identiﬁcation of oncogenic and drug-sensitizing mutations in the … · therapeutics, targets,...

Therapeutics, Targets, and Chemical Biology

Identification of Oncogenic and Drug-SensitizingMutations in the Extracellular Domain of FGFR2Junko Tanizaki1,2, Dalia Ercan1,2, Marzia Capelletti1,2, Michael Dodge1,2, Chunxiao Xu1,2,Magda Bahcall1,2, Erin M. Tricker1,2, Mohit Butaney1,2, Antonio Calles1,2, Lynette M. Sholl3,Peter S. Hammerman2,4, Geoffrey R. Oxnard1,2,5, Kwok-Kin Wong1,2,5,6, andPasi A. J€anne1,2,5,6

Abstract

The discovery of oncogenic driver mutations and the subse-quent developments in targeted therapies have led to improvedoutcomes for subsets of lung cancer patients. The identification ofadditional oncogenic and drug-sensitive alterationsmay similarlylead to new therapeutic approaches for lung cancer. We identifyand characterize novel FGFR2 extracellular domain insertionmutations and demonstrate that they are both oncogenic andsensitive to inhibition by FGFR kinase inhibitors.Wedemonstratethat the mechanism of FGFR2 activation and subsequent

transformation is mediated by ligand-independent dimerizationand activation of FGFR2 kinase activity. Both FGFR2-mutantforms are predominantly located in the endoplasmic reticulumand Golgi but nevertheless can activate downstream signalingpathways through their interactions with fibroblast growth factorreceptor substrate 2 (FRS2). Our findings provide a rationale fortherapeutically targeting this unique subset of FGFR2-mutantcancers as well as insight into their oncogenic mechanisms. CancerRes; 75(15); 3139–46. �2015 AACR.

IntroductionThe discovery of oncogenic driver genes has helped identify

druggable targets and the subsequent development of new ther-apies using small molecule tyrosine kinase inhibitors (TKI) for abroad range of cancers. Detection of epidermal growth factorreceptor (EGFR)mutations or anaplastic lymphoma kinase (ALK)rearrangements in non–small cell lung cancer (NSCLC) is cur-rently standard of care, leading to dramatically improved out-comes in patients with these genetic alterations when treated withan EGFR or ALK TKIs, respectively, as compared with systemicchemotherapy (1, 2). A number of new potentially targetableoncogenic gene alterations have been further characterized inrecent years (3). However, even with recent intense researchefforts, there still remain a significant proportion of cancerpatients whose cancers do not harbor any known oncogenicalterations. Given the fact that cancer is the leading cause of deathworldwide and accounts for 8.2 million deaths annually (4),identifying even a small subpopulation of patients who might

benefit from targeted therapy could have significant clinicalrelevance.

Fibroblast growth factor receptor 2 (FGFR2) is a member ofthe FGFR family. FGFRs have a core structure containing anextracellular domain (ECD), a hydrophobic transmembranedomain, and an intracellular kinase domain. FGF ligand bindingresults in FGFR dimerization, followed by receptor autopho-sphorylation and activation of downstream signaling, playingkey roles in multiple biologic processes such as tissue repair,angiogenesis, and embryonic development (5). Aside from theirnormal physiological roles, aberrant FGFR signaling has beenimplicated in a variety of diseases, including cancers (6). TheFGFR signaling pathway can be activated in various ways.Amplification of FGFR1 and mutations in FGFR2 and FGFR3have been identified in squamous cell lung carcinomas (7, 8).Recently identified FGFR gene rearrangements, involving FGFR1,FGFR2, and FGFR3, occur in diverse cancers, including in glio-blastomas, cholangiocarcinomas, and cancers of the bladder,thyroid, mouth, breast, head and neck, and lung (9–11).These FGFR genetic alterations are oncogenic and sensitive toFGFR-TKIs, suggesting that a new subset of cancers might betreatable with FGFR-targeted therapies.

In this study, we identify and study novel FGFR2 extracellulardomain insertion mutations. We characterize both the oncoge-nicity and FGFR inhibitor sensitivity of cells harboring thesemutations. Finally, we evaluate the underlying mechanism sup-porting oncogenic signaling.

Materials and MethodsSamples and FGFR2 exon 7 mutation detection

Targeted next-generation sequencing (NGS) of the indexpatient's tumor was performed in a commercial laboratory(12). Additional tumors from never smokers with lung cancer(n¼ 96) have been previously described (13, 14). Cancer patients

1LoweCenter for ThoracicOncology, Boston,Massachusetts. 2Depart-ment of Medical Oncology, Boston, Massachusetts. 3Department ofPathology, Brigham and Women's Hospital, Boston, Massachusetts.4The Broad Institute of Harvard and MIT, Cambridge, Massachusetts.5Department of Medicine, Boston, Massachusetts. 6Belfer Institute forApplied Cancer Science, Dana Farber Cancer Institute, Boston,Massachusetts.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

CorrespondingAuthor:Pasi A. J€anne, Dana-Farber Cancer Institute, 450Brook-line Avenue, HIM223, Boston, MA 02215. Phone: 617-632-6076; Fax: 617-582-7683; E-mail: [email protected]

doi: 10.1158/0008-5472.CAN-14-3771

�2015 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 3139

on April 18, 2021. © 2015 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst June 5, 2015; DOI: 10.1158/0008-5472.CAN-14-3771

http://cancerres.aacrjournals.org/

at Dana-Farber Cancer Institute (DFCI)/Brigham and Women'sHospital undergoing routine NGS (15) provided informed con-sent as part of an Institutional Review Board approved institu-tional wide sequencing effort.

For detection of FGFR2mutations in the 96 tumors from neversmokers, DNA was extracted from paraffin-embedded or frozentissues using the QIAamp DNA FFPE Tissue Kit (Qiagen) or theDNeasy Blood and Tissue Kit (Qiagen), respectively, and wasamplified via PCR. Primers used for the amplification are: forwardprimer 50-ACTGACAGCCCTCTGGACAACAC-30, reverse primer50-ACCCCAGTTGTGGGTACCTTTAGA-30.

Cell culture and reagentsNIH-3T3 cells were purchased from the ATCC in 2010. Authen-

tication of the cell line was performed using karyotyping at theBrigham and Women's CytoGenomics Core Laboratory in May2015. The cells with or without each gene alteration were main-tained in DMEM medium with 10% fetal calf serum, 100 units/mL penicillin, and 100 mg/mL streptomycin. BGJ398 wasobtained fromChemieTek andponatinib fromSelleckChemicals.FGF-1 ligand and EndoHfwere purchased fromR&DSystems andNew England Biolabs, respectively.

Expression constructs and lentiviral infectionFull-length FGFR2 cDNA from DF/HCC DNA Resource Core

was cloned into pDNR-Dual (BD Biosciences) as described pre-viously (16). To generate FGFR2 mutants, A266_S267ins,A266_S267ins K517R, 290_291WI>C, T730S, or V755I wereintroduced using site-directed mutagenesis (Agilent) withmutant-specific primers according to the manufacturer's instruc-tions. FGFR2–TACC3 fusions were created via a two-step PCRreaction (Phusion Kit, NEB). TACC3 and FGFR2 sequence frag-ments were amplified from existing A549 (ATCC: CCL-185)cDNA and a pDNR-dual plasmid stock respectively under themanufacture's recommended conditions. Anticipated ampliconfragments for each gene were gel purified, combined, and ream-plified to create FGFR2–TACC3. The PCR primers are availableupon request. All constructs were confirmed by DNA sequencing.Retroviral infection and culture of NIH-3T3 cells were done usingpreviously described methods (17). Polyclonal cell lines wereestablished by puromycin selection.

Soft-agar colony formation assayTwo milliliters of 0.5% noble agar (Sigma Aldrich Co.) and

media were plated in each well of a six-well plate and allowed tosolidify. A total of 1 � 104 cells were suspended in a finalconcentration of 0.35% agar with or without drug and platedonto the solidified bottom layer in triplicate. Plates were incu-bated for 21 days, then the colonies were stained with 0.005%crystal violet, and the numbers of colonies were counted.

AntibodiesAnti-phospho-FGFR (Tyr653/654), anti-phospho-FRS2

(Tyr436), anti-phospho Akt (Ser473), anti-total Akt, andanti-a-tubulin were obtained from Cell Signaling Technology.Total ERK1/2 and phopho-ERK1/2 (pT185/pY187), Alexa Fluor-488, and Alexa Fluor-549 antibodies were from Invitrogen. Anti-bodies against total FRS2, FGFR2 (C-17 and H-80), calnexin,GM130, and HSP90 were purchased from Santa Cruz Biotech-nology, Inc. Anti-Flag was obtained from Sigma Aldrich Co.

Western blottingCell lysis and western blotting were done as previously

described (17). Visualization of unreduced dimers was doneusing previously described methods (8). For ligand stimula-tion condition, cells expressing FGFR2 wild-type were culturedfor 8 hours in the presence of FGF-1 (40 ng/mL) and heparin (100ng/mL). For EndoHf digestions, 20 mg of lysate were subjected todigestion according to the manufacture's recommendations.

Xenograft experimentAll care of experimental animals were in accordance with

Harvard Medical School/DFCI institutional animal care andfollowed committee guidelines. Nu/Numice were purchase fromCharles River Laboratories International Inc. and housed in aDFCI animal facility. The expression level of FGFR2 in each NIH-3T3 cell line was confirmed by Western blot. Mice were injectedsubcutaneously with 1 � 106 cells per shot, two to three spots/mouse. The experimental mice were divided into two groupsaccording to the injected cell lines: (i) FGFR2 A266_S267ins,Parental 3T3, FGFR2 wild-type and (ii) FGFR2 A266_S267ins,FGFR2A266_S267insK517R,fivemiceper group. After 2weeks ofimplantation, tumors were monitored twice weekly and volumeswere calculated with the formula: (mm3) ¼ length � width �width� 0.5. Standard deviations were calculated usingGraphpadPrism (GraphPad Software, Inc.).

Cell surface protein isolation and immunoprecipitationCell surface protein isolation was done using a biotinylation-

based assay using the Pierce Cell Surface Protein Isolation Kit(Thermo Scientific). Flow-through lysates were examined byimmunoprecipitation as previously described (18) using ananti-Flag-tag antibody.

ImmunofluorescenceCells were fixedwith 4%paraformaldehyde and permeabilized

in 0.2% Triton X-100. The cells were then incubated in theindicated primary antibodies for 1 hour, followed by incubationwith fluorescent conjugated secondary antibodies for 1 hour.Nuclei were counterstained with 40,6-diamidino-2-phenylindole.Cells were visualized and photographed on a Nikon ECLIPSE 80iusing NIS-Elements AR software (Nikon Instruments Inc.).

ResultsNovel FGFR2 gene alterations detected in solid tumors

We first studied the tumor of a 37-year-old male never smokerwith advancedNSCLC. The patient received two cycles of cisplatinand etoposide as first-line systemic therapy but developed pro-gressive disease. While sequencing was being performed, he wastreatedwith second-line docetaxel chemotherapy. Thepatient hada dramatic response to docetaxel lasting over 8 months. A biopsyof a bone metastasis revealed NSCLC with neuroendocrine fea-tures. The tumor contained prominent glandular differentiationwith mucin production and positive staining by IHC for CD56and TTF-1 andwas negative for synaptophysin and chromogranin(Supplementary Fig. S1 and data not shown). Further analysis bytargeted NGS revealed the presence of a novel insertion in FGFR2.

This insertion mutation, A266_S267insSTVVGGD, is a dupli-cationof 21bp in the ECDof FGFR2 (Fig. 1A andB) and is somaticas determined by analyzing saliva extracted DNA (data notshown). No other genomic alteration was detected in this

Tanizaki et al.

Cancer Res; 75(15) August 1, 2015 Cancer Research3140




patient's sample using targeted NGS. We next analyzed 96 addi-tionalNSCLCs (18 tumor samples thatwere either carcinoids (n¼13) or NSCLCs with neuroendrocrine features (n ¼ 5) and 78adenocarcinomas from never smokers with no known oncogenicalterations (13, 14) for mutations in FGFR2 ECD but did notidentify any additional insertion mutations. We further queried aDFCI/Brigham and Women's Hospital database of all cancerpatients who had undergone targeted NGS for presence of FGFR2ECD mutations. Among 2,500 patients sequenced, we identified21 with FGFR2mutations (Supplementary Table S1). One of thepatients, with cholangiocarcinoma, had an FGFR2 ECDmutation(290_291WI>C) and was chosen for further study. In addition,twopatientswithNSCLChadT730S andV755Imutations locatedwithin the kinase domain of FGFR2 (Fig. 1A).We used an FGFR2–TACC3 fusion gene, recently discovered in cholangiocarcinoma(19), with most of tyrosine kinase domain of FGFR2, fused to thecoiled-coil domain of TACC3 (Fig. 1A), as a control in our studies.

FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 are oncogenic and sensitive to FGFR-TKIs

Todeterminewhether these FGFR2mutations are oncogenic,weexpressed each gene alteration in NIH-3T3 cells and evaluatedresultant growth in soft agar. FGFR2 A266_S267ins, FGFR2290_291WI>C, and FGFR2–TACC3 significantly increased colonyformation inanchorage-independent conditionsas comparedwithwild-type FGFR2 (Fig. 2A), whereas this effect was not observed incells expressing FGFR2 T730S or those with V755I (Fig. 2A).BGJ398, an FGFR tyrosine kinase inhibitor, blunted colony for-mation of FGFR2 A266_S267ins, FGFR2 290_291WI>C, andFGFR2–TACC3 cells in a dose-dependent manner (Fig. 2B). Wefurther evaluated the effect of BGJ398 on signal transduction in

these cells. Immunoblot analysis revealed greater phosphorylationof the FGFR2 substrate FRS2 in cells expressing FGFR2A266_S267ins, FGFR2 290_291WI>C, or FGFR2–TACC3. BGJ398inhibited phospho-FRS2, accompanied by a concomitant inhibi-tion of AKT and ERK1/2 phosphorylation in these cells. On thecontrary, FRS2 activation was hardly observed in cells expressingwild-type FGFR2, T730S, or V755I (Fig. 2C). A pan-FGFR inhibitorAP24534 (ponatinib) also had significant inhibitory effects onphospho-ERK1/2 and phospho-AKT as well as phospho-FRS2 inFGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3cells, but not in FGFR2 wild-type cells, T730S or V755I cells(Supplementary Fig. S2). These data suggest that cells expressingFGFR2 A266_S267ins, FGFR2 290_291WI>C, and those withFGFR2–TACC3 are dependent on FGFR signaling. Consistently,cells expressing the kinase dead version of FGFR2 A266_S267ins,FGFR2A266_S267ins K517R, formed fewer colonies (Fig. 2D).Wefurther confirmed that FGFR2A266_S267ins is sufficient to inducetumor formation in vivo using xenografts, whereas cells expressingFGFR2 A266_S267ins K517R had significantly smaller tumors(Fig. 2E), additionally confirming FGFR kinase-dependentoncogenicity.

Given that both FGFR2 A266_S267ins and FGFR2290_291WI>C alterations reside in the Ig3 ligand-bindingdomainof ECD,we further evaluated the effect of FGF stimulationin FGFR2A266_S267ins comparedwith cells expressingwild-typeFGFR2. In FGFR2 A266_S267ins cells, elevated basal FRS2 phos-phorylation levels precluded further increases following FGFstimulation as compared with cells expressing wild-type FGFR2(Supplementary Fig S3A). We also observed greater colony for-mation with FGF stimulation in FGFR2 wild-type cells, whereasFGFhad little additional effect in cells with FGFR2A266_S267ins,

Figure 1.FGFR2 genetic alterations detected inNSCLC and cholangioacrinoma. A,schematic representations of FGFR2gene alterations found in this study. Ig,immunoglobulin-like domain; TM,transmembrane domain; PTKC, proteinkinase. B, sequence tracing of FGFR2exon 7 from a 37-year-old neversmoker patient sample. There is aduplication of 21 bp at codon 797 of theFGFR2 protein, resulting in FGFR2A266_S267insSTVVGGD mutation.

Targetable FGFR2 ECD Gene Alterations in Solid Cancer

www.aacrjournals.org Cancer Res; 75(15) August 1, 2015 3141




consistent with the FRS2 phosphorylation results. Taken together,these data suggest that FGFR2 A266_S267ins leads to FGF-inde-pendent oncogenic transformation with little, if any, additionaleffect from exogenous FGF.

FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 form dimers

To identify the mechanistic basis for the oncogenicity of thesemutant FGFR2 proteins, we examined whether the receptorsdimerize in a ligand-independent fashion. Immunoblot analysisin nonreducing condition revealed that FGFR2 A266_S267insand FGFR2290_291WI>C formdimers as detectedusing either ananti-FLAG antibody or an anti-FGFR2 antibody (Fig. 3). FGFR2wild-type proteins only form dimers following FGF ligand expo-

sure (Fig. 3). In cells expressing FGFR2–TACC3, dimer bandswereobserved only with the anti-FLAG antibody, but not with the anti-FGFR2 antibody as this fusion protein lacks the C-terminalportion of FGFR2 that is recognized by this FGFR2 antibody.These data suggest that FGFR2 A266_S267ins, FGFR2290_291WI>C, and FGFR2–TACC3 are oncogenic throughligand-independent dimerization.

FGFR2 A266_S267ins, FGFR2 290_291WI>C primarily exist aspartially glycosylated forms

FGFR2 can be detected on a SDS-PAGE gel in two distinctforms: a fully glycosylated, 120-kDa form of the receptor(upper band), and a partially glycosylated, 110-kDa formof the receptor (lower band; ref. 20). Cells harboring FGFR2

Figure 2.FGFR2A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 are FGFR-TKI–sensitive driver gene alterations. A, soft-agar colony formation assayofNIH-3T3 cellsstably expressing the indicated gene alterations after 21 days. Data are means of triplicates from representative experiments and expressed as fold over thenumbers of colonies observed in cells expressing FGFR2 wild-type. B, soft-agar assay with increasing concentrations of BGJ398. After 21 days, the numbersof colonies were counted. Data are expressed relative to the corresponding value for nontreated cells and are means � SE from three independent experiments.C, cells with the indicated gene alterations were serum starved overnight and then treated with BGJ398 at the indicated concentrations for 6 hours. Cellextracts were immunoblotted to detect the indicated proteins. D, results of colony formation assay after 21 dayswith cells expressing FGFR2A266_S267ins, a kinasedead version (A266_S267ins K517R) or FGFR2 wild-type. Data are means of triplicates from representative experiments. E, tumor volume measurementsfrom mouse subcutaneous xenograft models. Each curve represents means from five to ten mice per group.

Tanizaki et al.





A266_S267ins and FGFR2 290_291WI>C were found to showhigher expression levels of the lower band (open arrow)detected by either an anti-FGFR2 antibody or an anti-FLAGantibody, compared with FGFR2 wild-type and the other muta-tions or translocations evaluated in this study (Fig. 4A). Toprove that the higher expressed lower band in cells with ECDmutations was due to a partially processed receptor, cell lysateswere digested with endoglycosidase (Endo Hf), which is spe-cific for the high mannose N-linked oligosaccharides charac-teristic of proteins that have not completed Golgi-mediatedmaturation of glycosylation. Endo Hf digestion results in a shiftin molecular mass from a primarily 110-kDa to a primarily 85-kDa band (open arrow) in both FGFR2 A266_S267ins cells andFGFR2 290_291WI>C cells (Fig. 4B), suggesting that FGFR2receptors with those ECD mutations mainly exist as partiallyglycosylated forms.

FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 express in cytoplasm and colocalize with Golgi and ERmarkers

To determine whether the FGFR2 mutations lead to alteredcellular localization of FGFR2 proteins, we evaluated the cellularlocalization using an immunofluorescence assay. In FGFR2wild-type cells, immunofluorescence signals detected by bothan anti-FGFR2 and anti-Flag antibodies were detected primarilyin the plasma membrane and also perinuclearly to a lesser

degree (Fig. 5A). However, in cells with FGFR2 A266_S267ins,FGFR2 290_291WI>C, or FGFR2–TACC3, the staining waslocalized primarily in a perinuclear pattern (Fig. 5A). Giventhis cytoplasmic distribution pattern, we next performed a co-immunofluorescence assay using antibodies against the ERmarker, calnexin, and the Golgi marker, GM130. The FGFR2immunofluorescence performed in cells with FGFR2A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3demonstrated colocalization with calnexin (Fig. 5B) andGM130 (Fig. 5C), suggesting that these mutant FGFR2 proteinspredominately localize in the ER/Golgi. In contrast, colocaliza-tion with calnexin and GM130 was not observed in cellsexpressing wild-type FGFR2. We next used biotin labeling asa complementary method to determine the cellular location ofwild-type and mutant FGFR2 proteins. Biotinylation and thesubsequent isolation of cell surface proteins revealed thatFGFR2 wild-type is mainly present at the cell surface, whereasFGFR2 A266_S267ins and FGFR2 290_291WI>C in the cyto-plasm (Fig. 6A). The specificity of cell surface labeling bybiotinylation was confirmed by the immunoblots for the cyto-plasmic protein, HSP90. Cells with FGFR2–TACC3 transloca-tion also showed cytoplasmic expression of FGFR2 (Fig. 6A).We finally evaluated whether these ER/Golgi-located mutantFGFR2 proteins are functional and can initiate signaling. Weperformed immunoprecipitation with the flow-through lysatesfrom the cell surface protein isolation assay, which contain

Figure 3.FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 form dimes in the absence of ligand. Unreduced lysates from NIH-3T3 cells expressing theindicated genetic alterations were probed for the formation of receptor dimerizations. For the ligand stimulation experiments, cells with FGFR2 wild-type werestimulated with FGF-1 and heparin for 8 hours before harvest. Open arrows, monomer bands; solid arrows, dimers.

Figure 4.Cells with FGFR2 A266_S267ins and those with FGFR2 290_291WI>C show stronger expression of partially glycosylated FGFR2. A, expression levels of FGFR2detected by FGFR2 antibody and Flag antibody. FGFR2 migrates on SDS-PAGE as two bands, a fully gycosylated 120-kDa band (solid arrow) and a partiallyglycosylated 110-kDa band (open arrow). B, lysates of cells with the indicated gene alteration were digested with Endo Hf and subsequently immunoblottedto detect FGFR2. Solid arrow, fully gycosylated 120-kDa FGFR2; open arrow, partially glycosylated 110-kDa FGFR2.






cytoplasmic proteins but not cell surface proteins. FLAG-taggedFGFR2 was immunoprecipitated from FGFR2 wild-type, FGFR2A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 celllysates, and the resulting precipitates were subjected to immu-noblot analysis with antibodies to FRS2. We found that FGFR2

A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3,which localize in the cytoplasm, coprecipitated with FRS2(Fig. 6B), suggesting that these proteins are in complex withFRS2 and hence could initiate signaling from intracellularcompartments.

Figure 5.Localization of FGFR2 wild-type, FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 by immunofluorescent staining. A, cells expressing the indicatedgene alterations were immunostained with Flag antibody (green) and with FGFR2 antibody (red). FGFR2 antibody H-80 was used for the experimentsin cells expressing FGFR2–TACC3. B, immunofluorescent analysis of colocalization of Flag (green) and the ER-resident marker protein, calnexin (red). C, costainingfor Flag (green) and the Golgi marker, GM130 (red). Magnification fields (�400) have been cropped and zoomed to facilitate the punctae visualization.

Figure 6.A, wild-type FGFR2 expresses on the cell surface, whereas FGFR2 A266_S267ins, FGFR2 290_291WI>C, and FGFR2–TACC3 express primarily in the cytoplasm.Cell surface proteins were biotinylated and precipitated with streptavidin beads (M). Streptavidin precipitates and the flow-through corresponding to theintracellular compartment (I) were immunoblotted and detected with FGFR2, Flag, and HSP90 (control for cytoplasmic proteins) B, FGFR2 A266_S267ins,FGFR2 290_291WI>C, and FGFR2–TACC3 interact with FRS2 in cytoplasm. Flow-through lysates fromAwere immunoprecipitatedwith antibodies to Flag antibody.The resulting precipitates were subjected to immunoblot analysis with antibodies against FRS2 and those against Flag.

Tanizaki et al.





DiscussionAberrant FGFR signaling has been implicated in the develop-

ment of several cancer types, including FGFR1 amplification inlung squamous cell carcinoma (7, 21), FGFR3 mutations inurothelial carcinomas (22), and recently reported fusions of FGFRfamily kinases with TACC or other genes in several cancer types(9, 11, 23, 24). In this study, we identify novel, albeit rare, FGFR2genetic alterations that are oncogenic and can be targeted withFGFR inhibitors in solid tumors, including lung cancer.

FGFR2 point mutations have been well described in studies ofdifferent diseases, such as craniosynotosis syndrome, gastric can-cer, and endometrial carcinomas (6, 25–27). Furthermore, recentstudies suggest that FGFR2 might define a molecular subset inother cancers as well, including NSCLC (8, 10, 19). FGFR2A266_S267insSTVVD found in an NSCLC patient and FGFR2290_291WI>C found in a cholangiocarcinoma patient localize tothe ECDof FGFR2. Althoughmutations in the ECDdoes not seemto promote direct activation of FGFR2 kinase activity, they couldbe oncogenic by changing the affinity to ligand (28, 29), or lead toligand independent dimerization and subsequent activation ofFGFR2 signaling (8, 30, 31). In this study, we detected elevatedbasal phosphorylation of FGFR and that of FRS2 in cells expres-sing these ECD mutations accompanied by dimer formation.Together with the in vivo results, our data indicate that FGFR2A266_S267ins and FGFR2 290_291WI>C are oncogenic muta-tions through dimer formation as a possible mechanism, but in aligand-independent manner.

We found that FGFR2A266_S267ins and FGFR2290_291WI>Chave altered receptor subcellular localization. Increased expressionlevels of the incompletely glycosylated receptor support this find-ing since the full N-glycosylation of FGFR2 is considered to beinvolved in normal trafficking of the receptor to the cell surface(32). Immunofluorescent costaining for ER or Golgi markersshowed that these mutated receptor forms are predominantlylocated within these intracellular compartments, whereas suchcolocalization was not detectable in cells expressing wild-typeFGFR2. Similar to our observation, other FGFR2 ECD mutationshave been reported to lead to diminished FGFR2 glycosylation,accompaniedwith altered localization of FGFR2 (32, 33). A crystalstructure study revealed that FGFR2 ECD mutation G271E cancause local structure perturbations and destabilization, whichcould interfere with dimers or receptor mislocalization by inter-fering with the correct receptor maturation and trafficking (33). Aprevious study has shown that incompletely glycosylated FGFR2ECD mutations are able to dimerize (8) and activate signal trans-duction in the ER/Golgi (32), consistent with our finding. Further-more, we showed that FGFR2–TACC3 localizes primarily in thecytoplasm, and more precisely in the ER/Golgi. Although thepapers reporting FGFR–TACC fusions are still limited, it has beenshown that FGFR3–TACC3 fusion results in enhanced expressionof this fusion gene (23) and that FGFR3–TACC3 localizes in thecytosol, in mitotic spindle polls (9). Conventional cell signalingmodels involve initiation of signaling cascades of receptor tyrosinekinases by ligand binding at the cells surface; however, recentfindings suggest that some of these signaling pathways can beinitiated within internal compartments (34, 35). Thus, an alteredexpression or localization pattern of FGFR2 protein could beinvolved in oncogenicmechanisms in addition to forming dimers.Taken together, our data suggest that these novel FGFR2 ECDmutations, A266_S277ins and 290_291WI>C, form dimers in a

ligand-independent fashion and activate signaling possiblythrough diminished glycosylation of the receptor and alteredlocalization in the ER/Golgi. However, it remains to be elucidatedwhether increased proportion of incompletely glycosylated FGFR2results either from less efficient trafficking to the cell surfaceor frommore rapid endocytosis and degradation after trafficking to the cellsurface. Moreover, the mechanism by which glycosylation isinvolved in receptor trafficking remains unclear, with furtherinvestigations needed.

In summary, we have identified and characterized several drug-sensitive FGFR2 gene alterations, providing a rationale for FGFR-targeted therapy in solid cancer. Reports of FGFR2 gene alterations,especially in NSCLC, are still limited. There are 314 FGFR2 muta-tions from 87 published studies in cBioPortal for cancer genomicsand one W290C mutation has been reported in a lung squamouscell carcinoma patient (36). Furthermore, recent study also reportedW290Cmutation in lung squamous cell carcinoma (8). These datasupport a rationale to investigate the frequency of these ECDalterations togetherwith the related clinic-pathologic features.Giventhe number of cancer patients all over the world, identifying even asmall subpopulation of patients who may respond to targetedtherapy can have significant therapeutic relevance.

Disclosure of Potential Conflicts of InterestP.S. Hammerman is a consultant/advisory board member for Astra Zeneca

and Janssen. G.R. Oxnard is a consultant/advisory board member for Novartis,Clovis, AstraZeneca, Boehringer-Ingelheim, and Genentech. P.A. Janne is aconsultant/advisory board member for Astra Zeneca, Boehriner Ingelheim,Pfizer, Genentech, Clovis Oncology, Merrimack Pharmaceuticals, Chugai Phar-maceuticals, and Sanofi. No potential conflicts of interest were disclosed by theother authors.

Authors' ContributionsConception and design: J. Tanizaki, M. Dodge, G.R. Oxnard, P.A. J€anneDevelopment of methodology: J. Tanizaki, M. Capelletti, C. Xu, M. Bahcall,E.M. TrickerAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J. Tanizaki, M. Capelletti, M. Dodge, M. Butaney,A. Calles, L.M. Sholl, G.R. Oxnard, P.A. J€anneAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J. Tanizaki, D. Ercan, M. Capelletti, M. Butaney,P.S. Hammerman, P.A. J€anneWriting, review, and/or revision of the manuscript: J. Tanizaki, M. Capelletti,M. Dodge, M. Bahcall, A. Calles, L.M. Sholl, P.S. Hammerman, G.R. Oxnard,K.-K. Wong, P.A. J€anneAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): J. Tanizaki, M. Butaney, A. Calles, K.-K. WongStudy supervision: K.-K. Wong, P.A. J€anne

AcknowledgmentsThe authors thank Dr. Mohammadi for helpful discussions and suggestions

on this article.

Grant SupportThis work is supported by The Uehara Memorial Foundation (J. Tanizaki),

the Cammarata Family Foundation Research Fund (M. Capelletti andP.A. J€anne), the Nirenberg Fellowship at the Dana-Farber Cancer Institute(M. Capelletti and P.A. J€anne), and by the National Cancer Institute (K08CA163677; P.S. Hammerman). P.S. Hammerman is supported by a V Foun-dation Scholar Award.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received December 22, 2014; revised March 22, 2015; accepted April 28,2015; published OnlineFirst June 5, 2015.






References1. Lynch TJ, Bell DW, Sordella R, Gurubhagavatula S, Okimoto RA, Brannigan

BW, et al. Activating mutations in the epidermal growth factor receptorunderlying responsiveness of non-small-cell lung cancer to gefitinib.N Engl J Med 2004;350:2129–39.

2. Solomon BJ, Mok T, Kim DW, Wu YL, Nakagawa K, Mekhail T, et al. First-line crizotinib versus chemotherapy in ALK-positive lung cancer. N EnglJ Med 2014;371:2167–77.

3. Oxnard GR, Binder A, Janne PA. New targetable oncogenes in non-small-cell lung cancer. J Clin Oncol 2013;31:1097–104.

4. GLOBOCAN 2012: Estimated cancer incidence, mortality and prevalenceworldwide in 2012. Available from: http://globocan.iarc.fr/Default.aspx.Accessed June 2014.

5. Grose R, Dickson C. Fibroblast growth factor signaling in tumorigenesis.Cytokine Growth Factor Rev 2005;16:179–86.

6. Kelleher FC, O'Sullivan H, Smyth E, McDermott R, Viterbo A. Fibroblastgrowth factor receptors, developmental corruption andmalignant disease.Carcinogenesis 2013;34:2198–205.

7. Weiss J, Sos ML, Seidel D, Peifer M, Zander T, Heuckmann JM, et al.Frequent and focal FGFR1 amplification associates with therapeuticallytractable FGFR1 dependency in squamous cell lung cancer. Sci Transla-tional Med 2010;2:62ra93.

8. Liao RG, Jung J, Tchaicha J, Wilkerson MD, Sivachenko A, Beauchamp EM,et al. Inhibitor-sensitive FGFR2 and FGFR3 mutations in lung squamouscell carcinoma. Cancer Res 2013;73:5195–205.

9. Singh D, Chan JM, Zoppoli P, Niola F, Sullivan R, Castano A, et al.Transforming fusions of FGFR and TACC genes in human glioblastoma.Science (New York, NY) 2012;337:1231–5.

10. Wu YM, Su F, Kalyana-Sundaram S, Khazanov N, Ateeq B, Cao X, et al.Identification of targetable FGFR gene fusions in diverse cancers.Cancer Discov 2013;3:636–47.

11. Capelletti M, Dodge ME, Ercan D, Hammerman PS, Park SI, Kim J, et al.Identification of Recurrent FGFR3-TACC3 Fusion Oncogenes from LungAdenocarcinoma. Clin Cancer Res 2014;20:6551–8.

12. Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J, et al.Development and validation of a clinical cancer genomic profiling testbased on massively parallel DNA sequencing. Nat Biotechnol 2013;31:1023–31.

13. Lipson D, Capelletti M, Yelensky R, Otto G, Parker A, Jarosz M, et al.Identification of new ALK and RET gene fusions from colorectal and lungcancer biopsies. Nat Med 2012;18:382–4.

14. Vaishnavi A, Capelletti M, Le AT, Kako S, Butaney M, Ercan D, et al.Oncogenic and drug-sensitive NTRK1 rearrangements in lung cancer.Nat Med 2013;19:1469–72.

15. Wagle N, Berger MF, Davis MJ, Blumenstiel B, Defelice M, Pochanard P,et al. High-throughput detection of actionable genomic alterations inclinical tumor samples by targeted, massively parallel sequencing.Cancer Discov 2012;2:82–93.

16. Sasaki T, Okuda K, Zheng W, Butrynski J, Capelletti M, Wang L, et al. Theneuroblastoma-associated F1174L ALK mutation causes resistance to anALK kinase inhibitor in ALK-translocated cancers. Cancer Res 2010;70:10038–43.

17. ZhouW, Ercan D, Chen L, Yun CH, Li D, Capelletti M, et al. Novel mutant-selective EGFR kinase inhibitors against EGFR T790M. Nature 2009;462:1070–4.

18. Tanizaki J, Okamoto I, Sakai K, Nakagawa K. Differential roles of trans-phosphorylated EGFR, HER2, HER3, and RET as heterodimerisation part-ners of MET in lung cancer with MET amplification. Br J Cancer 2011;105:807–13.

19. Borad MJ, Champion MD, Egan JB, Liang WS, Fonseca R, Bryce AH, et al.Integrated genomic characterization reveals novel, therapeutically relevantdrug targets in FGFR and EGFR pathways in sporadic intrahepatic cho-langiocarcinoma. PLoS Genet 2014;10:e1004135.

20. Feige JJ, Baird A. Glycosylation of the basic fibroblast growth factorreceptor. The contribution of carbohydrate to receptor function. J BiolChem 1988;263:14023–9.

21. Dutt A, Ramos AH, Hammerman PS, Mermel C, Cho J, Sharifnia T, et al.Inhibitor-sensitive FGFR1 amplification in human non-small cell lungcancer. PloS One 2011;6:e20351.

22. Cappellen D, De Oliveira C, Ricol D, de Medina S, Bourdin J, Sastre-GarauX, et al. Frequent activating mutations of FGFR3 in human bladder andcervix carcinomas. Nat Genet 1999;23:18–20.

23. Parker BC, Annala MJ, Cogdell DE, Granberg KJ, Sun Y, Ji P, et al. Thetumorigenic FGFR3-TACC3 gene fusion escapes miR-99a regulation inglioblastoma. J Clin Invest 2013;123:855–65.

24. Wang R, Wang L, Li Y, HuH, Shen L, Shen X, et al. FGFR1/3 tyrosine kinasefusions define a unique molecular subtype of non-small cell lung cancer.Clin Cancer Res 2014;20:4107–14.

25. Jang JH, Shin KH, Park JG. Mutations in fibroblast growth factor receptor 2andfibroblast growth factor receptor 3 genes associatedwith human gastricand colorectal cancers. Cancer Res 2001;61:3541–3.

26. Dutt A, Salvesen HB, 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.

27. Pollock PM, Gartside MG, Dejeza LC, Powell MA, Mallon MA, Davies H,et al. Frequent activating FGFR2 mutations in endometrial carcinomasparallel germline mutations associated with craniosynostosis and skeletaldysplasia syndromes. Oncogene 2007;26:7158–62.

28. Yu K, Herr AB, Waksman G, Ornitz DM. Loss of fibroblast growth factorreceptor 2 ligand-binding specificity in Apert syndrome. Proc Natl Acad SciU S A 2000;97:14536–41.

29. Anderson J, Burns HD, Enriquez-Harris P, Wilkie AO, Heath JK. Apertsyndrome mutations in fibroblast growth factor receptor 2 exhibitincreased affinity for FGF ligand. Hum Mol Genet 1998;7:1475–83.

30. Robertson SC, Meyer AN, Hart KC, Galvin BD, Webster MK, DonoghueDJ. Activating mutations in the extracellular domain of the fibroblastgrowth factor receptor 2 function by disruption of the disulfide bond inthe third immunoglobulin-like domain. Proc Natl Acad Sci U S A1998;95:4567–72.

31. Neilson KM, Friesel R. Ligand-independent activation of fibroblast growthfactor receptors by point mutations in the extracellular, transmembrane,and kinase domains. J Biol Chem 1996;271:25049–57.

32. HatchNE,HudsonM, SetoML,CunninghamML,BothwellM. Intracellularretention, degradation, and signaling of glycosylation-deficient FGFR2 andcraniosynostosis syndrome-associated FGFR2C278F. J Biol Chem 2006;281:27292–305.

33. GartsideMG, ChenH, Ibrahimi OA, Byron SA, Curtis AV,Wellens CL, et al.Loss-of-function fibroblast growth factor receptor-2 mutations in mela-noma. Mol Cancer Res 2009;7:41–54.

34. Bivona TG, Perez De Castro I, Ahearn IM, Grana TM, Chiu VK, Lockyer PJ,et al. Phospholipase Cgamma activates Ras on the Golgi apparatus bymeans of RasGRP1. Nature 2003;424:694–8.

35. Sato M, Ueda Y, Takagi T, Umezawa Y. Production of PtdInsP3 at endo-membranes is triggered by receptor endocytosis. Nat Cell Biol 2003;5:1016–22.

36. cBioPortal for cancer genomics. Available from: http://www.cbioportal.org/index.do. Accessed June 2014.


Tanizaki et al.




2015;75:3139-3146. Published OnlineFirst June 5, 2015.Cancer Res Junko Tanizaki, Dalia Ercan, Marzia Capelletti, et al. Extracellular Domain of FGFR2Identification of Oncogenic and Drug-Sensitizing Mutations in the

Updated version

10.1158/0008-5472.CAN-14-3771doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2015/06/05/0008-5472.CAN-14-3771.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/75/15/3139.full#ref-list-1

This article cites 34 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/75/15/3139.full#related-urls

This article has been cited by 1 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://cancerres.aacrjournals.org/content/75/15/3139To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-14-3771

http://cancerres.aacrjournals.org/content/suppl/2015/06/05/0008-5472.CAN-14-3771.DC1

http://cancerres.aacrjournals.org/content/75/15/3139.full#ref-list-1

http://cancerres.aacrjournals.org/content/75/15/3139.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/75/15/3139


identiﬁcation of oncogenic and drug-sensitizing mutations in the … · therapeutics, targets,...

Documents