The maturation of early endosomes into multivesicular bodies (MVBs) and subsequent trafficking to lysosomes is an important event for the control and silencing of endocytosed membrane receptors. The endosomal-sorting complex required for transport (ESCRT) proteins appear to play a key role in this event. Phosphatidylinositol lipids including PtdIns(3,5)P2 have been implicated in the MVB-lysosomal pathway and an ESCRT-III component CHMP3 binds to this lipid in vitro. The purpose of this thesis was to investigate the link between ESCRT proteins, PtdIns(3,5)P2 and endo-lysosomal trafficking. Firstly, a protein expressed by Salmonella, which is a phosphatase that acts on PtdIns(3,5)P2, was investigated as a potential tool for manipulating cellular PtdIns(3,5)P2 levels. Our results suggest that it is potentially a useful tool for this purpose and that expression of SopB perturbs endosome to lysosome trafficking. These findings provide further evidence for a role of PtdIns(3,5)P2 in endo-lysosomal trafficking.

Ubiquitination is a post-translational modification tht mediates sorting of integral membrane proteins to lysosomes for their degradation. ESCRTs (Endosomal Sorting Complex Required For Transport) bind and sequester ubiquitinated membrane proteins and direct them into multivesicular bodies (MVBs). ESCRTs themselves become covalently ubiquitinated, simply by virtue of non-covalently binding Ub. However, it is unclear whether this regulates a critical aspect of ESCRT function. In yeast, many MVB cargo proteins are ubiquitinated by the HECT-type Ub-ligase Rsp5, sometimes via the action of Rsp5 adaptor proteins. While many Rsp5 targets are modified by polyubiquitination, it remains unclear whether polyubiquitination is a necessary signal for their incorporation into MVBs. Despite years of research, these and related questions have been difficult to resolve because it is technically quite challenging to control the level of a given protein's ubiquitination. The aim of this research was to develop a novel technique, which can render proteins resistant to ubiquitination. The technique involved the fusion of the Ub-peptidase to a protein of interest via a flexible linker, essentially creating a "DUb module". The intent of this module would be to cleave any Ub form the target protein, essentially immunizing it from the effects of ubiquitination. This novel method was used in combination with several conventional methods to examine the role of ubiquitination within the endocytic pathway and in particular focus on the questions of what type of ubiquitin signal was sufficient for sorting into MVB vesicles and whether ubiquitination of ESCRTs was required for their sorting activity. We found that a single Ub was sufficient for membrane protein entry into MVBs in the absence of ESCRT ubiquitination.

The immune system is comprised of specific cell types that communicate and interact via a range of soluble and surface-bound molecules to defend the body against pathogens. Many gaps remain in our understanding of the subcellular trafficking pathways that regulate the diverse functions of the immune system. The central aim of this thesis was to investigate transport through the endocytic pathway, focussing in particular on the unique organelles and functions of this pathway in immune cells. Two subsets of immune cells were of particular interest in this thesis, macrophages and natural killer (NK) cells. These cell types both perform a range of functions that contribute to both innate and adaptive immunity. Another common thread between these cells is that they both perform functions involving specialised endocytic organelles and pathways. Macrophages utilise their endocytic pathways to perform several unique functions; phagocytosis, endocytosis and degradation of foreign proteins for presentation on MHC class II molecules, and signalling of Toll-like receptors from endosomes. Even secretion of cytokines such as tumour necrosis factor alpha (TNFα) by macrophages requires transport through an endosomal compartment, the recycling endosome, as recently discovered in this laboratory (Murray et al., 2005a). NK cells utilise specialised secretory lysosomes to deliver a lethal hit to carefully identified target cells, providing an alternative example of specialised endocytic trafficking in the immune system. Of the many protein families that regulate subcellular trafficking, the SNARE, Rab, Munc and exocyst proteins were focussed on during this thesis. The localisation and function of members of these families in the endocytic pathway were investigated. Novel results in macrophages concerned the role of Vti1b in endocytosis, a process with implications for MHC class II antigen presentation and TLR detection of endocytosed particles. Alteration of Vti1b protein levels in the cells significantly decreased uptake and degradation of endocytic cargo. A role for Rab11 and the recycling endosome in antigen presentation was also studied. MHC class II was detected in recycling endosomes, and overexpression of a mutant Rab11 protein altered the distribution of MHC class II, suggesting a role for Rab11 in subcellular trafficking during antigen presentation. Preliminary results also suggest a role for the exocyst protein Sec15 at the recycling endosome in macrophages, providing a new target for investigation of the regulation of TNFα secretion. The recycling endosome is emerging as a vital transport hub during cytokine secretion, phagocytosis and possibly other cellular functions in macrophages. This project also involved the unique opportunity to examine primary NK cells from patients with a number of genetic immunodeficiencies caused by mutations to trafficking proteins. The autosomal recessive immunodeficiencies Griscelli syndrome type 2 (GS2) and familial haemophagocytic lymphohistiocytosis types 3 (FHL3) and 4 (FHL4) are associated with loss-of-function of Rab27a, Munc13-4 and syntaxin 11 (Stx11), respectively. These diseases involve a loss of cytotoxic function by cytotoxic CD8+ T lymphocytes and NK cells, but the precise molecular role of these proteins in granule release is incompletely understood. In freshly isolated, resting NK cells from healthy subjects, PMA and ionomycin stimulation or conjugation to susceptible target cells induced colocalisation of endogenous Rab27a and Munc13-4 to perforin-containing granules. In Rab27a-deficient cells, which showed defective degranulation and cytotoxicity induced by signals for both natural and antibody-dependent cellular cytotoxicity, Munc13-4 failed to colocalise with perforin upon activation. Unexpectedly, Rab27a and Munc13-4 localisation to lytic granules was selectively induced by different receptor signals, demonstrating specificity for regulation of lytic granule maturation by target cell ligand expression. Recruitment of the SNARE protein Vti1b, which has not previously been associated with NK cell function or secretory lysosome release, to perforin granules was also discovered. Unexpectedly, Stx11 was not localised to perforin granules. These experiments have contributed to our understanding of the precise molecular roles of Munc13-4, Rab27a and Stx11 in NK cell granule release. Overall, this thesis presents novel and important results from studies of subcellular transport through the endocytic pathways of macrophages and NK cells. These results advance our understanding of several immune functions, and a number of human genetic immunodeficiencies. This new knowledge of the role of endocytic organelles and fusion machinery in these processes provides exciting targets for future research.

The Rab family of proteins are low molecular weight GTPases that have the ability to switch between GTP- (active) and GDP- (inactive) bound form, and in that sense act as molecular switches. Through distinct localization on various vesicles and organelles and by cycling through GTP/GDP bound forms, Rabs are able to recruit and activate numerous effector proteins, both spatially and temporally, and hence behave as key regulators of trafficking in both endocytic and biosynhtetic pathways. The Rab5 protein has been shown to regulate transport from plasma membrane to the early endosome as well as activate signaling pathways from the early endosome. This dissertation focused on understanding Rab5 activation via endocytosis of receptor tyrosine kinases (RTKs). First, tyrosine kinase activity of RTKs was linked to endosome fusion by demonstrating that tyrosine kinase inhibitors block endosome fusion and activation of Rab5, and a constitutively active form of Rab5 is able to rescue endosome fusion. However, depending on how much ligand is available at the cell surface, the receptor-ligand complexes can be internalized via a number of distinct pathways. Similarly, Rab5 was activated in a ligand-dependent concentration dependent manner via clathrin- and caveolin-mediated pathways, as well as a pathway independent of both. However, overexpression Rabex-5, a nucleotide exchange factor for Rab5, is able to rescue activation even when all of the pathways of EGF-receptor internalization were blocked. Next, the three naturally occurring splice variants of Rabex-5 selectively activated Rab5. Lastly, Rabex-5 inhibits differentiation of 3T3-L1 and PC12 cells through 1) degradation of signaling endosome via Rab5-dependent fusion with the early endosome, 2) and inhibition of signaling cascade via ubiquitination of Ras through the ZnF domain at the N-terminus of Rabex-5. In conclusion, these data shed light on complexity of the endosomal trafficking system where tyrosine kinase activity of the receptor is able to affect endosome fusion; how different endocytic pathways affect activation of one of the key regulators of early endocytic events; and how selective activation of Rab5 via Rabex-5 can control adipogenesis and neurogenesis.

The gram-negative endotoxin, lipopolysaccharide (LPS), has been extensively researched in high doses (10-200ng/ml) and is well-documented in the literature for its ability to result in devastating effects such as multi-organ failure, sepsis, and septic shock. In high doses, LPS signals through Toll-like-receptor 4 (TLR4) and triggers a cascade of events culminating in the release of pro- and anti-inflammatory cytokines and the activation of NF-κB. In contrast, super-low doses of LPS (1-100pg/ml) are able to trigger the persistent release of pro-inflammatory mediators while evading the compensatory activation of NF-κB. This mild yet persistent induction of inflammation may lie at the heart of numerous inflammatory diseases and disorders and warrants studies such as this to elucidate the novel mechanisms. In this study, we explored the novel mechanisms utilized by super-low dose LPS in cellular stress and low-grade inflammation. In the first study, the molecular mechanisms governing the role of super-low dose LPS on cellular stress and necroptosis were examined. We show that in the presence of super-low dose LPS (50pg/ml), the key regulators of mitochondrial fission and fusion, Drp1 and Mfn1 respectively, are inversely regulated. An increase in mitochondrial fragmentation and cell death which was not dependent on caspase activation was observed. In addition, super-low dose LPS was able to activate RIP3, a kinase responsible for inducing the inflammatory cell death, necroptosis. These mechanisms were regulated in an Interleukin-1 receptor-associated kinase 1 (IRAK-1) dependent manner. In the second study, the molecular mechanisms governing the role of super-low dose LPS on cellular stress and endosome/lysosome fusion were examined. In the presence of low-dose LPS (50pg/ml), endosomal-lysosomal fusion is inhibited and a loss of endosomal acidification required for the successful clearance of cellular debris and resolution of inflammation was observed. Additionally, super-low dose LPS induced the accumulation of p62 indicative of the suppression of autophagy. Tollip and Interleukin-1 receptor-associated kinase 3 (IRAK-M) appear to be critical regulators in this process. Collectively, these studies show that low-dose endotoxemia is capable of causing persistent cellular stress, not observed in the presence of high-dose LPS (10-200ng/ml), and that it promotes necroptotic cell death while suppressing mechanisms necessary for the resolution of inflammation such as endosome-lysosome fusion. This research reveals novel mechanisms utilized by low-dose endotoxemia which could aid future efforts to develop prevention and treatment for various debilitating inflammatory diseases. / Ph. D.

Lysosomes are acidic, hydrolase-rich vesicles capable of degrading most biological macromolecules. During the past several decades, much has been learned about different aspects of lysosome biogenesis. The selective phosphorylation of mannose residues on lysosomal enzymes, in conjunction with specific receptors for the mannose-6-phosphate recognition marker, has been found to be largely responsible for the targeting of newly synthesized lysosomal enzymes to lyzosomes. It is known that lysosomes receive input from both the endocytotic and biosynthetic pathways. Nevertheless the exact molecular mechanisms responsible for sorting of the biosynthetic imput involved in the lysosome biogenesis is still a matter of debate. Because osteoclast precursors do not secrete their lysosomal enzymes and osteoclasts do, the observation of modifications occuring during osteoclastogenesis is a good model to observe mechanisms responsible for lysosomal enzymes traffic. Osteoclasts are bone-degrading cells. To perform this specific task they have to reorganise the sorting of their lysosomal enzymes to be able to target them toward the bone surface in mature cells. Since few years, the differentiation of osteoclasts in vitro did help to study these cells. Osteoclast morphology has been therefore already well studied, and the nature of their specific membrane domains is now established. Sensing the proximity of a bone-like surface the cell reorganises its cytoskeleton, and creates specific membrane domains: an actin-rich ring-like zone (named actin ring) surrounded by highly ruffled membrane (named the ruffled border) where enzymes are secreted, while subsequent bone degradation products are endocytosed. Endocytosed material is then transported through the cell inside transcytotic vesicles and released at the top of the cell in an area named the functional secretory domain. Several molecular machineries are thought to control these different phenomena. The main purpose of this thesis was to identify the major regulators of lysosomal enzymes secretion and therefore to identify the molecular switches responsible for such a membrane traffic re-organisation.

No description available.

La maladie d'Alzheimer (MA) est la maladie neurodégénérative la plus répandue à travers le monde et reste actuellement incurable. Le peptide beta amyloïde (Abeta), composant principal des plaques séniles retrouvées dans le cerveau des patients, joue un rôle majeur dans le développement de la MA, d'où l'importance de contrôler sa production et/ou son élimination. Dans cette optique, nous travaillons sur des molécules nommées métalloprotéases matricielles (MMPs). Bien qu'ayant été impliquées à la fois dans de nombreux processus physiologiques et pathologiques dans système nerveux, leur rôle dans la MA reste encore relativement inexplorée. Nous avons utilisé comme modèle d'étude des souris qui développent les symptômes de la MA (déclin cognitif, mort des neurones). Nous montrons que deux MMPs, MMP-2 et MT1-MMP, augmentent leurs niveaux d'expression avec le vieillissement de l'animal et donc avec l'aggravation de la pathologie. Ceci a lieu dans l'hippocampe, une région du cerveau qui est particulièrement sensible car elle est impliquée dans l'apprentissage et la mémoire. Par la suite nous avons utilisé des cellules HEKswe qui produisent beaucoup d'Abeta et miment d'une certaine manière ce qui se passe dans le cerveau de la souris, afin de mieux appréhender la signification des augmentations de ces MMPs. Nous montrons que la surexpression de MT1-MMP dans ces cellules favorise la formation d'Abeta, alors que MMP-2 l'empêche. Ces résultats montrent pour la première fois une dualité fonctionnelle au sein de la famille des MMPs, et plus important, révèlent une nouvelle molécule amyloïdogénique (MT1-MMP) qui pourrait devenir à terme une cible thérapeutique. / We investigate the role of matrix metalloproteinases in the metabolism of beta amyloid peptide (Abeta) and its amyloid precursor protein (APP) in Alzheimer's disease (AD). Our results in the 5xFAD mouse model of AD indicate a cell-type and age-dependent upregulation of MMP-2 -and MT1-MMP active forms. This is concomitant with the increase of toxic forms of Abeta, but also of cytotoxic C99, a membrane fragment of APP generated by beta-secretase and that gives rise to Abeta after gamma-secretase cleavage. We show in HEK cells overproducing Abeta that while MT1-MMP interacts with APP and boosts C99 and Abeta production, MMP-2 does not interact with APP and degrades Abeta. These results uncover a MMP-specific regulatory crosstalk with amyloid and also MT1-MMP as a new pro-amyloidogenic proteinase. We want now to gain further insight into the mechanisms that support MT1-MMP effects, namely the possible modulation by MT1-MMP of beta- and gamma-secretase activities and/or APP trafficking.

Lysosomes are acidic, hydrolase-rich vesicles capable of degrading most biological macromolecules. During the past several decades, much has been learned about different aspects of lysosome biogenesis. The selective phosphorylation of mannose residues on lysosomal enzymes, in conjunction with specific receptors for the mannose-6-phosphate recognition marker, has been found to be largely responsible for the targeting of newly synthesized lysosomal enzymes to lyzosomes. It is known that lysosomes receive input from both the endocytotic and biosynthetic pathways. Nevertheless the exact molecular mechanisms responsible for sorting of the biosynthetic imput involved in the lysosome biogenesis is still a matter of debate. Because osteoclast precursors do not secrete their lysosomal enzymes and osteoclasts do, the observation of modifications occuring during osteoclastogenesis is a good model to observe mechanisms responsible for lysosomal enzymes traffic. Osteoclasts are bone-degrading cells. To perform this specific task they have to reorganise the sorting of their lysosomal enzymes to be able to target them toward the bone surface in mature cells. Since few years, the differentiation of osteoclasts in vitro did help to study these cells. Osteoclast morphology has been therefore already well studied, and the nature of their specific membrane domains is now established. Sensing the proximity of a bone-like surface the cell reorganises its cytoskeleton, and creates specific membrane domains: an actin-rich ring-like zone (named actin ring) surrounded by highly ruffled membrane (named the ruffled border) where enzymes are secreted, while subsequent bone degradation products are endocytosed. Endocytosed material is then transported through the cell inside transcytotic vesicles and released at the top of the cell in an area named the functional secretory domain. Several molecular machineries are thought to control these different phenomena. The main purpose of this thesis was to identify the major regulators of lysosomal enzymes secretion and therefore to identify the molecular switches responsible for such a membrane traffic re-organisation.

La maladie d'Alzheimer (MA) est une maladie neurodégénérative progressive et irréversible. Une étape précoce de la MA est la relâche neuronale excessive du peptide amyloïde-[bêta] (A[bêta]), qui s'accumule dans le cerveau, s'assemble et se dépose sous forme de plaques A[bêta] insolubles et neurotoxiques. L'A[bêta] est produit suite au clivage amyloïdogénique de la protéine APP, effectué par les sécrétases [bêta] et [gamma] au niveau des endosomes. Il est bien connu que le trafic intracellulaire de l'APP affecte son clivage. L'étude du trafic intracellulaire de cette protéine est donc cruciale pour comprendre ce qui régit la production d'A[bêta]. Certains membres de la famille des récepteurs de lipoprotéines de faibles densités (LDLR), dont SorLA/LR11, interagissent avec l'APP et modulent son clivage en régulant son trafic et/ou en s'associant avec les sécrétases. LRP10, un nouveau membre peu connu des LDLR, trafique entre le Golgi et les endosomes, tout comme SorLA/LR11. Conséquemment, nous avons émis l'hypothèse selon laquelle LRP10 serait un nouveau récepteur de la protéine APP, impliqué dans la régulation du trafic et du clivage de cette dernière ainsi que dans la relâche d'A[bêta]. Nos résultats démontrent que LRP10 et la protéine APP colocalisent au TGN (trans-Golgi network ) et interagissent de façon directe.La surexpression stable de LRP10 dans les cellules de neuroblastome humain SH-SY5Y, provoque une accumulation de la forme mature de l'APP, ainsi qu'une diminution de son clivage et de la production d'A[bêta].La déplétion de LRP10, par la technique d'ARN interférant, provoque l'augmentation de la production d'A[bêta]. De plus, l'expression d'un mutant de LRP10, redistribué aux endosomes précoces, induit la redistribution intracellulaire de l'APP au niveau de ces mêmes endosomes dans les cellules HeLa et SH-SY5Y, tel qu'observé en microscopie confocale.La surexpression stable du mutant de LRP10 dans les SH-SY5Y a aussi démontré une augmentation du clivage amyloïdogénique de l'APP normalement effectué aux endosomes et donc une augmentation de la production d'A[bêta]. Enfin, la comparaison des niveaux d'expression protéique de LRP10 retrouvés dans le cortex frontal et l'hippocampe de cerveaux de patients âgés sains ou atteints de la MA, révèle que l'expression de LRP10 est réduit dans le cerveau des patients atteints de la MA. En conclusion, LRP10 est un nouveau récepteur de l'APP participant à son triage entre le TGN et les endosomes, protégeant ainsi l'APP du clivage amyloïdogénique et de l'accumulation d'A[bêta]. Ainsi, la réduction de l'expression de LRP10 dans le cerveau pourrait augmenter la production de l'A[bêta] et représenter un facteur de risque dans la MA.