http://upload.etdadmin.com/etdadmin/pdfout/222759_supp_undefined_2A63E500-E9D7-11E2-925E-BE522E1BA5B1.PDF

La compréhension des mécanismes impliqués dans la formation des métastases est l’un des défis majeurs de la recherche sur le cancer. En effet, la formation de métastases est la cause principale de mortalité chez les patients atteints du cancer. L'influence du microenvironnement tumoral fait partie intégrante de la recherche et plusieurs études démontrent qu’il joue un rôle primordial dans l’invasion des cellules tumorales. L’une des caractéristiques du microenvironnement tumoral est l’hypoxie. Les cellules cancéreuses ont développé différentes stratégies afin de survivre dans ce microenvironnement. Des études récentes rapportent que des mécanismes posttranscriptionnels sont induits par l’hypoxie tels que le routage intracellulaire de molécules d'adhésion, de protéases et l’activation de facteurs de croissance, et qu’ils influencent le phénotype métastatique des cellules cancéreuses. L’étape importante dans l’initiation de la formation des métastases est la dégradation de la membrane basale de la tumeur et de la matrice extracellulaire. Les cellules cancéreuses ont développé des stratégies afin de faciliter leur migration dont l’une est la formation d'invadopodes. Malgré plusieurs études sur la biogenèse et les fonctions de ces structures, peu de travaux ont été accomplis concernant l’influence du microenvironnement tumoral hypoxique sur la formation et les fonctions des invadopodes. Les travaux présentés dans cette thèse portent sur l’étude des mécanismes induits par l’hypoxie dans l’invasion des cellules cancéreuses. Dans le premier chapitre de la section résultats, nous démontrons que l’hypoxie induit une relocalisation stratégique de la furine, une convertase de pro-protéines, dans une boucle de recyclage en périphérie cellulaire. Nous démontrons que la redistribution de la furine favorise l’invasion cellulaire en condition hypoxique. Dans le deuxième chapitre, nous avons étudié l'impact de la relocalisation de la furine en hypoxie sur la maturation de substrats tumorigéniques. Nos résultats indiquent que l’hypoxie favorise la maturation du TGFß par la furine dans des vésicules acides. Nous démontrons que la présence d’une histidine, sensible au pH, située au site de clivage du pro-TGFß par la furine influence la capacité de cette dernière à cliver la pro-protéine. Enfin, dans le dernier chapitre, nous démontrons que l’induction hypoxique de la formation des invadopodes est principalement causée par la signalisation dépendante de Smad3 du TGFß. Nous identifions la HDAC6 comme étant un régulateur de la signalisation du TGFß en hypoxie. La HDAC6 permettrait la libération de Smad3 du réseau de tubuline, et suite à la liaison du TGFß sécrété à son récepteur, permettrait la phosphorylation et la translocation nucléaire de Smad3 afin d'induire ses gènes cibles. L'ensemble de ces travaux a permis d'identifier des molécules clés impliquées dans la formation des invadopodes en condition hypoxique. Nos travaux ont contribué de manière substantielle à nos connaissances des mécanismes impliqués dans l’invasion cellulaire dans le microenvironnement hypoxique. Nos résultats permettront en outre l’identification de cibles thérapeutiques potentielles qui pourraient servir à inhiber l’invasion cellulaire et la formation de métastases.

Le cil primaire (CP) est une organelle présente chez l’Homme dans la grande majorité des cellules. Lors du développement le CP est d’une importance capitale, puisqu’il contrôle les voies de signalisation comme Hedgehog ou Wnt. Certaines pathologies génétiques affectant spécifiquement le CP, engendrent une obésité. Au cours de ma thèse je me suis intéressé à l’évolution du CP au cours de l’adipogenèse des cellules souches mésenchymateuses humaines. Les résultats que nous avons obtenus indiquent que le cil est présent dans les cellules indifférenciées, qu’il subit une élongation importante suite à l’induction de la différenciation, suivi d’une diminution de sa taille et fini par disparaitre dans les adipocytes. L’élongation de la taille du cil ne semble pas affecter la localisation des protéines qui lui sont associées comme Kif3-A ou Smoothened, une protéine importante de la voie Hedgehog. Néanmoins, il apparait que la voie de signalisation Hedgehog est inhibée après trois jours de différenciation et que les cellules ont développé une résistance à Sonic Hedgehog. La déacétylase de la tubuline acétylée HDAC6 est apparue comme étant une bonne cible puisque son expression augmente au cours de la différenciation et qu’elle est décrite pour être responsable de la perte du cil pendant la mitose. Les données que nous avons obtenues ont permis de montrer que l’inhibition, ou la surexpression d’HDAC6 au cours de l’adipogenèse engendrent une inhibition de l’élongation du cil associée à une forte inhibition de la différenciation adipocytaire. Ces résultats permettront, à terme de mieux comprendre les liens entre le cil primaire et la différenciation adipocytaire. / The primary cilium (PC) is an organelle present in almost all cell types of the organism. During development, the PC plays an important function by driving signaling pathways such as Hedgehog or Wnt. Some genetic syndromes affecting specifically the PC are associated with obesity. My project has consisted to analyze the evolution of the PC during adipocyte differentiation of human mesenchymal stems cells. Our results indicate that the PC is present in undifferentiated cells, then it undergoes a strong elongation at the beginning of the differentiation followed by a decreased of its size, and disappears in differentiated cells. This increase in the cilium size does not affect the localization of its associated proteins such as KIF3-A and Smoothened an important protein of the Hedgehog signaling pathway. However, this pathway is inhibited after three days of differentiation and cells have developed a Sonic Hedgehog resistance. The tubulin deacetylase HDAC6 appeared as a good target because its expression increases during differentiation and it is known to be responsible for the loss of the cilium during mitosis. Our data show that an inhibition or an overexpression of HDAC6 lead to a decrease in the cilium elongation associated with an inhibition of adipocyte differentiation. These results will ultimately lead to a better understanding of the connections between the PC and adipocyte differentiation.

Le gène suppresseur de tumeur LKB1 code une sérine/thréonine kinase qui régule le métabolisme énergétique et la polarité cellulaire. Son action biologique s'exerce en partie via la protéine kinase activée par l'AMP (AMPK), substrat de LKB1 dont la phosphorylation stimule l'activité catalytique. Nous avons récemment mis en évidence une interaction entre LKB1 et la déacétylase HDAC6. HDAC6 régule principalement l'état d'acétylation de protéines localisées dans le cytoplasme telles que la molécule chaperon HSP90, la tubuline α, et la cortactine. HDAC6 contrôle la stabilité des protéines liées à HSP90 mais agit aussi sur la polarité et l'adhérence des cellules. De plus, HDAC6 répond à différentes situations de stress cellulaire en favorisant le transport des protéines polyubiquitinées vers les aggrésomes, où celles ci sont dégradées, et en promouvant la formation des granules de stress, complexes ribonucléoprotéiques participant au stockage des ARNm et au blocage de la traduction. Mon projet de recherche a porté sur les conséquences fonctionnelles de l'interaction entre LKB1 et HDAC6. J'ai ainsi pu montrer que la formation de ce complexe est renforcée en condition de stress oxydatif et thermique. Dans cette situation biologique, LKB1 interfère avec la capacité de HDAC6 à fixer les protéines ubiquitinylées, et par conséquent prévient la formation des aggrésomes et des granules de stress. A l'inverse, LKB1 stimule l'activité déacétylase de HDAC6, et cette action de LKB1 est requise pour la migration orientée des cellules ainsi que pour la polarisation apico-basale dans un modèle de culture d'entérocytes. Ce travail nous a ainsi permis d'identifier un nouvel effecteur de LKB1 qui intervient dans la réponse au stress et dans la polarisation cellulaire. Il s'agit aussi de la première mise en évidence d'une régulation de l'activité de liaison à l'ubiquitine de HDAC6. Ces données suggèrent que LKB1, via son effet sur HDAC6, pourrait limiter la réponse adaptative des cellules soumises à des stress exogènes et endogènes, comme ceux que les cellules en voie de transformation rencontrent dans leur microenvironnement, une propriété qui pourrait s'avérer essentielle pour son activité de suppresseur de tumeur / The tumor suppressor LKB1 is a serine-threonine kinase that acts as a critical regulator of energy homeostasis and cell polarity 1,2. LKB1 relays its intracellular signal through the AMP-activated protein kinase (AMPK) as well as twelve additional members of the AMPK sub-family 3-5. However, despite the identification of these LKB1 effectors, the mechanisms that underlie LKB1-mediated biological effects remain incompletely understood. We now report that LKB1 interacts with and phosphorylates HDAC6, a deacetylase that protects cells against extrinsic insults through its ability to ligate polyubiquinated misfolded proteins and to dynamically associate with both the microtubule and the actin cytoskeleton networks 6. We further found that the formation of the LKB1-HDAC6 complex was promoted in response to diverse stressful stimuli. As a consequence, HDAC6 ubiquitin-binding activity was inhibited, thus impeding the formation of aggresomes and stress granules, two transient cellular structures that, respectively, prevent the accumulation of aggregated proteins 7 and remodel messenger ribonucleoprotein complexes following stresses that block translation 8. Collectively, these data identify HDAC6 as a key downstream component of the LKB1 signalling pathway. Our findings further suggest that LKB1, via its inhibitory effect on HDAC6 ubiquitin-binding activity, limits the cellular adaptive response to a protracted stress, a distinctive biological property that is likely to contribute to its tumor-suppressive function

<p>Advances in modern medicine have helped to prolong human life. These advancements coupled with an ever-increasing population means that diseases associated with aging will become more prevalent in the coming years. As such, it is critical to understand the pathogenesis of disease where aging is the main risk factor. While not widely known, age is in fact a large risk factor in development of obesity and metabolic syndrome. More widely known and discussed are the neurodegenerative diseases that occur late in life. While age as a risk factor is a common point between these types of pathology, there are other similarities, such as the interaction between lipid metabolism and mitochondrial health. </p><p>To study the overlap between obesity and neurodegeneration, we investigated two pathways that regulate both. First, we find that loss of cytoplasmic deacetylase HDAC6 leads to aberrant accumulation of lipid in vitro and in vivo. HDAC6 knock-out (KO) mice gain more weight than WT counterparts after a high-fat diet regimen. Additionally, the intermediary metabolism of cells lacking HDAC6 is disrupted as they increase glucose uptake while downregulating fatty acid oxidation. HDAC6 not only plays a role in lipid metabolism, but regulates mitochondrial dynamics. Upon glucose-withdrawal, HDAC6 KO cells fail to elongate their mitochondria and display increased levels of mitochondrial toxic by-products. Therefore, HDAC6 has critical roles in lipid homeostasis and mitochondrial health. </p><p>The other pathway we investigated is critical in neurodegenerative disease, Parkinson's disease. Parkin, an E3 ubiquitin ligase, flags damaged mitochondria for destruction so they do not poison the other functional organelles. We found that Parkin promotes lipid remodeling at the surface of the mitochondria. Phosphatidic acid (PA) accumulates shortly after mitochondrial damage while diacylglycerol (DAG) appears several hours later. This lipid accumulation is dependent upon Parkin's translocation and E3 ligase activity. Additionally, we found that lipin-1, a PA phosphatase, and endophilin B1 (EndoB1) are critical for DAG accumulation and effective mitochondrial clearance. </p><p>Through this work, we show that two proteins critical in quality control mechanisms also play significant roles in energy homeostasis. We aim to highlight this overlap and posit that common diseases of aging, though presenting differently, might have disruptions in the same basic process.</p> / Dissertation

The AC133 epitope of the pentaspan transmembrane glycoprotein CD133 has been used as a cell-surface marker for normal and cancer stem cells from a broad range of tissue types. Despite the utility of CD133 as a marker, little is known regarding its regulation and biological function. To study these poorly understood aspects of CD133, I took two main experimental approaches: RNA interference (RNAi) screening and affinity purification coupled with mass spectrometry (AP-MS) to identify CD133 regulatory genes and CD133 protein-protein interactions (PPIs), respectively. Both of these experimental approaches relied on a human embryonic kidney (HEK) 293 cell line that exogenously expresses affinity tagged CD133 (HEK293/AC133). This cell line allowed me to perform a large-scale RNAi screen to interrogate 11,248 genes for their involvement in cell-surface AC133 recognition. This resulted in the identification of the N-glycosylation pathway as a direct contributor to CD133 plasma membrane localization and cell-surface AC133 detection. I used the same RNAi screening approach on the colon adenocarcinoma cell line Caco-2, which express CD133 from its native promoter, to identify factors that regulate endogenous CD133 transcription. I was able to demonstrate that AF4 promotes CD133 transcription in a number of cancer cell lines. Furthermore, I showed that CD133 expression in an acute lymphoblastic leukemia (ALL) cell line SEM, which is dependent on the mixed-lineage leukemia (MLL)-AF4 gene fusion, is critical for the viability of these cells. To gain further insight into the function of CD133, I performed AP-MS using HEK293/AC133 cells to identify CD133 PPIs. I identified histone deacetylase 6 (HDAC6) as a CD133 protein interaction partner. I found that HDAC6 negatively regulates CD133 trafficking into the endosomal-lysosomal degradation pathway. CD133 binds HDAC6 to prevent inhibition of HDAC6 deacetylase activity by phosphorylation. Protection of HDAC6 from phosphorylation promotes HDAC6 deacetylation of β-catenin, which results in β-catenin dependent signalling and the suppression of cancer cell differentiation. My thesis provide functional roles for CD133 as a pro-proliferative protein and as a key signalling protein in certain cancer cell lines.

The AC133 epitope of the pentaspan transmembrane glycoprotein CD133 has been used as a cell-surface marker for normal and cancer stem cells from a broad range of tissue types. Despite the utility of CD133 as a marker, little is known regarding its regulation and biological function. To study these poorly understood aspects of CD133, I took two main experimental approaches: RNA interference (RNAi) screening and affinity purification coupled with mass spectrometry (AP-MS) to identify CD133 regulatory genes and CD133 protein-protein interactions (PPIs), respectively. Both of these experimental approaches relied on a human embryonic kidney (HEK) 293 cell line that exogenously expresses affinity tagged CD133 (HEK293/AC133). This cell line allowed me to perform a large-scale RNAi screen to interrogate 11,248 genes for their involvement in cell-surface AC133 recognition. This resulted in the identification of the N-glycosylation pathway as a direct contributor to CD133 plasma membrane localization and cell-surface AC133 detection. I used the same RNAi screening approach on the colon adenocarcinoma cell line Caco-2, which express CD133 from its native promoter, to identify factors that regulate endogenous CD133 transcription. I was able to demonstrate that AF4 promotes CD133 transcription in a number of cancer cell lines. Furthermore, I showed that CD133 expression in an acute lymphoblastic leukemia (ALL) cell line SEM, which is dependent on the mixed-lineage leukemia (MLL)-AF4 gene fusion, is critical for the viability of these cells. To gain further insight into the function of CD133, I performed AP-MS using HEK293/AC133 cells to identify CD133 PPIs. I identified histone deacetylase 6 (HDAC6) as a CD133 protein interaction partner. I found that HDAC6 negatively regulates CD133 trafficking into the endosomal-lysosomal degradation pathway. CD133 binds HDAC6 to prevent inhibition of HDAC6 deacetylase activity by phosphorylation. Protection of HDAC6 from phosphorylation promotes HDAC6 deacetylation of β-catenin, which results in β-catenin dependent signalling and the suppression of cancer cell differentiation. My thesis provide functional roles for CD133 as a pro-proliferative protein and as a key signalling protein in certain cancer cell lines.

Autophagy is compromised in Parkinson’s disease (PD) with a number of PD-associated genetic mutations leading to its dysregulation. Leucine-rich repeat kinase (LRRK2) mutations, causative of PD, aberrantly enhance autophagy. Our lab elucidated a LRRK2 gene regulatory network identifying transcripts showing coordinated expression level changes associated with PD. Histone deacetylase 6 (HDAC6) was found to be an important interactor with LRRK2, regulating many of the same transcripts. The majority of these transcripts associate with autophagy and the lysosomal complex. I hypothesized that LRRK2 interacts with HDAC6 to regulate autophagy. Silencing of HDAC6 in SH-SY5Y normalized the autophagosomal size altered by expression of PD-linked LRRK2 mutants. This work identified a key role for HDAC6 in mediating the autophagic dysfunction induced by the mutant LRRK2. In addition to autophagy, stress granule (SG) formation has emerged as a compelling mechanism in the pathogenesis of PD. RNA-binding proteins (RBPs), such as T-cell intracellular antigen-1 (TIA-1), are major component of SGs. I observed TIA-1 translocating from the nucleus to the cytoplasm in PD cortex without forming SGs. Hu antigen D (HuD) also showed changes, with the RBP more present in the cytoplasm than the nucleus in PD with no SGs observed. These preliminary studies lead to the hypothesis that low levels of SGs result from an inhibition by alpha-synuclein (syn), or hyperactive autophagy. For that purpose, brain tissues from a mouse model of PD (A53T-syn transgenic mouse) were examined by immunohistochemistry. There was no difference in TIA-1 expression in control and A53T-syn expressing mouse brains, or SG formation in primary neurons after treatment with recombinant A53T fibrils. To determine whether the lack of SGs in PD brain was due to activation of autophagy, BE-M17 cells were treated with rapamycin, an autophagy activator, which decreased SGs by 50%. Overexpression of TIA-1 in BE-M17 cells under arsenite treatment also increased autophagosomal size by 50%, indicating co-regulation of SGs and autophagy. My work indicates that the pathophysiology of PD is associated with a loss of SGs due to elevated activity of autophagy, presumably due to PD-linked LRRK2 mutations. This co-regulatory network may be a potential therapeutic target of PD.

Capsid of mouse polyomavirus (MPyV) is composed from three structural proteins: major structural protein VP1 and minor structural proteins VP2 and VP3. Posttranslational modifications may affect functions of proteins. This work deals with acetylation of MPyV structural proteins and its impact on the viral replication cycle. First part of the thesis is focused on acetylation of VP1. We showed that the VP1 protein is acetylated in viral particles and that interaction of VP1 with minor proteins supports VP1 acetylation. Further, we showed that cytoplasmatic deacetylase, histone deacetylase 6 (HDAC6), is important for virus infectivity. Overexpression of HDAC6 decreased MPyV infectivity, also decreased infectivity was exhibited by virus isolated from HDAC6 knock out cells. In addition, VP1 protein of virus from HDAC6 knock out cells was more acetylated in comparison with virus from parental cell line. These data suggest that VP1 is substrate for HDAC6. Second part of the thesis is focused on the characterization of N-terminal acetylation of VP3 minor structural protein. It has been previously shown that VP3 protein is N-terminally acetylated and MyPV with mutated (unacetylated) form of VP3 protein is non-infectious. The main aim of this part is to prove the hypothesis that N-terminal acetylation is...

Our understanding of non-histone acetylation as a means of cellular regulation is in its infancy. Using a mass spectrometry approach we identified acetylated lysine residues and monitored acetylation changes across the proteome as a consequence of metabolic stress (hypoxia). We observed changes in acetylation status of non-histone lysines in tumor cells. Through the use of small molecule inhibitors of histone deacetylase enzymes (HDACs) and siRNA screening identified HDAC6 as a pro-survival regulator of lysine acetylation during hypoxia. The phospho-binding protein 14-3-3ζ acts as a signaling hub controlling a network of interacting partners and oncogenic pathways. We show here that lysines within the 14-3-3ζ binding pocket and protein-protein interface can be modified by acetylation. The positive charge on two of these lysines, K49 and K120, is critical for coordinating 14-3-3ζ-phosphoprotein interactions. Through screening, we identified HDAC6 as the K49/K120 deacetylase. Inhibition of HDAC6 blocks 14-3-3ζ interactions with two well-described interacting partners, Bad and AS160, which triggers their dephosphorylation at S112 and T642, respectively. Expression of an acetylation-refractory K49R/K120R mutant of 14-3-3ζ rescues both the HDAC6 inhibitor-induced loss of interaction and S112/T642 phosphorylation. Furthermore, expression of the K49R/K120R mutant of 14-3-3ζ inhibits the cytotoxicity of HDAC6 inhibition. These data demonstrate a novel role for HDAC6 in controlling 14-3-3ζ binding activity.