Evolution of animal body plans/ Structures et décisions collectives dans les sociétés animales

This colloquium examined the molecular processes by which the diverse animal body forms arise during evolution. After eukaryotiс cells were formed, it was the different arrangements of cells in body plans that accounted for the innovation that allowed metazoans to colonize new environmental niches. Cells know their position within the body, .and whether they should proliferate or adopt special types of cell differentia­tion programs, through conserved systems of developmental genes. This presents an important question in evolutionary biology : How can con­served signalling mechanisms result in diversity of body shapes ? These are old questions that bring us to the beginnings of Natural History.

It has now become clear that all bilateral animals derived from a cor­plex common ancestor. This ancestor is called Urbilateria (Ur = primeval; bilateria = bilateral animals), which must have existed long before the Cambrian explosion. Cambrian fossils such as those of the Burgess shale or those at the edges of the Yangtze river (570 to 610 mil­lion years ago,Муa) already contained the main lineages of present dаy animals: the deuterostomes (mouth-second, with the anus forming close to the blastopore) and the protostomes (mouth-first, mouth formed in the vicinity of the blastopore). The new molecular phylogenes have fur­ther subdivided the protostomes into lophothrochozoans (LTZ, which include lophophorates, annelids and mollusks) and ecdysozoans (EDZ), including arthropods and nematods. Representatives of the three main branches — LTZ, EDZ and deuterostomes — can be found preserved in the Burgess shale. Thus, by considering the patterning mechanisms already in place in the Urbilateria, one can gain insights into early animal evolu­tion. These conserved mechanisms mаy have in turn placed important developmental constraints that may have influenced possible outcomes of Evolution.

Reconstructing Urbilateria was the main emphasis of the colloquium, and this was approached in a multidisciplinary way. Paleontologists, molecular biologists, zoologists and еmbryologists contributed to this exploration of evolutionary history. Participants originated from diverse geographical sources : Michael E. Akam, Wellcome/CRC Institute ; Detlef Arendt, EMBL; Guillaume Balavoine, Centre de Génétique Moléculaire du CNRS ; Michael Coates, University of Chicago; Stephen Cohen, EMBL ; Martin J. Cohn, Univ. of Reading ; Simon Conway Morris, Univ. of Cambridge, UK ; Eddy M. De Robertis, HHMI/UCLA; Claude Desplan, HHMI/Rockefeller Univ ; Denis Duboule, Université de Genève ; Brigitte Galliot, Université de Genève ; Nicholas Holland, Sсripps Institution of Oceanography; Peter W.H. Holland, University of Reading ; Phillipe Janvier, Muséum National d’Histoire Naturelle; Fotis C. Kafatos, EMBL; James Lake, UCLA; Nicole Le Douarin, CNRS – Institut d’Embryologie Cellulaire et Moleculaire; Patrick Lemaire, Institut de Biologie du Developpement; Nipam Patel, HHMI/Univ. of Chicago ; Fiuppo M. Rijli, CNRS ­INSERM and R. J. Sommer, Max-Planck Inst. fur Entwicklungsbiologie. The unique atmosphere provided by sparkling skies over wonderful Les Treilles ensured the discussions were lively, providing a delightful experi­ence for the fortunate individual participants. Summaries of the topics examined are listed below :

Historical Background

It is appropriate that this meeting was held in France, because the intellectual underpinnings of it can be traced back to a debate held in the French Aсadеmу in 1830 between Georges Cuvier and Étienne Geoffroy St-Hilaire. The effects of that debate have affected evolutionary thinking to this day, even though this dispute predated Darwin by several decades. Cuvier was the person most responsible for bringing together compara­tive anatomy and zoology, a combination that has proven vert’ effective up to the present day. Не is perhaps best known for his discovery of the existence of extinctions using stratigraphic analysis of the fossils of the Montmartre quarries. For Cuvier the functional purpose of animal struc­ture was primary. His was a teleological approach to nature, in which the “conditions of existence” or requirements of the environment determined animal structure. For example, an elephant had a trunk for the purpose of feeding. Cuvier had grouped animals in four “enbranchements”, the vertebrates, molluscs, articulates and radiates (including starfish, jеlly­fish). Within each embranchment there would be many different types of animals, but there could be no similarities in structure between animals of different embranchments.

Étienne Geoffroy Saint-Hilaire, like Cuvier, worked at the Museum of Natural History at the Jardin des Plantes in Paris. Hеintroduced the con­cept of homology between animal structures, although he used the term “analogues” rather than “homologues”. Geoffroy was thinking of a blue­print used by the Creator. The parts were vert’ similar in all animals but thеy could be used for different purposes. This turned the functional argument on its head : rather than structures developing for new func­tions, new functions could utilize old structures. Geoffroy proposed that there was a unity of plan among most animals. He traced homologous structures through the principle of connections, and was the first to study ossification centers in the fetus, which provided a better indicator of the relationships between neighboring bone structures.

The 1830 debate itself was concerned with problems such as the homologues of the sternum in fishes, or the origin of the opercular bone in fishes, which Geoffroy believed, incorreсtly, to derive from the hyoid bone. It was widely believed at the time that Cuvier was the clear winner of the debate on scientific terms. However, Geoffroy’s ideas resonated through Europe, as revealed by a statement several years later by Honoré de Balzac in the introduction to his La Comédie Humaine: “Il n’y a qu’un animal. Le Créateur ne s’est servi que d’un seul et même patron pour tous les êtres organisés” (“There is but one animal. The Creator used onlу a single pattern for ail organisms”). This debate of ideas contrasting the unity of plan to the primacy of function has continued to this day. Important questions of fundamental importance in evolutionary biology are, for example, how much of animal morphology is due to homology from structures derived from a common ancestor, or to convergent evolution. Is there a unity of plan among all bilateran animals, or perhaps a unity of patterning genes ? If this is so, what constraints if any, might this pose on the outcome of evolution ? Should evolutionary biologists, zoologiste and molecular biologists pursue studies of similarities among different animal phyla, or should we stress the importance of the molecular changes that have led to the enormous variety of life on earth ? These were the same questions analyzed at the colloquium, but in modem terms.

Simon Conway Morris

The early evolution of metazoans, or, the fossils aren’t silent

Life is old, but animals apparеntly are relatively young. The former had appeared probably by 3.8 billion years ago, the latter perhaps no earlier than 0.6 Byr. Such a view, at least so far as the metazoans are con-cerned is, however, decidedly controversial. The crucial question linges on whether the Cambrian “explosion” is to be interpreted as a genuine evolutionary event, as against an artifact dependent on the breaching of thresholds that allow the ready fossilization of larger animals, the rive-ment of which imprinted traces in or on the seafloor, and more obviously skeletons robust enough to resist deсay. In this view there is no obstacle to metazoans having a prior, and сryptic, history that could extend back­wards by hundreds of millions of years. This polarization of views has generated a healthy debate, but Conway Morris argued that the evidence is swinging decisively in favour of the evolutionary reality of the Cambrian “explosion”.

In brief, there is now doubt concerning a number of key arguments used in support of the supposed deep origination of the metazoans. These arguments turn on and include : the validity of molecular clicks, the notion that pre-Ediacaran metazoans were comparable to plank­totrophic larvae, or that one or more “Snowball Earths” had a material influence via bottleneck effects on eukaryotic evolution. In contrast, one could argue that to the first approximation, the fossil record offers a valid guide to early metazoan evolution. This is neither to deny gaps not the reality of “evolutionavy museums”, let alone the likelihood that a sub­stantial number of important groups were of millimetric size and apart from equivalent-sized traces (whose precise attribution is likely to remain unproven) maу lever be identified as fossils.

James Lake

The new phylogeny

Nearly one hundred and fifty years ago, Charles Darwin predicted that some dаy “… we will have fairly true trees of each great kingdorn of nature.” Today that dream is starting to come true. James Lake reviewed the progress around the world toward reaching Darwin’s goal and described the exciting prospects for the future that are being made possible by whole genome analyses. In this new view of the world, the bilateral ani­mals are divided into two groups, the Deuterostomia, and the Protostomia. The studies carried out by Lake at UCLA showed that Protostomia can be further subdivided into the Ecdysozoa, and the Lophotrochozoa. Representatives of these groups include humans (deuterostome), the fruit fly Drosophila (EDZ), and clams and worms (LTZ). At present, the positions of some phyla within these three pri­mary lineages are incompletely resolved. But complete genomes of many representative phyla are expected to be available in the near future, and these will help further to clarify the bilateral tree of life. White whole genome analyses hold much promise for the future, thеуalso present new challenges, such as correcting phylogenetic trees for artifacts caused by differing genome sizes (large genome attraction). The careful choice of genomes is important, both to surmount this artifact, and to help relate developmentally important genes found in model bilateral animals to their counterparts in other bilateral animals and in humans.

Edward De Robertis

Conservation of Urbilatevian dorso-ventral patterning mechanisms

In vertebrates, dorso-ventral patterning is mediated in part by a secreted protein called Chordin. Chordin binds to Bine Morphogenetic Protein (BMP) growth factors and inhibits their signalling. Chordin also uses two co-factors Twisted-gastrulation (Tsg) and the metalloprotease Tolloid that cleaves and inactivates Chordin. Whereas Chordin is expressed dorsally, BMP-4, Tsg and Tolloid are expressed ventrally in the Xenopus embryo.

In Drosophila, a conserved gene network also controls dorsal-ventral patterning. However, the Chordin homologue, called Short gastrulation (Sog) is expressed ventrally in the fruit fly еmbryo, rather than dorsallу. Homologues of the genes that are expressed ventrally in Xenopus, BMP, Tsg and Tolloid, are expressed dorsally in Drosophila. Thus, the entire dor­sal-ventral patterning mechanism has been conserved in bilaterian evolu­tion, but with inverted orientations. The central nervous system (CNS) is always formed on the side that expresses Chordin/Sog. This conserva­tion of developmental mechanisms provides strong support for the ideas of Geoffroy St -Hilaire, who proposed that the dorsal-ventral axis had been inverted between arthropods and vertebrates. The Urbilateria there­fore had not only a conserved system of genes providing antero-posterior positional information (provided by the Hox genes), but also a conserved dorso-ventral patterning system. The way in which these conserved sys­tems may have constrained the outcome of evolution was one of the cen­tral themes of the colloquium.

Detlev Arendt

Reconstucting Urbilateria

Current molecular phylogenies subdivide the Bilateria into three main branches, EDZ (e.g., insects, nematodes), Deuterostomia (e.g., verte­brates, sea urchin) and LTZ (polychaetеs, molluscs, flatworms, and vari­ous marine invertebrates). For the molecular study of development, few data are yet available for Lophotrochozoa as compared to the other two branches. Therefore Dr. Arendt established Platymereis dumerilii (Nereidae, Polychaeta) as a new LTZ molecular model system. Polychaeta worms are considered ancestral for annelids, and for LTZ as a whole. Nereidae show the prototype Lophotrochozoan life-cycle with trochophora-type larva, and prototype polychaete morphology. For example, thеу show prototype spiralian cleavage and protostome gastru­lation, and form a rope-ladder-like CNS characteristic for polychaetes and many other invertebrate groups.

Arendt presented current work on polychaete gastrulation, and out­lined some striking resemblances to vertebrate gastrulation. polychaete and vertebrate gastrulation are generally similar in that they involve a transition from an approximately radial symmetrical blastula towards an embryo with bilateral symmetry. In the vertebrates, the bilateral neurula forms by convergence and extension of tissues on one side of the radial gastrula. These movements are restricted to the more posterior regions of the forming body, as defined by Hox gene expression, and do not occur in the otx-positive, anterior tissue. A stripe of expression of the home­obox gene gbx defines this boundary by counteracting otx. Notably, a similar, gbx-positive morphogenetic boundary also exists in Platynereis, as a stripe of cells in between otx and hox expression territories. Strikingly, the thus defined polychaete embryonic regions show morphogenetic movements vеry similar to the corresponding vertebrate regions. This allows reconstructing the transition from radial to bilateral symmetry in Urbilateria, the last common ancestors of Proto- and Deuterostomia, and elucidates the evolution of the bilaterian body form.

Ralf J. Sommer

Evolution of Nematods

The evolution of developmental processes and developmental mecha­nisms in nematodes was analyzed. Ву comparing the development of the vulva between various free-living nematode species and the model system Caenorhabditis elegans, multiple changes during cell fate specification were observed. For instance, evolutionary changes involve the numbers of cells participating in vulval formation, the mode of vulva induction and the number of signaling centers involved. To understand the underlying genetic and molecular changes, Pyistionhus pacificas has been estab­lished as a new satellite organism by carrying out large-scale mutagenesis screens for vulval patterning mutants. In addition, genetic and molecular studies have also been initiated for sex determination, allowing a detailed analysis of how developmental pathways evolve.

Filippo Rijli

Evolution and development of the vertebrate jaw

Hохa2 is a homeotic selector of the segmental identity of the neural crest (NC) cells contributing to the 2nd (hyoid) branchial arch. A combi­natorial code of Нох expression patterns hindbrain segments, the rhom­bomeres (r), and rhombomere — derived neural crest (NC) cells coloniz­ing the branchial arches, and gives rise to most of the skeleton of the face and neck. The expression pattern of Нохa2, related to the proboscipedia gene of Drosophile, is conserved among vertebrates, and it is differentially regulated in the hindbrain and NC cells. In fact, Нoхa2 is expressed up to r2, though not in the r2 — derived NC cells of the 1st  (mandibular) branchial arch, devoid of Нох expression. Нохa2 is instead expressed in the NC cells of the 2nd and more posterior arches. The knockout of mouse Нохa2 suggested a tight causal link between Нох code and morphology of the branchial arches. In Нохa2 mutant mice, ²¹ arch NC derivatives were replaced by a mirror — image duplication of a subset of ^Pt arch-like structures.

To provide evidence of Нохa2 functional conservation in vertebrates, a knock — down approach was undertaken in Xenopus. Morpholino oligos were injected into frog embryos (that were shown to specifically interfere with the in vitro translation of the ХНоха2 protein). In morpholino — injected embryos, hyoid — to — mandibular arch mirror — image transfor­mations were observed that phenocopied the effects of the Нохa2 tar­geted inactivation in mice. These findings support predicted homologies between jaw and middle ear elements in amphibians and mammals. They

Evolution of animal body plans

further suggested that gnathostome mandibular and hyoid arches share a default developmental `ground pattern’, executed in the absence of lix expression and corresponding to the rl + r2 NC cell programme.

То investigate Ноха2 regulation in jawless vertebrates, an analysis of Ноха2 expression patterns was presented in the lamprey Lampetra japon-ka. The data show a conservation of LjHoxa2 expression domains along the anteroposterior axis, both in the hindbrain and NC cells. In particu­lar, the conserved downregulation of LjHoxa2 expression in 1st arch NC cells giving rise to the velum indirectly supports the proposal that lampreys might be evolutionary more closely related to gnathostomes than to hagfishes.

Philippe Janvier

Evolution of jawless vertebrates

Philippe Janvier first presented a brief account of the origin of the body plan concept, and the way it has been later recruited by the Darwinian evolutionists in the form of “common hypothetical ancestor”. Не highlighted the fact that “body plans”, “morphotypes”, or “common hypothetical ancestors” are essentially reconstructed on the basis of para- phyly and that, through the discovery of new, basal, fossil taxa, “body plans” are doomed to break down. Then, Philippe Janvier explained why the question of the monophyly vs. paraphyly of the living jawless verte­brates (cyclostomes) is crucial to the reconstruction of the vertebrate morphotype. For lack of a clear answer to this question, palaeontologists can but tutu to fossils. Palaeozoic jawless vertebrates, which unfortu- nately appear to be more informative as to the morphotype of the gnathostomes (jawed vertebrates) than as to that of the vertebrates in general. On1y the recently discovered Lower Cambrian (530 Муа) verte­brate from Chengjiang (China) may possiЫy be more basal than al other known vertebrates, yet it is surprisingly similar to a larval lamprey.

Finally Philippe Janvier spoke about his current research on the num­ber of gills in living and fossil vertebrata, showing some puzzling cases of “polybranchy” in fossil agnathans, in which the number of gills can reach to about 50. He addressed the question of the genetic control of the gill number and the possible role of the endoderm in regulating the number of gill pouches.

Brigitte Galliot

Hydra as a model system for Évolution and Development

The diblastic Cnidaria stand as a sister group of triploblastic animals and, as such, might illustrate a vеry primitive form of metazoan organi­zation and ontogeny. For example, cnidarian polyps are well known for both their common asexual way of reproduction, i.e. budding, and their astonishing capacity to regenerate. The analysis of expression patterns of several Paired-class and Antp-class homeobox genes suggest that apical patterning of cnidarians represents an ancestral process that led to ante­rior patterning in bilaterians. In contrast, expression of related molecular markers during development of the hydra apico-basal axis and the bila­terian antero-posterior axis do not appear to be similar. These results are in agreement with the radial head model, which proposes that the antero­posterior axis is a recent innovation, specific to bilaterians. In order to highlight the molecular mechanisms that underlie the reactivation of developmental programs in hydra, Gaillot compared the expression pat­terns of homeobox-containing genes and genes related to Hedgehog, Gli/ci, CREB, Thrombospondin, chordin-like and Follistatin-like at early stages of regeneration and budding. Interestingly, several of these genes show distinct regulations, either at the temporal level, like Thrombospondin, chordin-like, or at the spatial level, like Hh, Gli/ci. These data suggest that the initial processes that reactivate the developmental programs leading to head regeneration and budding rеly on distinct mechanisms.

Guillaume Balavoine

The segmentation of Urbilateria

The homologues of the engrailed and hairy genes of the fly are involved in the network regulating the segmentation of the mesoderm in some chordates. Based on these facts, several authors suggested that the ancestor of bilaterians was a segmented animal, while opponents of this theory see these functional similarities as mere convergence potentially favored by the regulative properties of these genes. Balavoine explained that different kinds of arguments coming from anatomy, embryology and paleontology make the idea of a metameric Urbilateria a very serious hypothesis. Balavoine extended the study of the functional conservation of segmentation to the third great branch of the bilaterians, the Lophotrochozoans. The marine annelid Platynereis is an obvious choice for two main reasons : first, as an annelid, it displays a characteristic metameric segmentation of the body that most zoologists have recog­nized as homologous to arthropod metamerism and second, it develops in an indirect way through a microscopic larva in a way that is also ances­tral to the bilaterians. Preliminary results suggest that the homologues of engrailed and wingless may also be involved in segmentation processes in Platynereis. The fact that the segment boundaries form between the stripes of wingless and engrailed in Platynereis further suggests that the segments of annelids may be homologous to the parasegments of arthro­pods.

Stephen Cohen

Cell growth and proliferation in morphological evolution

Stephen Cohen discussed the genetic control of growth, using the wing of the fruit-fly as an experimental mode! system. Evidence was pre­sented that a long-range gradient of a secreted signa]ling protein called Wingless (Wnt) controls the growth of the tissue as well as controlling spatial pattern. A genetic screen based on gene overexpression was used to identify genes that affect body size, but not body pattern. Tissue growth reflects control of the rates of cell growth and cell division to regulate cell proliferation and also involves an important balance between cell prolif­eration and cell survival to control cell numbers. A gene called bantam controls tissue growth rates. Evidence was presented that bantam coor­dinately controls the rate of cellular growth and the rate of cell cycle pro­gression, so that the result is coordinated cell proliferation as leen in nor­mal development. The molecular mechanism is not yet known. A second gene, slik, with a role in cell proliferation and cell death control was dis­cussed. slik encodes a protein kinase that causes abnormal tissue over­growth when cell death is prevented. The regulatory networks control­ling cell growth and survival are undoubtedly complex. The two genes discussed were presented as entrypoints that will allow researchers to learn more about these fundamental processes.

Michael I. Coates

The origin of Tetrapods

Body-plans or bauplans, unlike archetypes and ancestors, consist of conjectures about taxis homology, i.e. shared character states (advanced/primitive) at nodes within phylogenetic trees. Bauplans are hierarchical, and usually summarize conditions at the base of well-defined taxa, although phylogenies can be divided arbitrarily into any number of nested, bauplan-defined, subgroups. In contrast, archetypes are abstracts of generalised patterns, and, as such, represent idealised forms, likely to be incompatible with hypothesеs of anсestry and primitiveness. Significantly, ancestors and ancestral character states, whether morpho­logical, molecular or developmental, cannot be recovered using cladistics or other numerical methods of phylogenetic analyses. These methods can only recover hierarchies of relative relationships, and cladograms repre­sent no more than hypotheses of character distribution.

Tetrapods diverged from other sarcopterygian fishes at least 385 Муа. The tetrapod stem-group (all limbed/finned tetrapods externa! to the crown-group (amphibians and amniotes) includes a range of taxa encompassing the acquisition of all features associated with the tetrapod body-plan. Like all stem-groups, early members are barely distinguish­able as such — in this case resembling other early bony fishes. Later mem­bers are recognized as essentially complete. Coates’ results separate con­ventional adaptive scenarios of tetrapod body-plan evolution from terrestrialisation, and recent, large-scale analyses of fossil data deliver an unusually robust minimum date for the lissamphibian-amniote split, at around —340 Муаthat is potentially of value for calibration of molecu­lar phylogenies. At a more detailed level of anal)тsis, fin-limb evolution is similarly gradual. The implied transformation sequence is marked by repatterning into fewer, more individually distinct segments, and the on-gin of this arrangement precedes fin ray loss, digit acquisition, and wrist/ankle evolution. An alternative treatment, re-analysing limbs as dis­crete taxa, provides a means of investigating correlated variation among such serial homologues. Significantly, this demonstrates phylogenetically deep differences between fore- and hindlimb patterns, and that no simple bauplan for `the tetrapod limb’can summarise adequately conditions at both levels. Primitively, paired appendages differ, and morphological sim­ilarity is the specialised character state. Thus, anу explanatory develop­mental evolutionary scenario runs the risk of reducing a series of termi­nal, derived, character states into an over-simplified scala naturae.

Denis Duboule

Hox gene complexes, enhancers and the evolution of tetrapod limbs

The question of the evolution of regulatory mechanism was addressed by Denis Duboule, who discussed the emergence of an enhancer sequence in relation to the appearance of digits in the course of evolution.

The expression of several genes of the HoxD family is controlled by one and the same regulatory sequence located 200 Kb upstream the gene cluster. Ву using a method that generates targeted cross-over in mice, Duboule’s laboratory engineered a collection of mice carrying re-organ­ized Hoxd genes, in order to evaluate the enhancer/promoter relationship at work in this important mechanism. Duboule reported that Hox gene activation by this enhancer is poorly promoter-specific, hence it is sus­ceptible to great variations, following perturbations occurring within a genomic landscape of 300 to 500 Kilobases. In this way, the insertion of a transposon, for instance, at a large distance from the Hoxd cluster could lead to a regulatory mutation of this latter group of genes, simply due to enhancer titration. As a consequence, this would modify the morphology and, possible, the number of digits. Hе discussed the evolutionary impli­cation of this observation, raising once again the importance of Нох genes, both as an epistemic system and for their heuristic value. In this particular case, Нох collinear expression in digits exemplified the multiple use of gene networks, at different places and times during ontogеny, implying that such systems can not be modified without producing major effects upon the morphology of the animal. Once an enhancer is put to multiple uses in development distinct events such as digit formation and the formation of external genitalia become interlocked and cannot be changed without unexpected costs. The implication of these develop­mental constraints for our understanding of the relationships between ontogeny and phylogeny led to lively discussions on the splendid Les Treilles terrace.

Claude Desplan

The entero-posterior axis in insects

Although the Bicoid morphogen (Bcd) plays a pivotal role in pattern­ing the anterior of the Drosophile (Dm) еmbryо, no bcd homolog has been found outside Diptera. Desplan has proposed that hunchback (hb), orthodenticle (otd), and caudal (cad), which are functions that are evolutionar­ily conserved up to vertebrates, are the key components of an ancestral anterior patterning system whose function has been taken over by bcd in Drosophila. To better understand the evolution of hb, otd and cad, he asked whether thеу can act together to control axial patterning in the absence of bcd. These genes were also studied in another insect order, the wasp Nasonia vitripenis (Nv) . Sophisticated genetics has identified a mutation in Nv, hb that produces a zygotic phenotype much more severe than that of Dm hb, consistent with the hypothesis that hb was originally necessary for extensive control of axial patterning. In contrast to the beetle Tribolium (Tc) , which has a short germ band mode of development that distinguishes it from Nv and Dm, the morphological similarity of Nv and Dm embryogeasses facilitate comparative studies among related species. Since Bcd and Otd share the same DNA binding specificity, and otd has been reported to be maternal in Tc, otd might represent an ancestral ante­rior patterning gene. Desplan is studying the potential of Otd as a mor­phogen in Dm, and testing whether high levels of Otd, alone or in cor­bination with Hb, can pattern the anterior of the еmbryo in the absence of bcd.

This comparative characterization allows the understanding of the ancestral anterior patterning system of insects, using the power of Dm molecular genetics and the unique features of Nv. Using the plastiсity of developmental systems, one can reconstruct in Dm the ancestral mecha­nisms, to perform a “de-evolution” process from the derived fly embryо and stimulate the more generic type of insect where bcd might not have yet appeared to coordinate most anterior patterning functions.

Patrick Lemaire

Ascidian development

Ascidians are urochordates, which separated from the cephalochor­date/vertebrate lineages during the lower Cambrian. Patrick Lemaire presented a comparison of the strategies used by ascidians and vertebrates to form two typical chordate structures : the notochord and the dorsal cen­tral nervous system. Both ascidians and vertebrates make use of the Bmp/Chordin antagonism to form their notochord, but in different ways. Bmp signalling represses notochord formation in vertebrates, but is required for ascidian notochord specification. Lemaire also presented studies on the formation of the central nervous system in ascidians and vertebrates. In both cases, the FGF pathway is important before gastru­lation for the induction of at least some of the neural cells. In addition, signalling via ERK, a downstream component of the FGF pathway, is important for the formation of posterior neural tissue. These results argue for an ancestral role for FGF signals in the early induction and pat­terning of the chordate CNS.

Peter Holland

Choanoflagellates and the origin of metazoans

In trying to understand the genetic and developmental basic of animal evolution, it is helpful to know which genes are unique to animals and which are more widespread. Peter Нolland discussed recent molecular work on choanoflagellates — single celled eukaryotes long thought to be closely related to Metazoa. Mary housekeeping genes were highly con­served ; these proved useful links between choanoflagellates and animals. In contrast, two genes were discussed that were radically different to their metazoan homologues. In each case, it seems that the encoded protein was recruited for a novel function during the evolution of multicellular­ity : one was concerned with the basal lamina, the other with cell sig­nalling.

Clustered ANTP clans homeobox genes are another genetic character so far unique to animals. Peter Нolland contrasted the evolution of the lox and ParaHox clusters with the NK homeobox gene cluster. The first two are compact in vertebrates and amphioxus, but broken in Drosophile ; the opposite pattern is seen for the NK cluster. This emphasizes that dif­ferent gene families (and probably gene functions) can have very distinct еvolutionary histories in divergent bilaterian lineages.

Nicholas D. Holland

Amphioxus and die evolution of the neural crest

Nicholas Rolland discussed possible changes in gene regulation that mау have been involved in the evolutionary origin of vertebrate neural crest. Although only vertebrates have definitive neural crest, the inverte­brate chordate ancestor of the vertebrates is thought to have had an evo­lutionary precursor of neural crest — namely, the cells bordering the neu­ral plate. The living amphioxus is the best proxy for this ancestor. Cells at the neural plate border of amphioxus express homologues of some genes known to play roles in vertebrate neural crest development. These amphioxus homologues are not sufficient for the development of defini­tive neural crest (which requires genetic programs directing cell delami­nation, migration, and differentiation into a diversity of cell types). As a first step in finding what genetic programs might be missing from the presumed evolutionary precursor of the neural crest, Nicholas Rolland presented studies on the amphioxus AmphiFoxD gene, a homologue of vertebrate FoxD3, a gene known to activate aspects of cell migration in vertebrate neural crest. Importantly, AmphiFoxD is not transcribed at the neural plate border during amphioxus development, although it is expressed in close-by tissues (somites and nascent notochord) as thеy evaginate/delaminate from the mesendoderm. The basic assumption is that the activation of a Fo,D3 homologue in cells at the edge of the neu­ral plate was one event associated with the emergence of definitive neu­ral crest at the dawn of vertebrate evolution.

The change in tissue-specific expression of FoxD3 presumably resulted from changes in the promoter region of the gene. To study this, amphioxus eggs were injected with reporter constructs driven by the AmphiFoxD regulatory region. A 6.5 Kb stretch of the regulatory DNA linked to a lacZ reporter was expressed in patterns congruent with the expression domains seen by in situ hybridization. These successful tran­sient transgenies (the first ever obtained for developing amphioxus) open the way to characterize the regulatory DNA and to initiate reciprocal experiments between amphioxus and vertebrate genes.

Conclusion

The central theme that emerged from the colloquium is that an enor­mous number of developmental control genes were already present in the Urbilateria, the common ancestor of mort extant animals. Life on earth has an immense variety of animal forms. Considering this diversity, a main question remains unanswered : what were the contraints imposed by the conserved genetic mechanisms on the actual outcome of natural selection ? The Cuvier-Geoffroy debate continues to this day.