Aryl Hydrocarbon Receptor Nuclear Translocator and Upstream Stimulatory Factor Regulate Cytochrome P450 2a5 Transcription through a Common E-box Site

doi:10.1016/j.jmb.2007.03.075

J. Mol. Biol. (2007) 369, 640–652

Aryl Hydrocarbon Receptor Nuclear Translocator and Upstream Stimulatory Factor Regulate Cytochrome P450 2a5 Transcription through a Common E-box Site Satu Arpiainen 1 , Virpi Lämsä 1 , Olavi Pelkonen 1 , Sun Hee Yim 2 Frank J. Gonzalez 2 and Jukka Hakkola 1 ⁎ 1

Department of Pharmacology and Toxicology, University of Oulu, Oulu, Finland 2

Laboratory of Metabolism, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA

The aryl hydrocarbon receptor nuclear translocator (ARNT) belongs to the basic-helix-loop-helix (bHLH) transcription factors and regulates several genes as heterodimers with other bHLH proteins. ARNT is also able to homodimerize, but no mammalian target genes for the homodimer have been shown. We identified a palindromic E-box element in the 5′ regulatory region of the murine cytochrome P450 (Cyp) 2a5 gene that was found to be important for Cyp2a5 transcription in primary hepatocytes, and was found by chromatin immunoprecipitation assays to interact with ARNT. Electrophoretic mobility-shift assay experiments with in vitro translated ARNT showed binding without heterodimerization partner, indicating binding as a homodimer. Transfection studies in wild-type and ARNT-deficient Hepa-1 cells revealed that ARNT expression is necessary for full activity of the Cyp2a5 promoter. In the liver-specific Arnt-null mouse line, the level of hepatic CYP2A5 mRNA was decreased significantly. Co-transfection studies with an ARNT expression vector lacking the transactivation domain (TAD) demonstrated that the ARNT TAD is needed for Cyp2a5 activation, which suggests that ARNT transactivates Cyp2a5 as a homodimer. In primary hepatocytes, the mRNA levels of both CYP2A5 and ARNT splice variant 1 were increased during cultivation. Upstream stimulatory factors 1 and 2a were also able to bind to the same E-box as ARNT, indicating that there may be competition for DNA binding between these factors. Indeed, the upstream stimulatory factors activated the Cyp2a5 promoter through the E-box only in the presence of hepatocyte nuclear factor-4α, while ARNT transactivation was independent of hepatocyte nuclear factor-4α. In conclusion, these results indicate that ARNT controls Cyp2a5 transcription and thus, for the first time, suggest active involvement of the ARNT homodimer in mammalian gene regulation. © 2007 Elsevier Ltd. All rights reserved.

*Corresponding author

Keywords: ARNT; homodimer; USF; CYP2A5; regulation

Introduction The aryl hydrocarbon receptor nuclear translocator (ARNT) is a ubiquitously expressed nuclear Abbreviations used: ARNT, aryl hydrocarbon receptor nuclear translocator; bHLH, basic-helix-loop-helix; PAS, Per-Arnt-Sim; AHR, aryl hydrocarbon receptor; HIF, hypoxia inducible factor; E-box, enhancer box; USF, upstream stimulatory factor; NF, nuclear factor; HNF, hepatocyte nuclear factor; TAD, transactivation domain. E-mail address of the corresponding author: [email protected]

protein that belongs to the basic-helix-loop-helix (bHLH) class of transcription factors.1,2 It also has a second region of homology, termed the Per-ArntSim (PAS) domain. The ARNT protein is a wellknown heterodimerization partner for several members of the bHLH transcription factor family: The most extensively studied include the aryl hydrocarbon receptor (AHR), which responds to exogenous ligands such as dioxins,2 and hypoxia-inducible factor (HIF) 1α, which is regulated by cellular oxygen tension.3 In addition to heterodimerizing with a range of bHLH proteins, ARNT is able to homodimerize and bind the palindromic enhancer box (E-box) sequence CACGTG.4–7 ARNT homo-

0022-2836/$ - see front matter © 2007 Elsevier Ltd. All rights reserved.

641

ARNT and USF Regulate Cyp2a5 Transcription

dimer has been shown to activate transcription of reporter genes from the E-box located within the adenovirus major late promoter in mammalian cells in culture,4 but to date, no mammalian target genes for the ARNT homodimer have been reported and thus the in vivo relevance of its function is not clear. The upstream stimulatory factors (USFs) 1 and 2 are members of the bHLH-leucine zipper transcription factor family.8–10 They are both ubiquitously expressed, but distinct cell-specific ratios of homo- and heterodimers are found in different cell types.10,11 USFs are involved in stress and immune response, in cell-cycle control and in glucose and lipid metabolism.12 They were first identified because of their involvement in transcription of the adenovirus major late gene and they are able to bind to the palindromic E-box site both as homo- and heterodimers.13 The bHLH transcription factors bind variations of the core E-box sequence CANNTG. HLH domains mediate protein dimerization and the basic region binds to DNA. Dimerization of the bHLH factors is necessary, as two basic regions are required for the DNA-binding activity. When binding to DNA, each component of the dimers of the bHLH proteins recognizes three base-pairs on either side of the core sequence.14 bHLH transcription factors are divided into different classes according to the DNA half-sites they bind to. Binding site sequences depend on the basic region amino acids. Class B proteins, including ARNT and USF, have lysine or histidine at site 5 and arginine at site 13 of their basic region that recognizes 3′ half-site GTG.5,15 It is not known exactly how the target gene preference of dimers binding to the same E-box core sequence is determined. Binding affinity is shown to be at least partially dependent on flanking nucleotides on both sides of the E-box,16 but the transcriptional activity may be determined by other factors, such as concentrations or modifications of the dimer proteins and their co-regulators in the cell. The murine cytochrome P450 (CYP) 2A5 enzyme and its human orthologue CYP2A6 metabolize many toxic xenobiotic substrates, such as nitrosamines and aflatoxins.17,18 They are also the main metabolizers of nicotine,19,20 and according to the results reported by Abu-Bakar et al.,21 CYP2A5 participates in the degradation of bilirubin, a breakdown product of heme. CYP2A5 and CYP2A6 are expressed predominantly in hepatocytes, but are present also in some extrahepatic tissues, especially nasal mucosa.22–24 Cyp2a5 is regulated both at the transcriptional and post-transcriptional levels,25 and besides induction by several structurally different xenobiotics, CYP2A5 activity is shown to be upregulated in different pathophysiological states of the liver, such as tumors and inflammation caused by hepatitis B virus.26 We have shown earlier that hepatocyte nuclear factor (HNF)-4 plays a major role in the constitutive regulation of Cyp2a5 together with nuclear factor (NF)-I.27 HNF-4 and NF-I regulate the transcription of Cyp2a5 through the proximal region (−271/− 47) of the 5′ promoter. Another major activation region

of the Cyp2a5 promoter is located further upstream (− 3033/− 2014) from the transcription start site. Here, we show that this area includes a palindromic E-box site that is responsible for transcriptional activation of the region and is bound by both the ARNT and USFs. We present data indicating that the ARNT activates transcription of the Cyp2a5 gene without the need for a heterodimerization partner, and thus suggest active involvement of the ARNT homodimer in mammalian gene regulation.

Results Identification and functional analysis of the Cyp2a5 5′ E-box site In our previous study, regulation of the Cyp2a5 transcription was investigated by transfection of the Cyp2a5 promoter-luciferase reporter constructs to mouse primary hepatocytes. The results indicated that there are two areas in the Cyp2a5 5′ flanking region that activate transcription, the proximal promoter and a more distal region from − 3033 to − 2014. We could show that the proximal promoter contains binding sites for HNF-4 and NF-I that have a crucial role in the constitutive regulation of Cyp2a5 and transcriptional activation by the proximal promoter.27 We used the MatInspector professional program and Genomatix matrices† to search for transcription factor binding sites at the − 3033 to − 2014 area of the Cyp2a5 gene 5′ flanking region. A putative palindromic E-box element (CACGTG) was identified at the Cyp2a5 5′ − 2403 to − 2398 region. The functionality of the E-box and the significance of the palindromic structure was tested by sitedirected mutagenesis of the element in the Cyp2a5 5′ − 3033/+ 10-Luc construct. The C and G nucleotides in the middle of the element were changed to G and C, respectively (CACGTG → CAGCTG). The wildtype and mutated constructs were transfected into mouse primary hepatocytes. The mutation of the E-box decreased luciferase activity to 20% of that of the wild-type promoter (Figure 1), indicating that the E-box element is important in the regulation of the Cyp2a5 5′ promoter and that the palindromicity of the element is crucial. ARNT, USF-1 and HNF-4α bind to the Cyp2a5 promoter in vivo The symmetric E-box is known to be preferably bound by so-called class B bHLH proteins having certain amino acids in their basic region.5 Of these factors, we considered ubiquitously expressed ARNT and USF as the main candidates involved in regulation of the Cyp2a5 gene. A chromatin immunoprecipitation method was used to study Cyp2a5 promoter interacting transcription factors in the native chromatin structure. We used antibodies † http://www.genomatix.de

642

ARNT and USF Regulate Cyp2a5 Transcription

Figure 1. Significance of the E-box site on Cyp2a5 promoter activity. Cyp2a5 5′ −3033/+10-Luc and Cyp2a5 5′ −3033/+10-Luc with a mutated E-box site (core sequence CACGTG → CAGCTG) were transfected into DBA/2 mouse primary hepatocytes. The activities produced by the studied constructs were normalized against the co-transfected control plasmid (pRL-TK) activities. The values represent the means + SD of four individual samples. The activity of the mutated construct was compared with that of the non-mutated construct. *** The reduction in activity of the mutated E-box was statistically significant, p < 0.001. The experiment was repeated twice with similar results.

against ARNT, USF-1 and, as a control, HNF-4α to precipitate the protein–DNA complexes from murine primary hepatocytes. Two primer pairs were selected to test the binding of the factors to the promoter: one amplified region including the distal E-box and the other contained the proximal promoter HNF-4α binding site. Negative control primers amplifying the region between E-box and HNF-4α binding site were used to test that DNA is fragmented properly by sonication. No binding of ARNT, USF-1 or HNF-4α to the negative control region was seen. A schematic picture of the binding sites is shown in Figure 2(a). HNF-4α was found to bind on the proximal promoter, as expected on the basis of our previous studies.27 Both ARNT and USF1 interacted with the E-box sequence containing region (Figure 2(b)), but their quantitative involvement in the regulation of Cyp2a5 could not be compared on the basis of these results, since the

affinities of the antibodies used in immunoprecipitation are not identical. ARNT binds to the Cyp2a5 5′ E-box without a heterodimerization partner We used an electrophoretic mobility-shift assay to examine the ARNT binding to the Cyp2a5 promoter E-box. In vitro translated ARNT protein was incubated together with the labeled Cyp2a5 5′ E-box element and the complexes formed were separated on a non-denaturing polyacrylamide gel. ARNT alone, without a heterodimerization partner, was able to form a single complex. Since it is well established that ARNT is able to form a homodimer and bind to a palindromic E-box,4–7 this result indicates that ARNT binds to the Cyp2a5 5′ E-box as a homodimer. In the supershift assay, the ARNT antibody shifted the complex formed by the in vitro

Figure 2. In vivo binding of ARNT, USF-1 and HNF-4α to the Cyp2a5 5′ promoter. In chromatin immunoprecipitation assays, ARNT, USF-1 and HNF-4α antibodies (ab:s) were used to precipitate fixed DNA–protein complexes from murine primary hepatocytes. Extracted DNA fragments were amplified with specific primers in real-time PCR. (a) A schematic picture of the Cyp2a5 5′ promoter and the sites of the amplified DNA fragments. (b) The relative amounts of DNA copies amplified with Cyp2a5 5′ E-box and HNF-4 response element (RE) primers. The relative quantity of DNA was counted by comparing the sample fluorescence to the fluorescence values measured from total chromatin input dilution series. Rabbit IgG immunoprecipitation was used as a negative control. The experiment was repeated twice with similar results.

643

ARNT and USF Regulate Cyp2a5 Transcription

translated protein (Figure 3(a)). Similar to earlier findings,5 the anti-ARNT antibody seems to be able to stabilize the homodimer. The specificity of the binding was tested with competition reactions with unlabeled Cyp2a5 5′ E-box, consensus ARNT homodimer response element and mutated Cyp2a5 5′ Ebox site. Both the unlabeled E-box itself and the consensus binding site completely competed away the complex, while the mutated E-box affected binding only in high concentrations with 100-fold excess (Figure 3(b)). Thus, the ARNT is able to bind to the Cyp2a5 promoter E-box most likely as a homodimer. ARNT expression is necessary for Cyp2a5 promoter activity We examined the functional role of ARNT in the Cyp2a5 regulation by using mouse hepatoma cell lines Hepa-1 and its mutant strain deficient in ARNT. The Cyp2a5 5′− 3033/+10-Luc plasmid or the Cyp2a5

5′− 3033/+ 10-Luc with the mutated E-box site were transfected into wild-type or ARNT-defective Hepa1 cells. The mutation of the E-box site diminished the Cyp2a5 promoter mediated transcription by 60% in wild-type cells, but it did not have a significant effect in ARNT-defective cells. In addition, co-transfection of the ARNT expression vector to ARNT-deficient cells increased the Cyp2a5 promoter activity 2.5-fold, but it did not affect the activity of the reporter gene when the E-box site of the promoter was mutated (Figure 4). These results indicate that ARNT contributes to the regulation of Cyp2a5 transcription via the distal E-box site. ARNT transactivation domain is needed for regulation of Cyp2a5 transcription ARNT has a potent constitutively active transactivation domain (TAD) within its C terminus.28 Many domain analyses have revealed that the TAD of

Figure 3. Interaction of ARNT with the Cyp2a5 5′ E-box sequence in electrophoretic mobility-shift assay. (a) Binding of in vitro translated ARNT to the Cyp2a5 5′ E-box. Lane 1 is the binding reaction with no protein and lanes 2 and 3 represent the control reactions with the rabbit reticulosyte lysate (RRL) system only, without expression vector or with empty expression vector (pSP64), respectively. Lane 4 shows the reaction with in vitro translated ARNT protein (the retarded complex is indicated with an arrow on the left side of the Figure) and lane 5 represents the anti-ARNT antibodysupershifted DNA–protein complex (an arrow on the right side of the Figure). An unspecific complex present in the control reactions disappears when ARNT protein is added to the reaction. (b) The specificity of the binding tested with competition reactions. Lane 1 indicates the binding reaction with no protein, and lane 2 represents the reaction with in vitro translated ARNT protein (the retarded complex is indicated with an arrow on the left side of the Figure). Lanes 3–11 are competition reactions with the Cyp2a5 5′ E-box, the consensus ARNT response element and the mutated Cyp2a5 5′ Ebox. Fivefold, tenfold or 100-fold excess of the unlabeled oligonucleotides were used to compete with the labeled Cyp2a5 5′ E-box for the binding of in vitro translated ARNT protein as indicated.

644

ARNT and USF Regulate Cyp2a5 Transcription

Figure 4. Effect of E-box mutation and ARNT co-transfection on transcriptional activity of Cyp2a5 5′ promoter in Hepa-1 cell lines. The Cyp2a5 5′ −3033/+10 –Luc reporter construct or the reporter construct with mutated E-box was transfected into Hepa-1 wild-type (WT) and ARNT-defective (ARNT–) Hepa-1 cells. The ARNT expression vector was co-transfected with the reporter constructs into ARNT-defective cells. After transfection, the cells were incubated for 48 h, after which luciferase activity was measured. The activity was normalized against co-transfected control plasmid (pRLTK) activity. The values represent the means + SD of four individual samples. *** The difference between activity produced by mutated and non-mutated reporter gene constructs in wild-type cells and the difference of the ARNT co-transfection to the control with empty expression vector in ARNT-defective Hepa-1 cells were statistically significant (p < 0.001). The experiment was repeated, and the two independent experiments gave similar results.

ARNT is not essential for the transcriptional activity of the AHR-ARNT and HIF1α-ARNT heterodimers.1,29–33 Instead, AHR and HIF-1α act as the main transcriptional activators of the heterodimers by using their own TADs. ARNT functions as an active transcriptional activator using its TAD at least in association with itself as a homodimer,34 and in heterodimers with SIM135 or the HIF-1α variant HIF-1α417.36 To assess the importance of ARNT TAD in the regulation of the Cyp2a5 gene, we transiently transfected the truncated ARNT expression vector lacking the C-terminal transactivation domain (ARNT ΔC) together with the Cyp2a5 5′ promoter reporter constructs (Cyp2a5 5′− 3033/+ 10-Luc or Cyp2a5 5′−3033/+10-Luc with mutated E-box) into COS-1 cells. It has been reported that the E-box motif mediates transactivation in COS-1 cells only after cotransfection of ARNT or USF expression vectors, indicating that these transcription factors are not constitutively expressed in COS-1 cells.37 The fulllength ARNT, but not the TAD-deletion mutant, increased the activity of the Cyp2a5 5′ promoter 4.8fold in COS-1 cells, while the reporter plasmid with the mutated E-box site was not activated (Figure 5(a)). This indicates that ARNT is able to activate Cyp2a5 transcription through the Cyp2a5 5′ E-box site, and that ARNT TAD is needed for this activation. Similar results were obtained when transfecting ARNT and the ARNT TAD deletion construct into Hepa-1 ARNT-defective cells (data not shown). USF-1 and USF-2a activate the Cyp2a5 5′ promoter by an HNF-4α1 dependent mechanism The ability of USF-1 and USF-2a to activate the Cyp2a5 promoter was studied by co-transfection of them into COS-1 cells. No upregulation of Cyp2a5 promoter-mediated luciferase activity was seen by these factors. Several recent reports suggest that

USFs may act together with HNF-4.38–40 As Cyp2a5 is known to be regulated by HNF-4, we cotransfected the HNF-4α1 expression vector together with USFs. Synergic induction of Cyp2a5 promoter activity was seen with both USFs together with HNF-4α1: both USF-1 and USF-2a separately and together increased the transcriptional activity from 2.4-fold to 3.3-fold when compared to the HNF-4α1 co-transfection only (Figure 5(b)). In contrast, transcriptional activation by ARNT was not dependent on HNF-4 and no synergy was seen in the HNF-4α1/ARNT co-transfection studies (results not shown). When ARNT, USFs and HNF-4α were all co-transfected together, no significant change was detected compared with ARNT alone or USFs and HNF-4α together (data not shown). CYP2A5 mRNA level increases together with ARNT mRNA in murine hepatocyte cultures CYP2A5 mRNA level was found to be increased during the first day of cultivation of the primary hepatocytes. This could be a result of changes in expression of Cyp2a5 transactivating factors such as ARNT or USF. The mRNA levels of CYP2A5, ARNT, USF-1 and HNF-4α were measured by quantitative PCR at 0 h and 24 h time-points after liver perfusion and normalized against the 18 S control levels. The levels of CYP2A5, ARNT and HNF-4α mRNA increased 4.1-fold, 2.3-fold and 1.5-fold, respectively, during 24 h of cultivation. No change was seen in the level of USF-1 (Figure 6(a)). The primers used for ARNT quantification detected two ARNT splicing variants.41 We could see that only the level of the transcript variant 1 including an additional exon 5 was elevated during cultivation of the hepatocytes (Figure 6(b)). Similar increases in CYP2A5 (3.3-fold) and ARNT (2.5–fold) mRNAs were seen with Northern blot analysis (data not shown).

645

ARNT and USF Regulate Cyp2a5 Transcription

and control mice. The CYP2A5 protein expression was measured in the livers of the Arnt-null (n = 6) and control mice (n = 4) with immunoblotting. The average, hepatic, immunoreactive CYP2A5 protein level was lower in the Arnt-null mice compared with the control mice (mean 44% of the control) but there was a large variation in both groups and the difference was not statistically significant (data not shown).

Discussion

Figure 5. Effect of ARNT, ARNTΔC, USF and HNF-4α co-transfection on the function of the Cyp2a5 5′ E-box site in COS-1 cells. (a) The expression vectors of ARNT and the ARNT deletion mutant lacking the C-terminal transactivation domain (ARNTΔC) or (b) USF-1, USF-2a and HNF4α1 expression vectors were co-transfected with the Cyp2a5 5′ − 3033/+ 10 –Luc reporter plasmid or the reporter plasmid with mutated E-box into COS-1 cells. After 48 h, the cells were harvested and the luciferase activities measured. The activity produced by the studied constructs was normalized against the co-transfected control plasmid (pRL-TK) activity. The values represent means + SD of four individual samples. ### The difference from the control with empty expression vector, and *** the difference from the HNF-4α1 co-transfection are statistically significant (p < 0.001). The experiments were repeated twice with similar results.

Ten years ago, it was shown that, in addition to heterodimers, ARNT can form homodimers that are able to bind palindromic E-box sequences,4,5,34 but the physiological significance of these homodimers remains ambiguous. Recently, Wang et al. identified potential ARNT target genes using DNA microarray analysis after a reintroduction of ARNT expression in ARNT-deficient Hepa-1 cells.42 Although the mechanisms of regulation of these genes by ARNT are not known, many of them contained E-boxes in their promoter regions, suggesting possible involvement of the ARNT homodimer. In the current study, we present results suggesting that Cyp2a5 is a physiological mammalian target gene for the ARNT homodimer. We identified a symmetric Ebox sequence in the distal promoter of the Cyp2a5 gene that was necessary for Cyp2a5 transcriptional

Hepatic CYP2A5 mRNA level decreases in the liver-specific Arnt-null mouse To investigate in vivo the effect of ARNT on CYP2A5, we used liver-specific Arnt-null mouse. Cre-recombinase under control of the albumin promoter (AlbCre) and an Arnt-floxed allele were combined to produce Arnt(fl/fl)-AlbCre mice, that lack ARNT expression in the liver. The mRNA levels of ARNT and CYP2A5 were measured from the liver and kidney of the Arnt-null mice and compared with equivalent mice lacking the AlbCre transgene. mRNAs were normalized against the 18 S levels. In the livers of the Arnt-null mice, the level of the ARNT was almost totally abolished and the amount of CYP2A5 mRNA was significantly (over 50%) lower when compared with the control mice (Figure 7). No significant change was seen in ARNT or CYP2A5 mRNA levels between the kidneys of the Arnt-null

Figure 6. Expression of CYP2A5, HNF-4α, ARNT and USF-1 mRNAs in murine primary hepatocyte cultures at different time-points. (a) Quantitative PCR analysis from hepatocytes cultured 0 or 24 h after liver perfusion. Bars represent relative RNA levels normalized with 18 S control by using the comparative CT method. Values are means of two individual culture samples and three parallel Q-PCR analyses of each sample. (b) Separation of ARNT Q-PCR products on agarose gel. Two splicing variants were detected, as indicated with the arrows on the left side of the Figure. The difference from the 0 h sample is statistically significant (*** p < 0.001 or * p < 0.05).

646

Figure 7. The levels of ARNT and CYP2A5 mRNA in the liver and kidney of the liver-specific Arnt-null mouse line. mRNA levels were measured by quantitative PCR analysis from liver and kidney tissues of Arnt(fl/fl)-AlbCre (liver-specific Arnt-null) mice and mice lacking the AlbCre transgene (control). Bars represent relative mRNA levels normalized with 18 S control by using the comparative CT method. Values are means of four (control) or six (liverspecific Arnt-null) individual tissue samples and three parallel Q-PCR analyses of each sample + SD. The difference from the control samples is statistically significant (*** p < 0.001 or * p < 0.05).

activation in mouse primary hepatocytes. By chromatin immunoprecipitation, we showed that this Ebox interacts with ARNT, and the electrophoretic mobility-shift assay with in vitro produced ARNT revealed that the DNA binding does not require a heterodimerizing partner. Mutation of the E-box downregulated Cyp2a5 promoter activity in wildtype Hepa-1 cells, but not in a mutant, variant cell line deficient in ARNT suggesting that ARNT plays a key role in Cyp2a5 E-box mediated transactivation. Co-transfection of ARNT expression vector into ARNT-deficient Hepa-1 cells activated the Cyp2a5 promoter E-box-dependently. The physiological importance of the ARNT factor on maintenance of Cyp2a5 expression in liver was studied in a mouse line with liver-specific knockout of ARNT. The CYP2A5 mRNA level was significantly lower in the liver of mice lacking ARNT expression than in those of the control group. Cyp2a5 liver expression was down-regulated but not abolished by the ARNT deficiency, indicating that this transcription factor is contributing to, but not critical for, Cyp2a5 expression. This is in agreement with results from the reporter gene studies using the E-box-mutated Cyp2a5 promoter construct. Co-transfection studies with an ARNT expression vector lacking the transactivation domain showed that the ARNT TAD is needed for the activation of the Cyp2a5 promoter by ARNT. ARNT heterodimers requiring ARNT TAD for function, such as ARNTSIM, bind typically to the asymmetric E-boxes.5 The symmetric E-box is preferably bound by class B bHLH proteins, which are not known to form heterodimers with ARNT. In addition, the well-

ARNT and USF Regulate Cyp2a5 Transcription

known heterodimerization partners of ARNT, AHR and HIF-1α, are retained in the cytosol or degraded rapidly, respectively, without a stimulus outside the cell, while in our experiments the cell cultures were not treated with known AHR ligands or hypoxiainducing factors. Therefore, on the basis of previous information on binding specificities and exclusion of known heterodimerization partners, the results suggest strongly that ARNT regulates Cyp2a5, as a homodimer in this context. Yet, the involvement of heterodimers with unknown partners could not be rigorously excluded. In addition to ARNT, USF-1 and USF-2a were able to bind to the same E-box sequence. It is likely that ARNT and USF respond to different signals and compete on binding to the Cyp2a5 E-box depending on the physiological conditions in the cell. ARNT, USF-1 and USF-2a were all able to activate Cyp2a5 promoter activity through the E-box. However, USFs essentially required cooperation with HNF4α, while no synergy in transactivation was seen between ARNT and HNF-4α. Co-operation of USF and HNF-4 has been described,38,39 but Pastier et al. reported that they could not demonstrate physical interaction between USF and HNF-4 by immunoprecipitation.40 In the liver-specific Arnt-null mice, the CYP2A5 mRNA level was decreased significantly, suggesting that USF cannot fully compensate the role of ARNT. CYP2A5 is shown to be induced by fasting,43 and by glucagon through a cyclic adenosine monophosphate (cAMP)-mediated pathway, 44 suggesting that energy supply could be involved in CYP2A5 regulation also in cultured hepatocytes. Both USF and ARNT have been linked to the regulation of glucose and lipid metabolism. E-box sequences are shown to be involved in the glucose response of several genes and USFs mediate this response in many cases.45 For example, insulin, glucagon receptor, glucokinase and fatty acid synthase are regulated by USFs.46–49 Co-operative binding of USF and HNF-4 drives the transcription of the human apolipoprotein A-II and C-III genes.39,40 In the L-type pyruvate kinase promoter, HNF-1, NF-I, HNF-4 and USF binding sites are found, and HNF4-USF co-operation confers glucose responsiveness of the enzyme.38 Hepatic glucose-6-phosphatase (G6Pase) is induced by fasting and repressed by refeeding and insulin.50 Carrière et al. reported that in intestinal Caco-2/TC7 cells, maximal glucosedependent response of the G6Pase gene requires the distal (− 1155/− 1133) promoter region, which binds ARNT, cAMP-responsive element-binding protein (CREBP) and USFs.51 Gunton et al. found that loss of ARNT mediates altered gene expression and pancreatic islet dysfunction in human type 2 diabetes. 52 HNF-4α is one of the genes that appeared to be regulated by ARNT. No dimerization partner for ARNT was specified in the studies reported by Carrière and co-workers and Gunton and co-workers. Whether ARNT and/or USF are involved in the Cyp2a5 fasting response remains to be determined.

647

ARNT and USF Regulate Cyp2a5 Transcription

Cyp2a5 displays circadian expression in mouse liver. Transcription factor D-site binding protein (DBP) is known to contribute to the circadian regulation of CYP2A5. However, the circadian rhythm in CYP2A5 expression is not totally abolished in dbp−/−, mice indicating that some other factor is involved in this regulation.53 ARNT isoform brain and muscle ARNT-like 1 (BMAL1) is a well known circadian regulator that forms a heterodimer with the Clock protein and activates transcription through symmetric E-box elements. 54 ARNT itself has been reported to have a circadian rhythm in rat liver and lung and in mouse bone marrow,55,56 and therefore, it is possible that ARNT plays a role in the circadian regulation of the Cyp2a5 gene. In a previous study, we found that AHR-ARNT heterodimer is also able to regulate Cyp2a5 expression in response to AHR ligands via a xenobiotic response element (XRE, − 2514/− 2492).57 Together with the results of the present study, this indicates that ARNT plays a dual role in the regulation of Cyp2a5: constitutive regulation through the E-box as a homodimer and induction by environmental stimulus through the XRE as a heterodimer with AHR. The Cyp2a5 5′ E-box and the XRE are separated by only 100 bp. Whether there is functional interaction between these two regulatory regions is not known. Human CYP1A2 has also a distal enhancer region including XRE, E-boxes and AP-1 binding sites.58 Furthermore, similar to Cyp2a5, NF-I has been shown to participate in hepatic expression of CYP1A2.59 These two AHR-responsive genes may have several similarities in their regulation. CYP2A6, the human orthologue of Cyp2a5, also has a palindromic E-box site at − 4096 bp from the transcription start site, but the function of this E-box remains to be determined. CYP2A5 protein has a close relative in the CYP2A subfamily, CYP2A4. CYP2A5 and CYP2A4 have only 11 amino acid substitutions in their protein sequences, but the substrate preferences are different, with CYP2A4 having testosterone 15-α-hydroxylase activity.60 Recent sequencing results from the mouse genome project (GenBank accession number NT_039413, 27-APR-2006) indicate that besides the protein-coding regions of Cyp2a5 and Cyp2a4, the 5′ flanking sequences are over 90% homologous up to − 3000 bp and that the transcription factor binding sites known to be important for Cyp2a5 regulation are conserved.27,57 In the methods we have used, the primer pairs and probes are targeted to Cyp2a5 and promoter se quences cloned in the reporter gene vectors represent Cyp2a5. However, it is possible that ARNT and USF are also involved in the regulation of Cyp2a4. Indeed, HNF-4α has already been shown to participate in Cyp2a4 regulation.61 The regulation of ARNT is poorly understood. There are two alternatively spliced forms of ARNT, and both transcript variants 1 and 2 are found in mouse and human. The variant 1 contains 15 additional amino acid residues just N-terminal of the

basic region encoded by exon 5,1,29 and it activates slightly more than the variant 2, an E-box-driven but not an XRE-driven reporter gene.62 Furthermore, variant 1 binding to the E-box as a homodimer is inhibited by phosphorylation of Ser77 by protein kinase CKII.62 Yet, we could not find any study reporting in vivo roles for the different splicing variants or significance of the phosphorylation of the splice variant 1. The ARNT expression vector used in these studies encoded transcription variant 1. During hepatocyte cultivation, the splicing variants were found to be differentially expressed and the variant 1 was increased during the first day of culture. This suggests differential regulation and specific functions for the two variants. There are numerous factors that are changed in cell culture compared to the in vivo liver that can affect the expression of transcription factors, including nutrient and hormonal status, confluence of the cells, and growth matrices. Interestingly, together with increasing ARNT variant 1 expression, the CYP2A5 mRNA level was increased, suggesting, although not confirming, that the variant 1 could play a major role in the activation of E-box motifs. In conclusion, we have shown that the Cyp2a5 promoter contains a distal E-box motif that can bind both ARNT and USF. Both factors may have important roles in the regulation of Cyp2a5 but the significance of the factors may vary under different physiological conditions. ARNT was shown to contribute to transcriptional activation of the Cyp2a5 gene significantly both in cultured hepatocytes and in liver in vivo. These results indicate that ARNT controls Cyp2a5 transcription and thus suggest active involvement of the ARNT homodimer in mammalian gene regulation.

Materials and Methods Preparation of primary cultures of hepatocytes Hepatocytes were isolated from male DBA/2 (OlaHsd) mice (Center for Experimental Animals, University of Oulu, Finland) aged eight to ten weeks. Livers were perfused with collagenase solution (Worthington Biochemical Co., Lakewood, NJ, USA) as described.63 After filtration and centrifugation, the isolated hepatocytes were dispersed in William's medium E (Sigma Chemical Co., St. Louis, MO, USA) containing 20 ng/ml of dexamethasone (Sigma), ITS (5 mg/l of insulin, 5 mg/l of transferrin, and 5 μg/l of sodium selenate) (Sigma), 50 μg/ml of gentamicin (Invitrogen, Paisley, Scotland), and 10% (v/v) fetal bovine serum (Invitrogen) at a density of 1 × 107 cells/100 mm dish, 2 × 107 cells/175-cm2 flask and 3 × 105 cells/well in 12-well plates. The cultures were maintained at 37 °C in a humidified incubator for 1 h, after which non-attached cells were removed by aspiration and discarded, and the medium was replaced by serum-free William's E medium. The cultures were maintained for additional 24 h before transient transfection or chromatin immunoprecipitation, or for 23 h before RNA extraction.

648

ARNT and USF Regulate Cyp2a5 Transcription

The animal experiments have been approved by the local animal care and use committees.

activity with the Dual-Luciferase Reporter Assay System (Promega) 48 h after transfection.

Cell cultures

Site-directed mutagenesis

A subclone Hepa-1c1c7 of the mouse hepatoma cell line Hepa-1 and the Hepa1c1c7 mutant strain defective in ARNT (c4) were kindly provided by Dr Sirkku Saarikoski (Helsinki, Finland) (originally from Dr Oliver Hankinson, UCLA, Los Angeles, CA, USA).64 The cell lines were cultured in nucleoside-free α-minimal essential medium (Invitrogen) supplemented with 10% (v/v) fetal bovine serum, 100 U/ml of penicillin and 100 μg/ ml of streptomycin (Invitrogen). The monkey kidney cell line COS-1 (American Type Culture Collection, Rockville, MD, USA) was cultured in Dulbecco's modified Eagle's medium with GlutaMAX™ (Invitrogen), containing 10% (v/v) fetal bovine serum and 100 U/ml of penicillin and 100 μg/ml of streptomycin. The cells were seeded to 24-well plates on the day preceding the transfection assays.

The MatInspector professional program using Genomatix matrices‡ was used to search for putative binding sites at the − 3033 to +10 promoter region of the Cyp2a5 gene. A putative E-box element was identified at the Cyp2a5 5′ − 2403 to −2398 region. Site-directed mutagenesis of the element was performed using the QuikChange™ Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) according to the manufacturer's instructions. The mutations were introduced into the Cyp2a5 5′ −3033/+ 10-Luc plasmid using mutated oligonucleotides: 2417-GTCTGTGTCTCCCCCAGCTGACTTCAGTTTCTGC-2384. The core binding sequence is italicized, and the mutated nucleotides (−2401 CG → GC) are shown in the middle of the core sequence. Mutations were chosen on the basis of the results reported by Swanson et al.5 Correct assembly of the mutations was confirmed by sequencing.

Plasmids and transient transfection assays

Chromatin immunoprecipitation (ChIP)

The Cyp2a5 5′ −3033/+10 (from the transcription start site) fragment of the Cyp2a5 5′ flanking region cloned into the pGL3-Basic vector was used in reporter gene assays.27 The expression plasmid pcDNA/Neo-ARNT expressing mouse ARNT was provided by Dr. Oliver Hankinson (UCLA, Los Angeles, CA, USA).29 ARNT cDNA was restricted with BamHI and HindIII enzymes from the pcDNA/Neo construct and ligated to the pcDNA3.1(+) vector (Invitrogen). Sogawa et al.4 and Corton et al.31 have determined that the 34 C-terminal amino acid residues of the ARNT protein encode the transactivation domain. The ARNT deletion construct without the C-terminal transactivation domain (ARNT ΔC) was prepared by polymerase chain reaction (PCR), using the pcDNA/Neo-ARNT as a template and the following primers: forward mARNT+1 FW (5′ TT TTT AAG CTT GCC ACC ATG GCG GCG ACT ACA GCT AAC CCA G 3′, where italicized bases code for the HindIII restriction site) and reverse mARNT+1869 RV (5′ TT TTT GGA TCC TTA TAT CTG CCC TGC AGA AGA TGA TGG CTG G 3′, where BamHI site is italicized). The PCR product was subcloned into the pcDNA3.1(+) vector. The pCR3-hUSF-1 and pCR3-hUSF-2a expression vectors were kindly provided by Dr Benoit Viollet (Institut Cochin, Département de Genetique, Developpement et Pathologie Moleculaire, Université René Descartes Paris 5, INSERM U567, Paris, France). The rat HNF-4α1 expression plasmid was a generous gift from Dr Mary C. Weiss (Unité de Génétique de la Différenciation, Département de Biologie de Developpement, Institut Pasteur, Paris Cedex 15, France).65 The reporter gene constructs were transfected into mouse hepatocytes (12-well plates), Hepa-1 cells (24-well plates) or COS-1 cells 1 (24-well plates) together with Renilla luciferase reporter vector (pRL3-TK) (Promega, Madison, WI, USA) used as an internal control. 0.5 μg of Cyp2a5 5′-Luc and 0.1 μg of pRL3-TK were transfected per 3 × 105 hepatocytes, and 0.65 μg of Cyp2a5 5′-Luc and 0.05 μg of pRL3-TK were transfected per well of Hepa-1 or COS-1 cells using Tfx-20 reagent (Promega) according to the manufacturer's protocol in Opti-MEM I medium (Invitrogen). In the co-transfection assays of COS-1 or Hepa-1 cells, 0.05 μg of expression vector DNA was transfected per well. The cells were assayed for luciferase

Primary hepatocyte cultures (in 175 cm2 flasks) were used for immunoprecipitation of protein–DNA complexes. Chromatin immunoprecipitations were done according to Väisänen et al. with certain exceptions.66 At 24 h after liver perfusion, transcription factors were crosslinked to DNA with 1% (v/v) formaldehyde for 10 min at 37 °C. Cells were washed and collected, and pelleted cells were resuspended in 2 ml of SDS lysis buffer. Lysates were sonicated to reduce DNA length to between 200 and 1000 bp. Pre-cleared chromatin fractions were shared into 200 μl aliquots and diluted tenfold in ChIP dilution buffer. Then 1 μg of rabbit polyclonal anti-ARNT antibody (ab14829, Abcam Ltd, Cambridge, UK), rabbit polyclonal anti-USF-1 antibody (C-20, sc-229, Santa Cruz Biotechnology, Santa Cruz, CA, USA), goat polyclonal anti-HNF-4α antibody (C-19, sc-6556, Santa Cruz Biotechnology) or rabbit IgG (Zymed Laboratories, South San Francisco, CA, USA), used as a negative control, was added to the chromatin solution and incubated overnight at 4 °C. Immune complexes were collected with 60 μl of 50% (w/v) Protein A/agarose slurry and the beads were washed with immune complex wash buffer series. Protein–DNA complexes were eluted and reverse crosslinked at +65 °C overnight. The remaining proteins were digested with Proteinase K. DNAs were purified by extraction with phenol/chloroform/isoamylalcohol (25:24:1, v:v:v) and resuspended in 60 μl of water. A dilution series was prepared from input sample containing the total fragmented chromatin. The AmpliQ Universal Real Time PCR Master Mix Kit (Ampliqon, Copenhagen, Denmark) was used for realtime PCR. PCR mixtures contained 10 μl of the 2X master mix, 2 μl of Green DNA Dye diluted 1:2000 (v:v), 200 nM each PCR primer (Table 1) and 5 μl of precipitated DNA in a total volume of 20 μl. Samples were incubated at 95 °C for 15 min, followed by 45 cycles of 95 °C for 30 s, 55 °C for 1 min and 72 °C for 30 s in an Mx3000P QPCR system (Stratagene). Fluorescence values of the Q-PCR products were corrected with fluorescence signals of the passive reference dye (ROX). The specificity of the PCR-products ‡ http://www.genomatix.de

649

ARNT and USF Regulate Cyp2a5 Transcription Table 1. PCR primers Primer (Forward, FW; reverse, RV)

Sequence (5′-3′)

Amplicon size (bp)

A. Real-time PCR primers used in chromatin immunoprecipitation studies Cyp2a5 5′ E-box-2446FW GACTGCAGGCACTGACTCC Cyp2a5 5′ E-box-2319RV AGAGCTCGGAAGAGATCAAG Cyp2a5 5′ HNF-4 RE-131FW CAGTGTTGGCAATGTCCCAA Cyp2a5 5′ HNF-4 RE+10RV GATAGACAGACAGTGATGGC Cyp2a5 5′ NC-1028FW GTTGTCAGTGCATCCATAGG Cyp2a5 5′ NC-663RV TCTTGGAGTCAGCACCTGTG B. Quantitative PCR primers for detection of mRNA levels Cyp2a5+207FW CACCATCTACCTGGGACCT Cyp2a5+356RV CTGAAGGCTACGCCATAGC m18S+949FW CGCCGCTAGAGGTGAAATTC m18S+1098RV CCAGTCGGCATCGTTTATGG mHNF-4α+355FW GAAGCTGTCCAAAATGAGCG mHNF-4α+460RV GCCATTGATCCCAGAGATGG mARNT+182FW GATGCGATGATGACCAGATGTG a CAGTGAGGAAAGATGGCTTGTAGG mARNT+481(sv1)/+436(sv2)RV mUSF-1+581FW CAACACGAGATGAGAAACG mUSF-1+722RV CCTTTACTCTGGCCAGACT a

Gene Bank accession code

128 141

AY321510

366

150

NM_007812

150

X00686

141

NM_008261

300 (sv1) 255 (sv2) 142

NM_001037737.1(sv1) NM_009709.2(sv2) NM_009480.1

sv; splicing variant.

was confirmed with melting curve analysis and by size as determined by agarose gel electrophoresis. In vitro translation of ARNT For in vitro translation, ARNT cDNA was digested from the pcDNA/Neo-ARNT vector using HindIII and BamHI restriction enzymes and cloned into the pSP64 poly(A) vector (Promega). The TnT SP6 Quick Coupled Transcription/Translation System (Promega) was used to produce ARNT protein from pSP64/ARNT. Then 1 μg of pSP64/ ARNT, empty pSP64 or pSP64 Luciferase Control DNA was added to the rabbit reticulosyte lysate system and the reaction was performed according to the manufacturer's protocol. The translation results were analyzed by Western blot and by measuring the luciferase activity as a positive control. Electrophoretic mobility-shift assay Double-stranded DNA probes were prepared by annealing the desired sense and antisense oligonucleotides: Cyp2a5 5′-E-box; 2417-5′ GTCTGTGTCTCCCCCACGTGACTTCAGTTTCTGC 3′-2384 and the consensus ARNT binding site; 5′ GGGAGGTCACGTGATTGTGG 3′, where core binding sequences are underlined.5,16 The sequence of the mutated Cyp2a5 5′-E-box oligonucleotide was the same as that used in the site-directed mutagenesis. Double-stranded oligonucleotides were 5′ end-labeled with [γ-32P]ATP and bacteriophage T4 polynucleotide kinase and then purified using the QIAquick nucleotide removal kit (Qiagen, Venlo, The Netherlands). In a final volume of 15 μl, 1.5 μl of translation mix, 6 µl of binding buffer (25 mM Hepes (pH 7.9), 10% (v/v) glycerol, 50 mM KCl, 0.5 mM EDTA, Complete Mini protease inhibitor cocktail (Roche Diagnostics GmbH, Germany), 1 μg of single-stranded DNA, 1 μg of d(I-C) (Amersham Biosciences, Little Chalfont, UK) and 0.04 pmol (30,000 CPM) of labeled oligonucleotide probe were incubated at room temperature (22 °C) for 30 min. For competition experiments, unlabeled competitor oligonucleotides (fivefold to 100-fold excess) were added to the mixtures. For supershift experiments, 1 μg of rabbit anti-ARNT polyclonal

antibody (H-172, sc-5580, Santa Cruz Biotechnology) was added to the in vitro translation mix and kept on ice for 20 min. The samples were separated by electrophoresis through 6% (w/v) polyacrylamide gel, and the retarded complexes were detected by autoradiography. Liver-specific Arnt-null mouse line The Arnt-floxed (fl/fl) mouse, generated as described,67 was bred with a transgenic mouse expressing the Cre transgene under control of the liver-specific albumin enhancer-promoter sequence (AlbCre),68 kindly provided by Dr Derek Le Roith (NIDDK, NIH, Bethesda MD, USA). The liver-specific Arnt(fl/fl)-AlbCre (Arnt-null) mice and their Arnt(fl/fl) controls (without the AlbCre transgene) were on a mixed genetic background C57Bl/ 6N. The Arnt(fl/fl) mice were bred as littermates with the Arnt(fl/fl)-AlbCre mice in order to maintain the same genetic background. RNA preparation, quantitative PCR and Northern blot Murine hepatocytes were cultured for 24 h (100 mm wells) after liver perfusion, after which total RNA was isolated using the Tri-Reagent (Sigma) according to manufacturer's protocol for monolayer cells. A total of 1 × 107 cells were counted for 0 h samples, which were pelleted and used directly for RNA extraction. Total liver and kidney RNA of 12–16 weeks old Arnt(fl/fl) and Arnt (fl/fl)-AlbCre mice was isolated by the guanidine thiocyanate/CsCl method.69 RNA samples were treated with DNAse (Promega) to avoid contamination of the genomic DNA. A 1 μg portion of each RNA sample was reverse transcribed to produce cDNA using p(dN)6 random primers (Roche) and M-MLV reverse transcriptase (Promega). RT– samples were prepared similarly but without reverse transcriptase enzyme to test if any contaminating genomic DNA is still present in the samples. The AmpliQ Universal Real Time PCR Master Mix Kit (Ampliqon) was used for real-time quantitative PCR (Q-PCR). PCR mixtures contained 10 μl of the 2X master mix, 2 μl of Green DNA Dye diluted 1:2000 (v:v), 400 nM each PCR primer (Table 1) and 1 μl of cDNA in a total volume of

650 20 μl. Samples were incubated at 95 °C for 15 min, followed by 40 cycles of 95 °C for 30 s, 55 °C for 1 min and 72 °C for 30 s in an Mx3000P QPCR system (Stratagene). Fluorescence values of the Q-PCR products were corrected with fluorescence signals of the passive reference dye (ROX). The specificity of the PCR-products was confirmed with melting curve analysis and by size as determined by agarose gel electrophoresis. The RNA levels of CYP2A5, ARNT, USF-1 and HNF-4α were normalized against the 18 S control levels by using the comparative CT (ΔΔCT) method§. In Northern blot analysis, 20 μg of the total RNA was electrophoretically resolved and transferred onto a Hybond-N+nylon membrane (Amersham Biosciences). The RNA was fixed by UV-crosslinking, and the membrane was hybridized with [α-32P]dCTP-labeled probes. The full-length CYP2A5 cDNA was kindly provided by Dr Masahiko Negishi (NIEHS, Research Triangle Park NC, USA), the ARNT probe was prepared by digesting the pSP64/ARNT vector with SacI restriction enzyme yielding a 155 bp fragment and the 18 S probe was provided by Dr Heikki Ruskoaho (University of Oulu, Finland). Immunoblotting The 10,000 g supernatant fractions of mouse liver samples were isolated and the samples were transferred to a polyvinylidene fluoride filter (Millipore, Billerica, MA, USA) after SDS-PAGE using standard methods. Immunoblotting was then carried out as described earlier,57 using anti-CYP2A5 antibody (kindly provided by Dr Risto Juvonen, University of Kuopio, Kuopio, Finland) and anti-actin antibody (Sigma). The immunoreactive bands were quantified using Quantity One software (Bio-Rad, Hercules, CA, USA) and the CYP2A5 protein levels were normalized against the actin levels. Statistical analysis Student's t test was used for comparisons between two groups. Comparisons of several groups were done with one-way analysis of variance (ANOVA) followed by the least significant difference post hoc test. Differences were considered significant when p < 0.05.

Acknowledgements The skillful technical assistance of Päivi Tyni and Ritva Tauriainen is gratefully acknowledged. We thank Dr Sirkku Saarikoski (National Public Health Institute, Department of Mental Health and Alcohol Research, Helsinki, Finland) for her valuable help with Hepa-1 cells, and Dr Oliver Hankinson (Department of Pathology and Laboratory Medicine, Jonsson Comprehensive Cancer Center, Molecular Biology Institute, University of California, Los Angeles, CA, USA) for providing the ARNT expression plasmid and the permission to use the Hepa-1 § Described in the Methods and Application Guide, Introduction to Quantitative PCR (Stratagene, IN #70200-01/ Revision #105002).

ARNT and USF Regulate Cyp2a5 Transcription

mutant cell line. We are grateful to Dr Carsten Carlberg and Dr Sami Väisänen (Department of Biochemistry, University of Kuopio, Finland) for their helpful advice for the chromatin immunoprecipitation method, and to Esa Huusela (AH Diagnostics, Helsinki, Finland) for providing excellent assistance with real-time PCR. This study was supported by the Academy of Finland (contract 110591) and the Finnish Technological Research Agency.

References 1. Hoffman, E. C., Reyes, H., Chu, F. F., Sander, F., Conley, L. H., Brooks, B. A. & Hankinson, O. (1991). Cloning of a factor required for activity of the Ah (dioxin) receptor. Science, 252, 954–958. 2. Reyes, H., Reisz-Porszasz, S. & Hankinson, O. (1992). Identification of the Ah receptor nuclear translocator protein (Arnt) as a component of the DNA binding form of the Ah receptor. Science, 256, 1193–1195. 3. Wang, G. L., Jiang, B. H., Rue, E. A. & Semenza, G. L. (1995). Hypoxia-inducible factor 1 is a basic-helixloop-helix-PAS heterodimer regulated by cellular O2 tension. Proc. Natl Acad. Sci. USA, 92, 5510–5514. 4. Sogawa, K., Nakano, R., Kobayashi, A., Kikuchi, Y., Ohe, N., Matsushita, N. & Fujii-Kuriyama, Y. (1995). Possible function of Ah receptor nuclear translocator (Arnt) homodimer in transcriptional regulation. Proc. Natl Acad. Sci. USA, 92, 1936–1940. 5. Swanson, H. I., Chan, W. K. & Bradfield, C. A. (1995). DNA binding specificities and pairing rules of the Ah receptor, ARNT, and SIM proteins. J. Biol. Chem. 270, 26292–26302. 6. Huffman, J. L., Mokashi, A., Bachinger, H. P. & Brennan, R. G. (2001). The basic helix-loop-helix domain of the aryl hydrocarbon receptor nuclear transporter (ARNT) can oligomerize and bind E-box DNA specifically. J. Biol. Chem. 276, 40537–40544. 7. Card, P. B., Erbel, P. J. & Gardner, K. H. (2005). Structural basis of ARNT PAS-B dimerization: use of a common beta-sheet interface for hetero- and homodimerization. J. Mol. Biol. 353, 664–677. 8. Sawadogo, M., Van Dyke, M. W., Gregor, P. D. & Roeder, R. G. (1988). Multiple forms of the human gene-specific transcription factor USF. I. Complete purification and identification of USF from HeLa cell nuclei. J. Biol. Chem. 263, 11985–11993. 9. Sirito, M., Walker, S., Lin, Q., Kozlowski, M. T., Klein, W. H. & Sawadogo, M. (1992). Members of the USF family of helix-loop-helix proteins bind DNA as homo-as well as heterodimers. Gene Expr. 2, 231–240. 10. Sirito, M., Lin, Q., Maity, T. & Sawadogo, M. (1994). Ubiquitous expression of the 43- and 44-kDa forms of transcription factor USF in mammalian cells. Nucl. Acids Res. 22, 427–433. 11. Viollet, B., Lefrancois-Martinez, A. M., Henrion, A., Kahn, A., Raymondjean, M. & Martinez, A. (1996). Immunochemical characterization and transacting properties of upstream stimulatory factor isoforms. J. Biol. Chem. 271, 1405–1415. 12. Corre, S. & Galibert, M. D. (2005). Upstream stimulating factors: highly versatile stress-responsive transcription factors. Pigment Cell Res. 18, 337–348. 13. Sawadogo, M. & Roeder, R. G. (1985). Interaction of a gene-specific transcription factor with the adenovirus

ARNT and USF Regulate Cyp2a5 Transcription

14.

15. 16.

17.

18.

19.

20. 21.

22.

23.

24. 25.

26. 27.

28. 29.

30.

major late promoter upstream of the TATA box region. Cell, 43, 165–175. Blackwell, T. K. & Weintraub, H. (1990). Differences and similarities in DNA-binding preferences of MyoD and E2A protein complexes revealed by binding site selection. Science, 250, 1104–1110. Atchley, W. R. & Fitch, W. M. (1997). A natural classification of the basic helix-loop-helix class of transcription factors. Proc. Natl Acad. Sci. USA, 94, 5172–5176. Swanson, H. I. & Yang, J. H. (1999). Specificity of DNA binding of the c-Myc/Max and ARNT/ARNT dimers at the CACGTG recognition site. Nucl. Acids Res. 27, 3205–3212. Camus, A. M., Geneste, O., Honkakoski, P., Bereziat, J. C., Henderson, C. J., Wolf, C. R. et al. (1993). High variability of nitrosamine metabolism among individuals: role of cytochromes P450 2A6 and 2E1 in the dealkylation of N-nitrosodimethylamine and N-nitrosodiethylamine in mice and humans. Mol. Carcinog. 7, 268–275. Pelkonen, P., Lang, M. A., Negishi, M., Wild, C. P. & Juvonen, R. O. (1997). Interaction of aflatoxin B1 with cytochrome P450 2A5 and its mutants: correlation with metabolic activation and toxicity. Chem. Res. Toxicol. 10, 85–90. Nakajima, M., Yamamoto, T., Nunoya, K., Yokoi, T., Nagashima, K., Inoue, K. et al. (1996). Role of human cytochrome P4502A6 in C-oxidation of nicotine. Drug Metab. Dispos. 24, 1212–1217. Murphy, S. E., Raulinaitis, V. & Brown, K. M. (2005). Nicotine 5′-oxidation and methyl oxidation by P450 2A enzymes. Drug Metab. Dispos. 33, 1166–1173. Abu-Bakar, A., Moore, M. R. & Lang, M. A. (2005). Evidence for induced microsomal bilirubin degradation by cytochrome P450 2A5. Biochem. Pharmacol. 70, 1527–1535. Kaipainen, P. & Lang, M. (1985). Comparison of kidney, lung and liver coumarin 7-hydroxylases in phenobarbital pretreated DBA/2J and C57BL/6J mice. Biochem. Pharmacol. 34, 1987–1991. Koskela, S., Hakkola, J., Hukkanen, J., Pelkonen, O., Sorri, M., Saranen, A. et al. (1999). Expression of CYP2A genes in human liver and extrahepatic tissues. Biochem. Pharmacol. 57, 1407–1413. Su, T., Sheng, J. J., Lipinskas, T. W. & Ding, X. (1996). Expression of CYP2A genes in rodent and human nasal mucosa. Drug Metab. Dispos. 24, 884–890. Glisovic, T., Soderberg, M., Christian, K., Lang, M. & Raffalli-Mathieu, F. (2003). Interplay between transcriptional and post-transcriptional regulation of Cyp2a5 expression. Biochem. Pharmacol. 65, 1653–1661. Su, T. & Ding, X. (2004). Regulation of the cytochrome P450 2A genes. Toxicol. Appl. Pharmacol. 199, 285–294. Ulvila, J., Arpiainen, S., Pelkonen, O., Aida, K., Sueyoshi, T., Negishi, M. & Hakkola, J. (2004). Regulation of Cyp2a5 transcription in mouse primary hepatocytes: roles of hepatocyte nuclear factor 4 and nuclear factor I. Biochem. J. 381, 887–894. Li, H., Dong, L. & Whitlock, J. P., Jr (1994). Transcriptional activation function of the mouse Ah receptor nuclear translocator. J. Biol. Chem. 269, 28098–28105. Reisz-Porszasz, S., Probst, M. R., Fukunaga, B. N. & Hankinson, O. (1994). Identification of functional domains of the aryl hydrocarbon receptor nuclear translocator protein (ARNT). Mol. Cell. Biol. 14, 6075–6086. Fukunaga, B. N., Probst, M. R., Reisz-Porszasz, S. & Hankinson, O. (1995). Identification of functional domains of the aryl hydrocarbon receptor. J. Biol. Chem. 270, 29270–29278.

651 31. Corton, J. C., Moreno, E. S., Hovis, S. M., Leonard, L. S., Gaido, K. W., Joyce, M. M. & Kennett, S. B. (1996). Identification of a cell-specific transcription activation domain within the human Ah receptor nuclear translocator. Toxicol. Appl. Pharmacol. 139, 272–280. 32. Li, H., Ko, H. P. & Whitlock, J. P. (1996). Induction of phosphoglycerate kinase 1 gene expression by hypoxia. Roles of Arnt and HIF1alpha. J. Biol. Chem. 271, 21262–21267. 33. Wood, S. M., Gleadle, J. M., Pugh, C. W., Hankinson, O. & Ratcliffe, P. J. (1996). The role of the aryl hydrocarbon receptor nuclear translocator (ARNT) in hypoxic induction of gene expression. Studies in ARNT-deficient cells. J. Biol. Chem. 271, 15117–15123. 34. Antonsson, C., Arulampalam, V., Whitelaw, M. L., Pettersson, S. & Poellinger, L. (1995). Constitutive function of the basic helix-loop-helix/PAS factor Arnt. Regulation of target promoters via the E box motif. J. Biol. Chem. 270, 13968–13972. 35. Moffett, P. & Pelletier, J. (2000). Different transcriptional properties of mSim-1 and mSim-2. FEBS Letters, 466, 80–86. 36. Lee, K. H., Park, J. W. & Chun, Y. S. (2004). Nonhypoxic transcriptional activation of the aryl hydrocarbon receptor nuclear translocator in concert with a novel hypoxia-inducible factor-1alpha isoform. Nucl. Acids Res. 32, 5499–5511. 37. Long, W. P., Chen, X. & Perdew, G. H. (1999). Protein kinase C modulates aryl hydrocarbon receptor nuclear translocator protein-mediated transactivation potential in a dimer context. J. Biol. Chem. 274, 12391–12400. 38. Diaz Guerra, M. J., Bergot, M. O., Martinez, A., Cuif, M. H., Kahn, A. & Raymondjean, M. (1993). Functional characterization of the L-type pyruvate kinase gene glucose response complex. Mol. Cell. Biol. 13, 7725–7733. 39. Ribeiro, A., Pastier, D., Kardassis, D., Chambaz, J. & Cardot, P. (1999). Cooperative binding of upstream stimulatory factor and hepatic nuclear factor 4 drives the transcription of the human apolipoprotein A-II gene. J. Biol. Chem. 274, 1216–1225. 40. Pastier, D., Lacorte, J. M., Chambaz, J., Cardot, P. & Ribeiro, A. (2002). Two initiator-like elements are required for the combined activation of the human apolipoprotein C-III promoter by upstream stimulatory factor and hepatic nuclear factor-4. J. Biol. Chem. 277, 15199–15206. 41. Abbott, B. D., Schmid, J. E., Brown, J. G., Wood, C. R., White, R. D., Buckalew, A. R. & Held, G. A. (1999). RTPCR quantification of AHR, ARNT, GR, and CYP1A1 mRNA in craniofacial tissues of embryonic mice exposed to 2,3,7,8-tetrachlorodibenzo-p-dioxin and hydrocortisone. Toxicol. Sci. 47, 76–85. 42. Wang, F., Shi, S., Zhang, R. & Hankinson, O. (2006). Identifying target genes of the aryl hydrocarbon receptor nuclear translocator (Arnt) using DNA microarray analysis. Biol. Chem. 387, 1215–1218. 43. Bauer, M., Hamm, A. C., Bonaus, M., Jacob, A., Jaekel, J., Schorle, H. et al. (2004). Starvation response in mouse liver shows strong correlation with life-spanprolonging processes. Physiol. Genom. 17, 230–244. 44. Viitala, P., Posti, K., Lindfors, A., Pelkonen, O. & Raunio, H. (2001). cAMP mediated upregulation of CYP2A5 in mouse hepatocytes. Biochem. Biophys. Res. Commun. 280, 761–767. 45. Vaulont, S., Vasseur-Cognet, M. & Kahn, A. (2000). Glucose regulation of gene transcription. J. Biol. Chem. 275, 31555–31558. 46. Read, M. L., Clark, A. R. & Docherty, K. (1993). The

652

47.

48.

49.

50. 51.

52.

53.

54.

55.

56.

57.

ARNT and USF Regulate Cyp2a5 Transcription

helix-loop-helix transcription factor USF (upstream stimulating factor) binds to a regulatory sequence of the human insulin gene enhancer. Biochem. J. 295, 233–237. Portois, L., Tastenoy, M., Viollet, B. & Svoboda, M. (2002). Functional analysis of the glucose response element of the rat glucagon receptor gene in insulinproducing INS-1 cells. Biochim. Biophys. Acta, 1574, 175–186. Iynedjian, P. B. (1998). Identification of upstream stimulatory factor as transcriptional activator of the liver promoter of the glucokinase gene. Biochem. J. 333, 705–712. Wang, D. & Sul, H. S. (1997). Upstream stimulatory factor binding to the E-box at-65 is required for insulin regulation of the fatty acid synthase promoter. J. Biol. Chem. 272, 26367–26374. van Schaftingen, E. & Gerin, I. (2002). The glucose-6phosphatase system. Biochem. J. 362, 513–532. Carriere, V., Le Gall, M., Gouyon-Saumande, F., Schmoll, D., Brot-Laroche, E., Chauffeton, V. et al. (2005). Intestinal glucose-dependent expression of glucose-6-phosphatase: involvement of the aryl receptor nuclear translocator transcription factor. J. Biol. Chem. 280, 20094–20101. Gunton, J. E., Kulkarni, R. N., Yim, S., Okada, T., Hawthorne, W. J., Tseng, Y. H. et al. (2005). Loss of ARNT/HIF1beta mediates altered gene expression and pancreatic-islet dysfunction in human type 2 diabetes. Cell, 122, 337–349. Lavery, D. J., Lopez-Molina, L., Margueron, R., Fleury-Olela, F., Conquet, F., Schibler, U. & Bonfils, C. (1999). Circadian expression of the steroid 15 alphahydroxylase (Cyp2a4) and coumarin 7-hydroxylase (Cyp2a5) genes in mouse liver is regulated by the PAR leucine zipper transcription factor DBP. Mol. Cell. Biol. 19, 6488–6499. Gekakis, N., Staknis, D., Nguyen, H. B., Davis, F. C., Wilsbacher, L. D., King, D. P. et al. (1998). Role of the CLOCK protein in the mammalian circadian mechanism. Science, 280, 1564–1569. Richardson, V. M., Santostefano, M. J. & Birnbaum, L. S. (1998). Daily cycle of bHLH-PAS proteins, Ah receptor and Arnt, in multiple tissues of female Sprague-Dawley rats. Biochem. Biophys. Res. Commun. 252, 225–231. Garrett, R. W. & Gasiewicz, T. A. (2006). The aryl hydrocarbon receptor agonist 2,3,7,8-tetrachlorodibenzo-p-dioxin alters the circadian rhythms, quiescence, and expression of clock genes in murine hematopoietic stem and progenitor cells. Mol. Pharmacol. 69, 2076–2083. Arpiainen, S., Raffalli-Mathieu, F., Lang, M. A., Pelkonen, O. & Hakkola, J. (2005). Regulation of the Cyp2a5 gene involves an aryl hydrocarbon receptordependent pathway. Mol. Pharmacol. 67, 1325–1333.

58. Pickwell, G. V., Shih, H. & Quattrochi, L. C. (2003). Interaction of upstream stimulatory factor proteins with an E-box located within the human CYP1A2 5'-flanking gene contributes to basal transcriptional gene activation. Biochem. Pharmacol. 65, 1087–1096. 59. Zhang, J., Zhang, Q. Y., Guo, J., Zhou, Y. & Ding, X. (2000). Identification and functional characterization of a conserved, nuclear factor 1-like element in the proximal promoter region of CYP1A2 gene specifically expressed in the liver and olfactory mucosa. J. Biol. Chem. 275, 8895–8902. 60. Negishi, M., Lindberg, R., Burkhart, B., Ichikawa, T., Honkakoski, P. & Lang, M. (1989). Mouse steroid 15 alpha-hydroxylase gene family: identification of type II P-450(15)alpha as coumarin 7-hydroxylase. Biochemistry, 28, 4169–4172. 61. Yokomori, N., Nishio, K., Aida, K. & Negishi, M. (1997). Transcriptional regulation by HNF-4 of the steroid 15alpha-hydroxylase P450 (Cyp2a-4) gene in mouse liver. J. Steroid Biochem. Mol. Biol. 62, 307–314. 62. Kewley, R. J. & Whitelaw, M. L. (2005). Phosphorylation inhibits DNA-binding of alternatively spliced aryl hydrocarbon receptor nuclear translocator. Biochem. Biophys. Res. Commun. 338, 660–667. 63. Salonpaa, P., Pelkonen, O., Kojo, A., Pasanen, M., Negishi, M. & Raunio, H. (1994). Cytochrome P4502A5 expression and inducibility by phenobarbital is modulated by cAMP in mouse primary hepatocytes. Biochem. Biophys. Res. Commun. 205, 631–637. 64. Hankinson, O. (1994). A genetic analysis of processes regulating cytochrome P4501A1 expression. Advan. Enzyme Regul. 34, 159–171. 65. Torres-Padilla, M. E. & Weiss, M. C. (2003). Effects of interactions of hepatocyte nuclear factor 4alpha isoforms with coactivators and corepressors are promoter-specific. FEBS Letters, 539, 19–23. 66. Vaisanen, S., Dunlop, T. W., Frank, C. & Carlberg, C. (2004). Using chromatin immunoprecipitation to monitor 1alpha,25-dihydroxyvitamin D3-dependent chromatin activity on the human CYP24 promoter. J. Steroid Biochem. Mol. Biol. 89-90, 277–279. 67. Tomita, S., Sinal, C. J., Yim, S. H. & Gonzalez, F. J. (2000). Conditional disruption of the aryl hydrocarbon receptor nuclear translocator (Arnt) gene leads to loss of target gene induction by the aryl hydrocarbon receptor and hypoxia-inducible factor 1alpha. Mol. Endocrinol. 14, 1674–1681. 68. Yakar, S., Liu, J. L., Stannard, B., Butler, A., Accili, D., Sauer, B. & LeRoith, D. (1999). Normal growth and development in the absence of hepatic insulin-like growth factor I. Proc. Natl Acad. Sci. USA, 96, 7324–7329. 69. Chirgwin, J. M., Przybyla, A. E., MacDonald, R. J. & Rutter, W. J. (1979). Isolation of biologically active ribonucleic acid from sources enriched in ribonuclease. Biochemistry, 18, 5294–5299.

Edited by J. Karn (Received 26 January 2007; received in revised form 28 March 2007; accepted 28 March 2007) Available online 1 April 2007