The subcellular localisation of mRNA is a widespread mechanism to determine the fate of mRNAs in eukaryotes. Translationally repressed mRNAs localise to P-bodies and stress granules where their decay and storage, respectively, are directed. In a study from the Ashe lab, specific mRNAs were identified to localise, in actively growing S. cerevisiae, to cytoplasmic granules that do not seem to be related to P-bodies or stress granules but appear to be associated with active translation (Lui et al., 2014).It is possible that this might represent a strategy to co-regulate the expression of proteins from the same pathway. In the work of this thesis, microscopy techniques to visualise RNAs in live cells were used to extend the localisation analysis to several mRNAs encoding translation factors. The investigated transcripts were all found to localise to mostly one or two cytoplasmic granules per cell and would sometimes overlap with other transcripts, suggesting that each granule contains a mixture of mRNAs. Granules tend to migrate to the bud tip and may provide the daughter cell with a "start-up kit" of transcripts essential for rapid growth. A similar pattern can be observed in yeast cells growing undergoing filamentous growth, with granules harbouring translation factor transcripts often found in the apical quarter of the elongated cell. Although the mechanism by which the granules form and their protein composition are not yet known, high-throughput genetic screens performed as part of this work offer some insight into factors that might be involved in granule assembly and proteins that partially overlap with the granules. We propose that granules containing translation factor mRNAs might be functioning as a specialised factory for the translation machinery and are possibly being directed to the point in the cell where the rhythm of protein production is highest.

The dysfunction of autophagy pathways has been linked to the development and progression of numerous human diseases, in particular neurological disorders and cancer. Investigating these pathological autophagy mechanisms is essential to gain insights into the underlying disease mechanisms, identify novel biomarkers, and develop targeted therapies. In this thesis, I present three manuscripts that investigate the regulatory mechanisms of autophagy machinery in human diseases. In the first manuscript (Chitiprolu et al., 2018), we investigated the mechanism of p62-mediated selective autophagic clearance of RNA stress granules implicated in Amyotrophic Lateral Sclerosis (ALS). Repeat expansions in C9ORF72, the major cause of ALS, reduce C9ORF72 levels but how this impacts stress granules is uncertain. By employing mass spectrometry, high resolution imaging and biochemical assays, we demonstrated that the autophagy receptor p62 associates with C9ORF72 to eliminate stress granules by autophagy. This requires p62 to associate with proteins that are symmetrically methylated on arginines. Patients with C9ORF72 repeat expansions accumulate symmetric arginine dimethylated proteins which co-localize with p62. This suggests that C9ORF72 initiates a cascade of ALS-linked proteins (C9ORF72, p62, SMN, FUS) to recognize stress granules for degradation by autophagy and hallmarks of a defect in this process are observable in ALS patients. The second manuscript (Guo, Chitiprolu et al., 2014) describes the mechanism by which autophagy degrades retrotransposon RNA from both long and short interspersed elements, thereby preventing new retrotransposon insertions into the genome. By employing quantitative imaging tools, we demonstrated that retrotransposon RNA localizes to RNA granules that are selectively degraded by the autophagy receptors NDP52 and p62. Mice lacking a copy of Atg6/Beclin1, a gene critical for autophagy, also accumulate both retrotransposon RNA and genomic insertions. This suggests a mechanism for the increased tumorigenesis upon autophagy inhibition and therefore a role for autophagy in tempering evolutionary change. Finally, the third manuscript (Guo, Chitiprolu et al., 2017) examines the intersection of autophagy machinery with exosome release and function in cancer metastasis. By employing dynamic light scattering, Nanosight particle tracking, electron microscopy, super-resolution imaging and Western blotting, we robustly quantified exosome identity and purity in multiple cell lines. We demonstrated that exosome production is strongly reduced in cells lacking Atg5 and Atg16L1, but this is independent of Atg7 and canonical autophagy. The effect of Atg5 on exosome production promotes the migration and in vivo metastasis of orthotopic breast cancer cells. These findings delineate autophagy-independent pathways by which autophagy-related genes can contribute to metastasis. Taken together, data presented in the three manuscripts highlight the molecular mechanisms of autophagy core machinery proteins and selective receptors such as Atg5, p62 and NDP52, in the pathogenesis of cancer and neurodegeneration. In these diseases characterized by mutations in autophagy pathways, the mechanisms we uncover provide insights into their causes and serve as potential therapeutic targets.

Eukaryotic initiation factor 4E (eIF4E) is a key part of initiation and regulation of translation in human cells. Three members of human eIF4E proteins have been characterized: eIF4E1, eIF4E2 and eIF4E3. Cellular stress causes translation initiation inhibition followed by disassembly of the polysomes, those processes are accompanied by the assembly of cytoplasmic RNA granules, called stress granules (SG). Stress granules are dynamic structures whose composition may vary depending on the cell type and the stress stimulus. In this study, human cells were subjected to the following stress conditions: high temperature (HS), sodium arsenite (AS) or hypoxia. Using fluorescence microscopy, pairs of human translational initiation factors from the 4E protein family were visualized and their localization to SG was assessed with one GFP- 4E incorporated in the stable cell line and the other one detected endogenously. Here we show eIF4E1 being a part of all the SGs, both in HS and AS conditions. Next, the eIF4E1 and eIF4E3 proteins together form more SGs than proteins eIF4E1, respectively eIF4E3, with eIF4E2. And last, that the presence of the particular 4E protein has no effect on the composition of SGs. Furthermore, selected groups of proteins were assessed for their potential to localize to the SGs under HS...

La mémoire et l’apprentissage sont des phénomènes complexes qui demeurent encore incertains quant aux origines cellulaire et moléculaire. Il est maintenant connu que des changements au niveau des synapses, comme la plasticité synaptique, pourraient déterminer la base cellulaire de la formation de la mémoire. Alors que la potentialisation à long-terme (LTP) représente un renforcement de l’efficacité de transmission synaptique, la dépression à long-terme (LTD) constitue une diminution de l’efficacité des connexions synaptiques. Des études ont mis à jour certains mécanismes qui participent à ce phénomène de plasticité synaptique, notamment, les mécanismes d’induction et d’expression, ainsi que les changements morphologiques des épines dendritiques. La grande majorité des synapses excitatrices glutamatergiques se situe au niveau des épines dendritiques et la présence de la machinerie traductionnelle près de ces protubérances suggère fortement l’existence d’une traduction locale d’ARNm. Ces ARNm seraient d’ailleurs acheminés dans les dendrites par des protéines pouvant lier les ARNm et assurer leur transport jusqu’aux synapses activées. Le rôle des protéines Staufen (Stau1 et Stau2) dans le transport, la localisation et dans la régulation de la traduction de certains ARNm est bien établi. Toutefois, leur rôle précis dans la plasticité synaptique demeure encore inconnu. Ainsi, cette thèse de doctorat évalue l’importance des protéines Staufen pour le transport et la régulation d’ARNm dans la plasticité synaptique. Nous avons identifié des fonctions spécifiques à chaque isoforme; Stau1 et Stau2 étant respectivement impliquées dans la late-LTP et la LTD dépendante des récepteurs mGluR. Cette spécificité s’applique également au rôle que chaque isoforme joue dans la morphogenèse des épines dendritiques, puisque Stau1 semble nécessaire au maintien des épines dendritiques matures, alors que Stau2 serait davantage impliquée dans le développement des épines. D’autre part, nos travaux ont permis de déterminer que la morphogenèse des épines dendritiques dépendante de Stau1 était régulée par une plasticité synaptique endogène dépendante des récepteurs NMDA. Finalement, nous avons précisé les mécanismes de régulation de l’ARNm de la Map1b par Stau2 et démontré l’importance de Stau2 pour la production et l’assemblage des granules contenant les transcrits de la Map1b nécessaires pour la LTD dépendante des mGluR. Les travaux de cette thèse démontrent les rôles spécifiques des protéines Stau1 et Stau2 dans la régulation de la plasticité synaptique par les protéines Stau1 et Stau2. Nos travaux ont permis d’approfondir les connaissances actuelles sur les mécanismes de régulation des ARNm par les protéines Staufen dans la plasticité synaptique. MOTS-CLÉS EN FRANÇAIS: Staufen, hippocampe, plasticité synaptique, granules d’ARN, traduction, épines dendritiques. / Learning and memory are complex processes that are not completly understood at the cellular and molecular levels. It is however accepted that persistent modifications of synaptic connections, like synaptic plasticity, could be responsible for the encoding of new memories. Whereas long-term potentiation (LTP) is classically defined as a persistent and stable enhancement of synaptic connections, long-term depression (LTD) is a reduction in the efficacy of neuronal synapses. Numerous studies have identified some of the mechanisms of this phenomenon, in particular, the induction and expression mechanisms, as well as the changes in dendritic spine morphology. The most abundant type of synapse in the hippocampus is the excitatory glutamatergic synapse made on dendritic spines; the presence of the translational machinery in dendrites near spines strongly supports the concept of local mRNA translation. Moreover, those mRNA are transported in dendrites to activated synapses by RNA binding-proteins (RBP). Staufen proteins (Stau1 and Stau2) function in transport, localization and translational regulation of mRNA are now established. However, their precise roles in synaptic plasticity are still unknown. Thus, this Ph.D. thesis evaluates the importance of Staufen proteins in mRNA transport and regulation in synaptic plasticity. We have identified specific functions for each isoform; while Stau1 is implicated in late-LTP, Stau2 is required for mGluR-LTD. This specificity is also relevant for dendritic spine morphogenesis since Stau1 is involved in mature dendritic spine maintenance while Stau2 participates in dendritic spine morphogenesis at a developmental stage. Moreover, our studies have indicated that Stau1 involvement in spine morphogenesis is dependent on ongoing NMDA receptor-mediated plasticity. Finally, our results suggest that Stau2 is implicated in a particular form of synaptic plasticity through transport and regulation of specific mRNA granules required for mGluR-LTD such as Map1b. Our work uncovers specific roles of Stau1 and Stau2 in regulation of synaptic plasticity. These studies help to better understand mechanisms involving mRNA regulation by Staufen in long-term synaptic plasticity and memory. ENGLISH KEY WORDS: Staufen, hippocampus, synaptic plasticity, RNA granules, translation, dendritic spines

La mémoire et l’apprentissage sont des phénomènes complexes qui demeurent encore incertains quant aux origines cellulaire et moléculaire. Il est maintenant connu que des changements au niveau des synapses, comme la plasticité synaptique, pourraient déterminer la base cellulaire de la formation de la mémoire. Alors que la potentialisation à long-terme (LTP) représente un renforcement de l’efficacité de transmission synaptique, la dépression à long-terme (LTD) constitue une diminution de l’efficacité des connexions synaptiques. Des études ont mis à jour certains mécanismes qui participent à ce phénomène de plasticité synaptique, notamment, les mécanismes d’induction et d’expression, ainsi que les changements morphologiques des épines dendritiques. La grande majorité des synapses excitatrices glutamatergiques se situe au niveau des épines dendritiques et la présence de la machinerie traductionnelle près de ces protubérances suggère fortement l’existence d’une traduction locale d’ARNm. Ces ARNm seraient d’ailleurs acheminés dans les dendrites par des protéines pouvant lier les ARNm et assurer leur transport jusqu’aux synapses activées. Le rôle des protéines Staufen (Stau1 et Stau2) dans le transport, la localisation et dans la régulation de la traduction de certains ARNm est bien établi. Toutefois, leur rôle précis dans la plasticité synaptique demeure encore inconnu. Ainsi, cette thèse de doctorat évalue l’importance des protéines Staufen pour le transport et la régulation d’ARNm dans la plasticité synaptique. Nous avons identifié des fonctions spécifiques à chaque isoforme; Stau1 et Stau2 étant respectivement impliquées dans la late-LTP et la LTD dépendante des récepteurs mGluR. Cette spécificité s’applique également au rôle que chaque isoforme joue dans la morphogenèse des épines dendritiques, puisque Stau1 semble nécessaire au maintien des épines dendritiques matures, alors que Stau2 serait davantage impliquée dans le développement des épines. D’autre part, nos travaux ont permis de déterminer que la morphogenèse des épines dendritiques dépendante de Stau1 était régulée par une plasticité synaptique endogène dépendante des récepteurs NMDA. Finalement, nous avons précisé les mécanismes de régulation de l’ARNm de la Map1b par Stau2 et démontré l’importance de Stau2 pour la production et l’assemblage des granules contenant les transcrits de la Map1b nécessaires pour la LTD dépendante des mGluR. Les travaux de cette thèse démontrent les rôles spécifiques des protéines Stau1 et Stau2 dans la régulation de la plasticité synaptique par les protéines Stau1 et Stau2. Nos travaux ont permis d’approfondir les connaissances actuelles sur les mécanismes de régulation des ARNm par les protéines Staufen dans la plasticité synaptique. MOTS-CLÉS EN FRANÇAIS: Staufen, hippocampe, plasticité synaptique, granules d’ARN, traduction, épines dendritiques. / Learning and memory are complex processes that are not completly understood at the cellular and molecular levels. It is however accepted that persistent modifications of synaptic connections, like synaptic plasticity, could be responsible for the encoding of new memories. Whereas long-term potentiation (LTP) is classically defined as a persistent and stable enhancement of synaptic connections, long-term depression (LTD) is a reduction in the efficacy of neuronal synapses. Numerous studies have identified some of the mechanisms of this phenomenon, in particular, the induction and expression mechanisms, as well as the changes in dendritic spine morphology. The most abundant type of synapse in the hippocampus is the excitatory glutamatergic synapse made on dendritic spines; the presence of the translational machinery in dendrites near spines strongly supports the concept of local mRNA translation. Moreover, those mRNA are transported in dendrites to activated synapses by RNA binding-proteins (RBP). Staufen proteins (Stau1 and Stau2) function in transport, localization and translational regulation of mRNA are now established. However, their precise roles in synaptic plasticity are still unknown. Thus, this Ph.D. thesis evaluates the importance of Staufen proteins in mRNA transport and regulation in synaptic plasticity. We have identified specific functions for each isoform; while Stau1 is implicated in late-LTP, Stau2 is required for mGluR-LTD. This specificity is also relevant for dendritic spine morphogenesis since Stau1 is involved in mature dendritic spine maintenance while Stau2 participates in dendritic spine morphogenesis at a developmental stage. Moreover, our studies have indicated that Stau1 involvement in spine morphogenesis is dependent on ongoing NMDA receptor-mediated plasticity. Finally, our results suggest that Stau2 is implicated in a particular form of synaptic plasticity through transport and regulation of specific mRNA granules required for mGluR-LTD such as Map1b. Our work uncovers specific roles of Stau1 and Stau2 in regulation of synaptic plasticity. These studies help to better understand mechanisms involving mRNA regulation by Staufen in long-term synaptic plasticity and memory. ENGLISH KEY WORDS: Staufen, hippocampus, synaptic plasticity, RNA granules, translation, dendritic spines