Secreted proteins play important roles in many cellular functions and molecular processes. Because secreted proteins potentially enter the blood stream, they can serve as valuable measures of health and disease useful for disease diagnosis and prognosis, therapeutic target identification, and patient stratification in personalized medicine. Consequently, significant interest exists in secreted protein analysis within complex biospecimens, particularly blood but significant bioanalytical challenges including the wide protein dynamic range >10 orders of magnitude remain. The cellular secretome therefore represents a viable alternative to direct biomarker discovery in biofluids. Finally, cellular systems are amenable to labeling for the production of intact stable isotope labeled (SIL) proteins that can be used as global internal standards for quantitative proteomics. In this dissertation, two secretome-focused studies were undertaken. The first study involving candidate biomarker discovery in radiation-induced autophagy utilized the p53-null and inducible H1299 non-small cell lung cancer (NSCLC) secretome. The study identified 364 secreted proteins that were mainly associated with exosomes (N=224) and chaperone activity (N=21). CHGB and SCG2 were identified as potential population-based biomarkers (for patient stratification) due to their consistent overexpression in p53-null H1299 cell secretomes compared to p53-wt cells before and after radiation. FAM3C, CANX, EIF5A, GPI, and TXNRD1 were identified as candidate biomarkers for patient prognosis following radiotherapy due to their differential expression only in response to radiation treatment. In the second study, a comprehensive glycoproteomics characterization of the SILAC-labeled HepG2 secretome was undertaken. 1635 SIL proteins, 492 of which were major plasma proteins including 192 cancer biomarkers were identified with high sequence coverage spanning six orders of magnitude. EDTA plasma spiked with the SIL secretomes yielded 63 proteins that were quantified with H/L ratios in all samples out of 1405 total proteins identified. Additionally, LC-MS/MS analysis of the Con A and WGA enriched 72h secretome:plasma sample afforded an opportunity to clearly distinguish between glycoproteins in plasma and the HepG2 secretome that share/differ in N-glycan structures. Collectively, the two studies reveal the suitability of the H1299 cancer cell secretome as an experimental model for biomarker studies and support the HepG2 secretome as a viable platform for producing SIL glycoproteins.

Certains agents chimiothérapeutiques tels que les anthracyclines ou l'oxaliplatine induisent une mort cellulaire immunogène, ce qui implique que les cellules mourrantes du patient servent de vaccin thérapeutique en stimulant une réponse immunitaire antitumorale. La mort cellulaire immunogène est caractérisée par la libération de signaux d'alarme par la cellule tumorale mourante qui permettent l’activation du système immunitaire. En premier lieu, l'exposition de la calréticuline à la surface de la cellule tumorale mourante va agir comme un signal de type «eat-me» pour les cellules dendritiques. Une fois relâchée, la protéine nucléaire HMGB1 se lie au récepteur TLR4 afin de faciliter la présentation antigénique. Les cellules mourantes vont également libérer de l'ATP qui agit sur les récepteurs P2X7 et active l’inflammasomme NLRP3, conduisant à la libération d'IL-1β et ainsi à l’activation des cellules T CD8+ productrices d’IFN-γ. L’autophagie est un mécanisme cellulaire qui est activé en réponse à la chimiothérapie. L'autophagie signifie «self-ating», il s'agit d'un processus cellulaire activé par diverses conditions de stress, par lequel les cellules peuvent dégrader les protéines et les organites. Il peut aussi être induit par un stress du réticulum endoplasmique. Ce dernier étant également impliqué dans l'exposition de la calréticuline pendant la mort cellulaire immunogène, nous avons au cours de cette étude cherché à déterminer le rôle de l'autophagie dans la mort cellulaire immunogène. Nous avons constaté que l'autophagie est nécessaire pour la libération de l'ATP après un traitement par des chimiothérapies immunogènes, en observant que le nockdown de gènes essentiels de l'autophagie limitait la sécrétion d'ATP. Nous avons également observé que des cellules déficientes pour l'autophagie traitées par une chimiothérapie immunogène sont incapables d’immuniser des souris contre une injection de cellules vivantes. En outre, les tumeurs déficientes pour l’autophagie ne répondent pas à un traitement systémique immunogène dans des souris immunocompétentes et continuent à proliférer en comparaison à des tumeurs “wild-type”. De plus, nous avons montré que les cellules déficientes pour l'autophagie ne sont pas en mesure de recruter des cellules dendritiques dans le lit tumoral ou d'induire l’activation des cellules T CD8+. A l'inverse, l'inhibition des enzymes de dégradation de l’ATP extracellulaire accroit les concentrations d'ATP dans les tumeurs déficientes pour l'autophagie, ce qui rétablit le recrutement des cellules immunitaires dans le lit tumoral et restaure la réponse chimiothérapeutique des cancers déficients pour l'autophagie. Ainsi, cette étude a montré l'importance de l'autophagie dans la réponse anti-tumorale spécifique, après traitement par des chimiothérapies immunogènes. Ces résultats ouvrent de nouvelles perspectives dans le concept de la mort cellulaire immunogène. / In recent years it has been demonstrated that some chemotherapeutic agents such as anthracyclines or oxaliplatin can induce a type of tumor cell death that is immunogenic, implying that the patient’s dying cancer cells serve as a therapeuticvaccine that stimulates an antitumor immune response, which in turn can control or eradicate residual cancer cells. Immunogenic cell death is characterized by the emission of danger signals from the dying tumor cell, which activate the immune system. At first the exposure of calreticulin, acts as an «eat-me» signal for dendritic cells (DCs). Once released, the nuclear protein HMGB1 binds to TLR4 on DCs, facilitating antigen processing and presentation. The dying tumor cells also releases ATP, which acts on P2X7 receptors on DCs and activates the NLRP3 inflammasome, leading to IL-1β release, necessary for IFN-γ-producing CD8+ T cell activation. Autophagy literally ‘self-eating’ is a cellular process activated in response to various conditions of cellular stress, whereby cells can liberate energy resources via the degradation of proteins and organelles. Recently autophagy has been found activated in response to chemotherapy and in this project we aimed to determine the potential role of autophagy in immunogenic cell death. We found that autophagy isrequired for the release of ATP in response to immunochemotherapeutic treatment, as we observed that the knockdown of essential autophagy-related genes abolished its secretion. We observed that autophagy deficient cells treated with immunogenic cell death inducers failed to immunize mice against a re-challenge with living cells. Furthermore, autophagy deficient tumors growing on immunocompetent mice did not respond to systemic immunogenic treatment and continued proliferating in contrast to autophagy proficient tumors. We showed that autophagy deficient cells were neither able to recruit DCs into the tumor bed nor to activate CD8+ T cells. Conversely, the inhibition of extracellular ATP degrading enzymes increased extracellular ATP concentrations in autophagy deficient tumors, which reestablished the recruitment of immune cells into the tumor bed, and restored chemotherapeutic responses in autophagy-deficient cancers. Altogether, this study showed the importance of autophagy in tumor-specific immune response after treatment with chemotherapy, thus giving new insights into the concept of immunogenic cell death.

La LMMC est une hémopathie est caractérisée par une monocytose persistante (>1.10^9/L) ainsi que par la présence anormale de granulocytes immatures dans le sang des patients. Actuellement, il n’existe aucun traitement ciblé, il est donc essentiel de mieux caractériser et de comprendre les mécanismes qui causent cette monocytose afin d’établir de nouvelles stratégies thérapeutiques et de proposer des traitements ciblés. A partir de monocytes primaires humain stimulés par du CSF-1, nous avons observé que l’autophagie, est induite et qu’elle est dépendante des protéines Beclin1, ATG5, ATG7 et ULK1. Cette autophagie est également nécessaire à la différenciation macrophagique. Nous avons déterminé que l’AMP kinase (AMPK), est nécessaire à l’induction de l’autophagie et par conséquent à la différenciation macrophagique. Nous avons également découvert que le récepteur purinergique P2RY6 via la stimulation du CSF-1R active l’axe PLCβ3-CaMKKβ-AMPK-ULK1 conduisant à l’induction de l’autophagie nécessaire à la différenciation macrophagique. Enfin, dans le cadre de la LMMC, où la différenciation macrophagique est altérée, nous avons confirmé que la présence de granulocytes immatures est responsable de l’inhibition de l’axe P2RY6-AMPK. Dans ce contexte, nous avons observé que l’ajout d’agonistes du P2RY6 est capable d’une part de réinduire l’expression de l’AMPK et d’autre part de restaurer la différenciation macrophagique. L’ensemble de ces résultats souligne l’importance de l’induction du processus autophagique au cours de la différenciation macrophagique des monocytes mais également désigne l’axe P2RY6-AMPK comme cible thérapeutique potentielle dans le traitement de la LMMC. / CMML is a hematologic malignancy characterized by a persistent monocytosis (> 1.10^9/ L) and an abnormal presence of immature granulocytes in the blood of patients. Currently, there is no targeted therapy for this disease, so there is a real need to better understand the mechanisms leading to monocytosis in order to establish new therapeutic strategies and propose targeted treatments for CMML patients. From primary human monocytes, stimulated by CSF-1 to induce their differentiation into macrophages, we found that autophagy is induced during this process and is required for proper macrophagic differentiation. Then we were able to show that AMPK kinase (AMPK) is required for the induction of autophagy and therefore macrophage differentiation. We also found that the P2RY6 purinergic receptor via the CSF-1R stimulation, activate the PLCβ3-CaMKKβ-AMPK-ULK1 axis leading to the induction of autophagy, which is necessary for the macrophagic differentiation. Finally, in CMML context, where the macrophagic differentiation is impaired, we confirmed that the presence of immature granulocytes is responsible for the inhibition of AMPK P2RY6-axis. In this context, we observed that the addition of agonists P2RY6 is first capable of re-inducing the expression of AMPK and then restoring the macrophagic differentiation. All of these results emphasize the importance of autophagy induction during macrophage differentiation of monocytes and highlight the P2RY6 AMPK-axis as a potential therapeutic target in the treatment of CMML.

Autofagia e ativação dos inflamassomas são dois processos celulares autônomos, que podem interagir entre si. Estes processos participam ativamente do controle de infecções ocasionadas por diversos patógenos intracelulares. Anteriormente, descrevemos o inflamassoma NLRP3 no controle do T. cruzi, agente causador da Doença de Chagas. Entretanto, o papel da autofagia nesse controle não era conhecido. Neste trabalho, foi demonstrado que o T. cruzi induz aumento da expressão de LC3-II, formação de autofagossomas e autolisossomos em macrófagos peritoneais (MPs) de camundongos C57BL/6 selvagens. Ainda, a manipulação farmacológica da autofagia interferiu com a capacidade dos MPs em controlar a infecção pelo T. cruzi, apontando esse processo como um mecanismo efetor envolvido no controle do protozoário por macrófagos. Nesse contexto, NLRP3 parece funcionar como um modulador do processo autofágico. Na ausência de NLRP3, a manipulação farmacológica da autofagia não interferiu no controle do T. cruzi por MPs. Isso se correlacionou ao fato do fluxo autofágico se encontrar interrompido em MPs de camundongos deficientes para NLRP3 em resposta à infecção, mas não em resposta à rapamicina e starvation. A razão do bloqueio no fluxo autofágico parece ser a incapacidade de MPs deficientes em NLRP3 em formar autolisossomos, fato visualizado em microscopia confocal e eletrônica. Interessante, NLRP3 parece agir independente de caspase-1/11 na regulação da autofagia. Por outro lado, a análise da expressão de genes autofágicos por PCR-array revelou que MPs de animais deficientes em NLRP3 apresentam alta expressão basal de genes relacionados com a formação e maturação de autofagossomas e autolisossomas. Em contrapartida, a infecção pelo T. cruzi inibe a expressão desses genes na ausência de NLRP3, ao contrário da indução observada em MPs selvagens. Juntos, esses dados mostram que NLRP3 induz autofagia funcional em resposta ao T. cruzi, sendo sua presença fundamental para impedir o escape do parasita pela inibição de genes autofágicos. / Autophagy and inflammasome activation are two cell-autonomous and cross-regulated processes involved in host resistance against infections. Our group previously described that NLRP3 inflammasome is required for the control of T. cruzi, causative agent of Chagas disease. However, the involvement of autophagy in this process was largely unknown. Here, we demonstrated that T. cruzi is able to induce an increase in LC3II expression, formation of autophagosome and autolysosomes in peritoneal macrophages (PMs) from C57BL/6 mice. Moreover, the pharmacological modulation of autophagic machinery influenced the trypanocidal ability of PMs, pointing out autophagy as an effector mechanism to control T. cruzi infection. In this sense, NLRP3 seems to be involved in the modulation of the autophagic process. In the absence of NLRP3, the pharmacological modulation of autophagy did not interfere in the control of T. cruzi by PMs. Furthermore, autophagic flux is blocked in these cells in response to infection, but not in response to rapamycin and starvation. In fact, whereas T. cruzi induces the formation of large autolysosomes (LC3&#43 and Lysotracker&#43)-containing amastigotes in WT macrophages, only small and single positive vesicles are found in the absence of NLRP3. Interesting, NLRP3 appears to act independently of caspase-1/11 on the regulation of autophagy. On the other hand, the PCR-array analysis of autophagic genes demonstrated that NLRP3-/- PMs have higher basal expression of genes related to the formation and maturation of autophagosomes and autolysosomes in comparison to WT cells. In contrast, T. cruzi inhibited the expression of these genes in the absence of NLRP3, unlike the induction observed in WT PMs. Together, these data show that NLRP3 induces functional autophagy in response to T. cruzi being its presence required to overcome the escape of the parasite by preventing its inhibition of autophagic genes.

La maladie d'Alzheimer (MA) est une pathologie neurodégénérative principalement caractérisée par la présence de dépôts amyloïdes et d'enchevêtrements neurofibrillaires composés de protéines tau hyperphosphorylées. Tau est une protéine associée aux microtubules qui possède de nombreux sites de phosphorylation pouvant être phosphorylés par différentes kinases. En plus de la pathologie tau, on observe également dans le cerveau des patients atteints de la MA, une apparition précoce d’altérations métaboliques ainsi qu’une perte synaptique qui est à l’origine du développement des troubles cognitifs. En effet, les synapses sont des connexions neuronales essentielles pour la formation de la mémoire qui nécessitent une importante quantité d’énergie pour maintenir leurs fonctions. Depuis plusieurs années, des études suggèrent que l’AMPK – senseur métabolique essentiel des cellules – pourrait être impliquée dans le développement de la MA. En effet, des travaux réalisés in vitro ont permis de montrer que l’AMPK est une kinase de tau. Par ailleurs, il a été mis en évidence que chez les patients atteints de la MA, l’AMPK est dérégulée dans les neurones en dégénérescence où elle co-localise avec les protéines tau hyperphosphorylées. Enfin, des études menées dans notre équipe ont également permis de mettre en évidence que suite à une activation synaptique, l'AMPK restaure les niveaux d'énergie des neurones laissant ainsi supposer qu'une dérégulation de son activité pourrait avoir un impact néfaste sur le métabolisme neuronal. Au vu de ces données, les objectifs de mon projet de thèse ont donc été de déterminer l'impact d'une dérégulation de l’AMPK neuronale sur la pathologie tau, la perte synaptique et le métabolisme énergétique neuronal dans un modèle de culture primaire de neurones et in vivo chez la souris. La réalisation de ces objectifs nous a permis de démontrer (1) que l'AMPK régule la phosphorylation et la pathologie tau, (2) que la dérégulation de l’AMPK induit une diminution du nombre des synapses ainsi qu'une perte de la fonctionnalité du réseau neuronal, via une voie de signalisation impliquant l’autophagie et (3) que la dérégulation de l’AMPK entraine des perturbations du métabolisme énergétique neuronal. En conclusion, ce travail de thèse a permis d’apporter une meilleure compréhension sur le rôle de la dérégulation de l’AMPK dans le développement des différentes caractéristiques de la MA. Dans l’ensemble, ces données laissent fortement suggérer que l’AMPK pourrait faire le lien entre les dysfonctionnements métaboliques et l’ensemble des altérations qui se mettent en place au cours de la MA. / Alzheimer's disease (AD) is a neurodegenerative disorder mainly characterized by the presence of amyloid deposits and neurofibrillary tangles composed of hyperphosphorylated tau proteins. Tau is a microtubule-associated protein that bears many phosphorylation sites which can be phosphorylated by different kinases. Beside tau pathology, AD is also characterized by cerebral metabolic alterations and synaptic loss, the latter being responsible for the development of cognitive disorders. Indeed, synapses are essential for memory formation and require a large amount of energy to maintain their functions. Interestingly, studies have suggested that AMP-activated protein kinase (AMPK) – a crucial intracellular metabolic sensor – could be involved in the development of AD. Indeed, in vitro studies have shown that AMPK is a tau kinase. In addition, AMPK is deregulated in degenerating neurons of AD patients brain where it co-localizes with hyperphosphorylated tau proteins. Additionally, studies carried out in our team showed that upon synaptic activation, AMPK activity is essential to maintain neuronal energy levels thus suggesting that a deregulation of its activity could have harmful impact on neuronal metabolism. On the basis of these data, the objectives of this thesis were to determine the impact of neuronal AMPK deregulation on tau pathology, synaptic loss and neuronal energy metabolism in primary neurons and in vivo in mice. The achievement of these objectives allowed us to demonstrate (1) that AMPK regulated tau phosphorylation and pathology (2) that AMPK deregulation caused a decrease of synapses number as well as a loss of neuronal networks functionality, through a signaling pathway involving autophagy and (3) that AMPK deregulation impacted on neuronal energy metabolism. In conclusion, this thesis has provided a better understanding of the role of AMPK deregulation in the development of different hallmarks of AD. Altogether, these data strongly suggest that AMPK could be the link between neuronal metabolism dysfunctions and the development of the alteration that occur during AD.

京都大学 / 新制・課程博士 / 博士(医学) / 甲第24194号 / 医博第4888号 / 京都大学大学院医学研究科医学専攻 / (主査)教授 生田 宏一, 教授 伊藤 能永, 教授 中川 一路 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Modification by glycosylation gives proteins a range of diverse functions reflecting their structural variability. N-glycans regulate many biological outcomes in mammalian cells under both normal and pathological conditions. They play a major role in various pathologies such as cancer and lysosomal storage diseases. Interplay between N-glycans and other regulators, such as membrane lipid domains, in the control of signaling pathways remains poorly understood. My thesis therefore focuses on how N-glycans and membrane lipid domains oppose and/or work together at different cellular levels to regulate various processes such as receptor signaling and diffusion, endocytosis and lysosomal organelle biogenesis. Mgat5 encodes for ß1,6-N-acetylglucosaminyltransferase V that produces N-glycans, the preferred ligand for galectins. In tumor cells, galectins bind glycosylated receptors at the cell surface forming a lattice, that restricts receptor endocytosis and enhances its residency at the plasma membrane. In the first part of my thesis, I report that Galectin/receptor crosslinking opposes receptor sequestration by oligomerized caveolin-1 (Cav1) domains overriding its negative regulation of epidermal growth factor receptor (EGFR) signaling, cell surface diffusion and tumor growth. These results identify Cav1 as a conditional tumor suppressor. I also demonstrate that Cav1 is a negative regulator of lipid raft-mediated endocytosis. Cav1 indirectly regulates the internalization of cholera toxin b subunit to the Golgi apparatus independently of caveolae formation. That identifies a new role for caveolin-1 outside caveolae in the regulation of raft-dependent endocytosis Finally, Mgat5 overexpression in pneumocytes is associated with the expression of a lysosomal organelle, the multilamellar body (MLB), via autophagy. MLB expression is also a characteristic of various lysosomal storage diseases. I demonstrate that cholesterol accumulation can override the need for Mgat5 overexpression in MLB formation indicating that they may form via multiple mechanisms. However, I also demonstrate that a contribution of the autophagic pathway is a common determinant of biogenesis of MLB of various lipid compositions. In conclusion, Mgat5-dependent protein glycosylation and Cav1/raft domains therefore both function as regulators of plasma membrane interactions, endocytosis and lysosomal organelle biogenesis. Understanding of this interplay is crucial for the understanding of the mechanisms involve in various pathologies such as cancer and lysosomal storage diseases.

Endocrine resistance is a significant clinical problem in the treatment of estrogen (E2) receptor positive breast cancers. There are two ER subtypes, ERα and ERβ, which promote and inhibit breast cancer cell proliferation respectively. While ER positive breast cancers typically express a high ratio of ERα to ERβ, the acquisition of antiestrogen resistance in vitro and in vivo is associated with increased relative expression of the ERβ. On some gene enhancers ERβ has been shown to function in opposition to the ERα in the presence of E2. Here we demonstrate that exposure to two different ERβ agonists results in decreased cell viability, and produced a marked reduction in G2/M phase in antiestrogen resistant breast cancer cell line in conjunction with altered cyclin D1, and cyclin E expression relative to E2. ERβ agonists also strongly downregulated Bcl-2 expression and recruited both ERs to the Bcl-2 and pS2 E2-response elements resulting in a reduction in mRNA transcripts from both of these genes. Bcl-2 reduction correlated with increased lipidation of LC3-I to LC3-II, indicative of increased autophagic flux. Although ERβ agonist treatment alone did not induce apoptosis, remarkably, the coaddition of ERβ agonist and the autophagy inhibitor, chloroquine, resulted in robust cell death. Lastly, in vivo studies demonstrate that preferential-ERβ agonists are not estrogenic in the uterus or mammary gland. Together, these observations suggest that combined therapies including an ERβ agonist and an autophagy inhibitor may provide the basis for a safe, novel approach to the treatment of antiestrogen-resistant breast cancers.

Listeria monocytogenes is an intracellular bacterium that utilizes two phospholipases C (PLCs) and a pore-forming cytolysin (listeriolysin O, LLO) to escape the phagosome. However, prior to escape, the bacterium must overcome a number of phagosomal defenses, including autophagy and NOX2 NADPH oxidase production of reactive oxygen species (ROS). Autophagy, the cellular process of self-digestion, is a key component of innate immunity. Previously, it has been shown that L. monocytogenes is targeted by autophagy (LC3+) at 1 h post infection (p.i.) but the mechanism remains elusive. Here, I show that at 1 h p.i., diacylglycerol (DAG) and ROS production are required for autophagy targeting to the bacteria, which are predominantly in phagosomes. It has been shown that autophagy targeting of cytosolic L. monocytogenes is mediated via protein ubiquitination. However, protein ubiquitination is not associated with LC3+ bacteria at 1 h p.i.. Thus, my data suggest that distinct signals mediate autophagy targeting of L. monocytogenes depending on the location within host cells. Given that ROS mediate autophagy targeting to L. monocytogenes and that previous studies have demonstrated that ROS production limits bacterial escape, I investigated how L. monocytogenes overcomes ROS production prior to phagosomal escape. I found that LLO inhibits ROS production by preventing NOX2 NADPH oxidase localization to L. monocytogenes-containing phagosomes. LLO-deficient bacteria can be complemented by perfringolysin O, a related cytolysin, suggesting that other pathogens may also use pore-forming cytolysins to inhibit ROS production. While PLCs can activate ROS production, this effect is alleviated by LLO pore-formation. Therefore, the combined activities of PLCs and LLO allow L. monocytogenes to efficiently escape the phagosome while avoiding microbicidal ROS. Together, this thesis provides a clearer understanding of the balance between host defense versus bacterial evasion. Greater insight into host-bacterial interaction may lead to better therapeutics that can “tip the balance” in the host’s favour.

The innate immune system is important in both acute and chronic infection. In this thesis, I investigated the effect of H. pylori infection on 1) DCs, key orchestrators of the immune system, and 2) autophagy, recently identified as an important component of innate immunity. I determined that H. pylori activates the STAT3 pathway in DCs, increasing DC maturation and inducing production of IL-10, IL-12p40 and TNF-α, without IL-12p70. This cytokine profile may favour an immunoregulatory response, promoting persistent H. pylori infection. In addition I determined that H. pylori’s VacA toxin induced autophagy, ROS production and Parkin aggregation which has been implicated in mediating autophagy in response to mitochondrial damage. Thus H. pylori alters these key effectors of innate immunity which may play a role in promoting its chronic infection and disease.