Doublecortin-like kinase and neuronal differentiation Dijkmans, T.F.

Hele tekst

(1)Doublecortin-like kinase and neuronal differentiation Dijkmans, T.F.. Citation Dijkmans, T. F. (2009, October 14). Doublecortin-like kinase and neuronal differentiation. Retrieved from https://hdl.handle.net/1887/14055 Version:. Corrected Publisher’s Version. License:. Licence agreement concerning inclusion of doctoral thesis in the Institutional Repository of the University of Leiden. Downloaded from:. https://hdl.handle.net/1887/14055. Note: To cite this publication please use the final published version (if applicable)..

(2) Chapter4 NGFINDUCED NEURITOGENESISIS REGULATEDBY PHOSPHORYLATIONOFDCLK SHORTINNS1PC12CELLS DijkmansTF,VanHooijdonkLW,KamphorstJT,ZhangYP,SchoutenTG,Fitzsimons CP,VreugdenhilE. DivisionofMedicalPharmacology,Leiden/AmsterdamCenterforDrugResearch andLeidenUniversityMedicalCenter,TheNetherlands SubmittedatJournalofNeurochemistry.

(3) 92. Chapter4. ABSTRACT Nerve Growth Factor (NGF)induced differentiation of Ns1 PC12 cells involves complextranscriptional,translationalandposttranslationalregulation.Previously, we identified DCLKshort as a transcript that is strongly upregulated by NGF stimulation.WhiletheDCLKgeneisimplicatedinneuronalmigration,neurogenesis and neuronal plasticity, the function of DCLKshort is incompletely understood. Here, we report that DCLKshort is involved in NGF signaling and regulates neuritogenesis in Ns1 PC12 cells. We show that activation of Extracellularsignal regulated kinase 1/2 by NGF leads to specific phosphorylation of DCLKshort at serine 30. Phosphorylation of DCLKshort at serine 30 is important for neuronal differentiation, because overexpressed wild type and serine 30 phosphomutants had different effects on neuritogenesis and GAP43 induction. Interestingly, intracellular distribution of DCLKshort was dependent on an intact actin cytoskeleton. Moreover, DCLKshort colocalized with filamentous actin in growth cones. We show that DCLKshort is a phosphorylationdependent regulator of neuritogenesis,possiblyviainteractionwiththeactincytoskeleton. .

(4) Chapter4. INTRODUCTION The rat pheochromocytoma cell line PC12 is a widely used cell model to study neuronal differentiation. In response to Nerve Growth Factor (NGF) stimulation, PC12cellsexitcellcycle,developneuritesthatareelectricallyexcitableandhave the potential to form functional synapses with cocultured muscle cells [1]. On a molecular level, extracellular presentation of NGF to the cell leads to binding of NGFtothetransmembraneTropomyosinrelatedkinasereceptorA(TrkA;[2]).Asa consequence, TrkA produces an intracellular signal that activates different signaling cascades, including the Phosphatidylinositol 3Kinase, Phospholipase C and the RafRasMEKERK 1/2 pathway [3;4]. Within minutes, the earliest gene transcriptsareproducedandarefollowedbytranscriptionalregulationofdelayed responsegenes[5;6;6].Previously,weperformedtwoDNAmicroarraystudiesthat characterized the transcriptional response after up to two hours and up to four days of NGF stimulation [7;8]. Among the transcripts that were progressively induced by NGF stimulation was DCLKshort, suggesting a role for DCLKshort in neuronaldifferentiationofNs1PC12cells. DCLKshortisproducedbytheDoublecortinlikekinase1(DCLK1orDCLK) gene,amemberoftheDoublecortin(DCX)genefamily[9;10].BothDCXandDCLK are believed to be essential for neuronal migration, neurogenesis and neuronal plasticity [1114]. How DCX and DCLK function on a molecular level is mostly understood from their ability to bind microtubules through DCX domains [15]. Regulationofthisandotherproteininteractionsoccurinpartthroughtheserine, threonine and proline (SP)rich domain, including phosphorylationdependent crosslinkingofthemicrotubuleactincytoskeletonandmicrotubulebasedtransport [1618]. The DCLK gene, however, produces several proteins and DCLKshort, in contrasttootherDCLKproteins,lacksthemicrotubulebindingDCXdomain. DCLKshort rather consists of the Nterminal SPrich domain and a C terminalkinasedomain[9].Itsfunctionsremainincompletelyunderstoodandare inferredfromoverexpressionstudies.DCLKshortkinaseactivitymaybeaffectedby cAMPraising agents, although this observation failed to be reproduced [19;20]. Activated DCLKkinase, particularly by truncation of its autoinhibitory Cterminal domain, inhibits CREBmediated gene transcription in a TORC2dependent way [19;21]. Additionally, we and others have shown that the Nterminal, SPrich domain of DCLKshort is subject to cleavage by calpains and caspases during neuronal apoptosis [22;23]. DCLKshort appears to play neuroprotective roles, as an uncleavable DCLKmutant protected against neuronal apoptosis, and an N terminal cleavage fragment exacerbated apoptosis. Therefore, two important questionsthatneedfurtherelucidationare1)inwhichothercellularprocessesis DCLKshortinvolved?2)inwhichsignalingcascadesdoesDCLKshortfunction? Here,wereportthatDCLKshortisanendogenoustargetofNGFsignaling andmediatesNGFinducedneuritogenesisinNs1PC12cells.NGFstimulationleads to ERK 1/2dependent phosphorylation of DCLKshort at serine 30 (Ser30), which lies in its SPrich domain. Moreover, we show that Ser30 phosphorylation is. 93.

(5) 94. Chapter4. important for neuronal outgrowth. DCLKshort protein shows marked localization to growth cones of Ns1 PC12 cells. Within in the growth cone, DCLKshort colocalizes with filamentous actin. Together, we conclude that DCLKshort is a phosphorylationdependent regulator of neuritogenesis, possibly via interaction withtheactincytoskeleton. MATERIALSANDMETHODS Cellcultureandtransfection Neuroscreen1PC12cells(Ns1PC12;Cellomics,Pittsburgh,PA,USA)werecultured at37°Cat5%CO2inRPMI1640,supplementedwith5%FoetalBovineSerum,10% Horse Serum, penicillin (20 U/mL) and streptomycin (20 g/mL all Invitrogen, Carlsbad, CA, USA). For cell stimulation, 500 000 Ns1 PC12 cells per well were seededinpolyLlysinecoated(SigmaAldrich)6welldishes,glasscoverslipswere inserted for microscopy purposes. The following day, cells were placed on RPMI 1640with0,5%FoetalBovineSerum,1,0%HorseSerum,penicillin(20U/mL)and streptomycin (20 g /mL) and incubated overnight. The next day, the cells were stimulatedwith50ng/mlratNerveGrowthFactor(NGF;SigmaAldrich,StLouis, MO, USA) and resupplied every secondday. Transfection of Ns1 PC12cells were performed using the Amaxa Nucleofector (Amaxa Biosystems, Gaithersburg, MD, USA) with transfection kit Cell Line Nucleofector Kit V (Amaxa Biosystems), accordingtoprotocolwith2,5gplasmidper2millioncells.Foractincytoskeleton disruption,Ns1PC12cellswereincubatedfor3hourswith10μMCytochalasinD (SigmaAldrich)10μM. GenerationofpSer30DCLKantibody TodeterminethespatialandtemporalphosphorylationstateofDCLKshortSer30 in vivo, rabbit polyclonal pSer30DCLK was custom produced by using an intradermic multiimmunization protocol (Eurogentec Nederland, Maastricht, The Netherlands).Briefly,2rabbitswereimmunizedwiththepeptidePSPTpSPGSLRKQ, correspondingtopositions2438ofratDCLKprotein(AAC99476),wherepdenotes aninvitrophosphorylatedresidue.TheidentityofthepeptidewasverifiedbyHPLC and mass spectrometry. Thereafter, the peptide was coupled to keyhole limpet hemocyanin via the SHgroup of the terminal cystein using the m MaleimidobenzoylNhydroxysuccinimideestermethod,toprovidemolecularmass andincreasedantigenicity[24].Animalswereimmunizedwith5mgofthepeptide carrierconjugate/animalina3monthlongimmunizationprotocol(4injections,4 bleedings, 1 final bleed). Immune serum was affinity purified against a non phosphorylated peptide (PSPTSPGSLRKQ) to eliminate nonphosphospecific antibodies.ThespecificityofthefinalantiserumwastestedusingaspecificELISA testagainsttheoriginalphosphorylatedpeptidePSPTpSPGSLRKQ.Specificityofthe generatedantibodywasadditionallydeterminedasfollows.Pointmutationswere introduced into mouse wild type DCLK DNA sequence in pcDNA3.1+ (Invitrogen).

(6) Chapter4. vectorusingQuikChangeTMSitedirectedMutagenesisKit(Stratagene,LaJolla,CA, USA)accordingtothemanufacturer’sinstruction(Table1forprimers).Inone,we disruptedthepotentialserinephosphorylationsitebymutationofserinetoalanine (SA) which is assumed to mimic a nonphosphorylated state. In the other, conversion of serine to aspartate (SD) introduces a constitutive negative charged carboxylandthusmimicsthephosphorylatedstateofthisserine[25].Subsequent coincubationwithorwithoutrecombinantERK1/2alloweddeterminingspecificity ofthedevelopedantibody(pSer30DCLK). FractionationassaysandWesternblot Protein fractions were prepared using Fractionation Kit (Pierce Inc., Rockford, Ill, USA).Forwesternblot,totalproteinwasseparatedbyelectrophoresisonasodium dodecylsulphate–polyacrylamide gel (12%) and semidry electroblotted to a polyvinylidenedifluoridemembrane,ImmobilonP(MilliporeCorporation,Bedford, MA, USA). Blots were blocked with blocking buffer (Trisbuffered saline, 0.2% Tween, 10% milk) and incubated with primary antibodies for 1 hour at room temperature in blocking buffer. Primary antibodies were used against tubulin (clone DM1A; Sigma), Actin (C2; Santa Cruz), Histon 10, (Clone 27, Abcam, Cambridge, UK ) (ERK 1/2 (AntiERK 1/2 pAb, Promega, Madison, WI, USA), pERK (AntiactiveMAPK,Promega)DCLKshortA(CPG16/CaMkinaseVI,BDBiosciences, Palo Alto, CA, USA), DCLK (antiDCAMKL, Abnova Corporation, Taipei, Taiwan), pSer30DCLK (our lab, see results above for details, rabbit polyclonal). Three rinsings with washing buffer (Trisbuffered saline, 0.2% Tween) were performed. Then, blots were incubated with horseradish peroxidaseconjugated secondary antibodies(SantaCruzBiotechnology,SantaCruz,CA,USA;1:5000)for30minutes. Finally,blotswererinsedthreetimeswithwashingbufferandantibodycomplexes werevisualizedbyluminoldetectionusingautoradiographyfilms(Kodak).Accurate fractionation was controlled by detection of actin protein (cytoplasmic fraction) andhistoneprotein(nuclearfraction).Equalproteinloadingwasverifiedbyactin detectionandopticaldensitieswerequantifiedusingImageJ. RealtimeqPCR TotalRNAwasextractedusingTrizolreagent(Invitrogen).Purityandconcentration of isolated total RNA was measured by the ND1000 spectrophotometer (NanoDropTechnology,Wilmington,DE,USA).TotalRNAwastreatedwithDNAseI (Invitrogen),afterwhich1μgwasusedastemplateforcDNAsynthesis,usingthe iScriptcDNAsynthesiskit((Biorad,Hercules,CA,USA)bothaccordingtoprotocols providedbytherespectivemanufacturers.Foreachprimerpair,PCRefficiencywas determinedusingthestandardcurvemethod,usingtotalRNAfromexperimental samples as cDNA template. Primer sequences of DCLKshort, GAP43, Egr1, cFos andnormalizationgenesbetaactinandGAPDHwereaspreviouslydescribed[7;8]. Preparation of PCR reactions were performed with the LightCycler FastStart DNA Master PLUS SYBR Green I (Roche, Indianapolis, IN, USA) reagents, according to. 95.

(7) 96. Chapter4. manufacturer’sinstructions.Finalprimerconcentrationwas5Mperprimerand input cDNA per reaction was synthesized from 25 ng total RNA.All realtime PCR reactionswereperformedonaLightCycler2.0(Roche),withthefollowingthermal cycling parameters: annealing at 64 degrees for 10 seconds, amplification at 72 degrees for 10 seconds and dissociation at 95 degrees for 10 seconds. The PCR programwasfollowedbyameltingcurveinwhichthetemperaturerosefrom65 degrees to 95 degrees, with continuous SYBR Green emission measurement. Specificity of amplification of eachprimerpair was controlled byBLAST search of primer pair sequences against the rat genome, determination of actual PCR productsizebygelelectrophoresisandcomparisonwithpredictedproductsize. Immunocytochemistryandmicroscopy Forimmunocytochemistry,6wellplatescontainingglasscoverslipswithadherent cells were washed in phosphatebuffered saline (PBS; Invitrogen) and fixed with 80% acetonefor 5 minutes.After fixation,coverslips were washed three times in phosphatebuffered saline with 0.2% Tween (PBST) and blocked 60 min in 5% normal goat serum at room temperature. Incubations with antiDCLK (Abnova; mouse monoclonal, 1:400) and antipSer30DCLK (our lab; 1:400) primary antibodies were performed for 1 hour at room temperature and subsequently washedthreetimesinPBST.DetectionofDCLKorpSer30DCLKcellswasachieved with Alexa 488 or Alexa 596 conjugated antirabbit/mouse IgG (Santa Cruz biotechnologies;1:200).Forfilamentousactinvisualization,rhodaminephalloidin conjugate was coincubated with secondary antibodies (Invitrogen, 1:2000). After incubation, sections were washed in 1× PBST, counterstained for 10 min with Hoechst33528(SigmaAldrich)andwashedthreetimes(5minutes)inPBST.When no visualization other than EGFP, Factin (rhodaminephalloidin) and nucleus (Hoechst 33528) was required, primary antibody incubation and subsequent washingwereomitted,afterwhichsecondaryantibodyincubationwasreplacedby rhodaminephalloidin only incubation. All coverslips were embedded in polyaquamount (Polysciences, Inc., Warrington, PA, USA) and observed with an immunofluorescence microscope (Leica DM6000) or confocal laser scanning microscope for colocalization data (BioRad Radiance 2100 MP). Confocal image acquisition was performed with a Kalman filter (n=2) and processed in ImageJ to despeckleandsharpenandtomergepicturesusingthecolormergeplugin(Wayne Rasband,NationalInstituteofMentalHealth,USA). Imageanalysis Image analysis was performed using ImageJ software (Wayne Rasband, National InstituteofMentalHealth,USA).Inordertoquantifyneuritogenesis,fixatedcells were micrographed by fluorescence microscopy on Leica DM6000 (Leica Microsystems AG, Wetzlar, Germany) at 10x magnification. Quantification of the numberofneuriteswasperformedaspreviouslydescribed[26].Briefly,aneurite wasdefinedasanextensionfromthecellbodyequivalentorgreaterthan1xthe cell body diameter. From 5 micrographs per experimental group and using 3.

(8) Chapter4. experimental replicates (15 images with ~40 cells per image), cells were counted togetherwiththenumberofneuritebearingcells.Theratiobetweennuclearand cytoplasmiclocalizationwasquantifiedbyusingthefluorescenceintegratedoptical intensity [27] derived from (phospho)DCLK antibody and fluorescence derived fromHoechststaining.Atleast9imagesperexperimentalgroupatamagnification of20xweretaken.Aftercapturingimagesofcellsusingappropriatefilters,images were converted to 8bit grayscale images. Thresholds were set to remove background signal and retain specific cellular signal. Integrated optical density of (phospho)DCLKwasquantifiedwithintheareaasdefinedbyHoechst33528signal (nucleus) andoutside this area (cytoplasmic). The ratio between theserespective opticaldensitiesproducedthenuclear/cytoplasmiclocalizationratio. Plasmidconstruction InordertomanipulateDCLKshortinthiscellmodel,severalattemptsweremade toestablishaknockdownofDCLKshortbyshRNAvectorsandsyntheticantisense RNA.AlthoughmRNAlevelscouldbereducedbyseveraldifferentshRNAs,protein levels persisted and directed us in an alternative approach of overexpression. A fusion of DCLKshort A coding DNA sequence (NM_053343; [9]) and Cterminal FLAG sequence (coding for the amino acid sequence DYKDDDDL) was cloned into pcDNA 3.1(+) (Invitrogen), using EcoRI and BamH1 restriction sites. The encoded serinewithinthePTSPmotifsequence(28PTSP31,accessionnumberAAC99476) was mutated by means of QuickChange Sitedirected Mutagenesis (Stratagene) accordingtomanufacturer’sinstruction.ToyieldaserinetoAlaninemutant(S30A DCLK)theforwardandreverseprimersequenceswere5’CAAGTCCATCACCCA CCGCCCCAGGAAGCCTGCG3’ and 5’CGCAGGCTTCCTGGGTCGGTGGGT GATGGACTTG3’.Toyieldaserinetoaspartatemutant(S30DDCLK),theforward primer 5’CAAGTCCATCACCCACCGACCCAGGAAGCCTGCG3’ and reverse primer5’CGCAGGCTTCCTGGGTCGGTGGGTGATGGACTTG3wereused.For microscopy purposes, wild type and mutant DCLKshort sequences excluding the FLAGtag were cloned into pEGFPN1 (ClonTech Laboratories, Palo Alto, CA, USA) using Hind III and Sac II sites. Sequence integrity was verified by DNA sequence analysison3100AvantGeneticAnalyzer(AppliedBiosystems,FosterCity,CA,USA) using the BigDye® Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems), accordingtomanufacturer’sinstructions. RESULTS DCLKshort is upregulated during NGFinduced neuronal differentiation of Ns1 PC12cells Previously, we reported DCLKshort as an NGFresponsive mRNA transcript according to both microarray data and qPCR validation [7]. In order to further validate this observation we performed additional experiments with Ns1 PC12. 97.

(9) Chapter4. A. No NGF. 4 days NGF. B 6. Relative expression. 98. Synapsin DCLK-short. 5 4 3 2 1 0 0. 1. 2. 3. 4. T (days of NGF stimulation). C DCLK-short (53 kD) Actin (42 kD) 0. + 1. -. + 2. -. + 3. -. + 4. NGF T (days). Figure 1. NGF induces DCLK-short upregulation during neuronal differentiation. A). Ns-1 PC12 cells undergo morphological differentiation by extending neurites after NGF stimulation. Staining against filamentous actin was performed using by rhodamine-phalloidin. B). NGF induces biochemical differentiation of Ns-1 PC12 cells as quantified by relative expression levels of Synapsin I mRNA (qPCR). Relative DCLK-short mRNA expression levels rise in response to NGF treatment. Error bars represent SEM derived from two replicate culture dishes. C). DCLKshort protein (53 kD) expression is increased after NGF treatment. The beta-actin protein levels serve as loading control.. cells stimulated with NGF. As can be seen in Fig. 1A, NGF treatment caused Ns1 PC12 cells to adopt a neuronal phenotype, characterized by extensive neuritogenesis after 4 days of NGF treatment. Quantitative PCR for Synapsin I, a marker of neuronal differentiation [28], was upregulated 4.3±0.4fold (mean ± SEM,n=2)afterNGFtreatmentandfurtherillustratestheacquisitionofaneuronal identity (Fig. 1B). Concomitant with Synapsin I upregulation was a 4.6±0.0fold increaseofDCLKshortmRNAexpressionbyNGFstimulation,whichconfirmedour previousobservations.Finally,wetestedwhethertheriseinDCLKshortmRNAwas followed on the protein level, which was indeed the case. Fig. 1C shows a.

(10) Chapter4. representativewesternblot,whichreflectsanaverageproteininductionof4.8±1.5 after4daysNGFtreatment(mean±SEM,n=3,datanotshown).Interestingly,also withoutNGFstimulation,endogenousexpressionofDCLKshortispresent. NGFinducesphosphorylationofDCLKshortSer30viaERK1/2 Treatment of PC12 cells with NGF induces neuronal differentiation, which is dependentonsignalingoftheMAPKExtracellularsignalregulatedkinase(ERK)1/2 [4].WeidentifiedwithintheNterminal,SPrichdomainaProlineTyrosineserine Prolinesequence(28PTSP31;accessionnumberAAC99476).Thiscomplieswitha PXS/TP motif, which has been described as a generalized ERK 1/2 recognition motif, in which the serine is phosphorylated [29]. In order to measure the phosphorylation state of DCLKshort Ser30, we generated and tested the phosphospecificantibodypSer30DCLK(seeMaterialsandMethodsandFig.2).As canbeseeninFig.2A,total,purifiedrecombinantDCLKproteincouldbedetected byantiDCLKantibody.WhenincubatedwithrecombinantERK1/2(lane46),DCLK protein migrated slower than without ERK 1/2 (lane 13). Retardation of DCLK protein migration may be expected as a result of phosphorylation by ERK 1/2. Mutationofserine30toAlanine(S30A)abolishedthisphosphorylationsiteforERK 1/2andconsequentlyimmunoreactivityofpSer30DCLK(Fig.2B;lane5).Mutation of serine 30 to aspartate (S30D) mimicked constitutive phosphorylation of serine DCLK protein ERK. WT S/A S/D WT S/A S/D. -. - -. + + +. A. B. DCLK. 50. pSer30-DCLK Short Exposure. 37 C. pSer30-DCLK Long Exposure. 50 37. D. 50. 32P--ATP. 37 kD. 1 2 3 4 5 6. Figure 2. Specificity of antibody pSer30-DCLK. Antibody pSer30-DCLK was generated to specifically recognize the phosphorylation of Ser30 (within the 28 PTSP 31 motif, accession number AAC99476). Purified, recombinant wild type and serine substitution mutant (S30A and S30D) DCLK protein were incubated with or without ERK 1/2 and assayed for immunoreactivity A)-C). or radioactivity D). A). Total DCLK protein. B). and C). pSer30-DCLK-antibody short and long exposure, respectively. D). 32p-radioactivity as a measure for transfer of any phosphogroup, regardless target amino acid residues.. 99.

(11) 100. Chapter4. and maintained immunoreactivity with the pSer30DCLK antibody, although its immunoreactivity was much lower than wild type (Fig. 2C, lane 6). In conclusion, antibody pSer30DCLK recognizes Ser30specific phosphorylation and was therefore used to monitor Ser30 phosphorylation of DCLK in subsequent experiments.ThesealsodatashowedthatrecombinantERK1/2isabletodirectly phosphorylate recombinant DCLK protein in vitro. Radioactive kinase assays revealed that wild type, S30A and S30D protein remained susceptible to phosphorylationbyERK1/2(Fig.2D,lane46).Seemingly,alsoDCLKresiduesother thanSer30arephosphorylatedbyERK1/2. Subsequently, we tested whether Ser30 DCLKshort phosphorylation occursinNs1PC12cellsafterNGFstimulation(Fig.3).Ascanbeseen,DCLKshort andERK1/2areendogenouslyexpressedinNs1PC12cells.StimulationwithNGF for15minutesresultedintypicalactivationofERK1/2,butalsoinphosphorylation ofDCLKshortonSer30.AdditionofMEKinhibitorU0126priortoNGFapplication nearly abolished ERK 1/2 activation, and DCLKshort phosphorylation. Then, we overexpressedFLAGtaggedWTDCLK,S30ADCLKandS30DDCLKinNs1PC12cells (Fig. 4). As can be seen,high levels of recombinantDCLK were recognized by the DCLK and FLAG antibodies, as compared to the low levels of endogenous DCLK (faint, lower band). Treatment with NGF again induced ERK 1/2 activation (pERK and ERK) and phosphorylation of endogenous DCLKshort Ser30 (pSer30DCLK).. A. B. NGF. MEK. +. -. +. NGF -. -. +. +. 44 kD 42 kD. pERK. 44 kD 42 kD. ERK. 53 kD. pSer30-DCLK. 53 kD. DCLK. 50 kD. Tubulin. p. U0126 p. 1. U0126 -. ERK. PTSP. DCLK-short. 433. Figure 3. NGF induces phosphorylation of Ser30 DCLK-short via an ERK 1/2dependent pathway. A). Scheme of NGF signaling in Ns-1 PC12 cells. Stimulation with NGF is known to activate the Ras-Raf-MEK-ERK 1/2 pathway after TrkA-binding. As PTSP is a ERK 1/2 recognition motif and present in DCLK, DCLK was hypothesized to be an ERK 1/2 target. B). Western blot analysis shows phosphorylation of Ser30 DCLK, dependent on ERK 1/2 activity. 60 minutes of NGF stimulation leads to ERK 1/2 activation, whereas coincubation with U0126 inhibits NGF-induced ERK 1/2 activation. Phosphorylation of Ser30 DCLK occurs only when ERK 1/2 is activated and therefore is DCLK a downstream target of active ERK 1/2..

(12) Chapter4. EV. -. +. -. S30A. S30D. WT. DCLK. DCLK. DCLK. +. -. +. -. +. NGF. pSer30-DCLK. 54 kD 53 kD. DCLK. 54 kD 53 kD. FLAG. 54 kD. pERK. 44 kD 42 kD. ERK. 44 kD 42 kD. -actin. 43 kD. Figure 4. Overexpression of WT DCLK and Ser30 phosphomutants. Ns-1 PC12 cells were transfected with empty vector (EV; pcDNA 3.1 +), WT DCLK, S30A DCLK or S30D DCLK. Overexpressed DCLK-short is shown by anti-FLAG and anti-DCLK antibody (54kD), whereas endogenous DCLK-short (53 kD) is shown by anti-DCLK antibody only. Cells were stimulated with (+) or without (-) 1 hour NGF to determine the phosphorylation status of DCLK proteins, using pSer30-DCLK. Activated ERK 1/2 (phosphorylated ERK 1/2 or pERK 1/2) and total ERK 1/2 protein was used to verify NGF treatment and to determine activated-ERK 1/2 dependent DCLK-short phosphorylation. Beta-actin served as loading control. DCLK abbreviates DCLK-short.. Notably, pSer30DCLKshort recognized overexpressed wild type DCLKshort, with or without NGF stimulation. However, whereas wild type DCLKshort was recognizedbypSer30DCLK,specificmutationofSer30disruptedimmunoreactivity forDCLKshort.Basedonthesedatatogether,weconcludethatendogenousNGF signaling in Ns1 PC12 cells entails ERK 1/2mediated phosphorylation of DCLK shortSer30. DCLKshortresidespredominantlyinthecytoplasmiccompartment Previously, partial nuclear translocation of overexpressed DCLKshort upon cAMP stimulation was reported [19]. In addition, ERK 1/2 is known to translocate from thecytoplasmtothenucleusuponphosphorylation[3].Therefore,wedetermined possiblenucleartranslocationofDCLKshortinresponsetoNGFstimulation.After treatingthecellswithNGF,wepreparednuclearandcytosolicfractions.Then,we determined the relative content of phosphorylated DCLKshort and total DCLK shortinthesefractionsbywesternblot.Fig.5Aisarepresentativeblot,whileFig.. 101.

(13) Chapter4. A. 0. 2. 8. 24 Hours of NGF. pSer30DCLK (53 kD). Nuclear Cytoplasm. DCLK (53 kD). Nuclear. Actin (43 kD). Cytoplasm. Loading Control. C pSer30-DCLK. 3. *. DCLK. 0.5. Relative signal intensity. B Relative signal intensity. 102. 0.4. * 2. 0.3. *. 1. 0.2. *. 0.1. 0. 0.0. 0. 2. 8. Nucleus. 0. 2. 8. Cytoplasm. 0. 2. 8. Nucleus. 0. 2. 8. Cytoplasm. Figure 5. Cell fraction quantifications of DCLK-short translocation in response to NGF. A.) Immunoblot for detection of Ser30-phosphorylated DCLK-short protein and total DCLK-short protein in nuclear and cytoplasmic fractions after 0, 2 and 8 hours of NGF stimulation. B.) and C). Quantification of relative protein content (optical density) in the different subcellular fraction, based on triplo measurement (mean values ± SEM). Asterisks (*) indicate significant difference as compared to unstimulated group (p<0.05). DCLK abbreviates DCLK-short.. 5Band5Carecorrespondingquantifications(n=3pertimepoint). As can be seen in Fig. 5B, NGF treatment of 2 and 8 hours significantly induced phosphorylationofDCLKshortproteininthenucleusandinthecytoplasm(p<0.05; n=3, ANOVA). In contrast with the nucleus, some basal levels of phosphorylated DCLKshortwerepresentinthecytoplasmbeforeNGFstimulation.Measurements oftotalDCLKshortprotein(Fig.5AandC)suggestedamodestnuclearenrichment andcytoplasmicdepletionafter2hoursofNGFstimulation,whichneutralizedafter 8 hours. However, change in DCLKshort protein localization was not significant. Apparently, although NGF induced a nuclear DCLKshort phosphorylation, this.

(14) Chapter4. 103. phosphorylation resulted in no intracellular localization changes and left most DCLKshort protein cytoplasmic. In order to reinforce these observations, we performed immunocytochemistry. For identical periods of NGF stimulation, we quantifiedlocalizationofphosphorylatedandtotalDCLKshort(Fig.6AandB).Two hours after NGF addition, significant nuclear DCLKshort phosphorylation was detected, together with a significant increase in cytoplasmic DCLKshort phosphorylation. After 8 hours, most phosphorylated DCLKshort signal was present in the cytoplasmic compartment, as found by cell fractionation assays. Total DCLKshort protein showed no detectable nuclear translocation in response to2hoursofNGFstimulation,butasignificantincreaseincytoplasmicDCLKshort contentafter8hours.Inaddition,cytoplasmicandnuclearlocalizationratiosofWT DCLKEGFP and S30A DCLKEGFP or S30D DCLKEGFP showed no obvious differences(datanotshown). To further examine a possible role for DCLKshort in the nucleus, we determined the effects of wild type DCLKshort and phosphomutants on immediateearly gene expression. To this aim, we overexpressed the 3 different DCLKshortconstructsandcontrolvectorandmeasuredtheinductionofcFosand Egr1by1hourofNGFtreatment(Fig.7).Fig.7Adepictsthequantifiedinductions ofcFosbyNGFforemptyvector(EV;27.0±6.7),wildtypeDCLKshort(25.1±2.3), S30A DCLKshort (34.5±2.0) and S30D DCLKshort (32.2±4.0). Among these experimental groups, no significant differences were found (ANOVA, p>0.05). Fig. 7B depicts the quantified inductions of cFos by NGF for empty vector (EV; 520.4±127),wildtypeDCLKshort(417.3±88.6),S30ADCLKshort(381.7±44.0)and S30D DCLKshort (542.1±157.8). Neither for Egr1, presence of wild type or phosphomutant DCLKshort protein significantly affected its mRNA induction by NGF(ANOVA,p>0.05).Fromthesedatatogether,weconcludedthatafunctionof DCLKshortmaybecytoplasmic.. B. Integrated Pixel Density. pSer30-DCLK. *. 2000000. * * * 1000000. 0. Integrated Pixel Density. A. 1500000. DCLK. 1000000. 500000. 0. 0. 2 Nucleus. 8. 0. 2. 8. Cytoplasm. T(h). 0. 2 Nucleus. 8. 0. 2. 8. T(h). Cytoplasm. Figure 6. Immunocytochemical quantification of DCLK-short translocation in response to NGF. Quantification of signal intensities from Ser30-phosphorylated DCLK-short protein and total DCLK-short protein, for nuclear and cytoplasmic compartments. Ten images with at least 50 cells per experimental group were analyzed to obtain mean integrated optical density values (± SEM). Asterisks (*) indicate significant differences as compared to unstimulated group (p<0.05 ANOVA). A). Ser30-phosphorylated DCLK-short localization in nucleus and cytoplasm. B). Total DCLK-short protein localization in nucleus and cytoplasm. DCLK abbreviates DCLK-short..

(15) 104. Chapter4. 40. c-Fos induction. Egr1 induction. 800 600 400 200 0. EV. WT DCLK S30A DCLK S30D DCLK. 30 20 10 0. EV. WT DCLK S30A DCLK S30D DCLK. Figure 7. DCLK-short plays no role in induction of immediate-early genes c-Fos and Egr1 by NGF. Ns-1 PC12 cells were transfected with empty vector (EV; pcDNA 3.1 +), WT DCLK, S30A DCLK or S30D DCLK. Then, cells were treated without NGF or 1 hour with NGF to induce immediate-early gene expression. Relative mean induction (±SEM) for each construct was determined by qPCR for A). c-Fos and B). Egr1. No significant differences were found across the groups, indicating DCLKshort, or its phosphorylation, plays no role in the measured immediate-early gene induction (ANOVA, p>0.05). n=5/6 per group. DCLK abbreviates DCLK-short. . DCLKshortcolocalizeswithfilamentousactiningrowthcones TogainmoreinsightinthesubcellularlocalizationofendogenousDCLKshort,we performed confocal fluorescence microscopy. In general, total DCLKshort distribution was characterized by a punctuate appearance throughout the cell, bothinunstimulatedandNGFstimulatedcells(Fig.8A;seeAppendixofthisthesis forallcolorfigures).However,whenDCLKshortwasobservedacrossthedepicted time series, immunoreactivity was striking at the end of developing neurites. Because distal neurite ends typically bear actinsupported growth cones, we performed a costaining of filamentous actin and (phospho)DCLKshort. This revealed that DCLKshort indeed localized to growth cones, where it colocalized withfilamentousactin(Factin;Fig.8).PriortoNGFtreatment(whenSer30DCLK short is unphosphorylated) DCLKshort resided at Factin rich sites, from which neuritestypicallysprout(Fig.8A).Also,pSer30DCLKsignalcolocalizedwithFactin in growth cones after NGF stimulation (Fig.8B). After 24 hours of NGFtreatment, DCLKshort localization to growth cones was more explicit, suggesting Ser30 phosphorylation played a role in this process. Subsequently, we tested whether distribution of DCLKshort protein may be altered by disruption of the actin cytoskeleton by Cytochalasin D treatment. As can be seen in Fig. 9A and 9B, Cytochalasin D treatment resulted in partial collapse of the actin cytoskeleton, characterized by formation of Factin aggregates in the cell body. Interestingly, Cytochalasin D treatment resulted in redistribution of DCLKshort in the cell. Typically,DCLKshortremainedcolocalizedwithFactinandreplicateddistribution patternssimilartoFactinaggregates..

(16) Chapter4. To further validate DCLKshortFactin interaction and possible functional involvementofSer30,westudiedcolocalizationofFactinwithDCLK(short)EGFP fusion proteins. Fig. 10 shows images of NGFtreated Ns1 PC12 cells overexpressingWTDCLKEGFP,S30ADCLKEGFPandS30DDCLKEGFP(Fig.10A,B andC).Ingeneral,All3fusionproteinsappearedtobemoreuniformlydistributed throughoutthecell,withnoobservabledifferencesbetweenwildtypeormutant variants.InlinewiththeobservationsfromendogenousDCLKshort,recombinant DCLKshortcolocalizedwithFactin(Fig.10).Moreover,all3DCLKfusionproteins were affected by Cytochalasin D treatment as well. Under these circumstances, DCLK protein retained colocalization with Factin and showed indications of aggregate formation similar to Factin. Together, we conclude that DCLKshort colocalizes to Factin and that intracellular localization of DCLKshort depends on anintactactincytoskeleton. Phosphorylation of DCLKshort Ser30 affects neuritogenesis and GAP43 expression ThestrikingcolocalizationofbothnativeDCLKandDCLKshortSer30withFactin in cellular protrusions suggested a role for DCLKshort in neuritogenesis. To test this hypothesis we overexpressed WT DCLK, S30A DCLK and S30D DCLK in Ns1 PC12cellsandstudiedtheirconsequencesonneuritogenesis.First,overexpression of WT DCLKshort enhanced NGFinduced differentiation, as quantified by the degree of neuritogenesis (Fig. 11A). Compared to the empty vector, WT DCLK expression significantly elevated the fraction of neuritebearing cells by 13% (respectively,34%±2%to44%±2%oftotal;p<0.05;Fig.11A).Incontrast,S30A DCLKorS30DDCLKshowednosignificantlyenhancedneuritogenesiscomparedto. A. B *. 0.4 0.3 0.2 0.1 0.0. 6. Gap43 induction. Fraction neurite bearing cells. 0.5. EV. WT DCLK S30A DCLK S30D DCLK. *. 4. 2. 0. EV. WT DCLK S30A DCLK S30D DCLK. Figure 11. Phosphorylation of DCLK-short Ser30 regulates neuritogenesis and GAP43 expression. A). Overexpression of WT DCLK increases the fraction of cells that display neurite outgrowth after 24 hours of NGF stimulation, as compared to empty vector (EV: pcDNA 3.1 +). Mutation S30A or S30D diminished the neuritogenic effect of WT Bar values indicate mean fraction (±SEM) of neurite bearing cells as counted on 15 micrographs (with an average of >40 cells per image). A neurite was scored when the length appeared at least the cell soma diameter. B). WT DCLK increases GAP43 induction, whereas mutation S30A or S30D diminished this enhancement. Mean induction (±SEM) of GAP43 was determined by qPCR before and after 24 hours of NGF stimulation. n=3/4 per group. DCLK abbreviates DCLK-short. Asterisks (*) indicate significant difference as compared to EV (p<0.05).. 105.

(17) 106. Chapter4. empty vector. It appears that mutation of Ser30 disrupted the observed neuritogeniceffectofWTDCLK. Second, induction of growth associated protein 43, a marker for neuritogenesis, was used to determine effects of DCLKshort on neuronal outgrowth (GAP43; [30]). By means of qPCR, GAP43 mRNA induction was determined 1 day after NGF addition, using the same time frame as the morphologicalassessment.Ascomparedtotheemptyvector,GAP43inductionwas significantlyincreasedbywildtypeDCLKshortoverexpressionby53%(3.1±0.1to 4.7±0.5foldinduction;P<0.05;Fig.11B).However,whenSer30wasmutated,the NGFstimulated GAP43 induction remained unchanged, as no significant differenceswerefoundbetweenS30AorS30Dandthecontrolvector.Apparently, thedominantnegativemutantS30ADCLKfunctionedtoreducethedifferentiation promoting effects of wild type DCLKshort. Interestingly, the S30D mutant DCLK ratherdiminishedthanfurtherpotentiatedwildtypeDCLKshorteffects. DISCUSSION Neuronal differentiation of PC12 cells driven by NGF occ urs through extensive regulationofgeneexpressionandproteinactivity[1;3;7].How,andexactlywhich activated pathways and genes produce a cell of neuronal properties remains incompletely understood. Here, we show that in Ns1 PC12 cells, NGF promotes neuronal outgrowth by signaling through DCLKshort. In response to NGF stimulation, ERK 1/2 becomes activated and phosphorylates endogenous DCLK shortatSer30.Byusingphosphomutants,weshowedthatthisphosphorylationis important for subsequent neurite outgrowth. Clearly, regulation of cytoskeletal dynamics underlies neurite outgrowth. In this light, our observation that DCLK shortcolocalizeswithfilamentousactiningrowthconesisparticularlyinteresting. Given the Ser30dependent effects of DCLKshort on neuritogenesis, we hypothesize that DCLKshort is a phosphodependent regulator of the actin cytoskeleton. We conclude that DCLKshort provides a previously unknown connectionbetweengrowthfactorsignaling,ERK1/2andneuronaldifferentiation. WithintheNterminalSPrichdomainofDCLKshort,werecognizedSer30 withinthe28PTSP31motifasapotentialPXSPERK1/2phosphorylationsite[29]. ThiswasinterestingasERK1/2fulfillsacentralroleintransducingtheNGFsignal into a cellular response of differentiation [3]. Deployment of phosphospecific antibody pSer30DCLK showed that ERK 1/2 directly phosphorylates recombinant DCLKproteinattheidentifiedSer30residueinvitro.Moreinterestingly,exposure ofNs1PC12cellstoNGFresultedinclassicalactivationofendogenousERK1/2and inconcomitantphosphorylationofendogenousDCLKshortatSer30.Applicationof ERK inhibitor U0126, showed that ERK 1/2 activation corresponds to Ser30 phosphorylation. Therefore, these data show that DCLKshort is a substrate of activatedERK1/2inPC12cellsandtherebyadownstreamtargetofNGFsignaling..

(18) Chapter4. UponNGFstimulation,wefoundnoevidenttranslocationofDCLKshortto the nucleus. In response to NGF, ERK 1/2 translocates to the nucleus, where it is crucialfordifferentiationbyinductionofimmediateearlygenes[31;32].Moreover, Silvermanetal.previouslyreportedthatexogenousDCLKshortinCOS7andNIH3T cells partially translocates to the nucleus upon cAMP stimulation [19] and may function to inhibit CREdriven transcription [19;21]. Therefore, we investigated whether similar mechanisms apply to endogenous DCLKshort. Both cell fractionationandimmunocytochemistryexperimentsindicatedthatNGFproduces a significant nuclear signal of DCLKshort phosphorylation, however, not of DCLK shorttranslocation.Cellfractionationexperimentsshowedaweak,nonsignificant nucleartranslocationoftotalDCLKshortproteinafter2hoursofNGFtreatment (Fig. 5). Phosphorylated DCLKshort was significantly increased in the nucleus, although most phosphorylated DCLKshort remained cytoplasmic. Also, low levels of phosphorylated DCLKshort, in absence of NGF stimulation, were found in the cytoplasm. To support our findings, we performed immunocytochemistry experiments. These indicated a cytoplasmic translocation of total DCLKshort in responsetoNGF(Fig.6).AlthoughphosphorylatedDCLKshortwasobservedinthe nucleusinresponsetoNGF,thecytoplasmiccompartmentalwayscontainedmore phosphorylated DCLK. As stated by Silverman et al., nuclear localization of DCLK short is intriguing, however, nuclear translocation account ed for a small proportionuponcAMPtreatmentanditslocalizationremainedprimarilycytosolic [19]. AlthoughDCLKshortmayhavesomenuclearfunction,itplayednorolein measuredimmediateearlytranscription.ThetranscriptionalinductionofcFosand Egr1isdependentonERK1/2activationandpartiallyCREactivation[6].However, inductionofEgr1andcFosmRNAby1hourofNGFtreatmentwasnotsignificantly differentforNs1PC12cellstransfectedwithWTDCLK,S30ADCLKorS30DDCLK and empty vector (Fig. 7). Although two studies reported varying degrees of inhibition of CREdriven transcription by DCLK, Egr1 and cFos induction were unaffectedinourstudy[19;21].Thesedifferentobservationsmaybeexplainedby use of different cell lines, different stimuli, different readouts for gene transcription and importantly, use of a different DCLK protein by Ohmae et al, a truncated DCLKlong variant. This protein has an additional microtubulebinding domain and a hyperactive kinase activity, by removal of its Cterminal auto inhibitory domain [21]. Both may cause relevant changes in intracellular localizationandsignaling. MicroscopyindicatedthatthesubcellulardistributionofDCLKshorthada punctuate appearance in Ns1 PC12 cells. This was shown by using two different antibodies, against either total or Ser30phosphorylated DCLKshort. Because the DCLK gene produces several splice variants and because the DCLK gene is highly homologous to DCX and DCLK2, we controlled the specificity of these antibodies (datanotshown).WesternblotandPCRexperimentsindicatedthatDCLKshortis the only DCLK splice variant expressed in Ns1 PC12 cells. DCLK2, which is homologous with the DCLKshort Ser30 region, is expressed in Ns1 PC12 cells. 107.

(19) 108. Chapter4. DCXishomologoustocertainDCLKregions,butisnotexpressedinPC12cells[35]. Based on this, we considered the observed intracellular, speckled distribution as specificforDCLKshort.AlsothedistributionsofoverexpressedDCLKEGFPproteins showed punctuate structures, although they were in general more uniform. Together, these data indicate that DCLKshort forms discrete structures in Ns1 PC12cells. Previously,Deueletal.impliedDCLKgeneproductsinmicrotubulebased vesicletransport[36].DCLKshort,however,containsnoDCXdomainsthatenable bindingtomicrotubules[37].Inlinewiththis,DCLKshorthasbeenreportednotto bind microtubules [38]. Nor did our experiments suggest this. However, the conserved SPrich domain, in which Ser30 lies, is known to be a proteinprotein interaction domain and may mediate formation of discrete protein complexes in Ns1 PC12 cells. Specifically, the Nterminal SPrich domain, as contained in DCLK splice variant DCL, has previously been described as an interaction domain importantinglucocorticoidreceptortransport(GR;[17]).TheGRisanimportant modulatorofneuronalfunction,alsoin(differentiated)PC12cells[39].Therefore, DCLKshortmaybeinvolvedinneuronalGRsignalinginNs1PC12cells.Also,the SPrichdomainoftheDCLKsplicevariantCARP,waspreviouslyreportedtointeract withGrowthreceptorbound2(Grb2,[40]).Grb2isknowntobeincorporatedinto the signaling endosome, a protein complex that mediates NGF signaling [41]. It wouldbeofinteresttofurtherinvestigatewhethersuchinteractionsplayarolein neuronaldifferentiationofNs1PC12cellsaswell. Most strikingly, DCLKshort appeared enriched at cellular protrusions before NGF stimulation and after NGF stimulation in growth cones. These are typically actin supported structures [42]. Confocal microscopy studies indeed confirmedthatbothendogenousDCLKshortandDCLKshortEGFPcolocalizewith filamentous actin in growth cones. Also Cytochalasin D treatment showed that, intracellular DCLKshort distribution is clearly affected by inducing collapse of the actin cytoskeleton and apparently physically dependent on an intact actin cytoskeleton.WhiledoublecortindomaincontainingDCLKgeneproductstypically areconsideredmicrotubulebindingproteins(MAPs),Coquelleetal.reportedthat all11genesoftheDCXgenefamilyinteractwithmicrotubulesandthat7outof11 genes of the DCX gene family also interact with actin [43;44]. According to this study, DCLK did not interact with actin, however, DCLKlong and not DCLKshort was assayed here. Interestingly, endogenous DCLKshort immunoreactivity at F actinrich sites appeared to increase after NGF treatment (Fig. 8). This raises the possibility that DCLKshort regulates the actin cytoskeleton during NGFinduced neuritogenesis. Such functions have been found for other members of the DCX gene family. DCX is also known to interact with actin, either directly or indirectly throughNeurabinII[43;45;46].Moreover,DCXisphosphorylatedbyJNKintheSP rich domain at T321, T331 and S334 [47]. Phosphorylation increased filamentous actinaffinity,particularlyatgrowthcones.OverexpressionofDCXphosphomutants for these JNKsites affected NGFinduced neurite outgrowth in PC12 cells and neuronal migration. Hence, comparison with other members of the DCX gene.

(20) Chapter4. family supports a role for DCLKshort in regulating actin dynamics, possibly in a phosphodependentmanner. Overexpression of wild type DCLKshort and DCLKshort mutant for the ERK 1/2 phosphorylation site, showed that DCLKshort is functional in promoting NGFinducedneuritogenesisandGAP43expression.Althoughsubstitutionofserine into aspartate is often used to mimic a phosphoserine [25], in our study this mutationratherdisruptedwildtypeDCLKshortfunctionthanfurtherenhancingit. Possibly,conformationalchangesinducedbyserinephosphorylationinDCLKshort are not identical to those induced by aspartate substitution [48;49]. In line with this, the used Ser30 phosphospecific antibody lost immunoreactivity when serine was substituted with asparate. Alternatively, DCLKshort may require dynamic phosphorylation and dephosphorylation in order to perform its cellular function. These dynamics may be unable to occur with mutant proteins locked in their phosphomicking conformations. Previously, CDK5mediated phosphorylation of DCX, a DCLK paralogue, was shown to occur on serine 297 [50;51]. Here, overexpression of phosphorylated wild type DCX increased neuronal migration, whereas positive and negative phosphomutants abolished disrupted this effect. Notwithstanding, DCLKshort function seems specifically dependent on Ser30 phosphorylation, as mutation of this particular residue abolished the ability to enhanceNGFinducedneuronaldifferentiation. HowexactlyDCLKshortenhancesneuritogenesisandGAP43inductionby NGF treatment, remains elusive. GAP43 is a neuronal differentiation marker synthesizedduringneuriteoutgrowthandisenrichedingrowthcones[52].GAP43 exerts its function by assembling actin filaments, positively dependent on its phosphorylation [48]. Therefore, increased GAP43 induction by NGF, in the presence of overexpressed DCLKshort, correlates well with increased neuritogenesis.ThisphenotypeincludestranscriptionalregulationbyDCLKshort, namely enhanced GAP43 mRNA induction, and cytoskeletal remodeling, namely enhancedneuritogenesis.Basedonthecurrentdata,itispossiblethatDCLKshort has a primary transcriptional function and that the regulated genes caused differences in neuritogenesis. Alternatively, DCLKshort may have a primary functionasacytoskeletalproteinanditscytoskeletalmodelingcauseddifferences in transcription. Also a combination of these functions is possible. ERK 1/2, for instance,playsanimportantroleingeneinductioninthenucleus,butalsoinactin dynamicsoutsidethenucleus[3;53;54].Althoughnomechanismcanbeexcluded, DCLKshort promotes phosphorylationdependent neuritogenesis and is found to colocalizewithactininthegrowthcone.Therefore,thesedataprovidebasisfora so far unexplored function of DCLKshort, as a regulator of actin dynamics. In support, many other phosphoproteins have been described which regulate actin dynamics during neuritogenesis, including Cofilin, LIM kinase, Slingshot, Fes, Rac, Cdc42andothers[5558]. In conclusion, we report that DCLKshort has a regulatory role in NGF inducedneuronaldifferentiationofNs1PC12cells.InresponsetoNGFstimulation, ERK 1/2 is activated and phosphorylates DCLKshort at Ser30. Although some. 109.

(21) 110. Chapter4. subsequentnucleartranslocationoccurred,thiswasnotsignificantandDCLKshort played no apparent role in immediateearly gene transcription. Within the cell, mostDCLKshortproteinremainedoutsidethenucleusandclusteredinpunctuate structures, reminiscent of protein complexes. Moreover, DCLKshort colocalized with Factin before NGF stimulation and in growth cones after NGF stimulation. DCLKshortanditsphosphorylationwereshowntoparticipateinthegenerationof neurites; for which regulation of the actin cytoskeleton is a plausible underlying mechanism. Such a function is well in line with phosphoregulation of the microtubuleandactincytoskeletonbyothermembersoftheDCXgenefamily.So far, involvement of DCLKshort in mediating NGFinduced differentiation was unknown. Therefore, the present study contributes to a better understanding of neuronaldifferentiationandfunctionsoftheDCXgenefamily. ACKNOWLEDGEMENTS This research is supported by the Dutch Technology Foundation STW, applied sciencedivisionofNWO,andtheTechnologyProgramoftheMinistryofEconomic Affairs(projectnumberLFA6332). .

(22) Chapter4. REFERENCES 1. K.Fujita,P.Lazarovici,andG.Guroff,RegulationofthedifferentiationofPC12 pheochromocytomacells,Environ.HealthPerspect.80:12742.(1989)127142. . 2. A.Riccio,B.A.Pierchala,C.L.Ciarallo,andD.D.Ginty,AnNGFTrkAMediated RetrogradeSignaltoTranscriptionFactorCREBinSympatheticNeurons,Science277 (1997)1097.. . 3. D.Vaudry,P.J.S.Stork,P.Lazarovici,andL.E.Eiden,SignalingPathwaysforPC12Cell Differentiation:MakingtheRightConnections,Science296(2002)16481649.. . 4. D.R.KaplanandR.M.Stephens,NeurotrophinsignaltransductionbytheTrkreceptor, JournalofNeurobiology25(1994)14041417.. . 5. H.R.Herschman,Primaryresponsegenesinducedbygrowthfactorsandtumor promoters,Annu.Rev.Biochem.60:281319.(1991)281319.. . 6. M.ShengandM.E.Greenberg,Theregulationandfunctionofcfosandother immediateearlygenesinthenervoussystem,Neuron.4(1990)477485.. . 7. T.F.Dijkmans,L.W.vanHooijdonk,T.G.Schouten,J.T.Kamphorst,A.C.Vellinga, J.H.Meerman,C.P.Fitzsimons,E.R.deKloet,andE.Vreugdenhil,Temporaland functionaldynamicsofthetranscriptomeduringnervegrowthfactorinduced differentiation,J.Neurochem.(2008).. . 8. T.F.Dijkmans,L.W.vanHooijdonk,T.G.Schouten,J.T.Kamphorst,C.P.Fitzsimons,and E.Vreugdenhil,IdentificationofnewNerveGrowthFactorresponsiveimmediate earlygenes,BrainRes.1249:1933.(2009)1933.. . 9. E.Vreugdenhil,B.Engels,R.Middelburg,K.S.van,J.Knol,B.Veldhuisen,andE.R.de Kloet,MultipletranscriptsgeneratedbytheDCAMKLgeneareexpressedintherat hippocampus,BrainRes.Mol.BrainRes.(2001)6774.. 10. O.Reiner,F.M.Coquelle,B.Peter,T.Levy,A.Kaplan,T.Sapir,I.Orr,N.Barkai,G.Eichele, andS.Bergmann,Theevolvingdoublecortin(DCX)superfamily,BMC.Genomics. 7:188.(2006)188. 11. H.Koizumi,T.Tanaka,andJ.G.Gleeson,Doublecortinlikekinasefunctionswith doublecortintomediatefibertractdecussationandneuronalmigration,Neuron.49 (2006)5566. 12. T.Shu,H.C.Tseng,T.Sapir,P.Stern,Y.Zhou,K.Sanada,A.Fischer,F.M.Coquelle, O.Reiner,andL.H.Tsai,Doublecortinlikekinasecontrolsneurogenesisbyregulating mitoticspindlesandMphaseprogression,Neuron.49(2006)2539.. 111.

(23) 112. Chapter4. 13. E.Vreugdenhil,S.M.Kolk,K.Boekhoorn,C.P.Fitzsimons,M.Schaaf,T.Schouten, A.Sarabdjitsingh,R.Sibug,andP.J.Lucassen,Doublecortinlike,amicrotubule associatedproteinexpressedinradialglia,iscrucialforneuronalprecursordivision andradialprocessstability,Eur.J.Neurosci.25(2007)635648. 14. C.Kappeler,Y.Saillour,J.P.Baudoin,F.P.Tuy,C.Alvarez,C.Houbron,P.Gaspar, G.Hamard,J.Chelly,C.Metin,andF.Francis,Branchingandnucleokinesisdefectsin migratinginterneuronsderivedfromdoublecortinknockoutmice,Hum.Mol.Genet. 15(2006)13871400. 15. T.Sapir,D.Horesh,M.Caspi,R.Atlas,H.A.Burgess,S.G.Wolf,F.Francis,J.Chelly, M.Elbaum,S.Pietrokovski,andO.Reiner,Doublecortinmutationsclusterin evolutionarilyconservedfunctionaldomains,Hum.Mol.Genet.9(2000)703712. 16. G.Friocourt,A.Koulakoff,P.Chafey,D.Boucher,F.Fauchereau,J.Chelly,andF.Francis, Doublecortinfunctionsattheextremitiesofgrowingneuronalprocesses,Cereb. Cortex.13(2003)620626. 17. C.P.Fitzsimons,S.Ahmed,C.F.Wittevrongel,T.G.Schouten,T.F.Dijkmans, W.J.Scheenen,M.J.Schaaf,E.R.deKloet,andE.Vreugdenhil,Themicrotubule associatedproteindoublecortinlikeregulatesthetransportoftheglucocorticoid receptorinneuronalprogenitorcells,Mol.Endocrinol.22(2008)248262. 18. T.Tanaka,H.Koizumi,andJ.G.Gleeson,Thedoublecortinanddoublecortinlikekinase 1genescooperateinmurinehippocampaldevelopment,Cereb.Cortex.16Suppl 1:i6973.(2006)i69i73. 19. M.A.Silverman,O.Benard,H.Jaaro,A.Rattner,Y.Citri,andR.Seger,CPG16,anovel proteinserine/threoninekinasedownstreamofcAMPdependentproteinkinase,J. Biol.Chem.274(1999)26312636. 20. P.T.Lin,J.G.Gleeson,J.C.Corbo,L.Flanagan,andC.A.Walsh,DCAMKL1encodesa proteinkinasewithhomologytodoublecortinthatregulatesmicrotubule polymerization,J.Neurosci.20(2000)91529161. 21. S.Ohmae,S.TakemotoKimura,M.Okamura,A.chiMorishima,M.Nonaka,T.Fuse, S.Kida,M.Tanji,T.Furuyashiki,Y.Arakawa,S.Narumiya,H.Okuno,andH.Bito, Molecularidentificationandcharacterizationofafamilyofkinaseswithhomologyto Ca2+/calmodulindependentproteinkinasesI/IV,J.Biol.Chem.281(2006)20427 20439. 22. H.A.BurgessandO.Reiner,Cleavageofdoublecortinlikekinasebycalpainreleases anactivekinasefragmentfromamicrotubuleanchoragedomain,J.Biol.Chem.276 (2001)3639736403. 23. M.Kruidering,T.Schouten,G.I.Evan,andE.Vreugdenhil,Caspasemediatedcleavage oftheCa2+/calmodulindependentproteinkinaselikekinasefacilitatesneuronal apoptosis,J.Biol.Chem.%19;276(2001)3841738425..

(24)

(25) . !"# "$ &' #*#-'@' \\" ^ `{|}# " {! " ' #\ \ \ \"\$ # \ { "`}# ' $ $ * \ \ $' $\# # " \ " * \ `}# { "^ ' *# {\ \" \ \ `*# }\! ! \ \ \ \\ `}# |' '\! \ # \ {\ ""\ `}# & & ! \ $ \ " "^^ { "`} # -- { \ { * - * ! \#\ \ \ \ & $ {# \"\'"\ { `}# | '- " \! \ ' \ "*- \ `}# | \\' \ ' ! @\ # \! `}# ^ ""- ^ $ | $ \ #! $ \# \ " " @ { "`}# '| '\ ! \ \ \ @ $ \ \ "\ !* \ `}# |" '| ^* !!! \ ^{#" \ -" `} #. .

(26) 114. Chapter4. 36. T.A.Deuel,J.S.Liu,J.C.Corbo,S.Y.Yoo,L.B.RorkeAdams,andC.A.Walsh,Genetic interactionsbetweendoublecortinanddoublecortinlikekinaseinneuronal migrationandaxonoutgrowth,Neuron.49(2006)4153. 37. K.R.Taylor,A.K.Holzer,J.F.Bazan,C.A.Walsh,andJ.G.Gleeson,Patientmutationsin doublecortindefinearepeatedtubulinbindingdomain,J.Biol.Chem.275(2000) 3444234450. 38. B.M.Engels,T.G.Schouten,D.J.van,I.Gosens,andE.Vreugdenhil,Functional differencesbetweentwoDCLKsplicevariants,BrainRes.Mol.BrainRes.120(2004) 103114. 39. M.C.Morsink,M.Joels,R.A.Sarabdjitsingh,O.C.Meijer,E.R.deKloet,andN.A.Datson, Thedynamicpatternofglucocorticoidreceptormediatedtranscriptionalresponses inneuronalPC12cells,J.Neurochem.99(2006)12821298. 40. G.J.Schenk,B.Engels,Y.P.Zhang,C.P.Fitzsimons,T.Schouten,M.Kruidering,K.E.Ron de,andE.Vreugdenhil,Apotentialroleforcalcium/calmodulindependentprotein kinaserelatedpeptideinneuronalapoptosis:invivoandinvitroevidence,Eur.J. Neurosci.26(2007)34113420. 41. L.A.GreeneandJ.M.Angelastro,Youcan'tgohomeagain:transcriptionallydriven alterationofcellsignalingbyNGF,Neurochem.Res.30(2005)13471352. 42. S.OkabeandN.Hirokawa,Actindynamicsingrowthcones,J.Neurosci.11(1991) 19181929. 43. F.M.Coquelle,T.Levy,S.Bergmann,S.G.Wolf,D.BarEl,T.Sapir,Y.Brody,I.Orr, N.Barkai,G.Eichele,andO.Reiner,Commonanddivergentrolesformembersofthe mouseDCXsuperfamily,CellCycle.5(2006)976983. 44. M.Tsukada,A.Prokscha,E.Ungewickell,andG.Eichele,Doublecortinassociationwith actinfilamentsisregulatedbyneurabinII,J.Biol.Chem.280(2005)1136111368. 45. M.Tsukada,A.Prokscha,J.Oldekamp,andG.Eichele,IdentificationofneurabinIIasa noveldoublecortininteractingprotein,Mech.Dev.120(2003)10331043. 46. M.Tsukada,A.Prokscha,andG.Eichele,NeurabinIImediatesdoublecortin dephosphorylationonactinfilaments,Biochem.Biophys.Res.Commun.343(2006) 839847. 47. A.Gdalyahu,I.Ghosh,T.Levy,T.Sapir,S.Sapoznik,Y.Fishler,D.Azoulai,andO.Reiner, DCX,anewmediatoroftheJNKpathway,EMBOJ.23(2004)823832. 48. A.YuehandS.P.Goff,Phosphorylatedserineresiduesandanargininerichdomainof themoloneymurineleukemiavirusp12proteinarerequiredforearlyeventsofviral infection,J.Virol.77(2003)18201829..

(27) Chapter4. 49. K.Shen,A.C.Hines,D.Schwarzer,K.A.Pickin,andP.A.Cole,Proteinkinasestructure andfunctionanalysiswithchemicaltools,Biochim.Biophys.Acta.1754(2005)6578. 50. T.Tanaka,F.F.Serneo,H.C.Tseng,A.B.Kulkarni,L.H.Tsai,andJ.G.Gleeson,Cdk5 phosphorylationofdoublecortinser297regulatesitseffectonneuronalmigration, Neuron.41(2004)215227. 51. B.T.Schaar,K.Kinoshita,andS.K.McConnell,Doublecortinmicrotubuleaffinityis regulatedbyabalanceofkinaseandphosphataseactivityattheleadingedgeof migratingneurons,Neuron.41(2004)203213. 52. K.H.Pfenninger,B.A.delaHoussaye,S.M.Helmke,andS.Quiroga,Growthregulated proteinsandneuronalplasticity.Acommentary,Mol.Neurobiol.5(1991)143151. 53. Q.Lu,M.Paredes,J.Zhang,andK.S.Kosik,Basalextracellularsignalregulatedkinase activitymodulatescellcellandcellmatrixinteractions,Mol.CellBiol.18(1998) 32573265. 54. A.K.PullikuthandA.D.Catling,ScaffoldmediatedregulationofMAPKsignalingand cytoskeletaldynamics:aperspective,CellSignal.19(2007)16211632. 55. Z.Mi,Z.K.Mirnics,andN.F.Schor,Bcl2overexpressiondisruptsthemorphologyof PC12cellsthroughreducedERKactivation,BrainRes.1112(2006)4655. 56. M.Endo,K.Ohashi,andK.Mizuno,LIMkinaseandslingshotarecriticalforneurite extension,J.Biol.Chem.282(2007)1369213702. 57. C.E.LaurentandT.E.Smithgall,ThecFestyrosinekinasecooperateswiththe breakpointclusterregionprotein(Bcr)toinduceneuriteextensioninaRacand Cdc42dependentmanner,Exp.CellRes.299(2004)188198. 58. J.S.daSilvaandC.G.Dotti,Breakingtheneuronalsphere:regulationoftheactin cytoskeletoninneuritogenesis,Nat.Rev.Neurosci.3(2002)694704. . 115.

(28) . .

(29)

No results found