The aim of this project is to investigate genetic mechanisms of early development of vertebrate embryos using model fish species. Zebrafish (Danio rerio) and medaka (Orizias latipes) have been used extensively for molecular genetics and developmental biology studies because these fish produce many eggs, which can be manipulated from the 1 cell stage and are ideally suited for analysing gene expression, function, and embryonic phenotypes. These species have already been extensively used to generate many mutants which show clear phenotypes during early embryonic development. The development of other model species for mutant screening and analyses is likely to provide scope to analyse gene function from uncharacterised/under-characterised genes. Therefore we have developed and tested a small number of early developmental mutants from the mangrove killifish (Kryptolebias marmoratus). To achieve my aim, embryos from zebrafish, medaka and the mangrove killifish have been used as models to study gene function and understand the molecular mechanisms for early patterning genes. We focused in particular on development of neural ectoderm and non-neural ectoderm (epidermis) and anterior-posterior patterning (head, trunk and tail development). As different model animals have different advantages, we used these model animals for different purposes. Zebrafish and medaka were used with chemical treatment (specific inhibitors of target genes) and morpholino analyses because they give many synchronized eggs every morning allowing highly replicated analyses. On the other hand, the mangrove killifish were used for developing and testing novel mutants and associated loss (or gain) of gene function. Firstly, zebrafish was used to study maternal fibroblast growth factor (FGF) signaling at pre-maternal zygotic transition (Pre-MZT) and consequent neural induction at the gastrula stage (Chapter 3). This study found the important role of acquiring maternal FGF signaling in stem cells to achieve neural induction during the zygotic gene expression stage. An FGF signaling inhibitor SU5402 was tested using RNA–seq, ATAC-seq, in.situ hybridization and immunohistochemistry methods. Through these techniques, we found that the maternal FGF signaling provides competence to the ectodermal stem cells for neural induction possibly via epigenetic modification of histone trimethylation. To examine the role of a specific FGF molecule (FGF2), gene knockdown was conducted to study fgf2 gene function during early development in zebrafish (Chapter 4). In situ hybridization and immunostaining with tissue-specific markers at the gastrula stage were used to discover a novel role for fgf2 in development of the epidermis. The final stage of my project involved characterization of mutations underlying two mutant phenotypes (short tail/stl and ball tail/stl), that exhibit defects in tail development using the self-fertilizing mangrove killifish (Chapter 5). Using a small scale RNA-seq, the mutated genes responsible for the stl and btl mutations were instantly identified as noto and msgn1 respectively. The mutant phenotype was phenocopied by morpholino injections in medaka. This study revealed crucial roles of the two genes in tail bud development. Defects of these genes affected the motility of progenitor cells in the tail bud by suppressing cell translocation to the axial mesoderm in the noto mutation and to the paraxial mesoderm in the msgn1 mutation. The study demonstrated similarity of gene function and redundancy in the mangrove killifish and medaka that is different from the function of these genes in zebrafish, revealing the importance of research on different model animals to fully characterise the gene function. From these data, it can be considered that mangrove killifish is very powerful model for mutation screening, suggesting that this animal model can be applied in various genetic studies alongside or in addition to other vertebrate models.

Yes / The HOX genes are a group of homeodomain-containing transcription factors that play important regulatory roles in early development, including the establishment of cell and tissue identity. HOX expression is generally reduced in adult cells but is frequently re-established as an early event in tumour formation and supports an oncogenic phenotype. HOX transcription factors are also involved in cell cycle regulation and DNA repair, along with normal adult physiological process including stem cell renewal. There have been extensive studies on the mechanism by which HOX proteins regulate transcription, with particular emphasis on their interaction with cofactors such as Pre-B-cell Leukaemia Homeobox (PBX) and Myeloid Ecotropic Viral Integration Site 1 (MEIS). However, significantly less is known of how the activity of HOX proteins is regulated. There is growing evidence that phosphorylation may play an important role in this context, and in this review, we draw together a number of important studies published over the last 20 years, and discuss the relevance of phosphorylation in the regulation and function of HOX proteins in development, evolution, cell cycle regulation, and cancer.

La crête neurale (CN) est une population transitoire de cellules multipotentes qui émerge à la frontière entre l’ectoderme neural et non-neural, dans une région appelée la bordure neurale (BN). Lorsque la BN se soulève pour former le tube neural, les cellules de la CN subissent une transition épithélium-mésenchyme (TEM), et migrent de façon intensive dans l’ensemble de l’embryon pour atteindre leur destination finale et se différencier. Elles sont à l’origine de nombreux types de dérivés : neurones, cellules gliales, cartilage de la tête, os et tissus connectifs, cellules pigmentaires, cellules sympatho-adrenales. Tous ces processus sont régulés par l’action coordonnée de nombreux gènes qui forment un réseau de régulations génétiques complexe, au sein duquel de nombreuses interactions ont été décrites, même si de nombreuses relations restent à élucider à ce jour. Une mauvaise régulation de gènes normalement impliqués dans la formation de la CN provoque des malformations congénitales appelées neurocristopathies. Par ailleurs, la TEM subie par les cellules de CN avant leur migration est également observée dans les cellules cancéreuses acquérant des propriétés métastatiques. Les événements moléculaires et de nombreux gènes impliqués dans la TEM sont communs au développement de la CN et au cancer.Les liens existant entre le développement de la CN et les neurocristopathies, ainsi que les métastases, soulignent l’importance de l’étude du réseau de régulations génétiques permettant la formation de la CN et l’EMT.Au laboratoire, nous nous intéressons aux événements précoces d’induction et de spécification de la CN. Dans le but d’identifier les gènes préférentiellement impliqués dans le développement précoce de la CN et non dans la formation de l’ectoderme neural et non-neural, un crible a été effectué sur le transcriptome de différents tissus embryonnaires micro-disséqués. La validation des résultats de ce crible a permis d’identifier plusieurs gènes intéressants possédant une fonction potentielle dans la formation de la CN. Nous nous sommes particulièrement intéressés à deux d’entre eux, en raison de leur fonction originale comparée à la majorité des gènes impliqués dans le développement de la CN : serca1 et pfkfb4, un régulateur de l’homéostasie calcique et un régulateur de la glycolyse respectivement.Nous avons analysé les patrons d’expression des gènes des familles serca et pfkfb au cours du développement de Xenopus laevis. En raison de son expression spécifique dans la CN, nous avons étudié plus en détails le rôle de pfkfb4 dans la formation de la CN. Cette analyse a montré que pfkfb4 est nécessaire pour la spécification neurale et de la crête neurale.Toutefois, malgré son rôle documenté dans la glycolyse, le phénotype des morphants pfkfb4 dans l’embryon de Xenopus laevis n’est pas dû à une altération de la glycolyse.En conclusion, nos résultats démontrent l’existence d’un nouveau rôle non glycolytique pour Pfkfb4 au cours du développement embryonnaire de Xenopus Laevis. / Neural Crest (NC) is a transient population of multipotent cells that arises at the border between neural and non-neural ectoderm, in a region named the neural border (NB). As the neural border elevates to form the neural tube, NC cells undergo an Epithelial-To-Mesenchymal Transition (EMT), migrate extensively into the whole body to reach their final destinations and differentiate. They give rise to multiple derivatives: neurons and glia, head cartilage, bones and connective tissue, pigment cells, sympatho-adrenal cells. All these processes are regulated by the concerted actions of several genes that form a complex Gene Regulatory Network (GRN), in which many interactions have been elucidated, but even more relationships still need to be understood. Misregulation of genes normally involved in NC formation causes birth defects called neurocristopathies. Moreover, the EMT that NC cells undergo before migration also takes place when cancer cells become metastatic: the molecular events and many of the genes involved in EMT and migration are shared between NC development and cancer. The links with metastasis, neurocristopathies and the fact that still little is known about the earliest steps of NC formation, highlight the importance and the interest in understanding the Gene Regulatory Network (GRN) leading to NC formation and EMT.In the laboratory, we are interested in the early steps of NC induction and specification. In order to identify genes preferentially involved in early NC development compared to genes involved in neural and non-neural ectoderm formation, a transcriptome screen on different microdissected embryonic tissues has been performed. The validation of the results of the screen revealed several interesting genes with a potential function in NC formation. We focused particularly on two of them, due to their original function compared to the majority of the genes involved in NC development: serca1 and pfkfb4, a calcium homeostasis regulator and a glycolysis regulator respectively. We analysed the expression patterns of serca and pfkfb family genes during Xenopus laevis development. Then, due to its specific expression in NC, we studied more in details the role of pfkfb4 in NC formation. This analysis revealed that pfkfb4 is necessary for neural and neural crest specification. However, despite its known role in glycolysis, pfkfb4 morphant phenotype in Xenopus laevis embryos is not due to an alteration of the glycolytic pathway.In conclusion, our results reveal a novel extra-glycolytic role for Pfkfb4 during Xenopus laevis embryonic development.

Signaling cascades, such as the extracellular signal-regulated kinase (ERK) pathway, play vital roles in early vertebrate development. Signals through these pathways are initiated by a growth factor or hormone, are transduced through a kinase cascade, and result in the expression of specific downstream genes that promote cellular proliferation, growth, or differentiation. Tight regulation of these signals is provided by positive or negative modulators at varying levels in the pathway, and is required for proper development and function. Two members of the dual-specificity phosphatase (Dusp) family, dusp6 and dusp2, are believed to be negative regulators of the ERK pathway and are expressed in both embryonic and adult zebrafish, but their specific roles in gametogenesis and embryogenesis remain to be fully understood. Using CRISPR/Cas9 genome editing technology, we generated zebrafish lines harboring germ line deletions in dusp6 and dusp2. We do not detect any overt defects in dusp2 mutants, but we find that approximately 50% of offspring from homozygous dusp6 mutants do not proceed through embryonic development. These embryos are fertilized, but are unable to proceed past the first zygotic mitosis and stall at the one-cell stage for several hours before dying by 10 hours post fertilization. We demonstrate that dusp6 is expressed in the gonads of both male and female zebrafish, suggesting that loss of dusp6 causes defects in germ cell production. Notably, the 50% of homozygous dusp6 mutants that complete the first cell division appear to progress through embryogenesis normally and give rise to fertile adults. The fact that offspring of homozygous dusp6 mutants stall at the one-cell stage, prior to activation of the zygotic genome, suggests that loss of dusp6 affects gametogenesis. Further, since only approximately 50% of homozygous dusp6 mutants are affected, we postulate that ERK signaling is tightly regulated and that dusp6 is required to keep ERK signaling within a range that is permissive for gametogenesis. Lastly, since dusp6 is expressed throughout zebrafish embryogenesis, but dusp6 mutants do not exhibit defects after the first cell division, it is possible that other feedback regulators of the ERK pathway compensate for loss of dusp6 at later stages.