Downregulation of MicroRNA-644a Promotes...

Biology of Human Tumors

Downregulation of MicroRNA-644a PromotesEsophageal Squamous Cell CarcinomaAggressiveness and Stem Cell–like Phenotypevia Dysregulation of PITX2Jia-Xing Zhang1,2,3, Zhen-Hua Chen4, Yi Xu2, Jie-Wei Chen1,3,5, Hui-Wen Weng2,Miao Yun2, Zou-San Zheng2, Cui Chen2, Bing-Li Wu6, En-Min Li6, Jian-Hua Fu1,3,Sheng Ye2, and Dan Xie1,3,5

Abstract

Purpose:We previously reported the oncogenic role of paired-like homeodomain 2 (PITX2) in esophageal squamous cell car-cinoma (ESCC). In this study, we aimed to identify the miRNAregulators of PITX2 and the mechanism underlying the patho-genesis of ESCC.

Experimental Design: Using miRNA profiling and bioinfor-matics analyses, we identifiedmiR-644a as a negativemediator ofPITX2 in ESCC. A series of in vivo and in vitro assays wereperformed to confirm the effect of miR-644a on PITX2-mediatedESCC malignancy.

Results: ESCC cells and tissues expressed less miR-644a thannormal epithelial controls. In patient samples, lower expressionof miR-644a in ESCC tissues was significantly correlated withtumor recurrence and/or metastasis, such that miR-644a, PITX2,and the combination of the two were independent prognosticindicators for ESCC patient's survival (P < 0.05). Gain- and loss-

of-function studies demonstrated that miR-644a inhibited ESCCcell growth,migration, and invasion in vitro and suppressed tumorgrowth andmetastasis in vivo. In addition, miR-644a dramaticallysuppressed self-renewal and stem cell–like traits in ESCC cells.Furthermore, the effect of upregulation of miR-644a was similarto that of PITX2 knockdown in ESCC cells. Mechanistic studiesrevealed that miR-644a attenuates ESCC cells' malignancy andstem cell–associated phenotype, at least partially, by inactivationof the Akt/GSK-3b/b-catenin signaling pathway through PITX2.Furthermore, promoter hypermethylation caused downregula-tion of miR-644a in ESCC.

Conclusions: Downregulation of miR-644a plays an impor-tant role in promoting both aggressiveness and stem-like traits ofESCC cells, suggesting that miR-644a may be useful as a novelprognostic biomarker or therapeutic target for the disease. ClinCancer Res; 23(1); 298–310. �2016 AACR.

IntroductionEsophageal squamous cell carcinoma(ESCC) is among themost

lethalmalignanthead andneck tumors. It presentswitha spectrum

of aberrantly aggressive phenotypes (1) and despite recentadvances in multimodality therapies, the prognosis remains poor(2, 3). Like other solid tumors, the pathogenesis of ESCC is a longprocess involving activation of oncogenes and/or inactivation oftumor suppressor genes. Recently, tremendous efforts havefocused on identifying specific molecular markers associated withthe progression of ESCC while simultaneous work has sought tobetter understand the etiology of the disease (3, 4). To date,however, promising molecular genetic alterations with clinical orprognostic significance in ESCC have remained elusive.

TheWnt/b-catenin signaling pathway is essential for embryonicdevelopment and aberrant activation has been associated with theprogression of many cancer types, including ESCC (5, 6). Paired-like homeodomain transcription factor 2 (PITX2) is a member ofthe bicoid-related homeodomain family and a downstream effec-ter of Wnt signaling, which activates the expression of target genesrequired for cell proliferation and survival (7). In addition to itsregulation by Wnt, PITX2 itself could directly activate Wnt ligandgenes, activating the respective canonical Wnt/b-catenin signalingpathway, thus contributing to cancer progression (8). Overexpres-sion of PITX2 has been frequently studied in human cancers,including thyroid, colorectal, and ovarian cancers (9–11). Werecently found that PITX2 was frequently upregulated in ESCCtissues, as compared with the normal epithelial tissue. Further-more, overexpression of PITX2was correlatedwith poor prognosis

1The State Key Laboratory of Oncology in South China, Sun Yat-Sen UniversityCancer Center, Collaborative Innovation Center for Cancer Medicine, Guangz-hou, P.R. China. 2Department of Oncology, The First Affiliated Hospital, Sun Yat-Sen University, Guangzhou, P.R. China. 3Guangdong Esophageal Cancer Insti-tute, Guangzhou, P.R. China. 4Department of Urology, The First AffiliatedHospital, Sun Yat-Sen University, Guangzhou, P.R. China. 5Department ofPathology, Sun Yat-Sen University Cancer Center, Guangzhou, P.R. China.6Department of Biochemistry and Molecular Biology, Shantou University Med-ical College, Shantou, Guangdong, P.R. China.

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

J.X. Zhang, Z.H. Chen, and Y. Xu contributed equally to this article.

Corresponding Authors: Dan Xie, The State Key Laboratory of Oncology inSouth China, Sun Yat-Sen University Cancer Center, No. 651, Dongfeng RoadEast, Guangzhou 510060, P.R. China. Phone: 8620-8734-3193; Fax: 8620-8734-3170; E-mail: [email protected]; and Sheng Ye, Department of Oncology,The First Affiliated Hospital, Sun Yat-Sen University, No. 58, Zhongshan road II,510080 Guangzhou, P.R. China. Phone: 86-20-87755766; Fax: 86-20-87755766;E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-16-0414

�2016 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 23(1) January 1, 2017298

Research. on September 27, 2018. © 2017 American Association for Cancerclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst July 12, 2016; DOI: 10.1158/1078-0432.CCR-16-0414

http://clincancerres.aacrjournals.org/

and/or radiochemoresistance of the disease (12). Clearly, a betterunderstanding of the tumor-specific regulation of PITX2 and thetargets it regulates is crucial and has potentially important clinicalapplications. We therefore aimed to identify novel moleculescapable of regulating PITX2 at the transcriptional and post-tran-scriptional levels in ESCC.

Recently, small noncoding RNA molecules, miRNAs, haveemerged as key regulators of gene expression at the post-transla-tional level (13), and deregulation of miRNAs has been impli-cated in the development and progression of nearly all tumortypes. Multiple miRNAs have demonstrated critical roles duringthe development and/or progression of ESCC by regulatingvarious critical genes (14, 15). Given the important oncogenicrole of PITX2 in ESCC (12), we sought to determine whetherPITX2 expression is regulated by specific miRNAs, with thehypothesis that the regulatory miRNAs could also be importantin ESCC pathogenesis.

Herein, we provide evidence that miR-644a is a novel negativeregulator of PITX2 expression in ESCC, and that expression ofmiR-644a was significantly lower in ESCC that recurred or oth-erwise had a poor prognosis. The levels of miR-644a regulateimportant traits in ESCC cells, including tumorigenesis, invasion,and the stem cell–like phenotype both in vitro and in vivo.Importantly, we reveal that depleting miR-644a levels in ESCCcells potently activates the Akt/GSK-3b/b-catenin signaling path-way through directly upregulating PITX2, ultimately increasingthe malignancy of the tumor. The tumor-promoting downregula-tion of miR-644a in ESCC is caused by hypermethylation of itspromoter.

Materials and MethodsCell lines and clinical samples

Seven human ESCC cell lines (KYSE-30, TE-1, KYSE-510, KYSE-180, KYSE-140, KYSE-410, and KYSE-520) were obtained fromDSMZ, the German Resource Center for Biological Material. TheESCC line Eca109 and the normal epithelial cell line NE-1 werekindly provided by S.W. Tsao and G. Srivastava (University of

Hong Kong, Hong Kong, China). All cell lines were cultured inRPMI1640 supplemented with 10% FBS. A total of 190 formalin-fixed and paraffin-embedded (FFPE) ESCC tissue samples wereobtained from the Department of Pathology of Sun Yat-SenUniversity Cancer Center (Guangzhou, China) between January2007 and December 2008. The clinicopathologic characteristicsof these patients are summarized in Supplementary Table S1. Anadditional panel of 20 fresh ESCC tissues and matched adjacentnontumor esophageal tissues were collected from Sun Yat-SenUniversity Cancer Center between January 2013 and June 2013,and stored in liquid nitrogen until further use. All ESCC speci-mens included in this study were obtained from patients withstage I–III ESCC disease during surgical resection. The samplescontained matched tumors (percentage of tumor cells � 70%)and corresponding normal mucosal tissue (>5 cm laterally fromthe edge of the cancerous region). None of the recruited patientsreceived any preoperative treatment. Patients were monitoredonce every 3 months for the first 2 years after surgical resection,once every 6 months during the third and fourth years, and oncea year after the fifth year postoperatively. Diagnostic examina-tions included esophagography, computed tomography, chestX-ray, abdominal ultrasonography, and bone scanning as re-quired, to detect tumor recurrence and/or metastasis. Writteninformed consent was obtained from all patients before thestudy. This study was approved by Ethical Committee of SunYat-sen University Cancer Center (Guangzhou, China).

RNA isolation and quantitative real-time PCRTotal RNA was extracted from ESCC tissues and cell lines with

use of TRIzol reagent (Invitrogen). Real-time PCR (RT-PCR) wascarried out using SYBR Green SuperMix (Roche) and ABI7900HTFast Real-Time PCR system (Applied Biosystems). Either a-tubu-lin or U6 was used as an internal control. TaqMan probes wereused to detect miR-644a, U6, PITX2, ITCH, and a-tubulin (Gen-eCopoeia). The remaining qPCR primers are listed in Supple-mentary Table S2.

Vector construction and oligonucleotide transfectionThe miR-644a expression vector (HmiR0349-MR03), control

vector for miR-644a (CmiR0001-MR03), PITX2 coding sequenceexpression vector (EX-P0098-Lv105), and control vector forPITX2 (EX-NEG-Lv105) were purchased from GeneCopoeia.The target sequence of PITX2 used to construct a lentiviralshRNA was 50-GCCGACTCCTCCGTATGTTTA-30. Cells trans-fected with empty vector were used as controls. miR-644a mimicand antagomir-miR-644a were synthesized by Genepharma.Oligonucleotide transfection was performed using Lipofecta-mine 2000.

Lentivirus packaging and transductionVectors were packaged in 293FT cells using ViraPower Mix

(GeneCopoeia). After culturing for 48 hours, lentiviral particlesin the supernatant were harvested and filtered by centrifugationat 500 � g for 10 minutes, and then transfected into ESCC cells.The cells were cultured under puromycin (2 mg/mL) selection for2 weeks, at which point real-time PCR was used to determinethe level of miR-644a. Cell lines stably expressing miR-644a ornegative control (NC) vector were designated as KYSE-1400/Eca109-Lv-miR-644a and KYSE-1400/Eca109-Lv-miR-NC cells,respectively.

Translational Relevance

Recently, we demonstrated that paired-like homeodomaintranscription factor 2 (PITX2) functions as an oncogenicprotein in ESCC. Here, we provide evidence that miR-644ais a negative regulator of PITX2 in ESCC. miR-644a wasfrequently downregulated in ESCC cells and tissues, and inESCC patients, low levels of the miRNA were correlated withadvanced clinical stage, tumor recurrence, and/or poor prog-nosis. Our in vitro and in vivo studies showed that ectopicoverexpression of miR-644a in ESCC cells substantially sup-pressed cell growth, aggressiveness, and stem cell–like featuresby repressing PITX2expressionwhich leads to inhibitionof theAkt/GSK-3b/b-catenin signaling pathway. We further showedthat hypermethylation of the miR-644a promoter resulted indecreased expression in ESCC. These data collectively suggestthat miR-644a could be employed as a novel prognosticmarker, and that targeting miR-644a may represent a newtherapeutic strategy to improve the treatment and survival ofESCC patients.

miR-644a Functions as a Tumor Suppressor in ESCC

www.aacrjournals.org Clin Cancer Res; 23(1) January 1, 2017 299




MTT assayCell viability was measured by a 3-(4, 5-dimethylthiazol-2-yl)-

2, 5-diphenyl tetrazolium bromide (MTT) assay (Sigma). Briefly,cells were seeded in 96-well plates and cultured. Cell viability wasexamined by following standard procedures. Experiments wereperformed in triplicate.

Xenograft tumor growth and tumor formation assayTo assay xenograft tumor growth, miR-644a–expressing cells

(2 � 106) or empty vector–transfected control cells were subcu-taneously injected into the inguinal folds of 4-week-old BALB/cnudemice. For the tumor formation assay, three dosages (2�105,2 � 104, or 2 � 103) of the indicated cells in Matrigel (finalconcentration 25%) were inoculated subcutaneously into theinguinal folds of BALB/c nude mice. The mice were monitoreddaily for palpable tumor formation and tumors were measuredusing Vernier calipers, weighed, and photographed. All animalexperiments were conducted according to the standard institu-tional guidelines of Sun Yat-Sen University Cancer Center(Guangzhou, China).

In vivo metastasis assaysTen 4-week-old BALB/c nude mice in each experimental group

were injected with KYSE-140/Eca109-miR-644a or KYSE-140/Eca109-miR-control cells, respectively. Briefly, 2 � 105 cells wereinjected intravenously through the tail vein into each mouse in alaminar flow cabinet. Six weeks after injection, the mice weresacrificed and examined.

Immunohistochemical stainingIn brief, paraffin-embedded sections were deparaffinized and

incubated in retrieval buffer solution for antigen retrieval.Protein expression was visualized using a Dako Real EnvisionKit (K5007, Dako) after staining with the primary antibody.Staining intensity was scored manually by two independentexperienced pathologists as: 0 ¼ no staining, 1¼weak staining,2 ¼ moderate staining, and 3 ¼ strong staining. Tumor cells infive fields were selected randomly and scored on the basis of thepercentage of positively stained cells (0%–100%). The finalIHC score was calculated by multiplying the intensity score bythe percentage of positive cells.

Western blot, immunoprecipitation, and chromatinimmunoprecipitation assays

For Western blots, total cellular protein was extracted fromtissues or cells and separated by SDS-PAGE. Nuclear extractswere prepared using the Nuclear Extraction Kit (Active Motif),according to the manufacturer's instructions. Immunoprecipi-tation (IPs) were carried out using protein G agarose (Milli-pore). For ChIP analysis, ESCC cells (4� 107) were prepared byusing the chromatin immunoprecipitation (ChIP) Assay Kit(Cell Signaling Technology) according to the manufacturer'sinstructions. The following primary antibodies were used: anti-PITX2 (Abcam); anti-a-tubulin, anti-GSK-3b, anti-phospho-GSK-3b (p-GSK-3b), anti-Akt, anti-phospho-Akt (p-Akt), anti-b-catenin, anti-Lef-1, anti-Axin-2, anti-Survivin, anti-acetylated-lysine (Cell Signaling Technology); anti-active-b-catenin (Milli-pore); anti-c-myc, anti-cyclin D1, anti-Histone H1 (Santa CruzBiotechnology).

5-Aza-20-deoxycytidine treatmentESCC cells were treated with 5-aza-20-deoxycytidine (5-aza-dC;

50 mmol/L and 100 mmol/L; Sigma-Aldrich) for 72 hours with achange of culture medium every 24 hours.

Promoter methylation analysisGenomic DNA was extracted and subjected to bisulfite treat-

ment using the Epitect Bisulfite kit (Qiagen) according to themanufacturer's instructions. Bisulfite-treated DNA was then ana-lyzed by bisulfite genomic sequencing (BGS) based on the fol-lowing primers: forward, 50-TGGAGTTGAGGAAAATTGG-30; andreverse, 50- ATTCTCATCCGAACTCCC-30 (5 clones were pickedfor each sample).

Statistical analysesStatistical analyses were performed using SPSS software

(SPSS Standard version 16.0, SPSS Inc). Bivariate correlationsbetween study variables were calculated by Pearson correlationcoefficients. Differences between variables were analyzed by c2

or Fisher exact tests. Survival curves were plotted using theKaplan–Meier method and compared with log-rank tests.Multivariate survival analysis was performed for all parametersfound to be significant in univariate analysis using a Coxregression model. Comparisons between groups for statisticalsignificance were performed with a two-tailed Student t test.Data are presented as the mean � SD. P values <0.05 wereconsidered significant.

Additional methodsDetailed methods on colony formation assay, wound-healing

and invasion assays, luciferase reporter assay, immunofluorescent(IF) staining, sphere-forming assay, and flow cytometric analysisare reported in the Supplementary Data.

ResultsmiR-644a is a negative regulator of PITX2 expression in ESCC

To investigate the potential miRNA regulators of PITX2, wefirst performed bioinformatic analysis by using the miRANDApredicting program to compare the predicted regulators withmiRNAs downregulated in ESCC tumor tissues in publishedmiRNA profiles (NCBI/GEO/GSE23142; ref. 16). The twoscreening algorithms overlapped on only one miRNA, namely,miR-644a (Fig. 1A). We, therefore, focused our subsequentefforts on miR-644a as a putative regulator of PITX2 in ESCC.We examined the levels of miR-644a and PITX2 expression in20 pairs of fresh ESCC and nontumor esophageal tissues, aswell as in 8 ESCC cell lines and a normal epithelial cell line,NE-1. The results showed that levels of miR-644a were signif-icantly downregulated in ESCC tissues and cells as comparedwith the levels in nontumor tissues and NE-1 cells (Fig. 1B, iand C, i). Overall, there was a significant, inverse correlationbetween the levels of miR-644a and PITX2 mRNA either inESCC tissues (r ¼ �0.822, P < 0.001; Fig. 1B, ii and iii) or inESCC cell lines (r ¼ �0.960, P < 0.001; Fig. 1C, ii and iii).

Clinical significance and prognostic values of miR-644a andPITX2 in ESCC patients

We next examined the levels of miR-644a and PITX2 pro-tein expression to determine their potential clinical andprognostic significance in a large cohort of 190 ESCC FFPE

Zhang et al.

Clin Cancer Res; 23(1) January 1, 2017 Clinical Cancer Research300




Figure 1.

miR-644a is identified as a negative regulator of PITX2 expression in ESCC and correlates with poor prognosis of ESCC patients. A, A screen for negativeregulatory miRNAs of PITX2 in ESCC. Method #1 used the prediction algorithm of miRANDA analysis to identify 65 miRNAs, which could potentiallysuppress PITX2. Method #2 used miRNA profiling to identify 40 miRNAs that were downregulated significantly in ESCC (NCBI/GEO/GSE23142). A combinationof the two screening methods showed that miR-644a is a potential inhibitor for PITX2 and that it is downregulated in ESCC. B, The correlation betweenlevels of miR-644a and PITX2 mRNA in 20 fresh ESCC tissues: i, comparing differences in the levels of miR-644a between tumor and correspondingnontumor tissues; ii, miR-644a levels were significantly lower in ESCC tissues with high PITX2 expression compared with those that had low PITX2 expression;iii, miR-644a levels were inversely correlated with PITX2 mRNA in ESCC tissues. C, The correlation between the levels of miR-644a and PITX2 in ESCCcell lines: i, comparison of the expression levels of miR-644a between ESCC cell lines and nontransformed epithelial cell line NE1; ii, the expression of PITX2protein (top) and mRNA (bottom) in ESCC cell lines; iii, miR-644a levels were inversely correlated with mRNA levels of PITX2 in ESCC cell lines. D, Theexpression pattern of PITX2 in 190 ESCC FFPE samples: i, the expression levels of miR-644a and PITX2 protein were inversely correlated in 190 ESCCsamples. ii and iii, the differences of miR-644a levels between ESCC tissues from recurrent and nonrecurrent (ii), and distant metastatic and nonmetastatictumors (iii). E, Prognostic significance of miR-644a and PITX2 in 190 ESCC patients assessed by Kaplan–Meier analyses: i–vi, patients with higher miR-644a had(i) better overall survival (OS) and (ii) lower risk of tumor recurrence. Patients with higher PITX2 levels had (iii) poorer OS and (iv) higher risk of tumorrecurrence. Patients in subgroup I had the (v) longest OS and (vi) lowest risk of tumor recurrence among the four subgroups divided according tocombinations of miR-644a and PITX2: I, high miR-644a/low PITX2; II, high miR-644a/low PITX2; III, low miR-644a/high PITX2; IV, low miR-644a/highPITX2. For each cohort, the different subgroups were classified according to cut-off values of miR-644a and PITX2, defined by the median of thecohort. � , P < 0.05; �� , P < 0.01.






tissues. We found a significant inverse correlation betweenmiR-644a levels and PITX2 protein expression (r ¼ �0.669,P < 0.001; Fig. 1D, i). Furthermore, the level of miR-644a wassignificantly lower in ESCC patients with tumor recurrenceand/or distant metastasis after surgery, compared with patientswithout tumor recurrence and/or distant metastasis (P < 0.001;

Fig. 1D ii and iii). Expression level of miR-644a in ESCCswas negatively associated with tumor size (P ¼ 0.031), Tstatus (P ¼ 0.004), N status (P < 0.001), and clinical stage(P < 0.001; Supplementary Table S2). Moreover, ESCC pati-ents with high levels of miR-644a exhibited better overallsurvival (OS) and lower tumor recurrence rates than those

Figure 2.

Effects of miR-644a on ESCC cell growth, migration, and invasion in vitro and in vivo. A–D, Enforced overexpression of miR-644a in ESCC cellssignificantly inhibited cell proliferation (A), colony formation (B), migration (C) and invasion (D) abilities, as compared with that of control cells. E,miR-644a suppressed the in vivo tumor growth of ESCC cells: i, overexpression of miR-644a in ESCC cells suppressed tumor growth in a subcutaneousimplantation mouse model; ii and iii, tumor volume (ii) and tumor weight (iii) in ESCC xenografts were monitored; iv, H&E staining showing tumorboundary of ESCC xenografts. Arrow indicates extensive branch-like growth pattern spreading into surrounding tissue in tumors formed by vectorcontrol cells. F, miR-644a inhibited in vivo metastasis of ESCC: i, representative metastatic lesions stained by H&E in the lungs of mice 6 weeks after tailvein injection of indicated cells. "M" denotes the metastatic colonies in the lung, while "N" means the normal lung tissues; ii, quantitative analysis ofnumber of lung metastatic colonies. �� , P < 0.01; �, P < 0.05.

Zhang et al.





with low miR-644a (Fig. 1E, i and ii). Higher PITX2 expressionin ESCCs was positively associated with tumor size (P ¼0.031), N status (P ¼ 0.006), and clinical stage (P ¼ 0.009,

Supplementary Table S2) and the patients showed poorer OSand higher tumor recurrence rates than those with low PITX2(Fig. 1E, iii and iv).

Figure 3.

The levels of PITX2 influence the in vitro effects of miR-644a in ESCC cells. A–D, The inhibited cell growth (A), colony formation (B), migration (C), andinvasion (D) abilities in miR-644a–overexpressing ESCC cells were significantly rescued after the enforced overexpression of PITX2, while knockdownof PITX2 by specific shRNA had effects similar to those of miR-644a upregulation in ESCC cells (A–D). E, The expression levels of activated b-cateninand nuclear b-catenin, as well as the downstream targets examined by Western blot analysis in the indicated ESCC cell lines. F, The subcellularlocalization of b-catenin in the indicated ESCC cells examined by IF staining. G, The ChIP assays were conducted using chromatin isolated from theindicated ESCC cells. The following PCR primers for ChIP assays were used: Lef-1, 50-cctgaagggtgggaaaaa-30 , 50-cgggccgaggaaccaggac-30 ; and c-myc,50-gtatacgtggcaatgcgtt-30 , 50-tgagtataaatcatcgcag-30 . Normal IgG was used as a control, and 1% of the total cell lysates were subjected to PCR analysisbefore IP and used as input controls. H, b-catenin-TCF/LEF transcriptional activity determined as a function of miR-644a expression using the TOP/FOPflashassay in the indicated ESCC cell lines. Cells were transfected with TOPflash or FOPflash and Renilla plasmids, and assayed 48 hours later. Luciferasereporter activity was normalized to Renilla activity. � , P < 0.05; �� , P < 0.01.






Zhang et al.





In univariate analyses, miR-644a and PITX2 levels and TNMstage were significantly associated with OS and time to recur-rence (TTR) in patients with ESCC (Supplementary Table S3).Further multivariate analysis revealed that miR-644a, PITX2,and TNM stage were independent prognostic indicators forboth OS and TTR (Supplementary Table S3). We next dividedthe patients into four groups based on their levels of miR-644aand PITX2 expression. ESCC patients with high miR-644aand low PITX2 levels had the best OS and lowest frequencyof recurrence. In contrast, the ESCC patients with low miR-644aand high PITX2 levels had the lowest OS and highest frequencyof recurrence (P < 0.001; Fig. 1E, v and vi; SupplementaryTable S3). Together, these data revealed that miR-644a andPITX2 are independent prognostic indicators in patients withESCC and the combination of both factors has the strongestprognostic value.

The levels of miR-644a influence ESCC cell growth, colonyformation, migration, and invasion abilities in vitro

Subsequently, we investigated the role of miR-644a inESCC tumorigenesis and progression in vitro. Overexpressionof miR-644a in KYSE-140 and Eca109 ESCC cells sub-stantially reduced cell growth and colony formation andlargely inhibited both cell migration and invasive capacities(Fig. 2A–D). Consistent with a role for miR-644a in tumor-igenesis, knockdown of miR-644a with anti-miR-644a in theKYSE-410 ESCC cell line significantly increased cell growth,colony formation, migration, and invasion (SupplementaryFig. S1A–S1D).

Overexpression of miR-644a inhibits tumor growth andmetastasis of ESCC cells in vivo

We further studied the in vivo impact of miR-644a on ESCCcell growth and metastasis by injecting KYSE-140 or Eca109cells containing either a control or miR-644a-overexpressionvector into BALB/c nude mice, either subcutaneously or via thetail vein. We observed that overexpression of miR-644a signif-icantly inhibited tumor growth in vivo (Fig. 2E, i–iii). Hema-toxylin and eosin (H&E) staining revealed that tumors formedfrom control cells exhibited an extensive branch-like growthpattern that spread into the surrounding tissue, while miR-644a–overexpressing ESCC cells formed oval-shaped tumorswith smooth margins and a noninvasive front. This suggeststhat the invasive behavior of ESCC cells could be suppressed bymiR-644a in vivo (Fig. 2E, iv). In the mouse metastasis model,we did not detect tumor nodule formed in the liver of any miceexamined. However, the number and size of metastatic colo-nies in the lungs were dramatically decreased in mice injectedwith miR-644a–overexpressing cells, compared with the con-trol group (Fig. 2F).

The effect of miR-644a on ESCC cells is affected byPITX2 levels

On the basis of our in silico prediction, we cloned a fragmentof the 30UTR of PITX2 mRNA containing a site complementaryto the seed region of miR-644a (Supplementary Fig. S2A).Further studies showed that overexpression of miR-644a inESCC cells significantly suppressed the luciferase activity ofPITX2 containing wild-type 30UTR, but not that of PITX2-con-taining mutant 30UTR (Supplementary Fig. S2B). Depletion ofmiR-644a increased the luciferase activity of PITX2 (Supple-mentary Fig. S3A). Concordantly, overexpression of miR-644asubstantially decreased mRNA and protein levels of PITX2 inESCC cells (Supplementary Fig. S2C and S2D), while down-regulation of miR-644a dramatically increased their levels (Sup-plementary Fig. S3B and S3C). These results were verified byimmunohistochemical analysis of PITX2 in xenograft tumorsformed from miR-644a-overexpressing and control cells (Sup-plementary Fig. S2E).

To confirm that PITX2 is a functional target ofmiR-644a, stablemiR-644a-ESCC (KYSE-140 and EC109) cells were transfectedwith pEZ-Lv105-PITX2 encoding the entire PITX2 codingsequence but lacking the 30-UTR. As expected, overexpression ofPITX2 in miR-644a-ESCC cells abrogated miR-644a-suppressedESCC cell growth, colony formation, migration, and invasionabilities. Correspondingly, knockdown of PITX2 expression sig-nificantly inhibited ESCC cell growth, colony formation, migra-tion, and invasion ability (Fig. 3A–D), which was consistent withthe effect of miR-644a overexpression in ESCC cells. These datarevealed that the effect of miR-644a in ESCC cells is affected byPITX2 levels.

miR-644a inhibits Wnt/b-catenin pathway by targeting PITX2in ESCC cells

As previous study suggested that PITX2 crosstalks with theWnt/b-catenin pathway (7, 8), we sought to use miR-644a toprobe the role of PITX2 in Wnt/b-catenin signaling. We dem-onstrated that miR-644a overexpression in ESCC cells reducedthe pool of activated b-catenin (i.e., dephosphorylated at Ser37and Ser41), nuclear b-catenin, and the typical downstreamtarget genes of b-catenin (Fig. 3E). Furthermore, IF stainingshowed a clear redistribution of b-catenin from the nucleus tothe cytoplasm after miR-644a overexpression (Fig. 3F). Therewas also a marked decrease in binding of b-catenin to theTCF sites of both the Lef-1 and c-myc promoters, examinedby ChIP, in miR-644a–overexpressing ESCC cells (Fig. 3G). Inaddition, the TOP/FOP flash luciferase assay revealed thatreporter expression was suppressed by miR-644a overexpres-sion (Fig. 3H). Furthermore, we found that restoration ofPITX2 levels in miR-644a-overexpressing ESCC cells largelyblocked the inhibitory effect of miR-644a on the Wnt/b-cateninpathway (Fig. 3E–H). These findings indicated that miR-644a–

Figure 4.miR-644a inhibits the stemness of KYSE-140 cells by regulating PITX2 expression. A–D, Enforced overexpression of miR-644a in KYSE-140 cells substantiallydownregulated the levels of stemness-associated genes (Nanog, Oct-4, Bmi-1, Notch-1, and Smo), multiple drug-resistant transporter genes (ABCC2,ABCG2) and surface antigens associated with cancer stem cells (CD24, CD44, CD133, CD155, and CD166; A), reduced sphere-forming ability (B), andproportion of side population cells (C), and also largely increased chemosensitivity to cisplatin or radiosensitivity to IR (D). Restoration of PITX2 inmiR-644a–overexpressing KYSE-140 cells largely rescued the cells' stemness, while knockdown of PITX2 by shPITX2 decreased KYSE-140 cells' stemness(A–D). E, Tumor formation in nude mice shows reduced tumorigenicity in three groups of miR-644a–overexpressing KYSE-140 cells indicated, ascompared with that in matched control groups. � , P < 0.05; �� , P < 0.01.






mediated suppression of PITX2 expression inhibits the activa-tion of Wnt/b-catenin signaling.

Upregulation of miR-644a reduces ESCC cell stem-like traitsby targeting PITX2

It has been shown that stem cell–like properties in cancer cellscontribute to tumor recurrence and/or metastasis, and thatactivation of the Wnt/b-catenin pathway is thought to be amajor driver of the development of stem cell–like properties(17, 18). Thus, we sought to determine whether the levels ofmiR-644a affect ESCC cell stem-like characteristics. Our resultsshowed that overexpression of miR-644a in ESCC cells resultedin downregulation of many stemness-associated genes we exam-ined by qRT-PCR (Fig. 4A and Supplementary Fig. S4A). Furtherfunctional assays demonstrated that miR-644a–overexpressingESCC cells formed smaller and fewer spheres, with a reducedproportion of side population (SP) cells, and the cells were moresensitive to cisplatin and radiation, as compared with control cells(Fig. 4B–DandSupplementary Fig. S4B–S4D).However,when thelevels of PITX2 were restored in miR-644a–overexpressing ESCCcells, the inhibitory effect of miR-644a on stem-like traits waspartially blocked, whereas knockdown of PITX2 had a similareffect to miR-644a on ESCC stem-like traits (Fig. 4A–D andSupplementary Fig. S4A–S4D). These data revealed thatmiR-644a inhibits ESCC cell stem-like traits through regulationof the PITX2 gene.

In this study, we further assessed the in vivo effect of miR-644aon the tumorigenicity of ESCC cells in nude mice. Our resultsclearly showed that fewer and smaller tumors were formed afterthe injection of 2 � 105, 2 � 104, or 2 � 103 miR-644a–over-expressing ESCC cells comparedwith the tumors from injection ofcontrol cells. Moreover, the first palpable tumor in the groupinjected with miR-644a–overexpressing cells occurred later thanin the control group (Fig. 4E and Supplementary Fig. S4E).

Furthermore, we examined the levels of miR-644a and sixmarkers associated with cell stemness (ABCG2, SOX2, Oct4,Nanog, Bmi1, and CD44) by RT-PCR in the 20 fresh ESCC patientsamples. The levels of miR-644a expression were negativelycorrelated with the expressions of the 6 stemness-associatedmarkers (P < 0.05, Supplementary Fig. S5).

miR-644a inhibits Akt/GSK-3b/b-catenin pathway by targetingPITX2 in ESCC cells

Previous studies demonstrated that Akt activation inhibitsGSK-3b activity, leading to activation of Wnt/b-catenin signal-ing and the development of stem-like traits and tumor aggres-siveness (19, 20). Notably, after overexpression of miR-644a inESCC cells, the levels of PITX2, p-Akt, p-GSK-3b (inactive form)and active-b-catenin were all reduced (Fig. 5A). There was aconcomitant decrease in the cell invasion and sphere-formingabilities of the miR-644a–overexpressing cells (SupplementaryFig. S6). These malignant characteristics were substantially res-cued, by PITX2 transfection into the miR-644a–overexpressingESCC cells (Fig. 5A and Supplementary Fig. S6). These datasuggested that miR-644a could inhibit the Akt/GSK-3b/b-cateninpathway through PITX2. To probe this interaction, we used thePI3K/AKT inhibitor, LY294002, to inactivate the Akt/GSK-3b/b-catenin pathway (Fig. 5A). We observed that LY294002 rescuedboth the cell invasion and sphere-forming abilities in miR-644a–overexpressing ESCC (Supplementary Fig. S6). Moreover, knock-

down of PITX2 reduced p-Akt, p-GSK-3b and active-b-cateninlevels in ESCC cells (Fig. 5A).

In our study, we also found that the levels of miR-644a werenegatively correlated with the expressions of PITX2, p-AKT, p-GSK-3b, and nuclear b-catenin in 190 cases of FFPE ESCC tissues(P < 0.001, Fig. 5B; Supplementary Table S4). These results wereconfirmed in 20 fresh ESCC specimens, in which a negativecorrelation betweenmiR-644a levels and the expressions of PITX2(r¼�0.790, P¼ 0.007), p-AKT (r¼�0.857, P¼ 0.002), p-GSK-3b (r ¼ �0.744, P ¼ 0.014), and nuclear b-catenin (r ¼ �0.763,P ¼ 0.010, Fig. 5C) was observed. Furthermore, the expressionlevels of PITX2, p-AKT, p-GSK-3b, and nuclear b-catenin inxenograft tumors that developed from themiR-644a–overexpres-sing ESCC cells were significantly lower than those from thecontrol cells (Supplementary Fig. S7).

As the phosphorylation and activity of Akt are regulatedby its acetylation status, which regulate the downstream effec-tors (20, 21), we therefore investigated whether miR-644a isinvolved in the acetylation of Akt, GSK-3b, and b-catenin. Theseproteins were immunoprecipitated using the correspondingantibodies and their acetylation statuses were determined usinganti-acetylated-lysine antibodies. The results showed that b-cate-nin was not acetylated in either miR-644a–overexpressing orcontrol cells, whereas GSK-3b was constitutively acetylatedunder both conditions (Fig. 5D). In contrast, although Akt wasconstitutively acetylated in ESCC cells, the levels of acetylatedAkt were dramatically upregulated after overexpression ofmiR-644a; this could be largely inhibited by enforced expres-sion of PITX2 (Fig. 5D). These results revealed that miR-644arepression of PITX2 can regulate the levels of acetylated Akt inESCC cells.

miR-644a is downregulated by its promoter hypermethylationFinally, we explored themolecularmechanisms responsible for

the downregulation of miR-644a in ESCC. The miR-644a codingsequence is located in the intron of the ITCH gene, andmiR-644aexpression correlated with ITCH mRNA expression in ESCC celllines (r ¼ 0.967, P<0.001) and ESCC tissues (r ¼ 0.963, P <0.001, Fig. 6A), suggesting that miR-644a is cotranscribed withITCH mRNA. We next examined whether reduced miR-644aexpression in ESCCwas the result of promoter hypermethylation.Analysis of the ITCH promoter region using the UCSC genomebrowser (http://genome.ucsc.edu/) indicated a CpG island locat-ed between�285 bp and�56 bp relative to the transcription startsite (Fig. 6B). Furthermore, BGS PCR showed that the CpG islandswithin the ITCH promoter were hypermethylated in 2 miR-644alow–expressing ESCC cell lines (KYSE-140 and Eca109) and 3ESCC tumor samples. In contrast, the CpG islands were mostlynot methylated in the control NE-1 and miR-644a high–expres-sing ESCC cells (KYSE-410) and 3 paired normal esophagealtissues (Fig. 6C). Moreover, miR-644a expression was significant-ly increased in ESCC cell lines with low levels of miR-644a aftertreatment with the methylase inhibitor 5-aza-dC, but not in NE-1and KYSE-410 cells (Fig. 6D). These results indicated that reducedmiR-644a expression in ESCC is attributable to hypermethylationof its promoter.

DiscussionWe previously reported that PITX2 was frequently overex-

pressed in ESCCs and associated with ESCC patients'

Zhang et al.





chemoresistance and poor prognosis (12). However, the regula-tory mechanisms of PITX2 in ESCC remained elusive. In thecurrent study, we first identified that miR-644a functioned as animportant negative regulator of PITX2 in ESCC cells and tissues.Next, we provided evidence that decreased expression of miR-644a in ESCC is important in the acquisition of an aggressive and/or poor prognostic phenotype. Thus, we conclude that miR-644alevel appears to have the potential to predict certain outcomes

for ESCC patients. Specifically, miR-644a expression could beused as an additional tool in identifying those ESCC patients atincreased risk of tumor invasion and/or metastasis. These find-ings underscore a potentially important role of miR-644a down-regulation in the development and/or progression of ESCC.

The role of miR-644a in human cancers remains poorlystudied. Only one report has documented a significant asso-ciation between the levels of miR-644 and patients' survival in

Figure 5.

miR-644a exerts its functions by inhibiting the PITX2-Akt/GSK-3b/b-catenin signaling pathway. A, Western blotting shows that expression levels of p-Akt,p-GSK-3b, and active-b-catenin were significantly decreased in miR-644a–overexpressing ESCC cells, compared with control cells. Restoration ofPITX2 abrogated the decreased expression of p-Akt, p-GSK-3b, and active-b-catenin induced by miR-644a in ESCC cells, while the Akt inhibitor LY294002(LY) could effectively suppress the upregulated levels of these genes induced by PITX2, and moreover, knockdown of PITX2 by shPITX2 in ESCC cellsinhibited expression of these genes. B, The correlation between the levels of miR-644a and the PITX2-Akt/GSK-3b/b-catenin axis in 190 cases of ESCCFFPE tissues. Staining intensities for PITX2, p-Akt, p-GSK-3b, and active-b-catenin were significantly reduced in miR-644a high–expressing ESCCtissues compared with miR-644a low–expressing tissues. C, The levels of miR-644a and the expression of PITX2-Akt/GSK-3b/b-catenin axis in 20 freshESCC samples. Top, the levels of indicated proteins examined by Western blot analysis in ESCCs (H means high levels of miR-644a, L means low levelsof miR-644a). Bottom, significant negative correlations were evaluated between the levels of miR-644a and indicated proteins in ESCC samples. D,Overexpression of miR-644a enhanced Akt acetylation in ESCC cells, while restoration of PITX2 attenuated the effect of miR-644a on the acetylationof Akt. b-catenin, GSK-3b, and Akt were immunoprecipitated with the respective antibodies and probed with anti-acetylated-lysine antibodies. � , P < 0.05;�� , P < 0.01.






acute myeloid leukemias (22). In the current study, a series ofin vitro and in vivo assays were employed to investigate therole of miR-644a in regulating the characteristic malignantphenotype of ESCC. Our results clearly demonstrated thatmiR-644a could substantially inhibit ESCC cells growth and/oraggressiveness by inactivating the Wnt/b-catenin signalingthrough repressing PITX2. Various studies have documentedthat the Wnt/b-catenin signaling pathway is constitutivelyactivated in many types of human cancer, and plays an impor-tant role in triggering aggressive cellular features that constitutea malignant phenotype (23–25). It was reported that PITX2 andthe Wnt/b-catenin pathway exert positive-feedback regulation,while aberrant activation of Wnt/b-catenin signaling is stronglyinvolved in ESCC tumorigenesis and/or progression (8, 26).These results collectively suggest that the Wnt/b-catenin axis isone of the critical downstream pathways activated by PITX2accumulation in ESCC malignancy that we now propose ismiR-644a-mediated.

To date, however, the molecular mechanism by which PITX2regulates Wnt/b-catenin signaling have not been elucidated.PITX2 was previously shown to interact directly with the Wntgene promoters, thus inducing the activation of the canonicalWNT/b-catenin pathways in human ovarian cancer cells (8).Notably, PITX2 may function as a transcription factor and is thusexpected to be localized in the cell nucleus. However, consistentwith our previous study (12), we found that PITX2 was predom-inantly located in the cytoplasm of ESCC cells, which suggeststhat, in addition to its potential role as a transcription factor, thefunction and localization of PITX2 may be tumor tissue specific.In ESCC cells, PITX2 activates the WNT/b-catenin signalingthroughother unclearmechanisms. It is known that Akt activation

can inhibit GSK-3b activity, leading to nuclear b-catenin accumu-lation and ultimately triggering tumor progression (27). Wetherefore investigate whether PITX2 can activate Akt/GSK-3bsignaling, thus affecting theWnt/b-catenin pathway in ESCC cells.Our results demonstrate that in ESCC, the Akt/GSK-3b pathway isinvolved in themiR-644a–mediated regulation of theWnt/b-cate-nin signaling. Ectopic overexpression of PITX2 in ESCC cells couldnot only reverse miR-644a–inhibited Akt/GSK-3b/b-cateninactivity, but also substantially suppressedmiR-644a–induced Aktacetylation. These results provided evidence that in ESCC, PITX2activates the Wnt/b-catenin pathway through the Akt/GSK-3bsignaling cascade to exert its downstream oncogenic effect. Clear-ly, further studies are needed to clarify the precise molecularmechanisms governing how PITX2 regulates the activity of Akt/GSK-3b/b-catenin.

Recently, the cancer stem cell (CSC) hypothesis has attractedsignificant attention, and the development of stem cell–likeproperties in cancer cells has been recognized as a hallmark ofdisease progression, while the unlimited self-renewal capacity ofCSCs is thought to contribute to tumor progression and recur-rence (28, 29). Accumulating evidence supports the involvementof some miRNAs in the regulation of CSC function (30, 31); inparticular, reduced let-7 expression was associated with tumor-initiating cells after the initiation of chemotherapy in breastcancer, and is thought to result in the maintenance of theirundifferentiated status and proliferative potential (32). Similarly,miR-429 contributes to hepatocyte self-renewal, malignant pro-liferation, tumorigenicity, and chemoresistance, making it anattractive potential target to inactivate tumor-initiating cells inthe liver (33). We show that miR-644a functions as an upstreamactivator of the Wnt/b-catenin pathway in ESCC, a pathway well

Figure 6.

Downregulation of miR-644a in ESCC is caused by promoter hypermethylation. A, miR-644a expression was correlated with ITCH mRNA expression ineight ESCC cell lines (left) and in 20 fresh ESCC tissues (right). B, Schematic illustration of position of the miR-644a stem loop within the ITCHgenomic sequence. C, Bisulfite sequencing analysis of NE-1 cell line, 3 indicated ESCC lines, and 3 paired ESCC and nontumor tissue samples. Five clonesof PCR products from each sample of bisulfite-treated DNA were sequenced. Filled and open circles indicate methylation and nonmethylation, respectively.D, miR-644a was upregulated in the indicated ESCC cell lines after the treatment of 5-aza-dC for 72 hours. � , P < 0.05; �� , P < 0.01.

Zhang et al.





characterized in the development of stem cell–like properties in awide variety of cancer types, including ESCC (25, 26). Therefore,we hypothesized that the levels of miR-644a could influence thestemness of ESCC cells via regulation of PITX2. Our resultsdemonstrated a novel role for miR-644a in suppressing ESCCtumorigenesis by modulating ESCC cell's stem-like traits throughPITX2. We thus propose that miR-644a downregulation in ESCCcells, which increases PITX2 expression, results in aberrant acti-vation of the Wnt/b-catenin pathway and a resultant stem cell–like phenotype. This might promote ESCC cell survival followingadjuvant therapy, such as radiochemotherapy, thus mediatingresistance to therapy and resulting in tumor recurrence and/ormetastasis.

Previous studies have supported the notion that some aber-rantly expressedmiRNAs could be used as an ideal target in cancertherapy; in general, therapeutic modulation of miRNAs isachieved by inhibiting oncogenic miRNAs, or by reconstitutingtumor-suppressive miRNAs (34, 35). Multiple approaches havealso been in rapid development to open novel avenues forexploiting miRNAs in preclinical practice (36). For instance,restoring the expression of silenced miRNAs, such as Let-7,miR-16, and miR-31, has been shown to have strong anticancereffects in various preclinical models (37). These data, togetherwith our currentfindings, suggest that increasing the levels ofmiR-644a in ESCC cells by miR-644a replacement might be an effi-cacious approach for cancer treatment, leading to the inactivationof certain oncogenic pathways, specifically the Akt/GSK-3b/b-catenin signaling, thus generating strong suppressive effects onESCC stemness and/or progression.

In our study, one critical question was raised: how is miR-644adownregulated in ESCCs? It is well established that promotermethylation and genomic deletion are the most common path-ologic mechanisms for aberrant expression of certain miRNAs inhuman cancers (38). In the current study, we provided severallines of evidences that the promoter hypermethylation of miR-644a is a key regulator of miR-644a expression and that thismechanism appears to be used to decrease expression of miR-644a in ESCCs. It is noteworthy here that 5-aza-dCv, awidely usedmethylase inhibitor, could significantly increase the levels ofmiR-644a in ESCC cells with low endogenous miR-644a. Previousstudies have suggested that the reversal of some epigenetic pro-cesses, such as DNA demethylation, could prevent or reverse themalignant phenotype of human cancers by reactivating tumorsuppressor genes (39, 40). The potential anticancer activities ofDNA demethylation agents, including 5-aza-dC, have been stud-ied extensively and represent a promising avenue for cancerprevention and/or therapy (41, 42). Furthermore, 5-aza-dC has

shownprominent antitumor activity in vivo and is being evaluatedfor its potential clinical significance (43, 44). These data implythat in addition to the miR-644a replacement therapy strategy inESCC, demethylation of the miR-644a promoter by methylaseinhibitors is another approach that could prove effective as anESCC treatment.

In summary, our report describes, for the first time, the expres-sion dynamics of miR-644a in human ESCCs. Decreased expres-sion ofmiR-644a, caused by its promoter hypermethylation, maybe important in the tumorigenesis and acquisition of an aggres-sive/poor prognostic phenotype of ESCC. In addition, functionaland mechanistic studies of miR-644a provided in this studysuggest a mechanism by which downregulation of miR-644acould support ESCC cells' aggressiveness and/or stem cell–likephenotype by activating Akt/GSK-3b/b-catenin signaling throughPITX2, an activity thatmight be responsible, at least in part, for thedevelopment and/or progression of human ESCCs. Thus, miR-644a could be employed as a new prognostic marker and/or bedeveloped as a therapeutic for treatment of ESCC.

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

Authors' ContributionsConception and design: D. XieDevelopment of methodology: Z.-H. Chen, Y. Xu, Z.-S. Zheng, S. YeAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.-X. Zhang, Z.-H. Chen, Y. Xu, J.-W. Chen, H.-W.Weng, M. Yun, C. Chen, B.-L. Wu, E.-M. Li, J.-H. FuAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.-X. Zhang, Z.-H. Chen, Z.-S. ZhengWriting, review, and/or revision of the manuscript: J.-X. ZhangAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): Y. Xu, H.-W. Weng, M. Yun, C. Chen, B.-L. Wu,E.-M. Li, J.-H. FuStudy supervision: S. Ye , D. Xie

Grant SupportThis work was supported by the National Natural Science Foundation of

China (nos. 81401991, 81225018, and 81572359), the Natural Science Foun-dation of Guangdong (no. S2012010009466), the Foundation of GuangdongEsophageal Cancer Institute (no.M01513), and theNatural Science Foundationof China-Guangdong Joint Fund (no. U1301227).

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

Received February 15, 2016; revised June 2, 2016; accepted June 21, 2016;published OnlineFirst July 12, 2016.

References1. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer

statistics. CA Cancer J Clin 2011;61:69–90.2. Su XD, Zhang DK, Zhang X, Lin P, Long H, Rong TH. Prognostic factors in

patients with recurrence after complete resection of esophageal squamouscell carcinoma. J Thorac Dis 2014;6:949–57.

3. Mariette C, Balon JM, Piessen G, Fabre S, Van Seuningen I, Triboulet JP.Pattern of recurrence following complete resection of esophagealcarcinoma and factors predictive of recurrent disease. Cancer 2003;97:1616–23.

4. Luthra R, Wu TT, Luthra MG, Izzo J, Lopez-Alvarez E, Zhang L, et al. Geneexpression profiling of localized esophageal carcinomas: association withpathologic response to preoperative chemoradiation. J Clin Oncol2006;24:259–67.

5. Ninomiya I, Endo Y, Fushida S, Sasagawa T, Miyashita T, Fujimura T,et al. Alteration of beta-catenin expression in esophageal squamous-cellcarcinoma. Int J Cancer 2000;85:757–61.

6. Wang Y, Zhou X, Zhu H, Liu S, Zhou C, Zhang G, et al. Overexpression ofEB1 in human esophageal squamous cell carcinoma (ESCC)may promotecellular growth by activating beta-catenin/TCF pathway. Oncogene 2005;24:6637–45.

7. Suh H, Gage PJ, Drouin J, Camper SA. Pitx2 is required at multiple stagesof pituitary organogenesis: pituitary primordium formation and cellspecification. Development 2002;129:329–37.

8. Basu M, Roy SS. Wnt/beta-catenin pathway is regulated by PITX2 home-odomain protein and thus contributes to the proliferation of humanovarian adenocarcinoma cell, SKOV-3. J Biol Chem 2013;288:4355–67.






9. Huang Y, Guigon CJ, Fan J, Cheng SY, Zhu GZ. Pituitary homeobox 2(PITX2) promotes thyroid carcinogenesis by activation of cyclin D2. CellCycle 2010;9:1333–41.

10. Fung FK, Chan DW, Liu VW, Leung TH, Cheung AN, Ngan HY. Increasedexpression of PITX2 transcription factor contributes to ovarian cancerprogression. PLoS One 2012;7:e37076.

11. Hirose H, Ishii H, Mimori K, Tanaka F, Takemasa I, Mizushima T, et al. Thesignificance of PITX2 overexpression in human colorectal cancer. Ann SurgOncol 2011;18:3005–12.

12. Zhang JX, Tong ZT, Yang L, Wang F, Chai HP, Zhang F, et al. PITX2: apromising predictive biomarker of patients' prognosis and chemoradiore-sistance in esophageal squamous cell carcinoma. Int J Cancer 2013;132:2567–77.

13. Bartel DP. MicroRNAs: genomics, biogenesis, mechanism, and function.Cell 2004;116:281–97.

14. Kan T, Sato F, Ito T, Matsumura N, David S, Cheng Y, et al. The miR-106b-25 polycistron, activated by genomic amplification, functions as an onco-gene by suppressing p21 and Bim. Gastroenterology 2009;136:1689–700.

15. Kong KL, Kwong DL, Chan TH, Law SY, Chen L, Li Y, et al. MicroRNA-375inhibits tumour growth and metastasis in oesophageal squamous cellcarcinoma through repressing insulin-like growth factor 1 receptor. Gut2012;61:33–42.

16. Wu BL, Xu LY, Du ZP, Liao LD, ZhangHF, HuangQ, et al. MiRNA profile inesophageal squamous cell carcinoma: downregulation of miR-143 andmiR-145. World J Gastroenterol 2011;17:79–88.

17. Lee SH, Koo BS, Kim JM, Huang S, Rho YS, Bae WJ, et al. Wnt/beta-cateninsignalling maintains self-renewal and tumourigenicity of head and necksquamous cell carcinoma stem-like cells by activating Oct4. J Pathol2014;234:99–107.

18. Zhang N, Wei P, Gong A, Chiu WT, Lee HT, Colman H, et al. FoxM1promotes beta-catenin nuclear localization and controls Wnt target-geneexpression and glioma tumorigenesis. Cancer Cell 2011;20:427–42.

19. He K, Xu T, Xu Y, Ring A, Kahn M, Goldkorn A. Cancer cells acquire a drugresistant, highly tumorigenic, cancer stem-like phenotype through mod-ulation of the PI3K/Akt/beta-catenin/CBP pathway. Int J Cancer 2014;134:43–54.

20. Chen J, Chan AW, To KF, Chen W, Zhang Z, Ren J, et al. SIRT2 over-expression in hepatocellular carcinoma mediates epithelial to mesenchy-mal transition by protein kinase B/glycogen synthase kinase-3beta/beta-catenin signaling. Hepatology 2013;57:2287–98.

21. Sundaresan NR, Pillai VB, Wolfgeher D, Samant S, Vasudevan P, Parekh V,et al. The deacetylase SIRT1 promotes membrane localization and activa-tion of Akt and PDK1 during tumorigenesis and cardiac hypertrophy. SciSignal 2011;4:ra46.

22. Diaz-Beya M, Brunet S, Nomdedeu J, Tejero R, Diaz T, Pratcorona M, et al.MicroRNA expression at diagnosis adds relevant prognostic informationto molecular categorization in patients with intermediate-risk cyto-genetic acute myeloid leukemia. Leukemia 2014;28:804–12.

23. Clevers H, Nusse R. Wnt/beta-catenin signaling and disease. Cell 2012;149:1192–205.

24. Klaus A, Birchmeier W. Wnt signalling and its impact on development andcancer. Nat Rev Cancer 2008;8:387–98.

25. Reya T, Clevers H. Wnt signalling in stem cells and cancer. Nature2005;434:843–50.

26. Ge C, Wu S, Wang W, Liu Z, Zhang J, Wang Z, et al. miR-942 promotescancer stem cell-like traits in esophageal squamous cell carcinoma throughactivation of Wnt/beta-catenin signalling pathway. Oncotarget 2015;6:10964–77.

27. Zhou BP, Deng J, Xia W, Xu J, Li YM, Gunduz M, et al. Dual regulation ofSnail by GSK-3beta-mediated phosphorylation in control of epithelial-mesenchymal transition. Nat Cell Biol 2004;6:931–40.

28. Visvader JE, Lindeman GJ. Cancer stem cells in solid tumours: accu-mulating evidence and unresolved questions. Nat Rev Cancer 2008;8:755–68.

29. Dick JE. Looking ahead in cancer stem cell research. Nat Biotechnol2009;27:44–6.

30. Marson A, Levine SS, Cole MF, Frampton GM, Brambrink T, Johnstone S,et al. Connecting microRNA genes to the core transcriptional regulatorycircuitry of embryonic stem cells. Cell 2008;134:521–33.

31. Tay Y, Zhang J, Thomson AM, Lim B, Rigoutsos I. MicroRNAs to Nanog,Oct4 and Sox2 coding regions modulate embryonic stem cell differenti-ation. Nature 2008;455:1124–8.

32. Yu F, Yao H, Zhu P, Zhang X, Pan Q, Gong C, et al. let-7 regulatesself renewal and tumorigenicity of breast cancer cells. Cell 2007;131:1109–23.

33. Li L, Tang J, Zhang B, Yang W, LiuGao M, Wang R, et al. Epigeneticmodification of MiR-429 promotes liver tumour-initiating cell propertiesby targeting Rb binding protein 4. Gut 2015;64:156–67.

34. Naidu S, Magee P, Garofalo M. MiRNA-based therapeutic intervention ofcancer. J Hematol Oncol 2015;8:68.

35. Bader AG, Brown D, Stoudemire J, Lammers P. Developing therapeuticmicroRNAs for cancer. Gene Ther 2011;18:1121–6.

36. Melo SA, Kalluri R.Molecular pathways:microRNAs as cancer therapeutics.Clin Cancer Res 2012;18:4234–9.

37. Kim M, Kasinski AL, Slack FJ. MicroRNA therapeutics in preclinical cancermodels. Lancet Oncol 2011;12:319–21.

38. Suzuki H, Maruyama R, Yamamoto E, Kai M. DNA methylation andmicroRNA dysregulation in cancer. Mol Oncol 2012;6:567–78.

39. Santini V, Kantarjian HM, Issa JP. Changes in DNA methylation inneoplasia: pathophysiology and therapeutic implications. Ann InternMed2001;134:573–86.

40. Brown R, Plumb JA. Demethylation of DNA by decitabine in cancerchemotherapy. Expert Rev Anticancer Ther 2004;4:501–10.

41. Lubbert M. DNA methylation inhibitors in the treatment of leukemias,myelodysplastic syndromes and hemoglobinopathies: clinical results andpossible mechanisms of action. Curr Top Microbiol Immunol 2000;249:135–64.

42. Hsi LC, Xi X, Wu Y, Lippman SM. The methyltransferase inhibitor 5-aza-2-deoxycytidine induces apoptosis via induction of 15-lipoxygenase-1 incolorectal cancer cells. Mol Cancer Ther 2005;4:1740–6.

43. Wijermans P, LubbertM, Verhoef G, Bosly A, Ravoet C, AndreM, et al. Low-dose 5-aza-20-deoxycytidine, a DNA hypomethylating agent, for the treat-ment of high-risk myelodysplastic syndrome: a multicenter phase II studyin elderly patients. J Clin Oncol 2000;18:956–62.

44. Issa JP, Garcia-Manero G, Giles FJ, Mannari R, Thomas D, Faderl S, et al.Phase 1 study of low-dose prolonged exposure schedules of the hypo-methylating agent 5-aza-20-deoxycytidine (decitabine) in hematopoieticmalignancies. Blood 2004;103:1635–40.


Zhang et al.




2017;23:298-310. Published OnlineFirst July 12, 2016.Clin Cancer Res Jia-Xing Zhang, Zhen-Hua Chen, Yi Xu, et al. Phenotype via Dysregulation of PITX2

like−Squamous Cell Carcinoma Aggressiveness and Stem Cell Downregulation of MicroRNA-644a Promotes Esophageal

Updated version

10.1158/1078-0432.CCR-16-0414doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2016/07/12/1078-0432.CCR-16-0414.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/23/1/298.full#ref-list-1

This article cites 44 articles, 9 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/23/1/298.full#related-urls

This article has been cited by 1 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/23/1/298To 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-16-0414

http://clincancerres.aacrjournals.org/content/suppl/2016/07/12/1078-0432.CCR-16-0414.DC1

http://clincancerres.aacrjournals.org/content/23/1/298.full#ref-list-1

http://clincancerres.aacrjournals.org/content/23/1/298.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/23/1/298


Downregulation of MicroRNA-644a Promotes...

Documents

Transcript of Downregulation of MicroRNA-644a Promotes...