InVivo DisruptionofanRb E2F Ezh2SignalingLoopCauses ......Taq PCR master mix (Promega) and 1 mL of cDNA as a template.Melting curves were performed to verify speciﬁcity and absence

Molecular and Cellular Pathobiology

In VivoDisruption of an Rb–E2F–Ezh2 Signaling Loop CausesBladder Cancer

Mirentxu Santos1,2, M�onica Martínez-Fern�andez1,2, Marta Due~nas1,2, Ram�on García-Escudero1,2,Bego~na Alfaya1, Felipe Villacampa2,3, Cristina Saiz-Ladera1, Clotilde Costa1, Marta Oteo2,4, Jos�e Duarte2,3,Victor Martínez3, Ma Jos�e G�omez-Rodriguez2,3, Ma Luisa Martín3, Manoli Fern�andez3, Patrick Viatour5,Miguel A. Morcillo2,4, Julien Sage5, Daniel Castellano2,3, Jose L. Rodriguez-Peralto6, Federico de la Rosa2,3,and Jes�us M Paramio1,2

AbstractBladder cancer is a highly prevalent human disease in which retinoblastoma (Rb) pathway inactivation and

epigenetic alterations are common events. However, the connection between these two processes is stillpoorly understood. Here, we show that the in vivo inactivation of all Rb family genes in the mouse urotheliumis sufficient to initiate bladder cancer development. The characterization of the mouse tumors revealedmultiple molecular features of human bladder cancer, including the activation of E2F transcription factor andsubsequent Ezh2 expression and the activation of several signaling pathways previously identified as highlyrelevant in urothelial tumors. These mice represent a genetically defined model for human high-gradesuperficial bladder cancer. Whole transcriptional characterizations of mouse and human bladder tumorsrevealed a significant overlap and confirmed the predominant role for Ezh2 in the downregulation of geneexpression programs. Importantly, the increased tumor recurrence and progression in human patients withsuperficial bladder cancer is associated with increased E2F and Ezh2 expression and Ezh2-mediated geneexpression repression. Collectively, our studies provide a genetically defined model for human high-gradesuperficial bladder cancer and demonstrate the existence of an Rb–E2F–Ezh2 axis in bladder whosedisruption can promote tumor development. Cancer Res; 74(22); 6565–77. �2014 AACR.

IntroductionBladder cancer is a complex and heterogeneous disease

caused by both genetic and environmental factors (1). Mosttumors arise from the inner lining epithelial cells of the bladder

wall, being more than 90% transitional cell carcinomas. Atdiagnosis, two major types of transitional cell carcinomas canbe identified according to the pathologic characteristics: twothirds of patients present with superficial non–muscle-inva-sive bladder cancer (NMIBC) tumors, the remaining one thirdof patients present (or develops) a highly aggressive, muscle-invasive bladder cancer (MIBC) that leads to the death of 50%of patients. In general, NMIBCs have a favorable prognosis andare treated by transurethral resections and intravesical ther-apy. However, these tumors show a high rate of recurrence,which in some cases can progress into MIBC. This makesnecessary a regular surveillance with cystoscopy and urinecytology indefinitely (EAU guidelines; ref. 2). Therefore, NMIBCrepresents one of the most costly malignancies to health caresystems in developed countries (3). In MIBC, surgery, radio-therapy, and chemotherapy are effective treatments for thedisease, but there has been little progress in survival in the last20 years.

The interpretation of the genetic landscape of bladdercancer has largely been influenced by the current clinical(NMIBC vs. MIBC) and pathologic (stage and grade; papillaryvs. solid-invasive) classifications (4). Molecular taxonomy pro-vides a distinct viewof bladder cancer subphenotypes and theirrelationship with the disease progression (5). TP53 mutationsand RB1 inactivation are more prevalent in MIBC and mayfavor tumor progression and muscle invasion (6). However,

1Unidad deOncologíaMolecular, CIEMAT (ed70A),Madrid, Spain. 2Unidadde Oncogen�omica, Instituto de Investigaci�on, Iþ12 University Hospital "12de Octubre," UCM, Madrid, Spain. 3Unidad de Uro-Oncología, HospitalUniversitario "12 de Octubre", Madrid, Spain. 4Unidad de AplicacionesBiom�edicas y Farmacocin�etica, CIEMAT (ed 12), Madrid, Spain. 5Depart-ments of Pediatrics andGenetics, Stanford University, Stanford, California.6Servicio de Anatomía Patol�ogica, Hospital Universitario "12 de Octubre",Instituto de Investigaci�on Iþ12, Madrid, Spain.

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

M. Santos and M. Martínez-Fern�andez contributed equally to this articleand share first authorship.

Current address for P. Viatour: Center for ChildhoodCancer Research, TheChildren's Hospital of Philadelphia, Dept. of Pathology and LaboratoryMedicine, Perelman School of Medicine, University of Pennsylvania,Philadelphia, PA 19104.

Corresponding Author: Jesus M. Paramio, Department of BasicResearch, CIEMAT (ed 70A), Ave Complutense 40; Madrid 28040, Spain.Phone: 34-914962517; Fax: 34-913466484; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-14-1218

�2014 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 6565

on September 2, 2021. © 2014 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 24, 2014; DOI: 10.1158/0008-5472.CAN-14-1218

http://cancerres.aacrjournals.org/

there is insufficient evidence to use them as independentpredictors of poor outcome in the clinical practice of patientswith bladder cancer (7). Low-grade NMIBCs are genomicallystable, whereas high-grade NMIBCs and MIBCs display geno-mic instability (8). FGFR3 and PIK3CA mutations are moreprevalent in NMIBCs and can also be predictive of localrecurrences in NMIBCs (9). Bladder cancer is also character-ized by significant alterations in genes involved in chromatinregulation, affecting in particular those genes implicated inhistone modification (10).

In an attempt to reproduce human tumors, multiple genet-ically engineered mouse models have been generated (11). Inthese, the expression of large TAg from SV40 [leading to theelimination of all retinoblastoma (Rb) family members andp53] can induce bladder aggressive tumors in transgenic mice(11, 12). This is in contrast with the absence of bladder tumorsupon Rb1 ablation in the urothelium (13). In this regard, we didnot observe bladder tumor development in large cohorts ofmice bearing the specific elimination of Rb gene directed bykeratin K14cre expression (14), which is active in the adultmouse urothelium (15), even in the absence of p107 (16), E2F1(17), or p130 (18). These data indicate the existence of largeoverlapping roles for the Rb family members in bladderurothelial cells, similar to that reported for other mouseepithelial cells (19). To circumvent this possible problem andto specifically delete Rb family function in the urothelium ofadult mice, we took advantage of amouse strain bearing floxedalleles of Rb1 and Rbl2 genes and whole deficiency in Rbl1 gene(RbF/F;p130F/F;p107�/�; ref. 20). Conditional gene deletion inthe urothelium was induced by delivering an AdenoCre intothe bladder lumen of adult male mice (13). Here, we show thatthe ablation of all three Rb family members in the mousebladder urothelium leads to tumor development. The molec-ular characteristics of these triple knockout mouse tumorscombined with their genomic characterization provide newpossible molecular mechanisms of bladder cancer develop-ment. Collectively, the present data support a molecularconnection between Rb/E2F andEzh2 thatmay explain humanNMIBC recurrence and progression.

Materials and MethodsPatients

Tumor samples and medical records were analyzed from 77patients who had been consecutively evaluated at the UrologyDepartment of the University Hospital "12 de Octubre"between October 2009 and March 2011 and diagnosed witha Ta or T1 bladder cancer. Informed consent was obtainedfrom all patients and the study was approved by the EthicalCommittee for Clinical Research of University Hospital "12 deOctubre." The pathologic and clinical data, as well as thesample recollection and preservation procedures, have beenreported elsewhere (9). Samples and united data from patientsincluded in this study were provided by the Biobanco iþ12 inthe Hospital 12 de Octubre integrated in the Spanish HospitalBiobanks Network (RetBioH; www.redbiobancos.es) followingstandard operation procedures with appropriate approval ofthe RETHICAL AN Scientific Committees.

Retinoblastoma family–deficient bladder tumor mousemodel

All the animal experiments were approved by the AnimalEthical Committee and conducted in compliance with Centrode Investigaciones Energ�eticas, Medioambientales y Tec-nol�ogicas (CIEMAT) Guidelines. RbF/F;p130F/F;p107�/� micewere generated by breeding RbF/F;p107�/� (16) and p130F/F

(20) mice. Adenovirus expressing Cre recombinase wasobtained from University of Iowa's Vector Core Facility(www.uiowa.edu) and surgically delivered to the bladderlumen as previously described (13). At the time of sacrifice,tissues were collected and processed as previously reported(16, 17, 21).

Tissue microarrayThe construction and analysis of tissue microarray contain-

ing the human samples has been reported elsewhere (9). Atleast two representative duplicate cores for each case werescored.

ImmunohistochemistryImmunohistochemical analyses of human and mouse were

performed essentially as previously described (9). Antibodiesusedwere: anti-Ezh2 (Abnova; mAb9542 diluted 1:200), anti-K5(Covance; diluted 1:500), anti-K8 (TROMA, University of Iowa;rat mAb diluted 1:10), anti-laminin (Sigma; L9393 diluted1:100), anti-p63 (Santa Cruz Biotechnology; mAb 4A4), anti-pRb (Santa Cruz Biotechnology; sc50 diluted 1:50), and anti-p130 (Santa Cruz Biotechnology; sc317 diluted 1:50). Signal wasamplified using avidin-peroxidase (ABC Elite Kit; Vector Labs),and peroxidase was visualized using 3,30-diaminobenzidine asa substrate (DAB kit, Vector Labs). Negative control slides wereobtained by replacing primary antibodies with PBS (data notshown). Scoring of the results and selection of the thresholds,internal controls for reactivity of each antibody, and tissuecontrols for the series were done according to previouslypublished methods (9). Mice were injected intraperitoneallywith bromodeoxyuridine (BrdUrd; 0.1 mg/g weight in 0.9%NaCl; Roche) 1 hour before sacrifice. BrdUrd incorporationwasmonitored in formalin-fixed sections using an anti-BrdU anti-body (Roche) as described (16, 17, 21).

Determination of FGFR3 and PIK3CA gene mutationThe presence of mutations in the PIK3CA and FGFR3

genes was assessed in tumor gDNA by PCR test (QIAGEN)and/or by snapshot technique has been reported elsewhere(9).

RT-qPCRTotal RNAwas isolated frommouse andhuman samples and

analyzed by RT-qPCR as previously described (9, 16, 17, 21, 22).Briefly, total RNA was isolated using miRNeasy Mini Kit(Qiagen) according to the manufacturer's instructions andDNA was eliminated (RNAse-Free DNAse Set Qiagen). Reversetranscription was performed using the Omniscript RT Kit(Qiagen) and a primer mix specific for all genes of interest inthe case of human samples, and oligo dT primer for the mousesamples, using 10 ng and 1 mg of total RNA, respectively. PCR

Santos et al.

Cancer Res; 74(22) November 15, 2014 Cancer Research6566




was performed in a 7500 Fast Real Time PCR System using GoTaq PCR master mix (Promega) and 1 mL of cDNA as atemplate. Melting curves were performed to verify specificityand absence of primer dimerization. Reaction efficiency wascalculated for each primer combination, and TBP gene wasused as reference gene for normalization (23). The sequencesof the specific oligonucleotides used are listed in Supple-mentary Table S1. Discrimination between samples showingincreased or decreased relative expression was made usingthe median.

Whole transcriptome analysisGenome-wide transcriptome experiments were per-

formed using the Affymetrix HuGene-1_0-st-v1 microarrayor Mo Gene-1_0-st-v1 at the Genomics Facility of the CancerResearch Center (Salamanca, Spain) using standard proce-dures (see Supplementary Information). Datasets have beendeposited in GEO (GSE38264). Mouse to human comparisonwas performed essentially as described elsewhere (22).

ImmunoblotImmunoblotting was performed as described previously

(16, 21). Briefly, dissected bladder tumors were disrupted byfreeze-thawing cycles in lysis buffer [200 mmol/L4-(2-hydro-xyethyl)-1-piperazineethanesulfonic acid, pH 7.9, 25% glycerol,400 mmol/L NaCl, 1 mmol/L EDTA, 1 mmol/L ethylene glycoltetraacetic acid, 1 mg/mL aprotinin, 1 mg/mL leupeptin, 1mmol/L phenylmethylsulfonyl fluoride, 20 mmol/L NaF, 1mmol/L NaPPi, 1 mmol/L Na3VO4, 2.5 mmol/L dithiothreitol]and centrifuged to obtain supernatant containing total pro-tein. Thirty-five micrograms of protein per sample wasresolved in SDS-PAGE and transferred to nitrocellulose mem-branes (Amersham).Membraneswere blockedwith 5%non-fatmilk diluted in TBS and incubated with the appropriateantibodies diluted in TBS-T 0.5% BSA. Secondary antibodieswere purchased from Jackson ImmunoResearch. Super Sig-nal West Pico Chemiluminescence Substrate (Pierce) wasused according to the manufacturer's recommendations tovisualize the bands. Antibodies used are against E2F1, E2F2,E2F3, pRb, p130 (Santa Cruz Biotechnology); Ezh2 (Abnova);phosphoSer473-Akt (Epitomics or Cell Signaling); phospho-Erk1/2 (Thr202/Tyr204-P), p19Arf (Abcam); p53 (Novocas-tra); and Stat3-Tyr805-P and S6-P (Cell Signaling). Loadingwas controlled by using an anti-actin antibody (Santa CruzBiotechnology).

Statistical analysisComparisons were performed using the Wilcoxon–Mann–

Whitney test (for two groups) and the Student t test for pairedsamples showing normal distribution. Survival analyses (recur-rence-free or tumor progression in recurrence) according tovarious variables were performed using the Kaplan–Meiermethod, and differences between the patient groups weretested by the log-rank test. Discrimination between samplesshowing increased or decreased gene expression was madeusing the median. Overlapping significance was monitored byexact Fisher test. SPSS 17.0 and Graph prism 5.0 software wereused.

ResultsAblation of all Rb family members causes bladder tumordevelopment

To study the functional consequences of Rb family loss in theurothelium of adult mice, conditional gene deletion mediatedby delivering an AdenoCre into the bladder lumen (13) of adultmale RbF/F;p130F/F;p107�/� mice (20) was used. We observedthe development of bladder lesions in all the AdenoCre-infected mice (n ¼ 18; Fig. 1A and B) but not in control mice(noninfected littermates or infected with AdenoGFP, n ¼ 30).The presence of these lesions was easily monitored by CT(Supplementary Fig. S1) and in some cases was accompaniedby hematuria. The histologic analysis of these lesions revealedpapillary masses growing inward to the bladder lumen (Fig. 1Aand Supplementary Fig. S1), displaying histologic featuresof high-grade non–muscle-invasive carcinoma. The mousebladder tumors were keratin K5–positive, keratin K8–negative(Fig. 1C), displayed high proliferation (analyzed by BrdUrdincorporation; Fig. 1C), and invaded the basal lamina (asdetermined by laminin staining; Fig. 1C). When mice werefollowed up to 1 year after infection (Fig. 1B), no visiblemetastases were detected.

We observed the ablation of all Rb family members (Sup-plementary Fig. S2) and the increased expression of E2F1, 2,and 3a (Fig. 1D) in these tumors, by both immunoblotting andRT-qPCR analyses (Supplementary Fig. S3). In addition, tumorswere characterized by the upregulation of p19Arf and p53 andthe activation of Erk, S6, and Stat3 signaling pathways, asdemonstrated by immunoblot analysis using phospho-specificantibodies (Fig. 1D). In contrast, tumors did not showincreased Akt activity (Fig. 1D).

Tumors in Rb-deficient mouse resemble human bladdercancer

To further characterize the Rb familymutantmouse bladdertumors and to compare with human bladder cancer, a wholetranscriptome characterization of the mouse samples wasperformed. This analysis revealed a differential expression of3,053 transcripts (1,849 downregulated and 1,204 upregulated)in tumors compared with normal bladder (Fig. 2A and Sup-plementary Table S2). Gene ontology characterization of thesederegulated transcripts indicated that downregulated geneswere involved in oxidoreduction, small GTPases transductionsignaling, cell adhesion, and translation processes (Fig. 2B andSupplementary Table S3), whereas the upregulated genes wereprimarily involved in cell-cycle and DNA repair processes, aswould be expected in tumorswith high E2F activity (Fig. 2B andSupplementary Table S4). To characterize the putative tran-scription factors controlling these differentially expressedgenes, we performed chromatin immunoprecipitation (ChIP)enrichment analyses (ChEA; ref. 24). These revealed the pos-sible involvement of polycomb repressive group 2 (PRC2) indownregulated genes (Fig. 2C and Supplementary Table S5),whereas the upregulated genes showed a primary involvementof E2F and Myc, together with histone demethylases (Fig. 2Cand Supplementary Table S6). Gene set enrichment analysis(GSEA) showed significant overlap of the upregulated anddownregulated genes in mouse bladder tumors with various

Disruption of Rb–E2F–Ezh2 Loop in Bladder Cancer

www.aacrjournals.org Cancer Res; 74(22) November 15, 2014 6567




human bladder datasets and also with stem cell signatures,including genes silenced in embryonic stem cells byH3K27me3(Supplementary Table S7), thus reinforcing the possible role ofPRC2 in gene expression deregulation. The comparative anal-yses between the downregulated (Fig. 2D, top, and Supple-mentary Table S8) or upregulated (Fig. 2D, bottom, andSupplementary Table S9) mouse genes and human bladdertumors available in Oncomine database (25) showed a verysignificant overlapping with multiple external bladder cancerdatasets comparing normal versus tumor samples, tumorswith poor clinical outcome, and the presence of specific genemutations in human tumors, including KDM6A and RB muta-tions. Remarkably, this comparison (Fig. 2D and Supplemen-tary Table S9) also showed a statistically significant overlapbetween the upregulated mouse genes and upregulated genesin human recurrent bladder tumors and also between thedownregulated genes in mouse tumors and multiple datasets

identifying genes silenced by PRC2 in stem cells (Supplemen-tary Table S8).

Increased Ezh2 in Rb family–deficient mouse bladdertumors

The observed similarities between downregulated genes inmouse bladder tumors and those being silenced by PRC2prompted us to analyze possible changes in the expression ofPRC2 elements, Ezh2, Suz12, and Eed, in themouse tumors. RT-qPCR analysis (Fig. 3A) showed significantly increased expres-sion of Ezh2 and Suz12 genes, but not Eed in mouse tumorscompared with control samples, the increased expression ofEzh2 gene being more relevant. Immunoblot analysis con-firmed the increased expression of Ezh2 protein (Fig. 3B), thecatalytically active member of the PRC2 complex and respon-sible for trimethylation of H3K27. Finally, immunohistochem-ical analysis of Ezh2 expression (Fig. 3D) comparedwith that of

Figure 1. In vivo Rb family ablationcauses bladder cancerdevelopment in mice. A, exampleof external (top left) or internal (topright) view of a bladder from anRbF/F;p130F/F;p107�/� mouse11 months after inducing the generecombination by AdenoCreinfection. Bottom, an example ofH&E-stained section of the lesionsshown in top. Bar, 500 mm.B,Kaplan–Meier curve showing thekinetics of tumor development asassessed by CT scan in theAdenoCre-injected mice(n ¼ 18). C, representativeimmunohistochemistry of a mousetumor stained for keratin K5,BrdUrd, keratin K8, and lamininas indicated. Bar, 150 mm.D, immunoblot analyses of control(uninfected littermates bladdersamples) and mouse bladdertumors (TKO, triple knockout)showing the expression of thequoted proteins. Actin was used tonormalize protein loading.

Santos et al.





Figure 2. Transcriptome analysis of mouse bladder tumors. A, heatmap showing the distribution of genes (rows) identifying tumors and normal controluninfectedmouse bladder samples. Each column represents a sample. A red (overexpressed) toblue (downregulated) scheme following the above scale limits(in log2 scale) is shown.Numberson the right denote thenumberof transcripts of eachgroup (upregulatedor downregulated).B, representativeGeneOntologybiologic processes categories affecting the functions of the downregulated (top) or upregulated (bottom) genes in mouse bladder tumors. Numbers on theright of each bar indicate the genes on COBP category. C, relative relevance of transcription factors and histone-modifying enzymes in the genesdownregulated (top) or upregulated (bottom) in tumors as obtainedbyChEA. The relevance of each factor is providedby theP value (in�log10 scale). Numberson the right of eachbar represent genes boundby each transcription factor from thedatabase.D, summary of relevant overlap betweendownregulated (top) orupregulated (bottom) genes in mouse bladder triple knockout (TKO) tumors with human bladder tumors from the Oncomine database. The differentcomparison concepts (tumor vs. normal, mutation type, clinical outcome, and recurrence) are provided for each group. P values were obtained by the exactFisher test. The number of overlapping genes is shown for each dataset.






urothelium basal layer marker p63 (Fig. 3C) revealed theincreased expression of Ezh2 in mouse tumors, but not in theadjacent normal urothelium.

Genome-wide transcriptome analysis of humanrecurrent bladder tumor

These observations indicate that the mouse Rb familymutant tumors represent a bona fide model of human bladdercancer. The pathologic characteristics resemble those ofhuman high-grade NMIBC, a possibility also reinforced by thegenomic comparisons. To confirm this, a whole transcriptomestudy was performed using a series of human NMIBC samplesfocusing on differential gene expression between recurrent andnonrecurrent tumors. The analysis identified 351 transcripts(162 overexpressed and 299 underexpressed) differentiallyexpressed in recurrent tumors (Fig. 4A and SupplementaryTable S10), which also discriminated normal tissue. Remark-ably, this classification did not discriminate between Ta andT1 stages, high- and low-grade tumors, or tumors bearingFGFR3 mutations (Fig. 4A). On the contrary, tumors bearingPIK3CA gene mutation were predominantly associated withnonrecurrent samples (Fig. 4A). These findings may helpexplain our previously reported association between PIK3CAgene alterations and reduced recurrence in superficial blad-der tumors (9). Gene Ontology (GOBP; SupplementaryTables S11 and S12) and GSEA (Supplementary Table S13)revealed that the upregulated genes in recurrent tumorsplayed a major role in cell-cycle control, proliferation, andproteosomal protein degradation, whereas the downregu-lated genes displayed an association with ribosome, 30

untranslated region (UTR)-mediated translational regula-tion and protein translation. The GSEA also indicated sim-ilarities with stem cell transcription and PRC2 silencing(Supplementary Table S13). The comparison of these differ-entially expressed genes with other human bladder cancerdatasets in Oncomine database revealed a significant over-lap and trend with previously reported bladder cancerdatasets (Supplementary Table S14). Of note, we found thatthe downregulated genes in recurrent samples of our studywere able to identify early recurrence of NMIBC in anexternal dataset (Fig. 4B; ref. 26). ChEA of these deregulatedgenes revealed, besides the common involvement of multipletranscription factors in all datasets (Supplementary TablesS15 and S16), a statistically significant involvement of his-tone-modifying enzymes, affecting histone methyl trans-ferases and demethylases, and including the PRC2 members:Ezh2, Suz12, and Eed (Fig. 4C). In agreement, we alsoobserved a significant overlap with H3K27me3 in down-regulated genes in recurrent tumors by GSEA (Supplemen-tary Table S13).

Finally, we integrated our transcriptome data from bothmouse and human genes in a common dataset. The unsuper-vised classification of this common dataset according to (i) theexpression of genes discriminating mouse normal and tumorbladder samples (Fig. 4D, top) or (ii) human genes discrimi-nating human recurrent and nonrecurrent NMIBC samples(Fig. 4D, bottom) invariably showed that mouse tumors wereclustered with human recurrent tumors. Thus, the mouse

bladder tumors initiated by the ablation of all retinoblastomafamily members represent a putative model of human recur-rent NMIBC.

Increased EZH2 expression in human recurrent bladdertumors

Genomic data in mouse and human bladder tumors ana-lyzed pointed to amajor involvement of PRC2, and in particularEzh2, in the development and recurrence of NMIBC. To furthersupport this, those genes, identified by ChEA to be bound byany PRC2 element or specifically by Ezh2 (n ¼ 38 and 12,respectively), were loaded into Oncomine database. A statis-tically significant overlap was found with the Lee bladderdataset (26) in the case of PRC2-bound genes (overlappinggenes: n ¼ 4 of 38, P ¼ 0.00044, OR, 12.7). Moreover, we alsoobserved a significant overlap with both PRC2- and Ezh2-bound genes in another external dataset (overlapping genes:n¼ 5 of 38,P¼ 0.002, OR, 9.1; andn¼ 3 of 12,P¼ 0.008, OR, 13.6,respectively; ref. 27). Importantly, even these limited numbersof overlapping genes were able to discriminate patients ofNMIBCwith high likelihood of recurrence (Supplementary Fig.S4A–S4C). These observations point to a primordial role ofEzh2-mediated gene silencing in NMIBC early recurrence.

We next analyzed expression changes of PRC2members in aseries of human NMIBC samples previously described (9). RT-qPCR analysis revealed a significant increase of EZH2 expres-sion (Fig. 5A) but not of SUZ12 and EED (SupplementaryFig. S5A and S5B) in tumors compared with paired normalbladder samples. The EZH2 gene expression was also higher inrecurrent tumors than in nonrecurrent (Fig. 5A). When thepatients were stratified according to EZH2 expression in atissue microarray (Fig. 5C and data not shown), we found thatincreased expression of EZH2 correlated with tumor recur-rence (Fig. 5D, P ¼ 0.000021).

A possible involvement of E2F in Ezh2-deregulatedexpression

The expression of EZH2 gene is regulated by E2F activity(28). Accordingly, we monitored whether EZH2 expressioncorrelated with the expression of different E2Fs in our seriesof human bladder tumor samples. This analysis revealed astriking correlation between EZH2 and E2F3a gene expression(Fig. 6A). We also observed that E2F3a expression was signif-icantly higher in tumor than in normal samples and in recur-rent than in nonrecurrent tumors (Fig. 6B). Importantly,increased E2F3a expression levels stratified patients withrecurrent NMIBC (Fig. 6C). On the other hand, E2F1, 2, and3b gene expression did not discriminate between normal andtumor samples (Supplementary Fig. S6A) or between recurrentand nonrecurrent human NMIBC (Supplementary Fig. S6B).Furthermore, the expression of E2F1, 2, and 3b genes did notshow significant correlation with EZH2 gene expression(Supplementary Fig. S6C, S6E, and S6G) and did not allowstatistical stratification of recurrence in our series ofpatients with NMIBC (Supplementary Fig. S6D, S6F, andS6H). These results indicate that the early recurrence medi-ated by increased EZH2 expression could be facilitatedprimarily by increased E2F3a expression. However, a

Santos et al.





Figure 3. Mouse bladder tumors displayincreased Ezh2 expression A, expression ofEzh2, Suz12, and Eed genes in control(uninfected bladder samples) and mousebladder tumors as assessed by RT-qPCR (withrespect to Tbp). P value was obtained by theMann–Whitney t test; mean and SEM aredenoted in red. B, immunoblot analyses ofcontrol (uninfectedbladder samples) andmousebladder triple knockout (TKO) tumorsshowing the expression of Ezh2 protein.Actin was used to normalize protein loading.C, representative examples of theimmunohistochemical analysis of mousebladder tumor showing the expression ofp63. Bottom, higher magnifications of thecorresponding areas denoted in top.D, representative examples of theimmunohistochemical analysis of mousebladder tumor showing the expressionof Ezh2. Bottom, higher magnifications of thecorresponding areas denoted in top.Bar in top, 500 mm; bottom, 150 mm.






Figure 4. Transcriptome studies of human NMIBC recurrence. A, heatmap showing the distribution of genes (rows) identifying recurrent tumor samples. Eachcolumn represents a sample. ThemutationofPIK3CAorFGFR3genes is denotedby redandgreenarrows, respectively. The stage (TaandT1) andgrade (high,H; low, L) for each tumor sample is also provided. A red (upregulated) to blue (downregulated) scheme following the scale limits (in log2 scale) is shown.Numbers on right side denote the number of transcripts characterizing each class. B, Kaplan–Meier distribution of patients in Lee dataset (26) accordingto the expression levels of the downregulated genes identified in our study. P values were obtained by the log-rank test. n, the number of samples scored ofeach group. C, relative relevance of the different histone-modifying enzymes in the genes of each quoted group obtained by ChEA. The relevance of eachenzyme is provided by the P value. The number on the right of each bar represents the number of specific genes bearing this type of modificationand the total number of transcripts on each group: upregulated (left) or downregulated (right) in recurrent tumors. D, dendrograms of mouse humantranscriptome comparison upon unsupervised clustering (Pearson correlation and average linkage method) according to the expression of genesdifferentiating mouse control and bladder tumor samples (top) or genes discriminating human recurrent tumors (bottom).

Santos et al.





concerted action of other E2F family members in thisprocess could not be discarded at this point.

EZH2 affects tumor progression in recurrenceBesides the high recurrence rate, an important problem in

NMIBC is the relatively high rate of recurrent tumors appearing

with increased stage and/or grade compared with the primarytumors, indicating that tumor progression upon recurrenceoccurred. During this study, 11 of 33 patients suffering recur-rence in our series also displayed tumor progression. To studywhether similar events to those described for recurrence werealso affecting tumor progression, the differential gene expres-sion between tumors showing or not tumor progression uponrecurrence was studied. This analysis revealed a very limited

Figure 6. E2F3a expression in humanNMIBC. A, expression ofEZH2 geneas a function of E2F3a gene expression as assessed by RT-qPCR inNMIBC samples. The r and P values are provided according theSpearman correlation method. B, expression of E2F3a gene in normaland tumor bladder samples (left) and in recurrent and nonrecurrentbladder tumors (right). P values were obtained by the Mann–Whitney ttest; mean and SEM are denoted in red. C, Kaplan–Meier distribution ofNMIBC recurrence according to the increased expression ofE2F3a gene.P value was obtained by the log-rank test. n, the number of samplesscored for each group.

Figure 5. EZH2 expression in human NMIBC. A, expression of EZH2 geneas assessed by RT-qPCR (with respect to TBP) in normal versus tumorsamples and in recurrent versus nonrecurrent tumors. P values wereobtained by the Mann–Whitney t test; mean and SEM are shown. B,representative examples of human tumors showing positive (left) andnegative (right) staining for EZH2 protein. Bar, 100 mm. C, Kaplan–Meierdistribution of recurrence according to the expression of EZH2 proteinform tissue microarrays. P values were obtained by the log-rank test(n, the number of samples in each group).






number of transcripts (Supplementary Table S17). ChEA anal-ysis of these transcripts showed that a significant number ofunderexpressed genes in tumors showing progression werebound by PRC2 complex elements (Fig. 7A). In addition, thesetranscripts displayed a significant overlap with external data-sets of gene lists regulated by PRC2, including genes withH3K27me3 marks (Fig. 7B). Accordingly, samples of tumorsshowing progression in recurrence showed also a significantincrease in EZH2 andE2F3a expression (Fig. 7C).Moreover, theincreased expression of EZH2 protein was an independentpredictor of tumor progression in the recurrences of our seriesof NMIBC samples (Fig. 7D).

These data indicated that increased EZH2 expression med-iates not only early recurrence but also increased probability oftumor progression in human NMIBC.

DiscussionBladder cancer is one of the most common cancers in men

worldwide making it a current problem in terms of social andmedical relevance. The genomic landscape of these tumorsmay contribute to identify novel targets of possible therapeuticinterest. These studies have revealed multiple alterations

and mutations in bladder cancer, including chromatin remo-deling, cell-cycle, and specific signal transduction pathways(5, 10, 29–33). Regarding cell cycle, RB1mutations are relevant.However, these studies are biased by the analysis of mainlyaggressive MIBC tumors and few examples of NMIBC areincluded. Nonetheless, even in these few examples, RB1 altera-tions are present (29, 32). This, together with other possiblealterations leading to functional pRb inactivation, such asCycD or E2F amplification (26, 30, 32), may suggest that Rb-dependent pathway is also of relevance in NMIBC. Our presentdata showing that the complete ablation of Rb family in vivoleads to urothelial tumors support this possibility.

It has been previously shown that the inactivation of Rb1gene in mouse urothelium is insufficient to allow spontaneoustumor development (13). Regarding our observations of com-plete penetrance of bladder cancer development upon Rbfamily ablation, this apparent discrepancy could be attributedto functional compensation by the other family members, aspreviously reported inmouse epithelia (19). Our findings are inagreement with the reported results of urothelial expression ofT antigen in transgenic mice (12, 34). However, T antigentransgenic mice usually develop muscle-invasive tumors; this

Figure 7. EZH2-dependent signaling in tumor recurrence and progression. A, relative relevance of the binding of the different PRC2 elements to thedownregulated genes in tumors showing progression upon recurrence as assessed by ChEA. The relevance of each enzyme is provided by the P value. B,summary of relevant overlap of the downregulated genes in tumors showing progression upon recurrence and Oncomine database in the concept "literaturedefined" showing the relevance of PRC2 components. C, expression of EZH2 and E2F3a genes as assessed by RT-qPCR (with respect to TBP) in tumorsdisplaying or not progression upon recurrence. P values were obtained by the Mann–Whitney t test; mean and SEM are denoted in red. D, Kaplan–Meierdistribution of tumor progression in recurrences according to increased expression of EZH2. P values were obtained by the log-rank test.

Santos et al.





could be attributed to impaired p53 tumor-suppressive func-tions. Such possibility, which is also reinforced by our previousfindings indicating that p53 loss in Rb-deficient stratifiedepithelia facilitates the development of invasive metastaticdisease (21, 35), remains to be determined. The pathologiccharacteristics of themouse bladder tumors suggest that thesemice could represent a bona fide NMIBC model. Also, themouse tumors display activation of MAPK/Erk, Stat3, and S6pathways, which have been previously involved in humanbladder cancer (36–38).In viewof thefindings in themousemodel, we aimed touse it

as a tool to gain a better understanding of the molecularmechanisms responsible for the initiation, recurrence, andprogression of bladder cancer. This would also provide poten-tial biomarkers. In a first approach, we performed a wholetranscriptome analysis. This reinforced the similaritiesbetween mouse and human tumors and provided a possiblemechanism for the upregulation and downregulation of genesin the mouse tumors involving E2F and PRC2. We found thatthe removal of Rb family led to overexpression of E2F familymembers, and these mouse tumors invariably showedincreased Ezh2 expression. Furthermore, our GSEA and ChEAdata indicated a predominant role for E2Fs, in particularE2F3a, in gene upregulation in mouse tumors. Given that (i)E2F3 is frequently amplified and overexpressed in humanbladder cancer (39–42), (ii) the inactivation of the Rb pathwayis required in addition to E2F3 overexpression in humanbladder carcinogenesis (40), (iii) E2F3 is a primary regulatorof EZH2 expression (28), and (iv) there is emerging evidencesupporting the RB–E2F3–EZH2 pathway as a key oncogenicaxis in cancer development and aggressiveness (43), our find-ings provide a new molecular mechanism for bladder tumordevelopment. Importantly, various EZH2 inhibitors have beendeveloped and are being tested in various preclinical models(44–46). The possibility that these inhibitors may impairbladder tumor growth in our Rb family–deficient model willbe the subject of future investigations. In addition, the iden-tification of possible overlapping mutations between humanNMIBC and our mouse model will provide further evidence ofthe relevance of this model for preclinical studies.To confirm these findings and to validate the mouse model

as a suitable tool for understanding human NMIBC, we carriedout a whole transcriptome analysis in human bladder cancersamples. Furthermore, as recurrence and progression arecommon in this type of human tumors and constitute a severeclinical problem, this studywas designed to determine possiblemolecular factors affecting recurrence in NMIBC. The meta-genomic analyses also reinforced the extreme similaritiesbetween mouse and human recurrent tumors. We alsoobserved that recurrence is coupled to gene upregulationmediated by E2F activity and downregulation associated withincreased PRC2 activity. Importantly, we could also validatethe role of PRC2- and EZH2-mediated gene silencing in NMIBCrecurrence in external datasets, indicating that increasedEzh2 expression and activity could act as a predictor of earlyrecurrence. Furthermore, we could also associate the increasedEZH2 expression and activity with progression upon recur-rence in human NMIBC. In this regard, the increased expres-

sion of EZH2 had been previously reported associated withincreased malignancy in bladder cancer (47, 48). In agreement,KDM6A (which catalyzes the H3K27 demethylation and thusacting in oppositemanner to EZH2) had also been described asfrequently mutated in bladder cancer, in some cases, in asso-ciation with RB1 gene mutation in high-stage or -grade humanbladder tumors (29, 30, 33). Future research, aimed to detectpossible overlapping mutations in these genes in our mousemodel, will provide further data of the relevance of this modelfor preclinical studies. Nonetheless, to the best of our knowl-edge, this is the first report associating EZH2 activity withNMIBC recurrence and progression.

Collectively, these findings could be highly relevant in theclinical management and therapy of human bladder cancers.The possible therapeutic use of EZH2 inhibitors in the man-agement of bladder cancer is a crucial aspect thatmerits futureresearch. In this regard, themousemodel here described arisesas an essential and invaluable tool for such analyses.

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

DisclaimerThe funders had no role in study design, data collection and analysis, decision

to publish, or preparation of the article.

Authors' ContributionsConception and design: D. Castellano, J.L. Rodriguez-Peralto, F. de la Rosa,J.M. ParamioDevelopment of methodology: M. Santos, M. Martínez-Fern�andez, B. Alfaya,C. Saiz-Ladera, M. Oteo, M.A. Morcillo, D. Castellano, J.L. Rodriguez-Peralto,J.M. ParamioAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): M. Santos, M. Martínez-Fern�andez, M. Due~nas,B. Alfaya, F. Villacampa, C. Costa, M. Oteo, J. Duarte, V. Martínez,M.J. G�omez-Rodriguez, M.L. Martín, M. Fern�andez, M.A. Morcillo, J. Sage,J.L. Rodriguez-PeraltoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): M. Santos, M. Martínez-Fern�andez, M. Due~nas,R. García-Escudero, M. Oteo, M.A. Morcillo, D. Castellano, J.M. ParamioWriting, review, and/or revision of the manuscript: M. Martínez-Fern�an-dez, M. Due~nas, R. García-Escudero, F. Villacampa, C. Costa, M. Oteo,M.J. G�omez-Rodriguez, P. Viatour, M.A. Morcillo, J. Sage, D. Castellano,J.L. Rodriguez-Peralto, J.M. ParamioAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): M.J. G�omez-Rodriguez, P. ViatourStudy supervision: F. Villacampa, D. Castellano, J.L. Rodriguez-Peralto, F. de laRosa, J.M. ParamioOther (discussed the results and comments on themanuscript):M. Santos

AcknowledgmentsThe authors want to particularly acknowledge the patients enrolled in this

study for their participation and the Biobanco iþ12 in theHospital 12 deOctubreintegrated in the Spanish Hospital Biobanks Network (RetBioH; www.redbio-bancos.es). They appreciate F.X. Real from the CNIO for his suggestions andcritical reading of the article.

Grant SupportThe study was funded by the following: MICINN grants SAF2012-34378

and SAF2011-26122-C02-01; Comunidad Aut�onoma de Madrid grants S2006/BIO-0232 and S2010/BMD-2470 (Oncocycle Programs); MSyC grants ISCIII-RETIC RD06/0020/0029 and RD12/0036/0009; and from Fundaci�on SandraIbarra to J.M. Paramio. Grants AP99782012 and 40100017 from MMA Foun-dation to M. Due~nas and M.L. Martín, respectively. MSyC grants ISCIII-FISPI12/01959 to M. Santos. M. Martínez-Fern�andez is funded by a 'Juan dela Cierva' research fellowship (JCI-2010-06167) from MICINN. Work done onRb family mutant mice in the Sage lab (J. Sage and P. Viatour) is fundedby the NIH (R01 CA114102). Biobank is supported by Instituto de SaludCarlos III.






The costs of publication of this article were defrayed in part 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 April 24, 2014; revised August 18, 2014; accepted August 21, 2014;published OnlineFirst September 24, 2014.

References1. Wu X, Hildebrandt MA, Chang DW. Genome-wide association studies

of bladder cancer risk: a field synopsis of progress and potentialapplications. Cancer Metastasis Rev 2009;28:269–80.

2. Gallagher DJ, Milowsky MI. Bladder cancer. Curr Treat Options Oncol2009;10:205–15.

3. Burger M, OosterlinckW, Konety B, Chang S, Gudjonsson S, Pruthi R,et al. ICUD-EAU International consultation on bladder cancer 2012:non-muscle-invasive urothelial carcinoma of the bladder. EuropeanUrol 2013;63:36–44.

4. Knowles MA. Molecular subtypes of bladder cancer: Jekyll and Hydeor chalk and cheese? Carcinogenesis 2006;27:361–73.

5. Iyer G, Al-Ahmadie H, Schultz N, Hanrahan AJ, Ostrovnaya I, Balar AV,et al. Prevalence and co-occurrence of actionable genomic alterationsin high-grade bladder cancer. J Clin Oncol 2013;31:3133–40.

6. Rebouissou S, Herault A, Letouze E, Neuzillet Y, Laplanche A, Ofua-luka K, et al. CDKN2A homozygous deletion is associated withmuscleinvasion in FGFR3-mutated urothelial bladder carcinoma. J Pathol2012;227:315–24.

7. van Rhijn BW. Combining molecular and pathologic data to prognos-ticate non-muscle-invasive bladder cancer. Urol Oncol 2012;30:518–23.

8. Lindgren D, Sjodahl G, Lauss M, Staaf J, Chebil G, Lovgren K, et al.Integrated genomic and gene expression profiling identifies twomajorgenomic circuits in urothelial carcinoma. PLoS One 2012;7:e38863.

9. Duenas M, Martinez-Fernandez M, Garcia-Escudero R, Villacampa F,Marques M, Saiz-Ladera C, et al. PIK3CA gene alterations in bladdercancer are frequent and associate with reduced recurrence in non-muscle invasive tumors. Mol Carcinog 2013 Dec 18 [Epub ahead ofprint], doi: 10.1002/mc.22125.

10. Cancer Genome Atlas Research Network. Comprehensive molecularcharacterization of urothelial bladder carcinoma. Nature 2014;507:315–22.

11. Ahmad I, Sansom OJ, Leung HY. Exploring molecular genetics ofbladder cancer: lessons learned from mouse models. Dis Mod Mech2012;5:323–32.

12. He F, Mo L, Zheng XY, Hu C, Lepor H, Lee EY, et al. Deficiency of pRbfamily proteins and p53 in invasive urothelial tumorigenesis. CancerRes 2009;69:9413–21.

13. Puzio-Kuter AM, Castillo-Martin M, Kinkade CW, Wang X, Shen TH,Matos T, et al. Inactivation of p53 and Pten promotes invasive bladdercancer. Genes Dev 2009;23:675–80.

14. Ruiz S, Santos M, Segrelles C, Leis H, Jorcano JL, Berns A, et al.Unique and overlapping functions of pRb and p107 in the control ofproliferation and differentiation in epidermis. Development 2004;131:2737–48.

15. Liang FX, Bosland MC, Huang H, Romih R, Baptiste S, Deng FM,et al. Cellular basis of urothelial squamous metaplasia: roles oflineage heterogeneity and cell replacement. J Cell Biol 2005;171:835–44.

16. Costa C, Santos M, Segrelles C, Duenas M, Lara MF, Agirre X, et al. Anovel tumor suppressor network in squamous malignancies. Sci Rep2012;2:828.

17. CostaC, SantosM,Martinez-FernandezM,DuenasM, Lorz C, Garcia-Escudero R, et al. E2F1 loss induces spontaneous tumour develop-ment in Rb-deficient epidermis. Oncogene 2013;32:2937–51.

18. Lara MF, Garcia-Escudero R, Ruiz S, Santos M, Moral M, Martinez-Cruz AB, et al. Gene profiling approaches help to define the specificfunctions of retinoblastoma family in epidermis. Mol Carcinog2008;47:209–21.

19. Costa C, Paramio JM, Santos M. Skin tumors Rb(eing) uncovered.Front Oncol 2013;3:307.

20. Viatour P, Somervaille TC, Venkatasubrahmanyam S, Kogan S,McLaughlin ME, Weissman IL, et al. Hematopoietic stem cell quies-

cence is maintained by compound contributions of the retinoblastomagene family. Cell Stem Cell 2008;3:416–28.

21. Bornachea O, Santos M, Martinez-Cruz AB, Garcia-Escudero R,Duenas M, Costa C, et al. EMT and induction of miR-21 mediatemetastasis development in Trp53-deficient tumours. Sci Rep2012;2:434.

22. Duenas M, Santos M, Aranda JF, Bielza C, Martinez-Cruz AB, Lorz C,et al. Mouse p53-deficient cancer models as platforms for obtaininggenomic predictors of human cancer clinical outcomes. PLoS One2012;7:e42494.

23. Ohl F, JungM, Radonic A, SachsM, Loening SA, Jung K. Identificationand validation of suitable endogenous reference genes for geneexpression studies of human bladder cancer. J Urol 2006;175:1915–20.

24. Lachmann A, Xu H, Krishnan J, Berger SI, Mazloom AR, Ma'ayan A.ChEA: transcription factor regulation inferred from integratinggenome-wide ChIP-X experiments. Bioinformatics (Oxford, England)2010;26:2438–44.

25. Rhodes DR, Yu J, Shanker K, Deshpande N, Varambally R, Ghosh D,et al. ONCOMINE: a cancer microarray database and integrated data-mining platform. Neoplasia 2004;6:1–6.

26. Lee JS, Leem SH, Lee SY, Kim SC, Park ES, Kim SB, et al. Expressionsignature of E2F1 and its associated genes predict superficial toinvasive progression of bladder tumors. J Clin Oncol 2010;28:2660–7.

27. Dyrskjot L, Thykjaer T, Kruhoffer M, Jensen JL, Marcussen N, Hamil-ton-Dutoit S, et al. Identifying distinct classes of bladder carcinomausing microarrays. Nat Genet 2003;33:90–6.

28. Bracken AP, Pasini D, Capra M, Prosperini E, Colli E, Helin K. EZH2 isdownstream of the pRB-E2F pathway, essential for proliferation andamplified in cancer. Embo J 2003;22:5323–35.

29. Balbas-Martinez C, Sagrera A, Carrillo-de-Santa-Pau E, Earl J, Mar-quez M, Vazquez M, et al. Recurrent inactivation of STAG2 in bladdercancer is not associated with aneuploidy. Nat Genet 2013;45:1464–9.

30. Ross JS, Wang K, Al-Rohil RN, Nazeer T, Sheehan CE, Otto GA, et al.Advanced urothelial carcinoma: next-generation sequencing revealsdiverse genomic alterations and targets of therapy. Mod Pathol2014;27:271–80.

31. Guo G, Sun X, Chen C, Wu S, Huang P, Li Z, et al. Whole-genome andwhole-exome sequencing of bladder cancer identifies frequent altera-tions in genes involved in sister chromatid cohesion and segregation.Nat Genet 2013;45:1459–63.

32. Cazier JB,RaoSR,McLeanCM,Walker AL,Wright BJ, Jaeger EE, et al.Whole-genome sequencing of bladder cancers reveals somaticCDKN1A mutations and clinicopathological associations with muta-tion burden. Nat Commun 2014;5:3756.

33. Gui Y, Guo G, Huang Y, Hu X, Tang A, Gao S, et al. Frequent mutationsof chromatin remodeling genes in transitional cell carcinoma of thebladder. Nat Genet 2011;43:875–8.

34. Garcia-Espana A, Salazar E, Sun TT, Wu XR, Pellicer A. Differentialexpression of cell cycle regulators in phenotypic variants of transge-nically induced bladder tumors: implications for tumor behavior. Can-cer Res 2005;65:1150–7.

35. Martinez-Cruz AB, Santos M, Lara MF, Segrelles C, Ruiz S, Moral M,et al. Spontaneous squamous cell carcinoma induced by the somaticinactivation of retinoblastoma and Trp53 tumor suppressors. CancerRes 2008;68:683–92.

36. Ho PL, Lay EJ, Jian W, Parra D, Chan KS. Stat3 activation in urothelialstem cells leads to direct progression to invasive bladder cancer.Cancer Res 2012;72:3135–42.

37. Jebar AH, Hurst CD, Tomlinson DC, Johnston C, Taylor CF,Knowles MA. FGFR3 and Ras gene mutations are mutually exclu-sive genetic events in urothelial cell carcinoma. Oncogene2005;24:5218–25.


Santos et al.




38. Korkolopoulou P, LevidouG, Trigka EA, Prekete N, KarlouM, ThymaraI, et al. A comprehensive immunohistochemical and molecularapproach to the PI3K/AKT/mTOR (phosphoinositide 3-kinase/v-aktmurine thymoma viral oncogene/mammalian target of rapamycin)pathway in bladder urothelial carcinoma. BJU Int 2012;110:E1237–48.

39. OlssonAY, Feber A, EdwardsS, TePoeleR,Giddings I,MersonS, et al.Role of E2F3 expression in modulating cellular proliferation rate inhuman bladder and prostate cancer cells. Oncogene 2007;26:1028–37.

40. Hurst CD, Tomlinson DC, Williams SV, Platt FM, Knowles MA. Inac-tivation of the Rb pathway and overexpression of both isoforms ofE2F3 are obligate events in bladder tumours with 6p22 amplification.Oncogene 2008;27:2716–27.

41. Oeggerli M, Tomovska S, Schraml P, Calvano-Forte D, Schafroth S,Simon R, et al. E2F3 amplification and overexpression is associatedwith invasive tumor growth and rapid tumor cell proliferation in urinarybladder cancer. Oncogene 2004;23:5616–23.

42. Feber A, Clark J, Goodwin G, Dodson AR, Smith PH, Fletcher A, et al.Amplification and overexpression of E2F3 in human bladder cancer.Oncogene 2004;23:1627–30.

43. Foster CS, Falconer A, Dodson AR, Norman AR, Dennis N, Fletcher A,et al. Transcription factor E2F3 overexpressed in prostate cancerindependently predicts clinical outcome. Oncogene 2004;23:5871–9.

44. TanJ,YangX,ZhuangL, JiangX,ChenW,LeePL, et al. Pharmacologicdisruption of Polycomb-repressive complex 2-mediated gene repres-sion selectively induces apoptosis in cancer cells. Genes Dev2007;21:1050–63.

45. McCabe MT, Ott HM, Ganji G, Korenchuk S, Thompson C, Van AllerGS, et al. EZH2 inhibition as a therapeutic strategy for lymphoma withEZH2-activating mutations. Nature 2012;492:108–12.

46. Knutson SK, Wigle TJ, Warholic NM, Sneeringer CJ, Allain CJ, KlausCR, et al. A selective inhibitor of EZH2 blocks H3K27 methylation andkills mutant lymphoma cells. Nat Chem Biol 2012;8:890–6.

47. Weikert S, Christoph F, Kollermann J, Muller M, Schrader M, Miller K,et al. Expression levels of the EZH2polycomb transcriptional repressorcorrelate with aggressiveness and invasive potential of bladder car-cinomas. Int J Mol Med 2005;16:349–53.

48. Wang H, Albadine R, Magheli A, Guzzo TJ, Ball MW, Hinz S, et al.Increased EZH2 protein expression is associatedwith invasive urothe-lial carcinoma of the bladder. Urol Oncol 2012;30:428–33.






2014;74:6565-6577. Published OnlineFirst September 24, 2014.Cancer Res Mirentxu Santos, Mónica Martínez-Fernández, Marta Dueñas, et al. Bladder Cancer

Ezh2 Signaling Loop Causes−E2F− Disruption of an RbIn Vivo

Updated version

10.1158/0008-5472.CAN-14-1218doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2014/09/23/0008-5472.CAN-14-1218.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/74/22/6565.full#ref-list-1

This article cites 47 articles, 10 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/74/22/6565.full#related-urls

This article has been cited by 4 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/74/22/6565To 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-14-1218

http://cancerres.aacrjournals.org/content/suppl/2014/09/23/0008-5472.CAN-14-1218.DC1

http://cancerres.aacrjournals.org/content/74/22/6565.full#ref-list-1

http://cancerres.aacrjournals.org/content/74/22/6565.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/74/22/6565


InVivo DisruptionofanRb E2F Ezh2SignalingLoopCauses ......Taq PCR master mix (Promega) and 1 mL of cDNA as a template.Melting curves were performed to verify speciﬁcity and absence

Documents