SEARCH

SEARCH BY CITATION

Keywords:

  • chick;
  • somites;
  • somitocoele;
  • vertebrate column;
  • joints;
  • arthrotome

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Somitocoele cells previously have been shown to form the proximal part of the ribs, the intervertebral discs, and the intervertebral joints (synovial joints). To determine whether the somitocoele cells are necessary for the development of axial skeleton joints, we microsurgically ablated the somitocoele cells in epithelial somites of 2-day-old chick embryos. The operated embryos were analyzed after whole-mount skeletal preparations and in sections. Removal of the somitocoele cells led to two major outcomes: (1) Intervertebral joints failed to develop and resulted in the fusion of the superior articular process and the inferior articular process; (2) Adjacent vertebral bodies fused and lacked the intervertebral disc. These results demonstrate that somitocoele cells specifically give rise to intervertebral joints and discs. Furthermore, these results suggest that neighboring sclerotome cells cannot adapt to form vertebral joints in the absence of the somitocoele compartment. Thus, we provide for the first time experimental evidence for the existence of a joint forming compartment in the somites, which we term the “arthrotome.” Developmental Dynamics 234:48–53, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

In vertebrate embryos, somites bud off from the segmental plate mesoderm in a craniocaudal direction (Christ and Ordahl, 1995; Gossler and Hrabe de Angelis, 1998). The somites represent the first sign of overt segmentation along the rostrocaudal axis in the chick embryo.

The newly formed somite consists of at least two discrete groups of cells, with differing morphological characteristics and developmental fate (Christ and Wilting, 1992; Ordahl and Le Douarin, 1992). The cells constituting the outer spherical structure are epithelially organized, in contrast to the loose mesenchymal cells present within the central core of the somite called somitocoele cells (Williams, 1910; Trelstad et al., 1967; Solursh et al., 1979). During somite maturation, there is a significant increase in the density of somitocoele cells, which is dependent on the age of the embryo, whereas the density of the cells in the epithelial wall remains constant. The total number of somitocoele cells depends on the segment level and developmental stage. The central core of an epithelial somite at Hamburger and Hamilton (HH) stage 8 and 15 contains approximately 300 and 800 somitocoele cells, respectively (Bagnall and Berdan, 1994). Some cells of the epithelial sphere can contribute to the somitocoele population by undergoing an epithelial to mesenchymal transition, whereas a reciprocal transformation has not been detected (Wong et al., 1993).

Cells of the somitic epithelial wall and cells of the somitocoele differ in not only cell morphology but also in their prospective fate. The somitocoele cells represent a compartment separated from the surrounding epithelial wall with respect to cell communication (Bagnall et al., 1992). Furthermore, the somitocoele population is more susceptible to apoptosis than the epithelially organized cells (Christ et al., 1972; Hirano et al., 1995). Cell death is drastically increased within the somitocoele when the somites develop without the neighboring notochord (Hirano et al., 1995).

Somite maturation is characterized by the formation of a dorsal and a ventral compartment. The dorsal epithelial region of the somite develops into the dermomyotome, which in turn gives rise to dermis of the back and musculature of the body wall and the limbs (reviewed in Scaal and Christ, 2004). The remaining ventral epithelial region of the somite and the somitocoele cells form the sclerotome. The first and most significant morphological characteristic of sclerotome formation in the avian embryo is an epithelial–mesenchymal transition (EMT) in the ventral epithelial half of the somite, controlled by notochord derived signals such as Shh and Noggin (reviewed in Dockter, 2000). Thus, the somitocoele cells maintain their mesenchymal organization throughout this maturation process, from somite formation up to sclerotome development (reviewed in Christ et al., 2000, 2004). The somitocoele cells integrate into the mesenchymal cells derived from the ventral epithelial wall (Williams, 1910). Later on, most somitocoele cells are localized in the caudal half of the sclerotome close to the intervertebral fissure (von Ebner's fissure), which represents the first sign of resegmentation (Christ et al., 2000, 2004).

Two adjacent vertebrae are connected by intervertebral joints and the intervertebral discs. The intervertebral joints (also called zygapophysial joints) are synovial joints between the inferior articular process of superior vertebrae and the superior articular process of inferior vertebrae. These joints allow the spine to bend and twist, while at the same time limiting movement. The intervertebral discs link two adjacent vertebrae through the vertebral bodies. In mammals, each disc is composed of an external anulus fibrosus, which surrounds the internal nucleus pulposus. The anulus fibrosus is composed of lamellae built of thick linearly organized collagen fibers. Cells in the inner anulus have chondrocyte morphology (Roberts et al., 1991), whereas in the outer anulus, they resemble tendon cells (Postacchini et al., 1984). In mammals, the nucleus pulposus has been suggested to originate from the notochord (Walmsley, 1953). The nucleus pulposus contains high levels of aggrecan that attracts water and permits it to absorb force and dissipate it uniformly to the surrounding anulus fibrosus. In humans, the notochord cells die within the first few years of life (Walmsley, 1953), accompanied by a gradual decrease in the water content of the nucleus pulposus.

Using the chick/quail grafting technique, Huang et al. (1994, 1996) have shown that somitocoele cells in the somites of the prospective thoracic region mainly contribute to the ribs, the articular surface of the intervertebral joints, and the peripheral parts of the intervertebral discs. This observation raises the question of whether the somitocoele cells represent a specific joint forming compartment. To investigate this issue, we have removed the somitocoele cells of the epithelial somites of 2-day-old chick embryos (HH stage 12–16) and implanted inert beads to prevent their regeneration from the surrounding epithelial tissue (Wong et al., 1993). After a reincubation period of 6 days, we found that the intervertebral joint and disc of the vertebrae were missing, resulting in the fusion of the two adjacent articular processes and adjacent vertebral bodies. Thus, we demonstrate for the first time that somitocoele cells represent a joint forming compartment within the somite, which we have called the “arthrotome.” Moreover, our results indicate that neighboring sclerotome cells outside the arthrotome cannot adapt to form structures specific to somitocoele cells.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

To investigate the developmental fate of the somitocoele compartment, somitocoele cells of the epithelial somites II to IV (according to Christ and Ordahl, 1995) were microsurgically removed at HH stage 12–16 (n = 10; Fig. 1A). After a reincubation period of 6 days, the operated embryos were fixed and prepared for whole-mount skeletal preparations and frontal sections. Subsequent analysis of chick embryos in whole-mount skeletal preparations showed no effect in the vertebral column or other structures (data not shown). This result can be explained by the following: (1) the development of the vertebral joints and intervertebral discs does not depend on the somitocoele cells or (2) that the somitocoele population has regenerated from cells left after the operation or through the recruitment of mesenchymal cells from the epithelial wall (Wong et al., 1993).

thumbnail image

Figure 1. Somitocoele cells give rise to vertebral joints. A: Schematic illustration of removal of somitocoele cells. The somitocoele cells were aspirated through a mouth-controlled micropipette after opening the dorsal epithelial wall of the somite along with the ectoderm. B: An Affigel bead was implanted to replace the somitocoele cells. C: Whole-mount skeletal preparation of an 8-day-old chick embryo at thoracic level, dorsal view. Somitocoele cells were removed from three consecutive epithelial somites. Note the lack of intervertebral joints and the fusion of three adjacent articular processes in the operated region (red arrows). On the left control side, intervertebral joints (ij) are well formed. D: Frontal microtome sections of the specimen in C shows the missing of intervertebral joints and the fusions of the transverse processes of the three consecutive adjacent vertebrae (red arrows). The implanted beads are not visible in this section. E: Whole-mount skeletal preparation of an 8-day-old chick embryo at thoracic level, ventral view. Somitocoele cells of a single epithelial somite from a 2-day chick embryo were replaced with an Affigel bead. Note the lack of intervertebral disc and fusion of the two adjacent vertebral bodies (red arrow). F: Frontal microtome section showing the absence of the intervertebral disc. The blue region represents the Affigel bead at the site of the lacking intervertebral disc, which was placed at the time of microsurgery (arrowhead). aap, anterior articular process; ab, Affigel bead; es, epithelial somite; et, ectoderm; ij, intervertebral joint; ivd, intervertebral disc; na, neural arch; nc, notochord; nt, neural tube; pap, posterior articular process; R, Rib; sc, scapula; s, somitocoele; vb, vertebral body.

Download figure to PowerPoint

To prevent regeneration efficiently, a neutral Affigel bead was implanted into the somitocoele (Fig. 1A,B). Because the bead is spherical, a bead with appropriate size can occupy the entire central core of an epithelial somite. Therefore, considerable care was taken in selecting beads of appropriate size.

Unlike the sole removal of the somitocoele, the second type of procedure that involved the implantation of the inert bead led to a dramatic phenotype. All operated embryos showed an absence of intervertebral joints and/or fusion of adjacent articular processes (Table 1; Fig. 1C,D). Because the microsurgeries were performed at different stages, the operation regions in which intervertebral joints were affected were located at different thoracic levels.

Table 1. Overview of the Results of Replacement of Somitocoele Cells by an Inert Beada
Operation stage (HH)Number of operations% cases showing absence of intervertebral joints% cases showing absence of intervertebral discs
  • a

    The 30 chick embryos were operated at day 2 (Hamburger and Hamilton [HH] stage 12–16) and reincubated for analysis until day 8. Results show missing intervertebral joints in all cases and lack of intervertebral discs in many, but not all cases.

12–1312100 (n = 12)25 (n = 3)
14–1513100 (n = 12)61 (n = 8)
165100 (n = 5)60 (n = 3)

Many operated chick embryos (n = 14/30) also lacked intervertebral discs, resulting in the fusion of the two adjacent vertebral bodies (Fig. 1F; Table 1). The success rate of the operation with respect to the development of the intervertebral discs depended on the operation stages: 25% of embryos (n = 3/12) operated at HH stage 12–13, 60% of embryos operated at HH stage 14–15 (n = 8/13), and at HH stage 16 (n = 3/5) lacked intervertebral discs. This finding might be due to a stronger proliferative activity of the epithelial somites at HH stage 12–13 compared with HH stage 14–16 (Williams, 1910).

Frontal sections of the specimens (Fig. 1D,F) confirmed the loss of intervertebral discs and fusion of adjacent vertebral bodies at the site of surgery. In some cases, the implanted bead was seen exactly at the level of the missing intervertebral disc. In a few sections, some ectopic fibroblastic cells were found slightly cranial to the original axial level of the intervertebral disc. This finding could be due to accidentally remaining somitocoele cells that were pushed to a more cranial level by the bead implantation.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

To date, there have been very few studies focused on the development of the mesenchymal core of the epithelial somite, the somitocoele. In this study, we addressed the developmental fate of the somitocoele compartment in the avian embryo by microsurgically removing the somitocoele cells and physically preventing somitocoele cell regeneration by insertion of an inert bead. We show for the first time that the mesenchymal cells of the somitocoele specifically give rise to joint forming cells between the vertebrae.

Studies using autoradiography (Langman and Nelson, 1968; Summerbell et al., 1986) have shown that the somitocoele cells, like the cells of the outer epithelial sphere, have a high mitotic capacity. Single-cell labeling experiments with fluorescent dye (Wong et al., 1993) revealed that the somitocoele recruits additional cells from the epithelial wall through an epithelial to mesenchymal transition. Nevertheless, contribution of cells from the epithelial wall to the somitocoele seems to be of relatively minor importance (Wong et al., 1993). However, somitocoele cell recruitment is likely to be the reason why embryos developed normally after simple ablation of the somitocoele cells. Implantation of a neutral Affigel bead in the somitocoele after cell ablation was able to physically prevent regeneration and enabled the operated embryos to develop devoid of somitocoele cells.

We demonstrate for the first time that, in the absence of somitocoele cells, vertebral joint formation is severely disturbed. Interestingly, 100% of the operated embryos (n = 30/30) lack the intervertebral joints (Articulationes zygapophysiales), whereas only 46% (n = 14/30) showed an absence of intervertebral discs (Disci intervertebrales). The loss of intervertebral discs seemed to be stage dependent. More than 60% of the chick embryos lacked intervertebral discs if the microsurgery was performed at later stages (HH stage 14–16), whereas only 25% of the operated embryos lacked intervertebral discs (n = 3/12) if the operation was performed at earlier stages (HH stage 12–13). The rather high numbers of failed experiments could be due to remaining cells located on the bottom of the somitocoele, which had been missed during the operation. These cells had more time to proliferate in the epithelial somites at HH stage 12–13 than at HH stage 14–16. According to the histological observation by Williams (1910), the somites at HH stage 14–16 compartmentalize into dermomyotome and sclerotome very soon after their formation. If the remaining cells were sufficient in number to take over the function of the somitocoele cells, the intervertebral disc could be formed; if only very few cells were left, they could not form a complete intervertebral disc and differentiated only into a small population of fibroblastic cells. However, the reason why potential remaining cells participated in disc formation, but never in intervertebral joint formation, remains obscure. We propose a model based on the different distribution of disc and joint forming cells to explain these findings (Fig. 2). We suggest that there are two regions within the somitocoele: a ventral and a dorsal region. We suggest that cells in the dorsal region develop into intervertebral joints, whereas ventrally located cells form the intervertebral disc. In our experiments, the micropipette for somitocoele cell removal was inserted into the somite from dorsal, so that dorsal somitocoele cells were always completely removed, whereas ventral cells might sometimes have remained on the bottom of the somitocoele after bead insertion. In some cases, these remaining cells could have differentiated into an intervertebral disc. This possibility could explain the lower incidence of loss of intervertebral discs.

thumbnail image

Figure 2. Schematic illustration of somitocoele cell location. The discrimination between ventral (green) and dorsal (red) somitocoele cells is hypothetical. A: Transverse section of an epithelial somite. Dorsal is to the top. The somitocoele cells can be subdivided into a ventral subpopulation (red dots) and a dorsal subpopulation (green dots). B: Coronal section of dermomyotome and sclerotome of a mature somite. Cranial is to the top. After somite compartmentalization, the somitocoele cells are located adjacent to the intervertebral (von Ebner's) fissure (broken line) in the caudal compartment of the sclerotome (Huang et al. 1994; 1996). C: Lateral view of two adjacent vertebrae. Cranial is to the top. The ventral subpopulation of the somitocoele cells (red) contributes to the intervertebral disc (id), and the dorsal subpopulation (green) contributes to the intervertebral joint (green). aap, anterior articular process; ao, aorta; dm, dermomyotome; id, intervertebral disc; im, intermediate mesoderm; na, neural arch; no, notochord; nt, neural tube; pap, posterior articular process; scl, sclerotome; sp, spinous process; vb, vertebral body.

Download figure to PowerPoint

Differentiation of the somite results in the formation of additional compartments and is under the control of signals emanating from the adjacent structures. Dorsoventral compartmentalization of the somite leads to a dorsally located dermomyotome and ventrally located sclerotome (reviewed by Christ and Ordahl, 1995). The dermomyotome lies directly beneath the surface ectoderm. It subsequently forms a layer of primary muscle cells, the myotome, located between the dermomyotome and the sclerotome. The myotome gives rise to the skeletal muscle of the body wall. Moreover, the dermomyotome gives rise to the skeletal muscle of the limbs, the back dermis, endothelium, and cartilage of the scapula (reviewed in Scaal and Christ, 2004). Sclerotome development is characterized by the formation of three subcompartments, giving rise to different parts of the axial skeleton and ribs (reviewed in Christ et al., 2000, 2004). The lateral sclerotome gives rise to the laminae and pedicles of the neural arches and to the ribs. Its development depends on signals form the notochord and from the myotome (Huang et al., 2003). 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. Cells from the dorsomedial aspect of the sclerotome migrate into the space between the roof plate of the neural tube and the dermis (Monsoro-Burq et al., 1994). According to Halata et al. (1990), the sclerotome cells surrounding the spinal cord can be considered to form a meningotome, as these cells give rise to the meninges of the spinal cord. Another subset within the sclerotome is the syndetome, which has been proposed by Brent et al. (2003). It begins to develop at the cranial and caudal borders of the sclerotome just as the myotome forms. Its cells express scleraxis (Cserjesi et al., 1995; Brown et al., 1999), a specific marker for tendons and ligaments (Schweitzer et al., 2001), and give rise to epaxial tendons. Thus, different compartments of the sclerotome could be related to different structures of the definitive skeleton.

Our results show that the cells of the somitocoele represent another sclerotomal subdomain, an arthrotome, which is a specific compartment for vertebral joints in the chick embryo. An important aspect of this work is that the removal of the somitocoele did not interfere with the development of any other part of the vertebrae other than the joints. This result has an impact in resolving the molecular control of joint formation. To date, very few genes have been reported to be expressed in the somitocoele. In somite stage III of 2-day-old chick embryos, the prospective sclerotomal cells in the ventral wall of the epithelial somite are Pax1-positive (Ebensperger et al., 1995; Muller et al., 1996). The somitocoele cells are positive for both Pax1 and Pax9 (Huang et al., 1996). Pax1-deficient mice lack vertebral bodies and intervertebral discs (Wallin et al., 1994). Pax9 can only partially compensate the defects observed in Pax1 mutant mice. Furthermore, Pax1/9 double mutants possess neural arches but completely lack vertebral bodies, intervertebral discs, and proximal ribs (Peters et al., 1999). Nevertheless, the intervertebral joints still form in mice lacking Pax1 and Pax9. One possible explanation for these results could be due to the proposed heterogeneity within the somitocoele. Pax1 and Pax9 have been shown to control both cell proliferation and to a lesser extent cell survival (Peters et al., 1999).

Uncx 4.1 is expressed in the caudal somite halves of the newly formed somites but later becomes restricted to the somitocoele and the ventral epithelial wall (Schragle et al., 2004). Uncx 4.1 mutant mice lack the proximal part of the ribs and the pedicle of the neural arch (Leitges et al., 2000). However, in these mutants, the vertebral joints develop normally.

Thus, the deletion of three marker genes expressed in the somitocoele (Pax1, Pax9, and Uncx4.1) resulted in normal intervertebral joints and contrasts the observation of our study. Conversely, after somitocoele removal, none of our experimental data show missing of vertebral bodies, pedicles, and proximal parts of the ribs. This finding argues that these genes are not involved in the specification of somitocoele cells to a joint-forming fate. Further studies are under way in our laboratory to identify genes that regulate somitocoele development.

Taken together, our findings demonstrate that the mesenchymal somitocoele cells specifically give rise to intervertebral joints and intervertebral discs and that neighboring sclerotome cells cannot adapt to form somitocoele-specific structures if outside the somitocoele compartment. Therefore, we postulate that somitocoele cells represent a sclerotomal subdomain that specifically gives rise to intervertebral joints, the arthrotome.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Embryos

Eggs of White Leghorn chicken (Gallus gallus) were incubated at 37.8°C and 80% humidity. Staging was performed according to Hamburger and Hamilton (1951).

Preparation of Affigel Beads

Affigel beads (Bio-Rad) with 100-micron diameter were washed twice with phosphate buffer, and incubated in Locke's solution (Locke and Rosenheim, 1907) for 1 hr at 4°C.

Removal of Somitocoele Cells

The somitocoele cells of one to three consecutive epithelial somites HH stage 12–16 (somite II to IV according to Christ and Ordahl, 1995) were removed. Two slightly different procedures were performed. In both, the dorsal epithelial portion of the somite was opened, and the somitocoele cells were aspirated with a mouth-controlled micropipette without disturbing the ventral epithelial wall of the somite. In the first protocol, the dorsal epithelial cap was immediately replaced after somitocoele cell removal. In the second type of procedure, a neutral Affigel bead (ca. 100 microns in size) was implanted in place of the somitocoele cells. The bead was then covered with the dorsal epithelial somitic tissue. Embryos from both procedures were reincubated for a period of 6 days (8 days of total incubation).

Skeletal Preparations

Whole-mount Alcian blue staining was performed to investigate the skeletal pattern of the axial skeleton. Specimens were first stained with 0.015% Alcian blue in 80% ethanol and 20% acetic acid for 1–3 days, fixed, and dehydrated in ethanol for 1 day, and cleared and stored in 100% methylsalicylate (Kant and Goldstein, 1999). This method provides the possibility of consecutive paraffin sectioning followed by staining procedures.

Immunohistochemistry

After photographing the skeletal pattern, the operated embryos were serially sectioned at 8 μm. The sections were stained with a monoclonal anti-desmin antibody (Sigma, Deisenhofen, Germany) to identify differentiated muscle cells. The second antibody was peroxidase-conjugated goat anti-mouse antibody (Sigma) and diaminobenzidine as the chromogen. The color reaction leads to a brown signal for muscle cells. Finally, the sections were counterstained with Nuclear Fast Red (Sigma).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

We thank Mrs. Gimbel, Mrs. Koschny, and Mr. Frank for their excellent technical assistance. B.C. and M.S. were funded by the DFG.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES