Adaptor Molecule Crk Is Required for Sustained Phosphorylation of Grb2

Adaptor Molecule Crk Is Required for Sustained Phosphorylation of Grb2-Associated Binder 1 and Hepatocyte Growth Factor–Induced Cell Motility of Human Synovial Sarcoma Cell Lines

Takuya Watanabe,1,2 Masumi Tsuda,1 Yoshinori Makino,1 Shin Ichihara,1 Hirofumi Sawa,1 Akio Minami,2 Naoki Mochizuki,3 Kazuo Nagashima,1 and Shinya Tanaka1

1Laboratory of Molecular and Cellular Pathology and 2Department of Orthopaedic Surgery, Hokkaido University Graduate School of Medicine, Sapporo, Japan and 3National Cardiovascular Center Research Institute, Osaka, Japan

Abstract Importantly, the elimination of Crk in these cells induced Activation of the c-Met receptor tyrosine kinase through the disorganization of actin cytoskeleton and complete its ligand, hepatocyte growth factor (HGF), promotes abolishment of HGF-mediated Rac1 activation and cell mitogenic, motogenic, and morphogenic cellular motility. Time-lapse microscopic analysis revealed the responses. Aberrant HGF/c-Met signaling has been significant attenuation in scattering of Crk knockdown strongly implicated in tumor cell invasion and cells following HGF treatment. Furthermore, the metastasis. Both HGF and its receptor c-Met have been depletion of Crk remarkably inhibited the tumor shown to be overexpressed in human synovial sarcoma, formation and its invasive growth in vivo. These results which often metastasizes to the lung; however, little is suggest that the sustained phosphorylation of Gab1 known about HGF-mediated biological effects in this through Crk in response to HGF contributes to the sarcoma. Here, we provide evidence that Crk adaptor prominent activation of Rac1 leading to enhanced cell protein is required for the sustained phosphorylation of motility, scattering, and cell invasion, which may support c-Met-docking protein Grb2-associated binder 1 (Gab1) the crucial role of Crk in the aggressiveness of human in response to HGF, leading to the enhanced cell motility synovial sarcoma. (Mol Cancer Res 2006;4(7):499–510) of human synovial sarcoma cell lines SYO-1, HS-SY-II, and Fuji. HGF stimulation induced the sustained Introduction phosphorylation on Y307 of Gab1 where Crk was Cell motility is essential for a variety of biological and recruited. Crk knockdown by RNA interference disturbed physiologic processes, including embryonic development, this HGF-induced tyrosine phosphorylation of Gab1. wound healing, angiogenesis, and organogenesis, whereas an By mutational analysis, we identified that Src homology aberrant motility is observed in several human malignant 2 domain of Crk is indispensable for the induction of tumors that could be linked to promoting tumor cell invasion the phosphorylation on multiple Tyr-X-X-Pro motifs and metastasis correlating with poor prognosis. Several growth containing Y307 in Gab1. HGF remarkably stimulated factors have been shown to stimulate cell motility, and one of cell motility and scattering of synovial sarcoma cell the most potent inducers is hepatocyte growth factor (HGF). lines, consistent with the prominent activation of Rac1, HGF is a multifunctional factor that influences cell prolifera- extreme filopodia formation, and membrane ruffling. tion, migration, adhesion, and organ and tissue development through the activation of its receptor tyrosine kinase c-Met (1). Above all, HGF has also been known as scattering factor (2). Received 8/19/05; revised 3/16/06; accepted 5/1/06. Grant support: Ministry of Education, Science, Culture, Sports and Ministry of HGF triggers autophosphorylation on Y1234 and Y1235 in the Health, Labor, and Welfare grants-in-aid, Yasuda Medical Research Foundation, catalytic domain of c-Met (3). Moreover, two tyrosine residues and the Suhara Memorial Foundation. (Y1349 and Y1356) of c-Met are also autophosphorylated on The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in HGF stimulation (4), which provide a multisubstrate binding accordance with 18 U.S.C. Section 1734 solely to indicate this fact. site for several Src homology 2 (SH2) domain–containing Note: Supplementary data for this article are available at Molecular Cancer Research Online (http://mcr.aacrjournals.org/). cytoplasmic effectors. In addition to the several SH2 domain– Current address for M. Tsuda: Department of Pathophysiology, Hokkaido containing molecules, Grb2-associated binder 1 (Gab1) also University Graduate School of Medicine, Sapporo, Japan. interacts with the activated c-Met directly through its Met- Current address for Y. Makino and H. Sawa: Hokkaido University Research Center for Zoonosis Control, Sapporo, Japan. binding sequences (5) or indirectly via the adaptor protein Grb2 Current address for K. Nagashima: Sapporo Higashi-Tokushukai Hospital, (6, 7). Sapporo, Japan. Gab1 is a member of docking protein and plays a role Requests for reprints: Shinya Tanaka, Laboratory of Molecular and Cellular Pathology, Hokkaido University Graduate School of Medicine, N 15, W 7, Kita-ku, to integrate the signals of downstream of c-Met. Gab1 is phos- Sapporo 060-8638, Japan. Phone: 81-11-706-7806; Fax: 81-11-706-5902. phorylated on multiple tyrosine phosphorylation sites on c-Met E-mail: [email protected] Copyright D 2006 American Association for Cancer Research. kinase stimulation that are required for recruiting a variety of doi:10.1158/1541-7786.MCR-05-0141 SH2 domain–containing transducers, including the p85 subunit

of phosphatidylinositol 3-kinase, phospholipase Cg1, tyrosine SH2 domain binds several tyrosine-phosphorylated proteins, phosphatase SHP-2, SHIP, and adaptor protein Shc, CrkII, and including p130Cas, paxillin, Cbl, and Gab1. The SH3 domain Crk-like (CrkL; refs. 6, 8-14). Especially, Gab1 possesses six of Crk has been shown to interact with guanine nucleotide Tyr-X-X-Pro (YXXP) motifs those are potential binding sites for exchange factors (17), including Dock180, which activates adaptor protein CrkII and CrkL. Rac1 and regulates cell motility (18, 19), whereas the Crk has been shown to mediate a variety of biological activation of Rap1 through the another Crk SH3 domain responses involved in proliferation, cell differentiation, and target C3G (20, 21) is involved in cell adhesion (22). migration. Crk was originally identified as an avian sarcoma Recent studies show the involvement of Crk in tumorigen- virus CT10 encoding protein, v-Crk (15). Mammalian esis and progression. Human CrkI can induce transformation in homologue CrkII is composed of a SH2 domain and two rat fibroblasts (16), and Crk has been known to be overex- SH3 domains, and its alternative splicing form CrkI contains pressed in various human cancers, including brain tumors and the SH2 domain and only one SH3 domain (16). The Crk lung cancers (23-25). However, the precise role for Crk in human cancer is still under investigation. Abnormal HGF/c-Met signaling has been reported to be strongly implicated in the development and progression of a variety of human tumors. The amplification and point mutations in kinase domain of c-Met have been found in gastric carcinomas and gliomas (26, 27) and in hereditary and sporadic papillary renal cell carcinomas (28), respectively. Moreover, the overexpression of HGF and/or c-Met is associated with a poor prognosis of breast cancer (29) and multiple myeloma (30). Transgenic mice overexpressing HGF develop tumors with metastatic lesions (31). Synovial sarcoma is a malignant soft tissue tumor characterized by the oncogenic chimeric transcript SYT-SSX arising from chromosomal translocation t(X,18) (32, 33). Synovial sarcoma occurs primarily in the extremities of young adults, especially in the periarticular region, and often metastasizes to the lung. Interestingly, both HGF and c-Met have been found to be overexpressed in synovial sarcoma (34-36). The autocrine mechanism of HGF is thought to be involved in tumor growth and an epithelial-mesenchymal transition in this sarcoma (37, 38). However, the little is known about HGF/c-Met-mediated biological effects on cell motility and on metastasis. In this study, we show that the sustained tyrosine phosphorylation of Gab1 via coupling of Crk promotes HGF- dependent cell motility in synovial sarcoma cell lines. The interaction of Gab1 with Crk is necessary for HGF-mediated activation of Rac1 downstream of c-Met. Inhibition of Crk by RNA interference disturbed both the phosphorylation of Gab1 and the activation of Rac1, leading to the disorganized actin cytoskeleton, suppression of cell motility, and inhibition of scattering. In addition, the elimination of Crk remarkably suppressed the tumor formation with the invasive growth in vivo. These findings indicate a crucial role of Crk in HGF/c- Met-mediated activation of Rac1 that could be linked to the enhanced cell motility of human synovial sarcoma cell lines FIGURE 1. HGF-induced tyrosine phosphorylation of c-Met and through the sustained complexes of Crk to Gab1. Gab1 in human synovial sarcoma cell lines. A. Expression levels of c-Met and Gab1 in human synovial sarcoma cell lines SYO-1, HS-SY-II, and Fuji were examined by immunoblotting. B. HGF-induced tyrosine phosphorylation of c-Met and Gab1. SYO-1, HS-SY-II, and Fuji cells Results were treated with or without 50 ng/mL HGF for 30 minutes. The cell lysates were immunoprecipitated withor withoutanti-Met or anti-Gab1 HGF-Induced Tyrosine Phosphorylation of c-Met and antibody and subsequently probed withtheindicated antibody. C. HGF- Gab1 in Human Synovial Sarcoma Cell Lines induced interaction of c-Met and Gab1. HGF treatment was done as Human synovial sarcoma has been shown to overexpress described in B. The cell lysate was immunoprecipitated using anti-Gab1 antibody, and c-Met bound to Gab1 was detected by immunoblotting. both HGF and its receptor c-Met; therefore, we initially D. HGF-independent tyrosine phosphorylation of p130Cas and paxillin. examined the expression levels of c-Met protein in three Withor withoutHGF treatment, p130 Cas and paxillin were immunoprecipitated by eachantibody and probed withanti-phosphotyrosine independent human synovial sarcoma cell lines, SYO-1, antibody. HS-SY-II, and Fuji. c-Met was found in all cell lines,

Mol Cancer Res 2006;4(7). July 2006 Downloaded from mcr.aacrjournals.org on September 25, 2021. © 2006 American Association for Cancer Research. Requirement of Crk on Gab1 Phosphorylation and Cell Motility 501

particularly higher level in SYO-1 cells (Fig. 1A), consistent interfering RNA technique. These cells were confirmed to with the previous studies that the epithelial components in contain substantial amount of endogenous Crk (data not shown) biphasic synovial sarcoma highly express c-Met (39). as reported in various human tumors. We succeeded in Following HGF stimulation, the tyrosine-phosphorylated form establishing several vector control and Crk knockdown cell of c-Met became detectable in all cell lines (Fig. 1B, top), lines in three cell lines. In Crk knockdown cells, the expression particularly stronger in SYO-1 cells. Phosphorylation site- levels of CrkI and CrkII proteins were significantly suppressed specific antibody identified the autophosphorylation on Y1234 (Fig. 3A), whereas the expression levels of CrkL as well as and Y1235 in the catalytic domain of c-Met (Fig. 1B) that actin were constant (Fig. 3A). have been shown to be essential for the kinase activity of c-Met. Moreover, the phosphorylation on Y1349 of c-Met was Suppression of HGF-Dependent Sustained Phosphoryla- also induced on HGF stimulation (Fig. 1B), which is the direct tion of Gab1 in CrkKnockdownSynovial Sarcoma Cell binding site for Gab1. Gab1 was expressed in all three cell Lines lines (Fig. 1A), and HGF-dependent association of c-Met and To examine the involvement of Crk on the stabilized Gab1 was detected by immunoprecipitation analysis (Fig. 1C). phosphorylation of Gab1, we analyzed the time-dependent In addition, the tyrosine-phosphorylated forms of Gab1 became also detectable on HGF stimulation; indeed, Y307 residue, which is one of six Crk-binding consensus YXXP motifs, was identified as one of the phosphorylation sites (Fig. 1B). The results of immunoblotting of HGF-dependent phosphorylation of Met and Gab in all three synovial sarcoma cell lines by using total cell lysates were shown in Supplementary Data 1A. In contrast, HGF stimulation did not significantly enhance the tyrosine phosphorylation of p130Cas and paxillin, except for slightly enhanced phosphorylation of p130Cas in HS-SY-II cells (Fig. 1D). The tyrosine- phosphorylated levels of Crk and CrkL were constant in all cell lines, and the association of Crk and its SH3 domain targets, such as C3G and Dock180, were confirmed by immunoprecipitation (data not shown).

Crk-SH2 Domain–Dependent Regulation of Tyrosine Phosphorylation of Gab1 As expected from the result of HGF-induced phosphorylation on Y307 of Gab1 (Fig. 1B), HGF treatment induced the recruitment of Crk to Gab1 (Fig. 2A). To clarify the precise binding mechanism between Gab1 and Crk, the deletion mutant of Gab1 and the point mutants of CrkII were transiently overexpressed in 293T cells (Fig. 2B). We initially found that the exogenously expressed CrkII with functional SH2 domain, such as CrkII-WT and CrkII-W169L, induced the prominent phosphorylation within several YXXP motifs of Gab1 (Fig. 2B). In contrast, CrkII lacking the function of SH2 as CrkII-R38V completely failed to induce this phosphorylation (Fig. 2B). In immunoprecipitation analysis, we found that anti-Gab1 antibody efficiently coprecipitated wild-type CrkII but not CrkII-R38V; furthermore, HGF stimulation enhanced FIGURE 2. Crk-SH2 domain – dependent tyrosine phosphorylation the association of Crk and Gab1 (Fig. 2C). The endogenous and association of Gab1. A. HGF-induced recruitment of Crk to expression levels of c-Met were previously confirmed to be phosphorylated Y307 of Gab1. SYO-1, HS-SY-II, and Fuji cells were almost equal between 293T and SYO-1 cells (data not shown). treated with or without 50 ng/mL HGF for 30 minutes. The cell lysates were immunoprecipitated withanti-Crk antibody, and Gab1 bound to Thus, the mutational analyses of Gab1 and Crk indicate the Crk was detected by immunoblotting using anti-phosphorylated Gab1 crucial role of Crk on the stabilization of HGF-induced (Y307) antibody. B. Crk-SH2 domain – dependent tyrosine phosphorylation of Gab1. The 293T cells were transiently cotransfected with the tyrosine phosphorylation of Gab1. expression plasmids for Gab1 and Crk (top). The cell lysates were immunoprecipitated withanti-Gab1 antibody and subsequently probed withanti-phosphotyrosine antibody. Gab1-Wt, Gab1 wild-type; del, Establishment of CrkKnockdown Human Synovial Gab1-del-YXXP mutant; Crk-Wt, CrkII wild-type; RV, CrkII-R38V mutant; WL, CrkII-W169L mutant. C. Crk-SH2 domain – dependent Sarcoma Cell Lines association to Gab1. The 293T cells were transiently cotransfected To clarify the significance of Crk-mediated phosphorylation with the expression plasmids for Gab1 and Crk. Thirty-six hours after transfection, 50 ng/mL HGF treatment was done for 30 minutes. The of Gab1 in the biology of synovial sarcoma, we established Crk association of Gab1 and Crk was examined by immunoprecipitation knockdown SYO-1, HS-SY-II, and Fuji cell lines by small using anti-Gab1 antibody.

alteration on phosphorylation of c-Met and Gab1 following HGF stimulation in wild-type SYO-1 cells and its Crk knockdown cells. In parental SYO-1 cells, c-Met protein was significantly degraded within 3 hours after HGF stimulation (Fig. 3B). The kinase activity of c-Met, characterized by the phosphorylation on Y1234 and Y1235 residues in its catalytic domain, reached a maximum level at 20 minutes after HGF stimulation and gradually decreased (Fig. 3B). HGF treatment also triggered remarkable phosphorylation on Y307 of Gab1 (Fig. 3B) that was sustained >2 hours after HGF stimulation, contrary to rapid decline of c-Met phosphorylation. The association of Crk to Gab1 was also sustained over 3 hours after HGF stimulation (Fig. 3B). On the other hand, we detected no significant alteration on the endogenous expression levels of c-Met and its down- regulations following HGF-stimulation between wild-type SYO-1 cells and its Crk knockdown cells (Fig. 3B). However, in Crk knockdown SYO-1 cells, tyrosine phosphorylation on Y307 of Gab1 started to decrease after reaching maximum level at 10 minutes after HGF stimulation, and this phosphorylation was undetectable within 1 hour (Fig. 3B and C). These data suggest that the elimination of Crk caused the reduced phosphorylation of Gab1 and its rapid down-modulation (Fig. 3C). Indeed, in Crk-depleted SYO-1 cells, the phosphorylation levels of Gab1 at 30 and 60 minutes following HGF stimulation reduced to 0.75 and 0.6, respectively, relative to that of its wild-type cells (Fig. 3C). In established Crk knockdown cells, the expression levels of Crk-SH3 targets, such as Dock180 and C3G, were also analyzed. SYO-1 and HS-SY-II cell lines possess almost equal SH3 targets between control and Crk knockdown cells, whereas in Fuji cells the levels of Dock180 and C3G seemed to be weakened by Crk elimination (Fig. 3D).

Analysis of Rac1 Activity in CrkKnockdownCells Rac1 is one of the Rho family small GTPases that has been reported to regulate the cytoskeletal reorganization downstream of Crk/Dock180 association. Therefore, to identify the effect of Crk in HGF-induced cytoskeletal reorganization, we examined the activity of Rac1 in wild-type and Crk knockdown synovial sarcoma cell lines. HGF induced prominent activation of Rac1 in all parental cells. Especially, the biphasic activation of Rac1 within 90 minutes after HGF stimulation was shown with peak at 10 and 30 minutes in SYO-1 cells and at 5 and 30 minutes in FIGURE 3. Decreased phosphorylation of Gab1 on HGF in HS-SY-II cells (Fig. 4, top left and middle). This biphasic established Crk knockdown synovial sarcoma cells. A. Analysis of the expression levels of CrkI and CrkII in Crk knockdown synovial activation of Rac1 was remarkable in Fuji cells that exhibited sarcoma cell lines SYO-I, HS-SY-II, and Fuji. WT, wild-type sarcoma 4.0-fold increase at 20 minutes after HGF stimulation, cells; Vector, vector-control transfectants; Crki, Crk knockdown cells. reduction at 45 minutes, and a secondary 3.7-fold activation The endogenous levels of CrkL and actin were also examined by immunoblotting. B. HGF-induced tyrosine phosphorylation of c-Met and at 60 minutes (Fig. 4, bottom left). In contrast, HGF-mediated Gab1 in SYO-1 cells. Wild-type of SYO-1 and its Crk knockdown cells slight increase of Rac1 activity was detected in Crk knockdown were treated with50 ng/mL HGF for theindicated time. The SYO-1 cells (Fig. 4, top right). Furthermore, even mild phosphorylation of c-Met (Y1234/Y1235) and Gab1 (Y307) in whole- cell lysate was analyzed by immunoblotting. The association of Crk decrease of Rac1 activity was observed with HGF treatment in and phosphorylated Y307 of Gab1 was examined by immunoprecip- Crk knockdown HS-SY-II and Fuji cells (Fig. 4, middle and itation using anti-Crk antibody followed by immunoblotting withanti- bottom right). phosphorylated Gab1 (Y307) specific antibody. C. In wild-type of SYO-1 and its Crk knockdown cells, the phosphorylation levels at To identify the exact subcellular localization and its extent Y307 of Gab1 were compared on an identical membrane at 0, 10, 20, of HGF-induced Rac1 activation in individual cells, we 30, and 60 minutes after HGF stimulation. D. Expression levels of target molecules of Crk-SH3 domain, Dock180 and C3G, were employed fluorescent resonance energy transfer (FRET)– examined by immunoblotting. based time-lapse microscopy. After transient transfection of

the probe for monitoring Rac1 activity (Fig. 5A), synovial Crk-Dependent Cytoskeletal Reorganization on HGF sarcoma cell lines were treated with 50 ng/mL HGF. Positive Stimulation signals indicating the activation of Rac1 were observed in We proceeded to observe the appearances of actin cytoskel- all parental synovial sarcoma cell lines as red color, eton of synovial sarcoma cell lines to examine the effects of Rac1 particularly at the cellular membrane of SYO-1 cells after activation involved in cytoskeletal reorganization. In both HGF stimulation (Fig. 5B). In parental Fuji cells, positive parental SYO-1 and HS-SY-II cells, HGF stimulation attenuated signals were diffusely observed in the cytoplasm without the formation of actin stress fibers; instead, the extreme HGF (Fig. 5C, a). On HGF stimulation, the careful prolonged and the numerous filopodia formations were produced observation showed the additional activation of Rac1 as in SYO-1 cells (Fig. 6A, d, arrows) and HS-SY-II cells (Fig. 6A, spots around the cytoplasmic membrane (Fig. 5C, c, arrow) e, arrow), respectively. In Fuji cells, HGF induced the dynamic followed by subsequent cytoplasmic protrusion (Fig. 5C, d, morphologic changes with filopodia formation at its leading square), leading to the dynamic morphologic change edge (Fig. 6A, f, arrows). The prominent membrane ruffles were (Fig. 5C, e and f). In contrast, the activity of Rac1 was also detected in all cells on HGF stimulation (Fig. 6A, d-f, reduced by HGF stimulation in Crk knockdown Fuji cells arrowheads). On the other hand, in Crk knockdown cells, actin (Fig. 5C, g-l), correlating with the result of the pull-down was visualized by anti-actin antibody that more clearly assay (Fig. 4). delineated actin network compared with phalloidin recognizing

FIGURE 4. Analysis of Rac activity in Crk knockdown cells by pull-down assay. Pull-down assay for active form of Rac. Wild-type and Crk knockdown synovial sarcoma cell lines SYO-1, HS-SY-II, and Fuji were stimulated with50 ng/mL HGF for theindicated time, and cell lysates were subjected to pull-down assay using glutathione S -transferase-PAK-RBD as probe. The activity of Rac at each time point was normalized by that without HGF (indicated as 0 minute) and described as ratio value.

FIGURE 5. Analysis of Rac activity in Crk knockdown cells by FRET. A. Schematic structure of FRET probe for monitoring the activity of Rac. Rac-GTP bound to its effector PAK-RBD results in the closed localization between CFP and YFP. In this situation, the energy of the excited CFP is transferred to YFP, and signals from YFP are detected as 580 nm fluorescence. Red, positive signals. B. As a FRET probe for monitoring Rac activity, pRaichu-Rac1-CAAX was transiently transfected into wild-type SYO-1 cells. Thirty-six hours after transfection, cells were treated with 50 ng/mL HGF and FRET reaction was analyzed. Representative FRET image at 25 minutes after stimulation. C. pRaichu-Rac1-CAAX was transiently transfected into wild-type and Crk knockdown Fuji cells. Thirty-six hours after transfection, cells were treated with 50 ng/mL HGF and FRET reaction was analyzed for additional 80 minutes.

only F-actin. In Crk-depleted Fuji cells, actin fibers were cell motility in both parental and control cells at 48 hours after disorganized with random direction (Fig. 6B, b) and speckled scratching. However, Crk knockdown cells did not indicate pattern (Fig. 6B, c). To prove these aberrant organizations of any HGF-dependent enhancement on cell motility (Fig. 7A, actin were not caused by artificial error in the process of right). immunostaining, actin organization in the living cells were Furthermore, the time-lapse microscopic observation showed examined by monitoring the exogenously expressed green significant cell scattering in parental Fuji cells on HGF fluorescent protein–fused actin. Indeed, we confirmed the stimulation (Fig. 7B, a and b; Supplementary Data 2). Crk similar disorganized actin fibers and the speckled pattern in knockdown cells displayed flatter cell morphology compared Crk knockdown Fuji cells (Fig. 6B, e and f). HGF treatment did with parental cells and could not dissociate from cell assembly not induce any alterations on these actin disorganizations (data even on HGF stimulation that resulted in showing no significant not shown). cell movement (Fig. 7B, c and d; Supplementary Data 3).

Analysis of Cell Motility and Scattering in CrkKnockdown Suppression of Massive Tumor Formation by Crk Cells Depletion in Nude Mice Because HGF-dependent enhancement on the invasion was To examine the significance of Crk on tumorigenetic and/or observed in all three synovial sarcoma cell lines (Supplemen- metastatic phenotypes of synovial sarcoma in vivo, we have tary Data 1B), we further did the wound-healing assays to s.c. injected wild-type, empty, or Crk knockdown Fuji cells on analyze the effects of Crk elimination on HGF-induced cell the back of nude mice. One month after injection, wild-type motility. In the absence of HGF, Crk knockdown Fuji cells and empty cells developed the tumor masses sized of 20.0 and exhibited f40% and 50% reduction of motility relative to 14.6 mm, respectively, on the average diameter (Fig. 8A). The control cells at 24 and 48 hours after scratching, respectively microscopic analysis showed that newly formed tumors (Fig. 7A, left). HGF stimulation showed 1.4-fold increase on exhibited the typical morphologic features as synovial

sarcoma with the prominent mitosis (Fig. 8B, arrowheads) morphogenesis, epithelial-mesenchymal transition, cell spread- and the invaded cell growth into the peripheral tissues. ing, breakdown of epithelial adherence junction, and c-Jun NH2- Contrary, no tumor mass was formed in mice injected with terminal kinase activation for cell transformation (12, 40, 41). Crk knockdown Fuji cells (Fig. 8A). In its barely formed The overexpression of both HGF and c-Met has been shown in tumors, we hardly detected the images of condensed human synovial sarcoma (34-36), which possesses aggressive chromatins in contrast to that in tumors composed of wild- features, such as invasive growth into soft tissue around the knee type Fuji cells (Fig. 8B). joint and frequent metastasis to the lung (42). In this study, by the analysis of Crk knockdown synovial sarcoma cell lines, we Discussion found the crucial role of Crk in HGF/c-Met-mediated cell The adaptor protein Crk is involved in cellular transforma- motility of human synovial sarcoma cell lines through the tion, and currently, there are several reports that the over- sustained phosphorylation of Gab1. expression of Crk correlates with aggressiveness of various Gab1 is phosphorylated for a prolonged period of time human cancers (23-25). Crk also has been known to play a role (>60 minutes) on HGF stimulation in contrast to a transient in HGF/c-Met-mediated biological responses involved in phosphorylation in response to epidermal growth factor (43). Sustained phosphorylation of Gab1 is suggested to be required for the HGF-mediated responses. We also found that HGF stimulation induced the sustained phosphorylation (>3 hours) on Y307 residue of Gab1, which is the potential binding site for Crk, correlated with the recruitment of Crk to Gab1 (Fig. 3B). The sustained phosphorylation of Gab1 could be due to various mechanisms, including the kinetics of kinase activation (44, 45), half-life of the ligand-receptor complex at the cell surface (46, 47), and differential activation of phosphotyrosine phosphatases, such as SHPTP-2 (48, 49). In addition, we hypothesized that the stable complexes of Crk with Gab1 via its SH2 domain may induce the sustained phosphorylation of Gab1 probably by protecting its phosphotyrosines from phosphatases, such as SHPTP-2. In fact, overexpression of Crk-SH2 domain could induce the prominent phosphorylation on multiple YXXP motifs of Gab1 in HGF-independent manner (Fig. 2B). In addition, the phosphorylation of Gab1 evoked by HGF in parental SYO-1 cells was sustained >3 hours although c-Met was significantly degraded probably by the ubiquitin-proteasome pathway (Fig. 3B; refs. 50-52). In agreement with our hypothesis, Crk knockdown by RNA interference reduced the duration of HGF-induced phosphorylation on Y307 residue of Gab1 in synovial sarcoma cell lines (Fig. 3B). Y307 residue of Gab1, also known as the binding site for phospholipase Cg, can induce HGF-mediated branching tubulogenesis but not cell motility or scattering (43). Because the endogenous expression levels of CrkI and CrkII proteins were elevated in synovial sarcoma cell lines (data not shown), the sustained complexes of Crk to Gab1 following HGF stimulation may activate specific signaling pathway controlling cell motility and scattering of this sarcoma. FIGURE 6. Analyses of actin cytoskeletal reorganization in synovial As phosphatidylinositol 3-kinase has been reported to sarcoma cell lines A. Actin cytoskeleton in SYO-1 (a and d), HS-SY-II regulate Rac1 activity through the several guanine nucleotide (b and e), and Fuji cells (c and f) was visualized by phalloidin staining with(d-f ) or without (a-c) 50 ng/mL HGF. Sixty minutes after HGF exchange factors, such as SWAP70 and P-REX2 downstream of stimulation, filopodia formation was induced in all cell lines, particularly Gab1 (53-55), and Vav and Tiam-1 are also known to mediate the extreme prolonged in SYO-1 cells, the numerous in HS-SY-II cells, and at the leading edge in Fuji cells (arrows in d-f). Arrowheads in d to Rac1 activity involved in transformation (56, 57), we initially f, HGF-induced membrane ruffling. B. Disorganization of actin speculated that the implication of Crk in HGF-mediated Rac1 cytoskeleton in Crk knockdown Fuji cells. The actin cytoskeleton in activation may be partial. However, the significant suppression Crk knockdown Fuji cells (b and c) and its control cells (a) was analyzed using anti-actin antibody. In Crk knockdown cells, actin fibers of HGF-dependent activation of Rac1, cell motility, and were disorganized withrandom direction ( b) and speckled pattern (c). scattering was shown in Crk knockdown cells, which may Exogenous green fluorescent protein (GFP)-actin was transiently suggest the crucial role for Crk in HGF-evoked activation of expressed by transfection in control Fuji cells (d) and Crk knockdown cells (e and f) and observed by fluorescent microscopy. Control, vector- Rac1 and following cell motility in human synovial sarcoma control transfectants. cells.

subsequently in all regions of newly formed filopodia at 30 minutes (Supplementary Data 1C). Furthermore, glutathione S-transferase pull-down assay in buffer containing CHAPS, which favors the release of the protein out of endoplasmic reticulum and Golgi, seemed to exhibit no effects on Rac1 activation in endoplasmic reticulum and Golgi (Supplementary Data 1D). Alternatively, the activation of Rac-GAP, such as h2-chimaerin, or GC-GAP may be involved in the biphasic activation of Rac (59, 60). The precise mechanisms underlying the biphasic activation of Rac1 in synovial sarcoma cells should be continued the investigation. On the other hand, in Crk knockdown cells, we observed a decline of Rac1 activity after HGF stimulation (Fig. 4, right). This decreased activity of Rac1 may suggest that HGF provoked the activation of negative regulator for Rac1 activity, such as GC-GAP, by Crk-independent signaling mechanism. The activation of such negative signal in the absence of a Crk-dependent positive signal may induce down-regulation of Rac1 activity. Crk-depleted cells failed to induce HGF-dependent membrane rufflings as observed in parental synovial sarcoma cells, indicating that the functional disruption of Crk/Dock180/Rac1 signaling pathway impairs HGF-mediated cytoskeletal reorganization in synovial sarcoma cells. Rac1 and Cdc42 are known to cooperatively act for the spreading of Madin-Darby canine kidney epithelial cells stimulated with HGF (59). Because synovial sarcoma cells could induce filopodia formation on HGF stimulation (Fig. 6A, arrows), Cdc42 also may contribute to the remodeling of the actin cytoskeleton in this sarcoma. Furthermore, Crk knockdown cells remarkably disrupted the organization of actin cytoskeleton as represented as random FIGURE 7. Analyses of cell motility and scattering of Crk knockdown Fuji cells. A. Analysis of cell motility by wound-healing assay. Cells with direction of actin fibers (Fig. 6B, b and e) and speckled pattern confluent on the plate were scratched off and the distances of moved (Fig. 6B, c and f), which seemed to be similar to the cells were measured with( right) or without (left) 50 ng/mL HGF at 0, disorganization of actin network on the Rho inhibitor C3 or its 24, and 48 hours after wound formation. Points, averages of the distances of moved cells obtained in three independent experiments; downstream ROCK inhibitor Y27632 (data not shown). These bars, SD. Statistical analyses were done by Student’s t test. *, P < results may suggest the requirement of Crk on the activation of 0.05, Crk knockdown cells were significantly different from parent and control cells at 48 hours after wound formation. parent, parental Fuji Rho in synovial sarcoma cells (61), although we hardly detect cells; cont., vector-control transfected cells. B. Analysis of cell Crk-dependent alternations on the activation of Rho by scattering of parental Fuji cells and Crk knockdown cells on HGF glutathione S-transferase pull-down assay probably due to its stimulation. Images were detected at 5-minute intervals for 24 hours by phase-contrast time-lapse microscopy. The prominent cell scattering limited sensitivity (data not shown). Thus, Crk may compre- was observed in wild-type Fuji cells on HGF treatment for 24 hours (b); hensively contribute to the regulation of Rho family, including however, there was no significant cell movement detected in Crk Rac1, Cdc42, and Rho. knockdown cells (d). C3G, other target for Crk, is known to regulate cell adhesion through the activation of Rap1, but no significant change of the On HGF stimulation, a biphasic activation of Rac1 was cell attachment on culture dishes was observed in these Crk observed in synovial sarcoma cells (Fig. 4, left) in spite of knockdown sarcoma cell lines (data not shown). Previous study constant phosphorylation of Gab1. We hypothesized that the has shown that Gab1 association with CrkL correlates with differential activation of Rac1 in separate cell compartments c-Met-evoked activation of the Rap1 and decreased adhesion of may be involved in this biphasic activation of Rac1 as reported HEK293 cells (11). Because the expression of CrkL is retained about Ras, which is activated in both the cytoplasmic in Crk knockdown cells (Fig. 3A), CrkL may play a dominant membrane and the endomembranes of endoplasmic reticulum role in cell attachment. and Golgi (58). FRET-based time-lapse analysis was employed Wild-type and empty Fuji cells formed the tumor masses with to examine this points; however, we were unable to identify the the peripheral invasion in nude mice. However, the depletion of exact subcellular compartments arising the biphasic Rac1 Crk remarkably inhibited the massive tumor growth. Consistent activation because of its technical limitation. In our further with these results, we further found that Crk-depleted cells analysis, the monitoring of the exogenously expressed red exhibited the suppression of growth rate in vitro and the loss of fluorescent protein–fused Dock180 in SYO-1 cells did not colony formation in soft agar (data not shown). These results show HGF-mediated translocation of Dock180 to endoplasmic may suggest the crucial role of Crk on the tumorigenesis of reticulum, whereas we could find the induction of Dock180 to synovial sarcoma in addition to its function as the regulator of the cell surface at 10 minutes after HGF stimulation, cell motility and invasion.

SYO-1 and Fuji cells possess SYT-SSX2. In the original tumors, the typical monophasic morphologies were observed in HS-SY-II and Fuji cells, whereas SYO-1 cells exhibited the biphasic features composed of areas of the spindle cells and that of the epithelial cells arranged in glandular structures. SYO-1, HS-SY-II, and 293T human embryonic kidney cells were maintained in DMEM supplemented with 10% fetal bovine serum, and Fuji was maintained on type I collagen-coated dishes in RPMI 1640 with 10% fetal bovine serum.

Antibodies The antibodies against phosphotyrosine (PY20 and RC20H), p130Cas, paxillin, Crk, and Rac1 were purchased from Transduction Laboratories (Lexington, KY). Anti-Gab1 antibody was from Upstate (Lake Placid, NY). Anti-Met (C-12), C3G (C19), Dock180 (H4), and Crk-L (C20) antibodies were from Santa Cruz Biotechnology (Santa Cruz, CA). The antibodies against phosphorylated Met (Y1234/Y1235) and (Y1349) and phosphorylated Gab1 (Y307) were purchased from Cell Signaling (Beverly, MA). Anti-actin antibody was from Chemicon International (Temecula, CA). Anti-Flag antibody (M2) was from Sigma (St. Louis, MO). The phalloidin-594 for actin staining was purchased from Molecular Probes (Eugene, OR).

Plasmid cDNAs of CrkII-WT, SH2 mutant (R38V), and NH2- terminal SH3 mutant (W169L) were subcloned into pCXN2 expression vector with NH2-terminal Flag tag sequence (DYKDDDDK). pBat-Flag-Gab1 and pBat-Flag-Gab1-y YXXP were generous gifts from S.M. Feller (Oxford FIGURE 8. Suppressed tumor formation by Crk elimination in nude University, Oxford, United Kingdom). pGEX-PAX2-RBD mice. A. Tumor formation in nude mice. The wild-type Fuji cells formed the for pull-down assay of Rac was described previously (62). tumor masses with the peripheral invasion at the back of nude mice 1 monthafter cell inoculation, whereasno significant tumor mass was pEGFP-actin was purchased from Clontech (Mountain observed in mice injected withCrk knockdown cells. Columns, average View, CA). diameters of the formed tumors in each mice; bars, SD. Statistical analyses were done by Student’s t test. *, P < 0.05, Crk knockdown tumors on the average diameter was significantly different from parental and cont. tumors. B. Histologic features of tumors formed by parental Fuji Immunoprecipitation and Immunoblotting cells or its Crk-eliminated cells were examined on H&E staining. Arrows, Cells were lysed with a lysis buffer [10 mmol/L Tris-HCl parental tumor indicates the condensed chromatin, which is represented (pH 7.4), 150 mmol/L NaCl, 1 mmol/L EDTA, 0.5% NP40, as an indicator for high cell proliferation. 50 mmol/L NaF, 1 mmol/L phenylmethylsulfonyl fluoride, 1 mmol/L Na3VO4] for 10 minutes on ice and centrifuged at 15,000 rpm for 10 minutes at 4jC. Supernatants were Our study shows a requirement of Crk in HGF-mediated incubated with the appropriate antibody at 4jC with gentle constitutive activation of Rac1 through the sustained coupling shaking overnight followed by incubation with protein A/G- of Crk to Gab1 result in promoting cell motility and scattering Sepharose. After washing with lysis buffer, the immunocom- of human synovial sarcoma cells. Thus, Crk might be an plexes were subjected to SDS-PAGE, and separated proteins attractive therapeutic target for human synovial sarcoma. were transferred to a polyvinylidene difluoride filter (Immo- bilon, Millipore, Billerica, MA). A blocking buffer was used Materials and Methods to reduce nonspecific signals (1% bovine serum albumin in Cells TBS-Tween 20 for detecting phosphotyrosine, 5% skim milk Human synovial sarcoma cell lines SYO-1, HS-SY-II, and in TBS-Tween 20 for other proteins) before incubation with Fuji were established by A. Kawai (National Cancer Center, primary antibodies followed by peroxidase-labeled secondary Tokyo, Japan; ref. 39), H. Sonobe (Kochi Medical Univer- antibodies. Signals were detected by the enhanced chemilu- sity, Kochi, Japan), and T. Nojima (Kanazawa Medical minescence Western blotting reagent (Amersham Pharmacia University, Kanazawa, Japan), respectively. HS-SY-II cells Biotech, Piscataway, NJ) and quantified using a Lumino genetically occur as a SYT-SSX1 fusion transcript, whereas Image Analyzer (LAS1000, Fuji Film, Tokyo, Japan).

Establishment of CrkKnockdownSynovial Sarcoma Cell Analysis of Scattering Lines The scattering of cells followed by 50 ng/mL HGF treatment The plasmid containing small interfering RNAs for Crk was observed by phase-contrast microscopy using the Meta- (CrkI) was described previously (63). The CrkI corres- Morph software. Images were saved at 5-minute intervals for a ponding to the bases 277 to 296 of human c-CrkII encoding period of 24 hours. sequences as 5¶-GAGTTTGATTCATTGCCTG-3¶ and 5¶- CAGGCAATGAATCAAACTC-3¶ were subcloned into SalI In vivo Tumor Formation Assay in Nude Mice and KpnI sites of the pSUPER (suppression of endogenous Wild-type Fuji cells, empty-plasmid transfected cells RNA) vector (64) and transfected into synovial sarcoma (empty cells), and its Crk-depleted cells were employed to cell lines by Fugene 6 transfection reagent (Roche, Indian- examine the potential on the tumorigenesis and the invasive apolis, IN). Following selection with 0.5 Ag/mL puromycin growth into nude mice. Each 1 Â 107 cells in PBS were s.c. (Sigma), expression levels of Crk were confirmed by injected into the 6-week-old female nude mice, BALB/cA Jcl- immunoblotting. nu nu/nu (Clea Japan, Inc., Tokyo, Japan). Three mice each for wild-type and empty cells and six mice for Crk-depleted Analysis of the Actin Cytoskeleton cells were employed. One month after cell inoculation, the Cells were cultured on eight-chamber slides (Nunc, Naper- mice were euthanized, and the diameters of the formed tumors ville, IL). With or without 50 ng/mL HGF treatment for 60 were determined. The histologic features were identified on minutes, cells were fixed with 3.5% formaldehyde in PBS. H&E staining. After permeabilization with 0.1% Triton X-100 in PBS for four minutes at room temperature and blocking with 1% bovine serum albumin for 20 minutes at room temperature, Acknowledgments actin was stained by phalloidin-594 (1:40 dilution) in PBS for We thank Dr. M. Matsuda for setting up a FRET-based time-lapse microscopy and providing plasmids for monitoring the activity of small GTPases in a 1 hour at room temperature in the dark and observed using a single cell and Dr. S.M. Feller for the expression plasmid of wild-type Gab1 confocal laser scanning microscopy (FV-300, Olympus, and its mutant. Tokyo, Japan). In addition, green fluorescent protein–labeled actin cytoskeleton transfected in wild-type and Crk knockdown Fuji was observed using a fluorescent microscopy References 1. Bottaro DP, Rubin JS, Faletto DL, et al. Identification of the hepatocyte equipped with MetaMorph software (Universal Imaging, growth factor receptor as the c-met proto-oncogene product. Science 1991;251: Downingtown, PA). 802 – 4. 2. Stoker M, Gherardi E, Perryman M, Gray J. Scatter factor is a fibroblast-derived modulator of epithelial cell mobility. Nature 1987;327: Pull-down Assay for Rac Activity 239 – 42. The GTP-bound form of Rac was detected as described 3. Rodrigues GA, Park M. Autophosphorylation modulates the kinase activity previously (62). Synovial sarcoma parental cells and CrkI cells and oncogenic potential of the Met receptor tyrosine kinase. Oncogene 1994;9: treated with or without HGF were lysed with lysis buffer [1% 2019 – 27. NP40, 25 mmol/L HEPES (pH 7.4), 150 mmol/L NaCl, 10% 4. Ponzetto C, Bardelli A, Zhen Z, et al. A multifunctional docking site mediates signaling and transformation by the hepatocyte growth factor/scatter factor glycerol, 1 mmol/L EDTA, 10 mmol/L MgCl2,1Ag/mL apro- receptor family. Cell 1994;77:261 – 71. tinin, 1 mmol/L phenylmethylsulfonyl fluoride]. Cell lysates 5. Weidner KM, Di Cesare S, Sachs M, Brinkmann V, Behrens J, were centrifuged at 12,000 rpm at 4jC for 1 minute. The Birchmeier W. Interaction between Gab1 and the c-Met receptor ty- A rosine kinase is responsible for epithelial morphogenesis. Nature 1996; supernatant was incubated with 10 g purified glutathione 384:173 – 6. S-transferase-PAK2-RBD followed by incubation with glutathi- 6. Schaeper U, Gehring NH, Fuchs KP, Sachs M, Kempkes B, Birchmeier W. one-Sepharose beads for 1 hour at 4jC while rotating. The beads Coupling of Gab to c-Met, Grb2, and Shp2 mediates biological responses. J Cell were washed thrice with lysis buffer, and the precipitates were Biol 2000;149:1419 – 32. subjected to SDS-PAGE. After transfer to filter, the GTP-bound 7. Lock LS, Royal I, Naujokas MA, Park M. Identification of an atypical Grb2 carboxyl-terminal SH3 domain binding site in Gab docking proteins reveals Rac was detected by mouse monoclonal anti-Rac1 antibody. Grb2-dependent and -independent recruitment of Gab1 to receptor tyrosine kinases. J Biol Chem 2000;275:31536 – 45. FRET-Based Time-Lapse Analysis of Rac Activity 8. Maroun CR, Holgado-Madruga M, Royal I, et al. The Gab1 PH domain is required for localization of Gab1 at sites of cell-cell contact and epithelial The FRET probe for monitoring Rac activity, pRaichu-Rac- morphogenesis downstream from the met receptor tyrosine kinase. Mol Cell Biol CAAX, was a generous gift from M. Matsuda (Osaka 1999;19:1784 – 99. University, Osaka, Japan; ref. 65). The plasmid was transfected 9. Garcia-Guzman M, Dolfi F, Zeh K, Vuori K. Met-induced JNK activation is j mediated by the adaptor Crk and correlates with the Gab1-Crk signaling complex into cell lines and incubated at 37 C for 36 hours. After formation. Oncogene 1999;18:7775 – 86. treatment with HGF (50 ng/mL), FRET reaction in single cells 10. Sakkab D, Lewitzky M, Posern G, et al. Signaling of hepatocyte was analyzed using MetaMorph software. growth factor/scatter factor (HGF) to the small GTPase Rap1 via the large docking protein Gab1 and the adapter protein CRKL. J Biol Chem 2000; 275:10772 – 8. Wound-Healing Assay 11. Lamorte L, Kamikura DM, Park M. A switch from p130Cas/Crk to Gab1/ The wound-healing assay was done as described previously Crk signaling correlates with anchorage independent growth and JNK activation in cells transformed by the Met receptor oncoprotein. Oncogene (66). Briefly, 48 hours after plating on culture dishes, the cells 2000;19:5973 – 81. were scraped off/wounded using a yellow tip. After 24 and 12. Maroun CR, Naujokas MA, Holgado-Madruga M, Wong AJ, Park M. The 48 hours, movement of the cells was measured. tyrosine phosphatase SHP-2 is required for sustained activation of extracellular

signal-regulated kinase and epithelial morphogenesis downstream form the met 36. Kuhnen C, Tolnay E, Steinau HU, Voss B, Muller KM. Expression of c-Met receptor tyrosine kinase. Mol Cell Biol 2000;20:8513 – 25. receptor and hepatocyte growth factor/scatter factor in synovial sarcoma and 13. Gual P, Giordano S, Williams TA, Rocchi S, Van Obberghen E, epithelial sarcoma. Virchows Arch 1998;432:337 – 42. Comoglio PM. Sustained recruitment of phospholipase C-g to Gab1 is 37. Nishio J, Iwasaki H, Ishiguro M, et al. Establishment of a new human required for HGF-induced branching tubulogenesis. Oncogene 2000;19: synovial sarcoma cell line, FU-SY-1, that expresses c-Met receptor and its ligand 1509 – 18. hepatocyte growth factor. Int J Oncol 2002;21:17 – 23. 14. Lecoq-Lafon C, Verdier F, Fichelson S, et al. Erythropoietin induces the 38. Oda Y, Sakamoto A, Saito T, Kinukawa N, Iwamoto Y, Tsuneyoshi M. tyrosine phosphorylation of GAB1 and its association with SHC, SHP2, SHIP, Expression of hepatocyte growth factor (HGF)/scatter factor and its receptor and phosphatidylinositol 3-kinase. Blood 1999;93:2578 – 85. c-MET correlates with poor prognosis in synovial sarcoma. Hum Pathol 2000; 15. Mayer BJ, Hamaguchi M, Hanafusa H. A novel viral oncogene with 31:185 – 92. structural similarity to phospholipase C. Nature 1988;332:272 – 5. 39. Kawai A, Naito N, Yoshida A, et al. Establishment and characterization of 16. Matsuda M, Tanaka S, Nagata S, Kojima A, Kurata T, Shibuya M. Two a biphasic synovial sarcoma cell line, SYO-1. Cancer Lett 2004;204:105 – 13. species of human CRK cDNA encode proteins with distinct biological activities. 40. Lamorte L, Rodrigues S, Naujokas M, Park M. Crk synergizes with Mol Cell Biol 1992;12:3482 – 9. epidermal growth factor for epithelial invasion and morphogenesis and is 17. Feller SM. Crk family adaptors—signalling complex formation and required for the met morphogenic program. J Biol Chem 2002;277: biological roles. Oncogene 2001;20:6348 – 71. 37904 – 11. 18. Hasegawa H, Kiyokawa E, Tanaka S, et al. DOCK180, a major CRK-binding 41. Lamorte L, Royal I, Naujokas M, Park M. Crk adapter proteins promote an protein, alters cell morphology upon translocation to the cell membrane. Mol Cell epithelial-mesenchymal-like transition and are required for HGF-mediated cell Biol 1996;16:1770 – 6. spreading and breakdown of epithelial adhesions junctions. Mol Biol Cell 2002; 13:1449 – 61. 19. Kiyokawa E, Hashimoto Y, Kobayashi S, Sugimura H, Kurata T, Matsuda M. Activation of Rac1 by a Crk SH3-binding protein, DOCK180. Genes Dev 1998; 42. Ladanyi M, Antonescu CR, Leung DH, et al. Impact of SYT-SSX fusion type 12:3331 – 6. on the clinical behavior of synovial sarcoma: a multi-institutional retrospective study of 243 patients. Cancer Res 2002;62:135 – 40. 20. Tanaka S, Morishita T, Hashimoto Y, et al. C3G, a guanine nucleotide- releasing protein expressed ubiquitously, binds to the Src homology 3 43. Gual P, Giordano S, Williams TA, Rocchi S, Obberghen E, Comoglio PM. domains of CRK and GRB2/ASH proteins. Proc Natl Acad Sci U S A 1994; Sustained recruitment of phospholipase C-g Gab1 is required for HGF-induced 91:3443 – 7. branching tubulogenesis. Oncogene 2000;19:1509 – 18. 21. Gotoh T, Hattori S, Nakamura S, et al. Identification of Rap1 as a target for 44. Naldini L, Vigna E, Narsimhan RP, et al. Hepatocyte growth factor (HGF) the Crk SH3 domain-binding guanine nucleotide-releasing factor C3G. Mol Cell stimulates the tyrosine kinase activity of the receptor encoded by the proto- Biol 1995;15:6746 – 53. oncogene c-MET. Oncogene 1991;6:501 – 4. 22. Ohba Y, Ikuta K, Ogura A, et al. Requirement for C3G-dependent Rap1 45. Mclntyre BS, Birkenfeld HP, Sylvester PW. Relationship between activation for cell adhesion and embryogenesis. EMBO J 2001;20:3333 – 41. epidermal growth factor receptor levels, autophosphorylation and mitogenic- responsiveness in normal mouse mammary epithelial cells in vitro. Cell Prolif 23. Nishihara H, Tanaka S, Tsuda M, et al. Molecular and immunohistochemical 1995;28:45 – 56. analysis of signaling adaptor protein Crk in human cancers. Cancer Lett 2002; 180:55 – 61. 46. Kozu A, Kato Y, Shitara Y, Hanano M, Sugiyama Y. Kinetic analysis of transcytosis of epidermal growth factor in Madin-Darby canine kidney epithelial 24. Takino T, Nakada M, Miyamori H, Yamashita J, Yamada KM, Sato H. CrkI cells. Pharm Res 1997;14:1228 – 35. adaptor protein modulates cell migration and invasion in glioblastoma. Cancer Res 2003;63:2335 – 7. 47. Helin K, Beguinot L. Internalization and down-regulation of the human epidermal growth factor receptor are regulated by the carboxyl-terminal tyrosines. 25. Miller CT, Chen G, Gharib TG, et al. Increased C-CRK proto-oncogene J Biol Chem 1991;266:8363 – 8. expression is associated with an aggressive phenotype in lung adenocarcinomas. Oncogene 2003;22:7950 – 7. 48. Bardelli A, Longati P, Gramaglia D, Stella MC, Comoglio PM. Gab1 coupling to the HGF/Met receptor multifunctional docking site requires binding 26. Nakajima M, Sawada H, Yamada Y, et al. The prognostic significance of of Grb2 and correlates with the transforming potential. Oncogene 1997;15: amplification and overexpression of c-met and c-erbB-2 in human gastric 3103 – 11. carcinomas. Cancer 1999;85:1894 – 902. 49. Lehr S, Kotzka J, Herkner A, et al. Identification of tyrosine phosphorylation 27. Koochekpour S, Jeffers M, Rulong S, et al. Met and hepatocyte growth sites in human Gab-1 protein by EGF receptor kinase in vitro. Biochemistry 1999; factor/scatter factor expression in human gliomas. Cancer Res 1997;57: 38:151 – 9. 5391 – 8. 50. Jerrers M, Taylor GA, Weidner KM, Omura S, Vande Woude GF. 28. Schmidt L, Duh RM, Chen F, et al. Germline and somatic mutations in the Degradation of the Met tyrosine kinase receptor by the ubiquitin-proteasome tyrosine kinase domain of the MET proto-oncogene in papillary renal carcinomas. pathway. Mol Cell Biol 1997;17:799 – 808. Nat Genet 1997;16:68 – 73. 51. Petrelli A, Gilestro GF, Lanzardo S, Comoglio PM, Migone N, Giordano S. 29. Camp RL, Rimm EB, Rimm DL. Met expression is associated with poor The endophilin-CIN85-Cbl complex mediates ligand-dependent downregulation outcome in patients with axillary lymph node negative breast carcinoma. Cancer of c-Met. Nature 2002;416:187 – 90. 1999;86:2259 – 65. 52. Taher TE, Tjin EP, Beuling EA, Borst J, Spaargaren M, Pals ST. 30. Seidel C, Borset M, Turesson I, Abildgaard N, Sundan A, Waage A. c-Cbl is involved in Met signaling in B cells and mediates hepatocyte Elevated serum concentrations of hepatocyte growth factor in patients with growth factor-induced receptor ubiquitination. J Immunol 2002;169: multiple myeloma. The Nordic Myeloma Study Group. Blood 1998;91: 3793–800. 806 – 12. 53. Furge KA, Zhang YW, Vande Woude GF. Met receptor tyrosine kinase: 31. Otsuka T, Takayama H, Sharp R, et al. c-Met autocrine activation induces enhanced signaling through adapter proteins. Oncogene 2000;19:5582 – 9. development of malignant melanoma and acquisition of the metastatic phenotype. Cancer Res 1998;58:5157 – 67. 54. Shinohara M, Terada Y, Iwamatsu A, et al. SWAP-70 is a guanine- nucleotide-exchange factor that mediates signalling of membrane ruffling. Nature 32. Clark J, Rocques PJ, Crew AJ, et al. Identification of novel genes, SYT and 2002;416:759 – 63. SSX, involved in the t(X;18)(p11.2;q11.2) translocation found in human synovial sarcoma. Nat Genet 1994;7:502 – 8. 55. Rosenfeldt H, Vazquez-Prado J, Gutkind JS. P-REX2, a novel PI-3-kinase sensitive Rac exchange factor. FEBS Lett 2004;572:167 – 71. 33. Nagai M, Tanaka S, Tsuda M, et al. Analysis of transforming activity of human synovial sarcoma-associated chimeric protein SYT-SSX1 bound to 56. Collard JG, Habets GG, Michiels F, Stam J, van der Kammen RA, van chromatin remodeling factor hBRM/hSNF2a. Proc Natl Acad Sci U S A 2001; Leeuwen F. Role of Tiam 1 in Rac-mediated signal transduction pathways. Curr 98:3843 – 8. Top Microbiol Immunol 1996;213:253 – 65. 34. Motoi T, Ishida T, Kuroda M, et al. Coexpression of hepatocyte growth 57. Palmby TR, Abe K, Karnoub AE, Der CJ. Vav transformation requires factor and c-Met protooncogene product in synovial sarcoma. Pathol Int 1998; activation of multiple GTPases and regulation of gene expression. Mol Cancer 48:769 – 75. Res 2004;2:702 – 11. 58. Chiu VK, Bivona T, Hach A, et al. Ras signalling on the endoplasmic 35. Fukuda T, Ichimura E, Shinozaki T, et al. Coexpression of HGF and reticulum and the Golgi. Nat Cell Biol 2002;4:343 – 50. c-Met/HGF receptor in human bone and soft tissue tumors. Pathol Int 1998;48: 757 – 62. 59. Menna PL, Skilton G, Leskow FC, Alonso DF, Gomez DE, Kazanietz MG.

Inhibition of aggressiveness of metastatic mouse mammary carcinoma cells by the 63. Nagashima K, Endo A, Ogita H, et al. Adaptor protein Crk is required for h2-chimaerin GAP domain. Cancer Res 2003;63:2284 – 91. ephrin-B1-induced membrane ruffling and focal complex assembly of human 60. Zhao C, Ma H, Bossy-Wetzel E, Lipton SA, Zhang Z, Feng GS. GC-GAP, aortic endothelial cells. Mol Biol Cell 2002;13:4231 – 42. a Rho family GTPase-activating protein that interacts with signaling adapters 64. Brummelkamp TR, Bernards R, Agami R. A system for stable expression of Gab1 and Gab2. J Biol Chem 2003;278:34641 – 53. short interfering RNAs in mammalian cells. Science 2002;296:550 – 3. 61. Tsuda M, Tanaka S, Sawa H, Hanafusa H, Nagashima K. Signaling adaptor 65. Itoh RE, Kurokawa K, Ohba Y, Yoshizaki H, Mochizuki N, Matsuda M. protein v-Crk activates Rho and regulates cell motility in 3Y1 rat fibroblast cell Activation of rac and cdc42 video imaged by fluorescent resonance energy line. Cell Growth Differ 2002;13:131 – 9. transfer-based single-molecule probes in the membrane of living cells. Mol Cell Biol 2002;22:6582 – 91. 62. Nishihara H, Maeda M, Oda A, et al. DOCK2 associates with CrkL and regulates Rac1 in human leukemia cell lines. Blood 2002; 66. Mannino RJ, Ballmer K, Zeltner D, Burger MM. An inhibitor of animal 100:3968 – 74. cell growth increases cell-to-cell adhesion. J Cell Biol 1981;91:855 – 9.

Mol Cancer Res 2006;4(7). July 2006 Downloaded from mcr.aacrjournals.org on September 25, 2021. © 2006 American Association for Cancer Research. Adaptor Molecule Crk Is Required for Sustained Phosphorylation of Grb2-Associated Binder 1 and Hepatocyte Growth Factor−Induced Cell Motility of Human Synovial Sarcoma Cell Lines

Takuya Watanabe, Masumi Tsuda, Yoshinori Makino, et al.

Mol Cancer Res 2006;4:499-510.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/4/7/499

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2006/09/28/4.7.499.DC1

Cited articles This article cites 66 articles, 33 of which you can access for free at: http://mcr.aacrjournals.org/content/4/7/499.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/4/7/499.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/4/7/499. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.