Development of spontaneous multisystem autoimmune disease and ...

ARTHRITIS & RHEUMATISM Vol. 52, No. 10, October 2005, pp 3220–3229 DOI 10.1002/art.21344 © 2005, American College of Rheumatology

Development of Spontaneous Multisystem Autoimmune Disease and Hypersensitivity to Antibody-Induced Inflammation in Fc␥ Receptor IIa–Transgenic Mice Caroline Tan Sardjono,1 Patricia L. Mottram,1 Nicholas C. van de Velde,1 Maree S. Powell,1 David Power,1 Ronald F. Slocombe,2 Ian P. Wicks,3 Ian K. Campbell,3 Steven E. McKenzie,4 Mark Brooks,5 Andrew W. Stevenson,6 and P. Mark Hogarth1 transgenic C57BL/6 (H-2b) mice did not develop CIA when similarly immunized. Passive transfer of a single dose of anti-CII antibody induced a more rapid, severe arthritis in Fc␥RIIa-transgenic mice than in nontransgenic animals. In addition, most immune complex– induced production of tumor necrosis factor ␣ by activated macrophages occurred via Fc␥RIIa, not the endogenous mouse FcR. A spontaneous, multisystem autoimmune disease developed in aging (>20 weeks) transgenic mice (n ⴝ 25), with a 32% incidence of arthritis, and by 45 weeks, all mice had developed glomerulonephritis and pneumonitis, and most had antihistone antibodies. Elevated IgG2a levels were seen in mice with CIA and in those with spontaneous disease. Conclusion. The presence of enhanced passive and induced autoimmunity, as well as the emergence of spontaneous autoimmune disease at 20–45 weeks of age, suggest that Fc␥RIIa is a very important factor in the pathogenesis of autoimmune inflammation and a possible target for therapeutic intervention.

Objective. The major human Fc receptor, Fc␥RIIa, is the most widespread activating FcR. Our aim was to determine the role of Fc␥RIIa in a transgenic mouse model of immune complex–mediated autoimmunity and to characterize the development of spontaneous autoimmune disease. Methods. Arthritis was induced in normal and Fc␥RIIa-transgenic mice by immunization with type II collagen (CII) or by transfer of arthritogenic anti-CII antibodies. Also, mice that spontaneously developed autoimmune disease were assessed by clinical scoring of affected limbs, histology and serology, and measurement of autoantibody titers and cytokine production. Results. Fc␥RIIa-transgenic mice developed collagen-induced arthritis (CIA) more rapidly than did archetypal CIA-sensitive DBA/1 (H-2q) mice, while nonSupported by grants from the National Health and Medical Research Council and PrimaBiomed Ltd., Australia. Dr. Sardjono’s work was supported by PaperlinX Pty Ltd. Drs. Mottram and Powell’s work was supported by Nancy Prendergast fellowships from the Arthritis Foundation, Australia. 1 Caroline Tan Sardjono, PhD, Patricia L. Mottram, PhD, Nicholas C. van de Velde, BSc Hons, Maree S. Powell, PhD, David Power, PhD, P. Mark Hogarth, PhD: Austin Research Institute, Heidelberg, Victoria, Australia; 2Ronald F. Slocombe, PhD: University of Melbourne, Melbourne, Victoria, Australia; 3Ian P. Wicks, PhD, Ian K. Campbell, PhD: Walter and Eliza Hall Institute, Parkville, Victoria, Australia; 4Steven E. McKenzie, PhD: Jefferson Medical College, Philadelphia, Pennsylvania; 5Mark Brooks, MBBS: Austin Hospital, Heidelberg, Victoria, Australia; 6Andrew W. Stevenson, PhD: Commonwealth Scientific Industrial Research Organization, Clayton South, Victoria, Australia. Drs. Mottram, Powell, and Hogarth have stock options in PrimaBiomed. Dr. McKenzie has received consulting fees (less than $10,000 per year) from GlaxoSmithKline. Address correspondence and reprint requests to P. Mark Hogarth, PhD, Helen McPherson-Smith Laboratory, Austin Research Institute, Studley Road, Heidelberg, Victoria 3084, Australia. E-mail: [email protected]. Submitted for publication September 24, 2004; accepted in revised form June 30, 2005.

Antibody-induced inflammation is a major component of several autoimmune diseases (1,2). The role of cell surface receptors for antibodies, especially IgG Fc␥ receptors (Fc␥R), was recognized following amelioration of tissue destruction in type III hypersensitivity reactions after administration of soluble recombinant human Fc␥RIIa in vivo (3). Subsequent studies with FcR-deficient mice (4) showed that Fc␥R play significant roles in antibody-induced inflammatory disease models such as collagen-induced arthritis (CIA) (5), passive antibody-induced arthritis (6), and intraarticular antigen-induced arthritis (7). However, rodents lack an ortholog of Fc␥RIIa, the most abundant and widespread 3220

ROLE OF Fc␥RIIa IN AUTOIMMUNE INFLAMMATION

activating FcR in higher primates. Fc␥RIIa has unique structural, signaling, and biologic features (8–11). Unlike other FcR, Fc␥RIIa can signal without the homodimeric FcR ␥-chain used by Fc␧RI, Fc␥RI, Fc␥RIII, and Fc␣RI, since both the ligand binding site and the immunoreceptor tyrosine-based activation motif (ITAM) are in the same polypeptide (8). Moreover, Fc␥RIIa is a dimer, with the ITAM-containing cytoplasmic tails arranged in an FcR ␥-chain–like configuration (9,12). Studies of Fc␥RIIa, as well as other FcR, transfected into mouse or primate cells show that these FcR behave identically in both ligand binding and activation/ regulation (13,14). The interaction of Fc␥RII ITAMs and immunoreceptor tyrosine-based inhibition motifs (ITIMs) was seen in both transfected mouse and human cell lines, and ITIM sequences in mice and humans are highly conserved (15). Fc␥RIIa in transgenic mice is expressed under its own promoter and has the same expression pattern in mice and humans (16,17). Thus, Fc␥RIIa can interact appropriately with intracellular signaling pathways in mouse cells. Finally, genetic polymorphisms of Fc␥RIIa are associated with human autoimmune disease (1,18). In this study, we analyzed inflammatory responses in transgenic mice expressing Fc␥RIIa and confirmed a role for this receptor in passive, induced, and spontaneous autoimmune disease. MATERIALS AND METHODS Mice. DBA/1 (H-2q), C57BL/6 (H-2b), SJL/J (H-2s), (SJL ⫻ C57BL/6)F1 (H-2b/s), and Fc␥RIIa-transgenic mice (H-2b) derived from (SJL ⫻ C57BL/6)F2 embryos (17) were used. The Fc␥RIIa-transgenic mice were inbred for ⬎20 generations and were homozygous for the transgene under the control of its own promoter. They carried the high responder (Arg134) allele of Fc␥RIIa, which binds mouse IgG2a, Ig2b, and Ig1, as well as human IgG1, IgG2, and IgG3 (11). Induction of CIA. Mice were injected intradermally at the base of the tail with 100 ␮l of 2 mg/ml type II collagen (CII) emulsion in Freund’s complete adjuvant (Difco, Detroit, MI) that contained 2.5 mg/ml heat-killed Mycobacterium tuberculosis H37Ra (Difco). Mice were immunized a second time 21 days later (19). They were examined daily from days 1–60, and arthritis in each limb was graded on a scale of 0–3 (0 ⫽ normal, 1 ⫽ mild swelling and redness, 2 ⫽ severe swelling/redness, and 3 ⫽ severe swelling and redness and joint rigidity). The maximum possible score (arthritis index) was 12 for each mouse. Passively induced arthritis. Anti-CII monoclonal antibody (mAb) M2139 (2 mg) (20) was injected intraperitoneally into Fc␥RIIa-transgenic and nontransgenic C57BL/6 mice, and arthritis progression was monitored daily as described above. Joint histology. Joints were preserved in 10% formalin/ phosphate buffered saline (PBS), decalcified in 5% HCl, 3.5%

3221

glacial acetic acid, 95% ethanol, and 12.5% (volume/volume) chloroform, and then were embedded in paraffin. Sections (4–6 ␮m) were stained with hematoxylin and eosin (H&E). Enzyme-linked immunosorbent assay (ELISA) for anti-CII antibody. ELISA plates (96-well; Costar, Cambridge, MA) were coated with 50 ␮g/ml CII and blocked with 2% bovine serum albumin in PBS (1 hour at room temperature). Sera (serially diluted) were added, and antibody was detected using secondary horseradish peroxidase (HRP)–conjugated sheep anti-mouse IgG F(ab⬘)2 fragments (Amersham, Little Chalfont, UK). Development was performed for 10 minutes with ABTS (Boehringer Mannheim, Rockville, MD), and absorbance was read at 405 nm. IgG isotypes were detected with specific antibodies (see below). Bone marrow macrophage cultures. Bone marrow macrophages were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (CSL, Melbourne, Victoria, Australia) containing 2 mM glutamine, 50 mM 2-mercaptoethanol, 100 units/ml penicillin, 100 ␮g/ml streptomycin, 10% (v/v) fetal calf serum, and 30% (v/v) L cell–conditioned medium at 37°C in 10% CO2 for 5–7 days (21). Peritoneal exudate macrophage (PEM) cultures. Mice were injected intraperitoneally with 0.5 ml of 4% (weight/ volume) thioglycolate (CSL) 4 days prior to harvest. PEMs were cultured in DMEM, prepared as described above, for 1–3 days. Measurement of cytokine production. PEMs (1 ⫻ 106/ml) were incubated with heat-aggregated IgG (HAGG; 50 ␮g or 100 ␮g) or phorbol myristate acetate (PMA; 20 ng) (Sigma, St. Louis, MO) at 37°C for 24 hours. HAGG was prepared from 8–10 mg/ml of human gamma globulin (Sandoglobulin; Novartis, East Hanover, NJ) heated for 30 minutes at 63°C, then put on ice and brought to 1% (w/v) with polyethylene glycol 6000 (Sigma) in PBS and kept on ice for 30 minutes. The precipitated complexes were centrifuged (10,000g at 4°C for 10 minutes), the supernatant was discarded, and complexes were dissolved in PBS at 1 mg/ml. Production of tumor necrosis factor ␣ (TNF␣) and interleukin-10 (IL-10) by macrophages in the supernatant 2–24 hours after stimulation with HAGG or PMA was detected by ELISA using cytokine detection kits from eBioscience (San Diego, CA). The specificity of induced TNF␣ secretion was determined by preincubation of macrophages with antiFc␥RIIa mAb IV.3 Fab (20 or 50 ␮g) at 4°C for 1 hour. Detection of Fc␥R by flow cytofluorometry. The following Fc␥R were detected: mouse Fc␥RI using mAb X54-5/ 7.1 (mIgG1) (22), mouse Fc␥RIIb using mAb Ly-17.2 (mIgG2; Cedarlane Laboratories, Hornby, Ontario, Canada), and hFc␥RIIa using mAb IV.3 (mIgG2b) or mAb 8.7 (mIgG1). Monitoring of spontaneous autoimmunity. Mice were examined from 12 weeks of age, and the index of arthritis severity was assessed as described above. Organs were fixed in 10% formalin/PBS, embedded in paraffin, and histologic sections were stained with H&E. Immune complex deposition in kidney sections was detected with fluorescein isothiocyanate (FITC)–conjugated sheep anti-mouse IgG (Silenus Laboratories, Hawthorn, Victoria, Australia). Transmission electron microscopy. Mouse kidneys were cut into 1–2-mm cubes and fixed in 2–8% paraformalde-

3222

SARDJONO ET AL

hyde/2–5% glutaraldehyde in 0.15M cacodylate buffer (pH 7.4) for ⬎6 hours at 4°C. Tissues were rinsed in cacodylate buffer and postfixed in 1% osmium tetroxide in 0.15M cacodylate buffer (pH 7.4) for 2 hours at room temperature. Samples were washed in distilled water, dehydrated in 10% incremental concentrations of acetone, and then embedded in Procure–Araldite resin (ProSci Tech, Thuringowa, Queensland, Australia). During the dehydration, tissues were stained with 2% uranyl acetate in 70% acetone. Ultrathin sections were stained with 5% uranyl acetate in aqueous solution for 30 minutes at room temperature and then were stained with Reynolds lead citrate for 10 minutes. Sections were examined using a Philips 300 electron microscope (Philips, Mahwah, NJ) at 60 kV. Radiographic analysis. Polychromatic hard radiography was performed (23) on mice, and images were recorded using an FXE-225.20 microfocus x-ray source (Feinfocus, Stamford, CT) fitted with a tungsten target, operated at 60 kVp. Projection geometry with a large sample-to-detector distance resulted in high-resolution images. Exposures were ⬃3 mA, and images were recorded with imaging plates (XRT, Port Melbourne, Victoria, Australia) scanned with a Fuji BAS-5000 scanner (Fuji Photo Film, Tokyo, Japan). Detection of antihistone antibodies. ELISA plates were coated with 20 ␮g/ml purified histone (a mixture of H1, H2A, H2B, H3, and H4 from calf thymus) (Roche Laboratories, Basel, Switzerland). Serially diluted samples (50 ␮l) were added, incubated for 1 hour at room temperature, detected with HRP-conjugated sheep anti-mouse IgG F(ab⬘)2 fragments (Amersham), and then developed using ABTS. Quantitation of IgG subclasses. Total serum IgG concentrations were determined using ELISA plates coated with 50 ␮l (3 ␮g/ml) of rat anti-mouse IgG (BD PharMingen, San Diego, CA). Serially diluted serum samples (50 ␮l) were added and incubated for 1 hour at room temperature. Antibody was detected with biotin-conjugated, isotype-specific rat anti-mouse IgG1, anti-mouse IgG2a, and anti-mouse IgG2b mAb (BD PharMingen) (1 hour at room temperature), then with streptavidin–HRP (1 hour at room temperature), and developed with ABTS. Quantitation of IgG was by comparison with class-specific IgG standards. Statistical analysis. Results were expressed as the mean ⫾ SD. Statistical differences were analyzed using the Mann-Whitney 2-sample rank test, correlation coefficients, or Student’s t-test. All the statistical analyses were performed using Microsoft Excel software (Microsoft, Redmond, WA). P values less than 0.05 were considered statistically significant.

RESULTS Expression of transgenic and endogenous FcR. The expression of Fc␥R on bone marrow macrophages from Fc␥RIIa-transgenic and nontransgenic C57BL/6 mice was analyzed by flow cytometry. Using mAb IV.3 (anti-Fc␥RIIa) (Figure 1A), Fc␥RIIa was detected on macrophages from Fc␥RIIa-transgenic mice but not on macrophages from nontransgenic mice. The expression

Figure 1. Flow cytometric analysis of Fc␥ receptor (Fc␥R) expression in normal and human Fc␥RIIa–transgenic mice. Bone marrow macrophages from Fc␥RIIa-transgenic mice (solid histograms) or nontransgenic C57BL/6 mice (dotted line, open histograms) were stained for A, Fc␥RIIa with monoclonal antibody (mAb) IV.3, B, Fc␥RI with mAb X54-5/7.1, and C, Fc␥RIIb with mAb Ly-17.2. Studies were performed with fluorescein isothiocyanate–conjugated anti-mouse IgG (solid line, open histogram). FL-1 ⫽ fluorescence channel 1.

of the transgene did not significantly change the expression of endogenous mouse FcR, since similar levels of Fc␥RI (Figure 1B) and especially inhibitory Fc␥RIIb (Figure 1C) were observed on Fc␥RIIa-transgenic and nontransgenic macrophages. Similarly, expression of Fc␥RIIa on neutrophils did not alter endogenous FcR expression. Fc␥RIIa was not expressed on B or T cells (data not shown), as expected (16,17). Fc␥RIIa confers susceptibility to CIA. In mice, CIA susceptibility is associated with the major histocompatibility complex (MHC) genotype: H-2q and H-2r mice are highly susceptible, while H-2b, H-2d, and H-2s mice are less so (19,24). We compared CIA development in Fc␥RIIa-transgenic (H-2b) mice with the archetypal susceptible strain DBA/1 (H-2q) and the less susceptible C57BL/6 (H-2b) and (SJL ⫻ C57BL/6)F1 (H-2s/b) strains


Figure 2. Analysis of collagen-induced arthritis (CIA) in Fc␥RIIatransgenic mice. A, Development of CIA in Fc␥RIIa-transgenic mice (n ⫽ 41) compared with nontransgenic DBA/1 mice (n ⫽ 27), C57BL/6 mice (n ⫽ 28), and (SJL ⫻ B6)F1 mice (n ⫽ 8). Pooled data from 5 experiments are shown. B, Effects of treatment of arthritis with anti-Fc␥RIIa F(ab⬘)2 mAb 8.7 (intraperitoneal injection of 100 ␮g on days 21, 24, 27, and 30 versus treatment with phosphate buffered saline [PBS] alone [n ⫽ 8 mice per group]). The mAb 8.7 treatment caused a significant reduction in disease severity on day 40 (P ⬍ 0.05). Values in A and B are the mean ⫾ SD. C, A representative section showing histopathologic features of an Fc␥RIIa-transgenic mouse knee joint 36 days after CIA induction (hematoxylin and eosin stained; original magnification ⫻ 100). D, Histopathologic features of a knee joint from a C57BL/6 mouse 36 days after collagen injection (hematoxylin and eosin stained; original magnification ⫻ 40). See Figure 1 for other definitions.

(Figure 2A). The Fc␥RIIa-transgenic mice in this experiment were highly susceptible, with more rapid onset of arthritis (day 18) compared with other mice (days 22 and 24). Of the Fc␥RIIa-transgenic mice, 15% developed arthritis after 1 immunization, whereas 2 immunizations were always required for CIA development in DBA/1 and (SJL ⫻ C57BL/6)F1 mice. No arthritis occurred in C57BL/6 mice under these conditions. By day 26, ⬎90% of the Fc␥RIIa-transgenic mice developed arthritis, compared with ⬍10% of DBA/1 and (SJL ⫻ C57BL/ 6)F1 mice (P ⬍ 0.001). Disease incidence and severity at ⬎30 days were similar in Fc␥RIIa-transgenic and DBA/1 mice. In Fc␥RIIa-transgenic mice treated with a limited course of anti-Fc␥RIIa F(ab⬘)2 antibody (100 ␮g/mouse intraperitoneally on days 21, 24, 27, and 30), CIA was significantly reduced on day 35 (P ⬍ 0.05) compared with untreated CIA in Fc␥RIIa-transgenic mice. Although this dose of antibody is unlikely to have blocked all in

3223

vivo activity of Fc␥RIIa, the data show that this FcR plays a role in disease severity in these mice (Figure 2B). Histologic assessment of paws and ankle and knee joints from Fc␥RIIa-transgenic mice on day 36 after arthritis induction (when the maximal clinical index was observed) showed severe, destructive arthritis with pannus formation, infiltration of inflammatory cells (polymorphonuclear cells and macrophages) into the synovial space, and erosion of the cartilage (Figure 2C). No joint inflammation was seen in C57BL/6 controls (Figure 2D). Correlation of increased levels of IgG2a with the arthritis index in transgenic mice. Despite the accelerated development of severe CIA, Fc␥RIIa-transgenic mice had lower anticollagen antibody titers than susceptible DBA/1 mice, but similar to those of the less susceptible (SJL ⫻ C57BL/6)F1 mice (Figure 3A). The anticollagen antibodies were predominantly IgG2, and analysis of the serum IgG2 subclasses (Figure 3B) showed a significant increase in the IgG2a:IgG2b ratio in severely arthritic mice (P ⬍ 0.02) compared with mildly arthritic or unaffected Fc␥RIIa-transgenic mice. There was a positive correlation between the arthritis index and IgG2a levels (r ⫽ 0.57) and no correlation between arthritis index and IgG2b levels (r ⫽ ⫺0.16). This suggests a dominant Th1 response in Fc␥RIIatransgenic mice following induction of CIA and is consistent with previous data obtained from DBA/1 mice (25). Thus, expression of Fc␥RIIa, which binds mouse IgG2a avidly, may cause effector cells in the Fc␥RIIatransgenic mice to be sensitive to low levels of autoantibody/immune complex activation, possibly triggering the early release of inflammatory mediators. Production of TNF␣ by IgG-stimulated macrophages from Fc␥RIIa-transgenic mice. Because TNF␣ is a major inflammatory mediator in human RA and mouse CIA (24,26,27), we compared the production of TNF␣ from immune complex–stimulated Fc␥RIIatransgenic and nontransgenic macrophages. HAGGstimulated Fc␥RIIa macrophages produced significantly more TNF␣ (6.5 ng/ml) compared with nontransgenic macrophages (1 ng/ml; P ⫽ 0.001) (Figure 3C). Elevated TNF␣ production by Fc␥RIIa macrophages was blocked (although not completely) in a dose-dependent manner by anti-Fc␥RIIa mAb IV.3 Fab to levels not significantly different from those produced by nontransgenic macrophages, indicating that immune complex–induced TNF␣ production was due principally to activation by Fc␥RIIa, rather than through the endogenous mouse activating Fc␥R. PMA stimulation of nontransgenic and transgenic macrophages showed that these cells were equally responsive (Figure 3C). No difference in IL-10 production

3224

Figure 3. Sensitivity of human Fc␥ receptor IIa (Fc␥RIIa)–transgenic mice to antibody-mediated disease. A, Anticollagen antibody titers in Fc␥RIIa-transgenic (Tg) and nontransgenic (SJL ⫻ C57BL/6)F1 and DBA/1 mice (n ⫽ 5 per group), as determined on days 0, 21, and 36 postimmunization. OD ⫽ optical density. B, Comparison of changes in IgG2 subclasses in Fc␥RIIa-transgenic mice without collagen-induced arthritis (CIA) (index 0), with severe arthritis (index 5–12; high CIA), and with mild arthritis (index 1–4; low CIA) after the second injection of collagen (n ⫽ 5 mice per group). In the high CIA group, there was a significant increase in the IgG2a:IgG2b ratio (P ⫽ 0.016 versus mildly arthritic mice, by Student’s t-test), a positive correlation between the arthritis index and IgG2a levels (r ⫽ 0.57), and no correlation between the arthritis index and IgG2b levels (r ⫽ ⫺0.16). C, Comparison of tumor necrosis factor ␣ (TNF␣) production in heat-aggregated IgG (HAGG)–stimulated macrophages. Macrophages from Fc␥RIIatransgenic mice produced significantly higher levels of TNF␣ compared with nontransgenic C57BL/6 (B6) mice (P ⫽ 0.001) following stimulation with HAGG. There was no significant difference (P ⫽ 0.07) between Fc␥RIIa-transgenic and C57BL/6 mouse macrophages treated with HAGG following incubation with 20- or 50-␮g doses of monoclonal antibody (mAb) IV.3. Responses to phorbol myristate acetate (PMA) stimulation were equivalent in both strains of macrophages (P ⬎ 0.05). D, Arthritis index over time in Fc␥RIIa-transgenic and (SJL ⫻ C57BL/6)F1 mice treated with 2 mg of mAb M2139 given intraperitoneally on day 0 (n ⫽ 6 per group). Fc␥RIIa-transgenic mice treated with mAb M2139 were hyperresponsive to anti–type II collagen. Values are the mean and SD.

was observed in HAGG-stimulated macrophages (data not shown). Exaggerated antibody hypersensitivity of Fc␥RIIa-transgenic mice. The rapid CIA response in Fc␥RIIa-transgenic mice and increased sensitivity to HAGG implied an exaggerated response to pathologic antibodies. This possibility was tested using a passive antibody-transfer arthritis model described by Holmdahl and colleagues (20,28), wherein mice were normally given a single dose of a mixture of 2 anticollagen antibodies (4.5 mg each of M2139 and C1), followed by

SARDJONO ET AL

50 ␮g of lipopolysaccharide (LPS) 5 days later. These antibodies have been tested for arthritis induction in a number of mouse strains, including SJL, SJL ⫻ B6, BALB/c, and DBA/1 (28). None of these strains responded without LPS. However, in Fc␥RIIa-transgenic mice (Figure 3D), treatment with a single 2-mg dose of mAb M2139 alone, without LPS, caused a rapid onset of arthritis in 100% of mice, but had no effect in nontransgenic controls. Spontaneous development of autoimmunity in Fc␥RIIa-transgenic mice. Mice were housed for ⬎1 year, and we observed spontaneous progressive development of a systemic multiorgan autoimmune syndrome. Development of destructive arthritis. A proportion of aging Fc␥RIIa-transgenic mice spontaneously developed severe, destructive, symmetric arthritis. In a group of 25 Fc␥RIIa-transgenic mice monitored for ⬎1 year, none younger than 20 weeks of age developed arthritis, 7 (28%) developed arthritis between 20–45 weeks of

Figure 4. Analysis of arthritis in mice with spontaneous autoimmune disease. A, Cumulative percentage incidence in arthritis over time in human Fc␥ receptor IIa (Fc␥RIIa)–transgenic mice (n ⫽ 25). There was no difference in arthritis incidence between males and females. B and C, Photograph and radiographic image of the paw of a 36-weekold Fc␥RIIa-transgenic mouse with severe spontaneous arthritis (B) and a nonarthritic Fc␥RIIa-transgenic age-matched control mouse (C). Radiography was performed for each mouse category, i.e., nontransgenic, healthy Fc␥RIIa-transgenic, Fc␥RIIa-transgenic with mild arthritis, and Fc␥RIIa-transgenic with destructive arthritis (n ⫽ 4 per group) (results not shown). D–F, Representative sections showing the histopathologic features of hematoxylin and eosin–stained mouse joints. D, Early active destructive arthritis in an ankle from a 28-weekold mouse, with polymorphonuclear-dominant inflammatory cell infiltration. E, A later stage of destructive disease in a knee from a 36-week-old mouse, with predominantly mononuclear cell infiltration and advanced pannus. F, A normal ankle joint from an older (age 36 weeks) transgenic mouse. ca ⫽ cartilage; p ⫽ pannus; i ⫽ inflammatory cells. (Original magnification ⫻ 100.)


age, and 1 of the 25 developed arthritis thereafter, with a cumulative incidence of 32% (Figure 4A). Of the 8 mice with arthritis, 5 were severely affected (mean ⫾ SD index 10 ⫾ 2.38), with profoundly swollen joints and severe histologic and radiologic changes (Figure 4). The other 3 Fc␥RIIa-transgenic mice had less severe arthritis. Despite differences in arthritis severity, in both subsets of Fc␥RIIa-transgenic mice, the number of affected paws was similar. Age-matched nontransgenic mice never developed spontaneous disease. Radiologic examination of the affected limbs. The Fc␥RIIa-transgenic mice with severe arthritis developed marked radiologic changes that mirrored the florid edema and distortion of the foot seen clinically. Severe ankylosis of the tibiotarsal and tarsophalangeal cartilage was seen, with loss of joint space and prominent periarticular new bone formation (Figure 4B) compared with age-matched Fc␥RIIa-transgenic mice that did not develop arthritis (Figure 4C). Limbs from mice with mild arthritis revealed minimal radiologic changes (results not shown). Histologic evaluation of the joints. Joint histology of Fc␥RIIa-transgenic mice compared with that of nontransgenic, age-matched controls (Figures 4D–F) showed that mice with spontaneous severe arthritis had synovial hyperplasia and proliferation, cartilage erosion, pannus formation, and joint space infiltrate. The later stage of the disease showed more advanced destruction of bone and thinning of the cartilage, with the infiltrate changing from polymorphonuclear (PMN) cells in early active disease (Figure 4D) to macrophages in more advanced cases (Figure 4E). There was no evidence of disease in age-matched C57BL/6 or (SJL ⫻ C57BL/6)F1 controls (Figure 4F). Serum IgG2a elevation in affected mice. The levels of total IgG, IgG1, IgG2a, IgG2b, and IgG3 in all mice were determined by ELISA. Only the total level of IgG2a antibody was elevated (Figure 5A), and only in Fc␥RIIa-transgenic mice that developed destructive arthritis (mean ⫾ SD 6.6 ⫾ 2.5 ␮g/ml). IgG2a levels in mice with mild arthritis (2.0 ⫾ 0.8 ␮g/ml) were similar to those in Fc␥RIIa-transgenic mice with no disease (2.6 ⫾ 1.2 ␮g/ml) and in nontransgenic mice (1.7 ⫾ 1.3 ␮g/ml). Thus, there was a strong correlation between IgG2a levels and disease severity, particularly with the formation of pannus (r ⫽ 0.63). The levels of the other IgG subclasses were not altered (Figure 5). Other features of systemic autoimmune disease in Fc␥RIIa-transgenic mice. The Fc␥RIIa-transgenic mice were screened for further evidence of inflammation and autoimmunity by histologic examination of

3225

Figure 5. Serologic findings and incidence of kidney and lung disease in mice with spontaneous autoimmune disease. A, Levels of serum IgG2a in individual human Fc␥ receptor IIa (FcR␥IIa)–transgenic mice with destructive arthritis (n ⫽ 8) compared with unaffected FcR␥IIa-transgenic mice (n ⫽ 7), FcR␥IIa-transgenic mice with mild arthritis (n ⫽ 6), and normal age-matched nontransgenic mice (n ⫽ 6). Bars indicate the mean ⫾ SD concentration of IgG2a in each group. P ⫽ 0.0012 for IgG2a levels in mice with destructive disease versus unaffected nontransgenic mice. IgG2a levels correlated with disease severity, particularly with the formation of pannus (r ⫽ 0.63). Total IgG was increased in arthritic mice, and this correlated strongly with IgG2a levels (r ⫽ 0.96). The levels of the other Ig subclasses (IgM, IgG1, Ig2b, and Ig3) were not significantly different between the 4 groups of mice (P ⬎ 0.05 for all comparisons). B, Incidence of glomerulonephritis (GN) and pneumonitis (Pn) in Fc␥RIIa-transgenic mice at ages ⬍20 weeks (n ⫽ 8), 21–40 weeks (n ⫽ 14), and ⬎40 weeks (n ⫽ 9).

H&E-stained sections of skin, lymph nodes, gut, salivary glands, kidneys, eyes, brain, lungs, spleen, liver, pancreas, and heart at 14–60 weeks of age and compared with aged-matched, nontransgenic controls. Glomerulonephritis (GN) and pneumonitis were commonly observed in Fc␥RIIa-transgenic mice, with the disease incidence increasing with age. By 50 weeks of age, all mice were affected (Figure 5B). No abnormalities were found in other organs or in nontransgenic mice of any age. Glomerulonephritis. The time of onset and degree of severity of GN in individuals varied considerably. Few mice developed GN before 20 weeks of age, and up to 80% had the disease by 40 weeks. However, all mice ⬎40 weeks of age had moderate to severe GN, implying an age-related progression to severe GN (Figure 5B). Multifocal lymphoplasmacytic infiltrate in the renal interstitium, mainly around major arcuate vessels, was seen at 25–30 weeks (Figure 6A), with mild mesangial matrix deposition in the glomeruli and some tubular thickening. In mice ages ⬎40 weeks, more advanced disease was seen, with enlarged glomeruli, increased mesangial matrix (Figures 6B and C), and condensation of glomerular tufts. There were proliferative and sclerotic changes in Bowman’s capsule, indicative of crescent formation, and mild tubular proliferative changes (Figure 6B), although tubulointerstitial infiltrates were not present. The dis-

3226

Figure 6. Analysis of spontaneous autoimmune glomerulonephritis, pneumonitis, and antihistone antibodies in a group of older human Fc␥ receptor IIa (Fc␥RIIa)–transgenic mice (n ⫽ 31). A, Kidney section (hematoxylin and eosin [H&E] stained) taken at age 25 weeks, showing multifocal lymphoplasmacytic infiltration in the renal interstitium, mainly around major arcuate vessels. The glomerulus (arrow) shows mild mesangial matrix deposition. Tubules were thickened but otherwise normal (original magnification ⫻ 200). B, An H&E-stained kidney section taken at 45 weeks, showing enlarged glomeruli (thin arrows), increased mesangial matrix, and condensation of glomerular tufts. There are proliferative and sclerotic changes in Bowman’s capsule (thick arrow) indicative of crescent formation, and mild tubular proliferative changes (original magnification ⫻ 100). C, Fluorescence staining of immune complexes in the glomerulus. Kidney sections were stained with anti-mouse IgG, directly conjugated with fluorescein isothiocyanate, and immune complexes within the glomerulus appeared granular (original magnification ⫻ 200). Diffuse, lowlevel staining with no glomerular concentration of IgG was seen in age-matched, nontransgenic control mice (results not shown). D, Transmission electron microscopy of glomeruli, showing immune complexes deposited (D) on the glomerular basement membrane (BM, arrow) above the uriniferous space (U) and below the endothelial layer (En) (original magnification ⫻ 50). E, H&E-stained Fc␥RIIatransgenic mouse lung section taken at 40 weeks, showing infiltration of inflammatory cells (arrows) and local destruction of lung architecture (original magnification ⫻ 50). F, Enzyme-linked immunosorbent assay for the presence of antihistone antibodies in Fc␥RIIa-transgenic and nontransgenic mouse sera at 36 weeks of age. Antihistone antibodies were present in serum from many of the Fc␥RIIatransgenic mice at all ages, but were not present in serum from the nontransgenic mice. Horizontal line shows the cutoff for positivity.

ease was self-limiting, since mice ages ⬎40 weeks, with up to 65% of glomeruli affected, remained healthy, with normal urinary protein and serum creatinine levels (results not shown). Age-matched nontransgenic mice showed no evidence of disease (results not shown). FITC-conjugated anti-mouse IgG staining of kidney sections showed an accumulation of immune complexes within the glomeruli, which produced a dense, granular appearance (Figure 6C). In contrast, kidneys

SARDJONO ET AL

from age-matched nontransgenic control mice showed diffuse, low-level staining of the tubules and mesangium, with no concentration of immunoglobulin in the glomeruli (results not shown). Transmission electron microscopy also revealed immune complex deposition (Figure 6D), with features similar to those of lupus nephritis in humans, including small, electron-dense deposits forming wire-loop lesions in the subendothelial basement membrane. Lung histopathology. Pneumonitis was found in 25% of mice between ages 12 and 40 weeks, increasing to 100% of older mice (Figure 5B), and was characterized by patches of perivascular inflammation with cellular aggregates of macrophages, lymphocytes, plasma cells, and numerous PMN cells within alveolar walls (Figure 6E). In severe cases, up to 50% of the normal architecture of the lungs was obliterated, but the disease was self-limiting, and older mice remained healthy. Lungs of age-matched nontransgenic mice showed no inflammation. Analysis of antihistone and antinuclear antibodies (ANAs). The histologic features suggested that the tissue damage seen in the spontaneous autoimmune disease in Fc␥RIIa-transgenic mice was mediated, at least in part, by autoantibodies. Therefore, we evaluated mice for the presence of autoantibodies known to be associated with human autoimmune disease. Initial immunofluorescence studies with sera from most Fc␥RIIa-transgenic mice and many older nontransgenic C57BL/6 mice showed homogeneous nuclear staining of Chinese hamster ovary cells (results not shown). This staining pattern was similar to that of antihistone antibody huPIA3 (29). Sera were then tested for antihistone antibodies by ELISA, using a mixture of purified histones. In mice ages ⬎20 weeks, 13 of 23 had antihistone antibodies above background levels (Figure 6F). Histone antibody titers did not correlate with GN or arthritis incidence (r ⬍0.05). No antihistone antibodies above background levels were seen in age-matched C57BL/6 mice or (SJL ⫻ C57BL/6)F1 mice at ages ⬎25 weeks. None of the mice had the other common autoantibodies, such as anti–double-stranded or anti–single-stranded DNA or rheumatoid factor (data not shown). DISCUSSION Since it was found that recombinant soluble FcR inhibited immune complex vasculitis (3), there has been widespread interest in the role of FcR in antibodyinduced inflammation in autoimmune diseases (5,8). Although many studies have analyzed the role of mouse


FcR, including Fc␥RI, Fc␥RIIb, and Fc␥RIIIa, which are common to both mice and humans, humans have a unique FcR, Fc␥RIIa, which is absent from rodents and was therefore not analyzed in rodent models of autoimmunity. Our data demonstrate that Fc␥RIIa-transgenic mice are highly susceptible to passive antibody-induced inflammation and to active (collagen-induced) and spontaneous autoimmune disease. In CIA, there is variable susceptibility in mice, which is linked to MHC type (24). In this study, expression of Fc␥RIIa conferred susceptibility to CIA in strains of mice with low-susceptibility MHC (H-2b and H-2b/s). Indeed, CIA in Fc␥RIIa-transgenic mice developed more rapidly than in the archetypal CIAsusceptible DBA/1 strain, with almost 15% of Fc␥RIIatransgenic mice developing CIA after a single dose of collagen. These findings establish a role for Fc␥RIIa in enhancing inflammatory responses in CIA, especially since treatment with anti-Fc␥RIIa mAb reduced disease severity. The immune mechanisms involved in the development of CIA have been well described, with both T cells and anticollagen antibodies known to play major roles (30,31). The observation that anti-CII antibodies were lower on day 21 after CIA induction in Fc␥RIIatransgenic mice compared with DBA/1 or nontransgenic animals is evidence that Fc␥RIIa plays a role in increasing the sensitivity of effector cells to activation by immune complexes. This possibility was supported by data from the anti-CII antibody transfer model, wherein a single dose of anti-CII antibody M2139 induced disease in 100% of the Fc␥RIIa-transgenic mice. This contrasts with other strains, in which a mixture of 2 antibodies plus LPS was required to induce this level of disease (20). Although the use of 1 antibody to induce arthritis in susceptible DBA/1 mice has been reported (32), the highest incidence was only 50% after 2 doses of M2139 (total dose 9 mg), compared with the 100% incidence after a single 2-mg dose in the Fc␥RIIatransgenic mice. TNF␣ is a major clinically validated inflammatory mediator in human rheumatoid arthritis (RA) (33), and the majority of immune complex–induced TNF␣ production from transgenic macrophages could be attributed to Fc␥RIIa activation. Endogenous mouse FcR (Fc␥RIIIa and Fc␥RI) were responsible for the balance of TNF␣, which was equivalent to that produced from HAGG-stimulated nontransgenic macrophages. This also indicates that the endogenous receptors are functionally intact and unaffected by the presence of the transgenic receptor.

3227

A surprising characteristic of the Fc␥RIIatransgenic mice was the spontaneous development of disease with features of RA and systemic lupus erythematosus (SLE). Approximately 40% of older transgenic mice had features of both RA and SLE, similar to the “rhupus” overlap syndrome described in humans (34). The remaining 60% of the mice had features of SLE, e.g., endoproliferative GN, with intraglomerular accumulation of deposits (i.e., wire-loop lesions). Some Fc␥RIIa-transgenic mice also developed mild, nonerosive inflammatory arthritis similar to that seen in SLE. While most of the older Fc␥RIIa-transgenic mice had antihistone antibodies, the most interesting serologic observation was elevated IgG2a levels in mice with spontaneous severe destructive arthritis. Elevated IgG2a was seen in other autoimmune models in mice (35,36), and may reflect the cytokine profiles involved in leukocyte activation (37). The antibodies detected in our study were not directed against CII and, currently, the autoantigen remains undefined. The phenotype of the Fc␥RIIa-transgenic mice resembles that of mice deficient in the inhibitory Fc␥RIIb. These mice show increased susceptibility to CIA (36), elevated levels of TNF␣ and IgG2a (38), and, on a specific MHC background, spontaneously develop SLE-like symptoms, including ANAs to double-stranded DNA and DNA/histone complexes (39) and GN (40). However, there are some fundamental differences compared with Fc␥RIIa-transgenic mice. The Fc␥RIIbdeficient mice never developed spontaneous severe destructive arthritis; they showed exaggerated antibody responses, with elevated anti-CII antibodies in the CIA model, whereas Fc␥RIIa-transgenic mice had low antibody levels. Also, Fc␥RIIb-knockout mice have antiDNA antibodies, but Fc␥RIIa-transgenic mice developed only antihistone antibodies. Thus, analysis of Fc␥RIIb-deficient mice by other investigators suggests that their phenotype was due largely to dysregulation of B cell activation and loss of B cell tolerance (36,40) arising from unbalanced ITAM/ITIM signaling (38). While we cannot rule out such an imbalance, it must be relatively subtle, since Fc␥RIIa-transgenic mice had both activating and inhibitory receptors (Figure 1). In contrast, Fc␥RIIb-deficient mice and cells entirely lacked the inhibitory Fc␥RIIb, but retained a full complement of activating receptors. Furthermore, data from other studies (41) show that transfection of activating FcR into cells that already express both activating and inhibitory receptors left the inhibitory Fc␥RIIb still functional and potent. An alternative explanation for our observations is that Fc␥RIIa lowers the

3228

SARDJONO ET AL

threshold of immune complex activation or qualitatively changes the response induced by pathogenic antibodies. Indeed, evidence from human in vitro studies suggests such a role for Fc␥RIIa in activated macrophages, where Fc␥RI signals are partly dependent on Fc␥RIIa (42). Nonetheless, future studies of mechanisms will be informative in this regard. The development of destructive arthritis involves interlinked immunologic and cytokine pathways (43,44). It is clear that TNF␣ and IL-1␤ are major factors in active human disease (44), and recent clinical trials have suggested a role for B cells and possibly for immune complexes (45,46). Although animal models have been useful in suggesting possible mechanisms in human disease, analysis of the role of the major and unique activating human FcR (Fc␥RIIa) has been lacking. Our analysis strongly suggests a role for this receptor in RA. It is particularly interesting that Fc␥RIIa-transgenic mice spontaneously develop destructive arthritis. This is rare in mice, having been observed only in older males of the susceptible DBA/1 strain (47), in strains with alterations in T cell tolerance, selection, and/or activation (for example, K/BxN [6] and SKG [48] strains), or in mice with engineered cytokine disturbances, such as TNF␣-transgenic (26) and IL-1Ra–knockout (49) mice. These studies show that expression of the unique activating human FcR, Fc␥RIIa, is associated with spontaneous autoimmune inflammation and exaggerated reactivity to induced, antibody-dependent inflammation. Fc␥RIIa may lower the threshold for stimulation of effector cell function, resulting in hypersensitivity to immune complexes in vivo, and inducing enhanced cytokine secretion. Thus, Fc␥RIIa plays a critical role in determining the sensitivity to autoimmunity and, as such, provides an attractive target for the treatment of autoimmune disease. ACKNOWLEDGMENTS We thank Liliana Tatarczuch for assistance with the transmission electron microscopy, Bill Pispalliaris for useful discussion of the antihistone antibody analysis, Dacho Gao for the production of the radiographs, and Dr. Peck Sze Tan (ARI) for antibodies.

REFERENCES 1. Dijstelbloem HM, Bijl M, Fijnheer R, Scheepers RH, Oost WW, Jansson MD, et al. Fc␥ receptor polymorphisms in systemic lupus erythematosus: association with disease and in vivo clearance of immune complexes. Arthritis Rheum 2000;43:2793–800. 2. Takai T. Roles of Fc receptors in autoimmunity. Nat Rev Immunol 2002;2:580–92.

3. Ierino FL, Powell MS, McKenzie IF, Hogarth PM. Recombinant soluble human Fc␥RII: production, characterization, and inhibition of the Arthus reaction. J Exp Med 1993;178:1617–28. 4. Ravetch JV, Bolland S. IgG Fc receptors [review]. Annu Rev Immunol 2001;19:275–90. 5. De Stahl TD, Andren M, Martinsson P, Verbeek JS, Kleinau S. Expression of Fc␥RIII is required for development of collageninduced arthritis. Eur J Immunol 2002;32:2915–22. 6. Ji H, Ohmura K, Mahmood U, Lee DM, Hofhuis FM, Boackle SA, et al. Arthritis critically dependent on innate immune system players. Immunity 2002;16:157–68. 7. Van Lent PL, Nabbe K, Blom AB, Holthusyen AE, Sloetjes A, van de Putte LB, et al. Role of activatory Fc␥RI and Fc␥ III and inhibitory Fc␥RII in inflammation and cartilage destruction during experimental antigen-induced arthritis. Am J Pathol 2001;159: 2309–20. 8. Hogarth PM. Fc receptors are major mediators of antibody based inflammation in autoimmunity [review]. Curr Opin Immunol 2002;14:798–202. 9. Powell MS, Barton PA, Emmanouilidis D, Wines BD, Neumann GM, Peitersz GA, et al. Biochemical analysis and crystallisation of Fc␥RIIa, the low affinity receptor for IgG. Immunol Lett 1999;68: 17–23. 10. Chuang FY, Sassaroli M, Unkeless JC. Convergence of Fc␥ receptor IIA and Fc␥ receptor IIB signalling pathways in human neutrophils. J Immunol 2000;164:350–60. 11. Tan Sardjono C, Mottram PL, Hogarth PM. The role of Fc␥RIIa as an inflammatory mediator in rheumatoid arthritis and systemic lupus erythematosus [review]. Immunol Cell Biol 2003;81:374–81. 12. Maxwell KF, Powell MS, Hulett MD, Barton PA, McKenzie IF, Garrett TP, et al. Crystal structure of the human leukocyte Fc receptor, Fc␥RIIa. Nat Struct Biol 1999;6:437–42. 13. Powell MS, Hogarth PM. The role and use of recombinant receptors in the investigation and control of antibody-induced inflammation. In: van de Winkel JG, Hogarth PM, editors. Immunoglobulin receptors and their physiological and pathological roles in immunity. Dordrecht (The Netherlands): Kluwer Academic Publishers; 1998. 14. Hulett MD, Hogarth PM. Molecular basis of Fc receptor function [review]. Adv Immunol 1994;57:1–127. 15. Daeron M. Fc receptor biology [review]. Annu Rev Immunol 1997;15:203–34. 16. McKenzie SE. Humanized mouse models of FcR clearance in immune platelet disorders [review]. Blood Rev 2002;16:3–5. 17. McKenzie SE, Taylor SM, Malladi P, Yuhan H, Cassel DL, Chien P, et al. The role of the human Fc receptor Fc␥RIIA in the immune clearance of platelets: a transgenic mouse model. J Immunol 1999;162:4311–8. 18. Manger K, Repp R, Spriewald BM, Rascu A, Geiger A, Wassmuth R, et al. Fc␥ receptor IIa polymorphism in Caucasian patients with systemic lupus erythematosus: association with clinical symptoms. Arthritis Rheum 1998;41:1181–9. 19. Campbell IK, Hamilton JA, Wicks IP. Collagen-induced arthritis in C57BL/6 (H-2b) mice: new insights into an important disease model of rheumatoid arthritis. Eur J Immunol 2000;30:1568–75. 20. Svensson L, Nandakumar KS, Johansson A, Jansson L, Holmdahl R. IL-4-deficient mice develop less acute but more chronic relapsing collagen-induced arthritis. Eur J Immunol 2002;32:2944–53. 21. Barnes N, Gavin AL, Tan PS, Mottram P, Koentgen F, Hogarth PM. Fc␥RI-deficient mice show multiple alterations to inflammatory and immune responses. Immunity 2002;16:379–89. 22. Tan PS, Gavin AL, Barnes N, Sears DW, Vremec D, Shortman K, et al. Unique monoclonal antibodies define expression of Fc␥RI on macrophages and mast cell lines and demonstrate heterogeneity among subcutaneous and other dendritic cells. J Immunol 2003;170:2549–56. 23. Wilkins S, Gureyev T, Gao D, Pogany A, Stevenson A. Phase-


24. 25.

26.

27. 28.

29. 30. 31.

32.

33. 34. 35.

36.

contrast imaging using polychromatic hard X-rays. Nature 1996; 384:335–8. Courtenay JS, Dallman MJ, Dayan AD, Martin A, Mosedale B. Immunisation against heterologous type II collagen induces arthritis in mice. Nature 1980;283:666–8. Matsumoto T, Ametani A, Hachimura S, Iwaya A, Taguchi Y, Fujita K, et al. Intranasal administration of denatured type II collagen and its fragments can delay the onset of collagen-induced arthritis. Clin Immunol Immunopathol 1998;88:70–9. Butler DM, Malfait AM, Mason LJ, Warden PJ, Kollias G, Maini RN, et al. DBA/1 mice expressing the human TNF-␣ transgene develop a severe, erosive arthritis: characterization of the cytokine cascade and cellular composition. J Immunol 1997;159:2867–76. Feldmann M. Development of anti-TNF therapy for rheumatoid arthritis. Nat Rev Immunol 2002;2:364–71. Nandakumar KS, Svensson L, Holmdahl R. Collagen type IIspecific monoclonal antibody-induced arthritis in mice: description of the disease and the influence of age, sex, and genes. Am J Pathol 2003;163:1827–37. Cosgrove L. Monoclonal antibodies to platelet antigens. Melbourne (Australia): University of Melbourne; 1987. Myers LK, Rosloniec EF, Cremer MA, Kang AH. Collageninduced arthritis, an animal model of autoimmunity [review]. Life Sci 1997;61:1861–78. Andersson EC, Hansen BE, Jacobsen H, Madsen LS, Andersen CB, Engberg J, et al. Definition of MHC and T cell receptor contacts in the HLA-DR4 restricted immunodominant epitope in type II collagen and characterization of collagen-induced arthritis in HLA-DR4 and human CD4 transgenic mice. Proc Natl Acad Sci U S A 1998;95:7574–9. Nandakumar KS, Andren M, Martinsson P, Bajtner E, Hellstrom S, Holmdahl R, et al. Induction of arthritis by single monoclonal IgG anti-collagen type II antibodies and enhancement of arthritis in mice lacking inhibitory Fc␥RIIB. Eur J Immunol 2003;33: 2269–77. Feldmann M, Maini RN. Anti-TNF␣ therapy of rheumatoid arthritis: what have we learned? [review]. Annu Rev Immunol 2001;19:163–96. Hahn B. Systemic lupus erythematosus and related syndromes. In: Ruddy S, Harris E, Sledge B, editors. Kelley’s textbook of rheumatology. 6th ed. Philadelphia: W. B. Saunders; 2001. p. 1089–103. Quattrocchi E, Dallman MJ, Feldmann M. Adenovirus-mediated gene transfer of CTLA-4Ig fusion protein in the suppression of experimental autoimmune arthritis. Arthritis Rheum 2000;43: 1688–97. Yuasa T, Kubo S, Yoshino T, Ujike A, Matsumura K, Ono M, et

3229

37.

38.

39.

40.

41. 42.

43. 44.

45.

46.

47. 48.

49.

al. Deletion of Fc␥ receptor IIB renders H-2b mice susceptible to collagen-induced arthritis. J Exp Med 1999;189:187–94. Watson WC, Townes AS. Genetic susceptibility to murine collagen II autoimmune arthritis: proposed relationship to the IgG2 autoantibody subclass response, complement C5, major histocompatibility complex (MHC) and non-MHC loci. J Exp Med 1985; 162:1878–91. Clynes R, Maizes JS, Guinamard R, Ono M, Takai T, Ravetch JV. Modulation of immune complex-induced inflammation in vivo by the coordinate expression of activation and inhibitory Fc receptors. J Exp Med 1999;189:179–85. Bolland S, Yim YS, Tus K, Wakeland EK, Ravetch JV. Genetic modifiers of systemic lupus erythematosus in Fc␥RIIB⫺/⫺ mice. J Exp Med 2002;195:1167–74. Bolland S, Ravetch JV. Spontaneous autoimmune disease in Fc␥RIIB-deficient mice results from strain-specific epistasis. Immunity 2000;13:277–85. Daeron M. Structural bases of Fc␥R functions [review]. Int Rev Immunol 1997;16:1–27. Allen JM, Seed B. Isolation and expression of functional highaffinity Fc receptor complementary DNAs. Science 1989;243: 378–81. Firestein GS. Etiology and pathogenesis of rheumatoid arthritis. 6th ed. Philadelphia: W. B. Saunders; 2001. Feldmann M, Maini RN, Bondeson J, Taylor P, Foxwell BM, Brennan FM. Cytokine blockade in rheumatoid arthritis [review]. Adv Exp Med Biol 2001;490:119–27. Moore J, Ma D, Will R, Cannell P, Handel M, Milliken S. A phase II study of rituximab in rheumatoid arthritis patients with recurrent disease following haematopoietic stem cell transplantation. Bone Marrow Transplant 2004;34:241–7. Edwards JC, Szczepanski L, Szechinski J, Filipowicz-Sosnowska A, Emery P, Close DR, et al. Efficacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med 2004;350:2572–81. Nordling C, Karlsson-Parra A, Jansson L, Holmdahl R, Klareskog L. Characterization of a spontaneously occurring arthritis in male DBA/1 mice. Arthritis Rheum 1992;35:717–22. Hata H, Sakaguchi N, Yoshitomi H, Iwakura Y, Sekikawa K, Azuma Y, et al. Distinct contribution of IL-6, TNF-␣, IL-1, and IL-10 to T cell-mediated spontaneous autoimmune arthritis in mice. J Clin Invest 2004;114:582–8. Iwakura Y. Roles of IL-1 in the development of rheumatoid arthritis: consideration from mouse models [review]. Cytokine Growth Factor Rev 2002;13:341–55.