Abstract
The regulation mechanism of female-predominant expression of the mouse Cyp2b9 gene was investigated in vivo and in vitro. Luciferase reporter assay revealed that the –234/–194 region of the Cyp2b9 gene may be responsible for sexually dimorphic expression. There is a predicted forkhead box A2 (FoxA2) (hepatic nuclear factor 3β)-binding site in this region. Chromatin immunoprecipitation assay indicated that the binding protein to the site was FoxA2 in 5-week-old female mice, whereas this protein was found in both sexes at age 3 weeks, in accordance with our previous observation on the developmental expression of this gene. Mutation of the predicted FoxA2 site in the reporter construct containing the –234/+18 fragment led to complete elimination of luciferase activity, but deletion of the –234/–194 region resulted in considerable transcriptional activity, suggesting that by mutating the FoxA2-binding site a potent suppressor might bind to eliminate activity, whereas by deleting this region it could not. Sexually dimorphic secretion of growth hormone is involved in female-predominant expression of the gene, and the –234/–194 region was also responsible for suppressing the expression by male-type secretion.
Cytochrome P450 (P450) is involved in the metabolism of drugs, chemical carcinogens, and endogenous chemicals, such as fatty acids, prostaglandins, and steroid hormones (Gonzalez, 1991). It is generally recognized that the major reason for the adverse effect of a single drug or drug-drug interaction is P450-mediated metabolism. Furthermore, many endogenous and exogenous factors have been reported that modify the expression of P450, resulting in the alteration of pharmacological activity (Watkins et al., 1989; Jürgens et al., 2002). Although large individual variation of the expression is observed in almost all the P450s in humans, several P450s were found to be expressed in a sex-dependent manner. For example, CYP3A4, which is expressed as a major P450 in the liver and recognizes nearly half of the presently used drugs as substrates, is found to express female dominance by analyses of mRNA, substrate-metabolizing activity in surgical liver samples (Wolbold et al., 2003). Clinical investigations have also reported the existence of either male- or female-dominant P450 species (Lamba et al., 2003; Gandhi et al., 2004).
In laboratory animals, especially rodents, the expression of many P450s is found to be clearly dependent on sex, and factors and mechanisms for sexually dimorphic expression have been extensively investigated, indicating that the differences are regulated by sex-specific secretion profiles of growth hormone (GH) (Waxman and O'Connor, 2006). Male rats secrete GH in episodic bursts (∼200–300 ng/ml of plasma) every 3.5 to 4 h, and GH levels between the peaks are undetectable. In females, hormone pulses are more frequent and irregular and are of lower magnitude than those in males, whereas interpulse concentrations of GH are always measurable. Mice also show sexually dimorphic GH-secretory patterns, with females characterized by more frequent GH pulses and a shorter GH-free interpulse interval than males (MacLeod et al., 1991). These sex differences in circulating GH profiles, and not sexual differences in GH concentrations, per se, are responsible for the sexual dimorphism of P450 expression. After GH binds to its receptor localized in the cell membrane, Janus kinase 2 (JAK2), which is a tyrosine kinase and binds to the cytoplasmic domain of the GH receptor, is activated. Activated JAK2 then activates signal transducer and activator of transcription (STAT) 5b or insulin receptor substrates 1 and 2 (IRS1/2) by tyrosine phosphorylation. Thus, STAT5b recognizes the GH secretion profile and is activated by male patterns (Waxman and O'Connor, 2006).
Many Cyp genes were found to exhibit sexually dimorphic expression in mouse liver. Well characterized examples of male-specific mouse P450s include CYP2D9, CYP4A12, CYP7B1, and CYP8B1. Female-specific mouse P450s include CYP2A4, CYP2B9, CYP2B10, CYP2B13, CYP3A16, CYP3A41, and CYP3A44. CYP2B9 is a major P450 in the mouse liver, and its expression changes developmentally to female predominance in adults (Jarukamjorn et al., 2002). We previously observed that the expression of CYP2B9 was elevated in hypophysectomized male mice to the level seen in females, but the expression was reduced to the same level as in male liver after treatment with GH mimicking the male-specific profile (Sakuma et al., 2004). The expression increased in STAT5b-knockout male mice (Holloway et al., 2007). These observations suggest that female-predominant expression of CYP2B9 is regulated by the GH-STAT5b pathway and that the male-dependent secretion profile of GH works suppressively.
Furthermore, several kinds of hepatocyte-enriched nuclear factor (HNF) have been reported to act as regulating factors in the expression of sexually dimorphic P450s (Holloway et al., 2006). Among HNFs, HNF4α, known as an orphan receptor, is found to be a regulation factor for the expression of CYP3A4 (Jover et al., 2001). In HNF4α-knockout male mice, mRNAs of CYP2A4 and CYP2B9, which express female predominance in intact animals, increased significantly (Wiwi et al., 2004). In contrast, CYP3A41 and CYP3A44 mRNAs, the expressions of which are also female-predominant in intact mice, were not changed in males, whereas they decreased markedly in females (Wiwi et al., 2004); therefore, one transcription factor does not necessarily lead to the expression of different genes in an identical direction.
In the present study, we assayed reporter gene activity in the body because activity was not detected in cultured hepatocytes. The main reason for undetectable activity in vitro might be the low expression of essential factors for transcription. Establishing the same conditions as in vivo is generally difficult in cultured cells; therefore, the expression of the reporter gene was analyzed in vivo with the hydrodynamic method (Liu et al., 1999) to determine the responsible element in the 5∼-flanking region of the Cyp2b9 gene for sexually dimorphic expression.
Materials and Methods
Materials. Materials for the isolation and culture of hepatocytes were purchased from Wako Pure Chemical (Osaka, Japan) and Gibco-BRL (Grand Island, NY). Percoll and collagenase (type I) were products of GE HealthCare UK Ltd. (Buckinghamshire, UK) and Sigma-Aldrich (St. Louis, MO), respectively. TaKaRa RNA polymerase chain reaction (PCR) kit (AMV) version 2.1 was obtained from TAKARA Shuzo (Kyoto, Japan). Transfection reagent (Trans IT-EE hydrodynamic delivery solution) was obtained from Mirus Bio Corporation (Madison, WI). Recombinant human GH (rhGH) was obtained from Wako Pure Chemical, and an Alzet micro-osmotic pump was from Durect (Cupertino, CA). All the primers were commercially synthesized at Hokkaido System Sciences (Sapporo, Japan). Other chemicals of molecular biology grade were purchased from Wako Pure Chemical, Sigma-Aldrich, or Roche Diagnostics (Indianapolis, IN).
Animals. ddY mice (4 weeks old) of both sexes, weighing 25 to 30 g, were supplied by Sankyo Experimental Animals (Tokyo, Japan). Animals were housed for 1 week in the Laboratory Animal Center of University of Toyama under the supervision of certified laboratory veterinarians and were treated according to the research protocol approved by the university's Institutional Animal Care and Use Committee. At 5 weeks of age, some mice received twice-daily s.c. injections (50 μg/mouse) of rhGH for 7 days. Infusion with an osmotic minipump, Alzet 1007, implanted into the back of the mice, was designed to release contents at a rate of 1.5 μg/h. GH administration was checked by examining body weight gain. On the 7th day, a solution containing a reporter plasmid was injected into the tail vein, as described in a later section. Animals were sacrificed on the 8th day after the start of GH treatment, and the livers were excised immediately to prepare liver homogenates. Sankyo Experimental Animals supplied hypophysectomized or sham-operated 6-week-old C57BL/6 mice. At 7 weeks of age, some male mice received an s.c. injection of rhGH twice a day, and some female mice received the infusion with an osmotic minipump. Verification of hypophysectomy was confirmed by a lack of weight gain and the absence of pituitary fragments during necropsy after sacrifice. We have not observed any evident difference between two mouse strains for female-specific expression of the Cyp2b9 gene. Although GH treatment of each strain started at different ages, namely, 5 weeks old for ddY and 7 weeks old for C57BL/6, the expression of the Cyp2b9 gene in the liver had become sexually dimorphic by those ages.
Preparation of Primary Hepatocyte Cultures. The livers of ddY mice were perfused with collagenase-containing Hanks' solution, and viable hepatocytes were isolated by means of Percoll isodensity centrifugation as described (Nemoto and Sakurai, 1995). Hepatocytes isolated from two or three mice were pooled before Percoll isodensity centrifugation. Standard culture conditions were as follows: the cells were dispersed in Waymouth MB 752/1 medium containing bovine serum albumin (2 g/l), transferrin (0.5 mg/l), and selenium (0.5 μg/l), and seeded in dishes at a density of 5 × 105 cells/1 ml/35-mm dish. The Waymouth medium did not contain phenol red, a pH indicator, to exclude estrogen-like action. The culture dishes were maintained at 37°C in a CO2-humidified incubator. The medium was renewed 24 h after seeding. The cells were harvested on the day indicated in the figures to prepare the total RNA fraction.
Isolation of the Cyp2b9 Gene and Construction of Reporter Plasmids. A mouse Cyp2b9 gene fragment containing 2.4 kilobase pair of the 5∼-flanking region and part of the first exon (–2382/+18) was isolated using the Mouse Genome Walker Kit (Clontech Laboratories, Inc., Mountain View, CA) and then subcloned into a pGEM T-easy vector (Promega, Madison, WI). The resultant plasmid was digested with BsaWI (at +18), blunted by T4 DNA polymerase, digested with MluI (at multicloning sites in the vector), and then the modified gene fragment was inserted into the MluI/SmaI site of the pGL3-basic vector (Promega) (i.e., –2382/+18-Luc). A series of plasmids containing progressive deletion fragments (–1167/+18-Luc, –234/+18-Luc, –193/+18-Luc) was constructed using restriction endonuclease digestion at internal recognition sites. Two forkhead box A2 (FoxA2) site-mutated reporters (–234/+18-mut 1-Luc, –234/+18-mut 2-Luc) were constructed by insertion of PCR products amplified with mutation-containing primers into KpnI/HindIII site of the pGL3-basic vector. Sequences of the mutated sense primers were 5∼-GGTACCGCCCTAGGCTCGCATGTATCTGCC-3∼ and 5∼-GGT-ACCGCCCGGTCGTCGCATGTATCTGCC-3∼ for –234/+18-mut 1-Luc and –234/+18-mut 2-Luc, respectively. Italics or boldface indicates the KpnI site or mutated FoxA2 sites, respectively. Antisense primer used for amplification of the mutated fragments was commercially supplied (GLprimer2, Promega) and binds to the vector sequence downstream of a cloning site. The FoxA2 site 5∼-TATTT-3∼ (–207/–203) changed to 5∼-TAGGC-3∼ and 5∼-GGTCG-3∼ in –234/+18-mut 1-Luc and –234/+18-mut 2-Luc, respectively. The orientation of the six constructs was verified by restriction enzyme digestion or DNA sequencing. The sequences and positions of the FoxA2-binding site in the Cyp2b9 gene were predicted by Match-public (http://www.gene-regulation.com) and TFSEARCH (http://mbs.cbrc.jp/research/db/TFSEARCH.html).
Hydrodynamic Infusion of the Reporter Plasmid. Six-week-old ddY mice were given a rapid (5 s) tail vein injection of pGL3-basic vector (empty) or reporter constructs containing several sizes of the 5∼-flanking region of the Cyp2b9 gene (10 μg) and phRL-SV40 vector (0.5 μg), an internal standard, dissolved in transfection reagent in a volume equal to 10% of body weight. After 20 h, the mice were sacrificed, and the livers were homogenized in 5-times volume of the lysis buffer (0.1 M Tris-HCl, 2 mM EDTA, 0.1% Triton X-100). The homogenates were then centrifuged (15,000g, 4°C, 10 min.). An aliquot of the supernatant was diluted 60 times with HEPES, followed by determination of luciferase activity.
Luciferase Assay. Luciferase assay was performed by the dual-luciferase reporter assay system (Promega) using the method recommended by the supplier, and luminescence was determined with a TD-20/20 luminometer (Promega). Values of firefly luciferase activity were normalized by those of Renilla luciferase activity in individual experiments.
Construction and Transfection of a FoxA2 Expression Plasmid into Hepatocytes in Cultures. The FoxA2 expression plasmid was generated by replacing the DNA fragment between NheI and XbaI sites containing the coding sequence of Renilla luciferase of pRL-SV40 vector (Promega) with the 1946-base pair cDNA fragment of the entire coding region of mouse FoxA2. Mouse hepatocytes were cultured in Waymouth medium and transfected using Transpass D1 Transfection Reagent (New England Biolabs, Hercules, CA). Transfection mixtures consisted of Waymouth medium, empty plasmid or expression plasmid, and Transpass D1 at 2 ml, 5 μg, and 5 μl, respectively. Transfection continued for 3 h, and the medium was changed. Total RNA was prepared after 24 h.
Total RNA Preparation and Quantitative PCR. Total RNA was prepared from liver tissues or hepatocytes of mice using TRIzol reagent (Invitrogen, Carlsbad, CA). The expression of CYP2B9 mRNA was analyzed by quantitative reverse transcription-PCR (RT-PCR) using a TaKaRa RNA PCR kit (AMV) version 2.1 and a gene-specific TaqMan MGB gene expression detection kit. The forward primer, reverse primer, and TaqMan MGB probe of the TaqMan MGB gene expression detection kit for CYP2B9, designed by ourselves with assistance from Primer Express software, were 5∼-CACAGATGACCAGTTCCTTCATCT-3∼, 5∼-GTTCCTGCTGTTTTTTGACAATTT-3∼, and 5∼-FAM-CTCTGGTCA-GATGTTTGAG-MGB -3∼, respectively. The expression of FoxA2 mRNA was analyzed by TaKaRa RNA PCR kit (AMV) version 2.1 and a TaqMan gene expression assay for FoxA2 (Mm00839704_mH). Detection of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA amplification was carried out using the TaKaRa RNA PCR kit (AMV) version 2.1, SYBR green reagent, and the primer set specific to mouse GAPDH cDNA (Jarukamjorn et al., 2006). PCR was performed using the ABI Prism 7000 Sequence Detection System (Applied Biosystems, Foster City, CA) with ABI Prism 7000 SDS software. CYP2B9 or FoxA2 mRNA expression was normalized to the level of GAPDH.
Chromatin Immunoprecipitation Assay. The chromatin immunoprecipitation (ChIP) assay kit was purchased from Upstate Biotechnology (Lake Placid, NY). Anti-FoxA2 antibody (sc-9187x) and normal goat IgG (control) were from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-STAT5b antibody (RB-096-PO) was from NoeMarkers (Fremont, CA). ChIP assay was performed according to the manufacturer's protocol as follows. The livers of four ddY mice were perfused, and viable hepatocytes were isolated by Percoll isodensity centrifugation. Cells obtained from four mice were pooled before Percoll isodensity centrifugation. The cells were dispersed in Waymouth MB 752/1 medium, seeded in dishes at a density of 1 × 106 cells/10 ml/100-mm dish, and then cross-linked using 1% formaldehyde at 37°C for 15 min in a CO2-humidified incubator. After treatment, the hepatocytes were suspended in 200 μl of SDS lysis buffer (50 mM Tris-HCl, pH 8.1, 10 mM EDTA, 1% SDS, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, and 1 μg/ml pepstatin A) for 10 min on ice. The lysate was sonicated to shear the DNA and then centrifuged for 10 min at 13,000 rpm and 4°C. The supernatant was diluted to 1 ml with ChIP dilution buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl, pH 8.0, 167 mM NaCl, 1 mM phenylmethylsulfonyl fluoride, 1 μg/ml aprotinin, and 1 μg/ml pepstatin A), followed by preclearing with Salmon Sperm DNA/protein A-agarose beads, and then immunoprecipitated with 10 μl of antibody overnight at 4°C with rotation. Aliquots of the supernatant were stored to use for input control. Chromatin-antibody complexes were collected with 60 μl of Salmon Sperm DNA/protein A-agarose beads and washed once with 1 ml of each of two kinds of buffers, and then washed twice with 1 ml of Tris/EDTA buffer. Immune complexes were eluted using elution buffer (1% SDS and 0.1 M NaHCO3). Cross-links were reversed by heating at 65°C for6hinthe presence of NaCl, and then the eluates were treated with proteinase K at 45°C for 2 h. DNA was recovered from the eluate into 50 μl of distilled water by Wizard SV Gel and PCR Clean-up System (Promega). Conditions for PCR reactions were as follows: 2 μl of DNA sample, 0.3 μM each primer, 0.2 mM deoxynucleoside-5∼-triphosphate, 1× PCR buffer, and 0.625 units of Prime STAR DNA polymerase (TAKARA) in 25 μl of reaction volume. The forward and reverse primers to amplify a DNA fragment covering –374/–63 in the 5∼-flanking region of the Cyp2b9 gene were 5∼-ACATTTAGCAGAACACAGAGGAACACA-3∼ and 5∼-CCAACTGTTCCCACCTCCTGTCTCACACTG-3∼, respectively. PCR was performed as follows: denaturation at 98°C for 10 s, annealing at 60°C for 15 s, and extension at 72°C for 20 s for 35 cycles. Aliquots of the reaction mixture (20 μl) were applied to a 7% polyacrylamide gel and visualized by ethidium bromide staining for 15 min. The experiment was repeated twice to confirm the result.
Statistical Analysis. Data are plotted as the mean ± S.D. Statistical analyses of the data shown in Figs. 2, 3, 5, 6, and 10 were carried out by analysis of variance (ANOVA) with the Tukey's test. Statistical analyses of the data shown in Figs. 7 and 8 were carried out by unpaired Student's t test. Statistical analyses of the data shown in Figs. 1 and 9, taken from an expression vector- or reporter construct-transfected hepatocytes growing in different dishes, were carried out by nonparametric tests, the Kruskal-Wallis test and the Mann-Whitney U test for Figs. 1 and 9, respectively. A value of p < 0.05 was considered significant.
Results
Determination of Regulatory Region of Cyp2b9 Gene Expression by in Vivo Reporter Gene Assay. A 2.4-kilobase fragment of the 5∼-flanking region of the mouse Cyp2b9 gene was isolated, and then digested pieces were ligated to the firefly luciferase gene. Luciferase activity was determined after transfection to female or male mouse hepatocytes in primary culture. As shown in Fig. 1, no significantly different activity, compared with that of an empty vector, was found in any sets of constructs even in female hepatocytes, and a similar tendency was found in male hepatocytes (data not shown). This observation indicates that the transcription rate driven from the Cyp2b9 gene fragment in reporter constructs is extremely low in cultured cells. Although the primary cultured female hepatocytes expressed CYP2B9 mRNA, the level was far lower than that in vivo (Fig. 2), suggesting that certain factor(s) for maintaining actual transcription would be drastically reduced after transferring hepatocytes to a primary culture from the body, and would prove inappropriate to determine the regulatory region of the gene by the usual reporter assay technique using a cultured cell system.
To determine the regulatory element of the 5∼-flanking region of the Cyp2b9 gene, we introduced the indicated reporter constructs into male or female mouse liver by injection into the tail vein. Luciferase activities of –2382/+18-Luc construct were not different between male and female mice, but those of –1167/+18-Luc or –234/+18-Luc were more than twice in females than in males (Fig. 3), although those of –193/+18-Luc were almost the same in both sexes. Figure 3 also indicates that deletion between –234 and –194 decreased luciferase activity in female mice, whereas it increased 20% in males. These observations suggest that the Cyp2b9 gene is transcribed by sex-independent factors, which function by binding to the –193/+18 region, and that transcription is modified by factors that can bind to the –234/–194 region.
The regulatory factor binding site survey was performed using the Match-public and TRANFAC program. The result is shown in Fig. 4. The predicted sequence was found at –207/–203 for FoxA2 (HNF3β), which was isolated as a transcription factor highly expressed in the liver (Spear et al., 2006). We then attempted to prove that the sequence between –207 and –203 was critical for female-predominant expression in the liver. Two types of nucleotide substitution in the predicted FoxA2 site entirely eliminated luciferase activity in both sexes (–234/+18-mut 1-Luc, –234/+18-mut 2-Luc) (Fig. 5). This finding suggests the large contribution of FoxA2 to the promoter activity of Cyp2b9 gene expression in both sexes.
In the next experiment, the possibility that GH treatment might affect the expression of luciferase activity was examined because the secretion profile of GH is known to be an essential factor to determine the sexually dimorphic expression of several genes (Waxman and O'Connor, 2006). Thus, the reporter plasmid was transfected into the tail vein to mice that had been receiving GH treatment for 6 days, and then reporter activity was determined 20 h after transfection. Figure 6 shows that GH treatment mimicking the male-type secretion profile decreased luciferase activities of the –234/+18-Luc construct in female mice, whereas GH treatment mimicking the female-type secretion profile did not increase activity in males. These observations suggest that male-type secretion of GH greatly affects the expression of the reporter gene in female mice, but female-type treatment of GH had no effect in males.
Relationship between Expression of CYP2B9 and FoxA2. Different sex-dependent secretion profiles of GH are established after puberty, and hormonal change may reflect the expression of the Cyp2b9 gene. Figure 7 shows that the expression of CYP2B9 mRNA in the liver increased after birth to the same level in both sexes until 3 weeks; however, the expression decreased drastically thereafter in male mice, whereas female mice maintained their expression level during the observation period. In contrast, the expression of FoxA2 mRNA in the liver of both sexes was considerable at birth and did not alter markedly during the observation period (data not shown). We could find no relationship in the alteration of expression between FoxA2 and CYP2B9 mRNAs in the liver.
Liver tissue is composed of parenchymal and nonparenchymal cells. Drug metabolism is considered to be performed mainly by parenchymal cells (hepatocytes), which occupy 70 to 80% of the liver. Consistent with liver tissue composition (Fig. 7), the expression of CYP2B9 mRNA in parenchymal cells was almost equivalent between males and females at 3 weeks but decreased in male cells at 5 weeks (data not shown). Consistent with this finding, female-dominant expression was observed on FoxA2 mRNA in parenchymal cells isolated from 6-week-old mice (Fig. 8, day 0), suggesting a relationship between FoxA2 and CYP2B9 mRNA expression in hepatocytes.
Because the expression of CYP2B9 mRNA decreased markedly in the primary culture of hepatocytes (Fig. 2), it is conceivable that the expression of the major predictable transcription factor may be reduced in the culture. Figure 8 also shows that the expression of FoxA2 mRNA decreased markedly on day 1 of culture but increased thereafter; however, the expression did not exceed that in male hepatocytes. After introducing the expression vector of FoxA2 into hepatocytes, CYP2B9 mRNA expression was enhanced 2.3-fold (Fig. 9).
It is known that the FoxA2 gene is regulated by GH in rats (Lahuna et al., 2000). To clarify the role of GH in the expression of the FoxA2 gene in mouse liver, the effects of hypophysectomy and GH replacement were examined (Fig. 10). Hypophysectomy of female mice significantly reduced mRNA expression by approximately 50%, and GH replacement by continuous infusion, which mimics the female type of GH secretion, recovered it, suggesting that the expression of FoxA2 mRNA is under the control of female-type secretion.
Because a relationship between CYP2B9 and FoxA2 mRNA expression was observed in liver cells, ChIP assay was performed using purified hepatocytes in the next experiment. As shown in Fig. 11A, the PCR product from FoxA2 antibody precipitates was observed in both sexes at 3 weeks but not in male hepatocytes at 5 weeks, suggesting that differences in the expression of CYP2B9 mRNA between males and females might be regulated by FoxA2 protein. Because the involvement of STAT5b in the suppression of some female-specific genes in males is known, and because a STAT5b binding site-like sequence was found near the FoxA2 site, ChIP assay using anti-STAT5b antibody was carried out. Male-predominant amplification of the PCR product was observed in hepatocytes from 8-week-old mice (Fig. 11B).
Discussion
The present investigations indicate that a certain proportion of constitutive transcription of the mouse Cyp2b9 gene may be mediated by FoxA2, which also regulates sexually dimorphic expression in the liver.
Five functional mouse Cyp2b genes, namely, Cyp2b9, Cyp2b10, Cyp2b13, Cyp2b19, and Cyp2b23, have been identified to date, and their products show high amino acid sequence homology among them. No information is available about their substrate specificity; therefore, an assay based on differences of nucleotide sequences is the most reliable method to analyze the expression of each gene separately. In this report, we investigated the expression of the Cyp2b9 gene at the mRNA level using quantitative RT-PCR. However, we recognize that both the mRNA level and reporter gene activity do not always reflect protein or enzyme activity levels, whereas post-transcriptional regulation of CYP2B genes has not been reported to our knowledge. Thus, the following discussion has limitations because of the lack of support by findings at protein or enzyme activity levels.
Our conclusion that FoxA2 participates in the sexually dimorphic expression of the Cyp2b9 gene is based on the following observations. One, that reporter constructs containing the –234/–194 fragment showed clear sex-dependent luciferase activity in vivo, and deletion of the fragment eliminated this difference; two, that a relationship between CYP2B9 expression and FoxA2 expression/binding activity could be seen, i.e., female-dominant expression of FoxA2 in hepatocytes of adult mice, induction of FoxA2 by female-type GH administration, increased CYP2B9 mRNA expression in primary hepatocyte culture by transfection of a FoxA2 expression plasmid, and age- and sex-dependent detection of FoxA2-DNA complex in ChIP assay.
The transcription factor binding site responsible for the expression of the Cyp2b9 gene was identified by transfection of reporter constructs into mouse liver by the hydrodynamic method. We could not have determined this if we had used cultured cells as the host because CYP2B9 expression is generally very low in cultured cells, even in female hepatocytes in primary culture. In Fig. 5, nucleotide substitution of the FoxA2-binding site in the 5∼-flanking region of the Cyp2b9 gene entirely eliminated transcriptional activity in both sexes. Deletion of the –234/–194 fragment containing this site reduced luciferase activity in females, but the decreased activity was still much higher than that of mutated constructs. In contrast, the deletion increased activity in males. These observations suggest that FoxA2 may be important for sexually dimorphic expression by binding to –207/–203 of the Cyp2b9 gene.
There should be a suppressive factor, which competitively binds to the site between –234 and –194, and the predictive suppressive factor might superiorly bind to the site after mutation of the FoxA2 site, leading to the elimination of transcription activity. FoxA2 and the suppressive factor could not bind to the –193/+18 construct, resulting in an expression transcribed by sex-independent factors at the same level in male and female mice. At present, the predicted suppressive factor has not been identified yet. In Stat5b knockout mice, CYP2B9 expression was evident in male mice (Holloway et al., 2007), indicating that STAT5b might be involved in the suppression of CYP2B9 expression in males. Interestingly, a STAT5b binding site-like sequence, 5∼-TTCgcatGTA-3∼, overlapping the 3∼ end of the FoxA2 site was found, and the PCR product from STAT5b antibody precipitates was observed predominantly in males at 8 weeks (Fig. 11B). Thus, STAT5b may be a candidate in males, whereas it is unlikely in females because of low activity of the transcription factor in the female liver (Wiwi et al., 2004).
FoxA2 is reported to be involved in the female-predominant expression of rat CYP2C12 mRNA (Delesque-Touchard et al., 2000); however, transcriptional activation of the CYP2C12 gene by FoxA2 is not high, and synergistic transcription was observed by combination with HNF6. Enhanced transcription by FoxA2 and HNF6 has been shown in several genes in the liver (Rausa et al., 2003). With CYP2B9, enhancement of the expression after transfection of a FoxA2 expression plasmid into cultured hepatocytes was at most 2-fold, suggesting that other fundamental transcription factor(s) might participate in the liver and might be deficient in the present culture system; however, RNA expression was not enhanced by transfection of either HNF6 alone or in combination with FoxA2 and HNF6, suggesting that similar synergistic transcription might not occur in the expression of the Cyp2b9 gene (data not shown).
In this study, we observed female-dominant and female-type GH administration-dependent expression of FoxA2 mRNA in females. On the other hand, hypophysectomy and GH replacement with male-type administration did not show significant effects in males. This finding suggests that FoxA2 gene expression is enhanced by GH only in females, resulting in female-dominant expression in hepatocytes. These findings suggest that GH can be involved in female-specific expression of the Cyp2b9 gene by modulation of FoxA2 expression. Similar GH-dependent female-dominant expression is observed in rat HNF6 and is believed to be a factor causing female-specific expression of the rat CYP2C12 gene (Delesque-Touchard et al., 2000).
Insulin inhibits the transfer of FoxA2 into the nucleus by inducing it to the phosphorylated form, which cannot move into the nucleus (Wolfrum et al., 2003); therefore, in hyperinsulinemia mice, FoxA2 protein exists as the phosphorylated form in cytoplasm and cannot function (Wolfrum et al., 2004). In streptozotocin-induced diabetic mice, we observed that the expression of CYP2B9 mRNA in males and the administration of insulin depressed the expression (Sakuma et al., 2001). Streptozotocin induces type I diabetes by disrupting pancreatic β cells, indicating that circulating insulin is deficient in mice. The reason why the expression of CYP2B9 mRNA was observed in diabetic male mice might be, in part, because of the presence of FoxA2 protein in the nucleus without the influence of insulin. Additionally, it is known that GH shows insulin-like action (Dominici et al., 2005). This action is caused by the phosphorylation (activation) of members of the IRS family by JAK2, which is activated by GH-bound GH receptor (Souza et al., 1994; Argetsinger et al., 1996; Yamauchi et al., 1998; Thirone et al., 1999). Similar activation by JAK2 can be seen for STAT5b, and this activation is stronger in males than in females (Waxman and O'Connor, 2006). Therefore, it is likely that the insulin-like action of GH is stronger in males, resulting in stronger inhibition of FoxA2 transfer into the nucleus in male hepatocytes. Therefore, two mechanisms prohibiting nuclear translocation by insulin and GH might be involved in the lower binding of FoxA2 in male hepatocytes, as indicated in the ChIP assay.
In our previous study, we showed that male-type administration of rhGH to intact female mice suppressed CYP2B9 mRNA expression in the liver (Sakuma et al., 2004; Jarukamjorn et al., 2006). This finding is in accordance with the result in the present experiment, observing the suppression of –234/+18-Luc reporter activity in female mice treated with the male-secretion profile of rhGH (Fig. 6). However, we could not change luciferase activity in male mice treated with the female-secretion profile of the hormone, whereas the same treatment could induce CYP2B9 mRNA up to half of the intact female level (Sakuma et al., 2004; Jarukamjorn et al., 2006). This treatment could also induce CYP3A41 and CYP3A44 mRNA expression to 4 and 20 times higher than the level in intact female mice, respectively (Sakuma et al., 2004). The reason for the low response of the reporter gene to this GH treatment is not clear at present, but we can exclude the regimen of GH treatment as a reason.
One possibility for the participation of FoxA2 protein and GH in transcription of the Cyp2b9 gene may be as follows: competitive binding near the FoxA2 site of a suppressor protein, which is activated by male-type secretion of GH, may result in stabilization of the chromatin structure. Furthermore, because we could not observe the sexually different expression of –2382/+18-Luc in the mouse liver (Fig. 3), the possibility that sexually dimorphic expression of the Cyp2b9 gene might be because of differences in the chromatin structure cannot be excluded.
In conclusion, this study determined the regulatory region of the mouse Cyp2b9 gene and showed that FoxA2 is a predictably responsible nuclear factor in the female-predominant expression of the gene. Other transcription factor(s), sensitive to the sex-specific secretion profile of GH and possibly involved in transcription suppression, should be investigated in the future.
Acknowledgments
We thank Tatsuya Fujisawa and Yukihiro Furusawa for their assistance with the luciferase assay. We also thank Wattanaporn Bhadhprasit for assistance with statistical analysis.
Footnotes
-
This work was partly supported by Grants-in-Aid from the Japanese Ministry of Education, Culture, Sports, Science, and Technology and the Smoking Research Foundation.
-
Accession numbers of sequence data: Cyp2b9, AB365185; FoxA2, NM010446.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.107.019729.
-
ABBREVIATIONS: P450, cytochrome P450; GH, growth hormone; JAK2, Janus kinase 2; STAT, signal transducer and activator of transcription; IRS, insulin receptor substrate; HNF, hepatocyte-enriched nuclear factor; PCR, polymerase chain reaction; rhGH, recombinant human growth hormone; FoxA2, forkhead box A2; RT-PCR, reverse transcription-polymerase chain reaction; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; ChIP, chromatin immunoprecipitation; ANOVA, analysis of variance.
- Received November 14, 2007.
- Accepted March 10, 2008.
- The American Society for Pharmacology and Experimental Therapeutics