CXCR4 and Vascular Cell Adhesion Molecule 1 Are Key Chemokine/Adhesion Receptors in the Migration of Cytokine-Activated T Cells

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 64, No. 7, July 2012, pp DOI /art , American College of Rheumatology CXCR4 and Vascular Cell Adhesion Molecule 1 Are Key Chemokine/Adhesion Receptors in the Migration of Cytokine-Activated T Cells Jane Bryant, David J. Ahern, and Fionula M. Brennan Objective. To examine the migratory properties of cytokine-activated T (Tck) cells. Methods. Tck cells were generated by culture of peripheral blood T cells in the presence of interleukin-6 (IL-6), tumor necrosis factor, and IL-2. Changes in cell surface phenotype were analyzed by flow cytometry. Chemotactic responsiveness was measured using in vitro chemotaxis assays and transendothelial migration through human umbilical vein endothelial cell monolayers. Levels of vascular cell adhesion molecule 1 (VCAM-1) were measured by sandwich enzyme-linked immunosorbent assay. Results. Cytokine stimulation up-regulated the expression of chemokine receptors and integrins on Tck cells, including CXCR4, very late activation antigen 4 (VLA-4), and lymphocyte function associated antigen 1. Increased expression of CXCR4 and VLA-4 integrin resulted in concentration-dependent chemotaxis to their ligands, stromal cell derived factor 1 (SDF-1) and VCAM-1, which could be selectively blocked using a specific CXCR4 inhibitor and antibodies against VLA-4. Increased expression of VLA-4 also resulted in increased transendothelial migration of Tck cells, which could be abrogated using blocking antibodies against VLA-4. Tck cells also showed an increased chemotactic Supported by a core grant to the Kennedy Institute of Rheumatology from Arthritis Research UK. Jane Bryant, PhD, David J. Ahern, PhD, Fionula M. Brennan, PhD, FRCPath: Kennedy Institute of Rheumatology, University of Oxford, London, UK. The work was completed at the Kennedy Institute of Rheumatology when it was a division of Imperial College London; the Institute became a division of the University of Oxford on August 1, Dr. Brennan has received consulting fees from Wyeth, which is now part of Pfizer (less than $10,000). Address correspondence to David J. Ahern, PhD, Kennedy Institute of Rheumatology, University of Oxford, 65 Aspenlea Road, London W6 8LH, UK. david.ahern@kennedy.ox.ac.uk. Submitted for publication May 11, 2011; accepted in revised form January 12, response to rheumatoid arthritis (RA) fibroblast-like synoviocytes cultured in vitro, which could be blocked using inhibitors against VLA-4 and CXCR4. Conclusion. The activated phenotype of Tck cells results in increased migratory responsiveness to SDF-1 and soluble VCAM-1, which are among the chemokines and proteins found elevated in the RA synovial joint environment. Cytokine-dependent activation may contribute to RA pathogenicity by promoting T cell recruitment to and retention in the joint, perpetuating the inflammatory cascade in RA. Rheumatoid arthritis (RA) has a prevalence of 1% of the population. It results in reduced mobility and significant morbidity and is characterized by chronic inflammation arising from the activities of cells within the synovial tissue (1). In healthy individuals, the synovial tissue is relatively acellular; however, in RA, there is an influx of immune cells, which causes and maintains inflammation. These immune cells consist mainly of macrophages and T lymphocytes (2,3), which together with the resident cell types, such as synovial fibroblasts, contribute to the disease pathology. The importance of T cells in the disease is evidenced by the presence of the shared epitope in the HLA class II antigen, polymorphisms in T cell signaling molecules increasing RA susceptibility, and, more recently, T cell reactivity to citrullinated proteins (4 6). It has been shown that large numbers of T cells are recruited to the synovium in clinical disease (1) including both typical and atypical T cell phenotypes. These consist mainly of CD4 CD45RO CD45RB low memory T cells; however, there is evidence of CD45RO CD45RA CD45RB high naive T cells and double-positive CD45RO CD45RA T cells (7,8). These cells also show varying expression of the costimulatory molecules CD27 and CD28 as well as the cell adhesion marker CD62L, which is expressed mainly by naive T cells (7 9). 2137

2 2138 BRYANT ET AL RA synovial T cells exhibit differential expression of adhesion molecules and chemokine receptors compared to peripheral blood T cells from RA patients and healthy controls, which may affect their homing and migration to the joints (refs. 10 and 11, and for review, see ref. 12). These markers may also differentially affect their interaction with other cells present within the RA synovium, including M1-type macrophages, mature and immature plasmacytoid dendritic cells, endothelial cells, and the expanded resident fibroblast-like synoviocytes (FLS). It has also been shown that T cells can be activated independently of antigen by cytokine stimulation. The first finding that T cells could be activated independently of antigen was described in 1994 by Unutmaz et al (13) and was subsequently confirmed by our group and others (14 16). Our group hypothesized that the activation state of T cells within the joints of RA patients was maintained by cytokines present within the disease tissue in a bystander manner and independently of antigen. We subsequently showed that peripheral blood T cells cultured for 8 days in the presence of tumor necrosis factor (TNF ), interleukin-2 (IL-2), and IL-6 could stimulate production of proinflammatory cytokines from monocytes in a contact-dependent manner. Furthermore, these cytokine-activated T (Tck) cells demonstrated an effector function identical to that of T cells isolated from the RA synovium (16,17). More recently, we have shown that the most potent Tck effector cells reside within the CD4 CD45RO effector memory population (17). We reported the increased expression of a number of activation markers, chemokine receptors, and adhesion and integrin molecules on Tck cells, a phenotype that resembled that of T cells extracted from synovial tissue of RA patients. In that previous study we highlighted the significantly increased expression of the chemokine receptor CXCR4 and the adhesion molecules CD18, CD29, and CD49d on Tck cells, suggesting that Tck cells may show increased migration and trafficking to the synovial microenvironment. Cell migration has been implicated in the pathogenesis of RA due to an influx of cells into the joint, and it is controlled by the chemokine receptor/ligand network. Receptor/ligand pairs show high levels of promiscuity, which leads to heterogenous cell types migrating to and being retained in the RA synovium; however, there is a degree of specificity. Stromal cell derived factor 1 (SDF-1), the specific ligand for CXCR4, has been detected in the RA synovium and is of special interest since it predominantly attracts CD4 CD45RO memory T cells which express the receptor CXCR4 and are the main T cell subset implicated in RA pathogenesis (18,19). Cell migration also involves affinity up-regulation of integrins such as very late activation antigen 4 (VLA-4) and VLA-5, shown to be involved in the binding of inflammatory cells to extracellular matrix components such as fibronectin (11). VLA-4 is increased up to 10-fold in synovial cells compared to resting peripheral blood lymphocytes (PBLs) (10) and binds vascular cell adhesion molecule 1 (VCAM-1), which is expressed on both endothelium and macrophages in RA synovium. Furthermore, soluble VCAM-1 (svcam-1) is found circulating at high levels in RA and has been shown to induce migration of a number of cell types, including synovial fibroblasts, eosinophils, T cells, and monocytes (20 22). In the present study, we confirm and extend our previous findings on the up-regulation of chemokine receptors and adhesion molecules on Tck cells and investigate the chemotactic and migratory properties of these cells. We demonstrate that Tck cells undergo increased migration in response to SDF-1 and svcam-1 when compared to matched resting T cells or T cells activated through the T cell receptor (TCR), and this is dependent on expression of CXCR4 and VLA-4, respectively. Furthermore, Tck demonstrate increased transendothelial migration through TNF -stimulated human umbilical vein endothelial cell (HUVEC) monolayers, which is partially mediated through VLA-4/ VCAM-1. We also show increased chemotaxis of Tck cells to RA FLS. MATERIALS AND METHODS Patients. Synovial membranes were obtained from patients undergoing elective joint replacement procedures. All procedures received local ethics committee approval (Riverside Research Ethics Committee no. 1752). Tissue was obtained from RA patients fulfilling the American College of Rheumatology 1987 revised classification criteria (23). Singlecell suspensions were prepared by mechanical and enzymatic disruption as previously described (24). RA FLS were generated by subculture of synovial cells and removal of the nonadherent population. RA FLS were used between the first and second passages. Human umbilical cords were collected from Chelsea and Westminster Hospital (London, UK) with informed consent after approval (no. 2948) by the Riverside Research Ethics Committee. HUVECs were isolated by enzymatic digestion as previously described (25). Endothelial cells were used between the second and fourth passages for all experiments. Proteins and antibodies. TNF, IL-2, and IL-6 were kindly gifted by Boehringer Ingelheim, the National Institutes of Health, and Novartis, respectively. Flow cytometry anti-

3 MIGRATORY PROPERTIES OF Tck 2139 bodies to CXCR4, CXCR7, CCR6, CCR8, and CD49e were purchased from R&D Systems. Flow cytometry antibodies to CD49d, CD29, CD18, CD11a, 7 integrin, CXCR2, CXCR3, and CCR5 were from BD Biosciences. Blocking antibodies against CD29 (clone 4B4), CD49d (HP2/1), and VCAM-1 (P8B1) were from Beckman Coulter, Serotec, and Millipore, respectively. Isotype control antibodies IgG1 (107.3) and IgG2a (G ) were from BD Biosciences. CXCR4 antagonist AMD3100 was purchased from Sigma-Aldrich. Recombinant SDF-1 and recombinant svcam-1 were from Pepro- Tech and R&D Systems, respectively. Anti-CD3 and anti-cd28 antibodies were from BD Biosciences. Carboxyfluorescein succinimidyl ester (CFSE) was from Invitrogen. Cells were cultured in RPMI 1640/L-glutamine (PAA Laboratories) supplemented with 10% normal AB human serum (Sigma- Aldrich) or 10% fetal calf serum (Biosera), as appropriate. Lymphocyte purification and CD4 cell enrichment. Single-donor plateletpheresis residues were purchased from the North London Blood Transfusion service and then diluted 1:1 with sterile phosphate buffered saline and human peripheral blood mononuclear cells isolated by density-gradient centrifugation using Lympholyte (Cedarlane). Lymphocytes were then enriched by elutriation (Beckman Coulter). CD4 T cells were isolated from PBLs using CD4 immunomagnetic bead isolation kits (Invitrogen) in accordance with the manufacturer s instructions. CD4 T cell culture and phenotypic analysis. Tck cells were generated from CD4 T cells by culture in RPMI 1640 supplemented with 10% human AB serum in the presence of IL-2 (25 ng/ml), TNF (25 ng/ml), and IL-6 (100 ng/ml) for 8 days at 37 C in 5% CO 2. TCR-stimulated T (Ttcr) cells were prepared by stimulation of CD4 T cells with plate-bound anti-cd3 (2 g/ml) and soluble anti-cd28 (4 g/ml) for 2 days at 37 C in 5% CO 2. Cell surface marker expression was analyzed on a Becton Dickinson FACSCanto II flow cytometer after staining with fluorochrome-conjugated antibodies. In vitro chemotaxis assay. Cell migration was investigated using 96-well chemotaxis plates with 3- m pore size polycarbonate filters (Neuroprobe). Plate wells were blocked with 1% bovine serum albumin (BSA) followed by placing a range of concentrations of SDF-1 and svcam-1 diluted in assay buffer (RPMI 1640 with 0.1% BSA) in the lower plate wells. Cells ( ) in a 60- l volume were placed in the upper wells. In some cases, Tck cells were pretreated with neutralizing antibodies (anti-cd29 and anti-cd49d) at 10 g/ ml, or with a small molecule inhibitor (AMD3100) at 50 ng/ml for 30 minutes prior to addition to the assay. After a 2-hour incubation, the migrated cells in the lower wells were collected and counted using flow cytometry. Checkerboard analysis was used to determine if cell migration was gradient-dependent by placing a range of concentrations of SDF-1 or svcam-1 in the upper wells as well as in the lower wells of the assay. In vitro chemotaxis assay of Tck cell migration to cultured RA FLS. Tck cell migration to RA FLS was investigated using 96-well chemotaxis plates with 3- m pore size polycarbonate filters. FLS were cultured in the lower wells at a concentration of /well for 18 hours prior to the assay. T cells were stained with 5 M CFSE as previously described (17) and added to the upper wells. After a 2-hour incubation, CFSE migrated cells were collected and counted using flow cytometry. In vitro transmigration assay. HUVECs were stimulated for 18 hours in the presence of 10 ng/ml TNF prior to addition of cells to fibronectin-coated Transwell inserts (Corning) with 3- m pore size filters. HUVECs were allowed to adhere overnight to form a monolayer. Resting T, Ttcr, or Tck cells ( each) in a 500- l volume were added to each monolayer, and the assay was incubated for 24 hours at 37 C in 5% CO 2. Migrated cells in the lower chamber were sampled and counted using trypan blue exclusion at various time points as indicated. For some studies, Tck cells and HUVECs were pretreated with neutralizing antibodies at a final concentration of 10 g/ml for 30 minutes prior to addition of T cells to the monolayers. Soluble VCAM-1 produced by HUVECs was measured in cell culture supernatants by sandwich enzyme-linked immunosorbent assay (ELISA). Statistical analysis. Results were analyzed using GraphPad Prism 5 software, and statistical differences between groups were analyzed using Wilcoxon s rank sum test, paired t-test, or one-way analysis of variance with Bonferroni adjustment or Dunnett s post-test, as appropriate. Associations between groups of data were analyzed using Pearson correlation analysis. RESULTS Tck cells up-regulate expression of integrins and chemokine receptors. Stimulation of CD4 T cells in the presence of IL-6, TNF, and IL-2 up-regulated expression of a number of chemokine receptors and integrin molecules. In particular, cytokine stimulation resulted in a dramatic up-regulation of CXCR4 expression (9.5-fold; P ), as measured by flow cytometry (Table 1). In addition, we observed increased expression of CXCR2 (1.5-fold; P ), CXCR7 (1.9-fold; P ), CCR5 (2-fold; P ), CCR6 (2.5-fold; P ), and CCR8 (2.5-fold; P ), although to a lesser degree than that seen for CXCR4. Expression of the chemokine receptors CXCR3 and CCR7 was not significantly modulated by cytokine stimulation. Integrins are heterodimeric cell adhesion proteins consisting of an -chain and a -chain that are important in cell cell and cell extracellular matrix interactions. Analysis of integrin expression on cytokinestimulated CD4 T cells using flow cytometry revealed a significant up-regulation of expression of lymphocyte function associated antigen 1 (LFA-1), an integrin consisting of the L-chain CD11a and the 2-chain CD18. While both CD11a and CD18 were highly expressed on unstimulated CD4 T cells, cytokine stimulation increased the mean mean fluorescence intensity of both molecules (2.5-fold [P ] and 4.6-fold [P ], respectively). In addition, cytokine stimulation of CD4 T cells increased expression of VLA-4, an

4 2140 BRYANT ET AL Table 1. Expression of chemokine receptors and integrin chains on Tck cells* n MFI fold increase, mean SD P Chemokine receptor/ligand(s) CXCR4/CXCL12 (SDF-1) CXCR7/CXCL12 (SDF-1) CXCR2/CXCL1 (GRO ) CXCR3/CXCL10 (IP-10), CXCL11 (I-TAC) 16 NS CCR5/CCL5 (RANTES) CCR6/CCL20 (MIP-3 ) CCR7/CCL19 (ELC), CCL21 (SLC) 4 NS CCR8/CCL1 (I-309) Integrin/alternative nomenclature CD18/ 2 integrin CD11a/ L integrin CD29/ 1 integrin integrin CD49a/ 1 integrin CD49d/ 4 integrin CD49e/ 5 integrin NS * Lymphocytes were isolated from peripheral blood mononuclear cells by elutriation, followed by purification of the CD4 T cell population by positive immunomagnetic bead selection. CD4 T cells were cultured in the presence of tumor necrosis factor (25 ng/ml), interleukin-6 (IL-6) (100 ng/ml), and IL-2 (25 ng/ml) for 8 days. Cells were washed, stained, and analyzed by flow cytometry on the day of isolation (resting T cells) and after cytokine stimulation (cytokine-activated T [Tck] cells). Shown is the fold increase in mean fluorescence intensity (MFI) for each molecule, calculated with reference to resting T cell values. n number of donors tested or number of individual experiments; SDF-1 stromal cell derived factor 1; GRO growth-related oncogene ; IP-10 interferon- inducible 10-kd protein; I-TAC interferon-inducible T cell chemoattractant; NS not significant; MIP-3 macrophage inflammatory protein 3 ; ELC Epstein-Barr virus induced gene 1 ligand chemokine; SLC secondary lymphoid-tissue chemokine. By Wilcoxon s rank sum test. integrin consisting of the 4-chain CD49d (2.2-fold; P ) and the 1-chain CD29 (3.4-fold; P ). CD29 can also associate with -chains 1 9 to constitute other VLA molecules. Expression of the -chain CD49a was also increased in response to cytokine stimulation (2.7-fold; P ), while expression of CD49e was unaffected. In addition, expression of 7 integrin, also known to associate with 4 integrin chain CD49d, was up-regulated by cytokine stimulation (2.3-fold; P ). Increased expression of CXCR4 on Tck cells corresponds to increased chemotaxis in response to SDF-1. We used in vitro chemotaxis assays to determine if the observed increased expression of chemokine receptors on CD4 T cells after cytokine stimulation corresponded to an increased chemotactic response to ligands of their receptors. Interestingly, in the absence of any chemotactic gradient, Tck cells showed an increased ability to migrate in these assays (10% of total cells added) compared to donor-matched resting T cells or Ttcr cells ( 2% of total cells added) (P ) (Figure 1A). In response to increasing concentrations of SDF-1, Tck cells migrated to a higher level compared to donor-matched resting T cells in accordance with their increased expression of the receptor CXCR4 (Figure 1B). While resting T cells displayed a varied migratory response to SDF-1, Tck cells displayed a consistently increased chemotactic response. Tck cell chemotaxis to SDF-1 at concentrations of 3 and 30 ng/ml was significantly increased compared to background migration (P and P , respectively) (Figure 1C). Less chemotaxis was observed at higher concentrations of SDF-1, consistent with the characteristic bell-shaped chemotactic response. Pretreatment of Tck cells with a small-molecule inhibitor of CXCR4 abrogated migration to 20 ng/ml SDF-1 (P ), confirming that this effect was mediated through CXCR4 receptor expression on Tck cells (Figure 1D). However, blockade of CXCR4 in the absence of SDF-1 had no inhibitory effect on the background migration observed for Tck cells (Figure 1D). In light of the observed migration of Tck cells in the absence of a chemokine gradient (Figure 1A), we sought to ensure that the increased response to SDF-1 was truly chemotactic (directed movement) and not chemokinetic (random movement) in nature. Results of

5 MIGRATORY PROPERTIES OF Tck 2141 Figure 1. Stromal cell derived factor 1 (SDF-1) (CXCL12) is chemotactic for cytokine-activated T (Tck) cells and acts through CXCR4. Increasing concentrations of SDF-1 were added to the lower wells of a 96-well chemotaxis plate with 3- m pores. T cells ( ) were added to the top wells, and following a 2-hour incubation, migrated cells were removed from the lower wells and counted by flow cytometry. A, Shown is migration of T cell receptor stimulated T (Ttcr) cells, Tck cells, and resting T cells in response to media in the absence of chemokine (n 30 experiments). B, Shown is migration of Tck and resting T cells in response to increasing concentrations of SDF-1 (n 8 experiments). C, Shown is migration of Tck cells in response to increasing concentrations of SDF-1 (n 8 experiments). D, Tck cells were pretreated with 50 ng/ml of the CXCR4 antagonist AMD3100 for 30 minutes prior to addition to the top wells (n 5 experiments). Values in A and B are the mean SD percent migration of total cells added. Values in C and D are the mean SD percent chemotaxis. P 0.05; P 0.01; P NS not significant. a checkerboard assay investigating chemotaxis at different SDF-1 concentrations in the presence and absence of a chemokine gradient confirmed that Tck cells migrated to a positive gradient of SDF-1, but no migration above control levels was observed when this gradient was absent (Figure 2A). These results demonstrate that SDF-1 is indeed chemotactic and not chemokinetic for Tck. Despite the up-regulation of CCR5, CCR6, and, in some donors, CXCR3, no chemotaxis of Tck cells to their ligands RANTES, macrophage inflammatory protein 3, and interferon- inducible 10-kd protein was observed (data not shown). Increased expression of VLA-4 (CD49d/CD29) on Tck corresponds to increased migration in response to svcam-1. In light of previous studies documenting the ability of soluble integrin ligands to induce cell migration (20 22), coupled with our findings of increased VLA-4 expression on Tck cells, we investigated the migration of Tck in response to svcam-1 (Figure 3A). Results of in vitro chemotaxis assays revealed a dosedependent increase in Tck migration in response to svcam-1, while in contrast migration of resting T cells or Ttcr cells was minimal. Approximately 30% of Tck cells migrated in response to a concentration of 10 g/ml of svcam-1, significantly higher than that of Ttcr cells (P ) and Tck cells background migration (P ). Pretreatment of Tck cells with neutralizing antibodies against CD29 and CD49d abrogated this increased migration (P and P , respectively), confirming that this effect was mediated through VLA-4 expression on Tck cells (Figure 3B). However, VLA-4 blockade did not further inhibit the background levels of Tck cells migration in the absence of a chemoattractant gradient. The ability of svcam-1 to stimulate Tck cells migration was not shared by other integrin ligands such as soluble intercellular adhesion

6 2142 BRYANT ET AL molecule 1 (sicam-1) or sicam-2, despite upregulation of the counterreceptor LFA-1 on T cells in response to cytokine stimulation (Figure 3C). The nature of the Tck response to svcam-1 was investigated using checkerboard analysis (Figure 2B). Tck cell migration was observed in both the presence and the absence of a concentration gradient, indicating that the Tck cell response to svcam-1 is chemokinetic (random movement) in nature. Blockade of VLA-4 integrins (CD29/CD49d) selectively inhibits Tck cell transendothelial migration. The effect of increased integrin chain expression on Tck cells was further investigated using an in vitro transendothelial migration assay, involving the measurement of cell migration through a TNF -stimulated HUVEC monolayer at various time points. Tck cells demonstrated a significantly increased capacity to transmigrate Figure 2. Tck cell migratory response is chemotactic in response to SDF-1 and chemokinetic in response to soluble vascular cell adhesion molecule 1 (svcam-1). A and B, Checkerboard analysis of Tck cell migration in response to increasing concentrations of SDF-1 (A) or svcam-1 (B) in both the upper and lower wells in in vitro chemotaxis assays. Numbers in boxes are the percent migration in response to the various concentrations of SDF-1 or svcam-1. Shaded boxes are indicative of the same concentration in both wells and hence the absence of a chemokine gradient (data are representative of 3 independent experiments). See Figure 1 for other definitions. Figure 3. Soluble vascular cell adhesion molecule 1 (svcam-1) is a chemoattractant for Tck cells and acts through very late activation antigen 4. Increasing concentrations of svcam-1 were added to the lower wells of a 96-well chemotaxis plate with 3- m pores. T cells ( ) were added to the top wells, and following a 2-hour incubation, migrated cells were removed from the lower wells and counted by flow cytometry. A, Migration of Ttcr, Tck, and resting T cells in response to increasing concentrations of svcam-1 (n 5 experiments). B, Migration of Tck cells in response to 0 g/ml or 10 g/ml of svcam-1, following pretreatment of Tck cells with neutralizing antibodies against CD29 (10 g/ml) and CD49d (10 g/ ml) or with IgG1 control (10 g/ml) for 30 minutes prior to addition to the upper wells (n 6 experiments). C, Migration of Tck cells in response to 10 g/ml of soluble intercellular adhesion molecule 1 (sicam-1) or sicam-2. Values are the mean SD percent chemotaxis of total cells added. P 0.05; P 0.01; P 0.001, by paired t-test. See Figure 1 for other definitions.

7 MIGRATORY PROPERTIES OF Tck 2143 through an endothelial monolayer compared to resting T cells and Ttcr (P 0.01). Approximately 30% of Tck cells had migrated at the 24-hour time point compared to 7% of Ttcr cells and 2% of resting T cells (Figure 4A). We investigated the contribution of VLA-4 to Tck cell transmigration in this system using neutralizing antibodies against CD29 and CD49d. Pretreatment of Tck cells with neutralizing antibodies against CD29 inhibited migration by 70% (P ) (Figure 4B), while neutralizing antibodies against CD49d inhibited migration by 50% (P ) (Figure 4C). Coblockade of both molecules simultaneously did not result in further inhibition compared to blockade of each molecule alone (data not shown). Figure 5. Rheumatoid arthritis (RA) fibroblast-like synoviocytes (FLS) play a role in inducing migration of cytokine-activated T (Tck) cells. RA FLS were cultured overnight in the lower wells of 96-well chemotaxis plates with 3- m pores. Tck cells were stained with carboxyfluorescein succinimidyl ester and added to the upper wells. After a 2-hour incubation, migrated cells were collected and counted using flow cytometry. A, Shown is migration of Tck cells in response to media and RA FLS (n 12 experiments). P by Wilcoxon s rank sum test. B, Supernatants from cultured FLS were analyzed for levels of soluble vascular cell adhesion molecule 1 (svcam-1) by enzyme-linked immunosorbent assay (n 14 experiments). Values are the mean SD. C, Tck cells and FLS were pretreated with neutralizing antibodies against very late activation antigen 4 (CD29) (10 g/ml) and VCAM-1 (10 g/ml), respectively, or with IgG control (10 g/ml) for 30 minutes prior to addition to the upper wells (n 4 experiments). D, Tck cells were pretreated with 50 ng/ml of the CXCR4 antagonist AMD3100 for 30 minutes prior to addition to the upper wells (n 4 experiments). P 0.05 by paired t-test. Values in A, C, and D are the percent migration. Figure 4. Tck cells demonstrate increased transmigration, which is mediated through very late activation antigen 4. Human umbilical vein endothelial cells were stimulated for 18 hours with 10 ng/ml tumor necrosis factor prior to culture on Transwell inserts containing 3- m pores. T cells ( ) were added to the upper well and incubated for 24 hours. Migrated cells were removed from the bottom well and counted at the indicated time points. A, Shown is the percent transmigration of resting T cells, Ttcr cells, and Tck cells at the indicated time points. Values are the mean SD (n 4 experiments). P 0.01 versus resting T cells and Ttcr cells, by analysis of variance with Bonferroni adjustment. B and C, Tck cells were pretreated with neutralizing antibodies against CD29 (10 g/ml) (B) and CD49d (10 g/ml) (C) or with IgG1 control (10 g/ml) for 30 minutes prior to addition to the upper wells. Values are the percent of cells migrated at 24 hours (n 8 experiments). P See Figure 1 for definitions. RA FLS play a role in inducing migration of Tck cells. RA pathology has previously been shown to be perpetuated by the continuous immigration of T cells and other immune cells into the joint. Synovial fibroblasts have been reported to be involved in the chemokine production that attracts these cells (26 29). To investigate if Tck cells are attracted to RA FLS in vitro, we used modified chemotaxis assays of Tck migration to cultured RA FLS. The presence of RA FLS resulted in a modest yet significant increase of background Tck cell migration in the majority of donors (P ) (Figure 5A). ELISA analysis of FLS supernatants revealed svcam-1 production, with a mean concentration of 620 pg/ml (range

8 2144 BRYANT ET AL 19 1,400 pg/ml) (Figure 5B). To investigate if this svcam-1 was responsible for the increased Tck cell migration, Tck cells and FLS were pretreated with neutralizing antibodies against VLA-4 (CD29/CD49d) and VCAM-1, respectively. Migration was inhibited by up to 62% compared to the isotype control; however, this varied between donors (Figure 5C). Since RA FLS have previously been reported to produce SDF-1, we investigated whether CXCR4 inhibition could abrogate Tck cell migration. Pretreatment of Tck cells with AMD3100 could indeed inhibit Tck cell migration to FLS by up to 67% (P ); again, however, this differed between donors (Figure 5D). DISCUSSION The dysregulated production of inflammatory cytokines, particularly TNF, is acknowledged to play a key role in the pathology of RA. Evidence from our group and others has shown that synovial T cells play a crucial role in the induction of macrophage-derived TNF production through a cell contact dependent mechanism (15,16,30). We previously demonstrated that cytokine activation of peripheral blood T cells induced a phenotype that mimicked T cells isolated from synovial joint tissue in both effector function and cell surface expression (16,17). In a previous study, we noted the up-regulation of a number of adhesion molecules and chemokine receptors on Tck cells (17), and in the present study, we extend these findings and demonstrate that Tck cells exhibit increased migration in response to SDF-1 and svcam-1 due to increased expression of CXCR4 and VLA-4. This may have implications for the trafficking and retention of T cells to the inflamed synovium in RA. We analyzed the expression of chemokine receptors on Tck cells, which revealed a significant upregulation of CXCR4. Tck cells also showed a dosedependent migratory response to SDF-1, the ligand for CXCR4. Using a small-molecule inhibitor of CXCR4, we confirmed that the response to SDF-1 was CXCR4 dependent, as SDF-1 has recently been described as a ligand for a second chemokine receptor, CXCR7. Increased CXCR4 expression has previously been identified on RA synovial T cells (19). In addition, SDF-1 is expressed in the inflamed synovium and may play a role in the trafficking of T cells to this site in RA (19,27). Integrins and adhesion molecules are important in cell cell and cell extracellular matrix interactions and play a key role in cell trafficking. Analysis of adhesion molecule expression on Tck cells revealed up-regulation of VLA-4 (CD29/CD49d) and LFA-1 (CD11a/CD18) in response to cytokine stimulation. RA synovial T cells also express high levels of both CD29 and CD49d and show enhanced ability to bind the VLA-4 ligand fibronectin (31). In addition, VCAM-1, the principal ligand for VLA-4, is expressed selectively on RA endothelium and macrophages in the lining layer of the synovium, and its expression is decreased following anti-tnf therapy (32). A number of previous studies have shown that svcam-1, a ligand for VLA-4, can stimulate migratory responses in different cell types, including eosinophils, monocytes, and RA synovial T cells (20 22). We showed that VCAM-1 dose-dependently induced Tck cell migration. This migratory response was chemokinetic as opposed to chemotactic in nature and was mediated through VLA-4 expression on Tck cells, highlighted by the ability of neutralizing antibodies to CD29 and CD49d to inhibit the effect. Induction of VLA-4 expression on Tck cells also resulted in an increased ability to migrate through an endothelial cell monolayer, an in vitro model of extravasation. Taken together, these results suggest that SDF-1 and svcam-1 may play a role in the recruitment and retention of T cells to the joint in RA. In our in vitro system, T cells acquire an increased chemotactic response to SDF-1 and VLA-4 after an 8-day activation in the presence of TNF, IL-6, and IL-2 cytokines whose levels are increased in the rheumatoid joint. This would initially suggest that T cells acquire this phenotype after they have entered the synovium; therefore, what is the relevance of the increased ability of the cells to respond to SDF-1 and VLA-4 if they are already outside the vasculature within the synovium, which seemingly obviates their need to respond to a chemotactic signal to enter the joint? We believe that the increased responsiveness of these cells to SDF-1 and VLA-4 will play a major role in retention of these cells within the synovium, where they subsequently contribute to cell contact dependent induction of proinflammatory cytokine production from monocyte/macrophages and fibroblasts. The question remains, however, of whether SDF-1 and VLA-4 also play a role in the initial recruitment of these cells to the synovium in RA. It has previously been demonstrated that increased levels of both circulating SDF-1 and VLA-4 are present in the serum of RA patients, and levels are also increased in the synovium. This suggests the possibility that these molecules may also play a role in the initial recruitment of these cells to the synovium as well as in their retention at the site.

9 MIGRATORY PROPERTIES OF Tck 2145 To address this question, we previously attempted to quantify Tck cell numbers in the peripheral blood of RA patients. However, as Tck cells represent a general cell phenotype rather than a distinct lineage, such studies are difficult due to the lack of specific markers for these cells; nevertheless, we did find an increased expression of CXCR4 on CD4 T cells from the peripheral blood of RA patients as compared to healthy controls (data not shown). Increased SDF-1 production by the synovium in RA patients may therefore recruit these cells to the synovium in increased numbers, where they play a role in the induction of proinflammatory cytokine production contributing to the pathogenesis of RA. It is acknowledged that synovial fibroblasts are among the principal cell types producing chemokines in RA synovial tissue. Of note, RA fibroblasts produce both SDF-1 and VCAM-1. We show that Tck cells are attracted to RA fibroblasts in vitro, and this is dependent on CXCR4 and VLA-4 expression. A previous study has documented similar findings for RA synovial T cells and B cells. In that study, RA synovial T cells showed an increased ability to migrate beneath fibroblasts isolated from RA tissue, and this effect was due to SDF-1 and VCAM-1 production (33). Compared to resting T cells or Ttcr cells, Tck cells demonstrated an increased level of background, or basal, migration in the absence of chemoattractant in in vitro chemotaxis assays (Figure 1A). Increased basal migration has previously been reported for RA synovial T cells compared to RA peripheral blood T cells (20). This migration was not abrogated by blocking CXCR4 or VLA-4 and suggests that T cells activated by cytokines develop an increased level of basal migration, another phenotypic trait they share with T cells from RA synovial tissue. In conclusion, we have expanded our previous studies on Tck cells. We have previously shown that cytokine activation induced a phenotype and effector function that mimicked RA T cells (17). In the present study, we demonstrate that cytokine activation up-regulates expression of CXCR4 and VLA-4 on T cells, resulting in increased migratory responses to their ligands SDF-1 and VCAM-1. This increased responsiveness to SDF-1 and svcam-1 may play a role in trafficking and retention of T cells to the inflamed synovium in RA. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ahern had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Ahern, Brennan. Acquisition of data. Bryant, Brennan. Analysis and interpretation of data. Bryant, Ahern, Brennan. REFERENCES 1. Cush JJ, Lipsky PE. Phenotypic analysis of synovial tissue and peripheral blood lymphocytes isolated from patients with rheumatoid arthritis. Arthritis Rheum 1988;31: Kinne RW, Stuhlmuller B, Burmester GR. Cells of the synovium in rheumatoid arthritis: macrophages. Arthritis Res Ther 2007;9: Lundy SK, Sarkar S, Tesmer LA, Fox DA. Cells of the synovium in rheumatoid arthritis: T lymphocytes. Arthritis Res Ther 2007; 9: Gregersen PK, Silver J, Winchester RJ. The shared epitope hypothesis: an approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 1987;30: Klareskog L, Ronnelid J, Lundberg K, Padyukov L, Alfredsson L. Immunity to citrullinated proteins in rheumatoid arthritis. Annu Rev Immunol 2008;26: Lundstrom E, Kallberg H, Alfredsson L, Klareskog L, Padyukov L. Gene environment interaction between the DRB1 shared epitope and smoking in the risk of anti citrullinated protein antibody positive rheumatoid arthritis: all alleles are important. Arthritis Rheum 2009;60: Borthwick NJ, Akbar AN, MacCormac LP, Lowdell M, Craigen JL, Hassan I, et al. Selective migration of highly differentiated primed T cells, defined by low expression of CD45RB, across human umbilical vein endothelial cells: effects of viral infection on transmigration. Immunology 1997;90: Ponchel F, Morgan AW, Bingham SJ, Quinn M, Buch M, Verburg RJ, et al. Dysregulated lymphocyte proliferation and differentiation in patients with rheumatoid arthritis. Blood 2002;100: Goronzy JJ, Henel G, Sawai H, Singh K, Lee EB, Pryshchep S, et al. Costimulatory pathways in rheumatoid synovitis and T-cell senescence. Ann N Y Acad Sci 2005;1062: Van Dinther-Janssen AC, Horst E, Koopman G, Newmann W, Scheper RJ, Meijer CJ, et al. The VLA-4/VCAM-1 pathway is involved in lymphocyte adhesion to endothelium in rheumatoid synovium. J Immunol 1991;147: Rodriguez RM, Pitzalis C, Kingsley GH, Henderson E, Humphries MJ, Panayi GS. T lymphocyte adhesion to fibronectin (FN): a possible mechanism for T cell accumulation in the rheumatoid joint. Clin Exp Immunol 1992;89: Oppenheimer-Marks N, Lipsky PE. Adhesion molecules in rheumatoid arthritis. Springer Semin Immunopathol 1998;20: Unutmaz D, Pileri P, Abrignani S. Antigen-independent activation of naive and memory resting T cells by a cytokine combination. J Exp Med 1994;180: Vey E, Burger D, Dayer JM. Expression and cleavage of tumor necrosis factor- and tumor necrosis factor receptors by human monocytic cell lines upon direct contact with stimulated T cells. Eur J Immunol 1996;26: McInnes IB, Leung BP, Sturrock RD, Field M, Liew FY. Interleukin-15 mediates T cell-dependent regulation of tumor necrosis factor- production in rheumatoid arthritis. Nat Med 1997;3: Brennan FM, Hayes AL, Ciesielski CJ, Green P, Foxwell BM, Feldmann M. Evidence that rheumatoid arthritis synovial T cells

10 2146 BRYANT ET AL are similar to cytokine-activated T cells: involvement of phosphatidylinositol 3-kinase and nuclear factor B pathways in tumor necrosis factor production in rheumatoid arthritis. Arthritis Rheum 2002;46: Brennan FM, Smith NM, Owen S, Li C, Amjadi P, Green P, et al. Resting CD4 effector memory T cells are precursors of bystander-activated effectors: a surrogate model of rheumatoid arthritis synovial T-cell function. Arthritis Res Ther 2008;10:R Buckley CD. Michael Mason prize essay 2003: why do leucocytes accumulate within chronically inflamed joints? Rheumatology (Oxford) 2003;42: Nanki T, Hayashida K, El-Gabalawy HS, Suson S, Shi K, Girschick HJ, et al. Stromal cell-derived factor-1-cxc chemokine receptor 4 interactions play a central role in CD4 T cell accumulation in rheumatoid arthritis synovium. J Immunol 2000;165: Kitani A, Nakashima N, Izumihara T, Inagaki M, Baoui X, Yu S, et al. Soluble VCAM-1 induces chemotaxis of Jurkat and synovial fluid T cells bearing high affinity very late antigen-4. J Immunol 1998;161: Ueki S, Kihara J, Kato H, Ito W, Takeda M, Kobayashi Y, et al. Soluble vascular cell adhesion molecule-1 induces human eosinophil migration. Allergy 2009;64: Tokuhira M, Hosaka S, Volin MV, Haines GK III, Katschke KJ Jr, Kim S, et al. Soluble vascular cell adhesion molecule 1 mediation of monocyte chemotaxis in rheumatoid arthritis. Arthritis Rheum 2000;43: Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31: Beech JT, Brennan FM. Preparation of mononuclear cells from synovial tissue. Methods Mol Med 2007;135: Malik NM, Jin P, Raatz Y, Sumariwalla PF, Kiriakidis S, Shepard M, et al. Regulation of the angiopoietin Tie ligand receptor system with a novel splice variant of Tie1 reduces the severity of murine arthritis. Rheumatology (Oxford) 2010;49: Matsuyama T, Kitani A. The role of VCAM-1 molecule in the pathogenesis of rheumatoid synovitis. Hum Cell 1996;9: Kim KW, Cho ML, Kim HR, Ju JH, Park MK, Oh HJ, et al. Up-regulation of stromal cell derived factor 1 (CXCL12) production in rheumatoid synovial fibroblasts through interactions with T lymphocytes: role of interleukin-17 and CD40L CD40 interaction. Arthritis Rheum 2007;56: McGettrick HM, Smith E, Filer A, Kissane S, Salmon M, Buckley CD, et al. Fibroblasts from different sites may promote or inhibit recruitment of flowing lymphocytes by endothelial cells. Eur J Immunol 2009;39: Silverman MD, Haas CS, Rad AM, Arbab AS, Koch AE. The role of vascular cell adhesion molecule 1/very late activation antigen 4 in endothelial progenitor cell recruitment to rheumatoid arthritis synovium. Arthritis Rheum 2007;56: Dayer JM, Burger D. Cell-cell interactions and tissue damage in rheumatoid arthritis. Autoimmun Rev 2004;3 Suppl 1:S Laffon A, Garcia-Vicuna R, Humbria A, Postigo AA, Corbi AL, de Landazuri MO, et al. Upregulated expression and function of VLA-4 fibronectin receptors on human activated T cells in rheumatoid arthritis. J Clin Invest 1991;88: Tak PP, Taylor PC, Breedveld FC, Smeets TJ, Daha MR, Kluin PM, et al. Decrease in cellularity and expression of adhesion molecules by anti tumor necrosis factor monoclonal antibody treatment in patients with rheumatoid arthritis. Arthritis Rheum 1996;39: Bradfield PF, Amft N, Vernon-Wilson E, Exley AE, Parsonage G, Rainger GE, et al. Rheumatoid fibroblast-like synoviocytes overexpress the chemokine stromal cell derived factor 1 (CXCL12), which supports distinct patterns and rates of CD4 and CD8 T cell migration within synovial tissue. Arthritis Rheum 2003;48: