AfrequentcauseofconcernamonginvestigatorsperformingquantitativeRT-PCRisfalsepositivescausedbygenomicDNAcontaminationofRNApreparations.BecausePCRissuchasensitivetechnique,asinglecopyofagenecan,theoretically,bedetected.HerewetestvariousmethodsforremovingDNAcontamination.GenomicDNAfalsepositivesignalsareeasilyidentifiedbyperforminga"no-RT"controlduringRT-PCR.WecanthereforeassesstheeffectivenessofDNAremovalmethodsbyagarosegelanalysisofthe-RTreactions.

ThreeMethodsofRemovingContaminationfromRNA

ThedatapresentedhereresultsfromtestingseveralcommonmethodsofDNAremovalfromRNAsamples.EachDNAremovaltechniquewasanalyzedforeffectivenessbyPCRamplificationoftheRNAwithandwithoutpriorreversetranscription.Themethodsinclude:

DNasedigestionDNaseisanendonucleasethatcleavesDNAbybreakingphosphodiesterbonds.ItmustbeinactivatedorremovedfromthereactionpriortoPCR,otherwise,itmaydigestnewlyamplifiedDNA.Forthisstudy,wetested2concentrationsofDNaseI(10and50µlDNaseI/mlsample)andfourDNaseIinactivation/removalmethods:

• Chelationwith20mMEDTA
• Heatingat70°Cfor5minutes
• ProteinaseKdigestionfollowedbyphenol/chloroformextractionandNH₄OAc/EtOHprecipitation
• ProteinremovalusingAmbion"sRNAqueousÌKit

Acidphenol:chloroformextractionAcidphenol:chloroform(5:1phenol:CHCl₃;pH4.7)extractionpartitionsDNAintotheorganicphase.TheRNAremainsintheaqueousphaseandcanbesubsequentlyrecoveredbyprecipitation.

Lithiumchloride(LiCl)precipitationLiClprecipitationisaselectiveprecipitantofRNA.ItinefficientlyprecipitatesDNAwhichisdiscardedinthesupernatant.

AssessingDNAContaminationofRNASamples

ToconfirmthepresenceofcontaminatingDNA,twoRNAsampleswereassessedona1%denaturingagarosegelbyethidiumbromidestaining(Figure1a).OneRNAsampleshowedvisIBLeDNAcontamination(sampleB),whiletheotherdidnot(sampleA).ThesamesampleswerethensubjectedtoRT-PCR,alongwitha"no-RT"control,forthepresenceofribosomalproteinS15message.RegardlessofwhethercontaminatingDNAwasvisiblebygelassessment,bothRNAsamplesshowedamplifiableDNAintheno-RTcontrol(Figure1b).OnceDNAcontaminationwasverified,bothRNAsamplesweresubjectedtothethreeDNAremovalmethodsdescribedabove.

Results

DNasedigestionBothRNAsamplesA&Bweretreatedwith10and50UDNaseI(Ambion,Inc.)permlRNAsampleat37°Cfor30minutes.ThedigestedsamplesweresplitintofourtubesandeachwastreatedwithadifferentDNaseinactivationorremovalmethod.ThedatarevealthatDNasetreatmentfollowedbyanyoftheinactivationmethodswassufficienttoremovecontaminatingDNAastestedbytheno-RTcontrol(Figure2a)anddidnotinhibitsubsequentRT-PCR(Figure2b).

Acidphenol:chloroformextractionAliquotsofeachRNAsamplewereextractedwithanequalvolumeofacidphenol:chloroformfollowedbyprecipitationwith0.5MNH₄OAcandEtOH.TheRNApelletswereresUSPendedinnuclease-freewaterandthentestedinRT-PCRreactions.AswithDNasedigestion,acidphenol:chloroformextractionremovedcontaminatingDNAwithouteffectingRT-PCR.

LiClprecipitationRNAsamplesAandBwereprecipitatedwithequalvolumesof7.5MLiCl,resuspendedinnuclease-freewaterandusedinRT-PCRreactions.LiClprecipitationdidremoveDNAfromsampleAbutwasinsufficienttoremovethegreaterDNAcontaminationinRNAsampleB(Figures2aand2b).

Figure1.AssessmentofDNAContaminationinRNAPreparations.Figure1a.1µgofeachoftwoRNAsamples(labeledAandB)alongwith3µgofAmbion"sMillenniumÌMarkerswasassessedona1%denaturingagarosegelandstainedwithEtBr.NoticeinSampleB,thehighmolecularweightDNAcontamination. Figure1b.1µgofeachRNAfromFigure1awasusedinareversetranscriptionreactionperformedwiththeRETROscriptÌKit.1µlofeachRTreaction(RT-PCR)and0.5µgofeachRNA("no-RT"control)weresubsequentlyusedinastandardPCRreactionfortheamplificationoftheribosomalproteinS15message.One-tenthofeachPCRreactionwasassessedona1%agarosegelandstainedwithEtBr.

Conclusions

ThreecommonmethodsusedtoremovecontaminatingDNAfromRNApreparationsweretested.ThemethodsincludedDNaseIdigestion,acidphenol:chloroformextractionandLiClprecipitation.BothDNaseIdigestion(withsubsequentDNaseIinactivationorremoval)andacidphenol:chloroformextractionweresufficienttoremovecontaminatingDNAastestedbyno-RTcontrolPCRreactions.ThesemethodsworkedwellevenwhentheamountofDNAcontaminationwasvisiblebyEtBrstainingoftheRNAsample(sampleB).NeithermethodinhibitedsubsequentRT-PCRreactions.LiClprecipitationasamethodtoremovecontaminationDNAfromRNAsampleswasnotasefficient.LiClprecipitationwasadequatetoremovemoderatebutnotgrossDNAcontaminationfromRNAsamples.LiClprecipitationalsodidnotinhibitsubsequentRT-PCRreactions.

Figure2.AssessmentofMethodstoremoveDNAContaminationofRNA.ThetwoRNAsamplesshowninFigure1aweretreatedwithDNase,extractedwithacidphenolorprecipitatedwithLiClasamethodofDNAremoval.TheRNaseweretreatedwitheither10Uor50UDNaseI/mlRNAsample,splitintofouraliquotsandtheDNaseIwasinactivatedbyoneoffourmethods: AdditionofEDTAtoafinalconcentrationof20mM Heatingto70°Cfor5minutes Digestionwith150µgproteinaseK/mlsolutionat50°Cfor30minutes,followedbyphenol:chloroformextractionandNH4OAc/EtOHprecipitation PurificationusingAmbion"sRNAqueousÌKitaccordingtoprotocol. Samples: RNAsampleA(slightlycontaminated)treatedwith10UDNase/mlsample RNAsampleB(grosslycontaminated)treatedwith10UDNase/mlsample RNAsampleA(slightlycontaminated)treatedwith50UDNase/mlsample RNAsampleB(grosslycontaminated)treatedwith50UDNase/mlsample RNAsampleA(slightlycontaminated) RNAsampleB(grosslycontaminated) Figure2a.Approximately0.5µgofeachtreatedRNAwasusedinastandardPCRreactionfortheamplificationofribosomalproteinS15message.One-tenthofeachPCRreactionwasassessedona1%agarosegelandstainedwithEtBr. Figure2b.Approximately1µgofeachtreatedRNAwasusedinaRT-PCRreactionfortheamplificationofribosomalproteinS15messageusingAmbion"sRETROscriptÌKit.One-tenthofeachRTPCRreactionwasassessedona1%agarosegelandstainedwithEtBr.