Environmental and Molecular Mutagenesis 55:184^195 (2014) Research Article Bisphenol A-Associated Alterations in the Expression and Epigenetic Regulation of Genes Encoding Xenobiotic Metabolizing Enzymes in Human Fetal Liver Muna S. Nahar,1 Jung H. Kim,1 Maureen A. Sartor,2 and Dana C. Dolinoy1* 1Department of Environmental Health Sciences, University of Michigan, Ann Arbor, Michigan 2Center for Computational Medicine and Bioinformatics, University of Michigan, Ann Arbor, Michigan Alterations in xenobiotic metabolizing enzyme were evaluated in silico, putative binding sites (XME) expression across the life course, along for the E-twenty-six (ETS) and activator protein1 with genetic, nutritional, and environmental reg- (AP1) related transcription factor families were ulation, can influence how organisms respond identified and unique to 97% of all candidate to toxic insults. In this study, we investigated transcripts. Interestingly, many ETS binding sites the hypothesis that in utero exposure to the contain cytosine-guanine dinucleotides (CpGs) endocrine active compound, bisphenol A (BPA), within their consensus sequences. Thus, quantita- influences expression and epigenetic regulation tive analysis of CpG methylation of three candi- of phase I and II XME genes during develop- date genes was conducted across n 5 50 ment. Using healthy 1st to 2nd trimester human samples. Higher BPA levels were associated fetal liver specimens quantified for internal BPA with increased site-specific methylation at levels, we examined XME gene expression COMT (P < 0.005) and increased average using PCR Array (n 5 8) and RNA-sequencing methylation at SULT2A1 (P < 0.020) promoters. (n 5 12) platforms. Of the greater than 160 While toxicological studies have traditionally XME genes assayed, 2 phase I and 12 phase focused on high-dose effects and hormonal II genes exhibited significantly reduced expres- receptor mediated regulation, our findings sug- sion with higher BPA levels, including isoforms gest the importance of low-dose effects and from the carboxylesterase, catechol O-methyl- nonclassical mechanisms of endocrine disruption transferase, glutathione S-transferase, sulfotrans- during development. Environ. Mol. Mutagen. ferase, and UDP-glucuronosyltransferase families. 55:184–195, 2014. VC 2013 Wiley Periodicals, Inc. When the promoters of these candidate genes Key words: xenobiotic metabolism; bisphenol A; liver; DNA methylation; transcription factor INTRODUCTION A compound’s absorption, distribution, metabolism, and Grant sponsor: National Institute of Health; Grant number: ES017524. excretion via biotransformation pathways dictate its thera- Grant sponsor: University of Michigan (UM) National Institute of Envi- ronmental Health Science (NIEHS) Core Center; Grant number: P30 peutic or toxic potential. Specifically, changes in xenobiotic ES017885. metabolizing enzyme (XME) gene expression and activity Grant sponsor: UM NIEHS Institutional Training Grant; Grant number: can alter drug efficacy and toxicity, with the greatest dis- T32 ES007062. crepancies observed in children compared with adults *Correspondence to: Dana C. Dolinoy, 6638 SPH Tower, 1415 Wash- [Alcorn and McNamara, 2003]. In general, the oxidizing ington Heights, Ann Arbor, MI 48109-2029, USA. E-mail: ddolinoy@ phase I XMEs functionalize compounds into active metabo- umich.edu lites while the conjugating phase II XMEs increase the Received 23 July 2013; provisionally accepted 19 September 2013; and molecular weight of compounds for rapid excretion of toxic in final form 30 September 2013 metabolites [Caldwell et al., 1995]. Preliminary studies DOI 10.1002/em.21823 indicate that relative XME expression and activity among Published online 9 November 2013 in different classes and isoforms change drastically throughout Wiley Online Library (wileyonlinelibrary.com). VC 2013 Wiley Periodicals, Inc. Environmental and Molecular Mutagenesis. DOI 10.1002/em BPA and Xenobiotic Metabolism in Fetal Liver 185 the life course [Hines, 2008]. For example, the phase II limited to animal and in vitro models. For example, rat UDP-glucuronosyltransferase isoforms, UGT1A1 and hepatic microsome studies show that BPA inhibits CYP UGT1A6, are both expressed at low levels in the human enzymes, including activity for CYP1A2, CYP2C11, and fetus; while UGT1A1 attains adult levels within months CYP2E1 [Hanioka et al., 2000; Pfeiffer and Metzler after birth, the UGT1A6 isoform reaches adult levels at 10 2004]; yeast and human liver microsome studies associate years of age [McCarver and Hines, 2002]. Therefore, the BPA with both competitive and noncompetitive inhibition extent of metabolism of a xenobiotic chemical is dependent of UGT1A6 [Hanioka et al., 2008]; and human endome- on age and ontogeny, or the maturation of specific XMEs trial Ishikawa cell line studies show increased ALDH3A1 [Allegaert et al., 2007]. Much of our understanding of expression following BPA exposure [Naciff et al., 2010]. human XME ontogeny arises from adverse pharmaceutical Interspecies differences in metabolism and/or relatively exposures or diseases throughout birth and infancy, or high BPA exposure doses, however, limit the translation extrapolations from rodent models [Saghir et al., 2012]. of these findings for human risk assessment. Comprehensive studies examining metabolism in early The influence of xenoestrogens on the maturation and human fetal development are limited; for example, recent regulation of phase I and II xenobiotic metabolizing studies characterize only a handful of XMEs, such as ster- enzymes has yet to be studied in humans. Here, utilizing oidogenic enzymes in 2nd trimester fetal liver [O’Shaugh- a comprehensive approach evaluating XME expression, nessy et al., 2013] or cytochrome p450s (CYP) and we identify multiple XME genes that are associated with glutathione s-transferases in fetal liver and adrenals [Wang physiologically relevant concentrations of total BPA, et al., 2008]. Evaluating ontogeny, especially throughout assess relative abundance of these XME mRNAs in the human gestation, is thus of great importance in assessing developing human fetal liver, and investigate DNA meth- toxicity; however, obtaining suitable human specimens may ylation as a potential mechanism influencing biotransfor- pose ethical and technical challenges. mation response to BPA exposure. In addition to drug metabolism, XMEs play an impor- tant role in steroid homeostasis, neuroendocrine function, and growth. Both endogenous and exogenous compounds METHODS AND MATERIALS help regulate XME ontogeny and subsequently, metabolic function. More recently, studies demonstrating the impor- Tissue Samples tance of hormonal regulation on the establishment of Human fetal liver samples were procured from the NIH- hepatic metabolism throughout pregnancy and develop- funded University of Washington Birth Defects Research ment have emerged [Kennedy, 2008; Jeong, 2010]. Xeno- Laboratory fetal biobank (2R24HD000836-47), and charac- biotics that can mimic endogenous hormones, such as terized for BPA concentrations by the Kannan Laboratory endocrine active compounds (EAC), can potentially mod- at the Wadsworth Center (New York State Department of ify baseline hormonal regulation of xenobiotic metabo- Health). As previously described [Nahar et al., 2012], fol- lism and response to environmental stressors during lowing surgery and consent from volunteers undergoing critical windows of development. elective abortions during 1st and 2nd trimester of preg- Bisphenol A (BPA), a synthetic estrogen used in the nancy (gestational day 74–120), healthy tissue specimens production of polycarbonate plastics and epoxy resin, is a were flash frozen and immediately stored in polycarbonate- controversial EAC that is of concern primarily in the free tubing at 280C until processed for BPA analysis and developing organism [Rubin, 2011]. Presence of BPA in RNA/DNA extraction. No identifying clinical data were fetal tissue [Cao et al., 2012; Nahar et al., 2012], reduced available on samples except for sex and gestational age. capacity for BPA metabolism in the fetus [Nahar et al., Total BPA concentrations measured in liver tissue ranged 2012], and the ability for BPA transfer across the pla- from below the limit of quantification at 0.071 ng/g (LOQ/ centa [Schonfelder et al., 2002; Balakrishnan et al., 2010; ͱ2, where LOQ 5 0.1 ng/g) up to 96.8 ng/g [Nahar et al., Jimenez-Dıaz et al., 2010], place the developing human 2012]. fetus at a higher risk for BPA toxicity. Although the human health consequences of BPA exposures are dis- RNA Extraction and cDNA Synthesis puted, several rodent studies suggest that developmental exposure to BPA can alter susceptibility to disease later Total RNA was isolated from frozen liver tissue using in life by modifying the epigenome [Ho et al., 2006; the AllPrep DNA/RNA/Protein kit (Qiagen, Valencia, Bromer et al., 2010; Kundakovic and Champagne, 2011; CA) according to the manufacturer’s instructions. Anderson et al., 2012]. Thus, we hypothesize that BPA Approximately 10 to 20 mg of homogenized tissue was dependent changes to epigenetic regulation of XME dur- added to 600 mL of Buffer RLT (containing 1% b-mer- ing development may also result in latent effects on dis- captoethanol) in a 2 mL round bottom polypropylene ease susceptibility. To date, research addressing BPAs eppendorf tube with a 5 mm stainless steel bead. Samples influence on XME expression and activity has been were further homogenized in solution for 2 min at 20 Hz Environmental