Oncogene (2002) 21, 5301 – 5312 ª 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc
c-Maf negatively regulates ARE-mediated detoxifying enzyme genes expression and anti-oxidant induction Saravanakumar Dhakshinamoorthy1 and Anil K Jaiswal*,1 1
Department of Pharmacology, Baylor College of Medicine, One Baylor Plaza, Houston, Texas, TX 77030, USA
Anti-oxidant response element (ARE) and nuclear factors including Nrf2 and small Maf (MafG and MafK) proteins are known to regulate expression and induction of detoxifying enzyme genes including quinone oxidoreductase1 (NQO1). Nrf2 upregulates and small Maf proteins lacking the transcriptional activation domain down regulates ARE-mediated expression and induction. In this report, we have investigated the role of c-Maf (large Maf) containing the transcriptional activation domain in the regulation of ARE-mediated genes expression. The overexpression of c-Maf in human hepatoblastoma (Hep-G2) cells led to the repression of ARE-mediated NQO1 and GST Ya genes expression and induction in response to tert-butyl hydroquinone (tBHQ). This was in contrast to the role of c-Maf in the activation of Maf recognition element (MARE) mediated p53 gene expression. Deletion of transcriptional activation domain of c-Maf (^c-Maf) led to significant loss of MARE-mediated p53 gene expression but had no effect on the repression of ARE-mediated NQO1 gene expression. The overexpression of MafG in Hep-G2 cells repressed both ARE and MARE-mediated genes expression. The co-expression of c-Maf with MafG rescued the MafG repression of MARE but not AREmediated gene expression. Band and super shift assays showed the presence of c-Maf in the ARE-nuclear protein complex. Similar assays with in vitro translated proteins revealed that both c-Maf and ^c-Maf bound to NQO1 gene ARE as homodimers and heterodimers with small Maf but not as heterodimers with Nrf2. Mutational analysis of the NQO1 gene ARE indicated that core ARE sequence is essential for binding of c-Maf leading to repression of NQO1 gene expression. Northern analysis revealed that c-Maf expression increases 2 h after t-BHQ treatment. It reached a plateau at 4 h after t-BHQ treatment. The results together led to the conclusion that c-Maf negatively regulates AREmediated detoxifying enzyme genes expression and induction in response to anti-oxidants. Oncogene (2002) 21, 5301 – 5312. doi:10.1038/sj.onc. 1205642 Keywords: NQO1; ARE; t-BHQ; c-Maf
*Correspondence: AK Jaiswal; E-mail:
[email protected] Received 7 February 2002; revised 24 April 2002; accepted 29 April 2002
Introduction Anti-oxidant response element (ARE) was found in the promoter region of NAD(P)H : quinone oxidoreductase1 (NQO1), glutathione S-transferase Ya subunit (GST Ya), g-glutamyl cysteine synthetase (g-GCS), heme oxygenase 1 (HO-1) and ferritin-L genes (Dhakshinamoorthy et al., 2000). ARE is known to regulate basal expression and co-ordinated induction of these genes in response to xenobiotics, anti-oxidants and oxidants in all types of tissues (Dhakshinamoorthy et al., 2000). The Maf recognition element (MARE) is a cis element that regulates the transcription of many genes expressed in erythroid and megakaryocytic cells (Blank and Andrews, 1997; Motohashi et al., 2000). These include b-globin gene loci, erythroid 5-aminolevulinate synthase (ALAS) and porphobilinogen deaminase (PBGD). Recently, MARE element was also found in the 5’ flanking region of tumor suppressor p53 gene that regulates its expression (Hale et al., 2000). Many cap’n’collar (CNC) and leucine zipper (b-zip) transcription factors including Nrf1, Nrf2 and small Maf (MafG, K and F) proteins bind to the ARE and MARE elements and regulate the expression and induction of down stream genes (Dhakshinamoorthy et al., 2000; Blank and Andrews, 1997; Motohashi et al., 2000; Hale et al., 2000). Among these, the small Maf (MafG, MafK, MafF) and large Maf (c-Maf, MafB) proteins constitute a family of leucine zipper nuclear transcription factors that repress, as well as activate, transcription of many eukaryotic genes (Kataoka et al., 1994a; Kim and Andrews, 1997; Fujiwara et al., 1993; Matsushima-Hibiya et al., 1998). Their gene products are closely related to vMaf, especially in the structure of the DNA-binding and leucine zipper domains. The small Maf proteins, however, lack the amino-terminal transcriptional activation domain of v-Maf (Kim and Andrews, 1997; Fujiwara et al., 1993; Kataoka et al., 1993). Small Maf proteins are known to homodimerize and heterodimerize with Nrf1 and Nrf2 (Marini et al., 1997). Small Maf – Maf homodimers repressed, yet small Maf-Nrf heterodimers activated transcription of the MARE-mediated b-globin gene (Marini et al., 1997). Large Maf (c-Maf and MafB) proteins containing the amino-terminal transcriptional activation domain are also known to form homodimers and heterodimers. In contrast to the small Maf proteins,
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5302
homodimers of the large Maf protein c-Maf is required for transcriptional activation of several genes. This includes L7 (Kurschner and Morgan, 1995), IL4 (Ho et al., 1996) and p53 (Hale et al., 2000). c-Maf-c-Myb heterodimers, however, repressed transcription of early myeloid gene during differentiation (Hegde et al., 1998). In line with c-Maf, MafB heterodimerized with Ets and repressed transcription (Sieweke et al., 1996). The Small Maf proteins were also shown to be part of nuclear protein complexes that contain Nrf1, Nrf2, Jun, Fos and Fra that binds to NQO1 gene ARE (Dhakshinamoorthy and Jaiswal, 2000; Itoh et al., 1997; Nguyen et al., 2000; Wild et al., 1999). Overexpression of small Maf (MafG and MafK) proteins lacking the transcriptional activation domain have been shown to repress transcription of NQO1 (Dhakshinamoorthy and Jaiswal, 2000), GST Ya (Nguyen et al., 2000) and g-GCS genes (Wild et al., 1999). This raised an interesting question regarding the role of the large
Maf proteins (c-Maf and MafB) that contain the transcriptional activation domain in ARE-mediated expression and induction of detoxifying enzyme genes. It remains absolutely unknown whether the large Maf (c-Maf and Maf B), containing the transcriptional activation domain bind to the ARE and activate or repress ARE-mediated transcription of the NQO1, GST Ya and g-GCS genes. It is also not known whether these large Maf proteins bind to ARE as homodimers or heterodimers with other leucine zipper proteins. In the present report, we have investigated the role of c-Maf containing the transcriptional activation domain in the expression and anti-oxidant induction of ARE-mediated NQO1 and GST Ya genes. The overexpression of c-Maf in transfected human hepatoblastoma (Hep-G2) cells led to repression of AREmediated NQO1 and GST Ya genes expression and induction in response to t-BHQ. In similar experi-
Figure 1 (a) Nucleotide sequence of NQO1 gene ARE, GST Ya ARE and p53 gene MARE. The nucleotide sequences of human NQO1 gene ARE, GST Ya ARE, Core of the ARE and mouse p53 gene MARE are shown. The TRE and TRE-like elements are indicated with arrows. (b) NQO1 gene ARE mutants. The nucleotide sequences of mutant NQO1 gene AREs are shown. Mutant ARED1 contains mutations in 3’ TRE. Mutant ARED2 contains mutations in the ‘GC’ box and mutant ARED3 contains mutations in the 5’ TRE like element. The mutated bases are shown with *. (c) Schematic map of c-Maf, c-MafD1, Nrf2 and MafG showing the various activity domains. c-MafD1 lacks 125 amino acids of transcriptional activation domain. LZ, leuzine zipper region; B, basic region; CnC, cap’n’collar region; TA, Transcriptional activation domain; H, hydrophobic region Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5303
ments, the overexpression of c-Maf resulted in transcriptional activation of MARE-mediated p53 gene. Deletion analysis indicated that transcriptional activation domain of c-Maf is required for MAREmediated expression of p53 but not for the repression of ARE-mediated NQO1 gene expression. The band and super shift results show that c-Maf bound to the ARE as homodimer and heterodimer with small Maf protein. Mutational analysis of NQO1 gene ARE revealed that core of the ARE containing 3’ TRE and ‘GC’ box is essentially required for c-Maf binding and its role in negative regulation of ARE-mediated gene expression and induction. Northern analysis revealed that c-Maf expression increased between 2 – 4 h of t-BHQ treatment. Results The nucleotide sequences of the human NQO1 and rat GST Ya gene AREs and p53 MARE are shown in Figure 1a. The NQO1 and GST Ya gene AREs contain two TPA response element (TRE) and TRE-like elements followed by a ‘GC’ box (Figure 1a). p53 gene MARE contains a single TRE-like element with conserved flanking sequences (Figure 1a). It is known
that mutations in 3’ TRE and ‘GC’ box lead to significant reduction in ARE-mediated genes expression and induction in response to xenobiotics and antioxidants (Dhakshinamoorthy et al., 2000). The mutations in the 5’ TRE like element in the ARE also reduces its efficiency to mediate gene expression (Dhakshinamoorthy et al., 2000). The structure and various domains of c-Maf, c-MafD1, Nrf2 and MafG are shown in Figure 1c. These are b-zip proteins that contain leucine zipper (LZ) domain similar to Jun and Fos. Nrf2 contains cap’n’collar domain that was found conserved in Nrf and BACH proteins (Dhakshinamoorthy et al., 2000; Blank and Andrews, 1997; Motohashi et al., 2000). c-Maf and Nrf2 also contain transcriptional activation (TA) domain rich in acidic residues (Ho et al., 1996; Hegde et al., 1998; Moi et al., 1994). MafG lacks this transcriptional activation domain (Kim and Andrews, 1997; Dhakshinamoorthy and Jaiswal, 2000; Figure 1c). The mutated c-MafD1 also lacks the transcriptional activation domain (Figure 1c). The transfection of Hep-G2 cells with NQO1 gene ARE-Luc plasmid led to ARE-mediated luciferase gene expression (Figure 2). This expression was induced 2 – 3-fold in response to t-BHQ treatment of transfected Hep-G2 cells (P40.005; Figure 2). The overexpression of c-Maf resulted in c-Maf concentration dependent
Figure 2 Effect of overexpression of c-Maf on NQO1 gene ARE and p53 gene MARE-mediated luciferase gene expression and induction in response to tert-butyl hydroquinone (t-BHQ). 0.1 mg of the reporter plasmids NQO1 gene ARE-Luc, GST Ya gene ARE-Luc and p53 MARE-Luc was co-transfected with varying concentrations of expression plasmid pcDNA-c-Maf in Hep-G2 cells in separate experiments. 0.01 mg of plasmid pRL-TK encoding Renilla luciferase was included as the internal control in each transfection. After 36 h of transfection, the cells were treated with DMSO (control) or 100 mM t-BHQ. The cells were harvested 12 h after the treatment and assayed for luciferase activity. The values represent mean+s.e. of three independent transfection experiments Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5304
Oncogene
repression of NQO1 gene ARE-mediated luciferase gene expression. The c-Maf repression of NQO1 gene ARE-mediated luciferase gene expression was highly significant. The transfection of Hep-G2 cells with 0.05 mg c-Maf plasmid resulted in greater than 80% inhibition (P40.001) of NQO1 gene ARE-mediated luciferase gene expression. Interestingly, the overexpression of c-Maf also repressed t-BHQ induction of NQO1 gene ARE-mediated luciferase activity in similar proportions as basal expression. The negative role of c-Maf was also observed with GST Ya ARE (Figure 2). In contrast to NQO1 and GST Ya AREs, the p53 gene MARE-mediated luciferase gene expression was upregulated in Hep-G2 cells overexpressing cMaf (Figure 2). However, higher concentrations of cMaf reduced the magnitude of activation of p53 MARE-mediated luciferase gene expression but did not lead to repression of basal expression. The Hep-G2 cells transfected with p53 gene MARE-Luc alone or in
combination with c-Maf plasmid did not demonstrate increase in MARE-mediated luciferase gene expression in response to treatment with t-BHQ (Figure 2). The overexpression of MafG lacking the transcriptional activation domain resulted in repression of NQO1 gene ARE- and p53 gene MARE-mediated luciferase gene expression in transfected Hep-G2 cells (Figure 3). The co-expression of c-Maf with Maf G rescued the repression of p53 gene MARE-mediated luciferase gene expression but failed to demonstrate a similar effect on NQO1 ARE-mediated gene expression (Figure 3). In fact, the co-expression of c-Maf with MafG further repressed the NQO1 ARE-mediated luciferase activity and induction in response to tBHQ. In similar experiments, overexpression of Nrf2 upregulated NQO1 and GST Ya ARE-mediated luciferase gene expression (Figure 4). However, as reported earlier, induction by t-BHQ was diminished presumably because of very high overexpression of Nrf2 (Dhakshinamoorthy and Jaiswal, 2000). The overexpression of c-Maf with Nrf2 led to c-Maf concentration dependent repression of Nrf2-mediated
Figure 3 Effect of overexpression of MafG and c-Maf on NQO1 gene ARE and p53 gene MARE-mediated luciferase gene expression and antioxidant induction. 0.1 mg of the reporter plasmid NQO1 gene ARE-Luc or p53 MARE-Luc was co-transfected with expression plasmids pcDNA-MafG and pcDNA-c-Maf in concentrations as shown in Hep-G2 cells. 0.01 mg of plasmid pRL-TK encoding Renilla luciferase was included as the internal control in each transfection. The transfected cells were treated with DMSO (Control) or 100 mM of t-BHQ 36 h after transfection. The cells were harvested 12 h after the treatment and assayed for luciferase activity. The values represent mean+s.e. of three independent transfection experiments
Figure 4 Effect of overexpression of Nrf2 and c-Maf on NQO1 and GST Ya genes ARE-mediated luciferase gene expression and antioxidant induction. 0.1 mg of the reporter plasmid NQO1 gene ARE-Luc or GST Ya ARE-Luc was co-transfected with expression plasmids pcDNA-Nrf2 and pcDNA-c-Maf in concentrations as shown in Hep-G2 cells. 0.01 mg of plasmid pRL-TK encoding Renilla luciferase was included as the internal control in each transfection. The transfected cells were treated with DMSO (Control) or 100 mM of t-BHQ 36 h after transfection. The cells were harvested 12 h after the treatment and assayed for luciferase activity. The values represent mean+s.e. of three independent transfection experiments
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5305
activation and t-BHQ induction of NQO1 as well as GST Ya ARE-mediated gene expression. The overexpression of N-terminal truncated c-MafD1 lacking the transcriptional activation domain of c-Maf repressed NQO1 gene ARE-mediated luciferase gene expression in transfected Hep-G2 cells (Figure 5a). This repression was concentration dependent and comparable to wild-type c-Maf protein (compare Figure 5a with Figure 2). The c-MafD1 also repressed Nrf2 upregulation of ARE-mediated NQO1 gene expression as observed with wild-type c-Maf (Figures 5b and 4). In similar experiment c-MafD1 lacking transcriptional activation domain failed to activate MARE-mediated p53 gene expression (Figure 5c). It also failed to rescue MafG repression of MAREmediated p53 gene expression (Figure 5d). These observations with c-MafD1 were in contrast to wildtype c-Maf protein that activated MARE-mediated p53 gene expression (Figure 2) and rescued MafG repression of p53 gene expression (Figure 3). Mutational analysis of NQO1 gene ARE revealed that mutation of 3’ TRE within the ARE (ARED1) resulted in greater than 40-fold decrease in luciferase gene expression and complete loss of induction in response to t-BHQ (Figure 6a, compare with Figure 2 NQO1 ARE-Luc). Overexpression of c-Maf had no effect on expression and induction of mutant ARED1mediated luciferase gene expression (Figure 6a). Mutations in GC box in ARED2 led to the loss of tBHQ induction of luciferase gene expression (Figure
6b). In addition, the overexpression of c-Maf had no effect on mutant ARED2-mediated gene expression (Figure 6b). In contrast to mutant ARE ^1 and ^2, overexpression of c-Maf activated the mutant ARED3 (containing mutations in 5’ TRE like element)mediated expression of luciferase gene expression (Figure 6b). However, this activation was reduced at higher concentrations of c-Maf. These results are same as observed with p53 gene MARE-mediated gene expression (Figure 2, p53 MARE-Luc). Band shift analysis with nuclear extracts from HepG2 cells demonstrated the binding of two specific complexes to the NQO1 gene ARE as reported earlier (Figure 7a; Dhakshinamoorthy and Jaiswal, 2000). The treatment of cells with t-BHQ resulted in increased binding of both the complexes to the NQO1 gene ARE (Figure 7A). However, the increase in binding was more significant in the upper band as compared to the lower band. The super shift analysis with specific c-Maf antibody demonstrated supershifting of the lower band. The super shifted band moved along with the upper band (Figure 7b). In the same experiment preimmune serum failed to supershift any band (Figure 7b). The results of the band and super shift assays with NQO1 gene ARE and in vitro translated proteins are shown in Figures 8 and 9. MafG-V5 bound to the NQO1 gene ARE as homodimer, which was super shifted with V5 antibody. c-Maf also bound to the ARE as homodimer, which was super shifted with c-
Figure 5 Effect of overexpression of mutant c-MafD1 lacking the transcriptional activation domain of c-Maf on NQO1 gene ARE and p53 gene MARE-mediated luciferase gene expression and induction. 0.1 mg of the reporter plasmid NQO1 gene ARE-Luc or p53 MARE-Luc was cotransfected with expression plasmid pcDNA-c-MafD1 alone or in combination with Nrf2 or MafG in concentrations as shown in Hep-G2 cells. 0.01 mg of plasmid pRL-TK encoding Renilla luciferase was included as the internal control in each transfection. The transfected cells were treated with DMSO (control) or 100 mM of t-BHQ 36 h after transfection. The cells were harvested 12 h after the treatment and assayed for luciferase activity. The values represent mean+s.e. of three independent transfection experiments Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5306
Maf antibody. It may be noteworthy that super shifted band with c-Maf antibody remained very close to the well. However, the disappearance of c-Maf homodimer binding with c-Maf antibody was clearly visible (Figure 8, Lane ‘c-Maf+anti-c-Maf’). Further, this band was not super shifted with pre-immune serum (Figure 6, Lane ‘c-Maf+Preimmune Serum’). In the similar experiment, Nrf2 and Nrf2+c-Maf failed to bind to the ARE. The combination of c-Maf with MafG resulted in binding of c-Maf and MafG homodimers and c-Maf-MafG heterodimers that were supershifted with respective antibodies. Since c-Mar+anti-c-Maf produced a supershifted complex that migrated close to the gel wells (lane c-Maf+anti-c-Maf), the lower supershifted complex in lane MafG-V5+c-Maf+antic-Maf formed could represent the MafG-c-Maf heterodimer and the upper shifted complexes near the wells could be c-Maf homodimers. Band shift results with NQO1 gene ARE mutants are shown in Figure 9. The c-Maf-c-Maf and mutant c-MafD1-c-MafD1 homodimers bound to the wild-type ARE and ARE containing mutations in 5’ TRE like (ARED3). The supershift
Figure 7 Band and super shift assay. (a) NQO1 gene ARE was end labeled with g-P32-ATP. 50 000 c.p.m. of the labeled ARE was incubated with 10 mg of nuclear extract from Hep-G2 cells treated with DMSO (control) or t-BHQ (100 mM) and band shift assay performed. The two specific ARE-Nuclear protein complexes were indicated by an arrow. (b) Super shift assays were performed by incubating the ARE-Nuclear protein complexes with preimmune serum or with anti-c-Maf serum by procedures as described in Materials and methods. The super shifted c-Maf band is indicated by double arrow
assays also indicated that c-Maf and c-MafD1 also bound to NQO1 gene ARE as heterodimers with MafG (Figure 8, data for c-MafD1/MafG not shown because it is exactly similar as c-Maf in Figure 8). The ARE containing mutations in the core sequences i.e. 3’ TRE (ARED1) and GC box (ARED2) failed to bind to either c-Maf or truncated c-MafD1. Northern analysis of Hep-G2 cells treated with DMSO and t-BHQ are shown in Figure 10. The treatment of Hep-G2 cells with t-BHQ led to twofold increase in c-Maf transcripts as early as 2 h. It reached a plateau (fourfold increase) at 4 h after t-BHQ treatment.
Figure 6 Effect of mutations in NQO1 gene ARE on AREmediated luciferase gene expression and induction. Mutant NQO1 gene ARE (ARED1, ARED2 and ARED3) reporter plasmid were co-transfected with different concentrations of pcDNA-c-Maf in Hep-G2 cells. 0.01 mg of plasmid pRL-TK encoding Renilla luciferase was included as the internal control in each transfection. The transfected cells were treated with DMSO (control) or 100 mM of t-BHQ for 12 h and analysed for luciferase activity. The values represent mean+s.e. of three independent transfection experiments Oncogene
Discussion ARE-mediated expression and co-ordinated induction of detoxifying enzymes are known to protect cells against electrophilic and oxidative damage due to exposure to chemicals, ionizing radiations and UV light (Dhakshinamoorthy et al., 2000). The various genes ARE bind to a complex of nuclear proteins from cells of different origins (Dhakshinamoorthy et al.,
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5307
Figure 8 Band and super shift assays. NQO1 gene ARE was end labeled with g-P32-ATP. 50 000 c.p.m. of the labeled ARE was incubated with in vitro translated MafG-V5, c-Maf, Nrf2 individually or in combinations as shown. The protein mixture was incubated at 378C for 15 min before incubation with the labeled ARE. The band shift experiment was performed at room temperature. Super shift assays were performed by incubation of band shift reaction mixtures with preimmune serum, anti-V5, anti-c-Maf and anti-Nrf2 antibodies for 2 h at 48C. The band shift and super shift mixtures were analysed on 5% non-denaturing polyacrylamide gel. The gel was dried and autoradiographed. *Denotes the non-specific band from rabbit reticulocyte lysate
2000). Analysis of ARE-nuclear protein complexes have identified several nuclear transcription factors including c-Jun, Jun-B, Jun-D, c-Fos, Fra1, Nrf1, Nrf2, YABP, ARE-BP1, MafG and MafK (Dhakshinamoorthy and Jaiswal, 2000; Li and Jaiswal, 1992; Venugopal and Jaiswal, 1996; Liu and Pickett, 1996; Vasliou et al., 1995; Yoshioka et al., 1995; Venugopal and Jaiswal, 1998; Kwong et al., 1999; Alam et al., 1999). Recently, Nrf3, a third member of the Nrf family of transcription factors was cloned and sequenced (Kobayashi et al., 1999). However, the role of Nrf3 in the regulation of ARE-mediated expression and induction of detoxifying enzyme genes remains unknown. Among the various ARE-binding transcription factors, Nrf2 has been shown to positively regulate the ARE-mediated expression and induction of NQO1, GST Ya, g-GCS and HO-1 genes in response to antioxidants and xenobiotics (Venugopal and Jaiswal, 1996; Nguyen et al., 2000; Wild et al., 1999; Alam et al., 1999). Nrf2 does not homodimerize and requires another leucine zipper protein to be active (Moi et al., 1994; Venugopal and Jaiswal, 1996). Small Maf (Maf G and MafK) proteins lacking the transcriptional activation domain of v-Maf are known to form homodimers, bind to ARE and repress detoxifying enzyme genes (Dhakshinamoorthy and Jaiswal, 2000). MafG and MafK are also known to form heterodimers
with Nrf2 and bind to ARE (Marini et al., 1997; Dhakshinamoorthy and Jaiswal, 2000; Nguyen et al., 2000; Wild et al., 1999). However, it is not clear if Nrf2-small Maf heterodimers activate or repress transcription of detoxifying enzyme genes. In this paper, the role of c-Maf in ARE-mediated gene expression was investigated. Large Maf (c-Maf and MafB) proteins are different from small Maf (MafG and MafK) proteins in that these contain the acidic residues rich transcriptional activation domain (Ho et al., 1996; Hegde et al., 1998; Kataoka et al., 1994b). The various results revealed that c-Maf negatively regulates ARE-mediated expression and anti-oxidant induction of detoxifying enzyme genes. This was clearly evident from transfection experiments. Hep-G2 cells overexpressing c-Maf repressed ARE-mediated NQO1 and GST Ya genes expression and induction in response to t-BHQ. Interestingly, the c-Maf repressed ARE-mediated luciferase activity was induced to a similar magnitude as observed with unrepressed activity. This indicated that ARE-mediated expression was inducible in presence of c-Maf and suggested that a balance between positive and negative factors regulates ARE-mediated expression of detoxifying enzyme genes. The mutational analysis of NQO1 gene ARE revealed that c-Maf binds to the core of the ARE sequence containing 3’ TRE element and GC box. Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5308
Figure 9 Band shift assays with mutant ARE and in vitro translated c-Maf and c-Maf D1. NQO1 gene mutant AREs (ARED1, ARED2 and ARED3) were end labeled with g-P32-ATP in separate experiments. 50 000 c.p.m. of the labeled mutant AREs were incubated with in vitro translated c-Maf or mutant c-MafD1 and band shift experiment performed at room temperature. The band shift mixtures were analysed on 5% non-denaturing polyacrylamide gel. The gel was dried and autoradiographed. *Denotes the non-specific band from rabbit reticulocyte lysate
Mutational studies also indicated that 5’ TRE like element of the NQO1 gene ARE is not required for binding of c-Maf and its repression effect on AREmediated NQO1 gene expression. The results further indicated that transcriptional activation domain of cMaf is not required for c-Maf binding to ARE and ARE-mediated NQO1 gene expression. Band and super shift assays indicated that c-Maf binds to the NQO1 gene ARE as homodimer and heterodimer with MafG leading to repression in ARE-mediated NQO1 gene expression. c-Maf did not bind with NQO1 gene ARE as heterodimer with Nrf2 protein. The repressive effect of c-Maf on ARE-mediated gene expression was in contrast to its role in MARE-mediated transcription of p53 gene. c-Maf is known to activate transcription of MARE-mediated p53 gene expression (Hale et al., 2000; Present Report). Deletion analysis of c-Maf indicated that transcriptional activation domain of cMaf is required for transcription of MARE-mediated p53 gene expression. These results suggest that ARE is a distinct element than MARE in its response to cMaf. The role of positive factors in the mechanism of ARE-mediated gene expression and induction is clearly evident. However, the role of the negative factors like c-Maf in ARE-mediated gene expression, is not clear. It is not clear why to have negative regulatory mechanisms along with positive regulatory mechanisms in control of ARE-mediated gene expression. One Oncogene
hypothesis is that small amounts of superoxide and related reactive species are consistently required to keep cellular defenses active (de Macario and Macario, 2000). Since activation of detoxifying enzymes and other defensive proteins leads to a significant reduction in the levels of free radicals, the cell may require negative regulatory factors to keep the levels of superoxide from falling below a certain threshold. Therefore, the ARE-binding negative factors may work in parallel to positive factors and play an important role in maintaining the basal expression of NQO1 and other detoxifying enzyme genes. In addition, they may provide a check point for the induction of detoxifying enzymes and speed up the rate of repression of detoxifying enzyme genes once the chemical/oxidative stress is removed. Therefore, any alteration in expression or activity of positive and/or negative factors could lead to alterations in ARE-mediated detoxifying enzyme gene expression. Northern analysis of Hep-G2 cells treated with anti-oxidant for different periods indicated that c-Maf expression remained unchanged during the early (1 – 2 h) activation phase of NQO1 and other detoxifying enzyme genes. However, there was an increase in c-Maf expression during the late (2 – 4 h) phase of induction. NQO1 gene transcription is known to induce in the first hour of the treatment of cells with t-BHQ (Radjendirane and Jaiswal, 1999). Hence, the c-Maf may play an important role in bringing down the expression of induced detoxifying
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5309
Figure 10 Northern analysis. The total RNA from the HepG2 cells treated with DMSO or 100 mM t-BHQ for different time intervals were prepared using the RNAeasy Minikit (Qiagen). Ten mg of total RNA was run on a 1% formaldehyde agarose gel, transferred to the membrane and hybridized with 32P-labeled full-length c-maf cDNA. The membrane was washed and exposed to X-ray film for 48 h. The probe was removed and the membrane reprobed with GAPDH cDNA. The intensity of the RNA bands were measured by Phosphoimage Analyzer and compared to calculate fold-induction in c-Maf RNA. GAPDH RNA was used to normalize the fold induction
enzyme genes to their normal levels and to maintain the normal level until challenged with oxidative/ electrophilic stress. This hypothetical model also opens a new area of studying the mechanisms of the balance of positive and negative factors in the expression and induction of ARE-mediated detoxifying enzyme genes. Based on the available information, a hypothetical model is illustrated to demonstrate the sequence of events from anti-oxidants and xenobiotics to the coordinated activation of detoxifying enzyme genes including NQO1 and GST Ya (Figure 11). A cytosolic factor, Keap1/INrf2 was identified that under normal conditions, retains Nrf2 in the cytoplasm (Itoh et al., 1999; Dhakshinamoorthy and Jaiswal, 2001). Exposure of cells to xenobiotics and anti-oxidants leads to the release of Nrf2. Nrf2 then translocates to the nucleus resulting in the activation of ARE-mediated gene expression. However, the mechanism of the release of Nrf2 and the nature and mode of action of the cytosolic factor(s) remain unknown. Recent studies have shown the role of p38 and MEK kinases in AREmediated regulation of gene expression (Yu et al, 2000a,b). However, the molecular targets for these kinases in ARE-mediated genes expression and induction remain unknown. More recent studies have
Figure 11 Model to demonstrate the role of c-Maf-c-Maf homodimer in the mechanism of signal transduction from antioxidants to the ARE-mediated detoxifying enzyme genes expression. INrf2/ Keap1, a cytosolic inhibitor of Nrf2
disputed the involvement of p38 and MEK kinases in ARE-mediated gene expression and instead showed the role of PKC in phosphorylation of Nrf2 and induction of ARE-mediated gene expression (Huang et al., 2000). Based on the previous and present reports, we hypothesize that during the early/activation phase negative regulators including Jun-Fos (Venugopal and Jaiswal, 1996), Jun-Fra (Venugopal and Jaiswal, 1996), MafG – MafG (Dhakshinamoorthy and Jaiswal, 2000), MafK – MafK (Dhakshinamoorthy and Jaiswal, 2000) and c-Maf – c-Maf (present report) remain at a level that does not affect the process of activation. However, at later phase of activation, the negative regulators like c-Maf – c-Maf increase and achieve a threshold that competes with declining positive factors for binding to the ARE. This leads to the rapid return of the detoxifying enzymes to their normal level. At present, it is not clear how positive and negative factors balance each other and what factors contribute to their balancing that leads to ARE-mediated genes expression and induction. It is likely that half life, nuclear accumulation, redox status, and/or phosphorylation/ dephosphorylation contribute to relative affinity of positive and negative factors for ARE that regulate ARE-mediated genes expression at any given time. Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5310
In conclusion, our data demonstrated that c-Maf containing the transcriptional activation domain binds to the ARE as homodimers and heterodimers with MafG. This binding leads to the repression of NQO1, GST Ya and other detoxifying enzyme genes expression and anti-oxidant induction. We further demonstrated that c-Maf expression increases during later phase of the anti-oxidant induction of NQO1 gene. This presumably is to speed up the process of bringing down the NQO1 gene expression to its normal level. Materials and methods Materials The enzymes used in this study were purchased from Life Technologies GIBCO – BRL, (Rockville, MD, USA). pcDNA3.1/V-5-His-TOPO vector and Anti-V5 Antibody were purchased from Invitrogen, (Carlsbad, CA, USA). aMEM was purchased from Life Technologies GIBCO – BRL. Effectene transfection reagent and RNAeasy Minikit were purchased from Qiagen (Valencia, CA, USA). The Nrf2 and c-Maf antibodies were purchased from Santa Cruz Biotechnology, Inc (Santa Cruz, CA, USA). t-BHQ and all other chemicals were purchased from Sigma (St. Louis, MO, USA). The pGL2 promoter plasmid containing firefly luciferase gene, internal control plasmid pRL-TK that encodes Renilla luciferase, the dual luciferase kit and the TNT T7/T3 coupled rabbit reticulocyte lysate system were obtained from Promega (Madison, WI, USA). The Hybond ECL nitrocellulose membrane, ECL Western blot analysis kit and Amplify NAMP1000 were purchased from Amersham Pharmacia (Pitscataway, NJ, USA). The oligonucleotides used in this study were synthesized at IDT technologies (Coralville, IA, USA). Plasmid construction The construction of the plasmids pGL2-NQO1 ARE-Luc, pGL2-GST Ya ARE-Luc, pcDNA-Nrf2 and pcDNA-MafGV5 are described (Dhakshinamoorthy and Jaiswal, 2000). The nucleotide sequences of NQO1 and GST Ya genes ARE and p53 gene MARE are shown in Figure 1a. The cDNA encoding c-Maf was sub-cloned into pcDNA 3.1 vector at the HindIII/XbaI sites to generate pcDNA-c-Maf. 5’ and 3’ PCR primers (5’ Primer: 5’-ACCATGGGTGGCTTCGATGGCTATGCG-3’ and 3’ Primer: 5’-TCACATGAAAAATTCGGGAGAGGAAG-3’) were used to amplify a truncated c-Maf cDNA that lacked the N-terminal 125 amino acids transcriptional activation domain of c-Maf. The 5’ PCR primer was designed to include the codon for translation initiation. The truncated c-Maf cDNA was subcloned in pcDNA3.1 to generate the plasmid pcDNA-c-MafD1. The sense and antisense oligonucleotides corresponding to the mouse p53 MARE (Hale et al., 2000) was synthesized with 5’ NheI/BglII sites respectively. The oligonucleotides were annealed, phosphorylated using T4 polynucleotide kinase and cloned at the respective sites in the pGL2 promoter vector. Similar strategies were used to generate mutations in 3’ TRE, ‘GC’ box and 5’-TRE like element of the NQO1 gene ARE to generate pGL2-ARED1 (mutations in 3’-TRE), pGL2-ARED2 (mutations in ‘GC’ box) and pGL2-ARED3 (mutations in 5’ TRE like element). The nucleotide sequences of mutated NQO1 gene ARED1, ARED2 and ARED3 are shown in Figure 1b. All the resulting plasmids were confirmed by DNA sequencing. Oncogene
Cell culture and cotransfection of reporter and expression plasmids Human hepatoblastoma (Hep-G2) cells were grown in six well monolayer cultures containing a-MEM medium supplemented with Fetal Bovine Serum (Dhakshinamoorthy and Jaiswal, 2000). The Effectene Transfection Reagent Kit from Qiagen was used to perform the transfection by procedures as described in the manufacturer’s protocol. Briefly, 0.1 mg of reporter constructs (NQO1 ARE-Luc, GST Ya ARE-Luc, p53 MARE-Luc and mutant NQO1 ARE-Luc) were mixed with different concentrations of the pcDNA expression plasmids (c-Maf, c-MafD1, Nrf2 and MafG) and transfected into Hep-G2 cells in separate experiments. The plasmid pRLTK encoding Renilla luciferase was included as the internal control in each transfection. The various plasmids were mixed with DNA Condensation Buffer and Enhancer solution from the kit and incubated at room temperature for 5 min. This was followed by addition of Effectene reagent to the mixture. This mixture was incubated for 7 min at room temperature. The DNA-Enhancer-Effectene mixture was added drop wise onto the Hep-G2 cells and incubated at 378C with 5% CO2. Forty-eight hours after the transfection, the cells were washed with 16 PBS and lysed in 16 Passive Lysis buffer from the kit. The Dual – Luciferase Reporter Assay System from Promega was used to assay the samples for luciferase activity as described in the manufacturer’s protocol. First the cell lysate was assayed for the firefly luciferase activity using 100 ml of the substrate LARII. Then 100 ml of the STOP and GLO reagent was added to quench the firefly luciferase activity and activate the renilla luciferase activity, which was also measured. The assays were carried out in a Packard luminometer and the relative luciferase activity was calculated as follows: 100 000/Activity of Renilla luciferase (in units)6Activity of Firefly Luciferase (in units). Each set of transfection was repeated minimum of three times. For induction studies, the cells were treated with 100 mM t-BHQ, dissolved in DMSO for 12 h and analysed for luciferase activity by procedures as described above. The pcDNA-c-Maf, pcDNA-c-MafD1, pcDNA-Nrf2 and pcDNA-MafG plasmids expressed correct size of c-Maf, cMafD1, Nrf2 and Maf G proteins respectively upon transfection in Hep-G2 cells as determined by Western analysis (data not shown). Western analysis also indicated that amount of cDNA derived proteins was directly proportional to the amount of plasmids used for transfection (data not shown). Gel shift/super shift assays The nuclear extracts from Hep-G2 [DMSO control and tBHQ (100 mM for 12 h) treated] cells were prepared by previously described procedures (Dhakshinamoorthy and Jaiswal, 2000; Li and Jaiswal, 1992). The in vitro transcription/translation of the plasmids encoding Nrf2, c-Maf, cMafD1 and MafG-V5 were performed using the TNT Coupled Rabbit Reticulocyte Lysate Systems (Promega) by procedures as suggested in the manufacturer’s protocol. Redivue L-[35S] Methionine was substituted for methionine in the reactions. After the coupled transcription/translation, the proteins were checked for their correct size by SDS – PAGE and Western analysis. Briefly, 5 ml of the translated proteins were resolved on a 10% SDS – PAGE, treated with Amplify solution (NAMP 100, Amersham Pharmacia) to enhance the 35S signal, dried and exposed to X-ray film. In a similar experiment, the proteins were transferred onto a Hybond ECL nitrocellulose membrane and probed with
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5311
Nrf2, c-Maf and V5 antibodies. V5 antibodies were used to detect the V5 tagged small Maf proteins. c-Maf and Nrf2 antibodies were bought from Santa Cruz Biotechnology (CA, USA). The V5 peptide and antibodies are previously described (Dhakshinamoorthy and Jaiswal, 2000). All of the in vitro translated proteins gave the expected size products. The ARE was end labeled with g-P32ATP and T4 polynucleotide kinase. The labeled ARE was incubated in separate experiments with nuclear extracts from DMSO (Control) or t-BHQ treated HepG2 cells or in vitro translated proteins in different combinations. Band and super shift assays were then performed by previously described procedures (Dhakshinamoorthy and Jaiswal, 2000; Li and Jaiswal, 1992). Briefly, ten micrograms of the nuclear extract or equimolar concentrations of in vitro translated proteins were used in the gel shift and super shift experiments. 1.5 ml of anti-V5 antibody, 3 ml of Nrf2 antibody, 3 ml of c-Maf antibody or 3 ml of the preimmune serum was used in the super shift assays. Northern analysis Hep-G2 cells were treated for 0, 1, 2, 4, 8, 16 h with DMSO or t-BHQ (100 mM) dissolved in DMSO. The cells were washed three times with ice-cold Dulbecco’s PBS without calcium and magnesium. The cells were scraped and RNA isolated using the RNeasy mini kit from Qiagen, Inc. RNA was treated with DNase I to remove any DNA contamination, cleaned with phenol/chloroform, and precipitated with
ethanol by standard procedures. Ten micrograms of total RNA were run on a 1% formaldehyde agarose gel and blotted by procedures described (Radjendirane and Jaiswal, 1999). cDNAs encoding c-Maf and GAPDH were labeled in separate reactions using Prime-a-Gene labeling kit (Promega Corp.) and used as probe to hybridize the RNA blots. Prehybridization and hybridization were done according to a method described previously (Radjendirane and Jaiswal, 1999).
Abbreviations NQO1, NAD(P)H : Quinone Oxidoreductase1; GST Ya, Glutathione S-transferase Ya subunit; g-GCS, g-glutamylcysteine synthetase; ARE, Antioxidant Response Element; hARE, human NQO1 gene ARE; t-BHQ, tert-Butyl Hydroquinone; ROS, Reactive Oxygen Species; Hep-G2, Human hepatoblastoma cells
Acknowledgments We are grateful to Drs Jefferson Y Chan and Yuet W Kan, both from University of California, San Francisco for providing us the cDNAs encoding Nrf2 and Nrf1. We are also thankful to Dr Linda Shapiro (St. Judes Children’s Hospital, Memphis, TN, USA) for the gift of c-Maf cDNA. This investigation was supported by NIH grant GM47466.
References Alam J, Stewart D, Touchard C, Boinapally S, Choi MK and Cook JL. (1999). J. Biol. Chem., 274, 26071 – 26078. Blank V and Andrews NC. (1997). Trends Biol. Sci., 22, 437 – 441. de Macario EC and Macario AJ. (2000). Front. Biosci., 5, D780 – D786. Dhakshinamoorthy S, Long II DJ and Jaiswal AK. (2000). Curr. Top. Cell. Regul., 36, 201 – 216. Dhakshinamoorthy S and Jaiswal AK. (2000). J. Biol. Chem., 275, 40134 – 40141. Dhakshinamoorthy S and Jaiswal AK. (2001). Oncogene, 20, 3906 – 3917. Fujiwara KT, Ashida K, Nishina H, Iba H, Miyajima N, Nishizawa M and Kawai S. (1993). Oncogene, 8, 2371 – 2380. Hale TK, Myers C, Maitra R, Kolzau T, Nishizawa M and Braithwaite AW. (2000). J. Biol. Chem., 275, 17991 – 17999. Huang HC, Nguyen T and Pickett CB. (2000). Proc. Natl. Acad. Sci. USA, 97, 12475 – 12480. Itoh K, Chiba T, Takahashi S, Ishii T, Igarashi K, Katoh K, Oyake T, Hayashi N, Satoh K, Hatayama I, Yamamoto M and Nabeshima Y. (1997). Biochem. Biophys. Res. Commun., 236, 313 – 322. Hegde SP, Kumar A, Kurschner C and Shapiro LH. (1998). Mol. Cell. Biol., 18, 2729 – 2737. Ho IC, Hofge MR, Rooney JW and Glimcher LH. (1996). Cell, 85, 973 – 983. Itoh K, Wakabayashi N, Katoh Y, Ishii T, Igarashi K, Engel JD and Yamamoto M. (1999). Genes Dev., 13, 76 – 86. Kataoka K, Nishizawa M and Kawai S. (1993). J. Virol., 67, 2133 – 2141. Kataoka K, Noda M and Nishizawa M. (1994a). Mol. Cell. Biol., 14, 700 – 712.
Kataoka K, Fujiwara KT, Noda K and Nishizawa M. (1994b). Mol. Cell. Biol., 14, 7581 – 7591. Kim MJ and Andrews NC. (1997). Blood, 89, 3925 – 3935. Kobayashi A, Ito E, Toki T, Kogame K, Takahashi S, Igarashi K, Hayashi N and Yamamoto M. (1999). J. Biol. Chem., 274, 6443 – 6452. Kurschner C and Morgan JI. (1995). Mol. Cell. Biol., 15, 246 – 254. Kwong M, Kan YW and Chan JY. (1999). J. Biol. Chem., 274, 37491 – 37498. Li Y and Jaiswal AK. (1992). J. Biol. Chem., 267, 15097 – 15104. Liu S and Pickett CB. (1996). Biochemistry, 35, 11517 – 11521. Marini MG, Chan K, Casula L, Kan YW, Cao A and Moi P. (1997). J. Biol. Chem., 272, 16490 – 16497. Matsushima-Hibiya Y, Nishi S and Sakai M. (1998). Biochem. Biophys. Res. Commun., 245, 412 – 418. Moi P, Chan K, Asunis I, Cao A and Kan YW. (1994). Proc. Natl. Acad. Sci. USA, 91, 9926 – 9930. Motohashi H, Katsuoka F, Shavit JA, Engel JD and Yamamoto M. (2000). Cell, 103, 865 – 875. Nguyen T, Huang HC and Pickett CB. (2000). J. Biol. Chem., 275, 15466 – 15473. Radjendirane V and Jaiswal AK. (1999). Biochem. Pharmacol., 58, 597 – 603. Sieweke M, Tekotte H, Frampton J and Graf T. (1996). Cell, 85, 49 – 60. Vasliou V, Puga A, Chang C, Tabor MW and Nebert DW. (1995). Biochem. Pharm., 12, 2057 – 2068. Venugopal R and Jaiswal AK. (1996). Proc. Natl. Acad. Sci. USA, 93, 14960 – 14965. Venugopal R and Jaiswal AK. (1998). Oncogene, 17, 3145 – 3156. Oncogene
Role of c-Maf in ARE-mediated gene expression S Dhakshinamoorthy and AK Jaiswal
5312
Wild AC, Moinova HR and Mulcahy RT. (1999). J. Biol. Chem., 274, 33627 – 33636. Yoshioka K, Deng TL, Cavigelli M and Karin M. (1995). Proc. Natl. Acad. Sci. USA, 92, 4972 – 4976.
Oncogene
Yu R, Mandlekar S, Lei W, Fahl WE, Tan TH and Kong ANT. (2000a). J. Biol. Chem., 275, 2322 – 2327. Yu R, Chen C, Mo YY, Hebbar V, Owuor ED, Tan TH and Kong ANT. (2000b). J. Biol. Chem., 275, 39907 – 39913.
