The vertebrate dorsal aorta is the earliest intraembryonic blood vessel composed of an inner lining of ECs and a slightly later forming outer wall consisting of smooth muscle cells (SMCs) and fibrocytes. Whereas the vascular endothelium of the vertebrate body is exclusively derived from the mesoderm, the lineages leading to vascular smooth muscle cells are of mesodermal and neural crest origin. For instance, cephalic neural crest provides SMCs of the head, neck, and outflow tract of the heart (Le Lievre and Le Douarin,1975; Jiang et al.,2000; Etchevers et al.,2001), and coronary SMCs are derived from the splanchnic mesoderm (Mikawa and Gourdie,1996; Vrancken Peeters et al.,1999). Depending on their position within the embryo, developing vessels are thought to recruit SMC precursors from the local mesenchyme (Hungerford and Little,1999). Two recent reports in chick and mouse demonstrated that the paraxial mesoderm, i.e., the somites, contributes to the wall of the dorsal aorta (Esner et al.,2006; Pouget et al.,2006). However, these studies could not clarify if in the chick the cells that populate the wall of the aorta are of dermomyotomal or sclerotomal origin, which prompted us to investigate the somitic origin of SMC precursors in the wall of the avian dorsal aorta.
Precursor cells of the thoracic and lumbar aorta undergo vasculogenesis, an in situ differentiation of angioblasts (Risau and Flamme,1995; Risau,1997; Ambler et al.,2001). The aortic anlage is situated lateral to the segmental plate between the endoderm and the overlying splanchnic mesoderm (Fig. 1A,B). At the level of the last formed somite, corresponding to somite stage I (Christ and Ordahl,1995), paired aortae have formed, which progressively move under the now segmented paraxial mesoderm to eventually fuse in the midline of the embryo forming a single dorsal aorta (Fig. 1C,D). During this process, the aortic endothelium becomes dynamically remodelled (Wilting et al.,1995; Pardanaud et al.,1996; Pouget et al.,2006). Primary ECs of the dorsal aorta are derived from the splanchnic mesoderm, but just before fusion of the paired aortae, ECs in the aortic roof are replaced by a secondary EC lineage from the somites (Wilting et al.,1995; Pardanaud et al.,1996). Between embryonic days (E) 3 to 4, the aortic floor, which still is of splanchnic mesodermal origin at this stage, downregulates endothelial markers and acquires hematopoietic traits (Jaffredo et al.,1998). Subsequently, the ECs in the aortic floor are replaced by ECs of the secondary lineage from the somite as well, whereby the remodelling process of the aortic endothelium is completed (Pouget et al.,2006).
Concomitantly, somite-derived SMC precursors migrate around the aorta to form its wall (Pouget et al.,2006). Based on a retrospective clonal analysis carried out in mice, Esner et al. (2006) proposed that SMCs in the wall of the dorsal aorta and skeletal muscle cells of the myotome are derived from common progenitor cells. By analysing labelled cells in Pax3GFP/+ mice, they reasoned that these multipotent progenitor cells must be located in the hypaxial dermomyotome (Esner et al.,2006). According to this hypothesis, dermomyotome-derived SMC precursors colonize the wall of the dorsal aorta and adopt an SMC phenotype.
In the present study, we report by using quail–chick chimerization that α-smooth muscle actin (αSMA) expressing cells in the wall of the dorsal aorta originate from the sclerotome compartment. At E5, these cells constitute most of the aortic wall. We propose a two-step process of aortic wall formation: Before aortic fusion, the first αSMA-expressing cells that are present in the floor of the aorta are of non-somitic origin, likely originating from the splanchnopleure. SMCs in the dorsal and lateral part of the aorta are derived from the sclerotome. Later on, the primary SMCs in the aortic floor are replaced by SMCs derived from the sclerotome.
Formation of the Dorsal Aorta
To clarify the position of the forming aorta relative to the neighboring structures, we examined a series of semithin sections of chicken embryos at appropriate stages. We detected aortic precursor cells lateral to the segmental plate between the splanchnic mesoderm and the endoderm forming endothelial lacunae (Fig. 1A). These lacunae progressively fused to build larger vessels (Fig. 1B) that moved towards the embryonic midline at the level of the last formed epithelial somite (Fig. 1C). At this level, the aortic endothelium contacts the endoderm ventrally, the splanchnic mesoderm laterally, the intermediate mesoderm dorsally, and the ventrolateral side of the somite dorsomedially (Fig. 1C). At the mature somite level, the aorta has translocated further towards the embryonic midline, the dorsal half now contacts the sclerotome, while the ventral half contacts the splanchnic mesoderm (Fig. 1D). Subsequently, the paired aortae fuse in the embryonic midline.
Maturation of the Aortic Wall
We found the first expression of α-smooth muscle actin (αSMA), one of the earliest markers of differentiating SMCs (Frid et al.,1992; Hungerford et al.,1996), in a small number of cells close to the aortic endothelium at wing bud level in HH stage-16 embryos (Fig. 2A). At this stage, αSMA is restricted to cells in the aortic floor, which is closely associated with the endoderm and splanchnic mesoderm. At HH stage 17, more αSMA-expressing cells are found in the aortic floor, which contacts the splanchnic mesoderm. In addition, αSMA-expressing cells can be detected first at the lateral and then at the dorsal sides of the aorta, which are covered by sclerotomal tissue (Fig. 2B–E). By HH stage 18, αSMA-expressing cells have completely encircled the aorta at wing bud level (Fig. 2F). More than one layer of αSMA-expressing cells is present in the aortic floor. As development proceeds, the aortic endothelium gets covered by several layers of SMCs (Fig. 2G,H) and, by HH stage 24, expression of αSMA can also be found in the cardinal veins (Fig. 2I).
To verify that αSMA is expressed in mural cells, we looked at quail tissue sections that were double labeled with QH1 antibody, a quail-specific endothelial/hematopoietic cell marker (Pardanaud et al.,1987), and αSMA antibody, at various stages. We always detected distinct expression of αSMA in cells that did not stain for QH1 and were in a mural position (Fig. 3). Thus, αSMA marks early differentiating, aortic mural cells, and aortic SMCs during later stages.
The Paraxial Mesoderm Contributes to the Wall of the Dorsal Aorta
To elucidate the origin of SMCs in the wall of the dorsal aorta, we analyzed serially sectioned quail-chick chimeras after transplantation of single epithelial somites. In these chimeras, quail cells could be detected in the wall of the dorsal aorta on the operated side (n = 11/14) at E4. A fraction of host cells was still present among these donor cells (data not shown). However, after transplantation of anterior segmental plates (corresponding to 4–5 somites) (Figs. 4,5), nearly all cells in the wall of the dorsal aorta were of quail origin at E5. Interestingly, these cells did not cross the embryonic midline (Fig. 5) (n = 8/8). Thus, our quail-chick transplantation experiments are supportive of a somitic origin of the aortic SMCs.
The Sclerotome But Not the Dermomyotome Contributes to the Wall of the Dorsal Aorta
Within the somites, it has been controversial which compartment gives rise to aortic SMCs. We, therefore, investigated a possible contribution of the dermomyotomal and/or sclerotomal compartment to the formation of the aortic wall. After transplantation of hypaxial dermomyotomes (Fig. 6A), quail nuclei were detected in all derivatives of the hypaxial dermomyotome. However, in none of the cases examined were quail cells detected near the dorsal aorta (n = 7/7). The mesenchyme between the aorta and hypaxial myotome, the latter being of quail origin, was completely devoid of any cells derived form the graft (Fig. 6B,C). To exclude an early migration of SMC precursors form the dermomyotome towards the aorta, we transplanted dorsolateral somite quadrants, which give rise to hypaxial dermomyotome, from quail to chick embryos. Although all dorsolateral somite derivatives were stained, we never detected any quail cells in the aortic wall (n = 11/13) (data not shown). This demonstrates that in the chick there is no dermomyotomal contribution to the aortic wall.
In contrast, after transplantation of ventral somite halves (Fig. 7A), quail cells were detected in the wall of the aorta (n = 12/12). After a reincubation period of 1 day (at E3), we detected quail cells around the neural tube and notochord, in the central sclerotome, and in close association with the aortic roof (Fig. 7B–D). At this stage, the first cells in the aortic roof start to express αSMA (Fig. 2F). After reincubation periods of 2 days or longer, quail cells could be found in the entire wall of the dorsal aorta on the operated side (Fig. 7E–G). This demonstrates that the SMCs of the chick aorta arise from the sclerotome and not from the dermomyotome.
Maturation of blood vessels during embryonic development involves the recruitment of mural cells, i.e., SMCs and fibrocytes, to the developing vessel wall that forms around the endothelial vessel anlage. Vascular SMCs are recruited from different embryonic origins depending on the location in the body of the vessels concerned. For instance, the walls of the coronary vessels of the heart are formed by the splanchnic mesoderm (Mikawa and Gourdie,1996; Vrancken Peeters et al.,1999), whereas the mural cells in the head arise from the cephalic neural crest (Le Lievre and Le Douarin,1975; Etchevers et al.,2001).
Among the embryonic blood vessels, the aorta is of outstanding importance for the establishment of the embryonic vascular system and is the first intraembryonic vessel to form. The origin of the aortic wall cells has been of interest for embryologists for many decades. Based on morphological observations, Wilhelm His proposed as early as 1868 that the ventral part of the sclerotome forms the wall of the Aorta descendens. He observed that the ventral sclerotome grows dorsally and laterally around the aortic tube as it moves towards the embryonic midline (His,1868). The first experimental evidence for the origin of the aortic wall came more than one hundred years later with the use of quail-chick chimerization for cell lineage studies. In 1975, Le Lievre and Le Douarin showed that the proximal aorta forming the cardiac outflow tract is of neural crest origin. For the avian Aorta descendens, Pouget et al. (2006) presented evidence of a somitic origin of the aortic wall, even though the somitic compartment giving rise to these cells remained elusive.
In this report, we show by means of quail-chick chimerization, that SMCs of the avian dorsal aorta are derived from the sclerotome, whereas the dermomyotome does not contribute to the aortic wall. We analyzed the timing of SMC recruitment from the sclerotome to the aortic wall. In keeping with earlier data (Hungerford et al.,1996), we found that the first aortic SMCs appear in the floor of the aorta, and could show that these primary SMCs are not of somitic origin. The dorsal aorta develops in close association with the splanchnic mesoderm, which also gives rise to primary aortic ECs. This strongly argues that the early differentiating primary SMCs in the aortic floor also arise from the splanchnic mesoderm, which is in direct contact with the aortic floor at this stage. Alternatively, the primary SMCs might be derived from primary ECs of the aortic floor through a process known as transdifferentiation (DeRuiter et al.,1997). Further labelling studies are needed to clarify this question.
In a later stage, one day after transplantation of ventral somite halves (prospective sclerotome), we identified graft-derived SMC precursors in the aortic roof and lateral sides, but not in the aortic floor, which, at this stage, still contains primary SMCs. Subsequently, at E4 of development, when the aortic floor becomes remodelled by secondary ECs from the somite (Pouget et al.,2006), SMCs of sclerotomal origin constitute the entire aortic wall including the aortic floor. These secondary SMCs from the sclerotome differentiate in close association with the aortic endothelium. We argue that they migrate ventrally and displace the primary SMCs from the splanchnic mesoderm, thus paralleling the endothelial remodelling as shown by Wilting et al. (1995) and Pardanaud et al. (1996). Alternatively, it is possible that the primary wall cells stay in situ, but are dispersed by massive somatic cell migration. A definitive answer to this question requires lineage studies of primary wall cells.
The molecular control of aortic mural cell recruitment is largely unknown. It is likely that signals in the floor of the aorta and/or endoderm induce primary SMC recruitment from the splanchnic mesoderm to the aortic floor. Similarly, endothelium-derived signals, such as PDGF-BB or TGF-β, might be crucial in this context of migration and differentiation of secondary SMC precursors from the sclerotome (reviewed by Jain,2003). These questions await functional analysis by gain-and loss-of-function studies.
Our data do not support studies in mice arguing for the existence of multipotent progenitor cells located in the hypaxial dermomyotome, which give rise to both the skeletal muscle of the hypaxial myotome and the mural SMCs of the dorsal aorta (Esner et al.,2006). This discrepancy might be attributable to the different technique applied, e.g., timing problems due to very early recombination events prior to the compartmentalization of the paraxial mesoderm in mice carrying nlaacZ-reporter, or might reflect species differences between the mammalian and avian systems.
In summary, we could consistently show that in chick, only the sclerotome but not the dermomyotome contributes to αSMA-expressing mural cells of the dorsal aorta in the trunk. Based on our findings, we propose a two-step process of aortic wall formation. First, non-paraxial mesoderm-derived mural cells accumulate in the aortic floor. Due to the position of neighboring tissues during the early formation of the aorta, these primary SMCs very likely arise from the splanchnic mesoderm. Second, SMCs from the sclerotome are recruited to the roof and sides of the aorta and migrate ventrally, thereby replacing the primary lineage of SMCs in the aortic floor.
Preparation of Quail and Chick Embryos
Fertilized White Leghorn chick (Gallus domesticus) and Japanese quail eggs (Coturnix coturnix japonica) were incubated at 38°C under 80% humidity and the embryos were staged according to Hamburger and Hamilton (1951).
The following grafting experiments were performed between quail (donor) and chick (host) embryos at HH stages 10–18 (see Table 1): transplantation of (1) anterior segmental plates (an equivalent of 4–5 somites), (2) single somites (somite stages I and II according to Christ and Ordahl,1995), (3) dorsolateral somite quadrants (somite stages I and II), (4) hypaxial dermomyotomes (somite stages VI–VIII), and (5) ventral somite halves (somite stages I and II). All grafting experiments were performed homotopically and unilaterally (on the right side of the embryo). Details of the grafting procedure were described previously (Wilting et al.,1995; Zhi et al.,1996; Huang et al.,2000). Depending on the experiment, donors were placed either ventral- or dorsal-side down into a Petri dish covered with 4% agarose. Care was taken not to co-transplant any endothelial cells that might be associated with the graft. In some experiments, pancreatin (Gibco) was used to isolate the graft from the donor. A small window was made into the host egg shell that had been treated with wax, thereby allowing a drop of Locke's solution (Locke and Rosenheim,1907; Rugh,1962) to be placed on the window. This method, together with injection of Indian ink below the embryo and proper use of light, greatly enhanced contrast and visualization. Host tissue was removed by pipetting and in some cases pancreatin was used to help to prepare the site of grafting. Care was taken not to injure the dorsal aorta when the host tissue was removed. After successful transplantation, the eggs were sealed and re-incubated for 1 to 5 days.
Table 1. Summary of Grafting Experiments
Operation stage (HH)
Number of operations
Reincubation period (days)
Number of cases showing a contribution to the aortic wall
Anterior segmental plates
Dorsolateral somite quadrants
Ventral somite halves
Immunostaining was carried out according to standard procedures. Briefly, embryos were fixed overnight in Serra's fixative (Serra,1946). After dehydration, the embryos were embedded in paraffin and serially sectioned at 8 μm. Before immunostaining, sections were blocked with 1% bovine serum albumin. Incubation with the primary antibodies was overnight at room temperature, followed by washing with PBS, and a 90-min incubation with secondary antibodies.
Mouse monoclonal primary antibodies used were QCPN antibody (Developmental Studies Hybridoma Bank, DSHB; cell culture supernatant diluted 1:5), which detects quail cell nuclei; QH1 antibody (DSHB; cell culture supernatant undiluted), which detects quail ECs and hematopoietic cells (Pardanaud,1987); and anti-α-smooth muscle actin antibody (clone 1A4; Sigma; diluted 1:1,500).
Antibodies were used for single or double labelling. Goat secondary antibodies used were alkaline phosphatase-conjugated anti-mouse antibody (Dako; diluted 1:1,000) or fluorochrome-conjugated anti-mouse Ig-Cy3 (Jackson ImmunoResearch, diluted 1:200), anti-mouse IgG1-Alexa-Fluor 488 (Molecular Probes, diluted 1:300), and anti-mouse IgG2A-Alexa-Fluor 555 (Molecular Probes, diluted 1:300). For conventional light microscopy, sections were counterstained with nuclei red, mounted in entellane, and analyzed with an Axioskop microscope (Zeiss). For confocal microscopy, sections were counterstained with 4′,6-diamidino-2-phenylindole (DAPI) (Invitrogen) to visualize the cell nuclei, mounted in Fluoromount-G (SouthernBiotech), and examined with a confocal microscope LSM 510 (Zeiss).
Preparation of Semithin Sections
Embryos were fixed in 4% paraformaldehyde overnight, dehydrated, and embedded in Durcupan resin (Fluka). Polymerization was performed at 65°C. Semithin sections of 0.75-μm thickness were cut with an Ultracut S (Leica) and stained with 1% methylene blue and 1% azure II (Fluka).
We thank Ute Baur, Lidia Koschny, and Günther Frank for their excellent technical assistance. C.W. is a fellow of the research training program “From Cells to Organs: Molecular Mechanisms of Organogenesis” (GRK1104, Deutsche Forschungsgemeinschaft). This study was supported by the Deutsche Forschungsgemeinschaft (GRK1104 to B.C. and M.S.; SFB592 A1 to B.C. and M.S.) and the European Network of Excellence MYORES (to B.C. and M.S.).