The molecular basis of coupling between poly(A)-tail length ...
文章推薦指數: 80 %
These mRNAs were made by in vitro transcription from DNA templates that encoded the mRNA body followed by the poly(A) tail as well as the ... Sharethisarticle Doi Copytoclipboard Citethisarticle KehuiXiang DavidPBartel (2021) Themolecularbasisofcouplingbetweenpoly(A)-taillengthandtranslationalefficiency eLife10:e66493. https://doi.org/10.7554/eLife.66493 Copytoclipboard DownloadBibTeX Download.RIS Article Figuresanddata Abstract eLifedigest Introduction Results Discussion Materialsandmethods Dataavailability References Decisionletter Authorresponse Articleandauthorinformation Metrics Inanimaloocytesandearlyembryos,mRNApoly(A)-taillengthstronglyinfluencestranslationalefficiency(TE),butlaterindevelopmentthiscouplingbetweentaillengthandTEdisappears.Here,weelucidatehowthiscouplingisfirstestablishedandwhyitdisappears.Overexpressingcytoplasmicpoly(A)-bindingprotein(PABPC)inXenopusoocytesspecificallyimprovedtranslationofshort-tailedmRNAs,therebydiminishingcouplingbetweentaillengthandTE.Thus,strongcouplingrequireslimitingPABPC,implyingthatincoupledsystemslonger-tailmRNAsbettercompeteforlimitingPABPC.InadditiontoexpressingexcessPABPC,post-embryonicmammaliancelllineshadtwootherpropertiesthatpreventedstrongcoupling:terminal-uridylation-dependentdestabilizationofmRNAslackingboundPABPC,andaregulatoryregimewhereinPABPCcontributesminimallytoTE.Thus,theseresultsrevealedthreefundamentalmechanisticrequirementsforcouplinganddefinedthecontext-dependentfunctionsforPABPC,whichpromotesTEbutnotmRNAstabilityincoupledsystemsandmRNAstabilitybutnotTEinuncoupledsystems. Cellsaremicroscopicbiologicalfactoriesthatareconstantlycreatingnewproteins.Todoso,acellmustfirstconvertitsmastergeneticblueprint,theDNA,intostrandsofmessengerRNAormRNA.Thesestrandsaresubsequentlytranslatedtomakeproteins.Cellshavetwowaystoadjustthenumberofproteinstheygeneratesotheydonotproducetoomanyortoofew:bychanginghowmanymRNAmoleculesareavailablefortranslation,andbyregulatinghowefficientlytheytranslatethesemRNAmoleculesintoproteins. Inanimals,bothunfertilizedeggsandearly-stageembryoslacktheabilitytocreateordestroymRNAs,andconsequentlycannotadjustthenumberofmRNAmoleculesavailablefortranslation.ThesecellscanthereforeonlyregulatehowefficientlyeachmRNAistranslated.Theydothisbychangingthelengthoftheso-calledpoly(A)tailattheendofeachmRNAmolecule,whichismadeupofalongstretchofrepeatingadenosinenucleotides.ThemRNAswithlongerpoly(A)tailsaretranslatedmoreefficientlythanthosewithshorterpoly(A)tails.However,thisdifferencedisappearsinolderembryos,whenbothlongandshortpoly(A)tailsaretranslatedwithequalefficiency,anditislargelyunknownwhy. Tofindoutmore,XiangandBartelstudiedfrogeggs,anddiscoveredthatartificiallyraisinglevelsofaproteinthatbindspoly(A)tails,alsoknownasPABPC,improvedthetranslationofshort-tailedmRNAstocreateasituationinwhichbothshort-andlong-tailedmRNAsweretranslatedwithnear-equalefficiency.Thissuggestedthatshort-andlong-tailedmRNAscompeteforlimitedamountsofthetranslation-enhancingPABPC,andthatlong-tailedmRNAsarebetteratitthanshort-tailedmRNAs.FurtherinvestigationrevealedthateggsalsohadtoestablishtherightconditionsforPABPCtoenhancetranslationandhadtoprotectmRNAsnotassociatedwithPABPCfrombeingdestroyedbeforetheycouldbetranslated. Overall,XiangandBartelfoundthatineggsandearlyembryos,PABPCandpoly(A)tailsenhancedthetranslationofmRNAsbutdidnotinfluencetheirstability,whereaslaterindevelopment,theyenhancedmRNAstabilitybutnottranslation. Thisresearchprovidesnewinsightsintohowproteinproductioniscontrolledatdifferentstagesofanimaldevelopment,fromunfertilizedeggstoolderembryos.Understandinghowthisprocessisregulatedduringnormaldevelopmentiscrucialforgaininginsightsintohowitcanbecomedysfunctionalandcausedisease.Thesefindingsmaythereforehaveimportantimplicationsforresearchintoareassuchasinfertility,reproductivemedicineandraregeneticdiseases. MosteukaryoticmRNAsarepolyadenylatedattheir3′endsinaprocessassociatedwithtranscriptionaltermination.Inthenucleus,thesepoly(A)tailscanfacilitatemRNAnucleocytoplasmicexport(KühnandWahle,2004),whereasinthecytoplasm,theyserveasmoleculartimersformRNAdecay,withtheirlengthsbecomingprogressivelyshorterbydeadenylation,whicheventuallyleadstomRNAde-cappingandturnover(ChenandShyu,2011;Eisenetal.,2020;GoldstrohmandWickens,2008). Thelengthofapoly(A)tailcanalsoinfluencemRNAtranslationalefficiency(TE).Pioneeringstudiesinmaturingoocytesandearlyembryosshowthatlengtheningofpoly(A)tailsthroughcytoplasmicpolyadenylationiscriticalforregulatinggeneexpressionduringtheseearlystagesofanimaldevelopment(Richter,1999;Sallésetal.,1994;Sheetsetal.,1995).Resultsfromtheseandothersingle-genestudiesinoocytesandearlyembryoshadledtothenotionthatthelengthofapoly(A)tailgenerallycorrelateswithTE(Eckmannetal.,2011;Weilletal.,2012).Morerecenttranscriptome-widestudiesconfirmastrongglobalrelationshipbetweentaillengthandTEinoocytesandearlyembryos(Eichhornetal.,2016;Limetal.,2016;Subtelnyetal.,2014).However,infish,frogs,andflies,thiscorrelationdiminishesnearthetimeofgastrulation,andcouplingbetweenpoly(A)-taillengthandTEisessentiallynonexistentinpost-embryonicsystems(Eichhornetal.,2016;J.-E.Parketal.,2016;Subtelnyetal.,2014).Thus,theseglobalanalysesrevealadevelopmentaltransitioninhowtranslationisregulated(Subtelnyetal.,2014),whichcloselyfollowsthelong-knownmaternal-to-zygotictransitionintranscriptionalcontrol.Theexistenceofthistransitionintranslationalcontrolbringstotheforemechanisticquestionsastohowcouplingbetweenpoly(A)-taillengthandTEisestablishedinoocytesandearlyembryosandwhythiscouplingdisappearslaterindevelopment. Cytoplasmicpoly(A)-bindingproteins(PABPCs)arehighlyconservedRNA-bindingproteinsineukaryotes(Mangusetal.,2003).AlthoughSaccharomycescerevisiaehasonlyonePABPC(Pab1p),mostanimalscontainmultipleparalogsthathavespatiallyandtemporallyvariedexpressionpatterns(Smithetal.,2014;Wigingtonetal.,2014).PABPCshavehighaffinitytopoly(A)sequencesinvitro(Kd ~5nMforA25)andrequireatleast12Asforefficientbinding(KühnandWahle,2004).BindingofPABPCstomRNApoly(A)tailscanenhancetranslation,butthemechanismofthisenhancementisunclear.OnemodelpositsthatthemRNAformsaclosed-loopstructuremediatedbytheassociationoftheeukaryotictranslationinitiationfactoreIF4G(ascaffoldingprotein)withbothPABPCandthecap-bindingproteineIF4E(Hinnebusch,2014;ThompsonandGilbert,2017;Wellsetal.,1998).ThisassociationisproposedtostabilizetheinteractionbetweeneIF4EandthemRNA5′capandfacilitaterecruitmentand/orrecyclingofribosomestoincreasetranslationinitiation(Kahvejianetal.,2001).However,despitedirectvisualizationofloop-likeassembliesbothwithinsomecellsandinaninvitroreconstitutedsystem(Christensenetal.,1987;Wellsetal.,1998),resultsofseveralstudieshavequestionedtheuniversalityofthismodelamongdifferentmRNAsandbiologicalsystems(Adivarahanetal.,2018;Amranietal.,2008;Costelloetal.,2015;Risslandetal.,2017;ThompsonandGilbert,2017). PABPCscanalsoinfluencemRNAstability,asshowninyeast.GeneticablationofyeastPab1pislethalandcauseslengtheningofsteady-statepoly(A)-taillengths(SachsandDavis,1989),whichisattributedtopre-maturemRNAdecappingandcompromiseddeadenylation(CaponigroandParker,1995).BothyeastandmammalianPABPCscaninteractwithtwomRNAdeadenylationcomplexesPAN2-PAN3andCCR4-NOT,andeitherpromoteorinhibittheiractivitiesinvitro(Schäferetal.,2019;Uchidaetal.,2004;Websteretal.,2018;Yietal.,2018).BecausemRNAdecayiscoupledtodeadenylation(DeckerandParker,1993;Eisenetal.,2020),thedeadenylation-stimulatoryeffectsofPABPCwouldacceleratethedemiseofboundmRNAs,whichcontraststootherstudiessuggestingPABPCprotectsmRNAsfromdegradationincellextracts(Bernsteinetal.,1989;Wangetal.,1999).ThedichotomousandpotentiallyconflictingfunctionsofmetazoanPABPCexaminedinvitroraisethequestionoftheextentstowhichPABPCmightinfluencemRNApoly(A)-taillengthandstabilityinmetazoancells. PABPCsaregenerallythoughttocoatmRNApoly(A)tailsinthecytoplasm(KühnandWahle,2004;Mangusetal.,2003).However,thestoichiometrybetweenPABPCandpoly(A)sitesmightvaryindifferentbiologicalcontexts(Cossonetal.,2002;Voeltzetal.,2001),anditisunclearwhetherthispotentiallyvariablestoichiometrymightimpactgeneregulationincells.Moreover,thepossibilitythatPABPCmightinfluenceproteinsynthesisbyaffectingeithermRNAstabilityorTEcancomplicateanalysisofitsmolecularfunctionsindifferentbiologicalsystems,leavingitsmechanisticrolespoorlyunderstood. Here,weuncovermechanisticrequirementsforcouplingbetweenpoly(A)-taillengthandTEobservedinoocytesandearlyembryos,showingthatthiscouplingandthesubsequentuncouplingobservedlaterindevelopmentrelyonacontext-dependentswitchinthefunctionofPABPCs. Toassaytheinfluenceofpoly(A)-taillengthonTE,weusedaninvitrotranslationextractmadefromstageVIXenopuslaevisoocytes,wherecytoplasmicpolyadenylationleadstotranslationalactivationofthec-mos,cdk2andsomecyclinmRNAs (RichterandLasko,2011).IntothisextractweaddedNanolucluciferase(Nluc)reportermRNAswitheitherashort(29nt)oralong(139nt)poly(A)tail(Figure1A).ThesemRNAsweremadebyinvitrotranscriptionfromDNAtemplatesthatencodedthemRNAbodyfollowedbythepoly(A)tailaswellasthehepatitisdeltavirus(HDV)self-cleavingribozyme,whichcleavedduringinvitrotranscriptiontogeneratenotonlyadefined3′endatthedesiredpoly(A)-taillengthbutalsoa2′−3′-cyclicphosphatedesignedtoinhibitundesiredlengtheningorshorteningofthetail(Avisetal.,2012). Figure1with2supplementsseeall Downloadasset Openasset PABPCoverexpressionuncouplespoly(A)-taillengthandTEinfrogoocytes. (A)SchematicofcappedT7transcriptswithtwodifferenttaillengths,whichwereusedasreportermRNAs.AdditionalsequencesbeyondtheHDVsequencearenotdepicted.(B)Theeffectoftaillengthonrelativeyieldsofinvitrotranslationofshort-andlong-tailedNlucreportermRNAs,ineitherfrogoocyteextract(left)orrabbitreticulocytelysate(right).Thenumberaboveeachbracketindicatesthefolddifferenceofthemeannormalizedluciferasesignal(errorbars,standarddeviationfromthreetechnicalreplicates).(C)TheeffectofpurifiedPABPC1onrelativeyieldsofinvitrotranslationofshort-andlong-tailedNlucreportermRNAsinfrogoocyteextract.PurifiedeGFPandPABPC1wereeachaddedasindicated.Otherwise,thispanelisasin(B).(D)Experimentalschemeforserial-injectionofmRNAsintofrogoocytes.(E)TheeffectofoverexpressingPAPBC1andaPABPC1M161AmutantonrelativetranslationofNlucreportermRNAsinfrogoocytes.DifferentialPABPC1expressionwasachievedbyinjectingthe indicatedamountofmRNAinthefirstinjection(errorbars,standarddeviationfromthreebiologicalreplicates).Otherwise,thispanelisasin(B).(F)TheeffectofPAPBC1overexpressionontranslationofreportermRNAswithdifferent3′-endstructuresinfrogoocytes.ShownarerawluciferaseyieldsfromNlucreportersthathaveeitherashortpoly(A)tail,alongpoly(A)tail,ahistonemRNA3′-endstem-loop,oraMalat1triple-helix3′-endinoocytesoverexpressingeithereGFPorPABPC1(errorbars,standarddeviationsfromthreebiologicalreplicates).pvaluesarefromone-sidedt-tests(n.s.,notsignificant).Foroverexpression,2.4fmolmRNAwasinjectedperoocyte. Figure1—sourcedata1 SourcedataforluciferasevaluesshowninFigure1andFigure1—figuresupplement2. https://cdn.elifesciences.org/articles/66493/elife-66493-fig1-data1-v1.xlsx Downloadelife-66493-fig1-data1-v1.xlsx Whenaddedtothefrogoocyteextractstogetherwithafireflyluciferase(Fluc)mRNA,usedtonormalizeforoveralltranslationactivity,thelong-tailedreporterwastranslatedsubstantiallybetterthanwastheshort-tailedreporter(Figure1B).Incontrast,thesamereportermRNAsweretranslatednearlyequallywellinrabbitreticulocytelysate,apost-embryonicdifferentiatedsystemforwhichnocouplingbetweentaillengthandTEwasexpected(Subtelnyetal.,2014; Figure1B).SimilarresultswereobservedforananalogouspairofRenillaluciferasereportermRNAs(Figure1—figuresupplement2A).Inboththeoocyteandreticulocytesystems,thereportermRNAswerestablewithnodetectablechangestotheirtaillengths(Figure1—figuresupplement1A–C).Thus,thelargedifferenceinluciferasesignalobservedbetweentheshort-andlong-tailedreportersintheoocyteextractwasattributabletoadifferenceinTE.Theseresultsshowedthatthecausalrelationshipbetweenlongerpoly(A)-taillengthandgreaterTEobservedforsomematuration-specificmRNAsinfrogoocytes(Sheetsetal.,1995;Stebbins-BoazandRichter,1994)isnotuniquetothosemRNAs,andindicatedthatfrogoocyteextractsprovideasystemforprobingthemechanismthatcouplestaillengthtoTE. Whenconsideringthepotentialmechanismsforreadingouttaillengthandpromotingtranslation,aroleforPABPCseemedplausible.Forinstance,translationmightbesensitivetothenumberofPABPCmoleculesassociatedwithanmRNA.Inonemechanisticpossibility,PABPCmightbeinexcessoveritsbindingsiteswithintails,suchthattailsarecoatedwiththeprotein,asisgenerallythoughttooccur(KühnandWahle,2004;Mangusetal.,2003),inwhichcase,mRNAswithlongertailsmightbedetectedasthoseabletobindmorePABPCmolecules.Atanothermechanisticextreme,PABPCmightbelimiting,suchthatmRNAscompetewitheachotherforPABPCbinding,inwhichcase,thosewithlongpoly(A)-taillengthswouldcompetemoreeffectivelyandtherebypreferentiallybenefitfromanyenhancementinTEthatPABPCbindingconfers.Todistinguishbetweenthesepossibilities,weincreasedavailablePABPCinouroocyteextracts,reasoningthatifPABPCwerealreadycoatingthetails,addingmorewouldhavelittleeffect,whereasifPABPCwerelimiting,addingmorewoulddiminishthecompetitionforPABPCbindingandtherebyreducethedifferenceinTEobservedbetweenshort- andlong-tailedmRNAs.Accordingly,wepurifiedrecombinantXenopusPABPC1tonearhomogeneity(Figure1—figuresupplement1D)andexamineditsinfluencewhenaddedtotheinvitrotranslationextractderivedfromstageVIoocytes.AsmorePABPC1wasadded,translationoftheshort-tailedreporterincreased,withlittlechangeintranslationofthelong-tailedreporter,whereasaddingequivalentamountofeGFPhadlittleimpactontranslationofeitherreporter(Figure1C).Thisconcentration-dependentdiminutionofcouplingbetweentail-lengthandTEstronglysupportedthehypothesisthatlimitingPAPBCisrequiredforstrongcoupling. ToinvestigatewhetherthisrequirementoflimitingPABPCwasrestrictedtoourinvitroextractsorwhetheritalsoappliedtolivingoocytes,weperformedserial-injectionexperimentsinoocytes.StageVIfrogoocyteswerefirstinjectedwitheitherPABPC1mRNAoracontrol,andafterwaiting24hrtoallowPABPC1proteintoaccumulate(Figure1—figuresupplement1E),oocyteswereinjectedwiththereportermRNAsandassayedforluciferaseactivity(Figure1D).WhereasinjectingthecontrolmRNA,eGFP,hadnomoreinfluencethaninjectingwater,injectingPABPC1mRNAsignificantlyreducedtheextenttowhichpoly(A)-taillengthandTEwerecoupled(Figure1E).SimilarresultswereobservedforananalogouspairofRenillaluciferasereportermRNAsorwheninjectingePABmRNAratherthanPABPC1mRNA(Figure1—figuresupplement2B).Reporterpoly(A)-taillengthsdidnotchangeoverthecourseoftheexperiment(Figure1—figuresupplement1F),whichindicatedthattheincreasedrelativetranslationoftheshort-tailedreportermRNAwasnotduetoelongatedpoly(A)tails. IntroducingadditionalPABPCintofrogoocytesspecificallyimprovedtranslationoftheshort-tailedreporterwhilehavinglittleeffectontranslationofeitherthelong-tailedreporterorreportersforwhichtailswerereplacedwitheitherastem-loopfromthe3′endofahistonemRNA(Lingetal.,2002)oratriple-helixfromthe3′endoftheMalat1non-codingRNA(Wiluszetal.,2012; Figure1F).TheobservationthatmRNAsrequiredatailtobenefitfromaddedPABPCindicatedthattheeffectsofaddingPABPCweremediatedincisthroughtail-boundPABPCmolecules,andweredirectandnotsomesecondaryconsequenceofalteringtranslation.Moreover,theobservationthatPABPChadlittleeffectontranslationoflong-tailedmRNAssuggestedthatthesemRNAscompetedforthelimitingendogenousPABPCsoeffectivelythatbindingofadditionalPABPCimpartednodetectableadditionalbenefittotheirtranslation. IntroducingPABPC1(M161A),whichencodesaPABPC1mutantthatisunabletobindeIF4G(GroftandBurley,2002),alsodiminishedcouplingbutdidsobyrepressingtranslationofthelong-tailedreporter.Thereducedtranslationofthelong-tailedreporterwaspresumablyduetoadominant-negativeeffectofreplacingfunctionalendogenousPABPCmoleculeswithdefective ones.Theobservationthatthelong-tailedreporterwaspreferentiallyaffectedagreedwithourconclusionthatendogenousPABPCwaslimitingandpreferentiallybindingtolong-tailedmRNAs.TheideathattheM161AmutantwasunabletoenhancetranslationinfrogoocytesimpliedthattheabilityforPABPCtobindeIF4Gandformtheclosed-loopstructureisimportantforenhancingtranslationinthiscontext (Wakiyamaetal.,2000). Insummary,ourresultswithreportersinoocytesandoocyteextractsconfirmedboththepositiveeffectofPABPContranslationandthecausalrelationshipbetweenpoly(A)-taillengthandTEinthesesystems.Moreover,theseresultsrevealedthatstrongcouplingbetweenpoly(A)-taillengthandTErequireslimitingPABPC. ToexaminetheglobaleffectofincreasingPABPConthetranslationalregulatoryregimeactingintheoocyte,wemonitoredtherelationshipbetweentaillengthandTEforendogenousmRNAsoftheoocytes.Asexpectedfromresultsofsingle-geneexperimentsinfrogoocytes(Figure1E; Sheetsetal.,1995;Stebbins-BoazandRichter,1994)andthestrongcouplingbetweenpoly(A)-taillengthandTEobservedinbothfrogembryosandflyoocytes(Eichhornetal.,2016;Limetal.,2016;Subtelnyetal.,2014),wefoundthatpoly(A)-taillengthcorrelatedstronglywithTEinstageVIfrogoocytes(Figure2A).OverexpressingeitherPABPC1orePABintheseoocytessignificantlydiminishedthecoupling,withtheSpearmancorrelation(Rs)fortherelationshipbetweentaillengthandTEdroppingfrom0.62to0.36and0.38,respectively(Figure2A,bothp = 0,modifiedDunnandClark’sz-test[DiedenhofenandMusch,2015]).Incontrast,overexpressingeGFPhadnosignificantimpactonthecoupling(p = 0.11),whichindicatedthatthistranscriptome-wideeffectwasaresultofadditionalPABPCproteinratherthananon-specificeffectofaddingmoremRNA. Figure2with1supplementseeall Downloadasset Openasset IncreasedPABPCpromotestranslationofendogenousshort-tailedmRNAs,therebydiminishingcouplingbetweentaillengthandTE. (A)TheeffectofPABPConcouplingbetweentaillengthandTEinfrogoocytes.ShownistherelationshipbetweenTEandmedianpoly(A)-taillengthinoocytesinjectedwitheitherwaterormRNAsencodingeithereGFP,PABPC1,orePAB(injecting16fmolmRNAperoocyte).ResultsareshownformRNAsfromgeneswith ≥100poly(A)tags.Eachpoly(A)tagrepresentsapairofsequencingreadsthatidentifythemRNAanditspoly(A)taillength.RsistheSpearman’scorrelationcoefficient.(B)AglobalincreaseinTEobserveduponoverexpressingePAB.TEinePAB-overexpressingoocytesiscomparedwithTEineGFP-expressingoocytes.ExperimentwasasinA,exceptrRNAswerenotdepleted,andTEvaluesarenormalizedtothoseofmitochondrialmRNAs.Alsoshownareresultsfromaco-injectedshort-tailedNlucreporterandalong-tailedFlucreporter.(C)EffectofexpressingPABPC1onproteinsynthesisinoocytes.AttheleftisagelshowingtotalproteinafterinjectionofeitherwaterortheindicatedmRNA(injecting4fmolmRNAperoocyte),withorwithouttreatmentwithcycloheximide(CHX),asvisualizedbyCoomassiestaining.Attherightisthesamegelshowingproteinsynthesis,asvisualizedbyincorporationof35S-methionineand35S-cysteine.ProminentbandspresumablyrepresentPABPC1,eGFP,andanotfullydenaturedformofeGFP(asterisk)expressedfrominjectedmRNAs.(D)QuantificationoftheeffectofexpressingPABPC1onproteinsynthesisinoocytes,asmeasuredin(C)andtwoadditionalbiologicalreplicates.OnlyregionsabovethePABPC1band,andbetweenthePABPC1bandandthetopeGFPband,wereusedforquantification.Valueswerenormalizedtothatofthemeanvaluefromwater-injectedoocytes(errorbars,standarddeviation;pvalues,one-sidedt-tests;n.s.,notsignificant).(E)ThepreferentialeffectofoverexpressingePABontheTEofshort-tailedmRNAs.TEfoldchangesobservedbetweenePAB-overexpressingandeGFP-expressingoocytesareplottedasafunctionofmediantaillengthineGFP-expressingoocytes.TEandtail-lengthvalueswereobtainedfromdifferentbatchesofoocytes;resultsareshownformRNAsfromgeneswith ≥100poly(A)tags.TEvaluesarenormalizedtothoseofmitochondrialmRNAs.(F)EffectofoverexpressingePABonthedistributionofTEvaluesobservedinfrogoocytes.ShownistheTEdistributionobservedinePAB-overexpressingoocytesandthatobservedineGFP-expressingoocytes.Tail-lengthmeasurementsinthisfigurewereobtainedusingTAIL-seq. Figure2—sourcedata1 SourcedataforvaluesshowninFigure2D. https://cdn.elifesciences.org/articles/66493/elife-66493-fig2-data1-v1.xlsx Downloadelife-66493-fig2-data1-v1.xlsx AccompanyingthereducedcouplingobserveduponPABPCoverexpressionwasasignificantrelativeincreaseofTEforshort-tailedmRNAs,aneffectnotobservedineGFP-expressingoocytes(Figure2—figuresupplement1A).ThisTEincreasewasnotaccompaniedbycorrespondinglengtheningofpoly(A)tails(Figure2—figuresupplement1B),implyingthattail-lengthchangesdidnotcausethese relativeTEchanges.TomakecomparisonsofabsoluteTEchanges,werepeatedtheePAB-overexpressionexperimentbutomittedrRNAdepletionduringsequencinglibraryconstruction,therebyallowingustonormalizeTEusingmitochondrialmRNAs(Iwasakietal.,2016),whichwereotherwisedepletedbyIlluminaRibo-Zerokits(Figure2—figuresupplement1C).Inthisexperiment,wealsoinjectedoocyteswithashort-tailedNlucmRNAreporterandalong-tailedFlucmRNAreporterandmonitoredtheirabsoluteTEchangestogetherwiththoseofendogenousmRNAs.MostendogenousmRNAshadgreaterabsoluteTEinePAB-overexpressingoocytescomparedtoeGFP-expressingcontroloocytes(Figure2B).Thisresultwasconsistentwith35Smetabolic-labelingexperimentsshowingthatoverexpressionofPABPC1butnoteGFPsignificantlyincreasedglobalproteinsynthesisinoocytes(Figure2C–D).Moreover,themagnitudeoftheTEincreaseconferredbyePAB-overexpressionnegativelycorrelatedwithtaillength,whichshowedthattranslationofshort-tailedmRNAsimprovedsubstantiallymorethanthatoflong-tailedmRNAs(Figure2E),asobservedforourco-injectedreporters.Indeed,addingePABhadessentiallynooveralleffectonTEofendogenousmRNAswiththelongesttails(medianTEfoldchange = 1.06forthe54mRNAswithmediantaillengths > 80nt),asobservedforourlong-tailedreporters.ThepreferentialimprovementofTEsforshort-tailedmRNAsledtonotonlyanoverallshiftinTEbutalsonarrowingoftheTEdistribution(Figure2F)tomorecloselyresemblethedistributionsobservedincellsinwhichpoly(A)-taillengthandTEarenotcoupled(Subtelnyetal.,2014).TheseresultssupportedthehypothesisthatincreasingPABPCinoocytesincreasestheopportunityforshort-tailedmRNAstobindaPABPCmolecule,therebypromotingtranslation. Overall,theresultsofourglobalanalysesofmRNAsinfrogoocytesagreedwiththoseofreporterassays,therebyextendingtoendogenousmRNAssupportfortheconclusionthatlimitingPABPCplaysacriticalroleinconferringstrongcouplingbetweenpoly(A)-taillengthandTE. Ourglobalanalysisexaminingtherelationshipbetweenpoly(A)-taillengthandTEofendogenousmRNAsinoocytesdifferedfromourreporterassaysinthatthecomparisonwasmadebetweenmRNAsofdifferentgenes,whichcanbeconfoundedbyfeaturesotherthantaillengththatvarybetweenthesemRNAs.Toovercomethisissue,wedevelopedahigh-throughputmethodforcomparingeffectsondifferenttail-lengthisoformsfromeachgene.Thisapproachforintragenicanalyses,calledPAL-TRAP(Poly(A)tail-LengthprofilingfollowingTranslatingRibosomeAffinityPurification),resembledotherTRAPapproachesinthatribosomesweresparselytaggedsuchthattheirimmunoprecipitation(IP)preferentiallyisolatedmRNAisoformsassociatedwithmoreribosomes,whichwereinferredtobemorehighlytranslated(ChenandDickman,2017;Heimanetal.,2008).Inasysteminwhichpoly(A)-taillengthandTEwerecoupled,longer-tailmRNAswereexpectedtobeassociatedwithmoreribosomesandthereforeenrichedintheeluate(Figure3A),whereasinanuncoupledsystem,longer-tailmRNAswerenotexpectedtobeenrichedintheeluate. Figure3with1supplementseeall Downloadasset Openasset LimitingPABPCisrequiredforintrageniccouplingbetweenpoly(A)-taillengthandTE. (A)ThePAL-TRAPmethodformeasuringintrageniceffectsoftaillengthonTE.Ribosomesaresparselytagged(redstars)sothathighlytranslatedmRNAsaremorelikelytocontaintaggedribosomesandthusbeenrichedintheimmunoprecipitation(IP)eluate.TaillengthsofbothinputandeluatemRNAsweremeasuredandcomparedformRNAsofeachgene.Thedepictedenrichmentoflong-tailedisoformsintheeluateindicatesthatpoly(A)-taillengthandTEarecoupled,whereasnoenrichmentwouldindicateotherwise.(B)TheexperimentalschemeofPAL-TRAPinfrogoocytes.SeeFigure3—figuresupplement1B–CforresultsfrompulldownsusingtheeGFP-HAcontrol.(C)EffectofoverexpressingePABoncouplingbetweentaillengthandTE,asdetectedafterpoolingPAL-TRAPresultsformRNAsfromdifferentgenes.Plottedarecumulativedistributionsofpoly(A)-taillengthsinthePAL-TRAPinputandeluateobtainedafterexpressingeithereGFP(left)orePAB(right)inoocytes(injecting4fmolmRNAperoocyte).Medianvaluesareindicated(dashedlines)andlistedinparentheses.(D)TheeffectofoverexpressingePABonintrageniccouplingbetweentaillengthandTE.PlottedforeachmRNAisoformisthemedianpoly(A)-taillengthofmRNAsinthePAL-TRAPeluatecomparedtothatintheinput.ShownareresultsformRNAsfromoocyteseitherexpressingeGFPoroverexpressingePAB(leftandright,respectively).EachpointrepresentsanmRNAisoformwithaunique3′endrepresentedby ≥100poly(A)tagsinbothinputandeluate.Alsoindicatedareresultsfor(1)anRlucreportermRNApossessingavariable-lengthtail(reporter),whichwasco-injectedwithmRNAsexpressingeithereGFPorePAB,(2)anmRNAwithavariable-lengthtail,whichwasspikedintothelysateimmediatelybeforeIP(spike-in),and(3)syntheticRNAswithdefinedtaillengthsaddedtosamplespriortolibrarypreparation(tailstandards).Pointsforeightstandardswithlongertailsfelloutsidetheplotareas,asdidapointrepresentingthemRNA3′-endfromonegene(uqcrb.S)intheePABsample.(E)SummaryofdifferencesinmediantaillengthsobservedbetweentheeluateandtheinputofmRNAisoformsshownin(D).Boxandwhiskersindicatethe10th,25th,50th,75th,and90thpercentiles.(F)EffectofoverexpressingePABonintragenictail-lengthdistributionsinfrogoocytes.Shownaretail-lengthdistributionsofthereporterRluc(left),anendogenousoocytemRNAbtg4.S(middle),andthespike-inmRNANluc(right)ineGFP-expressing(top)orePAB-overexpressing(bottom)oocytes.Mediantail-lengthvaluesareindicated(verticallines)andlistedinparentheses.Tail-lengthmeasurementsinthisfigurewereobtainedusingPAL-seqv3. ToimplementPAL-TRAP,wefirstinjectedstageVIfrogoocyteswithanmRNAencodingC-terminalHA-taggedRPL3(ChenandDickman,2017)andallowedtimeforRPL3-HAproteinexpressionandincorporationintoribosomes(Figure3B).Asacontrol,HA-taggedeGFPwasexpressedinseparateoocytes.ToexaminetherequirementoflimitingPABPCforcouplingbetweenpoly(A)-taillengthandTE,wetheninjectedoocyteswithmRNAsthatexpressedeithereGFPorePAB.ConfirmingthatRPL3-HAwasincorporatedintofunctionalribosomes,HA-IPfromRPL3-HA-expressingoocytesenrichedforproteinsfrombothribosomalsubunits,cytoplasmicmRNAs,andePAB,whereasHA-IPfromeGFP-HA-expressingoocytesdidnot(Figure3—figuresupplement1A–C).IncontroloocytesexpressingeGFP,longer-tailmRNAswereenrichedintheeluatecomparedtotheinput(mediantaillengths41and36nt,respectively),whichreflectedthecouplingbetweentaillengthandTE(Figure3C).OverexpressionofePABreducedthisenrichmentforlonger-tailmRNAsintheeluatecomparedtotheinput(mediantaillengths37and35nt,respectively),asexpectediflimitingPABPCwasrequiredforthiscoupling(Figure3C).Wealsoanalyzedtheflowthroughfractionsfromthesepulldownexperiments.ForbotheGFP-andePAB-expressingoocytes,thecumulativedistributionofpoly(A)-taillengthsfromtheflowthroughwasnearlyidenticaltothatfromtheinput(Figure3—figuresupplement1D),asexpectedwhenconsideringthatonlyasmallfractionofinputwasdepletedbyHA-beads.Theseanalysesindicatedthatourmethodswereabletocapturesmalldifferencesintail-lengthdistributions. Whenanalyzing,foreachgene,mRNAswithdifferenttail-lengthisoforms,mRNAsfrommostgenes(84.4%)hadlongermedianpoly(A)-taillengthsintheeluatethanintheinput(Figure3D–E),implyingthatformRNAsfrommostgenes,long-tailedisoformsweretranslatedmoreefficientlythanshort-tailedones.Althoughthemedianofthemediantail-lengthdifferenceswasmoderate(+2.5nt),28.3%mRNAshada ≥ 5ntlongermediantaillengthintheeluate(Figure3E).Incontrast,wheninputandflowthroughwerecompared,themediantail-lengthdifferencescenteredon0nt,andmRNAsfromonly5.4%ofgeneshada ≥ 5ntlongermediantaillengthinoneofthesamples,asexpectedwhenconsideringthatonlyasmallfractionofinputwasdepletedbyHA-beads(Figure3—figuresupplement1E–F).OverexpressingePABreducedthenumberofgeneswithlong-tailedisoformsenrichedintheeluateandshiftedthedistributionofmediantail-lengthdifferencescloserto0nt(Figure3D–E),asexpectedifcouplingbetweenpoly(A)-taillengthandTEdiminished. InthesePAL-TRAPexperiments,amixtureofRlucreportermRNAmoleculeswithdifferentpoly(A)-taillengthswasco-injectedwithmRNAsthatexpressedeithereGFPorePAB.IneGFP-expressingoocytes,longer-tailRlucisoformswerehighlyenrichedintheeluatecomparedtotheinput,whereasinePAB-overexpressingoocytes,thisdifferencediminisheddramatically(Figure3F).Asexpected,NlucreportermRNA,whichwasaddedtothelysateasaspike-induringHApulldown,wasnotsignificantlyenrichedinlonger-tailspeciesintheeluate,regardlessofthetreatment,whichindicatedthatthechangesobservedforRlucmRNAreflectedchangesoccurringintheoocyte,priortolysis(Figure3F).AlthoughendogenousmRNAsfromsomegenes,suchasbtg4.S,alsounderwentlargechangesinmediantaildifferencesinresponsetoePAB-overexpression(Figure3F),changesweretypicallysmallerforendogenousmRNAs(Figure3E),whichwasatleastpartlyattributabletothenarrowerrangeofinitialtaillengthisoformsforendogenousmRNAscomparedtotheinjectedRlucmRNA. Takentogether,thePAL-TRAPresultsrevealedintrageniccouplingbetweenpoly(A)-taillengthandTEforendogenousmRNAsaswellasreportersinstageVIfrogoocytes.Moreover,thesePAL-TRAPanalysesshowedthatasobservedbothinourreporterassaysandinourglobalintergeniccomparisons,substantialcouplingbetweentaillengthandTErequireslimitingPABPC. HavingestablishedthenecessityoflimitingPABPCforcoupling,weinvestigatedifitcouldalsobesufficient,that is,whetherlimitingPABPCcouldconfercouplingbetweenpoly(A)-taillengthandTEincellsinwhichthesefeatureswerenormallyuncoupled.Tothisend,weknockeddownPABPCsinHeLacells.BasedonavailablemRNA-seqandmassspectrometryresults(Nagarajetal.,2011),PABPC1andPABPC4arethetwomajorPABPCparalogsinHeLacells,withPABPC1fourtimesmoreabundantthanPABPC4andthetwotogetheraccountingfor >95%ofallPABPCinHeLacells.ConsistentwiththeideathatPABPCisnotnormallylimitinginuncoupledsystems,PABPC1isestimatedtobepresentinthreefoldexcessoverthepoly(A)sitesinHeLacells(Görlachetal.,1994).ToreducePABPCtolimitinglevels,weusedsiRNAstoreduceeitherPABPC1orPABPC4aloneorbothPABPC1andPABPC4by >90%(Figure4—figuresupplement1A)andexaminedtherelationshipbetweenmediantaillengthsandTE.KnockingdownPABPC4alonehadlittleimpactonthecoupling(Figure4—figuresupplement1B),consistentwiththeinferencethatPABPC4constituteslessthan20%ofthetotalPABPCprotein.Althoughcorrelationbetweenmedianpoly(A)-taillengthandTEgraduallyincreasedasmorePABPCwasdepleted,itreachedanRsofonly0.18(p<10–19)indouble-knockdowncells(Figure4—figuresupplement1B),whichwasmuchweakerthanthatobservedinfrogoocytes(Figure2A)andfrogandfishearlyembryos(Subtelnyetal.,2014).MinorRsincreaseswereobservedinothermouseandhumanpost-embryoniccelllinesinwhichPABPCwasdepleted,butstrongcouplingbetweenpoly(A)-taillengthandTEwasnotestablished,withnoRsvaluesexceeding0.3(datanotshown).Thus,otherconditionsinadditiontolimitingPABPCmustalsobemettoconferstrongcouplingbetweenpoly(A)-lengthandTE. AstrikingconsequenceofdepletingPABPCinHeLacellswasasharpincreaseinmedianpoly(A)-taillengths,whichforHeLamRNAsincreasedanaverageof17and39ntinPABPC1-anddouble-knockdowncells,respectively(Figure4A).Substantialchangeswerealsoobservedinthedistributionsofglobalpoly(A)-taillength,whichshowedthatmRNAswithtailsrangingfrom10to50ntwere >2-folddepletedinthePABPC1-knockdowncellsandthatmRNAswithtailsrangingfrom10to135ntwere2-to20-folddepletedinthedouble-knockdowncells(Figure4B).SimilarresultswereobtainedinNIH3T3cells(Figure4—figuresupplement2A–B).Incontrast,lossofshort-tailedmRNAswasnotobservedformitochondria-encodedmRNAs,asexpectedforthesemRNAsthatneverencounterPABPC(Figure4A–B,Figure4—figuresupplement2A–B),andexaminationofinternalstandardsandreplicatesconfirmedthatthelossofshort-tailedcytoplasmicmRNAswasnotattributabletoinaccurateorvariablemeasurements(Figure4—figuresupplement2C–D).Moreover,knockingdowntheminorisoform(PABPC4)alonedidnothaveasimilareffect(Figure4—figuresupplement2E),suggestingthatthetail-lengthchangesobservedforcytoplasmicmRNAswereaconsequenceoflimitingPABPC. Figure4with3supplementsseeall Downloadasset Openasset PABPCdepletioncausesprematuredecayofshorter-tailcytoplasmicmRNAsinHeLacells. (A)TheeffectofPABPCknockdownonpoly(A)-taillength.Theplotscomparemedianpoly(A)-taillengthsineitherPABPC1-knockdowncells(left)orPABPC1andPABPC4double-knockdowncells(right)tothoseincontrolcells.ResultsareshownforcytoplasmicmRNAswith ≥100poly(A)tags(gray)andformitochondrialmRNAs(red),mergingdataforMT-ATP6andMT-ATP8andforMT-ND4andMT-ND4L,whicharebicistronicmitochondrialmRNAs.(B)TheeffectofPABPCknockdownontheabundanceofmRNAswithdifferenttaillengths.Shownaretail-lengthdistributionsofallcytoplasmic(left)andmitochondrial(right)mRNApoly(A)tagsincontrol,PABPC1-knockdown,anddouble-knockdowncells.Foreachdistribution,theabundanceoftagswasnormalizedtothatofthespike-intail-lengthstandards.Duetodepletionoftail-lengthcallingatposition50,whichwasassociatedwithachangeinlaserintensityatthenextsequencingcycle,thevaluesatthistaillengthwerereplacedwiththeaverageofvaluesattaillengths49and51nt.(C)TheeffectofPABPCknockdownontheabundanceofmRNAswithdifferenttaillengths,comparingtail-lengthsmeasuredbysequencing(left)withthoseobservedonRNaseHnorthernblots(right).ResultsareshownforacytoplasmicmRNAGAPDH(top)andamitochondrialmRNAMT-CYB(bottom).Relativetagdensitywascalculatedbylog-transforminglineartagdensityusingnormalizedpoly(A)tagcounts.Mediantail-lengthvaluesareindicated(horizontallines)andlistedinparentheses.ForRNaseHnorthernblots,aDNAoligonucleotidecomplimentarytothe3′-UTRwasusedtodirectcleavageofthetargetmRNAbyRNaseH,leavinga35-ntfragmentofthe3′-UTRappendedtothepoly(A)tail,whichwasresolvedonadenaturinggelanddetectedbyaradiolabeledprobe.Taillengthsindicatedalongtheleftsideofeachgelareinferredfromlengthsofsizemarkers.(D)TheeffectofPABPCknockdownontheabundanceofmRNAswithdifferenttaillengths,extendingtheintragenicanalysistotail-lengthdistributionsfromthousandsofgenes.Heatmapscomparepoly(A)-taglevelsinPABPC1-knockdown(left)ordouble-knockdown(right)cellstothoseincontrolcells,afternormalizingtospike-intail-lengthstandards,asmeasuredusingtail-lengthsequencing.EachrowrepresentsmRNAsfromadifferentgene,androwsaresortedbasedonfoldchangeofmRNAabundancemeasuredusingRNA-seq.Onlygeneswith ≥100poly(A)tagsineachoftwosamplesbeingcomparedwereincludedintheanalyses(n = 5504).Columnsrepresentvaluesfrom5-nttail-lengthbinsrangingfrom0to244ntanda6-ntbinrangingfrom245to250nt.Tilecolorindicatesthefoldchangeofnormalizedtagcounts(key).(E)TheeffectofPABPCknockdownontheabundanceofmRNAswithdifferenttaillengths,reanalyzingdatafrom(D)toshowdistributionsofpoly(A)-tagchangesobservedatdifferenttail-lengths.Eachbox-whiskershowsthe10th,25th,50th,75th,and90thpercentileoffoldchangesinnormalizedpoly(A)-tagcountsobservedforeachtail-lengthbinof(D).Thecolorofeachboxindicatesthemedianvalue(key).Tail-lengthmeasurementsinthisfigurewereobtainedusingTAIL-seq. WealsousednorthernblotstoexamineeffectsofPABPCdepletionontail-lengthdistributionsofmRNAsfromindividualgenesandfoundthattheresultscorrespondedwelltothoseobservedbytail-lengthsequencing.Bothnorthernblotsandsequencingshowedstrongdepletionofshort-tailedisoformsforacytoplasmicmRNA(GAPDH)afterPABPCknockdownbutnosubstantialchangeinthetail-lengthdistributionofamitochondrialmRNA(MT-CYB)(Figure4C).Usingsequencingdatatoexaminetheintragenictail-lengthdistributionsofcytoplasmicmRNAfromeachofmorethanfivethousandothergenesrevealedfindingsresemblingthoseobservedforGAPDH(Figure4D–E).AfterknockingdownPABPC1,thereductioninshort-tailedmRNAswastypicallymostsevereformRNAisoformswithtaillengthsof ~25nt,andinthedoubleknockdown,thisdipat ~25ntbecamemorepronounced,withreductionsextendingtoallbutthelongest-tailisoforms(Figure4D–E).Indeed,formorethanhalfofthegenesexamined, ≥2-foldreductionsextendedtoisoformswithtailsaslongas135nt(Figure4D–E).Similarresultswereobservedwhenexaminingtail-lengthdistributionsintheNIH3T3dataset(Figure4—figuresupplement2B),andwhenexaminingindividualtail-lengthdistributionsformRNAsofeachofthetop-expressednucleargenesbutnotwhenexaminingthoseformRNAsofmitochondrialgenes(Figure4—figuresupplement2F–G). Theseresults,whichshowedthatmRNAswithshorttailswerepreferentiallydestabilizedwhenPABPCwasdepletedprovidedgeneticloss-of-functionevidencethatPABPCstabilizesmRNAsinmammaliancelllines.AlthoughgeneticevidencefortheroleofPABPCinmRNAstabilityhasbeenreportedinyeast(CaponigroandParker,1995;Colleretal.,1998),thisfunctionhadnotbeenestablishedinmammaliancells.AlthoughinprinciplethemRNAdestabilizationthatweobserveduponPABPC-depletionmighthavebeenindirect,twolinesofevidencesupporttheconclusionthatthisdestabilizationwasadirectconsequenceofthelossofPABPCbindingtopoly(A)tails.First,destabilizationpreferentiallyoccurredforshort-tailedmRNAs,whichwereexpectedtobetheleastsuccessfulatcompetingforbindingunderconditionsoflimitingPABPC.Second,destabilizationsharplydiminishedattaillengthsof10–15nt,whichcorrespondedtothe12ntpoly(A)sequencereportedtobetheminimallengththatcanbeboundbyPABPC1withhighaffinity(KühnandPieler,1996).Indeed,themodestlossobservedformRNAswithveryshortpoly(A)tails(Figure4D–E)suggestedthatevenincontrolcellsthathadabundantPABPC,mRNAswithtailsshorterthan12ntwerepoorlyboundbyPABPCs,andthusPABPCdepletiondidnotsubstantiallyinfluencetheirabundance. Arecentstudyobservedsimilarpoly(A)tail-lengthchangesinPABPC1-depletedcellsbutattributedthesechangestoimpaireddeadenylation(Yietal.,2018).BecausePABPCcanpromotedeadenylationinvitro(Schäferetal.,2019;Uchidaetal.,2004;Websteretal.,2018;Yietal.,2018),itisconceivablethatthelossofPABPCwouldslowdeadenylation,therebyincreasingmRNAmediantaillengths,asobservedinPab1-knockoutyeast(CaponigroandParker,1995).However,ouranalyses,whichhadthebenefitofquantitativetailstandardsthatenabledmeasurementofabsoluteabundancechanges,revealedlittleaddedaccumulationoflong-tailedisoformsinPABPC-depletedcells(Figure4B–E,Figure4—figuresupplements2B, F),indicatingthatadeadenylationdefectwasnotthemajorcausefortheperturbedtail-lengthdistributions.Moreover,wefoundthatmRNAhalf-lifevalues,asdeterminedbymetaboliclabeling,reducedsignificantlywhenPABPCwasknockeddown(Figure4—figuresupplement3A–C),whichconcurredwiththeconclusionthatPABPCknockdowndestabilizedshort-tailedmRNAisoformsandarguedagainstthepreviousassertionthatPABPCknockdownimpaireddeadenylation,inthatimpaireddeadenylationwouldhavelengthenedmRNAhalf-lives. Takentogether,theseresultsshowthatPABPCbindingstabilizesmRNAsofculturedmammaliancells;ifPABPCbecomeslimitinginthesecells,theshort-tailedmRNAsbecomedestabilized,presumablybecausetheycompetelesseffectivelyforPABPC.Mostimportantly,thedestabilizationofmRNAsthatcompetedpoorlyforPABPCbindinghelpsexplainwhylimitingPABPCwasinsufficienttocausestrongcouplinginmammaliancelllines,inthatstrongcouplingbetweentaillengthandTEwouldbedifficulttoestablishinaregulatoryregimeinwhichshort-tailedmRNAmoleculesthatlackPABPCbindingaredegradedratherthantranslatedlessefficiently.Thus,theseresultsidentifyasecondmechanisticrequirementforstrongcouplingbetweentaillengthandTE:InadditiontolimitingPABPC,strongcouplingrequiresmetabolicstabilityofthemRNAsthatcompetepoorlyforPABPCbinding. TheidentificationofthissecondrequirementforstrongcouplingbroughttotheforethequestionofwhymRNAsthatcompetedpoorlyforlimitingPABPCweredestabilized.Toexplorethepossiblemechanisms,wesearchedforperturbationsthatcouldrestorestabilityofshort-tailedmRNAsinHeLacellsundergoingPABPCknockdown,monitoringtail-lengthdistributionsofendogenousGAPDHusingnorthernblots.Asapositivecontrol,expressingansiRNA-resistantPABPC1restoredstabilityofshort-tailedspecies,asdidfrogePAB,whichfurtherillustratedfunctionalconservationofPABPCfromdifferentspeciesanddevelopmentalstages(Figure5A).Interestingly,aPABPC1variantwithsubstitutionsthatdisruptitsinteractionwitheIF4G(Chorghadeetal.,2017)alsorestoredthestabilityofshort-tailedspecies,implyingthattheclassicalclosedloopisnotnecessaryforPABPC1toprotectshort-tailedmRNAsfromdegradation.BecausePABPChasbeenimplicatedininhibitingmRNAterminaluridylationinvitro(Limetal.,2014)andincells(Yietal.,2018),andterminaluridylationhasbeenlinkedtomRNAdecay(Limetal.,2014;RisslandandNorbury,2009),weaskedifterminaluridylationcontributedtothelossofshort-tailedmRNAs.KnockingdownbothTUT4andTUT7inPABPC-depletedcellspartiallyrestoredshort-tailedGAPDHmRNAs(Figure5B).SimilarresultswereobservedinHCT116cells,inwhichwetaggedendogenousPABPC1withanauxin-inducibledegron(AID)andinduceddepletionbyaddingindole-3-aceticacid(IAA,aformofauxin)(Figure5C; Natsumeetal.,2016). Figure5with1supplementseeall Downloadasset Openasset DepletionofTUT4andTUT7attenuatesprematuredecayofmRNAcausedbyPABPCdepletion. (A)Rescueoflossofshorter-tailmRNAsinPABPC-depletedHeLacellsbyexpressingeitheransiRNA-resistanthumanPABPC1,ansiRNA-resistantPABPC1codingforamutantthatdoesnotbindeIF4G(4G_mut),oransiRNA-resistantfrogePAB.AtthetopleftisanRNaseHnorthernblotprobedforGAPDH,asinFigure4C.Atthetoprightaretheintensitylevelsquantifiedfromtheblot.AtthebottomisawesternblotprobedfortheindicatedproteinsortheFLAGtagappendedtotheCterminusofeachoftheexpressedPABPCproteins.(B)Partialrescueoflossofshorter-tailmRNAsinPABPC-depletedHeLacellsbyknockingdownTUT4andTUT7.AtthetopleftisanRNaseHnorthernblotprobedforGAPDH,asinFigure4C.Atthetoprightaretheintensitylevelsquantifiedfromtheblot.Atthebottomisawesternblotprobedfortheindicatedproteins.(C)Partialrescueoflossofshorter-tailmRNAsinPABPC-depletedHCT116cellsbyknockingdownTUT4andTUT7.EndogenousPABPC1wastaggedwithAID,andIAAwasaddedfor24hrtotargetthefusionproteinfordegradation.Otherwise,thispanelisasin(B).(D)TheeffectofPABPCknockdownandTUTknockdownonterminaluridylation.PlottedisthefractionofuridinesneartheterminiofcytoplasmicmRNAsinHeLacellstransfectedwiththeindicatedsiRNAs,asmeasuredusingtail-lengthsequencing.(E)TheeffectofPABPCknockdownandTUTknockdownonterminaluridylationoftailswithdifferentlengths.ShownisterminaluridylationfrequencyofcytoplasmicmRNAsasafunctionofpoly(A)-taillengthinHeLacellstransfectedwiththeindicatedsiRNAs,asmeasuredusingtail-lengthsequencing.Theinsetshowsahigher-resolutionviewoftheshort-tailregion.(F)Partialrescueoflossofshorter-tailcytoplasmicmRNAsinPABPC-depletedHeLacellsbyknockingdownTUT4andTUT7,extendingtheanalysistoallpoly(A)tags.Plottedaretail-lengthdistributionsfromHeLacellstransfectedwiththeindicatedsiRNAs.Foreachdistribution,tagcountswerenormalizedtospike-intail-lengthstandards.Medianvaluesareindicated(dashedlines)andlistedinparentheses.Tail-lengthmeasurementswereobtainedusingPAL-seqv3. Toexaminetheglobalrescueofshort-tailedmRNAsandatthesametimemonitormRNAterminaluridylationlevels,wemodifiedourtail-lengthsequencingprotocolbyincludingintheadaptor-ligationstepasplintoligonucleotidedesignedtoaccommodatetailswitha3′terminalU(Eisenetal.,2020).KnockdownofPABPC1alonesignificantlyincreasedtheterminaluridylationlevelsacrossessentiallyalltail-lengthisoforms(Figure5D–E),consistentwithapreviousreport(Yietal.,2018).KnockdownofPABPC4inadditiontoPABPC1furtherincreaseduridylationofmRNAisoformswithlongertail lengths(Figure5E).KnockdownofTUT4andTUT7inPABPC-depletedcellsbroughtterminaluridylationofalltail-lengthisoformstobackgroundlevels(Figure5D–E)and,moreimportantly,preferentiallyrescuedshorter-tailisoforms,therebydecreasingmediantaillengths(Figure5F).Theseresultswereconsistentwiththoseofournorthernassays,andtogether,ourresultsindicatedthatinthesemammaliancells,limitingPABPCmakesshort-andmedium-tailedmRNAisoformsthatpoorlycompeteforPABPCbindingmoresusceptibletoterminaluridylation,therebyacceleratingtheirdecay. Havingfoundthatdestabilizationofshort-tailedmRNAsdampenedcouplingbetweenpoly(A)-taillengthandTEinPABPC-depletedmammaliancells,akeyquestionremainedregardinghowmRNATE,ifnotinfluencedstronglybypoly(A)-taillength,wasaffectedinthesecells.Toanswerthisquestion,weconductedglobalprofilingofHeLacells48hraftersiRNAtransfection,atimepointatwhichPABPC1andPABPC4knockdownsweresubstantialbutsecondaryeffectswerepresumablynotyettoosevere(Figure6—figuresupplement1A).WeagainimplementedRNA-seqandribosome-profilingprotocolsthatenabledabsoluteTEcomparisonbyusingmitochondrialmRNAsfornormalization(Rooijersetal.,2013).Surprisingly,near-completedepletionofPABPCshadnodetectableeffectonglobalmRNATEs(Figure6A–B).Althoughtheproteinsynthesisrateindouble-knockdowncellswasreducedto75.7%ofthatobservedincontrolcells,asmeasuredbyaveragingribosome-footprintchangesobservedforthe9697analyzedgenes(Figure6A–B),whichagreedwithresultsofaglobalpuromycin-basedtranslationassay(78.7%)(Figure6C),thisreducedproteinsynthesiswasfullyexplainedbythedecreaseinmRNAlevels,asindicatedbyadistributionofTEchangesthatcenterednearzero(Figure6A–B).Examinationofourribosome-footprintingdatarevealednoupregulationofotherPABPCparalogs,althoughtheTEofPABPC1mRNAincreased2.5-fold(Figure6A),consistentwithitsknownautoinhibitorytranslationalcontrol(BagandWu,1996).Resultsofpolysomeprofilingconfirmedthoseofoursequencing-basedmethods,inthatthereductionintranslationoutput,asmeasuredbytheheightofpolysomepeaksindouble-knockdowncells,wasattributabletoanoveralldecreaseofmRNAlevelsratherthantodecreasedTEsthatwouldotherwisecauseashiftofmRNAdistributionfromheavytolightfractions(Figure6—figuresupplement1B).Together,theseresultsindicatedthatPABPCdepletioninHeLacellshadnegligibleeffectonmRNATE. Figure6with2supplementsseeall Downloadasset Openasset DepletionofPABPCinmammaliancelllineshasminimaleffectonTE. (A)EffectofPABPCknockdownonmRNAabundance(left),ribosome-footprintabundance(middle)andTE(right)inHeLacells,comparingvaluesindouble-knockdowncellstothoseincontrolcells.Foreachgene,valuesformRNAandribosome-footprintreadsperkilobaseareplottedafternormalizingtomeasurementsformitochondrialmRNAs.(B)DistributionsoftheeffectsofPABPCdoubleknockdownonmRNAabundance,ribosome-footprintabundanceandTE.Eachbox-whiskershowsthe10th,25th,50th,75th,and90thpercentileofthefoldchangesobservedin(A).(C)EffectofPABPCknockdownonproteinsynthesisinHeLacells.Plottedarerelativelevelsofproteinsynthesismeasuredbypulsepuromycinincorporation48hraftertransfectionwiththeindicatedsiRNAs(Figure6—figuresupplement1A;errorbars,standarddeviationofthreebiologicalreplicates;pvalues,one-sidedt-tests;n.s.,notsignificant).(D)WesternblotshowingrapiddegradationofPABPC1-AIDfusionproteinafteraddingIAAtogenome-engineeredHCT116cells.PABPC1proteinlevelswerequantifiedbyaveragingsignalsforAIDandPABPC1,afternormalizingtothatforGAPDH.(E)EffectofPABPC1depletiononabundanceofmRNAswithdifferenttaillengths.Shownaretail-lengthdistributionsofallpoly(A)tagsobtainedfromcytoplasmicmRNAsofHCT116PABPC1-AIDcellsaftertreatmentwithIAAfortheindicatedtime.Foreachdistribution,theabundanceoftagswasnormalizedtothatofthespike-intail-lengthstandards.Medianvaluesareindicated(dashedlines)andlistedinparentheses.Duetodepletionof101-nttaillengths,whichwasassociatedwithachangeinlaserintensityatthenextsequencingcycle,thevaluesatthislengthwerereplacedwiththeaverageofvaluesattaillengths100and102nt.(F)SummaryoftheeffectsofrapidPABPC1depletion.Foreachgene,ateachtimepointafteraddingIAAtoHCT116PABPC1-AIDcells,valuesformRNAabundance,ribosome-footprintabundanceandTEwereeachcomparedtothevalueobservedincellsnottreatedwithIAA(Figure6—figuresupplement1C),andthedistributionoffoldchangesisplotted.Eachbox-whiskershowsthe10th,25th,50th,75th,and90thpercentile.Tail-lengthmeasurementswereobtainedusingPAL-seqv4. Figure6—sourcedata1 SourcedataforvaluesshowninFigure6CandFigure6D. https://cdn.elifesciences.org/articles/66493/elife-66493-fig6-data1-v1.xlsx Downloadelife-66493-fig6-data1-v1.xlsx TheslowdynamicsofsiRNA-mediatedknockdown,whichdictatedtherelativelylate48hrtimepointforourglobalmeasurements,mighthavepreventeddetectionofanyTEchangesthathappenedearlier,beforeanewsteadystatehadbeenreached.Toexaminethispossibility,wemonitoredthedynamicsoftail-length,mRNA-abundance,andtranslationchangessoonafterPABPCdepletion,usingtheHCT116PABPC1-AIDdegroncellline,inwhichPABPC1wasrapidlyandefficientlydepletedafteraddingIAA(85%within30min, >99%within1hr,Figure6D).BecausePABPC1istheprimaryPABPCisoforminHCT116cells,depletionofPABPC1alonecausedsubstantialdestabilizationofshorter-tailmRNAs1hrafterIAAaddition,andthisdestabilizationfurtherincreasedafter3hr(Figure6E).Accompanyingthelossofshorter-tailmRNAswasacorrespondingreductionofmRNAabundanceformostgenes(Figure6F,Figure6—figuresupplement1C).Importantly,ribosomefootprintsdeclinedinlockstepwithmRNAabundance,leadingtomedianTEchangesthatcenterednearzeroovertheentirecourseofPABPC1depletion(Figure6F,Figure6—figuresupplement1C).Thus,asPABPCbecamelimiting,mRNAsthatlostPABPChadnodetectablereductioninTEbeforetheyweredestabilized. ThesedatafromthePABPCdegronlineallowedustoexaminewhethersomecouplingbetweentaillengthandTEmighthaveoccurredverysoonafterPABPCdepletion—duringthetimewindowinwhichPABPChadbecomelimitingbutshort-tailedmRNAshadonlystartedtodegrade.However,nocouplingbetweentaillengthandTEwasdetectedoverthecourseofrapidPABPCdepletion(Figure6—figuresupplement2).Thus,inthiscontextinwhichtwoconditionsforstrongcouplingweresatisfied(i.e.PABPCwaslimitingandshort-tailedmRNAswerelargelyintact),nocouplingwasobserved,presumablybecausecouplingalsorequiresaregulatoryregimeinwhichPABPCenhancestranslation. TheseresultsshowthatincontrasttomRNAsoffrogoocytesandpresumablythoseofothercoupledsystems,mRNAsofHeLaandHCT116cellsdonotrequirePABPCforefficienttranslation,whichexplainswhypoly(A)-taillengthswerenotabletostronglyinfluenceTEafterwereducedPABPCofthesecellstolimitinglevels.Thus,theseresultsidentifyathirdmechanisticrequirementforcouplingbetweenpoly(A)-taillengthsandTE:couplingrequiresaregulatoryregimeinwhichPABPCaffectsmRNAtranslation. Wefindthatthreefundamentalmolecularconditionsmustbemetforcellstousepoly(A)-taillengthstoeffectivelyregulateTE.First,PABPCmustbelimitingcomparedtothenumberofpoly(A)sitesavailableforbinding.Underthiscondition,short-tailedmRNAisoformsthatpoorlycompeteforPABPCarelesslikelytohavePABPCboundtotheir3′ends(Figure7).TotheextentthatPABPCisnotbound,theseisoformslosethetranslation-activatingcapabilityofPABPCobservedincoupledsystems.CooperativebindingofadjacentPAPBCmolecules(Linetal.,2012;Schäferetal.,2019)wouldfurtherenhancepartitioningoflimitingPABPCawayfromshort-tailedmRNAsandontolonger-tailedmRNAs.WhenadditionalPABPCisintroducedintoacoupledsystem,short-tailedmRNAsbenefitmorefromPABPCbindingthanlong-tailedones,whicharemorelikelytoalreadypossessthenumberofPABPCmoleculesrequiredformoreefficienttranslation. Figure7 Downloadasset Openasset Modelforcouplingbetweenpoly(A)-taillengthandTE,andcontext-dependentrolesofPABPC. Seetextfordetails. TherequirementoflimitingPABPCforcouplingTEtopoly(A)-taillengthraisesthequestionofthestoichiometrybetweenPABPCanditssitesinthepoly(A)tailsandwhetherthisstoichiometrychangesduringtheembryonicswitchofgene-regulatoryregimes.OursequencingresultsindicatedthatthemRNAsofastageVIfrogoocytehave ~2.8×1011PABPCsites(~7×1010sitesper1µgtotalRNA, ~4µgtotalRNAperoocyte),whichconcurredwiththepreviousestimateof2 ~ 3×1011sitesperoocyte(Sagataetal.,1980).TheamountofPABPC(includingbothePABandPABPC1)infrogoocyteshasbeenestimatedateither1×1011(Cossonetal.,2002)or1.4×1011(Peuchenetal.,2017;Smitsetal.,2014)moleculesperoocyte.OurresultsshowingthatPABPCactivityislimitinginfrogoocytes,agreedwiththeseestimatesthatimplythatPABPClevelsaresufficienttobindnomorethanhalfoftheavailablePABPCsites.Poly(A)-siteoccupancywouldbeevenlowerifsomePABPCproteinsweresequesteredfromthemRNApoolorinaninactiveform,whichcouldbeconferredbyfactorsthateitherbindtoorpost-translationallymodifyPABPCtoaffectitsabilitytoeitherbindpoly(A)tailsorpromotetranslation(Brooketal.,2012;Khaleghpouretal.,2001).Untilstage15,developingfrogembryosmaintainthenumberofPABPCsitesatalevelresemblingthatofoocytes(Sagataetal.,1980).Incontrast,thetotalamountofPABPCmoleculesincreasessignificantly,nearlytriplingbystage12(Peshkinetal.,2019)—thestageatwhichthecouplingbetweentaillengthandTEstartstodisappear(Subtelnyetal.,2014).ThisincreasedPABPCwouldshiftthestoichiometrytowardPABPCbeinglesslimiting.Importantly,dysregulationofthistightlycontrolledstoichiometrynotonlydisruptsthenormalgene-regulatoryregimebutalsocancausesevereconsequencesduringoocytematurationandembryonicdevelopment(Gorgonietal.,2011;Wormingtonetal.,1996). Inouroverexpressionexperiment,weincreasedthelevelofPABPC1to >6timesitsendogenouslevel(Figure1—figuresupplement1E).WhenconsideringthattheratiobetweenendogenousPABPC1andePABisabout1:3(Wühretal.,2014),theestimatedincreaseinoverallPABPClevelwas >2.5-fold,whichmighthavebeensufficienttosaturatethepoly(A)sites.Despitethispotentialsaturationanddiminishedcouplingbetweenpoly(A)-taillengthandTE,couplingwasnotcompletelylost(Figures1E,2Aand3C–E).Totheextentthatpoly(A)sitesweresaturated,theresidualcouplingsuggeststhatoocytesmighthaveamechanismforcountingboundPABCmoleculesthatenablessomecouplingtopersistevenafterPABPCoverexpressionsaturatesthepoly(A)sites.OnewaytoachievethiscountingwouldbeforacriticalPABPC-interactingfactortoalsobelimitingsuchthatlong-tailedmRNAs,withtheirhighernumberofboundPABPCmoleculeswouldmoreeffectivelycompeteforbindingtothislimitingfactor.AtopcandidateforsuchafactoriseIF4G.Indeed,wheneIF4GwasoverexpressedtogetherwithPABPC1infrogoocytes,thecouplingbetweenpoly(A)-taillengthandTEwasfurtherreduced(Figure1—figuresupplement2C).TheideathatPABPCinteractionswithlimitingeIF4Gmightfavortranslationoflong-tailedmRNAsinoocytesalsohelpstoexplainwhyoverexpressingthePABPC1M161Amutantwasmostdetrimentaltotranslationfromthelong-tailedreporter(Figure1E). AnotherwaythatsomecouplingcouldpersistevenafteroverexpressingsaturatinglevelsofPABPCisthroughsequestrationofsomemRNAsawayfromthetranslationmachinery.Onelikelylocationforsuchsequestrationwouldbegermgranules,whichhavebeenimplicatedinregulatingmRNAtranslationinoocytesofdiverseanimalspecies(Voroninaetal.,2011).Thismechanismmightreinforcecouplingbetweenpoly(A)-taillengthandTE,perhapsbyselectivelysequesteringshort-tailedmRNAsfromtheactivetranslationpool. Thesecondconditionrequiredforstrongcouplingbetweenpoly(A)-taillengthandTEisthesurvivalofshort-tailedmRNAsunderconditionsinwhichPABPCislimiting(Figure7).Thisconditionwasnotmetinthepost-embryonicmammaliancelllinesweexamined.WhenPABPCwasdepletedintheseuncoupledsystems,manymRNAmolecules,particularlyshort-tailedonesthatpresumablycompetedpoorlyfortheremainingPABPC,weredegraded.ThepreferentiallossofmRNAsnotboundbyPABPCreducedtherangeoftaillengths,whichcorrespondinglyreducedtherangeofTEsthatcouldpotentiallybeimpartedbycouplingbetweentaillengthandTE.Thisreducedrangealsopresumablyreducedtheabilitytodetectcoupling,althoughsomeabilitywasexpectedtoberetained,asindicatedbyananalysisinwhichdatafromacoupledsystemwassampledtomatchthemorerestrictedtail-lengthdistributionofanuncoupledsystem(Subtelnyetal.,2014).Moreimportantly,thelossofmRNAsnotboundbyPABPCreducedthenumberofPABPC-bindingsites,therebyreducingtheextenttowhichthesesiteswereinexcessoverPABPCandthusreducingcouplingbetweentaillengthandTE. Inoocytesandearlyembryos,mRNAdecappingisuncoupledfromdeadenylation(Gillian-Danieletal.,1998),whichhelpsexplainwhyshort-tailedmRNAssurviveinthesesystemsdespiteourfindingthattheyhavelimitingPABPCactivity.Wesuggesttwononexclusivemechanisticexplanationsforthisunusualdecouplingofdecappingfromdeadenylation.First,oocytesandearlyembryoshaverelativelylowexpressionofdecappingenzymes(Maetal.,2013;Peshkinetal.,2019).Second,mRNAterminaluridylationactivityisverylowinfrogoocytes(Figure5—figuresupplement1; Changetal.,2018),anditremainslowthroughoutearlyembryonicdevelopmentandonlystartstoincreasedramaticallyafterzygoticgenomeactivation(Changetal.,2018).Thisdevelopmentaldelayofterminaluridylationmighthelpensurethesurvivalofshort-tailedmRNAswhenPABPCislimitingandtherebyenablestrongcouplingbetweenpoly(A)-taillengthandTEinoocytesandearlyembryos.Laterindevelopment,short-tailedmRNAsareprotectedfromterminaluridylationbysaturatingPABPC,whichhelpsexplainwhydeletionofTUT4andTUT7inmousesomaticcellshaslittleimpactonmRNAabundance(Morganetal.,2017). ThethirdconditionrequiredforcouplingbetweentaillengthandTEisthatPABPCmusthavetheabilitytoinfluenceTEofboundmRNAs(Figure7).Inthecoupledsystemoffrogoocytes,increasingPABPClevelssubstantiallyimprovedTEofnearlyallmRNAs(Figure2B–E).Incontrast,inuncoupledsystemssuchasHeLaandHCT116cells,severedepletionofPABPC,suchthatshort-andmedium-tailedmRNAsweremarkedlydestabilized,hadnoconsistentimpactonmRNATE(Figure6BandF).WesuspectthatthedifferentialeffectofPABPConTEobservedincoupledanduncoupledsystemsisrelatedtothedivergentlevelsofbasaltranslationinitiationobservedbetweenthesesystems.Indeed,theoveralltranslationmeasuredbypolysomeprofilesinoocytesandearlyembryosismuchlowerthanthatobservedineitherlaterdevelopmentalstages(Woodland,1974)orpost-embryonicmammaliancelllines(Figure3—figuresupplement1A,Figure6—figuresupplement1B),whichprovidestheopportunityforatranslation-activatingeffectofPABPCtobemoreprominentinthecoupledsystems. OurresultsshowingthatPABPCs,whileplayingacrucialroleinprotectingmRNAfromprematuredecay,haveminimalcontributiontotranslationinpost-embryonicmammaliancelllinesmightseemtocontradictthewell-acceptedfunctionofPABPCasatranslationalactivator.However,manypreviousstudiesthatestablishedtheroleforPABPCinpromotingtranslationwerecarriedoutinfrogoocytesorearlyembryonicsystems(Smithetal.,2014),wherewefoundPABPCtogloballyenhanceTE.Otherpreviousstudieswereconductedinvitro,withmixedresults:inreconstitutedsystems,PABPCisdispensablefortranslationinitiation(Mitchelletal.,2010),andinrabbitreticulocytelysates,PABPChasaminimaleffectontranslation(Hintonetal.,2007),whereasinsomeothercellextracts,PABPCactivatestranslation(Kahvejianetal.,2005;TarunandSachs,1995).Additionalexperimentswillberequiredtodeterminewhetherthisdiscrepancybetweenresultsweobtainedfromlivingcellsandthoseobtainedinsomepost-embryoniccellextractsareattributabletodifferencesbetweencelltypesortodifferencesbetweencellularcytoplasmandinvitroextracts—perhapsimpartedbydilutionoftranslationalcomponentsinextracts.Suchstudieswillneedtodifferentiatebetweenthetranslation-activationandmRNA-stabilizationactivitiesofPABPC.Inthemeantime,itishelpfultoknowthatPABPCstabilizesmRNAsofpost-embryonicmetazoancellsandthatthetwoactivitiesofPABPCcanbecontextdependent,suchthatinfrogoocytesPABPCstronglyactivatestranslationandhasnoeffectonmRNAstability,whereasinmammaliancelllinesitstabilizesmRNAsandhasnodetectableeffectonTE. ThedualpotentialofPABPCinstabilizingmRNAandpromotingtranslationbestowsPABPCwithdistinctandcontext-dependentrolesinregulatingproteinsynthesis.Inoocytesandearlyembryos,thelackofmRNAtranscriptionanddegradationleavesdifferentialTEastheprimaryoptionformodulatingproteinsynthesis.OurworkindicatesthatinthiscontextlimitingPABPCproteinsbindprimarilytomRNAswithlongerpoly(A)tailsandactivatetheirtranslation.Incontrast,inpost-embryonicmammaliancelllines,wheretranscription,mRNAdegradation,andtranslationareeachoperatingathighefficiency,PABPCsprotectmRNAsfrompre-maturedecay,enablingthemtocontributetheproperamountofproteinduringtheirlifetimes,butwithoutanyadditionalenhancementofTE.Thesecontext-dependentactivitiesarenotonlycrucialforunderstandinghowcouplingbetweentaillengthandTEisestablishedinoocytesandearlyembryosandwhyitislostlaterindevelopment,theyalsoprovidemechanisticinsightintotheeffectsofthemanyposttranscriptionalregulatoryphenomenathatalterpoly(A)-taillengths.Forexample,duringmiRNA-mediatedrepression,theArgonaute–miRNAcomplexbindstargetmRNAsandrecruitsfactorsthatdisplacePABPCfromthepoly(A)tailandacceleratetailshortening(Fabianetal.,2009;Giraldezetal.,2005;Morettietal.,2012;Risslandetal.,2017;Wuetal.,2006).Inearlyzebrafishembryos,theseeffectsareexpectedtodisadvantagetargetmRNAsatcompetingforlimitedPABPC,whichinthiscontextwouldreducetheirTEwithoutchangingtheirstability,therebyexplainingwhymiRNAsprimarilycausetranslationalrepressionintheseearlyembryos(Bazzinietal.,2012;Subtelnyetal.,2014).However,inlaterembryonicdevelopmentaswellasinpost-embryonicmammaliancells,displacementofPABPCandacceleratedtailshorteningwouldreducePABPCbindingtotargetmRNAs,whichinthiscontextwouldreducetheirstabilitywithoutchangingtheirTE,therebyexplainingwhymiRNAsprimarilycausemRNAdestabilizationinthesecells(Bazzinietal.,2012;Guoetal.,2010;Subtelnyetal.,2014). HowPABPCpromotestranslationremainsanenigma,althoughtheclosed-loopmodelhasofferedasoundmechanisticexplanation.TheinteractionbetweeneIF4GandPABPCiswellcharacterizedandprovidesaphysicallinkconnectingbothendsofthemRNA,andsomestudiesareabletocatchaglimpseofpossiblecircularstructuresofmRNAsinfixedtissues(Christensenetal.,1987)orinvitro(Wellsetal.,1998).However,recentsingle-moleculeimagingstudiesinmammaliancellsquestionthewidespreadexistenceofmRNAclosed-loopstructures(Adivarahanetal.,2018;Kochetal.,2020).OurfindingthatdepletionofPABPChadminimalimpactonTEinmammaliancelllinessupportsamodelinwhichpervasiveeIF4G–PABPC-associatedloopingofmRNAsisgenerallylackinginuncoupledsystems.ThisideaisconsistentwiththefindingthattheinteractionbetweeneIF4GandPABPCisdispensableinyeast(E.-H.Parketal.,2011)andHEK293cells(Adivarahanetal.,2018),bothofwhichareuncoupledsystems(Subtelnyetal.,2014).Incontrast,theeIF4G–PABPCinteractioniscriticalduringfrogoocytematuration(Wakiyamaetal.,2000),aprocessthatreliesoncouplingbetweenpoly(A)-taillengthandTE(RichterandLasko,2011),andourresultsshowedthatincreasingPABPChadaglobaleffectonupregulatingTEinfrogoocytes.Moreover,translationofalong-tailedreporterwassubstantiallyrepressedwhenthePABPCM161Amutantwasoverexpressedinfrogoocytes(Figure1E).Thus,theclosed-loopstructure,ifitexists,mightbemoreprevalentincoupledsystems,suchasoocytesorearlyembryos. OurexperimentsexaminedonlyafewsystemsthatinherentlypossessedeitherallornoneofthethreemolecularconditionsthatwefoundtoberequiredforstrongcouplingbetweentaillengthandTE,that is,limitingPABPCactivity,stabilizationofmRNAslackingaboundPABPC,andPABPC-sensitivetranslation.Whethertheseconditionsarebroadlyapplicabletomanyothersystemsremainstobeinvestigated.Forexample,somecellsmightfallinbetweenthetwoextremes,processingoneortwooftheseconditionsbutnotallthree.Theresultsofourexperimentsinwhichweremovedoneoftheconditionsfromfrogoocytesorimposedoneortwooftheconditionsinpost-embryonicmammaliancells,predictthatsuchcellsthatinherentlyfallbetweenthetwoextremeshaveminimalifanycoupling.Indeed,theconceptthatmultipleconditionsmustbemetbeforestrongcouplingcanbeestablishedhelpstoexplainwhycouplingbetweentaillengthandTEhasbeensoinfrequentlydetectedoutsidethegeneregulatoryregimeoperatinginoocytesandearlyembryos(Subtelnyetal.,2014). Keyresourcestable Reagenttype(species)orresourceDesignationSourceorreferenceIdentifiersAdditionalinformationCellline(human)HCT116PABPC1-AID(sC152-C16)ThisstudysC152-C16SeeMaterials and methodsfordetailsCellline(human)HCT116PABPC1-AID(sC278-C2)ThisstudysC278-C2SeeMaterials and methodsfordetailsCellline(human)HCT116OsTIRNatsumeetal.,2016Cellline(human)HeLaNametal.,2014Cellline(human)HeLaRPL3-3xHA(sC262-4)ThisstudysC262-4SeeMaterials and methodsfordetailsCellline(mouse)NIH3T3Eisenetal.,2020Cellline(zebrafish)ZF4ATCC(CRL-2050)AntibodyAnti-β-actin(Rabbitmonoclonal)CellSignallingTechnology(D6A8)1:1000dilutionAntibodyAnti-AID(Mousemonoclonal)MBLInternational(M214-3)1:1000dilutionAntibodyAnti-ePAB(Rabbitpolyclonal)Wilkieetal.,20051:2000dilutionAntibodyAnti-GFP(Mousemonoclonal)CellSignallingTechnology(2955)1:1000dilutionAntibodyAnti-FLAG(Mousemonoclonal)MilliporeSigma(F9291)1:1000dilutionAntibodyAnti-GAPDH(Mousemonoclonal)Proteintech(60004)1:1000dilutionAntibodyAnti-HA(Mousemonoclonal)CellSignallingTechnology(6E2)1:1000dilutionAntibodyAnti-PABPC1(Rabbitpolyclonal)CellSignallingTechnology(4992S)1:1000dilutionAntibodyAnti-PABPC4(Rabbitpolyclonal)NovusBiologicals(NB100-74594)1:1000dilutionAntibodyAnti-puromycin(Mousemonoclonal)MilliporeSigma(MABE343)1:1000dilutionAntibodyAnti-RPL3(Rabbitpolyclonal)Abcam(ab154882)1:1000dilutionAntibodyAnti-RPS15(Rabbitpolyclonal)Proteintech(14957–1-AP)1:1000dilutionAntibodyAnti-RPS3(Rabbitmonoclonal)CellSignallingTechnology(D50G7)1:1000dilutionAntibodyAnti-TUT4(Rabbitpolyclonal)ABclonal(A5972)1:1000dilutionAntibodyAnti-TUT7(Rabbitpolyclonal)Proteintech(25196–1-AP)1:1000dilutionAntibodyAnti-Vinculin(Mousemonoclonal)Proteintech(66305)1:1000dilutionAntibodyIRDye800CWGoatanti-RabbitIgG(H + L)LI-COR(926–32211)1:10,000dilutionAntibodyIRDye680RDGoatanti-MouseIgGLI-COR(926–68070)1:10,000dilutionSequencebasedreagentAlloligosSupplementaryfile1RecombinantDNAreagentAllplasmidsSupplementaryfile2OtherReferencesequencesSupplementaryfile3OtherMaskedmitochondrialpseudogenesSupplementaryfile4OthermRNA3′-endannotationsSupplementaryfile5 AllDNAplasmidswereassembledbyrestriction-freecloningunlessexplainedotherwise(Ungeretal.,2010).Site-directedmutagenesiswasalsocarriedoutwiththismethod.Forplasmidsusedformammaliancelltransfection,humanPABPC1(fromHeLacellcDNA)andX.laevisePAB(fromoocytecDNA)codingsequenceswereclonedintopcDNA5/FRT/TO(ThermoFisher,V652020).ForsiRNA-resistanthumanPABPC1,silentmutationswereintroducedatD107,K108,S109,I110,D111,N112,V131,C132,D249,E250,N252,andG253.AdditionalsubstitutionsweremadeatI110L,D111E,D117E,A121G,G139A,Y140F,T147S,andR166KtodisrupttheinteractionbetweenPABPC1andeIF4G(Chorghadeetal.,2017).ForsiRNA-resistantX.laevisePAB,silentmutationswereintroducedatC128,K129,V130,V131,T249,E250,andN252.SequencesofoligosusedformutagenesisareprovidedinSupplementaryfile1.AlistofplasmidsusedinthisstudyisprovidedinSupplementaryfile2.PlasmidsandtheirsequenceinformationwillbeavailableatAddgene. PlasmidsforpreparingDNAtemplatesforinvitrotranscriptionwereassembledusingthepGEM-11Zf(+)(Promega)backbone,insertingtheappropriatesequencesegmentsaftertheT7promoter.HDVribozymesequencewasobtainedfromtheplasmidp2RZ(Avisetal.,2012).X.laevisPABPC1(pabpc1.S),ePAB(pabpc1l.L),andRPL3(rpl3.L)codingsequenceswereamplifiedfromcDNAgeneratedfromX.laevisoocytes.Renilla(Rluc)andfireflyluciferase(Fluc)codingsequenceswereobtainedfrompIS2andpIS0,respectively(Farhetal.,2005).NanoLuc(Nluc)codingsequencewasobtainedfrompNL1.1.TK(Promega).X.laevisβ-globin5′-and3′-UTRsequenceswereobtainedfrompT7TS(Addgene#17091).MouseMalat13′sequencewasobtainedfromtheComp.25mutantplasmid(Wiluszetal.,2012).RlucandFlucreporterscontained5′-and3′-UTRsequencesinheritedfromthepGEM-11Zf(+)backbone,whereasNlucreportershadtheX.laevisβ-globin5′-and3′-UTRsequences.Fragmentscontainingvariablepoly(A)lengthswereputindesiredplasmidsafterallotherDNAfragmentswereassembled,alsousingrestriction-freecloning,exceptthatC3040Hcompetentcells(NEB)wereusedtoamplifytheassembledplasmids.BecauselonghomopolymerstendtobecomeshorterorgetlostwhenplasmidsarepropagatedinE. coli,individualcloneswereselectedandcheckedbyPCRandSangersequencingtoconfirmthedesiredlengthofeachpoly(A)region.TheseplasmidpreparationswerethenusedtogeneratetemplatesforinvitrotranscriptionwithoutfurtherpropagationinE. coli. ForgeneratingDNAtemplatesforinvitrotranscription,PCRreactionswerecarriedoutusingKAPAHiFiHotStartPCRKit,withacommon5′primerupstreamoftheT7promoteranddifferent3′primers.FormakingRNAsendingindefinedlengthsofpoly(A)sequence,aprimer300 ~ 600ntdownstreamoftheHDVcleavagesitewasusedtofacilitateseparatingthe5′cleavageproductfromthe3′cleavageproductandtheuncleavedtranscript.FormakingRNAsnotendingindefinedlengthsofpoly(A)sequence,aprimerpairingtotheendofthedesired3′UTRsequencewasused.AllDNAtemplatesforinvitrotranscriptionwerepurifiedonagarosegelsusingtheQIAquickGelExtractionKit(QIAGEN).SequencesofoligosusedforgeneratingDNAtemplatesareprovidedinSupplementaryfile1. InvitrotranscriptionwascarriedoutwithT7RNApolymerasepurifiedinhouseandusedat3.2ng/µlfinalconcentrationinastandard50µlreactioncontaining40mMTrispH8.0,21mMMgCl2,2mMSpermidine(Sigma),1mMdithiothreitol(DTT,GoldBio),5mMNTP(bufferedATP,UTP,CTP,GTPmix,ThermoFisher),0.1unitsyeastinorganicpyrophosphatase(NEB),40unitsSUPERase•In(ThermoFisher),and1µgDNAtemplate.Afterincubationat37°Cfor2hr,2unitsofTURBODNase(ThermoFisher)wereadded,followedbyanother20minincubationat37°C.ForconstructswiththeHDVribozyme,thermalcyclingwasperformedtoenhanceHDVcleavage(65°Cfor90sand37°Cfor5min,threecycles,50µlofreactionpertube).Beforegelloading,2µl0.5MEDTApH8.0and50µl2xRNAGelLoadingDye(ThermoFisher,R0641)wereaddedtoallinvitrotranscriptionreactions,regardlessofwhethertheHDVcleavagestepwasperformedornot.Afterincubationat65°Cfor5min,RNAswereseparatedon5%acrylamidedenaturinggels(NationalDiagnostics,EC-829).DesiredRNAbandswereidentifiedbyUV-shadowing,excised,maceratedandelutedin300mMNaClat23°Covernightonarotator.ThegelpieceswereremovedusingSpin-Xcolumns(Corning8160),andRNAswereprecipitatedwithisopropanolandresuspendedinwaterfordownstreamreactions. CappingofRNAswascarriedoutwiththeVacciniaCappingSystem(NEB,M2080S)followingthemanufacturer’sprotocol.RNAswerethenpurifiedbyphenol/chloroformextractionandethanolprecipitation.Water-resuspendedRNAswereappliedtoMicroBio-Spincolumns(Bio-Rad,7326250)fordesalting.AllRNAswerecheckedforintegritybyvisualizingonformaldehyde-agarosedenaturinggelsbeforebeingstoredat–80°C. X.laevisovarieswereobtainedfromNasco(LM00935).Ovarieswerebrokendownto2–3cmpieceswithtweezersandincubatedincalcium-omittedOR-2buffer(82.5mMNaCl,2.5mMKCl,1mMMgCl2,1mMNa2HPO4,5mMHEPES,pH7.5)with0.2%(w/v)collagenaseA(Roche,11088793001)onarockerat23°Cfor ~3hr,atwhichpointmostoocytesweredefolliculated.Theoocyteswerethenwashedextensivelywithcalcium-omittedOR-2bufferfollowedbycompleteOR-2buffer(with1mMCaCl2).StageVandVIoocyteswereseparatedfromtherestwitha ~ 0.8mmdiametermeshsieve.Forinjectionexperiments,healthystageVIoocyteswerehand-pickedandincubatedincompleteOR-2bufferat18°Covernight(>16hr)forrecoverybeforeinjection. Whenpreparingoocyteextracts,thebulkstageVandVIoocytes(500–1,000)werewashedthreetimeswithampleoocyteextractionbuffer(10mMHEPESpH7.5,10mMsodiumacetate,1mMmagnesiumacetateand2mMDTT).Afterremovingallexcessbuffer,oocyteswerecentrifugedat20,000gfor20min.Themiddlelayercontainingtheextractwascollectedwithapipetteandtransferredtoanewtube.Thecollectedextractwascentrifugedat20,000gfor10min,andthemiddlelayerwasagaincollected,avoidingthetoplipidlayerandthebottominsolublelayerasmuchaspossible.Theextractwaspassedthrougha0.45µmfilter,thenaliquotedandstoredat–80°C.TheconcentrationoftheoocyteextractwasmeasuredusingtheBradfordassay(ThermoFisher,23236)andwasusuallyat35–45mg/ml. Five hundredµlrabbitreticulocytelysateobtainedfromPromega(L4151,untreated)wassupplementedwith1µl12.5mMHemin(Sigma,51280),25µl1MHEPESpH7.5,and2.5µl10mg/mlyeasttRNA(Sigma,10109495001).Invitrotranslationreactionswerecarriedoutwith5 µlofeithersupplementedrabbitreticulocytelysateoroocyteextractintotalvolumeof20µlcontaining10µMaminoacidmix(Promega,L4461),0.5unitsSUPERase•In,10mMcreatinephosphate(Roche10621714001),200µg/mlcreatinekinase(Roche,10127566001),20mMHEPESpH7.5,75mMpotassiumacetate,1.5mMmagnesiumacetate,and1fmol/µlRlucorNlucreporterRNAwithindicatedpoly(A)tails.WhenusingRlucreporters,2.5fmol/µlFlucRNAwitha120ntpoly(A)regionfollowedbyamutantmouseMalat13′endwasincludedineachreactionforuseasanormalizationcontrol.WhenusingNlucreporters,2.5fmol/µlFlucRNAwithahistonemRNA3′-endstem-loopwasincludedineachreactionforuseasanormalizationcontrol.Whenexaminingtheeffectsofaddingpurifiedproteins,thereactionalsoincluded2µloftheindicatedamountofeitherX.laevisPABPC1,eGFP,orbufferG(20mMTrispH7.5,150mMNaCl,5%glycerol,5mMDTT). X.laevisPABPC1(pabpc1.S)codingsequencewasamplifiedfromcDNAgeneratedfromX.laevisoocytes.eGFPwasobtainedfromtheplasmidpCS2+-eGFP(Chenetal.,2017).AllcodingsequenceswereclonedintopET28a(Novagen)vectorbyrestriction-freecloning.TherecombinantproteincarriedahexahistidinetagatitsCterminus.TheplasmidwastransformedintoE.coliBL21(DE3)Starcells.Aftergrowingcellsat37°Ctoanopticaldensity(OD600)of0.6,expressionofrecombinantproteinwasinducedwith0.5mMisopropylβ-D-thiogalactoside(GoldBio).ForpurificationofPABPC1,thecellscontinuedtogrowat18°Cfor16hr,afterwhichtheywerecollectedbycentrifugation,resuspendedin10volumesoflysisbuffer(20mMTrispH7.5,2MNaCl,5%(v/v)glyceroland5mM2-mercaptoethanol)andlysedbysonication.Lysatewasclearedbycentrifugationat25,000gfor30minandincubatedwithNi-NTAagarose(Qiagen,30210)at4°Cfor1hr(0.5mlresinper50mlsupernatant).Theresinwaswashedwith20resinvolumesofbufferH(20mMTrispH7.5,2MNaCl,5%glycerol,5mM2-mercaptoethanoland10mMimidazolepH7.5)andthenwith20resinvolumesofbufferL(20mMTrispH7.5,150mMNaCl,5%glycerol,5mM2-mercaptoethanoland20mMimidazolepH7.5).ProteinswereelutedwithbufferE(20mMTrispH7.5,150mMNaCl,5%glycerol,5mM2-mercaptoethanoland200mMimidazolepH7.5).Theeluatewasdiluted1:9withbufferCL(20mMTrispH7.5,25mMNaCl,5%glycerol,5mMDTT)andloadedontoaMonoScolumn(GEHealthcare).BoundproteinswereelutedbylinearNaClgradientwithbufferCLandbufferCH(20mMTrispH7.5,500mMNaCl,5%glycerol,5mMDTT).Fractionsfromthedesiredpeakwerepooled,concentratedwithanAmiconfilterunit(Millipore,UFC805024)andappliedtoaSuperdex200column(GEHealthcare)inbufferG(20mMTrispH7.5,150mMNaCl,5%glycerol,5mMDTT).Fractionsfromthedesiredpeakwerepooled,concentratedwithanAmiconfilterunit,flash-frozeninliquidnitrogenandstoredat−80°C.eGFPproteinwaspurifiedsimilarly,exceptallbuffershad150mMNaCl,andthecation-exchangestepwasomitted. HealthyX.laevisstageVIoocytesthathadrecoveredfromdefolliculationwereselectedforinjection.WhenexaminingtheeffectsofexpressingadditionalPABPC,eitherwaterortheindicatedamountofcappedmRNA(0.25–2pmol/µl)codingforeithereGFP,frogPABPC1,frogPABPC1(M161A),frogePAB,orhumaneIF4Gwasinjectedintooocytesinavolumeof2–8nlperoocyte(PLI-100PlusPico-Injector,HarvardApparatus)at23°C.InjectedoocyteswereincubatedincompleteOR-2bufferat23°Cfor20hr,afterwhichtheywereeitherlysedformakingsequencinglibrariesorco-injectedwithanFlucreportermRNAwithahistonemRNA3′-endstem-loop(20fmol/µl,usedfornormalization),andeitheranRlucreportermRNAwithindicatedpoly(A)-taillength(10fmol/µl)oranNlucreportermRNAwithindicatedpoly(A)-taillength(10fmol/µl)inatotalvolumeof2nlperoocyte,exceptinFigure1F,whereonly2nlNlucreportermRNAs(10fmol/µl)withindicated3′endswereinjected.TheseoocyteswereincubatedincompleteOR-2bufferat23°Cfor6hrbeforelysisfordualluciferaseassays. ForPAL-TRAP,oocyteswereinjectedwithmRNA(1pmol/µl)codingforeitherX.laevisRPL3-HAoreGFP-HAatavolumeof4nlperoocyte.AfterincubationincompleteOR-2bufferat23°Cfor1day,theseoocyteswereinjectedagainwithamixtureofmRNAcodingforeithereGFPorePAB(1pmol/µl)andapopulationofRlucreportermRNAswithdifferentpoly(A)-taillengths(0,30,63,98,and120nt,equimolarratio;combinedconcentration,166.5fmol/µl)inavolumeof4nlperoocyte.AfterincubationincompleteOR-2bufferat23°Cforanotherday,oocyteswerecollectedandlysedforPAL-TRAPanalysis. Five to 10oocyteswerelysedbyvigorousshakingandpipettinginPassiveLysisBuffer(Promega,E1980),using20µlperoocyte.Lysateswereclearedbycentrifugationat5000gfor5minat4°C,afterwhich20µlofeachsupernatantwastransferredtoa96-wellmicroplateandequilibratedtoroomtemperature.LuciferaseassayswerecarriedoutwithDual-LuciferaseReporterAssaySystem(Promega,E1980)inaVeritasMicroplateLuminometeraccordingtothemanufacturer’sprotocol,regardlessofwhetherNlucorRlucwasusedasthereporter. TotalRNA(1–10µg)wasseparatedonanagarose-formaldehydegel(MansourandPestov,2013)andtransferredtoanylonmembraneusingtheWhatmanNytranSuPerCharge(SPC)TurboBlottersystem(Sigma,WHA10416300).Afterovernighttransfer,RNAwascrosslinkedtothemembraneusingaUVStratalinker2400atwavelength254nmforatotalof1200µJ.ForprobingRlucand18SRNAs,membraneswerepre-incubatedwithULTRAhybUltrasensitiveHybridizationBuffer(ThermoFisher,AM8670)at68°Cunderrotationfor1hrandthenhybridizedunderthesameconditionsovernightwithDNAprobes,whichwerebody-labeledwith[α-32P]dCTP(PerkinElmer)usingaRandomPrimerDNALabelingKit(TaKaRa,6045)accordingtothemanufacturer’sprotocol.ThetemplatesforthelabelingreactionswereDNAfragmentsgeneratedbyPCRwithgene-specificprimers.Afterprobehybridization,membraneswerewashedtwotimes(5mineach)withalow-stringencybuffer(2XSSCand0.1%SDS)at68°Cwithrotation,andtwotimes(15mineach)withhigh-stringencybuffer(0.1XSSCand0.1%SDS)at68°Cwithrotation.ForprobingendogenousmRNAs,membraneswerehybridizedtoradiolabeledgene-specificDNAoligonucleotideprobesinULTRAhyb-Oligobuffer(ThermoFisher,AM8663)overnightat42°C.TheDNAprobeswerelabeledwithT4PNK(NEB,M0201S)and[γ-32P]ATP(PerkinElmer).Followinghybridization,membraneswerewashedthreetimes(20mineach)withawashbuffer(2XSSCand0.5%SDS)at42°Cwithrotation.TheblotswerethenanalyzedusingaTyphoonFLA7000phosphor-imager(GEHealthcareLifeSciences).BeforeprobingforasecondmRNAonthesameblot,theblotswerestrippedthreetimes(20mineach)inaboilingstrippingbuffer(0.04%SDS)withgentleshakingandthencheckedforanyresidualradioactivitybyextendedphosphorimaging.SequencesofoligosusedfornorthernblotsareprovidedinSupplementaryfile1. TotalRNA(1–10µg)wasmixedwithaDNAoligo(100pmol,Supplementaryfile1)complimentarytoasegmentofthe3′-UTRinwaterinavolumeof15µl.Insomeexperiments,twoDNAoligos,eachcomplementarytoadifferentmRNA,wereaddedtogetherinonereaction.AfterincubatingtheRNA–DNAmixat85°Cfor5min,thetemperaturewasgraduallyloweredto42°C(0.1°Cpers),andthen2µl10xHybridaseBuffer(500mMHEPESpH7.5,1MNaCl,100mMMgCl2),0.5µlHybridase(Lucigen,H39500,fiveunits/µl)and2.5µlwaterwereaddedtothemix.Afterincubationat42°Cfor20min,nucleicacidswereextractedwithphenol/chloroform,precipitatedwithethanol,resuspendedin1xRNAGelLoadingDye(ThermoFisher,R0641),andresolvedonan8%acrylamidedenaturinggel(NationalDiagnostics,EC-829).RNAwastransferredontoHybond-NXmembranes(GEHealthcare,RPN303T)usingaTrans-BlotSDSemi-DryTransferCell(Bio-Rad).Membraneswereincubatedat60°Cfor1hrwithEDC(1-ethyl-3-(3-dimethylaminopropyl)carbodiimidehydrochloride),ThermoFisher,22981dilutedin0.17M1-methylimidazolepH8.0tochemicallycrosslink5′phosphatestothemembrane.BlotswerehybridizedtoradiolabeledDNAoligonucleotideprobesinULTRAhyb-Oligobuffer(ThermoFisher,AM8663)overnightat42°C.TheDNAprobeswerecomplimentarytothe3′-UTRregionsimmediatelyadjacenttothepoly(A)tailsandwerelabeledwithT4PNK(NEB,M0201S)and[γ-32P]ATP(PerkinElmer).Followinghybridization,membraneswerewashedthreetimes(20mineach)withawashbuffer(2XSSCand0.5%SDS)at42°Cwithrotation.TheblotswerethenanalyzedusingaTyphoonFLA7000phosphor-imager.BeforeprobingforasecondRNAonthesameblot,theblotswerestrippedasdescribedforconventionalnorthernblots.SequencesofoligosusedforRNaseHnorthernblotsareprovidedinSupplementaryfile1. SamplesforwesternblotswereboiledinNuPAGELDSSampleBuffer(ThermoFisher,NP0007),resolvedonNuPAGE4–12%Bis-Trisproteingels(ThermoFisher,NP0323BOX),andtransferredto0.2µmPVDFmembranes(ThermoFisher,LC2002)withaMiniGelTank(ThermoFisher,A25977),accordingtothemanufacturer’sprotocol.Membraneswerethenblockedwith5%(w/v)non-fatdrymilkinTBSbuffer(20mMTrispH7.5,150mMNaCl)for30min.Primaryantibodiesweredilutedat1:1000inTBSTbuffer(20mMTrispH7.5,150mMNaCl,0.1%(v/v)Tween20)andincubatedwiththeblotat4°Covernight.Secondaryantibodiesweredilutedat1:10,000inTBSTbufferandincubatedwiththeblotat23°Cfor1hr.BlotswereanalyzedusinganOdysseyClxmachine(LI-COR)andtheImageStudiosoftware(LI-COR).Fortotalproteindetection,Revert700TotalProteinStainKits(LI-COR,926–11010)wereusedbeforeincubationwithprimaryantibodies,accordingtothemanufacturer’sprotocol.TotalproteinlevelswerequantifiedwiththeImageQuantTLsoftware(GEHealthcareLifeSciences).IndividualproteinlevelswerequantifiedwiththeImageStudiosoftware(LI-COR,v5.2.5). Groupsoffiveoocyteswereincubatedwith0.5mCi/mlEasyTagEXPRESS35Sproteinlabelingmix(PerkinElmer,NEG772007MC)in100µlcompleteOR-2bufferat23°Cfor1hr.Foracontrol,agroupofoocyteswaspretreatedwith100µg/mlcycloheximidein100µlcompleteOR-2bufferfor5minbeforeaddingthelabelingmix.Afterlabeling,bufferswereremovedandoocyteswerewashedwith1mlcompleteOR-2buffer.Afterremovingallresidualwashbuffer,100µlicecoldlysisbuffer(20mMHEPESpH7.5,100mMKCl,5mMMgCl2,1%(v/v)TritonX-100,1xHaltproteinaseinhibitorcocktail(ThermoFisher,78429),2mMDTT)wasaddedtoeachgroup,andoocyteswerelysedbyvigorousshakingandpipetting.Lysateswereclearedbycentrifugingat5,000gat4°Cfor5min,andsupernatantswereboiledinNuPAGELDSSampleBufferandresolvedbySDS-PAGE.GelsweredriedandanalyzedusingtheTyphoonFLA7000phosphor-imagerandtheImageQuantTLsoftware(GEHealthcareLifeSciences). Injectedoocytes(~100)werewashedoncewithcompleteOR-2bufferandthreetimeswith1mlice-coldbufferRL(20mMHEPESpH7.5,100mMKCl,5mMMgCl2,1%(v/v)TritonX-100,100µg/mlcycloheximide,20units/mlSUPERase•In,cOmpleteproteaseinhibitorcocktail[Sigma,11836170001,1tabletper10mlbuffer]).Afterremovingallwashbuffer,oocyteswerelysedinbufferRL(10µl/oocyte)byvigorousshakingandpipetting.Lysateswereclearedbycentrifugationat5000gat4°Cfor10min.Aportionofeachsupernatant(5%)wassavedastheinputforwesternanalysisandRNAisolation.Theremainingsupernatantwasincubatedwithanti-HAmagneticbeads(ThermoFisher,88837)using5µlslurryper100µlsupernatant.ApopulationofNlucRNAswithvariedpoly(A)-taillengths(29,63,and139nt,equimolarratio)werespikedinat5ngper100µlsupernatant,andthebeadmixturewasincubatedat4°Cfor1hrwithend-to-endrotation.Beadswerethenimmobilizedusingamagneticstand,andthesupernatantwasremoved,withaportion(5%)savedastheflowthroughforwesternblotandRNAisolation.Beadswerewashedfourtimeswith1mlbufferRLandresuspendedin110µlbufferRL,10µlofwhichwastakenforwesternblotandtheremainderwasusedforRNAisolation.RNAwasisolatedwith10volumesofTriReagent(ThermoFisher,AM9738)accordingtothemanufacturer’sprotocol.OneµgpurifiedRNAwasincubatedat37°Cfor30minwith10unitsofT4PNK(NEB,M0201S)inPNKbuffercontaining20unitsSUPERase•Inina25µlreactiontoremove2′−3′-cyclicphosphatesofreporterRNAs.TwoPAL-seqv3librariesweremadefromeachinputandeluatesample,andaftermonitoringreproducibility,datafromeachpairofreplicatesweremergedduringanalyses,whereasonlyonelibrarywasmadefromtheeachflowthroughsample. Lysatepreparationandcentrifugationwereperformedthesameasforribosomeprofiling,exceptthatnonucleasewasaddedpriortocentrifugation.RNAwaspurifiedusing3volumesofTrizolLS(ThermoFisher,10296010)accordingtomanufacturer’sprotocolandproteinswereextractedbychloroform-methanolprecipitation. Allmammaliancellswereculturedat37°Cwith5%CO2.HeLacellswereobtainedfromtheBartellab(Nametal.,2014)andculturedinDMEM(VWR,45000–304)with10%FBS(TaKaRa,631106).HeLaRPL3-3xHA(sC262-4)cellsandtheirparentalcells,aFlp-InT-RexHeLacelllinethathadanOsTIR1geneobtainedfromAndrewHolland,wereculturedinDMEMwith10%FBS.HCT116OsTIR1cells,obtainedfromtheKanemakilab(Natsumeetal.,2016),andtheirderivativeHCT116PABPC1-AID(sC152-C16)cellswereculturedinMcCoy’s5Amedia(ThermoFisher,16600082)supplementedwith10%FBS,2mML-glutamine(ThermoFisher,25030081).HCT116PABPC1-AID(sC278-C2)cellswereculturedasHCT116OsTIR1cellsbutsupplementedwith600µg/mlG418(ThermoFisher,10131027).NIH3T3cellswereobtainedfromtheBartellab(Eisenetal.,2020)andculturedinDMEMwith10%BCS(Sigma,12133C).ZF4cellswereobtainedfromATCC(CRL-2050)andculturedinDMEMF-12media(ATCC,30–2066)and10%FBS,at28°Cwith5%CO2.Mycoplasmatestingwasperformedandnocontaminationwasobserved. ForsiRNAtransfection,cellswereplatedat2×106cellsper10cmplate,culturedfor12hr,andtransfectedwith15pmoltotalsiRNAsand45µlLipofectamineRNAiMAX(ThermoFisher,13778500)in1mlOpti-MEMmedia(ThermoFisher,31985062)accordingtomanufacturer’sprotocol.Cellsweresplit1:4ontonewplates24hraftertransfectionandthencollectedforanalysis48hr(Figure4—figuresupplement2A-B,Figure4—figuresupplement3,Figures5–6,Figure6—figuresupplement1)or72hr(Figure4,Figure4—figuresupplement1andFigure4—figuresupplement2C–G)aftertransfection.ForrescueofPABPC-knockdown,cellswereplatedat0.6×106cellsperwellin6-wellplates,andtransfectedwith2.5pmoltotalsiRNAsand7.5µlLipofectamineRNAiMAXin250µlOpti-MEMmedia.After24hr,cellsweretransfectedwith1µgDNAand3µlFuGENEHD(Promega,E2311)in150µlOpti-MEMmediaaccordingtomanufacturer’sprotocol,andcollectedforanalysis40hrafterDNAtransfection.ForotherDNAplasmidtransfections,cellswereplatedat0.6×106cellsperwellin6-wellplates,culturedfor24hr,transfectedwith1µgDNAand3µlFuGENEHD(Promega,E2311)in150µlOpti-MEMmediaaccordingtomanufacturer’sprotocol.siRNAswereallpurchasedfromDharmacon,includingsiControl(D-001810-10-05),humansiPABPC1(L-019598-00-0005),humansiPABPC4(L-011528-01-0005),mousesiPabpc1(L-060385-00-0005),humansiTUT4(L-021797-01-0005),andhumansiTUT7(L-026009-00-0005).PlasmidsandtheirsequenceinformationwillbeavailableatAddgene. Assayswereperformedasdescribed(Schmidtetal.,2009).Puromycin(ThermoFisher,A1113802)wasaddedtocellculturemediaatfinalconcentration1µg/ml.Cellswereincubatedat37°Cfor10minbeforeharvest.Foracontrol,cycloheximidewasaddedtothemediaatfinalconcentration100µg/ml5minbeforetheadditionofpuromycin. AIDwasintroducedattheCterminusofPABPC1usingCas9-mediatedgenomeengineeringinHCT116OsTIRcells.The5′and3′homologyarms(~500nt)flankingthestopcodonofPABPC1wereamplifiedfromgenomicDNAofHCT116OsTIR1cellsandclonedintoadonorplasmid(kindlyprovidedbyIainCheeseman).TheAIDcodingsequencefollowedbytheT2ApeptideandmCherry,whosesequenceswereobtainedfrompMK1221(Addgene,#84220),wasinsertedimmediatelybeforethePABPC1stopcodon.ThePAMregiontargetedbyCas9onthedonorplasmidwasmutated.TheCas9guideRNAwasclonedintopX330-BFP,asdescribed(McKinleyandCheeseman,2014).BoththedonorandtheCas9plasmidswereco-transfectedintoHCT116OsTIRcellsat0.5µgeachinasix-wellformat.After72hr,singlecellswithstrongmCherrysignalweresortedwithflowcytometryinto96-wellplates.CloneswereexpandedandgenotypedbyPCRandwesternblot.OnlyclonesthatwerehomozygousforAIDintegrationwereretained.Oneclonalline(sC152-C16)wasusedforresultsshowninFigure5C.SixhraftersiRNAtransfection,1µg/mldoxycyclinewasaddedtoinduceOsTIR1.Afteranother18hr,indole-3-aceticacid(IAA,GoldBio)wasdissolvedinethanolandaddedtocellsataconcentrationof0.5mM.Twenty-fourhrafterIAAaddition,cellswereharvestedforanalysis.SequencesofoligosusedformakingthedonorandtheCas9plasmidsareprovidedinSupplementaryfile1. IAA-induceddepletionofPABPC1inthePABPC1-AIDcellline(sC152-C16)wasrelativelyslowandgenerallyrequiredmorethan12hrfor90%depletion.Toachievefasterdepletiondynamics,morecopiesofOsTIR1wereintroduced.OsTIR1codingsequenceswereobtainedfrompBabePuroosTIR1-9Myc(Addgene#80074)andsubclonedintoPiggyBac-dCas9-Tet1(Liuetal.,2016)(kindlyprovidedbyShawnLiu),replacingthedCas9codingsequence.Thisplasmidwasco-transfectedwithpiggyBacTransposaseexpressionvector(SystemBiosciences,PB210PA-1)intothePABPC1-AIDcellline(sC152-C16).48hraftertransfection,cellswereselectedin600µg/mlG418for10d.SinglecloneswereexpandedandcheckedforOsTIR1expression.TheclonewiththehighestOsTIR1expression(sC278-C2)waspickedforIAA-inducedPABPC1-AIDdegradation.Tominimizebackgrounddegradation,cellswereincubatedwith1µg/mldoxycyclineand0.2mMauxinole(Aobious,AOB8812),aninhibitorofOsTIR1(Yesbolatovaetal.,2019)for6hr,afterwhichfreshmediawith1µg/mldoxycyclineand0.5mMIAAwereadded.Cellswerecollectedafter0.5,1,and3hrofinduction.Forthenon-inducedcondition,freshmediawith1µg/mldoxycyclinebutnoIAAwereadded,andcellswerecollectedafter1hr.PlasmidsusedformakingthePABPC1-AIDcelllinesandtheirsequenceinformationwillbeavailableatAddgene. Topreparelysatefrommammaliancells,cycloheximide(CHX)wasaddedtoeachplateatfinalconcentration100µg/ml.Plateswereimmediatelymovedtoacoldroomandculturemediawereremoved.Cellswerewashedtwicewithice-coldPBSsupplementedwith100µg/mlCHX.Afterthelastwash,1mlbufferRLL(20mMHEPESpH7.5,100mMKCl,5mMMgCl2,1%(v/v)TritonX-100,100µg/mlcycloheximide,500unit/mlRNasinPlus[Promega,N2615],2mMDTT,cOmpleteproteaseinhibitorcocktail[Sigma,11836170001,1tabletper10mlbuffer])wasaddedtoeach15cmplate,andcellswerescrapedoffandincubatedonicefor10min.Theresultinglysateswerepassedthrougha26-gaugeneedlesixtimesandclearedbycentrifugationat1300gat4°Cfor10min.Topreparelysatefromfrogoocytes,injectedoocyteswerewashedoncewithcompleteOR-2bufferandthreetimeswithbufferRLL.Afterremovingallwashbuffer,oocyteswerelysedinbufferRLL(10µl/oocyte)byvigorousshakingandpipetting.Lysateswereclearedbycentrifugationat5000gat4°Cfor10min.Onetenthofeachclearedlysatewasaddedto10volumesofTriReagentfortotal-RNApreparation,followingthemanufacturer’sprotocol.Anothersmallportion(~10µl)wastakenforwesternanalysis.Theremainderwasaliquoted,flashfrozeninliquidnitrogen,andstoredat–80°C. Forribosomeprofiling,RNaseI(ThermoFisher,AM2294)wasaddedtolysate,using30unitsperOD260unitforlysatefrommammaliancellsand10unitsperOD260unitforlysatefromfrogoocytes.Afterincubationat23°Cfor30min,lysateswereloadedontoa10–50%sucrosegradient(20mMHEPESpH7.5,100mMKCl,5mMMgCl2,10–50%(w/v)sucrose,100µg/mlcycloheximide,20units/mlSUPERase•In,2mMDTT),centrifugedinaBeckmanultracentrifugewithSW41Tirotorat36,000rpmfor2hr,andthenfractionedonaBioCompgradientfractionator.Fractionscorrespondingto80Sribosomeswerecollected,exceptinexperimentsthatgenerateddataforFigure6AandB,inwhichfractionscorrespondingto40S,60S,and80Sribosomeswerecollectedsoastoavoidlossofthesmaller-sizedmitochondrialribosomes(Rooijersetal.,2013).CollectedfractionswereconcentratedwithanAmiconfilterunit(Millipore,UFC810024),andribosome-protectedRNAfragmentswerereleasedinabuffercontaining20mMHEPESpH7.5,100mMKCl,5mMMgCl2,2mMEDTA,20units/mlSUPERase•In,2mMDTT.ReleasedRNAfragmentswereincubatedwith0.2mg/mlProteinaseK(ThermoFisher,AM2548)inthepresenceof1%SDSat42°Cfor20min,andthenpurifiedbyphenol/chloroformextractionandisopropanolprecipitation. Sequencinglibrarieswerepreparedasdescribedpreviously(Subtelnyetal.,2014),exceptinexperimentsthatgenerateddataforFigure6AandB,inwhichdifferentsizemarkerswereusedtoisolateRNAfragmentsfromalargersizerange(20–40ntinthefirstgel)soastoavoidlossofmitochondrialribosome-protectedfragments(Rooijersetal.,2013). FormatchedRNA-seq,rRNAswereeitherdepletedwiththeRiboZeroGoldKit(Illumina,Figure2A,Figure2—figuresupplement1A-B,andFigure4—figuresupplement1B),depletedwiththeNEBNextrRNADepletionKit(Human/Mouse/Rat,NEB,E6310L,Figure6andFigure6—figuresupplement1–2),ornotdepleted(Figure2B,EandF,andFigure2—figuresupplement1C).BecausetheRiboZeroGoldKitdepletesmitochondrialmRNAs(Figure2—figuresupplement1C),mitochondrialmRNAswerenotusedfornormalizationofdatageneratedwiththiskit.TotalRNA(withrRNAdepletedornotdepleted)wasfragmentedbyincubatingat95°Cfor20mininRNAFragmentationbuffer(2mMEDTA,12mMNa2CO3,88mMNaHCO3)andthenethanolprecipitated.RNAfragmentsweresize-selectedandsequencedinparallelwithribosome-protectedfragments.Adetailedprotocolforribosomeprofilingisavailableathttp://bartellab.wi.mit.edu/protocols.html. SequencingwasperformedonanIlluminaHiSeq2500,withstandardrunsofeither40or50cycles.ReadsweretrimmedatbothendstoremovedadaptersequencesusingCutadapt(v1.18)(Martin,2011)andthenmappedtotheirrespectivegenomereferenceusingSTAR(v2.4.2a)withtheparameters‘--runModealignReads--outFilterMultimapNmax1--outReadsUnmapped Fastx--outFilterTypeBySJout--outSAMattributesAll--outSAMtypeBAMSortedByCoordinate’.Mappedreadswerecountedforeachgenewithhtseq-count(0.11.0).ForbothribosomeprofilingandRNA-seq,onlyreadsthatuniquelymappedtocodingregionsofannotatedgenes(excludingthefirst15codonsandthelastfivecodons)wereincludedindownstreamanalyses.ForTEanalyses,anexpressioncutoffof30RNA-seqreadswasappliedforeachgene,withnocutoffforribosome-footprintreads.NormalizationsofRNA-seqandribosome-footprintreadswereperformedwithDESeq2(Loveetal.,2014),consideringreadsforallgenespassingthecutoff,exceptwhenabsoluteTEcomparisonsweremadebetweentwosamples—inwhichcase,RNA-seqandribosome-footprintsdatawerenormalizedbyonlyconsideringreadsfrommitochondrialgenesusingsoftwareDESeq2(Loveetal.,2014).WhenonlyrelativeTEswereconsidered,TEsweremanuallycenteredat0. OurimplementationofTAIL-seq(Eisenetal.,2020)resembledthatofmTAIL-seq(Limetal.,2016)inthatitusedsplintligationtoappendthe3′adapter,asinPAL-seqv1(Subtelnyetal.,2014).TotalRNA(1–30µg)wasmixedwithtwosetsoftail-lengthstandards(0.1ngper1µgtotalRNAforeachset)(Subtelnyetal.,2014),andtrace5′-radiolabeledmarkerRNAs(Eisenetal.,2020),whichwereusedtoevaluatetail-lengthmeasurementsand3′ligationefficiency,respectively.TheseRNAswereligatedtoa3′adapterina64µlreactioncontaining1.5µM3′adapter,1.25µM3′splintoligo,0.2mMATP,oneunit/µlRNasinPlus(Promega,N2615),10mMMgCl2,1xT4RNALigasetworeactionbuffer(NEB,M0239S),and0.5units/µlT4RNALigase2(NEB,M0239S),withtheenzymeaddedaftertheothercomponentshadbeenmixedandincubatedat23°Cfor5min.Theligationreactionwasincubatedat18°Cfor18hr.Smallportions(2µl)wereremovedatthestartandendofthereactionforexaminingtheligationefficiency.Afterligation,RNAwasextractedwithphenol/chloroform,precipitatedwithethanol,resuspendedin10µlwater,andmixedwith90µl1xRNASequencingBufferfromtheRNaseT1kit(ThermoFisher,AM2283).Afterincubationat50°Cfor5minandonicefor5min,1µlRNaseT1(oneunit/µl,ThermoFisher,AM2283)wasadded,andthereactionwasincubatedat23°Cfor15min,followedbyphenol/chloroformextractionandprecipitationwiththePrecipitation/InactivationBufferintheRNaseT1kit.RNAwasresuspendedin12µlRNAGelLoadingDyeandRNAfragmentsrangingbetween100and760ntwerepurifiedonan8%acrylamidedenaturinggel,capturedonstreptavidinbeads,5′phosphorylated,andligatedtoanequalmixoffourphased5′adaptersasdescribed(Eisenetal.,2020).cDNAwasgeneratedusingSuperScriptIII(ThermoFisher,18080044)withDNAprimerscontainingbarcodesusedformultiplexing,elutedfrombeadsbybasehydrolysisofRNAandresolvedon6%urea-acrylamidedenaturinggelsasdescribed(Subtelnyetal.,2014),exceptthatcDNAfragmentswithsizesbetween150ntand760ntwereselected.cDNAsweresequenceddirectlyusingaHiSeq2500machineinapaired-end50-by-250runinnormalmodewithav3kitasdescribed(Eisenetal.,2020).Poly(A)-taillengthsshowninFigure2,Figure2—figuresupplement1,Figure4,Figure4—figuresupplement1andFigure4—figuresupplement2C–Gweremeasuredwiththismethod.SequencesofadaptersandotheroligosusedforTAIL-seqareprovidedinSupplementaryfile1. BecauseTAIL-seqappearstorequirean8-laneflowcell(Eisenetal.,2020)andthusrequiresmanysamplesforefficientimplementation,westoppedusingthismethodandswitchedtoPAL-seq.PAL-seqv3differsfromPAL-seqv2(Eisenetal.,2020)intwomajorways:(1)Barcodeswereembeddedinthe3′adapterssothatdifferentsamplescouldbepooledfordownstreamstagesoflibraryconstruction.(2)SequencingwasperformedwithcDNAratherthanPCR-amplifiedcDNA.Eachreactionwassetupinatotalvolumeof20µlbymixingtotalRNA(0.3–3µg)withtwosetsoftail-lengthstandards(0.1ngofeachsetper1µgtotalRNA)(Subtelnyetal.,2014),twosplintDNAoligos(55pmolA-splintand2.75pmolU-splint,whichallowedpolyadenylatedRNAswithaterminaluridinetobemoreefficientlycaptured)andonebarcoded3′adapter(50pmol).Poly(A)-selectedmRNAfromzebrafishZF4cellline(0.1ngperµgtotalRNA)wasalsoaddedtomammalianRNAsamplestoenableadditionalassessmentoftail-lengthmeasurementreproducibility.TheRNA–DNAmixwasincubatedat65°Cfor5min,andthetemperaturewasgraduallyloweredto16°C(0.1°Cpersec)beforeothercomponentwereaddedtoavolumeof30µlcontaining0.2mMATP,oneunit/µlRNasinPlus,10mMMgCl2,1xT4RNALigase2reactionbufferand0.5units/µlT4RNALigase2.Afterincubationat16°Cfor16hr,EDTA(1µl,0.5M,pH8.0)wasaddedtostopligationandsamplestobesequencedtogether(withdifferentbarcodes)werecombined,phenol/chloroformextracted,andprecipitatedwithethanol.LigatedRNAwasresuspendedin20µlwater,mixedwith180µlRNaseT1buffer(20mMsodiumcitratepH5.0,1mMEDTA,7Murea),heatedto50°Cfor5min,andchilledonicefor5min,beforeadding1.6µlRNaseT1(oneunit/µl,ThermoFisher,AM2283).Afterincubatingthereactionat23°Cfor30min,RNAwasextractedwithphenol/chloroformandprecipitatedusingthePrecipitation/InactivationBufferintheRNaseT1kit.Fragmentsbetween150and760ntweregel-purified,selectedonbeads,5′-endphosphorylated,ligatedtoa5′adapter,andusedforcDNAsynthesisinthesamewayasinourimplementationofTAIL-seq,exceptoneofthefourphased5′adapterswasusedandcDNAswithlengthsbetween200and760ntwereselected.SequencingwasperformedsimilarlyasinPAL-seqv2(Eisenetal.,2020)butwithsomedifferences:(1)1fmolcDNAratherthanPCR-amplifiedcDNAwasusedperlane.(2)Asingleprimerwasusedtoobtaintworeads.(3)Afterclustergenerationandsequencing-primerhybridization,andbeforeextensionoftheprimerthroughthepoly(A)-tailregionusingtheKlenowfragment,16darkcycleswereperformedinordertoextendthesequencingprimerpastthebarcodeandconstantregionsofthe3′adaptersaswellaspasttwonucleotidescorrespondingtotheRNA3′termini.(4)40cyclesofstandardsequencing-by-synthesiswereperformedtoyieldthefirstsequencingread(read1).(5)Afterobtainingread1,theflowcellwasstripped,thesamesequencingprimerwasannealed,and260cyclesofstandardsequencing-by-synthesiswereperformedtoreadthebarcode(sixnt,whichrequiredsevencycles),aconstantsegmentofthe3′adapter(eightnt,whichrequiredonlysevencyclesbecauseofanextracycleinthebarcoderegion),andthesequenceoftheRNA,beginningatits3′terminus,whichrevealedwhethertheRNAhadaterminaluridineandprovidedinformationusedtomeasurethelengthofthepoly(A)tail.Poly(A)-taillengthsshowninFigure3,Figure3—figuresupplement1,Figure4—figuresupplement2A–BandFigure5weremeasuredwiththismethod.SequencesofadaptersandotheroligosusedforPAL-Seqv3areprovidedinSupplementaryfile1. WefoundthatinPAL-seqv3,somebarcoded3′adaptersgaverisetomuchlowernumbersofmappedreads,possiblyduetolowligationefficiency.Therefore,wedevelopedthePAL-seqv4protocol,inwhichwemadeseveralchanges.The3′-ligationreactionwasessentiallythesame,exceptthatone3′adapterwasusedforallsamples.Aftertheligation,eachsamplewasprocessedseparatelyratherthanmixed.LigatedRNAwasresuspendedin10µlwaterandfragmentedina100µlratherthan200µlvolume,andfragmentswithsizesbetween130and760ntwerethengelpurified.Forreversetranscription,differentprimerscontainingthemultiplexingbarcodesequencesaswellasaregionwithfiverandomnucleotideswasused.AftercDNAelutionfromthebeads,halfofeachsamplewassavedat–20°Cforfutureuse.cDNAsamplestobesequencedonthesamelaneinaflowcellwerecombined,ethanol-precipitatedandgel-purifiedasinPAL-seqv3,exceptcDNAfragmentswithsizesbetween190ntand800ntwereselected.SequencingclustergenerationwasperformedsimilarlytoPAL-seqv3,using0.3fmolcDNAmixtureforeachlane.Read1startedwith12cyclesofstandardsequencing-by-synthesisthatfirstsequencedthe5-ntrandomregion(usedtocallclusters)andthensequencedthe6-ntbarcoderegion(whichrequiredsevencycles)withacustomprimer.Theflowcellwasstripped,asecondsequencingprimerannealed,andtwodarkcycleswereperformedinordertoextendthisprimerpastthetwonucleotidescorrespondingtotheRNA3′termini.Thecustomextensionoftheprimerthroughthepoly(A)tailregionwithKlenowwasthenperformed,asinPAL-seqv2(Eisenetal.,2020)andv3,followedby40cyclesofstandardsequencing-by-synthesistocompletetheread1,whichwasgeneratedusingtwosequencingprimersandhadatotallengthof52nt(12cyclesbeforeand40cyclesaftertheKlenowreaction).Theflowcellwasthenstripped,andthesecondsequencingprimerwasusedfor255cyclesofstandardsequencing-by-synthesistogenerateread2.Poly(A)taillengthsshowninFigure6EandFigure6—figuresupplement2weremeasuredwiththismethod.SequencesofadaptersandotheroligosusedforPAL-Seqv4areprovidedinSupplementaryfile1. Bothhuman(release25,GRCh38.p7,primaryassembly)andmouse(release10,GRCh38.p4,primaryassembly)genomicsequencesweredownloadedfromtheGENCODEwebsite.Sequencesofmitochondrialpseudogenes(Supplementaryfile4)inhumanandmousegenomesweremaskedtoavoidlosingmitochondrialmRNAreadsduetomulti-mapping.X.laevisgenomicsequencesweredownloadedfromtheXenbasewebsite(v9.1assembly,repeatmasked),andscaffoldswereremovedsothatonlychromosomalsequenceswouldbeconsidered.TheX.laevismitochondrialgenomicsequencewasobtainedfromNCBIwebsite(NC_001573.1)andappendedtotheX.laevisgenome. Bothhuman(release25,GRCh38.p7,mainannotation)andmouse(release10,GRCh38.p4,primaryassembly)geneannotationsweredownloadedfromtheGENCODEwebsite.X.laevisgeneannotations(v9.1assembly,v1.8.3.2primarytranscript)weredownloadedfromtheXenbasewebsite,andonlychromosomalannotationswereused.X.laevismitochondrialgeneannotationswerecuratedbasedontheinformationobtainedfromtheNCBIwebsite(NC_001573.1)andappended.Forallspecies,annotationsforprotein-codinggeneswereextracted,andforeachgene,theisoformwiththelongestopenreadingframe(ORF)wasselectedtorepresentthatgene.IncasesinwhichmultipleisoformshadORFsofthesamelength,theisoformwiththelongesttranscriptwasselectedasthereferenceannotation.ForTAIL-seqandPAL-seqanalyses,thedatabasesweresupplementedwithsequencesandannotationsoftail-lengthstandards,eGFP,Rluc,Fluc,andNluc.ForribosomeprofilingandmatchedRNA-seqanalyses,thedatabasesweresupplementedwithcodingsequencesandannotationsofeGFP,Rluc,Fluc,andNluc.ForRNA-seqanalysesusedtodeterminehalf-lives,thedatabasesweresupplementedwithcodingsequencesandannotationsofAcGFP(Eisenetal.,2020),Rluc,andFluc.AllsupplementedsequencesareprovidedinSupplementaryfile3. EachPAL-seqtagcorrespondingtoanmRNAprovidesthesiteofcleavageandpolyadenylation,whichenabledannotationofmRNA3′-endisoforms.UniquelymappedtagsfromallPAL-seqdatasetsinthesamecelltype(eitherHeLaorfrogoocyte)weremerged.RNA-seqreadsgeneratedforcomparisontoribosomeprofilingweresimilarlymerged.SoftwareHOMER(Heinzetal.,2010)wasusedtocallpeaksbyusingthemergedRNA-seqdataasthebackgroundandthemergedPAL-seqdataasthesignal,withthefollowingparameters‘-stylefactor-oauto-strandseparate-fdr0.001-ntagThreshold50-fragLength40-size40-inputFragLength30-center’.ThepeaksthatintersectedwithannotatedproteinexonswereretainedasuniquemRNA3′ends.Theannotated3′endsareprovidedinSupplementaryfile5. Phasedconstantsequencesatthe5′endandpoly(A)sequencesatthe3′endofreadonewereremovedwithacustomscript(Eisenetal.,2020).ThetrimmedsequencesweremappedtoreferencegenomeswithSTAR(v2.4.2a)withtheparameters‘--runModealignReads--outFilterMultimapNmax1--outReadsUnmappedFastx--outFilterTypeBySJout--outSAMattributesAll--outSAMtypeBAMSortedByCoordinate’.Mappedsequenceswereintersectedwithsequencesofprotein-codinggenesbybedtools(v2.26.0)withtheparameters‘intersect-wa-wb-bed-s’,retainingonlythosesequencesthatwereassignedtoasinglegene.Remainingclusterswerethenfiltered,requiringatleastfivecombinedNandTbasesinthefirst6nucleotidesofread2.Foreachlibrary,1%ofthefilteredreadclusters(butnomorethan50,000andnolessthan5000)wererandomlypickedasthetrainingsetfordeterminingtheHiddenMarkovModel. ForeachIlluminasequencingcluster,averageintensitiesofeachchannelfromposition15–50ofreadonewereusedtonormalizeintensitiesofeachchannelinread2.ThenaT-signalvalueforeachcycleofreadtwoineachclusterwascalculatedbydividingnormalizedintensityfromTchannelbythesumofnormalizedintensitiesfromtheotherthreechannels.IftheT-signalwas0foraspecificcycle,theT-signalvaluesfromneighboringcycles(upto10,minimum5)wereaveragedtoinferthevalueforthatcycle.Ifaclusterhadmorethanfivecycleswitharead2T-signalvalueof0,theclusterwasdiscarded.Afive-statemixedGaussianHiddenMarkovModel(frompythonghmmpackage)wasthenusedtodecodethesequenceofstatesthatoccurredinread2.Itconsistedofaninitiationstate(state0),astrongpoly(A)-tailstate(state1),aweakpoly(A)-tailstate(state2),aweaknon-poly(A)-tailstate(state3)andastrongnon-poly(A)-tailstate(state4).Allreadsstartedinstate0,andallstateswereonlyallowedtogoforward(from0to4).Themodelwasinitializedwiththefollowingtransitionprobabilitymatrix(fromstateinrowtostateincolumn): 0.040.930.020.010.000.000.940.030.020.010.000.000.500.400.100.000.000.000.600.400.000.000.000.001.00 Theemissionmatrixforthemixedpopulationonewasinitializedwith(statesinrow,mean,varianceandpopulationfractionincolumn): 100.001.001.001.501.500.951.501.500.751.501.500.501.501.500.25 Theemissionmatrixforthemixedpopulationtwowasinitializedwith(statesinrow,mean,varianceandpopulationfractionincolumn): 0.001.001.00-1.001.500.05-1.001.500.25-1.001.500.50-1.001.500.75 Aftermodelinitialization,allclustersfromthetrainingsetwereusedtoperformunsupervisedtraining,andthenthetrainedmodelwasusedtodecodethesequenceofstatesforallretainedclusters.Foreachcluster,thepoly(A)taillengthwasdeterminedbysummingthenumberofstatesinstate1and2.Eachclusterassignedtoaspecificgenebyreadonewasconsideredasapoly(A)tag.Whenevaluatingtail-lengthdistributionsofmRNAsfromindividualgenes,onlyresultsfromgeneswithatleast100tagswereconsidered.NotethattheHiSeq2500machineinhigh-throughputmoderaisesthelaserintensitiesafter50cyclesinread2,causingirregularT-signalatthispositionandlower-than-expectedstatetransitionpredictedbytheHiddenMarkovModel.Thisledtoamilddepletionofpoly(A)tagswithtaillengthscalledat50nt,butitdidnotaffectresultsandconclusionsmadefromoveralltail-lengthdistributions. Fornormalizationofpoly(A)tagsamongsamples,DESeq2(Loveetal.,2014)wasusedwithallspike-intail-lengthstandardstoobtainthescalingfactorforeachdataset.Whenanalyzingmediantaillengths,onlygeneswithpoly(A)tagcountsexceedinganindicatedcutoffwereincludedintheanalyses. ForPAL-seqv3,readonesequencesweremappedtoreferencegenomeswithSTAR(v2.4.2a)withtheparameters‘--outFilterMultimapNmax1--outFilterTypeBySJout--outSAMattributesAll--outSAMtypeBAMSortedByCoordinate’.Mappedsequenceswereintersectedwithannotationsforprotein-codinggenesbybedtools(v2.26.0)withtheparameters‘intersect-wa-wb-bed-S’,retainingonlythosetagsforwhichthereadonesequencewasassignedtoasinglegeneortoanmRNAisoform,whenmRNA3′-endannotationswereused.Remainingclusterswerethenfilteredbyfirstremovingthefirst14basesofread2(whichcorrespondedtoaconstantregionandthebarcoderegion)andthenrequiringatleastfivecombinedNandTbaseswithinthenext6nucleotidesofread2.(ThisfilteringstepwasskippedforanalysesofterminaluridylationinFigure5andFigure5—figuresupplement1).Foreachlibrary,1%ofthefilteredreadclusters(butnomorethan50,000andnolessthan5000)wererandomlypickedasthetrainingsetfordeterminingtheHiddenMarkovModel.Thesequenceofpoly(A)stateswasdeterminedsimilarlyasforTAIL-seq,exceptasingleGaussianHiddenMarkovModelwasused,andtheemissionmatrixwasinitializedwith(statesinrow,meanandvarianceincolumn): 100.01.02.00.51.00.5-1.00.5-2.00.5 Fornormalizationofpoly(A)tagsamongsamples,DESeq2(Loveetal.,2014)wasusedwithallspike-intail-lengthstandardstoobtainthescalingfactorforeachdataset.Whenanalyzingmediantaillengths,onlymRNAisoformswithpoly(A)tagcountsexceedinganindicatedcutoffwereincludedintheanalyses.DataanalysesforPAL-seqv4werethesameasforPAL-seqv3,exceptbeforemapping,thefirst12ntofreadoneweretrimmedtoremovetherandomregionandthebarcoderegion,andnonucleotidesofreadtwoweretrimmed.NotethattheHiSeq2500machineinrapid-runmoderaisesthelaserintensitiesafter101cyclesinread2ofaPAL-seqv4run,causingirregularT-signalatthispositionandlower-than-expectedstatetransitionpredictedbytheHiddenMarkovModel.Thisledtoamilddepletionofpoly(A)tagswithtaillengthscalledat101nt,butitdidnotaffectresultsandconclusionsmadefromoveralltail-lengthdistributions. AvariantHeLacellline(sC262-4)wasusedfordatashowninFigure5D–FandFigure5—figuresupplement1.ThisFlp-InT-RexHeLacelllinehadanOsTIR1geneknocked-in.AnRPL3-3xHAgenewasinsertedusingtheFlp-In(ThermoFisher)systemandasingleclonewaspickedafterselectionwith400µg/mlHygromycinB(ThermoFisher10687010)for10days.Wehavenoreasontosuspectthattheuniquefeaturesofthislineaffectedanyresultsshown.DatawereprocessedasinPAL-seqv3,exceptthatmappedreadswereintersectedwithHeLamRNA3′-endannotations(seeAnnotationofmRNA3′-endisoforms).Allpoly(A)tagswithpoly(A)-taillengths ≥ 2ntwereusedwhenexaminingthepresenceofUnucleotidesneartheendsofpoly(A)tails. Forty-eighthraftersiRNAtransfection,HeLacellswereincubatedwithpre-warmedfreshmediawith400µM5-ethynyluridine(5EU,JenaBiosciences)for1,2,4,and8hr.Cellswereremovedfromplatesbytreatmentwithtrypsin,washedoncewithicecoldPBSandlysedwith200µlicecoldbufferRL.Lysateswereclearedbycentrifugingat1300gfor5min.Supernatantsweretransferredto2mlTriReagent,andtotalRNAwaspreparedaccordingtothemanufacturer’sprotocol. Biotinylationof5EU-labeledRNAandpurificationofbiotinylatedRNAwereperformedatdescribed(Eisenetal.,2020).RNA-seqlibrarieswerepreparedfrompurified5EU-labeledRNAandfromtotalRNAwiththeNEXTflexRapidDirectionalmRNA-seqKit(BiooScientific,5138–10).SequencingwasperformedonanIlluminaHiSeq2500withastandard40-cyclerun.SequencingreadsweremappedwithSTAR(v2.4.2a)withparameters‘--runModealignReads--outFilterMultimapNmax1--outReadsUnmappedFastx--outFilterTypeBySJout--outSAMattributesAll--outSAMtypeBAMSortedByCoordinate’.Exon-mappedreadswerecountedforeachgenewithhtseq-count(0.11.0). Relative5EU-labeledmRNAlevelsforeachgeneateachtimepointwereobtainedbynormalizingreadcountsbasedonthecountsfora5EU-containingGFPRNAstandardthathadbeenspikedintoeachsamplepriorto5EUbiotinylation(Eisenetal.,2020).Thesteady-statemRNAlevelsofeachgeneweremeasuredastheaverageofnormalizedreadcountsobtainedfromsequencingtheinputRNAforthe1and8hrtimepoints.Atotalof987geneswithvaluesthatdifferedby >2-foldatthesetwotimepointswereexcludedfromanalysis.ThenlspackageinRwasusedtofitthefollowingequationforeachgene: y=C+log21-e-kx-t0, wherexisthelabellingtimeinhours(using9999hasthelabelingtimeforthesteady-statedatapoint),yislog2valueofthenormalizedreadcount(leveloflabeledRNA),andt0isthetimeoffset.k isthedecayconstant,whichwasusedtodeterminethehalf-lifet1/2using, t1/2=ln2k,andC isacoefficientdeterminedbybothkandthesynthesisrate S,suchthat, C=log2Sk. Tofitt0foreachcondition,thevalueoft0wasvariedfrom0.05to0.7,withanintervalof0.05,andthevaluethatgavethesmallestmeansquare-lossofywhenfittingtodatafromallgeneswasused.Whenfittingtoresultsforeachgene,C wasboundby(0,Inf),andk wasboundby(0,Inf).C wasinitializedwithmax(y),andk wasinitializedwith0.23.Geneswithoutaconvergedfit(27of9644,17of9646,and12of9728inanalysisofthesiControl,siPABPC1,andsiPABPC1&4samples,respectively)wereomittedfromdown-streamanalyses.mRNAhalf-lifevaluesarereportedinFigure4—figuresupplement3—sourcedata1. GraphsweregeneratedandstatisticalanalyseswereperformedusingR(RDevelopmentCoreTeam,2013).Statisticalparametersincludingthevalueofn,statisticaltest,andstatisticalsignificance(pvalue)arereportedinthefiguresortheirlegends.Nostatisticalmethodswereusedtopredeterminesamplesize.Statisticaltestsforcorrelations(betweennonoverlappingdependentgroups)wereperformedbasedonapublishedmethod(SilverandHittner,2004)usingRpackagecocor(DiedenhofenandMusch,2015).Forluciferaseassaysofinjectedoocytes,eachreplicatereferstoagroupof7–10oocytes.For35Slabelingofinjectedoocytes,eachreplicatereferstoagroupof5–8oocytes.Forthepuromycin-basedtranslationassay,eachreplicatereferstoaseparatetransfectionexperiment. RawandprocesseddatafromsequencingweredepositedintheGEOdatabase(GSE166544).TAIL-seqandPAL-seqanalyseswereperformedusingacustomscriptwritteninPython2.7andavailableathttps://github.com/coffeebond/PAL-seq,(copyarchivedatswh:1:rev:b0aa1ba99cc89ce2080069b50cd441a7718b7b03, Xiang,2021a, Xiang,2021b). Addacomment +Openannotations.Thecurrentannotationcountonthispageisbeingcalculated. RawandprocesseddatafromsequencingweredepositedintheGEOdatabase(GSE166544).TAIL-seqandPAL-seqanalyseswereperformedusingacustomscriptwritteninPython2.7andavailableathttps://github.com/coffeebond/PAL-seq(copyarchivedathttps://archive.softwareheritage.org/swh:1:rev:b0aa1ba99cc89ce2080069b50cd441a7718b7b03). Thefollowingdatasetsweregenerated XiangK BartelDP (2021) NCBIGeneExpressionOmnibus IDGSE166544.Themolecularbasisofcouplingbetweenpoly(A)-taillengthandtranslationalefficiency. https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE166544 AdivarahanS LivingstonN NicholsonB RahmanS WuB RisslandOS ZenklusenD (2018) SpatialorganizationofsinglemRNPsatdifferentstagesofthegeneexpressionpathway MolecularCell72:727–738. https://doi.org/10.1016/j.molcel.2018.10.010 PubMed GoogleScholar AmraniN GhoshS MangusDA JacobsonA (2008) Translationfactorspromotetheformationoftwostatesoftheclosed-loopmRNP Nature453:1276–1280. https://doi.org/10.1038/nature06974 PubMed GoogleScholar AvisJM ConnGL WalkerSC (2012) Cis-actingribozymesfortheproductionofRNAinvitrotranscriptswithdefined5'and3'ends MethodsinMolecularBiology941:83–98. https://doi.org/10.1007/978-1-62703-113-4_7 PubMed GoogleScholar BagJ WuJ (1996) Translationalcontrolofpoly(A)-bindingproteinexpression EuropeanJournalofBiochemistry237:143–152. https://doi.org/10.1111/j.1432-1033.1996.0143n.x PubMed GoogleScholar BazziniAA LeeMT GiraldezAJ (2012) RibosomeprofilingshowsthatmiR-430reducestranslationbeforecausingmRNAdecayinzebrafish Science336:233–237. https://doi.org/10.1126/science.1215704 PubMed GoogleScholar BernsteinP PeltzSW RossJ (1989) Thepoly(A)-poly(A)-bindingproteincomplexisamajordeterminantofmRNAstabilityinvitro MolecularandCellularBiology9:659–670. https://doi.org/10.1128/MCB.9.2.659 PubMed GoogleScholar BrookM McCrackenL ReddingtonJP LuZL MorriceNA GrayNK (2012) Themultifunctionalpoly(A)-bindingprotein(PABP)1issubjecttoextensivedynamicpost-translationalmodification,whichmolecularmodellingsuggestsplaysanimportantroleinco-ordinatingitsactivities BiochemicalJournal441:803–816. https://doi.org/10.1042/BJ20111474 PubMed GoogleScholar CaponigroG ParkerR (1995) Multiplefunctionsforthepoly(A)-bindingproteininmRNAdecappinganddeadenylationinyeast Genes&Development9:2421–2432. https://doi.org/10.1101/gad.9.19.2421 PubMed GoogleScholar ChangH YeoJ KimJG KimH LimJ LeeM KimHH OhkJ JeonHY LeeH JungH KimKW KimVN (2018) Terminaluridylyltransferasesexecuteprogrammedclearanceofmaternaltranscriptomeinvertebrateembryos MolecularCell70:72–82. https://doi.org/10.1016/j.molcel.2018.03.004 PubMed GoogleScholar ChenGR SiveH BartelDP (2017) AseedmismatchenhancesArgonaute2-Catalyzedcleavageandpartiallyrescuesseverelyimpairedcleavagefoundinfish MolecularCell68:1095–1107. https://doi.org/10.1016/j.molcel.2017.11.032 PubMed GoogleScholar ChenX DickmanD (2017) Developmentofatissue-specificribosomeprofilingapproachinDrosophilaenablesgenome-wideevaluationoftranslationaladaptations PLOSGenetics13:e1007117. https://doi.org/10.1371/journal.pgen.1007117 PubMed GoogleScholar ChenCY ShyuAB (2011) Mechanismsofdeadenylation-dependentdecay WileyInterdisciplinaryReviews:RNA2:167–183. https://doi.org/10.1002/wrna.40 PubMed GoogleScholar ChorghadeS SeimetzJ EmmonsR YangJ BressonSM LisioM PariseG ConradNK KalsotraA (2017) Poly(A)taillengthregulatesPABPC1expressiontotunetranslationintheheart eLife6:e24139. https://doi.org/10.7554/eLife.24139 PubMed GoogleScholar ChristensenAK KahnLE BourneCM (1987) Circularpolysomespredominateontheroughendoplasmicreticulumofsomatotropesandmammotropesintheratanteriorpituitary AmericanJournalofAnatomy178:1–10. https://doi.org/10.1002/aja.1001780102 PubMed GoogleScholar CollerJM GrayNK WickensMP (1998) mRNAstabilizationbypoly(A)bindingproteinisindependentofpoly(A)andrequirestranslation Genes&Development12:3226–3235. https://doi.org/10.1101/gad.12.20.3226 PubMed GoogleScholar CossonB CouturierA LeGuellecR MoreauJ ChabelskayaS ZhouravlevaG PhilippeM (2002) Characterizationofthepoly(A)bindingproteinsexpressedduringoogenesisandearlydevelopmentofXenopuslaevis BiologyoftheCell94:217–231. https://doi.org/10.1016/S0248-4900(02)01195-4 PubMed GoogleScholar CostelloJ CastelliLM RoweW KershawCJ TalaveraD Mohammad-QureshiSS SimsPF GrantCM PavittGD HubbardSJ AsheMP (2015) GlobalmRNAselectionmechanismsfortranslationinitiation GenomeBiology16:10. https://doi.org/10.1186/s13059-014-0559-z PubMed GoogleScholar DeckerCJ ParkerR (1993) AturnoverpathwayforbothstableandunstablemRNAsinyeast:evidenceforarequirementfordeadenylation Genes&Development7:1632–1643. https://doi.org/10.1101/gad.7.8.1632 PubMed GoogleScholar DiedenhofenB MuschJ (2015) Cocor:acomprehensivesolutionforthestatisticalcomparisonofcorrelations PLOSONE10:e0121945. https://doi.org/10.1371/journal.pone.0121945 PubMed GoogleScholar EckmannCR RammeltC WahleE (2011) Controlofpoly(A)taillength WileyInterdisciplinaryReviews:RNA2:348–361. https://doi.org/10.1002/wrna.56 PubMed GoogleScholar EichhornSW SubtelnyAO KronjaI KwasnieskiJC Orr-WeaverTL BartelDP (2016) mRNApoly(A)-tailchangesspecifiedbydeadenylationbroadlyreshapetranslationinDrosophilaoocytesandearlyembryos eLife5:e16955. https://doi.org/10.7554/eLife.16955 PubMed GoogleScholar EisenTJ EichhornSW SubtelnyAO LinKS McGearySE GuptaS BartelDP (2020) ThedynamicsofcytoplasmicmRNAmetabolism MolecularCell77:786–799. https://doi.org/10.1016/j.molcel.2019.12.005 PubMed GoogleScholar FabianMR MathonnetG SundermeierT MathysH ZipprichJT SvitkinYV RivasF JinekM WohlschlegelJ DoudnaJA ChenCY ShyuAB YatesJR HannonGJ FilipowiczW DuchaineTF SonenbergN (2009) MammalianmiRNARISCrecruitsCAF1andPABPtoaffectPABP-dependentdeadenylation MolecularCell35:868–880. https://doi.org/10.1016/j.molcel.2009.08.004 PubMed GoogleScholar FarhKK GrimsonA JanC LewisBP JohnstonWK LimLP BurgeCB BartelDP (2005) ThewidespreadimpactofmammalianMicroRNAsonmRNArepressionandevolution Science310:1817–1821. https://doi.org/10.1126/science.1121158 PubMed GoogleScholar Gillian-DanielDL GrayNK AströmJ BarkoffA WickensM (1998) Modificationsofthe5'capofmRNAsduringXenopusoocytematuration:independencefromchangesinpoly(A)lengthandimpactontranslation MolecularandCellularBiology18:6152–6163. https://doi.org/10.1128/MCB.18.10.6152 PubMed GoogleScholar GiraldezAJ CinalliRM GlasnerME EnrightAJ ThomsonJM BaskervilleS HammondSM BartelDP SchierAF (2005) MicroRNAsregulatebrainmorphogenesisinzebrafish Science308:833–838. https://doi.org/10.1126/science.1109020 PubMed GoogleScholar GoldstrohmAC WickensM (2008) MultifunctionaldeadenylasecomplexesdiversifymRNAcontrol NatureReviewsMolecularCellBiology9:337–344. https://doi.org/10.1038/nrm2370 PubMed GoogleScholar GorgoniB RichardsonWA BurgessHM AndersonRC WilkieGS GautierP MartinsJP BrookM SheetsMD GrayNK (2011) Poly(A)-bindingproteinsarefunctionallydistinctandhaveessentialrolesduringvertebratedevelopment PNAS108:7844–7849. https://doi.org/10.1073/pnas.1017664108 PubMed GoogleScholar GörlachM BurdCG DreyfussG (1994) ThemRNApoly(A)-bindingprotein:localization,abundance,andRNA-bindingspecificity ExperimentalCellResearch211:400–407. https://doi.org/10.1006/excr.1994.1104 PubMed GoogleScholar GroftCM BurleySK (2002) RecognitionofeIF4GbyRotavirusNSP3revealsabasisformRNAcircularization MolecularCell9:1273–1283. https://doi.org/10.1016/S1097-2765(02)00555-5 PubMed GoogleScholar GuoH IngoliaNT WeissmanJS BartelDP (2010) MammalianmicroRNAspredominantlyacttodecreasetargetmRNAlevels Nature466:835–840. https://doi.org/10.1038/nature09267 PubMed GoogleScholar HeimanM SchaeferA GongS PetersonJD DayM RamseyKE Suárez-FariñasM SchwarzC StephanDA SurmeierDJ GreengardP HeintzN (2008) AtranslationalprofilingapproachforthemolecularcharacterizationofCNScelltypes Cell135:738–748. https://doi.org/10.1016/j.cell.2008.10.028 PubMed GoogleScholar HeinzS BennerC SpannN BertolinoE LinYC LasloP ChengJX MurreC SinghH GlassCK (2010) Simplecombinationsoflineage-determiningtranscriptionfactorsprimecis-regulatoryelementsrequiredformacrophageandBcellidentities MolecularCell38:576–589. https://doi.org/10.1016/j.molcel.2010.05.004 PubMed GoogleScholar HinnebuschAG (2014) Thescanningmechanismofeukaryotictranslationinitiation AnnualReviewofBiochemistry83:779–812. https://doi.org/10.1146/annurev-biochem-060713-035802 PubMed GoogleScholar HintonTM ColdwellMJ CarpenterGA MorleySJ PainVM (2007) FunctionalanalysisofindividualbindingactivitiesofthescaffoldproteineIF4G JournalofBiologicalChemistry282:1695–1708. https://doi.org/10.1074/jbc.M602780200 GoogleScholar IwasakiS FloorSN IngoliaNT (2016) RocaglatesconvertDEAD-boxproteineIF4Aintoasequence-selectivetranslationalrepressor Nature534:558–561. https://doi.org/10.1038/nature17978 PubMed GoogleScholar Conference KahvejianA RoyG SonenbergN (2001) ThemRNAclosed-loopmodel:thefunctionofPABPandPABP-interactingproteinsinmRNAtranslation ColdSpringHarborSymposiaonQuantitativeBiology.pp.293–300. https://doi.org/10.1101/sqb.2001.66.293 GoogleScholar KahvejianA SvitkinYV SukariehR M'BoutchouMN SonenbergN (2005) Mammalianpoly(A)-bindingproteinisaeukaryotictranslationinitiationfactor,whichactsviamultiplemechanisms Genes&Development19:104–113. https://doi.org/10.1101/gad.1262905 PubMed GoogleScholar KhaleghpourK KahvejianA DeCrescenzoG RoyG SvitkinYV ImatakaH O'Connor-McCourtM SonenbergN (2001) DualinteractionsofthetranslationalrepressorPaip2withpoly(A)bindingprotein MolecularandCellularBiology21:5200–5213. https://doi.org/10.1128/MCB.21.15.5200-5213.2001 PubMed GoogleScholar KochA AguileraL MorisakiT MunskyB StasevichTJ (2020) QuantifyingthedynamicsofIRESandcaptranslationwithsingle-moleculeresolutioninlivecells NatureStructural&MolecularBiology27:1095–1104. https://doi.org/10.1038/s41594-020-0504-7 PubMed GoogleScholar KühnU PielerT (1996) Xenopuspoly(A)bindingprotein:functionaldomainsinRNAbindingandprotein-proteininteraction JournalofMolecularBiology256:20–30. https://doi.org/10.1006/jmbi.1996.0065 PubMed GoogleScholar KühnU WahleE (2004) Structureandfunctionofpoly(A)bindingproteins BiochimicaEtBiophysicaActa(BBA)-GeneStructureandExpression1678:67–84. https://doi.org/10.1016/j.bbaexp.2004.03.008 PubMed GoogleScholar LimJ HaM ChangH KwonSC SimanshuDK PatelDJ KimVN (2014) UridylationbyTUT4andTUT7marksmRNAfordegradation Cell159:1365–1376. https://doi.org/10.1016/j.cell.2014.10.055 PubMed GoogleScholar LimJ LeeM SonA ChangH KimVN (2016) mTAIL-seqrevealsdynamicpoly(A)tailregulationinoocyte-to-embryodevelopment Genes&Development30:1671–1682. https://doi.org/10.1101/gad.284802.116 PubMed GoogleScholar LinJ FabianM SonenbergN MellerA (2012) Nanoporedetachmentkineticsofpoly(A)bindingproteinsfromRNAmoleculesrevealsthecriticalroleofC-terminusinteractions BiophysicalJournal102:1427–1434. https://doi.org/10.1016/j.bpj.2012.02.025 PubMed GoogleScholar LingJ MorleySJ PainVM MarzluffWF GallieDR (2002) Thehistone3'-terminalstem-loop-bindingproteinenhancestranslationthroughafunctionalandphysicalinteractionwitheukaryoticinitiationfactor4G(eIF4G)andeIF3 MolecularandCellularBiology22:7853–7867. https://doi.org/10.1128/MCB.22.22.7853-7867.2002 PubMed GoogleScholar LiuXS WuH JiX StelzerY WuX CzaudernaS ShuJ DadonD YoungRA JaenischR (2016) EditingDNAmethylationinthemammaliangenome Cell167:233–247. https://doi.org/10.1016/j.cell.2016.08.056 PubMed GoogleScholar LoveMI HuberW AndersS (2014) ModeratedestimationoffoldchangeanddispersionforRNA-seqdatawithDESeq2 GenomeBiology15:550. https://doi.org/10.1186/s13059-014-0550-8 PubMed GoogleScholar MaJ FlemrM StrnadH SvobodaP SchultzRM (2013) MaternallyrecruitedDCP1AandDCP2contributetomessengerRNAdegradationduringoocytematurationandgenomeactivationinmouse BiologyofReproduction88:11. https://doi.org/10.1095/biolreprod.112.105312 PubMed GoogleScholar MangusDA EvansMC JacobsonA (2003) Poly(A)-bindingproteins:multifunctionalscaffoldsforthepost-transcriptionalcontrolofgeneexpression GenomeBiology4:223. https://doi.org/10.1186/gb-2003-4-7-223 PubMed GoogleScholar MansourFH PestovDG (2013) SeparationoflongRNAbyagarose-formaldehydegelelectrophoresis AnalyticalBiochemistry441:18–20. https://doi.org/10.1016/j.ab.2013.06.008 PubMed GoogleScholar MartinM (2011) Cutadaptremovesadaptersequencesfromhigh-throughputsequencingreads EMBnet.journal17:10. https://doi.org/10.14806/ej.17.1.200 GoogleScholar McKinleyKL CheesemanIM (2014) Polo-likekinase1licensesCENP-Adepositionatcentromeres Cell158:397–411. https://doi.org/10.1016/j.cell.2014.06.016 PubMed GoogleScholar MitchellSF WalkerSE AlgireMA ParkEH HinnebuschAG LorschJR (2010) The5'-7-methylguanosinecaponeukaryoticmRNAsservesbothtostimulatecanonicaltranslationinitiationandtoblockanalternativepathway MolecularCell39:950–962. https://doi.org/10.1016/j.molcel.2010.08.021 PubMed GoogleScholar MorettiF KaiserC Zdanowicz-SpechtA HentzeMW (2012) PABPandthepoly(A)tailaugmentmicroRNArepressionbyfacilitatedmiRISCbinding NatureStructural&MolecularBiology19:603–608. https://doi.org/10.1038/nsmb.2309 PubMed GoogleScholar MorganM MuchC DiGiacomoM AzziC IvanovaI VitsiosDM PistolicJ CollierP MoreiraPN BenesV EnrightAJ O'CarrollD (2017) mRNA3'uridylationandpoly(A)taillengthsculptthemammalianmaternaltranscriptome Nature548:347–351. https://doi.org/10.1038/nature23318 PubMed GoogleScholar NagarajN WisniewskiJR GeigerT CoxJ KircherM KelsoJ PääboS MannM (2011) DeepproteomeandtranscriptomemappingofahumanCancercellline MolecularSystemsBiology7:548. https://doi.org/10.1038/msb.2011.81 GoogleScholar NamJW RisslandOS KoppsteinD Abreu-GoodgerC JanCH AgarwalV YildirimMA RodriguezA BartelDP (2014) GlobalanalysesoftheeffectofdifferentcellularcontextsonmicroRNAtargeting MolecularCell53:1031–1043. https://doi.org/10.1016/j.molcel.2014.02.013 PubMed GoogleScholar NatsumeT KiyomitsuT SagaY KanemakiMT (2016) RapidproteindepletioninhumancellsbyAuxin-Inducibledegrontaggingwithshorthomologydonors CellReports15:210–218. https://doi.org/10.1016/j.celrep.2016.03.001 PubMed GoogleScholar ParkEH WalkerSE LeeJM RothenburgS LorschJR HinnebuschAG (2011) MultipleelementsintheeIF4G1N-terminuspromoteassemblyofeIF4G1•PABPmRNPsinvivo TheEMBOJournal30:302–316. https://doi.org/10.1038/emboj.2010.312 PubMed GoogleScholar ParkJE YiH KimY ChangH KimVN (2016) Regulationofpoly(A)Tailandtranslationduringthesomaticcellcycle MolecularCell62:462–471. https://doi.org/10.1016/j.molcel.2016.04.007 PubMed GoogleScholar Preprint PeshkinL LukyanovA KalocsayM BioRxivRG (2019) Theproteinrepertoireinearlyvertebrateembryogenesis bioRxiv. https://doi.org/10.1101/571174 GoogleScholar PeuchenEH CoxOF SunL HebertAS CoonJJ ChampionMM DovichiNJ HuberPW (2017) PhosphorylationdynamicsdominatetheregulatedproteomeduringearlyXenopusdevelopment ScientificReports7:15647. https://doi.org/10.1038/s41598-017-15936-y GoogleScholar Software RDevelopmentCoreTeam (2013) R:Alanguageandenvironmentforstatisticalcomputing RFoundationforStatisticalComputing,Vienna,Austria. http://www.r-project.org RichterJD (1999) Cytoplasmicpolyadenylationindevelopmentandbeyond MicrobiologyandMolecularBiologyReviews63:446–456. https://doi.org/10.1128/MMBR.63.2.446-456.1999 PubMed GoogleScholar RichterJD LaskoP (2011) Translationalcontrolinoocytedevelopment ColdSpringHarborPerspectivesinBiology3:a002758. https://doi.org/10.1101/cshperspect.a002758 PubMed GoogleScholar RisslandOS SubtelnyAO WangM LugowskiA NicholsonB LaverJD SidhuSS SmibertCA LipshitzHD BartelDP (2017) TheinfluenceofmicroRNAsandpoly(A)taillengthonendogenousmRNA-proteincomplexes GenomeBiology18:211. https://doi.org/10.1186/s13059-017-1330-z PubMed GoogleScholar RisslandOS NorburyCJ (2009) Decappingisprecededby3′uridylationinanovelpathwayofbulkmRNAturnover NatureStructural&MolecularBiology16:616–623. https://doi.org/10.1038/nsmb.1601 GoogleScholar RooijersK Loayza-PuchF NijtmansLG AgamiR (2013) Ribosomeprofilingrevealsfeaturesofnormalanddisease-associatedmitochondrialtranslation NatureCommunications4:2886. https://doi.org/10.1038/ncomms3886 PubMed GoogleScholar SachsAB DavisRW (1989) Thepoly(A)bindingproteinisrequiredforpoly(A)shorteningand60Sribosomalsubunit-dependenttranslationinitiation Cell58:857–867. https://doi.org/10.1016/0092-8674(89)90938-0 PubMed GoogleScholar SagataN ShiokawaK YamanaK (1980) Astudyonthesteady-statepopulationofpoly(A)+RNAduringearlydevelopmentofXenopuslaevis DevelopmentalBiology77:431–448. https://doi.org/10.1016/0012-1606(80)90486-8 PubMed GoogleScholar SallésFJ LieberfarbME WredenC GergenJP StricklandS (1994) CoordinateinitiationofDrosophiladevelopmentbyregulatedpolyadenylationofmaternalmessengerRNAs Science266:1996–1999. https://doi.org/10.1126/science.7801127 PubMed GoogleScholar SchäferIB YamashitaM SchullerJM SchüsslerS ReicheltP StraussM ContiE (2019) Molecularbasisforpoly(A)RNParchitectureandrecognitionbythePan2-Pan3deadenylase Cell177:1619–1631. https://doi.org/10.1016/j.cell.2019.04.013 PubMed GoogleScholar SchmidtEK ClavarinoG CeppiM PierreP (2009) SUnSET,anonradioactivemethodtomonitorproteinsynthesis NatureMethods6:275–277. https://doi.org/10.1038/nmeth.1314 PubMed GoogleScholar SheetsMD WuM WickensM (1995) Polyadenylationofc-mosmRNAasacontrolpointinXenopusmeioticmaturation Nature374:511–516. https://doi.org/10.1038/374511a0 PubMed GoogleScholar Book SilverNC HittnerJB (2004) TestingDependentCorrelationswithNonoverlappingVariables:AMonteCarloSimulation Taylor&Francis. https://doi.org/10.3200/JEXE.71.1.53-70 GoogleScholar SmithRW BleeTK GrayNK (2014) Poly(A)-bindingproteinsarerequiredfordiversebiologicalprocessesinmetazoans BiochemicalSocietyTransactions42:1229–1237. https://doi.org/10.1042/BST20140111 PubMed GoogleScholar SmitsAH LindeboomRG PerinoM vanHeeringenSJ VeenstraGJ VermeulenM (2014) GlobalabsolutequantificationrevealstightregulationofproteinexpressioninsingleXenopuseggs NucleicAcidsResearch42:9880–9891. https://doi.org/10.1093/nar/gku661 PubMed GoogleScholar Stebbins-BoazB RichterJD (1994) MultiplesequenceelementsandamaternalmRNAproductcontrolcdk2RNApolyadenylationandtranslationduringearlyXenopusdevelopment MolecularandCellularBiology14:5870–5880. https://doi.org/10.1128/MCB.14.9.5870 PubMed GoogleScholar SubtelnyAO EichhornSW ChenGR SiveH BartelDP (2014) Poly(A)-tailprofilingrevealsanembryonicswitchintranslationalcontrol Nature508:66–71. https://doi.org/10.1038/nature13007 PubMed GoogleScholar TarunSZ SachsAB (1995) AcommonfunctionformRNA5'and3'endsintranslationinitiationinyeast Genes&Development9:2997–3007. https://doi.org/10.1101/gad.9.23.2997 PubMed GoogleScholar ThompsonMK GilbertWV (2017) mRNAlength-sensingineukaryotictranslation:reconsideringthe"closedloop"anditsimplicationsfortranslationalcontrol CurrentGenetics63:613–620. https://doi.org/10.1007/s00294-016-0674-3 PubMed GoogleScholar UchidaN HoshinoS KatadaT (2004) Identificationofahumancytoplasmicpoly(A)Nucleasecomplexstimulatedbypoly(A)-bindingprotein JournalofBiologicalChemistry279:1383–1391. https://doi.org/10.1074/jbc.M309125200 GoogleScholar UngerT JacobovitchY DantesA BernheimR PelegY (2010) Applicationsoftherestrictionfree(RF)cloningprocedureformolecularmanipulationsandproteinexpression JournalofStructuralBiology172:34–44. https://doi.org/10.1016/j.jsb.2010.06.016 PubMed GoogleScholar VoeltzGK OngkasuwanJ StandartN SteitzJA (2001) Anovelembryonicpoly(A)bindingprotein,ePAB,regulatesmRNAdeadenylationinXenopuseggextracts Genes&Development15:774–788. https://doi.org/10.1101/gad.872201 PubMed GoogleScholar VoroninaE SeydouxG Sassone-CorsiP NagamoriI (2011) RNAgranulesingermcells ColdSpringHarborPerspectivesinBiology3:a002774. https://doi.org/10.1101/cshperspect.a002774 PubMed GoogleScholar WakiyamaM ImatakaH SonenbergN (2000) InteractionofeIF4Gwithpoly(A)-bindingproteinstimulatestranslationandiscriticalforXenopusoocytematuration CurrentBiology10:1147–1150. https://doi.org/10.1016/S0960-9822(00)00701-6 PubMed GoogleScholar WangZ DayN TrifillisP KiledjianM (1999) AnmRNAstabilitycomplexfunctionswithpoly(A)-bindingproteintostabilizemRNAinvitro MolecularandCellularBiology19:4552–4560. https://doi.org/10.1128/MCB.19.7.4552 PubMed GoogleScholar WebsterMW ChenYH StowellJAW AlhusainiN SweetT GraveleyBR CollerJ PassmoreLA (2018) mRNAdeadenylationiscoupledtotranslationratesbythedifferentialactivitiesofCcr4-Notnucleases MolecularCell70:1089–1100. https://doi.org/10.1016/j.molcel.2018.05.033 PubMed GoogleScholar WeillL BellocE BavaFA MéndezR (2012) Translationalcontrolbychangesinpoly(A)taillength:recyclingmRNAs NatureStructural&MolecularBiology19:577–585. https://doi.org/10.1038/nsmb.2311 PubMed GoogleScholar WellsSE HillnerPE ValeRD SachsAB (1998) CircularizationofmRNAbyeukaryotictranslationinitiationfactors MolecularCell2:135–140. https://doi.org/10.1016/S1097-2765(00)80122-7 PubMed GoogleScholar WigingtonCP WilliamsKR MeersMP BassellGJ CorbettAH (2014) Poly(A)RNA-bindingproteinsandpolyadenosineRNA:newmembersandnovelfunctions WileyInterdisciplinaryReviews:RNA5:601–622. https://doi.org/10.1002/wrna.1233 PubMed GoogleScholar WilkieGS GautierP LawsonD GrayNK (2005) Embryonicpoly(A)-bindingproteinstimulatestranslationingermcells MolecularandCellularBiology25:2060–2071. https://doi.org/10.1128/MCB.25.5.2060-2071.2005 PubMed GoogleScholar WiluszJE JnBaptisteCK LuLY KuhnCD Joshua-TorL SharpPA (2012) AtripleHelixstabilizesthe3'endsoflongnoncodingRNAsthatlackpoly(A)tails Genes&Development26:2392–2407. https://doi.org/10.1101/gad.204438.112 PubMed GoogleScholar WoodlandHR (1974) ChangesinthepolysomecontentofdevelopingXenopuslaevisembryos DevelopmentalBiology40:90–101. https://doi.org/10.1016/0012-1606(74)90111-0 PubMed GoogleScholar WormingtonM SearfossAM HurneyCA (1996) Overexpressionofpoly(A)bindingproteinpreventsmaturation-specificdeadenylationandtranslationalinactivationinXenopusoocytes TheEMBOJournal15:900–909. https://doi.org/10.1002/j.1460-2075.1996.tb00424.x PubMed GoogleScholar WuL FanJ BelascoJG (2006) MicroRNAsdirectrapiddeadenylationofmRNA PNAS103:4034–4039. https://doi.org/10.1073/pnas.0510928103 PubMed GoogleScholar WührM FreemanRM PreslerM HorbME PeshkinL GygiS KirschnerMW (2014) DeepproteomicsoftheXenopuslaeviseggusinganmRNA-derivedreferencedatabase CurrentBiology24:1467–1475. https://doi.org/10.1016/j.cub.2014.05.044 PubMed GoogleScholar Software XiangK (2021a) PAL-seq,versionswh:1:rev:b0aa1ba99cc89ce2080069b50cd441a7718b7b03 SoftwareHost. https://archive.softwareheritage.org/swh:1:rev:b0aa1ba99cc89ce2080069b50cd441a7718b7b03 Software XiangK (2021b) PAL-seq,versionda6af39 GitHub. https://github.com/coffeebond/PAL-seq YesbolatovaA NatsumeT HayashiKI KanemakiMT (2019) Generationofconditionalauxin-inducibledegron(AID)cellsandtightcontrolofdegron-fusedproteinsusingthedegradationinhibitorauxinole Methods164-165:73–80. https://doi.org/10.1016/j.ymeth.2019.04.010 PubMed GoogleScholar YiH ParkJ HaM LimJ ChangH KimVN (2018) PABPcooperateswiththeCCR4-NOTcomplextopromotemRNAdeadenylationandblockprecociousdecay MolecularCell70:1081–1088. https://doi.org/10.1016/j.molcel.2018.05.009 PubMed GoogleScholar TimothyWNilsen ReviewingEditor;CaseWesternReserveUniversity,UnitedStates JamesLManley SeniorEditor;ColumbiaUniversity,UnitedStates TimothyWNilsen Reviewer;CaseWesternReserveUniversity,UnitedStates Oureditorialprocessproducestwooutputs:(i)publicreviewsdesignedtobepostedalongsidethepreprintforthebenefitofreaders;(ii)feedbackonthemanuscriptfortheauthors,includingrequestsforrevisions,shownbelow.Wealsoincludeanacceptancesummarythatexplainswhattheeditorsfoundinterestingorimportantaboutthework. Acceptancesummary: Thismanuscriptaddressesalong-standingquestion,namelyhowdoesthepoly(A)tailinfluencetranslationalefficiency?Itwillthereforebeofbroadinteresttoreadersfrommanyareasofmolecularbiologyincludingthoseinterestedintranslation,mRNAstability,developmentandgeneexpressioningeneral.Theauthorsconvincinglysetoutthreecriteriathatmustbemetforcouplingofpoly(A)taillengthwithtranslation. Decisionletterafterpeerreview: Thankyouforsubmittingyourarticle"Themolecularbasisofcouplingbetweenpoly(A)-taillengthandtranslationalefficiency"forconsiderationbyeLife.Yourarticlehasbeenreviewedby3peerreviewers,includingTimothyWNilsenastheReviewingEditorandReviewer#1,andtheevaluationhasbeenoverseenbyJamesManleyastheSeniorEditor. Thereviewershavediscussedtheirreviewswithoneanother,andtheReviewingEditorhasdraftedthistohelpyoupreparearevisedsubmission. Essentialrevisions: Allthreereviewerswerequitepositiveaboutthework.Neverthelessreviewer3raisedafewconcernsandreviewer2hadsomerelativelyminorsuggestions.Pleaseaddresstheseissuesasthoroughlyaspossible.Itseemslikerevisioninthetextwillbesufficient. Reviewer#1(Recommendationsfortheauthors): ThisisanexcellentmanuscriptinwhichXiangandBarteluseanabundanceofapproachestoprovidecompellingevidencerelevanttothecouplingbetweenpoly(A)-taillengthandtranslationalefficiency.Withoutreiteratingtheresults,thedataareconvincingandthepaperisclearlywritten.Anyconcernsaretootrivialtoarticulate.Publicationoftheworkisrecommendedasis. Reviewer#2(Recommendationsfortheauthors): Figure1–thereissubstantialvariabilityacrossdifferentextracts.Forexample,thedifferenceintranslationofshortvs.longtailreporteris15.2-fold(1B),33.8-fold(1C),and7.4-fold(1E).Whatisthesourceofthisvariability?Also,somepanelshaveerrorbarsandothersdonot.Canerrorbarsbeincludedinallpanels? Figure1Cand1E–HowdoestheoverexpressionlevelofPABPCcomparetoendogenouslevels? Pleasedefine"poly(A)tags"thefirsttimeitismentioned(Figurelegend2?) Figure2c–arethemoreprominentbandseGFPandPABPC1?Ifso,pleaselabelthem. Thereareseveraldifferentsequencingmethodsused(TAIL-seq,PAL-seqv3,PAL-seqv4)–whywasagivenmethodchosen?Itwouldbehelpfultomentionwhichmethodwasusedinthefigurelegends. Reviewer#3(Recommendationsfortheauthors): 1.NoerrorbarsarepresentedinFigure1A-B,whyisthisthecase? 2.TheauthorsconcludethelossofPABPCincelllinesresultsintheselectiveeliminationoftranscriptswithshortpoly(A)tails.Cantheauthorsexcludearoleforpoly(A)tailelongationintheabsenceofPABPC? 3.Theauthorsconcludebroadrulesfromasinglecoupledsystemandcelllineexperiments,wherethedepletionofPABPChasdrasticconsequencestothetranscriptome(~70%reductioninmRNAabundance)andlikelytothecell.Theauthorsshouldacknowledgetheselimitations. 4.Theauthorsfailtointegratepreviousfindingsinthemodelpresentedinfigure7.Morganetal.,2017haveshownthatlossofTUT4andTUT7inseveralcellsystemsandmousetissuesdoesnotimpactthetranscriptome.Theauthorsneedtodiscussthesefindingsandintegratethemintotheirmodel. https://doi.org/10.7554/eLife.66493.sa1 Reviewer#2(Recommendationsfortheauthors): Figure1–thereissubstantialvariabilityacrossdifferentextracts.Forexample,thedifferenceintranslationofshortvs.longtailreporteris15.2-fold(1B),33.8-fold(1C),and7.4-fold(1E).Whatisthesourceofthisvariability?Also,somepanelshaveerrorbarsandothersdonot.Canerrorbarsbeincludedinallpanels? Theexperimentthatproducedthe7.4-folddifferencewasperformedinoocytesratherthaninextracts,whichmighthelptoexplainwhyitsfold-changediffered.Forexperimentsdoneinextracts,onesourceofvariabilitymighthavebeenthedifferentbatchesofoocytelysates.Anothermighthavebeenthedifferentbufferconditionsthatresultedfromprotein-storagebufferbeingaddedtothereactionsthatdifferedby33.8-foldbutnottothosethatdifferedby15.2-fold. Toaddressthequestionofexperimentalvariabilityandlackoferrorbars,thesethreeexperimentswererepeatedwithreplicates.Whenrepeatingtheseexperiments,weusedNluc/RlucreportersandanFlucnormalizationcontrolthathadahistonemRNA3′-endstem-loop.Thenewresultsresembledtheoriginalonesandarenowshownasthe(Figure1B,C,EandFigure1—figuresupplement1B),whereasoneoftheoriginalpanelshasbeenmovedtothesupplement(Figure1—figuresupplement2A). Figure1Cand1E–HowdoestheoverexpressionlevelofPABPCcomparetoendogenouslevels? WetriedtoquantifytheoverexpressionofePABrelativeitsendogenouslevelbywestern,buttheePABantibodydidnotseemtorecognizethetaggedePABproteinweoverexpressed.WeobservedasignalfortheFLAGtagbutnotforoverexpressedePAB,whichmigratedsloweronthegel.Nonetheless,wewereabletodetectoverexpressedPABPC1(asshowninFigure1—figuresupplement1E)andestimatedthatitreachedalevelmorethansixtimesthatofendogenousPABPC1.Thisinformationisnowmentionedanddiscussedinourreviseddiscussion. Pleasedefine"poly(A)tags"thefirsttimeitismentioned(Figurelegend2?) We’veaddedthedefinitionintheFigure2legend. Figure2c–arethemoreprominentbandseGFPandPABPC1?Ifso,pleaselabelthem. Yes.We’venowlabeledtheseprominentbands. Thereareseveraldifferentsequencingmethodsused(TAIL-seq,PAL-seqv3,PAL-seqv4)–whywasagivenmethodchosen?Itwouldbehelpfultomentionwhichmethodwasusedinthefigurelegends. We’veaddedreasonsforchoosingthesequencingmethodstothemethodssectionandstatewhichmethodwasusedineachfigurelegend. Reviewer#3(Recommendationsfortheauthors): 1.NoerrorbarsarepresentedinFigure1A-B,whyisthisthecase? Whenperformingtheseexperimentswithdifferentextracts,weobservedconsistenteffectsofaddingPABPCbutvariablebaselineratios,presumablybecauseofvariablequalityoftheovaryweobtainedfromthevendor.Toassesstechnicalvariability,werepeatedtheexperimentsintriplicateusingthesamebatchofoocyteextract.We’vereplacedtheoriginalFigure1A–Bwiththeresultsofthesenewexperiments,witherrorbarsshowingstandarddeviationforthetechnicalreplicates,andwe’vemovedtheoriginalpanelAtoFigure1—figuresupplement2A. 2.TheauthorsconcludethelossofPABPCincelllinesresultsintheselectiveeliminationoftranscriptswithshortpoly(A)tails.Cantheauthorsexcludearoleforpoly(A)tailelongationintheabsenceofPABPC? Theideathatshortpoly(A)tailsmighthavegloballyelongatedwithnochangeinmRNAstabilityisnotsupportedbyourfindingsthat(1)absolutelevelsoflong-tailedmRNAs(determinedusingnormalizationtospike-instandards)remainedconstantanddidnotincrease,(2)mRNAhalf-livesdiddecrease,and(3)TUT4/7,whichhaveknownrolesinmRNAdegradationbutnottaillengthening,wererequiredforthelossofshorttails. 3.Theauthorsconcludebroadrulesfromasinglecoupledsystemandcelllineexperiments,wherethedepletionofPABPChasdrasticconsequencestothetranscriptome(~70%reductioninmRNAabundance)andlikelytothecell.Theauthorsshouldacknowledgetheselimitations. ThereasonwecreatedthePABPC-AIDfusionandexaminedtheeffectsofPAPBCdepletionbeforelargechangestothetranscriptomecouldoccurwastoavoidthesecondaryeffectsofconcerntothereferee.Nonetheless,wenowacknowledgethelimitationofexaminingonlyafewsystemsinthelastparagraphofourdiscussion. 4.Theauthorsfailtointegratepreviousfindingsinthemodelpresentedinfigure7.Morganetal.,2017haveshownthatlossofTUT4andTUT7inseveralcellsystemsandmousetissuesdoesnotimpactthetranscriptome.Theauthorsneedtodiscussthesefindingsandintegratethemintotheirmodel. WementiontheresultsofMorganetal.andintegratethemintoourmodelinourreviseddiscussion.ThefindingthatlossofTUT4andTUT7insomaticmousecellsdoesnotinfluencemRNAabundanceisconsistentwithourhypothesisthatPABPCisnotlimitinginthesecells.Accordingtoourmodel,inthesecellsinwhichPABPCisnormallynotlimiting,PABPCcoatsthemRNApoly(A)tailstherebyprotectingthemfromTUT4andTUT7. https://doi.org/10.7554/eLife.66493.sa2 Authordetails KehuiXiang HowardHughesMedicalInstitute,Cambridge,UnitedStates WhiteheadInstituteforBiomedicalResearch,Cambridge,UnitedStates DepartmentofBiology,MassachusettsInstituteofTechnology,Cambridge,UnitedStates Contribution Conceptualization,Resources,Datacuration,Software,Formalanalysis,Validation,Investigation,Visualization,Methodology,Writing-originaldraft,Projectadministration,Writing-reviewandediting Competinginterests Nocompetinginterestsdeclared "ThisORCIDiDidentifiestheauthorofthisarticle:" 0000-0002-3770-1367 DavidPBartel HowardHughesMedicalInstitute,Cambridge,UnitedStates WhiteheadInstituteforBiomedicalResearch,Cambridge,UnitedStates DepartmentofBiology,MassachusettsInstituteofTechnology,Cambridge,UnitedStates Contribution Conceptualization,Resources,Supervision,Fundingacquisition,Investigation,Projectadministration,Writing-reviewandediting Forcorrespondence [email protected] Competinginterests Nocompetinginterestsdeclared "ThisORCIDiDidentifiestheauthorofthisarticle:" 0000-0002-3872-2856 DavidPBartel DavidPBartel KehuiXiang Thefundershadnoroleinstudydesign,datacollectionandinterpretation,orthedecisiontosubmittheworkforpublication. WethankSEichhorn,TEisen,ASubtelny,SGupta,XWu,SMcGeary,WFang,KLin,JSmith,JKwasnieskiandHSiveforvaluablediscussions;SEichhorn,TEisen,KLin,andSGuptaforsharingimprovedmethodsforpoly(A)-tailprofiling;KMcKinley,KSu,MKanemaki,AHolland,A-BShyuforsharingcelllinesandplasmids;NGrayandMBrookforantibodies,andtheWhiteheadInstituteGenomeTechnologyCoreforsequencing.ThisworkissupportedbyNIHgrantGM118135.KXisaCancerResearchInstituteIrvingtonFellowsupportedbytheCancerResearchInstitute.DPBisaninvestigatoroftheHowardHughesMedicalInstitute. JamesLManley,ColumbiaUniversity,UnitedStates TimothyWNilsen,CaseWesternReserveUniversity,UnitedStates TimothyWNilsen,CaseWesternReserveUniversity,UnitedStates Received:January13,2021 Accepted:May21,2021 VersionofRecordpublished:July2,2021(version1) ©2021,XiangandBartelThisarticleisdistributedunderthetermsoftheCreativeCommonsAttributionLicense,whichpermitsunrestricteduseandredistributionprovidedthattheoriginalauthorandsourcearecredited. 4,160 Pageviews 625 Downloads 15 Citations Articlecitationcountgeneratedbypollingthehighestcountacrossthefollowingsources:Crossref,PubMedCentral,Scopus. Atwo-partlistoflinkstodownloadthearticle,orpartsofthearticle,invariousformats. Downloads(linktodownloadthearticleasPDF) ArticlePDF FiguresPDF Opencitations(linkstoopenthecitationsfromthisarticleinvariousonlinereferencemanagerservices) Mendeley Citethisarticle(linkstodownloadthecitationsfromthisarticleinformatscompatiblewithvariousreferencemanagertools) KehuiXiang DavidPBartel (2021) Themolecularbasisofcouplingbetweenpoly(A)-taillengthandtranslationalefficiency eLife10:e66493. https://doi.org/10.7554/eLife.66493 DownloadBibTeX Download.RIS Categoriesandtags ResearchArticle ChromosomesandGeneExpression regulationoftranslation regulationofmRNAstability terminaluridylation poly(A)tail PABPC1 Xenopusoocytes Researchorganisms Human Mouse Xenopus Relatedto Furtherreading Furtherreading Longerpoly(A)tailsimprovetranslationinearlydevelopment,butnotinmaturecellsthathavehigherlevelsoftheproteinPABPC. CohesinfoldschromosomesviaDNAloopextrusion.Cohesin-mediatedchromosomeloopsregulatetranscriptionbyshapinglong-rangeenhancer–promoterinteractions,amongothermechanisms.Mutationsofcohesinsubunitsandregulatorscausehumandevelopmentaldiseasestermedcohesinopathy.VertebratecohesinconsistsofSMC1,SMC3,RAD21,andeitherSTAG1orSTAG2.Toprobethephysiologicalfunctionsofcohesin,wecreatedconditionalknockout(cKO)micewithStag2deletedinthenervoussystem.Stag2cKOmiceexhibitgrowthretardation,neurologicaldefects,andprematuredeath,inpartduetoinsufficientmyelinationofnervefibers.Stag2cKOoligodendrocytesexhibitdelayedmaturationanddownregulationofmyelination-relatedgenes.Stag2lossreducespromoter-anchoredloopsatdownregulatedgenesinoligodendrocytes.Thus,STAG2-cohesingeneratespromoter-anchoredloopsatmyelination-promotinggenestofacilitatetheirtranscription.Ourstudyimplicatesdefectivemyelinationasacontributingfactortocohesinopathyandestablishesoligodendrocytesasarelevantcelltypetoexplorethemechanismsbywhichcohesinregulatestranscription. Invertebrates,condensinIandcondensinIIcooperatetoassemblerod-shapedchromosomesduringmitosis.AlthoughthemechanismofactionandregulationofcondensinIhavebeenstudiedextensively,ourcorrespondingknowledgeofcondensinIIremainsverylimited.ByintroducingrecombinantcondensinIIcomplexesintoXenopuseggextracts,wedissecttherolesofitsindividualsubunitsinchromosomeassembly.WefindthatoneoftwoHEATsubunits,CAP-D3,playsacrucialroleincondensinII-mediatedassemblyofchromosomeaxes,whereastheotherHEATsubunit,CAP-G2,hasaverystrongnegativeimpactonthisprocess.ThestructuralmaintenanceofchromosomesATPaseandthebasicaminoacidclustersofthekleisinsubunitCAP-H2areessentialforthisprocess.DeletionoftheC-terminaltailofCAP-D3increasestheabilityofcondensinIItoassemblechromosomesandfurtherexposesahiddenfunctionofCAP-G2inthelateralcompactionofchromosomes.Takentogether,ourresultsuncoveramultilayeredregulatorymechanismuniquetocondensinII,andprovideprofoundimplicationsfortheevolutionofcondensinII. Signupforemailalerts Privacynotice Menu Researchcategories Authorguide Reviewerguide About InsideeLife Community Innovation Backtotop
延伸文章資訊
- 1Poly A tail length analysis of in vitro transcribed ... - Springer
mRNA that is produced synthetically through in vitro transcription (IVT) gains a poly A tail eith...
- 2Segmented poly(A) tails significantly reduce recombination of ...
In the case of in vitro-transcribed mRNA, the poly(A) tail can be either encoded into the DNA tem...
- 3A convenient method to generate and maintain poly(A)
These data verify that the poly(A) tail of the in vitro transcribed transcript was functional in ...
- 4Poly A tail length analysis of in vitro ... - ResearchGate
The 3′-polyadenosine (poly A) tail of in vitro transcribed (IVT) mRNA was studied using liquid ch...
- 5Poly(A) Tailing Kit | Thermo Fisher Scientific - US
The Poly(A) Tailing Kit uses E. coli Poly(A) Polymerase I (E-PAP) to polyadenylate the 3' termini...