Somites represent the first visual evidence of segmentation in the developing vertebrate embryo and it is becoming clear that this segmental pattern of the somites is used in the initial stages of development of other segmented systems such as the peripheral nervous system. However, it is not known whether the somites continue to contribute to the maintenance of the segmental pattern after the dispersal of the somitic cells. In particular, the extent to which cells from a single somite contribute to all of the tissues of a single body segment and the extent to which they mix with cells from adjacent segments during their migration is not known. In this study, we have replaced single somites in the future cervical region of 2-day-old chick embryos with equivalent, similarly staged quail somites. The chimerae were then allowed to develop for a further 6 days when they were killed. The cervical region was dissected and serially sectioned. The sections were stained with the Feulgen reaction for DNA to differentiate between the chick and quail cells. The results showed that the cells from a single somite remained as a clearly delimited group throughout their migration. Furthermore, the sclerotome, dermatome and myotome portions from the single somites could always be recognised as being separate from similar cells from other somites. The somitic cells formed all of the tissues within a body segment excluding the epidermis, notochord and neural tissue. There was very little mixing of the somitic cells between adjacent segments. The segmental pattern of the somites is therefore maintained during the migration of the somitic cells and this might be fundamental to a mechanism whereby the segmentation of structures, such as the peripheral nervous system, is also maintained during development.

In a recent paper (Bagnall et al. 1988), we provided the first experimental evidence in support of the theory of resegmentation in vertebral formation. Using the chick-quail chimaera model, in which we replaced chick somites with equivalently staged donor quail somites, we were able to show that a single somite contributes to the caudal part of one vertebral body including part of its neural arch, the intervertebral disc, and the rostral part of the next caudal vertebra, again including part of its neural arch. However, this probably represents only part of the contribution made by a somite to the vertebral column. The tissues associated with the vertebral column include nerves, ligaments, muscles, blood vessels, general connective tissue, as well as the cartilage and fibrocartilage of the vertebrae and the intervertebral discs. The extent to which a single somite contributes to these tissues within a single body segment is not known. Furthermore, the extent to which cells from neighbouring somites mix with one another in the formation of these tissues is also not known (Keynes and Stern, 1988) although Verbout (1985) has suggested that there might be considerable mixing of the sclerotome cells from different segments during the formation of the vertebral column. If cells from adjacent somites do mix to any great extent, then any segmentation pattern represented by the somites would be destroyed. This would have significant implications for understanding the development of segmental structures, such as the peripheral nervous system, whose development appears to be determined by the characteristics of the somitic cells (Keynes and Stern, 1988).

This study was designed to determine the contribution made by a single somite to the various tissues associated with a body segment and, in particular, to determine the degree of mixing and integration of the somitic cells with similar cells from adjacent segments. We achieved this by replacing single somites in chick embryos with similarly staged, equivalent quail somites. We show that the segmental pattern of the somites is largely maintained after their dispersal and differentiation.

The methodology for this study has been described in detail in a previous paper (Bagnall et al. 1988). Briefly, a single chick somite in the future cervical region was replaced by an equivalent, similarly staged quail somite at approximately stage 12 according to the system of Hamburger and Hamilton (1951) for chick embryos. Attempts were made to maintain the correct orientation of the somite during the transplantation using carbon particles as markers, but this method was not always completely reliable. The chimaeric embryo was then incubated for a further 6 –8 days and killed, after which the cervical region was dissected. In several cases, the ventral portion of the cervical region was removed during this procedure, including the gut tube. Consequently, this study did not include assessment of the contributions made by the somites to this structure. The cervical region was embedded in paraffin, and serially sectioned. The sections were then stained with the Feulgen method for demonstration of DNA to distinguish the quail from the chick cells (Le Douarin, 1973). The position and distribution of the quail cells within the specific tissues was then noted in each section and recorded. Table 1 is a summary of the data relating to the host chick and donor quail embryos used during the transplantation and is taken from our previous paper (Bagnall et al. 1988).

Table 1.

A summary of the data relating to the host chick and donor quail embryos used during the transplantation

A summary of the data relating to the host chick and donor quail embryos used during the transplantation
A summary of the data relating to the host chick and donor quail embryos used during the transplantation

As described in our previous paper (Bagnall etal. 1988), the distribution of the donor quail cell population was clearly distinguished from the host chick cell population. In all the embryos, it was clear that the cells from the somite remained as a clearly delimited group as the somite dispersed and the cells migrated to their respective sites. Typical examples of this are shown in Fig. 1. In the coronal section shown in Fig. 1A and C, which is more dorsal than Fig. IB and D, the quail cells clearly form a repetitive or segmental portion of the vertebral column if the relative positions of the dorsal root ganglia within the segments are compared. In the more ventral section from the same embryo (Fig. ID), in which the intervertebral disc can be seen, the delimited area of quail cells is still clearly defined and the segmental nature of the area can be recognised again by the relative positions of the two nerves. Fig. ID can also be used to provide evidence in support of the theory of resegmentation in vertebral formation. The quail cells can be seen to contribute medially to the intervertebral disc, the caudal part of one adjacent vertebral body, and the rostral part of the next consecutive vertebral body (the notochord is still relatively large at this stage of development). Furthermore, the quail cells can also be seen contributing laterally to the caudal part of one neural arch (rostral vertebra) and the rostral part of the neural arch of the next caudal vertebra with the spinal nerve developing between them.

Fig. 1.

(⧣83/3) A and B are low power (×63) photomicrographs of the general areas shown in C and D, respectively. The precise areas shown in C and D are represented by the areas enclosed by the dashed lines in A and B. In both C and D, the area enclosed by the dashed lines approximately represents the area of distribution of the quail cells and delimits the segmental contribution of the somite. Note the clear delimitation of the quail cells in C and D. The sections are all oriented coronally. (C) The quail cells clearly form a segment of the body if the positions of the two dorsal root ganglia are compared. ×250. (D) In this more ventral section, the quail cells can be seen forming parts of two consecutive vertebrae. The quail cells also form a segment of the body if the relative positions of the two nerves are compared. ×250. g, dorsal root ganglion; ivd, intervertebral disc; n, spinal nerve; na, neural arch; nc, notochord; nt, neural tube; q, quail cells.

Fig. 1.

(⧣83/3) A and B are low power (×63) photomicrographs of the general areas shown in C and D, respectively. The precise areas shown in C and D are represented by the areas enclosed by the dashed lines in A and B. In both C and D, the area enclosed by the dashed lines approximately represents the area of distribution of the quail cells and delimits the segmental contribution of the somite. Note the clear delimitation of the quail cells in C and D. The sections are all oriented coronally. (C) The quail cells clearly form a segment of the body if the positions of the two dorsal root ganglia are compared. ×250. (D) In this more ventral section, the quail cells can be seen forming parts of two consecutive vertebrae. The quail cells also form a segment of the body if the relative positions of the two nerves are compared. ×250. g, dorsal root ganglion; ivd, intervertebral disc; n, spinal nerve; na, neural arch; nc, notochord; nt, neural tube; q, quail cells.

In older embryos (Fig. 2A and B) in which the vertebral anlage had chondrified, the quail cells retained their clearly defined boundaries but the tissues fused at these boundaries and mixed slightly with similar tissues made up of chick cells. In Fig. 2C, the quail cells can be clearly seen forming the cartilage of the vertebra) body right up to the notochord but not forming any part of the notochord. The quail cells (chondrocytes) can also be seen mixed with the chick chondrocytes at the junction between the two different types of cells. Significantly, the perichondrium surrounding the chick cartilage is made up of chick cells and the perichondrium surrounding the quail cartilage is made up of quail cells. This is also shown in Fig. 2D where the cartilage of the costal process and the surrounding perichondrium, is clearly a composite of both chick and quail cells in distinct, separate regions.

Fig. 2.

A and B are low power (×63) photomicrographs of the general areas shown in C and D respectively. The precise areas shown in C and D are represented by the areas enclosed by the dashed lines in A and B. The sections are transversely oriented. (C) (52/2) Note that the chick and quail cells are separate and form distinct areas of the vertebral body. The quail cells also lie immediately adjacent to the notochord but do not form any part of it. The perichondrium surrounding the chick cells is of chick origin while the perichondrium surrounding the quail cells is of quail origin. ×400. (D) (75/3) The costal process is made up of both chick and quail cells as is the surrounding perichondrium. The two areas of different cell origin blend with each other. ×400. c, chick cells; cp, costal process; nc, notochord; pc, perichondrium; q, quail cells.

Fig. 2.

A and B are low power (×63) photomicrographs of the general areas shown in C and D respectively. The precise areas shown in C and D are represented by the areas enclosed by the dashed lines in A and B. The sections are transversely oriented. (C) (52/2) Note that the chick and quail cells are separate and form distinct areas of the vertebral body. The quail cells also lie immediately adjacent to the notochord but do not form any part of it. The perichondrium surrounding the chick cells is of chick origin while the perichondrium surrounding the quail cells is of quail origin. ×400. (D) (75/3) The costal process is made up of both chick and quail cells as is the surrounding perichondrium. The two areas of different cell origin blend with each other. ×400. c, chick cells; cp, costal process; nc, notochord; pc, perichondrium; q, quail cells.

In those sites around the neural arch where the vertebral anlage from separate segments overlap and joints are destined to develop, the quail cells form the cartilage of both articulating processes as well as the intervening tissue that will eventually form the joint. In Fig. 3, the quail cells can be seen forming the chondrocytes of the two portions of the neural arches from two consecutive vertebrae as well as the intervening tissue, which will ultimately develop into a synovial joint.

Fig. 3.

(44/2) Sagittal section. The adjacent parts of the neural arches are both quail as is the intervening tissue that will ultimately form the synovial joint. ×400.

Fig. 3.

(44/2) Sagittal section. The adjacent parts of the neural arches are both quail as is the intervening tissue that will ultimately form the synovial joint. ×400.

More medially (Fig. 4A), the quail cells also appear to form all of the tissues intervening between the neural arch and the neural tube. In Fig. 4B, the quail cells can be seen filling all of the space lying dorsal to the notochord and extending to the neural tube. Clearly, this includes all layers of the meninges. The quail chondrocytic cells in the vertebral anlagen can also be seen merging with similar chick cells from the opposite side dorsal to the notochord. In Fig. 4C, which has been taken more laterally around the neural arch from the same embryo, the quail cells can be identified in both the chondrocytes of the neural arch and the layers of the meninges. It is possible that some of the spaces in the meningeal layers represent remains of blood vessels (as well as some artefacts of separation of the connective tissue created during histological preparation). If this is so, then they would presumably be composed of quail cells also.

Fig. 4.

A is a low power (×63) photomicrograph of the general area shown in C and D. The precise areas shown in C and D are enclosed by the dashed lines in A and labelled appropriately. The section is transversely oriented. (B) (75/3) Note that the distribution of the quail cells stretches from the notochord to the neural tube and that all tissues formed in this area are of quail origin. ×400. (C) (75/3) More laterally around the neural arch, note that the quail cells form the cartilaginous arch and fill the space between the arch and the neural tube. ×400. c, chick cells; j, joint; m, meninges; na, neural arch; nc, notochord; nt, neural tube; q, quail cells.

Fig. 4.

A is a low power (×63) photomicrograph of the general area shown in C and D. The precise areas shown in C and D are enclosed by the dashed lines in A and labelled appropriately. The section is transversely oriented. (B) (75/3) Note that the distribution of the quail cells stretches from the notochord to the neural tube and that all tissues formed in this area are of quail origin. ×400. (C) (75/3) More laterally around the neural arch, note that the quail cells form the cartilaginous arch and fill the space between the arch and the neural tube. ×400. c, chick cells; j, joint; m, meninges; na, neural arch; nc, notochord; nt, neural tube; q, quail cells.

It was clear from all the embryos that, at the level of the quail transplant, nearly all of the visible structures are composed of the quail cells and they are very prominent across the whole section. For example, in Fig. 5A, all of the tissues are made up of quail cells apart from the nervous tissues. Very clearly, the neural tube, the dorsal root ganglion and the cells interspersed along the associated spinal nerve are composed of chick cells. This isolation of the chick dorsal root ganglion from the surrounding quail cells is highlighted in other sections where the quail cells can be seen to surround completely the dorsal root ganglion but do not become part of it (Fig. 5B). These surrounding cells are situated to become the connective tissue of the developing ganglion. The cells along the spinal nerve associated with the dorsal root ganglion are also chick (Fig. 5C) and contrast sharply with the quail cells lying alongside. These chick cells are presumably the Schwann cells and the surrounding quail cells will eventually form the connective tissue associated with the developing nerve.

Fig. 5.

Transverse sections. (A) (75/3) All of the tissue except for the neural tissue is made up of quail cells. ×160. (B) (52/2) Note that the quail cells surround the chick dorsal root ganglion but do not become part of it. ×250. (C) (26/5) The emerging nerve has chick cells scattered along it. Presumably these will become the Schwann cells. The quail cells lie adjacent to the nerve in a position to form the connective tissue associated with it. ×250. g, dorsal root ganglion; n, spinal nerve; na, neural arch; nt, neural tube.

Fig. 5.

Transverse sections. (A) (75/3) All of the tissue except for the neural tissue is made up of quail cells. ×160. (B) (52/2) Note that the quail cells surround the chick dorsal root ganglion but do not become part of it. ×250. (C) (26/5) The emerging nerve has chick cells scattered along it. Presumably these will become the Schwann cells. The quail cells lie adjacent to the nerve in a position to form the connective tissue associated with it. ×250. g, dorsal root ganglion; n, spinal nerve; na, neural arch; nt, neural tube.

Perhaps the most interesting result was related to the blood vessels in the area of the quail cells. In regions that were dominated by the chick cells, the endothelium of the blood vessels was clearly chick. However, in the regions dominated by the quail cells, endothelium composed of quail cells was often seen (Fig. 6A). This region gave way to chick cell endothelium in the same blood vessel in later sections. In the dermis, where the cells are more diffuse and separated, very thin-walled blood vessels were present. These vessels could be identified by the presence of red blood cells in their lumen. Where quail cells were very prominent, many of these vessels had regions where quail cells had been incorporated into the endothelium (Fig. 6B). Even in some larger blood vessels, such as the aorta, which presumably had been present before the dispersal of the somite, the quail cells could be seen contributing to the wall of the blood vessel but probably not to the endothelium itself (Fig. 6C). Presumably, these quail cells had been incorporated into the tissue during the enlargement of these vessels.

Fig. 6.

(A) (40/1) The quail cells can be seen forming part of the endothelium (arrow) of the large blood vessel close to the vertebral column. ×400. (B) (92/3) The quail cells can be seen forming part of the endothelium (arrow) of the small blood vessel in the dermis. ×400. (C) (83/1) The quail cells can be seen forming the surrounding tissue, but probably not the actual endothelium, of the aorta (arrow), rbc, red blood cells. ×400.

Fig. 6.

(A) (40/1) The quail cells can be seen forming part of the endothelium (arrow) of the large blood vessel close to the vertebral column. ×400. (B) (92/3) The quail cells can be seen forming part of the endothelium (arrow) of the small blood vessel in the dermis. ×400. (C) (83/1) The quail cells can be seen forming the surrounding tissue, but probably not the actual endothelium, of the aorta (arrow), rbc, red blood cells. ×400.

The epidermis was recognisable in all sections and was always composed of chick cells (Fig. 7). However, in the dermis, there was always a band or group of quail cells that was very prominent. The quail cells in the dermis were widely separated from each other as were the dermal chick cells in later sections (Fig. 7A). In those embryos in which feather buds were forming, the quail cells could often be found associated with the developing buds (Fig. 7B). Again, however, the quail cells remained as a clearly identified group although they mixed with chick cells at their periphery.

Fig. 7.

(A) (92/3) The quail cells are not in the epidermis and are scattered in the dermis but are still recognisable as a group. ×250. (B) (38/6) The quail cells can be seen contributing to the formation of a feather bud. They are adjacent to some chick cells that are also forming part of the bud. ×400. c, chick cells; q, quail cells.

Fig. 7.

(A) (92/3) The quail cells are not in the epidermis and are scattered in the dermis but are still recognisable as a group. ×250. (B) (38/6) The quail cells can be seen contributing to the formation of a feather bud. They are adjacent to some chick cells that are also forming part of the bud. ×400. c, chick cells; q, quail cells.

Groups of cells surrounding the vertebral column were forming such that they were in the appropriate positions to be considered the precursors of developing muscle masses. However, within these muscle masses it was not possible to distinguish skeletal muscle cells from the developing associated connective tissue fibroblasts. Fig. 8A shows a group of quail cells forming a muscle mass close to the vertebral column that is clearly separate from an adjacent mass of chick cells. The quail cell muscle mass appears to have some chick cells scattered throughout it but there do not appear to be as many quail cells in the chick muscle cell mass. In Fig. 8B, a larger mass of muscle can be seen that is composed almost entirely of quail cells. This mass can be seen to blend with a mass of adjacent chick cells to form the complete muscle mass. In Fig. 8C, a more lateral and smaller muscle mass which lies further away from the vertebral column can be seen that is composed of a mixture of both chick and quail cells. This muscle mass could be traced over several segments and, in the region of the transplant, the chick and quail cells were always well integrated. At other levels, this same muscle mass consisted entirely of chick cells.

Fig. 8.

(A) (75/3) The quail cells can be seen forming a muscle mass that is adjacent to another muscle mass composed of chick cells. A nerve lies between them. ×250. (B) (75/3) The quail cells can be seen forming a much larger muscle mass. ×250. (C) (38/3) Further away from the vertebral column, the quail cells can be seen forming a smaller muscle mass that is a mixture of both chick and quail cells. ×250. c, chick cells; n, spinal nerve; q, quail cells.

Fig. 8.

(A) (75/3) The quail cells can be seen forming a muscle mass that is adjacent to another muscle mass composed of chick cells. A nerve lies between them. ×250. (B) (75/3) The quail cells can be seen forming a much larger muscle mass. ×250. (C) (38/3) Further away from the vertebral column, the quail cells can be seen forming a smaller muscle mass that is a mixture of both chick and quail cells. ×250. c, chick cells; n, spinal nerve; q, quail cells.

Closer to the vertebral column, it was again impossible to distinguish between skeletal muscle cells and simple fibroblasts in regions connecting adjacent vertebrae that would eventually become ligamentous or muscular. The two types of developing tissue looked very similar. The tissue that was forming between the vertebral elements in such a way that the two elements would ultimately be connected by the tissue was, therefore, considered to be either developing intervertebral muscle or developing ligament. In Fig. 9A, the cells going to form either muscle or ligament between the two neural arches can be recognised because the cells are aligned parallel to each other with their long axes being parallel to that of the vertebral column. In this example, most of the cells in this mass are of quail origin although some are chick. The quail cells also appear to be confined between the two neural arches and do not extend beyond them to any great extent. In contrast, at a similar site in other embryos, the proportion of quail cells was less, comprising only approximately 50 % of the population (Fig. 9B). Again, the few quail cells that were present were confined to the space between the two adjacent neural arches. In other embryos, the space between the two neural arches was filled mainly by chick cells although the adjacent parts of the neural arches were clearly of quail origin. However, again, any quail cells that were present were confined to the space between the two neural arches (Fig. 9C).

Fig. 9.

(A) (44/2) The quail cells can be seen between the neural arches. Here they are going to form muscle or ligament which is composed mostly of quail cells. ×250. (B) (40/4) The cells lying between the neural arches consist of both chick and quail cells. ×400. (C) (34/4) The cells lying between the neural arches consist mainly of chick cells. ×400. Note in all three Figs that few, if any, quail cells lie outside of the boundary between the two neural arches. Note also that the developing neural arches are made up of both chick and quail cells which are clearly separated from each other. The dashed lines in B and C represent the boundaries of the developing neural arches.c, chick cells; n, spinal nerve; na, neural arch; q, quail cells.

Fig. 9.

(A) (44/2) The quail cells can be seen between the neural arches. Here they are going to form muscle or ligament which is composed mostly of quail cells. ×250. (B) (40/4) The cells lying between the neural arches consist of both chick and quail cells. ×400. (C) (34/4) The cells lying between the neural arches consist mainly of chick cells. ×400. Note in all three Figs that few, if any, quail cells lie outside of the boundary between the two neural arches. Note also that the developing neural arches are made up of both chick and quail cells which are clearly separated from each other. The dashed lines in B and C represent the boundaries of the developing neural arches.c, chick cells; n, spinal nerve; na, neural arch; q, quail cells.

In some embryos, examples of aberrant tissue were found. This consisted of what appeared to be nephric tubules developing in a normal vertebral site and surrounded by vertebral anlage. For example, in Fig. 10A, a chick nephric tubule is clearly developing just lateral to the dorsal root ganglion but within the band of quail cells that will form the vertebra in this area. Similarly in Fig. 10B, several quail nephric tubules can be seen forming adjacent to both chick and quail cells that would ultimately have formed part of the vertebral column.

Fig. 10.

(A) (83/3) A chick nephric tubule can be seen developing in an area normally reserved for vertebrae. It is surrounded by quail cells that will form the vertebra. ×400. (B) (21/2) There are quail nephric tubules developing surrounded by both chick and quail cells. ×400.g, dorsal root ganglion; ne, nephric tubule.

Fig. 10.

(A) (83/3) A chick nephric tubule can be seen developing in an area normally reserved for vertebrae. It is surrounded by quail cells that will form the vertebra. ×400. (B) (21/2) There are quail nephric tubules developing surrounded by both chick and quail cells. ×400.g, dorsal root ganglion; ne, nephric tubule.

The production of chick-quail chimaerae by the method used in this study is well-established in the literature (Le Douarin, 1982; Balaban et al. 1988; Aoyama and Asamoto, 1988) and similar procedures involving spinal cord transplants have produced chimaerae that can hatch, stand, walk and fly (Kinutani and Le Douarin, 1985). However, it is open to some potential criticism as Sanders (1986) has shown that the quail cells are more adhesive than chick cells and this may account in part for the clearly defined aggregation of the quail cells found in this study. Furthermore, Newgreen et al. (1985) showed that disturbance of the local environment during the transplant can lead to abnormal displacement of at least the migrating neural crest cells in relation to the notochord. Nevertheless, the quail cells in this study appear to be well integrated into the developing vertebral column, which also appears to be normal. Similar results, showing good integration of the quail cells within the developing vertebral column following similar methodology, have also been produced in other studies (Beresford, 1983; Aoyama and Asamoto, 1988). Consequently, it is believed that the distribution pattern shown by the quail cells in this study is normal.

In general terms, the fate of the somites is already well documented. The cells forming the ventral and medial walls of the somite lose their epithelial shape and migrate to surround the notochord. These migrating sclerotome cells form a loosely woven tissue (mesenchyme) and eventually form the vertebral column. The cells of the remaining dorsal somite wall (dermatome) spread out under the overlying ectoderm and form the dermis and the subcutaneous tissue of the skin as well as giving rise to a new layer of cells (myotome) which provides the segmental muscle component. However, the precise contribution of the somites to the developing tissues and organs is rarely, if ever, described and the extent of integration of the cells from a single somite with similar cells from other somites to form these tissues and organs is not known.

Our results show that the medially migrating sclerotome cells form all of the tissues associated with the vertebral column excluding the skeletal muscle, nerve and notochord. The skeletal muscle is derived from the myotome cells of the somite, while the notochord is formed directly from the epiblast and the neural tissue from the ectoderm. The tissues formed from the sclerotome cells include the cartilaginous vertebrae (see Bagnall et al. 1988 for further detail), the annulus fibrosus of the intervertebral discs, the meninges, the endothelium of some blood vessels, all the general connective tissue, and presumably the ligaments. These results confirm and extend the results of other workers. For example, Haninec (1988) studied the origin of connective tissue sheaths of peripheral nerves in the limbs of avian embryos using chick-quail chimaera and found results similar to ours where the connective tissue sheaths were formed from the general mesenchyme while the Schwann cells were from the same origin as the nerves. Similarly, the results also support the work of Bunge et al. (1989) who suggested that the perineurium in vitro arises from fibroblasts.

The meninges represent another form of supportive tissue for neural derivatives and it is interesting to note that, at the cervical level, at least, all the meningeal layers were found to be of quail composition indicating that they are of sclerotomal origin. There is only a limited literature available on the development of the meninges which was thought by early investigators to arise from the neural primordium. The most recent and thorough study has been reported by Le Douarin (1982) on the unpublished work of Le Lievre who also used the chick-quail transplantation technique but who transplanted only neural primordia at various levels concentrating mainly on the brain. The results varied depending on the level studied but the meninges of the spinal cord were found to be entirely of the host type indicating that the meninges do not develop from the neural tissue. Our results extend these findings by showing that the meninges come specifically from the sclerotome.

Le Douarin (1982) also reported that the endothelial cells of the blood vessels irrigating the meninges following the transplantation of quail neural primordia were chimaeric in a similar way to that suggested by our own results. The quail endothelial cells in the meningeal vessels reported by Le Douarin (1982) presumably originated from the neural crest of the transplanted quail neural primordium since it has been established that neural crest cells participate in the formation of some blood vessels (Le Douarin, 1982; Kirby et al. 1983, 1985; Phillips et al. 1987). In our study, the quail endothelial cells, found in blood vessels of the dermis, blood vessels lying close to the vertebral column and in the blood vessels of the meninges, must have originated from the sclerotome of the transplanted quail somite. Jacobson (1988) reported that the elements of the vascular system appear early especially on the ventral surface of the mesoderm and Willier et al. (1955) reported that once the embryonic vascular system is fully established, the endothelium of new vessels arises only as an outgrowth from pre-existing vessels. In conjunction with this, our results and those reported by Le Douarin (1982) suggest that embryonic endothelial cells may develop from both neural crest and sclerotome cells during formation of the vessels. This is supported by further work reported by Le Douarin (1982) in which transplanted quail cells from the neural primordium were found only in the musculo-connective tissue wall of the larger arteries and not in the endothelium. Presumably such arteries were already established when the neural crest cells were incorporated and, consequently, the neural crest cells contributed to the musculo-connective tissue wall as it developed and not to the endothelial lining. Similarly in our study, the quail cells were presumably incorporated into the endothelium while the vascular system was being established and could then only make contributions to the musculo-connective tissue wall of the blood vessels at later stages of development.

In summary, the sclerotome cells remain as a very clearly defined band during their migration and do not mix to any great extent with cells from the adjacent segments. The only mixing is at the periphery of the clearly delimited area. In particular, the sclerotome cells remain separate from the developing neural tissue but are in position to develop into supportive tissue for these structures.

The dermatome cells migrate laterally to form the dermis in which the cells are distributed less densely than those of the sclerotome. Nevertheless, the dermal cells also remain as a delimited group with little mixing with similar cells from the adjacent segments. Within the dermis, the cells also form at least some of the endothelial cells that line the thin-walled blood vessels within the segment. In those regions of the dermis where feather buds form, the dermal cells from a somite are more compact and still remain as a group. They contribute to the formation of these buds but often the whole bud is a mixture of cells from at least two somites although the cells from the separate somites remain in clearly defined groups.

The myotome cells are perhaps the most widely scattered of the somitic cells and form the least well-defined group. This is, perhaps, not surprising as several studies have shown that the myotome cells are flexible in their determination and their innervation does not depend on their segmental origin (Chevallier et al. 1977: Keynes et al. 1987; Lance-Jones, 1988). When they are transplanted to a different vertebral level they form the correct muscle for the region to which they have been transplanted. In this study, the results have shown that the myotome cells often mix with myotome cells from other segments and become members of a larger muscle mass that spans several segments. Consequently, this larger muscle mass consists of cells from at least two somites (i.e. contained both chick and quail cells). However, within this muscle mass, the cells from one somite remain as a group to a large extent although there is always an infiltration of cells from other somites to a greater extent than in either the sclerotome or dermatome.

It has already been established that when appropriate quail somites are implanted into chick embryo hosts, the skeletal muscle of the limbs are of donor quail origin (Christ et al. 1977 and 1983; Chevallier et al. 1977; Beresford, 1983) but that the tendons and connective tissues are of host chick somatopleural origin (Chevallier et al. 1977). In contrast, the results from our study showed many instances in which regions of the muscle and accompanying connective tissue lying close to the vertebral column was totally of donor quail origin, implying that both tissues had developed from the implanted somite. Presumably the muscle cells were derived from the myotome while the connective tissue cells were derived from the sclerotome of the transplanted quail somite. This emphasises the similarities between the mesenchymal cells of somatopleural origin of the limbs and the mesenchymal cells of sclerotomal origin in that both can form the connective tissue of muscle. Also, it further confirms that skeletal muscle is of myotomal origin as described by Solursh etal. (1987). Our results are also in agreement with those of Noden (1983) who transplanted quail somites into chick embryos in the cranial and upper cervical region. He found that the craniofacial and upper cervical skeletal muscles in the chick trace their embryonic ancestry to the paraxial mesoderm and that those muscles in the cranial region of the neck that have attachment onto the vertebral column and occipital bones have connective tissue that is derived from the somites. However, Noden (1983) limited his study to the removal of somite no.7 as the most caudal somite removed which is more rostral than any of the somites removed in our study. Our results, therefore, extend the work of Noden (1983) to include all of the cervical region.

One intriguing aspect of the results relates to the band of tissue between the neural arches that will form either ligament or muscle. Only in a few embryos were there instances of all the cells in this region being from one somite (i.e. all quail), as, usually, the cells between the neural arches were a mixture of chick and quail indicating that they came from at least two somites. Significantly, the distribution of quail cells within this mixture did not often extend beyond the boundary of the neural arches and was confined between them. If the distribution of the myotome cells is based on a repetitive sequence of the vertebral column, as the sclerotome and dermatome have been shown to be, then quail cells might also be expected to be found between other adjacent neural arches. In this study, quail cells were found to span the gap between only two consecutive neural arches where they were invariably mixed with chick cells. There was not a reciprocal distribution of quail cells between other neural arches. This suggests that the distribution of the quail cells was less than expected as it was less than the equivalent chick distribution. This discrepancy is difficult to explain other than to suggest that the myotome cells did not fully integrate into the embryo.

In general, the results of this study show that the somitic cells stay as a clearly defined group as they disperse from the somite and migrate to form a variety of tissues. All of the tissues within a segment (defined by a pair of somites) of the embryo are formed by the cells from the somites except for the notochord, neural tissues, and epidermis (and presumably the lining of the gut which was not considered in this study). The neural tissues consist of the neural tube, the neurons of the dorsal root ganglion, the developing neurites from the neural tube and the dorsal root ganglion, and their accompanying Schwann cells. However, it must be emphasised that all of the connective tissue surrounding the neural tissue develops from the somites. These results suggest that the somite cells disperse within the segment, surround, or are invaded by, the developing neural tissues and contribute to the formation of the various other tissues. The precision of the morphogenesis of the somitic cells in forming these tissues and the intricacy involved in their integration with similar tissues from adjacent segments to form longitudinal structures connecting several segments of the body, strongly suggest that the somitic cells differentiate in response to cues they receive from the local environment during their migration. This is well exemplified by the distribution of quail cells within segments of the endothelium of small blood vessels both in the dermis and within the developing vertebral column. Such a suggestion is also supported by evidence that the somitic cells are not completely committed at the time of transplantation. For example, Gallera (1966) showed that the covering layers of endoderm and ectoderm are required for proper differentiation of the sclerotome and dermatome and more recent work by Aoyama and Asamoto (1988) has shown that the dorsal-ventral axis of the somites is only just being determined at this time. Such a mechanism of differentiation in which final determination is based on environmental cues would be in agreement with the results from other studies of this topic including neural crest cell differentiation (Perris et al. 1988; Bronner-Fraser and Lallier, 1988; Le Douarin, 1988), growth cone migration (Eisen, 1988) and general development of the zebra fish embryo (Kimmel and Warga, 1986, 1987, 1988).

This mechanism of response to environmental cues would seem to be in sharp contrast to the differentiation of the nephrotome, which forms from the intermediate mesoderm just lateral to the somite. In this study, nephric tissue was found forming at sites normally reserved for the vertebral column. Presumably, at the time of the transplant, some intermediate mesoderm was inadvertently transplanted along with the quail somite (in cases of developing quail nephrotome) or some intermediate mesoderm was inadvertently displaced medially along with the quail somite (in cases of developing chick nephrotome). This tissue from the intermediate mesoderm subsequently developed into nephrotome although it had been in an abnormal environment for normal nephrotome development. This demonstrates that these cells were committed to forming the nephric tissue at the time just after segmentation when transplantation occurred and were not subsequently influenced by the surrounding environment.

The results also clearly show that there is little mixing of the cells from a somite with similar cells from adjacent somites even during their migration. There were clearly defined limits to the boundaries of the quail cell distribution in all tissues, although these were less clear in the myotome portion than in the sclerotome. At the boundaries between similar tissues formed separately from chick and quail cells (e.g. parts of the vertebral body), there was some slight mixing of chick and quail cells but not sufficient to destroy the boundary line. These results support the finding made by Stern and Keynes (1987) that sclerotome cells from like halves mix with each other while those from unlike halves do not. At the boundaries between the migrating sclerotome cells from adjacent somites, adjacent sclerotome halves will be dissimilar. A similar relationship might also exist between cells from the two sclerotome halves present within the same somite. It is possible that the cells from the rostral sclerotome half of the transplanted quail somite do not mix during their migration with cells from the adjoining caudal sclerotome half. Unfortunately, this study did not allow us to examine this possibility but it is supported by additional results from our own recent work (Bagnall and Sanders, unpublished) using peanut agglutinin as a marker for caudal sclerotome cells (Stern et al. 1986) which have suggested that rostral and caudal sclerotome halves from the same somite do not mix during migration. This emphasises that sclerotome cells remain in their segmental position and do not mix with adjacent segments and provides further support for the importance of parasegments and their influence as a basic domain of genetic control in the developing embryo as described by Lawrence (1988). This lack of mixing between cells from the adjacent somites means that any segmental pattern developed by the somites is also maintained during the subsequent migration and differentiation of the cells. This might provide a means by which the segmental nature of structures such as the peripheral nervous system might be maintained. These results disagree with the suggestion made by Verbout (1985), in opposition to the theory of resegmentation in vertebral development, that the sclerotome cells from adjacent somites might mix freely during their migration.

This project was funded by grants from the Natural Science and Engineering Research Council of Canada (K.M.B.) and the Medical Research Council of Canada (E.J.S.).

Aoyama
,
H.
and
Asamoto
,
K.
(
1988
).
Determination of somite cells: independence of cell differentiation and morphogenesis
.
Development
104
,
15
28
.
Bagnall
,
K.
,
Higgins
,
S.
and
Sanders
,
E. J.
(
1988
).
The contribution made by a single somite to the vertebral column: experimental evidence in support of resegmentation using the chick-quail chimaera model
.
Development
103
,
69
85
.
Balaban
,
E.
,
Teillet
,
M.
and
Le Douarin
,
N.
(
1988
).
Application of the quail-chick chimera system to the study of brain development and behavior
.
Science
241
,
1339
1342
.
Beresford
,
B.
(
1983
).
Brachial muscles in the chick embryo: the fate of individual somites
.
J. Embryol. exp. Morph
.
77
,
99
116
.
Bronner-Fraser
,
M.
and
Lallier
,
T.
(
1988
).
A monoclonal antibody against a laminin-heparan sulfate proteoglycan complex perturbs cranial neural crest cell migration in vivo
.
J. Cell. Biol
.
106
,
1321
1330
.
Bunge
,
M.
,
Wood
,
P.
,
Tynan
,
L.
,
Bates
,
M.
and
Sanes
,
J.
(
1989
).
Perineurium originates from fibroblasts: demonstration in vitro with a retroviral marker
.
Science
243
,
229
231
.
Chevallier
,
A.
,
Kieny
,
M.
and
Mauger
,
A.
(
1977
).
Limb-somite relationship: origin of the limb musculature
.
J. Embryol. exp. Morph
.
41
,
245
258
.
Christ
,
B.
,
Jacob
,
H.
and
Jacob
,
M.
(
1977
).
Experimental analysis of the origin of the wing musculature in avian embryos
.
Anat. Embryol
.
150
,
171
186
.
Christ
,
B.
,
Jacob
,
M.
and
Jacob
,
H.
(
1983
).
On the origin and development of the ventrolateral abdominal muscles in the avian embryo, an experimental and ultrastructural study
.
Anat. Embryol
.
160
,
87
101
.
Eisen
,
J.
(
1988
).
Growth cone guidance and pathway formation
.
Trends Neurosci
.
11
,
333
336
.
Gallera
,
J.
(
1966
).
Mise en evidence du role de 1’ectoblaste dans la differenciation des somites chez les oiseaux
.
Revue Suisse de Zoologie
.
73
,
492
503
.
Hamburger
,
V.
and
Hamilton
,
H.
(
1951
).
A series of normal stages in the development of the chick embryo
.
J. Morph
.
88
,
49
92
.
Haninec
,
P.
(
1988
).
Study of the origin of connective tissue sheaths of peripheral nerves in the limb of avian embryos
.
Anat. Embryol
.
178
,
553
557
.
Jacobson
,
A.
(
1988
).
Somitomeres: mesodermal segments of vertebrate embryos
.
Development
104
Supplement
,
209
220
.
Keynes
,
R.
and
Stern
,
C.
(
1988
).
Mechanisms of vertebrate segmentation
.
Development
103
,
413
429
.
Keynes
,
R.
,
Stirling
,
R.
,
Stern
,
C.
and
Summerbell
,
D.
(
1987
).
The specificity of motor innervation of the chick wing does not depend upon the segmental origin of muscles
.
Development
99
,
565
575
.
Kimmel
,
C.
and
Warga
,
R.
(
1986
).
Tissue-specific cell lineages originate in the gastrula of the zebrafish
.
Science
231
,
365
368
.
Kimmel
,
C.
and
Warga
,
R.
(
1987
).
Cell lineages generating axial muscle in the zebrafish embryo
.
Nature, Lond
.
327
,
234
237
.
Kimmel
,
C.
and
Warga
,
R.
(
1988
).
Cell lineage and developmental potential of cells in the zebrafish embryo
.
Trends Genet.
4
,
68
74
.
Kinutani
,
M.
and
Le Douarin
,
N.
(
1985
).
Avian spinal cord chimeras. I. Hatching ability and post-hatching survival in horno- and heterospecific chimeras
.
Devi Biol
.
111
,
243
255
.
Kirby
,
M.
,
Gale
,
T.
and
Stewart
,
D.
(
1983
).
Neural crest cells contribute to aorticopulmonary septation
.
Science
220
,
1059
1061
.
Kirby
,
M.
,
Turnage
,
K.
and
Hays
,
B.
(
1985
).
Characterization of conotruncal malformations following ablation of ‘cardiac’ neural crest
.
Anat. Rec
.
213
,
87
93
.
Lance-Jones
,
C.
(
1988
).
The effect of somite manipulation on the development of motoneuron projection patterns in the embryonic chick hindlimb
.
Devi Biol
.
126
,
408
419
.
Lawrence
,
P.
(
1988
).
The present status of the parasegment
.
Development
104
Supplement
,
61
65
.
Le Douarin
,
N.
(
1973
).
A biological cell labelling technique and its use in experimental embryology
.
Devi Biol
.
30
,
217
222
.
Le Douarin
,
N.
(
1982
).
The Neural Crest
.
Cambridge University Press
:
Cambridge
.
Le Douarin
,
N.
(
1988
).
Embryonic chimeras: a tool for studying the development of the nervous and immune systems
.
Proc. R. Soc. Lond. B
235
,
1
17
.
Newgreen
,
D.
,
Scheel
,
M.
and
Kastner
,
V.
(
1985
).
Morphogenesis of sclerotome and neural crest in avian embryos
.
Cell Tiss. Res
.
244
,
299
313
.
Noden
,
D.
(
1983
).
The embryonic origins of avian cephalic and cervical muscles and associated connective tissues
.
Amer. J. Anat
168
,
257
276
.
Perris
,
R.
,
Von Boxberg
,
Y.
and
Lofberg
,
J.
(
1988
).
Local embryonic matrices determine region-specific phenotypes in neural crest cells
.
Science
241
,
86
88
.
Phillips
,
M.
,
Waldo
,
K.
and
Kirby
,
M.
(
1987
).
Neural crest ablation does not alter pulmonary vein development in the chick embryo
.
Anat. Rec
.
223
,
292
298
.
Sanders
,
E. J.
(
1986
).
A comparison of the adhesiveness of somitic cells from chick and quail embryos
.
In Somites in Developing Embryos
, (ed.
R.
Bellairs
,
D.
Ede
and
J.
Lash
), pp.
191
200
.
New York
,
Plenum Press
.
Solursh
,
M.
,
Drake
,
C.
and
Meier
,
S.
(
1987
).
The migration of myogenic cells from the somites at the wing level in avian embryos
.
Devi Biol
.
121
,
389
396
.
Stern
,
C.
and
Keynes
,
R.
(
1987
).
Interactions between somite cells: the formation and maintenance of segment boundaries in the chick embryo
.
Development
99
,
261
272
.
Stern
,
C.
,
Sisodiya
,
S.
and
Keynes
,
R.
(
1986
).
Interactions between neurites and somite cells: inhibition and stimulation of nerve growth in the chick embryo
.
J. Embryol. exp. Morph
.
91
,
209
226
.
Verbout
,
A.
(
1985
).
The development of the vertebral column
.
Advances in Anatomy, Embryology and Cell Biology
90
,
New York
,
Springer-Verlag
.
Willier
,
B.
,
Weiss
,
P.
and
Hamburger
,
V.
(
1955
).
Analysis of Development
.
Philadelphia
,
W.B. Saunders Co
.