THEJOURNALOFBIOLOGICALCHEMISTRYVOL.289,NO.48,pp.33258–33274,November28,2014 ©2014byTheAmericanSocietyforBiochemistryandMolecularBiology,Inc. PublishedintheU.S.A. (cid:2) Cyclopamine Modulates -Secretase-mediated Cleavage of Amyloid Precursor Protein by Altering Its Subcellular Trafficking and Lysosomal Degradation* Receivedforpublication,June30,2014,andinrevisedform,September23,2014 Published,JBCPapersinPress,October3,2014,DOI10.1074/jbc.M114.591792 AnnaG.Vorobyeva‡,RandallLee‡1,SeanMiller‡1,CharlesLongen§,MichalSharoni‡,PreetiJ.Kandelwal‡, FelixJ.Kim§,DanielR.Marenda‡¶,andAleisterJ.Saunders‡¶(cid:2)2 Fromthe‡DepartmentofBiology,DrexelUniversity,Philadelphia,Pennsylvania19104andtheDepartmentsof§Pharmacology andPhysiology,¶NeurobiologyandAnatomy,and(cid:2)BiochemistryandMolecularBiology,DrexelUniversityCollegeofMedicine, Philadelphia,Pennsylvania19102 Background:SterolscanalterAPPmetabolism. Results: Cyclopamine, a phytosterol, alters APP-CTF degradation rate, decreases APP-CTF bioavailability for (cid:2)-secretase cleavage,andreducesA(cid:3)andAICDgeneration. Conclusion:CyclopaminedecreasesA(cid:3)andAICDproductionbyalteringAPP-CTFretrogradetrafficking. Significance:CyclopamineisanovelmodulatorofAPPmetabolismandtrafficking,whichcanilluminatenewavenuesfor Alzheimerdiseasetreatment. Alzheimer disease (AD) is a progressive neurodegenerative cally characterized by senile plaques, composed of amyloid-(cid:3) diseaseleadingtomemoryloss.Numerouslinesofevidencesug- (A(cid:3)),andneurofibrillarytangles,composedofhyperphosphory- gestthatamyloid-(cid:3)(A(cid:3)),aneurotoxicpeptide,initiatesacas- lated Tau. The amyloid cascade hypothesis suggests neurotoxic cadethatresultsinsynapticdysfunction,neuronaldeath,and A(cid:3)initiatesaseriesofeventsthatresultinsynapticdysfunction eventuallycognitivedeficits.A(cid:3)isgeneratedbytheproteolytic leading to neuronal loss (3–9). The A(cid:3)peptide is produced by processingoftheamyloidprecursorprotein(APP),andaltera- proteolysisofamyloidprecursorprotein(APP)(10–15). tionstothisprocessingcanresultinAlzheimerdisease.Usingin APPisatypeItransmembraneproteincontaininganintra- vitroandinvivomodels,weidentifiedcyclopamineasanovel cellular C-terminal domain and a larger external N-terminal regulatorof(cid:2)-secretase-mediatedcleavageofAPP.Wedemon- domain(16,17).APPisproteolyticallyprocessedvianonamy- strate that cyclopamine decreases A(cid:3)generation by altering loidogenic or amyloidogenic pathways. After translation, N- APPretrogradetrafficking.Specifically,cyclopaminetreatment and O-glycosylated APP (mature APP) is trafficked to the reducedAPP-C-terminalfragment(CTF)deliverytothetrans- plasmamembranewhere(cid:4)-secretase(nonamyloidogenic)can Golgi network where (cid:2)-secretase cleavage occurs. Instead, cleavefull-lengthAPP(FL-APP)liberatingasolubleN-termi- cyclopamineredirectsAPP-CTFstothelysosome.Thesedata nal fragment (sAPP(cid:4)) and a membrane tethered C-terminal demonstrate that cyclopamine treatment decreases (cid:2)-secre- fragment(APP-CTF(cid:4)).Alternatively,FL-APPcanbeendocy- tase-mediatedcleavageofAPP.Inaddition,cyclopaminetreat- tosed and either recycled back to the plasma membrane or mentdecreasestherateofAPP-CTFdegradation.Together,our localizedtotheearlyendosomewhere(cid:3)-secretasecleavageof datademonstratethatcyclopaminealtersAPPprocessingand FL-APP initiates amyloidogenic cleavage. This (cid:3)-secretase A(cid:3)generationbyinducingchangesinAPPsubcellulartraffick- cleavageliberatesasolubleN-terminalfragment(sAPP(cid:3))anda ingandAPP-CTFdegradation. membrane-tethered C-terminal fragment (APP-CTF(cid:3)) (18). Followingthisinitialcleavagebyeither(cid:4)-or(cid:3)-secretase,ret- rogradetraffickingmechanismsdeliverAPP-CTFstothetrans- Alzheimer disease (AD)3 is the most common form of Golginetwork(TGN).Recentliteraturedemonstratesthatthe dementia.Itisanirreversibleneurodegenerativediseasecharac- majorityof(cid:2)-secretasecleavageofAPPoccursattheTGN(19). terizedbygradualcognitivedecline(1,2).ADisneuropathologi- Here,APP-CTFsarecleavedby(cid:2)-secretasetogeneratetheAPP intracellulardomain(AICD)andp3orAICDandA(cid:3)depen- *Thisworkwassupported,inwholeorinpart,byNationalInstitutesofHealth dingonwhethertheAPP-CTFwasderivedfrom(cid:4)-or(cid:3)-secre- GrantR01NS057295.ThisworkwasalsosupportedbyDrexelUniversity tase,respectively. funding(toA.J.S.). APPproteolysisisintimatelyassociatedwithitssubcellular 1Bothauthorscontributedequallytothiswork. 2Towhomcorrespondenceshouldbeaddressed:Dept.ofBiology,Drexel localization; therefore, APP trafficking plays a critical role in University,3245ChestnutSt.,Philadelphia,PA.Tel.:215-895-6772;Fax: amyloidogenesis.Cholesterolcanaltercellularmembraneflu- 215-895-1273;E-mail:[email protected]. idityandthereforecanaltertraffickingoftransmembranepro- 3Theabbreviationsusedare:AD,Alzheimerdisease;APP,amyloidprecursor protein; CTF, C-terminal fragment; A(cid:3), amyloid-(cid:3); AICD, APP intracellular teinslikeAPP(20,21).Cholesterolhasbeenshowntoplayan domain; BisTris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane- importantroleinADriskandpathogenesis(22–24).Highcho- 1,3-diol;CHAPSO,3-[(3-cholamidopropyl)dimethylammonio]-2-hydroxy-1- lesterollevelshavebeenimplicatedasanADriskfactor(25–27). propanesulfonicacid;TGN,trans-Golginetwork;sAPP,solubleAPP;MVB, multivesicularbody;NICD,Notchintracellulardomain. Clinically, patients treated with statins, inhibitors of cholesterol This is an Open Access article under the CC BY license. 33258 JOURNALOFBIOLOGICALCHEMISTRY VOLUME289•NUMBER48•NOVEMBER28,2014 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization synthesis,havereducedriskforAD(23,24,28).Increasingcholes- PrimaryNeuronCulture—Primarycorticalneuroncultures terollevelsininvitroandinvivoADmodelsexacerbatesA(cid:3)pro- were isolated from postnatal day 1 (P1) Sprague-Dawley rat duction(25,29,30).Conversely,inhibitingcholesterolsynthesisin pups.Briefly,corticesweredissectedout,minced,andtreated vitroorinvivoreducesA(cid:3)generation(31). withpapain(100units;Worthington)for15minat37°C.Next, Recently,phytosterolswerealsodemonstratedtomodulate tissuewastreatedwithtypeII-Otrypsininhibitorfromchicken A(cid:3)generation. Stigmasterol treatment decreased A(cid:3)genera- eggwhite(Sigma)for15minat37°C.Tissuewaswashedwith tion by modulating (cid:2)-secretase activity and (cid:3)-secretase traf- fresh Neurobasal medium (Invitrogen) supplemented with ficking(32).Cyclopamineisanaturallyoccurringplantphytos- B-27 (Invitrogen), 2 mM L-glutamine, 100 units/ml penicillin, terol from the corn lily (Veratrum californicum) plant (33). and100(cid:5)g/mlstreptomycin.Tissuewastriturated,centrifuged Cyclopamine and its analogs have been used to treat cancer, at1000rpmfor10min,andthenresuspendedinthefreshand specifically, in combination with lovastatin to treat medullo- complete Neurobasal medium. 2 (cid:3) 106 cells per 35-mm well blastoma(34,35). wereplatedontopoly-DL-lysine(50(cid:5)g/ml;Sigma)-coatedtis- Here, we demonstrate that cyclopamine modulates APP suecultureplates.Corticalneuronsweretreatedwithpharma- metabolism.Specifically,cyclopaminepreventsA(cid:3)generation cological agents on 6 days in vitro for 24 h, and lysates were by decreasing (cid:2)-secretase-mediated cleavage of APP-CTFs. collected for further biochemical analysis. All animals were ThisoccursduetoalterationinsubcellulartraffickingofAPP- used in accordance with animal protocols approved by the CTFs.Uponcyclopaminetreatment,APP-CTFsaccumulatein InstitutionalAnimalCareandUseCommittee(IACUCProto- lysosomesratherthanbeingtraffickedtotheTGN.Onceinthe colnumber19787).Animalsweredeliveredtoandmaintained lysosomes, APP-CTF degradation is decreased leading to its attheCalhounAnimalFacility(DrexelUniversity,PA).Animal accumulation.Together,ourdatademonstrateanoveluseof procedures were performed strictly in accordance with the cyclopamine and may open new avenues to treat Alzheimer NationalInstitutesofHealthGuideforthecareanduseofLab- disease. oratory Animals approved by the Drexel University Animal CareandUseCommittee. EXPERIMENTALPROCEDURES DrosophilaStocksandGenetics—Drosophilahusbandrywas Antibodies,Plasmids,andReagents—Antibodieswereobtained performedasdescribedpreviously(37).Forexperimentsutiliz- fromthefollowing:mouseC-terminalAPPclonec1/6.1(akind ingthe(cid:2)-secretasereporterGMR-APP-Gal4;UAS-Grim/Cyo gift from P. Mathews, Nathan Kline Institute, New York); model(38),flieswerecrossedonstandardcornmealagarfood mouse APP N-terminal 4A 22C11 (Millipore); rabbit APP supplemented with cyclopamine (100 nM) or DMSO vehicle C-terminalA8717,mouseanti-Mycclone9E10,rabbitLC3IIB, control(0.1%),andflieswerecollected24hpost-eclosion,and andmouse(cid:3)-actin(Sigma);mousecleavedNotch1cloneD3B8 theircompoundeyewasinspected.Assessmentofpenetrance and rabbit PSEN1 CTF clone D39D1 (Cell Signaling); mouse andseverityoftherough-eyephenotypewasaccomplishedby LAMP1, TGN38, and mouse EEA1 (BD Biosciences); rat photographing the compound eye using a Canon PowerShot LAMP1clone1D4BandmouseMP6Rclone22d4(DSHB);rab- S70 digital camera mounted to a Leica Mz 125 stereomicro- bitTGN46(AbDSerotec),andrabbitcathepsinD(kindgiftfrom scope. Severity of rough-eye phenotype was scored (cid:4) (mild) Dr. Stefan Höning). Fluorescent secondary antibodies (Alexa to(cid:4)(cid:4)(cid:4)(severe).One“(cid:4)”referstowherelessthan1⁄2oftheeye Fluor488and594)werefromJacksonImmunoResearchandIR- wasapoptoticandthereforeappears“rough”.Ascoreof“(cid:4)(cid:4)” conjugatedsecondaryantibodies(IRDye680andIRDye800)were (moderate) defined increased penetrance, where apoptosis fromLi-CorBiosciences.Peroxidase-conjugatedsecondaryanti- affectedapproximately1⁄2oftheeye.Severe“(cid:4)(cid:4)(cid:4)”rough-eye bodieswerefromCellSignaling.Cyclopamine(0.5–10(cid:5)M)was phenotype described when more than 1⁄2 of the eye appeared purchasedfromLCLaboratories,andL-685,458(2(cid:5)M),cyclohex- rough, and eye size was significantly reduced. For objective imide (50 (cid:5)g/ml), and DMSO were from Sigma. pCS2-Myc- quantification, five blinded laboratory personnel analyzed all (cid:2)ENotch was used for overexpression studies. pMst-APP-Gal4 experiments. was originally developed by Cao and Sudhof (36). Human Immunoblotting—Lysates were collected in complete RIPA Chmp2a-GFPconstructwasagiftfromDr.EliasSpiliotis. buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.5% sodium CellCultureandTransfection—HeLacellsweremaintained deoxycholate,0.1%SDS,1%NonidetP-40)supplementedwith at37°C,5%CO incompleteDMEM(CorningGlass)supple- HaltproteaseandphosphataseinhibitorandEDTA(Thermo- 2 mentedwith10%FBS(AtlantaBiologicals),100units/mlpen- Fisher).Lysateswerebrieflyclearedat20,000(cid:3)gat4°Cand icillin, and 100 (cid:5)g/ml streptomycin (Corning Glass), 2 mM stored at (cid:5)20°C. Protein concentrations were determined L-glutamine(CorningGlass).Cellsweregrownto80%conflu- usingtheBCAassaykitaccordingtothemanufacturer’sproto- ence and serum-starved (0.5% FBS/DMEM) for 24 h prior to col(Pierce).40(cid:5)goflysatewassupplementedwithNuPAGE pharmacologicalorgeneticmanipulation.Forpharmacological LDSsamplebuffer(Invitrogen)andheatedto75°Cfor10min. manipulation, drugs were diluted in 0.5% FBS/DMEM. For Proteinwasseparatedon4–12%NuPAGEBisTrisgels(Invit- geneticoverexpressionexperiments,cellsweregrownto80%con- rogen)usingMESrunningbuffer(Invitrogen)andthentrans- fluenceandthentransfectedusingwithTurboFecttransfection ferredontoImmobilonPVDFmembrane(Millipore).Odyssey reagent(ThermoScientific)accordingtothemanufacturer’spro- blocking buffer (Li-Cor Biosciences) was used for blocking tocol. Culture media were removed 24 h post-transfection, and and resuspending primary and secondary antibodies. Mem- cellswerecollectedorfurthertreatedwithpharmacologicalagents braneswerescannedusingLi-CorOdysseyinfraredscanning (0.5%FBS/DMEM)foranadditional24h. instrument. NOVEMBER28,2014•VOLUME289•NUMBER48 JOURNALOFBIOLOGICALCHEMISTRY 33259 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization A(cid:3)ELISA—HeLacellsandprimaryratcorticalneuronswere tocol. Briefly, samples were diluted 1:1 using kit diluent and treatedwithpharmacologicalagentsfor24h,andconditioned incubatedovernightat4°C.Sampleswerecomparedwiththe supernatants were collected and cleared at 20,000 (cid:3) g for 20 ELISAkitpositivecontrolandnegativecontrol(diluentalone). min at 4°C. Fresh cleared supernatants were used for A(cid:3) Samples were incubated and analyzed using a luminescence 40 ELISAkit(Wako,Japan)accordingtothemanufacturer’spro- platereader. 33260 JOURNALOFBIOLOGICALCHEMISTRY VOLUME289•NUMBER48•NOVEMBER28,2014 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization In Vitro (cid:2)-Secretase Assay—We utilized a well established Immunofluorescence—Cellswerefixedusing4%paraformal- cell-free (cid:2)-secretase activity assay that utilizes a fluorogenic dehyde,0.1%TritonX-100,blockedin2%BSAfor30min,and peptidesubstratecorrespondingtotheAPP(cid:2)-secretasecleav- incubatedwithprimaryantibodiesovernightat4°C.Cellswere age site (39, 40). HeLa cells grown to 100% confluence in rinsedwithPBSandstainedwithsecondaryantibodiesatroom 150-mmculturedisheswerecollectedinice-coldPBSandpel- temperaturefor1h,washedwithPBS,andmounted.Cellswere letedat5000rpmfor5min.Thepelletwashomogenizedin500 imagedusingOlympusFluoview1000invertedconfocalmicro- (cid:5)lofBufferB(20mMHEPES,pH7.5,150mMKCl,2mMEGTA, scope. Quantification of three-dimensional confocal image proteaseandphosphataseinhibitors)usinga27-gaugeneedle. stackswasaccomplishedusingSlideBook5.0orVolocityImage Theresultinghomogenatewasclearedat45,000rpmat4°Cfor analysissoftware(PerkinElmerLifeSciences). 1h.Supernatantwasstoredat(cid:5)80°C,andpelletwaswashed Surface Biotinylation—HeLa cells were treated with 5 (cid:5)M with500(cid:5)lBufferBandpassedthrough27gaugeneedleonice. cyclopamineorDMSOfor24h.Cellswereplacedonicetohalt Thesuspensionwasclearedagainat45,000rpmfor1hat4°C. membranedynamics,rinsedwithice-coldPBS,andincubated Supernatant was discarded and pellet resuspended in 75 (cid:5)l withsulfo-NHS-SS-biotin(1mg/mlinPBS;ThermoScientific) BufferB(cid:4)1%CHAPSOandpassedthrougha27-gaugeneedle for40minonicewithgentlerocking.Biotinwasquenchedwith onice.Theresultingmembranesamplesweresolubilizedona 100mMglycineinPBSfor15min.CellswerecollectedinPBS rotator at 4°C for 2 h. Solubilized samples were cleared at andpelletedat500(cid:3)gfor5minat4°C.Thepelletwaslysedin 45,000rpmfor1hat4°Cl;supernatant(totalcellmembrane) 200 (cid:5)l of standard RIPA lysis buffer containing protease and was collected and pellet discarded. Total protein was deter- phosphataseinhibitors.Lysatewasshearedusinga27-gaugenee- mined using BCA assay (Pierce) and 200 (cid:5)g of protein were dleoniceandsolubilizedfor2hat4°Conarotator.Lysatewas usedforinvitro(cid:2)-secretaseactivityassay.Briefly,membranes clearedbycentrifugationat10,000(cid:3)gfor5min,and50(cid:5)lfrom were resuspended in (cid:2)-secretase assay buffer (100 mM Tris- eachsamplewassetasidefor“total”proteinanalysis.Therestof HCl,pH6.8,4mMEDTA,0.5%CHAPSO),andpretreatedwith the supernatant was loaded into a capped spin column (Pierce; vehiclecontrol,L-685,458,orcyclopamine.Becausethemem- 69725) with NeutrAvidin-coated agarose resin (Thermo Scien- branepreparationenrichestotal(cid:2)-secretaseinthesample,the tific)ata1:1ratioandincubatedovernightat4°Conarotator. amount of pharmacological agent was increased accordingly. Columns were centrifuged at 10,000 rpm for 1 min, and flow- Therefore,20(cid:5)Mdruginatotalvehiclevolumeof1(cid:5)lperwas through(“unbound”control)wascollectedandsaved.Resinwas used.150(cid:5)loftotalvolumeperwellofa96-wellplatewasused. washedseveraltimeswithcompleteRIPA.Then50(cid:5)lofNuPAGE Membraneswerepretreatedfor3hat37°Cand5%CO ,and LDSSampleBuffer(Invitrogen)with5%(cid:3)-mercaptoethanolwas 2 thenfluorogenic(cid:2)-secretasesubstrate(Calbiochem,EMDMil- loadedintoeachcolumnandincubatedfor30minatroomtem- lipore)wasaddedtomembranesandfurtherincubatedat37°C peratureonashaker.Tocollectthesurface-biotinylatedprotein, and5%CO fortheindicatedtimepointsatwhichtimemem- columnswerecentrifugedatmaximumspeedfor2min.Biotiny- 2 braneswereremovedandfluorescencewasmeasuredusinga lated protein was separated on 4–12% NuPAGE BisTris gels plate reader (Promega). BSA was used as negative control in (Invitrogen)andthentransferred,andthemembranewasprobed placeofmembranes. forsurfaceAPP.Nonbiotinylatedlysateswerecollectedascontrol SubcellularFractionation—HeLacellsgrownto80%conflu- samples.Biotinylated“surface”sampleswerecomparedwithtotal encein100-mmculturedishesweretreatedwith5(cid:5)Mcyclo- lysatesamples. pamine or DMSO for 24 h, rinsed and collected in PBS, and Statistical Analysis—All graphs and diagrams represent thenclearedat1000rpmfor7min.Thecellpelletwasresus- mean values (cid:6) S.E. of all triplicates from at least three inde- pended in homogenization buffer (250 mM sucrose, 150 mM pendentexperiments.Eithertwo-tailedorone-tailedStudent’s NaCl, 25 mM Tris, 1 mM EDTA, protease and phosphatase ttestwasusedtocomparetwotreatmentgroupsandcalculate inhibitormixture)andhomogenizedusingball-bearing12-(cid:5)m significance from at least three independent experiments (*, clearancecellbuster.Homogenateswereclearedat1000(cid:3)gfor p(cid:7)0.05;**,p(cid:7)0.01;***,p(cid:7)0.005).ForinvivoDrosophila 15minat4°C,andpost-nuclearsupernatantwasloadedinto melanogasterexperiments,Gtest(goodnessoffit)wasusedto discontinuous density gradient (50, 30, and 10%) medium determine significance of phenotypic penetrance in experi- (Optiprep, Sigma) in Opti-Seal centrifuge tubes (Beckman). mentalpopulations.Degreeofsignificanceandcorresponding Homogenateswerespunat30,000rpmfor19hat4°Cand300 p value criteria for G test were identical to previously men- (cid:5)lfractionscollected. tionedStudent’sttest. FIGURE1.CyclopaminetreatmentaltersAPPmetabolism.A,primaryratcorticalneuronstreatedwith5(cid:5)Mcyclopamine(Cyc)for24h.EndogenousFL-APP andAPP-CTFsweredetectedusingWesternimmunoblottingandaC-terminalAPPantibody(c1/6.1).BandC,FL-APPproteinlevelsandAPP-CTFswere normalizedto(cid:3)-actinandFL-APP,respectively.D,HeLacellstreatedwiththeindicatedconcentrationsofcyclopaminefor24hfollowedbyFL-APPand APP-CTFdetectionincelllysatesviaWesternimmunoblotting.E,quantificationofdose-dependentaccumulationofAPP-CTFsnormalizedtoFL-APP.Relative proteinchangeswerecomparedwithvehicle(DMSO)control(E,dashedline).F,WesternblottimecourseanalysisofendogenousFL-APPandAPP-CTFsinHeLa cellstreatedwith5(cid:5)Mcyclopamine.G,normalizedAPP-CTFsincreaseinatime-dependentmannerascomparedwithvehicle(DMSO)control(G,dashedline). H,UsingWesternblotanalysisandN-terminalAPPantibody(22C11),endogenoussAPPlevelsweremonitoredinsupernatantsofHeLacellstreatedwith5(cid:5)M cyclopaminefor24h.I,sAPPlevelswerenormalizedtoproteinconcentrationsinlysates,determinedusingBCAassay.LackofchangeinsAPPproteinlevelsare illustratedincomparisonwithvehiclecontrol(DMSO).J,endogenousFL-APPandAPP-CTFsfromnaiveHeLacellstreatedwith5(cid:5)Mcyclopaminewere comparedwithcellstreatedwith2(cid:5)ML-685,458usingWesternimmunoblotanalysisandC-terminalAPPantibody(c1/6.1).Lowerpanelrepresentsincreased exposureofAPP-CTFs.KandL,cyclopamineincreasedAPP-CTFlevels(notFL-APP)comparedwithvehicle(DMSO)control.Valuesdenotemean(cid:6)S.E. Student’sttestwasusedforstatisticalanalysis:***,p(cid:7)0.005;**,p(cid:7)0.01;*,p(cid:7)0.05. NOVEMBER28,2014•VOLUME289•NUMBER48 JOURNALOFBIOLOGICALCHEMISTRY 33261 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization FIGURE2.Cyclopaminedecreases(cid:2)-secretasecleavageofAPPinvitroandinvivo.A,secretedA(cid:3) levelsfromHeLacellsandprimarycortical neurons(6daysinvitro)treatedwith5(cid:5)Mcyclopamine(Cyc)for24hweremeasuredusinganELISA4a0ndcomparedwithvehicle(DMSO)control. SecretedA(cid:3) levelswerenormalizedtototalproteininrespectivecelllysates.B,HeLacellstransientlyoverexpressingFL-APP-Gal4treatedwithvehicle (DMSO)con4tr0ol,2(cid:5)ML-685,458,or5(cid:5)Mcyclopaminefor24h.APP-CTF-Gal4andAICD-Gal4levelsanalyzedusingWesternimmunoblottinganda C-terminalAPPantibody(c1/6.1).C,APP-AICD-Gal4levelsnormalizedtoAPP-CTF-Gal4levels.D,HeLacellstransientlyoverexpressing(cid:2)ENotch-Myc treatedwithvehicle(DMSO)control,2(cid:5)ML-685,458,or5(cid:5)Mcyclopaminefor24h.NICDand(cid:2)ENotchlevelswereanalyzedusingWesternimmuno- blottingandanti-Mycandanti-NICD(Val-1744)antibody.E,NICDlevelsnormalizedto(cid:2)ENotchlevels.F,APP-CTF-Gal4and(cid:2)ENotchlevelsnormalized to(cid:3)-actin.G–J,representativeimagesofrough-eyephenotypefromGMR-APP-Gal4;UAS-Grim((cid:2)-secretasereporter)fliesraisedonnormalfood,100nM cyclopamine,orvehiclecontrol(DMSO).Flieswerescored1dayaftereclosionasfollows:mild((cid:4)),moderate((cid:4)(cid:4)),andsevere((cid:4)(cid:4)(cid:4)). K,relative changesinpenetranceofrough-eyephenotypeinfliesraisedonnormalfood(n(cid:8)168),vehicle-containingfood(n(cid:8)72),orcyclopamine-containing food(n(cid:8)112).Populationofflieswithmild,moderate,orsevererough-eyephenotypeisillustratedaspercentoftotalpopulationperexperimental group.Statisticalanalysisinvivoexperimentsareasfollows:Gtest,***,p(cid:7)0.005;**,p(cid:7)0.01,wasperformed.Student’sttestwasusedforstatistical analysisofcellbasedinvitrostudies.Valuesdenotemeans(cid:6)S.E.n.s.,notsignificant. RESULTS changeinthefull-lengthAPP(FL-APP)holoproteinafter24h CyclopamineTreatmentResultsinAPPC-terminalFragment of5(cid:5)Mcyclopaminetreatment(Fig.1,AandB).However,the Accumulation—TotestwhethercyclopaminemodulatesAPP 8–12-kDaAPPproductsof(cid:4)-and(cid:3)-secretase((cid:4)-and(cid:3)-CTFs; metabolism, we treated primary rat cortical neurons with collectivelyknownasAPP-CTFs)significantlyincreasedwhen cyclopamine (41, 42). We did not observe an appreciable compared with vehicle control-treated neurons (p (cid:8) 0.0190) 33262 JOURNALOFBIOLOGICALCHEMISTRY VOLUME289•NUMBER48•NOVEMBER28,2014 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization (Fig. 1, A and C). To determine whether these effects can be observedinothermodels,weutilizedHeLacellsbecausethey are easily manipulated and have been previously utilized to studyAPPprocessingandtrafficking(19,43).UsingnaiveHeLa cells, we performed cyclopamine time and dose dependence experiments.Cellsweretreatedfor24hwithincreasingcon- centrationsofcyclopaminefrom0.5to10(cid:5)M(Fig.1D).Com- paredwithvehiclecontrol(Fig.1E,dashedline),weobserveda significantincreaseinAPP-CTFlevelswithaslittleas0.5(cid:5)M cyclopamine (p (cid:8) 0.000615) (Fig. 1E). No change in FL-APP wasobservedincellsexposedto0.5,1,and5(cid:5)Mofdrug(Fig. 1E).Asmall,yetsignificantincreaseinFL-APPwasobserved upon10(cid:5)Mcyclopaminetreatment.Toaddresstimedepen- dence of cyclopamine’s effects on APP-CTF accumulation, we performed a time course experiment. Because we observed robust increases in APP-CTFs after 24 h with as littleas0.5(cid:5)Mdrug,wehypothesizedthatusing5(cid:5)Mcyclo- paminewouldsignificantlyincreaseAPP-CTFlevelswithin a shorter exposure time. Accumulation of APP-CTFs was evidentby3hofexposure(p (cid:8)0.000488),andfurtheraccu- mulationcontinuedfortheremainderofthetimecourse(by 24hp(cid:8)9.50(cid:3)10(cid:5)6)(Fig.1,FandG).Thelackofsignificant changesinFL-APPlevelsupon5(cid:5)Mcyclopamineexposure suggestsAPPgenetranscriptionisnotaltered.Infact,using quantitative PCR, we analyzed APP mRNA in naive HeLa cells upon cyclopamine treatment and did not observe changes in APP transcript levels as compared with vehicle control(datanotshown). APPproteolysisisinitiatedby(cid:4)-or(cid:3)-secretase.Thiscleav- age liberates soluble N-terminal APP ectodomains (sAPP). TreatmentofnaiveHeLacellswithcyclopaminedidnotalter FIGURE3.(cid:2)-Secretaseactivityisnotalteredbycyclopaminetreatment.A, sAPPlevels(Fig.1,HandI).ThissuggeststheincreaseinAPP- HeLacellstreatedwith5(cid:5)Mcyclopamine(Cyc)orvehiclecontrol(DMSO)for 24h.PSEN1-CTFlevelswerefromrespectivelysatesanalyzedviaWestern CTFsisnotduetomodulationof(cid:4)-or(cid:3)-secretasecleavageof immunoblotting.B,PSEN1-CTFslevelsnormalizedto(cid:3)-actin.C,fluorometric APPbycyclopamine.TheobservedincreaseinAPP-CTFsand (cid:2)-secretaseactivityassay.Fluorescenceintensityovertimeusingtotalmem- branes isolated from naive HeLa cells treated with cyclopamine (20 (cid:5)M), the lack of change in FL-APP levels resemble the effects of L-685,458 (20 (cid:5)M), or vehicle control (DMSO). Graph represents relative (cid:2)-secretaseinhibitorsbuttoadiminisheddegree(Fig.1,J–L) changesinfluorescenceaspercentactivityofcontrol(membranestreated withDMSO)overtime. (44,45). Cyclopamine Decreases (cid:2)-Secretase-mediated Cleavage of weremuchmoremodestthanthoseobservedupon(cid:2)-secretase APP in Vitro and in Vivo—Because cyclopamine treatment inhibitionwithL-685,458. increasedlevelsofAPP-CTFs,analogouslyto(cid:2)-secretaseinhib- To determine whether the observed effects were specific itor treatment, we hypothesized that cyclopamine would to APP, we monitored (cid:2)-secretase cleavage of Notch in decrease levels of (cid:2)-secretase cleavage products, namely A(cid:3) responsetocyclopaminetreatment.SimilartoAICD,endog- andtheAICD.BothA(cid:3)andAICDareproducedupon(cid:2)-secre- enousNICDisalsodifficulttodetect.Toovercomethisdif- tasecleavageofAPP-CTFs.Totestthishypothesis,weexposed ficulty,wetransientlyoverexpressedMyc-(cid:2)ENotchinHeLa naive HeLa and primary rat cortical neuron cells to cyclo- cellsandtreatedwithcyclopaminefor24h(47,48).Cyclo- pamine for 24 h. Cyclopamine-treated cells secreted signifi- pamine significantly decreased NICD levels (p (cid:8) 8.87 (cid:3) cantlylessA(cid:3)comparedwithvehiclecontrolinprimarycorti- 10(cid:5)5)toasimilarextentasobservedinA(cid:3)andAICDlevels calneuronsandHeLacells(p(cid:8)0.00567)(Fig.2A).Theother (Fig. 2, D and E). These effects on AICD and NICD were product of (cid:2)-secretase cleavage, AICD, is difficult to detect. much more modest than those observed upon (cid:2)-secretase Therefore,weusedapreviouslydescribedAPP-Gal4construct inhibition. Similar to endogenous APP-CTFs, cyclopamine toaidindetection(36,46).WeexposedHeLacellstransiently increased APP-CTF-Gal4 levels (Fig. 2F). Interestingly, no overexpressingAPP-Gal4tocyclopaminefor24handdetected change in (cid:2)ENotch levels was observed suggesting that the AICD-Gal4levelsusingWesternblotanalysis.The(cid:2)-secretase effectscouldbespecific(Fig.2F). inhibitorL-685,458servedasapositivecontrolbecauseitpre- Giventheseresultsandtheavailabilityofaninvivo(cid:2)-secre- ventsAICDgeneration.Aspredicted,incomparisonwithvehi- tasereporter,wetestedtheabilityofcyclopaminetomodulate cle control, cyclopamine significantly decreased AICD-Gal4 (cid:2)-secretasecleavageofAPPinvivo.Briefly,in2003,Guoetal. levels(p(cid:8)0.000228)(Fig.2,BandC).However,theseeffects (38)developedandcharacterizedaD.melanogaster(cid:2)-secretase NOVEMBER28,2014•VOLUME289•NUMBER48 JOURNALOFBIOLOGICALCHEMISTRY 33263 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization FIGURE4.CyclopamineinducessubcellularaccumulationofAPP-CTF.Confocalthree-dimensionalanalysisofHeLacellstreatedwithvehiclecontrol (DMSO)(A)or5(cid:5)Mcyclopamine(B)for0,6,or24h.APPwasdetectedusinganantibodyagainsttheCterminusofAPP(A8717).Scalebar,10(cid:5)m.C,numberof APPpunctanormalizedtoregionsofinterest(ROI).Regionsofinterest(cid:8)10(cid:3)10(cid:5)m,and20–30regionsofinterestpertreatmentwereanalyzed.Punctawere definedas2(cid:3)theintensityofbackgroundinthecytosolicnonpunctateregion,andobjectswererestrictedto0.2–2.0(cid:5)mskeletaldiameter.Cyc,cyclopamine. Confocalthree-dimensionalanalysisofnaiveHeLacellstreatedwithvehiclecontrol(DMSO)(D)or5(cid:5)Mcyclopamine(E)for24h.Cellswerestainedusing C-terminalAPP(A8717)(leftpanel)andN-terminalAPP(22C11)(middlepanel)antibodies.Rightpaneldenotesmergedchannels.Scalebar,10(cid:5)m.F,cellsurface FL-APPlevelsinHeLacellstreatedwith5(cid:5)Mcyclopamineorvehicle(DMSO)for24handthenbiotinylated.BiotinylatedsurfaceFL-APPlevelsweremeasured byWesternblotwithaC-terminalAPPantibody(c1/6.1).ToppanelistotalFL-APPfromwholecelllysate,middlepanelisbiotinylatedsurfaceFL-APPpurified withNeutrAvidin-coatedresin.G,surfaceFL-APPlevelsnormalizedtototalFL-APPaspercentofvehicle(DMSO)control.Valuesdenotemeans(cid:6)S.E.Student’s ttestwasusedforstatisticalanalysis;**,p(cid:7)0.01. reporter. These transgenic flies express the APP (cid:2)-secretase cleavage of APP-C99-Gal4, we raised APP-C99-Gal4;UAS- substrate,APP-C99-Gal4,specificallyintheflyeyeommatidia. GRIMfliesonnormal,vehicle,orcyclopaminesupplemented ThesefliesalsocarryaUASelementupstreamofGRIM,acell food. Flies were collected 1 day post-eclosion, and their eyes deathactivator.Upon(cid:2)-secretasecleavageofAPP-C99-Gal4, were scored for rough-eye phenotype. Flies raised on cyclo- the resulting AICD-Gal4 can bind to the UAS element and paminedisplayeddecreasedseverityoftherough-eyepheno- induce GRIM expression. GRIM expression leads to death of type(p(cid:8)2.40(cid:3)10(cid:5)34)(Fig.2,G–K).Morespecifically,10%of ommatidiaandresultsinarough-eyephenotype(38,49,50).To thefliesraisedoncyclopaminedisplayedsevererough-eyephe- test whether cyclopamine decreases (cid:2)-secretase-mediated notype compared with the 47% raised on vehicle food. 33264 JOURNALOFBIOLOGICALCHEMISTRY VOLUME289•NUMBER48•NOVEMBER28,2014 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization FIGURE5.EndogenousAPP-CTFdistributioninvehicle-treatedHeLacells.Confocalthree-dimensionalanalysisofHeLacellstreatedwithvehiclecontrol (DMSO)for24h.CellswerestainedforAPP-CTFsusinganAPPC-terminalantibody(A8717)andantibodies(EEA1,MP6R,LAMP1,TGN46,andLC3)for subcellularmarkers(middlecolumn).Right-handcolumnsarethemergedimages.Scalebar,10(cid:5)m. Althoughonly10%ofthevehicle-treatedfliesdisplayed“mild” CyclopamineDoesNotAlter(cid:2)-SecretaseActivity—Becausein rough-eye phenotypes, 53% of cyclopamine-treated flies dis- vivo and in vitro cyclopamine treatment leads to decreased playedthisphenotype(Fig.2K).Together,thesedatademon- (cid:2)-secretase cleavage of APP-CTFs, we investigated whether stratethatcyclopaminetreatmentdecreases(cid:2)-secretase-medi- cyclopamine inhibits (cid:2)-secretase activity. One major step in atedcleavageofAPP-CTFsinvitroandinvivo. (cid:2)-secretasecomplexmaturationistheautoproteolysisofpreseni- NOVEMBER28,2014•VOLUME289•NUMBER48 JOURNALOFBIOLOGICALCHEMISTRY 33265 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization FIGURE6.CyclopaminealtersretrogradetraffickingofAPP-CTFs.Confocalthree-dimensionalanalysisofHeLacellstreatedwith5(cid:5)Mcyclopaminefor24h. CellswerestainedforAPP-CTFsusinganAPPC-terminalantibody(A8717)andantibodies(EEA1,MP6R,LAMP1,TGN46,andLC3)forsubcellularmarkers(middle column).Right-handcolumnsarethemergedimages.Scalebar,10(cid:5)m. lin1 (PSEN1) to form the active N- and C-terminal fragments. Toassessoverall(cid:2)-secretaseactivity,weutilizedaninvitro, Therefore,detectionofthePSEN1-CTFisanindicatorofanactive fluorescence-basedactivityassay(40).Weisolatedtotalcellular (cid:2)-secretasecomplex(11).Tothisend,weexposednaiveHeLacells membranes from naive HeLa cells and treated these mem- tocyclopaminefor24handobservedanincreaseinAPP-CTFs branes with vehicle, L-685,458, or cyclopamine (51–54). As levels;however,PSEN1-CTFlevelsdidnotchangeinresponseto expected, treatment with L-685,458 decreased cleavage of cyclopaminetreatment(Fig.3,AandB). the fluorogenic (cid:2)-secretase peptide substrate resulting in 33266 JOURNALOFBIOLOGICALCHEMISTRY VOLUME289•NUMBER48•NOVEMBER28,2014 ReductionofA(cid:3)byAlteringAPPSubcellularLocalization FIGURE7.QuantificationofsubcellularmarkertotalintensityandAPP-CTFsubcellularlocalizationinHeLacells.A–E,singlecellsineachimagewere maskedoff,backgroundintensitiessubtractedfromeachchannel,andsumintensityofeachcellmeasuredandnormalizedtovolume.Also,Manders’ coefficientswerecalculatedforeachcellindependently,and26–51cellswereanalyzedperexperimentalgroup.F–J,Manders’coefficientsforcolocalization. Dotplotdiagramsrepresentrawdatapoints,andthehorizontallinerepresentsthemeansofManders’coefficientsofsubcellularmarkerandAPP-CTF colocalization.Cyc,cyclopamine.Student’sttestwasusedforstatisticalanalysis:***,p(cid:7)0.005;n.s.,notsignificant. decreased fluorescence intensity over time. Surprisingly, CTFsbyWesternblot(Fig.1G),herecyclopaminetreatment treatment with cyclopamine did not alter (cid:2)-secretase activity induced significant accumulation of APP-positive puncta (Fig. 3C). These results suggest that cyclopamine decreases detectedwithanantibodyraisedtotheAPPCterminus(p(cid:8) (cid:2)-secretase-mediatedcleavageofAPPwithoutdirectlyaffect- 0.00120) (Fig. 4, A–C). Visualization of APP distribution ing (cid:2)-secretase activity. One mechanism that could explain usinganantibodyspecifictotheN-terminalportionofAPP theseresultsisthatcyclopaminemediatesachangeinthesub- didnotrevealsimilarcyclopamine-inducedAPPpuncta.In cellularlocalizationofAPPand/or(cid:2)-secretase. fact, a lack of colocalization was observed between the N- Cyclopamine Alters APP-CTF Subcellular Localization— andC-terminalAPPantibodiesinthecyclopamine-induced Proteolytic processing of APP is dependent on its subcellular APP puncta (Fig. 4, D and E). This suggests that the cyclo- localization. To investigate whether cyclopamine alters APP pamine-inducedpunctaareAPP-CTFsandnotFL-APPnor subcellularlocalization,naiveHeLacellswereexposedtocyclo- sAPP. FL-APP is not a suitable substrate for (cid:2)-secretase paminefor0,6,or24h,andAPPsubcellulardistributionwas cleavage. The increase in APP-CTF subcellular puncta and visualizedusingimmunofluorescence.Analogoustothetime the lack of change in FL-APP and sAPP protein levels sug- course experiment in which we observed increased APP- geststhatcyclopaminedoesnotalterAPPbiosyntheticpath- NOVEMBER28,2014•VOLUME289•NUMBER48 JOURNALOFBIOLOGICALCHEMISTRY 33267