No description available.

O gene, uma seqüência de nucleotídeos necessária para a síntese de moléculas funcionais, é transcrito e regulado por um conjunto de processos e fatores extremamente complexos. Entender o momento e o tecido em que os genes são expressos, as isoformas funcionais, as regiões controladoras e os fatores envolvidos na regulação da expressão de cada gene é um dos grandes desafios da biologia molecular moderna. Hoje, com a enorme quantidade de informações de seqüências genômicas e de transcriptomas, aliado ao desenvolvimento de métodos computacionais para agrupar e analisar estes dados em larga escala, o estudo dos fenômenos relacionados à transcrição e regulação gênica está passando por uma revolução. Por exemplo, é possível medir, concomitantemente, a expressão gênica de milhares de genes em diferentes tecidos, assim como identificar diversos fenômenos que atuam nestes genes. Neste trabalho nós desenvolvemos e aplicamos métodos computacionais no estudo de quatro temas envolvendo aspectos chave da transcrição e regulação gênica. No primeiro trabalho, nós abordamos a expressão gênica tecido-específica através do estudo dos genes expressos no cérebro e em dez regiões cerebrais de camundongo. No segundo trabalho, nós identificamos seqüências potencialmente envolvidas no controle da transcrição gênica através do estudo de motivos sobre representados na região promotora dos genes de receptores olfativos. No terceiro trabalho, analisamos o transcriptoma humano quanto a presença de eventos de retenção de intron, um tipo de splicing alternativo. No quarto trabalho, nós abordamos a complexidade do transcriptoma e a regulação da expressão gênica através do estudo de pares de genes senso-antisenso em humanos e camundongos. Em todos os trabalhos, obtivemos resultados que nos permitiram tirar conclusões específicas sobre cada fenômeno estudado e nos mostraram a importância de estudá-los através de uma abordagem em larga escala. Adicionalmente, verificamos que os nossos métodos computacionais foram eficientes e adequados para o estudo da transcrição e regulação gênica em Homo sapiens e Mus musculus. / Genes, nucleotide sequences necessary for the synthesis of functional molecules, are transcribed and regulated by extremely complex cellular and molecular processes. To understand when and in which tissues the genes are expressed, their functional isoforms, control regions and the factors involved in gene regulation is one of major challenges of modern molecular biology. Today, the availability of complete genome sequences and transcriptomes, together with the development of new computational methods allows the study of phenomena related to the transcription and gene regulation in a large scale. For example, it is possible to quantify, concomitantly, gene expression of thousands of genes in different tissues and analyze different aspects of their regulation. In this work we developed and applied computational methods to the study of four key aspects of gene transcription and regulation. In the first study, we addressed tissue specific gene expression through the study of genes that are preferentially expressed in the brain and ten different mouse brain regions. In the second study, we identified sequences that are potentially involved in the control of gene transcription through the study of motifs that are over represented in the promoter region of olfactory receptor genes. In the third study, we browsed the human for the presence of intron retention, a type of alternative splicing. In the fourth study, we addressed the transcriptoma complexity and gene expression regulation through the study of pair of sense-antisense genes in human and mouse. In all studies, our results allowed us to make specific conclusions about each phenomenon analyzed which showed us the importance of a large scale approach. In addition, we verified that our computational methods can be efficiently applied to the study of transcription and gene regulation in Homo sapiens and Mus musculus.

O gene, uma seqüência de nucleotídeos necessária para a síntese de moléculas funcionais, é transcrito e regulado por um conjunto de processos e fatores extremamente complexos. Entender o momento e o tecido em que os genes são expressos, as isoformas funcionais, as regiões controladoras e os fatores envolvidos na regulação da expressão de cada gene é um dos grandes desafios da biologia molecular moderna. Hoje, com a enorme quantidade de informações de seqüências genômicas e de transcriptomas, aliado ao desenvolvimento de métodos computacionais para agrupar e analisar estes dados em larga escala, o estudo dos fenômenos relacionados à transcrição e regulação gênica está passando por uma revolução. Por exemplo, é possível medir, concomitantemente, a expressão gênica de milhares de genes em diferentes tecidos, assim como identificar diversos fenômenos que atuam nestes genes. Neste trabalho nós desenvolvemos e aplicamos métodos computacionais no estudo de quatro temas envolvendo aspectos chave da transcrição e regulação gênica. No primeiro trabalho, nós abordamos a expressão gênica tecido-específica através do estudo dos genes expressos no cérebro e em dez regiões cerebrais de camundongo. No segundo trabalho, nós identificamos seqüências potencialmente envolvidas no controle da transcrição gênica através do estudo de motivos sobre representados na região promotora dos genes de receptores olfativos. No terceiro trabalho, analisamos o transcriptoma humano quanto a presença de eventos de retenção de intron, um tipo de splicing alternativo. No quarto trabalho, nós abordamos a complexidade do transcriptoma e a regulação da expressão gênica através do estudo de pares de genes senso-antisenso em humanos e camundongos. Em todos os trabalhos, obtivemos resultados que nos permitiram tirar conclusões específicas sobre cada fenômeno estudado e nos mostraram a importância de estudá-los através de uma abordagem em larga escala. Adicionalmente, verificamos que os nossos métodos computacionais foram eficientes e adequados para o estudo da transcrição e regulação gênica em Homo sapiens e Mus musculus. / Genes, nucleotide sequences necessary for the synthesis of functional molecules, are transcribed and regulated by extremely complex cellular and molecular processes. To understand when and in which tissues the genes are expressed, their functional isoforms, control regions and the factors involved in gene regulation is one of major challenges of modern molecular biology. Today, the availability of complete genome sequences and transcriptomes, together with the development of new computational methods allows the study of phenomena related to the transcription and gene regulation in a large scale. For example, it is possible to quantify, concomitantly, gene expression of thousands of genes in different tissues and analyze different aspects of their regulation. In this work we developed and applied computational methods to the study of four key aspects of gene transcription and regulation. In the first study, we addressed tissue specific gene expression through the study of genes that are preferentially expressed in the brain and ten different mouse brain regions. In the second study, we identified sequences that are potentially involved in the control of gene transcription through the study of motifs that are over represented in the promoter region of olfactory receptor genes. In the third study, we browsed the human for the presence of intron retention, a type of alternative splicing. In the fourth study, we addressed the transcriptoma complexity and gene expression regulation through the study of pair of sense-antisense genes in human and mouse. In all studies, our results allowed us to make specific conclusions about each phenomenon analyzed which showed us the importance of a large scale approach. In addition, we verified that our computational methods can be efficiently applied to the study of transcription and gene regulation in Homo sapiens and Mus musculus.

La localisation des ARNm dans différents compartiments subcellulaires est conservée dans un large éventail d'espèces et de divers types cellulaires. Le trafic est médié par l'interaction entre les protéines de liaison à l'ARN (RBP) et l'ARNm. Les RBP reconnaissent les éléments cis-régulateurs de l'ARNm, également appelés éléments de localisation. Ceux-ci sont définis par leur séquence et/ou leurs caractéristiques structurelles résidant dans la molécule d'ARNm. La localisation des ARNm est essentielle pour la résolution subcellulaire et temporelle. De plus, les ARNm se sont avérés enrichis dans de nombreux compartiments cellulaires, notamment les mitochondries, l'appareil mitotique, et le réticulum endoplasmique. En outre, des études ont démontré que les RBP et les ARNm sont associés aux structures de l'appareil mitotique. Cependant, le rôle que joue la localisation de l'ARNm au cours de la mitose reste largement inexploré. Ma thèse de doctorat vise à comprendre comment le trafic d'ARNm est impliqué lors de la mitose. La première partie de cette thèse porte sur l'interaction post-transcriptionnelle qui se produit entre les deux ARNm, cen et ik2. Les gènes qui se chevauchent sont une caractéristique frappante de la plupart des génomes. En fait, il a été constaté que le chevauchement des séquences génomiques module différents aspects de la régulation des gènes tels que l'empreinte génomique, la transcription, l'édition et la traduction de l'ARN. Cependant, la mesure dans laquelle cette organisation influence les événements réglementaires opérant au niveau post-transcriptionnel reste incertaine. En étudiant les gènes cen et ik2 de Drosophila melanogaster, qui sont transcrits de manière convergente avec des régions 3' non traduites qui se chevauchent, nous avons constaté que la liaison physique de ces gènes est un déterminant clé dans la co-localisation de leurs ARNm aux centrosomes cytoplasmiques. Le ciblage du transcrit ik2 dépend de la présence et de l'association physique avec l'ARNm de cen, qui est le principal moteur de la co-localisation centrosomale. En interrogeant les ensembles de données de séquençage de fractionnement, nous constatons que les ARNm codés par des gènes qui se chevauchent en 3' sont plus souvent co-localisés par rapport aux paires de transcrits aléatoires. Ce travail suggère que les interactions post-transcriptionnelles des ARNm avec des séquences complémentaires peuvent dicter leur destin de localisation dans le cytoplasme. La deuxième partie de cette thèse consiste à étudier le rôle que jouent les RBP au cours de la mitose. Auparavant, les RBP se sont avérés être associés au fuseau et aux centrosomes. Cependant, leur rôle fonctionnel au niveau de ces structures reste à étudier. Grâce à un criblage par imagerie avec plus de 300 anticorps, nous avons identifié 30 RBP localisés dans les structures mitotiques des cellules HeLa. Ensuite, pour évaluer les rôles fonctionnels de ces RBP, nous avons utilisé l'interférence ARN (ARNi) pour évaluer si la fidélité du cycle cellulaire était compromise dans les cellules HeLa et les embryons de Drosophila melanogaster. Fait intéressant, nous avons identifié plusieurs candidats RBP pour lesquels le knockdown perturbe la mitose et la localisation de l'ARNm dans les cellules HeLa. De plus, la perte des orthologues a entraîné des défauts de développement chez l'embryon de mouche. Grâce à ce travail, nous avons démontré que les RBP sont impliquées pour assurer une mitose sans erreur. En résumé, les travaux que j'ai menés mettent en lumière l'implication de la régulation post-transcriptionnelle au cours de la mitose. En définissant les fonctions et le mécanisme de localisation des ARNm en mitose, ce travail permettra de définir de nouvelles voies moléculaires impliquées dans la régulation de la mitose. Puisque la division cellulaire non contrôlée peut mener à des maladies tel le cancer, étudier le contrôle du cycle cellulaire sous cet angle « centré sur l'ARN » peut aider à développer de nouvelles approches thérapeutiques pour trouver des solutions aux problèmes de santé. / The localization of mRNAs to different subcellular compartments is conserved in a wide range of species and diverse cell types. Trafficking is mediated by the interaction between RNA binding proteins (RBPs) and mRNA. RBPs recognize mRNA cis regulatory motifs, otherwise known as localization elements. These are defined by their sequence and/or structural features residing within the mRNA molecule. Localization of mRNAs is essential for subcellular and temporal resolution. Furthermore, mRNAs have been found to be enriched in many cellular compartments including the mitochondria, mitotic apparatus, and endoplasmic reticulum. Moreover, studies have demonstrated that RBPs and mRNAs are associated with mitotic apparatus structures. However, the role that mRNA localization plays during mitosis remains largely unexplored. My PhD thesis aims to understand how the trafficking of mRNAs is implicated during mitosis. The first part of this thesis encompasses the post-transcriptional interaction that occurs between the two mRNAs, cen and ik2. Overlapping genes are a striking feature of most genomes. In fact, genomic sequence overlap has been found to modulate different aspects of gene regulation such as genomic imprinting, transcription, RNA editing and translation. However, the extent to which this organization influences regulatory events operating at the post-transcriptional level remains unclear. By studying the cen and ik2 genes of Drosophila melanogaster, which are convergently transcribed with overlapping 3’untranslated regions, we found that the physical linkage of these genes is a key determinant in co-localizing their mRNAs to cytoplasmic centrosomes. Targeting of the ik2 transcript is dependent on the presence and physical association with cen mRNA, which serves as the main driver of centrosomal colocalization. By interrogating global fractionation-sequencing datasets, we find that mRNAs encoded by 3’overlapping genes are more often co-localized as compared to random transcript pairs. This work suggests that post-transcriptional interactions of mRNAs with complementary sequences can dictate their localization fate in the cytoplasm. The second part of this thesis involves investigating the role that RBPs play during mitosis. Previously, RBPs have been found to be associated with the spindle and centrosomes. However, their functional role at these structures was yet to be investigated. Through an imaging screen with >300 antibodies, we identified 30 RBPs localized to mitotic structures in HeLa cells. Then, to assess the functional roles of these RBPs, we used RNA interference (RNAi) to assess whether cell cycle fidelity was compromised in HeLa cells and Drosophila melanogaster embryos. Interestingly, we identified several RBP candidates for which the knockdown disrupted mitosis and mRNA localization in HeLa cells. Furthermore, loss of the orthologs led to developmental defects in the fly embryo. Through this work, we demonstrated that RBPs are involved in ensuring an error-free mitosis. In summary, the work that I have conducted sheds light on the involvement of post-transcriptional regulation during mitosis. By defining the functions and mechanism of mRNA localization in mitosis, this work will help define new molecular pathways involved in mitosis regulation. As uncontrolled cell division can lead to diseases such as cancer, studying cell cycle control from this ‘RNA-centric’ angle may help to develop new therapeutic approaches to find solutions to health problems.