Les interactions multivalentes sucre/protéine sont impliquées dans de nombreux processus biologiques tels que l'adhésion hôte-pathogène, la communication cellulaire ou les réponses immunitaires. La conception de glycoconjugués capables de présenter des motifs osidiques en cluster est essentielle, non seulement pour étudier ces phénomènes de reconnaissance complexes mais aussi pour développer des agents biologiquement actifs. Dans ce contexte, l'objectif de ma thèse a été de synthétiser de nouveaux glycoclusters et d'évaluer leurs propriétés biologiques. Ces glycoclusters ont été obtenus en conjuguant des sucres sur des cyclopeptides par des méthodes chimiosélectives (cycloaddition de Huisgen et ligation oxime). Des composés tetravalents, hexavalents et hexadécavalents d'architectures et de compositions variables ont été synthétisés, entièrement caractérisés puis évalués au laboratoire ou dans le cadre de collaborations avec différentes cibles. Un glycocluster hexadécavalent fucosylé a ainsi pu être identifié comme puissant inhibiteur de l'interaction de la lectine PA-IIL de la bactérie Pseudomonas aeruginosa à une concentration subnanomolaire. Une autre famille de composés associant des sucres et des peptides a montré des propriétés immunologiques uniques, notamment pour activer les cellules NK contre des cellules cancéreuses sans déclencher de phénomène d'autoapoptose. Mots clés : chimie des sucres, chimie des peptides, glycoclusters, ligations chimiosélectives, interactions sucre-protéine, lectine, cellule NK. / Multivalent carbohydrates-protein interactions are involved in a large variety of biological processes such as host-pathogen adhesions, cell communication or immune responses. The design of glycoconjugates displaying multiple copies of sugars as cluster is essential, not only to study these complex recognition processes but also to develop bioactive agents. In this context, the objective of my PhD was to synthesize new glycoclusters and evaluate their biological properties. These glycoclusters were obtained by conjugating carbohydrate moieties and peptides using chemoselective ligations (Huisgen cycloaddition and oxime ligation). Tetravalent, hexavalent and hexadecavalent glycoclusters with variable structures and compositions were synthesized, fully characterized and biologically evaluated with different targets in our laboratory or in collaborations. A hexadecavalent fucosylated glycocluster was thus identified as a strong inhibitor of the lectin PA-IIL from Pseudomonas aeruginosa at subnanomolar concentration. Another type of molecule containing carbohydrate and peptides has showed unique immunological properties, in particular for the activation of NK cells against tumors without inducing apoptotic effect Key words: carbohydrate chemistry, peptide chemistry, chimioselective ligations, glycoclusters, carbohydrate-protein interactions, lectin, NK cell.

La dystrophine est une protéine filamenteuse qui contribue à la structuration des cellules musculaires en créant un lien entre le cytosquelette et le sarcolemme. Avec l’actine du cytosquelette et les lipides membranaires, la dystrophine représente l’un des éléments d’un complexe macromoléculaire, localisé sous la membrane plasmique, dont le rôle est la prévention des dommages qui pourraient être induits à force de contractions-relâchements répétés. De tels dommages, notamment des ruptures du sarcolemme, sont observés chez des personnes atteintes de myopathies de Duchenne (DMD) et de Becker (BMD), des maladies causées par des mutations qui altèrent l’expression ou la fonction de la dystrophine. Ces myopathies sont actuellement incurables et une connaissance approfondie de la relation structure-fonction de la dystrophine et de ses interactions avec ses partenaires s’avère absolument nécessaire à la mise au point de nouvelles stratégies de thérapies géniques. Cette protéine se compose de quatre domaines fonctionnels, dont un domaine central filamentaire, constitué de 24 répétitions successives d’un même motif structural, un faisceau de trois hélices alpha ou « coiled-coil ». Or, il a été récemment montré que la structure de ce domaine central n’est pas strictement linéaire et que certaines régions inter-répétitions (linker) forment des coudes, délimitant ainsi des sous-domaines d’interaction spécifiques. Cette thèse a pour objectif de comprendre l’origine de cette diversité de conformations inter-répétition dans un domaine structuralement homogène, et d’explorer comment elle permet à certaines régions de se différencier afin d’interagir avec l’actine et/ou les lipides membranaires. / Dystrophin is a filamentous protein involved in muscular cells structure which links the cytoskeleton to the sarcolemma. Together with cytoskeletal actin and membrane lipids, dystrophin is a part of a macromolecular complex, located under the sarcolemma, which prevents damages induced during repeated muscle contractions and relaxations. Such damages, including sarcolemma disruption, are found in people with Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD), diseases caused by mutations altering dystrophin expression or function. There is currently no treatment to cure these myopathies, and a deep understanding of the structure-function of the dystrophin and its interactions with its partners is necessary to the development of gene therapy strategies. Structuraly, this protein is composed of four functionnal domains, including a long filamentous central domain, composed of 24 successive coiled-coil repeats. It was recently showed that the central domain is not rod shaped and some inter-repeats regions (linker) are kinked, delimiting specific interaction sub-domains. This thesis aims to bring knowledge about the basis of the conformationnal diversity of linkers in a structuraly homogenous domain in human dystrophin. We explore how dystrophin delimits some regions that interact with f-actin and/or membrane lipids.

Septinas abrangem uma família conservada de proteínas que ligam e hidrolisam GTP e formam heterofilamentos, anéis e redes para realizar as suas funções. Apresentam três domínios estruturais: o domínio N-terminal contendo uma sequência polibásica (para ligar membranas), o domínio de ligação ao nucleotídeo (G) e o domínio C-terminal que inclui uma sequência predita de formar um coiled-coil. Em humanos, as 13 septinas são classificadas em quatro grupos (I, II, III e IV) baseadas nas sequências de aminoácidos. O único filamento caracterizado estruturalmente, até hoje, é o formado por SEPT2-SEPT6-SEPT7, mostrando que as subunidades interagem através de duas interfaces (chamadas G e NC). Os determinantes estruturais da montagem correta do filamento são pouco conhecidos, sendo o estudo limitado pela complexidade em purificar e cristalizar complexos triméricos ou tetraméricos. Uma abordagem alternativa é estudar interfaces individuais de um filamento (G e/ou NC) por separado. Assim, o presente projeto objetivou estudar, utilizando uma abordagem biofísica e estrutural, a interface G formada por SEPT5 e SEPT8 para elucidar os fatores importantes em determinar a sua especificidade. Os domínios GTPase de SEPT5 e SEPT8 foram clonadas em vetor de expressão bicistrônico pET-Duet, co-expressas e co-purificadas. Estudos de análise do estado oligomérico e homogeneidade foram conduzidos utilizando cromatografia de exclusão molecular, espalhamento dinâmico de luz e ultracentrifugação analítica, revelando um complexo dimérico e monodisperso. O complexo apresenta uma mistura aproximadamente equimolar de nucleotídeos (GTP e GDP) ligados enquanto SEPT8(G) sozinha é incapaz de ligar qualquer um dos dois. Além disto o complexo apresenta uma termoestabilidade maior que SEPT8(G), verificado por um aumento em Tm de 5°C. Com o intuito de observar os determinantes estruturais da especificidade, ensaios de cristalização foram conduzidos e assim, cristais do complexo SEPT5-SEPT8(G) que difrataram apenas a muito baixa resolução foram obtidos. Na ausência de uma estrutura cristalográfica, modelagem por homologia foi realizada para analisar as interfaces G entre diferentes combinações de septinas. Identificamos uma interação entre aminoácidos característicos (aminoácidos únicos para cada grupo de septinas) para o complexo formado entre membros do grupo III, (incluindo SEPT5) e membros do grupo II, (incluindo SEPT8). Esta interação entre Phe131 (grupo III) e Thr19 (grupo II) pode explicar a especificidade na formação de uma interface G entre septinas destes grupos durante a formação do filamento e além disso, a importância da presença do GTP ligado ao septina do grupo II. Com isto, propomos pela primeira vez uma explicação plausível da relevância da perda de atividade catalítica das septinas deste grupo, um fato inexplicado até o momento. Mutação dos resíduos identificados levou a uma mudança no seu perfil de eluição do complexo durante purificação por exclusão molecular indicando alterações na formação do complexo mutante. / Septins are a conserved family of proteins that bind and hydrolyze GTP and form heterofilaments, rings and networks in order to carry out their functions. They have three structural domains: an N-terminal domain containing a polybasic sequence (for membrane binding), a nucleotide-binding (G) domain and a C-terminal domain including a sequence predicted to form a coiled-coil. In humans, 13 septins have been classified into four groups (I, II, III and IV) based on their amino acid sequences. The only structurally characterized filament described to date is formed by SEPT2-SEPT6-SEPT7, which reveals that the subunits interact through two different interfaces (G and NC). The structural determinants of correct filament assembly are poorly known, and this is limited by the complexity of purifying and crystallizing trimeric or tetrameric complexes. An alternative approach is to study a single filament interface (G or NC) on its own. Here, we aimed to study, using biophysical and structural approaches, the G interface formed between SEPT5 and SEPT8 to elucidate the factors relevant to determining its specificity. The GTPase domain of SEPT5 and SEPT8, were cloned into the bicistronic expression vector pET-Duet, co-expressed and co-purified. Studies to determine the oligomeric state and homogeneity of the complex were conducted using size exclusion chromatography, dynamic light scattering and analytical ultracentrifugation, revealing a monodisperse dimer for SEPT5-SEPT8(G). The complex elutes with an approximately equimolar mixture of bound nucleotides (GTP and GDP) whereas SEPT8(G) alone is shown to be unable to bind either. Furthermore, the complex has a greater thermostability than SEPT8(G), demonstrated by an increase of 5°C in Tm. In order to determine the structural determinants of specificity, crystallization trials were conducted and crystals of the SEPT5-SEPT8(G) complex were obtained, but these diffracted to only very low resolution. In the absence of a crystal structure, homology modeling was performed to analyze the potential G interfaces between different septin combinations. An interaction between characteristic amino acids (those which are unique to given septin group) was identified for the complex formed between group III septins (including SEPT5) and group II septins (including SEPT8). This interaction, between Phe131 (group II) and Thr19 (group III) may explain the specificity in the formation of a G interface between septins of these groups during filament formation and furthermore the importance of GTP bound to the group II septin. These observations allow us to propose for the first time a plausible explanation for relevance of the loss of catalytic activity by this septin group, an unexplained fact up until now. Mutation of the identified residues resulted in a change in the elution profile of the complex from the size exclusion column suggesting structural alterations in the mutants.

Septinas abrangem uma família conservada de proteínas que ligam e hidrolisam GTP e formam heterofilamentos, anéis e redes para realizar as suas funções. Apresentam três domínios estruturais: o domínio N-terminal contendo uma sequência polibásica (para ligar membranas), o domínio de ligação ao nucleotídeo (G) e o domínio C-terminal que inclui uma sequência predita de formar um coiled-coil. Em humanos, as 13 septinas são classificadas em quatro grupos (I, II, III e IV) baseadas nas sequências de aminoácidos. O único filamento caracterizado estruturalmente, até hoje, é o formado por SEPT2-SEPT6-SEPT7, mostrando que as subunidades interagem através de duas interfaces (chamadas G e NC). Os determinantes estruturais da montagem correta do filamento são pouco conhecidos, sendo o estudo limitado pela complexidade em purificar e cristalizar complexos triméricos ou tetraméricos. Uma abordagem alternativa é estudar interfaces individuais de um filamento (G e/ou NC) por separado. Assim, o presente projeto objetivou estudar, utilizando uma abordagem biofísica e estrutural, a interface G formada por SEPT5 e SEPT8 para elucidar os fatores importantes em determinar a sua especificidade. Os domínios GTPase de SEPT5 e SEPT8 foram clonadas em vetor de expressão bicistrônico pET-Duet, co-expressas e co-purificadas. Estudos de análise do estado oligomérico e homogeneidade foram conduzidos utilizando cromatografia de exclusão molecular, espalhamento dinâmico de luz e ultracentrifugação analítica, revelando um complexo dimérico e monodisperso. O complexo apresenta uma mistura aproximadamente equimolar de nucleotídeos (GTP e GDP) ligados enquanto SEPT8(G) sozinha é incapaz de ligar qualquer um dos dois. Além disto o complexo apresenta uma termoestabilidade maior que SEPT8(G), verificado por um aumento em Tm de 5°C. Com o intuito de observar os determinantes estruturais da especificidade, ensaios de cristalização foram conduzidos e assim, cristais do complexo SEPT5-SEPT8(G) que difrataram apenas a muito baixa resolução foram obtidos. Na ausência de uma estrutura cristalográfica, modelagem por homologia foi realizada para analisar as interfaces G entre diferentes combinações de septinas. Identificamos uma interação entre aminoácidos característicos (aminoácidos únicos para cada grupo de septinas) para o complexo formado entre membros do grupo III, (incluindo SEPT5) e membros do grupo II, (incluindo SEPT8). Esta interação entre Phe131 (grupo III) e Thr19 (grupo II) pode explicar a especificidade na formação de uma interface G entre septinas destes grupos durante a formação do filamento e além disso, a importância da presença do GTP ligado ao septina do grupo II. Com isto, propomos pela primeira vez uma explicação plausível da relevância da perda de atividade catalítica das septinas deste grupo, um fato inexplicado até o momento. Mutação dos resíduos identificados levou a uma mudança no seu perfil de eluição do complexo durante purificação por exclusão molecular indicando alterações na formação do complexo mutante. / Septins are a conserved family of proteins that bind and hydrolyze GTP and form heterofilaments, rings and networks in order to carry out their functions. They have three structural domains: an N-terminal domain containing a polybasic sequence (for membrane binding), a nucleotide-binding (G) domain and a C-terminal domain including a sequence predicted to form a coiled-coil. In humans, 13 septins have been classified into four groups (I, II, III and IV) based on their amino acid sequences. The only structurally characterized filament described to date is formed by SEPT2-SEPT6-SEPT7, which reveals that the subunits interact through two different interfaces (G and NC). The structural determinants of correct filament assembly are poorly known, and this is limited by the complexity of purifying and crystallizing trimeric or tetrameric complexes. An alternative approach is to study a single filament interface (G or NC) on its own. Here, we aimed to study, using biophysical and structural approaches, the G interface formed between SEPT5 and SEPT8 to elucidate the factors relevant to determining its specificity. The GTPase domain of SEPT5 and SEPT8, were cloned into the bicistronic expression vector pET-Duet, co-expressed and co-purified. Studies to determine the oligomeric state and homogeneity of the complex were conducted using size exclusion chromatography, dynamic light scattering and analytical ultracentrifugation, revealing a monodisperse dimer for SEPT5-SEPT8(G). The complex elutes with an approximately equimolar mixture of bound nucleotides (GTP and GDP) whereas SEPT8(G) alone is shown to be unable to bind either. Furthermore, the complex has a greater thermostability than SEPT8(G), demonstrated by an increase of 5°C in Tm. In order to determine the structural determinants of specificity, crystallization trials were conducted and crystals of the SEPT5-SEPT8(G) complex were obtained, but these diffracted to only very low resolution. In the absence of a crystal structure, homology modeling was performed to analyze the potential G interfaces between different septin combinations. An interaction between characteristic amino acids (those which are unique to given septin group) was identified for the complex formed between group III septins (including SEPT5) and group II septins (including SEPT8). This interaction, between Phe131 (group II) and Thr19 (group III) may explain the specificity in the formation of a G interface between septins of these groups during filament formation and furthermore the importance of GTP bound to the group II septin. These observations allow us to propose for the first time a plausible explanation for relevance of the loss of catalytic activity by this septin group, an unexplained fact up until now. Mutation of the identified residues resulted in a change in the elution profile of the complex from the size exclusion column suggesting structural alterations in the mutants.

Prion diseases are fatal neurodegenerative disorders of humans and animals. The prion hypothesis states that PrPSc, a misfolded conformational isoform of the cellular prion protein (PrPC), is the sole component of the infectious particle. Many open questions exist in prion biology including the cellular role of PrPC, the potential involvement of auxiliary factors in prion replication, and the mechanism of PrPSc-induced toxicity in prion disease. The identification of novel prion-like proteins and authentic in vivo prion protein-interacting proteins would certainly assist the process of demystifying these unsolved mysteries. Accordingly, two newly-identified proteins with potential relevance to prion protein biology, Shadoo and DPPX, were selected for biochemical and functional characterization. Shadoo, a hypothetical prion-like protein, is revealed as being a glycoprotein which possesses many overlapping properties with PrPC including neuronal expression, C1-like endoproteolytic processing, and the ability to protect against apoptotic stimuli in cerebellar neurons. Shadoo loosely resembles the disordered N-terminal domain of PrPC and consistent with this notion, Shadoo appears to lack a well-defined structure. Remarkably, Shadoo levels in the brains of mice with clinical prion disease are significantly decreased suggesting that Shadoo may be inherently linked to prion replication or prion disease pathogenesis. These experiments define Shadoo as the third member of the prion protein family and, because of its functional similarities to PrPC, Shadoo may be a useful tool for deciphering the in vivo function of PrPC. DPPX, a neuronal type II transmembrane protein, is demonstrated to be the first protein capable of interacting with all three members of the prion protein family (PrPC, Doppel, and Shadoo) in vivo. Complex formation between prion proteins and DPPX appears to be mediated by multiple binding sites. When coupled with high levels of DPPX expression in cerebellar granular neurons, DPPX is a strong candidate for mediating phenotypic interactions between prion proteins in cerebellar cells. Thus, Shadoo and DPPX comprise two new entry points for studying prion proteins. Further investigation of the roles of Shadoo and DPPX in both the cell biology of prion proteins and prion disease may yield important clues to these enigmatic topics.

Src homology 3 (SH3) domains are one family of the peptide recognition modules (PRMs), which bind peptides rich in proline or positively charged residues in the target proteins, and play important assembly or regulatory functions in dynamic eukaryotic cellular processes, especially in signal transduction and endocytosis. SH3 domains are conserved from yeast to human, and improper SH3 domain mediated protein-protein interaction (PPI) leads to defects in cellular function and may even result in disease states. Since commonly used large-scale PPI mapping strategies employed full-length proteins or random protein fragments as screening probes and did not identify the particular PPIs mediated by the SH3 domains, I employed a combined experimental and computational strategy to address this problem. I used yeast two-hybrid (Y2H) as my major experimental tool, as well as individual SH3 domains as baits, to map SH3 domain mediated PPI networks, “SH3 domain interactomes”. One of my important contributions has been the improvement for Y2H technology. First, I generated a pair of Y2H host strains that improved the efficiency of high-throughput Y2H screening and validated their usage. These strains were employed in my own research and also were adopted by other researchers in their large-scale PPI network mapping projects. Second, in collaboration with Nicolas Thierry-Mieg, I developed a novel smart-pooling method, Shifted Transversal Design (STD) pooling, and validated its application in large-scale Y2H. STD pooling was proven to be superior among currently available methods for obtaining large-scale PPI maps with higher coverage, high sensitivity and high specificity. I mapped the SH3 domain interactomes for both budding yeast Saccharomyces cerevisiae and nematode worm Caenorhabditis elegans, which contain 27 and 84 SH3 domains, respectively. Comparison of these two SH3 interactomes revealed that the role of the SH3 domain is conserved at a functional but not a structural level, playing a major role in the assembly of an endocytosis network from yeast to worm. Moreover, the worm SH3 domains are additionally involved in metazoan-specific functions such as neurogenesis and vulval development. These results provide valuable insights for our understanding of two important evolutionary processes from single cellular eukaryotes to animals: the functional expansion of the SH3 domains into new cellular modules, as well as the conservation and evolution of some cellular modules at the molecular level, particularly the endocytosis module.

The human ether-à-go-go related gene (hERG) encodes the pore-forming voltage-gated K+ channel that is essential for cardiac repolarization. Dr. Tsushima’s laboratory has previously characterized the endogenous expression of SNARE proteins in the mammalian heart, and the interaction of the SNARE protein syntaxin 1A (STX1A) with several cardiac ion channels. Here, we utilize a multi-disciplinary approach to describe the inhibitory effect of STX1A on hERG channel function. STX1A impairs hERG channel maturation and trafficking to the plasma membrane and induces a hyperpolarizing shift in the voltage-sensitivity of steady-state inactivation. We identify the residues involved in this protein-protein interaction through the use of hERG truncation mutations. We also describe the pharmacological and temperature-mediated rescue of hERG channel trafficking in the presence of STX1A. The regulation of cardiac ion channels by SNARE proteins represents a novel biological mechanism that may have universally intrinsic implications for normal and diseased heart function.

G-protein coupled receptors (GPCRs) have been shown to act as part of GPCR associated protein complexes (GAPCs) which are required to appropriately transduce downstream signaling pathways leading to specific cellular actions. I hypothesize that there are distinct molecular effectors that couple to the glucagon superfamily of B-class GPCRs (glucagon, GLP-1, GLP-2, GIP receptors) to effect the myriad of reported actions in numerous target cells including regulation of insulin secretion, intestinal growth and appetite suppression. GLP-1R, GIPR, GLP-2R and GCGR were screened using a newly developed membrane-based split-ubiquitin yeast two-hybrid (MYTH) system to reveal 181 novel candidate protein interactors associated with signal transduction, transport, metabolism and cell survival. Each candidate was validated using yeast two-hybrid prey retransformation tests and by co-purification to confirm coupling to each receptors. The present work is the first demonstration of a split-ubiquitin interaction screen using in situ membrane expressed GPCRs of the secretin-like B class.

The human ether-à-go-go related gene (hERG) encodes the pore-forming voltage-gated K+ channel that is essential for cardiac repolarization. Dr. Tsushima’s laboratory has previously characterized the endogenous expression of SNARE proteins in the mammalian heart, and the interaction of the SNARE protein syntaxin 1A (STX1A) with several cardiac ion channels. Here, we utilize a multi-disciplinary approach to describe the inhibitory effect of STX1A on hERG channel function. STX1A impairs hERG channel maturation and trafficking to the plasma membrane and induces a hyperpolarizing shift in the voltage-sensitivity of steady-state inactivation. We identify the residues involved in this protein-protein interaction through the use of hERG truncation mutations. We also describe the pharmacological and temperature-mediated rescue of hERG channel trafficking in the presence of STX1A. The regulation of cardiac ion channels by SNARE proteins represents a novel biological mechanism that may have universally intrinsic implications for normal and diseased heart function.

G-protein coupled receptors (GPCRs) have been shown to act as part of GPCR associated protein complexes (GAPCs) which are required to appropriately transduce downstream signaling pathways leading to specific cellular actions. I hypothesize that there are distinct molecular effectors that couple to the glucagon superfamily of B-class GPCRs (glucagon, GLP-1, GLP-2, GIP receptors) to effect the myriad of reported actions in numerous target cells including regulation of insulin secretion, intestinal growth and appetite suppression. GLP-1R, GIPR, GLP-2R and GCGR were screened using a newly developed membrane-based split-ubiquitin yeast two-hybrid (MYTH) system to reveal 181 novel candidate protein interactors associated with signal transduction, transport, metabolism and cell survival. Each candidate was validated using yeast two-hybrid prey retransformation tests and by co-purification to confirm coupling to each receptors. The present work is the first demonstration of a split-ubiquitin interaction screen using in situ membrane expressed GPCRs of the secretin-like B class.