RESEARCH Open Access HTLVinfected CD Tcells display alternative exon usages that culminate - PDF document

RESEARCHOpenAccess HTLV-1-infectedCD4+T-cellsdisplayalternative exonusagesthatculminateinadultT-cell leukemia MorganThénoz 1 ,CélineVernin 1 ,HusseinMortada 2 ,MarounKaram 1 ,ChristianePinatel 2 ,AntoineGessain 3 , ThomasRWebb 4 ,DidierAuboeuf 2 ,EricWattel 1,5,6* andFranckMortreux 1,6* Abstract Background: ReprogrammingcellulargenetranscriptionsustainsHTLV-1viralpersistencethatultimatelyleadsto thedevelopmentofadultT-cellleukemia/lymphoma(AT LL).Wehypothesizedthatbesidesthesequantitative transcriptionaleffects,HTLV-1qualitativelymo difiesthepatternofcellulargeneexpression. Results: Exonexpressionanalysisshowsthatpatients ’ untransformedandmalignantHTLV-1 + CD4 + T-cellsexhibit multiplealternateexonusage(AEU)events.Theseaffectei thertranscriptionallymodifiedorunmodifiedgenes, culminateinATLL,andunveilnewfuncti onalpathwaysinvolvedincancerandc ellcycle.Unsupervisedhierarchical clusteringofarraydatapermittedtoisolateexonexpressi onpatternsof3977exonsthat discriminateuninfected, infected,andtransformedCD4 + T-cells.Furthermore,untransformedinfectedCD4+clonesandATLLsamples shared486exonmodificationsdistributedin320 genes,therebyindicatingaroleofAEUsinHTLV-1 leukemogenesis.Exposingcellstosplicingmodulatorsre vealedthatSudemycinEredu cescellviabilityofHTLV-1 transformedcellswithoutaffectingprimarycontrolCD4+ cellsandHTLV-1negativecelllines,suggestingthatthe ewstargetsfortreatingATLL. Conclusions: Takentogether,thesedatarevealthatHTLV-1sig nificantlymodifiesthestructureofcellular transcriptsandunmasknewputativeleukemogeni cpathwaysandpossibletherapeutictargets. Keywords: HTLV-1,AdultT-cellleukemia,Alternativesplicing Background HTLV-1causeschronicinfectionthatreliesonthe persistentclonalexpansionofinfectedCD4 + andCD8 + T-cells.Reprogrammingofge netranscriptionsustains HTLV-1viralpersistenceandultimatelyleadstothe developmentofadultT-cellleukemia/lymphoma (ATLL)fromaCD4 + -infectedcloneinaminorityof carriersafteraprolongedlatency[1].Transcriptome analysesusingmicroarraytechnologyhasprovidedkey insightsintothebiologicalprocessesinvolvedinthe pre-malignantandmalignantexpansionofinfectedcells [2-14].HTLV-1hasbeenfoundtoderegulatetheexpres- sionofnumerousgenesinvolvedinkeycellularpathways suchasthecellcycle,apoptosis,telomeres/telomerase, DNArepair,andimmuneandinflammatoryresponses throughexpressionofthe Tax oncoprotein[2-6]. Recentworkshavehighlightedthatinadditionto theirquantitativeeffectsongeneexpression,numerous pathogenicprocesses,suchas persistentviralinfections [15]ortumordevelopment[16,17],relyonacquired ruses,alternativesplicingplaysapivotalroleinHTLV-1 expression.Fromits5 ’ LTR,HTLV-1transcribesasingle polycistronicpre-mRNAthatcodesforstructuraland enzymaticproteinsrequiredforviralparticleproduction. Thispre-mRNAalsoundergoesmultiplealternativespli- cingeventsthatgeneratemono-splicedtranscriptscoding *Correspondence: eric.wattel@ens-lyon.fr ; franck.mortreux@ens-lyon.fr 1 UniversitédeLyon1,CNRSUMR5239,OncovirologieetBiothérapies, LaboratoiredeBiologieMoléculairedelaCellule,FacultédeMédecineLyon Sud,PierreBénite,France 6 OncovirologieetBiotherapies,LaboratoiredeBiologieMoléculairedela Cellule,UMR5239CNRS/ENS,Lyon/UCBL/HCL;Ecolenormalesupérieurede Lyon,46,alléed ’ Italie;69364,Lyoncedex07,France Fulllistofauthorinformationisavailableattheendofthearticle ©2014Thénozetal.;licenseeBioMedCentral.ThisisanOpenAccessarticledistributedunderthetermsoftheCreative CommonsAttributionLicense(http://creativecommons.org/licenses/by/4.0),whichpermitsunrestricteduse,distribution,and reproductioninanymedium,providedtheoriginalworkisproperlycredited.TheCreativeCommonsPublicDomain Dedicationwaiver(http://creativecommons.org/publicdomain/zero/1.0/)appliestothedatamadeavailableinthisarticle, unlessotherwisestated. Thénoz etal.Retrovirology (2014) 11:119 DOI10.1186/s12977-014-0119-3 fortheregulatoryproteinsp21Rex,p12andp13,anddouble-splicedtranscriptscodingforTax,p27Rexandp30[18,19].Similarly,minus-strandtranscriptioninitiatedfromthe3LTRgeneratessplicedandunsplicedRNAisoformsofHBZthatsynthesizeHBZproteinswithdistinctpropertiesoncellproliferation[20-22].HowHTLV-1intervenesinal-ternativesplicingprocessesisstillincompletelyunderstood.IthasbeenreportedthattheRNA-bindingproteinRexregulatesviralsplicingthroughinteractingwithhostspli-cingmachinerytoinhibitviralRNAsplicingandexportunsplicedandsingle-splicedtranscripts[23-25].Gene-by-geneanalyseshavefurthershownthatHTLV-1mayaffectalternativesplicingofcellulargenesincludingCD44[26]andIL-6-andIL-2-receptors[27,28],proposingthatHTLV-1-inducedAEUsmightcontributetomolecularmechanismsthatunderlietheclonalexpansionandthema-lignanttransformationofinfectedcells.However,nosys-tematicstudyhasbeenhithertoconductedtoascertaintheextentofalternativesplicingmodificationsuponHTLV-1infection.Here,byusingintegrativeanalysisofexonexpres-sionprofilesandgeneontologyofCD4T-cellsderivedfrominfectedindividualswithandwithoutmalignancy,weshowthatHTLV-1inducesmultipleAEUalterationsthatunmasknewputativeleukemogenicpathwaysandpossibletherapeutictargets.ResultsanddiscussionComparativemicroarrayanalysisofexonexpressionpro-fileswasperformedwiththreeATLLsamplesand12untransformedCD4T-cellclones(sixinfected)derivedfromHTLV-1-infectedindividualswithnoclinicalsignsofmalignancy.ATLLsampleswereobtainedfrompatientswithanacuteformofATLL�(95%circulatingmalignantcells).CD4cloneswereobtainedthroughcloningbylimit-ingdilutionofperipheralbloodmononuclearcells(PBMCs)derivedfromthreeHTLV-1-infectedindividualswithtrop-icalspasticparaparesis/HTLV-1-associatedmyelopathywithadiseasedurationof6,11,and�26years.ThisstudywasconductedaccordingtotheprinciplesoutlinedintheDec-larationofHelsinki,andapprovedbytheInstitutionalRe-viewBoardoftheHospicesCivilsdeLyon(France).Aspreviouslydescribed[29,30],CD4clonesweresubmittedtoexvivocultureforonemonthpriortoHTLV-1screeningandRNAextraction.Atthistime,infectedanduninfectedcellsremainednon-immortalizedandrequiredIL-2forcon-tinuedgrowth.GiventhatinvitroT-cellactivationisknowntomodifyAEU[31,32],exonarrayanalysisofuntrans-formedCD4cloneswascarriedoutwithRNAsextractedfromunstimulatedandphytohemagglutinin(PHA)-stimu-latedCD4T-cells(Additionalfile1:TableS1).Bythisapproachthattakeintoaccounttheinvitrocellculture,weassumedthatsignificantchangesinexonexpressionbe-tweeninfectedanduninfectedclonesmainlyresultedfromtheglobalimpactofHTLV-1infection,irrespectiveofactivationstatus.Themicroarraydatahavebeendepos-itedintotointotheGeneExpressionOmnibusdatabaseandareavailableunderrecordnumberGSE52244.HTLV-1expressionwasassessedbyquantitativeRT-PCR(qRT-PCR)ofTaxtranscripts.Inaccordancewithpreviousreports,Taxtranscriptswereweaklydetectedin2outof3ATLLsamples(Additionalfile2:FigureS1)[33-35].Incomparison,therewasawiderangeofTaxbetweenclones(0.16-2.20;median,1.2;mean±standarddeviation[SD],1.172±0.799).Fromthesedata,weas-sumedthatHTLV-1samplesexhibitedtypicalfeaturesofuntransformed-andtransformed-infectedCD4T-cellsinvivo.Computationalanalysesof15exonarraysandannotationofexoneventswereachievedasdetailedinFigure1.Figure1Arepresentsthedistributionofquantita-tiveandqualitativegenemodificationsinuntransformedHTLV-1-infectedCD4clonesandATLLcells.Atotalof18/20(90%)array-predictedexonusageswerevalidatedbyexon-specificRT-PCR(Figures1CandD).Amongthose,wefoundalternativesplicingchangesinandCD44transcripts(Figure1C).ThesedataconfirmedandextendedpreviousobservationsindicatingthatIL2-RandCD44vari-antisoformsaredetectedinthePBMCofinfectedindivid-uals[26,28].Array-predictedchangesintranscriptabundancewerevalidatedbyqRT-PCRfor8/9(88%)genes(Additionalfile3:FigureS2).ToexaminetheglobalimpactofHTLV-1onAEUofT-cells,weinvestigatedAEUinuntransformedT-cellsbycomparingthepatternofexonexpres-sionininfectedversusuninfectedclones(Figure1A;Additionalfile1:TableS1).Overall,2201AEUeventswereidentifiedin1271genes,ofwhichonly656(51%)werefoundalteredatthelevelofwholegeneexpression.Thisdemonstratedthatinuntransformedcells,HTLV-1significantlymodifiestheexoncontentofmultipletran-scripts,consideringhalfofthesegenesweretranscrip-tionallyunmodified.Secondly,whenATLLsampleswerecomparedwithuninfectedclones,theoverallnumberofAEUeventswasfoundtobe11-fold(23504versus2201)higherthanthatdistinguishingHTLV-1-positivefrom-negativeuntransformedCD4T-cells(Figure1A).Inadditiontothisquantitativedifference,theproportionofAEUeventsfoundindifferentiallyexpressedgeneswassignificantlyhigherinATLLsamples(72%versus51%inuninfectedclones;p0.0001,PearsonsChi-squaredtest;Figure1A).Togethertheseresultsindicatethatthedis-tributionofAEUalterationswassignificantlydifferentbetweenATLLandHTLV-1-positiveuntransformedcells.ThecomparativeanalysisofAEUbetweenthetwocat-egoriesofinfectedcellsrevealedthatinfectedCD4clonesandATLLsamplesshared486AEUsin329genes(p=1.83×10,Hypergeometrictest),suggestingthat22%ofAEUsarisingatthechronicstageofinfectionmightpertaintoATLLdevelopment.ThenumberofAEUsetal.Retrovirology (2014) 11:119 Page2of9 pergenerangedfromoneto20(median,1;mean±SD, 1.47±1.42;Additionalfile4:TableS2).GOenrichment analysis(DAVIDresources,http://david.abcc.ncifcrf.gov/) performedwiththissetofgene sidentifiedseveralenriched termsthatfitwellwithHTLV-1infectionandrelated diseases,suchaspathwaysincancer(foldenrichment [FE]:2.71),endocytosis(F E:4.45),antigenprocessing andpresentation(FE:7.23),celladhesionmolecules (CAMs;FE:4.56),andnaturalkillercell-mediatedcyto- toxicity(FE:4.46;Figure1B;Additionalfile5:TableS3). Figure1 DistributionofAEUinATLLcellsandclonedCD4 + T-cellsderivedfromHTLV-1-infectedindividuals.(A) Distributionofquantitative andqualitativeHTLV-1modificationsinuntransformedHTLV-1-positiveCD4 + clonesandATLLsamples. TheVenn-diagramsshowthedistributionofgenes modifiedatthewholegeneexpressionlevel(white),AEU(darkgrey),orboth(lightgrey)inATLLsamples(bottom)anduntransformedinfectedclones (top;sixindependentdatasetsincludingthreePHA-activatedandthree unactivatedclones)versusuninfectedclones(sixindependentdata setsincludingthreePHA-activatedandthreeunactivatedclones).Grap hspresentthedistributionofeachclassofAEUannotatedaccordingto theFASTERDBdatabase[41].Acceptor:alternativeacceptorsite;deletio n:exondeletion;donor:alternativedonorsite;exon:exonskipping; intron-ret:intronretention;polya: alternativepolyadenylationsite;prom:alternativep romoter.Thetotalnumberofeachclassofexoneventis indicatedonthehistogrambar. (B) GOanalysisofgenes(n=329)presentingcommonAEUsinuntransformed-infectedclonesandinATLLsamples (comparedwithuninfectedclones). Thecompletesetofgenesfeaturedinmicroarrays(42304EnsemblgeneIDs)wasusedasareferencebackground. GeneannotationsarepresentedinAdditionalfile5:TableS3. (C) RT-PCRvalidationofmicroarray-predictedexoneventsinCD4 + T-cellsderived fromHTLV-1carrierscomparedwithuninfectedT-cellclones. Exon-specificRT-PCRwasperformedwithpooledRNAsamplesderivedfromsixinfected (I)andsixuninfected(UI)clones,therebyreflectingthedistributionofsplicevariantsinallclonesirrespectiveofactivationstatus.Numbers indicatethe expectedband(bp)ofPCRproducts,andtheSIandp-valueareindicatedforeachexonevent;aSI
1.2wasconsideredtobeasignificantchangein exonexpressionandusedforcomparisons. (D) RT-PCRvalidationofmicroarray-predictedexoneventsinATLLcells(array)andexon-specificRT-PCRofsix additionalATLLsamplescomparedtouninfectedT-cellclones . Thénoz etal.Retrovirology (2014) 11:119 Page3of9 Hierarchicalclusteringofexonarraydatashowed thatalternativeexpressionof3977exons(2218genes) permittedsamplegatheringinthreedistinctclusters, includinguntransformed-uninfectedandinfectedCD4 + clones,aswellasATLLsamples(p0.05,Kruskal-Wallis ANOVA;Figure2A).Foruntransformedcells,theseexon expressionprofileswereirrespectiveofPHAstimulation, implyingadirectroleofHTLV-1intheserecurrent alternativesplicingmodific ations.ATLLcellsexhibited markedchangesinAEUcomparedtountransformed HTLV-1-positiveor-negativeCD4 + clones.Suchdistri- butionsuggeststhatsomeAEUeventscouldrepresent anATLLmolecularsignature.Therefore,weusedexon- specificRT-PCRtoanalyzesixadditionalsampleshar- vestedfrompatientswithacuteformsofATLL.Over the11eventsidentifiedinFigure1D,theexpressionof 4AEUswasfoundtoberecurrentinATLL(Figure1D). WhethertheAEUsshowninFigures1B,D,and2Aare involvedinATLLdevelopmentandpossessdiagnosticand/ orprognosticimplicationsdeservesfurtherinvestigation. AGOanalysiswasperformedtogaininsightintothe functionalsignificanceoftheexonexpressionprofiles (3977exons)thatdiscriminatebetweenthethreegroups ofsamplespresentedinFigure2A.Significantenrich- mentwasdetectedforgenesinvolvedinimportant biologicalprocesses,su chasthoserelatedtocancer (50genes,FE=1.9;p=6.6×10  6 ),MAPKsignaling (46genes,FE:2.2;p=6.1×10  7 ),focaladhesion(34 genes,FE:2.2;p=5.5×10  5 ),cytokine-cytokinere- ceptorinteractions(34genes,FE:1.7;p=0.003), andendocytosis(32genes,FE:2.2;p=3.2×10  5 ; Figure2B).Officialsymbolannotationofgeneswith AEUsthatareenrichedinthesepathwaysarepre- sentedin(Additionalfile6:TableS4).Secondly,GO analysisofAEUeventsdetectedinATLLsamples (7616genes)anduntransformedHTLV-1-positiveCD4 + Figure2 Exon-basedhierarchicalclustering. HierarchicalclusteringanalysiswasperformedwithMev4.0software(http://www.tm4.org/)using thegene-normalizedexonintensities. (A) Theselectionofexonsusefulfordifferentiatingbetweenthe15cellsampleswasstatisticallyanalyzed usingaKruskal-WallisANOVAwithp0.05rep resentingstatisticalsignificance. (B) GOanalysis .Toppathwaysarepresented.Thegeneset correspondedto2000genesthatdisplayedthehighestsplicingindex(S I)values.Thecompletesetofgenesfeaturedinmicroarrays(42304 EnsemblgeneIDs)wasusedasareferencebackground.Officialge nesymbolannotationsarepresentedinAdditionalfile5:TableS3. Thénoz etal.Retrovirology (2014) 11:119 Page4of9 T-cellclones(1271genes)showeddistinctbiologicalchanges(Additionalfile7:FigureS3).Forexample,genesencodingspliceosomecomponentsappearedexclusivelyenrichedinATLLcells,whilecellcycle-relatedgeneswerefoundenrichedinuntransformedandtransformedHTLV-1positivesamples.Moreover,AEUontologicalenrichmentsweredifferentfromthoseofdifferentiallyexpressedgenes(Additionalfile7:FigureS3).Together,theseresultssuggestthatHTLV-1-relatedtranscriptionalandpost-transcriptionalchangesmightposedistinctfunctionalimpactsonuntransformedandmalignantcells.Theplethoraofsplicingeventsfoundintransformedanduntransformed-infectedcellssuggeststhatspliceo-somecomponentsmightconstitutenewtherapeutictargetsinthemanagementofHTLV-1infection.Thus,weexaminedwhetherthespliceosomemodulatorsSudemycinEandD1[36]exertedanti-proliferativeactivityagainstHTLV-1T-celllines.Tothisend,cellviabilityupondrugexposurewascomparedbetweenHTLV-1-transformedcelllines(MT2andHUT-102)anduninfectedcells,includingprimaryCD4T-cellsandtheHTLV-1-negativeacutelymphoblasticleukemiacellline,MOLT4.Figure3AshowsthatincreasingtheconcentrationofSudemycinD1ledtoarapiddecreaseintheviabilityofhumancelllineswhethertheycarriedHTLV-1provirusesornot,whilegrowthofprimaryT-cellswasinhibitedathigherdrugconcentra-tions.Incontrast,asimilarSudemycinEconcentrationrangewasfoundtosignificantlydecreasetheviabilityofHTLV-1celllines,whiledisplayingweakeffectsonprimaryCD4andMOLT4cells.HighconcentrationofSudemycinE�(5M)affectedcellviabilityofalltestedcells,carryingornotHTLV-1(notshown).Thesedatasuggestthat,comparedtobothtransformedanduntransformed-uninfectedcells,HTLV-1tumoralcellsarehighlysensitivetodrugsthattargetsplicingma-chinery.Time-courseanalysisconfirmedtheseresultsandfurtherrevealedthattheanti-proliferativeeffectofSudemycinEwaspersistentinMT2andHUT-102HTLV-1-positivecellsrelativetoHTLV-1-negativeCD4andMOLT4cells(Figure3B).FurtheranalysesexaminedtheimpactofSudemycinEontheexpressionofviralandcellulargenes.Figure3CshowsthatSudemycinEdose-dependentlyinducedtheexpressionofshortspliceoformsofMDM2andcaspases2and9thathavebeenpreviouslycorrelatedwithapop-tosisinduction[37,38].Incontrast,theexpressionofubiquitin,resultingfromconstitutivesplicingofintron3,remainedunaffected.ThisindicatesthatSudemycinEonlyaffectedalternativesplicing,ratherthanwholegenetranscription,ofapoptoticregulatorygenes.Inparallel,asTaxissynthetizedbydouble-splicedviraltranscripts,weexaminedwhetherSudemycinEcouldaffectTaxproteinlevels.AsshowninFigure3D,westernblotanalysisofMT2andHuT102celllinesrevealedthatincreasingtheconcentrationofSudemycinEledtoreducedTaxexpression.BecauseATLLcellsbarelyexpressTaxbutregularlyshowhighlevelsofHBZmRNA,wenextaimedattestingtheeffectsofSudemy-cinEonHBZexpression.IncontrasttoTax,HBZexpressionisnotstrictlydependentonsplicingevents.Infact,HTLV-1minusstrandencodesspliced(sHBZ)andunsplicedisoforms(unHBZ)ofHBZ,yetsHBZisthemostabundantinHTLV-1transformedcelllinesandinfreshATLLsamples[20,39].Incontrasttoitsunsplicedcounterpart,sHBZpromotescellproliferationinvitrotherebylikelyexplainingthepositivecorrelationbetweensHBZexpressionandHTLV-1proviralloadsinvivo[21,39].Thesedatapromptedustoassesstheex-pressionlevelsofbothsHBZandusHBZRNAisoformsinSudemycinEexposedcellsusingexon-specificquan-titativeRT-PCRanalysis.TheresultsillustratedintheFigure3EshowthatbothMT2andHUT102cellsexhibitedreducedratiosofsHBZ/unHBZuponSude-mycinEtogetherwithadramaticdecreaseintheoveralllevelofunHBZexpression,whichisconsistentwiththetransactivatingfunctionsofsHBZonHBZgeneexpres-sion[40].GiventhecrucialroleofbothTaxandsHBZinconferringproliferativeandanti-apoptoticproperties,thesedatahelpexplaintheheightenedsensitivityofHTLV-1-infectedcellstotheanti-proliferativepropertiesofSudemycinEcomparedtotheiruninfectedcounter-parts.Takentogether,ourresultsindicatethatbyimpair-ingsplicingand,inturn,geneexpressionofHTLV-1oncogenes,splicingmodulatorssuchasSudemycinEmightconstitutepromisingtoolsforATLLtreatment.ConclusionsOurinvestigationrevealedthatHTLV-1significantlyqualitativelymodifiesgenetranscripts,aswellasun-masksnewputativeleukemogenicpathwaysandpos-sibletherapeutictargets.BecausealternativesplicingcanresultinmRNAisoformscodingforproteinswithdifferentbiologicalactivity,ourresultssuggestthatexcessivesplicingalterationsinHTLV-1-positivecellsmightunderlietheirphenotypicplasticity.Futurestudiesareneededtoaddressthemolecularmechanismsunder-lyingHTLV-1-inducedAEUandtheirinvolvementinclonalpersistence,immuneescape,and/orcellulartrans-formationofinfectedCD4cells,aswellastheirputativeroleintreatmentresistanceandrelapseofATLL.TcelllimitingcloningPeripheralbloodmononuclearcells(PBMCs)wereob-tainedbyFicollseparationofwholebloodofHTLV-1in-fectedindividuals.ForT-celllimitingdilutioncloning,etal.Retrovirology (2014) 11:119 Page5of9 Figure3 Anti-proliferativeeffectofSudemycinEonHTLV-1-transformedcelllines.(A) CellviabilitywasdeterminedbyMTTassay(CellTiter96® Non-RadioactiveCellProliferationAssay,Promega)72hafterSudemycin(D1orE)exposure.Datarepresentnormalizedmean±SD;*p0.05was consideredstatisticallysignificantusingaMann – WhitneyU-test. (B) Time-courseanalysisover96husing1
MofSudemycin. Experimentswere performedtwice,intriplicateforeachcelltype,withDMSOusedascontrol.Datarepresentthenormalizedmean±SD. (C) SudemycinE-induced expressionofshortspliceoformsencodingcaspases2and9andMDM2. Exon-specificRT-PCRwascarriedout72hafterSudemycinEexposure. Theoligonucleotidesusedhavebeenpreviouslydescribed[36]. (D) SudemycinEreducesTaxexpression. Westernblotanalysisof Tax expression wascarriedoutwithamousemonoclonalanti-Taxantibody(clone474)72hafterSudemycinEexposure.Histogramsrepresentthequantificationof Taxsignalsnormalizedagainstbeta-actin. (E) SudemycinEdecreasesgeneexpressionlevelofHBZandtherelativeratiosHBZ/unHBZ. Isoform-specific qRT-PCRanalysiswascarriedoutafter72hofSudemycinexposure.HBZisoformexpressionswerenormalizedto hprt-1 transcriptsusedas internalcontrol. Thénoz etal.Retrovirology (2014) 11:119 Page6of9 PBMCswereseededat0.1cell/wellinTerasakiplatesafterremovalofadherentcells.T-lymphocyteswereculturedinRPMI1640containingpenicillinandstrep-tomycin,sodiumpyruvate,non-essentialaminoacids,2-mercaptoethanol,10%filteredhumanABserum,100U/mLrecombinantIL-2(ChironCorporation),andMHTLV-1integraseinhibitorL-731,988[30].T-lymphocyteswerere-stimulatedevery14dayswithPHA(1g/mL)andfreshfeedercells(5.10cells/mL)thatwerecomposedoflethallyirradiatedPBMCsfromthreedistinctallogenicHTLV-1-negativedonors.CloneswerephenotypedbyflowcytometryusingantibodiesagainstCD4andCD8(DakoCytomation)andisotype-matchedcontrolsonaFACScansystemusingCellQuestsoftware(BectonDickinson).HTLV-1-positivecloneswereassessedbyPCRusinglong-terminalrepeatre-specificprimersaspreviouslydescribed[29].Foreachclone,RNAextractionwasperformedbeforeand24hafterPHAstimulation.MicroarrayanalysisAffymetrixexonarrayhybridizationwascompletedbylabeling1gofTRIzol-purifiedtotalRNAwithAffyme-trixreagentsaccordingtothemanufacturersinstructions.Hybridizationcocktailscontaining55.5gcDNAwerepreparedandhybridizedtoAffymetrix-GeneChipHumanExon1.0STarrays(Affymetrix).AffymetrixExpressionConsoleSoftwarewasusedforqualityassessment,whileexonarraydatawerenormalizedusingquintilenormalizationandanalyzedusingFasterDBannotation(https://fasterdb.lyon.unicancer.fr/)[41].Backgroundcorrectionandprobeselectionwereperformedasde-scribedpreviously[42].StatisticalanalyseswereperformedusingaStudent-testonthesplicingindex(SI)thatcorrespondstocomparisonofgene-normalizedexonintensityvaluesbetweentwoexperimentalconditions[30,42].Accordingtoexon-specificRT-PCR(Figures1CandD),microarraydatawereconsideredstatisticallysignificantforp0.05andSI1.2.PolymerasechainreactionPCRwasperformedusingtheHerculaseIIFusionDNAPolymerase(AgilentTechnologies)in3035cycles.PCRproductswereanalyzedbyagarosegelelectrophoresiswithethidiumbromideunderUVlight.Primersweredesignedtoencompassalternativesplicingevents.Primersequencesareavailableuponrequest.QuantitativeRT-PCRExpressionofTaxmRNAwasquantifiedwithRotorGene(Qiagen)aspreviouslydescribed[29,43].ExpressionofsHBZandunHBZwasquantifiedusingprimersprevi-ouslydescribedbyMurataetal.[44].qRT-PCRwasper-formedintriplicateandrelativequantitation(RQ)wascalculatedbythe2ddCTmethodtonormalizegeneexpres-siontotheendogenouscontrol(ENSG00000206625)orhprt1(ENSG00000165704).DrugexposureandcellviabilityCellviabilitywasdeterminedbyMTTassay(CellTiter96®Non-RadioactiveCellProliferationAssay,Promega)72hafterSudemycin(D1orE)exposure.TheMTTdyewasaddedduringthelasthourofincubation.After,supernatantswereremovedandcellswereincubatedwith100L/wellSolubilizationSolution/StopMixfor1hatroomtemperature.TheODwasmeasuredusinganELISAreaderat570nm,with650nmasareference.Experimentswerecompletedtwice,intriplicateforeachcelltype,withDMSOusedascontrol.AdditionalfilesAdditionalfile1:TableS1.Listofsamples.Additionalfile2:FigureS1.TaxexpressionininfectedclonesandATLLsamples.TaxmRNAwasquantifiedbyreal-timeqRT-PCRasdescribedpreviously[29,43].*StatisticallysignificantdifferencesbyMannWhitneyU-test.Additionalfile3:FigureS2.qRT-PCRconfirmationofoverallgeneexpression.cDNAwasamplifiedbyqRT-PCRin20lreactionsusingaQuantiFastSYBRGreenPCRKit(Qiagen)with10Mofeachprimer.ReactionswererunonaRotor-Gene3000(CorbettLifeScience,Australia).HRPT1geneexpressionwasusedasaninternalcontrol.Ameltingcurve(5795°C)wasgeneratedattheendofeachruntoverifyprimerspecificity.ThePfafflmethodwasusedforrelativequantification.Primersequencesareavailableuponrequest.FC,indicatesmicroarray-predictedpositiveornegativefold-changesingeneexpressionbetweeninfectedanduninfectedCD4cells;CD4UI,uninfectedCD4clones;CD4I,infectedCD4clones.Additionalfile4:TableS2.CommonAEUsbetweenuntransformed-infectedCD4clonesandATLLsamples(comparedwithuninfectedAdditionalfile5:TableS3.GOtermsenrichedbygeneswithcommonAEUsinuntransformed-infectedclonesandATLLsamples(comparedwithuninfectedclones).Additionalfile6:TableS4.GOtermsenrichedbygenesdisplayingAEUsthatdiscriminatebetweenuninfectedandinfectedclonesandATLLAdditionalfile7:FigureS3.GOanalysisinATLLsamplesanduntransformedclonesaccordingtowholegeneexpressionandAEU.TheVenn-diagramsshowthedistributionofgenesmodifiedatthewholegeneexpressionlevel(white;FC1.2;p0.05),AEUlevel(darkgrey;SI1.2;p0.05),orboth(lightgrey).EachgenesetwasanalyzedusingtheDAVIDbrowser(http://david.abcc.ncifcrf.gov/;KEGGpathways).Eachgenesetwaslimitedto2000genesthatdisplayedthehighestfold-changeinexpressionandsplicingindex(SI)values.Thecompletesetofgenesfeaturedinmicroarrayswasusedasreferencebackground.Toppathwaysarepresented.CompetinginterestsTheauthorsdeclarethattheyhavenocompetinginterests.MT,CV,HM,MK,andCPperformedexperimentsandanalyzeddata.AGprovidedpatientsamplesandanalyzeddata.TWprovidedSudemycinEandanalyzeddata.EWandDAanalyzeddataandparticipatedinwritingthemanuscript.FMdesignedthestudy,analyzeddata,andwrotethemanuscript.Allauthorsreadandapprovedthefinalmanuscript.etal.Retrovirology (2014) 11:119 Page7of9 Acknowledgments TheauthorswouldliketothankJoelLachuerandNicolasNazaret (ProfilExpert,Lyon,France)forprovidingtranscriptomicfacilities.Thiswork wassupportedbytheLigueNationaleContreleCancer(Comitésdel ’ Ain,de laDrome,delaSaoneetLoire,andduRhône),FondationdeFrance, AssociationLauretteFugain,AssociationpourlaRecherchesurleCancer (ARC),AssociationGuillaumeEspoir,andtheAgenceNationalepourla Recherche(EPIVIR).M.T.andC.V.weresupportedbybursariesfromthe FrenchMinistryofHigherEducationandScience.F.M.wassupportedby INSERMandbytheHospicesCivilsdeLyon(AVIESANCHRTprogram2010). E.W.wassupportedbyHospicesCivilsdeLyonandLyonIUniversity (France).T.W.wassupportedinpartbyaUSNIHGrant(CA140474),the AmericanLebanese-SyrianAssociatedCharities(USA),andSt.JudeChildren ’ s ResearchHospital(USA). Authordetails 1 UniversitédeLyon1,CNRSUMR5239,OncovirologieetBiothérapies, LaboratoiredeBiologieMoléculairedelaCellule,FacultédeMédecineLyon Sud,PierreBénite,France. 2 CentredeRecherchesurleCancerdeLyon, FranceEpissagealternatifetprogressiontumorale,Lyon,France. 3 Institut Pasteur,Unitéd ’ EpidémiologieetPhysiopathologiedesVirusOncogènes, Paris,France. 4 SRIInternational,333RavenswoodAvenue,MenloPark,CA 94025-3493,USA. 5 UniversitéLyonI,Serviced ’ Hématologie,PavillonMarcel Bérard,CentreHospitalierLyon-Sud,PierreBénite,France. 6 Oncovirologieet Biotherapies,LaboratoiredeBiologieMoléculairedelaCellule,UMR5239 CNRS/ENS,Lyon/UCBL/HCL;EcolenormalesupérieuredeLyon,46,allée d ’ Italie;69364,Lyoncedex07,France. Received:20August2014Accepted:2December2014 References 1.MortreuxF,GabetAS,WattelE: MolecularandcellularaspectsofHTLV-1 associatedleukemogenesisinvivo. Leukemia 2003, 17: 26 – 38. 2.NgPW,IhaH,IwanagaY,BittnerM,ChenY,JiangY,GoodenG,TrentJM, MeltzerP,JeangKT,ZeichnerSL: Genome-wideexpressionchanges inducedbyHTLV-1Tax:evidenceforMLK-3mixedlineagekinase involvementinTax-mediatedNF-kappaBactivation. Oncogene 2001, 20: 4484 – 4496. 3.Pise-MasisonCA,RadonovichM,MahieuxR,ChatterjeeP,WhitefordC, DuvallJ,GuillermC,GessainA,BradyJN: Transcriptionprofileofcells infectedwithhumanT-cellleukemiavirustypeIcomparedwith activatedlymphocytes. CancerRes 2002, 62: 3562 – 3571. 4.TaylorJM,GhorbelS,NicotC: Genomewideanalysisofhumangenes transcriptionallyandpost-transcriptionallyregulatedbytheHTLV-I proteinp30. BMCGenomics 2009, 10: 311. 5.TsukasakiK,TanosakiS,DeVosS,HofmannWK,WachsmanW,GombartAF, KrebsJ,JauchA,BartramCR,NagaiK,TomonagaM,SaidJW,KoefflerHP: Identifyingprogression-associatedgenesinadultT-cellleukemia/ lymphomabyusingoligonucleotidemicroarrays. IntJCancer 2004, 109: 875 – 881. 6.ZaneL,SibonD,LegrasC,LachuerJ,WierinckxA,MehlenP,Delfau-Larue MH,GessainA,GoutO,PinatelC,LanconA,MortreuxF,WattelE: Clonal expansionofHTLV-1positiveCD8+cellsreliesoncIAP-2butnoton c-FLIPexpression. Virology 2010, 407: 341 – 351. 7.ChevalierSA,DurandS,DasguptaA,RadonovichM,CimarelliA,BradyJN, MahieuxR,Pise-MasisonCA: ThetranscriptionprofileofTax-3ismore similartoTax-1thanTax-2:insightsintoHTLV-3potentialleukemogenic properties. PLoSOne 2012, 7: e41003. 8.deLaFuenteC,DengL,SantiagoF,ArceL,WangL,KashanchiF: GeneexpressionarrayofHTLVtype1-infectedTcells:Up-regulationof transcriptionfactorsandcellcyclegenes. AIDSResHumRetroviruses 2000, 16: 1695 – 1700. 9.HarhajEW,GoodL,XiaoG,SunSC: GeneexpressionprofilesinHTLV-I- immortalizedTcells:deregulatedexpressionofgenesinvolvedin apoptosisregulation. Oncogene 1999, 18: 1341 – 1349. 10.MokSC,BonomeT,VathipadiekalV,BellA,JohnsonME,WongKK,ParkDC, HaoK,YipDK,DonningerH,OzbunL,SamimiG,BradyJ,RandonovichM, Pise-MasisonCA,BarrettJC,WongWH,WelchWR,BerkowitzRS,BirrerMJ: Agenesignaturepredictiveforoutcomeinadvancedovariancancer identifiesasurvivalfactor:microfibril-associatedglycoprotein2. Cancer Cell 2009, 16: 521 – 532. 11.Pise-MasisonCA,RadonovichM,DohoneyK,MorrisJC,O ’ MahonyD,Lee MJ,TrepelJ,WaldmannTA,JanikJE,BradyJN: Geneexpressionprofilingof ATLpatients:compilationofdisease-relatedgenesandevidencefor TCF4involvementinBIRC5geneexpressionandcellviability. Blood 2009, 113: 4016 – 4026. 12.RuckesT,SaulD,VanSnickJ,HermineO,GrassmannR: Autocrine antiapoptoticstimulationofculturedadultT-cellleukemiacellsby overexpressionofthechemokineI-309. Blood 2001, 98: 1150 – 1159. 13.SasakiH,NishikataI,ShiragaT,AkamatsuE,FukamiT,HidakaT,KubukiY, OkayamaA,HamadaK,OkabeH,MurakamiY,TsubouchiH,MorishitaK: Overexpressionofacelladhesionmolecule,TSLC1,asapossible molecularmarkerforacute-typeadultT-cellleukemia. Blood 2005, 105: 1204 – 1213. 14.TattermuschS,SkinnerJA,ChaussabelD,BanchereauJ,BerryMP,McNab FW,O ’ GarraA,TaylorGP,BanghamCR: Systemsbiologyapproachesreveal aspecificinterferon-induciblesignatureinHTLV-1associatedmyelopathy. PLoSPathog 2012, 8: e1002480. 15.DowlingD,Nasr-EsfahaniS,TanCH,O ’ BrienK,HowardJL,JansDA,Purcell DF,StoltzfusCM,SonzaS: HIV-1infectioninduceschangesinexpression ofcellularsplicingfactorsthatregulatealternativeviralsplicingand virusproductioninmacrophages. Retrovirology 2008, 5: 18. 16.PrzychodzenB,JerezA,GuintaK,SekeresMA,PadgettR,MaciejewskiJP, MakishimaH: PatternsofmissplicingduetosomaticU2AF1mutationsin myeloidneoplasms. Blood 2013, 122: 999 – 1006. 17.GermannS,GratadouL,DutertreM,AuboeufD: Splicingprogramsand cancer. JNucleicAcids 2012, 2012: 269570. 18.CavallariI,RendeF,D ’ AgostinoDM,CiminaleV: Convergingstrategiesin expressionofhumancomplexretroviruses. Viruses 2011, 3: 1395 – 1414. 19.CiminaleV,PavlakisGN,DerseD,CunninghamCP,FelberBK: Complexsplicing inthehumanT-cellleukemiavirus(HTLV)familyofretroviruses:novel mRNAsandproteinsproducedbyHTLVtypeI. JVirol 1992, 66: 1737 – 1745. 20.UsuiT,YanagiharaK,TsukasakiK,MurataK,HasegawaH,YamadaY, KamihiraS: CharacteristicexpressionofHTLV-1basiczipperfactor(HBZ) transcriptsinHTLV-1provirus-positivecells. Retrovirology 2008, 5: 34. 21.YoshidaM,SatouY,YasunagaJ,FujisawaJ,MatsuokaM: Transcriptional controlofsplicedandunsplicedhumanT-cellleukemiavirustype1bZIP factor(HBZ)gene. JVirol 2008, 82: 9359 – 9368. 22.GaudrayG,GachonF,BasbousJ,Biard-PiechaczykM,DevauxC,Mesnard JM: ThecomplementarystrandofthehumanT-cellleukemiavirustype 1RNAgenomeencodesabZIPtranscriptionfactorthatdown-regulates viraltranscription. JVirol 2002, 76: 12813 – 12822. 23.BakkerA,LiX,RulandCT,StephensDW,BlackAC,RosenblattJD: Human T-cellleukemiavirustype2Rexinhibitspre-mRNAsplicinginvitroatan earlystageofspliceosomeformation. JVirol 1996, 70: 5511 – 5518. 24.LiM,KannianP,YinH,KesicM,GreenPL: HumanTlymphotropicvirus type1regulatoryandaccessorygenetranscriptexpressionandexport arenotrexdependent. AIDSResHumRetroviruses 2012, 28: 405 – 410. 25.NakanoK,WatanabeT: HTLV-1Rex:thecourierofviralmessagesmaking useofthehostvehicle. FrontMicrobiol 2012, 3: 330. 26.MatsuokaE,UsukuK,JonosonoM,TakenouchiN,IzumoS,OsameM: CD44splicevariantinvolvementinthechronicinflammatorydiseaseof thespinalcord:HAM/TSP. JNeuroimmunol 2000, 102: 1 – 7. 27.HoriuchiS,YamamotoN,DewanMZ,TakahashiY,YamashitaA,YoshidaT, NowellMA,RichardsPJ,JonesSA,YamamotoN: HumanT-cellleukemia virustype-ITaxinducesexpressionofinterleukin-6receptor(IL-6R): SheddingofsolubleIL-6RandactivationofSTAT3signaling. IntJCancer 2006, 119: 823 – 830. 28.HoriuchiS,KoyanagiY,TanakaY,WakiM,MatsumotoA,ZhouYW, YamamotoM,YamamotoN: Alteredinterleukin-2receptoralpha-chainis expressedinhumanT-cellleukaemiavirustype-I-infectedT-celllines andhumanperipheralbloodmononuclearcellsofadultT-cellleukaemia patientsthroughanalternativesplicingmechanism. Immunology 1997, 91: 28 – 34. 29.SibonD,GabetAS,ZandeckiM,PinatelC,TheteJ,Delfau-LarueMH, RabaaouiS,GessainA,GoutO,JacobsonS,MortreuxF,WattelE: HTLV-1 propelsuntransformedCD4lymphocytesintothecellcyclewhile protectingCD8cellsfromdeath. JClinInvest 2006, 116: 974 – 983. 30.RabaaouiS,ZouhiriF,LanconA,LehH,d ’ AngeloJ,WattelE: Inhibitorsof strandtransferthatpreventintegrationandinhibithumanT-cell Thénoz etal.Retrovirology (2014) 11:119 Page8of9 leukemiavirustype1earlyreplication.AntimicrobAgentsChemother31.IpJY,TongA,PanQ,ToppJD,BlencoweBJ,LynchKW:GlobalanalysisofalternativesplicingduringT-cellactivation.32.WhistlerT,ChiangCF,LonerganW,HollierM,UngerER:Implementationofexonarrays:alternativesplicingduringT-cellproliferationasdeterminedbywholegenomeanalysis.BMCGenomics33.KannagiM,SugamuraK,SatoH,OkochiK,UchinoH,HinumaY:EstablishmentofhumancytotoxicTcelllinesspecificforhumanadultTcellleukemiavirus-bearingcells.JImmunol34.KinoshitaT,ShimoyamaM,TobinaiK,ItoM,ItoS,IkedaS,TajimaK,ShimotohnoK,SugimuraT:DetectionofmRNAforthetax1/rex1geneofhumanT-cellleukemiavirustypeIinfreshperipheralbloodmononuclearcellsofadultT-cellleukemiapatientsandviralcarriersbyusingthepolymerasechainreaction.ProcNatlAcadSciUSA35.YoshidaM,SeikiM,YamaguchiK,TakatsukiK:MonoclonalintegrationofhumanT-cellleukemiaprovirusinallprimarytumorsofadultT-cellleukemiasuggestscausativeroleofhumanT-cellleukemiavirusintheProcNatlAcadSciUSA36.FanL,LagisettiC,EdwardsCC,WebbTR,PotterPM:Sudemycins,novelsmallmoleculeanaloguesofFR901464,inducealternativegenesplicing.ACSChemBiol37.FushimiK,RayP,KarA,WangL,SutherlandLC,WuJY:Up-regulationoftheproapoptoticcaspase2splicingisoformbyacandidatetumorsuppressor,RBM5.ProcNatlAcadSciUSA38.LogetteE,WotawaA,SolierS,DesocheL,SolaryE,CorcosL:Thehumancaspase-2gene:alternativepromoters,pre-mRNAsplicingandAUGusagedirectisoform-specificexpression.39.SatouY,YasunagaJ,YoshidaM,MatsuokaM:HTLV-IbasicleucinezipperfactorgenemRNAsupportsproliferationofadultTcellleukemiacells.ProcNatlAcadSciUSA40.GazonH,LemassonI,PolakowskiN,CesaireR,MatsuokaM,BarbeauB,MesnardJM,PeloponeseJMJr:HumanT-cellleukemiavirustype1(HTLV-1)bZIPfactorrequirescellulartranscriptionfactorJunDtoupregulateHTLV-1antisensetranscriptionfromthe3longterminalJVirol41.MallinjoudP,VilleminJP,MortadaH,PolayEspinozaM,DesmetFO,SamaanS,ChautardE,TrancheventLC,AuboeufD:Endothelial,epithelial,andfibroblastcellsexhibitspecificsplicingprogramsindependentlyoftheirtissueoforigin.GenomeRes2014,51142.DutertreM,SanchezG,DeCianMC,BarbierJ,DardenneE,GratadouL,DujardinG,LeJossic-CorcosC,CorcosL,AuboeufD:exonskippinginthegenotoxicstressresponse.NatStructMolBiol43.SatouY,YasunagaJ,ZhaoT,YoshidaM,MiyazatoP,TakaiK,ShimizuK,OhshimaK,GreenPL,OhkuraN,YamaguchiT,OnoM,SakaguchiS,MatsuokaM:HTLV-1bZIPfactorinducesT-celllymphomaandsystemicinflammationinvivo.PLoSPathog44.MurataK,HayashibaraT,SugaharaK,UemuraA,YamaguchiT,HarasawaH,HasegawaH,TsurudaK,OkazakiT,KojiT,MiyanishiT,YamadaY,KamihiraS:AnovelalternativesplicingisoformofhumanT-cellleukemiavirustype1bZIPfactor(HBZ-SI)targetsdistinctsubnuclearlocalization.JVirol and take full advantage of: € Thorough peer review No space constraints or color “gure charges Research which is freely available for redistribution www.biomedcentral.com/submit etal.Retrovirology (2014) 11:119 Page9of9