The epaxial-hypaxial subdivision of the avian somite

In all jaw-bearing vertebrates, three-dimensional mobility relies on segregated, separately innervated epaxial and hypaxial skeletal muscles. In amniotes, these muscles form from the morphologically continuous dermomyotome and myotome, whose epaxial-hypaxial subdivision and hence the formation of distinct epaxial-hypaxial muscles is not understood. Here we show that En1 expression labels a central subdomain of the avian dermomyotome, medially abutting the expression domain of the lead-lateral or hypaxial marker Sim1. En1 expression is maintained when cells from the En1-positive dermomyotome enter the myotome and dermatome, thereby superimposing the En1-Sim1 expression boundary onto the developing musculature and dermis. En1 cells originate from the dorsomedial edge of the somite. Their development is under positive control by notochord and floor plate (Shh), dorsal neural tube (Wnt1) and surface ectoderm (Wnt1-like signalling activity) but negatively regulated by the lateral plate mesoderm (BMP4). This dependence on epaxial signals and suppression by hypaxial signals places En1 into the epaxial somitic programme. Consequently, the En1-Sim1 expression boundary marks the epaxial-hypaxial dermomyotomal or myotomal boundary. In cell aggregation assays, En1- and Sim1-expressing cells sort out, suggesting that the En1-Sim1 expression boundary may represent a true compartment boundary, foreshadowing the epaxial-hypaxial segregation of muscle.     

Sonic hedgehog controls epaxial muscle determination through Myf5 activation.

Sonic hedgehog (Shh), produced by the notochord and floor plate, is proposed to function as an inductive and trophic signal that controls somite and neural tube patterning and differentiation. To investigate Shh functions during somite myogenesis in the mouse embryo, we have analyzed the expression of the myogenic determination genes, Myf5 and MyoD, and other regulatory genes in somites of Shh null embryos and in explants of presomitic mesoderm from wild-type and Myf5 null embryos. Our findings establish that Shh has an essential inductive function in the early activation of the myogenic determination genes, Myf5 and MyoD, in the epaxial somite cells that give rise to the progenitors of the deep back muscles. Shh is not required for the activation of Myf5 and MyoD at any of the other sites of myogenesis in the mouse embryo, including the hypaxial dermomyotomal cells that give rise to the abdominal and body wall muscles, or the myogenic progenitor cells that form the limb and head muscles. Shh also functions in somites to establish and maintain the medio-lateral boundaries of epaxial and hypaxial gene expression. Myf5, and not MyoD, is the target of Shh signaling in the epaxial dermomyotome, as MyoD activation by recombinant Shh protein in presomitic mesoderm explants is defective in Myf5 null embryos. In further support of the inductive function of Shh in epaxial myogenesis, we show that Shh is not essential for the survival or the proliferation of epaxial myogenic progenitors. However, Shh is required specifically for the survival of sclerotomal cells in the ventral somite as well as for the survival of ventral and dorsal neural tube cells. We conclude, therefore, that Shh has multiple functions in the somite, including inductive functions in the activation of Myf5, leading to the determination of epaxial dermomyotomal cells to myogenesis, as well as trophic functions in the maintenance of cell survival in the sclerotome and adjacent neural tube.     

Epaxial-adaxial-hypaxial regionalisation of the vertebrate somite: evidence for a somitic organiser and a mirror-image duplication

During vertebrate neural tube formation, the initially lateral borders between the neural and epidermal ectoderm fuse to form the definitive dorsal region of the embryo, while the initially dorsally located notochord-floor plate complex is being internalised. Along the definitive dorso-ventral body axis, one can distinguish an epaxial (dorsal to the notochord) and a hypaxial (ventral to the notochord) body region. The mesodermal somites on both sides of the notochord and neural tube give rise to the trunk skeleton and skeletal muscle. Muscle forms from the somite-derived dermomyotomes and myotomes that elongate dorsally and ventrally. Based on gene expression patterns and comparative embryology, it is proposed here that the epaxial (dermo)myotome region in amniote embryos is subdivided into a dorsalmost and a centrally intercalated subregion. The intercalated subregion abuts to the hypaxial (dermo)myotome region that elongates ventrally via the hypaxial somitic bud. The dorsalmost subregion elongates towards the dorsal neural tube and is proposed to derive from an epaxial somitic bud. The dorsalmost and hypaxial somite derivatives share specific gene expression patterns which are distinct from those of the intercalated somite derivatives. The intercalated somite derivatives develop adaxially, i.e. at the level of the notochord-floor plate complex. Thus, the dorsalmost and intercalated (dermo)myotome subregions may be influenced preferentially by signals from the dorsal neural tube and from the notochord-floor plate complex, respectively. These (dermo)myotome subregions are sharply delimited from each other by molecular boundary markers, including Engrailed and Wnts. It thus appears that the molecular network that polarises borders in Drosophila and vertebrate embryogenesis is redeployed during subregionalisation of the (dermo)myotome. It is proposed here that cells within the amniote (dermo)myotome establish polarised borders with organising capacity, and that the epaxial somitic bud represents a mirror-image duplication of the hypaxial somitic bud along such a border. The resulting epaxial-intercalated/adaxial-hypaxial regionalisation of somite derivatives is conserved in vertebrates although the differentiation of sclerotome and myotome starts heterochronically in embryos of different vertebrate groups.     

The role of the notochord for epaxial myotome formation in the mouse.

The vertebrate somite is the source of all trunk skeletal muscles. Myogenesis in avian embryos is thought to depend on signals from notochord and neural tube for the epaxial muscles, and signals from lateral mesoderm and surface ectoderm for the hypaxial muscles. However, this hypothesis has to be tested because in mouse mutants lacking a notochord the presence of a fused myotome beneath the neural tube has been reported. We have compared the expression pattern of myogenic markers and markers for the hypaxial muscle precursors in the mutants Brachyury curtailed, truncate, Danforth's short tail and Pintail. In regions lacking notochord and sclerotome, we found small, ventrally located domains of Myf5 and MyoD expression, concomitant with ventrally expanded Pax3 signals and upregulated expression of the hypaxial marker Lbx1, suggesting that only the hypaxial program is active. We therefore hypothesise that in mammals, as in birds, the formation of the epaxial musculature depends on the presence of notochord derived signals.     

Clonal separation and regionalisation during formation of the medial and lateral myotomes in the mouse embryo.

In vertebrates, muscles of the back (epaxial) and of the body wall and limbs (hypaxial) derive from precursor cells located in the dermomyotome of the somites. In this paper, we investigate the mediolateral regionalisation of epaxial and hypaxial muscle precursor cells during segmentation of the paraxial mesoderm and myotome formation, using mouse LaacZ/LacZ chimeras. We demonstrate that precursors of medial and lateral myotomes are clonally separated in the mouse somite, consistent with earlier studies in birds. This clonal separation occurs after segmentation of the paraxial mesoderm. We then show that myotome precursors are mediolaterally regionalised and that this regionalisation precedes clonal separation between medial and lateral precursors. Strikingly, the properties of myotome precursors are remarkably similar in the medial and lateral domains. Finally, detailed analysis of our clones demonstrates a direct spatial relationship between the myocytes in the myotome and their precursors in the dermomyotome, and earlier in the somite and presomitic mesoderm, refuting several models of myotome formation, based on permanent stem cell systems or extensive cell mingling. This progressive mediolateral regionalisation of the myotome at the cellular level correlates with progressive changes in gene expression in the dermomyotome and myotome.     

BMP4 and noggin control embryonic blood vessel formation by antagonistic regulation of VEGFR-2 (Quek1) expression

Regulation of VEGFR-2 (Quek1) is an important mechanism during blood vessel formation. In the paraxial mesoderm, Quek1 expression is restricted to the lateral portion of the somite and later to sclerotomal cells surrounding the neural tube. By grafting of either intermediate mesoderm or BMP4 beads into the paraxial mesoderm, we show that BMP4 is a positive regulator of VEGFR-2 (Quek1) expression in the quail embryo. Separation of somites from intermediate mesoderm leads to down-regulation of Quek1 expression. The expression of Quek1 in the medial somite half is normally repressed by the notochord and becomes up-regulated and lateromedially expanded after separation of the notochord. Our results show that up-regulation of BMP4 leads to an increase of the number of blood vessels, whereas inhibition of BMP4 by noggin results in a reduction of blood vessels.     

Pax3 functions in cell survival and in pax7 regulation

In developing vertebrate embryos, Pax3 is expressed in the neural tube and in the paraxial mesoderm that gives rise to skeletal muscles. Pax3 mutants develop muscular and neural tube defects; furthermore, Pax3 is essential for the proper activation of the myogenic determination factor gene, MyoD, during early muscle development and PAX3 chromosomal translocations result in muscle tumors, providing evidence that Pax3 has diverse functions in myogenesis. To investigate the specific functions of Pax3 in development, we have examined cell survival and gene expression in presomitic mesoderm, somites and neural tube of developing wild-type and Pax3 mutant (Splotch) mouse embryos. Disruption of Pax3 expression by antisense oligonucleotides significantly impairs MyoD activation by signals from neural tube/notochord and surface ectoderm in cultured presomitic mesoderm (PSM), and is accompanied by a marked increase in programmed cell death. In Pax3 mutant (Splotch) embryos, MyoD is activated normally in the hypaxial somite, but MyoD-expressing cells are disorganized and apoptosis is prevalent in newly formed somites, but not in the neural tube or mature somites. In neural tube and somite regions where cell survival is maintained, the closely related Pax7 gene is upregulated, and its expression becomes expanded into the dorsal neural tube and somites, where Pax3 would normally be expressed. These results establish that Pax3 has complementary functions in MyoD activation and inhibition of apoptosis in the somitic mesoderm and in repression of Pax7 during neural tube and somite development.     

Sonic hedgehog is a survival factor for hypaxial muscles during mouse development

Sonic hedgehog (Shh) has been proposed to function as an inductive and trophic signal that controls development of epaxial musculature in vertebrate embryos. In contrast, development of hypaxial muscles was assumed to occur independently of Shh. We here show that formation of limb muscles was severely affected in two different mouse strains with inactivating mutations of the Shh gene. The limb muscle defect became apparent relatively late and initial stages of hypaxial muscle development were unaffected or only slightly delayed. Micromass cultures and cultures of tissue fragments derived from limbs under different conditions with or without the overlaying ectoderm indicated that Shh is required for the maintenance of the expression of myogenic regulatory factors (MRFs) and, consecutively, for the formation of differentiated limb muscle myotubes. We propose that Shh acts as a survival and proliferation factor for myogenic precursor cells during hypaxial muscle development. Detection of a reduced but significant level of Myf5 expression in the epaxial compartment of somites of Shh homozygous mutant embryos at E9.5 indicated that Shh might be dispensable for the initiation of myogenesis both in hypaxial and epaxial muscles. Our data suggest that Shh acts similarly in both somitic compartments as a survival and proliferation factor and not as a primary inducer of myogenesis.     

The mechanisms of dorsoventral patterning in the vertebrate neural tube

We describe the essential features of and the molecules involved in dorsoventral (DV) patterning in the neural tube. The neural tube is, from its very outset, patterned in this axis as there is a roof plate, floor plate, and differing numbers and types of neuroblasts. These neuroblasts develop into different types of neurons which express a different range of marker genes. Early embryological experiments identified the notochord and the somites as being responsible for the DV patterning of the neural tube and we now know that 4 signaling molecules are involved and are generated by these surrounding structures. Fibroblast growth factors (FGFs) are produced by the caudal mesoderm and must be down-regulated before neural differentiation can occur. Retinoic acid (RA) is produced by the paraxial mesoderm and is an inducer of neural differentiation and patterning and is responsible for down-regulating FGF. Sonic hedgehog (Shh) is produced by the notochord and floor plate and is responsible for inducing ventral neural cell types in a concentration-dependent manner. Bone morphogenetic proteins (BMPs) are produced by the roof plate and are responsible for inducing dorsal neural cell types in a concentration-dependent manner. Subsequently, RA is used twice more. Once from the somites for motor neuron differentiation and secondly RA is used to define the motor neuron subtypes, but in the latter case it is generated within the neural tube from differentiating motor neurons rather than from outside. These 4 signaling molecules also interact with each other, generally in a repressive fashion, and DV patterning shows how complex these interactions can be.     

Dorsal dermis development depends on a signal from the dorsal neural tube, which can be substituted by Wnt-1

To investigate the origin and nature of the signals responsible for specification of the dermatomal lineage, excised axial organs in 2-day-old chick embryos were replaced by grafts of the dorsal neural tube, or the ventral neural tube plus the notochord, or aggregates of cells engineered to produce Sonic hedgehog (Shh), Noggin, BMP-2, Wnt-1, or Wnt-3a. By E10, grafts of the ventral neural tube plus notochord or of cells producing Shh led to differentiation of cartilage and muscles, and an impaired dermis derived from already segmented somites. In contrast, grafts of the dorsal neural tube, or of cells producing Wnt-1, triggered the formation of a feather-inducing dermis. These results show that the dermatome inducer is produced by the dorsal neural tube. The signal can be Wnt-1 itself, or can be mediated, or at least mimicked by Wnt-1.     

The growth of the dermomyotome and formation of early myotome lineages in thoracolumbar somites of chicken embryos

Myotome formation in the epaxial and hypaxial domains of thoraco-lumbar somites was analyzed using fluorescent vital dye labeling of dermomyotome cells and cell-fate assessment by confocal microscopy. Muscle precursor cells for the epaxial and hypaxial myotomes are predominantly located in the dorsomedial and ventrolateral dermomyotome lips, respectively, and expansion of the dermomyotome is greatest along its mediolateral axis coincident with the dorsalward and ventralward growth directions of the epaxial and hypaxial myotomes. Measurements of the dermomyotome at different stages of development shows that myotome growth begins earlier in the epaxial than in the hypaxial domain, but that after an initial lag phase, both progress at the same rate. A combination of dye injection and/or antibody labeling of early and late-expressed muscle contractile proteins confirms the myotome mediolateral growth directions, and shows that the myotome thickness increases in a superficial (near dermis) to deep (near sclerotome) growth direction. These findings also provide a basis for predicting the following gene expression sequence program for the earliest muscle precursor lineages in mouse embryos: Pax-3 (stem cells), myf-5 (myoblast cells) and myoD (myocytes). The movements and mitotic activity of early muscle precursor cells lead to the conclusion that patterning and growth in the myotome specifically, and in the epaxial and hypaxial domains of the body generally, are governed by morphogenetic cell movements.     

FGFs, Wnts and BMPs mediate induction of VEGFR-2 (Quek-1) expression during avian somite development

Regulation of VEGFR-2 (Quek1) is an important mechanism during blood vessel formation. In the paraxial mesoderm, Quek1 expression is restricted to the lateral portion of the somite and later to sclerotomal cells surrounding the neural tube. By implanting FGF 8b/8c or SU 5402 beads into the paraxial mesoderm, we show that FGF8 in addition to BMP4 from the intermediate mesoderm (IM) is a positive regulator of VEGFR-2 (Quek1) expression in the quail embryo. The expression of Quek1 in the medial somite half is normally repressed by the notochord and Sfrps-expression in the neural tube. Over-expression of Wnt 1/3a also results in an up-regulation of Quek1 expression in the somites. We also show that up-regulation of FGF8/Wnt 1/3a leads to an increase in the number of endothelial cells, whereas inhibition of FGF and Wnt signaling by SU 5402 and Sfrp-2 results in a loss of endothelial cells. Our results demonstrate that the regulation of Quek1 expression in the somites is mediated by the cooperative actions of BMP4, FGF8 and Wnt-signaling pathways.     

Three developmental compartments involved in rib formation

When the thoracic somitic mesoderm was separated from the neural tube and the notochord with a piece of aluminum foil in two-day chick embryos, seven days after the operation ribs lacked their proximal part. The embryos were rescued by co-transplanting the notochord, the ventral half of neural tube, or QT6 cells transformed with Shh, on the somite side of the aluminum foil insert. Thus, proximal rib development depends on the notochord and the ventral neural tube, an effect which might be mediated through Shh secreted by these axial tissues. On the other hand, when the thoracic somitic mesoderm was separated from the surface ectoderm by a piece of polyethylene terephthalate film, the distal parts of the ribs were missing, suggesting that distal rib development depends on surface ectoderm. In these embryos, expression of Pax3 was weak and perturbed showing that the dermomyotome developed abnormally. It is not clear whether the development of distal rib is mediated by the dermomyotome, or the ectoderm. It has previously been shown that sternal rib development depends on lateral plate mesoderm. As to the distal rib, it is considered to be composed of two parts. Thus, the rib is composed of three developmental compartments, in agreement with a recently presented classification of somite derivatives as primaxial and abaxial.     

Sclerotomal origin of the ribs

The somites of vertebrate embryos give rise to sclerotomes and dermomyotomes. The sclerotomes form the axial skeleton, whereas the dermomyotomes give rise to all trunk muscles and the dermis of the back. The ribs were thought to be ventral processes of the axial skeleton and therefore to be derived from the sclerotomes; however, recently a dermomyotomal origin of the distal rib (the costal shaft) was suggested, with only the proximal parts (head and neck of the rib) being of sclerotomal origin. We have re-investigated the development of the ribs in quail-chick chimeras and carried out three experimental series. (1) Single dermomyotomes and (2) single sclerotomes were grafted homotopically, and (3) the ectoderm overlying the unsegmented paraxial mesoderm was removed in the prospective thoracic region. We found that the cells of the dermomyotome gave rise to epaxial and hypaxial trunk muscles, dermis of the back and endothelial cells, but not to ribs. Cells of the sclerotome formed the axial skeleton and all parts of the ribs. Ablation of the ectoderm, which affects dermomyotome development, results in severe malformations of the ribs, probably due to disturbed interactions between dermomyotome and sclerotome. Our results strongly confirm the traditional view of the sclerotomal origin of the ribs.