Induction of Osteoclast-Associated Receptor, a Key Osteoclast Costimulation Molecule, in Rheumatoid Arthritis

Transcription

1 ARTHRITIS & RHEUMATISM Vol. 58, No. 10, October 2008, pp DOI /art , American College of Rheumatology Induction of Osteoclast-Associated Receptor, a Key Osteoclast Costimulation Molecule, in Rheumatoid Arthritis Sonja Herman, 1 Ruediger B. Müller, 1 Gerhard Krönke, 1 Jochen Zwerina, 1 Kurt Redlich, 2 Axel J. Hueber, 1 Holger Gelse, 1 Elena Neumann, 3 Ulf Müller-Ladner, 3 and Georg Schett 4 Supported by the Austrian Science Fund Start Prize (SFB 643), the Interdisciplinary Center for Clinical Research at the University of Erlangen (IZKF projects C6 and C7), and Wyeth (grant 0881x1-4448). Dr. Schett s work was supported by the Doerfler- Stiftung. 1 Sonja Herman, MD, Ruediger B. Müller, MD, Gerhard Krönke, MD, Jochen Zwerina, MD, Axel J. Hueber, MD, Holger Gelse, MD: University of Erlangen Nuremberg, Erlangen, Germany; 2 Kurt Redlich, MD: Medical University of Vienna, Vienna, Austria; 3 Elena Neumann, PhD, Ulf Müller-Ladner, MD: Justus-Liebig- University of Giessen, Bad Nauheim, Germany; 4 Georg Schett, MD: University of Erlangen Nuremberg, Erlangen, Germany, and Medical University of Vienna, Vienna, Austria. Address correspondence and reprint requests to Georg Schett, MD, Department of Internal Medicine III and Institute for Clinical Immunology, University of Erlangen Nuremberg, Krankenhausstrasse 12, D Erlangen, Germany. georg.schett@ uk-erlangen.de. Submitted for publication December 12, 2007; accepted in revised form June 27, Objective. Osteoclast-associated receptor (OSCAR) is a newly identified osteoclast-specific receptor and is of key importance in the process of osteoclast costimulation. This study was undertaken to define the role of costimulation in osteoclast differentiation during inflammatory arthritis. Methods. OSCAR expression was assessed in the synovium and peripheral blood monocytes of patients with rheumatoid arthritis (RA), and associations with disease activity were assessed. Serum levels of OSCAR were determined by enzyme-linked immunosorbent assay. In vitro osteoclast assays were performed to characterize the role of OSCAR in human osteoclastogenesis. Cytokine regulation of OSCAR was investigated by reverse transcriptase polymerase chain reaction. Results. OSCAR was expressed by osteoclasts at the erosion front and by mononuclear cells around synovial microvessels. Flow cytometry revealed enhanced expression of OSCAR in peripheral blood monocytes of RA patients as compared with healthy controls. OSCAR expression was correlated with disease activity and acute-phase reactant concentrations. Serum levels of soluble OSCAR were lower in RA patients than in healthy controls. Monocytes with high OSCAR expression exhibited an enhanced potential to differentiate into osteoclasts. Tumor necrosis factor was identified as the main inducer of OSCAR expression in monocytes. Conclusion. These data suggest that the osteoclast costimulation pathway is activated in RA. OSCAR is induced in monocytes of RA patients, facilitating their differentiation into osteoclasts and bone resorption. Rheumatoid arthritis (RA) is characterized by profound bone loss. A hallmark of the disease is the destruction of periarticular bone, which leads to bone erosions and functional disability (1). Osteoclast formation is an essential step in inflammatory bone erosion, and it appears that the tight balance between bone resorption and bone formation is disturbed in RA (2 4): osteoclast formation is enhanced and is not balanced by increased activity of bone-forming osteoblasts. This unfavorable balance between bone resorption and bone formation is the basis of rapid bone loss in the joints, which breaks down the physiologic joint architecture and contributes to the characteristic clinical features of RA. Increasing evidence suggests that differentiation of osteoclasts occurs locally in the inflamed synovial membrane, where osteoclasts form from monocytic precursor cells. Cells of the monocyte/macrophage lineage are abundant in the inflamed synovium and constitute a rich cellular source for excessive osteoclast formation in the joint. Current evidence suggests that these cells are directed to the osteoclast lineage upon contact with 3041

2 3042 HERMAN ET AL cytokines such as macrophage colony-stimulating factor (M-CSF) and RANKL, which are locally expressed in the synovial tissue and are controlled by inflammatory cytokines (4 8). Although there is no doubt that the local cytokine milieu in the inflamed joint contributes to enhanced osteoclast formation, this mechanism may also be supported by increased susceptibility of the target cell to undergo differentiation into the osteoclast lineage. Thus, cells may be already primed, in the bloodstream, to be more susceptible to differentiate into osteoclasts before they enter the joint (9). Osteoclast-associated receptor (OSCAR) has been recently described as a key molecule in osteoclastogenesis (10). OSCAR is a member of the leukocyte receptor complex and associates with immunoreceptor tyrosine based activation motifs (ITAMs) on the common -chain of the Fc receptor. Thus, OSCAR is part of the costimulation axis of osteoclasts, which acts independently from cytokine stimulation and reflects the fact that osteoclasts as hematopoietic cells are also regulated by costimulation pathways. OSCAR is most prominently expressed on osteoclasts but can also be found in monocytes, which serve as osteoclast precursors, and dendritic cells, which share a common precursor with osteoclasts (11,12). Activation of OSCAR by stimulatory antibodies triggers an activated phenotype, prolongs survival, and increases the production of chemokines such as monocyte chemoattractant protein 1 (MCP-1), which is involved in the recruitment of leukocytes to sites of inflammation (12). Although a single-nucleotide polymorphism within the OSCAR gene is associated with increased risk of postmenopausal osteoporosis (13), little is known about the role of OSCAR in human inflammatory diseases. We hypothesized that OSCAR expression might be altered in the monocytes of RA patients as compared with healthy individuals, thus providing a molecular basis for increased susceptibility of monocytes to differentiate into the osteoclast lineage. We therefore investigated expression of OSCAR in the synovial tissue of RA patients and performed a comparative analysis of OSCAR expression in monocytes from peripheral blood of healthy individuals and patients with RA. In addition, we developed an assay to measure soluble OSCAR (soscar) in the serum of healthy subjects and RA patients. PATIENTS AND METHODS Patients. RA patients were selected at random from our outpatient clinic. All patients fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) classification criteria for RA (14), had a disease duration of 1 year, and had active disease treated with tumor necrosis factor (TNF ) blocker therapy. Age- and sexmatched healthy volunteers served as controls. A total of ml of whole blood was collected by venipuncture for routine laboratory investigations. In all RA patients, disease activity was measured with the Disease Activity Score 28-joint assessment (DAS28) (15). All procedures were approved by the local ethics committee, and all participants provided written informed consent. Flow cytometry. EDTA-treated blood samples were used for isolation of peripheral blood mononuclear cells (PBMCs) by standard gradient centrifugation with Ficoll- Hypaque (Biotest, Dreieich, Germany). PBMCs were washed twice with phosphate buffered saline (PBS) and used immediately for 3-color labeling. Cells were then incubated with biotinylated anti-human OSCAR monoclonal antibody (mab; R&D Systems, Minneapolis, MN) followed by streptavidin phycoerythrin (PE) conjugate (Becton Dickinson, Heidelberg, Germany). Subsequently, cells were labeled with fluorescein isothiocyanate conjugated anti-cd14 mab, PE-conjugated anti-cd19 mab, and peridin chlorophyll protein conjugated anti-cd3 (all from Becton Dickinson). In all experiments, control mab of the same IgG isotype was included. CD14, CD3, and CD19 cells were gated. Human osteoclast assay. Human PBMCs were obtained by standard gradient centrifugation with Ficoll- Hypaque (Biotest). Cells were then plated in 96-well plates ( cells/well) in the presence of 25 ng/ml M-CSF, 50 ng/ml RANKL (both from PeproTech, Hamburg, Germany), and 5 ng/ml transforming growth factor 1 (R&D Systems, Minneapolis, MN) in minimum essential medium (Gibco, Karlsruhe, Germany) that had been supplemented with 10% heat-inactivated fetal calf serum (FCS; Gibco) and 1% penicillin/streptomycin (R&D Systems, Wiesbaden, Germany) for 17 days. Additionally, recombinant human OSCAR fused to the human IgG constant region (OSCAR-Fc; a gift from Estelle Merck, Paris, France) was added in various concentrations (1 ng/ml, 10 ng/ml, or 100 ng/ml). Osteoclast differentiation was evaluated by staining for tartrate-resistant acid phosphatase (TRAP) using a leukocyte acid phosphatase kit (Sigma-Aldrich, Poole, UK). Osteoclasts were identified by the presence of 3 nuclei and purple color; osteoclast precursor cells were identified as TRAP-positive cells with 1 or 2 nuclei. Serum measurements. Serum from the same RA patients was stored at 80 C until the time of analysis. Soluble OSCAR was measured by coating microtiter plates with 1 g/ml anti-oscar mab (R&D Systems, Wiesbaden, Germany) before addition of human serum samples. Detection was performed by adding biotinylated polyclonal anti-human OSCAR antibody (1:100 [volume/volume]; R&D Systems, Wiesbaden, Germany). Enzyme-linked immunosorbent assay (ELISA) for soscar was performed with serum samples obtained from RA patients (n 10) before and after initiation of anti-tnf therapy. Immunoblotting. Synovial tissue extracts from 10 patients with RA undergoing joint replacement were used. All samples were collected by synovectomy and immediately fro-

3 OSCAR IN RHEUMATOID ARTHRITIS 3043 Table 1. Patient Characteristics of the rheumatoid arthritis patients studied* Age/sex Disease duration, years DAS28 HAQ score CRP, mg/liter ESR, mm/hour Treatment Radiographic erosions 1 28/F Inflix. 2 68/F Etan. 3 57/F Inflix. 4 73/F Inflix., MTX 5 48/F MTX 6 37/F Inflix. LEF 7 57/M Ada., MTX 8 59/F Inflix., MTX 9 62/F Ada., HCQ 10 66/F MTX 11 58/F Ada /F Inflix., MTX 13 47/F Etan /F Inflix., MTX 15 49/F MTX 16 57/M Inflix., MTX 17 65/M Ada. * DAS28 Disease Activity Score 28-joint assessment; HAQ Health Assessment Questionnaire; CRP C-reactive protein (normal 5 mg/liter); ESR erythrocyte sedimentation rate (normal 10 mm/hour); Inflix. infliximab; Etan. etanercept; MTX methotrexate; LEF leflunomide; Ada. adalimumab; HCQ hydroxychloroquine. zen in liquid nitrogen. Frozen synovial tissue samples were mechanically homogenized at 4 C in buffer containing 20 mm HEPES, 0.4M NaCl, 1.5 mm MgCl 2, 1 mm dithiothreitol, 1mM EDTA, 0.1 mm EGTA, 20% glycerol, and protease and phosphatase inhibitors (both from Sigma, St. Louis, MO), using an Ultra-Thurrax homogenizer. Tissue extracts were then separated from debris and fat by centrifuging for 15 minutes at 14,000 rpm. Protein content was measured by Bradford assay, and 50 ng/ l tissue protein was subjected to electrophoresis on 10% sodium dodecyl sulfate polyacrylamide gels, followed by transfer to nitrocellulose membrane. After blocking, the membranes were incubated with anti-human OSCAR as primary antibody. Horseradish peroxidase conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz, CA) was used for protein detection. Immunohistochemistry. Cryosections (4 m) were prepared from synovial tissue derived from RA and osteoarthritis (OA) patients. Sections were fixed with acetone for 10 minutes, before blocking of endogenous peroxidase with 3% hydrogen peroxide in PBS. Nonspecific binding was blocked by addition of 10% goat serum or 10% rabbit serum for 10 minutes at room temperature. Sections were then incubated with a polyclonal anti-human OSCAR antibody overnight at 4 C. The sections were then incubated with species-specific biotinylated immunoglobulins (Vector, Burlingame, CA) for 30 minutes at room temperature before detection with Vectastain ABC (Vector) and 3,3 -diaminobenzidine (Sigma, St. Louis, MO) as a chromogen, resulting in maroon staining of antigen-expressing cells. Counterstaining with hemalum was performed at the end of the procedure. Quantitative real-time PCR (polymerase chain reaction) analysis. Quantitative reverse transcriptase PCR was performed using SYBR Green as previously described (16). The expression of the target molecule was normalized to the expression of -actin (forward 5 -GAT-GAG-TAT-GCC-TG- C-CGT-GTG-3, reverse 5 -CAA-TCC-AAA-TGC-GGC-AT- C-T-3 ). The primers 5 -AAT-GGA-CCA-ATC-AGC-AGG- AC-3 (forward) and 5 -GAG-AAC-AAA-GCT-CCC-ACA- GC-3 (reverse) were used to amplify a fragment specific to OSCAR. Cells were stimulated for 48 hours with M-CSF, RANKL, or TNF or left unstimulated in Dulbecco s modified Eagle s medium supplemented with 1% FCS. Statistical analysis. Student s t-test was used to determine the significance of differences in results. P values less than 0.05 were considered significant. RESULTS Patient characteristics. The characteristics of the RA patients are summarized in Table 1. Fourteen patients were female and 3 were male. The mean age was 56 years, and the mean disease duration was 13.5 years, indicating longstanding RA. The mean erythrocyte sedimentation rate (ESR) was 29.3 mm/hour, and the mean DAS28 score was 5.0, indicating active disease. All patients were receiving disease-modifying antirheumatic drug and/or TNF blocker therapy. OSCAR-expressing cells in the synovial tissue of RA patients. Given the importance of OSCAR as a costimulation molecule in murine osteoclast differentiation and function (10,17), we investigated expression of OSCAR in the synovial tissue of RA patients, by immu-

4 3044 HERMAN ET AL Figure 1. Expression of osteoclast-associated receptor (OSCAR) in the synovium of patients with rheumatoid arthritis (RA). a, Immunoblotting of synovial tissue extracts from patients with RA and labeling with an antibody against human OSCAR. b, Immunohistochemistry examination for OSCAR in sections from metacarpal heads of patients with RA or with finger osteoarthritis (OA). Maroon color indicates positive staining at bone-resorbing multinucleated osteoclasts at the resorption front in the RA specimen (inset shows a highermagnification view). In contrast, no OSCAR expression is seen in the OA specimen (original magnification 100; 400 in inset). c, Immunohistochemistry examination of RA synovial tissue, showing mononuclear cells around synovial microvessels expressing OSCAR, as well as multinucleated osteoclasts at sites of resorption pits resembling small bone erosions, which also express OSCAR (original magnification 200). BV blood vessel. Graph shows quantification of the percentage of OSCAR-positive and OSCAR-negative synovial mononuclear cells (mean and SD). noblot and immunohistochemical analysis. Indeed, on immunoblot analysis of synovial tissue extracts, all tissue probes exhibited strong expression of OSCAR (Figure 1a). To better characterize the expression pattern of OSCAR within the synovium, we stained specimens obtained from RA and OA patients during total joint replacement surgery. Whereas we were unable to detect OSCAR-expressing cells in OA tissue, RA tissue exhibited expression of OSCAR by multinucleated osteoclasts attached to the bone surface (Figure 1b). Moreover, we detected OSCAR expression in mononuclear cells close to synovial microvessels. Approximately 30% of mononuclear cells in the synovial tissue stained positive for OSCAR (Figure 1c). High-level expression of OSCAR by circulating mononuclear cells in RA. Having observed OSCARexpressing mononuclear and multinucleated cells in the synovial membrane, we hypothesized that these cells derive from circulating PBMCs, which already express OSCAR. We therefore investigated the presence of OSCAR-expressing cells in the peripheral blood of patients with RA. The mean fluorescence intensity (MFI) of OSCAR was examined in freshly isolated monocytes. Expression was confined to peripheral blood monocytes, whereas it was not detectable on lymphocytes (Figure 2a). The MFI of OSCAR on the surface of monocytes from RA patients was significantly higher than in cells from healthy individuals (P ). Overall, RA patients showed a 2-fold increase in OSCAR expression (mean SD MFI ) as compared with healthy controls ( ) (Figure 2b), suggesting that up-regulation of OSCAR has already occurred before entry of cells into the synovium. Correlation between OSCAR expression on monocytes and RA disease activity. We next investigated whether the expression of OSCAR is related to levels of systemic inflammation in RA. Patients were divided into those with elevated C-reactive protein (CRP) levels ( 5 mg/liter) and those with normal levels ( 5 mg/liter). Interestingly, OSCAR expression was significantly higher in patients with elevated CRP levels as compared with those with normal levels (P 0.01) (Figure 2c). Similar results were obtained when ESR was used as a measure of systemic inflammation: patients with an ESR of 20 mm/hour exhibited significantly higher OSCAR expression on peripheral blood monocytes than was observed in those with an ESR of 20 mm/hour (P 0.05) (Figure 2c). Furthermore, OSCAR expression was related to clinical RA activity: patients with high OSCAR expression (defined as an MFI more than 2 SD above the mean in healthy donors) had significantly higher DAS28 scores than those with low OSCAR expression (P 0.05) (Figure 2c). These data suggest that surface expression of OSCAR is related to disease activity in RA. Low levels of soscar in the serum of RA patients, and inverse relationship with disease activity. Since shedding of surface molecules is a commonly observed regulatory phenomenon, we developed a sand-

5 OSCAR IN RHEUMATOID ARTHRITIS 3045 Figure 2. Enhanced expression of OSCAR in circulating monocytes from RA patients, and influence of inflammatory disease activity. a, Flow cytometric analysis of freshly isolated mononuclear cells from healthy donors and patients with RA. OSCAR was expressed on monocytes, but not on lymphocytes. SSC side scatter; FSC forward scatter. b, Mean fluorescence intensity (MFI) of OSCAR expression in circulating peripheral blood monocytes from RA patients and healthy donors. c, MFI of OSCAR expression in peripheral blood monocytes from RA patients with low C-reactive protein (CRP) levels versus those with high CRP levels, and from RA patients with low erythrocyte sedimentation rates (ESRs) versus those with high ESRs, and Disease Activity Score 28-joint assessment (DAS28) in RA patients with high OSCAR expression levels (MFI more than 2 SD above the mean in healthy controls) and those with low OSCAR expression levels. Values in b and c are the mean and SD. See Figure 1 for other definitions.

6 3046 HERMAN ET AL disease activity as observed with the membrane-bound protein. Influence of OSCAR on human osteoclast formation in vitro. Based on the observation that monocytes from RA patients show enhanced surface expression of Figure 3. Analysis of soluble OSCAR (soscar) in serum. a, Serum levels of soscar protein in patients with RA and healthy controls. Bars show the mean. b, Disease Activity Score 28-joint assessment (DAS28) in RA patients with low serum soscar levels ( 10 ng/ ml [2 SD below the mean in healthy controls]) and those with high serum soscar levels. Values are the mean and SD. See Figure 1 for other definitions. wich ELISA to investigate the presence of soscar in serum. OSCAR was found in the serum of both healthy individuals and patients with RA. However, serum levels were significantly reduced in RA patients (mean SD ng/ml, versus ng/ml in healthy controls; P ) (Figure 3a). We then investigated whether the level of soscar in serum is related to RA disease activity. RA patients with low serum soscar levels ( 10 ng/ml [2 SD below the mean in healthy subjects]) had significantly greater disease activity than did RA patients with soscar levels above 10 ng/ml (mean SD DAS versus ; P 0.05) (Figure 3b). This inverse relationship between disease activity and soscar level suggests that soscar might have a regulatory role, rather than representing a parameter of Figure 4. Influence of OSCAR on in vitro human osteoclast (OC) formation. a, Results of osteoclast differentiation assays in 3 RA patients with high surface expression of OSCAR and from 3 patients with low OSCAR expression. Graphs show the mean fluorescence intensity (MFI) in each individual patient, and the mean and SD number of osteoclasts per well. Osteoclasts appeared as tartrateresistant acid phosphatase (TRAP) positive multinucleated cells ( 3 nuclei). P 0.05 versus patients with high expression of OSCAR. Photomicrographs show TRAP-stained osteoclasts in representative high OSCAR-expressing and low OSCAR-expressing specimens (original magnification 20). b, Results of osteoclast differentiation assays with various concentrations of recombinant human OSCAR-Fc. Dose-dependent inhibition of osteoclast and osteoclast precursor (OCP) formation with addition of OSCAR-Fc is evident. Values are the mean and SD. P 0.05; P 0.01, versus no added OSCAR-Fc. Photomicrographs show TRAP-stained osteoclasts and osteoclast precursors in representative specimens with addition of OSCAR-Fc in various concentrations (original magnification 20). See Figure 1 for other definitions. Color figure can be viewed in the online issue, which is available at

7 OSCAR IN RHEUMATOID ARTHRITIS 3047 Induction of OSCAR expression in peripheral blood monocytes by TNF, and increase in soscar levels with inhibition of TNF. Since OSCAR protein is increased on monocytes from RA patients compared with healthy individuals, and since expression correlates with inflammatory disease activity, we speculated that OSCAR expression might be regulated by proinflammatory cytokines. Therefore, CD14 monocytes were challenged for 48 hours with TNF, RANKL, and M-CSF. TNF challenge led to a 4-fold increase of OSCAR messenger RNA expression as compared with that observed in unstimulated cells (P 0.01), as shown in Figure 5a. In contrast, M-CSF and RANKL did not affect OSCAR expression. To further establish the role of TNF, we measured soscar in the serum of RA patients before and 3 months after initiation of TNF blocker therapy. Interestingly, we found a significant increase in soscar levels after systemic blockade of TNF (mean SEM ng/ml, versus ng/ml before initiation of treatment; P 0.05) (Figure 5b). Figure 5. Regulation of OSCAR expression. a, Quantification of the results of reverse transcriptase polymerase chain reaction analysis of OSCAR in cultivated CD14 RA monocytes challenged with medium only or with 10 ng/ml tumor necrosis factor (TNF ), macrophage colony-stimulating factor (M-CSF), or RANKL. TNF caused a 4-fold increase in OSCAR mrna levels, whereas M-CSF and RANKL did not show any effect. b, Serum levels of soluble OSCAR (soscar) protein in RA patients before and 3 months after initiation of TNF blocker therapy. Values are the mean and SEM. See Figure 1 for other definitions. OSCAR, we were interested in whether this leads to enhanced osteoclastogenesis. We therefore performed osteoclast differentiation assays comparing patients with high and low surface expression of OSCAR. Differentiation of osteoclasts was significantly higher in PBMC cultures from individuals with high OSCAR surface expression (mean SD osteoclasts per well) than in those with low expression (48 17 osteoclasts per well) (P 0.05) (Figure 4a). Moreover, addition of recombinant human OSCAR-Fc reduced osteoclast formation, in a dosedependent manner. The formation of TRAP-positive multinucleated cells, as well as the number of TRAPpositive mononuclear osteoclast precursor cells, were both dose-dependently decreased with addition of OSCAR-Fc (Figure 4b), suggesting that OSCAR-Fc competes for the OSCAR ligand and inhibits stimulation of cells through surface OSCAR expression. DISCUSSION Joint destruction in rheumatoid arthritis results from an early and continuous attack of inflammatory tissue on mineralized structures such as bone and deep cartilage. Resorption of mineralized tissue is a hallmark of the disease, and virtually all drugs now being developed for RA have been tested for their ability to retard or arrest this bone-erosive process. Disease activity and structural damage are closely related in RA, and it is also thought that high disease activity drives increased formation of osteoclasts, which are bone-resorbing cells. This link between inflammation and increased bone destruction has been demonstrated for generalized as well as local periarticular bone loss. Thus, parameters such as the CRP level are excellent predictors of fracture risk as well as emergence of bone erosion in the inflamed joint (18,19). The molecular basis of this tight relationship between inflammation and bone destruction is only partly understood and most likely involves a variety of mechanisms, such as 1) the increased formation of osteoclast precursors in the bone marrow, 2) their enhanced migration to sites of inflammation, particularly the joint, and 3) an appropriate micro-milieu allowing their further differentiation into mature osteoclasts. It is clear that final differentiation of osteoclastlineage cells requires the presence of mineralized tissue,

8 3048 HERMAN ET AL which allows polarization of the cells and acquisition of specific structures to resorb bone. Indeed, mature osteoclasts are found only close to the bone surface in the joints of patients with RA (20). However, in their vicinity, numerous mononuclear osteoclast precursors, which have already acquired markers specific for osteoclast differentiation such as TRAP or cathepsin K, can be found. These cells, having undergone various differentiation steps toward the osteoclast lineage, are recruited from mononuclear precursors, which enter the joint through the bloodstream and synovial microvessels. The characteristics of the cells that invade the joint and are prone to differentiation into osteoclasts are not fully clear. It might be speculated that inflammation can drive certain phenotypic changes in monocytes, which occur before they enter the joint and make them susceptible to further differentiation into osteoclasts. Indeed, earlier studies have shown that there is a redistribution of CD11b mononuclear cells out of the bone marrow into the circulation and the spleen during inflammation (9). CD11b cells can serve as osteoclast precursor cells, and their increase in the circulation enhances the pool of cells that can enter the osteoclast lineage upon exposure to appropriate signals such as M-CSF and RANKL (9). However, there might be additional qualitative changes on monocytes that facilitate their susceptibility to osteoclastogenic stimuli. In that respect, surface molecules that are involved in osteoclast formation are of particular interest. OSCAR is a newly identified receptor molecule on the surface of osteoclast precursors and mature osteoclasts, which is part of the costimulation pathway and that signals through ITAM motifs and the common -chain of the Fc receptor. Our present findings show that OSCAR expression is increased in circulating monocytes of RA patients compared with healthy controls, and that expression is related to the inflammatory disease activity of RA. Mononuclear cells from patients with high OSCAR expression had an enhanced potential to differentiate into osteoclasts as compared with cells with low OSCAR expression, suggesting that OSCAR is an important enhancer of osteoclastogenesis. Moreover, levels of circulating soluble OSCAR are low in RA patients and inversely associated with inflammatory activity, suggesting that shedded soscar might exert a regulatory function by competing with membrane-bound OSCAR for its ligand. This concept is supported by the fact that addition of OSCAR-Fc fusion protein diminished osteoclastogenesis in a dosedependent manner. The overall process of regulation of OSCAR is poorly understood, although some transcription factors, such as microophtalmia-inducing transcription factor, induce OSCAR expression, whereas others, such as MafB and inhibitor of differentiation 2, suppress OSCAR transcription (21 23). Herein we have shown that TNF induces OSCAR expression in monocytes, which is an attractive explanation for the ability of this cytokine to trigger inflammatory bone loss. The inability of M-CSF and RANKL to induce OSCAR expression suggests that M-CSF/RANKL and OSCAR act in an additive manner, but independently from each other. OSCAR expression is driven mainly by inflammation, which allows priming of circulating monocytes to differentiate into osteoclasts upon contact with the appropriate signals, namely RANKL and M-CSF, which are provided by the synovial inflammatory tissue once these cells have entered the synovium. That this is indeed the case was demonstrated in our immunohistochemical studies, in which OSCARexpressing mononuclear cells were found within the inflamed synovium, and particularly at sites close to synovial microvessels. Although the ligand for OSCAR, as is the case for other immunoglobulin-like receptors, is unknown, these receptors make monocytes susceptible to stimulation with RANKL and enhance osteoclast differentiation. OSCAR can thus be regarded as a costimulation signal for osteoclasts, which acts in an additive manner with primary stimulation by RANKL, thus enabling a fine-tuning of osteoclast differentiation. OSCAR expression is also found in mature osteoclasts, which line along the erosion front between inflamed synovial tissue and bone. This finding highlights the specific role of OSCAR in the osteoclast lineage. Stimulation of OSCAR by crosslinking antibodies has been shown to drive chemokine expression in monocytes, including expression of interleukin-8 (CXCL8), MCP-1 (CCL2), neutrophil-activating peptide 78 (CXCL5), and CCL22 (12). These chemokines are involved in the recruitment of monocytes to sites of inflammation, and some of them, particularly MCP-1, also appear to have a role in osteoclast differentiation (24). Thus, expression and activation of OSCAR in the inflamed joint might additionally enhance the further recruitment and differentiation of osteoclast precursors by increased chemokine synthesis. OSCAR-mediated costimulation of osteoclasts may also play a role in inflammatory diseases other than RA in which there is involvement of increased bone resorption associated with bone erosions (e.g., psoriatic

9 OSCAR IN RHEUMATOID ARTHRITIS 3049 arthritis) or systemic osteoporosis (e.g., inflammatory bowel disease). In contrast, OSCAR expression was not found in OA synovial tissue, and there were no osteoclasts detectable in OA synovium. This is probably related to the bone phenotype observed in OA, which is primarily an osteoproliferative rather then a resorptive phenotype. There are several reasons for the absence of induced osteoclastogenesis in OA: 1) OA is not as inflammatory a disease as RA, and thus the cytokine load may not drive sufficient amounts of RANKL and OSCAR expression to induce osteoclasts, 2) the influx of monocytes as osteoclast precursors is lower in OA than in RA, and 3) prominent bone formation pathways such as Wnt proteins, which indirectly inhibit osteoclast formation through mediators such as osteoprotegerin, are active in OA (25). In summary, we have shown in the present study that expression of a key molecule in the costimulatory pathway of osteoclastogenesis is increased in circulating monocytes of RA patients, and that this molecule is also found in the synovial membrane of RA patients. These findings reinforce the hypothesis that in addition to a changed synovial microenvironment in RA, which provides pro-osteoclastogenic molecules such as RANKL, TNF, and interleukins-1, -6, and -17, there are systemic changes in the compartment of circulating monocytes that facilitate osteoclast formation and bone resorption. These changes involve the costimulatory pathway of osteoclast formation, particularly the immunoglobulinlike receptor OSCAR, which therefore might be considered as an interesting potential therapeutic target for inhibition of inflammatory osteoclastogenesis. ACKNOWLEDGMENT We thank Oskar Redlich for critical discussion of the manuscript. AUTHOR CONTRIBUTIONS Dr. Schett 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 design. Herman, Schett. Acquisition of data. Herman, Müller, Krönke, Hueber, Gelse, Neumann. Analysis and interpretation of data. Herman, Zwerina, Redlich, Müller-Ladner, Schett. Manuscript preparation. Herman, Schett. Statistical analysis. Herman. REFERENCES 1. McInnes IB, Schett G. Cytokines in the pathogenesis of rheumatoid arthritis [review]. Nat Rev Immunol 2007;7: Goldring SR. Pathogenesis of bone and cartilage destruction in rheumatoid arthritis. Rheumatology (Oxford) 2003;42 Suppl 2:ii Redlich K, Hayer S, Ricci R, David JP, Tohidast-Akrad M, Kollias G, et al. Osteoclasts are essential for TNF- -mediated joint destruction. J Clin Invest 2002;110: Gravallese EM, Manning C, Tsay A, Naito A, Pan C, Amento E, et al. Synovial tissue in rheumatoid arthritis is a source of osteoclast differentiation factor. Arthritis Rheum 2000;43: Udagawa N, Takahashi N, Akatsu T, Tanaka H, Sasaki T, Nishihara T, et al. Origin of osteoclasts: mature monocytes and macrophages are capable of differentiating into osteoclasts under a suitable microenvironment prepared by bone marrow-derived stromal cells. Proc Natl Acad Sci U S A 1990;87: Shalhoub V, Elliott G, Chiu L, Manoukian R, Kelley M, Hawkins N, et al. Characterization of osteoclast precursors in human blood. Br J Haematol 2000;111: Chu CQ, Field M, Feldmann M, Maini RN. Localization of tumor necrosis factor in synovial tissues and at the cartilage pannus junction in patients with rheumatoid arthritis. Arthritis Rheum 1991;34: Ritchlin CT, Haas-Smith SA, Li P, Hicks DG, Schwarz EM. Mechanisms of TNF- - and RANKL-mediated osteoclastogenesis and bone resorption in psoriatic arthritis. J Clin Invest 2003;111: Li P, Schwarz EM, O Keefe RJ, Ma L, Boyce BF, Xing L. RANK signaling is not required for TNF -mediated increase in CD11 hi osteoclast precursors but is essential for mature osteoclast formation in TNF -mediated inflammatory arthritis. J Bone Miner Res 2004;19: Kim N, Takami M, Rho J, Josien R, Choi Y. A novel member of the leukocyte receptor complex regulates osteoclast differentiation. J Exp Med 2002;195: Merck E, de Saint-Vis B, Scuiller M, Gaillard C, Caux C, Trinchieri G, et al. Fc receptor -chain activation via hoscar induces survival and maturation of dendritic cells and modulates Toll-like receptor responses. Blood 2005;105: Merck E, Gaillard C, Scuiller M, Scapini P, Cassatella MA, Trinchieri G, et al. Ligation of the FcR chain-associated human osteoclast-associated receptor enhances the proinflammatory responses of human monocytes and neutrophils. J Immunol 2006; 176: Kim GS, Koh JM, Chang JS, Park BL, Kim LH, Park EK, et al. Association of the OSCAR promoter polymorphism with BMD in postmenopausal women. J Bone Miner Res 2005;20: 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: 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: Kronke G, Bochkov VN, Huber J, Gruber F, Bluml S, Furnkranz A, et al. Oxidized phospholipids induce expression of human heme oxygenase-1 involving activation of camp-responsive elementbinding protein. J Biol Chem 2003;278: Koga T, Inui M, Inoue K, Kim S, Suematsu A, Kobayashi E, et al. Costimulatory signals mediated by the ITAM motif cooperate with RANKL for bone homeostasis. Nature 2004;428: Schett G, Kiechl S, Weger S, Pederiva A, Mayr A, Petrangeli M, et al. High-sensitivity C-reactive protein and risk of nontraumatic fractures in the Bruneck study. Arch Intern Med 2006;166: Smolen JS, van der Heijde DM, St.Clair EW, Emery P, Bathon

10 3050 HERMAN ET AL JM, Keystone E, et al, for the Active-Controlled Study of Patients Receiving Infliximab for the Treatment of Rheumatoid Arthritis of Early-Onset (ASPIRE) Study Group. Predictors of joint damage in patients with early rheumatoid arthritis treated with highdose methotrexate with or without concomitant infliximab: results from the ASPIRE trial. Arthritis Rheum 2006;54: Gravallese EM, Harada Y, Wang JT, Gorn AH, Thornhill TS, Goldring SR. Identification of cell types responsible for bone resorption in rheumatoid arthritis and juvenile rheumatoid arthritis. Am J Pathol 1998;152: Lee J, Kim K, Kim JH, Jin HM, Choi HK, Lee SH, et al. Id helix-loop-helix proteins negatively regulate TRANCE-mediated osteoclast differentiation. Blood 2006;107: So H, Rho J, Jeong D, Park R, Fisher DE, Ostrowski MC, et al. Microphthalmia transcription factor and PU.1 synergistically induce the leukocyte receptor osteoclast-associated receptor gene expression. J Biol Chem 2003;278: Kim K, Kim JH, Lee J, Jin HM, Kook H, Kim KK, et al. MafB negatively regulates RANKL-mediated osteoclast differentiation. Blood 2007;109: Binder NB. CCR2-deficient mice are osteopetrotic due to disturbed osteoclastogenesis [abstract]. Ann Rheum Dis 2007;66 Suppl II:ii Diarra D, Stolina M, Polzer K, Zwerina J, Ominsky MS, Dwyer D, et al. Dickkopf-1 is a master regulator of joint remodeling. Nat Med 2007;13: