Chicken Pax-1 gene: structure and expression during embryonic somite development

Engineering Musculoskeletal Grafts for Multi-Tissue Unit Repair: Lessons From Developmental Biology and Wound Healing

In the musculoskeletal system, bone, tendon, and skeletal muscle integrate and act coordinately as a single multi-tissue unit to facilitate body movement. The development, integration, and maturation of these essential components and their response to injury are vital for conferring efficient locomotion. The highly integrated nature of these components is evident under disease conditions, where rotator cuff tears at the bone-tendon interface have been reported to be associated with distal pathological alterations such as skeletal muscle degeneration and bone loss. To successfully treat musculoskeletal injuries and diseases, it is important to gain deep understanding of the development, integration and maturation of these musculoskeletal tissues along with their interfaces as well as the impact of inflammation on musculoskeletal healing and graft integration. This review highlights the current knowledge of developmental biology and wound healing in the bone-tendon-muscle multi-tissue unit and perspectives of what can be learnt from these biological and pathological processes within the context of musculoskeletal tissue engineering and regenerative medicine. Integrating these knowledge and perspectives can serve as guiding principles to inform the development and engineering of musculoskeletal grafts and other tissue engineering strategies to address challenging musculoskeletal injuries and diseases.

The function of DrPax1b gene in the embryonic development of zebrafish

Vertebrate Pax1 gene is a member of Pax gene family and encodes a transcription factor associated with crucial roles in the development of pharyngeal pouch, scletrotome and limb bud. In zebrafish, the genome contains two Pax1 paralogs, DrPax1a and DrPax1b, which share high sequence similarity with other Pax1 genes. To elucidate the function of zebrafish DrPax1b gene, we first examined the gene expression pattern and found that it was mainly expressed in the endodermal pharyngeal pouch, caudal somites, notochord, and fin bud. Then, we performed knockdown experiments using antisense morpholino oligonucleotides, which lead to the defects in the vertebral column, tail, pharyngeal skeleton, and pectoral fin. Additionally, we also found that the mouse MmPax1 mRNA, but not the amphioxus AmphiPax1/9 mRNA, could rescue the MO-induced defects. Furthermore, sequence alignment revealed that the N-terminal region of vertebrate Pax1 and amphioxus Pax1/9 were highly conserved, whereas their C-terminal regions were relatively divergent. However, the chimeric Am(N)Dr(C)Pax1, Mm(N)Dr(C)Pax1 and Dr(N)Mm(C)Pax1 mRNA could partially rescue the defects, while the Dr(N)Am(C)Pax1 mRNA could not. In conclusion, our data demonstrate a conserved function of DrPax1b in the development of the vertebral column, pectoral fin and pharyngeal skeleton formation in zebrafish and also provide critical insight into the functional evolution of Pax1 gene by changing its C-terminal sequence.

Application of synthetic photostable retinoids induces novel limb and facial phenotypes during chick embryogenesis in vivo

We have recently developed a range of synthetic retinoid analogues which include the compounds EC23 and EC19. They are stable on exposure to light and are predicted to be resistant to the normal metabolic processes involved in the inactivation of retinoids in vivo. Based on the position of the terminal carboxylic acid groups in the compounds we suggest that EC23 is a structural analogue of all-trans retinoic acid (ATRA), and EC19 is an analogue of 13-cis retinoic acid. Their effects on the differentiation of pluripotent stem cells has been previously described in vitro and are consistent with this hypothesis. We present herein the first description of the effects of these molecules in vivo. Retinoids were applied to the anterior limb buds of chicken embryos in ovo via ion-exchange beads. We found that retinoid EC23 produces effects on the wing digits similar to ATRA, but does so at two orders of magnitude lower concentration. When larger quantities of EC23 are applied, a novel phenotype is obtained involving production of multiple digit 1s on the anterior limb. This corresponds to differential effects of ATRA and EC23 on sonic hedgehog (shh) expression in the developing limb bud. With EC23 application we also find digit 1 phenotypes similar to thumb duplications described in the clinical literature. EC23 and ATRA are shown to have effects on the entire proximal–distal axis of the limb, including hitherto undescribed effects on the scapula. This includes suppression of expression of the scapula marker Pax1. EC23 also produces effects similar to those of ATRA on the developing face, producing reductions of the upper beak at concentrations two orders of magnitude lower than ATRA. In contrast, EC19, which is structurally very similar to EC23, has novel, less severe effects on the face and rarely alters limb development. EC19 and ATRA are effective at similar concentrations. These results further demonstrate the ability of retinoids to influence embryonic development. Moreover, EC23 represents a useful new tool to investigate developmental processes and probe the mechanisms underlying congenital abnormalities in vertebrates including man.

The generation of vertebral segmental patterning in the chick embryo

We have carried out a series of experimental manipulations in the chick embryo to assess whether the notochord, neural tube and spinal nerves influence segmental patterning of the vertebral column. Using Pax1 expression in the somite-derived sclerotomes as a marker for segmentation of the developing intervertebral disc, our results exclude such an influence. In contrast to certain teleost species, where the notochord has been shown to generate segmentation of the vertebral bodies (chordacentra), these experiments indicate that segmental patterning of the avian vertebral column arises autonomously in the somite mesoderm. We suggest that in amniotes, the subdivision of each sclerotome into non-miscible anterior and posterior halves plays a critical role in establishing vertebral segmentation, and in maintaining left/right alignment of the developing vertebral elements at the body midline.

Familial Chiari malformation: case series

Chiari malformations (Types I-IV) are abnormalities of the posterior fossa that affect the cerebellum, brainstem, and the spinal cord with prevalence rates of 0.1%-0.5%. Case reports of familial aggregation of Chiari malformation, twin studies, cosegregation of Chiari malformation with known genetic conditions, and recent gene and genome-wide association studies provide strong evidence of the genetic underpinnings of familial Chiari malformation. The authors report on a series of 3 family pairs with Chiari malformation Type I: 2 mother-daughter pairs and 1 father-daughter pair. The specific genetic causes of familial Chiari malformation have yet to be fully elucidated. The authors review the literature and discuss several candidate genes. Recent advances in the understanding of the genetic influences and pathogenesis of familial Chiari malformation are expected to improve management of affected patients and monitoring of at-risk family members.

Commitment of chondrogenic precursors of the avian scapula takes place after epithelial-mesenchymal transition of the dermomyotome

BACKGROUND

Cells of the epithelially organised dermomyotome are traditionally believed to give rise to skeletal muscle and dermis. We have previously shown that the dermomyotome can undergo epithelial-mesenchymal transition (EMT) and give rise to chondrogenic cells, which go on to form the scapula blade in birds. At present we have little understanding regarding the issue of when the chondrogenic fate of dermomyotomal cells is determined. Using quail-chick grafting experiments, we investigated whether scapula precursor cells are committed to a chondrogenic fate while in an epithelial state or whether commitment is established after EMT.

RESULTS

We show that the hypaxial dermomyotome, which normally forms the scapula, does not generate cartilaginous tissue after it is grafted to the epaxial domain. In contrast engraftment of the epaxial dermomyotome to the hypaxial domain gives rise to scapula-like cartilage. However, the hypaxial sub-ectodermal mesenchyme (SEM), which originates from the hypaxial dermomyotome after EMT, generates cartilaginous elements in the epaxial domain, whereas in reciprocal grafting experiments, the epaxial SEM cannot form cartilage in the hypaxial domain.

CONCLUSIONS

We suggest that the epithelial cells of the dermomyotome are not committed to the chondrogenic lineage. Commitment to this lineage occurs after it has undergone EMT to form the sub-ectodermal mesenchyme.

An evo-devo view on the origin of the backbone: evolutionary development of the vertebrae

Vertebral columns are a group of diverse axial structures that define the vertebrates and provide supportive, locomotive, protective, and other important functions. The embryonic origin of the first vertebral element in this subphylum, the lamprey arcualia, has remained a puzzle for more than a century although much developmental and genetic progress has been made. The comparative approach is a very powerful tool for studying vertebrate morphological variation and understanding how the novel structures were generated during evolution. Here, I first briefly describe the vertebral structures and their developmental processes in major taxa, and then analyze the most recently published data on the basal vertebrates. Finally, an ontogenetic and phylogenetic origin is proposed. The lamprey may have already evolved a sclerotome, which gave rise to arcualia ontogenetically; whole genome duplications likely promoted the establishment of sclerotomal core genetic program by gene co-options.

Notochord-derived BMP antagonists inhibit endothelial cell generation and network formation

Embryonic blood vessel formation is initially mediated through the sequential differentiation, migration, and assembly of endothelial cells (ECs). While many molecular signals that promote vascular development have been identified, little is known about suppressors of this process. In higher vertebrates, including birds and mammals, the vascular network forms throughout the embryonic disk with the exception of a region along the midline. We have previously shown that the notochord is responsible for the generation and maintenance of the avascular midline and that BMP antagonists expressed by this embryonic tissue, including Noggin and Chordin, can mimic this inhibitory role. Here we report that the notochord suppresses the generation of ECs from the mesoderm both in vivo and in vitro. We also report that the notochord diminishes the ability of mature ECs to organize into a primitive plexus. Furthermore, Noggin mimics notochord-based inhibition by preventing mesodermal EC generation and mature EC network formation. These findings suggest that the mesoderm surrounding the midline is competent to give rise to ECs and to form blood vessels, but that notochord derived-BMP antagonists suppress EC differentiation and maturation processes leading to inhibition of midline vessel formation.

Role of nitric oxide in chick embryonic organogenesis and dysmorphogenesis

BACKGROUND

Nitric oxide (NO), produced by the nitric oxide synthase family of enzymes, mediates multiple signaling functions, and when unchecked, NO causes pathological damage. Exposure of embryos to a variety of teratogens, including carbon monoxide (CO), has been shown to increase reactive intermediates, such as NO, and recent work showed that either the excess or absence of NO caused morphological defects. While endogenous NO is known to regulate many adult tissues, its role during embryonic organogenesis and/or in mediating responses to teratogen exposure has not been explored.

METHODS

We have examined here the presence of NO during normal chick embryonic organogenesis, and investigated the teratogenicity of NO through the application of sodium nitroprusside (SNP), which mimics NO overproduction, and NG-monomethyl-L-arginine (L-NMMA), which inhibits endogenous NOS activity.

RESULTS

Topical treatment with SNP or L-NMMA for 18 h resulted in morphological defects, specifically in the neural tube and somites, which corresponded to sites of altered apoptosis. The location of NO was histochemically correlated with the observed morphological defects. Coadministration of SNP or L-NMMA with CO showed functional coregulation and interaction between NO and CO in chick embryonic development.

CONCLUSIONS

Our results showed that regulation of NO is essential for normal axial development, that sites of altered NO expression correlate to those of altered apoptosis and dysmorphogenesis, and that CO coadministration resulted in a rectification of normal NO expression. Collectively, these results suggest that alteration in endogenous NO/CO signaling is responsible, at least in part, for the observed NO-induced teratogenesis.

Regionalized expression of ADAM13 during chicken embryonic development

ADAMs are a family of membrane proteins possessing a disintegrin domain and a metalloprotease domain, which have functions in cell-cell adhesion, cell-matrix adhesion, and protein shedding, respectively. ADAMs are involved in morphogenesis and tissue formation during embryonic development. In the present study, chicken ADAM13 was cloned and identified, and its expression was investigated by semiquantitative reverse transcriptase-polymerase chain reaction and in situ hybridization during chicken embryonic development. Our results show that ADAM13 expression is temporally and spatially regulated in chicken embryos. At early developmental stages, ADAM13 is expressed in the head mesenchyme, which later develops into the craniofacial skeleton, in the branchial arches, and in the meninges surrounding the brain. Furthermore, ADAM13 mRNA was also detected in several tissues and organs, such as the somites and their derived muscles, the meninges surrounding the spinal cord, the dorsal aorta, the developing kidney, and several digestive organs.

A proliferative role for Wnt-3a in chick somites

The proper patterning of somites to give rise to sclerotome, dermomyotome, and myotome involves the coordination of many different cellular processes, including lineage specification, cell proliferation, cell death, and differentiation, by intercellular signals. One such family of secreted signaling proteins known to influence somite patterning is the Wnt family. Although the participation of Wnt-3a in the patterning of dorsal structures in the somite is well established, no clear consensus has emerged about the cellular processes that are governed by Wnt-3a in the somite. The recent demonstration that Wnt-3a has a proliferative role in the neural tube [Development 129 (2002) 2087] suggested that Wnt-3a might also act to regulate proliferation in somites. To test this hypothesis, we first analyzed the effects of Wnt-3a on segmental plate and somite explants (from Hamburger and Hamilton stage 10 chick embryos) grown in culture. These studies indicate that Wnt-3a is capable of maintaining and/or inducing expression of both Pax-3 and Pax-7, transcription factors that have been implicated in proliferation. To directly test for a role in proliferation, explants were immunostained with antibodies against phospho-histone H3. Explants treated with Wnt-3a show an increase in the percentage of cells expressing phospho-histone H3 as compared to controls. To test the proliferative effect of Wnt-3a in vivo, we ectopically expressed Wnt-3a in chick neural tubes via electroporation. Consistent with previous studies, ectopic expression of Wnt-3a in vivo results in a mediolateral expansion of the dermomyotome and myotome. We now show that proliferation of dorsal/dermomyotomal cells is significantly enhanced by ectopic Wnt-3a. Collectively, our explant and in vivo studies indicate that an increase in proliferation plays an important role in the expansion of the dermomyotome and myotome in Wnt-3a-treated embryos. Furthermore, our results demonstrate that small changes in proliferation can dramatically influence patterning and morphogenesis.

Molecular and cellular biology of avian somite development

Much of our understanding of early vertebrate embryogenesis derives from experimental work done with the chick embryo. Studies of the avian somite have played a key role in elucidating the developmental history of this important structure, the source of most muscle and bone in the organism. Here we review the development of the avian somite including morphological and molecular data on the origin of paraxial mesoderm, maturation of the segmental plate, specification and formation of somite compartments, and somite cell differentiation into cartilage and skeletal muscle.

Pax1/Pax9-Related genes in an agnathan vertebrate, Lampetra japonica: expression pattern of LjPax9 implies sequential evolutionary events toward the gnathostome body plan

Among the transcription factor gene families, Pax genes play important and unique roles in morphological patterning of animal body plans. Of these, Group I Pax genes (Pax1 and Pax9) are expressed in the endodermal pharyngeal pouches in many groups of deuterostomes, and vertebrates seem to have acquired more extensive expression domains in embryos. To understand the evolution of Pax1/Pax9-related genes in basal groups of vertebrates, their cognates were isolated from the Japanese marine lamprey, Lampetra japonica. RT-PCR of larval lamprey cDNA yielded two different fragments containing vertebrate Pax1- and Pax9-like paired domains. The Pax9 orthologue was isolated and named LjPax9. Whole-mount in situ hybridization revealed that this gene was expressed in endodermal pharyngeal pouches, mesenchyme of the velum (the oral pumping apparatus) and the hyoid arch, and the nasohypophysial plate, but not in the somitic mesoderm of the lamprey embryo. These expression patterns could be regarded as a link between the basal chordates and the gnathostomes and are consistent with the phylogenetic position of the lamprey. Especially, the appearance of neural crest seemed to be the basis of velar expression. Homology of the velum and the jaw is also discussed based on the LjPax9 expression in the first pharyngeal pouch and in the velar mesenchyme. We conclude that Pax9 genes have sequentially expanded into new expression domains through evolution as more complicated body plans emerged.

A genetic hypothesis for Chiari I malformation with or without syringomyelia

In several reports the authors have suggested occasional familial aggregation of syringomyelia and/or Chiari 1 malformation (CM1). Familial aggregation is one characteristic of traits that have an underlying genetic basis. The authors provide evidence for familial aggregation of CM1 and syringomyelia (CM1/S) in a large series of families, establishing that there may be a genetic component to CM1/S in at least a subset of families. The authors observed no cases of isolated familial syringomyelia in their family studies, suggesting that familial syringomyelia is more accurately classified as familial CM1 with associated syringomyelia.

These data, together with the cosegregation of the trait with known genetic syndromes, support the authors' hypothesis of a genetic basis for some CM1/S cases.

Expression of the paired-box genes Pax-1 and Pax-9 in limb skeleton development

Vertebrate Pax genes encode a family of transcription factors that play important roles in embryonic patterning and morphogenesis. Two closely related Pax genes, Pax-1 and Pax-9, are associated with early axial and limb skeleton development. To investigate the role of these genes in cartilage formation we have examined the expression profiles of Pax-1 and Pax-9 in developing chick limb mesenchyme in vivo and in vitro. Both transcripts are detected by reverse transcription polymerase chain reaction and Northern blotting throughout chick limb development, from the early bud stages (Hamburger-Hamilton 20-23) to fully patterned appendages (stage 30). Whole-mount in situ hybridization reveals complex, nonoverlapping expression domains of these two genes. Pax-1 transcripts first appear at the anterior proximal margin of the limb buds, while Pax-9 is expressed more distally at what will be the junction of the autopod and the zeugopod. In situ hybridization to serial sections of the girdles reveals that in the pectoral region Pax-1 is expressed proximally in condensed mesenchyme surrounding the junction of the developing scapula, humerus, and coracoid. In the pelvis, Pax-1 is expressed between the femur and the developing acetabulum and along the ventral edge of the ischium; this transcript was also found in the distal hindlimb along the posterior edge of the fibula. Pax-9 transcripts were not detected in the pectoral girdle at any stage, and only weakly in the pelvis along the ventral ischial margin. In the distal parts of both wings and legs, however, Pax-9 is strongly expressed between the anterior embryonic cartilages (e.g., distal radius or tibia) and the anterior ectodermal ridge. The expression of both genes was strongest in undifferentiated cells of precartilage condensations or at the margins of differentiated cartilages, and was absent from cartilage itself. In micromass cultures of chondrifying limb bud mesenchyme expression of Pax-1 and Pax-9 is maintained for up to 3 days in vitro, most strongly at the end of the culture period during chondrogenic differentiation. As seen in vivo, transcripts are found in loose mesenchyme cells at the outer margins of developing cartilage nodules, and are absent from differentiated chondrocytes at the nodule center. Taken together, these investigations extend previous studies of Pax-1 and Pax-9 expression in embryonic limb development while validating limb bud mesenchyme culture as an accessible experimental system for the study of Pax gene function and regulation. Our in vivo and in vitro observations are discussed with reference to 1) the relationship between somitic and limb expression of these two Pax genes, 2) what regulates this expression in different regions of the embryo, and 3) the putative cellular functions of Pax-1 and Pax-9 in embryonic skeletogenesis.

Sonic hedgehog is required for survival of both myogenic and chondrogenic somitic lineages

In vertebrates, the medial moieties of the somites give rise to the vertebrae and epaxial muscles, which develop in close relationship with the axial organs, neural tube and notochord. The lateral moieties contribute to the ribs and to limb and body wall muscles (hypaxial muscles) after a phase of lateral and ventral migration. Surgical ablation of the neural tube and notochord in the chick embryo during segmentation and early differentiation of the somites (day 2 of incubation) does not affect primary development of the hypaxial muscles, but leads to a complete absence of epaxial muscles, vertebrae and ribs, due to cell death in the somites. Here we demonstrate that cell death, which occurs within 24 hours of excision of the axial organs, affects both myogenic and chondrogenic cell lineages defined, respectively, by the expression of MyoD and Pax-1 genes. In contrast, Pax-3 transcripts, normally present in cells giving rise to hypaxial muscles, are preserved in the excised embryos. Backgrafting either the ventral neural tube or the notochord allows survival of MyoD- and Pax-1-expressing cells. Similarly, Sonic hedgehog-producing cells grafted in place of axial organs also rescue MyoD- and Pax-1-expressing cells from death and allow epaxial muscles, ribs and vertebrae to undergo organogenesis. These results demonstrate that the ventral neural tube and the notochord promote the survival of both myogenic and chondrogenic cell lineages in the somites and that this action is mediated by Sonic hedgehog.

High sensitivity analysis of gene expression in single embryonic somites using coupled reverse transcription-polymerase chain reaction

We have developed a highly sensitive and reproducible method to detect the expression of specific genes in small tissue samples, such as a single embryonic somite. The procedure, which utilizes coupled reverse transcription-polymerase chain reaction (RT-PCR), was developed for evaluating the sequence of gene expression occurring in single somites during chick embryonic development. Comparisons of results obtained from using combinations of various RNA isolation methods and reverse transcription methods demonstrate that a protocol using a commercially available RNA isolation reagent (Tri Reagent) followed by optimized PCR, successfully detects low levels of mRNAs.

Cell death in the avian sclerotome

In this study the occurrence of apoptotic cells in chick embryo trunk somites, between 2.5 and 4 days of development, has been examined using an in situ nick-end-labeling method (TUNEL) to identify nuclei in which DNA is undergoing fragmentation. At 2.5 days of development, apoptotic cells were found in the sclerotome with a distribution that depended on the rostrocaudal level in the trunk. At the most rostral levels (somites 1-18), dying cells were present primarily in the rostral half of the ventral sclerotome; at midlevels (somites 19-26), they were present throughout the ventral sclerotome; and at caudal levels (somites 27-32), no dying cells were present. By 4 days of development, the number of dying cells in the sclerotome was sharply reduced, and those present were primarily distributed to the caudal side of the intrasclerotomal fissure. Double labeling of cells for both TUNEL and the HNK-1 epitope, at 2.5 days, indicated that the majority of the dying cells were not neural crest cells. Further, dying cells in the rostral somite half were present largely in regions of the sclerotome that labeled poorly for HNK-1. It was confirmed that apoptotic neural crest cells retain the HNK-1 epitope and therefore would have been observed if present. Neural crest cells only appeared to be apoptotic in relatively small numbers and only at the ventral border of the sclerotome. Examination of DiI-labeled neural crest cells confirmed that the dying cells in the body of the somite were not primarily neural crest cells. Two hypotheses regarding the TUNEL-positive cells in the sclerotome were experimentally tested. First, that they originate from the somitocoel compartment of the somite, because their distribution patterns at 4 days were similar to those of somitocoel cells. To test this, somitocoel cells were labeled with carboxyfluorescein and grafted into host embryos in ovo. Results showed that these cells did not become apoptotic and that the dying cells were therefore not derived from the somitocoel. Second, the hypothesis was tested that the distribution patterns of the dying cells in the sclerotome are determined by factors outside the somite itself. Somites and segmental plates were transplanted into hosts in ovo with reversed orientation, after which the patterns of dying cells were examined using nile blue sulfate staining. The results indicated that the patterns were unchanged after a further 2 days incubation, suggesting that the patterns of cell death in the sclerotome are not determined solely from within the somite. The distribution of the cell death-associated gene products, bcl-2, bax, and interleukin-1 beta converting enzyme, indicates that although these proteins are segmentally distributed in the dermomyotome and in the rostrodorsal quadrant of the sclerotome, their patterns are not directly correlated with the distribution of dying cells.

Cloning and characterization of chicken Paraxis: a regulator of paraxial mesoderm development and somite formation

To investigate the molecular regulation of embryonic somite formation and development, we have cloned the full-length cDNA and characterized the embryonic expression profile of chicken Paraxis, a member of a novel family of basic helix-loop-helix (bHLH) proteins, which has been suggested to play a role in paraxial mesoderm development. Chicken Paraxis encodes a 1.35-kb mRNA and contains a 53-amino-acid residue bHLH domain, identical in sequence to that found in the mammalian Paraxis genes of mouse, hamster, and human. Northern analysis revealed significant Paraxis expression in the early embryo up to the 30- to 35-somite stage, declining from Incubation Day 4 on and becoming undetectable by Day 5. By whole-mount in situ hybridization, Paraxis expression is first seen distinctly in the emerging paraxial mesoderm of the primitive streak stage chick embryo. During gastrulation, Paraxis expression in the mesoderm defines bilaterally symmetric crescents located immediately rostral to Hensen's node and appears to pre-configure the emerging somitic mesoderm. During somite development, Paraxis expression is evident in the rostral segmental plate and the newly formed somites, although the level of expression clearly decreases in the more mature somites. By the 10-12th pair of somites, counting from the caudal end, Paraxis expression appears to be preferentially localized to the medial aspect of individual somites. Histological analysis showed that Paraxis expression is evenly distributed in the newly formed caudal epithelial somites, then localized to the medial portion of maturing somites, and preferentially localized in the dermomyotome of more rostral somites before diminishing to undetectable levels in the most cranial somites. The functional involvement of Paraxis in somite development was assessed by perturbing its expression in somitic stage chick embryos using a Paraxis-specific antisense oligonucleotide. Disruption of somite formation from the paraxial mesoderm was observed in 67% of the surviving topically treated embryos, whereas control embryos treated with sense or random sequence oligonucleotides did not show similar effects. In addition, direct injection of Paraxis-specific antisense oligonucleotide into the paraxial mesoderm produced discrete segmentation anomalies which correlated spatially with the site of injection. Whole-mount in situ hybridization revealed that the regions defective in somite formation displayed perturbed Paraxis expression and a reduction of Pax-1 expression, a marker for epithelial somites and sclerotome. Histological analysis indicated poor condensation and/or epithelization of the somitic mesoderm. Finally, embryos treated with valproic acid, a known teratogen which affects somite segmentation, showed perturbed Paraxis expression, suggesting that the mechanism of action of this teratogen involves a pathway(s) requiring Paraxis activity. These data provide evidence that Paraxis acts as an important regulator of paraxial mesoderm and somite development and functions in axial patterning of the chick embryo.