Clostridium difficile 027 can be divided into several subclades which vary in their disease severity. Genomic studies of C. difficile 027 strains have established that they are rich in mobile genetic elements including prophages. Little is known about temperate bacteriophage carriage in C. difficile 027 clinical isolates. Therefore, this study was designed to induce and characterise temperate bacteriophages from C. difficile 027 subtypes’ clinical isolates as a first step to understanding their potential role in disease and diversity. Ninety-one C. difficile 027 clinical isolates were induced for prophage, and the bacteriophages present were characterised using transmission electron microscopy and pulsed-field gel electrophoresis. A correlation between phage morphology and subtype was established. Morphological and genetically distinct tailed bacteriophages belonging to Myoviridae and Siphoviridae were identified in 62 and three isolates, respectively. Dual phage carriage was observed in four isolates. There were inducible phage tail-like particles in all the isolates. The capacity of the two antibiotics norfloxacin and mitomycin C to induce prophages was compared and it was found that they induced specific prophages from C. difficile isolates. PCR assays targeting the capsid, holin and portal genes of the myoviruses were designed to examine molecular diversity of C. difficile myoviruses. Phylogenetic analysis of the genes sequences from ten ribotypes showed that all sequences found in the ribotype 027 strains were identical and distinct from other C. difficile ribotypes and other bacteria species. Sporulation which can be influenced by histidine kinase gene encoded by C. difficile phages was characterised in 41 isolates. The isolates sporulated within the first 96 h producing between 10[superscript 4] and 10[superscript 12] CFU/ml spores. The variation in sporulation characteristics did not correlate to the subtypes or prophage contents of the isolates examined. This study strongly suggests that phages and sporulation contribute to diversity, evolution and success of this pathogen.

Campylobacter jejuni constitutes the major cause of food-borne diarrheal disease in developed countries. The genome of this species comprises many surface genes having mononucleotide repeat tracts (PolyG/PolyC) which undergo reversible switching between ON and OFF phases, termed phase variation. Phase variation helps bacteria to colonize the host effectively as most phase variable genes are involved in expression of cell surface structures mediating interactions with the host. The mutation rates of phase variable genes are important determinants for genetic diversity and overall fitness of the bacterial population residing in various niches. The major aim of the project was to determine the phase variation rates for the simple sequence repeat tracts of varying lengths in cj0031 and capA of C. jejuni strain 11168. The PV rates were determined by using chromosomally-located reporter construct for cj0031 and by an immunoblotting assay for capA. The mutation rates of a G10 tract were a 1.5 fold higher than a G9 repeat tract in cj0031. Similarly, a 6-fold increase in the PV rate was recorded for G12 in capA over a G10 repeat tract. The mutational spectra for G9 and G10 were predominantly insertions and were shifted to mainly deletions for G11 and G12 repeats. Major shifts in the ON/OFF status of phase variable genes of C. jejuni strain 11168 were detected by performing a multiplex PCR on isolates following passage through chickens. Thirteen novel genotypes found in the output population indicated a high level of genetic diversity was generated by changes in repeat tract lengths of phase variable genes. Bioinformatics analysis of cj0031 revealed a homology to type IIG restriction modification systems. A Southern blot analysis demonstrated that cj0031 possessed methyltransferase activity and led to the conclusion that 5’ CCCGA 3’/5’ CCCGAA 3’ were putative recognition sequences of Cj0031 methyltransferase. An investigation of functional abilities showed that Cj0031 enhanced the capability of adhesion, invasion and biofilm formation without having any affect on the motility of C. jejuni. A 5-fold restriction activity was exerted by Cj0031 on one phage type, showing that this enzyme also possessed restriction activity although this was marginal in comparison to restriction endonucleases in E. coli. It is postulated that Cj0031 mainly controls the phasevarion of other genes in C. jejuni through methylation of target sequences located either in promoter region or intergenic regions near the promoters of such genes, rather than having active involvement in protection of hosts from phages. The high PV rates of cj0031 might be compatible with its role as phasevarion to facilitate the rapid adaptation of C. jejuni to the micro-environment of hosts.

Acanthamoeba is an opportunistic protozoan pathogen that can cause fatal granulomatous encephalitis and a sight-threatening keratitis. However, the pathogenic mechanisms associated with Acanthamoeba infections remain unclear. One of the primary requirement in Acanthamoeba encephalitis is the haematogenous spread, followed by invasion of the blood-brain barrier. Using human brain microvascular endothelial cells (HBMEC) which constitute the blood-brain barrier, here we studied Acanthamoeba interactions with HBMEC. The results revealed that Acanthamoeba binds to HBMEC and adhesion can be blocked using exogenous mannose suggesting a role for mannose-binding protein (MBP). Acanthamoeba mutants expressing reduced levels of MBP showed decreased binding, further suggesting that MBP plays a role in HBMEC interactions. Among 12 isolates belonging to 5 different genotypes, the clinical isolates belonging to T4 genotype exhibited optimum binding. Adhesion most likely leads to secondary processes such as phagocytosis and toxin secretion, which are crucial for the ability ofAcanthamoeba to produce host cell damage. Using heat-killed fluorescein isothiocyanate (FITC)-labelled Escherichia coli, phagocytosis in Acanthamoeba was studied. Genistein as well as cytochalasin D blocked E. coli uptake, suggesting that Acanthamoeba phagocytosis is dependent on protein tyrosine kinasemediated intracellular signalling pathways and involve cytoskeletal rearrangements. In support of this notion, Y27632 (RhoA inhibitor) reduced bacterial uptake. By using L Y294002, a phosphatidylinositol 3-kinase (PI3K) inhibitor, we identified PI3K as an important mediator ofAcanthamoeba phagocytosis. Next, we studied the role of proteases in Acanthamoeba traversal of the blood-brain barrier using in vitro permeability assays. Our findings revealed that Acanthamoeba conditioned medium produced approximately 80% increase in HBMEC permeability, by targeting tight junctions. Western blotting assays determined that extracellular proteases of Acanthamoeba degrade zonula-l and occ1udin, key molecules in the formation of tight junctions. Prior treatment with PMSF (a serine protease inhibitor) abolished permeability changes, indicating the role of serine proteases. Zymographic assays revealed thatAcanthamoeba produced two major proteases, one of which was inhibited with PMSF and the second with 1,1O-phenanthroline (metalloprotease inhibitor). We also studied the host defence mechanisms in case of AK. Using tears from healthy individuals and an AK patient, it was demonstrated that both subjects exhibited similar levels of Acanthamoeba-specific IgA as determined by Western blotting and enzymelinked immunosorbent assay. However, normal tears were slightly more potent in reducingAcanthamoeba binding to human corneal epithelial cells, compared with tears from AK patient. Neither normal tears nor AK tears had any protective effects on Acanthamoeba-mediated corneal epithelial cell cytotoxicity. Future studies should continue to identify mechanisms associated with Acanthamoeba pathogenesis, imperative for the rationale development of therapeutic interventions.

It is now widely accepted that most if not all open wounds will become contaminated with micro-organisms. This does not necessarily provide a barrier to wound healing, and it is only if the healing processes are disrupted by the presence and metabolic activities of these communities, that the wound will fail to heal. It was the aim of this thesis to develop in vitro microbiological models to simulate the growth of microbes within the wound environment, for testing and comparing the antimicrobial efficacy of novel or existing topical wound treatments. These models could then be applied towards the understanding of the behaviour of microbes when exposed to various biocidal or inhibitory agents, across a range of microbial physiological and environmental physicochemical conditions. An existing in vitro static diffusion model was modified to allow for 3 day repeat challenge capacity testing of topical wound treatments. This proved to be a powerful method able to differentiate the sustained antimicrobial potential of various test treatments. Association with surfaces is almost certainly the prevailing lifestyle in the wound bed, hence an in vitro perfusion biofilm model was developed. This enabled the growth of quasi-steady state P. aeruginosa (Jl = 0.13-1.16 h- l ) and S. aureus (Jl = 0.22-0.26 h- l ) biofilms, as determined by measuring the specific biofilm growth rate, facilitating topical antimicrobial therapies to be applied, and their effects continually monitored. This dynamic model produced antimicrobial kill kinetic profiles which could differentiate bactericidal from bacteriostatic effects, and could potentially predict the efficacy of a given therapy in vivo where a biofilm is present. A whole-cell bioluminescent P. aeruginosa reporter was integrated into both the static diffusion and perfusion biofilm models to monitor the effects of antimicrobial therapy in real time. There was found to be a strong correlation between bioluminescence readings and viable counting under the defined experimental conditions. This enabled the state of the systems to be continually monitored without disturbance, allowing more immediate and accurate calculations of antimicrobial kinetics. Microbial specific growth rate dramatically affects the structure and composition of microbial cells, therefore this was investigated using a chemostat, to ascertain whether it affected the sensitivity/resistance of cells to a range of bacteriostatic and bactericidal agents. The growth rate of target cells significantly affected their sensitivity/resistance to the test agents, although there were no general trends and the response was largely agent specific. Furthermore, biofilm cells grown within the flat bed perfusion model were growth rate matched to chemostat grown cells, and their antimicrobial sensitivity compared. The results largely showed that growth rate was the greatest predictor of antimicrobial sensitivity to the panel of test antimicrobial agents. The in vitro models developed, especially those that allow growth rate measurement and control, can further our understanding of the response of wound prevalent microorganisms to topical treatments, within an environment that models 'some' of the parameters that are believed to be present in at least some wounds. Therefore the antimicrobial kill kinetic data produced by these models could help direct research and development into novel antimicrobial therapies, as well as potentially helping to inform topical treatment selection in the clinical environment.

The Gram-negative bacterium, Moraxella ca tarrhalis, is an opportunistic human pathogen that colonises the nasopharynx of healthy individuals. The disease most often associated with this organism is otitis media in children, though it is increasingly implicated in lower respiratory tract infections and even some cases of meningitis, endocarditis and bacteraemia. The most commonly quoted sites of colonisation and infection are the nasopharynx, tympanic cavity and lower respiratory tract. The different microenvironments found at these infection sites suggest a range of conditions in which the bacterium is an effective pathogen. An investigation was undertaken to determine what effect changing between these environments had on the proteome of M. catarrhalis. The effect of temperature was examined by comparing protein expression between cultures grown at 26, 30, 35, 37 and 40°C and also following an up-shift in temperature from 30 to 37°C. The effect of pH was examined by shifting the growth media from pH 7.2 to either pH 6.5 or pH 8.0. The effect of oxygen was examined by growing M. catarrhalis in continuous culture with 5-6 % excess O2 and comparing this to M. catarrhalis grown with no excess O2 . Two dimensional electrophoresis and difference in-gel electrophoresis were used to separate out the proteins from whole cell Iysates and to measure the relative abundance of proteins that were affected by the test conditions. The number of proteins affected by temperature was: 51 at 26°C, 10 at 30°C, 9 at 35°C and 23 at 40°C. Up-shifting pH affected 12 proteins and down-shifting pH affected 12 separate proteins. 52 proteins were affected by restricting oxygen. In addition, the identities and positions of 60 proteins were found and used to produce the most complete annotated M. catarrhalis 20 proteomic map to date. Among those proteins found to be affected by changes in environmental conditions, isoforms of the adhesin Omp CO were found to increase when oxygen availability was reduced. Attempts were made to produce a knockout ompCD mutant but these proved unsuccessful, indicating that Omp CD may have a fundamentally important role in viability of M. catarrhalis strain NCTC 11020. The general porin M35 increased in reduced oxygen and during stationary phase, suggesting a role in modifying membrane permeability similar to that of its homolog Omp C from E. coli. An isogenic m35 knockout mutant showed reduced growth in nutrient poor media, was more tolerant of osmotic pressure, was killed more quickly by exposure to an anaerobic environment and was less able to autoagglutinate. The mutation affected little difference in ability to grow at different pH levels and there was no difference in susceptibility to gentamicin, or in cell morphology or membrane thickness. Also, there was no significant difference in adherence or invasion of human pharyngeal or alveolar epithelial cells compared to the wild type, indicating that M35 is not a virulence factor

Tigecycline is the first glycylcycline to enter clinical use and displays good in vitro activity against a broad range of Gram-positive and Gram-negative pathogens. It is often used as an agent of last resort for the treatment of infections caused by multidrug-resistant Gram-negative bacteria including some Enterobacteriaceae species and Acinetobacter baumannii. Therefore, the recent emergence of tigecycline resistance in some strains of these species is a serious public health concern. Efflux was investigated as a possible mechanism of tigecycline resistance using pre- and post-therapy pairs of clinical isolates and laboratory-selected, tigecycline-resistant mutants of A. baumannii and Enterobacter cloacae and a type strain, laboratory mutants, and a clinical isolate of Serratia marcescens. Minimum inhibitory concentrations (MICs) of tigecycline and other agents were determined by agar dilution. Pulsed-field gel electrophoresis was used to assign clones / determine isolate relatedness. Expression of efflux pump genes and genes thought to be implicated in their regulation was monitored by real-time reverse-transcriptase polymerase chain reaction and their role in tigecycline resistance was further investigated by knockout mutagenesis. There was an association between increased expression of specific resistance-nodulation-division (RND) efflux pump genes and elevated tigecycline MICs in all species studied. Insertional inactivation of RND efflux pump genes implicated the AdeABC, AcrAB and SdeXY-HasF systems of A. baumannii, E. cloacae and S. marcescens, respectively. The results of this study support the hypothesis that tigecycline resistance in clinical isolates of Gram-negative bacteria arises as a result of the up-regulated activity of intrinsic efflux systems of the RND family.

West Nile virus (\VNV) is a mosquito-borne flavivirus that has N-linked glycosylation sites on the premembrane (prM) envelope (E) and nonstructural 1 (NS1) proteins. NS1 contains three glycosylation (asparagine-X-threoine [NXT]) sites at NS1130, NS1175, and NSh07. The fIrst aim ofthe thesis was to attenuate WNV through the ablation of the NS1 glycosylation sites, individually and in all combinations, initially by replacing the asparagine (N) with alanine (A) in the glycosylation motifusing site-directed mutagenesis ofa WNV infectious clone. Two ofthe mutants were greatly attenuated, up to 50,000-fold, for lethal neuroinvasiveness in mice, had reduced viraemia at two and three days post-infection relative to the parental strain, but multiplied similarly to the parental strain in Vero cells. Although these viruses were highly attenuated, the most attenuated NS1 mutant virus reverted to virulence and reversion at the fIrst NS1 glycosylation site (NS1130) was seen in mice that succumbed to infection. The second aim was to diminish reversion. Thus, the fIrst NSI glycosylation site was mutated such that it contained two or three amino acid substitutions in the NS1130 glycosylation motif(NNT~SVT;NNT~QQA),while the second and third glycosylation sites remained a single N to A amino acid change. These viruses were highly attenuated for mouse neuroinvasiveness and neurovirulence, >1OO,OOO-foid and 3,000-fold, respectively, compared to the parental strain and did not revert to virulence. Multiplication kinetics ofthe attenuated viruses in mice showed a decrease in multiplication compared to the parental strain and a small plaque phenotype. The third aim was to examine a functional role for the NS1 protein using these mutant viruses. Electron microscopy studies ofthe most attenuated NS1 glycosylation mutant virus revealed changes in the virus-induced structures compared to the parent virusinfected Vero cells that may suggest involvement ofthe NSI protein in the fonnation ofthese structures. Cytokine expression experiments showed several down-regulated cytokines in the sera ofparent and attenuated NS1 mutant virus-infected mice compared to mock-infected serum suggesting that immune evasion by the suppression ofcytokines may contribute to the WNV virulence phenotype. Therefore, mutations in the NSI glycosylation sites were hypothesized to modify virus replication and the host response to infection leading to a mouse attenuated phenotype. Construction ofa virus with mutations ablating the glycosylation site in the E protein together with the ablation ofthe three NSI glycosylation sites completely attenuated this virus for mouse neuroinvasiveness with an ipLDso >100,000 PFU. It was hypothesized that the ablation ofthe prM glycosylation site would also attenuate WNV. Contrary to previous studies, however, the ablation ofthe prM glycosylation site did not attenuate the virus and multiplication kinetics in cell culture revealed that this mutant virus m~l~ipli.ed to higher titres at all time points when compared to the parental strain. UtIh~tlon ofthe mutant viruses generated in this thesis may have potential as effectIve components offuture vaccine candidates.

No description available.

The diagnosis of HIV infection in children born to HIV-1 infected mothers is hampered by the passive transfer of maternal IgG across the placenta and remain detectable up to 12 or even 15 months of age. We have therefore employed a simple, cheap, sensitive and specific test, the Antibody class A specific capture enzyme immuno assay (AAC-EIA) and class M specificapture enzyme immunoassay (MAC-EIA) for the detection of child's own antibodies, the IgA and IgM anti HIV-1. Contrary to the currently held dogma, I demonstrated that the IgA and IgM do cross the placental barrier. The maternally derived IgA and IgM clear and are undetectable by 3 months, but from this time, detection of IgA and IgM in infant samples is a strong indicator of HIV infection. Detection of postnatal transmission of HIV-1 through breast milk is also discussed. A highly sensitive and specific method for serotyping HIV-1 was developed and applied. The method provides a way of typing HIV-1 by a serological method. It also provides a way to measure the level of maternal antibodies in infecting and non-infecting mothers. 2) The sequencing of proviral DNA or plasma reverse transcribed RNA remains the method of choice for subtyping of HIV-1. The method was used in this study. A remarkable molecular heterogeneity was observed in a small number of patients examined. In Kimpese, at least 6 subtypes (A, C, D, F, G, H) and another possible new subtype of HIV-1 are co-circulating. This study is the first to have recruited and followed up a cohort of mothers and children from this rural part of Africa (DRC, ex Zaire). It provides a practical approach to early diagnosis of HIV-1 infection in children and the method can be transferred to laboratory services in developing countries. In addition, a simpler method of qualitative and quantitative analysis of antibodies against the principal neutralising domain of HIV-1 V3 loop and the subtyping of vertically transmitted HIV-1 isolates from Kimpese has been described.

No description available.