Abstract
Reduced levels of global DNA methylation, assessed in peripheral blood, have been associated with bladder cancer risk in European and American populations. Similar data are lacking in Asian populations where genetic differences, lifestyle factors, and different environmental exposures may affect DNA methylation and its risk relationship with bladder cancer. The association between global DNA methylation measured at long interspersed nuclear element (LINE-1) repeat regions through bisulfite pyrosequencing in lymphocyte DNA and bladder cancer risk was examined in a case-control study of 510 bladder cancer patients and 528 healthy control subjects in Shanghai, China. In an initial analysis restricted to control subjects, LINE-1 methylation was elevated among men, those who frequently consumed cruciferous vegetables, and those with a null genotype for either glutathione S-transferase M1 (GSTM1) or GSTT1. In contrast, reduced LINE-1 methylation was found in current smokers with a high cytochrome P450 1A2 (CYP1A2) phenotype index. In a case-control analysis, there was no significant association of LINE-1 methylation with case status, although reduced LINE-1 methylation was associated with increased risk of bladder cancer among never smokers (P for trend = 0.03); analysis by tertile revealed odds ratios (ORs) of 1.91 (lowest tertile; 95% CI = 1.17–3.13) and 1.34 (middle tertile; 95% CI = 0.79–2.28) when compared to the highest tertile. This association was strongest among nonsmokers null for either the GSTM1 or GSTT1 genotype (P for trend = 0.006). Further research is needed to understand the relationships between methyl group availability and LINE-1 methylation in relation to bladder cancer risk.
Introduction
Bladder cancer is the ninth most common form of cancer in the world, with men having three to four times the risk of women1. The major known risk factors include exposure to tobacco smoke, aromatic hydrocarbons, water chlorination byproducts, inorganic arsenic, and (in the case of squamous cell bladder cancer) Schistosoma hematobium infection1,2. Genetic variants in enzymes that metabolize bladder carcinogens such as N-acetlytransferase (NAT), cytochrome P450 (CYP) 1A2, and glutathione S-transferases (GSTM1, GSTT1) have been reported to modify the association of carcinogen exposure with bladder cancer risk1,3,4. Specifically, increased risk of bladder cancer is associated with GSTM1 and GSTT1 null genotypes and elevated CYP1A2 activity 2,3,5,6.
Bladder carcinomas exhibit somatic genetic and epigenetic alterations. Genetic changes are common and include point mutations as well as gene amplification and deletion7. Epigenetic alterations are heritable DNA modifications that do not involve changes in the DNA sequence. Epigenetic changes are associated with alterations in gene expression and are important in maintaining genomic stability8,9,10. DNA methylation-associated gene silencing occurs in tandem with modifications of chromatin, involving specifically the catalytic transfer of methyl groups to the 5-carbon of cytosine in a CpG dinucleotide9. Dietary folate and one-carbon compounds are an important source for methyl groups that are necessary for DNA methylation11,12.
It has long been known that global DNA hypomethylation, reflected in reduced levels of methylation in repeat regions, occurs in target tissues undergoing carcinogenic de-differentiation and can be used as a biomarker of malignant disease13,14. DNA hypomethylation and its coordinate epigenetic changes have been proposed as an integral component of cancer development, contributing to both the loss of genomic stability in regions that are generally heavily methylated, as well as being associated with alterations in gene expression (which can occur when gene promoter regions become abnormally hypomethylated). Finally, activation of transposable elements can occur upon reduced DNA methylation, resulting in insertional genetic mutations 9,10,13,15. In addition, global DNA methylation levels, assessed in repeat regions from lymphocyte-derived DNA, have been associated with risk of malignant solid tumors13,16, including bladder cancer in Spanish and American populations 17,18. Previous work suggests that the level of global DNA methylation in lymphocytes may be a surrogate for systemic global DNA methylation19,20. Here, we have investigated the relationship between LINE-1 DNA methylation, a measure of global methylation, and risk of bladder cancer in a Chinese population in Shanghai, China. In addition, we examined the association between LINE-1 methylation and potential risk factors for bladder cancer including tobacco smoking, genetic polymorphisms, and certain dietary exposures, as well as their modifying effect on the LINE-1-bladder cancer risk estimate.
Materials and Methods
Subjects
The present study included participants of the Shanghai Bladder Cancer Study. The study design has been described in detail elsewhere21. Briefly, bladder cancer cases were Han Chinese aged 25–74 at the time of diagnosis who were permanent residents of the city of Shanghai, China. All cases diagnosed between 1st July 1995 and 30th June 1998 who were registered in the Shanghai Cancer Registry were eligible for the Shanghai Bladder Cancer Study. Among 708 bladder cancer cases identified, 56 were either deceased or too ill to be interviewed, 29 refused to participate in the study, and 42 were unable to be located. The remaining 581 (82%) eligible patients were interviewed between July 1996 and June 1999. The diagnosis of bladder cancer for 531 (91%) patients was made based on histopathological evidence whereas the remaining 50 (9%) patients’ diagnoses were based on positive computerized axial tomography scan and/or ultrasonograph with consistent clinical history.
Control subjects were randomly selected from the urban population of Shanghai through the Residents Registry of the Shanghai Municipal Government. The control subjects were chosen to match the frequency distribution by sex and 5-year age groups of bladder cancer patients as ascertained by the Shanghai Cancer Registry during 1990–1994. Among the 750 potential control subjects chosen, 74 subjects could not be located due to change of home addresses. Seventy-two subjects refused to participate in the study. The remaining 604 (81%) subjects were interviewed during the same time period as the cases. All subjects provided informed consent following procedures approved by the appropriate institutional review boards.
Data Collection
An in-person interview with each eligible study subject was conducted for approximately one hour by a trained interviewer using a structured questionnaire in the subject’s home. The questionnaire gathered information on subject demographics, history of tobacco use, secondhand smoke (for lifelong nonsmokers only), history of beverage consumption (coffee, tea, soda, alcohol, and water), use of hormones (for women only), medical history, dietary history, and occupational history, two years prior to the diagnosis of bladder cancer for case patients and two years prior to the date of interview for control subjects (reference date). For lifelong nonsmokers, the questionnaire further asked for the smoking history of their mother, father, spouse(s), and other relatives who ever lived in the same household with the subject, as well as the smoking habits of coworkers in an indoor environment. For each subject, a composite index denoting total environmental tobacco smoke (ETS) exposure based on the 5 sources of ETS over subject’s lifetime was constructed and described in detail previously21.
All subjects were asked to donate blood and urine samples at the end of the in-person interview. A total of 513 (88%) of cases and 534 (88%) of controls provided a blood sample. Blood samples were collected in heparinized (10 ml) and non-heparinized (4 ml) tubes. Heparinized samples were fractioned into plasma, buffy coat, and erythrocytes on the day of the sample collection, and were stored at −80°C. Forty-six case patients and 61 control subjects refused to donate overnight urine samples. Prior to the collection of an overnight urine sample, each consenting subject was given two packets of Nestle instant coffee or two cans of Coca-Cola Classic drink (about 70 mg of caffeine) to be consumed between 3 and 6 pm. The subject then collected an overnight urine sample (ending with the first morning void) into a plastic jar that was picked up by the same interviewer in the following morning. The urine samples were processed, acidified (400 mg of ascorbic acid per 20 mg of urine), and stored at 80°C on the same day of urine pickup until analysis
Laboratory measurements
Genotyping
DNA was extracted from peripheral blood buffy coats using QIAmp DNA mini kit according to manufacturer’s protocol (Qiagen, Valencia, CA). A standard, multiplex polymerase chain reaction protocol was used to analyze for the presence or absence of the glutathione S-transferase M1 (GSTM1) and GSTT1 genes, as described in detail previously3. For subjects with available blood samples, we obtained GSTM1 genotype on 504 cases and 529 controls, and GSTT1 genotype on 503 cases and 528 controls.
Phenotyping
Urinary caffeine metabolites, namely 5-acetylamino-6-amino-3-methyluracil (AAMU), 1-methylxanthin (MX), 1-methyluric acid (MU) and 1,7-dimethylxanthin (17X), were measured by the following methods. Levels of AAMU in urine were determined by a modified procedure that was previously described22, using high-performance size exclusion chromatography. Quantification of MX, MU, and 17X in urine was performed according to a modified procedure23. These assays were performed with appropriate internal standards. Calibration curves were created during the analysis and used for calculation of concentrations of all analytes. Quality control urine samples spiked with a low, intermediate, and high range of the calibration concentrations were analyzed intermittently during the sample runs. The CYP1A2 phenotype scores were determined based on ratios of urinary caffeine metabolites, i.e. (AAMU + MX + MU)/17X. Higher ratio values of the CYP1A2 phenotype score reflect higher CYP1A2 activities. We used the median value of the CYP1A2 phenotype score in all control subjects to classify subjects into low (≤5.53) or high (>5.53) CYP1A2 phenotypic activity status. Of all urine samples, we were unable to detect caffeine metabolites in 45 samples (13 from case patients and 32 from control subjects).
DNA LINE-1 Methylation
One μg of peripheral lymphocyte DNA was sodium bisulfite modified using the EZ DNA Methylation Kit according to manufacturer’s protocol (Zymo Research, Orange, CA). LINE-1 region methylation extent was quantified using quantitative bisulfite Pyrosequencing24 as previously described25, which examines the cytosine methylation status at 4 CpG sites in the LINE-1 region. All PCR reactions were performed using Qiagen Hot Star Taq polymerase, and each batch included a no template control, unmodified DNA control, and a standardized methylation control. Each sample was run in triplicate, and each pyrosequencing reaction used 20μl of PCR product, and was run according to instrument/manufacturer’s protocols on a PyroMark™ MD System (Qiagen). The standard error of the averaged individual repeats was found to be the same as the standard error for each replicate, so the average measure (percentage) of LINE-1 methylation across the 4 CpG sites for each replicate was used to calculate an average of the replicates for each sample. The measure of methyl cytosine at position is relative to the total cytosine and thymine at that position in the amplified repeats. The assays for LINE-1 failed on 9 samples (3 cases and 6 controls).
Statistical Methods
For the present analysis, we included 510 (88% of eligible) case patients and 528 (87% of eligible) control subjects with available LINE-1 measurement. In the analyses stratified by CYP1A2 phenotypic status, we excluded an additional 13 cases and 32 controls with unknown CYP1A2 phenotype scores. In the analyses stratified by GSTM1 and GSTT1 genotypes, we excluded 9 cases and 4 controls with unknown GSTM1 and/or GSTT1 genotypes.
Chi-square test was used to examine the differences in the distributions of categorical variables and t-test for the differences in means of continuous variables between case patients and control subjects. The analysis of covariance (ANCOVA) method was applied to evaluate the effect of smoking, dietary, and other environmental factors and genetic determinants on LINE-1 methylation scores among control subjects only, separately for men and women.
Unconditional logistic regression models were used to examine the association between LINE-1 methylation scores and risk of bladder cancer. Study subjects were classified into tertiles based on the distributions of LINE-1 scores among all control subjects only. The strength of the association between exposure and bladder cancer risk was measured by odds ratios (ORs) and their corresponding 95% confidence intervals (CIs) and P-values. The association between the LINE-1 scores and bladder cancer risk was examined in total subjects as well as subgroups defined by gender and cigarette smoking status. Additional analyses for the association between LINE-1 score and bladder cancer risk were conducted in subjects stratified by intake frequency of cruciferous vegetables dichotomized to represent the lowest 10% of consumption (<4 or ≥4 times per week), the joint genotypes of GSTM1 and GSTT1, and CYP1A2 phenotype status. We did not show the null results from the latter subgroup analyses in tables. Gender, age at reference date (years) and family history of cancer (no, yes) were included in all logistic regression models. The composite index of ETS exposure was added to the covariate list when the analysis was restricted to nonsmokers only.
Statistical analyses were performed using SAS version 9.1 (SAS Institute, Cary, NC) statistical software package. All P values are two sided. P values less than 0.05 were considered statistically significant.
Results
Men accounted for 78% of cases and 77% of controls. The mean age (±standard deviation) of case patients at diagnosis of bladder cancer was 62.18 (±9.9) years while the mean age of control subjects at interview was 60.80 (±10.1) years (P=0.03). As expected, bladder cancer patients were more likely to smoke cigarettes than control subjects (Table 1). Case patients also were more likely to have an elevated level of CYP1A2 activity and a null genotype of both GSTM1 and GSTT1 (Table 1).
Table 1.
Distributions of selected characteristics in patients with bladder cancer (cases) and control subjects, The Shanghai Bladder Cancer Case-Control Study
| No. cases (%) | No. controls (%) | 2-sided P | |
|---|---|---|---|
| Total subjects | 510 (100) | 528 (100) | 0.312 |
| Age at reference year | 0.016 | ||
| <50 | 83 (16.3) | 59 (11.2) | |
| 50–<60 | 76 (14.9) | 94 (17.8) | |
| 60<70 | 275 (53.9) | 270 (51.1) | |
| ≥70 | 76 (14.9) | 105 (19.9) | |
| Gender | 0.505 | ||
| Male | 400 (78.4) | 405 (76.7) | |
| Female | 110 (21.6) | 123 (23.3) | |
| Education | 0.615 | ||
| No formal schooling | 43 (8.4) | 39 (7.4) | |
| Primary school | 122 (23.9) | 135 (25.6) | |
| Middle school | 283 (55.5) | 301 (57.0) | |
| College and above | 62 (12.2) | 53 (10.0) | |
| Body mass index, kg/m2 | 0.121 | ||
| <18.5 (underweight) | 44 (8.6) | 48 (9.1) | |
| 18.5–24.9 (normal) | 364 (71.4) | 400 (75.8) | |
| ≥25 (overweight and obese) | 102 (20.0) | 80 (15.2) | |
| Smoking status at reference date | 0.001 | ||
| Nonsmokers | 178 (34.9) | 236 (44.7) | |
| Former smokers | 79 (15.5) | 89 (16.9) | |
| Current smokers | 253 (49.6) | 203 (38.5) | |
| CYP1A2 phenotypic score levels | |||
| Low (≤5.53) | 216 (42.4) | 248 (47.0) | 0.003 |
| High (>5.53) | 281 (55.1) | 248 (47.0) | |
| Intake of total cruciferous vegetables * | 0.205 | ||
| Less than 4 times/week | 66 (12.9) | 55 (10.4) | |
| ≥4 times/week | 444 (87.1) | 473 (89.6) | |
| GSTM1/GSTT1 genotypes | 0.020 | ||
| Both Non-null | 95 (18.6) | 113 (21.4) | |
| Others | 233 (45.7) | 272 (51.5) | |
| Both null | 173 (33.9) | 139 (26.3) | |
| LINE-1 methylation in tertiles | 0.310 | ||
| 1st (82.52+) | 154 (30.2) | 177 (33.5) | |
| 2nd (81.22–<82.52) | 165 (32.3) | 176 (33.3) | |
| 3rd (<81.22) | 191 (37.5) | 175 (33.1) | |
| Mean (SD) | 81.86 (1.82) | 81.96 (1.89) | 0.382 |
Cruciferous vegetables include bokchoi, cabbage, and cauliflower.
The overall range of DNA LINE-1 methylation was 73.3% to 93.3%, with a mean of 82.1% for men and 81.5% for women in the control subjects. The difference between the two sexes was statistically significant (P= 0.004), and as a result, further analyses were stratified by genderor gender was included in the models (Table 2). Among male controls, high cruciferous vegetable intake (≥4 times/week) was associated with significantly elevated LINE-1 methylation levels (P = 0.002). Cigarette smoking was not associated with levels of LINE-1 methylation in this population (data not shown). In stratified analysis however, smokers with higher CYP1A2 phenotype scores (above median value of 5.53) had significantly reduced levels of LINE-1 methylation compared to smokers with lower CYP1A2 phenotype scores (P = 0.001 for both men and women combined with adjustment for gender). There was no statistically significant different difference in LINE-1 methylation between higher and lower CYP1A2 phenotype scores among nonsmokers at the time of urine sample collection. Men with GSTM1 null and/or GSTT1 null genotype had elevated LINE-1 methylation levels compared to men with non-null genotype of both GSTM1 and GSTT1 (P = 0.005). We did not observe a statistically significant relationship between LINE-1 methylation and age, body mass index, level of education, smoking intensity and duration, consumption of alcoholic beverage, tea and coffee, or N-acetyltransferase 2 acetylation status (data not shown).
Table 2.
The association between LINE-1 levels and potential risk factors among control subjects only, The Shanghai Bladder Cancer Case-Control Study
| Men | Women | |||
|---|---|---|---|---|
| No. | Mean (95% CI) | No. | Mean (95% CI) | |
| Total subjects | 405 | 82.09 (78.42–85.76) | 123 | 81.53 (77.79–85.26) |
| P | 0.004 | |||
| Intake of total cruciferous vegetables | ||||
| Less than 4 times/week | 46 | 81.31 (80.77–81.84) | 9 | 81.94 (80.69–83.19) |
| ≥4 times/week | 359 | 82.20 (82.00–82.39) | 114 | 81.50 (81.14–81.85) |
| P | 0.002 | 0.505 | ||
| CYP1A2 phenotypic score levels | ||||
| Among total control subjects | ||||
| Low (≤5.53) | 182 | 82.30 (82.03–82.57) | 66 | 81.46 (81.00–81.93) |
| High (>5.53) | 203 | 81.95 (81.69–82.20) | 45 | 81.80 (81.24–82.37) |
| P | 0.066 | 0.362 | ||
| Among current smokers at urine collection only | ||||
| Low (≤5.53) | 68 | 82.50 (82.09–82.91) | 3 | 82.54 (80.72–84.36) |
| High (>5.53) | 119 | 81.74 (81.43–82.05) | 4 | 79.56 (77.98–81.13) |
| P | 0.004 | 0.059 | ||
| Among nonsmokers at urine collection only | ||||
| Low (≤5.53) | 114 | 82.18 (81.81–82.55) | 63 | 81.41 (80.94–81.88) |
| High (>5.53) | 84 | 82.23 (81.81–82.66) | 41 | 82.02 (81.44–82.60) |
| P | 0.848 | 0.111 | ||
| GSTM1/GSTT1 genotypes | ||||
| Both Non-null | 89 | 81.54 (81.15–81.92) | 24 | 81.21 (80.44–81.97) |
| Others | 208 | 82.31 (82.05–82.56) | 64 | 81.49 (81.02–81.96) |
| Both null | 105 | 82.16 (81.80–82.51) | 34 | 81.84 (81.20–82.49) |
| P | 0.005 | 0.448 | ||
The mean percentage LINE-1 methylation values (±standard deviation) were comparable in cases (81.86±1.82) and controls (81.96±1.89) (Table 1; P = 0.38). Compared with the highest tertile of LINE-1, individuals in the lowest tertile of LINE-1 methylation had a statistically non-significant 28% increased risk of bladder cancer (Table 3). This inverse association between LINE-1 and bladder cancer risk was stronger in women than in men although neither association was significant. In a stratified analysis by smoking status, a statistically significant inverse association between LINE-1 methylation and bladder cancer risk was seen among nonsmokers (P for trend = 0.03). Compared with the highest tertile of LINE-1 methylation, the OR for bladder cancer was 1.91 (95% CI = 1.17–3.13) for the lowest tertile of LINE-1 methylation, and 1.34 (95% CI = 0.79–2.28) for the middle tertile of LINE-1 methylation after controlling for age, gender, family history of cancer, and ETS. There was no statistically significant association between LINE-1 methylation and bladder cancer risk among current or former smokers (Table 3). Further adjustment for CYP1A2 phenotype score (high versus low), combined genotypes of GSTM1 and GSTT1 (null of either one gene versus non-null of both genes), and intake frequency of cruciferous vegetables (≥4 versus <4 times per week) simultaneously did not materially alter the association between LINE-1 methylation and bladder cancer risk in all subjects as well as in subgroups stratified by gender or smoking status (data not shown).
Table 3.
LINE-1 levels in relation to risk of bladder cancer, The Shanghai Bladder Cancer Case-Control Study
| LINE1 in tertiles among total subjects
|
Ptrend | |||
|---|---|---|---|---|
| 1st (82.52+) | 2nd (81.22-<82.52) | 3rd (<81.22) | ||
| Total Subjects | ||||
| No. cases/no. controls | 154/177 | 165/176 | 191/175 | |
| OR (95% CI) * | 1.00 (reference) | 1.10 (0.81–1.50) | 1.28 (0.95–1.73) | 0.268 |
| By gender | ||||
| Men | ||||
| No. cases/no. controls | 132/143 | 134/143 | 134/119 | |
| OR (95% CI) * | 1.00 (reference) | 1.03 (0.73–1.42) | 1.23 (0.87–1.73) | 0.455 |
| Women | ||||
| No. cases/no. controls | 22/34 | 31/33 | 57/56 | |
| OR (95% CI) * | 1.00 (reference) | 1.57 (0.74–3.31) | 1.58 (0.82–3.04) | 0.360 |
| By smoking status at reference date | ||||
| Nonsmokers | ||||
| No. cases/no. controls | 40/75 | 50/74 | 88/87 | |
| OR (95% CI) * | 1.00 (reference) | 1.34 (0.79–2.28) | 1.91 (1.17–3.13) | 0.032 |
| Former smokers | ||||
| No. cases/no. controls | 25/33 | 32/31 | 22/25 | |
| OR (95% CI) * | 1.00 (reference) | 1.37 (0.66–2.83) | 1.10 (0.50–2.43) | 0.682 |
| Current smokers | ||||
| No. cases/no. controls | 89/69 | 83/71 | 81/63 | |
| OR (95% CI) * | 1.00 (reference) | 0.93 (0.59–1.45) | 0.99 (0.63–1.56) | 0.938 |
All odds ratios (ORs) were adjusted for age at reference date (continuous) and family history of cancer (yes/no); regression models for total subjects and by smoking status were also adjusted for gender (male/female); regression models for nonsmoking subjects at reference date were also adjusted for ETS exposure status (yes/no).
We next examined the LINE-1-bladder cancer association in high risk groups. In stratified analyses, we did not find any statistically significant associations between LINE-1 methylation and bladder cancer risk in subgroups of individuals regardless of intake frequency of cruciferous vegetables and CYP1A2 phenotype level (data not shown). There was a statistically borderline significant inverse association between LINE-1 value and bladder cancer risk among individuals with either the GSTM1 null or GSTT1 null genotype (P for trend = 0.054). This inverse association became stronger in lifelong nonsmokers. ORs (95% CIs) of bladder cancer for the middle and lowest tertile of LINE-1 methylation were 1.26 (0.69–2.29) and 2.36 (1.34–4.14) after adjustment for multiple potential confounders, respectively, compared with the highest tertile of LINE-1 methylation (P for trend = 0.006) among lifelong nonsmokers with null genotype of either GSTM1 or GSTT1 (Table 4). The interaction between LINE-1 methylation and GSTM1/GSTT1genotypes on bladder cancer risk was statistically significant among total subjects (P for interaction= 0.03) and borderline significant among lifelong nonsmokers (P for interaction= 0.07).
Table 4.
LINE1 levels in relation to risk of bladder cancer by GSTM1 and GSTT1 genotypes, The Shanghai Bladder Cancer Case-Control Study
| LINE1 in tertiles among total subjects
|
Ptrend | |||
|---|---|---|---|---|
| 1st (82.52+) | 2nd (81.22-<82.52) | 3rd (<81.22) | ||
| By GSTM1/GSTT1 genotypes | ||||
| Both Non-null genes | ||||
| No. cases/no. controls | 30/23 | 30/40 | 35/50 | |
| OR (95% CI) | 1.00 (reference) | 0.60 (0.28–1.25) | 0.51 (0.25–1.04) | 0.169 |
| One or both null genes | ||||
| No. cases/no. controls | 123/153 | 132/135 | 151/123 | |
| OR (95% CI) | 1.00 (reference) | 1.23 (0.87–1.72) | 1.58 (1.12–2.22) | 0.054 |
| One null gene only | ||||
| No. cases/no. controls | 73/99 | 70/96 | 90/77 | |
| OR (95% CI) | 1.00 (reference) | 1.02 (0.66–1.58) | 1.66 (1.07–2.57) | 0.037 |
| Both null genes | ||||
| No. cases/no. controls | 50/54 | 62/39 | 61/46 | |
| OR (95% CI) | 1.00 (reference) | 1.65 (0.94–2.90) | 1.49 (0.85–2.59) | 0.175 |
| Lifelong nonsmokers with one Or both null genes | ||||
| No. cases/no. controls | 31/60 | 38/59 | 70/58 | |
| OR (95% CI) | 1.00 (reference) | 1.26 (0.69–2.29) | 2.36 (1.34–4.14) | 0.006 |
All odds ratios (ORs) were adjusted for age at reference date (continuous) and family history of cancer (yes/no); regression models for total subjects and by smoking status were also adjusted for gender (male/female); regression models for nonsmoking subjects at reference date were also adjusted for ETS exposure status (yes/no).
Discussion
The present study demonstrated a statistically significant, inverse relationship between LINE-1 methylation and bladder cancer risk among lifelong nonsmokers in a Chinese population. This is consistent with Moore et al., although they employed a different measure of global methylation17. However, in our data there was a lack of association between DNA methylation at LINE-1 and bladder cancer risk in former or current smokers, which differs somewhat from what was previously observed in Caucasian populations17,18. GSTM1 and GSTT1 genotype modify this inverse relationship among lifelong nonsmokers (P for interaction= 0.07). The reasons for these differences in LINE-1 methylation level-bladder cancer risk associations between the Chinese and Caucasian populations are unknown, but may be due to the different lifestyle, environmental exposures, and/or the different genetic backgrounds between the two populations.
The LINE-1 DNA methylation levels observed in the present study population (mean 81.9%, ranging from 75.9% to 93.1%) were considerably higher than those in U.S. whites (mean 79.6%, ranging from 57.9% to 92.0%)18. Given the high incidence rate of bladder cancer in the latter than the former population, one could speculate that global hypomethylation, if confirmed as an underlying risk factor for bladder cancer, may contribute to the higher rates of bladder cancer in non-Hispanic whites in the U.S. than in Chinese in China26. This high level of global DNA methylation in the present study population may explain the lack of an overall association between LINE-1 methylation and bladder cancer risk given the fact that no subjects had a LINE-1 methylation value below 74.25%, a threshold level for elevated risk of bladder cancer found in a previous study conducted in New Hampshire which had 6% of controls and 14.4% of bladder cancer cases with such low level of LINE-1 methylation18. Although batch effects could play a role in these differences, the fact that samples from studies displaying these differences were processed in the same lab with identical protocol and technique makes these effects less likely to account for these differences.
The difference in global DNA methylation levels between populations also could be due to different environmental exposures, including diet. Associations of these exposures with LINE-1 methylation were examined in this study among controls only to avoid potential influence of bladder carcinogenesis on LINE-1 alterations. In the present study, a high intake of cruciferous vegetables was associated with an increased level of LINE-1 methylation. Thus, the higher LINE-1 methylation levels in Chinese in China compared to Caucasians in the U.S. could be in part due to the high consumption of cruciferous vegetables in the former than the latter population based on self-reported weekly consumtion27.
The present study also showed that smokers with elevated CYP1A2 phenotype score, a risk factor for bladder cancer in this population21, had reduced LINE-1 methylation levels, further supporting a role of DNA hypomethylation in bladder carcinogenesis. The modifying effect of GSTM1 and GSTT1 genotype on LINE-1 methylation could be due to the impact of these genes on the available pool of glutathione. Glutathione depletion negatively impacts methylation, and perhaps the deletion of GSTM1 and GSTT1 alters the methyl donor pool sufficiently to impact LINE-1 methylation levels28. Further studies are warranted to elucidate the underlying mechanism of genetic factors and one-carbon metabolites on global DNA methylation as well as on their modifying role in the global DNA methylation and bladder cancer association.
Consistent with previous studies, the present study found a significantly elevated level of LINE1 methylation in men when compared to women18,29. Although the mechanism for this difference is unknown, LINE-1 activity has been linked to the process of X chromosome inactivation, which may account for these differences30. Further studies are necessary to fully understand LINE-1 methylation gender differences.
The present study did not show a difference in LINE-1 methylation between smokers and nonsmokers. These findings were consistent with those in a similar study in the New Hampshire population18. Given that smoking can only account for approximately 50% of bladder cancer case burdens in the U.S. and other developed countries, the findings of the present study of the association between global DNA hypomethylation in LINE-1 and bladder cancer risk among lifelong nonsmokers shed some light on the biological mechanism of non-tobacco related bladder carcinogenesis. Although the etiologic agents causing bladder cancer for non-smokers remain to be ascertained, our data suggest that these agents are likely to be associated with altering the overall epigenetic state, thereby contributing to the risk of bladder cancer. Therefore, identification of factors that alter global DNA methylation would help to discover potential etiological factors for bladder cancer, especially among nonsmokers.
Strengths of this study included the population-based study design, quantitative pyrosequencing to determine LINE-1 methylation, relatively large sample size, and comprehensively collected data on environmental exposure and genetic determinants of study subjects. The chief limitation of the present study was the retrospective nature of the study design, i.e., the collection of blood samples and information on exposure from bladder cancer patients took place after their cancer diagnosis, and in some cases, therapeutic treatment for cancer. If the carcinogenesis process and/or therapeutic treatment had any direct or indirect impact on global DNA methylation through changing subject’s lifestyle or environmental exposure, we could have observed a confounded or biased association between LINE-1 methylation and bladder cancer risk.
In conclusion, the findings of the present study support DNA hypomethylation as a potential risk factor for bladder cancer, especially for lifelong nonsmokers. Consumption of cruciferous vegetables and certain genetically determined factors such as CYP1A2 and GSTs may have impact on global DNA methylation, whereby exerting their effect on bladder cancer risk.
Acknowledgments
We thank Charlotte Wilhelm and Devin Koestler for assistance with statistical methods, as well as Graham Poage for helpful discussions. We also thank the participants of the Shanghai Bladder Cancer Study. This study was funded by the NIH (R01 CA65726 3R01 and CA121147-04S1).
References
- 1.Parkin MP. The global burden of urinary bladder cancer. Scan J Urol Nephrol. 2008;42:12–20. doi: 10.1080/03008880802285032. [DOI] [PubMed] [Google Scholar]
- 2.Boffetta P. Tobacco smoking and risk of bladder cancer. Scand J Urol Nephrol Suppl. 2008;218:45–54. doi: 10.1080/03008880802283664. [DOI] [PubMed] [Google Scholar]
- 3.Yuan J, Chan KK, Coetzee GA, Castelao JE, Watson MA, Bell DA, Wang R, Yu MC. Genetic determinants in the metabolism of bladder carcinogens in relation to risk of bladder cancer. Carcinogenesis. 2008;29:1386–93. doi: 10.1093/carcin/bgn136. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hung RJ, Boffetta P, Brennan P, Malaveille C, Hautefeuille A, Donato F, Gelatti U, Spaliviero D, Placidi D, Carta A, Scotto di Carlo A, Porru S. GST, NAT, SULT1A1, CYP1B1 genetic polymorphisms, interactions with environmental exposures and bladder cancer risk in a high-risk population. Int J Cancer. 2004;110:598–604. doi: 10.1002/ijc.20157. [DOI] [PubMed] [Google Scholar]
- 5.Pavanello S, Mastrangelo G, Placidi D, Campagna M, Puilliero A, Carta A. CYP1A2 polymorphisms, occupational and environmental exposures and risk of bladder cancer. Eur J Epidemiol. 2010;25:491–500. doi: 10.1007/s10654-010-9479-8. [DOI] [PubMed] [Google Scholar]
- 6.Bartsch H, Nair U, Risch A, Rojas M, Wikman H, Alexadrov K. Genetic Polymorphism of CYP Genes, Alone or in Combination, as a RiskModifier of Tobacco-related Cancers. Cancer Epidem Biomar. 2000;9:3–28. [PubMed] [Google Scholar]
- 7.Volanis D, Kadiyska T, Galanis A, Delakas D, Logotheti S, Zoumpourlis V. Environmental factors and genetic susceptibility promote urinary bladder cancer. Toxicol Lett. 2010;193:131–7. doi: 10.1016/j.toxlet.2009.12.018. [DOI] [PubMed] [Google Scholar]
- 8.Davis CD, Uthus EO. DNA methylation, cancer susceptibility, and nutrient interactions. Soc for Exp Biol and Med. 2004;229(10):988–95. doi: 10.1177/153537020422901002. [DOI] [PubMed] [Google Scholar]
- 9.Jones PA, Baylin SB. The fundamental role of epigenetic events in cancer. Nat Rev Genet. 2002;3:415–28. doi: 10.1038/nrg816. [DOI] [PubMed] [Google Scholar]
- 10.Baylin SB, Ohm JE. Epigenetic gene silencing in cancer- a mechanism for early oncogenic pathway addiction. Nat Rev Cancer. 2006;6:107–16. doi: 10.1038/nrc1799. [DOI] [PubMed] [Google Scholar]
- 11.Ulrich CM, Reed MC, Nijhout HF. Modeling folate, one-carbon metabolism, and DNA methylation. Nutr Rev. 2008;66:S27–S30. doi: 10.1111/j.1753-4887.2008.00062.x. [DOI] [PubMed] [Google Scholar]
- 12.Stover PJ. One-Carbon Metabolism-Genome Interactions in Folate-Associated Pathologies. J Nutr. 2009;139:2402–5. doi: 10.3945/jn.109.113670. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wilson AS, Power BE, Molloy PL. DNA hypomethylation and human diseases. Biochim Biophys Acta. 2007;1775:138–62. doi: 10.1016/j.bbcan.2006.08.007. [DOI] [PubMed] [Google Scholar]
- 14.Feinberg AP, Vogelstein B. Hypomethylation distinguishes genes of some human cancers from their normal counterparts. Nature. 1983;301:89–92. doi: 10.1038/301089a0. [DOI] [PubMed] [Google Scholar]
- 15.Issa J. Aging, DNA methylation and cancer. Crit Rev Oncol. 1999;32:31–43. doi: 10.1016/s1040-8428(99)00019-0. [DOI] [PubMed] [Google Scholar]
- 16.Hsiung DT, Marsit CJ, Houseman EA, Eddy K, Furniss CS, McClean MD, Kelsey KT. Global DNA Methylation level in whole blood as a biomarker in head and neck squamous cell carcinoma. Cancer Epidem Biomar. 2007;16(1):108–14. doi: 10.1158/1055-9965.EPI-06-0636. [DOI] [PubMed] [Google Scholar]
- 17.Moore LE, Pfeiffer RM, Poscablo C, Real FX, Kogevinas M, Silverman D, García-Closas R, Chanock S, Tardón A, Serra C, Carrato A, Dosemeci M, et al. Genomic DNA hypomethylation as a biomarker for bladder cancer susceptibility in the Spanish Bladder Cancer Study: a case-control study. Lancet Oncol. 2008;9(4):359–66. doi: 10.1016/S1470-2045(08)70038-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Wilhelm CS, Kelsey KT, Butler R, Plaza S, Gagne L, Zens MS, Andrew AS, Morris S, Nelson HH, Schned AR, Karagas MR, Marsit CJ. Implications of LINE1 methylation for bladder cancer risk in women. Clin Cancer Res. 2010;16(5):1682–9. doi: 10.1158/1078-0432.CCR-09-2983. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Choi SH, Worswick S, Byun HM, Shear T, Soussa JC, Wolff EM, Douer D, Garcia-Manero G, Liag G, Yang AS. Changes in DNA methylation of tandem DNA repeats are different from interspersed repeats in cancer. Int J Cancer. 2009;125:723–9. doi: 10.1002/ijc.24384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Chalitchagorn K, Shuangshoti S, Hourpai N, Kongruttanachok N, Tangkijvanich P, Thong-ngam D, Voravud N, Sriuranpong V, Mutirangura A. Distinctive pattern of LINE-1 methylation level in normal tissues and the association with carcinogenesis. Oncogene. 2004;23:8841–6. doi: 10.1038/sj.onc.1208137. [DOI] [PubMed] [Google Scholar]
- 21.Tao L, Xiang YB, Wang R, Nelson HH, Gao YT, Chan KK, Yu MC, Yuan JM. Cytochrome P4501A2 phenotype and tobacco-related bladder cancer risk: The Shanghai Bladder Cancer Study. Cancer Epidem Biomar. 2010;19:3087–95. doi: 10.1158/1055-9965.EPI-10-0823. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Tang BK, Kadar D, Kalow W. An alternative test for acetylator phenotyping with caffeine. Clin Pharmacol Ther. 1987;42:509–13. doi: 10.1038/clpt.1987.189. [DOI] [PubMed] [Google Scholar]
- 23.Grant DM, Tang BK, Kalow W. Polymorphic N-acetylation of a caffeine metabolite. Clin Pharmacol Ther. 1983;33:355–9. doi: 10.1038/clpt.1983.45. [DOI] [PubMed] [Google Scholar]
- 24.England RPM. Pyro Q-CpG2: quantitative analysis of methylation in multiple CpG sites by Pyrosequencing. Nat Methods. 2005;2:1–2. [Google Scholar]
- 25.Bollati V, Baccarelli A, Hou L, Bonzini M, Fustinoni S, Cavallo D, Hyang-Min B, Jiang J, Marinelli B, Pesatori AC, Bertazzi PA, Yang AS. Changes in DNA methylation patterns in subjects exposed to low-dose benzene. Cancer Res. 2007;67:876–80. doi: 10.1158/0008-5472.CAN-06-2995. [DOI] [PubMed] [Google Scholar]
- 26.Curado MP, Edwards B, Shin HR, Storm H, Ferlay J, Heanue M, Boyle P. Cancer Incidence in Five Continents. IX. Lyon: International Agency for Research on Cancer; 2007. [Google Scholar]
- 27.Davis FG, Fischer ME, VanHorn L, Mermelstein RM, Sylvester JL. Self-reported dietary changes with respect to American Cancer Society nutrition guidelines (1982–1986) Nutr Cancer. 1993;20:241–9. doi: 10.1080/01635589309514292. [DOI] [PubMed] [Google Scholar]
- 28.Lertratanangkoon K, Orkiszewski RS, Scimec JM. Methyl-Donor Deficiency Due to Chemically Induced Glutathione Depleteion. Cancer Res. 1996;56:995–1005. [PubMed] [Google Scholar]
- 29.Zhu ZZ, Hou L, Bollati V, Tarantini L, Marinelli B, Cantone L, Yang AS, Vokonas P, Lissowska J, Fustinoni S, Pesatori AC, Bonzini M, et al. Predictors of global methylation levels in blood DNA of healthy subjects: a combined analysis. Int J Epidemiol. 2010:1–14. doi: 10.1093/ije/dyq154. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Chow JC, Claudo C, Fazzari MJ, Mise N, Servant N, Glass JL. LINE-1 Activity in Facultative Heterochromatin Formation during X Chromosome Inactivation. Cell. 2010;141:956–69. doi: 10.1016/j.cell.2010.04.042. [DOI] [PubMed] [Google Scholar]