Identification and Characterization of a Novel Retinoic Acid Response Element in Zebrafish cyp26a1 Promoter.код для вставкиСкачать
THE ANATOMICAL RECORD 295:268–277 (2012) Identification and Characterization of a Novel Retinoic Acid Response Element in Zebrafish cyp26a1 Promoter JINGYUN LI,1 PING HU,1 KUI LI,1 AND QINGSHUN ZHAO1,2* Model Animal Research Center, MOE Key Laboratory of Model Animal for Disease Study, Nanjing University, Nanjing, China 2 Zhejiang Provincial Key Lab for Technology & Application of Model Organisms, School of Life Sciences, Wenzhou Medical College, University Park, Wenzhou, China 1 ABSTRACT Cyp26A1 is a major enzyme that controls retinoic acid (RA) homeostasis by metabolizing RA into bio-inactive metabolites. Previously, we demonstrated that zebrafish cyp26a1 promoter possesses two conserved RA response elements (RAREs; proximal R1 and distal R2) in response to RA. Here, we report that it contains a novel RARE (R3) lying between R1 and R2. Mutagenesis analysis reveals that R3 works together with R1 and R2 to ensure the maximum RA inducibility of cyp26a1 promoter. Performing electrophoretic mobility shift assay and chromatin immunoprecipitation assay, we show that RA receptor alpha can bind the novel RARE. Creating and analyzing transgenic zebrafish of Tg(cyp26a1-R3mut:eYFP)nju3/+ that harbor enhanced yellow fluorescent protein reporter gene (eYFP) driven by cyp26a1 promoter with mutated R3, we demonstrate that the reporter is mainly expressed in tissues of endogenous RA independent but not regions of RA dependent. Like Tg(cyp26a1:eYFP)nju1/+, which harbor eYFP driven by wild-type cyp26a1 promoter, the reporter in Tg(cyp26a1R3mut:eYFP)nju3/+ responds to excessive RA dose dependently. However, it is expressed in a significantly lower level than the reporter in Tg(cyp26a1:eYFP)nju1/+ in response to exogenous RA. Taken together, our results demonstrate that zebrafish cyp26a1 promoter contains a novel RARE that plays crucial roles in regulating cyp26a1 expression during C 2011 early development of zebrafish. Anat Rec, 295:268–277, 2012. V Wiley Periodicals, Inc. Key words: zebrafish; cyp26a1; retinoic acid response element; promoter; transgenic zebrafish; development Retinoic acid (RA) plays crucial roles in vertebrate embryogenesis. However, it could be a potent teratogen if it is available to embryos in the wrong places or at the wrong times. To prevent the specific tissues and cells from inappropriate RA exposure, animals use Cyp26s (Cytochrome P450, family 26) to convert RA into bioinactive polar forms (Niederreither and Dolle, 2008). Three members of Cyp26 family (Cyp26A1, Cyp26B1, and Cyp26C1) are characterized in vertebrates (Gu et al., 2006; Hernandez et al., 2007). Cyp26A1 (Cytochrome P450, family 26, subfamily A, polypeptide 1) is the major enzyme that involves the maintenance of RA homeostasis during vertebrate early development C 2011 WILEY PERIODICALS, INC. V Grant sponsor: Ministry of Science and Technology of China; Grant numbers: 2007CB947101 and 2011CB943804; Grant sponsor: The National Natural Science Foundation of China; Grant number: 30871439; Grant sponsor: Natural Science Foundation of Jiangsu Province; Grant number: BK2010390. *Correspondence to: Qingshun Zhao, Model Animal Research Center, Nanjing University, 12 Xuefu Road, Pukou High-tech Development Zone, Nanjing 210061, China. Fax: þ86 25 58641500. E-mail: email@example.com Received 11 June 2011; Accepted 5 September 2011. DOI 10.1002/ar.21520 Published online 20 December 2011 in Wiley Online Library (wileyonlinelibrary.com). A NOVEL RARE IN ZEBRAFISH CYP26A1 PROMOTER (Dobbs-McAuliffe et al., 2004). Cyp26A1 null mouse embryos die during mid-late gestation, exhibiting a phenotype similar to wild-type embryos treated with teratogenic doses of RA (Abu-Abed et al., 2001; Sakai et al., 2001). In zebrafish, the mutant embryo with cyp26a1 null mutation (giraffe) displays a phenotype opposite to aldh1a2 mutated embryos (neckless or no fin) that are depleted the major RA synthesis enzyme, Aldh1a2 (Begemann et al., 2001; Grandel et al., 2002; Emoto et al., 2005). When Cyp26A1/ mice are bred into an Aldh1a2þ/ background, the Cyp26A1 null phenotypes are partially rescued (Niederreither et al., 2002). These results suggest that Cyp26A1 is essential to controlling RA levels during development. Other than the specific substrate for Cyp26A1, RA is also an inducer of the gene expression though certain tissues express Cyp26A1 in a RA-independent manner during embryogenesis (Dobbs-McAuliffe et al., 2004; Hu et al., 2008). During gastrulation of zebrafish embryos, the endogenous RA controls cyp26a1 expression in cells within or adjacent to the presumptive hindbrain to establish a robust RA gradient that assigns positional identities of hindbrain rhombomeres (White et al., 2007). At 24 hpf (hours post fertilization), endogenous RA directs cyp26a1 expression in retina and anterior dorsal spinal cord (Hu et al., 2008). When exposed with exogenous RA, embryos highly express cyp26a1 (DobbsMcAuliffe et al., 2004; Hu et al., 2008), suggesting that a feedback mechanism is involved in the maintenance of RA homeostasis by cyp26a1 in vertebrate embryogenesis (Dobbs-McAuliffe et al., 2004). It is known that RA regulates transcription through binding to the heterodimers of retinoic acid receptors (RARs) and retinoid-X-receptors (RXRs) that bind to RA response elements (RAREs) in promoters of target genes (Niederreither and Dolle, 2008). Previous researchers have demonstrated that Cyp26A1 promoters in mouse, human and zebrafish all contain two conserved DR5 RAREs, one lying in the proximal region (R1) and the other lying in the distal region (R2). The two RAREs function synergistically to provide maximal induction of Cyp26A1 expression in response to RA (Loudig et al., 2000; Loudig et al., 2005; Hu et al., 2008). In addition to the two conserved RARE sites, a recent research revealed that the Cyp26A1 promoter in rat owns a new RARE site (R3) lying upstream of R2 and a half RARE site (R4) lying between R1 and R2. The R3 and R4 work together with R1 and R2 to enhance the inducibility of Cyp26A1 expression in rat liver (Zhang et al., 2010). In this study, we report that zebrafish cyp26a1 promoter possesses a novel R3 site lying between R1 and R2. Both in vitro and in vivo analyses demonstrate that the novel RARE is essential for the promoter to respond to RA signal. MATERIALS AND METHODS Plasmid Constructs and Dual Luciferase Assay The firefly luciferase reporter constructs with different lengths or mutations of cyp26a1 regulatory sequence were made from pGL-cyp26a1-2533 (Hu et al., 2008). Truncated cyp26a1 promoters (cyp26a1-1735 and cyp26a1-1490) were generated by PCR amplifications with the primers containing SacI and NcoI sites, respectively. The forward primers were atccgagctcAACGCA 269 CTGCTTACTCAGAGGCA (for cyp26a1-1735) and atccga gctcTCATGTACCGGTAACCTTGCTCT (cyp26a1-1490), and the reverse primer is ggggtaccatggTTGAAGCGCGC AACTGATCGCC. The promoter containing mutations of R3 was generated by PCR amplification using QuikChange II-E Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA). The primers used to generate cyp26a12533-R3mut fragment were AGGATGAAgttTGAAATGCg ttGCAGGCACAC (forward) and GTGTGCCTGCaacGC ATTTCAaacTTCATCCT (reverse). The R3 fragment was generated by annealing two oligonucleotides with complementary 44 bp containing R3 RARE, SacI and SpeI sites. The sequences were gagctcAGGATGAAAGGTGAAATGCAGGGCAGGCACACactagt (sense strand) and actagtGTGTGCCTGCCCTGCATTTCACCTTTCATCCTga gctc (antisense strand). The 146 bp fragment upstream of zebrafish cyp26a1 translation start codon containing its basal promoter (cyp26a1-146) was PCR amplified using forward primer actagtAATTTATCTGAGGAAGTTA and reverse primer ggggtaccatggTTGAAGCGCGCAACT GATCGCC containing SpeI and NcoI sites, respectively. Both R3 RARE and cyp26a1-146 were subcloned into pGEM-T Easy vector (Promega, Madison, WI) and then recombined to produce pGEM-T-R3-cyp26a1-146. The amplified promoters (cyp26a1-1735, cyp26a11490, cyp26a1-2533-R3mut, and R3-cyp26a1-146) were sequenced to confirm their identities and then recombined into pGL3 basic luciferase reporter vector (Promega) between SacI and NcoI sites. The plasmids were named pGL-cyp26a1-1735, pGL-cyp26a1-1490, pGL-cyp26a1-2533-R3mut, and pGL-R3-cyp26a1-146, respectively. To obtain pGL-cyp26a1-R2mut-R3mut, the fragment containing mutated R2 was generated form pGL-cyp26a1-2533-R2mut (Hu et al., 2008) by digesting with NdeI and then recombined into pGL-cyp26a1R3mut. The coding sequences of zebrafish rars and rxrs were cloned by RT-PCR. The primers for cloning the nuclear receptors were ATGTATGAGAGTGTGGATGTGAACCC and TCATGGTGACTGTGTGGGCG (Q90271: raraa); ATGTTCGACTGTATGGAGGCTC and TCACTGAGCTC TTCCTCCG (Q91392: rarca); ATGCACCCGTCACTGCTGAG and TTATGTCATTTGGTGTGGAGC (A2T929: rxr aa); GATCATGCCACTGAACAGGA and CCGTAATGAA TGGGTTGGAC (Q90415: rxrab); GAGAGTCGGTGTGAA CAGCA and CAGGGAGAATGTGGAGGAGA (Q7SYN5: rxrba); TGGCCAGTCTAGTGTGCATC and AGAGGCTG ACCAATTGCTGT (Q90417: rxrbb); CTTCTTTCATATTC GTCCAAACA and GTGCGCTGGGGTTTATTTAC (Q90 416: rxrca); CCTCCAGGACAAGACCTGAC and TCGTC TTTTCCCTGCCAATA (Q6DHP9: rxrcb). The amplified genes were sequenced to confirm their identities and then further PCR amplified by new primers that were added sequences of XhoI recognition site and Kozak sequence (ctcgagGCCACC) immediately before above forward primers, and the sequences of KpnI recognition site immediately before above reverse primers, respectively. The modified sequences were finally recombined into pcDNA3.1(Invitrogen, Carlsbad, CA) between XhoI and KpnI or NotI sites to produce expression vectors of the nuclear receptors for transfection study. Dual luciferase assays were performed on 293T cells (American Type Culture Collection, Manassas, VA) and zebrafish embryos as reported previously (Hu et al., 2008; Xu et al., 2009) using a commercial kit (Promega). 270 LI ET AL. Briefly, 400 ng of tested promoter constructs, 8 ng of renilla luciferase expression plasmid, and 100 ng of zebrafish nuclear receptor expression vectors were cotransfected into one well of the 24-well plate using Lipofectamine 2000 (Invitrogen) for 6 hr. After transfection, each well was changed with 500 lL DMEM containing 10% charcoal stripped serum (Invitrogen) and various concentrations of all-trans RA or vehicle dimethylsulfoxide (DMSO; Sigma, St. Louis, MO). As to the assay on zebrafish embryos, 125 pg tested promoter constructs and 62.5 pg zebrafish RARab plus 2.5 pg renilla luciferase expression vector were co-microinjected into each zebrafish embryo at 1–2-cell stage. DMSO or 100 nM RA were then administrated to the embryo water immediately. The embryos were grown to 8 hpf. Three pools of 30 microinjected embryos were collected to detect luciferase activities. The luciferase activities were measured using a plate-reading GloMax 96 Microplate Luminometer (Promega) 24 hr after the treatment (for the cells) or at 8 hpf (for the embryos), respectively. Firefly luciferase activity was normalized by dividing by the activity of renilla luciferase. Relative luciferase activity was described as the fold change of each treatment to control experiment. Each treatment was repeated three times with two independent transfections. Results were subjected to the Student’s t-test. Electrophoretic Mobility Shift Assays To perform EMSA, we first transfected 293T cells with 4 lg of pWI-Ef1a-RARab (previously named RARa2.B; Perz-Edwards et al., 2001) or vehicle vector pWI-Ef1a in 100-mm dish using GenEscort II (Wisegen, Nanjing, China). After transfection for 24 hr, the cells were collected and then subjected to isolate nuclear proteins using the Nuclear Extract Kit (Active Motif, Carlsbad, CA). The protein concentration of the nuclear extracts was determined with BCA Protein Assay Kit (Pierce, Rockford, IL) for EMSA. The 32 bp double-stranded DNA fragment containing putative core sequences of R3 RARE was labeled with [c-32P]dATP as a probe by T4 polynucleotide kinase using Gel Shift Assay System Kit (Promega). Binding reactions were performed in a 10 lL buffer provided in the kit by incubating the labeled probes (10,000 cpm) with 20 lg nuclear extracts at room temperature for 20 min. For competition assay, 50-fold molar excess of 22 bp non-specific SP1 DNA fragments, 300-fold molar excess of unlabeled R3, 300-fold molar excess of unlabeled mutated R3 or 2 lg of the rabbit polyclonal antibody against RARa (Cat. #: sc-551X, human origin, Santa Cruz, CA) were added to above buffer and incubated for 20 min at room temperature before adding the labeled R3. After incubation with the labeled R3 for 30 min at room temperature, the reaction mixture was subjected to electrophoresis for 90 min under 12.5 V/cm in 4% native acrylamide gel. The gel was placed on a sheet of Whatman filter paper and covered with a plastic wrap and was subjected to autoradiography for 5 days in dark at 80 C. Chromatin Immunoprecipitation (ChIP) Assay To perform ChIP assay, we first recombined zebrafish RARab coding sequence into pCMV-3Tag-7 vector (Stra- tagene). The coding sequences of Myc tagged RARab (Myc-RARab) and zebrafish RXRab were then subcloned into the pBluescript SK(þ) vector (Stratagene) under T7 promoter direction. mRNAs of Myc-RARab and RXRab were synthesized and capped in vitro using mMESSAGE mMACHINE T7 Ultra Kit (Ambion, Austin, TX). About 2 nL of 0.1 ng/nL Myc-RARab mRNA and 0.1 ng/nL RXRab mRNA were co-microinjected into zebrafish embryos at 1–2-cell stage. The embryos were grown to 8 hpf for ChIP assay. ChIP analyses were performed using EZ-ChIP Chromatin Immunoprecipitation Kit (Cat. #: 17-371, Millipore, Billerica, MA) following the manufacture’s instructions. Briefly, 300 of microinjected embryos at 8 hpf (uninjected embryos were used as control) were fixed with 1% formaldehyde to allow cross-linking of macromolecules. 1 lg of anti-Myc Tag mouse monoclonal antibody (Millipore, Cat. #: 05-724) or mouse IgG (Millipore, Cat. #: 17-731) was used for Myc IP or control IP, respectively. The IP DNA was amplified by real-time PCR with primers (forward: TGTGTCATGCGTGAGTATTCC; reverse: AGA GACAGGATGGCTCCAAA) for detecting the zebrafish cyp26a1 promoter encompassing the R3 RARE (1686 to 1575; R3N-112bp) or with primers (forward: GAGA GTTTGGAGCCGCTTCT; reverse: CGCAGGTCTCCTCC TGTTA) for detecting the one encompassing the R1 RARE (223 to 113; R1-111bp) as a positive control. The Real-time PCR was performed using an ABI 7300 detection system (Applied Biosystems, Foster City, CA) with SYBR green I reagents (Takara, Katsushika, Tokyo, Japan). The relative enrichment of RARab on cyp26a1 R1 or R3 was calculated using Fold Enrichment Method (Invitrogen) by normalizing the PCR signals obtained from ChIP with anti-Myc Tag antibody to the signals obtained from control ChIP with mouse IgG. Results were subjected to the Student’s t-test. Generation of Transgenic Zebrafish To make a construct for generating transgenic zebrafish that express eYFP driven by cyp26a1-2533-R3mut, we first digested pGL-cyp26a1-R3mut with SacI and NcoI to get a 2533bp fragment containing the mutated R3 and then ligated the fragment into the vector of p2.5kcyp26a1pr_eYFP (Hu et al., 2008) between SacI and NcoI sites. The construct was then digested from p2.5kcyp26a1R3mut_eYFP with XhoI and BglII and recombined into TSG (Zhao et al., 2008) between XhoI and BglII sites. The resulting plasmid was named TSGcyp26a1-R3mut-eYFP. To make transgenic zebrafish, one-cell zebrafish embryos were microinjected with 1 nL of a DNA/RNA solution containing 25 ng/lL of TSG-cyp26a1-R3mut-eYFP plasmid and 25 ng/lL of transposase mRNA synthesized in vitro using pCS-TP (Kawakami et al., 2004) as a template with Message mMACHINE SP6 Kit (Ambion). The microinjected embryos were incubated at 28.5 C and examined for transient eYFP expression at 24 hpf under a fluorescent dissecting microscope (MVX10; Olympus, Tokyo, Japan). The embryos expressing eYFP ubiquitously (except in yolk) were raised to adults as transgenic founders. The embryos at 24 hpf produced from a founder and its wild-type partner were subjected to screen YFP signal using the dissecting microscope. A NOVEL RARE IN ZEBRAFISH CYP26A1 PROMOTER The transgenic offspring (F1) was raised to sexual maturity. To examine responses of the reporter to altered RA signaling, transgenic embryos were treated with 1 lM RA from 24 hpf to 30 hpf. The images of embryos expressing YFP were photographed under the MVX10 dissecting microscope equipped with a RetigaExit Fast 1394 CCD camera (Olympus). Embryos younger than 18 hpf were placed on a Petri dish to observe directly. Embryos older than 18 hpf were anaesthetized by 0.1% tricain (Sigma) before observation. Color micrographs were pseudo-colored to yellow with Image-Pro Plus Version 6.2 software (Media Cybernetics, Bethesda, MD). Quantitative PCR to Detect Integrated Copies of eYFP in Transgenic Zebrafish Genomes and Expression Levels of the Reporter in Transgenic Zebrafish Embryos Genomic DNAs isolated from eYFP-positive transgenic embryos at 24 hpf produced from a female wild-type zebrafish mated by a male transgenic zebrafish were used as template to detect integrated copies of eYFP in transgenic zebrafish genomes whereas total RNAs isolated from the same batch of transgenic embryos at 24.5 hpf that had been treated with vehicle control or different amount of RA for 30 min were reverse-transcribed to cDNA as templates for analyzing expression levels of the reporter in transgenic embryos. Real-time PCR was performed using an ABI 7300 detection system (Applied Biosystems, Foster City, CA) with SYBR green I reagents (Takara). Primers for detecting 133 bp eYFP genomic DNA fragment or cDNA reverse-transcribed from eYFP mRNA were CGACCACTACCAGCAGAACA (forward) and GAACTCC AGCAGGACCATGT (reverse). Primers for detecting 116 bp zebrafish b-actin promoter fragment were AACACAA CACAGGATCATGGA (forward) and CGTATTTTTCTGC GACAACG (reverse) and for detecting zebrafish b-actin cDNA were CGAGCAGGAGATGGGAACC (forward) and CAACGGAAACGCTCATTGC (reverse). Efficiency of amplification and detection were validated by determining the slope of CT versus dilution series. The copy number of the transgene was calculated by normalizing the amount of eYFP to that of internal control b-actin promoter according to standard procedures. Transcript level for eYFP was normalized to b-actin mRNA level according to standard procedures and then further normalized by dividing the integrated copy numbers of the transgene. Results were subjected to the Student’s t-test. RESULTS The Truncated cyp26a1 Promoter Without the Distal RARE has the Capacity to Respond to RA In Vitro Previously, we found that the zebrafish cyp26a1 promoter with mutated R2 is still RA inducible that cannot be attributed to R1 (Hu et al., 2008). To confirm that the promoter without R2 is able to respond to RA, we made different truncated promoters (Fig. 1A) and then tested their activities in 293T cells. As shown in Fig. 1B, the truncated promoter without R2 (1735) showed a good activity to 100 nM RA treatment though its activity was 271 significantly lower than that of the 2533 promoter. However, the truncated promoter (1490) showed no response to 100 nM RA treatment (Fig. 1B). To examine whether the response of the truncated promoter without R2 (1735) to RA is dose dependent, we tested the promoter activities with different concentrations of RA. As shown in Fig. 1C, the promoter activities at 10 nM, 100 nM, and 1000 nM RA treatment were significantly higher than control experiment (P < 0.05, P < 0.01, and P < 0.01, respectively). Statistical analysis also revealed that the promoter activity at 100 nM treatment was significantly higher than that at 10 nM treatment (P < 0.05) and that the promoter activity at 1,000 nM treatment was significantly higher than that at 100 nM treatment (P < 0.05). Taken together, the results suggest that some key sequences between 1735 and 1490 must play crucial roles in RA inducibility of the promoter. The Novel RARE is Required for the Maximum RA Inducibility of cyp26a1 To identify the key sequences between 1735 and 1490 that are essential to the RA inducibility of the promoter, we analyzed the 2533bp promoter of cyp26a1 using MatInspector (http://www.genomatix.de). We found the promoter has a presumptive novel RARE site (R3) between R1 and R2 (Fig. 2B). Like R1 and R2, the predicted R3 is also a DR5 RARE and its core sequences are AGGTGAAATGCAGGGCA locating between 1644 and 1628 of the promoter. Different from R1 and R2 that are positioned in reverse directions (Hu et al., 2008), R3 is present in forward direction. To confirm the role of the predicted R3 in RA inducibility, we mutated the core sequences (Fig. 2A), and then examined the promoter activity in 293T cells after 100 nM RA treatment. Like the R2-mutated promoter, R3mutated promoter exhibited a significantly reduced activity (P < 0.01) compared to the wild-type (WT) promoter (Fig. 2C). The activity of the promoter with mutated R2 and mutated R3 was significantly (P < 0.01) lower than that of either R2-mutated promoter or R3mutated promoter (Fig. 2C). To test whether the R3 is required for the activity of cyp26a1 promoter in zebrafish embryos, we microinjected the WT and R3mut constructs into zebrafish embryos and then test the promoters’ activities in the embryos at 8 hpf. The results showed that the activity of R3-mutated promoter was significantly lower (P < 0.01) than that of wild-type promoter in the embryos either treated with vehicle DMSO or 100 nM RA (Fig. 2D). These data demonstrate that the R3 is required for the maximum RA inducibility of cyp26a1 in both 293T cells and zebrafish embryos. Zebrafish RARab can Bind to the R3 Element In Vitro and In Vivo To identify whether R3 is a functional RARE, we performed EMSA to check whether zebrafish RAR could bind to this predicted RARE. To perform EMSA, we radiolabeled the 32-bp oligonucleotides that contain core sequences of R3 (32P-R3; Fig. 2A) and then incubated the probe with nuclear proteins extracted from 293T cells. Unlabelled oligonucleotides of R3, mutated R3 and non-specific control were used as competitors of 32P-R3. 272 LI ET AL. Fig. 1. The activity of the truncated promoter of zebrafish cyp26a1 without R2 is induced by RA in a concentration dependent manner. A: Schematic diagram showing the truncated promoter fragments of cyp26a1 driving firefly luciferase reporter gene. The first nucleotide upstream start codon site is numbered as 1. The numbers (2533, 1735, and 1490) denote the position of first nucleotide of the zebrafish cyp26a1 promoter upstream start codon site. Three RARE sites are shown as different shaded rectangles. B: Dual luciferase activity assay showing the relative activities of the different promoters were induced by 100 nM RA. C: Dual luciferase activity assay showing that the activity of the 1735 promoter fragment was RA concentration dependent. *P < 0.05; **P < 0.01. As shown in Fig. 3, a shift band was formed when the probe of 32P-R3 was incubated with the nuclear proteins extracted from 293T cells transfected with pWI-Ef1ararab to overexpress zebrafish RARab (Fig. 3A, Lane 3) whereas the band was not seen when the probe was incubated with the nuclear proteins extracted from 293T cells transfected with vehicle vector (Fig. 3A, Lane 2). The shift band was nearly gone (Fig. 3A, Lane 4, Lane 5) when the antibody that can recognize zebrafish RARab was added to deplete the overexpressed RARab or the unlabelled oligonucleotides of R3 was added to compete with 32P-R3 though we do not know why the non-specific band disappeared in Lane 5 after adding the cold R3 oligos and why there was a new band emerged in this lane (Fig. 3A, Lane 5). However, the band was only reduced its intensity when mutated R3 was added to compete with 32P-R3 (Fig. 3A, Lane 6) and it was not affected when non-specific control oligonucleotides were added to compete with 32P-R3 (Fig. 3A, Lane 7). The EMSA results demonstrate that zebrafish RARab is able to bind R3 in vitro. To further confirm that zebrafish RARab can bind to R3 in vivo, we performed ChIP assay. Because the antibody against human RARa did not work in ChIP assay (data not shown) and there is no ChIP grade antibody against zebrafish RARab commercially available, we added a Myc tag in the N-terminal of zebrafish RARab. We then overexpressed Myc-tagged zebrafish RARab into zebrafish embryos by microinjecting Myc-tagged RARab mRNA. When the embryos reached 8 hpf, we extracted nuclear proteins from the embryos and used anti-Myc Tag antibody to do ChIP and real-time PCR assay. The results showed that the fold enrichment value of RARab on cyp26a1 R1 (a canonical RARE conserved in vertebrates) is 1.25 in the WT control embryos and 2.55 in the Myc-RARab mRNA microinjected embryos (Fig. 3B). Similarly, the fold enrichment value of RARab on cyp26a1 R3 is 1.39 in the WT control embryos and 2.36 in the Myc-RARab mRNA microinjected embryos (Fig. 3B). Statistical analyses on differences between the fold enrichment values revealed that like binding to R1, zebrafish RARab could significantly bind to R3 (P < 0.01). RA Inducibility of the R3 Activity is Mediated by Heterodimers of Zebrafish RAR and RXR Zebrafish own three retinoic acid receptors (raraa, rarab, and rarca) and six retinoid X receptors (rxraa, A NOVEL RARE IN ZEBRAFISH CYP26A1 PROMOTER 273 Fig. 2. R3 in the 2533 bp promoter is required for RA inducibility of zebrafish cyp26a1 gene. A: Sequences of wild-type and mutated RAREs (R2 and R3) in zebrafish cyp26a1 promoter. The core sequences of wild-type RARE and mutated RAREs are underlined. B: Schematic diagram showing the different promoters of cyp26a1 driving firefly luciferase reporter gene. WT: the 2533 bp promoter with wildtype R1, R2, and R3; R2mut: the 2533 bp promoter with mutated R2, wild-type R1 and R3; R3mut: the 2533 bp promoter with mutated R3, wild-type R1 and R2; R2mut/R3mut: the 2533 bp promoter with mutated R2, mutated R3, and wild-type R1. C: Dual luciferase activity assay showing that the relative of the different promoters were induced by100 nM RA in 293T cells. **P < 0.01. D: Dual luciferase activity assay showing that the activity of cyp26a1 promoter with mutated R3 is significantly lower than the wild-type promoter in zebrafish embryos at 8 hpf treated with either DMSO or 100 nM RA. **P < 0.01. rxrab, rxrba, rxrbb, rxrca, and rxrcb; Waxman and Yelon, 2007). To test whether all the nuclear receptors involve RA induction of R3 activity, we made an expression vector by recombining the 32bp R3 with 146bp zebrafish cyp26a1 basal promoter that drives firefly luciferase and then examined the reporter activities in 293T cells that were overexpressed with different nuclear receptors and treated with 100 nM RA. As shown in Fig. 4, the R3 activity in the cells without receptors transfected (control experiment) was similar (P > 0.05) to the ones transfected with RARaa/RXRcb or RARca/ RXRba but significantly lower than (P < 0.01) the ones transfected with the other 16 combinations of receptors including RARaa/RXRaa, RARaa/RXRab, RARaa/RXRba, RARaa/RXRbb, RARaa/RXRca, RARab/RXRaa, RARab/ RXRab, RARab/RXRba, RARab/RXRbb, RARab/RXRca, RARab/RXRcb, RARca/RXRaa, RARca/RXRab, RARca/ RXRbb, RARca/RXRca, and RARca/ RXRcb. When induced by 100 nM RA, the R3 activity in the cells transfected with RARaa/RXRcb or RARca/RXRba was similar (P > 0.05) to the control experiment without receptors transfected, but significantly lower (P < 0.01) than in the cells transfected with the other 16 heterodimer receptors. Furthermore, the R3 activity in the cells transfected with all 18 heterodimer receptors or vehicle control was induced to increase significantly (P < 0.05 or P < 0.01) by 100 nM treatment when compared with DMSO treatment (Fig. 4). Taken together, the results demonstrate that RARaa/RXRcb and RARca/RXRba do not but the other 16 heterodimers of zebrafish RAR and RXR do mediate R3’s response to RA signal. R3 is Essential to RA Dependent Expression of cyp26a1 Gene in Zebrafish Early Development Previously, we reported that the 2533 bp promoter of zebrafish cyp26a1 almost fully recapitulates endogenous cyp26a1 expression (Hu et al., 2008). To identify the role of R3 in the expression of zebrafish cyp26a1 at early development, we mutated R3 in the 2533bp promoter and then fused the mutated promoter with the enhanced yellow fluorescent protein (eYFP) reporter gene to make transgenic zebrafish lines. In total, we obtained three transgenic zebrafish lines from three different founders. Generally, the three lines of zebrafish [Tg(cyp26a1R3mut:eYFP)nju3/+; Tg(cyp26a1-R3mut:eYFP)nju4/+; Tg(cyp26a1-R3mut:eYFP)nju5/+] display similar expression patterns. Figure 5 shows the early expression pattern of the reporter in the embryos [Tg(cyp26a1R3mut:eYFP)nju3/+] generated from a male transgenic zebrafish founder mated with a female wild-type zebrafish. Initially, very weak reporter signals were detected 274 LI ET AL. Fig. 3. The zebrafish RARab is able to bind to R3 element in vitro and in vivo. A: EMSA was performed to identify that R3 can be bound by zebrafish RARab. Nuclear proteins extracted from 293T cells that were overexpressed with vehicle expression vector (pWI-Ef1a; Lane 2) or zebrafish RARab expression vector (pWI-Ef1a-rarab; Lanes 3–7) were incubated with 32P-labeled R3 RARE probe (Lanes 2–7). Antibody against human RARa (Lane 4) was added to depleted zebrafish RARab, and 300-fold excess of unlabelled R3 (300 cold R3; Lane 5), 300-fold excess of unlabelled mutated R3 (300 cold mutant R3; Lane 6), or 50-fold excess of non-specific DNA sequences (50 cold non specific; Lane 7) were used to compete the binding to zebrafish RARab with 32P-labeled R3. The free probe of 32P-labeled R3 was run in Lane 1 as a control. In addition to non-specific binding bands (Lanes 2–7), a shift band was formed when the nuclear protein con- taining zebrafish RARab was incubated with 32P-labeled R3 (Lane 3). The binding that forms the shift band was gone when adding antibody against RARa (Lane 4) or 300-fold excess of unlabelled R3 (Lane 5), but only reduced when adding 300-fold excess of unlabelled mutated R3 (Lane 6), or even not affected when adding 50-fold excess of nonspecific DNA (Lane 7). B: ChIP – Real-time PCR assay showing the fold enrichment of zebrafish RARab binding to cyp26a1 R3 and R1. Myc tagged RARab (Myc-RARab) mRNAs plus with RXRab mRNAs were co-microinjected into zebrafish embryos at 1–2-cell stage. The ChIP analysis was performed on the embryos at 8 hpf. The relative enrichment of RARab on cyp26a1 R1 or R3 was calculated by normalizing the PCR signals obtained from ChIP with anti-Myc Tag antibody (Myc-RARab) to the signals obtained from control ChIP with mouse IgG (WT). **P < 0.01. Fig. 4. Effects of zebrafish RAR/RXR heterodimers on R3 activity in response to RA treatment in 293T cells. Different heterodimers of zebrafish RAR and RXR or vehicle control were overexpressed or transfected in 293T cells, respectively. The relative luciferase activity of R3 was induced by 100 nM RA or DMSO control. Units represent ratios of firefly luciferase activity to control renilla luciferase activity for each sample. The relative activity of R3 in the cells transfected with vehicle vectors (no receptors) was normalized to 1.00. The statistical significance of the difference between RA treatment and DMSO control for each tested RAR/RXR heterodimer was examined by Student’s t-test.*P < 0.05; **P < 0.01. Fig. 5. Expression pattern and RA inducibility of the reporter in early development of transgenic zebrafish Tg(cyp26a1-R3mut:eYFP)nju3/þ. Transgenic embryos were produced by a female wild-type zebrafish mated with a male transgenic zebrafish of Tg(cyp26a1R3mut:eYFP)nju3/þ (A–G) or Tg(cyp26a1:eYFP)nju1/þ (H–K). Fluorescent images were photographed by a CCD camera with 2 sec (A–G) or 1 sec (H–K) exposure time, respectively. Embryos are laterally viewed and positioned animal pole top (A), anterior top (B), or anterior left (C– K). YFP reporter in Tg(cyp26a1-R3mut:eYFP)nju3/þ embryo was initially detected at presumptive neural plate at 75% epiboly (A), the presumptive brain regions, somites and the tail bud at 11 hpf (B) and caudal notochord, pharyngeal arches, presumptive jaw and some regions of diencephalons at 24 hpf (C) or 30 hpf (D). The transgenic embryos produced from Tg(cyp26a1-R3mut:eYFP)nju3/þ (D–G) or Tg(cyp26a1:eYFP)nju1/þ (H–K) were treated with vehicle DMSO (D and H), 10 nM RA (E and I), 100 nM (F and J), or 1000 nM RA (G and K) for 6 hr from 24 hpf to 30 hpf, respectively. The reporter in embryos from both Tg(cyp26a1-R3mut:eYFP)nju3/þ (E–G) and Tg(cyp26a1:eYFP)nju1/þ (I– K) were induced or increased its expression in forebrain, retina, presumptive jaw, anterior dorsal spinal cord, proctodeum, and mesoderm of tail in a RA-dose dependent way. cn, caudal notochord; di, diencephalon; fr, forebrain; ja, presumptive jaw; mt, mesoderm of tail; pa, pharyngeal arches; re, retina; sc, anterior dorsal spinal cord. (L) Real-time PCR results showing the relative copy number of integrated eYFP gene in the genome of Tg(cyp26a1:eYFP)nju1/þ (Control) and Tg(cyp26a1R3mut:eYFP)nju3/þ (R3mut). The copy number of integrated eYFP gene in the genome of Tg(cyp26a1-R3mut:eYFP)nju3/þ was normalized to 1. (M) Real-time PCR results showing the relative level of eYFP mRNA in embryos of Tg(cyp26a1:eYFP)nju1/þ (Control) and Tg(cyp26a1R3mut:eYFP)nju3/þ (R3mut) treated with vehicle DMSO, 10 nM RA, 100 nM or 1000 nM RA for 30 min from 24 hpf to 24.5 hpf, respectively. The relative level of eYFP mRNA in embryos of Tg(cyp26a1:eYFP)nju1/þ was divided by the fold (7) of its integrated eYFP copy number to that of Tg(cyp26a1-R3mut:eYFP)nju3/þ. Results show that the cyp26a1 promoter with mutated R3 has a reduced response to RA signal compared with wild-type cyp26a1 promoter in vivo.*P < 0.05; **P < 0.01. 276 LI ET AL. from 75% epiboly (8 hpf) embryos at the presumptive neural plate (Fig. 5A). At 11 hpf, the transgene signals were mainly expressed at the presumptive brain regions, somites and the tail bud (Fig. 5B). At 24 hpf, the reporter expression was mainly detected in caudal notochord, pharyngeal arches, some regions of diencephalons and a presumptive jaw region (Fig. 5C) but not in retina, anterior dorsal spinal cord and proctodeum in which the reporter in Tg(cyp26a1:eYFP)nju1/þ is normally expressed in endogenous RA dependent way (Hu et al., 2008). At 30 hpf, the expression of cyp26a1 was maintained in a similar pattern to that in embryos at 24 hpf (Fig. 5D). When the Tg(cyp26a1-R3mut:eYFP)nju3/þ embryos were treated with different concentrations of RA for 6 hr from 24 hpf to 30 hpf, the reporter was induced or increased to express in forebrain, retina, presumptive jaw, anterior dorsal spinal cord, proctodeum, and mesoderm of tail in a RA-dose dependent way (Fig. 5D–G). The induced expression pattern was similar to Tg(cyp26a1:eYFP)nju1/þ embryos treated with different concentrations of RA (Fig. 5H–K) though the reporter expression level in the Tg(cyp26a1-R3mut:eYFP)nju3/þ embryos was much weaker than that in the Tg(cyp26a1:eYFP)nju1/þ embryos. However, the phenotype that the transgenic reporter of Tg(cyp26a1-R3mut:eYFP)nju3/þ embryos was not expressed in endogenous RA dependent regions (Fig. 5C,D) and lower response to excessive RA (Fig. 5E–G) might be due to the fewer integrated copies of transgenic reporter in Tg(cyp26a1-R3mut:eYFP)nju3/þ than in Tg(cyp26a1:eYFP)nju1/þ. To exclude the possibility, we divided the mRNA levels of reporters in the two transgenic zebrafish embryos by the number of integrated copies of reporters in genomes of the transgenic zebrafish and then compared their relative expression levels. Performing real-time PCR, we found the copy numbers of the reporter in Tg(cyp26a1:eYFP)nju1/þ are about 7 times as many as that in Tg(cyp26a1R3mut:eYFPnju)3/þ (Fig. 5L). Divided by the copy times of the integrated transgene, the expression level per reporter copy in Tg(cyp26a1:eYFP)nju1/þ embryos was significantly (P < 0.01) higher than that in Tg(cyp26a1R3mut:eYFP)nju3/þ embryos at 24.5 hpf when the embryos had been treated with vehicle control or different concentrations of RA for 30 min, respectively (Fig. 5M). Consistent with the results from the promoter activity assay in 293T cells (Fig. 1), the activities of the transgenic promoter with mutated R3 at 10 nM, 100 nM, and 1,000 nM RA treatment were significantly higher than control experiment (P < 0.05, P < 0.05, and P < 0.01, respectively; Fig. 5M). Statistical analysis also revealed that its activity at 100 nM treatment was significantly higher than that at 10 nM treatment (P < 0.05) and that the promoter activity at 1,000 nM treatment was significantly higher than that at 100 nM treatment (P < 0.01; Fig. 5M). Taken together, the results demonstrate that the cyp26a1 promoter with mutated R3 has a reduced response to RA signal when compared with wild type of cyp26a1 promoter in vivo. DISCUSSION Cyp26A1 is crucial to vertebrate early embryogenesis by playing essential roles in maintaining RA homeosta- sis in vivo (Abu-Abed et al., 2001; Sakai et al., 2001; Dobbs-McAuliffe et al., 2004; Emoto et al., 2005). Previously, researches demonstrated that vertebrate Cyp26A1 contains two canonical RAREs (proximal R1 and distal R2) that work synergistically to respond to RA signal directly (Loudig et al., 2000; Loudig et al., 2005; Hu et al., 2008). In addition to the two conserved RARE sites, a recent research revealed that the rat Cyp26A1 promoter contains a new RARE site (R3) lying upstream of R2 and a half RARE site (R4) lying between R1 and R2 (Zhang et al., 2010). Here, we report that the zebrafish contains a third RARE (R3) in its cyp26a1 promoter that is essential for the promoter to respond to RA signal both in vitro and in vivo. A RARE is typically composed of two direct repeats of the motif, PuG(G/T)TCA, or closely degenerate, separated by a 1-base pair (bp), 2-bp, or 5-bp spacer (DR1, DR2, or DR5), respectively. Like R1 and R2, R3 in both zebrafish and mammalian Cyp26A1 is also a DR5 element. However, these RAREs are positioned in different direction in chromosomes. In mammals, R1 in Cyp26A1 is positioned in reverse direction whereas both R2 and R3 are present in forward direction (Zhang et al., 2010). In contrast, zebrafish R3 in cyp26a1 is positioned in forward direction whereas R1 and R2 are positioned in reverse directions (Hu et al., 2008). The different orientation of these elements relative to the promoter region correspond their roles in controlling the species-specific expression of Cyp26A1 gene in mammals or fish. Previously we reported that the 2533 bp promoter of zebrafish cyp26a1 almost fully recapitulates controlling endogenous expression of cyp26a1 (Hu et al., 2008). Bioinformatics analysis on the sequences up to 3500 bp upstream of zebrafish cyp26a1 start codon reveal that no RARE are present upstream to R2. The results suggest R2 is the distal one and R1 is the proximal among the three RAREs in zebrafish cyp26a1 promoter. In contrast, R3 is the distal one and R1 is the proximal one in mammalian Cyp26A1 promoter (Zhang et al., 2010). However, the core sequences of R3 in mammalian Cyp26A1 promoter is the same as R2 in the zebrafish cyp26a1 promoter though the two RAREs are positioned in different direction. And the core sequences of R2 in mammalian Cyp26A1 promoter are more identical with that of zebrafish R3 than that of zebrafish R2 in cyp26a1 promoter. Additionally, the R2 and R3 double mutated promoter of zebrafish cyp26a1 is still RA responsive (Fig. 2), suggesting there might be other RAREs present in the promoter. This is consistent with the presence of a half RARE site (R4) lying between R1 and R2 in mammalian Cyp26A1 promoter. Taken together, our results suggest that the key cis-elements such as RAREs are conserved in the Cyp26A1 promoter of mammals and fish during evolution. It has been demonstrated in vitro that the multiple RAREs in rat Cyp26a1 promoter work together to maximize the gene’s response to RA signal (Zhang et al., 2010). However, Cyp26A1 responds to RA status differently in different tissues. It is highly responsive to RA level in liver but not in other tissues such as intestine, lung, and kidney (Ray et al., 1997; Wang et al., 2002; Xi and Yang, 2008; Zhang et al., 2010). ChIP assays reveal that both proximal and distal RAREs contribute significantly to the activity of Cyp26A1 promoter in HepG2 cells (from human adult liver) in response to RA whereas A NOVEL RARE IN ZEBRAFISH CYP26A1 PROMOTER only the proximal, but not the distal RARE of the promoter in HEK293T cells (from human embryonic kidney) in response to RA (Zhang et al., 2010). The results indicate that different RAREs may play different roles in the activity of a gene’s promoter in response to RA in different tissues. In this study, our in vitro assay shows that R3 works with R2 and R1 to maximize the response of zebrafish cyp26a1 promoter to RA treatment (Fig. 2). Consistently, our in vivo assays reveal that the reporter driven by cyp26a1 promoter with mutated R3 in Tg(cyp26a1-R3mut:eYFP)nju3/þ is not expressed in the tissues of RA dependent during zebrafish early development but responds to excessive RA in a dose dependent manner though much more weakly than the reporter driven by wild-type promoter in Tg(cyp26a1:eYFP)nju1/þ (Fig. 5). The results suggest that R3 is a key cis-element that controls cyp26a1’s expression in response to endogenous RA at zebrafish early development. The mutation of R3 in cyp26a1 promoter would abolish the expression of cyp26a1 in the regions of retina, anterior dorsal spinal cord, and proctodeum where cyp26a1 expression is required to limit RA activity. Therefore, the mutation of R3 could break RA homeostasis in the specific developmental organs and cause zebrafish embryos to develop abnormally due to RA toxicity. Taken together, our results demonstrate that R3 is crucial for zebrafish cyp26a1 to respond to endogenous RA level during zebrafish early development. LITERATURE CITED Abu-Abed S, Dolle P, Metzger D, Beckett B, Chambon P, Petkovich M. 2001. The retinoic acid-metabolizing enzyme, CYP26A1, is essential for normal hindbrain patterning, vertebral identity, and development of posterior structures. Genes Dev 15:226–240. Begemann G, Schilling TF, Rauch GJ, Geisler R, Ingham PW. 2001. The zebraﬁsh neckless mutation reveals a requirement for raldh2 in mesodermal signals that pattern the hindbrain. Development 128:3081–3094. Dobbs-McAulie B, Zhao Q, Linney E. 2004. Feedback mechanisms regulate retinoic acid production and degradation in the zebraﬁsh embryo. Mech Dev 121:339–350. Emoto Y, Wada H, Okamoto H, Kudo A, Imai Y. 2005. Retinoic acidmetabolizing enzyme Cyp26a1 is essential for determining territories of hindbrain and spinal cord in zebraﬁsh. Dev Biol 278: 415–427. Grandel H, Lun K, Rauch GJ, Rhinn M, Piotrowski T, Houart C, Sordino P, Kuchler AM, Schulte-Merker S, Geisler R, Holder N, Wilson SW, Brand M. 2002. Retinoic acid signalling in the zebraﬁsh embryo is necessary during pre-segmentation stages to pattern the anterior-posterior axis of the CNS and to induce a pectoral ﬁn bud. Development 129:2851–2865. Gu X, Xu F, Song W, Wang X, Hu P, Yang Y, Gao X, Zhao Q. 2006. A novel cytochrome P450, zebraﬁsh Cyp26D1, is involved in metabolism of all-trans retinoic acid. Mol Endocrinol 20:1661–1672. 277 Hernandez RE, Putzke AP, Myers JP, Margaretha L, Moens CB. 2007. Cyp26 enzymes generate the retinoic acid response pattern necessary for hindbrain development. Development 134: 177–187. Hu P, Tian M, Bao J, Xing G, Gu X, Gao X, Linney E, Zhao Q. 2008. Retinoid regulation of the zebraﬁsh cyp26a1 promoter. Dev Dyn 237:3798–3808. Kawakami K, Takeda H, Kawakami N, Kobayashi M, Matsuda N, Mishina M. 2004. A transposon-mediated gene trap approach identiﬁes developmentally regulated genes in zebraﬁsh. Dev Cell 7:133–144. Loudig O, Babichuk C, White J, Abu-Abed S, Mueller C, Petkovich M. 2000. Cytochrome P450RAI(CYP26) promoter: a distinct composite retinoic acid response element underlies the complex regulation of retinoic acid metabolism. Mol Endocrinol 14:1483–1497. Loudig O, Maclean GA, Dore NL, Luu L, Petkovich M. 2005. Transcriptional co-operativity between distant retinoic acid response elements in regulation of Cyp26A1 inducibility. Biochem J 392: 241–248. Niederreither K, Abu-Abed S, Schuhbaur B, Petkovich M, Chambon P, Dolle P. 2002. Genetic evidence that oxidative derivatives of retinoic acid are not involved in retinoid signaling during mouse development. Nat Genet 31:84–88. Niederreither K, Dolle P. 2008. Retinoic acid in development: towards an integrated view. Nat Rev Genet 9:541–553. Perz-Edwards A, Hardison NL, Linney E. 2001. Retinoic acid-mediated gene expression in transgenic reporter zebraﬁsh. Dev Biol 229:89–101. Ray WJ, Bain G, Yao M, Gottlieb DI. 1997. CYP26, a novel mammalian cytochrome P450, is induced by retinoic acid and deﬁnes a new family. J Biol Chem 272:18702–18708. Sakai Y, Meno C, Fujii H, Nishino J, Shiratori H, Saijoh Y, Rossant J, Hamada H. 2001. The retinoic acid-inactivating enzyme CYP26 is essential for establishing an uneven distribution of retinoic acid along the anterio-posterior axis within the mouse embryo. Genes Dev 15:213–225. Wang Y, Zolfaghari R, Ross AC. 2002. Cloning of rat cytochrome P450RAI (CYP26) cDNA and regulation of its gene expression by all-trans-retinoic acid in vivo. Arch Biochem Biophys 401:235–243. Waxman JS, Yelon D. 2007. Comparison of the expression patterns of newly identiﬁed zebraﬁsh retinoic acid and retinoid X receptors. Dev Dyn 236:587–595. White RJ, Nie Q, Lander AD, Schilling TF. 2007. Complex regulation of cyp26a1 creates a robust retinoic acid gradient in the zebraﬁsh embryo. PLoS Biol 5:e304. Xi J, Yang Z. 2008. Expression of RALDHs (ALDH1As) and CYP26s in human tissues and during the neural dierentiation of P19 embryonal carcinoma stem cell. Gene Exp Patterns 8:438–442. Xu F, Li K, Tian M, Hu P, Song W, Chen J, Gao X, Zhao Q. 2009. N-CoR is required for patterning the anterior-posterior axis of zebraﬁsh hindbrain by actively repressing retinoid signaling. Mech Dev 126:771–780. Zhang Y, Zolfaghari R, Ross AC. 2010. Multiple retinoic acid response elements cooperate to enhance the inducibility of CYP26A1 gene expression in liver. Gene 464:32–43. Zhao L, Zhao X, Tian T, Lu Q, Skrbo-Larssen N, Wu D, Kuang Z, Zheng X, Han Y, Yang S, Zhang C, Meng A. 2008. Heart-speciﬁc isoform of tropomyosin4 is essential for heartbeat in zebraﬁsh embryos. Cardiovasc Res 80:200–208.