The development of genome scale metabolic models have been aided by the increasing availability of genome sequences of microorganisms such as Geobacter sulfurreducens, involved in environmentally relevant processes such as the in-situ bioremediation of U(VI). Since microbial activities are the major driving forces for geochemical changes in the sub-surface, understanding of microbial behavior under a given set of conditions can help predict the likely outcome of potential subsurface bioremediation strategies. Hence, a model based lookup table was created to capture the variation in physiology of G. sulfurreducens in response to environmental perturbations. Thermodynamically feasible flux distributions were generated by incorporating thermodynamic constraints in the model. These constraints together with the mass balance constraints formed the thermodynamics based metabolic flux analysis model (TMFA). Metabolomics experiments were performed to determine the concentration of intracellular metabolites. These concentrations were posed as constraints in the TMFA model to improve the model accuracy.

An air cathode, membrane-less microbial fuel cell (MFC) containing a co-culture of Geobacter sulfurreducens and Escherichia coli was constructed and compared to pure culture MFCs of both organisms. The E. coli containing MFCs were unsparged and relied on E. coli for oxygen removal. The pure G. sulfurreducens MFC had a power output of 128 mW/m2, compared to 63 mW/m2 for the co-culture at an early stage and 56 mW/m2 for the late stage co-culture. The limiting current density is 404 mA/m2 for the pure G. sulfurreducens culture, 184 mA/m2 for the early co-culture, and 282 mA/m2 for the late co-culture, despite an increase in internal resistance between the early and late co-culture cells. Analysis of metabolites has shown that succinate production is likely to have negatively affected current production by G. sulfurreducens, and the removal of succinate is responsible for the increased current density in the late co-culture cell.

An air cathode, membrane-less microbial fuel cell (MFC) containing a co-culture of Geobacter sulfurreducens and Escherichia coli was constructed and compared to pure culture MFCs of both organisms. The E. coli containing MFCs were unsparged and relied on E. coli for oxygen removal. The pure G. sulfurreducens MFC had a power output of 128 mW/m2, compared to 63 mW/m2 for the co-culture at an early stage and 56 mW/m2 for the late stage co-culture. The limiting current density is 404 mA/m2 for the pure G. sulfurreducens culture, 184 mA/m2 for the early co-culture, and 282 mA/m2 for the late co-culture, despite an increase in internal resistance between the early and late co-culture cells. Analysis of metabolites has shown that succinate production is likely to have negatively affected current production by G. sulfurreducens, and the removal of succinate is responsible for the increased current density in the late co-culture cell.

The development of genome scale metabolic models have been aided by the increasing availability of genome sequences of microorganisms such as Geobacter sulfurreducens, involved in environmentally relevant processes such as the in-situ bioremediation of U(VI). Since microbial activities are the major driving forces for geochemical changes in the sub-surface, understanding of microbial behavior under a given set of conditions can help predict the likely outcome of potential subsurface bioremediation strategies. Hence, a model based lookup table was created to capture the variation in physiology of G. sulfurreducens in response to environmental perturbations. Thermodynamically feasible flux distributions were generated by incorporating thermodynamic constraints in the model. These constraints together with the mass balance constraints formed the thermodynamics based metabolic flux analysis model (TMFA). Metabolomics experiments were performed to determine the concentration of intracellular metabolites. These concentrations were posed as constraints in the TMFA model to improve the model accuracy.

Geobacter sulfurreducens possesses type IV pili that are considered to be conductive nanowires and a crucial structural element in biofilm formation, enabling electron transfer to insoluble metal oxides in anaerobic sediments and to graphite anodes in microbial fuel cells. The molecular mechanism by which electrons are transferred through the nanowires to the electron acceptor is not fully understood. Prior to the work described in this thesis, the gene (pilA) encoding the structural pilus subunit had been identified, but little was known about the functional translation start codon, the length of the mature secreted protein, or what renders the pili conductive. Using mass spectrometry, I found that a tyrosine residue (Y32) near the carboxyl terminus of the mature PilA protein is posttranslationally modified by attachment of glycerophosphate. I studied the significance of Y32 for biofilm formation on various surfaces and for growth of G. sulfurreducens with insoluble electron acceptors. A mutant in which Y32 was replaced by phenylalanine lacked the glycerophosphate; biofilm formation on graphite surfaces was severely diminished and current production in microbial fuel cells was initiated only after a long lag phase. Moreover, cells with Y32F mutation in the pilA gene exhibited growth deficiency when Fe(III) oxide was the sole electron acceptor. My data confirm the role of G. sulfurreducens pili in biofilm formation and electron transfer to Fe(III) oxide and identify an amino acid in the PilA protein that is essential for these two processes. I also confirmed the existence of two functional translation start codons for the pilA gene and identified two isoforms (short and long) of the PilA preprotein by series of genetic complementation experiments. The short PilA isoform is found predominantly in an intracellular fraction, and seems to stabilize the long isoform and influence the secretion of several outer surface c-type cytochromes. The long PilA isoform, on the other hand, is required for secretion of PilA to the outer surface of the cell, a process that requires co-expression of pilA and the nine genes on its 3’ side. The long isoform is essential for biofilm formation on various surfaces, for optimum current production in microbial fuel cells, and for growth on insoluble Fe(III) oxide. This study provides new insight concerning the function and biogenesis of Geobacter type IV PilA, as well as a foundation for further research that will be conducted on microbial nanowires.

L'industrie pétrolière et gazière subie d’importantes pertes économiques en raison de problèmes liés à la corrosion. Parmi ces problèmes, la corrosion induite par les micro-organismes (biocorrosion) fait toujours l’objet de recherche, le mécanisme le plus souvent évoqué et documenté étant lié aux bactéries sulfato-réductrices (BSR). Cependant certaines études ont montré que la biocorrosion pouvait se produire même en absence de BSR dans l'environnement corrosif ; le principal objectif de la thèse était donc de fournir un nouvel éclairage sur la corrosion anaérobie de l'acier au carbone en proposant des mécanismes différents de ceux impliquant les BSR. En premier lieu, l’influence d'une souche électro-active, G. sulfurreducens, sur la protection/corrosion de l'acier C1145 a été étudiée. Lorsque des espèces phosphate sont présentes dans le milieu, la bactérie favorise la formation d’une couche de Fer/Phosphate qui ensuite protège le matériau. En présence d’ammonium, les vitesses de corrosion sont plus élevées mais les bactéries réduisent la dissolution du métal. En deuxième partie, des échantillons de terrain issus des opérations de nettoyage des pipelines des systèmes d’injection ont été analysés d’un point de vue microbiologique et électrochimique. L’analyse moléculaire et l’identification de la communauté bactérienne montre la présence d'espèces sulfurogènes autre que les BSR. Ces bactéries peuvent stimuler la corrosion des métaux par la production d'acides organiques, de CO2 et de différentes espèces soufrées telles que H2S. De surcroît, il a été prouvé que le consortium contenu dans les échantillons de terrain accélérait la corrosion de l'acier au carbone, principalement par la production d'espèces sulfures. / The oil and gas industry is impacted by important economic losses due to corrosion problem. As part of this problem, microbially influenced corrosion (MIC) is still a subject of research. The most often evoked and well acknowledge MIC mechanism is linked to sulphate reducing bacteria (SRB). However, some studies have shown that MIC can occur even when SRB is not present in the corroding environment; in this framework, the main objective of the thesis is to provide new insights on corrosion of carbon steel caused by other mechanisms different to those described with SRB. First, the influence of an electroactive strain, G. sulfurreducens (an iron reducing bacteria, IRB) on the corrosion/protection of steel C1145 was studied. When phosphate species are present in the medium, bacteria promote the formation of an iron phosphate layer (vivianite) that afterwards protects the material. In presence of NH4+, corrosion rates are higher but bacteria decrease the dissolution of the material. In the second part, field samples from pigging operations performed in water injection pipelines were analysed from microbiological and electrochemical corrosion points of view. Molecular analysis and identification of the biofilm community show the presence of sulfidogenic species besides SRB. These bacteria can stimulate metal corrosion through production of organic acids, CO2 and different sulphur species such as H2S. Moreover, it was proved that the consortium contained in field samples accelerated corrosion of carbon steel mainly by production of sulphide species.

De multiples micro-organismes ont la capacité de catalyser l’oxydation électrochimique de matières organiques en s’organisant en biofilm à la surface d’anodes. Ce processus est à la base de procédés électro-microbiens très innovants tels que les piles à combustible microbiennes ou les électrolyseurs microbiens. L’interface biofilm/électrode a été l’objet de nombreuses étudesdont les conclusions restent difficiles à démêler en partie du fait de la diversité des paramètres interfaciaux mis en jeu. L’objet de ce travail de thèse est d’exploiter les technologies microsystèmes pour focaliser l’impact de la topographie de surface des électrodes sur le développement du biofilm et sur ses performances électro-catalytiques. La formation de biofilmsélectroactifs de Geobacter sulfurreducens a été étudiée sur des électrodes d’or présentant des topographies bien contrôlées, sous la forme de rugosité, porosité, réseau de piliers, à des échellesallant du nanomètre à quelques centaines de micromètres. La présence de microrugosité a permis d’accroitre les densités de courant d’un facteur 8 par rapport à une surface lisse et son effet a étéquantifié à l’aide du paramètre Sa. Nous avons tenté de distinguer les effets des différentes échelles de rugosité sur le développement du biofilm et la vitesse des transferts électroniques.L’intérêt de la microporosité a été discuté. L’accroissement de surface active par la présence de micro-piliers s’est avéré très efficace et une approche théorique a donné des clés de compréhension et d’optimisation. Les connaissances acquises dans les conditions de culture pure ont finalement été confrontées avec la mise en oeuvre de biofilms multi-espèces issus d’un inoculum complexe provenant de sédiments marins.

Outer membrane cytochromes OmcB and OmcS of Geobacter sulfurreducens are two important components of the respiratory chain for extracellular Fe(III) reduction. OmcS is a loosely bound cell surface protein involved in the reduction of insoluble Fe(III). OmcB is an outer membrane protein and required for insoluble and soluble Fe(III) reduction. The objective of this study was to understand better the mechanism of dissimilatory Fe(III) reduction, focusing on the cell surface proteins by further localization, identification of protein-protein interactions, and biochemical characterization of OmcB and OmcS. OmcB was found to be surface-exposed but embedded in the outer membrane because mild protease treatment of cells resulted in partial degradation of OmcB. Removal of surface-exposed proteins inhibited Fe(III) reduction, which is at least partially due to the degradation of OmcB. Co-immunoprecipitation studies with outer surface proteins using an antibody against OmcS revealed that OmcS interacts with several proteins, of which some are implicated in Fe(III) reduction, such as PilA, OmpJ, and OmpB, and in electricity production, such as OmcZ. Other OmcS-associated proteins, which have not been studied, include a cytochrome (GSU2887), a hypothetical and a conserved hypothetical protein, and a putative protease with a PDZ domain. The results suggest that co-immunoprecipitation with other antibodies would help to identify more elements of electron transport pathways related to extracellular Fe(III) reduction. OmcB was purified via preparative sodium dodecylsulfate polyacrylamide gel electrophoresis (SDSPAGE) and anion-exchange chromatography. The molecular mass was determined as 82 kDa, and 11.5 hemes per molecule were found. OmcB was able to transfer electrons to either soluble or insoluble Fe(III). OmcS was purified by detergent extraction. The molecular mass was 47 kDa and it contains 6 heme groups. UV-visible, EPR, and NMR spectroscopies determined that all hemes are bis-histidyl hexacoordinated and low-spin in both oxidized and reduced forms. OmcS has a –212 mV midpoint redox potential, and donates electrons to soluble and insoluble metals and quinones. Transient state kinetics showed that OmcS reduces anthroquinone-2, 6-disulfonate 10 times faster than it reduces Fe(III) citrate. This study revealed valuable further details about the mechanism of Fe(III) reduction by G. sulfurreducens by identifying the localization, protein-protein interactions and biochemical characteristics of the components of extracellular electron transport.

La corrosion induite par les micro-organismes (CIM) génère des pertes économiques mondiales chiffrées en milliards d’euros par an. Il est communément admis que les bactéries sulfato-réductrices (BSR) jouent un rôle clé dans la CIM anaérobie des aciers. Malgré cette unanimité, les essais en laboratoire peinent à reproduire la corrosion des aciers observées en milieu naturel; bien plus, ils n’expliquent pas quel est l’élément qui déclenche la corrosion, puisque les BSR présentes dans de nombreux environnements naturels n’induisent pas systématiquement de corrosion. L’objectif de ce travail est d’évaluer la pertinence dans le domaine de la CIM de nouveaux mécanismes de transferts électroniques entre aciers et protéines ou cellules microbiennes. La première partie de la thèse évalue l’effet d’une [Fe]-hydrogénase sur les processus de corrosion anaérobie des aciers au carbone. L’hypothèse d’une catalyse directe de la réduction des protons par des hydrogénases adsorbées a souvent été suggérée dans la bibliographie, elle est ici clairement démontrée. L’hydrogénase de Clostridium acetobutylicum, qu’elle soit active, désactivée ou dénaturée accélère la corrosion de l’acier au carbone. La présence de phosphate dans le milieu rend les interprétations plus complexes mais ne modifie pas le mécanisme. Une nouvelle hypothèse est avancée qui donne un rôle essentiel aux centres fer-soufre de la protéine. La catalyse de la corrosion par les hydrogénases pourrait donc être rapprochée des mécanismes bien connus de catalyse par le sulfure de fer. Dans ce cas l’état redox des centres fer-soufre serait une clé essentielle de l’apparition ou non de la corrosion. La deuxième partie élucide le rôle de Geobacter sulfurreducens sur la corrosion anaérobie de trois types de matériaux : aciers au carbone (1145), ferritique (403) et austénitiques (304L et 316L). Les résultats mettent en évidence pour la première fois que des cellules bactériennes adhérées induisent un anoblissement du potentiel libre des aciers et accélèrent la corrosion des aciers faiblement alliés par un mécanisme de transfert direct d’électrons. Suivant les concentrations d’accepteurs et de donneurs d’électrons en solution, G. sulfurreducens peut accentuer la propagation de la corrosion en catalysant directement la réduction cathodique ou, au contraire, en absence d’accepteurs et en excès de donneurs, protéger contre la corrosion. L’apparition de la corrosion ne peut donc être induite que par la conjonction défavorable de plusieurs paramètres. Ces résultats obtenus en laboratoire apportent de nouvelles voies d’investigations des phénomènes de CIM qui doivent maintenant être confrontées aux milieux naturels. / Microbially influenced corrosion (MIC) costs billions of euros per year. It is commonly agreed that sulphate-reducing bacteria (SRB) play a key role in anaerobic MIC of steels. In spite of this, laboratory experiments have difficulty in reproducing the corrosion of steels that is observed in natural environments. Moreover, they do not explain what triggers corrosion since SRB, ubiquitous in natural environments, do not systematically induce corrosion. The aim of this work was to evaluate the relevance of new electron transfer mechanisms between steels and proteins or microbial cells in the domain of MIC. The first part of the thesis evaluates the impact of [Fe]-hydrogenase on the anaerobic corrosion of mild steels. The direct catalysis of proton reduction by hydrogenases has often been suggested in the literature; here, it is clearly demonstrated. Hydrogenase from Clostridium acetobutylicum, whether it is active, deactivated on denatured, can accelerate the corrosion of mild steel. The presence of a phosphate medium makes the interpretations more complex without modifying the mechanism. A new hypothesis implying the crucial role of iron-sulphur clusters contained in the protein is brought to light. Corrosion catalysis by hydrogenases could be compared with well-known mechanisms of corrosion catalysis by iron sulphide. In this case, the redox state of iron-sulphur clusters would play a key role in the occurrence of corrosion. The second part elucidates the role of Geobacter sulfurreducens in anaerobic corrosion of three types of steels: mild steel (1145), ferritic (403) and austenitic steels (304L and 316L). Results show, for the first time, that adherent bacterial cells induce open circuit potential ennoblement of steels and accelerate the corrosion of slightly alloyed steels by a direct electron transfer mechanism. Depending on the concentrations of the electron acceptors and donors in the medium, G. sulfurreducens could either enhance corrosion propagation by direct catalysis of proton reduction or, in the absence of acceptors and with an excess of donors, protect against corrosion. Thus the occurrence of corrosion relies on the unfavourable conjunction of many parameters. These results obtained in laboratory conditions open new paths for investigating MIC in natural environments.

De multiples micro-organismes ont la capacité de catalyser l’oxydation électrochimique de matières organiques en s’organisant en biofilm à la surface d’anodes. Ce processus est à la base de procédés électro-microbiens très innovants tels que les piles à combustible microbiennes ou les électrolyseurs microbiens. L’interface biofilm/électrode a été l’objet de nombreuses étudesdont les conclusions restent difficiles à démêler en partie du fait de la diversité des paramètres interfaciaux mis en jeu. L’objet de ce travail de thèse est d’exploiter les technologies microsystèmes pour focaliser l’impact de la topographie de surface des électrodes sur le développement du biofilm et sur ses performances électro-catalytiques. La formation de biofilmsélectroactifs de Geobacter sulfurreducens a été étudiée sur des électrodes d’or présentant des topographies bien contrôlées, sous la forme de rugosité, porosité, réseau de piliers, à des échellesallant du nanomètre à quelques centaines de micromètres. La présence de microrugosité a permis d’accroitre les densités de courant d’un facteur 8 par rapport à une surface lisse et son effet a étéquantifié à l’aide du paramètre Sa. Nous avons tenté de distinguer les effets des différentes échelles de rugosité sur le développement du biofilm et la vitesse des transferts électroniques.L’intérêt de la microporosité a été discuté. L’accroissement de surface active par la présence de micro-piliers s’est avéré très efficace et une approche théorique a donné des clés de compréhension et d’optimisation. Les connaissances acquises dans les conditions de culture pure ont finalement été confrontées avec la mise en oeuvre de biofilms multi-espèces issus d’un inoculum complexe provenant de sédiments marins. / Many microorganisms have the ability to catalyze the electrochemical oxidation of organic matterby self-organizing into biofilm on the surface of anodes. This process is the basis of highlyinnovative electro-microbial processes such as microbial fuel cells or microbial electrolysis cells.The biofilm/electrode interface has been the subject of numerous studies whose conclusionsremain difficult to disentangle partly because of the diversity of the interfacial parameters involved.The purpose of this thesis work is to exploit microsystem technologies to focus the impact ofelectrode surface topography on biofilm development and electro-catalytic performance. Theformation of electroactive biofilms of Geobacter sulfurreducens was studied on gold electrodespresenting well-controlled topographies, in the form of roughness, porosity, pillar networks, atscales ranging from nanometer to a few hundred micrometers. The presence of micro-roughnessincreased the current densities by a factor of 8 compared to a smooth surface and its effect wasquantified using the Sa parameter. We have tried to distinguish the effects of different roughnessscales on biofilm development and electron transfer rates. The suitability of micro-porosity wasdiscussed. The increase of active surface area by the presence of micro-pillars has proved veryeffective and a theoretical approach has given keys to understanding and optimization. Theknowledge acquired under pure culture conditions was finally confronted with the use of multispeciesbiofilms formed from a complex inoculum coming from marine sediments.