Legionella pneumophila est la bactérie responsable de la légionellose, une pneumonie atypique dans les pays industrialisés. Les souches pathogènes sont issues de notre environnement après multiplication à l’intérieur d’amibes, sont disséminées par la technologie humaine, puis peuvent infecter les macrophages alvéolaires humains. Ce travail vise à caractériser une famille d’effecteurs du système de sécrétion de type IV Dot/Icm transloqués dans le cytoplasme de la cellule hôte, des protéine kinases, et en particulier à établir le rôle de la protéine kinase LegK2 dans la virulence. L’analyse in silico et des tests de phosphorylation in vitro ont permis d’identifier 5 protéine kinases fonctionnelles, LegK1-LegK5, codées par la souche épidémique L. pneumophila Lens. Des tests de translocation ont montré qu’à l’exception de LegK5, les protéine kinases de Legionella sont transloquées dans la cellule hôte de façon Icm/Dot dépendante. LegK2 joue un rôle clé dans la virulence, comme démontré par inactivation de gène. Les vacuoles contenant le mutant legK2 présentent un recrutement moins efficace de reticulum endoplasmique, ce qui entraine une réplication intracellulaire retardée. Un mutant de substitution déficient pour l’activité kinase présente les mêmes défauts de virulence, ce qui démontre le rôle central de la phosphorylation dans le contrôle de ce processus. Les mécanismes moléculaires contrôlés par LegK2 sont actuellement recherchés par identification de partenaires et/ou substrats protéiques / Legionella pneumophila is the most common causative agent of the severe pneumony legionellosis. Legionella pathogenic strains are emerging from the environment after intracellular multiplication in amoeba, are dissiminated by water aerosols technologies, and are able to infect alveolar macrophages of human lungs. This work aims to characterize one family of effectors translocated into the host cytoplasm, namely the protein kinase family, and particularly the role of LegK2 protein kinase in virulence. In silico analysis and in vitro phosphorylation assays allowed the identification of 5 functional protein kinases LegK1-LegK5 encoded by the epidemic L. pneumophila Lens strain. Translocation assays showed that except LegK5, the Legionella protein kinases are translocated. LegK2 plays a key role in bacterial virulence, as demonstrated by gene inactivation. The legK2 mutant containing vacuoles display less efficient recruitment of endoplasmic reticulum markers, which results in delayed intracellular replication. A kinase-dead substitution mutant of legK2 exhibits the same virulence defects. Molecular mechanisms controled by LegK2 have been investigated by searching LegK2 partner and substrate proteins

La multi-résistance aux drogues des bactéries est un problème majeur en clinique. L’un des mécanismes de résistance consiste à effluer les composés toxiques hors de la cellule grâce à des protéines de la membrane interne nommées pompes d’efflux. Ces protéines appartiennent à cinq familles (MFS, RND, MATE, SMR et ABC) et peuvent fonctionner en association avec deux autres types de protéines (protéine du périplasme et protéine de la membrane externe) pour former un canal. Dans le cadre d’une thématique de recherche basée sur l’étude des mécanismes de résistance auxdrogues de la bactérie pathogène Legionella pneumophila, une approche bioinformatique menée sur lesgénomes de trois souches séquencés (souches Lens, Paris et Philadelphia) a permis d’identifier des protéinespouvant participer à l’efflux. Notre but a été de vérifier l’implication de ces protéines dans la résistance auxdrogues et dans la virulence de Legionella en ciblant un ou plusieurs gènes codant pour des composants desystèmes d’efflux. Pour inactiver les gènes, nous avons choisi une stratégie de recombinaison homologue. Lesrecombinants ont été testés pour leur sensibilité à des composés toxiques afin de voir si les gènes ciblés jouentun rôle dans l’efflux d’E. coli. Un de ces mutants, le mutant MF201, altéré pour le gène codant pour une protéinehomologue à TolC chez E. coli s’est avéré être 2 à 16 fois plus sensible aux drogues testées comparé à lasouche sauvage. De plus, ce mutant présente un défaut important de virulence dans Acanthamoeba castellanii,Dictyostelium discoideum et les macrophages U937. Ce premier résultat implique que la protéine TolC-like deLegionella aurait un rôle clef dans la relation hôte pathogène et sous-tend un lien entre multi-résistance auxdrogues et virulence. Par ailleurs une étude de l’expression des gènes codant pour des pompes à efflux a étéinitiée afin de comprendre leur rôle au cours du cycle infectieux de Legionella. / Bacterial multi-drug resistance is of major concern in the case of clinic. One of the resistance mecanisms used by bacteria is the efflux of noxious compounds out of the cell thanks to inner membran proteins called efflux pumps. This proteins belong to five families (MFS, RND, MATE, SMR and ABC) and can function in close association with two partners (periplasmic protein and outer membrane protein) to form a canal. In our new research axis based on the study of the drug resistance of the bacterium Legionella pneumophila, we conducted a bioinformatical approach to identify efflux pumps proteins coded by the sequenced genome of three strains (strains Lens, Paris and Philadelphia). Our goal was to study the role of this proteins in Legionella drug resistance and in its virulence. The bioinformatic approach data allowed us to choose one or several genes coding for potential efflux pump components for genetic invalidation by an homologousrecombination strategy. The bacterial mutants were exposed to different noxious compounds in order to know ifthe target genes invalidated were implicated in the efflux of drugs. One of this mutants, strain MF201, which isdeleted for the gene encoding a protein homologous to E. coli TolC protein, revealed to be 2 to 16 times moresensitive to the drug tested compared to the wild-type strain. Furthermore, this mutant showed an importantvirulence defect in Acanthamoeba castellanii, Dictyostelium discoideum and U937 macrophages. This first resultsmeans that the TolC-like protein of Legionella could be a key factor in host-pathogen interaction and stronglysuggests a link between multi-drug resistance and virulence. We also initiated a transcriptomic approach to studyefflux pump genes expression in order to understand their role during the infectious cycle of Legionella.

Legionella pneumophila est une bactérie gram/négative de l'environnement qui infecte et se multiplie au sein des protozoaires aquatiques, comme les amibes. Elle peut également infecter les macrophages pulmonaires humains, causant une forme sévère de pneumonie appelée légionellose (ou maladie du légionnaire). L'importance du Système de Sécrétion de Type IV (SST4) Icm/Dot est clairement démontrée, de même que celle du Système de Sécrétion de Type II (SST2) Lsp. Des études bioinformatiques ont suggéré l'existence d'un Système de Sécrétion de Type I (SST1), mais aucune étude n'a à ce jour démontré sa fonctionnalité, ni même un éventuel rôle dans la virulence. Ce travail consiste à étudier la fonctionnalité et le rôle dans la virulence d'un SST1 potentiel de L. pneumophila. La fonctionnalité de pompes d'efflux potentielles, possédant une structure très proche d'un SST1, est également investiguée. Des mutants de L. pneumophila invalidés pour les gènes codant pour ce SST1 potentiel (lssB, lssD et tolC) ont été construits : ils possèdent une virulence fortement atténuée vis-à-vis de plusieurs types de cellules hôtes. L'entrée dans la cellule est affectée chez ces mutants, bien que le reste du cycle intracellulaire ne soit pas altéré. La fonctionnalité de ce SST1 a été démontrée par la sécrétion de protéines hybrides entre un rapporteur et la protéine RtxA, qui fait partie de la famille des protéines RTX (Repeat/in ToXins), classiquement sécrétées par des SST1. Enfin, la fonctionnalité de plusieurs pompes d'efflux potentielles a été démontrée : des composés toxiques expulsés par ces pompes ont été identifiés, ainsi qu'une pompe majeure (HelA/HelB/HelC). D'autres analyses sont en cours pour caractériser plus précisément l'importance de ces pompes d'efflux / Legionella pneumophila is a gram/negative pathogen that infects and survives within protozoans, such as amoebas. It can also infect human lung macrophages, causing a disease called Legionnaire’s disease. The Type IV Secretion System Icm/Dot is known to be involved in the virulence of L. pneumophila, as well as the Type II Secretion System Lsp. Bioinformatics studies suggested the presence of a Type I Secretion System (T1SS), but its functionality and importance have not been demonstrated to date. This work aims to study the functionality and the implication in the virulence of the putative T1SS of L. pneumophila. Investigation efforts also concerned putative efflux pumps, which share high similarity with T1SS. Mutant strains of L. pneumophila were constructed by deletion of genes encoding the T1SS (lssB, lssD and tolC): they are defective for the entry into the host cells. The creation of the Legionella replicative vacuole is not altered though. Then the functionality of the LssB/LssD/TolC T1SS was demonstrated in a heterologous host: the reconstructed T1SS allows the secretion of hybrid proteins created by fusing a reporter with parts of the RTX protein RtxA. Finally, the functionality of several putative efflux pumps was also demonstrated: several substrates were identified for those efflux pumps, as well as a major pump (HelA/HalB/HelC). Investigations are currently made to decipher the importance of such efflux systems

Legionella pneumophila peut causer une pneumonie fatale chez l’Homme dite maladie des légionnaires. L. pneumophila peut résister aux stress environnementaux en entrant dans un état VBNC (Viable mais non cultivable). La détection de L. pneumophila dans les milieux environnementaux est basée sur l’utilisation de méthodes normalisées : la culture (AFNOR T90-431, ISO 11731) et la PCR (AFNOR T90-471, ISO 12869). Mais, la culture ne détecte pas les formes VBNC et la PCR ne peut pas différencier les formes viables des mortes de la bactérie. Le premier objectif de notre étude était l’élaboration d’un immunocapteur pour la détection de L. pneumophila dans les échantillons artificiels. Notre immunocapteur a été obtenu par l’immobilisation d’un anticorps monoclonal anti-L. pneumophila (MAb) sur une électrode d’indium-tin oxyde (ITO), par la méthode de monocouches auto-assemblées d’époxysilane. Dans le second objectif, l’orientation des Mabs a été étudiée pour réduire la concentration à utiliser sans affecter sa sensibilité. Les Mabs ont été immobilisés sur l’ITO à l’aide des SAMs d’aminosilane pour la détection des formes VBNC. Les deux immunocapteurs ont été caractérisés par la mouillabilité (mesure de l’angle de contact), microscopie à force atomique (AFM), microscopie confocale à balayage laser (CLSM) et spectroscopie d’impédance électrochimique (EIS). Une limite de détection de 10 bactéries a été déterminée dans les échantillons artificiels. La performance de l’immunocapteur ITO-APTES a été évaluée sur des échantillons réels. Les deux méthodes de l’électrochimie et la culture ont été comparées, et les résultats suggèrent une sensibilité plus élevée par la méthode électrochimique. / Legionella pneumophila, may cause a fatal pneumonia in humans named Legionnaires’ disease. The strategies of L. pneumophila to resist to stressful environmental conditions include the ability to enter into a VBNC (Viable But Not ulturable) state. Detection of L. pneumophila in environmental samples benefit from the use of standardized methods: culture (AFNOR T90-431, ISO 11731) and PCR (AFNOR T90-471, ISO 12869). Culture is hampered by its inability to detect VBNC forms. PCR cannot discriminate between live and dead bacteria. The first objective of our study was the elaboration of an immunosensor for the detection of L. pneumophila in artificial samples. Our immunosensor was obtained by immobilization of a onoclonal anti-L. pneumophila antibody (MAb), on an indium-tin oxide (ITO) electrode – by Self Assembled Monolayers (SAMs) method using an epoxysilane. In the second objective, orientation of antibodies was studied for redusing the concentration of MAb used without affecting its sensitivity. Anti-L. pneumophila antibodies were immobilized into an ITO electrode modified by SAMs of aminosilane for the detection of VBNC. Tow immunosensors elaborated were characterized by wettability (contact angle measurement), atomic force microscopy (AFM), confocal laser scanning microscopy (CLSM) and electrochemical impedance spectroscopy (EIS). A detection limit of 10 bacteria was observed on artificial samples. The performance of the last immunosensor ITO-aminosilane electrode was evaluated in term of the bacteria detection in environmental samples. Electrochemical and culture methods were compared, and results suggest that the electochemical methods is more sensitive.

A doença dos legionários consiste em uma broncopneumonia severa e atípica, que acomete de 2 a 7% das pessoas infectadas com Legionella spp e que apresenta taxa de mortalidade que varia de 5 a 30%, sendo considerada uma importante causa de morbidade e mortalidade mundial. A patologia causada pela espécie L. pneumophila tem sido amplamente estudada em modelos experimentais e suas características clínicas foram extensivamente descritas. No entanto, este modelo não representa adequadamente a doença que acomete seres humanos, pois L. pneumophila não é letal aos camundongos como é para humanos. Recentemente, uma nova espécie de bactéria do gênero Legionella, denominada Legionella longbeachae, foi descrita como importante agente de doença dos legionários em países do hemisfério sul. A pneumonia induzida por L. longbeachae em humanos não difere da induzida por L. pneumophila. No entanto, L. longbeachae é letal para camundongos em doses baixas, o que torna esse modelo murino de doença dos legionários mais fidedigno ao que ocorre com humanos. Com a acentuada mudança dos hábitos de nossa sociedade, há o aumento do número de pessoas com fatores que predispõe a doença, como idade elevada ou tratamento imunossupressor. Assim, entender melhor a relação patógeno-hospedeiro no curso da doença dos legionários por meio da utilização de um modelo experimental adequado é importante para a descoberta de novos meios de combater este patógeno. Neste trabalho, geramos uma cepa de L. longbeachae mutante para rpsL, que se torna resistente à estreptomicina. Essa cepa pode ser utilizada para infecções in vivo nas quais a quantificação da CFU foi estimada em placas contendo antibiótico, o que culmina em maior eficiência experimental e menor quantidade de contaminações. Essa cepa foi utilizada em experimentos in vivo para avaliar os componentes do sistema imune que operam na resistência diante de uma dose letal bacteriana administrada pela via intranasal. Demonstramos que camundongos deficientes para as citocinas IFN ou TNF e para o receptor de quimiocinas CCR2 são mais susceptíveis à infecção do que os camundongos selvagens. No entanto, camundongos deficientes para o receptor de quimiocinas CCR5, para o receptor de IL-17, para a citocina IL-6 ou para o receptor citoplasmático NOD2 são mais resistentes à infecção quando comparados com animais selvagens. A descoberta destas moléculas em um modelo de infecção letal in vivo ressalta a importância de alguns componentes da imunidade para a resistência durante a doença dos legionários experimental e possíveis alvos terapêuticos para essa doença. / Legionnaires disease is a severe and atypical bronchopneumonia, which affects 2-7% people infected with Legionella spp and has a mortality rate of 5 to 30%, therefore it is considered an important cause of mortality and morbidity worldwide. Disease caused by Legionella pneumophila has been largely studied in experimental models and its clinical characteristics was extensively described. However this model does not adequately represent the disease that affects humans, because L. pneumophila is not lethal to mice, as it is to humans. Recently, a new species of bacterium from Legionella genus, called Legionella longbeachae, was described as an important agent of Legionnaires disease in the southern hemisphere. The pneumonia induced by L. longbeachae in humans is not different from pneumonia induced by L. pneumophila. However, a low dose of L. longbeachae is lethal to mice, which makes this murine infection model of Legionnaires disease more reliable than that which occurs in humans. Because our society is changing, there is an increase in the number of persons with predisposing factors, like higher age or immunosuppressive treatment. So, a better understanding of host-pathogen relationship by using a suitable experimental model is important to find new ways to fight this pathogen. Here, we generated a strain of rpsL mutant L. longbeachae, which becomes resistant to streptomycin. This strain could be used in in vivo infections, when CFU quantification was estimated in plates with antibiotic, culminating in greater experimental efficiency and lower contamination. This strain was used in in vivo experiments to evaluate components of the immune system that participates in resistance against lethal dose of bacteria administered intranasally. We showed that Tnf-/-, Ifn-/- or Ccr2-/- mice are more susceptible to infection than wild type mice. However Ccr5-/-, Il17r-/-, Il6-/- or Nod2-/- mice are more resistant to infection than wild type animals. The discovery of these molecules in a lethal infection model in vivo highlights the importance of some components of immunity to resistance during experimental Legionnaires disease and potential therapeutic targets to disease.

Legionellen sind stäbchenförmige Bakterien, die in nennenswerten Konzentrationen und als Krankheitserreger vorrangig in technischen Systemen in Erscheinung treten. Ziel dieser Arbeit ist es, die bisher erschienene Literatur sowohl aus dem mikrobiologisch-hygienischen Bereich als auch aus dem technischen Bereich zu sichten, zu vergleichen und auf dieser Basis, unter Berücksichtigung der ökonomischen Konsequenzen an dem konkreten Beispiel des Carl-Thiem-Klinikum Cottbus, ein sinnvolles Konzept zur Legionellenprävention vorzuschlagen. Das Konzept ist so aufgebaut, daß es durch Betrachtung verschiedener Risikobereiche leicht auf andere Pflegeeinrichtungen oder auch Wohnanlagen bzw. Industriebetriebe übertragbar ist. Regelmäßige Messungen werden von verschiedenen Arbeitsgruppen alternativ oder zusätzlich zu Präventionsmaßnahmen gefordert. In dieser Arbeit werden jährliche Messungen in jedem Kollektor zur Beurteilung des Gesamtsystems, 1/2-jährliche Kontrollen in den Gebäuden, in denen sich Pflegestationen befinden, und 1/4-jährliche Kontrollen in Bereichen mit hohem Risiko, wie zum Beispiel den Isolierzimmern der Hämatologie, empfohlen. Die tolerierten Grenzwerte variieren entsprechend des Riskobereiches, in dem sie auftreten. Das in dieser Arbeit am Beispiel des Carl-Thiem-Klinikum Cottbus dargestellte Präventionskonzept beruht im wesentlichen auf regelmäßigen Kontrollen und sieht die Durchführung einer differenziert angewandten Präventionsmaßnahme nur bei nachgewiesener erhöhter Belastung des Systems durch Legionellen vor. Hier hat sich die Dezentralisierung des Warmwassersystems mit Hilfe dezentral installierter Plattenwärmetauscher in das zentrale Heiznetz als die wirtschaftlichste der hygienisch sinnvollen Maßnahmen herausgestellt. Aber auch diese Art der Prävention entbindet nicht von den weiterhin durchzuführenden Kontrollmessungen - mindestens zur Qualitätssicherung. Insgesamt bleibt jedoch jede Festlegung von Grenzwerten, jede Anwendung einer Präventionsmaßnahme immer ein Kompromiß zwischen dem technisch Machbaren und dem ökonomisch Realisierbaren. Große Studien, die hierzu detailliertere Auskunft geben könnten fehlen bis heute, so daß auch weiterhin jeder Betreiber einer Warmwasseranlage seine eigenen Festlegungen treffen muß. / Legionella is a rod-shaped bacteria, that occur with priority in considerable numerus aswell as pathogens concentration in technical systems. Target of this work is to sight the literature appeared so far both from the micro-biological-hygenic area and from the technical area to compare and on this base to suggest with consideration of the economic consequences by the concrete example Carl Thiem clinical center (Cottbus), a meaningful concept to the Legionella - prevention. The concept is structured by view of different risk areas and therefor easily portable to other nursing facilities or also housing estates and industrial companies. Regular measurements are required by different working groups alternatively or additionally to prevention measures. Recommended in this work are annual measurements in each central warmwater line for the evaluation of the total system, 1/2-annual checks in the buildingswith maintenance stations, and 1/4-annual checks within areas with high risk, like for example, the isolating rooms of the Haematology. The tolerated limit values vary according to the risk-area, in which they occur. The prevention concept put up for Carl Thiem clinical center (Cottbus) has been based essentially on regular checks and more detailed means of prevention for areas with proven increased numbers of Legionella. Here the decentralization of the warm-water system by means of peripherally installed plate-type heat exchanger fed by the central heating network has prooven to be the most economical and aswell hygenically meaningful measure. Never the less, this concept of prevention does not relieve of the regularly check on Legionella - at least for quality assurance. Altogether however each definition of Legionella limit values as well as each prevention concept always is a compromise between the technically feasible means and the economically realizable ones. Large studies, to gain more information on this problem are missing until today, so that further on each operator of a warm water system must find his own definitions out of the broad possibilities of regulations.

Cheng Yee Wan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (leaves 95-112). / Abstracts in English and Chinese. / Acknowledgements --- p.i / Abstract --- p.ii / Table of Contents --- p.vi / List of Figures --- p.xi / List of Plates --- p.xiv / List of Tables --- p.xvi / Abbreviations --- p.xviii / Chapter 1. --- Introduction --- p.1 / Chapter 1.1 --- Legionella pneumophila --- p.1 / Chapter 1.1.1 --- Bacterial morphology and ultrastructure --- p.2 / Chapter 1.1.2 --- Microbial ecology and natural habitats --- p.4 / Chapter 1.1.2.1 --- Association with amoeba --- p.5 / Chapter 1.1.2.2 --- Association with biofilm --- p.5 / Chapter 1.2 --- Legionnaires' disease and clinical significance --- p.6 / Chapter 1.2.1 --- Epidemiology --- p.6 / Chapter 1.2.1.1 --- Worldwide distribution --- p.6 / Chapter 1.2.1.2 --- Local situation --- p.7 / Chapter 1.2.2 --- Clinical presentation --- p.7 / Chapter 1.2.3 --- Route of infection and pathogenesis --- p.8 / Chapter 1.2.4 --- Diagnosis --- p.10 / Chapter 1.2.4.1 --- Culture of Legionella --- p.10 / Chapter 1.2.4.2 --- Direct fluorescent antibody (DFA) staining --- p.13 / Chapter 1.2.4.3 --- Serologic tests --- p.13 / Chapter 1.2.4.4 --- Urine antigen testing --- p.14 / Chapter 1.2.4.5 --- Detection of Legionella nucleic acid --- p.15 / Chapter 1.2.5 --- Risk factors --- p.15 / Chapter 1.2.6 --- Treatment for Legionella infection --- p.16 / Chapter 1.3 --- Detection of Legionella in environment --- p.16 / Chapter 1.4 --- Disinfection methods --- p.17 / Chapter 1.4.1 --- Physical methods --- p.19 / Chapter 1.4.1.1 --- Filtration --- p.19 / Chapter 1.4.1.2 --- UV-C irradiation --- p.20 / Chapter 1.4.1.3 --- Thermal eradication (superheat-and-flush) --- p.21 / Chapter 1.4.2 --- Chemical methods --- p.21 / Chapter 1.4.2.1 --- Chlorination --- p.21 / Chapter 1.4.2.2 --- Copper-silver ionization --- p.22 / Chapter 1.4.3 --- Effect of biofilm and other factors on disinfection --- p.23 / Chapter 1.5 --- Photocatalytic oxidation (PCO) --- p.24 / Chapter 1.5.1 --- Generation of strong oxidants --- p.24 / Chapter 1.5.2 --- Disinfection mechanism(s) --- p.27 / Chapter 1.5.3 --- Major factors affecting the process --- p.28 / Chapter 2. --- Objectives --- p.30 / Chapter 3. --- Materials and Methods --- p.31 / Chapter 3.1 --- Chemicals --- p.31 / Chapter 3.2 --- Bacterial strains and culture --- p.31 / Chapter 3.3 --- Photocatalytic reactor --- p.33 / Chapter 3.4 --- PCO efficacy tests --- p.33 / Chapter 3.5 --- PCO sensitivity tests --- p.35 / Chapter 3.6 --- Optimisation of PCO conditions --- p.35 / Chapter 3.6.1 --- Optimization of TiO2 concentration --- p.36 / Chapter 3.6.2 --- Optimization of UV intensity --- p.36 / Chapter 3.6.3 --- Optimization of depth of reaction mixture --- p.36 / Chapter 3.6.4 --- Optimization of stirring rate --- p.37 / Chapter 3.6.5 --- Optimization of initial pH --- p.37 / Chapter 3.6.6 --- Optimization of treatment time and initial cell concentration --- p.37 / Chapter 3.6.7 --- Combinational optimization --- p.37 / Chapter 3.7 --- Transmission electron microscopy (TEM) --- p.38 / Chapter 3.8 --- Fatty acid profile analysis --- p.40 / Chapter 3.9 --- Total organic carbon (TOC) analysis --- p.42 / Chapter 3.10 --- UV-C irradiation --- p.44 / Chapter 3.11 --- Hyperchlorination --- p.44 / Chapter 3.12 --- Statistical analysis and replication --- p.45 / Chapter 3.13 --- Safety precautions --- p.45 / Chapter 4. --- Results --- p.46 / Chapter 4.1 --- Efficacy test --- p.46 / Chapter 4.2 --- PCO sensitivity --- p.47 / Chapter 4.3 --- Optimization of PCO conditions --- p.48 / Chapter 4.3.1 --- TiO2 concentration --- p.48 / Chapter 4.3.2 --- UV intensity --- p.48 / Chapter 4.3.3 --- Depth of reaction mixture --- p.51 / Chapter 4.3.4 --- Stirring rate --- p.56 / Chapter 4.3.5 --- Effect of initial pH --- p.56 / Chapter 4.3.6 --- Effect of treatment time and initial concentrations --- p.56 / Chapter 4.3.7 --- Combinational effects --- p.63 / Chapter 4.4 --- Transmission electron microscopy (TEM) --- p.66 / Chapter 4.4.1 --- Morphological changes induced by PCO --- p.66 / Chapter 4.4.2 --- Comparisons with changes caused by UV-C irradiation and chlorination --- p.67 / Chapter 4.5 --- Fatty acid profile analysis --- p.71 / Chapter 4.6 --- Total organic carbon (TOC) analysis --- p.73 / Chapter 4.7 --- UV-C irradiation --- p.74 / Chapter 4.8 --- Hyperchlorination --- p.74 / Chapter 5. --- Discussion --- p.76 / Chapter 5.1 --- Efficacy test --- p.76 / Chapter 5.2 --- PCO sensitivity --- p.76 / Chapter 5.3 --- Optimization of PCO conditions --- p.77 / Chapter 5.3.1 --- Effect of TiO2 concentration --- p.77 / Chapter 5.3.2 --- Effect of UV intensity --- p.78 / Chapter 5.3.3 --- Effect of depth of reaction mixture --- p.79 / Chapter 5.3.4 --- Effect of stirring rate --- p.79 / Chapter 5.3.5 --- Effect of initial pH --- p.80 / Chapter 5.3.6 --- Effect of treatment time and initial concentrations --- p.81 / Chapter 5.3.7 --- Combinational effect --- p.82 / Chapter 5.4 --- Transmission electron microscopy (TEM) --- p.83 / Chapter 5.4.1 --- Morphological changes induced by PCO --- p.83 / Chapter 5.4.2 --- Comparisons with changes caused by UV-C irradiation and chlorination --- p.85 / Chapter 5.5 --- Fatty acid profile analysis --- p.85 / Chapter 5.6 --- Total organic carbon (TOC) analysis --- p.86 / Chapter 5.7 --- Comparisons of the three disinfection methods --- p.88 / Chapter 6. --- Conclusion --- p.91 / Chapter 7. --- References --- p.95 / Chapter 8. --- Appendix --- p.113

L'implication des tours aéro-réfrigérantes dites « ouvertes », dans les cas de légionellose, demeure à l'heure actuel un problème de santé public malgré la multitude de composés biocides disponibles qui, parfois, sont écotoxiques et appliqués à de forte concentrations. L'efficacité de la majorité de ces biocides est étudiée vis-à-vis des formes planctoniques des microorganismes. Cette étude s'intéresse à la mise au point de deux procédés de traitement du risque Legionella pneumophila : H2O2/UV et ClO2, avec la prise en compte particulière des communautés microbiennes fixées et notamment des protozoaires qui constituent l'hôte naturel de cette bactérie. La démarche expérimentale adoptée pour l'étude de l'efficacité de ces traitements, s'est appuyée sur des biofilms environnementaux naturellement contaminés par des légionelles et issus d'une source thermale. La caractérisation des légionelles isolées des biofilms de cette source montre le maintien du sérogroupe 1 de L. pneumophila et l'apparition de deux autres sérogroupes non reportés dans des études précédentes (sg10 et sg12), avec toutefois une prédominance du sérogroupe 12. Quel que soit le sérogroupe, ces souches environnementales se sont avérées plus virulentes vis-à-vis de l'amibe Acantamoeba castellanii, que les souches cliniques de Lp1 répertoriées. A l'état planctonique elles se sont également avérées très sensibles aux traitements H2O2/UV et UV seul. A l'échelle expérimentale, les deux traitements, H2O2/UV et ClO2, montrent une performance élevée vis-à-vis des biofilms. L'étude a particulièrement mis en évidence le rôle du facteur trophique et de l'adaptation des bactéries au stress oxydatif dans la performance des traitements mais aussi dans l'apparition de la reviviscence. L'application des traitements à des installations en vraie grandeur a permis de conforter les résultats expérimentaux et de mettre l'accent, dans le contexte des sites étudiés, sur les limites de leur efficacité et sur la nécessité d'ajustements des doses appliquées (concentrations des biocides, flux d'irradiation UV, mode d'application) aux particularités des contextes industriels. / Cooling towers are considered as one of the major sources of Legionella pneumophila, the causal agent of Legionnaires' disease. Despite the high number of commercial disinfectants dedicated to it, the Legionella threat is still rising especially in connection with cooling towers. The efficacy of most disinfectants is demonstrated on planktonic bacteria while very few studies have been devoted to their effects on biofilms. The main objective of the study was to optimize two treatments dedicated to the control of biofilms in cooling towers, H2O2/UV and ClO2, so as to reduce the risk associated to Legionella pneumophila while minimizing environmental damages. Their efficacy was studied against environmental biofilms developed in a hot sulphur spring where presence of L. pneumophila is permanent. Their analysis revealed the occurrence of Lp serogroups 1 (Lp1) and of two serogroups not reported in previous studies that were oriented toward water samples, though (Lp10 and Lp12). The environmental strains we isolated display a higher cytotoxicity and virulence towards the amoeba Acantamoeba castellanii than those of known Lp1 epidemic strains and a higher sensitivity to UV and H2O2/UV. In the experimental part of the study, ClO2 and H2O2/UV display a high efficacy against biofilms. Furthermore the study showed the role of trophic parameters and of bacterial adaptation to oxidative stress on the performance of the treatments but also on biofilms regrowth. The experiments performed at the industrial scale corroborate the results gained at the pilot scale and focus on the relationships between the dose and the effectiveness of each treatment. Our results suggest the possibility to apply the process to an industrial scale with necessary adjustments about doses and injection modalities to the context of the considered sites.

A doença dos legionários consiste em uma broncopneumonia severa e atípica, que acomete de 2 a 7% das pessoas infectadas com Legionella spp e que apresenta taxa de mortalidade que varia de 5 a 30%, sendo considerada uma importante causa de morbidade e mortalidade mundial. A patologia causada pela espécie L. pneumophila tem sido amplamente estudada em modelos experimentais e suas características clínicas foram extensivamente descritas. No entanto, este modelo não representa adequadamente a doença que acomete seres humanos, pois L. pneumophila não é letal aos camundongos como é para humanos. Recentemente, uma nova espécie de bactéria do gênero Legionella, denominada Legionella longbeachae, foi descrita como importante agente de doença dos legionários em países do hemisfério sul. A pneumonia induzida por L. longbeachae em humanos não difere da induzida por L. pneumophila. No entanto, L. longbeachae é letal para camundongos em doses baixas, o que torna esse modelo murino de doença dos legionários mais fidedigno ao que ocorre com humanos. Com a acentuada mudança dos hábitos de nossa sociedade, há o aumento do número de pessoas com fatores que predispõe a doença, como idade elevada ou tratamento imunossupressor. Assim, entender melhor a relação patógeno-hospedeiro no curso da doença dos legionários por meio da utilização de um modelo experimental adequado é importante para a descoberta de novos meios de combater este patógeno. Neste trabalho, geramos uma cepa de L. longbeachae mutante para rpsL, que se torna resistente à estreptomicina. Essa cepa pode ser utilizada para infecções in vivo nas quais a quantificação da CFU foi estimada em placas contendo antibiótico, o que culmina em maior eficiência experimental e menor quantidade de contaminações. Essa cepa foi utilizada em experimentos in vivo para avaliar os componentes do sistema imune que operam na resistência diante de uma dose letal bacteriana administrada pela via intranasal. Demonstramos que camundongos deficientes para as citocinas IFN ou TNF e para o receptor de quimiocinas CCR2 são mais susceptíveis à infecção do que os camundongos selvagens. No entanto, camundongos deficientes para o receptor de quimiocinas CCR5, para o receptor de IL-17, para a citocina IL-6 ou para o receptor citoplasmático NOD2 são mais resistentes à infecção quando comparados com animais selvagens. A descoberta destas moléculas em um modelo de infecção letal in vivo ressalta a importância de alguns componentes da imunidade para a resistência durante a doença dos legionários experimental e possíveis alvos terapêuticos para essa doença. / Legionnaires disease is a severe and atypical bronchopneumonia, which affects 2-7% people infected with Legionella spp and has a mortality rate of 5 to 30%, therefore it is considered an important cause of mortality and morbidity worldwide. Disease caused by Legionella pneumophila has been largely studied in experimental models and its clinical characteristics was extensively described. However this model does not adequately represent the disease that affects humans, because L. pneumophila is not lethal to mice, as it is to humans. Recently, a new species of bacterium from Legionella genus, called Legionella longbeachae, was described as an important agent of Legionnaires disease in the southern hemisphere. The pneumonia induced by L. longbeachae in humans is not different from pneumonia induced by L. pneumophila. However, a low dose of L. longbeachae is lethal to mice, which makes this murine infection model of Legionnaires disease more reliable than that which occurs in humans. Because our society is changing, there is an increase in the number of persons with predisposing factors, like higher age or immunosuppressive treatment. So, a better understanding of host-pathogen relationship by using a suitable experimental model is important to find new ways to fight this pathogen. Here, we generated a strain of rpsL mutant L. longbeachae, which becomes resistant to streptomycin. This strain could be used in in vivo infections, when CFU quantification was estimated in plates with antibiotic, culminating in greater experimental efficiency and lower contamination. This strain was used in in vivo experiments to evaluate components of the immune system that participates in resistance against lethal dose of bacteria administered intranasally. We showed that Tnf-/-, Ifn-/- or Ccr2-/- mice are more susceptible to infection than wild type mice. However Ccr5-/-, Il17r-/-, Il6-/- or Nod2-/- mice are more resistant to infection than wild type animals. The discovery of these molecules in a lethal infection model in vivo highlights the importance of some components of immunity to resistance during experimental Legionnaires disease and potential therapeutic targets to disease.

L. pneumophila est une bactérie ubiquitaire des environnements aquatiques, responsable de la légionellose. Elle est principalement retrouvée au sein de protozoaires, mais aussi dans les biofilms. Il est admit que le fer est l'un des éléments indispensable à la croissance de ce pathogène. En 2008, une étude a été réalisée au sein de notre équipe montrant une modulation de l'expression des gènes entre L. pneumophila à l'état planctonique, et à l'état de biofilm. Dans cette même étude, l'ajout de forte concentration en fer (1,25 g/l), dans le milieu de culture des biofilms, a révélé une inhibition de leur développement. Le fer présente donc un rôle dans l'établissement de biofilms mono espèces de L. pneumophila. Nous avons développé un modèle de biofilms naturels, formés à partir d'eau de rivière, afin de tester l'établissement de L. pneumophila, dans des conditions où l'eau de rivière est supplémentée en fer ou au contraire appauvrit, par l'ajout de chélateurs, le deferoxamine mesylate (DFX) ou le dipyridyl (DIP). Les ajouts de fer et de DFX n'ont eu aucun impact sur l'établissement de L. pneumophila contrairement au DIP, qui a induit une augmentation de l'implantation de ces bactéries.Par ailleurs, nous avons effectué une analyse transcriptomique sur L. pneumophila cultivées en milieu liquide supplémenté en DFX. L'ajout du chélateur a entrainé une induction de l'expression de 113 gènes et la répression de 246 gènes. Parmi les gènes induits, certains sont déjà connus comme étant impliqués dans le métabolisme du fer ou contrôlés par le fer. Parmi eux, un gène a été surexprimé, il n'a jamais été associé au fer et sa fonction est encore inconnue à ce jour. Il s'agit du gène lpp2867. Des investigations ont été réalisées afin de caractériser et de comprendre le rôle de la protéine pour laquelle il code. Elle est notamment impliquée dans l'infection des amibes et des macrophages. Son rôle dans le transport du fer ferreux a également été mis en avant. Cette protéine a été nommée IroT pour « iron transporter ». / L. pneumophila is a ubiquitous bacterium found in aquatic environments, and responsible for legionellosis. It is mainly found into both protozoa and biofilms. Iron is a key nutrient for an optimal growth of this pathogen. In 2008, a transcriptomic analysis was carried out within our team, revealing a differential expression of genes involved in iron metabolism between sessile and planktonic bacteria. Also, this study showed that, for high iron concentrations (1.25 g/l), biofilm formation by L. pneumophila was inhibited. It suggested that iron is important for biofilm formation by L. pneumophila. To extend these observations in more natural conditions, a model of biofilm formation, using natural river water, was developed. Water was spiked with L. pneumophila and supplemented with iron or iron chelator, deferoxamine mesylate (DFX) or dipyridyl (DIP). Addition of iron and DFX did not have any effect on L. pneumophila establishment unlike the DIP which induced an increase of L. pneumophila concentration into the biofilms.Otherwise, we performed transcriptomic analysis on L. pneumophila grown in liquid medium supplemented with DFX. The addition of this chelator led to the induction of 113 genes and 246 genes were repressed significantly. Among the induced genes, some are involved in iron metabolism or controlled by iron. There was an induced gene, lpp2867, never associated with iron metabolism and whose function was still unknown. Investigations were realized to characterize and understand the role of the protein encoded by this gene. It is involved in L. pneumophila virulence and in ferrous iron assimilation. This new protein was named IroT for iron transporter.