Advertisement

Silencing NKD2 by Promoter Region Hypermethylation Promotes Esophageal Cancer Progression by Activating Wnt Signaling

Open AccessPublished:June 30, 2016DOI:https://doi.org/10.1016/j.jtho.2016.06.015

      Abstract

      Introduction

      Naked cuticle homolog 2 (NKD2) was found to be frequently methylated in human breast and gastric cancers. However, the epigenetic changes and mechanisms of NKD2 in human esophageal cancer remain unclear.

      Methods

      Nine esophageal cancer cell lines and 154 cases of primary esophageal cancer samples were analyzed using methylation-specific polymerase chain reaction, immunohistochemical analysis, Western blot, and xenograft mouse models.

      Results

      Loss of NKD2 expression and complete methylation were found in KYSE150 and TE1 cells. Reduced NKD2 expression and partial methylation of the promoter region were observed in KYSE30, KYSE70, KYSE410, KYSE140, and COLO680 cells. High levels of NKD2 expression and unmethylation were detected in KYSE450 and TE8 cells. Reexpression of NKD2 was induced by 5-aza-2′-deoxycytidine in cells in which NKD2 was not expressed or cells in which NKD2 expression was reduced. NKD2 was methylated in 53.2% of human primary esophageal cancer samples (82 of 154), and promoter region hypermethylation was significantly associated with reduced expression of NKD2 (p < 0.01). NKD2 methylation was associated with tumor, node, and metastasis stage and lymph node metastasis (p < 0.01). Our results suggest that NKD2 is regulated by promoter region methylation and that methylation of NKD2 may serve as a prognostic marker in esophageal cancer. Our further studies demonstrate that NKD2 suppresses cell proliferation, colony formation, cell invasion, and migration and also induces G1/S checkpoint arrest in esophageal cancer cells. NKD2 suppressed xenograft tumor growth and inhibited Wnt signaling in human esophageal cancer cells.

      Conclusions

      NKD2 is frequently methylated in human esophageal cancer, and the expression of NKD2 is regulated by promoter region methylation. NKD2 suppresses esophageal cancer progression by inhibiting Wnt signaling both in vitro and in vivo.

      Keywords

      Introduction

      Esophageal cancer is the eighth most common cancer and the sixth leading cause of cancer-related death worldwide.
      • Ferlay J.
      • Soerjomataram I.
      • Dikshit R.
      • et al.
      Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012.
      The overall 5-year survival remains less than 15%.
      • Enzinger P.C.
      • Mayer R.J.
      Esophageal cancer.
      Poor outcomes in patients with esophageal cancer are related to diagnosis at advanced (metastatic) stages and the propensity for metastases.
      • Pennathur A.
      • Farkas A.
      • Krasinskas A.M.
      • et al.
      Esophagectomy for T1 esophageal cancer: outcomes in 100 patients and implications for endoscopic therapy.
      Esophageal squamous cell carcinoma (ESCC) is the predominant histological type of esophageal carcinoma worldwide.
      • Pennathur A.
      • Gibson M.K.
      • Jobe B.A.
      • et al.
      Oesophageal carcinoma.
      Tobacco use and alcohol consumption are risk factors for development of ESCC, and the combination of tobacco and alcohol consumption further increases the risk for development of ESCC. Mutations in enzymes that metabolize alcohol have been associated with increased risk for development of ESCC.
      • De Stefani E.
      • Barrios E.
      • Fierro L.
      Black (air-cured) and blond (flue-cured) tobacco and cancer risk. III: oesophageal cancer.
      • Lee C.H.
      • Wu D.C.
      • Lee J.M.
      • et al.
      Carcinogenetic impact of alcohol intake on squamous cell carcinoma risk of the oesophagus in relation to tobacco smoking.
      • Vaughan T.L.
      • Davis S.
      • Kristal A.
      • et al.
      Obesity, alcohol, and tobacco as risk factors for cancers of the esophagus and gastric cardia: adenocarcinoma versus squamous cell carcinoma.
      Genomewide association analysis has demonstrated that gene-environment interaction promotes development of ESCC.
      • Wu C.
      • Kraft P.
      • Zhai K.
      • et al.
      Genome-wide association analyses of esophageal squamous cell carcinoma in Chinese identify multiple susceptibility loci and gene-environment interactions.
      Other studies have found that a second cancer develops in 2% of patients with esophageal cancer and 11% of patients with head and neck cancer on account of field cancerization.
      • Chuang S.C.
      • Hashibe M.
      • Scelo G.
      • et al.
      Risk of second primary cancer among esophageal cancer patients: a pooled analysis of 13 cancer registries.
      • Chuang S.C.
      • Scelo G.
      • Tonita J.M.
      • et al.
      Risk of second primary cancer among patients with head and neck cancers: a pooled analysis of 13 cancer registries.
      These studies support the idea that environment plays an important role in ESCC. Genetic and epigenetic alterations are involved in esophageal carcinogenesis.
      • Ahrens T.D.
      • Werner M.
      • Lassmann S.
      Epigenetics in esophageal cancers.
      Whole genome and whole exome sequencing in Chinese patients with ESCC revealed eight mutated genes, including six known tumor-associated genes (tumor protein p53 gene [TP53], RB transcriptional corepressor 1 gene [RB1], cyclin-dependent kinase 2 gene [CDKN2A], phosphatidylinositol 4,5-bisphosphate 3-kinase catalytic subunit alpha gene [PIK3CA], Homo sapiens notch 1[NOTCH1], and nuclear factor erythroid 2, like 2 gene [NFE2L2]) and two novel genes (ADAM metallopeptidase domain 29 gene [ADAM29] and family with sequence similarity 135 member B gene [FAM135B]).
      • Song Y.
      • Li L.
      • Ou Y.
      • et al.
      Identification of genomic alterations in oesophageal squamous cell cancer.
      Additional genes were found to be frequently methylated in ESCC in previous studies.
      • Yun T.
      • Liu Y.
      • Gao D.
      • et al.
      Methylation of CHFR sensitizes esophageal squamous cell cancer to docetaxel and paclitaxel.
      • Wu L.
      • Herman J.G.
      • Brock M.V.
      • et al.
      Silencing DACH1 promotes esophageal cancer growth by inhibiting TGF-beta signaling.
      • Lu D.
      • Ma J.
      • Zhan Q.
      • et al.
      Epigenetic silencing of RASSF10 promotes tumor growth in esophageal squamous cell carcinoma.
      • Jiang S.
      • Linghu E.
      • Zhan Q.
      • et al.
      Methylation of ZNF331 promotes cell invasion and migration in human esophageal cancer.
      • Jia Y.
      • Yang Y.
      • Zhan Q.
      • et al.
      Inhibition of SOX17 by microRNA 141 and methylation activates the WNT signaling pathway in esophageal cancer.
      • Jia Y.
      • Yang Y.
      • Brock M.V.
      • et al.
      Methylation of TFPI-2 is an early event of esophageal carcinogenesis.
      • Guo M.
      • Ren J.
      • House M.G.
      • et al.
      Accumulation of promoter methylation suggests epigenetic progression in squamous cell carcinoma of the esophagus.
      • Guo M.
      • Ren J.
      • Brock M.V.
      • et al.
      Promoter methylation of HIN-1 in the progression to esophageal squamous cancer.
      • Guo M.
      • House M.G.
      • Suzuki H.
      • et al.
      Epigenetic silencing of CDX2 is a feature of squamous esophageal cancer.
      • Guo M.
      • House M.G.
      • Akiyama Y.
      • et al.
      Hypermethylation of the GATA gene family in esophageal cancer.
      • Chen X.Y.
      • He Q.Y.
      • Guo M.Z.
      XAF1 is frequently methylated in human esophageal cancer.
      • Brock M.V.
      • Gou M.
      • Akiyama Y.
      • et al.
      Prognostic importance of promoter hypermethylation of multiple genes in esophageal adenocarcinoma.
      Aberrant expression levels of Wnt signaling pathway components are found in many types of cancers, including esophageal cancer, and the Wnt signaling pathway plays an important role in cancer progression.
      • Jia Y.
      • Yang Y.
      • Zhan Q.
      • et al.
      Inhibition of SOX17 by microRNA 141 and methylation activates the WNT signaling pathway in esophageal cancer.
      • He G.
      • Guan X.
      • Chen X.
      • et al.
      Expression and splice variant analysis of human TCF4 transcription factor in esophageal cancer.
      Despite recent advances in treatment strategies for esophageal cancer, there has been no significant improvement in overall survival rate for advanced and metastatic disease.
      • Mohamed A.
      • El-Rayes B.
      • Khuri F.R.
      • et al.
      Targeted therapies in metastatic esophageal cancer: advances over the past decade.
      New strategies are necessary for early detection and to improve treatment options in ESCC. Aberrant epigenetic changes can be induced by environmental factors, and epigenetic changes are reversible under certain circumstances.
      • Herman J.G.
      • Baylin S.B.
      Gene silencing in cancer in association with promoter hypermethylation.
      • Baylin S.B.
      • Esteller M.
      • Rountree M.R.
      • et al.
      Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer.
      Therefore, more effective therapeutic strategies based on epigenetics are being developed.
      The naked cuticle (NKD) family includes Drosophila naked cuticle and its two vertebrate orthologs, naked cuticle homolog 1 (NKD1) and naked cuticle homolog 2 (NKD2). NKD1 is located in human chromosome 16q12.1, which has frequent loss of heterozygosity in human breast and hepatocellular carcinoma.
      • Argos M.
      • Kibriya M.G.
      • Jasmine F.
      • et al.
      Genomewide scan for loss of heterozygosity and chromosomal amplification in breast carcinoma using single-nucleotide polymorphism arrays.
      • Sheu J.C.
      • Lin Y.W.
      • Chou H.C.
      • et al.
      Loss of heterozygosity and microsatellite instability in hepatocellular carcinoma in Taiwan.
      NKD2 is located in chromosome 5p15.3. Loss of heterozygosity has been frequently found in these regions in multiple tumors.
      • Arias-Pulido H.
      • Narayan G.
      • Vargas H.
      • et al.
      Mapping common deleted regions on 5p15 in cervical carcinoma and their occurrence in precancerous lesions.
      • Xu S.F.
      • Peng Z.H.
      • Li D.P.
      • et al.
      Refinement of heterozygosity loss on chromosome 5p15 in sporadic colorectal cancer.
      • Lu Y.
      • Yu Y.
      • Zhu Z.
      • et al.
      Identification of a new target region by loss of heterozygosity at 5p15.33 in sporadic gastric carcinomas: genotype and phenotype related.
      In both zebrafish and mice, NKD inhibits canonical and noncanonical Wnt signaling.
      • Van Raay T.J.
      • Coffey R.J.
      • Solnica-Krezel L.
      Zebrafish Naked1 and Naked2 antagonize both canonical and non-canonical Wnt signaling.
      • Katoh M.
      Molecular cloning, gene structure, and expression analyses of NKD1 and NKD2.
      • Yan D.
      • Wallingford J.B.
      • Sun T.Q.
      • et al.
      Cell autonomous regulation of multiple Dishevelled-dependent pathways by mammalian Nkd.
      The C-terminus of NKD2 is highly disordered, whereas the N-terminal region contains most of the functional domains, including myristoylation, an EF-hand motif, a Dishevelled binding region, and a vesicle recognition and membrane targeting motif.
      • Rousset R.
      • Mack J.A.
      • Wharton Jr., K.A.
      • et al.
      Naked cuticle targets dishevelled to antagonize Wnt signal transduction.
      • Zeng W.
      • Wharton Jr., K.A.
      • Mack J.A.
      • et al.
      Naked cuticle encodes an inducible antagonist of Wnt signalling.
      • Li C.
      • Franklin J.L.
      • Graves-Deal R.
      • et al.
      Myristoylated Naked2 escorts transforming growth factor alpha to the basolateral plasma membrane of polarized epithelial cells.
      NKD2 binds to multiple proteins and may function as a switch protein through its several functional motifs.
      • Hu T.
      • Krezel A.M.
      • Li C.
      • et al.
      Structural studies of human Naked2: a biologically active intrinsically unstructured protein.
      Both NKD1 and NKD2 have been proposed to interact with Dishevelled through their EF-hand–like motif. In addition, NKD2 has been reported to bind to Dishevelled through its transforming growth factor-α binding region.
      • Li C.
      • Franklin J.L.
      • Graves-Deal R.
      • et al.
      Myristoylated Naked2 escorts transforming growth factor alpha to the basolateral plasma membrane of polarized epithelial cells.
      • Hu T.
      • Krezel A.M.
      • Li C.
      • et al.
      Structural studies of human Naked2: a biologically active intrinsically unstructured protein.
      NKD2 was reported to suppress tumor growth and metastasis in osteosarcoma through negative regulation of Wnt signaling.
      • Zhao S.
      • Kurenbekova L.
      • Gao Y.
      • et al.
      NKD2, a negative regulator of Wnt signaling, suppresses tumor growth and metastasis in osteosarcoma.
      Our previous study found that methylation of NKD2 promotes breast cancer growth by activating Wnt signaling.
      • Dong Y.
      • Cao B.
      • Zhang M.
      • et al.
      Epigenetic silencing of NKD2, a major component of Wnt signaling, promotes breast cancer growth.
      The methylation status and the function of NKD2 in esophageal cancer have yet to be elucidated. Therefore, we investigated the epigenetic changes and functions of NKD2 in human ESCC.

      Materials and Methods

      Human Tissue Samples and Cell Lines

      Fifteen samples of human normal esophageal mucosa and samples from 154 cases of human esophageal cancer were collected from the Chinese People's Liberation Army (PLA) General Hospital in Beijing. The median age of the cancer patients was 62.1 years (range 46–87), and the ratio of male to female patients was 3.05:1. All cancer samples were classified according to tumor, node, and metastasis (TNM) staging (American Joint Committee on Cancer, 2010), including five cases of stage I, 99 cases of stage II, and 50 cases of stage III cancer. All samples were collected following the guidelines approved by the Institutional Review Board of the Chinese PLA General Hospital with written informed consent from patients (Reference No. 20090701-015).
      Nine esophageal cancer cell lines (KYSE450, KYSE30, KYSE150, KESE70, TE8, KYSE410, TE1, KYSE140, and COLO680) were previously established from primary esophageal cancer and maintained in 90% Roswell Park Memorial Institute 1640 medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum.

      5-AZA Treatment

      Esophageal cancer cell lines were split in a low-density confluence (30%) 12 hours before treatment. Cells were treated with 5-aza-2′-deoxycytidine (5-AZA [Sigma, St. Louis, MO]) at a concentration of 2 μM. Growth medium conditioned with 5-AZA at a concentration of 2 μM was exchanged every 24 hours for a total of 96 hours of treatment.

      RNA Isolation and Semiquantitative Reverse Transcriptase PCR

      Total RNA was isolated by Trizol reagent (Life Technologies, Carlsbad, CA). First-strand cDNA was synthesized according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). The polymerase chain reaction (PCR) primers used for NKD2 were 5′-ACAGGAGGTTGTCTGCACACG-3′ (F) and 5′-GACTTGAGGAACTGCTTCTCC-3′ (R). The primer sets for NKD2 were designed to span intronic sequences between adjacent exons to control for genomic DNA contamination. Semiquantitative reverse-transcriptase PCR (RT-PCR) was amplified for 33 cycles. Glyceraldehyde-3-phosphate dehydrogenase was used as an internal control.

      Mutation Detection, Bisulfite Modification, MSP, and BSSQ

      DNA was prepared by the proteinase K method. To detect NKD2 mutations (72C>T, 1231G>T, and 689C>T) in nine esophageal cancer cell lines and 154 cases of primary ESCC, the following sequencing primers were used: NKD2-72C-T- F: CACGGCGCGTCTCTTTCC, NKD2-72C-T- R: TACCCGCCACTCAGCATG, NKD2-689C-T-F: AGGGAGTTCACAGGGTTCT, NKD2-689C-T-R: AGGTCCAGGTAGTGGTTTCT, NKD2 1231G-T-F: GCAGCAAGTCCGGGAAAGC, and NKD2 1231G-T-R: GCTAGGACGGGTGGAAGTG. Bisulfite treatment was carried out as previously described.
      • Cao B.
      • Yang Y.
      • Pan Y.
      • et al.
      Epigenetic silencing of CXCL14 induced colorectal cancer migration and invasion.
      • Herman J.G.
      • Graff J.R.
      • Myohanen S.
      • et al.
      Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands.
      Methylation-specific PCR (MSP) primers were designed according to genomic sequences around transcription start sites and synthesized to detect unmethylated and methylated alleles. Bisulfite sequencing (BSSQ) was performed as previously described.
      • Jia Y.
      • Yang Y.
      • Zhan Q.
      • et al.
      Inhibition of SOX17 by microRNA 141 and methylation activates the WNT signaling pathway in esophageal cancer.
      BSSQ products were amplified by primers flanking the targeted regions and included MSP products. The locations and the sequences of BSSQ and MSP primers were listed in Supplementary Table 1 and shown in Supplementary Figure 1. To obtain more evidence supporting our discovery, the expression of NKD2 and the methylation status in The Cancer Genome Atlas database were analyzed by the Pearson correlation method in esophageal cancer and adjacent tissue samples (Supplementary Figs. 2 and 3).

      IHC Analysis

      Immunohistochemical (IHC) analysis was performed in human esophageal cancer samples and paired adjacent tissue samples. The NKD2 antibody (Novus Biologicals, Littleton, CO), matrix metalloproteinase 2 (MMP2) antibody (Proteintech, Chicago, IL), matrix metalloproteinase 9 (MMP9) antibody (Proteintech), phosphorylated β-catenin (p-β-catenin) antibody (Bioworld Technology Inc., St. Louis Park, MN), and total β-catenin antibody (Cell Signaling Technology, Danvers, MA) were diluted to 1:500, 1:200, 1:200, 1:100, and 1:200, respectively. For antigen retrieval, the slides were placed in citrate antigen-repairing solution and heated in a high-pressure cooker until steam arose. The slides were kept inside the cooker for 150 seconds and then cooled at room temperature for 15 minutes. The staining intensity and extent of the staining area were scored using the German semiquantitative scoring system as described previously.
      • Jia Y.
      • Yang Y.
      • Brock M.V.
      • et al.
      Methylation of TFPI-2 is an early event of esophageal carcinogenesis.
      • Yan W.
      • Wu K.
      • Herman J.G.
      • et al.
      Epigenetic regulation of DACH1, a novel Wnt signaling component in colorectal cancer.

      Plasmid Construction

      Human full-length NKD2 coding DNA sequence (GenBank accession number NM_033120) was amplified and subcloned as described previously
      • Dong Y.
      • Cao B.
      • Zhang M.
      • et al.
      Epigenetic silencing of NKD2, a major component of Wnt signaling, promotes breast cancer growth.
      using the following primers: 5′-GAGGATCCGCCACCATGGGGAAACTGCAGTCGAAG-3′ (F) and 5′-GATCTCGAGCTAGGACGGGTGGAAGTGGT-3′ (R). NKD2 expressing lentiviral or empty vectors was packaged using the ViraPower lentiviral expression system (Invitrogen, San Diego, CA). Lentivirus was added to the growing medium of KYSE150 and TE1 cells, and cells stably expressed by NKD2 were selected using blasticidin (Invitrogen, San Diego) at a concentration of 2 μg/mL.

      Cell Viability Detection

      Cells were plated into 96-well plates at a density of 2 × 103 cells per well, and cell viability was measured by the methyl thiazolyl tetrazolium assay (KeyGEN Biotech, Nanjing, People's Republic of China) at 0, 24, 48, and 72 hours. Absorbance was measured on a microplate reader (Thermo Multiskan MK3 [Thermo Fisher Scientific, Danvers, MA]) at a wavelength of 490 nm.

      Colony Formation Assay

      Cells in which NKD2 was unexpressed or stably expressed were seeded at 500 cells per well in six-well culture plates in triplicate. The complete growth medium conditioned with blasticidin at 2 ug/mL was exchanged every 72 hours. After 2 weeks, cells were fixed with 75% ethanol for 30 minutes and stained with 0.2% crystal violet (Beyotime, Nanjing, China) for visualization and counting.

      Flow Cytometry

      KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed were starved 12 hours for synchronization, and the cells were restimulated with 10% fetal bovine serum for 24 hours. Cells were fixed with 70% ethanol and treated using the Cell Cycle Detection Kit (KeyGEN Biotech). The cells were then sorted by a FACSCaliber flow cytometer (BD Biosciences, Mansfield, MA). The cell phase distribution was analyzed by Modfit software (Verity Software House, Topsham, ME).

      Transwell Assay

      KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed were suspended in serum-free medium. Cells (2 × 105) were placed into the upper chamber of an 8-μm pore size transwell apparatus (Corning, Corning, NY) and incubated for 20 hours. Cells that migrated to the lower surface of the membrane were stained with crystal violet and counted in three independent high-power fields (×200). For invasion analysis, KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed (2 × 105) were seeded into the upper chamber of a transwell apparatus coated with extracellular matrix (ECM) gel (BD Biosciences, San Jose, CA) and incubated for 36 hours. Cells that invaded the lower membrane surface were stained with crystal violet and counted in three independent high-power fields (×200).

      SiRNA Knockdown Technique

      Selected small interfering RNAs (siRNAs) targeting NKD2 and the RNAi negative control duplex were used in this study. The sequences of the siRNAs targeting NKD2 and the RNAi negative control were as follows: NKD2-F: 5′-GGGAUUGAGAACUACACGUTT-3′ and NKD2-R: 5′-ACGUGUAGUUCUCAAUCCCTT-3′; negative control-F: 5′-UUCUCCGAACGUGUCACGUTT-3′ and negative control-R: 5′-ACGUGACACGUUCGGAGAATT-3′. The RNAi oligonucleotide and RNAi negative control duplex were transfected into KYSE450 cells, which expressed high levels of NKD2.

      Esophageal Cancer Cell Xenograft Mouse Model

      KYSE150 cells (4×106 cells in 0.1 mL phosphate-buffered saline) in which NKD2 was stably expressed and unexpressed were subcutaneously injected into the dorsal flank of 5-week-old female BALB/c nude mice. The tumor sizes were measured every 3 days from the third day after implantation for 24 days. KYSE450 cells (4 × 106 cells in 0.1 mL phosphate-buffered saline), in which NKD2 was either highly expressed or knocked down, were subcutaneously injected into the dorsal flank of 5-week-old female BALB/c nude mice. As KYSE450 cells grow slowly, the tumor sizes were measured every 4 days from the fourth day after implantation for 24 days. The tumor volumes were calculated according to the following formula: V = L × W2/2, where V is the volume (mm3), L is the largest diameter (mm), and W is the smallest diameter (mm). All procedures were approved by the Animal Ethics Committee of the Chinese PLA General Hospital.

      Western Blot

      Protein samples from esophageal cancer cells were collected and Western blot was performed as described previously.
      • Yu Y.
      • Yan W.
      • Liu X.
      • et al.
      DACT2 is frequently methylated in human gastric cancer and methylation of DACT2 activated Wnt signaling.
      Antibodies were diluted according to manufacturer’s instructions. The primary antibodies were as follows: NKD2 (Cell Signaling Technology), MMP2, matrix metalloproteinase 7 (MMP7), MMP9, cyclin D1, c-myc, p-β-catenin, β-catenin, and β-actin (Bioworld Technology).

      Statistical Analysis

      SPSS 17.0 software (IBM, Armonk, NY) was used for data analysis. All data are presented as means plus or minus SD and analyzed using the Student’s t test. The chi-square test and the Fisher’s exact test were used to analyze the association between NKD2 methylation status and clinicopathologic factors and the association between NKD2 expression and methylation status. Correlations between NKD2 and MMP2, MMP9, and p-β-catenin were analyzed using Pearson's correlation coefficient. A p value less than 0.05 was considered statistically significant.

      Results

      NKD2 Expression Is Regulated by Promoter Region Methylation in Esophageal Cancer Cell Lines

      No mutations were found in the naked cuticle homolog 2 gene (NKD2) gene in human ESCC cell lines and primary tissue samples. The expression of NKD2 was detected by semiquantitative RT-PCR in human esophageal cancer cell lines. As shown in Figure 1A, loss of NKD2 expression was found in KYSE150 and TE1 cells. Reduced expression of NKD2 was observed in KYSE30, KYSE70, KYSE410, KYSE140, and COLO680 cells. High-level expression of NKD2 was detected in KYSE450 and TE8 cells. The methylation status of the NKD2 promoter was examined by MSP. Complete methylation was found in KYSE150 and TE1 cells. Partial methylation was detected in KYSE30, KYSE70, KYSE410, KYSE140, and COLO680 cells. Unmethylation was observed in KYSE450 and TE8 cells (Fig. 1B). These results demonstrate that loss or reduced expression of NKD2 was correlated with promoter region methylation in human esophageal cancer cells. To further reveal the methylation density and validate the MSP results, the BSSQ technique was used. As shown in Figure 1C, NKD2 was completely methylated in KYSE150 and TE1 cells, partially methylated in KYSE410 cells, and unmethylated in KYSE450 cells and normal esophageal mucosa. These results are consistent with the MSP results (Fig. 1C). To further analyze whether NKD2 expression is regulated by promoter region methylation, KYSE450, KYSE30, KYSE150, KYSE70, TE8, KYSE410, TE1, KYSE140, and COLO680 cells were treated with 5-AZA, a demethylating reagent. Restoration of NKD2 expression was induced by 5-AZA in KYSE150 and TE1 cells. Increased expression of NKD2 was observed in KYSE30, KYSE70, KYSE410, KYSE140, and COLO680 cells treated with 5-AZA, whereas no expression changes were found in KYSE450 and TE8 cells before and after 5-AZA treatment (Fig. 1A). These results suggest that the expression of NKD2 is regulated by promoter region methylation in human esophageal cancer cells.
      Figure thumbnail gr1
      Figure 1The expression and methylation status of naked cuticle homolog 2 (NKD2) in esophageal cancer cells and normal esophageal mucosa (NE). (A) Semiquantitative reverse-transcriptase polymerase chain reaction (RT-PCR) shows NKD2 expression levels in esophageal cancer cell lines. KYSE450, KYSE30, KYSE150, KESE70, TE8, KYSE410, TE1, KYSE140, and COLO680 are esophageal cancer cell lines. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is the internal control of RT-PCR. H2O is double-distilled water. Minus sign indicates absence of 5-aza-2′-deoxycytidine (5-AZA) and plus sign indicates presence of 5-AZA. (B) Methylation-specific polymerase chain reaction (MSP) results of NKD2 in esophageal cancer cell lines. U refers to unmethylated alleles, and M refers to methylated alleles. In vitro methylated DNA (IVD) serves as a methylation control; normal peripheral lymphocyte DNA (NL) serves as an unmethylation control. H2O is double-distilled water. (C) Bisulfite sequencing (BSSQ) results of NKD2. KYSE150, KYSE410, TE1, and KYSE450 are esophageal cancer cells. Double-headed arrow indicaets that the MSP PCR product spanned 103 base pairs (bp) in NKD2. Bisulfite sequencing focused on a 287-bp region of the CpG island (–287 bp to +38 bp) across the NKD2 transcription start site (TSS). Filled circles are methylated CpG sites, and open circles are unmethylated CpG sites.

      NKD2 Is Frequently Methylated in Primary Human Esophageal Cancer

      To further explore the methylation status of NKD2 in primary human esophageal cancer, the methylation status was examined by MSP in tissue samples from 154 cases of esophageal cancer and 15 samples of normal esophageal mucosa from patients without cancer. NKD2 was methylated in 53.2% of primary esophageal cancer samples (82 of 154), and no methylation was detected in normal esophageal mucosa (Fig. 2A and B). As shown in Table 1, NKD2 methylation was significantly associated with TNM stage and lymph node metastasis (both p < 0.01), but no association was found between NKD2 methylation and age, sex, tumor size, or differentiation (all p > 0.05). To further validate that NKD2 expression is regulated by promoter region methylation, 30 cases of available matched esophageal cancer and adjacent tissue paraffin samples were evaluated by IHC analysis. NKD2 staining was observed mainly in the cytoplasm of the esophagus. NKD2 is highly expressed in adjacent tissue samples and reduced in primary cancer tissue samples (Fig. 2C and D). Reduced expression of NKD2 is associated with promoter region hypermethylation (p < 0.01 [Fig. 2E]). These results demonstrate that NKD2 is regulated by promoter region methylation in primary esophageal cancer.
      Figure thumbnail gr2
      Figure 2Methylation status and expression of naked cuticle homolog 2 (NKD2) in primary esophageal cancer samples. (A) Methylation-specific polymerase chain reaction (MSP) results of NKD2 in normal esophageal mucosa (NE). (B) Representative results of MSP for NKD2 in primary esophageal cancer samples (EC). (C) Representative immunohistochemical analysis results showing NKD2 expression in esophageal cancer and matched adjacent tissue samples (upper images: ×100; lower images: ×400). (D) NKD2 expression scores are shown as box plots, with horizontal lines representing the median score; the bottom and top of the boxes represent the 25th and 75th percentiles, respectively, and vertical bars represent the range of data. The expression level of NKD2 was significantly different between adjacent tissue and esophageal cancer samples (***p < 0.001). (E) The bar diagram shows the expression and DNA methylation status of NKD2 in different cancer samples. Reduced expression of NKD2 was significantly associated with promoter region methylation (**p < 0.01).
      Table 1Clinical Factors and NKD2 Methylation in 154 Cases of Esophageal Cancer
      Clinical FactornNKD2 Methylation Statusp Value
      Methylated

      n = 82 (53.2%)
      Unmethylated

      n = 72 (46.8%)
      Age, y
       <501055p = 1.0
       ≥501447767
      Sex
       Male1166551p = 0.23
       Female381721
      Tumor size, cm
       <5995049p = 0.36
       ≥5553223
      Differentiation
       Well963p = 0.38
       Moderate975443
       Poor482226
      TNM stage
       Ⅰ + Ⅱ1044757p < 0.01
       Ⅲ + Ⅳ503515
      Lymph node metastasis
       N0903951p < 0.01
       N1644321
      Note: p Values are obtained from chi-square test and Fisher’s exact test; p less than 0.05 indicates a significant difference.
      NKD2, naked cuticle homolog 2; TNM, tumor, node, and metastasis.

      Restoration of NKD2 Expression Suppresses Cell Proliferation and Induces G1/S Arrest in Esophageal Cancer Cells

      To evaluate the effects of NKD2 on cell proliferation, cell viability was detected by methyl thiazolyl tetrazolium and colony formation assays. The optical density values were 0.892 ± 0.027 versus 0.763 ± 0.024 (p < 0.05) in KYSE150 cells and 0.551 ± 0.024 versus 0.438 ± 0.011 (p < 0.001) in TE1 cells before and after restoration of NKD2 expression (Fig. 3A). The results demonstrated that NKD2 inhibited esophageal cancer cell viability. The effect of NKD2 on cell proliferation was evaluated by colony formation assays. The clone numbers were 135.3 ± 6.8 versus 57.7 ± 4.0 (p <0.001) in KYSE150 cells and 58.3 ± 4.7 versus 29.7 ± 3.5 (p <0.001) in TE1 cells before and after restoration of NKD2 expression (Fig. 3B). These results suggest that NKD2 suppresses esophageal cancer cell growth.
      Figure thumbnail gr3
      Figure 3Naked cuticle homolog 2 (NKD2) inhibits esophageal cancer cell proliferation. (A) Growth curves represent the cell viability analyzed by the methyl thiazolyl tetrazolium assay in KYSE150 and TE1 cells in which NKD2 was reexpressed and unexpressed. The experiments were performed in triplicate (*p < 0.05 and ***p < 0.001). (B) Colony formation results show that colony number was reduced by reexpression of NKD2 in KYSE150 and TE1 cells. Each experiment was repeated three times. The average number of tumor clones is represented by a bar diagram (***p < 0.001). (C) Cell phase distribution in KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed. The ratio is presented by a bar diagram. Each experiment was repeated three times (**p < 0.01 and ***p < 0.001). G0/G1, G0/G1 phase; S, S phase; G2/M, G2/M phase.
      To further understand the mechanism of NKD2 in development of esophageal cancer, the role of NKD2 in the cell cycle was analyzed by flow cytometry. In KYSE150 cells, the cell phase distributions before and after reexpression of NKD2 were as follows: G0/1 phase: 33.46 ± 0.58% versus 41.82 ± 1.73%; S phase: 46.17 ± 2.21% versus 37.15 ± 1.46%; and G2/M phase: 20.37 ± 2.21% versus 21.03 ± 0.31%. The percentage of cells in G0/1 phase increased significantly whereas the percentage of cells in S phase decreased significantly after reexpression of NKD2 (all p < 0.01).
      In TE1 cells, the cell phase distributions before and after reexpression of NKD2 were as follows: G0/1 phase: 44.13 ± 2.60% versus 61.73 ± 1.28%; S phase: 44.21 ± 3.88% versus 22.93 ± 1.77%; and G2/M phase: 11.67 ± 1.38% versus 15.34 ± 0.82% (Fig. 3C). The percentage of cells in G0/1 phase increased whereas the percentage of cells in S phase decreased significantly after reexpression of NKD2 in TE1 cells (all p < 0.001). These results suggest that NKD2 induced G1/S checkpoint arrest in esophageal cancer cells.

      Restoration of NKD2 Expression Inhibits Cell Migration and Invasion in Human Esophageal Cancer Cells

      The transwell assay was performed in the absence of ECM gel coating to explore the effects of NKD2 on cell migration. The numbers of migrated cells for each high-power field under the microscope were 105.7 ± 5.1 versus 63.0 ± 4.0 in KYSE150 cells and 147 .0 ± 6.6 versus 52.3 ± 5.7 in TE1 cells before and after restoration of NKD2 expression. The cell numbers were reduced significantly after reexpression of NKD2 in esophageal cancer cells (all p < 0.001, Fig. 4A). These results demonstrate that NKD2 inhibits esophageal cancer cell migration.
      Figure thumbnail gr4
      Figure 4Restoration of naked cuticle homolog 2 (NKD2) expression inhibits cell migration and invasion. (A) Cell migration in KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed. The ratio is presented by a bar diagram. Each experiment was repeated three times (***p < 0.001). (B) Cell invasion in KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed. The ratio is presented by a bar diagram. Each experiment was repeated three times (***p < 0.001). (C) The expression levels of NKD2, matrix metalloproteinase-2 (MMP-2), matrix metalloproteinase-7 (MMP-7), and matrix metalloproteinase-9 (MMP-9) were detected by Western blot in KYSE150 and TE1 cells in which NKD2 was unexpressed and reexpressed. Knockdown of NKD2 by small interfering RNA (siRNA) was performed to validate the results in KYSE450 cells in which NKD2 was highly expressed. (D) Representative images of immunohistochemical analysis for NKD2, MMP2, and MMP9 in human esophageal squamous cell cancer. The expression levels of NKD2, MMP2, and MMP9 were evaluated by the German semiquantitative scoring system. The correlation of NKD2 and MMP2, or NKD2 and MMP9, was analyzed by Pearson correlation coefficient. The x axis represents levels of NKD2 and the y axis represents levels of MMP2 or MMP9. NC, KYSE450 cells in which NKD2 is highly expressed; SiNKD2, KYSE450 cells in which NKD2 is knocked down by siRNA.
      Next, the transwell assay with ECM coating was used to evaluate the effects of NKD2 on cell invasion. The numbers of invasive cells for each high-power field under the microscope were 114.7 ± 4.5 versus 79.7 ± 4.5 in KYSE150 cells and 137.0 ± 4.0 versus 65.0 ± 2.7 in TE1 cells before and after restoration of NKD2 expression. The cell numbers were reduced significantly after reexpression of NKD2 in KYSE150 and TE1 cells (all p < 0.001 [Fig. 4B]). These results suggest that NKD2 impedes esophageal cancer cell invasion.
      To further understand the mechanism of NKD2 in esophageal cancer migration and invasion, the expression levels of MMP2, MMP7, and MMP9 were detected by Western blot. As shown in Figure 4C, the expression levels of MMP2, MMP7, and MMP9 were reduced after reexpression of NKD2 in KYSE150 and TE1 cells. The inhibitory role of NKD2 on MMP2, MMP7, and MMP9 expression was further validated by knockdown of NKD2 in KYSE450 cells. Taken together, the aforementioned results suggest that NKD2 suppresses esophageal cancer cell migration and invasion. The expression levels of MMP2 and MMP9 in 30 cases of available human esophageal cancer samples were evaluated by IHC analysis. As shown in Figure 4D, the expression levels of MMP2 and MMP9 were negatively correlated with the expression levels of NKD2 in human esophageal cancer, with correlation coefficients of –0.417 (p < 0.05) and –0.618 (p < 0.001), respectively.

      NKD2 Inhibits Wnt/β-Catenin Signaling in Esophageal Cancer

      NKD2 has been reported to negatively regulate canonical Wnt signaling in multiple tumors.
      • Zhao S.
      • Kurenbekova L.
      • Gao Y.
      • et al.
      NKD2, a negative regulator of Wnt signaling, suppresses tumor growth and metastasis in osteosarcoma.
      • Dong Y.
      • Cao B.
      • Zhang M.
      • et al.
      Epigenetic silencing of NKD2, a major component of Wnt signaling, promotes breast cancer growth.
      To determine whether the canonical Wnt signaling pathway is regulated by NKD2 in human esophageal cancer, key downstream components of the Wnt signaling pathway were detected by Western blotting. The level of β-catenin was reduced and the level of phospho-β-catenin was increased after reexpression of NKD2 in KYSE150 and TE1 cells. The expression levels of the Wnt signaling targeting-genes, homo sapiens C-MYC proto-oncogene (c-myc), and homo sapiens cyclin D1 (CCND1, cyclin D1), were reduced after reexpression of NKD2 in KYSE150 and TE1 cells (Fig. 5A). These results demonstrate that NKD2 inhibits Wnt signaling in human esophageal cancer. To further validate the role of NKD2 on the Wnt signaling pathway, an siRNA knockdown technique was used. The expression levels of β-catenin, c-myc, and cyclinD1 were increased and the level of phospho-β-catenin was reduced after knockdown of NKD2 in KYSE450 cells (Fig. 5B). These results suggest that NKD2 represses proliferation of esophageal cancer cells by inhibiting Wnt signaling.
      Figure thumbnail gr5
      Figure 5Naked cuticle homolog 2 (NKD2) inhibits canonical Wnt signaling in human esophageal cancer cells. (A) The expression levels of β-catenin, cyclin D1, and c-myc were reduced, and the level of phosphorylated β-catenin (p-β-catenin) increased after reexpression of NKD2 in KYSE150 and TE1 cells. (B) The level of p-β-catenin was reduced and the expression levels of β-catenin, c-myc, and cyclin D1 increased after knockdown of NKD2 by small interfering RNA (siRNA) in KYSE450 cells. (C) Representative images of immunohistochemical analysis for NKD2 and phosphorylated β-catenin (p-β-catenin) in human esophageal squamous cell cancer. The expression levels of NKD2 and p-β-catenin were evaluated by the German semiquantitative scoring system. The correlation of NKD2 and p-β-catenin was analyzed with Pearson's correlation coefficient. The x axis represents levels of NKD2, and the y axis represents levels of p-β-catenin. NC, KYSE450 cells in which NKD2 is highly expressed; SiNKD2, KYSE450 cells in which NKD2 has been knocked down by siRNA.
      We further evaluated the expression of phospho-β-catenin in available samples from 30 cases of primary esophageal cancer by IHC analysis. The expression levels of phospho-β-catenin were correlated with the expression levels of NKD2 in human esophageal cancer, with a correlation coefficient of 0.798 (p < 0.001 [Fig. 5C]).

      NKD2 Suppresses Tumor Growth in Esophageal Cancer Cell Xenograft Mice

      To further validate the effects of NKD2 in esophageal cancer in vivo, KYSE150 cell xenograft mouse models in which NKD2 was unexpressed and reexpressed, as well as KYSE450 cell xenograft mouse models in which NKD2 was highly expressed and knocked down, were used (Fig. 6A). The mean tumor volume was 345.12 ± 18.42 mm3 in KYSE150 cells in which NKD2 was unexpressed and 96.78 ± 17.29 mm3 in KYSE150 cells in which NKD2 was reexpressed. The tumor volume was significantly smaller in KYSE150 cell xenograft mice in which NKD2 was reexpressed compared with in KYSE150 cell xenograft mice in which NKD2 was unexpressed (p < 0.001). The tumor volumes were 68.62 ± 7.39 mm3 in KYSE450 cells in which NKD2 was expressed and 113.38 ± 17.31 mm3 in KYSE450 cells xenograft mice in which NKD2 was knocked down. The tumor volume was significantly smaller in KYSE450 cell xenograft mice in which NKD2 was expressed compared with in KYSE450 cell xenograft mice in which NKD2 was knocked down (p < 0.001 [Fig. 6A and B]). The tumor weights were 220.62 ± 28.51 mg in KYSE150 cell xenograft mice in which NKD2 was unexpressed and 22.35 ± 5.19 mg in KYSE150 cell xenograft mice in which NKD2 was reexpressed. The tumor weight was significantly lower in KYSE150 cell xenograft mice in which NKD2 was reexpressed than in KYSE150 cell xenograft mice in which NKD2 was unexpressed (p < 0.001). The tumor weights were 13.77 ± 5.40 mg in KYSE450 cell xenograft mice in which NKD2 was expressed and 56.12 ± 11.86 mg in KYSE450 cell xenograft mice in which NKD2 was knocked down. The tumor weight was significantly lower in KYSE450 cell xenograft mice in which NKD2 was highly expressed than in KYSE450 cell xenograft mice in which NKD2 was knocked down (p < 0.001 [Fig. 6C]). To further validate whether NKD2 inhibits Wnt signaling in vivo, the expression levels of NKD2, β-catenin, and phospho-β-catenin were detected by IHC staining. The levels of phospho-β-catenin were increased and the levels of β-catenin were reduced in NKD2-expressing esophageal cancer cell xenografts (Fig. 6D). These results suggest that NKD2 suppresses esophageal cancer cell growth by inhibiting Wnt signaling in vivo.
      Figure thumbnail gr6
      Figure 6Naked cuticle homolog 2 (NKD2) suppresses esophageal cancer cell growth in xenograft mice. (A) Representative burdened nude mice with NKD2 reexpressed and unexpressed in KYSE150 cells and NKD2 expressed and knocked down in KYSE450 cells. The tumor locations are shown by red arrowhead. (B) Subcutaneous tumor growth curves for xenograft mice burdened with KYSE150 cells in which NKD2 is unexpressed and reexpressed and KYSE450 cells is which NKD2 is expressed and knocked down at different times (***p < 0.001). (C) Tumor weight in nude mice at the 24th day after inoculation with KYSE150 cells in which NKD2 is unexpressed and reexpressed and KYSE450 cells in which NKD2 is expressed and knocked down. Bars indicate mean of six mice (***p < 0.001). (D) Representative photographs of immunohistochemical analysis for NKD2, β-catenin, and p-β-catenin in xenografts. Staining of NKD2 and p-β-catenin was found in KYSE150 cell xenografts in which NKD2 is reexpressed. Staining of total β-catenin was reduced in KYSE150 cell xenografts in which NKD2 is reexpressed. Magnification: ×400. NC, KYSE450 cells in which NKD2 is highly expressed; SiNKD2, KYSE450 cells in which NKD2 has been knocked down by small interfering RNA.

      Discussion

      NKD2 was reported to be rarely mutated in human ESCC.
      • Dulak A.M.
      • Stojanov P.
      • Peng S.
      • et al.
      Exome and whole-genome sequencing of esophageal adenocarcinoma identifies recurrent driver events and mutational complexity.
      • Gao Y.B.
      • Chen Z.L.
      • Li J.G.
      • et al.
      Genetic landscape of esophageal squamous cell carcinoma.
      In our study, no mutations were found in nine esophageal cancer cell lines and 154 patients with ESCC. We demonstrated that NKD2 is frequently methylated in human ESCC and that expression of NKD2 is regulated by promoter region methylation. These results are supported by data from The Cancer Genome Atlas. Both DNA methylation and NKD2 expression were analyzed in tissue from 185 cases of esophageal cancer. The promoter region methylation was inversely associated with NKD2 expression in esophageal cancer (R = –0.34, p < 0.0001 [Supplementary Fig. 2]). NKD2 methylation is associated with TNM stage and lymph node metastasis, suggesting that NKD2 methylation may serve as a poor prognostic marker in human ESCC. In regard to information limitations, we did not analyze the influence of smoking and drug consumption in NKD2 methylation. Our further studies revealed that NKD2 inhibits esophageal cancer cell proliferation and colony formation and induces G1/S checkpoint arrest. In addition, NKD2 suppresses esophageal cancer cell migration and invasion. These results demonstrate that NKD2 is involved in progression and metastasis of esophageal cancer. The role of NKD2 in suppression of esophageal cancer growth was validated by an esophageal cancer cell xenograft model in vivo.
      We further explored the mechanism by which NKD2 suppresses progression and metastasis of esophageal cancer. NKD2 impedes ESCC metastasis by down-regulating MMP2, MMP7, and MMP9 expression, and it suppresses ESCC growth by inhibiting Wnt signaling. Degradation of the ECM and destruction of the basement membrane by cancer cells are important processes for direct invasion. There are three kinds of enzymes that effectively degrade the ECM: MMPs, serine proteinases, and cysteine proteinases. MMPs are known to play important roles in ECM remodeling during the process of tumor invasion and metastasis. The expression of MMPs was reported to be associated with tumor invasion and lymph node metastasis in ESCC.
      • Zeng R.
      • Duan L.
      • Kong Y.
      • et al.
      Clinicopathological and prognostic role of MMP-9 in esophageal squamous cell carcinoma: a meta-analysis.
      • Li Y.
      • Ma J.
      • Guo Q.
      • et al.
      Overexpression of MMP-2 and MMP-9 in esophageal squamous cell carcinoma.
      • Xu Y.B.
      • Du Q.H.
      • Zhang M.Y.
      • et al.
      Propofol suppresses proliferation, invasion and angiogenesis by down-regulating ERK-VEGF/MMP-9 signaling in Eca-109 esophageal squamous cell carcinoma cells.
      Our results demonstrate that the expression levels of NKD2 were negatively correlated with the expression levels of MMPs both in vitro and in vivo. These results suggest that NKD2 suppresses ESCC metastasis by inhibiting MMPs.
      The Wnt signaling pathway is conserved in various organisms, and it plays important roles in development and cellular proliferation and differentiation. Wnt signaling regulates various processes that are important for cancer development, including tumor initiation, progression, and metastasis.
      • Anastas J.N.
      • Moon R.T.
      WNT signalling pathways as therapeutic targets in cancer.
      Wnt signaling is suggested to inhibit β-catenin phosphorylation, thus inducing the accumulation of cytosolic β-catenin. Then, β-catenin is translocated into the nucleus, where it stimulates the expression of Wnt/β-catenin–responsive genes, including myc and cyclinD1.
      • Kikuchi A.
      Regulation of beta-catenin signaling in the Wnt pathway.
      • Rennoll S.
      • Yochum G.
      Regulation of MYC gene expression by aberrant Wnt/beta-catenin signaling in colorectal cancer.
      • Behrens J.
      • von Kries J.P.
      • Kuhl M.
      • et al.
      Functional interaction of beta-catenin with the transcription factor LEF-1.
      • Huber O.
      • Korn R.
      • McLaughlin J.
      • et al.
      Nuclear localization of beta-catenin by interaction with transcription factor LEF-1.
      • Molenaar M.
      • van de Wetering M.
      • Oosterwegel M.
      • et al.
      XTcf-3 transcription factor mediates beta-catenin-induced axis formation in Xenopus embryos.
      The c-myc proto-oncogene (myc), a Wnt/β-catenin target gene, may promote cell cycle progression.
      • He T.C.
      • Sparks A.B.
      • Rago C.
      • et al.
      Identification of c-MYC as a target of the APC pathway.
      • Oster S.K.
      • Ho C.S.
      • Soucie E.L.
      • Penn L.Z.
      The myc oncogene: marvelously complex.
      In most circumstances, cyclin D associates with Cdk4 and Cdk6 during early G1 phase.
      • King K.L.
      • Cidlowski J.A.
      Cell cycle regulation and apoptosis.
      In this study, we found that NKD2 represses proliferation of esophageal cancer cells by inhibiting Wnt signaling in human esophageal cancer cells both in vitro and in vivo when xenograft mouse models are used. In human primary esophageal cancer, the expression levels of phospho-β-catenin were correlated with the expression levels of NKD2. The aforementioned results strongly suggest that NKD2 represses esophageal cancer growth by inhibiting Wnt signaling in human esophageal cancer.
      At the time of diagnosis, more than 50% of patients with esophageal cancer have metastatic disease.
      • Wu A.H.
      • Wan P.
      • Bernstein L.
      A multiethnic population-based study of smoking, alcohol and body size and risk of adenocarcinomas of the stomach and esophagus (United States).
      Although there are many approaches to treating metastatic disease, the overall survival remains poor. Understanding the molecular events in ESCC may improve therapeutic strategies.
      • Guo M.
      • Liu S.
      • Herman J.G.
      • et al.
      Gefitinib-sensitizing mutation in esophageal carcinoma cell line Kyse450.
      • Guo M.
      • Liu S.
      • Lu F.
      Gefitinib-sensitizing mutations in esophageal carcinoma.
      Our findings provide more clues for epigenetic-based personalized medicine in esophageal cancer.

      Conclusion

      NKD2 is frequently methylated in human esophageal cancer, and the expression of NKD2 is regulated by promoter region methylation. Methylation of NKD2 is associated with TNM stage and lymph node metastasis. NKD2 suppresses growth of human esophageal cancer by inhibiting Wnt signaling.

      Acknowledgments

      This work was supported by the following grants: National Basic Research Program of China (973 Program No. 2012CB934002 , 2015CB553904 ), National High-Tech R&D Program of China (863 Program No. SS2012AA020314 , SS2012AA020821 , and SS2012AA020303 ), National Key Scientific Instrument Special Program of China (Grant No. 2011YQ03013405 ), and National Science Foundation of China (NSFC No. 81402345 , 81121004 , 81161120432 , 81490753 , and 81401950 ). Drs. Cao and Yang performed the experiments, analyzed the data, and wrote the manuscript. Drs. Jin, Zhang, He, and Zhan provided feedback and experimental advice. Drs. Herman and Zhong provided experimental advice and manuscript editing. Dr. Guo conceived the study design, supervised the experiments and edited the manuscript. All authors approved the final version of the submitted manuscript.

      References

        • Ferlay J.
        • Soerjomataram I.
        • Dikshit R.
        • et al.
        Cancer incidence and mortality worldwide: sources, methods and major patterns in GLOBOCAN 2012.
        Int J Cancer. 2015; 136: E359-E386
        • Enzinger P.C.
        • Mayer R.J.
        Esophageal cancer.
        N Engl J Med. 2003; 349: 2241-2252
        • Pennathur A.
        • Farkas A.
        • Krasinskas A.M.
        • et al.
        Esophagectomy for T1 esophageal cancer: outcomes in 100 patients and implications for endoscopic therapy.
        Ann Thorac Surg. 2009; 87 ([discussion 1054–1045]): 1048-1054
        • Pennathur A.
        • Gibson M.K.
        • Jobe B.A.
        • et al.
        Oesophageal carcinoma.
        Lancet. 2013; 381: 400-412
        • De Stefani E.
        • Barrios E.
        • Fierro L.
        Black (air-cured) and blond (flue-cured) tobacco and cancer risk. III: oesophageal cancer.
        Eur J Cancer. 1993; 29A: 763-766
        • Lee C.H.
        • Wu D.C.
        • Lee J.M.
        • et al.
        Carcinogenetic impact of alcohol intake on squamous cell carcinoma risk of the oesophagus in relation to tobacco smoking.
        Eur J Cancer. 2007; 43: 1188-1199
        • Vaughan T.L.
        • Davis S.
        • Kristal A.
        • et al.
        Obesity, alcohol, and tobacco as risk factors for cancers of the esophagus and gastric cardia: adenocarcinoma versus squamous cell carcinoma.
        Cancer Epidemiol Biomarkers Prev. 1995; 4: 85-92
        • Wu C.
        • Kraft P.
        • Zhai K.
        • et al.
        Genome-wide association analyses of esophageal squamous cell carcinoma in Chinese identify multiple susceptibility loci and gene-environment interactions.
        Nat Genet. 2012; 44: 1090-1097
        • Chuang S.C.
        • Hashibe M.
        • Scelo G.
        • et al.
        Risk of second primary cancer among esophageal cancer patients: a pooled analysis of 13 cancer registries.
        Cancer Epidemiol Biomarkers Prev. 2008; 17: 1543-1549
        • Chuang S.C.
        • Scelo G.
        • Tonita J.M.
        • et al.
        Risk of second primary cancer among patients with head and neck cancers: a pooled analysis of 13 cancer registries.
        Int J Cancer. 2008; 123: 2390-2396
        • Ahrens T.D.
        • Werner M.
        • Lassmann S.
        Epigenetics in esophageal cancers.
        Cell Tissue Res. 2014; 356: 643-655
        • Song Y.
        • Li L.
        • Ou Y.
        • et al.
        Identification of genomic alterations in oesophageal squamous cell cancer.
        Nature. 2014; 509: 91-95
        • Yun T.
        • Liu Y.
        • Gao D.
        • et al.
        Methylation of CHFR sensitizes esophageal squamous cell cancer to docetaxel and paclitaxel.
        Genes Cancer. 2015; 6: 38-48
        • Wu L.
        • Herman J.G.
        • Brock M.V.
        • et al.
        Silencing DACH1 promotes esophageal cancer growth by inhibiting TGF-beta signaling.
        PLoS One. 2014; 9: e95509
        • Lu D.
        • Ma J.
        • Zhan Q.
        • et al.
        Epigenetic silencing of RASSF10 promotes tumor growth in esophageal squamous cell carcinoma.
        Discov Med. 2014; 17: 169-178
        • Jiang S.
        • Linghu E.
        • Zhan Q.
        • et al.
        Methylation of ZNF331 promotes cell invasion and migration in human esophageal cancer.
        Curr Protein Pept Sci. 2015; 16: 322-328
        • Jia Y.
        • Yang Y.
        • Zhan Q.
        • et al.
        Inhibition of SOX17 by microRNA 141 and methylation activates the WNT signaling pathway in esophageal cancer.
        J Mol Diagn. 2012; 14: 577-585
        • Jia Y.
        • Yang Y.
        • Brock M.V.
        • et al.
        Methylation of TFPI-2 is an early event of esophageal carcinogenesis.
        Epigenomics. 2012; 4: 135-146
        • Guo M.
        • Ren J.
        • House M.G.
        • et al.
        Accumulation of promoter methylation suggests epigenetic progression in squamous cell carcinoma of the esophagus.
        Clin Cancer Res. 2006; 12: 4515-4522
        • Guo M.
        • Ren J.
        • Brock M.V.
        • et al.
        Promoter methylation of HIN-1 in the progression to esophageal squamous cancer.
        Epigenetics. 2008; 3: 336-341
        • Guo M.
        • House M.G.
        • Suzuki H.
        • et al.
        Epigenetic silencing of CDX2 is a feature of squamous esophageal cancer.
        Int J Cancer. 2007; 121: 1219-1226
        • Guo M.
        • House M.G.
        • Akiyama Y.
        • et al.
        Hypermethylation of the GATA gene family in esophageal cancer.
        Int J Cancer. 2006; 119: 2078-2083
        • Chen X.Y.
        • He Q.Y.
        • Guo M.Z.
        XAF1 is frequently methylated in human esophageal cancer.
        World J Gastroenterol. 2012; 18: 2844-2849
        • Brock M.V.
        • Gou M.
        • Akiyama Y.
        • et al.
        Prognostic importance of promoter hypermethylation of multiple genes in esophageal adenocarcinoma.
        Clin Cancer Res. 2003; 9: 2912-2919
        • He G.
        • Guan X.
        • Chen X.
        • et al.
        Expression and splice variant analysis of human TCF4 transcription factor in esophageal cancer.
        J Cancer. 2015; 6: 333-341
        • Mohamed A.
        • El-Rayes B.
        • Khuri F.R.
        • et al.
        Targeted therapies in metastatic esophageal cancer: advances over the past decade.
        Crit Rev Oncol Hematol. 2014; 91: 186-196
        • Herman J.G.
        • Baylin S.B.
        Gene silencing in cancer in association with promoter hypermethylation.
        N Engl J Med. 2003; 349: 2042-2054
        • Baylin S.B.
        • Esteller M.
        • Rountree M.R.
        • et al.
        Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer.
        Hum Mol Genet. 2001; 10: 687-692
        • Argos M.
        • Kibriya M.G.
        • Jasmine F.
        • et al.
        Genomewide scan for loss of heterozygosity and chromosomal amplification in breast carcinoma using single-nucleotide polymorphism arrays.
        Cancer Genet Cytogenet. 2008; 182: 69-74
        • Sheu J.C.
        • Lin Y.W.
        • Chou H.C.
        • et al.
        Loss of heterozygosity and microsatellite instability in hepatocellular carcinoma in Taiwan.
        Br J Cancer. 1999; 80: 468-476
        • Arias-Pulido H.
        • Narayan G.
        • Vargas H.
        • et al.
        Mapping common deleted regions on 5p15 in cervical carcinoma and their occurrence in precancerous lesions.
        Mol Cancer. 2002; 1: 3
        • Xu S.F.
        • Peng Z.H.
        • Li D.P.
        • et al.
        Refinement of heterozygosity loss on chromosome 5p15 in sporadic colorectal cancer.
        World J Gastroenterol. 2003; 9: 1713-1718
        • Lu Y.
        • Yu Y.
        • Zhu Z.
        • et al.
        Identification of a new target region by loss of heterozygosity at 5p15.33 in sporadic gastric carcinomas: genotype and phenotype related.
        Cancer Lett. 2005; 224: 329-337
        • Van Raay T.J.
        • Coffey R.J.
        • Solnica-Krezel L.
        Zebrafish Naked1 and Naked2 antagonize both canonical and non-canonical Wnt signaling.
        Dev Biol. 2007; 309: 151-168
        • Katoh M.
        Molecular cloning, gene structure, and expression analyses of NKD1 and NKD2.
        Int J Oncol. 2001; 19: 963-969
        • Yan D.
        • Wallingford J.B.
        • Sun T.Q.
        • et al.
        Cell autonomous regulation of multiple Dishevelled-dependent pathways by mammalian Nkd.
        Proc Natl Acad Sci U S A. 2001; 98: 3802-3807
        • Rousset R.
        • Mack J.A.
        • Wharton Jr., K.A.
        • et al.
        Naked cuticle targets dishevelled to antagonize Wnt signal transduction.
        Genes Dev. 2001; 15: 658-671
        • Zeng W.
        • Wharton Jr., K.A.
        • Mack J.A.
        • et al.
        Naked cuticle encodes an inducible antagonist of Wnt signalling.
        Nature. 2000; 403: 789-795
        • Li C.
        • Franklin J.L.
        • Graves-Deal R.
        • et al.
        Myristoylated Naked2 escorts transforming growth factor alpha to the basolateral plasma membrane of polarized epithelial cells.
        Proc Natl Acad Sci U S A. 2004; 101: 5571-5576
        • Hu T.
        • Krezel A.M.
        • Li C.
        • et al.
        Structural studies of human Naked2: a biologically active intrinsically unstructured protein.
        Biochem Biophys Res Commun. 2006; 350: 911-915
        • Zhao S.
        • Kurenbekova L.
        • Gao Y.
        • et al.
        NKD2, a negative regulator of Wnt signaling, suppresses tumor growth and metastasis in osteosarcoma.
        Oncogene. 2015; 34: 5069-5079
        • Dong Y.
        • Cao B.
        • Zhang M.
        • et al.
        Epigenetic silencing of NKD2, a major component of Wnt signaling, promotes breast cancer growth.
        Oncotarget. 2015; 6: 22126-22138
        • Cao B.
        • Yang Y.
        • Pan Y.
        • et al.
        Epigenetic silencing of CXCL14 induced colorectal cancer migration and invasion.
        Discov Med. 2013; 16: 137-147
        • Herman J.G.
        • Graff J.R.
        • Myohanen S.
        • et al.
        Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands.
        Proc Natl Acad Sci U S A. 1996; 93: 9821-9826
        • Yan W.
        • Wu K.
        • Herman J.G.
        • et al.
        Epigenetic regulation of DACH1, a novel Wnt signaling component in colorectal cancer.
        Epigenetics. 2013; 8: 1373-1383
        • Yu Y.
        • Yan W.
        • Liu X.
        • et al.
        DACT2 is frequently methylated in human gastric cancer and methylation of DACT2 activated Wnt signaling.
        Am J Cancer Res. 2014; 4: 710-724
        • Dulak A.M.
        • Stojanov P.
        • Peng S.
        • et al.
        Exome and whole-genome sequencing of esophageal adenocarcinoma identifies recurrent driver events and mutational complexity.
        Nat Genet. 2013; 45: 478-486
        • Gao Y.B.
        • Chen Z.L.
        • Li J.G.
        • et al.
        Genetic landscape of esophageal squamous cell carcinoma.
        Nat Genet. 2014; 46: 1097-1102
      1. National Cancer Institute Genomic Data Commons. TCGA data portal.. https://gdc-portal.nci.nih.gov/search/s?filters=%7B“op”:“and”,“content”:%5B%7B“op”:“in”,“content”:%7B“field”:“cases.project.disease_type”,“value”:%5B“Esophageal%20Carcinoma”%5D%7D%7D,%7B“op”:“in”,“content”:%7B“field”:“cases.project.project_id”,“value”:%5B“TCGA-ESCA”%5D%7D%7D%5D%7D. Accessed August 1, 2016.

        • Zeng R.
        • Duan L.
        • Kong Y.
        • et al.
        Clinicopathological and prognostic role of MMP-9 in esophageal squamous cell carcinoma: a meta-analysis.
        Chin J Cancer Res. 2013; 25: 637-645
        • Li Y.
        • Ma J.
        • Guo Q.
        • et al.
        Overexpression of MMP-2 and MMP-9 in esophageal squamous cell carcinoma.
        Dis Esophagus. 2009; 22: 664-667
        • Xu Y.B.
        • Du Q.H.
        • Zhang M.Y.
        • et al.
        Propofol suppresses proliferation, invasion and angiogenesis by down-regulating ERK-VEGF/MMP-9 signaling in Eca-109 esophageal squamous cell carcinoma cells.
        Eur Rev Med Pharmacol Sci. 2013; 17: 2486-2494
        • Anastas J.N.
        • Moon R.T.
        WNT signalling pathways as therapeutic targets in cancer.
        Nat Rev Cancer. 2013; 13: 11-26
        • Kikuchi A.
        Regulation of beta-catenin signaling in the Wnt pathway.
        Biochem Biophys Res Commun. 2000; 268: 243-248
        • Rennoll S.
        • Yochum G.
        Regulation of MYC gene expression by aberrant Wnt/beta-catenin signaling in colorectal cancer.
        World J Biol Chem. 2015; 6: 290-300
        • Behrens J.
        • von Kries J.P.
        • Kuhl M.
        • et al.
        Functional interaction of beta-catenin with the transcription factor LEF-1.
        Nature. 1996; 382: 638-642
        • Huber O.
        • Korn R.
        • McLaughlin J.
        • et al.
        Nuclear localization of beta-catenin by interaction with transcription factor LEF-1.
        Mech Dev. 1996; 59: 3-10
        • Molenaar M.
        • van de Wetering M.
        • Oosterwegel M.
        • et al.
        XTcf-3 transcription factor mediates beta-catenin-induced axis formation in Xenopus embryos.
        Cell. 1996; 86: 391-399
        • He T.C.
        • Sparks A.B.
        • Rago C.
        • et al.
        Identification of c-MYC as a target of the APC pathway.
        Science. 1998; 281: 1509-1512
        • Oster S.K.
        • Ho C.S.
        • Soucie E.L.
        • Penn L.Z.
        The myc oncogene: marvelously complex.
        Adv Cancer Res. 2002; 84: 81-154
        • King K.L.
        • Cidlowski J.A.
        Cell cycle regulation and apoptosis.
        Annu Rev Physiol. 1998; 60: 601-617
        • Wu A.H.
        • Wan P.
        • Bernstein L.
        A multiethnic population-based study of smoking, alcohol and body size and risk of adenocarcinomas of the stomach and esophagus (United States).
        Cancer Causes Control. 2001; 12: 721-732
        • Guo M.
        • Liu S.
        • Herman J.G.
        • et al.
        Gefitinib-sensitizing mutation in esophageal carcinoma cell line Kyse450.
        Cancer Biol Ther. 2006; 5: 152-155
        • Guo M.
        • Liu S.
        • Lu F.
        Gefitinib-sensitizing mutations in esophageal carcinoma.
        N Engl J Med. 2006; 354: 2193-2194