La protéine MLN51, surexprimée dans environ 30% des cancers du sein, est un facteur clé du métabolisme des ARNm, en tant que membre du complexe de jonction des exons (EJC). Ce graffiti moléculaire est un régulateur essentiel de l’expression génique de par son implication dans l’épissage, l’export, la traduction, la stabilité et la dégradation des ARNm. A l’échelle structurale, l’EJC est organisé autour d’un coeur protéique avec l’hélicase eIF4A3, MLN51 et l’hétérodimère Magoh/Y14. Ce tétramère sert de plateforme d’ancrage à d’autres facteurs périphériques. Les mécanismes de recrutement du coeur EJC sur l’ARNm ont été élucidés par des approches biochimiques. Dans ce contexte, nous avons initié un travail original destiné à mettre en évidence la localisation cellulaire de l’assemblage du coeur EJC in vivo. L’utilisation de techniques d’imagerie photonique et électronique a permis d’établir un lien véritable entre la localisation du coeur EJC et l’architecture nucléaire. Nous avons montré que la plupart des facteurs EJC sont localisés et interagissent à la périphérie des speckles nucléaires, lieux de stockage des facteurs d’épissage. Ces régions discrètes nucléaires ont été appelées «perispeckles» et sont des entités distinctes des speckles. De manière intéressante, la localisation des protéines coeur coïncide avec celle des ARNm dans les perispeckles et est spatialement reliée aux sites transcriptionnels. Ces données démontrent que l’assemblage du coeur EJC a lieu dans le compartiment nucléaire et définissent le perispeckle comme un territoire intermédiaire entre les speckles nucléaires et sites de transcription où s’opèrent des évènements post-transcriptionnels fondamentaux. / The MLN51 protein, overexpressed in around 30% of breast cancers, is a key factor for mRNA metabolism, as a member of the Exon Junction Complex (EJC). The EJC marks the splicing history of an mRNA and influences many stages of its subsequent metabolism: splicing, dynamic cytoplasmic export, efficient and localized translation, quality-control and stability. Structurally, the EJC is organized around a core complex that is formed by four proteins (eIF4A3, MLN51, Magoh, Y14). The core complex serves as a binding platform for more than a dozen peripheral factors. The EJC is not a pre-assembled complex; however, its assembly mode is well described in vitro using recombinant proteins and splicing extracts. Nevertheless, where this complex assembles in vivo was a matter of debate. By using light and electron microscopy approaches, we established an original link between the cellular distribution of the EJC core factors and the nuclear architecture. The core and most of the peripheral EJC factors are colocalized and interact together in discrete regions of the nucleus, located at the periphery of nuclear speckles. This doughnut-shaped region appears to be a novel nuclear territory that we termed “the perispeckle”. This territory is distinct from nuclear speckles; it contains nascent mRNAs and it is close to active transcription sites. Overall, this study supports a model in which the deposition of the EJC core takes place in the nucleus, and that assembled EJC core factors concentrate in discrete subnuclear territories termed perispeckles. These regions contribute to the compartmentalization of the nucleus as an active domain implicated in mRNP packaging.

<p>The six-layered neuron structure in the cerebral cortex is the foundation for human mental abilities. In the developing cerebral cortex, neural stem cells undergo proliferation and differentiate into intermediate progenitors and neurons, a process known as embryonic neurogenesis. Disrupted embryonic neurogenesis is the root cause of a wide range of neurodevelopmental disorders, including microcephaly and intellectual disabilities. Multiple layers of regulatory networks have been identified and extensively studied over the past decades to understand this complex but extremely crucial process of brain development. In recent years, post-transcriptional RNA regulation through RNA binding proteins has emerged as a critical regulatory nexus in embryonic neurogenesis. The exon junction complex (EJC) is a highly conserved RNA binding complex composed of four core proteins, Magoh, Rbm8a, Eif4a3, and Casc3. The EJC plays a major role in regulating RNA splicing, nuclear export, subcellular localization, translation, and nonsense mediated RNA decay. Human genetic studies have associated individual EJC components with various developmental disorders. We showed previously that haploinsufficiency of Magoh causes microcephaly and disrupted neural stem cell differentiation in mouse. However, it is unclear if other EJC core components are also required for embryonic neurogenesis. More importantly, the molecular mechanism through which the EJC regulates embryonic neurogenesis remains largely unknown. Here, we demonstrated with genetically modified mouse models that both Rbm8a and Eif4a3 are required for proper embryonic neurogenesis and the formation of a normal brain. Using transcriptome and proteomic analysis, we showed that the EJC posttranscriptionally regulates genes involved in the p53 pathway, splicing and translation regulation, as well as ribosomal biogenesis. This is the first in vivo evidence suggesting that the etiology of EJC associated neurodevelopmental diseases can be ribosomopathies. We also showed that, different from other EJC core components, depletion of Casc3 only led to mild neurogenesis defects in the mouse model. However, our data suggested that Casc3 is required for embryo viability, development progression, and is potentially a regulator of cardiac development. Together, data presented in this thesis suggests that the EJC is crucial for embryonic neurogenesis and that the EJC and its peripheral factors may regulate development in a tissue-specific manner.</p> / Dissertation

The function and evolution of introns have been topics of great interest since introns were discovered in the 1970s. Introns that interrupt protein-coding regions have the most obvious potential to affect coding sequences; therefore, their evolution have been studied most intensively. Splicing of introns within untranslated regions does not contribute directly to the diversity of proteins, yet ~35% of human transcripts contain introns within the 5' untranslated region (UTR). The evolution and possible functions of 5'UTR introns (5UIs) remain largely unexplored. Here we undertook a genome-wide functional analysis of 5UIs. Our main results are as follows: First, the distribution of these introns in the human genome is nonrandom. While genes with regulatory roles are enriched in having 5UIs, genes encoding proteins that are targeted to the endoplasmic reticulum and mitochondria are surprisingly depleted of these introns. Second, we offered and supported a model whereby gene encoding secretory and nuclear-encoded mitochondrial proteins share a common regulatory mechanism at the level of mRNA export, which is dependent on the absence of 5'UTR introns. Specifically, the upstream element in a given transcript, be it an intron or RNA elements near the 5' end of coding sequences (CDS), dictates the mRNA export pathway used. Finally, we discovered a strong correlation between existence of 5'UTR introns and sequence features near the 5' end of CDS. We developed an integrated machine-learning framework that can predict absence of 5UIs using solely the sequence near the 5' end of CDS. Our model achieved >80% accuracy when validated against nuclear-encoded mitochondrial transcripts. Specific RNA elements predictive of 5UI absence are found in ~40% of human transcripts spanning a wide spectrum of functions. By analyzing hundreds of large-scale datasets, we functionally characterized the transcripts with these RNA elements; revealing their association with translational regulation. These RNA elements were bound by proteins interacting with the Exon Junction Complex in vivo suggesting a molecular mechanism that links these elements to their downstream effects in mRNA export and translational regulation. While some 5'UTR introns might be evolving neutrally, our results, taken together, suggest that complex evolutionary forces are acting on this distinct class of introns.

Deep sequencing of mRNAs (RNA-Seq) is now the preferred method for transcriptome-wide quantification of gene expression. Yet many mRNA isoforms, such as those eliminated by nonsense-mediated decay (NMD), are inherently unstable. Thus a significant drawback of steady-state RNA-Seq is that it provides marginal information on the flux through alternative splicing pathways. Measurement of such flux necessitates capture of newly made species prior to mRNA decay. One means to capture nascent mRNAs is affinity purifying either the exon junction complex (EJC) or activated spliceosomes. Late-stage spliceosomes deposit the EJC upstream of exon-exon junctions, where it remains associated until the first round of translation. As most mRNA decay pathways are translation-dependent, these EJC- or spliceosome-associated, pre-translational mRNAs should provide an accurate record of the initial population of alternate mRNA isoforms. Previous work has analyzed the protein composition and structure of pre- translational mRNPs in detail. While in the Moore lab, my project has focused on exploring the diversity of mRNA isoforms contained within these complexes. As expected, known NMD isoforms are more highly represented in pre-translational mRNPs than in RNA-Seq libraries. To investigate whether pre-translational mRNPs contain novel mRNA isoforms, we created a bioinformatics pipeline that identified thousands of previously unannotated splicing events. Though many can be attributed to “splicing noise”, others are evolutionarily-conserved events that produce new AS-NMD isoforms likely involved in maintenance of protein homeostasis. Several of these occur in genes whose overexpression has been linked to poor cancer prognosis.

No description available.

No description available.

No description available.

Les fichiers accompagnant le document sont en format Microsoft Excel 2010. / Les modèles classiques de signalisation cellulaire eucaryotes sont généralement organisés en voies linéaires et hiérarchiques, impliquant un ensemble de facteurs restreint. Ces facteurs forment un circuit isolé qui transmet une information externe vers sa destination, d’où une réponse cellulaire sera alors engendrée. Or, ces modèles sont justement le fruit d’approches expérimentales réductionnistes qui ne permettent pas d’intégrer aisément la contribution de facteurs multiples, ni de faire une évaluation quantitative de l’apport des composantes du système. Le développement de techniques d’investigation plus holistiques, telles la génomique fonctionnelle et la protéomique, permettent d’examiner de manière systématique et quantitative l’apport d’ensembles larges de facteurs et de les mettre en relation avec d’autres systèmes cellulaires. Il y aurait donc lieu de réévaluer le modèle de voie de signalisation linéaire au profit d’un modèle de réseau de signalisation multiparamétrique, comportant plusieurs branches d’entrée et sortie de signal interagissant avec d’autres systèmes cellulaires. Cet ouvrage porte sur la voie RAS/MAPK, l’un des principaux axes de signalisation associé à la prolifération et la différenciation cellulaires. Le sujet y est d’abord abordé sous l’angle d’une perspective historique, en mettant l’emphase sur les contributions des études de génétique classique chez les organismes modèles D. melanogaster et C. elegans. Il fait ensuite état du développement du criblage par ARNi pan-génomique dans ces deux modèles en le comparant aux approches de criblage génétique classique. Le corps de l’ouvrage décrit ensuite les résultats expérimentaux d’une campagne de criblage par ARNi visant à dresser une carte globale des régulateurs de la voie chez la drosophile. Trois groupes de régulateurs identifiés dans ce crible ont été caractérisés de manière plus détaillée. Dans un premier article, nous démontrons que les composantes du complexe EJC ont un impact sur l’épissage de mapk; une découverte doublement intéressante puisque l’EJC était jusqu’alors associé qu’à la régulation post-épissage des ARNm. Une seconde publication fait état de l’ensemble des résultats du crible ARNi, mettant l’emphase sur un ensemble de facteurs d’épissage qui modulent également mapk. Nous y montrons que l’impact de ces facteurs sur l’épissage alternatif est différent de celui de l’EJC, suggérant ainsi deux modes de régulation distincts. Finalement, dans un troisième manuscrit, nous nous attardons au rôle d’Usp47, une déubiquitinase qui, contrairement aux autres facteurs identifiés dans le crible, régule l’expression de MAPK de manière post-traductionnelle. Nous y détaillons une stratégie de criblage d’interaction génétique par ARNi visant à identifier des facteurs reliés fonctionnellement à Usp47. Ce second crible a permis l’identification de trois facteurs reliés au « N-end rule », un mécanisme de dégradation des protéines caractérisé par la reconnaissance des résidus N-terminaux de protéines ou peptides. Il existait jusqu’alors très peu de données quant à la régulation de l’expression des composantes de la voie MAPK, ce qui rend la description d’un large réseau de régulateurs agissant sur l’expression de MAPK d’autant plus insoupçonnée. L’absence d’un réseau équivalent rattaché aux autres composantes de la voie laisse supposer que MAPK serait un noeud servant de point d’entrée à ce type de régulation dans le système RAS/MAPK. De plus, nos travaux témoignent de la capacité de la génomique fonctionnelle à mettre en relation différents systèmes cellulaires de manière plus globale et à quantifier les liens établis entre eux. / The classical model of eukaryotic cellular signalling generally involves hierarchically organized linear pathways involving a restricted set of elements. These generally function together as an insulated circuit, transmitting information from the outside to the intracellular compartment involved in eliciting a response. These models, often the fruit of reductionist experimental approaches, do not allow for the integration of multiple inputs nor for a gradation of responses. The recent emergence of more holistic investigation techniques has brought about the re-evaluation of these classical models in favor of multiparametric signalling networks. This thesis focuses on the RAS/MAPK pathway, one of the cell’s main proliferation and differentiation signalling conduits, beginning with a historical perspective covering the contributions of model organism genetics to the current pathway model. This provides context for the description of a whole-genome RNAi screen experiment that we carried out to obtain a global view of regulators in Drosophila. Three groups of factors emerging from this screen were then examined in more detail. A first article shows that the exon junction complex (EJC) plays a role in mapk alternative splicing, an observation that is unexpected given that this complex was not previously known to act on splicing. A second paper details the genome wide screening campaign and focuses on a large set of splicing factors that also regulate mapk, albeit in a distinct manner than the EJC’s. Finally, a manuscript in a third segment examines Usp47 function and finds it to control MAPK levels post-translationally. An RNAi-based genetic interaction screen is then used to identify factors functionally related to Usp47 capable of counteracting its impact on MAPK levels. Three such factors identified through this technique are linked to the N-end rule protein degradation pathway. Regulation of core pathway component expression is a poorly described process, which makes the identification of a large set of factors regulating MAPK expression all the more unusual. Moreover, the absence of such regulation linked to other pathway components suggests that MAPK may act as a node incorporating inputs of this type into RAS/MAPK signaling dynamics.