Nrf2 Plays an Important Role in Coordinated Regulation of Phase II Drug Metabolize Enzymes and Phase III Drug Transporters

Abstract

The nuclear transcription factor E2-related factor 2 (Nrf2) has been shown to play pivotal roles in preventing xenobiotics-induced toxicity and carcinogen- related tumorigenesis. These protective effects are mainly attributed to the induction of Phase II drug metabolizing/detoxification and antioxidant enzymes through the Nrf2-antioxidant response element (ARE) pathways. In this review, we will summarize the current research status on the identification of Nrf2-regulated drug metabolism enzymes (DMEs), especially Phase II DMEs, and Phase III drug transporters. In addition, the molecular mechanisms underlying the coordinated regulation of Phase II DMEs and Pharse III transporters are also discussed based on finding published in the literatures.

Introduction

Studies during the past decade have revealed that many drug metabolism enzymes (DME) are coordinately regulated by various nuclear receptors in response to the exposure of both endogenous substances and xenobiotics (reviewed in references [1, 2]). Among these nuclear receptors, aryl hydrocarbon receptor (AhR), constitutive androstane receptor (CAR) and pregenane X receptor (PXR) are well characterized for their roles in the induction of cytochrome P450 (CYP450) enzymes [1]. Binding of ligands to the ligand-binding domain (LBD) of a specific nuclear receptor will result in its nuclear translocation and trigger the transcription of its target DME genes. For example, AhR mediates the induction of human CYP1A1 enzyme upon the activation by its ligand 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) and its nuclear translocation is facilitated by a chaperone nuclear transporter Ah receptor nuclear translocator (ARNT) [3]. Similarly CAR and PXR are known to be involved in the coordinated regulation of human CYP2B6, CYP2C9 and CYP3A4 through cross-over mechanisms in response to inducers such as phenobarbital and rifampin [1]. Recently, new findings have shown that all these nuclear receptors are also involved in the regulation of Phase II DMEs as well as phase III drug transporters [1, 4]. Given their complementary role in the metabolism and disposition of xenobiotics and endogenous substances, it is not surprising that their regulation also shares similar mechanisms as those for phase I CYP450 system. .

In this review, we will discuss the important but less understood role of another nuclear transcription factor, nuclear transcription factor E2-related factor 2 (Nrf2), in the regulation of DMEs. Nrf2 belongs to the basic lucine zipper nuclear transcription factor family, which share regions of homology with that of Drosphila Cap’n’collar (CNC) protein [5]. The Nrf family transcription factors were identified in the search for proteins that regulate the expression of beta-globin genes during differentiation. Human Nrf2 was first isolated in 1994 from a hemin-induced K562 erythroid cell line and showed high sequence homology to the known p45 subunit of NF-E2 (nuclear factor erythroid 2) [5, 6]. However, unlike the p45 subunit which is only expressed in erythroid cells and megakaryocytes, expression of Nrf2 has been found in many tissues including liver, intestine, muscle, kidney and lung, suggesting a broad function of Nrf2 in these tissues [5]. Since then, the function of Nrf2 has been extensively investigated using various approaches including the Nrf2-deficient mouse model [7]. Interestingly, although mice having genetic deficient of Nrf2 gene appear to have normal phenotypes, they are more susceptible to butylated hydroxyltoluene (BHT) induced lung injury or acetoaminophen induced hepatoxicity at high doses [7, 8]. Therefore, Nrf2 is considered to play an important role in the cellular defense mechanisms against xenobiotic toxicity or carcinogenicity [9]. Further mechanistic studies demonstrated that the major mechanism of protection against carcinogenesis, mutagenesis and other form of toxicity elicited by carcinogen is through Nrf2-mediated induction of DME, particularly Phase II detoxification and antioxidant enzymes, such as glutathione S-transferase (GST), NAD(P)H:quinone oxidoreductase 1(NQO1), heme oxygenase-1 (HO-1) and UDP-glucuronosyltransferase (UGT); and more recently, transporters gene such as multidrug resistance-associated protein (MRP) [10]. Induction of these enzymes in cells is well known to confer resistance against carcinogen, reactive metabolites or reactive oxygen species by enhancing their elimination. Nrf2 regulates the transcription of these target genes through its binding to the consensus antioxidant response element (ARE) within the 5’-flanking promoter/enhancer region of these target genes [11, 12]. Because many cancer protective phytochemicals can elicit their cancer chemopreventive effects through Nrf2-dependent up-regulation of Phase II detoxification and antioxidant enzymes in many rodent cancer prevention studies [13], Nrf2 has been considered as an important molecular target to prevent cancer in human [14].

The focus of the current review will be on the Phase II DMEs and phase III drug transporters wherein their basal expression or induction are considered to be Nrf2-dependent. Additional focus will be on how these genes were coordinately regulated by Nrf2-dependent signaling pathway. Although a few recent studies also showed that the regulation of Cytochrome P450 genes such as mouse CYP2A5 and CYP3A11 also appears to be Nrf2-dependent [15, 16], the regulation of Phase I oxidative DME by Nrf2 will not be discussed in this review due to limited research in this field to date.

Mechanism of gene regulation by Nrf2

Unlike the general mechanism by which other nuclear receptors regulate majority of DMEs that introduced above, Nrf2-mediated gene transcription is not activated through ligand-receptor binding. Compare to nuclear receptors such as PXR, although Nrf2 also has a DNA-binding domain (DBD), it does not contain a ligand bindingg domain (LBD) that would bind to structurally specific ligands. This is consistent with previous findings that Nrf2 activators have very diverse chemical structures and at least nine structurally dissimilar classes of inducers have been found [17]. Interestingly, many Nrf2 chemical activators are electrophilic compounds.

Activation of Nrf2-meidated gene transcription involves very complex process and has been extensively investigated (reviewed in reference [9]). Under homeostatic condition, Nrf2 is mainly retained in the cytosol by the Keap1 protein, which is a homologue of the Drosphila actin binding protein Kelch that binds to the actin cytoskeleton. Keap1 can suppress the Nrf2 function by binding to the N-terminal Neh2 domain of Nrf2 and retains Nrf2 in the cytoplasma [18, 19]. Therefore, the Neh domain on the N-terminus of Nrf2 is distinct from the activation function 1 (AF-1) domain on other nuclear receptor such as PXR and plays a critical role in the regulation of basal activity of Nrf2 through the binding to Keap1. However, upon the challenge of oxidative or chemical stress, the Keapl’s sequestration can be reversed so that Nrf2 can be released from Keap1 and translocates into the nucleus to trigger the transcription of its target genes. Since many prototypical Nrf2 chemical activators or phase II inducers are electrophiles, it has been demonstrated that some of these electrophiles can directly modify the four most reactive cysteine residues on Keap1 [20]. Since Keap1 is a high cysteine-rich protein (mouse Keap1 contains 25 cysteines) that are very sensitive to oxidative/electrophile stress, conjugative modification of these cysteines on the Keap1 protein by Nrf2 activators could disrupt the binding between Keap1 and Nrf2, thereby to release the Nrf2 from the complex. Interestingly, studies have also shown that binding to the Keap1 protein also appears to accelerate the degradation of Nrf2 as compared to that of the unbound Nrf2 in the cytoplasm, which further explain the suppression effect of Keap1 binding under homeostatic condition [21, 22]. Once Nrf2 is released from Keap1 protein, it will translocate into the nucleus and form heterodimers with small Maf proteins and other Bzip transcription factors [23, 24]. Binding of the heterodimers, along with other nuclear transcriptional co-activators such as CBP [25], to the cis-acting element of ARE-containing target genes will result in the transcription of a battery of Nrf2 mediated genes.

It is noteworthy that in addition to the direct modification on the Keap1 protein to disrupt the Keap1/Nrf2 binding and activate Nrf2, several kinase signaling cascades including phosphatidylinositol 3-kinase (PI3K), protein kinase C (PKC) and mitogen-activated protein kinases (MAPKs) have also been proposed to control the release of Nrf2 from Keap1 binding and its nuclear translocation, especially when these kinases are activated by Phase II enzyme inducers or reactive electrophiles (reviewed in reference [26]). For example, PKC can directly phosphorylate Nrf2 N-terminus to facilitate its nuclear translocation [27]; while MAPK ERK, JNK and p38 pathways could play differential roles in the regulation of Nrf2 transactivation activity [25], probably through the regulation of its binding to AREs and other nuclear transcription co-activators such as CBP.

Interestingly, several findings also indicated that Nrf2 activity can be regulated at transcriptional level. Kwak et al. identified two ARE-like (AREL) elements in the 1 kb 5’-flanking promoter region of the mouse Nrf2 gene (Table 1) [28]. Based on deletion and mutagenesis studies, both AREL elements were found to be necessary to fully activate Nrf2 expression in response to a potent Nrf2 activator 3H-1,2-dithiole-3-thione (D3T). The binding of these two AREL elements to Nrf2 were also demonstrated by both electrophoretic mobility-shift assay (EMSA) and chromatin immunoprecipitation (ChIP) assay. Therefore, results from this study suggest that Nrf2 can auto-regulate its expression through the AREL elements in the proximal region of its promoter in response to its activators. The elevated nuclear level of Nrf2 from this mechanism will further enhance the expression of its target genes. In addition to this auto-regulation mechanism, Nrf2 expression was also demonstrated to be under the control of AhR [29]. Three AhR specific xenobiotic response element (XRE)-like elements (XREL) were identified after examining the DNA sequence of the promoter region of the mouse Nrf2 gene. Functionality of these XRELs in the induction of Nrf2 expression by AhR ligand TCDD was also confirmed using various approaches including site mutagenesis, EMSA, ChIP and siRNA of AhR. This new finding suggests that Nrf2-mediated gene transcription could be a downstream event of the AhR/ARNT pathway, and also suggests the existence of an integrated system to coordinately regulate both Phase I oxidation and Phase II DME to detoxify xenobiotics and carcinogens.

Table 1.

Summary of Nrf2-regulated drug metabolism enzymes (DME) and transporters

Genes	Species/Tissue	Reported ARE sequence	Inducer	References
Nrf2	mouse	AREL1: (−492) CCTGACTCCGCCAT	D3T	[28]
NQO1	human	(−472)5’-GCAGTCACAGTGACTCAGCAGAATCT-3’ (−445)	β-NF, t-BHQ	[57]
	rat	(−434)5’-TCTAGAGTCACAGTGACTTGGCAAAATCTGA-3’(−404)	β-NF, t-BHQ	[34]
	Mouse (liver & SIT)		BHA, D3T, SFN, I3C, OTZ, CMRN, AGLN	[23, 24, 35, 36]
GST
GSTA1	mouse	(−754)5’-TAGCTTGGAAATGACATTGCTAATGGTGACAAAGCAACTTT-3’(−714)	β-NF, t-BHQ, SFN; catechol, hydroquinone, 3-MC,EQ	[17, 37]
GSTA	rat	(−722)5’-GAGCTTGGAAATGGCATTGCTAATGGTGACAAAGCAACTTT-3’(−682)		[11]
GSTPiA	mouse	5’-TGACTCAGCATCCGGGGCGG-3’		[23, 39]
GSTPiB	mouse	5’-TGAGTCAGCATCCGGGGCGG-3’		[23, 39]
GSTPi	rat	5’-CAAAATAGTCAGTCACTATGATTCAGCAACAAACCC-3’		[40]
GSTM1/2/5	mouse (SIT)		curcumin	[24, 36, 48]
γGCS	human	5’-CGCACCGCCTCCCCGTGACTCAGCGCTTTGTGCGGG-3’	β-NA, EQ, BHA	[41]41]
	Mouse (liver & SIT)			[8, 24, 48]
UGT
UGT1A6	human	ARE1: 5’-CAGAAGCTCAGGTGAAGCTGACACGGCCATAGT-3’	t-BHQ	[44]
		ARE2: 5’-TTACCCACAACTTCTGTCTGACTTGGCAAAAAT-3’
		ARE3: 5’-AACTCGCGTGCCAGCCAGGTGTGCATGACTAGCTCTGG-3’ (XRE)
	mouse			[7, 8, 36]
	rat (liver)		EQ, OTZ	[46]
UGT1A7	rat (liver)		EQ, OTZ	[46]
UGT1A8	human		EGCG	[47]
UGT1A10	human		EGCG	[47]
UGT2B1	rat (liver)		EQ, OTZ	[46]
UGT2B3	rat (liver & SIT)		EQ, OTZ	[46]
UGT2B5	mouse (SIT)		curcumin	[36, 48]
UGT2B12	rat (liver & SIT)		EQ, OTZ	[46]
SULT
SULT1A1	mouse(liver)		BHA	[49]
SULT1B1	mouse(liver)		OTZ, EQ	[49]
SULT1C1	mouse(liver)		BHA	[49]
SULT1C2	mouse(liver)		EQ	[49]
SULT1D1	mouse(liver)		BHA	[49]
SULT1E1	mouse(liver)		BHA	[49]
SULT3A1	mouse(liver)		EQ, BHA	[49]
SULT5A1	mouse(liver)		OTZ, EQ	[49]
MRP
MRP1	human	ARE1: (−499) 5’-GTGACTCAGC-3’ (−490)	t-BHQ, quercetin	[52, 58]
		ARE2: (−1843) 5’-GTGACAAAGCA-3’(−1833)
	mouse	ARE1: (−1240) 5’-TTGAGTTAGCT-3’ (−1229)	DEM	[52]
		ARE2: (−2020) 5’-TTGAGACAGCA-3’(−2010)
MRP2	human	AREL1: (−340) 5’-TGAAAGACTGTGCACTCTTGA-3’	2AAF	[53, 54]
		AREL2: (−1090) 5’-AAGCAATTTAAGTGACAGTACAAAAGG-3’
		AREL3: (−2865) 5’-TACTGATGCTGCCCTTT-3’(−2856)
	mouse	ARE1: (−95) 5’-ATGACATAGCA-3’ (−85)	t-BHQ, BHA, β-NF, 2AAF, OTZ, EQ	[50, 51, 53]
		ARE2: (−1391) 5’-CTGACATGGCA-3’(−1381)
MRP3	mouse	ARE1: (−9919) 5’-GCAGTTAGTTAATGACTCTGCTTGTGAGTATCA-3’	t-BHQ, BHA, OTZ EQ	[50, 51]
	rat (liver)		GA, p-CA	[55]
MRP4	mouse	ARE1: (−3295) 5’-AGAGGCTCACAGTGACCTGGCAAAAAGCAG-3’	t-BHQ, BHA, OTZ, EQ	[50, 51]
		AREL2&3: (−3753) 5’-GAGCTGCTGAGTCACTCCTGACTCCAGCTTCTG-3’
		ARE4: (−9104) ’-TCTTAGACGTACAGTGACGTTGCTGTCTCCTAAAC-3’
		ARE5: (−10462) 5’-GTGCCAACCTCCTCTGACAAGGCCCTTGGCTCCTG-3’
MRP5	mouse (liver)		BHA, OTZ, EQ	[50, 51]
MRP6	mouse (liver)		BHA, OTZ, EQ	[50, 51]
OATP2B	mouse (liver)		BHA, EQ	[15]

The consensus ARE sequence has been characterized as (A/G) TGA(C/T)NNNGC(A/G) (N represents any nucleotides) and the core ARE sequence is TGACNNNGC. The core ARE sequences in the promoter region are underlined (imperfect nucleotides were bolded). The position of ARE or ARE like (AREL) sequences is also provided according to the original reports.

Abbreviations: β-naphthoflavone (β-NF); tert-butyl-hydroquinone (t-BHQ); 1,2-dithiole-3-thione (D3T); butylated hydroxyanisole (BHA); ethoxyquin (EQ); oltipraz (OTZ); indole-3-carbinol (I3C); sulforaphane (SFN); coumarin (CMRN); α-agelicalactone (AGLN); 3-methylcholanthrene (3-MC); diethyl maleate (DEM); (−)-epigallocatechin-3-gallate (EGCG); 2-acetylaminofluorene (2AAF); gallic acid (GA); p coumaric acid (p-CA)

Nrf2 regulated Phase II drug metabolize enzymes and drug transporters

Phase II DMEs and antioxidant enzymes such as NAD(P)H:quinone oxidoreductase 1 (NQO1) and glutathione S-transferases (GST) play critical roles in protecting cells from xenobiotic- or carcinogen-induced oxidative stress, cytotoxicity, mutagenicity and carcinogenecity. Because many of these enzymes were found to be induced by planar aromatic compounds and phenolic antioxidants in early studies [11], both AhR-dependent and -independent regulation mechanisms were proposed [30, 31]. Efforts to address the AhR-independent regulation mechanism lead to the identification of consensus ARE cis-element [11] in the 5’-flanking promoter region of these genes such as Rat GST Ya subunit and NQO1. However, the ARE-mediated regulation mechanism was not completely understood until the isolation of nuclear transcription factor Nrf2 that specifically binds to the cis-acting ARE component. Since then, many Nrf2 target genes were identified based on the existence and functional characterization of ARE or ARE-like elements in their 5’-flanking promoter region. Interestingly, although the core ARE sequence has been identified as 5’-GTGACnnnGC-3’ where n represents any nucleotide [11], but this sequence alone is not sufficient to mediate the induction [12]. In stead, the ARE sequence has been defined as 5’-TMAnnnRTGAYnnnGCRWWW-3’ (M = A or C, R = A or G, Y = C or T, W= A or T, S = G or C), in which Nrf2 is proposed to bind to the RTGAYnnn portion of the core sequence (underlined). In addition, the presence a tandem core sequence (underlined) at upstream 5’-flanking region of the consensus sequence will confer the maximum induction capability by inducers. Importantly, the establishment of genetic Nrf2-deficient mouse strain provided the critical link to identify Nrf2 target genes. With the aid of genome-wide microarray techniques, many Nrf2-dependent genes are characterized based on their differential basal/constitutive or inducible expression pattern between Nrf2 wild type and Nrf2 knockout (KO) mice.

Phase II drug metabolize enzymes

NQO1

NQO1 detoxifies quinones and its derivatives to protect cells against redox cycling and oxidative stress. Human NQO1 [32] gene contains a twenty-four base pairs of ARE in its promoter region that is essential for its high basal transcription as well as the induction by prototypical Nrf2 activator such as tert-butyl-hydroquinone (t-BHQ) and β-naphthoflavone (β-NF) (Table 1). Interestingly, the consensus ARE in human NQO1 contains one perfect and one imperfect AP1 element that may suppress the basal expression of NQO1 by the binding of c-Fos and Fra1 [33]. Nevertheless, both Nrf1 and Nrf2 were found to be involved in the ARE-mediated expression and induction of human NQO1 by these xenobiotics. Similarly, in addition to the Xenobiotic Response Element (XRE), a consensus ARE sequence was also characterized in rat NQO1 gene (Table 1) and was involved in the induction of rat NQO1 by t-BHQ and β-NF, but not by TCDD [34]. Although ARE hasn’t been reported in mouse NQO1 gene, but the induction of NQO1 by Phase II enzyme inducer including D3T and butylated hydroxyanisole (BHA) was largely abolished in both the liver and small intestine (SIT) of Nrf2-deficient mouse as compared to that in the Nrf2 wild type mouse [23, 35]. Interestingly, the basal level expression of NQO1 was only significantly decreased in the SIT of the Nrf2 knockout mouse, but not affected in the liver [24, 35, 36]. These findings from the Nrf2-deficient mouse studies suggest that the basal transcriptional regulation of NQO1 gene in the liver probably involves multiple pathways other than Nrf2-ARE pathway [24]. This hypothesisis very likely to be true considering the presence of numerous nuclear receptors controlling the expression of DMEs in the liver than in the SIT, which may result in the overall lower contribution from Nrf2-dependent mechanism in the liver than in SIT on the basal transcription of mouse NQO1.

GST

The GSTs are Phase II enzymes that catalyze the conjugation of reduced glutathione to electrophilic xenobiotics or reactive metabolites, products from Phase I oxidation. Mouse and rat alpha class GST Ya subunit gene is the first GST gene its expression and induction by Phase II inducers are found to be dependent on the ARE sequence in its 5’-flanking region (Table 1) [11, 17, 37, 38]. The promoter regions of several GSTs Pi class of genes in both mouse (GSTpiA and GSTpiB) and rat (GSTpi) have also been investigated [23, 39, 40]. Their ARE sequences containing the ARE core sequence differ slightly from the ARE consensus (Table 1).

In studies utilizing the homozygous Nrf2-deficient mouse model, researchers have found that the induction of GST alpha, Mu and pi isoforms was largely impaired in both the liver and the SIT of Nrf2-deficient mouse after BHA treatment [23]. In another similarly designed study [24], although the basal level of class alpha and Mu isoforms were 3–6 folds lower in the SIT of Nrf2-deficient mouse, the induction of some of these GSTs (GSTA1/2,4 and M1) by ethoxyquin (EQ) and BHA seems to be independent of Nrf2 status based on western blotting analysis. Although the promoter of GSTA1 contains a perfect ARE located 4 bp downstream from a closely related imperfect ARE (Table 1) [23], the region encompasses both elements can not form an effective palindromic structure but a head-to-tail tandem direct repeat which may not be able to bind Nrf2 complex tightly [12, 24]. In the same study, constitutional expression level of GSTpi class was not affected in SIT. However, study using microarray technique to compare the global gene expression profiles between Nrf2 wild-type and Nrf2-deficient mouse showed that basal level of few GSTs (GSTM1/2) were decreased by 3~6 folds in the Nrf2-deficient mouse [36]. When treated with a potent Phase II inducer sulforaphane (SFN), the inducibility of many GSTs (GSTA1/2/3 and GSTM1) was diminished due to the deficiency of Nrf2 in SIT. Based on studies described above, although incongruent results have been observed regarding the inducibility of individual GST isoforms (such as GSTPi), it is likely that the regulation of constitutive and inducible expression of GSTs by Nrf2 is dependent on multiple factors, including chemical inducers, tissues and species.

γGCS

In accordance with the induction of GSTs, expression of heavy chain of γGCS, an enzyme which catalyze the rate-limiting step in the de novo synthesis pathway of glutathione, was also found to be dependent on Nrf2 in the mouse. In the liver of Nrf2-deficient mouse, decreased basal level of γGCS heavy chain mRNA was observed [8]; furthermore, its induction was found to be largely abolished in SIT of Nrf2-deficient mouse in response to ethoxyquin (EQ) and BHA induction [24]. Two AREs (ARE3 and ARE4) separated by 34 bp were identified in the −3802 to −2752 bp promoter region of human γGCS heavy chain gene (Table 1) [41]. Interestingly, although ARE3 seems to be not involved to the β-NA induction, the tandem structure between ARE3 and ARE4 dramatically enhance the basal expression level of γGCS promoter-linked luciferase reporter activity. This result is consistent with the previous functional analysis of ARE sequences [12]. The identification γGCS and various GSTs as Nrf2 target genes is another good example of coordinated regulated mechanism on the glutathione-related detoxification process.

UGT

UDP-glucuronosyltransferases (UGTs) catalyze one of the major Phase II conjugation and detoxification process, the glucuronidation. Decreased basal expression level of UGT1A6 was observed in the liver, SIT and lung of Nrf2-deficient mouse [7, 8, 36, 42, 43]. In addition to the binding sites of AhR and PXR receptors, human UGT1A6 gene promoter region also has three ARE-like sequences (Table 1) [44]. However, the involvement of these three AREs in t-BHQ induction was not fully characterized, since the t-BHQ effects were not abolished by site-directed mutagenesis in both ARE3 and XRE (15bp upstream of ARE3), while the function of the other two AREs containing consensus ARE sequence were not investigated in that study. In contrast, rat UGT1A6 gene promoter has no ARE-like sequence but an AhR binding sequence [45], but significant induction of UGT1A6 and UGT1A7 were observed in rat liver after treatments of Nrf2 activators EQ and oltiparz (OTZ) [46]. Several other UGTs were also been implicated as Nrf2 target genes. For example, expression of UGT1A8 and UGT1A10 were inhibited by siRNA of Nrf2 in human caco-2 cells [47]. UGT2 family members such as UGT2B1, UGT2B3, UGT2B12 were all induced by prototypical Nrf2 activators EQ and OTZ in rat liver or duodenum [46]. UGT2B5 induction was also found to be largely diminished in Nrf2-deficient mouse after curcumin treatment [48]. Overall, some of the UGTs have been shown to be Nrf2 target genes, however, the contribution of Nrf2-ARE pathway on the regulation of specific UGTs may be species-dependent and may also not be as prominent due to involvement of other nuclear receptors such as PXR and AhR.

Sulfotransferases (SULT)

Most recently, the expression of several sulfotransferases (SULTs) and 3’-phosphoadenosine 5’-phosphosulfate synthase (PAPSs) isozymes has been demonstrated to be regulated by several prototypical Nrf2 activators such as OTZ, EQ and BHA (Table 1) [49]. The induction of SULTs and PAPS2 by these chemical activators is more prominent than AhR, CAR, PXR and PPAR activators, and the induction was only observed in the liver of male mice (not in females). This result is in agreement with the findings that many of these inducible SULTs contains ARE sequences in their promoter regions based on the authors’ in silico DNA analysis (unpublished observations). Interestingly, most of the AhR activators suppress the expression of both SULTs and PAPSs, which may explain the observation that OTZ induced much less SULTs than BHA and EQ since OTZ can also activate the AhR pathway. Since Nrf2-ARE pathway has been proposed to be downstream of the AhR pathway, the differential regulation of SULTs and PAPSs by AhR activators (mainly suppression) and Nrf2 activators (mainly induction) indicated a more complex regulatory mechanism. One possible explanation is that these chemical inducers may also affect many other pathways at high concentrations in vivo.

Drug Transporters

Multidrug Resistance-Associated Protein Transporters (MRPs)

Drug transporters can function synergistically with Phase II conjugating enzymes to facilitate the cellular excretion of conjugated metabolites, which is believed to be an important detoxification process to remove carcinogens, reactivate metabolites and xenobiotics. Since Nrf2 has been shown to play a pivotal role in the induction of many Phase II enzymes, it is not surprising recently to find that Nrf2-ARE pathway also regulates the expression of many drug transporters, such as multidrug resistance-associated protein transporters (MRPs), especially to the response of oxidative and electrophilic stress. MRP transporters are adenosine triphosphate-dependent drug transporters that efflux endogenous substances (such as bilirubin), xenbiotics and their metabolite (especially phase II conjugated metabolites). Using prototypical chemical activators for different nuclear receptors that regulate DME genes, Klaassen’s group has identified many MRPs that were induced by Nrf2 activators [10]. Maher et al. first found that mouse liver MPR2-6 were all induced by Nrf2 activators including OTZ, BHA and EQ [50]. In a more mechanistic follow-up study [51], the same group demonstrated that basal expression as well as the induction of mouse liver Mrp2, 3, 4 genes by Nrf2 activators BHA and OTZ are Nrf2-dependent. In vitro mechanistic studies in mouse Hepa1c1c7 cells demonstrated the specific binding of Nrf2 to the multiple putative ARE sequences in the promoter region of all three MRPs (Table 1). These results are consistent with the identification of ARE or ARE-like sequences in both mouse and human MRP1/2 genes (Table 1) [52–54]. The binding of Nrf2 to the mouse MRP2 AREs were also demonstrated in the previous study [53]. Regulation of rat MRP gene by Nrf2 has not been extensively investigated, however, induction of rat liver MRP3 gene by phenolic acid GA (gallic acid) and p-CA (p-coumaric acid) was observed and this induction was associated with the accumulation of Nrf2 level in liver cells [55]. Taken together, activation of Nrf2-ARE pathway by chemical inducers appears to be able to stimulate the coordinated induction of multiple hepatic MRPs.

OATPs

The regulation of organic anion-transporting polypeptides (OATPs) in the mouse liver was also investigated using similar approaches by Klaassen’s group [15] as MRP’s as described above. However, unlike MRPs, only mouse OATP2B1 genes was induced by Nrf2 activators BHA and EQ, suggesting the protential limited involvement of Nrf2 in the regulation of this class of transporters.

Summary

Research over the past decade utilizing mouse genetic knockout models, promoters/enhancers mapping technologies and prototypical chemical inducers have revealed that DMEs are coordinately regulated by a wide variety of nuclear receptors including AhR, CAR, PXR, FXR, LXR and PPAR [1]. However, the function of Nrf2 in the regulation of DMEs, especially Phase II DMEs and drug transporters, has not been fully investigated and gained much attention. Although some Phase II conjugation and transporter genes are characterized as Nrf2-dependent genes based on their inducibility by prototypical Nrf2 chemical activators or decreased basal expression levels in Nrf2-deficient mice, evidence of direct binding between Nrf2 and AREs in their promoter region is still necessary to support that conclusion. In addition, the contribution of Nrf2 in the regulation of these DMEs compared to other nuclear receptors (such as CAR and PXR) may highly dependent on the type of chemical inducers/activators, since numerous nuclear receptor binding elements were reported in the promoter regions of these genes such as UGT1A6 and MRP2. In addition, cross-talk between Nrf2-ARE pathway and other pathways such as AhR and CAR has also been proposed [29, 56], which makes the interpretation of induction studies using chemical inducers more complicated because of the lack of specificity of some of these chemical activators. Nevertheless, the coordinated regulation of phase II antioxidant and conjugation enzymes, as well as phase III drug transporters, and the potential cross-talk with other nuclear receptors, clearly indicates the pivotal role of Nrf2 in xenobiotics metabolism and detoxification process at these in these rodent models. Translating these obervations to human situations are clearly warranted.

Figure 1.