Highly infectious pathogenic strains of bacteria are becoming increasingly resistant to the current clinical antibiotics which have created a dire need for the development of novel antibiotics. O-Acetylpeptidoglycan esterase 1 (Ape1) is a periplasmic esterase present in several peptidoglycan (PG) O-acetylating pathogenic species of Gram-positive and all Gram-negative bacteria that perform this modification to this essential cell wall polymer. Inhibition of this growth-limiting enzyme may prove the principle that Ape1 has the potential to be the target for the development of a novel class of antibiotics. Ape1 plays a crucial role in bacterial growth by regulating PG turnover through catalytic removal of the C-6 acetyl group from O-acetylPG. This activity is required for the continued metabolism of PG because the major autolytic enzymes involved, the lytic transglycosylases, require a free C-6 hydroxyl group to produce their reaction product, 1,6-anhydromuramic acid. Several of the compounds that have been identified to effectively inhibit Ape1, were re-evaluated by determining their kinetic parameters. Work presented in this thesis explored the inhibitory potential of these compounds, belonging to the anthraquinone (alizarin, quinizarin, quinalizarin, emodin, sennoside A) or tannin (ellagic acid) families of compounds, both in vitro and in vivo, among species of bacteria that are known to O-acetylate their PG. Of the inhibitory compounds tested, ellagic acid was found to be most effective in vitro, with an IC50 value of 0.91 µM ± 0.06, Ki 1.18 ± 0.04 and in vivo it was shown to reduce bacterial growth.

<p>Bacteria recycle pre-existing peptidoglycan in order to minimize the de novo synthesis of peptidoglycan precursors. The recycling pathway is under study for its chemotherapeutic target potential. Anhydromuramic acid kinase (AnmK) is part of this recycling pathway and catalyzes the dual hydrolysis/phosphorylation of anhMurNAc to MurNAc-6-P. This enzyme has been discovered and introduced, but only minimally characterized. Therefore, the overarching goal of this work was to clone, express and purify AnmK to homogeneity; perform further kinetic characterization; solve the open, closed and transition state mimic-bound conformations of AnmK by x-ray crystallography; and develop a putative mechanism based on the accumulated research findings and <super>18</super>O-labeling studies.</p><p>The anmK gene was successfully cloned as a hexahistidine fusion protein and overexpression was optimized. After exhaustive trials, a final purification scheme was designed to yield homogeneous AnmK in three chromatographic steps and in less than 36 hours. Additionally, the synthesis of both anhMurNAc and a pseudosubstrate (anhGlcNAc) were carried out in 35% and 77% overall yield, respectively. The synthesis of these compounds allowed for both kinetic characterization and structural studies. </p><p>To this end, the structure of de novo AnmK was solved using SAD and high-resolution (1.9 &Aring;) data. Also, an ATP analog (ANP) and anhMurNAc substrate-bound, closed conformation structure (1.95 &Aring;) was solved. These structures elucidated an 11&deg; domain closure of the enzyme upon substrate binding and also revealed the active site geometry to be used to determine potential molecular determinants of specificity. </p><p>Insights into the kinetic mechanism of AnmK were then gathered using multiple techniques. First, the structure of AnmK (2.5 &Aring;) was solved the with a known transition state analog, the MgADP-vanadate complex. Following this structure, which sheds light on the potential importance of a residue other than the catalytic base (Asp187), isotopic labeling was performed with H₂<super>18</super>O. Using NMR and MS, the regiochemical selectivity of AnmK hydrolysis to impart the solvent derived oxygen at C1 was established. Additionally, this was carried out with stereochemical preference to create the &alpha;-anomer of the carbohydrate product. This regiochemistry and stereospecificity drove the design of our putative concomitant hydrolysis/phosphorylation mechanism but we are not able to rule out the formation of a transient phosphoenzyme intermediate.</p><p>This research can be applied to the immediate goal of understanding the function of a single, novel enzyme with unique chemistry and the clarification of the AnmK mechanism will facilitate future investigation into enzymes with dual hydrolase/kinase functionality. Furthermore, this research contributes to understanding of the complex bacterial peptidoglycan layer in order to harness new ideas for developing antibiotics.</p> / Dissertation

L'acquisition et le maintien de la forme bactérienne ont été consciencieusement étudiés pendant une très longue période. Néanmoins, il reste encore beaucoup de questions sans réponse. Les bactéries Gram-positives présentent une couche externe rigide (la paroi cellulaire) qui permet de préserver la pression osmotique interne et la morphologie cellulaire. La paroi cellulaire (CW) est principalement formée par un maillage de polymères de sucres, le peptidoglycane (PG), sur lequel sont accrochés des acides téichoïques. L'absence de cette barrière essentielle provoque la perte de forme et, finalement, la lyse de la cellule. L’intégrité du CW est par conséquent d'une importance vitale pour les bactéries. La structure ainsi que la synthèse correcte du CW dépendent de supposées machineries d'élongation du peptidoglycane (PGEM) chargées d’assembler le réseau du PG. Le fonctionnement et la composition des PGEMs restent incertains, mais on sait qu’ils dépendent d’une protéine essentielle : MreB, une protéine procaryote similaire à l'actine. MreB est suspectée de contrôler l’activité et/ou l’assemblage des PGEMs, mais sa fonction exacte comme son mode de régulation sont actuellement inconnus. J’utilise Bacillus subtilis, le modèle des bactéries Gram-positives, pour mieux comprendre les fonctions de MreB via i- le développement et l’utilisation d’un criblage génétique pour l’identification de mutants de mreB non fonctionnels et ii- l'étude d'un effecteur potentiel de MreB.(i) MreB a été étudié pendant près de deux décennies et pourtant, sa (ses) fonction(s) reste(nt) mal comprise(s). Comme les approches biochimiques se sont révélées particulièrement difficiles jusqu'à présent, la plupart des études se sont concentrées sur la localisation cellulaire et la dynamique de la protéine. Au cours de mes travaux, j’ai conçu un criblage génétique au moyen duquel j’ai obtenu une collection de mutants de mreB fonctionnellement déficients, chez B. subtilis. La caractérisation de ces mutants a révélé de nombreux résidus importants pour le fonctionnement de la protéine. De façon intéressante, mes résultats indiquent que certains mutants ont conservé leurs propriétés dynamiques (suggérant une association fonctionnelle aux PGEMs) en plus d'une morphologie de type sauvage, tout en étant fortement affectés pour la croissance. Des résultats préliminaires indiquent que ces mutants sont compromis dans leur capacité à utiliser certaines sources de carbone, reliant MreB au métabolisme cellulaire. Ceci suggère l'existence soit d'un point de contrôle, soit d'un couplage entre le métabolisme du carbone et l'expansion du CW chez B. subtilis.(ii) Des résultats non publiés de notre groupe ont révélé l'existence d'un opéron non caractérisé (ydcFGH) dont l'expression est fortement induite en absence de MreB, par comparaison à la souche sauvage. J’ai 1- mis en évidence la cause probable de l’induction de cet opéron en l’absence de MreB, révélant ainsi l’existence de nombreuses mutations dans la souche MreB et 2- réalisé une caractérisation poussée de chaque gène de l'opéron ydcFGH. Bien que le lien exact entre MreB et ydcFGH soit encore inconnu, nos résultats suggèrent un rôle potentiel d’YdcH dans le contrôle du métabolisme du carbone et l'adaptation à la phase stationnaire. À la lumière de mes données issues du criblage génétique (i), ces résultats indiquent un lien fort entre MreB et le métabolisme du carbone. / Acquisition and maintenance of the bacterial shape has been conscientiously studied for a long time. Nevertheless, there are still many unanswered questions. Gram-positive bacteria present a rigid external coating (cell wall) that allows them to preserve internal osmotic pressure and cell morphology. The cell wall (CW) is mainly formed by the peptidoglycan meshwork (PG), that confers its structure to the CW, to which are connected teichoic acids. The absence of this essential barrier causes the loss of shape and, ultimately, lysis of the cells. Integrity of the CW is, therefore, a matter of vital importance for bacteria. Proper CW synthesis and structure depends on the so-called peptidoglycan elongation machineries (PGEM) in charge of building the PG meshwork. The precise composition and functioning of the PGEM is not completely understood but they rely on a key player: MreB, a conserved prokaryotic actin-like protein. MreB is suspected to control PGEM activity and/or assembly but its precise function and mode of regulation are currently unknown. I used Bacillus subtilis, the model for Gram-positive bacteria, to gain a better understanding of MreB functions via i- the development and use of a genetic screen for loss-of-function mutants of mreB and ii- the study of a potential effector of MreB.(i) MreB has been studied for almost two decades now and still, little is known about its function(s). Since biochemical approaches proved to be difficult so far, most of the studies have focused on cellular localization and dynamics of the protein. Here, I have designed a genetic screen by means of which I have obtained a collection of functionally impaired mreB mutants in B. subtilis. Characterization of these mutants revealed numerous key residues for the functioning of the protein. Interestingly, my results indicate that some mutants have kept their dynamic properties (suggesting functional association to the PGEM) together with a wild type shape, while being strongly affected for growth. Preliminary results indicate an impaired ability to use certain carbon sources linking MreB to cellular metabolism. This suggests the existence of either a checkpoint or a coupling between carbon metabolism and CW expansion in B. subtilis.(ii) Unpublished results from our group revealed the existence of an uncharacterized operon (ydcFGH), whose expression is highly induced in the absence of MreB by comparison to the wild type. I have 1- deciphered the cause of ydcFGH induction in the absence of MreB, revealing the existence of multiple mutations in the MreB strain and 2- realized a thorough characterization of each gene of the ydcFGH operon. Although the exact link between MreB and ydcFGH is yet unknown, my results suggest a potential role of YdcH in the control of carbon metabolism and adaptation to stationary phase. In light of my mutagenesis screen data (i), these results are pointing towards a strong link between MreB and carbon metabolism.

The bacterial pathogen responsible for Lyme disease ¬— Borrelia burgdorferi— is an atypical Gram-negative spirochete that is transmitted to humans via the bite of an infected Ixodes tick. Like all Gram-negative bacteria the structural portion of the cell envelope known as peptidoglycan (PG) is sandwiched between the inner and outer membranes. Unlike virtually all bacteria, this PG layer is unique in B. burgdorferi in that the amino acid structure differs from most Gram-negative and Gram-positive bacteria by the addition of an Ornithine residue to the third amino acid location in the crosslinking structure. This unique motif is hypothesized to be responsible for the unusual clinical manifestations seen in Lyme disease, specifically Lyme arthritis, the most common late stage symptom of the disease in the United States. Peptidoglycan is only one component of the cell envelope in B. burgdorferi though; other portions of the cell envelope remain understudied specifically when viewed through the lens of the immune response they may elicit in addition to that of PG. The combined immunological effect of the unique bacterial antigen found in B. burgdorferi PG, as well as other potentially associated proteins contained within the cell wall, are explored here. These studies further our understanding of the B. burgdorferi cell envelope and provide critical information that underlies the elusive pathogenesis of Lyme disease. / Master of Science in Life Sciences / Lyme disease is a growing health concern, namely for the countries in the Northern Hemisphere. The bacterium responsible for this illness is Borrelia burgdorferi. B. burgdorferi can survive in the human body and is a threat in that as it replicates in the human host, it sheds pro-inflammatory fragments of its unique cell wall into the environment. This thesis will explore the consequences of this cell wall shedding and how the human immune response differs from the response seen in other more common bacteria. Additionally, I have found that the cell envelope fragments shed from B. burgdorferi may contain more than meets the eye. There is evidence here to support the discovery of a moonlighting protein that is bound to a portion of the cell wall in B. burgdorferi. This protein acts to bolster the structural integrity of the cell while also acting to modulate the host immune response.

Peptidoglycan (PGN) är en bakteriecellväggskomponent och känd för att känna igen olika receptorer eller enzymer för att leda aktiveringsimmunsystemet. Den allmänna strukturen för PGN består av sockerkedjor inklusive N-acetylglutamin (GlcNAc), N-acetylmuraminsyra (MurNAc) och tvärbundna peptidkedjor. PGN-fragment med D-Lac-ändpeptider har hittats från vankomycinresistenta enterokocker men ett kemiskt syntetiserat PGN-fragment med en D-Lac-ändpeptid har inte undersökts i detalj. Således fokuserade vi på syntesen av PGN-fragmentstrukturer som inkluderar en D-Ala-D-Lac-rest vid den terminala delen av peptidkedjan. För att syntetisera dessa fragmentstrukturer planerade vi att kombinera fastfassyntes (för Lac-peptiddelen) och lösningsfassyntes (för glykanberedning och kondensation). Detta tillvägagångssätt är fördelaktigt för framställning av peptidoglykanfragment med en komplex grenad peptiddel. Först beredde vi sockerdelen MurNAc-derivatet i lösningsfassyntes från ett glukosderivat. Medan den Lac-innehållande peptiden framställdes med fastfas-peptidsyntes med användning av 2-klortritylkloridharts. Med denna förening gav kondensationen av dessa två föreningar det önskade D-Lac-avslutade peptidoglykanfragmentet. / Peptidoglycan (PGN) is a bacterial cell wall component and known to be recognized by various receptors or enzymes to lead the activation immune system. The general structure of PGN consists of sugar chains including N-acetylglutamine (GlcNAc), N-acetylmuramic acid (MurNAc) and cross-linked peptide chains. PGN fragments having D-Lac terminus peptides have been found from vancomycin-resistant enterococcus, but a chemically synthesized PGN fragment having a D-Lac terminus peptide has not been examined in detail. Thus, we focused on the synthesis of PGN fragment structures that include a D-Ala-D-Lac residue at the terminal part of the peptide chain. In order to synthesize these fragment structures, we planned to combine solid-phase synthesis (for the peptide- Lac part) and solution-phase synthesis (for glycan preparation and the condensation). This approach is advantageous for the preparation of peptidoglycan fragments having complex branched peptide moiety. First, we prepared the sugar moiety MurNAc derivative in solution-phase synthesis from a glucose derivative. While, the Lac-containing peptide was prepared with solid-phase peptide synthesis using 2-chlorotrityl chloride resin. Having this compound, the condensation of these two compounds gave the desired D-Lac-terminated peptidoglycan fragment.

Diss. (sammanfattning) Umeå : Univ., 2007. / Härtill 4 uppsatser.

Le peptidoglycane (PG) est le composant majeur de la paroi des bactéries à Gram positif. Il assure la forme et l’intégrité de la cellule bactérienne. Le PG ou des fragments dérivés sont connus pour être des inducteurs du système d’immunité innée de l’hôte, en particulier au travers des récepteurs Nod2. Au cours de ce travail, nous avons étudié l’influence du métabolisme du PG sur les propriétés immunomodulatrices de Lactobacillus casei BL23, en étudiant principalement sa capacité à moduler la réponse des cellules dendritiques humaines. Nous avons tout d’abord caractérisé les hydrolases du PG (PGHs) majeures de L. casei BL23. Une recherche in silico a révélé que L. casei possède un système de PGHs relativement complexe comprenant treize enzymes putatives avec des domaines catalytiques variés. Une analyse protéomique d’extraits de paroi de L. casei BL23 a permis de détecter la production de sept d’entre elles pendant la croissance bactérienne. Quatre d’entre elles ont été étudié plus en détails. La PGH la plus fortement exprimée, Lc-p75, a une activité de -D-glutamyl-L-lysyl-endopeptidase et est responsable de la séparation des cellules après division. De plus, Lc-p75 associée à la paroi est localisée au niveau des septa cellulaires. Il s’agit également de l’une des protéines majeures secrétée dans le surnageant de culture de L. casei BL23. Lc-p75 possède la particularité d’être une glycoprotéine. La PGH Lc-p40 possède un domaine CHAP doué d’une activité endopeptidase avec un site de clivage situé au niveau des ponts interpeptidiques du PG. Lc-p40 parait localisée au niveau de la paroi latérale des cellules de L. casei. Lc-p45 est une deuxième -D-glutamyl-L-lysyl-endopeptidase avec un rôle dans le maintien de la forme de la bactérie. Enfin nous avons caractérisé deux enzymes de prophages, Lc-Lys et Lc-Lys2, codée par le génome de L. casei BL23, qui possède toute deux un domaine de liaison au PG d’un nouveau type qui possède la particularité de lier spécifiquement le D-Asp amidé présent dans les ponts interpeptidiques du PG de L. casei BL23. La délétion des deux gènes qui codent pour les endopeptidases Lc-p75 et Lc-p45 chez L. casei BL23 conduit à l’absence de disaccharide dipeptide dans la structure du PG du mutant, tandis que la délétion de Lc-p75 seulement conduit à une réduction de la quantité du disaccharide-dipeptide. Ce disaccharide dipeptide est un ligand des récepteurs Nod2. Les deux mutants obtenus par délétion de Lc-p75 ou bien par délétion des deux endopeptidases ont été comparés avec la souche sauvage BL23 pour leur capacité à activer in vitro des cellules dendritiques humaines dérivées de monocytes sanguins. Suite à l’activation des cellules dendritiques par les souches de L. casei, quatre cytokines pro-inflammatoires, les interleukines IL-6, IL-8, IL-12 et le TNF- ont été produites. La quantité de chaque cytokine sécrétée en réponse aux mutants simple Lc-p75 et double Lc-p75/Lc-p45 était diminuée par rapport à celle induite par la souche sauvage L. casei BL23.En conclusion, L. casei BL23 est doté d’un équipement complexe en PGHs. Les PGHs caractérisées au cours de ce travail présentent des caractéristiques uniques et jouent un rôle important dans la division des bactéries ainsi que dans le maintien de leur morphologie. Nos résultats indiquent que la souche sauvage de L. casei Bl23 et les mutants dérivés obtenus par inactivation d’enzymes à activité endopeptidase, qui diffèrent à la fois au niveau de leur contenu enzymatique ainsi qu’au niveau de la structure de leur PG, ont des effets différents sur les cellules dendritiques humaines, avec un caractère anti-inflammatoire plus élevé pour les mutants / Peptidoglycan (PG) is the major component of the Gram-positive bacteria cell wall. It ensures bacterial cell shape and integrity. PG or PG-derived fragments have been shown to stimulate the host innate immune system, through Nod-2 receptors. In this work, we studied the influence of PG metabolism on immunomodulatory properties of Lactobacillus casei BL23, mainly its ability to modulate the response of human dendritic cells (DCs).We have first characterized the main peptidoglycan hydrolases (PGHs) of L. casei BL23. In silico search revealed that L. casei BL23 has a rather complex PGH complement including thirteen predicted PGHs with various catalytic domains. Proteomic analysis of bacterial cell wall extracts revealed the expression of seven of them during bacterial growth. We characterized four of them in details. Lc-p75 is the major PGH with a γ-D-glutamyl-L-lysyl-endopeptidase specificity and is responsible for daughter cell separation. Lc-p75 associated to the cell wall localizes at the cell septa. It is also one of the major secreted proteins of L. casei found in culture supernatant. Besides, we showed that L. casei Lc-p75 is a glycosylated protein. Lc-p40 is a PGH with a CHAP-domain endowed with endopeptidase hydrolytic specificity toward peptidoglycan cross-bridges and appears to localize on lateral cell wall. Lc-p45 is a second γ-D-glutamyl-L-lysyl endopeptidase with a role in cell shape maintenance. We further demonstrated that two prophage endolysins Lc-Lys and Lc-Lys2, encoded in L. casei BL23 genome, share a common novel type peptidoglycan-binding domain that recognizes specifically D-Asn cross-bridge, present in L. casei BL23 peptidoglycan.Deletion of the two endopeptidases, Lc-75 and Lc-p45, resulted in a complete loss ofdisaccharide-dipeptide, which is a ligand of Nod-2 receptor, in the muropeptide structure of L. casei BL23, whereas deletion of Lc-p75 gene led only to a reduction of disaccharide dipeptide. The two PGH-mutants, obtained by deletion of Lc-p75 gene alone or both Lc-p75 and Lc-p45 endopeptidase genes were compared with wild type L. casei BL23 for their capacity to stimulate signaling pathways in vitro in DCs derived from human monocytes. As a consequence of DC activation by L. casei strains, four pro-inflammatory cytokines IL-6, IL-8, IL-12 and TNF-α were produced. The concentrations of secreted cytokines in response to the single Lc-p75 and Lc-p75/p45 double mutant were lower than those induced by wild type L. casei BL23.In conclusion, L. casei BL23 has a complex PGH complement. The PGHs described in this work present unique features and play important role in cell division and morphology of L. casei. Our results indicate that wild type L. casei and endopeptidase-negative mutants, which differ in their PGH content and in their PG structure, have distinct effects on human DCs, with a higher anti-inflammatory character of the endopeptidase-negative mutants.

<p>Bacterial outer membrane vesicles (OMVs) are spherical buds of the outer membrane (OM) containing periplasmic lumenal components. OMVs have been demonstrated to play a critical part in the transmission of virulence factors, immunologically active compounds, and bacterial survival, however vesiculation also appears to be a ubiquitous physiological process for Gram-negative bacteria. Despite their characterized biological roles, especially for pathogens, very little is known about their importance for the originating organism as well as regulation and mechanism of production. Only when we have established their biogenesis can we fully uncover their roles in pathogenesis and bacterial physiology. The overall goal of this research was to characterize bacterial mutants which display altered vesiculation phenotypes using genetic and biochemical techniques, and thereby begin to elucidate the mechanism of vesicle production and regulation. One part of this work elucidated a synthetic genetic growth defect for a strain with reduced OMV production (&#916;nlpA, inner membrane lipoprotein with a minor role in methionine transport) and envelope stress (&#916;degP, dual function periplasmic chaperone/ protease responsible for managing proteinaceous waste). This research showed that the growth defect of &#916;nlpA&#916;degP correlated with reduced OMV production with respect to the hyprevesiculator &#916;degP and the accumulation of protein in the periplasm and DegP substrates in the lumen of OMVs. We further demonstrated that OMVs do not solely act as a stress response pathway to rid the periplasm of otherwise damaging misfolded protein but also of accumulated peptidoglycan (PG) fragments and lipopolysaccharide (LPS), elucidating OMVs as a general stress response pathway critical for bacterial well-being. The second part of this work, focused on the role of PG structure, turnover and covalent crosslinks to the OM in vesiculation. We established a direct link between PG degradation and vesiculation: Mutations in the OM lipoprotein nlpI had been previously established as a very strong hypervesiculation phenotype. In the literature NlpI had been associated with another OM lipoprotein, Spr that was recently identified as a PG hydrolase. The data presented here suggest that NlpI acts as a negative regulator of Spr and that the &#916;nlpI hypervesiculation phenotype is a result of rampantly degraded PG by Spr. Additionally, we found that changes in PG structure and turnover correlate with altered vesiculation levels, as well as non-canonical D-amino acids, which are secreted by numerous bacteria on the onset of stationary phase, being a natural factor to increase OMV production. Furthermore, we discovered an inverse relationship between the concentration of Lpp-mediated, covalent crosslinks and the level of OMV production under conditions of modulated PG metabolism and structure. In contrast, situations that lead to periplasmic accumulation (protein, PG fragments, and LPS) and consequent hypervesiculation the overall OM-PG crosslink concentration appears to be unchanged. Form this work, we conclude that multiple pathways lead to OMV production: Lpp concentration-dependent and bulk driven, Lpp concentration-independent.</p> / Dissertation

La résistance bactérienne aux antibiotiques est un problème majeur de santé publique. Un moyen de la combattre est de viser des cibles non encore exploitées pour retarder l’apparition de la résistance. Dans ce contexte, nous avons entrepris la caractérisation sur les plans biochimique et structural de l’enzyme MraY, une protéine intégrale de membrane, membre d’une famille de transférases membranaires. MraY catalyse la première étape membranaire de la biosynthèse du peptidoglycane bactérien à savoir, le transfert du motif N-acétylmuramoyl-pentapeptide du précurseur cytoplasmique UDP-MurNAc-pentapeptide sur le transporteur membranaire, l’undécaprényl-phosphate aboutissant à la formation du lipide I. Aucune structure 3D de cette enzyme n’est disponible actuellement et aucun antibiotique en utilisation clinique ne la cible. D’une part nous avons entrepris la caractérisation structurale de cette enzyme par des approches de biophysique. Des essais de cristallisation 2D dans des systèmes membranaires ont permis d’observer au microscope électronique des dimères de MraY (taille de 70Å/50Å). Des expériences de diffusion des rayons X (SAXS) montrent un rayon de giration d’environ 42Å. Les résultats issus des expériences de SAXS ont été combinés à des approches de modélisation afin déterminer l’état d’oligomérisation de cette protéine en présence de détergents. Enfin, en vue de faciliter la cristallogenèse 3D, des chimères de MraY en fusion avec des protéines hydrosolubles de structure 3D résolues (mCherry et GFP) ont été construites. Des essais de cristallisation de la protéine seule et des chimères construites ont été effectués.D’autre part, nous avons élucidé le mécanisme catalytique de l’enzyme MraY et de son paralogue WecA. Au cours de ma thèse, j’ai pu montrer que cette famille de transférase membranaire présente un mécanisme catalytique commun qui procède en une seule étape par attaque directe d’un oxyanion du substrat lipidique, préalablement déprotoné par un résidu aspartate invariant, sur le phophate Beta du substrat nucléotidique. Cela conduit à la formation du produit lipidique et libération de l’UMP. / The growing emergence of multiresistance of pathogenic bacteria to currently used antibiotics is a major public health problem that requires the development of new therapeutic compounds and the identification and exploitation of novel targets. In this context, we undertook the biochemical and structural characterization of MraY enzyme, an integral membrane protein, member of the polyprenyl-phosphate N-acetylhexosamine 1-phosphate transferase superfamily. The MraY transferase catalyzes the first membrane step of bacterial cell wall peptidoglycan biosynthesis, namely the transfer of the N-acetylmuramoyl-pentapeptide moiety of the cytoplasmic precursor UDP-MurNAc-pentapeptide to the membrane transporter undecaprenyl phosphate, yielding C55-PP-MurNAc-pentapeptide (lipid I). To date, no crystal structure has been reported for this enzyme. On the one hand we have undertaken the structural characterization of this enzyme by differents biophysical approaches. Electron microscopic images after two-dimensional crystallization of the protein displayed a dimeric organisation of the MraY enzyme (size 70Å/50Å). Small X-ray scattering (SAXS) experiments have shown a radius of gyration of about 42A. The results of SAXS experiments were combined with modeling approaches to determine the oligomerization state of the protein in the presence of detergents. Finally, in order to facilitate 3D crystallisation of MraY, fusion proteins of MraY and mCherry/GFP were constructed. Crystallization trials of MraY alone and the constructed chimeras were made. On the other hand, we have elucidated the catalytic mechanism of the MraY transferase and its paralog WecA. In this study, we have shown that this family of membrane transferases has a common catalytic mechanism that proceeds by a single step displacement. During this “one-step” mechanism, the oxyanion of the poly-prenyl phosphate attacks the β phosphate of the nucleotide substrate, leading to the formation of lipid product and the liberation of UMP.

C. difficile est une bactérie anaérobie du tube digestif reconnue comme l’agent responsable de 25% des diarrhées nosocomiales post antibiotiques et de la plupart des colites pseudomembraneuses. Le métronidazole et la vancomycine sont les traitements de référence pour les infections liées à ce pathogène. A ce jour, aucune souche de C. difficile résistante à la vancomycine n’a été rapportée. Paradoxalement, 85 % des souches de cette espèce hébergent dans leur génome un groupe de gènes cryptiques vanGCd homologue à l’opéron de résistance à la vancomycine vanG de E. faecalis. Dans ce travail, nous avons étudié la capacité des gènes vanGCd à conférer la résistance à la vancomycine. L’étude transcriptionnelle a révélé que le groupe de gènes vanGCd est transcrit en deux opérons (i) l’un de régulation exprimé de façon constitutive et (ii) l’autre de résistance, inductible par la vancomycine. L’analyse enzymatique de VanGCd, VanXYCd et VanTCd in vitro a montré que ces protéines possèdent respectivement des activités ligase, D,D-peptidase et racémase nécessaires au pontage D-Ala-D-Ser et à l’hydrolyse des dipeptides naturels D-Ala-D-Ala. L’analyse des précurseurs du peptidoglycane par spectrométrie de masse en présence de vancomycine a permis l’identification de l’UDP-MurNAc-pentapeptide [Ser] témoin de l’activité in vivo de ces protéines. L’opéron de résistance vanGcd cloné chez E. coli NR698 sensible à la vancomycine exerce un très faible effet sur la CMI de cet antibiotique. En revanche, le clonage chez E. faecalis n’a pas été obtenu en raison de mutations spontanées. Pour ces raisons, nous avons introduit l’opéron vanC de E. gallinarum chez C. difficile et observé une faible augmentation de la CMI de la vancomycine. Ce résultat indique que l’expression de la résistance à cet antibiotique chez C. difficile implique des facteurs intrinsèques liés à la bactérie. Nous avons observé que la ligase MurF possède une meilleure affinité vis-à-vis du dipeptide D-Ala-D-Ala que pour le dipeptide D-Ala-D-Ser ce qui pourrait impacter la synthèse de l’UDP-MurNAc-pentapeptide[Ser]. Par ailleurs, nous avons observé qu’en présence de vancomycine, les précurseurs du peptidoglycane sont amidés. Cette modification affecte également le peptidoglycane mature. Toutefois, le rôle physiologique de l’amidation chez C. difficile reste à élucider. En conclusion, (i) l’absence d’expression de la résistance à la vancomycine de C. difficile est multifactorielle, (ii) le groupe de gènes vanGCd a peu de potentiel pour contribuer à l’émergence de la résistance à la vancomycine. / Clostridium difficile is a major enteric pathogen responsible for 25% of post antibiotics diarrhea and most cases of pseudomembranous colitis. Metronidazole and vancomycin are the standard treatments for infected patients. vanGCd, a cryptic gene cluster highly homologous to the vanG gene cluster of Enterococcus faecalis is largely spread in Clostridium difficile. Since emergence of vancomycin resistance would have dramatic clinical consequences, we have evaluated the capacity of the vanGCd cluster to confer vancomycine resistance. We showed that expression of vanGCd is inducible by vancomycin and that VanGCd, VanXYCd and VanTCd are functional, exhibiting D-Ala:D-Ser ligase, D,D-dipeptidase and D-Ser racemase activities, respectively. In other bacteria, these enzymes are sufficient to promote vancomycin resistance. Trans-complementation of C. difficile with the vanC resistance operon of Enterococcus gallinarum faintly impacted the MIC of vancomycin, but did not promote vancomycin resistance in C. difficile. Sub-lethal concentration of vancomycin led to production of UDP-MurNAc-pentapeptide [D-Ser], suggesting that the vanGCd gene cluster is able to modify the peptidoglycan precursors. Our results indicated amidation of UDP-MurNAc-tetrapeptide, UDP-MurNAc-pentapeptide[D-Ala] and UDP-MurNAc-pentapeptide[D-Ser]. This modification is passed on the mature peptidoglycan where a muropeptide Tetra-Tetra is amidated on the meso-diaminopimelic acid. We also demostrated that the ligase MurF has a better affinity for the D-Ala-D-Ala dipeptide than for the D-Ala-D-Ser, which could impact on the synthesis of UDP-MurNAc-pentapeptide [Ser].In conclusion, the lack of vancomycin resistance expression may be due to several factors that, in combinination with a gene cluster conferring a low level vancomycin resistancemay prevent the emergence of vancomycin resistance based on D-Ala- D-Ser modification of peptidoglycane precursors in C. difficile.