Fibroblast Growth Factor Receptors as Molecular Targets in Thyroid Carcinoma
http://www.100md.com
内分泌学杂志 2005年第3期
Department of Pathology (R.S.B., W.L., D.W., S.L.A.), University Health Network and University of Toronto, and the Department of Medicine (L.Z., S.E.), Mount Sinai Hospital and University of Toronto, Toronto, Ontario, Canada M5G 1X5; The Freeman Centre for Endocrine Oncology and the Ontario Cancer Institute (R.S.B., W.L., D.W., S.L.A., L.Z., S.E.), Toronto, Ontario Canada M5G 2M9
Address all correspondence and requests for reprints to: Dr. S. Ezzat, University of Toronto-Mount Sinai Hospital, 600 University Avenue #437, Toronto, Ontario, Canada M5G 1X5. E-mail: sezzat@mtsinai.on.ca.
Abstract
Several molecular abnormalities of potential therapeutic target value have been described in thyroid neoplastic transition. We report the expression of the fibroblast growth factor receptor family (FGFR-1–4) in normal thyroid tissues, human thyroid cancers of various types and behaviors, and cell lines representative of the spectrum of differentiation of tumors derived from follicular epithelial cells. FGFR-2 was the only receptor consistently detected in normal human thyroid tissue, and its expression diminished in all thyroid cancers and carcinoma cell lines, suggesting that it may have a protective role. FGFR-1 and FGFR-3 were expressed in most well-differentiated tumor types. FGFR-4, however, was expressed predominantly in aggressive tumor types and the most rapidly proliferative cell lines, indicating that it may promote the progression of these tumors. To specifically determine the function of FGFR-4 in thyroid carcinoma, gain- or loss-of-function studies were performed in cell lines representative of the spectrum of thyroid cancer behavior. Introduction of FGFR-4 resulted in enhanced cell proliferation, an effect that was more pronounced in cell lines derived from aggressive tumors than in those derived from more indolent neoplasms. Moreover, transduction of a dominant-negative FGFR attenuated cell proliferation in the aggressive poorly differentiated cell lines with no appreciable effect in well-differentiated cells. Pharmacologic FGFR-4 tyrosine kinase inhibition resulted in significant proliferation arrest in an aggressive cell line endogenously expressing the receptor. Furthermore, systemic administration of the FGFR tyrosine kinase inhibitor PD173074 resulted in significant inhibition of follicular thyroid carcinoma-derived cell growth in xenografted severe combined immunodeficient mice. These data indicate a role for FGFR-4 in human thyroid cancer cell progression and provide a rationale for FGFR manipulation as a potentially novel therapeutic approach.
Introduction
THYROID CANCER IS the most common form of endocrine malignancy with a significant contribution to endocrine cancer-related deaths (1). The most common tumor types are derived from follicular epithelial cells and show a spectrum of differentiation from the indolent well-differentiated papillary thyroid carcinoma (PTC) and follicular thyroid carcinoma (FTC) to more aggressive poorly differentiated carcinoma and the rare but rapidly lethal anaplastic (undifferentiated) thyroid carcinoma (2). Although several molecular abnormalities have been associated with the progression from normal thyroid tissue to thyroid carcinoma (3), this transformation is not well understood, and the identification of growth-controlling genes and their expression patterns in thyroid cancer progression has only recently gained interest (4).
One such gene family that is of potential interest is fibroblast growth factors (FGFs) and their receptors (FGFRs) (5). FGFs comprise a large family of heparin-binding growth factors that currently includes 23 members. FGFs are known to be expressed in thyroid cancer (6, 7, 8). These ligands signal through four high-affinity tyrosine kinase FGFRs (FGFR-1–4). Each receptor contains three Ig-like extracellular domains, a transmembrane region, an intracellular domain that contains a split tyrosine kinase, and a carboxy terminus (9). Components of the FGF system are potential oncoproteins as part of an autocrine/paracrine loop in which loss of regulation could potentially result in uncontrolled cell growth. FGFR signaling has also been implicated in differentiation and loss of these functions may play a role in neoplastic transformation in line with tumor suppressive functions.
Overexpression of FGFRs has been identified in malignancies of the brain (10), breast (11), prostate (12), skin (13), salivary gland (14), and thyroid gland (15). FGFR-1 has been reported to be up-regulated in thyroid follicular cells on goitrogen administration (16) and multinodular goiters (17). FGFR-1 expression has been detected in thyroid carcinoma (15, 18) and the expression of a dominant-negative (dn) FGFR-1 in thyroid cells reduces goitrogenesis in mice (19). FGFR-3 is expressed in PTC, and its overexpression in a human thyroid carcinoma cell line was shown to result in overgrowth of these cells in confluent cultures (20).
There are currently no studies that have systematically examined the expression of all FGFRs including FGFR-2 and FGFR-4 in thyroid carcinoma. We thus examined the expression of all members of the FGFR family of tyrosine kinases, FGFRs 1–4, in normal thyroid tissue, human thyroid tumors, and multiple thyroid carcinoma cell lines representative of the spectrum of biologic behavior. We used Western immunoblotting and immunohistochemistry to identify differences and similarities in patterns of FGFR protein expression. This information was used as the basis for gain- and loss-of-function approaches in vitro and in vivo to examine the potential value of targeting this family of tyrosine kinases in thyroid cancer therapy.
Materials and Methods
Thyroid tissue and cell lines
Normal human thyroid tissue was obtained after informed consent from seven different thyroidectomy specimens that had failed to identify any pathologic findings. Primary human thyroid carcinoma specimens were also obtained with informed consent and approval of the University Health Network Research Ethics Board. All tumors were characterized and classified according to accepted criteria (21, 22); they included 141 PTCs, 12 FTCs (six minimally invasive and five oncocytic) and six poorly differentiated and anaplastic carcinomas. Human thyroid carcinoma cell lines included two papillary carcinoma lines, TPC-1 (obtained from Dr. S. M. Jhiang, Ohio State University, Columbus, OH), and NPA (Dr. J. Fagin, University of Cincinnati, Cincinnati, OH), two follicular carcinoma lines WRO and MRO, and two lines derived from anaplastic carcinomas, ARO, and DRO (all kindly provided by Dr. J. Fagin, University of Cincinnati, Cincinnati, OH). These cell lines were originally established by Dr. G. Juillard (University of California, Los Angeles, Los Angeles, CA) except the TPC-1 cells, which were originally established by Dr. N. Satoh (Kanazawa University, Kanazawa, Japan).
Cell culture
TPC-1 cells were cultured in DMEM supplemented with 5% fetal calf serum, and 2 mmol/liter L-glutamine. WRO, NPA, ARO, DRO, and MRO cells were maintained in RPMI 1640 medium supplemented with 10% fetal calf serum, 2 mmol/liter L-glutamine, 1 mmol/liter sodium pyruvate, and 1x nonessential amino acid (Sigma-Aldrich Co. Ltd., Irvine, UK). Cells were cultured in 10-cm plates in a standard humidified incubator at 37 C in a 5% CO2-95% O2 atmosphere. Cell viability was assessed using trypan blue exclusion before and after all experiments.
Western blotting analysis
Cells were lysed in a lysis buffer [1x PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS)] with proteinase inhibitors [final concentrations 1 mM phenylmethylsulfonyl fluoride, 13.8 μgl/ml aprotinin (Sigma), and 1 mM sodium orthovanadate]. Samples were incubated on ice for 30 min and centrifuged at 10,000 x g for 15 min. Protein concentrations were determined using a protein assay (Bio-Rad Laboratories, Hercules, CA),. Equal amounts of protein (25 μg for cell lines and 50 μg for tissue samples) were solubilized in 2x SDS-sample buffer, separated on SDS-10% polyacrylamide gel and transferred to nitrocellulose. Membranes were blocked in Tris-buffered saline containing 1% Tween 20 and 5% nonfat dried milk and then incubated with primary antibody. Rabbit polyclonal antisera that recognize the cytoplasmic domains of FGFR-1 (SC-121), -2 (SC-122), -3 (SC-123), or -4 (SC-124) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were applied at a dilution of 1:1000. Human embryonic kidney (HEK) 293 cells transfected with FGFR-1, -2, -3, and -4 as previously described (23) served as positive controls, whereas empty-vector transfected cells served as negative controls. The specificity of each of the FGFR antibodies was further confirmed by examining lysates from HEK 293 cells transfected with the different FGFRs. Membranes were washed in Tris-buffered saline containing 1% Tween 20 and then incubated with horseradish peroxidase-conjugated goat antirabbit IgG secondary antibody at a dilution of 1:2000. Actin provided a loading control for protein and was detected using a mouse monoclonal antibody (Sigma) at a dilution of 1:500, and subsequently incubated with horseradish peroxidase-conjugated goat antimouse IgG secondary antibody at a dilution of 1:2000. Bands were visualized using the enhanced chemiluminescence detection system (Amersham, Buckinghamshire, UK).
Preparation of cell pellets
Semiconfluent plates were washed, gently scraped, and centrifuged into pellets that were coated in 2% bactoagar until solidified, fixed in 10% formalin, and embedded in paraffin.
Immunohistochemistry
Paraffin sections were dewaxed in five changes of xylene and brought down to water through graded alcohols. Heat-induced antigen retrieval was performed by heating the sections in a pressure cooker containing 10 mM citrate buffer (pH 6.0). Endogenous peroxidase and biotin activities were blocked respectively with 3% aqueous hydrogen peroxide and avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA) After blocking sections in normal goat serum for 10 min, sections were incubated with the FGFR antibodies (all at 1:300 dilution) overnight at room temperature in a moist chamber. Slides were washed in PBS and stained sequentially with biotinylated goat antirabbit IgG (Vector) and then with peroxidase-conjugated ultrastreptavidin labeling reagent (Signet Laboratories Inc., Seattle, WA). Color development was performed with freshly prepared NovaRed solution (Vector) and finally counterstained with Mayer’s hematoxylin. Sections were then dehydrated in alcohols, cleared in xylene, and mounted in Permount (Fisher, Burlington, Canada). Negative controls included omission of the primary antibody.
FGFR inhibition
To abrogate FGFR signaling, we used a genetic and a pharmacologic approach. The genetic approach used a soluble dominant-negative chimeric strategy. In addition to deletion of the cytoplasmic domain, the FGFR transmembrane domain is substituted with the human Ig heavy chain hinge and IgG1 Fc domains, creating a stable and secretable chimeric protein dnFGFR-HFc. This dnFGFR efficiently interferes with membrane-anchored FGFR signaling (24) and abrogates FGF-induced MAPK (Erk1/2) stimulation (23). To maintain stable expression in differentiating cells, we adopted an adenoviral approach for gene transfer. The dnFGFR-HFc or FGFR-4 cDNA (23) was subcloned into pACCMV-pLpA vector and cotransformed with pjM17 plasmid (25) into HEK 293 cells. Successful recombination between the two plasmids results in recombinant viruses encoding dnFGFR-HFc inserts and the control Ad-?gal. After plaque formation, viruses were amplified in HEK 293 cells and purified. Plaque-forming units were quantified using a commercial Adeno-X rapid titer kit (Boehringer; Indianapolis, IN). Chimeric dnFGFR-HFc protein expression was monitored by immunoblotting with anti-HFc antibody at 1:2000 (Dako, Carpinteria, CA). Inhibition of FGFR signaling was monitored by detection of phosphorylated forms of the Erk1/2 (1:1000; New England Biolabs, Beverly, MA).
The chemical approach used the small-molecular-weight pharmacological tyrosine kinase inhibitor PD173074 (Pfizer, Groton, CT) that selectively inhibits FGFR tyrosine kinase activity and autophosphorylation (26). Cells were seeded at 5 x 103 cells/well and maintained in the same serum-containing medium while treated with PD173074 at the following concentrations: 0, 0.25, 0.5, 1, 2, 4, 8, 16, and 32 μM diluted in sodium lactate buffer. Each PD173074 concentration was repeated in quadruplicate.
Adenoviral infection of thyroid carcinoma cell lines
To examine each type of thyroid carcinoma we chose TPC-1, MRO, and ARO, each of which displayed a distinct FGFR expression profile. TPC-1, MRO, and ARO cell lines were seeded in 96 multiwell plates each at variable densities (1 x 104, 5 x 103, 2.5 x 103, and 1.25 x 103) per well in the appropriate medium. Cells were infected with 107 viral particles per well 24 h after plating. The viral particle amount was selected based on levels of gene expression and cell viability in the first 24 h after infection. The viral particle load yielding FGFR-4 expression comparable with that in endogenously expressing ARO cells but without evidence of early cytotoxicity was chosen. This approach was also confirmed to result in successful infection of nearly 80% of cells as determined by immunohistochemistry. Cell lines were infected with an adenovirus expressing either FGFR-4, a soluble dnFGFR or ?-galactosidase.
Cell proliferation assay
Cell proliferation was assessed at 1, 3, and 6 d post infection using 3-(4,5-dimethylthiazol)-2,5-diphenyl tetrazolium (MTT) dye absorbance. The MTT assay was performed according to the manufacturer’s instructions (Roche Molecular Biochemicals, Indianapolis, IN). Each experimental condition was performed in at least triplicate wells on three separate occasions.
In vivo FGFR tumor inhibition assay
Ten female severe combined immunodeficient (SCID) mice, 9 wk old, were obtained from the Ontario Cancer Institute animal facility. Animal use and handling was in accordance with institutional guidelines and protocol approval. Mice were sc implanted on the abdominal right flank with 2.5 x 106 MRO cells suspended in PBS. Five days post implantation, mice were ip injected daily with either PD173074 50 mg/kg or an equal volume (0.2 ml) of the 50 mM sodium lactate buffer as vehicle. This dose was selected based on previous studies with this inhibitor (27). Five mice were included in each treatment group. Body weight and tumor size were recorded every 3–4 d. Tumor dimensions were measured using a vernier caliper (Fisher Scientific Ltd.). Tumor volumes were calculated as (length x width x depth)/2. Mice were killed 3 wk after tumor implantation, and complete autopsies were performed with microscopic examination of tissue from the site of implantation, lungs, and liver.
Statistical analysis
Data are presented as mean ± SE. In the experimental models, differences were assessed by the unpaired, two-sided t test. P < 0.05 was considered statistically significant. The analysis of human tumors applied Fisher’s exact t test.
Results
FGFR profile of human thyroid tissue
Only FGFR-2 protein was detected in normal thyroid tissue by Western blotting (Fig. 1). FGFR-1, FGFR-3, and FGFR-4 reactivity was not detected in normal tissues by this technique (Fig. 1). These data were also supported by immunohistochemistry (Fig. 2, A–D) in which again only FGFR-2 protein was detected in normal thyrocytes. Conversely, FGFR-1, -3, and -4 reactivity was noted in primary human thyroid tumors. Whereas FGFR-1 and FGFR-3 were found in hyperplastic goiters and 42 benign adenomas as well as in malignant thyroid lesions, FGFR-4 was restricted to more aggressive tumors including four of 12 follicular carcinomas (25%), 33 of 141 papillary carcinomas (23%) (Fig. 2E), and all six poorly differentiated and anaplastic thyroid carcinomas (100%) (Fig. 2F). Among the differentiated carcinomas, tumors with strong FGFR-4 positivity (Fig. 2E) exhibited a higher incidence of extrathyroidal extension (19%) compared with the group that was negative for FGFR-4 (9%) but this was not statistically significant (P = 0.19).
FIG. 1. Western blotting analysis of FGFR-1–4 expression in normal human thyroid tissue. Total cell lysates from seven thyroidectomy specimens with no evidence of pathology (lanes 1–7) were separated by SDS-PAGE and immunoblotted with antisera specific to human FGFR-1–4 as indicated. Positive controls were lysates of HEK 293 cells transfected with human FGFR-1–4 indicated as +; negative controls were lysates of HEK 293 cells transfected with empty vectors (not shown). Protein loading controls were immunoblotted with antiactin shown in the immediate lower panel of each pair. FGFR-2 is detected in four normal tissue specimens migrating slightly higher likely due to N-terminal glycosylation. FGFR-1, -3, and -4 are not detectable in normal thyroid tissue.
FIG. 2. Immunocytochemical detection of FGFR-1–4 in human thyroid tissue and tumors. Tissue from human thyroidectomy specimens was examined by immunohistochemistry using antibodies specific for human FGFR-1–4. Normal tissue is entirely negative for FGFR-1, -3, and -4 (A, C, and D) and stains positively only for FGFR-2 (B). Staining is restricted to thyrocytes, with negligible uptake by stromal supporting cells. In contrast, a papillary carcinoma that exhibited extrathyroidal extension and lymph node metastasis exhibits FGFR-4 positivity (E) and a poorly differentiated carcinoma with anaplastic dedifferentiation exhibits strong FGFR-4 positivity (F).
FGFR profile of thyroid carcinoma cell lines
FGFR-1 was detected strongly in TPC-1 cells and moderately in NPA and WRO cells as shown in Fig. 3. The more aggressive cell lines MRO, DRO, and ARO were negative for FGFR-1 (Fig. 3). In contrast to its expression in normal thyroid, FGFR-2 was not detectable in any of the neoplastic cell lines (Fig. 3). FGFR-3 was strong in TPC-1, NPA, and WRO cells. FGFR-4 was detectable and at relatively high levels in the aggressive ARO and MRO cell lines (Fig. 3).
FIG. 3. Western blotting analysis of FGFR-1–4 expression in thyroid cancer cell lines. Cell lysates from human thyroid cancer lines representative of the spectrum of biologic behavior were separated on SDS-PAGE and immunoblotted with specific antisera for FGFR-1–4. Positive controls were lysates of HEK 293 cells transfected with human FGFR-1–4 indicated as (+); negative controls were lysates of HEK 293 cells transfected with empty vectors (not shown). Protein loading controls were immunoblotted with antiactin shown in the immediate lower panel of each pair. FGFR-1 is detected in WRO, TPC-1, and NPA cells. FGFR-2 is not detectable in any of the cell lines. FGFR-3 is expressed in WRO, TPC-1, and NPA cells. FGFR-4 is detectable only in ARO and MRO cells but not the less aggressive cell lines.
The FGFR expression data derived from immunohistochemistry were largely consistent with those based on immunoblotting analyses and are summarized together in Table 1. Again, FGFR-1 was detected strongly in TPC-1 (Fig. 4A), WRO (Fig. 4C), and moderately in MRO cells (not shown). The cell lines DRO and ARO were negative for FGFR-1. FGFR-2 was not detected in any of the cell lines. FGFR-3 was detected in all cell lines with the exception of DRO; it was expressed weakly in ARO; moderately in TPC-1 (Fig. 4B), NPA, and WRO cells; and quite strongly in MRO cells (Fig. 4D). FGFR-4 was not detected in TPC-1, NPA, or WRO. As from the immunoblotting data, FGFR-4 was detected strongly in MRO and ARO cells (Fig. 4, E and F).
TABLE 1. Summary of FGFR-1-4 expression in normal thyroid and neoplastic thyroid cell lines
FIG. 4. Immunohistochemical detection of FGFR-1–4 in human thyroid carcinoma cell lines. Cell pellets of thyroid cancer cell lines representing the spectrum of biologic behavior were stained with specific antisera for each of the human FGFRs. A, TPC-1 cells stain strongly for FGFR-1. B, TPC-1 cells stain moderately for FGFR-3. C, WRO cells stain strongly for FGFR-1. D, WRO cells stain strongly for FGFR-3. E, MRO cells stain moderately for FGFR-4. F, ARO cells stain strongly for FGFR-4.
FGFR-4 gain and loss of function
Having determined that FGFR-4 is uniquely expressed in the more aggressive types of thyroid carcinoma but is absent in the well-differentiated less aggressive types, we elected to focus on the functional properties of this FGFR using gain- and loss-of function approaches. TPC-1 cells, which do not express detectable levels of FGFR-4, were infected with an adenovirus encoding FGFR-4 (Fig. 5A). The amount of virus required for levels of expression most closely reflective of endogenous FGFR-4 in ARO cells was used. Minimal if any effect on cell proliferation was evident after 1 and 3 d post infection (data not shown). However, by 6 d post infection, cells infected with FGFR-4 display a significantly higher proliferation rate, compared with control cells infected with the ?-galactosidase reporter. This effect was evident in the indolent TPC-1 cell line (2-fold increase; Fig. 5B) but was more pronounced in the MRO (Fig. 5C) and ARO cell lines (Fig. 5D) with a 4-fold increase in cell proliferation. In contrast, introduction of the soluble dnFGFR resulted in nearly 50% reduction in cell proliferation of MRO (Fig. 5C) and ARO cells (Fig. 5D) that each express only FGFR-4 but had no appreciable effect on TPC-1 cells that express FGFR-1 and FGFR-3 (Fig. 5B). Western immunoblotting detected the interruption of Erk1/2 phosphorylation in response to dnFGFR introduction in MRO and ARO cells (Fig. 5E).
FIG. 5. Effect of adenoviral-mediated transduction of FGFR-4 or dnFGFR on thyroid cancer cell proliferation. Thyroid cancer cell lines (2.5 x 103) of varying degrees of endogenous FGFR-4 expression were infected with adenoviruses (x 107 plaque-forming units) encoding either wild-type FGFR-4 (A) or a soluble dnFGFR or their ?-galactosidase control as indicated. Six days post viral infection, cell proliferation was examined by the colorimetric MTT assay. The results represent mean + SE of three separate experiments, each performed in at least triplicate wells in TPC-1 (B), MRO (C), and ARO cells (D). Note the significant increase in cell proliferation, an effect more pronounced in the aggressive ARO than the more indolent TPC-1 cell line. Note also the near 50% reduction in response to dnFGFR infection in ARO cells but lack of appreciable effect on TPC-1 cells that do not express FGFR-4. E, dnFGFR transduction abrogates Erk1/2 phosphorylation (upper panel) in ARO cells infected in duplicates as in C; corresponding total Erk1/2 is unchanged (lower panel).
Pharmacologic inhibition of FGFR in vitro
To further assess the effects of FGFR tyrosine kinase inhibition on thyroid cancer cell proliferation, the pharmacologic kinase inhibitor PD173074 was used to treat thyroid cancer cells. Changes in DNA content as a reflection of cell proliferation was compared in untreated cells and cells treated with different concentrations of the drug to identify a potential dose response. Cell proliferation of MRO cells that express FGFR-4 was lower when drug concentrations exceeded 8 μM (Fig. 6A). Cells treated with the highest concentration of PD173074 (32 μM) showed a nearly 90% decrease in cell proliferation (Fig. 6A). Similarly, ARO cells showed a steady decline in cell proliferation when PD173074 concentrations were 8 μM or greater, and by 32 μM cell proliferation was inhibited again by nearly 90% (data not shown).
FIG. 6. Effect of pharmacologic inhibition of FGFR on thyroid cancer cell growth. A, MRO cells were grown in the presence of increasing concentrations of the FGFR inhibitor PD173074 as indicated for 72 h with each concentration repeated in quadruplicate wells. Cell growth was examined by the MTT proliferation assay. The results are representative of triplicate experiments. B, Effect of the FGFR inhibitor PD173074 on MRO tumor cell growth in vivo. A total of 2.5 million MRO cells were injected under the right flank of xenografted SCID mice. Pharmacologic treatment (50 mg/kg) was commenced 5 d post cell injection to permit the development of a palpable tumor mass. Active compound or vehicle was administered ip 5 times per week as indicated. Shown are the effects on tumor volume expressed as a mean ± SE of measurements obtained from five mice in each treatment group. The results are representative of three independent experiments performed on separate occasions. Statistically significant differences (P < 0.001) in tumor volume were noted beginning at d 8 (3 d after injection). Corresponding total body animal weights are shown in C.
Because PD is an FGFR inhibitor that has been reported to inhibit multiple FGFRs (26, 27), we also examined cells that did not express FGFR-4. TPC-1 cells, which express only FGFR-1 and FGFR-3, showed similar dose-responsive inhibition of cell proliferation, reaching almost 90% at 32 μM PD173074 (not shown), consistent with the known growth-promoting effects of these FGFRs in thyroid (16, 17, 20).
Pharmacologic inhibition of FGFR in vivo
To determine the relationship between the in vitro findings and the potential role for FGFR inhibition on tumorous growth in vivo, we tested the effect of the small-molecular-weight FGFR inhibitor on MRO thyroid carcinoma cell growth in xenografted SCID mice (Fig. 6B). The MRO cell line was selected for in vivo studies as this FGFR-4-expressing cell line consistently formed tumors that were easily detectable and measurable. The systemic administration of PD173074 resulted in consistently smaller tumor sizes than control vehicle-treated animals with a reduction in tumor volume of approximately 55% (515 mm3± 89, compared with 1120 mm3± 102 in the vehicle-treated group; n = 10, P < 0.005) (Fig. 6B). Similarly, tumor weight was reduced by approximately 53% (0.33 g ± 0.07 vs. 0.69 ± 0.14 in the vehicle-treated group; P < 0.005) at the end of the 3-wk protocol. Body weights were not significantly different as a result of PD173074 treatment (Fig. 6C), suggesting that the reduction in tumor size was not due to reduction in body weight or as a result of a nonspecific toxic effect. Each of the treatments was well tolerated with no deleterious effect on food intake or body weight, compared with vehicle-treated animals. Autopsies revealed no significant pathology apart from tumor growth at the site of injection. The tumors had the typical morphology of FTC. However, tumors from PD173074-treated animals exhibited fewer mitoses and focal apoptosis as recognized by the presence of karyorrhexis and formation of apoptotic bodies, but no other morphologic alteration including alteration in angiogenesis was identified.
Discussion
We show significant differences in the expression of the FGFR family of tyrosine kinases among the different types of human thyroid carcinomas and compared with normal thyroid tissue. In particular, normal human thyroid tissue specimens but not primary tumors or neoplastic cell lines were found to express only FGFR-2, consistent with one previous study (28). The loss of FGFR-2 in transformed cells may indicate that FGFR-2 plays a protective role against thyroid cancer progression. Down-regulation of FGFR-2 is associated with malignant progression in the human prostate (29). Similarly, it has been shown that depression of the FGFR-2 signaling axis by expression of a dominant kinase-inactive construct of FGFR-2 contributes to the development of intraepithelial neoplasia in the mouse prostate (30). Collectively, these findings are supportive of a growth-restraining effect for FGFR-2. However, clearly detailed FGFR-2 gain- and loss-of-function studies in thyroid cancer are required.
FGFR-1 was expressed in benign and malignant thyroid tumors and the TPC-1, NPA, and WRO cell lines, which are representative of well-differentiated thyroid carcinomas. In agreement with this finding, Shingu et al. (15) detected FGFR-1 in PTCs using immunohistochemistry. The up-regulation of FGFR-1 in thyroid follicular cells exposed to goitrogens (16) and in multinodular goiters (17) as well as differentiated malignancies may indicate that FGFR-1 is a growth factor for differentiated thyroid cells. Similarly, FGFR-3 was strongly expressed in the differentiated thyroid cancer cell lines. In contrast, FGFR-1 and FGFR-3 were not expressed in the aggressive ARO line and not detected in DRO cells, which are derived from poorly differentiated tumors. MRO, which is a FTC cell line, is a well-differentiated yet faster-growing tumor than the other well-differentiated tumors. It is interesting to note that FGFR-1 was not detectable in MRO cells by Western blotting, and FGFR-3 was only weakly detected, suggesting that these two receptors may be markers of less aggressive tumor types. Clearly, further expression studies will be needed to determine whether FGFR-1 and FGFR-3 play a role in thyroid cell differentiation.
In marked contrast to the other FGFRs, FGFR-4 was strongly expressed in the more aggressive primary thyroid neoplasms and the anaplastic ARO tumor cell line and the rapidly proliferative MRO cell line. This receptor was not detected in hyperplastic lesions, less aggressive primary thyroid neoplasms, or any of the more differentiated cell lines. We thus chose to focus on the functional effects of FGFR-4 expression in thyroid cancer cells. To investigate the effect of gain of function, FGFR-4 was transduced using an adenovirus into cell lines of varying degrees of aggressive behavior. We show that cells infected with an adenovirus encoding FGFR-4 display a marked increase in cell proliferation, compared with control-infected cells. However, the proliferative effect of FGFR-4 was more pronounced in the aggressive ARO and MRO than the TPC-1 papillary carcinoma cells. Whereas clearly not sufficient to induce malignant transformation, these data are consistent with a role for FGFR-4 in mediating thyroid cancer progression.
Further evidence for the functional significance of FGFR-4 in thyroid cancer cell progression was obtained from adenoviral-mediated transduction of a dnFGFR chimeric receptor form. In addition to deletion of the cytoplasmic domain, the FGFR transmembrane domain is substituted with the human Ig heavy chain hinge and IgG1 Fc domains, thus creating a stable and secretable chimeric protein dnFGFR-HFc. This dnFGFR approach efficiently interferes with membrane-anchored FGFR signaling (24) and abrogates FGF1-induced MAPK (Erk1/2) stimulation without affecting endogenous FGFR expression (23). Again, the effect of FGFR antagonism was more pronounced in the aggressive ARO and MRO cells but was not evident in the indolent TPC-1 cells, consistent with the pattern of endogenous FGFR-4 expression in these lines.
To examine the potential pharmacologic utility of FGFR signal antagonism, we used the small-molecular-weight FGFR tyrosine kinase inhibitor PD173074. The FGFR selectivity of this tyrosine kinase inhibitor has been previously established (27) and is based on the high level of complementarity with the ATP-binding cavity of FGFRs (26, 27). Several other tyrosine kinases such as the insulin receptor, epidermal growth factor receptor, platelet-derived growth factor receptor-?, Src, and protein kinase C are not affected by this agent (26). PD173074 efficiently inhibits FGFR-2 as well as FGFR-4 and at similar concentrations demonstrated in breast cancer cells (26). In our studies, thyroid cancer cells treated with PD173074 showed a dose-dependent decrease in cell proliferation (31). This effect was similar in TPC-1 cells that express FGFR-1 and FGFR-3, as well as the more aggressive MRO cells that express FGFR-4. We therefore examined an in vivo model of aggressive thyroid carcinoma. SCID mice implanted with the aggressively growing MRO thyroid carcinoma cells that endogenously express significant amounts of FGFR-4 demonstrated a significant decrease in tumor size, compared with vehicle-treated animals. At the end of the study, tumor volume and weights were decreased by nearly half in PD173074-treated mice relative to control mice.
The molecular weights of the detected FGFRs in this study were mostly within the expected range. Minimal deviations, as with FGFR-3, are most likely the result of recognized N-terminal receptor glycosylation (32, 33). Thyroid carcinomas are well recognized to harbor genetic translocations (3). However, we could not identify any major molecular-weight changes corresponding to the recognized FGFR-1 (34) or FGFR-3 (35) translocations. Moreover, the possibilities of polymorphisms in the transmembrane domain of FGFR-4 as previously noted in breast carcinoma (36) will require specific elucidation in primary thyroid carcinomas.
In summary, we have demonstrated contrasting patterns in the expression of the tyrosine kinase family of FGFRs. In particular, normal adult thyroid tissue expresses mostly FGFR-2, which is curiously deficient in neoplastic cell lines of varying degrees of aggressiveness. Whereas FGFR-1 and FGFR-3 exhibited the largest degree of variation among the different tumor types, FGFR-4 expression appears to be restricted to the aggressive phenotype with conspicuous absence from the more indolent forms of thyroid carcinoma. Combined gain- and loss-of-function approaches provide further evidence of the contribution of FGFR-4 to thyroid cancer progression. Taken together, these data provide a framework emphasizing the significance of FGFR expression in neoplastic thyroid cells, rendering this family of tyrosine kinases as an attractive potential target for therapeutic intervention.
Acknowledgments
The authors gratefully acknowledge the technical assistance of Mr. Kelvin So and Ms. Ping Huang.
References
Brierley JD, Asa SL 2001 Thyroid cancer. In: Gospodarowicz MK, ed. Prognostic factors in cancer. New York: Wiley-Liss, Inc.; 195–208
Asa SL, Bedard YC 2000 Fine-needle aspiration cytology and histopathology. In: Clark OH, Noguchi S, eds. Thyroid cancer. St. Louis, MO: Diagnosis and Treatment Quality Medical Publishing, Inc.; 39–104
Fagin JA 1992 Genetic basis of endocrine disease 3. Molecular defects in thyroid gland neoplasia. J Clin Endocrinal Metab (Baltimore) 75:1398–1400
Sarlis NJ 2000 Expression patterns of cellular growth-controlling genes in non-medullary thyroid cancer: basic aspects. Rev Endocr Metab Disord 1:183–196
Jeffers M, LaRochelle WJ, Lichenstein HS 2002 Fibroblast growth factors in cancer: therapeutic possibilities. Expert Opin Ther Targets 6:469–482
Eggo MC, Hopkins JM, Franklyn JA, Johnson GD, Sanders DS, Sheppard MC 1995 Expression of fibroblast growth factors in thyroid cancer. J Clin Endocrinol Metab 80:1006–1011
Boelaert K, McCabe CJ, Tannahill LA, Gittoes NJ, Holder RL, Watkinson JC, Bradwell AR, Sheppard MC, Franklyn JA 2003 Pituitary tumor transforming gene and fibroblast growth factor-2 expression: potential prognostic indicators in differentiated thyroid cancer. J Clin Endocrinol Metab 88:2341–2347
Pasieka Z, Stepien H, Komorowski J, Kolomecki K, Kuzdak K 2003 Evaluation of the levels of bFGF, VEGF, sICAM-1, and sVCAM-1 in serum of patients with thyroid cancer. Recent Results Cancer Res 162:189–194
Givol D, Yayon A 1992 Complexity of FGF receptors: genetic basis for structural diversity and functional specificity. FASEB J 6:3362–3369
Morrison RS, Yamaguchi F, Saya H, Bruner JM, Yahanada AM, Donehower LA, Berger M 1994 Basic fibroblast growth factor and fibroblast growth factor receptor I are implicated in the growth of human astrocytomas. J Neurooncol 18:207–216
Yoshimura N, Sano H, Hashiramoto A, Yamada R, Nakajima H, Kondo M, Oka T 1998 The expression and localization of fibroblast growth factor-1 (FGF-1) and FGF receptor-1 (FGFR-1) in human breast cancer. Clin Immunol Immunopathol 89:28–34
Giri D, Ropiquet F, Ittmann M 1999 Alterations in expression of basic fibroblast growth factor (FGF) 2 and its receptor FGFR-1 in human prostate cancer. Clin Cancer Res 5:1063–1071
Ahmed NU, Ueda M, Ito A, Ohashi A, Funasaka Y, Ichihashi M 1997 Expression of fibroblast growth factor receptors in naevus-cell naevus and malignant melanoma. Melanoma Res 7:299–305
Myoken Y, Myoken Y, Okamoto T, Sato JD, Kan M, McKeehan WL, Nakahara M, Takada K 1996 Immunohistochemical study of overexpression of fibroblast growth factor-1 (FGF-1), FGF-2, and FGF receptor-1 in human malignant salivary gland tumours. J Pathol 178:429–436
Shingu K, Fujimori M, Ito K, Hama Y, Kasuga Y, Kobayashi S, Itoh N, Amano J 1998 Expression of fibroblast growth factor-2 and fibroblast growth factor receptor-1 in thyroid diseases: difference between neoplasms and hyperplastic lesions. Endocr J 45:35–43
Becks GP, Logan A, Phillips ID, Wang JF, Smith C, DeSousa D, Hill DJ 1994 Increase of basic fibroblast growth factor (FGF) and FGF receptor messenger RNA during rat thyroid hyperplasia: temporal changes and cellular distribution. J Endocrinol (Lond) 142:325–338
Thompson SD, Franklyn JA, Watkinson JC, Verhaeg JM, Sheppard MC, Eggo MC 1998 Fibroblast growth factors 1 and 2 and fibroblast growth factor receptor 1 are elevated in thyroid hyperplasia. J Clin Endocrinol Metab 83:1336–1341
Morikawa Y, Ishihara Y, Tohya K, Kakudo K, Seo MK, Matsuura N 1996 Expression of the fibroblast growth factor receptor-1 in human normal tissues and tumors determined by a new monoclonal antibody. Arch Pathol Lab Med 120:490–496
Davies EL, Ramsden JD, Cocks H, Searle PF, Watkinson JC, Ueno H, Logan A, Franklyn JA, Mautner V, Eggo MC 2003 Adenovirus-mediated 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 88:4472–4480
Onose H, Emoto N, Sugihara H, Shimizu K, Wakabayashi I 1999 Overexpression of fibroblast growth factor receptor 3 in a human thyroid carcinoma cell line results in overgrowth of the confluent cultures. Eur J Endocrinol 140:169–173
LiVolsi VA 1990 Surgical pathology of the thyroid. Philadelphia: W.B. Saunders Co.
Kleihues P, Burger PC, Scheithauer BW 1993 Histological typing of tumours of the central nervous system. World Health Organization international histological classification of tumours. 2nd ed. Berlin: Springer-Verlag
Ezzat S, Zheng L, Yu S, Asa SL 2001 A soluble dominant negative fibroblast growth factor receptor 4 isoform in human mcf-7 breast cancer cells. Biochem Biophys Res Commun 287:60–65
Celli G, LaRochelle WJ, Mackem S, Sharp R, Merlino G 1998 Soluble dominant-negative receptor uncovers essential roles for fibroblast growth factors in multi-organ induction and patterning. EMBO J 17:1642–1655
Gluzman Y, Ahrens B 1982 SV40 early mutants that are defective for viral DNA synthesis but competent for transformation of cultured rat and simian cells. Virology 123:78–92
Koziczak M, Holbro T, Hynes NE 2004 Blocking of FGFR signaling inhibits breast cancer cell proliferation through downregulation of D-type cyclins. Oncogene (Basinstoke) 23:3501–3508
Mohammadi M, Froum S, Hamby JM, Schroeder MC, Panek RL, Lu GH, Eliseenkova AV, Green D, Schlessinger J, Hubbard SR 1998 Crystal structure of an angiogenesis inhibitor bound to the FGF receptor tyrosine kinase domain. EMBO J 17:5896–5904
Hughes SE 1997 Differential expression of the fibroblast growth factor receptor (FGFR) multigene family in normal human adult tissues. J Histochem Cytochem 45:1005–1019
Naimi B, Latil A, Fournier G, Mangin P, Cussenot O, Berthon P 2002 Down-regulation of (IIIb) and (IIIc) isoforms of fibroblast growth factor receptor 2 (FGFR2) is associated with malignant progression in human prostate. Prostate 52:245–252
Jin C, McKeehan K, Guo W, Jauma S, Ittmann MM, Foster B, Greenberg NM, McKeehan WL, Wang F 2003 Cooperation between ectopic FGFR1 and depression of FGFR2 in induction of prostatic intraepithelial neoplasia in the mouse prostate. Cancer Res 63:8784–8790
Ezzat S, Huang P, Dackiw A, Asa SL, Dual inhibition of RET and FGFR4 restrains medullary thyroid cancer cell growth. Clin. Cancer Res, in press
Ezzat S, Zheng L, Zhu X-F, Wu GE, Asa SL 2002 Targeted expression of a human pituitary tumor-derived isoform of FGF receptor-4 recapitulates pituitary tumorigenesis. J Clin Invest 109:69–78
Tuominen H, Heikinheimo P, Loo BM, Kataja K, Oker-Blom C, Uutela M, Jalkanen M, Goldman A 2001 Expression and glycosylation studies of human fgf receptor 4. Protein Expr Purif 21:275–285
Grand EK, Grand FH, Chase AJ, Ross FM, Corcoran MM, Oscier DG, Cross NC 2004 Identification of a novel gene, FGFR1OP2, fused to FGFR1 in 8p11 myeloproliferative syndrome. Genes Chromosomes Cancer 40:78–83
Grand EK, Chase AJ, Heath C, Rahemtulla A, Cross NC 2004 Targeting FGFR3 in multiple myeloma: inhibition of t(4;14)-positive cells by SU5402 and PD173074. Leukemia (Baltimore) 18:962–966
Bange J, Prechtl D, Cheburkin Y, Specht K, Harbeck N, Schmitt M, Knyazeva T, Muller S, Gartner S, Sures I, Wang H, Imyanitov E, Haring HU, Knayzev P, Iacobelli S, Hofler H, Ullrich A 2002 Cancer progression and tumor cell motility are associated with the FGFR4 Arg(388) allele. Cancer Res (Baltimore) 62:840–847(Rosanne St. Bernard, Lei )
Address all correspondence and requests for reprints to: Dr. S. Ezzat, University of Toronto-Mount Sinai Hospital, 600 University Avenue #437, Toronto, Ontario, Canada M5G 1X5. E-mail: sezzat@mtsinai.on.ca.
Abstract
Several molecular abnormalities of potential therapeutic target value have been described in thyroid neoplastic transition. We report the expression of the fibroblast growth factor receptor family (FGFR-1–4) in normal thyroid tissues, human thyroid cancers of various types and behaviors, and cell lines representative of the spectrum of differentiation of tumors derived from follicular epithelial cells. FGFR-2 was the only receptor consistently detected in normal human thyroid tissue, and its expression diminished in all thyroid cancers and carcinoma cell lines, suggesting that it may have a protective role. FGFR-1 and FGFR-3 were expressed in most well-differentiated tumor types. FGFR-4, however, was expressed predominantly in aggressive tumor types and the most rapidly proliferative cell lines, indicating that it may promote the progression of these tumors. To specifically determine the function of FGFR-4 in thyroid carcinoma, gain- or loss-of-function studies were performed in cell lines representative of the spectrum of thyroid cancer behavior. Introduction of FGFR-4 resulted in enhanced cell proliferation, an effect that was more pronounced in cell lines derived from aggressive tumors than in those derived from more indolent neoplasms. Moreover, transduction of a dominant-negative FGFR attenuated cell proliferation in the aggressive poorly differentiated cell lines with no appreciable effect in well-differentiated cells. Pharmacologic FGFR-4 tyrosine kinase inhibition resulted in significant proliferation arrest in an aggressive cell line endogenously expressing the receptor. Furthermore, systemic administration of the FGFR tyrosine kinase inhibitor PD173074 resulted in significant inhibition of follicular thyroid carcinoma-derived cell growth in xenografted severe combined immunodeficient mice. These data indicate a role for FGFR-4 in human thyroid cancer cell progression and provide a rationale for FGFR manipulation as a potentially novel therapeutic approach.
Introduction
THYROID CANCER IS the most common form of endocrine malignancy with a significant contribution to endocrine cancer-related deaths (1). The most common tumor types are derived from follicular epithelial cells and show a spectrum of differentiation from the indolent well-differentiated papillary thyroid carcinoma (PTC) and follicular thyroid carcinoma (FTC) to more aggressive poorly differentiated carcinoma and the rare but rapidly lethal anaplastic (undifferentiated) thyroid carcinoma (2). Although several molecular abnormalities have been associated with the progression from normal thyroid tissue to thyroid carcinoma (3), this transformation is not well understood, and the identification of growth-controlling genes and their expression patterns in thyroid cancer progression has only recently gained interest (4).
One such gene family that is of potential interest is fibroblast growth factors (FGFs) and their receptors (FGFRs) (5). FGFs comprise a large family of heparin-binding growth factors that currently includes 23 members. FGFs are known to be expressed in thyroid cancer (6, 7, 8). These ligands signal through four high-affinity tyrosine kinase FGFRs (FGFR-1–4). Each receptor contains three Ig-like extracellular domains, a transmembrane region, an intracellular domain that contains a split tyrosine kinase, and a carboxy terminus (9). Components of the FGF system are potential oncoproteins as part of an autocrine/paracrine loop in which loss of regulation could potentially result in uncontrolled cell growth. FGFR signaling has also been implicated in differentiation and loss of these functions may play a role in neoplastic transformation in line with tumor suppressive functions.
Overexpression of FGFRs has been identified in malignancies of the brain (10), breast (11), prostate (12), skin (13), salivary gland (14), and thyroid gland (15). FGFR-1 has been reported to be up-regulated in thyroid follicular cells on goitrogen administration (16) and multinodular goiters (17). FGFR-1 expression has been detected in thyroid carcinoma (15, 18) and the expression of a dominant-negative (dn) FGFR-1 in thyroid cells reduces goitrogenesis in mice (19). FGFR-3 is expressed in PTC, and its overexpression in a human thyroid carcinoma cell line was shown to result in overgrowth of these cells in confluent cultures (20).
There are currently no studies that have systematically examined the expression of all FGFRs including FGFR-2 and FGFR-4 in thyroid carcinoma. We thus examined the expression of all members of the FGFR family of tyrosine kinases, FGFRs 1–4, in normal thyroid tissue, human thyroid tumors, and multiple thyroid carcinoma cell lines representative of the spectrum of biologic behavior. We used Western immunoblotting and immunohistochemistry to identify differences and similarities in patterns of FGFR protein expression. This information was used as the basis for gain- and loss-of-function approaches in vitro and in vivo to examine the potential value of targeting this family of tyrosine kinases in thyroid cancer therapy.
Materials and Methods
Thyroid tissue and cell lines
Normal human thyroid tissue was obtained after informed consent from seven different thyroidectomy specimens that had failed to identify any pathologic findings. Primary human thyroid carcinoma specimens were also obtained with informed consent and approval of the University Health Network Research Ethics Board. All tumors were characterized and classified according to accepted criteria (21, 22); they included 141 PTCs, 12 FTCs (six minimally invasive and five oncocytic) and six poorly differentiated and anaplastic carcinomas. Human thyroid carcinoma cell lines included two papillary carcinoma lines, TPC-1 (obtained from Dr. S. M. Jhiang, Ohio State University, Columbus, OH), and NPA (Dr. J. Fagin, University of Cincinnati, Cincinnati, OH), two follicular carcinoma lines WRO and MRO, and two lines derived from anaplastic carcinomas, ARO, and DRO (all kindly provided by Dr. J. Fagin, University of Cincinnati, Cincinnati, OH). These cell lines were originally established by Dr. G. Juillard (University of California, Los Angeles, Los Angeles, CA) except the TPC-1 cells, which were originally established by Dr. N. Satoh (Kanazawa University, Kanazawa, Japan).
Cell culture
TPC-1 cells were cultured in DMEM supplemented with 5% fetal calf serum, and 2 mmol/liter L-glutamine. WRO, NPA, ARO, DRO, and MRO cells were maintained in RPMI 1640 medium supplemented with 10% fetal calf serum, 2 mmol/liter L-glutamine, 1 mmol/liter sodium pyruvate, and 1x nonessential amino acid (Sigma-Aldrich Co. Ltd., Irvine, UK). Cells were cultured in 10-cm plates in a standard humidified incubator at 37 C in a 5% CO2-95% O2 atmosphere. Cell viability was assessed using trypan blue exclusion before and after all experiments.
Western blotting analysis
Cells were lysed in a lysis buffer [1x PBS, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS)] with proteinase inhibitors [final concentrations 1 mM phenylmethylsulfonyl fluoride, 13.8 μgl/ml aprotinin (Sigma), and 1 mM sodium orthovanadate]. Samples were incubated on ice for 30 min and centrifuged at 10,000 x g for 15 min. Protein concentrations were determined using a protein assay (Bio-Rad Laboratories, Hercules, CA),. Equal amounts of protein (25 μg for cell lines and 50 μg for tissue samples) were solubilized in 2x SDS-sample buffer, separated on SDS-10% polyacrylamide gel and transferred to nitrocellulose. Membranes were blocked in Tris-buffered saline containing 1% Tween 20 and 5% nonfat dried milk and then incubated with primary antibody. Rabbit polyclonal antisera that recognize the cytoplasmic domains of FGFR-1 (SC-121), -2 (SC-122), -3 (SC-123), or -4 (SC-124) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) were applied at a dilution of 1:1000. Human embryonic kidney (HEK) 293 cells transfected with FGFR-1, -2, -3, and -4 as previously described (23) served as positive controls, whereas empty-vector transfected cells served as negative controls. The specificity of each of the FGFR antibodies was further confirmed by examining lysates from HEK 293 cells transfected with the different FGFRs. Membranes were washed in Tris-buffered saline containing 1% Tween 20 and then incubated with horseradish peroxidase-conjugated goat antirabbit IgG secondary antibody at a dilution of 1:2000. Actin provided a loading control for protein and was detected using a mouse monoclonal antibody (Sigma) at a dilution of 1:500, and subsequently incubated with horseradish peroxidase-conjugated goat antimouse IgG secondary antibody at a dilution of 1:2000. Bands were visualized using the enhanced chemiluminescence detection system (Amersham, Buckinghamshire, UK).
Preparation of cell pellets
Semiconfluent plates were washed, gently scraped, and centrifuged into pellets that were coated in 2% bactoagar until solidified, fixed in 10% formalin, and embedded in paraffin.
Immunohistochemistry
Paraffin sections were dewaxed in five changes of xylene and brought down to water through graded alcohols. Heat-induced antigen retrieval was performed by heating the sections in a pressure cooker containing 10 mM citrate buffer (pH 6.0). Endogenous peroxidase and biotin activities were blocked respectively with 3% aqueous hydrogen peroxide and avidin/biotin blocking kit (Vector Laboratories, Burlingame, CA) After blocking sections in normal goat serum for 10 min, sections were incubated with the FGFR antibodies (all at 1:300 dilution) overnight at room temperature in a moist chamber. Slides were washed in PBS and stained sequentially with biotinylated goat antirabbit IgG (Vector) and then with peroxidase-conjugated ultrastreptavidin labeling reagent (Signet Laboratories Inc., Seattle, WA). Color development was performed with freshly prepared NovaRed solution (Vector) and finally counterstained with Mayer’s hematoxylin. Sections were then dehydrated in alcohols, cleared in xylene, and mounted in Permount (Fisher, Burlington, Canada). Negative controls included omission of the primary antibody.
FGFR inhibition
To abrogate FGFR signaling, we used a genetic and a pharmacologic approach. The genetic approach used a soluble dominant-negative chimeric strategy. In addition to deletion of the cytoplasmic domain, the FGFR transmembrane domain is substituted with the human Ig heavy chain hinge and IgG1 Fc domains, creating a stable and secretable chimeric protein dnFGFR-HFc. This dnFGFR efficiently interferes with membrane-anchored FGFR signaling (24) and abrogates FGF-induced MAPK (Erk1/2) stimulation (23). To maintain stable expression in differentiating cells, we adopted an adenoviral approach for gene transfer. The dnFGFR-HFc or FGFR-4 cDNA (23) was subcloned into pACCMV-pLpA vector and cotransformed with pjM17 plasmid (25) into HEK 293 cells. Successful recombination between the two plasmids results in recombinant viruses encoding dnFGFR-HFc inserts and the control Ad-?gal. After plaque formation, viruses were amplified in HEK 293 cells and purified. Plaque-forming units were quantified using a commercial Adeno-X rapid titer kit (Boehringer; Indianapolis, IN). Chimeric dnFGFR-HFc protein expression was monitored by immunoblotting with anti-HFc antibody at 1:2000 (Dako, Carpinteria, CA). Inhibition of FGFR signaling was monitored by detection of phosphorylated forms of the Erk1/2 (1:1000; New England Biolabs, Beverly, MA).
The chemical approach used the small-molecular-weight pharmacological tyrosine kinase inhibitor PD173074 (Pfizer, Groton, CT) that selectively inhibits FGFR tyrosine kinase activity and autophosphorylation (26). Cells were seeded at 5 x 103 cells/well and maintained in the same serum-containing medium while treated with PD173074 at the following concentrations: 0, 0.25, 0.5, 1, 2, 4, 8, 16, and 32 μM diluted in sodium lactate buffer. Each PD173074 concentration was repeated in quadruplicate.
Adenoviral infection of thyroid carcinoma cell lines
To examine each type of thyroid carcinoma we chose TPC-1, MRO, and ARO, each of which displayed a distinct FGFR expression profile. TPC-1, MRO, and ARO cell lines were seeded in 96 multiwell plates each at variable densities (1 x 104, 5 x 103, 2.5 x 103, and 1.25 x 103) per well in the appropriate medium. Cells were infected with 107 viral particles per well 24 h after plating. The viral particle amount was selected based on levels of gene expression and cell viability in the first 24 h after infection. The viral particle load yielding FGFR-4 expression comparable with that in endogenously expressing ARO cells but without evidence of early cytotoxicity was chosen. This approach was also confirmed to result in successful infection of nearly 80% of cells as determined by immunohistochemistry. Cell lines were infected with an adenovirus expressing either FGFR-4, a soluble dnFGFR or ?-galactosidase.
Cell proliferation assay
Cell proliferation was assessed at 1, 3, and 6 d post infection using 3-(4,5-dimethylthiazol)-2,5-diphenyl tetrazolium (MTT) dye absorbance. The MTT assay was performed according to the manufacturer’s instructions (Roche Molecular Biochemicals, Indianapolis, IN). Each experimental condition was performed in at least triplicate wells on three separate occasions.
In vivo FGFR tumor inhibition assay
Ten female severe combined immunodeficient (SCID) mice, 9 wk old, were obtained from the Ontario Cancer Institute animal facility. Animal use and handling was in accordance with institutional guidelines and protocol approval. Mice were sc implanted on the abdominal right flank with 2.5 x 106 MRO cells suspended in PBS. Five days post implantation, mice were ip injected daily with either PD173074 50 mg/kg or an equal volume (0.2 ml) of the 50 mM sodium lactate buffer as vehicle. This dose was selected based on previous studies with this inhibitor (27). Five mice were included in each treatment group. Body weight and tumor size were recorded every 3–4 d. Tumor dimensions were measured using a vernier caliper (Fisher Scientific Ltd.). Tumor volumes were calculated as (length x width x depth)/2. Mice were killed 3 wk after tumor implantation, and complete autopsies were performed with microscopic examination of tissue from the site of implantation, lungs, and liver.
Statistical analysis
Data are presented as mean ± SE. In the experimental models, differences were assessed by the unpaired, two-sided t test. P < 0.05 was considered statistically significant. The analysis of human tumors applied Fisher’s exact t test.
Results
FGFR profile of human thyroid tissue
Only FGFR-2 protein was detected in normal thyroid tissue by Western blotting (Fig. 1). FGFR-1, FGFR-3, and FGFR-4 reactivity was not detected in normal tissues by this technique (Fig. 1). These data were also supported by immunohistochemistry (Fig. 2, A–D) in which again only FGFR-2 protein was detected in normal thyrocytes. Conversely, FGFR-1, -3, and -4 reactivity was noted in primary human thyroid tumors. Whereas FGFR-1 and FGFR-3 were found in hyperplastic goiters and 42 benign adenomas as well as in malignant thyroid lesions, FGFR-4 was restricted to more aggressive tumors including four of 12 follicular carcinomas (25%), 33 of 141 papillary carcinomas (23%) (Fig. 2E), and all six poorly differentiated and anaplastic thyroid carcinomas (100%) (Fig. 2F). Among the differentiated carcinomas, tumors with strong FGFR-4 positivity (Fig. 2E) exhibited a higher incidence of extrathyroidal extension (19%) compared with the group that was negative for FGFR-4 (9%) but this was not statistically significant (P = 0.19).
FIG. 1. Western blotting analysis of FGFR-1–4 expression in normal human thyroid tissue. Total cell lysates from seven thyroidectomy specimens with no evidence of pathology (lanes 1–7) were separated by SDS-PAGE and immunoblotted with antisera specific to human FGFR-1–4 as indicated. Positive controls were lysates of HEK 293 cells transfected with human FGFR-1–4 indicated as +; negative controls were lysates of HEK 293 cells transfected with empty vectors (not shown). Protein loading controls were immunoblotted with antiactin shown in the immediate lower panel of each pair. FGFR-2 is detected in four normal tissue specimens migrating slightly higher likely due to N-terminal glycosylation. FGFR-1, -3, and -4 are not detectable in normal thyroid tissue.
FIG. 2. Immunocytochemical detection of FGFR-1–4 in human thyroid tissue and tumors. Tissue from human thyroidectomy specimens was examined by immunohistochemistry using antibodies specific for human FGFR-1–4. Normal tissue is entirely negative for FGFR-1, -3, and -4 (A, C, and D) and stains positively only for FGFR-2 (B). Staining is restricted to thyrocytes, with negligible uptake by stromal supporting cells. In contrast, a papillary carcinoma that exhibited extrathyroidal extension and lymph node metastasis exhibits FGFR-4 positivity (E) and a poorly differentiated carcinoma with anaplastic dedifferentiation exhibits strong FGFR-4 positivity (F).
FGFR profile of thyroid carcinoma cell lines
FGFR-1 was detected strongly in TPC-1 cells and moderately in NPA and WRO cells as shown in Fig. 3. The more aggressive cell lines MRO, DRO, and ARO were negative for FGFR-1 (Fig. 3). In contrast to its expression in normal thyroid, FGFR-2 was not detectable in any of the neoplastic cell lines (Fig. 3). FGFR-3 was strong in TPC-1, NPA, and WRO cells. FGFR-4 was detectable and at relatively high levels in the aggressive ARO and MRO cell lines (Fig. 3).
FIG. 3. Western blotting analysis of FGFR-1–4 expression in thyroid cancer cell lines. Cell lysates from human thyroid cancer lines representative of the spectrum of biologic behavior were separated on SDS-PAGE and immunoblotted with specific antisera for FGFR-1–4. Positive controls were lysates of HEK 293 cells transfected with human FGFR-1–4 indicated as (+); negative controls were lysates of HEK 293 cells transfected with empty vectors (not shown). Protein loading controls were immunoblotted with antiactin shown in the immediate lower panel of each pair. FGFR-1 is detected in WRO, TPC-1, and NPA cells. FGFR-2 is not detectable in any of the cell lines. FGFR-3 is expressed in WRO, TPC-1, and NPA cells. FGFR-4 is detectable only in ARO and MRO cells but not the less aggressive cell lines.
The FGFR expression data derived from immunohistochemistry were largely consistent with those based on immunoblotting analyses and are summarized together in Table 1. Again, FGFR-1 was detected strongly in TPC-1 (Fig. 4A), WRO (Fig. 4C), and moderately in MRO cells (not shown). The cell lines DRO and ARO were negative for FGFR-1. FGFR-2 was not detected in any of the cell lines. FGFR-3 was detected in all cell lines with the exception of DRO; it was expressed weakly in ARO; moderately in TPC-1 (Fig. 4B), NPA, and WRO cells; and quite strongly in MRO cells (Fig. 4D). FGFR-4 was not detected in TPC-1, NPA, or WRO. As from the immunoblotting data, FGFR-4 was detected strongly in MRO and ARO cells (Fig. 4, E and F).
TABLE 1. Summary of FGFR-1-4 expression in normal thyroid and neoplastic thyroid cell lines
FIG. 4. Immunohistochemical detection of FGFR-1–4 in human thyroid carcinoma cell lines. Cell pellets of thyroid cancer cell lines representing the spectrum of biologic behavior were stained with specific antisera for each of the human FGFRs. A, TPC-1 cells stain strongly for FGFR-1. B, TPC-1 cells stain moderately for FGFR-3. C, WRO cells stain strongly for FGFR-1. D, WRO cells stain strongly for FGFR-3. E, MRO cells stain moderately for FGFR-4. F, ARO cells stain strongly for FGFR-4.
FGFR-4 gain and loss of function
Having determined that FGFR-4 is uniquely expressed in the more aggressive types of thyroid carcinoma but is absent in the well-differentiated less aggressive types, we elected to focus on the functional properties of this FGFR using gain- and loss-of function approaches. TPC-1 cells, which do not express detectable levels of FGFR-4, were infected with an adenovirus encoding FGFR-4 (Fig. 5A). The amount of virus required for levels of expression most closely reflective of endogenous FGFR-4 in ARO cells was used. Minimal if any effect on cell proliferation was evident after 1 and 3 d post infection (data not shown). However, by 6 d post infection, cells infected with FGFR-4 display a significantly higher proliferation rate, compared with control cells infected with the ?-galactosidase reporter. This effect was evident in the indolent TPC-1 cell line (2-fold increase; Fig. 5B) but was more pronounced in the MRO (Fig. 5C) and ARO cell lines (Fig. 5D) with a 4-fold increase in cell proliferation. In contrast, introduction of the soluble dnFGFR resulted in nearly 50% reduction in cell proliferation of MRO (Fig. 5C) and ARO cells (Fig. 5D) that each express only FGFR-4 but had no appreciable effect on TPC-1 cells that express FGFR-1 and FGFR-3 (Fig. 5B). Western immunoblotting detected the interruption of Erk1/2 phosphorylation in response to dnFGFR introduction in MRO and ARO cells (Fig. 5E).
FIG. 5. Effect of adenoviral-mediated transduction of FGFR-4 or dnFGFR on thyroid cancer cell proliferation. Thyroid cancer cell lines (2.5 x 103) of varying degrees of endogenous FGFR-4 expression were infected with adenoviruses (x 107 plaque-forming units) encoding either wild-type FGFR-4 (A) or a soluble dnFGFR or their ?-galactosidase control as indicated. Six days post viral infection, cell proliferation was examined by the colorimetric MTT assay. The results represent mean + SE of three separate experiments, each performed in at least triplicate wells in TPC-1 (B), MRO (C), and ARO cells (D). Note the significant increase in cell proliferation, an effect more pronounced in the aggressive ARO than the more indolent TPC-1 cell line. Note also the near 50% reduction in response to dnFGFR infection in ARO cells but lack of appreciable effect on TPC-1 cells that do not express FGFR-4. E, dnFGFR transduction abrogates Erk1/2 phosphorylation (upper panel) in ARO cells infected in duplicates as in C; corresponding total Erk1/2 is unchanged (lower panel).
Pharmacologic inhibition of FGFR in vitro
To further assess the effects of FGFR tyrosine kinase inhibition on thyroid cancer cell proliferation, the pharmacologic kinase inhibitor PD173074 was used to treat thyroid cancer cells. Changes in DNA content as a reflection of cell proliferation was compared in untreated cells and cells treated with different concentrations of the drug to identify a potential dose response. Cell proliferation of MRO cells that express FGFR-4 was lower when drug concentrations exceeded 8 μM (Fig. 6A). Cells treated with the highest concentration of PD173074 (32 μM) showed a nearly 90% decrease in cell proliferation (Fig. 6A). Similarly, ARO cells showed a steady decline in cell proliferation when PD173074 concentrations were 8 μM or greater, and by 32 μM cell proliferation was inhibited again by nearly 90% (data not shown).
FIG. 6. Effect of pharmacologic inhibition of FGFR on thyroid cancer cell growth. A, MRO cells were grown in the presence of increasing concentrations of the FGFR inhibitor PD173074 as indicated for 72 h with each concentration repeated in quadruplicate wells. Cell growth was examined by the MTT proliferation assay. The results are representative of triplicate experiments. B, Effect of the FGFR inhibitor PD173074 on MRO tumor cell growth in vivo. A total of 2.5 million MRO cells were injected under the right flank of xenografted SCID mice. Pharmacologic treatment (50 mg/kg) was commenced 5 d post cell injection to permit the development of a palpable tumor mass. Active compound or vehicle was administered ip 5 times per week as indicated. Shown are the effects on tumor volume expressed as a mean ± SE of measurements obtained from five mice in each treatment group. The results are representative of three independent experiments performed on separate occasions. Statistically significant differences (P < 0.001) in tumor volume were noted beginning at d 8 (3 d after injection). Corresponding total body animal weights are shown in C.
Because PD is an FGFR inhibitor that has been reported to inhibit multiple FGFRs (26, 27), we also examined cells that did not express FGFR-4. TPC-1 cells, which express only FGFR-1 and FGFR-3, showed similar dose-responsive inhibition of cell proliferation, reaching almost 90% at 32 μM PD173074 (not shown), consistent with the known growth-promoting effects of these FGFRs in thyroid (16, 17, 20).
Pharmacologic inhibition of FGFR in vivo
To determine the relationship between the in vitro findings and the potential role for FGFR inhibition on tumorous growth in vivo, we tested the effect of the small-molecular-weight FGFR inhibitor on MRO thyroid carcinoma cell growth in xenografted SCID mice (Fig. 6B). The MRO cell line was selected for in vivo studies as this FGFR-4-expressing cell line consistently formed tumors that were easily detectable and measurable. The systemic administration of PD173074 resulted in consistently smaller tumor sizes than control vehicle-treated animals with a reduction in tumor volume of approximately 55% (515 mm3± 89, compared with 1120 mm3± 102 in the vehicle-treated group; n = 10, P < 0.005) (Fig. 6B). Similarly, tumor weight was reduced by approximately 53% (0.33 g ± 0.07 vs. 0.69 ± 0.14 in the vehicle-treated group; P < 0.005) at the end of the 3-wk protocol. Body weights were not significantly different as a result of PD173074 treatment (Fig. 6C), suggesting that the reduction in tumor size was not due to reduction in body weight or as a result of a nonspecific toxic effect. Each of the treatments was well tolerated with no deleterious effect on food intake or body weight, compared with vehicle-treated animals. Autopsies revealed no significant pathology apart from tumor growth at the site of injection. The tumors had the typical morphology of FTC. However, tumors from PD173074-treated animals exhibited fewer mitoses and focal apoptosis as recognized by the presence of karyorrhexis and formation of apoptotic bodies, but no other morphologic alteration including alteration in angiogenesis was identified.
Discussion
We show significant differences in the expression of the FGFR family of tyrosine kinases among the different types of human thyroid carcinomas and compared with normal thyroid tissue. In particular, normal human thyroid tissue specimens but not primary tumors or neoplastic cell lines were found to express only FGFR-2, consistent with one previous study (28). The loss of FGFR-2 in transformed cells may indicate that FGFR-2 plays a protective role against thyroid cancer progression. Down-regulation of FGFR-2 is associated with malignant progression in the human prostate (29). Similarly, it has been shown that depression of the FGFR-2 signaling axis by expression of a dominant kinase-inactive construct of FGFR-2 contributes to the development of intraepithelial neoplasia in the mouse prostate (30). Collectively, these findings are supportive of a growth-restraining effect for FGFR-2. However, clearly detailed FGFR-2 gain- and loss-of-function studies in thyroid cancer are required.
FGFR-1 was expressed in benign and malignant thyroid tumors and the TPC-1, NPA, and WRO cell lines, which are representative of well-differentiated thyroid carcinomas. In agreement with this finding, Shingu et al. (15) detected FGFR-1 in PTCs using immunohistochemistry. The up-regulation of FGFR-1 in thyroid follicular cells exposed to goitrogens (16) and in multinodular goiters (17) as well as differentiated malignancies may indicate that FGFR-1 is a growth factor for differentiated thyroid cells. Similarly, FGFR-3 was strongly expressed in the differentiated thyroid cancer cell lines. In contrast, FGFR-1 and FGFR-3 were not expressed in the aggressive ARO line and not detected in DRO cells, which are derived from poorly differentiated tumors. MRO, which is a FTC cell line, is a well-differentiated yet faster-growing tumor than the other well-differentiated tumors. It is interesting to note that FGFR-1 was not detectable in MRO cells by Western blotting, and FGFR-3 was only weakly detected, suggesting that these two receptors may be markers of less aggressive tumor types. Clearly, further expression studies will be needed to determine whether FGFR-1 and FGFR-3 play a role in thyroid cell differentiation.
In marked contrast to the other FGFRs, FGFR-4 was strongly expressed in the more aggressive primary thyroid neoplasms and the anaplastic ARO tumor cell line and the rapidly proliferative MRO cell line. This receptor was not detected in hyperplastic lesions, less aggressive primary thyroid neoplasms, or any of the more differentiated cell lines. We thus chose to focus on the functional effects of FGFR-4 expression in thyroid cancer cells. To investigate the effect of gain of function, FGFR-4 was transduced using an adenovirus into cell lines of varying degrees of aggressive behavior. We show that cells infected with an adenovirus encoding FGFR-4 display a marked increase in cell proliferation, compared with control-infected cells. However, the proliferative effect of FGFR-4 was more pronounced in the aggressive ARO and MRO than the TPC-1 papillary carcinoma cells. Whereas clearly not sufficient to induce malignant transformation, these data are consistent with a role for FGFR-4 in mediating thyroid cancer progression.
Further evidence for the functional significance of FGFR-4 in thyroid cancer cell progression was obtained from adenoviral-mediated transduction of a dnFGFR chimeric receptor form. In addition to deletion of the cytoplasmic domain, the FGFR transmembrane domain is substituted with the human Ig heavy chain hinge and IgG1 Fc domains, thus creating a stable and secretable chimeric protein dnFGFR-HFc. This dnFGFR approach efficiently interferes with membrane-anchored FGFR signaling (24) and abrogates FGF1-induced MAPK (Erk1/2) stimulation without affecting endogenous FGFR expression (23). Again, the effect of FGFR antagonism was more pronounced in the aggressive ARO and MRO cells but was not evident in the indolent TPC-1 cells, consistent with the pattern of endogenous FGFR-4 expression in these lines.
To examine the potential pharmacologic utility of FGFR signal antagonism, we used the small-molecular-weight FGFR tyrosine kinase inhibitor PD173074. The FGFR selectivity of this tyrosine kinase inhibitor has been previously established (27) and is based on the high level of complementarity with the ATP-binding cavity of FGFRs (26, 27). Several other tyrosine kinases such as the insulin receptor, epidermal growth factor receptor, platelet-derived growth factor receptor-?, Src, and protein kinase C are not affected by this agent (26). PD173074 efficiently inhibits FGFR-2 as well as FGFR-4 and at similar concentrations demonstrated in breast cancer cells (26). In our studies, thyroid cancer cells treated with PD173074 showed a dose-dependent decrease in cell proliferation (31). This effect was similar in TPC-1 cells that express FGFR-1 and FGFR-3, as well as the more aggressive MRO cells that express FGFR-4. We therefore examined an in vivo model of aggressive thyroid carcinoma. SCID mice implanted with the aggressively growing MRO thyroid carcinoma cells that endogenously express significant amounts of FGFR-4 demonstrated a significant decrease in tumor size, compared with vehicle-treated animals. At the end of the study, tumor volume and weights were decreased by nearly half in PD173074-treated mice relative to control mice.
The molecular weights of the detected FGFRs in this study were mostly within the expected range. Minimal deviations, as with FGFR-3, are most likely the result of recognized N-terminal receptor glycosylation (32, 33). Thyroid carcinomas are well recognized to harbor genetic translocations (3). However, we could not identify any major molecular-weight changes corresponding to the recognized FGFR-1 (34) or FGFR-3 (35) translocations. Moreover, the possibilities of polymorphisms in the transmembrane domain of FGFR-4 as previously noted in breast carcinoma (36) will require specific elucidation in primary thyroid carcinomas.
In summary, we have demonstrated contrasting patterns in the expression of the tyrosine kinase family of FGFRs. In particular, normal adult thyroid tissue expresses mostly FGFR-2, which is curiously deficient in neoplastic cell lines of varying degrees of aggressiveness. Whereas FGFR-1 and FGFR-3 exhibited the largest degree of variation among the different tumor types, FGFR-4 expression appears to be restricted to the aggressive phenotype with conspicuous absence from the more indolent forms of thyroid carcinoma. Combined gain- and loss-of-function approaches provide further evidence of the contribution of FGFR-4 to thyroid cancer progression. Taken together, these data provide a framework emphasizing the significance of FGFR expression in neoplastic thyroid cells, rendering this family of tyrosine kinases as an attractive potential target for therapeutic intervention.
Acknowledgments
The authors gratefully acknowledge the technical assistance of Mr. Kelvin So and Ms. Ping Huang.
References
Brierley JD, Asa SL 2001 Thyroid cancer. In: Gospodarowicz MK, ed. Prognostic factors in cancer. New York: Wiley-Liss, Inc.; 195–208
Asa SL, Bedard YC 2000 Fine-needle aspiration cytology and histopathology. In: Clark OH, Noguchi S, eds. Thyroid cancer. St. Louis, MO: Diagnosis and Treatment Quality Medical Publishing, Inc.; 39–104
Fagin JA 1992 Genetic basis of endocrine disease 3. Molecular defects in thyroid gland neoplasia. J Clin Endocrinal Metab (Baltimore) 75:1398–1400
Sarlis NJ 2000 Expression patterns of cellular growth-controlling genes in non-medullary thyroid cancer: basic aspects. Rev Endocr Metab Disord 1:183–196
Jeffers M, LaRochelle WJ, Lichenstein HS 2002 Fibroblast growth factors in cancer: therapeutic possibilities. Expert Opin Ther Targets 6:469–482
Eggo MC, Hopkins JM, Franklyn JA, Johnson GD, Sanders DS, Sheppard MC 1995 Expression of fibroblast growth factors in thyroid cancer. J Clin Endocrinol Metab 80:1006–1011
Boelaert K, McCabe CJ, Tannahill LA, Gittoes NJ, Holder RL, Watkinson JC, Bradwell AR, Sheppard MC, Franklyn JA 2003 Pituitary tumor transforming gene and fibroblast growth factor-2 expression: potential prognostic indicators in differentiated thyroid cancer. J Clin Endocrinol Metab 88:2341–2347
Pasieka Z, Stepien H, Komorowski J, Kolomecki K, Kuzdak K 2003 Evaluation of the levels of bFGF, VEGF, sICAM-1, and sVCAM-1 in serum of patients with thyroid cancer. Recent Results Cancer Res 162:189–194
Givol D, Yayon A 1992 Complexity of FGF receptors: genetic basis for structural diversity and functional specificity. FASEB J 6:3362–3369
Morrison RS, Yamaguchi F, Saya H, Bruner JM, Yahanada AM, Donehower LA, Berger M 1994 Basic fibroblast growth factor and fibroblast growth factor receptor I are implicated in the growth of human astrocytomas. J Neurooncol 18:207–216
Yoshimura N, Sano H, Hashiramoto A, Yamada R, Nakajima H, Kondo M, Oka T 1998 The expression and localization of fibroblast growth factor-1 (FGF-1) and FGF receptor-1 (FGFR-1) in human breast cancer. Clin Immunol Immunopathol 89:28–34
Giri D, Ropiquet F, Ittmann M 1999 Alterations in expression of basic fibroblast growth factor (FGF) 2 and its receptor FGFR-1 in human prostate cancer. Clin Cancer Res 5:1063–1071
Ahmed NU, Ueda M, Ito A, Ohashi A, Funasaka Y, Ichihashi M 1997 Expression of fibroblast growth factor receptors in naevus-cell naevus and malignant melanoma. Melanoma Res 7:299–305
Myoken Y, Myoken Y, Okamoto T, Sato JD, Kan M, McKeehan WL, Nakahara M, Takada K 1996 Immunohistochemical study of overexpression of fibroblast growth factor-1 (FGF-1), FGF-2, and FGF receptor-1 in human malignant salivary gland tumours. J Pathol 178:429–436
Shingu K, Fujimori M, Ito K, Hama Y, Kasuga Y, Kobayashi S, Itoh N, Amano J 1998 Expression of fibroblast growth factor-2 and fibroblast growth factor receptor-1 in thyroid diseases: difference between neoplasms and hyperplastic lesions. Endocr J 45:35–43
Becks GP, Logan A, Phillips ID, Wang JF, Smith C, DeSousa D, Hill DJ 1994 Increase of basic fibroblast growth factor (FGF) and FGF receptor messenger RNA during rat thyroid hyperplasia: temporal changes and cellular distribution. J Endocrinol (Lond) 142:325–338
Thompson SD, Franklyn JA, Watkinson JC, Verhaeg JM, Sheppard MC, Eggo MC 1998 Fibroblast growth factors 1 and 2 and fibroblast growth factor receptor 1 are elevated in thyroid hyperplasia. J Clin Endocrinol Metab 83:1336–1341
Morikawa Y, Ishihara Y, Tohya K, Kakudo K, Seo MK, Matsuura N 1996 Expression of the fibroblast growth factor receptor-1 in human normal tissues and tumors determined by a new monoclonal antibody. Arch Pathol Lab Med 120:490–496
Davies EL, Ramsden JD, Cocks H, Searle PF, Watkinson JC, Ueno H, Logan A, Franklyn JA, Mautner V, Eggo MC 2003 Adenovirus-mediated 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 88:4472–4480
Onose H, Emoto N, Sugihara H, Shimizu K, Wakabayashi I 1999 Overexpression of fibroblast growth factor receptor 3 in a human thyroid carcinoma cell line results in overgrowth of the confluent cultures. Eur J Endocrinol 140:169–173
LiVolsi VA 1990 Surgical pathology of the thyroid. Philadelphia: W.B. Saunders Co.
Kleihues P, Burger PC, Scheithauer BW 1993 Histological typing of tumours of the central nervous system. World Health Organization international histological classification of tumours. 2nd ed. Berlin: Springer-Verlag
Ezzat S, Zheng L, Yu S, Asa SL 2001 A soluble dominant negative fibroblast growth factor receptor 4 isoform in human mcf-7 breast cancer cells. Biochem Biophys Res Commun 287:60–65
Celli G, LaRochelle WJ, Mackem S, Sharp R, Merlino G 1998 Soluble dominant-negative receptor uncovers essential roles for fibroblast growth factors in multi-organ induction and patterning. EMBO J 17:1642–1655
Gluzman Y, Ahrens B 1982 SV40 early mutants that are defective for viral DNA synthesis but competent for transformation of cultured rat and simian cells. Virology 123:78–92
Koziczak M, Holbro T, Hynes NE 2004 Blocking of FGFR signaling inhibits breast cancer cell proliferation through downregulation of D-type cyclins. Oncogene (Basinstoke) 23:3501–3508
Mohammadi M, Froum S, Hamby JM, Schroeder MC, Panek RL, Lu GH, Eliseenkova AV, Green D, Schlessinger J, Hubbard SR 1998 Crystal structure of an angiogenesis inhibitor bound to the FGF receptor tyrosine kinase domain. EMBO J 17:5896–5904
Hughes SE 1997 Differential expression of the fibroblast growth factor receptor (FGFR) multigene family in normal human adult tissues. J Histochem Cytochem 45:1005–1019
Naimi B, Latil A, Fournier G, Mangin P, Cussenot O, Berthon P 2002 Down-regulation of (IIIb) and (IIIc) isoforms of fibroblast growth factor receptor 2 (FGFR2) is associated with malignant progression in human prostate. Prostate 52:245–252
Jin C, McKeehan K, Guo W, Jauma S, Ittmann MM, Foster B, Greenberg NM, McKeehan WL, Wang F 2003 Cooperation between ectopic FGFR1 and depression of FGFR2 in induction of prostatic intraepithelial neoplasia in the mouse prostate. Cancer Res 63:8784–8790
Ezzat S, Huang P, Dackiw A, Asa SL, Dual inhibition of RET and FGFR4 restrains medullary thyroid cancer cell growth. Clin. Cancer Res, in press
Ezzat S, Zheng L, Zhu X-F, Wu GE, Asa SL 2002 Targeted expression of a human pituitary tumor-derived isoform of FGF receptor-4 recapitulates pituitary tumorigenesis. J Clin Invest 109:69–78
Tuominen H, Heikinheimo P, Loo BM, Kataja K, Oker-Blom C, Uutela M, Jalkanen M, Goldman A 2001 Expression and glycosylation studies of human fgf receptor 4. Protein Expr Purif 21:275–285
Grand EK, Grand FH, Chase AJ, Ross FM, Corcoran MM, Oscier DG, Cross NC 2004 Identification of a novel gene, FGFR1OP2, fused to FGFR1 in 8p11 myeloproliferative syndrome. Genes Chromosomes Cancer 40:78–83
Grand EK, Chase AJ, Heath C, Rahemtulla A, Cross NC 2004 Targeting FGFR3 in multiple myeloma: inhibition of t(4;14)-positive cells by SU5402 and PD173074. Leukemia (Baltimore) 18:962–966
Bange J, Prechtl D, Cheburkin Y, Specht K, Harbeck N, Schmitt M, Knyazeva T, Muller S, Gartner S, Sures I, Wang H, Imyanitov E, Haring HU, Knayzev P, Iacobelli S, Hofler H, Ullrich A 2002 Cancer progression and tumor cell motility are associated with the FGFR4 Arg(388) allele. Cancer Res (Baltimore) 62:840–847(Rosanne St. Bernard, Lei )