Probing the different chaperone activities of the bacterial HSP70-HSP40 system using a thermolabile luciferase substrate.

During mild heat-stress, a native thermolabile polypeptide may partially unfold and transiently expose water-avoiding hydrophobic segments that readily tend to associate into a stable misfolded species, rich in intra-molecular non-native beta-sheet structures. When the concentration of the heat-unfolded intermediates is elevated, the exposed hydrophobic segments tend to associate with other molecules into large stable insoluble complexes, also called "aggregates." In mammalian cells, stress- and mutation-induced protein misfolding and aggregation may cause degenerative diseases and aging. Young cells, however, effectively counteract toxic protein misfolding with a potent network of molecular chaperones that bind hydrophobic surfaces and actively unfold otherwise stable misfolded and aggregated polypeptides. Here, we followed the behavior of a purified, initially mostly native thermolabile luciferase mutant, in the presence or absence of the Escherichia coli DnaK-DnaJ-GrpE chaperones and/or of ATP, at 22 °C or under mild heat-stress. We concomitantly measured luciferase enzymatic activity, Thioflavin-T fluorescence, and light-scattering to assess the effects of temperature and chaperones on the formation, respectively, of native, unfolded, misfolded, and/or of aggregated species. During mild heat-denaturation, DnaK-DnaJ-GrpE+ATP best maintained, although transiently, high luciferase activity and best prevented heat-induced misfolding and aggregation. In contrast, the ATP-less DnaK and DnaJ did not maintain optimal luciferase activity and were less effective at preventing luciferase misfolding and aggregation. We present a model accounting for the experimental data, where native, unfolded, misfolded, and aggregated species spontaneously inter-convert, and in which DnaK-DnaJ-GrpE+ATP specifically convert stable misfolded species into unstable unfolded intermediates.

Stimulation of hematopoiesis in vivo by recombinant bacterial murine interleukin 3.

Mouse interleukin 3 (IL-3) cDNA was cloned into a plasmid construction, allowing the synthesis of very high quantities of IL-3 in Escherichia coli. The recombinant (r) IL-3, purified to homogeneity, was active in vitro on the proliferation and differentiation of various hematopoietic progenitor cells at 1 pM. To maintain detectable blood levels of IL-3, osmotic pumps containing rIL-3 or control solutions were placed under the skin of normal and irradiated C3H/HeJ and (BALB X B10) F1 mice. The effect of IL-3 on hematopoietic progenitor cell numbers in spleen and bone marrow was evaluated 3 and 7 days later by using an in vitro clonal assay. The results demonstrated the following: (i) Doses of IL-3 infused at the rate of 2.5-5 ng per g of body weight per hr were sufficient to increase the numbers of hematopoietic progenitors in normal mice by at least 2-fold within 3 days. (ii) In mice with progenitor cell levels depressed by sublethal irradiation, 7-day treatment with IL-3 resulted in a 10-fold increase to near normal levels. (iii) The erythroid and myeloid lineages appeared to be enhanced to the same extent. (iv) Enhancement of hematopoiesis occurred primarily in spleen, but hematopoietic foci were also evident in the liver; in contrast, total cell and progenitor cell numbers were decreased in the bone marrow.

Transposase and cointegrase: specialized transposition proteins of the bacterial insertion sequence IS21 and related elements.

The bacterial insertion sequence IS21 shares with many insertion sequences a two-step, reactive junction transposition pathway, for which a model is presented in this review: a reactive junction with abutted inverted repeats is first formed and subsequently integrated into the target DNA. The reactive junction occurs in IS21-IS21 tandems and IS21 minicircles. In addition, IS21 shows a unique specialization of transposition functions. By alternative translation initiation, the transposase gene codes for two products: the transposase, capable of promoting both steps of the reactive junction pathway, and the cointegrase, which only promotes the integration of reactive junctions but with higher efficiency. This review also includes a survey of the IS21 family and speculates on the possibility that other members present a similar transpositional specialization.

Combining molecular evolution and environmental genomics to unravel adaptive processes of MHC class IIB diversity in European minnows (Phoxinus phoxinus)

Host-pathogen interactions are a major evolutionary force promoting local adaptation. Genes of the major histocompatibility complex (MHC) represent unique candidates to investigate evolutionary processes driving local adaptation to parasite communities. The present study aimed at identifying the relative roles of neutral and adaptive processes driving the evolution of MHC class IIB (MHCIIB) genes in natural populations of European minnows (Phoxinus phoxinus). To this end, we isolated and genotyped exon 2 of two MHCIIB gene duplicates (DAB1 and DAB3) and 1665 amplified fragment length polymorphism (AFLP) markers in nine populations, and characterized local bacterial communities by 16S rDNA barcoding using 454 amplicon sequencing. Both MHCIIB loci exhibited signs of historical balancing selection. Whereas genetic differentiation exceeded that of neutral markers at both loci, the populations' genetic diversities were positively correlated with local pathogen diversities only at DAB3. Overall, our results suggest pathogen-mediated local adaptation in European minnows at both MHCIIB loci. While at DAB1 selection appears to favor different alleles among populations, this is only partially the case in DAB3, which appears to be locally adapted to pathogen communities in terms of genetic diversity. These results provide new insights into the importance of host-pathogen interactions in driving local adaptation in the European minnow, and highlight that the importance of adaptive processes driving MHCIIB gene evolution may differ among duplicates within species, presumably as a consequence of alternative selective regimes or different genomic context.

Bacterial transcriptional regulators for degradation pathways of aromatic compounds.

Human activities have resulted in the release and introduction into the environment of a plethora of aromatic chemicals. The interest in discovering how bacteria are dealing with hazardous environmental pollutants has driven a large research community and has resulted in important biochemical, genetic, and physiological knowledge about the degradation capacities of microorganisms and their application in bioremediation, green chemistry, or production of pharmacy synthons. In addition, regulation of catabolic pathway expression has attracted the interest of numerous different groups, and several catabolic pathway regulators have been exemplary for understanding transcription control mechanisms. More recently, information about regulatory systems has been used to construct whole-cell living bioreporters that are used to measure the quality of the aqueous, soil, and air environment. The topic of biodegradation is relatively coherent, and this review presents a coherent overview of the regulatory systems involved in the transcriptional control of catabolic pathways. This review summarizes the different regulatory systems involved in biodegradation pathways of aromatic compounds linking them to other known protein families. Specific attention has been paid to describing the genetic organization of the regulatory genes, promoters, and target operon(s) and to discussing present knowledge about signaling molecules, DNA binding properties, and operator characteristics, and evidence from regulatory mutants. For each regulator family, this information is combined with recently obtained protein structural information to arrive at a possible mechanism of transcription activation. This demonstrates the diversity of control mechanisms existing in catabolic pathways.

Molecular characterization of HLA-C incompatibilities in HLA-ABDR-matched unrelated bone marrow donor-recipient pairs. Sequence of two new Cw alleles (Cw*02023 and Cw*0707) and recognition by cytotoxic T lymphocytes.

While the influence of HLA-AB and -DRB1 matching on the outcome of bone marrow transplantation (BMT) with unrelated donors is clear, the evaluation of HLA-C has been hampered by its poor serological definition. Because the low resolution of standard HLA-C typing could explain the significant number of positive cytotoxic T lymphocyte precursor frequency (CTLpf) tests found among HLA-AB-subtype, DRB1/B3/B5-subtype matched patient/donor pairs, we have identified by sequencing the incompatibilities recognized by CD8+ CTL clones obtained from such positive CTLpf tests. In most cases the target molecules were HLA-C antigens that had escaped detection by serology (e.g. Cw*1601, 1502 or 0702). Direct recognition of HLA-C by a CTL clone was demonstrated by lysis of the HLA class I-negative 721.221 cell line transfected with Cw*1601 cDNA. Because of the functional importance of Cw polymorphism, a PCR-SSO oligotyping procedure was set up allowing the resolution of 29 Cw alleles. Oligotyping of a panel of 382 individuals (including 101 patients and their 272 potential unrelated donors, 5 related donors and 4 platelet donors) allowed to determine HLA-C and HLA A-B-Cw-DRB1 allelic frequencies, as well as a number of A-Cw, B-Cw, and DRB1-Cw associations. Two new HLA-Cw alleles (Cw*02023 and Cw*0707) were identified by DNA sequencing of PCR-amplified exon 2-intron 2-exon 3 amplicons. Furthermore, we determined the degree of HLA-C compatibility in 287 matched pairs that could be formed from 73 patients and their 184 potential unrelated donors compatible for HLA-AB by serology and for HLA-DRB1/ B3/B5 by oligotyping. Cw mismatches were identified in 42.1% of these pairs, and AB-subtype oligotyping showed that 30% of these Cw-incompatible pairs were also mismatched for A or B-locus subtype. The degree of HLA-C incompatibility was strongly influenced by the linkage with B alleles and by the ABDR haplotypes. Cw alleles linked with B*4403, B*5101, B18, and B62 haplotypes were frequently mismatched. Apparently high resolution DNA typing for HLA-AB does not result in full matching at locus C. Since HLA-C polymorphism is recognized by alloreactive CTLs, such incompatibilities might be as relevant as AB-subtype mismatches in clinical transplantation.

Comparative genomics of epidemic versus sporadic Staphylococcus aureus strains does not reveal molecular markers for epidemicity.

Staphylococcus aureus, especially when it is methicillin resistant, has been recognised as a major cause of nosocomial and community-acquired infections. It has also been shown that certain strains were able to cause clonal epidemics whereas others showed a more incidental occurrence. On the basis of this behavioural distinction, a genetic feature underlying this difference in epidemicity can be assumed. Understanding the difference will not only contribute to the development of markers for the identification of epidemic strains but will also shed light on the evolution of clones. Genomes of strains from two independent collections (n=18 and n=10 strains) were analysed. Both collections were composed of carefully selected, genetically diverse strains with clinically well-defined epidemic and sporadic behaviour. Comparative genome hybridisation (CGH) was performed using an Agilent array for one collection (up to 11 probes per open reading frame - ORF), and an Affymetrix array for the other (up to 30 probes per ORF). Presence and absence information of probe homologues and ORFs was taken for analysis of molecular variance (AMOVA) at the strain and behaviour levels. Not a single probe showed 100% concordant differences between epidemic and sporadic strains. Moreover, probe differences between groups were always smaller than those within groups. This was also true, when the analysis was focussed on presence versus absence of ORF's or when probe information was transformed into allelic profiles. These findings present strong evidence against the presence or absence of a single common specific genetic factor differentiating epidemic from sporadic S. aureus clones.

Autoinduction of 2,4-diacetylphloroglucinol biosynthesis in the biocontrol agent Pseudomonas fluorescens CHA0 and repression by the bacterial metabolites salicylate and pyoluteorin.

The antimicrobial metabolite 2,4-diacetylphloroglucinol (2,4-DAPG) contributes to the capacity of Pseudomonas fluorescens strain CHA0 to control plant diseases caused by soilborne pathogens. A 2, 4-DAPG-negative Tn5 insertion mutant of strain CHA0 was isolated, and the nucleotide sequence of the 4-kb genomic DNA region adjacent to the Tn5 insertion site was determined. Four open reading frames were identified, two of which were homologous to phlA, the first gene of the 2,4-DAPG biosynthetic operon, and to the phlF gene encoding a pathway-specific transcriptional repressor. The Tn5 insertion was located in an open reading frame, tentatively named phlH, which is not related to known phl genes. In wild-type CHA0, 2, 4-DAPG production paralleled expression of a phlA'-'lacZ translational fusion, reaching a maximum in the late exponential growth phase. Thereafter, the compound appeared to be degraded to monoacetylphloroglucinol by the bacterium. 2,4-DAPG was identified as the active compound in extracts from culture supernatants of strain CHA0 specifically inducing phlA'-'lacZ expression about sixfold during exponential growth. Induction by exogenous 2,4-DAPG was most conspicuous in a phlA mutant, which was unable to produce 2, 4-DAPG. In a phlF mutant, 2,4-DAPG production was enhanced severalfold and phlA'-'lacZ was expressed at a level corresponding to that in the wild type with 2,4-DAPG added. The phlF mutant was insensitive to 2,4-DAPG addition. A transcriptional phlA-lacZ fusion was used to demonstrate that the repressor PhlF acts at the level of transcription. Expression of phlA'-'lacZ and 2,4-DAPG synthesis in strain CHA0 was strongly repressed by the bacterial extracellular metabolites salicylate and pyoluteorin as well as by fusaric acid, a toxin produced by the pythopathogenic fungus Fusarium. In the phlF mutant, these compounds did not affect phlA'-'lacZ expression and 2, 4-DAPG production. PhlF-mediated induction by 2,4-DAPG and repression by salicylate of phlA'-'lacZ expression was confirmed by using Escherichia coli as a heterologous host. In conclusion, our results show that autoinduction of 2,4-DAPG biosynthesis can be countered by certain bacterial (and fungal) metabolites. This mechanism, which depends on phlF function, may help P. fluorescens to produce homeostatically balanced amounts of extracellular metabolites.

Mechanism, regulation, and ecological role of bacterial cyanide biosynthesis.

A few bacterial species are known to produce and excrete hydrogen cyanide (HCN), a potent inhibitor of cytochrome c oxidase and several other metalloenzymes. In the producer strains, HCN does not appear to have a role in primary metabolism and is generally considered a secondary metabolite. HCN synthase of proteobacteria (especially fluorescent pseudomonads) is a membrane-bound flavoenzyme that oxidizes glycine, producing HCN and CO2. The hcnABC structural genes of Pseudomonas fluorescens and P. aeruginosa have sequence similarities with genes encoding various amino acid dehydrogenases/oxidases, in particular with nopaline oxidase of Agrobacterium tumefaciens. Induction of the hcn genes of P. fluorescens by oxygen limitation requires the FNR-like transcriptional regulator ANR, an ANR recognition sequence in the -40 region of the hcn promoter, and nonlimiting amounts of iron. In addition, expression of the hcn genes depends on a regulatory cascade initiated by the GacS/GacA (global control) two-component system. This regulation, which is typical of secondary metabolism, manifests itself during the transition from exponential to stationary growth phase. Cyanide produced by P. fluorescens strain CHA0 has an ecological role in that this metabolite accounts for part of the biocontrol capacity of strain CHA0, which suppresses fungal diseases on plant roots. Cyanide can also be a ligand of hydrogenases in some anaerobic bacteria that have not been described as cyanogenic. However, in this case, as well as in other situations, the physiological function of cyanide is unknown.

Common Variable Immunodeficiency: molecular pathways and clinical manifestations

Common variable immunodeficiency (CVID), is a disease that is characterized by hypogammaglobulinemia as well as a defect in T, B and dendritic cells. This leads to recurrent bacterial infection mainly caused by Streptococcus pneumoniae, Klebsiella pneumoniae and Haemophilus influenzae, as well as inflammatory manifestations, i.e. granulomateous disease, gastro-intestinal disorders and chronic lung disease. Intravenous Immunoglobulin (IVIg) therapy reduces CVID susceptibility to bacterial infections to some extend. We analyzed clinical aspects of patients from our database. We recently showed that bacteria-specific CD4 T cells of CVID patients were impaired. We therefor postulated that CVID patients may harbor an acquired T-cell deficiency also called exhaustion. To test this hypothesis, we performed a comprehensive investigation of the functional profiles of bacteria-specific CD4 T cells isolated from 31 healthy individuals and 30 CVID patients. In the present study, we demonstrated that bacteria-specific but not virus-specific CD4 T cells in CVID patients harbored reduced proliferation capacity and expressed high level of PD-1. Interestingly, the blockade of PD-1/PD-1 ligands interactions restored partially bacteria but not virus-specific CD4 T-cell proliferation. Finally, we showed that 1) the level of endotoxins inversely correlates with IgG concentration, 2) IVIG treated CVID patients harbored reduced endotoxemia and 3) IgG concentration exceeding 7 mg/mL strongly reduces both the proportion of CVID patients with detectable endotoxemia and the concentration of endotoxins in plasma. Taken together our observations, suggest that primary B-cell defect(s) in CVID patients leads to recurrent bacterial infections that are associated to an acquired (secondary) impairment of CD4 T cells which may in turn exacerbate the lack of protection against extracellular bacteria.

Force volume and stiffness tomography investigation on the dynamics of stiff material under bacterial membranes.

The determination of the characteristics of micro-organisms in clinical specimens is essential for the rapid diagnosis and treatment of infections. A thorough investigation of the nanoscale properties of bacteria can prove to be a fundamental tool. Indeed, in the latest years, the importance of high resolution analysis of the properties of microbial cell surfaces has been increasingly recognized. Among the techniques available to observe at high resolution specific properties of microscopic samples, the Atomic Force Microscope (AFM) is the most widely used instrument capable to perform morphological and mechanical characterizations of living biological systems. Indeed, AFM can routinely study single cells in physiological conditions and can determine their mechanical properties with a nanometric resolution. Such analyses, coupled with high resolution investigation of their morphological properties, are increasingly used to characterize the state of single cells. In this work, we exploit the capabilities and peculiarities of AFM to analyze the mechanical properties of Escherichia coli in order to evidence with a high spatial resolution the mechanical properties of its structure. In particular, we will show that the bacterial membrane is not mechanically uniform, but contains stiffer areas. The force volume investigations presented in this work evidence for the first time the presence and dynamics of such structures. Such information is also coupled with a novel stiffness tomography technique, suggesting the presence of stiffer structures present underneath the membrane layer that could be associated with bacterial nucleoids.

Molecular genetic approaches to the biology of the moss "Physcomitrella patens"

La mousse haplobiontique Physcomitrella patens est utilisée comme système génétique modèle pour l'étude du développement des plantes. Cependant, l'absence d'un protocole efficace de transformation a constitué jusqu'à présent un gros désavantage méthodologique pour le développement futur de ce système expérimental. Les résultats présentés dans le premier chapitre relatent la mise au point d'un protocole de transformation basé sur la technique de transfert direct de gènes dans des protoplastes par précipitation au PEG. Un essai d'expression transitoire de gènes a été mis au point. Ce protocole a été adapté afin de permettre l'introduction in vivo d'anticorps dans des protoplastes. Le protocole modifié permet d'introduire simultanément du DNA et des IgG dans les cellules, et nous avons démontré que ces anticorps peuvent inactiver spécifiquement le produit d'un gène co-introduit (GUS), ainsi que certaines protéines impliquées dans des processus cellulaires (tubuline). Cet essai, baptisé "essai transitoire d'immuno-inactivation in vivo", devrait être directement applicable à d'autres protoplastes végétaux, et permettre l'élaboration de nouvelles stratégies dans l'étude de processus cellulaires. Le second chapitre est consacré aux expériences de transformation de la mousse avec des gènes conférant une résistance à des antibiotiques. Nos résultats démontrent que l'intégration de gènes de résistance dans le génome de P. patens est possible, mais que cet événement est rare. Il s'agit là néanmoins de la première démonstration d'une transformation génétique réussie de cet organisme. L'introduction de gènes de résistance aux antibiotiques dans les protoplastes de P. patens génère à haute fréquence des clones résistants instables. Deux classes de clones instables ont été identifiés. La caractérisation phénotypique, génétique et moléculaire de ces clones suggère fortement que les séquences transformantes sont concaténées pour former des structures de haut poids moléculaire, et que ces structures sont efficacement répliquées et maintenues dans les cellules résistantes en tant qu'éléments génétiques extrachromosomaux. Ce type de transformation nous permet d'envisager des expériences permettant l'identification des séquences génomiques impliquées dans la replication de l'ADN de mousse. Plusieurs lignées transgéniques ont été retransformées avec des plasmides portant des séquences homologues aux séquences intégrées dans le génome, mais conférant une résistance à un autre antibiotique. Les résultats présentés dans le troisième chapitre montrent que les fréquences de transformation intégrative dans les lignées transgéniques sont 10 fois plus élevées que dans la lignée sauvage, et que cette augmentation est associée à une coségrégation des gènes de résistance dans la plupart des clones testés. Ces résultats génétiques indiquent que l'intégration de séquences d'ADN étranger dans le génome de P. patens a lieu en moyenne 10 fois plus fréquemment par recombinaison homologue que par intégration aléatoire. Ce rapport homologue/aléatoire est 10000 fois supérieur aux rapports obtenus avec d'autres plantes, et fournit l'outil indispensable à la réalisation d'expériences de génétique inverse dans cet organisme à haplophase dominante. THESIS SUMMARY The moss Physcomitrella patens is used as a model genetic system to study plant development, taking advantage of the fact that the haploid gametophyte dominates in its life cycle. But further development of this model system was hampered by the lack of a protocol allowing the genetic transformation of this plant. We have developed a transformation protocol based on PEG-mediated direct gene transfer to protoplasts. Our data demonstrate that this procedure leads to the establishment of an efficient transient gene expression assay. A slightly modified protocol has been developed allowing the in vivo introduction of antibodies in moss protoplasts. Both DNA and IgGs can be loaded simultaneously, and specific antibodies can immunodeplete the product of an expression cassette (GUS) as well as proteins involved in cellular processes (tubulins). This assay, named transient in vivo immunodepletion assay, should be applicable to other plant protoplasts, and offers new approaches to study cellular processes. Transformations have been performed with bacterial plasmids carrying antibiotic resistance expression cassette. Our data demonstrate that integrative transformation occurs, but at low frequencies. This is the first demonstration of a successful genetic transformation of mosses. Resistant unstable colonies are recovered at high frequencies following transformation, and two different classes of unstable clones have been identified. Phenotypical, genetic and molecular characterisation of these clones strongly suggests that bacterial plasmids are concatenated to form high molecular arrays which are efficiently replicated and maintained as extrachromosomal elements in the resistant cells. Replicative transformation in P. patens should allow the design of experiments aimed at the identification of genomic sequences involved in moss DNA replication. Transgenic strains have been retransformed with bacterial plasmids carrying sequences homologous to the integrated transloci, but conferring resistance to another antibiotic. Our results demonstrate an order of magnitude increase of integrative transformation frequencies in transgenic strains as compared to wild-type, associated with cosegregation of the resistance genes in most of these double resistant transgenic strains. These observations provide strong genetic evidence that gene targeting occurs about ten times more often than random integration in the genome of P. patens. Such ratio of targeted to random integration is about 10 000 times higher than previous reports of gene targeting in plants, and provides the essential requirement for the development of efficient reverse genetics in the haplodiplobiontic P. patens.

Computational problems in perfect phylogeny haplotyping: typing without calling the allele.

A haplotype is an m-long binary vector. The XOR-genotype of two haplotypes is the m-vector of their coordinate-wise XOR. We study the following problem: Given a set of XOR-genotypes, reconstruct their haplotypes so that the set of resulting haplotypes can be mapped onto a perfect phylogeny (PP) tree. The question is motivated by studying population evolution in human genetics, and is a variant of the perfect phylogeny haplotyping problem that has received intensive attention recently. Unlike the latter problem, in which the input is "full" genotypes, here we assume less informative input, and so may be more economical to obtain experimentally. Building on ideas of Gusfield, we show how to solve the problem in polynomial time, by a reduction to the graph realization problem. The actual haplotypes are not uniquely determined by that tree they map onto, and the tree itself may or may not be unique. We show that tree uniqueness implies uniquely determined haplotypes, up to inherent degrees of freedom, and give a sufficient condition for the uniqueness. To actually determine the haplotypes given the tree, additional information is necessary. We show that two or three full genotypes suffice to reconstruct all the haplotypes, and present a linear algorithm for identifying those genotypes.

Molecular epidemiology of methicillin-resistant Staphylococcus aureus in Switzerland: sampling only invasive isolates does not allow a representative description of the local diversity of clones.

We conducted a molecular study of MRSA isolated in Swiss hospitals, including the first five consecutive isolates recovered from blood cultures and the first ten isolates recovered from other sites in newly identified carriers. Among 73 MRSA isolates, 44 different double locus sequence typing (DLST) types and 32 spa types were observed. Most isolates belonged to the NewYork/Japan, the UK-EMRSA-15, the South German and the Berlin clones. In a country with a low to moderate MRSA incidence, inclusion of non-invasive isolates allowed a more accurate description of the diversity.

Molecular mechanisms of penicillin-induced killing and penicillin tolerance in Streptococcus gordonii

Abstract The main thesis topic relates to the 'molecular mechanisms of penicillin-induced bacterial death. Indeed, bacteria have developed two principal mechanisms to escape the killing effect of ß-lactam antibiotics: resistance and tolerance. Resistant bacteria are characterized by their ability to grow in the presence of drug concentrations higher than the one inhibiting the growth of susceptible members of the same species. Hence, resistant bacteria have an increased minimal inhibitory concentration (MIC) of the drug. Nevertheless, when exposed to antibiotic concentrations exceeding their new MIC, resistant bacteria remain sensitive to the antibiotic killing effect. In contrast, tolerant bacteria have an unchanged MIC. However, they have a considerably increased ability to survive drug-induced killing, even at concentrations exceeding their MIC by several orders of magnitude. In other words, in the presence of the antibiotic, tolerant bacteria become persister cells which stop growing but are not killed. In the present thesis, it is shown that the survival phenotype of a tolerant Streptococcus gordonii strain depends on two components belonging to sugar metabolism pathways. First, the transcription factor CcpA which mediates a global regulatory mechanism allowing bacteria to utilize the most efficient sugar source for their growth. We show that the inactivation of the ccpA gene leads to a partial loss of penicillin tolerance both in vitro and in a rat model of experimental endocarditis. Second, the Enzyme I of the phosphotransferase system which is involved in the uptake and phosphorylation of sugars. Here, we -show that a single nucleotide mutation in ptsI, the gene encoding the Enzyme I, is sufficient to confer a fully tolerant phenotype in S. gordonii both in vivo and in vivo. The mutation results in a radical proline to arginine substitution in the C-terminal domain of the protein, probably leading to a decrease in its homodimerization and subsequent activity. Taken together our results prove that tolerance is a global survival mechanism linked to sugar metabolism. We hypothesize that, in the presence of the antibiotic, the already altered metabolic processes of the tolerant strain are completely inactivated. Hence, bacteria may enter in a dormant state and become insensitive to the bactericidal effect of ß-lactams, which depends on actively dividing cells. This thesis manuscript also contains two other side-projects. The first one establishes that the ability to form a biofilm is not a requisite for the successful establishment of endocarditis due to S. gordonii. The second one characterizes the S. gordonii a-phosphoglucomutase gene, and shows that its inactivation results in a loss of in vitro fitness and in vivo virulence. Résumé Le sujet principal de cette thèse concerne les mécanismes moléculaires de la mort bactérienne induite par la pénicilline. En effet, les bactéries ont développé deux mécanismes principaux pour échapper à l'effet bactéricide des ß-lactamines : la résistance et la tolérance. Les bactéries résistantes sont caractérisées par leur capacité de croître en présence de concentration d'antibiotiques plus élevées que celles inhibant la croissance des organismes sensibles de la même espèce. Les bactéries résistantes ont donc une augmentation de leur concentration minimale inhibitrice (CMI) à l'antibiotique. Néanmoins, quand elles sont exposées à des concentrations dépassant leur nouvelle CMI, elles restent sensibles à l'effet bactéricide. Au contraire, les bactéries tolérantes ont une CMI inchangée. Toutefois, elles ont une très importante capacité à survivre à l'effet bactéricide des ß-lactamines, ceci même à des concentrations excédant leur CMI de plusieurs ordres de grandeur. En d'autres termes, en présence de l'antibiotique, les bactéries tolérantes deviennent des cellules persistantes qui arrêtent leur croissance mais ne sont pas tuées. Dans la présente thèse, il est montré que le phénotype de survie d'un Streptococcus gordonii tolérant dépend de deux composants appartenant aux voies du métabolisme des sucres. Premièrement, le facteur de transcription CcpA qui contrôle un système global de régulation permettant à la bactérie d'utiliser les sources de sucre les plus efficaces pour sa croissance. Il est montré que l'inactivation du gène ccpA résulte en la perte partielle de la tolérance à la pénicilline aussi bien in vitro que dans un modèle d'endocardite expérimentale chez le rat. Deuxièmement, l'Enzyme I du système de phosphotransfert impliqué dans l'import et la phosphorylation des sucres. Nous montrons qu'une mutation ponctuelle d'un nucléotide dans ptsl, le gène codant pour l'Enzyme I, suffit à complètement conférer un phénotype tolérant chez S. gordonii aussi bien in vitro qu'in vivo. La mutation induit la substitution radicale d'une proline en une arginine dans le domaine C-terminal de la protéine, résultant probablement en une diminution de sa capacité d'homodimérisation et donc d'activité. Dans leur ensemble, nos résultats prouvent que la tolérance est un mécanisme global de survie lié au métabolisme des sucres. Nous présentons l'hypothèse que, en présence de l'antibiotique, les processus métaboliques déjà altérés de la souche tolérante deviennent complètement inactifs. En conséquence, les bactéries entreraient dans un état dormant nonréplicatif, devenant ainsi insensibles à l'effet bactéricide des ß-lactamines qui nécessite des cellules en cours de division active. Le manuscrit de cette thèse contient également deux projets secondaires. Le premier montre que la capacité de former un biofilm n'est pas un prérequis pour le succès de l'initiation de l'endocardite à S. gordonii. Le second caractérise le gène de l'a-phosphoglucomutase de S. gordonii et montre que son inactivation résulte en une perte de fitness in vitro et de virulence in vivo.