Abstract
Rationale: Exposure to particulate air pollution has been related to increased hospitalization and death, particularly from cardiovascular disease. Lower blood DNA methylation content is found in processes related to cardiovascular outcomes, such as oxidative stress, aging, and atherosclerosis.
Objectives: We evaluated whether particulate pollution modifies DNA methylation in heavily methylated sequences with high representation throughout the human genome.
Methods: We measured DNA methylation of long interspersed nucleotide element (LINE)-1 and Alu repetitive elements by quantitative polymerase chain reaction–pyrosequencing of 1,097 blood samples from 718 elderly participants in the Boston area Normative Aging Study. We used covariate-adjusted mixed models to account for within-subject correlation in repeated measures. We estimated the effects on DNA methylation of ambient particulate pollutants (black carbon, particulate matter with aerodynamic diameter ≤ 2.5 μm [PM2.5], or sulfate) in multiple time windows (4 h to 7 d) before the examination. We estimated standardized regression coefficients (β) expressing the fraction of a standard deviation change in DNA methylation associated with a standard deviation increase in exposure.
Measurements and Main Results: Repetitive element DNA methylation varied in association with time-related variables, such as day of the week and season. LINE-1 methylation decreased after recent exposure to higher black carbon (β = −0.11; 95% confidence interval [CI], −0.18 to −0.04; P = 0.002) and PM2.5 (β = −0.13; 95% CI, −0.19 to −0.06; P < 0.001 for the 7-d moving average). In two-pollutant models, only black carbon, a tracer of traffic particles, was significantly associated with LINE-1 methylation (β = −0.09; 95% CI, −0.17 to −0.01; P = 0.03). No association was found with Alu methylation (P > 0.12).
Conclusions: We found decreased repeated-element methylation after exposure to traffic particles. Whether decreased methylation mediates exposure-related health effects remains to be determined.
Keywords: epigenetic processes, air pollution, inhalation exposure, interspersed repetitive sequences
AT A GLANCE COMMENTARY
Scientific Knowledge on the Subject
Changes in epigenetic markers, such as DNA methylation, may mediate environmental effects on human health. Short-term exposure to traffic pollution has been linked with cardiovascular diseases, which are characterized by a loss of blood DNA methylation.
What This Study Adds to the Field
Our data show that blood leukocyte DNA methylation in sequences with widespread genomic representation decreases rapidly following peaks of higher ambient levels of traffic particles. These changes may contribute to produce environmental effects on human health.
DNA methylation is a mechanism of epigenetic regulation that, in mammals, involves the addition of methyl groups to cytosine to form 5-methylcytosine. About 55% of the human genome consists of repetitive elements, including approximately 500,000 long interspersed nucleotide elements (LINE-1, also indicated as L1) and 1.4 million Alu repetitive elements, which are heavily methylated (1). Demethylation of LINE-1 and Alu elements increase their activity as retrotransposable sequences, which may induce genomic alterations by insertion and or homologous recombination (2), and deregulate gene transcription (3). Repetitive elements are activated during conditions of cellular stress, and LINE-1 expression has been recently identified as a mediator of ischemic heart damage (4). Because of their high representation throughout the genome, LINE-1 and Alu methylation has been shown to correlate with global genomic DNA methylation content (1, 5). Although changes in DNA methylation have been proposed to mediate the effects of environmental factors on human health (6), relations with environmental exposures and the time-related patterns of such variations are largely unexplored.
Epidemiologic time-series studies have identified an association between daily changes in concentration of ambient air pollution and daily number of deaths and hospitalizations, particularly from cardiovascular disease and after relatively short timelags after exposure peaks (7, 8). It has been suggested that traffic-derived particles are the most toxic component (9), and the ambient level of black carbon particles, used as a tracer for traffic pollution, has been consistently associated with a variety of adverse health outcomes (10–14). Processes related to cardiovascular disease, such as oxidative stress (15), atherosclerosis (16), and aging (17), have been associated with lower DNA methylation content in blood DNA. Initial observations from in vitro and animal models have shown that air particles, particularly those from traffic emissions such as black carbon or diesel exhaust (18), or metal components of air particles such as arsenic and cadmium (19) can induce changes in gene-specific or global DNA methylation. In human studies, particulate air pollution exposure, including exposure to black carbon and particulate matter ≤2.5 μm in aerodynamic diameter (PM2.5), has been related to hyperhomocysteinemia (14, 20), a marker of decreased availability of methyl-group donors that is associated with decreased global DNA methylation content (16). In addition, subjects with lower methyl-nutrient intakes or defective polymorphisms in the methyl-nutrient pathway have been recently found to be more susceptible to the negative effects of particulate pollution on heart rate variability (21). Whether exposure to particulate air pollution induces changes in DNA methylation in human subjects has not been investigated.
In the present study, we determined the effect of ambient particulate matter levels on blood DNA methylation in LINE-1 and Alu repetitive elements and the time-related patterns of DNA methylation variations after the exposure. Some of the results of this study were presented at the American Thoracic Society International Conference, Toronto 2008 (Publication page A266) and at the International Society for Environmental Epidemiology 2008 Annual Conference (Abstract #837).
METHODS
Study Subjects
The study included 718 elderly individuals (mean age, 73.3 yr; range, 55–100 yr) evaluated between January 1999 and June 2007 as part of the Normative Aging Study, a longitudinal investigation of aging established in 1963 by the US Veterans Administration. The participants of the Normative Aging Study are all male subjects who have reported for medical examination every 3 to 5 years. A blood sample was collected at each visit. During the study period, 368 subjects reported for examination only one time, 321 reported two times, and 29 reported three times, for a total of 1,097 samples collected. This study was approved by the Institutional Review Boards of all participating institutions. All participants gave written informed consent.
Air Pollution Measurement
Ambient PM2.5, black carbon, and sulfate were measured at a stationary monitoring site 1 km from the examination site. Sulfate concentrations were not measured between October 1, 2004 and December 31, 2006. Therefore, the total number of observations with available sulfate data was 671 (61%). We evaluated different time windows of exposure, using as exposure index the average pollutant concentrations taken from time periods of 4 hours to 7 days (moving averages) before the time of blood draw because these exposure windows have shown the strongest association with air pollution health effects in previous studies (22–24). Distributions of the average ambient pollutant levels are reported in Table 1.
TABLE 1.
DISTRIBUTION OF THE AVERAGE LEVEL OF AMBIENT PARTICULATE POLLUTANTS WITHIN 4 HOURS TO 7 DAYS BEFORE THE EXAM
| Percentile
|
|||||||
|---|---|---|---|---|---|---|---|
| Moving Average | Mean | SD | 10th | 25th | 50th | 75th | 90th |
| Black carbon | |||||||
| 4-h mean | 1.35 | 0.90 | 0.47 | 0.68 | 1.10 | 1.70 | 2.61 |
| 1-d mean | 0.89 | 0.46 | 0.39 | 0.53 | 0.79 | 1.17 | 1.56 |
| 2-d mean | 0.85 | 0.39 | 0.40 | 0.55 | 0.77 | 1.12 | 1.41 |
| 3-d mean | 0.79 | 0.33 | 0.40 | 0.52 | 0.73 | 1.01 | 1.27 |
| 4-d mean | 0.76 | 0.29 | 0.41 | 0.54 | 0.72 | 0.94 | 1.15 |
| 5-d mean | 0.76 | 0.26 | 0.44 | 0.56 | 0.73 | 0.94 | 1.10 |
| 6-d mean | 0.78 | 0.25 | 0.48 | 0.60 | 0.76 | 0.95 | 1.10 |
| 7-d mean | 0.79 | 0.24 | 0.48 | 0.62 | 0.78 | 0.98 | 1.12 |
| PM2.5 | |||||||
| 4-h mean | 12.2 | 7.7 | 5.1 | 6.7 | 10.3 | 15.3 | 22.8 |
| 1-d mean | 10.9 | 6.3 | 5.0 | 6.5 | 9.3 | 13.6 | 19.5 |
| 2-d mean | 10.6 | 5.2 | 5.3 | 6.9 | 9.5 | 13.1 | 17.2 |
| 3-d mean | 10.4 | 4.8 | 5.7 | 7.0 | 9.3 | 12.8 | 16.3 |
| 4-d mean | 10.3 | 4.3 | 5.99 | 7.3 | 9.5 | 12.3 | 15.5 |
| 5-d mean | 10.3 | 3.9 | 6.3 | 7.7 | 9.7 | 12.1 | 14.9 |
| 6-d mean | 10.3 | 3.5 | 6.7 | 7.8 | 9.8 | 12.2 | 14.5 |
| 7-d mean | 10.3 | 3.3 | 6.9 | 8.0 | 9.8 | 12.0 | 14.3 |
| Sulfate* | |||||||
| 4-h mean† | NA | NA | NA | NA | NA | NA | NA |
| 1-d mean | 3.24 | 2.79 | 0.88 | 1.50 | 2.53 | 4.02 | 5.67 |
| 2-d mean | 3.12 | 2.21 | 1.03 | 1.67 | 2.67 | 3.85 | 5.75 |
| 3-d mean | 3.01 | 1.84 | 1.22 | 1.86 | 2.56 | 3.62 | 5.23 |
| 4-d mean | 2.95 | 1.58 | 1.38 | 1.97 | 2.55 | 3.62 | 4.86 |
| 5-d mean | 2.93 | 1.40 | 1.47 | 1.98 | 2.62 | 3.60 | 4.75 |
| 6-d mean | 2.98 | 1.30 | 1.64 | 2.12 | 2.73 | 3.58 | 4.72 |
| 7-d mean | 2.98 | 1.24 | 1.70 | 2.19 | 2.77 | 3.51 | 4.57 |
Definition of abbreviation: PM2.5=particulate matter ≤2.5 μm in aerodynamic diameter.
Sulfate ambient measurements were not available between October 1, 2004 and December 31, 2006. The total number of observations with available sulfate data was 671 (61%)
Sulfate data were taken as 24-h integrated measurements before October 1, 2004. Consequently, 4-h moving averages were not computed.
Laboratory Methods
Methods for DNA methylation analysis have been previously described (25) and are reported in detail in the online supplement. Briefly, we performed DNA methylation analysis of Alu and LINE-1 repetitive elements on bisulfite-treated blood leukocyte DNA using highly quantitative polymerase chain reaction (PCR)–pyrosequencing technology. The degree of methylation was expressed for each DNA locus as the percentage of methylated cytosines over the sum of methylated and unmethylated cytosines. Each marker was tested in three replicates, and their average was used in the statistical analysis.
Statistical Analysis
Because our study included repeated measures of DNA methylation for many participants, the data are correlated. To deal with this, we fitted mixed effects models (PROC MIXED in SAS V9.0), assuming:
 |
where Yit is the DNA methylation level in either LINE-1 or Alu in subject i at time t, b0 is the overall intercept, ui is the separate random intercept for subject i, and X1it—Xpit are the p covariates for subject i at each time t. We evaluated the association between particle levels and DNA methylation in multivariable models adjusted for age, body mass index, cigarette smoking (never, former, current), pack-years, statin use, fasting blood glucose, diabetes mellitus, percent lymphocytes and neutrophils in differential blood count, day of the week, season, and outdoor temperature. We reported standardized regression coefficients (β) expressing the fraction of a standard deviation change in LINE-1 or Alu methylation associated with a standard deviation increase in particle level. All tests were two tailed. A P value of < 0.05 was considered statistically significant. All analyses were performed in SAS 9.1 (SAS Institute Inc., Cary, NC).
RESULTS
The general characteristics of the study participants are shown in Table 2. Mean blood DNA methylation levels, expressed as the percentage of 5-methylcytosine (%5mC), were 77.2 (SD = 2.2) for LINE-1 and 26.2 (SD = 1.2) for Alu.
TABLE 2.
CHARACTERISTICS OF THE STUDY POPULATION IN THE NORMATIVE AGING STUDY
| Variable | All visits (n = 1,097) | Visit #1 (n = 718) | Visit #2 (n = 350) | Visit #3 (n = 29) |
|---|---|---|---|---|
| Age, yr (SD) | 73.3 (6.7) | 72.3 (6.8) | 75.0 (6.1) | 75.9 (5.2) |
| Body mass index, kg/m2 (SD) | 28.2 (4.1) | 28.2 (4.0) | 28.0 (4.1) | 29.3 (4.7) |
| Smoking status, n (%) | ||||
| Never smoker | 338 (31) | 214 (30) | 115 (33) | 9 (31) |
| Former smoker | 714 (65) | 476 (66) | 219 (62) | 19 (65) |
| Current smoker | 45 (4) | 28 (4) | 16 (5) | 1 (4) |
| Cumulative smoking,* pack-years (SD) | 30.8 (26.7) | 31.4 (27.7) | 29.5 (24.3) | 31.6 (30.4) |
| Fasting blood glucose, mg/dL (SD) | 108.5 (26.6) | 109.5 (28.8) | 106.4 (21.6) | 109.8 (22.4) |
| Diabetes, n (%) | 215 (20) | 138 (19) | 71 (20) | 6 (21) |
| Treatment with statins, n (%) | 434 (40) | 246 (34) | 176 (50) | 12 (41) |
| Blood count | ||||
| White blood cells, cells/mm3 (SD) | 6,638 (2,955) | 6,605 (3,322) | 6,670 (2,002) | 7,093 (2,947) |
| Neutrophils, % (SD) | 62.2 (8.7) | 62.1 (8.6) | 62.5 (9.0) | 61.6 (8.0) |
| Lymphocytes, % (SD) | 25.5 (8.0) | 25.6 (7.9) | 25.2 (8.3) | 26.4 (7.7) |
| Monocytes, % (SD) | 8.6 (2.1) | 8.7 (2.1) | 8.5 (2.2) | 8.6 (2.4) |
| Basophils, % (SD) | 0.6 (0.5) | 0.6 (0.5) | 0.6 (0.5) | 0.8 (0.5) |
| Eosinophils, % (SD) | 3.3 (2.1) | 3.3 (2.1) | 3.4 (2.2) | 3.1 (2.1) |
| Day of the visit, n (%) | ||||
| Monday | 69 (6) | 69 (10) | 0 (0) | 0 (0) |
| Tuesday | 244 (22) | 152 (21) | 86 (25) | 0 (0) |
| Wednesday | 580 (53) | 321 (45) | 236 (67) | 6 (21) |
| Thursday | 204 (19) | 176 (24) | 28 (8) | 23 (79) |
| Season of the visit, n (%) | ||||
| Spring | 286 (26) | 180 (25) | 93 (27) | 13 (45) |
| Summer | 300 (27) | 203 (28) | 94 (27) | 3 (10) |
| Fall | 319 (29) | 210 (29) | 109 (31) | 0 (0) |
| Winter | 192 (18) | 125 (17) | 54 (15) | 13 (45) |
| Outdoor temperature, °C (SD) | 12.8 (9.2) | 13.1 (8.8) | 12.6 (9.8) | 5.8 (10.6) |
| DNA methylation, %5mC† (SD) | ||||
| LINE-1 | 77.2 (2.2) | 76.9 (1.9) | 77.5 (2.1) | 82.7 (1.8) |
| Alu | 26.2 (1.2) | 26.3 (1.1) | 26.2 (1.2) | 24.6 (1.2) |
Definition of abbreviation: LINE-1=long interspersed nucleotide elements.
Mean and SD among ever-smokers.
Percentage of 5-methylcytosine.
We first explored whether DNA methylation varied in association with time-related variables that might have been operating as confounders in the analysis of air pollution effects (Table 3). In unadjusted regression models, LINE-1 methylation was associated with day of the week (on exam days) and season. LINE-1 methylation was lowest in blood DNA samples taken on Mondays (Reference) and showed the highest levels on Wednesdays (β = 0.35; 95% confidence interval [CI], 0.10–0.59; P = 0.006). Exam days in the Normative Aging Study are Monday through Thursday; therefore, methylation data were not available for the other days of the week. LINE-1 methylation levels were highest in spring (β = 0.32; 95% CI, 0.16–0.49; P < 0.0001) and lowest in fall (Reference). Alu methylation showed a seasonal pattern opposite to that observed for LINE-1, with the lowest levels in spring (β = −0.26; 95% CI, −0.41 to −0.10; P = 0.002) and highest in fall (Reference), and had no significant association with day of the week. Outdoor temperature showed weak, nonsignificant associations with LINE-1 and Alu methylation (Table 3).
TABLE 3.
ASSOCIATION OF LINE-1 AND Alu METHYLATION WITH DAY OF THE WEEK, SEASON, AND OUTDOOR TEMPERATURE
| Effect on LINE-1 Methylation*
|
Effect on Alu Methylation*
|
|||
|---|---|---|---|---|
| Variable | Beta (95% CI) | P Value | Beta (95% CI) | P Value |
| Day of the week | ||||
| Monday | Reference | Reference | ||
| Tuesday | 0.04 (−0.22 to 0.30) | 0.77 | 0.04 (−0.22 to 0.30) | 0.77 |
| Wednesday | 0.35 (0.10 to 0.59) | 0.006 | −0.17 (−0.41 to 0.07) | 0.17 |
| Thursday | 0.23 (−0.04 to 0.50) | 0.10 | 0.00 (−0.27 to 0.26) | 0.97 |
| Season | ||||
| Fall | Reference | Reference | ||
| Winter | 0.16 (−0.03 to 0.34) | 0.093 | −0.05 (−0.23 to 0.13) | 0.56 |
| Spring | 0.32 (0.16 to 0.49) | <0.0001 | −0.26 (−0.41 to −0.10) | 0.002 |
| Summer | 0.08 (−0.07 to 0.24) | 0.29 | −0.05 (−0.21 to 0.10) | 0.49 |
| Temperature | −0.04 (−0.10 to 0.02) | 0.21 | 0.05 (−0.01 to 0.11) | 0.09 |
Definition of abbreviations: CI = confidence interval; LINE-1=long interspersed nucleotide elements.
Unadjusted standardized regression coefficients and 95% CIs expressing the fraction of a standard deviation change in LINE-1 or Alu methylation associated with a specific category of day of the week and season or a standard deviation increase in temperature.
LINE-1 methylation decreased in relation to higher black carbon and PM2.5 ambient level (Table 4), with significant associations for all the time windows evaluated and stronger effects at the longer time windows (2–7 d). Sulfate levels showed negative but not significant associations with LINE-1 methylation. Alu methylation showed no significant association with pollutant levels at any of the time windows evaluated (Table 4). To further investigate the effects of black carbon and PM2.5 on LINE-1 methylation, we ran models including the 4-hour and 7-day moving averages. In these models (Table 5), only the 7-day moving averages showed an independent effect on LINE-1 methylation.
TABLE 4.
EFFECTS OF AMBIENT LEVELS OF BLACK CARBON, PM2.5, AND SULFATE ON LINE-1 AND Alu METHYLATION
| Moving Average | Effects on LINE-1 Methylation
|
Effect on Alu Methylation*
|
||
|---|---|---|---|---|
| Coeff* (95% CI) | P Value | Coeff* (95% CI) | P Value | |
| Black carbon | ||||
| 4-h mean | −0.07 (−0.13 to −0.01) | 0.03 | −0.02 (−0.08 to 0.04) | 0.50 |
| 1-d mean | −0.09 (−0.15 to −0.02) | 0.007 | −0.02 (−0.08 to 0.05) | 0.58 |
| 2-d mean | −0.10 (−0.16 to −0.03) | 0.004 | −0.02 (−0.09 to 0.05) | 0.56 |
| 3-d mean | −0.10 (−0.16 to −0.03) | 0.005 | −0.02 (−0.09 to 0.05) | 0.62 |
| 4-d mean | −0.09 (−0.15 to −0.02) | 0.01 | −0.01 (−0.08 to 0.06) | 0.80 |
| 5-d mean | −0.09 (−0.15 to −0.02) | 0.0089 | 0.01 (−0.05 to 0.08) | 0.73 |
| 6-d mean | −0.10 (−0.17 to −0.04) | 0.002 | 0.01 (−0.06 to 0.07) | 0.87 |
| 7-d mean | −0.11 (−0.18 to −0.04) | 0.002 | 0.01 (−0.06 to 0.08) | 0.75 |
| PM2.5 | ||||
| 4-h mean | −0.07 (−0.13 to −0.01) | 0.03 | 0.03 (−0.03 to 0.09) | 0.28 |
| 1-d mean | −0.09 (−0.16 to −0.02) | 0.008 | −0.01 (−0.07 to 0.05) | 0.74 |
| 2-d mean | −0.10 (−0.17 to −0.03) | 0.003 | −0.01 (−0.07 to 0.05) | 0.82 |
| 3-d mean | −0.10 (−0.17 to −0.04) | 0.003 | −0.01 (−0.07 to 0.05) | 0.78 |
| 4-d mean | −0.10 (−0.16 to −0.03) | 0.004 | −0.01 (−0.07 to 0.05) | 0.75 |
| 5-d mean | −0.10 (−0.16 to −0.03) | 0.004 | −0.01 (−0.07 to 0.05) | 0.84 |
| 6-d mean | −0.11 (−0.17 to −0.04) | 0.001 | −0.01 (−0.07 to 0.05) | 0.74 |
| 7-d mean | −0.13 (−0.19 to −0.06) | <0.001 | −0.01 (−0.07 to 0.05) | 0.71 |
| Sulfate† | ||||
| 4-h mean‡ | NA | NA | NA | NA |
| 1-d mean | −0.08 (−0.16 to 0.01) | 0.07 | 0.02 (−0.06 to 0.10) | 0.66 |
| 2-d mean | −0.07 (−0.15 to 0.02) | 0.12 | 0.07 (−0.02 to 0.15) | 0.13 |
| 3-d mean | −0.07 (−0.15 to 0.02) | 0.11 | 0.06 (−0.02 to 0.15) | 0.16 |
| 4-d mean | −0.06 (−0.15 to 0.03) | 0.19 | 0.07 (−0.02 to 0.15) | 0.14 |
| 5-d mean | −0.06 (−0.15 to 0.02) | 0.14 | 0.07 (−0.02 to 0.15) | 0.12 |
| 6-d mean | −0.04 (−0.12 to 0.04) | 0.33 | 0.06 (−0.02 to 0.14) | 0.16 |
| 7-d mean | −0.04 (−0.12 to 0.04) | 0.36 | 0.05 (−0.03 to 0.13) | 0.25 |
Definition of abbreviations: CI = confidence interval; LINE-1= long interspersed nucleotide elements; PM2.5=particulate matter with aerodynamic diameter ≤2.5 μm.
Standardized correlation coefficients expressing the fraction of a standard deviation change in DNA methylation associated with a standard deviation change in pollutant level adjusted for age, body mass index, cigarette smoking (never, former, current), pack-years, statin use, fasting blood glucose, diabetes mellitus, percent lymphocytes, and neutrophils in differential blood count, day of the week, season, and outdoor temperature.
Sulfate ambient measurements were not available between October 1, 2004 and December 31, 2006. The total number of observations with available sulfate data was 671 (61.1%)
Sulfate data were taken as 24-h integrated measurements before October 1, 2004. Consequently, 4-h moving averages were were not computed.
TABLE 5.
EFFECTS OF BLACK CARBON AND PM2.5 ON LINE-1 METHYLATION IN MODELS INCLUDING SHORT-TERM (4-h) AND MID-TERM (7-d) MOVING AVERAGES OF AMBIENT POLLUTANT LEVEL
| Effects on LINE-1 Methylation
|
||
|---|---|---|
| Coeff* (95% CI) | P Value | |
| Black carbon (4-h + 7-d model) | ||
| 4-h mean | −0.05 (−0.11 to 0.02) | 0.16 |
| 7-d mean | −0.09 (−0.16 to −0.02) | 0.008 |
| PM2.5 (4-h + 7-d model) | ||
| 4-h mean | −0.04 (−0.11 to 0.03) | 0.24 |
| 7-d mean | −0.11 (−0.18 to −0.05) | 0.001 |
For definition of abbreviations, see Table 4.
Standardized correlation coefficients, expressing the fraction of a standard deviation change in DNA methylation associated with a standard deviation change in pollutant level, adjusted for age, body mass index, cigarette smoking (never, former, current), pack-years, statin use, fasting blood glucose, diabetes mellitus, percent lymphocytes and neutrophils in differential blood count, day of the week, season, and outdoor temperature.
Using the 7-day moving averages, we fit two-pollutant models that showed an independent significant effect of black carbon on LINE-1 methylation (β = −0.09; 95% CI, −0.17 to −0.01; P = 0.03), whereas the association between PM2.5 and LINE-1 methylation was not statistically significant (β = −0.03; 95% CI, −0.11 to 0.04; P = 0.43).
Because these results showed that air particle levels are time-varying determinants of LINE-1 methylation, we fit multivariable models to adjust for particulate levels the association of day of the week and season with LINE-1 methylation. In models adjusting for the 7-day moving average of black carbon level, day of the week (P ≥ 0.11) and season (P ≥ 0.18) were not significantly associated with LINE-1 methylation.
As a sensitivity analysis, we tested for nonlinear effects of ambient pollutants on LINE-1 methylation in a set of models in which air particle levels were fitted using nonlinear terms (i.e, penalized splines with degrees of freedom varying from 1 to 10). Our results showed that none of the models with penalized splines had a better fit for pollutant levels compared with the linear models, suggesting that the decreases in LINE-1 methylation associated with the exposure were linear.
DISCUSSION
Prior studies have linked particulate air pollution with oxidative stress production, activation of inflammatory pathways, endothelial injury and dysfunction, arterial vasoconstriction, and alterations in blood coagulation (26–28). However, the nature of the intermediate changes linking particulate air pollution to cardiovascular disease is not fully understood (8). In the present investigation on a large population of elderly individuals, we found that blood DNA methylation in repetitive elements was decreased in individuals with recent exposure to higher levels of ambient particulate matter. In particular, LINE-1 methylation was decreased in association with ambient levels of black carbon and PM2.5 measured in the 7 days before the examination, whereas no association was observed between Alu methylation and particle levels.
Previous in vitro experiments have shown that repetitive element hypomethylation and transcription occur in response to biological processes, such as cellular stress and inflammation (4), which are also induced by particulate pollution in exposed subjects (8, 26, 27). Air particles are known to increase the production of reactive oxygen species, perhaps in a catalytic fashion via redox cycling (27). Oxidative DNA damage can interfere with the ability of methyltransferases to interact with DNA (15), thus resulting in hypomethylation of cytosine residues at CpG sites. In addition, reactive oxygen species have been recently shown to alter the expression of genes belonging to DNA methylation machinery (29). Changes in DNA methylation through generation of reactive oxygen species may be induced by components of airborne particulate matter, such as metals (19). Takiguchi and colleagues showed that cadmium interferes with DNA methylation by inhibiting DNA methyltransferases, possibly due to an interaction of cadmium with the methyltransferase DNA binding domain (30). In addition, depletion of the cellular methyl pool has been indicated as a contributor to the global hypomethylation observed in arsenic-treated cells (19). Hypomethylation of repetitive DNA sequences is expected to lead to the transcriptional activation of those repetitive sequences that still contain active promoters, potentially resulting in disruption of transcription factor balance, sense or antisense transcriptional interference, and production of transcripts complementary to endogenous transcripts or to alterations in genomic organization and stability (2, 3). The decrease in methylation we observed may be part of the systemic events consequent to alveolar inflammation and the release of inflammatory mediators resulting from particle inhalation (27). Also, inhaled nanoparticles may move out of the lung into the bloodstream (8), where they may directly affect blood cells by inducing oxidative stress or inflammatory reactions (8, 15, 16).
In our study, the association of black carbon and PM2.5 with LINE-1 methylation was significant for all the time windows we examined, which included the average of hourly concentrations measured over 4 hour to 7 days before the examination. We selected those time windows because short-term health effects of air particle exposure, including increased numbers of hospitalizations and deaths resulting from cardiovascular disease, have been found to correlate with higher ambient pollutant levels measured within 1 to 7 days before the outcome (23), and 5- to 7-day cumulative exposures have been associated with outcomes related to cardiovascular disease, such as increased serum fibrinogen (24, 31) and C-reactive protein (26). The time patterns of changes in DNA methylation are largely unknown. In patients with myeloid leukemia treated with the demethylating analog 5-Aza-2′-deoxycytidine, LINE-1 methylation has been shown to decrease starting 2 to 4 days after the beginning of the treatment (32). Our data show that environmental exposures may operate on DNA methylation over a similar time frame, with effects from particles that were significant even in association with the shorter time windows evaluated (4 hours). However, when we fit models that included the shortest (4-h) and longest (7-d) moving averages that we used in our analyses, only the 7-day effects remained significant. This finding suggests that particles may need several days to operate changes on DNA methylation. Black carbon is a tracer of particles from traffic, which are a major component of fine particles measured by PM2.5. The finding that only the black carbon effect on LINE-1 methylation was statistically significant in two-pollutant models including black carbon and PM2.5 indicates traffic particles as the main determinant of the observed changes in DNA methylation. In these models, the standardized regression coefficient for the 7-day moving average of black carbon was equal to −0.09, indicating that a standard deviation increase in black carbon resulted in an estimated decrease in LINE-1 methylation equal to 9% of its standard deviation. To put this result into perspective, this decrease in LINE-1 methylation can be compared with the results of a recent analysis in this same study that showed that LINE-1 methylation decreases progressively with age (33). The effect on LINE-1 methylation determined by a standard deviation increase in black carbon concentration would be equivalent in size to a decline in LINE-1 methylation occurring over 3.4 years. In other words, assuming that a decrease in LINE-1 methylation reflects biological age (17), the black carbon effect would be similar to increased aging of nearly 3.5 years. However, whether this LINE-1 methylation decline is associated with increased risk of age-related diseases—including the increased risk of cardiovascular disease associated with air pollution exposure (7, 8)—remains to be determined. Also, experimental studies in humans or in vitro and animal models are warranted to confirm the exposure–methylation link and to clarify the epigenetic mechanisms of the action of traffic particles. Associations of LINE-1 methylation with sulfate, an indicator of coal combustion particles, never achieved statistical significance, although missing data gave that analysis less power.
In our data, we did not observe any correlation of Alu methylation with particle levels. Although LINE-1 and Alu repetitive element methylation have been shown to correlate with genomic DNA methylation content, the two repetitive elements are controlled through different mechanisms and have been shown to have different transcription patterns in response to cellular stressors (33). In addition, although our study had the advantage of high-precision quantitative Pyrosequencing measures, the lower availability of nonmutated CpG sites in Alu elements may have resulted in lower sensitivity, compared with LINE-1 methylation measures, to detect potential changes in global DNA methylation (1).
By examining potential confounders of the exposure-methylation relation, we observed differences in LINE-1 and Alu DNA methylation by season and day of the week, which we then controlled in multivariable models testing for particle effects. Epigenetic adaptive changes to season and temperature have been reported in plants and aquatic species (34, 35). In our data, LINE-1 and Alu methylation had associations with mean outdoor temperature of the day of the exam that, although not significant, were consistent with those observed with season. A previous study of 30 healthy volunteers reported cyclic variations in blood DNA methylation content observed over a 24-hour period (36), and recent investigations have shown transient cyclical methylation and demethylation of specific gene promoters (37, 38). The association of LINE-1 methylation with day of the week observed in our data may provide further evidence for the existence of cyclical patterns of variation in repetitive element DNA methylation. However, our analyses of temporal variations in DNA methylation were performed to identify potential confounders of air pollution effects because levels of air pollutants, particularly those from traffic emissions, show strong variations in association with day of the week and season (14, 21, 23). In multivariable models adjusting for air particle levels, the associations of day of the week and season with LINE-1 methylation were not statistically significant, suggesting that the cyclical variations in LINE-1 methylation we observed are dependent, at least in part, on changes in particle levels. Our study was not designed to evaluate temporal cycles of DNA methylation and provides only descriptive, nonconclusive evidence for the existence of such patterns, which should be confirmed in studies evaluating DNA methylation in multiple repeated measures taken in different days and seasons.
Human exposure to combustion particles is predominantly from traffic particles (39). In urban areas black carbon, which we used as a tracer for particles from traffic emissions, derives primarily from exhaust emissions from diesel-powered vehicles (40). Because black carbon stems from incomplete combustion, which usually contains toxic substances, it is considered a better indicator of the health effect of air particles than the weight measurement (11). A potential limitation of this study is that we used ambient black carbon levels, as well as levels of PM2.5 and sulfate, from a single monitoring site as a surrogate for recent exposure to air particles. PM2.5 and sulfate concentrations in Boston are very spatially homogeneous (14), so it is reasonable to conclude that PM2.5 and sulfate from a single monitoring site could characterize personal exposures to them. By contrast, black carbon concentration is considerably more spatially variable, depending on traffic exposure. However, a recent study comparing ambient concentrations at this site with 14 monitoring stations in metropolitan Boston ranging from downtown to rural showed correlation coefficients varying from 0.54 to 0.94 (median, 0.80) (14). Although these results suggest that some degree of misclassification is possible, due to the nondifferential nature of the error, this would be expected to bias the results toward the null. Our results can only be generalized to an aged population that consists of older men who are almost all white. The effect on women and children and on different ethnic groups should be addressed in future studies, particularly in relation to the exposure of different population groups with varying geographical location, occupation, socioeconomic status, and behavioral characteristics. Our study was based on a quantitative analysis of blood DNA methylation using pyrosequencing, which is highly reproducible and accurate for measuring small changes in DNA methylation (1, 25), and was repeated three times on each sample to minimize the assay variability. Whether the changes we observed in blood DNA reflect similar modifications of DNA methylation in tissues directly exposed to particles, such as the airway epithelium or different target tissues, and whether the amount of changes we observed is associated with negative health effects need to be assessed in future studies.
In conclusion, our results indicate that repetitive element methylation in blood DNA, particularly in LINE-1 sequences, varies in relation to environmental and time variables, including significant decreases after short-term exposure to air particles from traffic emissions. Our data indicate the existence of rapid methylation changes in vivo that may trigger adaptive responses and produce environmental effects on human health. Whether the alteration of these molecular mechanisms predicts the effects of traffic particles on human health remains to be determined.
Supplementary Material
Supported by the National Institute of Environmental Health Sciences grants ES015172–01, ES00002, and PO1 ES009825; the Environmental Protection Agency grants EPA R83241601 and R827353; and the CARIPLO Foundation grant 2007–5469. The VA Normative Aging Study, a component of the Massachusetts Veterans Epidemiology Research and Information Center, Boston, Massachusetts, is supported by the Cooperative Studies Program/Epidemiology Research and Information Center of the U.S. Department of Veterans Affairs.
This article has an online data supplement, which is accessible from this issue's table of contents at www.atsjournals.org
Originally Published in Press as DOI: 10.1164/rccm.200807-1097OC on January 8, 2009
Conflict of Interest Statement: None of the authors has a financial relationship with a commercial entity that has an interest in the subject of this manuscript
References
- 1.Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, Issa JP. A simple method for estimating global DNA methylation using bisulfite pcr of repetitive DNA elements. Nucleic Acids Res 2004;32:e38. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ostertag EM, Kazazian HH Jr. Biology of mammalian L1 retrotransposons. Annu Rev Genet 2001;35:501–538. [DOI] [PubMed] [Google Scholar]
- 3.Han JS, Szak ST, Boeke JD. Transcriptional disruption by the L1 retrotransposon and implications for mammalian transcriptomes. Nature 2004;429:268–274. [DOI] [PubMed] [Google Scholar]
- 4.Lucchinetti E, Feng J, Silva R, Tolstonog GV, Schaub MC, Schumann GG, Zaugg M. Inhibition of LINE-1 expression in the heart decreases ischemic damage by activation of Akt/PKB signaling. Physiol Genomics 2006;25:314–324. [DOI] [PubMed] [Google Scholar]
- 5.Weisenberger DJ, Campan M, Long TI, Kim M, Woods C, Fiala E, Ehrlich M, Laird PW. Analysis of repetitive element DNA methylation by methylight. Nucleic Acids Res 2005;33:6823–6836. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Jirtle RL, Skinner MK. Environmental epigenomics and disease susceptibility. Nat Rev Genet 2007;8:253–262. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Brook RD. Cardiovascular effects of air pollution. Clin Sci (Lond) 2008;115:175–187. [DOI] [PubMed] [Google Scholar]
- 8.Brook RD, Franklin B, Cascio W, Hong Y, Howard G, Lipsett M, Luepker R, Mittleman M, Samet J, Smith SC Jr, et al. Air pollution and cardiovascular disease: a statement for healthcare professionals from the expert panel on population and prevention science of the American Heart Association. Circulation 2004;109:2655–2671. [DOI] [PubMed] [Google Scholar]
- 9.Hoffmann B, Moebus S, Mohlenkamp S, Stang A, Lehmann N, Dragano N, Schmermund A, Memmesheimer M, Mann K, Erbel R, et al. Residential exposure to traffic is associated with coronary atherosclerosis. Circulation 2007;116:489–496. [DOI] [PubMed] [Google Scholar]
- 10.Gold DR, Litonjua AA, Zanobetti A, Coull BA, Schwartz J, MacCallum G, Verrier RL, Nearing BD, Canner MJ, Suh H, et al. Air pollution and ST-segment depression in elderly subjects. Environ Health Perspect 2005;113:883–887. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Sorensen M, Daneshvar B, Hansen M, Dragsted LO, Hertel O, Knudsen L, Loft S. Personal PM2.5 exposure and markers of oxidative stress in blood. Environ Health Perspect 2003;111:161–166. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Adar SD, Gold DR, Coull BA, Schwartz J, Stone PH, Suh H. Focused exposures to airborne traffic particles and heart rate variability in the elderly. Epidemiology 2007;18:95–103. [DOI] [PubMed] [Google Scholar]
- 13.O'Neill MS, Veves A, Sarnat JA, Zanobetti A, Gold DR, Economides PA, Horton ES, Schwartz J. Air pollution and inflammation in type 2 diabetes: a mechanism for susceptibility. Occup Environ Med 2007;64:373–379. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Park SK, O'Neill MS, Vokonas PS, Sparrow D, Spiro Iii A, Tucker KL, Suh H, Hu H, Schwartz J. Traffic-related particles are associated with elevated homocysteine: the VA normative aging study. Am J Respir Crit Care Med 2008;178:283–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Valinluck V, Tsai HH, Rogstad DK, Burdzy A, Bird A, Sowers LC. Oxidative damage to methyl-CpG sequences inhibits the binding of the methyl-CpG binding domain (MBD) of methyl-CpG binding protein 2 (MeCP2). Nucleic Acids Res 2004;32:4100–4108. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Castro R, Rivera I, Struys EA, Jansen EE, Ravasco P, Camilo ME, Blom HJ, Jakobs C, Tavares de Almeida I. Increased homocysteine and S-adenosylhomocysteine concentrations and DNA hypomethylation in vascular disease. Clin Chem 2003;49:1292–1296. [DOI] [PubMed] [Google Scholar]
- 17.Fuke C, Shimabukuro M, Petronis A, Sugimoto J, Oda T, Miura K, Miyazaki T, Ogura C, Okazaki Y, Jinno Y. Age related changes in 5-methylcytosine content in human peripheral leukocytes and placentas: an HPLC-based study. Ann Hum Genet 2004;68:196–204. [DOI] [PubMed] [Google Scholar]
- 18.Belinsky SA, Snow SS, Nikula KJ, Finch GL, Tellez CS, Palmisano WA. Aberrant CpG island methylation of the p16(INK4a) and estrogen receptor genes in rat lung tumors induced by particulate carcinogens. Carcinogenesis 2002;23:335–339. [DOI] [PubMed] [Google Scholar]
- 19.Wright RO, Baccarelli A. Metals and neurotoxicology. J Nutr 2007;137:2809–2813. [DOI] [PubMed] [Google Scholar]
- 20.Baccarelli A, Zanobetti A, Martinelli I, Grillo P, Hou L, Lanzani G, Mannucci PM, Bertazzi PA, Schwartz J. Air pollution, smoking, and plasma homocysteine. Environ Health Perspect 2007;115:176–181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Baccarelli A, Cassano PA, Litonjua A, Park SK, Suh H, Sparrow D, Vokonas P, Schwartz J. Cardiac autonomic dysfunction: effects from particulate air pollution and protection by dietary methyl nutrients and metabolic polymorphisms. Circulation 2008;117:1802–1809. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Park SK, O'Neill MS, Vokonas PS, Sparrow D, Schwartz J. Effects of air pollution on heart rate variability: the VA Normative Aging Study. Environ Health Perspect 2005;113:304–309. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Samet JM, Zeger SL, Dominici F, Curriero F, Coursac I, Dockery DW, Schwartz J, Zanobetti A. Distributed lag between air pollution and daily deaths. The national morbidity, mortality, and air pollution study. Part ii: morbidity and mortality from air pollution in the United States (appendix b). Res Rep Health Eff Inst 2000;94:54–61. [PubMed] [Google Scholar]
- 24.Zeka A, Sullivan JR, Vokonas PS, Sparrow D, Schwartz J. Inflammatory markers and particulate air pollution: characterizing the pathway to disease. Int J Epidemiol 2006;35:1347–1354. [DOI] [PubMed] [Google Scholar]
- 25.Bollati V, Baccarelli A, Hou L, Bonzini M, Fustinoni S, Cavallo D, Byun HM, Jiang J, Marinelli B, Pesatori AC, et al. Changes in DNA methylation patterns in subjects exposed to low-dose benzene. Cancer Res 2007;67:876–880. [DOI] [PubMed] [Google Scholar]
- 26.Peters A, Frohlich M, Doring A, Immervoll T, Wichmann HE, Hutchinson WL, Pepys MB, Koenig W. Particulate air pollution is associated with an acute phase response in men; results from the MONICA-Augsburg study. Eur Heart J 2001;22:1198–1204. [DOI] [PubMed] [Google Scholar]
- 27.Donaldson K, Stone V, Seaton A, MacNee W. Ambient particle inhalation and the cardiovascular system: potential mechanisms. Environ Health Perspect 2001;109:523–527. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Baccarelli A, Zanobetti A, Martinelli I, Grillo P, Hou L, Giacomini S, Bonzini M, Lanzani G, Mannucci PM, Bertazzi PA, et al. Effects of exposure to air pollution on blood coagulation. J Thromb Haemost 2007;5:252–260. [DOI] [PubMed] [Google Scholar]
- 29.Fratelli M, Goodwin LO, Orom UA, Lombardi S, Tonelli R, Mengozzi M, Ghezzi P. Gene expression profiling reveals a signaling role of glutathione in redox regulation. Proc Natl Acad Sci USA 2005;102:13998–14003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Takiguchi M, Achanzar WE, Qu W, Li G, Waalkes MP. Effects of cadmium on DNA-(cytosine-5) methyltransferase activity and DNA methylation status during cadmium-induced cellular transformation. Exp Cell Res 2003;286:355–365. [DOI] [PubMed] [Google Scholar]
- 31.Ruckerl R, Ibald-Mulli A, Koenig W, Schneider A, Woelke G, Cyrys J, Heinrich J, Marder V, Frampton M, Wichmann HE, et al. Air pollution and markers of inflammation and coagulation in patients with coronary heart disease. Am J Respir Crit Care Med 2006;173:432–441. [DOI] [PubMed] [Google Scholar]
- 32.Choi SH, Byun HM, Kwan JM, Issa JP, Yang AS. Hydroxycarbamide in combination with azacitidine or decitabine is antagonistic on DNA methylation inhibition. Br J Haematol 2007;138:616–623. [DOI] [PubMed] [Google Scholar]
- 33.Bollati V, Schwartz J, Wright RO, Litonjua A, Tarantini L, Suh H, Sparrow D, Vokonas P, Baccarelli A. Decline in genomic DNA methylation through aging in a cohort of elderly subjects. Mech Ageing Dev 2009;130:234–239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Sung S, Amasino RM. Vernalization and epigenetics: how plants remember winter. Curr Opin Plant Biol 2004;7:4–10. [DOI] [PubMed] [Google Scholar]
- 35.Pinto R, Ivaldi C, Reyes M, Doyen C, Mietton F, Mongelard F, Alvarez M, Molina A, Dimitrov S, Krauskopf M, et al. Seasonal environmental changes regulate the expression of the histone variant macroH2A in an eurythermal fish. FEBS Lett 2005;579:5553–5558. [DOI] [PubMed] [Google Scholar]
- 36.Bonsch D, Hothorn T, Krieglstein C, Koch M, Nehmer C, Lenz B, Reulbach U, Kornhuber J, Bleich S. Daily variations of homocysteine concentration may influence methylation of DNA in normal healthy individuals. Chronobiol Int 2007;24:315–326. [DOI] [PubMed] [Google Scholar]
- 37.Kangaspeska S, Stride B, Metivier R, Polycarpou-Schwarz M, Ibberson D, Carmouche RP, Benes V, Gannon F, Reid G. Transient cyclical methylation of promoter DNA. Nature 2008;452:112–115. [DOI] [PubMed] [Google Scholar]
- 38.Metivier R, Gallais R, Tiffoche C, Le Peron C, Jurkowska RZ, Carmouche RP, Ibberson D, Barath P, Demay F, Reid G, et al. Cyclical DNA methylation of a transcriptionally active promoter. Nature 2008;452:45–50. [DOI] [PubMed] [Google Scholar]
- 39.Hoek G, Brunekreef B, Goldbohm S, Fischer P, van den Brandt PA. Association between mortality and indicators of traffic-related air pollution in the Netherlands: a cohort study. Lancet 2002;360:1203–1209. [DOI] [PubMed] [Google Scholar]
- 40.Janssen NA, Schwartz J, Zanobetti A, Suh HH. Air conditioning and source-specific particles as modifiers of the effect of PM10 on hospital admissions for heart and lung disease. Environ Health Perspect 2002;110:43–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.