Protein kinases play essential roles in orchestrating almost every step during mitosis. Aberrant kinase activity often leads to errors in the cell cycle progression which consequently becomes the underlying cause for developmental defects or abnormal cell proliferation leading to cancer. Tank Binding Kinase 1 (TBK1) is overexpressed in certain cancer types and is activated on the centrosomes during mitosis. Loss of TBK1 impairs cell division resulting in growth defects and the accumulation of multinucleated cells. Therefore, proper activation and localization of TBK1 are essential for mitotic progression. Yet, the upstream regulation of TBK1 and the function of activated TBK1 on the centrosomes is unknown. Also, the cause and consequences of overexpression of TBK1 in cancers remain to be explored. Activation of TBK1 depends on its binding to an adaptor protein which induces a conformational change leading to trans autophoshorylation on serine 172 of its kinase domain. We identified that an established innate immune response protein, NAK Associated Protein1 (NAP1/AZI2), is the adaptor required for binding and activating TBK1 during mitosis. Loss of either NAP1 or TBK1 results in the accumulation of binucleated and multinucleated cells, possibly due to several mitotic and cytokinetic defects seen in these knockout (KO) cells. We establish NAP1 as a cell cycle regulated protein which colocalizes with activated TBK1 on the centrosomes during mitosis. Furthermore, by performing an unbiased quantitative phosphoproteomics analysis during mitosis, the substrates discovered reveal that TBK1 also regulates other known cell cycle regulating kinases such as Aurora A and Aurora B. TBK1 is also an established autophagy protein and since the autophagy machinery is often impaired or remodeled to facilitate rapid cell division, we evaluated the underlying connection between TBK1 activation and autophagy. The data shows that cells lacking the essential autophagy proteins FIP200 or ATG9A exhibit overactivation and mislocalization of TBK1. By using both genetic and pharmacological inhibition of autophagy processes, we found that impaired autophagy leads to a significantly higher number of micronuclei – a hallmark for tumorigenesis that correlates with defects in mitosis and cytokinesis. Taken together our work has uncovered a novel function for the NAP1-TBK1 complex during mitosis and establishes that overactivation and mislocalization of TBK1 is a direct consequence of impaired autophagy which causes micronuclei formation. / Doctor of Philosophy / Defective cell division is the underlying cause for many human health maladies such as birth defects and cancer. Investigation into the proteins that are abnormally expressed in cancer can help us identify their physiological roles in regulating the cell cycle. Tank Binding Kinase 1 (TBK1) is often overexpressed in several types of cancer such as glioblastomas, breast, and lung cancers. It has also been extensively studied in the process of removing damaged cytosolic components from cells called autophagy. During cancer progression, cells often hijack the autophagy machinery to their advantage for abnormal cell proliferation. However, we do not completely understand the role of TBK1 in cancer pathogenesis or during normal cell division. Each cell duplicates its genomic contents and divides its organelles and cytosolic components during cell division. Centrosomes organize microtubules to attach to the duplicated genomic material to equally segregate the DNA between two daughter cells. Previous studies have shown that TBK1 is active on the centrosomes during mitosis, and the loss of TBK1 leads to reduced cell proliferation. However, the function of TBK1 and what regulates its activation on the centrosomes are unknown. Using a combination of genetic, biochemical, and molecular biology techniques, we found that an immune response protein Nak Associated Protein 1 (NAP1/AZI2) binds to TBK1 and activates it on the centrosomes during cell division. Furthermore, our study demonstrates that the loss of either NAP1 or TBK1 exhibits a multitude of different types of defects in the process of cell division. We further identified TBK1 substrates in a phosphoproteomic screen indicating that TBK1 regulates the activity of other major cell division kinases. We show that defects in autophagy machinery result in the mislocalization and overactivation of TBK1 resulting in defects during chromosome segregation, and in the formation of micronuclei. Together our study shows that an established immune response protein NAP1 regulates the function of TBK1 during cell division and there exists a connection between TBK1 activity and disrupted autophagy.

L’Adénovirus (AdV) entre dans la cellule hôte par endocytose puis s’échappe de l’endosome en lysant la membrane de ces vésicules, empêchant ainsi sa dégradation via les lysosomes. Or, les membranes endommagées sont reconnues comme des signaux de danger par le système immunitaire et peuvent déclencher une réponse antivirale, telle que l’expression d’interféron (IFN). Dans nos conditions expérimentales, nous avons montré que l’infection par l’AdV n’induit pas l’expression d’IFNβ et qu’au contraire, le virus semble inhiber cette réponse. En revanche, l’entrée du virus active TBK1 (Tank Binding Kinase 1) qui est une kinase clef de la voie IFN mais qui est également impliquée dans la régulation de l’autophagie, une voie de dégradation cellulaire. Notre laboratoire a précédemment montré que l’autophagie est activée lors de l’entrée de l’AdV, par la rupture de la membrane endosomale. Nous avons donc étudié le mécanisme d’activation et le rôle de TBK1 lors de l’infection par l’AdV. Nos résultats montrent que la rupture de la membrane endosomale induite par le virus est nécessaire pour l’activation de TBK1 et que cette kinase est recrutée spécifiquement sur les sites de dommage membranaire. De plus, nous avons montré que TBK1 est impliqué dans l’activation de l’autophagie induite par l’AdV. Cependant, contrairement à ce qui est décrit pour l’autophagie dirigée contre les bactéries, cette activation de TBK1 est indépendante de NDP52 et d’autres adaptateurs conventionnels de l’autophagie. En résumé, nos travaux montrent que l’AdV est capable de contrôler la réponse IFN et que les ruptures de membrane induites par le virus activent TBK1 et l’autophagie par un nouveau mécanisme. Nos données suggèrent un rôle conservé de TBK1 dans l’activation de l’autophagie sélective contre les agents pathogènes. / Adenoviruses enter host cells by endocytosis and then escape from the endosomal compartment by lysing the endosomal membrane, thereby preventing its degradation via lysosomes. However, damaged membranes are recognized as danger signals by the cell intrinsic immune system and trigger an antiviral response, such as expression of interferon (IFN). In our experimental conditions we have shown that adenovirus infection does not induce the expression of IFNβ. On the contrary, our data suggest that the virus appears to inhibit the IFNβ response. However, adenovirus entry activates TBK1 (Tank Binding Kinase 1), which is a key kinase of the IFN pathway but is also involved in the regulation of autophagy, a cellular degradation pathway. Our laboratory previously showed that autophagy is activated upon rupture of the endosomal membrane during adenovirus entry. We therefore studied the activation mechanism and the role of TBK1 during adenovirus infection. Our results show that virus-induced endosomal membrane rupture is required for activation of TBK1 and that this kinase is specifically recruited at membrane damage sites. In addition, we show that TBK1 is involved in the activation of autophagy induced by adenovirus. TBK1 activation is independent of NDP52 and other conventional autophagic adapters, which is in contrast to membrane damaging bacteria. Thus, autophagy targeting membrane penetrating adenoviruses differs from the one induced by bacteria. In summary our work shows that adenovirus is able to control the IFN response and that membrane rupture induced by adenoviruses activates TBK1 and autophagy by a novel mechanism. In contrast our data suggest a conserved role for TBK1 in driving selective autophagy against invading pathogens.

Platelets play a crucial role for blood hemostasis, the process that prevents bleeding. In addition, platelets have been demonstrated to promote cancer progression and cancer related complications like metastasis and thrombosis. Platelets can affect cancer related diseases either directly or by interacting with other blood cells or molecules in the circulation of individuals with cancer. The current thesis addresses the role of platelets in tumor progression and tumor-induced systemic effects of cancer, with a special focus on the effects on the vasculature. In the first paper, the role of platelets in tumor progression in histidine-rich glycoprotein (HRG)-deficient mice was addressed. We report that HRG-deficient mice show enhanced tumor growth, epithelial to mesenchymal transition (EMT) and metastasis. The enhanced platelet activity in the absence of HRG is responsible for the accelerated tumor progression. In the second paper, we demonstrate that platelet-derived PDGFB is a central player to keep the tumor vessels functional. Moreover, in a pancreatic neuroendocrine carcinoma model with PDGFB-deficient platelets, spontaneous liver metastasis was enhanced. With this finding we identify a previously unknown role of platelet derived PDGFB. In the third paper, we found that TBK1 mediates platelet-induced EMT by activation of NF-kB signaling, which suggest that TBK1 contributes to tumor invasiveness in mammary epithelial tumors. In the last paper, we report that the vascular function in organs that are neither affected by the primary tumor, nor represent metastatic sites, is impaired in mice with cancer. We show that tumor-induced formation of intravascular neutrophil extracellular traps (NETs), a fibril matrix consisting of neutrophils with externalized DNA and histones, granule proteases and platelets, are responsible for the impaired peripheral vessel function.

Autophagy is a core cytoplasmic degradation process. It is established that KRas-mutant lung cancer cells require basal autophagy for survival. However, the mechanisms that govern this are poorly understood. It has recently been suggested that selective autophagic degradation of signalling complexes may regulate downstream cell signalling pathways. Primarily, this thesis aims to uncover molecular mechanisms through which selective autophagy can regulate signalling pathways that may impact upon cancer cell proliferation. Previous work in the lab identified a putative interaction between the signalling protein TRAF3 and the autophagy protein Ndp52 via mass spectrometric screening. In this thesis I have identified TRAF3 as a target of selective autophagy in both KRas-mutant lung cancer cells and in in vitro transformed MEFs. TRAF3 is a negative regulator of a gene expression regulation pathway called alternative NF-κB. As such, autophagy of TRAF3 promotes basal activation of the alternative NF-κB signalling pathway. This basal activity supports the proliferation of cancer cells. Investigation of TRAF2, a protein closely related to TRAF3, revealed that it too associates with the autophagy pathway, but is not degraded. This is promoted by the activity of TBK1, which itself can phosphorylate TRAF2. Both TBK1 and TRAF2 promote alternative NF-κB signalling, and I investigate possible mechanisms underlying this, including changes in TRAF3 mRNA and protein levels and binding to other alternative NF-κB regulators. This thesis therefore identifies mechanisms through which basal alternative NF-κB signalling is regulated in KRas-mutant lung cancer cells, with implications for cell proliferation. Ultimately, this work provides valuable mechanistic insight to inform the use of autophagy and/or TBK1 inhibition in future cancer therapies.

京都大学 / 0048 / 新制・課程博士 / 博士(生命科学) / 甲第22595号 / 生博第428号 / 新制||生||57(附属図書館) / 京都大学大学院生命科学研究科統合生命科学専攻 / (主査)教授 藤田 尚志, 教授 松田 道行, 教授 杉田 昌彦 / 学位規則第4条第1項該当 / Doctor of Philosophy in Life Sciences / Kyoto University / DFAM

京都大学 / 新制・課程博士 / 博士(生命科学) / 甲第23440号 / 生博第461号 / 新制||生||61(附属図書館) / 京都大学大学院生命科学研究科統合生命科学専攻 / (主査)教授 野田 岳志, 教授 朝長 啓造, 教授 千坂 修 / 学位規則第4条第1項該当 / Doctor of Philosophy in Life Sciences / Kyoto University / DFAM

Thèse numérisée par la Division de la gestion de documents et des archives de l'Université de Montréal.

Das Borna Disease Virus (BDV, Bornavirus) besitzt ein einzelsträngiges RNA-Genom negativer Polarität und ist innerhalb der Ordnung Mononegavirales der Prototyp einer eigenen Virusfamilie, die der Bornaviridae. Eine außergewöhnliche Eigenschaft des Virus ist seine nukleäre Transkription und Replikation, eine weitere besteht in seiner Fähigkeit, als neurotropes Virus sowohl in vivo als auch in vitro persistente Infektionen zu etablieren. Die zugrunde liegenden Mechanismen sowohl der Replikation als auch der Persistenz sind derzeit noch unzureichend verstanden, auch deshalb, weil das Virus noch relativ „jung“ ist: Erste komplette Sequenzen des RNA-Genoms wurden 1994 publiziert und erst vor einigen Monaten gelang die Generierung rekombinanter Viren auf der Basis klonierter cDNA. Im Mittelpunkt dieser Arbeit standen das p10 Protein und das Phosphoprotein (P), die von der gemeinsamen Transkriptionseinheit II in überlappenden Leserahmen kodiert werden. <br><br> Als im Kern der Wirtszelle replizierendes Virus ist das Bornavirus auf zelluläre Importmechanismen angewiesen, um den Kernimport aller an der Replikation beteiligten viralen Proteine zu gewährleisten. Das p10 Protein ist ein negativer Regulator der viralen RNA-abhängigen RNA-Polymerase (L). In vitro Importexperimente zeigten, dass p10 über den klassischen Importin alpha/beta abhängigen Kernimportweg in den Nukleus transportiert wird. Dies war unerwartet, da p10 kein vorhersagbares klassisches Kernlokalisierungssignal (NLS) besitzt und weist darauf hin, dass der zelluläre Importapparat offensichtlich flexibler ist als allgemein angenommen. Die ersten 20 N-terminalen AS vermitteln sowohl Kernimport als auch die Bindung an den Importrezeptor Importin alpha. Durch Di-Alanin-Austauschmutagenese wurden die für diesen Transportprozess essentiellen AS identifiziert und die Bedeutung hydrophober und polarer AS-Reste demonstriert. <br><br> Die Fähigkeit des Bornavirus, persistente Infektionen zu etablieren, wirft die Frage auf, wie das Virus die zellulären antiviralen Abwehrmechanismen, insbesondere das Typ I Interferon (IFN)-System, unterwandert. Das virale P Protein wurde in dieser Arbeit als potenter Antagonist der IFN-Induktion charakterisiert. Es verhindert die Phosphorylierung des zentralen Transkriptionsfaktors IRF3 durch die zelluläre Kinase TBK1 und somit dessen Aktivierung. Der Befund, dass P mit TBK1 Komplexe bildet und zudem auch als Substrat für die zelluläre Kinase fungiert, erlaubt es, erstmalig einen Mechanismus zu postulieren, in dem ein virales Protein (BDV-P) als putatives TBK1-Pseudosubstrat die IRF3-Aktivierung kompetitiv hemmt. / The Borna Disease Virus (BDV) harbors a single stranded RNA genome of negative polarity. Within the order of Mononegavirales it is the prototype of a new virus family named Bornaviridae. Unique features of this neurotrope virus are its nuclear transcription and replication as well as its ability to establish persistent infections both in vivo and in vitro. The underlying mechanisms of BDV replication and persistence are currently not well understood amongst others due to the fact that BDV is quite a young virus: First complete sequences of the RNA genome have been published in 1994. Only a few months ago the generation of a recombinant Bornavirus from cloned cDNA has been accomplished. <br><br> The work presented here focused on the viral p10 protein and the phosphoprotein P that are both encoded by two overlapping reading frames of the transcription unit II. <br><br> Nuclear replication of the Bornavirus relies on cellular import mechanisms to allow for nuclear import of viral proteins involved in viral replication. The p10 protein has been described as a negative regulator of the viral RNA dependent RNA polymerase (L). In vitro import experiments revealed that p10 translocates into the nucleus via the classical importin alpha/beta; dependent pathway. This was unexpected since p10 does not contain a predictable classical nuclear localization signal (NLS) suggesting that the cellular import machinery is more flexible than generally believed. The first 20 amino acids mediate nuclear import and binding to the import receptor importin alpha. Analysis of di-alanine-exchange mutants identified essential amino acids and furthermore revealed the impact of hydrophobic and polar side chains in receptor binding and nuclear import. <br><br> The ability of the Bornavirus to establish persistent infections rises the question of how the virus circumvents cellular antiviral defense mechanisms, in particular the type I interferon system. This work characterizes the viral P protein as a potent antagonist of IFN beta induction. It prevents the activation of the central transcription factor IRF3 by interfering with the cellular kinase TBK1. The finding that P forms complexes with TBK1 and moreover serves as a kinase substrate allows to postulate a mechanism for the first time, in which a viral protein (BDV-P) acts as a putative TBK1 pseudo-substrate and thereby competitively inhibits IRF3 activation.

Thèse numérisée par la Division de la gestion de documents et des archives de l'Université de Montréal

Contrairement aux maladies génétiques mendéliennes, qui dépendent d’un seul gène causal, les traits quantitatifs complexes sont des caractéristiques mesurables d’organismes vivants, qui résultent de l’interaction entre plusieurs gènes et les facteurs environnementaux. La génomique fonctionnelle nous a permis d’identifier de nombreux locus génétiques liés aux caractères complexes, qui sont appelés «locus de traits quantitatifs» (QTL). Cependant, de telles études ne permettent pas une caractérisation précise de l'architecture génétique des traits complexes. Plus récemment, il est devenu possible d’identifier des locus génétiques associés aux niveaux d'expression de gènes, appelés«locus de traits quantitatifs d’expression» (eQTLs). Dans de tels cas, les variants génétiques peuvent affecter l'expression soit des gènes qui se situent dans leur voisinage (cis-eQTLs), soit de ceux qui résident plus loin (trans-eQTLs). Dans des cas particuliers, un même locus peut affecter l’expression de plusieurs gènes situés dans différents chromosomes, formant ce qu’on appelle des ‘trans-eQTLs hotspots’. Ceux-ci peuvent avoir d’importants intérêts biologiques, car ils sont généralement enrichis en gènes fonctionnellement apparentés qui peuvent influencer le même trait phénotypique. Dans cette thèse, en analysant l'expression des gènes dans des échantillons de cœurs obtenus à partir d'un panel de souches consanguines recombinantes de souris AxB / BxA, nous avons détecté un QTL lié en trans- à l'expression de 190 transcrits, dont la majorité est connue pour être sensible aux interférons de type I. Le même locus correspondait également à celui d'un cis-eQTL pour le gène Ypel5, ce qui suggère que ce dernier peut être un régulateur commun des gènes trans-eQTL. Donc, le but principale de cette thèse fut de valider biologiquement le rôle du gène cis-eQTL dans la régulation du ‘trans-eQTLs hotspots’. Les travaux présentés dans cette thèse ont montrés que la réduction de l'expression de Ypel5 dans des macrophages de souris a stimulée l'expression de plusieurs gènes qui appartiennent au ‘trans-eQTL hotspot’, et ce d’une manière dépendante d’IFNB1. Le knockdown de YPEL5 a également augmenté l’induction d’IFNB1 dans les cellules humaines HEK293T. Lorsque ces dernières ont été soumis à des stimuli qui activent les kinases TBK1 / IKBKE, nous avons détecté des interactions fonctionnelles de YPEL5 avec l'activité de ces kinases, ainsi que des interactions physiques avec IKBKE. Nos résultats préliminaires (présentés dans le chapitre3) suggèrent aussi l’implication de YPEL5 dans la régulation du cycle cellulaire et /ou de la sénescence. En conclusion, nous sommes parmi les premiers groupes à fournir des preuves biologiques montrant le rôle d'un gène cis-eQTL en tant que régulateur commun de gènes appartenant à un ‘hotspot de trans-eQTL’. La validation biologique des analyses génomiques a ainsi permis de découvrir Ypel5 comme un nouveau régulateur négatif de la réponse antivirale innée qui agit (au moins en partie) au niveau des kinases TBK1 / IKBKE. / Unlike Mendelian genetic diseases, which depend on a single causal gene, complex quantitative traits are measurable characteristics of living organisms, which result from the interaction between several genes and environmental factors. Functional genomics has allowed us to identify many genetic loci linked to complex traits, which are called "quantitative trait loci" (QTL). However, such studies do not allow an accurate characterization of the genetic architecture of complex traits. More recently, it has become possible to identify genetic loci associated with gene expression levels, called "expression quantitative trait locus" (eQTLs). In such cases, the genetic variants can affect the expression of genes that are either located in their vicinity (cis-eQTLs) or that reside further away (trans-eQTLs). In particular cases, the same locus can affect the expression of several genes located on different chromosomes, forming so-called ‘trans-eQTLs hotspots’. These may have important biological interests, as they are generally enriched in functionally related genes, which may influence the same phenotypic trait. In this thesis, by analyzing the expression of genes in hearts from a panel of AxB / BxA mouse recombinant inbred strains, we detected a QTL linked in trans- to the expression of 190 transcripts, the majority of which are known to be sensitive to type I interferon. The same locus also corresponded to that of a cis-eQTL for the Ypel5 gene, suggesting that it could be a common regulator of the trans-eQTL genes. Therefore, the main purpose of this thesis was to biologically validate the role of the cis-eQTL gene in the regulation of the ‘trans-eQTL hotspot’. The work presented in this thesis showed that the silencing of Ypel5 expression in mouse macrophages stimulated the expression of several genes that belong to the ‘trans-eQTL hotspot’ in an IFNB1-dependent manner. YPEL5 knockdown also increased IFNB1 induction in human HEK293T cells. When the latter were subjected to stimuli that activate the TBK1/IKBKE kinases, we detected functional interactions of YPEL5 with the activity of these kinases and physical interactions with IKBKE. Our preliminary results (presented in Chapter 3) suggest also the involvement of YPEL5 in the regulation of cell cycle progression and / or senescence. In conclusion, we are among the first groups to provide biological evidence showing the role of a cis-eQTL gene as a common regulator of genes belonging to a ‘trans-eQTL hotspot’. The biological validation of genomic analysis thus revealed Ypel5 as a new negative regulator of the innate antiviral response that acts (at least in part) at the level of the TBK1 / IKBKE kinases.