The study of metazoan phylogeny using molecular data has led to major reorganizations in our conception of the animal phylogenetic tree. The bilaterians (bilaterally symmetrical triploblastic animals) are a monophyletic group, divided into three great branches: the deuterostomes and two large groups of protostomes, the ecdysozoans and the trochozoans. Between insects and vertebrates exist a striking number of similarities in the genetic regulation of some key-steps in the ontogeneses of structures and organs that were generally thought to have evolved several times independently in metazoans. To explain these similarities, it has been proposed that these structures and organs were present in the last common ancestor of all protostomes and deuterostomes, an annelid-like coelomate segmented animal called Urbilateria. We are testing this controversial theory of the complex Urbilateria and trying to understand how this animal lived and developed. Our team works on the development of the marine annelid Platynereis dumerilii, a representative of the trochozoan branch of the bilaterian tree, easily cultured in the lab and amenable to molecular biology techniques. We propose to investigate further two questions in the animal body plan evolution debate : 1°) How did Urbilateria make a nervous system ' What does it imply for the origin of the condensed vertebrate nervous system ' We are investigating the morphogenesis of the nervous system of Platynereis, both at the level of the fundamental process of neurogenesis and at the level of the patterning of neural architecture. Our preliminary results show that the trunk ectoderm of Platynereis is subdivided in longitudinal neurogenic columns similar to those existing in vertebrates. Expression patterns suggest that these columns are patterned by essentially the same set of homeobox genes as is known in vertebrates and these columns, according to their position are at the origin of both the central (ventral nerve cord) and peripheral (sensory cells and neurons) nervous system. We propose to explore the molecular architecture of the peripheral nervous system (PNS) of Platynereis. In vertebrates, the PNS is produced by two embryonic sources: the neural crest and the ectodermal placodes. We are going to investigate in a systematic way the expression patterns in Platynereis of the orthologues of vertebrate genes involved in the ontogenesis of the PNS, in order to determine whether embryonic territories homologous with placodes and neural crests do exist in annelids. The functions of some key genes will be tested experimentally using SiRNA or morpholino injections in fertilized eggs. With such approaches, we think we will be able to find key evidence that the complex vertebrate nervous system indeed derived from a simple annelid-like metameric pattern and even to find arguments for a scenario of the origin of this key vertebrate character, the neural crest. 2°) Was Urbilateria segmented ' Deciphering the genetic machinery that creates segmental periodicity in Platynereis by a transcriptome analysis approach. In our previous work, we have shown that the signalling pathways responsible for the individual axial patterning of ectodermal segments in Platynereis are similar to those involved in arthropod segment formation. Spatial relationships of the expression patterns of these segmentation genes suggest strongly that the ancestor of protostomes was a fully metameric animal. However, the genetic mechanisms that are responsible for creating segmental periodicity in the first place in posteriorly growing protostomes are unknown. Recent findings have shown that the mesoderm-based segmental pattern in vertebrates is regulated by cyclical waves of gene expression in the unsegmented mesoderm. These waves are controlled by the action of three signalling pathways (Notch, Wnt and FGF). In order to objectively determine the similarities and differences in segment-making genes in vertebrates and annelids, we propose to determine cyclical gene expression in the segment addition region of growing Platynereis juveniles, by transcriptome analysis using the massively parallel sequencing technology available at the GODMAP platform. We will then test hierarchical relationships between these genes and the signalling pathways potentially involved in cyclical expression such as Notch, Wnt, FGF by using specific inhibitors of these pathways. Therefore we believe our research will provide crucial elements for understanding the origins of the vertebrates. In the complex Urbilateria view, vertebrates evolved from an active annelid-like animal. We want to provide decisive arguments for two key aspects of this view : a segmented trunk and a regularly organized complex metameric nervous system.