Oncogene (2009) 28, 1110–1120 & 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www..com/onc ORIGINAL ARTICLE Activation of the non-canonical Dvl–Rac1–JNK pathway by Frizzled homologue 10 in human synovial sarcoma

C Fukukawa1, S Nagayama1, T Tsunoda2, J Toguchida3, Y Nakamura1 and T Katagiri1,4

1Laboratory of Molecular Medicine, Center, Institute of Medical Science, The University of Tokyo, Tokyo, Japan; 2Laboratory for Medical Informatics, SNP Research Center, RIKEN (Institute of Physical and Chemical Research), Yokohama, Japan and 3Institute for Frontier Medical Sciences, Kyoto University, Kyoto, Japan

We previously reported that Frizzled homologue 10 Szuhai et al., 2004; Williams et al., 2004; Hosono et al., (FZD10), a member of the Wnt signal receptor family, 2005; Shannon et al., 2005). Hence, efforts have been was highly and specifically upregulated in synovial made to understand the molecular basis of synovial sarcoma and played critical roles in its cell survival and sarcoma, but the origin tissue of synovial sarcoma is still growth. We here report a possible molecular mechanism not clarified. A SYT-SSX fusion , a product of of the FZD10 signaling in synovial sarcoma cells. We chromosomal translocation t(X;18)(p11;q11), is a well- found a significant enhancement of phosphorylation of the known feature of synovial sarcoma and it is considered Dishevelled (Dvl)2/Dvl3 complex as well as activation of to play a critical role in the oncogenesis of synovial the Rac1–JNK cascade in synovial sarcoma cells in which sarcoma (Turc-Carel et al., 1986; Clark et al., 1994; FZD10 was overexpressed. Activation of the FZD10– Brett et al., 1997; Lim et al., 1998; Soulez et al., 1999). Dvls–Rac1 pathway induced lamellipodia formation and Furthermore, several groups reported the gene expres- enhanced anchorage-independent cell growth cells. sion profile analysis using cDNA microarray and FZD10 overexpression also caused the destruction of the revealed that synovial sarcoma showed a distinct gene actin cytoskeleton structure, probably through the down- expression pattern compared with other spindle cell regulation of the RhoA activity. Our results have strongly sarcomas (Nagayama et al., 2002; Nielsen et al., 2002; implied that FZD10 transactivation causes the activation Segal et al., 2003). of the non-canonical Dvl–Rac1–JNK pathway and plays We previously analysed gene expression profiles critical roles in the development/progression of synovial of 47 soft tissue sarcomas using a genome-wide sarcomas. cDNA microarray consisting of 23 040 and found Oncogene (2009) 28, 1110–1120; doi:10.1038/onc.2008.467; that Frizzled homologue 10 (FZD10) was upregulated published online 12 January 2009 specifically in synovial sarcomas at a very high level (Nagayama et al., 2002). FZD10 is a member of Keywords: Frizzled; synovial sarcoma; JNK; cytoskele- the Frizzled family and considered as a seven-trans- ton; anchorage-independent growth; Rac membrane Wnt-signaling receptor (Koike et al., 1999). When FZD10 expression was suppressed using specific siRNAs, the growth of synovial sarcoma cells was significantly suppressed (Nagayama et al., 2005), indicating an essential role of FZD10 in the Introduction growth or survival of the sarcoma cells. As FZD10 expression was absent or hardly detectable in Synovial sarcoma accounts for 5–10% of all soft tissue any normal organs except the placenta, we considered sarcomas and its onset age is as young as 10–20 years that novel therapeutics targeting this molecule (Weiss and Goldblum, 2001). Although it is called as would be specifically effective to cancers and would ‘synovial’ sarcoma, it occurs in the head, neck, trunk, have no or little risk for adverse reactions (Nagayama heart, lung, esophagus, bone marrow, larynx, kidney et al., 2005). and prostate where no synovial tissue is present In this study, we report further biological mechanisms (Habu et al., 1998; Hiraga et al., 1999; Coindre et al., of FZD10 overexpression in synovial sarcoma cells. Our 2003; McGilbray and Schulz, 2003; Ochi et al., 2004; data presented here indicate that FZD10 enhances phosphorylation of the Dishevelled 2 (Dvl2) and Correspondence: Dr T Katagiri, Laboratory of Molecular Medicine, Dishevelled 3 (Dvl3) complex, and results in significant Human Genome Center, Institute of Medical Science, University of activation of the Rac1–JNK signaling pathway. As a Tokyo, 4-6-1 Shirokanedai, Minato-ku, Tokyo 108-8639, Japan. result, it causes rearrangement of the actin cytoskeleton E-mail: [email protected] structure and the enhancement of anchorage-indepen- 4Current address: Division of Genome Medicine, Institute for Genome Research, The University of Tokushima, Tokushima 770-8503, Japan dent cell growth. These results suggest the critical Received 14 July 2008; revised 1 December 2008; accepted 3 December involvement of FZD10 transactivation in development 2008; published online 12 January 2009 and progression of synovial sarcoma. FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1111 Results treated with various kinase inhibitors and effects on their phosphorylation levels were investigated. Among Regulation of Dvl2 and Dvl3 by FZD10 various kinase inhibitors, we found that D4476, which is FZD10 was previously reported to be involved in known to be a specific inhibitor of Casein kinase 1 canonical WNT-b/catenin signaling in human airway (CK1) (Rena et al., 2004; Bain et al., 2007), was able to smooth muscle cells (Wang et al., 2005) as well as in reduce the phosphorylation of Dvl2 and Dvl3 in mouse development (Nunnally and Parr, 2004). How- FZD10-overexpressing SYO-1 cells, suggesting a possi- ever, we found no activation of b-catenin/TCF (T-cell ble involvement of CK1 in the phosphorylation of Dvl2 factor) signaling in synovial sarcoma cells in which and Dvl3 in FZD10-positive cells (Figure 1d). FZD10 was expressed at a high level (data not shown), Synergistic phosphorylation of Dvl2 and Dvl3 in the suggesting that FZD10 activation is not related to the presence of FZD10 interested us in Dvl2/Dvl3 hetero- canonical WNT signaling pathway in synovial sarcoma oligomerization (Rothba¨ cher et al., 2000; Angers et al., cells. Therefore, we considered a possibility that the 2006). To investigate this possibility, we constructed the downstream signaling of FZD10 in synovial sarcoma myc-tagged-Dvl2 vector (Dvl2-myc) and transfected was different from the cases reported here. Dvl into FZD10-stably expressing cells (FZD10-2) or Mock are known to act as an important scaffold in the cells (Mock2). The results showed that Dvl2 and Dvl3 WNT-Frizzled signals (Wharton, 2003). A large number formed a complex, irrespective of the existence of of signaling proteins are suggested to bind to Dvl FZD10 (Figure 1e). proteins, through which signals in the divergent path- ways, such as cell motility, cell polarity and migration, Disruption of actin cytoskeleton by FZD10 are mediated through the non-canonical WNT signal- overexpression ing. Hence, we investigated possible correlations in the As the Dvl family is known to be involved in the expression levels of the FZD10 protein and the Dvl regulation of the actin cytoskeleton structure (Capelluto members, and found some correlations between FZD10 et al., 2002; Endo et al., 2005; Wechezak and Coan, 2005), expression and Dvl2 and Dvl3 expression levels in we examined actin fibers in cells with or without FZD10 synovial sarcoma cell lines by western blot analysis, expression by immunocytochemical analysis. The actin whereas no significant correlation was observed between stress fiber was completely disassembled in HuBM-Bmi1- FZD10 and Dvl1 (data not shown). Western blot hT-immortalized mesenchymal stem cells transfected with analysis using anti-Dvl2 and anti-Dvl3 antibodies FZD10 (Figure 2a; ii and iv) as well as in FZD10- detected enhancement of a higher-molecular-weight introduced 1273/99 cells (Figure 2a; vi and viii), whereas band of Dvl2 protein (polyclonal antibody) as well as it was well assembled in their parental cells (Figure 2a; i, the total protein level of Dvl3 in 1273/99-FZD10 stable- iii, v and vii). Interestingly, we observed lamellipodia at expressing cell lines (FZD10-1 and FZD10-2) the cell periphery in the FZD10-overexpressing cells (Figure 1a). Moreover, interestingly when we used an (Figure 2a; ii, iv, vi and viii), but not in the cells without anti-Dvl2 monoclonal antibody (MAb10B5), which is the introduction of FZD10 plasmid (Figure 2a; i, iii, v known not to recognize phosphorylated Dvl2 (Gonza´ - and vii). These findings indicate that overexpression of lez-Sancho et al., 2004), we could not detect Dvl2 FZD10 led to the disruption of the actin stress fiber protein in FZD10-positive YaFuSS cells, but could structure, possibly due to the regulation of Dvl proteins. detect Dvl2 after treatment of l-protein phosphatase Subsequently, to examine an effect of FZD10 on (Supplementary Figure S1). The result strongly implied RhoA activity, we transfected an FZD10-specific siRNA that Dvl2 was highly phosphorylated in FZD10-positive oligonucleotide into a synovial sarcoma cell line, SYO-1 cells. cells, and performed a RhoA activation assay. Suppres- Subsequently, to investigate a possible modification of sion of FZD10 expression resulted in a significant Dvl3, we treated cellular extracts from the FZD10-2 cell increase in RhoA activity as compared with the cells line with l-protein phosphatase and found the disap- transfected with a control siEGFP (Figure 2b). Further- pearance of Dvl3 upper bands, indicating that the Dvl3 more, we observed actin cytoskeleton in HS-SY-II protein was also phosphorylated in FZD10-positive cells synovial sarcoma cells treated with FZD10 siRNA (Figure 1b). Furthermore, disappearance of a non- (Figure 2c). These findings suggest that overexpression phosphorylated form of Dvl2 and enhancement of a of FZD10 suppressed the RhoA activation, although the phosphorylated form of Dvl3 were observed in synovial detailed mechanisms remain to be elucidated. sarcoma cell lines, SYO-1 and HS-SY-II, which showed a high level of endogenous FZD10 expression, whereas we found high levels of the non-phosphorylated forms Enhancement of anchorage-independent cell growth of Dvl2 and Dvl3 in FZD10-negative cell lines, 1273/99 by FZD10 through Rac1 and JNK activation and LoVo (Figure 1c). Small GTPase protein Rac1 is known to be an essential Phosphorylation of the Dvl family protein has been signal-transduction molecule in actin organization reported and a large number of kinases were implicated (Ridley et al., 1992) and in the mediation of the non- to be involved (Hino et al., 2003; Kinoshita et al., 2003; canonical WNT signaling (Habas et al., 2003). More- Wharton, 2003; Bryja et al., 2007a). To examine a kinase over, c-Jun N-terminal kinase (JNK) is a major effector possibly involved in the FZD10-induced phosphoryla- for an actin-related protein complex (Arp2/3) that tion of Dvl2 and Dvl3, FZD10-positive cell lines were regulates actin polymerization (Otto et al., 2000; Hamel

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1112

FZD10-2 FZD10 -1 Mock2 FZD10-2 Mock1

HA-FZD10 + PPase Dvl2 (polyclonal) -PPase P-Dvl3 Dvl3 Dvl2 (MAb10B5) Dvl3

Dvl3 -Actin -Actin SYO-1 HS-SY-II 1273/99 LoVo

Dvl2 (MAb10B5) DMSO D4476

Dvl3 Dvl2 WB

-Actin Dvl3

FZD10 β-Actin RT -Actin

Mock-2 FZD10-2 Mock Dvl2-myc Mock Dvl2-myc

α-myc α-myc IP P-Dvl3 α-Dvl3 Dvl3

α-myc

Input α-HA

α-Dvl3 P-Dvl3 Dvl3 Figure 1 Phosphorylation of Dvl2 and Dvl3 by FZD10 overexpression. (a) Western blot analysis of Dvl2 and Dvl3 proteins in 1273/ 99 cells in which HA-tagged FZD10 was stably expressed (HA-FZD10, 1273/99-FZD10-1 and 1273/99-FZD10-2) and those transfected with mock vectors (Mock1 and Mock2). Two different antibodies against Dvl2, a polyclonal antibody and a monoclonal antibody 10B5, were used in this study. b-Actin is served as an internal quantitative control. Arrows indicate the band shift of Dvl2 and Dvl3. (b) l-Protein phosphatase treatment of Dvl3. 1273/99-FZD10-2 cell lysates treated with l-protein phosphatase ( þ PPase) or sodium fluoride (ÀPPase) were analysed by western blot. The phosphorylated and unphosphorylated Dvl3 proteins are indicated as P- Dvl3 and Dvl3, respectively. b-Actin is served as an internal quantitative control. (c) Upper panels: western blot analysis (WB) of Dvl2 and Dvl3 proteins in synovial sarcoma cell lines, SYO-1, HS-SY-II, 1273/99, and a colon cancer cell line, LoVo. b-Actin is served as an internal quantitative control. Lower panels: Semiquantitative RT–PCR (RT) of FZD10 transcripts in the same cell lines as shown in upper panels. b-Actin is served as an internal control. (d) Treatment of D4476 (100 mM; a CK1 inhibitor) blocks the phosphorylation- dependent mobility shift of both Dvl2 and Dvl3 proteins in FZD10-positive SYO-1 cells. SYO-1 cells were treated with either D4476 (100 mM) or DMSO control for 1 h. The cells were lysed and a phosphorylation-dependent Dvl3 or Dvl2 band shift was examined by western blot. b-Actin is served as an internal quantitative control. (e) Co-immunoprecipitation of Dvl2 and Dvl3. 1273/99 stably FZD10-expressing clone (FZD10-2) or Mock control (Mock2) cells were transfected with either pcDNA3.1/myc-His (Mock) or pcDNA3.1/myc-His-Dvl2 (Dvl2-myc). Cells were lysed 48 h after transfection and subjected to immunoprecipitation with anti-c-myc antibody (9E10). Immunoprecipitates were analysed by western blot and bound endogenous Dvl3 was detected.

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1113 siEGFP siFZD10

FZD10 RT -Actin

Active RhoA WB Total RhoA i ii

i iii iv iv

v vi ii v

iii vi vii viii Figure 2 Disruption of actin cytoskeleton structure in FZD10-overexpressing cells, and restoration of RhoA activity and stress fiber in FZD10-knockdown cells. (a) HuBM-Bmi1-hT cells (i–iv) or 1273/99 cells (v–viii) were transfected with pCAGGS/neo HA FZD10 (ii, iv, vi and viii) or pCAGGS/neo empty vector (i, iii, v and vii). At 48 h after the transfection, HA-FZD10 was visualized with Alexa Fluor488-goat anti-rat IgG (green), and actin cytoskeleton was stained with Alexa Fluor594-Phalloidin (red). Nuclei were counterstained with 40,60-diamidine-20-phenylindole dihydrochloride (DAPI; blue). A scale bar indicates 10 mm. (b) Restoration of RhoA activation by knockdown of FZD10 in SYO-1 cells. Cells were transfected with siRNA targeting to either FZD10 (siFZD10-1) or EGFP (siEGFP). At 72 h after the transfection, RhoA assays were performed. b-Actin is served as an internal quantitative control for semi-quantitative RT–PCR. (c) HS-SY-II cells were transfected with siRNA targeting to either EGFP (i–iii) or FZD10 (iv–vi). At 72 h after the transfection, actin cytoskeleton was stained with Alexa Fluor594-Phalloidin (red). Nuclei were counterstained with 40,60-diamidine-20-phenylindole dihydrochloride (DAPI; blue). A scale bar indicates 10 mm. et al., 2006) and functions as a mediator in the Rac1 expression of HA-tagged FZD10 protein increased an signaling pathway (Coso et al., 1995; Minden et al., active GTP-bound form of Rac1 in 1273/99 cells, 1995; Rosso et al., 2005). Hence, we investigated the whereas the total amount of Rac1 remained unchanged effect of FZD10 on Rac1 and JNK pathways. Ectopic (Figure 3a). Conversely, knockdown of endogenous

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1114 FZD10 expression by siRNA in HS-SY-II cells com- We also investigated primary synovial sarcoma tissues pletely suppressed Rac1 activity, as compared with and found high levels of phosphorylation of Dvl2 and siEGFP-treated cells (Figure 3b). Moreover, the en- Dvl3 as well as enhancement of JNK phosphorylation dogenous JNK phosphorylation was increased in 1273/ by western blot (Figure 4a). The level of phosphorylated 99-FZD10-1 and 1273/99-FZD10-2 (FZD10-1 and JNK was correlated with the amount of FZD10 protein FZD10-2) cells, as compared with that in the cells in immunohistochemical analyses (Figure 4b). Taken transfected with the mock vector (Mock1) or the 1273/ together, overexpression of FZD10 is considered to 99 parental cell (Figure 3c). On the other hand, we cause activation of the Dvl–Rac1–JNK pathway in vivo confirmed that knockdown of FZD10 by siRNA in as well as in vitro. SYO-1 cells resulted in the suppression of endogenous JNK activity (P-JNK), as compared with siEGFP- treated cells (Figure 3d). These findings suggest that overexpression of FZD10 induced the activation of Discussion Rac1 and the phosphorylation of JNK. As the phosphorylation of Dvl proteins was regulated by We previously reported that FZD10, a member of the FZD10 expression (Figure 1), Dvls are suspected to be Frizzled family, was significantly transactivated in mediators for the Rac1–JNK activation, directly or synovial sarcoma (Nagayama et al., 2002), but little indirectly. Using siRNA, we showed that phosphoryla- was known about the mechanism it utilized to transmit tion of endogenous JNK induced by FZD10 over- oncogenic signals in synovial sarcoma cells. We first expression was completely inhibited in the cells in which examined the effect of FZD10 on nuclear b-catenin either Dvl2 or Dvl3 was depleted, whereas the FZD10 accumulation and Tcf/Lef reporter activity (the canoni- level was unchanged in these cells (Figures 3e and f). cal signaling pathway), but found no activation of them Taken together, these results strongly suggest that by FZD10 expression (data not shown). Hence, we FZD10 overexpression induces Rac1 and JNK activa- considered a possible involvement of FZD10 in the non- tion through phosphorylation of Dvl2 and Dvl3 canonical signaling and demonstrated in this study that proteins. Furthermore, as the Rac1 activation was overexpression of FZD10 enhanced the anchorage- indicated to be important for the inhibition of detach- independent cell growth through the activation of the ment-induced cell death (anoikis) (Frisch and Screaton, Rac1–JNK pathway as well as the destruction of the 2001; Cheng et al., 2004), we investigated the effect of actin stress fiber structure caused by downregulation of FZD10 on anchorage-independent cell growth. When the RhoA activity (Figure 5). Furthermore, we showed the cells were cultured in soft agar, the numbers of the remarkable increase of phosphorylation of both colonies in 1273/99-FZD10-1 and 1273/99-FZD10-2 Dvl2 and Dvl3 in the presence of FZD10. Among the were increased up to two-fold in comparison with the diverse signal transduction pathways mediated by Dvls, parental 1273/99 cells or those transfected with mock regulation of small GTPase activity and following control (Mock1) (Figure 3g). Conversely, HS-SY-II cells rearrangement of actin cytoskeleton is one of the most transfected with FZD10 siRNA showed significant important cellular processes, including development, suppression of cell growth in soft agar, compared with neurite retraction (Kishida et al., 2004), cell motility, the cells transfected with EGFP siRNA as a control migration (Endo et al., 2005), etc. As it has been reported (Figure 3h). These results suggest that overexpression of that several kinases for the phosphorylation of the Dvl FZD10 enhances anchorage-independent survival family are involved in response to ligand stimulation through activation of the Dvl2/3–Rac1–JNK pathway. (Hino et al., 2003; Kinoshita et al., 2003; Bryja et al.,

Figure 3 Activation of the Rac1–JNK signaling and the enhancement of anchorage-independent cell growth by FZD10. (a) Active Rac1 was pulled down with PAK-1 PBD agarose from 1273/99 cells in which HA-FZD10 was stably overexpressed (FZD10-1 and FZD10-2), those transfected with pCAGGS/neo vector (Mock1), or 1273/99 parental cell. (b) Downregulation of Rac1 activity by knockdown of FZD10 in HS-SY-II cells. The cells were transfected with siRNA targeting to either FZD10 (siFZD10-1) or EGFP (siEGFP). At 72 h after the transfection, Rac1 assays were performed. b-Actin is served as an internal quantitative control for semiquantitative RT–PCR. (c) An increase in JNK phosphorylation by FZD10 overexpression. FZD10-1, FZD10-2, Mock1 and 1273/ 99 parental cells were analysed by western blot analysis. (d) Inhibition of endogenous JNK phosphorylation in FZD10-siRNA transfected synovia sarcoma cells. SYO-1 cells were transfected with siRNA oligonucleotide targeting either EGFP (siEGFP) or FZD10 (siFZD10) using Lipofectamine RNAiMAX. b-Actin is served as an internal quantitative control for semi-quantitative RT–PCR. Total proteins were collected 72 h after transfection and subjected to western blot analysis. (e, f) Inhibition of JNK phosphorylation by depletion of Dvl2 (e) or Dvl3 (f) in HA-FZD10-overexpressing cells. HEK 293 cells were transiently transfected with a pCAGGS/neo mock vector (FZD10À) or a pCAGGS/neo HA FZD10 (FZD10 þ ) construct, and were subsequently treated with either a control siRNA against EGFP, siRNA against Dvl2 (e) or siRNA against Dvl3 (f). Western blot analysis was performed at 48 h after the transfection to detect the phosphorylation of JNK using specific phosphorylated-JNK antibody. A total amount of JNK was shown as a control. Phosphorylation of Dvl3 was shown by the mobility shift. (g) Anchorage-independent cell-growth. 1273/99- FZD10-1, 1273/99-FZD10-2, 1273/99-Mock1, or 1273/99 parental cells were seeded in 0.3% soft agar-containing medium, and layered on 0.5% hardened agar for 3 weeks. The cell cultures were stained and the numbers of colonies were counted. Means along with s.d. are shown (Student’s two-sided t-test: *P ¼ 0.017 and wP ¼ 0.044 vs parent 1273/99 control). (h) Anchorage-independent cell-growth of FZD10-knockdown cells. HS-SY-II cells transfected with siRNA targeting to EGFP (siEGFP) or FZD10 (siFZD10) were grown in 0.3% agar-containing medium layered on 0.5% agar for 3 weeks. The cell cultures were stained and the numbers of colonies were counted. Means along with s.d. are shown (Student’s two-sided t-test: wP ¼ 0.00013).

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1115 siEGFP siFZD10

FZD10 1273/99 parent Mock1 FZD10-1 FZD10-2 RT HA-FZD10 -Actin

Active Rac1 Active Rac1

Total Rac1 WB Total Rac1

SYO-1 siEGFP siFZD10

FZD10 1273/99 parent Mock1 FZD10-1 FZD10-2 RT -Actin HA-FZD10

p-JNK2 p-JNK2 p-JNK p-JNK p-JNK1 p-JNK1

JNK2 JNK2 Total JNK Total JNK JNK1 JNK1

Dvl2 siRNA --++ Dvl3 siRNA - -+ + FZD10 -+-+ FZD10 -+-+ Dvl3 Dvl2 (polyclonal)

HA-FZD10 HA-FZD10

p-JNK2 p-JNK2 p-JNK p-JNK p-JNK1 p-JNK1

JNK2 JNK2

Total JNK JNK1 Total JNK JNK1

1600 160 * 1400 † 140 1200 120

1000 100

800 80

600 60 † Colony number Colony number 400 40

200 20

0 0 Parent Mock1 FZD10-1 FZD10-2 siEGFP siFZD10 1273/99

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1116 FZD10 Phospho-JNK

1273/99 SS

FZD10 RT -Actin

Dvl2

Dvl3 Merge Phospho-JNK WB

Total JNK

-Actin

Figure 4 Phosphorylation of Dvl2 and Dvl3, and enhancement of JNK activation in primary synovial sarcoma tumors (SS). (a) Transcriptional levels of FZD10 were analysed by semi-quantitative RT–PCR (RT). b-Actin is served as an internal quantitative control for semiquantitative RT–PCR. Western blot (WB) analysis with anti-Dvl2, anti-Dvl3, anti-phospho-JNK and total JNK antibodies were performed. b-Actin is served as an internal quantitative control for western blot analysis. (b) Immunohistochemical analyses of synovial sarcoma tissue section using the anti-FZD10 monoclonal antibody, MAb 92-13 and phospho-JNK antibodies. Scale bar indicates 10 mm.

2007a), we screened a possible candidate kinase(s) by treating a variety of kinase inhibitors, and found that FZD10 the treatment of CK1-specific inhibitor, D4476, effec- tively suppressed the phosphorylation of Dvl2 and Dvl3. Involvement of CK1 in the Wnt-related signaling pathway has been implicated recently; CK1 enhances Wnt-3a-induced activation of the b-catenin signaling CK1 through Dvl-1 and Frat-1 binding (Hino et al., 2003). P P In addition, CK1 inhibitor, D4476, blocked the differenti- ation of primary dopaminergic neuronal precursor DVL3 DVL2 cells (Bryja et al., 2007a). Phosphorylated Dvl protein was reported to be involved in the full activation of b-catenin Rac1 indicated by active b-catenin-specific antibody (Bryja et al., 2007b). In our results, phosphorylated Dvl3 was RhoA JNK pulled down with an active Rac1 in the presence of FZD10 (Supplementary Figure S2). Therefore, it is suggested that Rac1-JNK activation under FZD10 Stress fiber Actin re-organization overexpression requires phosphorylation of both Dvl2 and Dvl3 by CK1, although a further detailed mechan- ism has to be elucidated. Anchorage-independent growth Rho GTPase family proteins are known to be the key regulators of actin cytoskeleton dynamics and to play METASTASIS critical roles in planar cell polarity/non-canonical WNT- Frizzled cascade (Paterson et al., 1990; Hall, 1998; Figure 5 Involvement of FZD10 in the regulation of actin cytoskeleton structure through the activation of Dvl2/Dvl3-Rac1- Wiggan and Hamel, 2002; Habas et al., 2003; Rosso JNK proteins in synovial sarcoma cells. et al., 2005; Wechezak and Coan, 2005). We also found

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1117 that depletion of endogenous FZD10 expression by with 30 units of DNase I (Roche, Basel, Switzerland) in the RNAi resulted in RhoA activation in synovial sarcoma, presence of 1 unit of RNase inhibitor (TOYOBO, Osaka, SYO-1 cells. Japan). After inactivation at 70 1C for 10 min, 3 mg each of Upregulation of JNK phosphorylation was observed RNA were reversely transcribed for single-stranded cDNAs in primary synovial sarcomas as well as in synovial using oligo (dT)12–18 primer and Superscript II (Invitrogen) at 42 1C for 60 min. We prepared appropriate dilutions of each sarcoma cell lines. The Rac1–JNK cascade was con- single-stranded cDNA for subsequent PCR by monitoring sidered to be activated in the presence of FZD10, but b-actin as a quantitative control. PCR amplification was this activation was blocked by the knockdown of either performed using EX-taq polymerase (TAKARA BIO, Ohtsu, Dvl3 or Dvl2 as well as by the knockdown of FZD10 Japan) and the cDNAs as templates with the following (Figure 3). We also observed that Rac1 activation by primers: 50-TATCGGGCTCTTCTCTGTGC-30 and 50-GA FZD10 possibly induced lamellipodia formation and CTGGGCAGGGATCTCATA-30 for FZD10; and 50-CATCC enhanced anchorage-independent growth. In addition, ACGAAACTACCTTCAACT-30 and 50-TCTCCTTAGAGA an active Cdc42 was also increased in FZD10-over- GAAGTGGGGTG-30 for b-actin. PCRs were optimized for expressing HEK293 cells (Supplementary Figure S3). the number of cycles to ensure product intensity within the Similarly to our findings, it was reported that Rho logarithmic phase of amplification. The PCR conditions were 95 1C for 30 s, 58 1C for 30 s and 72 1C for 30 s for 21 cycles for inactivation caused detachment from the substratum, b-actin and 28 cycles for FZD10. The PCR products were and transformed the adherent fibroblasts and Vero cells resolved by electrophoresis on 2.0% agarose gels, and their into a round shape, indicating that Rho and Rac have band intensities were quantified using NIH Image analysis opposing functions (Aepfelbacher et al., 1996). Taken software (http://rsb.info.nih.gov/nih-image/). together, our findings suggest that FZD10 has a critical role in the highly metastatic feature of synovial sarcoma. Western blot analysis In conclusion, we have shown that FZD10 is involved Cells and primary tumor samples were lysed with a lysis buffer in the development/progression of synovial sarcoma by (50 mM Tris-HCl, 0.5% Igepal, 0.5% Triton X-100, 2% glycerol, regulating actin reorganization and anchorage-indepen- 150 mM NaCl, 1 mM NaF, 1 mM sodium orthovanadate, pH 7.4) dent cell growth. Owing to the highly specific expression including 0.2% protease inhibitor cocktail III (Calbiochem, San in tumor and the involvement in tumor progression, Diego, CA, USA). After homogenization, the cell lysates were FZD10 is a promising molecular target to improve the incubated on ice for 30 min and centrifuged at 14 000 r.p.m. for prognosis of synovial sarcoma. 15 min to separate the supernatant from the cell debris. The amount of total protein was estimated by a protein assay kit (Bio-Rad, Hercules, CA, USA), and then the cell lysates were mixed with SDS sample buffer and boiled for 3 minutes before Materials and methods loading into SDS-polyacrylamide gels (Bio-Rad). After electro- phoresis, the proteins were blotted onto a nitrocellulose Cells lines and tissue specimens membrane (GE Healthcare, Buckinghamshire, UK). After A colon cancer cell line LoVo was purchased from American blocking with 4% BlockAce blocking solution (Dainippon Type Culture Collection (ATCC, Rockville, MD, USA). Pharmaceutical Co. Ltd, Osaka, Japan), the membranes were YaFuSS, derived from synovial sarcoma, and HuBM-Bmi1- incubated with the primary antibodies. Finally, membrane was hT, an immortalized mesenchymal stem cell line by transfec- incubated with horseradish peroxide-conjugated secondary anti- tion of Bmi and hTERT, were kindly provided by Professor body (1:10 000 dilution; GE Healthcare), and proteins were Toguchida (Faculty of Medicine, Kyoto University, Japan). visualized by the ECL detection reagent (GE Healthcare). b- SYO-1 was a gift from Dr A Kawai (National Cancer Center, Actin was used as a loading control. The primary antibodies used Japan), and 1273/99 from Dr O Larsson (Karolinska Institute, are as follows: b-Actin (1:30 000 dilution; clone AC-15, Sigma- Sweden). All cells were cultured under their respective Aldrich, St Louis, MO, USA), DVL2 (1:500 dilution; Dvl-2 H-75 depositors’ recommendation; that is, Ham’s F-12 nutrient or 10B5, Santa Cruz Biotechnology, Santa Cruz, CA, USA), mixture (Invitrogen, Carlsbad, CA, USA) for LoVo and 1273/ DVL3 (1:500 dilution; Dvl-3 4D3, Santa Cruz Biotechnology), 99 cell lines, Dulbecco’s modified Eagle’s medium (Invitrogen) HA (1:3000 dilution; anti-HA high affinity, clone 3F10; Roche), for HEK293, SYO-1, HS-SY-II, and HuBM-Bmi1-hT cell c-Myc (1:1000 dilution; 9E10, Santa Cruz Biotechnology), JNK lines. All cell lines were grown in monolayers in media (1:500 dilution; SAPK/JNK, Cell Signaling Technology, Dan- supplemented with 10% fetal bovine serum and 1% antibiotic/ vers, MA, USA), phosphorylated JNK (1:500 dilution; Phospho- antimycotic solution, and maintained at 37 1C in air containing SAPK/JNK Thr183/Tyr185, Cell Signaling Technology), RhoA 5% CO2. Primary synovial sarcoma samples were obtained (1:500 dilution; Santa Cruz Biotechnology), Rac1 (1:1000 with written informed consent, and snap-frozen in liquid dilution; anti-Rac1 monoclonal antibody, B&D Transduction, nitrogen immediately after resection and stored at À80 1C. The Lexington, KY, USA) and Cdc42 mouse monoclonal IgG3 use of these tissue samples was approved by the IRBs (Faculty (1:1000 dilution; Santa Cruz Biotechnology). of Medicine, Kyoto University and Institute of Medical Science, The University of Tokyo). Each tumor sample was l-Protein phosphatase assay fixed in 10% formalin and routinely processed for hematoxylin Cells were lysed with a lysis buffer (50 mM Tris-HCl, 0.5% and eosin staining to establish a pathological diagnosis by Dr Igepal, 0.5% Triton X-100, 2% glycerol, 150 mM NaCl, 0.2% J Toguchida. protease inhibitor cocktail Set III (Calbiochem), pH 7.4). Aliquots of 10 mg protein were supplemented with 2 mM Semiquantitative RT–PCR MnCl2, incubated with 800 units of l-protein phosphatase Total RNAs were extracted from each of the cell lines and (New England Biolabs, Beverly, MA, USA) or 25 mM of primary tumor samples using TRIzol reagent (Invitrogen). The sodium fluoride for 60 min at 30 1C. Incubation was termi- extracted RNAs were treated for 30 min at room temperature nated by the addition of SDS sample buffer and the samples

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1118 were boiled for 3 min, subjected to western blot analysis which a undetectable level of FZD10 expression was observed, according to the method described in the above section. using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s protocol. Transfected cells were incubated in the culture medium containing 0.2 mg/ml of neomycin Plasmid constructs (geneticin; Invitrogen). Three weeks later, 20 individual To construct the plasmid designed to express a full-length colonies were selected by limiting dilution and screened for FZD10 with HA-epitope tag at its N-terminus protein, we HA-FZD10 stably overexpressing clones. The expression level inserted an HA-tag sequence at the downstream of the signal of HA-FZD10 in each clone was examined by western blot and peptide sequence (1–22 residues of FZD10) because FZD10 is immunocytochemical staining analyses using an anti-HA a cytoplasmic membrane protein. Firstly, we generated a signal monoclonal antibody (Roche). We established two indepen- peptide sequence fragment of FZD10 protein (1–22 residues) dent clones with FZD10 overexpression and designated them by PCR, using PrimeSTAR HS DNA polymerase (TAKARA as 1273/99-FZD10-1 and 1273/99-FZD10-2. Two independent BIO) with the following set of primers: 50-AAGTCGACGC 0 0 cell lines transfected with mock vector were designated as CAGCATGCAGCGCCCG-3 and 5 -AACTCGAGGCT Mock1 and Mock2. Treatment of SYO-1 cells with D4476 was GATGGCGGCGCACGAG-30 (underline indicates restriction performed as follows: 10 ml of 10 mM D4476 was diluted in a Sal enzyme sites). The PCR product was digested with I and mixture of 100 ml serum-free medium and 3 ml of FuGENE 6 Xho XhoI I, and ligated into the site of pCAGGS/neovector transfection reagent (Roche) at room temperature before (pCAGGS/neo signal). Secondly, to generate the HA epitope adding this solution to the cultured cells. tag, a following set of oligonucleotides, 50-GCATGTCGAC TACCCATACGATGTTCCAGATTACGCTCTCGAGGCA T-30 and 50-ATGCCTCGAGAGCGTAATCTGGAACA Immunocytochemical and immunohistochemical staining TCGTATGGGTAGTCGACATGC-30 (underline indicates analyses restriction enzyme sites) was mixed, and denatured at 94 1C For immunocytochemical analysis, cells were plated onto a for 3 min and gradually cooled down to 37 1C (0.1 1C/s slope) COL I-coated cover glass (IWAKI, Tokyo, Japan) placed in for annealing. The annealed fragment was digested with SalI a six-well plate at a density of 5 Â 104 cells/well. Subsequently, and XhoI, and ligated into the XhoI site of pCAGGS/ the cells were fixed with phosphate-buffered saline (Sigma- neo signal (pCAGGS/neo signal HA). Thirdly, to amplify Aldrich) containing 4% paraformaldehyde for 15 min at room the cDNA fragment consisting of residues 23–581 of FZD10, temperature, and then permeabilized with 0.1% Triton X-100 we performed PCR using the following primers: 50-AA for 2 min at room temperature, followed by incubation with an GTCGACTCCATGGACATGGAGCGCCC-30 and 50-AA anti-HA monoclonal antibody (Roche). Bound antibody was CTCGAGTCACACGCAGGTGGGCGACT-30 (underline detected using goat anti-rat IgG (H þ L) conjugated with indicates restriction enzyme sites). The PCR fragment of Alexa Fluor 488 (1:1000 dilution; Molecular Probe, Eugene, FZD10 (23–581) was digested with SalI and XhoI and ligated OR, USA). Actin cytoskeleton was stained using phalloidin into the XhoI site of pCAGGS/neo signal HA vector conjugated with Alexa Fluor 594 (1:40 dilutions; Molecular (pCAGGS/neo HA FZD10). Probe). To construct the Dvl2 expression vector, the entire coding For immunohistochemical analysis, sections of 4-mm-thick sequence of Dvl2 was amplified by PCR using the following paraffin-embedded normal adult human tissues (lung, heart, primers: 50-AAGGTACCATGGCGGGTAGCAGCAC-30 liver, kidney, colon and placenta) (BioChain, Hayward, CA, and 50-AAGGATCCCATAACATCCACAAAGAACTCGC- USA) and clinical synovial sarcoma specimens were depar- 30 (underline indicates restriction enzyme sites). The PCR affinized and processed for antigen retrieval in Target fragment was digested with KpnI and BamHI, and ligated into Retrieval Solution (pH 9) (DAKO Cytomation, Carpinteria, pcDNA 3.1/myc-His (Invitrogen). DNA sequences of all CA, USA) at 125 1C for 30 s. After treatment with serum-free constructs were confirmed by DNA sequencing (ABI3700, protein blocking (DAKO Cytomation) for 30 min, slides were PE Applied Biosystems, Foster, CA, USA). incubated with 10 mg/ml of mouse anti-FZD10 monoclonal We used siRNA oligonucleotides (Sigma-Aldrich Japan antibody 92-13 (Fukukawa et al., 2008) and 1:50 dilution of KK, Tokyo, Japan) to further verify the knockdown effects of phospho-SAPK/JNK (Thr183/Tyr185) antibody (Cell Signal- FZD10, Dvl2 and Dvl3 on signaling pathways. The sequences ing Technology). Bound antibodies were detected using Alexa targeting each gene were as follows: 50-GAAGCAGCACGA Fluor 488 goat anti-mouse IgG (H þ L) and Alexa Fluor 594 CUUCUUC-30 (sense) and 50-GAAGAAGUCGUGCUGCU goat anti-rabbit IgG (H þ L) (Molecular Probe). Fluorescent UC-30 (antisense) for siEGFP (control); 50-GCCGUAG images were obtained with a TCS SP2 AOBS microscope GUUAAAGAAGAA-30 (sense) and 50-UUCUUCUUUAAC (Leica, Tokyo, Japan). CUACGGC-30 (antisense) for siFZD10; 50-GAUUACC AUCCCUAAUGCU-30 (sense) and 50-AGCAUUAGGGA RhoA and Rac1 activation assays UGGUAAUC-30 (antisense) for siDvl3; and 50-GCAAGU Activated forms of RhoA and Rac1 were detected using the GGGACUAGCGAUG-30 (sense) and 50-CAUCGCUA RhoA assay reagent (Rhotekin PBD, agarose conjugate; GUCCCACUUGC (antisense) for siDvl2. Upstate Biotechnology, Lake Placid, NY, USA) and the Rac/cdc42 assay reagent (PAK-1 PBD, agarose conjugate; Cell transfection and treatments Upstate biotechnology), respectively, according to the manu- HEK293, COS-7 and HuBM-Bmi1-hT cells were transfected facturer’s instructions. For RhoA assay, SYO-1 cells were with expression vector constructs as described above using transfected with siRNA oligonucleotides against either FZD10 FuGENE 6 transfection reagent (Roche). SYO-1 and HS-SY- or EGFP as a control. The cells were cultured for 72 h after II cells were transfected with siRNA oligonucleotides using transfection and subjected to the assay. For Rac1 and cdc42 Lipofectamine RNAiMAX reagent (Invitrogen) according to assays, 1273/99 or HEK293 cells were transfected with the manufacturer’s protocol. For the establishment of cell lines pCAGGS-HA-FZD10 vector or pCAGGS mock vector. In that stably overexpress FZD10, HA-tagged FZD10 expression addition, HS-SY-II cells were also transfected with siRNA vector (pCAGGS/neo HA FZD10) or mock vector (pCAGGS/ oligonucleotides against either FZD10 or EGFP as a control. neo) was transfected into 1273/99 synovial sarcoma cells, in Subsequently, for both assays, the cells were lysed in an MLB

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1119 buffer (25 mM HEPES (4-(2-hydroxyethyl)-1-piperazineetha- His-Dvl2 or an empty vector using FuGENE6 transfection nesulfonic acid), pH 7.5, 150 mM NaCl, 0.5% Igepal, 10% reagent (Roche). Transfected cells were lysed with the lysis glycerol, 10 mM MgCl2,1mM EDTA, 25 mM NaF, 1 mM buffer described above. The cell lysates were incubated with sodium orthovanadate and 0.1% protease inhibitor cocktail anti-c-myc antibody (9E10) and protein G-sepharose (Sigma- III). Appropriate amounts of RhoA assay or Rac/cdc42 assay Aldrich). Bound proteins were washed five times with the lysis reagent were mixed with the lysates and incubated at 4 1C for buffer and eluted with the SDS sample buffer. Eluted proteins 60 min. The conjugates were washed three times with the MLB were analysed by SDS–polyacrylamide gel electrophoresis and buffer and subjected to western blot analysis according to the western blot analysis. method described in the western blot analysis section.

Anchorage-independent growth assay Statistical analysis Anchorage-independent growth was examined by colony Data points for the anchorage-independent growth assay formation assay in soft agar. A cell suspension consisting of represent the means along with 95% confidence intervals (CIs) 1 Â 104 viable 1273/99 cells (stably overexpressing HA-FZD10, of two independent experiments. Student’s t-test was applied cells transfacted with a mock vector or parent cells) or HS-SY-II for comparisons of colony numbers in anchorage-independent cells, transfected with siRNA oligonucleotides against either growth assays. All statistical tests were two-sided, and FZD10 or EGFP in appropriate media containing 0.3% agar, P-values o0.05 were considered to be statistically significant. was plated onto a 3-ml base layer of 0.5% agar with the same media composition. Cultures were allowed to grow for 3 weeks, and colonies were stained with 4 mg/ml of iodonitrote- Acknowledgements trazolium chloride. We thank Dr Ryo Takata for statistical analysis, and Ms Immunoprecipitation Kie Naito, Ms Yoshiko Fujisawa, Ms Akiko Konuma, Ms 1273/99 stably FZD10-expressing clone (FZD10-2) or Mock Aya Sasaki and Ms Kyoko Kijima for excellent technical control (Mock2) were transfected with either pcDNA 3.1/myc- assistance.

References

Aepfelbacher M, Essler M, Huber E, Czech A, Weber PC. (1996). Rho Endo Y, Wolf V, Muraiso K, Kamijo K, Soon L, u¨ ren A et al. is a negative regulator of human monocyte spreading. J Immunol (2005). Wnt-3a-dependent cell motility involves RhoA activation 157: 5070–5075. and is specifically regulated by Dishevelled-2. J Biol Chem 280: Angers S, Thorpe CJ, Biechele TL, Goldenberg SJ, Zheng N, MacCoss 777–786. MJ et al. (2006). The KLHL12-Cullin-3 ubiquitin ligase negatively Frisch SM, Screaton RA. (2001). Anoikis mechanisms. Curr Opin Cell regulates the Wnt-b–catenin pathway by targeting Dishevelled for Biol 13: 555–562. degradation. Nat Cell Biol 8: 348–357. Fukukawa C, Hanaoka H, Nagayama S, Tsunoda T, Toguchida J, Bain J, Plater L, Elliott M, Shpiro N, Hastie CJ, McLauchlan H et al. Endo K et al. (2008). Radioimmunotherapy of human synovial (2007). The selectivity of protein kinase inhibitors: a further update. sarcoma using a monoclonal antibody against FZD10. Cancer Sci Biochem J 408: 297–315. 99: 432–440. Brett D, Whitehouse S, Antonson P, Shipley J, Cooper C, Goodwin G. Gonza´ lez-Sancho JM, Brennan KR, Castelo-Soccio LA, Brown AM. (1997). The SYT protein involved in the t(X;18) synovial sarcoma (2004). Wnt proteins induce Dishevelled phosphorylation via an translocation is a transcriptional activator localised in nuclear LRP5/6-independent mechanism, irrespective of their ability to bodies. Hum Mol Genet 6: 1559–1564. stabilize b-catenin. Mol Cell Biol 24: 4757–4768. Bryja V, Schulte G, Arenas E. (2007a). Wnt-3a utilizes a novel low Habas R, Dawid IB, He X. (2003). Coactivation of Rac and Rho by dose and rapid pathway that does not require casein kinase 1- Wnt/Frizzled signaling is required for vertebrate gastrulation. Genes mediated phosphorylation of Dvl to activate b-catenin. Cell Signal Dev 17: 295–309. 19: 610–616. Habu S, Okamoto E, Toyosaka A, Nakai Y, Takeuchi M. (1998). Bryja V, Schulte G, Rawal N, Grahn A, Arenas E. (2007b). Synovial sarcoma of the esophagus: report of a case. Surg Today 28: Wnt-5a induces Dishevelled phosphorylation and dopaminergic 401–404. differentiation via a CK1-dependent mechanism. J Cell Sci 120: Hall A. (1998). Rho GTPases and the actin cytoskeleton. Science 279: 586–595. 509–514. Capelluto DG, Kutateladze TG, Habas R, Finkielstein CV, Hamel M, Kanyi D, Cipolle MD, Lowe-Krentz L. (2006). Active stress He X, Overduin M. (2002). The DIX domain targets disheveled kinases in proliferating endothelial cells associated with cytoskeletal to actin stress fibres and vesicular membranes. Nature 419: structures. Endothelium 13: 157–170. 726–729. Hino S, Michiue T, Asashima M, Kikuchi A. (2003). Casein kinase Ie Cheng TL, Symons M, Jou TS. (2004). Regulation of anoikis by Cdc42 enhances the binding of Dvl-1 to Frat-1 and is essential and Rac1. Exp Cell Res 295: 497–511. for Wnt-3a-induced accumulation of b-catenin. J Biol Chem 278: Clark J, Rocques PJ, Crew AJ, Gill S, Shipley J, Chan AM et al. 14066–14073. (1994). Identification of novel genes, SYT and SSX, involved in the Hiraga H, Nojima T, Isu K, Yamashiro K, Yamawaki S, Nagashima t(X;18)(p11.2;q11.2) translocation found in human synovial sarco- K. (1999). Histological and molecular evidence of synovial sarcoma ma. Nat Genet 7: 502–508. of bone. A case report. J Bone Joint Surg Am 81: 558–563. Coindre JM, Pelmus M, Hostein I, Lussan C, Bui BN, Guillou L. Hosono T, Hironaka M, Kobayashi A, Yamasawa H, Bando M, Ohno (2003). Should molecular testing be required for diagnosing S et al. (2005). Primary pulmonary synovial sarcoma confirmed by synovial sarcoma? A prospective study of 204 cases. Cancer 98: molecular detection of SYT-SSX1 fusion gene transcripts: a case 2700–2707. report and review of the literature. Jpn J Clin Oncol 35: 274–279. Coso OA, Chiariello M, Yu JC, Teramoto H, Crespo P, Xu N et al. Kinoshita N, Iioka H, Miyakoshi A, Ueno N. (2003). PKCd is (1995). The small GTP-binding proteins Rac1 and Cdc42 essential for Dishevelled function in a noncanonical Wnt pathway regulate the activity of the JNK/SAPK signaling pathway. Cell 81: that regulates Xenopus convergent extension movements. Genes Dev 1137–1146. 17: 1663–1676.

Oncogene FZD10 signaling in synovial sarcoma cells C Fukukawa et al 1120 Kishida S, Yamamoto H, Kikuchi A. (2004). Wnt-3a and Dvl induce Rosso SB, Sussman D, Wynshaw-Boris A, Salinas PC. (2005). Wnt neurite retraction by activating Rho-associated kinase. Mol Cell Biol signaling through Dishevelled, Rac and JNK regulates dendritic 24: 4487–4501. development. Nat Neurosci 8: 34–42. Koike J, Takagi A, Miwa T, Hirai M, Terada M, Katoh M. (1999). Rothba¨ cher U, Laurent MN, Deardorff MA, Klein PS, Cho KW, Molecular cloning of Frizzled-10, a novel member of the Frizzled Fraser SE. (2000). Dishevelled phosphorylation, subcellular locali- gene family. Biochem Biophys Res Commun 262: 39–43. zation and multimerization regulate its role in early embryogenesis. Lim FL, Soulez M, Koczan D, Thiesen HJ, Knight JC. (1998). A EMBO J 19: 1010–1022. KRAB-related domain and a novel transcription repression domain Shannon BA, Murch A, Cohen RJ. (2005). Primary renal synovial in proteins encoded by SSX genes that are disrupted in human sarcoma confirmed by cytogenetic analysis: a lesion distinct sarcomas. Oncogene 17: 2013–2018. from sarcomatoid renal cell carcinoma. Arch Pathol Lab Med 129: McGilbray TT, Schulz TK. (2003). Clinical picture: primary cardiac 238–240. synovial sarcoma. Lancet Oncol 4: 283. Szuhai K, Knijnenburg J, Ijszenga M, Tanke HJ, Baatenburg de Jong Minden A, Lin A, Claret FX, Abo A, Karin M. (1995). Selective RJ, Bas Douwes Dekker P et al. (2004). Multicolor fluorescence in activation of the JNK signaling cascade and c-Jun transcription situ hybridization analysis of a synovial sarcoma of the larynx with a activity by the small GTPases Rac and Cdc42Hs. Cell 81: t(X;18)(p11.2;q11.2) and trisomies 2 and 8. Cancer Genet Cytogenet 1147–1157. 153: 48–52. Nagayama S, Fukukawa C, Katagiri T, Okamoto T, Aoyama T, Segal NH, Pavlidis P, Antonescu CR, Maki RG, Noble WS, Oyaizu N et al. (2005). Therapeutic potential of antibodies against DeSantis D et al. (2003). Classification and subtype prediction of FZD10, a cell-surface protein, for synovial sarcomas. Oncogene 24: adult soft tissue sarcoma by functional genomics. Am J Pathol 163: 6201–6212. 691–700. Nagayama S, Katagiri T, Tsunoda T, Hosaka T, Nakashima Y, Araki Soulez M, Saurin AJ, Freemont PS, Knight JC. (1999). SSX and the N et al. (2002). Genome-wide analysis of gene expression in synovial synovial-sarcoma-specific chimaeric protein SYT-SSX co-localize sarcomas using a cDNA microarray. Cancer Res 62: 5859–5866. with the human Polycomb group complex. Oncogene 18: 2739–2746. Nielsen TO, West RB, Linn SC, Alter O, Knowling MA, O’Connell JX Turc-Carel C, Dal Cin P, Limon J, Li F, Sandberg AA. (1986). et al. (2002). Molecular characterisation of soft tissue tumours: a Translocation X;18 in synovial sarcoma. Cancer Genet Cytogenet gene expression study. Lancet 359: 1301–1307. 23:93. Nunnally AP, Parr BA. (2004). Analysis of Fz10 expression in mouse Wang Z, Shu W, Lu MM, Morrisey EE. (2005). Wnt7b activates embryos. Dev Genes Evol 214: 144–148. canonical signaling in epithelial and vascular smooth muscle cells Ochi N, Uozumi M, Doi K, Tanimoto H, Ohbayashi C, Inagaki H through interactions with Fzd1, Fzd10, and LRP5. Mol Cell Biol 25: et al. (2004). Synovial sarcoma of the neck. ORL J Otorhinolaryngol 5022–5030. Relat Spec 66: 42–45. Wechezak AR, Coan DE. (2005). Dvl2 silencing in postdevelopmental Otto IM, Raabe T, Rennefahrt UE, Bork P, Rapp UR, Kerkhoff E. cells results in aberrant cell membrane activity and actin disorga- (2000). The p150-Spir protein provides a link between c-Jun nization. J Cell Physiol 202: 867–873. N-terminal kinase function and actin reorganization. Curr Biol 10: Weiss SW, Goldblum JR. (2001). Enzinger and Weiss’s Soft Tissue 345–348. Tumors, 4th edn, Mosby: St Louis. Paterson HF, Self AJ, Garrett MD, Just I, Aktories K, Hall A. (1990). Wharton Jr KA. (2003). Runnin’ with the Dvl: proteins that associate Microinjection of recombinant p21rho induces rapid changes in cell with Dsh/Dvl and their significance to Wnt signal transduction. Dev morphology. J Cell Biol 111: 1001–1007. Biol 253: 1–17. Rena G, Bain J, Elliott M, Cohen P. (2004). D4476, a cell-permeant Wiggan O, Hamel PA. (2002). Pax3 regulates morphogenetic cell inhibitor of CK1, suppresses the site-specific phosphorylation and behavior in vitro coincident with activation of a PCP/non-canonical nuclear exclusion of FOXO1a. EMBO R 5: 60–65. Wnt-signaling cascade. J Cell Sci 115: 531–541. Ridley AJ, Paterson HF, Johnston CL, Diekmann D, Hall A. (1992). Williams DH, Hua VN, Chowdhry AA, Laskin WB, Kalapurakal JA, The small GTP-binding protein rac regulates growth factor-induced Dumanian GA et al. (2004). Synovial sarcoma of the prostate. membrane ruffling. Cell 70: 401–410. J Urol 171: 2376.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene