Consistent and Differential Genetic Aberrations between ...clincancerres.aacrjournals.org/content/clincanres/19/21/5867.full.pdfPurpose: Our aim was to identify frequent genomic aberrations

Human Cancer Biology

Consistent and Differential Genetic Aberrations betweenEsophageal Dysplasia and Squamous Cell CarcinomaDetected By Array Comparative Genomic Hybridization

Zhi-Zhou Shi, Li Shang, Yan-Yi Jiang, Jia-Jie Hao, Yu Zhang, Tong-Tong Zhang, De-Chen Lin, Shu-Guang Liu,Bo-Shi Wang, Ting Gong, Qi-Min Zhan, and Ming-Rong Wang

AbstractPurpose: Our aim was to identify frequent genomic aberrations in both esophageal squamous cell

carcinoma (ESCC) and esophageal dysplasia and to discover important copy number-driving genes and

microRNAs (miRNA) in ESCC.

Experimental Design: We conducted array-based comparative genomic hybridization (array CGH) on

59ESCCresection samples and16dysplasia biopsy samples. Expressionof genes at 11q13.3was analyzedby

real-time PCR (RT-PCR) and immunohistochemistry (IHC). Integrated analysis was conducted to identify

genes or miRNAs with copy number-expression correlations.

Results: Array CGH identified 11 amplifications and eight homozygous deletions in ESCC. Integrated

analysis of array CGH data with matched gene expression microarray data showed that 90 overexpressed

genes and 24 underexpressed genes were consistent with DNA copy number changes, including 12 copy

number-driving miRNAs. In esophageal dysplasia, six gains, four losses, 12 amplifications, and four

homozygous deletions were detected. Amplifications of 7p11.2 and 11q13.2–11q13.3 (CCND1) and

homozygous deletion at 9p21.3 (CDKN2A) were consistent genomic changes in both dysplasia and

carcinoma. ANO1 at 11q13.3 was overexpressed at the mRNA and protein levels in tumors, and higher

mRNA expression was correlated with the copy number increase. In particular, ANO1 expression was

elevated in moderate dysplasia compared with normal esophageal epithelium. IHC revealed that ANO1

overexpression was positively correlated with lymph node metastasis and advanced clinical stage. Knock-

down of ANO1 significantly inhibited the proliferation of KYSE30 and KYSE510 cells.

Conclusion:Copynumber aberrations inboth esophageal dysplasia andESCCmaybeuseful as potential

biomarkers for early detection. In addition, ANO1 may be a candidate target gene in esophageal

tumorigenesis. Clin Cancer Res; 19(21); 5867–78. �2013 AACR.

IntroductionEsophageal cancer is the sixth leading cause of cancer-

related death in the world and the fourth leading cause ofcancer-related death in China (1). There are two types ofesophageal cancer, squamous cell carcinoma (ESCC) andadenocarcinoma (EAC), and most Chinese esophagealcancers are ESCCs (2). Because most patients are diagnosedat a later stage with distant or lymph node metastasis,esophageal cancer has a poor prognosis, with only a 17%

5-year survival rate (3). Therefore, identification of biomar-kers that can be used in early detection and understandingof the molecular changes underlying this cancer will lay thefoundations for reducing the mortality of patients withESCC and improve clinical management and outcomes.

ESCC develops in a multistep process from a precancer-ous dysplastic lesion through carcinoma in situ to invasivecarcinoma (4). The acquisition of genetic alterations isclosely related to the dysplasia-cancer sequence. Althoughextensive genomic changes including gains of 3q26-qter,5p15, 11q13.3, and 8q24.3-qter and losses of 16p13.3 and18q22-qter have been reported in ESCC, the alterations indysplasia remain unclear (5–8). Thus, identifying genomicalterations in dysplasia and squamous cell carcinoma of theesophaguswill behelpful in thediscovery of carcinogenesis-associated genes andof biomarkers for the early detection ofesophageal cancer.

Genomic aberrations are one of themechanisms that canresult in gene dysfunction and contribute to carcinogenesisand tumor progression. Differentially expressed genes asso-ciated with DNA copy number changes may be candidate

Authors' Affiliation: State Key Laboratory of Molecular Oncology, CancerInstitute and Hospital, Peking Union Medical College and Chinese Acad-emy of Medical Sciences, Beijing, China

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

Corresponding Author:Ming-RongWang, Peking UnionMedical Collegeand Chinese Academy of Medical Sciences, 17 Panjiayuan Nanli,Chaoyang District, Beijing 100021, China. Phone: 86-108-778-8788; Fax:86-108-777-8651; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-12-3753

�2013 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org 5867

on October 10, 2020. © 2013 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst September 5, 2013; DOI: 10.1158/1078-0432.CCR-12-3753

http://clincancerres.aacrjournals.org/

targets of amplifications or homozygous deletions.Although expressionmicroarrays have been used to analyzeESCC tissues and many genes have been associated withcarcinogenesis or tumor metastasis, the extent of genesaffected by copy number aberrations is not fully under-stood. Integrated analysis of comparative genomic hybrid-ization (CGH),mRNA expression, andmicroRNA (miRNA,miR) expression data can be used to screen and identifycopy number–associated genes at a whole-genome level.

To clarify recurrent genomic changes during the carcino-genesis of ESCC and identify potential biomarkers for earlydiagnosis of this disease, we compared the genetic changesin 59 ESCCs and 16 esophageal dysplasias by array-basedCGH (array CGH). We also evaluated the mRNA andmiRNA expression profiles of ESCCs and conducted inte-grated analysis to screen for copy number variation–asso-ciated genes and miRNAs. Validation was conducted toidentify genes underlying DNA copy number changes fromdysplasia to ESCC at the mRNA and protein expressionlevels and their oncogenic roles.

Materials and MethodsSample information

Fresh tissues containing ESCCs and adjacent histologi-cally normal epithelia from 233 patients with ESCC werecollected by theDepartment of Pathology, CancerHospital,Chinese Academy of Medical Sciences (Beijing, China).Primary tumor regions and the corresponding normalmucosa from the same patients were separated by experi-enced pathologists and immediately stored at �70�Cuntil use. All patients with ESCC were treated with radical

operation, and none of them received any preoperativetreatment.

Biopsy tissues were taken from symptom-free patientsduring endoscopic screening for esophageal cancer in Linz-hou, China, which is a well-recognized high-risk area forESCC.During endoscopy, the entire esophagus was visuallyexamined and biopsies were taken from all focal lesions. Ifno focal lesions were detected, a standard site in the mid-esophagus was sampled.

All the samples used in this studywere residual specimensafter diagnostic sampling. All operated or endoscopicallyscreened patients signed separate informed consent formsfor sampling andmolecular analysis. The clinicopathologiccharacteristics of patients used in the array CGH study areshown in Supplementary Tables S1 and S2.

Cell cultureThe human ESCC KYSE30 and KYSE510 cell lines were

generously provided by Dr. Y. Shimada (Kyoto University,Kyoto, Japan). Cells were cultured in RPMI-1640 (Invitro-gen) supplemented with 10% FBS.

Genomic DNA extractionGenomic DNAwas isolated from tumor tissues using the

Qiagen DNeasy Blood & Tissue Kit as described by themanufacturer (Qiagen). The tumor cell content of all thesamples was more than 50% by hematoxylin and eosin(H&E) staining.

Array CGHESCCs were tested using a 44K human genome CGH

microarray, and esophageal dysplasias were tested using a60K microarray. Array CGH experiments were carried outusing standard Agilent protocols (Agilent Technologies).Commercial humangenomicDNA(Promega)wasused as areference. For each CGH hybridization, 500 and 400 ng ofreference genomic DNA for the 44K and 60K microarrays,respectively, and the same amounts of tumor DNA weredigested with the AluI and RsaI restriction enzymes (Pro-mega). The digested reference DNA fragments were labeledwith cyanine 5-dUTP and the tumor DNA was labeled withcyanine 3-dUTP (Agilent Technologies). After cleanup,labeled reference and tumor DNA were mixed as probesand hybridized onto an Agilent 44K/60K human genomeCGH microarray (Agilent Technologies) for 40 hours/24hours. Washing, scanning, and data extraction procedureswere carried out following standard protocols.

Total RNA extractionTotal RNA was isolated from tissues using the RNeasy

Mini Kit as described by the manufacturer (Qiagen) andused for gene/miRNA expressionmicroarray assay and real-time PCR (RT-PCR) assay.

Gene expression microarray detectionAgilent Human Gene Expression Microarrays (Agilent

Technologies) were used to determine the gene expressionlevels in paired tumorous and paracancerous tissues. Total

Translational RelevanceEarly detection can greatly increase the chances of

successful treatment of human cancer. In the presentstudy, we carried out array-based comparative genomichybridization (CGH) to compare genomic aberrationsbetween esophageal dysplasia and squamous cell carci-noma (SCC) and identified amplifications of 7p11.2 and11q13.2-11q13.3 (CCND1) and homozygous deletionat 9p21.3 (CDKN2A) as consistent genomic changes inboth dysplasia and carcinoma. Copy number-drivinggenes/microRNAswere identifiedby integrative analysis.ANO1 on 11q13.3 was overexpressed as a result ofamplification in esophageal squamous cell carcinoma(ESCC), particularly with an elevation in moderate dys-plasia compared with normal esophageal epithelium.Overexpression of ANO1 was correlated with lymphnodemetastasis and advanced tumor stage. Knockdownof ANO1 significantly inhibited ESCC cell proliferation.These data revealed the molecular changes in precancer-ous lesions and ESCC. The genomic aberrations thatwere consistent or differed between dysplasia andESCC may be useful as candidate biomarkers for earlydetection.

Shi et al.

Clin Cancer Res; 19(21) November 1, 2013 Clinical Cancer Research5868




RNA was used as a template for cDNA synthesis using theOne Color Low Input Agilent Quick Amp Labeling Kit, andthe Spike-in Kit provided the positive controls. The AgilentOne-Color Microarray-based gene expression analysis fol-lowed themanufacturer’s instructions, and passed Agilent’squality control.

miRNA expression microarray detectionAgilent Human miRNA Microarrays (Agilent Technolo-

gies) were used to examine miRNA expression levels inpaired tumorous and paracancerous tissues. The input forthe AgilentmiRNA labeling systemwas 100ng of total RNA.After dephosphorylation and denaturation, the total RNAwas labeled with cyanine 3-pCp and then hybridized toAgilent Human miRNA Microarray V2.0 using the miRNAComplete Labeling and Hyb Kit. Following hybridizationfor 20 hours, the slides were washed using the GeneExpression Wash Buffer Kit and scanned using an AgilentScanner. The images were processed and analyzed withAgilent Feature Extraction Software. The raw data werenormalized using quantile normalization and then ana-lyzed in GeneSping GX software (zcomSilicon Genetics).

Microarray data analysisArray CGH data were analyzed using Genomic Work-

bench (Agilent Technologies), BRB-CGHTools (http://linus.nci.nih.gov/BRB-ArrayTools.html) and MD-SeeGH(www.flintbox.com). Genomic Workbench was used tocalculate the log2 ratio for every probe and to identifygenomic aberrations. The mean log2 ratio of all probes ina chromosome region between 0.25 and 1.0 was classifiedas genomic gain, more than 1.0 as high-level DNA ampli-fication, less than �0.25 as hemizygous loss, and less than�1.0 as homozygous deletion.Gene/miRNA expression microarray data were analyzed

by GeneSpring GX software (Agilent Technologies).The array CGH data set is available at Gene Expression

Omnibus (GEO; http://www.ncbi.nlm.nih.gov/geo/) underaccession number GSE46452.

RT-PCRRT-PCR was used to detect the copy number and mRNA

expression of ESCC and precancerous lesions. The PCRreactions were carried out in a total volume of 20 ml,including 10 ml of 2�Power SYBR Green PCR Master Mix(Applied Biosystems), 2 ml of cDNA/DNA (5 ng/ml), and 1ml of primer mix (10 mmol/L each). PCR amplification anddetection were conducted in a LightCycler 480 II (RocheApplied Science) as follows: an initial denaturation at 95�Cfor 10 minutes, 40 cycles of 95�C for 15 seconds and 60�Cfor 1minute. The relative gene expression and copy numberwere calculated using the comparative CT method (9). Thegene expression and copy number of the target gene werenormalized to an endogenous reference [glyceraldehyde-3-phosphate dehydrogenase (GAPDH)], and relative to the

calibrator were given by the formula 2�DDCT . DCT wascalculated by subtracting the average GAPDH CT from theaverage CT of the gene of interest. The ratio defines the level

of relative expression and copy number status of the targetgene to that of GAPDH. The primers for the examined genesare listed in Supplementary Table S3.

Immunohistochemical stainingFormalin-fixed, paraffin-embedded esophageal tumors

were placed on the tissue microarray. For each case, thecancer tissues were repeated three times and adjacent mor-phologically normal tissues were repeated two times. Theslides were deparaffinized, rehydrated, immersed in a 3%hydrogen peroxide solution for 10 minutes, heated incitrate buffer (pH 6) for 25 minutes at 95�C, and cooledfor 60 minutes at room temperature. The slides wereblocked with 10% normal goat serum for 30 minutes at37�C and then incubated with rabbit polyclonal antibodyagainst ANO1 (Abcam) and mouse monoclonal antibodyagainst MCM7 (Santa Cruz Biotechnology) overnight at4�C. After washing with PBS, the slides were incubated withbiotinylated secondary antibody (diluted 1:100) for 30minutes at 37�C, followed by streptavidin-peroxidase(1:100 dilution) incubation for 30minutes at 37�C. Immu-nolabeling was visualized with a mixture of 3,30-diamino-benzidine (DAB) solution. Tissues were counterstainedwith hematoxylin.

Expression level was determined on the basis of stainingintensity and the percentage of immunoreactive cells. Neg-ative expression (score ¼ 0) was considered no, or faintstaining, or moderate to strong staining in less than 25% ofcells.Weak expression (score¼ 1)was consideredmoderateor strong staining in 25% to 50% of cells. Strong expression(score ¼ 2) was considered strong staining in more than50% of cells.

Transfection assaysiRNAs specific to ANO1 (Silencer Select siRNA; s30185)

and a control siRNA (Silencer SelectNegativeControl no. 1)were purchased from Ambion). Cell transfections wereconducted using Lipofectamine 2000 (Invitrogen) accord-ing to the manufacturer’s instructions.

Western blot analysisCells were washed with ice-cold PBS and lysed in

radioimmunoprecipitation assay (RIPA) buffer [50mmol/L Tris (pH 7.5), 150 mmol/L NaCl, 1% NP-40,0.5% sodium deoxycholate, and 0.1% SDS] containingphenylmethylsulfonylfluoride (PMSF;1 mmol/L) andprotease inhibitors (2 g/mL; Protease Inhibitor CocktailSet III; Calbiochem) on ice for 30 minutes. The lysateswere clarified by centrifugation at 13,000 � g for 30minutes at 4�C. The total protein concentration wasestimated using a Protein Assay Kit (Bio-Rad). Proteinswere separated by SDS-PAGE, transferred to polyvinyli-dene difluoride (PVDF) membranes (Millipore), blocked,and probed with antibodies against ANO1 (1:1,000;Santa Cruz Biotechnology). After washing, blots wereincubated with horseradish peroxidase–conjugated sec-ondary antibodies and visualized using super enhancedchemiluminescence (ECL) detection reagent (Applygen).

Genomic Aberrations in Esophageal Carcinogenesis

www.aacrjournals.org Clin Cancer Res; 19(21) November 1, 2013 5869




Cell proliferation assayACell Counting Kit8 (CCK-8; Dojindo Laboratories) was

used to evaluate cell proliferation. At 24 hours after trans-fection, the cells were seeded into 96-well plates with 1,000cells per well and cultured for 24, 48, 72, 96, and 120 hours.Then, 10 ml of CCK-8 solution was added to each well andincubated for 1 hour. Optical density was examined at awavelength of 450 nm.

Statistical analysisThe correlation between genomic aberrations and clini-

copathologic parameters was analyzed in three steps. First,we identified candidate genomic aberrations with differentfrequencies among clinicopathologic statuses using MD-CGH software. We then narrowed the regions of thesecandidates using the aberration data produced by GenomicWorkbench. Finally, we analyzed the correlations betweenthe genomic aberrations and clinicopathologic parametersusing Pearson x2 test. The correlations between the copynumber and mRNA expression levels of CCND1, ANO1,FADD, and FGF19 were analyzed using the Spearman rankcorrelation test. Other statistical analyses were conductedusing Student t tests with the statistical software SPSS 15.0.Differences were considered statistically significant whenthe corresponding two-sided P value was less than 0.05.

ResultsRecurrent copy number alterations in ESCC by arrayCGH

In ESCCs, 19 gains and nine losses were detected inmorethan 50% of cases (Fig. 1A and Supplementary Table S4).Among them, the most common gain was detected at7q21.3-q22.1 (70%), whereas the most frequent region ofloss was 3p22.2-p24.3 (60%). Eleven regions including3q27.1 (EIF2B5), 7p11.2, 8p12-p11.21, 8q24.21 (MYC),11q13.2-q13.3 (CCND1), 12q14.3-q15, 13q22.1, 14q11.2(PARP2), and 19q13.11-q13.12 showed high-level amp-lifications, and eight homozygous deletions were iden-tified at 4p16.1-p15.1, 4q34.3-q35.1, 6p22.1, 9p21.3(CDKN2A), 1p15.4, 13q14.2 (RB1), 14q12, and 22q13.1(Table 1). Thirteen genes located in these amplificationregions, including AP2M1, ALG3, PSMD2, EIF4EBP1,CTTN, PPFIA1, ANO1, APEX1, USF2, FXYD3, LRFN3,HAUS5, and C14orf128, were overexpressed in esophagealtumors compared with paracancerous tissues (Table 1).Nearly half of ESCCs had 51 to 75 genomic changes (Fig.1B). GISTIC analysis, which scores the significance of recur-rent gains or losses anddetects peak regions likely to containdriver gene(s), revealed that gains of 7p11.2, 7q21.3,8q24.21, 11q13.2-q13.4, 12q14.3-q21.1, and 19q13.11-q13.12 and losses of 9p21.3 and 11p15.4 were the mostsignificant recurrent genomic aberrations (Fig. 1C; andSupplementary Table S5).

Frequency plot comparison analysis was applied to iden-tify genetic alterations associated with clinicopathologicparameters. The results showed that 11q was lost morefrequently in patients with lymph node metastasis and

that loss of 9p was more common in those at earlier stages(I and II; Supplementary Fig. S1A and S1B). Detailedfrequency and statistical analyses confirmed that the11q14.3 loss was more frequent in patients with lymphnode metastasis (P < 0.001). In addition, loss of 9p21.2-p21.1 was significantly associated with earlier clinical stagein ESCC (P < 0.001).

Integrated analysis of copy number and gene/miRNAexpression data

To explore the genes andmiRNAs that were differentiallyexpressed between tumorous andparacancerous tissues andalso consistent with genomic changes, we further analyzedeight and four ESCC samples (alsomeasured by arrayCGH)by cDNA expression microarray and miRNA expressionmicroarray, respectively. In total, 159 genes (170 probes)were identified when the cutoffs were a fold change morethan 5.0 and a P value less than 0.01. Among them, 135genes (141 probes) were overexpressed and 24 (29 probes)were underexpressed in tumor tissues compared with para-cancerous tissues (Supplementary Table S6). In addition, 90overexpressed genes (67% of all overexpressed genesincluding MCM7, MTDH, STAT2, MMP14, and RELB) andnineunderexpressed genes (38%)were consistentwith theirDNAcopynumber changes.We further validated themRNAand protein levels of the candidate gene MCM7. At themRNA level, MCM7 was overexpressed in 64% (9 of 14) ofESCCs. At the protein level,MCM7overexpression occurredin 88% (7 of 8) and 65% (13 of 20) of ESCCs examined byWestern blotting and immunohistochemistry (IHC),respectively (Supplementary Fig. S2).

Twenty-seven miRNAs were identified when the cutoffwas a fold change more than 2.0. Among them, 22miRNAswere overexpressed and five were underexpressed in tumortissues (Supplementary Table S7). Overexpression of miR-1280, miR-574-5p, miR-595, iR-548d-5p, miR-483-5p,miR-1260, miR-342-3p, miR-940, miR-21, miR-23a, andmiR-766 andunderexpression of let-7cwere consistentwiththe gains and losses, respectively. We then compared theselected genes/miRNAs with published data (Supplemen-tary Fig. S3) and found that COL1A1 and an additional 21genes, including MMP1 and SPP1, were selected in twostudies (10, 11). In studies by Kimura, Ogawa, and in ourstudy, miR-16, miR-107, andmiR-21 were overexpressed inesophageal tumor tissues. In addition, both we and Satofound that let-7c was underexpressed in ESCC (12, 13).

Differentially expressed genes were further analyzedusing the Gene Ontology (GO) option of GeneSpring GX11.5, yielding 22 significant GO terms (corrected P < 0.05).Detailed information on these GO terms is provided inSupplementary Table S8. The top five GO terms were asfollowing: (i) epidermis development (corrected P ¼0.0052); (ii) nuclear envelope (corrected P ¼ 0.0053); (iii)mRNA export from the nucleus (corrected P¼ 0.0249); (iv)mRNA transport (corrected P ¼ 0.0249); and (v) tissuedevelopment (corrected P ¼ 0.0249). Pathway enrichmentanalysis showed that six pathways were significantly chan-ged inESCC including theAlpha6Beta4Integrin, TNF-a/NF-

Shi et al.





kB, interleukin (IL-6), EGFR1, IL-4, and IL-3 pathways(Supplementary Table S9).

Recurrent copy number alterations in esophagealdysplasia by array CGHTo explore the early genomic changes in esophageal

carcinogenesis, we conducted array CGH on 16 precancer-ous lesions. The most frequent copy number alterationsin dysplasia with a frequency above 30% were gains of3q11.2-q29, 5p15.2-15.33, 7q21.3-q22.3, 8q11.21-q24.3,11q13.2-q13.3, and 22q13.33 and losses of 3p11.2-p26.3,9p24.3-p34.2, 13q21.1-q21.2, and 13q32.2-q32.3 (Fig. 2Aand Supplementary Table S10). Twelve high-level amplifi-

cations at 1q25.2 (RASAL2), 3q33-q34.1, 6p21.31, 7p11.2,7q21.3-q22.2, 8p12, 9p22.2-p21.3, 9p21.2-p21.1,11q13.2-q13.3 (CCND1), 14q21.1, 14q21.3-q23.1, and18q11.2 (RBBP8) and four homozygous deletions at2q22.1 (LRP1B), 7p14.1 (CDC2L5), 9p21.3 (CDKN2A),and 20p12.1 were identified (Table 1). GISTIC analysisidentified an 11q13.2-q13.3 gain and losses of 3q26.1 and11q11 in esophageal dysplasia (Fig. 2C SupplementaryTable S5).

By comparing the genomic profiles of ESCC and dys-plasia, we found that amplifications of 7p11.2 and11q13.2-11q13.3 and homozygous deletion at 9p21.3were consistent genomic changes and that 11q13.2-

7p11

.2

7q21

.3

9p21

.3

11p1

5.4

8q24

.21

11q1

3.2-

q13.

4

12q1

4.3-

q21.

1

19q1

3.11

-q13

.12

0 2 4

6

8 1

00 2 4

6

8 1

0

Am

plif

ication

-log10(F

DR

)

HD

-log10(F

DR

)

Number of aberrations

Num

ber

of cases

≤25 26–50 51–75 76–100 >100

25

20

15

10

5

0

1

A

B C

2 3 4 5

6 7 8 9

11 12 13 14 15

16 17

21 22

18 19 20

10

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

Figure 1. Genomic aberrations in ESCC. A, genome-wide frequency plot of ESCC by array CGH analysis. Line on the right of 0%-axis, gain; line on the left of0%-axis, loss. B, number of aberrations inESCC. X, number of aberrations; Y, number of cases. C, amplifications andhomozygousdeletions (HD) identifiedbyGISTIC.






q13.3 amplification was the only identified aberrationthat was common in both dysplasia and ESCC by GISTIC(Table 1 and Supplementary Table S5). A frequency plotcomparison analysis revealed that 3q, 5p, 7q, 8q, and11q13 and losses of 3p and 9p were shared by dysplasiaand cancer. Loss of 9q was more frequent in dysplasia, butgains of 16p, 17q, 19p, 19q, 20p, 20q, and 22q and lossesof 4p, 4q, 5q, 8p, 18q and 21q were more frequent inESCC (Fig. 2D). As the data obtained using all methodsshowed that 11q13.2-13.3 was frequently amplified inboth dysplasia and cancer, we validated the results usingFISH (Supplementary Fig. S4A). The number of genomic

aberrations in ESCC was significantly higher comparedwith esophageal dysplasia (Supplementary Fig. S4B).

Pathway enrichment analysis using the Kyoto Encyclo-pedia of Genes andGenomes (KEGG) database was appliedto the array CGH data, and we found that 14 and 11pathways were enriched for genes with genomic changesin ESCC and esophageal dysplasia, respectively (Supple-mentary Table S11). Interestingly, three pathways includingcytokine–cytokine receptor interaction, the Toll-like recep-tor signaling pathway, and regulation of autophagy wereenriched in genes with copy number changes in both ESCCand dysplasia.

Table 1. Amplification and homozygous deletion in squamous cell carcinoma and dysplasia of esophagus

ESCC Dysplasia

Aberration Cytoband RegionNo ofcases

Differentiallyexpressedgenesa Cytoband Region

No ofcases

Cancergenesb

Amp 1q25.2 176402810–177626551 1 RASAL23q27.1 185338220–185590314 5 AP2M1, ALG3,

PSMD2EIF2B5

4q33-q34.1 171236259–174696447 16p21.31 35311212–35335937 1

7p11.2 54572103–55680772 3 7p11.2 54725577–55414419 17p11.2 56304574–56754354 4

7q21.3-q22.2 92631344–105876199 18p12 35438363–35817762 1

8p12-p11.21 37715902–39068046 3 EIF4EBP18q24.21 127458201–127637605 58q24.21 128776700–129216964 5 MYC

9p22.2-p21.3 18464451–21895439 19p21.2-p21.1 28004311–28660535 1

11q13.2-q13.3 69131640–70182767 27 CTTN, PPFIA1,ANO1

11q13.2-q13.3 68943548–70182767 3 CCND1

12q14.3-q15 65950844–69024363 513q22.1 72630292–73097233 314q11.2 19895020–20767691 3 APEX1 PARP2

14q21.1 38288390–38571563 114q21.3-q23.1 45244742–58972895 118q11.2 18770696–19981761 1 RBBP8

19q13.11-q13.12 39868348–42263119 4 USF2, FXYD3,LRFN3, HAUS5

HD 2q22.1 141377244–142003772 1 LRP1B4p16.1-p15.1 10889900–28060348 24q34.3-q35.1 180569283–183227293 26p22.1 27208521–27210109 2

7p14.1 39952510–40008069 1 CDC2L59p21.3 21980522–21999029 11 9p21.3 21980522–22231139 4 CDKN2A11p15.4 9729948–9817397 213q14.2 47883983–47954006 3 RB114q12 31614404–31633839 2 C14orf128

20p12.1 14530401–14989538 122q13.1 37689058–37715431 4

NOTE: The number of ESCC and dysplasia in array CGH study are 59 and 16, respectively.Abbreviations: Amp, amplification; HD, homozygous deletion.In gray: the amplifications or homozygous deletions are shared by both groups.aGenes with different expression level between ESCC tissues and paracancerous tissues in our microarray assay.bThe cancer genes are identified by Cancer Gene resource http://cbio.mskcc.org/CancerGenes/.

Shi et al.





11q13.2-q13.3

3q26.1 11q11

0 2 4 6 8 1

00 2 4 6 8 1

0

Am

plif

ication

-log10(F

DR

)

HD

-log10(F

DR

)

8

6

4

2

0≤ 10 11–20 21–30 31–40 > 40

Number of aberrations

Num

ber

of cases

Dysplasia

ESCC

Both

1 2 3 4 5 6 7 8

9 10 11 12 13 14

17 18 19 20 21 22

15 16

1

A

B

C

D

2 3 4 5

6 7 8 9

11 12 13 14 15

16 17

21 22

18 19 20

10

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

100 80 60 40 20 0 20 40 60 80 100

Figure 2. Genomic aberrations detected in esophageal dysplasia. A, genome-wide frequency plot of esophageal dysplasia by array CGH analysis. Line on theright of 0%-axis, gain; line on the left of 0%-axis, loss. B, number of aberrations in esophageal dysplasia. X, number of aberrations; Y, number of cases. C,amplifications and homozygous deletions (HD) identified by GISTIC. D, frequency plot comparison of esophageal dysplasia with ESCC. The presentation isper array probe: gains are represented by the lines on the right and losses by the left. The vertical line represents 100% of the samples.






Candidate target genes in the sequence fromesophageal dysplasia to cancer

Wenext screened the candidate targets for 11q13.2-q13.3amplification. RT-PCR showed that the expression levels ofFGF19 and ANO1were significantly higher in tumor tissuesthan inparacancerous tissues, with a P value below0.05 andfold change above 2.0 (Fig. 3A). Therefore, we chose FGF19and ANO1 to examine the correlation between copy num-ber increase and mRNA overexpression in ESCC and theirmRNA expression status in normal esophageal epithelium,esophagitis, mild dysplasia, and moderate dysplasia. Wealso analyzed CCND1 and FADD because their overexpres-sion in ESCC has been previously reported. ANO1 andFADDwere overexpressed in ESCC with amplification (Fig.3B), and increased mRNA expression of ANO1 and FADDwas significantly correlated with copy number increases,according to a Spearman correlation analysis (Supplemen-tary Table S12). CCND1, ANO1, and FADD were signifi-cantly overexpressed in moderate dysplasia compared withnormal epithelium, but there was no difference in theexpression of these genes between normal epithelium andesophagitis (Fig. 3C). The IHC results further showed thatANO1 was overexpressed in 25% (n ¼ 88) of ESCCs andthat higher expression of ANO1was significantly correlatedwith lymphnodemetastasis and advanced tumor stage (Fig.

4A; Supplementary Table S13). In vitro studies revealed thatANO1 knockdown in the ESCC KYSE30 and KYSE510 celllines, which have increased ANO1expression, inhibited cellproliferation (Fig. 4B and C and Supplementary Figs. S5,S6A, and S6B).

DiscussionGenomic aberrations can contribute to carcinogenesis

and tumor progression. Earlier reports have identifiedmultiple abnormal regions in ESCC, including amplifica-tions at 1p34, 3q, 5p, 7p12, 8q, 11q13, 12p, 17q12, and22q as well as deletions at 2q, 3p, 4q, 5q13-q21, 9p21.3,and 13q (5, 7, 8, 14–21). Our study further narrowedthese altered chromosome regions and identified candi-date amplification- or homozygous deletion–associatedgenes. The known cancer genes EIF2B5, MYC, CCND1,and PARP2 were amplified, and CDKN2A and RB1 werehomozygously deleted in ESCC. Importantly, amplifica-tion of CCND1 and homozygous deletion of CDKN2Awere observed in both ESCC and dysplasia. It has beenreported that CCND1 is expressed in 20% of severeesophageal dysplasias and significantly amplified andoverexpressed in ESCC (22, 23). CDKN2A is deleted inESCC and underexpressed in 80% of esophageal

Figure 3. Expression analyses of candidate geneswithin the 11q13.3 byRT-PCR.A, expression of four genes (CCND1, ANO1, FADD, and FGF19) in ESCCandparacancerous tissues by RT-PCR (n ¼ 7). B, comparison of the expression of four genes (CCND1, ANO1, FADD, and FGF19) between patients withamplification and patients without amplification (n ¼ 23). C, expression of four genes (CCND1, ANO1, FADD, and FGF19) in normal epithelium, esophagitis,mild dysplasia, and moderate dysplasia by RT-PCR.

Shi et al.





dysplasias and 93% of ESCCs (23, 24). These findingsindicate that CCND1 and CDKN2A may play importantroles in esophageal carcinogenesis and that copy numberchange is one mechanism leading to the dysregulation ofthese two genes. In addition, the gene expression showedthat AP2M1 (3q27.1), ALG3 (3q27.1), PSMD2 (3q27.1),EIF4EBP1 (8p12), CTTN (11q13.3), PPFIA1 (11q13.3),ANO1 (11q13.3), APEX1 (14q11.2), USF2 (19q13.12),FXYD3 (19q13.12), LRFN3 (19q13.12), and HAUS5(19q13.12) were amplified and overexpressed in ESCC

and that C14orf128 at 14q12 were homozygously deletedand underexpressed.

ESCCarises fromdefinedhistopathologic lesions, includ-ing mild, moderate, and severe dysplasia. A followup studybyWang and colleagues indicated that in the high-risk ruralpopulation in Linzhou, China, the rates of mild, moderate,and severe dysplasias becoming ESCC after 13.5 years were23.7%, 50%, and 73.9%, respectively (25). Therefore, squa-mous dysplasia is a major risk factor and a precursorto ESCC. To date, many molecular changes have been

N

A

B C

T

×10 ×40 ×10 ×40

Case 1

ANO1 siRNA (d)

Case 2

KY

SE

30

No

nsilen

cin

g

1 3 5

ANO1

b-Actin

ANO1 siRNA (d)

KY

SE

510

No

nsilen

cin

g

1 3 5

ANO1

b-Actin

KYSE30 parental

Nonsilencing

ANO1 siRNA

KYSE510 parental

Nonsilencing

ANO1 siRNA

2.0

1.5

1.0

0.5

0.0

0 1 2 3 4 5 6

Time (d)

0 1 2 3 4 5 6

Time (d)

Cell

pro

lifera

tion

(absorb

ance a

t 450 n

m)

Cell

pro

lifera

tion

(absorb

ance a

t 450 n

m) 3

2

1

0

Figure 4. Overexpression of ANO1 in ESCCs and effect of ANO1 on cell proliferation. A, IHC assay of ANO1. N, paracancerous tissues; T, ESCC tissues. B andC, knockdown of ANO1 inhibited the proliferation of KYSE30 (B) and KYSE510 cells (C). The absorbance at 450 nm was measured.






identified in esophageal dysplasia. Yang and colleaguesfound that 21% of precancerous lesions and 38% of ESCCsexpress PTCH1, a target of hedgehog signaling, by IHC andsuggested that hedgehog signaling plays an important rolein esophageal carcinogenesis (26). Considering PTCH1located in the commonly deleted region 9q, further studyshould be conducted to analyze the correlation betweendeletion or no expression of PTCH1 and clinical factors inESCC. Another study showed that a panel of four genes(AHRR, P16INK4a, MT1G, and CLDN3) in esophagealballoon cytology specimens had a sensitivity and specificityof 50% and 68%, respectively, in detecting esophagealdysplasia (27). However, the understanding of genomicaberrations in esophageal dysplasia was quite limited. Thepresent study not only confirmed at themolecular level thatesophageal dysplasia was the precursor lesion of ESCC butalso revealed alterations of some genes in the precancerouslesions of the esophagus. Six gains, four losses, 12 ampli-fications, and four homozygous deletions were identified indysplasia. In particular, amplification of 7p11.2 and11q13.2-q13.3 and homozygous deletion at 9p21.3 weredetected both in dysplasia and ESCC. Interestingly, threepathways including cytokine–cytokine receptor interaction,the Toll-like receptor signaling pathway and regulation ofautophagy, were enriched in geneswith genomic changes inboth ESCC and dysplasia. Future research could focus onthe contribution of these three pathways to esophagealcarcinogenesis. In addition, the copy numbers of 3q26.1and 11q11 have often been reported to be increased intumors, including ESCC (28–30). Interestingly, we identi-fied 3q26.1 and 11q11 as novel regions of common loss inesophageal dysplasia, whichmay suggest that genes locatedin these two regions play different roles in dysplasia for-mation than in ESCC. We also found that loss of 9q wasmore frequent in dysplasia than in ESCC, and furtherstudies should be conducted to explore the tumorigenicrole of 9q loss.

By comparing our resultswith theCGHdata presented onthe Progenetix website (31, 32), we found that most geno-mic aberrations were consistent. However, there were somedifferences. For example, gains of chromosomes 16, 17, 19,and 22 were more common than their losses in our study,whereas the frequencies of gain and loss in chromosomes16 and17were similar, and the loss of chromosomes19 and22 was more dominant in the Progenetix database. In ourstudy, gain of 9q34.11-34.3 was detected in 53% of ESCCs,whereas the frequency of 9q gain is reported as less than20% in the Progenetix database. Considering the possibleexperimental artifacts, a large number of ESCCs need to bedetected to validate the differences between our results andthe Progenetix data.

11q13.2-q13.3, which was amplified in both ESCC anddysplasia in theCGHassay, is a commonly amplified regioninmany types ofmalignancies including esophageal, breast,liver, and ovarian cancers (8, 33–35). Hu and colleagueshave reported that gains of PSCA1, CCND1, CTTN, PPFIA1,and SHANK2 at 11q13 are correlated with increased RNAexpression in ESCC (36). Sugimoto and colleagues have

indicated that GAL is the target of 11q13 amplification inESCC (37).Our results showed that ANO1was significantlyoverexpressed as a result of amplification in esophagealcancer tissues and that overexpression of ANO1 was linkedto lymph node metastasis and advanced tumor stage. Inparticular, ANO1 knockdown significantly inhibited theproliferation of ESCC cells. On the basis of the elevatedexpression in moderate dysplasia compared with normalesophageal epithelium, ANO1 may play an oncogenic rolein esophageal carcinogenesis. ANO1was recently identifiedas a Ca(2þ)-activated Cl(�) channel. Overexpression ofANO1 has been reported in prostate carcinoma and squa-mous cell carcinoma of the head and neck (HNSCC). Inaddition, in prostate cancer, higher ANO1 expression hasbeen correlated with clinical tumor—node—metastasisstage and Gleason score (38, 39). Inhibition of ANO1expression significantly reduces the proliferation, metasta-sis, and invasion of prostate cancer and HNSCC cells.Moreover, in HNSCC, ANO1 promotes anchorage-inde-pendent growth by increasing extracellular signal–regulatedkinase (ERK)1/2 activation and cyclin D1 induction (40).However, the role of ANO1 in ESCC remains unclear. It willbe important to investigate the mechanisms underlying theinvolvement of ANO1 in esophageal carcinogenesis.

Overall, our study identified multiple copy number-altered chromosome regions and differentially expressedgenes and miRNAs. These findings provide importantinsights into the molecular alterations associated withESCC. Further studies should be conducted to explore thepossible tumorigenic roles of these candidate genes.

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

Authors' ContributionsConception and design: Z.-Z. Shi, Y. Zhang, D.-C. Lin, M.-R. WangDevelopment of methodology: Z.-Z. Shi, L. Shang, Y.-Y. JiangAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): Z.-Z. Shi, L. Shang, Y.-Y. Jiang, J.-J. Hao, D.-C. Lin,S.-G. Liu, T. GongAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): Z.-Z. Shi, L. Shang, J.-J. Hao, Y. Zhang, S.-G. Liu, B.-S. Wang, M.-R. WangWriting, review, and/or revisionof themanuscript:Z.-Z. Shi,M.-R.WangAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): Y.-Y. Jiang, T.-T. Zhang, B.-S. Wang,T. GongStudy supervision: Q.-M. Zhan, M.-R. Wang

AcknowledgmentsThe authors thank Kai-Tai Zhang (Department of Etiology and Carcino-

genesis, Peking Union Medical College, Beijing, China) for help in arrayCGH experiment.

Grant SupportThis study was funded by the Chinese Hi-Tech R&D Program Grant

(2012AA02A503 and 2012AA020206) and National Science Fund(30971482, 81021061).

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

Received December 8, 2012; revised July 25, 2013; accepted August 19,2013; published OnlineFirst September 5, 2013.

Shi et al.





References1. ParkinDM,BrayF, Ferlay J, Pisani P.Global cancer statistics, 2002.CA

Cancer J Clin 2005;55:74–108.2. HeYT,Hou J, Chen ZF,QiaoCY, SongGH,Meng FS, et al. Decrease in

the esophageal cancer incidence rate in mountainous but not levelparts of Cixian County, China, over 29 years. Asian Pac J Cancer Prev2005;6:510–4.

3. Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics,2009. CA Cancer J Clin 2009;59:225–49.

4. Ma X, Chen K, Huang S, Zhang X, Adegboyega PA, Evers BM, et al.Frequent activation of the hedgehog pathway in advanced gastricadenocarcinomas. Carcinogenesis 2005;26:1698–705.

5. Yen CC, Chen YJ, Chen JT, Hsia JY, Chen PM, Liu JH, et al. Com-parative genomic hybridization of esophageal squamous cell carci-noma: correlations between chromosomal aberrations and diseaseprogression/prognosis. Cancer 2001;92:2769–77.

6. Hirasaki S, Noguchi T, Mimori K, Onuki J, Morita K, Inoue H, et al. BACclones related to prognosis in patients with esophageal squamouscarcinoma: an array comparative genomic hybridization study.Oncologist 2007;12:406–17.

7. Ueno T, Tangoku A, Yoshino S, Abe T, Toshimitsu H, Furuya T, et al.Gain of 5p15 detected by comparative genomic hybridization as anindependent marker of poor prognosis in patients with esophagealsquamous cell carcinoma. Clin Cancer Res 2002;8:526–33.

8. Shi ZZ, Liang JW, Zhan T, Wang BS, Lin DC, Liu SG, et al. Genomicalterations with impact on survival in esophageal squamous cellcarcinoma identified by array comparative genomic hybridization.Genes Chromosomes Cancer 2011;50:518–26.

9. Wang G, Brennan C, RookM,Wolfe JL, Leo C, Chin L, et al. Balanced-PCR amplification allows unbiased identification of genomic copychanges in minute cell and tissue samples. Nucleic Acids Res2004;32:e76.

10. SuH,HuN, YangHH,WangC, TakikitaM,WangQH, et al. Global geneexpression profiling and validation in esophageal squamous cell car-cinoma and its association with clinical phenotypes. Clin Cancer Res2011;17:2955–66.

11. Zhang X, Lin P, Zhu ZH, Long H, Wen J, Yang H, et al. Expressionprofiles of early esophageal squamous cell carcinoma by cDNAmicro-array. Cancer Genet Cytogenet 2009;194:23–9.

12. Ogawa R, Ishiguro H, Kuwabara Y, Kimura M,Mitsui A, Katada T, et al.Expression profiling of micro-RNAs in human esophageal squamouscell carcinoma using RT-PCR. Med Mol Morphol 2009;42:102–9.

13. Kimura S, Naganuma S, Susuki D, Hirono Y, Yamaguchi A, Fujieda S,et al. Expression of microRNAs in squamous cell carcinoma of humanhead and neck and the esophagus: miR-205 and miR-21 are specificmarkers for HNSCC and ESCC. Oncol Rep 2010;23:1625–33.

14. ShinomiyaT,Mori T,AriyamaY,SakabeT, FukudaY,MurakamiY, et al.Comparative genomic hybridization of squamouscell carcinomaof theesophagus: the possible involvement of the DPI gene in the 13q34amplicon. Genes Chromosomes Cancer 1999;24:337–44.

15. Pack SD, Karkera JD, Zhuang Z, Pak ED, Balan KV, Hwu P, et al.Molecular cytogenetic fingerprinting of esophageal squamous cellcarcinoma by comparative genomic hybridization reveals a consistentpattern of chromosomal alterations. Genes Chromosomes Cancer1999;25:160–8.

16. Tada K, Oka M, Hayashi H, Tangoku A, Oga A, Sasaki K. Cytogeneticanalysis of esophageal squamous cell carcinoma cell lines by com-parative genomichybridization: relationship of cytogenetic aberrationsto in vitro cell growth. Cancer Genet Cytogenet 2000;117:108–12.

17. Noguchi T, Kimura Y, Takeno S, Chujo M, Uchida Y, Mueller W, et al.Chromosomal imbalance in esophageal squamous cell carcinoma: 3qgain correlateswith tumor progressionbut not prognostic significance.Oncol Rep 2003;10:1393–400.

18. Ishizuka T, Tanabe C, Sakamoto H, Aoyagi K, MaekawaM, MatsukuraN, et al. Gene amplification profiling of esophageal squamous cellcarcinomas by DNA array CGH. Biochem Biophys Res Commun2002;296:152–5.

19. Kwong D, Lam A, Guan X, Law S, Tai A, Wong J, et al. Chromosomalaberrations in esophageal squamous cell carcinoma among Chinese:

gain of 12p predicts poor prognosis after surgery. Hum Pathol2004;35:309–16.

20. Qin YR,Wang LD, Fan ZM, Kwong D, Guan XY. Comparative genomichybridization analysis of genetic aberrations associated with devel-opment of esophageal squamous cell carcinoma in Henan, China.World J Gastroenterol 2008;14:1828–35.

21. Sakai N, Kajiyama Y, Iwanuma Y, Tomita N, Amano T, Isayama F, et al.Study of abnormal chromosome regions in esophageal squamous cellcarcinoma by comparative genomic hybridization: relationship oflymph node metastasis and distant metastasis to selected abnormalregions. Dis Esophagus 2010;23:415–21.

22. Kawakubo H, Ozawa S, Ando N, Kitagawa Y, Mukai M, Ueda M, et al.Alterations of p53, cyclin D1, and pRB expression in the carcinogen-esis of esophageal squamous cell carcinoma. Oncol Rep 2005;14:1453–9.

23. Bass AJ, Watanabe H, Mermel CH, Yu S, Perner S, Verhaak RG, et al.SOX2 is an amplified lineage-survival oncogene in lung and esoph-ageal squamous cell carcinomas. Nat Genet 2009;41:1238–42.

24. Kim SG, Hong SJ, Kwon KW, Jung SW, Kim WY, Jung IS, et al. [Theexpression of p53, p16, cyclin D1 in esophageal squamous cellcarcinoma and esophageal dysplasia]. Korean J Gastroenterol2006;48:269–76.

25. Wang GQ, Abnet CC, Shen Q, Lewin KJ, Sun XD, Roth MJ, et al.Histological precursors of oesophageal squamous cell carcinoma:results from a 13 year prospective follow up study in a high riskpopulation. Gut 2005;54:187–92.

26. Yang L, Wang LS, Chen XL, Gatalica Z, Qiu S, Liu Z, et al. Hedgehogsignaling activation in the development of squamous cell carcinomaand adenocarcinoma of esophagus. Int J Biochem Mol Biol 2012;3:46–57.

27. Adams L, Roth MJ, Abnet CC, Dawsey SP, Qiao YL, Wang GQ, et al.Promoter methylation in cytology specimens as an early detectionmarker for esophageal squamous dysplasia and early esophagealsquamous cell carcinoma. Cancer Prev Res 2008;1:357–61.

28. Hao JJ, Shi ZZ, Zhao ZX, Zhang Y, Gong T, Li CX, et al. Character-ization of genetic rearrangements in esophageal squamous carcinomacell lines by a combination of M-FISH and array-CGH: further confir-mation of some split genomic regions in primary tumors. BMCCancer2012;12:367.

29. Katoh H, Shibata T, Kokubu A, Ojima H, Loukopoulos P, Kanai Y, et al.Genetic profile of hepatocellular carcinoma revealed by array-basedcomparative genomic hybridization: identification of genetic indicatorsto predict patient outcome. J Hepatol 2005;43:863–74.

30. Ying J, Shan L, Li J, Zhong L, Xue L, Zhao H, et al. Genome-widescreening for genetic alterations in esophageal cancer by aCGHidentifies 11q13 amplification oncogenes associated with nodalmetastasis. PLoS ONE 2012;7:e39797.

31. Baudis M, Cleary ML. Progenetix.net: an online repository for molec-ular cytogenetic aberration data. Bioinformatics 2001;17:1228–9.

32. Baudis M. Progenetix oncogenomic online resource 2012 [cited May31, 2013]. Available from: www.progenetix.net.

33. Sawey ET, Chanrion M, Cai C, Wu G, Zhang J, Zender L, et al.Identification of a therapeutic strategy targeting amplified FGF19 inliver cancer by oncogenomic screening. Cancer Cell 2011;19:347–58.

34. Brown LA, Kalloger SE, Miller MA, Shih Ie M, McKinney SE, Santos JL,et al. Amplification of 11q13 in ovarian carcinoma. Genes Chromo-somes Cancer 2008;47:481–9.

35. Kirkegaard T, Nielsen KV, Jensen LB, Campbell FM, Muller S, ToveySM, et al. Genetic alterations of CCND1 and EMSY in breast cancers.Histopathology 2008;52:698–705.

36. Hu N, Wang C, Ng D, Clifford R, Yang HH, Tang ZZ, et al. Genomiccharacterization of esophageal squamous cell carcinoma from a high-risk population in China. Cancer Res 2009;69:5908–17.

37. Sugimoto T, Seki N, Shimizu S, Kikkawa N, Tsukada J, Shimada H,et al. The galanin signaling cascade is a candidate pathway regulatingoncogenesis in human squamous cell carcinoma. Genes Chromo-somes Cancer 2009;48:132–42.






38. Liu W, Lu M, Liu B, Huang Y, Wang K. Inhibition of Ca(2þ)-activated Cl(-)

channel ANO1/TMEM16A expression suppresses tumor growth and in-vasiveness in human prostate carcinoma. Cancer Lett 2012;326:41–51.

39. Ayoub C, Wasylyk C, Li Y, Thomas E, Marisa L, Robe A, et al. ANO1amplification and expression in HNSCC with a high propensity for

future distant metastasis and its functions in HNSCC cell lines. Br JCancer 2010;103:715–26.

40. Duvvuri U, Shiwarski DJ, Xiao D, Bertrand C, Huang X, Edinger RS,et al. TMEM16A induces MAPK and contributes directly to tumori-genesis and cancer progression. Cancer Res 2012;72:3270–81.

Shi et al.





2013;19:5867-5878. Published OnlineFirst September 5, 2013.Clin Cancer Res Zhi-Zhou Shi, Li Shang, Yan-Yi Jiang, et al. Array Comparative Genomic HybridizationEsophageal Dysplasia and Squamous Cell Carcinoma Detected By Consistent and Differential Genetic Aberrations between

Updated version

10.1158/1078-0432.CCR-12-3753doi:

Access the most recent version of this article at:

Cited articles

http://clincancerres.aacrjournals.org/content/19/21/5867.full#ref-list-1

This article cites 39 articles, 7 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/19/21/5867.full#related-urls

This article has been cited by 6 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/19/21/5867To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-12-3753

http://clincancerres.aacrjournals.org/content/19/21/5867.full#ref-list-1

http://clincancerres.aacrjournals.org/content/19/21/5867.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/19/21/5867


Documents

Consistent and Differential Genetic Aberrations between ...clincancerres.aacrjournals.org/content/clincanres/19/21/5867.full.pdfPurpose: Our aim was to identify frequent genomic aberrations