Differential DNA Methylation by Hispanic Ethnicity Among Firefighters in the United States

Abstract

Firefighters are exposed to a variety of environmental hazards and are at increased risk for multiple cancers. There is evidence that risks differ by ethnicity, yet the biological or environmental differences underlying these differences are not known. DNA methylation is one type of epigenetic regulation that is altered in cancers. In this pilot study, we profiled DNA methylation with the Infinium MethylationEPIC in blood leukocytes from 31 Hispanic white and 163 non-Hispanic white firefighters. We compared DNA methylation (1) at 12 xenobiotic metabolizing genes and (2) at all loci on the array (>740 000), adjusting for confounders. Five of the xenobiotic metabolizing genes were differentially methylated at a raw P-value <.05 when comparing the 2 ethnic groups, yet were not statistically significant at a 5% false discovery rate (q-value <.05). In the epigenome-wide analysis, 76 loci exhibited DNA methylation differences at q < .05. Among these, 3 CpG sites in the promoter region of the biotransformation gene SULT1C2 had lower methylation in Hispanic compared to non-Hispanic firefighters. Other differentially methylated loci included genes that have been implicated in carcinogenesis in published studies (FOXK2, GYLTL1B, ZBTB16, ARHGEF10, and more). In this pilot study, we report differential DNA methylation between Hispanic and non-Hispanic firefighters in xenobiotic metabolism genes and other genes with functions related to cancer. Epigenetic susceptibility by ethnicity merits further study as this may alter risk for cancers linked to toxic exposures.

Keywords

Occupational health epigenome-wide analysis study xenobiotic metabolism health disparities occupational exposures

Introduction

Firefighters are a unique occupational group exposed to a variety of hazards including chemical exposures and physical and mental stressors. Exposure to combustion byproducts from fires is a major concern since a number are known or suspected human carcinogens.^1-3 Firefighters have an increased overall risk of cancer incidence and mortality compared to the general population, as well as increased incidence and mortality for specific cancers, including those of the digestive tract, bladder, prostate, testicles, thyroid, and blood, among others.^4-7

Ethnic and racial differences in firefighter cancer rates have been reported in the limited number of epidemiologic studies with a sufficient number of minority firefighters. In a cancer registry-based case control study in California, rates of 6 cancers (tongue, testicular, bladder, non-Hodgkin’s lymphoma, chronic leukocytic leukemia, and chronic myelogenous leukemia) were significantly elevated compared to the general population only in minority firefighters, most of whom were either Hispanic (62.2%) or black (27.7%).⁷ In addition, non-minority and minority firefighters both had significantly elevated risks for 6 other cancers (melanoma, prostate, kidney, brain, multiple myeloma, and overall leukemia). In US civilian populations, adults of Hispanic ethnicity have been found to have a lower incidence of lung cancer, but an increase in cervical, gall bladder, liver, and gastric cancer^8,9 compared to non-Hispanic whites.

The reasons for the observed cancer disparities among firefighters by race and ethnicity is not known. Potential explanations include but are not limited to differential occupational exposures, such as assignment to busier stations or differential environmental exposures (eg, from place of residence or diet). Of interest to this study, epigenetic or genetic differences between groups may also alter susceptibility to the impacts of exposures. Polycyclic aromatic hydrocarbons (PAHs) are a family of chemicals including known and suspected carcinogens which are produced during combustion. PAHs have a ubiquitous presence in the fire service, having been documented at the fireground during suppression, during overhaul activities, on firefighter personal protective equipment, and at fire stations.^10-13 Additionally, biomonitoring studies have observed significant PAH exposures across various firefighters and fire service personnel after fire events, regardless of job assignment.^2,14,15 Occupational exposure to mixtures of PAHs and other chemicals could contribute to the increased risk for lung, skin, and prostate cancers,^16-18 which are also increased in firefighters compared to the general population.^4,5,7,19

Epigenetic and genetic differences that alter the body’s ability to detoxify carcinogenic exposures, such as but not limited to PAHs, may modify susceptibility to cancers among firefighters with similar exposures. The epigenome consists of modifications to DNA and chromatin that do not alter the underlying DNA sequence yet are heritable, at least across cell divisions. DNA methylation is one major type of epigenetic regulation that is fairly stable across time and is typically associated with repression of gene expression.²⁰ It has been shown in adults to be responsive to a multitude of environmental factors including PAHs, other toxicant exposures,^21-24 and stress.²⁵ We have previously shown differential blood DNA methylation in incumbent compared to new recruit firefighters, adjusted for age, and other covariates, as well as differential whole blood microRNA expression.^26,27 DNA methylation has been shown to vary based on sex,^28,29 race,³⁰ and ethnicity³¹ at many genes. This may be due to differences in the underlying genetic sequence^32,33 and/or disproportionate exposures to toxicants or psychosocial stressors. In particular, one cross-sectional study showed that over 70% of the variance in DNA methylation between ethnic groups was explained by shared genomic ancestry.³⁴ Interestingly, they identified that CpG sites that were previously reported to be responsive to exogenous exposures were enriched among the loci associated with ethnicity.³⁴ Additionally, although exposures to smoking and other combustion byproducts,³⁵ race/ethnicity,⁸ and the epigenome^36-38 have all been associated with cancer, whether and how the intersection of the 3 relates to cancer risk is unclear.

In this pilot study, we profiled and compared DNA methylation in white firefighters of Hispanic (n = 31) and non-Hispanic (n = 163) ethnicity. We hypothesized that genes differentially methylated between the ethnic groups would have functions relevant to carcinogenesis. We quantified DNA methylation at >850 000 cytosine-guanine dinucleotides (CpG sites) via the Infinium MethylationEPIC³⁹ in blood leukocyte DNA. We utilized a two-tiered analysis approach. We first compared DNA methylation levels in 12 genes relevant to metabolism and detoxification of PAHs—exposures among firefighters that are known to be carcinogenic. Second, we compared DNA methylation at all CpG sites included on the EPIC array to discover genes in other pathways. We then report on genes that are differentially methylated according to ethnicity.

Methods

Cohort recruitment and study population

This study included participants from 2 larger cancer prevention studies: the first a 3-year research project working in partnership with the Tucson Fire Department (TFD) and the second the prospective multicenter Fire Fighter Cancer Cohort Study (FFCCS) involving multiple fire departments and universities as well as the National Institute for Occupational Safety and Health (NIOSH). Firefighters were recruited for this analysis from 5 fire departments in Arizona (from 2016 to 2018), California (2019), and Massachusetts (2018), 2 of which were volunteer fire departments. Inclusion criteria for enrollment in the current study included being an active duty firefighter (including emergency medical responder) responding to fires as part of normal duties. All study procedures were approved by the institutional review boards (IRB) of the University of Arizona (IRB approval No. 1509137073) and the University of Miami (IRB approval No. 20170997).

The research team delivered an in-depth explanation of the study design, including potential risks and responsibilities, before all subjects provided informed consent. The survey questions involved standard demographic information plus general information such as previous cancer diagnosis, body weight, height, working duration as firefighters (including at current and previous departments), and tobacco use. Blood samples for DNA methylation measurements were collected during the day by qualified phlebotomists in one 6.0 ml dipotassium ethylene diaminetetraacetic acid (K₂EDTA) tube (Beckton, Dickinson and Company, Franklin Lakes, NJ). Blood was stored frozen (temporarily at −20°C followed by long-term storage at −80°C) until use.²⁷

DNA methylation analysis

DNA was isolated from blood leukocytes. Concentration of double stranded DNA was measured via a QuantiFluor dsDNA System (Promega) or a Qubit Fluorometer (Thermo Scientific). DNA was bisulfite converted using Zymo kits. DNA methylation was quantified at >850 000 CpG sites throughout the genome using the Infinium MethylationEPIC array.³⁹ Samples were randomized and hybridized to chips during each batch, and personnel running the analysis were blinded to phenotype information about the samples (eg, ethnic group). Samples were run in 3 batches (one at the University of Utah DNA Sequencing and Genomics Core Facility, and 2 batches at the University of Michigan Advanced Genomics Core) and scanned by experienced personnel. Both Core facilities follow the same recommended protocol for the MethylationEPIC kit and scan the arrays using iScan instruments (Illumina, Inc). All batches included samples from both Hispanic and non-Hispanic firefighters. Raw image files were read with the R package minfi,⁴⁰ and quality control and normalization occurred using the package ENmix.⁴¹ Quality control included comparing recorded sex to estimated sex based on signals on X and Y chromosomes, examining median intensities, bisulfite conversion efficiency, and more. Probes were removed if at least 5% of samples were not detected (P-value >1e⁻¹⁶ compared to background), and samples with >5% of probes not detected were excluded. Background correction was performed with noob and dye bias correction with Regression on Logarithm of Internal Control probes (RELIC). Quantile normalization was used to normalize intensities separately for methylated and unmethylated for type I and type II probes.⁴² Probes that are known to be cross-reactive, have SNPs in the CpG or single-base extension site, or are on X and Y chromosomes were excluded. The final probe number used in downstream analyses covered 740 842 CpG sites; probes were annotated using the ilm10b4.h19 annotation R package. Analysis included 194 samples from white Hispanic or white non-Hispanic active firefighters with data passing all quality control.

The proportion of 6 cell types were estimated using reference data from sorted blood cells according to the algorithm established by Houseman et al.^43,44 Surrogate variable analysis was performed using the intensity values from the non-negative control probes to create variables representing technical variation influencing the DNA methylation data.⁴⁵ Three principal components (PCs) from this analysis explained 92% of the technical variance in the data and were used as covariates in downstream models.

Statistical analysis

All data pre-processing and statistical analyses were conducted in the R Project for Statistical Computing (version ⩾3.6.3). Descriptive statistics were first calculated for all demographic variables and compared between Hispanic and non-Hispanic firefighters using t-tests or chi-square tests as appropriate. In all described analyses, data were only included from self-identified white Hispanic and white non-Hispanic individuals. Singular Value Decomposition (SVD) analysis was performed with the ChAMP package in R to identify technical and biological covariates that correlate with variation in the DNA methylation data. Briefly, the correlation between PCs of the methylation data with covariates was determined using linear regression for continuous variables or Kruskal-Wallis for categorical variables.³⁸ Covariates considered included age, gender, ethnicity, BMI, estimated cell type proportions, smoking history (categorized as never smoker, former smoking, and not reported), and PCs explaining technical variation from the SVD. Based on this analysis, gender, age, cell types (represented by proportion of neutrophils), and 3 PCs for technical variation were selected for statistical models along with the independent variable of interest (ethnicity). BMI was not associated with DNA methylation, but was slightly higher in the Hispanic group (Table 1) and is considered in an additional model. We also ran models including adjustment for smoking history or total years working as a firefighter.

Table 1.

Study population characteristics by ethnicity.

	Hispanic		Non-hispanic
	n (%)	Mean (SD)	n (%)	Mean (SD)	P-value^c
Males	27 (87.1)		145 (89.0)		1
Females	4 (12.9)		18 (11.0)
BMI (kg m⁻²)^a		27.9 (3.6)		26.8 (3.5)	.16
Age (y)		38.4 (9.6)		38.6 (9.8)	.92
Years firefighting		13.6 (9.1)		15.2 (9.2)	.37
Smoking history
Never	23 (74.2)		122 (74.8)		.99
Past	5 (16.1)		25 (15.3)
Not reported	3 (9.7)		16 (9.8)
Estimated cell type proportions (%)
Neutrophils		58.9 (8.3)		57.2 (9.0)	.32
CD4+ T cells		12.7 (3.2)		13.8 (4.9)	.11^b
CD8+ T cells		10.0 (4.0)		10.6 (4.4)	.51
B cells		5.7 (2.4)		5.2 (2.4)	.32
Monocytes		7.5 (2.6)		8.0 (2.6)	.34
NK cells		6.3 (3.4)		6.1 (2.7)	.83

Some subjects were missing BMI. Thirty Hispanic and 158 non-Hispanic had BMI. For all other variables there are 31 Hispanic and 163 non-Hispanic participants. All participants are white.

Welch’s t-test used due to unequal variances.

t-Test for continuous variables; chi square test for categorical variables.

We used a 2-tiered approach: (1) a hypothesis-driven investigation of genes known to influence response to environmental exposures with links to cancer risk; and (2) an exploratory approach investigating all loci available from the EPIC data. In the hypothesis driven approach, we included 187 CpG sites that annotated to 12 genes according to UCSC “refGene”: CYP1A1, CYP1B1, EPHX1, GSTA1, GSTM1, GSTP1, GSTT1, PTGS1, PTGS2, UGT1A1, UGT1A6, and UGT1A9. These genes were selected because they are involved in the metabolism and detoxification of PAHs, which are one set of exposures firefighters face in the line of duty. In the exploratory epigenome-wide approach, we included 740 842 CpG sites that passed quality control.

For the hypothesis-driven and exploratory approaches, linear regression was used to identify CpG sites differentially methylated between the ethnic groups. Models were fit to beta values (which represent the proportion methylated) for each CpG site separately. An empirical Bayes method in the limma R package^46,47 was then used to shrink probe-wise variances towards a pooled estimate and calculate a moderated t-statistic prior to significance calling. For both approaches, CpG sites associated with ethnicity (Hispanic vs non-Hispanic) at a false discovery rate (FDR) adjusted P-value <.05 (referred to as q-value) are considered statistically significant.⁴⁸ Model 3 is considered the main model with all necessary potential confounders (eg, covariates associated with both DNA methylation and ethnicity). However, we also compared results with four other models (unadjusted, cell-free model, and models including additional adjustment for BMI, smoking history, or years firefighting):

Model 1 = ethnicity only (unadjusted model)

Model 2 = ethnicity + gender + age + PC1 + PC2 + PC3 (no-cell model)

Model 3 = Model 2 + Neutrophils (fully adjusted model)

Model 4 = Model 3 + BMI (n = 188 since some are missing BMI)

Model 5 = Model 3 + Smoking history

Model 6 = Model 3 + Years working as a firefighter

We performed a pathway analysis with the results from Model 3 to determine whether the top 10 000 CpG sites by raw P-value were enriched in certain biological or functional pathways. We used the gometh function in the missMethyl package, and ran pathway analysis separately for Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) terms.⁴⁹

Results

We profiled DNA methylation in blood leukocytes from 194 active US firefighters with an average of 15.0 (SD ± 9.2) total years of employment in the fire service. The firefighters were recruited from Massachusetts (n = 7), Southern California (n = 46), and Southern Arizona (n = 141). Of the 194, 188 were career firefighters and 6 (all from Arizona) were volunteer firefighters. Thirty-one of the firefighters included in this analysis identified as white Hispanic (16%) and the rest were white non-Hispanic. The average age was 38 years for both Hispanic and non-Hispanic firefighters. Smoking history, BMI, % female, years in the service, and estimated cell type proportions from the blood samples were similar among firefighters of both ethnicities (Table 1).

Hypothesis-driven approach

The EPIC array contained 187 CpG sites that annotated to 12 genes involved in PAH metabolism and detoxification. None of the CpG sites were associated with ethnicity using a 5% FDR significance cut-off (q < .05). Sixteen CpG sites in 5 of the genes were associated with ethnicity at an unadjusted P-value <.05 (Supplemental Table S1). These CpG sites, annotated to CYP1A1, CYP1B1, EPHX1, PTGS1, and UGT1A6 merit further exploration in larger multi-ethnic studies.

Epigenome-wide approach

In the main statistical model of all CpG sites, 54 had lower methylation in Hispanic firefighters compared to non-Hispanic firefighters and 22 had higher methylation (q < .05; Table 2). Effect sizes and significance values were similar in sensitivity analyses including in models without cell type adjustment and with adjustment for BMI, smoking history, or years working as a firefighter (Supplemental Table S2). Of the 76 significant loci, 21 were in intergenic regions and the remainder annotated to 46 genes. Effect estimates ranged from −15.2% lower in Hispanic firefighters (at chr1:210099448 in an intergenic region) to 19.1% higher (near the transcription start site of the CCDC66 gene; Table 2) compared with non-Hispanic participants. Two genes had 3 significant CpG sites each—FOXK2 and SULT1C2; the latter is a xenobiotic metabolizing enzyme (Figure 1). Five genes had 2 significant CpG sites each—ZBTB16, GYLTL1B, ARHGEF10, ZC2HC1A, and CRB2.

Table 2.

Differentially methylated CpG sites in Hispanic compared to non-Hispanic firefighters in the epigenome-wide analysis, fully adjusted model (q-value <.05).

ProbeID	Position	Gene^a	Effect estimate	SE	Average % methylation at CpG site	P-value	q-value
cg12545480	chr1: 210099448	NA	−0.152	0.018	61.0	2.88E−15	2.13E−09
cg04800768	chr17: 80545544	FOXK2	−0.009	0.001	98.4	3.57E−12	1.32E−06
cg24582990	chr5: 154248355	CNOT8	−0.082	0.011	86.7	1.98E−11	4.88E−06
cg12289926	chr11: 68153810	LRP5	−0.050	0.008	83.3	2.77E−10	5.13E−05
cg26748898	chr17: 74912327	MGAT5B	−0.079	0.013	70.8	2.90E−09	.000
cg04074945	chr11: 46071833	PHF21A	0.058	0.009	68.5	2.76E−09	.000
cg13227551	chr6: 144011222	PHACTR2	0.120	0.020	18.3	7.12E−09	.001
cg02796939	chr17: 80545125	FOXK2	−0.075	0.012	81.7	9.41E−09	.001
cg15386132	chr5: 133796338	NA	0.035	0.006	72.4	1.61E−08	.001
cg13968390	chr2: 108904812	SULT1C2	−0.138	0.024	77.2	1.90E−08	.001
cg12063639	chr9: 130184057	NA	−0.135	0.024	91.4	4.80E−08	.003
cg04238311	chr11: 45945630	GYLTL1B	−0.097	0.017	61.7	4.42E−08	.003
cg13962846	chr1: 7973298	NA	−0.134	0.024	75.1	1.20E−07	.006
cg09513263	chr17: 70283945	NA	−0.078	0.014	68.4	1.10E−07	.006
cg03142846	chr8: 1890946	ARHGEF10	−0.043	0.008	70.2	1.17E−07	.006
cg18912965	chr14: 91526996	RPS6KA5	−0.033	0.006	7.4	1.16E−07	.006
cg05194412	chr2: 137003533	NA	−0.078	0.015	42.0	2.19E−07	.009
cg07846855	chr10: 77845945	C10orf11	−0.025	0.005	86.8	2.20E−07	.009
cg12648759	chr1: 161812851	ATF6	−0.034	0.006	82.7	2.57E−07	.010
cg05401945	chr3: 56590734	CCDC66	0.191	0.036	30.0	2.46E−07	.010
cg26905489	chr7: 83551804	NA	−0.033	0.006	83.3	3.04E−07	.011
cg22139615	chr4: 141048272	MAML3	−0.065	0.012	84.1	3.33E−07	.011
cg23163573	chr2: 108905468	SULT1C2	−0.074	0.014	64.1	3.64E−07	.012
cg01297101	chr17: 37356023	RPL19	0.017	0.003	6.0	3.98E−07	.012
cg14720104	chr6: 118508739	SLC35F1	−0.064	0.012	36.6	4.40E−07	.013
cg12526471	chr11: 113929188	ZBTB16	−0.060	0.012	50.0	5.51E−07	.016
cg07737363	chr3: 53283930	TKT	−0.036	0.007	17.2	6.53E−07	.018
cg01770799	chr16: 85951663	IRF8	−0.039	0.008	81.6	7.20E−07	.019
cg05468483	chr2: 70563880	NA	−0.011	0.002	92.5	7.95E−07	.020
cg11031278	chr11: 113929114	ZBTB16	−0.061	0.012	58.0	1.08E−06	.025
cg03016722	chr9: 92517538	NA	−0.053	0.011	66.4	1.10E−06	.025
cg24846680	chr1: 228362309	C1orf69	−0.020	0.004	96.4	1.07E−06	.025
cg14076011	chr20: 23312790	NA	−0.069	0.014	82.2	1.16E−06	.026
cg02314394	chr5: 646139	CEP72	−0.066	0.013	92.6	1.23E−06	.026
cg24085707	chr17: 79615652	TSPAN10	−0.047	0.009	41.0	1.27E−06	.026
cg03582582	chr7: 139444053	HIPK2	0.038	0.008	67.5	1.25E−06	.026
cg21028142	chr17: 79581711	NPLOC4	0.098	0.020	66.7	1.29E−06	.026
cg04287289	chr16: 89883240	FANCA	0.021	0.004	1.3	1.33E−06	.026
cg20728173	chr3: 108791035	MORC1	0.055	0.011	49.3	1.36E−06	.026
cg08574105	chr9: 126137259	CRB2	−0.028	0.006	61.2	1.47E−06	.027
cg05945266	chr17: 80545020	FOXK2	−0.023	0.005	92.7	1.49E−06	.027
cg19662109	chr21: 45866122	NA	−0.025	0.005	75.4	1.66E−06	.029
cg08544606	chr8: 1891006	ARHGEF10	−0.027	0.005	91.1	2.02E−06	.035
cg14487892	chr2: 64558493	NA	−0.052	0.011	68.4	2.34E−06	.036
cg20674635	chr20: 44640803	MMP9	−0.036	0.007	16.3	2.40E−06	.036
cg16214826	chr2: 1065770	SNTG2	−0.025	0.005	92.7	2.25E−06	.036
cg24132527	chr5: 140019269	TMCO6	0.007	0.001	1.1	2.36E−06	.036
cg09353378	chr17: 34274396	NA	0.008	0.002	2.5	2.30E−06	.036
cg26245086	chr4: 87860957	AFF1	0.068	0.014	7.5	2.34E−06	.036
cg02786370	chr4: 2747928	TNIP2	−0.040	0.008	34.5	2.69E−06	.038
cg07162250	chr1: 7488166	CAMTA1	−0.040	0.008	54.9	2.58E−06	.038
cg02555772	chr16: 1079316	NA	−0.031	0.006	93.6	2.68E−06	.038
cg03364549	chr20: 44622696	NA	−0.030	0.006	91.8	2.72E−06	.038
cg16731079	chr21: 41758562	DSCAM	0.077	0.016	26.9	2.81E−06	.039
cg25838818	chr2: 108905173	SULT1C2	−0.103	0.021	39.6	2.87E−06	.039
cg14843888	chr3: 53530247	CACNA1D	−0.033	0.007	69.8	3.13E−06	.041
cg21140145	chr8: 79577641	ZC2HC1A	−0.098	0.021	32.6	3.28E−06	.042
cg00348244	chr17: 39222923	KRTAP2-4	0.010	0.002	93.6	3.37E−06	.042
cg13518537	chr8: 28618150	NA	0.076	0.016	39.0	3.33E−06	.042
cg07181209	chr5: 153585979	GALNT10	−0.083	0.017	60.6	3.44E−06	.042
cg02625222	chr9: 126135169	CRB2	−0.114	0.024	57.1	3.67E−06	.045
cg08655071	chr1: 209928895	TRAF3IP3	0.029	0.006	45.6	3.74E−06	.045
cg03983883	chr8: 79577618	ZC2HC1A	−0.133	0.028	39.7	3.97E−06	.046
cg02524205	chr6: 167559851	NA	0.136	0.029	54.3	3.96E−06	.046
cg07370361	chr6: 36395612	PXT1	−0.065	0.014	66.6	4.18E−06	.047
cg24021129	chr5: 114408148	NA	−0.014	0.003	88.0	4.12E−06	.047
cg20036516	chr17: 18006550	DRG2	0.017	0.004	86.4	4.22E−06	.047
cg23631229	chr1: 7962850	NA	0.004	0.001	1.2	4.35E−06	.047
cg23320965	chr2: 202850601	NA	−0.095	0.020	69.7	4.68E−06	.048
cg27274564	chr2: 55534947	CCDC88A	−0.054	0.011	92.0	4.66E−06	.048
cg24204282	chr11: 45944920	GYLTL1B	−0.039	0.008	9.5	4.69E−06	.048
cg14023573	chr8: 140971379	TRAPPC9	0.017	0.004	88.0	4.63E−06	.048
cg17079172	chr9: 130265152	LRSAM1	0.023	0.005	79.2	4.50E−06	.048
cg09373597	chr3: 138667775	C3orf72	−0.041	0.009	41.7	4.97E−06	.049
cg15551881	chr9: 123688715	TRAF1	0.053	0.011	19.8	4.97E−06	.049
cg21803548	chr3: 144483077	NA	−0.034	0.007	81.6	5.09E−06	.050

Effect estimates represent the change in proportion methylated in Hispanics (n = 31; compared to non-Hispanics, n = 163). Models adjusted for technical variation, age, gender, and estimated neutrophils.

NA means the CpG site is not within a gene or within a known feature (eg, promoter) of a specific gene.

Figure 1.

DNA methylation at SULT1C2. Three CpG sites near the transcription start site (TSS) of SULT1C2 (probe IDs cg13968390, cg25838818, and cg23163573) had lower blood leukocyte DNA methylation in Hispanic compared with non-Hispanic firefighters. They were statistically significant after adjustment for gender, age, technical variation, and proportion of neutrophils (q-values = .001, .04, and .01). Average DNA methylation for Hispanic (green triangles) and non-Hispanic (blue squares) firefighters at the significant loci (labeled with *) and neighboring CpG sites covered on the EPIC array in SULT1C2 are shown.

In the pathway analysis, the top 25 pathways (P < .005) by raw P-value for GO and also 5 KEGG pathways with raw P-value <.05 are shown in Supplemental Table S3. Processes and biological functions represented in these pathways included cytoskeleton organization and actin binding, sensory perception, ossification, alanine transport, other metabolic-related pathways, and circadian rhythms. However, no concepts were statistically significant using a 5% FDR.

Discussion

In this pilot study of firefighters recruited from 3 US states, we observed differences in DNA methylation comparing Hispanic to non-Hispanic participants at 76 loci, including in a gene involved in xenobiotic metabolism and other genes that may be related to risk for carcinogenesis. These results provide preliminary evidence for an epigenetic mechanism underlying differential cancer risks seen in minority firefighters. In this study, no CpG sites from among 12 xenobiotic metabolizing genes were significantly different at a FDR of 5% in the hypothesis-driven approach. In the exploratory approach, 76 differentially methylated CpG sites by ethnicity were identified (q < .05). These sites were in genes that can activate chemical exposures (SULT1C2)⁵⁰ and genes that have been linked to cancer or cancer-related processes in other studies (FOXK2, GYLTL1B, ZBTB16, MAML3, RPS6KA5, ATF6, ZC2HC1A, and ARHGEF10).^51-58

In the hypothesis-driven approach we focused on genes encoding biotransformation enzymes that are involved in handling chemical exposures to which firefighters are exposed. Cytochrome P450, epoxide hydrolase 1, glutathione-S-transferase, and prostaglandin-endoperoxide synthase are enzymes involved in PAH metabolism and detoxification, and variants and expression of the genes encoding these enzymes are associated with various types of cancer.^18,59,60 In our study, there was no strong evidence for DNA methylation differences at these genes by ethnicity. However, CpG sites within CYP1A1, CYP1B1, EPHX1, PTGS1, and UGT1A6 demonstrated differential methylation at a raw P-value <.05, and should be explored further in larger studies. After activation by cytochrome P450 enzymes, PAH metabolites can bind to DNA and form DNA adducts; this damage can have a carcinogenic effect.⁶¹ These genes may be dysregulated in minority firefighters more often compared to non-minority firefighters, and DNA adduct levels may increase risk for developing cancer at different rates between racial and ethnic groups. For example, although there was no difference in PAH-DNA adduct levels between African American and European American prostate cancer patients,⁶² one study showed elevated DNA adduct levels were significantly associated with prostate cancer only in African Americans.⁶³

In the epigenome-wide analysis, most statistically significant CpG sites had decreased methylation in Hispanic compared to non-Hispanic firefighters. Three CpG sites with decreased methylation were in a CpG island within the gene body of FOXK2, a transcriptional regulator involved in glucose metabolism and autophagy. FOXK2 also has an emerging role in cancer.⁵¹ DNA methylation at 3 CpG sites near the transcription start site of SULT1C2 was 7% to 14% lower in Hispanic compared to non-Hispanic firefighters (probe IDs cg13968390, cg25838818, and cg23163573; Figure 1). SULT1C2 was the only xenobiotic metabolizing gene to be identified in the epigenome-wide approach. This gene encodes an enzyme that activates carcinogenic hydroxylamines (such as N-hydroxy-2-acetylamino-fluorene). In an in vitro study, DNA methylation at one of the same CpG sites (cg13968390) was shown to inversely correlate with gene expression in a lung cell line.⁵⁰ Thus, reduced DNA methylation at this locus may increase SULT1C2 expression in Hispanic firefighters, which would increase the ability of the encoded enzyme to activate its substrates, including carcinogenic chemicals. Importantly, one of the differentially methylated CpG sites is also within an aryl hydrocarbon receptor (AhR) binding region, and AhR activation by exposures is thought to be a mechanism that promotes maintenance of hypomethylation in this region.⁵⁰ Thus, all firefighters would be expected to have reduced methylation at this AhR-binding region of SULT1C2 and increased expression due to AhR activation by occupational exposures (eg, PAHs; Figure 2). If Hispanic firefighters have a lower baseline methylation at this site to begin with due to genetic differences^32-34 or other environmental exposures, this may further increase their risk for SULT1C2-related pathogenesis. Genetic polymorphisms in several sulfotransferase family genes are associated with cancers for which firefighters have elevated rates compared to the general population.⁶⁴ For SULT1C2, a polymorphism in the gene is associated with response to therapy in prostate cancer patients.⁶⁵

Figure 2.

DNA methylation, occupational exposures, and disease risk: SULT1C2 as an example. DNA methylation at 3 CpG sites near the transcription start site of SULT1C2 was lower in Hispanic compared to non-Hispanic firefighters (1). Whether this is due to differences in genetics or life course environmental exposures or (2) due to differences in occupational exposures is not known. One of the differentially methylated CpG sites in SULT1C2 is within an AhR binding region. Activation of the AhR by occupational exposures (eg, PAHs) can also reduce methylation further. In an in vitro study, DNA methylation at the same CpG site inversely correlated with gene expression.⁵⁰ (3) Reduced DNA methylation at this gene may increase expression in Hispanic firefighters, which would increase the ability of SULT1C2 to activate carcinogenic exposures (4). Overall, these differences may lead to carcinogenesis later in life (5). In the figure: DNAm = DNA methylation; AhR = aryl hydrocarbon receptor; solid lines represent relationships with evidence from this or other studies; dashed lines represent hypothesized relationships.

We previously found AhR activation in dermal wipes and urine collected from firefighters following post-fire activities. The urinary AhR activity was somewhat correlated with urinary PAH metabolites, but also seems to be related to other unmeasured exposures, including other dioxin-like compounds.⁶⁶ In addition to the observation that the AhR binding region may be relevant for SULT1C2 methylation in our study sample, a large body of evidence implicates AhR as an important regulator of other genes that we report in Supplemental Table S1, including CYP1A1⁶⁷ and CYP1B1,^68,69 both dioxin-inducible genes that are responsible for detoxification of a wide range of xenobiotic compounds. Interactions between genetics, epigenetics, and environmental exposures at genes like these may play a role in cancer. For example, an in vitro study showed that prostate cancer is associated with differential methylation of CYP1A1, and that the chromatic structure of this gene influences how it responds to dioxin exposure.⁷⁰ In another example, there was an interaction between genetic ancestry and smoking on DNA methylation levels in lung sputum with implications for lung cancer.^71,72

Among genes annotated to CpG sites that differed by ethnicity, several of the genes have been implicated in carcinogenesis. Two CpG sites in the promoter region of ZBTB16 had 6% lower methylation in Hispanic firefighters compared to non-Hispanic. ZBTB16 encodes the promyelocytic leukemia zinc finger protein (PLZF), a transcription factor that has been implicated in acute childhood leukemia.⁷³ Whether PLZF plays a role in chronic leukemia, including leukemias that minority firefighters are at a higher risk for, is currently unknown. Recent studies suggest that ZBTB16 expression also inhibits metastasis of breast and prostate cancers.^55,58 Two CpG sites in the gene body of GYLTL1B had lower methylation (4% and 10%) in Hispanic firefighters. This gene, also known as LARGE2, is a glycosyltransferase that has been implicated in many cancers including chronic lymphocytic leukemia (which is increased in minority firefighters),⁷ renal cell carcinoma, and prostate cancer.^56,57,74,75

Among the differentially methylated genes, RPS6KA5 is a serine/threonine-protein kinase that is important for transcriptional regulation, including for proto-oncogenes c-fos/FOS and c-jun/JUN.⁵³ A CpG site near the transcription start site of RPS6KA5 has 3.3% lower methylation in Hispanic firefighters (q = .006). ATF6 is another transcription factor that may play a role in survival of quiescent proliferative squamous carcinoma cells, and a CpG site in the gene body of ATF6 has 3.4% lower methylation in Hispanic compared to non-Hispanic firefighters (q = .01). Two CpG sites in ARHGEF10 had lower methylation in Hispanic firefighters. ARHGEF10 functions as a tumor suppressor,⁵² and the methylation status of this gene may be responsive to environmental exposures.⁷⁶ ZC2HC1A expression has been implicated in an in vitro study of cancer,⁵⁴ and 2 CpG sites in this gene had lower methylation among Hispanic firefighters.

This study had several strengths and limitations. To assess differences in DNA methylation by Hispanic compared to non-Hispanic ethnicity, we leveraged samples from a cohort of firefighters. The unique study sample, which consisted of healthy white firefighters, enabled us to reduce confounding by occupation, occupational exposures, age, race, health status, and socioeconomic status. Our two-tiered analysis plan included a hypothesis-driven approach and identification of potential new targets via an epigenome-wide approach. While the total sample size was 194, only 16% were Hispanic, and thus statistical power was limited to detect all true associations after correcting for multiple comparisons. We report associations with a q-value <.05, which is a commonly used method to account for multiple testing in epigenome-wide studies. The top 12 loci in the exploratory analysis also remain significant at a P-value <9.4 × 10⁻⁸, which is an alternative cut-off recommended for EPIC data that considers the number of total sites and their intra-individual correlation.⁷⁷ We acknowledge that “Hispanic” is a broad classification that encompasses many ethnic groups and varying genetic ancestry³⁴; results should be followed up in larger cohorts with more granular classification of ethnicity and race and genetic data to estimate ancestry. The majority of study participants were male, and sex-specific differences in DNA methylation by ethnicity could not be assessed due to the sample size. Since the study is cross-sectional, we are uncertain if the differences in DNA methylation by ethnicity predated the study participants’ firefighting exposures, and/or if their exposures contributed to the observed differences. We lacked a control group of non-firefighters to help address this uncertainty. However, adjusting for years firefighting as a crude proxy of exposure did not change the results. Since exposure assessment (eg, for PAHs, etc.) was not performed, we cannot exclude the possibility that Hispanic and non-Hispanic firefighters were unequally exposed to epigenetic-modifying toxicants or stressors in the workplace. Environmental conditions in their past⁷⁸ or present residential environments could also contribute to the differences seen here.

Conclusion

In summary, this pilot study provides preliminary evidence for differences in DNA methylation by ethnicity, including in genes relevant to carcinogenesis, among firefighters who are at increased risk for multiple cancers due to occupational exposures. In the epigenome-wide approach, we report differential methylation at 76 CpG sites by ethnicity (q < .05). Some of the differentially methylated loci are in genes that have functions related to carcinogenesis (eg, GYLTL1B, ZBTB16, ARHGEF10, and RPS6KA5) or xenobiotic metabolism (SULT1C2). Given the increased risk firefighters already have for multiple cancers due to hazardous occupational exposures, it is imperative that we understand additional risk factors, including epigenetic susceptibility, that may alter this long-term risk among all firefighters and also specifically among ethnic or racial groups. Ultimately, the incorporation of such information into environmental or occupational risk assessment should be used to protect the most susceptible individuals, even within worker groups.

Supplemental Material

sj-xlsx-1-gae-10.1177_25168657211006159 – Supplemental material for Differential DNA Methylation by Hispanic Ethnicity Among Firefighters in the United States

Supplemental material, sj-xlsx-1-gae-10.1177_25168657211006159 for Differential DNA Methylation by Hispanic Ethnicity Among Firefighters in the United States by Jaclyn M Goodrich, Melissa A Furlong, Alberto J Caban-Martinez, Alesia M Jung, Ken Batai, Timothy Jenkins, Shawn Beitel, Sally Littau, John Gulotta, Darin Wallentine, Jeff Hughes, Charles Popp, Miriam M Calkins and Jefferey L Burgess in Epigenetics Insights

Footnotes

Acknowledgements

We first and foremost acknowledge the firefighter participants for graciously volunteering their time and samples to be in the study. We also acknowledge the partnering fire departments and their locals for their contribution to this research. We acknowledge Alisa Dewald at the University of Michigan for sample preparation. The University of Michigan Advanced Genomics Core and the University of Utah DNA Sequencing and Genomics Core Facility completed the EPIC analyses.

Funding:

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the National Institute of Environmental Health Sciences (NIEHS), specifically by supplements to the P30 Centers at the University of Arizona (Grant No. P30 ES006694) and the University of Michigan (Grant No. P30 ES017885). MF was supported by NIEHS Grant No. K99ES028743. The US Federal Emergency Management Agency (FEMA) also supported this work (Grant Nos. EMW-2014-FP-00200, EMW-2015-FP-00213, and EMW-2018-FP-00086). The findings and conclusions in this paper are those of the authors and do not necessarily represent the official position of the National Institute for Occupational Safety and Health (NIOSH), Centers for Disease Control and Prevention. Mention of trade names and commercial products does not constitute endorsement or recommendation for use by NIOSH.

Declaration of conflicting interests:

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Authors Contribution

JMG, JLB, and AJCM designed the research study; JLB, AJCM, and MMC obtained funding and designed protocols to develop the cohorts; JG, DW, JH, and CP recruited subjects and collected data; AMJ, SB, SL, MMC, and JLB managed data and samples; JMG and TJ conducted the DNA methylation analysis; JMG conducted statistical analysis and drafted the manuscript; and MAF, KB, TJ, MMC, and JLB provided interpretation of results. All authors contributed to the writing and/or editing of the manuscript.

ORCID iDs

Jaclyn M Goodrich

Alberto J Caban-Martinez

Supplemental material

Supplemental material for this article is available online.

References

Bolstad-Johnson

Burgess

Crutchfield

Storment

Gerkin

Wilson

JR.

Characterization of firefighter exposures during fire overhaul. Aihaj. 2000;61:636-641.

Fent

Toennis

Sammons

, et al. Firefighters’ absorption of PAHs and VOCs during controlled residential fires by job assignment and fire attack tactic. J Expo Sci Environ Epidemiol. 2020;30:338-349.

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans. Painting, firefighting, and shiftwork. In IARC Monographs on the Evalulation of Carcinogenic Risks to Humans, no 98. Lyon (FR): International Agency for Research on Cancer (IARC); 2010.

Daniels

Kubale

Yiin

, et al. Mortality and cancer incidence in a pooled cohort of US firefighters from San Francisco, Chicago and Philadelphia (1950-2009). Occup Environ Med. 2014;71:388-397.

Jalilian

Ziaei

Weiderpass

Rueegg

Khosravi

Kjaerheim

Cancer incidence and mortality among firefighters. Int J Cancer. 2019;145:2639-2646.

LeMasters

Genaidy

Succop

, et al. Cancer risk among firefighters: a review and meta-analysis of 32 studies. J Occup Environ Med. 2006;48:1189-1202.

Tsai

Luckhaupt

Schumacher

Cress

Deapen

Calvert

GM.

Risk of cancer among firefighters in California, 1988-2007. Am J Ind Med. 2015;58:715-729.

Haile

John

Levine

, et al. A review of cancer in US Hispanic populations. Cancer Prev Res. 2012;5:150-163.

Miller

Sauer

Ortiz

, et al. Cancer statistics for hispanics/latinos, 2018. CA Cancer J Clin. 2018;68:425-445.

10.

Fent

Alexander

Roberts

, et al. Contamination of firefighter personal protective equipment and skin and the effectiveness of decontamination procedures. J Occup Environ Hyg. 2017;14:801-814.

11.

Fent

Eisenberg

Snawder

, et al. Systemic exposure to PAHs and benzene in firefighters suppressing controlled structure fires. Ann Occup Hyg. 2014;58:830-845.

12.

Baxter

Hoffman

Knipp

Reponen

Haynes

, et al. Exposure of firefighters to particulates and polycyclic aromatic hydrocarbons. J Occup Environ Hyg. 2014;11:D85-D91.

13.

Stec

Dickens

Salden

, et al. Occupational exposure to polycyclic aromatic hydrocarbons and elevated cancer incidence in firefighters. Sci Rep. 2018;8:2476.

14.

Burgess

Hoppe-Jones

Griffin

, et al., Evaluation of interventions to reduce firefighter exposures. J Occup Environ Med. 2020;62:279-288.

15.

Laitinen

Mäkelä

Mikkola

Huttu

Fire fighting trainers’ exposure to carcinogenic agents in smoke diving simulators. Toxicol Lett. 2010;192:61-65.

16.

Stenehjem

Robsahm

Bråtveit

Samuelsen

Kirkeleit

Grimsrud

, et al. Aromatic hydrocarbons and risk of skin cancer by anatomical site in 25 000 male offshore petroleum workers. Am J Ind Med. 2017;60:679-688.

17.

Armstrong

Hutchinson

Unwin

Fletcher

Lung cancer risk after exposure to polycyclic aromatic hydrocarbons: a review and meta-analysis. Environ Health Perspect. 2004;112:970-978.

18.

Rybicki

Nock

Savera

Tang

Rundle

Polycyclic aromatic hydrocarbon-DNA adduct formation in prostate carcinogenesis. Cancer Lett. 2006;239:157-167.

19.

Lee

Koru-Sengul

Hernandez

, et al. Cancer risk among career male and female Florida firefighters: evidence from the Florida Firefighter Cancer Registry (1981-2014). Am J Ind Med. 2020;63:285-299.

20.

Jones

PA.

Functions of DNA methylation: islands, start sites, gene bodies and beyond. Nat Rev Genet. 2012;13:484-492.

21.

Furlong

Paul

Yan

, et al. An epigenome-wide association study of ambient pyrethroid pesticide exposures in California’s central valley. Int J Hyg Environ Health. 2020;229:113569.

22.

Zhu

, et al. Exposure to polycyclic aromatic hydrocarbons and accelerated DNA methylation aging. Environ Health Perspect. 2018;126:067005.

23.

Kim

Lee

Kim

, et al. Polycyclic aromatic hydrocarbons are associated with insulin receptor substrate 2 methylation in adipose tissues of Korean women. Environ Res. 2016;150:47-51.

24.

Dai

Mehta

Mordukhovich

, et al., Differential DNA methylation and PM2.5 species in a 450K epigenome-wide association study. Epigenetics. 2017;12:139-148.

25.

Lam

Emberly

Fraser

, et al. Factors underlying variable DNA methylation in a human community cohort. Proc Natl Acad Sci USA. 2012;109:17253-17260.

26.

Jeong

Zhou

Griffin

, et al. MicroRNA changes in firefighters. J Occup Environ Med. 2018;60:469-474.

27.

Zhou

Jenkins

Jung

, et al. DNA methylation among firefighters. PLoS One. 2019;14:e0214282.

28.

Zhang

Hong

, et al. Maternal smoking during pregnancy and cord blood DNA methylation: new insight on sex differences and effect modification by maternal folate levels. Epigenetics. 2018;13:505-518.

29.

McCarthy

Auger

Bale

, et al. The epigenetics of sex differences in the brain. J Neurosci. 2009;29:12815-12823.

30.

Adkins

Krushkal

Tylavsky

Thomas

Racial differences in gene-specific DNA methylation levels are present at birth. Birth Defects Res A: Clin Mol Teratol. 2011;91:728-736.

31.

Park

Patel

Loo

LWM

, et al. Association of internal smoking dose with blood DNA methylation in three racial/ethnic populations. Clin Epigenetics. 2018;10:110.

32.

McKennan

Naughton

Stanhope

, et al. Longitudinal data reveal strong genetic and weak non-genetic components of ethnicity-dependent blood DNA methylation levels. Epigenetics. 2020. doi: 10.1080/15592294.2020.1817290

33.

Bell

Pai

Pickrell

, et al. DNA methylation patterns associate with genetic and gene expression variation in HapMap cell lines. Genome Biol. 2011;12:R10.

34.

Galanter

Gignoux

, et al. Differential methylation between ethnic sub-groups reflects the effect of genetic ancestry and environmental exposures. Elife. 2017;6:e20532.

35.

Sasco

Secretan

Straif

Tobacco smoking and cancer: a brief review of recent epidemiological evidence. Lung Cancer. 2004;45:S3-S9.

36.

Dhir

Neste

, et al. Genomic and epigenomic integration identifies a prognostic signature in colon cancer. Clin Cancer Res. 2011;17:1535-1545.

37.

Widschwendter

Jones

Evans

, et al. Epigenome-based cancer risk prediction: rationale, opportunities and challenges. Nat Rev Clin Oncol. 2018;15:292.

38.

Teschendorff

Menon

Gentry-Maharaj

, et al. An epigenetic signature in peripheral blood predicts active ovarian cancer. PLoS One. 2009;4:e8274.

39.

Moran

Arribas

Esteller

Validation of a DNA methylation microarray for 850,000 CpG sites of the human genome enriched in enhancer sequences. Epigenomics. 2016;8:389-399.

40.

Fortin

Triche

Jr Hansen

KD.

Preprocessing, normalization and integration of the Illumina HumanMethylationEPIC array with minfi. Bioinformatics. 2016;33:558-560.

41.

Langie

SAS

De Boever

Taylor

Niu

, et al. RELIC: a novel dye-bias correction method for Illumina Methylation BeadChip. BMC Genomics. 2017;18:4.

42.

Pidsley

Zotenko

Peters

, et al. Critical evaluation of the illumina methylationepic beadchip microarray for whole-genome DNA methylation profiling. Genome Biol. 2016;17:208.

43.

Houseman

Accomando

Koestler

, et al. DNA methylation arrays as surrogate measures of cell mixture distribution. BMC Bioinformatics. 2012;13:86.

44.

Salas

Koestler

Butler

, et al. An optimized library for reference-based deconvolution of whole-blood biospecimens assayed using the Illumina HumanMethylationEPIC BeadArray. Genome Biol. 2018;19:64.

45.

Teschendorff

Zhuang

Widschwendter

Independent surrogate variable analysis to deconvolve confounding factors in large-scale microarray profiling studies. Bioinformatics. 2011;27:1496-1505.

46.

Ritchie

Phipson

, et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43:e47.

47.

Smyth

. Limma: Linear models for microarray data. In: Gentleman

, et al., eds. Bioinformatics and Computational Biology Solutions using R and Bioconductor, R. New York; 2005:397-420.

48.

Hochberg

Benjamini

More powerful procedures for multiple significance testing. Stat Med. 1990;9:811-818.

49.

Phipson

Maksimovic

Oshlack

missMethyl: an R package for analyzing data from Illumina’s HumanMethylation450 platform. Bioinformatics. 2016;32:286-288.

50.

Johnson

CJS

. CpG Poor Promoter SULT1C2 Regulated by DNA Methylation is Induced by Cigarette Smoke Condensate in Lung Cell Lines. University of Southern California: ProQuest Dissertation Publishing; 2015.

51.

Nestal de Moraes

da Torre Carneiro

Maia

Wing-Fai Lam

Sharrocks

AD.

FOXK2 transcription factor and its emerging roles in cancer. Cancers. 2019;11:393.

52.

Joseph

Radulovich

Wang

Raghavan

Zhu

C-Q

Tsao

M-S.

Rho guanine nucleotide exchange factor ARHGEF10 is a putative tumor suppressor in pancreatic ductal adenocarcinoma. Oncogene. 2020;39:308-321.

53.

Soloaga

Thomson

Wiggin

, et al. MSK2 and MSK1 mediate the mitogen- and stress-induced phosphorylation of histone H3 and HMG-14. EMBO J. 2003;22:2788-2797.

54.

Zhang

Cong

Zhang

Zeng

Song

Meng

Integrative analysis of ceRNA network and DNA methylation associated with gene expression in malignant pheochromocytomas: a study based on The Cancer Genome Atlas. Transl Androl Urol. 2020;9:344-354.

55.

Xiong

, et al. BTB/POZ zinc finger protein ZBTB16 inhibits breast cancer proliferation and metastasis through upregulating ZBTB28 and antagonizing BCL6/ZBTB27. Clin Epigenetics. 2020;12:82.

56.

Esser

Miller

Huang

, et al. Loss of LARGE2 disrupts functional glycosylation of α-dystroglycan in prostate cancer. J Biol Chem. 2013;288:2132-2142.

57.

Huang

Miller

Schappet

Henry

MD.

The glycosyltransferase LARGE2 is repressed by Snail and ZEB1 in prostate cancer. Cancer Biol Ther. 2015;16:125-136.

58.

Xiao

Unger

Yang

, et al. Loss of PLZF expression in prostate cancer by immunohistochemistry correlates with tumor aggressiveness and metastasis. PLoS One. 2015;10:e0121318.

59.

Ewa

Danuta

Polycyclic aromatic hydrocarbons and PAH-related DNA adducts. J Appl Genet. 2017;58:321-330.

60.

Wiese

Thompson

Kadlubar

FF.

Carcinogen substrate specificity of human COX-1 and COX-2. Carcinogenesis. 2001;22:5-10.

61.

Barnes

Zubair

John

Poirier

Martin

FL.

Carcinogens and DNA damage. Biochem Soc Trans. 2018;46:1213-1224.

62.

Rybicki

Neslund-Dudas

Bock

, et al. Polycyclic aromatic hydrocarbon–DNA adducts in prostate and biochemical recurrence after prostatectomy. Clin Cancer Res. 2008;14:750-757.

63.

Tang

Kryvenko

Wang

, et al. Elevated polycyclic aromatic hydrocarbon-DNA adducts in benign prostate and risk of prostate cancer in African Americans. Carcinogenesis. 2013;34:113-120.

64.

Daniels

Kadlubar

Sulfotransferase genetic variation: from cancer risk to treatment response. Drug Metab Rev. 2013;45:415-422.

65.

Deeken

Cormier

Price

, et al. A pharmacogenetic study of docetaxel and thalidomide in patients with castration-resistant prostate cancer using the DMET genotyping platform. Pharmacogenomics J. 2010;10:191-199.

66.

Beitel

Flahr

Hoppe-Jones

, et al. Assessment of the toxicity of firefighter exposures using the PAH CALUX bioassay. Environ Int. 2020;135:105207.

67.

Nebert

Roe

Dieter

Solis

Yang

Dalton

TP.

Role of the aromatic hydrocarbon receptor and [Ah] gene battery in the oxidative stress response, cell cycle control, and apoptosis. Biochem Pharmacol. 2000;59:65-85.

68.

Bock

Köhle

Ah receptor: dioxin-mediated toxic responses as hints to deregulated physiologic functions. Biochem Pharmacol. 2006;72:393-404.

69.

Helmig

Seelinger

Döhre

Schneider

RNA expressions of AHR, ARNT and CYP1B1 are influenced by AHR Arg554Lys polymorphism. Mol Genet Metab. 2011;104:180-184.

70.

Okino

Pookot

L-C

, et al. Epigenetic inactivation of the dioxin-responsive cytochrome P4501A1 gene in human prostate cancer. Cancer Res. 2006;66:7420-7428.

71.

Leng

Liu

Thomas

, et al. Native American ancestry affects the risk for gene methylation in the lungs of Hispanic smokers from New Mexico. Am J Respir Crit Care Med. 2013;188:1110-1116.

72.

Leng

Klinge

, et al. Gene methylation biomarkers in sputum as a classifier for lung cancer risk. Oncotarget. 2017;8:63978-63985.

73.

Wasim

Carlet

Mansha

, et al. PLZF/ZBTB16, a glucocorticoid response gene in acute lymphoblastic leukemia, interferes with glucocorticoid-induced apoptosis. J Steroid Biochem Mol Biol. 2010;120:218-227.

74.

Miller

Schappet

, et al. Downregulation of dystroglycan glycosyltransferases LARGE2 and ISPD associate with increased mortality in clear cell renal cell carcinoma. Mol Cancer. 2015;14:141.

75.

Mosquera Orgueira

Rodríguez

Díaz Arias

JÁ

González Pérez

Bello López

JL.

New recurrent structural aberrations in the genome of chronic lymphocytic leukemia based on exome-sequencing data. Front Genet. 2019;10:854.

76.

Mohanty

A.F.

, et al., Infant sex-specific placental cadmium and DNA methylation associations. Environ Res. 2015;138:74-81.

77.

Mansell

Gorrie-Stone

Bao

, et al. Guidance for DNA methylation studies: statistical insights from the Illumina EPIC array. BMC Genomics. 2019;20:366.

78.

Reuben

Sugden

Arseneault

, et al. Association of neighborhood disadvantage in childhood with DNA methylation in young adulthood. JAMA Netw Open. 2020;3:e206095.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.04 MB