To explore the role of aberrant hypermethylation of cancer-related genes, such as P16, MGMT, and hMLH1, in the esophageal squamous cell carcinoma (ESCC) as well as its relation to dietary folate intake and MTHFR C677T polymorphism, we conducted a molecular epidemiologic study in China. One hundred and twenty-five histologically confirmed ESCC patients having undergone surgery in the Yangzhong People's Hospital between January 2005 and March 2006 were recruited. The aberrant CpG island hypermethylation of P16, MGMT, and hMLH1 genes could be found in cancer tissues with frequency of about 88.0%, 27.2%, and 3.2%, respectively, and in remote normal-appearing esophageal tissues with frequency of about 36.8%, 11.2%, and 0.0%, respectively. No hypermethylation was found in the normal esophageal tissues from healthy controls. Compared with those patients without lymph node metastasis, MGMT gene showed a higher proportion of hypermethylation in cancer tissues, whereas P16 gene showed a higher proportion of hypermethylation in remote normal-appearing esophageal tissues in patients with lymph node metastasis. A significant association was found between MTHFR C677T genetic polymorphism and CpG island methylation status of MGMT gene. After adjustment for potential confounders, individuals carrying CT or TT genotype have higher frequency of hypermethylation in MGMT gene in cancer tissues, with odds ratio of 3.34 (95% confidence interval, 1.07-10.39) and 3.83 (95% confidence interval, 1.13-12.94), respectively. This study indicated that the aberrant CpG island hypermethylation of cancer-related genes was associated with ESCC and might be a promising biomarker in diagnosis and prognosis. (Cancer Epidemiol Biomarkers Prev 2008;17(1):118–25)

As the eighth commonest cancer, esophageal cancer usually presents as an aggressive and locally advanced disease with a poor prognosis (1). According to the global cancer statistics from the IARC, it was responsible for ∼462,000 new cases each year with age standardized incidence rate at 11.5 new cases per 100,000 person-years in men and 4.4 new cases per 100,000 person-years in women (2). There are two main common histologic types of esophageal cancer, each with distinct etiologic and pathologic characteristics: esophageal squamous cell carcinoma (ESCC) and adenocarcinoma (1). Although recently there seems to be an increase in adenocarcinomas of esophagus in Western countries, worldwidely, most esophageal cancers are squamous cell carcinomas (1-3).

Epidemiologic studies have shown that deficiency of specific micronutrient, such as folate, was a possible risk factor of esophageal cancer (3-7). Folate, a water-soluble B vitamin, mainly found in fresh vegetables and fruits, has been proved to play an important role in the pathogenesis of several human disorders, including neural tube defects, anemia, cardiovascular disease, adverse pregnancy outcome, and cancer (8). The negative association between dietary folate intake and the risk of esophageal cancer has been observed in several studies. A meta-analysis reported that the combined relative risk for high versus low dietary folate intake was 0.66 [95% confidence interval (95% CI), 0.53-0.83] for ESCC (based on four studies including 929 cases and 2,479 controls) and 0.50 (95% CI, 0.39-0.65) for esophageal adenocarcinoma (based on three studies including 501 cases and 1,268 controls; ref. 9). There are two possible major mechanisms that can explain the relation between folate deficiency and the risk of cancers (9). First, as an essential cofactor for the biosynthesis of purines and thymidylate, folate is important in DNA synthesis, integrity, and stability (9, 10). Deficiency of folate could lead to chromosomal breaks and mutations. Second, folate is the primary methyl group donor for DNA methylation, which is an important epigenetic determinant for gene silence (9-11). As one of the most common epigenetic events in the carcinogenic process, aberrant DNA methylation in the occurrence of cancers includes global hypomethylation in genome, which increases mutation rates and chromosomal instability, and hypermethylation in specific genes, which leads to transcriptional inactivation (12, 13). As the methyl group functioned in DNA methylation process mainly comes from folate metabolic pathway, it is believed that folate metabolism is involved in the DNA methylation process, thus influencing the risk of cancers by inappropriate activation of proto-oncogenes and inactivation of anti-oncogenes with the results of malignant transformation (11).

MTHFR (5,10-methylenetetrahydrofolate reductase), a central folate metabolic enzyme, is responsible for the circulation form of folate and 5-methyltetrahydrofolate, which is essential for DNA synthesis and methylation (14). The activity of MTHFR is controlled mainly by the genetic polymorphisms and can vary significantly between different individuals (9, 15). The commonest polymorphism for MTHFR gene is a single C→T substitution at nucleotide 677, which leads to the amino acid alanine being replaced by valine, resulting in a reduced activity of this enzyme (15). Theoretically, the varied activity of folate metabolic enzyme caused by genetic polymorphisms would influence the methylation status and individual susceptibility to cancers. However, currently, the available data about the role of MTHFR C677T polymorphism in relation to the aberrant DNA methylation are inconsistent and incomplete.

P16, MGMT, and hMLH1 are important tumor suppressor and DNA repair genes, which play significant roles in the carcinogenic process. Inactivation of these genes is believed to be involved in the progression of many human cancers. As one of the principal causes of gene inactivation, aberrant hypermethylation in the promoter of cancer-related genes is appealing more and more focuses (16, 17). However, corresponding information on ESCC is still limited (16). To explore the role of aberrant hypermethylation of cancer-related genes, such as P16, MGMT, and hMLH1, in the ESCC as well as its relation to dietary folate intake and MTHFR C677T polymorphism, we conducted a molecular epidemiologic study in a population with high incidence of ESCC in China.

Study Site

Yangzhong County is an island located on the middle of Yangtze River in the southeast part of Jiangsu Province, China, with a population of ∼0.28 million and an area of ∼332 km2. In the early 1980s, it had been found that the mortality rate of cancer, especially upper digestive tract cancer, was remarkably high in Yangzhong County (18). Since 1991, a population-based cancer registry has been established. Based on the newest report, the world standard population adjusted incidence rate of esophageal cancer was 93.8/100,000 for men and 73.7/100,000 for women in 2003 (19).

Study Subjects

One hundred and twenty-five ESCC patients having undergone esophagectomy between January 2005 and March 2006 in the Yangzhong People's Hospital were involved in the analysis on aberrant DNA hypermethylation, which accounted for about >80% of all eligible cases on esophageal cancer surgery in this hospital during the same period. Ten pathologically confirmed normal esophageal tissues as controls were obtained from those people who underwent gastroendoscopy examination in the outpatient department of Yangzhong People's Hospital.

Data Collection

Face-to-face interviews were conducted by trained investigators using a structured questionnaire, including information on socio-demographic factors, tobacco smoking, alcohol drinking, and dietary habits. A food frequency questionnaire was used to estimate the dietary intake during the last several years before their enrollment day into the study. Peripheral blood samples were collected with vacuum blood tubes at the same time. For those patients having undergone surgery, tissues in the center of the cancer lesion and remote normal-appearing esophagus were excised and stored in −70°C refrigerator immediately. Ten normal esophageal tissue samples were obtained under gastroendoscopy and stored until use.

DNA Methylation Detection and Genotyping

The methylation at the promoter region of P16, MGMT, and hMLH1 was determined by methylation-specific PCR after sodium bisulfite modification of DNA (20). In detail, 1.5 to 2.0 μg of genomic DNA dissolved in 50 μL H2O were incubated with 5.5 μL of 3 mol/L NaOH at 37°C for 10 min and then treated with freshly prepared 30 μL of 10 mol/L hydroquinone and 520 μL of 3 mol/L NaHSO3 (pH 5.0). Samples were incubated under mineral oil at 50°C for 16 h. Modified DNA was purified by Wizard DNA Clean-Up System (Promega) and eluted with 50 μL preheated Tris-EDTA solution. Modification was completed by the treatment of 5.5 μL of 3 mol/L NaOH (final concentration, 0.3 mol/L) for 15 to 20 min at room temperature. DNA was precipitated by 75% ethanol and resuspended in 20 μL Tris-EDTA. After treatment, the unmethylated cytosine would be converted to uracil and was recognized as thymine by Taq polymerase during PCR process. The methylated cytosine was unchangeable after chemical modification. The methylation status at the promoter region of P16, MGMT, and hMLH1 genes was assessed by methylation-specific PCR using methylation-specific primers and relevant annealing temperature (Table 1). A methylation-positive DNA control sample was made in vitro by SssI methylase (New England Biolabs). PCR products were loaded onto 3% gels, stained with ethidium bromide, and directly visualized under UV illumination (Fig. 1). MTHFR C677T genetic polymorphism was detected by PCR-RFLP with the forward primer TGAAGGAGAAGGTGTCTGCGGGA and the backward primer AGGACGGTGCGGTGAGAGTG. PCR products were digested overnight at 37°C with restriction endonuclease HinfI, which could recognize and cut the variant type sequence. The digestive products were resolved on 3% agarose gel and visualized under UV light after staining with ethidium bromide. The variant genotype TT has two bands (175 and 23 bp). The heterozygote CT has three bands (198, 175, and 23 bp), whereas the wild genotype CC has only one 198 bp band.

Table 1.

Primers and annealing temperature used for PCR

GenesM/US/ASPrimer sequences (5′→3′)Annealing temperature (°C)Product size (bp)
P16 TTATTAGAGGGTGGGGCGGATCGC 68 150 
  AS GACCCCGAACCGCGACCGTAA   
 TTATTAGAGGGTGGGGTGGATTGT 60 151 
  AS CAACCCCAAACCACAACCATAA   
MGMT TTTCGACGTTCGTAGGTTTTCGC 66 81 
  AS GCACTCTTCCGAAAACGAAACG   
 TTTGTGTTTTGATGTTTGTAGGTTTTTGT 60 93 
  AS AACTCCACACTCTTCCAAAAACAAAACA   
hMLH1 ACGTAGACGTTTTATTAGGGTCGC 60 124 
  AS CCTCATCGTAACTACCCGCG   
 TTTTGATGTAGATGTTTTATTAGGGTTGT 60 112 
  AS ACCACCTCATCATAACTACCCACA   
MTHFR  TGAAGGAGAAGGTGTCTGCGGGA 61 198 
  AS AGGACGGTGCGGTGAGAGTG   
GenesM/US/ASPrimer sequences (5′→3′)Annealing temperature (°C)Product size (bp)
P16 TTATTAGAGGGTGGGGCGGATCGC 68 150 
  AS GACCCCGAACCGCGACCGTAA   
 TTATTAGAGGGTGGGGTGGATTGT 60 151 
  AS CAACCCCAAACCACAACCATAA   
MGMT TTTCGACGTTCGTAGGTTTTCGC 66 81 
  AS GCACTCTTCCGAAAACGAAACG   
 TTTGTGTTTTGATGTTTGTAGGTTTTTGT 60 93 
  AS AACTCCACACTCTTCCAAAAACAAAACA   
hMLH1 ACGTAGACGTTTTATTAGGGTCGC 60 124 
  AS CCTCATCGTAACTACCCGCG   
 TTTTGATGTAGATGTTTTATTAGGGTTGT 60 112 
  AS ACCACCTCATCATAACTACCCACA   
MTHFR  TGAAGGAGAAGGTGTCTGCGGGA 61 198 
  AS AGGACGGTGCGGTGAGAGTG   

Abbreviations: M, methylated; U, unmethylated; S, sense; AS, antisense.

Figure 1.

Representative results of methylation-specific PCR analysis for three genes. Methylation-specific PCR analysis of each gene promoter by using both methylated (M) and unmethylated (U) specific primers. A, methylated sample; B, unmethylated sample; C, positive control; D, negative control.

Figure 1.

Representative results of methylation-specific PCR analysis for three genes. Methylation-specific PCR analysis of each gene promoter by using both methylated (M) and unmethylated (U) specific primers. A, methylated sample; B, unmethylated sample; C, positive control; D, negative control.

Close modal

Data Analysis

Data were entered into a database with EpiData software and analyzed by Statistical Package for the Social Sciences 11.0 software (SPSS). The dietary folate intake for each food item was calculated by multiplying the weight (grams) of food intake and the folate content (per gram) referring to a national food nutrition database (21, 22), and then the sum of all folate taken from various foods was calculated as the total folate intake. In this study, we used the folate from fresh fruits as a representative value for the total dietary folate intake. The continuous variables of folate intake were transferred to four categories as Q1, Q2, Q3, and Q4 using quartiles. To estimate the association between MTHFR genetic polymorphism, DNA hypermethylation, and the risk of ESCC, odds ratios (OR) and their 95% CI were calculated through unconditional logistic regression model. Dummy variables were used to estimate ORs for categorical variables of exposure.

Ethics Consideration

This project has been approved by the Institutional Review Board of the School of Public Health, Fudan University, China (IRB#05-06-0031). Ethics has been respected throughout the whole study period.

DNA Hypermethylation in P16, MGMT, and hMLH1 Genes

Totally, 125 ESCC patients having undergone esophagectomy between January 2005 and March 2006 in the Yangzhong People's Hospital were involved in the analysis. The average age was 61.8 ± 7.0 years and it included 81 (64.8%) men and 44 (35.2%) women. The proportions of hypermethylation for P16, MGMT, and hMLH1 genes in esophageal cancer tissues were 88.0% (110 of 125), 27.2% (34 of 125), and 3.2% (4 of 125), respectively. One hundred and thirteen (90.4%) of 125 cases had at least one CpG island hypermethylation in these three genes. In the remote normal-appearing esophageal tissues, 36.8% (46 of 125) of cases showed hypermethylation in P16 gene and 11.2% (14 of 125) showed hypermethylation in MGMT gene. The proportion of hypermethylation in at least one gene was 43.2% (54 of 125). No hypermethylation was found in the promoter of hMLH1 gene in remote normal-appearing esophageal tissues. In the 10 histologically confirmed normal esophageal tissues, all these three genes were found unmethylated.

The Association between DNA Methylation and Sex, Age, and Selected Exogenous Factors

Either in the cancer tissues or in the remote normal-appearing esophageal tissues, no significant association was found between the methylation status of P16, MGMT, and hMLH1 genes and the selected factors, including sex, age, tobacco smoking, and alcohol drinking. However, fruit folate intake was found to be significantly associated with the methylation of MGMT gene. In the individuals at highest quartile of fruit folate intake, 55.6% of the cancer tissues were methylated, whereas only 21.6% were methylated in the individuals with lowest fruit folate intake (Table 2).

Table 2.

Association between DNA methylation and sex, age, and selected exogenous factors in the ESCC

VariablesnFrequency of hypermethylation (%)
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Sex          
    Male 81 71 (87.7) 21 (25.9) 3 (3.7) 74 (91.4) 28 (34.6) 8 (9.9) 31 (38.3) 
    Female 44 39 (88.6) 13 (29.5) 1 (2.3) 39 (88.6) 18 (40.9) 6 (13.6) 23 (52.3) 
P  0.872 0.664 1.000* 0.752* 0.483 0.561* — 0.131 
Age (y)          
    <60 49 43 (87.8) 12 (24.5) 2 (4.1) 45 (91.8) 20 (40.8) 8 (16.3) 25 (51.0) 
    60+ 76 67 (88.2) 22 (28.9) 2 (2.6) 68 (89.5) 26 (34.2) 6 (7.9) 29 (38.2) 
P  0.946 0.585 0.645* 0.763* 0.455 0.144 — 0.156 
Tobacco smoking          
    Never 68 59 (86.8) 20 (29.4) 2 (2.9) 61 (89.7) 26 (38.2) 8 (11.8) 32 (47.1) 
    Ever 57 51 (89.5) 14 (24.6) 2 (3.5) 52 (91.2) 20 (35.1) 6 (10.5) 22 (38.6) 
P  0.643 0.544 1.000* 0.774 0.716 0.827 — 0.341 
Alcohol drinking          
    Never 71 64 (90.1) 21 (29.6) 2 (2.8) 65 (91.5) 31 (43.7) 8 (11.3) 36 (50.7) 
    Ever 54 46 (85.2) 13 (24.1) 2 (3.7) 48 (88.9) 15 (27.8) 6 (11.1) 18 (33.3) 
P  0.398 0.493 1.000* 0.617 0.068 0.978 — 0.052 
Fruit folate intake          
    Q1 37 32 (86.5) 8 (21.6) 33 (89.2) 12 (32.4) 3 (8.1) 14 (37.8) 
    Q2 45 41 (91.1) 17 (37.8) 43 (95.6) 16 (35.6) 8 (17.8) 20 (44.4) 
    Q3 29 24 (82.8) 4 (13.8) 1 (3.5) 24 (82.8) 11 (37.9) 1 (3.5) 12 (41.4) 
    Q4 9 (100) 5 (55.6) 2 (22.2) 9 (100) 3 (33.3) 2 (22.2) 4 (44.4) 
P  0.535* 0.027* 0.004* 0.230* 0.974* 0.139* — 0.933* 
VariablesnFrequency of hypermethylation (%)
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Sex          
    Male 81 71 (87.7) 21 (25.9) 3 (3.7) 74 (91.4) 28 (34.6) 8 (9.9) 31 (38.3) 
    Female 44 39 (88.6) 13 (29.5) 1 (2.3) 39 (88.6) 18 (40.9) 6 (13.6) 23 (52.3) 
P  0.872 0.664 1.000* 0.752* 0.483 0.561* — 0.131 
Age (y)          
    <60 49 43 (87.8) 12 (24.5) 2 (4.1) 45 (91.8) 20 (40.8) 8 (16.3) 25 (51.0) 
    60+ 76 67 (88.2) 22 (28.9) 2 (2.6) 68 (89.5) 26 (34.2) 6 (7.9) 29 (38.2) 
P  0.946 0.585 0.645* 0.763* 0.455 0.144 — 0.156 
Tobacco smoking          
    Never 68 59 (86.8) 20 (29.4) 2 (2.9) 61 (89.7) 26 (38.2) 8 (11.8) 32 (47.1) 
    Ever 57 51 (89.5) 14 (24.6) 2 (3.5) 52 (91.2) 20 (35.1) 6 (10.5) 22 (38.6) 
P  0.643 0.544 1.000* 0.774 0.716 0.827 — 0.341 
Alcohol drinking          
    Never 71 64 (90.1) 21 (29.6) 2 (2.8) 65 (91.5) 31 (43.7) 8 (11.3) 36 (50.7) 
    Ever 54 46 (85.2) 13 (24.1) 2 (3.7) 48 (88.9) 15 (27.8) 6 (11.1) 18 (33.3) 
P  0.398 0.493 1.000* 0.617 0.068 0.978 — 0.052 
Fruit folate intake          
    Q1 37 32 (86.5) 8 (21.6) 33 (89.2) 12 (32.4) 3 (8.1) 14 (37.8) 
    Q2 45 41 (91.1) 17 (37.8) 43 (95.6) 16 (35.6) 8 (17.8) 20 (44.4) 
    Q3 29 24 (82.8) 4 (13.8) 1 (3.5) 24 (82.8) 11 (37.9) 1 (3.5) 12 (41.4) 
    Q4 9 (100) 5 (55.6) 2 (22.2) 9 (100) 3 (33.3) 2 (22.2) 4 (44.4) 
P  0.535* 0.027* 0.004* 0.230* 0.974* 0.139* — 0.933* 
*

Fisher's exact test.

The cut point of fruit folate intake is quartile (μg/d): Q1, 0; Q2, 26.3; Q3, 99.5; and Q4, 311.1.

The Association between DNA Methylation and Clinical Characteristics

The hypermethylation of P16 gene in esophageal cancer tissues was related with the T stage of ESCC. In stage T1/T2, frequency of P16 gene was higher than that in stage T3/T4. Compared with those patients with N0 stage (without regional lymph node metastasis), MGMT gene showed a higher proportion of hypermethylation in cancer tissues, whereas P16 gene showed a higher proportion of hypermethylation in remote normal-appearing esophageal tissues from patients with N1 stage (with regional lymph node metastasis; Table 3).

Table 3.

Association between clinical characteristics and hypermethylation in the ESCC

VariablesnFrequency of hypermethylation (%)
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Site          
    Upper 23 19 (82.6) 8 (34.8) 1 (4.3) 21 (91.3) 8 (34.8) 3 (13.0) 10 (43.5) 
    Middle 68 61 (89.7) 14 (20.6) 2 (2.9) 61 (89.7) 27 (39.7) 7 (10.3) 29 (42.6) 
    Low 34 30 (88.2) 12 (35.3) 1 (2.9) 31 (91.2) 11 (32.4) 4 (11.8) 15 (44.1) 
P  0.653* 0.193 1.000* 1.000* 0.750 0.929* — 0.990 
TNM stage          
         
    T1/T2 67 64 (95.5) 18 (26.9) 2 (3.0) 64 (95.5) 23 (34.3) 8 (11.9) 27 (40.3) 
    T3/T4 58 46 (79.3) 16 (27.6) 2 (3.5) 49 (84.5) 23 (39.7) 6 (10.3) 27 (46.6) 
P  0.005 0.928 1.000* 0.037 0.538 0.778 — 0.482 
         
    N0 66 60 (90.9) 12 (18.2) 2 (3.0) 60 (90.9) 17 (25.8) 7 (10.6) 21 (31.8) 
    N1 59 50 (84.7) 22 (37.3) 2 (3.4) 53 (89.8) 29 (49.2) 7 (11.9) 33 (55.9) 
P  0.290 0.017 1.000* 0.838 0.007 0.824 — 0.007 
         
    M0 124 109 (87.9) 33 (26.6) 4 (3.2) 112 (90.3) 45 (36.3) 14 (11.3) 53 (42.7) 
    M1 1 (100) 1 (100) 1 (100) 1 (100) 1 (100) 
P  1.000* 0.272* 1.000* 1.000* 0.368* 1.000* — 0.432* 
VariablesnFrequency of hypermethylation (%)
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Site          
    Upper 23 19 (82.6) 8 (34.8) 1 (4.3) 21 (91.3) 8 (34.8) 3 (13.0) 10 (43.5) 
    Middle 68 61 (89.7) 14 (20.6) 2 (2.9) 61 (89.7) 27 (39.7) 7 (10.3) 29 (42.6) 
    Low 34 30 (88.2) 12 (35.3) 1 (2.9) 31 (91.2) 11 (32.4) 4 (11.8) 15 (44.1) 
P  0.653* 0.193 1.000* 1.000* 0.750 0.929* — 0.990 
TNM stage          
         
    T1/T2 67 64 (95.5) 18 (26.9) 2 (3.0) 64 (95.5) 23 (34.3) 8 (11.9) 27 (40.3) 
    T3/T4 58 46 (79.3) 16 (27.6) 2 (3.5) 49 (84.5) 23 (39.7) 6 (10.3) 27 (46.6) 
P  0.005 0.928 1.000* 0.037 0.538 0.778 — 0.482 
         
    N0 66 60 (90.9) 12 (18.2) 2 (3.0) 60 (90.9) 17 (25.8) 7 (10.6) 21 (31.8) 
    N1 59 50 (84.7) 22 (37.3) 2 (3.4) 53 (89.8) 29 (49.2) 7 (11.9) 33 (55.9) 
P  0.290 0.017 1.000* 0.838 0.007 0.824 — 0.007 
         
    M0 124 109 (87.9) 33 (26.6) 4 (3.2) 112 (90.3) 45 (36.3) 14 (11.3) 53 (42.7) 
    M1 1 (100) 1 (100) 1 (100) 1 (100) 1 (100) 
P  1.000* 0.272* 1.000* 1.000* 0.368* 1.000* — 0.432* 

Abbreviation: TNM, tumor-node-metastasis.

*

Fisher's exact test.

The Association between MTHFR C677T Polymorphism and Aberrant DNA Methylation

After adjustment for potential confounders, including age, sex, and dietary folate intake, the methylation status of MGMT gene in the cancer tissues was significantly associated with MTHFR C677T genotypes. Compared with the wild genotype of MTHFR 677CC, the ORs of MGMT hypermethylation for CT and TT genotypes were 3.34 (95% CI, 1.07-10.39) and 3.83 (95% CI, 1.13-12.94), respectively (Table 4).

Table 4.

The association between CpG island hypermethylation of cancer-related genes and selected factors analyzed by unconditional logistic regression

VariablesOR (95% CI)*
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Age (y) 1.01 (0.93-1.10) 1.01 (0.95-1.08) — 1.01 (0.92-1.11) 0.96 (0.91-1.02) 0.91 (0.83-1.00) — 0.94 (0.89-1.00) 
Sex         
    Men 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    Women 1.26 (0.36-4.38) 1.32 (0.53-3.24) — 0.85 (0.22-3.19) 1.57 (0.70-3.50) 2.37 (0.65-8.65) — 2.30 (1.03-5.13) 
Folate intake         
    Q1 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    Q2 1.56 (0.37-6.53) 2.53 (0.89-7.14) — 2.59 (0.43-15.50) 0.99 (0.38-2.56) 2.65 (0.58-11.99) — 1.11 (0.44-2.81) 
    Q3 0.74 (0.19-2.87) 0.58 (0.15-2.21) — 0.56 (0.14-2.34) 1.35 (0.48-3.83) 0.34 (0.03-3.87) — 1.23 (0.44-3.48) 
    Q4 — 5.09 (1.02-25.36) — — 1.09 (0.22-5.30) 4.23 (0.51-34.79) — 1.35 (0.29-6.21) 
MTHFR C677T         
    CC 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    CT 0.54 (0.12-2.31) 3.34 (1.07-10.39) — 0.80 (0.17-3.78) 0.66 (0.26-1.68) 2.90 (0.63-13.33) — 0.81 (0.33-2.00) 
    TT 0.89 (0.16-4.88) 3.83 (1.13-12.94) — 1.41 (0.21-9.55) 1.04 (0.38-2.85) 0.80 (0.11-5.72) — 0.95 (0.35-2.63) 
VariablesOR (95% CI)*
Cancer tissue
Remote normal-appearing tissue
P16MGMThMLH1Any geneP16MGMThMLH1Any gene
Age (y) 1.01 (0.93-1.10) 1.01 (0.95-1.08) — 1.01 (0.92-1.11) 0.96 (0.91-1.02) 0.91 (0.83-1.00) — 0.94 (0.89-1.00) 
Sex         
    Men 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    Women 1.26 (0.36-4.38) 1.32 (0.53-3.24) — 0.85 (0.22-3.19) 1.57 (0.70-3.50) 2.37 (0.65-8.65) — 2.30 (1.03-5.13) 
Folate intake         
    Q1 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    Q2 1.56 (0.37-6.53) 2.53 (0.89-7.14) — 2.59 (0.43-15.50) 0.99 (0.38-2.56) 2.65 (0.58-11.99) — 1.11 (0.44-2.81) 
    Q3 0.74 (0.19-2.87) 0.58 (0.15-2.21) — 0.56 (0.14-2.34) 1.35 (0.48-3.83) 0.34 (0.03-3.87) — 1.23 (0.44-3.48) 
    Q4 — 5.09 (1.02-25.36) — — 1.09 (0.22-5.30) 4.23 (0.51-34.79) — 1.35 (0.29-6.21) 
MTHFR C677T         
    CC 1.00 1.00 — 1.00 1.00 1.00 — 1.00 
    CT 0.54 (0.12-2.31) 3.34 (1.07-10.39) — 0.80 (0.17-3.78) 0.66 (0.26-1.68) 2.90 (0.63-13.33) — 0.81 (0.33-2.00) 
    TT 0.89 (0.16-4.88) 3.83 (1.13-12.94) — 1.41 (0.21-9.55) 1.04 (0.38-2.85) 0.80 (0.11-5.72) — 0.95 (0.35-2.63) 
*

Adjusted by age, sex, folate intake, and MTHFR C677T genotypes.

The cut point of fruit folate intake is quartile (μg/d): Q1, 0; Q2, 26.3; Q3, 99.5; and Q4, 311.1.

As one of the results of folate deficiency and metabolic disorder, aberrant methylation of DNA has been focused as a major event of epigenetic changes and plays an important role in the occurrence of various cancers. DNA methylation, a kind of covalent biochemical modification, results in the addition of a methyl (CH3) group at the carbon 5 position of the cytosine ring (12). In human cells, it primarily affects cytosine when they are parts of the symmetrical dinucleotide CpG (23) and this subsequent pattern is transmitted through mitosis and maintained after DNA replication (24). Aberrant CpG island methylation is common in the development of cancers and may play an important role in the carcinogenic process. Aberrant methylation occurring in cancers includes global hypomethylation in genomic DNA as well as hypermethylation in specific gene promoters (25). Global hypomethylation increases mutation rates and chromosomal instability and promoter hypermethylation usually results in transcriptional gene inactivation (16). As DNA methylation is an early event during the cancer pathogenesis and can be detected in many kinds of body fluids, it is supposed to be one of the potential alternative biomarkers for early detection of cancers and for determining the prognosis (12). On the other hand, as DNA methylation is reversible, development of relevant demethylating agents, such as antisense DNA methyl transferase and small interference RNA, is making the field of DNA methylation wider and more exciting (12, 26). With the progressively increasing data on the aberrant methylation in cancers, studies on ESCC, especially in Chinese population, are still limited (16).

In the current study, we found a high hypermethylation of P16 and MGMT genes in both ESCC tissues and remote normal-appearing esophageal tissues. Our results were in line with previous studies on esophageal cancer. A study in Linxian, Henan Province of China reported that aberrant methylation of P16 was detected in 17 of 34 surgically resected esophageal specimens from primary ESCC patients, whereas the homozygous deletion of P16 gene in ESCC was low (17%; ref. 27). It suggested that deletion is not a major cause of P16 inactivation in ESCC compared with aberrant hypermethylation. Another study in Huojia, a county neighboring Linzhou City of northern China, found that DNA hypermethylation could occur as early as basal cell hyperplasia or dysplasia in the development of ESCC and P16 and P14 were most frequently methylated, observed in 19% and 15%, respectively, of the samples examined (28). In Japan, Hibi et al. (29) reported the methylation of the P16 gene was detected in 31 of 38 (82%) tumor samples of ESCC and 7 of these 31 (23%) patients had the same methylation changes in the serum DNA. As P16 is an important tumor suppressor gene, it plays an important role as a potent inhibitor of the cyclin-dependent kinase CDK4A (16). The inactivation of P16 gene allows the tumor cells to progress through G1 checkpoint of the cell cycle and is proved to be associated with the development and metastasis in many cancers (30). The result in the present study indicated that the P16 aberrant methylation is a frequent event in the occurrence of ESCC and it also indicated that P16 methylation may be a good biomarker for early detection of ESCC. Besides tumor suppressor genes, DNA repair genes are also essential in the carcinogenesis. Defects in the normal function of DNA damage repair give rise to hypersensitivity to the mutations in the genome and finally to the development of cancers (31). Until now, up to 130 genes have been identified in humans that are associated with DNA repair (31). MGMT and hMLH1 are two major genes in the pathway of DNA repair and are frequently found to be silenced by CpG island hypermethylation in many cancers, such as lung cancer (32, 33), gastric cancer (34), and esophageal adenocarcinoma (35). However, the studies on the ESCC are still rare. Fang et al. (36) reported 29% (5 of 17) of normal esophageal tissues, 50% (10 of 20) of basal cell hyperplasia, 67% (8 of 12) of dysplasia, and 72% (13 of 18) of ESCC samples obtained from Linzhou City of Henan Province in northern China had DNA hypermethylation in the MGMT promoter region. With the increase of MGMT promoter hypermethylation, the frequency of MGMT mRNA and protein loss to express progressively decreased from normal to basal cell hyperplasia, dysplasia, and ESCC (36). In India, hMLH1 promoter hypermethylation in esophageal tissue was seen in 63.5% of patients with cancer and 53.8% of those with precancerous status, which was significantly increased when compared with controls (37). In the present study, however, hypermethylation of hMLH1 gene was a rare event. One possible explanation is that there might be some difference of carcinogenesis of esophageal cancer in different populations or some exogenous factors as well as genetic factors, which can influence the DNA methylation status.

The term epigenetics, when it was first introduced in the 1940s, was defined as the interactions of genes with their environment, which bring the phenotype into being (38). Several studies have proved that DNA methylation is related with age, diet, and other environmental factors (39-41). About the association between dietary folate and aberrant DNA methylation, we found that higher consumption of folate seemed to increase the hypermethylation rate of MGMT gene in the cancer tissues. It sounds plausible because folate is the major donor of one carbon unit, which is essential for DNA methylation. Although various epidemiologic studies showed the inverse association between folate supplement and cancers, the role of folate supplement for specific population in the carcinogenesis is still controversial (10). Recently, Mason et al. (42) reported a temporal association between folic acid fortification of enriched cereal grains and an increase in the incidence of colorectal cancer in the United States and Canada. Although it is an observational study and only creates a hypothetical foundation, it may be illuminating some important biological principles and arouse further consideration between folate and cancers. One potential harmful consequence of folate fortification is the risk of modification on DNA methylation (10). In some animal studies, dietary folate deficiency can inhibit rather than enhance the development of breast cancer in rats (43, 44), which is in contrast to the observation in epidemiologic studies. It suggested that folate possesses dual effects on carcinogenesis depending on the time and dose of folate intake. An emerging body of evidence suggests that folate supplement may promote the progression of preexisting, undiagnosed precancerous and cancer lesions (10). However, as we only used fruit folate values to estimate the total folate intake, it may not reflect the truth and further more studies are needed to prove the association between high-dose folate and hypermethylation of specific genes.

In addition to folate intake, the activity of folate metabolic enzyme is also involved in the methylation process of DNA. A recent study examined the effect of folate status and the MTHFR C677T polymorphism on genomic DNA methylation in peripheral blood mononuclear cell and found that the MTHFR C677T polymorphism influenced DNA methylation status through an interaction with folate status (45). The role of MTHFR genetic polymorphisms is still controversial in cancers. It might play a dual role in the carcinogenesis depending on the intake of folate. As reported by Friso et al., when compared with subjects with the CC or CT genotype, those with the TT genotype showed a decreased risk of colorectal adenomas when they had high levels of plasma folate (adjusted OR, 0.58; 95% CI, 0.21-1.61) and an increased risk when they had low folate levels (adjusted OR, 2.13; 95% CI, 0.82-5.54; ref. 46). To our knowledge, until now, few studies have been conducted to investigate the role of dietary folate, MTHFR C677T genotype, and their association in promoter methylation in cancers, especially in ESCC. In the present study, after adjustment by potential confounders, we found that individuals carrying variant genotypes CT or TT have a high risk for the methylation of MGMT rather than P16 and hMLH1 genes in esophageal cancer tissues. Why the interplay between MTHFR C677T polymorphism and aberrant hypermethylation is different in different genes is still unexplained. Only a more extensive understanding of the regulation of methylation in relation to gene expression and carcinogenesis will allow us to fully interpret our findings.

One of the major roles of the study on aberrant DNA methylation is to find and apply eligible biomarkers for early diagnosis of cancers because the aberrant DNA methylation is an early event in the occurrence of cancer. The results from our study give us an exciting clue to apply the aberrant hypermethylation status of these cancer-related genes as the biomarkers for early detection and diagnosis for ESCC. If we combine these three cancer-related genes (P16, MGMT, and hMLH1), >90% esophageal cancer samples were hypermethylated and 43.2% remote normal-appearing esophageal tissues were hypermethylated. Although our results were rooted in the DNA extracted from tissues, it can also be generalized to the circulating tumor DNA in peripheral fluids. Based on the newly developed molecular detection methods, circulating tumor DNA can now easily be extracted from serum, plasma, saliva, bronchoalveolar lavage fluid, and urine (47). This small amount of circulating aberrant DNA can be amplified and detected by the high specific and sensitive techniques (47). Ramirez et al. (48) found that methylation in tumor samples and plasma was closely correlated with the ratio of methylation in plasma versus cancer tissues of ∼93%. Unfortunately, in this study, we did not further test the methylation of these cancer-related genes in the blood and compare them with the results from the tissues.

Another goal of studying DNA aberrant methylation is to find the possible applicable biomarkers for clinical prognosis. As shown in the present study, aberrant hypermethylation of P16 and MGMT genes was related with regional lymph node metastasis. With the increase of methylation rate of P16 and MGMT genes, the proportion of regional lymph node metastasis increased simultaneously. As regional lymph node metastasis is associated with the patient's prognosis and treatment strategy, the methylation status of these genes might be used to assess the possibility of recurrence and metastasis of ESCC patients and also help to implement proper medications. More follow-up data are needed to support this hypothesis.

In conclusion, we found the aberrant hypermethylation of cancer-related genes, such as P16 and MGMT, was a frequent event in ESCC. It might be regarded as one of the promising biomarkers for cancer detection in the future. The aberrant hypermethylation of MGMT gene was found to be related with regional lymph node metastasis and also found to be associated with folate metabolic enzyme gene MTHFR C677T polymorphism in ESCC. More studies with large sample size involving more cancer-related genes and environmental factors on the interaction among folate intake, genetic polymorphisms, and DNA methylation are needed to extend our understanding of esophageal cancer biology.

Grant support: National Natural Science Foundation of China (30571603, 30600508) and Innovation Fund for graduate students in Fudan University (2006-2007). Part of the work was also done under a European Union Fellowship offered to J. Wang to complete the Erasmus Mundus Master's course: European Master of Science Programme in International Health (tropEd), University Victor Segalen Bordeaux 2 (Bordeaux, France), with field trainingship in the Team of Epidemiology for Cancer Prevention of the Institut National de la Sante et de la Recherche Medicale U593 research unit.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank the staff of the Yangzhong Cancer Research Institute and Yangzhong People's Hospital and, most importantly, the patients enrolled in this study.

1
Layke JC, Lopez PP. Esophageal cancer: a review and update.
Am Fam Physician
2006
;
73
:
2187
–94.
2
Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002.
CA Cancer J Clin
2005
;
55
:
74
–108.
3
Enzinger PC, Mayer RJ. Esophageal cancer.
N Engl J Med
2003
;
349
:
2241
–52.
4
Gallus S, La Vecchia C. Is there a link between diet and esophageal cancer?
Nat Clin Pract Gastroenterol Hepatol
2007
;
4
:
2
–3.
5
Parkin DM, Bray F, Ferlay J, Pisani P. Estimating the world cancer burden: Globocan 2000.
Int J Cancer
2001
;
94
:
153
–6.
6
Wang JM, Xu B, Rao JY, Shen HB, Xue HC, Jiang QW. Diet habits, alcohol drinking, tobacco smoking, green tea drinking, and the risk of esophageal squamous cell carcinoma in the Chinese population.
Eur J Gastroenterol Hepatol
2007
;
19
:
171
–6.
7
Wei WQ, Abnet CC, Qiao YL, et al. Prospective study of serum selenium concentrations and esophageal and gastric cardia cancer, heart disease, stroke, and total death.
Am J Clin Nutr
2004
;
79
:
80
–5.
8
Kim YI. Role of folate in colon cancer development and progression.
J Nutr
2003
;
133
:
S3731
–9.
9
Larsson SC, Giovannucci E, Wolk A. Folate intake, MTHFR polymorphisms, and risk of esophageal, gastric, and pancreatic cancer: a meta-analysis.
Gastroenterology
2006
;
131
:
1271
–83.
10
Kim YI. Will mandatory folic acid fortification prevent or promote cancer?
Am J Clin Nutr
2004
;
80
:
1123
–8.
11
Duthie SJ. Folic acid deficiency and cancer: mechanisms of DNA instability.
Br Med Bull
1999
;
55
:
578
–92.
12
Das PM, Singal R. DNA methylation and cancer.
J Clin Oncol
2004
;
22
:
4632
–42.
13
Deltour S, Chopin V, Leprince D. Epigenetics and cancer.
Med Sci (Paris)
2005
;
21
:
405
–11.
14
Song C, Xing D, Tan W, Wei Q, Lin D. Methylenetetrahydrofolate reductase polymorphisms increase risk of esophageal squamous cell carcinoma in a Chinese population.
Cancer Res
2001
;
61
:
3272
–5.
15
Zhang J, Zotz RB, Li Y, et al. Methylenetetrahydrofolate reductase C677T polymorphism and predisposition towards esophageal squamous cell carcinoma in a German Caucasian and a northern Chinese population.
J Cancer Res Clin Oncol
2004
;
130
:
574
–80.
16
Sato F, Meltzer SJ. CpG island hypermethylation in progression of esophageal and gastric cancer.
Cancer
2006
;
106
:
483
–93.
17
Macaluso M, Paggi MG, Giordano A. Genetic and epigenetic alterations as hallmarks of the intricate road to cancer.
Oncogene
2003
;
22
:
6472
–8.
18
Zhang ZL, Yang JL, Lu FL, Yin XS, Luo ZG, Wang MR. Mortality rate of malignant tumors in Yangzhong County.
Zhonghua Zhong Liu Za Zhi
1988
;
10
:
102
–4.
19
Wang JM, Xu B, Hsieh CC, Jiang QW. Longitudinal trends of stomach cancer and esophageal cancer in Yangzhong County: a high-incidence rural area of China.
Eur J Gastroenterol Hepatol
2005
;
17
:
1339
–44.
20
Herman JG, Graff JR, Myohanen S, Nelkin BD, Baylin SB. Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands.
Proc Natl Acad Sci U S A
1996
;
93
:
9821
–6.
21
Yang YX, Wang GY, Pan XC. Zhong Guo Shi Wu Cheng Fen Biao 2002. Beijing: Peking University Medical Press; 2002. p. 326–8.
22
Yang YX, He M, Pan XC. Zhong Guo Shi Wu Cheng Fen Biao 2004. Beijing: Peking University Medical Press; 2005. p. 78–215.
23
Issa JP. CpG island methylator phenotype in cancer.
Nat Rev Cancer
2004
;
4
:
988
–93.
24
Gius D, Bradbury CM, Sun L, et al. The epigenome as a molecular marker and target.
Cancer
2005
;
104
:
1789
–93.
25
Momparler RL, Bovenzi V. DNA methylation and cancer.
J Cell Physiol
2000
;
183
:
145
–54.
26
Egger G, Liang G, Aparicio A, Jones PA. Epigenetics in human disease and prospects for epigenetic therapy.
Nature
2004
;
429
:
457
–63.
27
Xing EP, Nie Y, Wang LD, Yang GY, Yang CS. Aberrant methylation of p16INK4a and deletion of p15INK4b are frequent events in human esophageal cancer in Linxian, China.
Carcinogenesis
1999
;
20
:
77
–84.
28
Nie Y, Liao J, Zhao X, et al. Detection of multiple gene hypermethylation in the development of esophageal squamous cell carcinoma.
Carcinogenesis
2002
;
23
:
1713
–20.
29
Hibi K, Taguchi M, Nakayama H, et al. Molecular detection of p16 promoter methylation in the serum of patients with esophageal squamous cell carcinoma.
Clin Cancer Res
2001
;
7
:
3135
–8.
30
Rocco JW, Sidransky D. p16(MTS-1/CDKN2/INK4a) in cancer progression.
Exp Cell Res
2001
;
264
:
42
–55.
31
Christmann M, Tomicic MT, Roos WP, Kaina B. Mechanisms of human DNA repair: an update.
Toxicology
2003
;
193
:
3
–34.
32
Liu Y, Lan Q, Siegfried JM, Luketich JD, Keohavong P. Aberrant promoter methylation of p16 and MGMT genes in lung tumors from smoking and never-smoking lung cancer patients.
Neoplasia
2006
;
8
:
46
–51.
33
Hsu HS, Wen CK, Tang YA, et al. Promoter hypermethylation is the predominant mechanism in hMLH1 and hMSH2 deregulation and is a poor prognostic factor in nonsmoking lung cancer.
Clin Cancer Res
2005
;
11
:
5410
–6.
34
Geddert H, Kiel S, Iskender E, et al. Correlation of hMLH1 and HPP1 hypermethylation in gastric, but not in esophageal and cardiac adenocarcinoma.
Int J Cancer
2004
;
110
:
208
–11.
35
Baumann S, Keller G, Puhringer F, et al. The prognostic impact of O6-methylguanine-DNA methyltransferase (MGMT) promotor hypermethylation in esophageal adenocarcinoma.
Int J Cancer
2006
;
119
:
264
–8.
36
Fang MZ, Jin Z, Wang Y, et al. Promoter hypermethylation and inactivation of O6-methylguanine-DNA methyltransferase in esophageal squamous cell carcinomas and its reactivation in cell lines.
Int J Oncol
2005
;
26
:
615
–22.
37
Vasavi M, Ponnala S, Gujjari K, et al. DNA methylation in esophageal diseases including cancer: special reference to hMLH1 gene promoter status.
Tumori
2006
;
92
:
155
–62.
38
Murrell A, Rakyan VK, Beck S. From genome to epigenome.
Hum Mol Genet
2005
;
14
Spec No 1:
R3
–10.
39
Rodenhiser D, Mann M. Epigenetics and human disease: translating basic biology into clinical applications.
CMAJ
2006
;
174
:
341
–8.
40
Sutherland JE, Costa M. Epigenetics and the environment.
Ann N Y Acad Sci
2003
;
983
:
151
–60.
41
Slattery ML, Curtin K, Sweeney C, et al. Diet and lifestyle factor associations with CpG island methylator phenotype and BRAF mutations in colon cancer.
Int J Cancer
2007
;
120
:
656
–63.
42
Mason JB, Dickstein A, Jacques PF, et al. A temporal association between folic acid fortification and an increase in colorectal cancer rates may be illuminating important biological principles: a hypothesis.
Cancer Epidemiol Biomarkers Prev
2007
;
16
:
1325
–9.
43
Baggott JE, Vaughn WH, Juliana MM, Eto I, Krumdieck CL, Grubbs CJ. Effects of folate deficiency and supplementation on methylnitrosourea-induced rat mammary tumors.
J Natl Cancer Inst
1992
;
84
:
1740
–4.
44
Kotsopoulos J, Sohn KJ, Martin R, et al. Dietary folate deficiency suppresses N-methyl-N-nitrosourea-induced mammary tumorigenesis in rats.
Carcinogenesis
2003
;
24
:
937
–44.
45
Friso S, Choi SW, Girelli D, et al. A common mutation in the 5,10-methylenetetrahydrofolate reductase gene affects genomic DNA methylation through an interaction with folate status.
Proc Natl Acad Sci USA
2002
;
99
:
5606
–11.
46
Marugame T, Tsuji E, Kiyohara C, et al. Relation of plasma folate and methylenetetrahydrofolate reductase C677T polymorphism to colorectal adenomas.
Int J Epidemiol
2003
;
32
:
64
–6.
47
Bremnes RM, Sirera R, Camps C. Circulating tumour-derived DNA and RNA markers in blood: a tool for early detection, diagnostics, and follow-up?
Lung Cancer
2005
;
49
:
1
–12.
48
Ramirez JL, Sarries C, de Castro PL, et al. Methylation patterns and K-ras mutations in tumor and paired serum of resected non-small-cell lung cancer patients.
Cancer Lett
2003
;
193
:
207
–16.