For a long time, the scientific community underestimated the value of carbohydrates and the approach of most scientists to the complex world of glycans was apprehensive. The scenario, however, has changed today. With the development of new research tools and methodologies the study of carbohydrates and glycoconjugates has progressed rapidly, increasing our understanding of these molecules. Carbohydrates are most abundant amongst biological polymers in nature and vital for life processes. In their simplest form, they serve as a primary source of energy to most living organisms. In generalis, they exist as complex structures (glycans), and as conjugates of protein (glycoproteins, proteoglycans), lipids (glycolipids) and nucleosides (UDP-Glucose). Defined in the broadest sense, the study of glycans in all their forms and their interacting partners is termed “Glycobiology”. Glycans are ubiquitously found in nature decorating cells of almost all types with a “sugar coat”. They are also present within the cytoplasm, as well as in the extra-cellular matrix. They have key roles in a broad range of biological processes, including signal transduction, cell development and immune responses. All living organisms have evolved to express proteins that recognize discrete glycans and mediate specific physiological or pathological processes. One major class of such proteins is “Lectins”. Found in all forms of life, they are characterized by their ability to recognize carbohydrates. They are proteins of non-immune origin that bind glycans reversibly with a high degree of stereo-specificity in a non-catalytic manner. It must be emphasized that they are a different class from glycan-specific antibodies. Lectins were first discovered in plants and a large amount of work has been carried out on plant lectins to decipher their structural organization, mode of interaction with substrate and as models to study protein stability and folding. Study on animal and microbial lectins, on the other hand, gathered momentum only recently. In spite of this, more is known about their function in animals and micro-organisms rather than in plants. Lectin-glycan binding is implicated in several important biological processes such as protein folding, trafficking, host-pathogen interactions, immune cell responses and in malignancy and metastasis. Most lectins have one or more carbohydrate recognition domains (CRDs) which often share either 3-D structural features or amino acid sequence. New members of a family can be identified using either sequence or structural homology. Interestingly, it turns out that several plant and microbial lectins have structural or sequential similarity with animal lectins , revealing that these CRDs are evolutionarily related. This thesis, entitled “Structural, Biophysical and Biochemical Studies on Mannose-specific Lectins”, focuses on three lectins, Banana lectin (Banlec), Calreticulin (CRT) and Peptide-N-Glycanase (PNGase). Although all three lectins have distinct biological functions, they share a common ligand specificity at the monosaccharide level i.e. mannose. This thesis, besides characterizing these lectins, studies in detail, the difference in the mode of interaction with their ligands. Chapter 1 is a general introduction on lectins, glycan-lectin interactions and the various techniques that are employed to characterize these interactions. Several principles have emerged about the nature of glycan–lectin interactions. It has been observed that the binding sites for low molecular weight glycans are of relatively low affinity (Kd values in the high micromolar to low millimolar range). Selectivity is mostly achieved via a combination of hydrogen bonds and by van der Waals packing of the hydrophobic faces of monosaccharide rings against aromatic amino acid side chains. Further selectivity and enhanced affinity can be achieved by additional contacts between the glycan and the protein. It is notable that the actual region of contact between the saccharide and the polypeptide typically involves only one to three monosaccharide residues. As a consequence of all of the above, these lectin-binding sites tend to be of relatively low affinity, although they can exhibit high specificity. It is intriguing to observe that such low-affinity sites have the ability to mediate biologically relevant interactions. There are many different ways to study binding of glycans to proteins, and each approach has its advantages and disadvantages in terms of thermodynamic rigor, amounts of protein and glycan needed, and the speed of analysis. In examining these interactions, two broad categories of techniques are applied: (1) kinetic and near-equilibrium methods, such as titration calorimetry; and (2) non-equilibrium methods such as glycan microarray screening and ELISA-based approaches. Two of the most widely used biophysical approaches for examining glycan-lectin interactions at the molecular level are X-ray crystallography and nuclear magnetic resonance (NMR). However, as small molecules often co-crystallize with a lectin better than large molecules, a lot of our knowledge about glycan–lectin interactions at the atomic level is based on co-crystals of lectins with unnatural ligands. Thus, a great challenge exists in attempting to understand glycan–lectin interactions in the context of natural glycans present as glycoproteins, glycolipids, or proteoglycans. Chapter 2 introduces Banana lectin and describes the stability studies carried out. The unfolding pathway of Banlec was determined using GdnCl induced denaturation. Analysis of isothermal denaturation provided information on its conformational stability and the high values of ΔG of unfolding at various temperatures indicated the strength of inter-subunit interactions. It was found that Banlec is a very stable protein and denatures only at high chaotrope concentrations. The basis of the stability may be attributed to strong hydrogen bonds at the dimeric interface along with the presence of water bridges. This is a very unique example in proteins where subunit association is not a consequence of the predominance of hydrophobic interactions. High temperature molecular dynamics simulations have been utilized to monitor and understand early stages of thermally induced unfolding of Banlec. The present study investigates the behavior of the dimeric protein at four different temperatures. The process of unfolding was monitored by monitoring the radius of gyration, the rms deviation of each residue, change in relative solvent accessibility and the pattern of inter- and intra-subunit interactions. The overall study demonstrates that the Banlec dimer is a highly stable structure, the stability in most part contributed by interfacial interactions. The pattern of hydrogen bonding within the subunits and at the interface across different stages has been analyzed and has provided the rationale for its intrinsic high stability. In Chapter 3 the conformational and dynamic behaviour of three mannose containing oligosaccharides, a tetrasaccharide with α1→2, and α1→3, and a penta- and a heptasaccharide with α1→2, α1→3, and α1→6 linkages has been evaluated. Molecular mechanics, molecular dynamics simulations and NMR spectroscopy methods were used for evaluation. It is found that they display a fair amount of conformational freedom, with one major and one minor conformation per glycosidic linkage. The evaluation of their recognition by Banlec has been performed by STD NMR methods and a preliminary view of their putative interaction mode has been carried out by means of docking procedures. In Chapter 4 the conformational behaviour of three mannose containing oligosaccharides, namely, the α1→3[α1→6] trisaccharide, the heptasaccharide with α1→2, α1→3, and α1→6 linkages and the tetrasaccharide consisting of α1→3 and α1→2 linkages, when bound to Banlec has been evaluated by trNOE NMR methods and docking calculations. It is found that the molecular recognition event involves a conformational selection process, with only one of the conformations, among those available to the sugar in free state, being recognised at the lectin binding site. It is known that many proteins, including members of the Jacalin-related lectin family (of which Banlec is a member), bind the high-mannose saccharides found on the surface of the HIV-associated envelope glycoprotein, gp120, thus interfering with the viral life cycle, potentially providing a manner of controlling a variety of infections, including HIV. These proteins are thought to recognize the high-mannose type glycans with subtly different structures, although the precise specificities are yet to be clarified. This study was carried out to gain a better understanding of these protein-carbohydrate recognition events. Chapter 5 reports interactions of Calreticulin (CRT) with the trisaccharide Glcα1-3Manα1-2Man. Previously in our laboratory it was established using modeling studies the residues in CRT important for sugar binding. Here, the relative roles of Trp-319, Asp-317 and Asp-160 for sugar binding have been explored by using site-directed mutagenesis and isothermal titration calorimetry (ITC). Residues corresponding to Asp-160 and Asp-317 in calnexin (CNX) are known to play important roles in sugar binding. The present study demonstrates that Asp-160 is not involved in sugar binding, while Asp-317 plays a crucial role. Further, it is also validated that hydroxyl-pi interactions of the sugar with Trp-319 dictate sugar binding in CRT. This study defines further the binding site of CRT and also highlights its subtle differences with that of CNX. Additionally, mono-deoxy analogues of the trisaccharide unit Glcα1-3Manα1-2Man have been used to determine the role of various hydroxyl groups of the sugar substrate in sugar-CRT interactions. Using the thermodynamic data obtained by carrying out ITC of CRT with these analogues, it is demonstrated that the 3-OH group of Glc1 plays an important role in sugar-CRT binding, whereas the 6-OH group does not. Also, the 4-OH, 6-OH of Man2 and 3-OH, 4-OH of Man3 in the trisaccharide are involved in binding, of which 6-OH of Man2 and 4-OH of Man3 have a more significant role to play. Therefore, the interactions between the substrate sugar of glycoproteins and the lectin chaperone CRT are further delineated. Chapter 6 introduces Peptide-N-Glycanase (PNGase) and delineates the various interactions involved in the binding of oligomannose structures of glycoproteins to the C-terminal domain (the carbohydrate recognition module) of PNGase. ITC is used to characterize the interaction to oligosaccharides in terms of affinity, stoichiometry, enthalpy, entropy and heat capacity changes with the mouse PNGase C-terminal domain. Using the thermodynamic data obtained, it was determined that PNGase requires the tri-mannoside moiety of the native glycan on glycoproteins as the basic minimum entity for recognition and binding. Additional mannose moieties on the glycan do not significantly interact with PNGase and therefore no enhancement in binding affinity is observed (unlike CRT) which is in concordance with its role of stripping glycans from misfolded glycoproteins targeted for degradation via the ERAD (Endoplasmic reticulum assisted degradation) pathway. Chapter 7 briefly summarizes all the findings of the research carried out and presents a comparative analysis of the three lectins studied. Appendix A: Protein folding in the ER is assisted by molecular chaperones. Lectin chaperones such as CRT and CNX assist the folding of glycoproteins by their N-linked oligosaccharide chains. Dynamic processing of the original glycan chain of (GlcNAc)2(Man)9(Glc)3 to remove the terminal glucose moieties is essential for accurate folding. Proteins that attain their native conformation are then transported to the Golgi complex for further glycan modifications. In case of aberrant folding the proteins are retrotranslocated into the cytosol, ubiquitinated, deglycosylated and degraded by the proteasome. Peptide-N-glycanase is a cytosolic enzyme that releases N-glycans from glycoproteins and glycopeptides. PNGase is now widely recognized as a major participant in protein quality control machinery for ERAD or the proteasomal degradation of retrotranslocated glycoproteins. It is therefore desirable to synthesize fluorescently labeled glycoprotein substrates which will provide direct understanding of how, when and where, the interaction between the substrate and the enzyme occurs. Towards this goal, cloning of GFP and RFP tagged full length mouse and human PNGase and CRT was carried out which is described in this section.

Pendant très longtemps, les traitements contre les cancers se sont basés sur la cytotoxicité des chimiothérapies, sur leur capacité à tuer directement les cellules malignes ou à induire leur senescence. Mais cette cytotoxicité accrue et non ciblée a également pour effet de tuer les cellules du système immunitaire du patient. Cependant, il a été montré que la radiothérapie, les anthracyclines ainsi que l’oxaliplatine étaient capables d’induire une apoptose décrite comme étant une mort cellulaire immunogène. De ce fait, les cellules tumorales mourantes agiront comme vaccin thérapeutique.La mort cellulaire immunogène se caractérise par trois grands marqueurs : un stress du réticulum endoplasmique pré-mortem qui va induire la translocation de la calréticuline de la lumière du réticulum endoplasmique vers la surface cellulaire, la libération d’ATP dans le milieu extracellulaire permettant le recrutement des cellules dendritiques et l’activation de l’inflammasome NLRP3 via le récepteur P2RX7, et enfin la libération de la protéine HMGB1 dans le milieu extracellulaire, qui va aller interagir avec TLR4 à la surface des cellules dendritiques pour stimuler leur fonction présentatrice d’antigène. La première partie de ce travail a consisté à comprendre les mécanismes moléculaires précis par lesquels l’ATP est sécrétée activement lors d’une mort cellulaire immunogène. En utilisant une combinaison de techniques impliquant des criblages pharmacologiques, des techniques de monitorage de la localisation intracellulaire de l’ATP entre autres, nous avons montré qu’après un traitement par les inducteurs de la mort immunogène, l’ATP était redistribué des lysosomes aux autolysosomes et que sa sécrétion requiert la protéine lysosomale LAMP1. Nous avons également montré qu’il existait d’autres voies de libération d’ATP telles que la voie de signalisation Rho, et également l’ouverture des hémicanaux pannexine 1 (PANX1). De façon surprenante, nous avons observé une implication de PANX1 dans la translocation de LAMP1 à la surface cellulaire. Ces résultats ont permis de comprendre un peu plus précisément les mécanismes de sécrétion d’ATP dans la mort cellulaire immunogène, mettant en évidence l’importance de l’exocytose lysosomale caspases dépendante et PANX1 dépendante.La seconde partie de ce travail s’est portée sur l’étude d’une autre caractéristique de la mort cellulaire immunogène, à savoir l’exposition de la calréticuline à la surface cellulaire. En partant du constat qu’après un traitement par la mitoxantrone, la calréticuline était relocalisée en périphérie à la fois dans les cellules humaines et les cellules de levure, il a été suggéré que la voie d’exposition de la calréticuline était conservée phylogénétiquement. Nous avons montré que les phéromones pouvaient agir comme inducteurs physiologiques de l’exposition de la calréticuline dans les cellules de levure. Un criblage d’ARN interférant et des analyses de transcriptome nous ont permis de montré que les chimiokines, en particulier CXCL8 chez l’humain (appelé également interleukine-8) et son orthologue Cxcl2 chez la souris étaient impliquées dans la translocation de la calréticuline à la surface cellulaire. En traitant les cellules cancéreuses par la mitoxantrone, nous observons une production de CXCL8 par les cellules cancéreuses humaines in vitro et de Cxcl2 par les cellules cancéreuses murines in vivo. Un « knockdown » des récepteurs pour CXCL8/Cxcl2 réduit de manière significative l’exposition de la calréticuline à la surface cellulaire. Ces résultats ont donc montré l’importance des chimiokines dans la voie d’exposition de la calréticuline.L’ensemble de ce travail a permis de comprendre plus en détails deux des trois grandes caractéristiques de la mort cellulaire immunogène. / Cytotoxic anti-neoplastic agents were considered for a long time to mediate their therapeutic effects via their capacity to directly kill malignant cells. Nevertheless, this high cytotoxicity is non-targeted and will eventually diminish immune cells. During the last years, it has been shown that radiotherapy and some anticancer agents, such as anthracyclines and oxaliplatin, can stimulate actively anti-tumor immune responses. In fact, they can induce an immunogenic type of apoptosis, which we termed immunogenic cell death (ICD). Thereby, dying cells can act as therapeutic vaccine against residual cancer cells that overcame the initial treatment.ICD is characterized by three major hallmarks: a pre-mortem stress of the endoplasmic reticulum (ER), which triggers the translocation of the ER chaperone protein called calreticulin (CRT) to the cell surface, the secretion of ATP from apoptotic cells, which acts as a signal for the recruitment of dendritic cells and for the activation of the NLRP3 inflammasome via its receptor P2RX7, and the release of HMGB1 into the extracellular space, allowing it to interact with TLR4 and thus stimulate the antigen-presenting functions of the DCs.The first part of my work focused on the precise molecular mechanisms by which ATP is actively secreted during ICD. Using a large panel of techniques, including chemical compounds screens and monitoring the subcellular localization of ATP, we showed that following treatment of various tumor cells with ICD inducers, ATP is redistributed from lysosomes to autolysosomes and the lysosomal protein LAMP1 is required for active ATP secretion. We also showed that Rho and pannexin 1 (PANX1) are indispensable for efficient ATP release in response to ICD inducers. Surprisingly, we observed an unexpected link between PANX1 and the exposure of LAMP1 at the cell surface. These results will help to understand the mechanisms necessary for ATP secretion during ICD.In the second part of this work we further studied the surface exposure of CRT during ICD. We observed that mitoxantrone (MTX), which belongs to the group of anthracyclines, can induce a peripheral relocalisation of CRT, both in human cells and yeast cells. In addition, we showed that pheromones can act as a physiological inducer of CRT translocation in yeast. Focused siRNA screening combined with transcriptome analyses revealed that human CXCL8 (also called interleukin-8) and its mouse ortholog Cxcl2 play an essential role in the translocation of CRT to the cell surface. Interestingly, MTX-treated human cancer cells displayed an elevated production of CXCL8 in vitro. These results were confirmed in vivo, with MTX treated murine tumors, which also displayed elevated Cxcl2 levels. The MTX-induced CRT exposure was significantly reduced when we performed a knockdown of CXCL8/Cxcl2 receptors. Altogether, these results showed the importance of chemokine signaling circuitries in immunogenic CRT exposure.This work allows for the detailed understanding of the mechanisms of ICD and might thus be useful for further targeted drug development.

Immunogenic cell death (ICD) is characterized by presence of specific molecules including surface exposed calreticulin (CRT) and the heat shock proteins HSP70 and HSP90. Release of ATP and high- mobility group box protein 1 (HMGB1) belongs to other typical characteristics. For induction of ICD in lung cancer cells high-hydrostatic pressure (HHP) was used. Treatment by HHP induces expression of immunogenic markers CRT, HSP70 and HSP90 on the cell surface. HHP also induces secretion of ATP to the extracellular milieu. Dendritic cells (DC) pulsed with HHP-treated tumor cells showed fenotypic maturation characterized by upregulation of maturation molecule CD83, costimulation molecules CD80 and CD86, chemokine receptor CCR7 and MHC class II molecule HLA-DR. Pulsed DCs have also higher rate of phagocytosis of HHP-treated tumor cells and they induce lower numbers of regulatory T cells compared to immature DCs. Moreover, activation of caspases (-8, -9, -3) and other proteins (phosphorylation of eIF2α) which are crucial in ER-stress mediated apoptotic pathway, was observed after HHP treatment. Using wide range of methods it was confirmed that HHP treatment is able to induce ICD in lung cancer cell lines, fenotypic and functional characteristics were described and the decreased induction of regulatory T-lymphocytes...