O-GlcNAcase酶与果蝇认知功能-蛋白质O-GlcNAc修饰在学习记忆中的作用

分子生物学O-GlcNAcase酶与果蝇认知功能：蛋白质O-GlcNAc修饰在学习记忆中的作用（神经科学/生物化学）作者：VilloMuha、MichaelaFenckova、AndrewFerenbach、MaricaCatinozzi、IlseEidhof、ErikStorkebaum许可：CCBY【内容简介】研究O-GlcNAc修饰酶在果蝇认知功能中的作用，证实O-GlcNAcase活性对果蝇正常学习记忆至关重要，发表于美国生物化学与分子生物学学会期刊。（学术资料，【关键词】O-GlcNAcase、果蝇、认知功能、蛋白质修饰、学习记忆、分子生物学【分类】科技>医药卫生>基础医学——————————————————————————————UniversityofDundeeO-GlcNAcasecontributestocognitivefunctioninDrosophilaMuha,Villo;Fenckova,Michaela;Ferenbach,Andrew;Catinozzi,Marica;Eidhof,Ilse;Storkebaum,ErikPublishedin:JournalofBiologicalChemistryDOI:10.1074/jbc.RA119.010312Publicationdate:2020Licence:CCBYDocumentVersionPublisher'sPDF,alsoknownasVersionofrecordLinktopublicationinDiscoveryResearchPortalCitationforpublishedversion(APA):Muha,V.,Fenckova,M.,Ferenbach,A.,Catinozzi,M.,Eidhof,I.,Storkebaum,E.,Schenck,A.,&vanAalten,D.(2020).O-GlcNAcasecontributestocognitivefunctioninDrosophila.JournalofBiologicalChemistry,295(26),8636-8646./10.1074/jbc.RA119.010312GeneralrightsCopyrightandmoralrightsforthepublicationsmadeaccessibleinDiscoveryResearchPortalareretainedbytheauthorsand/orothercopyrightownersanditisaconditionofaccessingpublicationsthatusersrecogniseandabidebythelegalrequirementsassociatedwiththeserights.TakedownpolicyIfyoubelievethatthisdocumentbreachescopyrightpleasecontactusprovidingdetails,andwewillremoveaccesstotheworkimmediatelyandinvestigateyourclaim.Downloaddate:16.Mar.2026O-GlcNAcasecontributestocognitivefunctioninDrosophilaReceivedforpublication,July23,2019,andinrevisedform,February7,2020Published,PapersinPress,February24,2020,DOI10.1074/jbc.RA119.010312XVilloMuha‡1,MichaelaFenckova‡§1,AndrewT.Ferenbach‡,XMaricaCatinozzi¶,IlseEidhof§,XErikStorkebaum¶,XAnnetteSchenck§,andXDaanM.F.vanAalten‡2Fromthe‡GeneRegulationandExpression,SchoolofLifeSciences,UniversityofDundee,DundeeDD15EH,UnitedKingdom,the§DepartmentofHumanGenetics,DondersInstituteforBrain,CognitionandBehaviour,RadboudUniversityMedicalCenter,6525GANijmegen,TheNetherlands,andthe¶DepartmentofMolecularNeurobiology,DondersInstituteforBrain,CognitionandBehaviourandtheFacultyofScience,RadboudUniversity,6525XZNijmegen,TheNetherlandsEditedbyGeraldW.HartO-GlcNAcylationisanabundantpost-translationalmodifica-tioninneurons.Inmice,anincreaseinO-GlcNAcylationleadstodefectsinhippocampalsynapticplasticityandlearning.O-GlcNAcylationisestablishedbytwoopposingenzymes:O-GlcNActransferase(OGT)andO-GlcNAcase(OGA).ToinvestigatetheroleofOGAinelementarylearning,wegener-atedcatalyticallyinactiveandpreciseknockoutOgaalleles(OgaD133NandOgaKO,respectively)inDrosophilamelanogaster.AdultOgaD133NandOgaKOflieslackingO-GlcNAcaseactivityshowedlocomotorphenotypes.Importantly,bothOgalinesexhibiteddeficitsinhabituation,anevolutionarilyconservedformoflearning,highlightingthattherequirementforO-GlcNAcaseactivityforcognitivefunctionispreservedacrossspecies.LossofO-GlcNAcaseaffectedanumberofsynapticboutonsattheaxonterminalsoflarvalneuromuscularjunction.Takentogether,wereportbehavioralandneurodevelopmentalphenotypesassociatedwithOgaallelesandshowthatOgacon-tributestocognitionandsynapticmorphologyinDrosophila.ProteinO-GlcNAcylation,adynamicmodificationofpro-teinswithGlcNAconserine/threonineresidues,isorches-tratedbytwoenzymes:O-GlcNActransferase(OGT)3andO-GlcNAcase(OGA).O-GlcNAcylationmaintainscellularhomeostasisbymodulatingtranslation(1),proteinstability(2,3),andsubcellularlocalizationofproteins(4,5).Further-more,itplaysakeyroleinregulatingtranscription(6–9)anddifferentiation(10,11).AlthoughthemechanismofO-GlcNAcylationishighlyevolutionarilyconserved,fromtheearlymetazoanTrichoplaxadhaerenstohumans(12),thereareconsiderabledifferencesintheextentverte-bratesandinvertebratestoleratealterationinproteinO-GlcNAcylation.OGA,theenzymethatremovestheO-GlcNAcmodification,istheproductoftheMGEA5(meningioma-expressedantigen5)geneinvertebrates.OGAisindispensableforlateembryonicdevelopmentandpostnatalsurvivalofmammals(13,14).MousepupslackingOGAproteinshowdelayeddevelopment,smallsize,abnormalityinlunghistology,andperinatallethality(13,14).DrosophilaOganullmutants,however,developnor-mallytoadulthood(15,16),makingDrosophilaanattractivesystemforuncoveringpreviouslyunappreciatedrolesofOGA.AsubstantialbodyofevidenceindicatesthatO-GlcNAcylationiscrucialfornormaldevelopmentandfunctionofthemammaliannervoussystem(17–21).Inmice,increasedO-GlcNAcylationinducedbyabrain-specificknockoutofOGAmanifestedinadelayinbraindevelopment,reducedolfactorybulbsize,missinganteriorpituitary,andenlargedbrainventriclesandrevealedthatOGAisrequiredforneuro-genesis(22).Ithasrecentlybeenestablishedthatcertainmuta-tionsinthehumanOGTgenecauseintellectualdisability(23–26).ThemutationsareassociatedwithreducedOGAmRNAandproteinlevels(23,27,28),suggestingthatalteredOGAexpressionmaycontributetothediversedevelopmentalandcognitivesymptomsinthesepatients.Furthermore,recentlyidentifiedSNPsintheintronicsequenceofOGAhaveassoci-atedthegenewithIQandintellectualdevelopment(29),togetherindicatingaroleforOGAinhumancognition.RecentstudieshavedemonstratedthatbothacuteandchronicincreasesofproteinO-GlcNAcylationcausehip-pocampus-associatedlearningandmemorydefectsinmiceandrats(30,31).DespitethesestudiessuggestingacrucialfunctionofOGAinnormallearning,ourknowledgeabouthowOGAaffectscognitiveabilityislimited.Therefore,weinvestigatedwhethertheroleofOGAinlearningisconservedinDrosophila.OGAisamultidomainprotein;itconsistsofanN-terminalO-GlcNAchydrolasecatalyticdomainthatbelongstotheGH84familyofglycosidehydrolases(32),amiddlehighlydisordered“stalk”domain,andaCterminuswithsequencehomologytohistoneacetyltransferases.Thehistoneacetyl-transferasedomainlackskeyaminoacidsresponsibleforacetyl-CoAbinding(33);thusOGAonlyexhibitsO-ThisworkwassupportedinpartbyTOPGrant912-12-109fromtheNether-landsOrganisationforScientificResearch(toA.S.)andbyWellcomeTrustSeniorResearchFellowshipWT087590MA(toD.M.F.v.A.).ThisworkwasalsosupportedbyfundsfromtheDondersCenterforNeuroscience(toE.S.)andtheMuscularDystrophyAssociation(toE.S.),EuropeanUnionJointProgramme–NeurodegenerativeDiseaseResearchGrantsZonMW733051075(TransNeuro)andZonMW733051073(LocalNMD)(toE.S.),andEuropeanResearchCouncilConsolidatorGrantERC-2017-COG770244(toE.S.).Theauthorsdeclarethattheyhavenoconflictsofinterestwiththecontentsofthisarticle.Author’sChoice—FinalversionopenaccessunderthetermsoftheCreativeCommonsCC-BYlicense.ThisarticlecontainsTablesS1andS2andFigs.S1–S5.1Theseauthorscontributedequallytothiswork.2Towhomcorrespondenceshouldbeaddressed.E-mail:dmfvanaalten@dundee.ac.uk.3Theabbreviationsusedare:OGT,O-GlcNActransferase;OGA,O-GlcNAcase;AUC,areaunderthecurve;TTC,trialstocriterion;VDRC,ViennaDrosophilaResourceCenter;NMJ,neuromuscularjunction;AMPAR,AMPAreceptor;ANOVA,analysisofvariance.croARTICLEAuthor’sChoice8636J.Biol.Chem.(2020)295(26)8636–8646©2020Muhaetal.PublishedbyTheAmericanSocietyforBiochemistryandMolecularBiology,Inc.byguestonJuly7,2020/DownloadedfromGlcNAcaseenzymaticactivity.However,ithasneverbeeninvestigatedbeforewhetherOGApossessesanynonenzy-maticroles.Therefore,wealsodevelopedtoolstodissectenzymaticornonenzymaticfunctionsofOgainnormalneu-ronaldevelopmentandcognition/learning.WegeneratedrationallydesignedcatalyticallyinactiveOga(OgaD133N)andnovelOgaknockout(OgaKO)allelesbyexploit-ingtheCRISPR/Cas9geneeditingtoolbox,resultinginelevatedlevelsofproteinO-GlcNAcylationinhomozygousflies.Wedis-coveredthatalossofO-GlcNAcaseactivityaffectslocomotionandcausesdeficitsinhabituationlearning,therebydemon-stratingaconservedroleofDrosophilaOgaincognitivefunc-tion.Additionally,weshowedthatsynapticboutoncountsatthelarvalneuromuscularjunctionsarealteredinOgaKOflies,indicatinganovelroleforOgainsynapticdevelopment.OurphenotypiccharacterizationofOgaD133NandOgaKOlinesalsorevealedthattheprimaryroleofOgaintheseprocessesisO-GlcNAcaseenzymeactivity.ResultsGenomeeditingofOgaresultsinincreasedproteinO-GlcNAcylationTodissectO-GlcNAcaseenzymaticandanynonenzymaticrolesofOGA,wegeneratedcatalyticallyinactiveandprecisenullDrosophilaOgaalleles(Fig.1A).TheDrosophilaOgapro-teinshows57%sequenceidentitywiththehumanenzyme(humanOGA).Previousstudieshaveidentifiedaconservedaspartate,Asp175inhumanOGAasakeycatalyticaminoacidinbothvertebrateandbacterialOGAs(34–36).ThisresidueisFigure1.GenerationandcharacterizationofOgaD133NandOgaKOalleles.A,schematicrepresentationofD.melanogasterOgaprotein.Purple,glycosylhydrolase(GH)domain;gray,linkerdomain;green,pseudo-histoneacetyltransferase(HAT-like)domain.TheOgaD133Nalleleexpressesthefull-lengthDrosoph-ilaOgawithasingleD133NmissensemutationintheGHdomain(themutationsiteisinorange).TheOgaKOalleleonlyproducesashort,truncatedpolypeptideof148aminoacidsterminatingbeforethecatalyticcoreofOgaGHdomain.B,Westernblottingon4–8-h-oldDrosophilaembryosamplesindicatealackofOgaexpressionintheOgaKOlineandanincreasedlevelofO-GlcNAcylationinhomozygousOgaKOandOgaD133NDrosophilaembryos.EmbryoswerecollectedfromcrossesofhomozygousOgaKOorOgaD133Nparentalflies.Westernblottingwasprobedwithananti-OGAantibodyandamonoclonalanti-O-GlcNAcantibody(RL2),raisedagainstO-GlcNAcmodifiednucleoporins,thatrecognizesasubsetoftheO-GlcNAcmodifiedproteome.Actinwasusedasaloadingcontrol.AcompletelossofthelowerbandcorrespondingtoOgaproteinwasapparentinhomozygous(OgaKO)samples.C,Westernblottingon1–4-day-oldmaleadultheadlysateswasprobedwithananti-OGAantibodyandmonoclonalanti-O-GlcNAcantibody(RL2).Actinwasusedasaloadingcontrol.Theanti-OGAantibodyrecognizestwoproteinsatthemolecularmassrangeof130–180kDa;thelowerbandisspecifictoDrosophilaOga,andtheupperbandisnonspecific.CompletelossofthelowerbandwasapparentinhomozygousOgaKOsamples.D,quantificationofOgaproteinlevelsinadultheadsamplesrevealedthatheterozygousOgaKO/flieshavereducedOgaproteincomparedwithgeneticbackgroundcontrol(one-wayANOVAwithDunnett’smultiplecomparisonstest,p0.036,n3).OgaproteinlevelsareunchangedinheterozygousOgaD133N/andhomozygousOgaD133Nsamples(p0.9914andp0.9013,respectively).MeansS.D.areshown.E,quantificationofO-GlcNAcylatedproteinsrevealedthatOgaKOandOgaD133NflieshaveincreasedO-GlcNAclevelscomparedwithgeneticbackgroundcontrolDrosophilaadultheadsamples(2.2-fold,one-wayANOVAwithDunnett’smultiplecomparisonstest,p0.0001,n3forbothlines;meansS.D.areshown).DrosophilaOgaisrequiredforhabituationlearningJ.Biol.Chem.(2020)295(26)8636–86468637byguestonJuly7,2020/DownloadedfromAsp133inDrosophilaOgaand,togetherwiththerestofthecatalyticmachinery,isconservedthroughoutevolution(Fig.S1).Mutationofthisaspartatetoanasparagineleadstoapro-teinspeciesincapableofhydrolysis,yetwithretentionofO-GlcNAcproteinbinding(37–39).WehaveusedaCRISPR/Cas9geneeditingapproachtointroducetheD133NmutationintotheendogenousOga,thusgeneratingthedesiredcatalyti-callyinactive(OgaD133N)allele.Inparallel,weisolatedanullallele(OgaKO)producedbytwonucleotidesframeshiftmuta-tionat3R:21219675[](OgamRNA547nucleotideposition),resultinginaprematureSTOPcodonandatruncated148–aminoacid–longpeptideproduct(Fig.S2).HomozygousOgaD133NandOgaKOanimalsdevelopedtoadulthoodwithoutapparentdefects.ToprobetheeffectofthenewlygeneratedOgaallelesonO-GlcNAchomeostasis,sam-plesfromOgaD133NandOgaKOembryosandadultheadsweresubjectedtoWesternblotting(Fig.1,BandC).Ogawasunde-tectableinhomozygousOgaKOsamples(Fig.1,BandC),whereasitwasexpressedatWTlevelinheterozygousOgaD133N(OgaD133N/)andhomozygousOgaD133Nflies(Fig.1C).HalfgenedosageofOgainOgaKO/fliesledto60%reductioninproteinlevel(normalizedexpression:w11181,OgaKO/0.400.1)(Fig.1D).ProteinO-GlcNAcylationlevelswere2.2-foldelevatedinsamplesfrombothhomozygousOgaD133NandOgaKOflies(Fig.1E).Insummary,wecreatedtwonewallelesofOgathatshowedincreasedproteinO-GlcNAcylationandenabledthedissectionofOgafunction.CompromisedOgafunctionleadstoareductioninlifespanandlocomotordefectsinadultfliesElevatedO-GlcNAclevelscausedbyhighsugarandhighglu-cosaminedietareknowntoshortenthemedianlifespanofadultflies(40).WefirsttestedwhethertheincreasedproteinO-GlcNAcylationinOgamutantfliesaffectssurvivalasamea-sureofoverallhealth.Batchesof20malefliescollectedwithin24hpasteclosionwereplacedintovialswithstandarddiet,andtheirlifespanwasmonitoredforover100days.Themeanlifespanforgeneticbackgroundcontrolflieswas73.01.4days,whereashomozygousOgaD133N(61.10.8days)andOgaKO(67.20.7days)adultmalesexhibitedsignificantmeanlife-spanreductionofapproximately12and6days,respectively(Fig.2A).LossofOGTinmousepostmitoticneuronscausedarapidincreaseofdailyfoodintakeaccompaniedwithhyperactivity(21),indicatingthattheO-GlcNAcsystempotentiallyinflu-enceslocomotoractivityofanorganism.Therefore,wetesteddailylocomotoractivityinOgaD133NandOgaKOmaleadultsusingDAM2Drosophilaactivitymonitors.Under12:12hoflight:darkcycleconditions,theOgaKOgroupconsistentlyshowedamodestbutsignificantdecreaseintotaldailyactivitycounts(meansS.D.,1088266totalactivitycounts/day)comparedwithcontrol(1359371)andOgaD133N(1350345)groups(Fig.2BandFig.S3,A–C).Furthermore,theOgaKOfliesexhibitedareductioninactivitycountswhileawake(Fig.S3,D–F),suggestingthattheobservedlowerdailyactivityispossiblyassociatedwithmotordefects.Hence,wenextfurtherexploredmotorphenotypesinOgaD133NandOgaKOfliesusingislandtestandnegativegeo-taxisassays.BothtestsarewidelyusedtoinvestigateneuronalimpairmentandmusculardefectsinDrosophilamodelsofintellectualdisability,neurodegeneration,andneuromusculardiseases(41–44).Theislandtestmeasuresescaperesponsesrequiringactivationandcoordinationoflegandwingmove-ments.Inthisassay,fliesarethrownontoawhiteflatplatformsurroundedwithwater,andthetimeeachindividualspendsontheplatformisdetermined.HealthyWTfliesnormallyexhibitanescaperesponseandquicklyflyawayfromtheplatform.Undertheislandassaytestconditions,homozygousOgaD133N(areaundercurveparameter(AUC),shownasthemeanS.D.,was37001300)andOgaKO(52001100)fliesremainedmarkedlylongerontheplatformthanthebackgroundcontrolline(19001000)(Fig.2,CandD).Inthenegativegeotaxisassay,young(2–4-day-old)malecontrolfliesclimbedatameanspeedof235mm/s.How-ever,climbingperformanceofhomozygousOgaD133N(177mm/s)andOgaKO(186mm/s)fliesshowedsignificantlyreducedclimbingspeedcomparedwithgeneticbackgroundcontrols,indicatinglocomotorimpairment(Fig.2E).Takentogether,thesedatashowthatcompromisedOgafunctionleadstoreducedlifespanandlocomotordefectsinadultflies.LossofDrosophilaOgaanditsO-GlcNAcaseactivitycausesdeficitsinhabituationlearningTodissectapossibleroleofOgaincognition,weinvestigatedtheeffectofOgamutationsonhabituation.Habituationisafundamental,evolutionarilyconservedformoflearningchar-acterizedbyatemporalattenuationofaninitialstrongresponsetoarepeated,irrelevantstimulus.Itisanimportantprerequi-siteforhighercognitivefunctioningandhasbeenfoundtobedefectiveinanumberofneurodevelopmentaldisordersinhumans(45,46)andanimalmodels(47–49).ToassesstheroleofOgaanditsO-GlcNAcasecatalyticactivityinthistypeoflearning,wesubjectedOgaD133NandOgaKOfliesandisogenicgeneticbackgroundcontrolfliesto100short(15ms)light-offstimuliwith1-sintervalsinthelight-offjumphabituationassay.Locomotordeficitsobservedinthenegativegeotaxisandislandassaydidnotprecludeassessmentofhabituation,asassessedbyafatigueassay(Fig.S4).Allflylineswereabletoexhibitgoodinitialjumpresponsestothefirstfivelight-offstimuli(50%initialjumpers;TablesS1andS2).Controlflieshabituatedquicklytotherepeatedlight-offstimulus(meanS.E.trialstocriterion(TTC)4.30.5,n65,Fig.3A;and7.20.8,n74,Fig.3B).BothhomozygousOgamutantlinesshowedslowhabituationandfailedtoadapttheirjumpresponsetotherepeatedstimulus(OgaD133N:53.83.7,meanTTCfoldchange7.5,n86;OgaKO:meanS.E.TTC10.12.3,meanTTCfoldchangecomparedwithcontrolflies2.3,n63;Fig.3,A–D,andTableS2).Wegeneratedandtestedalsopan-neuronalOgaknockdownflies(elavOgaRNAi),usinganinducibleOgaUAS-RNAiallele(ViennaDrosophilaResourceCenter(VDRC)catalogno.41822)andthepan-neu-ronalelav-Gal4drivertoaddresstheneuronal/glial-specificityofthesedefects.Progenyfromcrossesbetweentheelav-Gal4driverlineandthegeneticbackgroundoftheRNAiline(VDRCcatalogno.60000)wereusedascontrols.Theknockdownfliesshowedgoodinitialjumpresponse(70.8%initialjumpers)butDrosophilaOgaisrequiredforhabituationlearning8638J.Biol.Chem.(2020)295(26)8636–8646byguestonJuly7,2020/Downloadedfromfailedtohabituatecomparedwithcontrols(meanS.E.TTC12.92.5,meanTTCfoldchange4.2,n68;Fig.3,EandF,andTableS2).Takentogether,thesedatashowthatlossofDrosophilaOgaanditsO-GlcNAcaseactivityleadstoadeficitinhabituationlearningandthatOgaactivityinneuronsisrequiredforhabituation.O-GlcNAcasemodulatesthenumberofsynapticboutonsatthelarvalneuromuscularjunctionNormalsynapticdevelopmentandmorphologyarecrucialformotorbehavior,learning,andcognitivefunctioning.Importantly,asignificantnumberofproteinsthatorchestratesynapsestructureandsynaptictransmissionaremodifiedwithO-GlcNAc(50,51),includinginDrosophila(38).WenextinvestigatedsynapticmorphologyinOgaD133NandOgaKOflies.Thelarvalneuromuscularjunction(NMJ)isawell-establishedsystemtostudysynapticdevelopmentandmorphologyinDro-sophila.Type1bNMJsconsistofbranchedchainsofsynapticboutonscontainingglutamatergictransmissionsites,whichsharefundamentalmechanisticfeatureswiththeexcitatorysysteminthemammalianbrain(52).WevisualizedthelarvalNMJarchitecturebyimmunolabelingthepresynapticFigure2.LifespanandlocomotorbehaviorofOgaKOandOgaD133Nflies.A,lifespanofhomozygousOgaKOandOgaD133Nmaleflies.Survivalof20fliespervialwasfolloweduntilnoflieswereleftalive.SurvivorshipcurvesshowmeansS.E.ofdatarecordedfrom9–14vials/genotype,scoringthenumberoffliesaliveevery2days.Log-ranktestsindicatedecreasedmeanlifespanforOgaKO(n280)andOgaD133N(n280)fliescomparedwiththeirgeneticbackgroundcontrol(n179)(controlversusOgaKO,c241.5,p0.001)(controlversusOgaD133N,c2104.8,p0.001).B,totaldailyactivitycountsofcontrol,OgaKO,andOgaD133NmaleDrosophilaplottedin12:12-hlight:darkcycle.OgaKO(n101)exhibiteddecreaseddailyactivitycomparedwithcontrolflies(n97)(p0.001,one-wayANOVAwithBonferroni’smultiplecomparisonstest).C,locomotionandflightperformancewereassessedintheislandassay.15fliespermeasure-mentwerethrownonawhiteplatformsurroundedwithwater.Graphsshowthepercentageoffliesthatremainontheplatformovertime(10s;meansS.E.,forcontroln23,OgaKOn13,OgaD133Nn14repeats).Thedatawerecollectedover3daysofmeasurement.D,floatingbarsdepictingthemeanS.D.AUCbasedonthegraphsshowninD,one-wayANOVAwithHolm–Sidak’smultiplecomparisonsofmeanAUC.FlightescapeperformanceofOgaKOandOgaD133Nflieswasimpairedcomparedwithcontrol(OgaKOp0.0001;OgaD133Np0.0004).E,climbinglocomotorbehaviorofOga-deficientflieswasassessedbasedontheirclimbingspeed(mm/s)inanautomatednegativegeotaxisassay.OgaKOandOgaD133Ngroupsshowedsignificantlyreducedclimbingspeedcomparedwithbackgroundcontrol,indicatinglocomotordysfunction(meansS.D.;nonparametricMann–Whitneytest,controln14,OgaD133Nn15,p0.0001;controln15,OgaKOn15,p0.0002).DrosophilaOgaisrequiredforhabituationlearningJ.Biol.Chem.(2020)295(26)8636–86468639byguestonJuly7,2020/DownloadedfrommarkerSyt(synaptotagmin)andDlg1(postsynapticdiscslarge1)proteins(Fig.4A).Morphometricfeaturesofindi-vidualmusclefourNMJswerequantifiedsemiautomatically(53,54).Wedetectedmeanvaluesofarea,perimeter,andlengthinOgaD133NandOgaKONMJssimilartothoseincon-trols(Fig.S5).Thenumbersofsynapticislands,branches,andbranchingpointswerealsounaffectedintheOgamutants(Fig.S5).However,weobservedamodestincreaseinthenumberofboutonsinOgaD133NNMJs,whichdidnotreachstatisticalsignificance(meanS.D.,34.47.6),andasignificantincreaseinOgaKO(36.57.9)comparedwithbackgroundcontrol(30.77.3)larvalNMJs(Fig.4B).ThesedataindicatethatOgapotentiallyinfluencesthenumberofsynapticboutonsattheaxonterminalsatthelarvalneuro-muscularjunction.DiscussionEarlierstudieshaveuncoveredalinkbetweenO-GlcNAcaseandlearninginmouseandratmodels(30,31).HeterozygousOga/micewithincreasedO-GlcNAclevelsexhibitedhip-pocampal-dependentspatiallearningandmemorydefects(31),whereasratstreatedwithanOGAinhibitor,thiamet-G,showedreducedperformanceinnovelobjectandplacementtests(30).Theselearningphenotypeswereassociatedwithdys-regulationofsynapticplasticity,long-termsynapticpotentia-tion,and-amino-3-hydroxy-5-methyl-4-isoxazolepropionicacidreceptor(AMPAR)–dependentlong-termsynapticde-pression(30).SeveralO-GlcNAc-modifiedproteinswerefoundthatoperateatthemammaliansynapse,suchasBassoon,Pic-colo,andSynapsin(51);regulatetranscriptionalprogramsrel-evanttosynapticplasticityinneurons,suchasthecAMP-re-sponseelementbindingprotein(CREB)(55);controlneuronalmicrotubuledynamics,suchasTau(56)andCRMP2(57);ormediatesynaptictransmission,suchastheAMPARGlu2sub-unit(30).O-GlcNAcylationontheseandotherproteinstogetherpotentiallymodulatesneuronalfunctions.Althoughthemolecularmechanismbehindtheselearningandsynapticphenotypesarenotfullyunderstood,ourcurrentknowledgeindicatesthatsynergicresponseofmultiplevoltage-gatedionchannels(58)anddysregulationofAMPARareinvolved(31,58).IncontrasttomammalianorganismswhereMGEA5/OGAiscrucialforembryonicdevelopment(13,14),OgaisnotessentialinDrosophilamelanogaster(15,16)andCaenorhabditiselegans(59).However,thefactthattheOgageneisconservedacrossinvertebratessuggeststhatmaintenanceofhomeostaticO-GlcNAclevelsbyOGAmayprovideaconsiderableadvan-tagetoMetazoa.PreviousworkhasshownthatknockdownofOgaintheflyleadstoalteredmetabolismthrougheffectsoninsulin-producingcells(60,61).Someofthephenotypes,forexamplethelife-spaneffectsthatweobservehere(Fig.2A),couldbeamanifestationofthis.However,wealsodemon-stratedthatOgaKOandOgaD133Nmutationsledtodeficitsinhabituation,highlightingthattheroleofOGAinlearningisevolutionarilyconserved.OurdataalsoprovideevidencethattheDrosophilanervoussystemissensitivetoanincreaseofthelevelofO-GlcNAcylationandabsenceofOGA,establishingitasasuitablegeneticmodelsystemtostudyunderlyingmecha-nismsandsubstratesinvolved.IthasbeenreportedpreviouslythatincreasedproteinO-GlcNAcylationcausedimpairedsynapticplasticityinOga/micewithoutaffectingdendriticspinedensityinCA1pyramidalneurons(31).Here,wereportthatboutonnumberofthelarvalNMJisaffectedinOgaKOnullmutants,showingthatsynapticmorphologyisalteredinOgaKOanimals.AlthoughOgaD133NandOgaKOcausedchangesinproteinO-GlcNAcylationtothesameextent,wedescribedbehavioralandneuronalphenotypesthatmanifestedtoadifferentdegreeinthetwoOgalines.ReductionoftotaldailyactivityandanincreaseinNMJboutonnumberwasonlyapparentinOgaKO,whereasthisgenotypeappearedtoexhibitlessseverehabitua-tiondeficits.Apossibleexplanationofthisliesinthechoiceofinactivatingmutation.Previousworkinmammalianandbac-Figure3.LossofOgaactivityinDrosophilaaffectsnonassociativelearn-inginthelight-offjumpreflexhabituationparadigm.Jumpresponsesof3–7-day-oldindividualmaleflieswereinducedbyrepeatedlight-offpulses(100trials)witha1-sintertrialinterval.Habituationwasscoredasthemeannumberoftrialsrequiredtoreachtheno-jumpcriterion(TTC).Jumpresponsecurvesshowtheaveragejumpresponse(%ofjumpingflies)over100light-offtrialsat1-sintertrialintervals.Thenumberoftrialsneededtoreachtheno-jumpcriterionispresentedasmeanTTCS.E.AandB,habituationofhomozygousOgaKOmaleflies(n63)wassignifi-cantlyslowercomparedwithcontrolflies(n65)p0.030.CandD,habituationofhomozygousOgaD133Nmaleflies(n86)wassignificantlyslowercomparedwithcontrolflies(n74)p0.001.EandF,habituationofadultflieswithneuronalknockdownofOga(elav::GAL4/;UAS-OgaR-NAi41822/,n68)wassignificantlyimpairedcomparedwithcontrolflies(elav::GAL4/,n45;p0.002).DrosophilaOgaisrequiredforhabituationlearning8640J.Biol.Chem.(2020)295(26)8636–8646byguestonJuly7,2020/DownloadedfromterialO-GlcNAcaseshasshownthattheequivalentoftheD133Nmutationinactivatestheenzyme.However,recentworkhasuncoveredthatthismutationdoesnotlosetheabilitytobindO-GlcNAcproteins;indeedthisinactivemutantcanbeusedtoenrichtheO-GlcNAcproteome(38,62).Therefore,itispossiblethattheD133Nmutationcontributestothestrongerhabituationphenotypebybindingto(partsof)theO-GlcNAcproteomein