The critical role of kinase activity of interleukin-1 receptorassociated kinase 4 in animal models of joint inflammation.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 60, No. 6, June 2009, pp 1661–1671 DOI 10.1002/art.24552 © 2009, American College of Rheumatology The Critical Role of Kinase Activity of Interleukin-1 Receptor–Associated Kinase 4 in Animal Models of Joint Inflammation Magdalena Koziczak-Holbro, Amanda Littlewood-Evans, Bernadette Pöllinger, Jiri Kovarik, Janet Dawson, Gerhard Zenke, Christoph Burkhart, Matthias Müller, and Hermann Gram revealed that bone erosion, osteoclast formation, and cartilage matrix proteoglycan loss were reduced in IRAK-4 KD mice. Assessment of T cell response by MLR, by frequency of antigen-specific clones, and by production of antigen-specific IgG did not reveal substantial differences between IRAK-4 KD and wild-type mice. Conclusion. These results demonstrate that IRAK-4 is a key component for the development of proarthritis inflammation, but that it is not crucial for T cell activation. Therefore, the kinase function of IRAK-4 appears to be an attractive therapeutic target in chronic inflammation. Objective. We have previously reported that the kinase activity of interleukin-1 receptor–associated kinase 4 (IRAK-4) is important for Toll-like receptor and interleukin-1 receptor signaling in vitro. Using mice devoid of IRAK-4 kinase activity (IRAK-4 KD mice), we undertook this study to determine the importance of IRAK-4 kinase function in complex disease models of joint inflammation. Methods. IRAK-4 KD mice were subjected to serum transfer–induced (K/BxN) arthritis, and migration of transferred spleen lymphocytes into joints and cartilage and bone degradation were assessed. T cell response in vivo was tested in antigen-induced arthritis (AIA) by measuring the T cell–dependent antigenspecific IgG production and frequency of antigenspecific T cells in the spleen and lymph nodes. T cell allogeneic response was tested in vitro by mixed lymphocyte reaction (MLR). Results. Lipopolysaccharide-induced local neutrophil influx into subcutaneous air pouches was impaired in IRAK-4 KD mice. These mice were also protected from inflammation in the K/BxN and AIA models, as shown by reduced swelling of joints. Histologic analysis of joints of K/BxN serum–injected mice Interleukin-1 receptor (IL-1R)–associated kinase 4 (IRAK-4), a first proximal kinase downstream of IL-1R and most Toll-like receptors (TLRs), has been reported to be pivotal for receptor-induced signaling and proinflammatory mediator activation (1–3). Recently, genetically engineered mice expressing a kinasedeficient mutant of this protein (IRAK-4 KD) were generated (called IRAK-4 KD mice) (4–6). Using this approach, we and others have demonstrated that IRAK-4 kinase activity is crucial for IL-1R– and TLRmediated myeloid differentiation factor 88 (MyD88)– dependent signaling and expression of proinflammatory mediators in vitro. Furthermore, it has been shown that IRAK-4 kinase-inactive mice are completely resistant to lipopolysaccharide (LPS)– and CpG-induced septic shock, due to impaired TLR-mediated production of cytokines and chemokines (4,5). However, we have demonstrated recently that LPS can induce the expression of some macrophage gene products in an IRAK-4 kinase-independent manner, probably via TRIF and interferon regulatory factor 3 (7). Apart from the above-mentioned studies of TLR- Supported by Novartis Institutes for BioMedical Research, Switzerland. Magdalena Koziczak-Holbro, PhD, Amanda LittlewoodEvans, PhD, Bernadette Pöllinger, PhD, Jiri Kovarik, PhD, Janet Dawson, PhD, Gerhard Zenke, PhD, Christoph Burkhart, PhD, Matthias Müller, PhD, Hermann Gram, PhD: Novartis Institutes for BioMedical Research, Basel, Switzerland. Drs. Littlewood-Evans, Kovarik, Dawson, Zenke, Burkhart, Müller, and Gram own stock or stock options in Novartis. Address correspondence and reprint requests to Hermann Gram, PhD, Novartis Institutes for BioMedical Research, Novartis Pharma AG, Postfach, 4002 Basel, Switzerland. E-mail: email@example.com. Submitted for publication March 11, 2008; accepted in revised form February 25, 2009. 1661 1662 KOZICZAK-HOLBRO ET AL mediated acute lethality, there have been no in vivo studies identifying a role for IRAK-4 kinase activity in more complex and chronic autoimmune disorders largely dependent on innate immune system functions. In the present study, we investigated the role of IRAK-4 kinase activity in the antigen-induced arthritis (AIA) and KRNxNOD (K/BxN) serum transfer animal models of inflammation, which are T cell dependent and independent, respectively. TLRs, expressed predominantly on innate effector cells such as macrophages and dendritic cells, recognize pathogen-associated molecular patterns and initiate the innate inflammatory response. They can also bind host molecules, including breakdown products of the extracellular matrix such as hyaluronate and heparan sulfate, bind molecules that have been released from dead or damaged cells such as high mobility group box chromosomal protein 1 (HMGB-1), Hsp60, Hsp70, and fibronectin, and bind modified lowdensity proteins (8). TLR activation is also postulated to be involved in inflammatory reactions in rheumatoid arthritis (RA). TLR ligands of microbial origin (e.g., peptidoglycans and double-stranded DNA) have been detected in joints of RA patients (9). MATERIALS AND METHODS IRAK-4 KD mice. The IRAK-4 KD “knockin” mice were generated by replacing the wild-type (WT) gene with a gene containing a mutation in the IRAK-4 kinase domain and were described previously (6). The genetic background of the mice is BALB/c. WT control mice were derived from the initial heterozygous founder generation. All animal studies were performed in accordance with the animal experimentation laws and guidelines laid down by the Swiss Federal and Cantonal Authorities. Mouse air pouch model. Air pouches were produced in female WT and IRAK-4 KD mice by the subcutaneous injection of 3 ml of air (via a 0.22-m filter) into their backs according to the procedure described by Dawson et al (10). On day 6, 1 ml of pyrogen-free saline or 1 g/ml LPS (Escherichia coli serotype 0127:B8; Sigma, St. Louis, MO) was injected into the pouches. Six hours later, the mice were killed by CO2 asphyxiation, and the pouch contents were washed out with 1 ml of sterile phosphate buffered saline buffer. Total and differential cell counts were performed essentially as described previously (10). AIA. The mouse model of AIA was conducted and assessment of joint swelling was performed as described previously (11). Serum antibody titers to methylated bovine serum albumin (mBSA) were determined according to the method of Yang et al (12). Serum transfer model and arthritis scoring. Arthritis was induced by intraperitoneal injection of 250 g K/BxN serum into recipient mice on day 0. Severity of paw swelling was scored in the metatarsal region (range 0–3) and ankle region (range 0–3) of each paw to give a maximum score of 6 per paw and 24 per mouse. The individual sum scores of all the animals were averaged and SEMs were calculated. The scoring system used was as follows: 0 ⫽ no detectable sign of inflammation; 1.0 ⫽ entire paw swollen; 2.0 ⫽ swollen paw involving wrist or ankle; 3.0 ⫽ ankylosis or severely swollen paw. Histology. Joints were embedded in methylmethacrylate (Fluka, Buchs, Switzerland) (11), and sections of ⬃5 m thickness were stained with Giemsa or for tartrate-resistant acid phosphatase (TRAP) according to standard protocols. Histologic assessment of the joints was performed according to the following scoring system: 0 ⫽ no bone or cartilage loss, and 1–3 ⫽ mild (10–29%), 4–7 ⫽ moderate (30–59%), and 8–10 ⫽ strong (60–100%) proteoglycan (cartilage) loss and bone erosion. In vivo T cell activity assay. Each mouse was injected intraperitoneally with 100 g of 2,4-dinitrophenyl (DNP; Sigma) conjugated at a ratio of 20:1 to keyhole limpet hemocyanin (KLH; Calbiochem, La Jolla, CA) adsorbed to aluminum hydroxide (Alu-Gel-S; Serva, Heidelberg, Germany) as an adjuvant. On day 21, mice were boosted with DNP-KLH in Alu-Gel-S. Mice were bled 8 days after the first and second immunizations, respectively, and serum samples were analyzed for anti-DNP antibodies by enzyme-linked immunosorbent assay. IgG subclass–specific antibodies (against IgG1, IgG2a, IgG2b, and IgG3) were obtained from Southern Biotechnology (Birmingham, AL). Antibody titers are expressed as the dilution leading to half-maximal optical density at 405 nm. One-way mixed lymphocyte reaction (MLR). The MLR was performed as described previously (13). In vitro T cell recall responses to mBSA. Spleens and draining inguinal lymph nodes were harvested from AIA mice 1 day after the local injection of mBSA into a single knee joint cavity in each mouse. Single-cell suspensions were prepared in RPMI 1640 medium/10% fetal calf serum, and cells were seeded into 96-well enzyme-linked immunospot (ELISpot) assay microplates at densities of 2 ⫻ 106 cells/ml for lymph node and 4 ⫻ 106 cells/ml for spleen cells. The antigen-specific T cells were then restimulated for 72 hours with various concentrations of mBSA (25–100 g/ml; Sigma). The quantitative determination of the frequency of IL-2–secreting cells was then performed using the Dual-Color ELISpot kit (R&D Systems, Minneapolis, MN) according to the manufacturer’s instructions. The number of spots per well was counted using a dissection microscope. Adoptive transfer of 5,6-carboxyfluorescein succinimidyl ester (CFSE)–labeled splenocytes. Splenocytes from BALB/c WT or IRAK-4 KD donor mice were purified from red blood cells by hypotonic shock (0.83% NH4Cl for 5 minutes at room temperature), labeled with CFSE (Sigma) as described (see ref. 14), and adoptively transferred into BALB/c WT or IRAK-4 KD recipient mice which received K/BxN serum shortly afterward. After 5 days, the spleen, lymph nodes, and (arthritic) joints of recipient mice were analyzed. The absolute number of cells in the joint was determined by normalization to the total number of cells that were obtained by the above-described method, when an equal number of arthritic and healthy paws were processed. Isolation of cells from the joint. Cells from the joint were prepared as recently described (15). Briefly, the skin was removed, and whole paws were dissected into small pieces that were placed into 0.125% Dispase II (Roche, Indianapolis, IN), IRAK-4 KINASE ACTIVITY IN JOINT INFLAMMATION 0.2% collagenase 2 (Sigma-Aldrich, Bornem, Belgium), and 0.2% collagenase 4 (Sigma-Aldrich) and shaken for 75 minutes at 37°C. After digestion, joint pieces were passed through a cell strainer (BD Falcon, Bedford, MA). Nonspecific binding sites were blocked using mouse CD16/32 purified antibody (no. LMF-CR00-4; Caltag, South San Francisco, CA). Antibodies used for subsequent fluorescence-activated cell sorting staining were all purchased from BD PharMingen (San Diego, CA). These were phycoerythrin-conjugated anti-CD11b (no. 553311), peridinin chlorophyll A protein–conjugated antiB220 (no. 552771), and allophycocyanin-conjugated anti-CD3 (no. 553066). Cells were acquired using a FACSCalibur instrument (BD Biosciences, San Jose, CA) and analyzed using FlowJo 6.42 software (Tree Star, San Carlos, CA). Angiogenesis chamber assay. This assay has been described previously (16). Implanted chambers contained agar with or without 250 ng mouse IL-1␤ (R&D Systems). After 7 days, the vascularized tissue formed around each implant was weighed and homogenized in radioimmunoprecipitation assay buffer, and levels of tyrosine protein kinase receptor 2 (TIE-2) were determined (16). Statistical analysis. Data are reported as the mean ⫾ SEM and were analyzed by analysis of variance or Student’s t-test. P values less than 0.05 were considered significant. RESULTS LPS-induced influx of neutrophils into an air pouch is impaired in IRAK-4 KD mice. Since our previous data demonstrated that IRAK-4 kinase activity is critical for the maintenance of normal proinflammatory reactions in vitro (6,7), in the present studies we analyzed different aspects of the inflammatory response in vivo. Cell recruitment is a pivotal event in inflammation, and cell number and composition in the initial stages influence future inflammatory responses. To this end, we analyzed inflammatory cell migration in the mouse air pouch model of chemotaxis. Air pouches were formed in WT and IRAK-4 KD mice and injected with LPS or saline control. LPS induced a rapid recruitment of polymorphonuclear neutrophils (PMNs) into the pouch of WT animals, while the numbers of mononuclear cells (MNCs) were similar in both control and LPS-treated WT mice 6 hours after LPS injection (Figure 1). In contrast, the IRAK-4 KD animals were largely protected from LPS-induced inflammatory PMN influx into the pouch. IRAK-4 kinase activity has a critical role in mouse models of induced arthritis. To further determine the role of IRAK-4 kinase activity in inflammatory responses in vivo, WT and IRAK-4 KD mice were studied using the K/BxN animal model of RA (17). Transfer of serum containing arthritogenic Ig from K/BxN mice into healthy animals induces rapid development of arthritis (18). After transfer of K/BxN serum, 1663 Figure 1. Inhibition of neutrophil recruitment after lipopolysaccharide (LPS) stimulation in mice devoid of interleukin-1 receptor– associated kinase 4 (IRAK-4) kinase activity (IRAK-4 KD mice). Air pouches of wild-type (WT) and IRAK-4 KD mice (n ⫽ 4 per group) were injected with LPS or phosphate buffered saline (PBS) as a control. After 6 hours, neutrophil infiltration into the pouch was determined. Values are the mean and SEM. ⴱⴱ ⫽ P ⬍ 0.01. PMN ⫽ polymorphonuclear neutrophils; MN ⫽ mononuclear cells. WT animals developed severe joint inflammation, as assessed by swelling of the paws (Figure 2C). IRAK-4 KD mice essentially showed no clinical signs of disease. Histologic sections of ankle joints were stained with Giemsa or for TRAP, a marker of osteoclast differentiation. In WT mice injected with K/BxN serum, joint space was narrowed and infiltrated with inflammatory cells, bone contours were irregular, showing pannus invasion, and the proteoglycan layer was largely absent, which reflects cartilage damage (Figure 2A). Higher magnification revealed that the most abundant cell types invading the joints were PMNs and mononuclear macrophages. New blood vessel formation was also visible at the site of inflammation in WT animals. In contrast, IRAK-4 KD mice displayed no signs of joint inflammation or bone lesions. The bone surface appeared smooth, the joint space was wide and free of inflammatory cells, the synovium consisted of a thin lining layer 1–3 cells thick covering adipose and connective tissue, and the cartilage was relatively smooth. In line with the results from the Giemsa histology, TRAP stains showed abundant osteoclasts on the bone surface, especially in areas of pannus invasion only in WT animals treated with K/BxN serum (Figure 2B). TRAP staining was not detectable in the histologic sections of ankle joints of IRAK-4 KD animals. Independent scoring of multiple slides of stained ankle joint sections verified that IRAK-4 KD mice were completely 1664 KOZICZAK-HOLBRO ET AL Figure 2. Protection of IRAK-4 KD mice against K/BxN serum transfer–induced inflammation and bone destruction. WT and IRAK-4 KD mice (n ⫽ 4 per group) were injected intraperitoneally with K/BxN serum. A, Giemsa-stained sections and B, tartrate-resistant acid phosphatase (TRAP)–expressing osteoclast staining of ankle joints obtained from animals on day 8 after K/BxN serum injection. Boxed areas in top panels are depicted at higher magnification in bottom panels. C, Progression of inflammation monitored by scoring paw swelling up to day 8 (see Materials and Methods). Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱⴱ ⫽ P ⬍ 0.001, versus IRAK-4 KD mice. D, Histology scores for bone erosion, proteoglycan loss, and TRAP staining on day 8 in the animals treated as described above. Results shown are representative of 3 experiments with a total of 11–13 mice. Values are the mean ⫾ SEM. ⴱⴱ ⫽ P ⬍ 0.01; ⴱⴱⴱ ⫽ P ⬍ 0.001. See Figure 1 for other definitions. protected from inflammatory cell infiltration and bone erosion. IRAK-4 KD mice were partially protected against cartilage damage in the serum transfer model (Figure 2D). Taken together, these results indicate that in the K/BxN serum transfer model, IRAK-4 kinase activity plays an important role and is critically required for disease progression. We next investigated whether deficiency of IRAK-4 KINASE ACTIVITY IN JOINT INFLAMMATION IRAK-4 kinase activity in hematopoietic cells or stromal cells is responsible for the absence of joint inflammation. CFSE-labeled spleen cells from WT and IRAK-4 KD mice were intravenously injected into recipient mice which were subsequently treated with K/BxN serum to induce joint inflammation. Transferred cells from WT and IRAK-4 KD mice migrated to the same extent to the inflamed joint and spleen of WT recipients (Figure 3). Immunophenotyping revealed that T cells (CD3⫹), B cells (B220⫹), and monocyte/macrophages (CD11b⫹) migrated to the inflamed joint. In contrast, WT cells transferred into IRAK-4 KD recipient mice failed to induce joint swelling, and fewer transferred WT cells were detected in the joints. Homing to lymphoid tissue, such as the spleen, did not differ between WT and IRAK-4 KD recipients. We conclude from this experiment that IRAK-4 kinase deficiency does not ablate the recruitment of T cells, B cells, and macrophages to the site of inflammation, whereas the IRAK-4 defect in stromal cells influences recruitment to the site of inflammation. We next sought to determine whether IRAK-4 kinase activity is important in another experimental T cell–dependent murine joint inflammation model, AIA (19,20). AIA is induced by systemic immunization with mBSA in Freund’s complete adjuvant (CFA), followed by local injection of mBSA into a single knee joint cavity in each mouse. The inflammatory reaction in this model is considered to be initiated by T cells that respond to mBSA-activated antigen-presenting cells and that produce T cell–specific cytokines that thus induce innate inflammatory responses leading to joint edema (19,21). The mBSA-mediated induction of joint swelling was significantly suppressed in IRAK-4 KD mice compared with WT mice (Figure 4A), indicating that IRAK-4 is essential for formation of joint edema. Joint histology revealed a reduced, but not absent, influx of leukocytes into the joints of IRAK-4 KD mice (Figure 4B), and proteoglycan loss indicative of cartilage damage did not differ between WT and IRAK-4 KD mice (data not shown). To determine the humoral immune response in WT mice, serum IgG levels against the arthritogenic mBSA were measured. There was already a slightly decreased production of the mBSA-specific IgG in IRAK-4 KD mice compared with WT mice, after mBSA sensitization (Figure 4C). However, after the mBSA intraarticular challenge, IgG titers were comparably increased in both IRAK-4 KD and WT animals (8-fold and 11-fold, respectively). 1665 Figure 3. K/BxN serum transfer–induced arthritis after adoptive transfer of 5,6-carboxyfluorescein succinimidyl ester (CFSE)–labeled splenocytes. Shown is the progression of arthritis in WT or IRAK-4 KD recipient mice (WT donors into WT recipients [n ⫽ 6], IRAK-4 KD donors into WT recipients [n ⫽ 4], and WT donors into IRAK-4 KD recipients [n ⫽ 8]). A, Scoring of paw swelling is shown as in Figure 2 (see Materials and Methods). B and C, In a separate experiment, 5 days after transfer, mice were killed, and spleens and joints were prepared and analyzed for CFSE-positive cells. Shown are the percentages of CFSE-positive cells and cell subsets in the spleen (B) and the total numbers of CFSE-positive cells and cell subsets in the joints (C) (WT donors into WT recipients [n ⫽ 6], IRAK-4 KD donors into WT recipients [n ⫽ 4], and WT donors into IRAK-4 KD recipients [n ⫽ 4]). There were no statistically significant differences among the 3 donor/recipient groups (P ⬎ 0.05). Values are the mean ⫾ SEM. See Figure 1 for other definitions. IRAK-4 kinase activity is required for antigenspecific T cell activation. To assess the T cell response in this model more directly, we quantified the antigen- 1666 KOZICZAK-HOLBRO ET AL Figure 4. Protection of IRAK-4 KD mice against antigen-induced arthritis. A, Arthritis in WT and IRAK-4 KD animals (n ⫽ 9–10 per group) injected with methylated bovine serum albumin (mBSA) was monitored by measuring knee joint swelling over 7 days. Data are expressed as the mean ⫾ SEM of the ratio of the right (R) arthritic joint diameter to the left (L) control joint diameter. ⴱⴱⴱ ⫽ P ⬍ 0.001 versus WT mice. B, Cellular infiltration of the synovium was assessed by blinded scoring of histology slides, where normal synovium is scored 0 and inflamed tissue is scored up to a maximum of 5. Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05. C, The humoral immune response was determined by measuring levels of IgG antibodies raised against mBSA in sera on day –1 before challenge and on day 7 after challenge. The antibody titers are expressed as mean and SEM log10 values using 50% of the maximal extinction at 492 nm as an end point. ⴱⴱ ⫽ P ⬍ 0.01. D, Enzyme-linked immunospot assay was used to determine the frequency of interleukin-2 (IL-2)–secreting antigen-specific T cells in the spleen and draining lymph nodes of WT and IRAK-4 KD animals on day 1 after mBSA challenge. Results shown are from 1 experiment representative of 2 experiments (number of IL-2 spots/well) in which cell pools from 3 mice per group were analyzed in duplicate after in vitro restimulation for 72 hours using various concentrations of mBSA. Values are the mean. OD ⫽ optical density (see Figure 1 for other definitions). specific spleen cell response in mice immunized with mBSA. Spleen and lymph node lymphocytes were isolated from WT and IRAK-4 KD mice, and the frequency of mBSA-specific IL-2–producing T cells was determined by an ELISpot assay (Figure 4D). Both WT and IRAK-4 KD mice immunized with mBSA showed a sizable frequency of activated T cells in both organs, which could be slightly increased by in vitro stimulation with mBSA. The number of IL-2–secreting T cells in IRAK-4 KD mice was found to be somewhat lower than that in WT mice. We also assessed the antigen-specific proliferative response of splenocytes in WT and IRAK-4 IRAK-4 KINASE ACTIVITY IN JOINT INFLAMMATION KD mice, and this was similar in both strains (data not shown). Taken together, we found no evidence for a major defect in the T or B cell compartments with regard to immune responses to mBSA in IRAK-4 KD mice. Since recently reported studies in IRAK-4 KD and IRAK-4–/– mice showed conflicting results with respect to the role of IRAK-4 in T cell activation (4,22), we wanted to investigate the T cell response in more detail in IRAK-4 KD mice. First, we tested the IRAK-4 kinase activity requirement for T cell function in an antihapten immune response in vivo. Mice were immunized with DNP-KLH conjugate in aluminum hydroxide as adjuvant, and antibodies against DNP were measured 8 days later. Antibody formation against DNP is entirely dependent on T cell activation in this model (23). Both WT and IRAK-4 KD mice showed comparable production of IgG and IgM antibodies (Figure 5A), suggesting that T cell responses are independent of IRAK-4 kinase activity. Also, a detailed subclass analysis revealed no significant difference for IgG1, IgG2a, IgG2b, and IgG3 titers (data not shown). Alloantigen-induced proliferation of T cells isolated from the spleens of IRAK-4 KD animals as determined in a one-way MLR appeared slightly reduced compared with WT mice; however, the difference was not statistically significant (Figure 5B). In addition, proliferation of purified T cells in response to immobilized anti-CD3 was not impaired in IRAK-4 KD mice compared with WT mice (data not shown). Taken together, these results strongly suggested that IRAK-4 is not directly involved in T cell activation via the T cell receptor (TCR). The IL-1␤ –induced angiogenic response is blocked in IRAK-4 KD mice. Angiogenesis is a critical component of inflammation and has been shown to be implicated in RA progression in patients (24) as well as in experimental models of arthritis (25). Because IL-1␤ is a proangiogenic factor in inflammatory conditions (26), we tested the effects of the IRAK-4 kinase deficiency in an IL-1␤–driven angiogenesis agar chamber model. WT and IRAK-4 KD mice were implanted with agar chambers with or without IL-1␤. After 7 days, a new blood vessel–rich tissue was formed around the IL-1␤– containing chamber in WT animals (Figure 6A). In contrast, the tissue around chambers implanted in IRAK-4 KD mice was thin, and few vessels were visible. The tissue from around chambers was removed, weighed, homogenized, and analyzed for the total amount of TIE-2 protein, indicative of activated endothelial cells and, therefore, of vascularity. IL-1␤ increased the weight of tissue growing around the implanted chambers as well as the TIE-2 content of the 1667 Figure 5. T cell activation is not affected in IRAK-4 KD animals. A, In vivo T cell activation. WT and IRAK-4 KD animals (n ⫽ 6 per group) were immunized with 2,4-dinitrophenyl–keyhole limpet hemocyanin on days 1 and 21. IgG and IgM production was measured in blood serum on day 8 (first immunization) and on day 28 (second immunization). WT naive nonimmunized animals served as a control for background measurement. Antibody titers are expressed as mean ⫾ SD log10 dilution values using 50% of the maximal extinction at 405 nm as the end point. B, In vitro T cell activation. Responder T cells purified from the spleens of WT and IRAK-4 KD animals with a BALB/c background (n ⫽ 3 per group) were cocultured with irradiated stimulator spleen cells from CBA mice at the indicated ratios. T cell proliferation was assessed after 4 days using 3H-thymidine incorporation. Background values of responder T cells without stimulator cells were subtracted. Values are the mean ⫾ SEM of triplicate cultures from 1 of 3 experiments. OD ⫽ optical density (see Figure 1 for other definitions). tissue in WT animals, but not in IRAK-4 KD animals (Figures 6B and C). These data suggest that IRAK-4 kinase activity is also crucial in mediating IL-1– 1668 Figure 6. Impairment of interleukin-1␤ (IL-1␤)–mediated angiogenesis in IRAK-4 KD mice. WT and IRAK-4 KD mice (n ⫽ 4–6 per group) were implanted with agar chambers with or without IL-1␤. After 7 days, chambers were explanted. A, Blood vessel growth around agar chambers (original magnification ⫻4). B, Weight of tissue around agar chambers. C, Expression of tyrosine protein kinase receptor 2 (TIE-2), a marker of activated endothelial cells, in tissue growing around implanted chambers. Values are the mean ⫾ SEM. ⴱ ⫽ P ⬍ 0.05; ⴱⴱⴱ ⫽ P ⬍ 0.001. See Figure 1 for other definitions. dependent angiogenesis that is part of the inflammatory process. DISCUSSION Joint inflammation in murine models can be induced by different mechanisms. AIA induced by local KOZICZAK-HOLBRO ET AL mBSA injection appears to depend largely on a T cell response (21), while the serum transfer K/BxN model depends on complement, immune complexes, and IL-1 signaling. TLR-2 and TLR-4 have been shown to be critically involved in various murine models of induced arthritis (27,28) and may well be involved in AIA and K/BxN arthritis as well. TLR signaling not only might rely on the presence of bacterial products, but also might be induced by endogenous TLR ligands, such as Hsp22 (29) or HMGB-1, which might be released from necrotic cells at the site of inflammation. Using IRAK-4 KD knockin mice, we have previously shown that IL-1R– and TLR-mediated MyD88-dependent signaling and expression of proinflammatory cytokines are largely dependent on intact IRAK-4 kinase activity in vitro (6,7). Since both IL-1 and TLR signaling are involved in arthritis models, we reasoned that IRAK-4 kinase deficiency would lead to amelioration of experimentally induced joint inflammation in mice. However, IRAK-4 may play different roles in the mainly T cell–driven AIA and the antibody- and immune complex–driven K/BxN arthritis. Although LPS-mediated signaling in vitro is for the most part ablated, an IRAK-4/MyD88–independent pathway exists for TLR-4 signaling which leaves the induction of a number of messenger RNAs intact (7). IRAK-4 kinase deficiency significantly reduced, but did not completely block, the influx of neutrophils into the pouch in response to LPS. We did not observe migration of MNCs (e.g., macrophages) at 6 hours, which is probably due to the fact that these cells are typically recruited at later time points in this model. To investigate further the role of IRAK-4 in more chronic disease models of inflammation that reflect a complex interplay between the acquired and innate immune systems, we chose the immune complex– dependent K/BxN and the T cell–dependent AIA mouse models of joint inflammation. IRAK-4 KD mice were protected from joint inflammation in K/BxN serum transfer–induced arthritis. Detailed histologic examination of the ankle joints of K/BxN serum–injected IRAK-4 KD mice showed absence of inflammatory cell infiltration and pannus formation and no signs of bone destruction or cartilage proteoglycan loss compared with WT mice. Also, activated osteoclasts at the focal sites of bone erosion were only present in the WT animals. It is known that in the K/BxN serum transfer model of arthritis, IL-1 plays a central role and is critically required for disease progression (30), but other mechanisms such as complement activation have also been identified as relevant mechanisms in this model (31). IRAK-4 KINASE ACTIVITY IN JOINT INFLAMMATION Thus, the strong protective effect of kinase deficiency observed in this partly IL-1–dependent model is in accordance with our previous in vitro studies in which cytokine and chemokine expression were also impaired in IRAK-4 KD mouse cells stimulated with IL-1␤ or TLR-7 ligand (6). In addition to the absence of inflammatory cell infiltration, no new vessel formation was observed in joints of the K/BxN serum–injected IRAK-4 KD mice, in contrast to WT mice (data not shown). Interestingly, IRAK-4 kinase deficiency also severely affected vascularization of IL-1␤–loaded agar chambers, suggesting a strong correlation between functional IRAK-4 kinase activity and IL-1␤–induced inflammatory angiogenesis. Splenocyte transfer experiments revealed no general defect in the migratory capacity of IRAK-4 KD mouse cells of hematopoietic origin for homing or migration to the inflammatory site in WT mice, whereas the IRAK-4 KD defect in the recipient led to a reduced influx of lymphocytes from WT mice (Figure 3). This finding is reminiscent of similar observations in MyD88deficient bone marrow–reconstituted mice, in which MyD88–/– hematopoietic cells had retained some migratory capacity in WT mice, but WT cells did not migrate into MyD88⫺/⫺ recipients (32). Transfer of WT mouse splenocytes did not restore joint inflammation in IRAK-4 KD mice (Figure 3), even when the observation period was prolonged to 14 days after serum transfer (data not shown). This inability to induce even mild inflammation could be due to a reduced or impaired recruitment of leukocytes to the joint. We did not observe a selective effect on the recruitment of specific cell populations studied (CD3⫹, B220⫹, and CD11b⫹), but we cannot rule out a specific defect for specialized cell types, such as, for example, mature dendritic cells which may be crucial to establish joint inflammation in this model. Alternatively, IRAK-4 KD mouse stromal cells might provide important signals for the activation of joint leukocytes. To address the role of IRAK-4 in T cell–driven joint inflammation, we employed the murine AIA model. The absence of IRAK-4 kinase activity also prevents joint swelling in this model, although the T cell response to mBSA appeared largely undisturbed in IRAK-4 KD mice. The number of infiltrating leukocytes was significantly lower in IRAK-4 KD mouse joints, but leukocytes were not entirely absent (Figure 4B). In the absence of specific information on the activation state of infiltrating leukocytes, we speculate that the strong effect on edema formation might be due to the reduced presence and/or activation of monocytes or neutrophils 1669 in the joint. Ablation of IL-1 signaling in AIA only partly protects against joint swelling in AIA (33), and the strong effect we observe in IRAK-4 KD mice might be due to additional blockade of TLR signaling. The observed loss of proteoglycan did not correlate with edema or cellular infiltration in our experiment, a finding that has also been observed by others and that may not be uncommon in this model (33,34). These broader analyses of inflammatory responses in animal models are also in accordance with previous data describing the kinase activity of IRAK-4 as essential for the function of IRAK-4 in vivo. Initial reported in vivo studies have shown that cytokine production is suppressed in IRAK-4 KD mice challenged with IL-1␤ and different TLR ligands, including LPS and CpG (5,7). These mice were resistant to LPS- or CpGinduced septic shock (4,5), similar to IRAK-4–/– mice (35). Two studies revealed that IRAK-4–/– mice have impaired T cell responses to lymphocytic choriomeningitis virus infection in vivo, including T cell proliferation and virus-specific cytotoxicity (22,36). Suzuki et al (22) suggested that IRAK-4 is directly engaged in signaling downstream of the TCR. In contrast, Kawagoe et al showed intact T cell responses as well as TCR signaling in both IRAK-4–/– and IRAK-4 KD mice (4). Kawagoe et al suggest that a possible, but unlikely, explanation for this discrepancy might be the different genetic backgrounds of the mouse strains used in the 2 studies. The mice used for the present study are of BALB/c background, a background different from that used in the other studies cited above. Our studies concerning direct T cell activation in vitro by CD3 ligation or in an allogeneic response and in several in vivo assays revealed no significant defect in IRAK-4 KD mice. However, there appears to be a slightly lower frequency or activity of IRAK-4 KD mouse T cells in the mBSA-immunized mice or the MLR, respectively, although these experiments do not suggest a fundamental defect in the T cell function. A possible explanation for the slightly lower response in the MLR could be the absence of TLR signaling. Endogenous TLR ligands released in activated mixed cell cultures could also drive lymphocyte proliferation. The only notable difference we observed was a decrease in production of anti-mBSA IgG in IRAK-4 KD mice compared with WT mice after mBSA sensitization, in which CFA containing TLR ligands was used as the adjuvant. TLR ligands can contribute indirectly to T cell differentiation (37), and the reduction in IgG titers after mBSA sensitization is probably due to a partial activity 1670 KOZICZAK-HOLBRO ET AL of the adjuvant in IRAK-4 KD mice. Boosting with antigen in the absence of adjuvant produced a similar increase in anti-mBSA titer in both WT and IRAK-4 KD mice. Taken together, our data suggest that the reduced joint swelling observed for IRAK-4 KD mice in AIA is not due to an impairment of the initial T cell–mediated immune response to mBSA. Also, recently reported studies showed that only innate immunity, but not acquired immunity, has a functional defect in human IRAK-4–deficient patients, similar to that seen in IRAK-4 KD mice (38). While production of key cytokines was completely impaired in response to IL1R– and TLR-mediated IRAK-4–dependent activation, protein antigen–specific T and B cell responses were normal in cells from IRAK-4–deficient patients. As suggested by Ku et al (38), it is likely that in IRAK-4– deficient patients acquired immunity plays a greater role in the control of infections than does TLR-induced innate immunity. Thus, the discovery that IRAK-4 kinase activity is essential in innate immunity, without affecting acquired immune responses, is not only important for understanding of its physiologic function, but might also have implications for the development of future antiinflammatory drugs. Pharmacologic inhibition of IRAK-4 kinase activity could be beneficial in prevention of inflammation, bone and cartilage destruction, and inflammatory angiogenesis in arthritis patients, while sparing the adaptive immune system. ACKNOWLEDGMENTS The excellent technical assistance of Ulrike Strittmatter-Keller, Petra Kessler, Bernard Meyer, Tanja Kobel, Rita Nagele, Regina Santoro, and Bernhard Jost is gratefully acknowledged. AUTHOR CONTRIBUTIONS 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. Gram 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. Koziczak-Holbro, Littlewood-Evans, Pöllinger, Kovarik, Dawson, Gram. Acquisition of data. Koziczak-Holbro, Littlewood-Evans, Pöllinger, Kovarik, Dawson, Zenke, Burkhart, Müller. Analysis and interpretation of data. Koziczak-Holbro, LittlewoodEvans, Pöllinger, Kovarik, Dawson, Zenke, Gram. REFERENCES 1. Li S, Strelow A, Fontana EJ, Wesche H. IRAK-4: a novel member of the IRAK family with the properties of an IRAK-kinase. Proc Natl Acad Sci U S A 2002;99:5567–72. 2. Liu ZJ, Liu CA, Gong JP. IRAK-4: the key molecule of TLR/ IL-1R common signal transduction system. Sheng Li Ke Xue Jin Zhan 2005;36:276–9. In Chinese. 3. Suzuki N, Suzuki S, Yeh WC. IRAK-4 as the central TIR signaling mediator in innate immunity. Trends Immunol 2002;23:503–6. 4. Kawagoe T, Sato S, Jung A, Yamamoto M, Matsui K, Kato H, et al. Essential role of IRAK-4 protein and its kinase activity in Toll-like receptor-mediated immune responses but not in TCR signaling. J Exp Med 2007;204:1013–24. 5. Kim TW, Staschke K, Bulek K, Yao J, Peters K, Oh KH, et al. A critical role for IRAK4 kinase activity in Toll-like receptormediated innate immunity. J Exp Med 2007;204:1025–36. 6. Koziczak-Holbro M, Joyce C, Gluck A, Kinzel B, Muller M, Tschopp C, et al. IRAK-4 kinase activity is required for interleukin-1 (IL-1) receptor- and Toll-like receptor 7-mediated signaling and gene expression. J Biol Chem 2007;282:13552–60. 7. Koziczak-Holbro M, Gluck A, Tschopp C, Mathison JC, Gram H. IRAK-4 kinase activity-dependent and -independent regulation of lipopolysaccharide-inducible genes. Eur J Immunol 2008;38: 788–96. 8. Marshak-Rothstein A. Toll-like receptors in systemic autoimmune disease. Nat Rev Immunol 2006;6:823–35. 9. Van der Heijden IM, Wilbrink B, Tchetverikov I, Schrijver IA, Schouls LM, Hazenberg MP, et al. Presence of bacterial DNA and bacterial peptidoglycans in joints of patients with rheumatoid arthritis and other arthritides. Arthritis Rheum 2000;43:593–8. 10. Dawson J, Sedgwick AD, Edwards JC, Lees P. A comparative study of the cellular, exudative and histological responses to carrageenan, dextran and zymosan in the mouse. Int J Tissue React 1991;13:171–85. 11. Grosios K, Wood J, Esser R, Raychaudhuri A, Dawson J. Angiogenesis inhibition by the novel VEGF receptor tyrosine kinase inhibitor, PTK787/ZK222584, causes significant anti-arthritic effects in models of rheumatoid arthritis. Inflamm Res 2004;53: 133–42. 12. Yang YH, Morand EF, Getting SJ, Paul-Clark M, Liu DL, Yona S, et al. Modulation of inflammation and response to dexamethasone by annexin 1 in antigen-induced arthritis. Arthritis Rheum 2004;50:976–84. 13. Weckbecker G, Bruns C, Fischer KD, Heusser C, Li J, Metzler B, et al. Strongly reduced alloreactivity and long-term survival times of cardiac allografts in Vav1- and Vav1/Vav2-knockout mice. Transpl Int 2007;20:353–64. 14. Lyons AB, Parish CR. Determination of lymphocyte division by flow cytometry. J Immunol Methods 1994;171:131–7. 15. Roark CL, French JD, Taylor MA, Bendele AM, Born WK, O’Brien RL. Exacerbation of collagen-induced arthritis by oligoclonal, IL-17-producing ␥␦ T cells. J Immunol 2007;179:5576–83. 16. LaMontagne K, Littlewood-Evans A, Schnell C, O’Reilly T, Wyder L, Sanchez T, et al. Antagonism of sphingosine-1-phosphate receptors by FTY720 inhibits angiogenesis and tumor vascularization. Cancer Res 2006;66:221–31. 17. Kouskoff V, Korganow AS, Duchatelle V, Degott C, Benoist C, Mathis D. Organ-specific disease provoked by systemic autoimmunity. Cell 1996;87:811–22. 18. Ji H, Gauguier D, Ohmura K, Gonzalez A, Duchatelle V, Danoy P, et al. Genetic influences on the end-stage effector phase of arthritis. J Exp Med 2001;194:321–30. 19. Brackertz D, Mitchell GF, Mackay IR. Antigen-induced arthritis in mice. I. Induction of arthritis in various strains of mice. Arthritis Rheum 1977;20:841–50. 20. Pohlers D, Nissler K, Frey O, Simon J, Petrow PK, Kinne RW, et al. Anti-CD4 monoclonal antibody treatment in acute and early chronic antigen-induced arthritis: influence on T helper cell activation. Clin Exp Immunol 2004;135:409–15. 21. Brackertz D, Mitchell GF, Vadas MA, Mackay IR. Studies on IRAK-4 KINASE ACTIVITY IN JOINT INFLAMMATION 22. 23. 24. 25. 26. 27. 28. 29. 30. antigen-induced arthritis in mice. III. Cell and serum transfer experiments. J Immunol 1977;118:1645–8. Suzuki N, Suzuki S, Millar DG, Unno M, Hara H, Calzascia T, et al. A critical role for the innate immune signaling molecule IRAK-4 in T cell activation. Science 2006;311:1927–32. Braley-Mullen H. Antigen requirements for induction of B-memory cells: studies with dinitrophenyl coupled to T-dependent and T-independent carriers. J Exp Med 1978;147:1824–31. Walsh DA, Pearson CI. Angiogenesis in the pathogenesis of inflammatory joint and lung diseases. Arthritis Res 2001;3:147–53. Buma P, Groenenberg M, Rijken PF, van den Berg WB, Joosten L, Peters H. Quantitation of the changes in vascularity during arthritis in the knee joint of a mouse with a digital image analysis system. Anat Rec 2001;262:420–8. Coxon A, Bolon B, Estrada J, Kaufman S, Scully S, Rattan A, et al. Inhibition of interleukin-1 but not tumor necrosis factor suppresses neovascularization in rat models of corneal angiogenesis and adjuvant arthritis. Arthritis Rheum 2002;46:2604–12. Joosten LA, Koenders MI, Smeets RL, Heuvelmans-Jacobs M, Helsen MM, Takeda K, et al. Toll-like receptor 2 pathway drives streptococcal cell wall-induced joint inflammation: critical role of myeloid differentiation factor 88. J Immunol 2003;171:6145–53. Lee EK, Kang SM, Paik DJ, Kim JM, Youn J. Essential roles of Toll-like receptor-4 signaling in arthritis induced by type II collagen antibody and LPS. Int Immunol 2005;17:325–33. Roelofs MF, Boelens WC, Joosten LA, Abdollahi-Roodsaz S, Geurts J, Wunderink LU, et al. Identification of small heat shock protein B8 (HSP22) as a novel TLR4 ligand and potential involvement in the pathogenesis of rheumatoid arthritis. J Immunol 2006;176:7021–7. Ji H, Pettit A, Ohmura K, Ortiz-Lopez A, Duchatelle V, Degott C, 1671 31. 32. 33. 34. 35. 36. 37. 38. et al. Critical roles for interleukin 1 and tumor necrosis factor ␣ in antibody-induced arthritis. J Exp Med 2002;196:77–85. 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. Gasse P, Mary C, Guenon I, Noulin N, Charron S, SchnyderCandrian S, et al. IL-1R1/MyD88 signaling and the inflammasome are essential in pulmonary inflammation and fibrosis in mice. J Clin Invest 2007;117:3786–99. Van den Berg WB. Joint inflammation and cartilage destruction may occur uncoupled. Springer Semin Immunopathol 1998;20: 149–64. Van Lent PL, Nabbe K, Blom AB, Holthuysen AE, Sloetjes A, van de Putte LB, et al. Role of activatory Fc␥RI and Fc␥RIII and inhibitory Fc␥RII in inflammation and cartilage destruction during experimental antigen-induced arthritis. Am J Pathol 2001;159: 2309–20. Suzuki N, Suzuki S, Duncan GS, Millar DG, Wada T, Mirtsos C, et al. Severe impairment of interleukin-1 and Toll-like receptor signalling in mice lacking IRAK-4. Nature 2002;416:750–6. Lye E, Dhanji S, Calzascia T, Elford AR, Ohashi PS. IRAK-4 kinase activity is required for IRAK-4-dependent innate and adaptive immune responses. Eur J Immunol 2008;38:870–6. Schnare M, Barton GM, Holt AC, Takeda K, Akira S, Medzhitov R. Toll-like receptors control activation of adaptive immune responses. Nat Immunol 2001;2:947–50. Ku CL, von Bernuth H, Picard C, Zhang SY, Chang HH, Yang K, et al. Selective predisposition to bacterial infections in IRAK4–deficient children: IRAK-4–dependent TLRs are otherwise redundant in protective immunity. J Exp Med 2007;204:2407–22.