Human T Lymphotropic Virus Type 1 Increases T Lymphocyte Migration by Recruiting the Cytoskeleton Organizer CRMP2 This information is current as of September 28, 2021. Michel Varrin-Doyer, Adeline Nicolle, Romain Marignier, Sylvie Cavagna, Claire Benetollo, Eric Wattel and Pascale Giraudon J Immunol 2012; 188:1222-1233; Prepublished online 6 January 2012; Downloaded from doi: 10.4049/jimmunol.1101562 http://www.jimmunol.org/content/188/3/1222 http://www.jimmunol.org/ Supplementary http://www.jimmunol.org/content/suppl/2012/01/06/jimmunol.110156 Material 2.DC1 References This article cites 83 articles, 31 of which you can access for free at: http://www.jimmunol.org/content/188/3/1222.full#ref-list-1

Why The JI? Submit online. by guest on September 28, 2021

• 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

*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 © 2012 by The American Association of Immunologists, Inc. All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Human T Lymphotropic Virus Type 1 Increases T Lymphocyte Migration by Recruiting the Cytoskeleton Organizer CRMP2

Michel Varrin-Doyer,*,† Adeline Nicolle,*,† Romain Marignier,*,† Sylvie Cavagna,*,† Claire Benetollo,*,† Eric Wattel,‡ and Pascale Giraudon*,†

Recruitment of virus-infected T lymphocytes into the CNS is an essential step in the development of virus-associated neuro- inflammatory diseases, notably myelopathy induced by retrovirus human T leukemia virus-1 (HTLV-1). We have recently shown the key role of collapsin response mediator protein 2 (CRMP2), a phosphoprotein involved in cytoskeleton rearrangement, in the control of human lymphocyte migration and in brain targeting in animal models of virus-induced neuroinflammation.

Using lymphocytes cloned from infected patients and chronically infected T cells, we found that HTLV-1 affects CRMP2 activity, Downloaded from resulting in an increased migratory potential. Elevated CRMP2 expression accompanies a higher phosphorylation level of CRMP2 and its more pronounced adhesion to tubulin and actin. CRMP2 forms, a full length and a shorter, cleaved one, are also affected. Tax transfection and extinction strategies show the involvement of this viral protein in enhanced full-length and active CRMP2, resulting in prominent migratory rate. A role for other viral proteins in CRMP2 phosphorylation is suspected. Full-length CRMP2 confers a migratory advantage possibly by preempting the negative effect of short CRMP2 we observe on T lymphocyte migration. In addition, HTLV-1–induced migration seems, in part, supported by the ability of infected cell to increase the http://www.jimmunol.org/ proteosomal degradation of short CRMP2. Finally, expression in CD69+ cells selected from patients suggests that HTLV-1 has the capacity to influence the CRMP2/PI3K/Akt axis thus to positively control cytoskeleton organization and lymphocyte migration. Our data provide an additional clue to understanding the infiltration of HTLV-1–infected lymphocytes into various tissues and suggest that the regulation of CRMP2 activity by virus infection is a novel aspect of neuroinflammation. The Journal of Immunology, 2012, 188: 1222–1233.

ymphocyte infiltration in the CNS is at the basis of neu- paraparesis (HAM/TSP). This human retrovirus mainly targets

roinflammatory diseases, including those associated with CD4 and CD8 T lymphocytes and is the etiological agent of both by guest on September 28, 2021 virus infection. Besides CNS infection, infected immune adult T cell leukemia (ATL) (2) and HAM/TSP (3). The pre- L + + cells can affect neurons and glial cells through the bystander ef- dominant viral reservoir in the peripheral blood is CD4 CD25 fect of secreted inflammatory molecules, leading progressively to T cells (4). How HTLV-1 causes neurologic disorders is not demyelination and axonal loss (1). Dissection of the complex completely understood, but several works have focused on the mechanisms that control infiltration and migration of infected crucial role of infected T cells in pathogenesis. It is thought that, T lymphocytes into the CNS is of importance in understanding in HAM/TSP patients, HTLV-1–infected lymphocytes migrate the pathophysiology of virus-induced neuroinflammatory diseases. from the periphery into the CNS, because lymphocytes of the One prototype of such a disease is human T cell lymphotropic cerebrospinal fluid and PBMC shared the same HTLV-1 integra- virus type-1 (HTLV-1)–associated myelopathy/tropical spastic tion site in cellular DNA (5). Detection of HTLV-1 genome and the viral protein Tax in inflammatory T lymphocytes infiltrated in *INSERM U1028, Centre National de la Recherche Scientifique, Unite´ Mixte de the patients’ CNS has identified T lymphocytes as the main virus Recherche 5292, Centre de Recherche en Neuroscience de Lyon, Equipe Neuroon- reservoir in the brain and the main effectors of pathological cologie-Neuroinflammation, F-69000 Lyon, France; †University of Lyon, F-69000 Lyon, France; and ‡Oncovirologie et Biothe´rapies, Universite´ Claude Bernard Lyon 1, changes (6–10). In fact, Tax has the ability to promote or alter the Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche 5239, expression and activity of several cellular factors involved in Centre Le´on Be´rard, 69008 Lyon, France, France T cell behavior (11–13), as such, may contribute to pathogenic Received for publication May 27, 2011. Accepted for publication November 30, mechanism. In this context, we have previously shown that HTLV- 2011. 1–infected T cells disturb the metabolism of astrocytes through This work was supported by Agence Nationale de la Recherche and Agence de Recherche sur La Scle´rose en Plaques. the activity of TNF-a (14). Secreted by infected T lymphocytes, Address correspondence and reprint requests to Dr. Pascale Giraudon, INSERM TNF-a reduced the glutamate transporter EAAT2/GLT1 at as- U1028, Centre National de la Recherche Scientifique, Unite´ Mixte de Recherche trocyte membrane and, consequently, increased the extracellular 5292, Lyon Neuroscience Research Center, Faculty of Medicine Laennec, Rue Guil- level of glutamate, an excitotoxic amino acid deleterious for laume Paradin, 69372 Lyon, France. E-mail address: [email protected] neurons and oligodendrocytes at a high level. Other virus proteins The online version of this article contains supplemental material. are also involved in pathogenesis, notably the p8 protein, that Abbreviations used in this article: ATL, adult T cell leukemia; BMP, bone morpho- genetic protein; CRMP2, collapsin response mediator protein 2; HAM/TSP, human promote T cell contact and enhance virus transmission (15). Given T leukemia virus 1-associated myelopathy/tropical spastic paraparesis; HD, healthy the crucial role of infected T lymphocytes in pathogenesis, facil- blood donor; HTLV-1, human T leukemia virus 1; pep4, peptide 4; qPCR, quantita- itation of their recruitment into the CNS through changes induced tive PCR; shRNA, short hairpin RNA; siRNA, small interfering RNA. by HTLV-1 in the cell machinery of locomotion could be an im- Copyright Ó 2012 by The American Association of Immunologists, Inc. 0022-1767/12/$16.00 portant phase of the disease. www.jimmunol.org/cgi/doi/10.4049/jimmunol.1101562 The Journal of Immunology 1223

During neuroinflammation, leukocytes emigrate from the blood- microbeads. Cells were washed, suspended in 500 ml medium, and put 4 stream to the CNS across the blood–brain barriers through ei- through magnetic columns. Selected cells (1–5 3 10 cells) were used ther endothelial cells of inflamed meningeal veins or epithelial either for RNA amplification and microarray analysis, for quantitative PCR (qPCR) analysis on unamplified RNA, or for transmigration assays. cells of the choroids plexuses (16, 17). The continuous T lym- phocyte extravasation, migration into, and crawling within neural Antibodies tissue are orchestrated by the interplay of multiple adhesion re- CRMP2-A/B, generally termed CRMP2, was detected using purified ceptors whose expression and avidity are modulated by cytokines polyclonal rabbit Abs directed against peptide 4 (pep4) and C-ter epitopes, and chemokines (18–20). Activation of G protein-coupled receptor as previously described (29) and monoclonal C4G Ab from IBL-Japan. by chemokines induces a motile phenotype characterized by the CRMP2-C-ter and C4G Abs recognized only the full-length CRMP2. CRMP2–pep-4 Ab recognized both the full-length (62 kDa) and cleaved formation of lamellipodia at the front of migration and uropod at (58 kDa) CRMP2. Polyclonal anti–CRMP2-A Ab was described previ- the rear of T lymphocyte (21, 22). In parallel, a drastic reconfigu- ously (29). Ab specific of phosphorylated CRMP2 forms CRMP2-pSer522 ration of the microtubule and intermediate filament cytoskeletons and CRMP2-pThr509/514 was from Kinasource (Dundee, U.K.); anti– allows lymphocyte locomotion. However, the mechanism associ- CRMP2-pY479 was produced as described previously (25). Anti-GAPDH Ab was from Chemicon. Ab to vimentin was from Calbiochem (IF01), to ated with lymphocyte motility per se is not totally defined. To tubulin from Sigma-Aldrich (T8660), and to actin was from Santa Cruz extend our knowledge in this field, we recently focused our work Biotechnology (SC8432). Ab to Tax was from the National Institutes of on the phosphoprotein collapsin response mediator protein 2 Health (158A51-42). Small interfering RNA (siRNA) anti-CRMP2 and (CRMP2), first described to modulate the microtubule reorgani- control were used as described previously (24). zation during neural growth cone advance under semaphorin signal Flow cytometry

(23). Our experiments have highlighted the importance of CRMP2 Downloaded from Cells were washed in PBS and incubated (15 min at room temperature) with in immune cells, notably in migratory activity, in both physiolog- PE-labeled anti-CD69 (BD Pharmingen) or anti-CD4 (BD Biosciences) ical and neuroinflammatory situations. We showed that CRMP2 PE–cyanine 5-labeled Ab. Cells were then washed and fixed (4% parafor- was expressed in virtually all PBMC and had a crucial role in maldehyde in PBS [pH 7.4], 30 min at 4˚C), washed again, and processed T lymphocyte polarization and migration (24). Investigation of the further for intracellular CRMP2 protein detection. Purified polyclonal rabbit molecular mechanisms underlying CRMP2 activity in T lympho- anti-CRMP2 Ab (pep4) has been optimized for flow cytometry [signal re- duction following treatment with CRMP2-pep4 or siRNA-CRMP2 (24)]. cyte indicated that CRMP2 transduced CXCL12/SDF1 chemokine Cells were incubated with anti-CRMP2 Ab (PBS and 0.5% saponin, 30 min http://www.jimmunol.org/ signal. Differential phosphorylation by glycogen synthase - at room temperature), washed (PBS 0.05% saponin), incubated with FITC- 3b and Yes was supposed to modulate the contribution of CRMP2 labeled goat F(ab9)2 anti-rabbit Ab (20 min at room temperature; Roche), to cytoskeletal reorganization during chemokine-induced T cell washed (PBS 0.05% saponin), and analyzed by flow cytometry (EPICS XL; Beckman Coulter). migration (25). In addition, we showed that in mouse models of virus-induced neuroinflammation, the elevated CRMP2 expression Immunofluorescence in blood lymphocytes correlated with CNS infiltration and clinical T lymphocytes were grown on glass treated with polyLysin and then fixed signs, suggesting the potential use of CRMP2 as a peripheral in- with cold acetone for 10 min. Cells were incubated with anti-Tax and dicator of neuroinflammation (26). Extending our analysis to hu- -CRMP2pep4 Abs for 1 h at 37˚C, washed, and then incubated with FITC- man, we detected high CRMP2 expression in T lymphocytes of labeled anti-IgG (Alexa 488; Molecular Probes) for 1 h at room temperature. by guest on September 28, 2021 After washing, cells were examined with the Imager.Z1 Zeiss microscope. HTLV-1–infected patients suffering from neurologic disease com- pared with asymptomatic virus carriers (24). We now examined Tax and CRMP2 expression whether and how HTLV-1 modifies T cell motility through alter- T lymphocytes (4 3 106) were infected with the TRIP-Tax vector plasmid ation of CRMP2 activity. The present data demonstrate that the (600 ng in 50 ml, 30 min at 37˚C) and cultured for 2 d (30). CRMP2, Tax virus enhanced T lymphocyte migration by sustaining the presence expression, and the percentage of transfected cells were evaluated by GFP of active CRMP2 form in infected cells, in part though Tax ac- fluorescence and Tax immunodetection using flow cytometry. pSG5M- Tax1, CRMP2-FLAG-wt, and the C terminally truncated DC503 plas- tivity. More generally, we point to the mechanisms exploited by mids have been described previously (29, 31). For the transfection, Jurkat a virus to intensify, in infected patients, the migratory potential T cells were transfected with CRMP2-FLAG-wt, CRMP2-FLAG-DC503, of T lymphocyte. empty-FLAG, and empty-GFP plasmids using Amaxa Nucleofector tech- nology (Ko¨ln, Germany), according to the manufacturer’s instructions. T cells were used 24 and 48 h after transfection. Transfected cells were Materials and Methods visualized by immunostaining with anti-FLAG or anti-GFP Abs. The Cells percentage of transfection reached 40–50%. Tax knockdown was per- formed using lentiviral short hairpin RNA (shRNA) anti-Tax cloned in Study was performed on CD4+ T cell clones generated by plating PBMC 5 pKLO vector, as reported previously (32). Briefly, infected T cell line from two HTLV-1–infected patients (0.1 cell/well, 1 mg/ml PHA A, 10 HUT102 was treated as follows: cell medium was changed 1 h before allogenic PBMC/well, and 5 U/ml IL-2) and continuously maintained treatment, and cells were plated in a U-bottom 96-well plate. Cells were thereafter, as described previously (27). T cell clones were screened for then treated with Polybrene (5 mg/ml; Santa Cruz Biotechnology) and 30 HTLV-I proviral DNA by PCR amplification with LTR region-specific min later with control lentivirus (empty pKLO) and lentivirus expressing primers, whereas Tax expression was measured through quantitative RT- shRNA-anti Tax (10 multiplicities of infect). Medium was changed at 24 h PCR as described previously (25). The monoclonality of cultured cells was postinfection, and cells were cultured for another 2 d. assessed on DNA by TCR-g-chain gene rearrangement analysis as de- scribed previously (28). Abs recognizing CD4 T cells were identified using Western blotting fluorescence-activated cell scanner (FACScan) analysis and Abs from DakoCytomation. Cloned T cells were incubated with 5% filtered human Cells were lysed (20 mM Tris-HCl [pH 7.4], 10% sucrose, 1 mM EDTA, and serum and stained with Abs. Staining and scanning were performed in PBS 5 mM EGTA and Complete), subjected to ultrasound fragmentation, and with 2% FCS. Isotype-matched controls were used. Data were acquired protein concentration was determined using Lowry assay (Bio-Rad). Pro- on a FACScan and analyzed by means of CellQuest Software (BD Bio- teins (20 mg) were separated under reducing conditions by SDS-PAGE and sciences). An alternative study was performed on the CD4+ T cell lines transferred to nitrocellulose membrane (Schleider and Schuell). Proteins CEM, Jurkat/JK, and HTLV-1–infected T cells lines C8166, C91PL, HUT- were immunodetected with specific Ab and then revealed by the ECL 102, MT2, and MT4. T cell lines were cultured without IL-2. Cell selec- method (Pierce). GAPDH detection served as control of protein deposit. + tion: early-activated CD69 cells were selected from healthy blood donor Transmigration assay (HD) and HAM/TSP patients using MACS beads (Miltenyi Biotec). A total of 10 3 106 PBMCs were labeled with anti-CD69 mAb (20 min, 4˚C), T cell transmigration was performed in triplicate in micro-Transwell washed in medium, and incubated (20 min, 4˚C) with anti-CD69 magnetic (Boyden chamber) as described previously (24). T cell preparations (4 3 1224 RETROVIRUS MANIPULATES LYMPHOCYTE MOTILITY MACHINERY

105 cells/well) were added in the upper chamber and incubated at 37˚C for CRMP2+ cells compared with 34 6 13% in noninfected clones (p , 1 h and 30 min. Chemokines CXCL12 (20 ng/ml), CCL5 (100 ng/ml), and 0.001) (Fig. 1A), indicating that the CRMP2 expression level cor- CCL2 and CXCL10 (20 ng/ml) were added in the lower compartment. The related with the infection status of T cell clones. Because HTLV-1 is migratory T cells in the lower chambers were counted under microscopy (at least 10 fields). Data are expressed as the mean number of migratory known to chronically activate infected T lymphocytes, CRMP2 was T cells. analyzed in cells expressing CD69, a marker of T cell activation. Infected T cell clones displayed an elevated frequency of CD69+ Real-time PCR cells (72 6 4 versus 22 6 3% in uninfected clones). Colabeling RNA was isolated with the use of RNA PLUS (Qbiogen), residual genomic indicated that almost all activated T cells expressed CRMP2 (91 6 DNA was removed using Dnase I (DNase-free, Ambion), and reverse 6%) (Fig. 1B). Thus, CRMP2 expression level correlated with the transcription was performed [500 ng total treated RNA and 100 ng oligo activation status of T cell clones. This corroborated our previous [dT]12–18 primers (33)]. cDNA was amplified by real-time PCR (Light- Cycler; Roche) using FastStart DNA Master SYBR Green I (Roche). PCR analysis on PBMC of HTLV-1–infected patients, linking the high were separately performed with Tax, CRMP2, and b-actin primers as fol- CD69+ cells frequency with elevated CRMP2 expression (24). To go lows: 15 min at 95˚C (20 s, 95˚C; 20 s, 60˚C; and 10 s, 72˚C), 48 cycles. deeper into the cell locomotion machinery of activated T lympho- 9 9 9 Tax primers (J02029) were 5 -ATCCCGTGGAGACTCCTCAA-3 and 5 - cytes during HTLV-1 infection, we analyzed the association be- AACACGTAGACTGGGTATCC-39. Human b-actin primers (NM_001101) were 59-ACTGGAACGGTGAAGGTGAC-39 and 59-GTGGACTTGGGA- tween CRMP2 expression (cytometry), migratory rate (Transwell GAGGACTG-39. CRMP2 primers (NM_001386) were 59-TCACATCAG- system), and profile (transcriptome) in activated AACTCCTGTGG-39 and 59-GAAAGTTCCTCCACTCATGG-39. Speci- cells selected from HAM/TSP patients and HD (selection on CD69 ficity of amplification was achieved by a final melting step giving a single Ag). CRMP2 was higher in selected cells of patients (mean fluo- melting peak at 83.4, 77.2, and 62˚C for Tax, b-actin, and CRMP2 PCR 6 6 Downloaded from product, respectively. The level of Tax and CRMP2 mRNA was expressed rescence intensity = 72.2 2 versus 33.3 10 in HD) and mi- as relative units normalized to b-actin. gratory rate specifically elevated (22 6 2 versus 14 6 2 migrating cells per hour for HD; p = 0.02). Gene expression analysis (Sup- Gene expression profile plemental Table I) revealed the main presence of T lymphocytes in A transcriptomic approach was performed by the genomic platform facility selected cells (CD2, CD3dgz, CD7, and CD8ab mRNA) and their + ProfileXpert (Lyon, France; http://www.profilexpert.fr) on CD69 selected activation status (CD69, CD96, and CD97). Dendritic cell markers cells using oligonucleotide microarrays (CodeLink Uniset human 20k were also detected (CD1c,d, CD58, CD86, and CD207). These two http://www.jimmunol.org/ bioarrays; General Electric Health Care) and Assistant Pathway software, as reported previously (34). Of the 19,881 known and EST surveyed cell types are susceptible to HTLV-1 infection and have been on each array, 5,534 were expressed on the two chips for each human detected in HAM/TSP patients (Ref. 35 and reviewed in Ref. 36). group (two HD; Etablissement Franc¸ais du Sang; two HAM/TSP patients; Several genes were associated with immune cell homing (XCL2, A. Gessain, Pasteur Institute, Paris, France). These latter represented the CXCL8, CCL4L, CCL5, CCL22, CXCL10, CXCL16, CCR6, and genes commonly expressed in CD69+ selected cells, whatever the clinical status. Only those coding for cluster of differentiation (CD), cytokine, CCR7) and transendothelial migration (CXCR4, CD18, CD44, chemokine, and their receptors were shown (Supplemental Table I). In CD49a/VLA4, CD58, and CD164). HAM/TSP patients exhibited a contrast, some genes were exclusively expressed or displayed augmenta- distinctive profile, with 22 genes uniquely expressed and 10 others tion in infected patients (Table I). Validation of data was performed as with a $3-fold increased expression (Table I). Elevated expression follows. Data from microarray analysis were validated on the same RNA of CD69 and HLAG confirmed that lymphocytes were highly by guest on September 28, 2021 preparation using quantification by real-time qPCR of 11 genes selected for their specific behavior (present, absent, or upregulated, in each group). activated in HAM/TSP. In addition, several induced genes were In each case, results corroborated the microarray data. The specific ex- involved in molecular pathways acting downstream chemokine pression of gene products in CD69+ selected cells was confirmed using receptors, notably G protein and PI3K signaling (ARHGAP8, qPCR analysis of eight present genes and performed on mRNA isolated GNA13, GRASP, GRK5, GPR68, PIK3C2A, PI3KR1, RGS1, and from CD69+ selected cells of new human group (two HAM/TSP and two HD). Again, results corroborated observations in the microarray approach. RGS16). Other genes have been involved in intracellular trafficking Microarray data are in the public database Gene Expression Omnibus and secretion (HAMP, NUP37, NUP98, RAB28, and SENP1). This (Series record GSE33859; http://www.ncbi.nlm.nih.gov/geo/). approach showed that, in an HAM/TSP patient, a global activation Statistical analysis of pathways involved in lymphocyte response to chemokine signal was associated with high CRMP2 expression. Statistical significance in comparing two means was tested with the un- We further focused on cultured T cell clones and evaluated the , paired Student t test, and p 0.05 were considered significant (in figures relation between cell polarization, migratory rate, and CRMP2 ***p # 0.001 and **p # 0.01). expression. Polarization, the initial step of lymphocyte migration, was analyzed in 12 infected and 3 uninfected T cell clones by Results counting the cells exhibiting a bipolar shape (uropod formation). High migratory rate of blood T lymphocytes isolated from Lymphocytes of infected T cell clones displayed uropod at higher HTLV-1–infected patients correlates with virus infection, frequency (86 6 6%) than uninfected clones (41 6 7%) (p , activation status, and elevated CRMP2 expression 0.01). The migratory rate of T lymphocytes was specifically en- We previously reported that CRMP2 expression was elevated in hanced for infected T cell clones (Fig. 1C, left graph), and non- total PBMC of HTLV-1–infected patients, notably in CD4+,CD69+, linear regression analysis of data using a logarithmic representa- CD45RO+,HLADR+, and VLA4+ lymphocytes (24). To investigate tion indicated a degree of correlation between the migratory rate the role of a virus in CRMP2 modulation and its consequences on and the frequency of CRMP2+ cells in each T cell clone (R2 = T cell migration, we now studied T lymphocytes isolated from 0.6647) (Fig. 1C, right graph). Blockade experiment using anti- infected patients and clonally established T lymphocites in culture CRMP2 Ab (Fig. 1D) showed the involvement of CRMP2 in the (termed T cell clones). Among the 18 T cell clones studied, 12 migration and polarization of infected T cell clones (Fig. 1D1, harbored the virus and expressed mRNA coding for the viral pro- 1D2). In addition, treatment with CRMP2 siRNA reduced the tein Tax. First, we analyzed by flow cytometry the frequency of number of migrating cells (Fig. 1D3). These data confirmed the CRMP2+ cell in T cell clones. It must be noted that CRMP2 is role of CRMP2 in cell polarization status and migratory capacity generally expressed in almost all T lymphocytes analyzed ex vivo, of infected T cell clones. To conclude, HTLV-1 infection induced but the frequency of CRMP2+ cells is downregulated in vitro fol- a profound alteration of the T cell motility machinery in parallel lowing culture. Virus-infected T cell clones displayed 75 6 12% of modulation of CRMP2 expression. The Journal of Immunology 1225 Downloaded from http://www.jimmunol.org/

FIGURE 1. HTLV-1 infection conferred higher CRMP2 expression and migratory rate in virus-infected T lymphocytes. Analysis by flow cytometry of T cells cloned from HTLV-1–infected patients: A, elevated frequency of CRMP2+ cells in virus-infected T cell clones; B, elevated frequency of CRMP2+ and CD69+ activated cells in total lymphocytes in infected T cell clone compared with uninfected clone (left panel); elevated frequency of CRMP2+ in CD69+ cells in infected T cell clone (right panel). C, High migratory rate for infected versus uninfected T cell clones, evaluated in transmigration assay (Transwell chambers, triplicate, 1.5-h migration) toward chemokine mixture (C, left panel). Nonlinear regression analysis (C, right panel, logarithmic by guest on September 28, 2021 representation) showed a correlation between migratory rate and CRMP2+ cells frequency in each clone. D, Dose-dependent reduction of migration (D1) and polarization (D2) (infected T cell clone tested) by treatment with anti-CRMP2 blocking Ab compared with cells without treatment (control). Reduction of cell migration for two infected T cell clones treated with siRNA anti-CRMP2 or control.

The viral protein Tax is involved in the modulation of CRMP2 on day 2 (95% Tax+ cells) and elevated CRMP2+ cell frequency in level and lymphocyte migratory rate parallel (77–90%). Likewise, Tax transfection in the T cell line We suspected the virus protein Tax to regulate CRMP2 expression Jurkat enhanced CRMP2 expression level, as shown by Western in infected T cell clones, because this protein is known to maintain blotting performed at 24 and 48 h posttransfection (Fig. 2D). Thus, chronic activation of T lymphocytes and also modulates expression Tax protein per se was able to augment CRMP2 expression in of viral and cellular genes (11, 37). A set of experiments solidly T lymphocyte. Finally, to evidence a direct effect of Tax protein on T cell motility, TRIP-Tax–transduced T lymphocytes were exam- supported this hypothesis. Analysis of Tax at mRNA level was ined for migratory property (Fig. 2E). Tax+ lymphocytes exhibited a performed by qPCR in virus-infected T cell clones (11 analyzed) higher migratory rate compared with lymphocyte control (p =0.01). (Fig. 2A). Linear regression analysis, which combined tax ex- Altogether, these data implicated Tax in the motile phenotype dis- pression level and CRMP2+ cells frequency in each clone, re- played by HTLV-1–infected T lymphocytes via a modulation of vealed a degree of correlation between CRMP2 (positive cells in CRMP2 expression level. percentages) and tax expression level (R2 = 0.626). Using RT-PCR and Western blotting, we also compared CRMP2 expression in Virus infection affects CRMP2 phosphorylation and adhesion chronically infected Tax-expressing T cell lines (C91PL, C8166/ to cytoskeleton elements in T lymphocytes 45, and HUT102 cell lines) and uninfected T cell lines (Jurkat and In neural and immune cells, CRMP2 is known to mediate intra- CEM) (Fig. 2B). Increased CRMP2 mRNA and protein levels in cellular signaling of several membrane receptors, an activity reg- chronically infected T cells substantiated the association of Tax ulated by the concerted action of diverse , acting notably with elevated CRMP2 expression. However, Tax in T cell clones at the C terminus of the protein (reviewed in Refs. 25 and 38). In also reflected HTLV-1 proviral expression, suggesting that virus addition, CRMP2 activity also results from opposing activity of its proteins other than Tax could modulate CRMP2 in infected lym- two molecular subtypes, CRMP2-A and -2A/B, which occur from phocytes. The role of Tax in modulation of CRMP2 expression the alternative usage of coding exons. They indeed exert a con- level was confirmed in primary T lymphocytes transduced with trastive effect on microtubule pattern and neuron morphology HIV-based lentiviral vector (TRIP-Tax) to express Tax (Fig. 2C). because CRMP2A supports axonal elongation by antagonizing to TRIP-gfp was used as control (data not shown). Flow cytometry CRMP2A/B (39). We therefore investigated the consequence of analysis detected large amount of Tax in transduced lymphocytes virus infection on CRMP2 activity by analyzing CRMP2 subtypes 1226 RETROVIRUS MANIPULATES LYMPHOCYTE MOTILITY MACHINERY

Table I. Gene specifically expressed or with elevated expression in blood CD69+ cells isolated from HAM/TSP patients

Induced Expression

Gene Nomenclature Protein Function ARHGAP8 NM_181333 GTPase Motility BRDG1 NM_012108 BCR Proliferation CCNG1 NM_199246 Cyclin G1 Proliferation GNA13 NM_006572 G protein signaling Motility GRASP NM_181711 G protein signaling Motility GRK5 NM_005308 GPCR kinase Motility GPR68 NM_003485 G protein signaling Motility HAMP NM_021175 Hepcidin Antimicrobial peptide KLHL 5 AL049385 Kelch-like 5 Motility KLHL11 NM_018143 Kelch-like 11 Motility HLAG NM_002127 MHC class 1 Cell activation IL6 NM_000600 Cytokine Hematopoiesis NUP37 NM_024057 Nucleoporin Intracell trafficking NUP98 NM_139132 Nucleoporin Intracell trafficking RAB28 NM_004249 Ras family Intracell trafficking

PIK3C2A NM_002645 G protein signaling Motility Downloaded from SNX4 NM_003794 Sorting nexin 4 Endocytosis SENP1 NM_014554 Sumo1/sentrin Intracell trafficking TIAM2 AF120323 G protein signaling Motility TCTE3 BC063455 Dynein L chain Motility TNFRSF6 NM_152877 Fas cell Death TNFRSF8 NM_152942 Cytokine R/CD30 Activation http://www.jimmunol.org/ Enhanced Expression (Top 10 Genes)

Gene Nomenclature Protein Function Fold Increase RGS16 NM_002928 G protein signaling Motility 14 RGS1 NM_002922 G protein signaling Motility 6 PRC1 NM_199414 Regulator cytokinesis Proliferation 6 VIL2 NC_003071 Villine 2 adhesion, Migration 5 PI3KR1 NM_181524 PI3K signaling Motility 5 LTB4R NM_019839 G protein signaling Motility 4 CD69 NM_001781 Lectin Activation 3 by guest on September 28, 2021 RAB18 BO 015014 Ras associated protein Migration 3 RORA NM_002943 Retinoic-like Tfactor TH17 3 IL21R NM_021798 Cytokine R Proliferation 3 and phosphorylated forms in chronically infected T cell lines. lead to significant modification of phosphorylation of CRMP2 Western blotting was performed using polyclonal Abs directed on Ser522 and Tyr479, as shown by the ratio pCRMP2/CRMP2 against CRMP2-A, CRMP2-A/B (generally termed CRMP2), (0.37 and 0.35, respectively). Overall, these data showed the abil- CRMP2-pSer522, CRMP2-pThr509, and CRMP2-pTyr479 (Fig. ity of HTLV-1 infection to change the phosphorylation status of 3A). As expected, CRMP2-2(A/B) expression was elevated in CRMP2, thus to modify its activity. However, Tax was not in- virus-infected T cell lines. The most striking difference appeared volved in this posttranslational modification. in CRMP2-A expression, only detected in the uninfected cell line In neural cells, phosphorylation modulates CRMP2 interaction CEM. Interestingly, CRMP2-A is known for its blocking effect with cytoskeleton elements in response to external signals, re- toward CRMP2(-A/B) in the control of neural cell polarity (39). In sulting in cytoskeleton reorganization, axonal elongation, and addition, CRMP2-pSer522, CRMP2-pThr509, and at a lesser ex- cell migration (41, 42). We then compared CRMP2 subcellular tent, CRMP2-pTyr479, forms were enhanced in virus-infected localizationinuninfectedJurkat and HTLV-1–infected C8166 cells, notably in productive infected cell lines, as shown by the T cell lines. Western blotting performed on membrane/endosome, ratio of phosphorylated forms to total CRMP2 (CRMP2-A/B) nucleus, and cytoskeleton cell fractions showed, in C8166 cells, signal. Because Ser522, Thr509, and Tyr479 are target for Cdk5, a higher recruitment of CRMP2 to the cytoskeleton compartment GSK-3b, and the Src kinase Yes, respectively (25, 40), we ana- that paralleled its decrease in the membrane/endosomal fraction lyzed the active form of these kinases using Western blotting (Fig. 3C). Preferential CRMP2 adhesion to cytoskeleton in C8166 (data not shown). None or few modifications were observed in versus Jurkat T cells was also suggested by immunofluorescence infected versus uninfected T cell lines. The effect of Tax pro- analysis (Fig. 3D). Because CRMP2 is able to bind actin, tubulin, tein on CRMP2 phosphorylation was also investigated using tax and the intermediate filament vimentin (23), we investigated transfection of Jurkat cells. Tax affected the level of CRMP2 as whether virus infection modified CRMP2 binding to these pro- expected (1.6-fold increase), but only weakly changed CRMP2 teins. Immunoprecipitation of tubulin, actin, and vimentin was phosphorylation (weak increase in pSer522 form) (Fig. 3B). To performed on Jurkat and C8166 cells, and then, CRMP2 was confirm this aspect, the infected T cell line HUT102 was trans- evaluated in immunoprecipitates. It must be noted that CRMP2 duced with lentivirus-expressing shRNA anti-Tax and vector is often expressed as a 62-kDa full-length and a 58-kDa short- control, as described previously (32), and cells were examined in processed form resulting from calpain-mediated cleavage (a 50- Western blotting with anti-pCRMP2 Abs. Tax knockdown did not kDa product may appear). With Ab that recognizes the long and The Journal of Immunology 1227 Downloaded from http://www.jimmunol.org/

FIGURE 2. Tax expression in T lymphocyte enhanced CRMP2 level expression and migratory rate. A, CRMP2 and Tax expression in infected T cell clones: correlation between the frequency of CRMP2+ (flow cytometry) and Tax expression levels (qPCR) detected by linear regression analysis. Microcopy photo: example of Tax protein detection by immunofluorescence. Original magnification 3800. B, CRMP2 and Tax expression in HTLV-1 chronically infected and Tax-expressing T cell lines (C8166, C91PL, MT4, and HUT102) compared with uninfected cell lines (JK and CEM): elevated level of CRMP2 mRNA (qPCR, left panel) and protein (Western blotting, right panel) in infected cell lines, as shown by GAPDH/CRMP2 ratio. C, CRMP2 and Tax + expression in primary T lymphocytes transduced with TRIP-Tax vector (flow cytometry): increase in CRMP2 cell frequency (77–90%) and level (mean by guest on September 28, 2021 fluorescence intensity: 1.5-fold increased). D, CRMP2 and Tax expression (Western blot) in tax-transfected Jurkat cells: increased CRMP2 expression level (ratio to GAPDH). E, Migratory rate of TRIP-vector–transduced T lymphocytes: elevated migratory rate for Tax+ versus gfp+ lymphocytes (Transwell chamber system). short forms (CRMP2-pep4 Ab), Western blotting detected CRMP2 transfection with CRMP2wt and CRMP2D503 had opposite effect coimmunoprecipitation with tubulin, actin, and vimentin, as de- on T cell migratory rate (enhancement versus decrease, respec- scribed previously. However CRMP2 pattern differed in C8166 tively), validating in immune cells the dominant-negative effect of versus Jurkat cells (Fig. 3E). Notably, the full-length CRMP2 short CRMP2. We further examined the presence of full-length bound the three cytoskeleton components in C8166 cells but only and short CRMP2 in uninfected and HTLV-1–infected T cell lines vimentin in JK cells. To conclude, HTLV-1 modulated the phos- using Western blotting and CRMP2-pep4 Ab (Fig. 4B). The short phorylation status of CRMP2 and its recruitment to cytoskeleton (58 kDa) and full-length (62 kDa) products were detected in all elements in T lymphocyte. T cell lines, but the full-length CRMP2 was overexpressed in infected T cells, as shown by the 58:62-kDa ratio. The role of HTLV-1 acts on CRMP2 proteolytic processing, favoring the HTLV-1 in the modification of CRMP2 processing was then presence of active CRMP2 form in infected T lymphocytes evaluated in tax-transfected Jurkat cells (Fig. 4C). Western blot- We have recently reported the importance of proteolytic pro- ting showed that CRMP2 level expression was enhanced in Tax cessing of CRMP2 in controlling the protein activity in neural cells expressing cells, as expected. In addition, the 58:62-kDa ratio was (27). The short-processed CRMP2 form (58 kDa) antagonized the strongly reduced (51 to 9, in gfp- versus tax-transfected cells), full-length (62 kDa), acting as a physiological dominant-negative indicating that Tax promoted the full-length active form in trans- signal to reduce neural cell process extension. This observation fected lymphocytes. To confirm this aspect, infected T cell line and the preferential adhesion of the full-length/active CRMP2 HUT102 was transduced with lentivirus-expressing shRNA anti- form to actin and tubulin in infected T cells prompted us to test the Tax, and cells were examined for CRMP2 cleavage using Western hypothesis that HTLV-1 enhanced CRMP2 activity in T lympho- blotting. Knockdown expression of Tax protein was associated cyte by increasing the level of the CRMP2 active form. We first with the reduction of full-length CRMP2, as shown by the aug- tested whether short CRMP2 acted as a dominant-negative signal mentation of the 58:62-kDa ratio (3:21) in control lentivirus ver- in immune cells by testing the migratory rate of Jurkat cells sus shRNA anti–Tax-treated cells). transfected with plasmid coding either for the full-length CRMP2 As Tax is known to regulate the function of the proteasome, (CRMP2wt) or the short dominant-negative form, CRMP2D503, notably by binding specifically to two subunits of the 20S pro- as previously reported for neural cells study (29). Cells transfected teasome, HsN3 and HC9 (43), the evolving 58:62-kDa ratio of with gfp plasmid were used as control. As shown in Fig. 4A, CRMP2 that we observed led us to analyze CRPM2 proteasomal 1228 RETROVIRUS MANIPULATES LYMPHOCYTE MOTILITY MACHINERY Downloaded from http://www.jimmunol.org/

FIGURE 3. Modification of CRMP2 phosphorylation and subcellular localization of CRMP2 in HTLV-1–infected T cell lines. A, Analyze of CRMP2 subtypes, CRMP2-A and CRMP2-(A/B), and the phosphorylated CRMP2 forms: CRMP2-pSer522, -pThr509/514, and -pTyr479, using Western blotting: by guest on September 28, 2021 loss of CRMP2-A and increase in phosphorylated forms in HTLV-1–infected cell lines compared with uninfected CEM cells (ratio to CRMP2-[A/B] signal). B, CRMP2 phosphorylated forms in tax-transfected Jurkat cells: increased CRMP2 level in transfected cells (ratio to signal in gfp cells) and CRMP2 phosphorylated on pSer522 (ratio to CRMP2 signal in gfp cells). C, CRMP2 subcellular localization in Jurkat and C8166 T cell lines: enhanced recruitment to cytoskeleton compartment in HTLV-1–infected cells C8166 cells (vimentin as control of subcellular fraction). D, CRMP2 (red color) colocalizes (orange color, arrows) with cytoskleton elements vimentin and actin (green color) in HTLV-1–infected cells C8166. Original magnification 3500. E, Coimmunoprecipitation of CRMP2 with cytoskeleton elements tubulin, actin, and vimentin in Jurkat and C8166 T cells: modification of adhesion pattern to actin and tubulin in HTLV-1–infected cells (full-length, 62-kDa CRMP2-bound tubulin and actin only in infected cells; black arrow). degradation under Tax influence. Jurkat and C8166 cells were analysis of CRMP2 showed that transfection with CRMP2D503 treated with the proteasome inhibitor PS-341 (10 nM for 24 h), was associated with a decreased detection of the endogenous full- allowing us to evaluate the fraction of CRMP2 protein subjects length CRMP2 in Jurkat cells, whereas cotransfection with Tax to degradation. The expression of the short and full-length forms restored this expression (Fig. 5B). All these data strongly sug- of CRMP2 was analyzed by Western blot. We observed that the gested that the elevated migratory rate displayed by HTLV-1– fold increase of short CRMP2 expression after PS-341 treatment infected T cell clones reported above was associated with high was dramatically enhanced (108-fold more) in the infected cell level of full-length/active CRMP2. This hypothesis was examined line C8166 compared with the uninfected Jurkat one (Fig. 4D). in eight infected T cell clones established from patients using By contrast, the expression of full-length CRMP2 was simi- CRMP-2-Cter and -pep4 Abs in Western blot analysis. Interest- larly affected in the two cell lines. These results suggest that ingly, the full-length CRMP2 was detected only in infected T cell HTLV-1 infection may contribute to the modification of the 58: clones (Fig. 5C). We then examined the possible association be- 62-kDa ratio by enhancing the degradation of the short form of tween the full-length form in T lymphocytes and their migratory CRMP2. rate (three infected and three uninfected T cell clones studied) To substantiate a possible link between Tax, CRMP2 proteolytic following evaluation of the full-length/short (62:58 kDa) CRMP2 processing and T lymphocyte migration, we examined the mi- ratio (Fig. 5D, migration in lower panel, CRMP2 ratio in upper gratory rate of tax-transfected Jurkat cells following cotransfection panel). We observed that the more favorable to the full-length with full-length or short CRMP2 (CRMP2wt and CRMP2D503). form the ratio, the more elevated the migratory rate of infected Tax enhanced the migratory rate of cotransfected Jurkat cells, as T-cell clones was. Collectively, these data demonstrated that expected (Fig. 5A). In addition, in the presence of Tax, the cells HTLV-1 infection, possibly through Tax activity, profoundly al- transfected with CRMP2D503 showed no reduction of their mi- tered posttranslational processing of CRMP2, resulting in more gration ability compared with the control, contrary to what was active molecular form in T lymphocyte and elevated migration as observed with CRMP2D503 alone. Interestingly, Western blotting a consequence. The Journal of Immunology 1229

FIGURE 4. CRMP2 processing in HTLV-1– infected T cell lines and role of short CRMP2 in T lymphocyte migration. A, Migratory rate of JK cells transfected with full-length (CRMP2wt) and short (CRMP2-D503) CRMP2: elevation with CRMP2wt and reduction with the short/domi- nant-negative CRMP2-D503. B, Detection of the full-length/62kDa CRMP2 form (empty arrow) and cleaved/short product (58 kDa, black arrow) in T cell lines (western blot; anti-CRMP-2 pep4 Ab): full-length CRMP2 is elevated in infected cell lines, as shown by the decreased 58:62-kDa ratio. C, De novo Expression of Tax in JK cells (Tax plasmid versus empty vector) and extinc- tion of Tax in infected HUT102 cells (lentiviral shRNA anti-Tax versus empty lentivirus): the 62- kDa form is elevated in tax transfected Jurkat cells and reduced following Tax knocking-down, as shown by the short/full-length CRMP2 ratio.

D, Expression of CRMP2 full-length and short Downloaded from forms in Jurkat cells following treatment with proteasome inhibitor PS341 (fold increase): high increase in short CRMP2 expression (108-fold more) in C8166 cells compared with uninfected cells Jurkat. By contrast, full-length CRMP2 is similarly affected in the two cell lines. http://www.jimmunol.org/

Tax and CRMP2 colocalized at uropod and cell–cell contact in cooperate. Preliminary analyses using immunoprecipitation and HTLV-1–infected T cells CRMP2-GST pull-down strategy were unable to detect a direct Given 1) the crucial role of CRMP2 in remodeling neural cell and association between these proteins. Nevertheless, we further T lymphocyte cytoskeleton by direct binding to microtubules analyzed the localization of these two proteins in infected lym- (25, 44) and also in participation, as a cargo receptor, in the phocyte. Immunofluorescence study showed the presence of transport of specific vesicles (45, 46), 2) the ability of Tax protein Tax in lymphocyte cytoplasm, as small dots and its colocaliza- tion with CRMP2 at uropod in infected T cell clone and at the to modulate CRMP2 activity in infected lymphocyte, in partic- by guest on September 28, 2021 ular cytoskeleton binding (shown in this paper), 3) the capacity site of cell–cell contact in infected/productive HUT102 T cells of infected cell to transfer Tax though a well organized cell–cell (Fig. 6). This suggested a possible cooperation between these contact (the so-called virological synapse (47), and 4) the role two proteins for uropod formation and subsequent lymphocyte of microtubule in T cell signaling (48), we suspected CRMP2 migration and for molecule/signal delivery through membrane to bind Tax protein at specific sites in lymphocyte cytoplasm to connection.

FIGURE 5. Elevated expression of full-length CRMP2 in virus-in- fected T cell clones correlates their high migration competency. A,Mi- gratory rate of Jurkat cells cotrans- fected with tax and CRMP2wt or CRMP2-D503: increased migratory rate in tax-transfected cells in all conditions. Tax abolished the nega- tive effect of CRMP2-D503. B, Ex- pression of endogenous CRMP2 in transfected Jurkat cells: Tax restores the full-length CRMP2 (62 kDa) ex- pression decreased by CRMP2-D503. C, Patients’ T cell clone analysis using Western blotting: detection of full- length CRMP2 only in virus-infected T cell clones with anti–CRMP2-Cter Ab. D, Migration of uninfected and infected T cell clones (Transwell sys- tem): elevated migratory rate in virus- infected T cell clones is associated with higher level of full-length CRMP2 (62:58-kDa ratio). 1230 RETROVIRUS MANIPULATES LYMPHOCYTE MOTILITY MACHINERY

direct interaction. In addition, it is known that Sp1 activity depends, in part, on repression of PP2A phosphatase, a negative regulation that may be achieved by HTLV-1 (64–67). In contrast, Tax could modulate CRMP2 expression by interfering with TGF-b super- family members, including bone morphogenetic protein (BMP). SMAD elements are present in crmp2 promoter and BMP2 and BMP4 can suppress the expression of CRMP2 in neural cells through direct interaction of SMAD1 and SMAD4 (68). The abil- ity of Tax protein to block TGF-b signaling recently detected in HTLV-1–infected T cells (69) suggests a possible enhancement of CRMP2 expression by blockade of this negative regulation. It is now known that CRMP2 expression level is not sufficient FIGURE 6. CRMP2 and Tax colocalized at uropod and cell-cell contact to trigger cell motility. Posttranslational modifications, including in HTLV-1–infected lymphocyte. Immunodetection performed on infected phosphorylation and proteolytic processingm are crucial for CRMP2 T cell clone and HUT102 T cell line detected Tax (red) as cytoplasmic activity. We and others (40, 42, 70, 71) have previously shown that small dots, which colocalized (orange) with CRMP2 (green). Original sequential phosphorylation of CRMP2 is required for semaphorin magnification 3500. and chemokine signaling in neural and immune cells, respectively. In T lymphocyte, CXCL12/SDF1 signal decreases phosphorylation

Discussion on Thr509, allowing phosphorylation on Tyr479 presumably by the Downloaded from In the current study, we characterized a functional interaction be- kinase Yes, and induces T cell migration (25). In the current study, tween the human retrovirus HTLV-1 and CRMP2, a cytoskeleton we showed that HTLV-1 enhanced the phosphorylation of CRMP2 organizer that plays an important role in lymphocyte polariza- on several residues, a feature associated with CRMP2 recruitment tion and migration. Virus variously affected CRMP2, modifying on the cytoskeleton of infected lymphocyte. In addition, analysis the expression level, proteolytic processing, and phosphorylation of gene expression profile in activated lymphocytes of HAM/TSP

status of the protein. Such changes are strongly suspected to pro- patients revealed the activation of T cell locomotion machinery. http://www.jimmunol.org/ vide infected T lymphocytes with a highly motile phenotype. This Altogether, this suggests that HTLV-1 infection chronically acti- observation could give an explanation for the broad distribution vates the molecular pathways involved in lymphocyte motility, of infiltrated T lymphocytes and HTLV-1 proviral DNA in ATL notably those linked to CRMP2. Several observations support this and HAM/TSP patients who develop extraneural inflammatory hypothesis. First, CRMP2 belongs to the PI3K/Akt/GSK-3 path- disorders such as pulmonary alveolitis, dermatitis, Sjogren’s syn- way, shown essential for axon elongation in neurons (41) and for drome, thyroid disease, uveitis, arthritis, and polymyositis (49–52). cell migration in general (72). The ability of HTLV-1 to activate In addition, as a microtubule organizer and cargo receptor, CRMP2 PI3K/Akt signaling pathway in T lymphocyte has been demon- could be suspected to participate in HTLV-1 transmission via or- strated (73) and could explain the specific gene profile found in ganization of the microtubule organizer center. cells selected from HAM/TSP patients that involves the molecular by guest on September 28, 2021 The demyelinating and degenerative disease HAM/TSP induced G protein/PI3K pathway. Interestingly, CRMP2 is a binding partner by HTLV-1 often occurs in two phases, initiated first as an in- of several proteins of this pathway, including Fyn, PI3KR1/p85, flammatory disorder, followed by a chronic long-term degenerative and Vav-1 (25, 42). Finally, Tax may bind PI3KR1/p85 (37) and stage (reviewed in Ref. 53). Infiltrated lymphocytes have a crucial influence Vav phosphorylation (74). To conclude, HTLV-1 infec- role in pathogenesis as they constitute the major reservoir of tion has the ability to influence the CRMP2/PI3K/Akt/GSK-3 axis HTLV-1 in the CNS of patients, and they can produce inflam- in T lymphocyte thus to positively control the migratory activity matory cytokines and metalloproteases described for their del- of infected cells. The present observation on ex vivo lymphocytes eterious effect on neural cells (54–56). In addition, infiltrated complements two previous reports showing that HTLV-1 regulates T lymphocytes express and can release the viral protein Tax, a G protein signaling, in particular, activates the SDF1/CXCR4 axis strong inducer of proinflammatory cytokine secretion from the in T lymphocyte of ATL patients, lymphoblastoid cells from Tax resident CNS cells (57–59). transgenic mice, and HTLV-1–infected T cell lines (75, 76). The strategy developed by HTLV-1 to enhance T lymphocyte CRMP2 proteolytic cleavage is an additional posttranslational motility involves several aspects. HTLV-1 infection has been re- modification essential for cell motility/outgrowth. In fact, CRMP2 ported to augment expression of metalloproteases and membrane is expressed in neural cells as a full-length (62 kDa) and a short- molecule adhesion, as such was suspected to facilitate T lym- processed form (58 kDa), resulting from cleavage by the calcium- phocyte entry into the CNS through the blood–brain interfaces (56, dependent protease calpain (77). CRMP2 cleavage is enhanced 60–62). In this study, we describe a new strategy for HTLV-1 and in injured neural tissues displaying limited postinjury remodeling report a functional interplay between the virus and CRMP2, which and neurite regeneration (78). Our demonstration that the short activates the lymphocyte locomotion machinery. Using virus- CRMP2 molecular form negatively regulated axon elongation infected T cell clones established from infected individuals, could explain in part this feature (29). In the context of immune chronically infected T cell lines, and tax-transfected lymphocytes, cells, we have recently identified the full-length and short CRMP2 we showed a direct role of HTLV-1 in modifying CRMP2 activity forms in T lymphocytes (25) and showed, in the present work, that by altering its protein cellular level and posttranslational mod- the negative regulation of short CRMP2 (CRMP2D503) occurs ifications. Transfection experiments directly assign a role for Tax in lymphocyte migration. Interestingly, cotransfection experiments protein in the elevated expression of CRMP2 in T lymphocytes. pointed out the ability of Tax protein to overcome the negative effect As a potent regulator of cellular gene transcription (reviewed in of short CRMP2 on T cell migration. Such a feature could have an Ref. 37), Tax has the ability to control crmp2 expression. In fact, explanation at protein level as transfection with CRMP2D503 dra- several transcriptional sites located on the crmp2 gene promoter matically reduced the full-length CRMP2 in Jurkat cells, whereas were recently described and include a for the tran- cotransfection with Tax restored this expression. Given these data, scription factor Sp1 (63). Tax by itself can activate Sp1 through privileging the full-length/active CRMP2 form in T lymphocyte The Journal of Immunology 1231 could be a means for HTLV-1 to promote cell migration. The fol- 3. Gessain, A., F. Barin, J. C. Vernant, O. Gout, L. Maurs, A. Calender, and G. de The´. lowing observations support this hypothesis: 1) the ratio of CRMP2 1985. Antibodies to human T-lymphotropic virus type-I in patients with tropical spastic paraparesis. Lancet 2: 407–410. forms was always more in favor of the full-length protein in 4. Yamano, Y., C. J. Cohen, N. Takenouchi, K. Yao, U. Tomaru, H. C. Li, Y. Reiter, infected T cells clones, tax-transfected T cells, and infected T cell and S. Jacobson. 2004. Increased expression of human T lymphocyte virus type I (HTLV-I) Tax11-19 peptide-human histocompatibility leukocyte antigen A*201 lines; 2) knocking down Tax expression through lentiviral shRNA complexes on CD4+CD25+ T cells detected by peptide-specific, major histo- strategy resulted in decreased level of the full-length protein in compatibility complex-restricted antibodies in patients with HTLV-I–associated infected lymphocyte; 3) the full-length CRMP2 form, known as neurologic disease. J. Exp. Med. 199: 1367–1377. 5. Cavrois, M., A. Gessain, O. Gout, S. Wain-Hobson, and E. Wattel. 2000. Com- a direct modulator of microtubules (44), was bound to the micro- mon human T cell leukemia virus type 1 (HTLV-1) integration sites in cerebro- tubule protein tubulin in HTLV-1–infected T cell line and not spinal fluid and blood lymphocytes of patients with HTLV-1–associated in control T cells; and 4) the presence of full-length CRMP2 in myelopathy/tropical spastic paraparesis indicate that HTLV-1 crosses the blood– brain barrier via clonal HTLV-1–infected cells. J. Infect. Dis. 182: 1044–1050. infected T cell clones correlated with elevated migratory rate. The 6. Nagai, M., Y. Yamano, M. B. Brennan, C. A. Mora, and S. Jacobson. 2001. high level of full-length CRMP2 in infected T lymphocytes could Increased HTLV-I proviral load and preferential expansion of HTLV-I Tax- result either of a less effective protein cleavage or a modified specific CD8+ T cells in cerebrospinal fluid from patients with HAM/TSP. Ann. Neurol. 50: 807–812. stability of one or another form of CRMP2. We observed that 7. Osame, M. 2002. Pathological mechanisms of human T-cell lymphotropic virus HTLV-1 infection could result in an important increase of the type I-associated myelopathy (HAM/TSP). J. Neurovirol. 8: 359–364. proteasomal degradation of the short CRMP2, supporting the mod- 8. Yao, J., and B. Wigdahl. 2000. Human T cell lymphotropic virus type I genomic expression and impact on intracellular signaling pathways during neurodegen- ifications of the 58:62-kDa ratio observed. Although the precise erative disease and leukemia. Front. Biosci. 5: D138–D168. mechanism of CRMP2 modulation remains unclear, it obviously 9. Lezin, A., S. Olindo, S. Oliere, M. Varrin-Doyer, R. Marlin, P. Cabre, D. Smadja, appears that HTLV-1 exploits a fundamental cell regulatory mech- and R. Cesaire. 2005. Human T lymphotropic virus type I (HTLV-I) proviral load Downloaded from in cerebrospinal fluid: a new criterion for the diagnosis of HTLV-I–associated anism to enhance lymphocyte motility. myelopathy/tropical spastic paraparesis? J. Infect. Dis. 191: 1830–1834. The viral protein Tax appeared to be involved in the modulation 10. Hayashi, D., R. Kubota, N. Takenouchi, T. Nakamura, F. Umehara, K. Arimura, of CRMP2 level and cleavage but not in the enhanced phosphor- S. Izumo, and M. Osame. 2008. Accumulation of human T-lymphotropic virus type I (HTLV-I)–infected cells in the cerebrospinal fluid during the exacerbation ylation displayed by HTLV-1–infected lymphocytes, suggesting of HTLV-I–associated myelopathy. J. Neurovirol. 14: 459–463. that other viral proteins can modify CRMP2 expression and ac- 11. Alefantis, T., K. Mostoller, P. Jain, E. Harhaj, C. Grant, and B. Wigdahl. 2005. Secretion of the human T cell leukemia virus type I transactivator protein tax. tivity. Nevertheless, colocalization of Tax and CRMP2 at cell sites http://www.jimmunol.org/ J. Biol. Chem. 280: 17353–17362. important for cell migration (uropod) and molecule/signal delivery 12. Szymocha, R., H. Akaoka, M. Dutuit, C. Malcus, M. Didier-Bazes, M. F. Belin, (membrane cell–cell contact) strongly suggests the cooperation and P. Giraudon. 2000. Human T-cell lymphotropic virus type 1-infected of these two molecules to support functions that are important for T lymphocytes impair catabolism and uptake of glutamate by astrocytes via Tax- 1 and tumor necrosis factor alpha. J. Virol. 74: 6433–6441. HTLV-1 virus/protein transmission (79, 80). This interesting aspect 13. Ramirez, E., L. Cartier, M. Torres, and M. Barria. 2007. Temporal dynamics of of CRMP2 function in virus-infected lymphocyte remains to be human T-lymphotropic virus type I tax mRNA and proviral DNA load in pe- ripheral blood mononuclear cells of human T-lymphotropic virus type I- analyzed. associated myelopathy patients. J. Med. Virol. 79: 782–790. Other neuroinflammatory/neurodegenerative diseases share the 14. Szymocha, R., C. Brisson, A. Bernard, H. Akaoka, M. F. Belin, and P. Giraudon. importance of infiltrating T lymphocytes in pathogenesis. A link 2000. Long-term effects of HTLV-1 on brain astrocytes: sustained expression of Tax-1

associated with synthesis of inflammatory mediators. J. Neurovirol. 6: 350–357. by guest on September 28, 2021 between diffuse axonal loss, a key feature of multiple sclerosis, and 15. Van Prooyen, N., V. Andresen, H. Gold, I. Bialuk, C. Pise-Masison, and the presence of meningeal T lymphocytes in spinal cord of patients G. Franchini. 2010. Hijacking the T-cell communication network by the human was reported recently (81). In addition, recent studies assigned an T-cell leukemia/lymphoma virus type 1 (HTLV-1) p12 and p8 proteins. Mol. Aspects Med. 31: 333–343. unexpected role to infiltrated T lymphocytes in the pathophysiology 16. Engelhardt, B., and R. M. Ransohoff. 2005. The ins and outs of T-lymphocyte of Parkinson’s disease and Amyotrophic lateral sclerosis (82, 83). trafficking to the CNS: anatomical sites and molecular mechanisms. Trends Chronic activation of cell signaling programs that directs immune Immunol. 26: 485–495. 17. Engelhardt, B. 2006. Molecular mechanisms involved in T cell migration across cell motility and includes CRMP2 could be a common mechanism the blood–brain barrier. J. Neural Transm. 113: 477–485. leading to CNS infiltration and tissue damage in various neuro- 18. Springer, T. A. 1994. Traffic signals for lymphocyte recirculation and leukocyte inflammatory diseases. This point merits to be investigated. emigration: the multistep paradigm. Cell 76: 301–314. 19. Ley, K., C. Laudanna, M. I. Cybulsky, and S. Nourshargh. 2007. Getting to the To conclude, our findings provide an additional clue to under- site of inflammation: the leukocyte adhesion cascade updated. Nat. Rev. Immu- standing the migration and infiltration of HTLV-1–infected lym- nol. 7: 678–689. phocytes into various tissues and suggest that the regulation of 20. Luster, A. D., R. Alon, and U. H. von Andrian. 2005. Immune cell migration in inflammation: present and future therapeutic targets. Nat. Immunol. 6: 1182–1190. CRMP2 activity by retrovirus infection is a novel aspect of virus- 21. Sa´nchez-Madrid, F., and M. A. del Pozo. 1999. Leukocyte polarization in cell induced neuroinflammation. migration and immune interactions. EMBO J. 18: 501–511. 22. del Pozo, M. A., M. Vicente-Manzanares, R. Tejedor, J. M. Serrador, and F. Sa´nchez-Madrid. 1999. Rho GTPases control migration and polarization of Acknowledgments adhesion molecules and cytoskeletal ERM components in T lymphocytes. Eur. We thank Dr. Antoine Gessain, Dr. Simona Ozden (Pasteur Institute, Paris), J. Immunol. 29: 3609–3620. and Claudine Pique (Institute Cochin, Paris) for fruitful discussions, pro- 23. Goshima, Y., F. Nakamura, P. Strittmatter, and S. M. Strittmatter. 1995. Collapsin-induced growth cone collapse mediated by an intracellular protein viding immune cells of HAM/TSP patients, and the TRIP-Tax vector. We related to UNC-33. Nature 376: 509–514. also thank Joel Lachuer and Anne Wierinckx of the ProfileXpert platform 24. Vincent, P., Y. Collette, R. Marignier, C. Vuaillat, V. Rogemond, N. Davoust, (Lyon, France) for fruitful discussions and help. C. Malcus, S. Cavagna, A. Gessain, I. Machuca-Gayet, et al. 2005. A role for the neuronal protein collapsin response mediator protein 2 in T lymphocyte polar- ization and migration. J. Immunol. 175: 7650–7660. Disclosures 25. Varrin-Doyer, M., P. Vincent, S. Cavagna, N. Auvergnon, N. Noraz, The authors have no financial conflicts of interest. V. Rogemond, J. Honnorat, M. Moradi-Ame´li, and P. Giraudon. 2009. Phos- phorylation of collapsin response mediator protein 2 on Tyr-479 regulates CXCL12-induced T lymphocyte migration. J. Biol. Chem. 284: 13265–13276. 26. Vuaillat, C., M. Varrin-Doyer, A. Bernard, I. Sagardoy, S. Cavagna, References I. Chounlamountri, M. Lafon, and P. Giraudon. 2008. High CRMP2 expression 1. Ransohoff, R. M. 2007. Inflammatory disease: assault on the guardian. Nature in peripheral T lymphocytes is associated with recruitment to the brain during 448: 421–422. virus-induced neuroinflammation. J. Neuroimmunol. 193: 38–51. 2. Hinuma, Y., K. Nagata, M. Hanaoka, M. Nakai, T. Matsumoto, K. I. Kinoshita, 27. Sibon, D., A. S. Gabet, M. Zandecki, C. Pinatel, J. Theˆte, M. H. Delfau-Larue, S. Shirakawa, and I. Miyoshi. 1981. Adult T-cell leukemia: antigen in an ATL S. Rabaaoui, A. Gessain, O. Gout, S. Jacobson, et al. 2006. HTLV-1 propels cell line and detection of antibodies to the antigen in human sera. Proc. Natl. untransformed CD4 lymphocytes into the cell cycle while protecting CD8 cells Acad. Sci. USA 78: 6476–6480. from death. J. Clin. Invest. 116: 974–983. 1232 RETROVIRUS MANIPULATES LYMPHOCYTE MOTILITY MACHINERY

28. Delfau-Larue, M. H., L. Laroche, J. Wechsler, E. Lepage, C. Lahet, M. Asso- 53. Lepoutre, V., P. Jain, K. Quann, B. Wigdahl, and Z. K. Khan. 2009. Role of Bonnet, M. Bagot, and J. P. Farcet. 2000. Diagnostic value of dominant T-cell resident CNS cell populations in HTLV-1–associated neuroinflammatory disease. clones in peripheral blood in 363 patients presenting consecutively with a clin- Front. Biosci. 14: 1152–1168. ical suspicion of cutaneous lymphoma. Blood 96: 2987–2992. 54. Biddison, W. E., R. Kubota, T. Kawanishi, D. D. Taub, W. W. Cruikshank, 29. Rogemond, V., C. Auger, P. Giraudon, M. Becchi, N. Auvergnon, M. F. Belin, D. M. Center, E. W. Connor, U. Utz, and S. Jacobson. 1997. Human T cell + J. Honnorat, and M. Moradi-Ame´li. 2008. Processing and nuclear localization of leukemia virus type I (HTLV-I)–specific CD8 CTL clones from patients with CRMP2 during brain development induce neurite outgrowth inhibition. J. Biol. HTLV-I–associated neurologic disease secrete proinflammatory cytokines, che- Chem. 283: 14751–14761. mokines, and matrix metalloproteinase. J. Immunol. 159: 2018–2025. 30. Royer-Leveau, C., E. Mordelet, F. Delebecque, A. Gessain, P. Charneau, and 55. Kubota, R., F. Umehara, S. Izumo, S. Ijichi, K. Matsumuro, S. Yashiki, S. Ozden. 2002. Efficient transfer of HTLV-1 tax gene in various primary and T. Fujiyoshi, S. Sonoda, and M. Osame. 1994. HTLV-I proviral DNA amount + immortalized cells using a flap lentiviral vector. J. Virol. Methods 105: 133–140. correlates with infiltrating CD4 lymphocytes in the spinal cord from patients 31. Chiari, E., I. Lamsoul, J. Lodewick, C. Chopin, F. Bex, and C. Pique. 2004. with HTLV-I–associated myelopathy. J. Neuroimmunol. 53: 23–29. Stable ubiquitination of human T-cell leukemia virus type 1 tax is required for 56. Giraudon, P., S. Buart, A. Bernard, N. Thomasset, and M. F. Belin. 1996. Ex- proteasome binding. J. Virol. 78: 11823–11832. tracellular matrix-remodeling metalloproteinases and infection of the central 32. Zane, L., D. Sibon, C. Legras, J. Lachuer, A. Wierinckx, P. Mehlen, nervous system with retrovirus human T-lymphotropic virus type I (HTLV-I). M. H. Delfau-Larue, A. Gessain, O. Gout, C. Pinatel, et al. 2010. Clonal ex- Prog. Neurobiol. 49: 169–184. pansion of HTLV-1–positive CD8+ cells relies on cIAP-2 but not on c-FLIP 57. Moritoyo, T., T. A. Reinhart, H. Moritoyo, E. Sato, S. Izumo, M. Osame, and A. T. Haase. 1996. Human T-lymphotropic virus type I-associated myelopathy expression. Virology 407: 341–351. + 33. Giraudon, P., R. Szymocha, S. Buart, A. Bernard, L. Cartier, M. F. Belin, and and tax gene expression in CD4 T lymphocytes. Ann. Neurol. 40: 84–90. 58. Lindholm, P. F., R. L. Reid, and J. N. Brady. 1992. Extracellular Tax1 protein H. Akaoka. 2000. T lymphocytes activated by persistent viral infection differ- stimulates tumor necrosis factor-b and immunoglobulin k light chain expression entially modify the expression of metalloproteinases and their endogenous in lymphoid cells. J. Virol. 66: 1294–1302. inhibitors, TIMPs, in human astrocytes: relevance to HTLV-I–induced neuro- 59. Barmak, K., E. W. Harhaj, and B. Wigdahl. 2003. Mediators of central nervous logical disease. J. Immunol. 164: 2718–2727. system damage during the progression of human T-cell leukemia type I- ` 34. Fevre-Montange, M., J. Champier, A. Durand, A. Wierinckx, J. Honnorat, associated myelopathy/tropical spastic paraparesis. J. Neurovirol. 9: 522–529. J. Guyotat, and A. Jouvet. 2009. Microarray gene expression profiling in me- 60. Kambara, C., T. Nakamura, T. Furuya, K. Migita, H. Ida, A. Kawakami, Downloaded from ningiomas: differential expression according to grade or histopathological sub- S. Shirabe, S. Nakane, I. Kinoshita, and K. Eguchi. 1999. Vascular cell adhesion type. Int. J. Oncol. 35: 1395–1407. molecule-1–mediated matrix metalloproteinase-2 induction in peripheral blood 35. Macatonia, S. E., J. K. Cruickshank, P. Rudge, and S. C. Knight. 1992. Dendritic T cells is up-regulated in patients with HTLV-I–associated myelopathy. J. cells from patients with tropical spastic paraparesis are infected with HTLV-1 Neuroimmunol. 99: 242–247. and stimulate autologous lymphocyte proliferation. AIDS Res. Hum. Retro- 61. Uchiyama, T., T. Ishikawa, and A. Imura. 1996. Cell adhesion molecules in viruses 8: 1699–1706. HTLV-I infection. J. Acquir. Immune Defic. Syndr. Hum. Retrovirol. 13(Suppl. 1): 36. Matsuura, E., Y. Yamano, and S. Jacobson. 2010. Neuroimmunity of HTLV-I S114–S118.

infection. J. Neuroimmune Pharmacol. 5: 310–325. 62. Kim, S. J., A. M. Nair, S. Fernandez, L. Mathes, and M. D. Lairmore. 2006. http://www.jimmunol.org/ 37. Boxus, M., J. C. Twizere, S. Legros, J. F. Dewulf, R. Kettmann, and L. Willems. Enhancement of LFA-1–mediated T cell adhesion by human T lymphotropic 2008. The HTLV-1 Tax interactome. Retrovirology 5: 76. virus type 1 p12I1. J. Immunol. 176: 5463–5470. 38. Cole, A. R., M. P. Soutar, M. Rembutsu, L. van Aalten, C. J. Hastie, 63. Kodama, Y., Y. Murakumo, M. Ichihara, K. Kawai, Y. Shimono, and H. McLauchlan, M. Peggie, M. Balastik, K. P. Lu, and C. Sutherland. 2008. M. Takahashi. 2004. Induction of CRMP-2 by GDNF and analysis of the CRMP- Relative resistance of Cdk5-phosphorylated CRMP2 to dephosphorylation. J. 2 promoter region. Biochem. Biophys. Res. Commun. 320: 108–115. Biol. Chem. 283: 18227–18237. 64. Pallas, D. C., L. K. Shahrik, B. L. Martin, S. Jaspers, T. B. Miller, D. L. Brautigan, 39. Yuasa-Kawada, J., R. Suzuki, F. Kano, T. Ohkawara, M. Murata, and M. Noda. and T. M. Roberts. 1990. Polyoma small and middle T antigens and SV40 small t 2003. Axonal morphogenesis controlled by antagonistic roles of two CRMP antigen form stable complexes with protein phosphatase 2A. Cell 60: 167–176. subtypes in microtubule organization. Eur. J. Neurosci. 17: 2329–2343. 65. Garcia, A., S. Cereghini, and E. Sontag. 2000. Protein phosphatase 2A and 40. Cole, A. R., F. Causeret, G. Yadirgi, C. J. Hastie, H. McLauchlan, phosphatidylinositol 3-kinase regulate the activity of Sp1-responsive promoters. E. J. McManus, F. Herna´ndez, B. J. Eickholt, M. Nikolic, and C. Sutherland. J. Biol. Chem. 275: 9385–9389. 2006. Distinct priming kinases contribute to differential regulation of collapsin 66. Yang, S. I., R. L. Lickteig, R. Estes, K. Rundell, G. Walter, and M. C. Mumby. by guest on September 28, 2021 response mediator proteins by glycogen synthase kinase-3 in vivo. J. Biol. Chem. 1991. Control of protein phosphatase 2A by simian virus 40 small-t antigen. Mol. 281: 16591–16598. Cell. Biol. 11: 1988–1995. 41. Yoshimura, T., N. Arimura, Y. Kawano, S. Kawabata, S. Wang, and K. Kaibuchi. 67. Fu, D. X., Y. L. Kuo, B. Y. Liu, K. T. Jeang, and C. Z. Giam. 2003. Human T-lym- 2006. Ras regulates neuronal polarity via the PI3-kinase/Akt/GSK-3b/CRMP-2 photropic virus type I tax activates I-kB kinase by inhibiting I-kB kinase- pathway. Biochem. Biophys. Res. Commun. 340: 62–68. associated serine/threonine protein phosphatase 2A. J. Biol. Chem. 278: 1487– 42. Uchida, Y., T. Ohshima, Y. Sasaki, H. Suzuki, S. Yanai, N. Yamashita, 1493. F. Nakamura, K. Takei, Y. Ihara, K. Mikoshiba, et al. 2005. Semaphorin3A 68. Sun, Y., T. Fei, T. Yang, F. Zhang, Y. G. Chen, H. Li, and Z. Xu. 2010. The signalling is mediated via sequential Cdk5 and GSK3b phosphorylation of suppression of CRMP2 expression by bone morphogenetic protein (BMP)- CRMP2: implication of common phosphorylating mechanism underlying axon SMAD gradient signaling controls multiple stages of neuronal development. J. guidance and Alzheimer’s disease. Genes Cells 10: 165–179. Biol. Chem. 285: 39039–39050. 43. Rousset, R., C. Desbois, F. Bantignies, and P. Jalinot. 1996. Effects on NF-kB1/ 69. Grant, C., U. Oh, K. Yao, Y. Yamano, and S. Jacobson. 2008. Dysregulation of p105 processing of the interaction between the HTLV-1 transactivator Tax and TGF-b signaling and regulatory and effector T-cell function in virus-induced the proteasome. Nature 381: 328–331. neuroinflammatory disease. Blood 111: 5601–5609. ´ 44. Gu, Y., and Y. Ihara. 2000. Evidence that collapsin response mediator protein-2 70. Arimura, N., C. Menager, Y. Kawano, T. Yoshimura, S. Kawabata, A. Hattori, Y. Fukata, M. Amano, Y. Goshima, M. Inagaki, et al. 2005. Phosphorylation by Rho is involved in the dynamics of microtubules. J. Biol. Chem. 275: 17917–17920. kinase regulates CRMP-2 activity in growth cones. Mol. Cell. Biol. 25: 9973–9984. 45. Arimura, N., T. Kimura, S. Nakamuta, S. Taya, Y. Funahashi, A. Hattori, 71. Uchida, Y., T. Ohshima, N. Yamashita, M. Ogawara, Y. Sasaki, F. Nakamura, and A. Shimada, C. Me´nager, S. Kawabata, K. Fujii, et al. 2009. Anterograde Y. Goshima. 2009. Semaphorin3A signaling mediated by Fyn-dependent tyro- transport of TrkB in axons is mediated by direct interaction with Slp1 and sine phosphorylation of collapsin response mediator protein 2 at tyrosine 32. J. Rab27. Dev. Cell 16: 675–686. Biol. Chem. 284: 27393–27401. 46. Rahajeng, J., S. S. Giridharan, N. Naslavsky, and S. Caplan. 2010. Collapsin 72. Cain, R. J., and A. J. Ridley. 2009. Phosphoinositide 3-kinases in cell migration. response mediator protein-2 (Crmp2) regulates trafficking by linking endocytic Biol. Cell 101: 13–29. regulatory proteins to dynein motors. J. Biol. Chem. 285: 31918–31922. 73. Saito, K., M. Saito, N. Taniura, T. Okuwa, and Y. Ohara. 2010. Activation of the 47. Nejmeddine, M., V. S. Negi, S. Mukherjee, Y. Tanaka, K. Orth, G. P. Taylor, and PI3K-Akt pathway by human T cell leukemia virus type 1 (HTLV-1) oncoprotein C. R. Bangham. 2009. HTLV-1-Tax and ICAM-1 act on T-cell signal pathways to Tax increases Bcl3 expression, which is associated with enhanced growth of polarize the microtubule-organizing center at the virological synapse. Blood 114: HTLV-1–infected T cells. Virology 403: 173–180. 1016–1025. 74. Mahana, W., T. M. Zhao, R. Teller, M. A. Robinson, and T. J. Kindt. 1998. Genes 48. Babich, A., and J. K. Burkhardt. 2011. Lymphocyte signaling converges on in the pX region of human T cell leukemia virus I influence Vav phosphorylation microtubules. Immunity 34: 825–827. in T cells. Proc. Natl. Acad. Sci. USA 95: 1782–1787. 49. Gessain, A., and O. Gout. 1992. Chronic myelopathy associated with human T- 75. Kawaguchi, A., Y. Orba, T. Kimura, H. Iha, M. Ogata, T. Tsuji, A. Ainai, T. Sata, lymphotropic virus type I (HTLV-I). Ann. Intern. Med. 117: 933–946. T. Okamoto, W. W. Hall, et al. 2009. Inhibition of the SDF-1a-CXCR4 axis by 50. Tangy, F., M. Ossondo, J. C. Vernant, D. Smadja, O. Ble´try, A. C. Baglin, and the CXCR4 antagonist AMD3100 suppresses the migration of cultured cells S. Ozden. 1999. Human T cell leukemia virus type I expression in salivary from ATL patients and murine lymphoblastoid cells from HTLV-I Tax transgenic glands of infected patients. J. Infect. Dis. 179: 497–502. mice. Blood 114: 2961–2968. 51. Sueyoshi, K., M. Goto, M. Johnosono, E. Sato, and D. Shibata. 1994. Anatomical 76. Twizere, J. C., J. Y. Springael, M. Boxus, A. Burny, F. Dequiedt, J. F. Dewulf, distribution of HTLV-I proviral sequence in an autopsy case of HTLV-I–asso- J. Duchateau, D. Portetelle, P. Urbain, C. Van Lint, et al. 2007. Human T-cell leukemia ciated myelopathy: a polymerase chain reaction study. Pathol. Int. 44: 27–33. virus type-1 Tax oncoprotein regulates G-protein signaling. Blood 109: 1051–1060. 52. Ozden, S., M. Cochet, J. Mikol, A. Teixeira, A. Gessain, and C. Pique. 2004. 77. Zhang, X., K. Deguchi, T. Yamashita, Y. Ohta, J. Shang, F. Tian, N. Liu, Direct evidence for a chronic CD8+-T-cell–mediated immune reaction to tax V. L. Panin, Y. Ikeda, T. Matsuura, and K. Abe. 2010. Temporal and spatial within the muscle of a human T-cell leukemia/lymphoma virus type 1-infected differences of multiple protein expression in the ischemic penumbra after tran- patient with sporadic inclusion body myositis. J. Virol. 78: 10320–10327. sient MCAO in rats. Brain Res. 1343: 143–152. The Journal of Immunology 1233

78. Hou, S. T., A. Keklikian, J. Slinn, M. O’Hare, S. X. Jiang, and A. Aylsworth. 81. Androdias, G., R. Reynolds, M. Chanal, C. Ritleng, C. Confavreux, and S. Nataf. 2008. Sustained up-regulation of semaphorin 3A, Neuropilin1, and doublecortin 2010. Meningeal T cells associate with diffuse axonal loss in multiple sclerosis expression in ischemic mouse brain during long-term recovery. Biochem. Bio- spinal cords. Ann. Neurol. 68: 465–476. phys. Res. Commun. 367: 109–115. 82. Holmøy, T. 2008. T cells in amyotrophic lateral sclerosis. Eur. J. Neurol. 15: 79. Barnard, A. L., T. Igakura, Y. Tanaka, G. P. Taylor, and C. R. Bangham. 2005. 360–366. Engagement of specific T-cell surface molecules regulates cytoskeletal polari- 83. Brochard, V., B. Combadie`re, A. Prigent, Y. Laouar, A. Perrin, V. Beray-Berthat, zation in HTLV-1–infected lymphocytes. Blood 106: 988–995. O. Bonduelle, D. Alvarez-Fischer, J. Callebert, J. M. Launay, et al. 2009. Infil- 80. Nejmeddine, M., and C. R. Bangham. 2010. The HTLV-1 virological synapse. tration of CD4+ lymphocytes into the brain contributes to neurodegeneration in Viruses 2: 1427–1447. a mouse model of Parkinson disease. J. Clin. Invest. 119: 182–192. Downloaded from http://www.jimmunol.org/ by guest on September 28, 2021