Ablation of axial structures activates apoptotic pathways in somite cells of the chick embryo

It has been known for some time that ablation of the neural tube and/or the notochord in the chick embryo leads to a massive wave of cell death in the adjacent somites. It is postulated that in the normal embryo, survival signals emanate from the neural tube and/or notochord that suppress apoptosis in the cells of the somites, except for a small population of sclerotome cells that are programmed to die naturally. In this study we show that axial ablation results in the death of sclerotome and not somitic neural crest cells, and we have examined the apoptotic response of these cells to the ablation. We show that several elements of the apoptotic cascade become detectable in somite cells in response to the withdrawal of survival signals. We demonstrate the down-regulation of bcl-2 protein in the somites adjacent to, and caudal to, the site of ablation, corresponding to the region that displays an elevated level of cell death. Although caspase-9 appeared to be activated in somites at all levels of the trunk, caspase-2 showed a clear response to the ablation of the axial structures. Removal of the neural tube and notochord produced an up-regulation of caspase-2 activity in somites in the region of the operation. Cleavage of two down-stream substrates of these caspases was examined. The cleavage of poly (ADP-ribose) polymerase (PARP) was apparent in somites at all levels of the trunk, and showed only a modest up-regulation after ablation. By contrast, the cleavage of DNA fragmentation factor (DFF45) showed a marked up-regulation in response to ablation, suggesting that this is a primary substrate for a caspase-dependent apoptotic mechanism. Evidence was also found for a caspase-independent mechanism, since the expression of apoptosis-inducing factor (AIF) was found to be very sensitive to, and up-regulated in somites by, axial ablation. Because the wave of apoptosis that is precipitated in somites by removal of the axial structures may be mediated by BMP-4, we examined the levels of BMP-4 in somites in response to axial ablation. BMP-4 expression was clearly up-regulated in somites adjacent to, or close to, the site of operation. Accepted: 9 July 2001     

A resegmentation‐shift model for vertebral patterning

Segmentation of the vertebrate body axis is established in the embryo by formation of somites, which give rise to the axial muscles (myotome) and vertebrae (sclerotome). To allow a muscle to attach to two successive vertebrae, the myotome and sclerotome must be repositioned by half a segment with respect to each other. Two main models have been put forward: ‘resegmentation’ proposes that each half‐sclerotome joins with the half‐sclerotome from the next adjacent somite to form a vertebra containing cells from two successive somites on each side of the midline. The second model postulates that a single vertebra is made from a single somite and that the sclerotome shifts with respect to the myotome. There is conflicting evidence for these models, and the possibility that the mechanism may vary along the vertebral column has not been considered. Here we use DiI and DiO to trace somite contributions to the vertebrae in different axial regions in the chick embryo. We demonstrate that vertebral bodies and neural arches form by resegmentation but that sclerotome cells shift in a region‐specific manner according to their dorsoventral position within a segment. We propose a ‘resegmentation‐shift’ model as the mechanism for amniote vertebral patterning.     

Local signalling in dermomyotomal cell type specification

Summary The development and differentiation of the avian myotome was studied after removal of the neural tube, including neural crest, and after replacement of dorsal half-somites by ventral half-somites. Results show that in the absence of neural tissue myoblast differentiation within the somites does not take place. Ventral halfsomites are able to undergo muscle differentiation if they were grafted in place of dorsal half somites. It is suggested that local signals must be responsible for the dorsalisation of the newly formed somite including myoblast differentiation. Neural crest cells are discussed as possible sources of these signals.     

The migration and distribution of somite cells after labelling with the carbocyanine dye, Dil: the relationship of this distribution to segmentation in the vertebrate body

Summary The cells of individual somites in 2-day-old chick embryos were marked by injecting a fluorescent dye into the somitocoele. This procedure permanently marked the cells and allowed their subsequent development and distribution to be followed. The cells were found to remain in close association with each other within limited boundaries and did not mix to any great extent with similar cells from adjacent somites. Fluorescent cells from single somites were found in the intervertebral disc, connective tissue surrounding two adjacent neural arches, all the tissues between the neural arches, the dermatome, and the associated myotome. No fluorescent cells were found in the notochord or in any nervous tissue apart from accompanying connective tissue. Surprisingly, the vertebral bodies and neural arches did not contain any fluorescent cells apart from those in the connective tissue surrounding them, but this absence of fluorescent cells was thought to be due to the dilution of the fluorescence following cell proliferation. These results provide further experimental support for the theory of resegmentation in vertebral formation, and also provide evidence of a compartmental method of development along the rostrocaudal axis in vertebrates, similar to that already discovered in insects. On the basis of cell lineage criteria, the sclerotome might be considered as a developmental compartment.     

The binding pattern of peanut lectin associated with sclerotome migration and the formation of the vertebral axis in the chick embryo

Summary Lectins have been used extensively to detect changes in carbohydrate moieties on the surface of embryonic cells during early development. Peanut agglutinin (PNA) in particular has been used to investigate changes related to cell differentiation. PNA has also been used to differentiate between the rostral and caudal sclerotome halves which have been shown to be functionally different, with neural crest cells and neurites traversing only the rostral half during their migration. In this study, we have sectioned and stained chick embryos between 3 and 8 days of age with PNA to examine the distribution of PNA binding sites associated with the vertebral column during this period and also to determine the fates of the rostral and caudal sclerotome halves. Ultrastructural localisation of PNA-gold conjugate showed that binding sites for this lectin were present intracellularly and extracellularly both on cell surfaces and in the matrix. At the light microscope level, a clear banding pattern emerged after staining with PNA which consisted of alternating light and dark staining along the entire length of the vertebral axis of the embryo. In the younger embryos, a simple banding pattern emerged where the rostral sclerotome half of each segment stained only lightly while the caudal half stained darkly. This banding pattern was present throughout the 6 day period of development and could be traced continuously but grew more complex as the sclerotome cells migrated to surround the notochord and neural tube and as the dorsal root ganglia developed. The rostral sclerotome half was found to contribute to the caudal part of one vertebral body and its neural arch, while the caudal sclerotome half was found to contribute to the intervertebral disc, the rostral half of the next caudal vertebra, and part of its neural arch.     

背根神经节分节格式的形成和体节再分节

把分散的鹌鹑体节板细胞移植到鸡的体节板区。移植后１７ｈ，在移植区形成大小不等、排列杂乱的ＤＣ体节。有些体节靠近神经管，有些较靠外侧。移植后２９～３２ｈ，靠近神经管的ＤＣ体节向中壁破裂，外迁的间充质细胞在生肌节与神经管和脊索之间，形成生骨节（即原位生骨节）。每个生骨节再分为前后两半，位于外侧的ＤＣ体节，在移植后３３～３８ｈ，其侧壁或外侧壁破裂。外迁的间充质细胞在生肌节的外侧或生肌节之间，数量、大小和     

Morphological analysis of the role of the neural tube and notochord in the development of somites

The differentiation of avian somites and skeletal muscles, which themselves are derived from somites, was studied in ovo after the isolation of the unsegmented segmental plate from the notochord and/or neural tube by surgical operations at the level of the segmental plate. In each experiment, the newly formed somites had a normal histological structure, with an outer epithelial somite and core cells in the somitocoeles. Thereafter, the three derivatives of the somites (dermatome, myotome and sclerotome) reacted differently to the different operations. When the somites developed without the notochord, only somitocoele cells showed massive cell death, and muscles developed regardless of the presence or absence of the neural tube. On the contrary, no cell death was observed in any part of the somites that were formed with the neural tube or the notochord present, and muscle cells developed. However, in those embryos that retained only the notochord, striated muscles developed only in the lateral body wall. In each of the experimental operations, the surface ectoderm always covered the somites, and, regardless of the state of sclerotome and/or myotome differentiation, the dermatome always survived. These histological observations indicate that the first step in somite formation is independent of axial structures. The results further suggest that the notochord may produce diffusible factors that are necessary for the survival and further development of sclerotomal cells, and that both the neural tube and notochord can support muscle differentiation. However, it is likely that each structure has a relationship to the development of epaxial muscles and hypaxial muscles respectively. Furthermore, an intimate relationship may also exist between the surface ectoderm and the development of the dermatome.     

From somites to vertebral column.

We report on the development and differentiation of the somites with respect to vertebral column formation in avian and human embryos. The somites, which are made up of different compartments, establish a segmental pattern which becomes transferred to adjacent structures such as the peripheral nervous system and the vascular system. Each vertebra arises from three sclerotomic areas. The paired lateral ones give rise to the neural arches, the ribs and the pedicles of vertebrae, whereas the vertebral body and the intervening disc develop from the axially-located mesenchyme. The neural arches originate from the caudal half of one somite, whereas the vertebral body is made up of the adjacent parts of two somites. Interactions between notochord and axial mesenchyme are a prerequisite for the normal development of vertebral bodies and intervening discs. The neural arches form a frame for the neural tube and spinal ganglia. The boundary between head and vertebral column is located between the 5th and 6th somites. In the human embryo, proatlas, body of the atlas segment, and body of the axis fuse to form the axis.     

Immunophenotypic characterization of enteric neural crest cells in the developing avian colorectum

A chick–chick intraspecies chimera was created by removing the neural tube adjacent to somites 2–6 from a normal chick embryo at E1.5 and replacing it with equivalent tissue from an age‐matched chick‐GFP transgenic embryo. At E10, the colorectum was removed, sectioned, and stained with HNK‐1 antibody (red) to detect neural crest‐derived cells, and with DAPI (blue) to label nuclei. Vagal neural crest‐derived cells are HNK‐1+/GFP+, while sacral neural crest derived‐cells, which comprise the nerve of Remak, are HNK‐1+/GFP?. From Nagy et al., Developmental Dynamics 241:842–851, 2012. © 2012 Wiley Periodicals, Inc.     

The rostral level of origin of sympathetic neurons in the chick embryo,studied in tissue culture

Groups of three consecutive somites from the first to the eleventh somite from chick embryos of stages 17–18 were grown in tissue culture for seven days. Sympathetic neurons, identified both by phase contrast microscopy and FIF histochemistry, occurred only in cultures which included the sixth, or more caudal, somites. If it is assumed that sympathetic precursor cells (neural crest cells) have not undergone a caudal shift prior to stages 17–18, and taking into account the loss of one or two rostral somites, then the anterior sympathetic ganglia are derived from neural crest caudal to the sixth or seventh somite Thus, the vagal zone (level with somites 1–7) contributes little to the sympathetic nervous system.     

A segmented pattern of cell death during development of the chick embryo

Summary During the early development of the chick embryo, specific groups of cells die in characteristic patterns. In this study, Nile Blue sulphate staining was used to reveal a novel pattern of segmentally repeated cell death in the paraxial mesoderm of the chick prior to stage 23. This pattern varies according to the developmental stage of the embryo and shifts rostrocaudally, corresponding to progressing somite differentiation. Initially, during early somite differentiation, cell death is restricted to the rostral half of the somite (the rostral pattern of cell death). After the somite has differentiated into dermomyotome and sclerotome, dead cells appear in superficial tissues in a pyramidal pattern which lies in register (rostrocaudally) with the central part of the sclerotome. Finally, small bands of dying cells are seen between the neural tube and the expanding sclerotome. This third pattern (the ventral path) lies in register with the rostral part of the caudal half of the sclerotome. We show by fluorescent labelling of the migrating neural crest that these patterns of cell death correspond to the routes of neural crest migration. In addition, serial sectioning of stage 23 chick embryos confirms that the position of dying cells correlates with the known routes of neural crest migration and with the sites of development of certain neural crest-derived tissues.     

Analysis of hindbrain neural crest migration in the long-tailed monkey (Macaca fascicularis)

Neural crest cells make a substantial contribution to normal craniofacial development. Despite advances made in identifying migrating neural crest cells in avian embryos and, more recently, rodent embryos, knowledge of crest cell migration in primates has been limited to what was obtained by conventional morphological techniques. In order to determine the degree to which the nonhuman primate fits the mammalian pattern, we studied the features of putative neural crest cell migration in the hindbrain of the long-tailed monkey (Macaca fascicularis) embryo. Cranial crest cells were identified on the basis of reported distributional and morphological criteria as well as by immunocytochemical detection of the neural cell adhesion molecule (N-CAM) that labels a subpopulation of these cells. The persistent labeling of a sufficient number of crest cells with antibodies to N-CAM following their exit from the rostral, preotic and post-otic regions of the hindbrain facilitated tracking them along subectodermal pathways to their respective destinations in the first, second and third pharyngeal arches. Peroxidase immunocytochemistry was also employed to localize laminin and collagen-IV in neuroepithelial basement membranes. At stage 10 (8–11 somites), crest emigration occurred in areas of unfused neural folds through focal disruptions in the neuroepithelial basement membrane in both the rostral and pre-otic regions, although there was little evidence of crest migration in the post-otic hindbrain. By stage 11 (16–17 somites), the neural folds were fused (pre- and post-otic hindbrain) or in the process of fusing (rostral hindbrain), yet crest cell emigration was apparent in all three areas through discontinuities in the basement membrane. Emigration was essentially complete at stage 12 (21 somites) as indicated by nearly continuous cranial neural tube basement membranes. At this stage the pre-ganglia (trigeminal, facioacoustic and glossopharyngeal) were consistently stained with N-CAM. The current study has provided new information on mammalian neural crest in a well-established experimental model for normal and abnormal human development, including its use as a model for the retinoic acid syndrome. In this regard, the current results provide the basis for probing the mechanisms of retinoid embryopathy which may involve perturbation of hindbrain neural crest development.     

The incorporation and dispersion of cells and latex beads on microinjection into the amniotic cavity of the mouse embryo at the early-somite stage

Summary The ability of cells and latex beads to become incorporated into the cranial region of embryos after microinjection into the amniotic cavity was studied. Premigratory neural crest cells isolated from the lateral margins of the neuroepithelium, 3T3 fibroblast cells or H35 hepatoma cells were labelled with WGA-gold conjugates, and were then microinjected into the amniotic cavity of embryos with two to three somites in vitro. Latex beads were similarly microinjected into different groups of embryos. Incorporation of injected cells or latex beads was found in the neural crest of the midbrain and the hindbrain of 5–20% of the recipients 4 h after microinjection. At 6 and 12 h, increasingly more embryos (20–77%) were observed with labelled cells or latex beads in the crest region. While hepatoma cells and latex beads were restricted to the crest region, injected neural crest cells and fibroblasts were also found in the lateral mesenchyme, bounded laterally by the surface ectoderm and medially by the closing neural tube. By 24 h after microinjection, the injected cells or latex beads were found in 50–80% of the recipients. Neural crest cells and fibroblasts, which showed similar patterns of distribution in the embryos, were located on the dorsal aspect of the neural tube, the lateral mesenchyme, the pharyngeal arches and the regions for ganglia. Hepatoma cells and latex beads were limited to the dorsal regions of the neural tube. When microinjection was carried out in embryos with seven to eight somites, incorporation of cells or latex beads was found in 44–75% of embryos, but no dispersion of the incorporated cells or latex beads into the mesenchyme was found 24 h after microinjection. Incorporation and dispersion of cells and latex beads were not observed when embryos with 18–20 somites were used as recipients. The present study showed that neural crest or fibroblast cells when injected into the amniotic cavity could be incorporated into the neural crest, and then undergo migration along the neural crest pathways, whereas hepatoma cells and latex beads could only be incorporated. The incorporation and migration of the exogenous tissues are related to the formation and the accessibility of the neural crest in the recipients.     

Analysis of hindbrain neural crest migration in the long-tailed monkey (Macaca fascicularis)

Neural crest cells make a substantial contribution to normal craniofacial development. Despite advances made in identifying migrating neural crest cells in avian embryos and, more recently, rodent embryos, knowledge of crest cell migration in primates has been limited to what was obtained by conventional morphological techniques. In order to determine the degree to which the nonhuman primate fits the mammalian pattern, we studied the features of putative neural crest cell migration in the hindbrain of the long-tailed monkey (Macaca fascicularis) embryo. Cranial crest cells were identified on the basis of reported distributional and morphological criteria as well as by immunocytochemical detection of the neural cell adhesion molecule (N-CAM) that labels a subpopulation of these cells. The persistent labeling of a sufficient number of crest cells with antibodies to N-CAM following their exit from the rostral, preotic and post-otic regions of the hindbrain facilitated tracking them along subectodermal pathways to their respective destinations in the first, second and third pharyngeal arches. Peroxidase immunocytochemistry was also employed to localize laminin and collagen-IV in neuroepithelial basement membranes. At stage 10 (8–11 somites), crest emigration occurred in areas of unfused neural folds through focal disruptions in the neuroepithelial basement membrane in both the rostral and pre-otic regions, although there was little evidence of crest migration in the post-otic hindbrain. By stage 11 (16–17 somites), the neural folds were fused (pre- and post-otic hindbrain) or in the process of fusing (rostral hindbrain), yet crest cell emigration was apparent in all three areas through discontinuities in the basement membrane. Emigration was essentially complete at stage 12 (21 somites) as indicated by nearly continuous cranial neural tube basement membranes. At this stage the pre-ganglia (trigeminal, facioacoustic and glossopharyngeal) were consistently stained with N-CAM. The current study has provided new information on mammalian neural crest in a well-established experimental model for normal and abnormal human development, including its use as a model for the retinoic acid syndrome. In this regard, the current results provide the basis for probing the mechanisms of retinoid embryopathy which may involve perturbation of hindbrain neural crest development.     

Segmentation in staged human embryos: the occipitocervical region revisited

The first seven somites, the rhombomeres, and the pharyngeal arches were reassessed in 145 serially sectioned human embryos of stages 9-23, 22 of which were controlled by precise graphic reconstructions. Segmentation begins in the neuromeres, somites and aortic arches at stage 9. The following new observations are presented. (1) The first somite in the human, unlike that of the chick, is neither reduced in size nor different in structure, and it possesses sclerotome, somitocoel and dermatomyotome. (2) Somites 1-4, unlike those of the chick, are related to rhombomere 8 (rather than 7 and 8) and are caudal to pharyngeal arch 4 (rather than in line with 3 and 4). (3) Occipital segment 4 resembles a developing vertebra more than do segments 1-3. (4) The development of the basioccipital resembles that of the first two cervical vertebrae in that medial and lateral components arise in a manner that differs from that in the rest of the vertebral column. (5) The two groups of somites, occipital 1-4 and cervical 5-7, each form a median skeletal mass. (6) An 'S-shaped head/trunk interface', described for the chick and unjustifiably for the mouse, was not found because it is not compatible with the topographical development of the otic primordium and somite 1, between which neural crest migrates without hindrance in mammals. (7) Occipital segmentation and related features are documented by photomicrographs and graphic interpretations for the first time in the human. It is confirmed that the first somite, unlike that of the chick, is separated from the otic primordium by a distance, although the otic anlage undergoes a relative shift caudally. The important, although frequently neglected, distinction between lateral and medial components is emphasized. Laterally, sclerotomes 3 and 4 delineate the hypoglossal foramen, 4 gives rise to the exoccipital and participates in the occipital condyle, 5 forms the posterior arch of the atlas and 6 provides the neural arch of the axis, which is greater in height than the arches of the other cervical vertebrae. Medially, the perinotochord and migrated sclerotomic cells give rise to the basioccipital as well as to the vertebral centra, including the tripartite column of the axis. Registration between (1) the somites and (2) the occipital and cervical medial segments becomes interrupted by the special development of the axis, the three components of which come to occupy the height of only 2 1/2 segments.     

Timing of odontogenic neural crest cell migration and tooth-forming capability in mice.

The mammalian tooth develops through sequential and reciprocal interactions between cranial neural crest (CNC)- derived ectomesenchymal cells and the stomadial epithelium. Classic tissue recombination studies demonstrated that premigratory CNC cells and CNC-derived ectomesenchymal cells possess odontogenic capacity and can respond to oral epithelial signals to form a tooth, suggesting that the CNC cells contributing to odontogenic tissue are not prespecified. Here we show that, in mice, CNC cells have populated the forming first branchial arch before the 9-somite stage and continue to migrate into the arch by the 13-somite stage. Grafts of the first arch from the 10-somite embryo or earlier yielded membranous bone and cysts but no teeth after subrenal culture. However, grafts of the first arch with its dorsally adjacent tissue containing migrating neural crest cells from the same age embryos gave rise to teeth. In contrast, teeth formed in first arch grafts that do not contain migrating neural crest cells from embryos with 12 or more somites. Interestingly, the acquisition of tooth forming capability in the first arch coincides with the onset of Fgf8 expression in the oral epithelium. These results suggest that there exists a population of odontogenic neural crest cells that migrates into the first arch between the 10- and 12-somite stages. These cells either possess odontogenic potential and are able to initiate tooth development, or can respond to odontogenic signals derived from the oral epithelium to support tooth formation.     

Migratory patterns and developmental potential of trunk neural crest cells in the axolotl embryo.

Using cell markers and grafting, we examined the timing of migration and developmental potential of trunk neural crest cells in axolotl. No obvious differences in pathway choice were noted for DiI-labeling at different lateral or medial positions of the trunk neural folds in neurulae, which contributed not only to neural crest but also to Rohon-Beard neurons. Labeling wild-type dorsal trunks at pre- and early-migratory stages revealed that individual neural crest cells migrate away from the neural tube along two main routes: first, dorsolaterally between the epidermis and somites and, later, ventromedially between the somites and neural tube/notochord. Dorsolaterally migrating crest primarily forms pigment cells, with those from anterior (but not mid or posterior) trunk neural folds also contributing glia and neurons to the lateral line. White mutants have impaired dorsolateral but normal ventromedial migration. At late migratory stages, most labeled cells move along the ventromedial pathway or into the dorsal fin. Contrasting with other anamniotes, axolotl has a minor neural crest contribution to the dorsal fin, most of which arises from the dermomyotome. Taken together, the results reveal stereotypic migration and differentiation of neural crest cells in axolotl that differ from other vertebrates in timing of entry onto the dorsolateral pathway and extent of contribution to some derivatives.     

Ephrin-as cooperate with EphA4 to promote trunk neural crest migration

Trunk neural crest cells delaminate from the dorsal neural tube and migrate on two distinct pathways: a dorsolateral route, between the ectoderm and somites,and a ventromedial route, through the somitic mesoderm. Neural crest cells that migrate ventromedially travel in a segmental manner through rostral half-somites, avoiding caudal halves. Recent studies demonstrate that various molecular cues guide the migration of neural crest cells, primarily by serving as inhibitors to premature pathway entry orby preventing neural crest from entering inappropriate territories. Trajectories of migrating trunk neural crest are well organized and generally linear in nature, suggesting that positive, migration-promoting factors may be responsible for this organized cell behavior. However, the identity of these factors and their function are not well understood. Here we examine the expression of members of the EphA subclass of receptor tyrosine kinases and ephrins using RT-PCR and immunocytochemistry. Neural crest cells express ephrins and EphA4 at distinct stages during their migration. In functional analyses, addition of ephrin-A2-, ephrin-A5-, and EphA4-Fc disrupted the segmental organization of trunk neural crest migration in explants: neural crest cells entered rostral and caudal halves of somites. Finally, to test the specific effects of these factors on cell behavior, neural crest cells were exposed in vitro to substrate-bound EphA and ephrin-As. Surprisingly, neural crest cells avoided ephrin-A2 or ephrin-A5 substrates; this avoidance was abolished by the addition of EphA4. Together, these data suggest that ephrin-As and EphA4 cooperate to positively promote the migration of neural crest cells through rostral half somites in vivo.     

The development of the avian vertebral column

Segmentation of the paraxial mesoderm leads to somite formation. The underlying molecular mechanisms involve the oscillation of ”clock-genes” like c-hairy-1 and lunatic fringe indicative of an implication of the Notch signaling pathway. The cranio-caudal polarity of each segment is already established in the cranial part of the segmental plate and accompanied by the expression of genes like Delta1, Mesp1, Mesp2, Uncx-1, and EphA4 which are restricted to one half of the prospective somite. Dorsoventral compartmentalization of somites leads to the development of the dermomyotome and the sclerotome, the latter forming as a consequence of an epithelio-to-mesenchymal transition of the ventral part of the somite. The sclerotome cells express Pax-1 and Pax-9, which are induced by notochordal signals mediated by sonic hedgehog (Shh) and noggin. The craniocaudal somite compartmentalization that becomes visible in the sclerotomes is the prerequisite for the segmental pattern of the peripheral nervous system and the formation of the vertebrae and ribs, whose boundaries are shifted half a segment compared to the sclerotome boundaries. Sclerotome development is characterized by the formation of three subcompartments giving rise to different parts of the axial skeleton and ribs. The lateral sclerotome gives rise to the laminae and pedicles of the neural arches and to the ribs. Its development depends on signals from the notochord and the myotome. The ventral sclerotome giving rise to the vertebral bodies and intervertebral discs is made up of Pax-1 expressing cells that have invaded the perinotochordal space. The dorsal sclerotome is formed by cells that migrate from the dorso-medial angle of the sclerotome into the space between the roof plate of the neural tube and the dermis. These cells express the genes Msx1 and Msx2, which are induced by BMP-4 secreted from the roof plate, and they later form the dorsal part of the neural arch and the spinous process. The formation of the ventral and dorsal sclerotome requires directed migration of sclerotome cells. The regionalization of the paraxial mesoderm occurs by a combination of functionally Hox genes, the Hox code, and determines the segment identity. The development of the vertebral column is a consequence of a segment-specific balance between proliferation, apoptosis and differentiation of cells. Accepted: 25 May 2000