DNA methylation status is inversely correlated with green tea intake and physical activity in gastric cancer patients
Version of Record online: 18 DEC 2008
Copyright © 2008 Wiley-Liss, Inc.
International Journal of Cancer
Volume 124, Issue 11, pages 2677–2682, 1 June 2009
How to Cite
Yuasa, Y., Nagasaki, H., Akiyama, Y., Hashimoto, Y., Takizawa, T., Kojima, K., Kawano, T., Sugihara, K., Imai, K. and Nakachi, K. (2009), DNA methylation status is inversely correlated with green tea intake and physical activity in gastric cancer patients. Int. J. Cancer, 124: 2677–2682. doi: 10.1002/ijc.24231
- Issue online: 25 MAR 2009
- Version of Record online: 18 DEC 2008
- Accepted manuscript online: 18 DEC 2008 12:00AM EST
- Manuscript Accepted: 1 DEC 2008
- Manuscript Received: 18 JUL 2008
- Ministry of Education, Culture, Sports, Science, and Technology, Japan. Grant Number: 17015013
- Grant-in-Aid for Scientific Research on Priority Areas-Cancer
- DNA methylation;
- green tea;
- gastric cancer;
Epigenetic silencing of genes by aberrant DNA methylation is recognized as a crucial component of the mechanism underlying tumorigenesis. However, the relationship between DNA methylation and the past lifestyle in cancer patients remains largely unknown. We examined the methylation statuses of 6 tumor-related genes, CDX2 (homeobox transcription factor), BMP-2 (bone morphogenetic protein 2), p16 (INK4A), CACNA2D3 (calcium channel-related), GATA-5 (transcription factor) and ER (estrogen receptor), in 106 primary gastric carcinomas by methylation-specific PCR and compared them with the past lifestyles of the patients. The methylation frequencies of the genes were 23.6, 21.7, 9.4, 32.4, 40.8 and 59.1%, respectively. Significant association was found between a decreased intake of green tea and methylation of CDX2 and BMP-2. More physical activity was correlated with a lower methylation frequency of CACNA2D3. Of these 6 genes, the methylation statuses of CDX2, BMP-2 and p16 revealed a significant interrelationship and those of CACNA2D3, GATA-5 and ER did likewise. Thus, some epidemiological factors, such as green tea intake, could be important as to determination of the methylation statuses of selected genes and may influence the development of cancer, including that of the stomach. © 2008 Wiley-Liss, Inc.
Epigenetic changes, particularly methylation of cytosine in CpG dinucleotides in gene promoters, are found in almost every type of human neoplasm and are associated with transcriptional gene silencing.1, 2 Such promoter hypermethylation is as common as the disruption of tumor-suppressor genes in human cancer by mutation. Unlike irreversible genetic changes, epigenetic changes are thought to possibly be reversible by the environment, diet or pharmacological intervention. For example, monozygotic twins are considered genetically identical and are thus ideal for studying the effects of environmental and dietary factors on human health and diseases. In a study of a large cohort of identical twins, the patterns of DNA methylation across the genome were found to be very similar in young monozygotic twins in several cell types, but in older twins the patterns diverged.3 This strongly suggests that 1 or more environmental factors affect individuals throughout life, modifying gene expression through epigenetic mechanisms that have important implications for health.
Dietary factors are important determinants of cancer risk.4 Aberrant DNA methylation is associated with dietary factors and other lifestyle factors and may underlie carcinogenesis. The prevalence of promoter hypermethylation of 6 genes, such as APC, p14ARF, p16/INK4a (hereafter p16) and hMLH1, was higher in colorectal cancers derived from patients with a low folate/high alcohol intake than in ones with a high folate/low alcohol intake, but the differences were not statistically significant.5 The incidences of hypermethylation of D17S5 and p16 in lung cancer are significantly higher in cigarette smokers than in those who have never smoked.6–8 However, the relationship between DNA methylation and the past lifestyle in cancer patients remains largely unknown.
In 2000, gastric cancer was the second most frequent cause of cancer death worldwide.9 Infection with Helicobacter pylori is a strong risk factor for gastric cancer but is not a sufficient cause for its development.10 Epidemiological studies have strongly suggested that the risk may be increased with a high intake of salt and salt-preserved foods and decreased with a high intake of fruit and vegetables.11 The aberrant methylation of many genes has been reported in gastric cancer.12–14 We previously reported that CDX2 methylation in men was correlated with a decreased intake of green tea, suggesting that diet could be an important factor determining the methylation status of genes such as CDX2 and the resultant aberrant expression of genes involved in carcinogenesis.15 However, these effects may not be universal but gene-specific, and female patients have not been examined. Thus, we analyzed the methylation states of 6 genes in more gastric cancer patients. Five of the 6 genes, that is, CDX2,15BMP2,16p16,17CACNA2D318 and GATA5,19 were often methylated in gastric cancers but rarely in noncancerous epithelia. We, then, compared the relationship between DNA methylation and the past lifestyle in cancer patients including female ones.
Material and methods
Cancer tissue specimens were collected from 106 consecutive patients with primary gastric carcinoma in a hospital affiliated to Tokyo Medical and Dental University during 2000–2005. Informed consent was obtained from all patients, and the study was approved by the institutional review committee of Tokyo Medical and Dental University. A self-administered questionnaire was used in this study to assess the lifestyle before cancer onset, covering the disease history, familial history of cancer, medication, cigarette smoking, alcohol consumption, physical activity, intake frequencies of selected food groups and food items, daily consumption of tea (green tea, oolong tea and black tea), regularity of sleep and meals, eating quantity, bowel motion, height and body weight. The food groups were beef, pork, chicken, ham/sausage/bacon, grilled meat, all meat, grilled fish, salted/dried/other processed fish products, pickled vegetables, green leaf vegetables, yellow colored vegetables, cruciferous vegetables, all vegetables, fruits and probiotics-fermented milk. The intake frequencies of these food groups were categorized into not eaten, 1–2 times/month, 1–2 times/week, 3–4 times/week, almost every day and almost every meal. Most lifestyle factors in this questionnaire were selected from those which had previously been reported to be risk or preventive factors for gastric and colon cancers on epidemiological observation.
Methylation analyses by the methylation-specific PCR procedure
We extracted genomic DNA from paraffin-embedded tissues by the phenol-chloroform method, and then carried out bisulfite modification and the methylation-specific PCR (MSP) procedure as previously described.22 The primer sequences of the CDX2, BMP-2, p16, CACNA2D3, GATA5, and estrogen receptor (ER) genes for the MSP analyses are shown in Table I. The PCR reaction was performed for 35 cycles in a 25 μl mixture comprising bisulfite-modified DNA (∼50 ng), 2.5 μl of 10 × PCR buffer, 1.25 μl of 25 mM dNTP, 10 pmole of each primer and 1 U of JumpStart Red Taq polymerase (Sigma, St. Louis, MO). Each PCR cycle consisted of 95°C for 30 sec, 58°C for 30 sec and 72°C for 30 sec, followed by final extension at 72°C for 5 min. The PCR products were electrophoresed in 2.5% agarose gels. All the MSP procedures were repeated more than twice. The methylation statuses of CDX2 and CACNA2D3 in several gastric cancer samples were also analyzed by LightCycler real-time PCR using bisulfite-modified DNA and methylation-specific primers, and the results were concordant with the MSP results.
The promoter methylation status of specific genes, clinico-pathological parameters and lifestyle variables in the patients were computed. Differences in frequency by methylation status were tested using the χ2 test, and differences in mean values were tested using the t test. The association between the methylation status and dietary variables was also analyzed using a nonparametric test (Mann-Whitney U test). We further studied the association using the backward elimination (Wald test) method of logistic regression analysis. In this analysis, the intake frequencies of food groups were dichotomous as follows: ≤6 cups/day vs. ≥7 cups/day for green tea, ≤twice/week vs. ≥3 times/week for pickled vegetables and ≤twice/week vs. ≥3 times/week for drinking. Physical activity was defined as “recreational and voluntary physical exercise for health promotion” and primarily divided into 4 categories in questionnaire as follows: Never, 1–2 hr/week, 3–4 hr/week and ≥5 hr/week. However, in the logistic regression analysis, we combined these categories into 2 groups, ≥1 hr/week vs. never. Pearson's contingency coefficients for methylation status of an every pair of the 6 genes were calculated in 106 gastric carcinomas. p for trend was calculated by the Cochran-Armitage test. The statistical software used was SPSS software (version 14.0).
Methylation statuses of CDX2, BMP-2, p16, CACNA2D3, GATA5 and ER in primary gastric carcinomas
The methylation statuses of CDX2, BMP-2, p16, CACNA2D3, GATA5 and ER were determined in 106 primary gastric carcinomas by MSP analyses using the specific primers shown in Table I. The methylation frequencies of these genes were 23.6, 21.7, 9.4, 32.4, 40.8 and 59.1%, respectively. When we examined the CDX2 methylation statuses of noncancerous gastric tissues in 12 patients with methylation-positive gastric cancer and in 13 patients with methylation-negative one, and the CACNA2D3 methylation statuses of noncancerous tissues in 7 patients with methylation-positive gastric cancer and 19 patients with methylation-negative one, we found no methylation in any samples by MSP, indicating cancer-related methylation of these genes (data not shown).
The relationship between methylation frequencies of the 6 genes and clinicopathological parameters
The clinico-pathological characteristics of the studied patients by the methylation statuses of CDX2, BMP-2, p16, CACNA2D3, GATA5 and ER are shown in Table II. The methylation of p16 was significantly more frequent in diffuse type (8/49, 16.3%) than in intestinal (2/57, 3.5%) type gastric carcinomas (p = 0.04). CACNA2D3 methylation was more frequently found in lymph node metastasis-positive cases (20/45, 44.4%) than in negative ones (14/60, 23.3%) (p = 0.03). In contrast, there was no statistically significant correlation between methylation of 4 of the genes, CDX2, BMP-2, GATA5 and ER, and clinico-pathological parameters (Table II).
|CDX2 (n = 106)||BMP-2 (n = 106)||p16 (n = 106)|
|Methylated (n = 25)||Unmethylated (n = 81)||p value||Methylated (n = 23)||Unmethylated (n = 83)||p value||Methylated (n = 10)||Unmethylated (n = 96)||p value|
|Age (mean ± SD)||66.0 ± 10.6||63.8 ± 10.3||0.35||63.6 ± 8.8||64.5 ± 10.8||0.73||59.4 ± 12.1||64.8 ± 10.0||0.12|
|Size (cm, mean ± SD)||5.7 ± 4.3||5.6 ± 3.5||0.91||5.6 ± 4.6||5.7 ± 3.4||0.92||6.1 ± 4.0||5.6 ± 3.7||0.69|
|Depth of tumor invasion|
|mp - si||10||42||10||42||7||45|
|Lymph node metastasis|
|CACNA2D3 (n = 105)||GATA5 (n = 98)||ER (n = 93)|
|Methylated (n = 34)||Unmethylated (n = 71)||p value||Methylated (n = 40)||Unmethylated (n = 58)||p value||Methylated (n = 55)||Unmethylated (n = 38)||p value|
|Age (mean ± SD)||65.8 ± 11.2||63.8 ± 9.9||0.35||64.9 ± 9.8||64.1 ± 10.9||0.71||65.7 ±9.9||62.3 ± 10.3||0.12|
|Size (cm, mean ± SD)||5.9 ± 3.6||5.5 ± 3.7||0.61||6.4 ± 4.2||5.4 ± 3.4||0.23||5.5 ± 3.4||5.8 ± 3.8||0.72|
|Depth of tumor invasion|
|mp - si||18||34||19||33||27||19|
|Lymph node metastasis|
The relationship between methylation frequencies of the 6 genes and epidemiological parameters in gastric carcinoma patients
As shown in Table III, the methylation frequencies of CDX2 and BMP-2 were lower in patients consuming 7 cups or more per day of green tea than those consuming 6 cups or less per day (2/25 (8%) vs. 22/80 (27.5%), p = 0.06 and 1/25 (4%) vs. 22/80 (27.5%), p = 0.02, respectively). Patients consuming more pickled vegetables exhibited a higher methylation frequency of GATA5 than ones consuming less (p = 0.04). CACNA2D3 methylation was more frequently found in patients with no physical activity (20/44, 45.5%) than in those with more physical activity (14/59, 23.7%) (p = 0.03). In contrast, there was no statistically significant correlation between methylation of 2 of the genes, p16 and ER, and clinico-pathological parameters (Table III).
|CDX2 (n = 106)||BMP-2 (n = 106)||p16 (n = 106)|
|Menthylated (n = 25)||Unmethylated (n = 81)||Univariate p value||Menthylated (n = 23)||Unmethylated (n = 83)||Univariate p value||Menthylated (n = 10)||Unmethylated (n = 96)||Univariate p value|
|CACNA2D3 (n = 105)||GATA5 (n = 98)||ER (n =93)|
|Menthylated (n = 34)||Unmethylated (n = 71)||Univariate p value||Menthylated (n = 40)||Unmethylated (n = 58)||Univariate p value||Menthylated (n = 55)||Unmethylated (n = 38)||Univariate p value|
When the intake of green tea was stratified, the prevalence of aberrant methylation of CDX2 and BMP-2 decreased significantly with a higher intake of green tea (Mann-Whitney U-test, both p = 0.04; Cochran-Armitage test, ptrend = 0.03 and 0.02, respectively) (Figs. 1a and 1b). A distinct distribution of patients with methylated and unmethylated CACNA2D3 was also demonstrated for physical activity (Mann-Whitney U-test, p = 0.03; Cochran-Armitage test, ptrend = 0.03) (Fig. 1c). On the other hand, an increased intake of pickled vegetables was not associated with an increased methylation frequency of GATA5 (p = 0.11).
Since dietary factors are closely interrelated, we further performed the backward elimination (Wald) method of logistic regression analysis of the methylation status of each gene in gastric cancer patients including female ones (Table IV). A significant association was found between the intake of green tea and methylation of 2 of the genes, CDX2 and BMP-2. Increased daily consumption of green tea (7 cups or more per day) showed a significant association with decreased methylation frequencies of CDX2 and BMP-2 after adjustment (p = 0.04 and p = 0.049, respectively). On the other hand, an increased methylation frequency of CACNA2D3 was associated with less physical activity (negative versus positive), adjusting for confounding variables (p = 0.06) (Table IV). As for factors other than dietary ones, the logistic regression analysis also showed significant associations between CDX2 methylation and gender or histology, p16 methylation and histology, and CACNA2D3 methylation and lymph node metastasis (Table IV).
|Gender (men vs women)||−1.87||0.78||0.02|
|Histology (intestinal vs diffuse)||2.61||0.78||0.001|
|Green tea (≤6 cups/day vs. ≥7 cups/day)||−1.87||0.92||0.04|
|Green tea (≤6 cups/day vs. ≥7 cups/day)||−2.08||1.06||0.049|
|Histology (intestinal vs. diffuse)||2.06||6.03||0.01|
|Lymph node metastasis (negative vs. positive)||0.93||0.45||0.04|
|Physical activity (never vs. ≥1 hr/week)||−0. 86||0.45||0.06|
Interrelationship of the 6 genes relative to their methylation statuses in gastric carcinomas
The methylation statuses of the 6 genes in 106 gastric carcinomas are shown in Figure 2. The methylation patterns of 3 genes, CACNA2D3, GATA-5 and ER, were distinct from other 3 genes, CDX2, BMP-2 and p16. To determine the relationship of the methylation statuses among the 6 genes, Pearson's contingency coefficients for methylation status of an every pair of the 6 genes were calculated (Table V). On the basis of the contingency coefficients, we found that the 6 genes were divided into 2 groups, Group I (CDX2, BMP-2 and p16) and Group II (CACNA2D3, GATA-5 and ER), where a statistically significant interrelationship within each group but no intergroup association was noted. The methylation frequencies in gastric carcinomas were lower for Group I genes (CDX2, 23.6%; BMP-2, 21.7% and p16, 9.4%) than Group II ones (CACNA2D3, 32.4%; GATA-5, 40.8% and ER, 59.1%).
As described earlier, aberrant methylation of CDX2 and BMP-2 was inversely correlated with green tea intake. The prevalence of p16 methylation was also higher in patients with a lower green tea intake than those with a higher intake, although the difference was not significant. When we analyzed the relationship between green tea intake and the methylation of combinations of Group I genes by the multinominal logistic regression model, the odds ratios of methylation for any 1 gene and ≥2 genes vs. no methylation were 4.9 (confidence interval (CI) 1.0–24.3) and 14.8 (CI 1.1–206.7), respectively, in patients consuming 6 cups or less per day of green tea compared with those consuming 7 cups or more per day. On the other hand, no lifestyle factors were associated with the methylation of combinations of Group II genes.
The methylation frequencies of the 6 genes in 106 gastric carcinomas varied from 9.4 to 59.1%. The prevalence of promoter hypermethylation of CDX2 and BMP-2 was significantly higher in gastric carcinomas derived from patients with a low green tea intake than those with a high intake. When we analyzed the association between the methylation status and variables using a nonparametric test, increased intake of green tea was found to be significantly associated with decreased methylation frequencies of CDX2 and BMP-2. In a previous study,15 methylation of 1 of 3 genes, CDX2, was correlated with a decreased intake of green tea in 58 male gastric carcinoma patients. Since an inverse relationship with green tea intake was also found for BMP-2 promoter methylation in 106 gastric carcinoma patients including female ones in this study, the effect of green tea on the decrease of gene promoter methylation might be more common for many genes.
The evidence derived from epidemiologic studies on the relationship between drinking of green tea and cancer-preventive effects is inconclusive; some indicated preventive effects23, 24 and some did not.25, 26 In a detailed study, consumption of green tea was found to be associated with a decreased risk of gastric carcinoma in Japanese women after adjustment for potential confounding factors, whereas no association was observed among Japanese men.27 The difference between women and men might be explained by the much higher cigarette smoking rate in men than women in Japan, which may play a role as an effect modifier.28
Green tea contains several polyphenolic compounds, such as (−)-epigallocatechin-3-gallate (EGCG). A significant inhibitory effect of EGCG on chemical carcinogenesis in the rat stomach has been reported.29 As for its action on methylation, it was reported that EGCG dose-dependently inhibited DNA methyltransferase activity in several cancer cells, resulting in reactivation of methylation-silenced genes, such as retinoic acid receptor β, p16 and hMLH1.30, 31 Polyphenols are rapidly metabolized to forms with quite different bioactivities. But the epithelial surfaces of the gut, particularly those of the esophagus and the stomach, are exposed on the luminal side to high concentrations of tea polyphenols before they undergo metabolism. These characteristics may make gastrointestinal epithelial tissues particularly susceptible to what are probably the beneficial effects of DNA methyltransferase inhibitors.
In other studies, however, EGCG did not inhibit DNA methyltransferase activity or reactivate genes, whereas nucleoside analogue methylation inhibitors, such as 5-aza-2′-deoxycytidine, were far more effective.32, 33 We also analyzed the effect of EGCG on transcriptional levels of CDX2 and BMP-2 in 3 human gastric cancer cell lines. Upregulation of both genes was not found in any cell lines (Hashimoto et al., personal communication). Thus, further studies are necessary to determine how tea polyphenols act on DNA methylation.
CACNA2D3 methylation was more frequently found in gastric carcinoma patients with no physical activity than in those with physical activity. It is known from epidemiological studies that physical activity protects against cancers of the colon, breast (postmenopause) and endometrium.4 As for gastric carcinoma, a population based case-control study in Canada and a prospective cohort study in Norway indicated that recreational physical activity might have a protective effect against gastric cancer.34, 35 To determine the association of physical activity with promoter hypermethylation of APC and RASSF1A in breast tissue, a cross-sectional study on 45 women without breast cancer was performed, which revealed that physical activity was inversely associated with promoter hypermethylation of APC but not RASSF1A.36 It is, therefore, possible that physical activity may affect the methylation of genes, such as CACNA2D3, and gastric carcinogenesis.
There are 2 types of genes according to contingency coefficients for methylation status in gastric carcinomas. The methylation frequencies were lower for Group I genes (9.4–23.6%) than Group II ones (32.4–59.1%). The odds ratios of methylation for any 1 gene and ≥2 genes vs. no methylation among the Group I genes were much higher in patients consuming 6 cups or less per day of green tea than in those consuming 7 cups or more per day. These data suggest that green tea intake may be inversely related to the methylation of Group I genes, which may be involved in carcinogenesis. The logistic regression analysis also showed significant associations between histology and the methylation of CDX2 and p16. But there was no association between BMP-2 methylation and histology. Further investigations are required to clarify the significance of the 2 different types of genes as to methylation in gastric carcinomas.
In conclusion, there were inverse associations between the intake of green tea and the methylation of CDX2 and BMP-2, and between physical activity and CACNA2D3 methylation. We, therefore, hypothesized that some of the lifestyle factors, which have been reported to be preventive as to gastric cancer on epidemiological observation, may influence the development of gastric cancer through the demethylation or retaining of unmethylated status of selected genes. Because an epigenetic drift may contribute to the development of cancer, strategies involving changes in lifestyle including diet might be highly beneficial in preventing/reversing epigenetic alterations and counteracting cancer.
This work was supported by Grant-in-Aid for Scientific Research on Priority Areas-Cancer 17015013 from the Ministry of Education, Culture, Sports, Science, and Technology, Japan (Y. Yuasa). The authors thank Ms. Y. Ozawa for assistance in the preparation of the manuscript.
- 4Marmot M,Atinmo T,Byers T,Chen J,Hirohata T,Jackson A,James WPT,Kolonel LN,Kumanyika S,Leitzmann C,Mann J,Powers HJ, et al., eds. Food, nutrition, physical activity, and the prevention of cancer: a global perspective. Washington, DC: World Cancer Research Fund/American Institute for Cancer Research, 2007. 517p.