Induction of Notch signaling by tumor necrosis factor in ... - Nature

Oncogene (2003) 22, 7796–7803

& 2003 Nature Publishing Group All rights reserved 0950-9232/03 $25.00 www.nature.com/onc

Induction of Notch signaling by tumor necrosis factor in rheumatoid synovial fibroblasts Kiichiro Ando1,2, Satoshi Kanazawa1, Toshifumi Tetsuka1, Shusuke Ohta1,2, Xu Jiang1,2, Toyohiro Tada3, Masaaki Kobayashi2, Nobuo Matsui2 and Takashi Okamoto*,1 1

Department of Molecular Genetics, Nagoya City University Medical School, 1 Kawasumi, Mizuho-cho, Mizuho-ku, Nagoya 467-8601, Japan; 2Department of Orthopedics, Nagoya City University Medical School, Nagoya 467-8601, Japan; 3Department of Pathology, Nagoya City University School of Nursing, Nagoya 467-8601, Japan

Rheumatoid arthritis (RA) is characterized by progressive inflammation associated with abberrant proliferation of synoviocytes. In order to explore the characteristics of rheumatoid synovial fibroblasts (RSF), we performed the comparative gene expression profile analysis between RSF and normal synovial fibroblasts (NSF) upon tumor necrosis factor (TNF) stimulation. As an initial screening for the genes preferentially induced by TNF in RSF compared with NSF, we have adopted a cDNA array containing well-defined sets of genes responsible for cell growth, cell fate determination, and cellular invasiveness. Differentially expressed genes of interest were confirmed using real-time RT–PCR. We found that TNF induced the expression of Notch-1, Notch-4, and Jagged-2 in RSF. The expression of these proteins was detected in the RA synovial tissues. The nucleus of RA synoviocytes showed strong staining with anti-Notch-1 and Notch-4 antibody. TNF induced the nuclear translocation of Notch intracellular domain in RSF, indicating the elicitation of the Notch signaling. Notch-1, Notch-4, and Jagged-2 proteins were also detected in the developing synovium of mouse embryo. Thus, RSF may have re-acquired the primordial phenotype, accounting for the hyperproliferation and aggressive invasiveness, exhibiting tumor-like phenotype. Oncogene (2003) 22, 7796–7803. doi:10.1038/sj.onc.1206965 Keywords: rheumatoid arthritis; transformation; notch; Jagged; gene expression; TNF

Introduction Rheumatoid arthritis (RA) is a systemic inflammatory disease associated with progressive synovitis characterized by hyperplasia of the synovial cells with invasive characteristics often associated with destruction of cartilage and bone (Feldmann, 2001). The synoviocyte proliferation in RA mimics tumor cell growth. In the *Correspondence: T Okamoto; E-mail: [email protected] Received 18 February 2003; revised 9 July 2003; accepted 10 July 2003

rheumatoid synovium, proliferation of fibroblast-like synoviocytes (type B synoviocytes) is found adjacent to sites of joint destruction. This hyperproliferative nature of rheumatoid synovial fibroblasts (RSF) is considered as the final effector responsible for the pannus formation by mediating inflammatory responses and joint destruction (Alvaro-Gracia et al., 1991). The destruction of cartilage and other components of connective tissue associated with RA involves the actions of proteolytic enzymes including the matrix metalloproteinases (MMP) (Constantin et al., 2002), which also resembles tumor infiltration. As RSF exhibits the anchorage-independent growth and the loss of contact inhibition in vitro (Fassbender and Simmling-Annefeld, 1983), and activation of specific oncogenes such as ras and myc that promote cell cycling control and metabolic upregulation of MMPs (Muller-Ladner et al., 1995), these transformed-like (also described as ‘mesenchymal transformation’) cellular characteristics of RSF have long been proposed (Fassbender et al., 1980). The invasive nature of RSF has been demonstrated in SCID mice in which the transplanted RSF, not NSF or synovial fibroblasts from osteoarthritis (OA), invaded into human normal cartilage tissue that were coimplanted (Muller-Ladner et al., 1996). Interestingly, somatic mutations of the p53 tumor suppressor gene have been demonstrated in RA tissues and RSF (Yamanishi et al., 2002). In addition, we have recently found the upregulation of the stromal cell-derived factor 1 and platelet-derived growth factor receptor a, which are considered responsible for the formation of primordial joint tissue during the embryonic development, are overexpressed in RSF, even in the absence of inflammatory cytokines and growth factors (Watanabe et al., 2002). Thus, we performed comparative gene expression profile analysis in order to identify genes selectively induced in RSF by TNF and attempted to define the cellular characteristics of RSF in order to further decipher the molecular mechanism underlying RA. We show here that TNF selectively stimulates the cell fatedetermining Notch and Jagged genes in RSF, together with other genes involved in cell proliferation and tissue destruction. These findings indicate that RSF may have

Notch signaling in rheumatoid synovial fibroblasts K Ando et al

7797

acquired the revertant cellular phenotype during the disease process of RA, accounting for the tumor-like characteristics.

B1, and Jagged-2 were elevated by TNF only in RSF. The elevated mRNA levels of Notch-1, Jagged-2, and cyclin B1 were downregulated after 24 h of stimulation, suggesting the presence of a negative feedback mechanism.

Results Nuclear translocation of Notch 1 in RSF induced by TNF Identification of genes selectively induced by TNF in RSF As an initial screening for the genes preferentially induced by TNF in RSF, we have adopted a cDNA array containing six well-defined sets of representative genes. These categories include cell-cycle/growth regulators, oncogenes/tumor suppressors/apoptosis regulators, DNA damage response/repair and recombination/ cell fate and development regulators, cell adhesion and motility/angiogenesis, invasion regulators, and growth factors/receptors. The gene expression level of each gene was compared between RSF (n ¼ 5) and NSF (n ¼ 5) before and after 6 and 24 h of TNF stimulation at 200 pg/ml, the concentration often detected in rheumatoid synovial fluid (Yoshida et al., 1999a, b). Gene expression of both interleukin-1b (IL-1b) and interleukin-6 (IL-6) was augmented directly by TNF (Figure 1), as previously noted (Feldmann et al., 1996), and the extents of induction in RSF and NSF were similar. Thus, signal transduction pathways downstream of the TNF signaling leading to NF-kB activation (Okamoto et al., 1997; Ghosh and Karin, 2002) were comparable in both RSF and NSF. In contrast, the extents of induction of MMP-11, MMP-17, Wee1, cyclin B1, cyclin-dependent kinase 1 (Cdk1), Notch-1, and Jagged-2 were greater in RSF than in NSF. Selective induction of these MMPs and cell-cycle regulators (Wee1, cyclin B1, and Cdk1) may explain the invasive and proliferative characteristics of RSF upon inflammatory stimuli, although no common transcriptional signaling pathway among these genes is known. We were particularly interested in the concomitant induction of Notch-1, Notch-4, and Jagged-2 in RSF upon TNF stimulation as these genes encode cell fate determinant proteins involved in the bidirectional differentiation of adjacent primordial cells (Artavanis-Tsakonas et al., 1999). Notch-2, Notch-3, and human homologue of Drosophila Delta (delta-like protein (dlk)), the ligand for Notch-4, were not induced in both RSF and NSF. Confirmation for the differential induction of genes by real-time RT–PCR In order to confirm the results of the initial screening with cDNA array, we carried out real-time RT–PCR to determine more quantitatively the mRNA levels of IL-6, MMP-11, cyclin B1, Notch-1, Notch-4, and Jagged-2 before and after TNF stimulation (0, 6, and 24 h). We prepared mRNA samples of RSF (n ¼ 5) and NSF (n ¼ 5) at various time points and measured the mRNA levels in triplicates for each determination (Figure 2). The mRNA level of IL-6 was elevated in both RSF and NSF with a similar time course. Interestingly, the mRNA levels of Notch-1, Notch-4, MMP-11, cyclin

As TNF induced Notch-1 and its ligand Jagged-2 in RSF, we examined if the Notch signaling is elicited by TNF stimulation. RSF and NSF were stimulated with TNF, and the intracellular localization of Notch-1 intracellular domain (NICD) was examined by immunostaining. We found the nuclear translocation of Notch-1 NICD, a hallmark of the Notch signaling (Schroeter et al., 1998; Struhl and Greenwald, 1999), only in TNF-stimulated RSF (Figure 3). These results suggested that in response to TNF stimulation RSF expressed both Notch-1 and Jagged-2 proteins, which then interacted with each other between adjacent cells and elicited the signaling. Detection of notch-1, Notch-4, and Jagged-2 proteins in primary synovium from RA patients We then examined the expression of Jagged-2 and Notch proteins in fresh synovial tissues of RA and OA patients. Tissues were immunostained with antibodies against Notch-1 (NICD), Notch-4, and Jagged-2. In RA tissues, hyperproliferative synovial tissues were clearly stained by Notch-1 and Jagged-2 antibodies, which were predominantly found in the cytoplasm and the cytoplasmic membrane (Figure 4). Notch-4 was also stained in RA synovium. It is notable that some of the RA synovial cells showed the nuclear staining of Notch-1 and Notch-4. We also examined the expression of these proteins in fresh synovial tissues obtained from OA exhibiting proliferative characteristics. As shown in Figure 4, although there is slight cytoplasmic staining of Notch-1 in some synovial lining cells, neither the staining of Jagged-2 nor the nuclear staining of Notch proteins was detected. Normal synovial tissues, lacking the synovial proliferation, were not stained by these antibodies (data not shown). Expression of Notch-1, Notch-4, and Jagged-2 in joint tissues of mouse embryo Since Notch proteins and their ligands are known as developmental determinants for cell fate programming and subsequent differentiation in various body systems (Gaiano et al.,. 2000; Uyttendaele et al., 2001; Wolfer et al., 2001), we have examined the expression of these proteins in the developing joints of late embryonic mice (E15). Developing joints consist of somite cells committed to differentiate into the synovium and the cartilage, which were separated by the primordial joint space. As demonstrated in Figure 5a, Notch-1, Notch-4, and Jagged-2 proteins were detected in clusters of cells surrounding the joint space. Interestingly, a few cells with nuclear staining of Notch-1 and Notch-4 were Oncogene


7798

Figure 1 Comparative gene expression profile analysis of RSF vs NSF. (a) Spot images of cDNA array of representative cases (RSF and NSF) at various times after TNF stimulation. Excerpts of differentially expressed genes are shown. The positions of each gene are marked by oval circles. RSF and NSF were cultured with or without TNF (200 pg/ml). The mRNA was purified from each cell culture harvested at 0, 6, and 24 h after TNF stimulation, cDNA probe was synthesized and hybridized with a cDNA array membrane (Human Cancer cDNA Expression Array; Clontech) containing defined 588 cancer-related genes. (b) Quantitation of the hybridization signal intensities for indicated genes in (a). Intensity of gene expression was measured by a phosphoimager (BAS2500) and evaluated using a computer software (ArrayGauge, version 1.12). The quantitation of gene expression level was performed and standardized based on the average levels of nine housekeeping genes (partly shown in the lower part of the panel). For the sake of accuracy, genes with a hybridization signal greater than 10-fold of background level (nonspecific signal of hybridization membrane) were selected

Oncogene


7799

Figure 2 Time course of the mRNA induction of IL-6, MMP-11, cyclin B1, Notch-1, Notch-4, and Jagged-2 in RSF and NSF. The fold induction of each gene was measured by real-time detection RT–PCR with mRNA samples prepared from synoviocyte cultures of five RA patients (RSF; ) and five control subjects (NSF; J). The quantitative evaluation was made at the exponential phase of each amplification plot. Triplicate determination was made for each mRNA sample. The gene expression levels were normalized by b-actin for each mRNA preparation and the fold increase of individual gene was calculated by comparison with the result obtained without TNF stimulation. Data are presented as mean7s.d. of synoviocyte cultures from five RA patients (RSF) and five control subjects (NSF). *Po0.05; n.s., not significant

detected near the edge of the newly forming joint (indicated by arrows in Figure 5a). In more developed joints of new born mice (D1), expression of these proteins appeared to be restricted in the synovium (Figure 5b). No cells with typical nuclear staining of NICD were detected. These findings suggest that the Notch–Jagged system appears to play a role during the joint development.

Discussion The transformed-like phenotype of rheumatoid synoviocytes has been implicated in the hyperproliferative and invasive characteristics of synovial tissues of RA. It is not known, however, whether these cytological features are due to the intrinsic nature or rather an acquired phenotype endowed by the actions of cytokines (Mukaida et al., 1990) such as TNF. More specifically, it is yet to be clarified whether the cellular response of RSF to the inflammatory environment, as represented by the accumulation of TNF in the affected

Figure 3 Nuclear translocation of NICD in RSF after TNF stimulation. RSF and NSF were immunostained with anti-Notch-1 C-terminus polyclonal antibody (C-20) before and after 12 h of TNF stimulation and examined by fluorescent microscopy. Green, Notch-1 intracellular domain (detected by fluorescein isothiocyanate-conjugated rabbit anti-goat IgG as secondary antibody); blue, nuclear staining with DAPI

joint, is different from that of NSF. In order to answer these questions, we compared the gene expression profile of RSF with NSF upon stimulation with TNF. We found that the extents of gene induction of Notch-1, Notch-4, Jagged-2, cyclin B1, and MMP-11 were significantly higher in RSF than in NSF. Among these genes, MMP-11 and cyclin B1 are known to be responsible for tissue destruction and cell proliferation, respectively. Moreover, the selective induction of Notch families and the ligand of Notch-1 and Jagged-2 has prompted us to examine the expression of these proteins in the fresh synovial tissues of RA. We found that Notch-1, Notch-4, and Jagged-2 proteins were expressed in the synovial tissues of RA patients. We also detected expression of these proteins in the developing synovial and cartilage tissues of embryonic mice (Figure 5). In more developed joints of new born mice, expression of these proteins was restricted in the synovium, raising a possibility that Notch signaling pathway might control the differentiation and development of joints. In addition, it was recently reported that RSF expressed high levels of both Wnt and Frizzled, a receptor–ligand pair implicated in both limb bud and bone marrow stem cell development (Sen et al., 2002). In supporting our observations, Nakazawa et al. (2001) reported that Notch-1 was overexpressed in rheumatoid synoviocytes and TNF induced nuclear translation of NICD in RSF. Notch genes encode single-pass transmembrane receptors that elicit extracellular signals responsible for cell fate determination during the metazoan development (Matsuno et al., 1995). The large transmembrane Oncogene


7800

Figure 4 Immunohistochemical detection of Notch-1, Notch-4, and Jagged-2 in synovial tissues from patients with RA and OA. Expressions of Notch-1, Notch-4, and Jagged-2 were detected by immunohistochemical staining with each polyclonal antibody and examined by light microscopy. Fresh synovial tissues were obtained from RA and OA patients at total arthro-replacement. Representative results are shown. Arrows indicate the synovial cells with nuclear staining of Notch-1 and Notch-4. Hematoxylin–eosin was used as a counter stain

receptors encoded by Notch genes interact with membrane-bound ligands encoded by the Delta and Jagged (Serrate) genes at the extracellular surface of cells. The signal induced by this ligand binding leads to proteolysis of Notch, generation and nuclear translocation of NICD, and regulation of target gene expression. Genes homologous to members of the Notch signaling pathway have been cloned from numerous vertebrate organisms and many have been shown to be essential for normal embryonic development. The aberration of Notch signaling in human diseases is reported in T-lymphoblastic leukemia, in which normal function of Notch-1 is impaired by translocation (Ellisen et al., 1991), and in Alagille syndrome, an inherited disease involving affected organogenesis in skeletal system, liver, and heart, which is ascribable to mutations in Jagged-1, Notch-1 ligand, gene (Oda et al., 1997). Jagged-2 has previously been shown to activate Notch-1 in mammalian cells. Jiang et al. (1998) found that mice in which Jagged-2 gene is replaced by its mutant, defective for the interaction with Notch-1, exhibited syndactyly (digit fusions) of the fore- and hindlimbs, and died at birth because of defects in craniofacial morphogenesis, implicating that Jagged-2/ Oncogene

Notch signaling is indispensable for the development of the joint. Importantly, the similar phenotype, syndactyly, has been observed in mice lacking the IKKa subunit of IkB kinase complex (Hu et al., 1999), a crucial signal transducing effector kinase for the TNF signaling. Therefore, it is likely that the TNF-mediated NF-kB activation through IKKa is involved in expression of Jagged-2, either directly or indirectly, in the skeletal system development. These findings indicate that the Notch signaling plays an important role in the joint development. Although the target genes are yet to be elucidated, it is possible that the TNF-mediated NF-kB activation induces Jagged-2 and elicits the Notch signaling in the adjacent synovial cells of the developing joints, thus promoting differentiation of primordial synovial cells. We have recently reported the evidence suggesting that the genetic phenotype of synovial fibroblasts is ontogenically reversed, at least in part, during the rheumatoid inflammatory process and discussed that this revertant phenotype of RSF may be responsible for the selfperpetuating character of RA (Watanabe et al., 2002). These findings suggest the involvement of Notch signaling in RA pathophysiology as well as in the joint


7801

Figure 5 Immunohistochemical detection of Notch-1, Notch-4, and Jagged-2 in joints of mouse embryo and new-born mice. Expressions of Notch-1, Notch-4, and Jagged-2 protein were detected by immunohistochemical staining with each polyclonal antibody in the developing joints of mice at E15 (a) and D1 (b). Arrows indicate the synovial cells with nuclear staining of Notch proteins. Hematoxylin–eosin (HE) was used as a counter stain, syn, synovium; crtg, cartilage

development. However, several questions remain unanswered, such as how TNF selectively induces gene expression of Notch and Jagged genes in RSF and what are the targets of the NICD-mediated transcriptional activation. Since selective induction of genes in RSF was evident after 6 h of TNF stimulation and IL-6 induction was equally observed both in RSF and NSF, this effect cannot be ascribed solely to the activation of NF-kB. More extensive analysis of promoters for these genes and further gene expression profile analyses using updated cDNA arrays with much larger numbers of displayed genes should uncover these issues. Materials and methods Synoviocyte cultures RSF cultures were isolated from fresh synovial tissue biopsy samples from five RA patients at total arthro-replacement (Suzuki et al., 2000). Diagnosis of RA was made according to

the clinical criteria of the American College of Rheumatology. There were five female patients, between 37 and 63 years old, and all had active RA. NSF cultures were isolated from fresh and noninflammatory synovial tissue biopsy samples from five traumatic patients. Informed consent was obtained from each patient in conformity with the requirements of the ethics committee of the Nagoya City University Medical School. These tissue biopsy samples were minced into small pieces and treated with 1 mg/ml collagenase/dispase (Roche Diagnostics GmbH, Mannheim, Germany) for 10–20 min at 371C. The cells were cultured in F-12 (HAM) (Invitrogen Co., CA, USA) supplemented with 10% fetal calf serum, 100 U/ml of penicillin, 100 mg/ml of streptomycin, and 0.5 mm mercaptoethanol. The culture medium was changed every 3–5 days and nonadherent lymphoid cells were removed during the initial passages. Under these culture conditions, there was no difference in the growth property of RSF and NSF. All the experiments described here were conducted using synoviocyte cultures during the third to eighth passages. To characterize the cytological phenotype, the cells were stained with mouse monoclonal antibodies against human HLA-DR, von Oncogene


7802 Willebrand factor, desmin, smooth muscle actin, CD1a, CD68, and 5B5. Only 5B5 was positive for RSF and NSF, indicating their fibroblast-like phenotype. mRNA preparation Total cellular RNA was prepared from each RSF and NSF culture with TRIzol reagent (Invitrogen) as described previously. Purification of mRNA was carried out using the Oligotex-dT30 super RNA Purification Kit (Takara Shuzo, Kyoto, Japan) according to the manufacturer’s instructions. The mRNA samples were digested with DNase, ethanol precipitated, and further purified through gel filtration on a desalting column (Chroma-spin 200, Clontech Laboratories, Palo Alto, CA, USA). Gene expression profile analysis Gene expression profile was examined using the Atlas cDNA expression array (Human Cancer cDNA Expression Array, Clontech) containing cDNA fragments of cancer-related 588 human genes with known functions in duplicate. The genes plotted on the membrane are categorized into six functional groups associated with cancer pathophysiology such as cell proliferation, cell fate determination, and invasion (listed at http://www.clontech.com/products/catalog02/HTML/1047. shtml). The cDNA probe was synthesized by reverse transcription of 1 mg of each mRNA preparation, prepared from synovial cell culture using Superscript II reverse transcriptase (200 U/ml, Invitrogen) in the presence of a[33P]dATP (Amersham Pharmacia, Little Chalfont, UK) according to the supplier’s protocol. The labeled cDNA probe was then added to 10 ml of ExpressHyb hybridization solution (Clontech) with 100 mg/ml sheared salmon testis DNA (Sigma Chemical Co., St Louis, MO, USA) to reach a final probe concentration of 1 106 cpm/ml. The probe was freshly applied to the cDNA array membrane for hybridization at 681C in a roller bottle with continuous agitation for 16 h. After hybridization, membranes were washed for 20 min at 681C with 2 SSC/1% SDS four times and for 20 min at 681C with 0.5 SSC/0.5% SDS two times. The damp membranes were immediately covered with plastic wrap and exposed to Kodak BioMax MS X-ray film with an intensifying screen at 701C. Phosphorimaging analysis and quantitation To quantify the levels of gene expression, the membranes were exposed to a storage Imaging plate (BAS-SR-2040; Fuji Photo Film Co., Ltd, Tokyo, Japan) for 4 h and scanned using a phosphoimager (BAS2500, Fuji Photo Film). The quantitated gene expression levels were analysed using a computer software (Arraygauge version 1.12, Fuji Photo Film). The relative mRNA expression level of each gene was standardized based on the average expression level of nine housekeeping genes including ubiquitin, GAPDH, and b-actin. The average hybridization signal of each gene in RSF and NSF cultures was compared. RT–PCR reactions and analyses In order to carry out real-time quantitative RT–PCR, tTtb DNA polymerase was used to reverse transcribe and amplify 25 ng of total RNA in a single tube assay using the PerkinElmer TaqMan EZ RT–PCR kit (Perkin-Elmer Applied Biosystems, Foster City, CA, USA). Specific primers to IL-6, cyclinB1, MMP-11, Notch-1, Notch-4, Jagged-2, and b-actin were designed using Primer express software (version 1.0, Oncogene

Perkin-Elmer Applied Biosystems) and synthesized. Triplicate samples were reverse transcribed for 30 min at 601C and then subjected to 40 rounds of amplification for 15 s at 951C and 1 min at 601C using the ABI Prism 5700 sequence detection system (Perkin-Elmer Applied Biosystems) according to the manufacturer’s protocol. Sequence-specific amplification was detected as an increased fluorescent signal during the amplification cycle. Quantitative analysis of each genespecific mRNA level was based on comparison of the fluorescent intensity from a standard curve of the mRNA level of relevant genes. Amplification of b-actin mRNA was performed with all the samples to control for variation in RNA amounts. The expression levels of test genes were subsequently normalized by the mRNA level of b-actin in the same mRNA sample. Immunofluorescence staining Semiconfluent RSF and NSF cultures on Lab-Tek tissue culture chamber slides (Nunc, Inc., Naperville, IL, USA) were fixed with 0.5% glutaraldehyde in phosphate-buffered 0.1 m PIPES–KOH (pH 6.9) for 10 min at room temperature, rinsed twice with PBS, and incubated with PBS containing 0.5% Triton X-100 for 20 min at room temperature. They were subsequently incubated with the anti-Notch-1 goat polyclonal antibody (Santa Cruz Biotechnology, CA, USA) for 90 min at 371C, rinsed three times with PBS containing 0.05% Triton X100, and incubated with the secondary antibody, fluorescent isothiocyanate-conjugated rabbit anti-goat IgG antibody (Jackson ImmunoResearch Laboratories) for 90 min at 371C. The slides were rinsed three times with PBS and mounted with PBS-buffered glycerol for fluorescent microscopic examination. 40 ,6-Diamidino-2-phenylindole dihydrochloride (DAPI) was used for the nuclear counter staining. Primary and secondary antibodies were diluted at 1 : 400 and 1 : 50, respectively, in PBS containing 3% bovine serum albumin (Kajino et al., 2000). Immunohistochemistry Fresh synovial tissues were obtained from patients with RA and OA upon the total arthro-replacement operation under the informed consent and agreement of each patient. Developing synovial tissues of ICR mice (Charles-River Japan, Hino, Japan) were prepared from hindfeet under anesthesia at embryonic day 15 (E15) and 1 day after birth (D1). These tissues were fixed with Carnoy-methanol fixative for 3 h, dehydrated in ethanol, embedded in paraffin, and sectioned serially at 3 mm thickness for immunohistochemistry and hematoxylin–eosin staining. After deparaffinization and washing with PBS, the tissue sections were treated with 0.3% (v/v) hydrogen peroxide in methanol for 30 min to inactivate endogenous peroxidase. They were then subjected for immunohistochemical detection of Notch and Jagged proteins using polyclonal antibodies against Notch-1 (C-20), Notch-4 (H-225), and Jagged-2 (H-143) (Santa Cruz Biotechnology). Expression of these proteins was detected using the labeled streptavidin–biotin–peroxidase complex method. Control sections were treated with either nonimmune goat IgG or nonimmune rabbit IgG, instead of primary antibodies. The probed tissue sections were rinsed, incubated sequentially with secondary antibody (biotinylated rabbit anti-goat IgG or biotinylated goat anti-rabbit IgG antibody), and reacted with streptavidin–biotin–peroxidase complex (Histofine SAB-PO kit; Nichirei, Tokyo, Japan). After washing with PBS, the peroxidase reaction was performed by incubating the sections in 0.02% (w/v) 3,30 -diaminobenzidine tetrahydrochloride


7803 (Sigma Chemical Co.) solution containing 0.003% (v/v) hydrogen peroxide and 10 mm sodium azide. Hematoxylin was used for the counter staining as well as for the histological analysis.

Abbreviations RA, rheumatoid arthritis; OA, osteoarthritis; TNF, tumor necrosis factor; IL-1b, interleukin-1b; NF-kB, nuclear factor

kB; NICD, Notch-1 intracellular domain; RBP-Jk, recombinant signal binding protein Jk. Acknowledgements This work was supported by grants-in-aid from the Ministry of Health, Labor and Welfare (Grant No. H14-immunity-008), Ministry of Education, Culture, Sports, Science and Technology (Grant No. 14028052), and Japan Human Sciences Foundation (Grant No. SA14721).

References Alvaro-Gracia JM, Zvaifler NJ, Brown CB, Kaushansky K and Firestein GS. (1991). J. Immunol., 146, 3365–3371. Artavanis-Tsakonas S, Matsuno K and Fortini ME. (1995). Science, 268, 225–232. Artavanis-Tsakonas S, Rand MD and Lake RJ. (1999). Science, 284, 770–776. Constantin A, Lauwers-Cances V, Navaux F, Abbal M, van Meerwijk J, Mazieres B, Cambon-Thomsen A and Cantagrel A. (2002). Arthritis Rheum., 46, 1754–1762. Ellisen LW, Bird J, West DC, Soreng AL, Reynolds TC, Smith SD and Sklar J. (1991). Cell, 66, 649–661. Fassbender HG and Simmling-Annefeld M. (1983). J. Pathol., 139, 399–406. Fassbender HG, Simmling-Annefeld M and Stofft E. (1980). Verh. Dtsch. Ges. Pathol., 64, 193–212. Feldmann M. (2001). J. Clin. Invest., 107, 1353. Feldmann M, Brennan FM and Maini RN. (1996). Cell, 85, 307–310. Gaiano N, Nye JS and Fishell G. (2000). Neuron, 26, 395–404. Ghosh S and Karin M. (2002). Cell, 109 (Suppl.), S81–S96. Hu Y, Baud V, Delhase M, Zhang P, Deerinck T, Ellisman M, Johnson R and Karin M. (1999). Science, 284, 316–320. Jiang R, Lan Y, Chapman HD, Shawber C, Norton CR, Serreze DV, Weinmaster G and Gridley T. (1998). Genes Dev., 12, 1046–1057. Kajino S, Suganuma M, Teranishi F, Takahashi N, Tetsuka T, Ohara H, Itoh M and Okamoto T. (2000). Oncogene, 19, 2233–2239. Matsuno K, Diederich RJ, Go MJ, Blaumueller CM and Artavanis-Tsakonas S. (1995). Development, 121, 2633–2644. Mukaida N, Mahe Y and Matsushima K. (1990). J. Biol. Chem., 265, 21128–21133. Muller-Ladner U, Kriegsmann J, Franklin BN, Matsumoto S, Geiler T, Gay RE and Gay S. (1996). Am. J. Pathol., 149, 1607–1615.

Muller-Ladner U, Kriegsmann J, Gay RE and Gay S. (1995). Rheum. Dis. Clin. N. Am., 21, 675–690. Nakazawa M, Ishii H, Aono H, Takai M, Honda T, Aratani S, Fukamizu A, Nakamura H, Yoshino S, Kobata T, Nishioka K and Nakajima T. (2001). Arthritis Rheum., 44, 1545–1554. Oda T, Elkahloun AG, Pike BL, Okajima K, Krantz ID, Genin A, Piccoli DA, Meltzer PS, Spinner NB, Collins FS and Chandrasekharappa SC. (1997). Nat. Genet., 16, 235–242. Okamoto T, Sakurada S, Yang JP and Merin JP. (1997). Curr. Top. Cell Regul., 35, 149–161. Schroeter EH, Kisslinger JA and Kopan R. (1998). Nature, 393, 382–386. Sen M, Reifert J, Lauterbach K, Wolf V, Rubin JS, Corr M and Carson DA. (2002). Arthritis Rheum., 46, 2867–2877. Struhl G and Greenwald I. (1999). Nature, 398, 522–525. Suzuki M, Tetsuka T, Yoshida S, Watanabe N, Kobayashi M, Matsui N and Okamoto T. (2000). FEES Lett., 465, 23–27. Uyttendaele H, Ho J, Rossant J and Kitajewski J. (2001). Proc. Natl. Acad. Sci. USA, 98, 5643–5648. Watanabe N, Ando K, Yoshida S, Inuzuka S, Kobayashi M, Matsui N and Okamoto T. (2002). Biochem. Biophys. Res. Commun., 294, 1121–1219. Wolfer A, Bakker T, Wilson A, Nicolas M, Ioannidis V, Littman DR, Lee PP, Wilson CB, Held W, MacDonald HR and Radtke F. (2001). Nat. Immunol., 2, 235–241. Yamanishi Y, Boyle DL, Rosengren S, Green DR, Zvaifler NJ and Firestein GS. (2002). Proc. Natl. Acad. Sci. USA, 99, 10025–10030. Yoshida S, Kato T, Sakurada S, Kurono C, Yang JP, Matsui N, Soji T and Okamoto T. (1999a). Int. J. Immunol., 11, 151–158. Yoshida S, Katoh T, Tetsuka T, Uno K, Matsui N and Okamoto T. (1999b). J. Immunol., 163, 351–358.

Oncogene