modulation of gene expression and cell-cycle signaling...

Research Article

Modulation of Gene Expression and Cell-Cycle SignalingPathways by the EGFR Inhibitor Gefitinib (Iressa) in RatUrinary Bladder Cancer

Yan Lu1,2, Pengyuan Liu1,2, Francoise Van den Bergh3, Victoria Zellmer3, Michael James3,Weidong Wen1, Clinton J. Grubbs4, Ronald A. Lubet5, and Ming You1,3

AbstractThe epidermal growth factor receptor inhibitor Iressa has shown strong preventive efficacy in theN-butyl-

N-(4-hydroxybutyl)-nitrosamine (OH-BBN) model of bladder cancer in the rat. To explore its antitumor

mechanism, we implemented a systems biology approach to characterize gene expression and signaling

pathways in rat urinary bladder cancers treated with Iressa. Eleven bladder tumors from control rats, seven

tumors from rats treated with Iressa, and seven normal bladder epithelia were profiled by the Affymetrix Rat

Exon 1.0 ST Arrays. We identified 713 downregulated and 641 upregulated genes in comparing bladder

tumors versus normal bladder epithelia. In addition, 178 genes were downregulated and 96 genes were

upregulated when comparing control tumors versus Iressa-treated tumors. Two coexpression modules that

were significantly correlated with tumor status and treatment status were identified [r ¼ 0.70, P ¼ 2.80 �10�15 (bladder tumor vs. normal bladder epithelium) and r¼ 0.63, P¼ 2.00� 10�42 (Iressa-treated tumor

vs. control tumor), respectively]. Both tumor module and treatment module were enriched for genes

involved in cell-cycle processes. Twenty-four and twenty-one highly connected hub genes likely to be key

drivers in cell cycle were identified in the tumor module and treatment module, respectively. Analysis of

microRNA genes on the array chips showed that tumor module and treatment module were significantly

associated with expression levels of let-7c (r ¼ 0.54, P ¼ 3.70 � 10�8 and r ¼ 0.73, P ¼ 1.50 � 10�65,

respectively). These results suggest that let-7c downregulation and its regulated cell-cycle pathway may play

an integral role in governing bladder tumor suppression or collaborative oncogenesis and that Iressa

exhibits its preventive efficacy on bladder tumorigenesis by upregulating let-7 and inhibiting the cell cycle.

Cell culture study confirmed that the increased expression of let-7c decreases Iressa-treated bladder tumor

cell growth. The identified hub genes may also serve as pharmacodynamic or efficacy biomarkers in clinical

trials of chemoprevention in human bladder cancer. Cancer Prev Res; 5(2); 248–59. �2011 AACR.

Introduction

Bladder cancer is one of the most common cancers in theUnited States and occurs predominantly in men. In 2010,there were an estimated 70,530 new cases of bladder cancer,with 52,760 cases in males (1). It is also the most expensive

form of cancer to treat. Approximately 15% of bladdertumors become invasive tumors after infiltration throughthe basement membrane. Patients with muscle invasivedisease are at high risk for recurrence, progression, andmetastasis.

Treatment of mice or rats with N-butyl-N-(4-hydroxybu-tyl)-nitrosamine (OH-BBN) results in transitional andsquamous cell urinary bladder cancers that bear significanthistopathologic similarities to human bladder cancer andtends to be muscle invasive (2–4). This rodent model hasbeen used extensively to characterize the tumorigenic pro-cess and to assess the efficacy of potential chemopreventiveagents to inhibit the development of carcinogen-inducedbladder cancers (4–6). OH-BBN–induced bladder tumorshave increased production of survivin and glutathione S-transferase pi (GST-Pi), and decreased production of fragilehistidine triad (FHIT) due to promoter methylation byimmunohistochemical analysis (7, 8). Similar molecularchanges are observed in human bladder tumors (9). Micro-array analysis of OH-BBN–induced bladder tumorsrevealed a wide variety of induced genes, including cyclin

Authors' Affiliations: 1Department of Surgery and The Alvin J. SitemanCancer Center, Washington University School of Medicine, Saint Louis,Missouri; Departments of 2Physiology and 3Pharmacology andToxicology,Cancer Center, Medical College of Wisconsin, Milwaukee, Wisconsin;4Department of Surgery, University of Alabama at Birmingham, Birming-ham, Alabama; and 5Division of Cancer Prevention, National CancerInstitute, Bethesda, Maryland

Note:Supplementary data for this article are available atCancer PreventionResearch Online (http://cancerprevres.aacrjournals.org/).

Corresponding Author: Ming You, Department of Pharmacology andToxicology and the Cancer Center, Medical College of Wisconsin,Milwaukee, WI 53226. Phone: 414-955-2565; Fax: 414-955-6058; E-mail:[email protected]

doi: 10.1158/1940-6207.CAPR-10-0363

�2011 American Association for Cancer Research.

CancerPreventionResearch

Cancer Prev Res; 5(2) February 2012248

Research. on May 22, 2018. © 2012 American Association for Cancercancerpreventionresearch.aacrjournals.org Downloaded from

Published OnlineFirst October 7, 2011; DOI: 10.1158/1940-6207.CAPR-10-0363

http://cancerpreventionresearch.aacrjournals.org/

D1, Pcna, Cox-2, the calcium-binding proteins S100a4,S100a8, S100a10, and Annexin1 (10). Tumors which formin the rat bladder tumor model are similar at the level ofgene expression tomuscle invasive bladder tumors found inhumans (11).Recently, we examined the preventive effects of the

epidermal growth factor receptor (EGFR) inhibitor Iressa onthe inductionofbladder tumorsbyOH-BBN(12). Treatmentwith Iressa (10mg/kgbodyweight per day), beginning either2 weeks or 12 weeks after the last dose of OH-BBN (whenmicroinvasive carcinomas exist), decreased the incidence oflarge palpable bladder cancer bymore than90% at 6monthsafter the last dose of OH-BBN. To explore the mechanism ofthe antitumor action of Iressa in the treatment of bladdertumors, we implemented a systems biology approach tocharacterize gene expression and pathway modulation in ratbladder tumors treated with Iressa.

Materials and Methods

Rat bladder tumorsFemale Fischer-344 rats were obtained from Harlan Spra-

gue-Dawley, Inc. (virus-free colony202) at28daysof age andwere housed in polycarbonate cages (5 per cage). The ani-mals were kept in a lighted room 12 hours each day andmaintained at 22�C � 0.5�C. Teklad 4% mash diet (HarlanTeklad) and tap water were provided ad libitum. The carcin-ogen, OH-BBN, (TCI America; 150 mg/gavage, 2�/wk) wasstartedwhen the rats were 49 days of age and continued for 8weeks. The carcinogen vehicle was ethanol:water (20:80) in0.5 mL. All animals (unless sacrificed early because of pal-pable tumor burden greater than 2 cm, weight loss greaterthan 20% of initial body weight, or because animals becamemoribund), were sacrificed 8 months following the initialOH-BBN treatment. Starting6months after the initial doseofOH-BBN, animals were palpated twice per week for thedevelopment of lesions. when an animal had developed asmall palpable tumor (150–300 mg), rats were treated dailywith Iressa at a dose level of 10 mg/kg body weight andsacrificed 5 days later. At the time of sacrifice, the urinarybladder from each rat was tied off, and samples of controlbladder tumors, normal bladder epithelia, and bladdertumors fromrats treatedwith Iressawere excised, snap-frozenin liquid nitrogen, and stored at�80�C for subsequent RNAisolation and protein extraction. Additional bladder tissuefrom the same rat was fixed with 10%neutral formalin. Afterfixation, the bladders were examined under a high-intensitylight dissecting microscope for lesions. Each lesion wasseparately embedded in a paraffin block, and random trans-verse sections (5 mm) from 2 different levels were cut andstained with hematoxylin and eosin for histopathology. Allbladder tumors used in this study were diagnosed as bladdercancers with a mixed histology showing elements of bothtransitional and squamous cells. Normal bladder epitheliacame from age-matched controls. To isolate bladder epithe-lia, we separated epithelial cells from the stroma andmuscletissues by cutting the bladder into halves and scraping offthe epithelium with a surgical blade and forceps.

RNA isolation and amplificationTotal RNA from normal bladder epithelia, bladder

tumors, and bladder tumors treated with Iressa were iso-lated by TRIzol (Invitrogen) and purified by the RNeasyMini Kit and RNase-free DNase Set (Qiagen) according tothe manufacturer’s protocols. In vitro transcription–basedRNA amplification was carried out on each sample. cDNAfor each sample was synthesized by a Superscript cDNASynthesis Kit (Invitrogen) and a T7-(dT)24 primer as fol-lows: 50-GGCCAGTGAATTGTAATACGACT-CACTATAGG-GAGGCGG-(dT)24-30. The cDNA was purified by phase-lock gel (Fisher Cat ID E0032005101) phenol/chloroformextraction, and then the biotin-labeled cRNA was tran-scribed in vitro from cDNA by a BioArray High Yield RNATranscript Labeling Kit (ENZO Biochem) and further puri-fied by the RNeasy Mini Kit.

Affymetrix rat exon arraysThe labeled cRNAwas applied to the Affymetrix Rat Exon

1.0 ST Array (Affymetrix) according to the manufacturer’srecommendations at the Vanderbilt Gene Microarray Core.Routine quality assessment was conducted on each exonarray. The study used 10 arrays from bladder tumors, 7arrays from normal bladder epithelia, and 7 arrays fromIressa-treated bladder tumors that had passed quality con-trol criterion, as recommended by the manufactures. Genelevel and individual exon signal estimates for the CELfiles from the platform Rat Exon 1.0 ST Array were derivedby the Robust Multi-array Average (RMA) algorithm, asimplementedwith ExpressionConsole v1.1.1 (http://www.affymetrix.com/products_services/software/specific/expression_console_software.affx).

Identifying differentially expressed genesTwo-sample student t test was used to identify differen-

tially expressed genes (DEG) between 2 groups. To adjustformultiple testing in the study of high-dimensionalmicro-array data, the local false discovery rate (LFDR) was esti-mated (13), which was implemented in the R packagefdrtool (http://www.r-project.org/). TheDEGswere definedas genes with LFDR less than 0.05 and fold change morethan 1.5 between 2 groups.

Gene set enrichment analysisGene set enrichment analysis (GSEA) was conducted to

analyze the pattern of differential gene expression. GSEA is acomputational method that determines whether a set ofgenes shows statistically significant differences in expres-sion between 2 biological states and has proved successfulin discovering molecular pathways involved in humandiseases (http://www.broad.mit.edu/gsea). Using the Kol-mogorov–Smirnov statistic, GSEA assesses the degree of"enrichment" of a set of genes (e.g., a pathway) in the entirerange of the strength of associations with the phenotype ofinterest. GSEA was used to identify a priori defined sets ofgenes that were differentially expressed (14, 15). We usedcurated gene sets (c2) which contain genes in certainmolecular pathways and gene ontology (GO) gene sets

Effects of Iressa on Rat Urinary Bladder Cancers

www.aacrjournals.org Cancer Prev Res; 5(2) February 2012 249




(c5)which consist of genes annotatedby the sameGOtermsin theMolecular Signature Database (MSigDB; http://www.broad.mit.edu/gsea/msigdb/msigdb_index.html). Becauseof a small sample size, the GSEA with the gene set permu-tation option was conducted. Selected gene sets identifiedfrom GSEA were then visualized with MetaCore software(http://www.genego.com/).

Hierarchical clustering analysisHierarchical clustering was carried out as follows. For the

selected genes, expression indexes were transformed acrosssamples to anN (0, 1) distributionwith a standard statisticalZ-Transform. These values were input to the GeneClusterprogram of Eisen and colleagues (16), and genes wereclustered using average linkage and correlation dissimilarity.

Coexpression network analysisTo characterize gene expression and pathways in rat

bladder tumors treated with Iressa, we applied a systemsbiology approach using a weighted gene coexpression net-work analysis (WGCNA; ref. 17). The WGCNA convertsgene coexpression measures into connection weights astopology overlap measure (TOM). We chose expressionprofiles of 4,500genes in the coexpressionnetwork analysis.These geneswere either significantly differentially expressedbetween tumors with and without Iressa treatment (LFDR <0.05 and fold change > 1.5 between 2 groups) or showed alarge variability in expression. In weighted gene coexpres-sion networks, modules correspond to clusters of highlycorrelated genes.We definedmodules using a staticmethodby hierarchically clustering the genes using 1-TOM as thedistance measure with a height cutoff value of 0.95 and aminimum size (gene number) cutoff value of 40 for theresulting dendrogram.

To identify whichmodule is correlated with treatment, wefirst calculated the module eigengene (i.e., first principalcomponent of the expression values across subjects) usingall genes in each module. We then correlated the moduleeigengenes to treatment using the SpearmanCorrelation.Wedetermined intramodular connectivity for each gene by sum-mingtheconnectivityofthatgenewitheveryothergeneinthatmodule. All network analyses were implemented in the sta-tistical packageWGCNA in R environment (18). The VisANTprogram was used to construct gene–gene interaction (con-nections) networks (19). Genes within modules were ana-lyzed for GO term enrichment by the program DAVID (20).

Exon splicing analysisA number of splicing-related mutations have been

reported to be associated with malignancies, and some ofthese are within splice sites or splicing enhancers/silencersof cancer-related genes (21–23). To identify exon splicing,we first calculated the splicing index for each exon. Thesplicing index gives ameasureof thedifference in expressionlevel for each probe set in a gene between 2 sets of arrays,relative to the gene level average in each set. This is calcu-lated only for those probe sets that are defined as exontargeting and nonmultitargeted. Then, MIDAS (Microarray

Detection of Alternative Splicing) algorithm was used forthe detection of alternative splicing events. MIDAS P valuewas calculated by the "splanova" function. For genes con-taining at least one significant exon targeting probe set, wealso calculated the maximum splicing index and retrievedthe average fold change for each gene. The average foldchanges were used to partition the data set into (on average)up- and downregulated genes. All exon splicing analyseswere implemented in the statistical package exonmap(http://bioinformatics.picr.man.ac.uk).

Cell culture and treatmentThe Iressa sensitive UMUC-5 cell line (human squamous

cell carcinoma of the bladder; Sigma) were cultured inEagle’sMinimal EssentialMedium [EMEM; Earle’s balancedsalt solution (EBSS)] supplemented with 2 mmol/L gluta-mine, 1%nonessential amino acids (NEAA), 1%penicillin/streptomycin, and 10% FBS (Invitrogen). The Iressa (SantaCruz Biotechnology) was reconstituted in dimethyl sulfox-ide (DMSO) at a stock concentration of 70 mmol/L andstored at �80�C. This stock was diluted in medium justbefore use so that the concentration of DMSO neverexceeded 0.1%. For quantitative PCR (qPCR), the cells weregrown in 10-cm dishes and allowed to reach 50% to 70%confluency; at that point the cells were treated with theindicated concentration of Iressa for 24 hours. Followingthe incubation, the cells were collected by trypsinization,washed in cold PBS, counted, and frozen as cell pellets. ForMTS assay (CellTiter 96 AQueous One Solution Cell Pro-liferation Assay; Promega), 10,000 cells perwell were platedin 24-well plates. Twenty-four hours later, the medium waschanged, and the indicated amount of Iressawas added. TheMTS assay was conducted following manufacturer’s proto-col on that day (as day 0) and the following days.

miRNA qPCRThe protocol for miRNA isolation by the miRNeasy Mini

kit (Qiagen) was followed, including following Appendix Awhich deals specifically with the preparation of a miRNA-enriched fraction by the Qiagen RNeasy MinElute CleanupKit. The protocol for RT2 miRNA first strand kit was fol-lowed to obtain the cDNA template for the qPCR reaction.The SsoFast EvaGreen Supermix (Bio-Rad) protocol wasfollowed for the qPCR reactions-–carried out in triplicate, ano-DNA reaction and a primer set corresponding to ahousekeeping miRNA was included. The cycling programincluded a melt curve. The primer sets for let-7c(MPH0003A-200) and RNU6-2 (MPH01653A-200) werepurchased from Qiagen.

Results

Differentially expressed bladder cancer genes alteredwith Iressa treatment

We observed 713 downregulated and 641 upregulatedgenes in OH-BBN–induced rat bladder tumors (LFDR <0.05 and fold change > 1.5 between 2 groups). Down-regulated genes in bladder tumors are enriched in several

Lu et al.

Cancer Prev Res; 5(2) February 2012 Cancer Prevention Research250




metabolic processes such as lipid metabolic processes andcellular ketone metabolic processes (Fig. 1A and Supple-mentary Table S1) and upregulated genes are enriched incell proliferation and migration processes (Fig. 1B andSupplementary Table S2). Similarly, we compared geneexpression in control bladder tumors and Iressa-treatedtumors and identified 178 upregulated genes and 96 down-regulated genes in Iressa-treated versus control tumors.Most of the Iressa upregulated genes are involved in met-abolic processes (Fig. 1C and Supplementary Table S3). The3 most significant GO terms found in the downregulatedgenes after Iressa treatment were: cell-cycle phase, cell-cycleprocesses, and M phase (Fig. 1D and Supplementary TableS3).DEGswere input into hierarchical clustering analysis ofall the samples. Three distinct groups were identified incomplete concordance with the sample status as normalbladder epithelia, bladder tumor, and bladder tumor trea-ted with Iressa (Fig. 2A), implying that these DEGs capturebiological differences of the 3 tissue groups. In addition,Iressa-treated tumors clustered more closely to normalbladder epithelia than bladder tumors without Iressa treat-ment. Thus, short-term Iressa treatment reversed many ofthe gene changes associated with bladder neoplastic trans-formation and progression such asMki67,Bub1,Cdc20, andTop2a. This is further supported by hierarchical clustering

of cell-cycle–related DEGs which revealed that Iressa-trea-ted tumorsweremore similar to normal bladder epitheliumthan control tumors (Fig. 2B).

Gene coexpression networks and biological pathwaysSets of genes with expression levels that are highly cor-

related may share common biological process and regula-tory mechanisms. To examine this possibility, we analyzedglobal expression profiles and their interactions in the studysamples. To accomplish this, we constructed weighted genecoexpression networks based on pairwise Pearson correla-tions between the expression profiles. As indicated previ-ously, the identified DEGs captured biological significanceof the 3 tissue groups and thus were used to construct thenetwork. The analysis included an additional 2,872 geneswith greatest variability which determined by their coeffi-cient of variance from 13,855 remaining genes. We con-ducted coexpression network analyses in data sets ofcontrol tumors versus normal bladder epithelia and controlbladder tumors versus bladder tumors treated with Iressa,separately.

The WGCNA analysis identified a number of moduleswith highly coexpressed genes in both data sets (Fig. 3).These modules are significantly enriched for biologicallyimportant processes that are relevant to cancer, including

A B

C D

Lipid metabolic process Organ development

Regulation of cell proliferation

Response to wounding

System development

Cell migration

Positive regulation of biological process

Cell motion

Response to organic substance

Anatomical structure morphogenesis

Regulation of multicellular organismal...

M phase

Cell-cycle process

Cell-cycle phase

M phase of mitotic cell cycle

Mitosis

Nuclear division

Organelle fission

Mitotic cell cycle

Maintenance of organelle location

Regeneration

Cellular ketone metabolic process

Cellular ketone metabolic process

Cellular lipid metabolic process

Cofactor metabolic process

Lipid catabolic process



Regulation of cellular process

Regulation of transport

Vesicle-mediated transport

0

0 0 2 4 6 8 10 12 14 165 10 15 20 25

0 20 40 60 80 10012014016050 100 150 200 250

Organic acid metabolic process

Organic acid metabolic process

Oxoacid metabolic process

Oxoacid metabolic process

Lipid metabolic process

Response to xenobiotic stimulus

Xenobiotic metabolic process

Carboxylic acid metabolic process

Carboxylic acid metabolic process

Figure 1. GO of DEGs. A, genes downregulated in bladder tumors as compared with their normal bladder epithelia. B, genes upregulated in bladder tumors ascompared with their normal bladder epithelia. C, genes downregulated in bladder tumors after Iressa treatment. D, genes downregulated in bladder tumorsafter Iressa treatment. The top 10 GO terms are present in each category of DEGs. Abscissa ¼ numbers of DEGs.






those for the cell cycle, immune response, lymphocyte andT-cell activation, response to DNA damage, and RNA bind-ing and processing. We then examined whether individualexpressionprofileswithin eachof the identifiedmodules areassociated with clinically related traits by estimating themodule eigengene. In the data sets of control tumors versusnormal samples, we found significant correlation of tissuestatus (i.e., cancer vs. normal) with the salmon-coloredmodule (r¼ 0.70, P¼ 2.80� 10�15; Fig. 4A). Interestingly,

the salmon-colored module also showed a significant cor-relationwith expression levels of let-7cwhich is amicroRNAgene (r ¼ 0.54, P ¼ 3.70 � 10�8; Fig. 4B). In the data set ofcontrol bladder tumors versus bladder tumors treated withIressa samples, we found that the green-colored module isassociated with both tumor treatment status (i.e., tumorswith Iressa treatment vs. tumors without Iressa treatment;r ¼ 0.63, P ¼ 2.00 � 10�42; Fig. 4C) and let-7c expression(r¼ 0.73, P¼ 1.50� 10�65; Fig. 4D). Both the salmon- and

A B

TroapFoxmlKifc1RGD1564263PbkSpc25StilRGD1304693Uhrf1Mcm5PoleKntc1LOC683179Dt1Asf1bMybl2Mcm10Spag5SgoI1Kif4Kif15RGD1310335RGD1310376Plk1Tacc3Ncapd2Casc5Mast1Kif22MelkTtkCep55Sqol2RGD1562646Top2aCcna2Mk167Tpx2Ect2Nuf2Bub1Racgap1Cdc20Kif2cCcnb1Bub1bCitCdca2Fancd2AspmDepdc1Kif20aPrc1Iqgap3CenpIFam83dEspl1CenptCenpeArhgap11aKif20bAurkbAF303035Bard1AurkaIncenpCdca3Exo1TimelessPold1Pola2ENSRNOT00000045512Chaf1aCdca7Cdt1Ccne2Ccnel7164850ClspnOrcllCar9Cdca8Cdkn3E2f7Cdc25bLtb4rCalm3Spc24Ak3l1Eme1Kifl8aMcph1Gen1Serpind1Trpv3Trpm5Tgm1Pla2g4e7341030Saa4Serpina3kMatlaGuloSerpinalMug1HpdPlgItih4HpItih3FgbFgaPzpCyp2e1GcFggCyp2c22ApobApohCps1

Figure 2. Gene expression patternsamong the studied bladdersamples. Hierarchical clusteringanalysis was conducted using (A) allof the DEGs and (B) cell-cycle–related DEGs.

Lu et al.





green-colored modules are significantly enriched for cell-cycle processes (P ¼ 5.97 � 10�9 and P ¼ 1.90 � 10�16,respectively).Highly connected "hub" genes are hypothesized to play

an important role in organizing the behavior of biologicalmodules (24). To identify hub genes in the salmon- andgreen-colored modules, we estimated intramodular con-nectivity for each gene based on its Pearson correlationwith all of the other genes in the modules. The genes withhigh intramodular connectivity tended to have stronger

correlation with clinically related traits (Fig. 4). We set theweighted cutoff value of 0.40 to identify hub genes withthe strongest connections to other genes. As a result, weidentified 24 hub genes with at least 40 connections inthe data sets of control tumors versus normal samplesincluding Cdca2, Lmnb1, Sgol2, Tpx2, Anln, Nuf2, Kif4,Ect2, Bub1, Prc1, Kif20a, Cit, Uhrf1, Depdc1, Racgap1,RGD1310335, Casc5, Dtl, Ccna2, Bub1b, Tacc3, Iqgap3,Fancd2, and Cenpf; whereas 20 genes were identified inthe data set of control bladder tumors versus bladder

Figure 3. Gene coexpressionnetwork analyses. Unsupervisedhierarchical clustering was used todetect modules of highlycoexpressed genes in the 2 datasets: (A) control tumors versusnormal epithelia and (B) controltumors versus Iressa-treatedtumors. Each major branch in thefigure represents a color-codedmodule that contains a group ofhighly connected genes. Thearrows in (A) and (B) representmodules enriched for cell-cycleprocess.

1.0

0.9

0.8

0.7

0.6

0.5

Heig

ht

Dyn

am

ic t

ree

cu

t

A

B

1.0

0.9

0.8

0.7

0.6

0.5

0.4

0.3

Heig

ht

Dynam

ic tre

e c

ut






tumors treated with Iressa samples (Fancd2, Cep55,LOC683179, Kif2c, Ttk, Cenpf, Bub1, Ect2, Kif20a, Nuf2,Tpx2, Cdca2, Ccnb1, Depdc1, Ncapd2, Melk, RGD1562646,Mastl, Ccna2, and Sgol2). To illustrate their relationship,these genes are visualized in Fig. 5, suggesting a complexregulatory gene network with varying topology.

To explore the effects of let-7c expression on Iressa-treated bladder tumor growth, we checked the expressionof let-7c in an Iressa-sensitive human squamous cellcarcinoma of the bladder cell line UMUC-5 in the pres-ence of increasing amount of Iressa. We first confirmedthat the UMUC-5 cell line was Iressa sensitive followingtheir growth curve in 0.1, 0.2, and 0.5 mmol/L Iressacompared with no Iressa added. As previously described(25) and as seen in Fig. 6A, Iressa affected the growth ofUMUC-5 at concentration higher than 0.1 mmol/L. Thisexperiment was done in triplicate and was carried outtwice with similar results. We then collected themiRNA from UMUC-5 exposed for 24 hours to 0.2 and0.5 mmol/L of Iressa in parallel with a no treatmentcondition. These 3 samples of miRNA were used in qPCRreactions using either let-7c or RNU6-2 primer set. Theresults from the RNU6-2 reactions were used to normalize

let-7c expression. As observed in Fig. 6B, Iressa induced3- and 8-fold upregulation of let-7c when added at 0.2and 0.5 mmol/L, respectively. This qPCR was carried outtwice with similar results. These results confirmed theassociation of let-7c expression with Iressa treatmentobserved in our microarray analysis of rat model.

Exon splicingAlternative processing of pre-mRNA transcripts is amajor

source of protein diversity in eukaryotes and has beenimplicated in several disease processes including cancer(26). To identify alternative splicing variants, we calculatedthe splicing index (SI) for each exon; a measure of howmuch exon-specific expression, with gene induction fac-tored out, differs between 2 types of tissues (e.g., bladdertumors vs. normal bladder epithelia or bladder tumors withtreatment vs. without Iressa treatment). Using the cutoffvalue of maximum SI more than 1 and P value less than 1.0� 10�4, we identified 194 and 272 genes that have alter-native splicing that are down- and upregulated in bladdertumors versus normal epithelia, respectively. Supplemen-tary Figure S1 listed the top genes alternatively splicedbetween normal bladder epithelia and bladder tumors.

1.00.80.60.40.2

0.8

0.6

0.4

0.2

0.0

Module membership vs. gene significance cor = 0.73, P = 1.5e-65

Module membership in green module

Gene s

ignific

ance f

or

let-

7c

1.00.80.60.40.2

0.8

0.6

0.4

0.2

0.0

Module membership vs. gene significance cor = 0.63, P = 2e-42

Module membership in green module

Gene s

ignific

ance f

or

Iressa t

reatm

ent

1.00.90.80.70.6

0.8

0.6

0.4

0.2


Module membership in salmon module

Ge

ne

sig

nific

an

ce

fo

r b

lad

de

r tu

mo

r

1.00.90.80.70.60

.80

.70

.60

.50

.40

.30

.2


Module membership in salmon module

Ge

ne

sig

nific

an

ce

fo

r le

t.7

c

A B

C D

Figure 4. Associations of hub geneswith clinical traits. Scatter plotswere drawn for the scaledintramodular connectivity K (x-axis)against gene significance (y-axis).Salmon module from the data setsof control tumors versus normalbladder epithelia samples andgreen module from the data sets ofcontrol bladder tumors versusbladder tumors treated with Iressasamples were analyzed. Each pointcorresponds to a gene in themodules. The gene significance ofthe ith gene was defined as theabsolute correlation between the ithgene expression profile and thestudied trait. A, correlation (corr) ofsalmon module with cancer status(tumor or normal). B, correlation ofsalmon module with let-7cexpression. C, correlation of greenmodulewith Iressa treatment status(treated or not treated). D,correlation of greenmodulewith let-7c expression.

Lu et al.





These alternatively spliced genes are enriched in organmorphogenesis, regulation of cell proliferation, cell migra-tion, and immune systemprocesses.Using the same criteria,we also identified57genes that have alternative splicing thatare differentially expressed in control tumors and Iressa-treated tumors (Supplementary Table S4). Interestingly, 24of the 57 genes overlapped with the list of genes that havealternative splicing in bladder tumors versus normal blad-der epithelia, suggesting that Iressa treatment blocks aber-

rant splicing in these 24 genes in bladder tumors. Forexample, we identified exon splicing in Prkn2 when com-paring exon-specific expression between bladder tumorsand normal bladder tissues (Supplementary Fig. S2). Thesame exon splicing was observed between control tumorsand treated tumors. The rationale behind these 2 observa-tions is that Iressa treatment blocks aberrant splicing, andthus treated tumors have similar splicing status to normalsamples.

Figure 5. Gene interactions withinthe 2 significant modules. A,salmon module from the data setsof control tumors versus normalbladder samples. B, green modulefrom the data sets of controlbladder tumors versus bladdertumors treatedwith Iressa samples.Each node represents a gene andred nodes are hub genes; biggernodes indicate more connections.






Discussion

Wehave recently shown that EGFR inhibitors were highlyeffective in preventingOH-BBN–induced rat bladder cancer(12). Efficacywasobserved eitherwhen Iressa treatmentwasinitiated early in tumor progression (early hyperplasia) orwhen invasive microcarcinomas already existed (12). Thisindicates that Iressa is effective during the later stages oftumor progression. The traditional concept of preventionwould test the agents before cancer development. In ourstudy the animals in the Iressa treatment group were givenIressa after they developed a small palpable tumor (150–300 mg). The reasons of such design are as follows: (i)Human phase III clinical trials in prevention are relativelylate: The prime example of a "prevention trial" in bladder isto inhibit the recurrence of bladder cancer by the cancerchemopreventive retinoidN-(4-hydroxyphenyl)retinamide(4-HPR) within a period of 18 months. This is not an earlyintervention similar to the study with OH-BBN proposed.

Recently,wehave argued strongly that a later intervention inanimal models seems more relevant to apparent phase IIIclinical trials particularly given that the animal models aretypically not as genetically as advanced as human cancers(27). We showed that Iressa is highly effective in this modelwhen administered at the time that microcarcinomasalready exist implying that most of the effects of the agentare in the later stages of progression (12). (ii) Examining forbiomarkers in palpable lesions: The argument for examin-ing for biomarkers in early stage palpable lesions comesfrom the results observed in breast cancer. Results inhumans have shown clearly the 2 known positive "preven-tive" agents selective estrogen receptor modulators (SERM)examined in large prevention trials and aromatase inhibi-tors in prevention trials, but clear data from inhibition oftumors of the contralateral breast in an adjuvant settingyield major changes in biomarkers in a presurgical settingwith early stage cancer (28).Wehave used such an approachto examine a wide variety of effective and ineffective agentsin a breast cancer model and found a close correlationbetween alterations of biomarkers in this presurgical settingand efficacy in long-term prevention studies particularlywhen initiated further along in tumor progression (29).This is not an argument that there are agentswhichmight beeffective early in the process of tumor progression but areineffective at later times and will not give a strong signal inthis later intervention. (iii) Practical questions: For many ofthe earlier stages of tumor progression one either getslimited numbers of lesions or the lesions are fairly smalland can only be identified when tissue is fixed. Thesepractical problems virtually preclude the overall screeningby arrays or proteomics of changes in earlier lesions. Theobjective of part of this study is to define potential biomar-kers which we might examine in earlier lesions either byimmunohistochemistry or real-time PCR. It is not a thera-peutic study per se but rather a search for potentially usefulpharmacodynamic and perhaps efficacy biomarkers.

The efficacy of Iressa agrees with findings that genes andproteins associated with the EGFR pathway are overex-pressed in bladder tumors as compared with normal blad-der epithelia in both rodent model and humans (10, 30).Initial studies showed great promise for the use of EGFRinhibitors in the setting of advanced bladder cancer inhumans (31, 32). As expected, Iressa treatment strikinglydecreased levels of phosphorylated EGFR, AKT, and extra-cellular signal–regulated kinase (ERK; data not shown.).The present study examined the effects of Iressa on geneexpression in lesions with 2 objectives as follows: to (1)identify processes and pathways involved in themechanismof Iressa efficacy and (2) identify biomarkers which mightbe useful in clinical trials with this class of antitumor agents.

Studies suggest that highly centralized "hub" genes in thenetwork architecture are more likely to be key drivers forcellular function (33). Characterization of these hub genesmay help to give novel insight into bladder tumorigenesisand to develop additional novel drug targets for bladdercancer. Many of the hub genes we identified in the tumorand treatment modules are cell-cycle regulators such as

Time in d

Ce

ll p

rolif

era

tio

n b

y M

TS

assa

y (

OD

490 n

m)

5432100.0

0.2

0.4

0.6

0.8

1.0

no add 0.1 µmol/L0.2 µmol/L0.5 µmol/L

Iressa

0.50.200

1

2

3

4

5

6

7

8

9

Norm

aliz

ed let-

7c e

xpre

ssio

n

Iressa in µmol/L

A

B

Figure 6. The increased expression of let-7c decreases Iressa-treatedbladder tumor cell growth. A, MTT assay for the effect of increasingamount of Iressa on the growth of bladder cancer cells UMUC-5. B,quantitative real-time PCR to detect let-7c expression in UMUC-5 cellstreated with different amount of Iressa.

Lu et al.





Bub1, Casc5, Ccna2, Cenpf, and so on. Bub1, which wasidentified as a highly connected gene in both modules,encodes a kinase involved in spindle checkpoint function.This kinase functions in part by phosphorylating amemberof the mitotic checkpoint complex and activating the spin-dle checkpoint. p53 physically interacts with Bub1 at kine-tochores in response to mitotic spindle damage, suggestinga direct role for hBub1 in the suppression of p53-mediatedcell death (34). Altered expression of Bub1 is associatedwith therapy failure and death in patients with multipletypes of cancer (35). Casc5, identified as another hub genein both modules, is required for chromosome alignmentand the mitotic checkpoint through direct interaction withBub1 and BubR1 (36). Ccna2 binds and activates CDC2 orcyclin-dependent kinase 2 (CDK2) kinases and thus pro-motes both cell cycle G1–S and G2–M transitions (37, 38).Cdca2 is a cell-cycle regulator and selectively recruits PP1gonto mitotic chromatin at anaphase and into the subse-quent interphase. Depletion of Cdca2 by RNAi knockdownleads to large-scale cell death by apoptosis, primarily ininterphase; as did overexpression of this dominant-negativemutant (39). Cenpf is another cell-cycle regulator. Farne-sylation of Cenpf is required for G2–M progression anddegradation after mitosis (40).Several unique hub genes were identified in the module

associated with efficacy of Iressa treatment, includingCcnb1, Cep55, Kif2c, LOC683179,Mastl,Melk, Ncapd2, andRGD1562646. Cep55 encodes a centrosomal associatedprotein involved in centrosome duplication, cell-cycle pro-gression, and in the regulation of cytokinesis (41). Kif2cencodes a kinesin-like protein which has a role in bipolarspindle assembly (42). Melk is a potential regulator ofG2–M progression and may antagonize the CDC25B phos-phatase (43). Ncapd2 is an essential component of thehuman condensin complex required for mitotic chromo-some condensation (44). These hub genes might be usefulbiomarkers in clinical trials with this class of agents.Previous studies showed that several hub genes are asso-

ciated with human bladder cancer. For example, the expres-sion levels of NUF2 were increased in the majority of smallcell lung cancer, cholangiocellular cancer, urinary bladdercancer, and renal cell cancers (45). The MPHOSPH1/PRC1complex is likely to play a crucial role in bladder carcino-genesis and that inhibition of the MPHOSPH1/PRC1expression or their interaction should be novel therapeutictargets for bladder cancers (46). KIF20Awere upregulated inbladder tumors in both humans and rodents (47). Animmunohistochemistry-based UHRF1 detection in urinesediment or surgical specimens can be a sensitive andcancer-specific diagnostic and/or prognosis method andmay greatly improve the current diagnosis based on cytol-ogy (48). Immunocytochemical staining analysis detectedstrong staining of endogenous DEPDC1 protein in thenucleus of bladder cancer cells, and DEPDC1 expressionwas hardly detectable in any of 24 normal human tissuesexcept the testis. Suppression of DEPDC1 expression withsiRNA significantly inhibited growth of bladder cancer cells.DEPDC1 might play an essential role in the growth of

bladder cancer cells and would be a promising moleculartarget for novel therapeutic drugs or cancer peptide vaccineto bladder cancers (49). RacGAP1 are expressed at highlevels in superficial samples of bladder from patients withtransitional cell carcinoma (TCC; ref. 50).

We identified 2 modules that are significantly associatedwith either cancer status (control cancer vs. normal bladderepithelium)or Iressa treatment status (Iressa-treated tumorsvs. control tumors). An increase in proliferation-relatedgenes was observed in the cancer versus normal bladderepithelium, whereas a decrease in proliferation-relatedgenes was observed in Iressa-treated tumors versus controltumors. Interestingly, these 2 modules also showed signif-icant associationwith let-7c expression. Let-7 is amicroRNAand is a bona fide tumor suppressor. Let-7 inhibits theexpression of multiple oncogenes, including RAS, MYC,and HMGA2. Multiple important cell-cycle control genesare repressed by let-7, including cyclinD1, cyclinD3, cyclin A,CDK4 and CCNA2, CDC25A, CDK6, and CDK8 (51, 52).Decreased let-7 expression is linked to increased tumori-genesis and poor patient prognosis (53–56). This is inagreement with our observation that a decrease in let-7cexpression in tumors versus normal bladder epitheliumandan increase in expression in Iressa-treated tumors versuscontrol tumors. Cell-culture study also confirmed that theincreased expression of let-7c decreases Iressa-treated blad-der tumor cell growth. By searchingmiRNA target database,we found 2 hub genes differentially expressed with Iressatreatment, Uhrf1 and Tacc3, which are target of let-7c.

TCC is by far the most common form of human bladdercancer. It accounts for overall more than 90% of bladdercancer cases. A total of 70%of TCC cases are characterized assuperficial, meaning that the cancer is confined to the liningof the patient’s bladder, and therefore unlikely to spreadinto neighboring tissue (i.e., metastasize). The remaining30% of TCC cases are characterized as muscle invasive,meaning that they have invaded through the lining intothe muscular wall of the bladder, and therefore potentiallyinto other nearby organs. Squamous cell carcinomaaccounts for around 8% of bladder cancer cases. Thesecancers originate from the thin, flat cells that typically formas a result of bladder inflammation or irritation that hastaken place for manymonths or years. Nearly all squamouscell carcinomas (SCC) are invasive. The comparison ofgene expression profile of advanced squamous and TCCof the bladder showed that out of the 516 genes whichwere differentially expressed, only 30 showed significantdifferences between TCC and SCC (57). Another studyshowed that MHC class II, TNF, cytokines, and chemokinesare among the upregulated genes in bladder SCC (58),which are similar to the findings of others who foundthat the invading cells of the bladder TCC stimulate theimmune system. However, a group of immunology-relatedgenes were downregulated in SCC compared with normalurothelium including interferon, different immunoglobu-lins, human leukocyte antigens protein, and interleukin-8and interleukin-13. Cell adhesion molecules such as lami-nins, integrins, and 2 cadherins (E-cadherin and P-cadherin)






weredownregulated inbothSCCbladder cancerand invasivebladder TCC. These data indicated that the majority of geneexpression profiles of bladder TCC and SCC are similar. Inexperiments conducted using male and female Fischer-344rats, intragastric administration of OH-BBN has been shownto induce TCCs of the urinary bladder which were histolog-ically similar to the human counterpart (4). This should be agood animal model to find potential biomarkers of chemo-prevention agents.

In summary, we identified 2 coexpression modules thatare significantly correlated with tumor status and Iressatreatment status. We further showed that these 2 modulesare also significantly associated with expression levels oflet-7c. Both tumor module and treatment module areenriched for genes involved in cell-cycle control. Thetumor module and treatment module contain 24 and21 highly connected hub genes that are likely to be keydrivers of the cell cycle. These results suggest that let-7cand its regulated cell-cycle pathway might play a key rolein governing bladder tumor suppression or collaborativeoncogenesis and that Iressa exhibits its preventive efficacyon bladder tumorigenesis by regulating let-7 and its cell-cycle control. Cell-culture study confirmed that theincreased expression of let-7c decreases Iressa-treated

bladder tumor cell growth. The identified hub genes mayalso serve as pharmacodynamic or efficacy biomarkers inclinical trials of bladder cancer chemoprevention. Char-acterization of these hub genes may help to give novelinsight into bladder tumorigenesis and develop noveldrug targets. Finally, we showed that alternative exonusage correlates with bladder tumorigenesis and couldbe inhibited by the chemopreventive agent Iressa. Theshort-term Iressa treatment may operate through blockingof such effects.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Grant Support

The funding for the studies was provided in part byNCIContract NumberHHSN-261200433008C (N01-CN43308).

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

ReceivedDecember 2, 2010; revised July 24, 2011; accepted September 23,2011; published OnlineFirst October 7, 2011.

References1. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010. CA Cancer J

Clin 2010;60:277–300.2. Becci PJ, Thompson HJ, Strum JM, Brown CC, Sporn MB, Moon RC.

N-butyl-N-(4-hydroxybutyl)nitrosamine-induced urinary bladder can-cer in C57BL/6 X DBA/2 F1 mice as a useful model for study ofchemoprevention of cancer with retinoids. Cancer Res 1981;41:927–32.

3. McCormick DL, Ronan SS, Becci PJ, Moon RC. Influence of total doseanddose schedule on induction of urinary bladder cancer in themouseby N-butyl-N-(4-hydroxybutyl)nitrosamine. Carcinogenesis 1981;2:251–4.

4. Grubbs CJ, Moon RC, Squire RA, Farrow GM, Stinson SF, GoodmanDG, et al. 13-cis-Retinoic acid: inhibition of bladder carcinogenesisinduced in rats by N-butyl-N-(4-hydroxybutyl)nitrosamine. Science1977;198:743–4.

5. Grubbs CJ, Lubet RA, Koki AT, Leahy KM, Masferrer JL, Steele VE,et al. Celecoxib inhibits N-butyl-N-(4-hydroxybutyl)-nitrosamine-induced urinary bladder cancers in male B6D2F1 mice and femaleFischer-344 rats. Cancer Res 2000;60:5599–602.

6. Moon RC, Kelloff GJ, Detrisac CJ, Steele VE, ThomasCF, Sigman CC.Chemoprevention of OH-BBN-induced bladder cancer in mice bypiroxicam. Carcinogenesis 1993;14:1487–9.

7. Han SY, Iliopoulos D, Druck T, Guler G, Grubbs CJ, Pereira M, et al.CpG methylation in the Fhit regulatory region: relation to Fhit expres-sion in murine tumors. Oncogene 2004;23:3990–8.

8. Lubet RA, Huebner K, Fong LY, Altieri DC, Steele VE, Kopelovich L,et al. 4-Hydroxybutyl(butyl)nitrosamine-induced urinary bladder can-cers in mice: characterization of FHIT and survivin expression andchemopreventive effects of indomethacin. Carcinogenesis 2005;26:571–8.

9. Herman MP, Svatek RS, Lotan Y, Karakiewizc PI, Shariat SF. Urine-based biomarkers for the early detection and surveillance of non-muscle invasive bladder cancer. Minerva Urol Nefrol 2008;60:217–35.

10. YaoR, Yi Y, Grubbs CJ, Lubet RA, YouM. Gene expression profiling ofchemically induced rat bladder tumors. Neoplasia 2007;9:207–21.

11. WilliamsPD, Lee JK, TheodorescuD.Molecular credentialing of rodentbladder carcinogenesis models. Neoplasia 2008;10:838–46.

12. Lubet RA, Steele VE, Juliana MM, Grubbs CJ. Screening agents forpreventive efficacy in a bladder cancer model: study design, endpoints, and gefitinib and naproxen efficacy. J Urol 2010;183:1598–603.

13. Strimmer K. A unified approach to false discovery rate estimation.BMC Bioinformatics 2008;9:303.

14. Mootha VK, LindgrenCM,ErikssonKF, SubramanianA, SihagS, LeharJ, et al. PGC-1alpha-responsive genes involved in oxidative phos-phorylation are coordinately downregulated in human diabetes. NatGenet 2003;34:267–73.

15. SubramanianA, TamayoP,Mootha VK,Mukherjee S, Ebert BL,GilletteMA, et al. Gene set enrichment analysis: a knowledge-based approachfor interpreting genome-wide expression profiles. Proc Natl Acad SciU S A 2005;102:15545–50.

16. Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis anddisplay of genome-wide expression patterns. ProcNatl Acad Sci USA1998;95:14863–8.

17. Zhang B, Horvath S. A general framework for weighted gene co-expression network analysis. Stat Appl Genet Mol Biol 2005;4:Article17.

18. Li A, Horvath S. Network neighborhood analysis with the multi-nodetopological overlap measure. Bioinformatics 2007;23:222–31.

19. Hu Z, Snitkin ES, DeLisi C. VisANT: an integrative framework fornetworks in systems biology. Brief Bioinform 2008;9:317–25.

20. Huang da W, Sherman BT, Lempicki RA. Systematic and integrativeanalysis of large gene lists using DAVID bioinformatics resources. NatProtoc 2009;4:44–57.

21. Chen LL, Sabripour M, Wu EF, Prieto VG, Fuller GN, Frazier ML. Amutation-created novel intra-exonic pre-mRNA splice site causesconstitutive activation of KIT in humangastrointestinal stromal tumors.Oncogene 2005;24:4271–80.

22. MazoyerS, PugetN,Perrin-VidozL, LynchHT,Serova-SinilnikovaOM,Lenoir GM. A BRCA1 nonsense mutation causes exon skipping. Am JHum Genet 1998;62:713–5.

23. Wang XQ, Luk JM, Leung PP, Wong BW, Stanbridge EJ, Fan ST.Alternative mRNA splicing of liver intestine-cadherin in hepatocellularcarcinoma. Clin Cancer Res 2005;11:483–9.

Lu et al.





24. CarlsonMR, ZhangB, FangZ,Mischel PS,HorvathS,NelsonSF.Geneconnectivity, function, and sequence conservation: predictions frommodular yeast co-expression networks. BMC Genomics 2006;7:40.

25. Kassouf W, Dinney CP, Brown G, McConkey DJ, Diehl AJ, Bar-Eli M,et al. Uncoupling between epidermal growth factor receptor anddownstream signals defines resistance to the antiproliferative effectof Gefitinib in bladder cancer cells. Cancer Res 2005;65:10524–35.

26. Cartegni L, Chew SL, Krainer AR. Listening to silence and understand-ing nonsense: exonic mutations that affect splicing. Nat Rev Genet2002;3:285–98.

27. Umar A, Della'ZannaG, Lubet R. Further thoughts on preclinical animalmodels for cancer prevention: when is it best to start treatment? Whatare potential histopathologic endpoints? Semin Oncol 2010;37:339–44.

28. Cuzick J, Buzdar A, BaumM,BiancoR, ColemanR, ConstenlaM, et al.Adjuvant use of anastrozole in breast cancer. J Clin Oncol 2004;22:1524–6; author reply 1526–7.

29. Christov K, Grubbs CJ, Shilkaitis A, JulianaMM, Lubet RA. Short-termmodulation of cell proliferation and apoptosis and preventive/thera-peutic efficacy of various agents in a mammary cancer model. ClinCancer Res 2007;13:5488–96.

30. Izawa JI, Slaton JW, Kedar D, Karashima T, Perrotte P, Czerniak B,et al. Differential expression of progression-related genes in the evo-lution of superficial to invasive transitional cell carcinoma of thebladder. Oncol Rep 2001;8:9–15.

31. Pruthi RS,NielsenM,HeathcoteS,WallenEM,RathmellWK,GodleyP,et al. A phase II trial of neoadjuvant erlotinib in patients with muscle-invasive bladder cancer undergoing radical cystectomy: clinical andpathological results. BJU Int 2010;106:349–56.

32. WulfingC,Machiels JP,Richel DJ,GrimmMO, Treiber U,DeGrootMR,et al. A single-arm, multicenter, open-label phase 2 study of lapatinibas the second-line treatment of patients with locally advanced ormetastatic transitional cell carcinoma. Cancer 2009;115:2881–90.

33. Horvath S, Zhang B, Carlson M, Lu KV, Zhu S, Felciano RM, et al.Analysis of oncogenic signaling networks in glioblastoma identifiesASPM as a molecular target. Proc Natl Acad Sci U S A 2006;103:17402–7.

34. Gao F, Ponte JF, Levy M, Papageorgis P, Cook NM, Ozturk S, et al.hBub1 negatively regulates p53mediated early cell death uponmitoticcheckpoint activation. Cancer Biol Ther 2009;8:548–56.

35. Glinsky GV, Berezovska O, Glinskii AB. Microarray analysis identifies adeath-from-cancer signature predicting therapy failure in patients withmultiple types of cancer. J Clin Invest 2005;115:1503–21.

36. Kiyomitsu T, Obuse C, Yanagida M. Human Blinkin/AF15q14 isrequired for chromosome alignment and the mitotic checkpointthrough direct interaction with Bub1 and BubR1. Dev Cell 2007;13:663–76.

37. Koseoglu MM, Graves LM, Marzluff WF. Phosphorylation of threonine61 by cyclin a/Cdk1 triggers degradation of stem-loop binding proteinat the end of S phase. Mol Cell Biol 2008;28:4469–79.

38. Wheeler LW, LentsNH,Baldassare JJ.CyclinA-CDKactivity duringG1phase impairs MCM chromatin loading and inhibits DNA synthesis inmammalian cells. Cell Cycle 2008;7:2179–88.

39. Trinkle-Mulcahy L, Andersen J, Lam YW, Moorhead G, Mann M,Lamond AI. Repo-Man recruits PP1 gamma to chromatin and isessential for cell viability. J Cell Biol 2006;172:679–92.

40. Hussein D, Taylor SS. Farnesylation of Cenp-F is required for G2/Mprogression and degradation after mitosis. J Cell Sci 2002;115:3403–14.

41. Martinez-Garay I, Rustom A, Gerdes HH, Kutsche K. The novel cen-trosomal associated protein CEP55 is present in the spindle midzoneand the midbody. Genomics 2006;87:243–53.

42. Ganem NJ, Compton DA. The KinI kinesin Kif2a is required for bipolarspindle assembly through a functional relationship with MCAK. J CellBiol 2004;166:473–8.

43. Davezac N, Baldin V, Blot J, Ducommun B, Tassan JP. Human pEg3kinase associates with and phosphorylates CDC25B phosphatase: apotential role for pEg3 in cell cycle regulation. Oncogene 2002;21:7630–41.

44. Ball AR Jr, Schmiesing JA, Zhou C, Gregson HC, Okada Y, Doi T, et al.Identification of a chromosome-targeting domain in the human con-densin subunit CNAP1/hCAP-D2/Eg7.Mol Cell Biol 2002;22:5769–81.

45. Harao M, Hirata S, Irie A, Senju S, Nakatsura T, Komori H, et al. HLA-A2-restricted CTL epitopes of a novel lung cancer-associated cancertestis antigen, cell division cycle associated 1, can induce tumor-reactive CTL. Int J Cancer 2008;123:2616–25.

46. Kanehira M, Katagiri T, Shimo A, Takata R, Shuin T, Miki T, et al.Oncogenic role of MPHOSPH1, a cancer-testis antigen specific tohuman bladder cancer. Cancer Res 2007;67:3276–85.

47. LuY, Liu P,WenW,GrubbsCJ, TownsendRR,Malone JP, et al. Cross-species comparison of orthologousgene expression in humanbladdercancer and carcinogen-induced rodent models. Am J Transl Res2010;3:8–27.

48. Unoki M, Kelly JD, Neal DE, Ponder BA, Nakamura Y, Hamamoto R.UHRF1 is a novel molecular marker for diagnosis and the prognosis ofbladder cancer. Br J Cancer 2009;101:98–105.

49. Kanehira M, Harada Y, Takata R, Shuin T, Miki T, Fujioka T, et al.Involvement of upregulation of DEPDC1 (DEP domain containing 1) inbladder carcinogenesis. Oncogene 2007;26:6448–55.

50. Stone R 2nd, Sabichi AL, Gill J, Lee IL, Adegboyega P, Dai MS, et al.Identification of genes correlated with early-stage bladder cancerprogression. Cancer Prev Res 2010;3:776–86.

51. Johnson CD, Esquela-Kerscher A, Stefani G, Byrom M, Kelnar K,Ovcharenko D, et al. The let-7 microRNA represses cell proliferationpathways in human cells. Cancer Res 2007;67:7713–22.

52. Schultz J, Lorenz P, Gross G, Ibrahim S, Kunz M. MicroRNA let-7btargets important cell cycle molecules in malignant melanoma cellsand interferes with anchorage-independent growth. Cell Res2008;18:549–57.

53. Bussing I, Slack FJ, Grosshans H. let-7 microRNAs in development,stem cells and cancer. Trends Mol Med 2008;14:400–9.

54. Lee YS, Dutta A. The tumor suppressor microRNA let-7 represses theHMGA2 oncogene. Genes Dev 2007;21:1025–30.

55. Li Y, VandenBoom TG 2nd, Kong D,Wang Z, Ali S, Philip PA, et al. Up-regulation of miR-200 and let-7 by natural agents leads to the reversalof epithelial-to-mesenchymal transition in gemcitabine-resistant pan-creatic cancer cells. Cancer Res 2009;69:6704–12.

56. Takamizawa J, Konishi H, Yanagisawa K, Tomida S, Osada H, EndohH, et al. Reduced expression of the let-7 microRNAs in human lungcancers in association with shortened postoperative survival. CancerRes 2004;64:3753–6.

57. Khaled HM, Abu-Taleb F, Zekri A, Haggag R. Gene expression profileof advanced squamous and transitional cell carcinoma of the bladder.J Clin Oncol 28:15s, 2010 (suppl; abstr 4572).

58. Ewis AA, El-Samman E, Ali N, Kajimoto K, Shinohara Y, Ishikawa M,et al. Gene expressionprofile in squamous cell carcinomaof the urinarybladder using complementary deoxyribonucleic acid microarray. UrolOncol 2007;25:120–7.






2012;5:248-259. Published OnlineFirst October 7, 2011.Cancer Prev Res Yan Lu, Pengyuan Liu, Francoise Van den Bergh, et al. Cancerby the EGFR Inhibitor Gefitinib (Iressa) in Rat Urinary Bladder Modulation of Gene Expression and Cell-Cycle Signaling Pathways

Updated version

10.1158/1940-6207.CAPR-10-0363doi:

Access the most recent version of this article at:

Cited articles

http://cancerpreventionresearch.aacrjournals.org/content/5/2/248.full#ref-list-1

This article cites 56 articles, 21 of which you can access for free at:

Citing articles

http://cancerpreventionresearch.aacrjournals.org/content/5/2/248.full#related-urls

This article has been cited by 2 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://cancerpreventionresearch.aacrjournals.org/content/5/2/248To request permission to re-use all or part of this article, use this link



http://cancerpreventionresearch.aacrjournals.org/lookup/doi/10.1158/1940-6207.CAPR-10-0363

http://cancerpreventionresearch.aacrjournals.org/content/5/2/248.full#ref-list-1

http://cancerpreventionresearch.aacrjournals.org/content/5/2/248.full#related-urls

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

mailto:[email protected]

http://cancerpreventionresearch.aacrjournals.org/content/5/2/248


modulation of gene expression and cell-cycle signaling...

Documents