3%(139/554)(32/139)(37/138)(38/143)(58/134)The3base32.9%30.7%37.4%45.9%39.4%(189/575)(42/137)(52/139)(72/157)(56/142)The4base29.1%25.7%29.8%28.1%29.9%(162/557)(37/144)(42/141)(38/135)(41/137)Totalnumber29.1%26.4%31.3%34%37.5%(490/1686)(111/420)(131/418)(148/435)(155/413)etal.PlantMethodsPage4of9http://www.plantmethods.com/content/8/1/34 inmismatchsite(the2 nd ,3 rd and4 th siteclosesttothe SNPsite),thereexistthreedifferentmismatchstylesfor everykindofbase.Thebestmismatchstyleischosen accordingtothestatisticalresultsfromTable4.Forex- ample,ifthereisaGinthe3 rd site,themismatchesof CC,CAandCTwillbeformed.Thepolymorphismper- centofprimerswithCA(46.8%)mismatchishighestcom- paredtothatofCC(23.9%)andCT(32.1%). Wefurtheridentifiedtheusabilityoftheprimerdesign methodonrapeseedandsesame.Duringrapeseedoilcon- tentresearch,withDHlinesofF1generationbetween zy036and51070,whichhadbeenreportedbyHua etal [30],weidentifiedaQTLrelatedtooilcontentlocatedon A2chromosome.BecauseoflackofmarkersintheQTL interval,wedesignedatotalof20SNPprimersbetween markerO110C05andmarkerBrSF000036-9usingour Table4Effectofartificialbasemismatchesinthreemismatchsitesonthespecificityofallele-specificPCR Mismatchsites closesttothe3 ’ end ofprimers Mismatch types Polymorphism percentof primers Mismatch types Polymorphism percentof primers Mismatch types Polymorphism percentof primers Mismatch types Polymorphism percentof primers The2 nd baseTC26.2%GA22.5%CA25.7%AA23.8% (17/65)(16/71)(19/74)(15/63) TG26.4%GG24.2%CC26.1%TT25.6% (19/72)(15/62)(18/69)(20/78) The3 rd baseTC32.1%GA28.3%CA46.8%AA30.6% (26/81)(17/60)(37/79)(26/85) TG38.6%GG30.4%CC23.9%TT29.7% (27/70)(21/69)(16/67)(19/64) The4 th baseTC32.9%GA31.7%CA28.0%AA26.8% (24/73)(26/82)(21/75)(19/71) TG23.3%GG25.8%CC33.3%TT32.4% (14/60)(16/62)(22/66)(22/68) Figure2 AnalysisofspecificityforSNPprimersof B.oleracea .a ,1-24primerpairs,whichcorrespondedtoBo001-Bo024,wereintroduced withaCAbasepairmismatchinthe3 rd nucleotideclosestto3 ’ end(A/TSNPtype,AAmismatch). b ,24primerpairs,whichcorrespondedto Bo194-Bo217,wereintroducedwithaCAbasepairmismatchinthe3 rd nucleotideclosestto3 ’ end(A/TSNPtypes,TTmismatch).A,01-88;B,02-12. Liu etal.PlantMethods 2012, 8 :34 Page5of9 http://www.plantmethods.com/content/8/1/34 method(Table5).Resultsshowedthat15primerswerelocatedintheQTLinterval(Figure3).Thepolymorphismpercentofprimersis(15/20,75%).Another24primerscamefromA/TSNPsbetweentwosesamelines28-31andZZM2289werealsodesigned.Andamongthem,18SNPmarkersweredetectedtohavepolymorphism(18/24,75%)(Additionalfile3).BothresultsfromrapeseedandsesameindicatedthattheSNPprimermethodcouldbeusedinotherspecies.DiscussionInrecentyears,variousmethodsforhigh-throughputSNPanalysishavebeendescribed[28].AlthoughthesemethodsarehighlyefficientcomparedtoothertraditionalSNPgenotypingbyelectrophoresis,significantinvestmentsofexpensiveprobes,microchipsorspecialinstrumentationhavelimitedtheiruseinmostlaboratories.Fortraditionallow-throughputSNPgenotypingmethods,themaintimeandlabor,andlowefficiencyofspecificprimerarestillchallenges[27].TheAllele-specificPCRmethodwasdevelopedforalleleanalysisofclinicallysignificantmuta-tions.Tofacilitatereliablediscriminationbetweentwoalleleshighly,theadditionofartificialmismatcheswithinthethreebasesfrom3endoftheprimersmightbebenefi-cial[23].Althoughthethirdpositionfromthe3endhasbeendetectedasthebesttoplaceamismatchbaseinprimerpreviously[24],wereallydonotknowwhichkindofmismatch(foreverybase,therearethreekindsofmis-matches)isthebestchoiceinthe3position.Inthisstudy,differentfromthepreviousreportswhichusedonlyalimitedquantityofprimers,alargeamountofSNPpri-mersdesignedbyintroducingmismatcheswithinthethreebasesclosesttothe3endofprimerswereusedtosolvethisproblem.Generally,AS-PCRprimersdesignedrandomlyhadalowallelicspecificityrateofapproximately30%,whichwasconsistentwithourresults(29.1%).However,mis-matchsites(2,and4siteclosesttothe3end)haddifferenteffectsonthepolymorphicefficiencyofprimers.Inourstudy,wefoundprimerpolymorphicpercentwaslowestinthe2baselocationbecausemanyprimerscouldnotamplifyanybandsinbothofthelines01-88and02-12.Forthe4baselocation,thepolymorphismeffi-cienciesofallmismatchtypeswerealmostequivalent(under30%).Thehighestpolymorphicpercentwasfoundinthe3baselocatedclosesttotheSNPsite,whichwasobservedbyHayashietalsimilarly(2004)[24].Accordingtotheresultsofthermodynamicsofmis-matchesreportedbyPeyretetal(1999)andLittle(2001)[28,29],themismatchbasepairshaddifferentdestabilizationeffectsthatcouldbedividedintoweak,medium,andstrongstrengthofdestabilization.Therefore,duringdesignofAS- Table520primerspairsdesignedaccordingtoSNPsbetweenrapeseedzy036and51070MismatchsiteclosesttoendofbasepairsForwardprimer(5)Reverseprimer(5Br94494C/G2CTAGTTACATAGGTCCACAATCATAGAATAAACTTTTCTACCATTCGGAGCCTAAATAGAGGTAAAAGGTGBr27005T/A3CAAAGATTGTTTCAAACGCAAAAATATACAACAAAATCAACGAATTTCACACTTTAGTAATGCACTGAGATTTBr31054T/A3CACAAGTGAGACTGAATCCACAATAAAGGATGCTACTTCGGATAAAATCCCCAGCTCTACTATACATTCCBr43193T/A3CACCTTTTATTTGATCACAGGGGTTTGTAGGAACTTCAGCCAGTAACGTCCCCCACATCBr51190T/A3CAAAATAATGGCATGCTCCTCTTTAATCTACCAAACTTATTCGGTTCCGAAAATAATGCGATGCBr34590T/A3CAAAGTGACGGTTCTTTAAGTTATCAGAGTCTCCTAATAGATTTGGGATTAAAATCAAGTTGTGGGTTAGTTTTBr03637T/A3CAATTACAGAATGTGTGTGCAAACAGAAATACATTACTTGTGTCCCCATTTCGTGTAATCATAAAGCTAGBr39807T/A3CACGAGACTCGGGTCGTTGAGTGGAAATATCCTAAAGACTTCTCCCACAAATCCACCATBr68275T/A3CATGCCGCATGTATGTCGGAGATGATAATAACCGAAACCCTAGTAGGCTAGGCGCBr29253T/A3CATGGCGCTAAATCCAAGAAGAAGTCCATTAATTACCACCTTTCTTACCCTTGTTACTCATGACAGBr61715T/A3CATGAATAGATTCTTCCGCATCACCTTTTAAAGTTAATCTTGTTTCAAGAGAAATTGAACAAGCTGCAGTBr24494C/G2GTGTCAATAATACTAGCAAACATACAACAGCGAGATTCTTGCAAATTTTAGTCAAAGTCGGTAGAAAATAGATCBr06710T/A3CATCTTGTCGATGCTGAGCTGGCAAATACTGGTCAAGCTCACACACACTCCACGTCBr00855T/A3CACACTATGGGCTATGGTGGGTCCTTCAATTGATTGGAGTTCTGTGCTCGTAGTTTTGCBr71197C/G2GTTATGGCACACAGACAGAGTTCCAGGAAATCTCTTCCAGTTCGATATCTTGGTCTGTCCCBr07809C/G2GTCTCCGCCCACATGTTATAATATGTCAGTATATCTTCTAGTGAATGGAGAAAGAGAACAAAGCCTACAGTACABr02456T/A3CAGCCTTCAGAAGGTCTGGAAACTGGATTGTTCGATGGACTTCACTACCTCCCATAGCTBr77080T/A3CACGGATAGTTTCGGGTTCGGTTCGATTACCGAACGGGTACCCGAATATATAAAAATATTAATTBr77646T/A3CATCCACCAGAATTGTGTGATGGCACTTACTGAAAAACGTCAGGTCAATGTATCAACTTCGATAABr14184G/A3ACGATTAACCGATGAAAGTCTCAGTGCCACAGTCGTCTGTGACTCCCAAACACTTGGATAGetal.PlantMethodsPage6of9http://www.plantmethods.com/content/8/1/34 PCRprimers,theeffectsofmismatchesnomatterin3 ’ end orwithinthethreebasesclosesttothe3 ’ endofprimers shouldbebothconsidered[26,2 7].Inthisstudy,forconveni- entanalysis,wecompressedfourSNPtypesincludingA/T (T/A),A/G(T/C),A/C(T/G),andG/C(C/G)basedontheir destabilizationeffectsofmismatchbasepairsformedin 3 ’ endofprimers.Amongthem,primersgeneratedfromSNP typesA/G(T/C)hadthelowestdetectionefficiencyinall mismatchsites.ItwasreasonablebecauseACandGTmis- matcheshadweakdestabilizationstrength.Theprimersin- cludingthesespecificmismatchesat3 ’ endwereeasierto makeamplificationinbothalleles. Similarly,mismatchtypeswithinthethreebasesclosest tothe3 ’ endaffectspecificitiesofprimers.Inthe3 rd base, CAandTG(thehighestpolymorphicmismatches) belongedtoweakdestabilizationstrengthmismatches. ThemismatchesGA,TC,TT,andCC(thehigherpoly- morphicmismatches)locatedatthe4 th baseawayfrom theSNPsitebelongedtothestrongdestabilization strengthmismatches.Fromtheresults,wededucedthat SNPs(A/T),whichcontainedCAmismatchesinthe3 rd nucleotidefromthe3 ’ endoftheprimers,hadthehighest allele-specification.Accordingtothecombinationrules, polymorphicefficiencybetweenTT(mismatchin3 ’ endof primer,strongdestabilizationstrength)andCA(weak destabilizationstrength)aretypicallyhigherthanAA (mismatchin3 ’ endofprimer,mediumdestabilization strength)andCA.Ourresultsconfirmedthisdeduction. Basedontheseresults,weperformedtheprimerdesign principlewhichcouldformtheoneandonlybestprimer foreverySNPtype.Amongthem,mismatchesinthesec- ondpositionsweremoreappropriateforSNPtype(C/G andG/C),whichwasdifferentfromtheviewpointthat mismatchinthe3 rd positionwasthebestchoiceforAS- PCR.Withtheprimerdesignprinciple,wefurthertested theprimersdesignedbasedonSNPsofrapeseedandses- ame.Highefficientpolymorphismoftheprimersidenti- fiedtheusabilityofthemethodinotherspecies. Conclusion ASNPprimerdesignmethodwasdevelopedwhich improvedthepolymorphismefficiencyofAS-PCRpri- mershighly.Themodifiedprimerdesigncanhelpto identifythebesteffectiveprimerforeachSNPandpo- tentiallyisavaluabletoolforgenemapping,map-based cloningandmarker-assistedselectionincrops. Methods PlantmaterialsandSNPinformation Atleast20uggenomeDNAof B.oleracea lines01-88 and02-12ataconcentrationof
50ng/ul,wassentfor Solexasequencingasacommercialservice.TheDNA wasfragmentedintosmallpiecesusingdivalentcations atelevatedtemperature.ThecleavedshortDNAfrag- mentswerepreparedforSolexasequencinginBGI (China).REPEATMASTERwasusedforscreening repeatedsequenceswithdefaultparameterandlabeling thesequencesfromdifferentmaterials.Forgenomeloca- tionoffragments,SOAPadaptingthedefaultparameter valueswasusedfortheinitialalignmentandscreening Figure3 SNPmarkerdensityincreaseofrapeseedlinkagemap.a ,AnoilcontentQTLwasscannedbetweenO110C05andBrSF000036-9. b ,Amongtwentyprimers,fifteennewmakersweregeneratedforfurtherQTLlocation. Liu etal.PlantMethods 2012, 8 :34 Page7of9 http://www.plantmethods.com/content/8/1/34 toavoidtheeffectsofparalog[31].SNPprimerdesignwasperformedusingscreenedresults.SNPanalysisandverificationbySangersequencingToverifytheputativeSNPs,96SNPsitesderivedfromonly8-readsequenceswererandomlychosenbetweenB.olera-cealines01-88and02-12.Primers(Sangon,China)weredesignedtoamplifyabout500bpfragmentsinwhichcon-tainingthecorrespondingSNPs.ThePCRreactioncon-tained25ngDNA,0.2mMdNTP,0.5UTaq(MBI,USA)with1xbuffer,and5pMofeachprimer.PCRparameterswereasfollows:apre-denaturationof94°Cfor2min,35cyclesofamplification(94°Cfor30S,60°Cfor1minand72°Cfor1min)andafinalextensionreactionwasper-formedat72°Cfor5min.PCRproductsweredetectedon1.0%agarosegelbyelectrophoresisandligatedintoPMD18T-vector(Takara,Japan)forSNPidentification.PrimerdesignandtestingAllele-specificprimerscorrespondingto12kindsofSNPB.oleraceaweredesignedaccordingtodifferentcombi-nationsbetweenmismatchbaseandmismatchsite.Optimizationofmeltingtemperature,primerlengthandamplifiedproductslengthwereachievedusingprimerprogramWebSNAPER(http://pga.mgh.harvard.edu/cgi-bin/snap3/websnaper3.cgi)[23].PrimersequenceswerescreenedagainstB.oleraceagenomerepetitivesequencestominimizemis-priming.PolymorphismassayofSNPprimerswereperformedbyPCRanddetectedbyagarosegelelectrophoresis.Alltheforwardprimersareallele-specificforB.oleracealine02-12andthereverseprimerisnotallele-specific.AmplificationofSNPprimerswasperformedonC1000TMThermalCycler(Bio-Rad,USA)using20ulreactions.Beforecarryingoutthisstudy,wehadchosensomeTaqpolymerases:MBITaqDNAPolymeraseandTakaraTaq(twogeneralTaq),MBIDreamTaqPolymeraseandTakaraExTaq(whicharebetterinamp-lificationefficiencyandsensitivitycomparedtogeneralTaqs)toidentifytheireffectsonamplificationefficiency.Resultshowedbothofgeneralpolymerases(MBITaqDNAPolymeraseandTakaraTaq)hadsameandhighallele-specificamplificationefficiencycomparedtotheothertwoTaqs.Therefore,generalTaqpolymerasewouldbebestchoiceinallele-specificPCRandTaqDNAPolymerasefromMBIwaschoseinthisstudy.ThePCRreactioncontained25ngDNA,0.2mMdNTP,0.5UTaqDNAPolymerasewith1xbuffer,and5pMofeachprimer.PCRparameterswereasfollows:apre-denaturationof94°Cfor2min,35cyclesofamplifica-tion(94°Cfor30s,55°C-65°Cfor1minand72°Cfor30s)andafinalextensionreactionwasperformedat72°Cfor10min.PCRproductswereseparatedon2.5%agarosegelbyelectrophoresis.ApplicationinrapeseedandsesameRapeseedDNAsamplesincludingtwoparents(highoilcontentlinezy036andlowoilcontentline51070)andDHlines,whichhadbeenreportedbyHuaetalalwerepreparedusingtheDNAeasyplantkitminiprep(Qiagen,Valencia,CA).Zy036and51070werere-sequencedandblastedwithB.napusgenomesequence(unpublished).Additionally,twosesamelines28-31andZZM2289(genomesequencehasnotbeenpublished)werealsousedinourresearch.AllSNPswerechosenaccordingtothemethoddescribedinB.oleracea01-88and02-12.TheSNPprimersweredesignedaccordingtoourprimerdesignmethod.AdditionalfilesAdditionalfile1:1686primerpairsdesignedbasedontheSNPsB.oleracealines01-88and02-12.Additionalfile2:385primerpairsdesignedbasedontheSNPsB.oleracealines01-88and02-12whichareA/TSNPtypewithGandTbaseinthe3nucleotide.Additionalfile3:24primerpairsdesignedbasedontheSNPsbetweensesamelines28-31andZZM2289.SNP:SingleNucleotidePolymorphism;AS-PCR:Allele-SpecificPCR;SSR:SimpleSequenceRepeat;RFLP:Restrictionfragmentlengthpolymorphism;AFLP:Amplifiedfragmentlengthpolymorphism;CAPs:TheCleavedAmplifiedPolymorphicSequence;dCAPs:DerivedTheCleavedAmplifiedPolymorphicSequence.CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.JLandSMHcontributedtoprimerdesignandtooktheco-leadroleinwritingthemanuscript.SMHandMYSperformedtheprimeranalysis.SYLofferedsequencedataofB.oleraceaandB.napus.YMLandWWXprovidedtheDNAofB.oleracea01-88and02-12.XRZprovidedthegenomesequenceandDNAofS.indicumL.28-31andZZM2289.HZWparticipatedindiscussionsduringexperimentalwork.WHconceivedtheprojectandapprovedthefinalversionofthemanuscript.Allauthorsreadandapprovedthefinalmanuscript.AcknowledgementThisstudywassupportedbytheNationalKeyBasicResearchProgramofChina(2011CB109300),National863plansprojects(2012AA101107),andKeyProjectsintheNationalScience&TechnologyPillarProgram(2010BAD01B02).AuthordetailsKeyLaboratoryofBiologyandGeneticImprovementofOilCrops,MinistryofAgriculture,OilCropsResearchInstituteoftheChineseAcademyofAgriculturalSciences,Wuhan430062,PeoplesRepublicofChina.InstituteofVegetablesandFlowersoftheChineseAcademyofAgriculturalSciences,Beijing100081,PeoplesRepublicofChina.Received:15June2012Accepted:17August2012Published:24August20121.GutIG:Automationingenotypingofsinglenucleotidepolymorphisms.HumMutat2.KwokPY:Methodsforgenotypingsinglenucleotidepolymorphisms.AnnuRevGenomicsHumGenetetal.PlantMethodsPage8of9http://www.plantmethods.com/content/8/1/34 3.Flint-GarciaSA,ThornsberryJM,BucklerES: Structureoflinkage disequilibriuminplants. AnnuRevPlantBiol 2003, 54: 357 – 374. 4.WangC,LiuZ: Arabidopsisribonucleotidereductasesarecriticalforcell cycleprogression,DNAdamagerepair,andplantdevelopment. PlantCell 2006, 18: 350 – 365. 5.FeltusFA,WanJ,SchulzeSR,EstillJC,JiangN,PatersonAH: AnSNP resourceforricegeneticsandbreedingbasedonsubspeciesindicaand japonicagenomealignments. GenomeRes 2004, 14: 1812 – 1819. 6.HillierLW,MillerRD,BairdSE,ChinwallaA,FultonLA,KoboldtDC, WaterstonRH: ComparisonofC.elegansandC.briggsaegenome sequencesrevealsextensiveconservationofchromosomeorganization andsynteny. PLoSBiol 2007, 5: e167. 7.EberleMA,NgPC,KuhnK,ZhouL,PeifferDA,GalverL,Viaud-Martinez KA,LawleyCT,GundersonKL,ShenR,MurraySS: Powertodetectrisk allelesusinggenome-widetagSNPpanels. PLoSGenet 2007, 3: 1827 – 1837. 8.ShendureJ,MitraRD,VarmaC,ChurchGM: Advancedsequencing technologies:methodsandgoals. NatRevGenet 2004, 5: 335 – 344. 9.BarbazukWB,EmrichSJ,ChenHD,LiL,SchnablePS: SNPdiscoveryvia454 transcriptomesequencing. PlantJ 2007, 51: 910 – 918. 10.TrickM,LongY,MengJ,BancroftI: Singlenucleotidepolymorphism(SNP) discoveryinthepolyploidBrassicanapususingSolexatranscriptome sequencing. PlantBiotechnolJ 2009, 7: 334 – 346. 11.LiR,LiY,FangX,YangH,WangJ,KristiansenK,WangJ: SNPdetectionfor massivelyparallelwhole-genomeresequencing. GenomeRes 2009, 19: 1124 – 1132. 12.LivakKJ,MarmaroJ,ToddJA: Towardsfullyautomatedgenome-wide polymorphismscreening. NatGenet 1995, 9: 341 – 342. 13.MyakishevMV,KhripinY,HuS,HamerDH: High-throughputSNP genotypingbyallele-specificPCRwithuniversalenergy-transfer-labeled primers. GenomeRes 2001, 11: 163 – 169. 14.HuangJ,WeiW,ZhangJ,LiuG,BignellGR,StrattonMR,FutrealPA, WoosterR,JonesKW,ShaperoMH: WholegenomeDNAcopynumber changesidentifiedbyhighdensityoligonucleotidearrays. HumGenomics 2004, 1: 287 – 299. 15.ShaperoMH,ZhangJ,LoraineA,LiuW,DiX,LiuG,JonesKW: MARA:a novelapproachforhighlymultiplexedlocus-specificSNPgenotyping usinghigh-densityDNAoligonucleotidearrays. NucleicAcidsRes 2004, 32: e181. 16.MatsuzakiH,DongS,LoiH,DiX,LiuG,HubbellE,LawJ,BerntsenT, ChadhaM,HuiH,YangG,KennedyGC,WebsterTA,CawleyS,WalshPS, JonesKW,FodorSP,MeiR: Genotypingover100,000SNPsonapairof oligonucleotidearrays. NatMethods 2004, 1: 109 – 111. 17.ShenR,FanJB,CampbellD,ChangW,ChenJ,DoucetD,YeakleyJ,Bibikova M,WickhamGarciaE,McBrideC,SteemersF,GarciaF,KermaniBG, GundersonK,OliphantA: High-throughputSNPgenotypingonuniversal beadarrays. MutatRes 2005, 573: 70 – 82. 18.KoniecznyA,AusubelFM: AprocedureformappingArabidopsis mutationsusingco-dominantecotype-specificPCR-basedmarkers. Plant J 1993, 4: 403 – 410. 19.NeffMM,TurkE,KalishmanM: Web-basedprimerdesignforsingle nucleotidepolymorphismanalysis. TrendsGenet 2002, 18: 613 – 615. 20.ThielT,KotaR,GrosseI,SteinN,GranerA: SNP2CAPS:aSNPandINDEL analysistoolforCAPSmarkerdevelopment. NucleicAcidsRes 2004, 32: e5. 21.ChaRS,ZarblH,KeohavongP,ThillyWG: Mismatchamplificationmutation assay(MAMA):applicationtothec-H-rasgene. PCRMethodsAppl 1992, 2: 14 – 20. 22.KwokS,ChangSY,SninskyJJ,WangA: Aguidetothedesignanduseof mismatchedanddegenerateprimers. PCRMethodsAppl 1994, 3: S39 – S47. 23.DrenkardE,RichterBG,RozenS,StutiusLM,AngellNA,MindrinosM,Cho RJ,OefnerPJ,DavisRW,AusubelFM: Asimpleprocedurefortheanalysis ofsinglenucleotidepolymorphismsfacilitatesmap-basedcloningin Arabidopsis. PlantPhysiol 2000, 124: 1483 – 1492. 24.HayashiK,HashimotoN,DaigenM,AshikawaI: DevelopmentofPCR-based SNPmarkersforriceblastresistancegenesatthe Piz locus. TheorAppl Genet 2004, 108: 1212 – 1220. 25.HirotsuN,MurakamiN,KashiwagiT,UjiieK,IshimaruK: Protocol:asimple gel-freemethodforSNPgenotypingusingallele-specificprimersinrice andotherplantspecies. PlantMethods 2010, 6: 12. 26.WangkumhangP,ChaichoompuK,NgamphiwC,RuangritU,ChanprasertJ, AssawamakinA,TongsimaS: WASP:aWeb-basedAllele-SpecificPCR assaydesigningtoolfordetectingSNPsandmutations. BMCGenomics 2007, 8: 275. 27.BuiM,LiuZ: Simpleallele-discriminatingPCRforcost-effectiveandrapid genotypingandmapping. PlantMethods 2009, 5: 1. 28.TsuchihashiZ,DracopoliNC: ProgressinhighthroughputSNP genotypingmethods. PharmacogenomicsJ 2002, 2: 103 – 110. 29.LittleS: Amplification-refractorymutationsystem(ARMS)analysisof pointmutations. CurrProtocHumGenet 2001, 9: 9.8.1 – 9.8.12. 30.HuaW,LiRJ,ZhanGM,LiuJ,LiJ,WangXF,LiuGH,WangHZ: Maternal controlofseedoilcontentin Brassicanapus :theroleofsiliquewall photosynthesis. PlantJ 2012, 69: 432 – 444. 31.LiR,LiY,KristiansenK,WangJ: SOAP:shortoligonucleotidealignment program. Bioinformatics 2008, 24: 713 – 714. doi:10.1186/1746-4811-8-34 Citethisarticleas: Liu etal. : Animprovedallele-specificPCRprimer designmethodforSNPmarkeranalysisanditsapplication. Plant Methods 2012 8 :34. Submit your next manuscript to BioMed Central and take full advantage of: € Convenient online submission € Thorough peer review € No space constraints or color “gure charges € Immediate publication on acceptance € Inclusion in PubMed, CAS, Scopus and Google Scholar € Research which is freely available for redistribution Submit your manuscript at www.biomedcentral.com/submit Liu etal.PlantMethods 2012, 8 :34 Page9of9 http://www.plantmethods.com/content/8/1/34