Portal de Periódicos da CAPES

Linking Chordate Gene Networks to Cellular Behavior in Ascidians

2006; Cell Press; Volume: 124; Issue: 2 Linguagem: Inglês

10.1016/j.cell.2006.01.013

1097-4172

Brad Davidson, Lionel Christiaen,

Marine Biology and Environmental Chemistry

Embryos of simple chordates called ascidians (sea squirts) have few cells, develop rapidly, and are transparent, enabling the in vivo fluorescent imaging of labeled cell lineages. Ascidians are also simple genetically, with limited redundancy and compact regulatory regions. This cellular and genetic simplicity is now being exploited to link comprehensive gene networks to the cellular events underlying morphogenesis. Embryos of simple chordates called ascidians (sea squirts) have few cells, develop rapidly, and are transparent, enabling the in vivo fluorescent imaging of labeled cell lineages. Ascidians are also simple genetically, with limited redundancy and compact regulatory regions. This cellular and genetic simplicity is now being exploited to link comprehensive gene networks to the cellular events underlying morphogenesis. For more than 100 years, the value of simple ascidian embryos for investigating conserved chordate development has been recognized. Ascidians are marine invertebrates with a free-swimming tadpole larval stage. They have two defining chordate tissues: a rigid rod of notochord cells that underlies a dorsal neural tube. Ascidian embryos are highly amenable to experimentation: they develop rapidly and their blastomeres are easily separated and cultured. In addition, pigmented granules associated with maternal determinants (proteins derived from maternal transcripts in the oocyte) serve as innate lineage markers. These properties allowed Chabry and Conklin to record profound insights into the developmental function of cytoplasmic determinants that are now being characterized in molecular detail (Sardet et al., 2005Sardet C. Dru P. Prodon F. Biol. Cell. 2005; 97: 35-49Crossref PubMed Scopus (51) Google Scholar). More recent studies have focused on the gene networks underlying chordate structures, including the notochord and the dorsal neural tube. The lack of redundant paralogs in ascidians and the ease of characterizing regulatory DNA through transient transfection have permitted rapid progress in the delineation of gene networks. The availability of genome sequences for two related ascidian species, Ciona intestinalis and Ciona savignyi, has greatly increased the power and appeal of using ascidian embryos to study developmental gene networks. Although characterization of gene expression and its regulation in the ascidian embryo has reached an exceptional level of detail (Imai et al., 2004Imai K.S. Hino K. Yagi K. Satoh N. Satou Y. Development. 2004; 131: 4047-4058Crossref PubMed Scopus (332) Google Scholar), the cellular events by which these genetic networks influence developmental processes remains largely unexplored. In developmental studies, cells are traditionally treated as components of larger structures within the embryo (the animal hemisphere, the germ layers, the notochord, etc.). A growing number of studies, however, are highlighting the integral role of the cell as the ultimate effector of morphogenesis. Here, we discuss the value of ascidian embryos for studying development at the cellular level, linking comprehensive gene networks to the cellular mechanisms underlying tissue and organ formation. Ascidian embryogenesis involves a set of morphogenetic processes strikingly similar to those of vertebrates (see Figure 1) but involving extraordinarily low cell numbers (Nishida, 2005Nishida H. Dev. Dyn. 2005; 233: 1177-1193Crossref PubMed Scopus (125) Google Scholar). Fertilization leads to extensive rearrangements of cytoplasmic determinants that define the future embryonic axis of the ascidian larva (see Figure 1A), including the nuclear import of β-catenin on the future dorsal side. As in fish and frogs, gastrulation begins dorsally with involution of the endoderm precursor cells, followed by a ring of marginal mesoderm and completed by spreading of the epidermis to enclose the underlying layers. However, ascidians undergo gastrulation with only 110 cells, the majority having already undergone fate restriction (see Figure 1B). The ascidian neural plate consists of less than fifty cells and yet undergoes a highly conserved process of neural tube formation, consisting of elongation through medio-lateral intercalation (termed convergent extension) and elevation of the lateral borders to form neural folds (see Figure 1C–1E). The notochord rudiment, initially positioned just posterior to the neural lineage (Figure 1B), involutes anteriorly, coming to lie beneath the forming neural plate (Figure 1D). Formation of the 40-cell notochord consists of highly conserved cellular behaviors driving convergent extension (see Figure 2) . As the elongating notochord drives tail extension, the heart precursor cells (in Ciona there are only four) migrate to fuse along the ventral midline, mirroring the movement of vertebrate heart cells during vertebrate embryogenesis (Figure 1F).Figure 2Notochord Development in the Ascidian EmbryoShow full caption(A) Simplified version of the early gene network leading to (B) specification of the notochord by localized expression of Brachyury at the 64-cell stage. (C) Brachyury in turn activates regulators of cellular activity, such as prickle, which governs notochord cell polarity. The resulting cell behavior eventually leads to convergent extension of the notochord in the gastrula and neurula stages. (A and B: vegetal views of the 32- and 64-cell stage embryos; C: dorsal views of gastrula and neurula stage embryos, cell positions from confocal sections; bp, blastopore. Anterior is up in all schemes, and the notochord lineage is highlighted in pink.)View Large Image Figure ViewerDownload Hi-res image Download (PPT) (A) Simplified version of the early gene network leading to (B) specification of the notochord by localized expression of Brachyury at the 64-cell stage. (C) Brachyury in turn activates regulators of cellular activity, such as prickle, which governs notochord cell polarity. The resulting cell behavior eventually leads to convergent extension of the notochord in the gastrula and neurula stages. (A and B: vegetal views of the 32- and 64-cell stage embryos; C: dorsal views of gastrula and neurula stage embryos, cell positions from confocal sections; bp, blastopore. Anterior is up in all schemes, and the notochord lineage is highlighted in pink.) Rapid and invariant specification of ascidian cell lineages by maternal determinants has led to their traditional classification as mosaic embryos in contrast to the more regulative development of vertebrate embryos, in which specification is driven by contact with neighboring cells rather than by determinants. However, recent findings are blurring this distinction. Many ascidian tissues, including the notochord and neural tube, rely predominantly on early inductive interactions. It is also increasingly evident that regulative development in vertebrate embryos overlies substantial pre-patterning by maternal determinants (for example, see Dupont et al., 2005Dupont S. Zacchigna L. Cordenonsi M. Soligo S. Adorno M. Rugge M. Piccolo S. Cell. 2005; 121: 87-99Abstract Full Text Full Text PDF PubMed Scopus (259) Google Scholar). Thus studies of ascidian embryos can provide insights into conserved chordate mechanisms underlying both the localization of maternal determinants and subsequent regulative events. Conklin first characterized the posterior-vegetal egg cytoplasm, or myoplasm, which systematically segregates into tail muscle cells (see Figure 1). Blastomere isolation and cytoplasm transplantation experiments demonstrated that a muscle determination factor is associated with the myoplasm. As ascidian developmental biology entered the molecular age, differential screening and in situ hybridization techniques have led to the identification of a number of RNA molecules showing posterior-vegetal localization ("posterior end mark" or PEM genes). One of these, macho-1, was shown to act as the muscle determinant (Nishida and Sawada, 2001Nishida H. Sawada K. Nature. 2001; 409: 724-729Crossref PubMed Scopus (223) Google Scholar). However, macho-1 RNA does not cosegregate with the myoplasm and appears to exert its effect during initial cleavage stages, when it is present in muscle precursors. Using high-resolution in situ hybridization and organelle labeling techniques, Prodon et al., 2005Prodon F. Dru P. Roegiers F. Sardet C. J. Cell Sci. 2005; 118: 2393-2404Crossref PubMed Scopus (33) Google Scholar showed that macho-1 and other PEM RNAs colocalize with a thin layer of cortical rough endoplasmic reticulum that is tethered to the egg plasma membrane and becomes progressively concentrated at the posterior-vegetal pole. Recently, Nakamura et al., 2005Nakamura Y. Makabe K.W. Nishida H. Development. 2005; 132: 4731-4742Crossref PubMed Scopus (29) Google Scholar showed that a kinase, POPK-1, is required for concentration in the cortical rough endoplasmic reticulum and for subsequent localization of macho-1 and other PEM RNAs. Strikingly, disruption of either POPK-1 or macho-1 translation by injection of morpholinos (modified antisense oligonucleotides) leads to similar malformations, and macho-1 mRNA can rescue the POPK-1 phenotype. Thus, proper localization of macho-1 RNA is strictly required for axis determination as well as muscle formation. Although macho-1 is particular to the ascidians, other cytoplasmic determinants may have more conserved roles. As in vertebrates, localized activity of maternal β-catenin drives endoderm specification at the vegetal pole (see Figure 2). Possible homology extends to the induction of marginal mesoderm by the vegetal endoderm and particularly the activation of the crucial notochord transcription factor Brachyury. Because the notochord is a defining trait of the chordates, its development has been the subject of extensive research in ascidians. Much of this research has focused on the gene network surrounding the T box transcription factor Brachyury (see Figure 2). Inputs controlling Brachyury expression have been characterized extensively through detailed analysis of its cis-regulatory DNA and disruption of candidate regulators (Corbo et al., 2001Corbo J.C. Di Gregorio A. Levine M. Cell. 2001; 106: 535-538Abstract Full Text Full Text PDF PubMed Scopus (136) Google Scholar). Additionally, a subtractive hybridization screen identified 39 putative Brachyury target genes (Takahashi et al., 1999Takahashi H. Hotta K. Erives A. Di Gregorio A. Zeller R.W. Levine M. Satoh N. Genes Dev. 1999; 13: 1519-1523Crossref PubMed Scopus (170) Google Scholar). Many of these genes belong to molecular families with distinct roles in cell behavior. The seminal studies of Munro and Odell provided detailed characterization of the behavior of cells forming the notochord (Munro and Odell, 2002aMunro E.M. Odell G. Development. 2002; 129: 1-12PubMed Google Scholar, Munro and Odell, 2002bMunro E.M. Odell G.M. Development. 2002; 129: 13-24PubMed Google Scholar). The ability to isolate, culture, and combine specific embryonic fragments allowed the authors to separate the intrinsic versus extrinsic requirements for proper notochord formation. The transparency and simplicity of ascidian embryos permitted in-depth imaging of the in vivo behavior of notochord cells. Confocal analysis of phalloidin-stained embryos revealed that directional protrusive activity of notochord cells correlated precisely with oriented actin-rich extensions. This exceptional analysis enabled mathematical modeling of the mechanisms driving the fundamental developmental process of convergent extension of the notochord. Mutational analysis of prickle, a gene switched on by Brachyury, has provided a link between the notochord gene network and cell behavior (Jiang et al., 2005Jiang D. Munro E.M. Smith W.C. Curr. Biol. 2005; 15: 79-85Abstract Full Text Full Text PDF PubMed Scopus (129) Google Scholar). In-depth in vivo comparisons of mutant and wild-type notochord cell behavior demonstrated that prickle is required for the directional protrusive activity underlying convergent extension. The ability to readily generate transgenic embryos through transfection of zygotes by electroporation has permitted the use of a Brachyury enhancer to drive notochord-specific expression of tagged proteins. This technique revealed that an asymmetric distribution of prickle is required for localization of the planar cell polarity protein dishevelled and subsequently for directional motility underlying proper convergence (see Figure 2). Thus, ascidian notochord development can be understood as a cohesive process beginning with maternal determinants shaping localized expression of Brachyury and ending with directed cellular processes driving morphogenesis. Further studies of the precise roles of other predicted Brachyury target genes would make ascidian notochord formation one of the best characterized developmental processes. Convergent extension of the notochord is just one of many fundamental morphogenetic processes for which the genetic and cellular simplicity of ascidian embryos can be productively exploited. Asymmetric cell divisions are predominant during early embryogenesis. Coordinated changes in the shape and movement of epithelia are central to ascidian gastrulation, neural tube formation, and differentiation of the gastrointestinal tract. Interactions between the epithelium and mesenchyme occur during development of the placodes (the thickened ectodermal rudiments of sensory organs). Primordial heart, blood, and germ cells undergo extensive migrations. From a cellular perspective, these morphogenetic events result from the integration and coordination of subcellular processes including cytoskeletal and membrane dynamics, adhesion, and polarity. High-resolution imaging and labeling techniques must be used to achieve a detailed understanding of the cellular basis of these processes. Here lies a major advantage of ascidian embryos. Reduced cell numbers allow complex processes to be studied at the single-cell level within the context of the whole embryo. In Ciona intestinalis, targeted expression of fluorescent-tagged proteins can be readily achieved using a comprehensive set of tissue-specific enhancer elements. Thus, the detailed behavior of cells within distinct lineages can be characterized in vivo, either singly or in relation to other lineages (Rhee et al., 2005Rhee J.M. Oda-Ishii I. Passamaneck Y.J. Hadjantonakis A.K. Di Gregorio A. Genesis. 2005; 43: 136-147Crossref PubMed Scopus (21) Google Scholar). An emerging characterization of the genes involved in the formation of the ascidian neural tube, tail muscle, and heart make these developmental processes primary candidates for the type of comprehensive gene-to-cell analysis that has been carried out for the notochord. Efforts to decipher the gene network underlying ascidian neurogenesis have centered on the regulation of the conserved neural transcription factor Otx and inductive roles for fibroblast growth factor and for the Nodal signaling pathway (Bertrand et al., 2003Bertrand V. Hudson C. Caillol D. Popovici C. Lemaire P. Cell. 2003; 115: 615-627Abstract Full Text Full Text PDF PubMed Scopus (254) Google Scholar). In the tail muscles, macho-1 directly activates Tbx6 transcription factors, which appear to have a conserved role in the initial specification of chordate skeletal muscle (Yagi et al., 2005Yagi K. Takatori N. Satou Y. Satoh N. Dev. Biol. 2005; 282: 535-549Crossref PubMed Scopus (62) Google Scholar). The Tbx6 genes are also critical for ascidian, and possibly vertebrate, heart development, directly activating expression of the conserved early heart factor Mesp (Davidson et al., 2005Davidson B. Shi W. Levine M. Development. 2005; 132: 4811-4818Crossref PubMed Scopus (81) Google Scholar). The use of comprehensive microarrays to identify downstream targets of specification factors for the neural tube (Otx, neurogenin, Msx, Chordin, Pax 3/6, and Sox), tail muscles (macho-1, Zic-L, Tbx6b, myoD), and heart (Mesp, Nkx, Hand, Gata) will expand these genetic networks, connecting transcription factors to genes with direct roles in cell behavior. Systematic characterization of gene networks and resulting cellular behaviors provide the framework for in-depth analysis of gene function. Mutational analysis and morpholino knockdown experiments provide relatively unambigous results in Ciona due to a general lack of genetic redundancy. Additionally, characterized enhancers can be used to disrupt gene function in specific lineages through targeted expression of altered proteins. The effects of such manipulations on individual cell behaviors within the targeted lineages as well as in neighboring lineages can be simultaneously monitored in vivo by cotransfection with fluorescent lineage markers. Such an approach was recently used to study the function of Mesp in heart development, uncoupling heart cell migration and specification (Davidson et al., 2005Davidson B. Shi W. Levine M. Development. 2005; 132: 4811-4818Crossref PubMed Scopus (81) Google Scholar). Over the next several years, elaboration of complete ascidian gene regulatory networks will be accompanied by detailed visualization of dynamic cellular processes. This information will be integrated with targeted manipulations of relevant genes to assess their precise roles in cell behavior and resulting morphogenetic processes. A comprehensive understanding of ascidian developmental processes from gene networks to cell behaviors promises to generate profound insights into related processes in vertebrates.