All RA patients enrolled in this study fulfilled the 1987 revised criteria of the American College of Rheumatology 19. Patients were compared with age- and sex-matched control patients with OA.

Informed consent was obtained from all patients and the experimental protocol was approved by the Human Research Ethics Committee of the University of Ulsan College of Medicine (2010-871). SF was collected from the knee joint (with or without swelling) of each patient by sterile aspiration and centrifuged at 250g for 10 minutes. Cell-free RA synovial fluid (RA SF) was stored at -70°C until used. Synovia were isolated from seven patients with RA (mean age, 59.9 ± 12.3; range, 42 to 75 years) and seven patients with OA (mean age, 62.5 ± 8.2; range, 46 to 77 years). All patients were undergoing knee replacement surgery. For the one-year follow-up study, sera were collected from the patients listed in Additional file 1 at baseline and one year after treatment from ten patients with RA. RA activity was simultaneously assessed using Disease Activity Score 28 (DAS28) activity 20.

Pyrrolidine dithiocarbamate (PDTC), SP600125, SB203580, PD98059, and anti-β-actin antibody was purchased from Sigma (St. Louis, MO, USA). Rabbit anti-human IL-34 for immunoblotting was from ProSci Incorporated (Poway, CA, USA). Mouse anti-human IL-34 for blocking IL-34 activity, mouse polyclonal immunoglobulin G (IgG), IL-17, TNFα, and IL-1β were purchased from R&D Systems (Minneapolis, MN, USA). Recombinant M-CSF, RANKL, and IL-34 were purchased from PeproTech, Inc. (Rocky Hill, NJ, USA).

OA and RA synovia for immunohistochemistry analysis were collected in accordance with experimental protocols approved by the Stanford University Institutional Review Board and with the patients' informed consent as described 12. Sections of optimal cutting temperature (OCT) compound-embedded synovial tissue samples were cut on a cryostat (Thermo Scientific, Kalamazoo, MI, USA) and fixed in 4% paraformaldehyde for 10 minutes at 4°C. The slides were then washed twice (5 minutes each) with PBS and permeabilized with 0.1% Triton X-100 for 3 minutes at room temperature. After washing with PBS, endogenous peroxidase activity was quenched with 0.03% H₂O₂ solution. The slides were blocked with 7.5% normal goat serum for 20 minutes and thereafter incubated with 4 μg/ml of IL-34 antibody (LSBio, Seattle, WA, USA) or control rabbit IgG (Abcam, Cambridge, MA, USA) for 20 minutes at room temperature. Following two washes with PBS, the slides were incubated with 1:200 diluted biotinylated secondary antibody (Vectastain EliteABC Kit; Vector Laboratories, Orton Southgate, Peterborough, UK) for 30 minutes. After washing with PBS, the slides were incubated with complexed avidin-biotin horseradish peroxidase ABC reagent (Vectastain Elite ABCkit; Vector) for 30 minutes, and then exposed to 3,3-diaminobenzidine (DAB) substrate according to the manufacturer's instructions (Vector). The nuclei were subsequently counterstained with hematoxylin (Vector). Tissues were examined and photographed with an Olympus BX51 microscope outfitted with an Olympus DP72 digital camera (Olympus).

FLS were isolated by enzymatic digestion of synovial tissues obtained from RA and OA patients undergoing total joint replacement surgery, as previously described 21. Blood was collected from healthy volunteers and PBMCs were obtained from whole blood by centrifugation over Ficoll gradients (Pharmacia Biotech, Uppsala, Sweden) at 400g for 30 minutes and erythrocytes were removed by lysing in ACK buffer (0.15 M NH₄Cl, 1 mM KHCO₃ and 0.1 mM EDTA, pH 7.2).

Total RNA was extracted from synovial tissues obtained from three RA patients, as indicated, using the RNeasy Mini Kit (QIAGEN, Hilden, Germany). cDNA was synthesized from total RNA using murine Moloney leukemia virus reverse transcriptase (MMTV-RT, Gibco BRL, Gaithersburg, MD, USA). The cDNA was amplified using the Takara PCR amplification kit (Takara Biotechnology, Shiga, Japan); PCR was carried out in a Perkin Elmer thermal cycler. Quantitative RT-PCR (qRT-PCR) was performed using Power SYBR Green 1-Step Kit and an ABI 7000 Real Time PCR System (Applied Biosystems, Carlsbad, CA, USA) according to the manufacturer's instructions. The sequences of primers used were as follows: IL-34, 5'-TAC AGG AGC CGA CTT CAG-3' (sense) and 5'-CTC ACC AAG ACC CAC AGA-3' (antisense); M-CSF, 5'-GCT GCT TCA CCA AGG ATT ATG-3' (sense) and 5'- GGG TCA CTG CTA GGG ATG-3' (antisense); and GAPDH (glyceraldehyde 3-phosphate dehydrogenase), 5'-AGC CAC ATC GCT CAG ACA-3' (sense) and 5'-GCC CAA TAC GAC CAA ATC C-3' (antisense). Results are expressed as the ratio of target PCR product relative to GAPDH product. The amplification protocol consisted of an initial reverse transcription step at 48°C for 30 minutes, followed by 40 cycles of denaturation at 95°C for 15 seconds and annealing and extension at 60°C for 1 minute.

IL-34 concentration was measured using an IL-34-specific sandwich ELISA (R&D systems, Minneapolis, MN, USA) in accordance with the protocols of the manufacturer. Recombinant human IL-34 serially diluted in culture media was used as a standard. ELISA plates were incubated with mouse anti-human IL-34 Ab and then blocked with 3% BSA in PBS for 1 hour at room temperature. SF from patients with RA and OA, or supernatants from FLS cultures were added to the plates. After washing with PBS containing 0.05% Tween 20, each well was incubated with biotinylated mouse anti-human IL-34 (0.5 μg/ml) for 2 hours at room temperature. Plates were then washed with PBS containing 0.05% Tween 20 and incubated with streptavidin conjugated to horseradish peroxidase (HRP) for 1 hour at room temperature. A color reaction was developed with tetramethylbenzidine solution, and then stopped with 0.1 M H2SO4. The absorbance at 450 nm was measured using a Bio-Rad microtiter plate reader (Bio-Rad Laboratories, Hercules, CA, USA).

RA FLS were collected and proteins were resolved by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose paper (GE Healthcare, Piscataway, NJ, USA). The membranes were immunoblotted with an anti-human IL-34 rabbit polyclonal IgG or β-actin antibody. Bound antibodies were visualized with HRP-conjugated antibodies against rabbit or mouse IgG using an enhanced chemiluminescence Western blotting system (Millipore, Billerica, MA, USA).

Transmigration assays were performed as follows using PBMCs: conditioned media (CM) was prepared by collecting supernatant from RA FLS cultured in the presence of TNFα (10 ng/ml) for 24 hours. CM from RA FLS cultures, recombinant IL-34 (100 ng/ml), or CM plus anti-IL-34 Ab was added to the lower compartments of a 5-μm pore Transwell system (Costar, Corning, NY, USA). PBMCs (1.2 × 10⁶) were added to the upper chamber of each transwell and allowed to migrate for 6 hours. The number of PBMCs in the lower chamber was then counted.

For osteoclastogenesis, PBMCs were plated on 48-well plates and cultured in α-MEM supplemented with 10% fetal bovine serum (FBS) and cultured with RANKL (100 ng/ml) plus M-CSF (100 ng/ml) as a control as described 21 or RANKL (100 ng/ml) plus IL-34 (100 ng/ml). RANKL was added at a final concentration of 100 ng/ml to CM obtained from RA FLS cultured in the presence of TNFα (10 ng/ml) for 24 hours. Osteoclastic differentiation was maintained for 12 to 14 days prior to tartrate resistant acid phosphatase (TRAP) staining and scoring of multinucleated OCs, as previously described 22.

Means ± standard deviations (SDs) were calculated for all conditions, and differences between means were analyzed using Student's t-test. A P-value < 0.05 was considered significant. Patient data were tested by the Mann-Whitnney U test.

To demonstrate the involvement of IL-34 in RA pathogenesis, we first determined whether synovia and SF from patients with RA (RA SF) displayed high levels of IL-34. IL-34 mRNA expression was detected in synovia obtained from three RA patients at levels similar to those of M-CSF (Figure 1A); high levels of IL-34 mRNA (Figure 1A) and protein (Figure 1B) were observed in human RA synovium compared to OA synovium. The levels of IL-34 protein in RA SF were comparable to those of M-CSF (Figure 1C). In addition, we found that IL-34 protein was expressed in FLS from RA patients (Figure 1D). These observations indicate that IL-34 is expressed in synovial tissue, synovial fluid, and FLS from RA patients, suggesting that IL-34 is an additional factor involved in the pathogenesis of RA.

FLS act as the main cell mediating the cytokine-rich environment of the inflamed synovium in RA pathogenesis 23 and stimulate OC differentiation and activity in the inflammatory bone destruction by regulating RANKL expression 2425. Interestingly, we found that the level of soluble IL-34 in RA SF from inflamed RA synovium (351.7 ± 30.7 pg/ml) with joint swelling was significantly higher than that in RA SF from non-inflamed RA synovium (81.1 ± 24.5 pg/ml; P < 0.01) (Figure 2A). We therefore determined whether RA FLS produced IL-34 in response to pro-inflammatory cytokines. RA FLS (n = 9) or OA FLS (n = 4), used as a disease control, were cultured with TNFα, IL-1β, or IL-17 (each at 10 ng/ml) for 24 hours. Interestingly, the increase in IL-34 concentration was more marked in TNFα-stimulated RA FLS than in IL-1β- or IL-17-stimulated RA FLS (Figure 2B). In contrast, OA FLS showed no significant changes in those responses (Figure 2B). Moreover, an analysis of IL-34 mRNA levels using quantitative real-time RT-PCR showed that IL-34 mRNA expression was markedly increased in TNFα-stimulated RA FLS compared to controls (P < 0.01) (Figure 2C). IL-1β and IL-17 treatment induced a smaller increase in IL-34 mRNA expression (Figure 2C). However, the level of IL-34 expression in control OA FLS was not significantly modulated by these cytokines (Figure 2C). These results provide the first demonstration that RA FLS produce IL-34 and that this production is enhanced by TNFα.

IL-34 and M-CSF were recently reported to share the same receptor, c-Fms 13. Our data showed that IL-34 expression was enhanced after treatment with TNFα Figure 2), suggesting that differential production of IL-34, regulated by the pathological condition of RA, can replace M-CSF function. To test this possibility, we next examined whether TNFα differentially regulates IL-34 and M-CSF expression. RA FLS constitutively expressed and produced both M-CSF and IL-34 (Figure 3A). TNFα treatment caused a 12-fold increase in IL-34 mRNA expression above this basal level (Figure 3A, right), but had little effect on M-CSF mRNA expression (Figure 3A, left). Together, the level of IL-34 protein in RA FLS was much greater than M-CSF production in the presence of TNFα (Figure 3B), showing approximately a 1.5-fold increase in TNFα-treated RA FLS over that of the control RA FLS (Figure 3C). Thus, these data demonstrate the differential regulation of IL-34 and M-CSF expression by TNFα.

TNFα activates NF-κB and JNK, which are important for TNFα-mediated gene expression and are involved in the activation of FLS 26. To understand better the mechanism by which TNFα increases IL-34 production in RA FLS, we examined the effects of inhibitors of NF-κB or JNK on IL-34 expression. RA FLS were pretreated with the vehicle, NF-κB inhibitor, PDTC, or the JNK inhibitor, SP600125, at a concentration of 10 μM for 30 minutes followed by incubation with TNFα for 24 hours. As shown in Figure 3D, the NF-κB inhibitor, PDTC, effectively reduced TNFα-induced IL-34 mRNA expression; moreover, JNK inhibition modestly decreased the TNFα-induced elevation of IL-34. However, p38 or extracellular signal-regulated protein kinase (ERK) inhibition did not affect TNFα-induced IL-34 mRNA expression (Figure 3D). Collectively, these results suggest that the TNFα-induced increase in IL-34 expression in RA FLS is mediated by NF-κB and JNK pathways.

Given that IL-34 is up-regulated in inflamed synovium and in RA FLS in response to TNFα, (Figure 2) the levels of IL-34 may reflect the pathogenesis of inflammation in RA patients. To explore this correlation, we performed a comparative analysis of plasma IL-34 levels in RA patients (n = 10) using ELISAs. The patients with RA had higher plasma IL-34 levels than normal or OA controls (Figure 4A). One group of patients with rheumatoid arthritis, defined by a plasma DAS28 score at baseline of 2.9 to 7.4, had undergone a 1-year follow-up treatment with DMARDs, including methotrexate, sulfasalazine, leflunomide, minocycline, and/or hydroxychloroquine. Patients in this group (n = 10) showed an apparent decrease in DAS28 score of approximately 50% (to 1.7 to 3.4) after taking DMARDs (Figure 4B). Importantly, high levels of soluble IL-34 (6.573 ± 2.003 ng/ml) were detected in plasma from RA patients before DMARDs treatment (Figure 4C). However, a significant decrease (P < 0.01) in plasma IL-34 levels was observed after taking DMARDs in this group (Figure 4C). These data suggest that the plasma level of IL-34 correlates with the level of inflammation in RA patients.

Having shown that RA FLS produce IL-34 (Figure 2A and 3), we thus questioned whether IL-34 produced by RA FLS has functional activities to induce chemotactic migration of OCPs and to subsequently induce osteoclastogenesis as a substitute for M-CSF in RANKL-induced osteoclastogenesis. Human PBMCs as OCPs 27 in the upper chamber and CM from TNFα-treated RA FLS or recombinant IL-34 (positive control) in the lower chamber were incubated. Approximately 7.5 × 10⁵ human PBMCs migrated in the presence of recombinant IL-34; a similar level of cell migration was observed in the presence of CM, which contains soluble IL-34 secreted from RA FLS (Figure 5A). However, addition of a blocking Ab against IL-34 to CM reduced the number of migrated human mononuclear cells in a dose-dependent manner.

To further assess the effects of IL-34 produced by RA FLS on RANKL-induced OC differentiation, we cultured isolated human PBMCs with RANKL in the presence or absence of M-CSF, IL-34, CM, or an anti-IL-34 Ab for 12 to 14 days. M-CSF plus RANKL treatment induced OC differentiation (851.5 ± 17.7; Figure 5B). IL-34 or CM from RA FLS cultures containing IL-34 supported RANKL-induced OC formation to an extent similar to that of M-CSF. The addition of anti-IL-34 antibody to block functional IL-34 in CM (CM + R + IL-34 Ab) significantly reduced OC formation (472.4 ± 39; P = 0.042) (Figure 5C). These data demonstrate that IL-34 secreted by RA FLS is functionally active enough to stimulate the chemotactic migration of OCPs and subsequent OC formation.