Table 2.

Summary of hypermethylated genes during BE/EAC development

Number (%) of samples showing hypermethylation or study findings
GeneLocationFunctionMethodNormalBELGDHGDEACRef.
p16 (or CDKN2A or INK4A)9p21Cyclin-dependent kinase inhibitorMethylation-specific PCR5/9 (56%)14/18 (77%)18/21 (85%)(146)
Methylation-sensitive single-strand conformation analysis0/10 (0%)4/12 (33%)3/11 (27%)3/10 (30%)18/22 (82%)(147)
Methylation-specific PCR0/17 (0%)14/47 (30%)9/27 (32%)10/18 (56%)22/41 (54%)(148)
Methylation-specific PCR2/64 (3%)14/93 (15%)34/76 (45%)(149)
Methylation-specific PCR3/10 (30%)5/11 (45%)(150)
Methylation-specific PCR27/41(66%)21/45 (47%)17/21 (81%)65/107 (61%)(151)
Methylation-specific PCR0%1/15 (7%)4/20 (20%)12/20 (60%)8/15 (53%)(152)
Methylation-specific PCRSeparately determined exon 1 and 2 methylation. Five of 16 (31%) exon-1, 8/16 (50%) exon 2 in EAC patient samples showed hypermethylation. Exon 2 methylation correlates with stage of the tumor (P = 0.01)(153)
O6-Methylguanine-10q26DNA repairMethylight technique2/10 (20%)8/13 (62%)84/132 (64%)(154)
DNA Methyltransferase (or MGMT)Methylation-specific PCR6/29 (21%)24/27 (89%)13/13 (100%)37/47 (79%)(155)
APC5q21-q22Wnt/β-catenin signalingMethylation-specific PCR0/17 (0%)24/48 (50%)14/28 (50%)14/18 (78%)20/32 (63%)(148)
Methylation-sensitive single-strand conformation analysis and methylation-sensitive dot blot assay0/16 (0%)11/11 (100%)20/21 (95%)(156)
Eight of 14 histologically normal gastric mucosa adjacent to EAC showed significantly different methylation of APC promoter.(157)
GSTM21p13.3Antioxidants and protection against DNA damageBisulfite pyrosequencing (sample size: EAC-100, BE-11, dysplasia-11, normal esophageal/gastric mucosa-37)<10%∼50%∼55%69%(158)
GSTM3<10%∼13%∼37%15%(158)
GPX71p32<10%∼18%∼80%67%(158)
GPX35q23<10%∼90%∼88%62%(158)
Methylation-specific PCR2/12 (17%)13/21(62%)9/11 (82%)30/34 (88%)(159)
TIMP-322q12.3MMP inhibitorMethylight technique1/8 (13%)6/12 (50%)9/13 (69%)(160)
Death-associated protein kinase (DAPK)DAPK1: 9q21.33Tumor-suppressor and mediator of apoptosisMethylation-specific PCR4/20 (20%)14/28 (50%)11/21 (53%)21/35 (60%)(161)
DAPK2: 15q22.31
DAPK3: 19p13.3
Tachykinin-1 (TAC1)7q21-22Smooth muscle contractility, epithelial ion transport, vascular permeability and immune functionMethylation-specific PCR5/67 (7.5%)38/60 (63.3%)12/19 (63.2%)11/21 (52.4%)41/67 (61.2%)(162)
Reprimo2q23Regulates p53-mediated cell-cycle arrest in G2-phaseMethylation-specific PCR0/19 (0%)9/25 (36%)7/11 (64%)47/75 (63%)(163)
E-Cadherin16q22.1Ca+2-dependent intercellular adhesion and maintains normal tissue architectureMethylation-specific PCR0/4 (0%)26/31 (84%)(164)
SOCS-317q25.3Inhibits cytokine signalingMethylation-specific PCR0%4/30 (13%)6/27 (22%)20/29 (69%)14/19 (74%)(165)
SOCS-116p13.130%0/30 (0%)1/27 (4%)6/29 (21%)8/19 (42%)
Secreted frizzled-related proteins (SFRP)
SFRP18p11.21Wnt antagonistMethylation-specific PCR7/28 (25%)30/37 (81%)37/40 (93%)(166)
SFRP24q31.318/28 (64%)33/37 (89%)33/40 (83%)
SFRP18p11.21Methylation-sensitive single-strand conformation analysis and methylation-sensitive dot blot assay1/12 (8%)6/6 (100%)23/24 (96%)(156)
SFRP24q31.311/15 (73%)6/6 (100%)19/25 (76%)
SFRP47p14.1Methylation-specific PCR9/28 (32%)29/37 (78%)29/40 (73%)(166)
SFRP510q24.16/28 (21%)27/37 (73%)34/40 (85%)
Plakophilin-1 (PKP1)1q32Cell adhesion and intracellular signalingMethylation-specific PCR5/55 (9.1%)5/39 (12.8%)1/4 (25%)20/60 (33.3)(167)
GATA-48p23.1-p22Transcription factor and regulate cell differentiationMethylation-specific PCR0/17 (0%)31/44 (71%)(168)
GATA-520q13.330/17 (0%)24/44 (55%)
CDH13 (or H-cadherin or T-cadherin)16q24Cell adhesionMethylation-specific PCR0/66 (0%)42/60 (70%)15/19 (78.9%)16/21 (76.2)51/67 (76.1%)(169)
NELL-1 (nel-like 1)11p15Tumor suppressorMethylation-specific PCR0/66 (0%)28/60 (46.7%)8/19 (42.1%)13/21 (61.9%)32/67 (47.8%)(170)
Eyes Absent 46q23Apoptosis modulatorMethylation-specific PCR2/58 (3%)27/35 (77%)33/40 (83%)(171)
A-kinase anchoring protein 12 (or Gravin or AKAP12)6q24-25.2Cell-signaling, adhesion, mitogenesis, and differentiationMethylation-specific PCR0/66 (0%)29/60 (48.3%)10/19 (52.6%)11/21 (52.4%)35/67 (52.2)(172)
Vimentin10p13Cytoskeleton proteinMethylation-specific PCR0/9 (0%)10/11 (91%)5/5 (100%)21/26 (81%)(173)
RUNX31p36Transcription factorMethylation-specific PCR1/63 (2%)23/93 (25%)37/77 (48%)(149)
HPP119pter-p13.1Tumor suppressor2/64 (3%)41/93 (44%)55/77 (71%)
3-OST-216p12Sulfotransferase enzyme1/57 (2%)47/60 (78%)28/73 (38%)
Wnt inhibitory factor-1 (WIF-1)12q14.3Wnt antagonistMethylation-specific PCR81% of patients with Barrett's esophagus suffering from EAC showed hypermethylated WIF-1 as compared with 20% of patients with Barrett's esophagus without EAC(174)
CHFR (checkpoint with forkhead associated and ring finger)12q24Mitosis check point proteinBisulfite pyrosequencingEAC samples 31% (18/58) showed significantly higher CHFR promoter methylation as compared with normal samples (P = 0.01).(175)
Metallothionein 3 (or MT3)16q13Metal homeostasis and protection against DNA damageBisulfite pyrosequencing (sample size: normal-33, BE-5, EAC-78)Identified 2 regions (R2 and R3) of CpG nucleotides, which showed significantly higher methylation in EAC as compared with normal epithelium (FDR < 0.001). Increased DNA methylation of MT3 promoter R2 correlates with advanced tumor stage (P = 0.005) and lymph node metastasis (P = 0.03). DNA methylation of MT3 promoter R3 correlates with tumor staging (P = 0.03) but not with lymph node status (P = 0.4).(176)
Methylation marker panel
Sample sizeMethodFindingsRef.
EAC-35 undergoing chemoradiotherapyMethylation-specific PCRCombined mean of promoter methylation of p16, Reprimo, p57, p73, RUNX-3, CHFR, MGMT, TIMP-3, and HPP1 was lower in patients who responded to chemoradiotherapy (13/35) as compared with patients who did not respond (22/35; P = 0.003).(46)
BE-62 (28 patients with Barrett's esophagus progressed to EAC and remaining 34 patients with Barrett's esophagus were nonprogressors)Methylation-specific PCRThree-tiered stratification model was developed using methylation index (p16, HPP1, and RUNX3), Barrett's esophagus length and pathology. Combined model based on 2- (ROC: 0.8386) and 4-year (ROC: 0.7910) prediction was able to categorize patients with Barrett's esophagus into low-risk, intermediate-risk, and high-risk groups for EAC development.(44)
BE-195 (145 patients with Barrett's esophagus progressed to EAC and remaining 50 patients with Barrett's esophagus were nonprogressors)Methylation-specific PCRHPP1 (P = 0.0025), p16 (P = 0.0066), and RUNX3 (P = 0.0002) were significantly hypermethylated in progressors as compared with nonprogressors. In combination, panel of 8 methylation markers (p16, HPP1, RUNX3, CDH13, TAC1, NELL1, AKAP12, and SST) showed sensitivities of 0.443 and 0.629 at specificity of 0.9 and 0.8 for EAC progression in patients with Barrett's esophagus using combined model designed on the basis of 2 and 4 years of follow-up.(43)
EAC-41 (adjacent normal samples as control)Methylation-specific PCRPatients having more than 50% of their genes methylated (APC, E-cadherin, MGMT, ER, p16, DAP-kinase, and TIMP3) showed significantly poor 2-year survival (P = 0.04) and 2-year relapse-free survival (P = 0.03) as compared with the patients having less than 50% methylation.(45)
BE-18, EAC-38 (multiple biopsies were taken and classified into normal, Barrett's esophagus, HGD, and EAC)Bisulfite-modified DNA with PCRThe methylation frequencies of 9 genes (APC, CDKN2A, ID4, MGMT, RBP1, RUNX3, SFRP1, TIMP3, and TMEFF2) found to be 95%, 59%, 76%, 57%, 70%, 73%, 95%, 74%, and 83%, respectively, in EAC samples, whereas 95%, 28%, 78%, 48%, 58%, 48%, 93%, 88%, and 75%, respectively, in Barrett's esophagus samples, which was significantly higher as compared with normal squamous epithelium. The methylation frequency for CDKN2A and RUNX3 was significantly higher for EAC as compared with Barrett's esophagus biopsy samples.(177)
Normal-30, BE-29, HGD-8, EAC-29Illumina GoldenGate methylation bead arrayOverall median methylation at the total 706 numbers of most informative CpG sites gradually increased from normal-BE-HGD/EAC (P < 0.001). The authors differentiated between EAC vs. normal, HGD vs. normal, Barrett's esophagus vs. normal, EAC vs. Barrett's esophagus, and HGD vs. Barrett's esophagus based on 422, 225, 195, 17, and 3 numbers of CpG sites, which is showing differential methylation between respective groups.(178)
Identification phase (BE-22, EAC-24); retrospective validation phase (BE-60, LGD/HGD-36, EAC-90); prospective validation phase (98 patients under surveillance).Identification phase: Illumina Infinium assay; retrospective/prospective validation phase: pyrosequencingOn the basis of initial identification phase, 7 genes (SLC22A18, ATP2B4, PIGR, GJA12, RIN2, RGN, and TCEAL7) showing most prominent methylation changes were selected for validation. Combination of 4 genes (ROC 0.988) SLC22A18, PIGR, GJA12, and RIN2 showed sensitivity of 94% and specificity of 97%. This panel of 4 genes showing differential methylation, stratified patients into low-, intermediate-, and high-risk groups for EAC development in prospective validation.(179)
Nondysplastic Barrett's esophagus (not progressed to EAC)-16, Barrett's esophagus mucosa from patients progressed to EAC-12Methylation-sensitive single-strand conformation analysis and methylation-sensitive dot blot assayBarrett's esophagus samples collected from patients who progressed to EAC in 12 months time period showed 100%, 91%, and 92% hypermethylation of APC, TIMP-3, and TERT, respectively, as compared with 36%, 23%, and 17% in Barrett's esophagus mucosa collected from patients who did not progress to EAC.(180)