|Home | About | Journals | Submit | Contact Us | Français|
Development and maturation of the nascent cardiovascular system requires the recruitment of mural cells (MCs) around the vascular tree in a process called vascular myogenesis. Understanding the origin and development of vascular MCs has been hampered by difficulties in observing these cells in vivo and performing defined genetic and experimental manipulations in available model organisms. Here, we investigate the origin of vascular MCs using molecular and genetic tools in zebrafish. We show that vascular MCs are present around the lateral dorsal aortae (LDA) and anterior mesenteric arteries (AMA) of developing animals, and that they also contribute to the outflow tract of the developing heart and ventral aorta (VA). Genetic data indicate that the vascular MCs of the LDA and AMA do not arise from blood or endothelial progenitors but from other derivatives of the lateral plate mesoderm. We further show that zebrafish vascular MCs share many of the morphological, molecular and functional characteristics of vascular smooth muscle cells and pericytes found in higher vertebrates. These data establish the zebrafish as a useful cellular and genetic model to study vascular myogenesis as well as tumor angiogenesis and other MC-associated diseases.
In vertebrates, the development and maturation of the cardiovascular system require the recruitment of smooth muscle cells in a process called vascular myogenesis (Carmeliet, 2000; Jain, 2003). During this process, smooth muscle cell precursors, called mural cells (MCs), differentiate and surround vascular endothelial vessels promoting the maturation and specialization of the cardiovascular system (Carmeliet, 2003). During development, these vascular MCs undergo proliferation and/or differentiation depending on the size of the vessels. On large vessels, they become multi-layered and referred to as vascular smooth muscle cells (SMCs). On smaller vessels, mural cells remain as single sparse cells and usually referred to as pericytes (PCs). Large arterial vessels have strong, elastic vessel walls with dense populations of concentrically-arranged SMCs to withstand the higher blood pressures. SMCs in these large vessels support and regulate blood flow. Veins are generally irregularly covered by SMCs and have valves to prevent backflow of blood (Scultetus et al., 2001). Small blood vessels, which are composed of endothelial cells (ECs) surrounded by a basal lamina, are loosely covered by single PCs, which serve in both scaffolding and signaling functions, and can also regulate vasoconstriction and vasodilation within the vascular bed. It is not known whether PCs and SMCs have distinct progenitors or whether they represent phenotypic variants of the same lineage (Bergers and Song, 2005; Majesky, 2007).
In mammals SMCs are characterized by a set of contractile proteins that are distinct from those expressed in skeletal and cardiac muscles (Owens, 1995). Several dynamic molecular markers present in mammalian vascular SMCs are used for their detection, including smooth muscle-myosin heavy chain (MYH11) (Babij et al., 1991; Frid et al., 1993; Madsen et al., 1998; Miano et al., 1994), smooth muscle alpha-actin (ACTA2) (Gabbiani et al., 1981; Hungerford et al., 1996; Mack and Owens, 1999), Sm22alpha (aka TRANSGELIN) (Li et al., 1996; Kim et al., 1997), Smoothelin (van der Loop et al., 1996; Kramer et al., 1999), Desmin and Calponin (Duband et al., 1993; Owens et al., 2004). The expression patterns of these markers can vary depending on the developmental stage of the vascular system. In particular, Sm22alpha and smooth muscle alpha-actin are the first markers expressed in SMCs and for this reason are considered embryonic SMC markers in mammals (Majesky, 2007).
Intriguingly, unlike skeletal or cardiac muscle cells which permanently exit the cell cycle and are terminally (and essentially irreversibly) differentiated, SMCs retain remarkable plasticity and can undergo rather profound and reversible phenotypic changes in response to changes in local environmental cues. The high degree of mammalian smooth muscle plasticity, especially during postnatal development, predisposes it to phenotypic switching, and thus may play a major role in human disease (Halayko and Solway, 2001). Unregulated growth, differentiation and plasticity of mammalian vascular SMCs is a key feature of vascular diseases such as atherosclerosis, restenosis and hypertension, as well as monogenic diseases such as telangiectasia and retinopathies (Ross, 1999; Conway et al., 2001; Sata et al., 2002; Hao et al., 2003). Several genes identified to play a key role in vascular myogenesis are also associated with human disease (Carmeliet, 2003; Rensen et al., 2007). Furthermore, in response to specific vascular injuries, vascular SMCs dramatically increase their rate of proliferation, migration and synthetic capacity and play a critical role in vascular repair and tumor angiogenesis (Gerhardt and Semb, 2008).
A limitation in studying mammalian vascular myogenesis is the lack of an easily manipulatable in vivo system to analyse smooth muscle development and differentiation. The zebrafish has already been established as a powerful model system to study cardiovascular development (Stainier, 2001; Lawson and Weinstein, 2002; Beis et al., 2005; Santoro et al., 2007). The zebrafish is also a genetically accessible vertebrate with an optically clear embryo permitting high-resolution in vivo cell imaging. Surprisingly, the existence of vascular SMCs or PCs has not yet been fully reported in this model organism. We carried out a series of morphological, molecular and functional studies to visualize and characterize vascular myogenesis in the developing zebrafish. Since we were able to detect only a few cells surrounding early embryonic vessels, we decided to name these perivascular cells, mural cells (MCs). Our studies show that zebrafish embryos develop vascular MCs that share many of the characteristic features of embryonic SMCs found in higher vertebrates, and we provide evidence to support a functional role for zebrafish MCs during cardiovascular development. These studies identify the zebrafish as an embryological and genetic model system to further study the development, differentiation and plasticity of vascular MCs in vivo.
Understanding the molecular mechanisms regulating vascular myogenesis remains a major issue, in part because of the challenges of studying this process in amniotes (Majesky, 2007). We therefore decided to investigate the origin of vascular MCs in zebrafish, particularly focusing on their contribution to the trunk vasculature (mainly dorsal aorta and cardinal vein) during development. We started by comparing the ultrastructure of the dorsal aorta (DA) and posterior cardinal vein (PCV) at various stages. Light microscopy images of 3-months old zebrafish revealed striking differences in vessel wall caliber between the DA and PCV, with the former clearly characterized by thick layers of vascular SMCs around the ECs (Figure 1)(Miano et al., 2006). Interestingly, the PCV did not show the same extent of vascular SMC coverage as in mammals. To better characterize these vascular SMCs we analysed samples of DA and PCV by transmission electron microscopy (TEM) at earlier time points. Analyses of 20 day-old DA revealed that at least two to three layers of synthetic vascular SMCs are present in the DA, while single and dispersed SMCs were found in the PCV (Supplementary Figure 1). Interestingly, a dense layer of extracellular matrix, the internal elastic lamina (IEL), lies between the EC and the SMC layers, constituting the basement membrane. The synthetic SMCs in the DA at 20 dpf did not yet exhibit the ultrastructural features of fully differentiated adult vascular SMCs. Differentiated, contractile vascular SMCs are elongated, whereas undifferentiated, synthetic SMCs have a cobblestone morphology and contain a high number of organelles involved in protein synthesis. Moreover, synthetic SMCs show higher growth rates and higher migratory activity than contractile SMCs (Hao et al., 2003). Collectively, these results reveal a basic ultrastructural organization of the DA in zebrafish that is similar to that in mammals with the notable exception of a delay in zebrafish SMC maturation.
The transcriptional program controlling smooth muscle differentiation is poorly understood. Unlike skeletal and cardiac muscle, SMCs do not terminally differentiate but retain remarkable plasticity and exhibit a diverse range of phenotypes under various physiological and pathological conditions (Halayko and Solway, 2001). Due to their heterogeneous origin and transcriptional profiles, no single marker can be used to identify all SMCs. However, when these markers are assayed in combination, and are considered in the morphological context of developing organs, they can be used to identify cells of the SMC lineage. A variety of genes and gene products have been identified in mammals and other vertebrates as useful markers of the relative state of induction and differentiation of SMCs (Owens, 1995; Owens et al., 2004). These markers include mainly contractile, cytoskeletal and cytoskeleton-associated proteins such as smooth muscle alpha-actin, Sm22alpha, smooth muscle myosin heavy chain, and Smoothelin. Similarly, in zebrafish, the heterogeneity of MCs precludes their identification by single markers (Georgijevic et al., 2007; Davis et al., 2008). Therefore, we identified and cloned a panel of zebrafish MC markers including the genes encoding Transgelin (Sm22alpha-b), Transgelin2 (Sm22alpha-a), Acta2, Myh11, Smoothelin, Calponin, and Desmin. Among those we found that acta2 and transgelin were expressed specifically in the perivascular region of zebrafish larvae (Figure 2). Starting at 72 hpf these genes were not only expressed in visceral SMCs (Georgijevic et al., 2007) and the fin bud, but also in a specific region of the vasculature corresponding to the anterior portion of the DA, the lateral DA (LDA) as well as the primitive internal carotid arteries (PICA). Cross-section analyses revealed specific staining around the dorsal aorta and in the mesenchyme surrounding the gut epithelium. At this time, expression was relatively weak and discontinuous throughout the length of the dorsal aorta. However, this expression became more evident at 96 hpf especially around the DA and LDA. At this later stage, acta2 and transgelin expression also became evident in the heart, specifically the outflow tract and ventral aorta (VA) suggesting the “induction” of MC-like tissue in this region (Figure 2g-j). Altogether these data support the existence of zebrafish vascular MCs in specific regions of the developing vasculature starting at 72hpf. Progressively, these MCs became more evident in these anatomical regions and are possibly involved in cardiovascular maturation.
To better investigate zebrafish vascular MC development at cellular resolution we decided to use antibody staining followed by confocal imaging. As previously reported by Georgijevic et al., 2007 the epitope for the commonly used mammalian ACTA2 antibody is not conserved in zebrafish and thus the mammalian antibody is not useful to distinguish between the different muscle types. Therefore, we developed a specific antibody that recognizes zebrafish Transgelin (aka Sm22alpha-b), one of the earliest mammalian vascular SMC markers (Zhang et al., 2001). We investigated the localization of MCs throughout the length of the animal at 80 hpf (Figure 3a-c). Cross sections at the level of the 2nd somite showed Transgelin expression around the dorsal aorta (DA) and the gut (G). At more posterior positions (10th somite), the expression of Transgelin was lost around the vasculature (DA and PCV) and maintained only around the visceral organs (G). No expression was detected at the level of the 18th somite. Higher magnification imaging of DA sections at the level of the 2nd somite showed that Transgelin-positive cells are individual cells located in close proximity to the ventral and lateral sides of the DA (Figure 3d). Transgelin-positive MCs lined the outside of the vascular tube and are clearly different from endothelial cells (ECs) (Figure 3e). Interestingly, the PCV was not yet covered by MCs at this stage of development. We also investigated the morphology of the Transgelin-positive MCs using ultra-thin cross-sections and TEM (Figure 4). Representative images of the DA at 80 and 120 hpf at the level of the 2nd somite show a vascular lumen (L) and the lining ECs surrounded by tightly apposed MCs. At this stage, no internal elastic lamina (IEL) was observed between the ECs and the adjacent MCs, though tight junctions between ECs were readily evident. To further characterize the anteroposterior coverage of these MCs we performed whole-mount staining on 80 hpf larvae with the Transgelin antibody. We found that the MCs appear to form a bulge of cells in the ventral portion of the DA that appears to migrate towards the anterior portion of the DA as well as posteriorly towards the mesenteric artery (Figure 5a-e and supplementary movies). Cross-sections of the bulge region revealed a number of MCs in the ventral and lateral portion of the DA (Figure 5d-e). These data correlate well with the acta2 and transgelin in situ hybridization data indicating that zebrafish vascular MCs may start to differentiate on the ventral side of the DA at the 2nd somite level around 72 hpf.
It has been previously shown that the adult zebrafish retina vasculature features PCs that are located between the vascular endothelium and the basal lamina of the retinal vessels (Alvarez et al., 2007). In situ hybridization analyses with acta2 and transgelin as well as immunostaining with the Transgelin antibody staining revealed the presence of putative MCs in the eyes of developing larvae as early as 72 hpf (Supplementary Figure 2). These cells appeared to originate from a specific region of the retinal vasculature called the inner optic circle (IOC) in close association with the optic vein (OV). As the eye vasculature further developed, these MCs continued to cover the IOC (data not shown).
In zebrafish, as in other vertebrates, the dorsal aorta develops in close association with the splanchnic mesoderm, which also contributes to the primitive vascular tube (Lawson and Weinstein, 2002). This observation suggests that the early MCs in the zebrafish aortic floor also arise from the splanchnic mesoderm. Alternatively, these MCs might derive from the ECs of the aortic floor through a process called endothelial transdifferentiation (DeRuiter et al., 1997). To address this point, we took advantage of two zebrafish mutants, cloche (clos5) and hands off (hans6). cloche mutant embryos lack all endothelial precursors in the head and trunk regions and do not form functional blood vessels (Stainier et al., 1995). The hands off gene encodes the bHLH transcription factor Hand2 and is expressed in the lateral plate mesoderm during gastrulation and somitogenesis stages, and hans6 mutants have defects in lateral plate mesoderm-derived tissues (Yelon et al., 2000). We tested for the presence of vascular MCs in these mutants using the Transgelin antibody (Figure 6). Interestingly, while clos5 mutants showed Transgelin-positive cells in the region corresponding to the aortic floor, hans6 mutants completely lacked these cells. These experiments suggest that in zebrafish, MCs derive from the lateral plate mesoderm and not from the transdifferentiation of ECs.
The zebrafish is a widely accepted model for the study of early cardiogenesis (Stainier, 2001; Beis et al., 2005). Much of this research has been concentrated on the development of the atrium and ventricle, at stages long before the outflow tract is formed. As a result, this structure has been poorly characterized. Several reports have shown that the adult zebrafish outflow tract is invested with a considerable number of smooth muscle layers (Hu et al., 2000; Grimes et al., 2006). We therefore decided to investigate the onset of zebrafish MC induction in the bulbus arteriosus (BA) and outflow tract. Whole mount in situ hybridization analyses showed that acta2 and transgelin are expressed in the heart region at 96 hpf (Figure 2). To gain a better resolution of these MC-positive cardiac regions, we analyzed thin cross sections of whole mount stained larvae at 96hpf (Figure 7a-b). These analyses confirmed that the region involved is the BA and the proximal part of the VA. We confirmed the presence of this MC coverage in the outflow tract by Transgelin staining (Fig. 7c-e). Immunofluorescence on whole mount and cross sections with Transgelin antibody clearly showed a single layer of vascular MCs around the VA at 96 hpf. Interestingly, at 72 hpf, while the BA was already formed and several layers of cuboidal cells were evident, the outflow tract did not yet express these MC markers (data not shown). These data are in agreement with a previous characterization of the zebrafish outflow tract where a cellular structure committed to become the BA appears after 48 hpf (Grimes et al., 2006), a stage when the atrium and ventricle are already formed and pumping blood through a rudimentary circulatory system. We speculate that between 72 and 96 hpf these cells are induced to differentiate into vascular MCs and contribute to the future outflow tract.
Mammalian vascular SMCs are a highly heterogeneous cell population with diverse embryonic origins, properties, and gene expression profiles. A major difficulty in the smooth muscle field has been the lack of an easily manipulatable model system for the study of SMC biology in vivo. Surprisingly, the development of vascular SMCs, in particular vaSMCs and PCs, has not yet been characterized in zebrafish, even though the transparency and easy manipulation of embryos has established it as an excellent model system to study the development of a variety of organ systems. Here we analyzed the development and distribution of zebrafish vascular MCs and investigated the temporal and spatial expression patterns of several MC genes during cardiovascular development. Additionally, we examined the morphology of zebrafish vascular MCs with different microscopy techniques and identified these cells in the larval and adult vasculature. The goal of this study is to begin to provide an understanding of the ontogeny of zebrafish vascular SMCs and PCs and the morphological events that lead to their induction, differentiation and maturation. This understanding will help investigate the cellular and molecular mechanisms that regulate their development and to evaluate the importance of SMC biology as a susceptibility factor for vascular diseases like atherosclerosis, retinopathies and hereditary haemorrhagic telangiectasia (HHT). The importance of developing a model system to study SMCs resides on the extensive contribution of these cells to various pathological conditions. The zebrafish represents an interesting new approach to connect developmental biology and associated diseases (Dodd et al., 2000; Lieschke and Currie, 2007; Chico et al., 2008). Our data suggest an evolutionarily conserved developmental origin and function of SMCs indicating that the zebrafish can be used to study the genetics of vascular SMC-associated human diseases. Generation of transgenic zebrafish lines with fluorescently labelled vascular MCs, as well as the identification and characterization of mutant lines showing defects in zebrafish vascular MC development will be the next steps. In fact, some zebrafish mutant phenotypes identified in genetic screens for developmental regulators resemble human disease states and these mutations, once isolated molecularly, will provide candidate genes for evaluation in human disorders (Lieschke and Currie, 2007). The identification of mutant phenotypes resembling human vascular diseases will be very useful to i) dissect new genetic pathways involved in SMC-associated diseases, ii) study the genetic origin of pathological conditions associated with MCs and iii) facilitate the set-up of new therapeutic applications for these pathologies.
The origin of vertebrate vascular SMCs is complex with contributions from several independent cell lineages (Gittenberger-de Groot et al., 1998; Majesky, 2007). Neural crest cells (NCs) give rise to vascular (Wilm et al., 2005) SMCs in the cephalic region and in the cardiac outflow tract region, and are required for patterning the great vessels (Hutson et al., 2003). The pro-epicardial organ contributes to SMCs in the coronary arteries and ventrally emigrating neural tube cells may contribute to SMCs in the great vessels and the coronary vessels in chicken (Mikawa and Gourdie, 1996; Gittenberger-de Groot et al., 1998). The mesothelium also gives rise to SMCs in mesenteric vessels in mouse (Wilm et al., 2005). In posterior parts of the body, vascular MCs are assumed to derive primarily from the splanchnic lateral plate mesoderm (Gittenberger-de Groot et al., 1999), but the precise source of these cells has not been rigorously determined. Observations of the early pattern of SMC contractile protein expression in mouse and quail aorta suggest that vascular MC differentiation is induced in mesenchymal cells that line the endothelium (Hungerford et al., 1996; Takahashi et al., 1996). Vascular SMC contractile proteins are first observed on the ventral side of the vessel, and it has been hypothesized that MCs are only induced in the ventral area of the aorta and later migrate to populate the lateral and dorsal areas. Our findings are in agreement with these findings and suggest that vascular MCs in the trunk originate from the lateral plate mesoderm as in other vertebrate model organisms (Wiegreffe et al., 2007; Wasteson et al., 2008). However, cell lineage experiments need to be carried out to determine the precise origin of MCs in different anatomical regions. Establishment of transgenic lines for zebrafish MCs should also be a useful tool to study the perivascular origin of mesenchymal stem cells and their potency in myogenesis, adipogenesis, chondrogenesis and tissue regeneration (Cossu and Bianco, 2003; Crisan et al., 2008).
We also noticed that from early developmental stages until adulthood, zebrafish arteries and veins are differently covered by SMCs. While the aorta is covered with vascular MCs from an early time point, the cardinal vein is not. Lack of MCs in the posterior cardinal vein of zebrafish larvae is not surprising given that only a few scattered SMCs were found in the walls of the caudal vein and the posterior cardinal vein of 20 dpf animals (Figure 1 and Supplementary Figure 1). This situation resembles that found in the carp (Cyprinus carpio)(Satchell, 1992). Gravity exerts significant effect on venous return in terrestrial animals, affecting venous pressure and ventricular filling pressure. Hence, venous compliance must be continuously adjusted in terrestrial animals by MC contractility to ensure proper cardiac filling. In contrast, gravitational stress in water is much less than in a terrestrial environment. The specific gravity of water is similar to that of blood, resulting in constant transmural pressure along the length of the posterior cardinal vein (Scultetus et al., 2001; Satchell, 1992). Therefore, modulation of venous blood pressure by smooth muscle contractility may not be crucial in juvenile zebrafish. On the other hand, we found an early vascular MC coverage in the anterior mesenteric artery of zebrafish larvae suggesting that this is a crucial vascular region to control blood flow towards the gastro-intestinal tract region at early time points.
While the use of atrioventricular-specific molecular markers has allowed extensive characterization of the development of the atrium and ventricle, the lack of any bulbus arteriosus (BA)-specific markers has meant that this region of the zebrafish heart is poorly characterized and quite possibly misunderstood (Hu et al., 2000; Grimes et al., 2006). In zebrafish, the BA forms by 48 hpf as an independent structure, at a stage when the atrium and ventricle are already formed and pumping blood through a rudimentary circulatory system. Because it forms so late, it is usually excluded from studies of the true cardiac chambers, and detailed analyses of its development need to be carried out. The function of the BA in fish is to reduce the high ventricular pressure during systole, and to maintain constant blood flow into the ventral aorta and gills during diastole. The elastic walls of the BA maintain a strong and even pressure on the blood flowing to the gills, maintaining a steady flow. The ability of the bulbous wall to expand and hold much of the stroke volume is critical to this process (Bushnell et al, 1992). Our finding of extensive smooth muscle coverage in the outflow tract and closely associated vessels is consistent with a role for the BA in maintaining an even blood pressure. It will be important to identify the progenitors of these outflow tract MCs as their biology gets further investigated (Sato and Yost, 2003).
Our results provide the first analysis of zebrafish vascular myogenesis. Collectively, these data provide new insights into the emergence of differentiated vascular MCs and PCs in zebrafish, thus providing a foundation for future studies on the gene expression profile and function of these cells in wild-type and mutant animals.
Zebrafish, Danio rerio, were maintained and staged based on developmental time and morphological criteria as previously described (Santoro et al., 2007). Fish were kept under a 14-h light and 10-h dark photoperiod at approximately 28°C. Following fertilization, eggs were collected and embryos- wild type or mutant- were raised at 28°C under standard laboratory conditions in the presence of 0.003% 1-phenyl-2-thiourea to prevent formation of melanin pigments, greatly facilitating visualization of the vascular system.
Wild-type and mutant larvae were collected at 72 or 96 hpf, fixed in 4% paraformaldehyde (PFA) in PBS overnight at 4°C, then transferred into 100% Methanol (MeOH) for at least 2h at -20°C. Whole-mount in situ hybridization was performed as previously described (Santoro et al., 2007), using Digoxigenin-labeled antisense RNA probes of acta2 and transgelin. For acta2, a 3′-UTR–specific probe was generated using the following primers: 5′-gtcccgtaattctgctattg-3′ and 5′-aattattgctgagctttatt-3′. For transgelin2, a specific probe was generated using the following primers: 5′-atggcaaataaaggtccgtc-3′ and 5′-gtcgtccgtaaccggtcatt-3′. For transgelin, a specific probe was generated using the following primers: 5′-actatggcaaacaaggggcc-3′ and 5′-acgcccctcttttggagccc-3′. To perform the in situ hybridization, the larvae were rehydrated using successive dilutions of methanol (75%, 50%, 25%) in PBS. Then the larvae were permeabilized by digestion with proteinase K (10 μg/ml) at room temperature for 30 minutes to allow the RNA probe to penetrate. Proteinase K digestion was stopped by incubation for 30 minutes in 4% (w/v) paraformaldehyde (PFA) in PBS. Hybridization took place overnight at 65°C in the presence of approximately 150 ng antisense DIG-labeled RNA probes. After washing, the hybridized larvae were incubated overnight in a 1:10,000 dilution of alkaline phosphatase-conjugated anti-DIG antibody (Roche) at 4°C. Color reactions (using nitroblue tetrazolium and 5-bromo-4-chloro-3-indolyl phosphate as substrates) were used to detect the hybridized mRNA transcripts. The larvae were then embedded in 17% Gelatin-PBS-Azide, fixed in 4% paraformaldehyde for at least 24 h at room temperature and cut with a VT1000S vibratome (Leica) into 50 μm sections. Processed samples were mounted in Vectashield (Vector Laboratories) and images were acquired using a Zeiss Axiocam.
Wild type animals were fixed in 2% glutaraldehyde in 0.1M sodium cacodylate buffer (pH 7.4) for two hours at room temperature or overnight at 4 °C. Semi-thin sections were cut at 0.5 or 1.0 micron and a metal loop was used to collect thick sections, and transfer them to a drop of distilled water on a glass slide. Sections were dried down on a glass slide by placing the slide on a slide warmer lamp. After the sections were completely dry, they were covered with a few drops of staining solution (1% Toluidine Blue and 2% borate in Distilled Water) for 1-2 minutes depending on the darkness of staining desired. Excess stain was gently rinsed off with distilled water. Slides were air dried and mounted. For TEM analysis, wild type animals were fixed in 2% glutaraldehyde in 0.1M sodium cacodylate buffer (pH 7.4) for two hours at room temperature or overnight at 4 °C. Samples were dissected in cold PBS, and fixed in 2% formaldehyde (Polysciences) and 2% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) for 2 hours at room temperature. Embryos were post-fixed with 1% osmium tetraoxide, followed by staining in 1% uranyl acetate in 70% ethanol, dehydrated through a graded series of ethanol washes, followed by propylene oxide, and embedded in resin (EMBED 812 KIT, Electron Microscopy Sciences) for 72 hours at 60 °C in the vacuum oven. 80 nm sections were stained with uranyl acetate, followed by lead citrate and viewed on a Hitachi 7000 electron microscope.
A polyclonal anti-Transgelin (aka Sm22alpha-b) antibody was generated using the TGYGRPRQIMNP peptide sequence. Antigen-affinity purification was performed according to CovanceTM protocols. The larvae were washed at room temperature with PT buffer (0.3% Triton X-100 in phosphate-buffered saline) for 5 min, rinsed three times with phosphate-buffered saline (PBS), and then fixed in 3% paraformaldehyde (PFA) overnight at 4 °C. The fixed larvae were permeabilized with 1% BSA, 1% DMSO and 0.5% Triton X-100 in PBS for 1 hour at room temperature and then incubated with primary antibody overnight at 4°C. After washing in PBT-S (0.1% Triton X-100, 1% BSA in PBS) for 2 hours and incubated with secondary antibodies and TOPRO (Molecular Probes) overnight at 4°C, the larvae were washed in PBT-S and then in PBS, and finally embedded in 1.5% low melting agarose. Images were acquired using a Zeiss LSM5 Pascal confocal microscope and Zeiss confocal projection software.
Movies. Vascular mural cells are present around the dorsal aorta and anterior mesenteric artery. Confocal projections of transverse sections of 72 hpf Tg(flk1:EGFP)s843 wild-type larvae stained for Transgelin (red) show that mural cells form a bulge of precursor cells in the region of the DA and cover the AMA. Confocal projections were taken at the level of the1st-2nd somite. Scale bar, 20 μm. NT, neural tube; SB, swim bladder; G, gut. Movie a) is at 20× magnification and b) at 40× magnification.
We would like to thank members of the Stainier and Santoro laboratories for invaluable support and discussion. We thank S. Huling and the UCSF liver centre for help with electron microscopy. Support for this research came from the Human Frontier Science Program and Regione Piemonte (M.M.S.), as well as the American Heart Association, NIH and Packard Foundation (D.Y.R.S.). M.M.S. is currently supported by a Human Frontiers Science Program Career Development Award.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.