Disruption of Fibroblast Growth Factor Signal

Cancer Therapy: Preclinical

Disruption of Fibroblast Growth Factor Signal Pathway Inhibits the Growth of Synovial Sarcomas: Potential Application of Signal Inhibitors to MolecularTarget Therapy Ta t s u y a I s hi b e , 1, 2 Tomitaka Nakayama,2 Ta k e s h i O k a m o t o, 1, 2 Tomoki Aoyama,1Koichi Nishijo,1, 2 Kotaro Roberts Shibata,1, 2 Ya s u ko Shim a ,1, 2 Satoshi Nagayama,3 Toyomasa Katagiri,4 Yusuke Nakamura, 4 Takashi Nakamura,2 andJunya Toguchida 1

Abstract Purpose: Synovial sarcoma is a soft tissue sarcoma, the growth regulatory mechanisms of which are unknown.We investigatedthe involvement of fibroblast growth factor (FGF) signals in synovial sarcoma andevaluatedthe therapeutic effect of inhibiting the FGF signal. Experimental Design:The expression of 22 FGF and4 FGF receptor (FGFR) genes in18prima- ry tumors andfive cell lines of synovial sarcoma were analyzedby reverse transcription-PCR. Effects of recombinant FGF2, FGF8, andFGF18 for the activation of mitogen-activatedprotein kinase (MAPK) andthe growth of synovial sarcoma cell lines were analyzed.Growth inhibitory effects of FGFR inhibitors on synovial sarcoma cell lines were investigated in vitro and in vivo. Results: Synovial sarcoma cell lines expressedmultiple FGF genes especially those expressed in neural tissues, among which FGF8 showedgrowth stimulatory effects in all synovial sarcoma cell lines. FGF signals in synovial sarcoma induced the phosphorylation of extracellular signal ^ regulatedkinase (ERK1/2) andp38MAPK but not c-Jun NH 2-terminal kinase. Disruption of the FGF signaling pathway in synovial sarcoma by specific inhibitors of FGFR causedcell cycle arrest leading to significant growth inhibition both in vitro and in vivo.Growthinhibitionbythe FGFR inhibitor was associatedwith a down-regulation of phosphorylatedERK1/2 but not p38MAPK, andan ERK kinase inhibitor also showedgrowth inhibitory effects for synovial sarcoma, indicating that the growth stimulatory effect of FGF was transmitted through the ERK1/2. Conclusions: FGF signals have an important role in the growth of synovial sarcoma, andinhibi- tory molecules will be of potential use for molecular target therapy in synovial sarcoma.

Synovial sarcoma is the most frequent soft-tissue sarcoma chemotherapy is still a matter of debate, and the development (STS) among patients in the third to fourth decade of life (1) of a new therapeutic approach is required to improve the and accounts for about 7% to 10% of all human STSs (2). It prognosis. predominantly affects the lower extremities but can occur in Despite little progress in clinical treatment during the last any part of the body. Surgical resection with an adequate 20 years, cytogenetic and molecular genetic analyses have greatly surgical margin is the definitive choice of treatment for improved the understanding of this type of tumor, especially primary tumors and has been shown to control local with the discovery of the reciprocal translocation recurrence (3, 4). Disseminated distant metastasis is the major t(18;X)(q11;p11) creating an SYT-SSX fusion gene as a synovial cause of poor outcome, and several reports describing the sarcoma–specific genetic alteration (5, 6). Thus far, three SSX results of current therapy showed a 5-year survival rate of genes (SSX1, SSX2, and SSX4) have been identified as a partner around 50% to 60% (3, 4). The efficacy of adjuvant of the SYT gene, and >95% of synovial sarcoma tumors carried one of these fusion genes (7). Although the precise function and the mechanism of oncogenesis are not yet clearly shown, the Authors’ Affiliations: 1Institute for Frontier Medical Sciences, Departments of high sensitivity and specificity of the SYT-SSX fusion gene in 2 3 Orthopaedic Surgery, Surgery Surgical Basic Science, Graduate School of Medi- synovial sarcoma have proven to be useful for molecular cine, Kyoto University, Kyoto, Japan; and 4Institute of Medical Science, University o f To k y o , To k y o , Ja p a n diagnosis (8). In particular types of tumors with a specific Received10/7/04; revised1/4/05; accepted1/13/05. reciprocal translocation such as PML-RARa in acute promyelo- Grant support: Ministry of Education, Culture, Sports, Science, and Technology cytic leukemia (9) and BCR-ABL in chronic myelogenous scientific research grant 11177101 (J.Toguchida). leukemia (10), fusion gene products themselves serve as targets 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 accordance for the therapy. Although immunotherapy using a peptide with 18 U.S.C. Section 1734 solely to indicate this fact. derived from SYT-SSX protein as a specific vaccine has been Requests for reprints: Junya Toguchida, Department of Tissue Regeneration, investigated (11), no therapeutic approach has been discovered Institute for Frontier Medical Sciences, Kyoto University, 53 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto, 606-8507, Japan. Phone: 81-75-751-4134; Fax: 81-75- for directly targeting the fusion protein or its function. 751-4646; E-mail: [email protected]. Gene expression profiling of tumors has been shown a F 2005 American Association for Cancer Research. powerful tool with which to isolate a molecular target for

Clin Cancer Res 2005;11(7) April 1, 2005 2702 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Fibroblast Growth Factors in Synovial Sarcoma therapy (12). We have done a gene expression analysis of Reverse transcription-PCR. Total RNA was extracted using TRIzol synovial sarcoma using a genome-wide cDNA microarray and reagent (Invitrogen) following the manufacturer’s instructions. After found that synovial sarcoma shared its molecular signature with treatment with DNase I (Nippon Gene, Osaka, Japan), 1 Ag of total malignant peripheral nerve sheath tumor, of which the RNA was reverse transcribed for single-stranded cDNAs using oligo(dT) primer and Superscript II reverse transcriptase (Invitrogen). PCR was precursors were neural crest–derived cells, and also identified done using 1 AL of RT product in a final volume of 25 AL containing a set of genes commonly up-regulated in synovial sarcoma 20 pmol each of the sense and antisense primers, 2.5 mmol/L MgCl2, including the fibroblast growth factor 18 (FGF18) gene (13). The 0.2 mmol/L of each deoxynucleotide triphosphate, and 1 unit of rTaq FGF signaling pathway seems to play significant roles in tumor polymerase (TOYOBO, Osaka, Japan). All PCRs were done using development and progression (14–16), and recently FGF18 was GeneAmp 9700 (PE Applied Biosystems, Foster City, CA). Information identified as an autocrine growth factor involved in colon of the primers are available upon request. cancers (17). Thus far, 22 genes have been identified as members Quantitative reverse transcription-PCR (RT-PCR) analyses were done of the FGF family, and our cDNA microarray contained 10 FGF in selected FGF and FGFR genes with ABI PRISM 7700 Sequence genes (FGF2, FGF3, FGF4, FGF7, FGF9, FGF11, FGF12, FGF13, Detection System (PE Applied Biosystems). One microliter of RT FGF18, and FGF19). We found that some FGF genes other product in a final volume of 25 AL containing 12.5 ALof2Â SYBR than FGF18 were also expressed in synovial sarcoma (data Green Mastermix (PE Applied Biosystems). Information for primers are available upon requests. The mean of triplicated data was used to not shown). Based on these results, we have done an calculate the ratio of target gene/b-actin expression. Statistical analysis intensive analysis of FGF and its receptor (FGFR) genes in was done by t test after logarithmic transformation. synovial sarcoma and also investigated whether inhibition of Western blot analyses. To detect FGF18 protein in the supernatant, the FGF signal is a new therapeutic modality for synovial cells were cultured up to 80% confluency in 100-mm dish. After sarcoma. washing the dish, cells were further incubated in 5 mL OPTI-MEM I for 4 days. The supernatant was harvested, mixed with 8 mL ice-cold Materials and Methods acetone, and kept at À80jC for 1 hour. The mixture was centrifuged at 10,000 Â g for 15 minutes, and the precipitate was suspended by lysis Tissue samples and cell lines. Tumor tissues were obtained at either buffer (100 AL) containing aprotinin (1 Ag/mL), leupeptin (1 Ag/mL), biopsy or resection surgery and kept at À80jC. Informed consent was pepstatin A (1 Ag/mL), and phenylmethylsulfonyl fluoride (1 mmol/L) obtained from each patient, and tumor samples were approved for followed by sonication and centrifugation. Twenty microliters of the analysis by the Ethics Committee of the Faculty of Medicine, Kyoto supernatant were electrophoresed on a 10% SDS-polyacrylamide gel University. Five human synovial sarcoma cell lines (YaFuSS, HS-SY-II, and transferred onto a polyvinylidene difluoride membrane (Millipore, SYO-1, Fuji, and 1273/99) were used in this study. YaFuSS and HS-SY-II Bedford, MA). After blocking with 3% skim milk, membranes were cells have the SYT-SSX1 fusion gene and the others have the SYT-SSX2 probed with anti-FGF18 antibody at 1:1,000 dilutions for 1 hour and fusion gene (data not shown). YaFuSS was established in our laboratory with peroxidase-conjugated anti-mouse IgG 1:2,000 for 1 hour. from a monophasic synovial sarcoma in a 28-year-old male. HS-SY-II Immunoreactive bands were detected with Enhanced Chemilumines- was a gift from H. Sonobe (Kochi University, Japan; ref. 18), SYO-1 cence Plus (Amersham Biosciences, Little Chalfont Buckinghamshire, from A. Kawai (Okayama University, Japan; ref. 19), Fuji from S. United Kingdom). Tanaka (Hokkaido University, Japan; ref. 20), and 1273/99 from To detect FGFR3 protein in cell lysate, cells were harvested by the O. Larsson (Karolinska Institute, Sweden). Among control cell lines, same lysis buffer mentioned above. The lysate (40 Ag) was electro- NMS-2 (malignant peripheral nerve sheath tumor; ref. 21) was phoresed using 8% SDS-polyacrylamide gel, blotted, and blocked by provided by A. Ogose (Niigata University, Japan), and Saos2 5% skim milk. The membrane was probed with anti-FGFR3 antibody at (osteosarcoma), HT1080 (fibrosarcoma), COLO205 (colon cancer), 1:500 dilutions and with peroxidase-conjugated anti-rabbit immuno- and SW480 (colon cancer) cells were purchased from American Type globulin at 1:2,000. Culture Collection (Manassas, VA). Cells were maintained in RPMI Phosphorylation analyses. To evaluate the phosphorylation status 1640 (Sigma-Aldrich, St. Louis, MI) for Fuji, SW480, and COLO205 of FGFR3, serum-starved cells were treated by 20 Amol/L SU5402 for with 10% fetal bovine serum (FBS, HyClone, Road Logan, UT) or 30 minutes followed by the treatment with rhFGF18 or vehicle. The DMEM (Sigma, St. Louis, MO) for other cells with 10% FBS. OPTI- cells were harvested by 1 mL modified radioimmunoprecipitation assay MEM I, which contains insulin and transferrin as protein supplements buffer containing protease inhibitors and sodium orthovanadate, (Invitrogen, Carlsberg, CA), was used in serum-free cultures. sonicated, and centrifuged at 15,000 Â g for 10 minutes. After Reagents and antibodies. The FGFR-specific tyrosine kinase preclearing by 100 AL protein A agarose (Upstate, Lake Placid, NY), inhibitor SU5402 was purchased from Calbiochem (La Jolla, CA; 1 mg of the protein was added with 4 Ag anti-FGFR3 antibody and ref. 22). PD166866 was provided by Pfizer Global Research and rotated for 2 hours at 4jC, and the immunocomplex was mixed with Development (Groton, CT; ref. 23). The MAPK/extracellular signal– 100 AL protein A agarose and rotated for another 2 hours. The mixture regulated kinase (ERK) kinase (MEK) inhibitor U0126 was purchased was washed twice by PBS, and the precipitate was boiled for 5 minutes from Promega (Madison, WI; ref. 24). Recombinant human FGF18 with 60 ALof2Â sample buffer, centrifuged, and the supernatant was (rhFGF18) was purchased from Wako Pure Chemical Industry (Osaka, used for Western blotting using anti-FGFR3 antibody (1:500) or anti- Japan), rhFGF8 from PeproTech (London, United Kingdom), phosphotyrosine (1:500). and rhFGF2 from Oncogene Research Products (San Diego, CA). To evaluate the phosphorylation of MAPKs, cells (4 Â 105) were Anti-ERK1 (M12320), anti-phosphorylated-p38 (P39520), anti-pan- seeded on 60-mm dishes and incubated overnight with medium p38 (P19820), anti-phosphorylated-JNK1/SAPK1 (S37220), and anti- containing 10% FBS followed by serum starvation for 24 hours. After pan-JNK1/SAPK1 (M54920) antibodies were purchased from treatment with each reagent for 15 minutes in the serum-free medium, BD Biosciences PharMingen (San Diego, CA); anti-phosphorylated- cells were lysed in a buffer (100 AL) containing sodium orthovanadate ERK1/2 (sc-7383), anti-FGFR3 (sc-123), and peroxidase-conjugated (5 mmol/L) and protease inhibitors, sonicated, and centrifuged. anti-mouse IgG (sc-2005) from Santa Cruz Biotechnology (Santa Proteins (15 Ag) were used for Western blotting. Cruz, CA); anti-phosphotyrosine (PY20) from Zymed Laboratories Bromodeoxyuridine incorporation assay. Cells (4 Â 103) were (South San Francisco, CA); anti-FGF18 (MAB667) from R&D Systems seeded on 96-well plates and incubated overnight with the medium (Minneapolis, MN); peroxidase-conjugated anti-rabbitimmunoglobu- containing 10% FBS. Then the medium was replaced with the serum- lin (P0399) from DakoCytomation (Glostrup, Denmark). free medium, in which cells were further incubated with each reagent

www.aacrjournals.org 2703 Clin Cancer Res 2005;11(7) April 1, 2005 Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Preclinical for 48 hours. Bromodeoxyuridine (BrdUrd, 10 Amol/L) was added to (Becton Dickinson, San Jose, CA). The data from 10,000 cells were the culture medium during the last 4 hours, and the incorporated collected and analyzed. BrdUrd was detected with BrdUrd Detection Kit (Roche Molecular In vivo growth assay. All animal studies were approved by the Biochemicals, Manheim, Germany), according to the manufacturer’s Animal Research Committee, Graduate School of Medicine, Kyoto instructions. Experiments were done in triplicate at least. University and done according to the Guideline for Animal Experi- Cell cycle analysis. Cells (1 Â 105) were incubated overnight in ments of Kyoto University. First, the effect of PD166866 on 60-mm dishes in DMEM with 10% FBS, followed by serum phosphorylated ERK1/2 in vivo was investigated. SYO-1 (5 Â 106) starvation for 24 hours. After treatment with SU5402 (20 Amol/L) cells suspended in 100 AL of PBS were s.c. injected into the hind for 30 minutes, FBS (1%) was added to the medium and cells were flank region of male BALB/c nu/nu athymic mice at 5 weeks of age further incubated for 48 hours. Cells were then harvested, washed (Japan SLC, Hamamatsu, Japan). PD166866 (0.1 or 0.5 mg) once with PBS, and fixed in 70% ethanol overnight. After an suspended in 50 AL of DMSO was given i.p. when the tumor incubation in PBS containing propidium iodide (10 Ag/mL) and volume was f1,500 mm3. Tumors were dissected 30 minutes after RNaseA (10 Ag/mL) for 30 minutes at 37jC, the DNA content of the injection, and tissue blocks (5 Â 5 Â 5 mm) were prepared from these cells was analyzed using FACScan and CELL Quest Software the central and peripheral portion of tumors. Whole cell lysates were

Fig. 1. Expression of FGF andFGFR in cell lines. mRNA expression of 22 FGF genes (A)andfourFGFR genes including splicing variants (B) was analyzedby RT-PCR in five synovial sarcoma cell lines as well as cell lines of malignant peripheral nerve sheath tumor (NMS-2), fibrosarcoma (HT10 80), osteosarcoma (Saos2), andcolon carcinoma (COLO205 andSW480). C, Western blotting for FGF18 using concentratedsupernatant of synovial sarcoma cells. D, Western blotting for FGFR3 using cell lysates of synovial sarcoma cells andNMS-2.Two isoforms (125 and135 kDa) corresponding to FGFR3b andFGFR3c were detected in synovial sarcoma cells but not in NMS-2.

Clin Cancer Res 2005;11(7) April 1, 2005 2704 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Fibroblast Growth Factors in Synovial Sarcoma

Fig. 2. Expression of FGF and FGFR genes in tumor tissues. A, RT-PCR analysis of the expression of five FGF genes (FGF2 , FGF8, FGF9, FGF11,andFGF18) in tumor tissues. Among 18 synovial sarcoma (SS) tumors, seven were SYT-SSX1-positive biphasic, six were SYT-SSX1-positive monophasic, andfive were SYT-SSX2-positive monophasic tumors. Abbreviations: PLS, pleomorphic liposarcoma; LMS, leiomyosarcoma; MFH, malignant fibrous histiocytoma; MPNST, malignant peripheral nerve sheath tumor. B, quantitative RT-PCR analyses of the expression of five FGF genes. ., tumors other than synovial sarcoma; o,synovial sarcoma tumors. C, RT-PCR analysis of the expression of FGFR genes. D, quantitative RT-PCR analyses of the expression of FGFR2b, FGFR2c,and FGFR3 genes in tumor tissues. ., tumors other than synovial sarcoma; o, monophasic synovial sarcoma tumors; , biphasic synovial sarcoma tumors. **, P < 0.01.

prepared from each tissue block and used for Western blot. Growth microscopic images were digitally acquired, the color blue (hematox- inhibition experiments were done using SYO-1 and HT1080. Cells ylin) was subtracted, and the brown area (PCNA-positive nucleus) was (5 Â 106) were inoculated as described above. When the tumor calculated as a digital value using Scion Image software. Five different volume reached about 50 mm3 (usually 11-12 days after inocula- sites were randomly acquired and the average value for the PCNA- tion), PD166866 was given i.p., a treatment which was repeated positive area was calculated. thereafter once a day, 6 days a week, for 2 weeks. Tumor size was Statistical analysis. Significant differences between experimental measured with vernier calipers, and the volume was calculated as k/ values were determined using t test. 6 Â length Â width Â height. Proliferating cell nuclear antigen staining. Immunohistochemical staining of proliferating cell nuclear antigen (PCNA) was done using Results EPOS anti-PCNA/horseradish peroxidase system (DakoCytomation). On day 13 (the day of the last administration), tumor tissues were Synovial sarcoma cell lines expressed multiple fibroblast growth dissected, fixed in 10% formalin, and embedded in paraffin. Sections factor and fibroblast growth factor receptor genes. The mRNA (5-Am-thick) were probed with anti-PCNA antibody for 1 hour and expression of 22 subtypes of FGF genes was analyzed by RT- counterstained with hematoxylin. To quantify the PCNA-positive area, PCR in five synovial sarcoma and five other cell lines (Fig. 1A).

www.aacrjournals.org 2705 Clin Cancer Res 2005;11(7) April 1, 2005 Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Preclinical

None of the five synovial sarcoma cell lines expressed the nerve sheath tumor but at much lower levels, and the FGF5, FGF6, FGF12, FGF14, FGF16, FGF17, FGF22,orFGF23 expression profile of Saos2 (osteosarcoma) was quite different gene, whereas the FGF2, FGF8, FGF9, FGF11, and FGF18 genes being positive only for FGF1 and FGF13. Interestingly, synovial were expressed in all five synovial sarcoma cell lines. sarcoma shared with colon carcinoma cell lines the expression Expression of the remaining nine FGF genes varied among of some FGF genes such as FGF3 and FGF18 genes, which were cell lines. The expression profile in NMS-2 (malignant not expressed in mesenchymal cell lines. Secretion of these FGF peripheral nerve sheath tumor) was similar to that in synovial proteins by synovial sarcoma cell lines was confirmed by the sarcoma, which may relate to the overall similarity of the gene Western blotting of culture supernatant. FGF18 proteins were expression profile between synovial sarcoma and malignant clearly detected in culture supernatant of synovial sarcoma cell peripheral nerve sheath tumor (13). HT1080 (fibrosarcoma) lines, although the level of protein was not completely agreed expressed a similar set of FGF genes to malignant peripheral with the level of RNA expression (Fig. 1C).

Fig. 3. Growth stimulatory effect of FGFs in synovial sarcoma cell lines. Synovial sarcoma cells were treatedwith FGF8 ( A), FGF18 (B), or FGF2 (C) at the indicated concentration, andthe amount of incorporatedBrdUrdin each sample was presentedas a relative BrdUrduptake value using the uptake of a control sample treated with DMSO as a standard value (100%). *, P < 0.05 and**, P < 0.01comparedwith each control sample.

Clin Cancer Res 2005;11(7) April 1, 2005 2706 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Fibroblast Growth Factors in Synovial Sarcoma

In contrast with the heterogeneous expression pattern of (Fig. 2C), which was not detected in cell lines (Fig. 1B). In ligands, the expression profile of receptors (FGFR)was contrast with the uniform expression pattern observed in relatively homogeneous among synovial sarcoma cell lines, synovial sarcoma, expression of the FGFR genes in other types being positive for all subtypes analyzed except FGFR2b, which of tumors seemed to vary among tumors. For example, FGFR3 is an epithelia-specific FGFR expressed in colon carcinomas was expressed in one of three leiomyosarcomas and one of two (Fig. 1B). In addition to FGFR2b, NMS-2 was negative for malignant fibrous histiocytoma. The quantitative RT-PCR FGFR3, and HT1080 was negative for FGFR2c. Expression of analyses showed that the expression level of FGFR2b and FGFR3 in synovial sarcoma cell lines was confirmed at the FGFR3 but not FGFR2c genes were significantly higher in protein level by the Western blotting using cell lysates of each synovial sarcoma than those in other types of STS (Fig. 2D). As cell line (Fig. 1D). expected, the expression level of FGFR2b gene was higher in Expression of fibroblast growth factor and fibroblast growth biphasic than in monophasic tumors, although some mono- factor receptor genes in tumor tissues. The FGF2, FGF8, phasic tumors expressed this gene at the same level observed in FGF9, FGF11, and FGF18 genes, which were commonly ex- biphasic tumors. pressed in the five synovial sarcoma cell lines, were inves- Growth stimulatory effects of fibroblast growth factors in tigated in tumor tissues of 18 synovial sarcomas and 11 STSs synovial sarcoma cell lines. Expression of FGF and FGFR genes of other types (Fig. 2A). In accordance with the results in the and proteins suggested the involvement of FGFs as autocrine/ cell lines, almost all 18 synovial sarcoma tumors expressed paracrine growth factors in synovial sarcoma. Among FGF genes all of these FGF genes (Fig. 2A). Among five FGF genes, the expressed in synovial sarcoma both in vitro and in vivo,we FGF2, FGF9, and FGF11 genes were also expressed in some focused on the FGF2, FGF8, and FGF18 genes in subsequent tumors other than synovial sarcoma. In contrast, expression of experiments. The FGF2 gene was selected as a gene commonly FGF8 and FGF18 genes was only weakly detected in other up-regulated in STS (Fig. 2A). FGF8 and FGF18 genes were tumors. Expression profiles of these five FGF genes were further selected as synovial sarcoma–specific up-regulated FGF genes analyzed by the quantitative RT-PCR (Fig. 2B). The level of (Fig. 2A). expression of FGF8 and FGF18 genes in synovial sarcoma Growth stimulatory effects of each rhFGF protein were tumors were significantly higher than in tumor tissues of other analyzed under serum-starved conditions. Recombinant type of tumors, whereas those of FGF2, FGF9, and FGF11 genes hFGF8 showed a dose-dependent growth-promoting effect showed no difference. in all synovial sarcoma cell lines (Fig. 3A). On the treatment As for receptor genes, synovial sarcoma tumors expressed all with rhFGF18, only HS-SY-II showed a dose-dependent types of FGFR genes analyzed including the FGFR2b gene increase (Fig. 3B). A heterogeneous response to FGF was

Fig. 4. Activation of ERK1/2 andp38 by FGFs in synovial sarcoma cell lines. A, Western blot analysis of total ERK1and phosphorylatedERK1/2 (pERK1/2) in synovial sarcoma cell lines after treatment with FGF8, FGF18, or FGF2. B, Western blot analysis of total p38 andphosphorylated p38 (pp38) in synovial sarcoma cell lines after treatment with FGF8, FGF18, or FGF2.

www.aacrjournals.org 2707 Clin Cancer Res 2005;11(7) April 1, 2005 Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Preclinical

Fig. 5. Antiproliferative effect of FGFR inhibitor on synovial sarcoma cell lines. A, inhibition of phosphorylation of FGFR3 by SU5402. Immunoprecipitated FGFR3 was hybridized with anti- phosphotyrosine antibody (top)andanti- FGFR3 antibody (bottom) in three synovial sarcoma cell lines. Amount of incorporated BrdUrd was analyzed after the treatment with FGFR inhibitor SU5402 in synovial sarcoma cell lines as well as other cell lines, andthe relative uptake value of each sample was calculatedusing the value of a control as a standard (100%). Experiments were done in the culture medium with 10% FBS (B)or1%FBS(C). *, P < 0.05 and**, P < 0.01.

more clearly observed in the case of rhFGF2, in which the (Fig. 4A). FGF18 also induced phosphorylation in all cell lines growth of HS-SY-II showed a dose-dependent increase, (Fig. 4A). An induction by FGF2 was also observed in HS-SY-II whereas a growth inhibitory effect was observed in the case and 1273/99, but not remarkably in YaFuSS and SYO-1 (Fig. 4A). of YaFuSS (Fig. 3C). As in the case of ERK1/2, phosphorylated p38MAPK (pp38) Activation of mitogen-activated protein kinases by fibroblast was observed in all synovial sarcoma cell lines under serum- growth factor signals. One of major signal pathways located free conditions (Fig. 4B). The level of pp38 increased on downstream of FGFR is the MAPKs, for which three subtypes are treatment with FGF8 in all synovial sarcoma cell lines, among known: ERK, p38 MAPKs, and c-Jun NH2-terminal kinases which HS-SY-II showed the clearest induction (Fig. 4B). FGF18 (JNK; ref. 25). To investigate the events downstream of activated also induced pp38 in HS-SY-II and other cell lines but at much FGFRs, the phosphorylation of these MAPKs was investigated. In lower levels. A dose-dependent increase in pp38 upon all synovial sarcoma cell lines analyzed, ERK1/2 was phosphor- treatment with FGF2 was observed in SYO-1 (Fig. 4B). ylated under serum-free conditions, suggesting the presence of Phosphorylation of JNK was not observed under serum-free endogenous stimulatory signals (Fig. 4A). The amount of conditions or after the treatment with FGFs in any synovial phosphorylated ERK1/2 (pERK1/2) increased on treatment sarcoma cell lines (data not shown). Therefore, exogenous FGF with FGF8 in a dose-dependent manner in all four cell lines signals were transmitted through both ERK1/2 and p38MAPK analyzed, with the response of HS-SY-II being the highest in synovial sarcoma cell lines.

Clin Cancer Res 2005;11(7) April 1, 2005 2708 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Fibroblast Growth Factors in Synovial Sarcoma

Fibroblast growth factor receptor–specific inhibitors reduce the Growth inhibition of fibroblast growth factor receptor inhibitor growth of synovial sarcoma cell lines in accordance with associated with the reduction of phosphorylation of extracellular inactivation of fibroblast growth factor receptor. MAPKs in signal–regulated kinase 1/2 but not of p38. To investigate synovial sarcoma are activated by intrinsic pathways, and which signals may contribute to the growth of synovial exogenous FGFs can stimulate the growth of synovial sarcoma, sarcoma, the phosphorylation of ERK1/2 and p38 before and suggesting that inhibition of the intrinsic FGF signal pathway after the treatment with SU5402 was investigated. In three may inhibit the growth of synovial sarcoma. SU5402 is known synovial sarcoma cell lines (YaFuSS, HS-SY-II, and SYO-1), to inhibit FGFR autophosphorylation at an IC50 of 10 to which were relatively sensitive to the growth inhibitory effect of 20 Amol/L (22). To confirm the effect of SU5402 on synovial SU5402, phosphorylation of ERK1/2 under low-serum con- sarcoma cell lines, the phosphorylation status of FGFR3 was ditions was completely inhibited by SU5402, whereas the analyzed by the immunoprecipitation and Western blotting in amount of phosphorylated p38 showed no change (Fig. 6A). A three synovial sarcoma cell lines (YaFuSS, SYO-1, and 1273/ small amount of pERK1/2 was still observed after SU5402 99; Fig. 5A). Phosphorylated FGFR3 was detected in these cell treatment in 1273/99, whereas no significant decrease was lines with serum-starved condition, and the treatment with observed in Fuji, both of which showed no significant change in rhFGF18 protein increased the amount of phosphorylated the pp38 (Fig. 6A). In cell lines other than synovial sarcoma, FGFR3 (Fig. 5A), which was compatible with the autocrine/ SU5402 showed no significant changes in the amount of paracrine model of FGFs in synovial sarcoma cells. Treatment pERK1/2 or pp38 (data not shown). These results suggested that with SU5402 effectively decreased the phosphorylation of the signal to induce the phosphorylation of ERK1/2 in synovial FGFR3 into almost undetectable level (Fig. 5A), indicating that sarcoma was sent mainly through the FGF and FGFR pathway SU5402 can be used as an inhibitor of FGFR in synovial and that although the FGF signal induced the phosphorylation sarcoma cells. By the treatment with SU5402, the growth of all of both ERK1/2 and p38, the growth stimulatory effects were five synovial sarcoma cells was inhibited in a dose-dependent transmitted through ERK1/2. To further confirm that the growth manner (Fig. 5B), and the IC50 value ranged from 8.5 to 17.2 inhibition of SU5402 was due to the reduction in pERK1/2, the Amol/L (Table 1). The growth inhibitory effect of SU5402 for function of MAPK/ERK kinase 1/2 in synovial sarcoma, which is cell lines other than synovial sarcoma was not remarkable. The a kinase of ERK1/2, was inhibited by U0216. Phosphorylation same experiments were done under low-serum conditions (1% of ERK1/2 was effectively inhibited in all five cell lines, in which FBS, Fig. 5C), resulting in similar growth inhibition in all 1273/99 still had a small amount of pERK1/2 at the lower synovial sarcoma cell lines. The IC50 was almost identical in concentration (Fig. 6B). BrdUrd uptake was inhibited in a dose- the three cell lines (YaFuSS, HS-SY-II, and SYO-1), whereas the dependent manner in all five cell lines, in which 1273/99 was other two cell lines (Fuji and 1273/99) required a higher relatively resistant at the lower concentration, consistent with concentration of SU5402 (Table 1). Growth inhibitory effects the status of pERK1/2 (Fig. 6C). These results further indicated of another FGFR-specific inhibitor, PD166866 (23), was also that the growth inhibition of SU5402 in synovial sarcoma was investigated under the low-serum conditions. All five synovial achieved through the inhibition of signals through ERK1/2. sarcoma cell lines had an IC50 value (0.08-4.7 Amol/L) much Growth inhibition by fibroblast growth factor receptor inhibitor lower than those of other cell lines (>10 Amol/L, Table 1). in synovial sarcoma is associated with cell cycle arrest. The Similar to the data obtained using SU5402, IC50 values in fraction of each cell cycle phase was analyzed before and after YaFuSS, HS-SY-II, and SYO-1 were much lower than those in the treatment with SU5402 in YaFuSS cells. SU5402 signifi- Fuji and 1273/99. cantly increased the G1 and subG1 fractions, and decreased the S fraction (Fig. 6D). These data suggested that the growth inhibition of synovial sarcoma by SU5402 was due to its induction of G1 arrest as previously reported for other tyrosine Ta b l e 1. IC50 (Amol/L) of FGFR inhibitors in STS cell kinase inhibitors (26). lines Fibroblast growth factor receptor inhibitors reduced the growth of synovial sarcoma cells in vivo. For in vivo study, Inhibitor we used SYO-1 because of its consistency in developing (concentration SU5402 SU54 02 PD166866 tumor, and PD166866 was used because of its lower IC50 of serum*) (10%) (1%) (1%) value. To confirm its efficacy in vivo, tumor tissue was taken c Ya Fu SS 11. 5 F 1. 8 11. 4 F 3.6 0.08 F 0.04 30 minutes after the injection of PD166866. We found that c HS-SY-II 15.1 F3.4 4.9 F 0.5 0.77 F 0.52 0.1 mg of PD166866 was enough to reduce the phosphory- c SYO-1 8.5 F 2.5 8.6 F 1. 6 0.18 F 0.11 lated ERK1/2 in the tumor (Fig. 7A). Based on this result, c Fuji 17. 2 F 1. 8 2 5 . 3 F 3.6 4.0 F 1. 6 the growth inhibitory effect of PD166866 (0.1 and 0.5 mg/ c 1273 /9 9 12.8 F 6.1 29.9 F 16.6 2.7 F 0.4 body) was analyzed in vivo. Although PD166866 was unable NMS-2 46.4 F 6.9 60.2 F 20.6 >10 to stop the growth of the tumor completely, a significant HT1080 65.4 F 2.6 63.7 F 27.8 >10 inhibition of growth was observed during the period of Saos2 >80 55.8 F 19.1 >10 administration (P < 0.01), and the effect was retained even at COLO205 >80 >80 >10 day 21 at the higher dose (P = 0.02, Fig. 7B). No definite SW480 54.4 F 21.5 78.1 F20.3 >10 reduction of body weight or pathologic changes in vital organs such as liver and kidney was observed (data not *Concentration of fetal bovine serum in medium used for measuring IC50s. shown). No growth inhibitory effect was observed when cSynovial sarcoma cell lines. HT1080 was used instead of SYO-1 even at the higher dose (data not shown).

www.aacrjournals.org2709 Clin Cancer Res 2005;11(7) April 1, 2005 Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Preclinical

Fig. 6. Effect of FGFR inhibitor on down stream molecules andcell cycle profiles in synovial sarcoma cells. A, amount of ppERK1/2 andpp38MAPK after the treatment with SU5402 was analyzedby Western blotting in synovial sarcoma cell lines.Total ERK1andp38 were also analyzed. B, Western blot analysis of pERK1/ 2 andERK1after treatment with the MAPK/ ERK kinase inhibitor U0126 in synovial sarcoma cell lines. C, BrdUrd uptake of synovial sarcoma cell lines after treatment with U0126.The relative uptake value of each sample was calculatedusing the value of a control as a standard (100%). **, P < 0.01. D, effect of FGFR inhibitor on cell cycle profiles of synovial sarcoma cell line. DNA contents ofYaFuSS after treatment with vehicle (a)or20Amol/L SU5402 (b) were analyzedby fluorescence-activatedcell sorting. *, P < 0.05 and**, P < 0.01.

Histologic examination of tumor tissues showed no signifi- synovial sarcoma cells expressed a number of FGF genes, of cant morphologic changes or focal necrosis. The number of which FGF2, FGF8, FGF9, FGF11,andFGF18 were PCNA-positive cells, however, was clearly smaller in PD166866- commonly expressed. The involvement of FGF2 in malignant treated tumors (19.3 F 6.6%, Fig. 7D) than control tumors tumors as an autocrine growth factor and/or a paracrine (53.0 F 2.0%, Fig. 7C), indicating that PD166866 reduced the angiogenic factor was first shown in brain tumors (27), and number of cells in S phase. No such difference was observed subsequently in non–small cell carcinoma (14). FGF8 was when HT1080 was used instead of SYO-1 (data not shown). originally identified as an androgen-induced growth factor found in conditioned medium of androgen-dependent mouse Discussion mammary carcinoma SC-3 cells and reported to have a growth stimulatory effect (28). The human FGF8 gene was FGFs may activate genetic programs which promote cell expressed in breast and prostate cancers (15, 16, 29) and growth by at least one of three general mechanisms: first, as associated with anchorage-independent proliferation and mitogens for the tumor cells themselves; second, by invasion (30, 31). FGF9 was originally purified from the promoting angiogenesis to supply a growing tumor; and conditioned medium of the glial cell line NMC-G1 and third, by inhibiting apoptosis and allowing tumor cells to designated as glia-activating factor due to its mitogenic continue to grow beyond normal constraints. We found that activity toward glial cells (32). FGF11 was one of four FGF

Clin Cancer Res 2005;11(7) April 1, 2005 2710 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Fibroblast Growth Factors in Synovial Sarcoma homology factors, which shared structural homology with tissues and cell lines. The expression of FGFR2b gene, which was classic FGFs but failed to activate FGFRs (33). Recently, detected in tumor tissues but not in cell lines, is intriguing. FGF11 was identified as one of the gene products enriched in FGFR2b is a receptor for keratinocyte growth factor (FGF7), and neuronal precursors during the development of the nervous expressed in epithelial cells (40). As expected, all biphasic system (34). FGF18 is cloned by its homology to FGF8 and synovial sarcomas expressed the FGFR2b gene, whereas the FGF9 (35, 36). FGF18, in association with FGF8, is expressed expression was low in most of the monophasic synovial at various times and places during embryogenesis, especially sarcomas and other STSs. It is, however, intriguing that some neurogenesis (33, 37). Mitotic activity was shown in colon monophasic synovial sarcomas with no apparent epithelial carcinomas, in which the expression of FGF18 was up- structures expressed the FGFR2b gene at a level comparable with regulated by WNT signals (17). Overall, we found that that in biphasic synovial sarcoma (Fig. 2D). All five synovial synovial sarcoma expressed several FGF genes, particularly sarcoma cell lines used in this study including SYO-1, which was those expressed in neural tissues, which may further suggest established from a biphasic synovial sarcoma, were negative for the neural origin of precursor cells of synovial sarcoma as we the FGFR2b gene. These results suggested that synovial sarcoma previously proposed (13). Mitotic activity of exogeneous FGFs tumor cells may have an intrinsic mechanism to express FGFR2b was not remarkable in some cell lines (Fig. 3). The serum- in vivo, which was somehow inhibited or disappeared in vitro. free medium used in this study contained insulin, which is Alternatively, synovial sarcoma cells positive for FGFR2b may known to potentiate the effect of FGF, and the further have growth disadvantages, and therefore only negative cells addition of insulin showed no increase of the BrdUrd uptake were established as cell lines. Recently, single nucleotide (data not shown). It is possible that the extent of autocrine polymorphism in FGFR4 is reported to associate with poor stimulation may already be saturating so that further FGF prognosis in high-grade soft tissue sarcomas including synovial stimulation does not lead to additional growth. sarcomas (41) and both FGF8 and FGF18 bind to FGFR4. The expression of FGFR1, FGFR2c, FGFR3, and FGFR4, to Although the status of polymorphism was not investigated in our which FGF2 (1c, 3c and 4), FGF8 (2c, 3c, 4), and FGF18 (2c, samples, FGF signals through FGFR4 may contribute to the 3c, 4) preferentially bind (38, 39), was confirmed in both tumor aggressiveness of synovial sarcoma. The binding of FGF to FGFR leads to receptor dimerization and tyrosine autophosphorylation, followed by downstream activation of PLCg, Crk, and SNT-1/FRS2 signaling pathways among which the SNT-1/FRS2 is the major pathway for mitogenic signals (25). Phosphorylated SNT-1/FRS2 directly binds to the GRB2-SOS complex, and then membrane- associated RAS recruits RAF-1, which in turn activates MAPKs (42). MAPKs are composed of three well-characterized sub- families, ERK, p38 MAPK, and JNK (25), and we found that the growth-promoting signal of FGF was transmitted mainly through ERK1/2. Although ERK1/2 are activated by signals through other receptor tyrosine kinases (43), phosphorylation of ERK1/2 was completely inhibited by FGFR inhibitors, which indicated that the major signal to activate ERK1/2 in synovial sarcoma is FGF signaling, suggesting the important role of FGF in synovial sarcoma. We have no clear explanation for the heterogeneous response in synovial sarcoma cell lines to FGF2, which stimulated the growth of HS-SY-II, but inhibited the growth of YaFuSS. In contrast with many studies describing the role of FGF2 as a mitogen, several reports showed that FGF2 acted as a growth inhibitor of tumors of Ewing’s sarcoma family, which are supposed to be of neuroectodermal origin (44, 45). Multiple FGF signals are involved in the biological process in synovial sarcoma, and the response may depend on the cell specificity. Nevertheless, the disruption of FGF signals was found to cause growth inhibition in all synovial sarcoma cell lines, suggesting that an inhibitory molecule for FGFR will be a promising tool for molecular target therapy in synovial sarcoma. Molecular therapy targeted to FGFR has been investigated with hematologic malignancies carrying a translocation involv- Fig. 7. Effect of FGFR inhibitor on the growth of synovial sarcoma cells in vivo. ing the FGFR gene such as chronic myeloid leukemia with the A, Western blot of phosphorylatedERK1/2 obtainedusing tissue blocks dissected from the peripheral (P) andcentral ( C) portion of SYO-1tumors in vivo. B, growth BCR-FGFR1 gene (46) and multiple myeloma with the MMSET- curve of SYO-1tumors in vivo. PD166866 (0.1or 0.5 mg) was injected FGFR3 fusion gene (47). In both instances, SU5402 was found intraperitoneally on the days indicated by an arrow.C and D, immunohistochemistry to effectively inhibit the growth of cell lines with an IC50 for PCNA in SYO-1tumors.Tumor tissues treatedwith vehicle ( C) or 0.5 mg PD166866 (D) were dissected at day 13, and processed for the staining of PCNA. equivalent to that in synovial sarcoma. For experiments in vivo, Original magnification, Â400. we used PD166866 instead of SU5402 because of its lower IC50

www.aacrjournals.org 2711 Clin Cancer Res 2005;11(7) April 1, 2005 Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Cancer Therapy: Preclinical value and found that i.p. given PD166866 efficiently inhibited some instances (48, 49), and therefore further in vivo studies the phosphorylation of ERK1/2 in s.c. tumor causing an using the treatment combined with suitable anticancer drugs inhibition of growth without significant side effects. As in the may provide a promising approach for clinical application. case of other tyrosine kinase inhibitors, PD166866 did not strongly induce apoptosis or necrosis, and therefore no Acknowledgments complete tumor remission was observed. However, cell cycle arrest caused by the treatment with PD166866 may enhance We thank Drs. H. Sonobe, A. Kawai, S.Tanaka, O. Larsson, andA. Ogose for pro- the cytotoxic effects of anticancer drugs as previously shown in viding cell lines and Y.Furukawa for useful comments.

References 1. Weiss SW, Goldblum JR. Enzinger and Weiss’s soft tis- 19. Kawai A, Naito N, Yoshida A, et al. Establishment system development. Proc Natl Acad Sci U S A1996; sue tumors. 4th ed. St. Louis: Mosby; 2001. andcharacterization of a biphasic synovial sarcoma 93:9850 ^ 7. 2. Fisher C. Synovial sarcoma. Ann Diagn Pathol 1998; cell line, SYO-1. Cancer Lett 2004;204:105 ^ 13. 34. Luo Y, Cai J, Liu Y, et al. Microarray analysis of se- 2:401^21. 20. Nojima T, Wang YS, Abe S, Matsuno T, Yamawaki lectedgenes in neural stem andprogenitor cells. 3. Spillane AJ, A’Hern R, Judson IR, Fisher C, S, Nagashima K. Morphological andcytogenetic J Neurochem 2002;83:1481 ^ 97. Thomas JM. Synovial sarcoma: a clinicopathologic, studies of a human synovial sarcoma xenotrans- 35. Ohbayashi N, Hoshikawa M, Kimura S,Yamasaki M, staging, andprognostic assessment. J Clin Oncol plantedinto nudemice. Acta Pathol Jpn 1990; Fukui S, Itoh N. Structure andexpression of the mRNA 2000;18:3794^803. 40:486^93. encoding a novel fibroblast growth factor, FGF-18. 4. Lewis JJ, Antonescu CR, Leung DH, et al. Synovial 21. ImaizumiS,MotoyamaT,OgoseA,HottaT, J Biol Chem 1998;273:18161 ^ 4. sarcoma: a multivariate analysis of prognostic factors Takahashi HE. Characterization andchemosensitivity 36. Hu MC, Qiu WR, Wang YP, et al. FGF-18, a novel in112patients with primary localizedtumors of the ex- of two human malignant peripheral nerve sheath tu- member of the fibroblast growth factor family, stimu- tremity. J Clin Oncol 2000;18:2087 ^ 94. mour cell lines derived from a patient with neurofibro- lates hepatic andintestinal proliferation. Mol Cell Biol matosis type1.Virchows Arch1998;433:435^41. 5. Turc-Carel C, Dal Cin P, Limon J, et al. Involvement of 1998;18:6063^74. 22. Mohammadi M, McMahon G, Sun L, et al. Struc- chromosome X in primary cytogenetic change in hu- 37. OhuchiH,KimuraS,WatamotoM,ItohN.Involve- tures of the tyrosine kinase domain of fibroblast man neoplasia: nonrandom translocation in synovial ment of fibroblast growth factor (FGF)18-FGF8 growth factor receptor in complex with inhibitors. Sci- sarcoma. Proc Natl AcadSci U S A1987;84:1981 ^ 5. signaling in specification of left-right asymmetry and ence1997;276:955 ^ 60. 6. Clark J, Rocques PJ, Crew AJ, et al. Identifi- brain andlimb development of the chick embryo. 23. Panek RL, Lu GH, DahringTK, et al. In vitro biologi- cation of novel genes, SYT andSSX, involvedin Mech Dev 2000;95:55^66. cal characterization andantiangiogenic effects of the t(X;18)(p11.2;q11.2) translocation foundin human 38. Ornitz DM, Xu J, Colvin JS, et al. Receptor specific- PD166866, a selective inhibitor of the FGF-1receptor synovial sarcoma. Nat Genet 1994;7:502 ^ 8. ity of the fibroblast growth factor family. J Biol Chem tyrosine kinase. J Pharmacol Exp Ther 1998;286: 1996;271:15292^7. 7. Dos Santos NR, De Bruijn DR, Van Kessel AG. 569^77. Molecular mechanisms underlying human synovial 24. Favata MF, Horiuchi KY, Manos EJ, et al. Identifica- 39. Ellsworth JL, Berry J, Bukowski T, et al. Fibroblast sarcoma development. Genes Chromosomes Can- tion of a novel inhibitor of mitogen-activatedprotein growth factor-18 is a trophic factor for mature chon- cer 2001;30:1^ 14. kinase kinase. J Biol Chem1998;273:18623 ^ 32. drocytes and their progenitors. Osteroarthritis Carti- lage 2002;10:308 ^ 20. 8. Ladanyi M. Fusions of the SYTand SSX genes in sy- 25. Powers CJ, McLesky SW, Wellstein A. Fibroblast novial sarcoma. Oncogene 2001;20:5755 ^ 62. growth factors, their receptors, andsignaling. Endocr 40.Werner S. Keratinocyte growth factor: a unique play- 9. Huang ME,YeYC, Chen SR, et al. Use of all-trans ret- RelatCancer2000;7:165^97. er in epithelial repair processes. Cytokine Growth Fac- tor Rev 1998;9:153 ^ 65. inoic acidin the treatment of acute promyelocytic leu- 26. Shintani S, Li C, Mihara M, et al. Gefitinib (‘‘Iressa’’, kemia. Blood1988;72:567^72. ZD1839), an epidermal growth factor receptortyrosine 41. MorimotoY, Ozaki T, Ouchida M, et al. Single nucleotide polymorphism in fibroblast growth factor recep- 10. Druker BJ,Tamura S, Buchdunger E, et al. Effects of kinase inhibitor, up-regulates p27KIP1andinduces G 1 a selective inhibitor of the Abl tyrosine kinase on the arrest in oral squamous cell carcinoma cell lines. Oral tor 4 at codon 388 is associated with prognosis in growth of Bcr-Abl positive cells. Nat Med1996;2: Oncol 2004;40:43 ^ 51. high-grade soft tissue sarcoma. Cancer 2003;98: 2245 ^ 50. 561^6. 27. TakahashiJ, Mori H, Fukumoto M, et al. Gene expres- 42. Kouhara H, Hadari YR, Spivak-Kroizman T, et al. 11. Sato Y, Nabeta Y,Tsukahara T, et al. Detection and sion of fibroblast growth factors in human gliomas and A lipid-anchored Grb2-binding protein that links FGF- induction of CTLs specific for SYT-SSX-derived pep- meningiomas: demonstrationofcellular source ofbasic receptor activation to the Ras/MAPK signaling path- tides in HLA-A24(+) patients with synovial sarcoma. fibroblast growth factor mRNA andpeptidein tumor way. Cell 1997;89:693^ 702. J Immunol 20 02;169 :1611 ^ 8. tissues. Proc Natl AcadSci US A1990;87:5710^ 4. 43. Denhardt DT. Signal-transducing protein phosphor- 12 . Zanders ED. Gene expression analysis as an aid to 28. Tanaka A, Miyamoto K, Minamino N, et al. Cloning ylation cascades mediated by Ras/Rho proteins in the the identification of drug targets. Pharmacogenomics andcharacterization of an androgen-inducedgrowth mammalian cell: the potential for multiplex signaling. 2000;1:375 ^ 84. factor essential for the androgen-dependent growth Biochem J 1996;318:729 ^ 47. 13. Nagayama S, Katagiri T,Tsunoda T, et al. Genome- of mouse mammary carcinoma cells. Proc Natl Acad Sci U S A1992;89:8928 ^ 32. 44. Sturla LM, WestwoodG, Selby PJ, Lewis IJ, wide analysis of gene expression in synovial sarco- Burchill SA. Induction of cell death by basic fibro- 29.Tanaka A, Furuya A,Yamasaki M, et al. High frequen- mas using a cDNA microarray. Cancer Res 2002;62: blast growth factor in Ewing’s sarcoma. Cancer Res 5859^66. cy of fibroblast growth factor (FGF) 8 expression in 2000;60:6160^70. clinical prostate cancers andbreast tissues, immuno- 14. Berger W, Setinek U, Mohr T, et al. Evidence for a 45. Kim MS, Kim CJ, Jung HS, et al. Fibroblast growth role of FGF-2 andFGF receptors in the proliferation histochemically demonstrated by a newly established neutralizing monoclonal antibody against FGF 8. Can- factor 2 induces differentiation and apoptosis of Askin of non-small cell lung cancer cells. Int J Cancer tumour cells. JPathol 2004;202:103 ^ 12. cer Res 1998;58:2053 ^ 6. 1999;83:415^23. 46. Demiroglu A, Steer EJ, Heath C, et al.The t(8;22) in 15. Tanaka A, Miyamoto K, Matsuo H, Matsumoto K, 30. Ruohola JK, Viitanen TP,Valve EM, et al. Enhanced chronic myeloidleukemia fuses BCR to FGFR1: trans- invasion andtumor growth of fibroblast growth factor Yoshida H. Human androgen-induced growth factor in forming activity andspecific inhibition of FGFR1fusion prostate andbreast cancer cells: its molecular cloning 8b-overexpressing MCF-7 human breast cancer cells. proteins. Blood2001;98:3778 ^ 83. Cancer Res 2001;61:4229 ^ 37. andgrowth properties. FEBSLett1995;363:226 ^30. 47. PatersonJL, Li Z,Wen XY,et al. Preclinical studies of 16. Dorkin TJ, Robinson MC, March C, Bjartell A, 31. Song Z, Powell WC, Kasahara N, van Bokhoven A, fibroblast growth factor receptor 3 as a therapeutic Neal DE, Leung HY. FGF8 over-expression in pros- Miller GJ, Roy-Burman P. The effect of fibroblast target in multiple myeloma. Br J Haematol 2004;124: tate cancer is associatedwith decreasedpatient growth factor 8, isoform b, on the biology of prostate 595^603. carcinoma cells andtheir interaction with stromal cells. survival andpersists in androgen independent dis- 48. Garofalo A, Naumova E, Manenti L, et al. The com- ease. Oncogene 1999;18:2755 ^ 61. Cancer Res 2000;60:6730^6. bination of the tyrosine kinase receptor inhibitor 17. ShimokawaT, Furukawa Y, Sakai M, et al. Involve- 32. Miyamoto M, Naruo K, Seko C, Matsumoto S, SU6668 with paclitaxel affects ascites formation ment of the FGF18 Gene in colorectal carcinogenesis Kondo T, Kurokawa T. Molecular cloning of a novel andtumor spreadin ovarian carcinoma xenografts as a novel downstream target of the h-catenin/T-cell cytokine cDNA encoding the ninth member of the growing orthotopically. Clin Cancer Res 2003;9: factor complex. Cancer Res 2003;63:6116^ 20. fibroblast growth factor family, which has a unique 3476 ^ 85. 18. Sonobe H, Manabe Y, Furihata M, et al. Establish- secretion property. Mol Cell Biol 1993;13:4251 ^ 9. 49. Abdollahi A, Lipson KE, Sckell A, et al. Combined ment andcharacterization of a new human synovial 33. SmallwoodPM, Munoz-Sanjuan I,Tong P, et al. Fi- therapy with direct and indirect angiogenesis inhibi- sarcoma cell line, HS-SY-II. Lab Invest 1992;67: broblast growth factor (FGF) homologous factors: tion results in enhancedantiangiogenic andantitumor 498^505. new members of the FGF family implicatedin nervous effects. Cancer Res 2003;63:8890 ^8.

Clin Cancer Res 2005;11(7) April 1, 2005 2712 www.aacrjournals.org Downloaded from clincancerres.aacrjournals.org on September 28, 2021. © 2005 American Association for Cancer Research. Disruption of Fibroblast Growth Factor Signal Pathway Inhibits the Growth of Synovial Sarcomas: Potential Application of Signal Inhibitors to Molecular Target Therapy

Tatsuya Ishibe, Tomitaka Nakayama, Takeshi Okamoto, et al.

Clin Cancer Res 2005;11:2702-2712.

Updated version Access the most recent version of this article at: http://clincancerres.aacrjournals.org/content/11/7/2702

Cited articles This article cites 46 articles, 25 of which you can access for free at: http://clincancerres.aacrjournals.org/content/11/7/2702.full#ref-list-1

Citing articles This article has been cited by 13 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/11/7/2702.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://clincancerres.aacrjournals.org/content/11/7/2702. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.