1. Project summary
This project aims to develop genomic and post-genomic approaches to the genetics of development in two species which occupy key phylogenetic positions in the vertebrates, the lamprey, L. fluviatilis and the dogfish, S. canicula. The general goal of this study is to define the molecular basis for the unity of this group as well as the variations which could explain its diversification. Specifically we will attempt :

• to resolve the points which are still unclear in the phylogeny of the chordates,
• to explain the exact chronology of the mechanisms of gene duplications which have occurred in the vertebrates as well as the mode of evolution of the corresponding multigenic families and
• to compare the genetic networks which control several developmental processes between the large vertebrate groups.

Up

2. Previous findings of the groups involved
During the last few years we have been evaluating the possibility of using the lamprey and the dogfish for comparative analyses of the genetic mechanisms which control embryonic development; we have confirmed the relevance of these species. Our results concern two principal aspects, evolution of multigenic families in vertebrates, which will be analyzed by studying model systems, and the evolution of several developmental processes. In the first case, detailed comparative analyses have permitted us to define the composition and the mode of evolution of several model multigenic families which code for homeodomain transcription factors (Emx, Otx). The results obtained have demonstrated in two cases that the fixation of orthologic classes preceded the radiation of the gnathostomes. Comparative analyses of expression profiles have underlined the essential but non-exclusive role of regulatory signals in -cis in the functional evolution of these families and the complexity of the processes involved, which implicate " function shuffling ", such as recruitment to new expression territories. The study of another family of homeodomain genes (Evx) has also been initiated in the dogfish, in which several sequences have been cloned. Furthermore, our work has shown that the study of embryonic expression profiles of genes implicated in early regionalization of the forebrain (Otx, Emx, Pax6, Pax3, Lhx1/5, Lhx2/9) has shown that the majority of the neuromeric subdivisions characterized in the bony fish (osteichthyes) are found in these two groups. Important differences have nevertheless been observed: thus the variation present in jawed vertebrates, Nkx2.1, is not expressed in the telencephalon in the lamprey. We have also characterized the embryonic expression profile in lampreys of several genes which play essential roles during gastrulation in bony fish (Fox-a@, Lim1, Otx, Brachyury), and also provided the first molecular characterization of gastrulation in a cartiliaginous fish (Chondrichthyes). These first analyses have furnished the basis for much more detailed studies, which are currently being performed using new markers which have recently been amplified in our laboratories.

Up

Research project
The emergence of the vertebrates was marked by the appearance of numerous morphologic and physiologic innovations: presence of a skull, a chambered heart and kidneys, an elaborate brain with five well-defined subdivisions in the embryonic state, paired sensory organs, individualization of cell populations with specific migration properties or differentiation complexes during development, such as placodes or neural crests. This group of animals is further characterized by a very great morphologic diversity, associated with adaptations to very diverse environments and life-styles (appearance of jaws in gnathostomes, evolution of the structure of the inner ear in tetrapods during adaptation to life on land, various adaptations of the limb). The molecular bases of these changes, such as in the transition between protochordates and craniates as well as transitions in vertebrates, remain poorly understood. Several authors have advanced the hypothesis of a link between morphologic complexification of the vertebrates and the expansion of their genome by gene duplications. It is generally accepted today that the emergence of the vertebrates is correlated with the expansion of a very large number of gene families, which are present as single copies in the protochordates. According to classical scenarios for the evolution of multigenic families (neofunctionalization), the effect of gene duplication is to remove the constraints of selection on one of the gene copies, which can thus evolve to develop new functions as evolutive innovations, while the other gene conserves the ancestral function. No clear example of this mode of evolution has been described for the origin of the vertebrates in this specific context, however. On the other hand, other models, with very different functional consequences have been proposed. For instance, a study of gene duplications which have occurred recently in the actinopterygians has revealed that the first effect of this type of genetic event is a distribution of ancestral functions between paralogs, which does not involve any functional innovation. Furthermore, numerous molecular mechanisms other than gene duplications, such as modifications of expression territories, acquisition of new structural domains by transcription factors, or alternative splicing of transcripts, may lead to morphologic innovations as studies of several genetic systems in protostome models have demonstrated.
The study of genetic mechanisms of embryonic development are currently providing a renewed approach to the evolution of morphology in the metazoans. Although direct experimental approaches which can be used to study the processes of micro-evolution are inapplicable to the problem of the emergence and the diversification of vertebrates, comparative analyses between large groups from this phylum and the prochordates may provide the opportunity to clarify the changes which occurred at transition periods in their evolution. During the last 20 years, functional analyses of embryonic development in vertebrate model organisms have led to considerable progress in the comprehension of the genetic mechanisms of development in bony fish. This information constitutes an indispensable reference, but it is insufficient in a comparative perspective because two other large groups of vertebrates have been excluded, the cyclostomes and the chondrichthyes. Our project aims to develop the necessary tools for molecular characterization of embryonic development in a cyclostome, the lamprey, Lampetra fluviatilis, and a chondrichthian, the dogfish, Scyliorhinus canicula, by analysis of the transcriptome in both species and the construction of a BAC library in the dogfish. This strategy should allow us to attain three principal objectives :

• to resolve definitively the problems which remain in the phylogeny of the chordates,
• to characterize the modes of evolution of the multigenic families frequently encountered in the genomes of vertebrates during the study of model systems and
• to define the genetic networks which control the developmental processes studied by collaborating groups in order to evaluate their conservation and demonstrate eventual variations, notably those which may be linked to morphologic innovations.

Up

4. Objectives of the project

Phylogeny of the chordates
In order to understand the origin and the diversification of the metazoans, the evolutionary relationships between the animal phyla must be reliably established. In the taxon of the deuterostomes, several very important nodes are still poorly resolved, such as the monophyletic nature of the cyclostomes and the taxonomic relationships between the chordates, cephalochordates and urochordates. For several years we have been developing a phylogenomic approach in order to improve the phylogenetic inferences. Informatics tools have been elaborated and validated in previous studies of the phylogenetic position of organisms such as a choanoflagellate (Monosiga ovata) and an amoeba (Mastigamoeba balamuth) by using EST sequences of these organisms (Bapteste et al., 2002, Proc Natl Acad Sci USA, 99:1414-1419; Philippe et al., 2004, Mol Biol Evol. 21:1740-1752). The sequencing of ESTs from the lamprey (Lampetra fluviatilis) and the dogfish shark (Scyliorhinus canicula) will enrich our taxonomic sampling with two key organisms for which little information is presently available (project developed by D. Casane in collaboration with H. Philippe).

Up

Evolution of multigenic families in vertebrates.
Chronology and mechanism of gene duplications in vertebrates. The hypothesis of the role of gene duplications in the emergence of the vertebrates is mainly based on comparisons between bony fish and cephalochordates, and few studies are available in chondrichthyians and cyclostomes. However, some studies have indicated a good correspondence between classes of orthologous genes identified in bony fish and chondrichthyians. On the other hand, for the majority of genetic systems studied, no clear relationship has emerged between classes of genes identified in gnathostomes and genes characterized in the cyclostomes, which are generally very divergent. Although it is clear that the gene duplications observed in bony fish took place before the radiation of the gnathostomes and after the separation between the cephalochordates and the craniates, the precise chronology of these events is poorly understood. This uncertainty is mainly due to the lack of sequence data from cyclostomes and chondrichthyians. The mechanism of duplications is also subject to controversy because the current hypothesis of two successive phases of polyploidization has not yet been clearly demonstrated. The first part of the project will therefore have the objective of defining the phylogenetic relationships of several multigenic families from cyclostomes and gnathostomes, by determining and analysing much more extensive sets of sequences than those that are presently available (project developed by D. Casane in collaboration with H. Philippe).

Mode of evolution of model multigenic families. A second aspect of this study aims to study the mode of evolution of three model multigenic families, the Hox, Dlx and Evx genes, by a systematic comparative analysis of embryonic strucures and expression profiles across a large spectrum of vertebrates. These genes have been chosen for three principal reasons. Firstly, the corresponding families for which the composition is well characterized in osteichthyians (mammals and zebrafish) and the cephalochordates provide clear examples of duplications in the vertebrates. The data are currently very incomplete for the cyclostomes and chondrichthyians, however. Also, the avaliable sequences indicate large variations in the rate of evolution between the various classes, for which the biological significance is unknown. Finally, comparisons of embryonic expression profiles between cephalochordates and osteichthyians suggests considerable complexification of the function of these genes in vertebrates.A large number of these functions involve morphogenetic processes corresponding to innovations which occurred in certain groups of vertebrates (for example, the role of Hox genes in the development of paired limbs, the role of Evx genes in the formation of pharyngeal teeth and rayed fins, and the role of Elx genes in the formation of placode derivatives and neural crests). The challenge here is to correlate precisely gene duplications, modification of their embryonic expression territories and the morphogenetic events which depend on them (project developed by D. Casane).

Up

Evolution of developmental processes in the vertebrates
Gastrulation. Although it is now clear the the genetic control of gastrulation involves networks which have been conserved over a large evolutionary scale, the precise extent of conservation of genetic and cellular interactions is far from established, even within a group of relatively closely related species such as the vertebrates. This situation is linked to the extensive divergence of morphologies, the difficulties in comparing different functional approaches, and also in the complexity of the evolution of certain genetic systems characterized by duplications or losses of genes, and accelerations of the rate of evolution in certain taxons. In our previous studies, the little dogfish, S. canicula has proven to be the model of choice for the study of the evolution of the mechanisms of gastrulation in vertebrates, not only because of its transitional phylogenetic position in the gnathostomes but also because it has a morphology which is especially easy to interpret and it exhibits an unexpected similarity to embryos of certain amniotes (reptiles, chick). The study of this species may thus lead to a better understanding of the relationships and mechanisms of transition between the modes of gastrulation used by the principal model species that have been studied (Saula-Spengles et al., 2003, Dev. Biol. 264, 296-307). We will also seek to extend the molecular characterization of the gastrula of the dogfish in order to define eventual territorial homologies by focusing on organizer regions but also on extraembryonic tissues which, in amniotes, are sources of secreted signals which are essential for the development of the embryo. For this we will utilize a transcriptome approach, possibly complemented by the degenerate PCR technique. We will complete these molecular descriptions by the implementation of approaches based on heterologous transgenesis. In this case, BACs from the dogfish, which carry marker genes for well-defined territories in this species, will be integrated into the murine genome, and their expression profile will be tested in the murine context. The Otx5 gene of the dogfish will provide a particularly interesting system for this approach. This gene is expressed at very early stages of gastrulation at the level of the dorsal margin of the blastoderm, which may be homologous to Koller's (Rauber's) sickle or the posterior marginal zone in the chicken (project developed by S. Mazan and D. Casane).

Evolution of the anterior brain in vertebrates. We will attempt to develop a descriptive approach with the goal of systematically comparing the genetic mechanisms implicated in early specification and regionalization of the anterior brain in cyclostomes and chondrichthyians. Molecular studies have suggested that certain main subdivisions of the central nervous system, notably along the antero-posterior axis, are probably conserved in the ensemble of chordates. However, the brain, and in particular the anterior brain, presents a complexification and considerable diversification in the vertebrates, both morphologic and functional. The results obtained to date, essentially in osteichthyians, suggest good conservation of the early regional specification. This is accompanied, however, by subtle changes in the expression profiles of genes implicated in the process of cellular differentiation. The proposed project aims to define the genetic network conserved in the ensemble of the vertebrates, which determines the unity of the prosencephalon in this group, but also the variations which lead to the obvious morphologic differences between the vertebrate groups. This analysis will be performed by systematically studying expression profiles of genes in the dogfish and in the lamprey which code for transcription factors, which are well-characterized in the osteichthyes and implicated in early specification of the anterior neuroectoderm (Hex, Hesx-1, Fox-a2, goosecoid genes), in the regionalization of the neural tube (Pax, Dlx, Emx, Nkx, Tb Gsh, Lef, Ngn, TBr genes), in signaling pathways which play an important role in this regionalization (Bmp, Fgf, Wnt, Shh), and in the processes of neuronal differentiation (LIM-hd).This approach, using the dogfish and lamprey models, should lead to an evaluation of the conservation of embryonic expression profiles of the genes studied, to the identification of potential organizers of the prosencephalon, and to an understanding of how major innovations were produced at the base of the vertebrate group. To cite an example, the lamprey does not have a pallidum, which is a major component of the subpallium in tetrapods, implicated in the programming of voluntary movement. Analysis in our families of genes for regional and cellular specification in the subpallium of the lamprey and the dogfish will lead to proposal of a scenario for the emergence of a cerebral structure and of the function associated with it. Finally, the contribution of the genomic libraries will complement the study of regulatory sequences and their evolution, making it possible to refine the proposed evolutionary scenarios with the help of transgenesis and studies of the promoters (project developed by S. Retaux and S. Mazan, in collaboration with I. Rodriguez-Moldes, M.C. DiRodicio and M. Pombal).

Formation of cortical nervous structures in the Chondrichthyians. The structures of the brain of the vertebrates are spatially organized in two main modes: nuclear (the cell bodies are arranged in clusters) and cortical (arrangement in layers). The principal cortical structures are essentially dorsal, with the notable exception of the olfactory bulb: the cerebellum and the roof of the mesencephalon (optic tectum/superior colliculus and the semicircular torus/inferior colliculus) in all the gnasthostomes, and in mammals only, the pallium (dorsal telencephalon, which gives rise to the cerebral cortex in this group). Concerning the optic tectum at least, this structure is anatomically (arranged in layers) and physiologically (function in radial columns) similar in the actinopterygians and the sarcopterygians; on the other hand, the mechanisms are very different in the two groups (actinoptergyians have a marginal mitotic zone with continuous growth--Nguyen et al., 1999, J. Comp. Neurol. 413, 385-404, whereas sarcopterygians have a ventricular neuroepithelium underlying the structure, with a function which is limited in time). Nothing is known about tectal morphogenesis in the chondrichthyians. We have isolated a large number of genes implicated in tectal morphogenesis in the medaka (a teleost fish) during a systematic screening by hybridization in situ (Nguyen et al., 2001. Mech. Dev. 107, 55-67). Analysis of the expression of some of their orthologs in the dogfish will be very useful for the comprehension of the mechanisms of morphogenesis of cortical structures (optic tectum, therefore, and possibly the cerebellum) in this chondrichthyian, and will clarify the evolutionary origin of the mechanisms of corticalization in vertebrates. (Project developed by F. Bourrat, in collaboration with S. Retaux and I. Rodriguez-Moldes).

Organogenesis: origin and evolution of the molecular mechanisms of organogenesis of the internal skeleton and dental structures in vertebrates. The internal skeleton and teeth are highly informative characteristics for any study of the evolutionary history of the vertebrates. Indeed these hard tissues are evolutive innovations exclusively present in this taxon which, furthermore, demonstrates an extraordinary morphologic diversity and constitutes the essential feature of the paleontologic data. The phylogenetic position of the dogfish and its anatomy are key subjects for research into questions which have been controversial for a long time. For example, the classical hypothesis states that the origin of oral teeth involves recruitment in the jaw of genetic cascades which gave rise to dermal denticles (present at the surface of the skin of present chondrichthyians). However, a newly-proposed hypothesis tends to prefer a "pharyngeal origin" of the molecular mechanisms responsible for the dental structures (Smith et al., 2003. Evol Dev. 5:394-413). Comparison of the regulation of the development of oral teeth in the mouse (very well described) and pharyngeal teeth of the zebrafish (very poorly understood; we are studying this along with other groups) and dermal denticles (no data at present) should lead to very interesting clarifications of this problem. The ancestral organization of fins, their evolution (especially as the pentadactyl limb of tetrapods) and the problems of homology of associated structures are also very old questions which may progress enormously thanks to a knowledge of the molecular mechanisms implicated in the formation of the fins of a cartilaginous fish. The characterization of the transcriptome at the moment when the internal skeleton and the dental structures of the dogfish are developing will enable us to identify and study the expression patterns of candidate genes for a role in the development of these structures. (Project developed by D. Casane in collaboration with M. Smith).

Up