Epigenetically Controlled Fibroblast Growth Factor Receptor 2 ...

Research Article

Epigenetically Controlled Fibroblast Growth Factor Receptor 2 Signaling Imposes on the RAS/BRAF/Mitogen-Activated Protein Kinase Pathway to Modulate Thyroid Cancer Progression 1,3,4

2,3,4

Tetsuo Kondo, Lei Zheng, Wei Liu, 1,3,4 2,3,4 Sylvia L. Asa, and Shereen Ezzat

1,3,4

5

Junichi Kurebayashi,

Departments of 1Pathology and 2Medicine, and 3the Ontario Cancer Institute, Princess Margaret Hospital, University Health Network and 4University of Toronto, Toronto, Ontario, Canada and 5Departments of Breast and Thyroid Surgery, Kawasaki Medical School, Okayama, Japan

Abstract Fibroblast growth factor (FGF) signals play fundamental roles in development and tumorigenesis. Thyroid cancer is an example of a tumor with nonoverlapping genetic mutations that up-regulate mitogen-activated protein kinase (MAPK). Here, we show that FGF receptor 1 (FGFR1), which is expressed mainly in neoplastic thyroid cells, propagates MAPK activation and promotes tumor progression. In contrast, FGFR2 is down-regulated in neoplastic thyroid cells through DNA promoter methylation. Reexpression of FGFR2 competes with FGFR1 for the immediate substrate FGFR substrate 2 to impede signaling upstream of the BRAF/MAPK pathway. These data unmask an epigenetically controlled FGFR2 signal that imposes precisely on the intragenically modified BRAF/MAPK pathway to modulate thyroid cancer behavior. [Cancer Res 2007;67(11):5461–70]

Introduction Thyroid cancer is the most common endocrine malignancy and is increasing in incidence (1). The majority of thyroid carcinomas are derived from follicular epithelial cells and show a spectrum of differentiation from indolent well-differentiated papillary and follicular carcinomas to the more aggressive poorly differentiated carcinoma and the rare but rapidly lethal undifferentiated carcinoma (2). This spectrum of progression has been linked with a pattern of cumulative genetic defects that correlates with tumor differentiation, aggressiveness, and metastatic potential (3). Gene rearrangements involving the RET proto-oncogene or activating point mutations along the RAS/BRAF pathway account for the majority of these carcinomas (3). Cancer behavior and progression are modified by dysregulation of growth factor signaling. Alterations of several growth factors and their receptors identified in thyroid tumors include up-regulation of MET, epidermal growth factor receptor, platelet-derived growth factor, and vascular endothelial growth factor (VEGF; ref. 3). Fibroblast growth factors (FGF) and FGF receptors (FGFR) are also implicated in regulating endocrine neoplasia, including thyroid carcinoma (4, 5).

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). Conflict of interest: The authors have declared that no conflict of interest exists. Requests for reprints: Shereen Ezzat, Ontario Cancer Institute, 610 University Avenue #8-327, Toronto, Ontario, Canada, M5G-2M9. Phone: 416-586-8505; Fax: 416586-8834; E-mail: [email protected]. I2007 American Association for Cancer Research. doi:10.1158/0008-5472.CAN-06-4477

www.aacrjournals.org

FGFs comprise a family of heparin-binding proteins that currently includes 23 members. They signal through four highaffinity tyrosine kinase receptors (FGFR1–FGFR4; ref. 5). Each receptor has two or three immunoglobulin-like extracellular domains, a transmembrane domain, an intracellular split tyrosine kinase, and a carboxyl-terminus (6). After FGF binding and receptor dimerization, several signal transduction pathways are activated, mainly involving FRS2 and PLCg. Activated FRS2 recruits the GRB2/SOS complex and ultimately mitogen-activated protein kinase [MAPK or extracellular signal-regulated kinase 1 (ERK1)/ ERK2; ref. 6]. Combinations of FGFs, FGFR isoforms, and adaptor proteins comprise complex signaling networks that play fundamental roles in development, organogenesis, cell differentiation, angiogenesis, and tumor progression (5, 7). Up-regulation of FGFR1 has been identified in astrocytomas, breast carcinomas, prostate carcinomas, melanomas, and malignant salivary gland tumors (6, 7). In thyroid, increased expression of FGFR1 has been observed in benign and malignant tumors (4, 8). FGFR1 is up-regulated in rat thyroid follicular cells upon goitrogen administration, whereas a dominant-negative FGFR1 reduces goitrogenesis in mice (9, 10). We have previously shown that FGFR expression is dysregulated in human thyroid tumors and cell lines (4). FGFR2 was the only FGFR consistently detected in normal thyroid tissues, and its expression was diminished in a large tissue microarray of thyroid tumors and in six carcinoma cell lines (4). In contrast, FGFR1 was expressed in hyperplastic goiters, benign adenomas, and carcinomas (4). Thus far, no mutations or rearrangements involving FGFRs have been identified in thyroid cancers, suggesting that epigenetic factors are implicated in their dysregulated expression in thyroid tumors. In this study, we focused on the expression of two principal members of the FGFR family (1 and 2) in thyroid. We hypothesized that FGFR1 promotes thyroid cell growth and that FGFR2 plays a protective role against cancer progression in genetically transformed thyroid cells. We used loss-of-function and gain-of-function approaches to investigate the signaling effect of FGFR1 and FGFR2 on thyroid cancer growth in vitro and in vivo. We also examined the potential role of epigenetic modification as a mechanism implicated in FGFR2 down-regulation in thyroid carcinoma cells.

Materials and Methods Cell lines and cell culture. Human thyroid carcinoma–derived cell lines, WRO, NPA, and DRO (provided by Dr. J. Fagin, University of Cincinnati, originally established by Dr. G. Juillard, University of California at Los Angeles, Los Angeles, CA), TPC-1 (Dr. S.M. Jhiang, Ohio State University, Columbus, OH), KTC-1 (11), and 8505C (Cell Resource Center for

5461

Cancer Res 2007; 67: (11). June 1, 2007

Cancer Research Biomedical Research, Tohoku University, Sendai, Japan), were maintained in RPMI 1640 (Life Technologies) supplemented with 10% fetal bovine serum (FBS), streptomycin sulfate (100 units/mL), and penicillin (100 Ag/mL). H, N, or K-Ras mutations and BRAF V600E were examined in all cell lines by PCR direct sequencing using primers listed in Supplementary Table S1. The human embryonic kidney (HEK) 293 cells were propagated in DME (Life Technologies) supplemented with 10% FBS and antibiotics. Vector constructs and stable cell transfection. FGFR1 small interfering RNA (siRNA) target sequence, without homology to other FGFRs and other known human genes, was 5¶-AAGAAATTGCATGCAGTGCCG-3¶ situated in the second immunoglobulin-like domain (nucleotide positions 487–507) to cover most of the splice variants of FGFR1. The double-stranded oligonucleotide templates containing the sequence of the hairpin siRNA (top strand 5¶-GATCCGAAATTGCATGCAGTGCCGTCTGCAGGACGGCACTGCATGCAATTTCTTTTTTGGAAA-3¶; bottom strand 5¶-AGCTTTTCCAAAAAAGAAATTGCATGCAGTGCCGTCCTGCAGACGGCACTGCATGCAATTTCG-3¶) were synthesized and inserted into the expression vector (pSilencer 2.1-U6 neo, Ambion). A circular vector that expresses a control hairpin siRNA without homology to any known human sequence was used as a control. The cDNA encoding human FGFR2-IIIb, also known as Ksam-IIC1 (established by Drs. M. Terada and T. Yoshida, National Cancer Institute, Tokyo, Japan) in pcDNA1/Neo expression vector (Invitrogen) and a control empty vector (pcDNA1/Neo) were kindly provided by Dr. F. Radvanyi (Center National de la Recherche Scientifique, Paris, France). The full-length FGFR1 cDNA, kindly provided by Dr. J. Rossant (Hospital for Sick Children, Toronto, Canada), was inserted into pcDNA3.1/Neo (Invitrogen). The expression vectors were transfected into cells using LipofectAMINE (Invitrogen). Stable clones were selected and maintained in a growth medium containing 1 mg/mL of Geneticin (Life Technologies). Alterations in FGFR expression were confirmed by Western blotting. Two independent clones of each manipulation were used for all studies. Growth factor stimulation. After 24 h starvation in serum-free medium, cells were treated with the nonselective FGF1 ligand (25 ng/mL, Sigma) or the FGFR2-selective ligand FGF7 (25 ng/mL, Sigma), each with 10 units/mL of heparin (Sigma) in serum-free medium for 15 min at 37jC. Identical volume of vehicle served as control. In other experiments, 10% FBS was used as a source of multiple FGF ligands. RNA extraction and reverse transcription–PCR analysis. Total RNA was isolated from cultured cells and frozen human thyroid tissues using TRIzol (Invitrogen). cDNA was generated using the TaqMan reverse transcription (RT) reagent kit (Applied Biosystems). Specific PCR primers for FGFR1, FGFR2-IIIb, FGFR2-IIIc, FRS2, and phosphoglycerate kinase 1 (as an internal control) were used as listed in Supplementary Table S1. Amplicons were designed to cross exon/intron boundaries to exclude genomic DNA contamination. Amplification was done using HotStarTaq DNA polymerase kit (Qiagen). PCR conditions were as follows: (i) 95jC for 15 min; (ii) 30 cycles of 94jC for 30 s, 56jC or 58jC for 30 s, and 72jC for 1 min; (iii) 72jC for 10 min; and (iv) 4jC hold. Negative controls omitting RT and positive controls were included in each PCR reaction. Protein isolation and Western blotting analysis. Thyroid tissues were homogenized using a polytron homogenizer in radioimmunoprecipitation assay buffer (RIPA) lysis buffer with proteinase inhibitors [1 PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mmol/L phenylmethylsulfonyl fluoride (PMSF), 12 Ag/mL aprotinin, and 1 mmol/L sodium orthovanadate]. Cultured cells were lysed in a RIPA lysis buffer with proteinase inhibitors and 5 AL/mL of phosphatase inhibitor cocktail 1 and 2 (Sigma). Equal amounts of protein (40 Ag) solubilized in sample buffer were separated on 10% SDS polyacrylamide gels and transferred electrophoretically to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked in TBS containing 0.5% Tween 20 plus 5% nonfat dried milk for 1 h at room temperature, probed with primary antibodies at 4jC overnight. Primary antisera or monoclonal antibodies were used at the specified dilutions: anti-FGFR1 (Santa Cruz Biotechnology; 1:500), anti-FGFR2 (Santa Cruz Biotechnology; 1:500), anti–phospho-FRS2 (Cell Signaling Technology 1:1000), anti–total FRS2 (Santa Cruz Biotechnology; 1:500), anti–phospho-

Cancer Res 2007; 67: (11). June 1, 2007

BRAF (Santa Cruz Biotechnology; 1:200), anti–total BRAF (Santa Cruz Biotechnology; 1:1,000), anti–phospho-MAPK (Sigma; 1:1,000), anti–total MAPK (Sigma; 1:5,000), anti–phospho-AKT (Cell Signaling Technology; 1:1,000), anti–phospho-RB (Cell Signaling Technology; 1:1,000), phosphoserine (Zymed Laboratories; 1:1,000), and antiactin (Sigma; 1:1,000). Membranes were washed thrice for 10 min each in TBS containing 0.5% Tween 20 and incubated with horseradish peroxidase (HRP)–conjugated goat antirabbit or anti-mouse secondary antibody (Santa Cruz Biotechnology; 1:2,000) for 1 h at room temperature. Targeted proteins were visualized using an enhanced chemiluminescence detection system (Amersham). Immunoprecipitation. Protein lysates, 500 Ag in 1 mL RIPA lysis buffer, were incubated with 1 Ag primary antibody and 20 AL resuspended volume of protein A/G PLUS-Agarose (Santa Cruz Biotechnology) at 4jC with gentle shaking overnight. Immune complexes were collected by centrifugation at 2,500 rpm for 5 min, washed four times with 1 mL RIPA buffer by repeated centrifugation, then suspended in sample buffer, boiled at 95jC for 5 min, separated by 10% SDS-PAGE, and analyzed by Western blotting. BRAF kinase and Ras-GTP loading assays. The BRAF kinase activity was analyzed using an in vitro BRAF kinase assay kit (Upstate Biotechnology). Briefly, 500 Ag of cell lysate of each sample was immunoprecipitated with anti-BRAF polyclonal antibody (Santa Cruz Biotechnology). After extensive washing, immunoprecipitated BRAF was resuspended in a RIPA buffer. The immunoprecipitate was incubated with inactive recombinant MAP/ERK kinase 1 (MEK1) according to the manufacturer’s instructions. Active recombinant BRAF was used as a positive control. The phosphorylated MEK1 in reaction mixture was detected by Western blotting with anti–phospho-MEK1/MEK2 polyclonal antibody. Similarly, the RASGTP loading assay was done as per the manufacturer’s instructions (Upstate Biotechnology). Briefly, 500 Ag of cellular protein lysed in Mg2+ lysis/wash buffer was affinity precipitated with 10 Ag of glutathione S-transferase (GST)-RBD, a GST fusion protein containing RAS-binding domain of Raf-1, bound to glutathione agarose beads. After three washes with Mg2+ lysis/wash buffer, the beads were boiled in Laemmli reducing sample buffer, separated by 12% SDS-PAGE, transferred to PVDF membrane, and immunoblotted with anti-Ras monoclonal antibody. GDP or GTPgS was used as negative or positive controls, respectively. Cell pellets and immunohistochemistry. Cultured cells were pelleted as previously described (4). After deparaffinization of paraffin sections, indirect immunoperoxidase staining was carried out with antigen retrieval treatment. Primary antiserum against FGFR1 (Santa Cruz Biotechnology; 1:200) or FGFR2 (Santa Cruz Biotechnology; 1:200) was incubated at room temperature for 2 h, washed in PBS, then incubated with HRP-conjugated second antibody. Reactions were visualized with 3,3¶-diaminobenzidine (DAB) and counterstained with hematoxylin. Negative controls omitted primary antibody or used primary antiserum/antibody preabsorbed with purified antigen. Apoptosis assays. To examine the extent of DNA fragmentation characteristic of apoptosis, we stained 4-Am sections of cell pellets using the terminal dUTP nick-end labeling (TUNEL) technique (ApopTag kit, Oncor). Paraffin sections were treated with 2% hydrogen peroxide to quench endogenous peroxide for 30 min and exposed to 5 Ag/mL of proteinase K for 15 min at room temperature. Sections were washed and exposed to equilibration buffer for 10 min. Each slide was then incubated with terminal deoxytransferase and digoxigenin-labeled TdT at 4jC overnight followed by HRP-conjugated antidigoxigenin antiserum for 1 h. The peroxidase reaction was visualized with DAB. Control sections were stained without terminal deoxytransferase, digoxigenin-conjugated TdT, or antidigoxigenin antiserum. Cell proliferation assay. Cells were seeded in a 96-well plate and labeled with 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT; Sigma) as a measure of cell proliferation. Absorbance was measured with an OPTI max microplate reader (Molecular Devices) at 570 nm and reference wavelength of 650 nm. Cell cycle analysis. For cell cycle assessment, f50% confluent cells were starved in serum-free growth medium for 24 h before being exposed to growth medium containing 10% FBS for 24 h. After trypsinization, the cell suspension was centrifuged at 1,500 rpm for 5 min at 4jC. Cells were then

5462


FGFR1 and FGFR2 in Thyroid Cancer fixed in 20jC 80% ethanol overnight. Fixed cells were washed with icecold staining buffer [1 PBS, 0.2% Triton X-100, and 1 mmol/L EDTA (pH 8.0)] and resuspended in staining buffer containing 50 Ag/mL RNase A (Sigma) and 50 Ag/mL propidium iodide for 1 h. A FACScan (Becton Dickinson) coupled with CellQuest software was used to obtain fluorescence-activated cell sorting (FACS) data. Invasion and migration assays. Cell motility was examined in a transwell assay using 24-well plates with uncoated inserts to examine migration or Matrigel-coated inserts to assess invasiveness (Becton Dickinson). The upper and lower culture compartments were separated by polycarbonate filters with a pore size of 8 Am. After trypsinization, 2.5 104 cells were plated in each insert with 500 AL of serum-free medium. The growth medium containing 10% FBS was used as a chemoattractant in the bottom well. After 20 h of incubation, cells on the upper surface were removed by scrubbing with a cotton swab. Cells on the lower surface of the membrane were stained with Diff-Quik stain (Dade Behring) and quantified by light microscopy. Assays were done in triplicate. Human thyroid cancer cell xenografts in severe combined immune deficiency mice. Subconfluent cells were trypsinized, washed twice with PBS, and harvested by centrifugation. Cell pellets were resuspended in PBS, and 5 106 cells in 0.1 mL volume were injected s.c. into the flank of 6-week-old female severe combined immune deficiency (SCID) mice to generate s.c. tumors. Tumor volume was monitored every 2 days. Mice were sacrificed after 14 days after cell implantation, tumors were excised and weighed, and volume was measured. Excised tissue was fixed in 10% formalin and embedded in paraffin for light microscopy and immunohistochemical staining. The mouse protocol was approved by the Ontario Cancer Institute Animal Care and Utilization Committee. 5¶-Aza-deoxycytidine and trichostatin A treatment. Cells were plated at a density of 5 105 per 10 cm2 dish and incubated in growth medium without or with the demethylating agent 5¶-aza-deoxycytidine (10 Amol/L, Sigma) for 72 h. For histone deacetylase inhibition, trichostatin A (0.3 Amol/L, Sigma) was applied for 24 h. The equivalent volume of vehicle (50% acetic acid for 5¶-aza-deoxycytidine or 100% ethanol for trichostatin A) was applied as control. Bisulfite treatment and methylation-specific PCR (MSP) assay. Genomic DNA was extracted from 5¶-aza-deoxycytidine–treated or untreated cells by proteinase K digestion and phenol/chloroform extraction. Denatured DNA was modified by bisulfite under conditions that convert all unmethylated cytosines to uracils using CpGenome DNA modification kit (Chemicon International). Specific primers are listed in Supplementary Table S1. Amplification was done in a reaction volume of 50 AL containing 40 ng of bisulfite-treated DNA, 1 PCR buffer, 3.0 mmol/L MgCl2, 0.25 mmol/L of each deoxynucleotide triphosphate, 0.5 Amol/L of each primer, and 1.25 units of HotStarTaq DNA polymerase (Qiagen). PCR conditions were as follows: (a) 95jC for 15 min; (b) 30 cycles of 94jC for 30 s, 54jC for 30 s, and 72jC for 1 min, (c) 72jC for 10 min, and (d) 4jC hold. Both negative and positive controls were included for all PCR reactions. Statistical analysis. Data are presented as mean F SE. Statistical analysis was conducted by the Student’s t test. P values of 0.05 or less were considered statistically significant.

Results Expression of FGFR1 and FGFR2 in normal thyroid and thyroid carcinoma cell lines. We previously reported expression of FGFR1 in TPC-1, NPA, and WRO thyroid cancer cell lines and lack of FGFR2 expression in transformed compared with nontransformed thyroid follicular cells (4). In this study, FGFR1 mRNA was detected by RT-PCR in TPC-1, WRO, and 8505C carcinoma cell lines but not in KTC-1 cells; in contrast only a faint signal was found in normal thyroid tissue, a feature consistent with stromal contribution (Fig. 1A). Western blotting confirmed that endogenous FGFR1 protein was up-regulated in WRO and 8505C cells, although it was undetectable in all normal thyroid tissues (Fig. 1B).


We previously reported that FGFR2 protein was undetectable in six human thyroid carcinoma cell lines (TPC-1, NPA, WRO, MRO, DRO, and ARO; ref. 4). To complete a search for FGFR2 expression, we examined two additional thyroid carcinoma cell lines (papillary thyroid carcinoma–derived KTC-1 cells and undifferentiated thyroid carcinoma–derived 8505C cells). Alternative splicing of the third immunoglobulin-like extracellular domain of FGFR2 results in two major transcripts, FGFR2-IIIb and FGFR2-IIIc, that exhibit distinct ligand binding specificities (12). Thus, we also examined FGFR2 splicing using primers to specifically amplify either FGFR2-IIIb or FGFR2-IIIc. In contrast to FGFR1 expression, FGFR2-IIIb mRNA was identified in all normal thyroids but only in one thyroid carcinoma cell line, the KTC-1 cell line (Fig. 1A). FGFR2-IIIc mRNA was not detectable in normal thyroids or carcinoma cell lines. FGFR2 protein was detectable in all normal thyroid tissues but was not detectable in thyroid carcinoma cell lines with the exception of KTC-1 cells (Fig. 1B). The doublet migration of FGFR1 and FGFR2 is consistent with variable N-terminal glycosylation (4, 13). To examine the signaling potential of FGFRs in thyroid cells, we examined expression of the FGFR substrate adaptor protein FRS2. Although FRS2 mRNA was detected in normal and thyroid carcinoma cell lines (Fig. 1A), FRS2 was up-regulated at the protein level in all four thyroid cancer cell lines examined (Fig. 1B). FGFR1 silencing and FGFR2-IIIb reexpression in WRO cells. To examine the contribution of FGFR1 to cell growth, we established stable cell clones of WRO cells with down-regulated FGFR1 using siRNA-mediated gene silencing. We selected WRO cells based on their endogenous expression of FGFR1 but not FGFR2; these cells do not have confounding growth effects known to be characteristic of FGFR4 that is expressed in more aggressive tumors (4). We also confirmed that they do not harbor a BRAF mutation or an intragenic RAS mutation (data not shown). Western blotting documented silencing of FGFR1 expression, and this was corroborated by diminished immunoreactivity for FGFR1 using immunohistochemistry of cell blocks (Fig. 1C). Having determined that the FGFR2-IIIb isoform is expressed in normal thyroids but not in the majority of thyroid carcinoma cell lines, we examined the functional properties of FGFR2-IIIb using a gain-of-function approach. WRO cells, which do not endogenously express FGFR2, were forced to express FGFR2-IIIb. Western blotting showed strong expression of FGFR2 in FGFR2-IIIb/WRO cells, whereas pcDNA/WRO cells were negative (Fig. 1D). Similarly, immunohistochemistry confirmed the stable reexpression of FGFR2 (Fig. 1D). FGFR1 silencing impedes cell proliferation and invasion. To test the signaling consequences of FGFR1 silencing, we examined the response of FGFR1–down-regulated WRO cells to FGF1, a ligand that activates multiple FGFRs (6). As noted in control siRNA/WRO cells, FGF1 effectively induced phosphorylation of MAPK (ERK1/2), Akt, and Rb (Fig. 2A). In contrast, silencing of FGFR1 prevented FGF1-induced responses, highlighting the importance of FGFR1 in FGF-mediated signaling. To assess the effect of FGFR1 down-regulation on thyroid cancer cell behavior, we used transwell assays. FGFR1 underexpression unequivocally suppressed thyroid cancer cell invasion (P = 0.04) with no significant effect on cell migration (Fig. 2B). We also compared cell cycle progression in FGFR1 siRNA/WRO and control cells by FACS analysis. Cell cycle variables showed a consistent and significant decrease of the DNA synthetic S phase (P = 0.003) with corresponding induction of G2 phase arrest (P = 0.002) in FGFR1

5463

Cancer Res 2007; 67: (11). June 1, 2007

Cancer Research

Figure 1. Expression profiles of FGFR1 and FGFR2 and establishment of stable lines. A, RT-PCR. FGFR1 mRNA is detected in three thyroid carcinoma cell lines (TPC-1, WRO, 8505C), whereas only a faint signal can be seen in some normal thyroid (NT) samples. Negative and positive controls include omission of RT (RT( )) or cDNA (as indicated), respectively. FGFR2-IIIb amplification is detected in all normal thyroid tissues but not in three of four thyroid carcinoma cell lines including WRO cells. FGFR2-IIIc mRNA is not detectable in either normal or carcinoma cells. FRS2 mRNA is detected in both normal thyroid tissues and carcinoma cells. B, Western blotting. FGFR1 is detected in WRO and 8505C carcinoma cells. FGFR2 is detected in all normal thyroid tissues by Western blotting but not in three thyroid carcinoma cell lines (TPC-1, WRO, or 8505C). Low levels of FGFR2 are detectable in KTC-1 cells. FRS2 is not detectable in normal thyroid tissues but is up-regulated in all four thyroid carcinoma cell lines. C, endogenous expression of FGFR1 is detected in control WRO cells. Western blotting (left) and immunohistochemistry (right ) confirm down-regulation of FGFR1 in the siRNA/WRO clones. D, forced expression of FGFR2-IIIb was established in FGFR2-deficient WRO cells. Western blotting (left) and immunohistochemistry (right ) confirm sustained FGFR2-IIIb expression in the FGFR2-IIIb/WRO clones.

siRNA/WRO compared with control cells. Detailed cell cycle profiles are as follows: FGFR1 siRNA/WRO (G0-G1 phase, 41.9 F 1.4%; S phase, 20.0 F 0.7%; G2-M phase, 37.7 F 1.8%) and control (G0-G1 phase, 41.2 F 1.6%; S phase, 41.3 F 3.1%; G2-M phase, 14.2 F 2.9%). For assessment of tumor growth in vivo, we used a SCID mouse model of human thyroid cancer xenografts (14). Tumor volume was significantly reduced in FGFR1 siRNA/WRO tumors compared with control tumors (152.0 F 29.1 mm3 versus 390.6 F 107.9 mm3, P = 0.05; Fig. 2C). TUNEL assay identified no significant difference in apoptosis in the two types of xenografts [control 10.6/high power field (HPF) versus FGFR1 siRNA/WRO, 14.1/HPF, P = 0.6; Fig. 2D]. Restoration of FGFR2-IIIb attenuates RAS/BRAF/MAPK phosphorylation. Given the recognized importance of the

Cancer Res 2007; 67: (11). June 1, 2007

BRAF/MAPK signaling pathway in human thyroid tumorigenesis (3), we asked whether FGFR2-IIIb can affect this pathway. We chose serum as a stimulus source of multiple ligands. In control WRO cells, serum effectively induced BRAF serine phosphorylation. In contrast, WRO cells expressing FGFR2-IIIb failed to show BRAF phosphorylation (Fig. 3A, left). Immunoblotting analysis using a phospho-BRAF–specific antibody (Thr598/Ser601) also showed BRAF activation under serum stimulation in control cells but restoration of FGFR2-IIIb attenuated serum-induced BRAF phosphorylation. Consistent with its attenuating effect on BRAF activation, expression of FGFR2-IIIb diminished serum-induced MAPK activation (Fig. 3A, right). To determine whether FGFR2-IIIb can override the effect of the BRAF V600E mutation that is characteristic of nearly half of human thyroid carcinomas (3), we did these studies in 8505C cells that we confirmed to harbor this

5464


FGFR1 and FGFR2 in Thyroid Cancer

mutation by PCR direct sequencing. In these cells, expression of FGFR2-IIIb also attenuated serum-induced BRAF and MAPK phosphorylation (Fig. 3B). To determine the site of interruption of signaling mediated by FGFR2-IIIb, we did a Ras-GTP loading assay (Fig. 3C). This showed the ability of FGFR2-IIIb to significantly attenuate Ras activation, suggesting a level of control upstream of BRAF. Consistent with this finding, a corresponding BRAF kinase assay showed diminished intrinsic kinase activity in response to FGFR2-IIIb signaling (Fig. 3D). Restoration of FGFR2-IIIb impedes thyroid carcinoma growth and enhances apoptosis. To determine if FGFR2-IIIb can restrain tumor progression, we examined cell cycle progression by flow cytometry. Forced expression of FGFR2-IIIb significantly decreased the percentage of cell population in S phase (P = 0.006) with corresponding increases in G2-M phase (P = 0.009). Detailed cell cycle profiles are as follows: FGFR2-IIIb/WRO (G0-G1 phase, 34.8 F 0.9%; S phase, 28.4 F 3.3%; G2-M phase, 36.4 F 3.9%) and pcDNA/WRO (G0-G1 phase, 33.2 F 1.5%; S phase, 46.2 F 0.7%; G2-M phase, 16.8 F 1.1%). Cell proliferation as determined by an MTT assay showed similar reduction in cell growth in the presence of FGFR2-IIIb (Fig. 4A). Similarly, cell migration and invasion were significantly retarded in response to FGFR2-IIIb expression (Fig. 4B). In vivo studies using xenografts in SCID mice confirmed a significant reduction of tumor volume of cells with forced FGFR2-

IIIb expression compared with control FGFR2-IIIb–negative tumors (721.7 F 145.6 mm3 versus 1249.0 F 145.6 mm3, P = 0.03; Fig. 4C). TUNEL assay identified a significant increase in apoptosis in tumors expressing FGFR2-IIIb (pcDNA/WRO, 23.1/HPF versus FGFR2-IIIb/WRO, 93.6/HPF, P = 0.04; Fig. 4D). FGFR2-IIIb and FGFR1 compete with each other for FRS2 activation. To examine the functional signaling relationship between FGFR1 and FGFR2-IIIb, we tested their ability to engage FRS2. WRO cells expressing FGFR2-IIIb or their controls were treated after serum starvation with FGF1 or FGF7 (Fig. 5). As expected, the FGFR2-selective ligand, FGF7, activates FRS2 only in cells expressing FGFR2-IIIb (Fig. 5A). In contrast, the FRS2 response to FGF1 in WRO cells (which endogenously express FGFR1; Fig. 1A and B) transfected with FGFR2-IIIb was significantly diminished compared with control WRO cells. Forced expression of FGFR2-IIIb without or with FGF1 or FGF7 stimulation did not change the total amount of FRS2 in WRO cells. These studies were further extended in HEK293 cells in which FGFR1 and FGFR2-IIIb were independently introduced. In this system, FGF7 resulted in activation of FRS2 in the presence of FGFR2-IIIb alone (Fig. 5B). In contrast, the response to FGF7 was markedly blunted in the concomitant presence of FGFR1 and FGFR2-IIIb. Based on previous observations and these current data, we propose the following models (Fig. 5C). In the first model (i), FGFR2-IIIb directly mediates inhibitory signals in thyroid carcinoma cells. In the second model (ii), FGFR2-IIIb

Figure 2. FGFR1 silencing attenuates cell signaling and retards thyroid cancer cell growth. A, FGF1/FGFR1 cell signaling activation. After 24-h serum starvation, cells were incubated in serum-free medium with FGF1 (+) or vehicle ( ) for 15 min. In control WRO cells, FGF1 induces phosphorylation of MAPK (ERK1/2), AKT, and pRB. In contrast, silencing of FGFR1 prevents FGF1-induced responses in WRO cells. B, effect of FGFR1 silencing on thyroid carcinoma cell migration and invasion. FGFR1 siRNA/WRO cells and control cells were allowed to pass through Matrigel-coated (for invasion) or uncoated (for migration) filters with 8-Am pores. FGFR1 silencing significantly suppresses thyroid carcinoma cell invasion (P = 0.04). *, significant differences compared with controls. C, xenograft experiments involved the injection of FGFR1 siRNA/ WRO cells or control cells into the flank of SCID mice. Tumor volume is consistently lower in xenografted FGFR1 siRNA/WRO cells (n = 8) than in controls (n = 8; P = 0.05). D, TUNEL staining for apoptosis of xenografted tissue showed no significant differences in the two groups. Representative stains (top ) with quantitative assessment (bottom ) in bar graph.


5465

Cancer Res 2007; 67: (11). June 1, 2007

Cancer Research

Figure 3. FGFR2-IIIb expression attenuates RAS/BRAF/MAPK signaling in thyroid cancer cells. A, after 24 h of serum starvation, cells were incubated in 10% FBS or serum-free medium for 18 h (left ). Serine phosphorylation of immunoprecipitated BRAF increases after serum stimulation compared with serum-starved cells (right ). In contrast, FGFR2-IIIb–expressing WRO cells fail to show BRAF phosphorylation in response to serum. Western blotting using a phospho-specific BRAF antibody (Thr598/Ser601; p-BRAF ) also shows attenuation of serum-induced BRAF phosphorylation in FGFR2-IIIb–expressing WRO cells. Control cells show a MAPK response, which is attenuated in the presence of FGFR2-IIIb (right). B, 8505C cells that harbor BRAF V600E mutation were transfected with FGFR2-IIIb or pcDNA control vector as indicated. After overnight serum starvation, cells were stimulated with 10% FBS as indicated for assessment of BRAF and MAPK phosphorylation. Note down-regulated responses in FGFR2-IIIb transfectants. Columns, densitometric values; bars, SE. C, WRO control and FGFR2-IIIb– transfected cells were subjected to a Ras-GTP loading assay. GDP and GTP served as negative and positive controls, respectively. Columns, densitometric changes obtained from two independent experiments; bars, SE. D, WRO control and FGFR2-III–transfected cells as described under (A) were subjected to a BRAF kinase assay as detailed under Materials and Methods. Reaction products were resolved on 10% SDS-PAGE for immunoblotting with anti–phospho-MEK. Right, Negative (no BRAF ) and positive controls (active BRAF ). Columns, densitometric changes obtained from two independent experiments; bars, SE.

forms an inactive heterodimer complex with FGFR1, abrogating the tumor-promoting functions of FGFR1. In the third model (iii), FGFR2-IIIb recruits FRS2 and diverts signaling away from other tyrosine kinase receptors including FGFR1. FGFR2-IIIb is silenced through DNA methylation in thyroid carcinoma. Given the role for FGFR2-IIIb signaling in modulating thyroid cancer growth, we examined the mechanism underlying its down-regulation in thyroid carcinoma cells. Based on the observation that large CpG islands are present in the human promoter and noncoding region of FGFR2 (15), CpG methylation status was studied by MSP analysis. As shown in Fig. 6A, MSP showed strong amplification of CpG-unmethylated FGFR2 genomic DNA in normal thyroid tissues (which endogenously express FGFR2; Fig. 1A and B), whereas CpG-methylation of FGFR2 was observed in four of six cell lines with corresponding reduction in amplification of the unmethylated product. Next, we treated cells (WRO, TPC-1, NPA, and DRO) with the DNA demethylating agent 5¶-aza-deoxycytidine. Compared with vehicle treatment, 5¶-aza-deoxycytidine treatment restored unmethylated DNA and/or reduced the methylation-specific PCR product (Fig. 6B). Furthermore, DNA-demethylation restored FGFR2 protein expression without significantly affecting FGFR1 expression (Fig. 6C). Moreover, histone deacetylase inhibition by trichostatin A failed to alter FGFR1 or FGFR2 expression in these cells. These findings provide evidence of epigenetic silencing

Cancer Res 2007; 67: (11). June 1, 2007

through DNA methylation as a putative mechanism for FGFR2IIIb down-regulation.

Discussion In this study, we show divergent expression and actions of FGFR1 and FGFR2 in thyroid carcinoma cells, disclosing cancerpromoting roles of FGFR1 and cancer-suppressive properties of FGFR2. Down-regulated FGFR1 through siRNA knockdown decelerated carcinoma cell invasion and suppressed xenograft tumor growth in SCID mice. In marked contrast, reexpression of FGFR2IIIb interrupted signaling upstream of BRAF and consequently MAPK (ERK1/2) activation in thyroid carcinoma cells and significantly reduced tumor growth. Overexpression and/or gain-of-function mutations of FGFR1, FGFR3, and FGFR4 have been identified and implicated as oncogenes in a number of human neoplasms, including thyroid tumors (4, 7, 16). We have previously shown that transduction of a dominant-negative FGFR and pharmacologic FGFR tyrosine kinase inhibition using the PD173074 compound attenuates thyroid cancer cell proliferation (4). In the current study, we applied siRNA methodology for FGFR1-selective inhibition to clarify the oncogenic role of FGFR1 and to ask whether FGFR1 is a potential therapeutic target in thyroid carcinoma. Consistent with our hypothesis, FGFR1 silencing inhibited cell signal activation including AKT, cancer cell

5466



growth, and invasiveness of thyroid carcinoma cells. Invasionpromoting properties of FGFR1 are supported by several other studies, involving matrix metalloproteinase (MMP) regulation and cadherin modification. FGF1 increases MMP7 expression through FGFR1 signaling in normal prostate cells, and this is inhibited by an FGFR1-specific inhibitor or through dominant-negative FGFR1 transduction (17). In mouse cells, activation of FGFR1 results in invasive growth accompanied by induction of MMP3 and MMP9, and cleavage of adhesion factors including E-cadherin (18). FGFR1 signaling enhances N-cadherin signaling and activates MMP9 gene transcription to promote cellular invasion of human breast carcinoma cells (19). Our studies also support AKT involvement in FGFR1 signaling in thyroid cancer. Indeed, AKT is frequently overexpressed in thyroid carcinomas, in which it has been implicated in disease progression and invasion (20–22). These findings assign a role for FGFR1 in mediating invasive growth, thus providing a rationale for FGFR1-selective manipulation as a potential therapeutic target for human thyroid carcinoma.

We previously identified expression of FGFR2 in normal thyroid tissue (4). In the current study, we clarified that the FGFR2-IIIb isoform is expressed. FGFR2-IIIb expression is typically restricted to epithelial cells, whereas FGFR2-IIIc is characteristic of mesenchymal lineages (12, 23). Targeted disruption of FGFR2-IIIb causes agenesis of the lungs, anterior pituitary, thyroid, teeth, and limbs (24); in contrast, FGFR2-IIIc knockout mice show severe impairment of skull and bone development (25). Our data are consistent with these patterns of expression. The role of FGFR2 in tumorigenesis has recently gained interest. Down-regulation of FGFR2 has been noted with tumor progression in astrocytomas, bladder and prostatic carcinomas, pituitary tumors, and thyroid carcinomas (4, 5, 7, 15). Based on these observations, it is reasonable to propose a tumor-suppressive role for FGFR2. Our study clearly shows that forced FGFR2-IIIb expression significantly retards thyroid tumor progression while enhancing apoptosis (15, 26, 27). It should be noted, however, that FGFR2-IIIb down-regulation is not a universal feature in solid

Figure 4. FGFR2-IIIb restoration impedes thyroid cancer cell proliferation and retards tumor progression in mouse xenografts. A, WRO control and FGFR2-IIIb– transfected clones were subjected to MTT assays during 6 days of culture. Each curve represents an independently transfected clone (mean F SE, 8 wells per clone). Cell growth is significantly delayed by expression of FGFR2-IIIb; *, P < 0.05 versus pcDNA/WRO clone 1;c, P < 0.05 versus pcDNA/WRO clone 2. B, WRO control and FGFR2-IIIb–transfected clones as in (A) were subjected to invasion and migration assays as detailed under Materials and Methods. Cell migration and invasion were significantly retarded in response to FGFR2-IIIb expression; *, P < 0.05. C, xenograft experiments involved the injection of FGFR2-IIIb/WRO cells or control cells into the flank of SCID mice. Tumor volumes are consistently lower in xenografted FGFR2-IIIb–expressing WRO cells (n = 8) than in controls (n = 8, P = 0.04). D, tissue from mouse xenografts was examined by TUNEL staining for apoptosis assessment. Columns, mean of eight samples in each group; bars, SE.


5467

Cancer Res 2007; 67: (11). June 1, 2007

Cancer Research

tumors (28, 29). FGFR2 amplification was identified in gastric cancer cell lines with conspicuous absence of gain-of-function mutations (15, 30, 31). In fact, gastric carcinomas exhibit increased expression of the FGFR2-IIIb-C3 splice variant (also called KsamIIC3), in which the COOH terminus is shorter than that of wild-type FGFR2-IIIb and lacks the putative PLCg binding site (32). This C-terminally truncated FGFR2-IIIb-C3 isoform accelerates cancer cell growth and invasion (33, 34). Thus, alternate splicing of the C-terminal region of FGFR2-IIIb may clarify the controversy regarding FGFR2 expression and function in tumorigenesis. BRAF, located on chromosome 7q24, encodes a serine/threonine protein kinase that transduces regulatory signals through the MAPK signaling cascade. We show that restoration of FGFR2-IIIb inhibits BRAF phosphorylation, resulting in diminished MAPK activation even in the presence of activated BRAF signaling due to point mutation. Gain-of-function BRAF mutations, resulting in constitutive activation of MAPK signaling, are found in approximately one third of papillary thyroid carcinomas and one third of undifferentiated thyroid carcinomas (3, 35). BRAF is a putative therapeutic target, and transient siRNA–mediated down-regulation of mutant BRAF suppresses MAPK activation and cell growth in thyroid carcinoma cell lines (36). However, nearly half of human thyroid carcinomas and several thyroid cancer cell lines (including WRO cells) are negative for BRAF mutations (37, 38). Suppressive signals upstream of the BRAF kinase, as shown in the present study,

could represent an alternative or complementary therapeutic approach in thyroid carcinomas. The signaling mechanisms underlying FGFR2-IIIb antitumor action are summarized in three models proposed in Fig. 5C. In one model, FGFR2-IIIb directly mediates an inhibitory signal. Radvanyi’s group suggested that FGFR2-IIIb inhibits cancer cell growth by reducing insulin-like growth factor II via its C-terminal domain, independent of its tyrosine kinase domain (33). Although one group reported that restoration of FGFR2-IIIb inhibits FRS2 activation (27), we and others found that FGFR2-IIIb can engage FRS2 activation (39). These discrepancies may be due to cell specific responses or differences of FGFR profiles. In the second model, FGFR2-IIIb forms inactive heterodimers with FGFR1, abrogating the tumor-promoting functions of FGFR1. In the third model, FGFR2-IIIb and FGFR1 compete with each other for FRS2. FRS2 is up-regulated in thyroid carcinoma cells, and transduces signals for a number of kinases including FGFRs (6), RET and RET/PTC (40, 41), and NTRK1 and TRK-T1/T3 (42, 43). We note that RET and NTRK are not constitutively activated in WRO cells. We also observed stable amounts of FRS2 protein after FGFR2-IIIb reexpression. Therefore, the most plausible model is one in which FGFR2-IIIb sequesters limited amounts of FRS2 to divert signaling away from other receptor tyrosine kinases, including FGFR1 as well as the RET/RAS/BRAF/MAPK pathway, to retard tumor progression.

Figure 5. FGFR1 and FGFR2-IIIb compete with each other for FRS2 activation. A, WRO cells expressing FGFR2-IIIb or their controls (pcDNA) were treated after serum starvation with FGF1 or FGF7. As expected the FGFR2-selective FGF7 activates FRS2 only in the presence of FGFR2-IIIb. Note the diminished FRS2 response to FGF1 in cells expressing FGFR2-IIIb compared with those that do not. B, HEK293 cells were transfected with FGFR2-IIIb alone or with FGFR1. Note the robust effect of FGF7 on FRS2 activation in the presence of FGFR2-IIIb alone (left ) which is markedly diminished in the concomitant presence of FGFR1 (right ). C, the proposed models are based on previous observations and the current data. In model (i), FGFR2-IIIb directly mediates inhibitory signals in thyroid carcinoma cells. In model (ii), FGFR2-IIIb forms an inactive heterodimer complex with FGFR1, abrogating the tumor-promoting functions of FGFR1. In model (iii), FGFR2-IIIb recruits FRS2 and diverts signaling away from other tyrosine kinase receptors including FGFR1.

Cancer Res 2007; 67: (11). June 1, 2007

5468



Figure 6. DNA methylation regulates FGFR2 expression in thyroid carcinoma cells. A, the CpG methylation status of the FGFR2 gene was analyzed in normal human thyroids and KTC-1, TPC-1, WRO, 8505C, NPA, and DRO thyroid cancer cells as indicated. After bisulfite modification of genomic DNA, methylation-specific and unmethylation-specific primers were used for MSP analysis. Complete unmethylation of FGFR2 genomic DNA was observed in seven cases of normal thyroids, whereas CpG-methylation of FGFR2 genomic DNA was detected in TPC-1, WRO, NPA, and DRO cell lines with corresponding reduction in amplification of the unmethylated product. B, MSP analysis. The DNA demethylating agent 5¶-aza-deoxycytidine (Aza-dc ) treatment restored the unmethylated amplicon in WRO, TPC-1, and NPA cells and reduced the methylated product in DRO cells. C, Western blotting confirms that 5¶-aza-deoxycytidine treatment restored FGFR2 protein expression in WRO cells without significantly effecting FGFR1 expression. Histone deacetylase inhibition by trichostatin A treatment failed to significantly alter FGFR1 or FGFR2 expression.

Our data also point to a potential mechanism underlying the loss of expression of the FGFR2-IIIb tumor suppressor in thyroid cancer. One previous report implicated CpG methylation in the 5¶ region of the human FGFR2 gene in the process of FGFR2-IIIb down-regulation in human bladder carcinoma cell lines (15). We have directly shown this mechanism of FGFR2 gene silencing in thyroid cancer cells and shown restoration of FGFR2 protein expression after treatment with the DNA demethylating agent 5¶-aza-deoxycytidine. Epigenetic gene silencing of other tumorsuppressor genes, including E-cadherin, PTEN, and RASSF1A, and of differentiation-related genes, such as thyroid-stimulating hormone receptor and the sodium-iodide symporter, has been reported in thyroid cancers (3, 44). These data suggest that further studies should be pursued to determine the potential application of demethylating agents in the therapy of thyroid cancer. In conclusion, our data show a reciprocal expression profile of FGFR1 and FGFR2 in thyroid carcinomas. FGFR1 has tumor-

References 1. Brierley JD, Asa SL. Thyroid Cancer. In: Gospodarowicz MK, editor. Prognostic Factors in Cancer. New York: Wiley-Liss; 2001. 2. Asa SL, Bedard YC. Fine-needle aspiration cytology and histopathology. In: Clark OH, Noguchi S, editors. Thyroid Cancer. Diagnosis and Treatment. St. Louis: Quality Medical Publishing; 2000. p. 39–104. 3. Kondo T, Ezzat S, Asa SL. Pathogenetic mechanisms in thyroid follicular-cell neoplasia. Nat Rev Cancer 2006;6: 292–306. 4. St Bernard R, Zheng L, Liu W, Winer D, Asa SL, Ezzat S. Fibroblast growth factor receptors as molecular targets in thyroid carcinoma. Endocrinology 2005;146:1145–53. 5. Ezzat S, Asa SL. FGF receptor signaling at the crossroads of endocrine homeostasis and tumorigenesis. Horm Metab Res 2005;37:355–60. 6. Eswarakumar VP, Lax I, Schlessinger J. Cellular signaling by fibroblast growth factor receptors. Cytokine Growth Factor Rev 2005;16:139–49.


promoting actions, whereas FGFR2-IIIb plays a pivotal tumorsuppressive role. The growth suppressive functions are mediated upstream of the well-recognized RAS/BRAF/MAPK pathway. These findings underscore the complex network of the FGFR family of tyrosine kinases in modulating cancer cell growth and predict the need for highly selective inhibitors in the control of disease progression even in the context of distinct intragenic mutations.

Acknowledgments Received 12/7/2006; revised 3/5/2007; accepted 3/20/2007. Grant support: Canadian Institutes of Health Research grant MT-14404, Toronto Medical Laboratories, and Ministry of Education, Culture, Sports, Science and Technology, Japan grant 16-KAI-194 (T. Kondo). 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 with 18 U.S.C. Section 1734 solely to indicate this fact. We thank Kelvin So for his technical assistance.

7. Grose R, Dickson C. Fibroblast growth factor signaling in tumorigenesis. Cytokine Growth Factor Rev 2005;16: 179–86. 8. Thompson SD, Franklyn JA, Watkinson JC, Verhaeg JM, Sheppard MC, Eggo MC. Fibroblast growth factors 1 and 2 and fibroblast growth factor receptor 1 are elevated in thyroid hyperplasia. J Clin Endocrinol Metab 1998;83: 1336–41. 9. Becks GP, Logan A, Phillips ID, et al. Increase of basic fibroblast growth factor (FGF) and FGF receptor messenger RNA during rat thyroid hyperplasia: temporal changes and cellular distribution. J Endocrinol 1994; 142:325–38. 10. Davies EL, Ramsden JD, Cocks H, et al. Adenovirusmediated expression of dominant negative fibroblast growth factor (FGF) receptor 1 in thyroid cells blocks FGF effects and reduces goitrogenesis in mice. J Clin Endocrinol Metab 2003;88:4472–80. 11. Kurebayashi J, Tanaka K, Otsuki T, et al. All-transretinoic acid modulates expression levels of thyroglobulin and cytokines in a new human poorly differentiated

5469

papillary thyroid carcinoma cell line, KTC-1. J Clin Endocrinol Metab 2000;85:2889–96. 12. Orr-Urtreger A, Bedford MT, Burakova T, et al. Developmental localization of the splicing alternatives of fibroblast growth factor receptor-2 (FGFR2). Dev Biol 1993;158:475–86. 13. Feige JJ, Baird A. Glycosylation of the basic fibroblast growth factor receptor. The contribution of carbohydrate to receptor function. J Biol Chem 1988; 263:14023–9. 14. Dackiw AP, Ezzat S, Huang P, Liu W, Asa SL. Vitamin D3 administration induces nuclear p27 accumulation, restores differentiation, and reduces tumor burden in a mouse model of metastatic follicular thyroid cancer. Endocrinology 2004;145:5840–6. 15. Ricol D, Cappellen D, El Marjou A, et al. Tumour suppressive properties of fibroblast growth factor receptor 2-IIIb in human bladder cancer. Oncogene 1999;18:7234–43. 16. Ezzat S, Zheng L, Zhu XF, Wu GE, Asa SL. Targeted expression of a human pituitary tumor-derived isoform

Cancer Res 2007; 67: (11). June 1, 2007

Cancer Research of FGF receptor-4 recapitulates pituitary tumorigenesis. J Clin Invest 2002;109:69–78. 17. Udayakumar TS, Klein RD, Maliner MS, Nagle Rb, Bowden GT. Aberrant expression of fibroblast growth factor receptor-1 in prostate epithelial cells allows induction of promatrilysin expression by fibroblast growth factors. Int J Cancer 2001;91:187–92. 18. Xian W, Schwertfeger KL, Vargo-Gogola T, Rosen JM. Pleiotropic effects of FGFR1 on cell proliferation, survival, and migration in a 3D mammary epithelial cell model. J Cell Biol 2005;171:663–73. 19. Suyama K, Shapiro I, Guttman M, Hazan Rb. A signaling pathway leading to metastasis is controlled by N-cadherin and the FGF receptor. Cancer Cell 2002;2: 301–14. 20. Ringel MD, Hayre N, Saito J, et al. Overexpression and overactivation of Akt in thyroid carcinoma. Cancer Res 2001;61:6105–11. 21. Vasko V, Saji M, Hardy E, et al. Akt activation and localisation correlate with tumour invasion and oncogene expression in thyroid cancer. J Med Genet 2004;41: 161–70. 22. Shinohara M, Chung YJ, Saji M, Ringel MD. AKT in Thyroid Tumorigenesis and Progression. Endocrinology 2007;148:942–7. 23. Ornitz DM, Xu J, Colvin JS, et al. Receptor specificity of the fibroblast growth factor family. J Biol Chem 1996; 271:15292–7. 24. Revest JM, Spencer-Dene B, Kerr K, De Moerlooze L, Rosewell I, Dickson C. Fibroblast growth factor receptor 2-IIIb acts upstream of Shh and Fgf4 and is required for limb bud maintenance but not for the induction of Fgf8, Fgf10, Msx1, or Bmp4. Dev Biol 2001;231:47–62. 25. Eswarakumar VP, Monsonego-Ornan E, Pines M, Antopoulou I, Morriss-kay GM, Lonai P. The IIIc alternative of Fgfr2 is a positive regulator of bone formation. Development 2002;129:3783–93. 26. Feng S, Wang F, Matsubara A, Kan M, Mckeehan WL. Fibroblast growth factor receptor 2 limits and receptor 1

Cancer Res 2007; 67: (11). June 1, 2007

accelerates tumorigenicity of prostate epithelial cells. Cancer Res 1997;57:5369–78. 27. Zhang Y, Wang H, Toratani S, et al. Growth inhibition by keratinocyte growth factor receptor of human salivary adenocarcinoma cells through induction of differentiation and apoptosis. Proc Natl Acad Sci U S A 2001;98:11336–40. 28. Hattori Y, Itoh H, Uchino S, et al. Immunohistochemical detection of K-sam protein in stomach cancer. Clin Cancer Res 1996;2:1373–81. 29. Yamayoshi T, Nagayasu T, Matsumoto K, Abo T, Hishikawa Y, Koji T. Expression of keratinocyte growth factor/fibroblast growth factor-7 and its receptor in human lung cancer: correlation with tumour proliferative activity and patient prognosis. J Pathol 2004;204: 110–8. 30. Hattori Y, Odagiri H, Nakatani H, et al. K-sam, an amplified gene in stomach cancer, is a member of the heparin-binding growth factor receptor genes. Proc Natl Acad Sci U S A 1990;87:5983–7. 31. Hansen RM, Goriely A, Wall SA, Roberts IS, Wilkie AO. Fibroblast growth factor receptor 2, gain-offunction mutations, and tumourigenesis: investigating a potential link. J Pathol 2005;207:27–31. 32. Itoh H, Hattori Y, Sakamoto H, et al. Preferential alternative splicing in cancer generates a K-sam messenger RNA with higher transforming activity. Cancer Res 1994;54:3237–41. 33. Bernard-Pierrot I, Ricol D, Cassidy A, et al. Inhibition of human bladder tumour cell growth by fibroblast growth factor receptor 2b is independent of its kinase activity. Involvement of the carboxy-terminal region of the receptor. Oncogene 2004;23:9201–11. 34. Moffa AB, Tannheimer SL, Ethier SP. Transforming potential of alternatively spliced variants of fibroblast growth factor receptor 2 in human mammary epithelial cells. Mol Cancer Res 2004;2:643–52. 35. Kimura ET, Nikiforova MN, Zhu Z, Knauf JA, Nikiforov YE, Fagin JA. High prevalence of BRAF

5470

mutations in thyroid cancer: genetic evidence for constitutive activation of the RET/PTC-RAS-BRAF signaling pathway in papillary thyroid carcinoma. Cancer Res 2003;63:1454–7. 36. Salvatore G, De Falco V, Salerno P, et al. BRAF is a therapeutic target in aggressive thyroid carcinoma. Clin Cancer Res 2006;12:1623–9. 37. Namba H, Nakashima M, Hayashi T, et al. Clinical implication of hotspot BRAF mutation, V599E, in papillary thyroid cancers. J Clin Endocrinol Metab 2003;88:4393–7. 38. Xu X, Quiros RM, Gattuso P, et al. High prevalence of BRAF gene mutation in papillary thyroid carcinomas and thyroid tumor cell lines. Cancer Res 2003;63:4561–7. 39. Yasumoto H, Matsubara A, Mutaguchi K, Usui T, Mckeehan WL. Restoration of fibroblast growth factor receptor2 suppresses growth and tumorigenicity of malignant human prostate carcinoma PC-3 cells. Prostate 2004;61:236–42. 40. Kurokawa K, Iwashita T, Murakami H, Hayashi H, Kawai K, Takahashi M. Identification of SNT/FRS2 docking site on RET receptor tyrosine kinase and its role for signal transduction. Oncogene 2001;20:1929–38. 41. Melillo RM, Santoro M, Ong SH, et al. Docking protein FRS2 links the protein tyrosine kinase RET and its oncogenic forms with the mitogen-activated protein kinase signaling cascade. Mol Cell Biol 2001;21:4177–87. 42. Meakin SO, MacDonald JI, Gryz EA, Kubu CJ, Verdi JM. The signaling adapter FRS-2 competes with Shc for binding to the nerve growth factor receptor TrkA. A model for discriminating proliferation and differentiation. J Biol Chem 1999;274:9861–70. 43. Ranzi V, Meakin SO, Miranda C, Mondellini P, Pierrotti MA, Greco A. The signaling adapters fibroblast growth factor receptor substrate 2 and 3 are activated by the thyroid TRK oncoproteins. Endocrinology 2003; 144:922–8. 44. Xing M. Gene methylation in thyroid tumorigenesis. Endocrinology 2007;148:948–53.
