ARTICLE
Novel Regulatory Sequence
ⴚ82/ⴚ62 Functions as
a Key Element to Drive the Somite-Specificity of
Zebrafish myf-5
Yau-Hung Chen, Hung-Chieh Lee, Chia-Feng Liu, Cheng-Yung Lin, and Huai-Jen Tsai*
Myf-5, a transcription factor that controls muscle differentiation, is expressed in somites during early embryogenesis. However, gene regulation of myf-5 is poorly understood and detailed functional analysis of the regulatory cis-elements is needed. In zebrafish, the myf-5 upstream sequence fromⴚ82 to ⴚ62 (ⴚ82/ⴚ62) was fused with a basal promoter and transferred to fertilized zebrafish eggs. Theⴚ82/ⴚ62 cassette drove green fluorescent protein (GFP) reporter gene expression specifically in the somites. Moreover, GFP signals were detected exclusively in the somites of 28-hpf embryos derived from eggs injected with pCMV-5ⴛ(ⴚ82/ⴚ62), which contained five copies of the ⴚ82/ⴚ62 cassette inserted within cytomegalovirus promoter/enhancer. Thus, the ⴚ82/ⴚ62 cassette, conserved in mouse
myf-5, functions to drive somite-specific expression and to repress nonspecific expression during the early
development of zebrafish embryos. Mutated sequence analysis of ⴚ82/ⴚ62 cassette showed that the ⴚ70/ⴚ62 sequence was the key element for controlling myf-5 specificity. The putative CCAAT-like box, located atⴚ66/ⴚ62, could not direct somite-specific expression. A DNA-protein complex was specifically formed between theⴚ70/ⴚ62 probe and embryonic nuclear extracts. We conclude that the ⴚ70/ⴚ62 motif is essential for controlling somite-specific expression and the CCAAT-like box is essential for activating gene transcription. Developmental Dynamics
228:41–50, 2003.©2003 Wiley-Liss, Inc.
Key words: cis-element; EMSA; GFP; myf-5; somite-specific expression; transgenic zebrafish
Received 7 October 2002; Accepted 14 May 2003
INTRODUCTION
Transcription factors in the basic-he-lix-loop-helix family are essential for skeletal muscle determination and differentiation (Lorenzo-Puri and Sar-torelli, 2000). These myogenic regu-latory factors (MRFs) activate mus-cle-specific transcription by binding to a DNA consensus sequence, an E-box, present in the promoter of numerous muscle-specific genes (Rescan, 2001). Four MRFs, MyoD, Myogenin, Myf-5, and MRF4, have been characterized in humans (Pearson-White, 1991), mice (Pinney
et al., 1995), birds (Saitoh et al., 1993), frogs (Hopwood et al., 1991), and zebrafish (Chen et al., 2000, 2001).
Each MRF may play a different role during myogenesis. For exam-ple, the primary MRFs, MyoD and Myf-5, are involved in myoblast specification, whereas the second-ary MRFs, Myogenin and MRF4, are involved in myotube formation (Rud-nicki and Jaenisch, 1995). Gene knockout experiments in mice dem-onstrated that normal muscle devel-opment still occurred when the
myf-5 or myoD gene was inacti-vated. However, skeletal muscle did not form in transgenic mice in which both myf-5 and myoD were dis-rupted (Rudnicki et al., 1993). Thus, myf-5 and myoD are able to com-pensate for each other to regulate skeletal muscle development. In ze-brafish, myf-5 morphants displayed defects not only in somite pattern-ing, but also in brain formation and epiblast migration, indicating that Myf-5 protein has multiple biological functions during early embryonic development (Chen and Tsai, 2002).
Institute of Molecular and Cell Biology, National Taiwan University, Taipei, Taiwan Grant sponsor: NSC; Grant numbers: 90-2313-B002-260; 91-2313-B-002-331.
*Correspondence to: Huai-Jen Tsai, Institute of Molecular and Cell Biology, National Taiwan University, 1 Roosevelt Road, Sec. 4, Taipei, Taiwan. E-mail: hjtsai@ntu.edu.tw
DOI 10.1002/dvdy.10357
DEVELOPMENTAL DYNAMICS 228:41–50, 2003
Zebrafish myf-5 transcripts are de-tectable 7.5 hours postfertilization (hpf), increase substantially until 16 hpf, and then gradually decline to undetectable levels by 26 hpf (Chen et al., 2001; Coutelle et al., 2001). During somitogenesis, zebrafish myf-5 transcripts are found primarily in the somites and segmental plates (Chen et al., 2001; Coutelle et al., 2001). Prominent signals occurred transiently in adaxial cells, in two parallel rows, but did not extend be-yond the positive-signal somites. Thus, the expression pattern of ze-brafish myf-5 is dynamic and somite-specific.
In mice, it has been shown that myf-5 activation depends on signals from tissues surrounding the somite (Cossu and Borello, 1999). Wnt1, which is present in the dorsal neural tube, activated myf-5 in explant ex-periments (Tajbakhsh et al., 1998). Sonic hedgehog (shh), produced by the notochord, is also required for myogenesis, but only in the epaxial domain. In mice lacking shh, myf-5 is not expressed in the epaxial myo-tome, but hypaxial activation pro-ceeds normally (Borycki et al., 1999). By crossing shh knockout mice strain with germ-line transmission mice, Gustafsson et al. (2002) demon-strated that myf-5 is the direct target for shh signal transduction. These studies demonstrate that signal inte-gration for myf-5 activation is com-plex.
In mice, myf-5 promoter analysis has focused on the long-range or distal fragment of modulation. Had-chouel et al. (2000) cloned a 200-kb regulatory fragment from a YAC li-brary and reported that all essential sequences were located at approx-imately within 96 kb upstream of the transcriptional start site. BAC trans-genic analysis of the Mrf4/ Myf5 lo-cus revealed that interdigitated elements control activation and maintenance of gene expression during muscle development (Carva-jal et al., 2001). However, the actual relationship between cis-acting ele-ments and trans-acting factors on the distal or proximal regulatory re-gions in the mouse myf-5 gene are still unclear. In Xenopus myf-5, an in-terferon regulatory factor-binding el-ement within the Xenopus myf-5
promoter is responsible for the elimi-nation of myf-5 transcription in the mature somitic mesoderm of Xeno-pus embryos (Mei et al., 2001). Re-cently, Lin et al. (2003) found that a T-box binding site was able to medi-ate the dorsal activation of myf-5 in Xenopus gastrula embryos. How-ever, the mechanism regulating ze-brafish myf-5, particularly the inter-action between transcription factors and proximal regulatory elements, is completely unknown.
In zebrafish myf-5, the upstream sequence from nucleotide ⫺82 to ⫺1 (⫺82/⫺1) was able to direct the reporter gene expressed specifically in the somites, whereas the⫺62/⫺1 segment was not (Chen et al., 2001). In this study, we performed in vitro mutagenesis, transgenic experi-ments and an electrophoretic mo-bility shift assay (EMSA) to further study the functional roles of the ze-brafish myf-5, proximal regulatory cis-element ⫺82/⫺62. Interestingly, this short regulatory cassette drove somite-specific expression and re-pressed nonspecific expression dur-ing the early development of ze-brafish embryos.
RESULTS
Cassette
ⴚ82/ⴚ62 Is an
Important cis-Element
Controlling Somite-Specific
Expression of myf-5
We injected DNA fragments con-taining⫺290/⫺1 (pZMYP-290E), ⫺82/ ⫺1(⫺82E), ⫺62/⫺1(⫺62E), or ⫺22/ ⫺1(⫺22E) into one-celled zebrafish embryos. Embryos injected with ⫺290/⫺1 and ⫺82/⫺1 displayed green fluorescent protein (GFP) -positive signals in their somites at 28 hpf, but embryos injected with ⫺62/⫺1 or ⫺22/⫺1 did not (Fig. 1). A DNA fragment in which ⫺82/⫺62 was deleted from myf-5 ⫺290/⫺1 (pZMYP-290E-⌬(⫺82/⫺62)) was in-jected into one-celled embryos. None of these embryos were GFP-positive (Fig. 1).
To determine whether cassette ⫺82/⫺62 was able to direct somite-specific expression of the GFP reporter gene, first, plasmid pEGFPmTATA was constructed, which contained a cyto-megalovirus (CMV) mini-promoter
(TATA box only), fused with GFP and be used as a backbone plasmid for the following three constructs [pEG-FPm-1⫻(⫺82/⫺62), pEGFPm-6⫻(⫺82/ ⫺62), and pEGFPm-Non30fr]. Then, two plasmids, pEGFPm-1⫻(⫺82/⫺62) and pEGFPm-6⫻(⫺82/⫺62), were constructed, which contained one and six copies of ⫺82/⫺62 cassette, fused with pEGFPmTATA, respectively. Finally, a negative control plasmid, pEGFPm-Non30fr, was constructed, which contained a nonspecific cas-sette (Non30fr), fused with pEGFP-mTATA. Only 5.9% of the 197 embryos injected with pEGFPmTATA-injected were GFP-positive, and none were somite-specific (Fig. 2A,B-a,-b). In the nonspecific control group, only 2.9% of the embryos injected with pEG-FPm-Non30fr expressed faint, deep green signals in their somites (Fig. 2A,B-c,-d). However, somite-specific ex-pression rates in transgenic embryos injected with pEGFPm-1⫻(⫺82/⫺62) or pEGFPm-6⫻(⫺82/⫺62) were 16.7% and 24.4%, respectively (Fig. 2A,B-e,-f, -g,-h). Thus, cassette ⫺82/⫺62 was able to direct somite-specific expres-sion of zebrafish myf-5.
Cassette
ⴚ82/ⴚ62 Represses
the Ubiquitous Expression of
CMV Promoter/Enhancer
To investigate further the function of cassette ⫺82/⫺62 in vivo, we con-structed and injected three linear-ized expression plasmids that used GFP as a reporter gene. pCMVm, pCMVm-4⫻(Non30fr), and pCMVm-5⫻(⫺82/⫺62), contained CMV pro-moter/enhancer, CMV promoter/ enhancer with four inserted copies of Non30fr, and CMV promoter/en-hancer with five inserted copies of cassette ⫺82/⫺62, respectively. In zebrafish embryos injected with pC-MVm, GFP was expressed ubiqui-tously at 28 hpf (Fig. 3, bottom left). The pattern of GFP-fluorescence in embryos injected with the pCMVm-4⫻(Non30fr) was similar to that of embryos injected with pCMVm. Thus, the nonspecific sequence, Non30fr, inserted within the CMV promoter/enhancer did not affect transcription controlled by the CMV promoter/enhancer (Fig. 3, middle left). In contrast, green fluorescent signals were detected exclusively in
the somites of embryos injected with pCMVm-5⫻(⫺82/⫺62) (Fig. 3, top left). Thus, the zebrafish myf-5 cas-sette ⫺82/⫺62 suppressed non– somite-specific expression of the GFP reporter gene driven by the CMV promoter/enhancer.
ⴚ70/ⴚ62 Motif Within ⴚ82/ⴚ62
Is the Key Element for Somite
Specificity
Polymerase chain reaction (PCR) -based in vitro mutagenesis and transgenic assays were conducted to dissect cassette⫺82/⫺62. GFP ex-pression plasmids were constructed with sequentially mutated se-quences within ⫺82/⫺62. Plasmid pZMYP290E-M1 was mutated at ⫺82/⫺79. The other plasmids were mutated at ⫺78/⫺75 (-M2), ⫺74/⫺71 (-M3), ⫺70/⫺67 (-M4), and ⫺66/⫺62 (-M5; Fig. 4A, left). The somite-specific expression rates in transgenic embryos injected with pZMYP290E, pZMYP290E-M1, -M2, and -M3, were 68.8%, 71.4%, 71%, and 65.4%, respectively (Fig. 4A, right; 4B-a-d). The somite-specific
ex-pression rates in pZMYP290E-M4- and -M5-injected embryos were only 43.2% and 25.3%, respectively (Fig. 4A, right; 4B-e,-f). Moreover, the myocyte-specific expression rates in embryos injected with pZMYP290E-M4- and -M5 were only 37.5% and 7.7%, respectively (Fig. 4A, right). Thus,⫺70/⫺62 has an important role in the regulation of zebrafish myf-5. To determine whether the short, ⫺70/⫺62 motif had the same biolog-ical functions as the cassette ⫺82/ ⫺62, we constructed plasmid pZMYP-70E, which contains one ⫺70/⫺1 upstream element of ze-brafish myf-5. Based on transgenic analysis, the somite-specific expres-sion rate was 41.9% in pZMYP-70E-injected embryos (Fig. 5).
The ⫺64/⫺60 sequence of ze-brafish myf-5 is CCAAT (Chen et al., 2001), indicating that the ⫺66/⫺62 motif contains part of the CCAAT-like box core sequence. To study the regulation of the CCAAT-like box in detail, we mutated the core se-quence of the CCAAT-like box in pZMYP-290E from CAAT to AACC. The resulting plasmid, pZMYP-mcat,
was microinjected into zebrafish bryos. In pZMYP-mcat–injected em-bryos, the total expression rate, somite-specific expression rate, and myocyte-specific expression rate were 15.7%, 5.5%, and 4.3%, respec-tively (Fig. 4A,B-g). Then, we con-structed plasmid pZMYP-64E, which contains the CCAAT-like box core se-quence. No somite-specific GFP sig-nals were detected in the pZMYP-64E-injected embryos, indicating that the CCAAT-like box was not able to drive somite-specific expression of zebrafish myf-5 (Fig. 5). Finally, to determine whether the motifs ⫺66/⫺62 and ⫺70/⫺62 could direct somite-specific expression of the GFP reporter gene, the GFP expression plasmids pEGFPm-4⫻(⫺66/⫺62) and pEGFPm-3⫻(⫺70/ ⫺62) were constructed and injected. In pEGFPm-4⫻(⫺66/⫺62)-injected embryos, the total expression rate was 43.6%, but the somite-specific expres-sion rate was only 1.1% (Fig. 6A,B-a,-b). However, in pEGFPm-3⫻(⫺70/ ⫺62)-injected embryos, the total expression rate and somite-specific expression rates were 47% and 41.8%, respectively (Fig. 6A,B-c). Therefore,
Fig. 1. Cassette ⫺82/⫺62 is essential for zebrafish myf-5 gene expression. Microinjected plasmids are illustrated on the left. Plasmids pZMYP-290E, -82E, -62E, and -22E contain the upstream regions of zebrafish myf-5 from⫺1 to ⫺290 (⫺290/⫺1), ⫺1 to ⫺82 (⫺82/⫺1), ⫺1 to ⫺62 (⫺62/⫺1), and ⫺1 to ⫺22 (⫺22/⫺1), respectively. Plasmid pZMYP-290E-⌬(⫺82/⫺62) contains cassette ⫺82/⫺62, which was deleted from the ⫺290/⫺1 segment. The linearized plasmids (25 ng/l) were microinjected into fertilized zebrafish eggs. The total expression rate (filled bars), somite-specific expression rate (open bars), myocyte-specific expression rate (crossed bars), and nonspecific rate (hatched bars) are described in the Experimental Procedures section. For each construct, the total number of embryos that survived microinjection is shown (n).
minimal TATA-box derived from a CMV promoter fused with EGFP gene; a nonspecific DNA fragment, Non-30fr, fused with pEGFPmTATA; one copy of cassette⫺82/⫺62 fused with pEGFPmTATA; and six copies of cassette ⫺82/⫺62 fused with pEGFPmTATA, respectively. Right: Calculation of total expression (filled bars) and somite-specific expression rates (open bars) are described in the Experimental Procedures section. For each construct, the total number of embryos that survived microinjection is shown (n). B: Embryos were photographed under brightfield illumination (a, c, e, and g) and flourescent light (b,d,f,h). In pEGFPm-1⫻(⫺82/⫺62)- (f) and pEGFPm-6⫻(⫺82/⫺62)-injected zebrafish, EGFP signals appeared as bars with sharp edges (h).
⫺66/⫺62 sequence directs ubiquitous expression and⫺70/⫺62 sequence is necessary for somite-specific expres-sion. The short motif ⫺70/⫺62 is the key element within the zebrafish myf-5 cassette⫺82/⫺62.
Embryonic Nuclear Extract
Specifically Interacted With
Cassette
ⴚ82/ⴚ62
Electrophoretic mobility shift assay (EMSA) experiments were con-ducted to determine whether a pro-tein in the embryonic nuclear ex-tract specifically interacted with cassette ⫺82/⫺62. As expected, adding a nonspecific (Non30fr) probe did not cause the formation of a shifted band (Fig. 7, lanes 1–3). However, a complex did form be-tween embryonic nuclear extract and the⫺82/⫺62 probe, producing the shifted band shown on the gel (Fig. 7, lane 5). The intensity of the shifted complex increased with the amount of extract (Fig. 7, lane 6), but the shifted complex became very faint when excess amounts of cold⫺82/⫺62 oligonucleotide were added (Fig. 7, lanes 7 and 8). How-ever, addition of excess amounts of cold, nonspecific Non30fr oligonu-cleotide did not change the inten-sity of the shifted band (Fig. 7, lanes 9 and 10). Thus, a specific complex formed between embryonic nuclear extract and the⫺82/⫺62 probe.
Excess amounts of mutated DNA segments were added, including oli-gonucleotides M1 (mutation at⫺82/ ⫺79), M2 (mutation at ⫺78/⫺75), M3 (mutation at ⫺74/⫺71), M4
(muta-tion at⫺70/⫺67), and M5 (mutation at⫺66/⫺62). Only cold M1, M2, and M3 oligonucleotides were able to compete the formation of the shifted complex (Fig. 7, lanes 11–16), indicating the ⫺82/⫺71 segment was not involved in forming the spe-cific complex between the extract and the ⫺82/⫺62 probe. However, addition of unlabeled M4 oligonu-cleotides partially blocked complex formation, and addition of unla-beled M5 failed to block it (Fig. 7, lanes 17–20). Therefore, the⫺70/⫺62 element is bound by protein(s) in the embryonic nuclear extract, and the ⫺66/⫺62 motif is the key element for the formation of the complex be-tween embryonic nuclear extract and cassette⫺82/⫺62.
Mouse and Zebrafish myf-5
Gene Sequence Homology
and Conservation of Cassette
ⴚ82/ⴚ62 Functions
In addition, the short, mouse myf-5 motif ⫺151/⫺144 (CCTGGCCA) is identical to the zebrafish myf-5 motif ⫺69/⫺62 (CCTGGCCA) (Fig. 4A). To study the conservation of cassette function, we constructed plasmid pZMYP290E-m⫺82/⫺62m, in which the zebrafish cassette⫺82/⫺62 was replaced by the mouse cassette ⫺151/⫺144. This new plasmid was microinjected into zebrafish em-bryos. The somite-specific expression rate was 67.7%, and the myocyte-specific expression rate was 64.5%. These rates were not significantly dif-ferent than they those for pZMYP-290E-injected embryos (Fig. 4A,B-h). Thus, the functions of zebrafish⫺82/ ⫺62 and mouse ⫺161/⫺144 cas-settes have been conserved.
DISCUSSION
Myf-5 is a key MRF during vertebrate myogenesis. The mechanism regu-lating the myf-5 gene is extremely complicated and little known. In mice, promoter analysis of myf-5 has focused on the long-range or distal fragment of modulation (⬃200 kb). Discrete and dispersed enhancers specific to particular populations of skeletal muscle precursors have been found to control expression of myf-5 (Hadchouel et al., 2000;
Sum-merbell et al., 2000; Carvajal et al., 2001). In Xenopus, the proximal reg-ulatory elements of myf-5 from ⫺1869 to ⫺30 bp have been char-acterized, and an interferon regula-tory factor-binding element was found to eliminate myf-5 transcrip-tion (Mei et al., 2001). In fish, the ⫺82-bp sequence upstream from the transcription initiation site of ze-brafish myf-5 acts as a basal pro-moter (Chen et al., 2001). Here, we provide new insights about the prox-imal regulation of zebrafish myf-5. With regard to the results of dele-tion, replacement, mutagenesis, and regulatory cassette analysis, we suggest that cassette⫺82/⫺62 is es-sential for directing somite-specific expression of the zebrafish myf-5 gene and is capable of repressing ubiquitous expression of CMV pro-moter/enhancer (Fig. 3). A similar concept is proposed for Xenopus myf-5 by Polli and Amaya (2002), who identified HBX2, a 1.2-kb proxi-mal upstream element of Xenopus myf-5, which is necessary for both activation and repression of Xeno-pus myf-5 expression. Lin et al. (2003) identified a 42-bp T-box binding site containing DNA segment, which is able to mediate dorsal activation. However, this study is the first report to prove that a cassette as small as 21 bp has unique functions, espe-cially in the repression of CMV pro-moter/enhancer.
Interestingly, we found that the ze-brafish myf-5 cassette⫺82/⫺62 (CTCT-TAGCTCTGTCCTGGCCA) shared 62% nucleic acid identity with the mouse myf-5 cassette⫺161/⫺144 (CACTGAC-CGACCCTGGCCA). Transgenic analy-sis with plasmid pZMYP290E-m⫺82/ ⫺62m, in which the zebrafish cassette ⫺82/⫺62 was replaced by the mouse cassette ⫺151/⫺144, demonstrated the mouse cassette was able to direct somite-specific expression in zebrafish embryos. The biological functions of ze-brafish ⫺82/⫺62 and mouse ⫺161/ ⫺143 myf-5 cassettes are worth further study.
Nuclear protein complexes specif-ically interacted with cassette⫺82/ ⫺62; the motif ⫺70/⫺62 was the key element (Fig. 7). The sequence of zebrafish myf-5 ⫺64/⫺60 is CCAAT (Chen et al., 2001), indicating that the motif ⫺66/⫺62 contains part of
Fig. 3. Repression of nonspecific green
fluo-rescent protein (GFP) expression in embryos microinjected with cytomegalovirus (CMV) promoter fused with cassette⫺82/⫺62. Plas-mids pCMVm-, pCMVm-4⫻(Non30fr)- and pCMVm-5⫻(⫺82/⫺62) contained CMV pro-moter/enhancer, CMV promoter/enhancer with 4, inserted copies of Non30fr and CMV promoter/enhancer with 5, inserted copies of cassette ⫺82/⫺62, respectively. Each one-celled, fertilized zebrafish egg was microin-jected with one type of linearized plasmid. GFP-positive signals were observed through-out the body of pCMVm- (bottom) and pC-MVm-4⫻(Non30fr)-injected embryos (mid-dle) but only in the somites of pCMVm-5⫻(⫺82/⫺62)-injected embryos (top).
Fig. 4.
Fig. 6.
quences of cassette⫺82/⫺62. A: A sche-matic of plasmid pZMYP290E (290E) and de-rivates containing mutated sequences. Dotted bars represent the green fluorescent protein (GFP) reporter gene. Fertilized eggs were microinjected with linearized plas-mids 290E, 290E-M1 (mutation at⫺82/⫺79), 290E-M2 (⫺78/⫺75), 290E-M3 (⫺74/⫺71), 290E-M4 (⫺70/⫺67), 290E-M5 (⫺66/⫺62), -mcat (⫺64/⫺61), or -m⫺82/62m (in which zebrafish ⫺82/⫺62 was replaced with mouse⫺161/⫺144). Dots indicate nucleo-tides identical to those of 290E. Dashes rep-resent gaps created to maximize the iden-tity among the plasmid sequences. The total expression rate (filled bars), somite-specific expression rate (open bars), myo-cyte-specific expression rate (crossed bars), and nonspecific rate (hatched bars) are described in the Experimental Proce-dures section. For each construct, the total number of embryos that survived microin-jection is shown (n). B: Embryos were pho-tographed with fluorescence microscopy (a– h).
the CCAAT-like box core sequence. The CCAAT-like box sequence is conserved in zebrafish and mouse myf-5 genes. Typically, the trans-act-ing factor bound to the CCAAT-box is CCAAT-box binding factor (CBF) or nuclear factor Y (NF-Y; Lindahl et al., 2002). CBF and NF-Y function as ubiquitous transcription activators at the proximal promoter. Because nei-ther CBF nor NF-Y is known to repress nonspecific expression or direct somite-specific expression, it is highly unlikely that protein from the nuclear extract interacted with cassette ⫺82/⫺62 or that the overlapped CCAAT-like box is either CBF or NF-Y alone.
To investigate the functions of cas-sette ⫺70/⫺62 further, we con-structed and injected plasmid
pEG-FPm-3⫻(⫺70/⫺62). Cassette ⫺66/⫺62 was not able to drive somite-specific expression, and no myocyte-specific GFP signals were detected. Therefore, cassette ⫺66/⫺60 of zebrafish myf-5 (CCAAT-like box) is functional. It di-rects the transcriptional initiation of ze-brafish myf-5. The function of cassette ⫺70/⫺62 is to recruit a factor to drive somite-specific expression. Several studies have found that a ubiquitous transcription factor interacts with a cofactor to drive tissue-specific ex-pression. In neuron-specific expres-sion, a neuron-restricted transcription factor, MASH1, interacted with CBF to drive tissue-specific expression (Man-dolesi et al., 2002). Cardiac-specific expression was directed by the inter-action of a heart-specific factor, myo-cardin, with a ubiquitous serum re-sponse factor bound to a CArG box (Wang et al., 2001). We hypothesize that a somite-specific transcription factor(s) interacts with cassette⫺70/ ⫺67, or with CBF/NF-Y, to control the unique functions of cassette ⫺82/ ⫺62. This hypothesis merits further in-vestigation.
EXPERIMENTAL PROCEDURES
Fish
Zebrafish (AB strain) were reared at approximately 28.5°C, under a 14 hr light/10 hr dark photoperiod. After fertilization, eggs were collected and cultured in a fish tank.
Embry-onic cleavage number and somite formation were observed with a light microscope to determine the devel-opmental stage (Kimmel et al., 1995).
Plasmid Construction and
PCR-Based In Vitro Mutagenesis
For deletion experiments, pZMYP-290E (Chen et al., 2001) was used as a template for PCR reactions. All oli-gonucleotide sequences used in this study are shown in Table 1. Primers ZMfg-64F-HindIII (or ZMfg-70F-HindIII) and ZMfg-1R-BamHI were used to produce ⫺64/⫺1 (or ⫺70/⫺1) frag-ments with HindIII and BamHI restric-tion enzyme sites on both ends. Then, the HindIII-BamHI-digested ⫺64/⫺1 and ⫺70/⫺1 fragments were ligated to HindIII-BamHI-di-gested pEGFP-1 (Clontech) vector to produce 64E and pZMYP-70E, in which the GFP reporter gene was fused with⫺64/⫺1 and ⫺70/⫺1, respectively.
For internal deletion, replace-ment, and mutation experiments, pZMYP-290E was used as a template for three combination-PCR reactions to generate mutation and deletion constructs. Constructs were gener-ated by using a modified, PCR-based, in vitro mutagenesis method (Swaminathan et al., 2001). The first PCR product (240 bp) was pro-duced by using a forward (D30f) and a reverse primer (D260r). Then, another PCR product (117 bp) was produced with primers D255f and D381r. Equal amounts of the 240-and 117-bp PCR products were mixed together, denaturated for 5 min at 94°C, and annealed at 37°C for 2 hr. Then, 5 U of Klenow frag-ments were added, and the mixture was incubated at 37°C for 1 hr. The resultant product was ligated to SmaI-digested pEGFP-1 (Clonetech) vectors to generate pZMYP-290E-⌬(⫺82/⫺62), in which the ⫺82/⫺62 element was deleted from the ⫺290/⫺1 zebrafish myf-5 gene up-stream fragment. By using the pro-cedures described above, two pairs of single-stranded oligonucle-otides (D30f and Pmcat-r, Pmcat-f and D381r; D30f and Pm8262mr, Pm8262mf, and D381r) were used to construct pZMY-290E-mcat
(mu-Fig. 5. The short motif⫺70/⫺62 in cassette ⫺82/⫺62 directs green fluorescent protein (GFP) expression specifically in the somites. Plasmids pZMYP-82E, -70E, -64E, and -62E contain the upstream regions⫺82 (⫺82/⫺1), ⫺1 to ⫺70 (⫺70/⫺1), ⫺1 to ⫺64 (⫺64/⫺1), and⫺1 to ⫺62 (⫺62/⫺1), respectively. The linearized plasmids (25 ng/l) were microin-jected into fertilized zebrafish eggs. Total expression rate (filled bars), somite-specific expression rate (open bars). For each construct, the total number of embryos that survived microinjection is shown (n).
Fig. 6. Cassettes⫺66/⫺62 and ⫺70/⫺62
di-rect ubiquitous and somite-specific expres-sion, respectively. Plasmid pEGFPm-4⫻(⫺66/⫺62) included four copies of cassette⫺66/⫺62 fused with pEGFPmTATA, and pEGFPm-3⫻(⫺70/⫺62) contained three copies of cassette⫺70/⫺62 fused with pEGFPmTATA. A: Total expression rate (filled bars) and somite-specific expression rate (open bars). For each construct, the total number of embryos that survived microin-jection is shown (n). B: Embryos were pho-tographed with fluorescence microscopy (a– c). In pEGFPm-4⫻(⫺66/⫺62)- (a,b, ar-rowhead) and pEGFPm-3⫻(⫺70/⫺62)- (c, arrow) injected embryos, zebrafish EGFP signals appeared as bars with sharp edges.
tated at⫺64/⫺61) and pZMYP-290E-m-82/62m (to replace zebrafish⫺82/ ⫺62 with mouse ⫺161/⫺144), respectively. For mutagenesis analy-sis by using the same procedures, two pairs of single-stranded oligonu-cleotides (D30f and Pm1r, Pm1f and D381r; D30f and Pm2r, Pm2f and D381r; D30f and Pm3r, Pm3f and D381r; D30f and Pm4r, Pm4f and D381r; and D30f and Pm5r, Pm5f, and D381r) were used to construct pZMYP290E-M1, -M2, -M3, -M4, and -M5, respectively. The five constructs contained four sequentially mu-tated sequences in the⫺82/⫺62 re-gion.
For regulatory cassette analysis, forward 82F) and reverse (ZMF-62R) primers, were synthesized and annealed at 50°C for 30 min to gen-erate double-stranded DNA frag-ments of cassette ⫺82/⫺62. EcoRI-cut pEGFPmTATA, which contained a minimal TATA-box derived from a CMV promoter fused with EGFP gene (Ma et al., 2001; Wang et al.,
2002), was blunted. Then, cassette ⫺82/⫺62 was added to generate pEGFPm-1⫻(⫺82/⫺62), which con-tained one copy of cassette ⫺82/ ⫺62, and pEGFPm-6⫻(⫺82/⫺62), which contained six copies of cas-sette ⫺82/⫺62. By using the same procedures, plasmid pEGFPm-4⫻(⫺66/⫺62), with four copies of cassette ⫺66/⫺62, and plasmid pEGFPm-3⫻(⫺70/⫺62), with three copies of cassette ⫺70/⫺62, were constructed. A control plasmid, pEGFPm-Non30fr, which contained a minimal TATA-box fused with a nonspecific DNA fragment, Non-30fr, was constructed by using prim-ers Non-30f and Non-30r (Chen et al., 2002).
For nonspecific repression experi-ments, cassette ⫺82/⫺62 or a non-specific DNA fragment, Non-30fr, were ligated to a ScaI-digested pCMVm fragment. In pCMVm-5⫻(⫺82/⫺62), the CMV promoter/ enhancer sequences were fused with five copies of cassette
⫺82/⫺62. In pCMVm-4⫻(Non30fr), CMV was fused with four copies of nonspecific DNA fragments.
Microinjection and Green
Fluorescence Detection
EGFP cDNA was fused with different mutated sequences of the zebrafish myf-5 upstream region, linearized, recovered from gel, quantified, and diluted to appropriate concentra-tions with 0.1% phenol red in phos-phate-buffered saline. Fertilized eggs, collected before their first cleavage, received cytoplasmic mi-croinjections of EGFP cDNA con-structs (Wang et al., 2002). The eggs were incubated at 28.5°C.
Transgenic embryos were ob-served hourly, especially from 9 to 96 hpf, under a stereo dissecting micro-scope (MZ12, Leica) equipped with a fluorescent module having an en-hanced GFP filter cube (Kramer Sci-entific). Photographs were taken with a MPS 60 camera (Leica) and
Fig. 7. Cassette⫺82/⫺62 and embryonic nuclear extract complex formation were studied with electrophoretic mobility shift assay (EMSA) and competitive EMSA. Nuclear extract isolated from 24 hours postfertilization embryos (5g, lanes 2 and 5, indicated with a ⫹; or 20 g, lanes 3, 6 –20,⫹⫹) was reacted with32P-radiolabeled oligonucleotide probes (cassette⫺82/⫺62 and nonspecific DNA, Non-30fr). Radiolabeled probes (nonspecific DNA, Non-30fr, and cassette⫺82/⫺62) without added nuclear extract served as negative controls (lanes 1, 4). The arrow indicates shifted bands formed by double-stranded oligonucleotides and extract proteins. Competitors⫺82/⫺62, nonspecific DNA (Non30fr), mutated DNA M1, M2, M3, M4, and M5 were prepared as described in the Experimental Procedures section.
FUJI 400 ASA film when embryos de-veloped at 28 hpf. Four different EGFP expression rates were calcu-lated as described before (Chen et al., 2001). The total expression rate is the proportion of surviving embryos that expressed EGFP in any cell. The somite-specific expression rate is the proportion of surviving embryos that expressed EGFP exclusively in the somites. The myocyte-specific expres-sion rate is the proportion of surviving embryos that expressed EGFP
exclu-sively in the somites and exhibited green signals in the rod-like cells (myo-cyte). The nonspecific expression rate is the proportion of surviving embryos in which EGFP was expressed in the somites and in other cells.
Embryonic Nuclear Extracts
Nuclear proteins were extracted by us-ing the procedures described by Dig-nam et al. (1983) with some modi-fications. Approximately 200 28-hpf
embryos were collected, and 1 ml of buffer A (10 mM HEPES pH 7.9, 1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT, and 0.1
mM PMSF) was added. After homoge-nization, samples were centrifuged at 16,100⫻ g at 4°C for 30 min. The pellet was resuspended with 300l of buffer C (20 mM HEPES pH 7.9, 25% glycerol, 1.5 mM MgCl2, 0.42 M NaCl, 0.2 mM
ethyl-enediaminetetraacetic acid [EDTA], 0.5 mM dithiothreitol [DTT], and 0.5 mM phenylmethyl sulfonyl fluoride [PMSF]), shaken gently at 4°C for 30 min, and
TABLE 1. Oligonucleotide Sequences That Were Used for Mutagenesis, Internal Deletion, Deletion, Regulatory Cassette Analysis, and EMSAa
Experiments Symbols Sequences
Internal deletion and mutagenesis
D30f CTCGAGCTCAAGCTTGCATGCCTC D260r TGCCCCCATACCCCCCTCTCTCAC D255f GGGGGTATGGGGGCACGGTCATTAG D381r GGGGTGGATCCGATTGGTTT Pm1f AGAGTAGCTCTGTCCTGGCCAATC Pm1r GCTACTCTACCCCCCTCTCACCATC Pm2f GCTATCTGTCCTGGCCAATGGG Pm2r GACAGAGCAGAGACCCCCCTCT Pm3f GAGTTCCTGGCCAATGGGGGCA Pm3r GCCAGGAACTCGCTAAGAGACCCCC Pm4f GAAGGGCCAATGGGGGCACGGT Pm4r TTGGCCCTTCCAGAGCTAAGAGAC Pm5f TTAACAATGGGGGCACGGTCATTAG Pm5r CCCATGTTAAAGGACAGAGCTAAG Pmcat-f TAGCTCTGTCCTGGAACCTGGG Pmcat-r GGTTCCAGGACAGAGCTA Pm8262mf CACTGACCGACCCTGGCCAATGGGGGCACGGTCATTAGGGCTG Pm8262mr TCGGTCAGTGACCCCCCTCTCTCACCATCTAGATCCCCAC Deletion ZMf5g-70F-HindIII AAGCTTTCCTGGCCAATGGGGGCACG ZMf5g-64F-HindIII AAGCTTCCAATGGGGGCACGGTCATT ZMf5g-1R-BamHI GGATCCGATTGGTTTGGTGTTG Regulatory cassette analysis
ZMf7062f AATTTCCTGGCCATCCTGGCCATCCTGGCCA ZMf7062r AATTTGGCCAGGATGGCCAGGATGGCCAGGA ZMf6662f AATTGGCCAGGCCAGGCCAGGCCA ZMf6662r AATTTGGCCTGGCCTGGCCTGGCCG EMSA ZMF-82F CTCTTAGCTCTGTCCTGGCCA ZMF-62R TGGCCAGGACAGAGCTAAGAG Non-30f CACGTCACGAGCTATCGGTGATCATCTCTG Non-30r GTGCAGTGCTCGATAGCCACTAGTAGAGAC ZMF-M1F AGAGTAGCTCTGTCCTGGCCA ZMF-M1R TGGCCAGGACAGAGCTACTCT ZMF-M2F CTCTGCTATCTGTCCTGGCCA ZMF-M2R TGGCCAGGACAGAATGCAGAG ZMF-M3F CTCTTAGCGAGTTCCTGGCCA ZMF-M3R TGGCCAGGAACTCGCTAAGAG ZMF-M4F CTCTTAGCTCTGGAAGGGCCA ZMF-M4R TGGCCCTTCCAGAGCTAAGAG ZMF-M5F CTCTTAGCTCTGTCCTTTAAC ZMF-M5R GTTAAAGGACAGAGCTAAGAG
aEMSA, electrophoretic mobility shift assay.
centrifuged at 16,100⫻ g at 4°C for 30 min. The supernatant was dialyzed against 1 liter of buffer D (20 mM HEPES pH 7.9, 20% glycerol, 0.1 M KCl, 0.2 mM EDTA, 0.5 mM DTT, and 0.5 mM PMSF) at 4°C for 5 hr. Finally, all samples were centrifuged again at 16,100⫻ g at 4°C for 30 min. The supernatants were stored at⫺70°C.
EMSA
Two double-stranded oligonucleo-tides, cassette⫺82/⫺62 and nonspe-cific Non-30fr, were used as probes for the binding assay with embryonic nu-clear extracts. All probes were la-beled with ␥-[32P]ATP (3,000 Ci/ml)
by using T4 polynucleotide kinase (NEB), according to the supplier’s pro-tocols. Embryonic nuclear extract (5 or 20g) and 1 g of poly(dIdC) were added to the reaction buffer (10 mM Tris at pH 7.5, 50 mM NaCl, 0.5 mM EDTA pH 8.0, 0.5 mM DTT, 5% glycerol). For competitive inhibition, five pairs of single-stranded oligonucleotides, ZMF-M1F and -M1R; -M2F and -M2R; -M3F and -M3R; -M4F and -M4R; or -M5F and -M5R were annealed to form five double-stranded oligonucle-otides, which contained four sequen-tially mutated sequences in the⫺82/ ⫺62 region. The mixtures were placed on ice for 10 min. After adding 1l of probe with a specific radioactivity of 106cpm/g, each mixture was
incu-bated at 30°C for 30 min, and analyzed by 6% acrylamide gel elec-trophoresis (79:1 acrylamide:bisacryl-amide). Then, the gel was dried and exposed to X-ray film for 14 days.
REFERENCES
Borycki AG, Brunk B, Tajbakhsh S, Buck-ingham M, Chiang C, Emerson CP Jr. 1999. Sonic hedgehog controls epaxial muscle determination through Myf5 activation. Development 126:4953–4063. Carvajal JJ, Cox D, Summerbell D, Rigby
PWJ. 2001. A BAC transgenic analysis of the Mrf4/ Myf5 locus reveals interdigi-tated elements that control activation and maintenance of gene expression during muscle development. Develop-ment 128:1857–1868.
Chen YH, Tsai HJ. 2002. Treatment with myf5-morpholino results in somite pat-terning and brain formation defects in zebrafish. Differentiation 70:447–456. Chen YH, Lee WC, Cheng CH, Tsai HJ.
2000. Muscle regulatory factor gene: Zebrafish (Danio rerio) myogenin cDNA. Comp Biochem Physiol 127B:97–103.
Chen YH, Lee WC, Liu CF, Tsai HJ. 2001. Molecular structure, dynamic expres-sion and promoter analysis of zebrafish (Danio rerio) myf-5 gene. Genesis 29: 22–35.
Chen YH, Liang CT, Tsai HJ. 2002. Expres-sion, purification and DNA-binding ac-tivity of tilapia muscle-specific tran-scription factor, MyoD, produced in Escherichia coli. Comp Biochem. Physiol 131B:795–805.
Cossu G, Borello U. 1999. Wnt signaling and the activation of myogenesis in mammals. EMBO J 18:6867–6872. Coutelle O, Blagden CS, Hampson R,
Halai C, Rigby PWJ, Hughes SM. 2001. Hedgehog signaling is required for maintenance of myf5 and myoD ex-pression and timely terminal differenti-ation in zebrafish adaxial myogenesis. Dev Biol 236:136 –150.
Dignam JD, Lebovitz RM, Roeder RG. 1983. Accurate transcription initiation by RNA polymerase II in a soluble ex-tract from isolate mammalian nuclei. Nucleic Acids Res 11:1475–1489. Gustafsson MK, Pan H, Pinney DF, Liu Y,
Lewandowski A, Epstein DJ, Emerson CP Jr. 2002. myf5 is a direct target of long-range Shh signaling and Gli regu-lation for muscle specification. Genes Dev 16:114 –126.
Hadchouel J, Tajbakhsh S, Primig M, Chang THT, Daubas P, Rocancourt D, Buckingham M. 2000. Modular long-range regulation of Myf5 reveals unex-pected heterogeneity between skele-tal muscles in the mouse embryo. Development 127:4455–4467.
Hopwood ND, Pluck A, Gurdon JB. 1991. Xenopus Myf-5 marks early muscle cells and can activate muscle genes ec-topically in early embryos. Develop-ment 111:551–560.
Kimmel CB, Ballard WW, Kimmel SR, Ull-mann B, Schilling TF. 1995. Stages of embryonic development of the ze-brafish. Dev Dyn 203:253–310.
Lin GF, Geng X, Chen Y, Qu B, Wang F, Hu R, Ding X. 2003. T-box binding site me-diates the dorsal activation of myf-5 in Xenopus gastrula embryos. Dev Dyn 226: 51–58.
Lindahl GE, Chambers RC, Papakrivo-poulou J, Dawson SJ, Jacobsen MC, Bishop JE, Laurent GJ. 2002. Activation of fibroblast procollagen alpha 1 (I) transcription by mechanical strain is transforming growth factor-beta-de-pendent and involves increased bind-ing of CCAAT-bindbind-ing factor (CBF/ NF-Y) at the proximal promoter. J Biol Chem 277:6153–6161.
Lorenzo-Puri P, Sartorelli V. 2000. Regula-tion of muscle regulatory factors by DNA-binding, interacting proteins, and post-transcriptional modifications. J Cell Physiol 185:155–173.
Ma GC, Wang TM, Su CY, Wang YL, Chen S, Tsai HJ. 2001. Retina-specific cis-ele-ments and binding nuclear proteins of carp rhodopsin gene. FEBS Lett 508:265– 271.
Mandolesi G, Gargano S, Pennuto M, Illi B, Molfetta R, Soucek L, Mosca L, Levi A, Jucker R, Nasi S. 2002. NGF-dependent and tissue-specific transcription of vgf is regulated by a CREB-p300 and bHLH factor interaction. FEBS Lett 510:50 –56. Mei W, Yang J, Tao Q, Geng X, Rupp
RAW, Ding X. 2001. An interferon regu-latory factor-like binding element re-stricts Xmyf-5 expression in the posterior somites during Xenopus myogenesis. FEBS Lett 505:47–52.
Pearson-White SH. 1991. Human myoD: cDNA and deduced amino acid se-quence. Nucleic Acids Res 19:1148 – 1148.
Pinney DF, de la Brousse FC, Faerman A, Shani M, Maruyama K, Emerson CP Jr. 1995. Quail myoD is regulated by a complex array of cis-acting control se-quences. Dev Biol 170:21–38.
Polli M, Amaya E. 2002. A study of meso-derm patterning through the analysis of the regulation of Xmyf-5 expression. Development 129:2917–2927.
Rescan PY. 2001. Regulation and func-tions of myogenic regulatory factors in lower vertebrates. Comp Biochem Physiol 130B:1–12.
Rudnicki MA, Jaenisch R. 1995. The MyoD family of transcription factors and skel-etal myogenesis. Bioessays 17:203–209. Rudnicki MA, Schnegelsberg PN, Stead RH, Braun T, Arnold HH, Jaenisch R. 1993. MyoD or Myf-5 is required for the formation of skeletal muscle. Cell 75: 1351–1359.
Saitoh O, Fujisawa-Sehara A, Nabeshima Y, Periasamy M. 1993. Expression of myogenic factors in denervated chicken breast muscle: isolation of the chicken Myf5 gene. Nucleic Acids Res 21:2503–2509.
Summerbell D, Ashby PR, Coutelle O, Cox D, Yee SP, Rigby PWJ. 2000. The expres-sion of Myf5 in the developing mouse embryo is controlled by discrete and dis-persed enhancers specific for particular populations of skeletal muscle precur-sors. Development 127:3745–3757. Swaminathan S, Ellis HM, Waters LS, Yu D,
Lee EC, Court DL, Sharan SK. 2001. Rapid engineering of bacteria artificial chromosomes using oligonucleotides. Genesis 29:14 –21.
Tajbakhsh S, Borello U, Vivarelli E, Kelly R, Papkoff J, Duprez D, Buckingham M, Cossu G. 1998. Differential activation of Myf5 and MyoD by different Wnts in explants of mouse paraxial mesoderm and the later activation of myogenesis in the absence of Myf5. Development 125:4155–4162.
Wang DZ, Chang PS, Wang Z, Sutherland L, Richardson JA, Small E, Krieg PA, Ol-son EN. 2001. Activation of cardiac gene expression by myocardian, a transcriptional cofactor for serum re-sponse factor. Cell 105:851–862. Wang TM, Chen YH, Liu CF, Tsai HJ. 2002.
Functional analysis of the proximal pro-moter regions of fish rhodopsin and myf-5 genes using transgenesis. Mar Biotechnol 4:247–255.
