Antoine Gedeon

DNA gyrase is an essential bacterial enzyme and a clinically validated target for the treatment of tuberculosis. However, the discovery of new inhibitors remains limited by the many challenges regarding the manipulation on pathogenic mycobacteria. This study validates Corynebacterium glutamicum (Cglu) as a safe, non-pathogenic surrogate for Mycobacterium tuberculosis (Mtb) to investigate DNA gyrase and facilitate the identification of new inhibitors. Using Cglu as a target allows for fast whole-cell screening under safe conditions while ensuring efficient drug uptake. Cglu shares key physiological features with Mtb, including genome size, complex cell wall structure, and a single type I and type II topoisomerase. Structural and functional comparisons emphasize the similarity of Cglu and Mtb gyrases, which share 70% sequence identity and show comparable catalytic properties and responsiveness to known inhibitors. Thus, the cryo-EM structure of the Cglu gyrase-DNA complex at 3.2 Å resolution reveals highly conserved drug-binding pockets for known anti-gyrase inhibitors and the genetic depletion of gyrA or gyrB in Cglu causes severe growth and morphological defects, mirroring the effects of chemical inhibition and allowing to link gyrase function to cellular phenotypes. Comparative imaging of different inhibitor classes (fluoroquinolones, aminocoumarins, NBTIs) uncovers distinct morphological signatures that reflect the mode of action of each compound. Finally, cross-species complementation confirms functional conservation but also highlights subtle structural differences affecting efficiency. Together, these findings establish Cglu as a robust and biosafe model for dissecting gyrase function, visualizing DNA topology dynamics, and accelerating the discovery of gyrase-targeting antimicrobials. More generally, our studies demonstrate the feasibility of using Cglu as a cell-based screening platform to discover new anti-tuberculous compounds targeting conserved mechanisms, not only for validated TB drug targets such as DNA gyrase but also for new, yet to be identified, targets.

La biosynthèse des nucléotides puriques, regroupant la voie de sauvetage et la voie de novo, est essentielle chez tous les organismes vivants. Plusieurs modes de régulation coexistent pour assurer le maintien des concentrations physiologiques en nucléotides. Dans le cadre de ma thèse, deux niveaux de régulation de cette voie chez E. coli ont été étudiés. Premièrement, en utilisant une technique de double hybride bactérien (BACTH), les interactions binaires protéine-protéine des 14 enzymes de la voie de novo chez E. coli ont été cartographiées, mettant en évidence l'existence d'un assemblage supramoléculaire de ces enzymes. L'organisation de ce complexe apparaît différente de celle de son homologue eucaryote connu sous le nom de purinosome, avec notamment PurK, le hub du complexe chez E. coli et sans équivalent chez les eucaryotes. De plus un mutant d'E. coli du gène purK dont certaines des interactions protéine-protéine sont altérées a été créé et des tests de compétition entre les souches sauvage et mutante indiquent un avantage de fitness de la souche sauvage par rapport à la souche mutante, suggérant un rôle physiologique de la formation du complexe. Le deuxième volet de la thèse portait sur la compréhension de la régulation allostérique d'une enzyme-clé de la voie, l'inosine 5'-monophosphate déshydrogénase (IMPDH). Cette enzyme, catalysant la conversion dépendante du NAD de l'IMP en XMP, est formée de deux domaines structuraux, le domaine catalytique (DC) et le domaine Bateman (DB). Les études précédentes du laboratoire sur différentes IMPDHs bactériennes avaient conduit à la définition de deux classes d'IMPDHs, qui présentent des différences cinétiques et oligomériques suite à la fixation du MgATP sur le domaine Bateman. Pour mieux comprendre le rôle du domaine Bateman à l'échelle moléculaire, différents variants ont été caractérisés par des études biochimiques, biophysiques et structurales : des variants dont le domaine Bateman a été supprimé, et des protéines chimères issues de l'échange du domaine Bateman d'une classe par celui de l'autre classe. Tous les variants générés sont catalytiquement actifs. Plus surprenant encore, le transfert du domaine Bateman d'une classe I (IMPDH de P. aeruginosa, IMPDHpa) dans les domaines catalytiques d'une classe II (IMPDH de B. thailandensis, IMPDHbt et IMPDH d'E. coli, IMPDHec) conduit à un fort impact sur les propriétés catalytiques : les deux chimères (DCec DBpa et DCbt DBpa) sont activées par le MgATP, et leurs paramètres catalytiques sont similaires à ceux de l'enzyme sauvage, dont le domaine Bateman est dérivé, i.e. l'IMPDHpa. Il en est de même pour la troisième chimère (DCpa DBec), où le domaine Bateman d'une classe II (IMPDHec) a été inséré dans le domaine catalytique d'une classe I (IMPDHpa). Pour cette chimère, les paramètres cinétiques restent inchangés en présence de MgATP, mais l'état oligomérique est modifié vers une espèce octamérique. Enfin, des tests de compétition bactériennes d'une souche délétée du gène codant l'IMPDH et complémentée par différentes constructions thermosensibles codant les formes mutantes des IMPDHs sont en faveur d'un rôle physiologique du domaine Bateman.

The mfp conservon of Mycobacterium tuberculosis has been associated with fluoroquinolone resistance and encodes five conserved proteins, including the small GTPase MfpB and its regulatory partner MfpD. In this study, we combined phylogenetic, structural, and biophysical approaches to define the molecular basis of MfpD function. MfpD adopts a Roadblock/LC7-like alpha/beta fold and forms a stable dimer in solution, with hydrophobic alpha2-helix interactions stabilizing the interface. Additional biophysical analyses and AlphaFold3 modeling suggest that MfpD may promote GTP hydrolysis by MfpB through a noncanonical Switch I-dependent mechanism. These findings establish the first structural framework for MfpD-MfpB interactions, building on previously identified in vitro catalytic properties and proposing new insights into MfpD non-catalytic pathogenesis activity of MfpD in macrophages.

Most cellular processes are carried out by multiprotein assemblies. Although various molecular tools exist to visualize binary protein interactions in live cells, the visualization of multiprotein complexes remains a challenge. Here, we report the engineering of a complementation-based approach allowing one to visualize the interaction of three proteins through effective proximity-induced complementation of three fragments of pFAST, a chemogenetic fluorescent reporter that binds and stabilizes the fluorescent state of fluorogenic chromophores (so-called fluorogens). This tripartite-split-pFAST allowed the observation of dynamic ternary protein complexes in the cytosol, at the plasma membrane, in the nucleus and at the junction of multiple organelles, opening prospects to study the role and function of multiprotein complexes in live cells and in various biologically relevant contexts.

Despite a huge consumption of echinocandins, the emergence of resistance in Candida spp has remained overall limited. Here, we depicted the epidemiology of Candida spp in our center face to the echinocandins’ consumption. We postulate new hypotheses that may explain the shaping of candines resistance in the clinical setting. Epidemiology of Candida infections and echinocandin consumption were evaluated in our center over 12 years (2006-2018). Glucan synthase genes ( fks1, fks2) were sequenced. The in vitro fitness was assessed for couple of isogenic strains, of which one was resistant. Finally, the modelling of Candida FKS proteins was realized. Despite a three-fold increase in echinocandin consumption, no significant emergence of resistance was observed. In Candida albicans, fks1 mutations affect the three-dimensional conformation of the glucan synthase, which may result in altered export of the glucan chain, which could explain the observed reduced fitness. In contrast, cross-complementation between FKS1 and FKS2 in Nakaseomyces glabratus might circumvent the negative impact on fitness. Conclusions: Our results support the fact that N . glabratus is more prone to the spread of echinocandin-resistant strains. The link between mutations in the glucan synthase and fitness of Candida strains might explain the difference in the species-specific emergence of echinocandins-resistant strains.

De novo purine nucleotide biosynthesis (DNPNB) consists of sequential reactions that are majorly conserved in living organisms. Several regulation events take place to maintain physiological concentrations of adenylate and guanylate nucleotides in cells and to fine-tune the production of purine nucleotides in response to changing cellular demands. Recent years have seen a renewed interest in the DNPNB enzymes, with some being highlighted as promising targets for therapeutic molecules. Herein, a review of two newly revealed modes of regulation of the DNPNB pathway has been carried out: i) the unprecedent allosteric regulation of one of the limiting enzymes of the pathway named inosine 5′-monophosphate dehydrogenase (IMPDH), and ii) the supramolecular assembly of DNPNB enzymes. Moreover, recent advances that revealed the therapeutic potential of DNPNB enzymes in bacteria could open the road for the pharmacological development of novel antibiotics.

Most cellular processes are carried out by multiprotein assemblies. Although various molecular tools exist to visualize binary protein interactions in live cells, the visualization of multiprotein complexes remains a challenge. Here, we report the engineering of a complementation-based approach allowing one to visualize the interaction of three proteins through effective proximity-induced complementation of three fragments of pFAST, a chemogenetic fluorescent reporter that binds and stabilizes the fluorescent state of fluorogenic chromophores (so-called fluorogens). This tripartite-split-pFAST allowed the observation of dynamic ternary protein complexes in the cytosol, at the plasma membrane, in the nucleus and at the junction of multiple organelles, opening great prospects to study the role and function of multiprotein complexes in live cells and in various biologically relevant contexts.

Inosine 5′‐monophosphate (IMP) dehydrogenase (IMPDH) is an ubiquitous enzyme that catalyzes the NAD⁺‐dependent oxidation of inosine 5′‐monophosphate into xanthosine 5′‐monophosphate. This enzyme is formed of two distinct domains, a core domain where the catalytic reaction occurs, and a less‐conserved Bateman domain. Our previous studies gave rise to the classification of bacterial IMPDHs into two classes, according to their oligomeric and kinetic properties. MgATP is a common effector but cause to different effects when it binds within the Bateman domain: it is either an allosteric activator for Class I IMPDHs or a modulator of the oligomeric state for Class II IMPDHs. To get insight into the role of the Bateman domain in the dissimilar properties of the two classes, deleted variants of the Bateman domain and chimeras issued from the interchange of the Bateman domain between the three selected IMPDHs have been generated and characterized using an integrative structural biology approach. Biochemical, biophysical, structural, and physiological studies of these variants unveil the Bateman domain as being the carrier of the molecular behaviors of both classes.

In human cells, de novo purine nucleotide biosynthesis is known to be regulated through the formation of a metabolon called purinosome. Here, we employed a bacterial two‐hybrid approach to characterize the protein–protein interactions network among the corresponding enzymes of Escherichia coli. Our study revealed a dense network of binary interactions that connect most purine nucleotide biosynthesis enzymes. Notably, PurK, an exclusive prokaryotic enzyme, appears as one of the central hubs of this network. We further showed that modifications in PurK, which disrupted several interactions in the network, affected the purine nucleotide pools and altered the bacterial fitness. Our data suggest that the bacterial de novo purine nucleotide biosynthesis enzymes can assemble in a supramolecular complex and that proper interactions among the components of this complex can contribute to bacterial fitness.