SC5 mAb Represents a Unique Tool for the Detection of Extracellular Vimentin as a Specific Marker of Sézary Cells

This information is current as Delphine Huet, Martine Bagot, Denis Loyaux, Joël of September 28, 2021. Capdevielle, Laurence Conraux, Pascual Ferrara, Armand Bensussan and Anne Marie-Cardine J Immunol 2006; 176:652-659; ; doi: 10.4049/jimmunol.176.1.652

http://www.jimmunol.org/content/176/1/652 Downloaded from

References This article cites 36 articles, 9 of which you can access for free at: http://www.jimmunol.org/content/176/1/652.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on September 28, 2021

*average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

SC5 mAb Represents a Unique Tool for the Detection of Extracellular Vimentin as a Specific Marker of Se´zary Cells1

Delphine Huet,* Martine Bagot,* Denis Loyaux,† Joe¨l Capdevielle,† Laurence Conraux,† Pascual Ferrara,† Armand Bensussan,2,3* and Anne Marie-Cardine2*

Circulating malignant Se´zary lymphocytes result from a clonal proliferation of memory/activated CD4؉CD45RO؉ T lymphocytes primarily involving the skin. Recently, the CD158k/KIR3DL2 cell surface receptor has been identified to phenotypically charac- terize these cells. We previously described a mAb termed SC5 that identifies an unknown early activation cell membrane molecule. It is expressed selectively by T lymphocytes isolated from healthy individuals upon activation, and by circulating Se´zary syndrome lymphocytes. In addition, we found that SC5 mAb was reactive with all resting T lymphocytes once permeabilized, indicating that SC5 mAb-reactive molecule might present distinct cellular localization according to the T cell activation status. In this study, we show for the first time that SC5 mAb recognizes the intermediate filament vimentin when exported to the extracellular side Downloaded from of the plasma membrane of viable Se´zary malignant cells. We demonstrate that SC5 mAb is unique as it reacts with both viable malignant lymphocytes and apoptotic T cells. As vimentin is also detected rapidly at the cell membrane surface after normal T lymphocyte activation, it suggests that its extracellular detection on Se´zary cells could be a consequence of their constitutive activation status. Finally, as a probable outcome of vimentin cell surface expression, autoantibodies against vimentin were found in the sera of Se´zary syndrome patients. The Journal of Immunology, 2006, 176: 652–559. http://www.jimmunol.org/ utaneous T cell lymphomas (CTCL)4 are a heteroge- and to distinguish them from normal reactive CD4ϩ lympho- neous group of lymphomas primarily involving the skin cytes, was to identify characteristic cell membrane receptors. C (1). Mycosis fungoides (MF) and Se´zary syndrome (SS) Thus, the immunophenotypic analysis of peripheral blood T are the two major clinical variants of CTCL. MF, the most cells from patients with SS has revealed a frequent loss of T common form, is characterized by erythematous patches, lymphocyte receptor expression, such as CD7 or CD26, when plaques, or tumors of the skin constituted of clonally derived compared with normal T lymphocytes (7–12). Recently, we re- malignant T lymphocytes that phenotypically resemble mature/ ported the presence of the CD158k/KIR3DL2 inhibitory cell memory T cells. SS, a more aggressive leukemic and erythro- membrane receptor for HLA-A alleles on different CTCL lines by guest on September 28, 2021 dermic variant of CTCL, is characterized by the presence of a established from the skin or the blood of patients with SS (13). ϩ ϩ clonal CD4 CD45RO T lymphocyte population with atypical Furthermore, we found in a large group of patients with SS a cerebriform nuclei in the skin, lymph nodes, and peripheral strong positive correlation between the percentage of CD158kϩ blood (2). To understand the pathogenesis and to facilitate the blood lymphocytes, detected by flow cytometry, and the per- molecular diagnosis of Se´zary cells, a expression analysis centage of Se´zary cells evaluated by cytomorphology (14). In- approach was undertaken and revealed the selective expression terestingly, in situ analysis revealed that CD158k is expressed of the tyrosine kinase receptor EphA4 and of the transcription by cutaneous malignant lymphocytes from SS patients, but not factor Twist (3), an overexpression of JunB (4), and a complete from patients with early MF (15). This identified CD158k as the loss of Bcl2 (5). Furthermore, using a modified representational first specific cell surface marker available for the evaluation of difference analysis, Su et al. (6) reported the aberrant mRNA the circulating tumoral burden, and for the follow-up of patients and protein expression of T-plastin in Se´zary cells. A different with SS. In addition to CD158k, we further evidenced the pres- approach to better understand the pathogenesis of Se´zary cells, ence, on the cell surface of circulating Se´zary cells, of a struc- ture that was recognized by a mAb that we called SC5 (16, 17). *Institut National de la Sante´et de la Recherche Me´dicale Unite´659, University of However, in contrast to CD158k, which expression is highly Paris XII, Medical School of Cre´teil, Henri Mondor Hospital, Department of Der- ϩ matology, Cre´teil, France; and †Sanofi-Aventis, Labe`ge, France restricted to a minor circulating NK and CD8 T cell subset, the Received for publication June 30, 2005. Accepted for publication October 20, 2005. SC5-reactive molecule was found to be expressed in the cyto- The costs of publication of this article were defrayed in part by the payment of page plasm of all circulating lymphocytes, and at the cell surface of charges. This article must therefore be hereby marked advertisement in accordance normal T lymphocytes following activation. Thus, we postu- with 18 U.S.C. Section 1734 solely to indicate this fact. lated that this nonidentified protein might correspond to an in- 1 This work was supported by Institut National de la Sante´et de la Recherche Me´di- tracellular structure in normal resting T lymphocytes, and that cale and Grants 4604 (to A.B.) and 4220 (to M.B.) from the Association pour la Recherche contre le Cancer. its surface membrane expression increased rapidly after activa- 2 A.B. and A.M.-C. are senior authors on this paper. tion, as reported for CD152 (18, 19), and for a heterogeneous family of lysosome-associated membrane (20). 3 Address correspondence and reprint requests to Dr. Armand Bensussan, Institut National de la Sante´et de la Recherche Me´dicale Unite´659, Faculte´deMe´decine de In this study, we report that the SC5 mAb-reactive molecule, Cre´teil, 8 rue du Ge´ne´ral Sarrail, F-94010 Cre´teil Cedex. E-mail address: located on the extracellular side of the plasma membrane of acti- [email protected] vated normal T lymphocytes and of viable malignant SS cells, cor- 4 Abbreviations used in this paper: CTCL, cutaneous T cell lymphoma; LSP1, lym- phocyte-specific protein 1; MF, mycosis fungoides; PI, propidium iodide; SS, Se´zary responds to the class III intermediate filament protein vimentin. More syndrome. importantly, we identified SC5 mAb as a unique tool for the detection

Copyright © 2005 by The American Association of Immunologists, Inc. 0022-1767/05/$02.00 The Journal of Immunology 653

of vimentin at the cell membrane surface of viable lymphocytes, and an ESI-QTOF instrument (Waters). Mascot (Matrix Sciences) and Protein- therefore discuss whether this intermediate filament protein could be Lynx Global server (Matrix Sciences) softwares were used together for targeted as a tumor or an inflammatory Ag for mAb therapy. Swiss- and National Registar-National Center for Biotechnology Informa- tion- searches. Materials and Methods Generation of Flag-tagged vimentin constructs Patients The cDNA coding for a C-terminal Flag-tagged vimentin was generated by PCR amplification of the full-length coding region of vimentin. The re- After informed consent and approval by an ethics committee (Comite´Con- sulting PCR product was purified and ligated into the pcDNA3.1 vector, sultatif de Protection des Personnes dans la Recherche Biome´dicale, Henri according to the manufacturer’s recommendations (Invitrogen Life Tech- Mondor), sera and PBMC were isolated from blood samples of 10 patients nologies). Similarly, the vimentin-deleted constructs were raised by PCR with CTCL. The patients had not been previously treated with amplification of the coding regions corresponding to aa 101–466 (⌬1– chemotherapy. 100), 139–466 (⌬1–138), and 228–466 (⌬1–227), fused to a C-terminal Cells and cell lines Flag tag. All cDNA constructs were further inserted into pcDNA3.1 ex- pression vector. PBMCs were isolated by the technique of Ficoll-Isopaque (Pharmacia) density gradient centrifugation. To obtain T cell lines, cells were main- Generation of GST-vimentin fusion protein ␮ tained in RPMI 1640 supplemented with 2 mM L-glutamine, 100 g/ml Total RNAs were extracted from the human CTCL cell line HUT78, and penicillin/streptomycin (Invitrogen Life Technologies), 10% heat-inacti- vimentin cDNA was generated by RT-PCR. The PCR product was cloned vated human serum (Jacques Boy Institute), and 100 IU/ml human rIL-2 into the pGEX4T1 vector (Amersham Biosciences) at BamHI and EcoRI (produced by Sanofi-Aventis). The mycoplasma-free HUT78 cell line was sites. The cDNA sequence was confirmed by double-stranded sequencing grown in RPMI 1640 containing 10% FCS (Perbio Science Europe) and

analysis. Downloaded from antibiotics. Expression and purification of GST-vimentin protein were essentially Antibodies performed according to the manufacturer’s procedure and as previously described (22). Where indicated, the fusion protein was subjected to di- The mouse SC5 mAb (IgM) was raised in our laboratory (Institut National gestion with thrombin before analysis by SDS-PAGE and Western blotting. de la Sante´et de la Recherche Me´dicale Unite´659) against the leukemic NK cell line YTindi, as described elsewhere (17). SC5 hybridoma culture Immunoprecipitation and immunoblotting supernatant was used for Western blot analysis. The purified anti-vimentin Cells (2 ϫ 107 per sample) were lysed in 1% Nonidet P-40 lysis buffer and http://www.jimmunol.org/ goat polyclonal Abs (C-20) were generated against a peptide located in the processed as described above. After preclearing on protein A-Sepharose C terminus part of human vimentin (sc-7557; Santa Cruz Biotechnology), beads, lysates were incubated with SC5, anti-vimentin (clone V9), or iso- while the anti-vimentin mAb (clone V9, IgG1; Oncogene Research Prod- type-matched irrelevant mAb. Immunoprecipitates were collected on pro- ucts) was obtained by immunization of mice with purified vimentin. Both tein G-Sepharose beads and separated by SDS-8% PAGE under nonreduc- Abs were used at 1 ␮g/ml for Western blot analysis. The anti-TCR␣␤ mAb ing conditions. Proteins were electrotransferred on nitrocellulose (IP26, IgG1) was locally produced (Institut National de la Sante´etdela membrane and subjected to immunoblotting using SC5 or anti-vimentin Recherche Me´dicale Unite´659) and can be purchased from eBioscience. Abs and HRP-conjugated secondary Abs. Detection was realized with an The PE-conjugated anti-TCRV␤23 mAb was purchased from Beckman ECL detection system, according to the manufacturer’s recommendations Coulter. (Amersham Biosciences). For the mapping of SC5 mAb recognition site, COS 7 cells were trans-

Flow cytometry fected with the indicated plasmid using a DEAE-dextran transfection by guest on September 28, 2021 Double cell staining was performed according to standard procedure. method (23). After 48 h of culture, cells were harvested and lysed. Re- Briefly, cells (4 ϫ 105) were incubated with SC5 mAb (purified from combinant proteins were immunoprecipitated using an anti-Flag mAb and hybridoma culture supernatant) for 30 min at 4°C, followed by goat anti- protein G-Sepharose beads, and resolved by SDS-12% PAGE under non- mouse IgM FITC Abs (1 ␮l/test; Beckman Coulter). After washes, a sec- reducing conditions. Immunoblotting was performed as described above. ␣␤ ond labeling with the anti-TCR mAb (hybridoma culture supernatant) Confocal immunofluorescence microscopy and goat anti-mouse IgG1 PE (1 ␮l/test; Beckman Coulter) was realized. Alternatively, PE-conjugated anti-TCRV␤23 mAb was used (10 ␮l/test). HUT 78 cells were placed in V-shaped 96-well plates (3 ϫ 105cells/well). For single labeling, cells were left untreated or fixed for 30 min at 4°C After a washing step in PBS, cells were fixed in PBS/4% paraformalde- in PBS/4% paraformaldehyde. When indicated, an additional step of per- hyde/0.025% glutaraldehyde. For intracellular staining, cells were perme- meabilization was performed in PBS/0.1% Triton X-100. Cells were then abilized with PBS/0.1% Triton X-100. After quenching with 50 mM

labeled with SC5 mAb or anti-vimentin Ab, followed by the appropriate NH4Cl, and saturation of unspecific sites with PBS/1% BSA, cells were FITC-conjugated secondary Ab. incubated with SC5 and anti-vimentin (C-20) Abs. After washes, cells were Detection of apoptotic cells was performed by incubating the cells with subsequently incubated with goat anti-mouse PE (Beckman Coulter) and either annexin V directly coupled to FITC (5 ␮l/test; BD Biosciences) or donkey anti-goat FITC (Jackson ImmunoResearch Laboratories) secondary propidium iodide (PI; 1 ␮g/test; Sigma-Aldrich), according to the manu- Abs. Cells were transferred onto coverslips, and slides were mounted using facturer’s recommendations. Mowiol (Calbiochem). Analysis of cell labeling was performed on a Zeiss confocal microscope (LSM510; Zeiss). Two-dimensional gel electrophoresis and mass spectrometry protein sequencing Results PBMCs, cultured for 3 days in the presence of PHA (3 ␮g/ml; Sigma- SC5 mAb stains circulating SS lymphocytes and the Se´zary cell Aldrich), were washed in PBS and resuspended in Triton X-100 lysis buffer line HUT78 (20 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1% Triton X-100, 1 mM Na vanadate, 10 mM NaF, 1 mM PMSF, 1 ␮g/ml aprotinin, and 1 ␮g/ml The SC5 mAb raised against the functional cytotoxic NK tumor leupeptin). Following incubation for1hat4°Candcentrifugation, post- cell line YTindi was initially selected for its reactivity with both nuclear lysates were subjected to a preclearing step on protein A-Sepharose the immunizing tumor cell line and a minor population of normal beads. Immunoprecipitation was performed using SC5 mAb, or an isotype- PBLs (Fig. 1A, left panel). We previously reported that the PBMCs matched mAb as negative control, and protein G-Sepharose beads. After reactive with SC5 mAb, corresponding to 5–10% of gated lym- washes, precipitated proteins were released in 1% Triton X-100 lysis buffer phocytes, included CD4ϩ, CD8ϩ, as well as CD56ϩ cells (17). supplemented with 8 M urea at 37°C for 5 min. Samples were resolved by ϩ two-dimensional gel electrophoresis (isoelectrofocusing in the first dimen- Most of these SC5 circulating lymphocytes belonged to the ac- sion, followed by a SDS-8% PAGE in the second dimension), and proteins tivation/memory cell subset coexpressing CD45RO. We also dem- were detected by silver nitrate staining of the gel. Protein spots of interest onstrated that the expression of the SC5 mAb-reactive molecule were in-gel digested with trypsin, according to the method of Shevchenko et al. (21) on a MassPrep robot (Waters). Extracted peptide mixtures were was highly increased in peripheral blood T cells from patients with analyzed by a Voyager-DE STR MALDI-TOF instrument (Applied Bio- SS (16, 17), as shown in Fig. 1A (right panel), with PBMC from ϩ systems) and NanoHPLC (Switchos/Ultimate Pump; Dionex) on line with a representative patient presenting Ͼ90% of malignant CD4 654 CELL MEMBRANE VIMENTIN AS A MARKER OF SE´ ZARY CELLS

FIGURE 2. Purification of proteins coprecipitated by SC5 mAb. Ly- sates from PHA-activated PBMCs were subjected to SC5 precipitation. Proteins were separated by two-dimensional gel electrophoresis and de- tected by silver nitrate staining of the gel. The protein spots of interest were cut out from the gel and further processed for peptide sequencing. The identity of each polypeptide, as well as the position of the IgM H and L chains, and of the nonspecifically precipitated murine serum albumin are Downloaded from indicated.

complex in PHA-activated cells that at least encompasses ␣-acti- FIGURE 1. Se´zary cells present an increased reactivity toward SC5 nin 4, LSP1, and vimentin. To further assess which of these mAb. A, PBMCs were isolated by density gradient centrifugation from the blood of a healthy donor (left panel) or of a representative SS patient (right polypeptides was specifically recognized by SC5 mAb, immuno- http://www.jimmunol.org/ panel). Double labeling was performed by incubation of the cells with SC5 precipitates were prepared from HUT78 cell lysates using com- and anti-TCR␣␤ mAb, followed by goat anti-mouse IgM FITC and IgG1 mercially available Abs directed against each of the identified pro- PE secondary Abs, or anti-TCRV␤23 PE when indicated. Isotype-matched teins. An anti-vimentin Ab (C-20), raised toward the C-terminal controls were used for delimiting the positive regions in the two color moiety of the protein, was first tested (Fig. 3A). Immunoprecipi- histograms. B, Double staining of the Se´zary cell line HUT78 was per- tations with a control irrelevant Ab, and with SC5 mAb, were done formed as in A. in parallel. Immunoblot analysis with SC5 mAb allowed the de- tection of two protein bands at 53 and 60 kDa in the anti-vimentin cells. Furthermore, we found that the Se´zary cell line HUT78, precipitate (C-20). After stripping and reprobing of the blot with which was weakly recognized by an anti-TCR␣␤ mAb, was also the anti-vimentin Ab (C-20), an identical protein pattern was ob- by guest on September 28, 2021 dimly labeled by SC5 mAb (Fig. 1B). Thus, circulating SS T lym- tained. Conversely, Western blot analysis of SC5 mAb immune phocytes and the HUT78 cell line both presented an increased complex with SC5 or anti-vimentin Ab resulted in the visualization reactivity toward SC5 mAb when compared with normal resting T of a broad signal in the 50- to 60-kDa range (Fig. 3A). In contrast, cells. no signal was obtained when probing an anti-LSP1 or ␣-actinin 4 immunoprecipitate with SC5 mAb (Fig. 3B). Identification of vimentin as the Ag recognized by the SC5 mAb To definitely assess the specificity of SC5 mAb for vimentin, a We previously reported that stimulation of PBMCs with PHA re- cDNA construct coding for a GST-vimentin fusion protein was sulted in an increased cell staining with SC5 mAb when compared generated. Following expression and purification, a thrombin di- with resting cells, a maximum detection level being observed after gest was performed and the resulting cleaved polypeptide was sep- 72–96 h of activation (17). To identify SC5 mAb-reactive protein, arated by means of SDS-PAGE. Immunoblot analyses were then large-scale SC5 immunoprecipitates were obtained from lysates of realized using SC5 mAb or two anti-vimentin Abs of distinct or- PHA-treated cells. Immunoprecipitations with an isotype-matched igin (Fig. 3C). An identical recognition profile, with the detection mAb were done in parallel as control (data not shown). The im- of a 50- to 60-kDa protein, was obtained independently of the Ab munoprecipitates were further subjected to two-dimensional gel used for revelation. Altogether, these results demonstrated that electrophoresis, and proteins were detected by silver staining of the SC5 mAb was directed against vimentin. gel. Comparison between SC5 and control mAb protein patterns allowed the specific detection of three protein spots in SC5 mAb SC5 mAb specifically recognized vimentin when located on the immunoprecipitate (Fig. 2). Each protein spot was cut out from the outer side of the plasma membrane gel, digested with trypsin, and purified, and the resulting peptides The identification of SC5 mAb as an anti-vimentin Ab, combined were subjected to sequencing. Data-based assisted analysis of the with our previous observation that SC5 immunostaining was pos- obtained amino acid sequences led to the identification of all three itive on malignant Se´zary cells (16, 17), suggested that vimentin polypeptides. Thus, the 96-kDa protein was found to correspond to might present an atypical location at the plasma membrane in these the spectrin superfamily cytoskeleton protein ␣-actinin 4 (24). Fur- cells. To further investigate this possibility, immunofluorescence thermore, the proteins with an apparent molecular mass of 60 kDa, confocal microscopy was performed on the HUT78 cell line (Fig. and close isoelectric point of 4.7 and 5.1, were identified as the 4A). In a first set of experiments, cells were fixed and double lymphocyte-specific protein 1 (LSP1) (25) and vimentin, respec- stained with SC5 (red) and anti-vimentin (green) Abs (Fig. 4A, left tively. Note that the 70-kDa protein was identified as nonspecifi- panel). An overlay of the two labelings (yellow) showed that both cally precipitated murine serum albumin. Abs exerted a similar reactivity toward vimentin. Indeed, an im- The molecular characterization of the proteins present in SC5 munostaining polarized at one edge of the cell, and which appeared immunoprecipitate demonstrated that the Ab recognizes a protein to be tightly associated with the plasma membrane, was constantly The Journal of Immunology 655 Downloaded from http://www.jimmunol.org/

FIGURE 3. SC5 mAb is directed against vimentin. A, Immunoprecipi- tations were performed on HUT78 cell lysates using an irrelevant Ab (con- trol), SC5 mAb, or purified goat anti-vimentin Ab C-20. After separation by SDS-PAGE, proteins were first detected by Western blotting using SC5 mAb. The nitrocellulose membrane was then stripped and subjected to a by guest on September 28, 2021 second step of revelation with the anti-vimentin Abs C-20. B, Immuno- precipitations were performed on lysates of PHA-activated normal PBMCs using purified SC5, anti-LSP1, or anti-␣-actinin 4 mAb. Immune com- plexes were separated by SDS-PAGE, and proteins were detected by West- ern blotting using SC5 mAb. C, Vimentin was expressed as a GST fusion protein and digested with thrombin. The cut protein was then immunoblot- ted using SC5 mAb or the anti-vimentin Abs V9 or C-20. Protein detection was made with the appropriate HRP-conjugated secondary Ab and an ECL detection system. detected. An underneath diffuse intracellular labeling, linked to the one detected at the plasma membrane, was also observed. A second set of experiments was done, in which a step of per- meabilization of the cells was added after fixation and before im- munostaining (Fig. 4A, right panel). A complete and homogeneous FIGURE 4. SC5 mAb recognizes a pool of vimentin located on the staining of the plasma membrane was observed with both Abs outer face of the plasma membrane. A, The Se´zary cell line HUT78 was under these conditions. This membrane-bound staining was asso- either fixed (left panel) or fixed and permeabilized (right panel) before ciated to a deeper cytoplasmic labeling, when compared with the immunostaining. Double labeling was performed using SC5 mAb (red) and one observed in fixed and nonpermeabilized cells. These observa- the anti-vimentin Abs C-20 (green). An overlay of both staining is shown tions demonstrated that, according to the cell treatment (fixation (yellow) as well as a light microscopy view of the cell field. B, Flow alone vs fixation and permeabilization), one could distinguish a cytometry analyses were performed on HUT78 cells that were subjected to membrane-bound vimentin, presenting a polarized location, from a step of fixation alone (upper panel), fixation and permeabilization (mid- the total pool of cellular vimentin. dle panel), or left untreated (lower panel) before labeling. Cell staining was One limitation of the preparation of cells for their observation then performed with SC5 mAb or the commercial anti-vimentin Abs V9 or C-20, followed by the appropriate FITC-conjugated secondary Abs, as indi- by confocal microscopy relies on the use of a fixative agent before cated. Mouse (for SC5 and V9) or goat (for C-20) control Igs were used as immunoanalysis. This therefore did not allow us to establish negative controls. C, Cell surface immunolabeling was performed on resting or whether the immunofluorescent signal was associated with the PHA-activated lymphocytes from a healthy donor, or alternatively on Se´zary outer or the inner face of the plasma membrane. To overcome this cells isolated from a representative SS patient. Cell staining was performed difficulty, flow cytometry analyses were conducted using SC5 without any pretreatment of the cells, with SC5 mAb or the commercial anti- mAb as well as two distinct anti-vimentin Abs (Fig. 4B). In agree- vimentin Abs V9 or C-20, as described in B. Mouse (for SC5 and V9) or goat ment with the confocal microscopy data, all three Abs were found (for C-20) control Igs were used as negative controls. 656 CELL MEMBRANE VIMENTIN AS A MARKER OF SE´ ZARY CELLS to similarly label HUT78 cells when previously fixed, or fixed and permeabilized (Fig. 4B, upper and middle panels). However, a striking difference was observed when the immunostainings were performed on nontreated cells. Thus, while the cell labeling re- mained negative with both anti-vimentin Abs, a positive signal was still obtained with SC5 mAb (Fig. 4B, lower panel). A similar set of experiments performed on nontreated resting or PHA-activated normal PBMC, and on lymphocytes isolated from a representative SS patient (Fig. 4C), definitely confirmed the unique specificity of SC5 mAb. Indeed, as expected, the V9 (middle panel) and C-20 (right panel) anti-vimentin Abs remained unreactive with the cell types tested when not subjected to a fixation step. In contrast, SC5 mAb exhibited almost no stain with freshly isolated PBMCs from a healthy donor, but it showed an increased reactivity toward their PHA-activated counterparts, and more importantly with the Se´zary malignant cells obtained from a SS patient (Fig. 4C, left panel). Altogether, these results demonstrated that, unlike commercially available Abs, SC5 mAb specifically reacts with a peculiar fraction of vimentin. The observation that this pool of vimentin was con- Downloaded from centrated at one edge of the cell, and its association with the ex- tracellular leaflet of the plasma membrane, strongly suggested that FIGURE 5. A, Schematic representation of the vimentin constructs used vimentin was exported from the intracellular to the extracellular for transfection. C-terminal Flag-tagged full-length (FL) or deleted vimen- environment in both normal activated T cells and malignant SS tin constructs were ligated in frame in pcDNA3.1 expression vector. The lymphocytes. various domains of vimentin are as depicted. B, COS cells were transfected

with the empty expression vector pcDNA3.1 (Ctrl) or with the indicated http://www.jimmunol.org/ Mapping of SC5 mAb recognition site vimentin construct. After lysis, immunoprecipitation was performed using The observation that the anti-vimentin Abs used in this study ap- an anti-Flag mAb, and proteins were resolved by SDS-12% PAGE. The parently displayed different binding abilities toward the intra- and immunoprecipitated proteins were first detected by immunoblot analysis extracellular fractions of vimentin prompted us to map their cor- using SC5 mAb (top panel). The blot was then stripped and reprobed with responding recognition site. As the commercially available Ab V9 mAb (middle panel). A second dehybridization step was performed before analysis with the anti-Flag mAb (bottom panel). Arrows indicate the C-20 was generated against a peptide located within the C-terminal position of full-length or truncated vimentin. portion of the protein, we focused our attention on the specificity of recognition of V9 and SC5 mAb. cDNA constructs coding for full-length vimentin or deletion mutants, fused to a C-terminal by guest on September 28, 2021 Flag tag, were generated (Fig. 5A). Anti-Flag immunoprecipitates an inducible event specific to apoptotic cells. A T lymphocyte cell were prepared from lysates of transiently transfected COS cells line, developed by long-term growth of normal peripheral blood T and further analyzed by immunoblotting using V9 or SC5 mAb lymphocyte in the presence of IL-2, was tested for annexin V and (Fig. 5B). We observed that the vimentin mutants presenting a SC5 labeling. We observed that this cell line showed no reactivity deletion of the N-terminal head (mutant ⌬1–100) alone or together with SC5 mAb under basal conditions of cell growth (Fig. 6B). In with the 1A domain (mutant ⌬1–138) were still detected by V9 or contrast, a high number of cells was found to be positively labeled SC5 mAb. Moreover, the truncation of the full N-terminal half of with SC5 mAb when cells were deprived of IL-2 for 4 days (Fig. the protein (mutant ⌬1–227) resulted in a molecule recognized by 6C). Importantly, 45% of the SC5-positive cells were also stained V9 mAb, but which showed no more reactivity with SC5 mAb. We by annexin V. These results indicate that besides its reactivity with therefore concluded that both C-20 and V9 Abs are directed viable Se´zary malignant cells and activated T lymphocytes, the against the C-terminal moiety of vimentin, while SC5 mAb rec- anti-vimentin SC5 mAb might serve as an apoptotic marker as ognition site is part of the N-terminal half of the protein, and more efficiently as annexin V. precisely corresponds to a region located in the 1B domain. Autoantibodies against vimentin are present in the serum of SS SC5 mAb is reactive with apoptotic T lymphocyte cell line patients One of the earliest cellular events upon apoptosis is the appearance The unusual location of vimentin on the extracellular side of the of phosphatidylserine residues at the cell membrane, which can be plasma membrane of SS lymphocytes led to the possibility of an revealed by the binding of annexin V. To determine whether the autoantibodies production by Se´zary patients. This hypothesis was presence of vimentin at the cell surface of SS lymphocytes is a tested by performing Western blot analysis on vimentin using the consequence of an apoptotic cell status, we simultaneously tested serum corresponding to 10 Se´zary patients. The serum of two pa- annexin V and SC5 binding on the HUT78 cell line. The PI uptake, tients was found to recognize vimentin, at a dilution of 1/100 (Fig. reflecting the necrotic cell status, was also estimated. The results 7). In contrast, no signal was obtained when using the serum of shown in Fig. 6A indicated that the HUT78 cells were negative for healthy donors (Fig. 7, control lane), inferring the specificity of the annexin V binding, and for PI uptake, whereas they presented a detection. In addition, Abs against vimentin were detected for a significant amount of extracellular vimentin as identified by SC5 third patient when the serum was used at a 1/50 dilution for im- mAb. These data demonstrated that vimentin was detected at the munoblotting (data not shown). cell surface of viable Se´zary cells, and that this location did not derive from a process of cell death. Discussion We next investigated whether the relocation of vimentin from Se´zary cells correspond, in most patients, to lymphocytes exhib- the intracellular compartment to the extracellular matrix could be iting a CD4ϩCD45ROϩ phenotype. It is generally admitted that The Journal of Immunology 657

FIGURE 7. Presence of anti-vimentin autoantibodies in the serum of SS patients. Purified GST-vimentin was subjected to digestion with thrombin and separated by means of SDS-PAGE. After transfer onto a nitrocellulose membrane, samples were subjected to immunoblotting using SC5 mAb or sera prepared from the blood of SS patients (serum 1 and 2). The control corresponds to the use of the serum obtained from a healthy individual, and is representative of 10 tested. Protein revelation was performed with per- oxidase-conjugated secondary Abs and an ECL detection system.

Unlike the commercial anti-vimentin Abs tested, SC5 mAb ex- Downloaded from erted a unique ability of Ag recognition. Thus, identical patterns of immunolabeling were obtained on the Se´zary cell line HUT78 when fixed and/or permeabilized, regardless of the anti-vimentin FIGURE 6. SC5 mAb labels apoptotic T lymphocytes. A, HUT78 cells were double labeled with SC5 mAb and either PI (left panel) or annexin V Ab used for detection (Fig. 4). However, a striking difference was FITC (right panel). The percentages of single- and double-labeled cells are observed when cells were not subjected to any treatment before

indicated for each cell group. B and C, An IL-2-dependent T cell line was immunostaining. Under such conditions, SC5 mAb was the only http://www.jimmunol.org/ obtained by long-term culture of PBMCs in the presence of IL-2. Immu- Ab able to interact with a pool of vimentin otherwise not detected nolabeling was then performed with SC5 mAb and annexin V FITC on by the commercial Abs (Fig. 4, B and C). We therefore established cells grown with IL-2 (B) or after induction of apoptosis by deprivation of that, while recognizing intracellular vimentin, SC5 mAb could also IL-2 for 4 days (C). specifically target a fraction of the protein located on the outer face of the plasma membrane. By generating vimentin deletion mu- tants, we established that, while V9 and C-20 Abs were directed the lack of CD7 and CD26 expression characterizes the malignant against regions present within the C-terminal moiety of the pro- lymphocytes, although controversial reports were made concern- tein, SC5 mAb recognition site was located in the N-terminal ing the former receptor (14). Until now, and apart from the iden- half of the molecule (Fig. 5). This therefore suggests that the by guest on September 28, 2021 tification of specific TCRV␤ rearrangements, only the KIR3DL2/ protein sequence recognized by SC5 mAb was exposed in both CD158k cell membrane structure was described to phenotypically the intra- and extracellular form of the protein, while the one identify skin and blood malignant Se´zary cells, its expression cur- targeted by the commercial Abs became unaccessible in the rently allowing their efficient isolation from the patients’ blood soluble form, most likely as a consequence of conformational (14). To further identify cell surface receptors expressed by the ϩ modifications. Because SC5 mAb was found to specifically re- malignant CD4 CTCL cells, we developed a mAb called SC5. act with viable Se´zary cells or activated T lymphocytes (16, 17), We previously established that it recognized a plasma membrane the detection of extracellular vimentin on living cells using this Ag, which detection was increased on PBMCs from patients with Ab might represent a powerful tool for the identification of SS, when compared with PBMCs isolated from healthy individuals malignant transformed lymphocytes. (16, 17). In addition, we observed that the SC5 mAb-reactive mol- It is well established that vimentin is an intermediate filament ecule was mainly located in the cytoplasm of normal resting T protein that participates to the formation of the intracellular cy- lymphocytes, and that it rapidly became detectable at the cell sur- toskeletal structure. The comparison of the confocal microscopy face upon T cell activation (16, 17). Our first attempts to charac- data obtained on fixed cells vs fixed and permeabilized cells evi- terize the SC5 mAb-corresponding molecule, by performing im- denced the existence of a vimentin pool that is polarized at one munoprecipitation on lysates of cell surface biotinylated activated A T lymphocytes, led to the detection of a 96-kDa protein (17). In edge of the Se´zary cells (Fig. 4 ). Because this immunofluorescent this study, we report further biochemical studies indicating that signal was observed with both SC5 mAb and the purified poly- SC5 mAb precipitates a protein complex containing a 96-kDa and clonal goat anti-vimentin Abs C-20, it cannot only reflect the pres- two 60-kDa proteins from PHA-activated lymphocytes (Fig. 2). ence of vimentin on the external side of the cell membrane. Rather, These polypeptides were identified by peptide sequencing as ␣-ac- it might also correspond to a pool of protein undergoing a trans- tinin 4, LSP-1, and vimentin, respectively. In contrast to what we location from the intracellular to the extracellular compartment. initially thought, we established that the SC5 mAb is in fact di- The detection of vimentin on the outer leaflet of the plasma mem- rected against the 60-kDa molecule, exerting an isoelectric point of brane of Se´zary cells might thus result from a dynamic and in- 5.1 and corresponding to vimentin. In light of these new data, the duced process of exocytosis. Interestingly, it has been recently previously detected 96-kDa protein most likely corresponded to reported that vimentin, which is highly abundant in human acti- the coprecipitated ␣-actinin 4. The generation of a GST-vimentin vated macrophages, is secreted from ex vivo long-term cultured fusion protein provides definitive evidence for SC5 mAb specific- monocyte-derived macrophages (26). The pool of secreted vimen- ity toward vimentin. Indeed, the use of SC5 mAb in Western blot tin was then involved in the settings of two major activated mac- analysis allows the specific detection of vimentin as efficiently as rophage functions, namely bacterial killing and the generation of known anti-vimentin Abs (Fig. 3). oxidative metabolites. Furthermore, the release of vimentin in the 658 CELL MEMBRANE VIMENTIN AS A MARKER OF SE´ ZARY CELLS extracellular environment was found to be associated to its loca- In the past few years, the serological identification of recombi- tion on the extracellular side of the membrane of monocyte-de- nantly expressed has increasingly been used to screen new rived macrophages. More recently, Xu et al. identified the endo- tumor Ags. This serological identification of recombinantly ex- thelial cell-specific Ab PAL-E as an anti-vimentin Ab (27). PAL-E pressed genes method has been demonstrated to be a powerful tool mAb shares some common features with SC5 mAb. Indeed, to identify Ags such as tumor-suppressor genes, oncogenes, can- PAL-E was found to stain live endothelial cells (28), and this la- cer-testis genes, and differentiation Ags (33, 34). Interestingly, vi- beling was polarized along the luminal endothelial surface (29, mentin has been isolated as a potential CTCL-associated Ag (35). 30), thus resembling what we observed upon SC5 labeling of Se´z- In this study, our findings clearly confirm that vimentin represents ary cells. Further investigation established that blood endothelial a tumor Ag in SS patients. Interestingly, they also suggest that the cells exerted an unexpected vimentin metabolism, because the pro- presence of anti-vimentin autoantibodies in the serum of these pa- tein was expressed and secreted as a dimer. Strikingly, PAL-E was tient might result from the presence of the protein at the Se´zary cell found to be reactive with dimeric vimentin, but hardly detected the surface rather than from its overexpression, as reported for prostate protein in its monomeric form. In addition, further investigation carcinoma (36) and endometrial neoplasm (37). However, further demonstrated that the commercial V9 mAb was also able to detect studies are needed to determine whether the presence of autoanti- the vimentin dimers expressed by endothelial cells, or following bodies is concomitantly associated with the detection of soluble secretion. At this point, it is important to mention that we never vimentin in the serum of Se´zary patients. detected vimentin as a dimeric protein following immunoprecipi- In conclusion, we report for the first time that vimentin is lo- tation with SC5, V9, or C-20 Abs on lysates from Se´zary cells or cated at the cell membrane surface of malignant Se´zary cells and from activated normal T lymphocytes. In addition, only soluble apoptotic T lymphocytes. This observation was rendered possible Downloaded from monomeric vimentin was detected from the culture medium of by using the unique specificity of the anti-vimentin SC5 mAb. We HUT78 Se´zary cell line after immunoprecipitation (data not further suggest that anti-vimentin autoantibodies are present in the shown). Thus, while PAL-E specifically allows the detection of a serum of SS patients, which might represent a useful marker for dimeric pool of vimentin processed by blood endothelial cells, SC5 diagnosis or prognostic purposes. mAb remains the only mAb allowing the detection of extracellular vimentin at the surface of viable Se´zary cells or of activated nor- Acknowledgments http://www.jimmunol.org/ mal T lymphocytes. Further studies will now be needed to deter- We thank Abdelali Jalil and Fatia Mami-Chouaib (Institut National de la mine the role of secreted vimentin in the pathophysiology of SS. Sante´et de la Recherche Me´dicale Unite´487, Gustave Roussy Institute, Several hypotheses were made to explain how vimentin could Villejuif, France) for their help with confocal microscopy cell examination, be secreted, although it lacks a signal sequence. It was proposed and Jacques Bertoglio and Catherine Froin (Institut National de la Sante´et that its highly positively charged N-terminal sequence could react de la Recherche Me´dicale Unite´461, Chatenay-Malabry, France) for their with the endoplasmic reticulum hydophobic core of the lipid bi- assistance with the generation of the GST-vimentin construct. layer and direct the protein into membranes (26). In addition, the Disclosures C terminus moiety of vimentin contains a di-acidic motif preceded

The authors have no financial conflict of interest. by guest on September 28, 2021 by a Y-X-X-␾ motif (in which ␾ is a hydrophobic residue), known to be required for the selective export of some proteins from the References endoplasmic reticulum into the Golgi apparatus and found in many 1. Willemze, R., H. Kerl, W. Sterry, E. Berti, L. Cerroni, S. Chimenti, membrane-associated proteins (26). Although allowing the detec- J. L. Diaz-Perez, M. L. Geerts, M. Goos, R. Knobler, et al. 1997. EORTC clas- tion of the entire intracellular vimentin pool by confocal micros- sification for primary cutaneous lymphomas: a proposal from the Cutaneous Lymphoma Study Group of the European Organization for Research and Treat- copy or biochemical approaches, the conventional anti-vimentin ment of Cancer. Blood 90: 354–371. Abs did not highlight the presence of extracellular vimentin on 2. Rook, A. H., S. L. Gottlieb, J. T. Wolfe, B. R. Vowels, S. S. Sood, Z. Niu, Se´zary cells. Similarly, the quantification of vimentin transcripts S. R. Lessin, and F. E. Fox. 1997. Pathogenesis of cutaneous T-cell lymphoma: implications for the use of recombinant cytokines and photopheresis. Clin. Exp. within the cells would not have answered the question of vimentin Immunol. 107(Suppl. 1): 16–20. relocation from the cytoplasm to the outer side of the plasma mem- 3. Van Doorn, R., R. Dijkman, M. H. Vermeer, J. J. Out-Luiting, brane. Because of its original specificity, SC5 mAb currently rep- E. M. van der Raaij-Helmer, R. Willemze, and C. P. Tensen. 2004. Aberrant expression of the tyrosine kinase receptor EphA4 and the transcription factor resents a unique tool for the detection of vimentin translocation in twist in Sezary syndrome identified by analysis. Cancer Res. 64: Se´zary cells. It should be mentioned that another cytoplasmic pro- 5578–5586. 4. Mao, X., G. Orchard, D. M. Lillington, R. Russell-Jones, B. D. Young, and tein, T-plastin, involved in the regulation of assembly and S. J. Whittaker. 2003. Amplification and overexpression of JUNB is associated cellular motility, has also been identified as a potential Se´zary with primary cutaneous T-cell lymphomas. Blood 101: 1513–1519. cell-specific marker (6), pointing to a strong involvement of cy- 5. Mao, X., G. Orchard, D. M. Lillington, F. J. Child, E. C. Vonderheid, P. C. Nowell, M. Bagot, A. Bensussan, R. Russell-Jones, B. D. Young, and toskeleton-related proteins in Se´zary cells. S. J. Whittaker. 2004. BCL2 and JUNB abnormalities in primary cutaneous lym- Finally, we found that, apart from its ability to react with ex- phomas. Br. J. Dermatol. 151: 546–556. tracellular vimentin on viable Se´zary cells, SC5 mAb is also ca- 6. Su, M. W., I. Dorocicz, W. H. Dragowska, V. Ho, G. Li, N. Voss, R. Gascoyne, and Y. Zhou. 2003. Aberrant expression of T-plastin in Sezary cells. Cancer Res. pable of detecting cell surface vimentin on apoptotic T lympho- 63: 7122–7127. cytes. Apoptosis is characterized by cellular and nuclear shrinkage, 7. Dummer, R., F. O. Nestle, E. Niederer, E. Ludwig, E. Laine, H. Grundmann, P. Grob, and G. Burg. 1999. Genotypic, phenotypic and functional analysis of cytoplasmic blebbing, condensation of nuclear chromatin, and CD4ϩCD7ϩ and CD4ϩCD7Ϫ T lymphocyte subsets in Sezary syndrome. Arch. fragmentation of nuclear DNA (31). Thus, besides the analysis of Dermatol. Res. 291: 307–311. apoptosis-associated proteins (32), the detection of vimentin at the 8. Jones, D., N. H. Dang, M. Duvic, L. T. Washington, and Y. O. Huh. 2001. Absence of CD26 expression is a useful marker for diagnosis of T-cell lymphoma cell surface could represent an additional tool to evaluate an early in peripheral blood. Am. J. Clin. Pathol. 115: 885–892. stage of cell death. It is important to emphasize that the reactivity 9. Rappl, G., J. M. Muche, H. Abken, W. Sterry, W. Tilgen, S. Ugurel, and ϩ Ϫ of SC5 mAb with Se´zary cells is not a consequence of their apo- U. Reinhold. 2001. CD4 CD7 T cells compose the dominant T-cell clone in the peripheral blood of patients with Sezary syndrome. J. Am. Acad. Dermatol. 44: ptotic status. Indeed, we demonstrated that these cells were neg- 456–461. ative for annexin V binding and PI uptake (Fig. 6A), and that they 10. Vonderheid, E. C., R. D. Bigler, A. Kotecha, C. M. Boselli, S. R. Lessin, M. G. Bernengo, and M. Polansky. 2001. Variable CD7 expression on T cells in failed to react with Abs directed against cytoplasmic molecules, the leukemic phase of cutaneous T cell lymphoma (Sezary syndrome). J. Invest. such as ␣-actinin 4 (data not shown). Dermatol. 117: 654–662. The Journal of Immunology 659

11. Yagi, H., Y. Tokura, F. Furukawa, and M. Takigawa. 1996. CD7-positive Sezary 25. Jongstra, J., G. F. Tidmarsh, J. Jongstra-Bilen, and M. M. Davis. 1988. A new syndrome with a Th1 cytokine profile. J. Am. Acad. Dermatol. 34: 368–374. lymphocyte-specific gene which encodes a putative Ca2ϩ-binding protein is not 12. Bernengo, M. G., M. Novelli, P. Quaglino, F. Lisa, A. De Matteis, P. Savoia, expressed in transformed T lymphocyte lines. J. Immunol. 141: 3999–4004. ϩ Ϫ N. Cappello, and M. T. Fierro. 2001. The relevance of the CD4 CD26 subset 26. Mor-Vaknin, N., A. Punturieri, K. Sitwala, and D. M. Markovitz. 2003. Vimentin in the identification of circulating Sezary cells. Br. J. Dermatol. 144: 125–135. is secreted by activated macrophages. Nat. Cell Biol. 5: 59–63. 13. Bagot, M., A. Moretta, S. Sivori, R. Biassoni, C. Cantoni, C. Bottino, ϩ 27. Xu, B., R. M. de Wall, N. Mor-Vaknin, C. Hibbard, D. M. Markovitz, L. Boumsell, and A. Bensussan. 2001. CD4 cutaneous T-cell lymphoma cells and M. L. Kahn. 2004. The endothelial cell-specific anitbody PAL-E identifies express the p140-killer cell immunoglobulin-like receptor. Blood 97: 1388–1391. a secreted form of vimentin in the blood vasculature. Mol. Cell. Biol. 24: 14. Poszepczynska-Guigne, E., V. Schiavon, M. D’Incan, H. Echchakir, P. Musette, 9198–9206. N. Ortonne, L. Boumsell, A. Moretta, A. Bensussan, and M. Bagot. 2004. 28. Ades, E. W., F. J. Candal, R. A. Swerlick, V. G. George, S. Summers, CD158k/KIR3DL2 is a new phenotypic marker of Sezary cells: relevance for the D. C. Bosse, and T. J. Lawley. 1992. HMEC-1: establishment of an immortalized diagnosis and follow-up of Sezary syndrome. J. Invest. Dermatol. 122: 820–823. human microvascular endothelial cell line. J. Invest. Dermatol. 99: 683–690. 15. Wechsler, J., M. Bagot, M. Nikolova, S. Parolini, N. Martin-Garcia, L. Boumsell, A. Moretta, and A. Bensussan. 2003. Killer cell immunoglobulin-like receptor 29. Page, C., M. Rose, M. Yacoub, and R. Pigott. 1992. Antigenic heterogeneity of expression delineates in situ Sezary syndrome lymphocytes. J. Pathol. 199: vascular endothelium. Am. J. Pathol. 141: 673–683. 77–83. 30. Schlingemann, R. O., G. M. Dingjan, J. J. Emeis, J. Blok, S. O. Warnaar, and 16. Nikolova, M., M. Bagot, L. Boumsell, and A. Bensussan. 2002. Identification of D. J. Ruiter. 1985. Monoclonal antibody PAL-E specific for endothelium. Lab. cell surface molecules characterizing human cutaneous T-cell lymphomas. Leuk. Invest. 52: 71–76. Lymphoma 43: 741–746. 31. Martin, S. J., D. R. Green, and T. G. Cotter. 1994. Dicing with death: dissecting 17. Nikolova, M., A. Tawab, A. Marie-Cardine, M. Bagot, L. Boumsell, and the components of the apoptosis machinery. Trends Biochem. Sci. 19: 26–30. A. Bensussan. 2001. Increased expression of a novel early activation surface 32. Brockstedt, E., A. Rickers, S. Kostka, A. Laubersheimer, B. Dorken, membrane receptor in cutaneous T cell lymphoma cells. J. Invest. Dermatol. 116: B. Wittmann-Liebold, K. Bommert, and A. Otto. 1998. Identification of apopto- 731–738. sis-associated proteins in a human Burkitt lymphoma cell line: cleavage of het- 18. Linsley, P. S., J. Bradshaw, J. Greene, R. Peach, K. L. Bennett, and R. S. Mittler. erogeneous nuclear ribonucleoprotein A1 by caspase 3. J. Biol. Chem. 273: 1996. Intracellular trafficking of CTLA-4 and focal localization towards sites of 28057-28064. Downloaded from TCR engagement. Immunity 4: 535–543. 33. Scanlan, M. J., I. Gout, C. M. Gordon, B. Williamson, E. Stockert, A. O. Gure, 19. Linsley, P. S., and P. Golstein. 1996. Lymphocyte activation: T-cell regulation by D. Jager, Y. T. Chen, A. Mackay, M. J. O’Hare, and L. J. Old. 2001. Humoral CTLA-4. Curr. Biol. 6: 398–400. immunity to human breast cancer: antigen definition and quantitative analysis of 20. Goyert, S. 1987. CD68 workshop panel report. In Leukocyte Typing VI. Garland mRNA expression. Cancer Immun. 1: 4. Publishing, New York, pp. 1015–1016. 34. Old, L. J. 2001. Cancer/testis (CT) antigens: a new link between gametogenesis 21. Shevchenko, A., M. Wilm, O. Vorm, and M. Mann. 1996. Mass spectrometric and cancer. Cancer Immun. 1: 1. sequencing of proteins silver-stained polyacrylamide gels. Anal. Chem. 68: 850–858. 35. Hartmann, T. B., D. Thiel, R. Dummer, D. Schadendorf, and S. Eichmuller. 2004. SEREX identification of new tumor-associated antigens in cutaneous T-cell lym-

22. Marie-Cardine, A., H. Kirchgessner, C. Eckerskorn, S. C. Meuer, and http://www.jimmunol.org/ B. Schraven. 1995. Human T lymphocyte activation induces tyrosine phosphor- phoma. Br. J. Dermatol. 150: 252–258. ylation of ␣-tubulin and its association with the SH2 domain of the p59fyn protein 36. Lang, S. H., C. Hyde, I. N. Reid, I. S. Hitchcock, C. A. Hart, A. A. Bryden, tyrosine kinase. Eur. J. Immunol. 25: 3290–3297. J. M. Villette, M. J. Stower, and N. J. Maitland. 2002. Enhanced expression of 23. Da Silva, A. J., Z. Li, C. de Vera, E. Canto, P. Findell, and C. E. Rudd. 1997. vimentin in motile prostate cell lines and in poorly differentiated and metastatic Cloning of a novel T-cell protein FYB that binds FYN and SH2-domain-con- prostate carcinoma. Prostate 52: 253–263. taining leukocyte protein 76 and modulates interleukin 2 production. Proc. Natl. 37. Papadopoulos, N., A. Kotini, A. Cheva, T. Jivannakis, M. Lambropoulou, Acad. Sci. USA 94: 7493–7498. M. Bobos, S. Vavetsis, and D. Tamiolakis. 2002. Immunohistochemical expres- 24. Davison, M. D., and D. R. Critchley. 1988. ␣-Actinins and the DMD protein sion of vimentin and secretory component antigens in endometrial hyperplasia contain spectrin-like repeats. Cell 52: 159–160. and neoplasia. Eur. J. Gynaecol. Oncol. 23: 411–414. by guest on September 28, 2021