Peripheral blood T lymphocytes from patients with early rheumatoid ...

ARTHRITIS & RHEUMATISM Vol. 54, No. 4, April 2006, pp 1151–1164 DOI 10.1002/art.21731 © 2006, American College of Rheumatology

Peripheral Blood T Lymphocytes From Patients With Early Rheumatoid Arthritis Express RANKL and Interleukin-15 on the Cell Surface and Promote Osteoclastogenesis in Autologous Monocytes Marıá-Eugenia Miranda-Caru ´s, Marta Benito-Miguel, Alejandro Balsa, Tatiana Cobo-Ibań ˜ez, Carlos Pe´rez de Ayala, Dora Pascual-Salcedo, and Emilio Martıń-Mola the SF of patients with established RA and from the PB of patients with early RA, but not from the PB of healthy controls, resulted in osteoclast differentiation that was significantly inhibited by osteoprotegerin (OPG) and by neutralizing monoclonal antibodies to IL-15, IL-17, tumor necrosis factor ␣ (TNF␣), and IL-1␤. OPG, anti-TNF␣, and anti–IL-1␤ demonstrated a cooperative inhibitory effect. At 1-year followup, surface RANKL and IL-15 and ex vivo osteoclastogenesis were no longer observed on PB T cells or monocytes from patients with early RA in whom clinical remission had been achieved with treatment. Conclusion. T cells are important contributors to the pathogenesis of bone erosions in RA through interaction with osteoclast precursors of the monocyte/ macrophage lineage.

Objective. To investigate the osteoclastogenic potential of T cells from the peripheral blood (PB) and synovial fluid (SF) of patients with rheumatoid arthritis (RA) on autologous monocytes, and to study the cytokines implicated in this process. Methods. T cells and monocytes were isolated from the PB of 20 healthy subjects and 20 patients with early RA, and from the SF of 20 patients with established RA. Autologous T cell/monocyte cocultures were established in the absence of exogenous cytokines or growth factors in order to examine spontaneous ex vivo osteoclast differentiation by tartrate-resistant acid phosphatase staining and calcified matrix resorption activity. Results. Surface RANKL was expressed on freshly isolated T cells from the PB of patients with early RA and the SF of patients with established RA. In addition, surface interleukin-15 (IL-15) was detected on freshly isolated T cells and monocytes from the PB of patients with early RA and the SF of patients with established RA. Autologous T cell/monocyte cocultures derived from

Histopathologic studies of the bone–pannus junction and subchondral bone marrow of patients with rheumatoid arthritis (RA) (1) indicate that osteoclasts play a pivotal role in the focal marginal and subchondral bone loss of inflammatory arthritis. Osteoclasts are multinucleated cells formed by fusion of mononuclear precursors of the monocyte/macrophage family under the influence of cell interactions and cytokines (2). Physiologic osteoclastogenesis is driven by the tumor necrosis factor ␣ (TNF␣) family member RANKL (3), principally as a membrane-bound protein on the surface of marrow stromal cells or osteoblasts (2). Osteoprotegerin (OPG), a soluble decoy receptor, competes with RANK for binding to RANKL, preventing its osteoclastogenic effect (3). In the context of inflammation, activated T cells and RA synovial fibroblasts express RANKL (4–10) and have the capacity to induce osteoclast differentiation

Dr. Miranda-Caru ´s’s work was supported by the Ministerio de Educacio ń Programa Ramo ń y Cajal and the Ministerio de Educacio ń (grant SAF 2003-01670). Ms Benito-Miguel and Drs. Balsa and Martıń-Mola’s work was supported by the Ministerio de Educacio ń (grant SAF 2003-01670). Drs. Balsa, Pascual-Salcedo, and MartıńMola’s work was supported by the Fondo de Investigacio ń Sanitaria (grant FIS-PI-021176). Marıá-Eugenia Miranda-Caru ´s, MD, PhD, Marta BenitoMiguel, BS, Alejandro Balsa, MD, PhD, Tatiana Cobo-Ibań ˜ez, MD, Carlos Pe´rez de Ayala, MD, PhD, Dora Pascual-Salcedo, PhD, Emilio Martıń-Mola, MD, PhD: Hospital Universitario La Paz, Madrid, Spain. Address correspondence and reprint requests to MarıáEugenia Miranda-Caru ´s, MD, PhD, Department of Rheumatology, Hospital La Paz, Paseo de la Castellana, 261, 28046 Madrid, Spain. E-mail: [email protected]. Submitted for publication August 11, 2005; accepted in revised form December 30, 2005. 1151

´ S ET AL MIRANDA-CARU

1152

(6,8,10,11). In addition, several proinflammatory cytokines induce multinucleation of osteoclast precursors and/or commitment to the osteoclast phenotype (TNF␣, interleukin-1␤ [IL-1␤], IL-15, and IL-17 [12–16]) and may act synergistically with RANKL. In fact, the macrophage can serve both as osteoclast progenitor and as a source of osteoclastogenic cytokines (17,18). The cytokine IL-15 is expressed in both monocyte/ macrophages and T cells (19–22) and has recently been described to enhance osteoclast differentiation (13). IL-15 secretion to the extracellular space has rarely been demonstrated (23–25); in contrast, IL-15 can be expressed on the cell surface, where it is able to exert biologic functions through cell contact–dependent mechanisms (26–28). IL-15 has been detected in the synovial fluid (SF) (29) and synovial membrane (22,29) of RA patients. In addition, in vitro and in vivo studies (30,31) suggest that IL-15 may be a major player in the pathogenesis of RA. Administration of soluble IL-15 receptor ␣ (IL-15R␣) (30) or an antagonist mutant IL-15/Fc protein (31) prevents collagen-induced arthritis in mice and effectively reduces inflammation, synovial hyperplasia, and bone erosions. In addition, a phase I–II clinical trial in humans using a fully human anti–IL-15 monoclonal antibody (mAb) indicates that neutralization of IL-15 in patients with RA is effective and safe (32). Phenotypic and functional differences have been described between peripheral blood (PB) T cells and monocytes of RA patients and cells of healthy controls, whereas RA SF T cells and monocytes demonstrate a clearly activated phenotype (33,34). Our objective was to investigate the role of T cells from the PB and SF of RA patients in osteoclast differentiation of autologous monocytes and to study the cytokines implicated in this process. Our early arthritis clinic allowed the study of cells from patients with early RA who had not received disease-modifying antirheumatic drugs (DMARDs) or steroids, thereby minimizing interference of drugs with in vitro T cell and monocyte responses. PATIENTS AND METHODS Patients. SF was obtained from the knee joints of 20 patients with established RA who were receiving treatment with oral methotrexate (MTX) and low-dose prednisone. PB was obtained from 20 healthy controls and from 20 patients with early RA who had never received DMARDs or corticosteroids and who had a disease duration of ⬍6 months. All RA patients fulfilled at least 4 of the 1987 revised criteria of the American College of Rheumatology (formerly, the American Rheumatism Association) (35). La Paz University Hospital in

Madrid, Spain, has a monographic clinic in which patients with early arthritis (referred from a wide primary care area) receive care. This facilitated recruitment of untreated patients with early RA for the present study. The group with early RA comprised 20 patients (5 men and 15 women), 15 (75%) of whom tested positive for IgM rheumatoid factor (RF). The mean ⫾ SD age of these patients was 46.9 ⫾ 18.4 years (median 46 years, range 18–86 years). In this group, the mean ⫾ SD duration of symptoms at the first evaluation was 11.40 ⫾ 8.02 weeks (median 8 weeks, range 2–26 weeks), and the mean ⫾ SD Disease Activity Score in 28 joints (DSA28) (36) at the first evaluation was 6.05 ⫾ 0.90 (median 5.86, range 4.71–7.73). The group with established RA comprised 20 patients (6 men and 14 women), 16 (80%) of whom tested positive for IgM RF. Patients in this group ranged in age from 23 to 85 years (median 54 years), and the disease duration ranged from 14 months to 15 years (median 8.8 years). Among patients with established RA, the DAS28 at the time of SF aspiration ranged from 3.3 to 6.3 (median 4.4). All patients were receiving oral MTX; in addition, 14 patients were receiving low-dose prednisone (2.5–7.5 mg/day). All knee joints demonstrated signs of active synovitis at the time of aspiration. The SF white blood cell count was 2,500–18,300 cells/mm3 (median 8,108), with 72–87% polymorphonuclear cells (median 78%). The study was approved by the hospital ethics committee. T lymphocyte and monocyte purification. Mononuclear cells were separated from PB and SF by FicollHypaque (Amersham, Uppsala, Sweden) density-gradient centrifugation. Highly purified T cells and monocytes were prepared from mononuclear cells by exhaustive immunomagnetic negative selection in an autoMACS (Miltenyi Biotec, Bergisch Gladbach, Germany) using a T Cell Negative Isolation Kit and a Monocyte Negative Isolation Kit (Miltenyi Biotec). The purity of the T cell preparation was ⬎99% CD3⫹ by flow cytometry, and the purity of the monocyte preparation was ⬎95%. In some experiments, CD4⫹, CD8⫹, CD45RO⫹, and CD45RA⫹ T cells were further purified from CD3⫹ cells by positive magnetic selection using specific MACS MicroBeads (Miltenyi Biotec) and an autoMACS. Coculture conditions. All experiments were performed in triplicate, and variation between triplicates was ⬍5%. Monocytes (5 ⫻ 105/well) and T cells (2 ⫻ 106/well) were seeded in 24-well plastic tissue culture plates (Corning, Cambridge, MA). Alternatively, monocytes (1 ⫻ 105/well) and T cells (4 ⫻ 105/well) were seeded in BD BioCoat Osteologic 16-well slides (BioCoat Osteologic Bone Cell Culture System; BD Biosciences, Bedford, MA). Cocultures were maintained for 14 days in RPMI 1640 (Invitrogen Life Technologies, Carlsbad, CA) containing 10% fetal bovine serum, L-glutamine (2 mM), penicillin (100 units/ml), and streptomycin (100 ␮g/ml). Medium was replenished every 4 days, as follows. Culture medium was removed and centrifuged. T cells in the pellet were resuspended in 50% fresh culture medium and 50% of the previously removed supernatant and were then added back to the culture wells. On day 14, supernatants were harvested, filtered, and stored at –80°C for subsequent use in enzyme-linked immunosorbent assay (ELISA) and in coculture experiments. A Transwell system (Corning) was used to conduct some coculture experiments. The system consists of 2 compart-

RA T CELLS INDUCE OSTEOCLASTOGENESIS IN MONOCYTES

ments: a top well with a porous matrix (0.4 ␮m), and a bottom well. This set-up allows coculture of 2 types of cells in the same medium with soluble factors exchanged through the pores while preventing direct contact between the 2 types of cells. Monocytes were placed in the bottom well, and T cells were added either to the same well (allowing contact) or to the top well (avoiding contact). Determination of osteoclast differentiation. Nonadherent cells were removed after 14 days in culture. Adherent cells were stained for tartrate-resistant acid phosphatase (TRAP; Sigma, St. Louis, MO), and nuclei were counterstained with hematoxylin. Cells were viewed by light microscopy using an Olympus IX-51 microscope (Olympus, Hamburg, Germany), and TRAP⫹ cells with ⱖ3 nuclei were counted as osteoclasts. Cultures were photographed with a Coolpix 4500 digital camera (Nikon, Tokyo, Japan), and images were transferred with NikonView 5 software. Results are given as the number of TRAP⫹ multinuclear cells per well. In addition, calcified matrix resorption activity of the osteoclast-like cells was tested using BD BioCoat Osteologic calcium hydroxyapatite–coated slides (BD Biosciences). Cells were removed from BioCoat slides with 5% sodium hypochlorite. Slides were washed with distilled water, air-dried, and examined by light microscopy (IX-51 microscope). The resorbed area with cleared CaPO4 resorption pits was measured using Photoshop 7.0 software (Adobe Systems, San Jose, CA), and the percentage resorption area was calculated by dividing the total pitted area by the total surface area. Effect of soluble mediators. To test the biologic effect of soluble mediators present in coculture supernatants, T cell/monocyte cocultures from the PB of 5 healthy controls were established in 24-well plates (5 ⫻ 105 monocytes and 2 ⫻ 106 T cells per well) or 16-well Osteologic slides (1 ⫻ 105 monocytes and 4 ⫻ 105 T cells per well) with 50% RA PB T cell/monocyte coculture supernatant and 50% fresh culture medium; in parallel, T cell/monocyte cocultures from the PB of 5 healthy controls were established in 24-well plates or 16-well Osteologic slides with 50% RA SF T cell/monocyte coculture supernatant and 50% fresh culture medium. After 14 days, the number of TRAP⫹ multinucleated cells and the percentage of area resorbed were calculated. Inhibition of osteoclastogenesis by OPG and neutralizing mAb. Autologous T cell/monocyte cocultures were established as described above. Human OPG-Fc chimera (R&D Systems, Abingdon, UK) or a nonspecific control human IgG was added to the medium at 1 ␮g/ml. Alternatively, neutralizing mAb against IL-15 (catalog no. MAB247), TNF␣ (MAB610), IL-1␤ (MAB201), and IL-17 (MAB317) and an isotype control mAb (all from R&D Systems) or an anti–HLA class I mAb (W6/32; Sigma) were added to the medium at 10 ␮g/ml. In vitro treatment with MTX. Alternatively, cocultures were established in the presence of MTX (Sigma) at pharmacologically relevant doses varying from 1 nM to 500 nM (37). In some conditions, and to determine the contribution of adenosine release to the effect of MTX, adenosine deaminase (ADA) (type IV, calf intestinal; Sigma) (0.125 IU/ml), the adenosine A2A receptor antagonist ZM241385 (10 ␮M) (Tocris Crookson, Bristol, UK), the adenosine A2B receptor antagonist MRS 1706 (1 nM) (Tocris Crookson), or the adenosine A1 receptor antagonist 8-cyclopentyl-dipropylxanthine

1153

(DPCPX) (Sigma) (10 ␮M) was added to the medium together with MTX. ADA was dialyzed against phosphate buffered saline (PBS) overnight at 4oC before being used, as previously described (38). To study the effect of MTX on cytokine expression by cocultured cells, RNA was extracted from T cells and monocytes immediately after isolation and 3 days after initiation of cocultures. The level of T cell and monocyte cytokine messenger RNA (mRNA) expression was determined by real-time reverse transcription–polymerase chain reaction (RT-PCR), and conditions with or without MTX present were compared. In vivo effect of treatment. All 20 patients with early RA were treated with oral MTX at a dosage of 12.5–25 mg/week; in addition, 7 patients received oral prednisone at 2.5–5 mg/day, and 2 patients received oral leflunomide at 10 mg/day in combination with MTX and prednisone. Sixteen of these 20 patients donated blood a second time 1 year after the initiation of treatment with DMARDs. Blood was also obtained again from the 16 healthy subjects who had previously acted as controls for these patients. Cocultures of autologous T cells and monocytes were established as described above, and osteoclast formation was examined. Posteroanterior radiographs of hands and feet were evaluated by 2 rheumatologists who were unaware of the experimental results (AB and TC-I), using the method described by van der Heijde et al (39). The interobserver variation was 0.89, and the intraobserver variations were 0.93 and 0.88. Results for each patient were compared with results obtained before initiation of treatment. Flow cytometry. Freshly isolated T cells and monocytes were washed with PBS/2% fetal calf serum (FCS)/0.01% NaN3 and incubated on ice for 1 hour with an anti–IL-15 mAb (MAB247), an anti-RANKL mAb (MAB6261), or an irrelevant IgG1 isotype control mAb (MAB002; all from R&D Systems). Cells were then washed and incubated on ice for 30 minutes with an AlexaFluor 488–conjugated goat anti-mouse antibody (F[ab⬘]2 fragment) (Molecular Probes, Eugene, OR). After washing once with PBS/2% FCS/0.01% NaN3 and once with PBS, cells were resuspended in 1% paraformaldehyde and analyzed in a FACSCalibur flow cytometer using CellQuest software (Becton Dickinson, San Jose, CA). The following fluorochrome-conjugated mAb from BD PharMingen (San Jose, CA) were used to examine the expression of surface leukocyte antigens: fluorescein isothiocyanate–labeled antiCD25, anti–HLA–DR, and anti–CD40 ligand (anti-CD40L), phycoerythrin-labeled anti-CD69, anti–CTLA-4, antiCD45RO, and anti-CD14, and allophycocyanin-labeled antiCD45RA. Mean fluorescence intensity (MFI) values are given as the difference between the MFI of tested cells and the MFI of background staining. ELISAs. Cell-free coculture supernatants were collected and stored at ⫺80°C. ELISAs for RANKL were performed using a kit from Biomedica Medizinprodukte (Vienna, Austria). ELISAs for OPG, IL-15, TNF␣, IL-1␤, IL-17, IL-4, and interferon-␥ (IFN␥) were performed using DuoSet kits (R&D Systems) in accordance with the manufacturer’s instructions. Real-time RT-PCR. RNA was obtained from T cells and monocytes immediately after isolation and also 3 days after initiation of cocultures. Total cellular RNA was extracted using the RNeasy Mini Kit (Qiagen, Hilden, Germany) with in-column DNase treatment. For each sample, 1 ␮g of

1154


Figure 1. Cell surface expression of RANKL and interleukin-15 (IL-15), as determined by flow cytometry. A and B, Flow cytometry shows surface expression of RANKL and IL-15, respectively, on nonpermeabilized, freshly isolated T lymphocytes from peripheral blood (PB) of healthy controls (HCPBTL) (n ⫽ 20), PB of patients with early rheumatoid arthritis (RA) (RAPBTL) (n ⫽ 20), and synovial fluid (SF) of patients with established RA (RASFTL) (n ⫽ 20). C, Flow cytometric analysis of surface IL-15 expression on freshly isolated monocytes from the PB of healthy controls (HCPBMo) (n ⫽ 20), PB of patients with early RA (RAPBMo) (n ⫽ 20), and SF of patients with established RA (RASFMo) (n ⫽ 20). In the top rows of A–C are representative fluorescence-activated cell sorting (FACS) histograms, in which black lines represent specific staining with RANKL or IL-15 and gray lines represent staining with an isotype control monoclonal antibody. Below the FACS histograms, bar graphs at left show the mean and SD mean fluorescence intensity (MFI) of A, T cell RANKL, B, T cell IL-15, and C, monocyte IL-15 for each group of subjects. Each bar represents the arithmetic mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus expression on cells from healthy controls; † ⫽ P ⬍ 0.05 versus expression on PB T cells or monocytes of patients with early RA. Bar graphs at right represent the mean and SD surface expression of RANKL and IL-15, respectively, on T cell subsets (CD4⫹, CD8⫹, CD45RO⫹, and CD45RA⫹) from the PB of 20 patients with early RA. ⴱ ⫽ P ⬍ 0.05 versus expression on CD4⫹ T cells; † ⫽ P ⬍ 0.05 versus expression on CD45RO⫹ T cells.

total RNA was subjected to reverse transcription using the Advantage RT-for-PCR Kit (BD Clontech, Palo Alto, CA) in accordance with the manufacturer’s instructions. Aliquots (1 ␮l) of the reverse transcription products were used for quantitative PCR in the LightCycler PCR and detection system (Roche Molecular Biochemicals, Mannheim, Germany) using

the FastStart DNA Master SYBR Green I kit (Roche Diagnostics, Mannheim, Germany), as previously described (28). The PCR reactions were set up in microcapillary tubes in a volume of 20 ␮l. The following sense and antisense primers were used: for RANKL, sense 5⬘-AAA-TCC-CAA-GTT-CTC-ATA-


CCC-3⬘ and antisense 5⬘-TCT-CAT-AAG-GTC-AAC-CCGTAA-3⬘ (product size 490 bp); for IL-15, sense 5⬘-GGA-TTTACC-GTG-GCT-TTG-AGT-AAT-GAG-3⬘ and antisense 5⬘CAA-TCA-ATT-GCA-ATC-AAG-AAG-TG-3⬘ (product size 643/524 bp); for IL-17, sense 5⬘-TGG-AGG-CCA-TAG-TGAAGG-3⬘ and antisense 5⬘-GGC-CAC-ATG-GTG-GACAAT-3⬘ (product size 416 bp); for TNF␣, sense 5⬘-ATG-AGCACT-GAA-AGC-ATG-ATC-CGG-3⬘ and antisense 5⬘-CTACAA-CAT-GGG-CTA-CAG-GCT-TGT-3⬘ (product size 280 bp); for IL-1␤, sense 5⬘-GCC-CTA-AAC-AGA-TGA-AGTGCT-C-3⬘ and antisense 5⬘-AGA-AGG-TGC-TCA-GGTCAT-TCT-C-3⬘ (product size 198 bp); for OPG, sense 5⬘GGG-GAC-CAC-AAT-GAA-CAA-GTT-G-3⬘ and antisense 5⬘-AGC-TTG-CAC-CAC-TCC-AAA-TCC-3⬘ (product size 408 bp); and for RANK, sense 5⬘-GTA-CAC-ACA-CGGCAA-AC-3⬘ and antisense 5⬘-TGC-TCT-GTG-TCC-CCGTGA-AGC-3⬘ (product size 378 bp). Amplification of the PCR products was monitored by measuring SYBR Green I dye fluorescence. As an external standard, the transcript of 18S ribosomal RNA (rRNA) was amplified from the same complementary DNA (cDNA) samples using primers manufactured by Ambion (Austin, TX). Each sample was run in triplicate. Results were analyzed with LightCycler version 3.5.3 software (Roche Diagnostics). Quantities of specific mRNA in the sample were measured according to the corresponding gene-specific standard curve. The results are expressed as fold of induction: (cDNA sample cocultured cells/18S rRNA cocultured cells)/(cDNA sample freshly isolated cells/18S rRNA freshly isolated cells). The cDNA of freshly isolated cells normalized to the level of 18S rRNA mRNA was ascribed a fold induction of 1. Statistical analysis. Comparison between groups was by Mann-Whitney test. Paired samples were compared using Wilcoxon’s matched pairs signed rank sum test. When appropriate, Bonferroni correction for multiple comparisons was applied.

RESULTS Expression of RANKL and IL-15 on T cells from RA patients. We first sought to determine whether T cells from the PB and SF of RA patients express osteoclastogenic factors on the cell surface. Nonpermeabilized, freshly isolated PB T lymphocytes from patients with early RA (n ⫽ 20) and SF T lymphocytes from patients with established RA (n ⫽ 20) demonstrated significant surface expression of RANKL and IL-15 (Figures 1A and B). Expression of surface RANKL and IL-15 was higher on CD4 T cells than on CD8 T cells and higher on CD45RO T cells than on CD45RA T cells from the PB of patients with early RA (Figures 1A and B). In contrast, surface RANKL and surface IL-15 could not be detected on PB T lymphocytes from healthy controls (n ⫽ 20) (Figures 1A and B). An increased expression of HLA–DR was observed on PB T lymphocytes from patients with early RA (n ⫽ 20) (mean ⫾ SD MFI 7.1 ⫾ 3.9, median 5.5, range

1155

Figure 2. Unstimulated PB T cells from patients with early RA induce osteoclast differentiation in cocultured autologous PB monocytes. A, To study the formation of osteoclasts, autologous T cell/monocyte cocultures from A1, the PB of healthy controls, A2, the PB of patients with early RA, or A3, the SF of patients with established RA were established in 24-well plates in the absence of exogenously added macrophage colony-stimulating factor or RANKL. After 14 days, nonadherent cells were removed. Adherent cells were stained for tartrate-resistant acid phosphatase (TRAP) and viewed by light microscopy, and TRAP⫹ cells with ⱖ3 nuclei were counted as osteoclasts. Bars ⫽ 20 ␮m. A magnified image of a section from A3 shows in detail an osteoclast with prominent nuclei (bar ⫽ 40 ␮m). B, The number of TRAP⫹ multinucleated cells (mncs) after 14 days of culture was scored for each subject. Each bar represents the mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus healthy controls. C, CD4⫹, CD8⫹, CD45RO⫹, and CD45RA⫹ cell subsets were magnetically sorted from PB mononuclear cells of 5 patients with early RA, and cocultures with autologous monocytes were established in 24-well plates. Shown is the number of TRAP⫹ multinucleated cells per well for each subset. Each bar represents the arithmetic mean and SD of 5 patients with early RA. ⴱ ⫽ P ⬍ 0.05 versus CD4⫹ cell subset; † ⫽ P ⬍ 0.05 versus CD45RO⫹ cell subset, after Bonferroni correction for multiple comparisons. See Figure 1 for other definitions.

2–14) compared with that on PB T lymphocytes from healthy controls (n ⫽ 20) (mean ⫾ SD MFI 4.2 ⫾ 2.5, median 4.0, range 1–9) (P ⬍ 0.05). CD25 expression was observed in 13.5 ⫾ 1.8% of freshly isolated PB T lymphocytes from patients with early RA (MFI 4.5 ⫾ 0.7) and in 9.9 ⫾ 1.5% of freshly isolated PB T lympho-

1156


on the surface. Surface IL-15 expression could be detected by flow cytometry on freshly isolated monocytes from the PB of patients with early RA and from the SF of patients with established RA, but not from the PB of healthy controls (Figure 1C). Osteoclast differentiation in autologous T cell/ monocyte cocultures derived from the PB of patients with early RA and from the SF of patients with established RA. Numerous multinucleated TRAP⫹ cells were identified in autologous T cell/monocyte cocultures derived from the PB of patients with early RA and from the SF of patients with established RA, in the absence of exogenous RANKL or macrophage colony-stimulating

Figure 3. Bone resorption assay as functional evidence of osteoclast differentiation. A, To study the formation of resorption lacunae, autologous T cell/monocyte cocultures from A1, the PB of healthy controls, A2, the PB of patients with early RA, or A3, the SF of patients with established RA were established on Osteologic hydroxyapatitecoated slides. After 14 days, adherent cells were removed with 5% sodium hypochlorite. Slides were washed with distilled water, air-dried, and examined by light microscopy. Note resorption lacunae (arrows), together with leftover cell debris that remained attached to the slides despite extensive washing. Bars ⫽ 20 ␮m. B, The percentage area of resorption was calculated for each subject. Bars are the mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus healthy controls. C, CD4⫹, CD8⫹, CD45RO⫹, and CD45RA⫹ cell subsets were magnetically sorted from PB mononuclear cells of 5 patients with early RA, and cocultures with autologous monocytes were established on Osteologic hydroxyapatite-coated slides. Shown is the percentage area of resorption for each subset. Each bar represents the arithmetic mean and SD of 5 patients with early RA. ⴱ ⫽ P ⬍ 0.05 versus CD4⫹ cell subset; † ⫽ P ⬍ 0.05 versus CD45RO⫹ cell subset, after Bonferroni correction for multiple comparisons. See Figure 1 for definitions.

cytes from healthy controls (MFI 4.1 ⫾ 0.6). CD69 expression was negligible in freshly isolated PB T lymphocytes from patients with early RA (1.9 ⫾ 0.5% positive cells, MFI 1.27 ⫾ 0.4) as was CD40L expression (0.5 ⫾ 0.02% positive cells, MFI 1.1 ⫾ 0.01). Expression of IL-15 on monocytes from RA patients. PB and SF monocytes of RA patients have been described to show signs of activation (34). Therefore, we were interested in examining if these cells express IL-15

Figure 4. Soluble cytokine concentrations in supernatants from unstimulated autologous T cell/monocyte cocultures, as determined by enzyme-linked immunosorbent assay. A, Concentrations of soluble IL-15, IL-17, RANKL, osteoprotegerin (OPG), tumor necrosis factor ␣ (TNF␣), IL-1␤, IL-4, and interferon-␥ (IFN␥) in supernatants of autologous T cell/monocyte cocultures derived from the PB of healthy controls (HCPB), PB of patients with early RA (RAPB), and SF of patients with established RA (RASF). The concentration of soluble IL-15 was ⬍10 pg/ml in all cocultures. Each bar represents the arithmetic mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus PB of healthy controls; † ⫽ P ⬍ 0.05 versus PB of patients with early RA. B, Calculated OPG:RANKL ratio in supernatants of autologous T cell/monocyte cocultures derived from the PB of healthy controls, PB of patients with early RA, and SF of patients with established RA. Each bar represents the arithmetic mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus PB of healthy controls. See Figure 1 for other definitions.


1157

Figure 5. Effect of osteoprotegerin (OPG) and neutralizing anticytokine antibodies on ex vivo osteoclastogenesis in T cell/monocyte cocultures. A, The number of tartrate-resistant acid phosphatase–positive (TRAP⫹) multinucleated cells (mncs) and the percentage area of resorption were determined in T cell/monocyte cocultures derived from the PB of patients with early RA and from the SF of patients with established RA in the presence or absence of OPG, anti–tumor necrosis factor ␣ (a-TNF␣), anti–IL-1␤, anti–IL-15, anti–IL-17, or isotype control monoclonal antibodies (mAb). Bars represent the mean and SD of 20 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus cells cultured in medium; † ⫽ P ⬍ 0.05 for cells cultured with a combination of neutralizing mAb versus cells cultured in the presence of only 1 of the mAb. The effect of isotype control antibodies was negligible (not included in the bar graphs). B, The number of TRAP⫹ multinucleated cells and the percentage area of resorption were determined in T cell/monocyte cocultures derived from the PB of healthy controls and maintained for 14 days with 50% RA PB T cell/monocyte coculture supernatant (Sp) and 50% normal medium or with 50% RA SF T cell/monocyte coculture supernatant and 50% normal medium. After 14 days, the number of TRAP⫹ multinucleated cells and the percentage of area resorbed were calculated. The effect of OPG, anti-TNF␣, anti–IL-1␤, anti–IL-15, anti–IL-17, or isotype control mAb was tested. Bars represent the mean and SD of 5 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus cells cultured in medium; † ⫽ P ⬍ 0.05 versus cells cultured with supernatant alone. The effect of isotype control antibodies was negligible (not included in the bar graphs). See Figure 1 for other definitions.

factor (Figure 2). In contrast, osteoclasts were rare in cocultures derived from healthy controls (Figure 2). To assess the bone-resorbing capacity of these cells, unstimulated T cell/monocyte cocultures were established on hydroxyapatite-coated disks. The mean percentage

area resorbed was significantly higher in cocultures derived from the PB of patients with early RA compared with cocultures derived from the PB of healthy controls, and was highest in cocultures derived from the SF of patients with established RA (Figure 3). These

1158

data demonstrate the formation of functional osteoclasts, capable of enhanced bone-resorbing activity, in RA T cell/monocyte cocultures. Importantly, after 14 days in culture, 95% of the T cells remained viable as assessed by annexin V and 7-aminoactinomycin D staining, indicating that viable T cells contribute to the observed osteoclastogenic effect throughout the coculture period. To determine which subsets of T lymphocytes are involved in osteoclastogenesis, CD4⫹, CD8⫹, CD45RO⫹, and CD45RA⫹ T cells were magnetically isolated from the PB of 5 patients with early RA, and these subsets were tested in coculture experiments. CD4⫹ T cells demonstrated a higher capacity to induce osteoclastogenesis compared with CD8⫹ T cells, and CD45RO⫹ T cells demonstrated a higher osteoclastogenic capacity compared with CD45RA⫹ T lymphocytes (Figures 2 and 3). Observed differences persisted when applying Bonferroni correction for multiple comparisons. Concentrations of TNF␣, IL-1␤, IL-17, RANKL, OPG, IL-4, and IFN␥ in supernatants were significantly higher in cocultures derived from the PB of patients with early RA compared with cocultures derived from the PB of healthy controls and were highest in cocultures derived from the SF of patients with established RA (Figure 4). However, the OPG:RANKL ratio was decreased in cocultures from RA PB and RA SF compared with cocultures from PB of healthy controls (Figure 4). No soluble IL-15 could be detected in coculture supernatants. Cocultures of T cells and monocytes separated by 0.4-␮m inserts did not result in significant osteoclastogenesis, suggesting that the role of soluble mediators released from resting cells was minor. Effect of OPG and neutralizing anticytokine antibodies on ex vivo osteoclastogenesis from T cell/monocyte cocultures. Ex vivo osteoclastogenesis was significantly inhibited in the presence of OPG-Fc (1 ␮g/ml), indicating that RANKL plays a critical role in this phenomenon (Figure 5A). In addition, neutralizing antibodies to TNF␣, IL-1␤, IL-15, and IL-17 (10 ␮g/ml) were effective at decreasing osteoclast formation in our system (Figure 5A). Since TNF␣ has been described to strongly synergize with trace amounts of RANKL (12), we tested the combination of OPG-Fc and a neutralizing anti-TNF␣ mAb, which further suppressed osteoclastogenesis (Figure 5A). Anti-TNF␣ and anti–IL-1␤ also demonstrated a synergistic antiosteoclastogenic effect (Figure 5A). An irrelevant mouse IgG1 isotype control mAb (BD PharMingen), a nonspecific human IgG, and


Figure 6. Effect of methotrexate (MTX) on ex vivo osteoclastogenesis observed in 14-day–unstimulated T cell/monocyte cocultures derived from the PB of untreated patients with early RA. MTX inhibited coculture-induced formation of tartrate-resistant acid phosphatase (TRAP)–positive multinucleated cells (mncs) (A) and resorption lacunae (B) in a dose-dependent manner. This effect was reversed by adenosine deaminase (ADA; 0.125 IU/ml) and by the adenosine A2A receptor antagonist ZM241385 (10 ␮M), but not by the adenosine A2B receptor antagonist MRS1706 (1 nM), and not by the adenosine A1 receptor antagonist 8-cyclopentyl-dipropylxanthine (DPCPX; 10 ␮M). Values represent the mean ⫾ SD of 20 patients. ⴱ ⫽ P ⬍ 0.05 versus conditions without MTX present. See Figure 1 for other definitions.

an anti–HLA class I mAb (W6/32) had no effect on osteoclast formation. Effect of soluble cytokines on ex vivo osteoclastogenesis from T cell/monocyte cocultures. Nonpooled supernatants harvested from 5 early RA PB cocultures elicited osteoclastogenesis when added ex vivo (one to one) to 5 healthy control cocultures (Figure 5B). Similar results were obtained when adding nonpooled supernatants from 5 established RA SF cocultures to 5 healthy control cocultures (one to one) (Figure 5B). The addi-


tion of OPG-Fc and of neutralizing antibodies to TNF␣, IL-1␤, and IL-17, but not to IL-15, were effective at decreasing supernatant-induced osteoclast formation (Figure 5B). These experiments indicate that cocultures derived from the PB of patients with early RA and from the SF of patients with established RA secrete significantly greater quantities of biologically active cytokines than do cocultures from PB of healthy controls, and are consistent with the absence of soluble IL-15 in supernatants. In vitro effect of MTX on osteoclastogenesis. Because MTX is still the most commonly used drug in RA and because most of our patients require MTX as a disease-modifying drug immediately after the diagnosis of RA is confirmed, the in vitro effect of MTX on osteoclastogenesis was tested. A dose-dependent inhibition was observed. Inhibition was already seen with MTX at a dose of 1 nM, and the effect was maximal with MTX at a dose of 100 nM (Figure 6). The effect of MTX on osteoclastogenesis was reversed by ADA (0.125 IU/ml), suggesting that adenosine release mediates inhibition of osteoclastogenesis (Figure 6). The effect of MTX was also reversed by the adenosine A2A receptor antagonist ZM241385 (10 ␮M), but not by the adenosine A2B receptor antagonist MRS 1706 (1 nM), and not by the adenosine A1 receptor antagonist DPCPX (10 ␮M) (Figure 6). This suggests that adenosine released by MTX acts through A2A receptors. Real-time RT-PCR performed 3 days after initiation of RA PB T cell/monocyte cocultures confirmed that coculture induced an up-regulation of T cell RANKL, IL-15, IL-17, and TNF␣ (Figure 7). At the same time, we observed up-regulation of monocyte IL-15, TNF␣, IL-1␤, OPG, and RANK (Figure 7), which contributed further to the observed osteoclastogenic effect. Up-regulation of T cell and monocyte cytokines did not take place in the presence of MTX at 100 nM (Figure 7). Treatment with MTX did not result in decreased cell viability as determined by propidium iodide and annexin V staining. In vivo effect of treatment. We also sought to determine whether disease control with treatment results in decreased ex vivo osteoclastogenesis in cocultures derived from the PB of patients with early RA. Sixteen patients with early RA donated blood for a second time 1 year after the initiation of treatment with oral MTX; 6 of these 16 patients were additionally receiving oral prednisone at a dosage of 2.5–5 mg/day. Blood was also obtained from the 16 healthy subjects who had previously acted as controls for these patients.

1159

Figure 7. Effect of methotrexate (MTX) on cytokine mRNA expression in T cell/monocyte cocultures. MTX (100 nM) significantly prevented the up-regulation of cytokine mRNA expression observed in both T cells (A) and monocytes (B) 3 days after initiation of autologous cocultures. Each bar represents the mean and SD of 20 patients. ⴱ ⫽ P ⬍ 0.05 versus conditions without MTX present. IL-15 ⫽ interleukin15; TNF␣ ⫽ tumor necrosis factor ␣; OPG ⫽ osteoprotegerin.

In 8 of these 16 patients, remission had been achieved, as defined by a DAS28 ⬍2.6 (40). The remaining 8 patients had experienced a significant clinical improvement, with a decrease in the DAS28 of ⱖ2.0 points, but they still demonstrated significant disease activity associated with a DAS28 ⬎2.6. Among patients with disease in remission, surface expression of T cell RANKL and IL-15 and of monocyte IL-15 was no longer detected (Figure 8A). At the same time, ex vivo differentiation of monocytes to osteoclasts was significantly decreased and not different from healthy controls (Figure 8B). Among patients in whom remission had not been achieved, these changes were less marked (Figures 8A and B). Of note, basal expression of RANKL and IL-15 did not differ significantly between patients who achieved remission and patients who experienced a significant clinical improvement but whose disease remained active (Figure 8A). Experimental variation observed in healthy controls was minimal. These results indicate that the basal activated state displayed by PB T cells and monocytes from patients with early RA is down-regulated in vivo by disease-modifying agents while these agents are controlling disease activity. At the same time, the number of

1160


Figure 8. Control of disease activity in patients with early RA results in down-regulation of surface RANKL and IL-15 expression, decreased ex vivo osteoclastogenesis, and decreased bone erosion. A and B, Sixteen patients with early RA donated blood for a second time 1 year after the initiation of treatment (t.) with oral methotrexate (MTX); 6 of these 16 patients were additionally receiving oral prednisone at 2.5–5 mg/day. Blood was also obtained from the 16 healthy subjects who had previously acted as controls for these patients. A, Surface expression of T cell RANKL and IL-15 and of monocyte IL-15 was examined. B, Cocultures of autologous T cells and monocytes were established, and osteoclast formation was assessed. Results for each patient were compared with results obtained before initiation of treatment. A, Surface T cell RANKL, T cell IL-15, and monocyte IL-15 expression was undetectable in 8 of these 16 patients who had achieved clinical remission with treatment. B, In parallel, no significant ex vivo osteoclastogenesis was observed in T cell/monocyte cocultures derived from the PB of these subjects. In contrast, surface RANKL and IL-15 (A) and a lower but persistent degree of osteoclastogenesis (B) were still detected in the remaining 8 patients, who had demonstrated a partial response to treatment. Bars represent the mean and SD of 8 subjects per group. ⴱ ⫽ P ⬍ 0.05 versus healthy controls; † ⫽ P ⬍ 0.05 versus the same group of patients at initial evaluation, before treatment was initiated. C, Box plots indicating the number of new bone erosions in radiographs of the hands and feet at 1-year followup in patients with early RA who achieved remission (n ⫽ 10) and in patients with early RA who experienced improvement with treatment (n ⫽ 10). Horizontal lines represent upper and lower limits of the interquartile range. Whiskers represent maximum and minimum values. ⴱ ⫽ P ⬍ 0.05 versus patients with disease in remission. Note that while the basal RANKL expression did not differ significantly between patients who achieved remission and patients who experienced a significant clinical improvement but continued to have active disease (A), the number of new erosions at 1 year was significantly greater in the latter group (B). MFI ⫽ mean fluorescence intensity; TRAP ⫽ tartrate-resistant acid phosphatase; mncs ⫽ multinucleated cells (see Figure 1 for other definitions).

new erosions in radiographs of the hands and feet at 1-year followup was significantly lower among 10 patients with early RA in whom remission was achieved compared with the remaining 10 patients who had experienced a significant clinical improvement (Figure 8C). We were particularly interested in establishing whether baseline expression of surface RANKL and/or IL-15 and/or ex vivo osteoclastogenesis score are independent predictors of erosive disease for patients with early RA. Importantly, after adjusting for RA baseline

disease activity and persistent disease activity, no correlation was observed between baseline surface RANKL expression, surface IL-15 expression, or ex vivo osteoclastogenesis score and the number of new erosions in radiographs of the hands and feet at 1-year followup. DISCUSSION The role of T cells in RA has been controversial (33), and the presence of T cell–derived cytokines in the PB and SF of RA patients has been difficult to demon-


strate (33). Here we have shown that freshly isolated PB T cells from patients with early RA and SF T cells from patients with established RA, but not PB T cells from healthy controls, demonstrate a significant surface expression of the recently described T cell–derived cytokines RANKL and IL-15. In addition, we observed that PB T cells from patients with early RA and SF T cells from patients with established RA, but not PB T cells from healthy controls, can induce osteoclastogenesis in cocultured autologous monocytes. RANKL-mediated osteoclastogenesis has been described in cultures of unstimulated PB mononuclear cells from patients with psoriatic arthritis (41), multiple myeloma (42), and postmenopausal osteoporosis (43), but not in patients with RA. RANKL expression in synovial tissue (6,7) and SF (44) T cells from RA patients has previously been reported, together with raised levels of the soluble form of RANKL (sRANKL) and decreased levels of OPG in RA SF (6). In addition, Kotake et al demonstrated that mitogen-activated human T cells induce RANKL-dependent osteoclast formation from human monocytes (6). Here we have extended this observation by showing for the first time that osteoclastogenesis in monocytes is driven by autologous freshly isolated T cells from the PB of patients with early RA and from the SF of patients with established RA, without additional exogenous stimulation. Monocyte/macrophages may serve both as osteoclast progenitors and as a source of osteoclastogenic cytokines (17,18). In fact, we observed that freshly isolated PB monocytes from patients with early RA and SF monocytes from patients with established RA, but not PB monocytes from healthy controls, demonstrated significant surface expression of IL-15, indicating an activated state that may facilitate stimulation of cocultured T cells and osteoclastogenesis. In addition, the higher levels of TNF␣ and IL-1␤ observed in cocultures of T cells/monocytes from RA patients compared with those from healthy controls reflect an increased monocyte cytokine production favored by the preactivated state of both RA T cells and RA monocytes. In our system, experiments with OPG-Fc and neutralizing antibodies demonstrate that RANKL plays a pivotal role in ex vivo T cell–mediated osteoclastogenesis, and the cytokines IL-17, IL-15, TNF␣, and IL-1␤ are important contributors that potentiate this effect. This is consistent with previous observations indicating a cooperation of cytokines in inflammation-mediated osteoclastogenesis (12,14,16). A low-level constitutive RANKL seems to be mandatory for osteoclastogenesis to take place, as indicated by experiments with RANKL-

1161

knockout mice (45), and the effect of even trace amounts of RANKL is potentiated by proinflammatory cytokines (46). In contrast, it has been reported that TNF␣ is able to induce osteoclastogenesis independently of RANKL (47). TNF␣ is an important player in inflammatory osteolysis (41,48) and synergistically cooperates with RANKL (46). In fact, TNF␣ induces RANKL synthesis by marrow stromal cells (46), and RANKL prompts TNF␣ expression by osteoclast precursors (17). In addition, the action of TNF␣ on RANKL expression has been described to be mediated by IL-1␤ (14). The hypothesis of cooperation among osteoclastogenic cytokines is supported by in vivo observations. Blockade of either IL-1␤ or TNF␣ does not completely arrest the periarticular erosions of inflammatory arthritis, whereas combined inhibition of both cytokines is significantly more effective (49). IL-17 is a novel cytokine produced by activated T cells (15,50), and elevated levels of IL-17 have been described in SF from RA patients but not in SF from patients with osteoarthritis (50). In addition, CD4⫹,CD45RO⫹ T cells in synovial tissue of RA patients are immunoreactive with anti–IL-17 antibodies (15). IL-17 contributes to osteoclastogenesis by altering the RANKL/OPG balance (16), has the capacity to induce joint destruction in an IL-1–independent manner, and can bypass TNF␣-dependent arthritis (51). Anti–IL-17 is of interest as a new therapeutic option for RA (16), particularly in patients in whom elevated IL-17 might attenuate the response to other biologics such as anti-TNF␣ and anti–IL-1␤ agents. Experiments with Transwell inserts indicate that direct cell contact is mandatory to initiate the T cell/ monocyte crosstalk resulting in osteoclastogenesis. Once intercellular crosstalk is initiated, sRANKL, IL-17, TNF␣, and IL-1␤ are liberated that contribute to augmenting the osteoclastogenic effect, while OPG acts to neutralize RANKL. A decreased OPG:RANKL ratio in our RA coculture supernatants is an additional contributor (3). Soluble RANKL is liberated by a TNF␣converting enzyme-like protease that cleaves surface T cell RANKL (52), although membrane-bound RANKL has been demonstrated to be significantly more effective than its soluble form (53). The observed cocultureinduced up-regulation of monocyte RANK, which mediates RANKL action on osteoclast precursors, further contributed to augmenting osteoclastogenesis. In our system, antiosteoclastogenic cytokines IL-4 (54,55) and IFN␥ (56) present in coculture supernatants of RA patients were not able to counterbalance the effect of proosteoclastogenic factors.

1162

Consistent with previous reports (23–25), no secretion of IL-15 could be detected in our coculture supernatants. IL-15 acts through a heterotrimeric receptor consisting of a specific high-affinity binding ␣-chain (IL-15R␣) plus the IL-2R ␤- and common ␥-chain that mediate signaling (23). The high affinity of IL-15R␣ conditions an extremely rapid uptake of secreted IL-15, preventing detection of IL-15 in culture supernatants (24). Most of the IL-15 detected on cell surfaces is bound to IL-15R␣ (24) and can stimulate in trans both ␤␥- and IL-15R␣␤␥–bearing cells (24). The presence of surface IL-15R␣–bound IL-15 is synonymous with active IL-15 secretion, and the level of expression of surface IL-15 in a given cell population may reflect the rate of internalization of the IL-15–IL-15R␣ complex. Thus, in contrast to IL-2, IL-15 can be expressed on the cell surface, where it is able to exert biologic functions through cell contact–dependent mechanisms (26–28). Although not detected initially (57), T cells were later shown by more sensitive techniques to express IL-15 mRNA (19,20) and protein (20,21), and Thurkow et al described IL-15 protein expression in synovial tissue T cells from RA patients (22). IL-15 has been shown to enhance osteoclast differentiation, whereas IL-2, which shares receptor components with IL-15, has no effect on osteoclastogenesis (13). In our system, surface IL-15 on RA T cells and RA monocytes appears to be an important contributor to the observed ex vivo osteoclastogenesis. MTX was effective at decreasing ex vivo osteoclast formation, consistent with results presented by Lee et al (58). MTX is the most commonly used drug in RA, and its mechanism of action is still being investigated (59). Low-dose MTX, as used for RA treatment, induces the release of adenosine to the extracellular space, and this autacoid seems to mediate the pharmacologic effect of MTX (37,59). In our system, experiments with adenosine receptor antagonists suggest that the effect of MTX on ex vivo osteoclastogenesis is mediated through adenosine, acting on A2A receptors. An in vivo antiresorptive effect of MTX has been demonstrated in rat adjuvant-induced arthritis (60). Clinical studies indicate that changes in radiographic progression in RA patients are directly related to fluctuations in disease activity (61). In fact, early and aggressive antirheumatic drug treatment affects the association of HLA class II alleles with progression of joint damage in RA (62). This interaction was independent of other prognostic factors, such as RF and baseline disease activity, suggesting that early and aggressive DMARD treatment can modify the dysregulated immune process


(62). Accordingly, we observed that the number of new erosions at 1 year was significantly lower in patients in whom disease remission was achieved. Our experiments with PB T cells and monocytes from patients with early RA who subsequently received oral MTX with or without low-dose prednisone showed that control of disease activity is associated with a down-regulation of surface RANKL and IL-15 together with decreased ex vivo osteoclastogenesis. At the same time, the baseline expression of surface T cell RANKL, T cell IL-15, and monocyte IL-15, together with baseline ex vivo osteoclastogenesis, were not associated with the number of new bone erosions at 1-year followup. This reflects the fact that treatment with DMARDs is able to modify the natural course of the disease and prevent joint destruction (62), and indicates that surface RANKL and IL-15 expression together with ex vivo osteoclastogenesis are markers of disease activity rather than independent predictors of radiographic progression. The results presented here extend our understanding of the pathogenesis of bone erosion in RA, and provide an experimental basis for the use of biologic anticytokine agents or their combinations in patients who do not respond to conventional therapy. REFERENCES 1. 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;152:943–51. 2. Teitelbaum SL. Bone resorption by osteoclasts. Science 2000;289: 1504–8. 3. Lacey DL, Timms E, Tan HL, Kelley MJ, Dunstan CR, Burgess T, et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 1998;93:165–76. 4. 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;43:250–8. 5. Shigeyama Y, Pap T, Kunzler P, Simmen BR, Gay RE, Gay S. Expression of osteoclast differentiation factor in rheumatoid arthritis. Arthritis Rheum 2000;43:2523–30. 6. Kotake S, Udagawa N, Hakoda M, Mogi M, Yano K, Tsuda E, et al. Activated human T cells directly induce osteoclastogenesis from human monocytes: possible role of T cells in bone destruction in rheumatoid arthritis patients. Arthritis Rheum 2001;44: 1003–12. 7. Dai SM, Nishioka K, Yudoh K. Interleukin (IL) 18 stimulates osteoclast formation through synovial T cells in rheumatoid arthritis: comparison with IL1␤ and tumour necrosis factor ␣. Ann Rheum Dis 2004;63:1379–86. 8. Horwood NJ, Kartsogiannis V, Quinn JM, Romas E, Martin TJ, Gillespie MT. Activated T lymphocytes support osteoclast formation in vitro. Biochem Biophys Res Commun 1999;265:144–50. 9. Romas E, Bakharevski O, Hards DK, Kartsogiannis V, Quinn JM, Ryan PF, et al. Expression of osteoclast differentiation factor at


10.

11.

12.

13.

14. 15.

16.

17.

18.

19.

20.

21.

22.

23. 24. 25. 26.

sites of bone erosion in collagen-induced arthritis. Arthritis Rheum 2000;43:821–6. Takayanagi H, Iizuka H, Juji T, Nakagawa T, Yamamoto A, Miyazaki T, et al. Involvement of receptor activator of nuclear factor ␬B ligand/osteoclast differentiation factor in osteoclastogenesis from synoviocytes in rheumatoid arthritis. Arthritis Rheum 2000;43:259–69. Takayanagi H, Oda H, Yamamoto S, Kawaguchi H, Tanaka S, Nishikawa T, et al. A new mechanism of bone destruction in rheumatoid arthritis: synovial fibroblasts induce osteoclastogenesis. Biochem Biophys Res Commun 1997;240:279–86. O’ Gradaigh D, Ireland D, Bord S, Compston JE. Joint erosion in rheumatoid arthritis: interactions between tumour necrosis factor ␣, interleukin 1, and receptor activator of nuclear factor ␬B ligand (RANKL) regulate osteoclasts. Ann Rheum Dis 2004;63:354–9. Ogata Y, Kukita A, Kukita T, Komine M, Miyahara A, Miyazaki S, et al. A novel role of IL-15 in the development of osteoclasts: inability to replace its activity with IL-2. J Immunol 1999;162: 2754–60. Wei S, Kitaura H, Zhou P, Ross FP, Teitelbaum SL. IL-1 mediates TNF-induced osteoclastogenesis. J Clin Invest 2005;115:282–90. Kotake S, Udagawa N, Takahashi N, Matsuzaki K, Itoh K, Ishiyama S, et al. IL-17 in synovial fluids from patients with rheumatoid arthritis is a potent stimulator of osteoclastogenesis. J Clin Invest 1999;103:1345–52. Lubberts E, Koenders MI, van den Berg WB. The role of T cell interleukin-17 in conducting destructive arthritis: lessons from animal models. Arthritis Res Ther 2005;7:29–37. Zou W, Hakim I, Tschoep K, Endres S, Bar-Shavir Z. Tumor necrosis factor-␣ mediates RANK ligand stimulation of osteoclast differentiation by an autocrine mechanism. J Cell Biochem 2001; 83:70–83. Kitaura H, Sands MS, Aya K, Zhou P, Hirayama T, Uthgenannt B, et al. Marrow stromal cells and osteoclast precursors differentially contribute to TNF-␣-induced osteoclastogenesis in vivo. J Immunol 2004;173:4838–46. Azimi N, Brown K, Bamford RN, Tagaya Y, Siebenlist U, Waldmann TA. Human T cell lymphotropic virus type I Tax protein trans-activates interleukin 15 gene transcription through an NF-␬B site. Proc Natl Acad Sci U S A 1998;95:2452–7. Miranda-Carus ME, Benito-Miguel M, Llamas MA, Balsa A, Martin-Mola E. Human T cells constitutively express IL-15 that promotes ex vivo T cell homeostatic proliferation through autocrine/juxtacrine loops. J Immunol 2005;175:3656–62. Neely GG, Robbins SM, Amankwah EK, Epelman S, Wong H, Spurrell JC, et al. Lipopolysaccharide-stimulated or granulocytemacrophage colony-stimulating factor-stimulated monocytes rapidly express biologically active IL-15 on their cell surface independent of new protein synthesis. J Immunol 2001;167:5011–7. Thurkow EW, van der Heijden IM, Breedveld FC, Smeets TJ, Daha MR, Kluin PM, et al. Increased expression of IL-15 in the synovium of patients with rheumatoid arthritis compared with patients with Yersinia-induced arthritis and osteoarthritis. J Pathol 1997;181:444–50. Giri JG, Ahdieh M, Eisenman J, Shanebeck K, Grabstein K, Kumaki S, et al. Utilization of the ␤ and ␥ chains of the IL-2 receptor by the novel cytokine IL-15. EMBO J 1994;13:2822–30. Dubois S, Mariner J, Waldmann TA, Tagaya Y. IL-15R␣ recycles and presents IL-15 in trans to neighboring cells. Immunity 2002; 17:537–47. Lodolce JP, Burkett PR, Koka RM, Boone DL, Ma A. Regulation of lymphoid homeostasis by interleukin-15. Cytokine Growth Factor Rev 2002;13:429–39. Briard D, Brouty-Boye D, Azzarone B, Jasmin C. Fibroblasts from human spleen regulate NK cell differentiation from blood CD34(⫹) progenitors via cell surface IL-15. J Immunol 2002;168: 4326–32.

1163

27. Rappl G, Kapsokefalou A, Heuser C, Rossler M, Ugurel S, Tilgen W, et al. Dermal fibroblasts sustain proliferation of activated T cells via membrane-bound interleukin-15 upon long-term stimulation with tumor necrosis factor-␣. J Invest Dermatol 2001;116: 102–9. 28. Miranda-Carus ME, Balsa A, Benito-Miguel M, Perez de Ayala C, Martin-Mola E. IL-15 and the initiation of cell contact-dependent synovial fibroblast-T lymphocyte cross-talk in rheumatoid arthritis: effect of methotrexate. J Immunol 2004;173:1463–76. 29. McInnes IB, al-Mughales J, Field M, Leung BP, Huang FP, Dixon R, et al. The role of interleukin-15 in T-cell migration and activation in rheumatoid arthritis. Nat Med 1996;2:175–82. 30. McInnes IB, Gracie JA, Harnett M, Harnett W, Liew FY. New strategies to control inflammatory synovitis: interleukin 15 and beyond. Ann Rheum Dis 2003;62 Suppl 2:ii51–4. 31. Ferrari-Lacraz S, Zanelli F, Neuberg M, Donskoy F, Kim YS, Hancock WW, et al. Targeting IL-15 receptor-bearing cells with an antagonist mutant IL-15/Fc protein prevents disease development and progression in murine collagen-induced arthritis. J Immunol 2004;173:5818–26. 32. Baslund B, Tvede N, Danneskiold-Samsoe B, Larsson P, Panayi G, Petersen J, et al. Targeting interleukin-15 in patients with rheumatoid arthritis: a proof-of-concept study. Arthritis Rheum 2005; 52:2686–92. 33. Fox DA. The role of T cells in the immunopathogenesis of rheumatoid arthritis: new perspectives [review]. Arthritis Rheum 1997;40:598–609. 34. Stuhlmuller B, Ungethum U, Scholze S, Martinez L, Backhaus M, Kraetsch HG, et al. Identification of known and novel genes in activated monocytes from patients with rheumatoid arthritis. Arthritis Rheum 2000;43:775–90. 35. 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;31:315–24. 36. 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;38:44–8. 37. Cronstein BN, Eberle MA, Gruber HE, Levin RI. Methotrexate inhibits neutrophil function by stimulating adenosine release from connective tissue cells. Proc Natl Acad Sci U S A 1991;88:2441–5. 38. Gadangi P, Longaker M, Naime D, Levin RI, Recht PA, Montesinos MC, et al. The anti-inflammatory mechanism of sulfasalazine is related to adenosine release at inflamed sites. J Immunol 1996;156:1937–41. 39. Van der Heijde D, van Riel PL, Nuver-Zwart HH, Gribnau FW, van de Putte LB. Effects of hydroxychloroquine and sulphasalazine on progression of joint damage in rheumatoid arthritis. Lancet 1989;8646:1036–8. 40. Balsa A, Carmona L, Gonzalez-Alvaro I, Belmonte MA, Tena X, Sanmarti R, and the EMECAR Study Group. Value of disease activity score 28 (DAS28) and DAS28-3 compared to American College of Rheumatology-defined remission in rheumatoid arthritis. J Rheumatol 2004;31:40–6. 41. 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: 821–31. 42. Colucci S, Brunetti G, Rizzi R, Zonno A, Mori G, Colaianni G, et al. T cells support osteoclastogenesis in an in vitro model derived from human multiple myeloma bone disease: the role of the OPG/TRAIL interaction. Blood 2004;104:3722–30. 43. D’Amelio P, Grimaldi A, Pescarmona GP, Tamone C, Roato I, Isaia G. Spontaneous osteoclast formation from peripheral blood


1164

44.

45.

46.

47.

48.

49.

50.

51.

52.

mononuclear cells in postmenopausal osteoporosis. FASEB J 2005;19:410–2. Kanamaru F, Iwai H, Ikeda T, Nakajima A, Ishikawa I, Azuma M. Expression of membrane-bound and soluble receptor activator of NF-␬B ligand (RANKL) in human T cells. Immunol Lett 2004;94: 239–46. Pettit AR, Ji H, von Stechow D, Muller R, Goldring SR, Choi Y, et al. TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis. Am J Pathol 2001;159:1689–99. Lam J, Takeshita S, Barker JE, Kanagawa O, Ross FP, Teitelbaum SL. TNF-␣ induces osteoclastogenesis by direct stimulation of macrophages exposed to permissive levels of RANK ligand. J Clin Invest 2000;106:1481–8. Kobayashi K, Takahashi N, Jimi E, Udagawa N, Takami M, Kotake S, et al. Tumor necrosis factor ␣ stimulates osteoclast differentiation by a mechanism independent of the ODF/RANKLRANK interaction. J Exp Med 2000;191:275–86. 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;110:1419–27. Zwerina J, Hayer S, Tohidast-Akrad M, Bergmeister H, Redlich K, Feige U, et al. Single and combined inhibition of tumor necrosis factor, interleukin-1, and RANKL pathways in tumor necrosis factor–induced arthritis: effects on synovial inflammation, bone erosion, and cartilage destruction. Arthritis Rheum 2004;50: 277–90. Ziolkowska M, Koc A, Luszczykiewicz G, Ksiezopolska-Pietrzak K, Klimczak E, Chwalinska-Sadowska H, et al. High levels of IL-17 in rheumatoid arthritis patients: IL-15 triggers in vitro IL-17 production via cyclosporin A-sensitive mechanism. J Immunol 2000;164:2832–8. Koenders MI, Lubberts F, Oppers-Walgreen B, van den Bersselaar L, Helsen MM, Kolls JK, et al. Induction of cartilage damage by overexpression of T cell interleukin-17A in experimental arthritis in mice deficient in interleukin-1. Arthritis Rheum 2005;52: 975–83. Lum L, Wong BR, Josien R, Becherer JD, Erdjument-Bromage H, Schlondorff J, et al. Evidence for a role of a tumor necrosis factor-␣ (TNF␣)-converting enzyme-like protease in shedding of

53.

54.

55. 56. 57.

58.

59. 60.

61.

62.

TRANCE, a TNF family member involved in osteoclastogenesis and dendritic cell survival. J Biol Chem 1999;274:13613–8. Nakashima T, Kobayashi Y, Yamasaki S, Kawakami A, Eguchi K, Sasaki H, et al. Protein expression and functional difference of membrane-bound and soluble receptor activator of NF-␬B ligand: modulation of the expression by osteotropic factors and cytokines. Biochem Biophys Res Commun 2000;275:768–75. Wei S, Wang MW, Teitelbaum SL, Ross FP. Interleukin-4 reversibly inhibits osteoclastogenesis via inhibition of NF-␬B and mitogen-activated protein kinase signaling. J Biol Chem 2002;277: 6622–30. Mirosavljevic D, Quinn JM, Elliott J, Horwood NJ, Martin TJ, Gillespie MT. T-cells mediate an inhibitory effect of interleukin-4 on osteoclastogenesis. J Bone Miner Res 2003;18:984–93. Takayanagi H, Ogasawara K, Hida S, Chiba T, Murata S, Sato K, et al. T-cell-mediated regulation of osteoclastogenesis by signalling cross-talk between RANKL and IFN␥. Nature 2000;408:600–5. Grabstein KH, Eisenman J, Shanebeck K, Rauch C, Srinivasan S, Fung V, et al. Cloning of a T cell growth factor that interacts with the beta chain of the interleukin-2 receptor. Science 1994;264: 965–8. Lee CK, Lee EY, Chung SM, Mun SH, Yoo B, Moon HB. Effects of disease-modifying antirheumatic drugs and antiinflammatory cytokines on human osteoclastogenesis through interaction with receptor activator of nuclear factor ␬B, osteoprotegerin, and receptor activator of nuclear factor ␬B ligand. Arthritis Rheum 2004;50:3831–43. Chan ES, Cronstein BN. Molecular action of methotrexate in inflammatory diseases. Arthritis Res 2002;4:266–73. Segawa Y, Yamaura M, Aota S, Omata T, Tuzuike N, Itokazu Y, et al. Methotrexate maintains bone mass by preventing both a decrease in bone formation and an increase in bone resorption in adjuvant-induced arthritic rats. Bone 1997;20:457–64. Welsing PM, Landewe RB, van Riel PL, Boers M, van Gestel AM, van der Linden S, et al. The relationship between disease activity and radiologic progression in patients with rheumatoid arthritis: a longitudinal analysis. Arthritis Rheum 2004;50:2082–93. Lard LR, Boers M, Verhoeven A, Vos K, Visser H, Hazes JM, et al. Early and aggressive treatment of rheumatoid arthritis patients affects the association of HLA class II antigens with progression of joint damage. Arthritis Rheum 2002;46:899–905.