SEARCH

SEARCH BY CITATION

Keywords:

  • amphioxus;
  • ascidian;
  • evolution;
  • neural crest

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

The neural crest has long been regarded as one of the key novelties in vertebrate evolutionary history. Indeed, the vertebrate characteristic of a finely patterned craniofacial structure is intimately related to the neural crest. It has been thought that protochordates lacked neural crest counterparts. However, recent identification and characterization of protochordate genes such as Pax3/7, Dlx and BMP family members challenge this idea, because their expression patterns suggest remarkable similarity between the vertebrate neural crest and the ascidian dorsal midline epidermis, which gives rise to both epidermal cells and sensory neurons. The present paper proposes that the neural crest is not a novel vertebrate cell population, but may have originated from the protochordate dorsal midline epidermis. Therefore, the evolution of the vertebrate neural crest should be reconsidered in terms of new cell properties such as pluripotency, delamination–migration and the carriage of an anteroposterior positional value, key innovations leading to development of the complex craniofacial structure in vertebrates. Molecular evolutionary events involved in the acquisitions of these new cell properties are also discussed. Genome duplications during early vertebrate evolution may have played an important role in allowing delamination of the neural crest cells. The new regulatory mechanism of Hox genes in the neural crest is postulated to have developed through the acquisition of new roles by coactivators involved in retinoic acid signaling.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

Molecular biological methodologies have been applied to developmental biology with great success over the last decade. Developmental systems for morphologic structures can now be described in terms of gene functions as a result of the tremendous progress in molecular developmental biology, which has made the evolution of morphology a more accessible problem. One of the salient questions of evolutionary biology is how new structures or new cell types evolve, particularly with regard to genetic development.

The invertebrate to vertebrate transition was one of the most important steps in the course of human evolution, because many new cell types and tissues emerged at this stage. One example is the acquisition of vertebrae, and hence the name, by ancestral vertebrates. The evolution of vertebrae was accompanied by the evolution of new cell types; that is, chondrocytes and osteocytes.

Another new cell type that emerged in ancestral vertebrates was the neural crest. Neural crest cells arise at the boundary between the neural plate and the epidermis and migrate out, differentiating into fates that include pigment cells, peripheral neurons and skeletal tissue, depending on the final destination (LeDouarin & Kalcheim 1999). Hall (2000) has proposed that the neural crest be regarded as a fourth germ layer, because these properties of pluripotency and cell migration are comparable to those of mesoderm.

The neural crest has attracted the interest of evolutionary biologists because of its intimate relationship with the development of complex craniofacial structure, a characteristic of vertebrates, as well as representing a new cell type. Gans and Northcutt (1983) listed 20 new characteristics that have emerged during vertebrate evolution. Most of these are found in the head and relate to food detection and ingestion, which was essential for the transition from the filter-feeding protochordate state to that of the active predator. They proposed that many features of the vertebrate head derived from the neural crest and placodes. Regardless of whether one agrees with the detailed scenario of Gans and Northcutt (1983), the impact of the neural crest on the evolution of vertebrates is obvious, especially on craniofacial structure. In the present article I will review our current understanding of the origin and evolution of vertebrate neural crest, together with some speculative ideas. Evolutionary biology deals with history; thus, at times we need to speculate much more than in experimental sciences. Hypotheses have played and continue to play an important role in the progress of evolutionary biology; therefore, this paper presents speculations and includes suggestions on how they can be tested.

Brief overview of protochordate neural structure

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

Vertebrates belong to the phylum Chordata, which includes two protochordate groups: cephalochordates, represented by amphioxus, and urochordates, represented by ascidians. Chordates are a member of deuterostomes in which two other phyla, Echinodermata and Hemichordata, are included (Fig. 1). Protochordates (cephalochordates plus urochordates) share a similar body plan with vertebrates as well as several characteristics, such as pharyngeal gill slits, notochord and a dorsal tubular nervous system. Although protochordate neural tube structures are relatively simple, the basic structure and significant parts of the genetic mechanisms are conserved in protochordates and vertebrates (Holland & Chen 2001; Wada & Satoh 2001). Protochordates have an anterior neural tube region where Otx genes are expressed, which is homologous to the chordate forebrain and midbrain (Wada et al. 1996b; Williams & Holland 1996). The posterior part of the protochordate neural tube contains dichotomous regions that are comparable to the vertebrate hindbrain and spinal cord (Jackman et al. 2000; Wada & Satoh 2001), although the presence of a counterpart for the vertebrate isthmic organizer remains a matter of debate (Wada et al. 1998; Kozmik et al. 1999; Wada & Satoh 2001). Differentiation of the protochordate and vertebrate neural tubes along the dorsoventral axis appears to be controlled by similar genetic systems (Shimeld 1999a; Wada & Satoh 2001). Ventral expression of the hedgehog (hh) gene is observed in amphioxus (Shimeld 1999b), and bone morphogenetic protein (BMP) genes in the ascidian are expressed in epidermis abutting the neural tube (Miya et al. 1996, 1997); however, the contributions of ascidian hh or amphioxus BMP remain to be established.

image

Figure 1. Phylogenetic relationship of deuterostomes. The central nervous system is shown in blue, and the peripheral nervous system in red. The phylogenetic tree is based on Wada and Satoh (1994), Turbeville et al. (1994) and Castresana et al. (1998). The schematic illustration and distribution of peripheral nerves of the acornworm (hemichordate) are adopted from Knight-Jones (1952).

Download figure to PowerPoint

Despite conservation of the basic dorsoventral patterning system, protochordates are believed to lack definitive neural crest. Although neither ascidians nor amphioxus possess cells that fit the traditional definition of neural crest (i.e. cells that migrate out of their original location in the boundary between the neural tissue and epidermis), recent studies have shown the existence of protochordate genes homologous to those involved in vertebrate neural crest differentiation (Holland et al. 1996; Baker & Bronner-Fraser 1997; Holland & Holland 1998; Shimeld & Holland 2000; Wada & Satoh 2001).

Tetralogy

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

Before going on to discuss the origin of the neural crest, the genome duplications that may have occurred in early stages of vertebrate evolution will be reviewed briefly. Vertebrates often have two to four genes homologous to invertebrate counterpart genes. These are usually explained as a consequence of two rounds of genome duplication occurring in ancestral vertebrates; a feature called ‘tetralogy’ (Holland et al. 1994; Spring 1997), although there has been some opposition to this idea (Martin 2001). Early in the process, duplicated genes had redundant functions and some of the duplicated genes were lost relatively easily. Thus, in many cases, vertebrates have not retained a full complement of four genes, keeping only two or three counterpart genes. The supernumerary genes are believed to have stood as raw materials for the evolution of new cell types and structures (Holland et al. 1994; Holland & Garcia-Fernàndez 1996). Pax2, Pax5 and Pax8 have a single counterpart in invertebrates and are considered a result of genome duplication in ancestral vertebrates (Wada et al. 1998). Of these, Pax5 has a specific role in B-cell differentiation (Enver 1999), a process in which Pax2 or Pax8 are not involved. This is an example of recruitment of one of the duplicated genes for a new function. In some cases, the duplicated counterparts share a function in a vertebrate-specific structure; for example, both Pax1 and Pax9 are involved in sclerotome differentiation (Müller et al. 1996). Ascidians and amphioxus have single homologs of Pax1 and Pax9, which function in pharyngeal gill differentiation (Holland et al. 1995; Ogasawara et al. 1999). Because vertebrate Pax1 and Pax9 are both expressed in the pharyngeal gills (Müller et al. 1996), an ancestral function of the Pax1/9 probably involves pharyngeal gill differentiation. It is possible that the ancestral gene of Pax1 and Pax9 acquired a new role in sclerotome differentiation before genome duplication, which may have been essential for vertebral evolution. Genetic acquisition of new expression sites or new functions likely played an important role in morphologic evolution in the early evolution of vertebrates, without regard to the contribution made by gene duplication.

Origin of the neural crest in protochordates

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

Interaction between the neural plate and epidermis, together with signals from paraxial mesoderm, is important for the differentiation of the vertebrate neural crest (Selleck & Bronner-Fraser 1995; LaBonne & Bronner-Fraser 1999). Therefore, the neural crest is likely to have evolved in the region straddling the border between the neural plate and epidermis, not only because the neural crest occupies a homologous position, but also because this position is essential for its differentiation. In this section, the cells at the boundary between the neural plate and epidermis in protochordates are examined first, then gene expression patterns in those cells for which vertebrate homologs are involved in neural crest formation are compared.

The boundary between the neural plate and epidermis of protochordates

Protochordate surface ectoderm is believed to be relatively uniform, except for the presence of some epidermal sensory cells found in both ascidians and amphioxus, and the anterior adhesive papillae of ascidians. In the course of examining epidermis- specific gene expression, Ishida et al. (1996) found that ascidian surface ectoderm can be subdivided into six regions according to the sets of genes expressed (Fig. 2A,B). Of these, five regions are characterized as fields where epidermal peripheral neurons differentiate. Two of these regions correspond to the dorsal midline epidermis (regions 1 and 2 in Fig. 2A) and one to the ventral midline (region 5 in Fig. 2A). Two others correspond to regions where adhesive papilla and associated neurons differentiate (regions 3 and 4 in Fig. 2A). The dorsal midline populations abut the neural tube and thus are topologically homologous to the neural crest. Several cells from the dorsal midline population differentiate into primary sensory neurons, which are probably mechanosensory neurons with ciliary processes extending into the larval tunic (Torrence & Cloney 1982). Thus, ascidian larvae have a population of cells in the boundary between the neural tube and epidermis from which both neuronal cells and epidermal cells differentiate; however, little is still known about how each dorsal midline epidermal cell chooses one of the two fates. This striking similarity between the dorsal midline epidermis and the neural crest brings up the tantalizing possibility that the vertebrate neural crest may have originated from the dorsal midline epidermis of protochordates (Fig. 3).

image

Figure 2. Gene expression patterns of HrEpiA, HrBMPa and HrPax-37. (A) Schematic illustration of an ascidian tail-bud embryo (lateral view) indicating six epidermal territories. The dorsal midline epidermis is shown in orange. Numbering of the epidermal territories follows that used in Ishida et al. (1996). (B) Expression of HrEpiA (Ishida et al. 1996) in a tail-bud embryo (dorsal view; anterior to the left). Expression is exclusive of the dorsal midline epidermis. (C) Expression of HrBMPa in a tail-bud embryo (dorsal view; anterior to the left). Expression is observed in the dorsal midline epidermis (Miya et al. 1996) and thus is complementary to that of HrEpiA. (D–G) Expression of HrPax-37 in the gastrula (D,E; lateral view) and neurula stages (F,G; dorsal view). HrPax-37 expression at the gastrula stage is observed both in cells fated to the dorsal midline epidermis and in those fated to the dorsal neural tube, but is observed only in dorsal midline epidermis at the neurula stage (Wada et al. 1996a).

Download figure to PowerPoint

image

Figure 3. Hypothetical evolution of the neural crest.

Download figure to PowerPoint

In amphioxus, the dorsal midline epidermal cells also show similarity to the neural crest. Neurulation in amphioxus proceeds differently from that in ascidians or vertebrates and intercellular contact between the epidermis and neural plate is disrupted before the amphioxus neural plate rolls up (Fig. 4; Holland et al. 1996). The epidermal sheets from either side of the plate migrate across the dorsal surface by means of lamellipodia, so that epidermis surrounds the entire embryo prior to closure of the neural plate (Fig. 4; Holland et al. 1996). These migrating dorsal midline epidermal cells with lamellipodia are marked by expression of the Dlx gene (Holland et al. 1996), suggesting that the dorsal midline epidermal cells of the amphioxus are a distinct population that display a similar migratory property to cells in the neural crest. However, in amphioxus, sensory epidermal neurons are rather dispersed, not concentrated at the dorsal midline epidermis (Lacalli & Hou 1999). Details of this difference are discussed later.

image

Figure 4. Neurulation of the amphioxus embryo. (A) Schematic illustration of amphioxus neurulation. The neural plate–neural tube is shown in gray. (B) Dorsal view of a neurula embryo (anterior is to the bottom left). Epidermis is beginning to overgrow the neural plate (np). (C) Close view of cells at the edge of the epidermal sheet (top left). An arrowhead indicates lamellipodia from epidermal cells. (B,C) Reproduced from Holland et al. (1996), with permission of Nick Holland and the Company of Biologists Ltd.

Download figure to PowerPoint

Protochordate homologs of genes involved in the differentiation of the neural crest

The BMP genes, which belong to the transforming growth factor (TGF)-β superfamily, have been proposed to be involved in the induction of the neural crest. Some BMP functions during very early development are to allow differentiation of the ventral phenotype, promote blood and kidney differentiation in the mesoderm and promote epidermal cell differentiation through suppression of neural differentiation in the ectoderm. Neural induction occurs via inhibition of BMP activity (reviewed by Hogan 1996). After neural induction, the expression of BMP4 and BMP7 is upregulated in the boundary between the neural plate and the surrounding epidermis, where neural crest cells differentiate (Liem et al. 1995). The neural crest is induced through the interaction between the epidermis and the neural plate, and arises from both epidermis and neural plate (Selleck & Bronner-Fraser 1995). BMP4 and BMP7 have been shown to be able to substitute the activity of the epidermis in inducing neural crest from the neural plate (Liem et al. 1995). BMPa, an ascidian homolog of BMP5-8 (including BMP7), displays a expression pattern similar to vertebrate BMP7 (Miya et al. 1996). Expression begins predominantly in ectodermal cells at the gastrula stage. The expression of BMPa is upregulated in the dorsal midline epidermal cells of the neurula later in the differentiation process (Fig. 2C; Miya et al. 1996) and is consistent with the idea that the neural crest originates in the dorsal midline epidermis. BMPb, the ascidian homolog of vertebrate BMP2/4, is also expressed in dorsal midline epidermal cells (Miya et al. 1996). Thus, a similar genetic mechanism may operate for differentiation in both the ascidian dorsal midline epidermis and the vertebrate neural crest. However, amphioxus BMP is not upregulated in the dorsal midline epidermis, although it is expressed in epidermal cells of the gastrula (Panopoulou et al. 1998).

Several transcription factors are involved in differentiation of the vertebrate neural crest. The Pax3 mutant mouse gene, Splotch, produces a defect in neural crest differentiation (Franz 1993) and the closely related Pax7 also causes defects in the cephalic neural crest (Mansouri et al. 1996). Pax3 and Pax7 expression is first detected throughout the entire neural plate. Subsequently, their expression is downregulated in the ventral neural tube and maintained in the dorsal neural tube and neural crest cells (Goulding et al. 1993). Pax3 and Pax7 have a single counterpart gene in protochordates. HrPax-37, the ascidian homolog of Pax3 and Pax7, exhibits a pattern of expression similar to its vertebrate counterparts (Wada et al. 1996a, 1997). It is expressed both in cells destined to form the dorsal part of the neural tube and in those destined to form dorsal midline epidermis (Fig. 2D,E). This pattern is comparable with that of vertebrate genes in the dorsal neural tube and neural crest. Just prior to closure of the neural tube, HrPax-37 expression is maintained solely in the dorsal midline epidermis (Fig. 2F,G, Wada et al. 1996a, 1997). The amphioxus homolog of Pax3/7 is expressed in the lateral part of the neural plate, which subsequently occupies the dorsal part of the neural tube and is also found in the somitic mesoderm, although no expression is detected in the epidermis (Holland et al. 1999).

Dlx genes also stand as markers for the neural crest (Robinson & Mahon 1994). AmphiDll, the amphioxus homolog of Dlx, is expressed in the dorsal neural tube and in epidermis adjacent to the neural plate (Holland et al. 1996). An ascidian homolog of Dlx is also expressed in the dorsal midline epidermis (Wada et al. 1999b).

Snail and slug are two closely related zinc finger genes that are among the most widely used markers of the neural crest (Essex et al. 1993; Mayor et al. 1995). Slug is necessary for the induction and subsequent migration of neural crest cells (LaBonne & Bronner-Fraser 2000) and its overexpression leads to expansion of the neural crest cell population (LaBonne & Bronner-Fraser 1998). Snail and slug have one counterpart in protochordates (Corbo et al. 1997; Langeland et al. 1998). The ascidian counterpart gene is expressed in the dorsal part of the anterior neural tube, from which pigment cells differentiate. The posterior neural tube of ascidian larvae consists of four rows of cells. At this level, the snail/slug homolog is expressed predominantly in the lateral row (Corbo et al. 1997). In Ciona intestinalis, a species of ascidian, expression is restricted to the lateral row (Corbo et al. 1997). In another species of ascidian, Halocynthia roretzi, this gene is expressed weakly in the dorsal row (Wada & Saiga 1999). The snail gene is not expressed in ascidian epidermis (Corbo et al. 1997; Wada & Saiga 1999). The amphioxus homolog of the snail/slug gene is expressed in the dorsal region of the neural tube, but not in the epidermis (Langeland et al. 1998).

These comparisons of gene expression patterns are not completely consistent with the idea that the neural crest originates in the dorsal midline epidermis: expression patterns differ significantly between ascidians and amphioxus, as summarized in Table 1. The amphioxus BMP2/4 and Pax3/7 homologs are not expressed in the dorsal midline epidermis (Panopoulou et al. 1998; Holland et al. 1999), possibly correlating with the scattered distribution of the sensory neuronal cells in amphioxus (Lacalli & Hou 1999), while the ascidian sensory neurons are restricted in several fields (Fig. 1A). However, the strong similarities in the patterns of gene activity in the dorsal midline epidermis of ascidians and the vertebrate neural crest are unlikely to be the result of evolutionary convergence. In addition, Dll is expressed in amphioxus dorsal midline epidermis (Holland et al. 1996). Lack of dorsal expression of amphioxus BMP2/4 and Pax3/7 is described most parsimoniously as a secondary loss that took place after the divergence of vertebrates and amphioxus. However, in order to accept this hypothesis of neural crest origin in the dorsal midline epidermis, some reasonable explanation for the amphioxus exception should be made.

Table 1.  Expression of neural crest marker genes in protochordates
 AscidianAmphioxus
 Dorsal neural tubeDorsal midline epidermisDorsal neural tubeDorsal midline epidermis
  1. *Earliest expression is observed in whole epidermis, including the future dorsal midline epidermis. In Ciona intestinalis, the dorsal neural tube expression is only in the anterior part (Corbo et al. 1997). See text for details.

BMP2/4+–*
Pax3/7+++
Dlx+++
Snail++

Another point for consideration is whether the dorsal midline epidermis is the sole origin of the neural crest, or whether the dorsal neural tube also contributes to its development. Ascidian Pax3/7 and amphioxus Dll genes are also expressed in the dorsal part of the neural tube (Holland et al. 1996; Wada et al. 1996a, 1997). Amphioxus Pax3/7 and the snail/slug genes of both ascidians and amphioxus are only expressed in the dorsal neural tube (Corbo et al. 1997; Langeland et al. 1998; Holland et al. 1999; Wada & Saiga 1999). Langeland et al. (1998) suggest that the neural crest originated in the dorsal part of the neural tube, while Holland et al. (1996) suggest that its origin may be found in both the dorsal midline epidermis and the dorsal part of the neural tube. However, it may be inappropriate to look for the origin of the vertebrate neural crest in a distinct cell population. Cell lineage analyses of cells in the vertebrate neural fold reveal that a single cell can give rise to neuronal cells, neural crest cells and epidermal cells (Bronner-Fraser & Fraser 1988, 1989; Garcia-Castro & Bronner-Fraser 1999). Moreover, because neural crest cells generate a wide diversity of cell types and because neural crest cells in the cephalic region have a quite distinct character from those in trunk region, it is not certain whether the neural crest has a single origin (Shimeld & Holland 2000). The main point stressed here is that the boundary between the neural tube and epidermis in protochordates is less well defined than has been believed in the past. In addition, there are cells that show gene activity and cell properties comparable to the vertebrate neural crest. Thus, they perhaps originate the neural crest, possibly with some contribution from the dorsal neural tube (Fig. 3). Because we know that the interaction between neural tissue and the epidermis performs a central role for neural crest induction (Selleck & Bronner-Fraser 1995), we can ask whether a similar interaction is necessary for cell differentiation or specific gene expression in the dorsal midline epidermis of protochordates. This hypothesis can also be tested by comparing the genetic mechanisms for differentiation of the neural crest and dorsal midline epidermis in greater detail.

Evolution of the new cell property of the neural crest

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

Even accepting the presented hypothesis that the origin of the neural crest is in protochordate dorsal midline epidermis, possibly including the dorsal neural tube, remarkable differences remain between the vertebrate neural crest and its protochordate counterparts; for example, cell migratory behavior and pluripotency. Thus, I propose that the evolution of the vertebrate neural crest was not the birth of a new cell type; rather, it was the acquisition of these new cell properties by the protochordate precursors (Fig. 3). In addition to migration and pluripotency, the neural crest acquired another important new property: carriage of anteroposterior positional information. The neural crest cells used to be regarded as playing a pivotal role in patterning of the pharyngeal arch. However, recent studies have revealed that pharyngeal arch patterning occurs at least partially independently of the neural crest cells (Graham & Smith 2001). Even so, because neural crest contributes to skeletal tissues or cranial ganglia, which show a distinct shape or innervation pattern according to their rhombomeric origin, anteroposterior positional information of the neural crest is important for the development of the complex craniofacial structure in vertebrates.

This formation of a complex craniofacial structure is a characteristic feature of vertebrates. Pluripotency contributes to this development by allowing for facial skeleton and peripheral nerve differentiation. Transmission of anteroposterior positional information, together with cell migratory behavior, contributes to the building of a system that controls the coordinated differentiation of several cell types in the pharyngeal arches. In this way, the evolution of these new cell properties was a key innovation for the evolution of the complex craniofacial region in vertebrates. Molecular evolutionary processes involved in the acquistion of these new properties are suggested in the following section.

Pluripotency

The vertebrate neural crest differentiates into several cell types, including pigment cells, peripheral nerve cells inclusive of sensory neurons, peripheral glia and sympathetic neurons. In the cephalic region, it also differentiates into skeletal and connective tissues. Protochordates also possess peripheral nerves and pigment cells. The first stage of neural crest evolution likely affected peripheral nerves, because neural crest cells differentiate at the boundary between the epidermis and neural tissue. Indeed, sensory nerves differentiate from the dorsal midline epidermis in ascidians (Torrence & Cloney 1982). Interestingly, two BMP genes and the Dlx homolog in ascidians are expressed in the ventral midline epidermis, where primary sensory cells also differentiate (Miya et al. 1996, 1997; Wada et al. 1999b). This is rather similar to the nerve plexus in hemichordates. Although acornworms have a dispersed nerve plexus that resembles that of echinoderms, it is concentrated in the dorsal and ventral midlines of the epidermal layer of the trunk (Fig. 1; Knight-Jones 1952). In contrast, amphioxus has a rather scattered peripheral nervous system, with no observable concentration of the peripheral nerves in the dorsal midline (Lacalli & Hou 1999). This hypothesis regards ancestral chordates as having peripheral nerves in the dorsal and ventral midline, as seen in ascidians and acornworms. Subsequently, the dorsal midline evolved into the neural crest. Conversely, amphioxus changed its distribution of peripheral nerves and only Dlx retained its dorsal midline expression.

Pigment cells are found in the neural tubes of both ascidians and amphioxus. Two cell types, otoliths and ocelli, differentiate in the dorsal part of the ascidian brain (Nishida 1987). Ascidians and vertebrates share conserved genetic mechanisms for pigment cell differentiation. For example, Mitf regulates tyrosinase expression and subsequent melanization in both ascidians and vertebrates (I. Yajima & H. Yamamoto, pers. comm., 2001). In addition, the early dorsal neural tube marker gene HrPax-37 (Fig. 2D–G), when overexpressed in the ventral neural tube, has been shown to drive ectopic tyrosinase expression (Wada et al. 1997), which may be comparable to the abnormal pigmentation pattern seen in the mouse Pax3 mutant Splotch (Franz 1993). The fact that ascidian pigment cells differentiate from the dorsal lineage of the neural tube supports the idea that the dorsal part of the protochordate neural tube may also contribute to the neural crest. In addition, it implicates neural crest involvement in pigment cell differentiation at an early stage of evolution. However, amphioxus pigment cells in the neural tube (Hesse ocelli) sit ventrally in the neural tube (Ruppert 1997; S. M. Shimeld, pers. comm., 2001). The origin of pigmentation long predates that of the neural crest, but our current understanding cannot discern how or when the genetic machinery for pigmentation became associated with the neural crest.

Amphioxus has a visceral arch skeleton; however, because its chemical composition is not well understood, its homology with the vertebrate pharyngeal skeleton remains unclear (Holland & Chen 2001). More knowledge of protochordate homologs of genes, such as Cbfa1 or scleraxis, is needed in order to examine the relationships between the neural crest and skeletogenesis.

Shimeld and Holland (2000) suggest that the first step in neural crest evolution may have been the origin of a specific dorsal neural cell population that contributed to sensory processing; subsequently, the downstream genetic pathways for pigmentation and connective tissue were recruited in neural crest differentiation.

Migration

Amphioxus dorsal midline epidermal cells display migration during neurulation, as described previously. However, these cells do not delaminate and the migration proceeds as a sheet of cells. Thus, delamination is unique to the vertebrate neural crest. For this process, the switching of cell adhesion molecules is essential. Cadherin6 and N-cadherin are expressed in premigratory neural crest cells and these cell adhesion molecules must be downregulated in order for neural crest cells to delaminate. Subsequently, distinct cell adhesion molecules, such as cadherin7, are expressed in some neural crest cells (Nakagawa & Takeichi 1995, 1998). Examination of the molecular evolutionary histories of the genes encoding cell adhesion molecules provides a clue as to how this delamination process evolved. Cadherin6 and cadherin7 are closely related genes, which were derived from a single ancestral protochordate gene (Levi et al. 1997; H. Wada, unpubl. data, 2001). Because these two genes likely evolved during the genome duplications of ancestral vertebrates, the genome duplications may have been essential for the evolution of the genetic machinery of neural crest delamination. Similarly, the rhoB gene, which encodes a small G-protein and is upregulated by the BMP gene in the premigratory neural crest, is essential for delamination of the neural crest (Liu & Jessel 1998). Vertebrates contain three types of rho genes (rhoA, B and C), of which only rhoB is involved in delamination (Liu & Jessel 1998). These three genes have a single counterpart in protochordates and gene duplication must have also been essential for one of the rho genes to be co-opted to the delamination process (Wada, unpubl. data, 2001). The histories of the cadherin and rho genes suggest that the genetic machinery of delamination evolved through recruitment of duplicated and thus expendable genes.

Cells at the boundary between the larval neural tube and epidermis in ascidians do not migrate; however, during postmetamorphic development some cells in the nervous system may delaminate and migrate (i.e. from the dorsal strand of the ascidian adult nervous system; Baker & Bronner-Fraser 1997). Whether this behavior is related to the vertebrate neural crest is unclear. Based on the hypothesis that gene duplications were essential for the evolution of the neural crest delamination genetic system, this process may not be controlled by a system homologous with the neural crest.

Anteroposterior positional information

Anteroposterior positional information carried by the neural crest is important for development of the vertebrate craniofacial region. This information is encoded in the pattern of Hox gene expression (Hunt et al. 1991), but the Hox expression pattern in the neural crest is not simply transferred from the neural tube. The anterior limit of Hoxa-2 expression reaches to rhombomere 2, although neural crest cells that migrate from rhombomere 2 quickly inactivate expression of Hoxa-2 (Prince & Lumsden 1994). In addition, cis-regulatory analyses reveal that neural crest expression is, at least partially, regulated independently from that of the neural tube (Maconochie et al. 1999). In contrast, the Hox code in ascidians and amphioxus is restricted to the neural tube (Holland et al. 1992; Katsuyama et al. 1995; Gionti et al. 1998; Locascio et al. 1999; Wada et al. 1999a). Amphioxus Hox genes are expressed in the mesoderm, but this expression is always confined to the posterior region (Holland et al. 1992; Wada et al. 1999a): no Hox code is observed in the dorsal midline epidermis. These observations indicate that the evolution of the new regulatory system for the neural crest was essential to confer the capacity to carry anteroposterior positional information.

Manzanares et al. (2000) challenge this molecular evolutionary background of Hox regulatory mechanisms. They tested putative enhancers of amphioxus Hox genes in mice and chickens by performing transgenic assays. They identified four separate DNA elements with reproducible enhancer activities in transgenic mice in the genomic region around AmphiHox1–3 spanning ~30 kb. The element designated as 1A, which sits 3′ downstream of AmphiHox1, drives expression in the neural tube with an anterior limit in the posterior hindbrain, which is reminiscent of expression driven by the mouse endogenous Hox1 enhancer. Interestingly, element 1A also drives expression in the neural crest. Expression in both the neural crest and neural tube is dependent on retinoic acid (RA) signaling, as demonstrated in an inhibitory experiment using a dominant negative RA receptor as well as by induction of enhanced reporter expression by exogenous RA. This is consistent with the observation that exogenous RA affects AmphiHox1 expression (Holland & Holland 1996). Element 3B, which is located 5′ upstream of Hox3, drives expression in the neural tube, with its anterior limit at the boundary between rhombomeres 6 and 7. Signaling by RA also controls this activity; however, element 3B interestingly does not drive expression in the neural crest.

These results indicate that newly acquired Hox regulation in the neural crest had been achieved through incorporation of the RA signaling system by the neural crest. However, this did not appear to be achieved by the simple addition of new functions in the neural crest for retinoic acid receptor (RAR) or retinoid X receptor (RXR); otherwise, element 3B should also drive expression in the neural crest. Rather, these results can be interpreted as that the activities of elements 1A and 3A are regulated through the RA responsive element (RARE) and additional cis elements. It has been shown that some of the vertebrate RARE need to be accompanied by additional cis elements for their functions (Ogura & Evans 1995a,b). Three models can account for these results (Fig. 5). In the first case, coactivator interaction with element 1A (coactivator B in Fig. 5) drives expression in both the neural tube and neural crest, while coactivator interaction with element 3A (coactivator A in Fig. 5) only drives expression in the neural tube. In this case, it appears that neural crest expression of Hox genes was acquired by co- opting coactivator B for neural crest regulation. Alternatively, element 1A may contain a distinct RARE plus coactivator-binding site for the neural crest and neural tube, respectively. Elements 1A and 3B interact with the same coactivator (coactivator A) for neural tube expression and the other RARE and cis elements for coactivator B of element 1A may function for neural crest expression. Coactivator B may have its original function in the neural tube, because the amphioxus Hox gene is thought to only operate at that site. Alternatively, because Hox1 is expressed with a clear anteroposterior boundary in the epidermis of both amphioxus and ascidians (Katsuyama et al. 1995; Wada et al. 1999a), it may function in the epidermis. In either case, recruitment of the additional coactivator (coactivator B in Fig. C) must have been essential for the evolution of neural crest expression. The earlier results might also be explained in that element 3B contains a neural crest-specific repressor (factor C; Fig, 5C). In this case, we need to ask which kind of roles the neural crest specific repressor (factor C) in amphioxus would play.

image

Figure 5. Activities of putative amphioxus Hox enhancers in vertebrates. In these models, Hox genes have acquired neural crest expression by co-option of the hypothetical coactivator B.

Download figure to PowerPoint

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

A remarkable number of genes are expressed in similar patterns in the vertebrate neural crest and the dorsal midline epidermis of protochordates. The occurrence of these patterns is most pronounced in the ascidian dorsal midline epidermis, which differentiates into both epidermal cells and sensory neuronal cells. These observations suggest that the origin of the neural crest may be in the dorsal midline epidermis, possibly including the dorsal neural tube. Thus, the vertebrate neural crest is likely not a new cell population. The evolution of the neural crest should be re-evaluated and looked at as a series of acquisitions of new cell properties, which include pluripotency, delamination– migration and carriage of anteroposterior positional information. These novel cell properties were essential for the evolution of the finely patterned craniofacial structure, which is a defining vertebrate feature. Molecular evolutionary events that were required for these new cell properties are currently under study. Genome duplications during early vertebrate evolution may have been essential for the evolution of the neural crest cell delamination process. Acquisition of a new role by a coactivator that functions in cooperation with retinoic acid signaling may have achieved a new regulatory mechanism of Hox genes in the neural crest.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References

I thank Clare Baker, Nobue Itasaki, Linda Holland and Nick Holland for critical reading and helpful comments on the manuscript. I also thank Seb Shimeld, Ichiro Yajima and Hiroaki Yamamoto for communicating data prior to publication and Nori Satoh for providing probes of in situ hybridization. I also thank anonymous reviewers for their instructive suggestions. Part of the work presented here was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports and Culture of Japan, and Nissan Science Foundation.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Brief overview of protochordate neural structure
  5. Tetralogy
  6. Origin of the neural crest in protochordates
  7. Evolution of the new cell property of the neural crest
  8. Conclusions
  9. Acknowledgements
  10. References
  • Baker, C. & Bronner-Fraser, M. 1997. The origin of the neural crest. Part II: An evolutionary perspective. Mech. Dev. 69, 1329.
  • Bronner-Fraser, M. & Fraser, S. 1988. Cell lineage analysis reveals multipotency of some avian neural crest cells. Nature 335, 161164.
  • Bronner-Fraser, M. & Fraser, S. 1989. Developmental potential of avian trunk neural crest cells in situ. Neuron 3, 755766.
  • Castresana, J., Feldmaier-Fuchs, G., Pääbo, S. 1998. Codon reassignment and amino acid composition in hemichordate mitochondria. Proc. Natl Acad. Sci. USA 95, 37033707.
  • Corbo, J. C., Erives, A., Gregorio, A. D., Chang, A., Levine, M. 1997. Dorsoventral patterning of the vertebrate neural tube is conserved in a protochordate. Development 124, 23352344.
  • Enver, T. 1999. Pax5 is the deciding factor. Curr. Biol. 9, R933–R935.
  • Essex, L. J., Mayor, R., Sargent, M. G. 1993. Expression of Xenopus Snail in mesoderm and prospective neural fold. Dev. Dyn. 198, 108127.
  • Franz, T. 1993. The Splotch Sp1H and Splotch-delayed Spd alleles: Differential phenotypic effects on neural crest and limb musculature. Anat. Embryol. 187, 371377.
  • Gans, C. & Northcutt, R. G. 1983. Neural crest and the origin of vertebrates: A new head. Science 220, 268274.
  • Garcia-Castro, M. & Bronner-Fraser, M. 1999. Induction and differentiation of the neural crest. Curr. Opin. Cell Biol. 11, 695698.
  • Gionti, M., Ristoratore, F., DiGregorio, A., Aniello, F., Branno, M., DiLauro, R. 1998. CiHox5, a new Ciona intestinalis Hox-related gene, is involved in regionalization of the spinal cord. Dev. Genes Evol. 207, 515523.
  • Goulding, M. D., Lumsden, A., Gruss, P. 1993. Signals from the notochord and floorplate regulate the region-specific expression of two Pax genes in the developing spinal cord. Development 117, 10011016.
  • Graham, A. & Smith, A. 2001. Patterning the pharyngeal arches. Bioessays 23, 5461.
  • Hall, B. K. 2000. The neural crest as a fourth germ layer and vertebrates as quadroblastic not triploblastic. Evol. Dev. 2, 35.DOI: 10.1046/j.1525-142x.2000.00032.x
  • Hogan, B. L. M. 1996. Bone morphogenetic proteins: Multifunctional regulators of vertebrate development. Genes Dev. 10, 15801594.
  • Holland, L. Z. & Holland, N. D. 1996. Expression of AmphiHox1 and AmphiPax1 in amphioxus embryo treated with retinoic acid: Insights into evolution and patterning of the chordate nerve cord and pharynx. Development 122, 18291838.
  • Holland, L. Z. & Holland, N. D. 1998. Developmental gene expresion in amphioxus: New insights into the evolutionary origin of vertebrate brain regions, neural crest, and rostrocaudal segmentation. Am. Zool. 38, 647658.
  • Holland, L. Z., Schubert, M., Kozmik, Z., Holland, N. D. 1999. AmphiPax3/7, an amphioxus paired box gene: Insights into chordate myogenesis, neurogenesis, and the possible evolutionary precursor of definitive vertebrate neural crest. Evol. Dev. 1, 153165.DOI: 10.1046/j.1525-142x.1999.99019.x
  • Holland, N. D. & Chen, J. 2001. Origin and early evolution of the vertebrates: New insights from advances in molecular biology, anatomy, and palaeontology. Bioessays 23, 142151.
  • Holland, N. D., Holland, L. Z., Kozmik, Z. 1995. An amphioxus Pax gene, AmphiPax-1, expressed in embryonic endoderm, but not in mesoderm: Implications for the evolution of class I paired box genes. Mol. Mar. Biol. Biotech. 4, 206214.
  • Holland, N. D., Panganiban, G., Henyey, E. L., Holland, L. Z. 1996. Sequence and developmental expression of AmphiDll, an amphioxus Distal-less gene transcribed in the ectoderm, epidermis and nervous system: Insights into evolution of craniate forebrain and neural crest. Development 122, 29112920.
  • Holland, P. W. H. & Garcia-Fernàndez, J. 1996. Hox genes and chordate evolution. Dev. Biol. 173, 382395.DOI: 10.1006/dbio.1996.0034
  • Holland, P. W. H., Garcia-Fernàndez, J., Williams, N. A., Sidow, A. 1994. Gene duplications and origins of vertebrate development. Development (Suppl.), 125–133.
  • Holland, P. W. H., Holland, L. Z., Williams, N. A., Holland, N. D. 1992. An amphioxus homeobox gene: Sequence conservation, spatial expression during development and insights into vertebrate evolution. Development 116, 653661.
  • Hunt, P., Guilisano, M., Cooke, M. et al. 1991. A distinct Hox code for the branchial region of the vertebrate head. Nature 353, 861864.
  • Ishida, K., Ueki, T., Satoh, N. 1996. Spatio-temporal expression patterns of eight epidermis-specific genes in the ascidian embryo. Zool. Sci. 13, 699709.
  • Jackman, W. R., Langeland, J. A., Kimmel, C. B. 2000. islet reveals segmentation in the amphioxus hindbrain homolog. Dev. Biol. 220, 1626.
  • Katsuyama, Y., Wada, S., Yasugi, S., Saiga, H. 1995. Expression of the labial group Hox gene HrHox-1 and its alteration induced by retinoic acid in development of the ascidian Halocynthia roretzi. Development 121, 31973205.
  • Knight-Jones, E. W. 1952. On the nervous system of Saccoglossus cambrensis (Enteropneusta). Phil. Trans. R. Soc. Lond. 236, 315354.
  • Kozmik, Z., Holland, N. D., Kalousova, A., Paces, J., Schubert, M., Holland, L. Z. 1999. Characterization of an amphioxus paired box gene, AmphiPax2/5/8: Developmental expression patterns in optic support cells, nephridium, thyroid-like structures and pharyngeal gill slits, but not in the midbrain–hindbrain boundary region. Development 126, 12951304.
  • LaBonne, C. & Bronner-Fraser, M. 1998. Neural crest induction in Xenopus: Evidence for a two signal model. Development 125, 24032414.
  • LaBonne, C. & Bronner-Fraser, M. 1999. Molecular mechanisms of neural crest formation. Annu. Rev. Cell Dev. Biol. 15, 81112.
  • LaBonne, C. & Bronner-Fraser, M. 2000. Snail-related transcriptional repressors are required in Xenopus for both the induction of the neural crest and its subsequent migration. Dev. Biol. 221, 195205.
  • Lacalli, T. C. & Hou, S. 1999. A reexamination of the epithelial sensory cells of amphioxus (Branchiostoma). Acta Zool. (Stockholm) 80, 125134.
  • Langeland, J., Tomsa, J. M., Jackman, W. R., Kimmel, C. B. 1998. An amphioxus snail gene: Expression in paraxial mesoderm and neural plate suggests a conserved role in patterning the chordate embryo. Dev. Genes Evol. 208, 569577.
  • LeDouarin, N. M. & Kalcheim, C. 1999. The Neural Crest. Cambridge University Press, Cambridge.
  • Levi, L., Douek, J., Osman, M., Bosch, T. C. G., Rinkevich, B. 1997. Cloning and characterization of BS-cadherin, a novel cadherin from the colonial urochordate Botryllus schlosseri. Gene 200, 117123.DOI: 10.1016/s0378-1119(97)00391-0
  • Liem, K. F. J., Tremml, G., Roelink, H., Jessel, T. M. 1995. Dorsal differentiation of neural plate cells induced by BMP-mediated sugnals from epidermal ectoderm. Cell 82, 969979.
  • Liu, J. & Jessel, T. M. 1998. A role for rhoB in delamination of neural crest cells from the dorsal neural tube. Development 125, 50555067.
  • Locascio, A., Aniello, F., Amoroso, A., Manzanares, M., Krumlauf, R., Branno, M. 1999. Patterning the ascidian nervous system: Structure, expression and transgenic analysis of the CiHox3 gene. Development 126, 47374748.
  • Maconochie, M., Krishnamurthy, R., Nonchev, S. et al. 1999. Regulation of Hoxa2 in cranial neural crest cells involves members of the AP-2 family. Development 126, 14831494.
  • Mansouri, A., Stoykova, A., Torres, M., Gruss, P. 1996. Dysgenesis of cephalic neural crest derivatives in Pax7–/– mutant mice. Development 122, 831838.
  • Manzanares, M., Wada, H., Itasaki, N., Trainor, P. A., Krumlauf, R., Holland, P. W. H. 2000. Conservation and elaboration of Hox gene regulation during evolution of the vertebrate head. Nature 408, 854857.DOI: 10.1038/35048570
  • Martin, A. 2001. Is tetralogy true? Lack of support for the ‘one-to-four rule’. Mol. Biol. Evol. 18, 8993.
  • Mayor, R., Morgan, R., Sargent, M. G. 1995. Induction of the prospective neural crest of Xenopus. Development 121, 767777.
  • Miya, T., Morita, K., Suzuki, A., Ueno, N., Satoh, N. 1997. Functional analysis of an ascidian homologue of vertebrate Bmp-2/Bmp-4 suggests its role in the inhibition of neural fate specification. Development 124, 51495159.
  • Miya, T., Morita, K., Ueno, N., Satoh, N. 1996. An ascidian homologue of vertebrate BMPs5–8 is expressed in the midline of the anterior neuroectoderm and in the midline of the ventral epidermis of embryo. Mech. Dev. 57, 181190.
  • Müller, T. S., Ebensperger, C., Neubüer, A. et al. 1996. Expression of avian Pax1 and Pax9 is intrinsically regulated in pharyngeal endoderm, but depends on environmental influences in paraxial mesoderm. Dev. Biol. 178, 403417.DOI: 10.1006/dbio.1996.0227
  • Nakagawa, S. & Takeichi, M. 1995. Neural crest cell–cell adhesion controlled by sequential and subpopulation-specific expression of novel cadherins. Development 121, 13211332.
  • Nakagawa, S. & Takeichi, M. 1998. Neural crest emigration from the neural tube depends on regulated cadherin expression. Development 125, 29632971.
  • Nishida, H. 1987. Cell lineage analysis in ascidian embryos by intracellular injection of a tracer enzyme. III. Up to the tissue restricted stage. Dev. Biol. 121, 526541.
  • Ogasawara, M., Wada, H., Peters, H., Satoh, N. 1999. Developmental expression of Pax1/9 gene in urochordate and hemichordate gills: Insight into function and evolution of pharyngeal epithelium. Development 126, 25392550.
  • Ogura, T. & Evans, R. M. 1995a. Evidence for two distinct retinoic acid response pathway for HOXB1 gene regulation. Proc. Natl Acad. Sci. USA 92, 392396.
  • Ogura, T. & Evans, R. M. 1995b. A retinoic acid-triggered cascade of HOXB1 gene activation. Proc. Natl Acad. Sci. USA 92, 387391.
  • Panopoulou, G., Clark, M. C., Holland, L. Z., Lehrach, H., Holland, N. D. 1998. AmphiBMP2/4, and amphioxus Bone Morphogenetic Protein closely related to Drosophila decapentaplegic and vertebrate BMP2 and BMP4: Insights into evolution of dorsoventral axis specification. Dev. Dyn. 213, 130139.DOI: 10.1002/(sici)1097-0177(199809)213:1<130::aid-aja13>3.3.co;2-z
  • Prince, V. & Lumsden, A. 1994. Hoxa-2 expression in normal and transposed rhombomeres: Independent regulation in the neural tube and neural crest. Development 120, 911923.
  • Robinson, G. W. & Mahon, K. A. 1994. Differential amd overlapping expression domains of Dlx-2 and Dlx-3 suggest distinct roles for Distal-less homeobox genes in craniofacial development. Mech. Dev. 48, 199215.
  • Ruppert, E. E. 1997. Cephalochordata (Acrania). In Microscopic Anatomy of Invertebrates, vol. 15 (Eds F. W. Harrison & E. E. Ruppert), pp. 349–504. Wiley-Liss, New York.
  • Selleck, M. A. J. & Bronner-Fraser, M. 1995. Origins of the avian neural crest: The role of neural plate–epidermal interactions. Development 121, 525538.
  • Shimeld, S. M. 1999a. The evolution of dorsoventral pattern formation in the chordate neural tube. Am. Zool. 39, 641649.
  • Shimeld, S. M. 1999b. The evolution of the hedgehog gene family in chordates: Insights from amphioxus hedgehog. Dev. Genes Evol. 209, 4047.
  • Shimeld, S. M. & Holland, P. W. H. 2000. Vertebrate innovations. Proc. Natl Acad. Sci. USA 97, 44494452.
  • Spring, J. 1997. Vertebrate evolution by interspecific hybridization – are we polyploid? FEBS Lett. 400, 28.
  • Torrence, S. A. & Cloney, R. A. 1982. Nervous system of ascidian larvae: Caudal primary sensory neurons. Zoomorphology 99, 103115.
  • Turbeville, J. M., Schulz, J. R., Raff, R. A. 1994. Deuterostome phylogeny and the sister group of the chordates: Evidence from molecules and morphology. Mol. Biol. Evol. 11, 648655.
  • Wada, H., Garcia-Fernandez, J., Holland, P. W. H. 1999a. Colinear and segmental expression of amphioxus Hox genes. Dev. Biol. 213, 131141.DOI: 10.1006/dbio.1999.9369
  • Wada, H., Holland, P. W. H., Sato, S., Yamamoto, H., Satoh, N. 1997. Neural tube is partially dorsalized by overexpression of HrPax-37: The ascidian homologue of Pax-3 and Pax-7. Dev. Biol. 187, 240252.DOI: 10.1006/dbio.1997.8626
  • Wada, H., Holland, P. W. H., Satoh, N. 1996a. Origin of patterning in neural tubes. Nature384, 123.
  • Wada, H., Saiga, H., Satoh, N., Holland, P. W. H. 1998. Tripartite organization of the ancestral chordate brain and the antiquity of placodes: Insights from ascidian Pax-2/5/8, Hox and Otx genes. Development 125, 1131122.
  • Wada, H. & Satoh, N. 1994. Details of the evolutionary history from invertebrates to vertebrates, as deduced from the sequences of 18S rDNA. Proc. Natl Acad. Sci. USA 91, 18011804.
  • Wada, H. & Satoh, N. 2001. Patterning the protochordate neural tube. Curr. Opin. Neurobiol. 11, 1621.
  • Wada, S., Katsuyama, Y., Saiga, H. 1999b. Anteroposterior patterning of the epidermis by inductive influences from the vegetal hemisphere cells in the ascidian embryo. Development 126, 49554963.
  • Wada, S., Katsuyama, Y., Sato, Y., Itoh, C., Saiga, H. 1996b. Hroth, an orthodenticle-related homeobox gene of the ascidian, Halocynthia roretzi: Its expression and putative roles in the axis formation during embryogenesis. Mech. Dev. 60, 5971.
  • Wada, S. & Saiga, H. 1999. Cloning and embryonic expression of Hrsna, a snail family gene of the ascidian Halocynthia roretzi: Implication in the origins of mechanisms for mesoderm specification and body axis formation in chordates. Develop. Growth Differ. 41, 918.
  • Williams, N. A. & Holland, P. W. H. 1996. Old head on young shoulders. Nature383, 490.