Gastrulation and neurulation in the chick embryo, focusing on the mesodermal component. (A) Primitive streak region, showing migrating mesodermal and endodermal precursors. (B) Formation of the notochord and paraxial mesoderm. (C, D) Differentiation of the somites, coelom, and the two aortae (which will eventually fuse). A-C, 24-hour embryos; D, 48-hour embryo.
, the formation of mesodermal and endodermal organs is not subsequent to neural tube formation, but occurs synchronously. The notochord extends beneath the neural tube from the base of the head into the tail. On either side of the neural tube lie thick bands of mesodermal cells. These bands of paraxial mesoderm are referred to as the segmental plate (in birds) and the unsegmented mesoderm (in mammals). As the primitive streak regresses and the neural folds begin to gather at the center of the embryo, the paraxial mesoderm separates into blocks of cells called somites. Although somites are transient structures, they are extremely important in organizing the segmental pattern of vertebrate embryos. As we saw in the preceding chapter, the somites determine the migration paths of neural crest cells and spinal nerve axons. Somites give rise to the cells that form the vertebrae and ribs, the dermis of the dorsal skin, the skeletal muscles of the back, and the skeletal muscles of the body wall and limbs.
Mesoderm in the vertebrate embryo. The organization of the mesoderm in the neurula stage is similar for all vertebrates. You can see this organization by viewing serial sections of the chick embryo. [Click on Chick-Mid]
Neural tube and somites. Scanning electron micrograph showing well-formed somites and paraxial mesoderm (bottom right) that has not yet separated into distinct somites. A rounding of the paraxial mesoderm into a somitomere can be seen at the lower left, and neural crest cells can be seen migrating ventrally from the roof of the neural tube. (Photograph courtesy of K. W. Tosney.)
,D and 14.3). Somite formation begins as paraxial mesoderm cells become organized into whorls of cells called somitomeres. The somitomeres become compacted and bound together by an epithelium, and eventually separate from the presomitic paraxial mesoderm to form individual somites. Because individual embryos can develop at slightly different rates (as when chick embryos are incubated at slightly different temperatures), the number of somites present is usually the best indicator of how far development has proceeded. The total number of somites formed is characteristic of a species (50 in chicks, 65 in mice, and about 500 in some snakes).
Somite formation correlates with the expression of the hairy gene. (A) Schematic representation of the posterior portion of an X-somite chick embryo. Somite X is marked. The expression of the hairy gene (purple) is seen in the caudal half of this somite, as well as in the posterior portion of the presomitic mesoderm and in a thin band that will form the caudal half of the next somite. (B) A caudal fissure (small arrow) begins to separate the somite from the presomitic mesoderm. The posterior region of hairy expression extends anteriorly. (C) The newly formed X+1 somite retains the expression of hairy in its caudal half, as the posterior domain of hairy expression moves farther anteriorly and shortens. (D) The formation of somite X+1 is completed, and the anterior region of what had been the posterior hairy expression pattern is now the anterior expression pattern. It will become the caudal domain of somite X+2. The entire process takes 90 minutes. (After Palmeirim et al. 1997.)
.
Thus, the expression pattern of the hairy gene correlates with the positioning of the place where a somite will separate from the unsegmented mesoderm. Labeling of cells with diI shows that this wave of hairy expression is not caused by the migration of particular cells that express the hairy gene. Rather, the wavelike pattern of hairy expression is an autonomous property of the presomitic mesoderm. Even if the presomitic mesoderm is separated from the caudalmost mesoderm, the ectoderm, neural tube, notochord, Hensen's node, and lateral plate regions, the dynamic expression of the hairy gene remains (Palmeirim et al. 1997; Jouve et al. 2000).
Transition from somitomere to somite. (A) Expression pattern of receptor tyrosine kinase EphA4 (blue) and its ligand, ephrinB2 (red) as somites develop. The somite boundary forms at the junction between the region of ephrin expression on the posterior of the last formed somite and the region of Eph expression on the anterior of the next somite to form. In the presomitic mesoderm, the pattern is created anew as each somite buds off. The posteriormost region of the next somite to form does not express ephrin until that somite is ready to separate. (B) N-cadherin expression (white area) correlates with the conversion of loose mesenchyme cells into the epithelial somite. (C) In situ hybridization showing the expression of Paraxis mRNA (red) in the 6-somite chick embryo. It is seen both in the somites and in the region of the unsegmented mesoderm that will give rise to the next somite. (A after Durbin et al. 1998; B from Hatta et al. 1987, courtesy of M. Takeichi; C from Barnes et al. 1997, courtesy of R. Tuan.)
; Durbin et al. 1998). As somites form, this pattern of gene expression is reiterated caudally. Interfering with this signaling (by injecting embryos with RNA encoding dominant negative Ephs) leads to abnormal somite boundary formation. Eph signaling is thought to mediate cell shape changes, and these could be responsible for separating the presomitic mesoderm at the EphA4-ephrinB2 border. Thus, ephrin-Eph signaling may be responsible for converting the prepattern established by the Hairy protein in the presomitic mesoderm into actual somites.
, the cells of the somitomere are randomly organized as a mesenchymal mass, but the synthesis of two extracellular matrix proteins, fibronectin and N-cadherin, links them into arrays that will form tight junctions and generate their own basal laminae (Figure 14.5B; Ostrovsky et al. 1984; Lash and Yamada 1986; Hatta et al. 1987). These extracellular matrix proteins, in turn, may be regulated by the expression of the Paraxis gene. This gene encodes a transcription factor that is also expressed at the rostral (anterior) end of the unsegmented mesoderm of mouse embryos, and which is seen in precisely that region that will form the somite (Figure 14.5C). Injection of antisense oligonucleotides complementary to the Paraxis message produces defects of the paraxial mesoderm, and in the somites of Paraxis-deficient mice, no epithelial structures are formed. These defective somites have segregated from the segmental plate and their cells have differentiated, but they are completely disorganized (Burgess et al. 1995; Barnes et al. 1997). The Paraxis protein is therefore an essential part of the conversion from mesenchyme to epithelium (Burgess et al. 1996; Barnes et al. 1997; Tajbakhsh and Spörle 1998).
The segmental plate mesoderm is determined as to its position along the anterior-posterior axis before somitogenesis. When segmental plate mesoderm that would ordinarily form thoracic somites is transplanted into a region in a younger embryo (caudal to the first somite) that would ordinarily give rise to cervical (neck) somites, the grafted mesoderm differentiates according to its original position and forms ribs in the neck. Note that only one side is affected. (After Kieny et al. 1972.)
; Kieny et al. 1972; Nowicki and Burke 1999). As discussed in Chapter 11 (see Figure 11.41), the somites are specified in this manner according to the Hox genes they express. Mice that are homozygous for a loss-of-function mutation of Hoxc-8 will convert a lumbar vertebra into an extra ribbed thoracic vertebra (see Figure 11.39).
Diagram of a transverse section through the trunk of a chick embryo on days 2–4. (A) The 2-day somite can be divided into sclerotome cells and dermamyotome cells. (B) On day 3, the sclerotome cells lose their adhesion to one another and migrate toward the neural tube. (C) On day 4, the remaining dermamyotome cells divide. The medial cells form an epaxial myotome beneath the dermamyotome, while the lateral cells form a hypaxial myotome. (D) A layer of muscle cell precursors (the myotome) forms beneath the epithelial dermamyotome. (A, B after Langman 1981; C, D after Ordahl 1993.)
and 14.7).
Myotome derivatives of the mouse embryo. The epaxial muscles form from the region of the dermamyotome closest to the neural tube. The hypaxial muscles form from the region of dermamyotome furthest from the neural tube. The epaxial myotome will form the back muscles and the central portion of the rib. The hypaxial myotome will form the intercostal muscles and distal portion of the rib. The proximal (ventral) proteion of the rib is formed by sclerotome cells.* (After Kato and Aoyama 1998; Ordahl and Williams 1998.)
). The cells in the two lateral portions of the epithelium (those regions closest to and farthest from the neural tube) are muscle-forming cells. They divide to produce a lower layer of muscle precursor cells, the myoblasts. The resulting double-layered structure is called the dermamyotome, and the lower layer is called the myotome. Those myoblasts formed from the region closest to the neural tube form the epaxial muscles (the deep muscles of the back), while those myoblasts formed in the region farthest from the neural tube produce the hypaxial muscles of the body wall, limbs, and tongue (Figures 14.7 and 14.8; see Christ and Ordahl 1995; Venters et al. 1999). The central region of the dorsal layer of the dermamyotome is called the dermatome, and it generates the mesenchymal connective tissue of the back skin: the dermis. (The dermis of other areas of the body forms from other mesenchymal cells, not from the somites.) The dermamyotome may also produce the distal cartilage of the ribs, its lateral edge producing the most ventral portion of the rib (Figure 14.8; Kato and Aoyama 1998).
Like the proverbial piece of real estate, the destiny of a somitic region depends on three things: location, location, and location.
Model of major postulated interactions in the patterning of the somite. A combination of Wnts (probably Wnt1 and Wnt3a) are induced by BMP4 in the dorsal neural tube. These Wnt proteins, in combination with low concentrations of Sonic hedgehog from the notochord and floor plate, induce the epaxial myotome, which synthesizes the myogenic transcription factor Myf5. High concentrations of Sonic hedgehog induce Pax1 expression in those cells fated to become the sclerotome. Certain concentrations of neurotrophin-3 (NT-3) from the dorsal neural tube appear to specify the dermatome, while Wnt proteins from the epidermis, in conjunction with BMP4 and FGF5 from the lateral plate mesoderm, are thought to induce the hypaxial myotome. (After Cossu et al. 1996b.)
; Smith and Tuan 1996). They also express I-mf, an inhibitor of the myogenic bHLH family of transcription factors that initiate muscle formation (Chen et al. 1996).
The dermatome differentiates in response to another factor secreted by the neural tube, neurotrophin 3 (NT-3). Antibodies that block the activities of NT-3 prevent the conversion of the epithelial dermatome into the loose dermal mesenchyme that migrates beneath the epidermis (Brill et al. 1995).
In similar ways, the myotome is induced by at least two distinct signals. Studies involving transplantation and knockout mice indicate that the epaxial muscle cells coming from the medial portion of the somite are induced by factors from the neural tube, probably Wnt1 and Wnt3a from the dorsal region and low levels of Sonic hedgehog from the ventral region (Münsterberg et al. 1995; Stern et al. 1995; Ikeya and Takada 1998). The hypaxial muscles coming from the lateral edge of the somite are probably induced by a combination of Wnt proteins from the epidermis and bone morphogenetic protein 4 (BMP4) from the lateral plate mesoderm (Cossu et al. 1996a; Pourquié et al. 1996; Dietrich et al. 1998). These factors cause the myotome cells to express particular transcription factors that activate the muscle-specific genes.
In addition to these positive signals, there are inhibitory signals that prevent a signal from affecting an inappropriate group of cells. For example, Sonic hedgehog not only activates sclerotome and myotome development; it also inhibits BMP4 signal from the lateral plate mesoderm from extending medially and ventrally (thus preventing the conversion of sclerotome into muscle) (Watanabe et al. 1998). Similarly, Noggin is produced by the most medial portion of the dermamyotome and prevents BMP4 from giving these cells the migratory characteristics of hypaxial muscle (Marcelle et al. 1997).
And what happens to the notochord, that central mesodermal structure? After it has provided the axial integrity of the early embryo and has induced the formation of the dorsal neural tube, most of it degenerates. Wherever the sclerotome cells have formed a vertebral body, the notochordal cells die. However, in between the vertebrae, the notochordal cells form the tissue of the intervertebral discs, the nuclei pulposi. These are the discs that “slip” in certain back injuries.
14.1 Calling the competence of the somite into question. When the tbx6 gene was knocked out from mice, the resulting embryos had three neural tubes in the posterior of theirbodies. Without the tbx6 gene, the somitic tissue responded to the notochord and epidermal signals as if it were neural ectoderm. http://www.devbio.com/chap14/link1401.shtml
14.2 Cranial paraxial mesoderm. Most of the head musculature does not come from somites. Rather, it comes from the cranial paraxial mesoderm. These cells originate adjacent to the sides of the brain, and they migrate to their respective destinations. http://www.devbio.com/chap14/link1402.shtml
This model has recently been challenged by Huang et al. (2000), whose transplantation experiments confirm the traditional view that the entire rib originates from the sclerotome.
