1. Top of page
  2. Abstract
  6. Acknowledgements


Directional migration of leukocytes is orchestrated by the regulated expression of chemokine receptors and their ligands. The receptor CXCR6 is abundantly expressed by Th1-polarized effector/memory lymphocytes accumulating at inflammatory sites. This study was undertaken to examine the presence of CXCR6+ T cells and of CXCL16, the only ligand for CXCR6, in the joints of patients with rheumatoid arthritis (RA).


Flow cytometry analysis of the expression of CXCR6 by peripheral blood and synovial fluid (SF) T cells. In addition, by performing conventional and real-time reverse transcriptase–polymerase chain reaction, immunohistochemistry, and enzyme-linked immunosorbent assay, we determined the expression of CXCL16 and its protease ADAM-10 within synovium and by cultured macrophages. SF T cell migration was studied with the Transwell system.


Accumulation of CXCR6+ T cells within RA SF coincided with highly elevated levels of CXCL16+ macrophages. In vitro studies revealed that monocytes started to express CXCL16 upon differentiation into macrophages, and that RA SF and tumor necrosis factor (TNF) enhanced CXCL16 expression. Moreover, RA patients responding to anti-TNF therapy showed a strongly decreased CXCL16 expression, whereas nonresponding patients did not. Interestingly, ADAM-10, a recently identified protease of CXCL16, was abundantly expressed by CXCL16+ macrophages in vitro and in RA in vivo, which resulted in increased levels of cleaved CXCL16 in RA SF relative to controls. Finally, CXCR6+ T cells from RA SF were attracted by CXCL16.


These data provide evidence that enhanced production of CXCL16 in RA synovia leads to recruitment of CXCR6+ memory T cells, thereby contributing to the inflammatory cascade associated with RA pathology.

Rheumatoid arthritis (RA) is a systemic autoimmune disease which is characterized by a chronic inflammation of multiple synovial joints. Large numbers of leukocytes infiltrate and accumulate within the synovial tissue (ST) and synovial fluid (SF) (1–3). These leukocytes include T cells, especially CXCR6+ memory T cells, monocytes, plasma cells, and granulocytes. While in most patients these cells are dispersed throughout the synovium, in other patients highly organized lymphoid structures resembling germinal centers can be found (4). Although the cause of RA is still unknown, the recruitment and cytokine-induced activation of inflammatory cells is thought to be essential in perpetuation of the inflammatory response and, ultimately, in cartilage and bone destruction (5–7).

The trafficking of leukocytes is regulated through selective expression of an array of chemokines, adhesion molecules, and their corresponding receptors. Chemokines are secreted proteins that attract leukocytes via activation of 7-transmembrane–domain G-protein–coupled receptors (8, 9). Adhesion molecules provide adhesive capacity during cell–extracellular matrix or cell–cell contact, e.g., when leukocytes transmigrate the endothelium (10, 11). In this respect CXCL16 is an exceptional chemokine, because it has the potential to function as a chemoattractant and as an adhesion molecule. While classic chemokines are expressed as small soluble proteins, CXCL16 is first synthesized as a transmembrane protein expressed by macrophages, dendritic cells (DCs), and endothelial cells (12–14). Data from Shimaoka et al (15) have recently suggested that cell surface–expressed CXCL16 can indeed function as an adhesion molecule. However, upon cleavage by proteases, the extracellular domain is released as a soluble chemokine that attracts effector/memory T cells that express CXCR6, the receptor for CXCL16 (12, 14). Furthermore, CXCL16 also acts as a scavenger receptor for oxidized low-density lipoprotein and bacteria (13, 16, 17), confirming that CXCL16 is a multifunctional protein. Concerning structure and mechanism of action, CXCL16 resembles fractalkine, the other transmembrane chemokine. Fractalkine has been shown to mediate adhesion in its transmembrane form, and to mediate chemotaxis as a cleaved protein (18–20).

Kim and colleagues recently reported the accumulation of CXCR6+ T cells in SF of a small number of RA patients (21). As yet, however, nothing is known about the expression of CXCL16, the only known ligand for CXCR6, in RA joints. Therefore, we analyzed the expression of CXCL16, its recently characterized protease ADAM-10, and CXCR6 in vitro and in vivo, within healthy joints and in the joints of RA patients. Our data demonstrate that expression of CXCL16 and ADAM-10 is strongly enhanced in RA synovia, resulting in the recruitment and accumulation of CXCR6+ memory T cells in RA joints. These data imply an important role for CXCL16/CXCR6 in the synovial inflammation that is strongly associated with RA pathogenesis.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Patients and samples.

Synovial tissue, synovial fluid, and peripheral blood were obtained from a total of 43 RA patients attending either the Department of Rheumatology or the outpatient clinics of the University Medical Center (UMC) Nijmegen. All patients fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) criteria for RA (22) and gave informed consent for the study. Disease activity was assessed using the Disease Activity Score in 28 joints (DAS28) (23). For the present study, ST obtained from patients scored as having very active RA (DAS28 > 5.1) (n = 13) and was compared with ST from healthy individuals (n = 5). In addition, SF was obtained from 17 RA patients with active disease, 5 osteoarthritis (OA) patients, and 2 controls with trauma injury. Therapeutic regimens of all RA patients were recorded before blood sampling. Patients receiving either prednisolone or biologic therapies, such as anti–tumor necrosis factor (anti-TNF) or interleukin-1 (IL-1) receptor antagonist, within 6 weeks prior to the study were not included in the current analysis.

For immunohistochemical analysis, synovial biopsy samples from RA patients (n = 13) were obtained using small-needle arthroscopy. An average of 20 biopsy samples were obtained from the medial and lateral suprapatellar pouch on each occasion. For comparison, ST from healthy controls (n = 5) was obtained during arthroscopic procedures performed by the orthopedic surgeons. During these procedures, SF was isolated and collected when possible. In addition, we isolated peripheral blood mononuclear cells (PBMCs) from 18 healthy individuals. To analyze the effect of TNF blocking therapy, ST was isolated from RA patients (n = 6) before treatment and 6 weeks after treatment with the human anti-TNF monoclonal antibody (adalimumab; 40 mg subcutaneously every other week). This study was approved by the Ethics Committee of the UMC Nijmegen.

Isolation of synovial fluid mononuclear cells.

Synovial fluid mononuclear cells (SFMCs) from resuspended RA SF were obtained by density-gradient centrifugation over Lymphoprep (Axis-Shield, Oslo, Norway). SFMCs in the interphase were collected, washed extensively with citrated phosphate buffered saline (PBS), and used immediately for fluorescence-activated cell sorting (FACS) analyses and/or migration assays.

Generation of monocyte-derived macrophages.

To generate macrophages, PBMCs were isolated from buffy coats by density-gradient centrifugation over Lymphoprep (Axis-Shield). PBMCs in the interphase were collected, washed extensively with citrated PBS, and allowed to adhere to plastic for 1 hour. Next, the peripheral blood lymphocytes (PBLs) were washed away and the adhering monocytes were differentiated into macrophages by culturing in RPMI 1640 (“Dutch modification”) supplemented with L-glutamine and antibiotic–antimycotic (both from Invitrogen, Breda, The Netherlands) plus 5% human serum (Sigma, Steinheim, Germany) for up to 10 days. Culture medium was refreshed every 3 days. Analysis by flow cytometry demonstrated that these macrophages expressed high levels of CD14 and CD11c, intermediate levels of class II major histocompatibility complex and CD86, and no CD80 and CD209/DC-SIGN (data not shown). During some experiments, either freshly isolated monocytes or day 3 macrophages were cultured in the presence of RA SF or recombinant human TNF (PeproTech, London, UK) for 2 days.

RNA isolation, reverse transcriptase–polymerase chain reaction (RT-PCR), and real-time PCR.

Total RNA was isolated from synovial biopsy specimens or tonsil tissue sections using TRIzol according to the instructions of the manufacturer (Invitrogen). After treatment with DNase I (Roche, Almere, The Netherlands), first-strand complementary DNA (cDNA) synthesis was performed by a standard reverse transcription reaction, using Moloney murine leukemia virus reverse transcriptase (Invitrogen) and pd(N)6 random hexamers (Amersham Biosciences, Roosendaal, The Netherlands). Synthesis was performed at 20°C for 10 minutes, 42°C for 45 minutes, and 95°C for 10 minutes, followed by cool-down at 4°C. As a negative control, the reaction was also performed in the absence of reverse transcriptase. PCR was performed with AmpliTaq DNA Polymerase (Applied Biosystems, Niewerkerk a/d IJssel, The Netherlands), 100 ng primer, 200 μM dNTPs (both from Amersham Biosciences), and 1.5 mM MgCl2 in PCR Buffer (both from Applied Biosystems). The primers used were as follows: for CXCL16 amplification, 5′-CCCGCCATCGGTTCAGTTCA-3′ (forward) and 5′-GTGGACTGCAAGGTGGACAG-3′ (reverse); for ADAM-10, 5′-CGGAACACGAGAAGCTGTGA-3′ (forward) and 5′-AAGTCTGTGGTCTGGTAAATTGTATCA-3′ (reverse); and for actin, 5′-GCTACGAGCTGCCTGACGG-3′ (forward) and 5′-GAGGCCAGGATGGAGCC-3′ (reverse). PCR was started with a 5-minute denaturation step at 95°C, after which amplification was performed with 30 cycles of denaturation at 95°C for 30 seconds, annealing at 58°C for 30 seconds, and elongation at 72°C for 30 seconds. After a final elongation step of 10 minutes at 72°C, samples were cooled to 4°C and analyzed by electrophoresis in a 2% agarose gel containing ethidium bromide.

Quantitative real-time PCR was performed essentially as described previously (24). Briefly, amplifications were performed with SYBR Green Master Mix on an ABI/PRISM 7000 sequence detector system (both from Applied Biosystems). Quantification of the PCR signals was performed by comparing the cycle threshold (CT) value, in duplicate, of the gene of interest of each sample with the CT values of the reference housekeeping gene GAPDH. Furthermore, the housekeeping gene porphobilinogen deaminase (PBGD) was used as an internal control for the amount of cDNA in every individual. The primers (all from Eurogentec, Maastricht, The Netherlands) used for real-time PCR were 5′-CTTCATTTTTTGCTGATGGTTCC-3′ (forward) and 5′-GTCCCAGCACGGCACCT-3′ (reverse) for CXCL16, 5′-GAAGGTGAAGGTCGGAGT-3′ (forward) and 5′-AGATGGTGATGGGATTTC-3′ (reverse) for GAPDH, and 5′-GGCAATGCGGCTGCAA-3′ (forward) and 5′-GGGTACCCACGCGAATCAC-3′ (reverse) for PBGD.


In addition to the isotype controls mIgG1, mIgG2a, mIgG2b (all from BD Biosciences, Alphen aan den Rijn, The Netherlands), rbIgG (Jackson ImmunoResearch Laboratories, West Grove, PA), and gIgG (R&D Systems, Abingdon, UK), the following mouse antibodies were used (clone name given in parentheses): anti-human CXCR6 (56811.111; R&D Systems), phycoerythrin (PE)–conjugated anti-human CD3 (HIT3a; BD Biosciences), anti-human CD14 (RM052; Beckman Coulter, Mijdrecht, The Netherlands), PE-conjugated anti-human CD14 (RM052; Beckman Coulter), anti-CD208/DC-LAMP (104.G4; Beckman Coulter), anti-human CD31 (JC70A; Dako Cytomation, Glostrup, Denmark), anti-human CD45RO (UCHL-1; BD Biosciences), and anti-human CD68 (EBM11; Dako Cytomation). In addition, we used goat anti-human CXCL16 (R&D Systems), biotinylated rabbit anti-human CXCL16 (PeproTech), rabbit anti-human ADAM-10 (Serotec, Oxford, UK), biotinylated horse anti-mouse IgG, biotinylated rabbit anti-rat IgG, biotinylated horse anti-goat IgG, biotinylated goat anti-rabbit IgG (all from Vector, Burlingame, CA), fluorescein isothiocyanate–conjugated goat anti-mouse (Zymed, South San Francisco, CA), PE-conjugated goat anti-rabbit IgG (Caltag, Burlingame, CA), and AlexaFluor 647–conjugated donkey anti-goat and AlexaFluor 647–conjugated goat anti-mouse IgG2b (both from Molecular Probes, Leiden, The Netherlands).


For immunohistochemistry, frozen ST was cut into 7-μm sections, mounted on slides (Superfrost; Fisher Scientific, Pittsburgh, PA), air-dried, and stored at −80°C. Before staining, cryosections were air-dried, fixed in cold acetone for 10 minutes, air-dried again, and washed with PBS. Endogenous peroxidase was blocked with 1% H2O2 plus 0.2% NaN3 in PBS at room temperature for 10 minutes. After washing with 0.5% bovine serum albumin (BSA) plus 0.01% NaN3 in PBS, sections were stained with primary antibodies at 37°C for 1 hour, followed by incubation with biotin-conjugated secondary antibodies at room temperature for 30 minutes. Next, the samples were incubated with avidin–biotin–horseradish peroxidase (HRP) complex (Vector) at room temperature for 45 minutes. Color was then developed with aminoethylcarbazole (Zymed). Sections were counterstained with hematoxylin and mounted in Kaiser's glycerin–gelatin solution (Merck, Darmstadt, Germany). ST sections were analyzed with a DM LB microscope (Leica Instruments, Wetzlar, Germany) and photographed using a DC300 camera and Twain Driver software (IM500; both from Leica). All immunohistochemical stainings were accompanied by appropriate isotype-matched controls. For each antibody staining, at least 2 tissue sections per patient were microscopically examined by 2 independent observers who were unaware of the patient's identity and clinical status.

Flow cytometry.

Staining of cell surface proteins was essentially performed as described previously (25). Briefly, cells were incubated with the primary antibody at 4°C for 30 minutes. After washing, the cells were stained with fluorescence-labeled secondary antibody at 4°C for 30 minutes. Propidium iodide was added to exclude dead cells. Flow cytometric analyses were performed on either a FACSCalibur or a FACScan with CellQuest software (all from BD Biosciences).

CXCL16 sandwich enzyme-linked immunosorbent assay (ELISA).

For the detection of soluble CXCL16 in serum, SF, or culture supernatant, a sandwich ELISA was set up. Maxisorp ELISA plates (Nunc, Roskilde, Denmark) were coated overnight with 50 μl/well of 1 μg/ml goat anti-human CXCL16 (R&D Systems) in PBS at 4°C. Next, the plates were washed 3 times with PBS and blocked with 100 μl of 1% BSA in PBS at 37°C for 1 hour. After washing 3 times with PBS containing 0.05% Tween 20 (ELISA buffer), the plates were incubated with serial dilutions of the samples (50 μl/well) at 37°C for 1 hour. Serial dilutions of recombinant human CXCL16 (R&D Systems) were used to obtain a standard curve. Samples and recombinant protein were diluted in 1% BSA in PBS. After washing 3 times with ELISA buffer, the plates were incubated with 50 μl/well of 0.5 μg/ml biotinylated rabbit anti-human CXCL16 (PeproTech) in ELISA buffer at room temperature for 30 minutes. Next, the plates were washed 3 times with ELISA buffer and incubated with 50 μl/well of HRP-conjugated avidin–biotin complex (Vector) in ELISA buffer at room temperature for 30 minutes. After washing 3 times with ELISA buffer and 1 time with PBS, the presence of HRP was detected using 100 μl/well of 100 μg/ml 3,3′,5,5′-tetramethylbenzidine (Sigma) in DMSO (final percentage 1%) diluted in 100 mM NaAc buffer (pH 4.5). The reaction was stopped with 100 μl/well 800 mM H2SO4. Absorbance was measured at 450 nm using a 3550-UV Microplate Reader (Bio-Rad, Hercules, CA). The detection limit of this ELISA is ∼100 pg/ml.

Migration assays.

Migratory responses of SF T cells were evaluated by using Transwell polycarbonate inserts (6.5-mm diameter) with 5-μm pores (Costar Corning, Cambridge, MA). SFMCs were resuspended in RPMI 1640 containing 0.5% BSA (migration medium) and injected (250 × 103/100 μl) in the upper compartment of the Transwell. Serial dilutions of recombinant human CXCL16 were made in migration medium and added to the lower compartment (600-μl Transwell). SFMCs were allowed to migrate at 37°C in air with 5% CO2 for 90 minutes. Next, the inserts were discarded and 20% of the SFMCs that had migrated were counted using a flow cytometer as described previously (26). The remaining SFMCs were stained for CD3 to determine the percentage of T cells by flow cytometric analysis.

Statistical analysis.

For statistical analyses, we first logarithmically transformed the values from groups with a skewed distribution. Next, differences between groups were calculated by using the 2-sample t-test. P values less than 0.05 were considered significant.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Increased numbers of CXCR6+ memory T cells in RA SF.

Recent data demonstrated that the chemokine receptor CXCR6 is preferentially expressed by T effector/memory cells involved in Th1 responses and that CXCR6+ T cells are enriched at sites of inflammation (21). Therefore, we analyzed CXCR6 expression on T cells isolated from blood and SF of RA patients and healthy individuals. As shown in Figures 1A and B, both control PBMCs (mean 8%) and RA PBMCs (mean 5%) contained only small numbers of CXCR6+ T cells. In contrast, up to 80% of the T cells within RA SF expressed CXCR6 (mean 32%), whereas few or no leukocytes could be detected in SF of controls (Figures 1A and B, and data not shown). RT-PCR analysis further confirmed that CXCR6 messenger RNA (mRNA) was indeed expressed by RA SF cells (results not shown). In 2 patients analyzed, the CXCR6+ T cells displayed a memory phenotype (CD3+,CD45RO+), with ∼52% of them as CD4+ and 48% as CD8+ (results not shown). These results thus confirm and extend the data reported by Kim et al (21), and demonstrate that high numbers of CXCR6+ T cells specifically accumulate in SF of RA patients.

thumbnail image

Figure 1. Expression of CXCR6 on T cells from patients with rheumatoid arthritis (RA). A, Fluorescence-activated cell sorting results. While a small percentage of peripheral blood T cells from RA patients (RA blood) express CXCR6, this chemokine receptor is widely expressed by RA synovial fluid (SF) T cells. T cells were stained for CXCR6 and CD3. The quadrants indicate the expression above the background. Dead cells were excluded by gating on propidium iodide–negative cells. Representative examples are shown. B, Percentage of CXCR6+ T cells in control blood (C blood) (n = 10), RA blood (n = 4), and RA SF (n = 6). Values are the mean and SD. ∗∗ = P < 0.01. C, Expression of CXCL16 mRNA in control (C) synovial tissue and in synovia from RA patients as determined by reverse transcriptase–polymerase chain reaction (PCR) using specific primers. Tonsil mRNA (T) and water (W) were used as positive and negative controls, respectively. Results shown are representative of control synovia and RA synovia. D, Quantification of CXCL16 mRNA demonstrating significantly enhanced expression in RA synovia (n = 6) compared with control synovia (n = 4). Expression was determined by quantitative real-time PCR. CXCL16 levels were normalized to the levels of the housekeeping gene GAPDH. ∗ = P < 0.05 versus controls.

Download figure to PowerPoint

Increased CXCL16 expression by RA synovial macrophages and endothelial cells.

As yet nothing is known about the expression of CXCL16 (the only known ligand for CXCR6) in ST or SF. Therefore, we first determined the expression of CXCL16 mRNA in RA ST or control ST by RT-PCR. RNA from a tonsil was included as a positive control. As shown in Figure 1C, a CXCL16 PCR product of the expected size was readily detected both in control synovia (n = 3) and in RA synovia (n = 7), as well as in the tonsil sample. Subsequent quantitative analysis demonstrated that CXCL16 RNA levels were significantly increased in RA synovia (n = 6) compared with control ST (n = 4) (Figure 1D).

Next, we analyzed the expression of CXCL16 in situ by applying immunohistochemistry on tissue sections from 10 RA and 5 control synovia. The results demonstrated that in control synovia, the expression of CXCL16 was confined to the single layer of synovial lining macrophages (Figure 2A). Synovial tissue from RA patients, however, showed a strong increase in the number of CXCL16+ cells that was most predominant in the thickened synovial lining and sublining (Figures 2B and C). The number of CXCL16-expressing cells was directly related to the degree of cellularity of the ST (results not shown). The overlapping expression patterns of CXCL16 and the macrophage marker CD68 confirmed that the CXCL16+ cells in the synovial lining were CD68+ macrophages (Figures 2E and F). Interestingly, while essentially all blood vessels in control synovia appeared to be CXCL16 negative (results not shown), some CD31+ blood vessels in RA synovium expressed significant levels of CXCL16 (Figures 2G and H). In addition, CXCL16-expressing cells were also detected within perivascular lymphoid aggregates (Figure 2I). Staining of serial sections suggested that the CXCL16+ cells that were surrounded by CD45RO+ memory lymphocytes were also CD68+ macrophages (Figures 2I–K). These results demonstrate that, due to increased numbers of macrophages and activated endothelial cells, the expression of the chemokine CXCL16 is highly increased in RA synovia compared with control synovia.

thumbnail image

Figure 2. Enhanced expression of CXCL16 protein within rheumatoid arthritis (RA) synovia. CXCL16 was expressed in the thin synovial lining of healthy synovia (A) (arrowhead) but was much more pronounced in the hypertrophic lining of RA synovia (B). Many cells present within the sublining of the synovia of patients with RA expressed high levels of CXCL16 (B and C). Note that (cleaved) CXCL16 was also associated with filaments of the extracellular matrix (B). Staining of serial sections indicated that expression of CXCL16 (E) correlated with the presence of CD68+ synovial macrophages (F). We also detected CXCL16 within some vessels (G) (arrow). Staining of serial sections for CD31 confirmed that these cells were endothelial cells (H). Many lymphocyte aggregates contained CXCL16+ cells (I) (arrows). Analysis of stained serial sections suggested that these CXCL16+ cells were CD68+ macrophages (J) (arrows) amidst CD45RO+ memory lymphocytes (K). Cryosections were stained for CXCL16 (A–C, E, G, and I), CD68 (F and J), CD31 (H), CD45RO (K), or matched control antibodies (D and I). All sections were counterstained with hematoxylin. Sections shown are representative of synovial tissue from 10 RA patients and 5 controls. (Original magnification × 400 in AF and IL; × 630 in G and H.)

Download figure to PowerPoint

Enhancement of CXCL16 expression in macrophages by addition of RA SF or TNF.

Next, we investigated the expression of CXCL16 by blood-derived monocytes and macrophages. Whereas freshly isolated monocytes did not express CXCL16 on their cell surface, upon differentiation into macrophages, they rapidly began to express transmembrane CXCL16 and remained CXCL16+ for at least 10 days of culture (Figures 3A and B). Intriguingly, addition of RA SF to freshly isolated monocytes resulted in a significant and concentration-dependent increase in cell surface–expressed CXCL16 (Figure 3C). Also, TNF, one of the constituents of RA SF and a key protein in RA, was sufficient to increase the expression of transmembrane CXCL16.

thumbnail image

Figure 3. Strongly increased expression of CXCL16 by macrophages (MΦ) after exposure to RA SF or tumor necrosis factor (TNF). A, Peripheral blood monocytes were differentiated into macrophages and analyzed by flow cytometry. CD14+ monocyte/macrophages were gated and stained with anti-CXCL16 (bold lines) or isotype-matched control antibodies (thin lines). Dead cells were excluded by gating on propidium iodide–negative cells. B, Mean percentages of CXCL16+ macrophages at the time points indicated. C, Increasing levels of transmembrane CXCL16 on macrophages, with addition of increasing concentrations of RA SF or TNF to freshly isolated monocytes. On day 2, the percentage of CXCL16+,CD14+ macrophages was determined by flow cytometry. The mean and SD percentages from 2 donors are shown. Dead cells were excluded by gating on propidium iodide–negative cells. D, Significant levels of cleaved CXCL16 released by cultured macrophages. By sandwich enzyme-linked immunosorbent assay (ELISA), the supernatants of macrophages (MAC) or peripheral blood lymphocytes (PBL) cultured for 1, 2, or 3 days were analyzed for the presence of cleaved CXCL16. Values are the mean. E, TNF-induced expression of cleaved CXCL16. Culture supernatants of day-2 macrophages were analyzed by sandwich ELISA. Representative results of 2–5 experiments are shown. ∗ = P < 0.05; ∗∗ = P < 0.01; ∗∗∗ = P < 0.001 versus unstimulated group. See Figure 1 for other definitions.

Download figure to PowerPoint

Because cleavage of CXCL16 is necessary in order for it to exert its chemotactic activity toward CXCR6-expressing lymphocytes, we also analyzed the macrophage culture supernatants for the presence of cleaved CXCL16. As shown in Figure 3D, while PBL cultures were essentially negative, significant amounts of cleaved CXCL16 were readily detected in the culture supernatant of day-1–3 macrophages. Moreover, mimicking the RA inflammatory environment through the addition of TNF further significantly increased the amount of cleaved CXCL16 in the culture supernatant (Figure 3E). Because RA SF contains cleaved CXCL16 itself (see below), we did not measure soluble CXCL16 in the supernatant of RA SF–treated macrophages.

To investigate the effect of TNF on the expression of CXCL16 in vivo, we determined the expression of CXCL16 following anti-TNF therapy in synovia of 3 RA patients whose disease responded to the treatment as well as in synovia of 3 nonresponders. The synovial lining and sublining of all patients contained large amounts of CXCL16+ synovial macrophages prior to treatment (Figures 4A and C). Intriguingly, CXCL16 expression was severely reduced in the clinically responding patients (compare Figure 4A with Figure 4B) but not in the nonresponding patients (compare Figure 4C with Figure 4D). Taken together, these data demonstrate that TNF and RA SF, stimuli associated with synovial inflammation, increase the expression of CXCL16 in vitro and in vivo.

thumbnail image

Figure 4. Reduced CXCL16 expression upon successful anti–tumor necrosis factor (anti-TNF) treatment of rheumatoid arthritis (RA) patients. Before anti-TNF treatment, CXCL16 was strongly expressed within RA synovia (A and C). Expression of CXCL16 was severely reduced in clinically responding patients (B), but not in nonresponding patients (D). Synovia were isolated before (A and C) and after (B and D) anti-TNF treatment. All cryosections were stained for CXCL16 and counterstained with hematoxylin. Control stainings were negative. Results are representative of 3 responding and 3 nonresponding patients. (Original magnification × 100.)

Download figure to PowerPoint

ADAM-10 expression in RA synovia coincides with the presence of cleaved CXCL16 in RA SF.

ADAM-10 has very recently been implicated in the cleavage of CXCL16 (27, 28). Analysis of monocytes and monocyte-derived macrophages revealed that transmembrane ADAM-10, like CXCL16, was induced on day-2 macrophages and was further up-regulated by addition of RA SF (Figures 5A and B). Moreover, analysis of serial sections from RA ST showed that expression of the protease ADAM-10 overlapped completely with the CXCL16-expressing macrophages in the hypercellular synovial lining (Figures 5F and G). RT-PCR analysis further confirmed that ADAM-10 mRNA was expressed by RA ST (results not shown). In control ST, ADAM-10 staining was limited to the thin synovial lining but still overlapped with the CXCL16 staining (Figures 5C and D).

thumbnail image

Figure 5. Enhanced expression of the CXCL16 protease ADAM-10 by RA synovial macrophages (MΦ). ADAM-10–negative peripheral blood monocytes were differentiated into macrophages and analyzed by flow cytometry (A). CD14+ monocyte/macrophages were gated and stained with anti–ADAM-10 (bold lines) or isotype-matched control antibodies (thin lines). Dead cells were excluded by gating on propidium iodide–negative cells. Addition of RA SF to macrophages (day 3) increased their expression of ADAM-10 (B). After 2 days the percentage of ADAM-10+,CD14+ macrophages was determined by flow cytometry. The mean and SD percentages from 2 donors are shown. Dead cells were excluded by gating on propidium iodide–negative cells. ∗ = P < 0.05 versus controls. Synovial tissue from controls showed expression of both CXCL16 (C) and its protease ADAM-10 (D) in the thin synovial lining. In contrast, expression of both CXCL16 (F) and ADAM-10 (G) was highly enhanced in the hypercellular synovial lining of RA synovial tissue. Frozen sections were stained for CXCL16 (C and F), ADAM-10 (D and G), or a matched control antibody (E), and counterstained with hematoxylin. Representative stainings are shown. See Figure 1 for other definitions. (Original magnification × 400.)

Download figure to PowerPoint

To investigate whether the enhanced expression of both CXCL16 and ADAM-10 by synovial macrophages of RA patients results in elevated levels of cleaved CXCL16, we performed ELISA on serum and SF from controls or from patients with either OA or RA. Although cleaved CXCL16 was readily detectable in serum, we did not observe significant differences between sera from RA patients and healthy controls (data not shown). In contrast, analysis of SF demonstrated that RA SF contained significantly more cleaved CXCL16 than did SF from controls (Figure 6A). Interestingly, SF from patients with the milder inflammatory disease OA also contained less cleaved CXCL16 than did RA SF, although this difference was not significant. These data strongly suggest that the increase in CXCL16- and ADAM-10–expressing macrophages in RA ST also results in the release of high amounts of cleaved CXCL16 into the SF in vivo.

thumbnail image

Figure 6. Cleaved CXCL16 overexpression in RA SF and attraction of SF T cells. A, Expression of cleaved CXCL16 in RA SF (n = 17) was compared with that in SF from controls (n = 2) or from patients with osteoarthritis (OA; n = 5). CXCL16 concentrations were measured by sandwich enzyme-linked immunosorbent assay. Values are the mean and SD. ∗ = P < 0.02. B, Loss of cell surface–expressed CXCR6 on RA SF T cells induced by activation with cleaved CXCL16. The level of CXCR6 expressed by unstimulated T cells was set at 100%. C, SF T cells from RA patients are attracted by soluble CXCL16. Migration was concentration dependent and reached an optimum at 10 ng/ml. Migration of RA SF T cells was determined with a Transwell migration assay. Migrating cells were counted and stained for CD3 and CXCR6. Representative results from 3 experiments are shown. Values are the mean and SD. See Figure 1 for other definitions.

Download figure to PowerPoint

Cleaved CXCL16 attraction of memory T cells from RA SF.

Finally, we determined whether the large numbers of CXCR6+ T cells observed in RA SF were indeed functionally capable of responding to cleaved CXCL16. As shown in Figure 6B, RA SF–derived T cells rapidly lost their CXCR6 expression upon incubation with increasing concentrations of cleaved CXCL16. These data indicate that the CXCR6+ T cells derived from RA SF internalized CXCR6 upon binding of cleaved CXCL16 and thus were able to respond to CXCL16. Subsequent migration experiments further confirmed that cleaved CXCL16 attracted RA SF T cells in a concentration-dependent manner (Figure 6C). These results indicate that large numbers of RA SF T cells express functional CXCR6, allowing migration toward strong CXCL16 sources such as RA SF.


  1. Top of page
  2. Abstract
  6. Acknowledgements

Influx of leukocytes, including CXCR6+ leukocytes, into both ST and SF contributes to the pathogenesis of RA. Here, we report that the chemokine CXCL16, the ligand for CXCR6, is normally expressed by macrophages in the thin lining of healthy synovia. In RA synovia, CXCL16 expression is elevated and strongly increased due to the presence of large numbers of synovial macrophages. In addition, we show that these CXCL16+ macrophages strongly express the recently identified CXCL16 protease ADAM-10 in situ. In vitro studies demonstrated that monocyte-derived macrophages express both transmembrane and cleaved CXCL16 and that the expression is enhanced by RA SF and the proinflammatory stimulus TNF. Moreover, only successful anti-TNF therapy is associated with decreased CXCL16 expression in situ. Finally, elevated expression of both CXCL16 and ADAM-10 by RA ST macrophages in situ is associated with high amounts of cleaved CXCL16 in RA SF and with the presence of significantly increased numbers of CXCR6+ T cells in this SF.

CXCL16 is a recently identified transmembrane chemokine expressed by macrophages and DCs (12–14). Upon proteolytic cleavage, the NH2-terminal part of CXCL16 is released and functions as a soluble chemoattractant for CXCR6+ T cells and plasma cells (12, 14). Interestingly, Kim et al reported that CXCR6 is a marker for effector/memory T cells and that large numbers of CXCR6+ T cells were detected in SF from 3 RA patients (21). Here, we extended these observations and added novel data concerning the expression and function of the only ligand for CXCR6, CXCL16. First, we demonstrated that fresh control or RA PBMC contain only small numbers of CXCR6+ T cells (Figures 1A and B). In contrast, up to 80% of the T cells within RA SF expressed CXCR6. Immunohistochemical staining of RA ST for CXCR6 revealed no significant staining of T cells (results not shown). However, PCR analysis demonstrated that low levels of CXCR6 mRNA are also present within ST from RA patients (results not shown). Since chemokine receptors are generally expressed at relatively low levels, and we demonstrated that CXCR6 is rapidly down-regulated upon CXCL16 binding (Figure 6B), these data suggest that CXCR6 is difficult to detect by immunohistochemistry.

Quantitative analysis demonstrated that the expression of CXCL16 mRNA is increased in RA ST compared with control ST (Figure 1D). Immunostaining revealed that in healthy individuals, CXCL16 is expressed by the single layer of cells that make up the synovial lining (Figure 2). In RA patients, however, elevated expression of CXCL16 is detected in the hypercellular synovial lining, while additional CXCL16+ cells are detected in the sublining and within perivascular lymphocyte aggregates. Overall, CXCL16 expression correlated with increased cellularity of the RA synovia. Based on CD68/CXCL16 staining of serial sections and morphology, the CXCL16+ cells in the lining, sublining, and lymphocyte aggregates mainly represent macrophages. Although we did not observe colocalization of CXCL16 with the DC marker CD208/DC-LAMP (results not shown), we cannot exclude the possibility that some DCs or follicular DCs also express CXCL16. Finally, CD31/CXCL16 staining of serial sections implied that some endothelial cells express CXCL16 (Figures 2G and H). These data are consistent with previous results demonstrating that cardiac and umbilical endothelial cells can express CXCL16 (29, 30).

Furthermore, our in vitro studies show that a significantly increased population of macrophages expressed transmembrane CXCL16 upon addition of RA SF or TNF, one of the major cytokines in RA SF (Figure 3). Moreover, TNF-treated macrophages also released increased amounts of cleaved CXCL16. This could be of importance, since TNF is strongly expressed in RA joints and is considered to be a key player in RA pathogenesis (31). In fact, anti-TNF treatment is effectively used to treat RA (32–34). Interestingly, we observed severely reduced CXCL16 expression in ST from RA patients responding to anti-TNF treatment, but not in nonresponding patients (Figure 4). These data suggest that not only in vitro, but also in vivo the expression of CXCL16 is controlled by TNF. We are currently extending these studies with a larger cohort of RA patients receiving anti-TNF therapy.

Interestingly, highly elevated levels of cleaved CXCL16 were present in SF from RA patients, and this coincided with large numbers of ADAM-10+ macrophages (Figures 5 and 6). ADAM-10 has recently been described to be a major protease involved in the cleavage and release of CXCL16 (27, 28). Therefore, our data suggest that in RA, ADAM-10 expression by the thick layer of synovial lining macrophages is involved in the release of large amounts of cleaved CXCL16 in the SF. We note that significant amounts of CXCL16 are also detected in serum of healthy individuals, suggesting that cleavage of CXCL16 also occurs in steady-state conditions. This reasoning is in accordance with our observation that the synovial lining of healthy individuals does express low levels of both CXCL16 and ADAM-10. Since CXCL16 has been suggested to contain multiple restriction sites for proteases (12, 14), other proteases, e.g., matrix metalloproteinases (MMPs) or TNF-converting enzyme, are also likely to be involved in the cleavage of CXCL16. With respect to cleavage of CXCL16 in synovium, MMP-1 could be an interesting candidate since we have recently shown that this protease is abundantly expressed by RA synovial macrophages (35, 36).

Finally, we demonstrate that cleaved CXCL16 indeed activates CXCR6 expressed by RA SF T cells (Figure 6). First, addition of cleaved CXCL16 to these T cells leads to the loss of cell surface CXCR6, suggesting CXCL16-mediated CXCR6 internalization. Ligand-induced activation and subsequent internalization is a common feature of chemokine receptors (37). After being internalized, some chemokine receptors recycle back to the cell membrane, while others are degraded in the lysosomal compartment. As yet, it is not known how CXCR6 behaves after being internalized. However, we have demonstrated that CXCR6+ T cells isolated from RA SF are capable of migrating in response to CXCL16 in vitro (Figure 6C). Therefore, our data imply that CXCL16 and CXCR6 play an important role in the recruitment of activated T cells into RA joints.

Several mouse studies have confirmed the importance of chemokines in RA development in vivo. Administration of a CCL2/monocyte chemoattractant protein 1 (MCP-1) antagonist prevented the onset of arthritis in the MRL-lpr arthritis model (38), and neutralizing CXCL10/inducible protein 10 antibodies prevented adjuvant-induced arthritis (39). Also in RA patients, adhesion molecules and chemokines play important roles in synovial infiltration and disease pathogenesis. For instance, enhanced expression of adhesion molecules, e.g., E-selectin, vascular cell adhesion molecule 1, and intercellular adhesion molecule 1 (32, 40), and various chemokines, including CCL2/MCP-1, CCL5/RANTES, CCL18/DC-CK1, and CXCL8/IL-8, have been detected in RA tissue and/or SF (41–48). Interestingly, therapy with a CCR1 antagonist has recently been shown to be beneficial in RA (49). Oral administration of this antagonist significantly reduced the number of ST macrophages and T cells, and this was correlated with a trend toward clinical improvement compared with placebo-treated controls. Despite the apparent redundancy in the chemokine system, evidence is accumulating that chemokine and chemokine receptor antagonists have strong potential as therapeutic agents for patients with autoimmune disease (37, 50). Our data suggest that the use of either CXCR6 antagonists or protease inhibitors acting on CXCL16 cleavage could be additional novel approaches to treat patients with RA.

Based on the results of this study, the following model for the role of CXCL16/CXCR6 in RA pathogenesis can be envisaged. During inflammation of the joint, locally activated endothelial cells express increased levels of adhesion molecules and chemokines, resulting in enhanced immigration of monocytes. Within the synovial tissue, these monocytes now become attracted by chemokines released by the synovial lining, and they differentiate into macrophages. Differentiating macrophages start to express both CXCL16 and ADAM-10, expression of which is further enhanced by SF from the synovial cavity and/or by TNF released by the macrophages themselves. At the now thickened synovial lining, ADAM-10 cleaves transmembrane CXCL16, resulting in elevated concentrations of cleaved CXCL16 in the SF. Cleaved CXCL16 attracts large numbers of CXCR6+ memory T cells into the RA joint. These memory T cells release cytokines, such as TNF, that can now activate macrophages and other resident cells, thus sustaining the inflammatory cascade contributing to RA pathogenesis.

In conclusion, our data suggest that overexpression of CXCL16 targets CXCR6+ memory T cells to synovia from RA patients. Therefore, CXCL16 and CXCR6 could be intrinsically involved in the inflammation associated with RA pathology.


  1. Top of page
  2. Abstract
  6. Acknowledgements

We would like to thank Dr. M. C. de Waal Malefijt (Department of Orthopedic Surgery, University Medical Center Nijmegen) for collecting control SF and ST, Dr. A. B. van Spriel (Department of Tumor Immunology, Nijmegen Center for Molecular Life Sciences) for sharing FACS data, Mrs. M. Roelofs (Department of Rheumatology, University Medical Center Nijmegen) for providing RA PBMCs, Dr. H. Dolstra and Mr. B. de Rijke (Department of Hematology, University Medical Center Nijmegen) for CXCL16 and PBGD primers, and Mrs. N. Jansink for technical support.


  1. Top of page
  2. Abstract
  6. Acknowledgements
  • 1
    Bresnihan B. Pathogenesis of joint damage in rheumatoid arthritis. J Rheumatol 1999; 26: 7179.
  • 2
    Burmester GR, Stuhlmuller B, Keyszer G, Kinne RW. Mononuclear phagocytes and rheumatoid synovitis: mastermind or workhorse in arthritis? [review]. Arthritis Rheum 1997; 40: 518.
  • 3
    Tak PP, Bresnihan B. The pathogenesis and prevention of joint damage in rheumatoid arthritis: advances from synovial biopsy and tissue analysis [review]. Arthritis Rheum 2000; 43: 261933.
  • 4
    Weyand CM, Goronzy JJ. Ectopic germinal center formation in rheumatoid synovitis. Ann N Y Acad Sci 2003; 987: 1409.
  • 5
    Van den Berg WB, van Lent PL. The role of macrophages in chronic arthritis. Immunobiology 1996; 195: 61423.
  • 6
    Gravallese EM, Goldring SR. Cellular mechanisms and the role of cytokines in bone erosions in rheumatoid arthritis [review]. Arthritis Rheum 2000; 43: 214351.
  • 7
    Feldmann M, Maini RN. Anti-TNF α therapy of rheumatoid arthritis: what have we learned? Annu Rev Immunol 2001; 19: 16396.
  • 8
    Zlotnik A, Yoshie O. Chemokines: a new classification system and their role in immunity. Immunity 2000; 12: 1217.
  • 9
    Moser B, Wolf M, Walz A, Loetscher P. Chemokines: multiple levels of leukocyte migration control. Trends Immunol 2004; 25: 7584.
  • 10
    Springer TA. Traffic signals for lymphocyte recirculation and leukocyte emigration: the multistep paradigm. Cell 1994; 76: 30114.
  • 11
    Butcher EC, Picker LJ. Lymphocyte homing and homeostasis. Science 1996; 272: 606.
  • 12
    Matloubian M, David A, Engel S, Ryan JE, Cyster JG. A transmembrane CXC chemokine is a ligand for HIV-coreceptor Bonzo. Nat Immunol 2000; 1: 298304.
  • 13
    Shimaoka T, Kume N, Minami M, Hayashida K, Kataoka H, Kita T, et al. Molecular cloning of a novel scavenger receptor for oxidized low density lipoprotein, SR-PSOX, on macrophages. J Biol Chem 2000; 275: 406636.
  • 14
    Wilbanks A, Zondlo SC, Murphy K, Mak S, Soler D, Langdon P, et al. Expression cloning of the STRL33/BONZO/TYMSTR ligand reveals elements of CC, CXC, and CX3C chemokines. J Immunol 2001; 166: 514554.
  • 15
    Shimaoka T, Nakayama T, Fukumoto N, Kume N, Takahashi S, Yamaguchi J, et al. Cell surface-anchored SR-PSOX/CXC chemokine ligand 16 mediates firm adhesion of CXC chemokine receptor 6-expressing cells. J Leukoc Biol 2004; 75: 26774.
  • 16
    Shimaoka T, Nakayama T, Kume N, Takahashi S, Yamaguchi J, Minami M, et al. Cutting edge: SR-PSOX/CXC chemokine ligand 16 mediates bacterial phagocytosis by APCs through its chemokine domain. J Immunol 2003; 171: 164751.
  • 17
    Shimaoka T, Nakayama T, Hieshima K, Kume N, Fukumoto N, Minami M, et al. Chemokines generally exhibit scavenger receptor activity through their receptor-binding domain. J Biol Chem 2004; 279: 2680710.
  • 18
    Bazan JF, Bacon KB, Hardiman G, Wang W, Soo K, Rossi D, et al. A new class of membrane-bound chemokine with a CX3C motif. Nature 1997; 385: 6404.
  • 19
    Fong AM, Robinson LA, Steeber DA, Tedder TF, Yoshie O, Imai T, et al. Fractalkine and CX3CR1 mediate a novel mechanism of leukocyte capture, firm adhesion, and activation under physiologic flow. J Exp Med 1998; 188: 14139.
  • 20
    Imai T, Hieshima K, Haskell C, Baba M, Nagira M, Nishimura M, et al. Identification and molecular characterization of fractalkine receptor CX3CR1, which mediates both leukocyte migration and adhesion. Cell 1997; 91: 52130.
  • 21
    Kim CH, Kunkel EJ, Boisvert J, Johnston B, Campbell JJ, Genovese MC, et al. Bonzo/CXCR6 expression defines type 1-polarized T-cell subsets with extralymphoid tissue homing potential. J Clin Invest 2001; 107: 595601.
  • 22
    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: 31524.
  • 23
    Prevoo ML, van 't Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified Disease Activity Scores that include twenty-eight–joint counts: development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995; 38: 448.
  • 24
    Joosten LA, Koenders MI, Smeets RL, Heuvelmans-Jacobs M, Helsen MM, Takeda K, et al. Toll-like receptor 2 pathway drives streptococcal cell wall-induced joint inflammation: critical role of myeloid differentiation factor 88. J Immunol 2003; 171: 614553.
  • 25
    Krooshoop DJ, Torensma R, van den Bosch GJ, Nelissen JM, Figdor CG, Raymakers RA, et al. An automated multi well cell track system to study leukocyte migration. J Immunol Methods 2003; 280: 89102.
  • 26
    Campbell JJ, Qin S, Bacon KB, Mackay CR, Butcher EC. Biology of chemokine and classical chemoattractant receptors: differential requirements for adhesion-triggering versus chemotactic responses in lymphoid cells. J Cell Biol 1996; 134: 25566.
  • 27
    Abel S, Hundhausen C, Mentlein R, Schulte A, Berkhout TA, Broadway N, et al. The transmembrane CXC-chemokine ligand 16 is induced by IFNγ and TNFα and shed by the activity of the disintegrin-like metalloproteinase ADAM-10. J Immunol 2004; 172: 636272.
  • 28
    Gough PJ, Garton KJ, Wille PT, Rychlewski M, Dempsey PJ, Raines EW. A disintegrin and metalloproteinase 10-mediated cleavage and shedding regulates the cell surface expression of CXC chemokine ligand 16. J Immunol 2004; 172: 367885.
  • 29
    Hofnagel O, Luechtenborg B, Plenz G, Robenek H. Expression of the novel scavenger receptor SR-PSOX in cultured aortic smooth muscle cells and umbilical endothelial cells. Arterioscler Thromb Vasc Biol 2002; 22: 7101.
  • 30
    Yamauchi R, Tanaka M, Kume N, Minami M, Kawamoto T, Togi K, et al. Upregulation of SR-PSOX/CXCL16 and recruitment of CD8+ T cells in cardiac valves during inflammatory valvular heart disease. Arterioscler Thromb Vasc Biol 2004; 24: 2827.
  • 31
    Feldmann M, Maini RN. Lasker Clinical Medical Research Award: TNF defined as a therapeutic target for rheumatoid arthritis and other autoimmune diseases. Nat Med 2003; 9: 124550.
  • 32
    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 α monoclonal antibody treatment in patients with rheumatoid arthritis. Arthritis Rheum 1996; 39: 107781.
  • 33
    Taylor PC, Peters AM, Paleolog E, Chapman PT, Elliott MJ, McCloskey R, et al. Reduction of chemokine levels and leukocyte traffic to joints by tumor necrosis factor α blockade in patients with rheumatoid arthritis. Arthritis Rheum 2000; 43: 3847.
  • 34
    Smeets TJ, Kraan MC, van Loon ME, Tak PP. Tumor necrosis factor α blockade reduces the synovial cell infiltrate early after initiation of treatment, but apparently not by induction of apoptosis in synovial tissue. Arthritis Rheum 2003; 48: 215562.
  • 35
    Blom AB, Radstake TR, Holthuysen AE, Sloetjes AW, Pesman GJ, Sweep FG, et al. Increased expression of Fcγ receptors II and III on macrophages of rheumatoid arthritis patients results in higher production of tumor necrosis factor α and matrix metalloproteinase. Arthritis Rheum 2003; 48: 100214.
  • 36
    Van Lent PL, Figdor CG, Barrera P, van Ginkel K, Sloetjes A, van den Berg WB, et al. Expression of the dendritic cell–associated C-type lectin DC-SIGN by inflammatory matrix metalloproteinase–producing macrophages in rheumatoid arthritis synovium and interaction with intercellular adhesion molecule 3–positive T cells. Arthritis Rheum 2003; 48: 3609.
  • 37
    Proudfoot AE. Chemokine receptors: multifaceted therapeutic targets. Nat Rev Immunol 2002; 2: 10615.
  • 38
    Gong JH, Ratkay LG, Waterfield JD, Clark-Lewis I. An antagonist of monocyte chemoattractant protein 1 (MCP-1) inhibits arthritis in the MRL-lpr mouse model. J Exp Med 1997; 186: 1317.
  • 39
    Salomon I, Netzer N, Wildbaum G, Schif-Zuck S, Maor G, Karin N. Targeting the function of IFN-γ-inducible protein 10 suppresses ongoing adjuvant arthritis. J Immunol 2002; 169: 268593.
  • 40
    Koch AE, Burrows JC, Haines GK, Carlos TM, Harlan JM, Leibovich SJ. Immunolocalization of endothelial and leukocyte adhesion molecules in human rheumatoid and osteoarthritic synovial tissues. Lab Invest 1991; 64: 31320.
  • 41
    Radstake TR, van der Voort R, Ten Brummelhuis M, de Waal Malefijt M, Schreurs W, Looman M, et al. Increased expression of CCL18, CCL19, and CCL17 by dendritic cells from patients with rheumatoid arthritis and regulation by Fcγ receptors. Ann Rheum Dis. In press.
  • 42
    Koch AE, Kunkel SL, Burrows JC, Evanoff HL, Haines GK, Pope RM, et al. Synovial tissue macrophage as a source of the chemotactic cytokine IL-8. J Immunol 1991; 147: 218795.
  • 43
    Koch AE, Kunkel SL, Harlow LA, Johnson B, Evanoff HL, Haines GK, et al. Enhanced production of monocyte chemoattractant protein-1 in rheumatoid arthritis. J Clin Invest 1992; 90: 7729.
  • 44
    Villiger PM, Terkeltaub R, Lotz M. Production of monocyte chemoattractant protein-1 by inflamed synovial tissue and cultured synoviocytes. J Immunol 1992; 149: 7227.
  • 45
    Rathanaswami P, Hachicha M, Wong WL, Schall TJ, McColl SR. Synergistic effect of interleukin-1β and tumor necrosis factor α on interleukin-8 gene expression in synovial fibroblasts: evidence that interleukin-8 is the major neutrophil-activating chemokine released in response to monokine activation. Arthritis Rheum 1993; 36: 1295304.
  • 46
    Rathanaswami P, Hachicha M, Sadick M, Schall TJ, McColl SR. Expression of the cytokine RANTES in human rheumatoid synovial fibroblasts: differential regulation of RANTES and interleukin-8 genes by inflammatory cytokines. J Biol Chem 1993; 268: 58349.
  • 47
    Suzuki N, Nakajima A, Yoshino S, Matsushima K, Yagita H, Okumura K. Selective accumulation of CCR5+ T lymphocytes into inflamed joints of rheumatoid arthritis. Int Immunol 1999; 11: 5539.
  • 48
    Kraan MC, Patel DD, Haringman JJ, Smith MD, Weedon H, Ahern MJ, et al. The development of clinical signs of rheumatoid synovial inflammation is associated with increased synthesis of the chemokine CXCL8 (interleukin-8). Arthritis Res 2001; 3: 6571.
  • 49
    Haringman JJ, Kraan MC, Smeets TJ, Zwinderman KH, Tak PP. Chemokine blockade and chronic inflammatory disease: proof of concept in patients with rheumatoid arthritis. Ann Rheum Dis 2003; 62: 71521.
  • 50
    Kunkel SL, Godessart N. Chemokines in autoimmunity: from pathology to therapeutics. Autoimmun Rev 2002; 1: 31320.