Transcriptional, post-transcriptional and chromatin-associated regulation of pri-miRNAs, pre-miRNAs and moRNAs 
Chirag Nepal1,*, Marion Coolen2, Yavor Hadzhiev3, Delphine Cussigh2, Piotr Mydel4,5, Vidar 
M. Steen6, Piero Carninci7, Jesper B. Andersen1, Laure Bally-Cuif2, Ferenc M ¨
uller3,* and Boris Lenhard8,9,* 
1Biotech Research and Innovation Centre, University of Copenhagen, Ole Maales Vej 5, DK-2200 Copenhagen N, Denmark, 2Zebrafsh Neurogenetics Team, Paris-Saclay Institute of Neuroscience, CNRS UMR9197 – Universite´Paris Sud, 91198 Gif-sur-Yvette, France, 3School of Clinical and Experimental Medicine, College of Medical and Dental Sciences, University of Birmingham, Edgbaston B15 2TT, UK, 4Broegelmann Research Laboratory, Department of Clinical Science, University of Bergen, Bergen, Norway., 5Malopolska Centre of Biotechnology, Jagiellonian University, Krakow, Poland, 6Department of Clinical Medicine, University of Bergen, Norway, 7Division of Genomic Technologies, RIKEN Center for Life Science Technologies, Yokohama, Kanagawa 230-0045, Japan, 8Institute of Clinical Sciences MRC Clinical Sciences Centre, Faculty of Medicine, Imperial College London, Hammersmith Hospital Campus, Du Cane Road, London W12 0NN, UK and 9Department of Informatics, University of Bergen, Thormhlensgate 55, N-5008 Bergen, Norway 
Received May 26, 2015; Revised November 18, 2015; Accepted November 20, 2015 
ABSTRACT 
MicroRNAs (miRNAs) play a major role in the post-transcriptional regulation of target genes, especially in development and differentiation. Our understanding about the transcriptional regulation of miRNA genes is limited by inadequate annotation of pri-mary miRNA (pri-miRNA) transcripts. Here, we used CAGE-seq and RNA-seq to provide genome-wide identifcation of the pri-miRNA core promoter reper-toire and its dynamic usage during zebrafsh em-bryogenesis. We assigned pri-miRNA promoters to 152 precursor-miRNAs (pre-miRNAs), the majority of which were supported by promoter associated post-translational histone modifcations (H3K4me3, H2A.Z) and RNA polymerase II (RNAPII) occupancy. We validated seven miR-9 pri-miRNAs by in situ hybridization and showed similar expression patterns as mature miR-9. In addition, processing of an al-ternative intronic promoter of miR-9–5 was validated by 5 RACE PCR. Developmental profling revealed a subset of pri-miRNAs that are maternally inherited. Moreover, we show that promoter-associated H3K4me3, H2A.Z and RNAPII marks are not only present at pri-miRNA promoters but are also specifcally enriched at pre-miRNAs, suggesting chromatin level regulation of pre-miRNAs. Furthermore, we demonstrated that CAGE-seq also detects 3-end pro-cessing of pre-miRNAs on Drosha cleavage site that correlates with miRNA-offset RNAs (moRNAs) pro-duction and provides a new tool for detecting Drosha processing events and predicting pre-miRNA pro-cessing by a genome-wide assay. 
INTRODUCTION 
MicroRNAsareaclassofsmall(∼22nucleotide(nt))regulatorynon-codingRNAsfoundinplantsandanimals.ThesesmallRNAsareprocessedfromlargemiRNApri-marytranscripts(pri-miRNAs)into70∼90ntprecursors(pre-miRNAs)andfurtherintomaturemiRNA(1,2).PrimiRNAsaregenerallytranscribedbyRNApolymeraseII(RNAPII)(3),althoughafewmiRNAsaretranscribedbyRNAPIII(4).Rapidprocessingandrelativelylowcopynumberhavehinderedidentificationandcharacterizationofpri-miRNA,limitingthestudiestoonlyahandfulofexamples(3,5,6).However,highthroughputsequencingtechnologieshavefacilitatedannotationofpri-miRNAsonagenome-widescaleinhumanandmouse(7–12).Pri<BR>miRNApromotersarelocatedanywherefromaboutahun-dredofbasestoseveralkilobases(KB)upstreamofthe
Downloaded from https://academic.oup.com/nar/article-abstract/44/7/3070/2467725 by Jagiellonian University user on 07 June 2019 
*Towhomcorrespondenceshouldbeaddressed.Tel:+442083838353;Fax:+442083838577;Email:b.lenhard@imperial.ac.ukCorrespondencemayalsobeaddressedtoFerencM¨
uller.Tel:+441214142895;Email:f.mueller@bham.ac.ukCorrespondencemayalsobeaddressedtoChiragNepal.Tel:+4535335031;Email:chirag.nepal@bric.ku.dk
CTheAuthor(s)2015.PublishedbyOxfordUniversityPressonbehalfofNucleicAcidsResearch.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommonsAttributionLicense(http://creativecommons.org/licenses/by/4.0/),whichpermitsunrestrictedreuse,distribution,andreproductioninanymedium,providedtheoriginalworkisproperlycited.
pre-miRNA.Initialstudiesdetectedindependentpromotersonlyforafewintronicpre-miRNAs(i.e.miRNAswhosepre-miRNAisfoundintheintronofknownprotein-codinggenes)(7–11).However,asubsequentstudyshowedmanyintronicpre-miRNAshaveindependentpromotersthatex-hibittissue-specificregulation(13).Interestingly,somepre<BR>miRNAcanalsobetranscribedlocallybytranscriptioninitiationattheir5-ends(∼50basesfrommaturemiRNA)(14).
Transcriptionofpri-miRNAtranscriptsandpre-miRNAcleavagearenucleareventsthatarefacilitatedbymicroprocessorcomponentsDrosha/DGCR8(15).Ithasbeenpro-posedthatpre-miRNAsareprocessedco-transcriptionally(16,17).Thedetailsoftargetingandregulationofthiscleav-agearestillunclear,althoughitisbelievedtoinvolveregulatoryinformationembeddedintheprimarytranscripts(16).Flankingsingle-strandedRNAsegmentslocatedoutsidethepre-miRNAhairpinarecrucialforprecisecleavageofthestemloopandmaturationofmiRNA(18,19).Couplingbetweenpri-miRNAtranscriptionandDroshaprocessingpromotesefficientexpressionoftranscribedmiRNAs(20).Sequencemotifsassociatedwithco-transcriptionalcleav-ageremainelusive,thusitmaybemediatedviainterplaybetweentranscriptionandchromatinregulation.Genomewideanalysisofnucleosomepositioninghasrevealednu-cleosomesarepositionedalongpre-miRNAs(8).Analysesofpre-miRNAprocessinghaverevealedadistinctclassofsmallRNAs(miRNA-offsetRNAs(moRNAs))originatedfromthesequenceflanking5-or3-armsofpre-miRNAs(21,22).AlthoughtheexactbiogenesisofmoRNAsremainsunclear,theymightrepresentRNAproductsgeneratedbyDroshaduringpre-miRNAprocessing(21).
Thepredictionofpri-miRNApromotersremainsnoto-riouslydifficultduetotherapidandefficientnatureofDrosha/DGCR8processing.Thoughpri-miRNApromotershavebeenmappedinhumanandmousetovariousex-tents(7–13),inzebrafish––animportantvertebratemodelorganism––pri-miRNApromotersremainlargelyunannotated.Theavailabilityofcomprehensiveannotationofze-brafishpri-miRNAswillprovideavaluableresourcetostudymechanisticinsightsonpri-miRNAregulationdur-ingembryonicdevelopment.Recentfindingsreporttheex-istenceofbothdistalpromotersonhostgenesandlo-calpromotersatpre-miRNAs(14),suggestingalternativemechanismsofhowmiRNAsmaybeproduced.CAGE-seqtechnologycandetectprimaryandalternativepromoters(23,24)andalsodetectavarietyofprocessedRNAs(24,25).Inthispaper,weexploredtheutilityofthistechnologytostudytranscriptionalandpost-transcriptionalpro-cessingofpri-miRNAsduringdevelopment.Wehavecom-prehensivelyannotatedzebrafishpri-miRNAstranscriptsandunraveledprocessingeventsandtheirutilizationdur-ingearlyembryogenesis.WeusedCAGE-seqdefinedtran-scriptionstartsite(TSS)mapsatsinglenucleotideresolutiontoidentifypri-miRNApromoters.Furthermore,wedemonstratedthatourpredictedpri-miRNAsarebonafidetranscriptsbyvalidatingsevenparalogouscopiesofmiR9,bydescribingtheirprocessingandbyidentifyingtheirexpressionpatterns.Moreover,weshowthatpromoterassociatedH3K4me3,H2A.ZandRNAPIImarksarenotonlypresentatpri-miRNApromotersbutarealsospecificallyenrichedatpre-miRNAs,suggestingchromatinlevelregulationofpre-miRNAs.Wereportnovelinsightaboutpost-transcriptionalprocessingofpre-miRNAhairpinsthatcorrelatewiththepresenceofmiRNA-offsetRNAproduction.
MATERIALS AND METHODS 
Datasets
ProcessedCAGEtagsfrom12developmentalstages,RNA-seqdatafromfourdevelopmentalstagesandpost-translationhistonemodificationdata(H3K4me3(fourde-velopmentalstages)andH2A.Z)wereproducedinourre-centstudy(24,26).RawsequencereadsofRNAPII(27)andsmallRNA-seqdata(28)weredownloadedfromSRA.DetailedinformationregardingmappingandstatisticsofanalyzeddataisdescribedinSupplementaryTableS1.Sequencereadsweremappedtothezebrafishgenomeassembly(Zv9).Thegenomiccoordinatesof346pre-miRNAhairpinsweredownloadedfrommiRBase-V20(29).Weexcluded25unexpressedpre-miRNAsandre-tained321pre-miRNAsthatareexpressedabove1tagpermillion(tpm).BasedonEnsemblannotation(codingandnon-codinghostgenes),overlappingpre-miRNAswereclassifiedasintragenic,andremaining217(including59copiesofmiR-430incluster)pre-miRNAsasin-tergenic.Ensemblannotatedpre-miRNAs(30)weredown<BR>loaded,andweexcludedpre-miRNAsresidingoncodingexonsandthoseannotatedinmiRBase-V20,leaving227pre-miRNAs.CpGislands(CGIs)weredownloadedfromUCSCgenomebrowserdatabase(31)andin-housecustomannotatedCGIs.MappedsmallRNAandhistonemodification(alignedBAMfilesandpeakcalling)datafromh1ESCcelllinesweredownloadedfromENCODE(32).
CAGEtranscriptionstartsites
ACAGEtagmappedtothegenomeiscalledCAGEtran-scriptionstartsite(CTSS).ThefirstbaseofCAGEtagde-finestheTSSandthenumberofCAGEtagsmappedtothatnucleotidegivesquantitativemeasureofCTSS,whichismeasuredastpm.ProximalCTSSswithin20ntapart,onthesamestrand,areclusteredastranscriptclusters(TCs)(SupplementaryFigureS1A),whereTSSwithhighestexpressionvaluesdefinestherepresentativeCTSSs.ThewidthofTCanddistributionofCAGEtagsisusedtoclassifysharpandbroadpromoters.Initiatorsequenceisdefinedbythefirstbase(+1)ofCAGEtagandtheimmediateupstreambase(−1).Canonicalinitiatorsaredenotedbypyrimidineandpurineat−1,+1position.Aclassofnoncanonicalinitiatorsassociatedwithpost-transcriptionalprocessingaredenotedby‘Guanine’at−1,+1position(24).CAGETCswerefurtherclassifiedastranscriptional-associatedandpost-transcriptionalassociatedbasedoninitiatorsequence.Wemeasuredtheexpressionlevelofpri-miRNAsbysummingtheexpressionofallCAGETCsinthedefinedpromoterregion(±500bases).
Identificationofpri-miRNApromoters
WesystematicallyanalyzedallCAGETCs(≥0.5tpm)intheupstreamregions(upto40kB)ofintergenicpre-miRNAs,iterativelybyincreasingupstreamdistancewithstepsizeof5kB.However,ifannotatedcodinggenesareencounteredwithin40kB,theupstreamregionwasextendedupto3-endsofgenes.WefirstcompiledallCAGETCsintheas-signedupstreamregionandexcludedpost-transcriptionallyassociatedCAGETCs.OnremainingCAGETCs,ween-suredtheimmediateupstreamCAGETCswerenotanno-tated(bydefault)aspri-miRNApromoters.WeanalyzedallCAGETCsinthenextupstreamregionsandpriori-tizedtheirassociationwithpre-miRNAsbasedonRNAseq/ESTandH3K4me3evidence.IfonlyasingleCAGETCwasdetected,itwasassignedaspri-miRNApromoter.IfmultipleCAGETCsweredetected,weusedRNA-seq/ESTtranscriptstodeterminetheCAGETCsthatareassociatedwithfull-lengthtranscriptsorhaveconcordant5-ends.CAGETCswithoutevidenceofRNA-seq/ESTtranscriptsweresearchedforthepresenceofenrichedH3K4me3signals.CAGETCswithoutRNA-seq/ESTandH3K4me3evidencewereprioritizedbasedonhighexpressionlevelofCAGETCs.Pre-miRNAswithoutanyassignedpromoterwithin40kbupstreamwereexcludedasundetectablean-notatedpri-miRNAs.Afterannotationofpri-miRNAs,weanalyzedintronicCAGETCstoidentifyalternativepro-moters,wheresuchCAGETCshadtobeatleast±500basesawayfromassignedpri-miRNAs.ForintragenicpremiRNAsresidingonanon-codinghost,theannotatedtran-scriptswereassignedaspri-miRNAtranscripts.ForpremiRNAsresidingonacodinghost,upstreamregionswereextendeduptothe5-endsofhostgenes,andexaminedforevidenceofindependenttranscriptionthataredistinctfromhostpromoters.Forpre-miRNAswithundetectedal-ternativepromoters,itremainsunclearwhethertheyareco-transcribedbyhostgenesorhaveindependentpromotersthatremainundetectedinourdatasets.However,ifpremiRNAsarelocatedclose(<5kB)tothe5-endsofhostgenes,andhadnoevidenceofindependenttranscription,weconsideredthemtobetranscribedbyhostgene,resultingin13suchpre-miRNAs.

Micro-RNAoffsetRNAs
Micro-RNAoffsetRNAs(moRNAs)aregeneratedadja-centto5armand3armofmaturemiRNAs.Weinter-sectedCAGEtagswithrespecttoannotatedpre-miRNAsandidentifiedasubsetofpre-miRNAsthathaveCAGEtags.TodefineaCAGE-basedmoRNA,weensuredthatCAGEtagsoriginatedadjacent(1baseoffset)tothema-turemiRNAs,weredetectedinatleasttwodevelopmentalstagesandtranscribedfromthesamenucleotide.
Bioinformaticstoolsandanalyses
AllbioinformaticsanalyseswereperformedusingPerl,R/Bioconductor,bashscriptsandBedtoolssuite(33).Se-quencelogosweredrawnusingWebLogo(34).SmallRNA-seqreadsweremappedusingbowtie(35)withdefaultpa-rameters.
RNAextraction,in situ hybridizationandRACEPCR
TotalRNAfrom24hpost-fertilization(hpf)embryoswasextractedusingTrizol(Ambion),treatedwithDNaseI(Promega)andreversetranscribedwithSuperscriptIIreversetranscriptase(Invitrogen)usingrandomhexamerprimers.Sevenparalogouscopiesofassembledprimarytranscripts(describedbelow)ofmiR-9wereamplifiedbyPCRusingthefollowingprimers(SupplementaryTableS2),andclonedusingStrataclonePCRcloningkit(Agilenttech-nologies).5RACEPCRexperimentswereperformedusingthe5RACEsystemforrapidamplificationofcDNAends(Invitrogen),usingthelistedprimers(SupplementaryTableS2).AntisenseDIGlabeledRNAprobesforeachmiR-9transcriptsweresynthesizedandinsituhybridizationper-formedusingstandardprotocols(36).
RESULTS 
Highthroughputidentificationofpri-miRNApromoters
ToannotateTSSsofzebrafishpri-miRNApromoters,wesystematicallyanalyzedCAGEtagsintheupstreamregion(seeMaterialsandMethods)of321pre-miRNAsthatareexpressedduringearlyembryogenesis.Adoptingthecriteriafrompreviousstudies(7,8,13),weanalyzedupto40kBupstreamofpre-miRNAsasthelikelyregiontoencodeprimiRNAs(Figure1A).Weidentifiedpri-miRNApromot<BR>ersfor152(92intergenicand60intragenic)pre-miRNAsintheassignedupstreamregion(SupplementaryTableS3).Outof60intragenicpre-miRNAs,31pre-miRNAsresideonannotatednon-codinghosttranscriptsandtheremaining29pre-miRNAsareinsideproteincodinggenes.Outofthese29pre-miRNAs,weidentifiedindependentpromotersfor16pre-miRNAs,whiletheremaining13pre-miRNAslieproximalto5-endsofproteincodinggenes;thus,hostgenepromoterswereannotatedasthepri-miRNApro-moter.Asanillustrativeexample,primarytranscriptoftheintergenicmiR-17–92clusterisshownwithfull-lengthtranscript,CAGEtagsandH3K4me3peaks(Figure1B).CAGE-seqdefinedTSSislocated1.1KBupstreamfromthefirstpre-miRNA,wheretherepresentativeTSSisdeterminedbythepositionofdominantCAGEtagswithinthetagcluster(SupplementaryFigureS1A).Sixpre-miRNAsofthemiR-17–92clusterareencodedbyasinglepri-marytranscript,asreportedinhuman(6).Countingthe59copiesofmiR-430inaclusterasasinglegene,wean-notatedpri-miRNATSSfor∼58%oftotalpre-miRNAs(Figure1C).Wewereunabletodetectpri-miRNApro-motersfor55intergenicpre-miRNAs(excludingthemiR430cluster)and42intronicpre-miRNAsresidinginproteincodinghostgenes.Lackofevidenceforindependenttranscriptionofintragenicpre-miRNAssuggeststhattheyareeitherco-transcribedwiththeirhostgenesor,alternativelythatintragenicpre-miRNAsmighthavetissuespe-cificpri-miRNAs(13)thatremainundetectedintheana-lyzeddatasets.However,thiscannotbeconfirmedwithoutadditionaldata.Thelackofpri-miRNATSSsfor56in-tergenicpre-miRNAsmightbeduetooneofseveralrea-sons:(i)tissue-specificorlowexpressionthatremainsundetectableinourdatasetor(ii)transcribedbyPolIII(4)thatisundetectedbyCAGE-seq.Evenso,thedepthoftheanno-tatedpri-miRNAsiscomparabletopreviousstudiesinhumanandmouse(7,8).Wealsoanalyzedtheregionbetweentheannotatedpri-miRNApromotersandpre-miRNAsforpotentialalternativepromoters,andidentifiedadditional

Figure1.MappingofzebrafishmiRNAprimarytranscriptsduringembryonicdevelopment.(A)Schematicrepresentationoftheapproachusedtopredictpri-miRNAtranscripts(seeMaterialsandMethods).(B)Agenomebrowserviewofpri-miRNAtranscriptofintergenicmiR-17–92cluster.Full-lengthtranscriptisreconstructedfromRNA-seqandthepromotertranscriptionstartsiteisdefinedbyCAGE-seq.H3K4me3-modifiednucleosomesat5-endsprovidesadditionalevidenceforbeingatruepromoter.Maternalandzygotictranscriptomestagesarerepresentedbyverticalbarsinblueandred,respectively.(C)Distributionofpre-miRNAswithannotatedpri-miRNAs(red),otherpre-miRNAswithoutannotatedpri-miRNAs(green)andmiR-430pre-miRNAs(blue).(D)Distanceofpri-miRNATSSsrelativetopre-miRNA.(E)Numberofpri-miRNAsTSSsdefinedbyCAGE-seqthataresupportedbyH3K4me3peaksandRNA-seq/ESTtranscripts.(F–H)Alignmentofpromoter-associatedhistonemodifications(H3K4me3,H2A.Z)andRNAPIIsupportstheannotatedpri-miRNAs5-ends.Y-axisshowsthenormalizedreadcountintagspermillion(tpm).
CAGETCsassociatedwith47pre-miRNAs(Supplemen-taryTableS3),butthoseTCshadlowexpressionlevelscom-paredwiththemainidentifiedpromoter.TheTSSsofpredictedpri-miRNAsarelocatedfromafewhundredbasesuptotensofkBupstreamofpre-miRNAs(Figure1D),simi-lartopreviousobservationsinhumanandmouse(7,8).Themajorityofannotatedpri-miRNAspromotershaveadditionalevidenceofpromoteractivityfromcomplementarygenome-widedata(RNA-seq/ESTandH3K4me3)(Figure1E).Pri-miRNApromoterregionsof123pre-miRNAshadconcordant5-endswithRNA-seq/EST(87and118respectively)transcriptsand109pre-miRNAshadrobustH3K4me3peaks(Figure1E).Promoter-associatedsignalssuchasH3K4me3-modifiednucleosomes,histonevariantH2A.ZandRNAPIIoccupancyareallenricheddirectlydownstreamofaggregatedpredictedpri-miRNA5-ends(Figure1F–H),furthersupportingpredictionofgenuinepromotersbyCAGE-seq.
Validationofpri-miRNAstranscriptionstartsiteregionsandpre-miRNAprocessingsitesofthemiR-9family
Tovalidatetheannotatedpri-miRNApromoters,wese-lectedthemiR-9familywithsevenparalogouscopies.ThemiR-9familyishighlyconservedandacrucialregulatorofembryonicneurogenesisinvertebrates(37).Inzebrafishembryos,miR-9expressionisdetectedaround20hpf(30somitesstage)inthetelencephalon,andlaterspreadstoamoreposteriorregionofthebrain(38).DistinctCAGEtagsmappedtothepromoterofannotatedpri-miRNAtran-scriptsofmiR-9(SupplementaryFigureS2).ExceptformiR-9-1,CAGE-seqdetectedmiR-9pri-miRNATSSsat16hpf(14somitesstage)(SupplementaryFigureS3A)beforematuremiR-9becomesdetectablebyinsituhybridization.Inadditiontothehostpromoter,wedetectedanalternativeintronicpromoterbyCAGEintheupstream(∼1.1kB)regionofmiR-9-5(SupplementaryFigureS2).Notably,thealternativeintronicTSSandhosttranscriptTSSofmiR9-5aresimultaneouslyinitiatedatthe14somitesstageasdetectedbyCAGE-seqandvalidatedby5RACEPCR.

TheexpressionpatternofmaturemiR-9,determinedbymiR-9antisenseLNAprobe,isrestrictedtotelencephalonduringprim6stage(Figure2A).Spatiotemporaldistribu<BR>tionofeachmiR-9pri-miRNAwasdeterminedbyISHprobes(SupplementaryFigureS3B),andwashighlyrem-iniscentofthematuremiR-9LNAprobe,mostlyrestrictedtothetelencephalonatprim6(Figure2B-I)andatlaterstagesincludingtheposteriorbrain(SupplementaryFigureS3C),thusvalidatingthesetranscriptsasmiR-9precursors.However,differencesinfinepatternsofexpressionwerede-tectedbetweendifferentmiR-9familymembersmostno-tablyatprim6.Forinstance,miR-9-2ishighlyexpressedinthedevelopingretina(Figure2C,whiteasterisk),miR-9–4showedearlyexpressionintheposteriorbrainregions(Figure2E,whitearrowhead)andmiR-9-6showedadistinctivestrongexpressioninthehypothalamus(Figure2H,whitear-row).Expressionofthesepri-miRNAprecedestheappearanceofdetectablelevelsofmaturemiR-9inthesebrainre-gions,asseenat48hpf(SupplementaryFigureS3C).ThedifferentialdistributionofmiR-9pri-miRNAsmightgen-eratedifferentlevelofmaturemiR-9indifferentbrainar-easrelativetoitsmRNAtargets,andmightthusimpactitsfunctionaloutput.Incontrast,thealternativeintronicTSS(probe-1)andhosttranscriptTSS(probe-2)ofmiR-9-5primiRNAshowedanidenticalspatialdistribution(Figure2FandG).
ToaddressiftheobservedintronicCAGEtagsofmiR9-5actsasanalternativepromoter,weperformed5RACEPCR(primerslocatedupstreamofpre-miRNA(seeMate-rialsandMethods;SupplementaryTableS2))onmiR-9-5andcomparedittomiR-9-4inwhichnoalternativeintronicCAGEtagsweredetected.5RACEproductofmiR-9-5extendedexactlytotheintronicCAGEtags(Figure2JandK),thusconfirminganalternativeusageofintronicpro-moterthatisindependentofitshosttranscript.Incon-trast,5RACEPCRofmiR-9-4didnotrevealanydiscreteTSSintheupstreamregion(Figure2K)consistentwiththelackofalternativeintronicCAGEtags,furthersuggestingthatmiR-9-4isprocessedfromitshostgeneaspredictedbyCAGEanalysis.
Characteristicfeaturesofpri-miRNApromoters
Nextwecomputedtheexpressionlevelofpri-miRNApro-moters(seeMaterialsandMethods)andinvestigatedtheirdynamicsduringearlyembryogenesisbyhierarchicalclustering(Figure3A).Only12pri-miRNAs(encoding19pre<BR>miRNA;SupplementaryTableS3)arematernallyinheritedandremainactive(exceptformiR-30a)duringthecriticalphaseofvertebrateontogeny.Pri-miRNAsofmiR-30aisactiveonlyduringmaternalstagesandnozygotictranscriptionwasdetectedwithinthestagesexamined.Themajorityofremainingpri-miRNAsarezygoticallyinitiated.Com-paredtocodinggenes,pri-miRNAsareexpressedatlowlevels(SupplementaryFigureS4A),whichinadditiontotheirrapidprocessingmaybeduetothecell/tissue-specificac-tivityofmiRNAs(39,40),whichisdilutedinwhole-embryoRNAlibraries.
AlignmentofrepresentativeTSSsofpri-miRNAsre-vealedanenrichmentofcanonicalinitiators(Figure3B),similartogenome-wideenrichmentofpromoters(23,24).TATA-likemotifsenrichedinupstreamregionsweremostlynon-canonicalTATA-likeelements(W-boxmotifsasde-scribedin(26))asonly9pri-miRNAshadacanonicalTATAbox(characterizedbyTATAWA).Wefurtherclassifiedthecorepromotersinto‘sharp’and‘broad’promoters(SupplementaryFigureS4B)(23)andobservedapreferenceforbroadpromoters(Figure3C),althoughsharppromotersarepreferredduringearly(maternaltranscriptome)stages,similartothegenome-widepatterns(24,26).SharppromotersareenrichedforW-boxwhencomparedtobroadpro-moters(SupplementaryFigureS4C)asexpected(26).Pri<BR>miRNAspromotershavehighGCcontentcenteredaroundtheir5-ends(Figure3D)where54(51.4%)pri-miRNApro-moterregionsoverlapwithCGIs.
Promoter-associatedhistonemodificationsandPolIIareen-richedonpre-miRNAs
ItwasrecentlydemonstratedthatmiR-320aisdirectlytran-scribedatthe5-endofthepre-miRNA,suggestinganal-ternativetranscriptionalmechanismforthegenerationofmiRNAs(14).WehavenoticedH3K4me3-modifiedhis-tonesarepresentonpre-miRNAsitesofindividualmiR9inzebrafishandhuman(Figure4A,B;SupplementaryFigureS2).MotivatedbytheseobservationsweanalyzedchromatinfeaturesandCAGEsignalsonpre-miRNAsonagenome-widescale.WealignedpromoterassociatedH3K4me3,H2A.ZandRNAPIImarksalong5-endsofpre-miRNAsandobservedanoverallenrichment(Figure4C).Closerinspectionattheindividualpre-miRNAsrevealedaround47%(71outof152)ofpre-miRNAsshowenrichedH3K4me3peaks(SupplementaryTableS3).Similarly,alignmentofH3K4me3,H2A.Z,H3K27acandH3K9acmodifiedhistonesalong5-endsofpre-miRNAsinhumanembryonicstemcells(hESC)revealedanen-richedpeak(SupplementaryFigureS5A-D).Moreover,pre-miRNAsitesoverlappingwithCGIscarryhigherlevelofhistonemodifications(SupplementaryFigureS5E-H).Weidentified77(outof346expressedinhESC)premiRNAshadenrichedH3K4me3peaks,ofwhich19hadenrichedH3K4me3peaksinbothzebrafishandhuman(SupplementaryTableS4),indicatingthatitisanevolutionarilyconservedfeature.
Thesepromoter-associatedfeaturesdetectedatpremiRNAsmaybeduetoapre-miRNAspecificchromatinmodification,butcouldalsobeduetooverlapwiththeextendedregionoriginatingfromthehostgenepromoterandrunningintothebodyofthegeneforseveralkilo-

Figure2.ExpressionpatternrevealedbyinsituhybridizationofmaturemiR-9andpri-miRNAtranscriptsduringprim6stage.(A)ExpressionpatternofmaturemiR-9ismostlyrestrictedtotelencephalon.Leftpanelsrepresentlateralviewsandrightpanelsrepresentdorsalviews.(B–I)ExpressionpatternofmiR-9pri-miRNAtranscriptsisalsorestrictedtotelencephalonwithdifferencesinthespatiotemporaldetailsamongdifferentpri-miRNAs,suchas:miR-9-2(developingretina,whiteasterisk;Figure2C),miR-9-4(posteriorbrainregions,whitearrowhead;Figure2E)andmiR-9-6(hypothalamus,whitearrow;Figure2H).(J)Amplifiedregionof5RACEPCRproduct(correspondingtoprobe-1)ofmiR-9-5mapsexactlytothealternativeintronicCAGEtags.(K)Uniquediscreteamplificationbandobtainedfrom5RACEPCRproductsobtainedformiR-9-5(withintronicCAGEtags).Multiplenon-specificbandsand/orasmearwereobtainedformiR-9-4thatdidnothavealternativeintronicCAGEtags.
bases,asobservedindevelopmentalregulatorgenes(41,42).Totestiftheenrichmentofpromoter-associatedchromatinfeaturesispre-miRNAspecific,wealignedindividualpre-miRNAsaccordingtoincreasingdistancefromtheirpri-miRNAstartsites.Thisanalysisrevealedthatmod-ifiedhistonescanbealignedseparatelyonpre-miRNAsfromtheirpri-miRNApromoterregions(Figure4D),sug<BR>gestingthatbothpre-miRNAandpri-miRNAstartre-gionscanbeindependentlymarkedbypromoter-associatedmodifiedhistonesandRNAPII.Outof71pre-miRNAs,28pre-miRNAshadindependentH3K4me3peaksthatweredisjointedfromH3K4me3peaksassociatedwithprimiRNA.Theremaining43pre-miRNAshadH3K4me3peaksthatspannedfrompri-miRNAtopre-miRNAandhaddifferentenrichmentpatterns(SupplementaryFigureS5I).Sincepre-miRNAsaresplicedRNAs,specificenrichmentofpromoter-associatedmodifiedhistonesonasub-setofpre-miRNAswasratherunexpectedandcouldrep-resenteither(i)alocalpromoterundetectablebyCAGE-seq(similartotheonlyknownmiRNAwithproximalTSSs(14))or(ii)theinvolvementofchromatinregulationinco-transcriptionalprocessingofpre-miRNAs(16).
DetectionofDroshaprocessingsitesformiRNA-offsetRNA(moRNA)byCAGE
Tofurtheranalyzethetranscriptionalandpost-transcriptionalfeaturesofpre-miRNAs,wealignedCAGEtagsaround5-endsofpre-miRNAs.WedetecteddistinctenrichmentofCAGEtagsatthe3-endsofpre-miRNAs(Figure5A).UponinspectionofCAGEtagsmappedtothelociofpre-miRNAhairpinsgenomewide,weobservedthattheyareoftenpresentadjacenttothe3-endofthese


Figure3.Developmentaldynamicsofpri-miRNAsandtheircharacteristicsfeatures.(A)Clusteringofpri-miRNAsbasedontheactivityofannotatedpromotersacross12developmentalstages.Theheatmaprepresentstheexpressionlevel(log2(tpm))atindicatedstages.Blackrectangleindicatesmaternallyinheritedpri-miRNAs.(B)Alignmentofsequencesbasedonrepresentativetranscriptionstartsitesofpri-miRNAsrevealanenrichmentofcanonicalinitiatorsandassociatedTATA-likemotifs.(C)Distributionofsharpandbroadpromotersandtheirdynamicusageacrossdevelopmentalstages.X-axisindicates12developmentalstages,sameasin(A).(D)AverageG+Ccontentofpri-miRNAspromoters.Y-axisindicatestheaverageGCcontent.

Figure4.Promoter-associatedhistonemodificationsatpre-miRNAsdifferfromthoseatpri-miRNApromoters.(A)Agenomebrowserviewofzebrafishmir-9–2withCAGEtags,H3K4me3andH2A.Ztracks.H3K4me3andH2A.Zpeaksatpre-miRNAaredistinctfrompri-miRNA.(B)AgenomebrowserviewofhumanorthologousMIR-9-2withH3K4me3andH3K27actracks.(C)AlignmentofaverageH3K4me3(prim6stage),H2A.ZandRNAPII(bothDome/30%Epibolystage)marksalongthe5-endsofpre-miRNAsrevealanenrichedpeakassociatedwithpre-miRNAs.(D)HeatmapsshowingH3K4me3,H2A.ZandRNAPIIsignalsofeachpre-miRNAs.Pre-miRNAsoverlappingCGIsarestackedontopandpre-miRNAsnotoverlappingCGIsarestackedbelowintheincreasingorderofpri-miRNAdistance.

Figure5.CAGE-seqdetectspost-transcriptionalprocessingofpre-miRNAsonDroshacleavagesites.(A)AlignmentofCAGEtagsaroundpre-miRNAsrevealpost-transcriptionallygeneratedCAGEtagsareenrichedatDroshacleavagesiteduringzygoticstages.Pre-miRNAsareschematicallyrepresentedbyrectangularboxes.Brokenedgestowards3-endrepresentsvariablelength.RedandbluearrowsindicatetheDicerandDroshacleavagesites.(B)AgenomebrowserviewofmiR-125cwithCAGEtagsandsmallRNAtracksfromprim6stage.BlackhorizontalbarrepresentannotatedmaturemiRNA.SmallRNAreadsmaptomaturemiRNAarmsandflankingregions(blueblocks).(C)Comparisonoflengthoftranscriptioninitiationclusters(TCs)ofmoRNAsandpri-miRNAs.(DandE)SmallRNAreadsalignedto5-endsofDroshacleavagesitesdetectedbyCAGE-seqfrom(D)maternal(256cells)and(E)zygotic(Prim6)transcriptomerevealedthatmoRNAproductionoccurspredominantlyinzygotictranscription.
hairpins(Figure5B;SupplementaryFigureS6A–C),thusconfirmingthatCAGEtagsmappreferentiallytothenucleotideimmediatedownstreamofmaturemiRNA.TheCAGEtagsweredetectedat3-endofpre-miRNAsin61knownpre-miRNAgenes(SupplementaryTableS5)anddetectedonlyduringzygoticstages.TheunderlyingsequencesdonotcorrespondtoTSSs(SupplementaryFigureS6D)andthesignalisvery‘sharp’(nospreading,Figure5C).TheseobservationssuggestthatthedetectedRNArepresentsDroshacleavageproductdetectedbyCAGE-seq,similarlytootherpost-transcriptionallypro-cessedRNAproducts(24,25),andcorrespondsto5-endoftheso-calledmiRNA-offsetRNAs,detectedpreviouslyininCionaintestinalis(21)andmammals(22).
TovalidatethatthepresenceofCAGEtagsatmoRNAprecursorsitesisduetothepresenceofmoRNAsthem-selves,weanalyzedsmallRNAlibrariesgeneratedfromma-ternal(256cells)andzygotic(prim6)transcriptomes(seeMaterialsandMethods).Weidentified69pre-miRNAswithmoRNAs(50onthe5armand35onthe3arm;16pre-miRNAshavemoRNAsonbotharms)byusingatleast1readpersiteasthethreshold.Byincreasingtheminimumthresholdto2,46pre-miRNAswithmoRNAsaredetected(33on5armand21on3arm;8pre-miRNAshavemoR-NAsonbotharms)(SupplementaryFigureS6E).ThelowexpressionofmoRNAsisconsistentwithpreviousreports(21,22),andpartlyexplainswhyitisdifficulttouncoversucheventsunlesssequencedatunprecedenteddepth.Com-parisonofmoRNAsdetectedbyCAGE-seqandbysmallRNA-seqrevealedonly29moRNAsdetectedbybothtech-niques.SmallRNAsareproducedineitherofthearms(Figure5DandE;SupplementaryFigureS6A–C)withslightpreferenceforthe5arm.However,during256-cells(maternal)stageonlyfourpre-miRNAshadmoRNAs(≥2reads),andnomoRNAsweredetectedbyCAGE-seq.ThissuggeststhatmoRNAsareprimarilyzygoticallyinitiatedandaredenovoprocessedtranscriptsratherthaninheritedfromtheoocyte.

moRNACAGEaspredictivetoolforthevalidationofmiRNAdiscovery
TheresultspresentedaboveindicatethatmoRNAassociatedCAGEsignalisapredictorofactivelyprocessedpre-miRNAhairpins.TotestifCAGE-seqcouldbeusedasatoolforbiochemicalvalidationofcomputationallypredictedpre-miRNAhairpinprocessing,weanalyzed227predictedpre-miRNAs(seeMaterialsandMethods).WeidentifiedCAGEtagsonlyonfourpre-miRNAswhereonlyasinglepre-miRNAhadevidencewithhighconfidenceofactiveDroshaprocessing(SupplementaryFigureS6F).Furthermore,wealsoannotateditspri-miRNApromoterandobservedenrichedH3K4me3modifiednucleosomesalongpre-miRNAs.Thedrasticallylowerrateofactiveprocessingofpre-miRNAsobservedoncomputationallypredictedpre-miRNAssuggeststhatthelargemajorityofcomputationallypredictedmiRNAsmaynotbebiologicallyfunctional,andthatmoRNAdetectionprovidesstrongindependentevidencefortheirfunction.
ValidationofCAGEdetectedmoRNAsprocessingsitesbyRACEPCR
WenextsoughttoaskifCAGEtagsdetectedonpremiRNAscanbeverifiedbyanalternative5capdetectionmethod.Weperformed3RACEPCRexperimentsusingprimerslocateddownstreamofpre-miRNA(seeMaterialsandMethods)onthepredictedprimarytranscriptsofmiR9-4/5(locatedonintronofhostgene)andmiR-9-1/6(bothofwhichharboranexonichairpin)(Figure6).AnalysisofmiR-9–4/5productsindicatedthatthe5-endsmappedex-actlytotheannotatedDroshacleavagesite(redarrow;Fig-ure6AandB),furthersuggestingthatitisfirstprocessedfromthehostintronandsubsequentlycapped.Similarly,analysisofmiR-9–1/6productsindicatedthatthe5-endsmappedinclosevicinity(1baseoffset)tothepredictedDroshacleavagesite(redarrow;Figure6CandD).Inter-estingly,miR-9-1/6productscorrespondtothesplicedformoftheirprimarytranscript,suggestingthatprocessingbyDroshaoccurreddownstreamofprimarytranscriptsplicing.Consistently,unprocessedsplicedpri-miRNAsofmiR9-1/6canbeamplifiedbyRT-PCR,aswasdemonstratedwhenweclonedmiR-9-1andmiR-9-6ISHprobes(SupplementaryFigureS3B).
DISCUSSION 
Wehavepresentedacomprehensiveannotationofzebrafishpri-miRNApromoterrepertoiresthatcanserveasavaluableresourceforthestudyofmiRNA-mediatedregulatorynetworks.Overall,theannotationshowsthatprimiRNApromotersshowboththesequencepropertiesandfunctionaldiversitycomparabletothoseofprotein-codinggenes.Differentpri-miRNApromotershaveenrichedGCcontent(oftenoverlappingCGIs)andmanyalsohaveTATA-likemotifs,aswellascombinationsoffeatureschar-acteristicforthepromotersofgenesunderlong-rangede-velopmentalregulation(41).Thissuggeststhatthereisnomodeoftranscriptionalregulationspecificforpri-miRNAgenes.
Onenotableobservationfromtheanalysisisthepresenceofpotentialmaternallyinheritedpri-miRNAsdur-ingdevelopmentofthezebrafishembryo.Whilewede-tected12pri-miRNAsasmaternallyinherited,however,post-transcriptionallyassociatedCAGEtagsonDroshaprocessingsiteswerenotpresentonthemduringmaternalstages,butweredetectedonlyaftermid-blastulatransition.Thissuggeststhatmaternallydepositedpri-miRNAscouldrequireazygoticallyactivatedregulatorforprocessingpremiRNAsfromthem.Tocharacterizepotentialfunctionsofmaternallyinheritedpri-miRNAsfurtherexperimentsarerequired,suchastherecentstudyreportinguniquemechanismofadenylationofmaternallyinheritedmiRNAs(43).However,thematernallyinheritedmiR-17–92cluster,whichhassimilarseedsitesasmiR-430andhumanembryonicregulatormiRNA(miR-302/miR-467)(44),maybeofgreatimportancetodevelopmentalbiologistinstudyingtheirpo-tentialrolesindegradationofmaternaltranscriptsduringearlyembryogenesis.
Enrichmentofpromoter-associatedmodifiedhistonesandRNAPIIsignalsonpre-miRNAsraisesaninterestingquestiononwhethertheyreflectanoverlapwithaknownchromatinfeatureofgenesunderlong-rangedevelopmen-talregulation,orrepresentanadditionallayerofchromatinregulationinvolvedspecificallyinco-transcriptionalpro-cessingofpre-miRNAs.Basedontheresultsfromourownanalysisandpreviousstudies,wehaveevidencesupportingbothscenariosonvariouspre-miRNAs.Wedetectedspe-cificenrichmentofH3K4me3,H2A.ZandRNAPIIpeaksatpre-miRNAssuggestinglocalpromoteractivity,howeverCAGE-seqand5RACEPCRoftwopre-miR-9candi-datesdidnotsupportproductionofcappedRNAatthesepredictedlocalpromoterregions(Figure2JandK).Simi<BR>larly,3RACEPCRonfourpre-miR-9candidatesrevealednolocalpromoteractivity,butindicationofDroshacleav-age.Theseevidencecollectivelyarguesagainstactivityoflo-calpromotersatpre-miRNAs.Alternatively,thechromatinfeaturesraisethepossibilityofayetunexplainedchromatinmediatedmechanismforaco-transcriptionalpro-cessingofpre-miRNAsproducedfromadistalpri-miRNApromoter,similartochromatinassociationoftranscriptionandmRNAsplicing(45,46).Anewstudyusingadifferentsequencingtechnique(nativeelongationtranscript;NET-seq)inhumancelllinesshowednascenttranscriptsmappedattheexpectedDroshacleavagesitesthatareas-sociatedwithco-transcriptionalprocessingofpre-miRNAs(47).Furthermore,DNAsequencesencodingmaturemiR-NAswerepreferentiallyoccupiedbypositionednucleosomes(8),whichmaybeassociatedwithco-transcriptionalpre-miRNAcleavage.SuchascenariomaypartlyexplainwhymiRNA-encodingintronsaresplicedataslowerratethanadjacentintrons(48).
AlthoughourunderstandingofmoRNAbiogenesisandfunctionremainslacking,ithasbeenspeculatedthat

Figure6.3RACEPCRamplificationofmiR-9-(1/4/5/6).(A–D)SchematicrepresentationsofmiR-9-(1/4/5/6)locianddesignedprimersbasedondownstreamsequences(seeSupplementaryTableS3).Theresulting3RACEPCRproductsareshownindarkblue.MaturemiRNAssequencesarehighlightedinorangeandtheremainingsequencesofpre-miRNAhairpinsingrey.5-endofRACEPCRproducts(bluebars)preciselymappedDroshacleavagesite(redarrow).TheagarosegelimagesontherightshowthatuniqueRACEPCRproductswereobtained.
moRNAbiogenesismightbeassociatedwithpri-miRNAprocessing.Ifso,itisunclearwhymoRNAsaredetectedonlyonsubsetofpre-miRNAs.Toanswerthis,wefirstre-quirethecapacitytocomprehensivelydetectmoRNAsthatishinderedbythegenerallylowlevelsoftheirexpression.CombiningCAGE-seqandsmallRNA-seqpredictioncanfurtherstrengthenmoRNAdetection,aswedetectedsimi-larnumberofmoRNAsinzebrafishasinhuman(22),eventhoughthenumberofannotatedpre-miRNAsinhumanbeingalmost3-foldlarger.EvidenceforpresenceofmoR-NAsinhomologouspre-miRNAsinCiona,zebrafishandhumanindicatesthattheunderlyingmoRNAbiogenesisisevolutionarilyconservedandmayindicateafundamentalprocessofgenerationofanon-codingRNAclass,whichre-mainspoorlyunderstood.Thedataandannotationwehavegeneratedinthisworkprovideavaluableresourceforfutureinvestigationofthesephenomena.
AVAILABILITY 
Processeddata(CAGE-seq,RNA-seqandhistonesChIPseq)canbedownloadedfromhttp://promshift.genereg.net/zebrafish/.Alternatively,itcanbeviewedonUCSCgenomebrowserhttp://www.genome.ucsc.edu/goldenPath/customTracks/custTracks.html#Zebrafish.
SUPPLEMENTARY DATA 
SupplementaryDataareavailableatNAROnline.
ACKNOWLEDGEMENTS 
C.N.wassupportedbyB.L.,V.M.S.,P.M.,J.B.A.,duringdifferentstagesoftheproject.V.M.S.acknowledgesgranttoGenomicsCoreFacilityattheUniversityofBergen,whichispartoftheNorwegianGenomicsConsortium.AuthorContributions:C.N.,F.M.,B.L.conceivedtheproject.C.N.analyzeddata.C.N.,F.M.,B.L.interpretedresults.M.C.,D.C.,L.B.C.analyzedandinterpretedinsituand5RACEPCRresults.C.N.,F.M.,B.L.wrotethefirstdraftofmanuscriptwithinputandrevisionfromM.C.,L.B.C.,D.C.,P.M.,P.C.,V.M.S,J.B.A.
FUNDING 
NorwegianResearchCouncil(YFFtoB.L.);BergenRe-searchFoundationawarded(toB.L.);EuropeanCommission[FP1313-HEALTH-306029‘TRIGGER’toP.M.];NationalScienceCenter[2014/14/E/NZ6/00162,Poland];EUFP6largeintegratedprojectsEuTRACC,ZFHealthandBBSRC[BB/L010488/1toF.M.andB.L.];EUprojectZF-Health[FP7/2010–2015grantagreementno.242048toM.C.andL.B-C.];ANR[ANR-2012-BSV4-0004-01];EuropeanResearchCouncil[AdG322936];INSERM(toM.C.);DanishCancerSociety[R98-A6446toJ.B.A.];Re-searchGrantfromtheJapaneseMinistryofEducation,Culture,SportsandTechnology(MEXT)totheRIKENCenterforLifeScienceTechnologies(toP.C.).Fundingforopenaccesscharge:EUprojectZF-Health[FP7/2010-2015grantagreementno.242048].Conflictofintereststatement.Nonedeclared.

REFERENCES 
1.
Denli,A.M.,Tops,B.B.,Plasterk,R.H.,Ketting,R.F.andHannon,G.J.(2004)ProcessingofprimarymicroRNAsbytheMicroprocessorcomplex.Nature,432,231–235.

2.
He,L.andHannon,G.J.(2004)MicroRNAs:smallRNAswithabigroleingeneregulation.Nat.Rev.Genet.,5,522–531.

3.
Lee,Y.,Kim,M.,Han,J.,Yeom,K.H.,Lee,S.,Baek,S.H.andKim,V.N.(2004)MicroRNAgenesaretranscribedbyRNApolymeraseII.EMBOJ.,23,4051–4060.

4.
Borchert,G.M.,Lanier,W.andDavidson,B.L.(2006)RNApolymeraseIIItranscribeshumanmicroRNAs.Nat.Struct.Mol.Biol.,13,1097–1101.

5.
Liu,N.,Williams,A.H.,Kim,Y.,McAnally,J.,Bezprozvannaya,S.,Sutherland,L.B.,Richardson,J.A.,Bassel-Duby,R.andOlson,E.N.(2007)AnintragenicMEF2-dependentenhancerdirectsmuscle-specificexpressionofmicroRNAs1and133.Proc.Natl.Acad.Sci.U.S.A.,104,20844–20849.

6.
Woods,K.,Thomson,J.M.andHammond,S.M.(2007)Directregulationofanoncogenicmicro-RNAclusterbyE2Ftranscriptionfactors.J.Biol.Chem.,282,2130–2134.

7.
Marson,A.,Levine,S.S.,Cole,M.F.,Frampton,G.M.,Brambrink,T.,Johnstone,S.,Guenther,M.G.,Johnston,W.K.,Wernig,M.,Newman,J.etal.(2008)ConnectingmicroRNAgenestothecoretranscriptionalregulatorycircuitryofembryonicstemcells.Cell,134,521–533.

8.
Ozsolak,F.,Poling,L.L.,Wang,Z.,Liu,H.,Liu,X.S.,Roeder,R.G.,Zhang,X.,Song,J.S.andFisher,D.E.(2008)ChromatinstructureanalysesidentifymiRNApromoters.GenesDev.,22,3172–3183.

9.
Corcoran,D.L.,Pandit,K.V.,Gordon,B.,Bhattacharjee,A.,Kaminski,N.andBenos,P.V.(2009)FeaturesofmammalianmicroRNApromotersemergefrompolymeraseIIchromatinimmunoprecipitationdata.PLoSOne,4,e5279.

10.
Chien,C.H.,Sun,Y.M.,Chang,W.C.,Chiang-Hsieh,P.Y.,Lee,T.Y.,Tsai,W.C.,Horng,J.T.,Tsou,A.P.andHuang,H.D.(2011)IdentifyingtranscriptionalstartsitesofhumanmicroRNAsbasedonhigh-throughputsequencingdata.NucleicAcidsRes.,39,9345–9356.

11.
Gao,Y.,Schug,J.,McKenna,L.B.,Lay,J.L.,Kaestner,K.H.andGreenbaum,L.E.(2011)Tissue-specificregulationofmousemicroRNAgenesinendoderm-derivedtissues.NucleicAcidsRes.,39,454–463.

12.
Georgakilas,G.,Vlachos,I.S.,Paraskevopoulou,M.D.,Yang,P.,Zhang,Y.,Economides,A.N.andHatzigeorgiou,A.G.(2014)microTSS:accuratemicroRNAtranscriptionstartsiteidentificationrevealsasignificantnumberofdivergentpri-miRNAs.Nat.Commun.,5,5700.

13.
Marsico,A.,Huska,M.R.,Lasserre,J.,Hu,H.,Vucicevic,D.,Musahl,A.,Orom,A.U.andVingron,M.(2013)PROmiRNA:anewmiRNApromoterrecognitionmethoduncoversthecomplexregulationofintronicmiRNAs.GenomeBiol.,14,R84.

14.
Xie,M.,Li,M.,Vilborg,A.,Lee,N.,Shu,M.D.,Yartseva,V.,Sestan,N.andSteitz,J.A.(2013)Mammalian5-cappedmicroRNAprecursorsthatgenerateasinglemicroRNA.Cell,155,1568–1580.

15.
Gregory,R.I.,Yan,K.P.,Amuthan,G.,Chendrimada,T.,Doratotaj,B.,Cooch,N.andShiekhattar,R.(2004)ThemicroprocessorcomplexmediatesthegenesisofmicroRNAs.Nature,432,235–240.

16.
Morlando,M.,Ballarino,M.,Gromak,N.,Pagano,F.,Bozzoni,I.andProudfoot,N.J.(2008)PrimarymicroRNAtranscriptsareprocessedco-transcriptionally.Nat.Struct.Mol.Biol.,15,902–909.

17.
Pawlicki,J.M.andSteitz,J.A.(2008)PrimarymicroRNAtranscriptretentionatsitesoftranscriptionleadstoenhancedmicroRNAproduction.J.CellBiol.,182,61–76.


18.
Zeng,Y.andCullen,B.R.(2005)EfficientprocessingofprimarymicroRNAhairpinsbyDrosharequiresflankingnonstructuredRNAsequences.J.Biol.Chem.,280,27595–27603.

19.
Han,J.,Lee,Y.,Yeom,K.H.,Nam,J.W.,Heo,I.,Rhee,J.K.,Sohn,S.Y.,Cho,Y.,Zhang,B.T.andKim,V.N.(2006)MolecularbasisfortherecognitionofprimarymicroRNAsbytheDrosha-DGCR8complex.Cell,125,887–901.

20.
Ballarino,M.,Pagano,F.,Girardi,E.,Morlando,M.,Cacchiarelli,D.,Marchioni,M.,Proudfoot,N.J.andBozzoni,I.(2009)CoupledRNAprocessingandtranscriptionofintergenicprimarymicroRNAs.Mol.Cell.Biol.,29,5632–5638.

21.
Shi,W.,Hendrix,D.,Levine,M.andHaley,B.(2009)AdistinctclassofsmallRNAsarisesfrompre-miRNA-proximalregionsinasimplechordate.Nat.Struct.Mol.Biol.,16,183–189.

22.
Langenberger,D.,Bermudez-Santana,C.,Hertel,J.,Hoffmann,S.,Khaitovich,P.andStadler,P.F.(2009)EvidenceforhumanmicroRNA-offsetRNAsinsmallRNAsequencingdata.Bioinformatics,25,2298–2301.

23.
Carninci,P.,Sandelin,A.,Lenhard,B.,Katayama,S.,Shimokawa,K.,Ponjavic,J.,Semple,C.A.,Taylor,M.S.,Engstrom,P.G.,Frith,M.C.etal.(2006)Genome-wideanalysisofmammalianpromoterarchitectureandevolution.Nat.Genet.,38,626–635.

24.
Nepal,C.,Hadzhiev,Y.,Previti,C.,Haberle,V.,Li,N.,Takahashi,H.,Suzuki,A.M.,Sheng,Y.,Abdelhamid,R.F.,Anand,S.etal.(2013)Dynamicregulationofthetranscriptioninitiationlandscapeatsinglenucleotideresolutionduringvertebrateembryogenesis.GenomeRes.,23,1938–1950.

25.
Fejes-Toth,K.,Sotirova,V.,Sachidanandam,R.,Assaf,G.,Hannon,G.,Philipp,K.,Sylvain,F.,Willingham,A.T.,Duttagupta,R.,Dumais,E.etal.(2009)Post-transcriptionalprocessinggeneratesadiversityof5-modifiedlongandshortRNAs.Nature,457,1028–1032.

26.
Haberle,V.,Li,N.,Hadzhiev,Y.,Plessy,C.,Previti,C.,Nepal,C.,Gehrig,J.,Dong,X.,Akalin,A.,Suzuki,A.M.etal.(2014)Twoindependenttranscriptioninitiationcodesoverlaponvertebratecorepromoters.Nature,507,381–385.

27.
Zhang,Y.,Vastenhouw,N.L.,Feng,J.,Fu,K.,Wang,C.,Ge,Y.,Pauli,A.,Hummelen,P.V.,Schier,A.F.andLiu,X.S.(2014)Canonicalnucleosomeorganizationatpromotersformsduringgenomeactivation.GenomeRes.,24,260–266.

28.
Wei,C.,Salichos,L.,Wittgrove,C.M.,Rokas,A.andPatton,J.G.(2012)Transcriptome-wideanalysisofsmallRNAexpressioninearlyzebrafishdevelopment.RNA,18,915–929.

29.
Kozomara,A.andGriffiths-Jones,S.(2011)miRBase:integratingmicroRNAannotationanddeep-sequencingdata.NucleicAcidsRes.,39,D152–D157.

30.
Flicek,P.,Amode,M.R.,Barrell,D.,Beal,K.,Brent,S.,Carvalho-Silva,D.,Clapham,P.,Coates,G.,Fairley,S.,Fitzgerald,S.etal.(2012)Ensembl2012.NucleicAcidsRes.,40,D84–D90.

31.
Meyer,L.R.,Zweig,A.S.,Hinrichs,A.S.,Karolchik,D.,Kuhn,R.M.,Wong,M.,Sloan,C.A.,Rosenbloom,K.R.,Roe,G.,Rhead,B.etal.(2013)TheUCSCGenomeBrowserdatabase:extensionsandupdates2013.NucleicAcidsRes.,41,D64–D69.

32.
Consortium,E.P.(2012)AnintegratedencyclopediaofDNAelementsinthehumangenome.Nature,489,57–74.

33.
Quinlan,A.R.andHall,I.M.(2010)BEDTools:aflexiblesuiteofutilitiesforcomparinggenomicfeatures.Bioinformatics,26,841–842.

34.
Crooks,G.E.,Hon,G.,Chandonia,J.M.andBrenner,S.E.(2004)WebLogo:asequencelogogenerator.GenomeRes.,14,1188–1190.

35.
Langmead,B.,Trapnell,C.,Pop,M.andSalzberg,S.L.(2009)Ultrafastandmemory-efficientalignmentofshortDNAsequencestothehumangenome.GenomeBiol.,10,R25.

36.
Thisse,C.andThisse,B.(2008)High-resolutioninsituhybridizationtowhole-mountzebrafishembryos.Nat.Protoc.,3,59–69.

37.
Coolen,M.,Katz,S.andBally-Cuif,L.(2013)miR-9:aversatileregulatorofneurogenesis.Front.Cell.Neurosci.,7,220.

38.
Leucht,C.,Stigloher,C.,Wizenmann,A.,Klafke,R.,Folchert,A.andBally-Cuif,L.(2008)MicroRNA-9directslateorganizeractivityofthemidbrain-hindbrainboundary.Nat.Neurosci.,11,641–648.

39.
Lagos-Quintana,M.,Rauhut,R.,Yalcin,A.,Meyer,J.,Lendeckel,W.andTuschl,T.(2002)Identificationoftissue-specificmicroRNAsfrommouse.Curr.Biol.,12,735–739.

40.
Landgraf,P.,Rusu,M.,Sheridan,R.,Sewer,A.,Iovino,N.,Aravin,A.,Pfeffer,S.,Rice,A.,Kamphorst,A.O.,Landthaler,M.etal.(2007)A


mammalianmicroRNAexpressionatlasbasedonsmallRNAlibrarysequencing.Cell,129,1401–1414.
41.
Akalin,A.,Fredman,D.,Arner,E.,Dong,X.,Bryne,J.C.,Suzuki,H.,Daub,C.O.,Hayashizaki,Y.andLenhard,B.(2009)Transcriptionalfeaturesofgenomicregulatoryblocks.GenomeBiol.,10,R38.

42.
Benayoun,B.A.,Pollina,E.A.,Ucar,D.,Mahmoudi,S.,Karra,K.,Wong,E.D.,Devarajan,K.,Daugherty,A.C.,Kundaje,A.B.,Mancini,E.etal.(2014)H3K4me3breadthislinkedtocellidentityandtranscriptionalconsistency.Cell,158,673–688.

43.
Lee,M.,Choi,Y.,Kim,K.,Jin,H.,Lim,J.,Nguyen,T.A.,Yang,J.,Jeong,M.,Giraldez,A.J.,Yang,H.etal.(2014)AdenylationofmaternallyinheritedmicroRNAsbyWispy.Mol.Cell,56,696–707.

44.
Svoboda,P.andFlemr,M.(2010)TheroleofmiRNAsandendogenoussiRNAsinmaternal-to-zygoticreprogrammingandtheestablishmentofpluripotency.EMBORep.,11,590–597.


NucleicAcidsResearch,2016,Vol.44,No.73081 
45.
Schwartz,S.,Meshorer,E.andAst,G.(2009)Chromatinorganizationmarksexon-intronstructure.Nat.Struct.Mol.Biol.,16,990–995.

46.
Listerman,I.,Sapra,A.K.andNeugebauer,K.M.(2006)CotranscriptionalcouplingofsplicingfactorrecruitmentandprecursormessengerRNAsplicinginmammaliancells.Nat.Struct.Mol.Biol.,13,815–822.

47.
Nojima,T.,Gomes,T.,Grosso,A.R.,Kimura,H.,Dye,M.J.,Dhir,S.,Carmo-Fonseca,M.andProudfoot,N.J.(2015)MammalianNET-SeqRevealsGenome-wideNascentTranscriptionCoupledtoRNAProcessing.Cell,161,526–540.

48.
Kim,Y.K.andKim,V.N.(2007)ProcessingofintronicmicroRNAs.EMBOJ.,26,775–783.