Osteopontin (OPN) is expressed by fibroblast-like synoviocytes (FLS) in rheumatoid arthritis (RA), but its pathologic role is still obscure. The present study was undertaken to analyze the role of OPN in RA by focusing on its effects on cell–cell interactions between FLS and B lymphocytes.
FLS obtained from 10 patients with RA and 10 non-RA subjects and a B lymphocyte cell line were studied. The characteristics of OPN expression by FLS were analyzed by Western blotting, immunoprecipitation, and immunofluorescence studies. In cocultures of FLS and B lymphocytes, the effects of OPN on adhesion of B lymphocytes to FLS and the consequent production of interleukin-6 (IL-6) were analyzed in experiments involving overexpression and knockdown of OPN and inhibitory studies with an OPN-blocking antibody. In vivo, the expression of OPN in RA synovium was examined by immunohistochemistry.
A specifically modified 75-kd form of OPN was predominantly expressed in RA FLS, and this was associated with expression of >200-kd thrombin-cleaved OPN that was crosslinked with fibronectin and localized on the surface of the FLS. In FLS–B lymphocyte cocultures, 75-kd OPN–positive FLS produced a significantly higher amount of IL-6 than did 75-kd OPN–negative FLS. When the FLS were separated from B lymphocytes or cultured alone, the production of IL-6 was low and was not significantly different between these 2 culture conditions. Moreover, OPN overexpression enhanced production of IL-6 in 75-kd OPN–positive FLS–B lymphocyte cocultures. Addition of the OPN-blocking antibody inhibited the adhesion of B lymphocytes to FLS. Immunohistochemical analyses revealed that localization of IL-6–positive cells coincided with the sites at which OPN and B lymphocytes were colocalized.
Specifically modified 75-kd OPN was expressed by RA FLS. This form of OPN affected FLS–B lymphocyte interactions by supporting the adhesion of B lymphocytes to FLS and enhancing the production of IL-6.
In patients with rheumatoid arthritis (RA), the synovium of the inflamed joints is the site of a chronic inflammatory reaction, in which leukocytes and macrophages infiltrate, synoviocytes proliferate, and several proinflammatory cytokines (especially, interleukin-6 [IL-6] and tumor necrosis factor α), autoantibodies, and immune complexes are produced (1). In this situation, infiltrating leukocytes and proliferating synoviocytes are thought to interact with each other. Accordingly, several in vitro studies have focused on the interaction between fibroblast-like synoviocytes (FLS) and leukocytes by studying the findings in cocultures of these cells (2–7). Coculture of FLS obtained from RA synovium with B lymphocytes causes adhesion of B lymphocytes to FLS, and consequently this interaction supports the survival of B lymphocytes and also enhances the production of cytokines and immunoglobulin (8–12). The former mechanism is dependent on vascular cell adhesion molecule 1 (VCAM-1) and very late activation antigen 4 (VLA-4) (10), whereas the latter is yet to be determined.
Osteopontin (OPN) is abundant in the bone matrix, where it acts as a bridge between hydroxyapatite and osteoclasts to support bone resorption (13–15). It is also secreted by T lymphocytes and assists in the maturation of B lymphocytes and the migration of macrophages (16–18). These diverse functions are partly explained by the several functional domains of OPN, including integrin-binding, calcium-binding, and heparin-binding sites. Several studies have revealed that the function of OPN depends on posttranslational modifications (19–24). These modifications, which vary between different OPN-expressing cells, consist of phosphorylation at dozens of serine and threonine residues, along with glycosylation (25), sialylation (21), transglutamination (26), and cleavage (27, 28). However, the association between the details of posttranslational modification and the function of OPN is still poorly understood.
As for the relationship between OPN and arthritis, OPN-null mice are protected against inflammatory joint destruction in collagen-induced arthritis (29). A blocking antibody directed against the thrombin-cleaved neoepitope of OPN, which cooperates with several integrin receptors as a ligand (30, 31), also shows a curative effect on induced arthritis in mice and monkeys (32, 33). Indeed, in patients with RA, OPN, especially in the thrombin-cleaved form, is strongly detected in the synovium and synovial fluid of inflamed joints (34–36). In vitro studies on the function of OPN in arthritis have revealed that OPN stimulates the production of several proinflammatory cytokines by mononuclear cells in patients with RA (37), and also that monocytes obtained from mice with induced arthritis show increased migration toward thrombin-cleaved OPN (32). However, these studies were performed using recombinant OPN, which lacks various posttranslational modifications, and thus the posttranslational modification–dependent functions were not taken into account.
Considering these results and the fact that OPN is also strongly expressed in FLS from the RA synovium (35), we hypothesized that RA FLS express a unique form of OPN that acts to stimulate production of cytokines by creating a bridge between FLS and B lymphocytes in cocultures of these cells. The purpose of this study was to analyze the role of OPN in the development of RA, by focusing on the interaction between FLS and B lymphocytes.
PATIENTS AND METHODS
Patients and cell culture.
After obtaining the patients' informed consent and securing Institutional Review Board approval, synovial tissue samples were collected from 11 patients with RA, 8 patients with osteoarthritis of the knee, and 2 patients with medial meniscus degeneration who underwent meniscectomy and synovectomy. All of the patients with RA fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) revised criteria for RA (38). FLS were isolated by enzymatic digestion of the synovial tissue from 10 patients with RA and 10 non-RA subjects, as described previously (9). The FLS were maintained in Dulbecco's modified Eagle's medium (Gibco BRL, Grand Island, NY) containing 10% fetal bovine serum (FBS; Hyclone Laboratories, Logan, UT) and 1% penicillin/streptomycin (Gibco BRL). Cells from passages 4–10 were used for these experiments. A human B cell line, MC/car, was purchased from the American Type Culture Collection (Rockville, MD) and maintained in RPMI 1640 medium (Gibco BRL) containing 10% FBS and 1% penicillin/streptomycin.
For coculture experiments, 4 × 104 FLS were plated in 12-well plates and, on the following day, 3.3 × 105 B lymphocytes were added to each well. To prevent cell–cell contact between FLS and B lymphocytes, a Millicell culture insert with 0.4-μm pores (Millipore, Billerica, MA) was used. After incubation for 48 hours, enzyme immunoassay (EIA) was performed to measure the concentrations of IL-6 and OPN in the culture supernatant using a homogeneous time-resolved fluorescence human IL-6 assay kit (CIS Bio International, Saclay, France) and a human OPN assay kit (IBL Japan, Gunma, Japan).
Protein collection, sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), and Western blotting.
Total cell lysates of FLS were collected by placing the cells in lysis buffer (10× Cell Lysis Buffer; New England Biolabs, Beverly, MA), using 4 × 104 FLS plated in 12-well plates and grown in normal growth medium; some of the FLS had undergone transfection with small interfering RNA (siRNA) or lentiviral infection. We did not add any other reagent to the lysis buffer, which consisted of Tris HCl buffer with detergent, protease inhibitor (leupeptin), and phosphatase inhibitors (sodium pyrophosphate, β-glycerophosphate, and sodium orthovanadate). After 1 freeze–thaw cycle, samples were centrifuged and the supernatant was collected. An aliquot was obtained for protein quantification with the bicinchoninic acid assay, and the remaining supernatant was boiled in SDS sample buffer. As a control, recombinant human OPN was used.
Surface protein was collected from 1.2 × 106 FLS using a Cell Surface Protein isolation kit (Pierce, Rockford, IL). Equal amounts of each sample were separated by SDS-PAGE, blotted onto a polyvinylidene difluoride membrane (Hybond-P; Amersham, Piscataway, NJ), and blocked with 5% bovine serum albumin (BSA)/Tris buffered saline containing 0.1% Tween 20 (TBST). The membrane was then incubated overnight at 4°C with one of the following primary antibodies: rabbit anti-human OPN antibody (O-17; IBL Japan) or mouse anti-OPN N-Half antibody (34E3; IBL Japan) (each at 2 μg/ml), which detects the thrombin-cleaved neoepitope YGLR, mouse anti–β-actin antibody (AC-15; Sigma, St. Louis, MO) at 1:10,000, or rabbit antifibronectin antibody (H-300; Santa Cruz Biotechnology, Santa Cruz, CA) at 1:200. Specificity of the 34E3 antibody was confirmed by Western blotting of thrombin-treated recombinant OPN. After reaction with horseradish peroxidase–conjugated anti-rabbit or anti-mouse IgG and enhanced chemiluminescence Western blotting detection reagents (all from Amersham), images of the membrane were captured using a FAS-1000 Lumino Imaging Analyzer (Toyobo, Osaka, Japan).
Total cell lysates were prepared for immunoprecipitation using RA FLS from a single patient (RA sample 4 in Figure 1A). When the cells had reached 90% confluence, they were collected in a 10-cm dish, and after centrifugation, the resulting supernatant was precleared with ImmunoPure Immobilized Protein A Plus (Pierce), divided into 3 centrifuge tubes, and incubated overnight at 4°C with protein A–Sepharose bound with 10 μg of the anti-OPN antibody (O-17), the antifibronectin antibody, or normal rabbit IgG (Santa Cruz Biotechnology) as a negative control. On the following day, the samples were washed and eluted with SDS sample buffer, and then analyzed by Western blotting.
FLS cultured in Chamber Slides (Nunc, Rochester, NY) were fixed in 4% paraformaldehyde, permeabilized with TBST, blocked with 5% BSA/TBST, and incubated overnight at 4°C with the anti-OPN antibody (O-17) or normal rabbit IgG (Santa Cruz Biotechnology) (each at 2 μg/ml). After the samples were reacted with Alexa Fluor 488–conjugated anti-rabbit IgG at 1:500 (Molecular Probes, Eugene, OR), the fluorescence intensity of the samples was examined under an epifluorescence microscope (Nikon Eclipse 90i; Nikon, Tokyo, Japan).
Transfection with siRNA.
Commercially available, predesigned double-stranded RNA (nontargeting siRNA [siCONTROL Pool #1] and siRNA for human secreted phosphoprotein 1 [OPN] and IL-6 [siGENOME SMARTpool siRNA]; Dharmacon, Lafayette, CO) were used for the RNA interference experiments. One microliter of X-TremeGENE reagent (Roche Diagnostic, Penzberg, Germany) and 20 pmoles of the siRNA from either pool were diluted in serum-free medium, and the mixture was then incubated for 15 minutes and added to 4 × 104 FLS in a 12-well plate. Two days after transfection, messenger RNA (mRNA) was collected and reverse-transcribed to first-strand complementary DNA (cDNA) using a high-capacity cDNA reverse transcription kit. Real-time reverse transcription–polymerase chain reaction (PCR) was then performed to evaluate the OPN-knockdown efficacy, using TaqMan gene expression assays (Applied Biosystems, Foster City, CA). Three days after transfection, the cocultures were started.
B lymphocyte adhesion assay.
Adhesion of B lymphocytes to FLS was assessed as described previously (11), with some modifications. Briefly, 6.6 × 103 FLS were plated into a 96-well plate. After 3 days, the cells were incubated at 37°C for 2 hours with a polyclonal OPN-neutralizing antibody or normal goat IgG (both from R&D Systems, Minneapolis, MN) at 10 μg/ml. Subsequently, 3.3 × 105 B lymphocytes were added and incubated at 37°C for another 2 hours. B lymphocytes that did not adhere firmly to the FLS were removed by vigorous washing. Three separate fields were viewed under an inverted microscope at a magnification of 100× per well to count the adherent B lymphocytes, with results expressed quantitatively as the mean number of adherent B lymphocytes.
FLS were plated at 4 × 104 cells in 12-well plates, followed by incubation with cystamine sulfate (500 μM) for 2 days to inhibit transglutaminase activity. Subsequently, total cell lysates were collected for further analysis.
Plasmid construction and lentiviral infection.
The full-length OPN-coding sequence was amplified by PCR from a human synovial cell cDNA library using the following primers: 5′-CCCTCGAGATGAGAATTGCAGTGATTTGC-3′ (NM_001040058, nt166–186) and 5′-CGGGATCCTTAATTGACCTCAGAAGATGCAC-3′ (NM_001040058, nt1088–1110). The PCR product was digested with Xho I and Bam HI, and then purified and ligated to pcDNA3.1 (Invitrogen, Carlsbad, CA) for sequencing. The vector was cut with Pme I and the OPN insert was obtained by gel extraction/purification, which was then ligated to the lentiviral expression vector pWPI. Lentivirus infection was performed on 4 × 104 FLS in a 12-well plate, as described previously (39, 40). Three days after infection, coculture was started for measurement of IL-6 production.
Specimens of synovial tissue obtained from a patient with RA were fixed in 10% formaldehyde, dehydrated, and embedded in paraffin. Sections of the tissue were cut (4 μm thick) on a microtome, and then were analyzed by immunohistochemistry, as described previously (41), using the following primary antibodies: rabbit anti-OPN antibody (O-17) or normal rabbit IgG at 2 μg/ml, mouse anti-CD79α antibody (JCB117; Nichirei, Tokyo, Japan) at 1:50, normal mouse IgG1 (R&D Systems), and goat anti-human IL-6 antibody or normal goat IgG at 2.5 μg/ml.
Results are expressed as the mean ± SD of at least 3 independent experiments, if not specified. For statistical comparison, 2-way factorial analysis of variance (ANOVA) (sometimes followed by a Bonferroni post hoc test) was performed using SPSS software (version 16.0; SPSS, Chicago, IL). P values less than 0.05 were considered significant.
Detection of specifically modified 75-kd OPN in all 10 RA FLS and 3 non-RA FLS, in conjunction with significantly higher IL-6 production in FLS–B lymphocyte cocultures.
Western blotting of the total cell lysates from FLS, using the antibody for the N-terminus of OPN (O-17), showed several bands at 75 kd and also around 54 kd (Figure 1A), which is a well-known characteristic of OPN (42). Double bands around 54 kd, each at an equal density, were detected in every FLS sample. Recombinant human OPN also migrated at around 54 kd. However, 75-kd OPN was predominantly detected in all 10 RA FLS samples (RA samples 1–10 in Figure 1A) and in only 3 non-RA FLS samples (non-RA samples 3, 8, and 9). Thrombin-cleaved OPN was not detected around the reported molecular weight of 30 kd (43).
EIA analysis of the full-length OPN secreted in the culture supernatant showed that the concentration of OPN was 9–18 ng/ml, and this did not differ between RA and non-RA FLS (results not shown). Since the expression of 75-kd OPN was distinct, FLS were grouped into 75-kd OPN–positive and 75-kd OPN–negative FLS.
Production of IL-6 was evaluated by EIA for determination of the IL-6 concentration in the culture supernatants obtained in several culture conditions with FLS and B lymphocytes. In FLS monoculture (i.e., FLS cultured alone), the mean ± SD IL-6 production by the 75-kd OPN–positive FLS (n = 13) and the 75-kd OPN–negative FLS (n = 7) was 1,113.3 ± 818.9 pg/ml and 643.1 ± 576.5 pg/ml, respectively. In FLS–B lymphocyte cocultures without contact (i.e., FLS cocultured with B lymphocytes, but separated by the Millicell culture insert), the IL-6 levels were 1,363.3 ± 714.6 pg/ml in 75-kd OPN–positive FLS and 902.6 ± 772.5 pg/ml in 75-kd OPN–negative FLS. There were no significant differences in IL-6 production among these 4 groups.
However, in cocultures with contact (i.e., FLS and B lymphocytes cultured in contact with each other), IL-6 production was significantly elevated in 75-kd OPN–positive FLS, and showed a tendency to be elevated, although not to a significant extent, in 75-kd OPN–negative FLS. The IL-6 production by the 13 FLS positive for 75-kd OPN cocultured in contact with B lymphocytes was elevated to 19,928.5 ± 10,351.8 pg/ml, which was significantly higher than that in the other 2 culture conditions (P < 0.001, by Bonferroni t-test) and significantly higher than that in the 7 FLS negative for 75-kd OPN that were in contact coculture with B lymphocytes (910.9 ± 744.1 pg/ml; P < 0.001, by Bonferroni t-test) (Figure 1B).
Transfection of the FLS with IL-6 siRNA significantly reduced the levels of IL-6 in the 75-kd OPN–positive FLS that were in contact coculture with B lymphocytes. In contrast, when B lymphocytes were cultured alone, IL-6 was not detected in the culture supernatants (results not shown).
Production of >200-kd thrombin-cleaved OPN in the cell surface protein fraction of 75-kd OPN–positive FLS, and cell membrane distribution of OPN.
Analysis of cell surface proteins from FLS, using Western blotting with an antibody against the thrombin-cleaved neoepitope of OPN (34E3), detected the presence of >200-kd OPN in the surface fraction of 75-kd OPN–positive FLS. This band was poorly detected in 75-kd OPN–negative FLS or in the nonsurface fraction of FLS (Figure 2A). A similar result was obtained with the O-17 antibody, but the signals were much weaker (results not shown). As a loading control, Western blotting with the antifibronectin antibody probe was performed. When the anti-OPN antibody O-17 was used to analyze the 75-kd OPN–positive FLS by immunofluorescence assay, the distribution of OPN followed the cell outline, regardless of permeabilization (Figure 2B).
Detection of >200-kd OPN in the immunoprecipitant by antifibronectin antibody, and reduction by transglutaminase inhibitor treatment.
According to a previous study, OPN appears at >200 kd when it is covalently crosslinked to fibronectin by transglutamination through 2 widely conserved glutamine residues at its N-terminus (26). Therefore, immunoprecipitation of total cell lysates from 75-kd OPN–positive FLS with the anti-OPN or antifibronectin antibody was performed, and this was followed by Western blotting with these antibodies. The >200-kd band was detected when the immunoprecipitant obtained with the antifibronectin antibody was probed by the anti-OPN antibody, and vice versa (Figure 2C). A transglutaminase inhibitor, cystamine sulfate, reduced the expression of >200-kd OPN when the inhibitor was added to the FLS cultures, but did not alter the levels of 75-kd or 54-kd OPN (Figure 2D).
Enhanced IL-6 production by OPN overexpression in 75-kd OPN–positive FLS, but not in 75-kd OPN–negative FLS, in cocultures with B lymphocytes.
To elucidate the role of OPN in FLS–B lymphocyte cocultures, gain of function experiments were performed by inducing overexpression of OPN with a lentiviral vector. Western blotting of the cells after lentiviral infection showed that all 75-kd OPN–positive FLS showed up-regulation of the 75-kd, >200-kd, and 54-kd OPN bands, along with an extra band at 50 kd, whereas the 75-kd and >200-kd bands were not up-regulated in any of the 75-kd OPN–negative FLS (Figure 3A).
In addition, all 75-kd OPN–positive FLS showed a significant increase in IL-6 production after OPN overexpression followed by coculture in contact with B lymphocytes, when compared with cells without overexpression (P < 0.001 by 2-factorial ANOVA) (Figure 3B). In 75-kd OPN–negative FLS, however, IL-6 production was not altered by overexpression of OPN, which was consistent with the lack of a 75-kd band and the unaltered >200-kd band on Western blotting.
Significant reduction in IL-6 levels following loss of function of OPN in 75-kd OPN–positive FLS, and inhibition of adhesion of B lymphocytes on FLS in cocultures with B lymphocytes.
Loss of function experiments were performed using siRNA transfection and an OPN-neutralizing antibody. Transfection of OPN siRNA successfully knocked down OPN expression at the mRNA level as well as at the protein level in all of the 75-kd OPN–positive FLS (Figure 4A). After subsequent coculture of the FLS in contact with B lymphocytes, there was significantly lower IL-6 production by OPN–knocked down FLS compared with that by FLS transfected with nontargeting siRNA (P < 0.001 by 2-factorial ANOVA) (Figure 4B).
Adhesion of B lymphocytes to FLS occurred in FLS–B lymphocyte cocultures, as has been reported previously (9, 12), but was inhibited by the blocking antibody directed against OPN. The number of B lymphocytes adherent to FLS was significantly reduced by the anti-OPN antibody in all 75-kd OPN–positive FLS (P < 0.001 by 2-factorial ANOVA) (Figure 4C).
Detection of IL-6–positive cells in the sublining region of RA synovium, at the site of OPN–B lymphocyte colocalization.
Finally, to assess the distribution of OPN, IL-6, and B lymphocytes in the RA synovium, immunohistochemical analyses were carried out using anti–IL-6, anti-OPN, and anti–pan B lymphocyte antigen CD79α antibodies. One of the synovial villi located in the knee joint from an RA patient was studied. As reported previously, OPN was distributed in the fibroblastic cells and the matrix of the synovial lining and sublining regions (34, 35). Clusters of B lymphocytes were found in the deeper layers of the synovium as well as in the sublining region, and some of these B lymphocyte clusters were surrounded by OPN-positive cells. IL-6–positive cells were scattered in the sublining region, as has been reported previously (44).
Of note, some of the areas of distribution of IL-6–positive cells coincided with sites at which OPN and B lymphocytes colocalized (indicated by arrowheads in Figure 5). However, IL-6 was not detected in areas where only OPN-positive cells were distributed or at sites far from clusters of B lymphocytes (indicated by arrows in Figure 5).
The present study revealed a new role of OPN in the RA synovium, using primary cultured FLS derived from RA patients to analyze the function of FLS-expressed native OPN. A specifically modified 75-kd OPN was predominantly expressed by RA FLS and was associated with significantly elevated IL-6 production in FLS–B lymphocyte cocultures. The 75-kd OPN formed a >200-kd OPN/fibronectin–crosslinked molecule via transglutamination, which was detected on the synovial cell surface, resulting in exposure of its thrombin-cleaved neoepitope. This >200-kd cell surface OPN affected the interactions between FLS and B lymphocytes by supporting adhesion of B lymphocytes to FLS in FLS–B lymphocyte cocultures, and consequently IL-6 production was enhanced. This FLS–B lymphocyte interaction also appeared to occur in vivo.
OPN by itself weighs 37 kd, as determined previously using mass spectrometry (25), but due to its various posttranslational modifications, the migration of full-length OPN, detected on SDS-PAGE, differs among cell types within ∼45–80 kd (42). Accordingly, our results from Western blotting of OPN showed multiple bands with different molecular weights. In particular, a single band at 75 kd that was detected predominantly in RA FLS suggested that this specifically modified 75-kd OPN was worth investigating to determine the function of OPN in RA.
In fact, the localization of OPN was different between 75-kd OPN–negative and 75-kd OPN–positive FLS. The existence of >200-kd OPN in the cell surface fraction of 75-kd OPN–positive FLS on Western blotting, and the staining pattern of OPN on immunofluorescence assay of these FLS, showing the cell outline distribution, suggested that OPN was associated with the cell surface side of the plasma membrane. These distinct findings were considered attributable to 75-kd OPN, since they were negative in 75-kd OPN–negative FLS. Such posttranslational modification–dependent differences in subcellular localization of OPN have also been observed in previously published studies of OPN expressed by normal rat kidney (NRK) cells, which revealed that among phosphorylated and nonphosphorylated forms of OPN expressed by NRK cells, only phosphorylated OPN was associated with the cell surface (19, 23). Moreover, another study revealed that OPN was detected at >200 kd molecular weight and found to be localized on the surface of NRK cells when covalently crosslinked to fibronectin by a transglutaminase (26). Our results were consistent with that study, since the observations from immunoprecipitation and Western blotting of the FLS suggested that >200-kd OPN was an OPN/fibronectin-crosslinked molecule, and inhibition of a transglutaminase reduced the expression of >200-kd OPN. Considering these facts together, >200-kd OPN could be considered an OPN/fibronectin–covalently crosslinked molecule synthesized by a transglutaminase from 75-kd OPN.
In contrast, >200-kd cell surface OPN was poorly detected in 75-kd OPN–negative FLS. Moreover, overexpression of OPN by 75-kd OPN–negative FLS only increased the expression of 54-kd OPN, and did not induce expression of 75-kd OPN nor did it alter the expression of >200-kd OPN, suggesting that there might be an unknown enzyme in 75-kd OPN–positive FLS that performs the specific modifications of 75-kd OPN necessary for it to associate with the cell surface and to form the >200-kd OPN/fibronectin–crosslinked molecule.
To assess the relevance of 75-kd OPN in RA, we performed cocultures of FLS and B lymphocytes. Previous studies have revealed that FLS–B lymphocyte coculture allows the adhesion of B lymphocytes to FLS, which induces cell–cell interactions between FLS and B lymphocytes and, consequently, increases the production of several cytokines, including IL-6 (9, 12), but the mechanism has not been elucidated. We focused on IL-6 production in FLS–B lymphocyte cocultures and found that increased IL-6 production was associated with the existence of 75-kd OPN, which suggested that 75-kd OPN together with cell–cell interactions between FLS and B lymphocytes can enhance IL-6 production. We also revealed that among FLS and B lymphocytes, FLS were the dominant cell type for IL-6 production in coculture, as was shown in experiments involving knockdown of IL-6 in FLS.
To examine whether 75-kd OPN could enhance IL-6 production, we performed overexpression and knockdown of OPN in FLS, which demonstrated that IL-6 production was increased and decreased in accordance with the increase and decrease of 75-kd OPN and >200-kd OPN, respectively. Among these 2 OPN forms with different molecular weights, >200-kd OPN appeared to enhance IL-6 production, since the transglutaminase inhibitor that reduced the expression of only the >200-kd OPN significantly suppressed IL-6 production in 75-kd OPN–positive FLS–B lymphocyte cocultures (results not shown).
We then analyzed how 75-kd OPN or its crosslinked form, >200-kd OPN, enhanced IL-6 production in FLS–B lymphocyte cocultures. The >200-kd OPN on the cell surface exposed its thrombin-cleaved neoepitope, SVVYGLR (27), a ligand for integrin α4β1, also known as VLA-4, integrin α9β1, and integrin α4β7 (30). Considering the fact that VLA-4 is expressed by B lymphocytes and supports adhesion of B lymphocytes to FLS in FLS–B lymphocyte coculture (12), >200-kd OPN was indicated as the ligand for VLA-4, which acts as an adhesion molecule. This idea was supported by the results from our B lymphocyte adhesion assay with a blocking antibody against OPN, and by the fact that the transglutaminase inhibitor also suppressed adhesion of B lymphocytes to FLS (results not shown). Meanwhile, integrin α4β7, which was also expressed by B lymphocytes, did not mediate such adhesion (12).
VCAM-1, which was expressed by FLS and also supports B lymphocyte adhesion, possibly through VLA-4 (11), was detected in equal amounts among RA and non-RA FLS (results not shown). Therefore, we postulated that integrin α4β7 and VCAM-1 were not involved in the adhesion of B lymphocytes to FLS or subsequent IL-6 production, and that such adhesion was mediated by >200-kd OPN and VLA-4, which would further initiate the cell–cell interaction between FLS and B lymphocytes, leading to IL-6 production.
Our findings suggested that, in response to these FLS–B lymphocyte interactions, FLS boosted their IL-6 production. Similar studies on cell–cell interactions between FLS and T lymphocytes have shown that lymphocyte function–associated antigen 1, intercellular adhesion molecule 2, and the ezrin/Akt pathway are involved in their interaction, which also enhances IL-6 production (2). Thus, this pathway may also be involved in FLS–B lymphocyte interactions, although further investigation is needed to fully elucidate the mechanism.
In summary, the findings from these in vitro experiments showed that RA FLS are characterized by the expression of 75-kd OPN. This was the substrate of a transglutaminase that formed the >200-kd OPN/fibronectin–crosslinked molecule, a molecule localized on the surface of FLS in its thrombin-cleaved form. This surface OPN mediated cell–cell interactions between FLS and B lymphocytes, and enhanced IL-6 production (Figure 6). Moreover, such FLS–B lymphocyte interactions, or interactions between FLS-expressing OPN and B lymphocytes stimulating IL-6 production, appeared to take place in vivo, as shown by immunohistochemistry. Taking into account the previously reported findings on the pathogenic significance of IL-6 (45), B lymphocytes (46, 47), VLA-4 (48), and the thrombin-cleaved neoepitope of OPN (32, 33) in RA, the present study revealed a novel role of OPN in RA.
Regarding the weaknesses of this study, we need to mention the discrepancy in the findings between RA FLS and 75-kd OPN–positive FLS. Since 75-kd OPN was detected in all 10 RA FLS, but also in 3 non-RA FLS, it was not specific to RA FLS. There is no doubt that 75-kd OPN enhanced IL-6 production in FLS–B lymphocyte cocultures and appeared to aggravate chronic inflammation in vivo. However, it should not be considered the cause of RA; rather, we could postulate that it is one of several molecules with induced expression in arthritis and is involved in the chronic progression of arthritis by stimulating IL-6 production. This is supported by the observation that, in the joint tissue of a non-RA donor, the FLS expressed amounts of 75-kd OPN comparable with those in RA FLS, and showed severe synovitis at the time of surgery.
We also have to note that the mechanism involved in the modification of 75-kd OPN has not been defined, and the responsible transglutaminase has not been identified. Further investigations in this area would be required for better understanding of the pathology of RA. Nevertheless, our results show that a specifically modified 75-kd form of OPN was expressed by RA FLS. This form of OPN affected cell–cell interactions between FLS and B lymphocytes by supporting the adhesion of B lymphocytes to FLS. As a result, IL-6 production was enhanced in FLS–B lymphocyte cocultures.
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. Nakata 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. Nakata, Ochi, Yoshikawa.
Acquisition of data. Take, Nakata, Nishimoto.
Analysis and interpretation of data. Take, Nakata, Hashimoto, Tsuboi, Nishimoto.