La phophomannose isomérase de type I (PMI), une métalloenzyme à zinc, et la phosphoglucose isomérase (PGI), catalysent l’isomérisation réversible du β-D-fructose-6-phosphate (F6P), respectivement en β-D-mannose-6-phosphate (M6P) et en α-D-glucose-6-phosphate (G6P). Ces deux enzymes sont des cibles thérapeutiques potentielles. La phosphoglucose isomérase humaine (hPGI), connu également sous le nom de facteur de motilité autocrine (AMF), stimule, en plus de son activité glycolytique intracellulaire, la migration des cellules in vitro et le développement de métastases in vivo. D'autre part, Candida albicans est la principale levure impliquée en pathologie humaine. Ces dernières années, un problème de résistance du germe aux antifongiques classiques est apparu. En conséquence, la recherche se dirige vers de nouvelles cibles thérapeutiques, dont la PMI de C. albicans (CaPMI) qui joue un rôle important dans la biosynthèse de structures mannosylées nécessaires à la survie du pathogène. Les réactions catalysées par ces deux enzymes mettent en jeu le même intermédiaire de haute énergie (IHE) de type 1,2-cis-ènediolate, sauf qu’il est coordiné au zinc dans le cas de la PMI. La surexpression ainsi que la purification de CaPMI et de hPGI ont été réalisées au laboratoire. Une petite chimiothèque a été créée à partir du 5-phospho-D-arabinono-1,4-lactone (5-PAL) en modulant la partie tête de l’IHE. Un groupe chélatant du zinc (zinc binding group, ZBG) a été introduit dans plusieurs composés dans le but d’inhiber sélectivement CaPMI. De plus, deux composés possédant en partie tête une fonction amine terminale ont été synthétisés pour inhiber spécifiquement la PGI humaine en ciblant un résidu glutamate du site actif de l'enzyme (Glu357). Toutes ces molécules ont d’abord été testées sur la PGI du muscle de lapin et la PMI de E. coli commerciales, et par la suite sur la CaPMI et la hPGI surexprimées. Une série de bons voire très bons inhibiteurs de hPGI, et donc potentiellement anti-métastatiques, a été découverte. Ces composés ne sont cependant pas inhibiteurs de la CaPMI. Deux structures tridimensionnelles à haute résolution de complexes enzyme-inhibiteur ont été obtenues. Au delà des aspects thérapeutiques, mécanistiques et structuraux, un biocapteur électrochimique à base d'un des inhibiteurs synthétisés a été réalisé pour la détection de hPGI qui est un biomarqueur validé de cancers métastatiques. Ce biocapteur a démontré une limite de détection de 43 fM dans du tampon phosphate (PBS). / Phosphoglucose isomerase (PGI) and type I phosphomannose isomerase (PMI), a zinc metalloenzyme, catalyze the reversible isomerization of β-D-fructose 6-phosphate (F6P) to α-D-glucose 6-phosphate (G6P) and β-D-mannose 6-phosphate (M6P), respectively. These two enzymes are potential therapeutic targets. Human PGI (hPGI) often called as AMF-PGI (autocrine motility factor-PGI), in addition to its intracellular glycolytic activity, stimulates cell migration in vitro and metastasis in vivo. Inhibition of its extracellular activity is obviously interesting in oncology. On the other hand, Candida albicans is the main yeast involved in human pathology. During recent years, resistance of this pathogenic fungus to conventional antifungal drugs appeared. Consequently, research is moving towards new therapeutic targets, including C. albicans PMI (CaPMI) that plays an important role in the biosynthesis of mannosylated structures required for pathogen survival. The reactions catalyzed by these two enzymes involve the same high energy intermediate (HEI) type 1,2-cis-enediolate, except that it is coordinated to the zinc active site in the case of PMI. Overexpression and purification of both CaPMI and hPGI were performed in our laboratory. A small chemical library was created from the synthon 5-phospho-D-arabinono-1,4-lactone (5-PAL) by modulating the head part of the HEI. A zinc binding group (ZBG) was introduced in several compounds in order to selectively inhibit the CaPMI enzyme. Moreover, two compounds with a terminal amine function were designed to selectively inhibit hPGI by targeting a glutamate residue of the enzyme (Glu357). All these molecules were first tested on rabbit muscle PGI and PMI from E. coli, and later on CaPMI and hPGI. None of these compounds are good inhibitors of CaPMI. However, a series of strong inhibitors of hPGI, and therefore potentially anti-metastatic drugs, was discovered. High-resolved 3D structures of the two enzymes complexed with inhibitors have been successfully obtained. Beyond the therapeutic, mechanistic and structural aspects, an electrochemical biosensor based on one of the synthesized inhibitors was carried out for the detection of hPGI, which is a validated biomarker of metastatic cancers. This biosensor demonstrated a detection limit of 43 fM in phosphate buffer (PBS).

The genus Salmonella consists of pathogenic gram negative organisms which infect intestines of birds, animals and humans. They are the causative agents of salmonellosis which is characterised by diarrhoea, nausea, fever and abdominal cramps. If not treated in time, salmonellosis can also be fatal. Salmonella genus is divided into two species Salmonella bongori and Salmonella enterica. Salmonella enterica is further divided into six subspecies out of which the subspecies enterica has many of the pathogenic serovars of this species. Salmonella typhimurium is a server in the subspecies enterica of Salmonella enterica species. Transmission of salmonellosis takes place through contaminated food and water. When the organism enters a host, it encounters a range of hostile environments such as acidic pH, lack of oxygen as well as immune response of the host. In order to establish infection, the bacterium needs to survive under stressful conditions and propagate itself. Various proteins are induced in cells under unfavourable conditions that protect them in such situations. One such group of proteins belongs to the Universal Stress Protein (USP) family. Universal Stress Proteins are a set of proteins induced in organisms when it is exposed to a variety of environmental insults including heat shock, nutrient starvation, presence of toxic compounds, etc. Although survival in adverse conditions is mediated by induction of this group of proteins, the precise mechanism of cellular protection has not been elucidated yet. The functional role of a protein is directly related to its three-dimensional structure and hence important insights can be gained regarding the role of these proteins by determining their structures. The structures of two Universal Stress Proteins from S. typhimurium; a single domain protein, YnaF and another tandem USP domain protein, YdaA were determined by X-ray crystallography and biochemical analysis was carried out on them. Guided by structure, plausible roles for both the proteins in stress tolerance of S. typhimurium have been proposed. Additionally, work was also carried out on phosphomannose isomerise from S. typhimurium. Phosphomannose isomerase is a housekeeping enzyme which catalyses the interconversion of mannose-6-phosphate and fructose-6-phosphate. Mannose is important for mannosylation of various lipids and proteins which form an important component of bacterial and fungal cell walls. Presence of a functional phosphomannose isomerise enzyme is important as it helps the organism survive adverse conditions by forming a strong cell wall which shields it from harmful environments. Moreover, phosphomannose isomerase was also found to be essential for virulence of Leishmania mexicana and Cryptococcus neoformans. The structure of phosphomannose isomerase from S. typhimurium was determined in our laboratory in the year 2009. However, in the earlier studies, the catalytically important residues had not been identified and mechanism of isomerisation was not established. Structural analysis, site directed mutagenesis and biochemical assays were used to identify key residues in the active site of StPMI. Identification of these residues might help in deciphering the catalytic mechanism which will eventually be useful to develop inhibitors that arrest the growth of Salmonella as well as other microorganisms. The work reported in this thesis describes the efforts made to enhance our understanding of functional aspects of the two Universal Stress Proteins, YnaF and YdaA and phosphomannose isomerase from S. typhimurium. Chapter 1 begins with a brief introduction to the kinds of unfavourable environments encountered by microorganisms and their strategies of adaptation. This is followed by a review of the literature on Universal Stress Proteins, which are induced in many organisms in response to arrest of or perturbations in the growth rate. Structural, biochemical and evolutionary aspects of members of the family have also been discussed. Subsequently, a brief description of the earlier work carried out on another enzyme important in stress tolerance, phosphomannose isomerase, has been documented. A detailed account of mechanisms of isomerisation carried out by aldose ketose isomerases and identification of important strategies for determination of mechanism of phosphomannose isomerase catalysed reaction have then been provided. The chapter ends with a summary of aims and objectives of the present work. Chapter 2 describes the various experimental techniques and computational methods used during the course of this thesis work. Isolation of plasmids, overexpression and purification of protein, site directed mutagenesis, biochemical assays, crystallisation of proteins, X ray diffraction data collection form a part of the experimental aspect and have been described in detail. Brief descriptions of the programs used and principles behind computational methods used for structure determination (including data processing, phasing, model building and refinement), validation and analysis have also been provided. Chapter 3 includes the structural and functional studies carried out on YdaA, a tandem USP domain protein from S. typhimurium. Expression, purification, crystallisation and structure determination of YdaA in its native and ADP bound forms are described in the chapter. Biochemical assays with radiolabelled ATP showed that YdaA was an ATPase. The crystal structure of YdaA complexed with ATP revealed the presence of ADP (hydrolysis product of ATP) only in the C-terminal domain of the protein. Based on structural analysis and presence of ATP binding motif in the C-terminal domain, it could be hypothesized that ATP hydrolysis activity of the protein is confined to the C-terminal domain of the protein. The N-terminal domain of the protein was found to play another interesting role. A zinc binding site could be identified in the N terminal domain based on structural analysis and elemental X-ray absorption studies done at the synchrotron. Site directed mutagenesis and biochemical experiments suggested that zinc binding in the N-terminal domain was not related to ATPase activity of the C-terminal domain. Additionally, an intermediate of lipid A biosynthesis pathway UDP-(3-O-(R-3-hydroxymyristoyl))-N-acetyl glucosamine was found bound to the N-terminal domain of YdaA. Lipid A is the membrane anchor of polysaccharides in the outer membrane of gram negative organisms and the intermediate occurs at the committed step of the pathway. However, no similarities could be identified between YdaA and members of the relevant biosynthetic pathway. Therefore, YdaA is unlikely to play a catalytic role in the same pathway but can function as a carrier molecule. A plausible link between the N- and C-terminal domains of YdaA could be identified by structural analysis. Many catalytically suitable residues from the N-terminal domain were found to be close to the β-phosphate of ADP bound to the C-terminal domain. Hence YdaA was identified to be a zinc binding ATPase which might play some yet unidentified role in lipid A biosynthesis pathway. Chapter 4 describes the attempts made towards understanding the functional role of YnaF, a single domain USP from S. typhimurium. A description of the expression, purification, crystallisation and X ray diffraction techniques used for structure determination of YnaF and its single site mutant have been provided in detail. Gel filtration, dynamic light scattering studies and the crystal structure determination of YnaF showed a tetrameric organisation of four USP protomers stabilised in the centre by chloride ions. Additionally, YnaF crystallised with a bound ATP even though ATP was not included in the crystallisation cocktail. Biochemical assays on YnaF with radiolabelled ATP showed that it was inactive with respect to ATP hydrolysis. When selected mutations that disrupt chloride binding were made, YnaF was converted to an active ATPase. The crystal structure of the mutant complexed with an ATP analogue revealed key differences at the active site in comparison with that of the wild type and allowed identification of residues that might be important for ATP hydrolysis in this group of proteins. Hence YnaF might play the role of a sensor protein in some signal transduction pathway involving chloride ions in bacteria. A structure based analysis and comparison of USPs from the Protein Data Bank with the structures of YnaF and YdaA is summarised at the end of this chapter. Chapter 5 describes the efforts carried out towards determination of mechanism of isomerisation catalysed by phosphomannose isomerise (PMI). Earlier reports suggest that the enzyme catalyses the reversible isomerisation of mannose-6-phosphate and fructose-6-phosphate via formation of a cis-enediol intermediate. The structure of phosphomannose isomerase from S. typhimurium has been reported by our laboratory. The enzyme is a monomer with three domains; a catalytic domain, a carboxy terminal domain and an α-helical domain. Residues from the catalytic domain were found to coordinate a zinc ion. Overexpression, purification, co crystallisation experiments and soaking studies carried out on crystals of PMI and its single site mutants are outlined in this chapter. The structure of a complex of PMI with mannose-6-phosphate at pH 7.0 revealed the presence of a blob of density close to the zinc binding site which was confirmed to be the active site by analysis of conservation of residues in the site. Based on site directed mutagenesis, activity studies and analysis of structure of PMI, zinc was identified to play an important role in maintaining the structural integrity of the active site. Electrostatic surface analysis of the structure of PMI revealed that the zinc ion might also play the role of anchoring phosphate moiety of the substrate in a highly negatively charged active site pocket. Activity assays following site directed mutagenesis studies eliminated the role of Glu264 in catalysis and implicated two lysines, Lys86 and Lys132 as the possible base in the reaction. The plausible role of a highly conserved residue Arg274 was also proposed based on comparison of structures of wild type and mutant PMIs. The future prospects of the work are briefly discussed towards the end of the thesis. Further experiments and analysis required to obtain better understanding of the functions of these proteins have been discussed. The Appendix section describes extensive crystallisation attempts that were carried out on the enzyme sorbitol-6-phosphate-dehydrogenase from S. typhimurium which catalyses the isomerisation reaction between sorbitol-6-phosphate and glucose-6-phosphate using NADPH as the cofactor. Needle shaped crystals were obtained which diffracted to a poor resolution of 7-8 Å at our in house X ray facility. Attempts to improve the quality of the crystals like co crystallisation with substrate and its analogues, soaking in various compounds and seeding are briefly described. The following manuscripts based on work described in this thesis have been published or will be communicated for publication. 1. Structural and functional analysis of two universal stress proteins YdaA and YnaF from Salmonella typhimurium: possible roles in microbial stress tolerance. Bangera M., Panigrahi R., Sagurthi S.R., Savithri H.S., Murthy M.R.N. Journal of Structural Biology, 2015 Mar; 189 (3): 238-50. 2. Structural and functional insights into phosphomannose isomerise: role of zinc and catalytic residues. Bangera M., Savithri H.S., Murthy M.R.N. Manuscript under preparation

Les aldose-cétose isomérases sont des enzymes catalysant l’isomérisation réversible entre un aldose et un cétose. Nous avons étudiés trois d’entre-elles : les phosphoribose isomérases (RPI), les phosphomannose isomérases (PMI) et les phosphoglucose isomérases (PGI). Ces enzymes interviennent dans différentes voies métaboliques comme la glycolyse, la néoglucogenèse, la voie des pentoses phosphates ou le métabolisme du mannose. Il a été montré qu’elles jouent un rôle important pour assurer la survie et le développement de plusieurs parasites responsables de maladies comme la leishmaniose, la mucoviscidose, la tuberculose, le paludisme ou la maladie du sommeil. Ces enzymes sont donc des cibles thérapeutiques potentielles. Ainsi, les puissants inhibiteurs de ces enzymes peuvent donc être des agents thérapeutiques efficaces pour combattre ces maladies. Les réactions catalysées par ces enzymes impliquent des intermédiaires de haute énergie (IHE) de type 1,2-cis-ènediol(ate). La synthèse d’analogues de ces intermédiaires a permis d’obtenir au laboratoire, les meilleurs inhibiteurs connus de ces enzymes, l’acide 5-phospho-D-arabononohydroxamique (5PAH, meilleur inhibiteur des PMI et PGI) et le 5-phospho-D-ribonate (5PRA, meilleur inhibiteur des RPI). Cependant, ces inhibiteurs possèdent une fonction phosphate facilement hydrolysable en milieu physiologique. Ce qui les rend inactifs in vivo. Au cours de ce travail de thèse, des phosphoanalogues du 5PAH, du 5-phospho-D-ribose (R5P, le substrat des RPI) et du 5PRA possédant une fonction malonate, phosphonate, phosphorothiate, sulfate et sulfonate à la place de la fonction phosphate ont été obtenus par des voies de synthèse multi-étapes faisant intervenir le D-arabinose ou le D-ribose comme produit de départ. Les propriétés inhibitrices de ces composés ont ensuite été déterminées et leur stabilité en milieu physiologique évaluée. Le phosphoanalogue du 5PAH de type malonate, l’acide 5-désoxy-5-dicarboxyméthyl-D-arabinonohydroxamique (5DCAH) est un inhibiteur moyen et stable de la PMI d’Escherichia Coli. Parmi les phosphoanalogues du R5P, les composés de type sulfate et sulfonate, respectivement, le 5-sulfate-D-ribose (5SR) et 5-désoxy-5-sulfonométhyl-D-ribose (5SMR) sont de bons inhibiteurs de trois RPI (la RPI d’épinard, la RPI d’Escherichia Coli et la RPI de Micobacterium tuberculosis). Seul le composé de type sulfonate est stable en milieu physiologique. Le phosphoanalogue de type malonate, le 5-désoxy-5-dicarboxyméthyl-D-ribose (5DCR) est un inhibiteur moyen de ces trois RPI. En revanche, les phosphoanalogues de type phosphorothioate et phosphonate, respectivement, le 5-désoxy-5-phosphorothioate-D-ribose (5PTR) et 5-désoxy-5-phosphonométhyl-D-ribose (5PMR) sont de mauvais inhibiteurs. Le phosphoanalogue de type phosphonate du 5PRA, le 5-désoxy-5-phosphonométhyl-D-ribonate (5PMRA) est un bon inhibiteur de la RPI de Micobacterium tuberculosis. De plus, ce composé est stable en milieu physiologique. Il est en revanche un mauvais inhibiteur de la RPI d’épinard et d’Escherichia Coli. Ces résultats sont particulièrement prometteurs puisque le 5PMRA est à ce jour le meilleur inhibiteur stable et spécifique de la RPI de Micobacterium tuberculosis. / Aldose-ketose isomerases are enzymes which catalyze the interconversion of an aldose and a ketose. We have studied three of them: phosphoribose isomerase (RPI), phosphomannose isomerase (PMI) and phosphoglucose isomerase (PGI). These enzymes play a major role in various metabolic pathways as glycolysis, neoglucogenesis, the pentoses phosphates pathways or the mannose metabolism. It has been shown to have a crucial role for the survival and development of several microorganisms responsible for diseases as the leishmaniose, the cystic fibrosis, the tuberculosis, the malaria or the insomnia. These enzymes are thus potential therapeutic targets. Consequently, strong inhibitors of these enzymes could provide efficient therapeutic tools against these deseases. The reactions catalyzed by these enzymes involve intermediaries of high energy (IHE) of 1,2-cis-enediol(ate) type. The synthesis of analogues of these intermediaries allowed to obtain in the laboratory, the best inhibitors known for these enzymes, the acid 5-phospho-D-arabononohydroxamique (5PAH, the best inhibitor of the PMI and PGI) and the 5-phospho-D-ribonate (5PRA, the best inhibitor of the RPI). However, these inhibitors possess a phosphate group which is easily hydrolysable in physiological environment, what makes them inactive in vivo. During this work of thesis, phosphoanalogues of the 5PAH, the 5-phospho-D-ribose (R5P, the substrate of the RPI) and of the 5PRA possessing a malonate, phosphonate, phosphorothiate, sulphate and sulfonate were obtained by multi-steps synthesis bringing in D-arabinose or D-ribose as starting product. The inhibitive properties of these compounds were then determined and their stability in physiological environment evaluated. The phosphoanalogue of the 5PAH of malonate type, the acid 5-desoxy-5-dicarboxyméthyl-D-arabinonohydroxamique (5DCAH) is a modest and stable inhibitor of the PMI of Escherichia Coli. Among the phosphoanalogues of the R5P, the compounds of sulphate and sulfonate types, respectively, the 5-sulfate-D-ribose (5SR) and 5-desoxy-sulfonomethyl-D-ribose (5SMR), are good inhibitors of three RPI (the RPI of spinach, the RPI of Escherichia Coli and the RPI of Micobacterium tuberculosis). Only the compound of sulfonate type is stable in physiological environment. The phosphoanalogue of malonate type, the 5-desoxy-5-dicarboxymethyl-D-ribose (5DCR) is a modest inhibitor of this three RPI. On the other hand, the phosphoanalogues of phosphorothioate and phosphonate types, respectively, the 5-desoxy-5-phosphorothioate-D-ribose (5PTR) and the 5-desoxy-5-phosphonomethyl-D-ribose (5PMR), are bad inhibitors. The phosphoanalogue of phosphonate type of the 5PRA, the 5-desoxy-5-phosphonomethyl-D-ribonate (5PMRA), is a good inhibitor of the RPI of Micobacterium tuberculosis. Furthermore, this compound is stable in physiological environment. It is on the other hand a bad inhibitor of the RPI of spinach and Escherichia Coli. These results are particularly promising because the 5PMRA is this day the best stable and specific inhibitor of the RPI of Micobacterium tuberculosis.