The presence of cholesterol is critical in defining a dynamic lateral structure in pulmonary <p>surfactant membranes, including the segregation of fluid-ordered and fluid-disordered phases. <p>However, an excess of cholesterol has been associated with impaired surface activity both in <p>surfactant models and in surfactant from injured lungs. It has also been reported that surfactant <p>protein SP-C interacts with cholesterol in lipid/protein interfacial films. In the present study, we <p>have analyzed the effect of SP-C on the thermodynamic properties of phospholipid membranes <p>containing cholesterol and on the ability of lipid/protein complexes containing surfactant <p>proteins and cholesterol to form and re-spread interfacial films capable of producing very low <p>surface tensions upon repetitive compression-expansion cycling. We have also analyzed the effect of cholesterol on the <p>structure, orientation and dynamic properties of SP-C embedded in physiologically relevant <p>model membranes. <p><p> / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

The biological membrane surrounding the living cell provides a sealed barrier that tightly regulates the interactions with the outside environment. A large number of integral membrane proteins mediate these interactions and are involved in a wide variety of biological processes. An increasing number of studies have led to the conclusion that lipids provide more than a hydrophobic solvent for membrane proteins, and that interactions between lipids and proteins are required to allow protein function. ABC transporters are one of the most important family of membrane proteins. However, the importance of their lipidic environment is largely unknown. Only a few studies showed that their activity was dependent on the lipidic composition of the surrounding bilayer. The bacterial ABC transporter HorA was used as a model to probe the influence of the lipidic environment on that class of membrane proteins.<p><p> HorA is a multidrug transporter expressed in Lactobacillus brevis, a Gram-positive beer spoilage bacterium. It turned out that phosphatidylethanolamine (PE) was indispensable to maintain both the activity and the structural integrity of HorA.<p> Surprisingly, replacement of PE by the chemically related PC (phosphatidylcholine) did not led to the suppression of HorA activity, but to an unexpected phenotype. Whereas the cytoplasmic domains of HorA were still able to hydrolyze ATP, the membrane parts of the transporter were unable to use that energy to mediate substrate transport. Using several biophysical methods particularly adapted to the study of reconstituted systems, we showed that the structure of HorA is strongly altered by this lipid replacement. In particular, the structural organization of the transmembrane domains of the protein is strongly affected.<p> / Doctorat en Sciences agronomiques et ingénierie biologique / info:eu-repo/semantics/nonPublished

In the past fifteen years, ubiquitin has emerged as a central regulator of membrane protein trafficking. In this context, covalent attachment of this small protein to lysine residues of cargo proteins, a reversible modification termed ubiquitylation, provides a signal for their targeting to the vacuolar/lysosomal lumen where they are degraded, both in yeast and higher eukaryotes. Ubiquitylation is also used as a means of controlling the function of specific proteins in several trafficking machineries. The role of lipids - and in particular of membrane domains named lipid rafts - in controlling the intracellular trafficking of membrane proteins has also been the subject of intense investigation in recent years.<p>One of the membrane proteins of the yeast Saccharomyces cerevisiae whose intracellular trafficking has been extensively studied is the general amino acid permease Gap1. Yet some aspects of the function of ubiquitin in the nitrogen-dependent control of this protein remain controversial. Moreover, the potential role of lipid rafts in regulating the functional properties and traffic of the Gap1 permease had not been investigated before this thesis work. <p>The first part of our work readdresses the role of Gap1 ubiquitylation, and more precisely of the modification of the permease with polyubiquitin chains linked through the lysine 63 of ubiquitin, in controlling the fate of this protein in the secretory pathway. Our observations indicate that nitrogen-induced ubiquitylation of newly synthesised Gap1 occurs in the trans-Golgi complex. However, contrary to the generally accepted view, this modification is not necessary for the permease to exit this compartment en route to the endosome but only for its subsequent targeting to the vacuolar lumen via the multivesicular body (MVB) pathway. Our results also provide evidence that K63-linked polyubiquitylation is important mostly at the late endosomal level, for proper sorting of Gap1 into the MVB pathway, whether the permease comes from the cell surface by endocytosis or directly from the secretory pathway. <p>In the second part of this work, we present a set of data providing novel insights into the controversial question of the exact nature of lipid rafts in yeast. We first showed that the Gap1 permease is associated with detergent-resistant membranes (DRMs) - the proposed biochemical equivalent of lipid rafts - when it is located at the cell surface. Our data further suggest that this may be true for most if not all yeast plasma membrane proteins. Moreover, we found that Gap1 production must be coupled to de novo synthesis of sphingolipids (SLs), major constituents of rafts, in order for the newly synthesised permease to be correctly folded, active, associated with DRMs, and stable at the cell surface. We propose a model where Gap1 would associate with newly synthesised SLs during its biogenesis and/or secretion, this association shaping the permease into its native conformation and ensuring its incorporation and stabilisation in specific lipid domains at the plasma membrane. Failure of Gap1 to acquire this lipidic microenvironment in turns leads to its ubiquitin-dependent degradation by a quality-control mechanism. This model might be valid for many other plasma membrane proteins and might account for their lateral distribution between distinct membrane domains. <p> / Doctorat en Sciences / info:eu-repo/semantics/nonPublished

Degradation of membrane proteins in the vacuole/lysosome is dependent on their prior sorting into the multivesicular body (MVB) pathway. This sorting process involves incorporation of proteins into vesicles that are formed by budding of the limiting membrane of the endosome into the lumen of the organelle. The MVB sorting process on the whole is highly conserved from yeast to human, and depends on the Vps27/Hrs, ESCRT-I, -II, and -III protein complexes functioning sequentially on the endosomal membrane, as well as on additional factors, such as the ubiquitinating enzyme Rsp5/Nedd4. It has now been established that ubiquitin serves as a sorting signal for many cargoes into the MVB pathway. <p>In this thesis work, we provide evidence that Bro1 is not required for protein ubiquitination or early steps of endocytosis, but functions at the late endosome level as an integral component of the MVB pathway. Similarly to its human homologue Alix, Bro1 interacts with components of the ESCRT-I and ESCRT-III complexes. The putative role of Bro1/Alix in bridging an interaction between ESCRT-I and –III might be important to strengthen an association of these protein complexes to allow efficient sorting of cargo proteins. Deficiency in Bro1 results in recycling of the endocytosed Gap1 permease back to the plasma membrane, a process coupled to deubiquitination of the permease. This recycling is a non-classical phenotype for cells impaired in MVB pathway thus suggesting Bro1 to have a particular role in this sorting process. Furthermore, the conserved C-terminal proline-rich domain (PRD) of Bro1 is specifically important for MVB sorting of cargo proteins that are subject to ubiquitination. We show Bro1 (via its PRD) to play a highly important role in recruitment of the deubiquitinating enzyme Doa4 to the endosome. Consistent with this, Bro1 is required for deubiquitination of cargo proteins, a step occurring just before cargo incorporation into the endosomal vesicles, and similarly to Doa4, for ubiquitin recycling. In contrast to previous interpretations, we show that Doa4 has a direct role in sorting of ubiquitinated cargo proteins into the MVB pathway. We propose that Doa4 – via its association to Bro1 - achieves this role by catalyzing deubiquitination of cargo proteins and/or some components of the MVB sorting machinery. We further show Bro1 to interact with the ubiquitin ligase Rsp5, which, in addition to being required for cargo protein ubiquitination at the plasma membrane, apparently contributes to multiple steps of endocytosis and MVB sorting. Also the Bro1-Rsp5 interaction is dependent on the C-terminal PRD region of Bro1. We propose that this interaction is conserved. <p>A role for ubiquitin in regulation of the MVB sorting machinery is emerging: the function of factors recognizing and sorting ubiquitinated cargo proteins in the MVB pathway is suggested to be coupled to their cycling between ubiquitinated and deubiquitinated stages. A growing body of evidence indicates that ubiquitin ligases of the Rsp5/Nedd4 family play a central role in this regulation. We speculate the Bro1/Alix protein, through its ability to simultaneously interact with factors of the MVB sorting machinery and with ubiquitinating and deubiquitinating enzymes to play a central role in the successive rounds of ubiquitination and deubiquitination of specific factors along the MVB pathway. <p> / Doctorat en sciences, Spécialisation biologie moléculaire / info:eu-repo/semantics/nonPublished

La virulence de Plasmodium falciparum est fonction de sa capacité à séquestrer les érythrocytes infectés (iEs) dans les organes profonds de l’hôte. Elle est liée à la mise en place d’un trafic protéique conséquent au niveau membranaire et sub-membranaire de l’iE. Cet adressage protéique aboutit à l’expression d’antigènes variant de surface, dont P. falciparum Erythrocyte Membrane Protein-1, PfEMP-1. Cet adhésine présente une affinité pour divers récepteurs de l'hôte impliqués dans la physiopathologie des formes graves, telles que le paludisme associé à la grossesse (PAM) et le paludisme cérébral (CM). La diversité de liaison de PfEMP-1 est associée à la variabilité de séquence de son domaine extracellulaire. Ce domaine est constitué d’une succession de domaines DBL et CIDR en nombre variable. PfEMP-1 est codée par les gènes var classifiés en groupes Ups A, B, C, E et B/A, B/C. Des régions de PfEMP-1 constituées d’une succession établie de 2 à 4 domaines, nommées Domain Cassette, « DC » sont retrouvées dans différents isolats. L’obtention de données exhaustives par LC-MS/MS a permis d’établir un comparatif du protéome membranaire et hypothétique à partir des échantillons PAM et CM en comparaison à des accès simples de paludisme (UM). L’identification protéique des PfEMP-1 a montré la singularité de l’expression d’une PfEMP-1 particulière, VAR2CSA chez les PAM. Ces résultats corroborent les analyses des transcrits du gène var2csa comme surexprimé chez les PAM, transcrit qu’on retrouve également dans les infections de début et de fin de grossesse. Les PfEMP-1 identifiés au sein des CM étaient significativement associés aux groupes Ups A et B/A et la plupart des peptides identifiés étaient associés à la cassette DC8 dont le transcrit était également surexprimé. L’analyse de filter-based feature selection a permis d’identifier un sous ensemble de 13 et 14 protéines assignées comme protéines membranaires ou hypothétiques, prédictives des formes de paludisme PAM et CM, respectivement. Parmi ces protéines surexprimées chez les PAM, la présence de VAR2CSA valide l’approche d’analyse. PFI1785w a été identifiée et associée au PAM, quatre autres protéines apparaissent prédictives de cette forme clinique PFB0115w, PFF0325c, PFA_0410w et PF14_0018. Pour les CM, on distingue les protéines PfEMP-2 et antigen-332 comme spécifiques de cette forme clinique. Les autres protéines qui émergent de ce groupe sont des protéines de fonction inconnue, dont trois sont assignées comme des protéines exportées. L’obtention et l’analyse en LC-MS/MS d’extraits protéiques issus d’isolats de terrain font l’originalité de ce travail et permettent la corrélation entre les données cliniques et biologiques. Cette étude confirme les données de transcrits sur la famille des gènes var et suggère de nouvelles intéractions protéiques dans le cadre du paludisme associé à la grossesse et cérébral. / Plasmodium falciparum is responsible for severe malaria (cerebral malaria, CM and pregnancy associated malaria, PAM). During the intra-erythrocytic maturation of P. falciparum, parasite-derived proteins are expressed, exported and presented at the surface of the infected erythrocyte (iE) membrane. These include Plasmodium falciparum erythrocyte membrane protein-1 (PfEMP-1). PfEMP-1 is a highly polymorphic adherence receptor, variants of which have been assigned to four groups (A-E) based on their sequence homology. Semi-conserved types, defined by tandem runs of specific domains (“domain cassettes” (DC)), are also recognized. The PfEMP-1 type expressed determines the iE adherence phenotype, and is associated with the clinical outcome of infection. Parasite isolates from Beninese children or women presenting with, respectively, CM or PAM were collected along with samples from patients with uncomplicated malaria (UM). We assessed the transcript level of var genes by RT-qPCR and the expression of membrane and hypothetical proteins of Plasmodium by LC-MS/MS. Obtaining LC-MS/MS data enabled a comparison of hypothetical and membrane proteome samples from PAM and CM for comparison with UM samples. The proteomics-based identification of PfEMP-1 showed the expression of a particular PfEMP-1, VAR2CSA, in PAM isolates. These results corroborate the analysis of gene transcripts showing that var2csa is overexpressed in PAM, with transcripts found in isolates from infections both early and late in pregnancy. The PfEMP-1 variants identified in CM samples were predominantly from the Ups groups A and B/A, and most of the peptides identified by LC-MS/MS were associated with the DC8 cassette for which the transcripts were also overexpressed. Analysis using filter-based feature selection identified subsets of 13 and 14 proteins, assigned either as hypothetical or membrane proteins that were predictive of PAM and CM syndromes respectively. The presence of VAR2CSA amongst the proteins overexpressed in PAM samples validates the analytical approach. PFI1785w has previously been identified and associated with the PAM, whilst four other proteins, PFB0115w, PFF0325c, PFA_0410w and PF14_0018, appear to be also predictive of the syndrome. PfEMP-2 protein and antigen-332 were found to be specifically expressed in CM samples. Three other proteins, assigned as ‘exported’ but with unknown function, were also associated with CM samples. Obtaining and analyzing LC-MS/MS-derived data from protein extracts of field isolates represents the originality of this work and it allowed the identification of correlations between clinical and biological data. This study confirms the transcriptional data relating to the var gene family and provides evidence of new protein interactions in the context both of malaria associated with pregnancy and of cerebral malaria.

Notre projet vise à déterminer la structure 3D du transporteur BmrA de Bacillus subtilis. La protéine a été purifiée dans six détergents différents. L'utilisation de foscholine 12, a conduit à cristalliser OmpF, une porine de la membrane externe d'E. coli. Nous montrons que les conditions de cristallisation influencent directement l'empilement cristallin d'OmpF. Le protocole de purification de BmrA, optimisé en utilisant du triton X100 à l'extraction puis un mélange β-D-dodecyl maltoside-cholate pour les étapes chromatographiques nous a permis d'obtenir à 4°C des cristaux, pour lesquels nous avons vérifié qu'ils sont constitués de BmrA. Ces cristaux ont permis d'obtenir un jeu complet jusqu'à 7 Å. Ces données de diffraction constituent une avancée significative pour résoudre à court terme la structure 3D de BmrA. Nous avons développé une nouvelle méthode de dosage des détergents qui est basée sur la détermination par spectrométrie de masse de type MALDI du ratio d'isotopes deutérés/ protonés. La méthode a été validée avec la FC12, le DDM, le β-OG, le LMNG, le CHAPS, le cholate et des détergents calix[4]aréniques, en mesurant la concentration de ces détergents dans différentes conditions d'extraction/purification, de concentration, dialyse et gel filtration, de différentes protéines membranaires. Cette méthode nous a permis (i) d'estimer la taille de la ceinture de détergent associée à BmrA et d'autres protéines membranaires (ii) de moduler cette taille en fonction de mélange de détergents et (iii) d'apporter des informations sur le comportement des complexes protéine-détergent / Our project aims to determine the 3D structure of BmrA from Bacillus subtilis. The protein was purified in six different detergents. Using foscholine 12, led to crystallize OmpF, an outer membrane porin of E. coli. We show that the crystallization conditions directly influence the crystal packing of OmpF. The BmrA purification protocol optimized by using Triton X100 at the extraction and a mixture β-D-dodecyl-maltoside cholate for chromatographic steps allowed us to get to 4°C crystals, for which we verified they consist of BmrA. These crystals have yielded full data to 7 Å. These diffraction data are a significant advance in the short term to resolve the 3D structure of BmrA. We have developed a new detergents dosage assay which is based on the determination by MALDI-type mass ratio of deuterated isotopes / protonated. The method was validated with the FC12, the DDM, the β-OG, the LMNG, CHAPS, cholate detergents and calix [4] aréniques by measuring the concentration of these detergents in different conditions of extraction/ purification, concentration, dialysis and gel filtration, of different membrane proteins. This method allowed us (i) to estimate the size of the detergent belt associated to BmrA and other membrane proteins (ii) to modulate this size in terms of the detergent mixture and (iii) to provide information on the behavior of complex protein-detergent

Bilayer embedded proteins are affected by stress. This general affirmation is, in this thesis, embodied by two types of proteins: membrane spandex and voltage-gated sodium channels. In this work, we essentially explore, using methods from physics, the theoretical consequences of ideas drawn from experimental biology. Membrane spandex was postulated to exist and we study the theoretical implications and possible benefits for a cell to have such proteins embedded in its bilayer. There are no specific membrane spandex proteins, rather any protein with a transition involving a large enough area change between two non-conducting states could act as spandex. Bacterial cells have osmovalve channels which open at near-lytic tensions to protect themselves against rupture. Spandex expanding at tensions just below the osmovalves’ opening tension could relieve tension enough as to avoid costly accidental osmovalve opening due to transient bilayer tension excursions. Another possible role for spandex is a tension-damper: spandex could be used to maintain bilayer tension at a fixed level. This would be useful as many bilayer embedded channels are known to be modulated by tension. The Stress/shear experienced in traumatic brain injury cause an immediate (< 2 min) and irreversible TTX-sensitive rise in axonal calcium. In situ, this underlies an untreatable condition, diffuse axonal injury. TTX sensitivity indicates that leaky voltage-gated sodium (Nav) channels mediate the calcium increase. Wang et al. showed that the mammalian adult CNS Nav isoform, Nav1.6, expressed in Xenopus oocytes becomes “leaky” when subjected to bleb-inducing pipette aspiration. This “leaky” condition is caused by a hyperpolarized-shift (left-shift or towards lower potentials, typically 20 mV) of the kinetically coupled processes of activation and inactivation thus effectively degrading a well-confined window conductance into a TTX-sensitive Na leak. We propose experimental protocols to determine whether this left-shift is the result of an all-or-none or graded process and whether persistent Na currents are also left-shifted by trauma. We also use modeling to assess whether left-shifted Nav channel kinetics could lead to Na+ (and hence Ca2+ ) loading of axons and to study saltatory propagation after traumatizing a single node of Ranvier.

La métalloenzyme multimérique membranaire NADPH oxydase du neutrophile (NOX2), est impliquée dans plusieurs fonctions physiologiques vitales incluant la réponse immunitaire, en contribuant fortement à la destruction des pathogènes ou autres envahisseurs du corps humain. Les fonctions physiologiques de NOX2 sont assurées par sa fonction chimique de catalyseur de la production d’anions superoxyde via la réduction monoélectronique du dioxygène à une face de la membrane, simultanément à l’oxydation biélectronique du NADPH à l’autre face de la membrane. L’étude des caractéristiques biochimiques de l’enzyme entière, incluant ses mécanismes d’activation et de régulation, en lien avec l’assemblage macromoléculaire, est réalisée in vitro en utilisant des fractions membranaires de neutrophile (FM), petites vésicules contenant NOX2, ainsi que les protéines cytosoliques recombinantes (p67phox, p47phox et Rac1/2) indispensables à sa fonction, en présence d’une molécule activatrice comme l’acide arachidonique (AA), un acide gras. La technique historiquement privilégiée de mesure de l’activité enzymatique de NOX2 implique la détection des anions superoxyde par une sonde protéique, le Cytochrome c (Cytc). Dans ce système, les anions superoxyde, dont la production est catalysée par NOX2 vers l’intérieur des vésicules de FM, sont détectés à l’extérieur. En corrélation avec la littérature, ces recherches montrent que l’activité enzymatique déterminée via la détection des anions superoxyde par le Cytc est plus faible que lorsqu’elle est déterminée via la mesure de la consommation du NADPH. L’origine du problème inclut potentiellement des contraintes de perméabilité membranaire, de structure de la membrane et des protéines, d’interactions des protéines entre-elles et avec les lipides membranaires, de pertinence de la sonde utilisée et de réactions secondaires. Ces hypothèses ont été testées par différents moyens incluant notamment des mesures de cinétiques globales et de l’activité de NOX2 dans différentes conditions et avec différentes observables (NADPH, Cytc, dioxygène), en présence de détergent ou d’ionophore, en faisant varier la température de mesure, la concentration en Cytc, la concentration en substrat, la concentration en AA ou le temps de préincubation. La présence de réactions secondaires a également été testée par électrochimie. Cette étude montre que la mesure de la production des anions superoxyde est limitée par la perméabilité membranaire et par les réactions secondaires. Il a aussi été mis en exergue que la concentration en Cytc usuelle pour ces mesures n’est pas saturante et de façon inattendue que les FM catalysent intrinsèquement la dismutation du peroxyde d’hydrogène à l’aide d’un composant thermolabile. Il est aujourd’hui très compliqué de mesurer de façon concomitante l’activité de la NADPH oxydase et son assemblage. Le deuxième objectif de cette thèse était donc la mise au point de vésicules géantes unilamellaires (GUV) intégrant NOX2 dans leur membrane, ce qui permettrait d’observer par microscopie de fluorescence l’assemblage du complexe NADPH oxydase et de simultanément mesurer la production des anions superoxyde par électrochimie sous microscope. La formation de GUV possédant des FM (FM-GUV) à leur membrane est un succès mais sans confirmation de l’intégration de NOX2 à la membrane des GUV. L’assemblage des protéines cytosoliques à la face externe de la membrane a été étudié sur GUV et sur FM-GUV, ce qui a permis de montrer que l’ancrage membranaire de ces protéines est possible seulement en présence d’AA et dû de façon prépondérante aux lipides et que NOX2 joue un rôle minoritaire. L’étude des interactions entre les protéines cytosoliques et la face interne de la membrane des GUV reste à optimiser. Il a été possible de détecter en GUV qualitativement une activité de NOX2 par électrochimie et par fluorescence (Amplex-Red), mais ce point reste aussi à optimiser. / The membrane multimeric metalloenzyme NADPH oxidase (NOX2) from neutrophil is implied in several essential physiological functions including the immune response, by strongly contributing to the destruction of pathogens or other invaders of the human body. Physiological functions of NOX2 are fulfilled by its chemical function of catalyst of superoxide anion production via the monoelectron dioxygen reduction on one face of the membrane, simultaneously to the NADPH bielectron oxidation on the other face of the membrane. Studies of biochemical features of the whole enzyme, including its activation and regulation mechanisms linked to the macromolecular assembly, is done in vitro by using neutrophil membrane fractions (MF), which are small vesicles containing NOX2, and by using the recombinant cytosolic proteins (p67phox, p47phox, p40phox and Rac1/2) essential for its function, in presence of an activator molecule such as arachidonic acid (AA), a fatty acid. The historical technics to measure the NOX2 enzyme activity is the superoxide anion detection by a protein probe, the Cytochrome c (Cytc). In this system, NOX2 catalyses the production of superoxyde anions towards the inside of MF vesicles and the superoxide anions are detected outside. In correlation with literature, the present research shows that the enzyme activity determined via the detection of superoxide anions by the Cytc is lower that the activity determined from NADPH consumption measurement. The source of the problem includes potentially constraints of membrane permeability, of membrane and protein structure, of protein-protein and protein-membrane interactions, of the relevance of the probe and of secondary reactions. These hypotheses have been tested by various means including notably global kinetics measurements and NOX2 activity measurements in various conditions with three different observables (NADPH, Cytc, dioxygen), in presence of detergent or ionophore, by varying temperature, Cytc concentration, substrate concentration, AA concentration or still preincubation time. Secondary reactions existence has also been probed by electrochemistry. This study shows that the measurement of the superoxide anion production is limited by membrane permeability and secondary reactions and that the usual Cytc concentration is non-saturating, and unexpectedly that the MF catalyses the disproportionation of hydrogen peroxide by a thermolabile component. It is currently very hard to measure simultaneously the NADPH oxidase activity and the assembly of the whole complex. The second objective of my thesis was consequently to develop giant unilamellar vesicles (GUV) with NOX2 integrated into their membranes. This to be able to observe the complex assembly by fluorescence microscopy and simultaneously to measure the superoxide anion production by electrochemistry under microscope. The development of GUV with MF at the membrane (MF-GUV) has been successful, but without confirmation of NOX2 integration in the GUV membrane. The assembly of cytosolic proteins on the external face of the membrane was studied on GUV and on MF-GUV, leading to the discovery that membrane anchor of these proteins is possible only in presence of AA and is mostly due to lipids, NOX2 playing a minor role. Study of interactions between cytosolic proteins and internal face of the GUV membrane must be optimised. It was possible in GUV to detect qualitatively NOX2 activity by electrochemistry and by fluorescence, (Amplex-Red), but this point should still be optimised.

The Saccharomyces cerevisiae kinase Sat4p has been originally identified as a protein involved in salt tolerance and stabilization of plasma membrane transporters, implicating a cytoplasmic localization. Our study revealed an additional mitochondrial (mt) localization, suggesting a dual function for Sat4p. While no mt related phenotype was observed in the absence of Sat4p, its overexpression resulted in significant changes of a specific mitochondrial subproteome. As shown by a comparative two dimensional difference gel electrophoresis (2D-DIGE) approach combined with mass spectrometry, particularly two groups of proteins were affected: the iron-sulfur containing aconitase-type proteins (Aco1p, Lys4p) and the lipoamide-containing subproteome (Lat1p, Kgd2p and Gcv3p). The lipoylation sites of all three proteins could be assigned by nanoLC-MS/MS to Lys75 (Lat1p), Lys114 (Kgd2p) and Lys102 (Gcv3p), respectively. Sat4p overexpression resulted in accumulation of the delipoylated protein variants and in reduced levels of aconitase-type proteins, accompanied by a decrease in the activities of the respective enzyme complexes. We propose a regulatory role of Sat4p in the late steps of the maturation of a specific subset of mitochondrial iron-sulfur cluster proteins, including Aco1p and lipoate synthase Lip5p. Impairment of the latter enzyme may account for the observed lipoylation defects.

Chez les bactéries, l’undécaprényl-phosphate (C55-P) est utilisé comme transporteur lipidique de sous-unités glycanes à travers la membrane plasmique. Après la synthèse de son précurseur (C55-PP) par UppS, ce dernier doit être déphosphorylé. Le transfert final des sous-unités glycanes sur une molécule acceptrice du côté périplasmique libère également du C55-PP qui va être déphosphorylé pour être recyclé. Quatre C55-PP phosphatases membranaires ont été identifiées chez E. coli : trois enzymes appartenant à la famille des PAP2 (PgpB, YbjG et LpxT) et une protéine appartenant à une nouvelle famille de phosphatases (BacA). Au cours de cette étude, nous avons caractérisé les propriétés biochimiques et la topologie membranaire de BacA. Les conditions optimales de son activité (pH, détergent, cations, température) ont été déterminées et une étroite spécificité de substrats, avec une préférence pour le C55-PP, a été observée. Trois résidus essentiels à son activité, Glu21, Ser27 et Arg174, ont été identifiés par mutagénèse, nous permettant de proposer un mécanisme catalytique basé sur l’attaque nucléophile du C55-PP par un résidu sérine. La topologie membranaire de BacA, déterminée expérimentalement à l’aide de fusions de protéines, n’a validé aucun des modèles in silico précédemment proposés. Ainsi, BacA comprend 7 segments transmembranaires et présente deux larges boucles périplasmiques portant les résidus hautement conservés du site actif. Nos résultats démontrent que toutes les C55-PP phosphatases d’E. coli identifiées à ce jour (BacA et PAP2) catalysent la déphosphorylation du C55-PP du même côté de la membrane plasmique (côté périplasmique), nous interrogeant sur l’identité de l’enzyme catalysant la déphosphorylation du C55-PP synthétisé de novo. LpxT est une PAP2 qui possède une activité kinase assurant le transfert du phosphate β du C55-PP sur une molécule de lipide A, pour former du lipide A-1-PP. Nous avons cartographié par mutagénèse dirigée le site actif de LpxT et mis en évidence l’importance d’une triade catalytique caractéristique des PAP2 (His150, His190, Asp194) et d’autres résidus spécifiques à LpxT et ses proches homologues. L’activité de LpxT est inhibée par un petit peptide, PmrR, dont l’expression est sous contrôle du système à deux composants PmrA-PmrB. Notre étude a montré que cette inhibition intervenait via une interaction directe entre ces deux partenaires. Nous montrons que l’induction du système PmrA-PmrB conduit à une résistance à la polymyxine B (peptide antimicrobien cationique) et à une sensibilité au déoxycholate (composant majeur de la bile) et que la modification catalysée par LpxT produit un effet opposé. La résistance à la polymyxine B est corrélée à la force des signaux induisant le système PmrA-PmrB, mais également le système PhoP-PhoQ et nous avons clairement identifié les signaux nécessaires à cette résistance chez E. coli. / In bacteria, the undecaprenyl-phosphate (C55-P) is used as a lipid carrier of glycans subunits across the plasma membrane. After synthesis of its precursor (C55-PP) by UppS, this latter must be dephosphorylated. The transfer of the glycans subunit onto a final acceptor molecule at the periplasmic side also releases C55-PP that will be dephosphorylated to be recycled. Four C55-PP membrane phosphatases have been identified in E. coli : three enzymes belonging to the PAP2 family (PgpB, YbjG and LpxT) and a protein belonging to a new family of phosphatases (BacA).In this study, we characterized the biochemical properties and membrane topology of BacA. The optimal conditions for its activity (pH, detergent, cation, temperature) were determined and narrow substrate specificity, with a preference for the C55-PP, was observed. Three essential residues to its activity, Glu21, Ser27 and Arg174 were identified by mutagenesis, allowing us to propose a catalytic mechanism based on the nucleophilic attack of the C55-PP by a serine residue. The membrane topology of BacA determined experimentally using protein fusions did not validated previous in silico models. Thus, BacA has 7 transmembrane segments and contains in particular two large periplasmic loops carrying the highly-conserved active site residues. Our results demonstrate that all C55-PP phosphatases of E. coli identified to date (BacA and PAP2) catalyze the dephosphorylation of C55-PP on the same side of the plasma membrane (periplasmic side), questioning us about the identity of the enzyme catalyzing the dephosphorylation of C55-PP synthesized de novo. LpxT is a PAP2 enzyme with a specific kinase activity, transferring the β-phosphate group of C55-PP on a molecule of lipid A, to generate lipid A-1-PP. We mapped, by directed mutagenesis, the active site of LpxT and highlighted the importance of a catalytic triad characteristic to the PAP2 enzymes (His150, His190, Asp194) and other specific residues of LpxT and its closer homologues. The activity of LpxT is inhibited by a small membrane peptide, called PmrR, whose expression is under the control of the two-component system PmrA-PmrB. Our study showed that this inhibition occurred via a direct interaction between these two partners. We showed that the induction of PmrA-PmrB system leads to resistance to the polymyxin B (cationic antimicrobial peptide) and sensitivity to deoxycholate (component major bile) and that the modification catalyzed LpxT produces an opposite effect. The robustness of the resistance to the polymyxin B is connected to the force of the signals inducing PmrA-PmrB system, but also the system PhoP-PhoQ and we clearly identified the signals needed to this resistance in E. coli.