Chaque année, Mycobacterium tuberculosis (Mtu) est responsable d‘environ 2 million de morts dans le monde. L’apparition de souches multi-résistantes aux antibiotiques rend nécessaire le développement de nouveaux antituberculeux. Dans ce contexte, les voies de biosynthèse des composés essentiels de l’enveloppe, tels que les acides mycoliques (AM), représentent des cibles thérapeutiques de choix. Les AM, cruciaux pour l’architecture et la perméabilité de l’enveloppe, jouent également un rôle important dans la virulence et dans la persistance bactérienne. La biosynthèse des AM fait appel à une Fatty Acid Synthase de type II (FAS-II), système multienzymatique qui produit les chaînes méromycoliques des AM et qui contient au moins deux déshydratases, HadAB et HadBC. Lors d’une analyse approfondie du génome de Mtu dans l’équipe, onze hydratases/déshydratases (R)-spécifiques potentielles ont été identifiées, dont la protéine Rv0504c, nommée "HadDMtu". L’objectif de mes travaux a été de caractériser la fonction enzymatique et le rôle physiologique de la protéine HadDMtu chez Mtu et de son orthologue potentiel chez M. smegmatis (Msm), un modèle d’étude mycobactérien, non pathogène et à croissance rapide. Les tests d’activité enzymatique in vitro sur la protéine HadDMsm ont montré qu'elle possède une activité hydratase/deshydratase. Lorsque le gène hadD est délété, chez Msm et chez Mtu, les profils et les structures fines des AM sont modifiés. Chez Msm, le mutant n’est plus capable de produire d’AM à longues chaînes α et epoxy. Par contre, chez Mtu, c’est la structure fine et la proportion des AM keto qui sont affectées. En parallèle de mes travaux, il a été montré que la protéine HadDMsm de Msm interagit spécifiquement avec HadAB. L'ensemble de nos données a apporté la preuve que HadD fait partie du système FAS-II. Il suggère fortement que dans les deux espèces, l’enzyme intervient dans les derniers cycles d’élongations de la chaîne méromycolique. En accord avec ces résultats, des expériences de complémentation croisée des souches mutantes de Mtu ΔhadDMtu et de Msm ΔhadDMsm, ont montré que les deux protéines HadD ont une fonction proche in vivo mais elles ne sont pas orthologues. Des tests phénotypiques ont montré que les propriétés de surface et l’intégrité de l’enveloppe des mutants ΔhadD sont profondément altérées. De plus, chez la souris, la souche mutante de Mtu apparait beaucoup moins virulente avec une baisse des charges bactériennes dans les poumons et dans la rate. / Mycobacterium tuberculosis (Mtu) kills 2 million people every year in the world. The development of new antituberculous drugs is urgently needed due to the emergence of drug-resistant strains. In this context, the biosynthesis pathways of essential compounds from the mycobacterial envelope, such as mycolic acids (MAs), constitute relevant therapeutic targets. The MAs are crucial for both the architecture and the permeability of the envelope and play important roles in the bacterial virulence and persistence. The MA biosynthesis involves a Fatty Acid Synthase type II (FAS-II), a multi-enzyme system that produces the main meromycolic chains of MA and contains at least two dehydratases, HadAB and HadBC. A thorough analysis of the Mtu genome led to the identification of eleven potential (R)-specific hydratases/dehydratases, including the protein Rv0504c, named “HadDMtu”. The objective of my work was to characterize the enzymatic function and the physiological role of HadDMtu protein in Mtu and of its putative orthologue in M. smegmatis (Msm), a non-pathogenic and fast growing mycobacteria model. The in vitro enzymatic activity tests with HadDMsm protein showed that it has a hydratase/dehydratase activity. When hadD gene is deleted, in Msm and in Mtu, the MA distribution and fine structures were changed. In Msm, the mutant could not produce long chain α- and epoxy-MAs while in Mtu, it is the fine structure and the proportion of the keto-MAs that were affected. In parallel of my work, it was that HadDMsm protein specifically interacts with HadAB. All these data show that HadD belongs to the FAS-II system. They strongly suggest that, in the both two strains, the enzyme occurs in the late steps of the meromycolic chain elongation. According to these results, cross-complementation assays of Mtu and Msm ΔhadD mutants showed that both HadD proteins have a similar function in vivo but they are not orthologs. Phenotypic tests showed that the surface properties and the envelope integrity of the ΔhadD mutants were profoundly altered. Also, in the mouse model of infection, Mtu mutant exhibited a significant loss of virulence with a decrease of bacterial loads in the lungs and in the spleen.

This thesis describes the target selection, cloning, expression, purification, crystallization, structure and biochemical characterization of five essential Mycobacterium tuberculosis (Mtb) proteins. The search for drugs against the causal agent of tuberculosis is urgently needed and the targeting of essential genes is necessary to fulfill this goal. The crystal structures of carbonic anhydrases (CA) Rv1284 and Rv3588c have been determined to 2.0 and 1.7 Å resolution, respectively. Rv3588c, in contrast to Rv1284, is an active β-CA that shows two different active site conformations and pH-dependent oligomerization states. Rv1295 is an active threonine synthase with an unusually high pH optimum; the structure has been solved to 2.5 Å resolution, based on which a modification to the reaction mechanism published previously is proposed. Mtb has a thick and impermeable cell envelope that constitutes an efficient barrier against drugs. One of the essential components of the envelope is mycolic acid (MA). The inhibition of enzymes participating in its synthesis would be lethal for Mtb. Rv0636, a formerly unknown-function protein has β-hydroxyacyl-ACP dehydrase activity which is essential for MA synthesis. Co-expression with partners notably improves its solubility. Around 55% of Mtb proteins have unknown function. Rv3778c is one of them and its three-dimensional structure has been determined to 1.8 Å resolution. Studies aimed at the elucidation of its biochemical function are shown. A pathway not yet reported in Mtb is also suggested.