Identiﬁcation of FGFR4 as a Potential Therapeutic...

Human Cancer Biology

Identification of FGFR4 as a Potential Therapeutic Target forAdvanced-Stage, High-Grade Serous Ovarian Cancer

Tarrik M. Zaid1, Tsz-Lun Yeung1, Melissa S. Thompson1, Cecilia S. Leung1, Tom Harding5, Ngai-Na Co1,Rosie S. Schmandt1, Suet-Ying Kwan1, Cristian Rodriguez-Aguay2, Gabriel Lopez-Berestein2,3,4,Anil K. Sood1,3,4, Kwong-Kwok Wong1, Michael J. Birrer6, and Samuel C. Mok1

AbstractPurpose: To evaluate the prognostic value of fibroblast growth factor receptor 4 (FGFR4) protein

expression in patients with advanced-stage, high-grade serous ovarian cancer, delineate the functional role

of FGFR4 in ovarian cancer progression, and evaluate the feasibility of targeting FGFR4 in serous ovarian

cancer treatment.

Experimental Design: Immunolocalization of FGFR4 was conducted on 183 ovarian tumor samples.

The collected FGFR4 expression data were correlated with overall survival using Kaplan–Meier and Cox

regression analyses. The effects of FGFR4 silencing on ovarian cancer cell growth, survival, invasiveness,

apoptosis, and FGF1-mediated signaling pathway activation were evaluated by transfecting cells with

FGFR4-specific siRNAs. An orthotopic mouse model was used to evaluate the effect of injection of FGFR4-

specific siRNAs and FGFR4 trap protein encapsulated in nanoliposomes on ovarian tumor growth in vivo.

Results: Overexpression of FGFR4 protein was significantly associated with decreased overall survival

durations. FGFR4 silencing significantly decreased the proliferation, survival, and invasiveness and

increased apoptosis of ovarian cancer cells. Also, downregulation of FGFR4 significantly abrogated the

mitogen-activated protein kinase (MAPK), nuclear factor-kB (NF-kB), andWNT signaling pathways, which

are activated by FGF1. Targeting FGFR4with the FGFR4-specific siRNAs and FGFR4 trapprotein significantly

decreased ovarian tumor growth in vivo.

Conclusions: FGFR4 is a prognostic marker for advanced-stage, high-grade serous ovarian carcinoma.

Silencing FGFR4and inhibiting ligand-receptor binding significantly decrease ovarian tumor growthboth in

vitro and in vivo, suggesting that targeting ovarian cancer cells with high levels of FGFR4 protein expression is

a new therapeutic modality for this disease and will improve survival of it. Clin Cancer Res; 19(4); 809–20.

�2012 AACR.

IntroductionHigh-grade, late-stage metastatic serous carcinoma

accounts for themajority of annual epithelial ovarian cancerdeaths in theUnited States (1).More than15,000deaths peryear result from ovarian cancer, making it the most lethal

gynecologic malignancy. Researchers have made minimaladvances in extending overall survival durations over thepast 3 decades. Currently, the cornerstone of treatment ofthis cancer is surgery with the aim of reducing the tumorburden to microscopic disease (2). This is usually followedby adjuvant combination chemotherapywith platinumandtaxane, which produces initial complete responses in 80%of patients and disease-free periods ranging from 12 to 65months (3, 4). However, abdominal and pelvic recurrencerates approach 80%, and response to further chemotherapyis limited (5). Attempts at using biologic agents to improveoutcomes of this disease, including trap proteins, siRNAencapsulated in nanoparticles, and humanized antibodies,are ongoing (6–8). Thus far, although only the antivascularEGF antibody bevacizumab has shown clinical activity asreported in a phase 3 clinical trial (7). PARP inhibitors alsohave shown promising activity in pre-phase 3 trials (9).Prognostic and predictive biomarkers that can stratifypatients for treatment are still lacking.

Using a 60-mer 22K oligonucleotide-based array com-parative genome hybridization platformwithDNA isolated

Authors' Affiliations: Departments of 1Gynecologic Oncology andRepro-ductive Medicine, 2Experimental Therapeutics, and 3Cancer Biology; 4TheCenter for RNA Interference andNon-Coding RNA, TheUniversity of TexasMD Anderson Cancer Center, Houston, Texas; 5Five Prime Therapeutics,Inc, South San Francisco, California; and 6Department of Medicine, Mas-sachusetts General Hospital, Harvard Medical School, Boston,Massachusetts

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Corresponding Author: Samuel C. Mok, Department of GynecologicOncology and Reproductive Medicine, Unit 1362, The University of TexasMD Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, TX77030. Phone: 713-792-1442; Fax: 713-792-1459; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-12-2736

�2012 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org 809

Research. on September 12, 2018. © 2013 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst January 23, 2013; DOI: 10.1158/1078-0432.CCR-12-2736

http://clincancerres.aacrjournals.org/

from microdissected tumor tissue samples, Birrer and col-leagues have shown that a gain frequency of 5q31 to 35.3 inovarian cancer cells was a negative prognostic indicator foradvanced-stage, high-grade serous ovarian cancer, with anoverall prevalence of 25% (10). Further studies showed thatfibroblast growth factor 1 (FGF1) in 5q31 to 35.3 may beone of the genes that drive ovarian cancer progression (10).Fibroblast growth factor receptor 4 (FGFR4), one of the keyreceptors for FGF1, is located on the same chromosomalsegment and has exhibited preferential binding to FGF1(11, 12). An alteration in FGFR4 expression would suggestfurther activation of the FGF ligand and receptor axis, whichwould subsequently impact ovarian cancer progression. Itwould also suggest that targeting the FGF ligand and recep-tor axis can be a new regimen in ovarian cancer treatment.

In the present retrospective study, we evaluated the clin-ical significance of FGFR4 expression in patients with ovar-ian cancer and delineated the receptor’s signaling pathwaysin conferring an aggressive phenotype to ovarian cancercells. Furthermore, we evaluated the feasibility of targetingFGFR4with siRNAdelivered in nanoparticles and an FGFR4trap protein in the treatment of ovarian cancer.

Materials and MethodsPatient samples

A total of 183 paraffin-embedded, high-grade, Interna-tional Federation of Gynecology andObstetrics stage IIIB toIV (advanced-stage) serous ovarian tumor samples wereused in this study. They were obtained from the ovariancancer repository at The University of Texas MD AndersonCancer Center. The samples were collected from patientsundergoing primary cytoreductive surgery for ovarian can-cer from 1995 to 2006. After surgery, patients received

platinum-based combination chemotherapy. Optimal sur-gical cytoreduction was defined by a residual tumor nomore than 1 cm in diameter. The overall survival durationwasmeasured from thedate of diagnosis to thedate of deathor censored at the date of the last follow-up examination.Clinical data, including age, cytoreduction status (optimalvs. suboptimal), and overall survival were abstracted fromthe patients’ medical records. Ten normal ovarian and 10normal fallopian tube tissue samples were collected frompatients with benign gynecologic diseases for use as con-trols. All samples and clinical data were collected with theapproval of the MD Anderson Institutional Review Board.

Evaluation of FGFR4 overexpression byimmunohistochemistry

Immunolocalization of FGFR4 was carried out on 183serous carcinoma, 10 normal ovarian tissue, and 10 normalfallopian tube tissue samples. Slides were stained with acommercially available anti-FGFR4 antibody (1:250; sc-124, Santa Cruz Biotechnology; Supplementary TableS1). FGFR4 protein expression was visualized by 3,30-dia-minobenzidine (DAB). Normal rabbit IgG applied to high-grade serous carcinoma samples with high levels of FGFR4expression was used as a negative control. Digital photo-micrographs of representative areas of each slide were takenat �20 magnification. Quantitative FGFR4 staining inten-sity was determined, and localization measurements wereobtained using the Image-Pro Plus software (version 5.1;MediaCybernetics). Briefly, areas of interest were drawnaround tumor regions. Color segmentation was then usedto isolate DAB from background staining. A luminescencegrayscale filter was then used with each pixel carrying anintensity value from0 (black,maximumDAB saturation) to255 (white, no DAB). The FGFR4 staining intensity scorewas calculated by dividing the sumof the intensity values ina tumor area of interest by the number of pixels. This scorewas used to group the patients according to low and highFGFR4 expression using the median score as the cut offpoint. Survival analyses using both a Kaplan–Meier model-ing (with log-rank significance testing) and a Cox propor-tional hazards model were conducted to determine theeffect of FGFR4 expression levels on overall survival andrisk of death. A P-value less than 0.05 was consideredsignificant.

Effects of FGFR4 silencing on ovarian cancer cellproliferation, invasion potential, survival, andapoptosis

Endogenous FGFR4 expression was first evaluated in 6ovarian cancer cell lines (A2780, SKOV3, OVCA3, OVCA5,OVCA432, and OVCA433) and a normal ovarian surfaceepithelial (OSE) cell line using Western blot analysis.FGFR4 was knocked down in ovarian cancer cells by using2 validated, commercially available FGFR4-specific siRNAs(Hs_FGFR4_5 and Hs_FGFR4_6) and a nontarget, scram-bled siRNA sequence as a control (Qiagen Sciences). Trans-fection of ovarian cancer cells with the siRNAs was carriedout using siRNA duplexes at a final concentration of

Translational RelevanceHigh-grade serous carcinoma accounts for the major-

ity of epithelial ovarian cancers. Prognostic or predictivebiomarkers that can stratify patients for treatment arelacking. This study seeks to validate the prognostic valueof FGFR4 in high-grade serous carcinoma overexpres-sion, to delineate the functional role of FGFR4 in ovariancancer progression, and to evaluate the feasibility oftargeting FGFR4 in serous ovarian cancer treatment. Theresults of this study have shown that FGFR4 protein is aprognostic marker for advanced-stage, high grade serouscarcinoma, and targeting ovarian cancer expressinghigh levels of the protein with FGFR4-specific siRNAsand FGFR4 trap protein suppressed ovarian tumorgrowth in a xenograft mouse model. These studies pro-vide evidence that silencing FGFR4, or inhibiting ligand-receptor binding, can significantly inhibit ovarian cancerprogression, suggesting that targeting ovarian cancerexpressing high levels of the protein may be a newmodality in the treatment of the disease and improvesurvival.

Zaid et al.

Clin Cancer Res; 19(4) February 15, 2013 Clinical Cancer Research810




10 nmol/L and Lipofectamine RNAiMAX reagent (LifeTechnologies). Validation of successful FGFR4 knockdownwas carried out at the protein and mRNA levels usingWestern blot analysis and quantitative real-time PCR anal-ysis, respectively (Supplementary Fig. S1A).The effect of FGFR4 expression on ovarian cell survival

was determined by incubating high-grade serous ovariancarcinoma cell lines transfected with siRNAs in mediumcontaining 2% FBS for 3 days, after which the medium wasremoved, and water-soluble tetrazolium salt (WST-1)reagent (Roche Applied Science) diluted in serum-freemedia was added to the cells. After 2 hours of incubation,the assay was read using a FLUOstar Omega plate reader(BMG LABTECH). All data were normalized to respectivecontrols.A real-time cell proliferation assay was conducted using

the xCELLigence system (Roche Applied Science). Thissystem uses measurement of electrical impedance, createdby cells attached to the microelectrode-integrated cell cul-ture plated to measure cell proliferation in real time (13).Cells were plated at 10,000 cells per well in serum-freeOpti-MEM (Life Technologies) and allowed to attach for 12hours. A unit-less parameter termed as the cell index wasderived and used to represent the cell number based on themeasured relative change in electrical impedance thatoccurred in the presence and absence of cells in the wells(14). The cell index was normalized to the baseline readingat time point 0 following attachment and transfection. Cellproliferation as measured according to the cell index wasobserved for 3.5 days.The invasion potential of 3 high-grade serous ovarian

carcinoma cell lines (SKOV3, OVCA3, and OVCA433) wasassessed using the Oris cell invasion assay (Platypus Tech-nologies) as described by the manufacturer (15). In brief,cells were seeded at a density of 30,000 per well in serum-free Opti-MEM on a 96-well plate previously coated with acollagen I layer and with silicon stoppers covering thecentral areas of the wells to create a cell-free gap. The cellswere then transfected with Hs_FGFR4_5 and Hs_FGFR4_6and with the nontarget, scrambled siRNA sequence for 24hours, after which fresh culturemedium containing 2%FBSwas added to the cells. Twelve hours after the addition offresh medium, the stoppers were removed, and cells wereimmediately overlaid with a thick layer of the same matrix.The cells were then allowed to invade into the collagenmatrix at the central area of the wells for another 48 hours.The cells were then stained with calcein AM (Life Technol-ogies). The invasion potential of the cells was assessed bymeasuring the fluorescence of labeled cancer cells invadedinto the center of the collagen matrix with a FLUOstarOmega plate reader. Photomicrographs were taken undera fluorescent microscope for each of the wells.Apoptosis assays were conducted with 2 high-grade

serous ovarian carcinoma cell lines (SKOV3 andOVCA432)harvested after transfection with Hs_FGFR4_5 andHs_FGFR4_6 and with the nontarget, scrambled siRNAsequence for 72 hours using the Cell Death Detection cellELISA system (Roche Diagnostics), which quantified the

relative number of apoptotic cancer cells by colorimetricallymeasuring cell-free nucleosomes (16).

Effects of FGFR4 overexpression on ovarian cancer cellproliferation

To determine the effect of FGFR4 overexpression oncancer cell proliferation, FGFR4 was overexpressed in ovar-ian cancer cell lines OVCA429 andOVCA5, which have lowendogenous FGFR4 expression levels. The full-lengthFGFR4 gene was delivered into the cells via transient trans-duction using FGFR4 lentiviral particles (GeneCopoeia).Forty-eight hours after transduction, cells were harvested,and overexpression of FGFR4 in transduced cancer cells wasconfirmed. A cell proliferation assay was conducted withFGFR4-overexpressing cells or control mock-transducedcells seeded on a 96-well plate at 2,500 cells per well.Twenty-four hours after cell seeding, cell proliferation wasassayed by incubating cells with WST-1 for 1 hour, and theabsorbance at 450 nm was measured.

Effect of FGFR4 silencing on FGF1-mediated cellsignaling

OVCA432 cells, stably transduced with different report-er–response elements from the Cignal Lenti reporter system(SABiosciences), were transfected with Hs_FGFR4_5 andHs_FGFR4_6 and with the nontarget, scrambled siRNAsequence. They were then plated in 96-well plates at adensity of 30,000 cells per well and allowed to attach tothe plate for 24 hours. After cell attachment, cell culturemediumwas replacedwith freshmedium containing 10ng/mL FGF1 and incubated for 6 hours. Next, the cells werelysed with cell lysis buffer and assayed according to themanufacturer’s instructions (Promega) using a FLUOstarOmega plate reader. Data were normalized to the respectivenegative control groups.

Effect of FGFR4 silencing on transcriptome profiles inovarian cancer cells

Total RNAs were isolated from SKOV3 and OVCA432cells transiently transfected with Hs_FGFR4_5 andHs_FGFR4_6 and with the nontarget, scrambled siRNAsequence for 36 hours. The isolated RNA samples were thensubjected to whole-genome transcriptome profiling usingthe GeneChip Human Genome U133 Plus 2.0 array (Affy-metrix). Differentially expressed genes (>2-fold) commonto both FGFR4-specific siRNAs and both cell lines wereselected for further analysis using the Ingenuity PathwayAnalysis program (Ingenuity Systems). Upregulation of 2 ofthe identified genes, CXCR4 and BNIP3, was validated inOVCA432, OVCA433, and SKOV3 cells using TaqMan geneexpression assays (Life Technologies).

To identify the key canonical pathways most relevant tothe set of differentially expressed genes identified in theFGFR4-silenced ovarian cancer cells, canonical pathwayanalysis was conducted using the Ingenuity Pathway Anal-ysis program.Differentially expressed geneswith expressionlevels that differed more than 2-fold from those in bothOVCA432 and SKOV3 cells were input into the program for

A New Therapeutic Target for Advanced-Stage Serous Ovarian Carcinoma

www.aacrjournals.org Clin Cancer Res; 19(4) February 15, 2013 811




analysis. P-values less than 0.05 for key pathways wereconsidered significant, and association of a specific pathwaywith the uploaded data set was unlikely to result fromchance.

Targeting FGFR4 with FGFR4 trap proteins and siRNAin an orthotopic mouse model

The FGFR4 fusion trap protein FTP-091 (FGFR4mut:Fc:R4-Trap; FGFR4:Fc) generated by Five Prime Therapeuticswas used for targeting FGFR4 in the in vivo study. FGFR4mut:Fc is a chimeric fusion protein consisting of the 3 Ig-likeextracellular domains of FGFR4 (D1–3) with the acid boxlinker between D1 andD2 replaced with the correspondingacid box region of FGFR1. The Biacore profile for the affinityof FGFR4:Fc to different FGF ligands is shown in Supple-mentary Table S2. Intraperitoneal tumor growth in amurine orthotopic model was used as described previously(17). Briefly, a firefly luciferase-expressing serous ovariancancer cell line (OVCA432-Luc) was generated by transfect-ing the paternal cell line with commercially availablelentiviral particles (GenTarget) and selected for luciferase-expressing cells using puromycin. Mice received intraperi-toneal injections of 1 � 106 OVCA432-Luc cells and weresubsequently placed in experimental and control groups of12 mice each. The experimental group received 20 mg/kgFTP-091 twice weekly, whereas the control group receivedpooled human IgGs (Sigma-Aldrich) at the same dose andschedule. Weekly bioluminescent imaging, carried out onthe IVIS 100 imaging system (Caliper Life Sciences), wasused to follow intraperitoneal tumor nodule developmentfor 4 weeks, after which the animals were euthanized byCO2 inhalation followed by cervical dislocation. All visibletumors were resected and weighed. Ex vivo bioluminescentimaging of all of the tumors in themicewas carried out, andthe signal intensity of ex vivo tumors was quantified usingthe Living Image software (Caliper Life Sciences).

In addition, an in vivo siRNA silencing study was con-ducted. Specifically, 2� 106 OVCA432-Luc cells suspendedin PBS were intraperitoneally injected into 30 nude mice.The mice were then randomized and placed in 3 groups forinjection of dioleoyl phosphatidylcholine (DOPC) nano-liposomes containing a control siRNA (control-siRNA-DOPC), Hs_FGFR4_5 (siRNA-Hs_FGFR4_5-DOPC), orHs_FGFR4_6 (siRNA-Hs_FGFR4_6-DOPC). DOPC nano-liposomes were prepared as described previously (18).Briefly, DOPC and siRNA were mixed and lyophilized. Thelyophilized preparations were hydrated in PBS at a concen-tration of 5 mg/100 mL before administration. The DOPCnanoliposomes were determined to have a mean diameterof 30 nm. Injection of nanoliposomes began 1 week afterthe initial cancer cell injection. Each mouse was intraper-itoneally injected with 5 mg of the corresponding siRNA–DOPC reconstituted in 200mL of Caþþ- andMgþþ-free PBS.Intraperitoneal injectionswere conducted twice aweek for 7weeks, after which the mice were euthanized by CO2 inha-lation followed by cervical dislocation and submitted tonecropsy. The mice and their tumors were weighed, andresected tumors were processed for histologic evaluation.

Immunostainingof tissue sections from theorthotopicmouse model

Ki-67 staining was conducted to visualize proliferatingovarian cancer cells using a polyclonal antihuman Ki-67antibody (1:200; Life Technologies) with paraffin-embed-ded ovarian tumor sections obtained in the FGFR4 trapprotein and in vivo siRNA silencing studies. Slides wereincubated with the antibody and staining was visualizedby incubating with fast red chromogen. For each tissuesection, the Ki-67–positive cancer cells in 5 random fieldsunder a � 20 objective were counted, and the averagenumber of Ki-67–positive cells per unit tumor area wascalculated.

To validate the differential expression of the candidategenes identified from our microarray study, immunolo-calization of the 2 genes with upregulated expressionidentified in the FGFR4 knockdown cancer cells, CXCR4and BNIP3, was carried out using polyclonal antihumanCXCR4 (1:200; LifeSpan BioSciences) and polyclonalantihuman BNIP3 (1:200; Sigma-Aldrich) antibodieswith tumor tissue sections obtained in the in vivo FGFR4trap protein study. The staining intensity in each sectionin both treatment groups was quantified and presented ina box plot.

Statistical analysisThe SPSS software (version 17; IBM Corporation) was

used to conduct all statistical tests. All in vitro experimentswere repeated independently in triplicate. Data werepooled with mean and SE of the mean values calculatedfor all end points where appropriate. The 2-tailed Studentt test was used to test the significance of differences insample means for data with normally distributed means.Alternatively, the Mann–Whitney U test was used for theanalysis of nonparametric data. Kaplan–Meier survivalcurves were generated and compared using a 2-sided log-rank statistic. The Cox proportional hazards model wasused for the multivariate analysis of patient survival. ThePearson correlation coefficient was used to test linearassociations. A P value less than 0.05 was consideredsignificant.

ResultsFGFR4 overexpression is associated with overallsurvival

Previous studies showed that overexpression of FGF1,which is located on chromosome 5q31 to 35.3, wassignificantly associated with poor survival (10). BecauseFGFR4 is one of the key receptors for FGF1, we evaluatedthe association between FGFR4 expression and outcomesin patients with high-grade serous ovarian carcinoma.Staining of both normal ovarian tissue and ovarian tumorsamples has shown cell surface expression of FGFR4.Immunolocalization of FGFR4 in the 183 high-gradeserous ovarian carcinoma samples showed that FGFR4expression was significantly higher in the tumor samplesthan in the normal ovarian and fallopian tube tissuesamples, which are postulated to be the tissues of origin

Zaid et al.





for the majority of high-grade serous ovarian carcinomas(P < 0.01; Fig. 1A–C; ref. 19). Furthermore, using themedian staining intensity as a cut off, Kaplan–Meieranalysis showed a significant correlation between highFGFR4 protein expression and poor overall survival (log-rank test; P < 0.001; Fig. 1D). We further confirmed theprognostic significance of FGFR4 expression using Coxregression analysis after adjusting for age and debulkingstatus [HR, 2.1; 95% confidence interval (CI): 1.4–4.3; P <0.01]

FGFR4 silencing decreases ovarian cancer cell survival,proliferation, and invasion potential in vitroThe association between FGFR4 expression and survival

suggests that FGFR4 expression plays a key role in ovariancancer progression. To determine the effect of FGFR4expression on ovarian cancer cells, we first evaluated FGFR4overexpression usingWestern blot analysis with the ovariancancer cell lines A2780, SKOV3, OVCA3, OVCA5,OVCA432, and OVCA433 and compared it with that inHOSE cells. All 6 ovarian cancer cell lines exhibited higherFGFR4 expression (Fig. 2A). We confirmed successfulFGFR4 silencing in all of the cell lines by the FGFR4-specificsiRNAs Hs_FGFR4_5 and Hs_FGFR4_6 using quantitativePCR and Western blot analyses (Supplementary Fig. S1Band C). Cell survival analysis showed that transfection withboth FGFR4-targeting siRNAs in all 6 cell lines significantly

decreased cell survival in serum-reduced media more sothan did transfection with the nontarget scrambled siRNAsequence and untransfected control (Fig. 2B). Furthermore,cell proliferation assays showed that all 6 ovarian cancer celllines transfected with FGFR4-specific siRNAs have shownsignificant decreases in proliferation after 18 to 24 hours(Fig. 2C). Moreover, invasion assays with 3 cell lines thatexhibited invasion potential in vitro have shown that thenumber of FGFR4-specific siRNA-transfected cells invadinginto the central zone of the Oris assay was significantlylower than that of the nontarget siRNA-transfected controlcells (Fig. 3A and B). Finally, the number of SKOV3 andOVCA432 cells transfected with Hs_FGFR4_5 andHs_FGFR4_6 that were apoptotic was significantly higherthan that of the cells transfected with the nontarget scram-bled siRNA sequence (Fig. 3C).

FGFR4 overexpression promotes ovarian cancer cellproliferation

Our results showed that FGFR4 overexpression inOVCA429 and OVCA5 cells, which have low levels ofendogenous FGFR4 expression, promoted cell prolifera-tion. The rate of proliferation of both cell lines was morethan 50% higher than that of their corresponding controlcells at 24 hours after initial cell seeding (SupplementaryFig. S2A and S2B), suggesting that overexpression of FGFR4led to an aggressive phenotype.

Normal ovaryA B

D

C

HGSC (low FGFR4)

Low FGFR4 expression

Perc

ent

surv

ival

Survival (mo)

0 50 100

28 55

150 200 250

100

80

60

40

20

0

High FGFR4 expression

HGSC (high FGFR4)

Normal fallopian tube

60.0

50.0

40.0

30.0

20.0

10.0

0.0

4.00

3.00

2.00

1.00

0.00

P < 0.001N = 183

P < 0.001

P < 0.001

Normal ovaries1.00 1.50 2.00 2.50 3.00 3.50 Normal

fallopian tube

Ovarian cancer

Fre

qu

en

cy

Rela

tive F

GF

R4 s

tain

ing in

tensi

ty (

arb

itrary

unit)

Relative FGFR4 staining intensity (arbitrary unit)

Figure 1. Overexpression of FGFR4 protein is associated with overall survival. A, immunolocalization of FGFR4 showing low FGFR4 protein expression innormal ovarian and fallopian tube tissue samples and low and high FGFR4 expression in an advanced-stage, high-grade serous ovarian carcinoma sample.S, stroma; T, tumor cells; bar, 10 mm. B, histogram showing the distribution of FGFR4 signaling intensity in the 183 high-grade serous ovarian carcinomasamples used in the survival analysis. C, box plot showing significantly higher FGFR4 protein expression in high-grade serous ovarian tumor samplesthan in normal ovarian and fallopian tube tissue samples. The bottom of the box indicates the 25th percentile, the top indicates the 75th percentile, and thewhiskers represent 95% CI. D, Kaplan–Meier analysis of the 183 study patients with advanced-stage, high-grade serous ovarian carcinoma showing asignificant correlation between FGFR4 protein expression and overall survival with use of the median FGFR4 staining intensity as the cut off (log rank test;P < 0.001). Themedian survival in the high FGFR4 expression group was 24months, whereas that in the low expression group was 55months. Correlation ofFGFR4 protein expression with survival was maintained after stratification according to age and debulking status.






FGFR4 silencing abrogates the effect of FGF1 onovarian cancer cell growth and downstream signalingpathway activation

Researchers have shown that FGFR4 preferentially bindsto FGF1 (11, 12), suggesting that FGF1-activated down-stream signaling pathways are mediated by FGFR4. Toconfirm this in ovarian cancer cells, we first treatedOVCA432 cells with exogenous FGF1 (1–10 ng/mL), whichproduced a significant increase in cell proliferation rates(Supplementary Fig. S2C). The cells have also shown asignificant increase in reporter activity, which reflectedactivation of the mitogen-activated protein kinase (MAPK),NF-kB, andWNT signaling pathways (Fig. 4A). In addition,Western blot analysis has shown a significant increase inphosphorylated form of extracellular signal-regulatedkinase (ERK) 1/2 and glycogen synthase kinase 3b (GSK-3b; Fig. 4B). However, when we treated OVCA432 cellstransfected with FGFR4-specific siRNAs with FGF1 (10 ng/mL), their growth rates were significantly lower than thoseof cells transfected with the nontarget scrambled siRNA (P <0.001; Fig. 4C). Furthermore, the reporter activities of theMAPK, NF-kB, and WNT signaling pathways activated byFGF1 were abrogated (Fig. 4D).

FGFR4 silencing alters transcriptome profiles inovarian cancer cells

To further delineate the underlying molecular mecha-nism of FGFR4’s control of the malignant phenotypes of

ovarian cancer cells, we conducted transcriptome profil-ing analysis on SKOV3 and OVCA432 cells transientlytransfected with Hs_FGFR4_5 and Hs_FGFR4_6 and withthe nontarget scrambled siRNA sequences. A total of 24genes were identified to be altered more than 2-fold inexpression, which are common for both OVCA432 andSKOV3 cells transiently transfected with the siRNAs thanin those transfected with the scrambled siRNA sequence(Fig. 5A; Supplementary Table S3; Gene ExpressionOmnibus, GES34828). Network functions analysisshowed that of these 24 genes, 13 were associated withfree radical scavenging, cell death, cellular growth, andproliferation (Table 1). Two of these genes were CXCR4and BNIP3 (Fig. 5B). We further validated upregulation ofthe expression of CXCR4, which regulates cell prolifera-tion and apoptosis upon binding of its ligand CXCL12 inT cells and cancer cells (20, 21), and of BNIP3, whichinduces apoptosis (22), using quantitative PCR analysisof 3 ovarian cancer cell lines in which FGFR4 was silencedby siRNA (Supplementary Fig. S3A and B). In addition,pathway analyses showed that several canonical signalingpathways, including ERK/MAPK (P ¼ 0.03), NF-kB (P ¼0.05), and TNF receptor 1 (TNFR1; P ¼ 0.04), weresignificantly associated with the differentially expressedgene signature identified in the transcriptome profilingstudy. Identification of the MAPK and NF-kB pathways inthe array data was consistent with the results of thereporter assay.

Figure 2. FGFR4 silencing in vitro decreases ovarian cancer cell survival and proliferation. A,Western blot analysis showing higher levels of FGFR4 expressionin the ovarian cancer cell lines A2780, SKOV3, OVCA3, OVCA5, OVCA432, and OVCA433 than in the normal ovarian epithelial cell. Fifty microgramsof protein was loaded in each lane. The same blot was probed with an anti-b-actin antibody for the normalization of protein loading and transfer. B, bar chartsshowing the survival of high-grade serous ovarian cancer cell lines after transfection with Hs_FGFR4_5 and Hs_FGFR4_6 and with the nontarget, scrambledsiRNA and incubation for 72 hours in serum-reduced media. Cells transfected with the FGFR4-specific siRNAs had significantly decreased survivalthan did those transfected with the control siRNA asmeasured usingWST-1 assays (P < 0.001). C, real-time cell proliferation assay using electric impedanceas a measure of ovarian cancer cell proliferation. Electrical impedance was normalized according to the background measurement at time point 0. Resultsshowasignificantly lower rate of proliferation of high-gradeserousovarian carcinomacells transiently transfectedwithHs_FGFR4_5andHs_FGFR4_6 thanofcells transfected with the nontarget, scrambled siRNA (P < 0.001).

Zaid et al.





Targeting FGFR4 suppresses ovarian tumor growth invivoThe antitumor effect of silencing of FGFR4 in vitro sug-

gests that targeting FGFR4 is a new strategy for treatment ofovarian cancer. We therefore examined the effect of block-ing the binding of ligands to FGFR4 using the novel FGFR4fusion trap protein FTP-091 (FGFR4mut:Fc:R4-Trap;FGFR4:Fc) and downregulating FGFR4 expression by injec-tion of DOPC nanoliposomes containing FGFR4-specific

siRNAs in a murine xenograft OVCA432 ovarian tumormodel. The results showed that mice injected with 20mg/kg FTP-091 had significantly lower luciferase activityin ex vivo tumors than did those given control human IgG(Fig. 6A–C), illustrating a decrease in the number of viabletumor cells in mice given the trap protein. Furthermore,mice injected with FGFR4 HS_5 and FGFR4 HS_6 in DOPCnanoliposomes had significantly smaller tumors than didthose injected with scrambled siRNA (Fig. 6D). Histologicexamination of the resected tumor tissue collected from thexenografts showed that the tumors had histologies consis-tent with serous ovarian carcinoma. Ki-67 staining of tumorsections obtained in the FGFR4 trap protein and siRNAstudies has shown significantly fewer Ki-67–positive cancercells per unit tumor area in the FGFR4 trap protein-injectedmice and in the FGFR4 knockdown mice (via injection ofsiRNAs inDOPCnanoliposomes) than in their correspond-ing controls (Supplementary Fig. S4A and S4B), suggestingthat targeting FGFR4 in vivo can suppress ovarian tumorgrowth via inhibition of cancer cell proliferation. Further-more, immunolocalization ofCXCR4 andBNIP3protein intumor samples collected in the FGFR4 trap protein studyshowed significantly lower expression of both proteins inthe FGFR4 trap protein-treated mice than in their controls,suggesting that blockage of FGFR4 signaling with the trapfusionprotein upregulated the expressionof theCXCR4 andBNIP3 genes (Supplementary Fig. S4C and S4D),whichwasconsistent with our findings of transcriptome profiling ofFGFR4-silenced ovarian cancer cells in vitro.

DiscussionIn a previous study, we showed decreased survival in

patients with high-grade serous carcinomas harboringamplification of 5q31 to 35.3. Specifically, we observedthat overexpression of FGF1, located in this amplicon, was apoor prognostic indicator for these tumors (10). One of thekey receptors for FGF1 is FGFR4, which is located on thesame amplicon andpreferentially binds to FGF1 (11, 12). Inthe present study, we have shown that FGFR4 overexpres-sion is associated with a more aggressive high-grade serouscarcinoma phenotype in vitro and in vivo, suggesting thatFGF axis activation through overexpression of both FGFRsand FGF ligands may represent a targetable autocrine sig-naling loop associated with poor overall survival in patientswith ovarian cancer.

Birrer and colleagues observed a 25% gain frequency ofchromosome segment 5q31 to 35.3 harboring the FGFR4gene in ovarian cancer cases (10). Their data also showed asignificant correlation between FGFR4 copy number andoverall survival. However, recently reported data from TheCancer Genome Atlas (TCGA) showed a 6% amplification/overexpression rate for FGFR4 with no significant correla-tion with survival in ovarian cancer cases (23). The discrep-ancy between these 2 sets of data may result from severalfactors. First, the TCGA tumor samples were bulk tissuesamples with stromal cell contamination rates ranging from5% to 50%. Tissue samples with high levels of stromal

Figure 3. FGFR4 silencing decreases ovarian cancer cell invasionpotential in vitro. A, fluorescent micrographs showing SKOV3, OVCA3,and OVCA433 cells transiently transfected with Hs_FGFR4_5 andHs_FGFR4_6 and the nontarget, scrambled siRNA invading into thecentral zone of the Oris assay at 0 and 12 hours. The cells were thenstained with calcein AM for visualization. B, bar charts showingsignificantly lower invasion potential of SKOV3, OVCA3, and OVCA433cells after transfection with FGFR4-specific siRNAs than of control cells.�, P ¼ 0.024; ��, P ¼ 0.003; ��, P ¼ 0.001. Error bars represent SEM.C, bar chart showing significantly higher apoptosis rates in response toserumstarvation asmeasured according to increased number of cell-freenucleosomes in SKOV3 and OVCA432 cells after transfection withFGFR4-specific siRNAs than in the control cells. ��, P < 0.01;��, P < 0.001. Error bars represent SEM.






contamination will show significantly low levels of FGFR4gene amplification and/or expression, as the FGFR4 geneand mRNA copies are diluted by the stromal DNA andmRNA. This may impact survival correlation studies. Incomparison, in our analyses, we used DNA and RNAextracted from microdissected ovarian tumor cells, whichhad minimal stromal DNA and RNA contamination. Sec-ond, the TCGA samples were collected from multiple insti-tutions, whereas we collected our samples at a single insti-tution, which may have given us a more homogenouspatient population. Third, we found a significant correla-tion between FGFR4 protein expression and survival. WhileFGFR4 amplification may not correlate with survival asindicated in the TCGA dataset, FGFR4 protein expressionmay correlate with it, as gene amplification and mRNAexpressionmay not correlate with protein expression owingto mechanisms that regulate mRNA and/or protein expres-sion. However, researchers have not immunohistochemi-cally evaluated FGFR4 protein expression in the TCGAdataset.

FGFR4 differs from the other 3 members of the FGFRfamily in genomic structure, ligand binding, and signaltransduction (24). FGFR activation, either by activatingmutations (25–29) or overexpression (30–35), occurs in

multiple solid tumors. FGFR2 and FGFR3 mutations arecommon in endometrial cancer (36) and bladder cancer(31). In comparison, FGFR1 and FGFR4 mutations are notcommon in carcinoma cells; instead, overexpression ofFGFR1 and FGFR4 is more prevalent (37). To exclude thepresence of several rare activating mutations of FGFR4, wesequenced DNA isolated from microdissected high-gradeserous carcinoma samples. We did not identify any muta-tions in either the kinase or intermembrane domain exceptfor the Gly388Arg polymorphism, which has exhibited noeffect on cancer prognosis (38), in 6 of 43 (14%)high-gradeserous carcinoma samples. Our data suggest that, similar tobreast cancer (39), FGFR overexpression is the main mech-anism implicated in high-grade serous ovarian carcinoma.

Unlike other FGFR familymembers, FGFR4preferentiallybinds to acidic FGF (FGF1; refs. 11, 12). Our prior workshowed that poor survival and related phenotypic changesinduced by FGF1 in high-grade serous carcinoma parallelthose in FGFR4. Hence, overexpression of both FGF1 andFGFR4 may provide an autocrine loop that drives high-grade serous ovarian carcinoma growth. This may resultfrom activation of the MAPK/ERK signaling pathway byFGF1 as described previously (40) and confirmed by thepathway reporter and Western blot analysis data in the

Control

A

D

B C

MAPK AKT JNK NFKB

Pathway reporter

3

2

1

0

4

3

2

1

0

WNT NOTCH

No

rma

lize

d lu

min

esce

nce

Norm

aliz

ed lum

inescence

Norm

aliz

ed lum

inescence

Norm

aliz

ed lum

inescence

Norm

aliz

ed c

ell

index

2.5

2.0

1.5

1.0

0.5

0.0

2.0

1.5

1.0

0.5

0.0

10 ng/mL FGF1FGF1

FGF1

FGF1 10 ng/mL

Treatment

Time in hours

Hs_FGFR4 5

FGFR4 HS_5 siRNA FGFR4 HS_6 siRNA

Control (nontarget) siRNA

Nontarget siRNA

MAPK Pathway reporter WNT Pathway reporter NF-κB Pathway reporter

Control FGF1 10 ng/mL

Treatment

Control FGF1 10 ng/mL

Treatment

Control

Hs_FGFR4 6

0 6 12 18 14 30 36 42 48 54 62 68 74 80 86 92 98104

110

116

8

6

4

2

0

Phospho-Erk1/2

Phospho-GSK3β

Phospho-p65 NH-κB

Total-p65 NH-κB

Total-GSK3β

Total-Erk1/2

Figure 4. FGFR4 silencing abrogates the effect of FGF1 on ovarian cancer cell growth and downstream signaling pathway activation. A, bar chart showing asignificant increase in reporter activity in FGF1-treatedOVCA432 cells stably transfectedwith theCignal Lenti reporter system for theMAPK,NF-kB, andWNTsignaling pathways. �, P < 0.001. B, Western blot analysis showing significantly higher level of phosphorylated ERK1/2, GSK-3b, and NF-kB protein inOVCA432 cells treated with FGF1 at 10 ng/mL for 1 hour (þ) versus control cells (�). C, real-time cell proliferation assay results showing a significantly lowerproliferation rate in OVCA432 cells transfected with Hs_FGFR4_5 and Hs_FGFR4_6 than in those transfected with the nontarget, scrambled siRNA undertreatmentwith FGF1 at 10 ng/mL (P < 0.001). D, bar charts showing the effect of FGF1 on reporter activity, whichwas significantly lower in ovarian cancer cellstransfected with Hs_FGFR4_5 and Hs_FGFR4_6 than in those transfected with the scrambled siRNA (P < 0.001).

Zaid et al.





present study. In addition to activation of the MAPK/ERKpathway, we observed activation of the proliferative NF-kBand WNT signaling pathways in FGF1-treated cells, whichour pathway reporter and Western blot analysis confirmed.In addition, our canonical pathway analysis of transcrip-tome profiling data showed a significant correlationbetween these pathways and the differentially expressedgenes identified bymanipulating the level of FGFR4 expres-sion in ovarian cancer cells. These data suggest thatmultiplepathways canbe activatedby FGF1,most likelymediated viaFGFR4, because the effect of FGF1 on pathways activationcan be abrogated by downregulation of FGFR4 expression.Furthermore, our data suggest that the effects of FGFR4 on

ovarian cancer cell proliferation and survival result fromupregulation of expression of CXCR4, which can regulatecell proliferation and apoptosis upon binding of its ligandCXCL12 in T cells and cancer cells (20, 21), and fromupregulation of expression of BNIP3, which is known toinduce apoptosis (22). We confirmed upregulation of these2 proteins in the ovarian tumor samples obtained frommice given the FGF trap protein. However, the molecularmechanisms involved by these 2 proteins in mediating theeffect of the FGFR4 signaling pathway on ovarian cancer cellgrowth remain to be elucidated.

In the absence of activating mutations of FGFR4, down-regulation of FGFR4 expression and prevention of

OVCA 432A B

SKOV3

FGFR4 siRNA5

1852 370

24

1069

1590 563 999

FGFR4 siRNA6

FGFR4 siRNA5 FGFR4 siRNA6

Figure 5. Transcriptome analysis on FGFR4-silenced ovarian cancer cell lines. A, Venn diagram showing the number of differentially expressed genes (>2-fold)common to both OVCA432 and SKOV3 cells transiently transfectedwith Hs_FGFR4_5 andHs_FGFR4_6 for 36 hours when comparedwith those transfectedwith the nontarget, scrambled siRNA. B, the most significant network associated with FGFR4 silencing. Genes with upregulated and downregulatedexpression are shown in red and green, respectively.

Table 1. Networks involved by the differentially expressed genes in FGFR4-silenced ovarian cancer cells

Top Networks

ID Associated Network Functions Score Focus Molecules

1 Free radical scavenging. Cell death. Cellular growth and proliferation 34 132 Cell death, free radical scavenging. Lipid metabolism 13 63 Amino acid metabolism, small molecule biochemistry, cellular growth and proliferation 2 14 Gene expression, cell-to-cell signaling and interaction, molecular transport 2 1

NOTE: Function networks involved by the differentially expressed genes (>2-fold) common to OVCA432 and SKOV3 cells. Networkscores were calculated by the Ingenuity Pathway Analysis software program using the Fisher exact test according to the number ofnetwork-eligiblemolecules in a specific network,molecules analyzed in the submitteddata set, andmolecules in the IngenuityPathwayAnalysis database. The network score equals -log (Fisher exact test result): the higher the score, the lower the chance of getting thecorresponding network when randomly picking molecules in the Ingenuity Pathway Analysis database.






binding of ligands to FGFR4 may be effective in thetreatment of serous ovarian cancer. Using 2 strategies fortargeting FGFR4 with an orthotopic mouse model ofhigh-grade serous carcinoma, we showed that decreasingFGFR4 expression leads to a decrease in tumor growth.Although the use of FGFR4 siRNAs in vitro can completelysilence FGFR4 expression, leading to decreased prolifer-ation and survival of ovarian cancer cell lines, ovariantumors, although small, still developed in mice givenFGFR4-specific siRNAs delivered in DOPC nanolipo-somes. In addition, the tumors had FGFR4 expression,suggesting that the nanoliposomes may not efficientlydeliver the siRNAs to all tumor cells. In addition to theuse of siRNAs, our data showed that use of the solublefusion protein FPT-091 containing the extracellulardomain of FGFR4 and an IgG Fc fragment may havesignificant clinical applications. Replacement of the

FGFR4 acid box region with the corresponding FGFR1region improved the bioavailability and pharmacokinet-ics of the fusion protein in vivo. The chimeric FGFR4/FGFR1 extracellular domain in FGFR4mut:Fc is fused tothe Fc fragment of IgG1 and can bind to several membersof the FGF protein family, including FGF1. The ability tosequester several FGF ligands and inhibit the pathwayupstream of the FGF receptor enables the trap protein toinhibit all of the possible effects of overexpression of FGFsand FGFR4. Given the high degree of genetic heteroge-neity of high-grade serous carcinomas, activation of mul-tiple pathways and/or targets, such as Notch3 (41) andNAC1 (42), may be involved in the pathogenesis ofdifferent patients with ovarian cancer. A challengeremains in identifying patients with ovarian cancer whomay benefit from targeting a specific pathway or path-ways, possibly via individual tumor expression profiling.

Human IgG control groupA

B

C D

Wee

k 0

Wee

k 1

Wee

k 2

Wee

k 3

Wee

k 4

Wee

k 0

Wee

k 1

Wee

k 2

Wee

k 3

Wee

k 4

1000,000

800,000

600,000

400,000

200,000

0

Lu

min

esce

nce

(co

un

t)Lum

inescence (

count)

Tum

or

weig

ht

(g)

1000,000

800,000

600,000

400,000

200,000

0

Lu

min

esce

nce

(co

un

t)

Luminescence4,000

3,000

2,000

1,000

800000.00

600000.00

400000.00

200000.00

.00

Human IgG control FGFR4 trap treatment FGFR4

HS_5 siRNA

Scrambled siRNA

17

22

P = 0.015

P = 0.027

N = 10/group P = 0.012, N = 12/group

FGFR4

HS_6 siRNA

8,000

6,000

4,000

2,000

0,000

CountsColor scaleMin = 100Max = 4,000

FGFR4 trap protein treatment group

N = 12 N = 12

Luminescence4,000

3,000

2,000

1,000

CountsColor scaleMin = 100Max = 4,000

Figure 6. Targeting FGFR4decreases ovarian cancer cellgrowth in vitro. A, bioluminescencesignals in 12 mice given 20 mg/kgFGFR4:Fc and 12 mice givencontrol IgG from week 0 toweek 4. B, bioluminescent imagesof mice given FGFR4:Fc and thosegiven human IgG at week 4 andtheir corresponding ex vivotumor samples. C, box plotshowing significantly lowerbioluminescence signals in the 12mice given FGFR4:Fc than in thosegiven control IgG in week 4(P ¼ 0.012). The bottom of the boxindicates the 25th percentile, thetop indicates the 75th percentile,and whiskers indicate 95% CI.D, box plot showing significantlylower tumor weights in OVCA432tumor-bearing mice given siRNA-Hs_FGFR4_5-DOPC and siRNA-Hs_FGFR4_6-DOPC than in thosegiven control-siRNA-DOPC (P ¼0.015 andP¼ 0.027, respectively).Values greater than one and a halfbox lengths from either end of thebox are denoted by circles andidentified as outliers. All values,including the outliers in allexperimental groups, wereincluded in statistical analyses.

Zaid et al.





In conclusion, the present study has shown that theoverexpression of FGFR4 is an indicator of poor prognosisfor high-grade serous ovarian carcinoma. It also identifiedmechanisms by which activation of FGFR4 by ligands suchas FGF1 may lead to an aggressive cancer phenotype. Suc-cessful targeting of FGF axis activation in our orthotopicmouse model suggests that this approach is feasible inclinical trials.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: T. Zaid, T. Yeung, C. Leung, G. Lopez-Berestein,K. Wong, M.J. Birrer, S.C. MokDevelopment of methodology: T. Zaid, M.S. Thompson, C. Leung,T. Harding, C. Rodriguez-Aguayo, G. Lopez-Berestein, A.K. Sood, M.J. Birrer,S.C. MokAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): T. Zaid, T. Yeung, M.S. Thompson, C. Leung,T. Harding, S. Kwan, A.K. Sood, M.J. Birrer, S.C. Mok

Analysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): T. Zaid, T. Yeung, C. Leung, N. Co,G. Lopez-Berestein, A.K. Sood, K. Wong, S.C. MokWriting, review, and/or revision of the manuscript: T. Zaid, T. Yeung,M.S. Thompson, C. Leung, R.E. Schmandt, G. Lopez-Berestein, A.K. Sood,K. Wong, M.J. Birrer, S.C. MokAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): T. Zaid, C. Leung, N. Co, G. Lopez-Berestein, M.J. Birrer, S.C. MokStudy supervision: G. Lopez-Berestein, K. Wong, S.C. Mok

Grant SupportThis study was supported in part by grants R01 CA133057 (S.C. Mok),

RC4 CA156551 (M.J. Birrer), and U54 CA151668 (A.K. Sood); the GilderFoundation (A.K. Sood); and MD Anderson Ovarian Cancer SPORE grantP50 CA083639 (S.C. Mok) from the National Institutes of Health. Thisresearch is supported in part by the MD Anderson Cancer Center SupportGrant CA016672.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received August 21, 2012; revisedNovember 9, 2012; acceptedNovember16, 2012; published OnlineFirst January 23, 2013.

References1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer

J Clin 2012;62:10–29.2. NIH consensus conference. Ovarian cancer. Screening, treatment,

and follow-up.NIHConsensusDevelopmentPanel onOvarianCancer.JAMA 1995;273:491–7.

3. ArmstrongDK, BundyB,Wenzel L, HuangHQ,BaergenR, Lele S, et al.Intraperitoneal cisplatin and paclitaxel in ovarian cancer. N Engl J Med2006;354:34–43.

4. McGuire WP, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, LookKY, et al. Cyclophosphamide and cisplatin compared with paclitaxeland cisplatin in patients with stage III and stage IV ovarian cancer.N Engl J Med 1996;334:1–6.

5. Markman M, Rothman R, Hakes T, Reichman B, Hoskins W, Rubin S,et al. Second-line platinum therapy in patients with ovarian cancerpreviously treated with cisplatin. J Clin Oncol 1991;9:389–93.

6. Yap TA, Carden CP, Kaye SB. Beyond chemotherapy: targeted ther-apies in ovarian cancer. Nat Rev Cancer 2009;9:167–81.

7. Burger RA. Overview of anti-angiogenic agents in development forovarian cancer. Gynecol Oncol 2011;121:230–8.

8. Landen CN Jr, Chavez-Reyes A, Bucana C, Schmandt R, Deavers MT,Lopez-Berestein G, et al. Therapeutic EphA2 gene targeting in vivousing neutral liposomal small interfering RNA delivery. Cancer Res2005;65:6910–8.

9. Fong PC, Yap TA, Boss DS, Carden CP, Mergui-Roelvink M, GourleyC, et al. Poly(ADP)-ribose polymerase inhibition: frequent durableresponses in BRCA carrier ovarian cancer correlating with platinum-free interval. J Clin Oncol 28:2512–9.

10. Birrer MJ, JohnsonME, Hao K,Wong KK, Park DC, Bell A, et al. Wholegenome oligonucleotide-based array comparative genomic hybridiza-tion analysis identified fibroblast growth factor 1 as a prognosticmarker for advanced-stage serous ovarian adenocarcinomas. J ClinOncol 2007;25:2281–7.

11. KanM,Wu X,Wang F, McKeehanWL. Specificity for fibroblast growthfactors determined by heparan sulfate in a binary complex with thereceptor kinase. J Biol Chem 1999;274:15947–52.

12. Partanen J, Makela TP, Eerola E, Korhonen J, Hirvonen H, Claesson-Welsh L, et al. FGFR-4, a novel acidic fibroblast growth factor receptorwith a distinct expression pattern. EMBO J 1991;10:1347–54.

13. Irelan JT, Wu MJ, Morgan J, Ke N, Xi B, Wang X, et al. Rapid andquantitative assessment of cell quality, identity, and functionality forcell-based assays using real-time cellular analysis. J Biomol Screen2011;16:313–22.

14. Xing D, Orsulic S. A genetically defined mouse ovarian carcinomamodel for the molecular characterization of pathway-targeted therapyand tumor resistance. Proc Natl Acad Sci U S A 2005;102:6936–41.

15. Carragher NO, Frame MC. Modelling distinct modes of tumour inva-sion and metastasis. Drug Discov Today: Disease Models 2011;8:103–12.

16. Holdenrieder S, Stieber P, Chan LY, Geiger S, Kremer A, Nagel D, et al.Cell-free DNA in serum and plasma: comparison of ELISA and quan-titative PCR. Clin Chem 2005;51:1544–6.

17. Liao S, Liu J, Lin P, Shi T, Jain RK, Xu L. TGF-beta blockade controlsascites by preventing abnormalization of lymphatic vessels in ortho-topic human ovarian carcinoma models. Clin Cancer Res 2011;17:1415–24.

18. Gray MJ, Van Buren G, Dallas NA, Xia L, Wang X, Yang AD, et al.Therapeutic targeting of neuropilin-2 on colorectal carcinoma cellsimplanted in the murine liver. J Natl Cancer Inst 2008;100:109–20.

19. Kuhn E, Kurman RJ, Shih IM. Ovarian cancer is an imported disease:fact or fiction? Curr Obstet Gynecol Rep 2012;1:1–9.

20. Colamussi ML, Secchiero P, Gonelli A, Marchisio M, Zauli G, CapitaniS. Stromal derived factor-1 alpha (SDF-1 alpha) induces CD4þ T cellapoptosis via the functional up-regulation of the Fas (CD95)/Fas ligand(CD95L) pathway. J Leukoc Biol 2001;69:263–70.

21. Drury LJ, Wendt MK, Dwinell MB. CXCL12 chemokine expression andsecretion regulates colorectal carcinoma cell anoikis through Bim-mediated intrinsic apoptosis. PLoS ONE 2010;5:e12895.

22. Ray R, Chen G, Vande Velde C, Cizeau J, Park JH, Reed JC, et al.BNIP3 heterodimerizes with Bcl-2/Bcl-X(L) and induces cell deathindependent of a Bcl-2 homology 3 (BH3) domain at both mitochon-drial and nonmitochondrial sites. J Biol Chem 2000;275:1439–48.

23. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609–15.

24. Vainikka S, Partanen J, Bellosta P, Coulier F, Birnbaum D, Basilico C,et al. Fibroblast growth factor receptor-4 shows novel features ingenomic structure, ligand binding and signal transduction. EMBOJ 1992;11:4273–80.

25. Chin K, DeVries S, Fridlyand J, Spellman PT, Roydasgupta R, KuoWL,et al. Genomic and transcriptional aberrations linked to breast cancerpathophysiologies. Cancer Cell 2006;10:529–41.

26. Behrens C, Lin HY, Lee JJ, Raso MG, Hong WK, Wistuba II, et al.Immunohistochemical expressionof basic fibroblast growth factor andfibroblast growth factor receptors 1 and 2 in the pathogenesis of lungcancer. Clin Cancer Res 2008;14:6014–22.






27. Allerstorfer S, Sonvilla G, Fischer H, Spiegl-Kreinecker S, GauglhoferC, SetinekU, et al. FGF5 as anoncogenic factor in humanglioblastomamultiforme: autocrine and paracrine activities. Oncogene 2008;27:4180–90.

28. Baird K, Davis S, Antonescu CR, Harper UL, Walker RL, Chen Y, et al.Gene expression profiling of human sarcomas: insights into sarcomabiology. Cancer Res 2005;65:9226–35.

29. Toyokawa T, Yashiro M, Hirakawa K. Co-expression of keratinocytegrowth factor andK-sam is an independent prognostic factor in gastriccarcinoma. Oncol Rep 2009;21:875–80.

30. Tomlinson DC, Hurst CD, Knowles MA. Knockdown by shRNA iden-tifies S249Cmutant FGFR3 as a potential therapeutic target in bladdercancer. Oncogene 2007;26:5889–99.

31. van Rhijn BW, van Tilborg AA, Lurkin I, Bonaventure J, de Vries A,Thiery JP, et al. Novel fibroblast growth factor receptor 3 (FGFR3)mutations in bladder cancer previously identified in non-lethal skeletaldisorders. Eur J Hum Genet 2002;10:819–24.

32. vanRhijnBW,Montironi R, Zwarthoff EC, Jobsis AC, vanderKwast TH.Frequent FGFR3 mutations in urothelial papilloma. J Pathol 2002;198:245–51.

33. Stephens P, Edkins S, DaviesH, GreenmanC,CoxC,Hunter C, et al. Ascreen of the complete protein kinase gene family identifies diversepatterns of somatic mutations in human breast cancer. Nat Genet2005;37:590–2.

34. Jang JH, Shin KH, Park JG. Mutations in fibroblast growth factorreceptor 2 and fibroblast growth factor receptor 3 genes associated

with human gastric and colorectal cancers. Cancer Res 2001;61:3541–3.

35. Dutt A, SalvesenHB, Chen TH, Ramos AH, Onofrio RC, Hatton C, et al.Drug-sensitive FGFR2 mutations in endometrial carcinoma. Proc NatlAcad Sci U S A 2008;105:8713–7.

36. PollockPM,GartsideMG,Dejeza LC,PowellMA,MallonMA,DaviesH,et al. Frequent activating FGFR2mutations in endometrial carcinomasparallel germline mutations associated with craniosynostosis andskeletal dysplasia syndromes. Oncogene 2007;26:7158–62.

37. Haugsten EM, Wiedlocha A, Olsnes S, Wesche J. Roles of fibroblastgrowth factor receptors in carcinogenesis. Mol Cancer Res 2010;8:1439–52.

38. Spinola M, Leoni VP, Tanuma J, Pettinicchio A, Frattini M, Signoroni S,et al. FGFR4 Gly388Arg polymorphism and prognosis of breast andcolorectal cancer. Oncol Rep 2005;14:415–9.

39. Gelsi-Boyer V, Orsetti B, Cervera N, Finetti P, Sircoulomb F, Rouge C,et al. Comprehensive profiling of 8p11–12 amplification in breastcancer. Mol Cancer Res 2005;3:655–67.

40. Luo Y, Ye S, Kan M, McKeehan WL. Control of fibroblast growth factor(FGF) 7- and FGF1-induced mitogenesis and downstream signaling bydistinctheparinoctasaccharidemotifs.JBiolChem2006;281:21052–61.

41. Shih IeM,Wang TL. Notch signaling, gamma-secretase inhibitors, andcancer therapy. Cancer Res 2007;67:1879–82.

42. Yap KL, Fraley SI, Thiaville MM, Jinawath N, Nakayama K, Wang J,et al. NAC1 is an actin-binding protein that is essential for effectivecytokinesis in cancer cells. Cancer Res 2012;72:4085–96.

Zaid et al.





2013;19:809-820. Published OnlineFirst January 23, 2013.Clin Cancer Res Tarrik M. Zaid, Tsz-Lun Yeung, Melissa S. Thompson, et al. Advanced-Stage, High-Grade Serous Ovarian CancerIdentification of FGFR4 as a Potential Therapeutic Target for

Updated version

10.1158/1078-0432.CCR-12-2736doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2012/12/31/1078-0432.CCR-12-2736.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/19/4/809.full#ref-list-1

This article cites 41 articles, 17 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/19/4/809.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/19/4/809To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-12-2736

http://clincancerres.aacrjournals.org/content/suppl/2012/12/31/1078-0432.CCR-12-2736.DC1

http://clincancerres.aacrjournals.org/content/19/4/809.full#ref-list-1

http://clincancerres.aacrjournals.org/content/19/4/809.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/19/4/809


Identiﬁcation of FGFR4 as a Potential Therapeutic...

Documents

Transcript of Identiﬁcation of FGFR4 as a Potential Therapeutic...