Portal de Periódicos da CAPES

Exposure on Cell Surface and Extensive Arginine Methylation of Ewing Sarcoma (EWS) Protein

2001; Elsevier BV; Volume: 276; Issue: 22 Linguagem: Inglês

10.1074/jbc.m011446200

1083-351X

Larisa Belyanskaya, Peter Gehrig, Heinz Gehring,

RNA modifications and cancer

In contrast to the knowledge regarding the function of chimeric Ewing sarcoma (EWS) fusion proteins that arise from chromosomal translocation, the cellular function of the RNA binding EWS protein is poorly characterized. EWS protein had been found mainly in the nucleus. In this report we show that EWS protein is not only found in the nucleus and cytosol but also on cell surfaces. After cell-surface biotinylation, isoelectric focusing of membrane fraction, avidin-agarose extraction of biotinylated proteins, and SDS-polyacrylamide gel electrophoresis, EWS protein was identified by matrix-assisted laser desorption ionization and nanoelectrospray tandem mass spectrometry of in-gel-digested peptides. These analyses revealed that the protein, having repeated RGG motifs, is extensively asymmetrically dimethylated on arginine residues, the sites of which have been mapped by mass spectrometric methods. Out of a total of 30 Arg-Gly sequences, 29 arginines were found to be at least partially methylated. The Arg-Gly-Gly sequence was present in 21 of the 29 methylation sites, and in contrast to other methylated proteins, only 11 (38%) methylated arginine residues were found in the Gly-Arg-Gly sequence. The presence of Gly on the C-terminal side of the arginine residue seems to be a prerequisite for recognition by a protein-arginine N-methyltransferase (PRMT) catalyzing this asymmetric dimethylation reaction. One monomethylarginine and no symmetrically methylated arginine residue was found. The present findings imply that RNA-binding EWS protein shuttles from the nucleus to the cell surface in a methylated form, the role of which is discussed. In contrast to the knowledge regarding the function of chimeric Ewing sarcoma (EWS) fusion proteins that arise from chromosomal translocation, the cellular function of the RNA binding EWS protein is poorly characterized. EWS protein had been found mainly in the nucleus. In this report we show that EWS protein is not only found in the nucleus and cytosol but also on cell surfaces. After cell-surface biotinylation, isoelectric focusing of membrane fraction, avidin-agarose extraction of biotinylated proteins, and SDS-polyacrylamide gel electrophoresis, EWS protein was identified by matrix-assisted laser desorption ionization and nanoelectrospray tandem mass spectrometry of in-gel-digested peptides. These analyses revealed that the protein, having repeated RGG motifs, is extensively asymmetrically dimethylated on arginine residues, the sites of which have been mapped by mass spectrometric methods. Out of a total of 30 Arg-Gly sequences, 29 arginines were found to be at least partially methylated. The Arg-Gly-Gly sequence was present in 21 of the 29 methylation sites, and in contrast to other methylated proteins, only 11 (38%) methylated arginine residues were found in the Gly-Arg-Gly sequence. The presence of Gly on the C-terminal side of the arginine residue seems to be a prerequisite for recognition by a protein-arginine N-methyltransferase (PRMT) catalyzing this asymmetric dimethylation reaction. One monomethylarginine and no symmetrically methylated arginine residue was found. The present findings imply that RNA-binding EWS protein shuttles from the nucleus to the cell surface in a methylated form, the role of which is discussed. cyclophilin Ewing sarcoma protein-arginineN-methyltransferase phosphate-buffered saline 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonic acid matrix-assisted laser desorption ionization mass spectrometry tandem mass spectrometry polyacrylamide gel electrophoresis peripheral blood mononuclear cells erythroblastosis virus-transforming sequence heterogeneous nuclear ribonuclear protein While investigating a 90-kDa anti-cyclophilin (anti-CyP)1 immunoreactive band we noticed that anti-CyP antibodies recognized the RNA-binding Ewing sarcoma (EWS) protein and not a cyclophilin. The EWS gene is involved in tumor-related chromosomal translocations that associate part of EWS gene with various genes encoding transcription factors (1Delattre O. Zucman J. Melot T. Garau X.S. Zucker J.M. Lenoir G.M. Ambros P.F. Sheer D. Turc-Carel C. Triche T.J. et al.N. Engl. J. Med. 1994; 331: 294-299Crossref PubMed Scopus (933) Google Scholar). The N-terminal transcriptional activation domain of EWS is fused to C-terminal DNA binding domains of corresponding partners. The translocation produces chimeric EWS proteins with transforming potential (2Brown A.D. Lopez-Terrada D. Denny C. Lee K.A. Oncogene. 1995; 10: 1749-1756PubMed Google Scholar, 3Fujimura Y. Ohno T. Siddique H. Lee L. Rao V.N. Reddy E.S. Oncogene. 1996; 12: 159-167PubMed Google Scholar, 4May W.A. Lessnick S.L. Braun B.S. Klemsz M. Lewis B.C. Lunsford L.B. Hromas R. Denny C.T. Mol. Cell. Biol. 1993; 13: 7393-7398Crossref PubMed Scopus (453) Google Scholar, 5Ohno T. Rao V.N. Reddy E.S. Cancer Res. 1993; 53: 5859-5863PubMed Google Scholar, 6Prasad D.D. Ouchida M. Lee L. Rao V.N Reddy E.S. Oncogene. 1994; 9: 3717-3729PubMed Google Scholar, 7Zucman J. Melot T. Desmaze C. Ghysdael J. Plougastel B. Peter M. Zucker J.M. Triche T.J. Sheer D. Turc-Carel C. EMBO J. 1993; 12: 4481-4487Crossref PubMed Scopus (510) Google Scholar). The EWS gene of Ewing sarcoma and primitive neuroectodermal tumor is translocated to one of different members of the ETS (erythroblastosis virus-transforming sequence) family that contains the highly conserved DNA binding ETS domain. Often the ETS domain is derived from FLI-1 (Friend leukemia integration-1) and in rare cases from ERG (ETS-related gene), ETV-1 (ETS translocation variant-1), E1AF (E1A factor), or FEV (fifth Ewing variant). In malignant melanoma of soft parts, EWS is fused to ATF-1, in intra-abdominal desmoplasmic small round-cell tumor to WT-1, in myxoid liposarcoma to CHOP, and in myxoid chrondrosarcoma to CHN (8de Alava E. Gerald W.L. J. Clin. Oncol. 2000; 18: 204-213Crossref PubMed Google Scholar). The cellular role of wild-type EWS protein remains less clear. The EWS protein is a nuclear protein with unknown function containing a C-terminal RNA binding motif and a N-terminal activation domain (9Lessnick S.L. Braun B.S. Denny C.T. May W.A. Oncogene. 1995; 10: 423-431PubMed Google Scholar, 10Ohno T. Ouchida M. Lee L. Gatalica Z. Rao V.N. Reddy E.S. Oncogene. 1994; 9: 3087-3097PubMed Google Scholar, 11Ouchida M. Ohno T. Fujimura Y. Rao V.N. Reddy E.S. Oncogene. 1995; 11: 1049-1054PubMed Google Scholar). The IQ domain of the EWS protein is involved in calmodulin binding and protein kinase C phosphorylation (12Deloulme J.C. Prichard L. Delattre O. Storm D.R. J. Biol. Chem. 1997; 272: 27369-27377Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar). EWS protein interacts with an SH3 domain of Bruton's tyrosine kinase and has been identified in B cells as a phosphotyrosine-containing protein (13Guinamard R. Fougereau M. Seckinger P. Scand. J. Immunol. 1997; 45: 587-595Crossref PubMed Scopus (62) Google Scholar). G-coupled receptor signaling and other stimuli of tyrosine kinase Pyk2 block the interaction between EWS protein and Pyk2. Partitioning of the EWS protein into a ribosome-associated fraction indicated that the role for EWS in gene expression includes an extranuclear action (14Felsch J.S. Lane W.S. Peralta E.G. Curr. Biol. 1999; 9: 485-488Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar). In the present investigation we show the EWS protein is not only localized in the nucleus and cytosol but also on the surface of cells and that it is posttranslationally methylated at arginine residues. The identified ω-N G,N G-dimethylarginine residues of EWS protein let us modify a previously reported consensus sequence for asymmetric dimethylarginine formation in proteins. Stock cultures of Jurkat cells, a tumor T lymphoma cell line, were kindly provided by Dr. J. Kemler-Carraneo (Abteilung Klinische Immunologie, Universitätsspital Zürich). Peripheral blood mononuclear (PBM) cells and human cutaneous T lymphoma cell line (H9) were obtained from the National Center for Retroviruses (University of Zürich). Purified polyclonal antibodies raised against human CyPA and CyPB (anti-CyP) were kindly provided by G. Woerly (Toxikologisches Institut der Universität Zürich, Schwerzenbach). Polyclonal antibody 677 against the N terminus of EWS protein (anti-EWS) was a generous gift from Dr. Olivier Delattre (Institut Curie, Pathologie Moléculaire des Cancers, Paris Cedex). Human Jurkat cells were grown in RPMI 1640 medium (Sigma) supplemented with 5% newborn calf serum (Life Technologies), 15 mm HEPES, 2 mml-glutamine, 50 μm β-mercaptoethanol, and 1% (w/v) penicillin and streptomycin (Life Technologies) in a humidified 5% CO2 atmosphere at 37 °C. PBM cells were isolated by Ficoll-Hypaque gradient centrifugation. The cells were washed with phosphate-buffered saline, pH 7.4 (PBS), and stimulated with 2 μg/ml phytohemagglutinin in RPMI 1640 medium supplemented with 20% fetal calf serum (Life Technologies), 15 mm HEPES, 2 mml-glutamine, 50 μmβ-mercaptoethanol, and 1% (w/v) penicillin and streptomycin in a humidified 5% CO2 atmosphere at 37 °C. Human H9 cell line was grown in RPMI 1640 medium supplemented with 20% fetal calf serum, 15 mm HEPES, 2 mml-glutamine, 50 μm β-mercaptoethanol, and 1% (w/v) penicillin and streptomycin in a humidified 5% CO2 atmosphere at 37 °C. Cell-surface biotinylation was performed as described (15Altin J.G. Pagler E.B. Anal. Biochem. 1995; 224: 382-389Crossref PubMed Scopus (117) Google Scholar) with some modification. 2 × 108 cells were washed three times with ice-cold PBS, suspended in PBS (25 × 106 cells/ml), and incubated with 0.5 mg/ml sulfosuccinimidobiotin (Calbiochem) for 1 h at 4 °C. The biotinylation procedure was stopped by replacing the labeling solution with culture medium. After incubation for 10 min at 4 °C, cells were washed twice with PBS. Biotinylated or non-labeled cells were solubilized (25 × 106 cells/ml) in lysis buffer containing 1% (w/v) Triton X-100, PBS, protease inhibitor mixture (20 μg/ml pancreas extract, 5 μg/ml Pronase, 0.5 μg/ml thermolysin, 3 μg/ml chymotrypsin, 330 μg/ml papain; Roche Molecular Biochemicals), 2 mm EDTA, 0.5 mm phenylmethylsulfonyl fluoride, 0.5 mmdithiothreitol for 40 min on ice with occasional vortexing. The cell lysates were centrifuged at 480 × g for 5 min followed by 15000 × g for 15 min. Obtained supernatants were then subjected to isoelectric focusing followed by avidin-agarose extraction or directly to avidin-agarose extraction and immunoprecipitation experiments. The Bio-Rad protein assay kit was used to determine protein concentrations. All steps were performed at 4 °C. Biotinylated or nonlabeled cells were harvested and washed three times with ice-cold PBS. Pelleted cells were suspended (5 × 108 cells/ml) in homogenizing buffer containing 25 mm Tris-HCl, pH 7.6, 10% (v/v) glycerol, 0.5 mm dithiothreitol, 0.5 mm phenylmethylsulfonyl fluoride, 2 mm EDTA, and protease inhibitor mixture. Cells were homogenized with a Dounce homogenizer (20 strokes), and the suspension was centrifuged at 500 × g. The supernatant was then centrifuged for 10 min at 15,000 × g, and a membrane pellet obtained at 150000 × g for 1 h was used for further purification. This supernatant was designated as cytosolic fraction. Pellets were solubilized in 8 m urea, 2% (w/v) CHAPS, 15 mm dithiothreitol, 2% (v/v) Bio-Lyte ampholytes (Bio-Rad), pH 3–10, for 40 min at room temperature. Insoluble material was removed by centrifugation at 10,000 ×g for 15 min, and the supernatant was designated as membrane fraction. Isoelectric focusing using a Mini Rotofor cell (Bio-Rad) was done as recommended by the supplier. After cell-surface biotinylation, part of the detergent lysates (17 mg) or solubilized membrane fraction (15 mg) were added to the Rotophor cell containing 13 ml of 8 m urea, 2% (w/v) CHAPS, 15 mm dithiothreitol, 2% (v/v) Bio-Lyte ampholytes, pH 3–10, after prefocusing for 1 h at 12 W. Focusing was carried out at 12 W for 5 h at 10 °C. The 20 fractions were harvested, and their pH values were measured. After Western blot analysis, the fractions of interest were pooled and concentrated with Centricon (YM-30; Millipore), and biotinylated proteins were extracted with avidin-agarose. Either the detergent lysate or membrane fraction pooled after isoelectric focusing was incubated with avidin-agarose (Sigma) overnight on a rotary device at 4 °C. The affinity gel was then washed 3 times with 1% (w/v) Triton X-100, 0.2% (w/v) SDS, 5 mm EDTA in PBS followed by a wash without SDS and one with water. The captured biotinylated proteins were eluted from the agarose with reducing 2-fold Laemmli buffer heated for 5 min at 95 °C. Proteins were separated by SDS-PAGE (7.5%, 0.75 mm thickness) and stained with 0.2% (w/v) Coomassie Brilliant Blue R250 in 50% (v/v) ethanol and 10% (v/v) acetic acid for 30 min. Protein bands of interest were excised from the gel and in-gel-digested with trypsin (Promega) or chymotrypsin (Roche Molecular Biochemicals) following the procedure of Shevchenkoet al. (16Shevchenko A. Wilm M. Vorm O. Mann M. Anal. Chem. 1996; 68: 850-858Crossref PubMed Scopus (7885) Google Scholar). Upon reduction of the disulfide bonds of the protein with tris(2-carboxyethylphosphine) hydrochloride (Pierce), cysteines were alkylated with iodoacetamide (Sigma). Digestions with trypsin (400 ng) were carried out overnight at 37 °C in 100 mm ammonium bicarbonate buffer, pH 8.3, and 4 mm calcium chloride. Digestions with chymotrypsin (500 ng) were performed overnight at 25 °C in the same buffer, pH 8.0. The resulting peptides were extracted from the gel pieces and desalted using pipette tips with C18 resins (Millipore). Matrix-assisted laser desorption ionization time-of-flight (MALDI-TOF) mass spectra of the entire digests were recorded on a Bruker Biflex instrument in the reflector mode employing pulsed ion extraction. α-Cyano-4-hydroxycinnamic acid (Fluka) was used as the matrix. The MALDI mass spectra were mass calibrated using ion signals mainly from autoproteolytic fragments of trypsin and chymotrypsin, respectively. Electrospray ionization mass spectra (MS) and tandem mass spectra (MS/MS) were acquired on an API III+ triple-quadrupole instrument (PE-Sciex, Ontario, Canada) equipped with a nanoelectrospray ion source (Protana, Odense, Denmark). The detergent lysates of cells were preincubated with protein A-Sepharose (Amersham Pharmacia Biotech) for 2 h. Anti-CyP antiserum or anti-EWS antiserum were added to the supernatants, and the mixture was incubated at 4 °C for at least 12 h, and then protein A-Sepharose was added for 2 h. The agarose pellet was washed three times with 1% (w/v) Triton X-100, 2 mm EDTA, and 10 mm Tris-HCl, pH 7.4, and the antigen was eluted from the gel with reducing 2-fold Laemmli buffer by heating for 5 min at 95 °C. Proteins separated on 7.5% SDS-PAGE were transferred electrophoretically to a nitrocellulose membrane (Schleicher and Schuell) as described (17Towbin H. Staehelin T. Gordon J. Proc. Natl. Acad. Sci. U. S. A. 1979; 76: 4350-4354Crossref PubMed Scopus (46639) Google Scholar). After transfer, the membrane was blocked for 40 min with 5% (w/v) milk powder dissolved in PBS and 0.25% (w/v) Tween 20 and incubated with the first antibodies anti-CyP or anti-EWS, respectively, for 1 h at room temperature, followed by incubation with goat anti-rabbit antibody coupled to horseradish peroxidase (Sigma) for 1 h at room temperature. The detection was performed by enhanced chemiluminescence (ECL, Amersham Pharmacia Biotech). Cell-surface proteins of Jurkat and PBM cells were labeled under physiological conditions with the impermeable biotinylation reagent. Labeled surface proteins were extracted with avidin-coupled agarose and separated by SDS-PAGE. Western blot analysis with anti-CyP antibodies showed a strong signal of a protein band with a molecular mass of about 90 kDa (Fig.1 A). To obtain a pure protein in amounts high enough for identification, a membrane fraction of Jurkat and PBM cells was prepared by differential centrifugation after cell-surface biotinylation. The solubilized proteins were further purified by isoelectric focusing under denaturing conditions. The anti-CyP immunoreactive protein was found in a pH range from 7.2 to 9.3. These fractions were pooled, and the biotinylated proteins extracted with avidin-agarose and subjected to SDS-PAGE. The Coomassie-stained band (Fig. 1 B) containing the immunoreactive 90-kDa protein of interest (Fig. 1 C) was clearly separated from other proteins and was sufficiently pure to be subjected for an identification procedure. The 90-kDa Coomassie-stained gel band containing the protein of interest was excised and in-gel-digested with trypsin, and the resulting peptides were analyzed by MALDI-MS (see “Experimental Procedures”). A protein sequence data base search (Mascot Search) performed with the obtained peptide masses (Table I) did not provide an unequivocal result. RNA-binding protein EWS (1Delattre O. Zucman J. Melot T. Garau X.S. Zucker J.M. Lenoir G.M. Ambros P.F. Sheer D. Turc-Carel C. Triche T.J. et al.N. Engl. J. Med. 1994; 331: 294-299Crossref PubMed Scopus (933) Google Scholar) was retrieved with the highest score; however, some of the most intense signals of the MALDI mass spectrum could not be assigned to this protein. To confirm the identification of the protein, nanoelectrospray MS/MS sequencing of selected peptides was performed. Four MS/MS spectra were in complete agreement with the predicted fragmentation pattern of peptides 269–292, 411–424, 425–439, and 472–486 from RNA-binding EWS protein and, thus, unambiguously verified the identification of this protein (Fig. 2, Table I).Table ISummary of mass spectrometric data of proteolytic fragments from the in-gel-digested protein (see Fig. 1B)Observed massPredicted peptide massResidue numbers in EWS proteinNotesDaDaDigestion with trypsin2479.002479.04QDHPSSMGVYGQESGGFSGPGENR269–292Assignment confirmed by MS/MS.2055.072055.08TGQPMIHIYLDKETGKPK393–4101449.621449.66GDATVSYEDPPTAK411–424Assignment confirmed by MS/MS.1683.701683.79AAVEWFDGKDFQGSK425–439Assignment confirmed by MS/MS.1692.881692.92KKPPMNSMRGGLPPR447–461766.46766.42GMPPPLR465–471Cleavage at Arg-464, which is a potential methylation site; Arg-471 is not modified in this peptide.1983.001982.99GMPPPLRGGPGGPGGPGGPMGR465–486Cleavage at Arg-464 (potential methylation site); dimethylation of Arg-471 confirmed by MS/MS.1968.991968.98GMPPPLRGGPGGPGGPGGPMGR465–486Arg-471 is mono-methylated in this peptide.1206.531206.55GGPGGPGGPGGPMGR472–486Cleavage at Arg-471 (potential methylation site); assignment confirmed by MS/MS.3326.633326.72MGGRGGDRGGFPPRGPRGSRGNPSGGGNVQHR487–5181380.691380.73GGDRGGFPPRGPR491–503Cleavage at Arg-490 (potential methylation site); Arg-503 was not modified in this peptide.2897.462897.50GGDRGGFPPRGPRGSRGNPSGGGNVQHR491–5181506.661506.74GSRGNPSGGGNVQHR504–518Cleavage at Arg-503, which is a potential methylation site.1777.831777.87RGGPGGPPGPLMEQMGGR615–632Dimethylation of Arg-615 confirmed by MS/MS.896.55896.54RGGRGGPGK633–641Digestion with chymotrypsin3156.493156.47GQQPAATAPTRPQDGNKPTETSQPQSSTGGY128–1581606.691606.72GQQPPTSYPPQTGSY233–247Assignment confirmed by MS/MS.1275.551275.56RQDHPSSMGVY268–2781633.831633.81SMSGPDNRGRGRGGF293–3071415.721415.75SGPDNRGRGRGGF295–3071960.031959.99SGPDNRGRGRGGFDRGGM295–3121187.591187.64SRGGRGGGRGGM313–3243225.433225.39NKPGGPMDEGPDLDLGPPVDPDEDSDNSAIY334–3641741.901741.98RGGLPPREGRGMPPPL455–4701149.561149.54RGGPGGPGGPGGPM471–484Arg-471 was not methylated in this peptide.1177.531177.57RGGPGGPGGPGGPM471–484Arg-471 was dimethylated in this peptide.1971.071971.02LPPPFPPPGGDRGRGGPGGM552–571727.43727.45RGGRGGL572–578858.50858.49RGGRGGLM572–5791760.931760.90RGGRGGLMDRGGPGGMF572–5881051.471051.47MDRGGPGGMF579–588Dimethylation of Arg-581 confirmed by MS/MS.920.45920.43DRGGPGGMF580–588Dimethylation of Arg-581 confirmed by MS/MS.1174.581174.64RGGRGGDRGGF589–5991248.601248.66RGGRGMDRGGF600–6101315.681315.72GGGRRGGPGGPPGPL611–6251834.941834.90GGGRRGGPGGPPGPLMEQM611–629988.55988.55RGGPGGPPGPL615–625Dimethylation of Arg-615 confirmed by MS/MS.1507.741507.74RGGPGGPPGPLMEQM615–629Arginines found to be dimethylated in the corresponding peptides are in bold and underlined, and the only monomethylated arginine at position 471 is in italics. Open table in a new tab Arginines found to be dimethylated in the corresponding peptides are in bold and underlined, and the only monomethylated arginine at position 471 is in italics. The tandem mass spectrum shown in Fig. 3 appeared to match peptide 615–632 from this protein, but the observed mass was 28 Da higher than the mass calculated from the amino acid sequence. Detailed analysis of the spectrum indicated that this peptide most likely contains a dimethylated arginine residue located at position 615 (Table I). Another MS/MS spectrum was found to be consistent with peptide 465–486 containing a dimethylarginine residue at position 471. Re-examination of the MALDI mass spectrum of the tryptic digest revealed the presence of eight additional dimethylarginines located in the C-terminal arginine- and glycine-rich domain of the protein (Fig. 2). Arginines 464, 471, 490, and 503 appear to be dimethylated to a large extent but not completely. Trypsin does not cleave after methylated arginines (18Baldwin G.S. Carnegie P.R. Science. 1971; 171: 579-581Crossref PubMed Scopus (147) Google Scholar, 19Merrill B.M. LoPresti M.B. Stone K.L. Williams K.R. Int. J. Pept. Protein Res. 1987; 29: 21-39Crossref PubMed Scopus (17) Google Scholar). Analyses of tryptic peptides indicated cleavages at these four arginines that were found to be modified in other peptides. In addition, peptides containing the non-methylated arginines 471 and 503 were identified. A small fraction of arginine 471 (≤15%) also occurs in the monomethylated form (Table I, peptide 465–486). To analyze the sites of arginine methylation more completely, RNA-binding protein EWS was in-gel-cleaved with chymotrypsin. MALDI-MS of the entire chymotryptic digest allowed the identification of 19 additional dimethylarginines (Table I and Fig. 2). Mainly peptides resulting from specific cleavage after the known chymotryptic cleavage sites phenylalanine, tyrosine, methionine, and leucine were assigned; however, arginine 292 and 614 were also found to be susceptible to chymotryptic cleavage. Only uniquely assignable mass peaks were reported. The identification of all peptides listed in TableI was confirmed by the excellent agreement between the observed molecular masses with those predicted from the known sequences (within 0.006%). The assignment of three dimethylarginine-containing peptides and of peptide 233–247 was ascertained by nanoelectrospray MS/MS analyses as noted in Table I. A qualitative amino acid analysis (data not shown) of an EWS protein hydrolysate revealed the presence of asymmetric dimethylarginine residues, whereas symmetric dimethylarginine, if at all present, or monomethylarginine residues were below the detection limit, although monomethylarginine is present in small amounts according to the MS data. To check the methylation state of EWS protein in the nucleus, the protein was isolated from the nucleus as described (12Deloulme J.C. Prichard L. Delattre O. Storm D.R. J. Biol. Chem. 1997; 272: 27369-27377Abstract Full Text Full Text PDF PubMed Scopus (72) Google Scholar), in-gel digested with either trypsin or chymotrypsin, and subjected to MALDI-MS analysis (data not shown). The peptide map of nuclear EWS protein was found to be comparable with the peptide map of EWS protein isolated from the plasma membrane, indicating a similar dimethylation pattern of arginine residues of the nuclear EWS protein. The detection of an immunoreactive band with anti-CyP antibodies at the same position in SDS-PAGE as the EWS protein indicated either cross-reactivity of the antibodies or cross-contamination of the proteins. Thus immunoprecipitations with anti-CyP antibodies and anti-EWS antibody were performed from cell lysates followed by Western analysis with anti-CyP antibodies (Fig.4 A) as well as with anti-EWS antibody (Fig. 4 B). In any case a stronger signal was obtained in the Western with anti-EWS antibody even if immunoprecipitation was done with anti-CyP antibodies, indicating cross-reactivity of both antibodies with EWS protein. No signals were observed if (when) goat anti-rabbit antibody alone or rabbit antibodies against several other proteins was used. Excision of the immunoprecipitated 90-kDa Coomassie-stained bands, in-gel digestion with trypsin, and MALDI-MS analysis of tryptic peptides confirmed the presence of methylated RNA-binding protein EWS in anti-CyP and anti-EWS-precipitated samples. To ascertain that EWS protein is indeed present on the cell surface, Jurkat, PBM, but also H9 cells were surface-labeled with the impermeable biotinylation reagent, and biotinylated proteins were extracted with avidin-coupled agarose. The Western blot, now performed with anti-EWS antibody, revealed the presence of EWS protein on all three cell types (Fig. 5 A). Densitometry showed an identical high content of EWS protein on the surface of the tumor cell lines Jurkat and H9, whereas in PBM cells the content of EWS protein was drastically lower. On purifying the cell-surface biotinylated proteins by isoelectric focusing of membrane-enriched fractions as described in Fig. 1 B, a strong signal was obtained again in the Western blot analysis with anti-EWS antibody (Fig. 5 B), showing that EWS protein is indeed present on the cell surface. To verify that the 90-kDa protein represents a surface-exposed molecule and not an unspecifically extracted protein, proteins from nonbiotinylated cells were extracted and analyzed by the same procedure. Signals in the anti-CyP or anti-EWS Western blot analysis were not found (not shown). Our data show that the anti-CyP immunoreactive protein located on the surface of T cells is not a cyclophilin but the EWS protein. The anti-CyP antibodies cross-react with the EWS protein as demonstrated by immunoprecipitation experiments. The cause of the cross-reactivity is not obvious. Global alignment of the EWS protein sequence either with human CyPA or CyPB using the program LALIGN revealed a low degree of identity (8.8%) in both cases, and some of the identity seems to be due to numerous glycines present in the proteins. The cross-reactivity led us, however, to the finding that the EWS protein is not only exposed on the cell surface of different cells but also that its arginine residues are extensively and asymmetrically dimethylated. These properties of the EWS protein shed a new light on the functionality of this unusual multidomain protein. The previously reported localization of EWS protein in the nucleus (20Bertolotti A. Melot T. Acker J. Vigneron M. Delattre O. Tora L. Mol. Cell. Biol. 1998; 18: 1489-1497Crossref PubMed Scopus (222) Google Scholar) and the cytosol (14Felsch J.S. Lane W.S. Peralta E.G. Curr. Biol. 1999; 9: 485-488Abstract Full Text Full Text PDF PubMed Scopus (44) Google Scholar), both of which we confirmed (data not shown), together with the present finding of the EWS protein to be accessible on the cell surface means that this protein shuttles between the nucleus, cytosol, and the cell surface. A similar behavior was reported for nucleolin. This major nucleolar protein shuttles between the cytosol and the nucleus and has also been detected on the cell surface of different cells. The C-terminal domain of nucleolin is, as in the EWS protein, rich in glycine residues and interspersed with dimethylarginines. It was suggested as a potential receptor in the human immunodeficiency virus binding processes by interaction with the V3 loop of gp120 (21Srivastava M. Pollard H.B. FASEB J. 1999; 13: 1911-1922Crossref PubMed Scopus (436) Google Scholar). So far we found cell-surface-exposed EWS protein in all investigated cells, i.e. Jurkat, H9, C816645 T cell lines, and PBM cells, but also NIH/3T3 fibroblasts (not shown). Remarkably, tumor cell lines showed a higher level of EWS protein expression on the cell surface (∼4-fold) than PBM cells and fibroblasts. Arginine methylation is a post-translational modification found mainly in nuclear proteins that interact with RNA (22Gary J.D Clarke S. Prog. Nucleic Acid Res. Mol. Biol. 1998; 61: 65-131Crossref PubMed Google Scholar). This modification is catalyzed by protein-arginine N-methyltransferases (PRMTs), utilizing S-adenosyl-l-methionine as the donor of methyl groups (23Lee H.W. Kim S. Paik W.K. Biochemistry. 1977; 16: 78-85Crossref PubMed Scopus (74) Google Scholar). Type I protein-arginineN-methyltransferases (EC 2.1.1.23) catalyze the formation of N G-monomethylarginine and asymmetric ω-N G,N G-dimethylarginine residues, whereas Type II enzymes catalyze the formation ofN G-monomethylarginine and symmetric ω-N G,N′G-dimethylarginine residues (22Gary J.D Clarke S. Prog. Nucleic Acid Res. Mol. Biol. 1998; 61: 65-131Crossref PubMed Google Scholar). The Type III enzyme found in yeast catalyzes the monomethylation of the internal δ-guanidino nitrogen atom of arginine residues (24Zobel-Thropp P. Gary J.D. Clarke S. J. Biol. Chem. 1998; 273: 29283-29286Abstract Full Text Full Text PDF PubMed Scopus (56) Google Scholar). Three Type I PRMTs from mammalian cells PRMT1 (25Lin W.J. Gary J.D. Yang M.C. Clarke S. Herschman H.R. J. Biol. Chem. 1996; 271: 15034-15044Abstract Full Text Full Text PDF PubMed Scopus (401) Google Scholar), PRMT3 (26Tang J. Gary J.D. Clarke S. Herschman H.R. J. Biol. Chem. 1998; 273: 16935-16945Abstract Full Text Full Text PDF PubMed Scopus (277) Google Scholar), and coactivator-associated arginine methyltransferase 1 (CARM1) (27Chen D. Ma H. Hong H. Koh S.S. Huang S.M. Schurter B.T.