RESEARCHARTICLESUMMARY◥NEUROBIOLOGYLossofamammaliancircularRNAlocuscausesmiRNAderegulationandaffectsbrainfunctionMonikaPiwecka,*PetarGlažar,*LuisR.Hernandez-Miranda,*SebastianMemczak,SusanneA.Wolf,AgnieszkaRybak-Wolf,AndreiFilipchyk,FilipposKlironomos,ClediAliciaCerdaJara,PascalFenske,ThorstenTrimbuch,VeraZywitza,MireyaPlass,LuisaSchreyer,SalahAyoub,ChristineKocks,RalfKühn,ChristianRosenmund,CarmenBirchmeier,NikolausRajewsky†INTRODUCTION:Recently,aspecialclassofRNAshasexcitedresearchersandtriggeredhundredsofnow-publishedstudies.KnownascircularRNAs(circRNAs),theseRNAsarepro-ducedbyregulartranscriptionfromgenomicDNA,butthetwoendsofthe(usually)exonictranscriptsarecovalentlyclosed,probablyinmostcasesbynoncanonicalsplicereactions.MostcircRNAsareexpressedinthecytoplasmandareunusuallystable,suggestingthattheymayhavefunctionsthatdivergefromthoseofcanonicalmessengerRNAs(mRNAs)orlongnoncodingRNAs(lncRNAs).CircRNAstendtobeweaklyexpressed,butthereareexceptionsinanimalbrains.Forexample,inthemousebrain,afewhundredcircRNAsarehighlyexpressed,oftenwithde-velopmentallyspecificexpressionpatternsthatareconservedinthehumanbrain.Weprevious-lyproposedthatcircRNAsmay,atleastsome-times,serveasregulatoryRNAs.AcircRNAdiscoveredbytheKjemslaboratory,CDR1as,caughtourattentionbecauseitwascoveredwith>70bindingsitesforthemicroRNA(miRNA)miR-7.OurdatasuggestedthatCDR1asmightservetoalterthefreeconcen-trationofmiR-7.Butwhatreallyisthefunc-tionofCDR1as?RATIONALE:WefirstdeterminedwhichmiRNAsspecificallybindCdr1asinpostmor-temhumanandmousebrainsandcharac-terizedCdr1asexpressionpatterns.Oncewehadthatinformation,weremovedCdr1asfromthemousegenometostudythemolec-ularandbehavioralconsequences.RESULTS:WeshowthatCdr1asis,inthehumanbrain,directlyandmassivelyboundbymiR-7andmiR-671.Infact,Cdr1asisoneofthemostcommontranscriptstargetedbymiRNAsoutofallbrainmRNAsorlncRNAs.TheexpressionofmiRNAswasgenerallyunperturbedinCdr1asknockout(KO)mice,withtheexceptionofthetwomiRNAsthatdi-rectlyinteractwithCdr1as,miR-7andmiR...
