Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease characterized by preferential motor neuron death in the brain and spinal cord. The rapid disease progression results in death due to respiratory failure, typically within 3-5 years after disease onset. While ~90% of cases occur sporadically, remaining 10% of ALS cases show familial inheritance, and the number of genes linked to ALS has increased dramatically over the past decade. FUS/TLS (Fused in Sarcoma/ Translocated to liposarcoma) is a nucleic acid binding protein that may regulate several cellular functions, including RNA splicing, transcription, DNA damage repair and microRNA biogenesis. More than 50 mutations in the FUS gene are linked to 4% of familial ALS, and many of these may disrupt the nuclear localization signal, leading to variable amounts of FUS accumulation in the cytoplasm. However, the mechanism by which FUS mutants cause motor neuron death is still unknown. The studies presented in this dissertation focused on investigating the properties of FUS mutants in the absence and presence of stress conditions. We first examined how ALS-linked FUS mutants behaved in response to imposed stresses in both cell culture and zebrafish models of ALS. We found that FUS mutants were prone to accumulate in stress granules in proportion to their degree of cytoplasmic mislocalization under conditions of oxidative stress, ER stress, and heat shock. However, many FUS missense mutants are retained predominantly in the nucleus, and this suggested the possibility that these mutants might also perturb one or more nuclear functions. In a human cell line expressing FUS variants and in human fibroblasts from an ALS patient, mutant FUS expression was associated with enlarged promyelocytic leukemia nuclear bodies (PML-NBs) under basal condition. Upon oxidative insult with arsenic trioxide (ATO), PML-NBs in control cells increased acutely in size and were turned over within 12-24 h, as expected. However, PML-NBs in FUS mutant cells did not progress through the expected turnover but instead continued to enlarge over 24 h. We also observed a persistent accumulation of the transcriptional repressor Daxx and the 11S proteasome regulator in association with these enlarged PML-NBs. Furthermore, the peptidase activities of the 26S proteasome were decreased in FUS mutant cells without any changes in the expression of proteasome subunits. These results demonstrate that FUS mutant expression may alter cellular stress responses as manifested by (i) accumulation of mutant FUS into stress granules and (ii) inhibition of PML-NB dynamics. These findings suggest a novel nuclear pathology specific to mutant FUS expression that may perturb nuclear homeostasis and thereby contribute to ALS pathogenesis.

The central dogma of biology asserts that DNA is transcribed into RNA and RNA is translated into protein. However, this overtly simplistic assertion fails to portray the highly orchestrated and regulated mechanisms of transcription and translation. During the process of transcription, RNA provides the template for translation and protein synthesis as well as the structural and sequence specificity of many RNA and protein-based machines. While only 1-5% of the genome will escape the nucleus to be translated as mRNAs, complex, parallel, highly-conserved mechanisms have evolved to regulate specific mRNAs. Trans-acting factors bind cis-elements in both the 5" and 3" untranslated regions of mRNA to regulate their stability, localization, and translation. While a few salient examples have been elucidated over the last few decades, mRNA translation can be reversibly regulated by the shortening and lengthening of the 3" polyadenylate tail of mRNA. CPEB, an important factor that nucleates a complex of proteins to regulate the polyadenylate tail of mRNA, exemplifies a major paradigm of translational control during oocyte maturation and early development. CPEB function is also conserved in neurons and somatic foreskin fibroblasts where it plays an important role in protein synthesis dependent synaptic plasticity and senescence respectively. Focusing on the function of CPEB and its role in mRNA polyadenylation during human cellular senescence, the following dissertation documents the important finding that CPEB is required for the normal polyadenylation of p53 mRNA necessary for its normal translation and onset of senescence. Cells that lack CPEB have abnormal levels of mitochondria and ROS production, which are demonstrated to arise from the direct result of hypomorphic p53 levels. Finally, in an attempt to recapitulate the model of CPEB complex polyadenylation in human somatic cells, I unexpectedly find that Gld-2, a poly(A) polymerase required for CPEB-mediated polyadenylation in Xenopus laevis oocytes, is not required for p53 polyadenylation, but instead regulates the stability of a microRNA that in turn regulates CPEB mRNA translation. Furthermore, I demonstrate that CPEB requires Gld-4 for the normal polyadenylation and translation of p53 mRNA.

RNA-binding proteins (RBPs) are important for a wide variety of biological processes involved in gene regulation. However, the structural and dynamic contributions to their biological activity are poorly understood. The tristetraprolin (TTP) family of RBPs, including TTP, TIS11b and TIS11d, regulate the stability of mRNA transcripts encoding for key cancer-related proteins, such as tumor necrosis factor- and vascular endothelial growth factor. Biophysical studies have shown that the RNA binding domain, consisting of two CCCH zinc fingers (ZFs), is folded in the absence of RNA in TIS11d and TIS11b. In TTP, however, only ZF1 adopts a stable fold, while RNA is required to completely fold the tandem zinc finger (TZF). The focus of this research was to understand the origin and biological significance of the structural differences observed for the TZF domains of TTP and TIS11d. Three residues were shown to control the affinity for the structural Zn2+ and determine the folding of ZF2 in the absence of RNA. The partially-folded TZF domain of TTP has greater selectivity for RNA sequences than the fully folded TZF domain of TIS11d. The mRNA destabilizing activity of TTP was increased when the partially disordered RBD of TTP was replaced with the fully structured TZF domain of TIS11d. Disruption of the structure and/or dynamics of the TZF domain observed in the disease-associated mutations of TIS11d, P190L and D219E, results in aberrant cytoplasmic localization. This work demonstrates that the extent of RBD folding in the TTP family is important for differential RNA recognition, mRNA turnover, and protein localization in vivo.

Frontotemporal dementia (FTD) is the second most common early-onset dementia. A rare mutation in CHMP2B gene was found to be associated with FTD linked to chromosome 3. Previous studies have shown that mutant CHMP2B could lead to impaired autophagy pathway and altered RNA metabolism. However, it is still unknown what genes mediate the crosstalk between different pathways affected by mutant CHMP2B. Genetic screens designed to identify genes interacting with mutant CHMP2B represents a key approach in solving the puzzle. Expression of mutant CHMP2B (CHMP2Bintron5) in Drosophila eyes leads to a neurodegenerative phenotype including melanin deposition and disrupted internal structure of ommatidia. The phenotype is easily quantified by estimating the percentage of black dots on the surface of the eyes. Using this established Drosophila model, I searched for genes encoding RNA binding proteins that genetically modify CHMP2Bintron5 toxicity. I found that partial loss of Pumilio, a translation repressor, mitigates CHMP2Bintron5 induced toxicity in the fly eyes. Western blot analysis showed that down regulation of Pumilio does not significantly decrease CHMP2Bintron5 protein level, indicating indirect regulation involved in suppression of the phenotype. The molecular targets regulated by Pumilio and the mechanism underlying CHMP2Bintron5 toxicity suppression by Pumilio down-regulation requires further investigation.

Le cycle cellulaire eucaryote se divise en une série de phases ordonnées qui ont pour finalité la division cellulaire. Ce processus est primordial dans la prolifération des cellules normales et le développement, mais il est aussi très fortement dérégulé dans les cellules cancéreuses. Les phases de cycle cellulaire sont différenciées par les tâches effectuées au cours de celles-ci et requièrent l’expression de gènes spécifiques à chacune des phases. Chez l’humain, il existe environ 1000 gènes dont l’expression est dépendante de la phase du cycle cellulaire. Les mécanismes impliqués dans le contrôle de l’expression périodique de ces gènes ont principalement été étudié aux niveaux transcriptionnels et post-traductionnels. Cependant, la régulation post-transcriptionnelle demeure encore peu étudiée dans le contexte du cycle cellulaire, malgré son importance dans le contrôle de l’expression génique. Afin d’identifier des régulateurs post-transcriptionnels du cycle cellulaire, nous avons analysé la corrélation existante entre l’expression des gènes périodiques du cycle cellulaire et celle de 687 protéines liant l’ARN (RNA-binding protein; RBP) sur plus de 1000 spécimens de cancer du sein. Cette analyse nous a permis d’identifier 39 RBP dont les protéines Nuclear Factor 45 (NF45) et Nuclear Factor 90 (NF90). NF45 et NF90 forment un hétérodimère qui lie des structures d’ARN double brin et qui contrôle l’expression génique à différents niveaux de l’épissage à la stabilisation des ARNm. La déplétion de NF45 ou NF90 inhibe la prolifération des cellules en induisant de nombreux défauts mitotiques qui résultent d’une baisse d’expression de plusieurs gènes essentiels à la mitose. D’autre part, à l’aide d’une méthode de protéomique nous avons réalisé l’interactome du complexe NF45-NF90 afin de déterminer à quels niveaux ce mécanisme de régulation prend place et avons identifié une interaction avec le complexe Staufen1/2-UPF1 responsable de la dégradation des ARNm. Ainsi, les niveaux d’expression de certains ARNm importants à la mitose sont conditionnés par une compétition entre NF45-NF90 et Staufen1/2 pour la liaison à ces ARNm. Dans une seconde étude, nous avons recherché les régulations potentielles sur NF45 et NF90 au cours du cycle cellulaire et avons i découvert des évènements de phosphorylation sur NF90 prenant place en phase G2/M. Nous avons montré que cette phosphorylation est médiée par CDK1, et l’activation de CDK1 provoque la translocation de NF90 du noyau vers le cytoplasme. Enfin, au vu de l’implication du complexe NF45-NF90 dans la prolifération des cellules cancéreuses, nous avons réalisé un essai de criblage à haut débit de 120 000 molécules sur l’interaction entre NF45 et NF90. Cet essai nous as permis d’identifier plus de plus 1000 molécules pouvant potentiellement interférer avec le complexe NF45-NF90. Parmi celles-ci, nous avons retrouvé 14 molécules de la famille des glycosides cardiaques, qui sont des composés antiarythmiques mais qui par ailleurs possèdent des effets anticancéreux décrits depuis plusieurs décennies. De façon intéressante, le traitement des cellules à ces composés mène à un phénotype mitotique très similaire à la déplétion de NF45 ou NF90, suggérant une implication du complexe NF45-NF90 dans les effets antimitotiques induits par les glycosides cardiaques. En conclusion, ces études nous ont permis d’éclairer le rôle du complexe NF45-NF90 dans la prolifération cellulaire, mais aussi d’approfondir la compréhension des différents mécanismes impliqués dans le contrôle du cycle cellulaire. / The eukaryotic cell cycle is divided into a series of ordered phases leading to cell division. This process is essential in normal cell proliferation and development, but it is also largely deregulated in cancer cells. Cell cycle phases are differentiated by the different molecular processes performed and expression of specific genes at each phase is determinant. In humans, there are approximately 1000 genes that are periodically expressed throughout the cell cycle. Control of this periodic expression has been well characterized at the transcriptional and post-translational levels. However, post-transcriptional regulation remains little studied in the context of the cell cycle, despite its importance in the control of gene expression. In order to identify post-transcriptional cell cycle regulators, we have correlated the expression of cell cycle genes with the expression of 687 RNA-binding proteins (RBP) in more than 1000 breast cancer specimens. This analysis allowed us to identify 39 RBPs, including Nuclear Factor 45 (NF45) and Nuclear Factor 90 (NF90). NF45 and NF90 form a heterodimer that binds double-stranded RNA structures and controls gene expression at various levels, from splicing to stabilization of mRNAs. Depletion of NF45 or NF90 inhibits cell proliferation by inducing several mitotic defects resulting from decreased expression of many genes essential for mitosis. In order to determine at which levels this regulatory mechanism takes place, we performed a proteomic method to identifyNF45-NF90 proximal interactors and identified an interaction with the Staufen1/2-UPF1 complex responsible for the degradation of mRNAs. Thus, it appears that the expression of some mitotic mRNAs is controlled by a competition between NF45-NF90 and Staufen1/2 for binding to these mRNAs. In a second study, we looked for potential regulations on NF45 and NF90 during the cell cycle and found phosphorylation events on NF90 taking place in the G2/M phase of the cell cycle. We have shown that this phosphorylation is CDK1-dependent, and that CDK1 activation leads to the translocation of NF90 from the nucleus to the cytoplasm. Finally, based on the involvement of the NF45-NF90 complex in cancer cell proliferation, we carried out a high-throughput screening assay of 120,000 ii. Abstract ii molecules on the interaction between NF45 and NF90. This assay allowed us to identify more than 1000 molecules that could potentially interfere with the NF45-NF90 complex. Amongst them, we found 14 molecules belonging to the cardiac glycoside family, which are antiarrhythmic drugs that also display anticancer effects. Interestingly, treatment with cardiac glycosides leads to a mitotic phenotype very similar to the depletion of NF45 or NF90, suggesting an involvement of the NF45-NF90 complex in the antimitotic effects induced by cardiac glycosides. In conclusion, these studies have shed light on the role of the NF45-NF90 complex in cell proliferation, but also deepened our understanding of the different mechanisms involved in cell cycle control.

No description available.

This work focused on the further characterisation of Idhp and of the Krebs cycle enzymes citrate synthase 1 (Cit1p) and malate dehydrogenase 1 (Mdh1p) both of which have been identified as RNA-binding proteins without known RNA recognition motifs. Besides analysing their effects on mitochondrial translation and their organisation in protein complexes the work focused on the characterisation of the RNA-binding properties of recombinant Cit1p and Mdh1p: · Cit1p and Mdh1p play no essential role in mitochondrial protein synthesis. · Idhp is in a complex of molecular weight larger than the cytochrome c oxidase (250 kDa). · Cit1p and Mdh1p are in mitochondrial complexes smaller than 250 kDa. · 1000-fold molar excess of tRNA referring to COX2 leader RNA did not inhibit the RNA-binding of Cit1p and Mdh1p. · Cit1p and Mdh1p bind mitochondrial mRNAs (sense and antisense). The influence of cofactors and substrates on RNA-binding was analysed in order to reveal a possible link between the enzymatic function and the property of RNA-binding: · Acetyl-CoA and ATP inhibited the RNA-binding of Cit1p and Mdh1p at a concentration of 5 mM.

No description available.

No description available.

Schizophrenia is a severe psychiatric disorder with unsatisfactory treatment options and poorly under- stood etiology. Genetic models are a suitable tool for studying this disorder with its high heritability. However, currently available animal models do not cover the broad range of schizophrenia symptoms and are not disorder-specific. Ribonucleic acid binding motif protein 12 gene (RBM12), a novel, high- risk gene for schizophrenia, was recently identified. This thesis aimed to assess the social behavior of schizophrenia-like phenotype in RBM12 zebrafish mutants. The social behavior of mutated adult zebrafish was assessed during free-swimming. Trajectories of each zebrafish were obtained from recordings by the usage of idtracker.ai. Parameters selected to quantify the social behavior of the zebrafish were chosen based on common symptoms of humans with schizophrenia. Inter-fish distance was examined as an indicator of preferred personal space since humans diagnosed with schizophre- nia have an increased need for a greater personal space compared to mentally healthy individuals. Wall-hugging, increased speed and bottom-dwelling were studied as indicators of anxiety, a common comorbid symptom of schizophrenia. The RBM12 mutants exhibited a greater inter-fish distance than their wild-type siblings during three-dimensional recordings. They however, did not demonstrate an increased inter-fish distance during two-dimensional recordings. The mutated zebrafish displayed a higher average speed and greater wall-hugging, indicating anxiety. It can be concluded that RBM12 mutation produces partial symptomatology consistent with humans diagnosed with schizophrenia, providing a promising animal model. The current work provided novel insight into the neural substrates of schizophrenia and for potential drug screening for this disorder. Further research is needed to fully characterise schizophrenia-like symptoms in this RBM12 animal model.