<p>Mycobacterium abscessus (Mab) is a non-tuberculous mycobacterium that is highly resistant to many antibiotics. Mab causes pulmonary infections in immunocompromised individuals. The presence of efflux pumps to pump out antibiotics and its ability to form biofilms makes Mab a virulent pathogen. Studies have been done on Mab antibiotic tolerance but there are still a lot of gaps in knowledge about the effects of efflux pump inhibitors (EPIs) on antibiotic resistance and lipid biosynthesis in this bacterium during biofilm formation. In this study, we investigated the effects of the EPIs chlorpromazine (CPZ), 1-(1-naphthylmethyl)-piperazine (NMP), thioridazine (TRZ), Phenylalanine-arginine β-naphthylamide (PBN) and plumbagin (PLU) on antibiotic resistance, efflux, biofilm formation and lipid biosynthesis associated with log-phase growth and biofilm formation in Mab. We used the resazurin assay to determine the minimum inhibitory concentration (MIC) of the EPIs. We investigated the effects of the EPIs during biofilm-forming growth conditions on the MICs of antibiotics such as clarithromycin, amikacin, cefoxitin, ciprofloxacin which are the frontline antibiotics used to treat non-tuberculous mycobacterial infections. We also assessed the effects of the EPIs on the accumulation and efflux activities of the Mab cells through ethidium bromide (EtBr) assay. Furthermore, we evaluated the effects of the EPIs at sub-MIC concentrations on Mab biofilm formation under normoxic and hypoxic conditions. We utilized metabolic radiolabeling methods using 14C-palmitic acid and 14C- acetic acid which are precursors of lipid biosynthesis and analyzed lipids by silica-thin layer chromatography and autoradiography. We observed that Mab cells developed higher tolerance to the EPIs in a biofilm-stimulating medium. Furthermore, a decrease in the MICs of antibiotics was observed in the presence of the EPIs. Also, in the presence of the EPIs, there was less efflux activity within the Mab cells. In addition, EPIs inhibited biofilm formation significantly. We also noticed that NMP and PBN inhibited 14C-palmitic acid and 14C- acetic acid incorporation into polar lipids such as glycopeptidolipids, trehalose monomycolate, phosphatidylethanolamine, phosphatidylglycerol/cardiolipin, phosphatidylinositol mannosides at specific tested conditions. Our findings suggest that the EPIs inhibited the activities of the efflux pumps associated with the efflux of the antibiotics and lipid biosynthesis involved in biofilm formation. In conclusion, the results from this study gives insights on possible therapeutic opportunities.</p>

Multidrug efflux pumps are membrane proteins that actively transport foreign objects out of cells. The active efflux of these pumps is a critical self-defense mechanism that enables the survival of bacteria under hostile environments. Efflux pump AcrB is a member of the Resistance-Nodulation-Division (RND) super family. In E. coli, it associates with periplasmic protein AcrA and outer membrane channel TolC to extrude a variety of noxious compounds out of cell from both the cytoplasm and the periplasm. My dissertation research focused on two aspects of this multidrug efflux pump: the oligomerization process during the biogenesis of AcrB and its degradation. Oligomerization is an important aspect of the structure and function for many proteins and has been the subject of many studies. However, most of such studies focused on soluble proteins. The oligomerization process of membrane proteins, including AcrB, is rarely explored. In chapter 2, the co-assembly of AcrB variants co-expressed in the same cell was used as a tool to investigate the assembly of AcrB trimers during its bio-production. By monitoring the portion of pure trimers containing only one type of subunit and hybrid trimers containing a mixture of the two kinds of subunits, it was found that the oligomerization of membrane proteins is not a random process as the formation of pure trimer is favored. In chapter 3, the GALLEX system was used to monitor AcrB oligomerization in cells under the native condition. Previously GALLEX has only been used to monitor the oligomerization of small transmembrane proteins. By constructing a series of fusion proteins with different linker length between LexA and AcrB, and optimizing inducer concentration, we finally developed a system that could be used to differentiate AcrB trimers of different oligomerization affinities. While chapters 2 and 3 focus on the trimerization of AcrB, a critical step of its biogenesis, chapters 4 and 5 focus on its life time and degradation. In chapter 4, the life time of AcrB was measured by incorporating non-natural amino acid azidohomoalanine (AHA) into protein translation. Using this method, it was determined that that the half-life of both AcrA and AcrB in E. coli were six days. The surprisingly long lifetime of these detoxification proteins might represent a strategy by the bacteria to conserve energy and maximize their competition niche for survival in a hostile environment. In chapter 5, the degradation process of ssra tagged AcrB was investigated. In-vivo degradation test showed that properly inserted AcrB can be digested after addition of ssra-tag to its C-terminus. It was found that cytoplasmic unfoldase-protease complex ClpXP and chaperone SspB are involved in the degradation. In vitro assay is still being optimized to quantitatively analyze the activity of ClpXP in the degradation of AcrB.

Multiple drug resistance (MDR) has been a severe issue in treatment and recovery from infection.Gram-negative bacteria intrinsically exhibit higher drug tolerance than Gram-positive microbes. In this thesis, two proteins involved in Gram-negative bacterial MDR were studied, AcrB and SurA. Resistance-nodulation-cell division pump AcrAB-TolC is the major MDR efflux system in Gram-negative bacteria and efficiently extrudes a broad range of substances from the cells. To study subtle conformational changes of AcrB in vivo, a reporter platform was designed. Cysteine pairs were introduced into different regions in the periplasmic domain of the protein, and the extents of disulfide bond formation were examined. Using this platform, an inactive mutant, AcrB∆loop, was created that existed as a well-folded monomer in vivo. Next, random mutageneses were performed on a functionally compromised mutant, AcrBP223G, to identify residues that restored the function loss. The mechanism of function restoration was examined. SurA is a periplasmic molecular chaperone for outer membrane biogenesis. Deletion of SurA decreased outer membrane density and bacterial drug resistance. The dependence of SurA function on structural flexibility and stability was examined. In addition, the effect of molecular crowding on SurA interaction with its outer membrane protein substrates was examined.

Chemotherapeutic treatment of Pseudomonas aeruginosa, a Gram-negative opportunistic pathogen, is substantially challenged by several membrane-spanning, multidrug-efflux pumps of the three-component RND family. Of these pumps, MexXY-OprM contributes to the intrinsic resistance of this organism by exporting clinically relevant antibiotics, most notably the ribosome-targeting aminoglycosides. Overproduction of MexXY-OprM is the most common mechanism providing pan-aminoglycoside resistance to P. aeruginosa cystic fibrosis clinical isolates. The mexXY genes are located in an operon, the expression of which is induced by ribosome-targeting antimicrobials. The mexXY operon is negatively regulated by MexZ, a repressor protein encoded by the divergently-transcribed gene mexZ. A second gene, PA5471, is also induced by ribosome-targeting antibiotics and is required for antibiotic induction of mexXY expression. One possibility is that PA5471 interacts with MexZ to alleviate repression of mexXY, thereby providing a mechanism for PA5471-dependent drug inducibility of mexXY. PA5471 interaction with MexZ was confirmed using a bacterial two-hybrid assay. To identify residues/regions of PA5471 important for interaction with MexZ, random chemical mutagenesis of the mexZ and PA5471 genes was carried out and the effects of these mutations on interaction of their protein products was assessed using the bacterial two-hybrid assay. Mutations of PA5471 that compromised interaction with MexZ included P68S, G76C, R216C, R221W, R221Q, G231D, and G252S, which occur within or in close proximity to a predicted surface-exposed α-helix of a PA5471 structural model that may contribute to the MexZ-interaction domain. Representative mutations P68S, G76C, R216C and R221W were placed into the chromosome of P. aeruginosa to assess their impact on drug-inducible mexXY expression. All of these mutations significantly reduced mexX upregulation in the presence of spectinomycin, where mutations R216C and R221W resulted in the near complete ablation of this antibiotic induction. These data suggest that PA5471 acts as a direct antirepressor of MexZ and that this interaction is key to mexXY upregulation in response to ribosome-targeting induction signals. / Thesis (Master, Microbiology & Immunology) -- Queen's University, 2013-02-26 13:32:39.307

La maladie d’Alzheimer (MA) touche environ 35 millions de personnes à travers le monde ce qui en fait la première forme de démence. Le principal facteur de risque de cette maladie est l’âge et donc, avec l’augmentation de l’espérance de vie, on pense que ce nombre va doubler d’ici 2025. C’est donc un problème de santé publique. Le mécanisme à l’origine de la MA n’est pas connu, et il n’existe aucun traitement permettant de guérir ou de ralentir efficacement l’évolution des symptômes. Il est donc urgent de mieux comprendre cette maladie afin de proposer aux patients des traitements efficaces. Certaines études épidémiologiques ont montré que l’hypercholestérolémie est un facteur de risque pour la MA. A l’inverse, la prise de statines diminuerait les risques de développer cette maladie. De plus, des études réalisées in vitro et in vivo ont montré que le cholestérol peut favoriser la production des peptides amyloïdes qui sont à l’origine de la formation des plaques, caractéristiques de la MA. Enfin, certaines études ont montré qu’une baisse des niveaux de HDL peut être un facteur de risque pour la MA. Tous ces résultats nous amènent à penser que le métabolisme du cholestérol puisse être altéré dans la MA. La PON1 est une enzyme associée aux HDL. Elle possède des propriétés anti-inflammatoires et anti-oxydantes. Certaines études ont montré que son activité est réduite dans la MA, et que certains de ses polymorphismes sont associés à cette maladie. La PON1 pourrait donc jouer un rôle dans la MA. Mon projet de thèse a pour but d’investiguer si l’efflux du cholestérol est altéré dans la MA, et de comprendre le mécanisme à l’origine de cette éventuelle altération. Il vise également à étudier la PON1 dans la MA. Nos résultats montrent que l’efflux du cholestérol est diminué dans la MA. Cette baisse est due à une perturbation de la fonctionnalité des HDL et non, à une baisse de l’expression du transporteur ABCA1 qui est impliqué dans la première étape de l’efflux du cholestérol. Nos résultats suggèrent que la baisse de fonctionnalité des HDL soit due à leur oxydation, et à une altération de leur structure. Notre étude montre que l’activité paraoxonase est réduite de manière non significative dans la MA supposant donc, une légère baisse de la fonctionnalité de la PON1. De plus, les polymorphismes 192Q/R et 55L/M de la PON1 ne sont pas associés à cette maladie. Enfin, nos résultats montrent que les HDL des personnes Alzheimer sont plus sensibles à l’oxydation probablement du fait d’une baisse de la fonctionnalité de la PON1. En conclusion, dans la MA, les HDL sont oxydés et leur structure est altérée ce qui entraîne une baisse de l’efflux du cholestérol. Une altération de la fonctionnalité de la PON1 pourrait expliquer, en partie, la sensibilité des HDL à l’oxydation.

Burkholderia pseudomallei et Burkholderia mallei sont respectivement les agents biologiques responsables de la mélioïdose et de la morve. Pour déterminer si les échecs thérapeutiques étaient dus à l'émergence d'une résistance acquise durant le traitement antibiotique, nous avons sélectionné des souches de Burkholderia thailandensis, modèle d'étude, B. pseudomallei et B. mallei, avec différents antibiotiques : le chloramphénicol, la doxycycline et le triméthoprime-sulfaméthoxazole. Les Burkholderia ont montré qu'elles étaient capables de développer une multirésistance in vitro en réponse à chaque antibiotique utilisé dans le traitement oral de la mélioïdose ou de la morve. Pour comprendre les mécanismes de résistance impliqués, nous avons étudié les aspects moléculaires et génétiques de la résistance chez B. thailandensis par des méthodes protéomiques et transcriptomiques. Nous avons développé une méthode pour quantifier l'expression des gènes de pompes d'efflux par RT-PCR quantitative après normalisation sur plusieurs gènes de référence. Ces méthodes nous ont permis d'identifier la surproduction séquentielle de trois pompes d'efflux de type RND : BpeAB-OprB, AmrAB-OprA et BpeEF-OprC, toutes induites par le chloramphénicol ou la doxycycline chez les souches multirésistantes. L'étude de mutants déficients en pompe d'efflux nous a permis de mieux appréhender les relations étroites entre ces trois pompes et a confirmé que l'efflux actif était le principal mécanisme impliqué cette résistance induite. / Burkholderia pseudomallei and Burkholderia mallei are respectively the causative agents of melioidosis and glanders. To determine whether treatment failures were due to the emergence of acquired resistance during antibiotic treatment, we selected strains of B. pseudomallei, B. mallei, and Burkholderia thailandensis, used as a study model of these two pathogenic bacteria, with structurally unrelated antibiotics: chloramphenicol, doxycycline and trimethoprim-sulfamethoxazole. We showed that Burkholderia were able to develop multidrug resistance in vitro in response to each of theses antibiotics used in the oral treatment of melioidosis and glanders. To understand the resistance mechanisms involved, we studied the molecular and genetic aspects of resistance in B. thailandensis by proteomic and transcriptomic methods. We have developed a method to quantify efflux pumps gene expression by quantitative RT-PCR after normalization with several reference genes. These methods allowed us to identify sequential overproduction of three RND efflux pumps: BpeAB-OprB, AmrAB-OprA and BpeEF-OprC, all induced by chloramphenicol or doxycycline in multiresistant strains. The study of mutants respectively defective in one of these efflux pumps has allowed us to better understand the close relationship between these three pumps and confirmed that active efflux acted as a major mechanism involved in the induced resistance.

Objetivos: O objetivo deste trabalho foi desenvolver modelo farmacocinético (popPK) populacional para descrever simultaneamente as concentrações das fluoroquinolonas (levofloxacino – LEV e ciprofloxacino – CIP) no plasma, pulmão e próstata na presença e ausência do inibidor da P-gp tariquidar (TAR) visando determinar a contribuição desse transportador de efluxo na distribuição tecidual desses antimicrobianos. Método: Para alcançar este objetivo as seguintes etapas foram realizadas: i) foi validado o método analítico de HPLC-fluorescência para quantificação de CIP em amostras de plasma e microdialisado; ii) foram estabelecidas as condições para microdiálise para o CIP e as taxas de recuperação in vitro, por diálise e retrodiálise, e em tecido pulmonar e prostático in vivo por retrodiálise; iii) foi avaliada a farmacocinética do LEV após administração a ratos Wistar via i.v. bolus e por nebulização intratraqueal na dose de 7mg/kg na ausência e após administração prévia de TAR (15 mg/Kg i.v.); iv) foi desenvolvido um modelo popPK para prever as concentrações do LEV simultaneamente no plasma, pulmão e próstata após administração intravenosa e intratraqueal na presença e ausência do TAR; v) foi desenvolvido o modelo popPK para descrever as concentrações de CIP simultaneamente no plasma, pulmão e próstata após administração a ratos Wistar da dose de 7 mg/kg i.v. bolus na presença e ausência de TAR (15 mg/kg i.v.); vi) Para ambos os fármacos os dados foram avaliados por análise não-compartimental e modelados por modelo de quatro compartimentos modificado, com ajuda do software NONMEN®. Resultados e Conclusões. i) Método analítico foi desenvolvido e validado com sucesso para quantificação de CIP em HPLC/fluorescência mostrando-se linear na faixa de 10–2000 ng/mL em plasma e 5–1000 ng/mL em microdialisado com coeficientes de determinação (r2) superiores a 0,99. Os valores obtidos de erro padrão relativo para ensaios de precisão intra e inter-dia foram entre 8,8 e 6,0 %, para microdialisado entre 11,1 e 7,4 % para plasma, respectivamente. Os valores de exatidão foram 86,1% entre 114.3% para microdialisado e 85,6% entre 108,2% para plasma; ii) A avaliação do CIP por microdiálise mostrou recuperação concentração independente (0,25 - 1,5 μg/mL). Além disso, não houve diferença entre as recuperações obtidas por diálise e retrodiálise para o mesmo fluxo. No fluxo selecionado para os experimentos (1,5 μL/min) as recuperações médias por diálise e retrodiálise foram 23,0 ± 2,8% e 22,8 ± 1,6 %, respectivamente. A recuperação relativa das sondas in vivo foi de 11,3 ± 1,9 e 13,1 ± 2,7 % para pulmão e próstata, respectivamente; iii) A análise dos perfis plasmáticos e teciduais LEV após dose intravenosa do grupo controle (sem TAR) mostrou boa penetração tecidual na próstata (ƒT = 0,68) e no pulmão (ƒT = 0,69). Para a mesma via de administração, o grupo TAR mostrou uma penetração praticamente inalterada para o pulmão (ƒT = 0,81) e um aumento de mais de 2 vezes na penetração prostática (ƒT= 1,64). Na dose intratraqueal houve um aumento significativo na biodisponibilidade para o grupo TAR (F = 0,86) em relação ao controle (F = 0,4). Nessa via de administração foi detectado um aumento significativo na exposição (ASC) do pulmão ao LEV no grupo TAR demonstrando que o transporte por efluxo no pulmão é mais relevante quando o fármaco é administrado pela via intratraqueal; iv) Para o LEV, o modelo popPK de quatro compartimentos foi capaz de descrever simultaneamente os dados no plasma, pulmão e próstata na presença e ausência do TAR. Além disso, o modelo para administração intravenosa foi estendido e adaptado para administração intratraqueal. Foi possível analisar o impacto do transporte por efluxo sobre a penetração tecidual do LEV por diferentes vias de administração utilizando o modelo popPK; v) A avaliação do perfil farmacocinético plasmático do CIP após administração intravenosa, na presença e ausência de TAR, demonstrou diferença significativa entre todos os parâmetros calculados por análise não-compartimental, exceto para a constante de velocidade de eliminação (= 0,05). Em relação à penetração tecidual do CIP na próstata e pulmão, não houve alteração significativa nos parâmetros de eliminação e exposição tecidual do fármaco na presença do inibidor de efluxo TAR ( = 0,05), demonstrando que o transporte por efluxo possui papel minoritário no processo de distribuição do fármaco para os tecidos estudados. O modelo popPK de quatro compartimentos foi capaz de descrever as concentrações plasmáticas totais, livres no pulmão e próstata em presença e ausência de TAR, simultaneamente; vi) O modelo popPK desenvolvido permitiu o estudo mais profundo do processo de distribuição do LEV e do CIP no pulmão e próstata. / Objectives: The aim of this study was to develop a population pharmacokinetic model (popPK) able to simultaneously describe fluoroquinolones (levofloxacin – LEV and ciprofloxacin – CIP) concentrations in plasma, lung and prostate in the presence and absence of the inhibitor of P-gp tariquidar (TAR) to determine the contribution of this efflux transporter on the tissue distribution of these antimicrobials. Methods: To achieve this goal the following steps were taken: i) An analytical method by HPLC-fluorescence was developed and validated for CIP analysis in plasma and microdialysate samples; ii) microdialysis conditions were established for CIP including determination of in vitro relative recovery by dialysis and retrodialysis. The relative recovery was also determined in vivo, in lung and prostate, by retrodialysis; iii) LEV pharmacokinetics was evaluated after intravenous (i.v.) bolus and intratracheal (i.t.) administration of 7 mg/kg dose alone and following TAR administration (15 mg/kg i.v.) to Wistar rats; iv) a popPK model was developed to describe and predict LEV concentrations in plasma, lung and prostate following i.v. and i.t. dosing with and without TAR co-administration; v) the popPK model developed was used to describe CIP concentrations in plasma, lung and prostate after i.v. bolus administration of 7 mg/kg in presence and absence of TAR; vi) For both drugs non-compartmental analysis was performed besides data modeling by four compartment model using NONMEN®. Results and Conclusions i) The analytical method was developed and successfully validated for quantification of CIP by HPLC/fluorescence. The method was linear in the range of 10-2000 ng/mL in plasma and 5-1000 ng/mL in tissues microdialysate samples with coefficients of determination (r2) higher than 0.99. The relative standard error (RSD) obtained for intra and inter-day precision were lower than 8.8% and 6.0% for microdialysate and lower than 11.1 and 7.4% for plasma, respectively. The accuracy was 86.1% to 114.3% for microdialysate and 85.6 to 108.2 % for plasma samples; ii) the evaluation of CIP microdialysis probes relative recovery in vitro showed that the recovery was concentration independent (0.25 to 1.5 μg/mL). In addition, there was no statistical difference between the recoveries determined by dialysis and retrodialysis at the same flow rate. Using the selected flow rate (1.5 μL/min) the recoveries by dialysis and retrodialysis were 23.0 ± 2.8% and 22.8 ± 1.6%, respectively. CIP relative recoveries in vivo by retrodialysis were 11.3 ± 1.9 and 13.1 ± 2.7% for lung and prostate, respectively; iii) the analysis of LEV plasma and tissues concentration-time profiles after i.v. dosing showed a good tissue penetration of LEV in the prostate (ƒT = 0.68) and lung (ƒT = 0.69). For the same route of administration, TAR group showed virtually the same penetration into lung (ƒT = 0.81) and an increase of over 2 fold in drug levels in prostate (ƒT = 1.64). For the i.t. dose, there was a significant increase on LEV bioavailability for TAR group (F = 0.86) compared to control (F = 0.4). Furthermore, a significant increase was detected on lung exposure to LEV for TAR group indicating that efflux transport in the lung is more relevant when the drug is administered by the i.t. route; iv) For LEV, a four compartment model was able to describe the data simultaneously in plasma, lung and prostate in the presence and absence of TAR. Moreover, the intravenous model was extended to adapt the intratracheal dosing route. The popPK model allowed to analyze the impact of efflux transport on tissue LEV penetration of different routes of administration; v) the evaluation of plasma CIP profiles after i.v. dosing with and without TAR showed a significant difference in all parameters determined by non-compartmental analysis in the TAR group, except the elimination rate constant (α = 0.05). The CIP tissue penetration in prostate and lung, no significant difference was observed in tissues exposure and elimination rate when TAR was present demonstrating that efflux transporter play a minor role on CIP distribution to tissues investigated (α = 0.05). The popPK model with four compartments was able to describe CIP concentrations in plasma, lung and prostate in the presence and absence of TAR, simultaneously; vi) the popPK model developed allowed a more detailed investigation of LEV and CIP distribution process in lung and prostate.

In this dissertation, I employed a combination of in vitro and in silico techniques to characterize two bacterial membrane proteins: the MtrE efflux conduit from Neisseria gonorrhoeae and the AmtB ammonia transporter from Escherichia coli. MtrE is an efflux conduit located in the outer membrane of N. gonorrhoeae. It can form a protein complex with MtrC (a periplasmic adapter) and MtrD (an inner membrane active transporter). This tripartite machinery actively extrudes a diverse set of compounds from the periplasm to the exterior of the bacteria. The MtrCDE complex and homologous tripartite efflux pumps are amongst the major contributors to the emergence of super-resistant Gram-negative bacteria. The efforts to abolish the activity of these protein complexes have so far concentrated on the inhibition of the active pump, but had little clinical success to date. I carried out a combination of Planar Lipid Bilayer electrophysiology experiments, Molecular Dynamics simulations, and protein homology modeling on the MtrE and the MtrC proteins. I characterized the MtrE conduit as being slightly cation selective, as opposed to other homologous proteins which exhibit a much stronger selectivity. Additionally, I demonstrated that the opening of MtrE is modulated by the binding of the adapter protein MtrC. These results have a critical importance because they imply that the tripartite pump activity can be diminished not only by the MtrD inhibition, but by targeting the interface between MtrC and MtrE. The AmtB ammonia transporter is a protein embedded in the inner membrane of E. coli in situations of growth-limiting low levels of ammonia. The transport has been shown to be electrogenic in similar orthologous proteins, but the exact mechanism has so far remained elusive. The only inner pathway that has been identified in AmtB is lined by hydrophobic amino acids and for this reason it has been proposed to carry neutral ammonia. My Molecular Dynamics simulations reveal the opening of a water wire separated from the well known hydrophobic pore. This finding suggests a transport mechanism in which the neutral ammonia and the proton travel along different conduits. My simulations on several AmtB mutants gave additional important insights into the protein function. Notably my results explain the structural determinants of the switching of substrate from NH4+ to K+ observed for the H168D/H318E double mutant. Lastly, I identified several 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphoglycerol (POPG) binding sites on AmtB, which could possibly contribute to modulating the transport. Importantly, these in silico experiments were guided and confirmed by Solid Supported Membrane-electrophysiology measurements carried out in collaboration with the Javelle group. The data presented in this Thesis highlights the strength of a coordinated approach, in which experimental and computational findings direct, and integrate with each other.

abstract: Synechocystis sp. PCC 6803 is a readily transformable cyanobacteria used to study cyanobacterial genetics, as well as production of biofuels, polyesters, and other industrial chemicals. Free fatty acids are precursors to biofuels which are used by Synechocystis cells as a means of energy storage. By genetically modifying the cyanobacteria to expel these chemicals, costs associated with retrieving the products will be reduced; concurrently, the bacteria will be able to produce the products at a higher concentration. This is achieved by adding genes encoding components of the Escherichia coli AcrAB-TolC efflux system, part of the resistance-nodulation-division (RND) transporter family, to Synechocystis sp. PCC 6803. AcrAB-TolC is a relatively promiscuous multidrug efflux pump that is noted for expelling a wide range of substrates including dyes, organic solvents, antibiotics, and free fatty acids. Adding components of the AcrAB-TolC multidrug efflux pump to a previously created high free fatty acid producing strain, SD277, allowed cells to move more free fatty acids to the extracellular environment than did the parent strain. Some of these modifications also improved tolerance to antibiotics and a dye, rhodamine 6G. To confirm the function of this exogenous efflux pump, the genes encoding components of the AcrAB-TolC efflux pump were also added to Synechocystis sp. PCC 6803 and shown to grow on a greater concentration of various antibiotics and rhodamine 6G. Various endogenous efflux systems have been elucidated, but their usefulness in expelling products currently generated in Synechocystis is limited. Most of the elucidated pumps in the cyanobacteria are part of the ATP-binding cassette superfamily. The knowledge of the resistance-nodulation-division (RND) family transporters is limited. Two genes in Synechocystis sp. PCC 6803, slr2131 and sll0180 encoding homologs to the genes that encode acrB and acrA, respectively, were removed and the modifications resulted in changes in resistance to various antibiotics and a dye and also had an impact on free fatty acid secretion. Both of these deletions were complemented independently with the homologous E. coli gene and the resulting cyanobacteria strains had some of the inherent resistance restored to chloramphenicol and free fatty acid secretion was modified when compared to the wild-type and a high free fatty acid producing strain. / Dissertation/Thesis / Doctoral Dissertation Microbiology 2018

Mature trees help to offset urban area problems caused by impervious pavement. Trees in paved areas remain unhealthy due to a poor root zone environment. The objective of this experiment was to test if soil under pervious concrete, with greater water and gas infiltration, would be more beneficial to existing mature trees during urban development. Root activity, root growth and soil chemistry of American sweetgum under standard concrete, pervious concrete and no concrete were measured. Soil CO2 efflux rates and soil CO2 concentrations were extremely high under both concrete treatments. Soil under standard concrete had lower oxygen concentrations than soil under pervious concrete and control treatments, particularly under wet conditions. There was no pavement effect on soil water content or soil chemistry. Under control treatment standing live root length was greater than under both concrete treatments. There were no major differences in soil conditions between impervious and pervious concrete treatments. The soil under the plots, a Ships clay, with very low permeability may have prevented soil water infiltration. Likely this overrode any potential treatment effects due to porosity of the concrete. To obtain root zone benefits out of pervious concrete, a different base soil with a higher permeability would be a better alternative.