Dioxin-like compound exposures and DNA methylation in the Anniston Community Health Survey Phase II

https://doi.org/10.1016/j.scitotenv.2020.140424Get rights and content

Highlights

  • Anniston, Alabama has high levels of environmental dioxin-like compounds.

  • Using 850K arrays, we found 10 genome-wide significant exposure/DNA CpG methylation associations.

  • Seven methylation associations were with polychlorinated dibenzo-p-dioxins.

  • Four methylation associations had an absolute differential methylation ≥1.00%.

  • Eight associations were in genes reported to interact with dioxin exposures.

Abstract

The Anniston Community Health Survey (ACHS-I) was initially conducted from 2005 to 2007 to assess polychlorinated biphenyl (PCB) exposures in Anniston, Alabama residents. In 2014, a follow-up study (ACHS-II) was conducted to measure the same PCBs as in ACHS-I and additional compounds e.g., polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and dioxin-like non-ortho (cPCBs) substituted PCBs. In this epigenome-wide association study (EWAS), we examined the associations between PCDD, PCDF, and PCB exposures and DNA methylation. Whole blood DNA methylation was measured using Illumina EPIC arrays (n=292). We modeled lipid-adjusted toxic equivalencies (TEQs) for: ΣDioxins (sum of 28 PCDDs, PCDFs, cPCBs, and mPCBs), PCDDs, PCDFs, cPCBs, and mPCBs using robust multivariable linear regression adjusting for age, race, sex, smoking, bisulfite conversion batch, and estimated percentages of six blood cell types. Among all exposures we identified 10 genome-wide (Bonferroni p≤6.74E−08) and 116 FDR (p≤5.00E−02) significant associations representing 10 and 113 unique CpGs, respectively. Of the 10 genome-wide associations, seven (70%) occurred in the PCDDs and four (40%) of these associations had an absolute differential methylation ≥1.00%, based on the methylation difference between the highest and lowest exposure quartiles. Most of the associations (six, 60%) represented hypomethylation changes. Of the 10 unique CpGs, eight (80%) were in genes shown to be associated with dioxins and/or PCBs based on data from the 2019 Comparative Toxicogenomics Database. In this study, we have identified a set of CpGs in blood DNA that may be particularly susceptible to dioxin, furan, and dioxin-like PCB exposures.

Introduction

The major US producer of polychlorinated biphenyls (PCBs) operated in Anniston, Alabama from 1929 until 1971. Due to poor environmental controls at the facility, releases of dioxins, furans, and PCBs over time caused wide-spread environmental contamination in the Anniston community. Because these compounds are highly lipophilic and have long half-lives, they bioaccumulate in the food chain and persist in human tissues. The Anniston Community Health Survey (ACHS-I) was established in 2005 to assess the health effects of PCB exposure in the general population of the community (Birnbaum and Staskal-Wikoff, 2010; Pavuk et al., 2014a) and in 2014, a follow-up study was conducted (ACHS-II) (Birnbaum et al., 2016). Prior ACHS-I research has found links between PCB exposures and hypertension (Goncharov et al., 2010), blood pressure (Goncharov et al., 2011), race (Pavuk et al., 2014b), liver disease (Clair et al., 2018), metabolic syndrome (Rosenbaum et al., 2017), serum lipid levels (Aminov et al., 2014), diabetes (Silverstone et al., 2012), and leukocyte telomere length (Callahan et al., 2017).

A number of studies have found associations between altered DNA methylation and dioxin, furan, and PCB exposures in several populations, including Faroe Islanders (Leung et al., 2018), Dutch (van den Dungen et al., 2017), Taiwanese (Su et al., 2019), Greenlandic Inuits (Rusiecki et al., 2008), Koreans (Kim et al., 2010), and Japanese (Kobayashi et al., 2017). We have recently reported on 28 genome-wide significant associations between serum PCB concentrations in ACHS-I and whole blood DNA methylation (Pittman et al., 2019). We carried out a look-up analysis of the 369 FDR significant ACHS-I associations in the ACHS-II methylation dataset reported in the present project and determined that the most significantly altered ACHS-I CpG, cg00475490 in the PRSS23 gene, was still significantly associated with the group, summed tri/tetra-ortho substituted, nondioxin-like PCBs (p = 1.33E−04) in ACHS-II.

The current ACHS-II study focuses on dioxins and dioxin-like compounds and their toxicities have generally been attributed to sustained aryl hydrocarbon receptor (AHR) activation as measured in model systems, and this property is used in human risk assessment (Theobald and Peterson 1994). Ligand-mediated AHR activation results in upregulation of genes in the AHR pathway, including cytochrome P450s (e.g. CYP1A1 and CYP1B1) and the aryl hydrocarbon receptor repressor (AHRR), a negative regulator of AHR (Hahn et al., 2009). DNA methylation levels at the CpG cg05575921 in AHRR are strongly associated with adult and prenatal tobacco smoke exposure (Joehanes et al., 2016; Joubert et al., 2012; Reynolds et al., 2017; Su et al., 2016; Wan et al., 2018), and this could potentially be related to tobacco smoke polyaromatic hydrocarbons (PAHs), which are also AHR ligands. We recently explored the hypothesis that PCB exposures in ACHS might be associated with whole blood DNA methylation levels in AHRR or other AHR pathway genes, but found no relationship (Pittman et al., 2019). Dioxins are considered stronger ligands for AHR (Theobald and Peterson, 1994), so in this work we also tested if dioxin and dioxin-like exposures might lead to methylation changes in AHRR cg05575921. Because dioxin exposures have been linked with altered immune system effects, we used the Houseman model (Houseman et al., 2012; Houseman et al., 2016) to test if exposure was associated with changes in methylation-based estimated cell-type percentages in whole blood. However, the primary objective was to carry out an epigenome-wide association study (EWAS) assessing if exposures to AHR ligands such as PCDDs, PCDFs, cPCBs, and mPCBs were associated with altered whole blood DNA methylation at CpGs across the genome.

Section snippets

Study population

ACHS-II was conducted in 2014 as a follow-up study to ACHS-I (2005–2007). Both study designs have been previously reported (Birnbaum et al., 2016; Pavuk et al., 2014a). In ACHS-I, two-stage stratified random sampling was used to select households and adults within the household of Anniston, Alabama. We contacted over 1800 households; individuals living in west Anniston (the area closest to the PCB manufacturing facility) were over-sampled (two-thirds of eligible participants). 1100 participants

Demographics

Table 1 provides the demographics for subjects with available DNA methylation and exposure data (n = 292). The cohort was primarily female (74.3%) and had a higher mean BMI (31.6 ± 0.5) relative to that reported (26.0 ± 0.1) in the National Health and Nutrition Examination Survey 2015–2016 (CDC National Center for Health Statistics, 2018). White study participants were older than African-American participants (64.3 vs 60.9 years). African-American participants were more likely to live in west

Look up analysis

We searched the literature for array-based methylation data and associations with dioxins, furans and dioxin-like PCBs. We identified three studies with 450 k methylation array data (Leung et al., 2018; Su et al., 2019; van den Dungen et al., 2017). These studies examined the associations between whole blood DNA methylation and PCBs 105 and 118; 23,478-PeCDF, 123,478-HxCDF, PCBs 153 and 156; and PCBs 28, 52, 105, 118, 138, 153, 156, 170, 180, and 187, respectively (Supplemental File 1 Table

Discussion

In this current study we have examined the associations between dioxin, furan, and dioxin-like PCB exposures and whole blood DNA methylation in the second phase of the Anniston Community Health Study. Among all exposures we identified 10 (Bonferroni p ≤ 6.74E−08) and 116 (FDR p ≤ 5.00E−02) significant associations representing 10 and 113 unique CpGs, respectively. The PCDDs exposure group had the most Bonferroni and FDR significant CpG associations (seven and 107, respectively), and the

Conclusions

This study contributes to the understanding of the effects of dioxin and dioxin-like exposures on whole blood DNA methylation in a human population study. This is the first report to examine dioxin-like exposures and alterations in DNA methylation using the Illumina 850 K methylation array. We identified one highly significant DMR and 10 genome-wide significant CpG/exposure associations with 40% having differential methylation ≥1.00%, based on the methylation difference between the highest and

Author credit statement

GSP: Methodology, Formal data analysis, Visualization, Writing – Original Draft, review and editing.

XW: Formal data analysis.

MRC: Sample preparation, Investigation.

SJC: Sample preparation.

JRO: Resources.

MP: Conceptualization, Resources, Writing- Review and Editing, Project Administration.

LSB: Conceptualization, Resources, Writing – Review and Editing, Funding.

DAB: Conceptualization, Supervision, Resources, Writing- Review and Editing, Project Administration.

Funding details

The baseline study (ACHS-I) was conducted using a grant from ATSDR to Jacksonville State University, #5U50TS473215. The follow up study (ACHS II) was funded by the National Cancer Institute through interagency agreements with the Centers for Disease Control and Prevention (CDC) (IAA#: 11-AT1-001-00; IAA#: 12-AT-12-ANNISTON) and by ATSDR. Data collection was supported via contract from ATSDR to the University of Alabama at Birmingham (UAB) (CDC Contract No. 200-2011-40834). The contents of this

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

J.R. Olson served as an expert witness for the plaintiffs in legal actions regarding the residents of Anniston, Alabama being exposed to PCBs. The other authors declare that they have no competing interests.

Acknowledgements

The coauthors would like to dedicate this work to the memory of Gary S. Pittman, MSPH, National Institute of Environmental Health Sciences, who died unexpectedly after a short illness just as the final edits on this manuscript were completed. We recognize Dr. Stephen T. Mennemeyer of University of Alabama Birmingham for his study contributions. We express gratitude to the study participants in the ACHS-I and ACHS-II cohorts. We thank Andreas Sjödin and Richard Jones, National Center for

References (58)

  • P.F. Rosenbaum et al.

    Metabolic syndrome is associated with exposure to organochlorine pesticides in Anniston, AL, United States

    Env. Int

    (2017)
  • K.Y. Su et al.

    Perinatal polychlorinated biphenyls and polychlorinated dibenzofurans exposure are associated with DNA methylation changes lasting to early adulthood: findings from Yucheng second generation

    Environ. Res.

    (2019)
  • P. Viatour et al.

    GSK3-mediated BCL-3 phosphorylation modulates its degradation and its oncogenicity

    Mol. Cell

    (2004)
  • E. Yang et al.

    Exposure of dioxin-like chemicals in participants of the Anniston community health survey follow-up

    Sci Total Env

    (2018)
  • M.J. Aryee et al.

    Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium DNA methylation microarrays

    Bioinformatics

    (2014)
  • L.S. Birnbaum et al.

    Anniston community health survey: follow-up and dioxin analyses (ACHS-II)—methods

    Env. Sci Pollut Res Int

    (2016)
  • C.E. Breeze et al.

    Eforge v2.0: updated analysis of cell type-specific signal in epigenomic data

    Bioinformatics

    (2019)
  • CDC National Center for Health Statistics

    National Health and Nutrition Examination Survey: Analytic Guidelines, 2011–2014 and 2015–2016

    (2018)
  • E.Y. Chen et al.

    Enrichr: interactive and collaborative HTML5 gene list enrichment analysis tool

    BMC Bioinformatics

    (2013)
  • H.B. Clair et al.

    Liver disease in a residential cohort with elevated polychlorinated biphenyl exposures

    Toxicol. Sci.

    (2018)
  • C. Consales et al.

    Exposure to persistent organic pollutants and sperm DNA methylation changes in Arctic and European populations

    Env. Mol Mutagen

    (2016)
  • A.P. Davis et al.

    The comparative toxicogenomics database: update 2019

    Nucleic Acids Res.

    (2019)
  • M.W. van den Dungen et al.

    Association between DNA methylation profiles in leukocytes and serum levels of persistent organic pollutants in Dutch men

    Env. Epigenet

    (2017)
  • M.D. Ericksson

    PCBs: Recent Advances in Environmental Toxicology and Health Effects

    (2001)
  • J.P. Fortin et al.

    Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi

    Bioinformatics

    (2017)
  • A. Goncharov et al.

    Blood pressure and hypertension in relation to levels of serum polychlorinated biphenyls in residents of Anniston, Alabama

    J. Hypertens.

    (2010)
  • A. Goncharov et al.

    Blood pressure in relation to concentrations of PCB congeners and chlorinated pesticides

    Env. Heal. Perspect

    (2011)
  • M.A. Harris et al.

    The gene ontology (go) database and informatics resource

    Nucleic Acids Res.

    (2004)
  • R.W. Hornung et al.

    Estimation of average concentration in the presence of nondetectable values

    Appl. Occup. Environ. Hyg.

    (1990)
  • Cited by (5)

    • Epigenetic alterations induced by genotoxic occupational and environmental human chemical carcinogens: An update of a systematic literature review

      2022, Mutation Research - Reviews in Mutation Research
      Citation Excerpt :

      Global methylation levels were significantly decreased following dioxin-like PCB exposure within several cross-sectional human in vivo studies, demonstrating that exposure to carcinogenic environmental pollutants induces change in the methylome [192,218]. Further investigation into the epigenetic modifications associated with exposures to dioxin-like PCBs in humans showed significant changes in CpG methylation [219]. In contrast to human in vivo results, human carcinoma cell lines (HepG2 and HeLa) treated with PCB 126 showed no apparent differences within methylation levels of specific CpG locations following exposure [220].

    1

    Deceased

    View full text