- Open Access
Exposure to NO2, CO, and PM2.5 is linked to regional DNA methylation differences in asthma
© The Author(s). 2018
- Received: 9 August 2017
- Accepted: 12 December 2017
- Published: 5 January 2018
DNA methylation of CpG sites on genetic loci has been linked to increased risk of asthma in children exposed to elevated ambient air pollutants (AAPs). Further identification of specific CpG sites and the pollutants that are associated with methylation of these CpG sites in immune cells could impact our understanding of asthma pathophysiology. In this study, we sought to identify some CpG sites in specific genes that could be associated with asthma regulation (Foxp3 and IL10) and to identify the different AAPs for which exposure prior to the blood draw is linked to methylation levels at these sites. We recruited subjects from Fresno, California, an area known for high levels of AAPs. Blood samples and responses to questionnaires were obtained (n = 188), and in a subset of subjects (n = 33), repeat samples were collected 2 years later. Average measures of AAPs were obtained for 1, 15, 30, 90, 180, and 365 days prior to each blood draw to estimate the short-term vs. long-term effects of the AAP exposures.
Asthma was significantly associated with higher differentially methylated regions (DMRs) of the Foxp3 promoter region (p = 0.030) and the IL10 intronic region (p = 0.026). Additionally, at the 90-day time period (90 days prior to the blood draw), Foxp3 methylation was positively associated with NO2, CO, and PM2.5 exposures (p = 0.001, p = 0.001, and p = 0.012, respectively). In the subset of subjects retested 2 years later (n = 33), a positive association between AAP exposure and methylation was sustained. There was also a negative correlation between the average Foxp3 methylation of the promoter region and activated Treg levels (p = 0.039) and a positive correlation between the average IL10 methylation of region 3 of intron 4 and IL10 cytokine expression (p = 0.030).
Short-term and long-term exposures to high levels of CO, NO2, and PM2.5 were associated with alterations in differentially methylated regions of Foxp3. IL10 methylation showed a similar trend. For any given individual, these changes tend to be sustained over time. In addition, asthma was associated with higher differentially methylated regions of Foxp3 and IL10.
- Ambient air pollution
- Immune system
- Regulatory T cell
Asthma is the most frequent chronic disease in children , and its prevalence continues to increase, raising global public health concerns [2–4]. Previous studies found that high prenatal ambient air pollutant (AAP) exposure alters epigenetic programming in utero  and that children are at greater risk of developing asthma when exposed to higher concentrations of AAPs . This further raises concern and increases the urgency of understanding the relationship between AAP exposures and asthma and finding effective ways to prevent, treat, and/or cure asthma. Although the mechanisms by which exposures to AAPs increase asthma prevalence remain poorly understood, evidence suggests that AAPs can mediate T cell polarization and cytokine dysregulation via epigenetic modifications [7–10].
DNA methylation at least partially controls the expression of certain key genes that are known to be involved in immune tolerance [11–13]. We previously demonstrated that exposure in children to high levels of polycyclic aromatic hydrocarbons (PAHs), a component of AAPs, is associated with higher overall percent methylation of the transcriptional regulatory region of the forkhead box transcription factor 3 (Foxp3) locus . In a subsequent study, in addition to the presence of higher Foxp3 methylation levels, we found that exposure to higher levels of PAHs is associated with decreased expression of IL10, an anti-inflammatory cytokine. These associations were stronger in children with asthma, suggesting disease-linked effects . Other studies have also associated air pollution with other immune parameters, such as the association of exposure to NOx with decreases in IL10 cytokine expression and exposure to diesel exhaust with higher differentially methylated regions (DMRs) of Foxp3 . A recent review of ambient air pollutants (AAPs), epigenetic regulation, and asthma highlights the challenge that asthma is heterogeneous, responsiveness to traffic-related air pollution, and may be a distinct phenotype . In addition, a study by Clifford et al.  found DNA methylation changes after diesel exhaust exposure in more than 500 CpG sites in the bronchial epithelial cells. Finally, AAPs and DNA methylation have been linked to other health problems, such as cardiovascular disease .
We sought to identify CpG sites in the Foxp3 and IL10 genes that have higher DMRs in children with asthma exposed to AAPs. Our primary hypothesis was that exposure to high levels of CO, O3, NO2, and PM2.5 would be associated with higher DMRs within the Foxp3 gene and the intron 4 region of the IL10 gene. The Foxp3 gene regulates transcription of Tregs, and the Il10 gene regulates transcription of various cell types, including CD4+ T cell populations, because it has been shown to act as an enhancer element . In addition, our secondary hypotheses were that methylation would be sustained following continued exposure to these pollutants and that Treg levels would be negatively associated with Foxp3 methylation at specific CpG loci in the enhancer and promoter regions.
We recruited subjects from Fresno, California, where there is consistently high AAP. According to the 2017 American Lung Association “State of the Air” report, Fresno ranked third in the country for year-round and short-term particle pollution and O3 exposure . Fresno County also has a high childhood asthma prevalence in comparison to other counties in California (24 vs. 17%, ages 5–17) . Recruiting from the Fresno Unified School district from 2010 to 2015 resulted in a convenience sample of 198 children with complete data, 188 of which were included in the analysis after quality control, with a median age of 14.7 years. In addition, we retested 33 subjects after approximately 2 years (mean 820 days, SD 243 days; again, as a convenience sample).
At each clinical visit, information on BMI, a detailed health and general history questionnaire, and blood samples were obtained. Subjects were excluded if they had taken oral immunosuppressants within 5 days of the blood draw, had a history of allergen immunotherapy within 1 year of the clinical visit, had a chronic disease other than allergies or asthma, or had an acute infection. In addition, no blood specimen was taken during an infection, sickness, or during the use of oral steroids. Subjects with asthma were defined by the participant’s report of a physician’s diagnosis of current asthma. Secondhand smoke exposure (SHS) was defined as exposure to cigarette smoke from other household members. Participant smoking was defined as smoking more than 100 cigarettes in their lifetime.
Collection and processing of blood specimens
Human peripheral blood mononuclear cells (PBMCs) and plasma were extracted from the blood via Ficoll procedure and stored in liquid nitrogen and at − 80 °C as per published techniques .
We first selected CpG sites within the upstream promoter (− 138, − 126, − 77, − 65, − 58, and − 15) and enhancer (− 4506, − 4500, − 4494, and − 4484) regions of Foxp3 (on the X chromosome) located on the human genome (hg38) chrX at nucleotides 49264838, 49264826, 49264777, 49264765, 49264758, 49264715, and 49269206, 49,69200, 49269194, 49269184, respectively, and within the intron 4 region of IL10 (+ 2888, + 2907, + 2921, + 3261, + 3265, and + 3281 on chromosome 1), located on the human genome (hg38) chr1 at nucleotides 206769607, 206769588, 206769574, 206769234, 206769230, 206769214, respectively, as these sites have been shown in the literature to be key in modulating regulatory T cell (Treg) response and overall immune tolerance [11, 19, 23]. We then examined associations between exposures to CO, O3, NO2, and PM2.5 and methylation levels at these CpG sites. To determine whether methylation of CpG sites is maintained with long-term exposure, we repeated measurements on a subset of samples obtained at approximately 2 years after the initial blood samples were obtained. Methylation analysis using the pyrosequencing approach was performed on the following gene loci: ten CpG sites for Foxp3 and six CpG sites for IL10. Bisulfite modification and subsequent DNA amplification of specific CpG regions within the Foxp3 and IL10 genes were conducted by PCR using the different pairs of primers for each region enumerated in Additional file 1: Table S1. Please see Additional file 1 for further details.
CyTOF mass cytometry
We also examined Treg, Foxp3, and IL10 protein expression in a subset of the subjects using mass cytometry (CyTOF). CyTOF uses antibodies tagged with heavy metal ions and allows for identifying many more cell markers in a single sample than flow cytometry and without issues of spectral overlap. Here, we performed CyTOF on a randomly selected subset of subjects (n = 74; 24.3% asthmatics, 75.7% non-asthmatics) for which additional blood was available from the same blood draw as that used for methylation analysis to assess for cell composition differences. Briefly, PBMCs were thawed, stimulated for 4 h with 20 ng/mL phorbol myristate acetate + 1 μg/mL ionomycin (PMA-ION), and stained for surface markers, intracellular markers, and transcription factors using CyTOF. Please see Additional file 1 for further details.
Cell composition analysis
Nine cell subtypes were measured by gating on cell surface markers obtained from the CyTOF analysis: B Cells, monocytes, CD3+ cells, CD4+, CD8+, Th17 cells, Tregs, Th1, and Th2 cells. The cell counts were converted into non-overlapping proportions, with an additional miscellaneous category (the remaining fraction of unmeasured cell types, typically < 5%) accounted for to normalize the distribution to 100%. The resulting cell distributions were compared between the asthmatic and healthy groups and used as covariates to confirm the reproducibility of differences in methylation (see Additional file 1: Table S2).
Ambient air pollution exposure estimation
Exposure to AAPs was estimated using hourly concentration data  of four common pollutants (CO, O3, NO2, and PM2.5) measured at four air quality monitoring stations located within Fresno city limits. These data were acquired from the US Environmental Protection Agency’s Air Quality System online database (https://www.epa.gov/aqs).
In order to capture both the spatial and temporal variation in the AAP exposure of each participant, we computed an index of the average outdoor residential AAP exposure for each individual using inverse distance weighting (IDW) [25, 26] of AAP measurements from the air quality monitoring stations according to their distance from each participant’s residence. For each blood sample and air pollutant, exposure was defined as the mean of the IDW values over the given time period (1, 15, 30, 90, 180, or 365 days) before the sample was taken. See online Additional file 1 for further details of these data processing choices.
Statistical analysis was performed using the R statistical computing software (R 3.4.1). We used linear regression to adjust methylation values for sex, age, and BMI. Using multivariate linear regression, we evaluated associations of resulting residuals with asthma status and ambient air pollution.
Due to the presence of related subjects within our study, we also considered linear mixed-effects models fit using generalized least squares, as implemented in the “gls” function in the “nlme” R package (version 3.1.131) for non-linear mixed-effects models. Unless explicitly stated otherwise, the results from fixed-effects models are described in the main text, and the results from mixed-effects models are described in Additional file 1. Detailed descriptions of all analyses are available in Additional file 1, along with de-identified data, exposure data, and R code (see Additional files 2, 3, 4, and 5).
Multivariate linear regression was repeated on a random subset of the data for which CyTOF-derived cell counts were obtained. The proportions of different cell types were incorporated into the model to adjust for any differences. Additionally, we performed t tests for each cell type to look for statistically significant differences in cell distribution. Family-wise multiple testing correction was performed by using the Benjamini-Hochberg procedure  with a false discovery rate of 0.05.
Demographic characteristics of study subjects*
Number of subjects
Total number of male
Median age at enrollment (years) (quartile 1; quartile 3)
14.7 (12.9; 16.9)
14.6 (13.1; 16.9)
Median BMI (kg/m2) (quartile 1; quartile 3)
21.7 (19.1; 25.6)
22.0 (18.8; 28.0)
Total number of rhinitis
Total number of eczema
Total number of African-American
Total number of Hispanic or Latino
Total number of Caucasian
Total number of mixed race
Total number of not reported
Total number of exposed to secondhand smoke
Total number of smokers
Total number of household income < $15,000
Total number of household oncome $15,000–30,000
Total number of household income $31,000–50,000
Total number of household income > $50,000
Total number of household income not reported
Normalization of methylation data
For the rest of the analysis, we used the residuals from fitting methylation to a regression model with the following covariates: sex, age, BMI, and age-sex interaction. For much of the analysis, we used averages across multiple sites, so these residuals are divided by the estimated standard deviation of the error to obtain “normalized” methylation, giving equal weight to all sites in these averages.
Association between methylation and asthma status
We tested whether asthma was significantly associated with normalized methylation at each CpG site studied for Foxp3 (Fig. 3a, Additional file 1: Table S4) using an ANOVA model. We found that asthma was significantly associated with methylation at four of the six CpG sites in the promoter region of Foxp3 (p < 0.05). We also found that averaging over all six of these sites gave a stronger association than that for any individual site, so we used this average in the rest of the analysis. To aid in interpretation of the effect sizes in Fig. 3a, we note that the estimated standard deviation of the error for each sex at each CpG site in the Foxp3 promoter was generally 4–6%. Hence, an effect size of 0.3 for the normalized values is roughly equivalent to 1.2–1.8 for percent methylation.
Similarly, we found that asthma was significantly associated with normalized methylation at all three CpG sites in region 3 of intron 4 of IL10 (p < 0.05) (Fig. 3b, Additional file 1: Table S4). Furthermore, averaging over all three of these sites gave a stronger association than for any individual site, so we used this average in the rest of the analysis. The estimated standard deviation of the error for each sex at each CpG site in IL10 was generally 9–11%.
Because the measured PBMC methylation is the average of all constituent cell types in a sample weighted by their proportion, an observed difference in methylation between healthy and asthmatic subjects could potentially be an artifact of differences in the underlying cell distribution. Conversely, because Treg cells constitute a small percentage of the PBMC fraction, a large difference in methylation within Treg could be masked by a small difference in cell composition. To rule out these possibilities, we performed two kinds of analyses on a random subset of the data for which cell counts of nine different cell types were obtained using CyTOF. First, we compared the proportion of each cell type. The proportion of activated Tregs was found to be statistically indistinguishable between asthmatic and healthy subjects (p = 0.528). After multiple testing corrections at a false discovery rate (FDR) of 0.05, Th17 cells were the only type of cell among those tested with a statistically significant difference (see Additional file 1: Table S2 for adjusted p values). We then repeated multiple linear regression to fit the average methylation in the promoter area of the Foxp3 gene and the average methylation of region 3 of intron 4 of the IL10 gene. The cell type composition was included as an additional covariate to adjust for any differences. For IL10, there was a higher DMR of IL10 by 8.9±7.7% (p = 0.026) for asthmatic subjects.
Association between methylation and ambient air pollution
We investigated the association between AAP exposure (CO, O3, NO2, PM2.5) and methylation of CpG sites in the Foxp3 and IL10 genes. The general characteristics of our estimates of AAP exposure were described previously in this “Results” section.
p values and regression coefficients for 90-day pollution exposure prior to blood draw and Foxp3 promoter methylation models
Pollutant in model
The p values and regression coefficients from models analyzing the associations between 90-day average pollutant exposure and normalized methylation averaged over the sites in the promoter of Foxp3. Each row is for one model that includes 90-day exposure for the specified pollutant. Asthma and season are covariates in every model. Since season is modeled as a factor with four levels, it has a coefficient for each season. In the last row, the pollutants are aggregated by their first principal component. PC1 is scaled much differently than the exposure values used in the other rows and its sign is arbitrary, so we omit its coefficient.
Differences in percent methylation across repeated testing of methylation and estimation of ambient air pollution
In a convenience subset of the subjects, we evaluated changes in DMRs on the Foxp3 (n = 33) and IL10 (n = 30) genes after a period of time (mean 820 days, SD 243 days). We evaluated whether our model could explain changes in normalized methylation associated with exposure to AAPs. We predicted percent methylation for each participant at the time of repeat measure using data from 188 subjects (or 179 for IL10) and average normalized methylation to covariates asthma, season, and 90-day average of exposure to a single pollutant. Additional file 1: Figure S5 compares the observed values from the second time point to both the predicted values and to the original values for each gene and each pollutant for the subset of subjects with two visits. For any given individual, the actual normalized values generally change between the two time points. Furthermore, the prediction error for the second time point is generally large, but the model does correctly predict for most individuals whether an individual will have higher or lower DMRs at the second time point than at the first sample time.
Treg: Foxp3 and IL10 protein expression
Cell type and protein expression were measured by CyTOF. The association between both activated (a result of the PMA-ION used to stimulate IL10 cytokine production) Tregs (CD25 high, CD127 low) in CD4+ cells and Foxp3 protein expression vs. Foxp3 methylation was analyzed at each CpG site for the 74 subjects for which we had performed CyTOF analysis. When analyzed using multivariate linear regression with the covariates of asthma status, age, BMI, and sex, Tregs were inversely proportional to the average Foxp3 methylation of the CpG sites in the promoter site (p = .039; Additional file 1: Figure S6a). There was no significant association between Foxp3 methylation and Foxp3 expression (Additional file 1: Figure S6b).
The association between IL10 cytokine expression in CD4+ T cells and average IL10 methylation at region 3 of intron 4 was also analyzed for a subset of subjects with available cytokine data (n = 53). When analyzed using multivariate linear regression with the covariates of asthma status, age, BMI, and sex, IL10 cytokine expression was positively associated with IL10 methylation (p = 0.030; Additional file 1: Figure S6c).
In this cohort study of children from Fresno, an area located in the San Joaquin Valley of California and known for high AAP exposure, we examined DMRs of Foxp3 and IL10, which are genes involved in immune tolerance [28, 29]. Foxp3 plays a key role in maintaining tolerance to common antigens in asthma and allergy . IL10 secretion by T cells can promote tolerance and suppress the production of IgE by B cells . Additionally, prior studies have found an association between exposure to AAPs and Foxp3 or IL10 modifications [14, 31]. Overall, the results were consistent with our primary hypothesis (i.e., higher DMR with increased pollution exposure).
Our results indicate that DMRs in the Foxp3 promoter and IL10 intronic regions are preferentially methylated in asthmatic patients. These results are supported by other studies that have similarly found DNA methylation differences between asthmatic and control subjects at CpG loci in other genes [10, 32]. However, this DMR may be associated with asthma, regardless of the time point of pollution exposure. Furthermore, recent work indicates that heritable epigenetic alterations may mediate the increasing prevalence of atopic diseases [33, 34]. These epigenetic markers may serve as potential biomarkers to support clinical diagnosis.
Several variables measured in this study have been shown to be associated with asthma, such as age, BMI, and sex. We therefore adjusted for these factors in our analyses and found that exposure to AAPs is associated with increased DMRs in the promoter region of Foxp3. These results are consistent with prior studies that have demonstrated that the Foxp3 gene is involved in overall immune tolerance , and more specifically that Tregs, which are suppressors of immune responses involved in asthma pathogenesis, are impaired in pediatric asthma . Foxp3 can also be present in activated CD4+ T effector cells; therefore, its presence is not a sine qua non of a Treg. We have also shown in two prior studies that increased exposure to ambient PAHs is associated with generalized higher differential DMR of the Foxp3 locus [8, 14].
This study provides novel information by identifying specific gene regions that are sensitive to AAPs. We evaluated methylation at the enhancer and promoter regions of the Foxp3 gene. We found methylation at the upstream promoter region of Foxp3 to be associated with both asthma status and exposure to AAPs. However, we found no correlation between methylation and asthma status at the enhancer region. Importantly, our findings of a positive association between AAP exposure and Foxp3 methylation in our initial cross-sectional analyses were consistent with the results of our longitudinal analyses in a subset of subjects.
We also found that individual pollutants differ in their effects on CpG methylation. Exposure to higher levels of NO2, CO, and PM2.5 are associated with higher DMRs of the promoter region of the Foxp3 gene. However, O3 exposure was inversely associated with Foxp3 methylation. This is not surprising as O3 is a regional pollutant with a seasonal pattern opposite to that of the other pollutants (Additional file 1: Figure S7), and we may not be observing the effect of high ozone on methylation, but, rather, the fact that the other pollutants were low.
This is the first study to our knowledge that has evaluated the association between IL10 methylation and air pollution. We found that DMRs in the intronic region 3 are preferentially methylated in patients with asthma. However, we did not find significant associations between AAP and intron 4 methylation. In future studies, additional CpG sites, such as those in the promoter region of IL10, could be evaluated.
While previous investigations have suggested that exposure to higher levels of AAPs are associated with increased prevalence of asthma and allergy [36–39] and that protein expression of IL10 decreased as PAH exposure increased from 24 h to 1 year , previous studies have not specified a critical time of exposure duration. Here, we determined that there was a significant positive association of Foxp3 promoter methylation and pollution for all subjects when there was exposure to certain AAPs (NO2, CO, and PM2.5) over the previous 90 days prior to the blood draw. In addition, AAP exposure averaged over the 90 days was more strongly associated with Foxp3 methylation than it was with season. For the IL10 gene, there was not a significant association between pollutant exposure levels and the amount of intro region methylation, prompting our suggestion that additional IL10 gene regions be evaluated in future studies.
By using CyToF, we were able to link methylation and protein levels in high dimensional cell immune monitoring for these subjects. In addition, we demonstrated that methylation of the promoter region of the Foxp3 gene is negatively associated with activated Treg levels, and IL10 methylation at region 3 of intron 4 is positively associated with IL10 cytokine expression. Foxp3 is produced by both Treg and activated T cells. However, IL10 is synthesized by many immune cells and given there was a wide standard of error, it is difficult to comment on the biological significance of the IL10 results.
We also recognize that our study is subject to a number of limitations. Blood collection occurred across all four seasons, and visits were not evenly distributed within each season. It is possible that the visit peak in April is driven by seasonal variations in airborne pollen and that the regression analyses might not remove the impact of this effect. The 90-day exposure finding could represent opposing effects (pollen/pollutant seasonal variations), which negate each other so that pollutant exposure impact can be detected. Future analyses are planned to address this. Not only do the AAP pollutants vary by season, aeroallergens such as ragweed, oak/maple, and Pinaceae pollen can affect the severity of asthma and pediatric asthma emergency department and clinic visits and have been shown to be associated with daily aeroallergen concentrations . Repeat visits are limited in number and did not necessarily occur in the same season. To compensate, seasonality was included in the models used to investigate associations between methylation and AAP exposure, and 90-day exposure prior to the blood draw had a stronger effect than season, but simply using four seasons to model seasonality might not accurately reflect actual pollen levels in a given year. Furthermore, individual exposure was evaluated with the inverse distance weighting (IDW) method, rather than directly measured [25, 26]. It is also possible that other pollutants, dietary exposure, and toxins from the air, water, and other sources, not measured in this study, may contribute to DNA methylation. The generalizability of the results may be limited because the population was predominantly Hispanic and a convenience sample. And ethnicity was not adjusted in the models. Asthma was defined by self-report of physician diagnosis of asthma, rather than by clinical testing. However, in publications using the same technique, it was found that the agreement of patient self-report and physician diagnosis was correlated  with the questions about “physician-diagnosed asthma” having a specificity of 99% ; therefore, we used this method as part of our study definition of asthma. Our failure to find a significant association between Foxp3 protein expression and Foxp3 methylation may be related to the ongoing debate whether all Treg are Foxp3 positive or whether all Foxp3 positive T cells are Treg or if transcription and protein expression are directly related in Treg . Given that the Treg population is a small fraction of the overall cell count, the measurement of Foxp3 protein expression may be lost in the margin of error when measuring both it and the percentage of Tregs.
In summary, our data demonstrate that (1) CpG methylation in the promoter region of Foxp3 and in the intronic region of IL10 is greater in asthmatic than non-asthmatic subjects; (2) methylation levels at specific CpG sites in the Foxp3 promoter region are associated with level of exposure to the pollutants CO, NO2, PM2.5, and O3 over the previous 90 days prior to the blood draw; and (3) activated Treg levels are negatively associated with average Foxp3 methylation in the promoter region, and IL10 methylation at region 3 of intron 4 is positively associated with IL10 cytokine expression. The findings from this study increase our understanding of the epigenetic effects of exposure to AAPs and may impact the development of preventative regimens and/or therapies to alter methylation of these sites and possibly prevent asthma associated with AAP exposure .
Our study suggests that asthma is significantly associated with higher DMRs within the promoter region of Foxp3 as well as region 3 of intron 4 of IL10. In addition, DMRs in the promoter region of Foxp3 are positively linked with average exposure to CO, NO2, and PM2.5 during the 90 days prior to the blood draw. These findings may provide important insight into understanding how pollution could exert detrimental effects relating to asthma, but need to be confirmed in future studies. Ultimately, this research may be useful in elucidating epigenetic biomarkers of asthma and in developing therapeutic targets by which to prevent or correct epigenetic damage resulting from pollution.
The authors gratefully acknowledge Dr. Robert Hamilton for his input and the assistance of Dr. Vanitha Sampath in editing the manuscript. We also thank Timothy Tyner and the clinical team at UCSF-Fresno for their help in enrolling the subjects. We also wish to thank the subjects and their families in Fresno, CA, for their involvement in this study.
Supported by: 5R01ES020926 (PI, Nadeau), Supplement 3R01ES020926 (Prunicki), Austin Family Foundation (PI, Nadeau), and some use of partial funds from Grant NIEHS/EPA Grant P20 to the University of California, Berkeley/Stanford Children’s Environmental Health Center (Overall PI, Hammond; NIEHS 1P20ES018173 and EPA RD-83459601-0).
Availability of data and materials
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
This publication was developed with R01 funding awarded to Stanford University by NIEHS (PI, Nadeau), R01 Supplement (Awardee Prunicki), and in part with funds awarded to Stanford/UC Berkeley (P20; Overall PI Hammond; NIEHS 1P20ES018173 and EPA RD-83459601-0.) Its contents are solely the responsibility of the grantee and do not necessarily represent the official views of the USEPA. It has not been formally reviewed by EPA. The views expressed in this document are solely those of the authors and do not necessarily reflect those of the Agency. Further, USEPA does not endorse the purchase of any commercial any products or services mentioned in this publication.
KN conceived and designed the study and prepared and wrote the manuscript. MP coordinated the study, processed the samples, performed the CyTOF, and prepared and wrote the manuscript. RM and MdP-S designed and performed the methylation studies. LS, CS, XZ, and DD analyzed the data and created the figures. RL provided the pollution data and assisted in modeling. XZ processed the samples, performed the CyTOF, and assisted with the manuscript tables. TP assisted with the manuscript tables. JB and KH assisted in the interpretation of the data and writing of the manuscript. All authors read and approved the final manuscript.
Ethics approval and consent to participate
All methods and procedures were approved by the Stanford University Institutional Review Board for the Protection of Human Subjects. Written informed consent was received from parents of subjects prior to inclusion in the study.
Consent for publication
The authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Fu LS, Tsai MC. Asthma exacerbation in children: a practical review. Pediatr Neonatol. 2014;55:83–91.View ArticlePubMedGoogle Scholar
- Platts-Mills TA. The allergy epidemics: 1870–2010. J Allergy Clin Immunol. 2015;136:3–13.View ArticlePubMedPubMed CentralGoogle Scholar
- World Health Orgnization. Asthma. 2013. http://www.who.int/mediacentre/factsheets/fs307/en/. Accessed 1 June 2017.Google Scholar
- Carlsten C, Rider CF. Traffic-related air pollution and allergic disease: an update in the context of global urbanization. Curr Opin Allergy Clin Immunol. 2017;17:85–9.View ArticlePubMedGoogle Scholar
- Goodrich JM, Reddy P, Naidoo RN, Asharam K, Batterman S, Dolinoy DC. Prenatal exposures and DNA methylation in newborns: a pilot study in Durban, South Africa. Environ Sci Process Impacts. 2016;18:908–17.View ArticlePubMedPubMed CentralGoogle Scholar
- Alhanti BA, Chang HH, Winquist A, Mulholland JA, Darrow LA, Sarnat SE. Ambient air pollution and emergency department visits for asthma: a multi-city assessment of effect modification by age. J Expo Sci Environ Epidemiol. 2015;26:180–8.View ArticlePubMedGoogle Scholar
- Saunders V, Breysse P, Clark J, Sproles A, Davila M, Wills-Karp M. Particulate matter-induced airway hyperresponsiveness is lymphocyte dependent. Environ Health Perspect. 2010;118:640–6.View ArticlePubMedPubMed CentralGoogle Scholar
- Nadeau K, McDonald-Hyman C, Noth EM, Pratt B, Hammond SK, Balmes J, et al. Ambient air pollution impairs regulatory T-cell function in asthma. J Allergy Clin Immunol. 2010;126:845–52.View ArticlePubMedGoogle Scholar
- Iwanaga K, Elliott MS, Vedal S, Debley JS. Urban particulate matter induces pro-remodeling factors by airway epithelial cells from healthy and asthmatic children. Inhal Toxicol. 2013;25:653–60.View ArticlePubMedGoogle Scholar
- Somineni HK, Zhang X, Biagini Myers JM, Kovacic MB, Ulm A, Jurcak N, et al. Ten-eleven translocation 1 (TET1) methylation is associated with childhood asthma and traffic-related air pollution. J Allergy Clin Immunol. 2015;137:797–805.View ArticlePubMedPubMed CentralGoogle Scholar
- Brunst KJ, Leung YK, Ryan PH, Khurana Hershey GK, Levin L, Ji H, et al. Forkhead box protein 3 (FOXP3) hypermethylation is associated with diesel exhaust exposure and risk for childhood asthma. J Allergy Clin Immunol. 2013;131:592–4.View ArticlePubMedGoogle Scholar
- Prescott SL, Tulic M, Kumah AO, Richman T, Crook M, Martino D, et al. Reduced placental FOXP3 associated with subsequent infant allergic disease. J Allergy Clin Immunol. 2011;128:886–7.View ArticlePubMedGoogle Scholar
- Kohli A, Garcia MA, Miller RL, Maher C, Humblet O, Hammond SK, et al. Secondhand smoke in combination with ambient air pollution exposure is associated with increasedx CpG methylation and decreased expression of IFN-gamma in T effector cells and Foxp3 in T regulatory cells in children. Clin Epigenetics. 2012;4:17.View ArticlePubMedPubMed CentralGoogle Scholar
- Hew KM, Walker AI, Kohli A, Garcia M, Syed A, McDonald-Hyman C, et al. Childhood exposure to ambient polycyclic aromatic hydrocarbons is linked to epigenetic modifications and impaired systemic immunity in T cells. Clin Exp Allergy. 2015;45:238–48.View ArticlePubMedPubMed CentralGoogle Scholar
- Mostafavi N, Vlaanderen J, Chadeau-Hyam M, Beelen R, Modig L, Palli D, et al. Inflammatory markers in relation to long-term air pollution. Environ Int. 2015;81:1–7.View ArticlePubMedGoogle Scholar
- Ji H, Biagini Myers JM, Brandt EB, Brokamp C, Ryan PH, Khurana Hershey GK. Air pollution, epigenetics, and asthma. Allergy Asthma Clin Immunol. 2016;12:51.View ArticlePubMedPubMed CentralGoogle Scholar
- Clifford RL, Jones MJ, MacIsaac JL, McEwen LM, Goodman SJ, Mostafavi S, et al. Inhalation of diesel exhaust and allergen alters human bronchial epithelium DNA methylation. J Allergy Clin Immunol. 2017;139:112–21.View ArticlePubMedGoogle Scholar
- Zhong J, Agha G, Baccarelli AA. The role of DNA methylation in cardiovascular risk and disease: methodological aspects, study design, and data analysis for epidemiological studies. Circ Res. 2016;118:119–31.View ArticlePubMedPubMed CentralGoogle Scholar
- Tsuji-Takayama K, Suzuki M, Yamamoto M, Harashima A, Okochi A, Otani T, et al. The production of IL-10 by human regulatory T cells is enhanced by IL-2 through a STAT5-responsive intronic enhancer in the IL-10 locus. J Immunol. 2008;181:3897–905.View ArticlePubMedGoogle Scholar
- American Lung Association. Most polluted cities. 2017. http://www.lung.org/our-initiatives/healthy-air/sota/city-rankings/most-polluted-cities.html. Accessed 30 May 2017.Google Scholar
- Fresno County asthma profile. 2016. https://www.cdph.ca.gov/Programs/CCDPHP/DEODC/EHIB/CPE/CDPH%20Document%20Library/County%20profiles/Fresno2016profile.pdf. Accessed 1 June 2017.
- Fuss IJ, Kanof ME, Smith PD, Zola H. Isolation of whole mononuclear cells from peripheral blood and cord blood. Curr Protoc Immunol. 2009;Chapter 7:Unit7 1.PubMedGoogle Scholar
- Janson PC, Winerdal ME, Marits P, Thorn M, Ohlsson R, Winqvist O. FOXP3 promoter demethylation reveals the committed Treg population in humans. PLoS One. 2008;3:e1612.View ArticlePubMedPubMed CentralGoogle Scholar
- Environmental Protection Agency. Air data. 2016. http://www3.epa.gov/airdata/. Accessed 1 June 2017.Google Scholar
- Shepard D. A two-dimensional interpolation function for irregularly-spaced data. Proceedings of the 1968 23rd ACM national conference: ACM; 1968. p. 517–24.Google Scholar
- Pebesma EJ. Multivariable geostatistics in S: the gstat package. Comput Geosci. 2004;30:683–91.View ArticleGoogle Scholar
- Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B (Methodological). 1995;57:289–300.Google Scholar
- Runyon RS, Cachola LM, Rajeshuni N, Hunter T, Garcia M, Ahn R, et al. Asthma discordance in twins is linked to epigenetic modifications of T cells. PLoS One. 2012;7:e48796.View ArticlePubMedPubMed CentralGoogle Scholar
- Akdis CA, Akdis M. Mechanisms of immune tolerance to allergens: role of IL-10 and Tregs. J Clin Invest. 2014;124:4678–80.View ArticlePubMedPubMed CentralGoogle Scholar
- Alroqi FJ, Chatila TA. T regulatory cell biology in health and disease. Curr Allergy Asthma Rep. 2016;16:27.View ArticlePubMedPubMed CentralGoogle Scholar
- Latzin P, Frey U, Armann J, Kieninger E, Fuchs O, Roosli M, et al. Exposure to moderate air pollution during late pregnancy and cord blood cytokine secretion in healthy neonates. PLoS One. 2011;6:e23130.View ArticlePubMedPubMed CentralGoogle Scholar
- Acevedo N, Reinius LE, Greco D, Gref A, Orsmark-Pietras C, Persson H, et al. Risk of childhood asthma is associated with CpG-site polymorphisms, regional DNA methylation and mRNA levels at the GSDMB/ORMDL3 locus. Hum Mol Genet. 2015;24:875–90.View ArticlePubMedGoogle Scholar
- Dolinoy DC, Das R, Weidman JR, Jirtle RL. Metastable epialleles, imprinting, and the fetal origins of adult diseases. Pediatr Res. 2007;61:30R–7R.View ArticlePubMedGoogle Scholar
- Yang IV, Pedersen BS, Liu A, O’Connor GT, Teach SJ, Kattan M, et al. DNA methylation and childhood asthma in the inner city. J Allergy Clin Immunol. 2015;136:69–80.View ArticlePubMedPubMed CentralGoogle Scholar
- Hartl D, Koller B, Mehlhorn AT, Reinhardt D, Nicolai T, Schendel DJ, et al. Quantitative and functional impairment of pulmonary CD4+CD25hi regulatory T cells in pediatric asthma. J Allergy Clin Immunol. 2007;119:1258–66.View ArticlePubMedGoogle Scholar
- Kim J, Han Y, Seo SC, Lee JY, Choi J, Kim KH, et al. Association of carbon monoxide levels with allergic diseases in children. Allergy Asthma Proc. 2016;37:1–7.View ArticleGoogle Scholar
- Kim BJ, Hong SJ. Ambient air pollution and allergic diseases in children. Korean J Pediatr. 2012;55:185–92.View ArticlePubMedPubMed CentralGoogle Scholar
- Miller RL, Garfinkel R, Horton M, Camann D, Perera FP, Whyatt RM, et al. Polycyclic aromatic hydrocarbons, environmental tobacco smoke, and respiratory symptoms in an inner-city birth cohort. Chest. 2004;126:1071–8.View ArticlePubMedPubMed CentralGoogle Scholar
- Jung KH, Hsu SI, Yan B, Moors K, Chillrud SN, Ross J, et al. Childhood exposure to fine particulate matter and black carbon and the development of new wheeze between ages 5 and 7 in an urban prospective cohort. Environ Int. 2012;45:44–50.View ArticlePubMedPubMed CentralGoogle Scholar
- Zhong W, Levin L, Reponen T, Hershey GK, Adhikari A, Shukla R, et al. Analysis of short-term influences of ambient aeroallergens on pediatric asthma hospital visits. Sci Total Environ. 2006;370:330–6.View ArticlePubMedPubMed CentralGoogle Scholar
- Baumeister H, Kriston L, Bengel J, Harter M. High agreement of self-report and physician-diagnosed somatic conditions yields limited bias in examining mental-physical comorbidity. J Clin Epidemiol. 2010;63:558–65.View ArticlePubMedGoogle Scholar
- Toren K, Brisman J, Jarvholm B. Asthma and asthma-like symptoms in adults assessed by questionnaires. A literature review. Chest. 1993;104:600–8.View ArticlePubMedGoogle Scholar
- Ziegler SF. FOXP3: of mice and men. Annu Rev Immunol. 2006;24:209–26.View ArticlePubMedGoogle Scholar
- Kawasaki H, Taira K. Induction of DNA methylation and gene silencing by short interfering RNAs in human cells. Nature. 2004;431:211–7.View ArticlePubMedGoogle Scholar