Rhythmic expression of DEC1 and DEC2 in peripheral tissues: DEC2 is a potent suppressor for hepatic cytochrome P450s opposing DBP
2004; Wiley; Volume: 9; Issue: 4 Linguagem: Inglês
10.1111/j.1356-9597.2004.00722.x
ISSN1365-2443
AutoresMitsuhide Noshiro, Takeshi Kawamoto, Masae Furukawa, Katsumi Fujimoto, Yuzo Yoshida, Eri Sasabe, S. Tsutsumi, Taizo Hamada, Sato Honma, Ken‐ichi Honma, Yukio Kato,
Tópico(s)Circadian rhythm and melatonin
ResumoThe mammalian master molecular clock consisting of several clock gene products in the suprachiasmatic nucleus (SCN) drives circadian rhythms in behaviour and physiology. Molecular clocks consisting of the same components also exist in various peripheral organs. DEC1 and DEC2, basic helix-loop-helix transcription factors, were recently reported to be involved in the central clock in the SCN. We examined the expression profile of DEC1 and DEC2 in the periphery and their roles in the regulation of oscillating target genes in the liver. Levels of DEC1 and DEC2 mRNA exhibited a day-night variation in various peripheral tissues of rats. In the liver, their expression was high during the subjective night. Transfection assays showed that DEC2, but not DEC1, suppressed the transcription of the cholesterol 7α-hydroxylase gene (CYP7A), overwhelming the potent enhancement by d-site binding protein (DBP). Electrophoretic mobility shift assays indicated that DEC2 binds to the E-box (CACATG) at the -219/-214 region of CYP7A. The transcriptional activities of the other sterol metabolizing cytochorme P450s (Cyps), CYP8B and CYP51, were also suppressed by DEC2 but not DEC1. DEC2, but not DEC1, works as a direct output mediator that transmits the circadian signals to the hepatic functions, including the CYP7A, CYP8B, and CYP51 expression. DEC1 and DEC2, basic helix-loop-helix transcription factors, were isolated in the course of exploring the novel genes involved in the proliferation and differentiation of human chondrocytes (Shen et al. 1997; Fujimoto et al. 2001). Human DEC1 seems to be orthologous to mouse Stra13 (Boudjelal et al. 1997) and rat SHARP-2 (Rossner et al. 1997), isolated from retinoic acid-treated mouse P19 cells and rat brain, respectively. On the other hand, human DEC2 may be orthologous to SHARP-1 cloned from rat brain (Rossner et al. 1997). DEC1/Stra13/SHARP-2 has been shown to be involved in chondrogenesis (Shen et al. 1997, 2002), neurogenesis (Boudjelal et al. 1997), cell growth arrest (Sun & Taneja 2000), carcinogenesis, and hypoxia (Miyazaki et al. 2002). Stra13 null mice have been shown to develop age-induced autoimmunity as a result of impaired T-lymphocyte activation (Sun et al. 2001). Unlike most other basic helix-loop-helix transcription factors, DEC2/SHARP-1 was expressed in the late stage of neurogenic differentiation (Rossner et al. 1997), and suppressed the transcription of M1 muscarinic acetylcholine receptor (Garriga-Canut et al. 2001). The molecular mechanism of gene regulation by DECs has been shown to be through E-box (CANNTG) elements in target promoters (Honma et al. 2002; Azmi et al. 2003; Li et al. 2003). Recently, we found that in the suprachiasmatic nucleus (SCN) the expression of both DEC111 Although SHARP1/2 and Stra13 were designated for the rat and mouse orthologous genes, respectively, DEC1 and DEC2 are hereafter used in this paper irrespective of the animal species, to avoid confusion. The human gene nomenclature committee (http://www.gene.ucl.ac.uk/nomenclature/) has designated BHLHB2 and BHLHB3 as the gene symbols for DEC1 and DEC2. and DEC21 mRNA exhibited a robust circadian rhythm with a peak in the subjective day, and we speculated that DECs are involved in molecular feedback loops generating the circadian rhythms in the central clock (Honma et al. 2002), where several clock gene products, such as PER1, PER2, CLOCK, BMAL1, CRY1, and CRY2 are involved (Reppert & Weaver 2001). In addition, DECs may regulate clock-controlled genes by binding to E-box enhancer (CACGTG) (Honma et al. 2002). Circadian rhythms in behaviour and physiology are driven by the master clock located in the SCN, which entrains to the environmental light cycle. Recently, the existence of peripheral clocks has been suggested in the liver, kidney, lung, and other organs, which serve as an endogenous oscillator to regulate the peripheral circadian rhythms on the one hand, and are synchronized with the master clock on the other hand (Sakamoto et al. 1998; Panda et al. 2002a; Storch et al. 2002). The liver is crucial for energy metabolism, detoxification, and nutrient absorption such as bile secretion, and more than 300 circadian rhythms have been found in hepatic transcripts (Panda et al. 2002b; Ueda et al. 2002). CYP7A is a rate-limiting enzyme for bile acid production and is expressed in a circadian-dependent fashion (Noshiro et al. 1990). DBP, a proline and acidic amino acid rich basic leucine zipper transcription factor, amplified the circadian CYP7A rhythm, and multiple DBP-responsive elements in the 5′-upstream of CYP7A have been identified (Lavery & Schibler 1993; Lee et al. 1994). Consequently, DBP is thought to be a clock output gene, which regulates oscillating hepatic genes such as CYP7A, Cyp2a4, and Cyp2a5 (Lavery & Schibler 1993; Lavery et al. 1999). CLOCK/BMAL1 heterodimer is also a positive regulator and had been proved to regulate, as the clock output genes, the circadian expression of vasopressin in the brain (Ghorbel et al. 2003). However, except DBP, the clock output transcription factor, which regulates oscillating genes in the liver, is unknown. The mRNA levels of other sterol metabolizing CYPs, CYP8B (sterol 12α-hydroxylase) and CYP51 (lanosterol 14-demethylase), showed a circadian rhythm with different patterns from that of CYP7A (Ishida et al. 2000). Furthermore, DNA sequencing of the rat CYP7A, mouse CYP8B, and rat CYP51 genes have revealed the existence of consensus E-boxes (CANNTG) in the 5′-upstream of the genes (Nishimoto et al. 1991; Noshiro et al. 1997; Gafvels et al. 1999), suggesting that DECs may be involved in the circadian regulation of these CYPs. To investigate this hypothesis, we examined the effects of DEC1 and DEC2, as well as other clock gene products, on the transcription of hepatic CYP7A, CYP8B, and CYP51 genes. Here we demonstrate that in the peripheral tissues DEC1 and DEC2 showed higher expression during the subjective night, in contrast to their expression pattern in the SCN, and that DEC2, but not DEC1, is a potent suppresser for the expression of hepatic CYP7A, CYP8B, and CYP51 genes, overwhelming the up-regulation of these genes by DBP. We hypothesize that, in the liver, DEC2 is an output mediator from a molecular feedback loop of the peripheral clock. Total RNA samples were prepared at 10:00 h and 22:00 h from rat tissues under 12 h light and 12 h dark (LD) condition to determine DEC1 and DEC2 transcript levels. DEC1 mRNA was detected in the liver, heart, kidney, and skeletal muscle with the mRNA levels higher at 22:00 h than at 10:00 h (Fig. 1A). DEC2 mRNA was more abundantly detected in the heart, kidney, and skeletal muscle (Fig. 1B), as previously reported (Rossner et al. 1997; Fujimoto et al. 2001). DEC2 mRNA levels were also higher at 22:00 h than at 10:00 h in all tissues examined. Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA showed no significant difference between 10:00 h and 22:00 h, although the mRNA levels were higher in the heart and skeletal muscle, which are active in glycolysis (Fig. 1C). Similar day-night differences of DEC1 and DEC2 mRNA expression were observed in other tissues such as the adrenal gland, duodenum, lung, testis, cerebrum, and cerebellum (data not shown). Levels of mRNA of DEC1 and DEC2 in various tissues at 10:00 h and 22:00 h. Rats were kept under LD condition and sacrificed at 10:00 h and 22:00 h. RNA preparations were obtained from the liver, heart, kidney, and skeletal muscle of the individual rats (5 rats at each time point). The levels of mRNA for DEC1 (A), DEC2 (B), and GAPDH (C) were determined by Northern blot analysis as described in 'Experimental Procedures'. Methylene blue staining of the blotted membrane confirmed the equal loading of the RNA sample on each lane (data not shown). The results are shown both in the representative film images (upper panels) and corresponding quantified values (lower panels) using a Bio-imaging Analyser System BAS2000. The sizes of the hybridized mRNA are shown on the right side of the respective panels. The values for the liver, heart, kidney, and skeletal muscle were the mean ± SD (n = 5). *P < 0.05, **P < 0.005 vs. 22:00 h. Li, liver; He, heart; Kd, kidney; Mu, skeletal muscle. Since the circadian variation of DEC1 and DEC2 mRNA levels was observed in various rat tissues, the liver samples were prepared at 4-h intervals under both LD and constant dark (DD) conditions. A marked circadian rhythm was observed on the mRNA levels of DEC1 by Northern blot analysis (Fig. 2A, upper panel). Quantitative analysis by real time RT-PCR (Fig. 2A, lower panel) also showed that the level of DEC1 mRNA reached the maximum at 20:00 h, and the high level was prolonged to 04:00 h; it decreased to the minimum at 08:00-12:00 h in LD. The ratio of the maximum to minimum was c. 3 in LD. Expression pattern of DEC1 mRNA in DD was similar to that in LD. No significant variation was observed in the levels of GAPDH mRNA by Northern blot and real time RT-PCR analyses (data not shown). Levels of mRNA of DEC1 and DEC2 in the liver at various times of day. Rats were kept under LD or DD conditions and sacrificed at the indicated clock times. RNA preparations were obtained from the livers of 3 rats for each time point and poly(A)+ RNA samples were prepared as described in 'Experimental Procedures'. The levels of mRNA for DEC1 (A), DBP (C), CYP7A (D), CYP8B (E), and CYP51 (F) were analysed by Northern blot as described in 'Experimental Procedures'. Methylene blue staining of the blotted membrane confirmed the equal loading of the RNA sample on each lane (data not shown). The sizes of the hybridized mRNA are shown on the right side of the respective upper photos. The quantified values (lower graphs) were obtained by quantitative real time RT-PCR. The marked circadian rhythms were observed for DEC1 (P < 0.0001 and P < 0.005 for LD and DD, respectively) and DEC2 (P < 0.0001 and P < 0.005 for LD and DD, respectively). All the values are the mean ± SD (n = 3) for each time point. All the experiments were repeated at least three times, and gave reproducible results. Since the hepatic DEC2 mRNA expression was hardly detectable by Northern blot analysis, it was determined by quantitative real time RT-PCR analysis. The rhythm of DEC2 mRNA showed a peak at 16:00 h (Fig. 2B) in LD, whereas two peaks were observed at 16:00 h and 24:00 h in DD. The ratio of the maximum to minimum was c. 5–7. Quantitative real time RT-PCR using respective standard cDNA samples for DEC1 and DEC2 revealed that the DEC1/DEC2 ratio varied from c. 3–6 at various times of the day; the maximum amounts of DEC1 and DEC2 mRNA in the liver in LD were calculated as 680 and 180 ng/g poly(A)+ RNA fraction, respectively. Using the same RNA preparations, the levels of other hepatic transcripts were determined for comparison. The transcript levels of DBP exhibited a robust rhythm in the liver, showing a maximum at 20:00 h and a minimum at 04:00–12:00 h, with a 50-fold difference in amplitude in LD as previously reported (Fig. 2C) (Lavery & Schibler 1993; Lee et al. 1994). A more than 20-fold increment was observed at 16:00 h in LD, which was slightly earlier than those of CYP7A and CYP51 (see below). Under DD condition, the peak of DBP mRNA shifted to 16:00 h. A marked circadian rhythm was observed on the mRNA levels of CYP7A as well as that of CYP8B (Ishida et al. 2000) in LD, although their phases differed with each other (Fig. 2D,E) (Noshiro et al. 1990; Ishida et al. 2000). Under DD condition, the peaks of CYP7A and CYP8B mRNA were dampened, but their rhythmicity remained. Three distinct bands (2.3, 2.7, and 3.1 kb) were observed for CYP51 mRNA in the liver (Noshiro et al. 1997), and they varied in parallel (Fig. 2F). The circadian rhythm of CYP51 mRNA was observed, and the high level of CYP51 mRNA was prolonged from 20:00 h to 04:00 h. Under DD condition, the CYP51 mRNA levels showed no circadian variation. Using the same RNA preparations, mRNA levels of other clock genes, CLOCK, BMAL1, PER2, and E4BP4, were determined, and their expression patterns previously reported were confirmed (Oishi et al. 1998a; Mitsui et al. 2001) (data not shown). Since DEC1 and DEC2 exhibited marked circadian rhythms in the liver, we examined whether they affect the transcription of CYP7A, CYP8B, and CYP51. Luciferase reporter constructs containing the 5′-flanking region of CYP7A (Fig. 3A) were used for luciferase reporter assay in HepG2 cells. The basal transcriptional activity of the rat CYP7A promoter (1.8 kbp) was low, and was not appreciably affected by human DEC2 (hDEC2) or mouse DEC2 (mDEC2) (Fig. 3B, left panel). In contrast, the transcriptional activity of the promoter (1.8 kbp) was 40–60 fold enhanced by addition of rat DBP (rDBP) expression plasmid (Fig. 3B, right panel), since this promoter contains multiple DBP-binding sites (Fig. 3A) (Lavery & Schibler 1993; Lee et al. 1994). DEC2 completely suppressed the promoter activity of CYP7A enhanced by rDBP (Fig. 3B, right panel). The suppression was dose-dependent, and a low amount (12.5 ng) of hDEC2 suppressed more than 90% of the rDBP-enhanced promoter activity. Similar suppression was observed with mDEC2, indicating that suppression of the rat CYP7A promoter activity by DEC2 was independent of the source of DEC2 preparations, although the size of human DEC2 protein (50.5 kD, 482 amino acids) is larger than that of murine DEC2 (44 kD, 410 amino acids), with the extra 72 amino acids inserted in C-terminal half of human DEC2 protein (Fujimoto et al. 2001). The reporter assay using the pGL3 carrying a shorter fragment (397-bp) of the CYP7A 5′-flanking region showed a similar suppressive effect by hDEC2 or mDEC2 on the enhancement by rDBP (Fig. 3C), which indicates that the response element(s) for DEC2 are located within this short region. DEC2, but not DEC1, suppresses the transcriptional activity of the rat CYP7A promoter. (A), The 5′-upstream DNA fragments (1823 and 397 bp) of the rat CYP7A were incorporated into the pGL3-Basic vector to construct the reporter plasmids. Multiple DBP binding sites (b–e) in CYP7A are shown in the figure (Lee et al. 1994). The locations of the E-box element (CANNTG) are also shown. (B), HepG2 cells were co-transfected with the reporter plasmid pCYP7A (1.8)-Luc (20 ng), internal control pRL-TK (20 ng), and the expression plasmid (200 ng) of rDBP, hDEC2, and/or mDEC2. (C), The reporter plasmid pCYP7A (0.4)-Luc (20 ng), pRL-TK (20 ng) and the expression plasmids were co-transfected. (D), pCYP7A (1.8)-Luc (20 ng), pRL-TK (20 ng), and the expression plasmid (200 ng) for rDBP, hDEC1, or mCLOCK/mBMAL1 were co-transfected (left panel). The reporter plasmid pDEC1-Luc (20 ng), pRL-TK (20 ng), and hDEC1 expression plasmid (200 ng) were co-transfected (right panel). (E), The reporter plasmid (20 ng) for pCYP7A (0.4)-Luc and increasing amounts (50–200 ng) of the expression plasmids carrying various clock genes (rDBP, mDEC2, rE4BP4, or rPER2) were co-transfected. The luciferase activities were normalized by the activities of the internal control pRL-TK, and mean values+/− SD in the duplicate assay, and are shown as the ratio to the basal luciferase activity by the reporter plasmid alone in each experiment, except that the results of pCYP7A (0.4)-Luc are expressed as the ratio to the basal activity of pCYP7A (1.8)-Luc. All the experiments were repeated at least three times and gave reproducible results. In contrast to DEC2, human DEC1 (hDEC1) had little effect on the enhanced promoter activity of CYP7A by rDBP or its basal promoter activity (Fig. 3D, left panel). It has been reported that DEC1 negatively auto-regulates its own promoter activity (Sun & Taneja 2000), and the negative auto-regulation of the human DEC1 promoter by hDEC1 was confirmed in this study (Fig. 3D, right panel), indicating that the transfected hDEC1 expression plasmid was functional. The addition of the expression plasmids for mouse CLOCK (mCLOCK) and mouse BMAL1 (mBMAL1) did not affect the transcriptional activity of the pCYP7A (1.8)-Luc reporter construct (Fig. 3D, left panel), although they markedly enhanced the promoter of mouse Per1, as reported elsewhere (Gekakis et al. 1998; Honma et al. 2002) (data not shown). The responsive element (CACGTG) required for the binding of CLOCK and BMAL1 was absent in the 5′-flanking region in the CYP7A gene (1.8 kb). The effect of the expression of other clock genes, E4BP4 and PER2, on the transcriptional activity of CYP7A was compared with that of DEC2: The addition of rat E4BP4 (rE4BP4) or rat PER2 (rPER2), as well as mDEC2, had little effect on the basal transcriptional activity of pCYP7A(0.4)-Luc reporter construct (data not shown). E4BP4 dose-dependently suppressed the enhanced transcriptional activity by rDBP, but to a lesser extent than did mDEC2 (Fig. 3E). Based on the common response elements to DBP, E4BP4 had been predicted to antagonize DBP in the transcriptional regulation of CYP7A (Reppert & Weaver 2002): The present findings are empirical evidence for the antagonizing function of E4BP4 and DBP on the regulation of CYP7A. PER2 had much less effect on the enhanced transcriptional activity than did mDEC2. The 5′-flanking region (1.8 kbp) of the CYP7A gene examined in this study contains seven E-box consensus elements (CANNTG), as shown in Fig. 3A. The two E-boxes located at -214 and -300 of CYP7A, designated as E-box1 and E-box2 (Fig. 4A), respectively, may be responsive for DEC2, since the pCYP7A(0.4)-Luc containing them responded to DEC2 (Fig. 3C). Consequently, the oligonucleotides (probes E1 and E2) containing Ebox-1 and Ebox-2 were synthesized (Fig. 4A) and subjected to electrophoretic mobility shift assay (EMSA) with GST-mDEC2 fusion protein: GST-mDEC2 fusion protein produced two DNA-protein complex bands (Fig. 4B, lane 1). The formation of multiple bands is probably due to partial proteolytic digestion of the fusion protein during the preparation of the protein, which was confirmed by SDS-polyacrylamide gel electrophoresis of the final preparation of GST-mDEC2 (data not shown). The DNA-fusion protein complex was further shifted by the addition of the anti-DEC2 antibodies (lane 2), whereas GST itself did not show any protein-DNA complex (lane 3). The GST-mDEC2 fusion protein did not form any complex with the Probe E2 at all (lanes 4 and 5). The complex of GST-mDEC2 fusion protein and Probe E1 decreased with the addition of an excess amount of cold Probe E1 (Fig. 4C, lane 2), whereas a mutated E1m did not affect the DNA-protein complex (lane 3). DEC2 binds to the proximal E-box1 of the rat CYP7A gene. (A) Double stranded oligonucleotide probes (E1 and E2) containing E-boxes (CANNTG) of rat CYP7A gene were prepared. Putative E-box elements are shown by dashes. Overhang TCGA and GATC were added to the 5′ end of upper strand and lower strand of the oligonucleotides, respectively, for labelling with 32P-dCTP and Klenow fragment. The mutation was introduced into E-box1 of probe E1, and the mutated oligonucleotide was named as E1m. (B) EMSA was performed with 32P-labelled double-stranded probes for E1 and E2 and GST-fused mDEC2 protein (100 ng/reaction mixture). Specific antibodies (5 µg anti-DEC2 antibodies) were added to the reaction mixtures. (C) EMSA was performed with a 32P-labelled double-stranded probe E1 and GST-fused mDEC2 protein (100 ng/reaction mixture). The excess un-labelled oligonucleotides (E1 or E1m, 100 ×) as competitors were added to the reaction mixture. Luciferase reporters carrying 1.8 kb of 5′-flanking region of the mouse CYP8B gene or 0.9 kb of 5′-flanking region of the rat CYP51 gene were constructed (Fig. 5A). The basal transcriptional activity of the CYP8B promoter (Fig. 5B, left panel) was high as compared with that of CYP7A. The addition of hDEC1 had little effect on the basal activity of the CYP8B promoter (Fig. 5B, left panel), whereas hDEC2 and mDEC2 suppressed the basal transcriptional activity of the CYP8B promoter. DBP expression plasmid enhanced the promoter activity by 4-fold (Fig. 5B, right panel). The enhanced transcriptional activity by rDBP was markedly suppressed by hDEC2 or mDEC2, but not by hDEC1. Like CYP8B, the basal transcriptional activity of the CYP51 promoter was high as compared with CYP7A, and was suppressed by hDEC2 or mDEC2 (Fig. 5C). Enhancement of the CYP51 promoter by rDBP (3-fold) and its suppression by DEC2 were also observed. The basal activity and the enhanced transcriptional activity of CYP51 by rDBP were not affected by hDEC1. In accordance with the absence of CACGTG E-box in the regions examined, addition of mCLOCK and mBMALl1 had little effect on the transcriptional activity of CYP8B or CYP51 (Fig. 5B,C, left panels). DEC2, but not DEC1, suppresses the transcriptional activity of the CYP8B and CYP51 promoters. (A), DNA fragments of mouse CYP8B and rat CYP51 genes were incorporated into pGL3 basic vector to construct the reporter plasmids. Putative DBP sites (consensus: RTTAYGTAAY) in CYP8B and CYP51 are shown in the figure. The locations of E-box element (CANNTG), putative cAMP response element (CRE, TGAGGTCA), and insulin response sequence-like elements (IRS, TGTTTTG) are also shown. (B), HepG2 cells were co-transfected with pCYP8B-Luc (20 ng), internal control pRL-TK (20 ng), and the expression plasmid (200 ng) of rDBP, hDEC1, hDEC2, mDEC2, or mCLOCK/mBMAL1. In the presence of rDBP plasmid (200 ng), hDEC2 expression plasmid (12.5 or 200 ng), mDEC2 expression plasmid (200 ng), or hDEC1 expression plasmid (200 ng) was added. (C), HepG2 cells were co-transfected with pCYP51-Luc (20 ng), internal control pRL-TK (20 ng) and other expression plasmids as indicated. Other conditions were the same as described in the legend to Figure 3. In the present study, DEC2 exhibited potent suppressive effects on all three hepatic CYPs examined. However, neither DEC2 nor DEC1 is a general inhibitor of transcription, since DEC2 co-transfection had little effect on the reporter activity of hypoxia-response elements enhanced by hypoxia inducible factor-1α, as previously described (Honma et al. 2002). Since the findings in the luciferase reporter assay using limited lengths of the promoter regions may not reflect the in vivo regulation, we examined the effects of transfection of mouse DBP (mDBP) and mDEC2 expression plasmids on the expression of endogenous CYP7A and CYP8B in hepatoma cells. Mouse Hepa 1c1c7 hepatoma cells were transfected with mDBP and mDEC2 expression plasmids, and the expression of endogenous CYP7A and CYP8B mRNA levels were determined by quantitative real time RT-PCR. The endogenous CYP7A mRNA level was elevated 5-fold (P < 0.05) when DBP was fortified, and the enhanced CYP7A mRNA level was suppressed by the addition of DEC2 (Fig. 6A). DEC2 alone did not affect significantly the CYP7A expression in the absence of mDBP expression plasmid. Similar results were obtained for the endogenous expression of CYP8B mRNA (Fig. 6B). The mRNA level of DBP was increased 2-fold (P < 0.05) by the transfection of mDBP expression plasmid (Fig. 6C), whereas the DEC2 mRNA level increased more prominently (> 300-fold) with the transfection of mDEC2 expression plasmid because of a low endogenous DEC2 mRNA level (Fig. 6D). Effects of transient transfection of DBP and DEC2 on endogenous expression of CYP7A and CYP8B in mouse hepatoma cells. Confluent cultures of mouse hepatoma Hepa-1c1c7 in 6-multiwell plastic plates were transfected with the expression plasmids for mDBP (200 ng) and mDEC2 (200 ng). Three wells were used for each point. Cells were harvested 9 h after the transfection and the total RNA preparations were examined for the expression levels of endogenous mRNA for CYP7A (A) and CYP8B (B) and the fortified mRNA levels of DBP (C) and DEC2 (D), using the quantitative real time RT-PCR method. The relative levels of mRNA are expressed as the ratio to the respective control level at time 0. *P < 0.05. The molecular circadian clock consists of an auto-regulatory feedback loop associated with Per transactivation. The clock gene products, CLOCK/BMAL1 heterodimer, serve as a positive limb, and PER products, a negative limb (Gekakis et al. 1998; Jin et al. 1999). Moreover in the SCN, mRNA levels of DEC1 and DEC2 showed circadian rhythms similar to that of PER, and both DEC1 and DEC2 suppressed the transcriptional activity of the PER1 gene activated by a heterodimer CLOCK/BMAL1 (Honma et al. 2002). In the present study, the hepatic expression of DEC1 and DEC2 mRNA exhibited robust circadian rhythms, although their phases were sifted from those in the SCN. The peripheral circadian variation of DEC expression was observed in various tissues such as the liver, heart, kidney, and skeletal muscle (Fig. 1). The expression patterns of the DEC2, DBP, E4BP4 and CYP7A genes and the reporter assays suggest cooperation of these transcription factors in the regulation for circadian rhythm of CYP7A (Fig. 7). The level of hepatic CYP7A mRNA showed a circadian peak at 24:00 and decreased subsequently to the basal level during daytime. The DBP gene expression increased slightly ahead of the circadian rise of CYP7A transcripts. According to a recent work (Yamaguchi et al. 2000), the peak of DBP protein level is two hours later than that of DBP mRNA level. Therefore, the circadian rise in DBP may trigger the nocturnal increase in CYP7A gene transcription (Lavery & Schibler 1993; Lee et al. 1994). In contrast, DEC2 suppressed the transcription of CYP7A in the presence of DBP (Fig. 3B), and the expression pattern of DEC2 similar to that of DBP suggests that DEC2 suppresses excess enhancement of the transcription of CYP7A by DBP, sharply diminishing the transcription of CYP7A. On the other hand, the circadian rhythm of E4BP4 is 180° out of phase with that of DBP (Mitsui et al. 2001) suggesting that the role of E4BP4 in the suppression of CYP7A expression differs from that of DEC2. CLOCK and BMAL1 may not be involved directly in the regulation of CYP7A, although they do modulate DEC2, DBP, and E4BP4 expression (Fig. 7). Peripheral clock and regulation of clock-controlled CYP7A gene expression in the liver. DEC1 and DEC2 are involved in the peripheral clock. DBP enhances the expression of the CYP7A gene, whereas DEC2 suppresses the action of DBP through some E-box or other mechanism involving protein-protein interaction. E4BP4 antagonizes DBP action by binding to the DBP responsive site. These factors coordinate the rhythmic transcription of CYP7A. Basic helix-loop-helix transcription factors regulate the expression of the target genes by binding to DNA consensus sequences, such as the E-box and N-box. DEC1 and DEC2 bind to a specific E-box (CACGTG) (Honma et al. 2002; St-Pierre et al. 2002; Azmi et al. 2003; Li et al. 2003), which is a response element for the CLOCK/BMAL1 complex. The CYP7A gene contains seven E-boxes (CANNTG), but no CACGTG E-box for CLOCK/BMAL1 complex binding in 1.8 kbp of its 5′-flanking region (Fig. 3A). Based on the results of luciferase reporter assay of the CYP7A promoter, we focused on E-box1 and E-box2, and found that DEC2 protein binds to the proximal E-box1 (CACATG at -219). DEC2 protein probably formed a homodimer (Garriga-Canut et al. 2001) for the binding to the E-box1 element, and protein-protein interactions with other transcription factors or cofactors, such as histone deacetylase corepressor complex or the factors of the basal transcriptional machinery (Boudjelal et al. 1997; Sun & Taneja 2000; Garriga-Canut et al. 2001), may suppress the transcription of CYP7A gene. Interference of DBP binding to the DBP sites by DEC2 may be less possible, since putative E-boxes and DBP-binding sites are separated in the CYP8B and CYP51 genes (Fig. 5). CYP8B expression was rhythmic both under LD and DD conditions with a major peak at 13:00 h–16:00 h and a minor one at 24:00 h–04:00 h under LD condition, with both peaks a phase-ahead of that of CYP7A (Ishida et al. 2000). Nonetheless, CYP8B transcription was enhanced by DBP and suppressed by DEC2. Because hepatic cAMP and serum insulin, which exhibit circadian rhythms, have profound effects on CYP8B expression (Ishida et al. 2000), the contribution of DBP and DEC2 could be less than that of the hormonal signals. CYP51 mRNA levels exhibited two peaks, at the early dark phase (20:00 h) and the late dark phase (04:00 h), under LD condition. The expression of CYP51 may also be regulated via two mechanisms: a combination of clock gene products such as DEC2 and DBP, and hormonal signals producing cAMP (Yoshida et al. 1996). Considering all the above, DEC2 and DBP are involved in the circadian control of hepatic sterol metabolizing CYPs to various extents. The physiological significance of the circadian rhythm of CYP7A with high enzyme levels at night may be to cope with the nocturnal feeding pattern of rats, since CYP7A is a rate-limiting enzyme for the biosynthesis of the bile acids that are indispensable for the digestion of lipid (Noshiro et al. 1990). The master circadian clock in the SCN regulates the nocturnal behaviours of rats (Stephan & Zucker 1972). However, the peripheral clock must directly modulate the circadian rhythms of the hepatic enzymes, and thus the peripheral clock and the central clock can be dissociated by limiting food access to a particular time of day (Damiola et al. 2000). Many clock genes are involved in the peripheral clock in the liver (Fig. 7). Among these, DEC2, but not DEC1, is a direct mediator that transmits the circadian signals from the core molecular loops of the peripheral clock(s) to hepatic functions such as CYP7A. In addition to DBP and E4BP4, DEC2 may play a key role in the output pathway from the peripheral clock. Two-month-old male Wistar rats were kept under 12 h light: 12 h dark (LD, light on 06:00 h–18:00 h) or under constant dark (DD) for three days before use and fed ad libitum normal laboratory rat chow and tap water. The experimental procedures on animal care and treatment were performed with permission, and following the rules and guidelines of Hiroshima University and Hokkaido University. The rats were killed by decapitation at different times of day (10:00 h and 22:00 h for various tissues and 04:00 h, 08:00 h, 12:00 h, 16:00 h, 20:00 h and 24:00 h for the liver). In the dark phase, decapitation was conducted with the aid of a red-dim light (less than 0.1 lx). Tissues (liver, heart, kidney, and skeletal muscle) were removed quickly and total RNA was extracted by Trizol reagent (Invitrogen). The liver preparations were subjected to oligo (dT)-cellulose column chromatography to enrich the poly(A)+ RNA (Aviv & Leder 1972). Rat cDNA clones used as hybridization probes for CYP7A, CYP8B, CYP51, and GAPDH were prepared as previously described (Noshiro et al. 1990; Aoyama et al. 1996; Ishida et al. 1999). The rat cDNAs for DEC1 and DEC2 were generated by reverse transcription-polymerase chain reaction (RT-PCR) from the mRNA of rat liver and brain, respectively, by two pairs of primers, 5′-TCCTACCCGAACATCTC-3′ and 5′-GACTCACAGACCAAGACTTTGAAG-3′, and 5′-AAAATCTCTCCAGGCGACCGT-3′ and 5′-TCTTGTCTAGCCAGGGCTGTA-3′, for DEC1 and DEC2, respectively. The primers were synthesized based on the sequences of human DEC1 (EMBL/GenBank/DDBJ No. AB004066) (Shen et al. 1997) and mouse DEC2 (EMBL/GenBank/DDBJ No. AB044090) (Fujimoto et al. 2001). The PCR products were subcloned into the pGEM T-Easy vector (Promega) and subjected to sequencing for identification. Poly(A)+ RNA (1-2 µg/lane) or total RNA (10 µg/lane) preparations were electrophoresed on 1% agarose gels containing 2.2 m formaldehyde as described by Thomas (Thomas 1980) and transferred to Nytran membranes (Schleicher & Schuell GmbH). The membranes were hybridized with the 32P-labelled cDNA probes in hybridization solution containing 6 × SSC, 5 × Denhardt's solution, 10 mm EDTA, 1% SDS, and 0.5 mg/ml sonicated salmon sperm DNA at 68 °C overnight. DNA probes were labelled using α-[32P]dCTP (111 TBq/mmol, Du Pont-New England Nuclear) and random primer DNA labelling kit (Takara Co. Tokyo, Japan). The membranes were washed with 0.1 × SSC containing 0.5% SDS at 50 °C and exposed to Kodak BMX films at −80 °C. The radioactivities of the hybridized areas were quantified using a Bio-imaging Analyser System BAS2000 (Fuji Photo Film Co. Ltd, Tokyo). Real time quantitative RT-PCR analysis was performed using an ABI PRISM 7900 Sequence Detection System instrument and software (Applied Biosystems). The principles and protocols for RT-PCR analyses using this system have been previously described (Gibson et al. 1996). The first strand cDNA was synthesized using a SuperScript II reverse transcriptase (Invitrogen) with rat poly(A)+ RNA (0.1 µg) preparations. GAPDH was chosen as an internal standard to control for variability in amplification. The sequences for all the primers and probes used in these analyses were as follows; 5′-GCAAGGAAACTTACAAACTGCC-3′ and 5′-CAATGCACTCGTTAATCCGGT-3′ for DEC1 cDNA amplification; 5′-ATTGCTTTACAGAATGGGGAGCG-3′ and 5′-AAAGCGCGCGAGGTATTGCAAGAC-3′ for DEC2 cDNA amplification; 5′-ACTCTCTGAAGCCATGATGCAAA-3′ and 5′-TCCCAGACAGCGCTCTTTGAT-3′ for CYP7A cDNA amplification; 5′-GGCTGGCTTCCTGAGCTTATT-3′ and 5′-ACTTCCTGAACAGCTCATCGG-3′ for CYP8B cDNA amplification; 5′-AGGAACTGAAGCCTCAACCAATC-3′ and 5′-CTCCGGCTCCAGTACTTCTCAT-3′ for DBP cDNA amplification; and 5′-GTATGATTCTACCCACGGCAAGT-3′ and 5′-GTTTCCCATTGATGACCAGCTT-3′ for GAPDH cDNA amplification. The sequences of TaqMan™ fluorogenic probes used were 5′-FAM-CACCGGCTGATTGAGAAAAAGAGACGT-TAMRA-3′, 5′-FAM-CGACTTGGATGCGTTCCACTCGG-TAMRA-3′, 5′-FAM-TGCAAAACCTCCAATCTGTCATGAGACCTC-TAMRA-3′, 5′-FAM-CAAGGACAAGCAGCAAGACCTGGATGAG-TAMRA-3′, 5′-FAM-TGAAGAAGGCAAGGAAAGTCCAGGTGC-TAMRA-3′, and 5′-VIC-CAATGGCACAGTCAAGGCTGAGAATGG-TAMRA-3′ for DEC1, DEC2, CYP7A, CYP8B, DBP, and GAPDH, respectively. The DNA fragments of hDEC1, hDEC2 and mDEC2, containing full coding regions were inserted into pcDNA 3.1/Zeo (+) expression plasmid vector (Invitrogen) (Shen et al. 1997; Fujimoto et al. 2001). The expression plasmids of mCLOCK, mBMAL1, mDBP, and rE4BP4 were also prepared by subcloning the full-length of the coding regions of the mCLOCK (AF000998), mBMAL1 (AB012601), mouse DBP (NM_016974) and rat E4BP4 (AY004663), respectively. The expression plasmids for the rat DBP and rat PER2 were generously supplied by Dr Ueli Schibler (Mueller et al. 1990) and Dr Takahiro Nagase (Oishi et al. 1998b), respectively. The DNA fragment (1823 bp) for the reporter plasmid pCYP7A(1.8)-Luc was amplified by PCR using the primers (5′-GGACGCGTTCTATATTGTGATTTGATGC-3′) and (5′-GGAGATCTGGGGAGACTCTTTGCCTAGC-3′) and the rat genomic clone λG4 (Nishimoto et al. 1991) as the template, and then cloned into the MluI-BglII site of the pGL3-Basic vector (Promega) (see Fig. 3A). A DNA fragment containing 397 bp of the CYP7A gene was used to construct pCYP7A(0.4)-Luc as shown in Fig. 3A. To construct the reporter plasmid pCYP8B-Luc, the DNA fragment (1771 bp) was obtained from the mouse total genomic DNA as the template by PCR using the first primers (5′-GTATTGAGTAGAGGGATGTC-3′) and (5′-AGCGTCATGGCTAGGCTCCA-3′), and the nested primers (5′-ACGCGAGTCTCCCCACTGGATACT-3′) and (5′-AGGCTGAGCTCCACTTGTCAGTCGAC-3′), and cloned into the MluI-SaII site of the pGL3-Basic vector (see Fig. 5A). The PCR primers were synthesized based on the sequence of mouse CYP8B gene (Gafvels et al. 1999). The DNA fragment (925 bp) from rat genomic clone CYP51 (Noshiro et al. 1997) digested with HindlIII was cloned into the HindlIII site of the pGL3-Basic vector to construct a reporter plasmid pCYP51-Luc (see Fig. 5A). The DNA fragment (3 kbp) from human genomic clone DEC1 (Teramoto et al. 2001) was cloned into the pGL3-Basic vector to construct a reporter plasmid pDEC1-Luc. Human hepatoma HepG2 cells were inoculated at 5 × 104 cells/16 mm diameter well in 24-multiwell plastic plates, and cultured in Dulbecco's modified Eagle's medium (Sigma-Aldrich) containing 10% foetal bovine serum, 50 µg/ml ascorbic acid, 32 U/ml penicillin, and 40 µg/ml streptomycin. DNA transfection of the luciferase constructs (20 ng/assay) with internal control plasmid pRL-TK (20 ng/assay) and the expression plasmids to HepG2 cells was performed using TransIT1 (Mirus Corp. Madison, WI) 24 h after cell seeding. Next, 48 h after transfection, these cells were harvested and the cell lysate was analysed using dual luciferase assay reagents (Promega). Two wells were measured for each construct and the luciferase activity was normalized to the internal control values. At least two independent transfection experiments were performed. Mouse hepatoma Hepa 1c1c7 cells (American Type Culture Collection, Manassas, VA) were inoculated at 3 × 105 cells/35 mm diameter well in 6-multiwell plastic plates, and cultured in the same medium described above. DNA transfection of the expression plasmids for mDBP and mDEC2 to Hepa1c1c7 cells was performed using TransIT1 (Mirus Corp. Madison, WI) when cell cultures became confluent. The cells were harvested 9 h later, and total RNA was extracted by Trizol reagent (Invitrogen). Expression levels of endogenous mRNA for CYP7A and CYP8B were measured for each point (three wells) by the quantitative real time RT-PCR method. The GST-mouse DEC2 fusion protein (GST-mDEC2) was expressed in the BL21 (DE3) cells using plasmid pGST-mDEC2, which was prepared by inserting mouse DEC2 cDNA into pET-41a(+) (Novagen). The produced fusion protein was purified by chromatography using a HiTrap affinity column (Amersham Pharmacia). Rabbit antibodies to DEC2 were produced by immunizing a synthetic peptide fragment, Cys-Asn-Pro-Glu-Ser-Ser-Gln-Glu-Asp-Ala-Thr-Gln-Pro-Ala, corresponding to the amino acid positions of 394-406 of the rat/mouse DEC2. The immunoglobulin fractions were obtained by ammonium-sulphate fractionation (Noshiro & Omura 1978). Double-stranded synthetic probes (shown in Fig. 4A) for EMSA were prepared by heating equal molar amounts of complementary oligonucleotides at 85 °C in an annealing buffer (20 mm Tris-HCl pH 7.5, 10 mm MgCl2, 50 mm NaCl), and were gradually cooled to room temperature. The annealed oligonucleotides were labelled by filling in the overhang added to the synthetic oligonucleotides with α-[32P]dCTP with a Klenow fragment of DNA polymerase I (Takara Co. Tokyo, Japan). The cold competitors were prepared by filling in the annealed oligonucleotides with non-labelled dNTP and Klenow fragment. The labelled fragments were purified through a G-25 spin column (Amersham Pharmacia). The binding reaction mixture contained 20 000 cpm of labelled oligonucleotide probe and the protein factor(s) in 15 µl of 10 mm Tris-HCl (pH 8.0), 50 mm NaCl, 25 mm MgCl2, 5 mm dithiothreitol, 0.2 µg/µl poly (dI-dC), and 10% glycerol. The mixtures were incubated at room temperature for 10 min and applied to 5% acrylamide gel electrophoresis. Electrophoresis was performed at room temperature for 1 h at constant 15 W. Significance of difference between the two groups was analysed by Student's t-test. Circadian rhythms were statistically analysed by one-way analysis of variance. This work was supported by grants-in-aid for science from the Ministry of Education, Culture, Sports, Science and Technology of Japan. We thank the Research Centre for Molecular Medicine, Hiroshima University School of Medicine, for the use of their facilities.
Referência(s)