Transcriptional Regulation of the Human Microsomal

The Pennsylvania State University

The Graduate School The Huck Institutes of the Life Sciences

TRANSCRIPTIONAL REGULATION OF THE HUMAN MICROSOMAL

EPOXIDE HYDROLASE GENE (EPHX1) DRIVEN BY A FAR UPSTREAM

ALTERNATIVE PROMOTER

A Dissertation in

Molecular Toxicology

Shengzhong Su

 2013 Shengzhong Su

Submitted in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy

August 2013

The dissertation of Shengzhong Su was reviewed and approved* by the following:

Curtis J. Omiecinski Professor of Veterinary and Biomedical Sciences and Hallowell Chair Dissertation Advisor Chair of Committee Chair of Intercollege Graduate Degree Program in Molecular Toxicology

Adam B. Glick Associate Professor of Veterinary and Biomedical Sciences

K. Sandeep Prabhu Associate Professor of Immunology and Molecular Toxicology

John Vanden Heuvel Professor of Molecular Toxicology

Joshua D. Lambert Associate Professor of Food Science

*Signatures are on file in the Graduate School

iii

ABSTRACT

Microsomal epoxide hydrolase (mEH) is an important metabolizing enzyme that plays roles in both detoxification and bioactivation of xenobiotics. Human mEH gene expression is subjected to the regulation of alternative promoter usage generating multiple transcripts, including the most prevalent, termed E1b and E1. These transcripts possess distinct untranslated first exons but encode identical mEH protein given that the second exon contains the translation initiation site. E1b is ubiquitously expressed at high levels in all tissues while E1 is selectively expressed in the liver. Although several liver-specific transcription factors were characterized previously as involved in the regulation of E1 transcription, little is known regarding the molecular mechanism regulating E1b expression. Study of those underlying processes is the principle focus of these investigations. Initially in these studies we sought to identify the key transcription factors responsible for controlling the constitutive expression of the E1b transcript. Sequence analysis of E1b proximal promoter revealed several potential Sp1/Sp3 binding sites. Site-directed mutagenesis of these motifs established their roles in regulating E1b promoter activities. Chromatin immunoprecipitation (ChIP) analyses demonstrated that both Sp1 and Sp3 are bound to the proximal promoter region of E1b. Silencing, or knockdown of Sp1 expression using siRNA had no detectable effect on the endogenous E1b transcriptional level. However, knockdown of Sp3 greatly diminished E1b expression in several different human cell lines. These results demonstrated that Sp3 in particular was involved in regulating the basal expression patterns of the mEH E1b variant transcript. Secondly, following analysis of DNase I hypersensitivity data available in the ENCODE project, we identified and characterized two intronic DNA elements in the mEH genomic region. This led to the discovery that the master oxidative stress regulator, Nuclear factor erythroid-derived 2 (NF-E2)-related factor 2 (Nrf2) functioned as a mediator of E1b upregulation in lung cancer-derived cells. Results obtained from both luciferase gene reporter and ChIP assays indicated that Nrf2 interacts with the 2nd intronic DNA element following its activation with the antioxidants, sulforaphane or tBHQ. DNA sequence analysis of the enhancer region together with electrophoretic mobility shift assays (EMSA) enabled the identification of a conserved antioxidant-response element within the enhancer that appeared to mediate these transcriptional responses. Finally in these studies, we sought to characterize differences in the transcriptional responses of the E1b and E1 promoters in hepatoma cell lines and human normal hepatocytes to chemical mediators. Nrf2 siRNA knockdown studies in hepatoma C3A cells were performed to identify the Nrf2 signaling pathway as functional in mediating the activation effects contributed by the monofunctional inducers, sulforaphane and tBHQ, to both of E1b and E1 promoters. Luciferase reporter assays demonstrated that these effects were localized to the 2nd intronic enhancer element. However, bifunctional inducers exhibited considerable differences in their regulation of E1b and E1 transcript expression. Treatment with 3-MC, an aryl hydrocarbon receptor (AhR) agonist, induced E1b expression but inhibited E1. Another bifunctional inducer, β-naphthoflavone (β-NF) upregulated both E1b and E1 whereas 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) treatment resulted in decreased E1 expression iv

(similar to 3-MC), yet had no discernible effect on expression of E1b. Further studies with 6,2',4'- trimethoxyflavone (TMF), an AhR antagonist, together with Nrf2 siRNA, suggested that both AhR and Nrf2 signaling pathways contribute to the regulation of the E1b and E1 promoters mediated by bifunctional AhR agonists. As well, these effects were comparatively evaluated in human hepatoma HepG2 and Huh7 cells, in human primary hepatocytes and in human lung BEAS-2B cells. Overall, this thesis research successfully identified and defined multiple levels of molecular interaction that contribute to the regulation of the dual gene promoter usage characterizing human mEH gene expression. The results contributed through these investigations provide important advances in our core understanding of the complex genetic regulatory schemes controlling the dual promoter usage and tissue-selective expression character underlying the functional biology and toxicological roles of human mEH.

TABLE OF CONTENTS

LIST OF FIGURES ...... vii LIST OF TABLES ...... viii ABBREVIATIONS ...... ix ACKNOWLEDGEMENTS ...... xi DEDICATION ...... xii Chapter 1 Introduction ...... 1 GENERAL INTRODUCTION TO XENOBIOTIC METABOLISM ...... 1 EPOXIDE HYDROLASE ...... 2 The physiological functions of mEH ...... 4 Gene organization of human mEH gene (EPHX1) ...... 4 Polymorphisms and their clinical implications ...... 5 Transcriptional regulation of Human mEH gene ...... 7 Tissue-specific regulation of human EPHX1 gene by the selective use of alternative promoters ...... 7 TRANSCRIPTIONAL REGULATION OF GENE EXPRESSION ...... 9 The core promoter ...... 9 Focused versus dispersed core promoters ...... 10 CpG islands (CGI) ...... 11 Proximal Promoter Elements ...... 12 Transcription factor Sp1 ...... 13 Distal regulatory elements ...... 13 Identification of distal enhancer elements ...... 15 TRANSCRIPTION FACTORS REGULATING DRUG-METABOLIZING ENZYMES...... 16 OXIDATIVE STRESS AND NRF2-ARE SIGNALING PATHWAY ...... 17 The structures of Nrf2 and Keap1 ...... 18 Keap1 and Nrf2 mutation in cancers ...... 20 The crosstalk between Nrf2 and AhR signaling pathways ...... 20 HYPOTHESES AND AIMS ...... 21 Chapter 2 Transcription factors Sp1 and Sp3 contribute to the basal expression of human microsomal epoxide hydrolase alternative E1b mRNA variant by interacting with the proximal E1b promoter ...... 31 ABSTRACT ...... 31 INTRODUCTION ...... 31 MATERIALS AND METHODS ...... 32 Materials ...... 32 Plasmids ...... 33 Cell culture, transient transfection and luciferase reporter assays ...... 33 Sp1 and Sp3 siRNA knockdown studies ...... 34 RNA isolation, reverse transcription and quantitative real-time PCR ...... 34 Western blotting ...... 35 Electrophoretic mobility shift assays (EMSA) ...... 35 Chromatin immunoprecipitation (ChIP) assay ...... 36 Statistical analyses ...... 37 RESULTS ...... 37 Identification of the critical promoter region for the basal expression of E1b ...... 37 Analysis of the E1b proximal promoter region for the transcription factor Sp1/Sp3 binding sites ...... 37 Sp1/Sp3 is involved in activation of the E1b proximal promoter ...... 38 Sp1 and Sp3 bind and interact with the E1b proximal promoter region ...... 39 Knockdown of Sp1 and Sp3 regulated expression of E1b variant ...... 39 DISCUSSION ...... 40 Chapter 3 The role of intronic DNA elements in the regulation of the human microsomal epoxide hydrolase (mEH, EPHX1) driven by a far upstream alternative E1b promoter ...... 48 ABSTRACT ...... 48 vi

INTRODUCTION ...... 48 MATERIALS AND METHODS ...... 50 Materials ...... 50 Plasmids ...... 50 Cell culture, transient transfection and luciferase reporter assays ...... 51 RNA isolation, reverse transcription and quantitative real-time PCR ...... 51 Western blotting ...... 52 Nrf2 siRNA knockdown studies ...... 53 Chromatin immunoprecipitation (ChIP) assay ...... 53 Electrophoretic mobility shift assays (EMSA) ...... 54 Statistical analyses ...... 54 RESULTS ...... 55 Nrf2 activators induce E1b expression in human lung epithelial cell lines ...... 55 Identification of an Nrf2-responsive DNA enhancer element ...... 55 Nrf2 binding to an antioxidant response element (ARE) in HS-2 Enhancer ...... 56 Characterization of a TPA-response element overlapping with the ARE within HS-2 ...... 57 DISCUSSION ...... 57 Chapter 4 The transcriptional regulation of human microsomal Epoxide Hydrolase (mEH) driven by two alternative promoters in human hepatoma C3A cells ...... 72 ABSTRACT ...... 72 INTRODUCTION ...... 73 MATERIALS AND METHODS ...... 74 Materials ...... 74 Plasmids ...... 75 Cell culture, transient transfection and luciferase reporter assays ...... 75 RNA isolation, reverse transcription and quantitative real-time PCR...... 76 Nrf2 gene knockdown by siRNA ...... 76 Western blotting ...... 77 Statistical analyses ...... 77 RESULTS ...... 77 Regulation of E1b and E1 by antioxidants and Nrf2 in C3A cells ...... 77 Identification of antioxidant-responsive regions in E1b and E1 promoters ...... 78 Regulation of E1b and E1 transcripts by AhR agonists ...... 78 Involvement of AhR and Nrf2 in regulation of E1b and E1 expression ...... 79 Regulation of E1b and E1 promoter activities by AhR agonists ...... 79 Induction of mEH expression by xenobiotics in other cell lines ...... 80 DISCUSSION ...... 81 Chapter 5 Conclusions ...... 91 Chapter 6 Future perspectives ...... 94 1. TISSUE-DEPENDENT HUMAN MEH EXPRESSION ...... 94 2. DISTAL REGULATORY ELEMENT-MEDIATED E1B EXPRESSION ...... 95 3. SINGLE NUCLEOTIDE POLYMORPHISMS (SNP) AND THEIR POTENTIAL ROLE FOR REGULATING ENHANCER AND PROMOTER FUNCTIONS ...... 98 4. IN VIVO MODELS FOR THE STUDY OF REGULATORY MECHANISMS AND FUNCTIONS OF HUMAN MEH GENE ...... 98 References ...... 105

vii

LIST OF FIGURES

Figure 1-1. Metabolism of xenobiotics by phases I, II, and III of xenobiotic metabolizing systems...... 24 Figure 1-2. The gene structure of EPHX1 and conservation among 46 vertebrate species...... 25 Figure 1-3. The genomic region encompassing the human mEH gene is highly polymorphic...... 26 Figure 1-4. The genomic region of human EPHX1 contains intronic regulatory elements...... 30 Figure 1-5. Schematic of a typical gene regulatory region...... 27 Figure 1-6. Focused versus dispersed transcription initiation...... 27 Figure 1-7. A schematic structure of Sp1 protein...... 27 Figure 1-8. The Keap1–Nrf2 pathway...... 28 Figure 1-9. The structure of Nrf2 and Keap1...... 29 Figure 2-1. Location of E1b proximal promoter...... 42 Figure 2-2. Identification and mutational analysis of Sp1/Sp3 binding sites within the E1b proximal promoter region...... 43 Figure 2-3. Sp1 and Sp3 regulated E1b proximal promoter activities...... 44 Figure 2-4. EMSA and supershift analyses show Sp1 binding to putative Sp1/Sp3 binding sites on E1b-300 promoter...... 45 Figure 2-5. ChIP assay for Sp1 and Sp3 binding to the E1b proximal promoter in BEAS-2B and C3A cells...... 46 Figure 2-6. Effect of Sp1 and Sp3 knockdown on E1b transcript and mEH protein...... 47 Figure 3-1. Antioxidants induce E1b expression...... 63 Figure 3-2. Effect of overexpression and siRNA knockdown of NRF2 on E1b expression...... 64 Figure 3-3. Identification of regulatory elements involved in E1b induction...... 66 Figure 3-4. Role of DNase HS sites in Nrf2-mediated induction of E1b...... 68 Figure 3-5. Nrf2 binds to HS-2 enhancer element...... 69 Figure 3-6. Identification of Antioxidant response element (ARE) within HS-2...... 70 Figure 3-7. A TRE binding site influences ARE-driven enhancer activity...... 71 Figure 4-1. Regulation of E1b and E1 by antioxidants and Nrf2 in C3A cells...... 85 Figure 4-2. Involvement of intronic regulatory regions in antioxidant and Nrf2-mediated E1b and E1 inductions...... 86 Figure 4-3. Regulation of E1b and E1 transcripts by AhR agonists...... 87 Figure 4-4. Involvement of AhR and Nrf2 in regulation of E1b and E1 expression by AhR agonists...... 88 Figure 4-5. Regulation of E1b and E1 promoter activities by AhR agonists...... 89 Figure 4-6. Induction of mEH expression by xenobiotics in other cell lines...... 90 Figure 6-1. Studies on interaction between intronic enhancer elements and the proximal promoters of human mEH gene...... 100 Figure 6-2. Overview of 3C and 3C-derived methods...... 101 Figure 6-3. Distribution of Single nucleotide polymorphisms (SNP) in the genomic regions of human mEH gene (EPHX1)...... 102 Figure 6-4. Electrophoretic mobility shift assay (EMSA) and supershift analyses of AP-2α binding to alleles at the loci of rs12741681(A/C) and rs12727007(C/T)...... 103 Figure 6-5. Differential effect of rs12403480 (C/T) polymorphism on the enhancer activity of DNase I HS-2 to E1b and E1 promoter activities...... 104

viii

LIST OF TABLES

Table 1-1. Modification of Cysteine residues of Keap1 by electrophilic reagents...... 24 Table 3-1. Alignment of HS-2 ARE with known AREs...... 61

ABBREVIATIONS

3-MC 3-Methylcholantrene AhR Aryl hydrocarbon receptor AP-1 Activator protein 1 ARE Antioxidant response element, also known as the electrophile response element (EpRE or ERE) ARNT Aryl hydrocarbon receptor nuclear translocator BHA Butylated hydroxyanisole bHLH Basic-helix-loop-helix BRE TFIIB recognition element bZIP Basic region-leucine zipper C/EBP CCAAT/enhancer binding protein CAR Constitutive active/androstane receptor CGI CpG island CBP CREB-binding protein ChIP Chromatin immunoprecipitation CNC Cap ‘n’ collar CTCF CCCTC-binding factor CTF CCAAT transcription factor, also known as nuclear factor-1, NF-1 D3T 3H-1,2-dithiole-3-thione DCE Downstream core element DPE Downstream core promoter element EH Epoxide hydrolase EMSA Electrophoretic mobility shift assay ERα Estrogen receptor-α GCS γ-Glutamyl cysteine synthethase GR Glucocorticoid receptor GST Glutathione S-transferase HNF Hepatic nuclear factor HO-1 Heme oxygenase 1 Inr Initiator element Keap1 Kelch-like ECH-associated protein 1, also known as inhibitor of Nrf2 (INrf2) KLF Kruppel-like factor MAPK Mitogen-activated protein kinase MDR1 Multidrug resistant protein-1 mEH Microsomal epoxide hydrolase MRP Multidrug resistance-associated protein MTE Motif ten element NAT N-acyltransferase NCoR Nuclear hormone receptor corepressor Neh Nrf2-ECH homology NF-κB Nuclear factor-κB NF-Y CCAAT box binding factor/nuclear factor Y NMO NAD(P)H:menadione reductase x

NQO1 NAD(P)H:quinone oxidoreductase, also known as quinone reductase Nrf-1 nuclear respiratory factor-1 Nrf2 Nuclear factor erythroid-derived 2 (NF-E2)-related factor 2 OATP2 Organic anion transporting polypeptide 2 P-gp P-glycoprotein PAH Polycyclic aromatic hydrocarbon PIC Preinitiation complex PPAR Peroxisome proliferator-activated receptor PXR Pregnane X receptor ROS Reactive oxygen species RXR Retinoid X receptor SFN Sulforaphane SNP Single nucleotide polymorphism SP1 Specificity protein 1 SP3 Specificity protein 3 SULT Sulfotransferase TAF TBP-associated protein tBHQ tert-Butylhydroquinone TBP TATA-binding protein TCDD 2,3,7,8-Tetrachlorodibenzo-p-dioxin TPA 12-O-tetradecanoylphorbol-13-acetate, also known as phorbol 12-myristate 13-acetate (PMA) TRE TPA responsive element TSS Transcription start site UGT UDP-glucuronosyltransferase XCPE X-gene core promoter element XRE Xenobiotic response element, also known as dioxin response element (DRE) β-NF β-Naphthoflavone

ACKNOWLEDGEMENTS

This thesis would never been possible without the unending help of so many. My first gratitude goes to Dr. Scott Auerbach who taught me a great deal of molecular biology techniques so that I could have a quick start in my project. Also, I would like to thank the current and former members in Dr. Omiecinski’s lab: Denise Coslo, Dr. Tao Chen, Hong Loan Thi Nguyen, Ben Niu, Dr. Fengming Chen, Dr. Xiaokun Cai, Will Hedrich, Dr. Kaarthik John, Brian Sell, Dr. Joshua G. DeKeyser, Dr. Katy Olsavsky Goyak, Mary Hutchinson, Dr. Stephanie M. Zamule, Dr. Matthew Stoner, Dr. Jeanine Page, Dr. Shun-Hsin Liang, Dr. Xi Yang, for the stimulating discussions, for their support, and for all the fun we have in the past years. In particular, I am grateful to Dr. Elizabeth Laurenzana for the critical reading and editing of my thesis. I wish to thank Janice Kennedy in the office of Huck Institute of Life Sciences who have help me a lot, in areas such as class registration and application for Sumer Tuition Assistance Program. I would like to express the deepest appreciation to my advisor, Dr. Curtis Omiecinski, for his excellent training, untiring supervision, and continuous support during this course of my research in his lab. I would like to thank him for providing me the opportunity to work with him and also for his understanding and patience he has offered this phase of my professional career. I would like to thank my committee members, Dr. Adam Glick, Dr. Sandeep Prabhu, Dr. Jack Vanden Heuvel, and Dr. Joshua D. Lambert, for their guidance, encouragement and constructive criticism. Last but not least, I would like to thank my wife, Lijun Wang, and my children, Jeffery and Katherine, who offered me unconditional love and support throughout the course of this thesis endeavor.

xii

DEDICATION

This thesis is dedicated to my wife Lijun, and my children, Jeffery and Katherine, who have always stood by me. Also, this thesis is dedicated to my parents who have never stopped believing in me and for giving me moral support from the other side of the globe. Finally, this thesis is dedicated to the almighty GOD who gave me the strength to complete it.

Chapter 1 Introduction

GENERAL INTRODUCTION TO XENOBIOTIC METABOLISM Humans are continuously exposed to numerous xenobiotics, such as drugs, food additives, industrial chemicals, and environmental pollutants. To protect the body from the potential harmful exposure to these chemicals, xenobiotics undergo a broad range of biotransformation processes that in general, render them less toxic and more water-soluble and allow for their elimination. The metabolism of xenobiotics is mediated by a collection of xenobiotic metabolizing enzymes (XME) including Phase I and Phase II metabolizing enzymes as well as Phase III transporters (Figure 1-1) (Xu et al., 2005). Major reactions catalyzed by Phase I enzymes include oxidation (including cytochrome P450, flavin monooxygenase, amine oxidases, lipoxygenases, alcohol dehydrogenase, aldehyde and xanthine oxidases, and peroxidase), reduction (such as nitroreductases and azoreductases), and hydrolysis (including the carboxylesterases and epoxide hydrolases) (Chen, 2012a). These reactions introduce a functional group (e.g. –OH,

–COOH, –SH, –O–, or NH2) to the parent compound that leads to an alteration in its biological properties that may render the substance toward further metabolism and also facilitate elimination. However, in many cases, the acquirement of a functional group results in the formation of an intermediate metabolite that is more chemically active and more toxic than the parent compound. These metabolites may react with cellular molecules, e.g., proteins, DNA and/or lipids, causing cell and organ toxicity. Primary metabolites generated from Phase I reactions are subsequently modified further by the Phase II enzymes. The Phase II enzymes include UDP-glucuronosyltransferases (UGTs), glutathione S-transferases (GSTs), sulfotransferases (SULTs), acetyltransferases, methyltransferases, N-acyltransferases (NATs), NAD(P)H:quinone oxidoreductase (NQO) and NAD(P)H:menadione reductase (NMO) (Chen, 2012b; Xu et al., 2005). The Phase II reactions are conjugation reactions involving the transfer of endogenous substances (e.g., glutathione, sulfate, and glucuronic acid) to a functional group of the primary metabolite. Phase II reactions generally inactivate potentially toxic metabolites generated in Phase I reactions and greatly improve water solubility of xenobiotics, facilitating the elimination of xenobiotics through urinary excretion or via the transporter-mediated Phase III elimination pathways. Phase III transporters include P-glycoprotein (P-gp; also known as multidrug resistant protein-1, MDR1), multidrug resistance-associated protein (MRP) and organic anion transporting polypeptide 2 (OATP2) (Xu et al., 2005). They play critical roles in xenobiotic absorption, distribution and excretion. Both P-gp and MRP are members of ATP binding cassette (ABC) transporter family. ABC transporters import or export a wide range of substrates, such as amino acids, metal ions, sugars, lipids, and xenobiotics, across the cell membrane. OATP2 is a member of the OATP family that mediates sodium- and ATP-independent transport of a variety of endogenous and exogenous compounds, including conjugated and unconjugated bilirubin, conjugated steroids, neutral 2 compounds, some type II organic cations, thyroid hormones, and bile salts. With these transporters, conjugated metabolites from Phase II reactions are able to be eliminated from mammalian organisms. A functioning detoxification system is critical in humans for protection against a variety of diseases. Factors altering the expression levels and activities of XMEs could alter the balance between detoxication and bioactivation functions of the xenobiotic metabolism system and potentially lead to adverse effects arising from xenobiotic exposure. Modulatory factors include genetic polymorphism, enzyme inducibility and posttranslational modifications. Epoxide hydrolases (EH) are important Phase I enzymes as chemical epoxides may be direct substrates for epoxide hydrolase metabolism. Because they also carry out the hydrolysis of epoxides generated from Phase I oxidative activation, such as CYP450-catalyzed epoxidation, epoxide hydrolases can also be classified as Phase II enzymes. This dissertation research is primarily focused on the study of molecular mechanisms involved in the transcriptional regulation of human microsomal epoxide hydrolase (mEH) expression. The following section provides a background review on the types of epoxide hydrolase, their physiological roles, genetic polymorphism and known aspects of their transcriptional regulation.

EPOXIDE HYDROLASE An epoxide is a three-membered cyclic ether containing an oxygen atom attached to two adjacent carbon atoms of a hydrocarbon. Epoxide-containing compounds are typically formed during the metabolic process of monooxygenation of arene or alkene in endogenous and xenobiotic compounds catalyzed by the cytochrome P450 enzymes. The highly strained ring structure with a highly polarized oxygen-carbon bond renders many epoxides chemically reactive (Parker and Isaacs, 1959). Consequently, such metabolites can interact with cellar macromolecules such as DNA and proteins, leading to mutagenic, toxic, and carcinogenic effects (Gelboin, 1969; Grover and Sims, 1968). Therefore, the concentrations of reactive epoxides in living organisms must be controlled to ensure the normal cellular function. EHs are a special class of hydrolase that transform epoxides to the corresponding vicinal diols by the catalytic addition of water. The dihydrodiol reaction products are often less reactive, less mutagenic and more water soluble than the parental epoxides. To date, five epoxide hydrolases with different subcellular localizations and substrate specificities have been characterized in mammals (Decker et al., 2009). These include: 1) Microsomal cholesterol epoxide hydrolase that catalyzes the hydration of cholesterol-5α, 6α-epoxide, specifically; 2) Hepoxilin A3 hydrolase that was purified from rat liver cytosol and exhibits a substrate preference for hepoxilin A3 derived from arachidonic acid metabolism; 3) Leukotriene A4 hydrolase that is a bifunctional metalloprotein functioning as an epoxide hydrolase and an aminopeptidase and is localized in the cytosolic compartment; 4) Soluble epoxide hydrolase (sEH, EPHX2) that is mainly localized in the cytosolic compartment and involved in the metabolism of arachidonic epoxides and other fatty acid-derived epoxides; and 5) Microsomal 3 epoxide hydrolase (mEH, EPHX1) that is attached to the cytosolic side of the endoplasmic reticulum membrane and plays an important role in xenobiotic metabolism. sEH and mEH are the most studied mammalian EHs (Morisseau and Hammock, 2005). The two enzymes are members of the α/β-hydrolase fold family of proteins and although not yet crystallized, likely share a similar three-dimensional structure. However, they differ in their subcellular distributions, substrate specificities and physiological roles. sEH is found mostly in the cytosol (Decker et al., 2009) and exists as a homodimer, in which each monomer is composed of two domains linked together by a proline-rich linker (Newman et al., 2005). The C-terminal domain has an α/β hydrolase fold and harbors the catalytic center for the EH activity. The N-terminal domain has a different α/β fold and belongs to the haloacid dehalogenase enzyme superfamily, functioning as a phosphatase (Cronin et al., 2003; Newman et al., 2003). The two monomers form a domain-swapped dimer in which the hydrolase domain of one monomer binds to the phosphatase domain of the other monomer (Decker et al., 2009). A recent report shows that dimerization is required for sEH hydrolase activity (Nelson et al., 2013). A mutation on the dimerization interface was shown to decrease the hydrolase activity of human sEH by disrupting dimerization. sEH has been established as a mediator of physiological processes rather than detoxification of xenobiotics (Morisseau and Hammock, 2005). The C-terminal sEH is involved in the metabolism of endogenous epoxides from arachidonic acid and other unsaturated fatty acids. In particular, arachidonic acid-derived epoxyeicosatrienoic acids (EETs) are the most studied substrates of sEH. EETs are important signaling molecules that regulate multiple biological processes such as vasodilation, inflammation, cell proliferation, angiogenesis and analgesia (Spector and Norris, 2007). The breakdown of EETs by sEH is believed to be a deactivation process. Inhibition of EET metabolism by selective sEH inhibitors results in anti-inflammatory, anti-hypertensive, neuroprotective, and cardioprotective effects in animal models (Morisseau and Hammock, 2013). The N-terminal phosphatase acts on lysophosphatidic acids and isoprenoid phosphates (Enayetallah and Grant, 2006; Morisseau et al., 2012). Lysophosphatidic acid is well-recognized as an important signaling molecule, affecting many different cellular responses, such as proliferation, survival, cytoskeletal changes and calcium influx (Lin et al., 2010). Isoprenoid phosphates play a major role in the regulation of cholesterol levels and cell signaling (Holstein and Hohl, 2004). Therefore, it is possible that the phosphatase activity of sEH also has a physiological role. On the other hand, mEH is an endoplasmic reticulum (ER)-resident enzyme that is attached to the cytosolic surface membrane of the ER through a single N-terminal transmembrane anchor (Holler et al., 1997). Unlike sEH, mEH exists as a monomer (Decker et al., 2009). It is best known for its role in foreign compound metabolism. The substrate selectivity of mEH is broader and quite different from that of sEH. mEH substrates include epoxides derived from environmental toxins such as styrene, butadiene, benzene, naphthalene, anthracene and other polycyclic aromatic hydrocarbons (Decker et al., 2009) as well as clinical drugs, such as phenytoin and carbamazepine (Fretland and Omiecinski, 2000). Its exceptionally broad substrate spectrum against a diverse group of epoxides demonstrates its importance as a xenobiotic metabolizing enzyme. Of importance, mEH is the 4 only hydrolase known to participate in the metabolism of xenobiotic-derived epoxides (Newman et al., 2005). In the following sections, mEH will be addressed in this review for its function, regulation and genetic polymorphism.

The physiological functions of mEH The principle physiological function of mEH involves the metabolism of epoxide intermediates generated by cytochrome P450 enzymes. These epoxide intermediates are often highly reactive and may interact with cellular macromolecules (DNA, RNA and proteins) and have the potential to participate in mutagenic and carcinogenic processes. The hydrolysis of an epoxide generally results in an increased water solubility of the metabolites and the termination of its genotoxic potential. Therefore, mEH is usually thought to play a pivotal role in protection against the toxicity of reactive epoxide intermediates. However, under some circumstances, mEH is involved in the bioactivation of procarcinogens, such as polycyclic aromatic hydrocarbons (PAHs). PAHs are ubiquitous environmental pollutants formed as a result of incomplete combustion of organic matter and are present in automobile exhaust and smoked or charbroiled food. They are also found in cigarette smoke and are established lung carcinogens. However, PAHs are procarcinogens in that they must be metabolically activated to highly reactive bay-region diol-epoxides to cause their mutagenic and carcinogenic effects. For example, in the bioactivation pathway of 7,12-dimethylbenz[a]anthracene (DMBA), P450 CYP1B1 oxidizes DMBA to the 3,4-epoxide which is hydrolyzed by mEH to DMBA-3,4-diol. DMBA-3,4- diol is further oxidized by either CYP1A1/1B1 to the principal ultimate carcinogenic metabolite, DMBA-3,4-diol- 1,2-epoxide, that is capable of interacting with DNA with higher potency than its parental compound. The requirement of mEH involvement in bioactivation of PAHs was established in mEH-null mice which were highly resistant to DMBA-induced carcinogenesis, compared with wild type mice (Miyata et al., 1999). In addition to the detoxification and bioactivation of xenobiotics, mEH may play a physiological role in steroid metabolism (Hattori et al., 2000), bile acid transport (Alves et al., 1993) and vitamin K1 oxide reduction (Guenthner et al., 1998). However, the fact that mEH null mice do not show any obvious phenotype when compared to the wild type controls indicates that the potential endogenous roles of the enzyme need to be further studied.

Gene organization of human mEH gene (EPHX1) The human EPHX1 gene is located on Chromosome 1q42.1 and composed of 9 exons, of which exons 2- 9 are protein coding (Figure 1-2). The human EPHX1 protein is composed of 455 amino acids and has a molecular weight of ~50kDa. EPHX1 protein sequences are highly conserved among human, rat and rabbit as well as other vertebrates (Hassett et al., 1994; Yang et al., 2009). Interestingly, the transcription of the human EPHX1 gene is initiated from two different genomic regions, resulting in two unique transcripts (Liang et al., 2005). The two transcripts, denoted E1 and E1b respectively, differ from each other only at the non-coding first exon. Consequently, the two transcripts encode a same protein since the protein translation start site is located on 5

Exon 2. The E1 and E1b transcripts exhibit a tissue-dependent expression pattern (Liang et al., 2005; Yang et al., 2009). The E1 transcript is almost exclusively expressed in liver and only very low levels of E1 are detected in the ovary and small intestine, while the E1b transcript is expressed ubiquitously in all tissues examined (Liang et al., 2005; Yang et al., 2009). The non-coding regulatory regions of the EPHX1 gene are not well conserved between primate species and rodents (Yang et al., 2009). In particular, the proximal E1 promoter and its corresponding exon 1 are highly conserved in primates and placental mammals, while the far upstream E1b proximal promoter and the E1b exon 1 are only highly conserved among primates. These findings strongly suggest that primates and rodents have evolved different regulatory mechanisms for mEH expression. The EPHX1 gene is expressed in all human tissues that have been evaluated. It is highly expressed in the liver and, to a lesser extent, other organs such as lung, kidney, intestine, brain, prostate, heart and tests (Coller et al., 2001; Liang et al., 2005; Newman et al., 2005; Yang et al., 2009).

Polymorphisms and their clinical implications The genomic region encompassing the human mEH gene is highly polymorphic (Figure 1-3A). With data retrieved from dbSNP database (Build 137) using UCSC Genome Browser, we identified 178 validated SNPs within the 35.4kb EPHX1 gene region. Among these SNPs, two major polymorphic loci have been well described for the EPHX1 gene affecting the amino acid composition of mEH at positions 113 (SNP rs1051740, Figure 1- 3B) in exon 3 and 139 (SNP rs2234922, Figure 1-3C) in exon 4 (Hassett et al., 1994). The polymorphism (TC) at nucleotide position 337 in exon 3 leads to tyrosine (Tyr or T, TAC) to histidine (His or H, CAC) substitution at amino acid position 113. A polymorphism (AG) at nucleotide 416 in exon 4 results in a histidine (His or H, CAT) to arginine (Arg or R, CGT) substitution at amino acid residue 139 (Hassett et al., 1994). According to dbSNP database (Build 137), the frequencies of H113 and R139 alleles are 31.6% and 18.5%, respectively. The haplotype frequency varies among studies with different data sources. In the initial study by Hassett et al., the haplotype frequencies at residue 113 were 36.2% for Tyr/Tyr, 56.2% for Tyr/His and 7.6% for His/His, and the haplotype frequencies at residue 139 were 58.7% for His/His, 36.7% for His/Arg and 4.6% for Arg/Arg (Hassett et al., 1994). In another study, the average distribution of each genotype was 48.9% for Tyr/Tyr, 37.7% for Tyr/His and 13.4% for His/His at residue 113 in exon 3, and 65.3% for His/His, 30.7% for His/Arg and 4.0% for Arg/Arg at residue 139 in exon 4 (Lee et al., 2002). The association of disease incidence with human mEH genetic polymorphisms has been intensely investigated. In a pooled analysis of eight studies with a total of 986 lung cancer patients and 1633 controls, exon 3 His/His genotype showed a significant decrease in lung cancer risk (OR=0.70, 95% CI = 0.51-0.96) after adjustment for age, sex, smoking and study center, indicating a protective effect of exon 3 polymorphism (Lee et al., 2002). However, there was no apparent association between lung cancer and polymorphism at exon 4. Another systematic review and meta-analysis of 13 published epidemiologic studies indicated that, compared with the Tyr/Tyr genotype, the His/His genotype of EPHX1 polymorphism at exon 3 was associated with decreased risk of lung cancer (odds ratio = 0.65; 95% confidence interval = 0.44–0.96) in lung cancer risk among whites. 6

The Arg/Arg genotype of EPHX1 polymorphism at exon 4 was associated with a modest increase in risk of lung cancer in white populations (OR = 1.22; 95% CI = 0.79 –1.90), as compared with the Tyr/Tyr genotype (Kiyohara et al., 2006). These results further imply the protective effect of exon 3 polymorphism. The His139Arg polymorphism was indicated to be significantly associated with decreased colorectal cancer risk (Arg/His vs. His/His, OR = 0.90, 95%CI = 0.83–0.98) (Liu et al., 2012). A meta-analysis of mEH genetic polymorphism and risk of hepatocellular carcinoma indicated that the His/His homozygote genotype at exon 3 was significantly associated with increased risk of hepatocellular carcinoma (OR = 1.65, 95% CI = 1.07–2.54, p = 0.02), while individuals carrying the Arg139Arg exon 4 mEH genotype had no association with increased or decreased risk of hepatocellular carcinoma (Zhong et al., 2013). Meta-analysis of the association between mEH polymorphisms and the susceptibility to esophageal cancer showed that EPHX1 Tyr113His and His139Arg polymorphisms were not associated with esophageal cancer risk (Zhao et al., 2013). These results suggest that the impact of EPHX1 polymorphisms on cancer susceptibility is complex, as is the disease of cancer, and certain of the EPXH1 genotypes may have a risk association, albeit almost certainly along with other genetic factors contributing to disease phenotypes. Despite abundant epidemiologic studies on the association between mEH polymorphisms and risk of human diseases, detailed analysis of functional differences between EPHX1 polymorphic variants is scarce. In the initial characterization of mEH variants, the variant containing His113/His139 showed 40% less in vitro enzyme activity than the Tyr113/His139 variant, while the Tyr113/Arg139 variant exhibited 25% more activity than the Tyr113/His139 variant. However, when the enzymatic activities of mEH variants were normalized to the corresponding protein levels, there was no significant difference in specific activities of the variant enzymes (Hassett et al., 1994). These results indicated that the substitutions at amino acid residue 113 and 139 did not impart a major effect on the enzymatic activity. The observation that the protein levels of mEH appeared dependent on the variant genotypes suggested that polymorphisms at residues 113 and 139 might affect the mEH protein stability. However, further studies on post-transcriptional regulation of the human mEH variants indicated that EPHX1 polymorphisms at exon 3 and exon 4 did not have a major impact on either mRNA stability or translational efficiency (Laurenzana et al., 1998). The protein half-lives for the polymorphic variants did exhibit a modest difference, suggesting that polymorphic amino acid substitution may result in altered protein stability. However, the differences were not statistically significant and therefore perhaps not likely to result in differential enzymatic activities among the mEH protein variants. In conclusion, the answers to the question as to how mEH polymorphisms appear to differentially associate with disease susceptibility remain largely unknown. In addition to the two polymorphisms that affect the amino acid composition of mEH, non-coding polymorphisms also exist in the upstream and downstream regions of the human mEH gene, as shown in the UCSC genome browser-derived Figure 1-3A. In early studies, 7 polymorphic sites within the proximal E1 promoter region were identified (Raaka et al., 1998). Reporter assays demonstrated that these polymorphic motifs differentially regulated E1 promoter activity, suggesting that genetic variation in the promoter region of the mEH 7 gene could be a factor contributing to mEH expression level variation in human populations. The biological functions of other non-coding polymorphisms have not yet been explored.

Transcriptional regulation of Human mEH gene The regulation of mEH has been well studied in animals. In rodents, mEH can be highly induced by a variety of compounds, such as phenobarbital (PB), 3-methylcholanthrene (3-MC), polychlorinated biphenyls, trans-stilbene oxide, peroxisome proliferators, radiation, heavy metals, and certain steroids, including estroxide and its precursor estratetraenol (Newman et al., 2005). However, the chemically-induced expression of human EPHX1 appears modest. In primary human hepatocytes from several donors, treatments with common inducers, such as PB, butylated hydroxyanisole, dexamethasone, Arochlor 1254 and β-naphthoflavone increased mEH expression by 2-3 fold (Hassett et al., 1998). In addition, two benzo[a]pyrene metabolites, benzo[a]pyrene 1,6- quinone and benzo[a]pyrene 3,6-quinone, were capable of inducing mEH expression in MCF-10A human mammary epithelial cells to a similar extent (Burchiel et al., 2007). Some chemopreventative agents and antioxidants also induce mEH expression. For example, mEH levels were increased by tert-butylhydroquinone in human MCF-10A cells (Burchiel et al., 2007) and oltipraz in rat liver (Merrell et al., 2008). Although the functional Nrf2-binding DNA motif, antioxidant response element, is not found in the promoter of mEH gene, some studies with Nrf2 null mice suggest that the transcription factor Nrf2 regulates mEH expression induced by these types of chemicals (Hu et al., 2006a; Kensler et al., 2007; Ramos- Gomez et al., 2001). Constitutive mEH mRNA expression in multiple tissues is lower in Nrf2-null mice than in wild-type mice. Furthermore, Nrf2 activators, sulforaphane, oltipraz, and 3H-1,2-dithiole-3-thione (D3T), induce mEH expression in wild-type but not in Nrf2-null mice, demonstrating that Nrf2 is indispensable for chemical- mediated induction of mEH (Kwak et al., 2001; Ramos-Gomez et al., 2001; Thimmulappa et al., 2002). However, a recent investigation showed that hepatic mEH mRNA expression was not increased in Keap1-null mice (Reisman et al., 2009). This finding suggests that other regulatory mechanisms, aside from Nrf2 nuclear translocation, are important for mEH mRNA induction in the mouse liver.

Tissue-specific regulation of human EPHX1 gene by the selective use of alternative promoters The human EPHX1 gene is subject to regulation by alternative promoter usage (Liang et al., 2005; Yang et al., 2009). EPHX1 transcription is driven by two unique promoters that are located at different genomic regions approximately 15.2 kb apart (Figure 1-2). The two promoters cause tissue-specific expression of human EPHX1. The proximal E1 promoter is exclusively active in the liver, while the far upstream E1b promoter drives mEH expression in all tissues, including the liver. The existence of alternative promoters provides the possibility of using differential regulatory mechanisms to control tissue-specific mEH expression. The E1 promoter is located about 3.4 kb upstream from Exon 2. Promoter reporter analysis with stepwise deletions of upstream regions of E1 suggested that the core 208 bp upstream region contributes the highest level of transcriptional activity for the E1 promoter under basal conditions (Liang et al., 2005). Further sequence 8 analysis of this region revealed potential transcription factor-binding motifs for GATA, HNF3 and CCAAT binding factors. EMSA analysis indicated that both GATA-4 and HNF3 bind to their identified binding sites in the E1 proximal promoter (Liang et al., 2005; Zhu et al., 2004a). Site-directed mutagenesis and transient transfection assays showed that GATA-4 plays a critical and positive role in regulating EPHX1 expression, while HNF3 downregulates the E1 promoter activity by antagonizing the function of GATA-4 on the E1 promoter (Liang et al., 2005). The CCAAT/enhancer-binding protein α (C/EBPα) was also established as a critical regulator of the E1 promoter. Results of EMSA and ChIP assays demonstrated that C/EBPα does not bind to DNA directly, but regulates the E1 promoter activity by interacting with the CCAAT box binding factor/nuclear factor Y (NF-Y) that itself binds to the CCAAT motif in the E1 promoter (Zhu et al., 2004b). In addition, a putative Nrf2-binding ARE motif was identified in the E1 proximal promoter (Zhu et al., 2004a; Zhu et al., 2004b). However, no further characterization has been published. The E1b promoter is located approximately 18.6 kb 5’ from exon 2. Since its discovery in 2005, there exists only limited new information regarding the regulatory mechanisms that control E1b transcription. Promoter reporter analysis with stepwise deletions of upstream regions of E1b suggested that its most proximal 320 bp region contributes the highest transcriptional activity for the E1b promoter under basal conditions, indicating the presence of proximal and core E1b promoters (Liang et al., 2005). A recent study identified the presence of transposable elements in a ~2kb upstream region of E1b, the presence of which decreases basal E1b promoter activity. Luciferase reporter assays demonstrated that these elements rendered a significant reduction on E1b promoter activity in 4 different cultured human cell lines (Yang et al., 2009). The mechanisms accounting for how the E1 and E1b promoters are selectively used to direct tissue- specific expression is unclear, at least in part due to the fact that the identity of transcription factors that regulate E1b and E1 promoter activities are largely unknown. As discussed above, the transcription factors C/EBPα and HNF3 were reported to regulate E1 basal promoter activity. These latter factors are liver-enriched and have been reported to control liver-specific transcription for other genes (Antonson and Xanthopoulos, 1995; Kaestner, 2000). Therefore, it is possible that they play a role in for selective use of the E1 promoter and controlling the liver-specific constitutive expression of E1 transcription. Compared to the E1 proximal promoter, the 5’-flanking regions of the E1b promoter are GC rich and host a putative CpG island (CGI). CGIs have been identified in the promoters of all house-keeping genes and in approximately 40% of tissue-specific genes (Illingworth and Bird, 2009). These findings suggest that the CGI in the E1b promoter region may play a role in the regulation of E1b transcript – a transcript that exhibits a ubiquitous expression pattern across tissues. This discovery of a CGI in the proximal promoter region of E1b also indicates a possibility that GC-rich DNA motif-binding transcription factors, such as Sp1, may be involved in the regulation of E1b. In addition, the presence of a CGI in the E1b promoter raises the possibility that epigenetic modification on E1b or E1 promoter regions may contribute to the tissue-specific selection of alternative promoters. Since promoter methylation at CG dinucleotides leads to gene silencing (Hatchwell and Greally, 2007), differential 9 promoter methylation statuses can result in differential alternative promoter activity. The proximal and core regions of E1b promoter lack DNA methylation, which indicates that E1b promoter is constitutively active and correlates with its ubiquitous expression pattern (Cai, Hedrich, Su; Omiecinski laboratory, unpublished data). In contrast, the E1 proximal and core promoters are methylated in non-liver cells, but not in hepatoma cell lines and human normal hepatocytes (Omiecinski laboratory, unpublished data), indicating that the E1 promoter is inactive in non-liver cells. This observation is consistent with the liver-restricted expression pattern of the E1 transcript. Most CGIs are hypomethylated (Illingworth and Bird, 2009). Therefore, the CGI in the E1b promoter appears to play a role in maintaining the active status of the E1b promoter through its largely unmethylated status. Since little is known about the regulatory mechanisms that control E1b transcript expression, additional studies are necessary to explore the mechanistic basis of human EPHX1 E1b variant regulation and to better understand how xenobiotics may contribute to transcriptionally regulate mEH expression and mEH involvement in mediating the toxic effects associated with xenobiotic exposures. To this end, the following section reviews various classes of transcriptional regulatory elements and the current knowledge of how they function to provide a more comprehensive understanding of the experiments which were performed to delineate the molecular basis of human EPHX1 gene regulation.

TRANSCRIPTIONAL REGULATION OF GENE EXPRESSION In higher eukaryotes, proper control of gene expression in response to developmental, differentiating or environmental signals requires three types of regulatory DNA elements that are cis-acting: core promoters, proximal promoter elements and distal regulatory elements (Figure 1-4) (Maston et al., 2006). Core promoters function by providing binding sites for the basic transcription factors, facilitating the proper assembly of preinitiation complexes (PIC) and defining the start site of mRNA synthesis (Butler and Kadonaga, 2002). Core promoters are responsible for initiation of transcription and basal levels of gene expression. Proximal promoter elements are located immediately upstream of the core promoter. They typically reside approximately 100-200 bp upstream of the transcription initiation site and contain transcription factor binding motifs. Transcription factors bind to the proximal promoter elements and stimulate the rate of transcription by promoting PIC formation or/and modifying chromatin structure (Lodish, 2004; Maston et al., 2006). Distal regulatory elements, on the other hand, can be located either upstream or downstream of the promoter and may work from great genomic distances (Noonan and McCallion, 2010). Not only do the distal sites fine-tune promoter activity, but also play an important role in establishing the appropriate levels and expression patterns of their cognate genes (Merika and Thanos, 2001).

The core promoter The core promoter is classically defined as the minimal DNA fragment that is sufficient to direct accurate initiation of transcription by the RNA polymerase II machinery (Butler and Kadonaga, 2002). It is typically found within 40-50 bp upstream or downstream of the transcriptional start site (TSS) and can encompass several distinct 10 core promoter sequence elements for the binding of regulatory proteins. Some commonly found core promoter elements include the TATA box (consensus, TATAWAAR; located about 25 to 30 bp upstream of the TSS), the initiator element (Inr; consensus, YYANWYY; the A residues at position +1 correspond to the TSS), the TFIIB recognition element (BRE; consensus, SSRCGCC; located immediately upstream of some TATA boxes), and the downstream core promoter element (DPE; consensus, RGWYVT; located between positions +28 and +33 relative

to the A+1 position in the Inr) (Smale and Kadonaga, 2003). Other elements are less common, such as the downstream core element (DCE) (Lewis et al., 2000), motif ten element (MTE) (Lim et al., 2004), and the X-gene core promoter element (XCPE) (Anish et al., 2009; Tokusumi et al., 2007). These motifs each have specific functions related to the transcription process. It is important to note that there are no universal core promoter elements present in all core promoters (Smale and Kadonaga, 2003). The formation of PIC does not depend on one single element, such as a TATA box. Instead, many of the core promoter elements can interact with TFIID components to stabilize the PIC (Maston et al., 2006). The TATA box was the first described and represents the best-studied core promoter element (Breathnach and Chambon, 1981). The TATA box binds the TATA-binding protein (TBP), which is a key subunit of transcription factor TFIID. The TFIID itself is a multi-subunit protein that consists of TBP and approximately a dozen TBP-associated proteins (TAFs) (Thomas and Chiang, 2006). With TATA box-dependent core promoters, a stable PIC can be formed in vitro by orderly association of the basal transcriptional factors: TFIID/TFIIA, RNA polymerase II/TFIIF, TFIIE, and then TFIIH (Butler and Kadonaga, 2002). For TATA-free promoters, accurate transcription of genes is mediated by a TFIID-associated TAF binding to other core promoter elements. The TAF1 and TAF2 subunits contact Inr, the TAF1 subunit recognizes DCE, and TAF6 and TAF9 subunits interact with DPE (Maston et al., 2006).

Focused versus dispersed core promoters There are two types of core promoters that show distinct distribution patterns of transcription initiation sites: focused and dispersed core promoters (Juven-Gershon et al., 2008). In focused core promoters, transcription initiates at a single nucleotide or within a narrow region of several nucleotides (Figure 1-5). Whereas, in dispersed core promoters, there are multiple transcription start sites over a broad region of about 50 to 100 nucleotides (Figure 1-5). As a result, such genes give rise to mRNAs with multiple alternative 5’ ends. In vertebrates, only about 30% of core promoters are focused core promoters while the majority of genes have dispersed core promoters. It generally appears that focused promoters are associated with highly regulated, cell-specifically expressed genes, while dispersed promoters serve as promoters for many house-keeping or commonly expressed genes (Juven-Gershon and Kadonaga, 2010). The two types of core promoters exhibit different promoter structures (Carninci et al., 2006; Sandelin et al., 2007). Analyses of focused core promoters have found common core promoter elements such as, the TATA box, BRE, Inr, MTE, DPE and DCE (Juven-Gershon and Kadonaga, 2010). In contrast, dispersed core promoters 11 do not contain these core promoter elements and typically contain CpG islands, which are stretches of GC-rich DNA sequences.

CpG islands (CGI) The CGIs are 0.5-2 kb stretches of DNA sequences that are GC-rich and possess a relatively high density of CpG dinucleotides that are mostly unmethylated (Bird, 1987; Smale and Kadonaga, 2003). In contrast, the majority of CpG dinucleotides in other genomic regions are methylated by DNA methyltransferases, which add a methyl group to the fifth carbon of the cytosine nucleotide. The methylated cytosine in CpG dinucleotides spontaneously deaminates to thymine resulting in the under representation of CpG dinucleotides in genomes (21% of that expected in the human genome) (Bird, 2002). CGIs were defined by Gardiner-Garden and Frommer as a DNA sequence with a length of at least 200 bp, a G+C content in that region of greater than 50%, and a ratio of observed versus expected number of CpG dinucleotides above 0.6 (Gardiner-Garden and Frommer, 1987). Another definition was also suggested with more stringent thresholds by Takai and Jones (Takai and Jones, 2002). It defines CGI as regions with a minimum length of 500 bp, a G+C content of 55% or greater, and an observed CpG/expected CpG ratio of 0.65. These CGI criteria effectively exclude false positives from short repetitive sequences, such as Alu elements (Illingworth and Bird, 2009). CGIs associate with the promoter region of 70% of human genes (Saxonov et al., 2006). CGI-associated promoters are the most common promoter type in the vertebrate genome and are found in all house-keeping genes and approximately 40% tissue-specific genes (Illingworth and Bird, 2009). CGI promoters typically lack TATA and other core promoter elements, but contain multiple Sp1 binding sites (Butler and Kadonaga, 2002). Another feature of CGI promoters is the presence of multiple transcription start sites over a stretch of about 50 to 100 nucleotides within them (Juven-Gershon et al., 2008; Smale and Kadonaga, 2003). In contrast, TATA-box- containing promoters have either a single transcription start site or tightly clustered sites within several nucleotides. Although the broad distribution of TSSs in CGI promoters is well documented, the mechanism underlying the distributive patterns of transcription initiation from them is not quite clear (Müller et al., 2007). One of the potential explanations is that several pre-initiation complexes (PIC) may be formed at different sites in these promoters. The ubiquitous transcription factor Sp1 facilitates the assembly of PIC by interacting with TFIID (Safe and Abdelrahim, 2005). Hence, multiple Sp1 binding sites within the CGI promoters serve as a deck for Sp1 to form multiple PICs and initiate transcription from an array of TSSs (Smale and Kadonaga, 2003). In addition, studies of TSS usage with genome-wide approaches have shown tissue-dependent utilization of TSSs, indicating that TSS selection within CGI promoters can be regulated (Kawaji et al., 2006). Many studies show that CGIs contain binding sites for transcription factors that recognize GC-rich motifs. One common feature of CGIs is the presence of multiple binding sites for Sp1, which allow for TBP recruitment and transcriptional activation (Butler and Kadonaga, 2002; Smale and Kadonaga, 2003). Other studies show that Sp1 can bind to the CGI-containing promoters of human sEH and secretin genes and regulate their expression (Lee et al., 2004; Zhang et al., 2010). Computational analysis of GC-rich promoters indicated an 12 overrepresentation of binding motifs for Sp1, Nrf-1 (nuclear respiratory factor-1), E2F and ETS (Landolin et al., 2010). Consistently, E2F4 was shown to bind CGIs in the promoters of its target genes (Weinmann et al., 2002). Similarly, ETS, Nrf-1, BoxA, Sp1, CRE, and E-Box motifs are enriched in the CGI-containing promoters of housekeeping genes (Rozenberg et al., 2008). Subsequently, EMSA assays demonstrate that methylation of the CpG in the ETS, Nrf-1, and Sp1 motifs prevents transcription factor binding. CGI methylation of gene promoters plays a functional role in transcriptional regulation. CGIs are typically hypomethylated. However, methylation of CpG sites within CpG islands in the promoter region of genes is observed during normal physiological conditions and plays a key role in X chromosome inactivation and genomic imprinting (Latham, 1996; Reik, 2007). In addition, DNA methylation on CGIs results in silencing of the associated promoters (De Smet et al., 1999; Zhang et al., 2010). One simple explanation is that CpG methylation impairs the motif binding ability of transcription factors, resulting in promoter inactivity. For example, methylation of CpG dinucleotides within the promoter region has been shown to lead to transcriptional repression of a number of genes regulated by Sp1 and Sp3 (Chan et al., 2004; Lee et al., 2004; Zelko et al., 2010; Zhang et al., 2010). CpG methylation is also reported to inhibit the binding function of CRE, ETS, Nrf-1, E-Box, AP2 and CTCF motifs (Bell and Felsenfeld, 2000; Choi et al., 2004; Comb and Goodman, 1990; Gaston and Fried, 1995; Prendergast and Ziff, 1991; Tate and Bird, 1993; Weih et al., 1991). CGI methylation is well documented in cancer. Gene promoter CGI methylation is associated with the silencing of tumor-suppressor genes and tumor-related genes (Baylin et al., 1998). Aberrant methylation of several tumor-suppressor and tumor-related genes such as P16, COX-2, RASSF1A, hMLH1, and SOCS1 is frequently observed in human cancers (Esteller, 2002). However, it is unclear whether DNA methylation causes gene silencing or is acquired at already silenced genes. Genome-wide studies have demonstrated that many of the CGIs that acquire aberrant methylation in cancer are not associated with tumor suppressor genes (Keshet et al., 2006; Ruike et al., 2010).

Proximal Promoter Elements Proximal promoter elements are located in the region immediately upstream (up to several hundred base pairs) of the core promoter (Figure 1-4) (Maston et al., 2006). They typically serve as binding sites for activators to stimulate transcriptional activity of PIC on the core promoter, which is low at default. In addition, these proximal promoter elements can serve as tether elements for distant enhancers, enabling these enhancers to interact with the core promoter (Calhoun et al., 2002; Nolis et al., 2009; Su et al., 1991). Two typical elements associated with proximal promoters are GC box and the CAAT box (Epstein, 2003). GC boxes contain the sequence GGGCGG and bind GC-rich DNA motif-binding transcription factors, such as Sp1/Sp3 (Wierstra, 2008). CAAT boxes contain the sequence CCAAT. Several transcription factors can potentially bind to CCAAT or similar sequences, such as the CCAAT/enhancer-binding proteins (C/EBPs), the CCAAT transcription factor (CTF, also known as nuclear factor-1, NF-1), and the CCAAT box binding factor/nuclear factor Y (NF-Y) (Mantovani, 1998; Mantovani, 1999). 13

Transcription factor Sp1 Transcription factor Sp1 belongs to the Sp/Kruppel-like factor (Sp/KLF) superfamily (Wierstra, 2008). It possesses three Cys2His2 zinc fingers and binds GC-rich motifs with the consensus sequence 5’-G/T-GGGCGG- G/A-G/A-C/T-3’ or 5’-G/T-G/A-GGCG-G/T-G/A-G/A-C/T-3’ (Figure 1-6). In addition to the three zinc finger DNA-binding domain, Sp/KLF proteins possess two glutamine-rich modules that together make up their transcriptional activation domains (AD). Adjacent to the two ADs is a serine/threonine (S/T)-rich module that may be targeted for posttranslational modifications (e.g. phosphorylation or acetylation). Sp1 is expressed ubiquitously in various mammalian cells and is implicated in the transcription of many genes that lack a TATA-box and contain GC boxes in their promoter, particularly housekeeping genes and those involved in cell growth and development, such as cyclin D1, E2F, c-fos, TGF-α (Black et al., 2001; Marin et al., 1997; Safe and Abdelrahim, 2005). In addition to initiating transcription from TATA-less promoters, Sp1 is also involved in the activation of the distal enhancer elements of many genes by regulating chromatin looping (Bouwman and Philipsen, 2002; Deshane et al., 2010; Nolis et al., 2009; Suske, 1999). Sp1 interacts directly with chromatin-modifying factors, such as p300, SWI/SNF, and histone deacetylases (HDACs) (Kadam and Emerson, 2003; Suzuki et al., 2000; Zhao et al., 2003). In addition, Sp1 transactivates synergistically with a large variety of transcription factors, such as, c-MYC, E2F-1, HNF-4, AP-2, NF-YA, c-Jun/c-Fos, p53, GATA-1, NFκB (Safe and Kim, 2004; Solomon et al., 2008; Wierstra, 2008). These transcription factors can synergize either by binding DNA cooperatively or by recruiting the basal transcription machinery synergistically. Moreover, some transcription factors, such as AP-2 and c-Jun, activate Sp1-mediated transactivation by directly interacting with DNA-bound Sp1, but not binding to DNA (Kardassis et al., 1999; Pena et al., 1999; Zhang et al., 2008). The zinc-finger DNA binding domains of Sp/KLF proteins are highly conserved (Kaczynski et al., 2003). For example, the DNA binding domain of Sp1 and Sp3 share over 95% similarity. DNA-binding studies have shown that most Sp/KLF members have similar affinities for GC-rich motifs. Competition for DNA binding for some of Sp/KLF family members has been demonstrated. For example, Sp1 competes for the same sites with Sp3, KLF4, KLF6, KLF9 and KLF13 (Grande et al., 2012; Kaczynski et al., 2003).

Distal regulatory elements There are four types of cis-regulatory elements that regulate gene transcription over a considerable genomic distance from the core promoters of the target genes. They are enhancers, silencers, insulators, and locus control regions (LCRs) (Figure 1-4) (Maston et al., 2006; Noonan and McCallion, 2010). Enhancers are cis-acting DNA sequences that can enhance promoter activity in a location- and orientation-independent manner (Maston et al., 2006; Noonan and McCallion, 2010). They are commonly found proximal to the promoter of their target genes (e.g., within the introns of the genes) or distal over a large genomic distance. One of the extreme examples is a limb bud enhancer for the mouse Sonic hedgehog (Shh) gene, which is located more than 1Mb from the Shh gene promoter (Lettice et al., 2003). An enhancer usually harbors binding sites for multiple transcription factors and these transcription factors act cooperatively to regulate the tissue- and developmental stage-specific patterns 14 of their target genes (Maston et al., 2006). Enhancers are also highly modular, such that a single promoter can be regulated by different enhancers at different times, or in different tissues, or in response to different stimuli (Atchison, 1988; Maston et al., 2006). The precise mechanisms through which enhancers regulate promoter activity over such long physical distances are not well understood (Noonan and McCallion, 2010). It is widely accepted that enhancers act via a chromatin loop which brings enhancers and promoters into close proximity (Bulger and Groudine, 2011). The loop model has received a lot of support from studies of spatial organization of chromatin and physical interactions between genomic elements with chromosome conformation capture (3C) technology (Dekker et al., 2002; Kadauke and Blobel, 2009; Miele and Dekker, 2008). For example, with the 3C technique, the LCR of the murine beta-globin locus was found to be in direct physical contact with the expressed globin gene within the locus (Tolhuis et al., 2002). As to how DNA intervening between enhancers and promoters forms a chromatin loop, several mechanistic models have been proposed (Bulger and Groudine, 2011; Miele and Dekker, 2008; Nolis et al., 2009). It is hypothesized that enhancer-bound complexes migrate along the chromatin fiber by free or facilitated diffusion, or an active scanning mechanism, until they encounter a functional promoter (Hatzis and Talianidis, 2002; Wang et al., 2005). The intervening chromatin between the enhancer and the promoter loops out as the enhancer complex moves along the chromatin fiber toward the promoter. The close proximity between the enhancer and the promoter stimulates the assembly of a more stable PIC on the promoter, causing enhancement of transcription. Given the complexity of the nucleus, and the number of genes and regulatory elements present in it, it seems that the active scanning mechanism is more likely than the free diffusion mechanism. Interestingly, studies have also suggested that RNA polymerase II brings the enhancer and the promoter into proximity by binding to the enhancer and then actively moving along the DNA via a scanning mechanism until reaching a promoter (Koch et al., 2008; Szutorisz et al., 2005). In contrast, silencers are regulatory elements that confer a negative (silencing or repressing) effect on promoter activity of target genes. Silencers contain binding sites for negative transcription factors called repressors. The repressor proteins can function in a few ways (Maston et al., 2006). Firstly, they can block the binding of activators to nearby motifs. Secondly, a repressor may recruit corepressors to establish a repressive chromatin structure, preventing activators from accessing a promoter (Privalsky, 2004). Finally, a repressor may inhibit the assembly of PIC and block gene transcription. Silencers can be present within enhancers or can act as independent modules with binding sites for repressors (Ogbourne and Antalis, 1998). Two classes of silencers are known: silencer elements, which are short, position-independent elements that recruit repressors and actively interfere with the PIC assembly, and negative regulatory elements (NREs), which are position-dependent elements that prevent transcription factors from binding to their respective cis-regulatory elements to indirectly repress promoter activity (Ogbourne and Antalis, 1998). Insulators (also known as boundary elements) are special cis-acting regulatory DNA sequences that flank the regulatory regions of genes and block unwanted interaction between enhancers or repressors and promoters, 15 thus creating boundaries for these regulatory elements and preventing inappropriate gene activation or repression (Bushey et al., 2008). Two types of insulators are known: enhancer-blocking insulators and barrier insulators (Gaszner and Felsenfeld, 2006). The enhancer-blocking insulators block enhancer-promoter communication preventing gene activation. Barrier insulators can prevent the spread of repressive chromatin to other regions and restrict chromatin-mediated gene silencing within the boundaries. In vertebrates, the only known insulator protein is CCCTC-binding factor (CTCF) (Valenzuela and Kamakaka, 2006). Locus control regions (LCRs) are groups of regulatory elements involved in regulating an entire locus or gene cluster (Li et al., 2002). LCRs are typically composed of multiple cis-acting elements, including enhancers, silencers, insulators, and nuclear-matrix or chromosome scaffold-attachment regions, that function together to confer tissue-specific and developmental stage-specific expression to genes. One best-studied example of LCRs is the human β-globin LCR (Levings and Bungert, 2002). The human β-globin locus contains five genes (ε-, Gγ-, Aγ, δ-, and β-globin) that are arranged linearly on chromosome 11 and are expressed in a developmental stage- specific manner in erythroid cells. The human β-globin LCR is composed of 5 domains that exhibit extremely high sensitivity to DNase I in erythroid cells. These domains are located from about 6 to 22 kb upstream of the ε- globin gene and confer high-level globin gene expression at all developmental stages.

Identification of distal enhancer elements One of the most useful methods for identification and characterization of transcriptional regulatory elements is the luciferase reporter assay (Maston et al., 2006). In this assay, the element to be tested for regulatory activity is subcloned into a plasmid upstream of a reporter gene, such as luciferase. The resulting constructs are transfected into cultured cells and the luciferase activity is measured to determine if the test element changes the reporter gene expression. If a genomic segment is confirmed to contain enhancer elements, it can be further characterized by serial deletions and site-directed mutagenesis. Although this approach is convenient and successful in the characterization of the intronic regulatory elements and identification of transcription factors that associate with these elements, the tested enhancer elements are removed from their normal genomic context and put into a plasmid that lacks native chromatin structure. Thus, luciferase reporter assays cannot reveal effects on long-range regulation by enhancers and associated transcription factors. Most genes are thought to be regulated by more than one enhancer. In cases where activation of a gene requires multiple enhancers, alteration of reporter activity may not be observed when only a single enhancer is tested. The development of chromatin immunoprecipitation (ChIP) coupled with microarray (ChIP-chip) or massively parallel sequencing (ChIP-seq) has made possible genome-wide mapping of histone modifications and association between chromatin and transcription factors or other chromatin-bound proteins (Barrera and Ren, 2006). These genome-wide studies suggest that enhancers exhibit characteristic chromatin signatures. ChIP-chip analysis of the chromatin architecture along 44 human loci selected by the ENCODE consortium showed that enhancers are marked by the presence of monomethylation of histone H3 lysine 4 (H3K4Me1), but absence of trimethylation (H3K4Me3). In contrast to enhancers, promoters are marked by H3K4Me3 (Heintzman et al., 16

2007). Further, enhancers are often characterized by binding of the acetyltransferase p300 and DNase I hypersensitivity (Heintzman et al., 2007; Xi et al., 2007). Combinatorial analysis using these distinct chromatin signatures to identify new regulatory elements is very effective (Blow et al., 2010; Ghisletti et al., 2010; Visel et al., 2009). While putative enhancers are identified based on chromatin signatures, one of the challenges in characterizing enhancer function is to determine the target genes they control. The issue is that enhancers can be located quite far from their target genes or even be in the gene body of other genes or on different chromosomes (Lomvardas et al., 2006). In principle, a more direct approach for determining the genes associated with enhancers is to experimentally determine the long-range chromatin interaction between enhancers and their target promoters. This can be accomplished by the chromosome conformation capture (3C) assay and its derivative technologies (Dekker et al., 2002; Dostie et al., 2006; Simonis et al., 2006; Zhao et al., 2006). Chromosome conformation capture (3C) technology allows analysis of the physical interaction between a promoter and a distal regulatory element in the native cellular state. In this assay, interactions between two genomic regions are captured by crosslinking followed by restriction endonuclease digestion and intramolecular ligation of the crosslinked chromatin. The resulting ligated products are analyzed by PCR using primers located in the promoter and the predicted enhancer sequence. The relative abundance of a particular ligation product indicates the frequency with which the two chromatin segments interact. This technology has been further developed to generate a genome- wide interaction map between a given locus (“4C”) or several regions (“5C”) and any other region in the genome (Bulger and Groudine, 2011).

TRANSCRIPTION FACTORS REGULATING DRUG-METABOLIZING ENZYMES The expression of xenobiotic metabolizing enzymes (XMEs) is coordinately controlled by various transcription factors in response to exposure to xenobiotics (Xu et al., 2005). Several key transcription factors include constitutive androstane/active receptor (CAR), pregnane X receptor (PXR), peroxisome proliferator activation receptors (PPARs), nuclear factor-erythroid 2-related factor 2 (Nrf2), aryl hydrocarbon receptor (AhR) and others (Nakata et al., 2006; Tirona and Kim, 2005; Xu et al., 2005). CAR and PXR are activated by a wide range of xenobiotics and collaboratively control the transcription of a broad spectrum of distinct and overlapping genes encoding Phase I, Phase II xenobiotic-metabolizing enzymes, as well as Phase III transporters, such as CYP2A, CYP2B, CYP2C, CYP3A, aldo-keto reductase, carboxylesterase, aldehyde dehydrogenase, UGT, SULT, GST, MDR1 and OATPs (Tolson and Wang, 2010). Although all of the transcription factors are important in regulating the expression of XMEs, specific emphasis has been placed in the following section on Nrf2 because of its central role in regulating the expression of regulating Phase II enzymes and other cellular defensive enzymes such as heme oxygenase-1 (HO-1) in response to electrophiles and oxidative stress (Figure 1-7). There is also large amount information on transcriptional regulation of the mEH gene in rodent models as discussed above. However, it is unknown whether 17 it still is true in human. The following section will attempt to explain the basic biology underlying Nrf2 function as well as the crosstalk between Nrf2 and AhR signaling pathways. Our goal is to try and unravel the potential role of Nrf2 as a mediator of signal transduction cascades required for human EPHX1 gene regulation.

OXIDATIVE STRESS AND NRF2-ARE SIGNALING PATHWAY Nrf2 is a transcription factor that positively regulates the expression of genes encoding antioxidants and xenobiotic detoxification enzymes and confers cytoprotection against oxidative stress in normal cells (Sinha et al., 2013). It is ubiquitously expressed in a wide range of tissue and cell types, but relatively abundant in tissues where detoxification reactions occur routinely, such as liver and kidney, and tissue that are exposed to the external environment, such as, skin, lung and gastrointestinal tract (Itoh et al., 1997a; Motohashi et al., 2002). Under normal physiological conditions, reactive oxygen species (ROS), such as superoxide, hydrogen peroxide, and hydroxyl radicals, are generated as metabolic by-products. ROS can be produced from endogenous sources, such as mitochondria, peroxisomes, and inflammatory cell activation (Babior, 1999; Klaunig and Kamendulis, 2004; Schrader and Fahimi, 2006; St-Pierre et al., 2002); and exogenous sources, including environmental agents, pharmaceuticals and industrial chemicals, such as metals, benzene, benzo[a]pyrene and other PAHs (Goetz and Luch, 2008; Kasprzak, 1995; Klaunig et al., 2010). Oxidative stress resulting from the accumulation of ROS may lead to DNA, protein, and/or lipid damage, changes in chromosome stability, genetic mutation, degeneration of tissues, apoptotic cell death, cellular transformation and cancer (Kaspar et al., 2009). As cellular antioxidant defense, a number of genes are activated to limit oxidative stress to a level not harmful to cells (Dhakshinamoorthy et al., 2000; Jaiswal, 2000). These defensive genes include antioxidant enzymes capable of inactivating ROS, such as, superoxide dismutase, catalase, and glutathione peroxidase (Auten et al., 2006; Ho et al., 2004; Koo et al., 2005), and phase 2 detoxifying (conjugating) enzymes facilitating the excretion of oxidized and reactive metabolites through reduction and conjugation reactions, such as GST, NQO1, γ-glutamyl cysteine synthethase (GCS), and UGTs (Holtzclaw et al., 2004; Klaassen and Reisman, 2010). The signal transduction pathways responsible for sensing oxidative stress and activating the appropriate defense genes includes Nrf2, mitogen-activated protein (MAP) kinase/AP-1, and NFκB pathways (Klaunig et al., 2010). Among these pathways, the Nrf2 pathway is presumably the most important in the cell to protect from oxidative stress (Dhakshinamoorthy et al., 2000). Unlike other xenosensing transcription factors, Nrf2 is not a receptor that binds directly to and is activated by xenobiotics or their metabolites. The activation of Nrf2 is dependent on Keap1 (Kelch-like ECH-associated protein 1) which serves as a sensor of xenobiotic-induced oxidative and electrophilic stress (Figure 1-7) (Zhang, 2010). Under normal homeostatic conditions, Nrf2 is bound to Keap1 and sequestered in cytoplasm. Keap 1 serves as a substrate adaptor for Cul3/Rbx1 E3 ubiquitin ligase forming complex with Nrf2 leading to ubiquitination and proteasomal degradation of Nrf2 (Cullinan et al., 2004). On exposure to oxidative stress, Nrf2 dissociates from Keap1 and translocates into the nucleus. Two possible mechanisms for Keap1-Nrf2 dissociation 18 have been identified (Niture et al., 2010; Osburn and Kensler, 2008). Modification of cysteine residues on the BTB (Broad-complex, Tramtrack, Bric-ά-brac) and intervening domains of Keap1 leads to disruption of the Keap1-Nrf2 interaction and release of Nrf2 (McMahon et al., 2006). Secondly, phosphorylation of Nrf2 enhances the stability and release of Nrf2 from Keap1. This process involves multiple kinase pathways: MAPKs, PKC, PI3K, casein K2 GSK3K/Fyn K and PERK (Bloom and Jaiswal, 2003; Huang et al., 2002; Yu et al., 1999). After translocation into nucleus, Nrf2 forms a heterodimer with small Maf proteins and binds to a consensus sequence called the antioxidant response element (ARE) in the promoter region of genes encoding various antioxidant and phase 2 detoxifying enzymes, including heme oxygenase-1, NAD(P)H:quinone oxidoreductase 1, glutathione reductase, glutathione peroxide, catalase, and glutathione S-transferase (Boutten et al., 2010; Osburn and Kensler, 2008). AREs are cis-acting regulatory elements present in the promoter region of cytoprotective genes targeted by Nrf2. The locations of AREs identified so far are in the upstream region of transcriptional start sites (Chorley et al., 2012; Hayes et al., 2010). These functional AREs contain a core sequence of RTGACnnnGC, where ‘n’ represents any nucleotide (Rushmore et al., 1991). Further characterization of nucleotides surrounding the core ARE sequences showed that these flanking sequences are required for both basal and inducible activity of the reporter gene. As a result, the consensus sequence of AREs was extended to TMAnnRTGAYnnnGCRwwww (Wasserman and Fahl, 1997). Other minor revisions of this consensus sequence have also been suggested (Erickson et al., 2002; Nioi et al., 2003). It is notable that some AREs contain an AP-1 binding site (TPA- response element, 5’-TGAGTCA-3’). Based on the length of ARE sequences and the presence of AP-1 binding site, AREs are divided into 4 categories (Hayes et al., 2010). Class 1 AREs have an extended ARE consensus sequence as well as an embedded AP-1-binding site, while Class 2 AREs have an extended ARE sequence without an AP-1 site. Class 3 and 4 AREs have a minimal core ARE sequence. The difference between Class 3 and 4 is that an AP-1 site is present in Class 3 AREs, but not in Class 4 AREs. AREs in different classes display differential responses to chemopreventive agents. Class 1 and 2 AREs are likely to be more responsive than those in Class 3 and 4. In addition, the activation of AREs in Class 1 and 2 with the extended sequences may be less likely to be subject to negative regulation by various basic-region leucine zipper (bZIP) transcriptional factors. The embedded AP-1 sites in Class 1 and 3 AREs may also respond to other factors that activate AP-1, such as UV radiation and TPA (Hayes et al., 2010).

The structures of Nrf2 and Keap1 Nrf2 is a member of the Cap’n’Collar family of basic leucine zipper genes (CNC-bZIP) (Moi et al., 1994). As shown in Figure 1-8, Nrf2 consists of six conserved domains, Neh1-Neh6, each of which plays a unique role in the functioning of Nrf2 (Hayes and McMahon, 2009; Kaspar et al., 2009). The C-terminal Neh1 and Neh3 domains contain a bZIP region fused to the CNC region and are responsible for nuclear localization, DNA binding, and heterodimerization with small Maf proteins. The Neh4 and Neh5 domains in the central region of Nrf2 are transactivation domains that recruit transcriptional co-activators, such as CBP (CREB-binding 19 protein). The Neh6 and Neh2 domains independently function to control ubiquitin-dependent degradation of Nrf2. The Neh6 is located in the central region of Nrf2 and renders moderate instability of Nrf2 independent of stress signals. The Neh2 domain negatively controls Nrf2 through its ability to interact with Keap1 through DLG and ETGE amino acid motifs. The Neh2 domain contains seven lysine residues that are ubiquitinated in a Keap1- dependent manner (Tong et al., 2006). The Keap1 protein contains an N-terminal Broad-complex, Tramtrack, Bric-ά-brac (BTB) domain, a cysteine-rich intervening/linker domain (IVR) in the central region, and a C-terminal Kelch repeat domain (Figure 1-3). Keap1, via its N-terminal BTB protein-protein interaction domain, binds to Culin3 E3 ubiquitin ligase and, via its C-terminal Kelch repeat domain, binds to the Neh2 domain of Nrf2, leading to the ubiquitination and degradation of Nrf2 through the 26S proteasome(Cullinan et al., 2004). Keap1 is a cysteine-rich protein that contains 25 and 27 cysteine residues in mice and human, respectively. These cysteine residues act as sensors of electrophilic or oxidative stress and are responsible for Nrf2 activation. Studies have shown that cysteine residues, Cys151, Cys273 and Cys288 are crucial for the repression of Nrf2 by Keap1 under basal conditions and their modification by inducers reduces the rate of ubiquitination and degradation of Nrf2, subsequently resulting in activation of Nrf2. Besides these cysteine residues, other cysteine residues are also modified by electrophilic reagents, resulting in Nrf2 activation. Different electrophilic reagents modify a unique set of cysteine residues (Table 1-1). More and more studies support the idea of a multiple sensor mechanism in which activation of Nrf2 involves modification of a combination of cysteine residues within Keap1 (Bryan et al., 2013). Two models have been proposed to describe how modification of Keap1cysteine residues by electrophilic reagents stabilizes Nrf2 (Baird and Dinkova-Kostova, 2011; Taguchi et al., 2011). In the hinge and latch model, Nrf2 binds to a Keap1 homodimer under basal conditions. One of the Keap1 monomers binds to the ETGE motif in the Neh2 domain of Nrf2, while the other monomer binds to the DLG motif. The two motifs have a different affinity with Keap1. The binding affinity of DLG to Keap1 is much lower than that of ETGE (Tong et al., 2006). The two-site binding structure between Nrf2 and Keap1 allows efficient ubiquitination of the lysine residues in the Neh2 domain of Nrf2. Ubiquitinated Nrf2 is subsequently degraded by the proteasome. When cells are challenged by electrophilic or oxidative stress, modification of cysteine residues results in a conformational change in Keap1 and the disruption of the weaker interaction with the DLG motif. Although Keap1 still binds to the ETGE motif of Nrf2, Nrf2 is no longer efficiently ubiquitinated, leading to a reduced rate of Nrf2 degradation. As Nrf2 is still bound to Keap1, newly synthesized Nrf2 is not able to bind Keap1 and therefore can escape the regulation of Keap1, allowing it to accumulate in the cell and translocate to the nucleus. Alternatively, the modification of Keap1 by electrophilic or oxidative stress disrupts the Keap1-Cul3 interaction. Dissociation of Keap1 and Cul3 leads to inhibition of Nrf2 ubiquitination and degradation. 20

Keap1 and Nrf2 mutation in cancers Somatic mutations in Keap1 have been identified in human cancers, such as lung cancer, breast cancer, and gallbladder cancer (Nioi and Nguyen, 2007; Ohta et al., 2008; Shibata et al., 2008a; Singh et al., 2006). These mutations inactivate Keap1 and lead to constitutive activation of Nrf2. About 65% of the Keap1 mutations occur in the Kelch domain, which is responsible for binding to Nrf2. Mutations are also found in the NTR domain (3%), BTB (3%), and IVR (29%) (Mitsuishi et al., 2012). Mutations in the Nrf2 gene were also identified in lung cancer, head and neck, and esophageal squamous cancer (Kim et al., 2010; Shibata et al., 2011; Shibata et al., 2008b). Notable, all the mutations in the Nrf2 gene are within the DLG (43%) and ETGE (57%) motifs (Mitsuishi et al., 2012), which are critical for the binding of Nrf2 to the Kelch domain of Keap1. Mutations in these two motifs result in accumulation and activation of Nrf2. Constitutive activation of Nrf2 due to mutations within the Kelch domain of Keap1 or the Neh2 domain of Nrf2 is associated with many types of cancers, indicating that constitutive activation of Nrf2 can be beneficial for tumor survival (Hayes and McMahon, 2009). Experimental studies show that active Nrf2 can protect tumor cell lines from chemotherapeutic drugs (Shibata et al., 2008b). Because the Nrf2 pathway can protect both normal cells and cancers, the consequences of Nrf2 activation are dependent on the cell context.

The crosstalk between Nrf2 and AhR signaling pathways The transcription factor AhR is a member of the basic helix-loop-helix/Per-Arnt-Sim (bHLH/PAS) family. In its nonliganded state, AhR is sequestered in the cytosol as part of a multiprotein complex containing two molecules of heat shock protein 90 (HSP90), the HSP90 co-chaperone p23, and X-associated protein 2 (XAP2) (Omiecinski et al., 2011). Upon ligand binding (e.g. 2,3,7,8-tetrachlorodibenzo-p-dioxin ,TCDD), the liganded AHR complex translocates to the nucleus where AhR is released from its protein complex and dimerizes with aryl hydrocarbon receptor nuclear translocator (ARNT). The liganded AhR/ARNT complex binds to a dioxin response element (DRE) in the promoters of many XMEs to control their transcriptional activity. Many inducers of XMEs are engaged in the cross-talk of different signaling pathways. Some inducers, such as polycyclic aromatic hydrocarbons (PAH), β-NF and TCDD, induce the expression of phase I and phase II enzymes in a coordinating manner. These inducers are called bifunctional inducers. The induction of XMEs by these inducers requires both AhR and Nrf2 signaling pathways as manifested by AhR-, Arnt- and Nrf2-deficident cells or mice (Ma et al., 2004; Noda et al., 2003; Yeager et al., 2009). PAHs and β-NF interact with the AhR directly to transcriptionally activate phase I enzymes and then undergo transformation by these enzymes to reactive intermediates that lead to oxidative stress and in turn trigger Nrf2 signaling. Besides this indirect interaction between the AhR-XRE and Nrf2-ARE signaling pathways, AhR and Nrf2 interact also by other mechanisms. Mouse Nrf2 gene expression is directly increased by TCDD-mediated AhR activation (Miao et al., 2005). Direct binding of AhR to its target motif, xenobiotic response element (XRE), in the Nrf2 promoter was observed, and siRNA knockdown of AhR abolished Nrf2 induction by TCDD. Interestingly, AhR is also a target of Nrf2 (Shin et al., 2007). Nrf2 genotype affects the basal expression of AhR. Activation of Nrf2 resulted in an 21 enhanced expression of AhR, CYP1A1 and CYP1B1 in Nrf2+/+ mouse embryonic fibroblasts (MEF), but not in Nrf2-deficient MEFs. Further analysis showed that Nrf2 binds to one ARE in the proximal promoter region of AhR. Furthermore, both functional XREs and AREs are present in the regulatory region of some Nrf2 target genes, such as NQO1, GSTs and UGTs (Kalthoff et al., 2010; Köhle and Bock, 2007). The fact that these two responsive elements are located closely to each other raises the possibility of a protein-protein interaction between the two transcription factors or their associated proteins, e.g. coregulators. EMSA analyses have showed that both AhR and Nrf2 bind to the XRE site, as well as the ARE site in the proximal promoter region of human UGT1A10 (Kalthoff et al., 2010).

HYPOTHESES AND AIMS As discussed above, human microsomal epoxide hydrolase (mEH) is an important biotransformation enzyme involved in detoxification and bioactivation of xenobiotics. Its expression level in non-small-cell lung carcinoma (NSCLC) has been linked to cancer risk and disease progression. The general aim of this thesis research is to delineate the molecular mechanisms regulating the basal and inducible expression of human mEH. We anticipate that detailed knowledge of human mEH gene regulation will be of value to assist in better risk prediction in humans exposed to xenobiotic agents, and at a future juncture benefit prevention and treatment of diseases involving aberrant mEH expression or activity. The predominant transcript of human mEH gene in all tissues is the E1b transcript, which is driven by the alternative far upstream E1b promoter. Although the E1b transcript is ubiquitously expressed in all tissues, its expression level varies considerable among tissues. Early studies using luciferase reporter assays revealed that the proximal 300 nucleotides 5’ of the E1b promoter contain important motifs that control the basal transcriptional activity of the E1b promoter. Analysis of ENCODE datasets projecting genome-wide mapping of open chromatin by DNase I hypersensitivity (a marker of open chromatin structure) indicated that an open chromatin structure associates with the proximal promoter region of E1b constitutively in all cell types. Therefore, we hypothesized that the E1b proximal promoter is constitutively active and responsible for the basal expression of the E1b transcript in all cell types. Sequence analysis of this region demonstrated that it is GC-rich and contains a CpG island, indicating a possible link between E1b transcriptional regulation and the GC-rich DNA motif-binding transcription factors, such as Sp1/Sp3. Sp1 and Sp3 are ubiquitously expressed and are linked to expression of many housekeeping genes. They have also been shown to interact with other cell-type specific transcription factors to regulate differential gene expression according to cellular backgrounds. Therefore, we hypothesized that Sp1 and Sp3 are involved as core regulators of the basal and constitutive expression of E1b transcription through their activities and that they serve an important role in regulating the cell-type specific patterning of mEH through their involvement in directing protein-protein interactions with cell-type specific transcription factors. The transcription factor Nrf2 is a master regulator of detoxification enzymes. It induces expression of these enzymes through antioxidant response elements (ARE) present in their promoter regions. Results of 22 microarray analyses have indicated that Nrf2 regulates mEH expression differentially in normal and in Nrf2-null mice. In addition, tBHQ, an Nrf2 activator, was shown to induce mEH expression in human mammary epithelial MCF-10A cells. Therefore, we hypothesized that Nrf2 functions to regulate the inducible expression of E1b through its interaction with AREs within E1b genomic regulatory regions. Genome-wide mapping of open chromatin revealed two open chromatin regions located on the intronic regions between E1b and E1. The status of these regions was cell-type dependent. These two regions are also associated with characteristic chromatin signatures of enhancers, such as H3K4Me1 and H3K27Ac (Figure 1-9). Based on these findings, we hypothesized that these regions are functional enhancers that facilitate the regulation of E1b induction by antioxidants through the Nrf2 pathway. The human mEH gene is subject to regulation by alternative promoter usage. Two transcripts, E1b and E1, are transcribed using alternative promoters. E1b is expressed in all tissues whereas E1 is restrictively expressed in liver. Thus, liver is the only organ expressing both E1b and E1 transcripts. The differential expression profiles of E1b and E1 transcripts indicate they are controlled through distinct regulatory mechanisms. Earlier studies indicated that E1 promoter activity is under combinatorial regulation through several liver-enriched transcription factors. However, little is known regarding the mechanisms of E1b transcriptional regulation. In the course of this research, two putative intronic enhancer elements, localized between the E1b and E1 promoters, were identified by genome-wide mapping of DNase I hypersensitivity sites. This discovery led us to hypothesize that these elements function to enhance the promoter activities of E1b and E1, either selectively or in combination. Previous reports using human primary hepatocyte models indicated that several prototypic chemical agents are capable of inducing human mEH expression at both RNA and protein levels. However, those early investigations did not assess their potential differential impact in regulating E1b vs. E1 transcription. Here we hypothesized that these prototypic inducers differentially regulate E1b and E1 expression. As some of these agents are now recognized as Nrf2 and AhR activators, we further hypothesized that both Nrf2 and AhR signaling pathways are involved in the differential induction of the E1b and E1gene promoters. The overall aim of this thesis investigation was to characterize the molecular regulatory mechanisms underlying the basal and inducible expression of human mEH and to identify potential crosstalk between different signaling pathways involved in these processes. Specifically, the aims of this thesis research were to: 1. Characterize the E1b proximal promoter and to determine the transcription factors involved in regulating the constitutive expression of the E1b transcript (Chapter 2); 2. Characterize the modulation of the E1b expression by antioxidants in lung cancer cells and assess the molecular basis of potential Nrf2-mediated induction of the E1b transcript (Chapter 3); 3. Assess the modulation of the E1b and E1 expression by monofunctional and bifunctional chemical activators and investigate potential mechanisms of crosstalk between the Nrf2 and AhR signaling pathways underlying the differential expression of E1b and E1 transcripts driven by two alternative promoters (Chapter 4). 23

In summary, the transcriptional regulation of human mEH gene is quite complex, involving multiple levels of control. mEH expression is programmed through protein-DNA interactions positioned at multiple regulatory regions localized within the proximal promoters as well as within enhancer elements at distal genomic locations. This research contributes a detailed characterization of these control features and defines specific interactions of selective transcription factors and the domains of their interaction within the mEH genomic locus. These features were further defined as a function of xenobiotic exposure and differential cell type. The results of these studies will contribute ultimately toward the better prediction of tissue-specific chemical toxicities and provide mechanistic basis to better understand and mitigate the risks in humans associated with xenobiotic exposures.

Table 1-1. Modification of Cysteine residues of Keap1 by electrophilic reagents. Each reagent displays a unique modification pattern of Cysteine residues in Keap1. Adapted from Taguchi K., et al. Genes to Cells. 2011, 16(2), 123-140. Keap1 Domain BTB IVR Kelch CTR Electrophilic reagents Dexamethasone 21-mesylate Cys257, 273, 288, 297 Cys613 Biotinylated iodoacetamide Cys151 Cys257, 273, 288, 297, 319 Cys613 N-iodoacetyl-N- Cys196, 226, 241, 257, 288, 319 biotinylhexylenediamine Sulforaphane Cys489, 583 Cys624 8-nitro-cGMP Cys434

Excretion Phase I Phase II Phase III Primary Conjugated Xenobiotics Oxidation Metabolites Conjugation with Metabolites Transporters Reduction cysteine, glycine, (e.g., MRPs and OATPs) Hydrolysis GSH, sulfates, etc

Figure 1-1. Metabolism of xenobiotics by phases I, II, and III of xenobiotic metabolizing systems.

Figure 1-2. The gene structure of EPHX1 and conservation among 46 vertebrate species. Genome browser display of the hg19 human assembly shows that human EPHX1 gene is located on Chromosome 1q42.1 and composed of 9 exons, of which exons 2-9 are protein coding. EPHX1 protein-encoding sequences are highly conserved among human, rat and rabbit as well as other vertebrates. However, the non-coding regulatory regions of EPHX1 gene are not well among human and non-primate vertebrates, in particular, mouse and rat.

Figure 1-3. The genomic region encompassing the human mEH gene is highly polymorphic. A.) Genome browser display of the hg19 human assembly showing the data track of dbSNP database (Build 137). Among these SNPs, two major polymorphic loci have been well described for the EPHX1 gene affecting the amino acid composition of mEH at positions 113 (SNP rs1051740) in exon 3 (B) and 139 (SNP rs2234922) in exon 4 (C).

Figure 1-4. Schematic of a typical gene regulatory region. The promoter is composed of a core promoter and proximal promoter elements. Distal regulatory elements include enhancers, silencers, insulators and locus control regions. These distal regulatory elements regulate the promoter activity by a mechanism of chromatin looping. Adapted from Maston, G. A., et al., Annual Review of Genomics and Human Genetics. 2006, 7(1), 29-59.

Focused versus dispersed core promoters Focused

Dispersed

Figure 1-5. Focused versus dispersed transcription initiation. In focused transcription, there is a single major transcription start site. In dispersed transcription, there are several weak transcription start sites over a broad region of about 50 to 100 nucleotides. Adapted from Juven-Gershon, T. and J. T. Kadonaga, Dev Biol. 2010, 339(2), 225-229.

Figure 1-6. A schematic structure of Sp1 protein. Sp1 contains three zinc fingers, an Sp box, a Buttonhead domain, two transcriptional activation domains (AD), a serine/threonine-rich (S/T) domain and an inhibitory domain (ID). Adapted from Solomon, S. S., et al., Life Sci. 2008, 83(9-10), 305-312. 28

Figure 1-7. The Keap1–Nrf2 pathway. The transcription factor Nrf2 is a master regulator of many cytoprotective genes in response to oxidative and electrophilic stresses. Keap1 is a cytoplasmic protein regulating Nrf2 activity. Under unstressed conditions, Nrf2 is constantly degraded via the ubiquitin–proteasome pathway in a Keap1-dependent manner. When oxidative or electrophilic stress modify the cysteine residues of Keap1, Nrf2 is stabilized by two possible mechanisms (hinge and latch mechanism and Cul3 dissociation) and translocates into nuclei. There, Nrf2 heterodimerizes with small Maf proteins and activates target genes through antioxidant/electrophile response element (ARE/EpRE). Pathways and genes targeted by Nrf2 are involved in 1) glutathione synthesis (Glutamate-cysteine ligase, catalytic subunit (Gclc), glutamate-cysteine ligase, modifier subunit (Gclm)), 2) elimination of ROS (Thioredoxin reductase 1 (Txnrd1), Peroxiredoxin 1 (Prdx1), 3) detoxification of xenobiotics (NAD(P)H dehydrogenase, quinone 1 (Nqo1), Glutathione S-transferase (Gst) gene family) and 4) drug transport (Multidrug resistance-associated protein (Mrp) gene family). E, electrophile. Adapted from Taguchi K., et al. Genes to Cells. 2011, 16(2), 123-140.

Keap1 binding Tansactivation CNC-type bZIP domain domain DNA binding domain 1 16 86 112 134 183 201 338 388 435 562 605

Nrf2 Neh2 Neh4 Neh5 Neh6 Neh1 Neh3

29 31 79 82 DLG ETGE

Binding with Cul3 Binding with Nrf2

1 78 178 327 609 624

Keap1 NTR BTB IVR Kelch (DGR domain) CTR

C151 C273 C288

Figure 1-8. The structure of Nrf2 and Keap1. Abbreviations: CNC, Cap'n'collar; bZIP: Basic leucine zipper; CTR, C-terminal region; NTR, N-terminal region; BTB, Broad comples, tramtrack, bric-a-brac; IVR, Intervening region. Adapted from Baird, L. and A. Dinkova-Kostova, Archives of Toxicology. 2011, 85(4), 241-272.

Figure 1-9. The genomic region of human EPHX1 contains intronic regulatory elements. The snapshot of the genomic region in the human mEH locus reveals that the proximal promoter region of E1b is associated with an open chromatin structure in a variety of cell types, indicating the E1b proximal promoter is constitutively active and accessible for transcription factors. There is also a putative CpG island in the same region. This discovery of a CGI in the proximal promoter region of E1b indicates a possibility that GC-rich DNA motif-binding transcription factors, such as Sp1 and Sp3, may be involved in the regulation of E1b promoter. Two open chromatin regions are located in the intronic regions between E1b and E1. These two regions are also associated with characteristic chromatin signatures of enhancers, such as monomethylation of histone H3 lysine 4 (H3K4Me1) and H3K27 acetylation (H3K27Ac), indicating these regions could contain enhancer elements.

Chapter 2 Transcription factors Sp1 and Sp3 contribute to the basal expression of human microsomal epoxide hydrolase alternative E1b mRNA variant by interacting with the proximal E1b promoter

ABSTRACT Microsomal epoxide hydrolase (mEH) is an important biotransformation enzyme, catalyzing xenobiotic metabolism. Human mEH is transcribed using alternative promoters. The proximal E1 promoter is active in liver while the far upstream E1b promoter drives the expression of mEH in all tissues including liver. Although several liver-specific transcription factors have been identified in the regulation of E1 transcription, little is known regarding the mechanisms of E1b transcriptional regulation. Genome-wide mapping of DNase I hypersensitive sites revealed an open chromatin region between nucleotide -300 upstream and +400 downstream of E1b. This area coincides with a previously described promoter region responsible for maintaining high basal promoter activity. In silico analysis of this location revealed several Sp1/Sp3 binding sites. Site-directed mutagenesis of these motifs suppressed the transactivation activity of the E1b proximal promoter, indicating their importance as contributors to E1b promoter regulation. Further, E1b promoter activities were increased significantly following Sp1 and Sp3 overexpression, while Mithramycin A, a selective Sp1 inhibitor, reduced the promoter activities. EMSA studies demonstrated that Sp1 bound to two putative Sp1/Sp3 binding sites. ChIP analysis confirmed that both endogenous Sp1 and Sp3 were bound to the proximal promoter region of E1b. Knockdown of Sp1 expression using siRNA did not change endogenous E1b transcriptional level, while knockdown of Sp3 greatly decreased E1b expression in different human cell lines. Taken together, these results support the concept that Sp1 and Sp3 are functionally involved as integrators regulating the basal expression of the derived mEH E1b variant transcript.

INTRODUCTION Human microsomal epoxide hydrolase (mEH, EPHX1) is an critical biotransformation enzyme (Fretland and Omiecinski, 2000), catalyzing the hydrolysis of electrophilic epoxides generated from oxidative metabolism contributed by cytochrome P450 (CYP450) enzymes. Epoxides are potentially highly reactive and may react covalently with cellular DNA and other macromolecules resulting in mutagenesis and carcinogenesis. In contrast, dihydrodiols, the products of hydrolysis of epoxides, tend to exhibit less reactivity and greater water soluble and therefore more readily eliminated. In this sense, mEH is a detoxifying enzyme. However, in the metabolism of polycyclic aromatic hydrocarbons (PAH), mEH plays an opposite role (Lu and Miwa, 1980). PAHs are ubiquitous pollutants and known human carcinogens. The metabolism of PAHs involves the action of specific enzymes in a multistep process. CYP450 enzymes first convert PAHs to epoxides that are further hydrolyzed by mEH to form 32

PAH dihydrodiols. These dihydrodiols may be further oxidized to yield diol epoxides, which are often more reactive than the original epoxides. The necessity of mEH in the bioactivation of PAH procarcinogens was confirmed in mEH-null mice, which are highly resistant to PAH-induced carcinogenesis compared with wild type mice (Miyata et al., 1999). The balance between detoxification and bioactivation by mEH is important for protecting against many chemically-initiated diseases, such as cancer. Any aberrant change affecting protein levels and subsequent enzymatic activities of mEH may therefore represent a risk factor for various diseases (Omiecinski et al., 2000). Human mEH is encoded by a single gene on Chromosome 1. Driven by alternative promoters, it is transcribed from two distinct locations approximately 15 kb apart (Liang et al., 2005). Two resulting transcripts are termed as E1 and E1b. The E1 promoter is exclusively active in liver, while the E1b promoter drives mEH expression in all tissues, including liver (Liang et al., 2005; Yang et al., 2009). Previous studies have shown that several liver-enriched transcription factors, in particular the C/EBPα, HNF3 and GATA transcription factors, are involved in regulating E1 transcription (Liang et al., 2005; Zhu et al., 2004a; Zhu et al., 2004b). However, with regard to E1b, since its discovery in 2005 there has been little progress elucidating its regulatory mechanisms. A recent study from our laboratory demonstrated that the presence of genetically polymorphic transposable elements within the promoter region of E1b functions to decrease luciferase reporter-based transcription activity (Yang et al., 2009). These data appear to explain some of the interindividual variability noted in mEH expression. However, they do not explain why human mEH levels vary across tissues within the same individual. In addition, how exactly these elements affect mEH promoter activity is unknown. In this study, we sought to identify the transcription factors that are mechanistically involved in maintaining the basal expression of the human mEH alternative transcript variant, E1b. The discovery of a CpG island in the proximal promoter region of E1b indicated a potential link between E1b transcriptional regulation and GC-rich DNA motif-binding transcription factors, such as Sp1/Sp3. The results generated demonstrate that Sp1 and Sp3 are bound to the E1b proximal promoter region and functionally regulate its promoter activity.

MATERIALS AND METHODS

Materials Mithramycin A was from Enzo Life Sciences; the protease inhibitor mixtures were from Calbiochem. FuGENE 6 Transfection Reagent and the dual luciferase reporter assay system were from Promega. All cell culture media and supplies were from Invitrogen. The TRIzol Reagent was from Invitrogen and all other chemicals were purchased from Sigma unless otherwise indicated. Small interfering RNA (siRNA) targeting Sp1 mRNA (siRNA ID: s13319), Sp3 mRNA (siRNA ID: s13326), or Silencer® Select Negative Control No. 1 siRNA and the Lipofectamine RNAiMAX reagent were all purchased from Invitrogen. The mouse monoclonal anti-mEH antibody, rabbit polyclonal anti-Sp1 (H-255) and anti-Sp3 (D-20) antibodies were from Santa Cruz 33

Biotechnology. Another rabbit polyclonal anti-Sp1 antibody was from Millipore. The rabbit polyclonal anti- GAPDH was from Sigma and the normal rabbit IgG was from Cell Signaling Technology.

Plasmids The E1b promoter luciferase reporter construct (E1b-320/+46-pGL3) containing 320bp of the 5’-flanking region upstream of E1b was constructed as described previously (Liang et al., 2005). Site-directed mutagenesis of the putative Sp1/Sp3 sites in the E1b proximal promoter region was carried out with the use of the QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent, Santa Clara, CA) with the following primers: mut Sp#1 (5’- GCGGAGACTGaaaCGGGGCTGCTGA-3’ and 5’-TCAGCAGCCCCGtttCAGTCTCCGC-3’), mut Sp#2 (5’- AGGCCGGGCTTacatGGAGACTGCGC-3’ and 5’-GCGCAGTCTCCatgtAAGCCCGGCCT-3’), mut Sp#3 (5’- AGGCCGGGGAACaaaCCGCTCGGAGGC-3’ and 5’-GCCTCCGAGCGGtttGTTCCCCGGCCT-3’), mut Sp#4 (5’-GGAGCCTTAttCAttCCTAGAGACT-3’ and 5’-AGTCTCTAGGaaTGaaTAAGGCTCC-3’), mut Sp#5 (5’- GGCCGCGGACCaaaCTTTAAGTAGCCCG-3’ and 5’-CGGGCTACTTAAAGtttGGTCCGCGGCC-3’), and mut Sp#6 (5’-TCTGGCCGCGGaaaCGCGGACCGCCC-3’ and 5’-GGGCGGTCCGCGtttCCGCGGCCAGA-3’). The mutated nucleotides are in lowercase. The basal E1b promoter construct (E1b-320/+46-pGL3) was used as template for amplification. Expression plasmids were generated by inserting full length cDNA of Sp1 (NM_138473.2), Sp3 (NM_003111.4) and ZBTB10 (NM_001105539.1) into the p3XFLAG-CMV10 expression vector (Sigma, St. Louis, MO). Primers for amplifying these genes were as follows: Sp1 (5’- GATCGAATTCAAGCGACCAAGATCACTCCATG-3’ and 5’- GATCTCTAGAATCAGAAGCCATTGCCACTGAT-3’), Sp3 (5’- GATCGAATTCAACCGCTCCCGAAAAGCCCGTG-3’ and 5’- GATCGGATCCTTACTCCATTGTCTCATTTCCAG-3’), and ZBTB10 (5’- ATCGTCGTTCAGTGAAATGAACCGC-3’ and 5’- GATCGGATCCTTAATCATCTAGAGACATACAAACTTCTCC-3’). All constructs were confirmed by DNA sequencing.

Cell culture, transient transfection and luciferase reporter assays Human bronchial epithelial BEAS-2B cells and human hepatoma HepG2-derived C3A cells were purchased from American Type Culture Collection (Manassas, Virginia). Both cell lines were cultured in 5% CO2 incubator at 37°C in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% FBS, 2 mM L- glutamine, 0.1 mM Non-Essential Amino Acids, 1.0 mM Sodium Pyruvate , 10 mM HEPES, 0.15% sodium bicarbonate, 100 units/ml penicillin G, and 100 µg/ml streptomycin. Cells were cultured in 24-well or 6-well plates and 60 mm or 100 mm petri dishes and harvested according to the requirements of the experiments. BEAS-2B and C3A cells were seeded a day before transfection in 24-well plates at a density of 5×104 cells per well. For assessing E1b promoter activities in response to ectopic expression of Sp1, Sp3 and other transcription factors, cells were co-transfected with E1b-320/+46-pGL3 reporter plasmid and the corresponding 34 expression plasmid of the given transcription factor using a FuGENE 6 transfection protocol according to the manufacturer’s instructions. The pRL-CMV plasmid containing Renilla Luciferase cDNA was also co-transfected as an internal control for transfection efficiency. Cells were harvested 24 h post transfection and luciferase activity was measured and analyzed in a Veritas Microplate Luminometer (Turner Biosystems) using the Dual Luciferase Reporter Assay System (Promega) as described previously (Auerbach et al., 2005). For Mithramycin A treatment, the cells were transfected with E1b-320/+46-pGL3 and pRL-CMV reporter plasmids for 6 h and were incubated for 24 h in culture medium containing the indicated concentration of Mithramycin A or vehicle (0.1% DMSO). Luciferase activity was measured in the same manner as described above. All transfections were performed in triplicate and the results were expressed as means ± standard deviations (SD) of triplicates. The experiments were repeated three times and the most representative results were shown.

Sp1 and Sp3 siRNA knockdown studies To reduce endogenous Sp1 or Sp3 and assess the effect on E1b promoter activity, BEAS-2B and C3A cells were transfected with the respective siRNAs at 25nM using the Lipofectamine RNAiMAX reagent and assessed with a Reverse Transfection Protocol according to the manufacturer’s instructions. Briefly, the transfection complexes of the Lipofectamine RNAiMAX reagent and the given siRNA were prepared in 24-well plates before medium and cells at a density of 5×104 cells per well were added to each well. Following transfections, cells were allowed to recover for 24 h and sequentially transfected with E1b-320/+46-pGL3 and pRL-CMV reporter plasmids using FuGENE 6 as described above. Luciferase activities were measured and analyzed after 24 h as mentioned previously. To assess endogenous E1b transcription and mEH protein level in response to the knockdown of Sp1 or Sp3, BEAS-2B and C3A cells were transfected with these siRNAs at 25 nM using the Lipofectamine RNAiMAX reagent with a Forward Transfection Protocol according to the manufacturer’s instructions. Briefly, cells were seeded a day before transfection in 6-well plates at a density of 3×105 cells per well or in 60 mm petri dishes at a density of 7×105 cells per dish. The transfection complexes of the Lipofectamine RNAiMAX reagent and the given siRNA were added to each well containing cells. After 48 h, siRNA-transfected cells in 6-well plates were harvested for RT-PCR analysis and cells in 60 mm petri dishes were collected for western blotting.

RNA isolation, reverse transcription and quantitative real-time PCR Total RNA from siRNA-transfected BEAS-2B and C3A cells in 6-well plates was extracted with TRIzol Reagent according to the manufacturer’s instructions. Total RNA (2 µg) was converted to cDNA using the High- Capacity cDNA Archive Kit (Applied Biosystems). cDNAs were analyzed with CFX96 Real-Time PCR Detection System (Bio-Rad) using PerfeCTa SYBR Green SuperMix (Quanta Biosciences). The final concentration of primers in each reaction was 0.2 µM. The PCR conditions consist of an initial denaturation for 3 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. Each sample was run in duplicate and the results were normalized to the level of GAPDH mRNA. The primers used for quantitative real-time PCR were as 35 follows: E1b, 5'-GAGCCTGCGAGCCGAGAC-3' (forward)/ 5'-CGTGGATCTCCTCATCTGACGTTT-3' (reverse); Sp1, 5’-ATTGAGTCACCCAATGAGAACAG-3’ (forward)/ 5’- CAGCCACAACATACTGCCC-3’ (reverse); Sp3, 5’-CACTGGTCAGTTGCCAAATC-3’ (forward)/ 5’-GAGCTGCCACTCTTCAGGAT-3’ (reverse); and GAPDH, 5'-CCCATCACCATCTTCCAGGAG-3' (forward)/5'- GTTGTCATGGATGACCTTGGC-3' (reverse).

Western blotting BEAS-2B and C3A cells were plated in 60 mm petri dishes and transfected with siRNA as described above. Cells were washed with PBS, trypsinized and centrifuged at 1000×g for 3 min. For preparation of whole cell lysates, cells were lysed in RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with 1× protease inhibitor cocktail (Cat # 539131, Calbiochem). The cell lysates were centrifuged at 16,000×g for 10 min at 4°C and the supernatants were collected as whole-cell lysate. Protein concentrations were determined by Pierce 660 nm Protein Assay (Thermo Scientific). The extracted proteins (30 μg) were separated on a 10% denaturing polyacrylamide gel (Bio-Rad) and transferred to a PVDF membrane (Bio-Rad). After blocking in 5% skim milk for 30 min, the blots were incubated sequentially with primary antibodies at the dilution of 1:1000 and horseradish peroxidase-conjugated secondary antibodies at the dilution of 1:5000. The membranes were washed three times with 1×TBS/0.1% Tween 20, treated with Pierce ECL Western Blotting Substrate (Thermo Scientific), and exposed to ImageTek-H X-ray films (American X-Ray & Medical Supply). The antibodies used for immunoblotting were as follows: anti-mEH (sc-135984, Santa Cruz Biotechnology), anti-Sp1 (17-601, Millipore), anti-Sp3 (sc-644, Santa Cruz Biotechnology) and anti-GAPDH (G9545, Sigma).

Electrophoretic mobility shift assays (EMSA) BEAS-2B cells in 100 mm petri dishes at 80% confluence were transfected with the Sp1 expression plasmid using FuGENE 6 as described above. After 24 h, the nuclear extracts were prepared with NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific) according to the manufacturer’s instructions. Double-stranded probes containing putative Sp1/Sp3 sites were end-labeled with [γ-32P] ATP by T4 polynucleotide kinase ((New England Biolabs). For EMSA, the DNA-binding reactions, containing 2 µg of nuclear extracts, 20 fmol of labeled probes, 0.01 mg/ml sonicated salmon sperm DNA (D7656, Sigma), 2 µl of 5×

binding buffer [20% (v/v) glycerol, 5 mM MgCl2, 2.5 mM EDTA, 2.5 mM dithiothreitol, 250 mM NaCl, 50 mM Tris-HCl (pH 7.5), 0.25 mg/ml poly(dI-dC)] in a final volume of 10 µl, were incubated with or without unlabeled competitor for 20 min at room temperature. For supershift assays, 2 µg of anti-Sp1 antibody (17-601, Millipore) or normal rabbit IgG (2729s, Cell Signaling Technology) were added to the binding reaction mixture without the labeled probe and incubated at 4°C for 30 min before addition of the labeled. The DNA-protein complexes were resolved by electrophoresis through a nondenaturing 4% polyacrylamide gel in 0.5× TBE buffer. Subsequently, gels were dried and exposed to X-ray film with intensifying screens at -70°C. The sense sequence of probes were 36 as follows: Sp site#1, 5’- GCGGAGACTGCGCCGGGGCTGCTGA-3’; Sp site#2, 5’- AGGCCGGGCTTGGGCGGAGACTGCGC-3’; Sp site#3, 5’-AGGCCGGGGAACGCCCCGCTCGGAGGC-3’; Sp site#4, 5’-GGAGCCTTAGGCAGGCCTAGAGACT-3’; Sp site#5, 5’- GGCCGCGGACCGCCCTTTAAGTAGCCCG-3’; Sp site#6, 5’-TCTGGCCGCGGGGCCGCGGACCGCCC-3’; Sp1 consensus oligos, 5’-ATTCGATCGGGGCGGGGCGAGC-3’; and mutant Sp1 consensus oligos, 5’- ATTCGATCGGttCGGGGCGAGC-3’. The mutated nucleotides in the mutant Sp1 consensus oligos are in lowercase.

Chromatin immunoprecipitation (ChIP) assay BEAS-2B and C3A cells, grown to 80-90% confluence in 100 mm dishes, were harvested by trypsinization and fixed in 1% formaldehyde at room temperature for 10 min with slow agitation. The fixation was stopped by addition of glycine to a concentration of 0.125 M. After a 5 min incubation at 25ºC, cells were pelleted by centrifugation at 1000×g for 5 min and then washed twice with ice-cold phosphate-buffered saline. Cells were lysed for 10 min on ice in SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8) with protease inhibitor cocktail (539131, Calbiochem). Cells were then sonicated with a Bioruptor sonicator (Diagenode, Liège, Belgium) for 5 cycles of 30 sec ON and 30 sec OFF at HIGH setting in a refrigerated water bath. Sheared cross-linked chromatin was centrifuged at 12,000×g for 10min at 4°C and diluted 10-fold in ChIP Dilution Buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH8, and 167 mM NaCl) with protease inhibitor cocktail. The diluted chromatin was pre-cleared overnight at 4°C with 35 μl protein A/G Plus-agarose beads (sc-2003, Santa Cruz Biotechnology) which were pre-blocked with sonicated salmon sperm DNA (201190, Stratagene) and BSA (2930, EM Science). Pre-cleaned chromatin was then incubated overnight at 4°C with 4 μg of anti-Sp1 antibody (17-601, Millipore), anti-Sp1 antibody (sc-14027, Santa Cruz Biotechnology), anti-Sp3 antibody (sc-644, Santa Cruz Biotechnology) or normal rabbit IgG (2729s, Cell Signaling Technology). To collect the antibody-chromatin complex, 75 μl protein A/G Plus-agarose beads pre-blocked as above were added, incubated for 3 h with rotation at 4°C and pelleted by centrifugation at 5000×g for 1 min. The pelleted complexes were then washed sequentially with Low Salt Immune Complex Wash Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8, 150 mM NaCl), High Salt Immune Complex Wash Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8, 500 mM NaCl), and LiCl Immune Complex Wash Buffer (0.25 M LiCl, 1% IGEPAL CA630, 1% deoxycholic acid, 1 mM EDTA, 10 mM Tris-HCl, pH 8), followed by two washes with TE buffer. Precipitated Protein-DNA complexes were eluted twice with 100µl elution buffer

(1% SDS, 0.1 M NaHCO3) for 15 min at room temperature. To reverse crosslinks, the eluates were incubated at 65°C for 4 h in the presence of 8 µl of 5 M NaCl and 1 µl of 10 mg/ml RNase A. Proteins were digested with 2 µl of 10 mg/ml proteinase K for 2 h at 45°C in the presence of 4 µl of 0.5 M EDTA, pH 8, and 8 µl of 1 M Tris-HCl, pH 8. DNA was purified with ChIP DNA Clean & Concentrator kit (D5205, Zymo Research). Immunoprecipitated DNA was amplified and the PCR amplicons were analyzed on 1.5% agarose gels. PCR amplification with appropriate primers was performed to analyze immunoprecipitated DNA. The PCR amplicons 37 were subjected to 1.5% agarose gel electrophoresis and visualized by ethidium bromide staining. The immunoprecipitated DNA was also subjected to the analysis of quantitative RT-PCR. The primers used for detecting E1b proximal promoter were: forward (5’- ACCGCCCTTTAAGTAGCCCGTTT-3’) and reverse (5’- TTACGGTCTCGGCTCGCA-3’).

Statistical analyses Data are expressed as means ± standard deviations (SD). The statistical significance of the differences between samples was determined using one-way analysis of variance (ANOVA) in combination with Dunnett's test or one-tailed Student's t test, dependent on the design of experiments. Differences were considered significant for samples with p-values <0.05.

RESULTS

Identification of the critical promoter region for the basal expression of E1b To investigate the basal transcriptional regulation of the E1b variant, we analyzed the promoter region of the gene using the UCSC Genome Browser. Genome-wide mapping of DNase I hypersensitive sites revealed an open chromatin region in several cell types containing the alternative first exon, E1b (Figure 2-1A). This region, which spans from -300 bp upstream to +400 bp downstream of E1b, coincides with a previously described promoter region that contributes to high basal promoter activity (Liang et al., 2005). The co-localization of an active promoter sequence proximal to the alternative first exon and a nucleosomal structure highly accessible to transcription factors justifies the notion that the E1b core and proximal promoters are located at this region. As indicated in Genome Browser display of the E1b flanking region (Figure 2-1A), the area maps to a putative CpG island (CGI). This CpG-rich region is characterized with a GC content of 66.3% and an Obs/Exp ratio of CpG dinucleotide of 0.937 (Figure 2-1B).

Analysis of the E1b proximal promoter region for the transcription factor Sp1/Sp3 binding sites To assess the proximal promoter region of E1b, the nucleotide sequence between position -320 bp and +46 bp relative to the transcriptional start site (TSS) was retrieved using the UCSC Genome Browser (Figure 2- 2A). The TSS for E1b variant was determined by the 5’ end of Exon E1b as reported previously (Liang et al., 2005) and defined as +1 (Figure 2-2A). We noticed a discrepancy on the start position of E1b as reported by Liang, et al (termed as TSS1) with GenBank entry NM_001136018 (termed as TSS2). TSS1 is ~ 30bp downstream of TSS2. To resolve this inconsistency, we searched the DataBase of Transcription Start Sites (DBTSS) (Yamashita et al., 2012). Both TSS1 and TSS2 were detected by TSS-seq, but the usage of the former is greater than the latter. Therefore, TSS1 represents the major start site of E1b. The discovery of a CpG island in the region of the TSS indicated possible regulation of E1b promoter by transcription factors that bind to GC boxes. A search for transcription factor binding sites in this proximal 38 promoter region using MatInspector (http://www.genomatix.de) and TESS (http://www.cbil.upenn.edu) indicated that the -320 bp/+46 bp region lacked canonical CCAAT and TATA boxes, but contained 6 potential Sp1/Sp3 binding sites (Figure 2-2A). To evaluate the contribution of these Sp1/Sp3 binding sites to the basal promoter activity of E1b proximal promoter, we made point mutations and measured their effects on transcriptional activity in BEAS-2B and C3A cells with the luciferase assay (Figure 2-2A). As shown in Figure 2-2B, all of site mutations had a strong negative effect on E1b promoter activity; especially mutations at sites #2, #3 and #6, which reduced the promoter activity to less than 35% of the wild-type promoter activity. The results indicated that these putative Sp1/Sp3 sites in the E1b proximal promoter play an important role in the basal transcription of E1b but not all of the Sp1/Sp3-binding sites are functionally equivalent.

Sp1/Sp3 is involved in activation of the E1b proximal promoter The presence of canonical Sp1/Sp3 binding sites in the proximal promoter region of E1b variant suggested that Sp1/Sp3 was involved in the regulation of E1b promoter activity. Therefore, we co-transfected the E1b −320/+46 luciferase construct and Sp1 or Sp3 expression plasmid into BEAS-2B and C3A cells. Overexpression of Sp1 and Sp3 significantly increased reporter activity in BEAS-2B cells (Figure 2-3A). Interestingly, Sp1 and Sp3 had only a small effect on reporter activity in C3A cells. This may be due to the presence of endogenous Sp1 or Sp3 proteins, as well as other Sp family proteins which may virtually saturate the reporter read out. We next evaluated the effect of siRNA knockdown of Sp1 and Sp3 on E1b promoter activity. In BEAS-2B cells, knockdown of Sp1 and Sp3 significantly attenuated E1b proximal promoter activity, while in C3A cells only knock down of Sp1 had a significant effect on E1B promoter activity (Figure 2-3B). Overall, these results suggest that both Sp1 and Sp3 can regulate the basal activity of the E1b proximal promoter. To further confirm the involvement of Sp1/Sp3 for the basal activity of E1b proximal promoter, we performed a series of transient transfection experiments using the E1b -320/+46 promoter-luciferase reporter construct and tested the converse hypothesis that interference in the interaction of Sp1 or Sp3 with the E1b promoter would reduce its promoter activity. First, we treated cells with Mithramycin A that binds to GC rich DNA sequences and prevents Sp1 binding to its target DNA motif (Snyder et al., 1991). Treatment of Mithramycin A significantly blocked E1b promoter activity in a dose-dependent manner in both BEAS-2B and C3A cells (Figure 2-3C). Next, we co-transfected the cells with the E1b-300 luciferase construct and increasing amounts of a ZBTB10 expressing plasmid. ZBTB10 is a suppressor of Sp-dependent transactivation. It has been shown to compete with Sp1 binding sites and decrease the expression of Sp1 and Sp3 (Mertens-Talcott et al., 2007; Tillotson, 1999). Over-expression of ZBTB10 resulted in a dose-dependent reduction in reporter activity (Figure 2-3D). These data suggest that inhibition of Sp1/Sp3 protein binding to the E1b proximal promoter inhibits basal promoter activity. 39

Sp1 and Sp3 bind and interact with the E1b proximal promoter region To extend our promoter transactivation analysis and determine whether Sp1 or Sp3 binds to the E1b proximal promoter region via the putative binding sites for Sp1 and Sp3, we performed EMSA with oligonucleotide probes containing these sites. When these labeled probes were incubated with nuclear extracts from BEAS-2B cells transfected with Sp1 expression plasmid, several DNA-protein complexes were produced (Figure 2-4A). The specificity of binding was confirmed by loss of labeled binding in the presence of 50-fold excess of unlabeled Sp1 consensus oligonucleotide but not unlabeled mutated Sp1 consensus oligonucleotide. Sp1/Sp3 binding sites #2 and #3 displayed a substantial competition effect, which suggested that Sp1 preferentially binds to these two sites compared to the other 4 sites (Figure 2-4A). Furthermore, addition of anti- Sp1 antibody, but not normal rabbit IgG, disrupted the DNA-protein complexes formed with Site#2 and #3, indicating that Sp1 is involved in the formation of these DNA-protein complexes (Figure 2-4B). These results suggested that Sp1 protein specifically binds to the sites in the E1b proximal promoter in vitro. In these studies, we did not test whether Sp3 interacts with these putative binding sites in E1b proximal promoter. Because Sp1 and Sp3 share more than 90% sequence homology in the DNA-binding domain and bind to the same cognate DNA-element, it is likely that Sp3 can also bind to sites #2 and #3. Next, we examined whether Sp1 or Sp3 binds to the E1b proximal promoter using the chromatin immunoprecipitation (ChIP) assay. Chromatin from BEAS-2B and C3A cells was sonicated and immunoprecipitated using antibodies against Sp1 and Sp3. The precipitated DNA was subjected to PCR (Figure 2-5A) and qPCR (Figure 2-5B) analyses using primers for the -320/+46bp region of E1b proximal promoter. We found that E1b promoter-specific primers amplified this promoter region from DNA that was immunoprecipitated by either Sp1 or Sp3 antibody in BEAS-2B and C3A cells. In the same experiment, no signal was observed when chromatin was immunoprecipitated with control rabbit IgG. These results clearly showed that Sp1 and Sp3 directly interact with E1b proximal promoter region.

Knockdown of Sp1 and Sp3 regulated expression of E1b variant Because luciferase reporter-based promoter characterization using EMSA and ChIP analyses indicated an important role for Sp1 and Sp3 in transactivation of the E1b proximal promoter, we determined whether Sp1 and Sp3 siRNA knockdown would reduce the endogenous expression of the E1b variant. Transfection with Sp1 and Sp3 siRNAs in BEAS-2B and C3A cells resulted in significant knockdown of Sp1 and Sp3 mRNA and protein levels (Figure 2-6A and 2-6B). Surprisingly, down-regulation of Sp1 did not significantly affect the mRNA level of E1b in either cell line (Figure 2-6A). In contrast, siRNA knockdown of Sp3 resulted in a 65~75% reduction in E1b transcript levels, as well as mEH protein levels in BEAS-2B and C3A cells (Figure 2-6B). These data suggested that Sp1/Sp3 interactions do influence basal endogenous expression of E1b variant. 40

DISCUSSION This investigation characterized the E1b proximal promoter and provided new insights into the regulation of human mEH gene expression driven by the alternative E1b promoter. We showed that the E1b proximal promoter is located within the region of first 300 bp upstream of E1b and contains at least two TSSs, CGI and several potential Sp1/Sp3 binding sites. Further characterization revealed that the Sp1/Sp3 sites are important for basal promoter activity and that Sp1 and Sp3 bind to the E1b promoter. We also found that Sp1 and Sp3 overexpression and knockdown influenced E1b promoter activity. Collectively, our results strongly suggest that Sp1 and Sp3 are key regulators contributing to the basal transcription of E1b. Bioinformatics analysis of the E1b proximal promoter revealed a CGI overlap with this region. CGIs have been shown to associate with promoter regions in 70% of human genes (Saxonov et al., 2006) and are found in all house-keeping genes and approximately 40% of tissue-specific genes (Illingworth and Bird, 2009). CGI promoters typically lack TATA and other core promoter elements, but contain multiple Sp1 binding sites (Butler and Kadonaga, 2002). CGI promoters are also known to contain multiple TSS over a stretch of about 50 to 100 nucleotides (Juven-Gershon et al., 2008; Smale and Kadonaga, 2003). The E1b promoter has no TATA box or other core promoter elements, but several potential Sp1/Sp3 binding sites within the upstream region of E1b and perhaps more within the rest of the CGI. Through literature review and DBTSS searching, we identified two TSSs within the E1b proximal promoter that are about 30 nucleotides apart. The DBTSS inquiry results also indicated that there are additional TSSs in this region and these TSSs show tissue-dependent patterns. The mechanistic basis of the differential TSS usage on E1b promoter is unknown. It is possibly due to variations in the availability of transcription factors at different sites on the E1b promoter or/and the changes on the patterns of epigenetic modifications, such as DNA methylation (Kawaji et al., 2006). In addition, how tissue-specific usage of TSSs contributes to the total transcription level of E1b needs further investigation. It is tempting to speculate that various transcriptional factors and cofactors may utilize distinct TSSs to control the expression of E1b depending on cell context. Six potential Sp1/Sp3 binding sites were identified in the bioinformatics analysis of the E1b promoter DNA sequence. The mutation of any site significantly reduced the transactivation activity of E1b promoter, indicating these sites are important for basal promoter activity. EMSA established that Sp1 binds two out of the six sites. The rest of these binding sites may interact with other related transcription factors that are also important for the maintenance of the basal transcription of E1b. Possible candidates are transcription factors from Sp- like/KLF family, as they share more than 65% sequence homology on the DNA binding-domains composed of

three adjacent Cys2His2-type zinc finger motifs (Kaczynski et al., 2003). The structural similarities among family members enable them to bind similar DNA sequences. For example, Sp1 has been shown to compete for the same sites with Sp3, KLF4, KLF6, KLF9 and KLF13 (Grande et al., 2012; Kaczynski et al., 2003). In addition, there are potential binding sites for other transcription factors that overlap with putative Sp1/Sp3 binding sites. For example, site #6 overlaps with a potential binding site for transcription factor AP-2α (TFAP2A). Mutation on site 41

#6 also destroys AP-2α binding sites. These mutagenesis data indicate that AP-2α plays an important role in the basal expression of E1b. Sp1 and Sp3 are ubiquitous transcription factors that bind to GC-rich motifs in the proximal promoter of a wide variety of genes, such as housekeeping and tissue-specific genes (Li and Davie, 2010). Sp1 acts as a transcriptional activator, whereas Sp3 can be either a transcriptional activator or repressor (Suske, 1999). In the context of the E1b promoter, Sp1 and Sp3 both serve as activators as indicated by the data generated from overexpression of Sp1 and Sp3, mutagenesis analysis and Mithramycin A treatment coupled with luciferase reporter assays. Further, siRNA knockdown of Sp3 expression resulted in a reduction of the endogenous E1b transcription level, confirming Sp3 is an activator. However, the E1b transcription level remained unchanged when Sp1 was knocked down by Sp1 siRNA, indicating Sp1 does not regulate E1b expression. Nevertheless, we interpreted this inconsistency on the lack of effect of Sp1 on E1b expression from two aspects. First, the cancer cell lines used have high basal expression level of Sp1 and incomplete knockdown of Sp1 by siRNA leaves sufficient levels of Sp1 for normal function. Second, the transcription factors from the Sp-like/KLF family might bind to the E1b proximal promoter and have the same function as Sp1. Knockdown of Sp1 by siRNA results in occupancy by these related transcription factors on these sites that normally are bound by Sp1. Therefore, it is likely that E1b expression was not changed due to functional redundancy of these transcription factors. Sp1 has been shown to regulate target gene transcription cooperatively through interaction with other transcription factors that bind to DNA motifs proximal to Sp1-binding sites (Safe and Abdelrahim, 2005). Although this study focused on Sp1/Sp3 sites, in our analysis of the E1b promoter sequence we also identified putative binding sites for other cell type-specific transcription factors, such as AP-2, E2F, NFkB and GATA-1. The interaction between Sp1/Sp3 and these other transcription factors may be important determinants contributing to the regulation of E1b expression through differential cellular expression of the respective proteins. Thus, E1b expression is likely subject to combinatorial regulation in both ubiquitous and cell-type-specific manner. Further work is required to understand how the E1b promoter activity is altered with regard to cell background. The link between E1b expression and Sp1/Sp3-mediated transcriptional regulation via E1b proximal promoter is an intriguing development, given that little is known about the transcriptional regulation of the E1b promoter. The features identified in the context of the E1b proximal promoter indicate that its regulation is complex. The complexity is manifested in at least two aspects: 1) transcription from multiple TSSs and 2) combinatorial regulation by Sp1/Sp3 and likely other transcription factors whose binding sites are in close proximity within the E1b proximal promoter. Therefore, the E1b proximal promoter provides a platform that can accommodate diverse transcription factors to control the expression of E1b in different cells. The characterization of the E1b promoter and the fundamental insights provided here with respect to E1b transcriptional regulation will serve as a valuable platform for further elucidation of normal and pathological conditions involving mEH expression.

Figure 2-1. Location of E1b proximal promoter. A.) Genome browser display of the hg19 human assembly showing the alternative E1b promoter region of human mEH. Data tracks shown are DNaseI Digital Genomic Footprinting from ENCODE/University of Washington for A549, HepG2, K562 and NHLF cells and CpG islands. B.) A schematic structure of the putative E1b promoter CpG island. CpG sites within the CpG island are shown as short vertical lines.

-320 GGTGATAGAGTGAGACTCTGTCTCCAAAAAAAAAAAGGACATACATCATGCAAGTTTG -262 GATTTTTGTTTTTAGATTCAACTGACGAAGTTACGGGATCAAACTGCTGTAGGAGCTG -204 CCAAACGCTTCTCTCCATTTCTGGCCGCGGGGCCGCGGACCGCCCTTTAAGTAGCCCG Site #6: Sp1/Sp3 Site #5: Sp1/Sp3 -146 TTTTATCCCTGGCAGAGGTGGAGCCTTAGGCAGGCCTAGAGACTTTCCCGGGTCCTCC Site #4: Sp1/Sp3 -88 AGGCCGGGGAACGCCCCGCTCGGAGGCCGGGCTTGGGCGGAGACTGCGCCGGGGCTGC Site 3: Sp1/Sp3 Site #2: Sp1/Sp3 Site #1: Sp1/Sp3 -30 TGAAAACTAGCCGAGGAGAGCCAGGGAGCCGGAGAGATCGCGCGCCTGCCGCCGCCGG +1 +29 AGCCTGCGAGCCGAGACC

Figure 2-2. Identification and mutational analysis of Sp1/Sp3 binding sites within the E1b proximal promoter region. A.) Nucleotide sequence of the E1b -300bp proximal promoter. The predicted binding sites of Sp1 are underlined. The site indicated with +1 denotes the transcription start site (TSS) of the E1b transcript as previously reported by Liang, et al. B.) Mutation analysis of Sp1/Sp3 binding sites in the E1b proximal promoter. Individual Sp1/Sp3 sites were mutated and their effects on the E1b -300 promoter activity were measured by luciferase assays. Statistically significant differences on the luciferase activity compared to the wild type (WT) are indicated by an “*” (p<0.05).

A. B.

C. D.

Figure 2-3. Sp1 and Sp3 regulated E1b proximal promoter activities. A.) Overexpression of Sp1 and Sp3 activated E1b proximal promoter activity. BEAS-2B and C3A cells were co-transfected with Sp1 or Sp3 expressing plasmid and E1b-300 promoter-luciferase reporter. After 24h, luciferase activity was measured. B.) Knockdown of Sp1/Sp3 by siRNAs reduced E1b promoter activity. Cells were transfected with 25nM Sp1 or Sp3 siRNA and E1b-320/+46-pGL3 as described under “Materials and Methods”. C.)The Sp1/Sp3 inhibitor Mithramycin A attenuates E1b-300 promoter activity. Luciferase assay was performed after cells transfected with E1b -300 promoter-luciferase reporter were treated with Mithramycin A for 24h. D.) Overexpression of Sp protein repressor ZBTB10 decreased E1b-300 promoter activity. BEAS-2B and C3A cells were co-transfected with ZBTB10 expressing plasmid and E1b-300 promoter-luciferase reporter. After 24 h, luciferase activity was measured. Statistically significant differences between treatment groups compared to the empty vector control, the control siRNA, or the vehicle-treated control are indicated by an “*” (p<0.05).

Figure 2-4. EMSA and supershift analyses show Sp1 binding to putative Sp1/Sp3 binding sites on E1b-300 promoter. A.) An EMSA was performed using 32P-labeled oligonucleotide probes containing the putative E1b- 300 Sp1/Sp3 binding sites. The labeled probes were incubated with nuclear extracts from BEAS-2B cells overexpressing Sp1. A fifty fold excess of unlabeled competitor oligonucleotide was used for competition assays. Competitors used were the consensus Sp1 oligonucleotide ( Sp1, Lane 2, 5, 8, 11, 14, and 17) and a mutant Sp1 oligonucleotide (Sp1m, Lane 3, 6, 9, 12, 15, and 18). The arrows show specific binding. B.) Supershift analysis was performed using a polyclonal antibody against Sp1 (more detail about experimental conditions) (Lane 3 and 6). Supershift with a rabbit normal IgG was used as a negative control (Lane 2 and 5). The arrows show specific binding which was interrupted by addition of anti-Sp1 antibody.

B. BEAS-2B C3A 80 50 * * 60 40 30 40 30 20 * 15 20

(Fold change) * (Fold change) (Fold 10

Relative Enrichment 10 * Relative Enrichment 5

0 0 t t u G X re X u G X re X p g 5 0 g 5 0 I 5 o -2 p I 5 o -2 In -2 ip In 2 ip ll D - ll D H i 3 H i 3 1 -M p 1 M p p 1 S p 1 S S p S p S S Figure 2-5. ChIP assay for Sp1 and Sp3 binding to the E1b proximal promoter in BEAS-2B and C3A cells. ChIP assay was performed to confirm the binding of Sp1 and Sp3 to the E1b proximal promoter. The DNA- protein complexes were incubated with polyclonal antibodies against Sp1 or Sp3 and isolated by immunoprecipitation. For the negative controls, the DNA-protein complexes were incubated without antibodies or with normal IgG. The immunoprecipitated DNA fragments were analyzed by conventional PCR analysis (A) and qPCR analysis (B). Statistically significant differences on the binding to the E1b promoter region compared to the rabbit normal IgG are indicated by an “*” (p<0.05).

A. BEAS-2B C3A 1.5 1.5 Control siRNA Control siRNA Sp1 siRNA Sp1 siRNA 1.0 1.0 *

0.5 0.5 * * Relative mRNA fold change Relative mRNA fold change fold mRNA Relative 0.0 0.0 Sp1 E1b Sp1 E1b

B. BEAS-2B C3A 1.5 Control siRNA 1.5 Sp3 siRNA Control siRNA Sp3 siRNA 1.0 1.0

0.5 0.5 * * * * Relative mRNAfold change

0.0 Relative mRNA change fold Sp3 E1b 0.0 Sp3 E1b

Figure 2-6. Effect of Sp1 and Sp3 knockdown on E1b transcript and mEH protein. Cells were transfected with Sp1- (A) or Sp3- (B) specific siRNAs. After 48 h, total RNA was collected for assessment of transcript levels by real time quantitative PCR (upper panels) and total protein extracts were immunoblotted with anti-Sp1, Sp3 and mEH antibodies (lower panels). GAPDH was used as an internal control for real-time qPCR and a loading control for western blotting. Statistically significant differences compared to the negative control siRNA are indicated by an “*” (p<0.05). 48

Chapter 3 The role of intronic DNA elements in the regulation of the human microsomal epoxide hydrolase (mEH, EPHX1) driven by a far upstream alternative E1b promoter

ABSTRACT Microsomal epoxide hydrolase (mEH) plays an important role in the detoxification and bioactivation of xenobiotic epoxide intermediates generated by cytochrome P450 (CYP450) enzymes. Our laboratory discovered that human mEH gene transcription is initiated from alternative promoters, resulting in the mRNA transcripts termed E1 and E1b. These transcripts display tissue-specific expression, however, little is known about their regulation. In these studies, we discovered a positive effect of Nrf2 activators, sulforaphane (SFN) and tert- butylhydroquinone (tBHQ), on E1b expression in human lung cancer cell lines. Using data available from UCSC genome browser, we identified two DNase I hypersensitive regions (HS-1 and HS-2) located within the ~15 kb genomic region separating E1 and E1b, and investigated their involvement in the Nrf2-mediated up-regulation of E1b. Both elements stimulated transcription when incorporated within an E1b promoter-luciferase gene construct, however, the enhancer activities were cell-type dependent. In the human lung adenocarcinoma A549 epithelial cell line, both HS-1 and HS-2 were functional, while in the human bronchial epithelial BEAS-2B cell line, only HS-2 led to enhanced transactivation. Further, in BEAS-2B cells, the positive effect of SFN and tBHQ on the E1b promoter-driven luciferase activity was dependent on the more distal HS-2. Within the HS-2 region, we identified a conserved antioxidant-response element (ARE) that contributes to Nrf2 inducer mediated up-regulation of E1b. Furthermore, we identified an activator protein 1 (AP-1)/12-O-tetradecanoylphorbol-13-acetate (TPA) response element (TRE), overlapping the HS-2 enhancer ARE, that modulated ARE-mediated induction of E1b. These studies suggest complex, tissue specific mechanisms of mEH regulation and expression that likely influence its protective vs. protoxic metabolic contributions.

INTRODUCTION Microsomal epoxide hydrolase (mEH, EPHX1) is a key biotransformation enzyme that catalyzes the hydrolysis of electrophilic epoxides to dihydrodiols (Fretland and Omiecinski, 2000). Epoxides derived from xenobiotic metabolism by cytochrome P450 (CYP450) enzymes may interact with cellular macromolecules (DNA, RNA and proteins) and lead to certain genotoxicities. Metabolism of epoxides by mEH often generates less reactive and therefore less toxic dihydrodiol intermediates that also tend to be more easily eliminated. Thus, mEH is usually thought to play a protective role against the deleterious effects of reactive epoxide metabolites. However, it plays a role in the bioactivation of polycyclic aromatic hydrocarbons (PAHs), which are important components of cigarette smoke and established agents in the etiology of human lung cancers. In concert with CYP450 enzymes, mEH participates in the formation of stable PAH dihydrodiol epoxide DNA adducts. The 49 necessity of mEH in the bioactivation of PAHs was established in mEH-null mice, which are highly resistant to PAH-induced carcinogenesis compared to control mice (Miyata et al., 1999). Furthermore, data from epidemiological studies suggest that genetic polymorphisms in the coding regions of mEH correlate with lung cancer risk (Kiyohara et al., 2006). Besides its dual roles in xenobiotic metabolism, the transcriptional regulation of human mEH is interestingly complex. The transcription of mEH is regulated by alternative promoter usage and generates transcripts with unique noncoding first exons, termed E1 and E1b (Liang et al., 2005). E1 is initiated from a transcription start site ~3.2kb upstream of coding exon 2, while E1b is driven by a far upstream alternative promoter located ~18.5kb upstream of exon 2. These two mRNA isoforms exhibit highly distinctive tissue- specific distribution patterns (Liang et al., 2005; Yang et al., 2009). The E1 isoform is exclusively expressed in liver, while the E1b isoform is expressed in all tissues, including liver. The transcription factor, Nrf2 (NFE2L2), functions as a master regulator of cytoprotective genes that combat cellular oxidative stress. Activation of these genes involves Nrf2 binding to antioxidant response elements (ARE) in their respective promoters. Under normal non-stressed conditions, Nrf2 is constantly subject to proteasomal degradation mediated by KEAP1. KEAP1 functions as a cytoplasmic anchor of Nrf2 and an adaptor of Cullin-3 E3 ubiquitin ligase complex, leading to continuous ubiquitination of Nrf2 and its proteasomal degradation. Under oxidative stress conditions, Nrf2 dissociates from KEAP1 and escapes proteolysis, which in turn leads to rapid accumulation of Nrf2 in the nucleus and the subsequent transcriptional induction of target genes. Constitutive activation of Nrf2-mediated genes has been documented in non-small-cell lung cancer-derived cell lines and tumor samples, as well as other cancers (Ohta et al., 2008; Shibata et al., 2008b; Singh et al., 2006). In these studies, aberrant activation of Nrf2 was linked to mutations in KEAP1 and Nrf2 that impair the interaction between the two proteins, leading to nuclear accumulation of Nrf2. Consequently, these mutations result in up-regulation of drug metabolizing enzymes and drug efflux pumps and conceivably confer enhanced resistance of cancer cells to chemotherapeutic agents (Ohta et al., 2008; Shibata et al., 2008b; Wang et al., 2008). Nrf2 has been indicated as a regulator of mEH expression in mice. The knockout of Nrf2 reduces the basal expression of mEH in mouse liver and small intestine (Hu et al., 2006b; Kwak et al., 2001; Ma et al., 2006; Reisman et al., 2009; Thimmulappa et al., 2002). Similarly, Nrf2-null mice showed minimal inducible expression of mEH by Nrf2 inducers (Hu et al., 2006a; Hu et al., 2006b; Kwak et al., 2001; Lee et al., 2003; Ma et al., 2006; Reisman et al., 2009; Thimmulappa et al., 2002). Although a functional ARE is believed to exist in the regulatory region of mouse mEH, as in other phase II detoxifying enzymes (Lee et al., 2003; Thimmulappa et al., 2002), its exact location is unknown. In spite of the fact that Nrf2 is implicated in mEH regulation in mice, very little is known about human mEH gene regulation. Further, DNA sequence alignment data demonstrate that the upstream region of E1b promoter is poorly conserved between primates and rodents (Yang et al., 2009), suggesting that the mEH regulatory mechanisms existing in human and rodents are quite different. 50

In the present study, we characterized two unique intronic DNA enhancer elements, located between the E1 and E1b promoters, for their ability to modulate mEH expression in human lung cancer cell lines. These experiments identified a novel mechanism by which Nrf2 induces expression of the E1b variant transcript, providing an important step towards understanding the complicated transcriptional regulation of the human mEH gene.

MATERIALS AND METHODS

Materials The protease inhibitor mixtures were from Calbiochem, FuGENE 6 Transfection Reagent and the dual luciferase reporter assay system were from Promega. All cell culture media and supplies and the TRIzol Reagent were from Invitrogen. Sulforaphane (Cat#574215) was from CalBiochem; tert-Butylhydroquinone (tBHQ, Cat#B0833) was from TCI America; TPA (12-O-tetradecanoylphorbol-13-acetate, Cat#445-004-M001) was from Alexis Biochemicals. All other chemicals were purchased from Sigma unless otherwise indicated. Small interfering RNA (siRNA) targeting Nrf2 mRNA (siRNA ID: s9493) or Silencer® Select Negative Control No. 1 siRNA and the Lipofectamine RNAiMAX reagent were purchased from Invitrogen. The mouse monoclonal anti- mEH (sc-135984) and anti-Actin (sc-81178) antibodies and rabbit polyclonal anti-Nrf2 (sc-13032) were from Santa Cruz Biotechnology; the rabbit polyclonal anti-GAPDH (G9545) was from Sigma; and the normal rabbit IgG was from Cell signaling Technology.

Plasmids The E1b promoter luciferase reporter constructs containing 300bp, 1kb, 2kb or 2.2kb of the 5’-flanking region upstream of E1b were constructed as described previously (Liang et al., 2005). To generate the intronic enhancer/E1b-300 promoter luciferase reporter constructs, the regions containing DNase I hypersensitive sites located between E1b and E1 were amplified by PCR from human genomic DNA using primers as follows: DNase I HS-1(5'-ATGCGCTAGCTCAGGAAAGGAATGTGTAGGAGGG-3' and 5'- ATGCCTCGAGAAATGCTGGGATTACAGGTGTGCG-3'); and DNase I HS-2 (5'- GATCGGTACCTCCCTTTCCACTGGATGTTCCCTT-3' and 5'- GATCTCTAGAACCATAAGATGCAGGAAGAGGGCT-3'). The PCR products were inserted to the E1b-300 proximal promoter-containing pGL4 luciferase reporter vector. Site-directed mutagenesis of the putative ARE in DNase I HS-2 was carried out by QuikChange Lightning Site-Directed Mutagenesis Kit (Agilent, Santa Clara, CA) with the following primers: mutARE (5'-ATGGCAAGTaaTGAaTtATGGCCGGCTAGCA-3' and 5'- TGCTAGCCGGCCAtaAtTCAttACTTGCCAT-3'), mutARE-1 (5'- ATGGCAAGTGCTGAaTtATGGCCGGCTAGCA-3' and 5'-TGCTAGCCGGCCAtaAtTCAGCACTTGCCAT- 3'), and mutARE-2 (5'-ATGGCAAGTaaTGAGTCATGGCCGGCTAGCA-3' and 5'- TGCTAGCCGGCCATGACTCAttACTTGCCAT-3'). The mutated nucleotides are in lowercase. Expression 51 plasmids were generated by inserting full length cDNA of Nrf2 (NM_006164.4), KEAP1 (NM_203500.1) and JUN (NM_002228.3) into the p3XFLAG-CMV-10 expression vector (Sigma, St. Louis, MO). Primers for amplifying these genes were as follows: Nrf2 (5'-GATCGCGGCCGCAATGGACTTGGAGCTGCCGCCG-3' and 5'-GATCTCTAGACTAGTTTTTCTTAACATCTGG-3'), KEAP1 (5'- ATCGCAGCCAGATCCCAGGCCTAGC-3' and 5'-GATCTCTAGAATCAACAGGTACAGTTCTGCTGG-3') and JUN (5'-GATCGAATTCAACTGCAAAGATGGAAACGACC-3' and 5'- GATCTCTAGAATCAAAATGTTTGCAACTGCTGCG-3'). All constructs were confirmed by DNA sequencing.

Cell culture, transient transfection and luciferase reporter assays Human lung carcinoma epithelial A549 cells and human bronchial epithelial BEAS-2B cells were

purchased from American Type Culture Collection (Manassas, Virginia). Both cell lines were cultured in 5% CO2 incubator at 37°C in Dulbecco’s modified Eagle medium supplemented with 10% FBS, 2 mM L-glutamine, 0.1 mM Non-Essential Amino Acids, 1.0 mM Sodium Pyruvate , 10 mM HEPES, 0.15% sodium bicarbonate, 100 units/ml penicillin G, and 100 µg/ml streptomycin. Cells were cultured in 24-well or 6-well plates and 100 mm petri dishes and harvested according to the requirements of the experiments. For transient transfection, A549 and BEAS-2B cells were seeded a day before transfection in 24-well plates at a density of 5×104 cells per well. To assess the enhancer properties of intronic regulatory elements on E1b promoter activity, cells were transfected with corresponding reporter plasmids using a FuGENE 6 transfection protocol according to the manufacturer’s instructions. The pRL-CMV plasmid containing Renilla Luciferase cDNA was also co-transfected as an internal control for transfection efficiency. Cells were harvested 24 h post transfection and luciferase activity was measured and analyzed in a Veritas Microplate Luminometer (Turner Biosystems) using the Dual Luciferase Reporter Assay System (Promega) as described previously (Auerbach et al., 2005). For studying the effect of chemical treatment on E1b promoter activities and the enhancer activities of DNase I HS sites, cells were transfected with E1b promoter reporter or enhancer constructs and pRL- CMV reporter plasmids for 6 h and were incubated for 24 h in culture medium containing the indicated concentration of chemicals or vehicle (0.1% DMSO). For assessing E1b promoter activities in response to ectopic expression of Nrf2 and JUN, cells were co-transfected with reporter plasmids containing E1b proximal promoter and intronic enhancers and the corresponding expression plasmid of the give transcription factor. Luciferase activity was measured in the same manner as described above. All transfections were performed in triplicate and the results were expressed as means ± standard deviations (SD). The experiments were repeated three times and the most representative results were shown.

RNA isolation, reverse transcription and quantitative real-time PCR For investigating inducible expression of various genes by chemical treatment, A549 and BEAS-2B cells in 6-well plates were treated with chemicals for 24 h. To assess the effect of overexpression of Nrf2 on E1b expression in BEAS-2B cells, cells were transfected with Nrf2-expressing plasmid for 24 h and 48 h as described 52 above. Total RNA was extracted with TRIzol Reagent according to the manufacturer’s instructions. Total RNA (2μg) was converted to cDNA using the High-Capacity cDNA Archive Kit (Applied Biosystems). cDNAs were analyzed with CFX96 Real-Time PCR Detection System (Bio-Rad) using PerfeCTa SYBR Green SuperMix (Quanta Biosciences). The final concentration of primers in each reaction was 0.2µM. The PCR conditions consisted of an initial denaturation for 3 min at 95°C, followed by 40 cycles of 15s at 95°C and 1 min at 60°C. Each sample was run in duplicate and the results were normalized to the level of β-actin or GAPDH mRNA. The primers used for quantitative real-time PCR were as follows: E1b, 5'-GAGCCTGCGAGCCGAGAC-3' (forward)/5'-CGTGGATCTCCTCATCTGACGTTT-3' (reverse); Nrf2, 5'-CAGCGACGGAAAGAGTATGAG-3' (forward)/5'-GGGCTGGCTGAATTGGGAG-3' (reverse); HMOX1, 5'-CAGTGCCACCAAGTTCAAGC-3' (forward)/5'-GTTGAGCAGGAACGCAGTCTT-3' (reverse); NQO1, 5'-GGCAGAAGAGCACTGATCGTA-3' (forward)/5'-TGATGGGATTGAAGTTCATGGC-3' (reverse); β-Actin, 5'-CATGTACGTTGCTATCCAGGC-3' (forward)/5'-CTCCTTAATGTCACGCACGAT-3' (reverse); and GAPDH, 5'-CCCATCACCATCTTCCAGGAG- 3' (forward)/5'-GTTGTCATGGATGACCTTGGC-3' (reverse). The experiments were repeated three times and the most representative results were shown.

Western blotting To assess the effect of chemical treatments on mEH protein level, A549 and BEAS-2B cells in 100 mm dishes were treated with chemicals at various concentrations for 24 h. To assess the effect of overexpression of Nrf2 on mEH protein level in BEAS-2B cells, cells were transfected with Nrf2-expressing plasmid for 24 h and 48 h as described above. At the time of harvest, cells were washed with PBS, trypsinized and centrifuged at 1000×g for 3 min. For making whole cell lysates, cells were lysed in RIPA buffer (50 mM Tris, pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with 1× protease inhibitor cocktail (539131, Calbiochem). The cell lysates were centrifuged at 16,000×g for 10 min at 4°C and the supernatants were collected as whole-cell lysate. Protein concentrations were determined by Pierce 660 nm Protein Assay (Thermo Scientific). The extracted proteins (30 μg) were separated on a 10% denaturing polyacrylamide gel (Bio-Rad) and transferred to a PVDF membrane (Bio-Rad). After blocking in 5% skim milk for 30 minutes, the blots were incubated sequentially with primary antibodies at the dilution of 1:1000 and horseradish peroxidase-conjugated secondary antibodies at the dilution of 1:5000. The membranes were washed three times with 1×TBS/0.1% Tween 20, treated with Pierce ECL Western Blotting Substrate (Thermo Scientific), and exposed to ImageTek-H X-ray films (American X-Ray & Medical Supply). The antibodies used for immunoblotting were as follows: anti-mEH (sc-135984, Santa Cruz Biotechnology), anti-Nrf2 (sc-13032, Santa Cruz Biotechnology), anti-β-Actin (sc-81178, Santa Cruz Biotechnology), and anti-GAPDH (G9545, Sigma). The experiments were repeated three times and the most representative results were shown. 53

Nrf2 siRNA knockdown studies To reduce endogenous Nrf2 and assess its effects on the E1b expression in A549 cells, cells were transfected with either control or Nrf2 siRNA at 25nM using the Lipofectamine RNAiMAX reagent with a Forward Transfection Protocol according to the manufacturer’s instructions. Briefly, cells were seeded a day before transfection in 6-well plates at a density of 3×105 cells per well or in 60 mm petri dishes at a density of 7×105 cells per dish. At the day of transfection, the transfection complexes of the Lipofectamine RNAiMAX reagent and the given siRNA were added to each well containing cells. After 48 h, cells in 6-well plates were harvested for RT-PCR analysis and cells in 60 mm petri dishes were collected for western blotting. To assess the effect of Nrf2 knockdown on tBHQ-mediated induction of E1b expression in BEAS-2B cells, cells in 6-well plates were transfected with 25nM Nrf2 siRNA in the same manner as described above. After 36 h, siRNA- transfected cells were treated with tBHQ at 40 µM for an additional 24 h and then harvested for RT-PCR analysis. The experiments were repeated three times and the most representative results were shown.

Chromatin immunoprecipitation (ChIP) assay A549 and BEAS-2B cells were grown to 80-90% confluence in 100 mm dishes. Cells were harvested by trypsinization and fixed in 1% formaldehyde (252549, Sigma) at 25ºC for 10 min with slow agitation. The fixation was stopped by addition of glycine to a concentration of 0.125M. After a 5 min incubation at 25ºC, cells were pelleted by centrifugation at 1000×g for 5 min and then washed with ice-cold phosphate-buffered saline twice. Cells were lysed for 10min on ice in SDS lysis buffer (1% SDS, 10 mM EDTA, and 50 mM Tris-HCl, pH 8) with protease inhibitor cocktail (539131, Calbiochem). Cells were then sonicated with a Bioruptor sonicator (Diagenode, Liège, Belgium) for 5 cycles of 30 sec ON and 30 sec OFF at HIGH setting in a refrigerated water bath. Sheared cross-linked chromatin was centrifuged at 12,000×g for 10 min at 4°C and diluted 10-fold in ChIP Dilution Buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH8, and 167 mM NaCl) with protease inhibitor cocktail. The diluted chromatin was pre-cleaned overnight at 4°C with 35 μl protein A/G Plus-agarose beads (sc-2003, Santa Cruz Biotechnology) which were pre-blocked with sonicated salmon sperm DNA (201190, Stratagene) and BSA (2930, EM Science). Pre-cleaned chromatin was then incubated overnight at 4°C with 4 μg of anti-Nrf2 antibodies (sc-13032, Santa Cruz Biotechnology) as well as normal rabbit IgG (2729s, Cell Signaling Technology). To collect the antibody-chromatin complex, 75 μl protein A/G Plus-agarose beads pre-blocked as above were added, incubated for 3 h with rotation at 4°C, and pelleted by centrifugation at 5000×g for 1 min. The pelleted complexes were then washed sequentially with Low Salt Immune Complex Wash Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8, 150 mM NaCl), High Salt Immune Complex Wash Buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl, pH 8, 500 mM NaCl), and LiCl Immune Complex Wash Buffer (0.25 M LiCl, 1% IGEPAL CA630, 1% deoxycholic acid, 1 mM EDTA, 10 mM Tris-HCl, pH 8), followed by two washes with TE buffer. Precipitated Protein-DNA complexes were eluted

twice with 100 µl elution buffer (1% SDS, 0.1 M NaHCO3) for 15 min at 25ºC. To reverse crosslinks, the eluates were incubated at 65°C for 4 h in the presence of 8 µl of 5 M NaCl and 1 µl of 10 mg/ml RNase A. Proteins were 54 digested with 2 µl of 10 mg/ml proteinase K for 2h at 45°C in the presence of 4 µl of 0.5 M EDTA, pH8.0, and 8 µl of 1 M Tris-HCl, pH8. DNA was purified with ChIP DNA Clean & Concentrator kit (D5205, Zymo Research). Immunoprecipitated DNA was amplified and the PCR amplicons were analyzed on 1.5% agarose gels. The primer sets used for detection were: E1b proximal promoter (5’-GCGGACCGCCCTTTAAGTAGCC-3’ and 5’- GATCTCTCCGGCTCCCTGGCTC-3’), DNase I HS-1 #1 (5’-GCCACAGGGTAGGGGAGGCAAA-3’ and 5’- GGCTGGATCCTTGGCAACGCTT-3’), DNase I HS-1 #2 (5’-AGCTCACACTGAGCAGCCTCCC-3’ and 5’- ACCCACCAGGCAACAGACGGAA-3’), DNase I HS-2 (5’-CCCAACTGTCGCAGGGCTGG-3’ and 5’- CCGCCTGCCCAAAGACTCCC-3’), E1 proximal promoter (5’-CTGGTAGTGCTGGTGGGGGC-3’ and 5’- CCCACTCCCCACGGCCTTCT-3’), and Exon 3 flanking region (5’-CACCCCACCTTTGGAGGACAGC-3’ and 5’-ACATCCCTCTCTGGCTGGCGTT-3’).

Electrophoretic mobility shift assays (EMSA) The nuclear extracts from untreated A549 cells or SFN-treated BEAS-2B cells were prepared with NE- PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific) according to the manufacturer’s instructions. Double-stranded E1b HS-2 ARE probe was end-labeled with [γ-32P] ATP by T4 polynucleotide kinase (New England Biolabs, Ipswich, MA). For EMSA the DNA-binding reactions containing 2 µg of nuclear extracts, 20 fmol of labeled probes, 0.01 mg/ml sonicated salmon sperm DNA(201190, Stratagene), 2 µl of 5×

binding buffer [20% (v/v) glycerol, 5 mM MgCl2, 2.5 mM EDTA, 2.5 mM dithiothreitol, 250 mM NaCl, 50 mM Tris-HCl (pH 7.5), 0.25 mg/ml poly(dI-dC)] in a final volume of 10µl, were incubated with or without unlabeled competitor for 20 minutes at room temperature. The DNA-protein complexes were resolved by electrophoresis through a nondenaturing 4% polyacrylamide gel in 0.5× TBE buffer. Subsequently, gels were dried and exposed to X-ray film with intensifying screens at -70ºC. The probe sequences were as follows: E1b HS-2 ARE: 5’- ATGGCAAGTGCTGAGTCATGGCCGGCTAGCA-3’; mutant E1b HS-2 ARE: 5’- ATGGCAAGTaaTGAaTtATGGCCGGCTAGCA-3’; human NQO1 ARE: 5’- CAGTCACAGTGACTCAGCAGAATCT-3’; and mutant NQO1 ARE: 5’- CAGTCACAtaGttTCAcaAGAATCT-3’. The mutated nucleotides are in lowercase.

RESULTS

Nrf2 activators induce E1b expression in human lung epithelial cell lines To investigate a potential role for Nrf2 in human mEH regulation, we treated the human bronchial epithelial cell line, BEAS-2B, and, the human lung adenocarcinoma epithelial cell line, A549, with various concentrations of Nrf2 inducers for 12 or 24 h. Treatment of BEAS-2B cells with SFN or tBHQ increased E1b transcription in dose- and time-dependent manners (Figure 3-1A). In A549 cells SFN and tBHQ did not induce E1b transcription (Figure 3-1B). However, this result was not surprising because A549 cells possess a KEAP1 mutation that leads to constitutive Nrf2 activation. The mEH protein levels in SFN- or tBHQ-treated BEAS-2B and A549 cells correlated well with observed E1b transcript levels (Figure 3-1C). To further explore a role for Nrf2 in induction of E1b expression, BEAS-2B cells were transiently transfected with Nrf2 expression plasmid. Overexpression of Nrf2 significantly increased E1b transcription and protein levels of mEH compared with cells transfected with empty vector (Figure 3-2A and 3-2B). Further, siRNA knockdown of Nrf2 in BEAS-2B cells blocked tBHQ-mediated induction of E1b mRNA and NQO1, a known Nrf2 target gene (Figure 3-2C). siRNA-mediated knockdown of Nrf2 in A549 cells resulted in a ~5-fold decrease constitutive Nrf2 expression, and a similar effect was observed for mEH and NQO1 (Figure 3-2D). The mEH and Nrf2 protein expressions were also decreased in Nrf2 siRNA-transfected A549 cells (Figure 3-2E). Together, these results suggested that Nrf2 is involved in the tBHQ-mediated E1b induction in BEAS-2B cells and constitutive E1b expression in A549 cells.

Identification of an Nrf2-responsive DNA enhancer element To investigate whether Nrf2 regulatory sites exist in the E1b promoter, we first tested the ability of SFN or tBHQ to activate transcription in a series of truncated E1b promoter-luciferase reporter constructs. Surprisingly, transcriptional activity driven by the E1b promoter was decreased by both SFN and tBHQ in BEAS- 2B cells (Figure 3-3A). The E1b proximal promoter was not able to drive reporter expression in a manner that mimicked the endogenous E1b transcript. Based on these results, we postulated that these E1b promoter regions lack Nrf2 regulatory elements. Thus, we explored the potential for regulatory elements in other regions. Genome- wide mapping of DNase I hypersensitive (HS) sites reveals two DNase I hypersensitive sites (HS-1 and HS-2) located in the 15kb region between the E1b and E1 promoters (Figure 3-3B). These sites are present in HepG2, normal human lung fibroblasts and other cell lines. Considering that DNase hypersensitivity is an indicator of active DNA cis-regulatory elements, we tested if these two elements are the missing regulatory elements that may restore a more endogenous-like E1b expression pattern to the reporter construct in response to Nrf2 activator. HS- 1 and HS-2 were cloned and inserted 5’ of the -300 bp E1b proximal promoter sequence in the pGL4 luciferase reporter plasmid. This promoter was chosen because it gave the highest basal transcriptional activity in previous studies (Liang et al., 2005). In A549 cells, the incorporation HS-1 stimulated transcriptional activity about 2-fold, while HS-2 resulted in a 14-fold increase, indicating that both elements were likely functional enhancers (Figure 56

3-3C). In BEAS-2B cells, HS-1 had no effect, while HS-2 increased transcriptional activity about 2-fold. (Figure 3-3C). In the next series of experiments, we tested whether the HS-1 and HS-2 DNA elements were responsive to Nrf2 activation. Luciferase reporter assays showed that both overexpression of Nrf2 and tBHQ treatments produced a significant induction of reporter activity in BEAS-2B cells, and the induction pattern closely resembled expression of endogenous E1b (Figure 4-4A, upper 2 panels). In contrast, A549 cells did not exhibit any transactivation to Nrf2 overexpression or tBHQ treatment, which is consistent with the expression profile of E1b transcript in this cell line (Figure 4-4A, lower panels). Since A549 cells display constitutive Nrf2 activation, we sought to determine whether the high enhancer activity of HS-2 was due to Nrf2 by testing the ability of KEAP1, an Nrf2 inhibitor, to inhibit transactivation. Overexpression of KEAP1 significantly reduced the enhancer activity in a dose dependent manner, indicating that constitutively activated Nrf2 contributes to increased luciferase activity (Figure 4-4B). Furthermore, KEAP1 expression led to the abrogation of antioxidant- inducted luciferase activity in BEAS-2B cells (Figure 4-4C). Based on these results, we suspected that the HS-2 enhancer contains a site that confers Nrf2 responsiveness.

Nrf2 binding to an antioxidant response element (ARE) in HS-2 Enhancer To investigate Nrf2 binding to HS-2 in vivo, chromatin immunoprecipitation (ChIP) assays were performed. Primers used in ChIP assays were specifically designed to amplify E1b proximal promoter, DNase I hypersensitive sites, E1 proximal promoter and Exon 3 regions (Figure 3-5A). In A549 cells, an Nrf2 antibody specifically enriched the region containing the HS-2 site in comparison to a control rabbit IgG (Figure 3-5B). The binding of Nrf2 to E1b proximal promoter was slightly enhanced over the control IgG. No binding was detected in other regions. In BEAS-2B cells, ChIP assays were performed to determine whether Nrf2 binds to HS-2 in response to Nrf2 activators. As shown in Figure 3-5C, SFN induced a significant increase in Nrf2 binding to the 2nd enhancer element while tBHQ treatment had a marginal induction. No binding was detected in the exon 3 region. Taken together, the ChIP analysis indicated an interaction between Nrf2 and HS-2 enhancer element, which further supports a critical role of HS-2 in Nrf2-mediated E1b induction. We next used Genomatix Software Suite to identify a putative ARE in the HS-2 enhancer element. DNA sequence alignment with known AREs from several human genes confirmed the presence of a perfect ARE (Table 3-1). Site-directed mutagenesis of the ARE site resulted in significantly decreased luciferase reporter activity in the A549 and BEAS-2B cells (Fig. 6A). Interestingly, mutation of either the nucleotides “TGA” at the 5’-end or “GC” at the 3’ end in the core ARE motif abolished the luciferase activity (Figure 3-6A). The results indicated that both motifs are essential for ARE function. In addition, mutation on the ARE abolished the responsiveness of BEAS-2B cells to SFN treatments (Figure 3-6B). These data suggested that this ARE within the HS-2 enhancer is essential for not only the enhancer activity, but also Nrf2-mediated induction of the E1b isoform. To verify the specific interaction of Nrf2 with this ARE in the 2nd enhancer, EMSA analysis was performed. Incubation of nuclear extracts prepared from untreated A549 cells and SFN-treated BEAS-2B cells 57 with double stranded oligonucleotides containing this ARE, resulted in three protein-DNA complexes (Figure 3- 6C, lane 1). Unlabeled probe containing the ARE from the 2nd enhancer competed out the first and third complex completely and weakened the intensity of the second complex significantly (Figure 3-6C, lane 2), whereas a mutant ARE probe was not able to compete with the wild-type probe in the formation of the first complex, but weakened the intensity of the second band (Figure 3-6C, lane 3). The results suggested that the first complex was specific for Nrf2 binding and two other binding sites were nonspecific. Furthermore, an oligonucleotide containing human NQO1 ARE prevented the formation of the first and second complexes, but not the third (Figure 3-6C, lane 4), while the mutated NQO1 ARE oligonucleotide had no effect on the first complex formation (Figure 3-6C, lane 5). In summary, these data strongly suggest a specific Nrf2 interaction with the ARE in the HS-2 element.

Characterization of a TPA-response element overlapping with the ARE within HS-2 Our analysis of the HS-2 element also revealed a consensus TPA (12-O-tetradecanoylphorbol-13- acetate)-response element (TRE) that overlaps with the ARE site. Thus, we tested if this TRE affects Nrf2- mediated reporter activity in A549 cells. TPA treatments significantly decreased the transactivation of the HS-2 enhancer in a dose-dependent manner (Figure 3-7A). Furthermore, A549 cells treated with TPA exhibited a decreased level of mEH (Figure 3-7B). Overexpression of JUN (AP-1), a transcription factor that binds TRE, repressed the transcriptional activity of the HS-2/E1b-300 luciferase reporter plasmid in A549 cells (Figure 3-7C). In addition, JUN blocked the tBHQ-induced transactivation of the HS-2/E1b-300 reporter construct in BEAS-2B cells (Figure 3-7D). These data suggested that the binding of JUN to the TRE embedded in the ARE interfered with the binding activity of Nrf2 to its target motif in HS-2 and prevented the up-regulation of E1b promoter activity and E1b expression.

DISCUSSION Our previous studies characterized the expression of two human mEH transcripts, E1 and E1b, which result from the use of alternative promoters. These transcripts are expressed in a tissue-specific manner, yet little is known about their regulation. The results presented in the current study demonstrate a role for the transcription factor Nrf2 in the regulation of the human mEH E1b transcript. The regulatory elements involved in Nrf2- mediated regulation of E1b transcript are unique in that they are located in the intronic region between the two alternative first exons, downstream of E1b. A DNase hypersensitive site identified in the intronic region, HS-2, that is present in human cell lines and normal human liver cells, conferred responsiveness to Nrf2 inducers and contained an Nrf2 ARE binding element. AREs are cis-acting regulatory elements present in the promoter region of cytoprotective genes targeted by Nrf2. Functional AREs contain a core sequence of RTGACnnnGC, where ‘n’ represents any nucleotide (Rushmore et al., 1991). Further characterization of nucleotides flanking to the core ARE sequences demonstrates 58 that the flanking sequences are required for both basal and inducible activity of AREs. Thus, the consensus sequence of AREs was extended to TMAnnRTGAYnnnGCRwwww (Wasserman and Fahl, 1997). Other minor revisions of this consensus sequence have also been suggested (Erickson et al., 2002; Nioi et al., 2003). It is notable that some AREs contain a JUN binding site (also known as TPA-response element or TRE, 5’- TGAGTCA-3’). Based on the length of ARE sequences and the presence of a JUN binding site, AREs have been divided into four categories (Hayes et al., 2010). Class 1 AREs have an extended ARE consensus sequence, as well as an embedded JUN-binding site, while Class 2 AREs have an extended ARE sequence without a JUN site. Class 3 and 4 AREs have a minimal core ARE sequence. The difference between Class 3 and 4 is that a JUN site is present in Class 3 AREs, but not in Class 4 AREs. The intronic ARE of mEH identified in this study is a Class 3 ARE which contains only minimal core sequence and an embedded JUN site. AREs in different classes display differential responses to chemopreventive agents. Class 1 and 2 AREs are likely more responsive than those in Class 3 and 4 (Hayes et al., 2010). Therefore, it is not surprising that antioxidants only induced a moderate change on mEH expression (about 2-4 fold). The embedded JUN binding sites in Class 1 and 3 AREs may also respond to other factors that activate JUN, such as UV radiation and TPA. Indeed, we found that JUN overexpression in A549 cells attenuated the transactivation activity of the HS-2 enhancer in luciferase reporter assays. Therefore, further study of mEH regulation by bZIP transcriptional factors and AP-1 activators through this enhancer element is warranted. Although site-directed mutagenesis of the ARE in HS-2 completely abolished its responsiveness to Nrf2 activators, the mutated ARE still enhanced the promoter activity of E1b proximal promoter. These results indicate that other transcription factors may bind to the HS-2 region and confer transcriptional enhancement of the E1b proximal promoter. Further studies are needed to explore this possibility. Thus far, identified AREs are located in the upstream regions of transcriptional start sites (Chorley et al., 2012; Hayes et al., 2010). Interestingly, the HS-2 ARE identified in this study is located in the intronic region between E1b and E1, downstream of the E1b promoter that it appears to regulate. The large distance (~7kb) separating the HS-2 from the E1b transcription start site suggests that intrachromosomal interaction resulting from chromatin looping may be involved in promoting transcriptional activation of E1b mediated by Nrf2. It is well established that chromatin looping brings transcription factors binding to the distal regulatory elements into close proximity with the general transcription machinery at the transcription start site (Kadauke and Blobel, 2009). This enables the distal response element and its associated transcription factors to direct the transcription activity of proximal promoter. Many studies have demonstrated interactions between promoters and intronic regulatory elements (Deshane et al., 2010). In these studies, the Chromatin Conformation Capture (3C) assay was an essential tool used to dissect this mechanism (Miele and Dekker, 2009). To determine whether a similar looping interaction between the E1b promoter and the HS-2 enhancer occurs, and the potential role of Nrf2 in this interaction, our laboratory efforts are currently vigorously focused on pursuing the 3C technology to define the nature of these interactions within the context of mEH gene regulation. 59

The intronic enhancers in human mEH gene appear to be important drivers of the transcriptional activity exhibited by the E1b proximal promoter. Based on the chromatin looping model for the enhancer-promoter interaction, there likely exists a region within the E1b promoter that communicates with the distal enhancers. The studies described in Chapter 2 identified Sp1 and Sp3 as factors that interact with the E1b proximal promoter region to regulate basal expression of E1b. In other research, Sp1 has been shown to regulate chromatin looping between intronic enhancers and upstream promoter regions of several genes (Deshane et al., 2010). Thus, we speculate that Sp1 participates in the formation of a chromatin loop between the intronic enhancers and the E1b proximal promoter by interacting with Nrf2 or other factors binding to these enhancers. However, other E1b promoter-binding factors may also mediate the DNA looping interaction. Therefore, although the factors contributing to the E1b chromatin looping remain to be elucidated, they represent unique targets as E1b expression modulators that may in turn be regulated through the action of xenobiotics at this distal enhancer. Nrf2 is a therapeutic target for cancer chemoprevention (Kwak and Kensler, 2010). Nrf2 induces the expression of cytoprotective genes, including NQO1, several UGT and GST isoforms, HO-1, multidrug resistance-associated proteins (MRPs) and others (Klaassen and Reisman, 2010). Nrf2 is essential in protecting cells from the toxic and carcinogenic effects of electrophiles, oxidants and inflammatory agents, as it has been demonstrated that Nrf2-deficient mice are more sensitive to oxidative stress and reactive electrophiles than their wild type counterparts (Slocum and Kensler, 2011). The protective role of Nrf2 has also been established in many type of diseases, including cancers, Alzheimer’s and Parkinson’s diseases, chronic obstructive pulmonary disease, asthma, diabetes, and inflammatory diseases (Zhang, 2010). However, recent clinical studies using lung tumor tissues and cell lines showed that Nrf2 is excessively expressed due to mutations in KEAP1 or Nrf2 (Ohta et al., 2008; Shibata et al., 2008b; Singh et al., 2006). Further, enhanced Nrf2 activity promotes cancer cell survival and contributes to chemo- and radio-resistance (Zhang, 2010). Therefore, the consequences of Nrf2 activation in carcinogenesis are biological context-dependent. In these respects, Nrf2 appears as s clinical target to direct the inhibition of tumor growth and for enhancing the efficacy of anti-cancer drugs. Similarly, mEH functions in a dual role in the metabolism of xenobiotics. On one hand, mEH activity is pivotal in the protection against toxicities elicited by reactive epoxide intermediates. On the other, for example in the metabolism of polycyclic aromatic hydrocarbons, mEH activity facilitates DNA adduct formation and subsequent processes of mutagenesis and cancer. To fully understand the consequence of mEH up-regulation by Nrf2 activation in human cancers, it will be necessary to establish a model that closely resembles the in vivo human situation. The current rodent models are not suitable in these respects since the rodent mEH regulatory regions are not conserved in humans (Yang et al., 2009). Further, the human mEH regulatory mechanisms are different from rodents in that the human mEH gene is subject to alternative promoter usage. The presence of the two functional enhancers identified in this research additionally contribute to the species differences in play, as they likely represent host-specific response elements for transcription factors that could be modulated by many stimuli. Therefore, the results of these studies provide a compelling rationale for need to develop humanized 60 rodent models that maintain regulatory features of the human mEH gene. These models do not yet exist but their future development will provide the necessary platform for more detailed study of the mEH molecular regulation as well as the related disease implications to human systems. In summary, the results derived from the studies presented demonstrate that Nrf2 contributes a critical role in the antioxidant mediated transcriptional regulation of human mEH through the gene’s distal enhancer region The data obtained from these investigations provide new insights regarding the nature of human mEH gene regulation, its modulation by xenobiotic inducers and the identification of Nrf2 as a potential therapeutic target for cancer chemoprevention.

Table 3-1. Alignment of HS-2 ARE with known AREs. ARE Sequence E1b DNaseI HS-2 AGCCGGCCATGACTCAGCACTTGCC human NQO1 CAGTCACAGTGACTCAGCAGAATCT human GCLC CCTCCCCGTGACTCAGCGCTTTG human GCLM GAAGACAATGACTAAGCAGAAAA ARE core motif GTGACnnnGC ARE consensus TMAnnRTGAYnnnGCRwwww AP-1 (TRE) TGACTCA MAF-recognition element TGCTGACTCAGCA

A. BEAS-2B B. A549

Figure 3-1. Antioxidants induce E1b expression. A.) BEAS-2B cells and B.) A549 cells were treated with DMSO, Sulforaphane or tBHQ for different time periods and E1b mRNA expression was analyzed by the real time quantitative PCR. GAPDH was used as an internal control. Results are expressed as the mean ± SD from two separate experiments. Significant differences from DMSO control are indicated by “*” (p<0.05). C.) The induction of mEH protein in BEAS-2B and A549 cells was evaluated by western blotting. Whole cell lysates were immunoblotted to detect mEH protein. GAPDH was used as a loading control. All experiments were performed three times, and one representative set of data are showed.

A. D. 3 1.5 Empty Vector Control siRNA Nrf2 Nrf2 siRNA

2 * 1.0

(Fold Change) 1 0.5

E1b mRNA Expression E1b mRNA * Relative mRNA change fold * * 0 0.0 24 H 48 H E1b NQO1 Nrf2 B. E.

C. 10 * 8 E1b 6 NQO1 4 * Nrf2 3

2 (Fold Change) 1 *

Relative mRNA Expression mRNA Relative * 0 * DMSO 40M tBHQ DMSO 40M tBHQ Control siRNA Nrf2 siRNA

Figure 3-2. Effect of overexpression and siRNA knockdown of NRF2 on E1b expression. BEAS-2B cells were transfected with either empty expression vector (Ctrl) or NRF2-expressing plasmid. At 24 and 48 h post- transfection, total RNA and proteins were extracted. A.) Real time qPCR analysis of E1b expression. GAPDH was used as an internal control and results were normalized to empty vector control. B.) Western blot analysis of mEH protein levels with GAPDH as a loading control. C.) qPCR analysis of t-BHQ-induced expression of E1b in BEAS-2B cells transfected with Nrf2 siRNA. BEAS-2B cells were transfected with either control or Nrf2 siRNA for 48 h and then treated with tBHQ for another 12 h. D.) A549 cells were transfected with either control or Nrf2 siRNA. Forty-eight hours after siRNA transfection, cells were harvested for total RNA or total proteins. The expressions of mEH, NQO1 and Nrf2 genes were measured with Real time quantitative PCR analysis. β-actin was used as an internal control and all values were normalized to DMSO-treated control siRNA values. E.) Western blot analysis of mEH protein levels with β-actin as a loading control. All qPCR values are expressed as the mean ± SD. Significant differences from respective control were indicated by “*” (p<0.05).

A. -2.2kb * DMSO * 2M SFN * 20M tBHQ -2.0kb * 40M tBHQ ** -1.0kb * ** -300bp * * 0.00.51.01.5 Relative Luciferase Activity (Fold Change)

Figure 3-3. Identification of regulatory elements involved in E1b induction. A.) Reporter analysis on E1b upstream sequence. BEAS-2B cells were transfected with luciferase reporter constructs driven by E1b promoters containing -300bp, -1.0kb, -2.0kb or -2.2kb promoter region. After 24h treatment with Nrf2 inducers, cells were harvested for luciferase reporter assays. B.) A map of the intronic region located between E1b and E1 harboring DNase hypersensitive sites in A549, HepG2, K562 and NHLF cells (modified from the UCSC genome browser). C.) Reporter analysis on DNase I hypersensitive sites. A549 and BEAS-2B cells were transfected with the E1b - 300 or E1b DNase I HS/E1b-300 plasmids for 24hr and harvested for luciferase assay. All luciferase reporter activity values are expressed as the mean ± SD. Significant differences from DMSO control or the E1b-300 plasmid were indicated by “*” (p<0.05).

A. BEAS-2B

A549

Figure 3-4. Role of DNase HS sites in Nrf2-mediated induction of E1b. A.) Effect of Nrf2 overexpression or tBHQ treatment on transactivation activity of E1b DNase I HS sites in BEAS-2B and A549 cells. For Nrf2 overexpression, cells were co-transfected with E1b promoter reporter constructs containing HS-1 or HS-2 and varying amounts of Nrf2. Luciferase activity was determined 24 h later. For tBHQ induction experiments, cells were transfected with promoter reporter constructs containing HS-1 or HS-2 and treated with varying amounts of tBHQ 18 h later. Luciferase activity was determined 6 h after treatment. KEAP1 down regulates HS-2 enhancer activity in A549 cells (B) and SFN-treated BEAS-2B cells (C). Cells were co-transfected with the KEAP1 expression plasmid and the reporter vectors containing E1b-300 promoter or E1b DNase I HS-2/E1b-300 promoter for 24 h. A549 cells were harvested for luciferase assay while BEAS-2B cells were treated with 2µM SFN for extra 24 h before luciferase assay. All luciferase values represent mean ± SD. *p<0.05 compared to empty vector controls or DMSO control.

Figure 3-5. Nrf2 binds to HS-2 enhancer element. A.) Locations of specific primers in mEH gene used for ChIP assay. The binding of Nrf2 to the enhancer regions of human mEH gene was evaluated by ChIP assays in untreated A549 cells (B) and BEAS-2B cells (C) treated with Nrf2 activators for 2 h as described under “Material and Methods”. Cross-linked chromatin was isolated from formaldehyde-fixed A549 and BEAS-2B cells and incubated with anti-Nrf2 or control IgG. Nrf2 binding to E1b and E1 promoters and enhancer regions was analyzed by PCR with specific primers for these regions. I, Input, C, control IgG, N, anti-Nrf2 antibody.

E1b -300 BEAS-2B A549 ARE LUC

HS-2 E1b -300 LUC

mutARE HS-2 E1b -300 LUC * * mutARE-1 HS-2 E1b -300 LUC * * mutARE-2 HS-2 E1b -300 LUC * * 012340 5 10 15 Relative Luciferase Activity Relative Luciferase Activity (Fold Change) (Fold Change)

B. C. BEAS-2B cells DMSO 0.5M SFN E1b -300 LUC * * 1M SFN * 2M SFN HS-2 E1b -300 LUC * * HS-2 E1b -300 LUC * * ARE 012345 Relative Luciferase Activity (Fold Change)

Figure 3-6. Identification of Antioxidant response element (ARE) within HS-2. A.) BEAS-2B and A549 cells were transfected with E1b promoter constructs containing HS-2 (wild-type) or HS-2 with mutations in the core ARE motif. Luciferase values were determined 24 h later. B.) Mutation of the putative Nrf2-binding site influences SFN responsiveness of BEAS-2B cells. Cells were transfected with the reporter vectors containing E1b-300 promoter, E1b DNase I HS-2/E1b-300 promoter, or E1b DNase I mutARE/E1b-300 promoter for 18hr and treated with DMSO or SFN for 6hr before harvested for luciferase assay. All luciferase values represent mean ± SD. *p<0.05 compared to empty vector controls. C.) EMSA analysis of Nrf2 binding to the putative ARE in E1b DNase I HS-2 in SFN-treated BEAS-2B and A549 cells. E1b DNase I HS-2 ARE was end-labeled with [γ- 32P]ATP and T4 kinase and incubated with nuclear extracts from untreated A549 cells or BEAS-2B cells treated with 2 μM SFN for 2 h. The binding of Nrf2 to E1b DNase I HS-2 ARE was competed with cold E1b DNase I HS-2 ARE (Lane 2), mutated E1b DNase I HS-2 ARE (Lane 3), ARE from human NQO1 gene (Lane 4), or mutated NQO1 ARE (Lane 5). The arrows show specific binding.

A. C. 25 15 TPA-0nM JUN-0ng TPA-0.1nM JUN-60ng 20 TPA-0.5nM JUN-120ng TPA-1.0nM 10 JUN-180ng * 15 * * 10 5 (Fold Change) (Fold Change) * 5 * Relative LuciferaseActivities Relative Luciferase Activities 0 * 0

E1b -300 LUC HS-2 E1b -300 LUC E1b -300 LUC HS-2 E1b -300 LUC

B. D. 10 DMSO 8 40M tBHQ *

4 (Fold Change) 2 *

Relative Luciferase Activities Luciferase Relative 0 Vector JUN Vector JUN

E1b -300 LUC HS-2 E1b -300 LUC

Figure 3-7. A TRE binding site influences ARE-driven enhancer activity. A.) TPA treatments down-regulated the enhancer activity in A549 cells. Cells were transfected with the E1b -300 or E1b DNase I HS-2/E1b-300 plasmids for 18hr, treated with TPA for 6hr and harvested for luciferase assay. B.) Western blot analysis of mEH expression in A549 cells after treated with TPA for 24 and 48 h. GAPDH was used as a loading control. C.) Regulation of the E1b DNase HS-2 enhancer activity by JUN overexpression in A549 cells. Cells were co- transfected with the JUN expressing plasmid and the reporter vectors containing E1b-300 promoter or E1b DNase I HS-2/E1b-300 promoter. After 24 h, luciferase activity was measured. D.) JUN blocks tBHQ induced activation of E1b DNase I HS-2 enhancer activity in BEAS-2B cells. Cells were co-transfected with the JUN expressing plasmid and the reporter vectors containing E1b-300 promoter or E1b DNase I HS-2/E1b-300 promoter for 18 h and treated with DMSO or tBHQ for 6 h before harvested for luciferase assay. All luciferase values represent mean ± SD. *p<0.05 compared to empty vector controls.

Chapter 4 The transcriptional regulation of human microsomal Epoxide Hydrolase (mEH) driven by two alternative promoters in human hepatoma C3A cells

ABSTRACT The transcription factors AhR and Nrf2 are well studied for their roles in regulating induction of xenobiotic metabolizing and detoxifying enzymes. Some xenobiotics, designated as bifunctional inducers, can induce phase I and phase II enzymes by activating both AhR and Nrf2 signaling pathways, while monofunctional inducers activate Nrf2 selectively and induce only phase II enzymes. Human microsomal epoxide hydrolase (mEH) is a metabolic enzyme contributing an important role in xenobiotic detoxification and bioactivation of polycyclic aromatic hydrocarbons (PAH). Its expression is regulated in part by the use of two alternative promoters, resulting in the production of unique transcripts, in particular the transcripts termed E1b and E1. In this study, we investigated the regulation of E1b and E1 transcript expression by both monofunctional and bifunctional inducers in human hepatoma HepG2 C3A cells. Quantitative real-time PCR results demonstrated that two monofunctional inducers, sulforaphane (SFN) and tert-butylhydroquinone (tBHQ), upregulated E1b and E1 transcript levels in both dose- and time-dependent manners. Knockdown of Nrf2 with siRNA revealed that Nrf2 was involved in tBHQ-mediated induction of E1b and E1. Luciferase reporter assays suggested that intronic regulatory elements were responsible for the inducible expression of E1b and E1 by SFN. Bifunctional AhR agonists displayed a considerably significant variation in regulation of E1b and E1 transcripts. Treatment with 3- methylcholanthrene (3-MC) activated E1b expression at 24 h, but not 12 h. In contrast, E1 expression was suppressed by 3-MC treatment both at 12 and 24 h. Another bifunctional inducer, β -naphthoflavone (β-NF), upregulated E1b and E1 expression at both 12 and 24 h time points while treatment with the arylhydrocarbon receptor (AhR) agonist,TCDD, resulted in no effect on E1b expression, but decreased E1 levels, similar to 3-MC. An AhR antagonist, 6,2',4'-trimethoxyflavone (TMF), reversed the effects of 3-MC treatment on E1b and E1 expression, indicating AhR involvement. Results from Nrf2 siRNA studies suggested that induction of E1b and E1 by β-NF was dependent on Nrf2, while 3-MC-mediated induction of E1b was not dependent on Nrf2. Luciferase reporter assays were also conducted and showed that 3-MC treatment inhibited the promoter activities of both E1b and E1, but TMF additions only inhibited the E1 promoter activities. Finally, we assessed the inducible expression profiles of E1b and E1 following xenobiotic treatments in various cell types. Results of real- time PCR analyses demonstrated that the chemically induced E1b and E1 expression profiles in HepG2 cells more closely resembled C3A cells than Huh7 cells. When primary cultures of human hepatocytes were treated with 3- MC and TCDD, induction of both E1b and E1 transcript levels was observed. Overall, the results of these studies indicate that the regulation E1b and E1 by chemical inducers occurs in a cell-type specific manner, and is complex, likely involving multiple pathways. 73

INTRODUCTION Humans are exposed rather continuously to xenobiotics, such as drugs, food additives, industrial chemicals, and environmental pollutants. To mitigate the effects of these exposures, , xenobiotics undergo a broad range of biotransformation processes that in general, render them more water-soluble to facilitate more rapid elimination, and in the process, less toxic. The metabolism of xenobiotics is carried out by a collection of xenobiotic metabolizing enzymes (XMEs) including the phase I and phase II metabolizing enzymes as well as phase III chemical transporters (Xu et al., 2005). Expression of XMEs is coordinately controlled by various transcription factors that act as xenobiotic sensors (Xu et al., 2005). The aryl hydrocarbon receptor (AhR) and nuclear factor (erythroid-derived 2)-like 2 (Nrf2) are two well characterized transcription factors that regulate expression of XMEs. AhR is a member of the basic helix-loop- helix/Per-Arnt-Sim (bHLH/PAS) family. In its nonliganded state, AhR is sequestered in the cytosol as part of a multiprotein complex (Omiecinski et al., 2011). Upon ligand binding, the AhR complex translocates to the nucleus where AhR is released and dimerizes with the aryl hydrocarbon receptor nuclear translocator (ARNT). The liganded AhR/ARNT complex in turn binds to xenobiotic response elements (XREs) in the promoters of many XMEs to control their transcriptional activity. Unlike other xenosensing transcription factors, Nrf2 is not activated by direct ligand binding. The activation of Nrf2 is dependent on Keap1 (kelch-like ECH-associated protein 1) (Zhang, 2010). Keap1 functions as an adaptor protein between Nrf2 and the Cullin-3-based E3-ligase ubiquitination complex, which leads to continuous ubiquitination and proteasomal degradation of Nrf2 (Kobayashi et al., 2004). Upon exposure to electrophilic /oxidative stress, Keap1 reactive cysteine residues are modified, leading disruption of the complex, thereby decreasing the degradation of Nrf2. Consequently, Nrf2 translocates to the nucleus, binds to antioxidant response elements (ARE) in the promoter region of target genes, and activates their transcription (Itoh et al., 1997b). Many XME inducers engage in cross-talk with different signaling pathways. Polycyclic aromatic hydrocarbons (PAH), β-naphthoflavone (β-NF) and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), are classified as bi-functional inducers as they induce the expression of phase I and phase II enzymes in a coordinated manner. The induction of XMEs by these inducers requires both AhR and Nrf2 signaling pathways as delineated through studies using AhR-, ARNT- and Nrf2-deficident cells or mice (Ma et al., 2004; Noda et al., 2003; Yeager et al., 2009). PAHs and β-NF interact with the AhR directly to induce phase I enzymes, and then undergo transformation by these enzymes to reactive intermediates that lead to oxidative stress and in turn, trigger Nrf2 signaling. Besides this indirect interaction between the AhR-XRE and Nrf2-ARE signaling pathways, mouse Nrf2 gene expression is directly up-regulated by TCDD-mediated AhR activation (Miao et al., 2005). In contrast, tBHQ and sulforaphane are monofunctional inducers that function to more specifically upregulate certain phase II enzymes through the Nrf2 signaling pathway. Interestingly, activation of Nrf2 enhances expression of AhR, CYP1A1 and CYP1B1 in Nrf2+/+ mouse embryonic fibroblasts (MEF), and Nrf2 binding to an ARE in the 74 proximal promoter region of AhR has been demonstrated (Shin et al., 2007). Thus, the regulation of XMEs expression is influenced at multiple levels by AhR and Nrf2. Microsomal epoxide hydrolase (mEH) is an important biotransformation enzyme involved in both detoxification and bioactivation of xenobiotics (Fretland and Omiecinski, 2000). Human mEH is encoded by a single gene in Chromosome 1 and its expression is driven by two alternative promoters, residing approximately 15 kb apart (Liang et al., 2005). The two major transcripts, termed E1 and E1b, respectively, encode the same protein since the protein translation start site is located in Exon 2. It is noteworthy that the E1 promoter is exclusively active in liver, while the E1b promoter drives the mEH expression in all tissues, including liver (Liang et al., 2005; Yang et al., 2009). A previous report demonstrated that several prototypic chemical inducers, including phenobarbital (PB), butylated hydroxyanisole (BHA), dexamethasone (DEX), Arochlor 1254 and β-NF, induced mEH expression in human primary hepatocytes (Hassett et al., 1998). Interestingly, antioxidants have also been reported to up-regulate mEHexpression in mice and certain mammalian cell lines (Burchiel et al., 2007; Cha et al., 1978; Hayes et al., 2010). However, the previous studies did not characterize these chemical effectors with respect to their selective effects on the human mEH alternative promoters, nor were any molecular mechanisms of their mEH induction clarified. In the current investigation, we examined the ability of prototypic chemical inducers and antioxidants to differentially regulate the expression of the two prominent mEH transcripts, E1 and E1b. Using both bifunctional and monofunctional inducers, we studied the potential mechanistic roles contributed by AhR and Nrf2 in regulation of human mEH expression. Although other cell models were also used in these studies, the majority of the analyses were conducted using human hepatoma C3A cells as these cells express both the E1 and E1b mEH transcripts and exhibit reasonably well differentiated characteristics of human hepatocytes.

MATERIALS AND METHODS

Materials The protease inhibitor mixtures were from Calbiochem; FuGENE 6 Transfection Reagent and the dual luciferase reporter assay system were from Promega; all cell culture media and supplies were from Invitrogen; TRIzol Reagent was from Invitrogen; sulforaphane (Cat#574215) was from CalBiochem; tert-butylhydroquinone (tBHQ, Cat#B0833) was from TCI America; TCDD was a gift from Dr. Gary Perdew, Penn State University; 3- MC (Cat#21394), β-NF (Cat#N3633), 6,2',4'-trimethoxyflavone (TMF, Cat#T4080) and all other chemicals were purchased from Sigma unless otherwise indicated. Small interfering RNA (siRNA) for Nrf2 mRNA (siRNA ID: s9493) or Silencer® Select Negative Control No. 1 siRNA and the Lipofectamine RNAiMAX reagent were purchased from Invitrogen. The mouse monoclonal anti-mEH (sc-135984) and anti-Actin (sc-81178) antibodies and rabbit polyclonal anti-Nrf2 (sc-13032) and anti-TBP (TFIID, sc-273) antibodies were from Santa Cruz Biotechnology; the rabbit polyclonal anti-GAPDH (G9545) was from Sigma. 75

Plasmids The E1b promoter luciferase reporter constructs containing 300bp (E1b-300/+46), 1kb (E1b-1kb/+46), or 2kb (E1b-2kb/+46) of the 5’-flanking region upstream of E1b and the E1 promoter luciferase reporter constructs containing 128bp (E1-128/+72), 208bp (E1-208/+72), or 1428bp (E1-1428/+72) of the 5’-flanking region upstream of E1 were constructed as describe previously (Liang et al., 2005). The -208/+72 region of E1 has been shown to contribute the majority of the basal transcriptional activity for the E1 promoter whereas the -300/+46 region of E1b contributes the majority of E1b’s promoter activity (Liang et al., 2005). Therefore, these core promoter regions were used to study the additive transcriptional activities contributed by DNase I enhancer elements localized to distal gene regulatory regions surrounding the human mEH gene. To generate the intronic enhancer/E1b-300 or E1-208 promoter luciferase reporter constructs, the regions containing DNase I hypersensitive sites located between E1b and E1 were amplified by PCR from human genomic DNA using primers as follows: DNase I HS-1(5'-ATGCGCTAGCTCAGGAAAGGAATGTGTAGGAGGG-3' and 5'- ATGCCTCGAGAAATGCTGGGATTACAGGTGTGC G-3'); and DNase I HS-2 (5'- GATCGGTACCTCCCTTTCCACTGGATGTTCCCTT-3' and 5'- GATCTCTAGAACCATAAGATGCAGGAAGAGGGCT-3'). The PCR products were inserted to the E1b-300 or E1-208 proximal promoter-containing pGL3 luciferase reporter vector. The Nrf2 expression plasmid was generated by inserting the full length cDNA of Nrf2 (NM_006164.4) into the p3XFLAG-CMV-10 expression vector (Sigma). Primers for amplifying Nrf2 were as follows: Forward, 5'- GATCGCGGCCGCAATGGACTTGGAGCTGCCGCC.G-3', and reverse, 5'- GATCTCTAGACTAGTTTTTCTTAACATCTG G-3'. All constructs were confirmed by DNA sequencing.

Cell culture, transient transfection and luciferase reporter assays Human bronchial epithelial BEAS-2B cells, human hepatoma HepG2-derived C3A cells, HepG2 cells and Huh7 cells were purchased from American Type Culture Collection (Manassas, Virginia). All cell lines were

cultured in 5% CO2 incubator at 37°C in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% FBS, 2 mM L-glutamine, 0.1 mM Non-Essential Amino Acids, 1 mM Sodium Pyruvate , 10 mM HEPES, 0.15% sodium bicarbonate, 100 units/ml penicillin G, and 100 µg/ml streptomycin. Cells were cultured in 24-well or 6- well plates and 100 mm petri dishes and harvested according to the requirements of the experiments. Normal human hepatocytes were obtained and maintained as described previously (Goyak et al., 2008; Laurenzana et al., 2012). For transient transfections, C3A cells were seeded a day before transfection in 24-well plates at a density of 5×104 cells per well. To assess enhancement on E1b or E1 promoter activity by intronic regulatory elements, cells were transfected with reporter plasmids using a FuGENE 6 transfection protocol according to the manufacturer’s instructions. The pRL-CMV plasmid containing Renilla Luciferase cDNA was also co-transfected as an internal control for transfection efficiency. Cells were harvested 24 h post transfection and luciferase activity was measured and analyzed in a Veritas Microplate Luminometer (Turner Biosystems) using the Dual 76

Luciferase Reporter Assay System (Promega) as described previously (Auerbach et al., 2005). For studying the effect of chemical treatment on E1b and E1 promoter activities, the C3A cells were transfected with E1b and E1 promoter reporter constructs and pRL-CMV reporter plasmids for 6 h and were incubated for 24 h in culture medium containing the indicated concentration of chemicals or vehicle (0.1% DMSO). Luciferase activity was measured in the same manner as described above. All transfections were performed in triplicate and the results were expressed as means ± standard deviations (SD). The experiments were repeated three times and the most representative results are shown.

RNA isolation, reverse transcription and quantitative real-time PCR. For investigating inducible expression of various genes by chemical treatment, cultured cells in 6-well plates or normal human hepatocytes in 12-well plates were treated with chemicals for 24 h. Total RNA was extracted with TRIzol Reagent according to the manufacturer’s instructions. Two µg of total RNA was converted to cDNA using the High-Capacity cDNA Archive Kit (Applied Biosystems). cDNAs were analyzed with CFX96 Real-Time PCR Detection System (Bio-Rad) using PerfeCTa SYBR Green SuperMix (Quanta Biosciences). The final concentration of primers in each reaction was 0.2µM. The PCR conditions consisted of a 3 min initial denaturation at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. Each sample was run in duplicate and the results were normalized to the level of β-actin mRNA. Results are expressed as the mean ± SD from two separate experiments. The primers used for quantitative real-time PCR were as follows: E1b, 5'- GAGCCTGCGAGCCGAGAC-3' (forward)/5'-CGTGGATCTCCTCATCTGACGTTT-3' (reverse); E1, 5’- CAGAGGGTGAGAACGTGGAG-3’(forward)/5’- CGTGGATCTCCTCATCTGACGTTT-3' (reverse); Nrf2, 5'- CAGCGACGGAAAGAGTATGAG-3' (forward)/5'-GGGCTGGCTGAATTGGGAG-3' (reverse); CYP1A1, 5’- TCTTCCTTCGTCCCCTTCAC-3’(forward)/5’-TGGTTGATCTGCCACTGGTT-3’(reverse); HMOX1, 5'- CAGTGCCACCAAGTTCAAGC-3' (forward)/5'-GTTGAGCAGGAACGCAGTCTT-3' (reverse); and β-Actin, 5'-CATGTACGTTGCTATCCAGGC-3' (forward)/5'-CTCCTTAATGTCACGCACGAT-3' (reverse).

Nrf2 gene knockdown by siRNA To reduce endogenous Nrf2 and assess its effect on inducible expression of E1b and E1 by chemicals, C3A cells were transfected with either control or Nrf2 siRNA at 25 nM using the Lipofectamine RNAiMAX reagent with a Forward Transfection Protocol according to the manufacturer’s instructions. Briefly, cells were seeded a day before transfection in 6-well plates at a density of 3×105 cells per well. On the day of transfection, transfection complexes of the Lipofectamine RNAiMAX reagent and the given siRNA were added to each well. After 36 h, siRNA-transfected cells were treated with chemicals for an additional 24 h and then harvested for RT- PCR analysis as described above. The experiments were repeated three times and the most representative results are shown. 77

Western blotting To assess the effect of chemical treatments on mEH protein level, C3A cells in 100 mm dishes were treated with the chemical reagents at various concentrations for 24 h. Cells were washed with PBS, trypsinized and centrifuged at 1000×g for 3 min. For preparation of whole cell lysates, cells were lysed in RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, and 0.1% SDS) supplemented with 1× protease inhibitor cocktail (Cat # 539131, Calbiochem). The cell lysates were centrifuged at 16,000×g for 10 min at 4°C and the supernatants were collected as whole-cell lysate. For assessing nuclear translocation and accumulation of Nrf2, C3A cells were treated with Sulforaphane for 2 h and proteins from nuclear fraction were isolated with NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific,) according to the manufacturer’s instructions. Protein concentrations were determined by Pierce 660 nm Protein Assay (Thermo Scientific). The extracted proteins (30 μg) were separated on a 10% denaturing polyacrylamide gel (Bio-Rad) and transferred to a PVDF membrane (Bio-Rad). After blocking in 5% skim milk for 30 min, the blots were incubated sequentially with primary antibodies at the dilution of 1:1000 and horseradish peroxidase-conjugated secondary antibodies at the dilution of 1:5000. The membranes were washed three times with 1×TBS/0.1% Tween 20, treated with Pierce ECL Western Blotting Substrate (Thermo Scientific), and exposed to ImageTek-H X-ray films (American X-Ray & Medical Supply). The antibodies used for immunoblotting were as follows: anti-mEH (sc- 135984, Santa Cruz Biotechnology), anti-Nrf2 (sc-13032, Santa Cruz Biotechnology), anti-β-Actin (sc-81178, Santa Cruz Biotechnology), anti-TBP (sc-273, Santa Cruz Biotechnology) and anti-GAPDH (G9545, Sigma). The experiments were repeated three times and the most representative results are shown.

RESULTS

Regulation of E1b and E1 by antioxidants and Nrf2 in C3A cells To investigate the role of monofunctional antioxidant inducers in the regulation of human mEH transcribed from two alternative promoters, we treated human hepatoma C3A cells, a model cell line expressing both the mEH E1b and E1 transcripts, with various concentrations of sulforaphane (SFN) for 12 and 24 h. Treatment of C3A cells with SFN up to 10 μM increased both E1b and E1 transcription levels in dose- and time- dependent manners (Figure 4-1A and 4-1B). The magnitude of induction was lower in the 24 h treatment groups than that in the 12 h treatment groups, especially for E1 transcription, possibly due to the metabolism of SFN. Consistent with the RNA data, the protein expression levels of mEH in C3A cells also were increased by SFN 78 treatments (Figure 4-1C). In addition, the treatment of SFN led to the accumulation of Nrf2 in the nucleus (Figure 4-1D). Similarly, a single dose of another antioxidant, tBHQ, induced the transcription of E1b and E1 (Figure 4- 1E and 4-1F). These results implied that Nrf2, a direct target of antioxidants, was involved in induction of human mEH from E1b and E1 promoters. To further investigate the role of Nrf2 in up-regulation of E1b and E1, Nrf2 siRNA was used to silence the expression of Nrf2 in C3A cells (Figure 4-1G). Transfection of Nrf2 siRNA in C3A cells inhibited the Nrf2 expression level. Knockdown of Nrf2 blocked tBHQ-mediated induction of E1b mRNA as well as HMOX1, a known Nrf2 target gene. Interestingly, knocking down Nrf2 in C3A cells led to an increase in basal E1 transcript levels, but abrogated further induction of E1 by tBHQ. Together, these results suggested that Nrf2 mediated E1b and E1 transcription induction in response to the antioxidant tBHQ.

Identification of antioxidant-responsive regions in E1b and E1 promoters We next investigated E1b and E1 promoter regions for antioxidant response elements involved in induction of these transcripts. Two regulatory elements located in the intronic region between E1b and E1 were identified by genome-wide mapping of DNase I hypersensitive (HS) sites (ENCODE Project Consortium). These two elements were demonstrated as functional in studies conducted with human lung cancer cell lines (Su and Omiecinski, unpublished data). To determine the ability of these regions to confer transcriptional activation in response to antioxidants, C3A cells were transiently transfected with reporter constructs containing enhancer segments and DNA fragments that span the proximal E1b or E1 promoter. In C3A cells, both enhancer elements increased the basal E1b and E1 promoter activities, indicating they are functional enhancers (Figure 4-2A). SFN treatment of C3A cells transiently transfected with the enhancer element-containing reporter constructs showed a significant induction of reporter activity in the presence of the HS-2 element, but not the HS-1 element (Figure 4- 2B). Furthermore, overexpression of Nrf2 led to activation of both E1b and E1 proximal promoters by the HS-1 and HS-2 elements (Figure 4-2C).

Regulation of E1b and E1 transcripts by AhR agonists As previous studies indicated that β-NF induces human mEH mRNA and protein expression in primary hepatocyte cultures (Hassett et al., 1998), we next investigated the ability of AhR agonists to affect the expression of two human mEH isoforms, i.e., E1b and E1 transcripts. Therefore, C3A cells were treated with various AhR agonists, including 3-MC, β-NF and TCDD (bifunctional XME inducers) and the expression profiles of the E1b and E1 transcripts were analyzed with quantitative real-time PCR. Treatment of C3A cells with 3-MC for 24 h significantly stimulated E1b mRNA expression approximately 3-fold compared with the solvent control (Figure 4-3A). However, the effect of 3-MC on E1b transcription was less pronounced at 12 h. Interestingly, 3-MC exerted a negative influence on E1 expression, that was diminished by ~40-60% at 12 h and 24 h (Figure 4-3A). The net effect of 3-MC on mEH protein level was negative, as shown with western blotting analysis of the whole cell lysates (Figure 4-3B). β-NF induced 79 transcription of both E1b and E1 at both 12 h and 24 h time points (Figure 4-3C). Induction of E1b by β-NF at 24 h (~ 3.5-fold) was more robust than that of E1 (~ 2.5-fold). In contrast to 3-MC and β-NF, TCDD treatment had no effect on the transcription of E1b (Figure 4-3D). With regard to E1 transcript, TCDD treatment decreased E1 expression at 12 h and 24 h by ~30-40% (Figure 4-3D). These results demonstrate that the induction of the mEH E1b and E1 transcripts by different AhR agonists varies considerably.

Involvement of AhR and Nrf2 in regulation of E1b and E1 expression In the next series of experiments, we investigated potential crosstalk between AhR and Nrf2 pathways in the regulation of E1b and E1 expression. First, to determine if CYP450 induction was accompanied by an increase in Nrf2 expression, we examined the effect of mono- and bi-functional inducers on CYP1A1 and Nrf2 expression. Clearly, 3-MC, β-NF and TCDD stimulated the expression of CYP1A1 significantly in C3A cells (Figure 4-4A). By contrast, the monofunctional inducer tBHQ was able to elicit only a mild induction of CYP1A1. All chemicals except tBHQ appeared to induce a very modest increase in Nrf2 expression levels, however, this increase was only significant for TCDD. (Figure 4-4A). Next, we investigated the involvement of AhR in regulation of E1b and E1 in C3A cells using TMF, an AhR competitive antagonist (Murray et al., 2010), to suppress AhR activity. Treatment with TMF abolished the induction of E1b expression and suppression of E1 expression by 3-MC (Figure 4-4B). These results indicate that AhR plays a role in 3-MC mediated regulation of E1b and E1 expression. We next investigated the involvement of Nrf2 in the regulation of E1b and E1 by 3-MC and β-NF using C3A cells transiently transfected with Nrf2 siRNA. The real-time PCR analysis demonstrated that Nrf2 siRNA decreased the basal expression of Nrf2, as well as NQO1, a known Nrf2 target gene (Figure 4-4C and 4-4D). In addition, siRNA-mediated reduction in Nrf2 resulted in decreased NQO1 induction by 3-MC or β-NF (Figure 4- 4C and 4D). However, knocking down Nrf2 had no effect on 3-MC inducible expression of E1b, suggesting that Nrf2 is not required for 3-MC-mediated induction of the E1b transcript (Figure 4-4C). In contrast, knocking down Nrf2 significantly decreased the mRNA levels of E1b and E1 induced by β-NF (Figure 4-4D), suggesting that Nrf2 is required for β-NF-mediated E1b and E1 transcript induction.

Regulation of E1b and E1 promoter activities by AhR agonists To further investigate the mechanistic basis of E1b and E1 regulation by AhR agonists, we tested whether the effects of AhR agonists on mEH gene expression were dependent on the 5’-flanking regions of E1b and E1. In the C3A cell-based luciferase gene reporter assay, transient transfections with reporter constructs containing truncated DNA fragments of E1b and E1 promoter in various sizes were performed, followed by 3-MC treatments (Figure 4-5A). Because β-NF has been shown to inhibit firefly, but not Renilla luciferase enzyme activity (Wang, 2002), it was excluded from the luciferase reporter assays. Treatment with 3-MC suppressed E1 promoter-driven luciferase activity, a result that was consistent with its effect on E1 expression. However, 3-MC inhibited E1b 80 promoter activity, which was inconsistent with its effect on E1b expression (shown with quantitative real-time PCR). This result indicated that other regulatory elements are likely involved in its regulation on E1b expression. To assess whether the effect of 3-MC on E1b and E1 promoter activity was mediated by AhR, C3A cells were exposed to TMF in the presence and absence of 3-MC. TMF treatment reversed the 3-MC-mediated repression of E1 promoter activity, but was incapable of diminishing the repression on E1b promoter activity exerted by 3-MC (Figure 4-5B). We also tested whether 3-MC activated the intronic enhancer elements which have been demonstrated to encompass AREs and respond to Nrf2 activators. Results from luciferase reporter assays showed that 3-MC induced a moderate increase on E1b or E1 promoter activity driven by the HS-2 enhancer element, but not HS-1 (Figure 4-5C).

Induction of mEH expression by xenobiotics in other cell lines In the final series of experiments, we assessed the variability among the inducible expression profiles of E1b,E1, Nrf2 and CYP1A1 following xenobiotic treatments in human hepatoma HepG2 and Huh7 cell lines, human hepatocytes, and the human lung cancer BEAS-2B cell line. The chemical inducers chosen represent both monofunctional Nrf2 inducers (tBHQ) and bifunctional AhR and Nrf2 inducers (3-MC, β-NF and TCDD). The results obtained with HepG2 cells revealed a remarkable similarity in the gene expression profiles of E1b, E1, Nrf2 and CYP1A1 induced by 3-MC, β-NF, tBHQ and TCDD, as compared to C3A cells (Figure 4-6A). 3-MC treatment induced expression of E1b, but suppressed E1 expression. β-NF and tBHQ activated both E1b and E1 expression, while TCDD treatment had no effect. The magnitude of induction by these chemicals in HepG2 cells was in the same range as observed in C3A cells. (Figure 4-6B). Compared with C3A and HepG2 cells, Huh7 cells exhibited dramatic differences in E1 and E1b expression in response to chemical treatment. No significant changes in E1b transcript levels were observed following any treatment, while expression of the E1 transcript was decreased by tBHQ and increased by TCDD. Nrf2 expression profiles were similar across the liver derived cell lines, but a significant increase in expression was only observed in β-NF-treated Huh7 cells. With regard to CYP1A1, similar significant induction patterns were observed in hepatic-derived cell lines with all chemical treatments, although the magnitude of CYP1A1 induction was lower in Huh7 cells. In human hepatocytes, 3-MC treatment significantly induced expression of E1b, E1, Nrf2 and CYP1A1, while TCDD significantly induced E1, Nrf2 and CYP1A1 (Figure 4-6C). Interestingly, the effect of 3-MC on E1 expression in human hepatocytes were opposite those observed in C3A and HepG2 cells, where E1 expression was suppressed. In addition, Nrf2 was more responsive to TCDD induction in human hepatocytes than in hepatoma cell lines. Similarly, CYP1A1 induction by 3-MC and TCDD was much higher in human hepatocytes than hepatoma cell lines. Finally, inducible E1b expression by 3-MC and TCDD in lung BEAS-2B cancer cells was analyzed since only E1b transcripts are expressed in BEAS-2B cells (Figure 4-6D). Treatment with 3-MC induced a moderate increase in E1b expression while TCDD did not change the expression level of E1b. 3-MC treatment resulted in a 4-fold induction on Nrf2 expression, whereas TCDD treatment increased its expression only moderately. A 81 striking difference between BEAS-2B cells and the liver-derived cells is with respect to the induction of CYP1A1. In BEAS-2B cells, treatments with AhR agonists resulted in 40-60 fold induction of this endpoint, compared with a 600-700 fold induction in human hepatocytes and 200-500 fold induction in C3A or HepG2 cells.

DISCUSSION Alternative promoter (AP) usage is now recognized as a common mechanism whereby multiple mRNA isoforms are transcribed from a single gene locus. It has been estimated that over 50% of human genes have two or more APs (Carninci et al., 2006; Kimura et al., 2006). The use of APs creates diversity and flexibility in the control of gene expression under various contexts (Ayoubi and Van De Ven, 1996; Davuluri et al., 2008; Landry et al., 2003). For example, individual promoters are selectively utilized to regulate the expression of corresponding mRNA isoforms in cell-type, tissue or developmental stage-specific manners. Aberrant activation or silencing of disease-associated genes through AP usage due to genetic and/or epigenetic defects has been linked to cancers and other diseases (Davuluri et al., 2008). The human mEH gene has two APs, E1b and E1, that are located ~15kb apart (Liang et al., 2005). The results of the current study clearly demonstrate that the E1b and E1 promoters are selectively impacted in response to xenobiotic exposures and that both Nrf2 and AhR play a role in their differential expression character. Effects of xenobiotic exposure on human mEH alternative promoter usage are apparently complex, as demonstrated by the data obtained with studies of human hepatoma C3A cells. In these cells, the monofunctional inducers, SFN and tBHQ, activated both E1b and E1 transcript expression with comparable magnitude, and expression of both transcripts was lower after 24 h treatment, compared to 12 h. Interestingly, maximal expression of E1b was observed at 10 µM SFN, while this dose appeared to attenuate the maximal induction of E1 that was observed at 2-5 µM, suggesting that high doses of SFN trigger repressive pathways other than Nrf2 pathway that could specifically act on the E1 promoter and antagonize Nrf2-mediated induction of E1 (Cheung and Kong, 2010; Zhou et al., 2007). Treatment with the bifunctional inducer β-NF increased expression of both the E1b and E1 transcripts, similar to the monofunctional inducers; however, the time course was distinct, with E1b expression higher at 24 h than 12 h, and E1 expression increasing to similar levels at both 12 and 24 h. However, the bifunctional AhR agonists, 3-MC and TCDD, induced divergent transcriptional responses on E1b and E1 expression. 3-MC activated E1b expression, but decreased E1 expression, while TCDD did not change E1b expression, but inhibited E1 expression. The protein levels of mEH resulting from the net effects of chemical inducers on E1b and E1 transcription varied significantly among these chemicals. SFN treatment increased mEH protein levels, consistent with the transcriptional regulation results. The overall mEH protein level was decreased by treatment with 3-MC. Since β-NF increased expression of both E1b and E1 transcripts, and mEH RNA levels tend to correlate closely with protein levels (Fig 1, Fig 2, Hassett et al, 2008) we expect that β-NF correspondingly would upregulate mEH protein expression, whereas TCDD would down-regulate mEH protein expression. These specific measures will require further study, but overall it appears clear that the xenobiotic 82 exposures do have the capacity to alter human mEH gene expression profiles and that these effects are dependent on the nature of the chemicals. Mechanistic studies suggested that the transcription factor Nrf2 was involved in regulation of mEH promoter usage in response to chemical exposure in C3A cells. For the monofunctional inducers SFN and tBHQ and bifunctional inducer β-NF, as demonstrated by Nrf2 siRNA knockdown experiments, Nrf2 appeared to play a critical role in upregulating both the E1b and E1 transcripts. In addition, SFN induced nuclear translocation of Nrf2 and activated luciferase activity of reporter constructs containing both mEH promoters and intronic DNase I HS sites that are also activated by overexpression of Nrf2. These data suggest that Nrf2 enhances both E1b and E1 promoter activity through the same shared intronic enhancer element (DNase HS-2). How Nrf2 and the DNase HS-2 enhancer interact with E1b and E1 promoters and whether the interactions involve promoter-specific regulatory mechanisms, such as diverse core-promoter structure and cis-regulatory elements require further study. Studies with the bifunctional AhR agonists 3-MC and TCDD suggested AhR involvement in regulating expression of E1 and E1b. However, since these compounds elicited divergent effects, the mechanisms of regulation appear complex. From the results, it is likely that AhR involvement in regulating mEH transcription is predominantly an indirect effect. In this respect, recent reports using ChIP-seq or ChIP-chip to study the mapping of genomic regions bound by agonist-activated AhR failed to identify any AhR binding sites within 50kb upstream or 30kb downstream of the human mEH gene (Ahmed et al., 2009; Lo and Matthews, 2012; Pansoy et al., 2010). Experiments with 3-MC clearly revealed that two distinct regulatory mechanisms are applied to E1b and E1 promoters. Treatment with 3-MC induced E1b expression, but decreased E1 levels. The involvement of an AhR signaling pathway contributing to the 3-MC-induced differential regulation of E1b and E1 was confirmed using the AhR antagonist, TMF, which reversed the effects of 3-MC on both E1b and E1 expression. In addition, both 3-MC and TCDD inhibited E1 transcription in C3A cells, indicating that this effect was AhR-mediated. Despite the fact that 3-MC is a bifunctional inducer, it does not appear to act through Nrf2 in regulating the E1b and E1 transcript levels. In these respects, our studies with SFN and tBHQ suggest that Nrf2 activation induces both E1b and E1 expression. Further, knockdown of Nrf2 did not interfere with induction of the E1b transcript by 3-MC, lending additional evidence excluding the involvement of the Nrf2 signaling pathway in this mechanistic process. Given that TCDD and 3-MC are two prototypical AhR agonists (Denison and Nagy, 2003), it is surprising that their effects on E1b transcript expression differed. We hypothesize that this may be due to differences in metabolism of TCDD and 3-MC. While TCDD is resistant to biotransformation, 3-MC is metabolized much more rapidly (Riddick et al., 1994). These differences in metabolism have been implicated in differential cellular responses. For example, TCDD and 3-MC exert distinct effects on estrogen receptor (ER) signaling (Swedenborg et al., 2008). Activation of AhR with TCDD is associated with a repression of the ER signaling. However, 3-MC exposure leads to either activation or repression of the ER signaling dependent on the cellular context. In cells with low metabolic capabilities, 3-MC inhibited ER activity through AhR. The authors 83 proposed that 3-MC-derived metabolites, but not parental 3-MC, activated ERα. In our studies, exposure of C3A cells to 3-MC increased E1b expression, whereas TCDD treatment did not alter E1b’s expression, suggesting that these differential outcome results from inherent differences between TCDD and 3-MC. We did confirm that both of these compounds effectively induced CYP1A1 mRNA levels both in the hepatoma cell models and in primary human hepatocytes. In addition, the activation of E1b expression by 3-MC required an exposure of more than 12 h, suggesting that metabolism of 3-MC may be required for its regulation of E1b expression, and that activated AhR is not directly involved in E1b regulation. It is also important to note that the ability of 3-MC to activate E1b expression was greatly diminished by TMF treatment, suggesting that the activation of AhR, and perhaps induction of AhR target genes, are necessary for 3-MC-induced E1b expression. Taken together, the results suggest that biotransformation of 3-MC is a prerequisite for E1b induction and that the ability of a cell to metabolize an AhR agonist is a determinant of the outcomes to exposure. In addition to potential metabolic differences that affect AhR-mediated E1b regulation, ligand dependent differences in promoter binding may also be involved in the differential effects observed for 3-MC and TCDD. Recent studies using ChIP-chip techniques have indicated that 3-MC and TCDD have discrete promoter binding activities in T-47D human breast cancer cells (Pansoy et al., 2010). Further studies are required to explore these possibilities. Due to the seemingly complex effects of 3-MC on E1b and E1 promoter usage, transient transfection assays using constructs containing varying lengths of the E1b or E1-proximal promoter to drive a luciferase reporter were conducted. These studies showed that treatment with 3-MC decreased the activities of both the E1b and E1 promoters. TMF treatment did not affect the suppression of luciferase activity driven by the E1b promoter, indicating AhR was not involved in this process. It appears therefore that the suppression of luciferase activities contributed by 3-MC result from other signaling pathways that are activated by 3-MC or a metabolites. In these respects, 3-MC has been shown to act as a direct activator of ERα without the involvement of AhR (Abdelrahim et al., 2006). In addition, since the reporter assay result with E1b proximal promoter was not consistent with that of endogenous E1b transcript expression, it is possible that AhR regulates E1b induction through remote cis- regulatory regions. For E1 proximal promoter, the result of the reporter assay was consistent with that of endogenous E1 induction in C3A cells. Further, TMF treatment attenuated the 3-MC mediated suppression of luciferase activity driven by E1 promoter. These results support the idea that 3-MC inhibition of E1 expression involves AhR regulation of the E1 promoter. However, since no AhR binding sites have been identified in mEH upstream regions (Ahmed et al., 2009; Lo and Matthews, 2012; Pansoy et al., 2010), it is possible that AhR modulates E1 expression by interacting with other regulatory proteins by protein-protein interaction. Recent studies have shown that AhR interacts with various transcription factors and modulate their activities. For example, AhR blocks E2F1-dependent gene expression by forming AhR-E2F1 protein complexes (Marlowe et al., 2008). The interaction between AhR and E2F1 was indicated by ChIP analysis to be dependent on the presence of E2F1 binding sites but not AhR binding sites. It is possible that a similar mechanism exists involving the 3-MC mediated repression of E1 expression. Further experiments are necessary to address this idea. 84

E1 and E1b transcription profiles exhibited cell-line dependent and chemical dependent differences. The bifunctional AhR inducers all significantly increased CYP1A1 expression in HepG2 and Huh7 hepatoma cell lines, the BEAS-2B human bronchial epithelial cell line, and in human primary hepatocytes, with primary hepatocytes exhibiting the most responsiveness. In general, tBHQ and β-NF, which appear to act through Nrf2, exhibited similar effects on both E1b and E1 in all of the human hepatoma cell lines. In the cells tested, 3-MC and TCDD also exhibited similar effects on E1b transcript levels. However, the effect of 3-MC and TCDD on E1b and E1 transcript levels was variable across all cell lines tested. In normal human hepatocytes, 3-MC and TCDD failed to inhibit E1 expression. In fact, these compounds enhanced activity ofE1 expression. Again, we speculate that differences in metabolic capacity may account for cell type-specific differences on E1 expression in the response to 3-MC and TCDD exposure. The expression of xenobiotic metabolizing enzymes is lower in hepatoma cell lines than in normal human hepatocytes (Wilkening et al., 2003). The biotransformation of 3-MC in human hepatocytes may efficiently decrease its concentration and prevent persistent activation of AhR. A much greater activation on Nrf2 expression was observed in primary human hepatocytes than in other cells. The elevated expression of Nrf2 in response to 3-MC and TCDD exposure may also contribute to the E1 induction by activating the distal DNase HS2 intronic enhancer element. Taken together, these results suggest that the ability of 3-MC and TCDD to activate or repress E1 transcription is dependent on the cellular context. Studies on the AhR agonist-mediated modulation of human mEH in this investigation clearly established a diverse response dependent on chemical types and cellular backgrounds. These analysis should ultimately assist in predictive understanding of tissue-specific toxicities resulting from xenobiotic exposure. For example, induction of mEH and CYP450 enzymes in hepatocytes by 3-MC could result in an enhancement of the bioactivation of xenobiotics and the production of DNA adducts. However, exposing C3A cells to 3-MC would be predicted to result in fewer adducts but generally exert more direct toxicity due to less inherent metabolic capacity of C3A cells compared with normal hepatocytes. In summary, these investigations have compared the effects of several monofunctional and bifunctional inducers on differential usage of E1b and E1 promoters. The results demonstrate that crosstalk between Nrf2 and AhR signaling pathways contributes importantly to the differential responses to these compounds. The findings also show that bifunctional AhR agonists display distinct abilities to activate or repress the expression of the mEH E1b and E1 transcripts in a cell type-dependent manner. These results highlight the critical importance of taking into account the characteristic of individual cell type in predicting outcomes of exposure to a given xenobiotic compound.

A. B. 6 4 12hr * 12hr 24hr * 3 24hr * * 4 * * * * * 2 2 * (Fold Change) (Fold Change) * * 1 * * E1 mRNA Expression E1b mRNA Expression mRNA E1b * * 0 0 00.512510 00.51 2 510 Sulforaphane(M) Sulforaphane(M)

C. D.

E. F. 3 3 12hr 12hr * 24hr * 24hr *

2 2

1 1 (Fold Change) (Fold Change) E1 mRNA Expression E1b mRNA Expression E1b

0 0 DMSO 40M tBHQ DMSO 40M tBHQ DMSO 40M tBHQ DMSO 40M tBHQ

G. 3 E1b * E1 * HMOX1 2 * Nrf2

1 (Fold Change) Relative mRNAExpression Relative 0 DMSO 40M tBHQ DMSO 40M tBHQ Control siRNA Nrf2 siRNA Figure 4-1. Regulation of E1b and E1 by antioxidants and Nrf2 in C3A cells. E1b (A) and E1 (B) mRNA expression in C3A cells treated with DMSO or SFN for different time periods. β-actin was used as an internal control. C.) mEH protein expression in C3A cells treated with DMSO or SFN for 24 h. β-actin was used as a loading control. D.) Nuclear translocation of Nrf2 in C3A cells following 2 h treatment with DMSO or 5μM SFN. TATA-binding protein (TBP) was used as a loading control. E1b (E) and E1 (F) mRNA expression in C3A cells 86 treated with DMSO or 40μM tBHQ for different time periods. β-actin was used as an internal control. G.) Effects of Nrf2 siRNA knockdown on E1b, E1 and HMOX1 mRNA expressions. C3A cells were transfected with either control or Nrf2 siRNA for 36 h and then treated with either DMSO or tBHQ for an additional 24 h as described under “Materials and Methods”. The mRNA levels of E1b, E1, HMOX1 and Nrf2 were quantified by real time PCR. β-actin was used as an internal control. Statistically significant differences between treatment groups compared to the DMSO-treated control are indicated by an “*” (p<0.05).

A. B. HS-2+E1-208 * HS-1+E1-208 * E1-208 HS-2+E1b-300 * HS-1+E1b-300 * E1b-300

01234

Figure 4-2. Involvement of intronic regulatory regions in antioxidant and Nrf2-mediated E1b and E1 inductions. A.) Luciferase activity of the E1b or E1 proximal promoter/DNase HS-1 or HS-2-driven reporter vectors. B.) Effect of SFN treatments on luciferase activity of the E1b or E1 proximal promoter/DNase HS-1 or HS-2-driven reporter vectors. C.) Effects of Nrf2 overexpression on luciferase activity of the E1b or E1 proximal promoter/DNase HS-1 or HS-2-driven reporter vectors. C3A cells were co-transfected with 100 ng Nrf2 expressing plasmid and E1b or E1 promoter reporter/DNase HS-1 or HS-2 constructs. After 24 h, luciferase activity was measured. Statistically significant differences between treatment groups compared to the empty vector control or the DMSO-treated control are indicated by an “*” (p<0.05).

A. B. 3-MC 4 12hr * 24hr 3 E1b E1

(Fold Change) 1 * * Relative mRNA Expression mRNA Relative 0 DMSO 3-MC DMSO 3-MC

C. D. -NF 4 12hr 1.5 TCDD 12hr * 24hr E1b E1 24hr 3 E1b E1 * * * 1.0 2 * * 0.5 (Fold Change) 1 (Fold Change) Relative mRNA Expression Relative Expression mRNA 0 0.0 DMSO -NF DMSO -NF DMSO TCDD DMSO TCDD

Figure 4-3. Regulation of E1b and E1 transcripts by AhR agonists. E1b and E1 mRNA expression in C3A cells treated with 2.5µM 3-MC (A), 20µM β-NF (C) or 10nM TCDD (D) for different time periods was analyzed with real time PCR. β-actin was used as an internal control. B.) Suppression of mEH protein expression in C3A cells treated with 3-MC for 24 h. GAPDH was used as a loading control. Statistically significant differences between treatment groups compared to the DMSO-treated control are indicated by an “*” (p<0.05).

A. B.

C. D.

Figure 4-4. Involvement of AhR and Nrf2 in regulation of E1b and E1 expression by AhR agonists. A.) CYP1A and Nrf2 mRNA expression in C3A cells treated with DMSO or 2.5µM 3-MC, 20µM β-NF, 40μM tBHQ or 10nM TCDD for 24 h. β-actin was used as an internal control. B.) Antagonization of TMF on 3-MC-mediated E1b and E1 mRNA expressions. C3A cells were pre-treated with DMSO or 10μM TMF for 1 h and then treated with DMSO or 2.5µM 3-MC for 24 h. Effects of Nrf2 siRNA knockdown on E1b, E1 and NQO1 mRNA expressions in C3A cells treated with 2.5µM 3-MC (C) or 20µM β-NF (D). C3A cells were transfected with either control or Nrf2 siRNA for 36 h and then treated with either DMSO, 3-MC or β-NF for an additional 24 h as described under “Materials and Methods”. The mRNA levels of E1b, E1, NQO1 and Nrf2 were quantified by real time PCR. β-actin was used as an internal control. Statistically significant differences between treatment groups compared to the DMSO-treated control are indicated by an “*” (p<0.05). For Nrf2 siRNA knockdown experiments, statistically significant differences between the control siRNA and Nrf2 siRNA groups are indicated by an “*” (p<0.05).

A. B.

Figure 4-5. Regulation of E1b and E1 promoter activities by AhR agonists. A.) Effect of 3-MC treatment on luciferase activity of the E1b or E1 proximal promoter-driven reporter vectors. Cells were transfected with E1b or E1 proximal promoter luciferase reporter vectors as described under “Material and Methods”. Six hours after transfection, cells were treated with 2.5µM 3-MC for 24 h. B.) Antagonization of TMF on 3-MC-mediated luciferase activity of the E1b or E1 proximal promoter-driven reporter vectors. Cells were transfected with E1b- 300 or E1-208 promoter luciferase reporter vectors. Five hours after transfection, cells were treated with 10µM TMF. One hour after TMF treatment, cells were treated with DMSO or 2.5µM 3-MC for 24 h. C.) Effect of 3-MC treatment on luciferase activity of the E1b-300 or E1-208 promoter/HS-2-driven reporter vectors. Statistically significant differences between treatment groups compared to the DMSO-treated control are indicated by an “*” (p<0.05). For the TMF antagonization experiment, statistically significant differences between DMSO-treated and TMF-treated groups are indicated by an “*” (p<0.05).

A. HepG2 B. Huh7

C. Human hepatocyte D. BEAS-2B

Figure 4-6. Induction of mEH expression by xenobiotics in other cell lines. Human hepatoma HepG2 (A) and Huh7 (B) cells were treated with DMSO, 2.5µM 3-MC, 20µM β-NF, 40μM tBHQ or 10nM TCDD for 24 h. Normal human hepatocytes (C) and human bronchial epithelial BEAS-2B cells (D) were treated with DMSO, 2.5µM 3-MC or 10nM TCDD for 24 h. The expression levels of E1b, E1, Nrf2 and CYP1A1 mRNA were analyzed with real time PCR. β-actin was used as an internal control. Statistically significant differences between treatment groups compared to the DMSO-treated control are indicated by an “*” (p<0.05).

Chapter 5 Conclusions

The general aim of this thesis was to characterize molecular regulatory mechanisms of human mEH gene expression. The transcription of human mEH gene initiates from two unique genomic locations using two alternative promoters. The E1b promoter drives the expression of E1b transcript ubiquitously in all tissues, whereas the E1 promoter drives the expression of E1 restrictedly in liver. Firstly, the molecular basis of constitutive expression of E1b isoform was investigated. We demonstrated that the first 300 bp upstream promoter region of E1b confers the constitutive expression of E1b variant. Inspection of the DNA sequence of this E1b promoter region identified a putative CpG island. The discovery of the CpG island in the E1b promoter could at least partially explain the ubiquitous expression of E1b variant because CpG islands are found in the promoters of all housekeeping genes. Promoter analysis using luciferase reporter and site-directed mutagenesis, as well as siRNA knockdown, EMSA and ChIP assays showed that Sp1 and Sp3 participate in regulating E1b promoter activity and maintaining constitutive and basal expression of E1b variant. Identification of Sp1 and Sp3 as modulators of E1b expression through E1b promoter provides a plausible mechanistic basis for tissue-specific expression magnitude of E1b variant because these two transcription factors have been shown to regulate target genes through protein-protein interaction with tissue-specific transcription factors. Secondly, the molecular mechanism of antioxidant-mediated inducible expression of E1b isoform in human lung cancer cells was investigated. We demonstrated that siRNA knockdown of Nrf2 results in a decrease in E1b expression in A549 cells, which have constitutive Nrf2 expression, and blocks antioxidant-mediated induction of E1b transcript in BEAS-2B cells which have a constitutively low Nrf2 expression. Overexpression of Nrf2 significantly increased E1b mRNA and protein levels in BEAS-2B cells. Promoter analysis using luciferase reporter established that activation of the reporter gene by antioxidant or Nrf2 requires not only the E1b proximal promoter but also an intronic enhancer element which locates in the genomic region between E1b and E1. ChIP assay showed that Nrf2 binds to the enhancer element after being activated by antioxidant treatment. A conserved antioxidant response element (ARE) was identified in the enhancer element with EMSA. The ARE is essential for the enhancer activity and Nrf2-mediated E1b induction because mutagenesis of the ARE abolished the enhancer activity and blocked antioxidant-induced transactivation of E1b promoter. These data suggest that antioxidants stimulate E1b expression by activating Nrf2 and prompting interaction between Nrf2 and the ARE in the enhancer element. Considering the great distance between the intronic enhancer and E1b promoter, we speculate an introchromosomal interaction between the two regions via a chromatin looping which brings Nrf2 binding to the enhancer into close proximity with the general transcription machinery at the transcription start site. Since Sp1 was shown to constitutively bind to the E1b proximal promoter as mentioned above and has been demonstrated to 92 regulate chromatin looping in other studies, it is possible that Sp1 participates in the formation of the chromatin looping between E1b promoter and the distal enhancer by interacting with Nrf2. Finally, the differential regulation of E1b and E1 transcriptions through alternative promoter usage by antioxidants and AhR agonists was explored. The E1b and E1 variants exhibit distinct tissue-dependent expression patterns. The liver is unique in that it expresses both E1b and E1 isoforms. Therefore, human hepatoma C3A cells were chosen to study the differential regulation of human mEH alternative promoters by various chemicals. Antioxidants, SFN and tBHQ, also known as monofunctional inducers, enhanced the expression of both E1b and E1 variants. Mechanistic studies with promoter-reporter assays and Nrf2 siRNA knockdown suggested that they activate E1b and E1 transcription through Nrf2 signaling pathway and that the intronic HS-2 enhancer element is involved for both promoters. In contrast, AhR agonists, 3-MC, β-NF and TCDD, also known as bifunctional inducers, showed considerable differences in the regulation of E1b and E1 expressions. Treatment with β-NF, like monofunctional inducers, upregulated both of E1b and E1 expression after as early as 12 h exposure, which results from activation of Nrf2 signaling pathway, as shown with real-time PCR and Nrf2 siRNA knockdown. Treatment with 3-MC enhanced E1b expression, but decreased E1 expression, indicating that two distinct regulatory mechanisms happened to E1b and E1 promoters. Further studies using Nrf2 siRNA knockdown and AhR antagonist, TMF, demonstrated that AhR signaling pathway, but not Nrf2, was involved in 3-MC-mediated E1b induction. A long exposure time (24 h) was required for 3-MC-mediated E1b induction, which suggested that metabolites of 3-MC may activate transcription factors that can upregulate E1b expression. Both 3-MC and TCDD inhibited E1 expression. Mechanistic studies revealed that the AhR signaling pathway was involved because TMF diminished the repression of E1 expression and promoter activity by 3-MC treatment. Examination of E1 expression profiles in other hepatoma cell lines and normal human hepatocytes showed that the ability of 3-MC and TCDD to activate or repress E1 expression is dependent on the cellular context, such as metabolic capacity. Studies on antioxidant-mediated human mEH induction suggest a new mechanism by which an intronic enhancer regulates mEH expression. Most of the previous mechanistic studies on human mEH transcription have been focused on its proximal promoters. This thesis research revealed that xenobiotics may alter mEH expression via not only the proximal promoter but also distal regulatory elements. These results have moved beyond the simplistic mechanism of proximal promoter-regulated transcription to include distal regulatory elements that are critical for proper transcriptional control of human mEH gene in response to xenobiotic exposure. Since enhancer elements can be located far from their target genes, we expect that, besides the two intronic enhancer element identified in this study, there may be more enhancer elements which contribute to the human mEH gene regulation. These distal regulatory elements may be involved in not only xenobiotic-mediated mEH induction but also the tissue-dependent constitutive expression of mEH by differentially influencing the activities of alternative mEH promoters. In addition, they can affect the expression magnitude of mEH variants by recruiting tissue- 93 specific transcription factors. Therefore, these distal regulatory elements are likely to constitute a primary basis for basal and inducible expression of human mEH gene. Studies on the AhR agonist-mediated modulation of human mEH in this thesis research revealed a diverse response dependent on chemical types and cellular backgrounds. For example, 3-MC differentially regulated E1b and E1 transcription in C3A cells, but eventually downregulated mEH protein expression. In normal human hepatocytes, it enhanced both E1b and E1 mRNA levels and a higher mEH protein expression would be expected. With higher elevated levels of CYP1A1 in human hepatocytes than C3A cells, we expect that more DNA adducts would be produced from PAH bioactivation resulted from the enzyme activities of mEH and CYP450s in human hepatocytes than C3A cells. However, C3A cells and other cells with less metabolic capacity may be subject to the direct adverse-effects of xenobiotics. Thus, the consequence of xenobiotic exposure regarding mEH modulation is subject to further analysis in in vivo animal model and in clinical studies. These analysis will ultimately help us predict tissue-specific toxicities of xenobiotic exposure and provide useful information to minimize the risk of xenobiotic-mediated toxic effects. In summary, we have defined multiple mechanisms regulating expression of E1b variant as well as E1 variant. The basal expression of E1b variant is maintained by transcription factors Sp1 and Sp3 through interacting with E1b proximal promoter. Antioxidant-induced E1b expression is regulated by Nrf2 through an intronic enhancer element. Differential expression of E1b and E1 variants in response to xenobiotic exposure is mediated by AhR and Nrf2 pathways. The mechanistic insights obtained extend our understanding of the regulation of human mEH gene expression by endogenous and xenobiotic inducers and are of value for further elucidation of normal and pathological conditions involving mEH expression.

Chapter 6 Future perspectives

Xenobiotic exposure can be an important contributor to a variety of diseases, including cancer. The human body has developed an intricate system of xenobiotic metabolizing enzymes to counter the deleterious consequences of chemical exposure. The contribution of human mEH to xenobiotic metabolism can have two outcomes. Often, mEH acts as a detoxifying enzyme, hydrolyzing reactive epoxides derived from cytochrome P450 enzyme-mediated xenobiotic metabolism to less reactive dihydrodiols. In contrast, in the metabolism of polycyclic aromatic hydrocarbons (PAH), mEH plays a critical role in formation of ultimately carcinogenic of bay-region or fjord-region dihydrodiol epoxides which are highly genotoxic. However, the consequence of mEH enzymatic activity is not always protective. For PAHs and similar compounds that require metabolic activation for their toxicities, inhibition of mEH is likely to reduce xenobiotic toxic effects and induction of mEH is expected to increased toxicity. Clinically, mEH overexpression in non-small cell lung cancer (NSCLC) patients is inversely correlated with patient survival (Lin et al., 2007). Therefore, detailed molecular studies on the mEH gene regulation in humans will be beneficial for risk predictions and potentially of value in preventing xenobiotic- mediated toxic effects in normal and pathological conditions. While the studies included in this dissertation research have advanced our understanding of human mEH gene regulation, there are several unresolved issues for future investigation.

1. TISSUE-DEPENDENT HUMAN mEH EXPRESSION Human mEH has a broad tissue distribution although its expression level varies in a tissue-dependent manner. These tissue specific expression patterns are determined from the use of alternative gene promoters. The studies presented in this thesis utilized different human cell lines from different tissue origins and show that the constitutive expression of human mEH as driven by the principle E1b promoter is regulated, at least partially, by a common mechanism. The transcription factors Sp1 and Sp3 are involved in maintaining the basal E1b promoter activity and constitutive expression of the E1b variant. Sp1 and Sp3 are well known to interact with other tissue- specific transcription factors to direct the tissue-dependent transcription of adjacent genes. We identified several candidate binding sites for other transcription factors. How these factors integrate with Sp1/Sp3 to regulate E1b promoter activity requires further investigation. Future studies to test the ability of these factors to regulate Sp1/Sp3-mediated E1b promoter activity using luciferase reporter assays will aid in understanding how and when E1b promoter activity is altered in reference to cell background. However, during the study of regulation of E1b expression we found that endogenous expression of Sp1 and Sp3 are very high in BEAS-2B and C3A cells. We were not able to observe any change on endogenous E1b expression after expression vectors encoding Sp1 and Sp3 or siRNA specific to Sp1were transfected into these cells. Therefore, a cell line which lacks the Sp family of transcription factors is required for further addressing 95 transcriptional functions of Sp1/Sp3 and other Sp/KLF family transcription factors in the regulation of E1b transcription driven from this far upstream promoter. Schneider’s Drosophila cell line 2 (SL2), derived from Drosophila embryos, are devoid of many ubiquitous transcription factors including Sp1/Sp3 and has been used to study transcriptional properties of Sp1 and Sp3 without interference by endogenous factors (Jin et al., 2011; Suske, 2000; Wan et al., 2005). Thus, SL2 cells are potentially suited for the task of studying Sp1 and Sp3- mediated transcriptional regulation of E1b, as well as inherent activities of cell type or tissue-specific transcription factors in regulating E1b promoter activity.

2. DISTAL REGULATORY ELEMENT-MEDIATED E1B EXPRESSION Distal regulatory elements are involved in the regulation of human mEH promoter activity and gene expression. As previously published by our laboratory, a transposon element located approximately 2 kb 5’- upstream of E1b downregulates E1b proximal promoter activity. In the current study, we demonstrated that antioxidants, such as sulforaphane and tBHQ, induce both E1b and E1 expression through an intronic enhancer that contains an antioxidant response element and binds to activated Nrf2. Because these elements are distant from either the E1b or E1 promoters, we speculate that chromatin looping is involved in their transcriptional activation effects (Figure 6-1A). How Nrf2 participates in the formation of a chromatin loop and whether other transcriptional factors and chromatin modifying factors are involved in the actions of the distal enhancer to induce E1b or E1 transcription remain to be clarified. In the ENCODE project, whole genome analysis of accessible chromatin in human cells using the DNase- array approach has revealed two open chromatin regions located in the intervening region between E1b and E1 (Sabo et al., 2006). The accessibility of these regions are cell-type dependent. In some cells, both regions are constitutively open while in other cells, only one is open. Reporter analyses showed that these regions contain regulatory DNA elements that are able to transactivate both the E1b and E1 promoters. The regulatory activity of these DNA elements is also dependent on cell background. For example, the first element transactivates E1b promoter in human hepatoma C3A cells, but not in human bronchial epithelial BEAS-2B cells. Moreover, the second element, but not the first element, is involved in Nrf2-mediated induction of human mEH gene expression. It is possible that these regions host various binding motifs for transcription factors that regulate cell-type dependent expression of human mEH gene. These transcriptional factor binding motifs and functional interactions remain to be identified. Our initial studies suggested that CAR and HNF4α may be two of the transcription factors that regulate the transactivation activity of the intronic enhancer elements (data not shown). The approach used in this thesis research to characterize the distal regulatory elements heavily relied upon transient luciferase reporter assays in cultured cell lines. Based on the locations of the two putative intronic enhancers, we speculate that they could regulate the promoter activities of E1b and E1 as they represent the gene promoters in closest genomic proximity (Figure 6-1A). We subcloned the corresponding genomic segments for the given enhancers upstream of a luciferase reporter gene driven by the E1b or E1 proximal and core promoter. 96

The resulting constructs was then transfected into cultured cells and the luciferase activity was measured to determine if the test segment changed the transcription activity of a given promoter. Although this approach was convenient and successful in characterizing the intervening regulatory elements and in identifying the transcription factors that associate with these elements, the tested enhancer elements were removed from their normal genomic context and placed into a plasmid which lacks the native chromatin structure. Thus, luciferase reporter assays may not reveal effects on long-range regulation by enhancers and transcription factors that bind to them. However, these long range interaction can be revealed using techniques such as the chromosome conformation capture (3C) assay. The 3C technology allows analysis of the physical interaction between a promoter and a distal regulatory element in the native cellular state (Figure 6-2) (Dekker, 2006; Dekker et al., 2002; Hagege et al., 2007; Simonis et al., 2007). This method relies on the notion that distal enhancers are brought in close proximity to target promoters through DNA looping (Dekker, 2003). In this assay, interactions between two genomic regions are captured in intact cells directly by formaldehyde crosslinking. The crosslinked chromatin is digested with an appropriate restriction enzyme to create sticky ends. Subsequently, the sticky ends of the digested, crosslinked DNA fragments are religated at a low DNA concentration. Under such conditions, intramolecular ligations between crosslinked DNA fragments are strongly favored over intermolecular ligations between random fragments. In this way, DNA fragments that may be distant in the linear genome, but close in space, are ligated to each other. The crosslinks are then reversed and ligated DNA fragments are purified. The resulting DNA fragments contain a large number of ligation products, each representing a specific interaction between two genomic regions, which is referred to as a 3C library. Conventional or quantitative PCR using primers spanning the ligated fragments is used to detect or measure the amount of specific ligation products. The relative abundance of a particular ligation product indicates the frequency with which the two chromatin segments interact. To conclude a chromatin loop is formed between two segments, it must be demonstrated that they interact more frequently each other than with neighboring segments. In a proposed experimental design, we select endonuclease PmeI to digest crosslinked chromatin (Figure 6-1B). PmeI digests the EPHX1 genomic locus into several segments that will allow study of the interactions between these segments, such as interactions between E1b or E1 proximal promoter region and DNase I HS-1 or HS-2 site. Specific PCR primers are designed to anneal to each segment. If there is any physical interaction between any two segments, a PCR product will be generated. In this way, the status of chromatin looping present in the EPHX1 genomic region is surveyed to detect any change in response to environmental stimuli. Going forward, we expect that the 3C assay will provide useful information on how chromatin looping forms between enhancers and proximal promoters and distinguish factors which contribute to the formation of looping from factors that are not involved in the loop formation but regulate the promoter activities. Although ChIP-seq studies performed by others have shown that AhR does not bind to the proximal promoter region of E1b, our studies demonstrated that AhR ligands modulate E1b expression. Therefore, it 97 appears that the transcriptional regulation of E1b by AhR may involve unidentified distal regulatory elements. We also speculate that there exist additional distal enhancers associated with the regulation of human mEH expression that response to environmental stimuli. While we can identify putative enhancers within the flanking region of human mEH gene based on chromatin signatures, such as nuclease hypersensitivity and histone methylation, it is difficult to determine whether these enhancers associate with human mEH gene. Enhancers can be located a long distance from their target genes and do not always regulate all genes nearby. Therefore, the current strategies, such as promoter-reporter and 3C assays, which depend on prior knowledge of putative enhancers and test only a few proposed interactions between specific loci in a single experiment, are not suitable for genome-wide mapping of regulatory elements interacting with human mEH gene promoters. Recent developments of 3C-based technology coupled with microarray or next generation sequencing will enable examination of the interaction between human mEH gene promoters and enhancers on a genome-wide scale. The method of circular chromosome conformation capture (4C, also known as chromosome conformation capture-on-chip) reveals interactions between one given promoter and all of the other regulatory sequences from the rest of the genome (Figure 6-2) (Simonis et al., 2006; Zhao et al., 2006). In 4C technology, the 3C library is processed with a second round of digestion and religated under conditions that favor the formation of self-ligated circles. The DNA circles containing 3C ligation junctions of the fragment of interest and its interaction partner are amplified by inverse PCR with primers specific to the fragment of interest. The resulting 4C library is analyzed by microarrays or next generation sequencing. Similarly, the chromosome conformation capture carbon copy (5C) technology offers concurrent determination of interactions between multiple selected loci and can generate a comprehensive ‘many-to-many’ interaction network for given genomic regions (Figure 6-2) (Dostie et al., 2006). In 5C technology, the 3C library is copied by ligation-mediated amplification (LMA) using primers flanking the predicted ligation sites to generate a quantitative carbon copy of a part of the initial 3C library, which is subsequently analyzed with microarray or high throughput sequencing. LMA involves using a combination of test and fixed 5C primers that can anneal on the 3C template, facing outward and inward, so that each primer covers exactly a half restriction site. Such a pair of primers can be ligated tail-to-head across the ligated junction of 3C joined fragments by ligase. Therefore, applications of these technologies will aid in the identification of long-range regulatory elements interacting with human mEH promoters and help promote understanding of their roles in regulation of the human mEH gene. These enhancer elements are not only likely to function as a basis for differential mEH gene expression in various tissues, but also will likely contribute functionally in the development of human diseases related to xenobiotic exposure. 98

3. SINGLE NUCLEOTIDE POLYMORPHISMS (SNP) AND THEIR POTENTIAL ROLE FOR REGULATING ENHANCER AND PROMOTER FUNCTIONS The genomic region encompassing the human mEH gene is highly polymorphic. With data retrieved from dbSNP database (Build 137) using the UCSC Genome Browser, we identified 178 validated SNPs within the EPHX1 gene region (Figure 6-3). Among these, there are 67 SNPs in the region between E1b and E1 and 111 SNPs in the region between intron 1 and exon 9. In addition, the regions flanking the EPHX1 gene are highly polymorphic. There are 78 SNPs in the -10kb upstream region of E1b and 57 SNPs in the -20kb to -10kb upstream region of E1b. In the downstream region of EPHX1 gene, there are 50 SNPs in the first 10kb region and 33 SNPs in the 2nd 10kb region. The impact of these SNPs on the transcriptional levels of EPHX1 gene is of toxicological interest and potential importance. SNPs in the promoter regions and distal enhancer elements have the potential to alter gene transcription by several mechanisms. Mutations within these regulatory regions can either eliminate or create a binding motif for a transcription factor. SNP alleles may also affect the binding affinity of a transcription factor to its target DNA motif. In the initial characterization of two SNPs, rs12727007 (C/T) and rs12741681 (A/C), from the proximal promoter region of E1b, we observed differential binding of the transcription factor AP-2α (TFAP2A) on these SNP sites as manifested with EMSA and Supershift assays (Figure 6-4A and 6-4B). AP-2α had a greater affinity for allele “C” than “T” or “A”. In addition, alleles in the rs12403480 locus, located at the 2nd DNase I HS site, exhibited differential influence on its enhancer activity for the E1 proximal promoter, but not on the E1b proximal promoter (Figure 6-5). As shown with the luciferase reporter assay in C3A cells, the enhancer element containing allele ‘T’ in the rs12403480 locus yields a significantly lower transactivation activity on the E1 proximal promoter. No significant difference was observed between alleles in the rs12403480 on the enhancer activity to E1b proximal promoter in BEAS-2B, A549 and C3A cells. These preliminary studies indicate that SNPs in the regulatory regions of EPHX1 may play a role in its gene expression. The development of high-throughput methodologies has allowed genome-wide assay for genetic variation. A high density SNP array is able to detect up to one million SNPs (LaFramboise, 2009). We propose to use custom SNP arrays to genotype all SNPs in the regulatory regions of EPHX1simultaneously. The association between EPHX1 expression and these SNPs in population might pinpoint SNPs affecting EPHX1 expression.

4. IN VIVO MODELS FOR THE STUDY OF REGULATORY MECHANISMS AND FUNCTIONS OF THE HUMAN mEH GENE The transcriptional regulation of human mEH gene is very complex and impacted at multiple levels. There is no doubt that modulation of mEH is the result of combinations of different regulatory mechanisms, making it challenging to correlate individual mechanisms characterized using cell-based methods with downstream transcriptional outcomes. In addition, mEH plays both detoxification and bioactivation roles in xenobiotic metabolism. With cultured cell lines that lack the in vivo environment, it is difficult to predict whether mEH 99 activity results in lower or higher carcinogenicity following specific xenobiotic exposures. Therefore, an animal model is needed for in vivo research on the regulation and function of human mEH. The poor conservation in the genomic regulatory features among human and rodents limits the usage of rodent animal models for study of the regulatory mechanisms of human mEH expression. The human mEH gene is transcribed from two different genomic locations using two unique promoters, while the mouse mEH gene is transcribed from a single location, indicating that the mEH gene is subject to different regulatory mechanisms in human and rodents. To study in vivo regulation of human mEH with a rodent animal model, we propose to establish a transgenic mouse model, perhaps using a bacmid gene replacement method, integrating the human genomic DNA region comprising human mEH gene, as well as its upstream and downstream regulatory regions into the genome of mEH-null mice. In this way, most of the regulatory mechanisms of human mEH should be retained. This future transgenic mouse model would provide a useful platform for the study of in vivo regulation of human mEH as well as the characterization of the inducible properties and toxicities of xenobiotics.

100

A. HS-1 HS-1

HS-2 HS-2

E1b HS-1 HS-2 E1

B. DNase I HS +1 site#1 site#2 +1 P1 P2 P3 P4 P5 P6 P7-9

E1b E1 E2

Figure 6-1. Studies on interaction between intronic enhancer elements and the proximal promoters of human mEH gene. A.) The proposed interaction between DNase I HS sites and E1b or E1 proximal promoter region by chromatin looping. Transcription factors (indicated as red or blue ovals in this schematic) binding to these regions can interact with each other and regulate the transcription of E1b and E1 transcripts. B.) The proposed 3C experimental design. The interaction between DNase I HS sites and promoter regions of E1b and E1 will be assessed. Digestion sites of PmeI that will be used in the 3C assay are depicted as small vertical bars and the letter ‘P’ and numbered. The relative positions of PCR primers (blue arrows) that will be used for detecting physical interaction between two regions are also depicted.

101

Enhancer Promoter

Transcription factors

Formaldehyde crosslinking

Digestion

Intramolecular ligation

Reverse crosslink The 3C library

3C 4C 5C

Secondary digestion Ligation-mediated and ligation amplification

Quantitative PCR

Microarray or sequencing Inverse PCR

Microarray or sequencing

Figure 6-2. Overview of 3C and 3C-derived methods. In 3C, 4C and 5C methods, chromatin interactions are captured by formaldehyde crosslink, followed by digestion and ligation of the cross-linked chromatin. The crosslinks are then reversed and ligation products can be analyzed by quantitative PCR in 3C. In 4C, the ligation products are digested by a secondary restriction enzyme and ligated to form a circle. Inverse PCR is used to identify the chromatin interactions using primers specific to the promoters, followed by microarray analysis or next generation sequencing. In 5C, the ligation products are copied by ligation-mediated amplification using primers flanking the predicted ligation site and are analyzed with microarray analysis or next generation sequencing.

102

Figure 6-3. Distribution of Single nucleotide polymorphisms (SNP) in the genomic regions of human mEH gene (EPHX1). Information of SNPs was retrieved from dbSNP database (Build 137) using UCSC Genome Browser. Numbers above bars indicate the total number of SNPs in each region.

103

Figure 6-4. Electrophoretic mobility shift assay (EMSA) and supershift analyses of AP-2α binding to alleles at the loci of rs12741681(A/C) and rs12727007(C/T). A.) EMSA was performed with nuclear extracts from BEAS-2B cells transfected with empty p3XFLAG-CMV10 expression vector or p3XFLAG-AP-2α expression vector and 32p-labeled oligonucleotide probes containing alleles of rs12741681(A/C) or rs12727007(C/T). A 200- fold molar excess of the respective unlabeled competitor oligonucleotide was used for competition assays. The specific binding was framed and shown with an arrow. B.) Supershift analysis was performed using anti-FLAG antibody and nuclear extracts from BEAS-2B cells transfected with p3XFLAG-AP-2α expression vector. In competition assays, a 200-fold molar excess of the respective unlabeled competitor oligonucleotides were used. The supershifted and specific binding bands were framed and shown with arrows.

104

Figure 6-5. Differential effect of rs12403480 (C/T) polymorphism on the enhancer activity of DNase I HS-2 to E1b and E1 promoter activities. Luciferase reporter constructs containing two rs12403480 alleles were transfected in BEAS-2B, A549 and C3A cells. Luciferase activity was determined 24 h later. Fold change of relative luciferase activity was calculated by defining the activity of constructs containing allele “C” as 1. Statistically significant differences on the luciferase activity compared to constructs containing allele “C” are indicated by an “*” (p<0.05).

105

References

Abdelrahim M, Ariazi E, Kim K, Khan S, Barhoumi R, Burghardt R, Liu S, Hill D, Finnell R, Wlodarczyk B, Jordan VC and Safe S (2006) 3-Methylcholanthrene and other aryl hydrocarbon receptor agonists directly activate estrogen receptor alpha. Cancer Res 66(4): 2459-2467. Ahmed S, Valen E, Sandelin A and Matthews J (2009) Dioxin Increases the Interaction Between Aryl Hydrocarbon Receptor and Estrogen Receptor Alpha at Human Promoters. Toxicological Sciences 111(2): 254-266. Alves C, von Dippe P, Amoui M and Levy D (1993) Bile acid transport into hepatocyte smooth endoplasmic reticulum vesicles is mediated by microsomal epoxide hydrolase, a membrane protein exhibiting two distinct topological orientations. Journal of Biological Chemistry 268(27): 20148-20155. Anish R, Hossain MB, Jacobson RH and Takada S (2009) Characterization of Transcription from TATA-Less Promoters: Identification of a New Core Promoter Element XCPE2 and Analysis of Factor Requirements. PLoS ONE 4(4): e5103. Antonson P and Xanthopoulos KG (1995) Molecular Cloning, Sequence, and Expression Patterns of the Human Gene Encoding CCAAT/Enhancer Binding Protein α (C/EBPα). Biochemical and Biophysical Research Communications 215(1): 106-113. Atchison ML (1988) Enhancers: mechanisms of action and cell specificity. Annual review of cell biology 4: 127- 153. Auerbach SS, Stoner MA, Su S and Omiecinski CJ (2005) Retinoid X receptor-alpha-dependent transactivation by a naturally occurring structural variant of human constitutive androstane receptor (NR1I3). Mol Pharmacol 68(5): 1239-1253. Auten RL, O'Reilly MA, Oury TD, Nozik-Grayck E and Whorton MH (2006) Transgenic extracellular superoxide dismutase protects postnatal alveolar epithelial proliferation and development during hyperoxia. AJP - Lung Cellular and Molecular Physiology 290(1): L32-40. Ayoubi TA and Van De Ven WJ (1996) Regulation of gene expression by alternative promoters. The FASEB Journal 10(4): 453-460. Babior BM (1999) NADPH oxidase: an update. Blood 93(5): 1464-1476. Baird L and Dinkova-Kostova A (2011) The cytoprotective role of the Keap1–Nrf2 pathway. Archives of Toxicology 85(4): 241-272. Barrera LO and Ren B (2006) The transcriptional regulatory code of eukaryotic cells--insights from genome-wide analysis of chromatin organization and transcription factor binding. Curr Opin Cell Biol 18(3): 291-298. Baylin SB, Herman JG, Graff JR, Vertino PM and Issa JP (1998) Alterations in DNA methylation: a fundamental aspect of neoplasia. Advances in cancer research 72: 141-196. Bell AC and Felsenfeld G (2000) Methylation of a CTCF-dependent boundary controls imprinted expression of the Igf2 gene. Nature 405(6785): 482-485. Bird A (2002) DNA methylation patterns and epigenetic memory. Genes Dev 16(1): 6-21. Bird AP (1987) CpG islands as gene markers in the vertebrate nucleus. Trends in Genetics 3(0): 342-347. Black AR, Black JD and Azizkhan-Clifford J (2001) Sp1 and kruppel-like factor family of transcription factors in cell growth regulation and cancer. J Cell Physiol 188(2): 143-160. Bloom DA and Jaiswal AK (2003) Phosphorylation of Nrf2 at Ser40 by Protein Kinase C in Response to Antioxidants Leads to the Release of Nrf2 from INrf2, but Is Not Required for Nrf2 Stabilization/Accumulation in the Nucleus and Transcriptional Activation of Antioxidant Response Element-mediated NAD(P)H:Quinone Oxidoreductase-1 Gene Expression. Journal of Biological Chemistry 278(45): 44675-44682. Blow MJ, McCulley DJ, Li Z, Zhang T, Akiyama JA, Holt A, Plajzer-Frick I, Shoukry M, Wright C, Chen F, Afzal V, Bristow J, Ren B, Black BL, Rubin EM, Visel A and Pennacchio LA (2010) ChIP-Seq identification of weakly conserved heart enhancers. Nat Genet 42(9): 806-810. Boutten A, Goven D, Boczkowski J and Bonay M (2010) Oxidative stress targets in pulmonary emphysema: focus on the Nrf2 pathway. Expert Opin Ther Targets 14(3): 329-346. Bouwman P and Philipsen S (2002) Regulation of the activity of Sp1-related transcription factors. Molecular and Cellular Endocrinology 195(1-2): 27-38. 106

Breathnach R and Chambon P (1981) Organization and expression of eucaryotic split genes coding for proteins. Annu Rev Biochem 50: 349-383. Bryan HK, Olayanju A, Goldring CE and Park BK (2013) The Nrf2 cell defence pathway: Keap1-dependent and - independent mechanisms of regulation. Biochem Pharmacol 85(6): 705-717. Bulger M and Groudine M (2011) Functional and Mechanistic Diversity of Distal Transcription Enhancers. Cell 144(3): 327-339. Burchiel SW, Thompson TA, Lauer FT and Oprea TI (2007) Activation of dioxin response element (DRE)- associated genes by benzo(a)pyrene 3,6-quinone and benzo(a)pyrene 1,6-quinone in MCF-10A human mammary epithelial cells. Toxicology and Applied Pharmacology 221(2): 203-214. Bushey AM, Dorman ER and Corces VG (2008) Chromatin insulators: regulatory mechanisms and epigenetic inheritance. Mol Cell 32(1): 1-9. Butler JEF and Kadonaga JT (2002) The RNA polymerase II core promoter: a key component in the regulation of gene expression. Genes & Development 16(20): 2583-2592. Calhoun VC, Stathopoulos A and Levine M (2002) Promoter-proximal tethering elements regulate enhancer- promoter specificity in the Drosophila Antennapedia complex. Proc Natl Acad Sci U S A 99(14): 9243- 9247. Carninci P, Sandelin A, Lenhard B, Katayama S, Shimokawa K, Ponjavic J, Semple CAM, Taylor MS, Engstrom PG, Frith MC, Forrest ARR, Alkema WB, Tan SL, Plessy C, Kodzius R, Ravasi T, Kasukawa T, Fukuda S, Kanamori-Katayama M, Kitazume Y, Kawaji H, Kai C, Nakamura M, Konno H, Nakano K, Mottagui- Tabar S, Arner P, Chesi A, Gustincich S, Persichetti F, Suzuki H, Grimmond SM, Wells CA, Orlando V, Wahlestedt C, Liu ET, Harbers M, Kawai J, Bajic VB, Hume DA and Hayashizaki Y (2006) Genome- wide analysis of mammalian promoter architecture and evolution. Nat Genet 38(6): 626-635. Cha YN, Martz F and Bueding E (1978) Enhancement of liver microsome epoxide hydratase activity in rodents by treatment with 2(3)-tert-butyl-4-hydroxyanisole. Cancer Res 38(12): 4496-4498. Chan Y, Fish JE, D'Abreo C, Lin S, Robb GB, Teichert AM, Karantzoulis-Fegaras F, Keightley A, Steer BM and Marsden PA (2004) The cell-specific expression of endothelial nitric-oxide synthase: a role for DNA methylation. J Biol Chem 279(33): 35087-35100. Chen C-H (2012a) Metabolic Conversion of Lipophilic Compounds, in Activation and Detoxification Enzymes: Functions and Implications (Chen C-H ed) pp 17-24, Springer, New York. Chen C-H (2012b) Phase II Enzymes, in Activation and Detoxification Enzymes: Functions and Implications (Chen C-H ed) pp 37-48, Springer, New York. Cheung KL and Kong AN (2010) Molecular targets of dietary phenethyl isothiocyanate and sulforaphane for cancer chemoprevention. The AAPS journal 12(1): 87-97. Choi YS, Kim S, Kyu Lee H, Lee KU and Pak YK (2004) In vitro methylation of nuclear respiratory factor-1 binding site suppresses the promoter activity of mitochondrial transcription factor A. Biochem Biophys Res Commun 314(1): 118-122. Chorley BN, Campbell MR, Wang X, Karaca M, Sambandan D, Bangura F, Xue P, Pi J, Kleeberger SR and Bell DA (2012) Identification of novel NRF2-regulated genes by ChIP-Seq: influence on retinoid X receptor alpha. Nucleic acids research. Coller JK, Fritz P, Zanger UM, Siegle I, Eichelbaum M, Kroemer HK and Murdter TE (2001) Distribution of microsomal epoxide hydrolase in humans: an immunohistochemical study in normal tissues, and benign and malignant tumours. The Histochemical journal 33(6): 329-336. Comb M and Goodman HM (1990) CpG methylation inhibits proenkephalin gene expression and binding of the transcription factor AP-2. Nucleic Acids Res 18(13): 3975-3982. Cronin A, Mowbray S, Dürk H, Homburg S, Fleming I, Fisslthaler B, Oesch F and Arand M (2003) The N- terminal domain of mammalian soluble epoxide hydrolase is a phosphatase. Proceedings of the National Academy of Sciences 100(4): 1552-1557. Cullinan SB, Gordan JD, Jin J, Harper JW and Diehl JA (2004) The Keap1-BTB Protein Is an Adaptor That Bridges Nrf2 to a Cul3-Based E3 Ligase: Oxidative Stress Sensing by a Cul3-Keap1 Ligase. Molecular and Cellular Biology 24(19): 8477-8486. Davuluri RV, Suzuki Y, Sugano S, Plass C and Huang THM (2008) The functional consequences of alternative promoter use in mammalian genomes. Trends in genetics : TIG 24(4): 167-177. 107

De Smet C, Lurquin C, Lethe B, Martelange V and Boon T (1999) DNA methylation is the primary silencing mechanism for a set of germ line- and tumor-specific genes with a CpG-rich promoter. Mol Cell Biol 19(11): 7327-7335. Decker M, Arand M and Cronin A (2009) Mammalian epoxide hydrolases in xenobiotic metabolism and signalling. Arch Toxicol 83(4): 297-318. Dekker J (2003) A closer look at long-range chromosomal interactions. Trends in biochemical sciences 28(6): 277-280. Dekker J (2006) The three 'C' s of chromosome conformation capture: controls, controls, controls. Nat Methods 3(1): 17-21. Dekker J, Rippe K, Dekker M and Kleckner N (2002) Capturing chromosome conformation. Science 295(5558): 1306-1311. Denison MS and Nagy SR (2003) Activation of the aryl hydrocarbon receptor by structurally diverse exogenous and endogenous chemicals. Annu Rev Pharmacol Toxicol 43: 309-334. Deshane J, Kim J, Bolisetty S, Hock TD, Hill-Kapturczak N and Agarwal A (2010) Sp1 regulates chromatin looping between an intronic enhancer and distal promoter of the human heme oxygenase-1 gene in renal cells. J Biol Chem 285(22): 16476-16486. Dhakshinamoorthy S, Long DJ, 2nd and Jaiswal AK (2000) Antioxidant regulation of genes encoding enzymes that detoxify xenobiotics and carcinogens. Curr Top Cell Regul 36: 201-216. Dostie J, Richmond TA, Arnaout RA, Selzer RR, Lee WL, Honan TA, Rubio ED, Krumm A, Lamb J, Nusbaum C, Green RD and Dekker J (2006) Chromosome Conformation Capture Carbon Copy (5C): a massively parallel solution for mapping interactions between genomic elements. Genome Res 16(10): 1299-1309. Enayetallah AE and Grant DF (2006) Effects of human soluble epoxide hydrolase polymorphisms on isoprenoid phosphate hydrolysis. Biochemical and Biophysical Research Communications 341(1): 254-260. Epstein RJ (2003) Gene expression, in Human Molecular Biology: An Introduction to the Molecular Basis of Health and Disease pp 77-95, Cambridge University Press. Erickson AM, Nevarea Z, Gipp JJ and Mulcahy RT (2002) Identification of a Variant Antioxidant Response Element in the Promoter of the Human Glutamate-Cysteine Ligase Modifier Subunit Gene. Journal of Biological Chemistry 277(34): 30730-30737. Esteller M (2002) CpG island hypermethylation and tumor suppressor genes: a booming present, a brighter future. Oncogene 21(35): 5427-5440. Fretland AJ and Omiecinski CJ (2000) Epoxide hydrolases: biochemistry and molecular biology. Chem Biol Interact 129(1-2): 41-59. Gardiner-Garden M and Frommer M (1987) CpG Islands in vertebrate genomes. Journal of Molecular Biology 196(2): 261-282. Gaston K and Fried M (1995) CpG methylation and the binding of YY1 and ETS proteins to the Surf-1/Surf-2 bidirectional promoter. Gene 157(1-2): 257-259. Gaszner M and Felsenfeld G (2006) Insulators: exploiting transcriptional and epigenetic mechanisms. Nat Rev Genet 7(9): 703-713. Gelboin HV (1969) A microsome-dependent binding of benzo[a]pyrene to DNA. Cancer Res 29(6): 1272-1276. Ghisletti S, Barozzi I, Mietton F, Polletti S, De Santa F, Venturini E, Gregory L, Lonie L, Chew A, Wei CL, Ragoussis J and Natoli G (2010) Identification and characterization of enhancers controlling the inflammatory gene expression program in macrophages. Immunity 32(3): 317-328. Goetz ME and Luch A (2008) Reactive species: A cell damaging rout assisting to chemical carcinogens. Cancer Letters 266(1): 73-83. Goyak KM, Johnson MC, Strom SC and Omiecinski CJ (2008) Expression profiling of interindividual variability following xenobiotic exposures in primary human hepatocyte cultures. Toxicol Appl Pharmacol 231(2): 216-224. Grande L, Bretones G, Rosa-Garrido M, Garrido-Martin EM, Hernandez T, Fraile S, Botella L, de Alava E, Vidal A, Garcia del Muro X, Villanueva A, Delgado MD and Fernandez-Luna JL (2012) Transcription Factors Sp1 and p73 Control the Expression of the Proapoptotic Protein NOXA in the Response of Testicular Embryonal Carcinoma Cells to Cisplatin. Journal of Biological Chemistry 287(32): 26495-26505. Grover PL and Sims P (1968) Enzyme-catalysed reactions of polycyclic hydrocarbons with deoxyribonucleic acid and protein in vitro. The Biochemical journal 110(1): 159-160. 108

Guenthner TM, Cai D and Wallin R (1998) Co-purification of microsomal epoxide hydrolase with the warfarin- sensitive vitamin K1 oxide reductase of the vitamin K cycle. Biochem Pharmacol 55(2): 169-175. Hagege H, Klous P, Braem C, Splinter E, Dekker J, Cathala G, de Laat W and Forne T (2007) Quantitative analysis of chromosome conformation capture assays (3C-qPCR). Nature protocols 2(7): 1722-1733. Hassett C, Alcher L, Sidhu JS and Omieclnskl CJ (1994) Human microsomal epoxide hydrolase: genetic poloymorphism and functional expression in vitro of amino acid variants. Human Molecular Genetics 3(3): 421-428. Hassett C, Laurenzana EM, Sidhu JS and Omiecinski CJ (1998) Effects of Chemical Inducers on Human Microsomal Epoxide Hydrolase in Primary Hepatocyte Cultures. Biochemical Pharmacology 55(7): 1059-1069. Hatchwell E and Greally JM (2007) The potential role of epigenomic dysregulation in complex human disease. Trends Genet 23(11): 588-595. Hattori N, Fujiwara H, Maeda M, Fujii S and Ueda M (2000) Epoxide hydrolase affects estrogen production in the human ovary. Endocrinology 141(9): 3353-3365. Hatzis P and Talianidis I (2002) Dynamics of enhancer-promoter communication during differentiation-induced gene activation. Mol Cell 10(6): 1467-1477. Hayes JD and McMahon M (2009) NRF2 and KEAP1 mutations: permanent activation of an adaptive response in cancer. Trends in biochemical sciences 34(4): 176-188. Hayes JD, McMahon M, Chowdhry S and Dinkova-Kostova AT (2010) Cancer chemoprevention mechanisms mediated through the Keap1-Nrf2 pathway. Antioxid Redox Signal 13(11): 1713-1748. Heintzman ND, Stuart RK, Hon G, Fu Y, Ching CW, Hawkins RD, Barrera LO, Van Calcar S, Qu C, Ching KA, Wang W, Weng Z, Green RD, Crawford GE and Ren B (2007) Distinct and predictive chromatin signatures of transcriptional promoters and enhancers in the human genome. Nat Genet 39(3): 311-318. Ho Y-S, Xiong Y, Ma W, Spector A and Ho DS (2004) Mice Lacking Catalase Develop Normally but Show Differential Sensitivity to Oxidant Tissue Injury. Journal of Biological Chemistry 279(31): 32804-32812. Holler R, Arand M, Meckey A, Oesch F and Friedberg T (1997) The Membrane Anchor of Microsomal Epoxide Hydrolase from Human, Rat, and Rabbit Displays an Unexpected Membrane Topology. Biochemical and Biophysical Research Communications 236(3): 754-759. Holstein SA and Hohl RJ (2004) Isoprenoids: remarkable diversity of form and function. Lipids 39(4): 293-309. Holtzclaw WD, Dinkova-Kostova AT and Talalay P (2004) Protection against electrophile and oxidative stress by induction of phase 2 genes: the quest for the elusive sensor that responds to inducers. Adv Enzyme Regul 44: 335-367. Hu R, Xu C, Shen G, Jain MR, Khor TO, Gopalkrishnan A, Lin W, Reddy B, Chan JY and Kong A-NT (2006a) Gene expression profiles induced by cancer chemopreventive isothiocyanate sulforaphane in the liver of C57BL/6J mice and C57BL/6J/Nrf2 (-/-) mice. Cancer Letters 243(2): 170-192. Hu X, Roberts JR, Apopa PL, Kan YW and Ma Q (2006b) Accelerated ovarian failure induced by 4-vinyl cyclohexene diepoxide in Nrf2 null mice. Mol Cell Biol 26(3): 940-954. Huang HC, Nguyen T and Pickett CB (2002) Phosphorylation of Nrf2 at Ser-40 by Protein Kinase C Regulates Antioxidant Response Element-mediated Transcription. Journal of Biological Chemistry 277(45): 42769- 42774. Illingworth RS and Bird AP (2009) CpG islands – ‘A rough guide’. FEBS letters 583(11): 1713-1720. Itoh K, Chiba T, Takahashi S, Ishii T, Igarashi K, Katoh Y, Oyake T, Hayashi N, Satoh K, Hatayama I, Yamamoto M and Nabeshima Y-i (1997a) An Nrf2/Small Maf Heterodimer Mediates the Induction of Phase II Detoxifying Enzyme Genes through Antioxidant Response Elements. Biochemical and biophysical research communications 236(2): 313-322. Itoh K, Chiba T, Takahashi S, Ishii T, Igarashi K, Katoh Y, Oyake T, Hayashi N, Satoh K, Hatayama I, Yamamoto M and Nabeshima Y (1997b) An Nrf2/small Maf heterodimer mediates the induction of phase II detoxifying enzyme genes through antioxidant response elements. Biochem Biophys Res Commun 236(2): 313-322. Jaiswal AK (2000) Regulation of genes encoding NAD(P)H:quinone oxidoreductases. Free Radic Biol Med 29(3- 4): 254-262. 109

Jin H, Kanthasamy A, Anantharam V, Rana A and Kanthasamy AG (2011) Transcriptional regulation of proapoptotic protein kinase Cdelta: implications for oxidative stress-induced neuronal cell death. J Biol Chem 286(22): 19840-19859. Juven-Gershon T, Hsu J-Y, Theisen JWM and Kadonaga JT (2008) The RNA polymerase II core promoter — the gateway to transcription. Current Opinion in Cell Biology 20(3): 253-259. Juven-Gershon T and Kadonaga JT (2010) Regulation of gene expression via the core promoter and the basal transcriptional machinery. Developmental biology 339(2): 225-229. Kaczynski J, Cook T and Urrutia R (2003) Sp1- and Kruppel-like transcription factors. Genome Biology 4(2): 206. Kadam S and Emerson BM (2003) Transcriptional Specificity of Human SWI/SNF BRG1 and BRM Chromatin Remodeling Complexes. Molecular Cell 11(2): 377-389. Kadauke S and Blobel GA (2009) Chromatin loops in gene regulation. Biochim Biophys Acta 1789(1): 17-25. Kaestner KH (2000) The Hepatocyte Nuclear Factor 3 (HNF3 or FOXA) Family in Metabolism. Trends in Endocrinology & Metabolism 11(7): 281-285. Kalthoff S, Ehmer U, Freiberg N, Manns MP and Strassburg CP (2010) Interaction between Oxidative Stress Sensor Nrf2 and Xenobiotic-activated Aryl Hydrocarbon Receptor in the Regulation of the Human Phase II Detoxifying UDP-glucuronosyltransferase 1A10. Journal of Biological Chemistry 285(9): 5993-6002. Kardassis D, Papakosta P, Pardali K and Moustakas A (1999) c-Jun Transactivates the Promoter of the Human p21WAF1/Cip1 Gene by Acting as a Superactivator of the Ubiquitous Transcription Factor Sp1. Journal of Biological Chemistry 274(41): 29572-29581. Kaspar JW, Niture SK and Jaiswal AK (2009) Nrf2:INrf2 (Keap1) signaling in oxidative stress. Free Radical Biology and Medicine 47(9): 1304-1309. Kasprzak KS (1995) Possible Role of Oxidative Damage in Metal-Induced Carcinogenesis. Cancer Investigation 13(4): 411-430. Kawaji H, Frith M, Katayama S, Sandelin A, Kai C, Kawai J, Carninci P and Hayashizaki Y (2006) Dynamic usage of transcription start sites within core promoters. Genome Biology 7(12): R118. Kensler TW, Wakabayashi N and Biswal S (2007) Cell Survival Responses to Environmental Stresses Via the Keap1-Nrf2-ARE Pathway. Annual Review of Pharmacology and Toxicology 47(1): 89-116. Keshet I, Schlesinger Y, Farkash S, Rand E, Hecht M, Segal E, Pikarski E, Young RA, Niveleau A, Cedar H and Simon I (2006) Evidence for an instructive mechanism of de novo methylation in cancer cells. Nat Genet 38(2): 149-153. Kim YR, Oh JE, Kim MS, Kang MR, Park SW, Han JY, Eom HS, Yoo NJ and Lee SH (2010) Oncogenic NRF2 mutations in squamous cell carcinomas of oesophagus and skin. The Journal of pathology 220(4): 446- 451. Kimura K, Wakamatsu A, Suzuki Y, Ota T, Nishikawa T, Yamashita R, Yamamoto J, Sekine M, Tsuritani K, Wakaguri H, Ishii S, Sugiyama T, Saito K, Isono Y, Irie R, Kushida N, Yoneyama T, Otsuka R, Kanda K, Yokoi T, Kondo H, Wagatsuma M, Murakawa K, Ishida S, Ishibashi T, Takahashi-Fujii A, Tanase T, Nagai K, Kikuchi H, Nakai K, Isogai T and Sugano S (2006) Diversification of transcriptional modulation: large-scale identification and characterization of putative alternative promoters of human genes. Genome Res 16(1): 55-65. Kiyohara C, Yoshimasu K, Takayama K and Nakanishi Y (2006) EPHX1 Polymorphisms and the Risk of Lung Cancer: A HuGE Review. Epidemiology 17(1): 89-99. Klaassen CD and Reisman SA (2010) Nrf2 the rescue: Effects of the antioxidative/electrophilic response on the liver. Toxicology and Applied Pharmacology 244(1): 57-65. Klaunig JE and Kamendulis LM (2004) THE ROLE OF OXIDATIVE STRESS IN CARCINOGENESIS. Annual Review of Pharmacology and Toxicology 44(1): 239-267. Klaunig JE, Kamendulis LM and Hocevar BA (2010) Oxidative Stress and Oxidative Damage in Carcinogenesis. Toxicologic Pathology 38(1): 96-109. Kobayashi A, Kang MI, Okawa H, Ohtsuji M, Zenke Y, Chiba T, Igarashi K and Yamamoto M (2004) Oxidative stress sensor Keap1 functions as an adaptor for Cul3-based E3 ligase to regulate proteasomal degradation of Nrf2. Mol Cell Biol 24(16): 7130-7139. Koch F, Jourquin F, Ferrier P and Andrau JC (2008) Genome-wide RNA polymerase II: not genes only! Trends in biochemical sciences 33(6): 265-273. 110

Köhle C and Bock KW (2007) Coordinate regulation of Phase I and II xenobiotic metabolisms by the Ah receptor and Nrf2. Biochemical Pharmacology 73(12): 1853-1862. Koo H-c, Davis JM, Li Y, Hatzis D, Opsimos H, Pollack S, Strayer MS, Ballard PL and Kazzaz JA (2005) Effects of transgene expression of superoxide dismutase and glutathione peroxidase on pulmonary epithelial cell growth in hyperoxia. AJP - Lung Cellular and Molecular Physiology 288(4): L718-726. Kwak M-K and Kensler TW (2010) Targeting NRF2 signaling for cancer chemoprevention. Toxicology and Applied Pharmacology 244(1): 66-76. Kwak MK, Itoh K, Yamamoto M, Sutter TR and Kensler TW (2001) Role of transcription factor Nrf2 in the induction of hepatic phase 2 and antioxidative enzymes in vivo by the cancer chemoprotective agent, 3H- 1, 2-dimethiole-3-thione. Molecular medicine (Cambridge, Mass) 7(2): 135-145. LaFramboise T (2009) Single nucleotide polymorphism arrays: a decade of biological, computational and technological advances. Nucleic Acids Res 37(13): 4181-4193. Landolin JM, Johnson DS, Trinklein ND, Aldred SF, Medina C, Shulha H, Weng Z and Myers RM (2010) Sequence features that drive human promoter function and tissue specificity. Genome Res 20(7): 890-898. Landry J-R, Mager DL and Wilhelm BT (2003) Complex controls: the role of alternative promoters in mammalian genomes. Trends in genetics : TIG 19(11): 640-648. Latham KE (1996) X chromosome imprinting and inactivation in the early mammalian embryo. Trends Genet 12(4): 134-138. Laurenzana EM, Chen T, Kannuswamy M, Sell BE, Strom SC, Li Y and Omiecinski CJ (2012) The Orphan Nuclear Receptor DAX-1 Functions as a Potent Corepressor of the Constitutive Androstane Receptor (NR1I3). Molecular Pharmacology 82(5): 918-928. Laurenzana EM, Hassett C and Omiecinski CJ (1998) Post-transcriptional regulation of human microsomal epoxide hydrolase. Pharmacogenetics 8(2): 157-167. Lee J-M, Calkins MJ, Chan K, Kan YW and Johnson JA (2003) Identification of the NF-E2-related Factor-2- dependent Genes Conferring Protection against Oxidative Stress in Primary Cortical Astrocytes Using Oligonucleotide Microarray Analysis. Journal of Biological Chemistry 278(14): 12029-12038. Lee LT, Tan-Un KC, Pang RT, Lam DT and Chow BK (2004) Regulation of the human secretin gene is controlled by the combined effects of CpG methylation, Sp1/Sp3 ratio, and the E-box element. Molecular endocrinology (Baltimore, Md) 18(7): 1740-1755. Lee WJ, Brennan P, Boffetta P, London SJ, Benhamou S, Rannug A, To-Figueras J, Ingelman-Sundberg M, Shields P, Gaspari L and Taioli E (2002) Microsomal epoxide hydrolase polymorphisms and lung cancer risk: a quantitative review. Biomarkers : biochemical indicators of exposure, response, and susceptibility to chemicals 7(3): 230-241. Lettice LA, Heaney SJ, Purdie LA, Li L, de Beer P, Oostra BA, Goode D, Elgar G, Hill RE and de Graaff E (2003) A long-range Shh enhancer regulates expression in the developing limb and fin and is associated with preaxial polydactyly. Hum Mol Genet 12(14): 1725-1735. Levings PP and Bungert J (2002) The human beta-globin locus control region. European journal of biochemistry / FEBS 269(6): 1589-1599. Lewis BA, Kim TK and Orkin SH (2000) A downstream element in the human beta-globin promoter: evidence of extended sequence-specific transcription factor IID contacts. Proc Natl Acad Sci U S A 97(13): 7172- 7177. Li L and Davie JR (2010) The role of Sp1 and Sp3 in normal and cancer cell biology. Annals of Anatomy - Anatomischer Anzeiger 192(5): 275-283. Li Q, Peterson KR, Fang X and Stamatoyannopoulos G (2002) Locus control regions. Blood 100(9): 3077-3086. Liang SH, Hassett C and Omiecinski CJ (2005) Alternative promoters determine tissue-specific expression profiles of the human microsomal epoxide hydrolase gene (EPHX1). Mol Pharmacol 67(1): 220-230. Lim CY, Santoso B, Boulay T, Dong E, Ohler U and Kadonaga JT (2004) The MTE, a new core promoter element for transcription by RNA polymerase II. Genes Dev 18(13): 1606-1617. Lin M-E, Herr DR and Chun J (2010) Lysophosphatidic acid (LPA) receptors: Signaling properties and disease relevance. Prostaglandins & Other Lipid Mediators 91(3–4): 130-138. Lin TS, Huang HH, Fan YH, Chiou SH and Chow KC (2007) Genetic polymorphism and gene expression of microsomal epoxide hydrolase in non-small cell lung cancer. Oncology reports 17(3): 565-572. 111

Liu F, Yuan D, Wei Y, Wang W, Yan L, Wen T, Xu M, Yang J and Li B (2012) Systematic review and meta- analysis of the relationship between EPHX1 polymorphisms and colorectal cancer risk. PLoS One 7(8): e43821. Lo R and Matthews J (2012) High-Resolution Genome-wide Mapping of AHR and ARNT Binding Sites by ChIP-Seq. Toxicological Sciences 130(2): 349-361. Lodish HF (2004) Transcriptional control of gene expression, in Molecular cell biology (Lodish HF ed) pp 447- 491, W. H. Freeman and Company, New York. Lomvardas S, Barnea G, Pisapia DJ, Mendelsohn M, Kirkland J and Axel R (2006) Interchromosomal Interactions and Olfactory Receptor Choice. Cell 126(2): 403-413. Lu AYH and Miwa GT (1980) Molecular Properties and Biological Functions of Microsomal Epoxide Hydrase. Annual Review of Pharmacology and Toxicology 20(1): 513-531. Ma Q, Battelli L and Hubbs AF (2006) Multiorgan Autoimmune Inflammation, Enhanced Lymphoproliferation, and Impaired Homeostasis of Reactive Oxygen Species in Mice Lacking the Antioxidant-Activated Transcription Factor Nrf2. The American Journal of Pathology 168(6): 1960-1974. Ma Q, Kinneer K, Bi Y, Chan JY and Kan YW (2004) Induction of murine NAD(P)H:quinone oxidoreductase by 2,3,7,8-tetrachlorodibenzo-p-dioxin requires the CNC (cap 'n' collar) basic leucine zipper transcription factor Nrf2 (nuclear factor erythroid 2-related factor 2): cross-interaction between AhR (aryl hydrocarbon receptor) and Nrf2 signal transduction. Biochem J 377(1): 205-213. Mantovani R (1998) A survey of 178 NF-Y binding CCAAT boxes. Nucleic Acids Res 26(5): 1135-1143. Mantovani R (1999) The molecular biology of the CCAAT-binding factor NF-Y. Gene 239(1): 15-27. Marin M, Karis A, Visser P, Grosveld F and Philipsen S (1997) Transcription factor Sp1 is essential for early embryonic development but dispensable for cell growth and differentiation. Cell 89(4): 619-628. Marlowe JL, Fan Y, Chang X, Peng L, Knudsen ES, Xia Y and Puga A (2008) The aryl hydrocarbon receptor binds to E2F1 and inhibits E2F1-induced apoptosis. Molecular biology of the cell 19(8): 3263-3271. Maston GA, Evans SK and Green MR (2006) Transcriptional Regulatory Elements in the Human Genome. Annual Review of Genomics and Human Genetics 7(1): 29-59. McMahon M, Thomas N, Itoh K, Yamamoto M and Hayes JD (2006) Dimerization of Substrate Adaptors Can Facilitate Cullin-mediated Ubiquitylation of Proteins by a “Tethering” Mechanism: A TWO-SITE INTERACTION MODEL FOR THE Nrf2-Keap1 COMPLEX. Journal of Biological Chemistry 281(34): 24756-24768. Merika M and Thanos D (2001) Enhanceosomes. Current opinion in genetics & development 11(2): 205-208. Merrell MD, Augustine LM, Slitt AL and Cherrington NJ (2008) Induction of drug metabolism enzymes and transporters by oltipraz in rats. Journal of Biochemical and Molecular Toxicology 22(2): 128-135. Mertens-Talcott SU, Chintharlapalli S, Li X and Safe S (2007) The Oncogenic microRNA-27a Targets Genes That Regulate Specificity Protein Transcription Factors and the G2-M Checkpoint in MDA-MB-231 Breast Cancer Cells. Cancer Research 67(22): 11001-11011. Miao W, Hu L, Scrivens PJ and Batist G (2005) Transcriptional Regulation of NF-E2 p45-related Factor (NRF2) Expression by the Aryl Hydrocarbon Receptor-Xenobiotic Response Element Signaling Pathway: DIRECT CROSS-TALK BETWEEN PHASE I AND II DRUG-METABOLIZING ENZYMES. Journal of Biological Chemistry 280(21): 20340-20348. Miele A and Dekker J (2008) Long-range chromosomal interactions and gene regulation. Molecular bioSystems 4(11): 1046-1057. Miele A and Dekker J (2009) Mapping Cis - and Trans - Chromatin Interaction Networks Using Chromosome Conformation Capture (3C), pp 105-121. Mitsuishi Y, Motohashi H and Yamamoto M (2012) The Keap1-Nrf2 system in cancers: stress response and anabolic metabolism. Frontiers in oncology 2: 200. Miyata M, Kudo G, Lee YH, Yang TJ, Gelboin HV, Fernandez-Salguero P, Kimura S and Gonzalez FJ (1999) Targeted disruption of the microsomal epoxide hydrolase gene. Microsomal epoxide hydrolase is required for the carcinogenic activity of 7,12-dimethylbenz[a]anthracene. J Biol Chem 274(34): 23963-23968. Moi P, Chan K, Asunis I, Cao A and Kan YW (1994) Isolation of NF-E2-related factor 2 (Nrf2), a NF-E2-like basic leucine zipper transcriptional activator that binds to the tandem NF-E2/AP1 repeat of the beta- globin locus control region. Proceedings of the National Academy of Sciences of the United States of America 91(21): 9926-9930. 112

Morisseau C and Hammock BD (2005) Epoxide hydrolases: mechanisms, inhibitor designs, and biological roles. Annu Rev Pharmacol Toxicol 45: 311-333. Morisseau C and Hammock BD (2013) Impact of soluble epoxide hydrolase and epoxyeicosanoids on human health. Annu Rev Pharmacol Toxicol 53: 37-58. Morisseau C, Schebb NH, Dong H, Ulu A, Aronov PA and Hammock BD (2012) Role of soluble epoxide hydrolase phosphatase activity in the metabolism of lysophosphatidic acids. Biochemical and Biophysical Research Communications 419(4): 796-800. Motohashi H, O'Connor T, Katsuoka F, Engel JD and Yamamoto M (2002) Integration and diversity of the regulatory network composed of Maf and CNC families of transcription factors. Gene 294(1-2): 1-12. Müller F, Demény MA and Tora L (2007) New Problems in RNA Polymerase II Transcription Initiation: Matching the Diversity of Core Promoters with a Variety of Promoter Recognition Factors. Journal of Biological Chemistry 282(20): 14685-14689. Murray IA, Flaveny CA, DiNatale BC, Chairo CR, Schroeder JC, Kusnadi A and Perdew GH (2010) Antagonism of Aryl Hydrocarbon Receptor Signaling by 6,2′,4′-Trimethoxyflavone. Journal of Pharmacology and Experimental Therapeutics 332(1): 135-144. Nakata K, Tanaka Y, Nakano T, Adachi T, Tanaka H, Kaminuma T and Ishikawa T (2006) Nuclear receptor- mediated transcriptional regulation in Phase I, II, and III xenobiotic metabolizing systems. Drug Metab Pharmacokinet 21(6): 437-457. Nelson JW, Subrahmanyan RM, Summers SA, Xiao X and Alkayed NJ (2013) Soluble epoxide hydrolase dimerization is required for hydrolase activity. J Biol Chem 288(11): 7697-7703. Newman JW, Morisseau C and Hammock BD (2005) Epoxide hydrolases: their roles and interactions with lipid metabolism. Progress in lipid research 44(1): 1-51. Newman JW, Morisseau C, Harris TR and Hammock BD (2003) The soluble epoxide hydrolase encoded by EPXH2 is a bifunctional enzyme with novel lipid phosphate phosphatase activity. Proceedings of the National Academy of Sciences 100(4): 1558-1563. Nioi P, McMahon M, Itoh K, Yamamoto M and Hayes JD (2003) Identification of a novel Nrf2-regulated antioxidant response element (ARE) in the mouse NAD(P)H:quinone oxidoreductase 1 gene: reassessment of the ARE consensus sequence. Biochemical Journal 374(2): 337-348. Nioi P and Nguyen T (2007) A mutation of Keap1 found in breast cancer impairs its ability to repress Nrf2 activity. Biochemical and Biophysical Research Communications 362(4): 816-821. Niture SK, Kaspar JW, Shen J and Jaiswal AK (2010) Nrf2 signaling and cell survival. Toxicology and Applied Pharmacology 244(1): 37-42. Noda S, Harada N, Hida A, Fujii-Kuriyama Y, Motohashi H and Yamamoto M (2003) Gene expression of detoxifying enzymes in AhR and Nrf2 compound null mutant mouse. Biochemical and Biophysical Research Communications 303(1): 105-111. Nolis IK, McKay DJ, Mantouvalou E, Lomvardas S, Merika M and Thanos D (2009) Transcription factors mediate long-range enhancer–promoter interactions. Proceedings of the National Academy of Sciences 106(48): 20222-20227. Noonan JP and McCallion AS (2010) Genomics of long-range regulatory elements. Annu Rev Genomics Hum Genet 11: 1-23. Ogbourne S and Antalis TM (1998) Transcriptional control and the role of silencers in transcriptional regulation in eukaryotes. The Biochemical journal 331 ( Pt 1): 1-14. Ohta T, Iijima K, Miyamoto M, Nakahara I, Tanaka H, Ohtsuji M, Suzuki T, Kobayashi A, Yokota J, Sakiyama T, Shibata T, Yamamoto M and Hirohashi S (2008) Loss of Keap1 Function Activates Nrf2 and Provides Advantages for Lung Cancer Cell Growth. Cancer research 68(5): 1303-1309. Omiecinski CJ, Hassett C and Hosagrahara V (2000) Epoxide hydrolase — polymorphism and role in toxicology. Toxicology Letters 112–113(0): 365-370. Omiecinski CJ, Vanden Heuvel JP, Perdew GH and Peters JM (2011) Xenobiotic Metabolism, Disposition, and Regulation by Receptors: From Biochemical Phenomenon to Predictors of Major Toxicities. Toxicological Sciences 120(suppl 1): S49-S75. Osburn WO and Kensler TW (2008) Nrf2 signaling: an adaptive response pathway for protection against environmental toxic insults. Mutat Res 659(1-2): 31-39. 113

Pansoy A, Ahmed S, Valen E, Sandelin A and Matthews J (2010) 3-Methylcholanthrene Induces Differential Recruitment of Aryl Hydrocarbon Receptor to Human Promoters. Toxicological Sciences 117(1): 90-100. Parker RE and Isaacs NS (1959) Mechanisms Of Epoxide Reactions. Chemical Reviews 59(4): 737-799. Pena P, Reutens AT, Albanese C, D'Amico M, Watanabe G, Donner A, Shu IW, Williams T and Pestell RG (1999) Activator Protein-2 Mediates Transcriptional Activation of the CYP11A1 Gene by Interaction with Sp1 Rather than Binding to DNA. Molecular Endocrinology 13(8): 1402-1416. Prendergast GC and Ziff EB (1991) Methylation-sensitive sequence-specific DNA binding by the c-Myc basic region. Science 251(4990): 186-189. Privalsky ML (2004) The role of corepressors in transcriptional regulation by nuclear hormone receptors. Annual review of physiology 66: 315-360. Raaka S, Hassett C and Omiencinski CJ (1998) Human microsomal epoxide hydrolase: 5'-flanking region genetic polymorphisms. Carcinogenesis 19(3): 387-393. Ramos-Gomez M, Kwak M-K, Dolan PM, Itoh K, Yamamoto M, Talalay P and Kensler TW (2001) Sensitivity to carcinogenesis is increased and chemoprotective efficacy of enzyme inducers is lost in nrf2 transcription factor-deficient mice. Proceedings of the National Academy of Sciences of the United States of America 98(6): 3410-3415. Reik W (2007) Stability and flexibility of epigenetic gene regulation in mammalian development. Nature 447(7143): 425-432. Reisman SA, Yeager RL, Yamamoto M and Klaassen CD (2009) Increased Nrf2 Activation in Livers from Keap1-Knockdown Mice Increases Expression of Cytoprotective Genes that Detoxify Electrophiles more than those that Detoxify Reactive Oxygen Species. Toxicol Sci 108(1): 35-47. Riddick DS, Huang Y, Harper PA and Okey AB (1994) 2,3,7,8-Tetrachlorodibenzo-p-dioxin versus 3- methylcholanthrene: comparative studies of Ah receptor binding, transformation, and induction of CYP1A1. J Biol Chem 269(16): 12118-12128. Rozenberg JM, Shlyakhtenko A, Glass K, Rishi V, Myakishev MV, FitzGerald PC and Vinson C (2008) All and only CpG containing sequences are enriched in promoters abundantly bound by RNA polymerase II in multiple tissues. BMC Genomics 9: 67. Ruike Y, Imanaka Y, Sato F, Shimizu K and Tsujimoto G (2010) Genome-wide analysis of aberrant methylation in human breast cancer cells using methyl-DNA immunoprecipitation combined with high-throughput sequencing. BMC Genomics 11: 137. Rushmore TH, Morton MR and Pickett CB (1991) The antioxidant responsive element. Activation by oxidative stress and identification of the DNA consensus sequence required for functional activity. Journal of Biological Chemistry 266(18): 11632-11639. Sabo PJ, Kuehn MS, Thurman R, Johnson BE, Johnson EM, Cao H, Yu M, Rosenzweig E, Goldy J, Haydock A, Weaver M, Shafer A, Lee K, Neri F, Humbert R, Singer MA, Richmond TA, Dorschner MO, McArthur M, Hawrylycz M, Green RD, Navas PA, Noble WS and Stamatoyannopoulos JA (2006) Genome-scale mapping of DNase I sensitivity in vivo using tiling DNA microarrays. Nat Methods 3(7): 511-518. Safe S and Abdelrahim M (2005) Sp transcription factor family and its role in cancer. European Journal of Cancer 41(16): 2438-2448. Safe S and Kim K (2004) Nuclear Receptor-Mediated Transactivation Through Interaction with Sp Proteins, in Progress in Nucleic Acid Research and Molecular Biology (Kivie M ed) pp 1-36, Academic Press. Sandelin A, Carninci P, Lenhard B, Ponjavic J, Hayashizaki Y and Hume DA (2007) Mammalian RNA polymerase II core promoters: insights from genome-wide studies. Nat Rev Genet 8(6): 424-436. Saxonov S, Berg P and Brutlag DL (2006) A genome-wide analysis of CpG dinucleotides in the human genome distinguishes two distinct classes of promoters. Proceedings of the National Academy of Sciences of the United States of America 103(5): 1412-1417. Schrader M and Fahimi HD (2006) Peroxisomes and oxidative stress. Biochim Biophys Acta 1763(12): 1755- 1766. Shibata T, Kokubu A, Gotoh M, Ojima H, Ohta T, Yamamoto M and Hirohashi S (2008a) Genetic Alteration of Keap1 Confers Constitutive Nrf2 Activation and Resistance to Chemotherapy in Gallbladder Cancer. Gastroenterology 135(4): 1358-1368.e1354. Shibata T, Kokubu A, Saito S, Narisawa-Saito M, Sasaki H, Aoyagi K, Yoshimatsu Y, Tachimori Y, Kushima R, Kiyono T and Yamamoto M (2011) NRF2 mutation confers malignant potential and resistance to 114

chemoradiation therapy in advanced esophageal squamous cancer. Neoplasia (New York, NY) 13(9): 864- 873. Shibata T, Ohta T, Tong KI, Kokubu A, Odogawa R, Tsuta K, Asamura H, Yamamoto M and Hirohashi S (2008b) Cancer related mutations in NRF2 impair its recognition by Keap1-Cul3 E3 ligase and promote malignancy. Proceedings of the National Academy of Sciences 105(36): 13568-13573. Shin S, Wakabayashi N, Misra V, Biswal S, Lee GH, Agoston ES, Yamamoto M and Kensler TW (2007) NRF2 Modulates Aryl Hydrocarbon Receptor Signaling: Influence on Adipogenesis. Molecular and Cellular Biology 27(20): 7188-7197. Simonis M, Klous P, Splinter E, Moshkin Y, Willemsen R, de Wit E, van Steensel B and de Laat W (2006) Nuclear organization of active and inactive chromatin domains uncovered by chromosome conformation capture-on-chip (4C). Nat Genet 38(11): 1348-1354. Simonis M, Kooren J and de Laat W (2007) An evaluation of 3C-based methods to capture DNA interactions. Nat Methods 4(11): 895-901. Singh A, Misra V, Thimmulappa RK, Lee H, Ames S, Hoque MO, Herman JG, Baylin SB, Sidransky D, Gabrielson E, Brock MV and Biswal S (2006) Dysfunctional KEAP1-NRF2 interaction in non-small-cell lung cancer. PLoS Med 3(10): e420. Sinha D, Biswas J and Bishayee A (2013) Nrf2-mediated redox signaling in arsenic carcinogenesis: a review. Arch Toxicol 87(2): 383-396. Slocum S and Kensler T (2011) Nrf2: control of sensitivity to carcinogens. Archives of Toxicology 85(4): 273- 284. Smale ST and Kadonaga JT (2003) THE RNA POLYMERASE II CORE PROMOTER. Annual Review of Biochemistry 72(1): 449-479. Snyder RC, Ray R, Blume S and Miller DM (1991) Mithramycin blocks transcriptional initiation of the c-myc P1 and P2 promoters. Biochemistry 30(17): 4290-4297. Solomon SS, Majumdar G, Martinez-Hernandez A and Raghow R (2008) A critical role of Sp1 transcription factor in regulating gene expression in response to insulin and other hormones. Life Sci 83(9-10): 305- 312. Spector AA and Norris AW (2007) Action of epoxyeicosatrienoic acids on cellular function. American Journal of Physiology - Cell Physiology 292(3): C996-C1012. St-Pierre J, Buckingham JA, Roebuck SJ and Brand MD (2002) Topology of superoxide production from different sites in the mitochondrial electron transport chain. J Biol Chem 277(47): 44784-44790. Su W, Jackson S, Tjian R and Echols H (1991) DNA looping between sites for transcriptional activation: self- association of DNA-bound Sp1. Genes Dev 5(5): 820-826. Suske G (1999) The Sp-family of transcription factors. Gene 238(2): 291-300. Suske G (2000) Transient transfection of Schneider cells in the study of transcription factors. Methods in molecular biology (Clifton, NJ) 130: 175-187. Suzuki T, Kimura A, Nagai R and Horikoshi M (2000) Regulation of interaction of the acetyltransferase region of p300 and the DNA-binding domain of Sp1 on and through DNA binding. Genes Cells 5(1): 29-41. Swedenborg E, Ruegg J, Hillenweck A, Rehnmark S, Faulds MH, Zalko D, Pongratz I and Pettersson K (2008) 3- Methylcholanthrene displays dual effects on estrogen receptor (ER) alpha and ER beta signaling in a cell- type specific fashion. Mol Pharmacol 73(2): 575-586. Szutorisz H, Dillon N and Tora L (2005) The role of enhancers as centres for general transcription factor recruitment. Trends in biochemical sciences 30(11): 593-599. Taguchi K, Motohashi H and Yamamoto M (2011) Molecular mechanisms of the Keap1–Nrf2 pathway in stress response and cancer evolution. Genes to Cells 16(2): 123-140. Takai D and Jones PA (2002) Comprehensive analysis of CpG islands in human chromosomes 21 and 22. Proc Natl Acad Sci U S A 99(6): 3740-3745. Tate PH and Bird AP (1993) Effects of DNA methylation on DNA-binding proteins and gene expression. Current opinion in genetics & development 3(2): 226-231. Thimmulappa RK, Mai KH, Srisuma S, Kensler TW, Yamamoto M and Biswal S (2002) Identification of Nrf2- regulated Genes Induced by the Chemopreventive Agent Sulforaphane by Oligonucleotide Microarray. Cancer research 62(18): 5196-5203. 115

Thomas MC and Chiang CM (2006) The general transcription machinery and general cofactors. Critical reviews in biochemistry and molecular biology 41(3): 105-178. Tillotson LG (1999) RIN ZF, a Novel Zinc Finger Gene, Encodes Proteins That Bind to the CACC Element of the Gastrin Promoter. Journal of Biological Chemistry 274(12): 8123-8128. Tirona RG and Kim RB (2005) Nuclear receptors and drug disposition gene regulation. Journal of pharmaceutical sciences 94(6): 1169-1186. Tokusumi Y, Ma Y, Song X, Jacobson RH and Takada S (2007) The new core promoter element XCPE1 (X Core Promoter Element 1) directs activator-, mediator-, and TATA-binding protein-dependent but TFIID- independent RNA polymerase II transcription from TATA-less promoters. Mol Cell Biol 27(5): 1844- 1858. Tolhuis B, Palstra R-J, Splinter E, Grosveld F and de Laat W (2002) Looping and Interaction between Hypersensitive Sites in the Active β-globin Locus. Molecular Cell 10(6): 1453-1465. Tolson AH and Wang H (2010) Regulation of drug-metabolizing enzymes by xenobiotic receptors: PXR and CAR. Advanced Drug Delivery Reviews 62(13): 1238-1249. Tong KI, Katoh Y, Kusunoki H, Itoh K, Tanaka T and Yamamoto M (2006) Keap1 recruits Neh2 through binding to ETGE and DLG motifs: characterization of the two-site molecular recognition model. Mol Cell Biol 26(8): 2887-2900. Valenzuela L and Kamakaka RT (2006) Chromatin insulators. Annual review of genetics 40: 107-138. Visel A, Blow MJ, Li Z, Zhang T, Akiyama JA, Holt A, Plajzer-Frick I, Shoukry M, Wright C, Chen F, Afzal V, Ren B, Rubin EM and Pennacchio LA (2009) ChIP-seq accurately predicts tissue-specific activity of enhancers. Nature 457(7231): 854-858. Wan J, Carr BA, Cutler NS, Lanza DL, Hines RN and Yost GS (2005) Sp1 and Sp3 regulate basal transcription of the human CYP2F1 gene. Drug metabolism and disposition: the biological fate of chemicals 33(8): 1244- 1253. Wang Q, Carroll JS and Brown M (2005) Spatial and temporal recruitment of androgen receptor and its coactivators involves chromosomal looping and polymerase tracking. Mol Cell 19(5): 631-642. Wang TTY (2002) β-Naphthoflavone, an Inducer of Xenobiotic Metabolizing Enzymes, Inhibits Firefly Luciferase Activity. Analytical Biochemistry 304(1): 122-126. Wang X-J, Sun Z, Villeneuve NF, Zhang S, Zhao F, Li Y, Chen W, Yi X, Zheng W, Wondrak GT, Wong PK and Zhang DD (2008) Nrf2 enhances resistance of cancer cells to chemotherapeutic drugs, the dark side of Nrf2. Carcinogenesis 29(6): 1235-1243. Wasserman WW and Fahl WE (1997) Functional antioxidant responsive elements. Proceedings of the National Academy of Sciences 94(10): 5361-5366. Weih F, Nitsch D, Reik A, Schutz G and Becker PB (1991) Analysis of CpG methylation and genomic footprinting at the tyrosine aminotransferase gene: DNA methylation alone is not sufficient to prevent protein binding in vivo. The EMBO journal 10(9): 2559-2567. Weinmann AS, Yan PS, Oberley MJ, Huang TH and Farnham PJ (2002) Isolating human transcription factor targets by coupling chromatin immunoprecipitation and CpG island microarray analysis. Genes Dev 16(2): 235-244. Wierstra I (2008) Sp1: Emerging roles--Beyond constitutive activation of TATA-less housekeeping genes. Biochemical and biophysical research communications 372(1): 1-13. Wilkening S, Stahl F and Bader A (2003) Comparison of primary human hepatocytes and hepatoma cell line Hepg2 with regard to their biotransformation properties. Drug metabolism and disposition: the biological fate of chemicals 31(8): 1035-1042. Xi H, Shulha HP, Lin JM, Vales TR, Fu Y, Bodine DM, McKay RD, Chenoweth JG, Tesar PJ, Furey TS, Ren B, Weng Z and Crawford GE (2007) Identification and characterization of cell type-specific and ubiquitous chromatin regulatory structures in the human genome. PLoS genetics 3(8): e136. Xu C, Li CY and Kong AN (2005) Induction of phase I, II and III drug metabolism/transport by xenobiotics. Archives of pharmacal research 28(3): 249-268. Yamashita R, Sugano S, Suzuki Y and Nakai K (2012) DBTSS: DataBase of Transcriptional Start Sites progress report in 2012. Nucleic Acids Research 40(D1): D150-D154. 116

Yang X, Liang SH, Weyant DM, Lazarus P, Gallagher CJ and Omiecinski CJ (2009) The expression of human microsomal epoxide hydrolase is predominantly driven by a genetically polymorphic far upstream promoter. J Pharmacol Exp Ther 330(1): 23-30. Yeager RL, Reisman SA, Aleksunes LM and Klaassen CD (2009) Introducing the “TCDD-Inducible AhR-Nrf2 Gene Battery”. Toxicological Sciences 111(2): 238-246. Yu R, Lei W, Mandlekar S, Weber MJ, Der CJ, Wu J and Kong ANT (1999) Role of a Mitogen-activated Protein Kinase Pathway in the Induction of Phase II Detoxifying Enzymes by Chemicals. Journal of Biological Chemistry 274(39): 27545-27552. Zelko IN, Mueller MR and Folz RJ (2010) CpG methylation attenuates Sp1 and Sp3 binding to the human extracellular superoxide dismutase promoter and regulates its cell-specific expression. Free Radic Biol Med 48(7): 895-904. Zhang D, Ai D, Tanaka H, Hammock BD and Zhu Y (2010) DNA methylation of the promoter of soluble epoxide hydrolase silences its expression by an SP-1-dependent mechanism. Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms 1799(9): 659-667. Zhang DD (2010) The Nrf2-Keap1-ARE signaling pathway: The regulation and dual function of Nrf2 in cancer. Antioxid Redox Signal 13(11): 1623-1626. Zhang Y, Li Y, Shibahara S and Takahashi K (2008) Synergistic activation of the human adrenomedullin gene promoter by Sp1 and AP-2[alpha]. Peptides 29(3): 465-472. Zhao S, Venkatasubbarao K, Li S and Freeman JW (2003) Requirement of a specific Sp1 site for histone deacetylase-mediated repression of transforming growth factor beta Type II receptor expression in human pancreatic cancer cells. Cancer Res 63(10): 2624-2630. Zhao W, Luo J and Cai X (2013) Association between microsomal epoxide hydrolase 1 polymorphisms and susceptibility to esophageal cancer: a meta-analysis. Tumour biology : the journal of the International Society for Oncodevelopmental Biology and Medicine. Zhao Z, Tavoosidana G, Sjolinder M, Gondor A, Mariano P, Wang S, Kanduri C, Lezcano M, Sandhu KS, Singh U, Pant V, Tiwari V, Kurukuti S and Ohlsson R (2006) Circular chromosome conformation capture (4C) uncovers extensive networks of epigenetically regulated intra- and interchromosomal interactions. Nat Genet 38(11): 1341-1347. Zhong JH, Xiang BD, Ma L, You XM, Li LQ and Xie GS (2013) Meta-analysis of microsomal epoxide hydrolase gene polymorphism and risk of hepatocellular carcinoma. PLoS One 8(2): e57064. Zhou C, Poulton EJ, Grun F, Bammler TK, Blumberg B, Thummel KE and Eaton DL (2007) The dietary isothiocyanate sulforaphane is an antagonist of the human steroid and xenobiotic nuclear receptor. Mol Pharmacol 71(1): 220-229. Zhu Q-s, Qian B and Levy D (2004a) Regulation of human microsomal epoxide hydrolase gene (EPHX1) expression by the transcription factor GATA-4. Biochimica et Biophysica Acta (BBA) - Gene Structure and Expression 1676(3): 251-260. Zhu QS, Qian B and Levy D (2004b) CCAAT/enhancer-binding protein alpha (C/EBPalpha) activates transcription of the human microsomal epoxide hydrolase gene (EPHX1) through the interaction with DNA-bound NF-Y. J Biol Chem 279(29): 29902-29910.

VITA

Shengzhong Su

Education Ph.D. Molecular Toxicology, 08/2013 Huck Institute of Life Sciences, Pennsylvania State University, University Park, PA. Advisor: Dr. Curtis Omiecinski

M. S. Entomology, 12/2002 Department of Entomology, Pennsylvania State University, University Park, PA. Advisor: Dr. Kelli Hoover

M. S. Insect Toxicology, 06/1996 Institute of Elemento-Organic Chemistry, Nankai University, Tianjin, China Advisor: Prof. Zhi-Zhen Shang

B. S. Environmental Biology, 06/1993 Department of Environmental Science, Nankai University, Tianjin, China

Awards and honors 2011 Society of Toxicology Molecular Biology Specialty Section (MBSS) Student Award, 3rd place 2010 Society of Toxicology Graduate Student Travel Award

Publications 1. SHENGZHONG SU, Xi Yang, and Curtis J. Omiecinski. The roles of intronic DNA elements in the regulation of the human microsomal epoxide hydrolase (EPHX1) driven by a far upstream alternative promoter. (In preparation) 2. SHENGZHONG SU, and Curtis J. Omiecinski. Transcription factors Sp1 and Sp3 contribute to the basal expression of human microsomal epoxide hydrolase driven by a far upstream promoter. (In preparation) 3. Scott S. Auerbach, Matthew A. Stoner, SHENGZHONG SU, and Curtis J. Omiecinski. Retinoid X Receptor- α-Dependent Transactivation by a Naturally Occurring Structural Variant of Human Constitutive Androstane Receptor (NR1I3). MOLECULAR PHARMACOLOGY, 2005,68:1239-1253 4. Kelli Hoover, Michael J. Grove, and SHENGZHONG SU. Systemic component to intrastadial developmental resistance in Lymantria dispar to its baculovirus. BIOLOGICAL CONTROL, 2002, 25: 92-98 5. SHENGZHONG SU, Zhisheng Jiang, and Zhizhen Shang. Comparative studies of insecticidal effects of IGRs on Calospilos suspecta. In THE 9TH JAPAN-CHINA SYMPOSIUM ON PESTICIDE SCIENCE, 1998, 286-92. 6. SHENGZHONG SU, Zhizhen Shang. Insecticidal effects of RH5849 on Semioihisa cinerearia Bremer et Grey and Calospilos suspecta Warren. ENTOMOLOGICAL KNOWLEDGE, 1997, 34(4): 225- 8(Chinese)