Figures
Abstract
Growing evidence suggests that maternal exposures to endocrine disrupting chemicals during pregnancy may lead to poor pregnancy outcomes and increased fetal susceptibility to adult diseases. Polybrominated diphenyl ethers (PBDEs), which are ubiquitously used flame-retardants, could leach into the environment; and become persistent organic pollutants via bioaccumulation. In the United States, blood PBDE levels in adults range from 30–100 ng/g- lipid but the alarming health concern revolves around children who have reported blood PBDE levels 3 to 9-fold higher than adults. PBDEs disrupt endocrine, immune, reproductive and nervous systems. However, the mechanism underlying its adverse health effect is not fully understood. Epigenetics is a possible biological mechanism underlying maternal exposure-child health outcomes by regulating gene expression without changes in the DNA sequence. We sought to examine the relationship between maternal exposure to environmental PBDEs and promoter methylation of a proinflammatory gene, tumor necrosis factor alpha (TNFα). We measured the maternal blood PBDE levels and cord blood TNFα promoter methylation levels on 46 paired samples of maternal and cord blood from the Boston Birth Cohort (BBC). We showed that decreased cord blood TNFα methylation associated with high maternal PBDE47 exposure. CpG site-specific methylation showed significantly hypomethylation in the girl whose mother has a high blood PBDE47 level. Consistently, decreased TNFα methylation associated with an increase in TNFα protein level in cord blood. In conclusion, our finding provided evidence that in utero exposure to PBDEs may epigenetically reprogram the offspring’s immunological response through promoter methylation of a proinflammatory gene.
Citation: Dao T, Hong X, Wang X, Tang W-Y (2015) Aberrant 5’-CpG Methylation of Cord Blood TNFα Associated with Maternal Exposure to Polybrominated Diphenyl Ethers. PLoS ONE 10(9): e0138815. https://doi.org/10.1371/journal.pone.0138815
Editor: Toshi Shioda, Massachusetts General Hospital, UNITED STATES
Received: February 24, 2015; Accepted: September 3, 2015; Published: September 25, 2015
Copyright: © 2015 Dao et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This work was in part supported by National Institutes of Health grants R00ES016817 (WYT) and R21HD066471 (XBW). Other supports were come Discretionary Fund (WYT) from Department of Environmental Health Sciences, Johns Hopkins Bloomberg School of Public Health. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Polybrominated diphenyl ethers (PBDEs), known as endocrine disrupting compounds, affecting the endocrine, immune, reproductive and nervous systems [1]. PBDEs, until recently, are ubiquitously used flame-retardants, in many consumer products, that have leached into the environment and bioaccumulated to become persistent organic pollutants (POPs) [2, 3]. There are over 209 congeners of PBDEs, but five are significant to humans, since they account for 90% of the total body burden; PBDE-47, -99, -100, -153 and -154 [2]. Human PBDE exposure occurs through multiple exposure pathways; inhalation, ingestion and dermal contact [4, 5]. Many of the lower PBDE congeners are lipophilic and accumulate in fatty tissues. They can easily cross the placenta into the fetal circulation [6]. Nursing mothers can transfer PBDEs to infants through breast milk [4]. PBDEs also accumulate in fetal lipid-rich tissues such as brain, liver and adipose. PBDEs are persistent in the body with estimated human half-lives of 2–12 years [6]. Epidemiological and experimental studies suggest that PBDEs are developmental neurotoxicants, at least partly impairing thyroid hormone homeostasis, oxidative damage or dysregulation of neurotransmitter signaling in the brain [7].
2,2’,4,4’-Tetrabromodiphenyl ether (PBDE47) accounts for 50% of the PBDE body burden [7] and is the PBDE congener found at the highest levels in human serum and breast milk [8–10]. Female mice exposed prenatally or postnatally to repeated low doses of PBDE47 resulted in the maternal transfer of PBDE47 from dams to the fetus blood, brain, liver and fat tissues. The offspring showed growth retardation and behavioral deficiencies [3, 5] [11]. These animal findings are supported by epidemiological evidence of the relationship between PBDE47 levels and cognitive delay [12, 13]. Furthermore, animals exposed to a commercial PBDE mixture (PBDE47 is the dominant congener) showed decreased lymphocyte proliferation and antibody production [14], suggesting that PBDEs may pose a risk of immunotoxicity on specific target tissues. PBDE47 is postulated to disrupt thyroid function because it shares a similar structure to thyroid hormone [1]. However, studies have reported that PBDE47 does not interact with the thyroid hormone receptor [15] suggesting that alternative mechanisms may underlie the toxicity of PBDE47 in the cells.
It has been proposed that maternal PBDE exposure induces epigenetic reprogramming of the offspring and associates with increased disease risks in children. The “Barker Hypothesis” postulated that organs undergo developmental programming in the womb that predetermines subsequent physiological and metabolic adaptations during childhood or as adults [16, 17]. Epigenetics now underpins the developmental reprogramming by demonstrating the molecular relationship between the exposure to environmental pollutants/toxicants and gene expression changes that influence disease susceptibility [18–23]. Epigenetics mechanisms (DNA methylation, histone modifications and microRNAs) act singularly or conjointly to regulate gene expression without altering DNA sequences in response to environmental exposures. Therefore, they produce an array of unique phenotypes that control cell differentiation and organ development. DNA methylation is the covalent addition of a methyl group from S-adenosyl-methionine (SAM), by DNA methyltransferases (DNMTs), to the fifth position of cytosine in CpG dinucleotides to generate 5-methylcytosine (5-mC). DNA methylation at CpG dinucleotides mostly occurs in the 5’ flanking region and often associates with repression of gene activity. The biological contribution of DNA methylation may vary according to the position of the specific CpG sites relative to the transcription unit and the binding of transcriptional factors at the promoter [24]. Decreased global DNA methylation associates with high blood POPs levels [25]. A longitudinal birth cohort study, Center for the Health Assessment of Mothers and Children of Salinas (CHAMACOS) reported increased parental exposure to PBDEs associates with decreased global DNA methylation, measured on the Alu (Arthrobacter luteus) and LINE-1 (Long Interspersed Elements) sequences, in the cord blood of newborns [26]. By stratifying the children by sex, they revealed that there is a dynamic relationship between the PBDEs exposure and global DNA methylation changes that result from differences in sex, age, exposure-dose and multiple exposures. Mice prenatally exposed to PBDE47 showed decreased global methylation in their brains that persists into adulthood [1]. Recently, low perinatal PBDE47 exposure has been demonstrated to influence the methylation of mitochondrial cytochrome c oxidase (Mt-co2) and genes related to brain function (Bdnf and Nr3c1) [27]. All in all, it suggests exposure to PBDE47 may affect cell and tissue functions by regulating gene expression through DNA methylation.
Exposure to PBDEs in rodents was shown to impair the immune function [28, 29]. Their interaction with the circulating immune cells may be driving the immunological signals to the target tissues. It could be expected that early life exposure to PBDEs may disrupt immune cell functions via epigenetic modifications of gene regulation. Once the epigenetic program is set, epigenetic disruption of the immune system can persist throughout life, resulting in increased susceptibility to inflammation-related diseases later in life. Additionally, dysregulation of immune responses has been suggested to be associated with neurological dysfunctions. Production of cytokines like Interleukin 1 beta (IL1β) and Tumor Necrosis Factor alpha (TNFα) from the peripheral blood cells are extensively studied and suggested to be affecting the central nervous system [30–32]. However, there are no studies investigating the relationship between the PBDEs exposure and DNA methylation patterns of these two cytokines. We first utilized an in vitro system to test the epigenetic effect of PBDEs on these cytokines by exposing peripheral blood mononuclear cells (PBMCs) to PBDE47. TNFα methylation, not IL1β, showed a linear relationship to the doses of PBDEs. As a proof of concept, we utilized paired maternal and cord blood samples from the Boston Birth Cohort (BBC) and examined if maternal PBDE47 exposure was associated with cord blood TNFα promoter methylation. Our findings may provide the insight to understand the adverse effects of maternal PBDEs exposure on human health through novel epigenetic mechanisms, affecting immunity and inflammation and resulting in higher susceptibility of immune diseases like food allergy, asthma, or other metabolic symptoms [33, 34].
Materials and Methods
Boston Birth Cohort (BBC)
This study included a subset of mother-infants pairs from the BBC, a cohort consisting of multiethnic mother-infant pairs (predominantly African American) enrolled 24 to 72 hours post delivery and followed prospectively from birth onward as detailed previously [34–36]. We collected the biospecimens including maternal blood and cord blood just after delivery. We obtained written informed consent from each mother. The study protocol was approved by the Institutional Review Boards (IRB) of Boston University Medical Center and by the IRB of Johns Hopkins University Bloomberg School of Public Health.
We measured the concentrations of lipid normalized to PBDEs in peripartum maternal serum samples of the BBC (Table 1). The median level of PBDE47 (12.1ng/g-lipid) was used as the rough cut-off for sample selection. To select samples for the high-level maternal PBDE47 exposed group, we sorted all the available samples and chose samples with PBDE47 levels higher than the median and with available clinical data and DNA samples. Similarly, to select samples for the low-level maternal PBDE47 exposed group, we sorted all the available samples and chose samples with the PBDE47 level lower than the median and with available clinical data and DNA samples. As a result, subjects in the low-level PBDE47 exposed group have maternal PBDE47 < 8.4 ng/g-lipid (about 36th percentile in the BBC) and subjects in the high-level PBDE47 exposed group have maternal PBDE47 > 20 ng/g-lipid (about 70th percentile in the BBC). The distribution of PBDE47 level in each group was illustrated in Table 2. The subjects in the low (n = 23) and high (n = 23)–level PBDE47 exposed groups also fall into the same low and high-level of the sum of all PBDEs measured.
PBDEs measurement in maternal serum
Maternal serum samples from each subject were prepared for shipment using Center for Disease Control (CDC)’s protocols and were sent to the CDC in Atlanta for PBDE measurement using an established and validated PBDE assay. The analytical method and quality control procedures have been described previously [37]. The method used for sample processing included automatic fortification of the samples with internal standards as well as addition of formic acid and water for denaturation and dilution of the samples using a Gilson 215 liquid handler (Gilson Inc.; Middleton, WI). The samples were thereafter extracted by solid phase extraction (SPE) using a Rapid Trace (Caliper Life Sciences; Hopkinton, MA) modular SPE system. Removal of co-extracted lipids was performed on a silica/sulfuric acid column using the Rapid Trace equipment for automation. Final analytical determination of the target analytes was performed by gas chromatography isotope dilution high resolution mass spectrometry (GC-IDHRMS) employing a MAT95XP (ThermoFinnigan MAT, Bremen, Germany) instrument. The samples were analyzed for 10 PBDE congeners: PBDE-17, -28, -47, -66, -85, -99, -100, -153, -154, -183 (comprising tri-, tetra-, penta-, hexa- and hepta-brominated congeners, but not deca).
Measurement of TNFα in cord blood serum sample
Cord blood sample was collected at delivery, using a BD Vacutainer® Plus Plastic K20EDTA tube (purple top). The sample was centrifuged at 2500g (4°C) for 10 minutes. Cord blood serum (the supernatant) was split into 3 aliquots and stored at −80°C. Cord blood cell fraction was transferred into a single cryotube and frozen at -80°C if DNA isolation was not performed immediately after collection. Serum sample was analyzed simultaneously for TNFα by immunoassay using flowmetric Luminex xMAP technology (Luminex Corp, Austin, TX), as reported previously [38].
DNA isolation of cord blood leukocytes
Genomic DNA was extracted from cord blood leukocytes and quantified with SpectraMax M2 (www.moleculardevices.com). In brief, the cord blood cell fraction was subject to red blood cell (RBC) lysis and shaken for 10 minutes before centrifugation at 3000g (4°C) for 10 minutes. The cell pellet presented as the leukocytes and subjected to DNA isolation to collect the genomic DNA.
Bisulfite genomic sequencing
Genomic DNA (200ng) extracted from cord blood leukocytes was bisulfite-treated (EZ DNA methylation kit, Zymo Research) before PCR. The CpG rich region of the 5’ promoter region of TNFα (NC_000006.11: 31541344–31545344) was revealed by analyses of genomic sequences at MethPrimer (http://www.urogene.org/methprimer/) (Fig 1). Primers were designed to amplify a 404 base pair fragment (-354 to +50) including a total of 12 CpG sites from bisulfite-treated DNA: hBS-TNFα-F1:5’-TTGGTTTTTAAAAGAAATGGAGGTA-3’; hBS-TNFα-R1: 5’-TCTCCCTCTTAACTAATCCTCTACTATC-3’. PCR, using GoTaq PCR mix (Promega, MO), was performed on the bisulfite-treated DNA (95°C for 10 minutes, 35 cycles of 95°C for 30 seconds, 60°C for 1 minute and 72°C for 2 minutes, followed by an extension at 72°C for 12 minutes). PCR products were purified (GeneJET Gel Extraction Kit, Thermo Scientific, NY) and subcloned into pCR 2.1 vector (Invitrogen, CA). 4–6 clones from each sample were sequenced (Macrogen, MD) to obtain direct measures of DNA methylation at each CpG site in the TNFα promoter region. Sequencing data was analyzed with the BiQ analyser. Genomic DNA with 0 and 100% methylation was used as a quality control to ensure bisulfite conversion quality on sample.
The sequence of the primers used for bisulfite genomic sequencing (BS- TNFα-F1, BS- TNFα-R1) are underlined and in italic bold type. There are a total of 12 CpG nucleotides (shaded in gray) located at the promoter region of TNFα encompassing the transcription start site (TSS).
Data analysis
We analyzed the average % methylation of TNFɑ promoter as well as % methylation of each of the 12 CpG sites in the TNFα promoter region, separately. % methylation at each CpG site was calculated using the mean value obtained from 4–6 clones for each subject. Average % methylation of the TNFɑ promoter was calculated by taking the average of the methylation level of all the analyzed 12 CpG sites within the TNFɑ promoter region. A scatter plot with smooth fitted line was generated in R (Version 3.01) to explore the relationship between the average % methylation of TNFɑ promoter and log10-transformed maternal blood PBDE47 level (Fig 2). We plotted the distribution of average % methylation of the TNFɑ promoter in the low- and high- level PBDE47 exposed groups, separately (Fig 3). We then compared the difference between these two groups using an unpaired t-test (two-tailed p-value) with Welch’s correction. In addition, we generated the plots to describe the distribution of % methylation at each of 12 CpG sites in low- and high-level PBDE47 exposed groups (boys and girls, separately), and then compared the difference between the two groups by two-way ANOVA (Fig 4). We presented p-value without adjustment for multiple comparisons due to the fact that 1) the sample size was small; 2) these CpG sites were not independent of each other, in which, Bonferroni adjustment appeared to be very stringent. The difference of cord blood TNFα protein (in log-transformation) between low- and high-level PBDE47 exposed groups was compared using an unpaired t-test. Spearman correlation coefficient was calculated to determine the relationship between cord blood TNFα protein level and average % methylation of TNFɑ promoter. Graphs were generated using Prism 6 (GraphPad Software, CA).
Methylation of each CpG site of the TNFɑ promoter was assayed in 46 cord blood DNA samples by bisulfite genomic sequencing. Average % methylation of the TNFɑ promoter (y-axis) was calculated by taking an average of the methylation level of a total 12 CpG sites within the TNFɑ promoter region.
Average % methylation of the TNFɑ promoter (y-axis) was calculated by taking an average of the methylation level of a total 12 CpG sites within the TNFɑ promoter region in 46 cord blood DNA samples. Results were compared between low- and high-level PBDE47 exposed groups. Each circle and square represented the average percent methylation level for each subject in low-level PBDE47 exposed group and high-level PBDE47 exposed group, respectively. Boys (in black) and girls (in red) are marked differently. The error bars represented means (±standard error of mean, SEM).
Percentage of methylation of each CpG site of the TNFα promoter represented as mean ± SEM in low- (gray open circles) and high-level (black squares) PBDE47 exposed group. The difference in % methylation in these 12 CpG sites between the two groups was compared by two-way ANOVA. * p<0.05, meant % methylation at a CpG site was significantly different between low- and high- level PBDE47 exposed groups (based on t-test).
Results
PBDE47 shows the most abundance among PBDEs congers in BBC
In the BBC cohort, a total of 247 maternal blood samples had available data for blood PBDEs level (of 10 PBDE congeners), which showed a wide range of variability (Table 1). PBDE47 is the most prevalent PBDE (accounting for 58% of the total PBDEs level) in this cohort (Table 1) and thus we focused on the epigenetic effect of PBDE47 in our study. A total of 46 mother-infant pairs (23 in low-level maternal PBDE47 exposed group and 23 in high-level maternal PBDE47 exposed group) were selected for this study (Table 2). There was no significant difference in baby’s gender, smoking status, parity, preterm status, maternal education, maternal age and pre-pregnancy BMI between the selected 46 samples and the rest of the 201 samples (Table 3). Among the 46 selected samples, mothers with high PBDE47 exposure were significantly younger than those with low PBDE47 exposure (p<0.01, Table 3).
Decreased cord blood TNFα promoter methylation associated with high maternal PBDE47 exposure
We first utilized an in vitro system to test the epigenetic effect of PBDE47 on these cytokines by exposing peripheral blood mononuclear cells (PBMCs) to PBDE47 (0.1–100nM). TNFα methylation, not IL1β, showed a linear relationship to the doses of PBDEs (data in S1 Fig). DNA methylation patterns of a total of 12 CpG sites located at the promoter region of TNFα (Fig 1) were analyzed in cord blood leukocytes. Previous studies have shown the significance of CpG site-specific DNA methylation in TNFα transcription although the TNFα gene does not have traditional CpG islands (GC%>60%) at its promoter region [39]. We first examined the linear relationship between average % methylation of the TNFα promoter in cord blood and maternal blood PBDE47 level (log10-transformed). As shown in Fig 2, their relationship was not linear. Instead, TNFɑ methylation remained unchanged when maternal blood PBDE47 level (log10-transformed) was less than 0.7 (or PBDE47 level <5 ng/g-lipid) but TNFɑ methylation decreased with increasing maternal PBDE47 exposure when maternal blood PBDE47 level (log10-transformed) is greater than 0.7 (or PBDE level > 5ng/g-lipid).
Compared to that in the low-level PBDE47 exposed group, average % methylation of the TNFα promoter was lower in the high-level PBDE47 exposed group (p = 0.046) (Fig 3). When stratified by the gender of offspring, a similar association was found in girls (p = 0.004), but not in boys (p>0.99) (Fig 3). To determine if one or more CpG sites at the TNFα promoter were more sensitive to PBDE47 exposure, we compared % methylation at each individual CpG site between low- and high-level PBDE47 exposed groups. In boys, there was no significant difference in % methylation of any CpG sites between low- and high-level PBDE47 exposed groups. In girls, three CpG sites, CpG_-172 (p = 0.04), CpG_-75 (p = 0.05), and CpG_+8 (p = 0.05) showed significantly decreased % methylation in the high-level PBDE47 exposed group as compared to that of low-level PBDE47 exposed group (Fig 4).
Inverse relationship between cord blood TNFα promoter methylation and cord blood TNFα protein level in girls
To examine if the decreased cord blood TNFα promoter methylation contributed to increased gene expression, we measured TNFα protein level in 23 (out of 46 selected samples) cord blood serum as RNA sample was not available. The 23 selected samples were not chosen by either maternal PBDE47 level or cord blood methylation status. The analysis of TNFα protein level in serum was primarily designed to study preterm birth and related traits in other BBC-related studies. Therefore, we would expect more cases of preterm birth among the 23 selected samples we measured cord blood TNFα protein. When compared with those unselected samples (without TNFα protein analysis), these 23 selected samples were more apt to preterm birth (65.2% vs 8.7%, p = 0.0003). Nonetheless, they were comparable with unselected samples on other variables, such as, gender, nullipartiy, cesarean section, maternal education, maternal age and maternal BMI. The distribution of cord blood TNFα protein level in each PBDE47 exposed group was shown in Table 4. The median of TNFα protein level in the low-level PBDE47 exposed group (n = 13) (28.2 pg/mL) was higher than that of high-level PBDE47 exposed group (n = 10) (19.3pg/mL). A similar trend was found in boys (n = 15). In girls (n = 8), cord blood TNFα protein level tended to be higher in the high-level PBDE47 exposed group than in the low-level PBDE47 exposed group. However, none of the differences were statistically significant.
As shown in Table 5, average % methylation of the TNFα promoter was not significantly correlated to cord blood TNFα protein level in 23 samples, as indicated by the Spearman’s rank coefficient (ρ, rho). When we focused this analysis to each specific CpG site at TNFα promoter, we did not identify any significant association in the total 23 samples and in boys (n = 15). In girls, we found that % methylation at CpG_+8 negatively correlated with cord blood TNFα level (ρ = -0.733, p-value 0.039). It suggested increases in cord blood TNFα protein level associates with hypomethylation at a specific CpG site (CpG_+8) at the TNFα promoter.
Discussion
Our results suggested that maternal PBDE47 exposure altered CpG site specific DNA methylation at the TNFα promoter, which may contribute to the aberrant TNFα gene expression in the offspring. There are intensive studies that have reported the toxic effect of PBDEs on neural and thyroid systems [7, 12–13]. Strikingly, PBDEs can also influence the functions of the immune system [14, 28–29]. Our previous studies reported a number of CpG sites (located within the genes with immunological functions) showing significant longitudinal changes in methylation levels [34]. Recently, we have demonstrated that pre-pregnancy maternal BMI associated with the epigenetic alterations in offspring’s genes related to the inflammatory and lipid metabolism response [40]. Here, we provide the evidence showing the relationship between maternal PBDE47 exposure and gene-specific methylation of a proinflammatory gene at birth. TNFα is a proinflammatory cytokine produced mainly by macrophages but also found in T cells, B cells and fibroblasts. This cytokine has been implicated in a variety of diseases including immune-related disease [41–43] and cancer [44]. The proximal region of the TNFα promoter (~200 bp upstream of the TSS) regulates the transcription of TNFα in multiple cell types that respond to various stimuli including the T cell and B cell activation, infection and cytokines [45]. Sullivan et al demonstrated that TNFα production correlated to the CpG site-specific methylation of the TNFα promoter in monocytes and THP-1 cells in the presence of lipopolysaccharide (LPS) [39]. By using a subset of the BBC, we showed that TNFα hypomethylation in cord blood correlated to high maternal PBDE47 exposure. Remarkably, we found that girls are more susceptible to TNFα promoter hypomethylation in response to maternal PBDE47 exposure. Further, we showed, in girls, % methylation at the specific CpG site (CpG_-172, CpG_-75 and CpG_+8) as previously described by Sullivan et al [39], significantly associated with maternal PBDE47 exposure. In addition, methylation status of CpG_+8 (girls) inversely correlated to TNFα protein level in cord blood. It suggests that maternal PBDE47 exposure might alter TNFα production in the offspring via CpG site-specific DNA methylation. We did not note a statistically significant correlation between cord blood TNFα protein level and maternal PBDE47 exposure in girls, although we found the median of TNFα protein level in the high-level PBDE47 exposed group was higher than that of low-level PBDE47 exposed group. We acknowledged that this TNFα protein analysis is limited by a relatively small sample size and so these findings warrant further validation in other samples enrolled in BBC.
Studies showed that transcriptional control at the TNFα promoter associated with its production and with the disease risk. In Jurkat T cells or human monocytic cells, promoter methylation [39, 46], histone acetylation and methylation [47–49] together with other transcription factors [50] could regulate transcription of TNFα gene. Long noncoding RNAs and their binding proteins were shown to regulate gene expression of TNFα [51]. Histone modifications at the TNF promoter also associated with the disease states of diabetes [52] and systemic lupus erythematosus [53]. Women with higher adiposity showed decreased methylation of the TNFα promoter and higher plasma TNFα protein level in their peripheral white blood cells [54]. Further, epigenetic regulation of the monocyte and T cell lineages by TNFα production showed to play a key role in immunological diseases [55]. These results suggested that epigenetic regulation of TNFα may contribute to the modulation of inflammatory responses.
No studies have examined the association between the maternal PBDE exposure and TNFα promoter methylation in the developing fetus. Newborns with growth restriction showed increased TNFα expression in the placenta [56]. It suggests that TNFα cytokine expression may influence developmental outcomes of the offspring. Our study, although limited by a small sample size and no adjustment for multiple testing, may represent the first step to demonstrate the potential influence of prenatal PBDE exposure on the fetal epigenome. We acknowledged that cord blood leukocytes utilized in the current study consist of many functionally and developmentally distinct cell populations in varying proportions. Because DNA methylation is cell-type specific, it is possible that cellular heterogeneity may confound our methylation measurements. It suggests that differential cell counting and/or sorting of cells from whole blood is required to adjust the findings from the methylation patterns by regression model (using the estimated cell type composition as a covariate in the regression model [57]) or access the methylation patterns in different types of blood cells [58, 59]. Unfortunately, it is not feasible to obtain the cell distribution of the immune cells in the archived samples we utilized in this study. Given the fact that cord blood leukocytes may contain stem cells that can populate the brain in later life [60] and also provide a rich source of immune cells which are important producers of cytokines [61], it suggests that cord blood leukocytes (if only archived cord blood samples are available) may provide a reasonable surrogate for our target organ/tissues (brain/immune cells). Our previous study demonstrated the feasibility of using cord blood DNA methylation as a surrogate biomarker to associate the risk of childhood asthma [21, 23]. All-in all, in the future, we will take this step further by performing replicated studies in a larger sample pool and by following up the developmental outcomes of the offspring from infancy, childhood, to adolescence. We will assess the disease phenotypes like food allergy, asthma as well as metabolic and neurological disorders. If we further confirm our findings on a relatively large sample size at the BBC, we may be able to evaluate the use of cord blood TNFα methylation patterns as biomarkers for the susceptibility of immunological and neurological diseases. Or we can measure cord blood TNFα methylation to assess the adverse health effects of maternal PBDEs exposure.
In the present study, we adjust the BBC samples by mother’s age and birth implications and outcomes (low birth weight). Nevertheless, profound factors for TNFα methylation may include social-economics status, psychological stress, ethnicity, and exposures to other endocrine disrupting chemicals and PBDEs congers. PBDE47 showed sex-, age-, species- and dose-dependent effects on its metabolism and elimination in experimental animal studies [62–65]. Therefore, we must carefully choose proper data modeling and multiple testing when we employ the future epigenetic analyses. We will need a set of follow-up studies, in vitro or in vivo, to show the epigenetic effect of the PBDEs toxicity in humans. Taken together, our findings provide evidence that maternal exposure to PBDEs may induce epigenetic reprogramming of the offspring’s immune response by aberrant DNA methylation of a proinflammatory gene.
Supporting Information
S1 Fig. Exposure to PBDE47 decreased TNFɑ but not IL1ß methylation status of peripheral blood mononuclear cells.
Peripheral blood mononuclear cells (PBMCs, HemaCare, CA) were exposed to a range of PBDE47 (AccuStandard, NH) exposures for 3 days. DNA isolated from cells was undergone bisulfite treatment before PCR. Promoter methylation status for TNFɑ and IL1ß was assayed by bisulfite genomic sequencing. Results were collected from three independent sets of experiment and analyzed by 2-tailed ttest. Error bar represented standard error of mean. p<0.05 presented result was statistically significant in PBDE47-exposed group when compared with that of untreated control (PBDE47_0nM).
https://doi.org/10.1371/journal.pone.0138815.s001
(TIFF)
Author Contributions
Conceived and designed the experiments: WYT XBW. Performed the experiments: TD XMH. Analyzed the data: XMH WYT. Contributed reagents/materials/analysis tools: WYT XBW. Wrote the paper: TD XMH XBW WYT.
References
- 1. Woods R, Vallero RO, Golub MS, Suarez JK, Ta TA, Yasui DH, et al. Long-lived epigenetic interactions between perinatal PBDE exposure and Mecp2308 mutation. Human molecular genetics. 2012;21(11):2399–411. pmid:22343140
- 2. Costa LG, de Laat R, Tagliaferri S, Pellacani C. A mechanistic view of polybrominated diphenyl ether (PBDE) developmental neurotoxicity. Toxicology letters. 2014;230(2):282–94. pmid:24270005
- 3. Ta TA, Koenig CM, Golub MS, Pessah IN, Qi L, Aronov PA, et al. Bioaccumulation and behavioral effects of 2,2',4,4'-tetrabromodiphenyl ether (BDE-47) in perinatally exposed mice. Neurotoxicology and teratology. 2011;33(3):393–404. pmid:21334437
- 4. Abdallah MA, Harrad S. Polybrominated diphenyl ethers in UK human milk: implications for infant exposure and relationship to external exposure. Environment international. 2014;63:130–6. pmid:24291653
- 5. Koenig CM, Lango J, Pessah IN, Berman RF. Maternal transfer of BDE-47 to offspring and neurobehavioral development in C57BL/6J mice. Neurotoxicology and teratology. 2012;34(6):571–80. pmid:23022914
- 6. Geyer HJ, Schramm K, Darnerud PO, Aune M, Feicht EA, Fried KW, et al. Terminal elimination half-lives of the brominated flame retardants TBBPA, HBCD, and lower brominated PBDEs in humans. Organohalogen Compounds. 2004;66:3820–5.
- 7.
(EPA) USEPA. An Exposure Assessment of Polybrominated Diphenyl Ethers. Washington, DC 2010.
- 8. Koh TW, Chih-Cheng Chen S, Chang-Chien GP, Lin DY, Chen FA, Chao HR. Breast-milk levels of polybrominated diphenyl ether flame retardants in relation to women's age and pre-pregnant body mass index. Int J Hyg Environ Health. 2010;213(1):59–65. pmid:19781992
- 9. Costa LG, Giordano G. Developmental neurotoxicity of polybrominated diphenyl ether (PBDE) flame retardants. Neurotoxicology. 2007;28(6):1047–67. pmid:17904639
- 10. Suvorov A, Girard S, Lachapelle S, Abdelouahab N, Sebire G, Takser L. Perinatal exposure to low-dose BDE-47, an emergent environmental contaminant, causes hyperactivity in rat offspring. Neonatology. 2009;95(3):203–9. pmid:18799892
- 11. Suvorov A, Takser L. Delayed response in the rat frontal lobe transcriptome to perinatal exposure to the flame retardant BDE-47. Journal of applied toxicology: JAT. 2011;31(5):477–83. pmid:21394737
- 12. Herbstman JB, Sjodin A, Kurzon M, Lederman SA, Jones RS, Rauh V, et al. Prenatal exposure to PBDEs and neurodevelopment. Environmental health perspectives. 2010;118(5):712–9. pmid:20056561
- 13. Eskenazi B, Chevrier J, Rauch SA, Kogut K, Harley KG, Johnson C, et al. In utero and childhood polybrominated diphenyl ether (PBDE) exposures and neurodevelopment in the CHAMACOS study. Environmental health perspectives. 2013;121(2):257–62. pmid:23154064
- 14. Martin PA, Mayne GJ, Bursian FS, Tomy G, Palace V, Pekarik C, et al. Immunotoxicity of the commercial polybrominated diphenyl ether mixture DE-71 in ranch mink (Mustela vison). Environ Toxicol Chem. 2007;26(5):988–97. pmid:17521147
- 15. Suvorov A, Bissonnette C, Takser L, Langlois MF. Does 2,2',4,4'-tetrabromodiphenyl ether interact directly with thyroid receptor? Journal of applied toxicology: JAT. 2011;31(2):179–84. pmid:20737425
- 16. Hales CN, Barker DJ, Clark PM, Cox LJ, Fall C, Osmond C, et al. Fetal and infant growth and impaired glucose tolerance at age 64. BMJ (Clinical research ed). 1991;303(6809):1019–22.
- 17. Barker DJ, Osmond C, Simmonds SJ, Wield GA. The relation of small head circumference and thinness at birth to death from cardiovascular disease in adult life. BMJ (Clinical research ed). 1993;306(6875):422–6.
- 18. Tang WY, Ho SM. Epigenetic reprogramming and imprinting in origins of disease. Reviews in endocrine & metabolic disorders. 2007;8(2):173–82.
- 19. Jirtle RL, Skinner MK. Environmental epigenomics and disease susceptibility. Nature reviews Genetics. 2007;8(4):253–62. pmid:17363974
- 20. Ho SM, Tang WY, Belmonte de Frausto J, Prins GS. Developmental exposure to estradiol and bisphenol A increases susceptibility to prostate carcinogenesis and epigenetically regulates phosphodiesterase type 4 variant 4. Cancer research. 2006;66(11):5624–32. pmid:16740699
- 21. Tang WY, Levin L, Talaska G, Cheung YY, Herbstman J, Tang D, et al. Maternal exposure to polycyclic aromatic hydrocarbons and 5'-CpG methylation of interferon-gamma in cord white blood cells. Environmental health perspectives. 2012;120(8):1195–200. pmid:22562770
- 22. Tang WY, Newbold R, Mardilovich K, Jefferson W, Cheng RY, Medvedovic M, et al. Persistent hypomethylation in the promoter of nucleosomal binding protein 1 (Nsbp1) correlates with overexpression of Nsbp1 in mouse uteri neonatally exposed to diethylstilbestrol or genistein. Endocrinology. 2008;149(12):5922–31. pmid:18669593
- 23. Perera F, Tang WY, Herbstman J, Tang D, Levin L, Miller R, et al. Relation of DNA methylation of 5'-CpG island of ACSL3 to transplacental exposure to airborne polycyclic aromatic hydrocarbons and childhood asthma. PloS one. 2009;4(2):e4488. pmid:19221603
- 24. Jones PA. Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat Rev Genet. 2012; 13(7):484–92. pmid:22641018
- 25. Rusiecki JA, Baccarelli A, Bollati V, Tarantini L, Moore LE, Bonefeld-Jorgensen EC. Global DNA hypomethylation is associated with high serum-persistent organic pollutants in Greenlandic Inuit. Environmental health perspectives. 2008;116(11):1547–52. pmid:19057709
- 26. Huen K, Yousefi P, Bradman A, Yan L, Harley KG, Kogut K, et al. Effects of age, sex, and persistent organic pollutants on DNA methylation in children. Environmental and molecular mutagenesis. 2014;55(3):209–22. pmid:24375655
- 27. Byun HM, Benachour N, Zalko D, Frisardi MC, Colicino E, Takser L, Baccarelli AA. Epigenetic effects of low perinatal doses of flame retardant BDE-47 on mitochondrial and nuclear genes in rat offspring. Toxicology. 2015; 328:152–159. pmid:25533936
- 28. Du P, Li Z, Du L, Zhang H, Zhou Y, Sun W, Xiao X, He Y, Sun B, Yu Y, Chen D.The effects of PBDE-209 exposure during pregnancy on placental ET-1 and eNOS expression and the birth weight of offspring. Int J Dev Neurosci. 2015 Mar 27.
- 29. Liu X, Zhan H, Zeng X, Zhang C, Chen D. The PBDE-209 exposure during pregnancy and lactation impairs immune function in rats. Mediators of inflammation. 2012;2012:692467. pmid:22619485
- 30. Viviani B, Boraso M, Marchetti N, Marinovich M. Perspectives on neuroinflammation and excitotoxicity: a neurotoxic conspiracy? Neurotoxicology. 2014;43:10–20. pmid:24662010
- 31. Evans HL. Markers of neurotoxicity: from behavior to autoantibodies against brain proteins. Clin Chem. 1995;41(12 Pt 2):1874–81. pmid:7497648
- 32. Galic MA, Riazi K, Pittman QJ. Cytokines and brain excitability. Front Neuroendocrinol. 2012;33(1):116–25. pmid:22214786
- 33. Hong X, Wang X. Early life precursors, epigenetics, and the development of food allergy. Seminars in immunopathology. 2012;34(5):655–69. pmid:22777545
- 34. Wang D, Liu X, Zhou Y, Xie H, Hong X, Tsai HJ, et al. Individual variation and longitudinal pattern of genome-wide DNA methylation from birth to the first two years of life. Epigenetics: official journal of the DNA Methylation Society. 2012;7(6):594–605.
- 35. Wang X, Zuckerman B, Pearson C, Kaufman G, Chen C, Wang G, et al. Maternal cigarette smoking, metabolic gene polymorphism, and infant birth weight. Jama. 2002;287(2):195–202. pmid:11779261
- 36. Liu X, Wang G, Hong X, Tsai HJ, Liu R, Zhang S, et al. Associations between gene polymorphisms in fatty acid metabolism pathway and preterm delivery in a US urban black population. Human genetics. 2012;131(3):341–51. pmid:21847588
- 37. Sjodin A, McGahee EE 3rd, Focant JF, Jones RS, Lapeza CR, Zhang Y, et al. Semiautomated high-throughput extraction and cleanup method for the measurement of polybrominated diphenyl ethers and polybrominated and polychlorinated biphenyls in breast milk. Anal Chem. 2004;76(15):4508–14. pmid:15283595
- 38. Matoba N, Yu Y, Mestan K, Pearson C, Ortiz K, Porta N, et al. Differential patterns of 27 cord blood immune biomarkers across gestational age. Pediatrics. 2009;123(5):1320–8. pmid:19403498
- 39. Sullivan KE, Reddy AB, Dietzmann K, Suriano AR, Kocieda VP, Stewart M, et al. Epigenetic regulation of tumor necrosis factor alpha. Molecular and cellular biology. 2007;27(14):5147–60. pmid:17515611
- 40. Liu X, Chen Q, Tsai HJ, Wang G, Hong X, Zhou Y, et al. Maternal preconception body mass index and offspring cord blood DNA methylation: exploration of early life origins of disease. Environmental and molecular mutagenesis. 2014;55(3):223–30. pmid:24243566
- 41. Golikova EA, Lopatnikova JA, Kovalevskaya-Kucheryavenko TV, Nepomnyashih VM, Sennikov SV. Levels of TNF, TNF autoantibodies and soluble TNF receptors in patients with bronchial asthma. The Journal of asthma: official journal of the Association for the Care of Asthma. 2013;50(7):705–11.
- 42. Blandizzi C, Gionchetti P, Armuzzi A, Caporali R, Chimenti S, Cimaz R, et al. The role of tumour necrosis factor in the pathogenesis of immune-mediated diseases. International journal of immunopathology and pharmacology. 2014;27(1 Suppl):1–10. pmid:24774503
- 43. Nasiri R, Amirzargar AA, Movahedi M, Hirbod-Mobarakeh A, Farhadi E, Behniafard N, et al. Single-nucleotide polymorphisms of TNFA and IL1 in allergic rhinitis. Journal of investigational allergology & clinical immunology. 2013;23(7):455–61.
- 44. Braumuller H, Wieder T, Brenner E, Assmann S, Hahn M, Alkhaled M, et al. T-helper-1-cell cytokines drive cancer into senescence. Nature. 2013;494(7437):361–5. pmid:23376950
- 45. Goldfeld AE, Tsai E, Kincaid R, Belshaw PJ, Schrieber SL, Strominger JL, et al. Calcineurin mediates human tumor necrosis factor alpha gene induction in stimulated T and B cells. The Journal of experimental medicine. 1994;180(2):763–8. pmid:8046352
- 46. El Gazzar M, Yoza BK, Chen X, Hu J, Hawkins GA, McCall CE. G9a and HP1 couple histone and DNA methylation to TNFalpha transcription silencing during endotoxin tolerance. The Journal of biological chemistry. 2008;283(47):32198–208. pmid:18809684
- 47. Barthel R, Tsytsykova AV, Barczak AK, Tsai EY, Dascher CC, Brenner MB, et al. Regulation of tumor necrosis factor alpha gene expression by mycobacteria involves the assembly of a unique enhanceosome dependent on the coactivator proteins CBP/p300. Molecular and cellular biology. 2003;23(2):526–33. pmid:12509451
- 48. Li Y, Reddy MA, Miao F, Shanmugam N, Yee JK, Hawkins D, et al. Role of the histone H3 lysine 4 methyltransferase, SET7/9, in the regulation of NF-kappaB-dependent inflammatory genes. Relevance to diabetes and inflammation. The Journal of biological chemistry. 2008;283(39):26771–81. pmid:18650421
- 49. Ranjbar S, Rajsbaum R, Goldfeld AE. Transactivator of transcription from HIV type 1 subtype E selectively inhibits TNF gene expression via interference with chromatin remodeling of the TNF locus. Journal of immunology (Baltimore, Md: 1950). 2006;176(7):4182–90.
- 50. Tsytsykova AV, Falvo JV, Schmidt-Supprian M, Courtois G, Thanos D, Goldfeld AE. Post-induction, stimulus-specific regulation of tumor necrosis factor mRNA expression. The Journal of biological chemistry. 2007;282(16):11629–38. pmid:17303559
- 51. Li Z, Chao TC, Chang KY, Lin N, Patil VS, Shimizu C, et al. The long noncoding RNA THRIL regulates TNFalpha expression through its interaction with hnRNPL. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(3):1002–7. pmid:24371310
- 52. Tsytsykova AV, Rajsbaum R, Falvo JV, Ligeiro F, Neely SR, Goldfeld AE. Activation-dependent intrachromosomal interactions formed by the TNF gene promoter and two distal enhancers. Proceedings of the National Academy of Sciences of the United States of America. 2007;104(43):16850–5. pmid:17940009
- 53. Sullivan KE, Suriano A, Dietzmann K, Lin J, Goldman D, Petri MA. The TNFalpha locus is altered in monocytes from patients with systemic lupus erythematosus. Clinical immunology. 2007;123(1):74–81. pmid:17276734
- 54. Hermsdorff HH, Mansego ML, Campion J, Milagro FI, Zulet MA, Martinez JA. TNF-alpha promoter methylation in peripheral white blood cells: relationship with circulating TNFalpha, truncal fat and n-6 PUFA intake in young women. Cytokine. 2013;64(1):265–71. pmid:23796695
- 55. Falvo JV, Jasenosky LD, Kruidenier L, Goldfeld AE. Epigenetic control of cytokine gene expression: regulation of the TNF/LT locus and T helper cell differentiation. Advances in immunology. 2013;118:37–128. pmid:23683942
- 56. Almasry SM, Eldomiaty MA, Elfayomy AK, Habib FA. Expression pattern of tumor necrosis factor alpha in placentae of idiopathic fetal growth restriction. Journal of molecular histology. 2012;43(3):253–61. pmid:22461198
- 57. Houseman EA, Accomando WP, Koestler DC, Christensen BC, Marsit CJ, Nelson HH et al. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC bioinformatics. 2012;13:86 pmid:22568884
- 58. Toperoff G, Aran D, Kark JD, Rosenberg M, Dubnikov T, Nissan B et al. Genome-wide survey reveals predisposing diabetes type 2-related DNA methylation variations in human peripheral blood. Hum Mol Genet. 2012;21(2):371–83. pmid:21994764
- 59. Reinius LE, Acevedo N, Joerink M, Pershagen G, Dahlén SE, Greco D et al. Differential DNA methylation in purified human blood cells: implications for cell lineage and studies on disease susceptibility. PLoS One. 2012;7(7):e41361. pmid:22848472
- 60. Singh AK, Kashyap MP. An Overview on Human Umbilical Cord Blood Stem Cell-Based Alternative In Vitro Models for Developmental Neurotoxicity Assessment. Mol Neurobiol. 2015. [Epub ahead of print].
- 61. Stock W, Hoffman R.White blood cells 1: non-malignant disorders. Lancet. 2000. 355: 1351–57. pmid:10776761
- 62. Sanders JM, Lebetkin EH, Chen LJ, Burka LT. Disposition of 2,2',4,4',5,5'-hexabromodiphenyl ether (BDE153) and its interaction with other polybrominated diphenyl ethers (PBDEs) in rodents. Xenobiotica. 2006;36(9):824–37. pmid:16971346
- 63. Staskal DF, Diliberto JJ, DeVito MJ, Birnbaum LS. Toxicokinetics of BDE 47 in female mice: effect of dose, route of exposure, and time. Toxicological sciences: an official journal of the Society of Toxicology. 2005;83(2):215–23.
- 64. Staskal DF, Hakk H, Bauer D, Diliberto JJ, Birnbaum LS. Toxicokinetics of polybrominated diphenyl ether congeners 47, 99, 100, and 153 in mice. Toxicological sciences: an official journal of the Society of Toxicology. 2006;94(1):28–37.
- 65. Emond C, Sanders JM, Wikoff D, Birnbaum LS. Proposed mechanistic description of dose-dependent BDE-47 urinary elimination in mice using a physiologically based pharmacokinetic model. Toxicology and applied pharmacology. 2013;273(2):335–44. pmid:24055880