SEARCH

SEARCH BY CITATION

Keywords:

  • somitogenesis;
  • sclerotome;
  • vertebra;
  • bHLH factor;
  • Pax genes

Abstract

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

Mesp2 and Paraxis are basic helix–loop–helix (bHLH) -type transcription factors coexpressed in the presomitic mesoderm (PSM) and are required for normal somite formation. Here, we show that Mesp2/Paraxis double-null mice exhibit a distinct phenotype unexpected from either Mesp2 or Paraxis single-null mice. In the posterior region of the body, most of the skeletal components of both the vertebral body and neural arches are severely reduced and only a rudimental lamina and ribs remain, indicating a strong genetic interaction in the sclerotomal cell lineage. However, yeast two-hybrid analyses revealed no direct interaction between Mesp2 and Paraxis. The Mesp2/Paraxis double-null embryo has caudalized somites, revealed by expanded Uncx4.1 expression pattern observed in the Mesp2-null embryo, but the expression level of Uncx4.1 was significantly decreased in mature somites, indicative of hypoplasia of lateral sclerotome derivatives. By focusing on vertebral column formation, we found that expressions of Pax1, Nkx3.1, and Bapx1 are regulated by Paraxis and that Pax9 expression was severely affected in the Mesp2/Paraxis double-null embryo. Furthermore, the expression of Pax3, a crucial factor for hypaxial muscle differentiation, is regulated by both Mesp2 and Paraxis in the anteriormost PSM and nascent somite region. The present data strongly suggest that patterning events by bHLH-type transcription factors have deep impacts on regional chondrogenic and myogenic differentiation of somitic cells, mainly by means of control of Pax genes. Developmental Dynamics 236:1484–1494, 2007. © 2007 Wiley-Liss, Inc.


INTRODUCTION

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

Somitogenesis consists of several steps involved in the generation and segmentation of paraxial mesoderm, followed by cellular differentiation, depending on the position of cells within somites (Pourquie, 2003). Before segmental border formation, the presomitic mesoderm derived from primitive steak or tail bud region acquires their periodicity and rostocaudal polarity as a future somite compartment. After segmental border formation and epithelialization, cells located in the dorsal region start to differentiate to dermatomal and myotomal lineages, while cells in ventromedial region de-epithelialize and take on a sclerotomal cell fate for the future axial skeleton. These morphogenetic changes are accompanied with the sequential expression of a great number of genes. Roles of these genes have been revealed by studies in mouse, chick, and zebrafish embryos, by using molecular genetic and experimental embryological techniques.

Mesp1 and Mesp2 are members of the Mesp family, a group of basic helix–loop–helix (bHLH) transcription factors, which is expressed in the anterior presomitic mesoderm (PSM) just before somite formation (Saga et al., 1997). While the Mesp1-null mouse shows no phenotype in somitogenesis, the Mesp2-null mouse shows defects in segment border formation and the rostrocaudal patterning of a somite. Our recent genetic analyses have revealed that Mesp2 plays a central role in the activation of rostral genes Notch2, FGFR1, Cer1, EphA4, Tbx18, and in suppression of caudal genes such as a Notch ligand Dll1 and the homeobox transcription factor Uncx4.1 (Takahashi et al., 2000, 2003; Nakajima et al., 2006). As a result, Mesp2 specifies rostral half property at the expense of caudal half property in the somites. So far, the effects of Mesp2 on skeletal development have been attributed to rostrocaudal patterning and not to general chondrogenic, osteogenic, or myogenic differentiation of somitic cells. While spatially disorganized dermomyotomes are formed, all myogenic factors are expressed in the Mesp2-null embryo, indicating that the general myogenic program is unaffected in the absence of Mesp2 (Saga et al., 1997). However, our chimera analysis strongly suggests that the Mesp family genes are essential for epithelialization of somitic mesodermal cells (Kitajima et al., 2000; Takahashi et al., 2005).

Paraxis is also a member of the bHLH-type family of transcription factors and is expressed in the anterior two thirds of the presomitic mesoderm, throughout the entire forming epithelial somite, and later the epithelial dermomyotome (Burgess et al., 1995; Barnes et al., 1997; Tseng and Jamrich, 2004). The Paraxis-null embryo has segmented somites, but the newly formed somites are not fully epithelialized and fail to form a normal epithelial dermomyotome. This results in the abnormal patterning of many paraxial mesoderm-derived tissues, including the chondrocranium, the axial skeleton, and the ribs (Burgess et al., 1996). In addition, defects in the development of somite-derived nonmigratory hypaxial muscles have been reported (Wilson-Rawls et al., 1999). Although the rostrocaudal polarity is generated in the presomitic mesoderm of paraxis-null embryos, the role of Paraxis in epithelialization is required for the maintenance of the rostrocaudal polarity of somites and normal resegmentation of the sclerotome (Johnson et al., 2001). However, the genetic interactions between Paraxis and other somite-expressed genes, required for the normal formation of the vertebral column, are not well understood.

Because both Mesp2 and Paraxis are members of bHLH-type transcription factors family, their expression domains in the presomitic mesoderm overlap and the functions of both genes are implicated in somite patterning and mesenchymal–epithelial transition, it is very likely that these genes have a cooperative function. To address this question, we have generated Mesp2/Paraxis double-knockout mice. The double-knockout mice exhibited a distinct phenotype that was not expected by the parental single-knockout mice. Foremost, axial skeletal components were extremely reduced in the posterior body and only rudimental lamina of neural arches and ribs remained, which is different from the patterning defects exhibited in the single knockouts. We show that this is the result of a loss of cartilage differentiation from sclerotomal cells. In addition, myotomal components are also severely reduced. Consistent with the perinatal phenotype, we found that the expression of Pax family genes were consistently reduced or lost in the double-knockout embryo. Thus, we conclude that Mesp2 and Paraxis may play a cooperative role to initiate differentiation of both myotomal and sclerotomal cells by regulating Pax family genes.

RESULTS

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

Mesp2 and Paraxis Exhibit Strong Genetic Interaction

An initial indication of a genetic interaction between Mesp2 and Paraxis was obtained by the analysis of skeletal specimens of the double-knockout mouse stained for cartilage and bone. The embryonic day (E) 18.5 wild-type control fetus exhibits a metameric pattern of vertebrae (neural arches and vertebral bodies) and ribs throughout the anterior/posterior (A/P) axis of the body (Fig. 1A,E). The E18.5 fetuses of Mesp2 and Paraxis single-knockout mice show their distinct phenotypes in axial skeletal morphology. The Mesp2-null fetus has vertebrae with completely fused pedicles of neural arches (Fig. 1B) and shows a proximal fusion of ribs as previously described (Saga et al., 1997; Takahashi et al., 2000). In a ventral view of the vertebrae, the vertebral bodies are segmented in the Mesp2-null fetus, although the spatial patterning is somewhat disorganized (Fig. 1F). Thus, Mesp2-null fetus exhibit ectopic formation of skeletal elements derived from the caudal-half of somites, but shows no sign of reduction or loss of skeletal elements. In the Paraxis-null fetus, vertebral bodies are not normally formed at the ventral midline and dual ossification centers are observed in the thoracic and anterior lumbar region (Fig. 1G; Burgess et al., 1996). In addition, the cartilaginous precursors of vertebral bodies are not segmented but are fused along the A/P axis (Fig. 1G; Johnson et al., 2001). In the posterior lumbar region, only rudimental cord-like cartilages, widely separated to left and right sides, are seen instead of the vertebral bodies. A rudimental rod-like cartilage is also seen at the midline. The neural arches are segmented but the lower part of the pedicle is occasionally missing (Fig. 1C). This finding indicates that the Paraxis-null fetus cannot complete formation of the vertebral body at the ventral midline, but the neural arches are relatively normal. In the Mesp2/Paraxis double-null fetus, the skeletal defects are much more severe than the Paraxis-null fetus. In the lumbar region, most parts of the neural arches are not formed, except the dorsal cartilaginous elements corresponding to the lamina (Fig. 1D). From a ventral view, it is apparent that the vertebral bodies are spatially disorganized and hypoplastic in the thoracic and anterior lumbar region. In the posterior lumbar region, only a rudimental rod-like cartilage is observed at the midline (Fig. 1H). Thus, the Mesp2/Paraxis double-null fetus shows extensive loss of skeletal elements both in the vertebral body and neural arches not seen in either the Mesp2- or Paraxis-null fetuses.

thumbnail image

Figure 1. Mesp2/Paraxis double-null fetuses exhibit unexpected defects in vertebral morphology. A–P: Alcian blue/Alizarin red double-stained skeletal specimens at embryonic day (E) 18.5 (A–H) and E13.5 Alcian blue-stained cartilage specimens of lumbar vertebrae (I–P) in fetuses from Mesp2/Paraxis intercross. Left lateral view (A–D,I–L) and ventral view (E–H,M–P) of indicated genotypes. Note that the Mesp2/Paraxis double-null fetus shows extensive loss of pedicles of the neural arches (D,L) and enhancement of defects in vertebral body formation in Paraxis mutants (H,P). lm, lamina; p, pedicle; t, transverse process; o, ossification center in the vertebral body; na, neural arches; vb, vertebral body. The arrowhead indicates the rudimental rod-like cartilage.

Download figure to PowerPoint

To trace the origin of the observed skeletal defects, we analyzed early cartilage formation in E13.5 embryos using Alcian blue staining. We analyzed embryos in the same litter for the comparison. Cartilage formation of all components of the vertebrae and ribs was observed in the wild-type and Mesp2-null embryos, although the rostrocaudal pattern was perturbed in the Mesp2-null embryo (Fig. 1I,J,M,N). In the Paraxis-null embryo, formation of presumptive vertebral bodies was limited to the cervical, thoracic, and anterior lumbar regions (Fig. 1K,O). Cartilage formation of the neural arches was not largely affected in the Paraxis-null embryo. However, in the double-null embryo, cartilaginous precursors of vertebral bodies appeared more hypoplastic (Fig. 1P). In addition, most of the cartilaginous components of the neural arches were absent and only the precursors of laminae were seen (Fig. 1L). The extent of cartilage formation in both the forelimbs and hindlimbs was comparable among these genotypes, indicating that the above skeletal defects are not due to general developmental retardation. Examination of Alcian blue-stained histological sections also confirmed that cartilage formation was greatly reduced both in Paraxis-null and double-null embryos, and cell density appeared to be reduced especially in the double-null embryo (Fig. 4A–D). In summary, the Mesp2/Paraxis double-null mouse exhibits more severe defects in the axial skeleton than either single-null mice.

thumbnail image

Figure 4. Histological sections confirm the severe defects in vertebral column formation in the Mesp2/Paraxis double-null embryos. A–H: Alcian blue-stained (A–D) and X-gal stained (for detection of Uncx4.1-LacZ expressing cells; E–H) transverse sections at the lumbar region at embryonic day (E) 13.5. Note that Paraxis-null fetuses show defects in midline fusion of vertebral body primordia (C,G) and that vertebral column is rudimental in the double-null embryos (D,H). In wild-type and Mesp2-null embryos (E,F), the pedicle (p), transverse process (t), and part of vertebral body (vb) are stained. The staining outlining the notochord of Paraxis-null and Mesp2/Paraxis double-null embryos may represent very rudimental vertebral body as a cord-like structure (Fig. 1, arrowhead). Uncx4.1 is also expressed in the metanephric mesenchyme, as previously reported (Neidhardt et al., 1997). vb, vertebral body; n, notochord; p, pedicle; t, transverse process; lm, lamina; mt, metanephros.

Download figure to PowerPoint

Mesp2 and Paraxis Do Not Directly Interact With Each Other

Because both Mesp2 and Paraxis belong to bHLH-type transcription factors and their expression in the PSM is overlapping, we examined the possibility that these two factors might form a heterodimer when they coexist. A yeast two-hybrid assay system was used to address this question (Fields and Bartel, 2001). As bHLH transcription factors usually dimerize with each other through the bHLH domain, and Mesp2 exhibits self-transactivation activity when it was used as a bait, only bHLH domains were used for either factors. As shown in Figure 2, neither Paraxis nor Mesp2 form homodimers or heterodimers with each other, but as a control, do generate heterodimers with the ubiquitously expressed E47. Analysis of mutual regulation between these two genes in single-knockout mice revealed that Paraxis expression is not altered in Mesp2-null embryos and Mesp2 expression is also not altered in Paraxis-null embryos (data not shown). Therefore, these two genes may share the same target genes or lie in the same genetic network.

thumbnail image

Figure 2. No direct interactions between Mesp2 and Paraxis proteins. Yeast two-hybrid assay to investigate the interaction specificities between basic helix–loop–helix (bHLH) proteins. Five clones of double transfectants harboring pBTM118 variants and pGAD10 variants were assayed to detect of beta-galactosidase activity. A–E: The vector combinations are vacant pGAD10, pGAD10-Mesp1, pGAD10-Mesp2, pGAD10-Paraxis, and pGAD10-E47, against pBTM118-E47 (A); pBTM118-Mesp1 (B); pBTM118-Mesp2 (C); pBTM118-Paraxis (D); and original pBTM118 (E). The results demonstrated that Mesp1, Mesp2, and Paraxis do not form heterodimers with each other, and each of them forms a heterodimer only with E47, a ubiquitous partner.

Download figure to PowerPoint

Mesp2 Establishes and Paraxis Maintains Rostrocaudal Polarity of Somites

Mesp2-null mice exhibit a strongly caudalized vertebral morphology, which is characterized by an almost complete fusion of the pedicles of the neural arches (Fig. 1B). However, the Mesp2/Paraxis double-null fetus showed no pedicles at all (Fig. 1D). This phenotype is quite similar to that of Psen1-null fetus, which shows a rostralization of somites (Takahashi et al., 2000). In addition, Paraxis is also implicated in maintenance of the rostrocaudal polarity (Johnson et al., 2001). To test the possibility that the somites in the double-null mice might lack the caudal property, we examined the rostrocaudal polarity of somites in Mesp2/Paraxis double-knockout using several molecular markers.

We have previously shown that the expression pattern of a Notch ligand Dll1 (Bettenhausen et al., 1995) reflects the rostrocaudal polarity of somites (Takahashi et al., 2000) and that stripes of Dll1 expression in the caudal-half of somites are not formed in Paraxis-null embryos (Johnson et al., 2001; Fig. 3C). In Mesp2-null embryos, strong expression of Dll1 in the PSM was expanded to the putative somite region and gradually faded away (Fig. 3B). In the Mesp2/Paraxis double-null embryos, Dll1 expression was expanded but disappeared relatively suddenly in the mature somite region (Fig. 3D). This finding suggests that maintenance of Dll1 expression in the mature somite region in both wild-type and Mesp2-null embryos requires Paraxis.

thumbnail image

Figure 3. Somites in Mesp2/Paraxis double-null embryos show significant down-regulation of both rostral and caudal gene expression. A–D:Dll1. E–H:Uncx4.1. I–L:Tbx18 expression patterns in embryonic day (E) 11.5 embryos with indicated genotypes. While strong expression of Dll1 and Uncx4.1 is expanded in Mesp2-null embryos (A,B,E,F), expression of these genes is significantly reduced in both Paraxis-null (C,G) and double-null embryos (D,H), especially in the mature somite region (brackets in G,H). Tbx18 expression is severely reduced in both Mesp2-null and Paraxis-null, and in double-null embryos (I–L).

Download figure to PowerPoint

Uncx4.1 is not only a caudal molecular marker (Mansouri et al., 1997) but is required for the formation of the pedicles of neural arches, proximal ribs, and transverse processes, which are derived from the caudal-half of somites (Leitges et al., 2000; Mansouri et al., 2000). As shown in Figure 3, Uncx4.1 expression in the caudal-half of somites (Fig. 3E) was expanded in Mesp2-null embryos, and strong expression was retained in the putative somite region over the length of 10 somites (Fig. 3F). Paraxis-null embryos showed a normal stripe pattern of Uncx4.1 expression, but expression level was significantly lower than that in the wild-type embryo, especially anterior to the fourth to fifth recently formed somite (Fig. 3G). In Mesp2/Paraxis double-null embryos, the Uncx4.1 expression pattern was uniform, and the expression level was reduced especially in the putative mature somite region (Fig. 3H). These observations suggest that Paraxis has roles in maintenance of caudal half property, and expression of a key gene for skeletal element specification. To trace descendants of Uncx4.1-positive cells, we crossed these mice with Uncx4.1-lacZ transgenic mouse (Fig. 4E–H). As Uncx4.1 is finally expressed in the caudal lateral sclerotome, the Uncx4.1-LacZ mainly labels its derivatives, the pedicle of neural arch, proximal rib, and transverse process. In the lumbar vertebrae of wild-type and Mesp2-null embryos, the pedicle and transverse process were stained (Fig. 4E,F). In Paraxis-null embryos, the transverse process was missing but the pedicle was observed (Fig. 4G). In the Mesp2/Paraxis double-null embryos, staining for neither pedicle nor transverse process was observed (Fig. 4H). Thus, these results confirm observations from the skeletal specimen.

Tbx18 is expressed in the rostral-half of somites and is implicated in the maintenance of the rostral compartment of somites (Bussen et al., 2004). Tbx18 expression was severely reduced in both Mesp2-null and Paraxis-null, and the double-null embryos (Fig. 3J–L). These results indicate that both Mesp2 and Paraxis are necessary for maintenance of rostral property.

Mesp2/Paraxis Double-Null Embryo Exhibits Severely Reduced Expression of Both Pax1 and Pax9

Significant hypoplasty of the vertebral body derived from the medial sclerotome was found in the Paraxis-null fetus. To understand this phenotype, we examined the differentiation of the medial sclerotome. It is well known that Sonic hedgehog (Shh) expressed in the notochord and floor plate is responsible for both sclerotome induction and survival by means of inducing the target gene Pax1 (Fan and Tessier-Lavigne, 1994; Johnson et al., 1994). The paired box transcription factors Pax1 and Pax9 (Schnittger et al., 1992; Neubuser et al., 1995) are essential for cell proliferation and chondrogenic differentiation of the sclerotome (Wilm et al., 1998; Peters et al., 1999). Other transcription factors expressed in all cartilaginous tissues are Nkx3.2 (Bapx1; Tribioli et al., 1997; Lettice et al., 1999; Tribioli and Lufkin, 1999) and Sox9 (Wright et al., 1995; Bi et al., 1999; Zeng et al., 2002). Bapx1 is known to be a direct target of Pax1 and Pax9 (Rodrigo et al., 2003; see also Fig. 8). In the lumbar region at E11.5, expression of Pax1, Nkx3.1 (not shown), and Bapx1 in the entire sclerotome and expression of Sox9 in the ventral sclerotome was severely down-regulated in the Paraxis-null and double-null embryos, indicating that induction of these genes are dependent on Paraxis function (Fig. 5A–D,I–P). The expression pattern of the cartilage-specific marker Col2a largely reflected the above expression patterns and results from Alcian blue staining (Fig. 5Q–T). Of interest, these genes did not show abnormal expression in the Mesp2-null embryos. However, a difference between the Paraxis-null and double-null embryos was detected in the expression pattern of another Pax gene, Pax9 (Fig. 5E–H). Pax9 expression appeared slightly reduced in the Paraxis-null embryo (Fig. 5G), but spatial distribution was normal. In contrast, Pax9 expression was significantly decreased, especially in the lateral sclerotome in the double-null embryo (Fig. 5H). Whole-mount in situ analysis of E9.5 embryos also confirmed the severe reduction of Pax9 expression (Fig. 5U–X). Because the Pax1/Pax9 double-null mouse completely lacks vertebral bodies in the vertebrae (Peters et al., 1999), this additional lack of Pax9 may contribute to the more severe defects in formation of the vertebral bodies in the Mesp2/Paraxis double-null fetus. In addition, this severe reduction of Pax9 expression suggests that the level of Uncx4.1 expression in the double mutant is functionally insufficient, because a high level of Pax9 expression in the caudal lateral sclerotome is dependent on Uncx4.1 (Mansouri et al., 2000). Taken together, the observed loss of skeletal elements in the double-null fetus may be attributed to a severe down-regulation of both Pax1 and Pax9 (loss of vertebral body), in addition to dysfunction of Uncx4.1 (loss of pedicles).

thumbnail image

Figure 8. A schematic illustration for the putative genetic network, including interactions between Mesp2 and Paraxis. Red arrows indicate new findings in the current study. Paraxis is involved in expression of both Pax1 and Uncx4.1. Mesp2 suppresses Uncx4.1, which activates Pax9, but might also positively regulate Pax9. In addition, Mesp2 and Paraxis redundantly regulate Pax3 expression in the nascent somites.

Download figure to PowerPoint

thumbnail image

Figure 5. Analysis of sclerotomal and chondrogenic gene expression in the Mesp2/Paraxis double-null embryos. A–T: Expression of Pax1 (A–D), Pax9 (E–H), Sox9 (I–L), Bapx1 (M–P), and Col2a (Q–T) in transverse sections at the lumbar region at embryonic day (E) 11.5. Note that Pax1, Sox9, and Bapx1 expression is reduced in the absence of Paraxis, and Pax9 expression is severely decreased in the double mutant. U–X: Whole-mount specimen at E9.5 confirms the severe reduction of Pax9 expression.

Download figure to PowerPoint

Pax3 Expression in the PSM and Nascent Somites Is Regulated by Both Mesp2 and Paraxis

We have also examined expression of several genes involved in myogenesis and tendon formation (Fig. 6). Expression patterns of myogenin, myod (Sassoon et al., 1989), and Pax7 (Jostes et al., 1990) in transverse sections at the E11.5 lumbar region revealed that mislocalized myotome formation is seen mainly in Paraxis-null embryos and is further enhanced in the Mesp2/Paraxis double-null embryos. Scleraxis expression (Cserjesi et al., 1995; Brent et al., 2003) also confirmed that the normal localization of sclerotome–myotome–syndetome is severely impaired in the double-null embryos (Fig. 6I–L).

thumbnail image

Figure 6. Analysis of myotomal and syndetomal gene expression in the Mesp2/Paraxis double-null embryos. A–P: Expression of myogenin (A–D), myod (E–H), scleraxis (I–L), and Pax7 (M–P) in transverse sections at the lumbar region at embryonic day (E) 11.5. Note that the spatial organization of myotome is perturbed in both Paraxis-null and double-null embryos.

Download figure to PowerPoint

Because Paraxis regulates the expression of Pax3 and MyoD, two crucial factors in hypaxial muscle differentiation (Wilson-Rawls et al., 1999), we examined the genetic interactions of Mesp2 and Paraxis on the expression of these genes in the tail somites (Fig. 7). Both epaxial and hypaxial expression of MyoD was observed in Mesp2-null myotome as well as that in wild-type, although MyoD expression was delayed and poor (Fig. 7A,B). Hypaxial myotome formation was delayed and mislocalized in the Paraxis-null embryo (Fig. 7C). The Mesp2/Paraxis double-null embryo shows a more severe reduction in MyoD-expressing cell populations and hypaxial myotome formation was seldom detected (Fig. 7D). To determine whether the defects in myotome formation of the double-null embryo result from an additive effect from a lack of Mesp2 and Paraxis, we examined the expression pattern of Pax3, a gene upstream of MyoD (Maroto et al., 1997; Tajbakhsh et al., 1997). In the wild-type embryo, Pax3 expression was observed in the anteriormost PSM and somites and was localized to the dermomyotome in the mature somites (Fig. 8E; Williams and Ordahl, 1994). In the Mesp2-null embryo, Pax3 expression in the recently formed somite region was down-regulated, while that in the anteriormost PSM and dermomyotome of the mature somite region was relatively unaffected (Fig. 8F). The Paraxis-null embryo showed severely decreased Pax3 expression in the dermomyotome, while expression in the anteriormost PSM and forming somites was maintained (Fig. 8G and Wilson-Rawls et al., 1999). Expression of Pax3 was severely decreased in the Mesp2/Paraxis double-null embryo, resulting in an almost complete loss of Pax3 expression throughout the paraxial mesoderm (Fig. 8H). This finding suggests that Mesp2 and Paraxis cooperatively regulate Pax3 expression in the anteriormost PSM and forming somites, and that the observed effects on myotome formation are not only additive but also include a synergistic effect on hypaxial myogenesis.

thumbnail image

Figure 7. Pax3 expression in the nascent somite region is regulated by both Mesp2 and Paraxis. A–H:Myod (A–D) and Pax3 (E–H) expression of embryonic day (E) 11.5 in the tail region. Note that myod expression appears delayed and severely disrupted (D), and Pax3 expression in the nascent somite region is severely reduced (H) in the double-null embryo.

Download figure to PowerPoint

DISCUSSION

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

Paraxis is a bHLH transcription factor expressed in the anterior PSM, entire somite, and then dermomyotomal compartment of mature somites in mouse and chick embryos, and is implicated in multiple aspects of somitogenesis. Although its roles in epithelialization of somites (Burgess et al., 1996), hypaxial myotome formation (Wilson-Rawls et al., 1999) and maintenance of rostrocaudal polarity (Johnson et al., 2001) have been reported, less attention has been paid to its genetic interaction with other genes during axial skeletogenesis. In this study, we have shown that Paraxis is required for the expression of Pax1, Nkx3.1, Bapx1, and Sox9, crucial factors in vertebral column formation, in the ventral sclerotome in the posterior body. These findings suggest that Paraxis has significant roles in formation of the vertebral body at the ventral midline. Mesp2-null fetuses exhibit almost complete fusion of neural arches and proximal rib elements, and this defect is attributed to ectopic formation of skeletal elements derived from caudal half compartment of somites. So far, no sign of a loss of skeletal elements has been suggested from previous genetic analyses using double-knockout studies with Mesp2-null mice. Therefore, the reduction of vertebral elements observed in the Mesp2/Paraxis mutant was surprising and led us to examine the caudal gene expression. The most striking phenotype in the Mesp2/Paraxis double-null vertebrae is the extensive loss of pedicles. We have found that expression level of Uncx4.1 is significantly reduced in both Paraxis-null and Mesp2/Paraxis double-null embryos, especially in the mature somite region. The reduction of Uncx4.1 expression in the Paraxis-null embryo may contribute to the partial loss of the pedicles and loss of proximal portion of ribs. Although the low level of Uncx4.1 stripe expression in Paraxis-null embryo is sufficient for formation of pedicles, the diffuse and low expression of Uncx4.1 in the Mesp2/Paraxis double-null embryo may be insufficient for formation of pedicles. Because Pax1 expression was severely reduced in the Paraxis-null mutant, we initially asked whether the lack of Pax1 and the additional loss of Mesp2 might be responsible for the compromised sclerotomal program by generating Mesp2/Pax1 double-null mice. These compound-null mice, however, exhibited only additive skeletal phenotype and did not show the loss of vertebral bodies and pedicles observed in the Mesp2/Paraxis double-mutants (data not shown). Our further analysis of sclerotomal gene expression clarified that, in addition to Pax1, Pax9 expression is significantly down-regulated in the Mesp2/Paraxis double-null embryo. Because Pax1/Pax9 double-null mouse completely lacks vertebral bodies in the vertebrae (Peters et al., 1999), this additional lack of Pax9 may contribute to the severer defects in formation of the vertebral bodies in the Mesp2/Paraxis double-null fetus (Fig. 8).

The extent of reduction of Pax9 appears much more severe than that of Uncx4.1 in the double mutants. Although Pax1 and Pax9 are initially expressed in the entire sclerotome, later Pax1 becomes predominantly expressed in the ventromedial sclerotome and Pax9 expression becomes stronger in the ventrolateral sclerotome, suggesting some differential usage of the two genes. In addition, strong Pax9 expression in the caudal lateral sclerotome is dependent on Uncx4.1, suggesting its involvement in pedicle formation (Mansouri et al., 2000; Fig. 8). Thus, in the absence of Uncx4.1, the potential functions of Pax9 are also lost. It is noted that, in the Pax1/Pax9 double-null embryos, the proximal parts of ribs are missing in addition to the vertebral bodies and intervertebral discs (Peters et al., 1999), suggesting roles of Pax9 in the lateral sclerotome. In the nascent somite, Mesp2 suppresses Uncx4.1, which activates Pax9, and thus Uncx4.1 and Pax9 expression is stronger in the Mesp2-null condition. In the absence of Paraxis, however, this up-regulation of Uncx4.1 does not occur and additional loss of Mesp2 causes down-regulation of Pax9. Therefore, Mesp2 may positively regulate Pax9 independently of Uncx4.1 (Fig. 8).

Mesp2 is necessary for the normal spatial patterning of the myotome (Saga et al., 1997). Mesp2-null embryos show significant delay and perturbation of myogenesis (Fig. 7), and the current study has revealed that Pax3 expression is regulated by both Paraxis and Mesp2 in the anteriormost PSM and nascent somite region. As the Pax3 expression domain overlaps with those of Paraxis and Mesp2 at this region, it is possible that these two factors directly regulate the Pax3 gene. Pax9 appears to be expressed in the formed somites, after robust Mesp2 expression ceases, suggesting indirect regulation. The current study has revealed previously unrecognized interactions between Mesp2 and Paraxis in the formation of axial skeleton and musculature. Analysis of molecular interactions between these factors and Pax genes, as well as Uncx4.1, will contribute to understanding of genetic network underlying somitogenesis.

EXPERIMENTAL PROCEDURES

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

Animals

The Paraxis-null mice (Burgess et al., 1996) and Mesp2-LacZ knockin mice (Takahashi et al., 2000) were maintained as heterozygotes with an ICR background in the National Institute of Health Sciences, Tokyo. The double heterozygous mice were mated to obtain the Mesp2/Paraxis double-null embryos. The Mesp2-LacZ knockin mice has impaired Mesp1 function in addition to lack of Mesp2, thus were suitable for analyzing roles of the Mesp family in somitogenesis (Morimoto et al., 2006).

Yeast Two Hybrid Assay

A cDNA fragment of Mesp2 encoding a bHLH domain (from Q80 to L139 in MESP2) was fused in-frame to the lexA-coding sequence in vector pBTM118 (Fields and Bartel, 2001). The L40, a yeast strain containing lexA-HIS3 and lexA-lacZ reporter genes, was first transformed with pBTM118-Mesp2, and then with a mouse E11 cDNA library constructed in pGAD10 (Gal4 activation domain fusion vector, Clontech), using the lithium acetate method. Six hundred independent colonies were isolated from 1 × 106 transformants on the selective media plate lacking leucine, tryptophan, and histidine and supplemented with 5 mM 3-aminotriazole. All these clones were restreaked onto the selection medium and assayed for beta-galactosidase activity by a filter assay (Vojtek et al., 1993). The plasmids containing cDNA fragments were isolated from yeast culture, and the cDNAs were characterized by sequencing analysis and similarity search using NCBI programs (BLASTn). We identified E47 as an interacter for Mesp2.

To test for interaction specificity, we constructed the pBTM118 and pGAD10 variants expressing fusion proteins of Mesp1, Mesp2, and Paraxis and E47. Although full-length cDNAs were inserted into pGAD10, the fragmental cDNAs including a bHLH domain were used to construct pBTM118 variants, because the full-length cDNAs exhibited cognate activity of Gal4. cDNA fragments coding for Mesp1 (G74 to S137), Mesp2 (Q80 to L139), Paraxis (R67 to L126), and E47 (410A to 650G) were inserted. Then, the resulting pBTM118 and pGAD10 variants were cotransfected into L40 and selected on the selective media plate supplemented with histidine. Aroused colonies were restreaked on the same plate and used for a filter assay to detect beta-galactosidase activity.

In Situ Hybridization, Histological Analysis, and Skeletal Preparations

The methods used for whole-mount and section in situ hybridization, histology, and skeletal preparation by Alcian blue/Alizarin red staining are as described in our previous reports (Saga et al., 1997; Takahashi et al., 2000). For Alcian blue staining of paraffin sections, deparaffinized and rehydrated sections were treated with 3% acetic acid for 2 min, and then stained with 1% Alcian blue/3% acetic acid for 30 min. For detection of beta-galactosidase activity in Uncx-LacZ–expressing cells, embryos were fixed in 2% paraformaldehyde, 0.2% glutaraldehyde, and 0.02% NP-40 in phosphate buffered saline for 30–60 min at room temperature. After embedded in OCT compound, frozen sections were cut at 8 μm, dried overnight, and stained with X-gal solution for several hours. The sections were briefly counterstained with eosin.

Acknowledgements

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

We thank Randy Johnson for providing us the Uncx4.1-LacZ transgenic mouse line and the following researchers for cDNA clones encoding Dll1 (A. Gossler), Uncx4.1, Pax7 (P. Gruss), Sox9, and Col2a (K.S.E. Cheah). We also thank M. Ikumi, C. Fujihira, and E. Ikeno for technical assistance and Douglas Anderson for editorial assistance. This work was supported by Grants-in-Aid for Science Research on Priority Areas (B), the Organized Research Combination System, and National BioResource Project of the Ministry of Education, Culture, Sports, Science and Technology, Japan.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES
  • Barnes GL, Alexander PG, Hsu CW, Mariani BD, Tuan RS. 1997. Cloning and characterization of chicken Paraxis: a regulator of paraxial mesoderm development and somite formation. Dev Biol 189: 95111.
  • Bettenhausen B, Hrabe de Angelis M, Simon D, Guenet JL, Gossler A. 1995. Transient and restricted expression during mouse embryogenesis of Dll1, a murine gene closely related to Drosophila Delta. Development 121: 24072418.
  • Bi W, Deng JM, Zhang Z, Behringer RR, de Crombrugghe B. 1999. Sox9 is required for cartilage formation. Nat Genet 22: 8589.
  • Brent AE, Schweitzer R, Tabin CJ. 2003. A somitic compartment of tendon progenitors. Cell 113: 235248.
  • Burgess R, Cserjesi P, Ligon KL, Olson EN. 1995. Paraxis: a basic helix-loop-helix protein expressed in paraxial mesoderm and developing somites. Dev Biol 168: 296306.
  • Burgess R, Rawls A, Brown D, Bradley A, Olson EN. 1996. Requirement of the paraxis gene for somite formation and musculoskeletal patterning. Nature 384: 570573.
  • Bussen M, Petry M, Schuster-Gossler K, Leitges M, Gossler A, Kispert A. 2004. The T-box transcription factor Tbx18 maintains the separation of anterior and posterior somite compartments. Genes Dev 18: 12091221.
  • Cserjesi P, Brown D, Ligon KL, Lyons GE, Copeland NG, Gilbert DJ, Jenkins NA, Olson EN. 1995. Scleraxis: a basic helix-loop-helix protein that prefigures skeletal formation during mouse embryogenesis. Development 121: 10991110.
  • Fan CM, Tessier-Lavigne M. 1994. Patterning of mammalian somites by surface ectoderm and notochord: evidence for sclerotome induction by a hedgehog homolog. Cell 79: 11751186.
  • Fields S, Bartel PL. 2001. The two-hybrid system. A personal view. Methods Mol Biol 177: 38.
  • Johnson RL, Laufer E, Riddle RD, Tabin C. 1994. Ectopic expression of Sonic hedgehog alters dorsal-ventral patterning of somites. Cell 79: 11651173.
  • Johnson J, Rhee J, Parsons SM, Brown D, Olson EN, Rawls A. 2001. The anterior/posterior polarity of somites is disrupted in paraxis-deficient mice. Dev Biol 229: 176187.
  • Jostes B, Walther C, Gruss P. 1990. The murine paired box gene, Pax7, is expressed specifically during the development of the nervous and muscular system. Mech Dev 33: 2737.
  • Kitajima S, Takagi A, Inoue T, Saga Y. 2000. MesP1 and MesP2 are essential for the development of cardiac mesoderm. Development 127: 32153226.
  • Leitges M, Neidhardt L, Haenig B, Herrmann BG, Kispert A. 2000. The paired homeobox gene Uncx4.1 specifies pedicles, transverse processes and proximal ribs of the vertebral column. Development 127: 22592267.
  • Lettice LA, Purdie LA, Carlson GJ, Kilanowski F, Dorin J, Hill RE. 1999. The mouse bagpipe gene controls development of axial skeleton, skull, and spleen. Proc Natl Acad Sci U S A 96: 96959700.
  • Mansouri A, Yokota Y, Wehr R, Copeland NG, Jenkins NA, Gruss P. 1997. Paired-related murine homeobox gene expressed in the developing sclerotome, kidney, and nervous system. Dev Dyn 210: 5365.
  • Mansouri A, Voss AK, Thomas T, Yokota Y, Gruss P. 2000. Uncx4.1 is required for the formation of the pedicles and proximal ribs and acts upstream of Pax9. Development 127: 22512258.
  • Maroto M, Reshef R, Munsterberg AE, Koester S, Goulding M, Lassar AB. 1997. Ectopic Pax-3 activates MyoD and Myf-5 expression in embryonic mesoderm and neural tissue. Cell 89: 139148.
  • Morimoto M, Kiso M, Sasaki N, Saga Y. 2006. Cooperative Mesp activity is required for normal somitogenesis along the anterior-posterior axis. Dev Biol 300: 687698.
  • Nakajima Y, Morimoto M, Takahashi Y, Koseki H, Saga Y. 2006. Identification of Epha4 enhancer required for segmental expression and the regulation by Mesp2. Development 133: 25172525.
  • Neidhardt LM, Kispert A, Herrmann BG. 1997. A mouse gene of the paired-related homeobox class expressed in the caudal somite compartment and in the developing vertebral column, kidney and nervous system. Dev Genes Evol 207: 330339.
  • Neubuser A, Koseki H, Balling R. 1995. Characterization and developmental expression of Pax9, a paired-box-containing gene related to Pax1. Dev Biol 170: 701716.
  • Peters H, Wilm B, Sakai N, Imai K, Maas R, Balling R. 1999. Pax1 and Pax9 synergistically regulate vertebral column development. Development 126: 53995408.
  • Pourquie O. 2003. Vertebrate somitogenesis: a novel paradigm for animal segmentation? Int J Dev Biol 47: 597603.
  • Rodrigo I, Hill RE, Balling R, Munsterberg A, Imai K. 2003. Pax1 and Pax9 activate Bapx1 to induce chondrogenic differentiation in the sclerotome. Development 130: 473482.
  • Saga Y, Hata N, Koseki H, Taketo MM. 1997. Mesp2: a novel mouse gene expressed in the presegmented mesoderm and essential for segmentation initiation. Genes Dev 11: 18271839.
  • Sassoon D, Lyons G, Wright WE, Lin V, Lassar A, Weintraub H, Buckingham M. 1989. Expression of two myogenic regulatory factors myogenin and MyoD1 during mouse embryogenesis. Nature 341: 303307.
  • Schnittger S, Rao VV, Deutsch U, Gruss P, Balling R, Hansmann I. 1992. Pax1, a member of the paired box-containing class of developmental control genes, is mapped to human chromosome 20p11.2 by in situ hybridization (ISH and FISH). Genomics 14: 740744.
  • Tajbakhsh S, Rocancourt D, Cossu G, Buckingham M. 1997. Redefining the genetic hierarchies controlling skeletal myogenesis: Pax-3 and Myf-5 act upstream of MyoD. Cell 89: 127138.
  • Takahashi Y, Koizumi K, Takagi A, Kitajima S, Inoue T, Koseki H, Saga Y. 2000. Mesp2 initiates somite segmentation through the Notch signalling pathway. Nat Genet 25: 390396.
  • Takahashi Y, Inoue T, Gossler A, Saga Y. 2003. Feedback loops comprising Dll1, Dll3 and Mesp2, and differential involvement of Psen1 are essential for rostrocaudal patterning of somites. Development 130: 42594268.
  • Takahashi Y, Kitajima S, Inoue T, Kanno J, Saga Y. 2005. Differential contributions of Mesp1 and Mesp2 to the epithelialization and rostro-caudal patterning of somites. Development 132: 787796.
  • Tribioli C, Lufkin T. 1999. The murine Bapx1 homeobox gene plays a critical role in embryonic development of the axial skeleton and spleen. Development 126: 56995711.
  • Tribioli C, Frasch M, Lufkin T. 1997. Bapx1: an evolutionary conserved homologue of the Drosophila bagpipe homeobox gene is expressed in splanchnic mesoderm and the embryonic skeleton. Mech Dev 65: 145162.
  • Tseng HT, Jamrich M. 2004. Identification and developmental expression of Xenopus paraxis. Int J Dev Biol 48: 11551158.
  • Vojtek AB, Hollenberg SM, Cooper JA. 1993. Mammalian Ras interacts directly with the serine/threonine kinase Raf. Cell 74: 205214.
  • Williams BA, Ordahl CP. 1994. Pax-3 expression in segmental mesoderm marks early stages in myogenic cell specification. Development 120: 785796.
  • Wilm B, Dahl E, Peters H, Balling R, Imai K. 1998. Targeted disruption of Pax1 defines its null phenotype and proves haploinsufficiency. Proc Natl Acad Sci U S A 95: 86928697.
  • Wilson-Rawls J, Hurt CR, Parsons SM, Rawls A. 1999. Differential regulation of epaxial and hypaxial muscle development by paraxis. Development 126: 52175229.
  • Wright E, Hargrave MR, Christiansen J, Cooper L, Kun J, Evans T, Gangadharan U, Greenfield A, Koopman P. 1995. The Sry-related gene Sox9 is expressed during chondrogenesis in mouse embryos. Nat Genet 9: 1520.
  • Zeng L, Kempf H, Murtaugh LC, Sato ME, Lassar AB. 2002. Shh establishes an Nkx3.2/Sox9 autoregulatory loop that is maintained by BMP signals to induce somitic chondrogenesis. Genes Dev 16: 19902005.