Relative quantitation goes viral: An RT-qPCR assay for a grapevine virus

ContentslistsavailableatScienceDirect

Journal of Virological Methods

jou rn al h om ep ag e :w w w . e l s e v i e r . c o m / l o c a t e / j v i r o m e t

Relative quantitation goes viral: An RT-qPCR assay for a grapevine virus

R.Bester^a,P.T.Pepler^a,J.T.Burger^a,H.J.Maree^a,b,∗

aDepartmentofGenetics,StellenboschUniversity,PrivateBagX1,Matieland7602,SouthAfrica

bARCInfruitec-Nietvoorbij(TheFruit,VineandWineInstituteoftheAgriculturalResearchCouncil),PrivateBagX5026,Stellenbosch7599,SouthAfrica

Articlehistory:

Received2July2014 Receivedinrevisedform 16September2014 Accepted24September2014 Availableonline5October2014

Keywords:

GLRaV-3

Relativequantification RT-qPCR,SubgenomicRNA Grapevineleafrolldisease SYBRgreen

a b s t r a c t

Accuratedetectionandquantitationofvirusescanbebeneficialtoplant–virusinteractionstudies.In thisstudy,threeSYBRgreenreal-timeRT-PCRassaysweredevelopedtoquantitategrapevineleafroll- associatedvirus3(GLRaV-3)ininfectedvines.Threegenomicregions(ORF1a,coatproteinand3^UTR) weretargetedtoquantitateGLRaV-3relativetothreestablyexpressedreferencegenes(actin,GAPDH and␣-tubulin).TheseassayswereabletodetectallknownvariantgroupsofGLRaV-3,includingthe divergentgroupVI,withequalefficiency.Nolinkcouldbeestablishedbetweentheconcentrationratios ofthedifferentgenomicregionsandsubgenomicRNA(sgRNA)expression.However,asignificantlower virusconcentrationratioforplantsinfectedwithvariantgroupVIcomparedtovariantgroupIIwas observedfortheORF1a,coatproteinandthe3^UTR.Significanthigheraccumulationofthevirusinthe growthtipwasalsodetectedforbothvariantgroups.Thequantitationofviralgenomicregionsunder differentconditionscancontributetoelucidatingdiseaseaetiologyandenhanceknowledgeaboutvirus ecology.

©2014ElsevierB.V.Allrightsreserved.

1. Introduction

Acomplexnetworkofcellularprocessesregulatesgeneexpres- sion in plants to ensure normal development and appropriate responsestoenvironmentalstresses.Bioticstressesfromfungal, bacterialandviralpathogensareamajorconstrainttotheproduc- tionofagriculturalcrops.Itisthereforeimperativetounderstand plant–pathogeninteractionsbeforetranslatingthisknowledgeinto managementstrategies.Researchintoplant–pathogeninteractions canbeapproachedfromtheperspectiveofthehostplantorthe pathogen(Boydet al.,2013).Studyingthehostwilllead tothe identificationofgenesinvolvedinpartialorpermanentpathogen resistancewhilestudyingthepathogenleadstotheidentification of factorsthat could triggertheplant’s defence response.Both approachesbenefitfromaccuratedetectionandquantitationofthe pathogen.Specificallyforviruses,thequantitationofnotonlythe viralparticleswithELISA,butalsodifferentviralgeneswithRT- qPCRcancontributetoourunderstandingofthediseaseaetiology.

Reverse transcription quantitative PCR (RT-qPCR) is cur- rently one of themost sensitivetechniques for analysinggene

∗ Correspondingauthorat:DepartmentofGenetics,StellenboschUniversity,Pri- vateBagX1,Matieland7602,SouthAfrica.Tel.:+27218089579.

E-mailaddresses:hjmaree@sun.ac.za,mareeh@arc.agric.za(H.J.Maree).

expressionandhasbeenappliedforviralquantitation(Eunetal., 2000;Robertsetal.,2000).RT-qPCRassayscanutilisefluorescent dyesorprobe-basedchemistry(Bustin,2000)andquantitationwill involveeitheranabsoluteorrelativequantitationstrategy(Pfaffl, 2001).For viruses,the detectionand quantitation canbe com- plicatedbylowvirusconcentrationandthepresenceofdiverse variants.Therefore,asensitiveRT-qPCRassaythatcandetectall virusvariantswithequalefficiencycanaidresearchfocussedat plant–virusinteractions.

Grapevineisahighlyvaluableagriculturalcommoditythatis hosttothelargestnumberofvirusesofanycropplant(Martelliand Boudon-Padieu,2006;Prosseretal.,2007).Forthepurposeofthis studywefocusedonGrapevineleafroll-associatedvirus3(GLRaV- 3),believedtobethemainaetiologicalagentofGrapevineleafroll disease(GLD)(Mareeetal.,2013).GLRaV-3isthetypespeciesof thegenusAmpelovirusinthefamilyClosteroviridae(Martellietal., 2012).Currently,thecompletegenomesofonlytendistinctGLRaV- 3isolatesareavailablethatcanbedividedintofourmajorgenetic variantgroups(Engeletal.,2008;Mareeetal.,2008;Jarugulaetal., 2010;Joosteetal.,2010;Gouveiaetal.,2011;Sharmaetal.,2011;

Wangetal.,2011;Besteretal.,2012a;Seahetal.,2012;Feietal., 2013).TheisolatesofGLRaV-3are91%similarwhenvariantgroup IiscomparedtovariantgroupIIand88%whenvariantgroupIis comparedtovariantgroupIII.However,isolatesfromvariantgroup VIarelessthan70%identicaltoisolatesfromvariantgroupsI–III

http://dx.doi.org/10.1016/j.jviromet.2014.09.022 0166-0934/©2014ElsevierB.V.Allrightsreserved.

(Besteretal.,2012a).Threeadditionalvariantgroupshavealsobeen identified,butareonlyrepresentedbypartialsequences(Gouveia etal.,2011;Sharmaetal.,2011;Chooietal.,2013a).Recently,two newGLRaV-3isolates, GH24(GenBank: KM058745) andGTG10 (Goszczynski,2013),wereidentified.Theywerefoundtobemore diversecomparedtoknownvariantgroupsandhavenotyetbeen assignedtoagroup.Thisfindingfurtherhighlightsthegreatextent ofgeneticvariabilitybetweenvariantgroupsofGLRaV-3andwar- rantsthesearchforauniversaldetectionsystem.

Currently, the industry standard for GLRaV-3 detection is enzyme-linkedimmunosorbentassays (ELISA)andconventional end-pointRT-PCRs.ELISAs canbetimeconsuming andlackthe sensitivityneeded forthedetection oflow virusconcentration.

StandardisedELISAprotocolsforGLRaV-3arealsonotabletodetect allGLRaV-3variants withequal sensitivity(Cohenet al.,2012).

Improvement in thespecificity and sensitivity of detectionhas beenachievedbytheintroductionofqPCRassaysbasedonfluo- rescentdetectionsystems.BothSYBRgreenandhydrolysisprobe qPCRhavebeenusedforthediagnosisand/orquantitationofsev- eralgrapevineviruses,includingGLRaV-3(OsmanandRowhani, 2006;Osmanetal.,2007,2008;Margariaetal.,2009;Pacificoetal., 2011;Besteretal.,2012b;Tsaietal.,2012;Chooietal.,2013b;

López-Fabueletal.,2013).

InthisstudythreesensitiveSYBRgreenRT-qPCRassayswere developedthatareabletodetectallvariantgroups(groupsI–III, VI and GH24-like) of GLRaV-3 known to be present in South Africawithequalefficiency.Theseassaysenabledtheevaluation ofdifferent virusgenome regions for theirsuitabilityfor accu- ratecalculationofGLRaV-3virusconcentrationininfectedphloem material.TheRT-qPCRassaysdescribedinthisstudyprovidetools forthestudyofvirusecology.

2. Materialsandmethods

2.1. Plantmaterialandsamplepreparation

SixVitisviniferacultivarCabernetSauvignonplants,fromavirus isolatecollection(VitisLaboratory,StellenboschUniversity,South Africa),wereprunedbackandlefttogrowfor60daysinthegreen- house.Onlyoneshootwasallowedtogrowandallside shoots wereconstantlyremoved.Threeplantseach,infectedwithGLRaV- 3variantgroupIIandVI,respectivelywereselected.Plantswere negativeforfrequentlyoccurringgrapevineviruses,exceptGLRaV- 3.GLRaV-3variantgroupstatusofallplantswasconfirmedusing thepreviouslydesignedreal-timeRT-PCRhigh-resolutionmelting curveanalysisassay(Besteretal.,2012b).Theshootfromeachplant wasdividedinto5equalsegmentstorepresentdifferentgrowth stageswithsegment1representingtheolder(bottom)partofthe shootandsegment5theactivelygrowingyoungmaterialatthetop oftheplant.DuetoGLRaV-3beingaphloem-limitedvirus,phloem materialofeachsegmentwascollectedandstoredat−80^◦C.

2.2. TotalRNAextraction

TotalRNAwasextractedfrom2gofphloemmaterialusinga modifiedCTABextractionprotocol(Carraetal.,2007).TheCTAB buffercontained2%CTAB,2.5%PVP-40,100mMTris–HCL(pH8), 2MNaCl,25mMEDTA(pH8)and2%␤-mercaptoethanol.TotalRNA wasprecipitatedbyadding2.5volumes100%ethanoland0.1vol- umes3Msodiumacetate(pH5.2)totheupperphaseofthe5MNaCl andchloroform–isoamylalcohol(24:1)extractionstep(Carraetal., 2007).RNAwasprecipitatedfor1hat−20^◦Candcentrifugedat 13,500rpmfor30minat4^◦C.Pelletswerewashedwith80%ethanol andresuspendedin100␮lMilli-QH₂O(MilliporeCorporation,Bil- lerica,USA).Integrityandpuritywasassessedusingagarosegel

electrophoresisandspectrophotometry(NanoDropND-100,Nano- DropProducts,Wilmington,USA).

10␮g of total RNA wastreated withRQ1RNase-free DNase (Promega,Madison,USA)in50␮lreactionsaccordingtomanufac- turer’sinstructions.450␮lof10mMTris–HCl(pH8.5)wasadded totheDNAsetreatmentmixtureandanacidicphenol:chloroform- isoamylalcohol(5:1)extractionwasperformedwithanethanol andsodiumacetateprecipitation(2.5volumesof100%ethanoland 0.1 volumesof3MSodium acetate(pH5.2)).After awash step with80% ethanol, pelletswere driedand resuspended in 30␮l Milli-QH2O.Integrityandpuritywasassessedusingagarosegel electrophoresisandspectrophotometry.

2.3. cDNAsynthesis

Complementary DNA(cDNA) were synthesised from1␮g of totalRNAusing0.15␮grandomhexamers(Promega)andAvian myeloblastosisvirus(AMV)reversetranscriptase(ThermoScien- tific,Massachusetts,USA)inafinalvolumeof20␮laccordingto manufacturer’sinstructions.10␮lofeachcDNAsamplewaspooled anda 5-folddilutionserieswaspreparedtoconstructstandard curvesforallprimersets.TheremainingcDNAwasdiluted1:24 andtreatedastheunknownsamplesforquantitation.AllcDNA dilutionswerestoredat−20^◦C.

2.4. Primerdesign

Primersweredesignedtargetingthreedifferentregionsofthe GLRaV-3 genome. Open reading frame 1a (ORF1a), ORF6 (coat protein)andthe3^UTRwereselectedtorepresentgenes/regions with different levels of subgenomic RNAs (sgRNAs) (Jarugula et al., 2010; Maree et al., 2010). By constructing a multiple sequence alignment using CLC main workbench 6.5 (CLC bio, Aarhus,Denmark),conservedregionsinthechosengenes/regions ofGLRaV-3wereidentified.AlltheGLRaV-3complete genomes available(GenBank:AF037268.2,GenBank:JQ423939.1,GenBank:

JQ655296.1,GenBank: JQ655295.1,GenBank: EU259806.1,Gen- Bank:EU344893.1,GenBank:JX559645.1,GenBank:JQ796828.1, GenBank: GQ352633.1, GenBank: GQ352632.1, GenBank:

GQ352631.1, GenBank: GU983863.1, GenBank: KM058745) were includedin themultiple sequence alignment in order to designprimersthatwereabletodetectallvariantgroupsknown to be present in South Africa. All primers were subjected to anNCBI BLAST screen for specificity. Fivedifferent primer sets targetingVitisviniferareferencegeneswereselectedfromtheReid etal.(2006)studytoevaluatetheirexpressionstabilityacrossall samplesusedinthisstudy.

2.5. RT-qPCR

2.5.1. PCRandcycleconditions

RT-qPCRs were performed using the Rotor-Gene Q thermal cycler (Qiagen, Venlo, Netherlands) and the SensiMix^TM SYBR No-ROX Kit (Bioline, Taunton, USA). Reactions contained 2× SensiMix^TMSYBRNo-ROX,Milli-QH₂Oand0.4,0.48or0.56␮M forwardandreverseprimers(IDT,Coralville,USA),dependingon theprimerset(Table1).2.5␮lcDNAwasaddedtoeachreactiontoa finalreactionvolumeof12.5␮l.ThesamecDNAdilutionserieswas usedtoconstructalleightprimer-specificstandardcurvesandthe same1:24dilutionofeachofthe“unknown”sampleswasscreened withtheeightprimersetsforquantitation.No-templatecontrols, negativeplantcontrols(negativeforGLRaV-3)andthethirddilu- tionpoint(1/25)ofthefive-folddilutionserieswereincludedin allruns. To test for theextent of genomicDNA contamination

“no-reversetranscription”controlqPCRswereperformedforall samplesusinganintron-spanningprimersetfortheactingene.

Table1

PrimersetstargetingGLRaV-3genomicregionsofinterestandVitisviniferareferencegenes.

Primer Sequence Ampliconsize Target Primer

concentration (␮M)

Annealing temperature(^◦C)

Reference

LR36995F GGGRACGGARAAGTGTTACC 144 GLRaV-3

ORF1a

0.4 53 Thisstudy

LR37138R TCCAAYTGGGTCATRCACAA

LR314586F ATGAAYGARAARGTYATGGC 140 GLRaV-3Coat

protein(CP)

0.48 50 Thisstudy

LR314725R CTAAACGCYTGYTGYCTAG

LR318345F CCTCACGGTTTAATACTCTG 144 GLRaV-33^untranslated region(3^UTR)

0.4 54 Thisstudy

LR3 18488R ATTGTCGATAAGTTAGCCTC

Vv actin F CTTGCATCCCTCAGCACCTT 82 Vitisvinifera

actin

0.4 55 Reidetal.(2006)

VvactinR TCCTGTGGACAATGGATGGA

Vv␣-tubulinF CAGCCAGATCTTCACGAGCTT 119 Vitisvinifera

alpha-tubulin

0.4 55 Reidetal.(2006)

Vv␣-tubulinR GTTCTCGCGCATTGACCATA

VvUBCF GAGGGTCGTCAGGATTTGGA 75 Vitisvinifera

Ubiquitin-conjugatingenzyme

0.4 55 Reidetal.(2006)

VvUBCR GCCCTGCACTTACCATCTTTAAG

VvGAPDHF TTCTCGTTGAGGGCTATTCCA 70 VitisviniferaGlyceraldehyde

3-phosphatedehydrogenase

0.4 55 Reidetal.(2006)

VvGAPDHR CCACAGACTTCATCGGTGACA

VvEF1␣ F GAACTGGGTGCTTGATAGGC 150 VitisviniferaElongationfactor

1-alpha

0.56 55 Terrieretal.(2005)

Vv EF1␣ R AACCAAAATATCCGGAGTAAAAGA

AllreactionswereperformedintriplicateinQiagenRotor-GeneQ 0.1mltube-and-capstrips.Cyclingparametersincludedaninitial activationof95^◦Cfor10minand45cyclesof95^◦Cfor15s,primer- specificannealingtemperaturefor15s(Table1)and72^◦Cfor15s.

Acquisitiononthegreenchannelwasrecordedattheendofthe extensionstep.MeltingcurveanalysisofPCRampliconswasper- formedwithtemperaturesrangingfrom65^◦Cto95^◦Cwitha1^◦C increaseintemperatureevery5stoidentifyprimer-dimersand non-specificamplification.

2.5.2. Single-variantefficiency

Tobeabletocalculateaccuratelytherelativequantityofviral genomes,irrespectiveofGLRaV-3variantstatus,thePCRefficiency of different GLRaV-3 single-variant infections were compared.

GLRaV-3single-variant infected sampleswere collectedfrom a virusisolatecollection(VitisLaboratory,StellenboschUniversity, SouthAfrica),processedandtreatedasdescribedabove.Thesesam- plesrepresentedGroupsI–VI,aswellasisolateGH24.Afive-fold dilutionserieswaspreparedfromthecDNAandstandardcurves constructedforeachvariantgroupseparately.PCRefficiencieswere calculatedforallGLRaV-3primersets(Rasmussen,2001,Eq.(1)).

2.5.3. Relativelimitofdetection

ThesensitivityofeachGLRaV-3primersetwasevaluatedand comparedbyusingthefive-foldcDNAdilutionseriesusedtocon- structstandardcurvesinthisstudy.Thesamedilutionserieswas usedforallprimersetsin bothreal-timePCRand conventional end-pointPCRinstruments.Thereal-timePCRswereperformed accordingtotheconditionsdescribedabove.Intheend-pointPCRs, a2.5␮laliquotofcDNAwasaddedtoa25␮lPCRreactionmixture containing1×KAPATaqbufferA(KAPABiosystems,CapeTown, SouthAfrica),0.4mMdNTPmix(ThermoScientific),0.4–0.48␮M forwardandreverseprimers(IDT)(Table1)and0.08U/␮lKAPATaq DNApolymerase(KAPABiosystems).Cycleconditionsincludedan initialdenaturationstepat94^◦Cfor5min,followedby35cyclesof 94^◦Cfor15s,primer-specificannealingtemperaturefor15sand elongationat72^◦Cfor15s.Finalextensionwasat72^◦Cfor7min.

Ampliconswerevisualisedonanethidiumbromide-stained2%TAE agarosegel(2MTris,1Mglacialaceticacid,0.05MNa2EDTA,pH 8).

2.5.4. Referencegenestabilitytest

InordertofindreferencegenemRNAthatisstablyexpressed inphloemmaterialacrossthelengthofagrowinggrapevineshoot, fiveV.viniferareferencegeneswereselected(Table1).Thequanti- tationcycle(Cq)data(seeSection2.5.5forCqcalculation)fromall

30samples(sixplants,fivesampleseach)wereusedtocalculate stabilityofreferencegeneexpressionusingtwoExcel-basedappli- cations,Normfinder(Andersenetal.,2004)andBestKeeper(Pfaffl etal.,2004).

2.5.5. Dataanalysis

ThePCR efficiency(E)foreach ofthetargetswascalculated usingtheslopeofthestandardcurveconstructedwiththefive-fold dilutionseriesoverthelineardynamicrangeconsistentbetween the primer sets (2nd (1/5) to 6th (1/3125)dilution point) (Eq.

(1)).Therelativevirusconcentrationratio(VCR)wascalculated byanefficiencycorrectionmathematicalmodelusingreference genenormalisation(Pfaffl,2001).Theoriginal(Pfaffl,2001)model utilisesonlytheslopeandCqvariablesduetothey-intercept(b) beingequalforthecontrolandsamplegroup(Eq.(2)).However, sincenoCqvaluewillbeobtainedforsamplesnegativeforGLRaV- 3,thecontrolgroupwillbeabsentandtheinfluenceofbshouldbe considered.TheCqvaluesforallsampleswerecalculatedusingthe thresholdcyclemethod(Pfaffl,2004).Afterimportingtheprimer- specificstandardcurveineachRotor-GeneQrun,bwasadjusted usingthethirddilutionofthestandardcurve.Thisstandardwas includedineachrun,tocompensateforinter-assayvariation.The geometricmeanofthetriplicatereactionsofeachsamplewasused forallcalculations.Thethreemoststablereferencegenes(ref)were selectedfornormalisationandareferencegeneindex(RGI)wascal- culatedusingtheirgeometricmean(Eq.(3)).Normalisationofthe viralgenequantitationvalues,tocalculatetherelativeVCR,was performedusingtheRGIandnotasinglereferencegene(Eq.(4)) (Vandesompeleetal.,2002).TheRotor-geneQsoftwareversion 2.2.3(Qiagen)wasusedtocalculateprimerefficiencies,C_qvalues andgenequantitationvaluesforalltargets.TherelativeVCRswere calculatedusingEq.(4):

E=10^[−1/slope] (1)

Quantitationvalue(conc)=10

(Cq−b)

slope (2)

RGI=³

(conc_ref1)×(conc_ref2)×(conc_ref3) (3) VCR_{(multiplereferencegenes)}=conctarget

RGI (4)

2.5.6. Statisticalmodels

AnumberofstatisticalmodelswerefittedusingSASEnterprise Guideversion 5.1(SASInstitute, NC,USA).Due tothedistribu- tionofVCRsbeingpositivelyskewed,thedataweretransformed usingthenaturallogarithm(logbasee)tomeetthedistributional assumptions (normality andhomoscedasticity) ofthestatistical