Transcriptional regulation of chicken cytochrome P450 2D49 basal expression by CCAAT/enhancer-binding protein α and hepatocyte nuclear factor 4α

Authors

  • Qi Yang,

    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    Search for more papers by this author
  • Shulin Tang,

    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    Search for more papers by this author
  • Linfeng Dong,

    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    Search for more papers by this author
  • Qingmei Chen,

    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    Search for more papers by this author
  • Xin Liu,

    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    Search for more papers by this author
  • Jun Jiang,

    Corresponding author
    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    • Correspondence

      Y. Deng, Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, 510642 Guangdong, China

      Fax: +86 20 38604967

      Tel: +86 20 38294890

      E-mail: yqdeng@scau.edu.cn

      J. Jiang, Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, 510642 Guangdong, China

      Fax: +86 20 38604967

      Tel: +86 20 38604967

      E-mail: jiangjun@scau.edu.cn

    Search for more papers by this author
  • Yiqun Deng

    Corresponding author
    1. Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, China
    • Correspondence

      Y. Deng, Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, 510642 Guangdong, China

      Fax: +86 20 38604967

      Tel: +86 20 38294890

      E-mail: yqdeng@scau.edu.cn

      J. Jiang, Guangdong Provincial Key Laboratory of Protein Function and Regulation in Agricultural Organisms, College of Life Sciences, South China Agricultural University, Guangzhou, 510642 Guangdong, China

      Fax: +86 20 38604967

      Tel: +86 20 38604967

      E-mail: jiangjun@scau.edu.cn

    Search for more papers by this author

Abstract

Chicken cytochrome P450 (CYP)2D49 is structurally and functionally related to human CYP2D6, which is an important drug-metabolizing enzyme. To date, little is known about the transcriptional regulation of this cytochrome. Through deletion analysis of the CYP2D49 promoter, we identified two putative degenerate CCAAT/enhancer-binding protein (C/EBP)-binding sites and an imperfect DR1 element (the site contains direct repeats of the hexamer AGGTCA separated by a one-nucleotide spacer motif) within regions –296/–274, –274/–226, and –226/–183, respectively, which may play critical roles in the transcriptional activation of the CYP2D49 gene. Electrophoretic mobility shift assays and chromatin immunoprecipitation assays showed that the putative C/EBP boxes and DR1 element in the CYP2D49 promoter are functional motifs that bind to C/EBPα and hepatocyte nuclear factor 4α (HNF4α), respectively. Furthermore, we studied the functional importance and relationships of these transcription factor-binding sites by examining the effects of mutation and deletion of these regions on promoter activity. These studies revealed that the two C/EBP-binding sites show a compensatory relationship and work cooperatively with the DR1 element to modulate the transcription of CYP2D49. The results of overexpressing C/EBPα and HNF4α in culture cells further confirmed that both C/EBPα and HNF4α contribute significantly to sustaining a high level of CYP2D49 transcription. In conclusion, the data indicate that the constitutive hepatic expression of CYP2D49 is governed by both C/EBPα and HNF4α. Further studies will be required to fully characterize the molecular mechanisms that modulate CYP2D49 expression.

Abbreviations
C/EBP

CCAAT/enhancer-binding protein

CAR

constitutive androstane receptor

ChIP

chromatin immunoprecipitation

COUP-TF

chicken ovalbumin upstream promoter transcription factor

CYP

cytochrome P450

DR1

direct repeat 1

EMSA

electrophoretic mobility shift assay

HNF

hepatocyte nuclear factor

PXR

pregnane X receptor

SD

standard deviation

Introduction

Cytochrome P450s (CYPs) constitute a superfamily of heme-thiolate proteins that play critical roles in the activation of chemical carcinogens, the detoxification of xenobiotics, and the metabolism of endogenous compounds, such as steroid hormones, vitamins, and fatty acids [1-3]. Members of the CYP1–3 protein families are predominantly expressed in the liver, and show broad and overlapping substrate specificities; these enzymes metabolize 70–80% of all drugs in clinical use [4]. Their expression levels and activities show great interindividual variation, owing to complex regulatory networks involving induction/inhibition by xenobiotics, regulation by cytokines or hormones, several disease states, stress, sex, and age, which may ultimately influence drug–drug interactions, carcinogen activation, and detoxification [5]. Thus, understanding the regulatory mechanisms affecting hepatic CYPs is vital for predicting pharmacokinetic variability and the development of personalized medicine.

Intrinsic and extrinsic factors (e.g. xenobiotics exposure, age, and sex) influence the expression and function of hepatic CYPs, mainly through a number of liver-enriched transcription factors that cooperate with ligand-activated nuclear receptors. Each of them generally binds to specific regulatory elements in the 5′-upstream regions of genes [5, 6]. To date, important advances have been made in understanding the transcriptional regulation of human hepatic CYPs through promoter analyses. Liver-enriched transcription factors, such as hepatocyte nuclear factor (HNF)1 (HNF1), HNF3, HNF4, CCAAT/enhancer-binding proteins (C/EBPs), and albumin D-site binding protein are known to be involved in the constitutive and tissue-specific expression of CYP1A, CYP2A, CYP2B, CYP2C, CYP2D, and CYP3A [5, 6]. Also, endogenous or exogenous ligands (e.g. phenobarbital, rifampicin, and 2,3,7,8-tetrachlorodibenzo-p-dioxin) activate nuclear receptors, such as constitutive androstane receptor (CAR), pregnane X receptor (PXR), and aryl-hydrocarbon receptor, which then cooperate with the liver-enriched transcription factors to regulate the induction or repression of CYP1A, CYP2B, CYP2C, and CYP3A [5, 7].

The chicken is one of the most common food-production animals, and chickens are regularly exposed to high levels of xenobiotics, such as veterinary drugs, food additives, and pollutants. Understanding the function and regulation of chicken CYPs is important not only for poultry pharmacology and toxicology, but also for human health, owing to the potential persistence of drug residues in edible tissues. To date, several important CYPs, including CYP1A4/5 [8], CYP2H1/2 [9], CYP2C45 [10], and CYP3A37 [11], have been cloned and identified in chickens; the catalytic profiles of these enzymes are similar, but not identical, to those of their human counterparts. Moreover, the molecular mechanisms involved in the induced expression of CYP2H1/2, CYP2C45 and CYP3A37 have been well studied. Chicken xenobiotic receptor, which is closely related to mammalian PXR and CAR, heterodimerizes with retinoid X receptor γ and binds to direct repeat 4 to transactivate CYP2H1 [12, 13], CYP2C45 [10], and CYP3A37 [14]. The chicken and mammalian drug induction mechanisms are essentially identical [12].

We previously reported the cloning and functional characterization of CYP2D49, an important drug metabolism enzyme in chickens that is structurally and functionally related to human CYP2D6 [15]. CYP2D49 is an essentially noninducible gene that is expressed mainly in the liver. However, in comparison with the large amount of available data regarding the inductive regulation of other chicken CYP genes, little is known about the molecular mechanisms regulating the constitutive and tissue-specific expression of CYP2D49. In the present study, we examined the promoter activity of CYP2D49 by using luciferase reporter assays, and we ultimately identified the cis-acting elements and trans-acting factors involved in its constitutive expression.

Results

The chicken CYP2D49 promoter contains three potential activating cis-elements

To identify potential regulatory elements in the chicken CYP2D49 promoter, the 5′-upstream sequence was obtained, and progressive 5′-deletions were generated. The various promoter fragments were then fused to the LUC gene in the pGL3-Basic vector, and transiently transfected into LMH cells to detect luciferase activity. As shown in Fig. 1A, transfection of the –2300-Luc construct into LMH cells resulted in a 10-fold induction of luciferase activity as compared with the pGL3-Basic vector. However, no activity was observed when the same construct was transfected into COS-7 cells (data not shown). The deletion of the –2300 to –1900 and –350 to –296 sequences caused significant increases in luciferase activity, indicating the existence of negative regulatory elements within these two regions. The –296-Luc construct showed the highest promoter activity. Further deletions to –274, –226 and –183 elicited substantial decreases in luciferase activity (30%, 50%, and 90%, respectively). These results revealed the presence of positive regulatory elements within the –296/–274 region (designated site A), the –274/–226 region (site B), and the –226/–183 region (site C). These sequences were examined with the matinspector professional program (http://www.genomatix.de/onlinehelp/helpmatinspector/matinspectorhelp.html), and three putative transcription factor-binding sites were identified. C/EBP-binding sites are known to consist of palindrome or palindrome-like sequences with half-sites bearing the pentanucleotide sequence 5′-ATTGC-3′. We found two putative C/EBP-binding sites localized within site A and site B that contain degenerate sequences as compared with the perfect sequences of C/EBP-binding sites (Fig. 1B). Additionally, a binding site similar to the site containing direct repeats of the hexamer AGGTCA separated by a one-nucleotide spacer [designated direct repeat 1 (DR1)] was found within site C (Fig. 1B). These three putative regulatory elements may play critical roles in the transcriptional activation of CYP2D49. Moreover, the promoter activities of the (–296 to –183)-Luc, (–274 to –183)-Luc and (–226 to –183)-Luc constructs, in which the sequences down to –183 were removed, were dramatically reduced as compared with their counterparts. This may be attributable to the elimination of the transcriptional start site.

Figure 1.

The chicken CYP2D49 promoter contains three potential activating cis-elements. (A) Deletion analysis of the CYP2D49 promoter. A series of deleted CYP2D49 promoter fragments were generated and transfected into LMH cells as described in 'Experimental procedures'. Schematic representations of the 5′-flanking region and the reporter constructs used are shown on the left. The firefly luciferase activity for each construct was normalized to the Renilla luciferase activity, and expressed as the fold change relative to the pGL3-Basic plasmid. All experiments were performed in triplicate, and the values given represent the mean ± SD of three independent experiments, each of which represents the average of three wells. Statistical significance was determined with ANOVA, and significance was defined as **P < 0.01. (B) Schematic diagram showing the putative binding sites in the CYP2D49 proximal promoter. The positions relate to the translation start site (ATG) as +1. The –296/–274, –274/–226 and –226/–183 sequences were designated as sites A, B, and C, respectively. The two putative C/EBP boxes and the DR1 element located in sites A, B and C are underlined.

The putative C/EBP boxes and DR1 element in the CYP2D49 promoter are functional motifs that bind to C/EBPα and HNF4α, respectively

The C/EBP boxes and DR1 element are well-known binding motifs for the liver-enriched transcription factors C/EBPα and HNF4α, respectively. To determine whether the putative C/EBP boxes and DR1 element in the CYP2D49 promoter are functional motifs that bind to C/EBPα and HNF4α, gel shift assays were carried out with the nuclear extracts both from C/EBPα or HNF4α overexpressed COS-7 cells and from chicken liver tissues.

Biotinylated probes (A, B, and C) were generated on the basis of the sequences of sites A, B, and C, respectively. When incubated with the nuclear extracts from pcDNA-C/EBPα-transfected COS-7 cells, both probe A and probe B migrated slowly in the gel, forming single and specific shifted bands; however, no shifted band was detected when probe A or probe B was incubated with the extracts from pcDNA3.1(–)-transfected COS-7 cells (Fig. 2A, lanes 2 and 3; Fig. 2B, lanes 2 and 3). Additionally, the intensities of the shifted bands were specifically reduced when a 100-fold excess of unlabeled probe A or probe B was added (Fig. 2A, lane 4; Fig. 2B, lane 4); however, this did not occur when competition experiments were performed with the same amount of mutated probes (Fig. 2A, lane 5; Fig. 2B, lane 5). These results indicate that C/EBPα can specifically bind to the two adjacent C/EBP boxes in the CYP2D49 promoter. Using the same method, we further evaluated the interaction between DR1 and HNF4α. As shown in Fig. 2C, a shifted band was observed in the gel (Fig. 2C, lane 3), and the formation of this protein·DNA complex was specifically competitively disrupted by the addition of a 100-fold excess of unlabeled probe C (Fig. 2C, lane 4), but the complex remained when competition experiments were performed with the same amount of unrelated probe A (Fig. 2C, lane 5) or the mutated probe C (Fig. 2C, lane 6). These results suggest that the putative DR1 element in the CYP2D49 promoter can bind to HNF4α.

Figure 2.

EMSA analyses demonstrate the binding of overexpressed C/EBPα and HNF4α to the proximal CYP2D49 promoter in COS-7 cells. EMSAs were performed as described in 'Experimental procedures'. Nuclear extracts from COS-7 cells transfected with pcDNA3.1, C/EBPα (A, B) or HNF4α (C) were used. The wild-type (WT) and mutant (MT) DNA probes used in each experiment are shown above the gel. For competition experiments, a 100-fold molar excess of the appropriate unlabeled oligonucleotides was added to the binding reaction.

To further confirm the results described above, gel shift analysis was performed with nuclear extracts from chicken liver tissues. As shown in Fig. 3A–C (lane 2), the addition of nuclear proteins from chicken liver tissues resulted in the formation of single protein·DNA complexes with sites A, B, and C, respectively. All of these complexes were observed to bind specifically to the labeled probes, because their formation was abolished by the addition of unlabeled probes but not by the addition of their mutated counterparts (Fig. 3A–C, lanes 3–5). Moreover, the complexes were shown to be dependent on the interactions of C/EBPα and HNF4α, respectively, as the complexes were supershifted by the inclusion of antibodies against C/EBPα or HNF4α in the DNA-binding reactions (Fig. 3A–C, lanes 6 and 7). The supershifts were observed to be specific, as the addition of goat IgG to the DNA-binding reactions had no effect on complex mobility (data not shown).

Figure 3.

EMSAs and ChIP assays further demonstrate the binding of C/EBPα and HNF4α to the upstream region of the CYP2D49 promoter. EMSAs and ChIP assays were performed as described in 'Experimental procedures'. Nuclear extracts from chicken liver tissues were used in the EMSAs. The DNA probes used are the same as those used in Fig. 2. The DNA protein complexes and supershifts are indicated by arrows. (A, B) C/EBPα present in the chicken liver nuclear extracts was able to bind to probes A and B. (C) HNF4α present in chicken liver extracts recognized probe C. (D) ChIP assays demonstrated that both C/EBPα and HNF4α physiologically interact with their corresponding elements. Upper panel: schematic diagram showing the positions of the three PCR primer sets. Replicon 1 and replicon 2 encompass the two C/EBP boxes and the DR1 element, respectively. Replicon 3 encompasses a genomic fragment upstream of the CYP2D49 promoter that does not contain C/EBP boxes or DR1 elements. Crosslinked chromatin collected from LMH cells was immunoprecipitated with antibodies against C/EBPα, or HNF4α, or IgG. Lower panel: representative electrophoretic images of PCR products are shown.

The gel shift assays described above provided in vitro evidence that the two adjacent C/EBP boxes and one DR1 element in the CYP2D49 promoter are binding sites for C/EBPα and HNF4α, respectively. To validate the interaction of C/EBPα and HNF4α with the CYP2D49 locus in vivo, chromatin immunoprecipitation (ChIP) assays were performed with LMH cells. In addition to the genomic fragments containing the two adjacent C/EBP boxes (replicon 1) and the DR1 element (replicon 2), a genomic fragment derived from the upstream portion of the CYP2D49 promoter that did not contain the C/EBP boxes and DR1 element (replicon 3) was monitored as a negative control. When the chromatin input was used as the template, each PCR amplification produced a band corresponding to the respective genomic fragment (Fig. 3D). The sizes of these bands were the same as the size predicted for the particular set of primers (Table 1). Moreover, specific DNA bands were also detected in the C/EBPα-immunoprecipitated and HNF4α-immunoprecipitated chromatin when the CYP2D49 promoter region that contains the two adjacent C/EBP boxes and the DR1 element was amplified (Fig. 3D). In contrast, no DNA band was amplified from the IgG-immunoprecipitated chromatin with the same primer pairs (Fig. 3D). These results provided further evidence that the putative C/EBP boxes and DR1 element in the CYP2D49 promoter are functional motifs that bind to C/EBPα and HNF4α in vivo.

Table 1. Primers used for ChIP and real-time PCR.
PrimerSequence (5′- to 3′)Usage
Replicon 1 forwardAAGACAGCACAGCTGCCTCTGTTChIP
Replicon 1 reverseTGAGGGCCAATCCTTTTCCTCCT
Replicon 2 forwardAAAGGAGGAGGAGCTCTGGATCC
Replicon 2 reverseTTTGTAGTTTGTCTACTCCTTGGTC
Replicon 3 forwardAGGAGCCTTTCTCTGATCATGTTG
Replicon 3 reverseTGGGTAGGACTGGATGATGG
β-Actin forwardGGCTGTGCTGTCCCTGTAReal-time PCR
β-Actin reverseCGGCTGTGGTGGTGAAG
CYP2D49 forwardGGCAAAGGGTAAGGAGGCT
CYP2D49 reverseTGACGGCATTG GTGTAGGG

The two C/EBP-binding sites show a compensatory relationship and work cooperatively with the DR1 element to modulate the transcription of CYP2D49

To study the functional importance and relationships of the transcription factor-binding sites identified by the electrophoretic mobility shift assays (EMSAs) and ChIP assays, we examined the effects of their mutations and deletions on promoter activities. As demonstrated in Fig. 4A, mutagenesis of both the 5′ and 3′ hexamer half-sites of the DR1 element had the same effect as deletion of the DR1 element; both caused a decrease of ~ 50% in luciferase activity relative to the wild-type promoter. These results are in accordance with our previous finding, obtained from deletion analysis of the CYP2D49 promoter (Fig. 1A), that the DR1 element is the key regulatory element in the CYP2D49 promoter. However, the mutation of site A or site B did not change the luciferase activity significantly, although the deletion of site A or site B moderately reduced the luciferase activity (Fig. 4A). These results indicated that the two C/EBP sites may compensate for each other during the regulation of CYP2D49 expression. Also, all C/EBP boxes and their adjacent sequences are required for the maximal function of C/EBPα. Taken together, these data suggest that the two C/EBP boxes work in a compensatory but not redundant manner. To further confirm our speculation, we mutated both site A and site B. As expected, a 55% drop in luciferase activity relative to the wild-type construct was observed (Fig. 4A). Additionally, we performed competition assays to study the binding affinity of C/EBPα for site A and site B. As shown in Fig. 4B, probe A or probe B formed one DNA·protein complex with chicken liver nuclear extracts (lanes 2 and 6), which was disappeared when a 10-fold excess of unlabeled probe B or probe A was added, respectively (lane 3 and 7); however, the complex remained even a 100-fold excess of unlabeled probe C was added in the mixture (lane 10 and 11). These results indicate that C/EBPα binds to site A and site B with equal affinity, and that the two C/EBP-binding sites have a compensatory relationship.

Figure 4.

The two C/EBP-binding sites have a compensatory relationship, and work cooperatively with the DR1 element to modulate the transcription of CYP2D49. (A) Mutation and deletion analyses of the CYP2D49 proximal promoter in LMH cells. A series of mutated and deleted CYP2D49 promoter fragments were generated and transfected into LMH cells as described in 'Experimental procedures'. The nucleotides selected for mutagenesis were the same as those used in the EMSAs. Schematic illustrations of the CYP2D49 reporter constructs are shown on the left; the different binding sites for the transcription factors are shown as shaded when mutated. The firefly luciferase activity for each construct was normalized to the Renilla luciferase activity, and is expressed as the fold change relative to the pGL3-Basic plasmid. All experiments were performed in triplicate, and the values given represent the mean ± SD of three independent experiments, each of which represents the average of three wells. Statistical significance was defined as **P < 0.01 or not significant (NS). (B) Competition assays showed that C/EBPα has similar binding affinities for site A and site B. EMSAs were performed as described in 'Experimental procedures'. Nuclear extracts from chicken liver tissues were incubated with biotinylated probe A, B, or C. For competition experiments, a 10-fold or 100-fold molar excess of the indicated unlabeled oligonucleotide was added to the binding reaction.

Mutagenesis of the DR1 element or the two C/EBP-binding sites alone resulted in only a 50% decrease in luciferase activity as compared with the wild-type construct, whereas mutation of the three elements together led to a further decrease in luciferase activity (Fig. 4A). The results of the mutagenesis analysis suggest that the DR1 element and the two C/EBP-binding sites are important regulatory elements in the CYP2D49 promoter, and cooperatively regulate the expression of CYP2D49.

C/EBPα and HNF4α transactivate the CYP2D49 promoter

To investigate whether C/EBPα and HNF4α were capable of activating the CYP2D49 promoter, cotransfection experiments were performed in COS-7 cells. In contrast to the results obtained in LMH cells, the transfection of A-Luc, B-Luc or C-Luc into COS-7 cells did not induce luciferase activity (Fig. 5A–C). However, cotransfection of C/EBPα or HNF4α with A-Luc, B-Luc or C-Luc led to significant increases in luciferase activity, which was not observed when the constructs were cotransfected along with the empty pcDNA3.1 vector (Fig. 5A–C). Furthermore, the activating effect of C/EBPα or HNF4α was impaired by the mutation of the core C/EBPα-binding or HNF4α-binding site, but not by the mutation of adjacent unrelated sites (Fig. 5A–C). These results demonstrate that C/EBPα and HNF4α transactivate CYP2D49 by binding to the two C/EBP boxes and the DR1 element, respectively.

Figure 5.

C/EBPα and HNF4α transactivate the CYP2D49 promoter. (A–C) C/EBPα and HNF4α overexpressed in COS-7 cells transactivate the CYP2D49 promoter. COS-7 cells were cotransfected with a series of reporter constructs and their mutant counterparts, as well as the pcDNA–C/EBPα (A, B) and pcDNA–HNF4α (C) expression plasmids. The wild-type constructs (A-Luc, B-Luc, and C-Luc) contain the two C/EBP boxes and the DR1 element, respectively. The nucleotides mutagenized in mut1 were the same nucleotides that were mutated in the EMSAs, and the mut2 constructs, in which the two C/EBP boxes and DR1 element remain intact, were used as controls. Each value represents the mean ± SD of three experiments. Statistical significance was defined as **P < 0.01 or not significant (NS). (D) The effects of C/EBPα and HNF4α on CYP2D49 expression in LMH cells. LMH cells were first transfected with different expression plasmids. After 24 h, the cells were collected for RNA extraction and real-time PCR. The ratio of CYP2D49 to β-actin in the control cells was set to 1, and the values of all treated cells were normalized to this value. The experiments were conducted in triplicate, and the data are expressed as the mean ± SD. Statistical significance was determined with ANOVA, and significance was defined as *P < 0.05 or **P < 0.01.

Next, we further examined the roles of C/EBPα and HNF4α in the activation of CYP2D49 mRNA in a physiological environment. LMH cells were transiently transfected with different expression vectors. As shown in Fig. 5D, the mRNA level of CYP2D49 was elevated significantly when cells were transfected with C/EBPα or HNF4α expression plasmids (P < 0.05 for both). We next investigated the effect of the two factors together, and found additive upregulation of CYP2D49 mRNA (Fig. 5D). These results strengthen our hypothesis that both C/EBPα and HNF4α contribute significantly to sustaining a high level of CYP2D49 transcription.

Discussion

Chicken CYP2D49, similarly to human CYP2D6, is an important drug-metabolizing and noninducible gene, and is expressed predominantly in the liver [15]. The goal of the present study was to gain insights into the molecular mechanism of the constitutive expression of CYP2D49 in the liver. Despite the complexity of this mechanism, our results offer solid evidence that both C/EBPα and HNF4α mediate the transcriptional activation of CYP2D49. To our knowledge, this is the first study to comprehensively show how multiple transcription factors regulate the basal expression of CYP2D49.

HNF4α is a member of the nuclear receptor superfamily that is involved in the liver-specific regulation of many CYP genes encoding major drug-metabolizing enzymes, including CYP2A, CYP2C, CYP2D, and CYP3A [16]. HNF4α acts as a homodimer, and binds to sequences in the proximal promoter regions that contain direct repeats of the hexamer AGGTCA, separated by a single base (DR1). Moreover, HNF4α is also critical for the appropriate drug-mediated induction of CYP genes, such as CYP3A4 or CYP2C9, through crosstalk with CAR or PXR [16]. A positive DR1 element was identified in the human CYP2D6 promoter, and this element bound and responded to HNF4α [17]. The conditional inactivation of HNF4α in a CYP2D6-humanized transgenic mouse line decreased debrisoquine 4-hydroxylase activity by > 50%, further confirming that HNF4α regulates the expression and function of CYP2D6 [18]. Moreover, HNF4α also mediates the downregulation of CYP2D6 during inflammation caused by nitric oxide through the same proximal DR1 site [19]. Similarly, an imperfect DR1 element was identified in the proximal promoter region of CYP2D49, and this motif specifically binds to HNF4α and mediates the transcriptional activation of CYP2D49 to a large extent.

The DR1 element can be recognized not only by HNF4α, but also by other members of the steroid receptor family, including chicken ovalbumin upstream promoter transcription factors (COUP-TFs), peroxisome proliferator-activated receptor, and retinoid X receptor [20, 21]. In the human CYP2D6 promoter, two other protein·DNA complexes, excluding HNF4α·DR1, can be detected, which indicates that other transcription factors interact with the DR1 element [17]. Moreover, a negative effect of COUP-TFI on the transactivating activity of HNF4α was detected in the CYP2D6 promoter; this was most likely attributable to the competitive binding of COUP-TFI to the same HNF4α-binding sequence (DR1) [17]. In fact, Lutz et al. [22] reported that high expression levels of chicken COUP-TFII were detected during embryonic development, especially in the developing spinal motor neurons, whereas relatively low expression levels were observed in tissues, such as the liver and heart. These authors also found that the COUP-TFII present in the developing spinal motor neurons is capable of binding to a perfect DR1 element, and is ultimately involved in motor neuron development [22]. However, in the present study, we did not detect the binding of any other proteins to the DR1 element when chicken liver nuclear extracts were used in the gel shift assay. Moreover, chicken COUP-TFII cannot bind to the DR1 element in the CYP2D49 promoter when overexpressed in COS-7 cells (data not shown). On the basis of these data, we suggest that chicken COUP-TFII may not be involved in the regulation of CYP2D49 in the liver, and that the DR1 element may undergo differential occupancy by various transcription factors in a tissue-specific manner. Recently, Fang et al. identified an HNF4α-specific binding motif, xxxxCAAAGTCCA, which is bound by HNF4α but not by other DR1-binding factors, including COUP-TFII and retinoid X receptor α [23]. We found that the imperfect DR1 element in the CYP2D49 promoter fits nearly completely (one mismatch) with the newly identified HNF4α-specific binding motif, which may explain the observed specificity for HNF4α as well.

In addition to the DR1 element, two adjacent imperfect C/EBP boxes were identified in the proximal promoter region of CYP2D49. Both bind equally to C/EBPα and work in a compensatory but not redundant manner to mediate the transcriptional expression of CYP2D49. However, in the human CYP2D6 promoter, no C/EBP box was identified in the proximal promoter region [17]. Only one C/EBP box was found in the –1231/–1220 region, and this motif mediates the transactivation of CYP2D6 through the binding of C/EBPα [24]. To our knowledge, C/EBPα is a member of the basic region leucine zipper family of transcription factors, which is critical for normal cellular differentiation and function in a variety of tissues [25]. The regulation of CYP genes by C/EBPα has not been thoroughly explored. In the liver, C/EBPα has been previously shown to bind to the multiple C/EBP boxes in the promoter and regulate the constitutive or induced expression of CYP2A6, CYP2B6, CYP2C9, CYP2D6, and CYP3A4 [26-29]. Cooperative binding and interaction between C/EBPα and other transcription factors is a common characteristic in the regulation of CYP genes. For example, the interaction between HNF4α and C/EBPα leads to a cooperative effect on human CYP2A6 transcription [29]. C/EBPα and HNF4α synergistically cooperate with CAR to transactivate human CYP2B6. Differentiated embryo chondrocyte-2 interacts with C/EBPα to regulate the time-dependent expression of human CYP2D6 [24]. Here, we not only discovered the involvement of C/EBPα in the transcriptional activation of CYP2D49 through binding to the two C/EBP boxes, but also found that C/EBPα can work with HNF4α to transactivate CYP2D49. However, whether these two transcription factors interact and produce a synergistic effect on the transcriptional regulation of CYP2D49 is not yet known.

As we know, the C/EBP family consists of six isoforms – C/EBPα, C/EBPβ, C/EBPγ, C/EBPδ, C/EBPε, and C/EBPζ – which bind to the cognate C/EBP consensus sequences [25]. As well as C/EBPα, C/EBPβ has also been found to be involved in the transcriptional regulation of CYP genes, including human CYP3A4, rat CYP2D5, and CYP7A1, through binding to the C/EBP boxes in their promoters [6, 30]. However, we did not detect the binding of any proteins other than C/EBPα to the C/EBP boxes found in the promoter of CYP2D49 when chicken liver nuclear extracts were used in the gel shift assay, as the single protein·DNA complex was completely supershifted by the inclusion of an antibody against C/EBPα. This indicates that other C/EBP family members, such as C/EBPβ, may not bind to the C/EBP boxes in the promoter of CYP2D49. In fact, we also cotransfected C/EBPβ with A-Luc or B-Luc in COS-7 cells, and did not find significant increases in luciferase activities as compared with the control (data not shown). All of these findings indicate that C/EBPβ may not be involved in the regulation of CYP2D49.

In addition to C/EBPα and HNF4α, several additional factors may participate in the regulation of CYP2D49. For example, we found that regions –2300/–1900 and –350/–296 of the CYP2D49 promoter may contain several elements that negatively regulate the expression of CYP2D49. Indeed, overexpressed C/EBPα and HNF4α induced the endogenous expression of CYP2D49 to a lower than expected extent; this may be attributable to negative regulation. Additionally, we cannot rule out the possibility that elements outside of the studied region may also contribute to gene transcription. Therefore, further studies are needed to identify additional cis-acting elements and trans-acting factors that may be involved in the transcriptional regulation of CYP2D49.

In conclusion, the present study constitutes the first comprehensive functional analysis of the chicken CYP2D49 promoter, and clearly demonstrates that C/EBPα and HNF4α cooperatively regulate CYP2D49 transcription. Further studies are required to examine the molecular mechanisms modulating CYP2D49 expression.

Experimental procedures

Ethics statement

The experiments were performed in strict accordance with the recommendations in the Regulations for the Administration of Affairs Concerning Experimental Animals of Guangdong Province, China. All surgery was performed under sodium pentobarbital anesthesia, and all efforts were made to minimize suffering.

Animals and cell culture

Three-yellow broiler chickens (7–8 weeks of age, 1.2–1.5 kg) were purchased from the College of Veterinary Medicine at South China Agricultural University. Chicken liver tissues were collected, snap-frozen in liquid nitrogen, and stored at − 80 °C until use.

LMH cells (ATCC, CRL-2117) and COS-7 cells (ATCC, CRL-1651) were maintained at 37 °C in William's E medium and DMEM, respectively, supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 mg/mL streptomycin (all from Invitrogen, Carlsbad, CA, USA).

Construction of reporter plasmids and expression vectors

To construct the luciferase reporter vector for chicken CYP2D49, a 2.3-kb fragment (–2300 to –1; +1 indicates the translational start site) derived from the 5′-flanking region of chicken CYP2D49 (GenBank accession no. NW_003763476) was amplified by PCR from the genomic DNA of chicken liver tissues. The PCR product was then purified and ligated into the pGL3-Basic vector (Promega, Madison, WI, USA) to generate the –2300-Luc construct. A series of 5′-promoter deletion constructs were made on the basis of the –2300-Luc construct: –1900-Luc, –1400-Luc, –1000-Luc, –730-Luc, –550-Luc, –350-Luc, –296-Luc, –274-Luc, –226-Luc, and –183-Luc. Additional deleted constructs, including (–296 to –183)-Luc, (–274 to –183)-Luc, (–226 to -183)-Luc, AC-Luc, AB-Luc, A-Luc, B-Luc and C-Luc, were prepared on the basis of the –296-Luc construct with the QuikChange Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA) and the corresponding primer pairs. All mutations were then introduced into the constructs above by the use of appropriate mutated oligonucleotides. The pcDNA–C/EBPα and pcDNA–HNF4α expression plasmids were generated by PCR with corresponding primers and cDNAs prepared from chicken liver tissues. The resulting amplicons were subcloned into the XhoI and HindIII sites of the pcDNA3.1/myc-His(−) A vector (Invitrogen). All plasmids were verified with DNA sequencing, and were prepared with the EndoFree Plasmid Kit (Omega, Norcross, GA, USA).

Transfection and luciferase activity detection

LMH cells and COS-7 cells were grown in monolayers in 24-well plates at a density of 2 × 104 cells per well at 37 °C in 5% CO2 for 1 day prior to transfection. The cells in each well were transfected with 0.6 μg of the reporter construct, 60 ng of the pRL-TK control plasmid (Promega) and 60 ng of the expression vectors (if appropriate) with 2 μL of Lipofectamine 2000 (Invitrogen), according to the manufacturer's instructions. After 24 h, the cells were lysed, and the luciferase activities were measured with the Dual-Luciferase Reporter Assay System (Promega) on a Turner Designs TD-20/20n luminometer (Promega). The amount of firefly luciferase activity was normalized to the amount of Renilla luciferase activity. The results shown are representative of at least three independent experiments (each performed in triplicate).

To determine the amount of endogenous CYP2D49 present at the mRNA level, LMH cells were transfected with 60 ng of pcDNA, pcDNA–C/EBPα, pcDNA–HNF4α, or a mixture of pcDNA–C/EBPα and pcDNA–HNF4α (equal amounts of both). After 24 h, the cells were collected for RNA extraction and real-time PCR. For EMSA analysis, COS-7 cells were first transfected with the different expression vectors and then subjected to nuclear protein extraction.

RNA isolation and real-time PCR

Total RNA was isolated from LMH cells with the TRIZOL reagent (Invitrogen). First-strand cDNAs were synthesized with the Prime Script RT reagent kit (TaKaRa, Qingdao, China). Gene-specific primer pairs, which were designed to span exon–exon junctions, were used in the real-time PCR reactions to avoid potential genomic DNA contamination. The sequences of the primer pairs are listed in Table 1. Real-time PCR was performed with a BioRad CFX96 real-time PCR detection system (Bio-Rad, Hercules, CA, USA), according to the manufacturer's recommendations. Reactions were performed in 20-μL volumes containing SYBR Green I Dye. The following cycling parameters were used: 94.0 °C for 2 min, followed by 40 cycles of 94.0 °C for 30 s, 58.0 °C for 30 s, and 72.0 °C for 30 s. All samples were analyzed in triplicate, and the expression of the target gene was calculated relative to the expression of β-actin according to the 2−ΔCT method [31].

EMSA

Nuclear extracts from chicken liver tissues or COS-7 cells were prepared with the Nuclear Extraction Kit (Beyotime, Haimen, China). EMSA analysis was performed with the EMSA Assay Kit (Beyotime). In brief, binding reaction mixtures containing nuclear extracts (4 μg) and binding buffer were preincubated at 20 °C for 10 min in a total volume of 10 μL. Double-stranded biotinylated probes (0.1 pmol) were then added, and the mixtures were further incubated at 20 °C for 20 min. The protein·DNA complexes and the unbound free probe were separated on 4% nondenaturing polyacrylamide [acrylamide/bisacrylamide 29 : 1 (v/v)] gels, and were detected by the use of chemiluminescence (Millipore, Bedford, MA, USA). For competition experiments, a 100-fold molar excess of unlabeled probe was incubated with the nuclear extracts for 10 min before addition of the labeled probes to verify the specificity of the protein–DNA interactions. For supershift experiments, 1 μg of polyclonal antibody against HNF4α (sc-6557X; Santa Cruz Biotechnology, Santa Cruz, CA, USA), polyclonal antibody against C/EBPα (sc-9315X; Santa Cruz Biotechnology) or goat IgG was added to the preincubation mixtures.

ChIP assay

ChIP assays were performed with the ChIP Assay Kit (Beyotime) according to the manufacturer's recommendations. Briefly, LMH cells cultured to 80% confluence were fixed at 37 °C for 15 min in the presence of 1% (w/v) formaldehyde, and this was followed by neutralization with 125 mm glycine. Then, the cells were collected, suspended in SDS lysis buffer (1% SDS, 10 mm EDTA, 150 mm NaCl, 50 mm Tris/HCl, 1 mm phenylmethanesulfonyl fluoride, pH 8.0) and sonicated on ice to shear the genomic DNA into 200–1000-bp fragments. The sonicated samples were centrifuged at 14 000 g for 5 min at 4 °C to clear the supernatants. The DNA content was quantified and diluted to maintain an equivalent amount of DNA in all of the samples (input DNA). For immunoprecipitation of the HNF4α·DNA and C/EBPα·DNA complexes, 2 μg of the specific antibody was incubated with 60 mg of protein A+G agarose/salmon sperm DNA and 2 mL of precleared lysate at 4 °C overnight. DNA was isolated from the immunoprecipitates with the standard phenol/chloroform method, and subjected to PCR with the primers listed in Table 1. Amplification was monitored in real time, stopped when exponential amplification was detected, and analyzed with 1.5% agarose gel electrophoresis. Negative control ChIPs were performed in the absence of antibody or with goat IgG. No PCR products were amplified from these samples, as determined by ethidium bromide staining. ChIP assays were performed three times independently, and the representative electrophoretic images of PCR products were used in the figures.

Statistical analysis

All experiments were performed independently at least three times. All data were expressed as the mean ± standard deviation (SD). Statistical significance was evaluated by ANOVA, and significance was defined as P < 0.05 or P < 0.01.

Acknowledgements

This work was supported by the National Basic Research Program of China (973 program) (2009CB118802), the National Natural Science Foundation of China (31172087 and 31201716), the Specialized Research Fund for the Doctoral Program of Higher Education (SRFDP) (20114404110010), and the Guangdong Natural Science Foundation (S2012040007667).

Ancillary