A systems biology approach identified different regulatory networks targeted by KSHV miR-K12-11 in B cells and endotheli...

MISSING IMAGE

Material Information

Title:
A systems biology approach identified different regulatory networks targeted by KSHV miR-K12-11 in B cells and endothelial cells
Physical Description:
Mixed Material
Language:
English
Creator:
Yang, Yajie
Boss, Issac W.
McIntyre, Lauren M.
Renne, Rolf
Publisher:
Bi Med Central (BMC Genomics)
Publication Date:

Notes

Abstract:
Background: Kaposi’s sarcoma associated herpes virus (KSHV) is associated with tumors of endothelial and lymphoid origin. During latent infection, KSHV expresses miR-K12-11, an ortholog of the human tumor gene hsa-miR-155. Both gene products are microRNAs (miRNAs), which are important post-transcriptional regulators that contribute to tissue specific gene expression. Advances in target identification technologies and molecular interaction databases have allowed a systems biology approach to unravel the gene regulatory networks (GRNs) triggered by miR-K12-11 in endothelial and lymphoid cells. Understanding the tissue specific function of miR-K12-11 will help to elucidate underlying mechanisms of KSHV pathogenesis. Results: Ectopic expression of miR-K12-11 differentially affected gene expression in BJAB cells of lymphoid origin and TIVE cells of endothelial origin. Direct miRNA targeting accounted for a small fraction of the observed transcriptome changes: only 29 genes were identified as putative direct targets of miR-K12-11 in both cell types. However, a number of commonly affected biological pathways, such as carbohydrate metabolism and interferon response related signaling, were revealed by gene ontology analysis. Integration of transcriptome profiling, bioinformatic algorithms, and databases of protein-protein interactome from the ENCODE project identified different nodes of GRNs utilized by miR-K12-11 in a tissue-specific fashion. These effector genes, including cancer associated transcription factors and signaling proteins, amplified the regulatory potential of a single miRNA, from a small set of putative direct targets to a larger set of genes. Conclusions: This is the first comparative analysis of miRNA-K12-11’s effects in endothelial and B cells, from tissues infected with KSHV in vivo. MiR-K12-11 was able to broadly modulate gene expression in both cell types. Using a systems biology approach, we inferred that miR-K12-11 establishes its GRN by both repressing master TFs and influencing signaling pathways, to counter the host anti-viral response and to promote proliferation and survival of infected cells. The targeted GRNs are more reproducible and informative than target gene identification, and our approach can be applied to other regulatory factors of interest.
General Note:
Yang et al. BMC Genomics 2014, 15:668 http://www.biomedcentral.com/1471-2164/15/668; Pages 1-17
General Note:
doi:10.1186/1471-2164-15-668 Cite this article as: Yang et al.: A systems biology approach identified different regulatory networks targeted by KSHV miR-K12-11 in B cells and endothelial cells. BMC Genomics 2014 15:668.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
© 2014 Yang et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
System ID:
AA00026986:00001

Full Text

PAGE 1

RESEARCHARTICLEOpenAccessAsystemsbiologyapproachidentifieddifferent regulatorynetworkstargetedbyKSHVmiR-K12-11 inBcellsandendothelialcellsYajieYang,IsaacWBoss,LaurenMMcIntyre*andRolfRenne*AbstractBackground: Kaposi ’ ssarcomaassociatedherpesvirus(KSHV)isassociatedwithtumorsofendothelialandlymphoid origin.Duringlatentinfection,KSHVexpressesmiR-K12-11,anorthologofthehumantumorgenehsa-miR-155.Both geneproductsaremicroRNAs(miRNAs),whichareimportantpost-transcriptionalregulatorsthatcontributetotissue specificgeneexpression.Advancesintargetidentificationtechnologiesandmolecularinteractiondatabaseshave allowedasystemsbiologyapproachtounravelthegeneregulatorynetworks(GRNs)triggeredbymiR-K12-11in endothelialandlymphoidcells.UnderstandingthetissuespecificfunctionofmiR-K12-11willhelptoelucidate underlyingmechanismsofKSHVpathogenesis. Results: EctopicexpressionofmiR-K12-11differentiallyaffectedgeneexpressioninBJABcellsoflymphoidorigin andTIVEcellsofendothelialorigin.DirectmiRNAtargetingaccountedforasmallfractionoftheobserved transcriptomechanges:only29geneswereidentifiedaspu tativedirecttargetsofmiR-K12-11inbothcelltypes. However,anumberofcommonlyaffectedbiologicalpathw ays,suchascarbohydratemetabolismandinterferon responserelatedsignaling,wererevealedbygeneontologyanalysis.Integrationoftranscriptomeprofiling, bioinformaticalgorithms,anddatabasesofprotein-proteininteractomefromtheENCODEprojectidentifieddifferent nodesofGRNsutilizedbymiR-K12-11inatissue-specificfashion.Theseeffectorgenes,includingcancerassociated transcriptionfactorsandsignalingproteins,amplifiedtheregulatorypotentialofasinglemiRNA,fromasmallsetof putativedirecttargetstoalargersetofgenes. Conclusions: ThisisthefirstcomparativeanalysisofmiRNA-K12-11 ’ seffectsinendothelialandBcells,fromtissues infectedwithKSHV invivo .MiR-K12-11wasabletobroadlymodulategeneexpressioninbothcelltypes.Using asystemsbiologyapproach,weinferredthatmiR-K12-11establishesitsGRNbybothrepressingmasterTFsand influencingsignalingpathways,tocounterthehostanti-viralresponseandtopromoteproliferationandsurvival ofinfectedcells.ThetargetedGRNsaremorereproducibleandinformativethantargetgeneidentification,and ourapproachcanbeappliedtootherregulatoryfactorsofinterest.BackgroundKaposi ’ ssarcoma(KS)isanendothelialtumoranda majorcauseofAIDSpatientdeath.Itsassociatedherpes virus(KSHV,HHV-8)isadoublestrandDNAvirusand amemberofthe subfamilyofhumanherpesviruses [1].KSHVcanalsoinfectlymphocytes,promotingtransformationintoprimaryeffusionlymphoma(PEL)or MulticentricCastleman ’ sdisease(MCD)inimmunodeficientpatients[2,3].Thedistinctpathologicaloutcome ofKSHVintwotypesofhumantissuesservesasamodel systemforstudyingcelltypespecificgeneregulation. InKStumorsandPELs,themajorityofcellsare latentlyinfectedandexpressviralgenesonlywithina specificregionoftheviralgenome:theKSHVlatencyassociatedregion(KLAR)[4-6].Thisregionencodesthe latency-associatednuclearantigen(LANA,involvedin latentDNAreplicationandepisomalmaintenance), v-Cyclin(cyclinDhomologthatpromotesSphaseentry), v-Flip(promotescellsurvival),thekaposingenefamily(involvedincytokinemRNAstabilizationand celltransformation),and12microRNAs(miRNAs). *Correspondence: mcintyre@ufl.edu ; rrenne@ufl.edu DepartmentofMolecularGeneticsandMicrobiology,UniversityofFlorida, 2033MowryRoad,Gainesville,FL32610,USA 2014Yangetal.;licenseeBioMedCentralLtd.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.Yang etal.BMCGenomics 2014, 15 :668 http://www.biomedcentral.com/1471-2164/15/668

PAGE 2

MiRNAsaresmallRNAsof19 – 24nucleotidesthatinhibit translation[7,8]andinducedegradationofmRNAs[9-11]. ThegenomiclocationofKSHVmiRNAsandtheirabundantexpressioninKSHV-associatedtumorssuggeststhey playanimportantroleinestablishinglatencyandpromotingKSHVpathogenesis. Thefirststepindecipheringthefunctionalroleofa miRNA,istoidentifyitstargetgenes.The5 sequence (especiallybases2 – 7,termedthe  seedsequence Ž )ofa miRNA,guidesitscomplementarybindingtothe3 UTRsofitstargetmRNAsandfacilitatestherepression ofthelatterintheRNA-inducedsilencingcomplex (RISC)[12-15].Therefore,analysisofmiRNAsequence propertiescancomputationallypredictitstargets[16,17]. Duetotheshortlengthoftheseedsequenceandthe generaldisregardfortissuespecifictarget-geneexpression,bioinformaticapproachestypicallyreportlarge numbersofgenesasputativetargetsofindividualmiRNAsreviewedby[18-20].Greaterthanhalfofallprotein codinggenesinmammaliancellsareestimatedtocontainmultiplemiRNAtargetsites[21].Restrictedbytissue specificgeneexpression,onlyasmallfractionofputative targetsarepresentinaspecificcellularcontext(the direct targets)[22,23].Thedirecttargetsfrequentlydonot functioninisolationbutinteractwithothermoleculesto formgeneregulatorynetworks(GRNs).Accordingly, genesthatarepositionedatalowerlevelofthenetwork hierarchymayalsobefunctionaltargetsevenwithoutthe miRNAtargetsiteintheirsequences(theindirecttargets) (Figure1). Thisglobalregulatoryeffectcanbecapturedbygene expressionprofilingafterperturbingspecificmiRNA levels.Thedifferentiallyexpressedgenes(DEG)reflect theglobaloutcomeofthemiRNAregulation[13,24]. Apriori knowledgeofmolecularinteractionsisnecessarytoplacetheDEGsinthecontextforinterpreting thejointeffectofdirectandindirecttargetsfromanetworkperspective.Asystemsapproach,whichintegrates secondarydatawithprimarymeasurementsofgeneexpression,canconnectdifferentlayersofregulatorsfrom sparseandnoisyexpressionprofiles[25].Thisapproach isenabledbyavarietyofdatabasesonDNA-proteinand protein-proteininteractions[26-28]. KSHVmiR-K12-11providesauniquemodelforstudyingtissuespecificGRNswithregardtoviralinfection andpathogenesis.ItsseedsequenceisidenticaltocellularmiR-155.Previousstudieshaveidentifiedsimilar functionaltargetsofthetwomiRNAs[29,30].MiR-155 isawell-studied  oncomiR Ž ,beingassociatedwithimmuneactivation[31-33]andimplicatedintumorigenesis [34-38].MiR-K12-11andmiR-155showmutuallyexclusiveexpressioninKSHVinfectedtissues:miR-K12-11is abundantlyexpressedinPELcells,whilemiR-155was detectedinKSHVinfectedendothelialcells[30]. Inthisstudy,miR-K12-11wasexpressedinKSHV negativehumanendothelialandBcells,closetophysiologicallevelsobservedduringKSHVinfection.Tissue specific,aswellascommontargetgenesandpathways, wereidentifiedandtheresultswereintegratedwithtranscriptionnetworks,protein-proteininteractomeandsignalingpathways.Thissystemsapproach(Figure2)revealed thatmiR-K12-11opposeshostdefensesandcontributesto theproliferationandsurvivalofKSHVinfectedcellsby influencingkeyelementsincellularGRNslikeTFsandsignalingproteins.Ourapproachisapplicabletoabroader rangeofregulatorsofinterestforunderstandingtheGRNs inwhichtheyoperate.ResultsanddiscussionTargetomesofmiR-K12-11inendothelialandBcellshad littleoverlapindirecttargetgenes,butsharedmany indirecttargetsincommonpathwaysTomimicthecellularcontextofmiR-K12-11,wemoderatelyexpressedmiR-K12-11incellsoflymphatic origin(BJAB)andendothelialorigin(TIVE),usingarecombinantretroviralvectorwithbi-cistronicmiRNA andGFPgenes.TheconstantdetectionofGFPinthe transducedcellsindicatedstableexpressionofthemiRNA gene(Figure3Aand3B).QuantitativePCRresultsfurther confirmedtheectopicexpressionofmiR-K12-11inboth BJABandTIVEcells(Figure3).Specifically,theretroviral transductionapproachimitatesmiRNAexpressionunder physiologicalconditions,unliketransfectionexperiments thatexcessivelyover-expressthemiRNAandtriggerofftargeteffects[39-42].Inourexperiment,thecopynumbersofectopicmiR-K12-11werelowerthaninBCBL-1 cells(KSHVinfectedBcelllineisolatedfromcancerpatientswithPEL),indicatingthatitwasnotexpressedat superphysiologicallevels(Figure3C).Tocomparethe GRNsofmiR-K12-11tothoseofmiR-155,wealsocarried outretroviraltransductionformiR-155.InBJABcells, miR-155wassignificantlyexpressedovertheendogenous level.ThemiR-155transducedTIVEcells,however,did notshowsignificantlyincreasedmiR-155levelsover endogenousexpression,preventingfurtheranalysison miR-155inthiscellline. Inaddition,theover-expressionofmiR-K12-11did notaffectthebaselineexpressionofmiR-155inBJAB cellsbutwasrepressiveinTIVEcells(Additionalfile1: TableS4). RNAsamplesformicroarrayanalysiswerecollected fromfourbiologicalreplicatesofBJABcellsexpressing miR-K12-11ormiR-155,TIVEcellsexpressingmiR-K1211,andcorrespondingmockcontrols.Allsampleswere successfullyhybridizedandshowedstatisticalagreement amongbiologicalreplicates(Pearsoncorrelation>0.9, Spearmancorrelation>0.9,weightedkappa>0.7).Differentiallyexpressedgenes(DEGs)weredeterminedusingYang etal.BMCGenomics 2014, 15 :668 Page2of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 3

pairedcomparisonswithFDR<0.05asthesignificance cutoff.Amongthetotal13,793genessurveyedbythe array,141wereDEGsresponsivetomiR-155inBJAB cells,andmiR-K12-11affected1,215and3,189genesin BJABandTIVEcells,respectively(Table1;Additional file2:TableS1).EndogenousexpressionofmiR-155is expectedtoaffectitstargetgenes,andthereforefew geneswereexpectedtobedifferentiallyregulatedbythe additionofectopicmiR-155.This,andthetargetspecificitybeyondtheseedsequence,ledtofewoverlapping DEGsbetweenmiR-155andmiR-K12-11inBJABcells (Figure4).ThefoldchangesoftheDEGsweremostly modest:91%oftheDEGscausedbymiR-K12-11had lessthana50%changeattheRNAlevelinTIVEcells (Figure4A).Theeffectwasevenmoremoderatein BJABcells,with97%oftheDEGschanginglessthan 50%.Thesmallfoldchangeswereconsistentwithpreviousreports[7,11]thatmiRNAsactasfinetunersof geneexpression. GenescommonlyaffectedbymiR-K12-11between BJABandTIVEwererelativelyfew(<20%;Figure4B and4C).WealsocomparedourDEGswithmultiple miR-155/miR-K12-11perturbationstudies(Additional file2:TableS1).AsimilarstudyexpressingmiR-K12-11 inBJABtransductants[29]had40%oftheDEGs(19out of48)sharedbyourmiR-12-11targetsinBJAB.Nosuch studieshaveyetbeencarriedoutinendothelialcells.In Figure2 Analysispipeline. Bycomparingthemicroarrayprofiles ofmiRNA-expressingcellsandmocktransducedcells,geneswith significantchangeswereidentified.Thedown-regulatedgeneswith predictedmiRNAbindingsiteswerecategorizedasputativedirect targetsofmiR-K12-11/miR-155.Fordirecttargetsthatareknown transcriptionfactors,transcriptionfactorbindingsites(TFBS)were searchedinthepromoterregionsofotheraffectedgenes.Forthose indirecttargets,motifanalysiswithi ntheirsequencesidentifiedpotential regulators.Inaddition,GeneOntologyandknownprotein-protein interactionshelptobuildthegeneregulatorynetworks(GRNs). Figure1 MicroRNAscanaffectGRNsdirectlyandindirectly. TheregulatoryeffectsofamiRNAarenotlimitedtothedirectRISC-dependent targeting.BothdirectandindirecttargetsareintegralcomponentsofGRNsandshouldbeincludedinfunctionalanalysis.WhenamiRNAis over-expressed,itsdirecttargetsshouldbedown-regulated.Ifthedirecttargetisarepressorofdownstreamgenes,thenasaresultofmiRNA regulation,thesegeneswillbede-repressedandtheirlevelswillgoup(UpregulateddifferentiallyexpressedgenesorDEGs).Ontheother hand,genesdownstreamofactivatorsandtranscriptionfactorswillgodownaccordinglywiththedirecttargets.Inaddition,proteinsthat physicallyassociatewithdirecttargetstofunctiontogetherinacomplexmayalsobeaffected. Yang etal.BMCGenomics 2014, 15 :668 Page3of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 4

othercelltypes,fewoverlappinggeneswereidentified, likelybecausethetissuespecifictranscriptomesaredifferent(Evidenceontissuespecifictranscriptomeprofiles isabundant,e.g.in[43,44]).Theseresultsdemonstrated thetissuespecificityofmiRNAtargetgenesandthe importanceoftargetomeidentificationinrelevantcell types. DirecttargetsofmiRNAsareexpectedtoberepressed throughsequencecomplementarity.Weidentifiedthese genesasdown-regulatedDEGsthatalsocontainedseed matches,aspredictedbyaunionofbioinformaticsalgorithms(Additionalfile2:TableS2).Therepressionoffour suchgeneswasverifiedbyqPCR.TheyareAGTRAP (angiotensin),APOBEC3G(c ontrolsRNAprocessing), AB 0 200 400 600 800 1000 1200 miR-K12-11 Copy Number per CellC 0 0.5 1 1.522.5 3 3.5 4 4.5 5 Expression levels of miR-155 in BJAB cellsD 0 0.5 1 1.5 2 2.5 3 3.5 44.55 Expression level of miR-K12-11 in BJAB cellsE Figure3 EctopicmiR-K12-11andmiR-155expression.A and B .BJAB (A) andTIVE (B) cellsstablyexpressGFPafterfoamyvirustransduction andpurificationbyFluorescenceActivatedCellSorter. C .ExpressionandcopynumberanalysisofmiR-K12-11intransducedcellscomparedto thePELcelllineBCBL-usingstem-loopqRT-PCR.TheabsolutenumbersofmiR-K12-11fromtransducedcellswaslowerthaninBCBL-1indicating thatectopicexpressionwasnottosuper-physiologicallevels. D and E .ExpressionlevelsofmiR-155inBJABcells.Therewasendogenousexpression ofmiR-155,althoughtheectopicmiRNAexpressionwashigher.Multiplicityofinfection(MOI,i.e.copiespercell)didnotresultinconsistentand significantchangesinthemiRNAexpressionlevels,andwasthereforenotseparatelyconsideredinfurtheranalysis.Yaxis:relativequantitytoth e referenceRNU66. Yang etal.BMCGenomics 2014, 15 :668 Page4of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 5

SAMHD1(regulatesTNFproinflammatoryresponses) andSOCS1(cytokinesuppressor)(Figure4D).AGTRAP andSAMHD1arevalidatedtargetsofmiR-K12-11[29]. MiR-155isabletosuppressSOCS1,asuppressorofcytokinesignaling[45]andAID,amemberofthesamefamily ofdeaminaseswithcriticalfunctionsinadaptiveandinnate immunityasAPOBEC3G[46-48]. ComparisonbetweenthecomputationaltargetpredictionandDEGsfoundonlyasmallportionoftheDEGs attributabletodirecttargeting.Thenumberofupregulatedgeneswasaboutthesameasthenumber down-regulated.Down-regu latedgenesandpredicted Table1Number,directionandfoldchange(FC)of differentiallyexpressedgenes(DEGs)miRNACell type DirectionFDR <0.05 FDR<0.05 andFC>1.2 FDR<0.05 andFC>2 miR-K12-11 TIVEDown16071332151 Up1582 miR-K12-11 BJABDown60832521 Up607 miR-155 BJABDown52374 Up89DEGswereidentifiedusingapairedtestwithsignificancecutoffFDR<0.05. TIVE miR-K12-11BJAB miR-155 BJAB miR-K12-11 Figure4 OveralleffectonthetranscriptomeafterectopicmiRNAexpression.A .miRNAeffectsarequantitativelymoderate.Thefold changeofexpressionlevelsformostDEGswasbelow2-fold. B :Venndiagramshowingcommongeneexpressionchangesbetweencelllines. C .Heatmapshowingtheexpressionchangecomparedtothemocksamplesforalldown-regulateddifferentiallyexpressedgenes(DEGs)thatare alsopredictedtobemiR-155/miR-K12-11targets.MostDEGsshowstrongtissuespecificity. D .VerificationofmicroarraymeasurementsbyqPCR onfourpreviouslyreportedmiR-155/miR-K2-11targets. Yang etal.BMCGenomics 2014, 15 :668 Page5of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 6

targetswereassociatedinTIVEcells(chi-squaretest p=0.0128forTIVE,p=0.3227forBJAB)(Additional file1:TableS4).Severalfactorsmaycontributetothe predictedbutnotobservedtargets:falsepredictionsby thebioinformaticsalgorithms;truetargetsthataretissuespecific,falsenegativesforthetestsofdifferential expression;ortargetssubjec ttotranslationalcontrol notmeasuredbymRNAprofiling. DespitethelimitedoverlapbetweenDEGsinTIVE andBJABcells,miR-K12-11targetedmanycommon pathwaysinthesetwocelltypes(Additionalfile2: TableS3).BycomparingGeneOntology(GO)terms withDEGsusingFisher ’ sexacttest(significancecutoffP<0.05;GEOaccession:GSE59412) Ž ,wefound carbohydratemetabolismamongthetopenriched pathwaysinbothcelltypes(A dditionalfile2:TableS3). Delgadoetal.[49]reportedthatKSHVlatentinfection ofendothelialcellsstronglyinducedtheWarburgeffect, thephenomenonthatcancercellsincreasedglycolysisto meettheirenergyneeds[50,51].GlycolysiswasalsoidentifiedasthetopenrichedbiologicalprocessinacomprehensivemiRNAtargetomeanalysisinKSHVinfected PELcells[52].Takentogether,thisevidencesuggeststhat miR-K12-11isanimportantregulatorforthemetabolic changeafterKSHVinfectioninbothendothelialand Bcells.EffectofmiR-K12-11wasamplifiedbytranscription factorsandproteininteractionsGOenrichmentanalysisidentifiedsequence-specific transcriptionfactors(TFs)andproteinbindingamong thetopmolecularfunctionsofdirectmiR-K12-11targetsinbothBJABandTIVEcells(Fisher ’ sexacttest p<0.05),leadingustohypot hesizethattheindirect targetswereproducedbytranscriptionalregulationand proteininteractions.EnrichmentofTFsinmiRNAtargetshavebeenreportedforplants[53],insects[54]and human[55].MiRNAregulationcancontrolTFlevels [56-59]andexplainstheimportanceofthe3 UTRfor thestabilityofTFs[60,61].Bybindingtopromoterelementsandinteractingwithcofactors,TFsregulatethe expressionofalargenumberofgenesandareableto amplifytheeffectoftheinitialmiRNAtargetingevent. WhilemiRNAregulationcanresultinanindirecteffect ofbothup-regulationanddown-regulation(Figure1), negativeregulatorsofgeneexpressionaremorecontextdependentanddifficulttoprove.Herewefocusedonthe feed-forwardGRNsinwhichthecomponentsconsistentlychangetowardsthesamedirection. InTIVEcells,weidentifiedmultiplecancerassociated TFsthatweredown-regulatedandtherebyamplifiedthe regulatoryeffectsofmiR-K12-11.WeidentifiedCEBP E2F1,PAX6,RELA(alsoknownasNFBp65),and STAT1usingacombinationofDEGsandtargetprediction. CEBP isapreviouslyconfirmedtargetforbothmiR-155 andmiR-K12-11inBcellsandinthecontextofhuman hematopoiesis[62,63].E2F1isamasterregulatorofcell cycle.PAX6isinvolvedintissuespecificationduringearly development.RELApromotesDNArepairandresistance toapoptosisthroughtheregulationofanti-apoptotic proteins.STAT1isrequiredforantiproliferativeactivity, immunesurveillanceandtumorsuppression.Repression ofthesekeyregulatorsinvolvedincancerbymiR-K12-11 mayhelptheestablishmentoflatencyandplayaroleinKS tumorigenesis.Moderatedown-regulationofthesefiveTFs bymiR-K12-11shouldresultindecreasedexpressionof theirdownstreamgenes. PutativedownstreamtargetsofCEBP ,E2F1,PAX6, RELAandSTAT1wereidentifiedbasedonscreeningfor correspondingtranscriptionfactorbindingsiteswithin promoterregionsusingHMMalgorithms[64].Initially, 3000to8000putativeTFBSwerecatalogued.Genesthat werenotonthearray,orwerenotexpressedinmock transducedcells(i.e.lowintensityspotsonthearray) wereomitted.Genesnotdifferentiallydownregulatedin thecontrolvsmiRK12-11werealsoremovedinorder tofocusspecificallyongenesthatwereresponsiveto ectopicmicroRNAexpression.Duetothespatialand temporaldynamicsofgeneexpression,TFbindingis predominantlycelltypespecific[65].TheDNase-seq dataonHUVECcells(primaryendothelialcells)from theENCODEprojectenabledidentificationofactive chromatinregions.GenesthatdidnotshowDNase hypersensitivitywerealsofilteredfromourlistofgenes withTFBSastheylackTFaccessibility.Thesefiltering stepswereappliedtoeachofthelistsgeneratedfrom thepreliminarypredictionresultsinconsiderationofthe cellularcontextandthelackoftissuespecificityincomputationalprediction.Afterfiltering,480geneswere deemedpossibletargetsofCEBP ,240forE2F1,274for PAX6,499forRELA,and571forSTAT1.Whileallof thesegenescontainedTFBSforthecorrespondingTF, morethan66%ofthesegenesdidnotcontainseedsequencematchesformiR-K12-11.Thereforetheirdownregulationwasunlikelytobeduetodirecttargetingby miR-K12-11,butthroughtherepressionoftheTFsby miR-K12-11.Thisanalysisconstructedtheextended GRNsofmiR-K12-11,includingthecandidatedirecttargetsofasmallnumberofTFsandhundredsofdownstreamgenes(Figure5). Co-occupancyofdifferentTFsonpromoterscanform distinctfunctionalregulatorycomplexesinacelltype specificmanner.Thesecomplexesorregulatorymodules areamechanismespeciallycommontopleiotropicTFs suchasE2FsandSTATs[66].Weexaminedourcontext specificTFBSprediction,andfoundthatco-localization ofmultipleTFsonpromoterswasfrequent(Table2). PutativeCo-bindingofSTAT1andE2F1wasidentifiedYang etal.BMCGenomics 2014, 15 :668 Page6of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 7

for92down-regulatedgenes(14.96%ofdown-regulated genes;chi-squaretestp<0.05).RELAandSTAT1 coocupiedparticularlyfrequent(n=200;32.52%of thedown-regulatedgenes;chi-squaretestp<0.001), consistentwithdatathatactivationofsomegenesrequiresbindingofbothSTAT1andNFKB[67]. Aprotein-proteininteraction(PPI)paircantransmit theexpressionchangeofoneproteinthatwasrepressed bythemiRNAtoitsinteractingpartner(Figure1).CombiningTFBSwiththePPImapprovidedmoredetailsfor extendingregulatoryeffects.Forthispurpose,weassembledthecompletehumanproteininteractomefromIntAct [26]andBioGrid[27,28].Thecompleteinteractomecontains173,609interactingpairsrepresentedby11,494genes. Theconnectivityandtheneighbornumbersfollowed Figure5 Transcriptionfactorbindingsites(TFBS)predictionforTFsdirectlytargetedbymiR-K12-11inTIVEcells. MAPPER2predicted thousandsofgeneswithbindingsitesforeachofthefiveTFs. A .Allpredictedsiteswiththegenesnotonthearray,notexpressed,orwerenot differentiallyexpressedbetweenmiRK12-11inductionandcontrol,orlocatedininactivechromatinregionsaccordingtoENCODEdatainblue andthegenesthataretargetsofTFsinred(containingseedsequences)orgreen. B :GenescontainingTFBSanddown-regulatedcanbefurther dividedintotwogroups:thosecontainingbindingsitesforbothTFandmiR-K12-11(red),andthoseonlycontainingTFBSbutnotseed sequencesformiR-K12-11(green). Table2Co-bindingofmultipleTFsonsamepromotersCobindofFrequencyPercent CEBPBXE2F16911.22% CEBPBXPAX612820.81% CEBPBXRELA19131.06% CEBPBXSTAT118129.43% E2F1XPAX6467.48% E2F1XRELA8814.31% E2F1XSTAT19214.96% PAX6XRELA11518.70% PAX6XSTAT19315.12% RELAXSTAT120032.52%Percentagewasbasedon615down-regulatedgenes.Yang etal.BMCGenomics 2014, 15 :668 Page7of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 8

powerlawdistribution(Additionalfile2:FigureS3).This comprehensivehumanPPInetworkcontainsallavailable geneidentifiersasthefocalgenesandallgenesthatphysicallybindtoeachfocalgeneasitsinteractinggenes.Afocal geneanditsdirectlyinteractinggenesweredefinedasa subnetwork. TorefinethePPIforthespecificbiologicalcontextin thisstudy,weintegratedthecuratedinteractomewith ourexpressiondata,andremovednodesforgenesnot onourarrayandnon-expressedgenesfromthePPInetwork.Foreachsub-networkconsistingofanodeandall itsinteractinggenes,theenrichmentfordownregulated targetsofmiR-K12-11wastested.Wefoundthatthe neighboringgenesofE2F1wereenrichedwithgenes down-regulatedbymiR-K12-11,indicatingthatthesubnetworkwastargeted(Figure6).Similarlocalenrichmentfordownregulatedtargetswasalsoidentifiedfor non-TFproteins,liketheapoptosiseffectorCASP9 (Additionalfile2:FigureS2). Thedegreeofexpressionlevelchangesfortheeffectors inBJABcellsweremoresubtle(Table1).Still,miR-K12-11 overexpressioncausesexpressionchangesinmorethan 1000genesinadditionto197directlytargetedgenes.TFs wereidentifiedfromtheputativedirecttargets,including E2F1,aTFdirectlytargetedbymiR-K12-11alsoinTIVE cells.ToexamineTF-dependentregulationaffectedby miR-K12-11inBJABcells,weanalyzedthepromotersequencesofDEGsusingRSAT[68]andTOMTOM[69]. Fromthesetofdown-regulatedgenes,E2F,SP1andKLF wereidentifiedasenrichedmotifs(Figure7).TheseTFs containtheseedsequenceofmiR-K12-11,supportingtheir rolesaseffectorgenesdirectlytargetedbymiR-K12-11. TheseTFsarealsotranscriptionalactivatorsandtheregulatoryeffectofmiR-K12-11isexpectedtocauseacascade ofrepressionoftranscription.miR-K12-11synergisticallyregulatedmultiplesignaling pathwaystorepresstheactivationofinterferon responsesMiR-K12-11alsoregulatesinterferonresponsesanda varietyofsignalingpathways(Figure8).SignalingpathwayshavebeensuggestedaslogicaltargetsofmiRNA regulation,wheresmallchangesintheexpressionlevel ofupstreamgenescanaffectthesignaltransductioncascadesignificantly[70].IndividualmiRNAsareableto targetseveralcomponentsofasinglesignalingpathway, asinthecasesofmiR-8forWntsignaling[71],miR-21 forRTKsignaling[72,73]andmiR-126forVEGFsignaling[74,75].WeidentifiedmultiplelayersofJAK-STAT signalingthatwereaffectedbymiR-K12-11,withdirect targetsdifferingbetweenBJABandTIVEcells(Additional file1:TableS4).InBJABcells,theputativedirecttargets includethecytokinereceptorIFNGR1(fc>1.2),which isaconfirmedtargetofmiR-155[76].InTIVEcells, miR-K12-11directlytargetedSOCS1(foldchange>1.4, FDR<0.05)andthetranscriptionactivatorSTATs(STAT1 andSTAT2foldchange>2andSTAT3foldchange>1.4 FDR<0.05forall)(Figure8;Additionalfile1:TableS4). Interferonsarepotentcytokinesproducedinresponse toviralinfectionthatmediatebothinnateimmuneresponseandsubsequentdevelopmentofadaptiveimmunity.Modulationofinterferonpathwaysisrequiredto suppresstheinnateimmuneresponseandestablishsuccessfullatentinfection.AlongwithJAK-STATsignaling, multipleothersignalingpathwaysassociatedwithinterferonresponsesweretargetedbymiR-K12-11.InTIVE cells,miR-K12-11targetsPTENandAKT1S1ofthe AKTpathway,SKIandSMAD4oftheTGFsignaling pathway,MYD88oftheTLR-MYD88pathwaywhich regulateshostdefense,andRELAoftheNFBsignaling pathway(Additionalfile1:TableS4). Theaffectedsignalingpathwaysarenotindependent fromeachotherbutknowntobecoordinatedthrough cross-talking[77].ThecooperationofSTATsandNFB canactivatedownstreamantiviralgenessuchasthe IRFs,afamilyoftranscriptionfactors(TFs)(Additional file3:TableS5)[67,78].IRFsandotherTFssuchasNFB andAP-1complex(ATF-FOS-JUN)regulatetheexpressionofinterferons.Besidestheirtranscriptionalactivation property,STATsalsomediatetheIFNresponsethrough competitionwithAP-1[79].InBJABcells,IRF3,ATF1, ATF4andATF5weredown-regulatedbymiR-K12-11 (Additionalfile1:TableS4;Additionalfile3:TableS5),but notlikelythroughdirectbindingbecausetheydonotcontaintheseedsequencematchsites. InTIVEcells,aconsistentdecreaseofexpression levelswasobservedforSTAT1,STAT2,STAT3,and theirtranscriptionallyregulatedgenes(Figure8).RELA andATF7,whichcontaintheseedsequenceofmiRK12-11andaredownregulatedareputativedirecttargetsbymiR-K12-11(Additionalfile1:TableS4).JUND (memberofJUN,protectscellfromapoptosis)andmultipleIRFswerealsodown-regulatedthroughindirecteffects.ThedecreasedexpressionofIRF1,IRF7andIRF9 (alsoknownasp48)maybeduetoreducedSTATlevels sincenoneoftheseIRFscontainseedsequencematches (Additionalfile1:TableS4).WhileRELAexpressionis subjecttothenegativeregulationofIRF7,weshowthat itisdirectlydownregulatedbymiR-K12-11.Asimilar functionalloophasbeenreportedformiR-155,whichby attenuatingNFBactivity,contributestostabilization ofEBVlatency[80].IRF9canalsointeractwithSTAT dimerstoformaproteincomplextobindpromoter sequences[81].AsimportantTFs,thesereducedIRFs likelyaffectedavarietyofdownstreamgenes.Anumber ofwellcharacterizedinterferonstimulatedgenes(ISGs) suchasISG15,USP18andtheOASgenefamilyallexhibitedsignificantdown-regulationbymiR-K12-11,stronglyYang etal.BMCGenomics 2014, 15 :668 Page8of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 9

Figure6 Changeofexpressionintheinteractinggeneswiththefivetranscriptionfactors. Amongthegenesthatdirectlyinteractwith E2F1 (A) ,CEBPB (B) ,PAX6 (C) ,RELA (D) andSTAT1 (E) ,thereisanenrichmentofdown-regulationinaccordancewiththecenternodeTFgenes. Proteininteractions,aswellasdirecttargetingofmiR-K12-11(genesofthecircles)maycontributetothecoordinateddown-regulation. Yang etal.BMCGenomics 2014, 15 :668 Page9of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 10

supportinginhibitionofinterferonresponsesinendothelialcells(Additionalfile2:TableS3;Additionalfile1: TableS4). Liang etal. [82]hasidentifiedIKK asamiR-K12-11 targetinlungcancercells.ThoughIKK levelwasunchangedinthisexperiment,itsdownstreameffectorIRF andNFBwerereduced.ItislikelythatmiR-K12-11 attenuatesIFNsignalingbydown-regulatingmultiple possiblecomponents,IKK inlungcancercells,IFNGR1 inBcells,andSTAT1inendothelialcells(Figure5; Additionalfile2:FigureS2).TargetingofthesekeycomponentsnotonlyeliminatedtheactivationofIFNresponse, butalsoincreasedkeyproliferativeandsurvivalsignalsthat arebeneficialforKSHVlatencyestablishment. InadditiontomiR-K12-11,KSHVexpresseshomologs tocellularIRFs,thatpreventtheassociationofIRFswith Figure7 MotifenrichmentanalysisfromDEGsofBJABcells. Motifsidentifiedfromthepromotersequencesofgenesdown-regulatedby miR-K12-11inBJABcellsmatchedtheconservedbindingmotifsofSP1,KLF4andMYC.SP1andKLF4weredown-regulatedthemselvesandmay thereforerelaytheregulatoryeffecttoalargergroupofgenes.Forup-regulatedgenesinresponsetomiR-K12-11,motifofFOXA1andFOXA2 wereidentifiedasputativemediatorsfortheincreaseofgeneexpressioninBJABcells. Yang etal.BMCGenomics 2014, 15 :668 Page10of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 11

theirco-activators[83,84].Theinhibitionimposedby miR-K12-11andvIRFtocellularIRFsmayreinforce eachotherthroughafeedforwardloop.Whilewecannot estimatetherelativecontributionofmiR-K12-11versus vIRFsignaling,expressingamiRNAcomeswiththe addedadvantageofnotelicitinghumoralhostimmune responsesliketheproteinproductsdo.OtherKSHV geneproductssuchasv-cyclinandvIL-6arealsocytokinesignalinggenesthatcanblocktheactivityofhost homologs[85].TakentogetherandinpartduetomiRK12-11,KSHVisabletomanipulatecellcycleandapoptosis,toevadeimmuneresponse,andpromoteproliferation,andsurvivalofinfectedcells.ConclusionsAnalyzingGRNsprovidesinsightsintotheregulatory networksofmiRNAregulationthatcannotbefoundby studyingsinglegenes.ExaminingmiRNAtargetgenesin thecontextofcellularGRNscanseparatetargetsthat drivephenotypicconsequencesfromnon-functionalones. GRNsarehighlytissuespecific[43,44,86,87],thereforeit isimperativetorecognizethetissuespecificityanddefine theGRNsofthemiRNAonlyintherelevantcelltypes. WedemonstratedasystemsapproachtoinferthecombinatorialGRNsutilizedbymiR-K12-11incellularcontextsthatareclosetoKSHVinfection invivo .Thisstudy includedthefirsttargetidentificationofKSHVmiRNAs inTIVEcells,afrequentlyusedendothelialcellculture systemforstudyingKSHVinfection[88-90]. WefoundthatmiR-K12-11functionedatdifferent hierarchicallevelsoftheGRNs.Putativedirecttargets ofmiR-K12-11wereunderrepresentedinthealtered transcriptomes.Bytargetingafeweffectorgenes,five timesmoregeneswereaffectedbeyonddirectsequence pairing.Differentcomponents,butfrequentlyofcommonbiologicalpathways,weretargetedinBJABand TIVEcells.TherewasapreferencetotargetingTFs,includingCEBP ,PAX6,RELA,andSTAT1inTIVEcells, FOXA,KLFandSP1inBJABcells,andE2F1commonto both.DecreaseintheTFlevelssignificantlyamplifiedthe effectofmiR-K12-11tomanymoregenesdownstream, whichcouldpotentiallyresultinbroadphenotypiceffects suchasinducingendothelialcelldifferentiationinthe contextofKSHVinfection.SinceviralmiRNAscoevolvewithhostgenesandcanbefunctionalorthologs, wefoundthatlikeitscellularhomologmiR-155[29,30], miR-K12-11isalsoinvolvedininnateandadaptiveimmunefunctionsbymodulatingtheinterferonresponse andcarbohydratemetabolism.PreviouslyvalidatedtargetsofmiR-155suchasCEBP andSOCS1werealso identified. MiR-K12-11alsoregulatedgenesatthemiddleand bottomofthewell-knownsignalingcascades,likesignalingproteinsandcaspases,andmodulatedkeybiological processeslikecellcyclecontrolandvarioussignaling pathways,allofwhichwereaccomplishedbytargeting distinctsetsofgeneswithineachcelltype.Hostresponsestoviralinfection,suchasinnateimmunityand Figure8 Interferonresponseswererepressedviatheinterplayof multiplesignalingpathwaysandtranscriptionfactors. MiR-K12-11targeted cytokinereceptorsandTFs,bothofwhichaffectedavarietyofinterferonstimulatedgenes(ISGs).Throughdirectandindirectimpact,miR-K12-11i sable tomodulatethehostinnateimmuneresponseandtohelpKSHVtoestablishlatency. Yang etal.BMCGenomics 2014, 15 :668 Page11of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 12

apoptosis,arecounteredbymiR-K12-11andadditional viralgeneproducts,enablingtheestablishmentoflatency.Themultilevelregulationallowedoneindividual miRNAtoprofoundlyaffectthegeneexpressionprogramtoadapttospecificneeds. Finally,theapproachwehavetakenheretoidentifying miR-K12-11GRNscanbeappliedtoinvestigatingthe viralandcellularmiRNAsindifferenttissuesandsystems.Withananticipatedexpansionofgenomewide dataonshortRNAprofiles,ChIP,ribonomics,andproteomicsinthenearfuture,ourstrategycouldbeapplied torevealconditionalregulatorypathwaysinahighlytissueandcelltypespecificmanner.MethodsTheexperimentaldesignallowscomparisonofmiR-155 transducedcells,miR-K12-11treatedcells,andmock transducedcells.Theexperimentwasconductedinfour subsequenttimeperiodssuchthatalltheexperimental conditionswereindependentlyrepeated.VectorsystemThefoamyvirusvectorplasmidpCEGFPLwasconstructedasdescribedbefore[62].Thegag,poland envgenesarereplacedbyamiRNAgenefollowinga minimalhumancytomegalovirus(CMV)immediateearlypromoteratthetranscr iptionstartsitelocated inthe5 -LTRandaGFPgeneasthereporter.The replicationabilityoftheviralvectorcanberestored byco-transfectionwiththepackagingplasmidpCI env3.5.RecombinantvirusvectorsexpressingmiR155,miR-K12-11andemptyvectorwithoutinsertas thecontrolwereproducedbytransientcotransfection withMirustransfectionreagentfollowingthemanufacturer ’ sinstructions.Thesupernatantwasfiltered, concentratedbycentrifugation.Resultingfoamyvirusesweretitratedonfresh293Tandgreencellswere evaluatedforGFPexpressionusingfluorescencemicroscopy.Notably,emptyvectorsmaylackthecontrol overthenon-specificeffectoftheprecursortranscriptsbuttheywereabletoreducetheoff-targeteffectsofascrambleinsert.CellcultureBJABisaBurkitt ’ slymphomahumanBcelllinethatis uninfectedandEpstein-Barrvirus-negative.BJABcells weregrowninculturesuspensionincompleteRPMI mediumwith10%fetalbovineserum(FBS). T elomerasei mmortalizedhumanumbilicalv ein e ndothelial(TIVE) cells[88]havebeenspeciallydevelopedforthepurposeof studyingtheeffectsofKHSVlatentinfectioninendothelial cells.TIVEcellsareadherentcellsgrowninMedium199 supplementedwith20%FBSand60 g/mLEndothelial CellGrowthFactor(ECGF).TransductionandvalidationTIVEandBJABcellswereretrovirallytransducedattwo levelsofMultiplicityofInfection(MOI):1and10.72hr posttransduction,positivecellsweresortedaccordingto theirGFPsignal.Cellswerealiquotedin1millioncells pertubeandfrozendowninliquidnitrogen.EmptyvectorswithoutmiRNAexpressioncassetteswereusedfor mocktransductiontocontrolfortheimpactofretroviral integrationonthecellulartranscriptomes.Theaimof thefreezingistosynchronizethegrowthstatusofthe cellsacrosssamples,andtoreducenoisetomicroarray profiling.Later,cellswereremovedfromliquidnitrogen andgrownforthesamenumberofpassages.RNAwas extractedusingtheRNA-Beereagentaccordingtothe manufacturer ’ sinstructions.Thequantityandqualityof RNAwasconfirmedbyNanoDropspectrometerand agarosegelelectrophoresis.TheintegrityoftotalRNA wasassessedwithAgilentBioanalyzer.Expressionof miRNAswasexaminedusingTaqManqPCR.Expression levelsofmiR-155andmiR-K12-11werenormalizedto RNU66levels.TheMOIdidnotresultindifferencesin miRNAexpressionlevels.Therefore,allsampleswere treatedasbiologicalreplicates.MicroarrayanalysisForeachHG-133plus2.0chip,200ngRNAwas usedasthestartingmaterial.RNAwassynthesized andlabeledusingGeneChip3 IVTExpressKitand chipswerehybridizedacco rdingtomanufacturerinstructions(Affymetrix).Rawdata(cellintensityfiles,CEL) weresummarizedusingAffymetrixExpressionConsole software(v1.1).Chipswereexaminedforsuccessful hybridizationbyensuringt hatthemarginaldistributionofallslideswassimilar.Sampleswerecompared fortheglobaleffectofmiRNAtreatmentatapopulationlevelusingprincipalcomponentanalysis[91]. Probesetswereflaggedas  absent ’ iftheywereabsent accordingtoAffymetrixprobedetectionalgorithm (AffymetrixStatisticalAlgorithmsDescriptionDocument. http://media.affymetrix.com/ support/technical/whitepapers/sadd_whitepaper.pdf)inmorethanhalfofthesamples.ThedataweredepositedintheGEOdatabasewith accessionnumberGSE59412. ThefollowingmodelwasfitYij= + i+ ij,whereYijisthedifferenceofthelog2signalsforeachprobeset betweenthemiRNAtransducedandcontrolvectorfor theithconditionandthejthreplicate; isthedifference fortheoverallexpressionmean. ij~N(0, i 2).The signaldifferencesbetweenmiRNAtransducedsamples andtheircorrespondingcontrolsampleswereusedas thispaireddesignreflects theexperimentaldesign. Thetestofthenullhypothesisthat i=0isadirect testofthemiRNAcondition.Ftestsforeachofthe miRNAconditions(miR-155inBJAB,miR-K12-11inYang etal.BMCGenomics 2014, 15 :668 Page12of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 13

BJAB,miR-K12-11inTIVE)wereconducted.AnFDR of0.05wasusedtodeterminestatisticalsignificance fortheprobeset[92]. Theprobesetswereannotatedbycomparingthegenomepositionsofhumangenesandofprobesethits. Agenewasconsidereddifferentiallyexpressed(DEG) whenatleastoneprobesetwassignificant.Thechange inexpressionlevelswasthedifferenceinthemeanofall probesetsbetweentreatmentandcontrol.DEGswere examinedforpotentialfunctionalgroupsbyenrichment analysis[93].EnrichedGeneOntologyterms[94]ofthe DEGsandknownbiologicalpathwayswerecompared usingFisher ’ sexacttest.IdentificationofdirectmiRNAtargetsToincreasethespecificityofourGRNinference,we focusedonthecanonicaltargets,forwhicharangeof targetingruleshavebeendefinedandmostprediction algorithmsaredeveloped.Evenso,ouranalysisandpreviousreports[19,95]foundthelackofconcordance acrossthemiRNAtargetpredictionofdifferentalgorithms.Thisistheresultofusingdifferenttrainingset oftargetgeneswhenthealgorithmsweredeveloped.A comprehensivelistofputativetargetsofmiR-155/miRK12-11wascreatedbyusingtheunionoftargetpredictionfrommultiplealgorithms:EMBL-EBImirBase[96], TargetScan[21],PITA[97],DIANA[98],miRDB[99], RNA22[100],mirWalk[101],mirZ[102]andPicTar [103].Inaddition,SylArray[104]wasusedtoidentify enrichmentofmiRNAseedsequencematches.Thepredictedtargetswerealsocomparedtovalidatedtarget genesintheliterature.IdentificationoftranscriptionfactorregulationAlistofhumantranscriptionalfactor(TF)geneswas obtainedfromtheJASPARdatabase[105]andaTFcensusstudy[106].DEGsonthislistaswellonthemiRNA targetlistwereexaminedi ndetailforexpression changesandbiologicalimplications,astheywerethe primarytargetsofthemiRNA.WeusedMAPPER [64,107],whichusesbindingsiteinformationfrom TRANSFACandJASPARdatabasesderivedHidden MarkovModels,todetectputativetranscriptionfactor bindingsites(TFBS).GenescontainingTFBSwithinthe upstream2kbregionoftranscriptionstartsiteswereidentifiedasgenesthatmightbeunderTFregulation. ForDEGswiththesamedirectionofexpression change,enrichedmotifsintheirpromoterregionswere identifiedusingRSAToligoanalysis[68].Themotifs werecomparedtothebindingmotifsofTFsusingthe TOMTOMprogramoftheMEMEsuite[69].Motifs identifiedfromup-anddown-regulatedsetofDEGs werecompared,anduniquemotifsforeachsetwere identified.AdditionalevidenceforTFregulationwas obtainedfromliteraturesearchandtheTranscriptional RegulatoryElementDatabase(TRED)[108].ChIP-seq (measuringDNA-proteininteraction)andDNase-seq (measuringDNAaccessibilitytoregulatoryproteins) profilesoftheENCODEproject[65]fromcorrespondingcelltypeswereusedtoconstraintheTFregulated genestobetissuespecific.IdentificationofsignalinggenesHumansignalingpathwaydatawasobtainedfromthe NationalCancerInstitutePathwayInteractionDatabase (NCIPID)[109],whichisamanuallycuratedcollection ofbiomolecularinteractionsandkeycellularprocesses assembledintosignalingpathways.NCIPIDholds128 pathwaysincluding47sub-networks.Allsubnetworks withtheirparentnetworkswerecombinedtogenerate thesetofsignalingpathways.Pathwayscuratedinthe BioCartadatabase(http://www.biocarta.com/)wereused forcross-referencingtoreduceambiguity.Inaddition, allpathwaysthathavemorethanonepredictedmicroRNAtargetgenewerekept,leadingtoafinaldatasetof 79humansignalingpathwayscontaining1573unique humanproteins.Thedatabasealsoprovidesinformation onsubcellularlocationtermsfromtheGeneOntology Consortium.Processtypeinformationwasextractedfor eachbiologicalprocess,whichcanbeinput,output, positiveornegativeregulator.Intotal,thereare1120interactionsofwhich765areactivating,74inhibitingand 281proteinsactingasactivatorsaswellasinhibitors.IdentificationoffunctionalinteractionAbinaryinteractomewasassembledenablinganoverviewofallphysicalinteractionsthatcanoccurbetween humanproteins.Geneassociationdataweredownloaded fromGeneRIF(GeneReferencesintoFunction)database atNCBI[110]andtheIntActdatabase[26]atEBIon Febuary282011.TheinteractionsinGeneRIFare sourcedfromBind[111,112],BioGrid[27,28],EcoCyc [113],andHPRD[114].TheIntActdatabaseincludes interactionsfromliteraturecurationatEBIaswellas usersubmission.Onlyprotein -proteininteractiondata forhumanwasretained.Theformatteddatacontaina listoffocalgenesthatcoversallavailablevaluesofgene identifiers,theinteractinggenesforeachfocalgene, thedetectionmethodandthesourceoftheinteraction. Secondaryinteractionsarederivedfromtheinteractionsofthegenesidentifiedasinteractorsoftheinitial focalgene. ThehumanPPInetworkswereplottedasundirected graphs,wherethenodesareproteinsandtwonodesare connectedbyanundirectededgeifthecorresponding proteinsphysicallybindtoeachother.DEGswere mappedtotheinteractomestoidentifytheinteractants oftheindirecttargets.TheexpressionlevelsofgenesYang etal.BMCGenomics 2014, 15 :668 Page13of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 14

belongingtothemapwereexaminedandabsentgenes wereremoved.Up-anddown-regulatedDEGswere flaggedtodisplayindifferentcolors.Afocalgeneandits neighboringgenesweredefinedasasubnetwork.The percentageofDEGsinthesubnetworkforeachfocal genewascalculated.IfDEGswerepresentmoreoften thanintheexperimentasawhole,thefocalgenewas identifiedasanenrichedregulatoranditssubnetwork wasconsideredasresponsive.GOenrichmentwasalso examinedontheenrichedreg ulators,todetermineif transcriptionallyregulat edsub-networkssharedGO termsindicativeofknownorrelatedbiologicalfunctions. ThesubnetworkswereviewedinCytoscape[115,116]for activebiologicalpathways.AdditionalfilesAdditionalfile1:TableS4. Overviewofalldifferentiallyexpressed genes. Additionalfile2: Supplementaltables(TableS1-S3)andfigures (FigureS1-S3). Additionalfile3:TableS5. Resultsoftranscriptionfactorbinding prediction. Competinginterests Theauthorsdeclarethattheyhavenocompetinginterest. Authors ’ contributions YYcarriedouttheexperimentsanddataanalysis,andwrotethepaper.IWB assistedinthelabwork.LMMandRRdesignedthestudyandhelpedwith thedataanalysisandmanuscriptpreparation.Allauthorsreadandapproved thefinalmanuscript. Acknowledgements ThisworkwassupportedbygrantsfromNationalInstitutesofHealth (CA88763,CA119917andRC2CA148407toRRandGM102227toLMM).We wouldliketothankmembersoftheRenneandMcIntyrelabfordiscussions andPeterTurnerforcarefulreadingandeditingofthemanuscript. Received:24April2014Accepted:1August2014 Published:8August2014 References1.ChangY,CesarmanE,PessinMS,LeeF,CulpepperJ,KnowlesDM,MoorePS: Identificationofherpesvirus-likeDNAsequencesinAIDS-associatedKaposi ’ s sarcoma. Science 1994, 266 (5192):1865 – 1869. 2.CesarmanE,KnowlesDM: TheroleofKaposi ’ ssarcoma-associated herpesvirus(KSHV/HHV-8)inlymphoproliferativediseases. Semin CancerBiol 1999, 9 (3):165 – 174. 3.SoulierJ,GrolletL,OksenhendlerE,CacoubP,Cazals-HatemD,BabinetP, d ’ AgayMF,ClauvelJP,RaphaelM,DegosL,SigauxF: Kaposi ’ ssarcomaassociatedherpesvirus-likeDNAsequencesinmulticentricCastleman ’ s disease. Blood 1995, 86 (4):1276 – 1280. 4.DittmerD,LagunoffM,RenneR,StaskusK,HaaseA,GanemD: Aclusterof latentlyexpressedgenesinKaposi ’ ssarcoma-associatedherpesvirus. JVirol 1998, 72 (10):8309 – 8315. 5.RenneR,LagunoffM,ZhongW,GanemD: Thesizeandconformationof Kaposi ’ ssarcoma-associatedherpesvirus(humanherpesvirus8)DNAin infectedcellsandvirions. JVirol 1996, 70 (11):8151 – 8154. 6.ZhongW,WangH,HerndierB,GanemD: RestrictedexpressionofKaposi sarcoma-associatedherpesvirus(humanherpesvirus8)genesinKaposi sarcoma. ProcNatlAcadSciUSA 1996, 93 (13):6641 – 6646. 7.BaekD,VillenJ,ShinC,CamargoFD,GygiSP,BartelDP: Theimpactof microRNAsonproteinoutput. Nature 2008, 455 (7209):64 – 71. 8.SelbachM,SchwanhausserB,ThierfelderN,FangZ,KhaninR,RajewskyN: WidespreadchangesinproteinsynthesisinducedbymicroRNAs. Nature 2008, 455 (7209):58 – 63. 9.FarhKK-H,GrimsonA,JanC,LewisBP,JohnstonWK,LimLP,BurgeCB, BartelDP: Thewidespreadimpactofm ammalianMicroRNAson mRNArepressionandevolution. Science 2005, 310 (5755):1817 – 1821.10.EulalioA,HuntzingerE,NishiharaT,RehwinkelJ,FauserM,IzaurraldeE: DeadenylationisawidespreadeffectofmiRNAregulation. RNA 2009, 15 (1):21 – 32. 11.GuoH,IngoliaNT,WeissmanJS,BartelDP: MammalianmicroRNAs predominantlyacttodecreasetargetmRNAlevels. Nature 2010, 466 (7308):835 – 840. 12.LewisBP,BurgeCB,BartelDP: Conservedseedpairing,oftenflankedby adenosines,indicatesthatthousandsofhumangenesaremicroRNA targets. Cell 2005, 120 (1):15 – 20. 13.LimLP,LauNC,Garrett-EngeleP,GrimsonA,SchelterJM,CastleJ,BartelDP, LinsleyPS,JohnsonJM: MicroarrayanalysisshowsthatsomemicroRNAs downregulatelargenumbersoftargetmRNAs. Nature 2005, 433 (7027):769 – 773. 14.StarkA,BrenneckeJ,RussellRB,CohenSM: IdentificationofDrosophila MicroRNAtargets. PLoSBiol 2003, 1 (3):E60. 15.LewisBP,ShihIH,Jones-RhoadesMW,BartelDP,BurgeCB: Predictionof mammalianmicroRNAtargets. Cell 2003, 115 (7):787 – 798. 16.GrimsonA,FarhKK,JohnstonWK,Garrett-EngeleP,LimLP,BartelDP: MicroRNAtargetingspecificityinmammals:determinantsbeyondseed pairing. MolCell 2007, 27 (1):91 – 105. 17.BartelDP: MicroRNAs:targetrecognitionandregulatoryfunctions. Cell 2009, 136 (2):215 – 233. 18.MazireP,EnrightAJ: PredictionofmicroRNAtargets. DrugDiscovToday 2007, 12 (11 – 12):452 – 458. 19.AlexiouP,MaragkakisM,PapadopoulosGL,ReczkoM,HatzigeorgiouAG: Lostintranslation:anassessmentandperspectiveforcomputational microRNAtargetidentification. Bioinformatics 2009, 25 (23):3049 – 3055. 20.ZhuY,HaeckerI,YangY,GaoS-J,RenneR: -Herpesvirus-encodedmiRNAs andtheirrolesinviralbiologyandpathogenesis. CurrOpinVirol 2013, 3 (3):266 – 275. 21.FriedmanRC,FarhKK,BurgeCB,BartelDP: MostmammalianmRNAsare conservedtargetsofmicroRNAs.GenomeRes 2009, 19 (1):92 – 105. 22.MukherjiS,EbertMS,ZhengGX,TsangJS,SharpPA,vanOudenaardenA: MicroRNAscangeneratethresholdsintargetgeneexpression. NatGenet 2011, 43 (9):854 – 859. 23.GraurD,ZhengY,PriceN,AzevedoRB,ZufallRA,ElhaikE: Onthe immortalityoftelevisionsets: “ function ” inthehumangenome accordingtotheevolutio n-freegospelofENCODE. GenomeBiolEvol 2013, 5 (3):578 – 590. 24.ParkSM,GaurAB,LengyelE,PeterME: ThemiR-200familydeterminesthe epithelialphenotypeofcancercellsbytargetingtheE-cadherinrepressors ZEB1andZEB2. GenesDev 2008, 22 (7):894 – 907. 25.IdekerT,ThorssonV,RanishJA,ChristmasR,BuhlerJ,EngJK,BumgarnerR, GoodlettDR,AebersoldR,HoodL: Integratedgenomicandproteomic analysesofasystematicallyperturbedmetabolicnetwork. Science 2001, 292: 929 – 934. 26.KerrienS,Alam-FaruqueY,ArandaB,BancarzI,BridgeA,DerowC, DimmerE,FeuermannM,FriedrichsenA,HuntleyR,KohlerC,KhadakeJ, LeroyC,LibanA,LieftinkC,Montecchi-PalazziL,OrchardS,RisseJ,RobbeK, RoechertB,ThorneycroftD,ZhangY,ApweilerR,HermjakobH: IntAct – open sourceresourceformolecularinteractiondata. NucleicAcidsRes 2007, 35 (Databaseissue):D561 – 5. 27.BreitkreutzBJ,StarkC,RegulyT,BoucherL,BreitkreutzA,LivstoneM, OughtredR,LacknerDH,BahlerJ,WoodV,DolinskiK,TyersM: TheBioGRID InteractionDatabase:2008Update. NucleicAcidsRes 2008, 36 (Databaseissue):D637 – 40. 28.StarkC,BreitkreutzB-J,RegulyT,BoucherL,BreitkreutzA,TyersM: BioGRID: ageneralrepositoryforinteractiondatasets. NucleicAcidsRes 2006, 34 (Databaseissue):D535 – 9. 29.GottweinE,MukherjeeN,SachseC,FrenzelC,MajorosWH,ChiJT,BraichR, ManoharanM,SoutschekJ,OhlerU,CullenBR: AviralmicroRNAfunctions asanorthologueofcellularmiR-155. Nature 2007, 450 (7172):1096 – 1099. 30.SkalskyRL,SamolsMA,PlaisanceKB,BossIW,RivaA,LopezMC,BakerHV, RenneR: Kaposi ’ ssarcoma-associatedherpesvirusencodesanortholog ofmiR-155. JVirol 2007, 81 (23):12836 –12845.Yang etal.BMCGenomics 2014, 15 :668 Page14of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 15

31.MoffettH,NovinaC: AsmallRNAmakesaBicdifference. GenomeBiol 2007, 8 (7):221. 32.ThaiTH,CaladoDP,CasolaS,AnselKM,XiaoC,XueY,MurphyA, FrendeweyD,ValenzuelaD,KutokJL ,Schmidt-SupprianM,RajewskyN, YancopoulosG,RaoA,RajewskyK: Regulationofthegerminalcenter responsebymicroRNA-155. Science 2007, 316 (5824):604 – 608. 33.RodriguezA,VigoritoE,ClareS,WarrenMV,CouttetP,SoondDR,van DongenS,GrocockRJ,DasPP,MiskaEA,VetrieD,OkkenhaugK,EnrightAJ, DouganG,TurnerM,BradleyA: Requirementofbic/microRNA-155for normalimmunefunction. Science 2007, 316 (5824):608 – 611. 34.CostineanS,ZanesiN,PekarskyY,TiliE,VoliniaS,HeeremaN,CroceCM: Pre-Bcellproliferationandlymphoblasticleukemia/high-grade lymphomainE(mu)-miR155transgenicmice. ProcNatlAcadSciUSA 2006, 103 (18):7024 – 7029. 35.EisPS,TamW,SunL,ChadburnA,LiZ,GomezMF,LundE,DahlbergJE: AccumulationofmiR-155andBICRNAinhumanBcelllymphomas. Proc NatlAcadSciUSA 2005, 102 (10):3627 – 3632. 36.KluiverJ,PoppemaS,deJongD,BlokzijlT,HarmsG,JacobsS,KroesenBJ, vandenBergA: BICandmiR-155arehighlyexpressedinHodgkin, primarymediastinalanddiffuselargeBcelllymphomas. JPathol 2005, 207 (2):243 – 249. 37.O ’ ConnellRM,RaoDS,ChaudhuriAA,BoldinMP,TaganovKD,NicollJ, PaquetteRL,BaltimoreD: SustainedexpressionofmicroRNA-155in hematopoieticstemcellscausesamyeloproliferativedisorder. JExpMed 2008, 205 (3):585 – 594. 38.vandenBergA,KroesenBJ,KooistraK,deJongD,BriggsJ,BlokzijlT, JacobsS,KluiverJ,DiepstraA,MaggioE,PoppemaS: Highexpressionof B-cellreceptorinduciblegeneBICinallsubtypesofHodgkinlymphoma. GenesChromosomesCancer 2003, 37 (1):20 – 28. 39.JeyapalanZ,DengZ,ShatsevaT,FangL,HeC,YangBB: Expressionof CD443 -untranslatedregionregulatesendogenousmicroRNA functionsintumorigenesisandangiogenesis. NucleicAcidsRes 2011, 39 (8):3026 – 3041. 40.PolisenoL,SalmenaL,ZhangJ,CarverB,HavemanWJ,PandolfiPP: A coding-independentfunctionofgeneandpseudogenemRNAsregulates tumourbiology. Nature 2010, 465 (7301):1033 – 1038. 41.TayY,KatsL,SalmenaL,WeissD,TanSM,AlaU,KarrethF,PolisenoL,ProveroP,DiCuntoF,LiebermanJ,RigoutsosI,PandolfiPP: CodingindependentregulationofthetumorsuppressorPTENbycompeting endogenousmRNAs. Cell 2011, 147 (2):344 – 357. 42.SumazinP,YangX,ChiuHS,ChungWJ,IyerA,Llobet-NavasD,RajbhandariP, BansalM,GuarnieriP,SilvaJ,CalifanoA: AnextensivemicroRNA-mediated networkofRNA-RNAinteractionsregu latesestablishedoncogenicpathways inglioblastoma. Cell 2011, 147 (2):370 – 381. 43.GersteinMB,KundajeA,HariharanM,LandtSG,YanKK,ChengC,MuXJ, KhuranaE,RozowskyJ,AlexanderR,MinR,AlvesP,AbyzovA,AddlemanN, BhardwajN,BoyleAP,CaytingP,CharosA,ChenDZ,ChengY,ClarkeD, EastmanC,EuskirchenG,FrietzeS,FuY,GertzJ,GrubertF,HarmanciA, JainP,KasowskiM, etal : Architectureofthehumanregulatorynetwork derivedfromENCODEdata. Nature 2012, 489 (7414):91 – 100. 44.ThurmanRE,RynesE,HumbertR,VierstraJ,MauranoMT,HaugenE, SheffieldNC,StergachisAB,WangH,VernotB,GargK,JohnS,SandstromR, BatesD,BoatmanL,CanfieldTK,DiegelM,DunnD,EbersolAK,FrumT, GisteE,JohnsonAK,JohnsonEM,KutyavinT,LajoieB,LeeBK,LeeK, LondonD,LotakisD,NephS, etal : Theaccessiblechromatinlandscapeof thehumangenome. Nature 2012, 489 (7414):75 – 82. 45.JiangS,ZhangHW,LuMH,HeXH,LiY,GuH,LiuMF,WangED: MicroRNA-155functionsasanOncomiRinbreastcancerby targetingthesuppressorofc ytokinesignaling1gene. CancerRes 2010, 70 (8):3119 – 3127. 46.DemorestZL,LiM,HarrisRS: Phosphorylationdirectlyregulatestheintrinsic DNAcytidinedeaminaseactivityofactivation-induceddeaminaseand APOBEC3Gprotein. JBiolChem 2011, 286 (30):26568 – 26575. 47.DorsettY,McBrideKM,JankovicM,GazumyanA,ThaiTH,RobbianiDF,Di VirgilioM,ReinaSan-MartinB,HeidkampG,SchwickertTA,EisenreichT, RajewskyK,NussenzweigMC: MicroRNA-155suppressesactivationinducedcytidinedeaminase-mediatedMyc-Ightranslocation. Immunity 2008, 28 (5):630 – 638. 48.TengG,HakimpourP,LandgrafP,RiceA,TuschlT,CasellasR,PapavasiliouFN: MicroRNA-155isanegativeregulatorofactivation-inducedcytidine deaminase. Immunity 2008, 28 (5):621 – 629. 49.DelgadoT,CarrollPA,PunjabiAS,MargineantuD,HockenberyDM, LagunoffM: InductionoftheWarburgeffectbyKaposi ’ ssarcoma herpesvirusisrequiredforthemaintenanceoflatentlyinfected endothelialcells. ProcNatlAcadSciUSA 2010, 107 (23):10696 – 10701. 50.WarburgO: Ontheoriginofcancercells.Science 1956, 123 (3191):309 – 314. 51.WarburgO,PosenerK,NegeleinE: UeberdenStoffwechselderTumoren. BiochemZ 1924, 152: 319 – 344. 52.HaeckerI,GayLA,YangY,HuJ,MorseAM,McIntyreLM,RenneR: Ago HITS-CLIPexpandsunderstandingofKaposi ’ ssarcoma-associated herpesvirusmiRNAfunctioninprimaryeffusionlymphomas. PLoS Pathog 2012, 8 (8):e1002884. 53.RhoadesMW,ReinhartBJ,LimLP,BurgeCB,BartelB,BartelDP: Prediction ofplantmicroRNAtargets. Cell 2002, 110 (4):513 – 520. 54.EnrightAJ,JohnB,GaulU,TuschlT,SanderC,MarksDS: MicroRNAtargets inDrosophila. GenomeBiol 2003, 5 (1):R1. 55.TuK,YuH,HuaY-J,LiY-Y,LiuL,XieL,LiY-X: Combinatorialnetworkof primaryandsecondarymicroRNA-drivenregulatorymechanisms. Nucleic AcidsRes 2009, 37 (18):5969 – 5980. 56.LiangH,LiWH: MicroRNAregulationofhumanproteininteraction network. RNA 2007, 13 (9):1402 – 1408. 57.ShalgiR,LieberD,OrenM,PilpelY: Globalandlocalarchitectureofthe mammalianmicroRNA-transcriptionfactorregulatorynetwork. PLoS ComputBiol 2007, 3 (7):e131. 58.MartinezNJ,OwMC,BarrasaMI,Hamme llM,SequerraR,Doucette-StammL, RothFP,AmbrosVR,WalhoutAJ: AC.elegansgenome-scalemicroRNA networkcontainscompositefeedbackmotifswithhighfluxcapacity. Genes Dev 2008, 22 (18):2535 – 2549. 59.CuiQ,YuZ,PanY,PurisimaEO,WangE: MicroRNAspreferentiallytarget thegeneswithhightranscriptionalregulationcomplexity. Biochem BiophysResCommun 2007, 352 (3):733 – 738. 60.KabnickKS,HousmanDE: Determinantsthatcontributetocytoplasmic stabilityofhumanc-fosandbeta-globinmRNAsarelocatedatseveral sitesineachmRNA. MolCellBiol 1988, 8 (8):3244 – 3250. 61.YeildingNM,RehmanMT,LeeWM:Identificationofsequencesinc-myc mRNAthatregulateitssteady-statelevels. MolCellBiol 1996, 16 (7):3511 – 3522. 62.BossIW,NadeauPE,AbbottJR,YangY,MergiaA,RenneR: AKaposi ’ s sarcoma-associatedherpesvirus-encodedorthologofMicroRNAmiR-155 induceshumansplenicB-cellexpansioninNOD/LtSz-scidIL2R nullmice. JVirol 2011, 85 (19):9877 – 9886. 63.YinQ,WangX,FewellC,CameronJ,ZhuH,BaddooM,LinZ,FlemingtonEK: MicroRNAmiR-155inhibitsbone morphogeneticprotein(BMP) signalingandBMP-mediatedEps tein-Barrvirusreactivation. JVirol 2010, 84 (13):6318 – 6327. 64.RivaA: TheMAPPER2Database:amulti-genomecatalogofputative transcriptionfactorbindingsites. NucleicAcidsRes 2012, 40 (Databaseissue):D155 – D161. 65.NephS,VierstraJ,StergachisAB,ReynoldsAP,HaugenE,VernotB, ThurmanRE,JohnS,SandstromR,JohnsonAK,MauranoMT,HumbertR, RynesE,WangH,VongS,LeeK,BatesD,DiegelM,RoachV,DunnD, NeriJ,SchaferA,HansenRS,KutyavinT,GisteE,WeaverM,CanfieldT, SaboP,ZhangM,BalasundaramG, etal : Anexpansivehuman regulatorylexiconencodedintranscriptionfactorfootprints. Nature 2012, 489 (7414):83 – 90. 66.KiuchiN,NakajimaK,IchibaM,FukadaT,NarimatsuM,MizunoK,HibiM, HiranoT: STAT3isrequiredforthegp130-mediatedfullactivationofthe c-mycgene. JExpMed 1999, 189 (1):63 – 73. 67.LiQ,VermaIM: NF-kappaBregulationintheimmunesystem. NatRev Immunol 2002, 2 (10):725 – 734. 68.Thomas-ChollierM,DefranceM,Medina-RiveraA,SandO,HerrmannC, ThieffryD,vanHeldenJ: RSAT2011:regulatorysequenceanalysistools. NucleicAcidsRes 2011, 39 (WebServerissue):W86 – W91. 69.BaileyTL,BodenM,BuskeFA,FrithM,GrantCE,ClementiL,RenJ,LiWW, NobleWS: MEMESuite:toolsformotifdiscoveryandsearching. Nucleic AcidsRes 2009, 37 (suppl2):W202 – W208. 70.InuiM,MartelloG,PiccoloS: MicroRNAcontrolofsignaltransduction. NatRevMolCellBiol 2010, 11 (4):252 – 263.71.KennellJA,GerinI,MacDougaldOA,CadiganKM: ThemicroRNAmiR-8isa conservednegativeregulatorofWntsignaling. ProcNatlAcadSciUSA 2008, 105 (40):15417 – 15422.Yang etal.BMCGenomics 2014, 15 :668 Page15of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 16

72.MengF,HensonR,Wehbe-JanekH,GhoshalK,JacobST,PatelT: MicroRNA21regulatesexpressionofthePTENtumorsuppressorgeneinhuman hepatocellularcancer. Gastroenterology 2007, 133 (2):647 – 658. 73.ThumT,GrossC,FiedlerJ,FischerT,KisslerS,BussenM,GaluppoP,JustS, RottbauerW,FrantzS,CastoldiM,SoutschekJ,KotelianskyV,RosenwaldA, BassonMA,LichtJD,PenaJT,RouhanifardSH,MuckenthalerMU,TuschlT, MartinGR,BauersachsJ,EngelhardtS: MicroRNA-21contributesto myocardialdiseasebystimulatingMAPkinasesignallinginfibroblasts. Nature 2008, 456 (7224):980 – 984. 74.FishJE,SantoroMM,MortonSU,YuS,YehRF,WytheJD,IveyKN,BruneauBG, StainierDY,SrivastavaD: miR-126regulatesangiogenicsignalingandvascular integrity. DevCell 2008, 15 (2):272 – 284. 75.KuhnertF,MancusoMR,HamptonJ,StankunasK,AsanoT,ChenCZ,KuoCJ: AttributionofvascularphenotypesofthemurineEgfl7locustothe microRNAmiR-126. Development 2008, 135 (24):3989 – 3993. 76.BanerjeeA,SchambachF,DeJongCS,HammondSM,ReinerSL: Micro-RNA155inhibitsIFN-gammasignalinginCD4+Tcells. EurJImmunol 2010, 40 (1):225 – 231. 77.ShuaiK,LiuB: RegulationofJAK-STATsignallingintheimmunesystem. NatRevImmunol 2003, 3 (11):900 – 911. 78.HondaK,YanaiH,NegishiH,AsagiriM,SatoM,MizutaniT,ShimadaN, OhbaY,TakaokaA,YoshidaN,TaniguchiT: IRF-7isthemaster regulatoroftype-Iinterferondependentimmuneresponses. Nature 2005, 434 (7034):772 – 777. 79.WangG,WangY,TengM,ZhangD,LiL,LiuY: Signaltransducersand activatorsoftranscription-1(STAT1)regulatesmicroRNAtranscriptionin interferongamma-stimulatedHeLacells. PLoSOne 2010, 5 (7):e11794. 80.LuF,WeidmerA,LiuCG,VoliniaS,CroceCM,LiebermanPM: Epstein-Barr virus-inducedmiR-155attenuatesNF-kappaBsignalingandstabilizes latentviruspersistence. JVirol 2008, 82 (21):10436 – 10443. 81.DunnJJ,McCorkleSR,EverettL,AndersonCW: Paired-endgenomic signaturetags:amethodforthefunctionalanalysisofgenomesand epigenomes. GenetEng(NY) 2007, 28: 159 – 173. 82.LiangD,GaoY,LinX,HeZ,ZhaoQ,DengQ,LanK: Ahumanherpesvirus miRNAattenuatesinterferonsignalingandcontributestomaintenance ofvirallatencybytargetingIKKvarepsilon. CellRes 2011, 21 (5):793 –806. 83.LinR,GeninP,MamaneY,SgarbantiM,BattistiniA,HarringtonWJ,BarberGN, HiscottJ: HHV-8encodedvIRF-1repressestheinterferonantiviralresponse byblockingIRF-3recruitmentoftheCBP/p300coactivators. Oncogene 2001, 20 (7):800 – 811. 84.JooCH,ShinYC,GackM,WuL,LevyD,JungJU: Inhibitionofinterferon regulatoryfactor7(IRF7)-mediatedinterferonsignaltransductionbythe Kaposi ’ ssarcoma-associatedherpesvirusviralIRFhomologvIRF3. JVirol 2007, 81 (15):8282 – 8292. 85.DamaniaB: ModulationofcellsignalingpathwaysbyKaposi'ssarcomaassociatedherpesvirus(KSHVHHV-8). CellBiochemBiophys 2004, 40 (3):305 – 22. 86.ChengC,AlexanderR,MinR,LengJ,YipKY,RozowskyJ,YanKK,DongX, DjebaliS,RuanY,DavisCA,CarninciP,LassmanT,GingerasTR,GuigR, BirneyE,WengZ,SnyderM,GersteinM: Understandingtranscriptional regulationbyintegrativeanalysisoftranscriptionfactorbindingdata. GenomeRes 2012, 22 (9):1658 – 1667. 87.PaikYK,HancockWS: UnitingENCODEwithgenome-wideproteomics. NatBiotechnol 2012, 30 (11):1065 – 1067. 88.AnFQ,FolarinHM,CompitelloN,RothJ,GersonSL,McCraeKR,FakhariFD, DittmerDP,RenneR: Long-term-infectedtel omerase-immortalized endothelialcells:amodelforKaposi ’ ssarcoma-associatedherpesvirus latencyinvitroandinvivo. JVirol 2006, 80 (10):4833 – 4846. 89.RoyD,SinSH,LucasA,VenkataramananR,WangL,EasonA,ChavakulaV, HiltonIB,TamburroKM,DamaniaB,DittmerDP: mTORinhibitorsblock Kaposisarcomagrowthbyinhibitingessentialautocrinegrowthfactors andtumorangiogenesis. CancerRes 2013, 73 (7):2235 – 2246. 90.DiBartoloDL,CannonM,LiuYF,RenneR,ChadburnA,BoshoffC, CesarmanE: KSHVLANAinhibitsTGF-betasignalingthrough epigeneticsilencingoftheTGF-betatypeIIreceptor. Blood 2008, 111 (9):4731 – 4740. 91.JohnsonRA,WichernDW: AppliedMultivariateStatisticalAnalysis. Englewood,NewJersey:PrenticeHall;1992. 92.BenjaminiY,HochbergY: Controllingthefalsediscoveryrate:apractical andpowerfulapproachtomultipletesting. JRoyalStatSocB(Method) 1995, 57 (1):289 – 300. 93.MoothaVK,LindgrenCM,ErikssonK-F,SubramanianA,SihagS,LeharJ, PuigserverP,CarlssonE,RidderstraleM,LaurilaE,HoustisN,DalyMJ, PattersonN,MesirovJP,GolubTR,TamayoP,SpiegelmanB,LanderES, HirschhornJN,AltshulerD,GroopLC: PGC-1[alpha]-responsivegenes involvedinoxidativephosphorylati onarecoordinatelydownregulatedinhumandiabetes. NatGenet 2003, 34 (3):267 – 273. 94.SubramanianA,TamayoP,MoothaVK,MukherjeeS,EbertBL,GilletteMA, PaulovichA,PomeroySL,GolubTR,LanderES,MesirovJP: Geneset enrichmentanalysis:Aknowledgebasedapproachforinterpreting genome-wideexpressionprofiles. ProcNatlAcadSciUSA 2005, 102 (43):15545 – 15550. 95.RitchieW,FlamantS,RaskoJEJ: PredictingmicroRNAtargetsand functions:trapsfortheunwary. NatMeth 2009, 6 (6):397 – 398. 96.BetelD,WilsonM,GabonA,MarksDS,SanderC: ThemicroRNA.org resource:targetsandexpression. NucleicAcidsRes 2008, 36 (Databaseissue):D149 – D153. 97.KerteszM,IovinoN,UnnerstallU,GaulU,SegalE: Theroleofsiteaccessibility inmicroRNAtargetrecognition. NatGenet 2007, 39 (10):1278 – 1284. 98.KiriakidouM,NelsonPT,KouranovA,FitzievP,BouyioukosC,MourelatosZ, HatzigeorgiouA: Acombinedcomputational-experimentalapproach predictshumanmicroRNAtargets. GenesDev 2004, 18 (10):1165 – 1178. 99.WangX: miRDB:amicroRNAtargetpredictionandfunctionalannotation databasewithawikiinterface. RNA 2008, 14 (6):1012 – 1017. 100.MirandaKC,HuynhT,TayY,AngYS,TamWL,ThomsonAM,LimB, RigoutsosI: Apattern-basedmethodfortheidentificationofMicroRNA bindingsitesandtheircorrespondingheteroduplexes. Cell 2006, 126 (6):1203 – 1217. 101.DweepH,StichtC,PandeyP,GretzN: miRWalk – database:predictionof possiblemiRNAbindingsitesby “ walking ” thegenesofthreegenomes. JBiomedInform 2011, 44 (5):839 – 847. 102.GaidatzisD,vanNimwegenE,HausserJ,ZavolanM: InferenceofmiRNA targetsusingevolutionaryconservationandpathwayanalysis. BMC Bioinformatics 2007, 8: 69. 103.KrekA,GrunD,PoyMN,WolfR,RosenbergL,EpsteinEJ,MacMenaminP, daPiedadeI,GunsalusKC,StoffelM,RajewskyN: CombinatorialmicroRNAtargetpredictions. NatGenet 2005, 37 (5):495 – 500. 104.BartonicekN,EnrightAJ: SylArray:awebserverforautomated detectionofmiRNAeffectsfromexpressiondata. Bioinformatics 2010, 26 (22):2900 – 2901. 105.SandelinA,AlkemaW,EngstrmP,WassermanWW,LenhardB: JASPAR:an open accessdatabaseforeukaryotictranscriptionfactorbindingprofiles. NucleicAcidsRes 2004, 32 (suppl1):D91 – D94. 106.VaquerizasJM,KummerfeldSK,TeichmannSA,LuscombeNM: Acensusof humantranscriptionfactors:function,expressionandevolution. NatRev Genet 2009, 10 (4):252 – 263. 107.MarinescuVD,KohaneIS,RivaA: TheMAPPERdatabase:amulti-genome catalogofputativetranscriptionfactorbindingsites. NucleicAcidsRes 2005, 33 (Databaseissue):D91 – D97. 108.JiangC,XuanZ,ZhaoF,ZhangMQ: TRED:atranscriptionalregulatory elementdatabase,newentriesandotherdevelopment. NucleicAcidsRes 2007, 35 (suppl1):D137 – D140. 109.SchaeferCF,AnthonyK,KrupaS,BuchoffJ,DayM,HannayT,BuetowKH: PID:thepathwayinteractiondatabase. NucleicAcidsRes 2009, 37 (Databaseissue):D674 – D679. 110.BensonDA,Karsch-MizrachiI,LipmanDJ,OstellJ,WheelerDL: GenBank. NucleicAcidsRes 2005, 33: D34 – D38. 111.BaderGD,HogueCWV: BIND – adataspecificationforstoringand describingbiomolecularinteractions,molecularcomplexesand pathways. Bioinformatics 2000, 16 (5):465 – 77. 112.BaderGD,DonaldsonI,WoltingC,OuelletteBFF,PawsonT,HogueCWV: BIND – TheBiomolecularInteractionNetworkDatabase. NucleicAcidsRes 2001, 29 (1):242 – 245. 113.KarpPD,RileyM,SaierM,PaulsenIT,PaleySM,Pellegrini-TooleA: The EcoCycandMetaCycdatabases. NucleicAcidsRes 2000, 28 (1):56 – 59.114.PeriS,NavarroJD,KristiansenTZ,AmanchyR,SurendranathV,Muthusamy B,GandhiTKB,ChandrikaKN,DeshpandeN,SureshS,RashmiBP,ShankerK, PadmaN,NiranjanV,HarshaHC,TalrejaN,VrushabendraBM,RamyaMA, YatishAJ,JoyM,ShivashankarHN,KavithaMP,MenezesM,ChoudhuryDR, GhoshN,SaravanaR,ChandranS,MohanS,JonnalagaddaCK,PrasadCK, etal : Humanproteinreferencedatabaseasadiscoveryresourcefor proteomics. NucleicAcidsRes 2004, 32 (suppl1):D497 – D501.Yang etal.BMCGenomics 2014, 15 :668 Page16of17 http://www.biomedcentral.com/1471-2164/15/668

PAGE 17

115.ClineMS,SmootM,CeramiE,KuchinskyA,LandysN,WorkmanC,ChristmasR, Avila-CampiloI,CreechM,GrossB,HanspersK,IsserlinR,KelleyR,KillcoyneS, LotiaS,MaereS,MorrisJ,OnoK,PavlovicV,PicoAR,VailayaA,WangPL, AdlerA,ConklinBR,HoodL,KuiperM,SanderC,SchmulevichI,SchwikowskiB, WarnerGJ, etal : Integrationofbiologicalnetworksandgeneexpressiondata usingCytoscape. NatProtocols 2007, 2 (10):2366 – 2382. 116.ShannonP,MarkielA,OzierO,BaligaNS,WangJT,RamageD,AminN, SchwikowskiB,IdekerT: Cytoscape:asoftwareenvironmentfor integratedmodelsofbiomolec ularinteractionnetworks. GenomeRes 2003, 13 (11):2498 – 504.doi:10.1186/1471-2164-15-668 Citethisarticleas: Yang etal. : Asystemsbiologyapproachidentified differentregulatorynetworkstargetedbyKSHVmiR-K12-11inBcells andendothelialcells. BMCGenomics 2014 15 :668. 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 Yang etal.BMCGenomics 2014, 15 :668 Page17of17 http://www.biomedcentral.com/1471-2164/15/668



PAGE 1

1 Table S1. Gene sets retri e ved from Geo or ArrayExpress using search terms K12 list Accession Gene # Description Reference 1 GSE13296 45 regulated genes were identified in LPS activated moDCs after miR 155 knockdown [1] 2 GSE14477 261 changed genes after overexpress microRNA 155 in lung fibrobl asts [2] 3 GSE10467 10 genes regulated by mi R 155 in a mouse macrophage cell line [3] 4 GSE10863, GSE10864, GSE10868 78 genes expressed more than twofold lower or twofold higher in miR 155 expressing cells [4] 5 GSE8867 64 genes respond ing to miR K12 11 overexpression in BJAB cells [5] 6 GSE9264 66 genes changed by both miR 155 and miR K12 11 in 293 cells [6]

PAGE 2

2 Table S2. Algorithms for miRNA target prediction Algorithm Criteria for Prediction and Ranking Website Reference TargetScan Stringent seed pairing, site number, site type, site context (which includes factors that influence site accessibility); option of ranking by likelihood of preferential conservation rather than site context http://targetscan.org [7] EMBL Stringent seed pairing, site number, overall predicted pairing stability http://russell.embl heidelberg.de [8] PicTar Stringent seed pairing for at least one of the sites for the miRNA, site number, overall predicted pairing stability http://pictar.mdc berlin.de [9] EIMMo Stringent seed pairing, site number, likelihood of preferential conservation http://www.mirz.uniba s.ch/ElMMo2 [10] Miranda Moderately stringent seed pairing, site number, pairing to most of the miRNA http://www.microrna.o rg [11] miRBase Targets Moderately stringent seed pairing, site number, overall pairing http://microrna.sanger. ac.uk [12] PITA Moderat ely stringent seed pairing, site number, overall predicted pairing stability, predicted site accessibility http://genie.weizmann. ac.il/pubs/mir07/mir07 _data.html [13] mirWIP Moderately stringent seed pair ing, site number, overall predicted pairing stability, predicted site accessibility http://146.189.76.171/ query [14]

PAGE 3

3 RNA22 Moderately stringent seed pairing, ma tches to sequence patterns generated from miRNA set, overall predicted pairing and predicted pairing stability http://cbcsrv.watson.ib m.com/rna22.html [15] RNAhybrid thermodynamic stability, Moderately stringent seed pairing http://bibiserv.techfak. uni bielefeld.de/rnahybrid/ [16] Targetboost Moderately stringent seed pairing; site number, conservation; thermodynamic stability http://www.in teragon. com/demo/ [17] Table S3 Enrichment pathways and associated genes. Genes in bold are also putative direct targets of miR K12 11. Biological process Genes in TIVE Genes in BJAB IFN HILA A, HLA B, HLA C, HLA DMA, HLA F, HLA G, IFI30, IRF1, IRF7, IRF9, OAS1, OAS2 OAS3, OASL, SOCS1 STAT1 CAMK2G, HLA C, HLA E, IFNGR1 IRF3, IRF9, PTPN6, SP100 Response to glucose stimulus / carbohydrate metabolic process AKR7A2 B4GALT1, CS, GALT, GBA, GOT1, GYG1, NPL, NUP160, NUP43, PFKFB2 PFKFB3, PGK1, PRPS1 CTSB, EP300 PFKL, RHOC, SREBF1, TCF7L2 UCP2

PAGE 4

4 induction of apoptosis / Regulation of apoptotic process ACSL5 APP, ATPIF1, BAD, BCL2A1, BEX2, BIRC5, BTG1,CASP3, CASP9 CD70, CEBPB DEDD, DUSP1 ERN1 FEM1B, FOXO3 GCH1 HIP1, IRF1, JMY KLF10, MAPK1, MUL1 MX1, NACC1, NDUFA13, NDUFS3, NUDT2, PRMT2, PRMT2, PRDX2 PSMB2, PSMB3 PSMB6, PSMB8 PSMB9, PSMC3, PSMD13, PSMD8, PSMD13, PSMG2, PTEN RNF41, SKI, SKIL, STAT1, STK17A STK17B, USP7 -Figure S1 Different components of the same IFN pathway were targeted in TIVE and BJAB cells. Green : unchanged ; Blue: up regulated; Pink: down regulated; Pink boxes with red words: down regulated genes that are potential direct targets. Up: in BJAB cells, the cytokine receptor may be directly targeted by miR K12 11, leading to

PAGE 5

5 reduced levels of downstr eam factors. Down: in TIVE cells, the transcription factor STAT and AKT are directly targeted, amplifying the effect to a large set of genes. Figure S2 Enrichment of down regulated genes in the neighboring genes of CASP9 centered network of protein in teraction.

PAGE 6

6 Figure S3 Connectivity of human protein protein interactions. The distribution follows the power law. Few proteins have many neighbors, while most genes are sparsely connected.

PAGE 7

7 1. Ceppi M, Pereira PM, Dunand Saut hier I, Barras E, Reith W, Santos MA, Pierre P: MicroRNA 155 modulates the interleukin 1 signaling pathway in activated human monocyte derived dendritic cells Proc Natl Acad Sci U S A 2009, 106 (8):2735 2740. 2. Pottier N, Maurin T, Chevalier B, Puissgur MP, Lebrigand K, Robbe Sermesant K, Bertero T, Lino Cardenas CL, Courcot E, Rios G et al : Identification of keratinocyte growth factor as a target of microRNA 155 in lung fibroblasts: implication in epithelial mesenchymal interactions PLoS One 2009, 4 (8): e6718. 3. O'Connell RM, Rao DS, Chaudhuri AA, Boldin MP, Taganov KD, Nicoll J, Paquette RL, Baltimore D: Sustained expression of microRNA 155 in hematopoietic stem cells causes a myeloproliferative disorder J Exp Med 2008, 205 (3):585 594. 4. Yin Q, McBride J, Fewell C, Lacey M, Wang X, Lin Z, Cameron J, Flemington EK: MicroRNA 155 is an Epstein Barr virus induced gene that modulates Epstein Barr virus regulated gene expression pathways J Virol 2008, 82 (11):5295 5306. 5. Gottwein E, Mukherjee N, Sachse C, Frenzel C, Majoros WH, Chi JT, Braich R, Manoharan M, Soutschek J, Ohler U et al : A viral microRNA functions as an orthologue of cellular miR 155 Nature 2007, 450 (7172):1096 1099. 6. Skalsky RL, Samols MA, Plaisance KB, Boss IW, Riva A, Lopez MC, Baker HV, Renne R: Kaposi's sarcoma associated herpesvirus encodes an

PAGE 8

8 ortholog of miR 155 Skalsky RL, Samols MA, Plaisance KB, Boss IW, Riva A, Lopez MC, Baker HV, Renne R 2007, 81 (23):12836 12845. 7. Friedman RC, Farh KK, Burge CB, Bartel DP: Most mammalian mRNAs are conserved targets of microRNAs Genome Res 2009, 19 (1):92 105. 8. Stark A, Brennecke J, Bushati N, Russell RB, Cohen SM: Animal MicroRNAs confer robustness to gene expression and have a significant impact on 3'UTR evolution Cell 2005, 123 (6):1133 1146. 9. Lall S, Grn D, Krek A, Chen K, Wang Y L, Dewey CN, Sood P, Colombo T, Bray N, MacMenamin P et al : A Genome Wide Map of Conserved MicroRNA Targets in C. elegans Current Biology 16 (5):460 471. 10. Gaidatzis D, van Nimwegen E, Hausser J, Zavolan M: Inference of miRNA targets using evolutionary conservation and pathway analysis BMC Bioinformatics 2007, 8 :69. 11. Betel D, Wilson M, Gabon A, Marks DS, Sander C: The microRNA.org res ource: targets and expression Nucleic Acids Res 2008(36 Database):D149 D153. 12. Griffiths Jones S: The microRNA Registry Nucleic Acids Res 2004, 32 (Database issue):D109 111. 13. Kertesz M, Iovino N, Unnerstall U, Gaul U, Segal E: The role of site acce ssibility in microRNA target recognition Nat Genet 2007, 39 (10):1278 1284. 14. Hammell M, Long D, Zhang L, Lee A, Carmack CS, Han M, Ding Y, Ambros V: mirWIP: microRNA target prediction based on microRNA containing ribonucleoprotein enriched transcripts Nat Methods 2008, 5 (9):813 819. 15. Miranda KC, Huynh T, Tay Y, Ang YS, Tam WL, Thomson AM, Lim B, Rigoutsos I: A pattern based method for the identification of MicroRNA binding sites and their corresponding heteroduplexes Cell 2006, 126 (6):1203 1217. 16. Krger J, Rehmsmeier M: RNAhybrid: microRNA target prediction easy, fast and flexible Nucleic Acids Res 2006, 34 (Web Server issue):W451 454. 17. Saetrom O, Snove O, Jr., Saetrom P: Weighted sequence motifs as an improved seeding step in microRNA target p rediction algorithms Rna 2005, 11 (7):995 1003.