To investigate the role of Crk-associated substrate lymphocyte type (Cas-L), a downstream signaling molecule of β1 integrins, in the pathophysiology of rheumatoid arthritis (RA).
To investigate the role of Crk-associated substrate lymphocyte type (Cas-L), a downstream signaling molecule of β1 integrins, in the pathophysiology of rheumatoid arthritis (RA).
We analyzed human T lymphotropic virus type I (HTLV-I) tax transgenic mice as well as samples from human RA patients. Splenocytes from tax transgenic mice were cultured on mouse endothelial cell–covered Transwell inserts, and cells migrating through the endothelial monolayer were counted. Biochemical studies were performed to analyze the protein expression and tyrosine phosphorylation of Cas-L. Immunohistochemical analysis was performed to detect Cas-L–positive cells that had infiltrated into the joints.
Migratory activity of splenocytes from tax transgenic mice with arthritis (ATg) was much higher than that of tax transgenic mice without arthritis (NTg) and littermate control mice. The expression of Cas-L protein and its tyrosine phosphorylation were increased in ATg mice compared with NTg and control mice, and this was accompanied by enhanced autophosphorylation of Fyn and Lck. Immunohistochemical analysis demonstrated a large number of Cas-L–positive lymphocytes migrating into the affected joints. Furthermore, in human RA, Cas-L–positive lymphocytes were shown to infiltrate to the inflammatory lesions.
These results strongly suggest that Cas-L plays an important role in the pathophysiology of RA.
There is accumulating evidence suggesting that β1 integrin–dependent cell activation and migration pathways are critical points of intervention in several inflammatory and autoimmune diseases, such as rheumatoid arthritis (RA) (1). Indeed, in RA patients, increased expression of β1 integrins and their ligands on the surface of synovial fluid mononuclear cells (SFMCs) (2–5) and synovium cells (6–9) has been reported, suggesting that these integrins play an important role in triggering and maintaining the inflammatory response in the disease.
It has been shown that β1 integrins exert a variety of biologic functions, such as cytokine production, proliferation, cell differentiation, cell survival, apoptosis, and cell migration, as well as cell adhesion through the interaction of their ligands (extracellular matrix and vascular cell adhesion molecule 1) (10–17). For understanding the molecular mechanisms of these numerous biologic effects, it is particularly important to analyze cell signaling through the β1 integrins. In this regard, it has been demonstrated that interaction of fibronectin (FN) and its receptor very late activation antigen 5 (VLA-5) or the CS-1 domain of FN and VLA-4 can induce costimulatory signals to the CD3/T cell receptor (TCR) pathway (18–20). It has been reported that tyrosine phosphorylation of cellular proteins is an early obligatory event in cell activation and signal transduction. Subsequent studies showed that phospholipase Cγ, focal adhesion kinase (FAK), paxillin, Fyn, Lck, extracellular signal–related kinase 1/2, and pp105 are phosphorylated on their tyrosine residues upon engagement of β1 integrins in T cells (21–24). The 105-kd protein pp105 has been shown to associate with FAK that is autophosphorylated and activated upon the engagement of β1 integrins (25). Isolation of complementary DNA encoding pp105 has revealed that this protein belongs to the Crk-associated substrate (Cas) family, and hence it is designated as Cas-L (Cas lymphocyte type) (25, 26).
Pp105/Cas-L is a docking protein that is heavily tyrosine phosphorylated by FAK and Src-family kinases upon the engagement of β1 integrins in T cells (27, 28). Transfection of Cas-L into Jurkat cells markedly enhances cell motility (29) and interleukin-2 (IL-2) production (30) upon engagement of β1 integrins, through its tyrosine phosphorylation. These results clearly indicate the involvement of Cas-L in β1 integrin–mediated costimulation of signal transduction and cell migration.
To investigate the role of Cas-L in the pathophysiology of RA in vivo, we studied cell migration and protein tyrosine phosphorylation in a mouse model of RA (using human T lymphotropic virus type I [HTLV-I] tax transgenic mice) as well as in RA patients. Notably, Iwakura et al (31, 32) and Saijo et al (33) have demonstrated that HTLV-I transgenic mice that carry the env-pX gene develop inflammatory arthropathy in high incidence (31–33). In addition to RA-like disease, tax transgenic mice are reported to develop a disease resembling Sjögren's syndrome in humans (34). Furthermore, tax transgenic mice produce autoantibodies against IgG (rheumatoid factor), type II collagen, and heat-shock proteins, all of which are accompanied by IgG hypergammaglobulinemia. Therefore, these mice are regarded as an animal model of human RA (31–33).
Herein we show the possible involvement of Cas-L in the pathogenesis of RA-like disease in tax transgenic mice, mediated by its overexpression and hyperphosphorylation, resulting in remarkably enhanced cell motility of lymphocytes. Furthermore, we demonstrate the infiltration of Cas-L–positive, CD3-positive T cells into the affected lesion in humans with RA. These findings strongly suggest that Cas-L plays a role in the pathophysiology of RA.
Antiphosphotyrosine antibody (4G10; IgG2b) was purchased from Upstate Biotechnology (Lake Placid, NY). Monoclonal antibodies (mAb) against pp125FAK, Pyk2, p130Cas, Lck, and Fyn were obtained from Transduction Laboratories (Lexington, KY). Polyclonal antibodies (pAb) specific for Cas-L and p130Cas have been described previously (28). Briefly, rabbits were immunized with glutathione S-transferase (GST) fusion protein of Cas-L (amino acids 419–524) and p130Cas. Those portions of each protein were selected because of relatively low homology of amino acid sequences with each other. After antiserum against anti–Cas-L and anti-p130 were absorbed with GST–p130Cas and GST–Cas-L, respectively, specific antibodies for Cas-L and p130Cas were affinity purified with GST–Cas-L and GST–p130Cas. Basically, immunoprecipitates with anti–Cas-L pAb do not contain p130Cas, and those with anti-p130Cas do not contain Cas-L (28). Affinity-purified rabbit anti-mouse antibody was purchased from Jackson Laboratories (West Grove, PA). All chemicals were purchased from Sigma (St. Louis, MO) unless otherwise stated. Fetal calf serum was purchased from Tissue Culture Biologicals (Tulare, CA). Human serum FN was obtained from Life Technologies (Grand Island, NY). All radioisotopes were purchased from Amersham Biosciences (Uppsala, Sweden).
Transgenic mice with the HTLV-I env-pX region of the viral genome with its own long terminal repeat promoter were used (31–33). Original transgenic mice with (C3H/He × C57BL/6J)F1 background were backcrossed with BALB/c mice. Female mice of 8–11 generations after backcrossing were used for experiments. These mice were kept under specific pathogen–free conditions in a clean room at the Animal Research Center, Institute of Medical Science, University of Tokyo. Paw joints were examined macroscopically for swelling and redness once weekly. Transgenic mice were classified as either those that developed arthritis (ATg), i.e., with obvious swelling of the joints, or those without the disease (NTg).
Synovial tissue specimens were obtained from the affected joints of patients with RA. Synovial tissue specimens from osteoarthritis (OA) patients were used as controls. Samples were obtained during surgical procedures, e.g., total joint arthroplasty and arthroscopic synovectomy. RA or OA was diagnosed according to American College of Rheumatology criteria (35, 36). Written informed consent was obtained from each participant. For histopathologic evaluation, synovial tissue was fixed with 4% paraformaldehyde at 4°C immediately after resection, and paraffin sections stained with hematoxylin and eosin (H&E) were observed under a light microscope.
In the transendothelial migration assay of murine splenocytes, murine endothelial cell line SVEC4-10EE2 (37) (obtained from American Type Culture Collection, Rockville, MD) was cultured on Transwell inserts with a 3.0-μm pore size. Confluent monolayers were established on the inserts by incubating 1 × 105 SVEC4-10EE2 cells overnight at 37°C in 5% CO2/95% air. The Transwell chambers were inserted into wells filled with 600 μl of 0.6% bovine serum albumin (BSA)–RPMI 1640 medium, and splenocytes resuspended at 1 × 106/ml were added to the upper chamber in a final volume of 100 μl. After incubation at 37°C for an appropriate time period, cells that had migrated into the bottom chambers were harvested and counted using a FACSCalibur (Becton Dickinson, Franklin Lakes, NJ).
For immunoprecipitation, cells were lysed in 1% Nonidet P40 lysis buffer. Cellular lysates were immunoprecipitated with mAb-conjugated protein A–Sepharose beads (Amersham Biosciences). The washed beads were boiled for 5 minutes in the presence of 2% sodium dodecyl sulfate (SDS) and 0.1M dithiothreitol. The supernatants of boiled samples were loaded onto SDS–7.5% polyacrylamide gels, electrotransferred onto polyvinylidene difluoride membranes (Millipore, Bedford, MA), and blocked for 2 hours with 5% skim milk or 5% BSA in Tris buffered saline containing 0.05% Tween 20. After blocking, membranes were incubated with appropriate concentrations of first antibody, washed, and incubated with horseradish peroxidase–conjugated anti-mouse IgG antibody. The membranes were developed by the enhanced chemiluminescence system (Amersham Biosciences).
Total RNA from mouse spleen was prepared using an Isogen kit according to the instructions of the manufacturer (Nippon Gene, Tokyo, Japan). Approximately 10 μg of RNA was separated by electrophoresis through 1% agarose gel containing MOPS-formaldehyde, transferred to nylon membrane filters (Amersham Biosciences) in 10× saline–sodium citrate (SSC), and filters were dried and irradiated on an ultraviolet transilluminator. Filters were prehybridized for 1 hour at 68°C in PerfectHyb hybridization solution (Toyobo, Tokyo, Japan). Hybridization with 32P-labeled probes was carried out at 68°C for 1 hour in hybridization solution. Blots were washed in 2× SSC containing 0.1% SDS at 68°C and exposed for autoradiography. Densitometry analysis was performed on a Macintosh computer using the public domain NIH Image program (available online at http://rsb.info.nih.gov/nih-image/).
Mice were killed by CO2 asphyxiation, and paw joints were fixed in 10% phosphate buffered formalin for further processing and analysis. After fixation, the joints were decalcified in 14% EDTA for 7–10 days, dehydrated in graded ethanol, and embedded in paraffin. Serial 3 mm–thick sections were cut. Immunohistochemical staining was performed on paraffin sections that had been deparaffinized in xylene and rehydrated in graded ethanol and phosphate buffered saline. After blocking with normal goat serum, the sections were incubated with primary antibody for 30 minutes at room temperature. After washing, they were treated with 0.3% H2O2 solution for 5 minutes at room temperature to quench endogenous peroxidase activity, then incubated with biotinylated secondary antibody and streptavidin-linked peroxidase reagent for 30 minutes at room temperature, followed by incubation with substrate-chromogen solution. Finally, specimens were counterstained with Mayer's hematoxylin. As a negative control, the primary mAb was replaced with rabbit IgG.
To determine the percent positivity of various surface markers among the total cell population, splenocytes were stained with fluorescein isothiocyanate (FITC)–conjugated anti-CD3 mAb, phycoerythrin-conjugated anti-CD4 or anti-CD8 mAb, or purified anti-CD29, anti-CD49d, anti-CD11a, and anti-CD44 mAb, followed by FITC-conjugated secondary antibody. Flow cytometric analysis was then performed.
Statistical analysis of the transendothelial migration assay and flow cytometry results was performed by Student's 2-tailed t-test.
HTLV-I tax transgenic mice are well characterized as a murine model for human RA, since they are known to develop autoimmune polyarthritis resembling the human disease (31–33). Our preliminary finding that HTLV-I–transformed T cell lines expressed Cas-L at high levels led us to further investigate these mice to determine the clinical relevance of Cas-L in the pathophysiology of arthritis. Figure 1 shows transendothelial migratory activity of splenocytes from transgenic mice and littermate controls. Among 4-week-old mice, spontaneous migratory activity through the endothelium was much higher in tax transgenic mice without arthritis (NTg) than in littermate controls. Moreover, 14-week-old mice with arthritis (ATg) showed greater migratory activity than those without arthritis (P < 0.05).
Since Cas-L appears to play a role in cell migration, we examined the expression and tyrosine phosphorylation of Cas-L protein in those mice. Figure 2 shows the protein amount and degree of tyrosine phosphorylation of Cas-L precipitated from lysates of splenocytes. Among 4-week-old mice, both the amount and the tyrosine phosphorylation of Cas-L were higher in NTg mice than in littermate controls. Among 14-week-old mice, both the amount and the tyrosine phosphorylation of Cas-L were highest in ATg mice. Moreover, anti–Cas-L pAb unexpectedly precipitated a tyrosine-phosphorylated band with a putative molecular weight of 78 kd, in addition to the 105-kd band. Interestingly, the extent of tyrosine phosphorylation of both proteins seemed to parallel the migratory activity of splenocytes. To define the specificity of pp78, we compared immunoprecipitates, using specific antibodies against Cas-L and p130Cas. The specific pAb for Cas-L and antiphosphotyrosine mAb precipitated pp78, whereas specific pAb for p130Cas and control rabbit IgG failed to do so (results not shown).
It has been reported that Cas-L and p130Cas are cleaved into smaller fragments by caspases in the process of apoptosis (38). To examine the possibility that this 78-kd protein is the degradation product of Cas-L, we next prepared lysates directly from tissue kept frozen in liquid nitrogen immediately after resection, without isolating the individual cells. Figure 3A shows the protein amount and degree of tyrosine phosphorylation of Cas-L precipitated from tissue lysates from lymphoid organs such as spleen, lymph node, and thymus. Both the amount and the tyrosine phosphorylation of Cas-L in these organs were highest in tax transgenic mice with arthritis and next-highest in tax transgenic mice without arthritis compared with littermate controls. It should be noted that the pp78 protein band was not observed.
We next evaluated the level of Cas-L mRNA in spleens of tax transgenic mice. We found that Cas-L mRNA was elevated in spleens of tax transgenic mice compared with littermate controls (Figure 3B). Of greater interest, spleens from ATg mice showed remarkably elevated levels of Cas-L mRNA compared with those of NTg and control mice, indicating a close parallel with the levels of Cas-L protein and its tyrosine phosphorylation, and also with the migratory behavior of splenocytes. Moreover, the above findings support the notion that the pp78 band was the degradation product of Cas-L.
Since several signaling molecules are involved in β1 integrin–mediated signaling events, we next determined the expression and phosphorylation of tyrosine kinases in spleen and lymph node. We have recently demonstrated that there are two pathways in the tyrosine phosphorylation of Cas-L; one is mediated through FAK and Pyk2, and the other through the Src-family tyrosine kinases, Fyn and Lck (25, 27, 28). As shown in Figure 4, protein expression of FAK and Pyk2 in ATg mice was increased, but the phosphorylation level of these tyrosine kinases was not changed except for a slight increase of Pyk2 in lymph nodes. Since alterations in the tyrosine phosphorylation of FAK and Pyk2 were unexpectedly subtle, we next evaluated the tyrosine phosphorylation of Fyn and Lck. We found that the level of tyrosine phosphorylation as well as protein amounts of those kinases were remarkably elevated in tax transgenic mice, especially those with arthritis.
To determine whether the increased protein amount of Lck and Fyn was due to differences in cell population among the groups of mice, we next evaluated the surface markers of splenocytes from tax transgenic and littermate control mice. As shown in Table 1, the percent positivity of various surface markers was not altered significantly among those mice (P > 0.05). In particular, expression of β1 integrins was similar among mice in the different groups, indicating that enhanced motile behavior of splenocytes from tax transgenic mice was not caused by a change in surface expression of β1 integrins.
|Surface antigen||Control mice (n = 3)||NTg mice (n = 3)||ATg mice (n = 3)|
|CD3/CD4||24.1 ± 1.7||27.2 ± 7.3||27.1 ± 0.3|
|CD3/CD8||9.3 ± 1.8||9.7 ± 2.9||7.4 ± 0.5|
|CD29||20.4 ± 2.4||22.2 ± 2.6||19.9 ± 0.7|
|CD49d||22.8 ± 1.0||23.0 ± 0.6||18.6 ± 0.4|
|CD11a||96.6 ± 1.2||97.5 ± 1.6||96.5 ± 0.6|
|CD44||64.7 ± 5.1||71.1 ± 0.8||62.0 ± 7.2|
Based on the finding that the protein level of Cas-L was generally elevated in tax transgenic mice with arthritis, we investigated the in situ expression of Cas-L in inflamed joints and synovial tissue from those mice, by immunohistochemical analysis. As shown in Figures 5A and B, Cas-L–positive cells of lymphoid origin, as well as those of leukocyte origin, infiltrated into the affected synovium in tax transgenic mice with arthritis. It should be noted that the number of Cas-L–positive cells in these animals was increased not only in synovial tissue, but also in lymph node and spleen (results not shown).
Finally, we analyzed human tissue samples obtained from patients with RA or OA. In the RA samples, the cluster of infiltrated lymphocytes observed on H&E-stained specimens appeared to correspond to the stained regions in the specimens treated with anti–Cas-L pAb. To determine the cell populations, we examined two serial sections: one treated with anti–Cas-L antibody and the other with anti-CD3 antibody (Figure 6A). Findings in these sections clearly indicated that infiltrated mononuclear cells were Cas-L and CD3 double-positive. In H&E-stained specimens from OA patients, infiltrated lymphocytes were not found, so Cas-L–positive cells were not observed (Figure 6B). These results suggest that Cas-L is also involved in human RA.
In this study, we showed that expression of Cas-L was elevated in HTLV-I tax transgenic mice, the mouse model of autoimmune polyarthritis. More importantly, Cas-L appears to play a role in the pathophysiology of human RA. It is known that HTLV-I is an etiologic agent for a unique type of hematologic malignancy, adult T cell leukemia (39, 40). HTLV-I has been shown to be associated with a number of diseases other than malignancies arising from the lymphoid system, affecting various organs. These chronic inflammatory diseases may involve the nervous system (HTLV-I–associated myelopathy/tropical spastic paraparesis [41, 42]), the eyes (HTLV-I–associated uveitis ), the salivary glands (Sjögren's syndrome ), or the joints (HTLV-I–associated arthropathy [45, 46]).
It should be noted that splenocytes as well as lymphoid organs from NTg mice expressed more Cas-L protein than those from control mice, but less than those from ATg mice. This positive correlation between incidence of disease and amount of Cas-L may be of particular interest, although the exact pathologic role of enhanced Cas-L expression remains to be elucidated.
In this study, a 78-kd band, in addition to pp105, was observed in splenocytes. Both the protein amount and the level of tyrosine phosphorylation of pp78 were highest in tax transgenic mice with arthritis. We previously showed that Cas-L and p130Cas had >50% homology of amino acids, and mAb against p130Cas (Transduction Laboratories) also reacts with Cas-L (25). Furthermore, p130Cas has multiple forms (Cas-A, Cas-B, and Cas-C), with molecular sizes varying from 115 kd to 135 kd (47). Pp105Cas-L also has multiple forms (doublet bands) with molecular sizes of approximately pp105 and pp120 (23, 38). It has been shown that these differences in molecular size may be due to the extent of the protein phosphorylation (38). Therefore, we developed polyclonal antibodies specific for Cas-L and p130Cas, by affinity purification following cross-absorption of immunized serum using GST fusion proteins. The immunoreactivity of pp78 suggests that it may be derived from Cas-L.
The immediate lysis of various tissues clearly showed the predominance of pp105Cas-L rather than the pp78 band, strongly suggesting that pp78 was derived from protein degradation of Cas-L. Alternatively, it is possible that the p78 band might be a nonspecific band that is preferentially observed in tax transgenic mice. As reported by Law et al, overexpression of Cas-L/human enhancer of filamentation 1 resulted in its degradation mediated by activation of caspases (38). It appears that the degradation of Cas-L was accelerated in tax transgenic mice with arthritis. Ishino et al reported that Efs/Sin, a member of the Cas-family proteins, has an alternative spliced form that lacks the SH3 domain (48). Therefore, another possibility is that pp78 might be derived from alternative splicing of Cas-L gene product. Further studies are needed to define the nature of pp78 in tax transgenic mice.
It has been reported that Tax can activate transcription of numerous host cellular genes associated with lymphocyte proliferation, e.g., protooncogenes and cytokines (49) including IL-2 receptor (IL-2R) (50, 51), c-Fos (52, 53), transforming growth factor β (54), gp34/OX-40/CD134 (55), IL-6 (56), tumor necrosis factor α (TNFα) (57), IL-8 (58), as well as the integrated provirus of HTLV-I (59), indicating its close association with T cell immortalization and transformation. It has been shown that this transactivation of host genes is mediated through the interaction of p40tax with several cellular transcription factors, such as cAMP response element binding protein/activating transcription factor, nuclear factor κB, and serum response factor (40, 60–62). Serum levels of a variety of inflammatory cytokines, including TNFα, IL-1β, and IL-6, are elevated in tax transgenic mice (32, 33). These cytokines (or chemokines) have been regarded as playing a role in executing the inflammatory responses in the affected joints of the mice.
We cannot exclude the possibility that cofactors or accelerators other than Cas-L may contribute to the development of autoimmune arthritis. Nonetheless, our speculation is that Cas-L functions, at least in the accelerated phase of the disease, together with those cytokines or chemokines to maintain or promote the inflammatory response. It should be noted that induction of Cas-L by Tax might be secondarily regulated by those inflammatory cytokines, possibly explaining the difference in the amount of Cas-L between NTg mice and ATg mice. Further detailed studies on the transcriptional regulation of Cas-L in mice and humans are indicated. Moreover, our recently established results (30) indicate that Cas-L may augment IL-2 production by the engagement of β1 integrins and CD3/TCR. Consequently, it is possible that Cas-L may cooperate with Tax to form the autocrine loop of IL-2 and IL-2R, resulting in clonal expansion of infiltrating lymphocytes.
Of particular importance is the finding that not only the amount of Cas-L, but also its extent of tyrosine phosphorylation, differs between littermate controls, tax transgenic mice without arthritis, and those with arthritis. Biochemical studies have demonstrated that Src-family tyrosine kinases such as Fyn and Lck, as well as FAK, phosphorylate tyrosine residues on Cas-L (27), and this phosphorylation is induced by crosslinking of β1 integrins and/or CD3/TCR (28). The results of the present study strongly suggest that tyrosine phosphorylation of Cas-L was mainly induced through the activation of the Src-family tyrosine kinases Fyn and Lck in tax transgenic mice with arthritis. With regard to the cause of increased phosphorylation of Cas-L, we believe elevations in protein level may partly reflect the apparently increased levels in the immunoblot with antiphosphotyrosine mAb (4G10).
Several lines of evidence indicate that tyrosine phosphorylation as well as expression of Cas-family proteins may be essential for the migratory behavior of the cells. Using Jurkat T cells, we previously showed that gene transfer of Cas-L promotes a cell migratory response on the ligand for β1 integrins, FN, and/or mAb against CD3 (29). The strikingly enhanced motility of splenocytes from tax transgenic mice with arthritis may be attributed to the elevated expression of Cas-L in those cells. The enhancement of motility was dependent on tyrosine phosphorylation of Cas-L, since the phosphorylation-deficient mutant of Cas-L lacking the SH3 domain failed to enhance cell migration (29). The exact mechanism by which Src-family protein tyrosine kinases and/or FAK and Pyk2 are autophosphorylated is another important question that remains to be elucidated.
Evaluation of clinical relevance (such as correlation between Cas-L positivity and disease activity in RA patients) is a very important purpose for determining the exact pathophysiologic role of Cas-L in RA. In breast cancer, elevated expression of p130Cas, a homolog of Cas-L that is mainly expressed in adherent cell types, is correlated with aggressive disease or poor prognosis (63). In another study examining the role of Cas-L in the pathophysiology of human RA, we have found enhanced transendothelial migration of peripheral blood lymphocytes (PBLs) and SFMCs from RA patients (64). Lymphocytes from RA synovial fluid showed more enhanced migratory capacity than did PBLs from healthy subjects and RA patients. In the present study, we have shown that Cas-L–expressing lymphocytes actually infiltrate the affected joints of RA patients. Notably, we have recently demonstrated that tyrosine phosphorylation of Cas-L protein was markedly enhanced in SFMCs from patients with RA (64). Further study on the clinical relevance of Cas-family proteins in autoimmune and inflammatory diseases is needed.
Together with previous results, the present findings strongly suggest that Cas-L protein plays a crucial role in the pathophysiology of human RA as well as in polyarthritis in tax transgenic mice, especially by harboring T cells with extensively motile behavior. We are currently further addressing the significance of Cas-L in the pathogenesis of RA, in studies using a series of Cas-L mutants. This approach should be beneficial for investigation of possible future therapeutic applications of Cas-L gene products.
The authors thank Ms Fumiki Nojima for excellent secretarial assistance, and Mr. Haruo Onoda, Ms Mamiko Sato, and Ms Nozomi Yusa for invaluable technical support.