Ms Willis, Ms Alcorn, and Ms Nelson own stock or stock options in Amgen Inc.
Blockade of the interleukin-7 receptor inhibits collagen-induced arthritis and is associated with reduction of T cell activity and proinflammatory mediators
Article first published online: 24 MAY 2010
Copyright © 2010 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 62, Issue 9, pages 2716–2725, September 2010
How to Cite
Hartgring, S. A. Y., Willis, C. R., Alcorn, D., Nelson, L. J., Bijlsma, J. W. J., Lafeber, F. P. J. G. and van Roon, J. A. G. (2010), Blockade of the interleukin-7 receptor inhibits collagen-induced arthritis and is associated with reduction of T cell activity and proinflammatory mediators. Arthritis & Rheumatism, 62: 2716–2725. doi: 10.1002/art.27578
- Issue published online: 31 AUG 2010
- Article first published online: 24 MAY 2010
- Manuscript Accepted: 18 MAY 2010
- Manuscript Received: 6 JUL 2009
To study the effects of interleukin-7 receptor α-chain (IL-7Rα) blockade on collagen-induced arthritis (CIA) and to investigate the effects on T cell numbers, T cell activity, and levels of proinflammatory mediators.
We studied the effect of anti–IL-7Rα antibody treatment on inflammation and joint destruction in CIA in mice. Numbers of thymocytes, splenocytes, T cell subsets, B cells, macrophages, and dendritic cells were assessed. Cytokines indicative of Th1, Th2, and Th17 activity and several proinflammatory mediators were assessed by multianalyte profiling in paw lysates. In addition, T cell–associated cytokines were measured in supernatants of lymph node cell cultures.
Anti–IL-7Rα treatment significantly reduced clinical arthritis severity in association with reduced radiographic joint damage. Both thymic and splenic cellularity were reduced after treatment with anti–IL-7Rα. IL-7Rα blockade specifically reduced the total number of cells as well as numbers of naive, memory, CD4+, and CD8+ T cells from the spleen and significantly reduced T cell–associated cytokines (interferon-γ, IL-5, and IL-17). IL-7Rα blockade also decreased local levels of proinflammatory cytokines and factors associated with tissue destruction, including tumor necrosis factor α, IL-1β, IL-6, matrix metalloproteinase 9, and RANKL. IL-7Rα blockade did not significantly affect B cells, macrophages, and dendritic cells. B cell activity, indicated by serum anticollagen IgG antibodies, was not significantly altered.
Blockade of IL-7Rα potently inhibited joint inflammation and destruction in association with specific reductions of T cell numbers, T cell–associated cytokines, and numerous mediators that induce inflammation and tissue destruction. This study demonstrates an important role of IL-7R–driven immunity in experimental arthritis and indicates the therapeutic potential of IL-7Rα blockade in human arthritic conditions.
Interleukin-7 (IL-7), a member of the IL-2 family of cytokines, is a potent pleiotropic immunostimulatory cytokine produced by stromal cells at lymphopoietic sites in the bone marrow, gut, spleen, and thymus, and has a pivotal role in T cell development (1, 2). IL-7 potently promotes T cell development in mice and humans. Although B cell development in mice is strongly dependent on IL-7, in humans this seems not to be the case since reduced T cell numbers, but no reduction in B cell numbers, are observed in IL-7 receptor (IL-7R)–deficient humans (3). Reduced B cell activity (Ig production) in these IL-7R–deficient individuals is therefore suggested to be T cell driven (4). In accordance with this, IL-7 has been shown to induce T cell–dependent activation of other cell types such as monocytes and osteoclasts (1, 2, 5–7). In the absence of inflammation, IL-7 in ovariectomized mice induces T cell–mediated and RANKL- and tumor necrosis factor α (TNFα)–dependent generalized bone loss (8). In addition, overexpression of IL-7 in mice induces increased bone resorption leading to osteopenia (9), and administering IL-7 to healthy mice induces bone loss (8, 10).
High levels of IL-7 are found in several arthritic conditions including rheumatoid arthritis (RA) (6, 7, 11–13). Serum concentrations of IL-7 are increased in individuals with arthritis compared with healthy controls (6, 11, 12) and correlate with markers of disease activity in RA patients (12). Increased concentrations of IL-7 are present in synovial fluid from RA patients compared with osteoarthritis (OA) patients. In RA synovial tissue, IL-7 is abundantly expressed by macrophages, endothelial cells, and fibroblasts, and the expression correlates with numbers of CD68+ macrophages (7). TNFα, important in the induction of tissue destruction and inflammatory processes, correlates with IL-7 levels in RA (6). In addition, we have reported that IL-7 levels persist in RA patients whose disease is refractory to anti-TNFα treatment (6). These results point toward a significant role for IL-7 in immunopathology of RA, a role that might be independent of TNFα in a specific group of RA patients.
The immunostimulatory capacities of IL-7 also suggest a significant role in the immunopathology of RA. In vitro, IL-7 induces proinflammatory activities of several cells of the immune system, but primarily activates T cells. IL-7 stimulates primarily Th1 and Th17 cytokine secretion upon stimulation of mononuclear cells from RA peripheral blood and synovial fluid (12, 14). It increases production of TNFα and interferon-γ (IFNγ) by isolated peripheral blood T cells from RA patients (15). In addition, IL-7 stimulates T cell–dependent expression of costimulatory molecules on monocyte/macrophages, resulting in contact-dependent activation of T cells (6, 7). T cell–dependent activation of monocyte/macrophages by IL-7 is also associated with TNFα production that is mainly monocyte derived (6, 7). In addition to TNFα production, IL-7 stimulates the production of a number of other proinflammatory cytokines by monocytes (IL-1α, IL-1β, IL-6, IL-8, macrophage inflammatory protein 1β [MIP-1β]) (16–18). Finally, IL-7 blockade prevents gp130-dependent autoimmune arthritis in mice (19). Taken together, these results indicate the importance of IL-7 in promoting inflammation and tissue destruction in several inflammatory diseases including RA.
IL-7 effects are mediated through the high-affinity IL-7R α-chain (IL-7Rα) in conjunction with the common γ-chain. Recently, we demonstrated increased intraarticular IL-7R expression in synovium of RA patients and showed that this receptor is present on highly proliferating synovial T cells but not on regulatory FoxP3+ T cells (20). Furthermore, soluble human IL-7R inhibited IL-7–induced Th1 activity of cultured mononuclear cells from RA patients (20).
Because IL-7 activity is critically dependent on signaling through IL-7Rα, in the present study we investigated the effects of prophylactic and therapeutic IL-7R blockade in the collagen-induced arthritis (CIA) model of RA. Additionally, we evaluated the mechanisms by which IL-7R blockade regulates experimental arthritis.
MATERIALS AND METHODS
Induction, treatment, and assessment of CIA.
Chicken type II collagen (CII) (no. C9301; Sigma) was dissolved at a concentration of 8 mg/ml in 0.1N acetic acid and emulsified in an equal volume of Freund's complete adjuvant (CFA) (Freund's incomplete adjuvant [IFA] plus 5 mg/ml heat-killed Mycobacterium tuberculosis). Eight-week old male DBA/1 mice (Harlan) were immunized intradermally at the base of the tail with 50 μl emulsion (200 μg of CII). On day 21 mice were given a booster injection intradermally with 200 μg of CII dissolved in IFA. Mice were examined for onset and severity of disease, in a blinded manner. Arthritis symptoms were graded using the following scoring system: grade 0 = normal appearance; grade 1 = slight erythema/edema (1–3 digits); grade 2 = erythema/edema in >3 digits, or mild swelling in ankle/wrist joint; grade 3 = erythema/edema in entire paw; grade 4 = massive erythema/edema of entire paw extending into proximal joints, ankylosis, loss of function. Each limb was graded, giving a maximum possible score of 16 per mouse.
In the prophylactic treatment experiments, rat anti-mouse IL-7Rα monoclonal antibody (mAb) (M595, rat IgG2b; Amgen) was administered starting at the time of collagen boost (experimental day 21) and on days 24, 27, and 30. Intraperitoneal (IP) injections of anti–IL-7Rα were given at a dosage of either 100 μg or 500 μg.
In addition, mice with established arthritis were treated with anti–IL-7Rα mAb. When immunized mice started to show clinical symptoms, arthritic mice (mean arthritis score of 3 on day 27) were injected IP with 100 μg anti–IL-7Rα mAb on days 27, 30, 34, 37, and 41.
As a control in both experiments, mice were given IP injections with phosphate buffered saline (PBS) or isotype control (nonbinding rat anti-mouse IgG2b antibody; Amgen) at the same time points at which anti–IL-7Rα was given. Mice in the prophylactic treatment experiment were killed on day 33, and mice in the therapeutic experiment were killed on day 42.
Assessment of radiographic joint damage in CIA.
After autopsy, ankles were fixed in formalin for 24 hours. Subsequently, lateromedial radiographs of both ankles were obtained. All individual ankles were scored for incidence of radiographic lesions and severity according to the following index: grade 0 (normal) = well-defined joint spaces, smooth periosteum, homogeneous bone density, no erosions or lytic lesions; grade 1 (mild) = roughened periosteum, a few poorly defined joint spaces, homogeneous bone density, no erosions or lytic lesions; grade 2 (intermediate) = loss of clearly defined joint spaces, periosteal reaction (roughened periosteal surface), variability in bone density, minimal osteophyte formation; grade 3 (severe) = loss of defined joint spaces, bone/joint erosions, more prominent osteophyte formation, more severe periosteal reaction, lytic lesions. Both ankles were graded, and mean values per ankle were determined for each treatment group.
Spleen and thymus cell preparation and flow cytometry.
On experimental day 33 of the prophylactic treatment experiment, spleen and thymus were collected and weighed. Spleens and thymuses were pressed through a 70-μm cell strainer (Falcon; BD Biosciences) into Hanks' balanced salt solution (HBSS; Invitrogen) with 5% fetal bovine serum (FBS; Invitrogen) and prepared as single-cell suspensions. Red blood cells in the spleen were lysed by 5 minutes of incubation in 5 ml lysis buffer (Invitrogen). The reaction was stopped by the addition of 10 ml cold PBS. Cells were counted and thereafter pelleted and resuspended in HBSS/5% FBS/1 μg/ml 2.4G2 (CD32/16, Fc block; BD Biosciences) prior to flow cytometric staining. For identification of different cell types by flow cytometry, the following mAb were used: Alexa 647–labeled anti-CD3 (clone UCHT1; BD Biosciences), peridinin chlorophyll protein (PerCP)–Cy5.5–labeled anti-CD4 (clone RM4-5; BD Biosciences), allophycocyanin (APC)–labeled anti-CD8 (clone 53-6.7; BD Biosciences), PerCP-Cy5.5–labeled anti-CD11b (clone M1/70; BD Biosciences), anti-CD11c (clone HL3; BD Biosciences), fluorescein isothiocyanate (FITC)–labeled anti-CD44 (clone IM7; BD Biosciences), and FITC-labeled anti–class II major histocompatibility complex (MHC). For control staining, APC- or FITC- or phycoerythrin- or PerCP-Cy5.5–labeled rat IgG2a, rat IgG2b, or mouse IgG2a isotype antibodies were used. A total of 105 events were collected, gated on viable lymphocytes based on forward and side scatter pattern, and specific staining was performed using a FACSCalibur instrument (BD Biosciences) and FlowJo software (Tree Star).
Detection of anti-CII antibodies.
Serum concentrations of anti–chicken CII antibodies were measured by enzyme-linked immunosorbent assay (ELISA) using a mouse IgG1, IgG2a, or IgG2b anti–type II collagen antibody assay kit (Chondrex) and detected according to the manufacturer's instructions.
Investigation of paw lysates for cytokine and chemokine content.
Front paws of each mouse were collected and frozen in liquid nitrogen directly upon removal from the animal. Using a tissue lyser machine (Qiagen), paws were shredded together into homogeneous lysates in digestion buffer (50 mM Tris HCl buffer, pH 7.4, containing 0.1M NaCl, 0.1% Triton X-100, and protease inhibitors [Roche]) with the use of 5-mm stainless steel beads. The lysates were run over a Qiashredder column (Qiagen). The protein content of the resulting lysates was normalized by assaying the lysate with a BCA total protein quantitation kit (Pierce). Concentrations of specific proteins were assayed using rodent multianalyte profiling (Rodent MAP version 1.1; Rules-Based Medicine), and concentrations of RANKL were measured using a mouse bone panel 2B Lincoplex kit (Millipore). Serum concentrations of IL-6 were measured by ELISA (quantikine kit; R&D Systems) according to the manufacturer's instructions.
In vitro restimulation of lymph node cells.
Draining (axial, brachial, inguinal) lymph nodes were collected at the end of the prophylactic treatment experiment and pooled for each treatment group. Cells were crushed through a 70-μm cell strainer and prepared as a single-cell suspension in Dulbecco's modified Eagle's medium with 10% FBS. Cells were cultured for 48 hours in triplicate wells in the absence or presence of CII (100, 50, 25, or 12.5 μg/ml). Supernatants were harvested and tested for IL-17 (and IL-2) concentrations by ELISA according to the instructions of the manufacturer (R&D Systems).
Statistical analysis of arthritis severity and radiographic damage was done using an independent sample t-test. Total cell numbers, subsets of thymocytes and splenocytes, and paw lysate concentrations of IFNγ, IL-5, IL-17, and other proinflammatory mediators in anti–IL-7Rα–treated mice compared with those in isotype control–treated mice were analyzed by one-way analysis of variance with Dunnett's post-test.
IL-7R blockade inhibits development of CIA and CIA-induced joint destruction.
To examine the role of IL-7R in CIA development, mice were treated with an anti-mouse IL-7Rα mAb. Treatment with an isotype control antibody did not, at any time point, significantly alter arthritis severity as compared with the PBS-treated group (Figure 1A). Treatment with 500 μg anti–IL-7Rα significantly reduced arthritis compared with the isotype control– and PBS-treated groups (P < 0.01 and P < 0.05, respectively, for area under the curve), resulting in mean ± SEM arthritis inhibition of 47 ± 3% and 48 ± 6%, respectively (mean ± SEM inhibition from days 27–33) (data not shown). The lower dose of anti–IL-7Rα mAb (100 μg) was more effective than the 500 μg dose in reducing arthritis severity compared with the control groups (both P ≤ 0.005 for area under the curve). Mean ± SEM arthritis inhibition from the 100 μg dose was 69 ± 6.8% (mean ± SEM inhibition from days 27–33) compared with isotype control–treated mice and 66 ± 10.4% compared with PBS-treated mice (Figure 1A).
The incidence of arthritis did not differ significantly among the groups. Thirteen of 15 PBS-treated mice (86.7%) and 100% of isotype control–treated mice developed clinical arthritis, compared with 86.7% and 80.0% of mice treated with anti–IL-7Rα at 100 μg (Figure 1B) and 500 μg (data not shown), respectively.
On day 33, for further ex vivo examination, radiographs were obtained of each ankle joint for grading of radiographic joint damage (Figure 1C). Both control groups showed joint space narrowing, variability in bone density, and roughened periosteal surface, and some mice showed osteophyte formation in the ankle joint (mean ± SEM score 1.37 ± 0.16 in the PBS-treated group and 1.03 ± 0.15 in the isotype control–treated group). Mice treated with 100 μg anti–IL-7Rα showed significantly less radiographic joint damage (mean ± SEM score 0.37 ± 0.14) (P < 0.001 versus either control group) (Figure 1C). Also, prophylactic treatment with the higher dose (500 μg) significantly reduced radiographic joint damage (mean ± SEM score 0.6 ± 0.16) (P < 0.05 versus either control group) (data not shown).
Anti–IL-7R treatment decreases thymic and splenic cell numbers.
Thymic cell numbers, expressed as cells per tissue weight, were decreased in anti–IL-7Rα–treated mice (29% reduction versus isotype control–treated mice). Absolute splenocyte numbers were also decreased with anti–IL-7Rα treatment (31% reduction versus isotype control–treated mice) (Figure 2A). We used flow cytometry to determine which splenic immune cells were affected by anti–IL-7Rα treatment. No significant effect on numbers of B220+ B cells, CD11c+class II MHChigh dendritic cells, or CD11b+class II MHChigh macrophages was found (Figure 2B). In contrast, IL-7Rα blockade significantly decreased total numbers of CD4+ and CD8+ T cells compared with isotype control treatment (by an average of 49% and 58%, respectively) (Figure 2C), resulting in a significant (54%) reduction of CD3+ T cells (P < 0.05 versus isotype control treatment). Within the CD4+ T cell population, numbers of both CD44low naive and CD44high memory T cells were significantly reduced by IL-7Rα blockade (by an average of 52% and 39%, respectively) (Figure 2C). Similar reductions in numbers of naive and memory CD8+ T cells were found in the spleens of anti–IL-7Rα–treated mice compared with isotype control–treated mice (by an average of 66% and 37%, respectively).
IL-7Rα blockade does not significantly change anticollagen-specific IgG levels.
Not only decreased T cell immunity but also decreased B cell activity could play a role in the efficacy of IL-7Rα blockade. Therefore, antigen-specific B cell responses were examined by measuring anti-CII IgG1, IgG2a, and IgG2b concentrations in the serum of the mice. Comparable concentrations of anti-CII IgG antibody isotypes were found in mice treated with isotype control and anti–IL-7Rα mAb (Figure 3). Serum concentrations of anti-CII IgG antibodies also did not differ significantly between PBS-treated and anti–IL-7Rα–treated mice (data not shown).
IL-7Rα blockade reduces Th1, Th2, and Th17 activity.
Because IL-7Rα blockade led to significant decreases in T cell numbers, and IL-7 is known to induce proinflammatory T cell–associated cytokine production and naive and memory T cell proliferation, we investigated whether blockade of the IL-7/IL-7R pathway interferes with T cell activity. We examined the effect of IL-7Rα blockade on regulation of several proinflammatory mediators at the local level in paws and from lymph node cells. Local cytokine concentrations indicative of Th1 (IFNγ), Th2 (IL-5), and Th17 (IL-17) cell activity were significantly reduced by anti–IL-7Rα treatment compared with isotype control treatment (Figure 4A).
Because IL-17 has been described to play a pivotal role in CIA, we tried to confirm whether IL-7Rα blockade can prevent Th17 activity. IL-17 production was analyzed upon culture of draining lymph node cells from arthritic mice restimulated with CII. In the supernatant of CII-stimulated lymph node cells from anti– IL-7Rα–treated mice, significantly (80%) lower IL-17 production was detected compared with isotype control–treated mice (Figure 4B). Moreover, anti–IL-7Rα treatment significantly reduced IL-2 production (by 75%) upon restimulation with CII compared with isotype control treatment (data not shown).
Reduction of proinflammatory and tissue-destructive mediators upon IL-7Rα blockade.
In addition to T cell–related cytokines (Figure 4), the immunomodulatory effect of IL-7Rα blockade on proinflammatory cytokines, chemokines, and tissue factors was further analyzed with multicytokine analysis of lysates from the inflamed paws. Anti–IL-7Rα treatment significantly reduced IL-1β, IL-11, IL-18, TNFα, and leukemia inhibitory factor (LIF) concentrations compared with isotype control treatment (at least P < 0.05) (Figure 5). IL-6 concentrations were reduced in the inflamed paws (P = 0.059) and in serum (P = 0.002). IL-7Rα blockade also reduced concentrations of chemokines, including IFNγ-inducible 10-kd protein (IP-10), cytokine-induced neutrophil chemoattractant (KC; an IL-8 homolog), lymphotactin, monocyte chemotactic protein 5 (MCP-5), MIP-2, MIP-3β, and stem cell factor (SCF) (Figure 5).
Also, growth factors associated with tissue destruction/remodeling were regulated by IL-7Rα blockade; significantly reduced concentrations of RANKL, oncostatin M (OSM), and matrix metalloproteinase 9 (MMP-9) were found in the lysates from anti–IL-7Rα–treated mice. In addition, fibroblast growth factor 9 (FGF-9) was significantly reduced by anti–IL-7Rα treatment compared with isotype control treatment. Conversely, basic FGF (bFGF) was significantly increased upon IL-7Rα blockade (Figure 5). IL-7Rα blockade was also associated with a significant reduction (P < 0.05) in local concentrations of serum amyloid P (data not shown). In addition, cytokines indicative of vascular activation such as von Willebrand factor (vWF) and vascular cell adhesion molecule 1 (VCAM-1) were reduced (Figure 5).
Cytokines that were detectable but that were not significantly altered by anti–IL-7Rα treatment included CD40, eotaxin, IL-10, MCP-1, MCP-3, macrophage colony-stimulating factor, macrophage-derived chemokine, MIP-1β, MIP-1γ, myeloperoxidase, tissue factor, factor VII, and vascular endothelial growth factor.
Therapeutic administration of anti–IL-7Rα diminishes established arthritis in CIA.
Prophylactic anti–IL-7Rα treatment is relevant from a scientific standpoint, but may be less relevant from a clinical standpoint. Therefore, the therapeutic effect of IL-7Rα blockade on established arthritis was studied as well. When the mean clinical arthritis scores of a group of 15 mice reached 3, we initiated treatment with 100 μg anti–IL-7Rα or control treatments. Treatment with isotype control antibody did not yield a significant difference in arthritis scores compared with PBS treatment. Treatment with anti–IL-7Rα resulted in significantly less arthritis severity during the course of arthritis (P = 0.034 and P = 0.036 versus isotype control treatment and PBS treatment, respectively, for area under the curve on days 27–42) (Figure 6).
In this study we demonstrated that prophylactic blockade of IL-7Rα decreased arthritis severity and joint destruction in CIA. This was associated with reductions in T cell numbers, T cell–associated cytokines, and several mediators that are well known for their capacity to induce inflammation and tissue destruction. Furthermore, anti–IL-7Rα treatment was effective in reducing established CIA.
These current results corroborate our findings and those of others indicating the importance of IL-7 and IL-7Rα in immunopathology of RA. IL-7Rαbright T cells from RA patients are highly proliferative and largely lack FoxP3 expression, compared with poorly proliferating IL-7Rαlow/–FoxP3+ T cells (20–22). This strongly suggests that IL-7 primarily activates arthritogenic IL-7Rα+ T cells and promotes arthritis. IL-7 has been shown to strongly activate Th1 and Th17 cells and, to a lesser extent, Th2 cells in RA patients (12, 14, 15). In addition, blockade of IL-7R–mediated immune activation by soluble human IL-7R inhibited IFNγ production (indicative of Th1 cell activity) of ex vivo–cultured mononuclear cells from RA patients (20). Consistent with these findings, the present study demonstrates that IL-7Rα blockade reduces levels of IFNγ, IL-5, and IL-17, cytokines indicative of Th1, Th2, and Th17 activity.
In addition to inducing T cell–related cytokines, IL-7 induces several proinflammatory cytokines largely produced by monocytes, including TNFα, IL-1β, IP-10, monokine induced by IFNγ, MIP-1α, thymus and activation–regulated chemokine, and IL-8 (7, 12, 14, 16–18). In accordance with these capacities of IL-7, anti–IL-7Rα treatment significantly reduced a number of cytokines and chemokines including TNFα, IL-1β, IL-6, IL-11, IL-18, LIF, OSM, KC (the mouse equivalent of IL-8), IP-10, lymphotactin, SCF, MCP-5, MIP-2, and MIP-3β. The immunomodulatory effect of IL-7Rα blockade was also reflected by a significant reduction in proinflammatory mediators such as C-reactive protein and vWF, which have been shown to be increased in arthritic joints (23, 24) and are suggested to play a role in activation of monocyte/macrophages and endothelial cells. Taken together, these results indicate that these latter proteins also play an important role in cytokine production and leukocyte recruitment to sites of inflammation. In addition, we observed a significant reduction in cytokines indicative of vascular activation, such as VCAM-1 and vWF.
Complementary to clinical arthritis severity, anti–IL-7Rα treatment clearly reduced radiographic joint damage. This is in accordance with previous observations demonstrating that overexpression of IL-7 in transgenic mice induces T cell–dependent osteoclast activity leading to osteopenia and increased bone resorption (9). IL-7–induced activation of human effector CD4+ T cells up-regulates expression and production of RANKL, which is associated with induction of osteoclast differentiation and activation (25, 26). Corroborating these results, in the present study blockade of IL-7Rα diminished local RANKL protein expression. Decreased protein expression could prevent RANK/RANKL signaling, resulting in a reduction of osteoclast formation and maturation and in reduced bone resorption. The reduction in MMP-9 levels that was observed in this study are consistent with this finding, since RANKL-induced MMP-9 has been shown to be involved in bone resorption (27).
Not only osteoclasts, but also chondrocytes, have been shown to be influenced by IL-7/IL-7R signaling. IL-7 is produced by chondrocytes that are treated with fibronectin fragments, and increased IL-7 production has been noted in cells from aging and OA cartilage compared with healthy cartilage. Chondrocytes express IL-7R, and they respond to IL-7 stimulation with production of MMP-13 and with proteoglycan release from cartilage explants (28). Our analysis of radiographs demonstrated that loss of joint space, indicative of cartilage destruction, was prevented by anti–IL-7Rα treatment. Although a direct effect of IL-7R–mediated signaling on cartilage cannot be demonstrated, the present study demonstrated that IL-7Rα blockade prevents cartilage degradation as well as bone damage. In addition to direct effects on bone and cartilage as well as through inhibition of catabolic cytokines, prevention of joint destruction by IL-7Rα blockade may be caused by inhibition of tissue-destructive fibroblast activity. FGF-9 was significantly reduced by anti–IL-7Rα treatment compared with isotype control treatment. Conversely, bFGF was significantly increased by IL-7Rα blockade. It remains to be demonstrated to what extent these mediators contribute to synovial hyperplasia and tissue destruction.
Associated with strong suppression of cytokines indicative of T cell and monocyte/macrophage activation, IL-7Rα blockade also selectively reduced the numbers of naive and memory CD4+ and CD8+ T cells (while not significantly affecting the numbers of B cells, macrophages, and dendritic cells). This study did not analyze whether reduction of T cell numbers results from apoptosis induction. However, evidence from earlier studies indicates that effective inhibition of IL-7–dependent activation of lymphocytes by anti–IL-7R antibodies results from increased apoptosis (29). Furthermore, it has been shown that anti–IL-7R treatment reduces allogeneic T cell numbers (after bone marrow transplantation) by inhibition of IL-7R function, not antibody-mediated cell lysis (29). In addition, Liu et al showed that anti–IL-7R treatment specifically decreased Th17 expansion and survival, not differentiation, due to inhibition of STAT-5 phosphorylation. Importantly, that study also showed that anti–IL-7R treatment did not affect Treg cells (30).
Because the effects of IL-7Rα blockade on T cell activity may therefore be dependent on IL-7–induced regulation of T cell numbers, it is difficult to tell from this study whether the antiarthritic effects of IL-7Rα blockade are also induced by inhibition of T cell effector functions induced by IL-7Rα ligands. Although we have previously shown that T cell–associated cytokine expression per cell (IFNγ, TNFα) was significantly increased by IL-7 (12), such cytokine-secreting effector T cells still can result from enhanced selective survival, as was recently shown for IL-7–induced expansion of Th17 cells in experimental autoimmune encephalomyelitis (30). Taken together, our data demonstrate that IL-7Rα blockade selectively inhibits expansion of proinflammatory T cells, and that this is associated with prevention of arthritis. Inhibition of T cell activity seems to result primarily in the prevention of T cell–dependent monocyte/macrophage activation, because levels of circulating anti-CII antibodies, indicative of B cell activity, were not affected. This was also described for anti-TNFα treatment in CIA (31).
IL-7–induced T cell activation via IL-7R involves signaling via JAK-3, and inhibition of the IL-7R pathway might therefore contribute partially to the mechanism of action of the JAK-3 inhibitor that has been recently tested in RA. Although good antiarthritic effects are observed upon JAK-3 inhibition, specificity of some of these inhibitors is controversial (32). Changelian et al reported that JAK-3 inhibition by several inhibitors does not selectively block JAK-3 but does inhibit the epidermal growth factor receptor family of kinases (32). Therefore, when interpreting the results of studies with JAK-3 inhibitors, the (lack of) specificity of these inhibitors must be taken into consideration. In accordance with this, the use of JAK-3 inhibitors is associated with numerous side effects, such as neutropenia, anemia, increased levels of high-density lipoprotein/low-density lipoprotein and creatinine, headache, and nausea. In this respect it might be advantageous that IL-7R blockade selectively targets T cells that respond strongly to T cell receptor triggering (3, 20). Furthermore, JAK inhibition decreases FoxP3 expression in Treg cells, thereby affecting Treg cell function (33, 34). Targeting IL-7R, however, is not likely to affect Treg cell function, since Treg cells have been shown to have very low or no expression of IL-7R (20–22). Thus, the use of JAK-3 inhibitors results in broad immunosuppression and may be partly associated with nonselective effects, in contrast to a more selective immunomodulation by specific anti–IL-7R treatment.
Thymic stromal lymphopoietin (TSLP) is an IL-7–related stromal-derived cytokine that is also produced by epithelial cells and fibroblasts (35) and shares IL-7Rα for signaling (36). Although in vitro our anti–IL-7Rα antibody was 100-fold more effective at blocking IL-7–induced chemokine secretion than TSLP-induced chemokine secretion by mouse dendritic cells (Willis CR: unpublished observations), IL-7Rα blockade might also block TSLP-induced immune activation in CIA. In accordance with the observation that TSLP has been shown to promote Th2-mediated responses in mice (37, 38), IL-7Rα blockade also reduced local concentrations of IL-5, a cytokine that is indicative of Th2 cell activity. Consistent with a possible role of TSLP in arthritis, it has recently been shown that TNFα and Toll-like receptor ligands up-regulate TSLP production by RA- and OA-derived synovial fibroblasts (39). Despite this, the functional properties of TSLP in RA are unknown. Recently, we have demonstrated that TSLP does not activate mononuclear cells from RA patients, but that it might activate dendritic cells to activate CD4+ T cells from RA patients (20). However, the exact role of TSLP in arthritis remains to be demonstrated.
Our results demonstrate that blockade of IL-7Rα potently inhibits joint inflammation and destruction. This was associated with specific reductions of T cell numbers, T cell–associated cytokines, and several mediators that are able to induce inflammation and tissue destruction. Although further studies on the mechanisms of reduction in T cell activity are needed, our data indicate that targeting IL-7R–expressing T cells is a potential strategy for selective immunomodulation of RA as well as other rheumatic diseases.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. van Roon had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Hartgring, Willis, Bijlsma, Lafeber, van Roon.
Acquisition of data. Willis, Alcorn, Nelson.
Analysis and interpretation of data. Hartgring, Willis, Nelson, van Roon.