Asymptomatic bacteriuria (ABU) represents the long term bacterial colonization of the urinary tract, frequently caused by Escherichia coli (E. coli), without typical symptoms of a urinary tract infection (UTI). To investigate characteristics of ABU E. coli isolates in more detail, the geno- and phenotypes of eleven ABU isolates have been compared. Moreover, consecutive in vivo re-isolates of the model ABU strain 83972 were characterized with regard to transcriptomic, proteomic and genomic alterations upon long term in vivo persistence in the human bladder. Finally, the effect of the human host on bacterial adaptation/evolution was assessed by comparison of in vitro and in vivo-propagated strain 83972. ABU isolates represent a heterologous group of organisms. The comparative analysis of different ABU isolates elucidated the remarkable genetic and phenotypic flexibility of E. coli isolates. These isolates could be allocated to all four major E. coli phylogenetic lineages as well as to different clonal groups. Accordingly, they differed markedly in genome content, i.e., the genome size as well as the presence of typical UPEC virulence-associated genes. Multi locus sequence typing suggested that certain ABU strains evolved from UPEC variants that are able to cause symptomatic UTI by genome reduction. Consequently, the high E. coli genome plasticity does not allow a generalized view on geno- and phenotypes of individual isolates within a clone. Reductive evolution by point mutations, DNA rearrangements and deletions resulted in inactivation of genes coding for several UPEC virulence factors, thus supporting the idea that a reduced bacterial activation of host mucosal inflammation promotes the ABU lifestyle of these E. coli isolates. Gene regulation and genetic diversity are strategies which enable bacteria to live and survive under continuously changing environmental conditions. To study adaptational changes upon long term growth in the bladder, consecutive re-isolates of model ABU strain 83972 derived from a human colonisation study and from an in vitro long term cultivation experiment were analysed with regard to transcriptional changes and genome rearrangements. In this context, it could be demonstrated that E. coli, when exposed to different host backgrounds, is able to adapt its metabolic networks resulting in an individual bacterial colonisation strategy. Transcriptome and proteome analyses demonstrated distinct metabolic strategies of nutrients acquisition and energy production of tested in vivo re-isolates of strain 83972 that enabled them to colonise their host. Utilisation of D-serine, deoxy- and ribonucleosides, pentose and glucuronate interconversions were main up-regulated pathways providing in vivo re-isolates with extra energy for efficient growth in the urinary bladder. Moreover, this study explored bacterial response networks to host defence mechanisms: The class III alcohol dehydrogenase AdhC, already proven to be involved in nitric oxide detoxification in pathogens like Haemophilus influenzae, was shown for the first time to be employed in defending E. coli against the host response during asymptomatic bacteriuria. Consecutive in vivo and in vitro re-isolates of strain 83972 were also analysed regarding their genome structure. Several changes in the genome structure of consecutive re-isolates derived from the human colonisation study implied the importance of bacterial interactions with the host during bacterial microevolution. In contrast, the genome structure of re-isolates from the in vitro long term cultivation experiment, where strain 83972 has been propagated without host contact, was not affected. This suggests that exposure to the immune response promotes genome plasticity thus being a driving force for the development of the ABU lifestyle and evolution within the urinary tract. / Asymptomatische Bakteriurie (ABU) stellt eine bakterielle Infektion der Harnblase über einen langen Zeitraum dar, die häufig von Escherichia coli hervorgerufen wird, ohne dass typische Symptome einer Harnwegsinfektion auftreten. Um die Charakteristika von ABU E. coli Isolaten genauer zu untersuchen, wurden die Geno- und Phänotypen von 11 ABU-Isolaten verglichen. Außerdem wurden in mehreren aufeinanderfolgenden in vivo-Reisolaten des Modell-ABU Stammes 83972 die Veränderungen im Transkriptom, Proteom und Genom während einer langfristigen Persistenz in der menschlichen Blase charakterisiert. Schließlich wurde der Effekt des menschlichen Wirtes auf die bakterielle Adaptation durch einen Vergleich von in vitro- mit in vivo-kultivierten Stämmen abgeschätzt. ABU-Isolate stellt eine heterogene Gruppe von Organismen dar. Diese können den vier phylogenetischen Hauptgruppen von E. coli sowie unterschiedlichen klonalen Gruppen zugeordnet werden. Dementsprechend unterscheiden sie sich erheblich bezüglich der Zusammensetzung des Genomes, der Genomgröße und auch der Ausstattung mit UPEC-typischen Virulenz-assoziierten Genen. Multi-Lokus-Sequenz-Typisierung legt nahe, dass bestimmte ABU Stämme sich durch Genomreduktion aus UPEC Stämmen entwickelt haben, die eine Harnwegsinfektion mit charakteristischen Symptomen auslösen konnten. Folglich erlaubt die hohe Genomplastizität von E. coli keine generalisierte Betrachtung einzelner Isolate eines Klons. Genomreduktion über Punktmutationen, Genom-Reorganisation und Deletionen resultierte in der Inaktivierung einiger Gene, die für einige UPEC Virulenz-Faktoren kodieren. Dies stützt die Vorstellung, dass eine verminderte bakterielle Aktivierung der Entzündung der Wirtsschleimhaut den Lebensstil von ABU (bei diesen E. coli-)Isolaten fördert. Genregulation und genetische Diversität sind Strategien, die es Bakterien ermöglichen unter sich fortlaufend ändernden Bedingungen zu leben bzw. zu überleben. Um die anpassungsbedingten Veränderungen bei einem langfristigen Wachstum in der Blase zu untersuchen, wurden aufeinanderfolgende Reisolate, denen eine langfristige in vivo-Kolonisierung im menschlichen Wirt beziehungsweise eine in vitro-Kultivierung vorausgegangen ist, im Hinblick auf Veränderungen Genexpression und Genomorganisation analysiert. In diesem Zusammenhang konnte gezeigt werden, dass E. coli in der Lage ist, seine metabolischen Netzwerke verschiedenen Wachstumsbedingungen anzupassen und individuelle bakterielle Kolonisierungsstrategien entwickeln kann. Transkriptom- und Proteom-Analysen zeigten verschiedene metabolische Strategien zur Nährstoffbeschaffung und Energieproduktion bei untersuchten in vivo-Reisolaten vom Stamm 83972, die es ihnen ermöglichen, den Wirt zu kolonisieren. Das Zurückgreifen auf D-Serin, Deoxy- und Ribonucleoside sowie die bidirektionale Umwandlung zwischen Pentose und Glucuronat waren hoch-regulierte Stoffwechselwege, die die in vivo-Reisolate mit zusätzlicher Energie für ein effizientes Wachstum in der Blase versorgen. Zudem wurden in dieser Studie die Netzwerke für eine Reaktion auf Abwehrmechanismen des Wirtes erforscht: Erstmals wurde hier die Rolle der Klasse-III-Alkoholdehydrogenase AdhC, bekannt durch ihre Bedeutung bei der Entgiftung von Stickstoffmonoxid, bei der Wirtsantwort während einer asymptomatischen Bakteriurie gezeigt. Aufeinanderfolgende in vivo- und in vitro-Reisolate vom Stamm 83972 wurden ebenfalls bezüglich ihrer Genomstruktur analysiert. Einige Veränderungen in der Genomstruktur der aufeinanderfolgenden Reisolate, die von einer humanen Kolonisierungsstudie stammen, implizieren die Bedeutung einer Interaktion der Bakterien mit dem Wirt bei der Mikroevolution der Bakterien. Dagegen war die Genomstruktur von Reisolaten eines langfristigen in vitro-Kultivierungsexperiments, bei dem sich der Stamm 83972 ohne Wirtskontakt vermehrt hat, nicht von Veränderungen betroffen. Das legt nahe, dass die Immunantwort eine Genomplastizität fördert und somit eine treibende Kraft für den ABU Lebensstil und die Evolution im Harnwegstrakt ist.

Polyketide (PK) und nichtribosomale Peptide (NRP) sind zwei grosse Klassen von Naturstoffen, die eine grosse Vielfalt hinsichtlich ihrer Struktur und Funktion aufweisen. Sie werden von einer Reihe von Bakterien, Pilzen und Pflanzen als Sekundärmetabolite produziert und besitzen eine Vielzahl pharmakologisch wichtiger Aktivitäten, wie z.B. antimikrobielle, antimykotische, antitumorale oder antiparasitische Wirkungen. Ein Grossteil der bakteriellen Produzenten findet sich im Phylum Firmicutes, innerhalb der Gattungen Bacillus, Streptomyces und Mycobacterium. In E. coli sind Polyketide und nichtribosomale Proteine von eher geringer Bedeutung, mit Ausnahme der Siderophore Enterobactin und Yersiniabactin. Unerwartet war daher die Identifizierung eines neuen PKS/ NRPS-Gencluster in verschiedenen E. coli-Stämmen. Das 2006 durch NOUGAYRÈDE et al. zuerst beschriebene Colibactin-Gencluster kodiert für ein hybrides System aus modularen Polyketidsynthasen und nichtribosomalen Peptidsynthetasen sowie für zusätzliche editierende Enzyme und einen möglichen transkriptionellen Regulator (ClbR). Das Produkt der PKS/NRPS-Synthasen, Colibactin, übt in vitro einen zytopathischen Effekt (CPE) auf Säugerzelllinien aus. Die zytopathische Aktivität Colibactins zeichnet sich u.a. durch die Induktion von Doppelstrangbrüchen in der DNA der eukaryotischen Zellen aus. Darüber hinaus kommt es zu einer Unterbrechung des Zellzyklus in der G2-Phase nach einer transienten in vitro Infektion mit Colibactin-positiven Bakterienstämmen. Im Rahmen der vorliegenden Arbeit war besonders die weitere Aufklärung der Struktur des Colibactinclusters sowie die regulatorischen Mechanismen, die die Exression des hybriden nichtribosomalen Peptid-Polyketids von Interesse. Eine Transkriptionsanalyse führte zur Identifizierung der Transkriptionsstartpunkte der meisten relevanten Gene des Colibactinclusters. Basierend auf diesen neugewonnenen Informationen war eine Sequenzanalyse der upstream-Bereiche der Gene möglich, in deren Ergebnis neben den Elementen eines Sigma70-abhängigen Promotors, putative Bindestellen für mehrere Transkriptionsfaktoren identifiziert wurden. Untersuchungen zur Regulation der Colibactinsynthese zeigten, dass die Expression der Colibactin-Gene sowohl unter Kontrolle des Transkriptionsfaktors H-NS als auch des Colibactin-spezifischen Regulators ClbR stehen. Neben der Aufklärung der Struktur und Regulation der Colibactin-Gene bestand das Ziel dieser Arbeit in der Optimierung der Synthese des nichtribosomalen Peptid-Polyketids. Hierfür durchgeführte Expressionstudien zeigten einen Einfluss von Fettsäuren und Indol sowie von der Sauerstoffverfügbarkeit auf die Promotoraktivität einzelner Gene des Colibactin-Genclusters. Darüberhinaus konnte das pks-Genclusters erfolgreich in Pseudomonas putida KT2440 transferiert werden sowie der Nachweis der Funktionsfähigkeit Colibactins in diesem Wirtsorganismus nachgewiesen werden. Wenngleich die Stabilität des für diesen Zweck konstruierten Shuttle-Vektors nicht von Dauer ist, konnte gezeigt werden dass Pseudomonas putida prinzipiell als Wirtssystem für die Realisierbarkeit der heterologen Expression von Colibactin, geeignet ist. Zusätzlich zur Strukturanalyse des pks-Clusters und den Studien zur Expression der Colibactin-Gene befasste sich die hier vorliegende Arbeit mit der Fragestellung nach der biologischen Funktion Colibactins. Phänotypische Untersuchungen zeigen sowohl eine Beeinflussung der Eisenaufnahme als auch der Biofilmbildung durch das nichtribosomale Peptid-Polyketid. Dies sind die ersten Hinweise die zur Aufklärung der Funktion Colibactins beitragen könnten. / Polyketides (PK) and nonribosomal peptides (NRP) are two large classes of natural products showing a great variety in structure and function. They are produced as secondary metabolites by a range of bacteria, fungi and plants and exhibit a wealth of pharmacologically important activities, including antimicrobial, antifungal, antitumor or antiparasitic properties. The vast majority of bacterial producers belong to the phylum Firmicutes, especially to the genera Bacillus, Streptomyces and Mycobacterium. With the exception of the siderophores enterobactin and yersiniabactin polyketides and nonribosomal peptides are of minor relevance within E. coli. Therefore unexpected was the identification of a new PKS/ NRPS gene cluster in several E. coli strains. The colibactin gene cluster being described for the first time in 2006 by NOUGAYRÈDE et al. is coding for a hybrid system of modular polyketide synthases and nonribosomal petid synthetases as well as editing enzymes and a putative transcriptional regulator (ClbR). The product of these PKS/ NRPS synthases, termed colibactin, induces in vitro a cytopathic effect (CPE) on mammalian cell lines. The cytopathic activity of colibactin is characterized by the induction of double strand breaks in the DNA of eukaryotic cells as well as the arrest of the cell cycle in G2 phase after transient infection with E. coli strains expressing colibactin. In context of this thesis especially the elucidation of the regulation of clb operon transcription and the organisation of transcriptional units within the colibactin-encoding genomic island were of main interest. A transcriptional analysis led to the identification of the transcriptional starting points of most of the relevant genes within the colibactin cluster. Based on these newly obtained information it was possible to perform a sequence analysis of the upstream regions of the genes resulting in the detection of sigma70 depending promoter elements and several putative transcription factor binding sites. Studies on the regulation of the colibactin synthesis could also demonstrate that the expression of colibactin genes are under control of the transcription factor H-NS as well as the colibactin specific regulator ClbR. Beside the studies concerning the structure and regulation of colibactin genes optimization of the nonribosomal peptid-polyketid was object of this work. Therefor performed expression analysis showed an influence of fatty acids and indole, as well as the oxygen availability on the promoter activities of single genes within the colibactin gencluster. Further investigations belonging the transcriptome and the proteome of the Colibactin expressing strain E. coli Nissle 1917 showed an over all influence of Colibactin synthesis on the amino acid and carbohydrate metabolism of this strain. Further more a successful transfer of the pks gene cluster into Pseudomonas putida KT2440 was carried out as well as the demonstration of functionality of colibactin in this host organism. Even though long term stability of the constructed shuttle vector was not given it was shown that Pseudomonas putida is a suitable host for realizing the heterologous expression of colibactin. Additionally to the structural analysis of the pks cluster and the studies on expression of the colibatin genes this thesis questioned on the biological function of Colibactin. Phenotypical examination showed an influence of the iron upake as well as on biofilm formation due to the nonribosomal peptid – polyketide. These are the first evidences that could contribute the elucidation of Colibactin function.

Escherichia coli wird als der häufigste Erreger von Harnwegsinfektionen des Menschen beschrieben. Um die Krankheit auslösen zu können, benötigen die Bakterien ganz bestimmte Eigenschaften, die als Virulenzfaktoren bezeichnet werden und durch die sie sich von apathogenen Stämmen unterscheiden. Der uropathogene E. coli Stamm 536 (O6:K15:H31) exprimiert verschiedene Virulenzfaktoren wie a-Hämolysin, die Adhäsine S-, P-related (Prf) und Typ 1-Fimbrien sowie das Kapselantigen K15. Außerdem wurden auch Enterobaktin- und Yersiniabaktin-Produktion sowie Serumresistenz nachgewiesen. Die Ausprägung der Virulenz hängt unter anderem mit dem Vorhandensein von Pathogenitätsinseln (PAI I536-V536) zusammen, die mit seltenen tRNA-Genen assoziiert sind. Die Deletion der Inseln PAI I536 und PAI II536 führt zum Verlust der Virulenz und zur Zerstörung der entsprechenden tRNA Gene. Es wurde festgestellt, daß die leuX-kodierte tRNA5Leu, die mit der PAI II536 assoziiert ist, einen Einfluß auf die Expression von Typ 1-Fimbrien sowie auf die Serumresistenz, Motilität und die Enterobaktin- und a-Hämolysin-Produktion hat. Ziel dieser Arbeit war die Analyse der Bedeutung der leuX-kodierten tRNA5Leu und der Pathogenitätsinseln I536 und II536 für die Expression von Proteinen sowie Typ 1-Fimbrien bei dem E. coli Stamm 536. Dazu wurde zunächst eine Proteomanalyse durchgeführt. Mit der Hilfe von 2D-Gelen wurde der Einfluß von leuX-kodierten tRNA5Leu und PAI I536 und PAI II536 auf die Expression verschiedener Proteine im E. coli Stamm 536 (PAI I536+, PAI II536+, leuX+) und seinen Mutanten: 536-21 (PAI I536-, PAI II536-, leuX-), 536D102 (PAI I536+, PAI II536+, leuX-) und 536R3 (PAI I536-, PAI II536-, leuX+) untersucht. Mit Hilfe von präparativen Gelen konnten in den zytosolischen Fraktionen 39 Unterschiede in der Expression von Proteinen nachgewiesen werden. Von diesen differentiell exprimierten Proteinen wurden 37 mit Hilfe von MALDI-TOF-MS identifiziert. Die zwei weiteren Proteine konnten nicht identifiziert werden. In den Kulturüberstandsfraktionen der untersuchten Stämme konnten drei Unterschiede in der Expression von Proteine nachgewiesen werden. Außerdem wurde der Einfluß der leuX-kodierten tRNA5Leu auf die Expression der Membranproteine der jeweiligen Stämme untersucht. Zu diesem Zwecke wurden sowohl ein- wie auch zweidimensionale Gele durchgeführt. Mit Hilfe von zweidimensionalen Gelen konnten zwischen den untersuchten Stämmen mehrere Unterschiede in der Expression von Proteinen festgestellt werden. Dabei konnten vier Unterschiede in der Proteinexpression detektiert werden. Mit Hilfe der 2 D-Gelelektrophorese wurde ein leuX-abhängiges Protein (YgaG), das ein Analogon des LuxS-Proteins ist, gefunden. Das LuxS-Protein ist ein Bestandteil des "Quorum sensing"-Systems von Vibrio harveyi und wird in ähnlicher Form bei vielen Bakterien beschrieben. Sein Einfluß auf die Pathogenität wurde bei vielen Bakterien beschrieben. Aus diesem Grund wurde eine ygaG-Mutante im E. coli Stamm 536 hergestellt. Anschließend wurde der Einfluß des YgaG-Proteins auf die Expression von Virulenzfaktoren überprüft und ein Proteomvergleich zwischen dem Wildtyp Stamm E. coli 536 und der ygaG-Mutante wurde durchgeführt. Die Expression der bekannten Virulenzfaktoren wurde von YgaG nicht beeinflußt. Weiterhin wurden Transkriptionsfusionen zwischen den Promotoren der fimB- und fimE-Gene mit dem promotorlosen ß-Galaktosidase-Gen (lacZ) konstruiert. Es sollte damit die Frage beantwortet werden, ob die leuX-kodierte tRNA5Leu als potentieller Regulator die Transkription von Typ 1-Fimbrien beeinflußt. Es konnten jedoch keine signifikanten Unterschiede in der Transkription von fimB und fimE zwischen dem Wildtyp Stamm 536 und der leuX-Mutante 536D102 festgestellt werden. / Escherichia coli is one of the common causes of urinary tract infections. Pathogenic E. coli differ from non-pathogenic E. coli variants by the presence of certain virulence factors which contribute to their ability to cause disease. The uropathogenic E. coli strain 536 (O6:K15:H31) is able to produce different virulence factors such as a-hemolysin, fimbrial adhesins (P-, type 1 and S-fimbriae) and specific capsules. Furthermore the bacteria produce iron uptake systems like enterobactin and yersiniabactin and have the capacity to survive in human serum. The development of pathogenicity is connected to pathogenicity islands (PAI I536-V536). All four PAIs are associated with tRNA genes. The deletion of PAI I536 and PAI II536 results in the truncation of the associated tRNA gene and loss of virulence. The deletion of PAI II536 results in the truncation of the leuX gene. PAI II536 as well as leuX deletion mutants show a reduced production of type 1 fimbriae, flagellae and of the iron uptake system enterobactin. Furthermore, they show delayed hemolysin production and a reduced serum resistance. Additionally, the leuX-encoded tRNA5Leu may also regulate the expression of various other genes. The aim of this work was the further characterization and analysis of the role of PAI I536/II536 and of the tRNA5Leu for protein expression as well as for the regulation and expression of two site-specific recombinases, FimB and FimE, of the E. coli strain 536. Firstly, the protein expression patterns of E. coli 536 and different derivatives were studied. Differences in the protein expression pattern of the wild-type strain E. coli 536, its mutants 536-21 (PAI I536-, PAI II536-, leuX-), 536D102 (PAI I536+, PAI II536+, leuX-) and strain 536R3 (PAI I536-, PAI II536-, leuX+) were analyzed by two-dimensional polyacrylamide gel electrophoresis. With the help of preparative 2 D-gel electrophoresis 39 differentially expressed intracellular proteins could be identified. The identities of 37 proteins have been determined by MALDI-TOF mass spectrometry in Halle. Two of these proteins show no matches in sequence databases. The preparations of the culture supernatants resulted in 2 D proteinpatterns from which three protein spots whose expression is markedly altered in the different strains are identified. The influence of the tRNA5Leu on the expression of outer membrane proteins was also studied. Differences in the protein expression patterns of the wild-type strain 536 and the mutants were analyzed by one- and two-dimensional gel electrophoresis. The analysis of 2 D protein patterns of outer membrane preparation resulted in the detection of some proteinspots whose expression was altered in the different strain backgrounds. It was identified during the proteom analysis that tha expression of the YgaG protein was shown to be reduced in a leuX-negative background. This protein shows a strong homology to the LuxS protein of Vibrio harveyi. The LuxS protein is a component of the quorum sensing system. It has been shown that many bacteria express proteins similar to LuxS of V. harveyi and that quorum sensing may play a role in pathogenicity. To investigate whether YgaG contributes to the virulence of the E. coli strain 536, a ygaG deletion mutant was made. In addition, differences in the protein expression pattern of the wild-type strain E. coli 536 and the ygaG mutant were analyzed by 2 D gel electrophoresis. Whether YgaG contributes to the virulence of the E. coli Strain 536 has not be investigated. In order to get a deeper insight into the role of the tRNA5Leu as a potential regulator of the expression of the typ 1-fimbriae, a transcriptional fusion between the promotor area of fimB and fimE and the lacZ gene lacking its own promotor was constructed. However the transcription of fimB and fimE did not show any significant differences between the wild-type strain 536 and the leuX-mutant 536D102.

In this study pore forming proteins of the gram-negative bacteria B. burgdorferi, B. duttonii and E.coli were investigated. Therefore the study is subdivided into three parts. In the first part outer membrane preparation of three relapsing fever Borrelia were investigated. In the second part the putative TolC homologue BB0124 of B. burgdorferi, the Lyme borreliosis agent, was studied. In the last part the influence of point mutants within the greasy slide of the maltose specific porin (LamB) of E. coli were shown. In the first part of this study outer membrane preparations of three Borrelia relapsing fever strains have been studied for pore-forming activity in the black lipid bilayer assay. Histograms of conductance fluctuations were obtained from single-channel experiments with outer membrane preparations of B. hermsii, B. recurentis and B. duttonii. All strains had a different conductance fluctuation pattern with a broad range of single-channel conductance values varying from 0.5 nS – 11 nS. Common for all three strains was a high pore-forming activity at around 0.5 nS. Furthermore the proteins of the outer membrane of B. duttonii were separated by chromatographic methods. Some eluate fractions contained a channel-forming protein, which was forming stable channels with a single-channel conductance of 80 pS in 1 M KCl. Characterization of this channel showed that it is slightly anionic selective and voltage independent. The small single-channel conductance suggests that it is a specific pore. However, a substrate specificity could not be determined. In the second part, for the B. burgdorferi HB19 and p66 knock out strain HB19/K02, their outer membrane preparations were characterized in the black lipid bilayer assay. Comparing the histograms of single-channel conductions fluctuations of both strains showed no single-channel activity at 11.5 nS for the p66 knock out strain. This verifies earlier studies that P66 is a pore-forming protein in B. burgdorferi. Furthermore, one fraction obtained by anion exchange chromatography of the p66 knock out outer membrane protein preparation showed a uniform channel-forming activity with a single channel conductance of 300 pS. The electrophysically characterization of the 300 pS channel showed that it is not ionselective or voltage dependent. By mass spectrometry using peptide mass finger prints, BB0142 could be identified as the sole channel forming candidate in the active fraction. A BLAST search and a conserved domain search showed that BB0142 is a putative TolC homologue in B. burgdorferi. Furthermore the location of the bb0142 gene within the chromosome is in an operon encoding a multidrug efflux pump. In this study the expression of an outer membrane component of a putative drug efflux system of B. burgdorferi was shown for the first time. In the third part functional studies of the maltooligosaccharide-specific LamB channel were performed. The 3D-structure of LamB suggests that a number of aromatic residues (Y6, Y41, W74, F229, W358 and W420) within the channel lumen is involved in carbohydrate and ion transport. All aromatic residues were replaced by alanine (A) scanning mutagenesis. Furthermore, LamB mutants were created in which one, two, three, four and five aromatic residues were replaced to study their effects on ion and maltopentaose transport through LamB. The purified mutant proteins were reconstituted into lipid bilayer membranes and the single-channel conductance was studied. The results suggest that all aromatic residues provide some steric hindrance for ion transport through LamB. Highest impact is provided by Y6 and Y41, which are localized opposite to Y118, which forms the central constriction of the LamB channel. Stability constants for binding of maltopentaose to the mutant channels were measured using titration experiments with the carbohydrate. The mutation of one or several aromatic amino acids led to a substantial decrease of the stability constant of binding. The highest effect was observed when all aromatic amino acids were replaced by alanine because no binding of maltopentaose could be detected in this case. However, binding was again possible when Y118 was replaced by tryptophane (W). The carbohydrate-induced block of the channel function could also be used for the study of current noise through the different mutant LamB-channels. The analysis of the power density spectra of some of the mutants allowed the evaluation of the on- and off-rate constants (k1 and k-1) of carbohydrate binding to the binding-site inside the channels. The results suggest that both on- and off-rate constants were affected by the mutations. For most mutants k1 decreased and k-1 increased. / In dieser Studie wurden porenformende Proteine der gram-negativen Bakterien B. burgdorferi, B. duttonii und E. coli untersucht. Daher wurde die Arbeit in drei Teile untergliedert. Im ersten Teil wurden zuerst die Außenmembranpräparationen von drei Rückfallfieber auslösenden Borrelien Stämmen untersucht. Der zweite Teil der Arbeit behandelt das TolC homologe Protein BB0142 des Erreger der Lyme Borreliose, B. burgdorferi. Im letzten Teil wurde der Einfluss von Punktmutationen innerhalb der „greasy slide“ auf die zuckerspezifische Pore (LamB) von E. coli untersucht. Die Außenmembranpräparationen der drei Rückfallfieber Borrelien wurden auf porenformende Aktivität im Black Lipid Bilayer untersucht. Leitfähigkeitshistograme wurden nach Einzelkanalmessungen der Außenmembranepräparation von B. hermsii, B. recurrentis und B. duttonii erstellt. Alle Stämme zeigten ein anderes Leitfähigkeitsfluktuationsmuster, wobei die Werte der Einzelleitfähigkeit von 0,5 nS bis 11 nS variierten. Auffällig war das alle drei Stämme eine hohe Aktivität um 0,5 nS gemeinsan hatten. Darauffolgend wurden die Proteine der Außenmembran von B. duttonii durch verschiedene chromatographische Methoden aufgetrennt. Einige Eluatfraktionen enthielten ein kanalformendes Protein, das stabile Kanäle mit einer Einzelleitfähigkeit von 80 pS in 1 M KCl formt. Die Bestimmung der Kanaleigenschaften zeigte, dass er leicht anionenselektive und spannungsunabhängig ist. Die geringe Einzelleitfähigkeit suggeriert, dass der Kanal zudem eine spezifische Pore sein könnte. Bisher konnte aber keine Substratespezifität festgestellt werden. Im zweiten Abschnitt erfolgte die Untersuchung von Außenmembranpräparationen von B. burgdorferi HB19 und dem p66 knock-out Stamm HB19/K02 mit Hilfe des Black Lipid Bilayers. Im Vergleich der beiden erhalten Einzelleitfähigkeitshistogramme konnte gezeigt werden, dass die Einzelleitfähigkeit von 11,5 nS nicht mehr im p66 knock out Stamm vorkam. Dies belegt frühere Studien, dass P66 ein poreformendes Protein von B. burgdorferi ist. Durch Anionenaustauscher-Chromatographie der Außenmembranpräparation des p66 knock out Stammes wurde zudem eine Proteinfraktion mit einer einheitlichen porenformenden Aktivität von 300 pS in 1 M KCl erhalten. Die elektrophysikalische Charakterisierung der 300 pS Pore zeigte, dass der Kanal nicht ionenselektiv oder spannungsabhängig ist. Mit Hilfe von Massenspektrometrie und Peptidemassenbestimmung, konnte BB0142 als einziger möglicher kanalformender Kandidat identifiziert werde. Eine BLAST Suche sowie eine „conserved domain“ Suche zeigten, dass BB0142 ein putatives TolC homologes Protein von B. burgdorferi ist. Darüber hinaus ist das bb0142 Gen in einem Operon lokalisiert, das eine putative Efflux Pumpe codiert. In der vorliegenden Arbeit konnte das erste Mal gezeigt weden, dass in B. burgdorferi die Außenmembrankomponente einer Efflux Pumpe expremiert wird. Im dritten Teil wurden Funktionsstudien an dem zuckerspezifischen Kanal LamB durchgeführt. Die 3D Struktur zeigte, dass einige aromatische Aminosäuren (Y6, Y41, W74, F229, W358 und W420) innerhalb des Kanallumen in den Kohlenhydrat- und Ionentransport involviert sind. Alle aromatischen Reste wurden bei Punktmutation durch Alanin ersetzt. Zudem wurden LamB Mutanten erzeugt in denen ein, zwei, drei, vier oder fünf aromatische Reste durch Alanin ersetzt. Die aufgereinigten LamB Mutanten wurden im Black Lipid Bilayer untersucht und ihre Einzelleitfähigkeit bestimmt. Die Ergebnisse zeigen, dass alle aromatischen Reste eine gewisse sterische Hinderung beim Ionentransport durch LamB bewirken. Den größten Einfluss haben die Reste Y6 und Y41, die gegenüber dem Rest Y188 liegen, der die zentrale Verengung des LamB Kanal darstellt. Stabilitätskonstanten der Zuckerbindung in den Mutanten wurde durch Titrationsexperimente mit Maltopentaose und -heptaose bestimmt. Mutation an einem oder mehreren aromatischen Resten führt zu einer deutlichen Abnahme der Bindungsstabilitätskonstanten. Den stärksten Effekt zeigte die Mutante, in der alle aromatischen Reste durch Alanin ersetzt wurden. Es konnte dort keine Zuckerbindung mehr festgestellt werden. Die Bindung wurde wieder hergestellt nachdem Y118 durch ein Tryptophan ersetzt wurde. Durch das zuckerinduzierte Öffnen und Schließen des Kanals ergiebt sich ein Stromrauschen. Die Analyse des Rausch Spektrum von einigen Mutanten erlaubte die Bestimmung der Dissoziations- (k-1) und der Assoziationskonstaten (k1) der Zuckerbindung an der Bindestelle innerhalb des Kanals. Die Ergebnisse zeigten, dass die Geschwindigkeitskonstanten durch die Mutationen beeinflusst wurden. Für die meisten Mutanten sank der k1 Wert während der k-1 Wert anstieg.

Das Kerngenom pathogener Escherichia coli Isolate wird von zahlreichen variablen Regionen unterbrochen, die meist durch horizontalen Gentransfer erworben wurden und über das ganze Chromosom verteilt sind. Diese variablen Bereiche tragen häufig Gene für Virulenz- und Fitnessfaktoren und sind oftmals nur instabil in das Chromosom integriert. Um die Verbreitung variabler Bereiche, die insbesondere Virulenzfaktoren kodieren, innerhalb verschiedener klinischer Isolate näher untersuchen zu können, wurde im Rahmen dieser Arbeit ein spezieller DNA-Array entwickelt. Dieser enthielt zahlreiche Sonden für Gene, die für die Virulenz von verschiedenen Erregern der Gattung E. coli als auch der Untergruppe Shigella charakteristisch sind. Mit diesem "Pathoarray" wurde die Verbreitung von Virulenzgenen in unterschiedlichen E. coli Isolaten untersucht. Zusätzlich wurden Unterschiede im Kerngenom mit Hilfe eines kommerziell erwerbbaren DNA-Arrays bestimmt. Ein Vergleich des Kerngenoms von uropathogenen Stämmen mit Derivaten, bei denen Pathogenitätsinseln deletiert sind, bestätigte die Auffassung, dass der Deletion von Pathogenitätsinseln ein spezieller Mechanismus zu Grunde liegt, von dem das Kerngenom nicht betroffen ist. Das Kerngenom der untersuchten Stämme war prinzipiell sehr konserviert und unterschied sich lediglich durch wenige Gene aus Bakteriophagen. Die größten Unterschiede wurden bei Genen beobachtet, die zum variablen Teil des Genoms gehören und charakteristisch für das jeweilige Isolat waren. Mit Hilfe der DNA-Array Technologie lassen sich auch Änderungen von Expressionsprofilen studieren, die von Mutationen oder durch Umwelteinflüsse bedingt werden. Im zweiten Teil dieser Arbeit wurde durch Transkriptomanalysen das RfaH-abhängige Regulon untersucht, insbesondere im Hinblick auf solche Gene, die die Biofilmbildung beeinflussen. Beim Vergleich der Transkriptome von E. coli 536rfaH mit dem Wildtyp wurde eine signifikant erhöhte Expression von Antigen 43 festgestellt. Im E. coli K-12 Stammhintergrund konnte dieses Oberflächenprotein als Hauptfaktor für die RfaH-abhängige Biofilmbildung identifiziert werden. Das verkürzte LPS-Kernoligosaccharid im Stamm MG1655rfaH hatte ebenfalls einen großen Einfluss auf die verstärkte Biofilmbildung. Vermutlich verstärkte die verbesserte Präsentation von Agn43 durch ein verkürztes LPS die Biofilmbildung signifikant. Andere Oberflächenstrukturen, wie die Colansäure-Kapsel, zeigten keinen Effekt auf die Biofilmbildung von E. coli MG1655rfaH. Neben den Expressionsprofilen der Stämme 536 und 536rfaH bei 37 Grad C wurden auch die Expressionsprofile bei 30 Grad C sowie von Biofilmen analysiert. Prinzipiell konnten bei allen untersuchten Wachstumsbedingungen nur geringe Unterschiede zwischen 536 und 536rfaH festgestellt werden. Beim Vergleich der unterschiedlichen Wachstumsbedingungen (Temperatureffekt und planktonische Zellen vs. Biofilm) wurden jedoch deutliche Unterschiede beobachtet. Sowohl Gene des Kerngenoms als auch Gene von Pathogenitätsinseln waren temperaturabhängig reguliert. Bei E. coli Isolaten lassen sich neben genomischen Unterschieden auch phänotypische Unterschiede beobachten. Es wurde festgestellt, dass die Biofilmbildung von E. coli Isolaten abhängig von verschiedenen Faktoren und molekularen Mechanismen ist. Zudem konnte dargelegt werden, wie Unterschiede in der Zusammensetzung der äußeren Membran durch eine veränderte LPS-Struktur und die Expression von Adhäsinen die Biofilmbildung beeinflussen können. / Evolutionary adaptation is the driving force for the variability observed within genomes of all different Escherichia coli pathotypes. Beside the core genome, shared by all strains, also variable regions, which can be strain-specific, are scattered on the chromosome. These flexible regions mostly encode virulence factors as well as other fitness factors and are acquired through horizontal gene transfer but are also characterised by their instability. To shed a closer light on the distribution of these virulence factors among different clinical isolates, DNA-arrays were used for genome comparisons. One aim of this Ph.D. thesis was the design of a DNA-array with specific probes for typical virulence-related genes of pathogenic E. coli. With this tool in hand, the distribution of virulence genes among several E. coli isolates was analysed. In addition to virulence related genes, also differences in the core genome of well-known uropathogenic isolates of E. coli were detected using DNA-array technology. Furthermore, the core genome of different wildtype strains was compared with derivatives shown to have lost pathogenicity islands. The results confirmed the assumption that PAI deletion from core genome is a specific process and underlies a specific mechanism. The core genome itself was very conserved and the observed small differences related to genes derived from different bacteriophages. The majority of differences were detected for the flexible regions, which differ in a strain-specific manner. DNA-array technology is also a versatile tool to gain insights in changes of gene expression levels caused by gene function disorders or environmental differences. The expression profiling approach using DNA-arrays was employed in the second part of this thesis. In this work, the RfaH-related regulon was studied by transcriptome analysis. Besides the already known effect of RfaH on facilitated capsule, LPS and alpha-haemolysin expression, the focus was set on genes involved in biofilm formation. Comparison of the 536rfaH and the wildtype strain transcriptomes revealed a significant upregulation of agn43 transcript levels. In order to study the underlying mechanisms of RfaH-dependent increased biofilm formation, selected mutants of E. coli K-12 and 536 were generated and tested for their biofilm forming capacities. In MG1655rfaH we could show that Agn43 is the major factor leading to biofilm formation. In addition, this phenotype was dependent on the LPS core truncation coming along within rfaH deficient strains. In conclusion, these results demonstrated that the LPS core truncation leads to unshielding of Agn43 in this strain, thus supporting autoaggregation and biofilm formation. Other surface structures like colanic acid had no influence on this effect. Besides the expression profiles of strains E. coli 536 and 536rfaH at 37 degrees, also expression profiles at 30 degrees and in biofilms were analysed. Typical differences in either stage were characterised. In general, when comparing the expression profiles from wildtype and mutant the changes observed were very small. However, the influence of temperature and also the mode of growth (planktonic cells vs. biofilm) affected the expression profiles in both strains more severely. In conclusion, E. coli strains not only differ in their genotypes, moreover complex phenotypic differences are observable. The obtained phenotypic differences in biofilm formation were shown to be multifactorial. The phenotype was attributed to variances in the composition of the outer membrane. In this context, the influence of LPS structures and adhesin expression was pointed out.

E. coli isoladas de lesões de celulite aviária em frangos de corte produzem uma citotoxina, denominada ECVF (E. coli Vacuolating Factor), que causa intensa vacuolização citoplasmática em células aviárias, mas não em células mamárias. A importância de ECVF na patogenia da celulite foi avaliada neste estudo. ECVF purificado foi inoculado subcutaneamente em frangos de corte, e induziu sinais de inflamação nos tecidos subcutâneo, adiposo e conjuntivo. Em ensaios de citotoxicidade, foi verificado que ECVF induz alterações citoplasmáticas e nucleares que podem afetar diretamente o metabolismo celular, entre elas condensação da cromatina e fragmentação nuclear, intensa vacuolização citoplasmática e desorganização do citoesqueleto, conduzindo à apoptose. Também foi verificada interação de ECVF com proteínas de células aviárias, em detrimento das de células de mamíferos, sugerindo uma especificidade da toxina a este tipo celular. Nossos resultados, apoiados por dados de estudos anteriores, permitem sugerir um importante papel de ECVF na patogenia da celulite aviária. / E. coli isolated from cellulitis lesions in broiler chickens produce a citotoxin, called ECVF (Escherichia coli Vacuolating Factor), which causes intense cytoplasm vacuolization in avian cells, but not in mammalian cells. The importance of ECVF in the pathogenesis of avian cellulitis was assessed in this study. Purified ECVF was inoculated subcutaneously in broiler chickens, and induced signs of inflammation on subcutaneous, adipose and connective tissues. In citotoxicity assays, we verified that ECVF induced cytoplasmic and nuclear alterations, which can affect cellular metabolism directly, such as chromatin condensation and nuclear fragmentation, intense cytoplasm vacuolization, and disorganization of cytoskeleton, leading to apoptosis. It was also verified the interaction of ECVF with proteins of avian cells, instead of those from mammalian cells, suggesting the specificity of this toxin to this cells. Our results, supported by data from previous studies, suggest an important role of ECVF in the pathogenesis of avian cellulitis.

De todas as influências exógenas que possam alterar a microbiota intestinal, os antimicrobianos são capazes de causar as mais rápidas e drásticas mudanças. O impacto da exposição aos antimicrobianos na microbiota intestinal causa diminuição no número de microrganismos ou mesmo supressão, dependendo do antimicrobiano utilizado, da dose e do tempo de exposição. Assim, o objetivo deste estudo foi analisar de forma comparativa alguns microrganismos que compõem a microbiota fecal de crianças com e sem antibioticoterapia em idade escolar; bem como avaliar a susceptibilidade aos antimicrobianos e os genes de resistência envolvidos. Foram coletadas amostras fecais não diarreicas de 30 crianças sem antibiótico (controle) e 31 de crianças com antibioticoterapia. Na análise quantitativa foi observada redução no número de cópias por g/fezes de: Bifidobacterium spp., B. fragilis, C. perfringens, E. coli, M. smithii e do filo Firmicutes nas amostras das crianças com antibióticos em relação ao grupo controle, exceto para Lactobacillus spp. e P. distasonis que apresentaram quantificação maior no grupo antibióticos quando comparados com o controle. E. coli foi isolada em 26 (86,7%) crianças controles e em 23 (74,2%) tratadas com antibióticos. A resistência foi verificada para diversas drogas no grupo controle exceto para ciprofloxacina, meropenem e tigeciclina; entretanto o grupo com antibioticoterapia apresentou elevada resistência para todas as drogas avaliadas, caracterizando os isolados desse estudo como MDR. Todos os isolados do grupo controle e antibióticos albergaram diversos genes de resistência, entretanto o gene blaKPC foi o único não detectado nos isolados do grupo controle. Desta forma, nossos dados demonstram que a antibioticoteria causa alterações qualitativas e quantitativas na microbiota intestinal; além disso, a elevada resistência as diversas classes de antimicrobianos das cepas de E. coli, bem como a presença de diversos genes de resistência ressalta a importância de cepas comensais serem MDR e albergarem esses genes. / Of all the exogenous influences that may alter the intestinal microbiota, antimicrobial agents are able to cause the more rapid and dramatic changes. The impact of exposure to antimicrobial agents on intestinal microbiota causes a decrease in the number of certain genera and species, depending on the antimicrobial agent used, dose and duration of exposure. Thus, the aim of this study was to analyze comparatively some microorganisms that composing the fecal microbiota of children with and without antibiotic therapy in school age; and evaluates the antimicrobial susceptibility and resistance genes involved. Stool samples (not diarrhea) were collected of 30 children without antibiotic (control) and 31 children with antibiotic therapy. In quantitative analysis was observed decrease in the number of copies per g/feces: Bifidobacterium spp., B. fragilis, C. perfringens, E. coli, M. smithii and the phylum Firmicutes in samples of children with antibiotic therapy in relation to control group, except Lactobacillus spp. and P. distasonis that showed a higher quantification in the antibiotics group when compared with control group. E. coli was isolated in 26 (86.7 %) children controls and in 23 (74.2 %) children treated with antibiotics. The resistance was verified for several drugs in the control group except for ciprofloxacin, meropenem and tigecycline; however the group with antibiotic therapy showed high resistance to all drugs evaluated, characterizing isolates of this study as MDR. All isolates from control group and antibiotics harbored several resistance genes, however blaKPC gene was the only one not detected in isolates from the control group. Thus, our data demonstrate that the antibiotic therapy cause qualitative or quantitative changes in intestinal microbiota leading to a decrease in the diversity and the elimination of microorganisms; in addition, the high resistance the various classes of antimicrobial of the strains of E. coli, as well as the presence of several genes of resistance highlights the importance of commensal strains are MDR and harboring these genes.

Geok-yen Yeo. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1994. / Includes bibliographical references (leaves 225-233). / Title page --- p.i / Members of Thesis Advisory Committee --- p.ii / Abstract --- p.iii -iv / Acknowledgments --- p.v / Dedication --- p.vi / Table of Contents --- p.vii -xi / Chapter CHAPTER 1 --- INTRODUCTION --- p.1-31 / Chapter 1.1 --- General Introduction --- p.1 / Chapter 1.2 --- Fermentation --- p.1 / Chapter 1.3 --- Growth in Escherichia coli --- p.3 / Chapter 1.3.1 --- Aerobic growth in Escherichia coli --- p.3 / Chapter 1.3.2 --- The regulation of enzyme synthesis during cell metabolism --- p.7 / Chapter 1.3.3 --- Anaerobic growth in E. coli --- p.8 / Chapter 1.3.4 --- Anaerobic regulation by the transcriptional regulator Fnr --- p.12 / Chapter 1.3.5 --- "The case for ""Pasteur Control Proteins"" (PCP)" --- p.13 / Chapter 1.4 --- The family of alcohol dehydrogenases : An overview --- p.15 / Chapter 1.4.1 --- Molecular characteristics of alcohol dehydrogenases --- p.17 / Chapter 1.4.2 --- Residue conservation in alcohol dehydrogenases --- p.24 / Chapter 1.4.3 --- The effect of amino acid substitution on substrate specificity --- p.25 / Chapter 1.5 --- Alcohol dehydrogenases in bacteria --- p.28 / Chapter 1.5.1 --- Alcohol dehydrogenase in E. coli --- p.28 / Chapter 1.6 --- Aims of this study --- p.30 / Chapter CHAPTER 2 --- MATERIALS & METHODS --- p.32 -90 / Chapter 2.1 --- Bacterial strains --- p.32 / Chapter 2.2 --- Plasmids --- p.32 / Chapter 2.2.1 --- "Low copy number plasmid, pTJS75Km" --- p.32 / Chapter 2.2.2 --- "High copy number plasmid, pUC18" --- p.33 / Chapter 2.3 --- Bacterial culture media and solutions --- p.39 / Chapter 2.3.1 --- Luria Bertani (LB) medium --- p.39 / Chapter 2.3.2 --- L-Broth + MOPS --- p.39 / Chapter 2.3.3 --- "R medium, containing Triphenyltetrazolium chloride-ethanol (TTC-EtOH)" --- p.40 / Chapter 2.3.4 --- SOB and SOC media --- p.41 / Chapter 2.3.5 --- M9 Glucose medium --- p.42 / Chapter 2.3.6 --- Terrific Broth (TB) --- p.42 / Chapter 2.3.7 --- Rich Broth (RB) --- p.43 / Chapter 2.3.8 --- Antibiotic solutions --- p.43 / Chapter 2.4 --- Restriction endonucleases and other enzymes --- p.44 / Chapter 2.5 --- Isolation of chromosomal DNA --- p.45 / Chapter 2.5.1 --- Preparation of chromosomal DNA by spooling --- p.45 / Chapter 2.5.2 --- Preparation of chromosomal DNA by cesium chloride density gradient --- p.48 / Chapter 2.6 --- Isolation of plasmid DNA --- p.50 / Chapter 2.6.1 --- Large-scale preparation of plasmid by CsCl density gradient --- p.50 / Chapter 2.6.2 --- Small-scale preparation of plasmid DNA --- p.54 / Chapter 2.6.2. --- A Boiling method --- p.54 / Chapter 2.6.2. --- B Alkaline Lysis method --- p.55 / Chapter 2.6.3 --- Preparation of plasmid DNA by Qiagen columns --- p.56 / Chapter 2.7 --- Purification of DNA --- p.59 / Chapter 2.7.1 --- Ethanol precipitation --- p.59 / Chapter 2.7.2 --- Concentration and desalting using Centricon columns --- p.59 / Chapter 2.7.3 --- Purification of DNA by Geneclean procedure --- p.61 / Chapter 2.8 --- DNA cloning techniques --- p.63 / Chapter 2.8.1 --- Restriction endonuclease digestion --- p.63 / Chapter 2.8.2 --- Agarose-ethidium bromide gel electrophoresis --- p.65 / Chapter 2.8.2. --- A Gel loading buffer --- p.66 / Chapter 2.8.2. --- B Electro-elution of DNA --- p.67 / Chapter 2.8.3 --- Size fractionation --- p.68 / Chapter 2.8.3. --- A Salt gradient fractionation --- p.68 / Chapter 2.8.3. --- B Sucrose gradient --- p.70 / Chapter 2.8.4 --- Dephosphorylation of restriction-enzyme digested vector plasmid using calf intestinal phosphatase (CIP) --- p.71 / Chapter 2.8.5 --- Ligation of vector and insert --- p.72 / Chapter 2.8.6 --- Preparation of competent cells --- p.73 / Chapter 2.8.7 --- DNA transformation --- p.75 / Chapter 2.8.7.A --- By heat shock --- p.75 / Chapter 2.8.7.B --- By electroporation --- p.75 / Chapter 2.9 --- Screening for adhC transformants --- p.78 / Chapter 2.9.1 --- Screening for adhC clones --- p.78 / Chapter 2.9.2 --- Screening for pUC18 transformants --- p.79 / Chapter 2.10 --- Confirmation of adhC clones --- p.80 / Chapter 2.10.1 --- Reproduction of red colonies on R plates and antibiotic resistance --- p.80 / Chapter 2.10.2 --- T7 phage test for E. coli strains --- p.80 / Chapter 2.10.3 --- Plasmid size determination --- p.82 / Chapter 2.10.4 --- Re-transformation into E. coli host strains --- p.82 / Chapter 2.10.5 --- Physiological study of adhC clones --- p.83 / Chapter 2.10.6 --- Alcohol dehydrogenase assay --- p.84 / Chapter 2.11 --- The dye-binding method of protein determination --- p.87 / Chapter 2.12 --- Special procedures --- p.88 / Chapter 2.12.1 --- Generation of adh clones with deletions --- p.88 / Chapter 2.12.2 --- Sequencing reactions --- p.89 / Chapter CHAPTER 3 --- RESULTS: PART I Cloning and Restriction Mapping of the adhC mutation in a low copy number plasmid vector --- p.91 -122 / Chapter 3.1 --- Introduction: Cloning strategy --- p.91 / Chapter 3.2 --- Cloning of the adh mutation from strain CC2807B (an ADH overproducing mutant strain) in pTJS75Km --- p.93 / Chapter 3.2.1 --- Construction of the 'HK' clones --- p.93 / Chapter 3.3 --- Restriction mapping of the adh clones --- p.101 / Chapter 3.4 --- Subcloning the adhC insert --- p.110 / Chapter 3.4.1 --- Construction of plasmid pHK14 --- p.110 / Chapter 3.4.2 --- Construction of plasmid pHK15 --- p.115 / Chapter 3.4.3 --- Construction of plasmid pSS22 --- p.121 / Chapter 3.5 --- Remarks concerning the clones --- p.121 / Chapter CHAPTER 4 --- RESULTS:PART II Cloning and Sequencing of the adhC mutation in a high copy number plasmid vector --- p.123 -148 / Chapter 4.1 --- Introduction --- p.123 / Chapter 4.1.1 --- Choice of sequencing strategy --- p.123 / Chapter 4.1.2 --- An attempt to eliminate clone instability --- p.124 / Chapter 4.2 --- Subcloning of adh insert in pUC18 --- p.125 / Chapter 4.2.1 --- Study of adh clone EPR --- p.125 / Chapter 4.2.2 --- Re-construction of plasmid pEPR ( = pEE5) --- p.126 / Chapter 4.2.3 --- Construction of plasmids pEH2 and pEH3 --- p.127 / Chapter 4.2.4 --- Construction of a nested deletion library --- p.138 / Chapter CHAPTER 5 --- RESULTS : PART III Sequencing of the Mutation --- p.149 -177 / Chapter 5.1 --- Nucleotide sequencing --- p.149 / Chapter 5.2 --- Sequencing of the cloned adhC gene insert --- p.150 / Chapter 5.3 --- Analysis of the sequenced DNA by DNASIS computer software --- p.151 / Chapter 5.3.1 --- Search for codons associated with initiation and termination of transcription using the open reading frame (ORF) search --- p.151 / Chapter 5.3.2 --- Translation of the nucleotide sequence at the open reading frame (start 223 - end 2896) --- p.152 / Chapter 5.4 --- Search for DNA sequence homology with known DNA sequences --- p.152 / Chapter 5.4.1 --- Sequence homology of the structural gene (nucleotide # 223- #28%) : Two nucleotide changes revealed in DNA sequence of the structural gene adhE of Escherichia coli --- p.153 / Chapter 5.4.2 --- adhC mutation is due to changes in two amino acids --- p.153 / Chapter 5.4.3 --- The DNA sequence 5' of the mutated structural gene (upstream sequence) --- p.155 / Chapter 5.4.4 --- The DNA sequence 3' of the mutated structural gene (downstream sequence) --- p.156 / Chapter 5.5 --- Comparisons between the computer-predicted properties of the mutant and wild-type protein --- p.156 / Chapter 5.5.1 --- Prediction of the alcohol dehydrogenase protein secondary structure by the Robson Method --- p.156 / Chapter 5.5.2 --- Isoelectric point prediction --- p.156 / Chapter CHAPTER 6 --- RESULTS : PART IV Comparative Studies of Alcohol Dehydrogenase Expressionin adhC Strains and Clones --- p.178 -203 / Chapter 6.1 --- Introduction --- p.178 / Chapter 6.1.1 --- Basis for the alcohol dehydrogenase assay --- p.178 / Chapter 6.1.2 --- Choice of assay method --- p.179 / Chapter 6.1.3 --- Points to consider for ADH assay --- p.179 / Chapter 6.2 --- General growth characteristics of bacterial strains --- p.181 / Chapter 6.2.1 --- Plate cultures --- p.181 / Chapter 6.2.2 --- Overnight liquid cultures --- p.183 / Chapter 6.2.3 --- Batch liquid cultures --- p.183 / Chapter 6.2.4 --- ADH activity of strain CC2807B --- p.190 / Chapter 6.2.5 --- Comparison of ADH activity --- p.192 / Chapter 6.3 --- Investigating the mutated ADH enzyme --- p.197 / Chapter 6.3.1 --- Oxygen inactivation of the mutated enzyme --- p.197 / Chapter 6.3.2 --- Thermostability of the mutated enzyme --- p.201 / Chapter CHAPTER 7 --- DISCUSSION --- p.204 -220 / Chapter 7.1 --- Cloning of the adhC mutation --- p.204 / Chapter 7.1.1 --- Instability of clones in plasmid vector pUC18 --- p.204 / Chapter 7.1.2 --- Eliminating 'toxic' genes adjacent to adh locus --- p.207 / Chapter 7.1.3 --- Cloning in pTJS75Km low copy number vector --- p.208 / Chapter 7.2 --- DNA sequence of the adhC clones --- p.211 / Chapter 7.2.1 --- The basis for sequencing pUC 18-derived clones --- p.211 / Chapter 7.2.2 --- Homology to known alcohol dehydrogenases (ADH) sequences --- p.213 / Chapter 7.3 --- Findings concerning the adhC mutation --- p.217 / Chapter 7.3.1 --- How amino acid substitutions may affect an enzyme --- p.217 / Chapter 7.3.2 --- Physiological aspects of the bacterial cell due to the mutated enzyme --- p.218 / Chapter 7.4 --- Conclusions --- p.220 / APPENDICES --- p.221 -224 / REFERENCES --- p.225 -233

by Lee Kai Fai, Calvin. / Thesis (Ph.D.)--Chinese University of Hong Kong, 1995. / Includes bibliographical references (leaves 159-167). / ACKNOWLEDGMENTS --- p.i / ABSTRACT --- p.ii / CONTENTS --- p.iv / ABBREVIATIONS --- p.xi / Chapter CHAPTER ONE --- General Introduction --- p.1 / Chapter 1.1 --- The Phenomenon of Host-controlled Restriction --- p.1 / Chapter 1.2 --- Classification of Restriction and Modification Systems --- p.2 / Chapter 1.2.1 --- Type I Restriction-Modification Systems --- p.2 / Chapter 1.2.2 --- Type II Restriction-Modification Systems --- p.3 / Chapter 1.2.3 --- Type III Restriction-Modification Systems --- p.4 / Chapter 1.2.4 --- Type IV Restriction-Modification Systems --- p.5 / Chapter 1.3 --- Occurrence of Restriction-Modification Systems --- p.6 / Chapter 1.4 --- Effect of Methylation --- p.7 / Chapter 1.5 --- Alternation of Recognition Specificities --- p.7 / Chapter 1.5.1 --- Cross Protection by DNA Methyltransferase --- p.8 / Chapter 1.5.2 --- A-Assisted Restriction Endonuclease (RARE) Cleavage --- p.9 / Chapter 1.5.3 --- Site-specific Cleavage mediated by Triple-helix formation --- p.9 / Chapter 1.5.4 --- Site-specific Cleavage of Duplex DNA with a λ repressor- Staphylococcal Nuclease Hybrid --- p.10 / Chapter 1.5.5 --- Achilles' heel Cleavage --- p.10 / Chapter 1.5.6 --- Chimeric Restriction Endonuclease --- p.11 / Chapter 1.6 --- Cloning of Restriction and Modification Systems --- p.11 / Chapter 1.6.1 --- Selection based on Modification --- p.11 / Chapter 1.6.2 --- Other Cloning Strategies --- p.12 / Chapter 1.6.2.1 --- Sub-Cloning of Plasmids --- p.12 / Chapter 1.6.2.2 --- Selection based on Restriction --- p.13 / Chapter 1.6.2.3 --- Multi-step Cloning --- p.13 / Chapter 1.6.2.4 --- Cloning in AP1-200 and AP1-200-9 strain --- p.13 / Chapter 1.6.2.5 --- Direct Cloning of Restriction gene by 'endo-blue' method --- p.14 / Chapter 1.7 --- Genetic Location of Restriction-Modification Systems --- p.14 / Chapter 1.8 --- Sequences of Restriction-Modification Systems --- p.15 / Chapter 1.9 --- Catalytic Properties of Type II Restriction-Modification Systems --- p.17 / Chapter 1.10 --- Crystallography of Type II Restriction and Modification Enzymes --- p.19 / Chapter 1.11 --- Evolution of Type II Restriction and Modification Enzymes --- p.22 / Chapter 1.12 --- Aim of Study --- p.23 / Chapter CHAPTER TWO --- Materials and Methods --- p.24 / Chapter 2.1 --- Bacterial Strains --- p.24 / Chapter 2.2 --- General Techniques --- p.25 / Chapter 2.2.1 --- Phenol/Chloroform Extraction --- p.25 / Chapter 2.2.2 --- Ethanol Precipitation --- p.25 / Chapter 2.2.3 --- Spectrophotometry --- p.25 / Chapter 2.2.4 --- Restriction digestion of DNA --- p.26 / Chapter 2.2.5 --- Agarose Gel Electrophoresis of DNA --- p.26 / Chapter 2.2.6 --- Recovery of DNA fragment from Agarose gel --- p.26 / Chapter 2.2.7 --- Minipreparation of Plasmid --- p.27 / Chapter 2.2.8 --- Large-Scale Preparation of Plasmid DNA --- p.28 / Chapter 2.2.8A --- By Equilibrium Centrifugation in Cesium Chloride- Ethidium Bromide Gradient --- p.28 / Chapter 2.2.8B --- By Using Qiagen-tip 100 Cartridge --- p.29 / Chapter 2.2.9 --- Preparation of Competent Cells --- p.30 / Chapter 2.2.10 --- Transformation of Competent Cells --- p.31 / Chapter 2.2.11 --- Screening of Recombinant Plasmids --- p.32 / Chapter 2.2.11A --- Using Selective media --- p.32 / Chapter 2.2.11B --- Rapid Alkaline Lysis Method --- p.32 / Chapter 2.2.12 --- Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis (SDS-PAGE) --- p.33 / Chapter 2.2.13 --- Size Exclusion Chromatography --- p.34 / Chapter 2.2.14 --- Electroblotting of Protein on Polyvinylidene Difluoride (PVDF) membrane --- p.35 / Chapter 2.2.15 --- Isoelectric Focusing (EEF) --- p.36 / Chapter 2.2.16 --- Protein Assay --- p.37 / Chapter 2.3 --- DNA Sequencing --- p.37 / Chapter 2.3.1 --- Isolation of a template DNA --- p.38 / Chapter 2.3.2 --- DNA Denaturation and Annealing Reaction --- p.38 / Chapter 2.3.3 --- Labeling and Termination Reaction --- p.38 / Chapter 2.3.4 --- DNA Sequencing Electrophoresis --- p.39 / Chapter 2.3.5 --- Autoradiography --- p.40 / Chapter CHAPTER THREE --- Purification and Characterization of Restriction Endonuclease from Escherichia coli HK31 --- p.41 / Chapter 3.1 --- Introduction --- p.41 / Chapter 3.2 --- Materials and Methods --- p.42 / Chapter 3.2.1 --- Preparation of Crude enzyme Extract --- p.42 / Chapter 3.2.2 --- Purification of R.EcoHK31I --- p.42 / Chapter 3.2.3 --- Characterization of Restriction endonuclease --- p.43 / Chapter 3.2.3.1 --- Enzyme Activity assay --- p.43 / Chapter 3.2.3.2 --- "Optimal pH, Temperature, Metal Ion and Salt concentration of R.EcoHK31I" --- p.43 / Chapter 3.2.3.3 --- Assay for the Purity of R.EcoHK31I --- p.43 / Chapter 3.2.3.4 --- Determination of Recognition Specificity --- p.44 / Chapter 3.2.3.5 --- Determination of the Cleavage Specificity --- p.44 / Chapter 3.3 --- Results and Discussion --- p.45 / Chapter 3.3.1 --- Purification ofR.EcoHK31I from Escherichia coli HK31 --- p.45 / Chapter 3.3.2 --- "Optimal pH,Temperature, Metal ions and Salt concentration of R.EcoHK31I" --- p.46 / Chapter 3.3.3 --- Unit Definition --- p.51 / Chapter 3.3.4 --- Purity of the R.EcoHK31I --- p.51 / Chapter 3.3.5 --- Recognition Site of the R.EcoHK31I --- p.51 / Chapter 3.3.6 --- Sensitivity of the R.EcoHK31I to dcm Methylation --- p.52 / Chapter 3.3.7 --- Cleavage Specificity of R.EcoHK31I --- p.52 / Chapter CHAPTER FOUR --- Cloning of EcoEK31I Restriction and Modification (R-M) System from Escherichia coli HK31 --- p.57 / Chapter 4.1 --- Introduction --- p.57 / Chapter 4.2 --- Materials and Methods --- p.58 / Chapter 4.2.1 --- Extraction of genomic DNA from E. coli HK31 --- p.58 / Chapter 4.2.2 --- Extraction of Extra-Chromosomal DNA from E. coli HK31 --- p.59 / Chapter 4.2.3 --- Restriction Digestion of the Total DNA --- p.59 / Chapter 4.2.4 --- Preparation of Linearized and Dephosphorylated Vector --- p.60 / Chapter 4.2.5 --- Fill-in Reaction --- p.60 / Chapter 4.2.6 --- Ligation between Vector and Digested Chromosomal DNA --- p.61 / Chapter 4.2.7 --- Selection of Clones Harboring Methyltransferase gene --- p.61 / Chapter 4.2.8 --- Screening of the Survival Clones --- p.62 / Chapter 4.3 --- Results --- p.62 / Chapter 4.3.1 --- Construction of Genomic Libraries --- p.62 / Chapter 4.3.2 --- Selection of the Methyltransferase Gene --- p.66 / Chapter 4.3.3 --- In vitro Detection of R.EcoHK31I activity --- p.67 / Chapter 4.3.4 --- Functional Localization of EcoHK31I --- p.67 / Chapter 4.3.5 --- Subcloning of the Complete EcoHK31I R-M System --- p.72 / Chapter 4.4 --- Discussion --- p.72 / Chapter 4.4.1 --- Construction of Genomic Libraries --- p.72 / Chapter 4.4.2 --- Cloning of EcoHK31I Restriction and Modification System --- p.75 / Chapter 4.4.2.1 --- Selecting Endonuclease --- p.75 / Chapter 4.4.2.2 --- Detection of Restriction Endonuclease Activity --- p.76 / Chapter 4.4.3 --- Functional Localization of the R-M System --- p.76 / Chapter CHAPTER FIVE --- The Nucleotide Sequences of the EcoHK31I R-M System --- p.78 / Chapter 5.1 --- Introduction --- p.78 / Chapter 5.2 --- Materials and Methods --- p.79 / Chapter 5.2.1 --- Sequencing Strategies --- p.79 / Chapter 5.2.2 --- DNA Sequencing --- p.80 / Chapter 5.2.3 --- Sequence Analysis --- p.80 / Chapter 5.3 --- Results and Discussion --- p.80 / Chapter 5.3.1 --- Nucleotide Sequences and Deduced Amino Acid sequences --- p.80 / Chapter 5.3.2 --- Comparison of Amino Acid Sequences --- p.85 / Chapter CHAPTER SIX --- Purification and Characterization of EcoHK31I Methyltransferase from E. coli K802 [pEcoHK31E] --- p.91 / Chapter 6.1 --- Introduction --- p.91 / Chapter 6.2 --- Materials and Methods --- p.92 / Chapter 6.2.1 --- Preparation of Crude enzyme Extract --- p.92 / Chapter 6.2.2 --- Purification of M.EcoHK31I --- p.92 / Chapter 6.2.3 --- Characterization of EcoHK31I Methyltransferase --- p.93 / Chapter 6.2.3.1 --- Enzyme Activity assay --- p.93 / Chapter 6.2.3.2 --- Determination of Methylation specificity --- p.93 / Chapter 6.2.3.3 --- Determination of Molecular weight of M.EcoHK31I --- p.94 / Chapter 6.2.3.4 --- Determination ofM.EcoHK31I Kinetics --- p.94 / Chapter 6.3 --- Results and Discussion --- p.96 / Chapter 6.3.1 --- Purification of EcoHK31I Methyltransferase --- p.96 / Chapter 6.3.2 --- M.EcoHK31I Modification Specificity --- p.99 / Chapter 6.3.3 --- "Determination of Molecular Weight ofM,EcoHK31I" --- p.99 / Chapter 6.3.4 --- Catalytic Properties of EcoHK31I Methyltransferase --- p.103 / Chapter 6.3.5 --- A Novel m5C-MTase M.EcoHK31I --- p.103 / Chapter CHAPTER SEVEN --- Over-expression and Characterization of EcoHK31I Restriction and Modification Enzymes --- p.106 / Chapter 7.1 --- Introduction --- p.106 / Chapter 7.1.1 --- Expression Vector pTrc series --- p.107 / Chapter 7.1.2 --- Expression Vector pET series --- p.107 / Chapter 7.2 --- Materials and Methods --- p.109 / Chapter 7.2.1 --- General technique --- p.109 / Chapter 7.2.2 --- Polymerase Chain Reaction --- p.110 / Chapter 7.2.3 --- Construction of plysSM13 --- p.110 / Chapter 7.2.4 --- Construction of pTrc99A-R36 --- p.110 / Chapter 7.2.5 --- Construction of pET3a-M38 --- p.111 / Chapter 7.2.6 --- Construction of pET3a-C23 --- p.111 / Chapter 7.2.7 --- Expression of Recombinant Proteins in E. coli hosts --- p.115 / Chapter 7.2.8 --- Purification of Recombinant R.EcoHK31I --- p.115 / Chapter 7.2.9 --- Determination of Molecular Weight of Recombinant R. EcoHK31I --- p.115 / Chapter 7.2.10 --- Polyclonal Antibodies against R.EcoHK31I --- p.116 / Chapter 7.2.11 --- Western Blotting --- p.116 / Chapter 7.2.12 --- Purification of Recombinant M.EcoHK31I polypeptide α --- p.117 / Chapter 7.2.13 --- Purification of Recombinant M.EcoHK31I polypeptide β --- p.118 / Chapter 7.2.14 --- In vitro Complementation Methylation Activity --- p.118 / Chapter 7.2.15 --- Incorporation of [3H]-AdoMet to non-methylated Lambda DNA --- p.119 / Chapter 7.3 --- Results and Discussion --- p.119 / Chapter 7.3.1 --- Expression of Recombinant R. EcoHK31I --- p.119 / Chapter 7.3.2 --- Purification and Characterization of Recombinant R.EcoHK31I --- p.120 / Chapter 7.3.2.1 --- Purification of Recombinant R.EcoHK31I --- p.120 / Chapter 7.3.2.2 --- Characterization of Recombinant R.EcoHK31I --- p.122 / Chapter 7.3.2.2.1 --- Molecular Weight and Isoelectric point of the Recombinant R.EcoHK31I --- p.122 / Chapter 7.3.2.2.2 --- Antibodies to Recombinant R.EcoHK31I --- p.125 / Chapter 7.3.3 --- Expression and Purification of M.EcoHK31Ipolypeptide α --- p.127 / Chapter 7.3.4 --- Expression and Purification of M.EcoHK31I polypeptide β --- p.127 / Chapter 7.3.5 --- Characterization of M.EcoHK31I polypeptides a and β --- p.129 / Chapter 7.3.5.1 --- Molecular Weight Determination --- p.129 / Chapter 7.3.5.2 --- Isoelectric Point Determination --- p.132 / Chapter 7.3.5.3 --- In vivo and in vitro Methylation Activity --- p.132 / Chapter CHAPTER EIGHT --- Generation and Activity Assay of Q193G Mutein of M.EcoHK31I Polypeptide a --- p.138 / Chapter 8.1 --- Introduction --- p.138 / Chapter 8.2 --- Materials and Methods --- p.139 / Chapter 8.2.1 --- Construction of pET3a-M38 (Q193G) --- p.139 / Chapter 8.2.2 --- Expression and Purification of Q193G protein --- p.140 / Chapter 8.2.3 --- In vivo and in vitro methylation activity of Q193G Mutein --- p.140 / Chapter 8.3 --- Results and Discussion --- p.145 / Chapter 8.3.1 --- "Construction, Expression and Purification of Q193G Mutein" --- p.145 / Chapter 8.3.2 --- Determination of Molecular Weight and Isoelectric point of Q193G --- p.145 / Chapter 8.3.3 --- In vivo and in vitro methylation activity of Q193G Mutein --- p.145 / Chapter 8.3.4 --- Recognition Specificity of Q193G Mutein --- p.147 / Chapter CHAPTER NINE --- General Discussion --- p.151 / REFERENCES --- p.159 / APPENDIX A --- p.168

由於潔淨用水日漸短缺，科學家著力研究各種水淨化方法，其中以光催化技術作水淨化處理為可行的方法之一。光催化是以半導體光催化劑在光照射下所產生的活性物種（reactive oxidative species）進行消毒，其中的失活原理、各活性物種的作用和活性物種對細菌的攻擊方位，雖然已有廣範的研究，但當中仍有不清之處，比如說過氧化氫(H₂O₂)在光催化失活的作用便是其中之一，在光催化系統中所產生的H₂O₂濃度一般較低，因此其對細菌失活的效能仍然存有爭議。 / 本研究設計一種新的反應器去研究H₂O₂在連續供應模式中的失活動力學。在 8 mM 的H₂O₂下，10⁵的大腸桿菌(Escherichia coli)在8小時內完全失活。而在 2 mM 的H₂O₂ 下，並無出現顯著失活，由於該濃度遠遠高於一般光催化系統所產生的濃度（<50 μM），因此可以推斷，即使一般光催化系統所產生的H₂O₂是連續供應，也不會使細菌失活。然而在光照的情況下，其失活動力學大為不同，在強光照射(200 mW cm⁻²)下，H₂O₂的失活效率顯著增強，證明光照和過氧化氫之間存有協同效應。這現象亦出現於光預處理過(light pretreated)的大腸桿菌，進一步證實了光照改變細菌的生理機能，從而使其易於被H₂O₂失活。 / 其後我們使用RNA測序(RNA sequencing)去檢測的大腸桿菌的基因表達水平在光照下的變化，以便研究光照和H₂O₂之間的協同作用的機理。大多數涉及抵抗氧化的基因，包括過氧化氫酶(catalase, CAT)和超氧化物歧化酶(superoxide dismutase,SOD)的表達、DNA修復及細菌內的鐵含調控等等，其mRNA 水平沒有顯著的增加或減少，只有dps、fes和sodB有明顯的變化。此外，還有幾種調控細胞內的銅合量（cutA和cueR）和細胞膜組成（ompA、ompC、resx和gnsB）的基因在光照下產生顯著變化。 經RNA測序後，我們選定了10個目標基因，並選擇相對的大腸桿菌變異體(mutants)，對比他們和母體(E. coli BW25113)經過光預處理後被H₂O₂的失活效能。雖然這次研究並未找到相關基因，但研究結果表示，光照和H₂O₂的協同效應，應該是光照增加細胞膜的通透性和提高細菌內Fenton劑含量，使細菌內的羥基自由基(·OH)的濃度增加，因此加強對細菌DNA的損傷。 / 最後，我們亦比較了AgBr/Ag/Bi₂WO₆在不同的光源的照射下的對大腸桿菌的光催化失活效率。雖然發光二極管(light emitting diode)和熒光管都常用於室內照明，但AgBr/Ag/Bi₂WO₆的細菌失活效率在兩者的光照下表現出顯著的差異，而不同的發射波長下的細菌失活效率和AgBr/Ag/Bi₂WO₆光學吸收表現出良好的相關性。此外，相對其他顏色的發光二極管，綠色發光二極管照射下在犧牲劑研究(scavenger study)的結果大為不同，進一步表明了光照的發射波長(emissionwavelength)對光催化失活機制的影響。 / 本研究揭示了H₂O₂和光照在光催化失活中的重要性，並演示了H₂O₂和光照射之間的協同作用，也闡明了光照的屬性如何影響光催化下各活性物種的產生。本研究不僅提供了一個新的角度去探討的光照、H₂O₂和細菌的生理狀態在光催化失活中的重要性，也提供了新的方向和方法去研究光催化失活機制的。 / Due to the increasing concern for the need of clean drinking water, different methods for water purification have been developed. Photocatalysis, which makes use of semiconductor photocatalyst for the generation of reactive charged and oxidative species (ROSs) under light irradiation, is one of the most promising methods for water disinfection. The mechanisms of the photocatalytic inactivation have been extensively investigated. Different factors, including the roles of ROSs and the ROSs target site(s) of bacterial cell, were elaborated by different studies. However, there are still controversial issues on the role of H₂O₂ in photocatalytic inactivation. The effectiveness of the low concentration of H₂O₂ in the bacterial inactivation process is still under question. / This study designs a new reactor to study the kinetic of H₂O₂ inactivation in continuous supply mode. Complete inactivation of 5-log Escherichia coli within 8 h is achieved when 8 mM of H₂O₂ is applied. No significant inactivation was observed when 2 mM H₂O₂ is applied, this concentration of H₂O₂ is much higher than that detected in common photocatalytic system (< 50 μM). The results show that H₂O₂ produced by common photocatalytic system is not harmful to bacterial cell, even they are produced continuously. However, when light irradiation of 200 mW cm⁻² , using Xenon lamp as lighting source, was applied to the system, the inactivation efficiency of H₂O₂ was significantly enhanced, which demonstrate the synergistic effect between the light irradiation and H₂O₂. The enhancement of inactivation by H₂O₂ can also be observed with light pretreated E. coli K-12, further confirms that light irradiation alter the physiology of the bacterial cell which increases their sensitivity to H₂O₂. / In order to find out the mechanism(s) of the synergism between the light irradiation and H₂O₂, RNA sequencing (RNA-Seq) was used to reveal the change of gene expression level of the E. coli under light irradiation. The mRNA level of most of the genes involve in catalase (CAT) and superoxide dismutase (SOD) expression, DNA repairing and intracellular iron regulation did not have significant increase or decrease. Only dps, fes and sodB showed significantly changes. Moreover, some genes that related to regulation of intracellular copper (cutA and cueR) and membrane composition (ompA, ompC, resX and gnsB) also showed significantly changes under light irradiation. After the RNA-Seq, ten genes were chosen as the possible target genes that related to the mechanism(s). Then the inactivation of E. coli BW25113 (parental strain) and the isogenic deleted mutants by H₂O₂ with light pretreatment were conducted and compared. Although the gene(s) that directly involved in the mechanisms of the synergy between H₂O₂ and light irradiation are not identified in the study, the results show that genes that are important to bacterial defense of oxidative damages, such those responsible for CAT and SOD expression and DNA repairing, are not involved in the mechanism(s). Increase of cell permeability and intracellular Fenton’s reagent content should be the main causes for the enhancement of H₂O₂ under light irradiation. / Finally, the inactivation efficiency of E. coli K-12 using AgBr/Ag/Bi₂WO₆ under different lighting sources is compared. The results show that inactivation efficiency under different emission wavelength are highly correlated with the optical absorption of the AgBr/Ag/Bi₂WO₆. Photocatalytic inactivation under two indoor lighting sources, LED lamps and Fluorescence tubes, also showed significant difference. The result of scavenger study under green LED lamps is completely different from those under other colour of LED lamps, indicates that emission wavelength also has great influence in photocatalytic inactivation mechanisms. / This study reveals the roles of H₂O₂ and light irradiation in photocatalytic inactivation and demonstrates the synergism between the H₂O₂ and light irradiation. The influence of the properties of light irradiation, including the light intensity and major emission wavelength, on the ROSs production by photocatalyst is also reported as well. This study not only provides a new perspective to the importance of light irradiation, H₂O₂ and the physiology of bacteria in photocatalytic inactivation, but a new approach in the investigation of photocatalytic inactivation mechanisms as well. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Ng Tsz Wai. / Thesis (Ph.D.) Chinese University of Hong Kong, 2014. / Includes bibliographical references (leaves 111-131). / Abstracts also in Chinese.