La maladie de Crohn (MC) est une maladie inflammatoire chronique intestinale (MICI). Son incidence et sa prévalence ont augmenté en Europe au cours des dix dernières années (150 pour 100000 habitants) constituant ainsi un problème de santé majeur. L’inflammation chronique dans la MC favorise la mise en place d’une angiogenèse pathophysiologique. Inflammation et angiogenèse sont deux réponses cellulaires suspectées dans la survenue des cancers coliques associés au MICI. Même si les facteurs favorisant la mise en place de la MC restent non élucidés, la contribution des bactéries exogènes est fortement suspectée. Parmi ces bactéries, les E.coli adhérentes et invasives (AIEC), isolées à partir de la muqueuse iléale de patients porteurs de la MC, sont un bon candidat. Les objectifs de mon projet de thèse étaient de caractériser les mécanismes moléculaires induits par les AIEC et impliqués dans la mise en place des réponses pro inflammatoire et pro angiogénique des cellules intestinales épithéliales. Le facteur de transcription hypoxique (HIF-1) est au cœur de l’immunité innée et de l’angiogenèse. J’ai émis l’hypothèse que les AIEC pouvaient moduler le niveau d’expression de HIF-1α et ainsi contrôler les réponses pro inflammatoire et pro angiogénique. Dans mon premier article, j’ai montré que HIF-1α est maximalement exprimé au niveau de l’épithélium iléal des patients porteurs de la MC. Ensuite, j’ai montré sur un modèle murin compétent pour l’infection par les AIEC, les souris CEABAC10, que les bactéries induisent l’augmentation du niveau protéique de HIF-1 α ainsi que l’activation de la voie de signalisation du VEGF, le facteur angiogénique le plus puissant. / Non communiqué.

Mounting concerns about drug-resistant pathogenic bacteria have rekindled the interest in bacteriophages (bacterial viruses). As bacteria’s natural predators, bacteriophages offer a critical advantage over antibiotics, namely that they can be highly specific. This means that phage therapeutics can be designed to destroy only the infectious agent(s), without causing any harm to our microbiota. However, the potential secondary effects on the balance of microbiota through bacteriophage-induced genome evolution remains as one of the critical apprehensions regarding phage therapy. There exists a significant gap in knowledge regarding the direct and indirect effect of phage therapeutics on the microbiota. The aim of this thesis was to: (1) establish an in vivo model for investigation of the evolutionary dynamics and co-evolution of therapeutic phage and its corresponding host bacterium in the gut; (2) determine if phage therapy can affect the composition of the gut microbiota, (3) observe the differences of phage-resistant bacteria mutants evolved in vivo in comparison to those evolved in vitro. We used germ-free mice colonized with a consortium of eight known bacteria, known as the altered Schaedler flora (ASF). The colonizing strain of choice (mock infection) was a non-pathogenic strain E. coli K-12 (JM83) known to co-colonize the ASF model, which was challenged in vivo with T7 phage (strictly lytic). We compared the composition of the gut microbiota with that of mice not subject to phage therapy. Furthermore, the resistant mutants evolved in vivo and in vitro were characterized in terms of growth fitness and motility. / Thesis / Master of Applied Science (MASc) / Bacteriophages are viruses that infect bacteria. After their discovery in 1917, bacteriophages were a primary cure against infectious disease for 25 years, before being completely overshadowed by antibiotics. With the rise of antibiotic resistance, bacteriophages are being explored again for their antibacterial activity. One of the critical apprehensions regarding bacteriophage therapy is the possible perturbations to our microbiota. We set out to explore this concern using a simplified microbiome model, namely germ-free mice inoculated with only 8 bacteria plus a mock infection challenged with bacteriophage. We monitored this model for 9 weeks and isolated a collection of phage-resistant bacterial mutants from the mouse gut that developed post phage challenge, maintaining the community of mock infection inside the gut. A single dose of lytic phage challenge effectively decreased the mock infection without causing any extreme long-term perturbations to the gut microbiota.

In Bakterien werden Phospholipide auf der cytoplasmatischen Seite der Plasmamembran synthetisiert. Damit ein gleichmäßiges Wachstum und somit die Stabilität biogener Membranen, d.h. Membranen, an bzw. in denen Lipidsynthese stattfindet, gewährleistet ist, muss zumindestens die Hälfte neu synthetisierter Lipide auf die entgegengesetzte Membranhälfte gelangen. Aus früheren Untersuchungen ist bereits bekannt, dass dieser transversale Phospholipidaustausch, auch als Flip-Flop bezeichnet, sehr schnell, kopfgruppenunabhängig und möglicherweise proteinabhängig ist. Dennoch sind die genauen Mechanismen dieser Prozesse noch weitgehend unverstanden. Um die oben erwähnten grundlegenden Phospholipidtransportprozesse zwischen beiden Membranhälften genauer untersuchen zu können, wandten wir einen neuartigen, sogenannten stopped-flow BSA back-extraction Assay an. Mit Hilfe dieses Assays, waren wir in der Lage, die transversale Bewegung und die Verteilung von kurzkettigen, fluoreszenzmarkierten Phospholipidanaloga über beide Membranhälften in ex vivo-Membranen zu charakterisieren. Der stopped-flow BSA back-extraction Assay basiert auf der Technik der stopped-flow-Spektroskopie und der Tatsache, dass BSA in der Lage ist, kurzkettige, fluoreszenzmarkierte Lipidanaloga aus der äußeren Leaflet von (biologischen) Membranen zu extrahieren. Wir entschieden uns für invertierte Membranvesikel der Plasmamembran (IIMV) vom E.coli Wildtypstamm MG1655 als Untersuchungsobjekt, einerseits, weil diese Vesikel nur eine Membran besitzen und zum Anderen, weil IIMV sich sehr gut als Modell für den Flip-Flop von Phospholipiden nutzen lassen. Wir beobachteten, dass kurzkettige, fluoreszenzmarkierte Analoga der beiden am häufigsten in E.coli vorkommenden Phospholipide, Phosphatidylethanolamin (PE) und Phosphatidylglycerol (PG), sehr schnell, d.h. mit Halbwertzeiten von weniger als drei Minuten, über die Membran von IIMV verteilten. Weiterhin verhielten sich kurzkettige, fluoreszenzmarkierte Analoga von den E.coli-fremden Phospholipiden, Phosphatidylcholin (PC) und Phosphatidylserin (PS), ähnlich wie die Analoga von PE und PG. Überraschenderweise, fanden wir heraus, dass alle oben genannten Phospholipidanaloga im Gleichgewichtszustand nicht gleichmässig über beide Membranhälften verteilt waren. Inwiefern Proteine an dieser transversalen Bewegung der Phospholipidanaloga beteiligt sind, sollten Messungen des Flip-Flop von Analoga an unbehandelten und mit Proteinase K inkubierten Vesikeln zeigen, die aus einem Detergenzextrakt von IIMV rekonstituiert wurden. Zunächst konnten wir zeigen, dass die schnelle Bewegung der Phospholipidanaloga über die Membran von rekonstituierten, nicht mit Proteinase K behandelten Vesikeln (Proteoliposomen) erhalten blieb. Nach Inkubation mit Proteinase K wurde jedoch der Flip-Flop von PE- und PG-Analoga vollständig inhibiert. Untersuchungen an rekonstituierten Serien von Proteoliposomen mit ansteigendem bakteriellen Proteingehalt zeigten, dass in Proteoliposomen ohne bakterielle Proteine kein Flip-Flop stattfand und somit nur 50% der fluoreszenten Analoga extrahiert wurden. In Proteoliposomen, die bakterielle Proteine enthielten, stieg das Ausmass der Extrahierbarkeit der untersuchten Analoga mit steigendem Proteingehalt. Diese Daten zeigten sehr deutlich, dass die transversale Bewegung von Phospholipiden über die innere Membran von E.coli durch Proteine vermittelt wird. Schlussfolgernd aus unseren Analysen konnten wir zeigen, dass die transversale Bewegung von Phospholipidanaloga über die Membran von IIMV sehr schnell, proteinabhängig, bidirektional und kopfgruppenunbhängig ist. Zur Identifizierung der molekularen Grundlagen der proteinvermittelten, schnellen Transversalbewegung von Phospholipiden über IIMV-Membranen, nutzen wir Ionenaustauschchromatografie. Zur unserer Überraschung mussten wir feststellen, dass in keiner der rekonstituierten Fraktionen eine nennenswerte Anreicherung der Flippaseaktivität auftrat. Möglicherweise sind mehrere Proteine, mit unterschiedlichen Nettoladungen, oder aber auch Untereinheiten, die sich nicht durch Anionenaustauscher trennen liessen, am Flip-Flop von Phospholipiden beteiligt. Weitergehende Analysen mit anderen Proteinfraktionierungsmethoden sind notwendig, um den oder die Flippasekomplex(e) zu identifizieren. / In the plasma membrane of bacteria, phospholipids are synthesized on the cytoplasmic leaflet of the plasma membrane. To ensure balanced growth and thus, stability of biogenic membranes, half of the newly synthesized lipids must move to the opposing leaflet. It is known that this phospholipid transmembrane movement (flip-flop) is rapid, head-group independent and possibly protein mediated. However, the exact mechanism of this process remains elusive. To investigate these fundamental transbilayer phospholipid transport processes in biogenic membranes, a novel stopped-flow BSA back-exchange assay was utilized to characterize the transmembrane movement and transbilayer distribution of fluorescent labeled, short-chain phospholipid analogues in ex vivo membranes. This approach is based on stopped-flow fluorescence spectroscopy, and the fact that BSA is able to extract fluorescent labeled, short-chain phospholipid analogues from the outer leaflet of (bio)membranes. We chose isolated inverted inner membrane vesicles (IIMV) derived from E.coli wild type MG1655, both for their simple membrane organization and for their suitability as a simple model organism for phospholipid flip-flop. We observed that fluorescent-labeled, short-chain analogues of the major phospholipids in E.coli, phosphatidylethanolamine (PE) and phosphatidylglycerol (PG), rapidly redistributed across the IIMV bilayer with half-times of less than three minutes. Furthermore, fluorescent, short-chain phospholipid analogues of phosphatidylcholine (PC) and phosphatidylserine (PS), which are not naturally occurring phospholipids in E.coli membranes, behaved similar to the PE and PG analogues. To analyze the relevance of proteins for the transmembrane movement of fluorescent analogues, we measured flip-flop of phospholipid analogues in reconstituted and/or untreated and proteinase K treated vesicles generated from protein detergent extracts of IIMV. The amount of extractable fluorescent phospholipids analogues correlated with the amount of protein reconstituted into the proteoliposomes, strongly indicating, that protein concentrations below 100 µg/ml were not sufficient to equip every vesicle with proteins that facilitate the transmembrane movement of the fluorescent analogues. We found that the rapid transbilayer movement of phospholipid analogues across the membrane was maintained in untreated reconstituted vesicles. However, the flip-flop of fluorescent PG and PE analogues was eliminated in proteinase K treated vesicles. In conclusion, our analysis showed that the transmembrane movement of the phospholipid analogues across the membrane of IIMV was protein-mediated, very rapid, bi-directional and head-group independent. To identify the molecular basis of the protein-mediated, rapid transmembrane movement of phospholipids across IIMV membranes, we used ion exchange chromatography (IEC) to separate the IIMV proteins. To our surprise, we did not observe an enhanced flip-flop activity in any of the fractions, indicating that at least two proteins with possibly opposite net charges or several subunits, which were not separable by AEC, are involved. Further analysis using different protein separation techniques will be necessary to identify the putative flippase complex.

La cellule répond aux contraintes physiques exercées par son environnement par un ensemble de mécanismes regroupés sous le terme de mécanotransduction. Ces processus font appel aux molécules impliquées dans l’adhésion cellulaire, au cytosquelette et au noyau. Ces contraintes environnementales, qu’elles soient liées à la rigidité du support, à sa topographie ou à la nature de sa chimie de surface, vont moduler la morphologie cellulaire et impacter le comportement de la cellule. Afin d’étudier cette influence du support, nous avons ensemencé des cellules souches mésenchymateuses (CSMs) de moelle osseuse (MO) issues d’une culture primaire sur des surfaces de mica vierges ou fonctionnalisées de façon homogène avec des molécules naturelles (la fibronectine FN et le peptide RGD cyclique) ou avec des multicouches de polyélectrolytes PEM (cinq cycles de Chitosan/PAA ou de Chitosan/PSS). Nous avons ensuite étudié la morphologie, la prolifération et la différenciation de ces cellules après 12 jours de culture. Il en résulte que les CSMs de MO adhèrent sur toutes les surfaces, traitées ou non, et bien que leur étalement soit moindre sur les surfaces vierges, elles adoptent une morphologie de type fibroblastique similaire à leur phénotype physiologique. Leur pourcentage de confluence varie significativement en fonction du traitement de surface utilisé. En effet la confluence maximale a été observée pour les surfaces greffées avec la FN (93.25 ± 2.75 %) alors que les surfaces traitées avec les PEM présentent des pourcentages de confluence bien plus faibles (61.00 ± 4.08 % pour le couple chitosan/PAA et 54.75 ± 1.75 % pour le couple Chitosan/PSS), s’expliquant principalement par une latence cellulaire en début de culture. Enfin, les cellules cultivées sur nos surfaces ne réagissent à aucune des trois colorations Oil Red O, Alcian Blue ou Alizarin Red S, suggérant une absence de différenciation dans les voies adipogénique, chondrogénique ou ostéogénique induite par ces surfaces. Ainsi, le contrôle de la chimie du support ne permet pas à lui seul un contrôle de la différenciation cellulaire. Cette étude ouvre la voie à l’étape suivante au cours de laquelle l’influence des supports à chimie et géométrie contrôlées. De même, la souche E.coli (bactérie pathogène) répond aux contraintes physiques et chimiques qui lui ont été imposées. Ces contraintes qu’elles soient liées à la topographie ou la nature de la chimie de surface font appel à des molécules naturelles impliquées dans le comportement des bactéries et leur morphologie en particulier sur leur taille. Pour étudier cet impact, nous avons mis en contact la souche E.coli E2146 avec des surfaces de mica vierges ou traitées de façon homogène ou patternée avec des molécules naturelles (la FN et le peptide RGD cyclique). Ensuite, nous avons étudié le taux de recouvrement et la taille des bactéries. Il en résulte que les bactéries adhèrent sur l’ensemble des surfaces bien que l’adhésion soit moindre sur les surfaces de mica vierges. Leur taux de recouvrement varie significativement pour une surface donnée. En effet, le taux de recouvrement et la taille maximaux sont observés sur des surfaces patternées greffées avec la FN, ce qui prouve leur efficacité et l’impact qu’elles ont sur le comportement de E.coli. Nous avons donc démontré dans ce travail de thèse l’influence des propriétés de surfaces sur la croissance de cellules vivantes telles que les cellules souches ou les bactéries. / The cell responds to the physical constraints exerted by its environment by a set of mechanisms grouped under the term of mechanotransduction. These processes involve the molecules involved in cell adhesion, the cytoskeleton and the nucleus. These environmental constraints, whether related to the rigidity of the support, to its topography or to the nature of its surface chemistry, will modulate the cellular morphology and impact the behavior of the cell. In order to study this influence of the support, we have seeded bone marrow mesenchymal stem cells from a primary culture on virgin mica surfaces or functionalized homogeneously with natural molecules (fibronectin and the cyclic RGD peptide) or with polyelectrolyte multilayers (five cycles of Chitosan/PAA or Chitosan/PSS). We then studied the morphology, proliferation and differentiation of these cells after 12 days of culture. As a result, bone marrow mesenchymal stem cells adhere to all surfaces, whether treated or not, and although they are less spread on virgin surfaces, they adopt a fibroblastic type morphology similar to their physiological phenotype. Their percentage of confluence varies significantly depending on the surface treatment used. Indeed the maximum confluence was observed for the surfaces grafted with fibronectin (93.25 ± 2.75%) whereas the surfaces treated with the polyelectrolyte multilayers have much lower confluence percentages (61.00 ± 4.08% for the chitosan/PAA couple) and 54.75 ± 1.75% for the Chitosan/PSS couple), mainly due to cell latency at the beginning of culture. Finally, cells cultured on our surfaces do not respond to any of the three Oil Red O, Alcian Blue or Alizarin Red S stains, suggesting a lack of differentiation in the adipogenic, chondrogenic or osteogenic pathways induced by these surfaces. Thus, the control of the support chemistry alone does not allow control of cell differentiation. This study paves the way for the next step in which the influence of controlled chemistry and geometry media will be studied. Similarly, the E. coli strain (pathogenic bacterium) responds to the physical and chemical constraints imposed on it. These constraints, whether related to the topography or the nature of surface chemistry, involve natural molecules involved in the behavior of bacteria and their morphology, in particular their size. To study this impact, we contacted E.coli strain E2146 with virgin mica surfaces or treated homogeneously or patterned with natural molecules (fibronectin and cyclic RGD peptide). Then we studied the recovery rate and the size of the bacteria. As a result, the bacteria adhere to all surfaces although adhesion is less on virgin mica surfaces. Their recovery rate varies significantly for a given area. Indeed, the recovery rate and the maximum size are observed on patterned surfaces grafted with fibronectin which proves their effectiveness and the impact they have on the behavior of E. coli. We have therefore demonstrated in this thesis the influence of surface properties on the growth of living cells such as stem cells or bacteria.

Made available in DSpace on 2015-03-26T13:28:10Z (GMT). No. of bitstreams: 1 texto completo.pdf: 3335039 bytes, checksum: 432a8a4734747e35f11922a30da30b08 (MD5) Previous issue date: 2012-02-08 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / This study aimed at evaluating the disinfection process of sewage sludge from a UASB reactor by greenhouse solar drying. Seven lots of sewage sludge were monitored between February 2010 and May 2011, each lot referring to an event of sludge discharge from the UASB reactor. The monitoring period of time of the different sludge lots varied from 50 to 130 days, with sampling every week. The following variables were monitored: pH, moisture, total solids, total coliforms, Escherichia coli, total and viable helminth eggs. In addition, somatic coliphages spiking experiments were conducted in sludge samples dried in the same field greenhouse, as well as in samples dried in a laboratory incubator kept under controlled temperature. A high sludge dewatering during the drying treatment in the greenhouse was observed, with moisture and total solids contents reaching levels, respectively, below 10% and above 90% when the sludge was subjected to periods of treatment longer than 75 days. Microbial decay during sludge drying was variable, but effective. Exponential decay was quite clear for total coliforms and Escherichia coli, but not so much for total and viable helminth eggs. Regarding the bacterial removal, an influence of the temperature was noticed, with higher decreasing of total coliforms and E.coli in the warmer periods. E.coli concentrations below 10³ MPN/gTS (the CONAMA Resolution 375/2006 standard for Class A biosolids) were obtained with more than 60 days of treatment in the greenhouse. Calculated values of Kb 20 for E.coli varied widely, between 0.004 and 0.142 d-1. Wide variations in the decay of helminth eggs were also observed, with calculated values of Koh 20 for total eggs ranging between 0.008 and 0.035 d-1. Concentrations of viable helminth eggs were always low, requiring only 15 days to reach counts below 0.25 eggs/gTS (the CONAMA Resolution 375/2006 standard for Class A biosolids). The results of the spiking experiments conducted in the greenhouse also showed intense exponential decay of somatic coliphages, with high Kcs 20 values ranging between 0.187 and 0.248 d-1. In contrast, in the spiking experiments performed in laboratory under conditions of controlled temperature (30ºC) the decay was much less intense, with Kcs 20 values between 0.028 and 0.078 d-1. Finally, this work presents a model for estimating the microbiological quality of biosolids treated by greenhouse drying based on the microbiological quality of raw sewage. Using stochastic modeling (taking into account variations on the input variables) it was estimated that, with 60 days of sludge drying in the greenhouse, the probability of reaching the bacteriological quality established in COMANA Resolution 375/2006 for biosolids classes A and B (103 and 106 E.coli/gST, respectively) was 66.2% and 97% respectively. In the case of helminth eggs these chances were respectively 64.3% and 99.8%. That is, the results suggest high stability and reliability of the greenhouse drying treatment process of the UASB sewage sludge. / O presente trabalho teve como objetivo o acompanhamento da higienização de lodo de esgoto proveniente de reator UASB por secagem em estufa. Foram monitorados sete lotes de lodo, entre fevereiro de 2010 e maio de 2011, sendo cada lote referente a um evento de descarte de lodo do reator UASB. O tempo de acompanhamento dos vários lotes de lodo variou de 50 a 130 dias, com realização de coletas semanais. Foram monitoradas as seguintes variáveis: pH, umidade, sólidos totais, coliformes totais, Escherichia coli e ovos totais e viáveis de helmintos. Foram ainda realizados experimentos de inoculação de colifagos somáticos em lodos submetidos à secagem na mesma estufa de campo e também sob condições de temperatura controlada (estufa laboratorial). Foi observada elevada desidratação do lodo durante a secagem em estufa, sendo obtidos teores de umidade menores que 10% e de sólidos totais acima de 90% quando os lotes foram submetidos a períodos de tratamento superiores a 75 dias. O decaimento microbiano durante a secagem do lodo se mostrou variável, mas eficiente. Decaimento exponencial foi bastante nítido para coliformes totais e Escherichia coli, porém essa tendência não foi tão clara para os dados de ovos viáveis e totais de helmintos. Para a remoção das bactérias, foi possível perceber influência da temperatura (sazonalidade), com quedas das concentrações de coliformes totais e E.coli mais acentuada nos períodos mais quentes. Valores de E.coli abaixo de 10³ NMP/gST, limite estabelecido pela Resolução CONAMA 375/2006 para biossólido Classe A, foram obtidos com tempos de secagem acima de 60 dias. Os valores de Kb 20 calculados para E.coli foram bastante variáveis, entre 0,004 e 0,142 d-1. Variações amplas também foram observadas no decaimento de ovos de helmintos, com valores de Koh 20 calculados para ovos totais entre 0,008 e 0,035 d-1. As concentrações de ovos viáveis de helmintos foram sempre baixas, sendo necessários apenas 15 dias para se obter contagens abaixo de 0,25 ovos/gST (biossólido Classe A de acordo com a Resolução CONAMA 375/2006). Os resultados dos experimentos de campo de inoculação de colifagos somáticos também mostraram intenso decaimento exponencial para esse bacteriófago em estufa, com valores de Kcs 20 elevados e variando entre 0,187 e 0,248 d-1. Em contrapartida, nos experimentos de inoculação realizados em laboratório sob condições temperatura controlada (30ºC) o decaimento foi bem menos intenso, com valores de Kcs 20 entre 0,028 e 0,078 d-1. Por fim, o trabalho apresenta um modelo de estimativa da qualidade microbiológica de biossólidos tratados por secagem em estufa a partir da qualidade microbiológica do esgoto bruto. Por meio de modelagem estocástica (que leva em consideração variações em torno das variáveis de entrada) estimou-se que, com 60 dias de secagem em estufa, a probabilidade de alcançar a qualidade bacteriológica preconizada para as Classes A e B na Resolução CONAMA 375/2006 (103 e 106 E.coli/gST, respectivamente) seria de 66,2% e 97%, respectivamente. No caso dos ovos de helmintos essas chances seriam, respectivamente, de 64,3% e 99,8%. Ou seja, os resultados sugerem estabilidade e confiabilidade consideráveis do processo de tratamento por secagem em estufa do lodo de reator.

Master of Science / Department of Grain Science and Industry / Praveen V. Vadlani / Synthesis of chemical derivatives from finite fossil fuels requires considerable energy inputs and leaves an undesirable environmental footprint. The emerging biorefinery approach leads to sustainable processing of biomass into a wide spectrum of bio-based products, catering to food, feed, chemicals, materials, and bioenergy industries. Cyanophycin (multi-L-arginyl-poly-L-aspartic acid, CGP) is a non-ribosomally synthesized reserve polypeptide, which consists of equimolar amounts of arginine and aspartic acid arranged as a polyaspartate backbone and arginine as the side chain. Cyanophycin is a source of the constituent N-functionalized platform chemical, which can be further processed into many other chemicals of importance. It can be hydrolyzed in mild condition to its constituent amino acids - aspartic acid and arginine. These amino acids may be utilized directly in food and pharmaceutical applications. Based on the chemical structure of these amino acids and the presence of functionalized nitrogen-containing groups, it is conceivable that a number of industrial chemicals can be synthesized, for example: 1, 4-butanediamine， a co-monomer in the production of nylon-4, 6. Other chemicals which could be obtained from cyanophcyin, that are currently prepared from fossil resources, include 1,4-butanediol and urea. Cyanophycin can also be hydrolyzed to a derivative with reduced arginine content or even to poly-aspartic acid, and used as a biodegradable substitute for synthetic polyacrylate in various technical process, such as water treatment (water softeners) and plastics. Cyanophycin is produced by most cyanobacteria in nature; however, these microbes are not suitable for large-scale production due to slow growth and low polymer content. Biosynthesis of cyanophycin is catalyzed by a single enzyme - cyanophycin synthetase (CphA), which is encoded by cyanophycin synthetase structure gene (cphA). The cphA gene can be expressed in several bacteria and plants. E.coli is one of the most commonly used bacterial hosts for the production of recombinant proteins. The recombinant culture has the ability to produce considerably large quantities of cyanophycin in a shorter period of time compared to cyanobacteria. Genome of Anabaena variabilis ATCC 29413 has been sequenced, and it contains the structural gene (cphA) for cyanophycin synthetase. The native enzyme-cyanophycin synthetase from this cyanbacterium culture had been purified and it consists of identical subunits of 98kDa. Polymerization of the amino acids to cyanophycin in vitro is dependent on the presence of ATP, K+, Mg2+, a (cyanophycin) primer and a thiol reagent such as β-mercaptoethanol in the reaction mixture. Our research is the first time that cphA gene from A. variabilis ATCC 29413 was cloned into E.coli. The 2.7 kb cphA gene was amplified by PCR, ligated to the vector pET45b+ and cloned into BL21 (DE3) pLysS and BL21 (DE3). Characterization of cyanophycin was performed by SDS-PAGE, HPLC, mass spectrometry and amino acid analysis. Results showed that protein with molecular weight of 21.5 to 31 kDa did not match any E.coli proteins when compared with E.coli protein data base, thereby showing expression of a protein foreign to host strain. Amino acids analyses showed that the cyanophycin produced by recombinant E.coli contained aspartic acid and arginine, and small amount of lysine, in the ratio: 1.05: 1: 0.2 (mass basis), thereby confirming cyanophycin biosynthesis. Experiments for high cyanophycin synthesis was performed at shake flask and 2-L fermentor level using recombinant BL21 (DE3) pLysS, LB broth as carbon and nutrient source, and casamino acids as primer. The maximum yield of cyanophycin obtained in flask level was 7.6% of cell dry mass, and the yield increased to 12.6% of cell dry mass at 2-L fermentor level. Cyanophycin is also referred to as “structural granules” because of substructures visible in electron micrographs. Phase contrast photomicrograph was able to depict cyanophycin inclusions in the cytoplasm, and transmission electron microscopy depicted finer details inside cell after IPTG induction.

Expressing foreign proteins in E.coli is a major challenge because they often tend to develop into unsolvable and inactive proteins. They aggregate into so called  inclusion bodies which prevent expression of the protein. This problem might be avoided by fusing the gene of the foreign protein with a soluble protein called solubility tags, which  function is to enhance the solubility of the foreign protein. This report investigates whether Sterol Carrier Protein-2 (SCP-2) could function as a solubility tag. The experiment was carried out by fusing SCP-2 to two recombinant proteins, Green fluorescent protein (GFP) and a form of chloroamphenicol acetyl transferase (CATΔ9). The gene fusion was then inserted into a pET-15 vector and transformed into  the E.coli strain BL21(DE3) to be expressed. The results obtained from Western blot and PageBlue staining indicates that SCP-2 does not enhance the solubility of GFP or CATΔ9 since neither of them was expressed.  Furthermore, previous studies have shown that GFP can in fact be expressed  usingmaltose binding protein (MBP) as a solubility tag. Unfortunately, no success has been made regarding CATΔ9. In conclusion, regarding the results from this report, SCP-2 does not function as a solubility tag. However, further studies should be carried out on SCP-2 with more experiments before rejecting the possibility to use SCP-2 as a solubility tag.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

La lysine décarboxylase (LDC) appartient à une famille d'enzymes décamériques dépendantes du cofacteur PLP qui sont connus pour catalyser la réaction transformant la L-Lysine en cadavérine tout en consommant un proton. Dans les entérobactéries comme Escherichia coli, nous trouvons deux paralogues, LdcI et LdcC. LdcI permet de faire face à la bactérie au conditions hostile de pH acide lors du passage à travers du tract gastro-intestinale. LdcC est produite pendant la phase stationnaire et aussi quand les bactéries font face aux traitements antibiotiques. La cadavérine produite par les LDCs est connue pour protéger les bactéries du stress oxydant. Cela s’explique par le fait que la cadavérine bloque les porines de la membrane externe, réduisant ainsi la perméabilité des molécules responsables du stress acides et oxydant. L'activité des LDCs chez E. coli est coordonnée avec la réponse stringente qui est mise en place lorsque les microorganismes sont dans des conditions pauvres en nutriments, afin d’éviter l’épuisement intracellulaire de la L-Lysine nécessaire pour la synthèse des protéines. Cependant, cette inhibition peut être levée par la formation d'un complexe en forme de cage avec son partenaire RavA, permettant ainsi aux bactéries de faire face aux stress multiples. Etant donné que la réponse au stress est importante pour que les bactéries puissent exhiber leur pathogénicité, nous nous sommes demandés si la bactérie opportuniste Pseudomonas aeruginosa pourrait employer LdcA pour contrer des conditions de stress qui ont déjà été décrites pour LdcI chez les entérobactéries. Au cours de ma thèse, nous avons abordé cette question en utilisant différentes approches complémentaires. Tout d'abord, nous avons utilisé des fusions promoteur-gène et de l'analyse par Western-blot pour déterminer les conditions dans lesquelles le gène ldcA a été exprimé et sa protéine synthétisée. Nous avons pu observer que ldcA est exprimé sur la phase stationnaire de croissance dans des conditions aérobies en milieux riches et également pendant des conditions anaérobies de respiration avec nitrate. Nous avons également confirmé que l'expression de ldcA est régulée par ArgR et elle est induite complètement lorsque l’acide aminé L-arginine est présente dans le milieu de croissance. Même si nous avons trouvé que les conditions de stress n'induisent pas l'expression de ldcA, nous avons obtenu de nouvelles données suggérant que d'autres mécanismes de régulation tels que le système de quorum sensing dépendant des quinolones (PQS) pourraient être impliqués dans l'expression de ldcA. En utilisant des souches mutantes de ldcA et son complémentée, nous avons évalué si LdcA était impliqué dans la réponse au stress acide et oxydatif. Bien que les données obtenues à l'aide des expériences dans notre laboratoire et des technologies à haut débit (Biolog) aient révélé que LdcA ne présente pas les mêmes fonctions que LdcI, nous avons découvert que la cadavérine produite par LdcA est nécessaire pour la croissance en milieu minimal avec L- Glutamate comme source de carbone. Nous avons également examiné si la présence de LdcA modifie la résistance aux antibiotiques et nous montré que les rends moins persistants face aux carbenicillines. Enfin, en combinant l'analyse phylogénétique et structurelle, nous avons découvert que LdcA appartient à un sous-groupe différent de LDCs bactériennes. Les alignements de séquences montrent que les résidus clés nécessaires pour lier le ppGpp ne sont pas présents dans le site de liaison prédit ce qui a été confirmer par l'analyse biochimique. Notre travail montre que, malgré le fait que LdcA catalyse la même réaction enzymatique et partage les mêmes caractéristiques structurelles que LdcI et LdcC, elle ne joue pas le même rôle que ses homologues. Son rôle est lié aux effets physiologiques de la cadavérine et à la relation entre la L-lysine et le catabolisme de la L-arginine. / The lysine decarboxylase (LDC) belongs to a family of decameric PLP-dependent enzymes that catalyse the reaction transforming L-Lysine into cadaverine while consuming a proton. They are known to be involved in polyamine metabolism and during acid and oxidative stress responses.In enterobacteria like Escherichia coli, two paralogs are present, LdcI and LdcC. LdcI takes part in acid stress response by buffering bacterial cytoplasm. LdcC is produced during stationary phase and also when bacteria face fluoroquinolone treatment. The cadaverine produced by LDCs is known to scavenge reactive oxygen species (ROS) and is capable of blocking outer membrane proteins, thus reducing the permeability of molecules responsible for acid and oxidative stresses. The activity of the LDCs from E. coli is coordinated with the stringent response (nutrient starvation) in order to prevent intracellular L-Lysine depletion. The stringent response signal molecule ppGpp is able to bind directly to LDCs and inhibit their enzymatic activity. However, the inhibition of the LdcI can be prevented by the formation of a cage-like complex with its partner RavA allowing bacteria to face the challenge of both acid and nutrient stresses.Since mechanisms allowing bacteria to counter stress challenges are important for displaying full virulence, we wondered if the opportunistic bacterium Pseudomonas aeruginosa could be using LdcA to counter stress conditions that have already been described for LdcI in enterobacteria. During my PhD, we addressed this question by using different but complementary approaches.First of all, we used promoter-gene fusions and western-blot analysis to determine the conditions in which ldcA was expressed and its product synthesized. We could observe that ldcA is expressed on stationary phase under aerobic conditions in rich media and also during nitrate-respiring anaerobic conditions. As previously described in literature, we also confirmed that ldcA expression is regulated by ArgR and fully induced when L-Arginine is present in the growth medium. Even though we found out that acid and oxidative stress conditions do not induce the expression of ldcA, we obtained new data suggesting that other regulation mechanisms such as the quinolone signal system (PQS) could be involved in ldcA expression.In paralell, we constructed an ldcA mutant and its complemented strain to understand whether LdcA was involved in acid and oxidative stress response. Although the data obtained by using manual screenings and high-throughput technologies (Biolog) revealed that LdcA is not displaying the same functions as LdcI, we discovered that the cadaverine produced by LdcA is needed for full growth fitness when growing in minimal medium using L-glutamate as carbon source. Since slow growing phenotypes are linked to heightened bacterial persistence and because cadaverine has been shown to reduce the persisters population, we also examined if the presence of LdcA is modifying the amount of persisters during carbenicillin treatment. Our data has confirmed that this is indeed the case.Finally, by combining phylogenetic and structural analysis, we discovered that LdcA belongs to a different subgroup of bacterial LDCs. Sequence alignments show that key residues needed for binding ppGpp are not present in the predicted binding site which also suggests that the enzymatic activity is not inhibited by this molecule. And biochemical analysis has confirmed that this is indeed the case as it is the case for Arginine decarboxylases.Our work shows that, in spite of the fact that LdcA catalyses the same enzymatic reaction and shares the same structural fold than LdcI and LdcC, it is not implicated in acid stress or oxidative stress responses. Its role is linked to physiological effects of cadaverine and to the relationship between L-lysine and L-Arginine catabolism.

O DNA de plasmídeos naturais de uma cepa de B. stearothermophilus foi clivado com a enzima de restrição Hind III e os fragmentos resultantes foram ligados com T 4 DNA ligase ao vetor pBR 322 (Bolívar et. al., 1977) que já havia sido previamente tratado com Hind III e fosfatase alcalina. A terça parte desta mistura de ligação foi usada para transformar células de E. coli HB 101. Foram obtidos cerca de 3.500 transformantes, dos quais 46% eram recombinantes. Duas cepas que mostraram caracter amilolítico consideravelmente maior que a doadora do gene continham o plasmídeo pBR 322 com uma inserção de 5.4 Kb. O mapa de restrição, tratamento com a enzima BAL 31 e, sucessivas subclonagens (Silva, E.F. et. al.,1986)mostraram que o gene que codifica e expressa a enzima amilolítica está contido em um fragmento de 2 Kb. A enzima produzida pelas células transformadas tem peso molecular de 60.000, é estabilizada por Ca+2, tem um ótimo de temperatura de 72ºC e retém 90% da atividade original após aquecimento a 85°C por 1 hora. Estes resultados, em conjunto com a análise dos produtos de hidrólise desta enzima em cromatografia de papel, sugerem que foi clonada a alfa - amilase de B. stearothermophilus em células de E. coli. Células de duas cêpas de B. subtilis, IA 289 e BD 241 foram transformadas respectivamente com os plasmídeos p USP 33.2 (Silva, E.F. et al., 1986;1987) e p BU 217 ami 2 (Silva, E.F. & Pueyo, M.T.,1988) para produzir em ambos os casos colônias fortemente amiloliticas. Os mecanismos pelos quais as 2 cêpas passaram a apresentar o fenótipo AMY + , são provavelmente diferentes. / The DNA of natural plasmids. from a B. stearothermophilus strain was cut with Hind III endonuclease and the resulting fragments were joined with T4 DNA ligase to the vector pBR 322 (Bolivar et al. , 1977), which had previously been treated with Hind III and alkaline phosphatase. One-third of the ligation mixture was used to transform E. coli HB 101 cells. It was obtained about 3.500 transformants, which included 46% recombinans. Two strains displayng amylolytic act ivity remarkably higher than the donor gene strain, harbored the plasmid pBR 322 with an insertion of 5.4 kb. The restriction map, Bal 31 treatment and successive subcloning (Silva, E.F, et al.,1986) showed that the entire gene which codifies and allows the expression of the amylolytic enzyme is contained in a 2 Kb fragment. The enzyme has a molecular weight of 60.000, is stabilized by Cata, has a temperature optimum at 72°C and retains 90% of the original activity after heating for 1h at 85°C. These features, together with the analysis of hydrolysis produts carried on paper chromatography , suggests that we succeded in cloning the amylase from B. stearothermophilus in E. coli cells. Cells from two B. subtilis strains, IQ 289 and BD 241 were transformed with the plasmid sp USP 33.2 (Silva E. F. et al., 1986 1987) and pBU 271 ami 2 (Silva, E. F. & Pueyo, M.T., 1988) , and produce in both strains, amyiolytic colonies. The methods in which the two strains have got the AMY + fenotype, may be very different.