SEARCH

SEARCH BY CITATION

Keywords:

  • Epstein-Barr virus;
  • gastric cancer;
  • epigenome;
  • promoter methylation;
  • tumor suppressor gene

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

BACKGROUND:

Aberrant methylation of tumor-related genes has been reported in Epstein-Barr virus (EBV)-associated gastric cancers. This study sought to profile EBV-driven hypermethylation in EBV-infected cells.

METHODS:

The EBV-positive AGS gastric cancer cell line (AGS-EBV) and EBV-negative AGS cells were used in this study. DNA methyltransferase-3b (DNMT3b) activity was assessed by EpiQuick activity assay, and genome-wide DNA methylation profiles were assessed by methyl-DNA immunoprecipitation microarray assay.

RESULTS:

EBV infection was confirmed in AGS-EBV cells by EBV-encoded RNA in situ hybridization. Expression and activity of DNA methyltransferase-3b (DNMT3b) was significantly increased in AGS-EBV compared to AGS. Ectopic expression of LMP2A (latent membrane protein 2A) in AGS increased activity of DNMT3b. A total of 1065 genes were differentially methylated by EBV infection (fold-changes ≥ 2, P < .05) in AGS-EBV compared to AGS cells. The majority of the differentially methylated genes (83.2%, 886 of 1065 genes) had cytosine-guanine dinucleotide (CpG) hypermethylation in AGS-EBV (fold-changes 2.43∼65.2) versus that found in AGS cells. Gene ontology analysis revealed that hypermethylated genes were enriched in the important cancer pathways (≥ 10 genes each, P ≤ .05) including mitogen-activated protein kinase signaling, cell adhesion molecules, wnt signaling pathway, and so forth. Six novel hypermethylated candidates (IL15RA, REC8, SSTR1, EPHB6, MDGA2, and SCARF2) were further validated. Higher levels of DNA methylation were confirmed for all these genes in AGS-EBV cells by bisulfite genomic sequencing. Furthermore, these candidates were silenced or down-regulated in AGS-EBV cells, but can be restored by demethylation treatment.

CONCLUSIONS:

EBV infection in AGS cells induced aberrant CpG hypermethylation of 886 genes involving in important cancer-related pathways. Induction of promoter methylation by EBV is regulated by up-regulation of DNMT3b through LMP2A. Cancer 2013. © 2012 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Epstein-Barr virus (EBV) is a human herpesvirus that infects more than 90% of the world population before adolescence. This oncogenic virus has been identified in a wide variety of human malignancies, in particular, the epithelial malignancies.1, 2 EBV-associated gastric cancer accounts for 8% to 10% of gastric cancers3, 4 and is estimated to occur more than 90,000 patients annually.5 EBV is usually cytotoxic and induces cell death if EBV infects the gastric epithelial cells. However, once latent EBV infection is triggered, this persistent latent infection is established and initiates the progression into gastric cancer.

During persistent EBV infection, the cellular methylation system methylates the EBV genome as a host defense mechanism to suppress viral gene expression. However, overdriving DNA methylation mechanisms may also cause repression of cancer-related genes, resulting in transformation of host cells.6 A recent report demonstrated the increased activity of DNA methyltransferase 1 (DNMT1) in EBV-associated gastric cancer.7 Increasing evidences showed significantly higher frequencies of promoter cytosine-guanine dinucleotide (CpG) methylation of tumor suppressor genes (eg, p14, p15, p16, p73, APC, E-cadherin, CDH1, and PTEN) in EBV-associated gastric cancer than EBV-negative gastric cancer.8-10 EBV-driven aberrant promoter methylation will lead to the transcriptional silence of tumor suppressors in EBV-associated gastric cancer, thus resulting in uncontrolled cell expansive growth.11 Aberrant promoter methylation contributes to human gastric carcinogenesis. However, the contribution of EBV to methylation changes has not been fully clarified. In this study, we used a high-throughput technique, named methylated DNA immunoprecipitation coupled with hybridization on high-resolution microarray (MeDIP-chip), for a genome-wide evaluation of EBV-related methylation genes in gastric cancer using a well-established EBV-infected gastric cancer cell model (AGS-EBV)12 and EBV-negative AGS cells. A total of 886 genes were found to be CpG hypermethylated in AGS-EBV cells, as compared with AGS cells, and were associated with cancer pathways. Among them, 6 novel EBV-driven hypermethylated genes (IL15RA, REC8, SSTR1, EPHB6, MDGA2, and SCARF2) were demonstrated and further validated.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

Cancer Cell Lines and Culture Condition

The gastric cancer cell line AGS was purchased from the ATCC, Manassas, Va, and grown in Roswell Park Memorial Institute-1640 medium supplemented with 10% fetal bovine serum in a humidified incubator at 37°C. The AGS-EBV cell model was a gift from Dr. Shannon C. Kenney (Department of Oncology and Medicine, McArdle Laboratory for Cancer Research, University of Wisconsin School of Medicine and Public Health, Madison, Wis). AGS-EBV cells were maintained in F-12 medium supplemented with 10% fetal bovine serum and 100 ng/mL hygromycin (Invitrogen, Carlsbad, Calif).

Semiquantitative Reverse Transcription Polymerase Chain Reaction (RT-PCR) and Real-Time Quantitative RT-PCR

Total RNA was extracted from cell pellets using Trizol reagent (Invitrogen) according to the manufacturer's protocol. The complementary DNA was synthesized from 2 μg of total RNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Foster City, Calif). Semiquantitative RT-PCR and quantitative RT-PCR analyses were performed using primers listed in Table 1.

Table 1. Primers Used for Reverse Transcription Polymerase Chain Reaction and Bisulfite Genomic Sequencing in This Study
NameSequence (5′-3′)Application
  1. Abbreviations: BGS, bisulfite genomic sequencing; RT-PCR, reverse transcription polymerase chain reaction.

DNMT3b.FGAATTACTCACGCCCCAAGGART-PCR
DNMT3b.RACCGTGAGATGTCCCTCTTGTC
EPHB6.FCCCCGGACTGGAGAAGAC
EPHB6.RGAAGTCTGTGAATTGGGTGGA
IL15RA.FCCTTCAAAATCACCTTCCAC
IL15RA.RGCCTTGACTTGAGGTAGCAT
LMP2A.FCGGGATCACTCATCTGAACACATA
LMP2A.RGGCGGTCACAACGGTACTAACT
MDGA2.FGGGAACAAGAAAATGTTTGG
MDGA2.RTTGTCGTCAGGTAGCTCAAA
REC8.FGTTGTTCAGAACCCCAACTC
REC8.RAAGACACCATAAGGGGAACA
SCARF2.FACGACCTGGATAACACACTC
SCARF2.RCATGGGGTACACAGTACACA
SSTR1.FTAACAGACCGCACATGCACT
SSTR1.RCGGCTCTGGACTGGTAAATG
β_Actin.FGTCTTCCCCTCCATCGTG
β_Actin.RAGGGTGAGGATGCCTCTCTT
EPHB6-B.FGGGTGTTTGTGTTTTTTAGAAAAATTABGS
EPHB6-B.RCTACACCTCTAAATACCTCCTCCCT
IL15RA-B.FATATTTGGGGATTTGGAAATAATTT
IL15RA-B.RTATTTTCCTCCTTCCTCTAACTACC
MDGA2-B.FGGTTTAAGGTGGTAGTTGGTTTTTTA
MDGA2-B.RTTTACTCTACCTTCCTATATCAACC
REC8-B.FGTAAAATTTTTATGATTGGTTTGTTG
REC8-B.RCCCCTAAACCTTACACTAACT
SCARF2-B.FGGTTAGGTTAAGGTTGGGTTTTAGT
SCARF2-B.RCCTTCAAAAATACCTTTAAAAACTCC
SSTR1-B.FGAATGGTTTATTTAGGTTTTTGAATA
SSTR1-B.RCTAAAAACCCCAACTTCTCTC

Western Blot Analysis

Total proteins were extracted using CytoBuster Protein Extraction Reagent (Novagen, Madison, Wis). The protein concentration was measured by the Bradford method (Bio-Rad Laboratories, Hercules, Calif). Total extracted proteins (20 μg) from each sample were separated on 7.5% sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto nitrocellulose membrane (GE Healthcare, Piscataway, NJ). After blocking with 5% skimmed milk, the protein expression was assessed with primary antibodies: DNMT3b (Imgenex, San Diego, Calif), DNMT1 (Imgenex), latent membrane protein 2A (LMP2A; Abcam, Cambridge, UK), and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Santa Cruz Biotechnology, Santa Cruz, Calif). GAPDH was used as internal control. The proteins were visualized using ECL Detection Reagent (GE Healthcare).

LMP2A Plasmid Construct and Transfection

LMP2A plasmid was constructed as described13 and transfected into AGS cells using Lipofectamine 2000 reagent (Invitrogen). Transient transfectant of LMP2A plasmid was collected for RNA and protein extraction after 48 hours of incubation.

DNMT1 and DNMT3b Activity Assays

DNMT1 and DNMT3b activities were measured using EpiQuik DNA Methyltransferase 1 and 3B Activity/Inhibition Assay Kits according to the manufacturer's instructions (Epigentek, Brooklyn, NY). Briefly, 2 μg proteins isolated from AGS-EBV, AGS, or AGS transfected with LMP2A were incubated with methylation substrates for 2 hours at 37°C. After washing, the capture antibody was added and incubated for 1 hour, followed by detection antibody for additional 30 minutes at room temperature. DNMT1 and DNMT 3b activity was then read at 450 nm absorbance, or optical density (OD), using a SpectraMax Gemini dual-scanning microplate spectrophotometer (Bio-Tek Instruments, Winooski, Vt). Relative activity of DNMT1 or DNMT 3b was calculated as (treatment OD450 − blank OD450)/(control OD450 − blank OD450).

Genome-Wide Profiling of EBV-Driven DNA Methylation by MeDIP-Chip

DNA methylation profiling of AGS-EBV and AGS cells was performed using MeDIP-chip. We used the Agilent custom Human CpG Island Microarray and Human Promoter array, which has 1 million features per array covering ∼17,000 refined human genes (Agilent Technologies, Santa Clara, Calif). Briefly, 5 μg of genomic DNA was sonicated into fragments ranging from 300 to 1000 base pairs in 250 μL phosphate-buffered saline. Sheared DNA (50 μL) was then transferred into a fresh microtube as an input sample. The remaining fragmented DNA was denatured at 95°C for 10 minutes and immunoprecipitated at 4°C overnight with 5 μg of polyclonal antibody against 5-methylcytosine (Eurogentec SA, Liege, Belgium) per immunoprecipitation. The mixture was incubated with magnetic beads conjugated with anti-mouse immunoglobulin G (Life Technologies, Stafford, Tex) to bind anti-5-methylcytosine antibodies at 4°C overnight. The immunoprecipitated DNA (MeDIP-enriched DNA) and the input DNA were purified by phenol-chloroform extraction and labeled with Cyanine 5 (Cy5) and Cyanine 3 (Cy3), respectively. Cy5- and Cy3-labeled DNA samples were combined and hybridized on the Agilent high-resolution microarray. Microarray slides were scanned by GenePix 4000B scanner (Axon, Sunnyvale, Calif) and exported images were analyzed with GenePix Pro 6.0 (Axon). The methylation status of specific CpG sites was presented as the ratio of methylated DNA-Cy5 signal to unmethylated input DNA-Cy3 signal.14

Bioinformatics Analysis

Hierarchical clustering analysis of MeDIP-chip data was performed using Gene Cluster 3.0, with data adjusted using the “center arrays” method by subtracting the column-wise mean from the values in each column. Hierarchical clustering analysis result was displayed using Java TreeView software. Histograms of gene numbers according to log2 fold-change ranges were drawn using SPSS software. Involvement of hypermethylated genes in KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways was analyzed using the Genecodis2 Web-based tool (http://genecodis.cnb.csic.es).

Gastric Cancer Samples

Six EBV-positive and 5 EBV-negative gastric cancer samples were collected in the First Affiliated Hospital of Sun Yat-sen University, Guangzhou, China, from January 1999 to December 2006. The presence of EBV was determined by in situ hybridization analysis of EBV-encoded small RNA (EBER). This study was approved by the Clinical Research Ethics Committee of Sun Yat-sen University.

Bisulfite Genomic Sequencing and Combined Bisulfite Restriction Analysis

Genomic DNA from cell pellets and primary gastric cancer tissues was extracted using QIAamp DNA Mini Kit (Qiagen, Hilden, Germany). DNA (2 μg) was subjected to bisulfite modification using EZ DNA Methylation-Gold Kit (Zymo Research, Irvine, Calif) following the manufacturer's instructions. Primers for bisulfite genomic sequencing (BGS) PCR and combined bisulfite restriction analysis (COBRA) were designed using MethPrimer (http://www.urogene.org/methprimer/index1.html) and listed in Table 1. PCR products for BGS were confirmed by electrophoresis on 2% agarose gel and then sequenced directly. The sequencing data were analyzed by Bioedit (http://www.mbio.ncsu.edu/BioEdit/bioedit.html). PCR products for COBRA were digested with BstUI (New England Biolabs, Ipswich, Mass) at 60°C for 4 hours and then separated on 2% agarose gel.

Demethylation With 5-Aza-2′Deoxycytidine Agent Treatment

AGS-EBV cells were seeded at a density of 1 × 106 cells/mL in 100-mm dishes and grown for 24 hours. Cells were then treated with 2 μM 5-aza-2′-deoxycytidine (5-Aza) (Sigma-Aldrich, St Louis, Mo) for 5 days, with culture medium containing 5-Aza replenished every day. Cells were then harvested and gene expression was analyzed by RT-PCR.

Statistical Analysis

Data are presented as mean ± standard deviation. The difference of DNMT3b activity in 2 groups was analyzed by Mann-Whitney U test. Methylation percentage of the target genes was compared between AGS-EBV and AGS cells using the Student t test. A value of P < .05 was taken as statistical significance.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

EBV Infection in AGS-EBV Cell Model

Successful EBV infection in AGS-EBV cells was demonstrated by in situ hybridization detection of the viral gene EBER (EBV-encoded RNA) as described in our previous study.15 Expression of the key EBV viral gene LMP2A was examined in AGS-EBV and AGS cells by RT-PCR, with cell lines that were naturally infected with EBV (SNU719 and YCC10) as positive controls (Fig. 1A). Results indicated positive expression of LMP2A in AGS-EBV, SNU719, and YCC10 cells, but not in AGS cells, confirming successful EBV infection in AGS-EBV cells.

thumbnail image

Figure 1. Gels show DNA methyltransferase 3b (DNMT3b) activation by Epstein-Barr virus (EBV) infection via LMP2A expression. (A) Messenger RNA expression of latent membrane protein 2A (LMP2A) and DNMTs in EBV-positive AGS-EBV cells and EBV-negative AGS cells by reverse transcription polymerase chain reaction (RT-PCR). (B) Protein expression is shown for LMP2A, DNMT3b, and DNMT1 in AGS-EBV and AGS cells by western blot. (C) DNMT1 and DNMT3b activities in AGS-EBV and AGS cells were evaluated by EpiQuick DNMT1 and DNMT3b activity assays, respectively. (D through F) DNMT3b expression, DNMT1 and DNMT3b activities were evaluated in LMP2A-transfected AGS cells in comparison with empty vector-transfected AGS cells by (D) RT-PCR, (E) western blot, and (F) DNMT1/DNMT3b activity assays. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) was used as a control.

Download figure to PowerPoint

Activation of DNMT3b in AGS-EBV Cells

In order to determine whether EBV infection has an effect on DNMTs, expression of DNMT1, DNMT3a, and DNMT3b were compared between AGS-EBV and AGS cells by semiquantitative RT-PCR. The results indicated up-regulation of DNMT1 and DNMT3b, but not DNMT3a in AGS-EBV cells as compared with AGS cells (Fig. 1A). Western blot analysis demonstrated enhanced protein expression of DNMT3b but not DNMT1 in AGS-EBV cells as compared with AGS cells (Fig. 1B). Quantitative real-time RT-PCR was further performed to compare the messenger RNA (mRNA) expression of DNMT1 between AGS-EBV and AGS cells. However, no significant difference in DNMT1 levels between AGS-EBV and AGS cells was observed (1.2 ± 0.18 vs 1 ± 0.0, P = .25). We further examined DNMT1 and DNMT3b activities using EpiQuick DNMT1 and DNMT3b activity assays respectively. DNMT3b activity was significantly higher in AGS-EBV cells as compared with AGS cells, whereas no difference in DNMT1 activity was observed between AGS-EBV and AGS cells (Fig. 1C). Collectively, these results indicated that DNMT3b was the only DNMT activated by EBV infection in AGS cells.

LMP2A-Induced DNMT3b Activity in AGS Cells

Among the several EBV viral genes expressing in EBV-associated gastric cancer (EBNA1, EBER, BARF0, and LMP2A), only LMP2A has been reported to be associated with clinical outcomes of EBV-associated gastric cancer and mediate the up-regulation of DNMTs.7, 11 In order to determine whether LMP2A regulated the activation of DNMT3b, LMP2A expression vector was transfected into AGS cells. Ectopic expression of LMP2A in AGS was confirmed by RT-PCR (Fig. 1D). LMP2A increased the expression of DNMT3b in LMP2A-transfected AGS cells as compared with empty vector-transfected AGS cells, at the mRNA and protein levels (Fig. 1D,E). Moreover, LMP2A significantly induced the activity of DNMT3b in AGS cells, although DNMT1 activity was not changed (Fig. 1F). These results indicated that EBV infection activated DNMT3b in AGS cells at least in part through LMP2A expression.

Genome-Wide Profiling of DNA Methylation Driven by EBV Infection, Using MeDIP-Chip

To identify genes regulated by EBV-driven aberrant methylation, genome-wide DNA methylation profiles were compared between AGS-EBV and AGS cells by MeDIP-chip assay, which covered 17,000 best-refined genes. A total of 1065 candidates were detected with differential DNA methylation between AGS-EBV and AGS cells, with fold-changes ≥ 2 (normalized log2 ratios ≥ 1) and P ≤ 0.05. As expected, 83.2% (886 of 1065 genes) of the differentially methylated genes were CpG hypermethylated genes in AGS-EBV cells compared with AGS cells, with fold-changes ranging from 2.43 to 65.2 (normalized log2 ratios of 1.28∼6.03) (Fig. 2A,B). In concordance with activation of DNMT3b, these results revealed that genome-wide aberrant methylation driven by EBV infection was composed mainly of CpG hypermethylation in AGS-EBV cells.

thumbnail image

Figure 2. Differential methylation profiling is shown between AGS cells positive for Epstein-Barr virus (AGS-EBV) and non-EBV AGS cells (AGS). (A) Hierarchical clustering analysis of MeDIP-chip (methylated DNA immunoprecipitation coupled with hybridization on high-resolution microarray) data. Data was adjusted by the “center arrays” method embedded in Gene Cluster 3.0 by subtracting the column-wise mean from the values in each column. (B) Histograms of gene numbers are shown according to log2 fold-change ranges. Nhypo, number of hypomethylated genes in AGS-EBV; Nhyper, number of hypermethylated genes in AGS-EBV. (C) Enrichment of hypermethylated genes in KEGG (Kyoto Encyclopedia of Genes and Genomes) pathways is shown by pie chart of pathways with ≥ 3 genes involved. Each pie section stands for one pathway. The number of genes involved in each pathway was indicated by both the color of the pie section and the numeral beside the pie section (left panel). Pathways with ≥ 10 genes enriched were further indicated by histograms (right panel), with corresponding pathways labeled on the right. MAPK indicates mitogen-activated protein kinase.

Download figure to PowerPoint

EBV-Associated Cancer Pathways Defined by EBV-Driven Promoter Methylated Genes

KEGG pathway analysis was performed to identify EBV-associated cancer pathways with the 886 CpG hypermethylated genes driven by EBV infection. EBV-driven hypermethylated genes were found to be enriched in 13 cancer pathways (with ≥ 10 genes involved in each pathway and P < .05) including: neuroactive ligand-receptor interaction, pathways in cancer, mitogen-activated protein kinase (MAPK) signaling pathway, cytokine-cytokine receptor interaction, axon guidance, regulation of actin cytoskeleton, insulin signaling pathway, cell adhesion molecules, endocytosis, calcium signaling pathway, wnt signaling pathway, glutamatergic synapse, and focal adhesion (Table 2 and Fig. 2C).

Table 2. Enrichment of Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathways in Hypermethylated Genes Driven by Epstein-Barr Virus Infection
KEGG PathwaysNGaHyp P ValuebHypermethylated Genes in AGS-EBV
  • a

    NG, number of annotated genes in the input list (number of annotated genes in the reference list).

  • b

    Hyp P value, corrected hypergeometric P value.

  • MAPK indicates mitogen-activated protein kinase.

Neuroactive ligand-receptor interaction22 (269)6.18E-05GRM6,CALCR,NTSR1,CHRM2,GRIN2A,CRHR2,GABRQ, ADRA1D,DRD4,GRIA1,ADRB3,P2RY2,GPR83,MLNR, SSTR1,GABRG3,GABRE,GRIA4,NPBWR1,GIPR,SSTR4,P2RY11
Pathways in cancer22 (325).000687WNT11,AXIN1,FZD9,WNT3,GSTP1,ZBTB16,MMP2, FGF19,WNT2,RAC3,PDGFRA,RUNX1,FOXO1,FGF3,ARNT, FGF1,CDKN2A,AKT2,GLI3,RASSF5,RARA,SHH
MAPK signaling pathway19 (265).000831CACNA1B,DUSP5,NR4A1,MAP3K14,HSPB1,FGF19,PAK2, RAC3,PDGFRA,DUSP8,RASGRP2,FGF3,FGF1,NFATC2,AKT2, RASGRF1,CACNG4,CACNB2,PLA2G2E
Cytokine-cytokine receptor interaction13 (263).027313CNTFR,BMP7,TNFRSF18,FLT1,IL15RA,IL10RA,ACVR2A, IL13RA1,PDGFRA,TNFSF11,IL17B,CXCR4,CD40
Axon guidance12 (128).001165SLIT3,EFNA4,EPHB6,EPHA8,PAK2,RAC3,CXCR4,NFATC2, SEMA6D,EFNA1,EPHA2,ARHGEF12
Insulin signaling pathway11 (135).004505SORBS1,IRS2,PDE3B,HK1,IRS1,SLC2A4,GCK, FOXO1,FBP1,AKT2,PPARGC1A
Regulation of actin cytoskeleton11 (211).030246CHRM2,FGF19,PAK2,RAC3,PDGFRA,FGF3,DOCK1, FGF1,ACTB,ARHGEF12,ARPC1B
Cell adhesion molecules11 (134).004647NLGN4X,NCAM1,NRXN2,CDH15,NRCAM,ICAM1,CD6, NRXN1,CDH4,CLDN3,CD40
Glutamatergic synapse10 (124).00762GRM6,GLUL,GRIN2A,ADRBK1,GRIA1,GNG7, SLC1A2,GRIA4,GNAS,PLA2G2E
Focal adhesion10 (199).039355TNXB,FLT1,PAK2,RAC3,PDGFRA,DOCK1, AKT2,ACTB,RASGRF1,CCND2
Wnt signaling pathway10 (150).012994WNT11,SFRP4,AXIN1,FZD9,WNT3,SFRP5,WNT2, RAC3,NFATC2,CCND2
Endocytosis10 (200).039484FLT1,RAB31,AP2A2,ADRBK1,SMAD6,ADRB3,PDGFRA, CXCR4,GRK6,CHMP2A
Calcium signaling pathway10 (175).024864CACNA1B,PDE1C,ATP2B2,NTSR1,CHRM2,GRIN2A, ADRA1D,ADRB3,PDGFRA,GNAS

CpG Hypermethylation and Transcriptional Silencing of EBV-Driven Methylated Genes in AGS-EBV Cells

To confirm EBV-driven differentially methylated candidates, we performed targeted methylation analysis in the promoters of 6 novel candidates with higher methylation levels by MeDIP-chip analysis, including EPH receptor B6 (EPHB6) (29-fold), interleukin-15 receptor alpha (IL-15RA) (24.4-fold), REC8 (23.5-fold), MAM domain containing glycosylphosphatidylinositol anchor 2 (MDGA2) (20-fold), scavenger receptor class F, member 2 (SCARF2) (16.1-fold), and somatostatin receptor 1 (SSTR1) (8.4-fold). Bisulfite genomic sequencing of the CpG sites in the promoter region of each gene confirmed significant higher methylation levels in EBV positive AGS-EBV cells compared with the EBV negative AGS cells for EPHB6 (P < .0001), IL15RA (P = .002), REC8 (P < .0001), MDGA2 (P < .0001), SCARF2 (P < .01) and SSTR1 (P < .05) (Fig. 3A). To investigate if EBV-driven promoter methylation of these candidates might regulate their expression, mRNA expression of these genes was compared between AGS-EBV and AGS cells by semiquantitative RT-PCR. All 6 genes were found to be silenced or down-regulated in AGS-EBV cells as compared with AGS cells (Fig. 3B). In addition, we treated the AGS-EBV cells with 5-Aza, a chemical inhibitor of DNA methyltransferases. The transcripts of all these 6 genes were up-regulated after 5-Aza treatment (Fig. 3C). These results indicated that EBV-driven CpG hypermethylation is likely to regulate gene transcription silence, which can be restored following demethylation treatment.

thumbnail image

Figure 3. Validation of genes identified by methylated DNA immunoprecipitation coupled with hybridization on high-resolution microarray (MeDIP-chip). (A) Bisulfite genomic sequencing confirmed that novel candidates identified by MeDIP-chip were methylated at significantly higher levels in AGS cells positive for Epstein-Barr virus (AGS-EBV) than in EBV-negative AGS cells. (B) Transcriptional silence or down-regulation of the candidate genes was demonstrated in AGS-EBV cells by reverse transcription polymerase chain reaction (RT-PCR). (C) Expression of these genes was successfully restored in AGS-EBV cells by demethylation treatment using 5-aza-2′-deoxycytidine (5-Aza). (D) Promoter methylation of 2 newly identified EBV-driven methylated genes (SSTR1 and REC8) in primary EBV-positive and EBV-negative primary gastric cancers is shown by combined bisulfite restriction analysis (COBRA).

Download figure to PowerPoint

Validation of Novel EBV-Driven Methylated Genes in Primary Gastric Cancers

The methylation status of 2 identified novel EBV-driven methylated genes (SSTR1 and REC8) was examined in 6 EBV-positive gastric cancer and 5 EBV-negative gastric cancer samples by COBRA. Results show that SSTR1 was methylated in 4 of 6 (66.7%) EBV-positive gastric cancers, but none in 5 EBV-negative gastric cancers. Similarly, REC8 was methylated in 4 of 6 (66.7%) EBV-positive gastric cancers, but only in 1 of 5 (20%) EBV-negative gastric cancers.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

To gain insight into the effect of EBV infection on the aberrant DNA methylation in gastric cancer, we sought to understand the DNA methylation profiling of EBV-positive AGS-EBV cells in which EBV infection was confirmed by EBER in situ hybridization assay. Before performing the DNA methylation profile analysis in AGS-EBV and in AGS cells, we evaluated if EBV plays a role in DNMTs expression and activation. Our results showed an enhanced DNMT3b mRNA and protein expression in AGS-EBV cells as compared with AGS cells. This was further confirmed by a significantly induced DNMT3b activity in AGS-EBV cells as compared with AGS cells. Similar to our finding, it has been reported that recombinant EBV infection in gastric cancer cells MKN-1 and MKN-7 increased the expression of DNMT1.7 Although apparent enhanced DNMT1 mRNA level was shown by semiquantitative RT-PCR in AGS-EBV cells compared with AGS cells, further quantitative real-time RT-PCR analysis revealed that this difference is not significant. Moreover, no increased protein expression or enzymatic activity of DNMT1 was demonstrated in AGS-EBV as compared with AGS cells (Fig. 1B,C), indicating DNMT1 was not activated in EBV-AGS. These data suggest that aberrant DNA methylation at the genomic level may be induced by EBV infection through activating DNMTs in gastric cancer. In addition, the effect of the EBV gene LMP2A on DNMT3b expression and activity was examined, because LMP2A has been reported to be the only EBV gene that can mediate the up-regulation of DNMTs.7, 11 Our results clearly show that ectopic expression of LMP2A in LMP2A-transfected AGS cells caused an enhanced protein expression and induced activity of DNMT3b compared with vector-transfected AGS cells. These results suggested that EBV infection plays a role in gastric carcinogenesis by inducing DNA methylation, which is mediated by the activation of DNMT3b at least in part through LMP2A expression. We were thus interested to compare the methylation profiles of AGS-EBV and AGS cells.

Using a genome-wide methylation array (MeDIP-chip), we quantitatively assessed the DNA methylation status of 17,000 refined genes between AGS-EBV and AGS cells. We found that the DNA methylation profile of the EBV-positive AGS-EBV was distinctly different from those of the EBV-negative AGS cells. In concordance with the activation of DNMT3b, the majority of the differentially methylated genes (83.2%, or 886 of 1065 genes) were characterized by CpG hypermethylation in AGS-EBV cells as compared with that of AGS cells (fold-changes ranging from 2.43-65.2) (Fig. 2A,B). These identified CpG hypermethylated genes were found to be involved in important cancer-related pathways including MAPK signaling, cell adhesion molecules, the wnt signaling pathway, and the like (Table 2; Fig. 2C), where dysregulation of these genes through CpG hypermethylation could significantly contribute to gastric carcinogenesis. As reported, EBV-driven methylated genes can be involved in different aspects of tumorigenesis, including both tumor suppressors and genes with oncogenic function.16 Among the hypermethylated genes identified in this study, numerous genes are potential tumor suppressors (eg, RASSF5, FOXO1, TNXB),17-19 and some genes have oncogenic functions (eg, RAC3, ARHGEF12).20, 21 These complicated effects of EBV infection on the genome-wide methylation of the host genes might be associated with the generally special characteristics of EBV-associated gastric cancer.3, 22

Furthermore, we performed validation experiments to verify the methylation data from the arrays. Six novel hypermethylated genes, including IL15RA, REC8, SSTR1, EPHB6, MDGA2, and SCARF2, were confirmed by BGS, which showed significantly higher methylation levels of these genes in AGS-EBV than in EBV-negative AGS cells (Fig. 3A). In concordance with DNA hypermethylation, transcription of these genes was silenced or down-regulated in AGS-EBV cells, but readily expressed in AGS cells (Fig. 3B), confirming the effect of EBV infection on cellular gene transcriptional repression through mediating promoter methylation. Moreover, the EBV-driven promoter methylation genes were confirmed in primary gastric cancers. These results encouraged further investigation whether epigenetic changes are reversible by use of a demethylating agent. In this study, we demonstrated that demethylation treatment of AGS-EBV using 5-Aza could restore the expression of all 6 of these methylated genes in AGS-EBV cells (Fig. 3C). These results suggest that DNA methylation is a mechanism by which EBV alters cellular gene expression and contributes to gastric carcinogenesis. Because EBV infection is associated with CpG hypermethylation in vitro, demethylating agents could provide a potential therapeutic approach to overcoming viral-associated oncogenic effects in tumor cells. Demethylating agents have been used clinically for treatment of certain cancers.23 Our study provided potential implication of their efficacy in treatment of EBV-related gastric cancer.

In conclusion, this study provides for the first time a comparative DNA methylation profiling between EBV-positive AGS-EBV and EBV-negative AGS cells. EBV infection induced significantly specific hypermethylation patterns by up-regulation of DNMT3b through LMP2A, with 886 hypermethylated genes involved in important cancer-related pathways. This excessive methylation in EBV-related gastric cancer suggests a unique mechanism of gastric carcinogenesis.

FUNDING SOURCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES

This project was supported by research funds of RFCID (11100022, and 10090942), China 973 Program (2010CB529305), and CUHK Focused Investments (2041423).

CONFLICT OF INTEREST DISCLOSURE

The authors made no disclosure.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SOURCES
  8. REFERENCES