Palmitoylation by DHHC58 Targets GRIP1_图文

2024-07-29 作者:

Article

PalmitoylationbyDHHC5/8TargetsGRIP1toDendriticEndosomes

toRegulateAMPA-RTrafficking

,1,2TakashiHayashi,1,3Shu-LingChiu,1Chih-MingChen,1,r1,*

Neuron

ofNeuroscienceandHowardHughesMedicalInstitute,JohnsHopkinsUniversitySchoolofMedicine,Hunterian1001,

treet,Baltimore,MD21205,USA

2Presentaddress:ShrinersHospitalPediatricResearchCenter(CenterforNeurorehabilitationandNeuralRepair),andDepartmentofAnatomyandCellBiology,TempleUniversityMedicalSchool,treet,Philadelphia,PA19140,USA

3Presentaddress:DepartmentofMolecularNeurobiologyandPharmacology,GraduateSchoolofMedicine,TheUniversityofTokyo,7-3-1Hongo,Bunkyo-ku,Tokyo113-0033,Japan

4Presentaddress:GraduateProgram,DepartmentofBiology,JohnsHopkinsUniversity,sStreet,224MuddHall,Baltimore,MD21218,USA

*Correspondence:rhuganir@10.1016/.2011.11.021

1Department

SUMMARY

Palmitoylation,akeyregulatorymechanismcontrol-lingproteintargeting,iscatalyzedbyDHHC-familypalmitoylacyltransferases(PATs).ImpairedPATactivityislinkedtoneurodevelopmentalandneuro-psychiatricdisorders,r,fewsubstratesforspecificPATsareknown,andfunc-tiona,weidentifythecloselyrelatedPATsDHHC5andDHHC8asspecifig,palmitoylation,anddendritictargetingofGRIP1brequireaPDZliganduniquetoDHHC5/oylatedGRIP1bistargetedtotraffitentwiththistraffickingrole,GRIP1b’spalmi-toylationturnoverrateapproachesthehighestofallreportedproteins,andpalmitoylationincreasesGRIP1b’nowledge,thesefindingsidentifythefirstneuronalDHHC5/8substrate,definenovelmecha-nismscontrollingpalmitoylationspecificity,andsuggestfurtherlinksbetweendysregulatedpalmi-toylationandneuropathologicalconditions.

INTRODUCTION

Precisetargetingofproteinstospecificsubcellularlocationsiscriticalinallcells,butitsimportanceisespeciallyapparentinhighlyspecialized,alproteintargetingmustbepreciselyregulatedtocontrolneuro-transmissionatspecificsynapses,whichinturnunderlieshigherbrainfunctionssuchassynapticplasticity,learning,andmemory

482Neuron73,482–496,February9,2012ª2012ElsevierInc.

(ShepherdandHuganir,2007;NewpherandEhlers,2008;LauandZukin,2007).

Amajormechanismthatcontrolsproteintargetingtospecificsubcellularlocationsisdirectlipidmodification,whichfacilitatesproteininteractionswithintracellularorplasmamembranes(Johnsonetal.,1994;ZhangandCasey,1996;el-HusseiniandBredt,2002;FukataandFukata,2010).Ofthethreemostcommonlipidmodifications,myristoylation,prenylationandpalmitoylation,lowsaddi-tionaldynamicregulationandmaybeonereasonwhypalmitoy-lationismorefrequentlyobservedinneuronsthanotherlipidmodifications(FukataandFukata,2010).Indeed,palmitoylationisrapidlyemergingasacriticalmodulatorofneuronalfunction,whosedisruptionislinkedtoneurodevelopmentalandneuro-psychiatricconditions(FukataandFukata,2010;Mukaietal.,2004,2008;Mansourietal.,2005;Raymondetal.,2007).

Inmammaliancells,palmitoylationiscatalyzedbyafamilyofpalmitoylacyltransferases(PATs),eachcontainingaconservedAsp-His-His-Cys(DHHC)motif(Fukataetal.,2004).ManyPATsareexpressedinneurons(Heimanetal.,2008;Doyleetal.,2008),buttwoPATs,DHHC5andDHHC8,aredetectedfarmorefrequentlythanothersatboththemRNAandproteinlevelsinneuronalstudies(Trinidadetal.,2006,2008;Muntonetal.,2007;Heimanetal.,2008;Doyleetal.,2008).ThissuggeststhatDHHC5/tentwiththishypothesis,DHHC5isimplicatedinhigherbrainfunction,sincemicewithreducedDHHC5levels(ahypomorphic‘‘genetrap’’line)showimpairedperformanceinalearningtask(Lietal.,2010).Moreover,neuronsfromDHHC8knockoutmicehaveareduceddensityofdendriticspinesandglutamatergicsynapses(Mukaietal.,2008).Inadditiontotheirphysiologicalroles,HC5gene,whichcodesforDHHC5,liesinaregionofchromosome11associatedwithbipolardisorder(Fallinetal.,2004),whiletheZDHHC8geneliesinaregionofchromosome22repeatedlyimplicatedinschizophrenia(Mukaietal.,2004;Chenetal.,2004).PalmitoylationofneuronalproteinsbyDHHC5/8is,

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

therefore,likelyesr,littleisknownregardingthedirectneuronalsubstratesofDHHC5/8.

Here,weidentifyaspecificspliceformofthemulti-PDZdomaincontainingproteinGRIP1basanovelneuronalsubstrateforDHHC5/oylatedGRIP1b,whichistargetedtotraffickingendosomes,servesasaspecificalizationplacespalmitoylatedGRIP1binaperfectpositiontomediateactivity-dependentAMPA-Rtrafficking,,palmitoylationenhancesGRIP1b’ngly,binding,palmitoylation,anddendritictargetingofGRIP1findingsnotonlyidentifyaneuronalDHHC5/8substrate,butalsodefineadditionalmechanismscontrollingpalmitoylationspecifiS

DHHC5andDHHC8BindandPalmitoylateGRIP1b

DHHC5andDHHC8arecloselyrelatedbutdiffermarkedlyinstructurefromallotherPATsbecausetheypossessgreatlyextendedC-terminaltails(Fukataetal.,2004;Ohnoetal.,2006).WehypothesizedthatthesetailsmightprovidecluestothepossiblespecificrolesandtargetsofDHHC5/icular,wenoticedthatbothtailsendwithamotifthatispredictedtobindtoPDZdomain-containingproteins(KimandSheng,2004;FengandZhang,2009).PDZdomainproteinsareheavilyimplicatedinmanyaspectsofneuronalregulationbutareespeciallyknowntocontrolthetargetingandtraffickingofgluta-matereceptors(Kornauetal.,1995;Dongetal.,1997;Srivastavaetal.,1998;Steinbergetal.,2006;Dawetal.,2000;Ostenetal.,2000;Wyszynskietal.,2002;Terashimaetal.,2008;Hanley,2008).We,therefore,hypothesizedthatDHHC5/8mightusetheirC-terminalmotiftobindspecificPDZdomainproteinsandpotentiallytorecognizethemassubstratesforpalmitoylation.

TheDHHC5andDHHC8CterminiareidenticalandconformtoatypeIIPDZligand(EISV;Figure1A;Songyangetal.,1997).AsafirststeptoaddressthepossibilitythatDHHC5/8usethisC-terminalmotiftobindspecificsubstrates,weperformedayeasttwo-hybridscreenofarathippocampalcDNAlibraryusingaC-terminalbaitthatincludedthesharedDHHC5/83106clonesscreened,four‘‘hits’’encodedaniden-ticalcentralregion(PDZdomains4–6:‘‘GRIP1-456’’)ofthemulti-PDZdomainadaptorproteinGRIP1(Dongetal.,1997).ConsistentwiththefactthattheirPDZligandsareidentical,GSTfusionsofbothDHHC5andDingofGRIP1-456wasobservedwithGSTalone,orwithDHHC5orDHHC8CterminilackingthePDZligand(Figures1Band1C).PDZligand-dependentbindingofbothDHHC5andDHHC8C-terminaltailswasalsoobservedwhentheexper-imentwasperformedinthereversedirection(todetectDHHC5/8tailsinmyc-GRIP1-456immunoprecipitates;seeFiguresS1AandS1Bavailableonline).Moreover,thesharedDHHC5/8C-terminal15AAsequencewassufficienttorobustlybindGRIP1-456(FigureS1C).

AlternativesplicingproducestwoGRIP1isoforms,GRIP1aandGRIP1b,whichdifferinauniqueN-terminalsequence(Fig-ure1D;Yamazakietal.,2001).ItwaspreviouslyreportedthatGRIP1bisspecificallypalmitoylated,althoughthePAT(s)responsiblewasnotidentified(Yamazakietal.,2001).TotestwhetherDHHC5and/orDHHC8specificallypalmitoylatesGRIP1b,weoptimizedanonradioactiveacyl-biotinylexchange(ABE)assay(Hayashietal.,2009;Wanetal.,2007;Drisdeletal.,2006).ABEisachemicalexchangeofbiotinforthio-ester-linkedacylmodifi,palmitoylation),withthere-sultingbiotinylatedproteinbeingaffinitypurifiidsthelongexposuretimesrequiredfor[3H]palmitateincorporationexperimentsandwasusedroutinelyforthisstudy,althoughmajorfindingswerealsoshownby[3H]palmitateincorporation,withessentiallyidenticalresults(Fig-ureS1DandFigures2E).

GRIP1bexpressedinHEK293Tcellswassignificantlypalmi-toylated,asdetectedbyABE,butGRIP1bpalmitoylationwasrobustlyincreasedbycoexpressionofeitherDHHC5orDHHC8(Figures1Eand1F).IncontrasttheGRIP1asplicevariantwasnotpalmitoylated,eitherwhenexpressedalone,orwithDHHC5orDHHC8(Figures1Eand1F).BecauseGRIP1adiffersfromGRIP1bonlyatitsNterminus(Figure1F),thissug-gestedthatDHHC5andDHHC8specificallypalmitoylatetheuniqueN-terminalcysteine,Cys11,,pointmutationofGRIP1bCys11toanonpalmitoylatableserineabol-ishedpalmitoylationbyDHHC5andDHHC8(datanotshown).WenextexaminedtheabilityofspecificDHHC5/cted,GRIP1bwasnotpalmitoy-latedbycatalyticallyinactivePATmutants(catalyticCysmutatedtoSer;DHHS5,DHHS8;Figures1Gand1H).Strikingly,GRIP1bwasalsonotpalmitoylatedbyDHHC5andDHHC8mutantslackingtheC-terminalPDZligand(Figures1Gand1H).Quantificationofpalmitoylated:totalGRIP1levelsfrommultipleexperimentsconfirmedtheseresults(FiguresS1EandS1F).ThesefindingssuggestthatDHHC5/8canbindandpalmi-toylateGRIP1binheterologouscells,andrequirebothcatalyticactivityandPDZdomainbindingtorecognizeGRIP1asasubstrate.

DHHC5IsaMajorNeuronalPATforGRIP1b

LittleisknownregardingtheendogenoussubcellulardistributionofDHHC5andDHHC8andtheirspecifiinsightintowhetherDHHC5and/orDHHC8palmitoylatesGRIP1inneurons,wefirstusedspecificantibodies(FigureS2A)HC5andDHHC8wereclearlydetectedindendrites(Figure2A).DHHC8waslargelysynapticallylocalized,asshownbycolocalizationwiththesynapticactivezoneproteinBassoon(Figure2A).Incontrast,DHHC5colocalizedonlyrarelywithBassoonbutwasstronglydetectedwithindendriticshafts(Figure2A).ToconfirmDHHC5distribution,c-andHA-taggedDHHC5im-munostainingmirroredthepatternseenforendogenousDHHC5,beingdetectedoccasionallyindendriticspines,butfrequentlyindendriticshafts(Figures2BandS2B).Consistentwiththisdistribution,myc-DHHC5punctacolocalizedonlyocca-sionallywiththesynapticmarkerPSD-95(Figure2B).

Neuron73,482–496,February9,2012ª2012ElsevierInc.483

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

5andDHHC8SpecificallyBindandPalmitoylateGRIP1

(A)SchematicofthestructureofDHHC5andDHHC8,showingpredictedtransmembranedomains(blue),catalyticDHHC-CysteineRichDomain(red),andidenticalC-terminal15AAsequence(yellow)terminatinginidenticalPDZligand(EISV,orange).TheregionoftheDHHC8Cterminususedasbaitforyeasttwo-hybridscreeningisindicated.

(B)HEK293Tcellsweretransfectedwithmyc-taggedGRIP1-PDZdomains4–6,togetherwitheitherGSTalone(GST),aGSTfusionoftheCterminusofDHHC5(GST-5wt),oraGSTfusionoftheCterminusofDHHC5lackingtheC-terminalPDZ-bindingmotif(GST-5DC).Inputs(leftpanels)andGSTpull-downs(rightpanels)wereimmunoblottedwithanti-mycandanti-GSTantibodies.

(C)isthesameas(B),exceptthatindicatedconstructsofDHHC8werecotransfected.

(D)quecysteine11ofGRIP1bisreportedtobepalmitoylatedinheterologouscells,butthePATwasnotidentified.

(E)HEK293Tcellsweretransfecperformedtoisolatepalmitoylatedproteins,andGRIP1levelsinABEsamplesweredetectedbyimmunoblotting(toppanel).CelllysateswereblottedtodetecttotallevelsofGRIP1(middlepanel)andHA-DHHC5(lowerpanel).NotethatpalmitoylatedGRIP1signalisonlydetectedfollowingtreatmentwithhydroxylamine(NH2OH),anessentialstepoftheABEreaction.

(F)isthesameas(E),exceptthatcellsweretransfectedwithmycHis-DHHC8andeitherGRIP1aorGRIP1b,andlysateswereblottedtodetecttotallevelsofGRIP1(middlepanel)andmycHis-DHHC8(lowerpanel).

484Neuron73,482–496,February9,2012ª2012ElsevierInc.

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

ThisextensivedendriticdistributionofDHHC5andDHHC8contrastedmarkedlytotheER/GolgilocalizationreportedformanyotherPATs(Ohnoetal.,2006).Tofurtherexplorethisdifference,wecomparedDHHC5distributionwithtwootherPDZligand-containingPATs,HC3andDHHC7localizedexclusivelywithaGolgimarker(FigureS2B)andwereabsentfromdendrites,andquantitativecomparisonofDHHC3andDHHC5dendriticdistributionconfirmedthishighlysignificantdifference(FigureS2C).DHHC5signalsalsoextendedfarbeyondthesomaticsignalseenwiththeERmarkerKDEL-CFP(FigureS2D).Together,thesedatasuggestthatDHHC5andDHHC8arepresentindendriticlocationsinneuronsthatdifferfromotherPATs,wheretheymayplayuniqueroles.

BiochemicalanalysisofDHHC5andDHHC8distributionsup-portedtheseimmunostainingdata:DHHC8wasenrichedinpostsynapticdensity(PSD)fractions,consistentwithitssynapticlocalization,whileDHHC5,thoughdetectableinPSDfractions,wasmarkedlylessenriched,consistentwithitsmoreprominentdendriticdistribution(Figure2C).FidelityofthePSDpreparationwasconfirmedbyimmunoblottingwithpre-andpostsynapticmarkers(FigureS2E).

ThedendriticlocalizationofDHHC5resemblesthepreviouslyreporteddistributionofGRIP1,whichispresentthroughoutdendriticshafts,butonlyrarelyindendriticspines(Wyszynskietal.,1999;Maoetal.,2010;Figure2D).However,previousre,therefore,developedaGRIP1b-specificantibody(characterizedinFigureS2F).TheGRIP1banti-bodyrecognizednumerousdendriticpuncta(Figure2D),whichresembledthepreviouslyreporteddistributionofGRIP1(Maoetal.,2010)andoverlappedalmostentirelywithsignaldetectedbyapan-GRIP1monoclonalantibody(Figure2D).Bycontrast,GRIP1bcolocalizedwithneitherthesynapticmarkerPSD-95(FigureS2G)northeGolgimarkerGM130(FigureS2H).Together,thesedatasuggestthatGRIP1bislargelypresentindendriticpuncta,micaldataalsosup-portedthisconclusion,assubcellulardistributionofbothGRIP1andGRIP1bwasbroadlysimilartoDHHC5(FigureS2E).

AlthoughGRIP1bispalmitoylatedinheterologouscells,[3H]palmitatelabeling,wefirstconfirmedthatGRIP1isindeedpalmitoylatedinprimaryneurons(Figure2E).ImmunoblottingofABEsampleswitha‘‘pan-GRIP1’’antibodyalsoshowedarobustsignal,confirmingGRIP1palmitoylationinbothculturedneuronsandinintactbrain(Figure2E).Tospecif-icallydetectGRIP1bpalmitoylation,ngly,thissuggestedthatahigherpercentageofGRIP1bispalmitoylatedinneuronsthanthewell-knownpalmitoyl-proteinPSD-95(Figure3A).

TheirsimilarsubcellularlocalizationsuggeststhatDHHC5isappropriatelypositionedtopalmitoylateGRIP1binneurons.

Indeed,althoughGRIP1bproteinwasdetectedasearlyas5daysinvitro[DIV](Figure2F),GRIP1bpalmitoylationwasde-tectedonlyatlatertimes(12–19DIV),coincidentwiththeappearanceofDHHC5(Figure2F).Moreover,neuronsinfectedwithlentivirusencodingasmallhairpinRNA(shRNA)thatspecif-icallytargetsDHHC5(FigureS2I)showedmarkedlyreducedlevelsofpalmitoylated,butnottotal,GRIP1(Figures2Gand2H).Importantly,asacontrolforpotentialoff-targeteffectsofshRNA,bothDHHC5levelsandGRIP1palmitoylationwererescuedbycoexpressionofshRNA-resistantDHHC5(Figures2Gand2H).DHHC5knockdownandrescuedidnotaffecteitherpalmitoylatedortotallevelsoftheknownpalmitoylproteinsFyn(Figure2G)orSNAP25(datanotshown).

DHHC5knockdowndidnotcompletelyeliminateGRIP1bpal-mitoylation,suggestingthatotherPATs,inparticularDHHC8,,whileinfectionofaDHHC8-specific(FiguresS2JandS2K)shRNAonlyslightlyreducedGRIP1palmitoylation,coinfectionwithshRNAstarget-ingDHHC5andDHHC8togetherreducedGRIP1palmitoylationtoalmostundetectablelevels(Figures2Gand2H).LevelsoftotalandpalmitoylatedSNAP25andFynwereunaffected,esultsstronglysuggestthatbothDHHC5andDHHC8canpalmitoylateGRIP1inneuron,anantibodythatrecognizesbothDHHC5andDHHC8equallyrevealedthatDHHC5isbyfarthemajorofthesetwoPATsinourneuronalcultures(FigureS2L).Thus,insubsequentexperimentswefocusedourattentiononDHHC5.

RapidGRIP1bPalmitateTurnoverSuggestsaRoleinDynamicTrafficking

Althoughpalmitoylationisreversible,ratesofpalmitateturnoveronneuronalproteinsvarywidely(HuangandEl-Husseini,2005;Kangetal.,2008).Palmitateturnoverratecanprovideinsightintothepossiblefunctionofpalmitoylation;rapidturnoversuggestsaroleindynamiceventssuchasregulatedproteintrafficking,whileslowfirststeptoaddressthefunctionalconsequenceofGRIP1bpalmitoylation,ngneuronswiththebroad-spectrumpalmitoylationinhibitor2-Bromopalmi-tate,whichblockspalmitateaddition(Webbetal.,2000;Resh,2006),allowspalmitateturnoverratetobemeasuredbytrackingkineticsofpalmitoylation‘‘rundown’’ngly,almostallpalmitateonGRIP1wasremovedafteronly1hrof2-Bromopalmitatetreatment(Figure3A).Indeed,GRIP1palmi-tatecyclingwasfaster(half-time[T1/2]approximately35min,Figure3B)thananyotherproteinthatweexamined,includingthewell-knownreversiblypalmitoylatedproteinPSD-95(El-Hus-seinietal.,2002;Figures3Aand3B)andapproachesthefastestturnoverratesreportedforanyknownpalmitoyl-protein(Magee

(G)Palmitoylat293TcellsweretransfectedwithGRIP1bpluseitheremptyvector,HA-taggedDHHC5wild-type(HA-DHHC5wt),catalyticallyinactivemutant(HA-DHHS5)orPDZligandmutant(HA-DHHC5DC).LysatesweresubjectedtoABEandimmunoblottedtodetectpalmitoylatedGRIP1(upperpanel),GRIP1expression(middlepanel),andDHHC5expression(lowerpanel).(H)isthesameas(G),exceptthatcellsweretransfectedwithGRIP1bpluseitherDHHC8wt,oFigureS1foradditionalcontrolexperimentsandquantifieddata.

Neuron73,482–496,February9,2012ª2012ElsevierInc.485

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

5/8ControlGRIP1PalmitoylationinPrimaryNeurons

(A)Top-rowpanelsshowprimaryhippocampalneuronsimmunostainedwithantibodiesagainstendogenousDHHC5(firstpanel),DHHC8(secondpanel),andthesynapticmarkerBassoon(thirdpanel).Asinglehighlighteddendrite(dashedrectangleineachupper-panelimage)isshownathighermagnifiesofcolocalizedDHHC8andBassoonpunctaareindicatedwitharrowsintheoverlaidimage.

(B)Immunostainingofprimaryhippocampalneuronscotransfectedwithmyc-taggedDHHC5andthesolublemarkermCherry,andiehighlighteddendrite(dashedrectangleineachupper-panelimage)isshownathighermagnifipleofmyc-DHHC5colocalizedwithaPSD-95-positivedendriticspineisshown(arrow),butmyc-DHHC5ismorefrequentlyfoundindendriticshafts.

(C)EqualproteinamountsoftheindicatedtyofthepreparationwasconfirmedbyimmunoblottingwithPSD-95antibody.

(D)ampalneuronswereimmunostainedwithGRIP1monoclonalandGRIP1b-specifiehighlighteddendrite(dashedrectangleineachupper-panelimage)isshownathighermagnifi1dendriticpunctathatareGRIP1b-positiveareindicatedwitharrowsintheoverlaidimage.

486Neuron73,482–496,February9,2012ª2012ElsevierInc.

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

etal.,1987;Rocksetal.,2005).ThissuggestedthatGRIP1bpal-mitoylationmostlikelyregulatesdynamiceventssuchasrapidchangesinproteintrafficking.

PalmitoylationbyDHHC5TargetsGRIP1btoDendriticEndosomes

EndogenousGRIP1bishighlypalmitoylated(Figure2E),butthesignaldetectedbyGRIP1bimmunostaining(Figure2D)doesnotd,therefore,soughttocomparetheneuronaldimitoylatedGRIP1bwasmadebymutatingthesinglepalmitoylatedcysteineresidue,Cys11,toanonpalmitoy-latableSerine(GRIP1b-C11S;Figure3C).Tomimicpalmitoyla-tion,weaddedaconsensussequencetotheGRIP1bNterminusthatdirectsadditionofmyristate(C14,fullysaturated),analmostidenticallipidtopalmitate(C16,fullysaturated),whichisattachedatanalmostidenticalpositionintheGRIP1bprotein(Figure3C).Importantly,however,myristatemodificationisirre-versible(Johnsonetal.,1994),somyristoylatedGRIP1b(Myr-GRIP1b)mimicsconstitutivelypalmitoylatedGRIP1b.

ThedistributionofGRIP1b-CSimmu-nofluorescencewasrestrictedtothecellsomaandproximaldendrites,myr-GRIP1bimmunofluorescenceextendedfarintodistaldendrites(Figure3D;quantifiedinFigure3E).Evenmoredramatically,whileGRIP1b-CSimmunofluorescencewasalmostentirelydiffuse,Myr-GRIP1bwasstrikinglypunctate(Fig-ure3D;quantifiedinFigure3F).SimilartoendogenousGRIP1b,Myr-GRIP1bpunctawerepresentthroughoutdendriticshafts,butonlyrarelypresentindendriticspines(FigureS3A,quantifiedinFiguresS3BandS3C).NumerousMyr-GRIP1bpunctaweredetectedfar(>60mm)intodistaldendrites,andtheirsizeanddistributionresembledpreviouslydescribedendogenousGRIP1/GRIP1bpuncta,whichcolocalizewithmarkersforrecy-clingendosomes,butnotwithearlyendosome,synaptic,orGolgimarkers(Maoetal.,2010;FiguresS2GandS2H).Indeed,dendriticMyr-GRIP1bpunctacolocalizedextensivelywithAlexa555transferrin(Figure3G),butonlyrarelywiththeearlyendoso-malmarkerEEA1(FigureS3D),confier,liveimagingofGFP-taggedMyr-GRIP1brevealedthatasubsetoftheserecyclingendo-somesishighlymotile(MovieS1).Together,thesedatasuggestthatmimickingN-terminalpalmitoylationtargetsGRIP1btomotiledendritictraffiand-DependentPalmitoylationbyDHHC5TargetsGRIP1btoDendriticVesicles

Thesefindingssuggestedthatwild-typeGRIP1bmightdistributebetweenr,dendriticpunctaoftrans-fectedGRIP1bwtwerefarlessnumerousthanthoseseenwithMyr-GRIP1b(Figures3Dand3F).Wehyptentwiththisnotion,trans,DHHC5wtincreasedthelevelofGRIP1bwtdetectedindistaldendrites(Figure4A;quantitatedinFigure4B).Second,DHHC5wttransformedGRIP1bwtstainingfromalargelydiffusepatterntoonethatwasstrikinglypunctate(Figure4A;quantitatedinFig-ure4C).Indeed,thenumberofGRIP1bwtpunctaindistaldendrsinGRIP1bwtdistributionwerelikelyduetodirectpalmitoylationgly,neitherDHHS5norDHHC5DCincreasedGRIP1btargetingtodendrites(Figures4A–4C),despitethenormaldendritictargetingofthesemutants(FigureS4).Together,theseresultssuggestthatpalmitoylationbyDHHC5targetsGRIP1btorecyclingendo-somesandthat,asinheterologouscells(Figure1G),thispheno-typiceffectrequiresboththePATactivityandPDZbindingabilityofDHHC5.

TherapidturnoverofpalmitateonGRIP1suggestedthatGRIP1vesicu,acutetreatment(90min)with2-Bro-mopalmitatedramaticallydispersedGRIP1punctainbothproximalanddistaldendrites(Figure5A).ThesefindingsareconsistentwithpalmitoylationreversiblytargetingGRIP1btodendriticendosomesandsuggestedthatpalmitoylationmightmodulateinteractionswithotherGRIP1partnersthatcontrolvesicletraffihtraffickingproteinisthedendriticki-nesinmotorproteinKIF5,whoseinteractionwithGRIP1iscriticalforGluA2traffickingwithindendrites(Setouetal.,2002).We,therefore,addressedwhetherGRIP1palmitoylationmightmodulateGRIP1interactionswithKIF5,bycoexpressingKIF5Cwithwild-type,nonpalmitoylatable,ngly,myristoylatedGRIP1boundmoreKIF5Cthandidwild-typeGRIP1,whileGRIP1b-C11SboundKIF5Conlyminimally(Figure5B).Moreover,inneuronsaMyr-GRIP1bmutantlackingthepreviouslyreportedKIF5-bindingdomainofGRIP1(Setouetal.,2002;myr-GRIP1b-delKBD)

(E)Highselsshowculturedneuronsmetabolicallylabeledwith[3H]palmitate,lysedandimmunoprecipitatedwiththeindicatedantibodies.[3H]signal(upperpanel)andGRIP1proteinlevels(lowerpanel)panelsillustrateculturedneuronssubjectedtoABEandimmunoblottedwithpan-GRIP1(upperpanel)orGRIP1b-specificantibodies(lowerpanel).

(F)LysatesfromcorticalneuronsculturedfortheindicatednumberofDIVwereimmunoblottedforexpressionofGRIP1b(toppanel)orDHHC5(middlepanel).LysatessubjectedtoABEwereblottedtodetectpalmitoylatedGRIP1b(bottompanel).

(G)CorticalculturedneuronswereinfectedwiththeindicatedlentivirusescontainingshRNAsequencestoknockdownexpressionofDHHC5,DHHC8,ionofeachlysatewasimmunoblottedtodetecttotallevelsofDHHC5(toppanel),GRIP1(thirdpanel),andFyn(fifthpanel).TheremaininglysatewassubjectedtoABEtodetectpalmitoylatedGRIP1(secondpanel)andFyn(fourthpanel).

(H)QuantitationofmultipleexperimentsconfirmsthatDHHC5knockdownsignificantlyreducesGRIP1palmitoylation(56.8%±7.2%ofcontrolvirus[mean±SEM],n=8,p<0.05comparedtocontrolvirus[FUGW]alone,ttest),shRNA-resistantDHHC5(‘‘rescue’’)restoresGRIP1palmitoylationlevels(95.5%±7.5%ofcontrolvirus,n=8,notsignificantfromFUGWalone),andGRIP1palmitoylationisalmostcompletelyeliminatedbyknockdownofbothDHHC5andDHHC8(6.9%±2.6%ofcontrolvirus,n=sksindicatesignificantdifferencefromcontrol(FUGW)-infectedneurons(p<0.05,ttest).SeealsoFigureS2.

Neuron73,482–496,February9,2012ª2012ElsevierInc.487

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

oylationTargetsGRIP1btoSpecificDendriticEndosomalVesicles

(A)Primaryneuronsweretreatedwith100mM2-Bromopalmitatefortheindicatedtimesorwith0.1%(v/v)EtOH(solventcontrol).ABEreactionswereperformedtodetectpalmitoylatedandtotallevelsofGRIP1andPSD-95,asindicated.

(B)Quantitationofpalmitateturnoverfrommultipleexperimentsasin(A),plottedasmean±SEMforfourdeterminationspertimepoint.

(C)SchematicofpalmitoylatedGRIP1bwild-type,comparedtononpalmitoylatable(GRIP1b-C11S)andconstitutivepalmitoylation-mimic(Myr-G1b).Themyristoylationconsensusaddsasimilarlipid(C14,saturated)ataminoacid2(Gly),followingcleavageoftheinitiatingmethionine.

(D–F)DendritictargetingofGRIP1bbylipidmodification.(D)RepresentativeimagesofprimaryhippocampalneuronstransfectedwithGFPasamorphologymarker,pluseitherGRIP1bwt-myc(GRIP1wt,toprow),GRIP1b-C11S-myc(GRIP1CS,middlerow)orMyr-GRIP1b-myc(Myr-GRIP1,bottomrow),followingfixationandstainingwithantibodiesagainstGFP(firstcolumn)andmyc(secondcolumn).Therightcolumnshowsmycsignals,thresholdedatanidenticalvalueforeachtransfectedneuron.(E)Quantitationofaveragefluorescence(mean±SEM)incellsomaandindendriticsegmentsattheindicateddistancesfromthesoma,inneuronsexpressingGRIP1bwt-myc(n=14),GRIP1b-C11S-myc(n=16),orMyr-GRIP1b-myc(n=15).Asterisksindicatesignificantdifference(p<0.05)fromGRIP1bwt.(F)NumberofGRIP1bpunctaperdendriticsegment(mean±SEM)foreachneuronfrom(E)(plottedforthreedendriteseachfromn=14,n=16andn=15neurons)attheindicateddistancesfromthecellsoma.

488Neuron73,482–496,February9,2012ª2012ElsevierInc.

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

showedmarkedlyreducedtargetingtodistaldendrites(Fig-ures5Cand5D).ThesefindingssuggestfirstthatGRIP1bpalmi-toylationfavorsGRIP1b-KIF5binding,thattheseinteractionsactinconcerttocontrolGRIP1bdendritictargeting,andmayenhancetheabilityofGRIP1btobridgevesicularcargoes,particularlyglutamatereceptors,ne-AssociatedGRIP1bAcceleratesAMPAReceptorRecycling

GRIP1regulatesAMPAreceptortargetingtodendritesandtherecyclingofAMPAreceptorstotheplasmamembranefollowingNMDAreceptor(NMDAR)activation(Setouetal.,2002;Maoetal.,2010).We,therefore,hypothesizedthatGRIP1bpalmitoy-lationmightinturnaffectGRIP1b’essthispossibility,wetransfectedhippocampalneuronswithwild-type,nonpalmitoylatable,orconstitutivelymembrane-targetedformsofGRIP1b,togetherwithapHluorin-taggedGluA2AMPAreceptor,towhichGRIP1directlybinds(Dongetal.,1997;Maoetal.,2010).ThepHluorintagfluorescesbrightlyatneutralpH,reatmentwithNMDAdrivesinternalizationofpHGluA2torecyclingendosomes,whoseacidity(pH<6.6)dramaticallyquenchespHGluA2fluores-cence,whileNMDAwashoutinducespHGluA2recyclingtotheplasmamembraneandfluorescencerecovery(Ashbyetal.,2004,LinandHuganir,2007;Thomasetal.,2008;Maoetal.,2010;Figure6A).FluorescenceofpHGluA2,therefore,actsasareadoutofreceptordistributionandciculartheT1/2offluorescencerecoverytime,derivedfromasingleexponentialfitoftherecyclingphase,providesaquantitativemeasureofrecyclingrate.

InneuronstransfectedwithGRIP1bwtorGRIP1bC11S,ratesofpHGluA2internalizationandrecyclingwerehighlysimilartoneuronstransfectedwithvectoralone(Figure6B).However,pHGluA2recyclingwasmarkedlyacceleratedinneuronstrans-fectedwithMyr-GRIP1b(Figure6C).ThisacceleratedrecyclingwasalsoseeninneuronstransfectedwithDHHC5,whichispre-dictedtoincreasepalmitoylationofendogenousGRIP1b(Figures6DandFiguresS5A).BothMyr-GRIP1bandDHHC5causedacceleratedrecyclingofbothsomaticanddendriticpHGluA2(Figures6andFiguresS5B–S5E).Theeffectoftrans-fectedDHHC5islikelyduetodirectpalmitoylationofGRIP1b,asalthoughGluA2isaknownpalmitoylatedprotein(Hayashietal.,2005),itisnotdetectablypalmitoylatedbyDHHC5(Fig-ureS5B).AMPAreceptorrecyclingis,therefore,significantlyacceleratedunderconditionswhereGRIP1bmembraneattach-mentisenhanced(Figures6EandFiguresS5E).Myr-GRIP1b,whichistargetedtotraffickingvesicles,alsocolocalizedexten-sivelywithpH-GluA2indendriticpunctainfixedneurons(Fig-ureS5C),suggestingthateffectsontraffickingwerelikelyduetoadirectGRIP1b-pHGluA2interaction.

DISCUSSION

Here,wereportthattwoPATsuseanovelPDZdomainrecogni-tionmechanismtopalmitoylateandcontrolthedistributionandtraffieofGRIP1bpalmitoylationisdistinctfromthatobservedformanypalmitoyl-proteins:palmi-toylationtargetsGRIP1btomotiletraffickingvesiclesinneuronaldendrites,findingsareconsistentwithboththedendriticlocalizationofthemajorGRIP1PAT,DHHC5,andtheknownroleofGRIP1inthedendritictraffickingofitsinteractingpartners,mostnotablyAMPA-typeglutamatereceptors(Setouetal.,2002;Maoetal.,2010).Why,though,ispalmitoylatedGRIP1bnotdetectedattheplasmamembrane,asobservedforseveralotherpalmitoylatedproteins?AlikelyexplanationisthattheGRIP1bNterminuslacksasecondmembrane-targetingsignal,suchasanadditionallipidmodificationsiteorapolybasicsequence(Sigaletal.,1994;DunphyandLinder,1998;Resh,2006).‘‘Twosignal’’modificationofthistypeisessentialforplasmamembranetargetingofGFP,whileGFPmodifiedwithonlyasinglelipidandlackingapolybasicsequencelocalizestointracellularvesiclesthataremostlikelyendosomes(McCabeandBerthiaume,2001).The‘‘singlesignal’’presngdatabasesforconservedN-terminalcysteinessur-roundedbynonbasicresiduesmaywellrevealfurtherproteinsthataretargetedtovesiclesbypalmitoylation.

Severallinesofevidencesupporttheconclusionthatpalmi-toylatedGRIP1bistargetedtodendriticendosomes;endoge-nousGRIP1b,whichishighlypalmitoylated,showsadener,DHHC5targetsGRIP1bwt,butnotthepalmitoylationmutantGRIP1b-C11S,y,though,theendosomaltargetingofpalmitoylatedGRIP1bisdistinctfromthesynaptictargetingdescribedforthecloselyrelatedpalmitoylatedGRIP2b(DeSouzaetal.,2002;Misraetal.,2010).Consistentwiththesereports,wealsoobservedprominentGRIP2btargetingtodendriticspines,whichdidnotrequireDHHC5orDHHC8coexpression(datanotshown).AlthoughGRIP1andGRIP2cancompensateforoneanotherincerebellarPurkinjeneurons(Takamiyaetal.,2008),tworelatedissuesl,plasmamembrane/synaptictargetingofGRIP2bisconsistentwiththeadditionalbasicresiduesthatsurroundthepalmitoylatedcysteineattheGRIP2bNterminus,,whilethePDZdomainsofGRIP1andGRIP2arehighlyhomologous,theKIF5-bindingregionofGRIP1(betweenPDZ6andPDZ7;Setouetal.,2002)ispoorlyconservedinGRIP2,suggestingthatGRIP1isuniqueinihersupportoftheirdistinctregulation,wealsoobservednoincreaseinGRIP2bpalmitoylationbyeither

(G)-GRIP1b-myctransfectedneuronswerelivelabeledwithAlexa555transferrin,fixed,anelsshowhigh-magnifiivelycolocalioFigureS3.

Neuron73,482–496,February9,2012ª2012ElsevierInc.489

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

ticActivityandPDZDomainRecognitionbyDHHC5TargetGRIP1bwttoDendriticVesicles

(A)Representativeimagesofprimaryhippo-campalneuronstransfectedwithGFPasamorphologymarker,GRIP1bwt-mycorGRIP1b-C11S-mycandeitheremptyvector,HA-DHHC5wt,HA-DHHS5orDHHC5DC,followingfixationandstainingwithantibodiesagainstGFP(firstcolumn),myc(secondcolumn),andHAtags(fourthpanels).Thethirdcolumnshowsmycsignal,thresholdedatanidenticalvalueforeachtransfectedneuron.

(B)AveragefluorescenceintensityofGRIP1b-myc(mean±SEM)wasplottedasinFigure3EforGRIP1bwt-myc(n=14),GRIP1bwt-mycplusHA-DHHC5wt(n=19),GRIP1bC11SplusHA-DHHC5wt(n=12),GRIP1bwtplusHA-DHHS5(n=15),andGRIP1bwtplusHA-DHHC5DC(n=14).(C)DendriticpunctaofGRIP1b-mycwerecountedandplotted(mean±SEM)rGRIP1bwt-mycfromFigure3arereplottedin(B)and(C).Asterisksindicatesignificantdifference(p<0.05,ttest)tativeanalysisofDHHC5dendriticdistributionconfirmedthatDHHC5mutantstargetequallywell,ifnotbetter,thanDHHC5wttodistaldendrites(seealsoFigureS4);thus,impairedtar-getingdoesnotaccountfortheirinabilitytoregulateGRIP1bwtdistribution.

DHHC5orDHHC8intransfectedheterologouscells(datanotshown),andamuchslowerGRIP2palmitoylationturnoverrateinneurons,comparedtoGRIP1b(datanotshown).Together,

490Neuron73,482–496,February9,2012ª2012ElsevierInc.

thesefindingssuggestthatpalmitoylatedGRIP1bplaysauniqueroleinendosomaltraffickingandcouplingtokinesinmotorproteins.

OurfindingthatGRIP1bpalmitoylationspecificallyaffectsactivity-dependentAMPA-Rrecyclingwouldappeartodifferfromarecentreport(HanleyandHenley,2010),r,wesuspectthatexperimentaldifferenceslikelyunderliethisdiscrep-ancyandthatourfindingsmoreaccu-ratelyreflicular,HanleyandHenleyusedSindbisvirusinfectiontoexpressGRIP1b,asystemthathastwokeyissueswhenusedtostudyintracellulartraf-fi,hostcellproteinsynthesisisshutdown,complicatingtheanalysisofintracellulartraffi,GRIP1bisoverexpressedathighlevels,leadingtointracellularaggre-gation(visibleinsomeimagesfromthisreport;HanleyandHenley,2010).More-over,theauthorsusedalarge,N-terminal

(YFP)tagclosetothesiteofGRIP1bpalmitoylation,whichmaywellaffectregulationofGRIP1bpalmitoylationand/orfunctionaldownstreameffectsthatdependonthismodification.

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

urnoverofPalmitateonGRIP1ModulatesDendriticTargeting

(A)elsshowunprocessepanelsillustrateGFPsignal(morphologymarkertooutlineindividualdendrites).LowerpanelsindicatethresholdedpunctawithintheboundaryofanindividualGFP-transfectedneuronfanelisaplotofthenumberofendogenousGRIP1puncta(mean±SEM)countedatindicateddistancesfromthecellsomaforthreedendritespercellfromn=19cells(control)andn=21cells(2-Br).

(B)Membrane-associatedGR293cellsweretransfeunoprecipitateswereimmunoblottedforcoimmunoprecipitatedHA-KIF5C.

(C)elsarerepresentativeimagesofprimaryhippocampalneuronstransfectedwithGFPasamorphologymarker,pluseitherfull-lengthMyr-GRIP1b-myc(Myr-GRIP1b)orMyr-GRIP1blackingKBD(Myr-GRIP1b-delKBD),followingfixationandstainingwithantibodiesagainstGFP(firstrow)andmyc(secondrow).Thebottomrowshowsmycsignals,thresholdedatanidenticalvalueforeachtransfectedneuron.(D)Averagefluorescenceintensity(mean±SEM)ofmyc-taggedconstructsinsomaanddendrites,plottedasinFigure3EforMyr-GRIP1b(n=13)andMyr-GRIP1b-delKBD(n=12).Doubleasterisksindicatesignificantdifferences(p<0.01,ttest)betweenconditionsatalldistancesfromsoma.

Werecentlydevelopedamorephysiologicalgeneticmanipu-lationapproach(Maoetal.,2010)proachallowedustorevealaspecificroleforGRIP1inactivity-dependentrecy-clingofbothendogenousandexogenous(pHluorin-tagged)rast,weobserfindingsreportedherearehighlyconsistentwiththereportbyMaoetal.(2010)

Neuron73,482–496,February9,2012ª2012ElsevierInc.491

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

ingGRIP1btoMembranesIncreasesActivity-DependentGluA2Recycling

(A)RepresentativeimagesoffluorescencechangeofahippocampalneuronexpressingpH-GluA2,exposedatt=10–20mMNMDAfor5minandthenallowedtorecoverfollowingNMDAwashout.

(B)Fluorescencechange(mean±SEM)duringinitialincubation,perfusionatt=10minwith20mMNMDAfor5min(bluebar)andwashout,plottedforhippocampalneuronstransfectedwithpHluorin-taggedGluA2(pH-GluA2)pluseitheremptyvector(blackcircles,n=11),myc-taggedGRIP1bwild-type(G1bwt,redcircles,n=5)orGRIP1b-C11S(G1bCS,bluecircles,n=7).

(C)and(D)arethesameas(B),butforneuronstransfectedwithpH-GluA2plusMyr-GRIP1b(Myr-G1b,orangecircles,n=6),orwithHA-taggedDHHC5(greencircles,n=8),inginDHHC5-transfectedneuronsisstillacceleratedwhenthefluorescencedecreaseisscaledtomatchvectorcontrols(FigureS5A).

(E)Timeconstants(T1/2)offluorescencerecovery,plottedasmean±SEMforthecurvesin(B)–(D).AsterisksindicatesignificantdifferencefrompHGluA2alone(p<0.05,ttest).SeealsoFigureS5.

andwithrecentworkfromourcollaborators(Mejiasetal.,2011),er,inthisstudywealsodeliberatelytrans-fectedonlysmallamountsofplasmidDNA,expressingGRIP1fromaweakpromoter(seeExperimentalProcedures)P1bconstructscarriedonlyasmallC-terminalmyctagfarfromthesiteofpalmitoylation,cefrommultiplereadouts,usingbothendogenousandexogenousAMPA-Rs,therefore,suggeststhatthepredominantphysiologicalroleof

492Neuron73,482–496,February9,2012ª2012ElsevierInc.

GRIP1istocontrolactivity-dependentAMPA-Rrecycling,andthatpalmitoylatedGRIP1benhancesthisprocess.

Wenotethat,inadditiontoGRIP1bdescribedhere,theirprominentdendriticdistributionsuggeststhatDHHC5/8arewellghDHHC5doesnotpalmitoylateGluA2(FigureS6B),itstargetsmayincludeotherAMPA-Rsubunits(Hayashietal.,2005),NMDARs(Hayashietal.,2009),orotherPDZdomainadaptorproteins(FukataandFukata,2010),,otherDHHC5

Neuron

GRIP1PalmitoylationandTargetingbyDHHC5/8

substratesmayu,theslightlyreducedpHGluA2fluo-rescencedecreaseseenwithDHHC5transfection;Figure6D).Inaddition,thepresenceoftransfectedDHHC5inlong,aspinyneuritesthatarelikelyaxons,suggeststhatDHHC5maypalmi-toylateadditionalaxonal/presynapticsubstratesinadditiontoitsdendriticregulationofGRIP1bdescribedhere.

TheidentificationofadditionalDHHC5/withinterestthatotherPATscannotcompensateforlossofDHHC5/8topal-mitoylateGRIP1inneurons,andintransfectedcellsevenPATsthatdisplaybroadsubstratespecificity(DHHC3,DHHC7;Fukata

etal.,2004;Ferna

´ndez-Hernandoetal.,2006;Greavesetal.,2008;Ponimaskinetal.,2008;Tsutsumietal.,2009)orpreferen-tiallypalmitoylatecysteineslocatedclosetotheNterminioftheirsubstrates(DHHC20;DraperandSmith,2010)donotpalmitoy-lateGRIP1b(FigureS1).ThesefindingssuggestthatGRIP1bpalmitoylationbyDHHC5/8hasdistinctrequirements,namelythatthePDZdomaininteractionuniqu5,inparticular,isamajorGRIP1bPATinneuronsbutcannotpalmitoylateseveralotherpalmitoyl-proteins(Fukata

etal.,2004;Ferna

´ndez-Hernandoetal.,2006;Greavesetal.,2008;Tsutsumietal.,2009),suggestingthatPDZdomain-dependentrecognitionisakeydeterminantofDHHC5substratespecificity.

MultiplestudieslinkDHHC5andDHHC8tobothnormalhigherbrainfunctionandneuropsychiatricdisease(Mukaietal.,2004,2008;Lietal.,2010).However,toourknowledge,noneuronalsubstrateshavebeenidentifiedforDHHC5,andalthoughPSD-95palmitoylationisreducedinDHHC8knockoutmice(Mukaietal.,2008),otherPATsarealsoreportedtodirectlypalmitoylatePSD-95inneurons(Noritakeetal.,2009),,ouridentificationofGRIP1basthefirstbonafideneuronalsubstrateforDHHC5/8hasbroadimplications,sinceGRIP1isalsogeneticallylinked

toneuropsychiatricconditionsandtoautism(Grataco

`setal.,2009;Mejiasetal.,2011).Thisraisesthepossibilitythatabnormaldendriticand/orsynapticpalmitoylationofPDZdomainp,anotherPDZdomainproteinlinkedtoneuropsychiatricdiseaseisalsopalmitoylatedbyDHHC5andDHHC8inaPDZligandmanner(G.M.T.,T.H.,andR.L.H.,unpublisheddata).ThesefindingsraisethehopethattherapeutictargetingofspecificPATsand/ortheirinteractionswithspecificsubstratesmayprovideanewapproachtobettertherapeutictreatmentsforthesediseases.

EXPERIMENTALPROCEDURES

Antibodies

Thefollowingantibodies,fromtheindicatedsources,onalantibodieswereGRIP1,Fyn,GM130(BDBiosciences),PSD-95(K28/43;NeuromAb),GFP(3E6;Invitrogen),myc,HA11(Covance),andHAF-7(SantaCruzBiotechnology).PolyclonalantibodieswereDHHC5(Sigma-Aldrich),ZDHHC8(EverestBiotech),andrabbitanti-HA(QEDBioscience).AntibodyagainsttheCterminusofGRIP1hasbeenpreviouslydescribed(Dongetal.,1997).AnantibodyraisedagainsttheuniqueNterminusofGRIP1b(aminoacids5–19;KKNIPICLQAEEEQER)wasaffinitypurifiedusing

ye-conjugatedfluorescentsecondaryantibodiesandAlexatransferrinwerefromInvitrogen.

MolecularBiologyandcDNAClones

AllmammalianDHHC5andDHHC8sequencesreportedshareanidenticalC-terminal15aminoacids,-terminal109aminoacid‘‘bait’’fromhumanDHHC8(Ohnoetal.,2006)wassubclonedintothepPC9thatgrewonquadruple-deficientplates(Leu-,Trp-,His-,Ade-)wereselected,-tiveclonesweresubclonedintomyc-taggedpRK5mammalianexpressionvector,andCterminiofbothDHHC5andDHHC8weresubclonedintoamammalianGSTfusionvector(Thomasetal.,2005)forbindingexperimentsinmammaliancells.

Full-lengthuntaggedratGRIP1aandmouseGRIP1bcDNAsinpBKexpres-sionvectorhavebeenpreviouslydescribed(Dongetal.,1997;Yamazakietal.,2001).toylationconsensussequence(MGQSLTT;Wyszynskietal.,2002)istoylationconsensuscontainsnopolybasicsequencethatmightaffectmembranetargeting,andMyr-GRIP1bcontainedamutatedCys11->Ser,sothatonlyasinglelipidmodificationoccurs,eimaging,full-lengthMyr-GRIP1bsequencewasamplifi-taggedmouseDHHC5andDHHC8andmycHis-taggedhumanDHHC8cDNAhavebeenpreviouslydescribed(Fukataetal.,2004;Ohnoetal.,2006).Catalyticallyinac-tive(DHHC->DHHS)anddeltaC(DC)mutants(lackingthelastfiveaminoacidsthatconstitutethePDZligand)viouslyreportedkinesin-bindingdomain(KBD;Setouetal.,2002)ofGRIP1bwasdeletedbySplicingbyOverlapExtension(SOE)-basedPCRusingtheMyr-GRIP1b-myccDNAastemplatetogenerateMyr-GRIP1b-myc-deltaKBD.

LentiviralInfectionandshRNAKnockdown

shRNAs(invectorpLKO;MissionshRNAlibrary)targetingsequencesidenticalinbothratandmouseDHHC5(50-CCTCAGATGATTCCAAGAGAT-30)orDHHC8(50-CTTCAGTATGGCTACCTTCAT-30)weretestedfortheirabilitytoreduceexpressioonfirmingthatthesesequenceseffectivelyandspecificallysuppressedexpressionofDHHC5andDHHC8,respectively,eachsequencewasamplifiedbyPCR,ultantH1-shRNAcassettesweresubclonedintothePacIsiteofthelentiviralFUGWvector(Loisetal.,2002)andverifi-resistantDHHC5wasgeneratedbymutatingfivenucleotideswithintheshRNAtargetsequence,sultant‘‘rescue’’cDNAwasamplifiedbyPCRwithSalIandNotIprimersandinsertedintoamodifiedFUGWvectorbyreplacingtheGFPcassettewithmyc-taggedDHHC5.

fly,HEKatantcontainingviruswasharvestedat48and72hrposttransfection,concentratedbyultracentrifugation,resuspendedinNeurobasalmedium,swerelysedat16DIV.

Biochemistry

Allbiochemicalexperimentswereperformedatleastthreetimes,fiedanalysisofcertainexperimentsispresentedinFigureS1.

RadioactiveandNonradioactivePalmitoylationAssays

[3H]palmitatelabelingof293Tcellsandculturedneuronswasperformedasdescribed(Hayashietal.,2005,2009).ABEassaywasperformedasdescribed(Hayashietal.,2009),similartopublishedprotocols(Drisdeletal.,2006;Wanetal.,2007).ForneuronalABEexperiments,neuronswerelyseddirectlyinbuffercontaining2%SDSand20mMmethyl-methanethiosulfonate(MMTS,toblockfreethiols).2-Bromopalmitatewaspreparedasa100mM

Neuron73,482–496,February9,2012ª2012ElsevierInc.493

stockinethanolandaddedtoneuronsataficulturesweretreatedwithsolventcontrol(0.1%[v/v]ethanol).

ForABEanalysisofforebrain,onemouse(P21)forebrainwashomogenizedinice-coldbuffercontaining10mMHEPES(pH7.4),0.32Msucrose,20mMMMTS,genizedtissuewaspelletedbycentri-fugationat2,1003g,andthesupernatantwasrapidlywarmedtoroomtemperature,adjustedto1%(w/v;finalconcentration)SDS,centrifugedat27,0003gtoremoveinsolublematerial,ection

HEK293Tcellsweretransfectedusingacalciumphosphate-basedmethodaspreviouslydescribed(Thomasetal.,2005).NeuronsweretransfectedusingaLipofectamine-basedmethodandusedforliveimagingorfixedwithparafor-maldehyde(seebelow)either10–16hrlater(forGRIP1transfections)or72hrlater(forpHluorin-GluA2transfections).

CoprecipitationinHEK293TCells

HEK293TcellsweretransfectedwithpCISvectorconstructstoexpressGSTalone,GSTfusionsofDHHC5andDHHC8wild-typeorDCCtermini,erelysedinimmunopreciptationbuffer(IPB;Thomasetal.,2005)blemate-rialwaspelletedbycentrifugation,andthesupernatants(termedlysates)wereincubatedwithGlutathioneSepharose(GEHealthcare).BeadswerewashedextensivelywithIPB,denaturedinSDSsamplebuffer,retransferredP1-KIF5binding,eachGRIP1construct,containingaC-terminalmyctag,erelysed16hraftertransfectionandlysatesprocessedasabove.

CulturedPrimaryNeurons

E18embrmalsweretreatedinaccordancewalneuronswerepreparedasdescribed(Thomasetal.,2008)andusedat16–ampalneuronsoncoverslipswerepreparedbythemethodofGoslinandBanker(1998)forfixedimmunostainingofendogenousproteins;oraspreviouslydescribed(LinandHuganir,2007)ectionwasperformedat15–ronalexperimentswereper-formedfromtheindicatednumbersofindividualneurons,datafromeachconditionareplottedasmean±SEM,andstatisticalsignifistaining

NeuronsoncoverslipswerefixedinPBScontaining4%(w/v)sucroseand4%(w/v)lipswerewashedwithPBSandcellspermea-bilizedwithPBScontaining0.25%(w/v)ingbriefwashingwithPBS,coverslipswereblockedovernightat4

本文标签： Palmitoylation by DHHC58 Targets GRIP1图文