Promoter architecture and transcriptional regulation of musculoskeletal embryonic nuclear protein 1b (mustn1b) gene in zebrafish
Abstract
Background: Mustn1 is a specific musculoskeletal protein that plays a critical role in myogenesis and chondrogenesis in vertebrates. Whole-mount in situ hybridization revealed that mustn1b mRNAs are specifically expressed in skeletal and cardiac muscles in Zebrafish embryos. However, the precise function and the regulatory elements required for its muscle-specific expression are largely unknown. Results: The purpose of this study was to explore and uncover the target genomic regions that regulate mustn1b gene expression by in vivo functional characterization of the mustn1b promoter. We report here stable expression analyses of eGFP from fluorescent transgenic reporter Zebrafish line containing a 0.8kb_mustn1b-Tol2-eGFP construct. eGFP expression was specifically found in the skeletal and cardiac muscle tissues. We show that reporter Zebrafish lines generated replicate the endogenous mustn1b expression pattern in early Zebrafish embryos. Specific site directed-mutagenesis analysis revealed that promoter activity resides in two annotated genomic regulatory regions, each one corresponding to a specific functional transcription factor binding site. Conclusions: Our data indicate that mustn1b is specifically expressed in skeletal and cardiac muscle tissues and its muscle specificity is controlled by the 0.2-kb promoter and flanking sequences and in vivo regulated by the action of two sequence-specific families of transcription factors. Developmental Dynamics 246:992–1000, 2017. © 2017 Wiley Periodicals, Inc.
Introduction
MUSTN1, musculoskeletal embryonic nuclear protein 1 (previously known as Mustang, musculoskeletal temporally activated novel gene), is a small nuclear protein linked to the skeletal muscle system in vertebrates. It was first identified studying the transcriptional profile of the bone repair process on a rat fracture model, where Mustn1 was strongly up-regulated in osteoprogenitor cells of periosteum, proliferating chondrocytes, and osteoblasts during early stages of the healing (Hadjiargyrou et al., 2002). Subsequently, expression studies revealed that Mustn1 is specifically expressed in mesenchymal tissue and areas of cartilage, bone, and muscle formation in mouse embryos (Gersch et al., 2012; Gersch and Hadjiargyrou, 2009; Lombardo et al., 2004), as well as in skeletal and cardiac muscle (Li et al., 2013) and tendon in adult (Lombardo et al., 2004).
Functional analyses have shown a key role of MUSTN1 in both cartilage and muscle development in mammals. Silencing of Mustn1 via RNA interference (RNAi) affected the in vitro proliferation and differentiation of chondrocytes, and decreased the expression of chondrogenic gene markers such as Sox9, Col II, and Col X in mouse embryos (Gersch and Hadjiargyrou, 2009). Furthermore, a critical role of Mustn1 has also been reported during myogenic differentiation and myofusion. Mustn1-silenced myoblasts were unable to differentiate in order to form multinucleated myotubes and displayed a decrease in the expression of known myogenic markers (MyoD, Myog, Des) (Liu et al., 2010). Thus, it has been hypothesized that Mustn1 is an early marker of myogenic differentiation and regeneration, as Mustn1:eGFP expression was found in newly activated satellite cells of transgenic mice (Krause et al., 2013).
Very few studies exist on the non-mammalian mustn1. It has been shown in amphibians that mustn1 knockdown causes craniofacial cartilage malformations as well as reduced expression of sox9 in Xenopus embryos, supporting its importance for in vivo chondrogenesis as well (Gersch et al., 2012). In fish, the embryonic and larval developmental expression of mustn1 co-orthologs (i.e., mustn1a and mustn1b) has recently been reported (Camarata et al., 2016). Its expression was detected in the Kupffer's vesicle, paraxial mesoderm, and otic vesicle, however expression in heart and musculoskeletal tissues was restricted to Zebrafish mustn1b (Camarata et al., 2016). Furthermore, loss-of-function studies using Morpholinos showed the involvement of Zebrafish mustn1a in the cilia formation (Choksi et al., 2014). However, the precise function and the regulatory elements required for its expression are largely unknown. Thus, comprehensive information on the molecular mechanisms regulating mustn1 gene activity is still lacking, and a promoter study has not yet been performed in non-mammalian vertebrates. Moreover, MUSTN1 displays a muscle and skeletal tissue-specific expression very conserved across vertebrates, and, therefore, observations in non-mammalian vertebrates might reveal key functions of MUSTN1 and its regulatory mechanisms.
In this work, we describe the generation of Zebrafish transgenic (Tg) lines, which harbor Zebrafish mustn1b gene promoter modified to drive enhanced green fluorescent protein (eGFP) expression. The resulting transgenic animals showed eGFP expression in a pattern that mirrors that of endogenous mustn1b gene expression in embryos. By using specific site directed-mutagenesis analysis, we also functionally characterized the mustn1b promoter activity and identified the potential transcriptional factors regulating its in vivo expression in skeletal and cardiac muscle tissues. Additionally, we should mention that our work exemplifies that the transgenic Zebrafish is advantageous for identifying the conserved regulatory elements required for normal mustn1b expression within embryos and adult fish.
Results
Analysis of mustn1b Genomic Region
We first characterized the mustn1b promoter and 5′-untranslated region (UTR). Results from 5′RACE revealed that the 5′-UTR sequences comprise 24 bp upstream from the ATG start codon (Fig. 1A,B). Sequence analysis of the 3245 bp promoter region identified a TATA box at 104 bp upstream from translation start site (Fig. 1A).
A: Schematic diagram of the Zebrafish mustn1b locus. It shows a three-exon gene structure and 3221 bp promoter sequence upstream of transcription start site denoted by +1. Boxes depict exons with white regions indicating 5'-UTR and 3'-UTR. B: The first exon comprises the 5'-UTR of 24 bp and 9 bp of the coding sequence.
The coding sequence of mustn1b (GenBank accession number: NM_001197053.2) was isolated and cloned from Zebrafish cDNA, resulting in an open reading frame (ORF) of 237 bp. Comparison with the in silico genomic sequence revealed a gene structure of three exons and two introns, where exon 1 encodes the 5′-UTR region plus 9 bp of the coding sequence (Fig. 1B).
Muscle Specificity of mustn1b Promoter
To determine the tissue specificity of mustn1b promoter and 5′ flanking region at embryonic and larval stages, we cloned 3245, 1764, and 760 bp DNA fragments upstream of mustn1b ATG start site, including the 5′-UTR, into Tol2 vector (3.2kb_mustn1b-Tol2-eGFP, construct P1; 1.8kb_mustn1b-Tol2-eGFP, construct P2; 0.8kb_mustn1b-Tol2-eGFP, P3). The constructs were microinjected into Zebrafish embryos for transient expression analysis. eGFP expression in injected embryos was monitored by direct observation under a fluorescence stereoscope. Promoter activities of P1, P2, and P3 reporter gene constructs are shown in Figure 2. All injected embryos exhibited eGFP expression in skeletal and cardiac muscle at 2 days post fertilization (dpf) (Fig. 2B,C,D). Conversely, almost no expression was detected in non-muscle tissues. Observed eGFP expression from transient expression of P1, P2, and P3 constructs was equivalent, indicating that the 0.8kb promoter and flanking sequences contain all of the regulatory elements required for normal mustn1b expression. However, the transgenic expression was mosaic, as it was observed only in some muscle fibers.
Transient expression analysis of mustn1b:eGFP in 2-dpf Zebrafish embryos. A: Schematic diagram of the Zebrafish eGFP reporter gene construct containing a mustn1b promoter region of 3.2 kb (P1), 1.8 kb (P2), 0.8 kb (P3), and 5′-flanking sequences. B–D: eGFP expression in skeletal and cardiac muscles of Zebrafish embryos injected with mustn1b:eGFP constructs that contain the 3.2 kb (P1) (B), 1.8 kb (P2) (C), or 0.8 kb (P3) (D) mustn1b gene promoter and 5′-flanking sequences, respectively. Scale bars = 200 μm.
Generation and Expression Analysis of Transgenic Reporter Zebrafish
Having shown that eGFP was just transiently expressed in skeletal and cardiac muscles, we generated stable transgenic Zebrafish lines in order to characterize the conserved muscle-specific mustn1b expression found in Zebrafish (Fig. 3). We created eGFP reporter transgenic lines, containing the 0.8kb mustn1b promoter and flanking sequences (Tg(mustn1b:eGFP)iim01). In the stable transgenic lines, the eGFP expression was first detectable in 30-hours post fertilization (hpf) embryos, when the stripes of muscle pioneer cells adjacent to the notochord and the somites were the initial site of eGFP expression (Fig. 3A1–A4). At 2 dpf, transgenic embryos showed eGFP expression in skeletal muscle fibers in the trunk and cardiac muscle, which was visualized from 2 dpf onward (Fig. 3C–F). From 3 dpf, eGFP expression was also detected in craniofacial muscles and pectoral fins (Fig. 3E,G), unlike the previous transient analysis, where no expression was observed. To determine whether the endogenous mustn1b gene was specifically expressed in the same domains, in situ hybridization was performed with a mustn1b antisense probe in Zebrafish embryos. Results showed that mustn1b mRNA was specifically expressed in muscle pioneer cells and somatic mesoderm during early embryonic development (Fig. 3B1–B4), as well as in cardiac and craniofacial muscles in later stages (data no shown), therefore showing that mustn1b mRNA was essentially expressed in almost all the same domains as in stable eGFP expression studies.
Whole-mount confocal imaging and in situ hybridization of the stable Tg(mustn1b:eGFP)iim01 line embryos and larvae. A: Expression of eGFP reporter gene in muscle pioneer cells and somites in 30-hpf transgenic embryos. The white arrowhead indicates the strong eGFP expression in a marked stripe of muscle pioneer cells adjacent to the notochord. A1–A4: Confocal imaging of cross-sections from 30-hpf transgenic embryo. B: Whole-mount in situ hybridization showing expression of mustn1b mRNA in muscle pioneer cells and somites in 30-hpf embryos. The black arrowhead indicates mustn1b gene expression in the stripe of muscle pioneer cells adjacent to the notochord. B1–B4: Cross-sections from 30-hpf embryo with in situ hybridization staining. C–F: eGFP expression in mononucleated skeletal and heart muscles of transgenic Zebrafish at 2, 4, and 7 dpf. E,G: eGFP expression in craniofacial and fin muscles of transgenic larvae at 4 and 7 dpf. e, eye; am, adductor mandibularis; hh, hyohyal; ih, interhyal; ima, intermandibularis anterior; imp, intermandibularis posterior; io, inferior oblique; ir, inferior rectus; sh, sternohyoideus; lap, levator arcus palatine; ah, adductor hyoideus; ao, adductor operculae; do, dilator operculae; lo, levator operculae; bl5, 5th branchial levator; bl1-4, 1st–4th branquial levators; tv, transversus ventralis; pfm, pectoral fin muscles. Scale bars A,B,C,E = 200 μm. Scale bars A1–A4 = 50 μm. Scale bars B1–B4 = 25 μm. Scale bars D,F,G = 100 μm.
We next investigated the eGFP expression in adult Tg(mustn1b:eGFP)iim01 Zebrafish. Transverse cuts across the fish body section revealed that eGFP is also expressed in a muscle-specific manner in head and trunk at 180 dpf (Fig. 4). Expression of eGFP showed a strong signal in cranial, jaw, cardiac muscles, and the tongue (Fig. 4A,B). Strong eGFP expression was also detected in the esophagus muscles (Fig. 4C,D) and the external muscle layer of the intestine (Fig. 4E). A relatively weak eGFP expression was also observed in the supracarinalis anterior (Fig. 4C–E) and lateral superficialis (Fig. 4D–F) of the epaxial muscles, as well as in the hypaxial muscles (Fig. 4D–E).
Expression of eGFP reporter gene in adult Tg(mustn1b:eGFP)iim01 Zebrafish. Transverse body cryosections were analyzed for eGFP fluorescence. A,B: A marked eGFP expression is detected in jaw, cranial muscles, tongue, and heart. C–E: eGFP expression is also found in esophagus and the external muscle layer of intestine. D–F: Trunk transverse sections show slight eGFP expression in supracarinalis anterior, lateralis superficialis, and hypaxial muscles. e, eye; am, adductor mandibularis; ih, interhyal; cm, cranial muscles; hh, hyohyal; t, tongue; h, heart; es, esophagus; sca, supracarinalis anterior; ls, lateralis superficialis; hyp, hypaxial muscles; int, intestine. Scale bar = 400 μm.
Slow- and Fast-twitch Myofibers of Skeletal Muscles Express mustn1b
To further characterize the eGFP expression in the skeletal muscles of the transgenic fish trunk, we determined the muscle fiber type identity. The mustn1b:eGFP transgenic line was crossed with the Myomesin-3-RFP line to generate double transgenic embryos. The Myomesin-3-RFP Zebrafish line was generated by an enhancer trap that expresses a Myomesin-3-RFP fusion protein specifically in slow muscles (Clark et al., 2011). In addition, Myomesin-3-RFP fusion protein is localized specifically at the M-lines of sarcomeres in slow muscle fibers (Xu et al., 2012). The mustn1b:eGFP; Myomesin-3-RFP double transgenic embryos were analyzed by confocal microscopy at 2 and 6 dpf. The results revealed that mustn1b:eGFP is expressed in both slow and fast myofibers (Fig. 5). In slow myofiber, mustn1b:eGFP expression co-localized with Myomesin-3-RFP-positive slow fibers at the superficial layer (Fig. 5A–C). Compared with day 2, mustn1b:eGFP expression in some of the slow fibers was missing or significantly reduced in embryos at 6 dpf (Fig. 5A,D). This is in contrast to the pattern of Myomesin-3-RFP expression in all slow fibers (Fig. 5E). In addition to slow muscles, confocal analysis revealed mustn1b:eGFP-positive fibers in the deeper myotome, where fast muscle fibers are located (Fig. 5D,H). However, the number of mustn1b:eGFP-positive fast fibers was significantly reduced in 6-dpf fish embryos, and not every fast fiber expressed the mustn1b:eGFP reporter gene (Fig. 5H). Strikingly, we noted a sarcomeric localization of mustn1b:eGFP in myofibers (Fig. 5A,a). Close examination revealed that eGFP bands were broader than the Myomesin-3-RFP band localized at the M-line (Fig. 5a,b). Myomesin-3-RFP appeared to be localized in the center of mustn1b:eGFP bands (Fig. 5a–c). This pattern of eGFP localization resembled that of thick filaments in the sarcomere.
Expression of eGFP reporter gene in slow and fast muscle fibers of Tg(mustn1b:eGFP)iim01 embryos and larvae. eGFP expression was analyzed in the mustn1b:eGFP; Myomesin-3-RFP double transgenic fish embryos at 2 and 6 dpf. A–C: mustn1b:eGFP expression co-localized with Myomesin-3-RFP-positive slow muscles at 2 dpf. In the enlarged pictures (a–c), eGFP protein showed a sarcomeric localization in the myofibers. D–F: mustn1b:eGFP expression in Myomesin-3-RFP-positive slow muscles at 6 dpf. G,H; Localization of eGFP expression in fast muscle fibers of transgenic Zebrafish at 2 and 6 dpf, respectively. Scale bars A, D, G = 30 μm. Scale bars a = 6 μm.
Functional Analysis of Zebrafish mustn1b Promoter
To identify the key regulatory region within the P3 construct sequence responsible for mustn1b expression, third and fourth constructs containing 0.4 kb (P4) and 0.2 kb (P5) mustn1b promoter and flanking sequences were also generated. The promoter activities of P3, P4, and P5 gene constructs are shown in Figure 6. Observed eGFP expression from transient expression of P3, P4, and P5 constructs were equivalent (Fig. 6A), indicating that the 0.2 kb promoter and flanking sequences contain all of the potential regulatory elements required for normal mustn1b expression. In the transient expression studies of the reporter gene constructs, eGFP expression was not detected in craniofacial muscles and pectoral fins (Fig. 6B). eGFP expression was not detectable in embryos injected with the vector eGFP backbone construct. TRANSFAC analysis of the 211 bp promoter fragment revealed consensus sequences for potential transcription factor binding sites for myogenic differentiation 1 (MYOD) and myocyte-specific enhancer binding factor (MEF2) (Fig. 6C). To study the possible role of predicted transcription factors involved in mustn1b expression in the musculoskeletal system, we performed the targeted mutagenesis to disrupt these consensus-binding sequences. We first designed a mutation targeting MYOD core sequence (Region 1), and the 97% injected Zebrafish embryos (n = 117) did not show eGFP transient expression in the skeletal muscle fibers along the trunk; however, eGFP expression was observed in the cardiac muscle (Fig. 6A–C). Subsequently, we performed a second mutation (Region 2), which targeted consensus-binding sequence for MEF2 (Fig. 6C). Observed eGFP expression from transient expression of Region 2 mutated construct was observed only in the trunk skeletal muscle (data not shown). We finally performed a combined double mutation targeting Regions 1 + 2, which resulted in a complete loss of eGFP expression in injected embryos (n = 217) (Fig. 6A,B). These data indicate that the transcriptional activity of mustn1b promoter is independently modulated by different cis-regulatory factors in skeletal and cardiac muscle. However, additional regulatory sequences might also be involved in regulating mustn1b gene expression in craniofacial muscles and pectoral fins.
Functional analysis of the mustn1b promoter through eGFP expression. A: DNA fragments containing 0.8 (P3), 0.4 (P4), and 0.2 kb (P5) sequences upstream from the ATG of mustn1b are in gray. Black circles mark conserved regions selected for mutations. Region 1, positions -138 -141 (transcription factor binding site [TFBS]: MYOD) and Region 2, positions -86 -89 (TFBS: MEF2). Arrow and hatched section denote the start of transcription and 5'UTR, with eGFP coding sequences substituting at the native ATG start site. Promoter activity was assessed as the proportion of embryos displaying one or more fluorescent mustn1b cells and normalized to the 0.8kb-mustn1b promoter results. B: Transient expression analysis of 3-dpf Zebrafish embryos injected with 211bp_mustn1b-Tol2-eGFP, Region 1, and Region1 + 2 site-directed mutagenesis constructs. The white arrowheads indicate the eGFP expression in skeletal muscles and heart. Counterstaining with DAPI was used to visualize nuclei (blue). C: Nucleotide sequence of the Zebrafish mustn1b gene showing the distribution in the promoter region of targeted transcription-factor binding boxes for site-directed mutagenesis. Nucleotides are shown in uppercase letters and numbered on the left of the sequence. The transcription start site (TSS) in the 5'-UTR is denoted by (+1). and the coding region, in italics, shows the ATG start site denoted by (+25). Transcription-factor binding regions of interest in white box and mutated nucleotides in red. The TATA box is marked by black overhead line. Scale bars = 250 μm.
Discussion
Mustn1 is a musculoskeletal specific protein involved in cellular proliferation and differentiation processes during vertebrate chondrogenesis, myogenesis, and regeneration. Some studies have investigated the role of Mustn1 in mammals; however, very few studies exist on non-mammalian mustn1. Therefore, the precise function and the regulatory elements required for its expression are largely unknown.
Combining both transient and transgenic techniques in Zebrafish, we report here the first in vivo functional analysis of mustn1b promoter in a non-mammalian species and illustrate that the transgenic Zebrafish is advantageous for identifying the conserved regulatory elements required for normal mustn1b expression within embryos and adult fish. We describe the generation of eGFP reporter transgenic Zebrafish lines, which harbor Zebrafish mustn1b promoter and flanking sequences. The resulting transgenic lines showed eGFP expression in a pattern that mirrors that of endogenous mustn1b gene expression in Zebrafish embryos.
The genomic structure of the mustn1b locus in Zebrafish appeared to be highly conserved to that in mammals (Liu and Hadjiargyrou, 2006). The structure consists of three exons separated by two introns and a 5′-UTR region belonging to exon 1. Upstream from the ATG start site, a TATA box and consensus sequences for potential transcription factor binding were identified and the minimal promoter region determined.
Transient expression profiles for several different-sized mustn1b gene promoter constructs were determined in Zebrafish embryos following microinjection. The constructs analyzed consisted of the eGFP reporter gene driven by three different lengths of mustn1b promoter regions. For all mustn1b-eGFP constructs analyzed, expression of eGFP was strictly muscle-specific during early embryonic development. To confirm conclusions from transient assays, we generated stable reporter transgenic Zebrafish lines, containing 0.8kb mustn1b promoter and flanking sequences (Tg(mustn1b:eGFP)iim01). In the Tg(mustn1b:eGFP)iim01 Zebrafish line, the eGFP reporter gene expression was restricted to somatic mesoderm, skeletal, and cardiac muscle throughout Zebrafish development, which correspond to the Mustn1b expression pattern found in other vertebrates (Gersch et al., 2012; Gersch and Hadjiargyrou, 2009; Li et al., 2013). Muscle pioneer cells flanking the notochord and somites showed a strong eGFP expression at 30 hpf. Subsequently, expression of eGFP was found in cardiac and in skeletal muscle fibers. At 3 dpf, eGFP expression was also found in craniofacial muscles and pectoral fins. Therefore, transgenic reporter lines expressing eGFP under the Zebrafish 0.8 kb mustn1b promoter and flanking successfully demonstrate the recapitulation of developmental expression of mustn1b gene (Camarata et al., 2016). The temporal activation of the mustn1b promoter is also consistent with the timing of the activation of endogenous mustn1b gene during normal development.
Using the stable eGFP-expressing Zebrafish transgenic line, we were able to determine the fiber-type specific expression of mustn1b gene. Our data show that mustn1b is expressed in cells of both slow- and fast-twitch pathway initiated at an early stage of fiber differentiation. However, mustn1b-eGFP expression remained strong in slow muscles but reduced significantly in fast muscles as embryos develop. This is also consistent with the stronger eGFP signal in slow muscles than in fast muscles in adult mustn1b-eGFP transgenic fish. Interestingly, we noted that mustn1b-eGFP was not expressed in all slow or fast myofibers of the transgenic fish embryos when analyzed at day 6. This differs dramatically with the Myomesin-3-RFP expression in all slow fibers. The reason for the mosaic pattern of mustn1b-eGFP expression in only some of the slow myofiber is not clear. Given that this expression analysis was performed on the stable mustn1b-eGFP transgenic line at the F3 and F4 generations, it cannot be due to the mosaic pattern of DNA integration. It is possible that mustn1b-eGFP is expressed only in a subset of muscle cells.
Strikingly, our data showed a sarcomeric localization of eGFP in myofibers of mustn1b:eGFP transgenic fish embryos. The eGFP localization resembled the thick filaments. Given that mustn1b:eGFP transgenic fish expressed eGFP as reporter, not a fusion protein, the reason behind the sarcomeric localization of eGFP protein in myofibers is not clear. Our observation of eGFP sarcomere localization would highlight cautions that should be taken in using eGFP for subcellular protein localization studies in muscle fibers.
To understand the molecular mechanisms that regulate mustn1b gene expression during the myogenesis, we functionally characterized the mustn1b gene promoter activity in vivo. Promoter dissection characterization identified a region -187/+24 nt that is critical for mustn1b expression. Results of TRANSFAC sequence prediction, promoter bashing approach, and directed mutagenesis uncovered two potential regulatory elements of the mustn1b expression. MEF2 factor binding site appeared to be critical for mustn1b expression in cardiac muscle, while MYOD factor binding site was crucial for mustn1b expression in skeletal muscles.
In vertebrates, skeletal muscles forming trunk derive from somatic paraxial mesoderm, unlike the head and cardiac muscles, which come from cranial paraxial mesoderm and splanchnic mesoderm (Tzahor, 2015). Thus, the genetic programs that promote distinct muscle types are supposed to be led by lineage-specific transcription factors. In vertebrates, myocyte enhancer factor-2 (MEF2) family of transcription factors has been shown to regulate early heart development (Breitbart et al., 1993). MEF2C in mouse and two MEF2 paralogs in Zebrafish (i.e., mef2ca and mef2cb) control cardiomyocyte differentiation and heart formation (Hinits et al., 2012; Verzi et al., 2005). MyoD is a known early myogenic marker and it is first expressed in Zebrafish adaxial cells to promote the differentiation into the slow muscle lineage (Coutelle et al., 2001; Jackson and Ingham, 2013). Subsequently, a second wave of myoD expression takes place in non-adaxial cells in the somites for fast muscles commitment (Jackson and Ingham, 2013; Weinberg et al., 1996). Previous studies reported that myoD Zebrafish mutants exhibit only a delay in the slow muscle myogenesis (Hinits et al., 2011), but double mutants myod/myf5 cause a complete inhibition in the skeletal muscle differentiation, which implies a certain cooperation between both myogenic regulatory factors (Coutelle et al., 2001; Hinits et al., 2011). On the other hand, Mustn1 was reported to act upstream of MyoD during myoblast differentiation and myofusion in cell culture (Liu et al., 2010), as well as during skeletal muscle regeneration in mouse (Krause et al., 2013). Our results suggest that MyoD would transcriptionally activate the mustn1b gene expression in Zebrafish adaxial and non-adaxial cells to promote the cellular differentiation into both slow and fast muscles.
Although the endogenous mustn1b gene was also expressed in the craniofacial muscles and pectoral fins, transient expression of the reporter gene constructs was not detected in craniofacial muscles and pectoral fins. So, the transient eGFP expression was mosaic and variable among the embryos injected with the same reporter gene construct. This event is mainly attributable to the differential segregation of the injected DNA during embryogenesis (Westerfield et al., 1992). Despite that, the transient reporter gene expression remains an effective and reliable system for identifying the conserved regulatory elements required for normal gene expression.
In summary, our findings show that mustn1b has a distinctive expression pattern restricted to muscle tissues during Zebrafish embryonic development, as well as in adult fish. Thus, mustn1b could be involved in early myogenic differentiation and also in maintenance and repair of muscle tissue. Functional characterization of mustn1b promoter sequence revealed that mustn1b is involved in independent molecular signaling pathways for cardiac and skeletal muscle differentiation during early development. Interestingly, Myod and Mef2 appeared to be crucial myogenic transcriptional factors for skeletal and cardiac muscle differentiation acting upstream of mustn1b. Additionally, our results suggest that the mustn1 promoter activities could be remarkably conserved from fish to mammals.
Experimental Procedures
Zebrafish
Fish were maintained as previously described (Westerfield, 2007) and staged following stablished criteria (Kimmel et al., 1995) by hours post fertilization or days post fertilization. Experiments were performed with a TU-WT strain (Tuebingen [TU], Nüsslein-Volhard Lab). Embryo medium was supplemented with 0.003% (w/v) 1-phenyl-2-thiourea to inhibit pigmentation. Ethical approval (Ref: AGL2014-52473R) for all studies was obtained from the Institutional Animal Care and Use Committee of the IIM-CSIC Institute in accordance with the National Advisory Committee for Laboratory Animal Research Guidelines licensed by the Spanish Authority (RD53/2013) and conformed to European animal directive (2010/63/UE) for the protection of experimental animals.
Determination of Transcription Start Site by 5′ RACE
To identify the transcription start site of mustn1b gene, we performed 5′ RACE reactions using SMART RACE cDNA amplification kit (Clontech) following manufacturer's instructions. 5′RACE-ready cDNA was used as a template for polymerase chain reaction (PCR) amplification with the universal primer mix (according to the protocol) and 5′ gene-specific primer (5′- CTTTTTCTTAACCTCCGGCTGTGACAT-3′). PCR product was subcloned into pGEM-T Easy Vector (Promega) and sequenced.
DNA Constructs
Promoter fragments spanning 3245, 1764, 760, 421, and 211 bp (GenBank accession number: BX548064.12) upstream of the translation start site, including 5′UTR region, were amplified by PCR using primers listed in Table 1. Six-dpf genomic DNA was used for PCR amplification using PfuUltra II Fusion HS DNA Polymerase (Agilent) for longer promoter sequences than 1 kb, or DreamTaq DNA Polymerase (Thermo Scientific) for those below 1 kb, according to the manufacturer's instructions. Amplified fragments were cloned into pSpark I vector (Canvax) or pGEM-T Easy Vector (Promega), depending on type of polymerase used in the previous amplification step, then subcloned into restricted Tol2 transposon-based vector (pT2AL200R150G), provided by Kochi Kawakami (Kawakami, 2007). Tol2 reporter constructs generated were as follows: 3.2kb_mustn1b-Tol2-eGFP (construct P1), 1.8kb_mustn1b-Tol2-eGFP (construct P2), 0.8kb_mustn1b-Tol2-eGFP (construct P3), 0.4kb_mustn1b-Tol2-eGFP (construct P4), and 0.2kb_mustn1b-Tol2-eGFP (construct P5). Cis-acting transcription factor binding sites (TFBSs) located in the 211 bp mustn1b promoter sequence were identified using MatInspector software (Cartharius et al., 2005). Substitution mutations were generated as described by Wang and Wilkinson (Wang and Wilkinson, 2001) and using the QuikChange Site-Directed Mutagenesis Kit (Stratagene). Resulting genomic clones were verified by sequencing.
| Construct | Forward primer sequence (5'-3') | Reverse primer sequence (5'-3') |
|---|---|---|
| 3.2kb-mustn1b | ctcgagCGGCAGGTAAACTGAAGCTC | ctgcagTTTGGCTGAGTTGACTCCTTG |
| 1.8kb-mustn1b | ctcgagACCGCGGTATACCTTGAAAAC | aagcttTTTGGCTGAGTTGACTCCTTG |
| 0.8kb-mustn1b | ctcgagATCTTGAATGTTTTTCAATG | aagcttTTTGGCTGAGTTGACTCCTTG |
| 0.4kb-mustn1b | ctcgagCCCCGTCATTATGCACATTAG | aagcttTTTGGCTGAGTTGACTCCTTG |
| 0.2kb-mustn1b | ctcgagGCATGGGTGAATCAGGAGTAG | aagcttTTTGGCTGAGTTGACTCCTTG |
| Region 1 | GTAATGTGTTCAGAGGCTCATAGTGGTCATG | CATGACCACTATGAGCCTCTGAACACATTAC |
| Region 2 | CTGACAGCAGAGCGCGTAATATATAATGATAAGG | CCTTATCATTATATATTACGCGCTCTGCTGTCAG |
- a Restriction sites are denoted in lowercase.
Transgenic Zebrafish Lines
eGFP reporter gene Tol2 constructs containing different lengths of mustn1b promoter and specific point mutations were dissolved in distilled RNAse-free water to a final concentration of 50 μg/ml. A total of 250 pg of construct and synthetic 5’-capped mRNA (150 pg) encoding a transposase were co-injected into TU-WT embryos at the one- or two-cell stage, with 1% of phenol red as tracer. Microinjection was carried out under a dissection microscope (MZ8, Leica) fitted with a MPPI-2 pressure injector (ASI Systems). eGFP transient expression was analyzed by direct observation under a fluorescent Leica M165FC stereoscope. The number of embryos showing eGFP fluorescence was determined, and the Tol2 vector constructs were compared to score activity and tissue specificity.
Stable expression transgenic Zebrafish lines with the 0.8bp_mustn1b-Tol2-eGFP construct (P3) were generated Tg(mustn1b:eGFP)iim01. Founder fish (F0) were mated and reporter-expressing F1 offspring was selected, grown until adulthood, and mated. F2 progeny were raised to sexual maturity and crossed to TU-WT. F3 progeny were analyzed for the transgene to identify homozygous F2 individuals. Homozygous F2s were mated to generate a stable homozygous population without loss of the integrated transgene (F3).
Double mustn1b:eGFP and myomesin-3:RFP transgenic Zebrafish lines were crossed, and double-positive embryos were directly analyzed for localization of green and red markers.
Whole-mount in situ Hybridization
Whole-mount in situ hybridization was performed using digoxigenin-labeled antisense riboprobes as previously described (Rotllant et al., 2008). First, mustn1b coding sequence (GenBank accession number: NM_001197053.2) was isolated from Zebrafish-pooled cDNA with primers (5′- ggatccAGCCAAAATGTCACAGCTGGAG-3′ and 5′- ctcgagTCATTTCCCAAACACACTTCCAG-3) and cloned into pGEM-T Easy Vector (Promega). Antisense and sense riboprobes were made from linearized vector. Fish (n = 20) were fixed in 4% paraformaldehyde (PFA) in 1X PBS overnight at 4 °C and stored in 100% methanol until hybridization. Embryos were fixed in 4% PFA in 1X PBS at 4 °C, then dehydrated and embedded in Paraplast (Sherwood, St Louis, MO). Serial 12 μm transverse anatomical sections were cut using a rotary microtome. Sections were mounted with Eukitt on (3-aminopropyl)triethoxylane-treated slides and air-dried at room temperature for imaging.
Cryostat Sectioning and eGFP Observation
For cryostat sectioning, Tg(mustn1b:eGFP)iim01 adult fish (n = 2) were fixed in 4% paraformaldehyde overnight at 4 °C, washed with PBS, transferred to 15% sucrose, followed by 30% sucrose, then embedded and frozen in optimal cutting temperature medium. Cross-cryosections of 25 μm thickness were collected on poly-l-lysine slides (Thermo Fisher Scientific) and allowed to dry. Slides were washed in 1X PBS, mounted in Vectashield, and imaged on a fluorescent Leica M165FC stereoscope for transgenic adult fish sections. Slides from transgenic embryos were mounted in Mowiol and imaged on a Leica TCS SP5 confocal microscope.
Whole Tg(mustn1b:eGFP)iim01 fish embryos (n = 50) were examined at 30 hpf and 2, 4, 6, and 7 dpf. eGFP expression was analyzed by direct observation under a fluorescent Leica M165FC stereoscope, or specimens we fixed in 4% paraformaldehyde overnight at 4 °C and imaged on a Leica TCS SP5 confocal microscope.
Acknowledgments
We thank Professor N. Lawson and Professor K. Kawakami for providing the iTol2 constructs. We would also like to thank Inés Pazos (CACTI, University of Vigo) and Fátima Adrio (University of Santiago de Compostela) for their advice and assistance with the confocal microscope and cryosectioning. This work was funded by the Spanish Science and Innovation Ministry grants ALG2011-23581 and AGL2014-52473R (to J.R.). P. Suarez-Bregua was supported by a Campus do Mar PhD grant, Xunta de Galicia.
