Huntingtin Phosphorylation Sites Mapped by Mass Spectrometrycnc.cj.uc.pt/BEB/BEB/private/pdfs... · Huntingtin Phosphorylation Sites Mapped by Mass Spectrometry MODULATION OF CLEAVAGE

Huntingtin Phosphorylation Sites Mappedby Mass SpectrometryMODULATION OF CLEAVAGE AND TOXICITY*□S

Received for publication, December 19, 2005, and in revised form, May 23, 2006 Published, JBC Papers in Press, June 16, 2006, DOI 10.1074/jbc.M513507200

Birgit Schilling‡, Juliette Gafni‡, Cameron Torcassi‡, Xin Cong‡, Richard H. Row‡, Michelle A. LaFevre-Bernt‡,Michael P. Cusack‡, Tamara Ratovitski§, Ricky Hirschhorn§¶, Christopher A. Ross§!, Bradford W. Gibson‡**,and Lisa M. Ellerby‡1

From ‡The Buck Institute for Age Research, Novato, California 94945, the §Division of Neurobiology, Department of Psychiatry,!Departments of Neurology and Neuroscience, Johns Hopkins University School of Medicine, Baltimore, Maryland 21287,¶Hood College, Frederick, Maryland 21701, and the **Department of Pharmaceutical Chemistry, University of California,San Francisco, California 94143

Huntingtin (Htt) is a large protein of 3144 amino acids, whosefunction and regulation have not been well defined. Polyglu-tamine (polyQ) expansion in the N terminus of Htt causes theneurodegenerative disorder Huntington disease (HD). Thecytotoxicity of mutant Htt is modulated by proteolytic cleavagewith caspases and calpains generating N-terminal polyQ-con-taining fragments. We hypothesized that phosphorylation ofHtt may modulate cleavage and cytotoxicity. In the presentstudy, we have mapped the major phosphorylation sites of Httusing cell culture models (293T and PC12 cells) expressing full-length myc-tagged Htt constructs containing 23Q or 148Qrepeats. Purifiedmyc-taggedHtt was subjected tomass spectro-metric analysis including matrix-assisted laser desorp-tion/ionization mass spectrometry and nano-HPLC tandemmass spectrometry, used in conjunction with on-target alkalinephosphatase and protease digestions. We have identified morethan six novel serine phosphorylation sites within Htt, one ofwhich lies in the proteolytic susceptibility domain. Three of thesites have the consensus sequence for ERK1 phosphorylation,and addition of ERK1 inhibitor blocks phosphorylation at thosesites. Other observed phosphorylation sites are possibly sub-strates for CDK5/CDC2 kinases. Mutation of amino acid Ser-536, which is located in the proteolytic susceptibility domain, toaspartic acid, inhibited calpain cleavage and reduced mutantHtt toxicity. The results presented here represent the firstdetailedmapping of the phosphorylation sites in full-lengthHtt.Dissection of phosphorylationmodifications inHttmay provideclues to Huntington disease pathogenesis and targets for thera-peutic development.

Huntington disease (HD)2 is a neurodegenerative diseasecaused by an elongated polyglutamine (polyQ) stretch in thehuntingtin protein (Htt) and is characterized by involuntarymovements, personality changes, dementia, and early death (1).Under non-pathological conditions, this stretch of glutaminesranges from 2 to 34 repeats. Repeat lengths of 36 or greatercause HD. The threshold of polyQ length for Htt aggregation isquite similar to the threshold for disease suggesting that a con-formational change in Htt triggers the disease. However, itshould be pointed out that the aggregates forming inclusionsare not likely toxic themselves (2).Htt is a protein with 3144 amino acids with amolecular mass

of !350 kDa. Its function and regulation are incompletelyunderstood. Upon polyQ expansion in the N terminus of Htt, aconformationally altered protein is produced. Structural pre-dictions suggest that Htt is composed of four HH-HEATdomains with two unstructured proteolytic susceptibilitydomains (these areas contain PEST sequence) (Fig. 1). In addi-tion Htt harbors a potential nuclear localization signal at theend of the secondHH2 domain and a confirmed nuclear exportsignal in the third HH3 domain (3).Posttranslational modifications, especially phosphorylation

and proteolytic cleavage, may facilitate the initial conversion ofHtt from a normal to an abnormal conformation, and may beinitiating steps in a pathogenic cascade. Thus understandinghowHtt is posttranslationally modified in the context of polyQexpansion may give us insight into HD pathogenesis. Someposttranslational modifications of Htt have been described,such as ubiquination (4) and SUMOylation (5). One impor-tant previous report has shown that phosphorylation of Httat Ser-421 by Akt1 is neuroprotective against HD cellulartoxicity (6, 7).Proteolysis of Htt has been well characterized and occurs in

the proteolytic susceptibility region of Htt (Fig. 1). We havepreviously shown thatHtt is cleaved in several places by caspase

* This work was supported by National Institutes of Health Grant NS40251Aand HighQ (to L. M. E. and B. W. G.), National Institutes of Health Postdoc-toral Fellowship F32 NS043937 (to J. G.), National Institutes of HealthGrants NS 16375 and 38144, and the Huntington’s Disease Society ofAmerica and Hereditary Disease Foundation (to C. A. R.). The costs of pub-lication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement” inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Tables S1–S5 and Figs. S1–S10.

1 To whom the correspondence should be addressed: 8001 RedwoodBlvd., Novato, CA 94945. Tel.: 415-209-2088; Fax: 415-209-2230; E-mail:[email protected].

2 The abbreviations used are: HD, Huntington disease; MEK, mitogen-activatedprotein kinase/extracellular signal-regulated kinase kinase; CID, collision-induced dissociation; BisTris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxy-methyl)propane-1,3-diol; MALDI-MS, matrix-assisted laser desorption ioniza-tion mass spectrometry; ERK, extracellular signal-regulated kinase; PMF,peptide mass fingerprinting; HPLC, high performance liquid chromatography;ACN, acetonitrile; MOPS, 4-morpholinepropanesulfonic acid.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 33, pp. 23686 –23697, August 18, 2006© 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

23686 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 33 • AUGUST 18, 2006

at Ecole Polytechnique Fédérale de Lausanne on December 7, 2006

www.jbc.orgDownloaded from

http://www.jbc.org/cgi/content/full/M513507200/DC1Supplemental Material can be found at:

http://www.jbc.org

http://www.jbc.org/cgi/content/full/M513507200/DC1

enzymes (8, 9) and cleavage contributes to toxicity (10, 11).Caspase enzymes that can cleave Htt include caspase-2, -3, -6,and -7, and cleavage occurs between amino acids 513 and 587(9, 11, 12). Htt is cleaved in several places by calpains as well(13, 14) and this may also contribute to toxicity (15). Themajor calpain sites are located between amino acids 469 and537 (13, 15).Systematic analysis of phosphorylation or other types of

modifications of Htt protein and its consequences for Htt bio-chemistry and function have not previously been described. ToidentifyposttranslationalmodificationsinHtt,specificallyphos-phorylation, we generated myc-tagged full-length Htt23Q/148Q constructs for expression in 293T cells, PC12 cells, andtransgenics to use for our investigations (16). Htt was purifiedby immunoprecipitation using an anti-myc antibody, separatedby one-dimensional SDS-PAGE, and then subjected to proteo-lytic in-gel digestion andmass spectrometric analysis includingMALDI-MS peptide mass fingerprinting (PMF) and tandemmass spectrometry (nano-HPLC-ESI-MS, MS/MS). In addi-tion, a hypothesis driven multiple stage mass spectrometrystrategy was employed for phosphopeptide analysis using avMALDI-LTQmass spectrometer as described (17). This studydescribes the first extensivemass spectrometric sequencemap-ping and characterization of full-length Htt. During this workwe identified six novel phosphorylation sites. We describe thebiological significance of one of the phosphorylation sites interms of proteolysis of Htt and cytotoxicity in cell culture.

EXPERIMENTAL PROCEDURES

Materials—Reagents for protein chemistry includingiodoacetamide and dithiothreitol were obtained from Sigma.Sequencing grade modified trypsin (porcine, Promega), chy-motrypsin (bovine pancreas, Roche Molecular Biochemicals),andAsp-N protease (Pseudomonas fragi, RocheMolecular Bio-chemicals) were utilized for in-gel digestion reactions. HPLCsolvents, acetonitrile (ACN), and water were obtained fromBurdick & Jackson. A matrix solution of !-cyano-4-hydroxy-cinnamic acid (33 mM) in ACN/methanol (Agilent Technolo-gies) was used forMALDI-MS experiments. For vMALDI-MS/

MS, 2,5-dihydroxybenzoic acid (LaserBio Labs) in 50% ACNwas used as matrix.Cell Culture and Transfection—Full-length Htt constructs

with polyQ repeats of either 23Q or 148Q (pTet-c-myc-FL23Qand pTet-c-myc-FL148Q) were generated containing N-termi-nal c-myc epitopes (pTet-splice vector, Invitrogen) (16). pTet-c-myc-FL23Q and pTet-c-myc-FL148Q were transfected into293T cells (10-cm plate, Dulbecco’s modified Eagle’s mediumwith 10% fetal bovine serumwith Superfect (Qiagen) accordingto themanufacturer’s directions). Cells were harvested 48–72 hafter transfection. In addition, a stable inducible PC12 cellmodel expressing full-length Htt (pTet-c-myc-FL148Q) wasestablished as previously described (18). PC12 cells were grownin Dulbecco’s modified Eagle’s medium with 5% fetal bovineserum, 10% horse serum, 100 "g/ml G418, 200 "g/ml hygro-mycin, 200 ng/ml doxycycline, 100 units/ml penicillin, and 100units/ml streptomycin. Full-length Htt148Q was expressed byremoval of doxycycline. Cells used to obtain Htt for mass spec-trometric analysis were cultured under conditions of differen-tiation in the presence of nerve growth factor (50 ng/ml) andlow serum (0.5% fetal bovine serum, 1% horse serum) for 8 daysin the absence of doxycycline. Cells were cultured in serum-freemedia for 24 h prior to harvesting to reduce protein contami-nation for mass spectrometric analysis.Kinase Inhibitor Studies—Kinase inhibitors U0126 (10 and

50 "M, Cell Signaling) and roscovitine (25, 100, and 200 "M,Calbiochem) were solubilized in Me2SO (Sigma) and added tocells 24 h after transfection for 24 h. Cells were treated withMe2SO as a control. Similarly, lithium chloride (5, 25 mM) wasadded to cells with sodium chloride utilized as a control. Cellswere harvested by scraping and centrifugation at 500 " g.Immunoprecipitation andWestern blot analysis on cell lysateswere carried out as described below.Site-directed Mutagenesis—Site-directed mutagenesis of Htt

constructs was performed using the QuikChange kit (Strat-agene) and the following primers: serine at amino acid 1201wasconverted to an alanine, forward, 5#-CAAGCATCTGTACC-GTTGGCTCCCAAGAAAGGCAGTGAGG-3#, reverse, 5#-CCTCACTGCCTTTCTTGGGAGCCAACGGTACAGATG-

FIGURE 1. Schematic of Htt structure with four HEAT repeat domains (HH) with the indicated number of HEAT repeats and the caspase/calpain domainof Htt. Within the sequence we indicated regions of a putative aspartic endopeptidase site, unstructured proteolytic susceptibility domains (caspase/calpain),and a nuclear export site (NES). The red arrows indicate phosphorylation sites identified in these studies.

Mass Spectrometry of Huntingtin

AUGUST 18, 2006 • VOLUME 281 • NUMBER 33 JOURNAL OF BIOLOGICAL CHEMISTRY 23687



http://www.jbc.org

CTTG-3#; serine at amino acid 536 was converted to asparticacid, forward 5#-GGATATCTTGAGCCACAGCGACAGCC-AGGTCAGCGCCGTCC-3#, reverse, 5#-GGACGGCGCTGA-CCTGGCTGTCGCTGTGGCTCAAGATATCC-3#; serine atamino acid 535 was converted to aspartic acid, forward, 5#-GGAGGATATCTTGAGCCACGACTCCAGCCAGGTCAG-CGCCG-3#, reverse, 5#-CGGCGCTGACCTGGCTGGAGTC-GTGGCTCAAGATATCCTCC-3#; and serine at amino acid533 was converted to an aspartic acid, forward, 5#-GGATGA-GGAGGATATCTTGGACCACAGCTCCAGCCAGGTCAG-3#, reverse, 5#-CTGACCTGGCTGGAGCTGTGGTCCAAG-ATATCCTCCTCATCC-3#. PCRwas performed using 50 ng ofDNA, 5 "l of 10" Pfu buffer (Stratagene), 0.2 mM dNTPs(Roche Molecular Biochemicals), 125 ng each of forward andreverse primers (Integrated DNA Technologies), 5% Me2SO(Sigma), and 1 "l pf Pfu polymerase (Stratagene) at 96 °C for 1min, 18 cycles at 96 °C for 50 s, 55 °C for 1min, and 68 °C for 24or 45min, and 68 °C for 7 or 10min. PlasmidswereDpnI (Strat-agene)-treated, transformed into XL1-Blue Supercompetentcells (Stratagene), and purified using the Qiagen Plasmid MiniKit. Mutations, CAG repeat length, and construct integritywere confirmed by DNA sequencing. Mutated constructs ofdifferent amino acid length were generated (Htt residues1–1212 and full-length Htt including a myc tag), and expressedas described above.Immunoprecipitation andWestern Blot Analysis—Harvested

cells were lysed with M-PER (Mammalian Protein ExtractionReagent, Pierce) or RIPA lysis buffer (50mMTris, pH 8, 150mMNaCl, 0.1% SDS, 1% SDOC, 1% Nonidet P-40) with proteaseinhibitors (Mini Complete, Roche) and phosphatase inhibitors(1 mM NaF, 1 mM Na3VO4). Lysates were sonicated and thenspun to remove debris (16,000" g, 20min). Protein concentra-tion was determined with the BCA Protein Assay kit (Pierce).Htt was immunoprecipitated overnight from cell lysates (500–2000"g) using the ProFoundMammalianC-mycTag IP/Co-IPkit (Pierce) following the manufacturer’s protocol with a finalwash in 10 mM Tris, pH 7.4, to remove excess salt and elutionProtocol 2 with 25 "l of non-reducing sample buffer (0.3 MTris-HCl, pH 6.8, 5% SDS, 50% glycerol, dye). Immunoprecipi-tated sample (3 "l for Western blot analysis or 22 "l for massspectrometric analysis) was resolved by SDS-PAGE on aNuPAGE 4–12% BisTris gel (Invitrogen) in MOPS runningbuffer (Invitrogen) for 1.5 h at 200 V. ForWestern blot analysis,proteins were transferred to polyvinylidene difluoride (Bio-Rad) or Optitran BA-S nitrocellulose (Schleicher & Schuell)typically for 14 h at 20 volts at 4 °C.Membraneswere blocked in3% bovine serum albumin (Sigma) (for serine phospho-specificantibodies) or 5% milk in Tris-buffered saline Tween-20 andprobed overnight with polyclonal or monoclonal N-terminalHtt BKP1 (1:500), polyclonal c-myc (Upstate, 1:1000), ormono-clonal anti-phospho-Ser-Pro (16B4) and anti-phospho-Ser(1C8) antibody (Chemicon, 1:100). Secondary anti-rabbit oranti-mouse antibody (1:3000, Amersham Biosciences/GEHealthcare) was applied for 1 h at room temperature. ECL(Amersham Biosciences/GE Healthcare) was used for detec-tion. For mass spectrometric analysis, SDS gels were fixed with10% methanol, 7% acetic acid for 30 min and then stained with

Sypro Ruby (Molecular Probes/Invitrogen) followed bydestaining in 10% methanol, 7% acetic acid.In-gel Proteolytic Digestion of Proteins—Proteins were

digested manually to maximize sensitivity and efficiency. Pro-tein bands were destained and dehydrated with ACN. Subse-quently, proteins were reduced with 10 mM dithiothreitol in 25mM NH4HCO3 at 56 °C for 1 h and alkylated with 55 mMiodoacetamide in 25mMNH4HCO3 at room temperature for 45min. For chymotrypsin digestion, 4 M guanidine hydrochloride,25 mMNH4HCO3 was used in the reduction step to fully digestHtt in the gel. Samples were incubated overnight with trypsin(125 ng, 37 °C), chymotrypsin (200 ng, 25 °C), or Asp-N (100–200 ng, 37 °C). The resulting proteolytic peptides were sub-jected to aqueous (100 "l H2O, sonication, 10 min) and hydro-phobic extraction (50 "l of 50% ACN, 5% formic acid), andanalyzed by mass spectrometry after concentration under vac-uum to a 10–15-"l final volume.Mass Spectrometry—Mass spectra of digested protein gel

bands were obtained by matrix-assisted laser desorption ioni-zation time-of-flight mass spectrometry on a Voyager DESTRplus instrument (Applied Biosystems) and proteolytic peptideextracts were analyzed by reverse-phase nano-HPLC-MS/MSwith anUltimateHPLC (Dionex) connected to aQSTARPulsarI quadrupole orthogonal TOF mass spectrometer (MDS Sciex)as previously described (19, 20). In addition, peptides obtainedfrom Htt were analyzed by vacuum MALDI-MS, MS/MS, andMS3 on a vMALDI-LTQTM linear ion trap (Thermo Electron,San Jose, CA). For these experiments, samples were analyzedfirst by MS/MS for the neutral loss of phosphoric acid (98 Da)(17). Peptides showing this neutral loss were interrogated fur-ther byMS3 by selecting the (M $ H % 98)$ peak as precursor.For further experimental detail using vMALDI-MSn for phos-phopeptide analysis see the supplementary materials and pre-vious reports (17, 21).Phosphatase Treatment—ProteolyzedHtt samples were sub-

jected to phosphatase treatment. Calf intestine alkaline phos-phatase (1"l, NewEngland Biolabs) in 50mMNH4CO3, pH 8.0,was added on top of the dried analyte/matrix MALDI spot for1 h at 37 °C as described in Ref. 22. After incubation, 1 "l ofmatrix was added to recrystallize the sample. Alternatively, ali-quots of the Htt digestions were incubated in solution with 1"lof alkaline phosphatase for 1–2 h in the presence of NE Buffer 3(50 mM Tris-HCl, 100 mM NaCl, 10 mM MgCl2, 1 mM dithio-threitol, pH 7.9). Samples were subsequently analyzed byMALDI mass spectrometry.Bioinformatics and Protein Data Base Searches—Mass spec-

trometric data were analyzed with the bioinformatics data basesystem RADARS (Genomic Solutions) (23), Mascot (Ma-trix Sciences, London) (24), and Sequest (25). Routinely,MALDI-MSdatawere analyzedwith RADARSusing the searchengine ProFound for PMF, matching against peptides fromknown protein sequences as previously described (20). For ESI-MS/MS data sets, tandem mass spectra were submitted to thein-house licensed data base search engine Mascot (version 2.1)(Matrix Sciences). A custom-designed protein data base forhuman full-length Htt protein was constructed and incorpo-rated intoMascot that could be searchedmore exhaustively forextended sequence coverage and identification of a more com-





http://www.jbc.org

plete set of posttranslational modifications. To extract andcompile peak lists from ESI-MS data generated during an LC-MS/MS run, Mascot Distiller (1.1.1) and Distiller MDRO Soft-ware Developer’s Package (Matrix Sciences) were used in con-junction with an in-house Java program, “MS-Assign” (20).Therefore the generated ESI-MS peak lists were submitted totheMascot search engine for PMFanalysis at amass accuracy of&50 ppm. MALDI-MSn data were searched using Sequest (fordetails see Supplement). A Java program was developedin-house, “SeqDisp,” to graphically display the observed proteinsequence coverage of peptides obtained fromESI-MS/MS, ESI-MS, and MALDI-MS data.Toxicity Measurements—Caspase activity was measured

using theApoAlert Caspase-3 Fluorescent Assay Kit (Clontech,Palo Alto, CA) as described in our previous work (15).

RESULTS

Htt Is Constitutively Phosphorylated—Htt is a challenging pro-tein to analyze for posttranslational modifications with a molecu-lar mass of !350 kDa (3144 amino acid residues) and long polyQstretches that are susceptible to aggregation. Furthermore, humanfull-length Htt contains 307 serine residues, 170 threonine resi-dues, and 62 tyrosine residues, many of which have predictedkinase consensus sequencemotifs (Table 1).

In the present study, we expressed myc-tagged human full-length Htt constructs containing 23 and 148 polyQ repeats in293Tor PC12 cells.Htt23QorHtt148Qcell lysateswere immu-noprecipitated using a myc tag antibody and subsequently sub-jected to one-dimensional SDS-PAGEorWestern blot analysis.Western blots of immunoprecipitated Htt23Q and Htt148Qwere probed with different antibodies as shown in Fig. 2. TheN-terminal Htt antibody BKP1 revealed immunoreactivity forHtt23Q and Htt148Q migrating at a molecular mass corre-sponding to the theoretical molecularmass of full-lengthHtt of!350 kDa (Fig. 2A). In addition, an anti-phospho-Ser-Pro anti-body (16B4) was used that specifically recognizes phosphoryl-ated serines that are adjacent to a proline residue. Htt23Q andHtt148Qwere immunoreactive to this antibody (Fig. 2B, lanes 2and 3) indicating that Htt was phosphorylated by proline-di-rected serine kinases. Htt23Q andHtt148Qwere immunoreac-tive to an anti-phospho-Ser (1C8) antibody as well. This is amonoclonal antibody immunoreactive to a broad range ofserine-phosphorylated proteins, preferring positively chargedamino acids directly neighboring the phosphoserine residues(Fig. 2C, lanes 2 and 3). These results suggest that we were ableto isolate full-length humanHtt that is endogenously phospho-rylated by several different kinases during protein expression

TABLE 1Predicted phosphorylation sites in human HttSee Yaffe et al. (31) for predictions.

Sequence Position Score Prediction percentile Kinasea

Basophilic serine/threonine kinase group (Baso_ST_kin)GGRSRSGSIVELIAG Ser-421 0.3742 0.030% AKT1NLKSSSPTIRRTAAG Thr-267 0.3950 0.167% PRKCAAHLLKNMSHCRQPSD Ser-636 0.4076 0.126% PRKCMLDRFRLSTMQDSLSP Thr-2068 0.4476 0.104% AKT1TSTTRALTFGCCEAL Thr-1024 0.4817 0.180% AKT1

Acidophilic serine/threonine kinase group (Acid_ST_kin)LVMEQEESPPEEDTE Ser-2489 0.3843 0.130% CSNK2BASSPPTSPVNSRKH Ser-2657 0.4047 0.043% GSK3AKAALPSLTNPPSLSP Thr-1175 0.4633 0.146% GSK3AMQDSLSPSPPVSSHP Ser-2076 0.4759 0.184% GSK3A

Proline-dependent serine/threonine kinase group (Pro_ST-kin)EVVAAPGSPYHRLLT Ser-3126 0.3267 0.054% CDK5LTNPPSLSPIRRKGK Ser-1181 0.3343 0.062% CDK5WWAEVQQTPKRHLSL Thr-1859 0.3483 0.080% CDK5EVVAAPGSPYHRLLT Ser-3126 0.3531 0.034% CDC2EQASVPLSPKKGSEA Ser-1201 0.3718 0.122% CDK5LTNPPSLSPIRRKGK Ser-1181 0.3766 0.054% CDC2QPGEQLLSPERRTNT Ser-2330 0.3997 0.194% CDK5ADAPAPSSPPTSPVN Ser-2653 0.4314 0.124% EPHB2EQASVPLSPKKGSEA Ser-1201 0.4438 0.195% CDC2STMQDSLSPSPPVSS Ser-2074 0.4486 0.171% EPHB2

Phospho-serine/threonine binding group (sST_bind)ESGGRSRSGSIVELI Ser-419 0.4473 0.712% YWHAZVTTSKSSSLGSFYHL Ser-1228 0.4544 0.777% YWHAZEFIYRINTLGWTSRT Thr-2457 0.4575 0.807% YWHAZ

Phosphotyrosine binding group (Y_kin)LDGTDNQYLGLQIGO Tyr-592 0.4056 0.356% FGRGALHGVLYVLECDLL Tyr-2790 0.4495 0.972% ITKESFFSKLYKVPLDTT Tyr-737 0.4160 0.293% CRKPRLQLELYKEIKKNG Tyr-173 0.4587 0.546% FYN

Protein kinase A, PKC #TRFGRKLSIIRGIVE Ser-2550 0.3627 0.258% PRKACGQQTPKRHSLSSTKLL Ser-1864 0.4273 0.557% PRKACGAPKSLRSSWASEEEA Ser-1110 0.4311 0.569% PRKCEAALGMLTCMYTGKE Thr-2926 0.4478 0.366% PRKCM

a Kinase sites searched and abbreviations: AKT1, Akt kinase; PRKCA, protein kinase !/$/%; PRKCM, protein kinase "; CSNK2B, casein kinase 2; GSK3A, glycogen synthasekinase 3; CDK5, Cdk5 kinase; CDC2, Cdc2 kinase; EPHB2, ERK1 kinase; YWHAZ, 14-3-3 mode; FRG, FGR kinase; ITK, Itk Src homology domain 3; CRK, Crk Src homologydomain 3; FYN, Fyn Src homology domain 2; PRKACG, protein kinase A; PRKCE, protein kinase #.





http://www.jbc.org

and these posttranslational modifications are constitutivelypresent in Htt.Htt Is Constitutively Phosphorylated by MEK1 Kinases—Ki-

nase inhibitor studies were performed to determine whichgroup of kinases might be involved in Htt phosphorylation.293T cells expressing Htt23Q and Htt148Q were treated withkinase inhibitors U0126 (MEK1/2 inhibitor), roscovitine(CDK5 inhibitor), and lithium chloride (GSK3 inhibitor).MEK1/2 inhibitor, U0126, decreased the overall level of Httphosphorylation as detected by Western blots probed with ananti-phospho-Ser-Pro antibody (Fig. 3). Studies with roscovi-tine and lithium chloride, also using Western blot analysis and

phosphoserine antibodies, were inconclusive. This is mostlikely due to the presence of several other abundant phospho-rylation sites that are not affected by these kinase inhibitors(data not shown).Immunopurification ofHtt forMass Spectrometry—Todeter-

mine the unique sites of phosphorylation within Htt by massspectrometry, we optimized the immunopurification of Httusing several different cell lysis buffers (Fig. 3; see “Experimen-tal Procedures”). Htt was immunoprecipitated from relativelylarge amounts of starting material (500–2000 "g of cell lysateper immunoprecipitation) from either 293T or PC12 cells.After release of Htt from the myc tag antibody, the proteinswere separated by one-dimensional SDS-PAGE and stainedwith the fluorescent dye Sypro Ruby (Fig. 4). Microgram quan-tities of Htt were obtained, sufficient for our analysis of post-translational modifications by mass spectrometry.Mass Spectrometric Characterization of Full-length Htt—To

determine the sites of phosphorylation in Htt, we digested theHtt protein with trypsin, chymotrypsin, or Asp-N, and sub-jected the resulting peptides to both MALDI-MS and HPLC-ESI-MS/MS analysis. Mass spectrometry can give sequencedata from the Sypro Ruby-stained band or gel spot. The proteinis digested with a protease in the gel, peptides are eluted andintroduced into themass spectrometer. Themass spectrometerdetermines the mass of the peptides and the sequence (by col-lisionally induced dissociation). From themasses of the peptidefragments, sequence data are determined by comparison withknown sequences. Posttranslational modifications of thesepeptides can be identified by shifts in masses (i.e. 80 kDa forphosphorylation). As shown in Fig. 5A, protonated molecularions, [M $ H]$, for 31 peptides were detected by MALDI-MS(i.e. PMF) for Htt23Q after tryptic digestion, yielding an overallprotein sequence coverage of 15%. Chymotryptic digestion ofHtt23Q generated 68 peptides yielding an overall proteinsequence coverage of 27% (Fig. 5B). We also digested Htt23Qwith the protease Asp-N that cleaves N-terminal to Asp andGlu residues, and MALDI-MS analysis resulted in 13%

FIGURE 2. Immunoprecipitation and Western blot analysis of Htt withanti-Htt and phospho-specific antibodies. Myc-tagged full-length Httconstructs (Htt23Q and Htt148Q) were expressed in 293T cells. Westernblot of immunoprecipitated Htt23Q and Htt148Q probed with (A) poly-clonal N-terminal Htt antibody BKP1, (B) monoclonal anti-phospho-Ser-Pro 16B4, and (C) monoclonal anti-phospho-Ser 1C8 antibody.

FIGURE 3. MEK1/2 inhibitor (U0126) decreases Htt phosphorylation. 293Ttransfected cells were treated with U0126 kinase inhibitor (10 "M, 24 h). West-ern blot of input cell lysates was probed with a N-terminal anti-Htt antibody(BKP1). Western blot of anti-myc immunoprecipitated (IP) Htt was probedwith anti-Htt BKP1 antibody to show comparable protein expression levelsfor the different kinase inhibitor treatments, and a decrease in phosphoryla-tion was detected with anti-phospho-Ser-Pro antibody (16B4) in the presenceof U0126.





http://www.jbc.org

sequence coverage (data not shown). Proteolytic samples werethen subjected to an on-line separation by nano-HPLCreversed-phase C18 chromatography directly coupled to ahybrid quadrupole time-of-flightmass spectrometer (QSTAR).Peptideswere selected forMS/MS and fragmented by collision-induced dissociation (CID) providing peptide sequence infor-mation. Analysis of Htt using nano-HPLC-ESI-MS/MS andMALDI-MS gave good sequence coverage of Htt. A completelist of all Htt peptides that were sequenced and confirmed bymass spectrometry is provided in supplemental Table S1 (forgraphic display of Htt sequence coverage see supplementalFig. S1).Mass Spectrometric Identification of Phosphorylation Sites in

Full-lengthHttUsingNano-HPLC-ESI-MS/MS—Todeterminespecific sites of phosphorylation within Htt, we searched theESI-MS/MS spectra with Mascot using a custom-designed Httdata base. With this analysis, we mapped six novel phosphoryl-ation sites of Htt. The results are summarized in Table 2 andrepresent phosphorylations identified by bothMALDI-MS andnano-HPLC-ESI-MS/MS (see Table S2 for details). Other typesof Htt amino acid modifications, including oxidation anddeamidation, are noted in Table S3.Asarepresentativeexampleofhowweassignedspecificphos-

phorylation sites inHtt, Fig. 6 shows the tandemmass spectrumof monophosphorylated peptide EKEPGEQASVPLpS1201PK(residues 1189–1203) (the bold indicates amino acids that arepotential sites of phosphorylation), generated from trypticdigestion of Htt23Q. A precursor ion [M $ 2H]2$ at m/z

838.412$ (M ' 1674.82 Da) was selected for MS/MS, and thefragmentation pattern was indicative of the presence of onephosphate group on serine residue 1201. A nearly completeseries of y-ions and several b-ion fragments were observed,including several yn-98 ions that result from the neutral loss ofphosphoric acid (-H3PO4). The y-ion pairs observed at m/z411.2/313.2 (y3/y3-98) and 621.3/523.3 (y5/y5-98) are indicativeof the phosphorylation site at Ser-1201 as these fragment ionsare shifted by 80 Da (mass of one phosphate group) comparedwith the corresponding fragments generated from the non-phosphorylated peptide atm/z 331.2 (y3) andm/z 541.3 (y5). Inaddition, other fragment ions and y-ion pairs were observedthat could be easily assigned according to the tryptic peptideEKEPGEQASVPLpS1201PK, bearing the phosphorylation siteon Ser-1201 (Fig. 6). A phosphorylation level of (50% of thecorresponding non-phosphorylated peptides (residues 1189–1203 and 1187–1203) was estimated based on relative peakabundance as observed by HPLC-ESI-MS. These estimates arenot quantitative but do indicate that Ser-1201 is phosphoryla-ted at high levels in Htt. Several other novel phosphorylationsites were determined in a similar manner using nano-LC-ESI-MS/MS (Table 2).In addition, we mutated a number of the identified phos-

phorylation sites and recorded the mass spectra. As a repre-sentative example, the ESI-MS/MS spectrum of peptideEPGEQASVPLA1201PK (residues 1191–1203) obtained aftertryptic digestion of Htt23Q demonstrated the successful site-directed mutagenesis from Ser-1201 to Ala-1201. The molecu-lar ion [M $ 2H]2$ at m/z 661.852$ (M ' 1321.68 Da) wasselected for CID (data not shown). The corresponding phos-phopeptide containing Ser(P)-1201 previously observed inwild-typeHttwas not detected in thisHtt23QS1201A (data notshown). In addition, we identified Ser-1181 as a unique phos-phorylation site after tandemESI-MS/MSanalysis of the trypticpeptide A1169ALPSLTNPPSLpS1181PIR (Fig. 9).Mass Spectrometric Identification of Phosphorylation Sites in

Full-length Htt using MALDI-MS and Nano-HPLC-ESI-MS/MS—Phosphopeptides were also observed by MALDI-MS andwere decreased upon phosphatase treatment. A representativeMALDI mass spectrum of Htt23Q digested with Asp-N proteaseis shown (Fig. 7). The protonated molecular ions of peptideDAPAPSS2653PPTS2657PVNSRKHRAGV (residues 2647–2668)were detected in its non-phosphorylated form ([M $ H]$ atm/z2228.4) and in its mono- and diphosphorylated forms ([M$H]$at m/z 2308.4 and 2388.3, respectively) (Fig. 7A). The observedmass shifts of 80 and 160 Da correspond to the addition of one ortwo phosphate groups to the corresponding non-phosphorylatedpeptide. Alkaline phosphatase treatment on the MALDI targetresulted in the disappearance of the diphosphorylated peptidewith [M$H]$ atm/z 2388.4 due to dephosphorylation (Fig. 7B).The monophosphorylated peptide with [M $ H]$ atm/z 2308.4decreased significantly in intensity, whereas the non-phosphoryl-ated peptide with [M $ H]$ at m/z 2228.4 increased. Observedpeptides at m/z 2398.5 (residues 1527–1549, contained oxidizedmethionine) and atm/z 2444.7 (residues 110–131 or 1394–1415)were not affected by the phosphatase treatment and served aslandmarks to compare relative peak abundance.

FIGURE 4. Immunoprecipitation and one-dimensional SDS-PAGE of Htt.Myc-tagged full-length Htt expressed in 293T or PC12 cells was immunopre-cipitated from cellular lysates using the Pierce ProFound Mammalian C-myctag IP/Co-IP kit in M-PER buffer. Samples were separated by one-dimensionalSDS-PAGE on 4 –12% BisTris gels. Gels were stained with fluorescent SyproRuby protein gel stain (Invitrogen). The lower panel demonstrates that theexpanded form of Htt immunoprecipitates at higher levels in a buffer com-posed of 20% RIPA and 80% M-PER (labeled as RIPA). Gel bands at !50 kDacorrespond to immunoglobulin. Protein samples were analyzed by massspectrometry after in-gel digestion. Arrows indicate identified Htt protein.





http://www.jbc.org

To determine the site of phosphorylation within this spe-cific peptide we performed online nano-HPLC-MS/MS.Tandem mass spectra (ESI-MS/MS) were obtained ofthe mono- and diphosphorylated forms of peptide DAPAPSS-2653 PPTS2657PVNSRKHRAGV (Fig. 8, A and B). The frag-mentation pattern of the monophosphorylated peptide

proved that the unique site of phosphorylation was on Ser-2653 (Fig. 8A). Several y-type ions were observed, such asions at m/z 610.92$, 654.42$, 704.92$, 753.42$, and 802.02$

(y11–y15), with no evidence for phosphorylation. Fragmentions containing Ser(P)-2653 yielded characteristic ion pairsof yn/yn-98 ions due to the neutral loss of phosphoric acid

(-H3PO4) from the yn ion; charac-teristic ions were observed at m/z557.93$ and 836.52$ (y16-98),652.03$ and 977.52$/928.52$ (y18/y18-98), and 708.03$/1012.52$

(y20/y20-98). For the correspond-ing diphosphorylated peptide, sev-eral identical ions were observed,such as the b2-, b3-, and b4-ions,and also the y11-ion (Fig. 8B). Ay15/y15-98 ion pair was observed atm/z 842.02$/793.02, which indi-cated a new phosphorylation siteon either Ser-2657 or Thr-2656,and this was not observed in theMS/MS spectrum for the mono-phosphorylated peptide. Furtherfragment ions at m/z 584.63$

(y16-98) and 1017.52$/968.52$

(y18/y18-98) showed the secondphosphate group to be onSer-2653 (Fig. 8B).Several other novel phosphoryla-

ted peptides generated after proteo-lytic digestion of Htt protein wereidentified by tandemmass spectrom-etry as summarized in Table 2. AnESI-MS/MS spectrum was recordedfor monophosphorylated peptideDSLSPpS2076PPVSSHPL (residues2071–2084) as shown in supplemen-tary Fig. S2. The molecular ion [M $2H]2$ atm/z 750.382$ (M' 1498.74Da) was selected for CID, and a seriesof y-ions, (yn/yn-P) ion pairs and sev-

FIGURE 5. MALDI-TOF mass spectra of immunoprecipitated Htt23Q. The MALDI-MS TOF spectra displaymolecular ions [MH]$ of peptides obtained after in-gel proteolytic digestion of immunoprecipitated Htt23Q. A,Htt PMF resulting from tryptic digestion. An overall protein sequence coverage of 15% was obtained. B, HttPMF resulting from chymotryptic digestion. An overall protein sequence coverage of 27% was obtained.

TABLE 2Huntingtin phosphorylation sites identified by mass spectrometryFurther mass spectrometric details are in Table S2.

Observed proteolytic phosphopeptidea Residue Kinase Mascot scoreb Xcorr valuec

TAAKEESGGRSRSGpSIVEL 421d AKT-1 14Phospho-(DILSHSSSQVSAVPS) 533/5/6e 41AALPSLTNPPSLpSPIR 1181 CDC2/CDK5 32 2.08 (MS3)EKEPGEQASVPLpSPK 1201 CDC2/CDK5 58 3.29 (MS3)GKEKEPGEQASVPLpSPK 1201 CDC2/CDK5 13 4.23 (MS3)EQASVPLpSPKKGSEASAASRQS 1201 CDC2/CDK5 51DSLSPpSPPVSSHPL 2076 ERK1 22DAPAPSpSPPTSPVNSRKHRAGV 2653 ERK1 46 1.52 (MS3)DAPAPSpSPPTpSPVNSRKHRAGV 2653 ERK1 44 2.37 (MS3)

2657 f Pro-dir.a All phosphopeptides were confirmed by tandem mass spectrometry.b Mascot scores of individual peptides were observed by nano-LC ESI-MS/MS and were determined as described in Perkins et al. (24).c SEQUESTXcorr value of individual peptides thatwere observed byMALDI-MS (MS2/MS3)were determined as described in Eng et al. (25). The abundant neutral loss fragmention (MH$-98) observed in MS2 spectra of the phosphopeptides was subjected to further fragmentation (MS3).

d Phosphorylation site: Ser-413, Ser-417, Ser-419, or Ser-421 (based on ESI-MS/MS); phospho-Ser-421 was previously identified as P-site in Htt.e Phosphorylation site: Ser-533, Ser-535, or Ser-536 (based on ESI-MS/MS).f Phosphorylation site: Ser-2657 or Thr-2656, with higher prediction score for Ser-2657.





http://www.jbc.org

eralb-fragment ionswereobserved that indicated thephosphoryl-ation of Ser-2076 (see supplementary Fig. S2).We also identified the phosphorylated peptide (T407AAK-

EESGGRSRSGpS421IVEL425) by tandem mass spectrometrycontaining a single phosphorylation site at Ser-421. Themolec-ular ion, [M $ 3H]3$ at m/z 672.03$ (M ' 2012.98 Da) wasselected for CID (see supplementary Fig. S3). Ser(P)-421 wasthe first phosphorylation site of Htt described and was identi-fied using biochemical and molecular methods (6, 7).

Hypothesis Driven Investigation of Phosphorylation Sites inHttUsingNeutral Loss ScanningMass Spectrometry—To inves-tigate Htt for additional phosphorylation sites we adapted ahypothesis driven multiple stage mass spectrometric approachoriginally described by Chang et al. (17). Analysis was per-formed on a FinniganTM LTQTM vMALDI quadrupole linearion trap. Briefly, peptides with highly predicted kinase consen-sus sites were examined using vMALDI-MS/MS for the pres-ence (or absence) of phosphoserine and phosphothreonineresidues based on the preferential neutral loss of phosphoricacid (%98 Da). Subsequent MS3 analysis of the correspond-ing (M $ H % 98)$ peaks provided sequence data to confirmor reject the hypothetical assignments. The phosphorylationsites that were identified by ESI-MS/MS in this study werealso investigated as positive controls. An example is dis-played in Fig. 9, A and B, showing tandem mass spectra ofpeptideAALPSLTNPPSLpS1181PIR obtained fromHtt23Q andHtt148Q, respectively. In both cases, a loss of 98 Da from the

FIGURE 6. ESI-MS/MS spectrum of monophosphorylated peptideEKEPGEQASVPLpS1201PK (residues 1189–1203) obtained after trypticdigestion of Htt23Q. The molecular ion [M $ 2H]2$ at m/z 838.412$ (M '1674.82 Da) was selected for CID. Series of y fragment ions and several b fragmentions were observed; several y-ions featured loss of 98 Da due to loss of phos-phoric acid.

FIGURE 7. MALDI mass spectra displaying molecular ions, [M ! H]!

of PMF obtained after Asp-N digestions of Htt23Q. A, peptideDAPAPSS2653PPTS2657PVNSRKHRAGV (residues 2647–2668) present in itsnon-phosphorylated form, [M $ H]$ at m/z 2228.4, and in its mono-and diphosphorylated forms, [M $ H]$ at m/z 2308.4 for DAPAPS-pS2653PPTSPVNSRKHRAGV and [M $ H]$ at m/z 2388.3 for DAPAPS-pS2653PPTpS2657PVNSRKHRAGV. B, the same sample after incubation withalkaline phosphatase for 1 h.

FIGURE 8. ESI-MS/MS tandem mass spectra of mono- and diphosphorylatedpeptides obtained after Asp-N digestion of immunoprecipitated Htt23Q. A,monophosphorylated peptide DAPAPSpS2653PPTSPVNSRKHRAGV (residues2647–2668). The molecular ion [M$3H]3$ at m/z 770.073$ (M'2307.20 Da) wasselected for CID. Series of y fragment ions and several b fragment ions wereobserved; several y ions featured loss of %98 Da due to loss of phosphoric acid,proving serine phosphorylation on residue Ser-2653. B, diphosphorylated pep-tide DAPAPSpS2653PPTpS2657PVNSRKHRAGV (residues 2647–2668). The molec-ular ion [M $ 3H]3$ at m/z 796.743$ (M ' 2387.20 Da) was selected for CID. Thepeptide fragmentation pattern proves one phosphorylation site at Ser-2653, andsuggests the second phosphorylation site to be either Ser-2657 or Thr-2656, withSer-2657 being the more likely site according to predictions.





http://www.jbc.org

[M $ H]$ precursor ion at m/z 1713.7$ (M ' 1712.7) wasobserved to yield an abundant neutral loss of 98 Da at m/z1615.8. This peptide was selected for MS3 analysis (Fig. 9C),revealing a characteristic set of sequence ions (y8/y8-17, y9–11,and y13) that confirmed the phosphorylation site as Ser(P)-1181. Supplementary Figs. S5–S10 show additional vMALDIMS/MS and MS3 spectra of peptides containing the phospho-rylation sites Ser(P)-2653, Ser(P)-2657, Ser(P)-1181, Ser(P)-1201, and Ser(P)-421, confirming our results.Using the hypothesis driven approach, we also investigated a

predicted set of Htt peptides containing Ser/Thr sites that hadbeen identified as potential kinase consensus motifs with highprediction scores (see Table 1 and Table S4). However, no newphosphorylation sites were identified from these additionalanalyses.Mass Spectrometric Identification of a Phosphorylation Site

in the Proteolytic Susceptibility Domain of Htt—We detectedanother novel monophosphorylated peptide by LC-MS/MS,phospho-(DILS533HS535S536S537QVSAVPS), corresponding toHtt residues 530–544 obtained from an Asp-N protease digest.As shown in supplementary Fig. S4, a tandem mass spectrumwas recorded for themonophosphorylated peptidewith a [M$2H]2$ atm/z 797.402$ (M' 1592.78Da), 80Dahigher than thenon-phosphorylated peptide counterpart at m/z 757.402$

(M ' 1512.78 Da). This particular phosphorylation site lies inthe proteolytic susceptibility domain of Htt, and our previouswork has identified amino acid 536 as a site of calpain cleavagein Htt (15). The peptide fragmentation patterns could not

uniquely determine which of the serines is phosphorylated(Ser-533, Ser-535, or Ser-536).Biological Significance of Phosphorylation at Amino Acid 536

ofHtt—Oneof themore interesting phosphorylation sites iden-tified in our studies lies in the putative proteolytic susceptibilitydomain of Htt (PEST sequence) (Fig. 1). We have previouslyshown that calpains cleave Htt to produce N-terminal cleavageproducts (13, 15). In those studies, we found that calpaincleaves Htt15Q-(1–1212) to yield a 72-kDa N-terminal prod-uct, and the site of cleavage was eliminated by mutation at the536-amino acid region ofHtt. Thiswould suggest that the phos-phorylation site identified in our current studies at amino acid536 controls proteolysis of Htt by calpains. To test this, wemutated Ser-536 to aspartic acid to mimic phosphorylation(Fig. 10,A andB).We found that Htt15Q orHtt138Q-(1–1212)S536D, when expressed in 293T cells, was resistant to calpaincleavage. Production of the 72-kDa calpain-derived Htt frag-ment (92 kDa for Htt138Q-(1–1212)) was eliminated (Fig.10A). Mutation of serine to aspartic acid at amino acids 533 or535 did not alter cleavage (Fig. 10A). The remaining immuno-reactivity is due to cleavage ofHtt at amino acid 513 by caspases(data not shown). We also evaluated whether calpain cleavageof polyQ-expanded Htt influenced toxicity through phospho-rylation of this site.We expressed the calpain-resistant Htt15Qor 138Q-(1–1212) S536D constructs in 293T cells and evalu-ated cytotoxicity. As shown in Fig. 10B, mutation of Ser-536 toaspartic acid to mimic phosphorylation reduced cellular toxic-ity as measured by caspase activation and correlated withdecreased proteolysis of mutant Htt.Kinase Consensus Sites for Htt—We have identified six novel

Htt phosphorylation sites in Htt23Q andHtt148Q expressed in293T cells. These include Ser(P)-536, Ser(P)-1181, Ser(P)-1201,Ser(P)-2076, Ser(P)-2653, and Ser(P)-2657 (Table 2). Knowingthe corresponding kinases for these sites would provide tools toinvestigate the cellularmechanisms ofHtt phosphorylation andpotential signaling pathways in more detail. As each kinaseclass recognizes specific protein consensus sequences (26, 27),it is possible to predict the kinases might be responsible for thephosphorylation of the Htt residues that we have identified andsummarized in Table 2. We have used several different web-based phosphorylation prediction programs includingNetPhos(Technical University of Denmark) (28, 29), ScanSite (MIT)(30, 31), PhosphoBase (EMBL), and ELM (EMBL) (32). Thekinase prediction results are summarized in supplementaryTable S4. Proline-directed kinases (CDK5/CDC2 and ERK1)are predicted to phosphorylate Htt. Phosphorylation sites atSer-1181 and Ser-1201 are likely substrates for cyclin-depend-ent kinases p35/CDK5 and/or p34/CDC2 (top 0.2% ScanSitepercentile). The phosphorylation site at Ser-2653 is predictedas a highly likely ERK1 substrate (ERK1 is a member of themitogen-activated protein kinase pathway) (33). Phosphoryla-tion sites at Ser-2657 and Ser-2076 are predicted as ERK1 sub-strates using ScanSite prediction programs (using lower strin-gency searches) or substrates for other Pro-directed kinases (forfurther details regarding predictions see supplementary TableS4). The phosphorylation site at Ser-536, which is discussedabove, does not contain a known kinase consensus site.

FIGURE 9. vMALDI-MS/MS and MS3 identification of phosphorylated pep-tide AALPSLTNPPSLpS1181PIR (residues A-1169 to R-1184) from Htt23Qand Htt148Q for Ser(P)-1181. A, the precursor ion [M $ H]$ at m/z 1713.7$

(M ' 1712.7) was selected for MS/MS analysis from Htt23Q; and B, Htt148Qdigests. In both cases, a strong neutral loss of 98 Da was observed due to lossof phosphoric acid, yielding an abundant fragment ion at m/z 1615.8 (A) and1615.7 (B), respectively. In panel A a shorter collision energy activation time of10 ms (instead of 30 ms) was used. C, the neutral loss fragment ion at m/z1615.8 was isolated and collisionally activated to yield the resulting confirm-atory MS3 spectrum where Ser(P)-1181 is now dehydroalanine.





http://www.jbc.org

DISCUSSION

This study presents the first detailed mass spectrometriccharacterization of expressed Htt including mapping of theprotein sequence by tandemmass spectrometry and PMF. Themajor focus of this work was to apply mass spectrometric tech-niquestoHttproteincharacterizationandidentificationofphos-phorylation sites in Htt. This was accomplished by analyzingmultiple protease digests of Htt byHPLCESI-MS/MS aswell astargeted neutral loss experiments using tandem vMALDI iontrap experiments of potential or hypothetical phosphorylationsites suggested by mass data or kinase prediction algorithms.These newly identified phosphorylation sites may be involvedin the normal function of Htt as well as HD pathogenesis. Ourcurrent study provides a set of tools to investigate the role ofposttranslational modifications of Htt in HD. We were able tomap phosphorylation sites common to both wild-type (23Qrepeat) and mutant (148Q repeat) Htt protein expressed in293T and PC12 cells (Table 2, supplemental Table S5). Giventhe 42% sequence coverage obtained by ESI-MS/MS, and anoverall 88% sequence coverage of Htt additionally consideringPMF (ESI-MS and MALDI-MS) in our studies, it is likely wehave identified the major constitutive phosphorylation sites inHtt. The methods employed in our studies, however, did notquantify the absolute level of the individual phosphorylationsites in Htt.

In addition to the data presentedhere, we also carried out a compre-hensive analysis of two tryptic pep-tide digests of full-length Httafter selective enrichment of phos-phopeptides using immobilizedmetal ion affinity chromatographyas described in the supplementarymaterials (34). Although we wereable to confirm several phosphoryl-ation sites already identified, noadditional sites were detected.Fig. 1 displays the location of the

phosphorylation sites within Httidentified in this study. Htt contains4majorHEATdomains indicated asHH1–HH4, several known proteo-lytic susceptibility domains, and apotential nuclear localization signal.The identified novel phosphoryla-tion sites are distributed throughoutthe protein sequence, i.e. Ser(P)-536in the proteolytic susceptibilitydomain, Ser(P)-1181 and Ser(P)-1201 in proximity to a potential Httnuclear localization signal (35, 36),Ser(P)-2076 within HEAT domain 3(HH3), and Ser(P)-2653 and Ser(P)-2657 overlapping with a predictedC-terminal proteolytic susceptibil-ity domain as well as a predicted cal-cineurin binding motif.Kinase inhibitor studies with

U0126, a MEK1/2 inhibitor, confirmed the phosphorylation ofHtt protein by either MEK1/2 or ERK1 kinases (ERK1 is akinase downstream ofMEK1/2 in theMAP kinase pathway andwould be equally inhibited upon U0126 inhibitor treatment).Htt phosphorylation sites Ser(P)-2076, Ser(P)-2653, andSer(P)-2657 are predicted to be phosphorylated by ERK1 kinaseconsistent with decreased levels of Htt phosphorylation withU0126 treatment (Fig. 3). U0126 treatment did not completelyeliminate phosphorylation of all serines adjacent to prolines,consistent with other types of Htt phosphorylation sites.Ser(P)-1181 and Ser(P)-1201 are likely CDC2/CDK5 kinaseconsensus sites and should not be affected byMEK1/2 inhibitortreatment.It has been suggested that proteolytic cleavage of the

mutant Htt may play an important role in the pathophysiol-ogy of HD. We have previously shown that the toxicity ofcaspase-resistant or calpain-resistant expanded Htt wasmarkedly reduced in transfected cells (15). This suggests thatposttranslational modifications that prevent the productionof toxic polyQ-containing products would be neuroprotec-tive. In the present study we identified a phosphopeptide(D530ILS533HS535S536SQVSAVPS544) by ESI-MS/MS thatis located within an N-terminal proteolytic susceptibilitydomain targeted by caspases and calpains (8, 9, 11, 13, 15,37). This region of Htt is particularly important because it

FIGURE 10. A, Western blot of Htt-(1–1212) expressed in 293T cells probed with a N-terminal Htt antibody.Mutation of Ser-536 to Asp dramatically decreases cleavage, whereas the same mutation at Ser-535 or Ser-533does not. B, Htt138Q-(1–1212) S536D mutation decreases toxicity in 293T cells, whereas the S535D mutationhas no effect (**, p ) 0.005).





http://www.jbc.org

contains both the calpain and the caspase cleavage sites ofHtt. We have recently demonstrated that cleavage at aminoacid Ser-536 is mediated by calpains (15). Functional analysisof this site (Fig. 10) suggests that phosphorylation of this siteblocks cleavage by calpains and modulates toxicity of mu-tant Htt. The site of phosphorylation in the sequenceDILSHSS536SQVSAVPS of Htt does not contain a knownkinase consensus site. Future work will be directed at iden-tifying the kinase responsible for phosphorylating Htt at thissite.Our findings demonstrate that phosphorylation of Htt at

Ser-536 decreases Htt cellular cytotoxicity and the productionof cytotoxic fragments bymodifying a site of calpain cleavage inHtt. Other work on Htt as well as the androgen receptor hasshown that phosphorylation at more distal sites can modulatecleavage. In a related polyglutamine disease, Kennedy disease,we have shown that the phosphorylation of androgen receptormodulates proteolysis and toxicity (38). Our studies found thatphosphorylation at Ser-514 of the androgen receptor, a site dis-tal to the caspase cleavage site at amino acid 154, enhancedcellular toxicity and the production of cytotoxic fragments.Recent work onHtt suggests that CDK5-mediated phosphoryl-ation of Ser-434 reduces cleavage of Htt at a distal caspase-3cleavage site (amino acid 513) (39). A key role for phosphoryl-ation of Ser-776 of ataxin-1 was shown to trigger toxicity inSCA1 in vivo (40, 41). In this case protein interactions andataxin-1 localization are altered by phosphorylation and notcleavage of the protein.In summary, we have mapped the major constitutive phos-

phorylation sites of Htt and identified a critical phosphoryla-tion event modulating Htt cleavage and toxicity. Future studieswill be directed at understanding how these phosphorylationevents affect the function and regulation of Htt and what theirrole is in HD.

Acknowledgments—We thank Thermo Electron Corporation for pro-viding a FinniganTM vMALDI-LTQTM linear ion trap mass spec-trometer for evaluation purposes, and Dr. Rosa Viner for experimen-tal advice. We thank Chip Witt for bioinformatics support and ChrisYoo for experimental support. We thank Michael Hayden for provid-ing BKP1 antibody to Htt.

REFERENCES1. Huntington Disease Collaborative Research Group (1993) Cell 72,

971–9832. Arrasate, M., Mitra, S., Schweitzer, E. S., Segal, M. R., and Finkbeiner, S.

(2004) Nature 431, 805–8103. Xia, J., Lee, D. H., Taylor, J., Vandelft,M., and Truant, R. (2003)Hum.Mol.

Genet. 12, 1393–14034. Kalchman, M. A., Graham, R. K., Xia, G., Koide, H. B., Hodgson, J. G.,

Graham, K. C., Goldberg, Y. P., Gietz, R. D., Pickart, C. M., and Hayden,M. R. (1996) J. Biol. Chem. 271, 19385–19394

5. Steffan, J. S., Agrawal, N., Pallos, J., Rockabrand, E., Trotman, L. C., Slepko,N., Illes, K., Lukacsovich, T., Zhu, Y. Z., Cattaneo, E., Pandolfi, P. P.,Thompson, L. M., and Marsh, J. L. (2004) Science 304, 100–104

6. Humbert, S., Bryson, E. A., Cordelieres, F. P., Connors, N. C., Datta, S. R.,Finkbeiner, S., Greenberg, M. E., and Saudou, F. (2002) Dev. Cell 2,831–837

7. Rangone, H., Poizat, G., Troncoso, J., Ross, C. A., MacDonald, M. E.,Saudou, F., and Humbert, S. (2004) Eur. J. Neurosci. 19, 273–279

8. Goldberg, Y. P., Nicholson, D.W., Rasper, D.M., Kalchman,M. A., Koide,H. B., Graham, R. K., Bromm,M., Kazemi-Esfarjani, P., Thornberry, N. A.,Vaillancourt, J. P., and Hayden, M. R. (1996) Nat. Genet. 13, 442–449

9. Wellington, C. L., Ellerby, L. M., Gutekunst, C. A., Rogers, D., Warby, S.,Graham, R. K., Loubser, O., van Raamsdonk, J., Singaraja, R., Yang, Y. Z.,Gafni, J., Bredesen, D., Hersch, S. M., Leavitt, B. R., Roy, S., Nicholson,D. W., and Hayden, M. R. (2002) J. Neurosci. 22, 7862–7872

10. Ellerby, L. M., Andrusiak, R. L., Wellington, C. L., Hackam, A. S.,Propp, S. S., Wood, J. D., Sharp, A. H., Margolis, R. L., Ross, C. A.,Salvesen, G. S., Hayden, M. R., and Bredesen, D. E. (1999) J. Biol. Chem.274, 8730–8736

11. Wellington, C. L., Singaraja, R., Ellerby, L., Savill, J., Roy, S., Leavitt, B.,Cattaneo, E., Hackam, A., Sharp, A., Thornberry, N., Nicholson, D. W.,Bredesen,D. E., andHayden,M. R. (2000) J. Biol. Chem. 275, 19831–19838

12. Hermel, E., Gafni, J., Propp, S. S., Leavitt, B. R., Wellington, C. L., Young,J. E., Hackam, A. S., Logvinova, A. V., Peel, A. L., Chen, S. F., Hook, V.,Singaraja, R., Krajewski, S., Goldsmith, P. C., Ellerby, H.M., Hayden,M. R.,Bredesen, D. E., and Ellerby, L. M. (2004) Cell Death Differ. 11, 424–438

13. Gafni, J., and Ellerby, L. M. (2002) J. Neurosci. 22, 4842–484914. Kim, Y. J., Yi, Y., Sapp, E., Wang, Y., Cuiffo, B., Kegel, K. B., Qin, Z. H.,

Aronin, N., and DiFiglia, M. (2001) Proc. Natl. Acad. Sci. U. S. A. 98,12784–12789

15. Gafni, J., Hermel, E., Young, J. E., Wellington, C. L., Hayden, M. R., andEllerby, L. M. (2004) J. Biol. Chem. 279, 20211–20220

16. Tanaka, Y., Igarashi, S., Nakamura,M., Gafni, J., Torcassi, C., Schilling, G.,Crippen, D.,Wood, J. D., Sawa, A., Jenkins, N. A., Copeland, N. G., Borch-elt, D. R., and Ellerby, L. M. (2005) Neurobiol. Dis. 21, 381–391

17. Chang, E. J., Archambault, V., McLachlin, D. T., Krutchinsky, A. N., andChait, B. T. (2004) Anal. Chem. 76, 4472–4483

18. Igarashi, S., Morita, H., Bennett, K. M., Tanaka, Y., Engelender, S., Peters,M. F., Cooper, J. K., Wood, J. D., Sawa, A., and Ross, C. A. (2003) Neuro-report. 14, 565–568

19. Schilling, B., Murray, J., Yoo, C. B., Row, R. H., Cusack, M. P., Capaldi,R. A., and Gibson, B. W. (2006) Biochim. Biophys. Acta 1762, 213–222

20. Schilling, B., Bharath, M. M. S., Row, R. H., Murray, J., Cusack, M. P.,Capaldi, R. A., Freed, C. R., Prasad, K. N., Andersen, J. K., and Gibson,B. W. (2005)Mol. Cell Proteomics 4, 84–96

21. Krutchinsky, A. N., Kalkum, M., and Chait, B. T. (2001) Anal. Chem. 73,5066–5077

22. Larsen, M. R., Sorensen, G. L., Fey, S. J., Larsen, P. M., and Roepstorff, P.(2001) Proteomics 1, 223–238

23. Field, H. I., Fenyo, D., and Beavis, R. C. (2002) Proteomics 2, 36–4724. Perkins, D. N., Pappin, D. J., Creasy, D. M., and Cottrell, J. S. (1999) Elec-

trophoresis 20, 3551–356725. Eng, J. K., McCormack, A. L., and Yates, J. R., III (1994) J. Am. Soc. Mass

Spectrom. 5, 976–98926. Kennelly, P. J., and Krebs, E. G. (1991) J. Biol. Chem. 266, 15555–1555827. Pinna, L. A., and Ruzzene, M. (1996) Biochim. Biophys. Acta 1314,

191–22528. Blom, N., Gammeltoft, S., and Brunak, S. (1999) J. Mol. Biol. 294,

1351–136229. Blom, N., Sicheritz-Ponten, T., Gupta, R., Gammeltoft, S., and Brunak, S.

(2004) Proteomics 4, 1633–164930. Obenauer, J. C., Cantley, L. C., and Yaffe, M. B. (2003) Nucleic Acids Res.

31, 3635–364131. Yaffe, M. B., Leparc, G. G., Lai, J., Obata, T., Volinia, S., and Cantley, L. C.

(2001) Nat. Biotechnol. 19, 348–35332. Puntervoll, P., Linding, R., Gemund, C., Chabanis-Davidson, S., Mattings-

dal, M., Cameron, S., Martin, D.M., Ausiello, G., Brannetti, B., Costantini,A., Ferre, F., Maselli, V., Via, A., Cesareni, G., Diella, F., Superti-Furga, G.,Wyrwicz, L., Ramu, C., McGuigan, C., Gudavalli, R., Letunic, I., Bork, P.,Rychlewski, L., Kuster, B., Helmer-Citterich, M., Hunter, W. N., Aasland,R., and Gibson, T. J. (2003) Nucleic Acids Res. 31, 3625–3630

33. Pearson, G., Robinson, F., Beers Gibson, T., Xu, B. E., Karandikar, M.,Berman, K., and Cobb, M. H. (2001) Endocr. Rev. 22, 153–183

34. Corthals, G. L., Aebersold, R., and Goodlett, D. R. (2005) Methods Enzy-mol. 405, 66–81





http://www.jbc.org

35. Aronin, N., Kim, M., Laforet, G., and DiFiglia, M. (1999) Philos. Trans. R.Soc. Lond. B Biol. Sci. 354, 995–1003

36. Peters, M. F., Nucifora, F. C., Jr., Kushi, J., Seaman, H. C., Cooper, J. K.,Herring, W. J., Dawson, V. L., Dawson, T. M., and Ross, C. A. (1999)Mol.Cell. Neurosci. 14, 121–128

37. Wellington, C. L., Ellerby, L. M., Hackam, A. S., Margolis, R. L., Trifiro,M. A., Singaraja, R., McCutcheon, K., Salvesen, G. S., Propp, S. S., Bromm,M., Rowland, K. J., Zhang, T., Rasper, D., Roy, S., Thornberry, N., Pinsky,L., Kakizuka, A., Ross, C. A., Nicholson, D. W., Bredesen, D. E., and

Hayden, M. R. (1998) J. Biol. Chem. 273, 9158–916738. LaFevre-Bernt, M. A., and Ellerby, L. M. (2003) J. Biol. Chem. 278,

34918–3492439. Luo, S., Vacher, C., Davies, J. E., and Rubinsztein, D. C. (2005) J. Cell Biol.

169, 647–65640. Kaytor, M. D., Byam, C. E., Tousey, S. K., Stevens, S. D., Zoghbi, H. Y., and

Orr, H. T. (2005) Hum. Mol. Genet. 14, 1095–110541. Emamian, E. S., Kaytor, M. D., Duvick, L. A., Zu, T., Tousey, S. K., Zoghbi,

H. Y., Clark, H. B., and Orr, H. T. (2003) Neuron 38, 375–387





http://www.jbc.org

Documents

Huntingtin Phosphorylation Sites Mapped by Mass Spectrometrycnc.cj.uc.pt/BEB/BEB/private/pdfs... · Huntingtin Phosphorylation Sites Mapped by Mass Spectrometry MODULATION OF CLEAVAGE