Outer membrane vesicles (OMVs) are nanosized proteoliposomes ubiquitously released from the outer membrane of Gram negative bacteria, and are known to contribute to immune priming and disease pathogenesis. However, the current understanding of their interactions with host cells is limited by a lack of methods to study the rapid kinetics of vesicle entry and cargo delivery. This work has developed a highly sensitive method to study vesicle entry into host cells in real-time using a genetically encoded, vesicle targeted probe. Using this approach, it was found that the route of vesicular uptake, and thus entry kinetics and efficiency, are shaped by bacterial cell wall composition. The presence of O polysaccharide in lipopolysaccharide creates a bias towards non-receptor mediated endocytosis, which enhances both the rate and efficiency of entry into host cells. This work indicates that the composition of the bacterial cell wall influences the behaviour of OMVs, and is therefore implicated in secretion-system independent delivery of bacterial virulence factors during Gram negative infection.

Tuberculosis (TB) is an infectious bacterial disease mainly infecting the pulmonary system of the human body. It affects around 1.5 million people every year, most of whom live in developing countries. The incidence of TB has increased in line with the rise in incidences of Human Immunodeficiency Virus (HIV) infections and Acquired immune deficiency syndrome (AIDS). Due to the pressing concerns of TB, the World Health Organisation (WHO) came up with the Direct Observed Treatment (DOTS) programme. Unfortunately, the development of several resistant strains against first-line drugs and consequently second and third-line drugs have developed. As the current TB drug regimen is inadequate, a good screening strategy, discovery of newer drugs and identification of the mode of action would help in developing better treatment routines and determining bacterial pathways more clearly. Drug discovery follows two major routes, one leading from the drug to the target and the other from target to the drug. Both methods have been applied in this work in order to identify new drugs effective against mycobacteria. Screens performed against a drug library approved by the Food and Drug Administration (FDA) have resulted in some promising hits. Functional characterisation of a putative enoyl CoA hydratase EchA12, which was targeted by florfenicol, revealed a novel lipid chaperone functionality associated with cell wall lipid biosynthesis. Furthermore, a target based phenotypic drug screen of the GSK177 box set against Mtb-PrsA provided further evidence that this enzyme as a viable drug target (Ballell et. al., 2013).

The bacterium Vibrio cholerae is the causative agent of the acute diarrhoeal disease cholera. V. cholerae is naturally found in aquatic environments but can switch lifestyles to cause disease in humans. The lifestyle switch requires modulation of genetic systems. Much of the regulation occurs at the level of gene expression and is controlled by transcription factors. In this work, I show that the global transcription regulator, cAMP receptor protein (CRP), plays an integral role in the regulatory network that controls lifestyle switching. I have identified two sites for CRP in the intergenic region between rtxHCA and rtxBDE, a locus which encodes the multifunctional-autoprocessing repeats-in-toxin (MARTX) toxin and toxin transport system respectively. Using a combination of genetics, biochemistry and in vivo animal studies, I have determined a CRP dependent regulation of gene expression for toxin transport in response to host infection. This work shows that rtxHCA is constitutively expressed and not subject to regulation by CRP whist CRP acts as a repressor of rtxBDE transcription. Examination of further CRP targeted genes reveals similar behaviour upon host colonisation. These findings suggest that toxin export occurs in nutritionally rich environments, where the MARTX toxin can exert cytopathic and cytotoxic effects on host cells.

The family Hantaviridae, of the Bunyavirales order, contains many important human pathogens of which Puumala virus (PUUV) is the most widely distributed member in Europe. It causes nephropathia epidemica, a milder form of haemorrhagic fever with renal syndrome and mortality rates of up to 1% have been reported. They are enveloped viruses, with a tripartite single-stranded negative sense RNA genome, that replicate solely in the cell cytoplasm. Several factors have been proposed to play a role in hantavirus pathogenicity, including regulation of innate immune responses, cell signalling and enhancement of endothelial cell permeability. The work presented in this thesis describes biological and molecular characterisation of the mechanisms behind a hantavirus infection. Transcriptome analysis was a valuable tool that allowed the investigation of the broader picture of the effect of PUUV on the host cell. 549 and as many as 7,283 genes were differentially expressed at 24 and 48 hours post infection, respectively, in PUUV-infected cells, revealing extensive transcriptional change. By 48 hours normal cellular function appeared severely disrupted. Most genes involved in mitochondrial functioning were down-regulated, suggesting a reduced cellular energy level. Dysregulation of an important signalling hub such as mitochondria might have a more global impact on cellular functions, consistent with findings in this study. Intrinsic apoptosis pathway, which is mediated by mitochondria, appeared inhibited. Whereas, death receptor signalling was activated. Pathways associated with actin formation, organisation and signalling also appeared inhibited. Members of Rho family of GTPases, which are key regulators of actin dynamics, were down-regulated overall. Furthermore, integrin signalling, which mediates Rho GTPase activity, was also inhibited. Immunofluorescence studies revealed marked morphological changes in mitochondria and substantial remodelling of the actin cytoskeleton. Further analysis revealed a direct interaction between PUUV N protein and anillin, a scaffolding protein that mediates formation and organisation of actin filaments, suggesting a potential novel mechanism behind actin cytoskeleton reorganisation. Biological interferon (IFN) assays enabled the identification of two IFN antagonists encoded by PUUV, the cytoplasmic tail of the Gn glycoprotein and the non-structural protein, NSs. The Gn tail inhibited type I IFN induction at the level of TRAF3-TBK, in agreement with previous studies on other hantaviruses. Whereas, NSs was found to block IFN induction downstream of IRF3, suggesting it was able to disrupt transcription or translation. Utilising immunofluorescence and chromatin immunoprecipitation methods, it was found that PUUV NSs possessed a potential mechanism to inhibit transcription by blocking serine 2 phosphorylation at the C-terminal domain of RNA polymerase II in a similar manner to the previously described Bunyamwera virus NSs. The data presented in this thesis illustrates the broad range of mechanisms employed by PUUV to alter cell function to aid virus replication and subvert innate immune responses.

Microbial communities play important roles in organismal and ecosystem health. High throughput sequencing has revolutionized our understanding of host-associated microbial communities, but the viral component of these communities remains poorly characterized relative to microbes such as bacteria, particularly in non-human hosts. This knowledge gap has implications for global health, as viruses originating in wildlife are responsible for recent disease outbreaks in humans and domestic animals. Although studies have identified factors differentiating viral communities between species, we have little understanding of the variability of viral communities within species. Comparative studies of viral communities are therefore necessary to characterize novel taxa and to evaluate the ecological factors influencing intraspecific viral diversity and distribution. Bats are recognized as “special” reservoirs for viruses because they are associated with diverse viral communities and display deep evolutionary relationships with individual viral taxa. Common vampire bats (Desmodus rotundus) represent a particularly interesting system in which to investigate viral communities, as they are obligate blood feeders that interact ecologically with many different host species, providing opportunities for the acquisition of diverse viruses. The overall objective of this thesis was to advance our understanding of intraspecific wildlife-associated viral communities using an established field network of common vampire bat colonies across Peru. Specifically, I developed a novel method for comparative viral community studies, characterized the viral communities of vampire bats, and examined the ecological correlates of vampire bat viral diversity across Peru. Metagenomic sequencing is a promising technique for comparative studies of viral communities in wildlife, but there is a need to first develop standardized methods that can be applied to samples collected in the field. In Chapter 2 I developed a shotgun metagenomic sequencing approach to characterizing viral communities from non-invasive samples. Specifically, I optimized extraction and sequencing protocols using fecal and oropharyngeal swabs collected from common vampire bats in Peru. Two preliminary sequencing runs were performed, the results of which motivated four pilot studies in which I tested how different storage media, nucleic acid extraction procedures, and enrichment steps affect the viral community detected. Metagenomic sequencing revealed viral contamination of fetal bovine serum, a component of viral transport medium, suggesting that swabs should be stored in RNALater or another non-biological medium. Extraction and qPCR tests were performed on swabs inoculated with known concentrations of virus, which revealed that nucleic acid should be directly extracted from swabs rather than from supernatant or pelleted material. Metagenomic sequencing of paired samples was used to test enrichment by ribosomal RNA depletion and light DNAse treatment, which both reduced host and bacterial nucleic acid in samples and improved virus detection. A bioinformatic pipeline was developed specifically for processing vampire bat shotgun viral metagenomic data. Finally, the optimized protocol was applied to twelve pooled samples from seven localities in Peru, and read subsampling demonstrated that the viral communities detected were consistent at commonly attained depths of sequencing. The protocol developed in this chapter enables minimally biased comparative viral community studies in non-invasive samples collected from wildlife. Having a detailed understanding of viral diversity in key wildlife hosts is an important first step in evaluating the risk of zoonotic disease emergence, but we still lack a holistic view of viral communities in many species including vampire bats. In Chapter 3, I used the metagenomic sequencing protocol developed in Chapter 2 to thoroughly characterize viral communities in the saliva and feces of vampire bats captured across Peru. Viruses were detected from a range of natural host groups including vertebrate-associated taxa that were potentially infecting vampire bats, bacteriophages associated with gut bacteria, and plant- or insect-infecting viruses potentially acquired from the environment. There were broad differences between fecal and saliva viral communities, showing evidence of body habitat compartmentalization. Overall, results established that vampire bat viral communities differ between body habitats and suggested that, for the vertebrate-infecting families analyzed, novel viruses mostly fall within bat-specific clades, without evidence of livestock or humans acting as a major source of viral diversity in vampire bats. Interspecific differences in ecological and life history traits are known to impact viral richness in bats, but the factors structuring viral communities within bat species are less well understood. In Chapter 4, I examined the spatial, demographic and environmental correlates of intraspecific viral diversity in vampire bats. Three measures of viral diversity were calculated at the colony level: richness, a novel measure of taxonomic diversity, and community composition. Generalized linear models were then used to test the effects of broad scale and local ecological variables on saliva and fecal viral diversity. The results showed for the first time that ecological variables can influence intraspecific viral diversity. In summary, the work presented in this thesis advances our understanding of wildlife-associated viral communities in an ecologically important bat host. Future directions in comparative wildlife viral metagenomics, as discussed in Chapter 5, will include exploring the determinants of viral communities across host species, environments and time.

Microbial lab-based evolution is a technique to study evolutionary theory. It is a method which can provide insights into the ability of a microbe to adapt to a biological process such as low pH. To investigate pathways that could lead to an acid resistant phenotype in E. coli, we evolved six independent lines or populations of E. coli K-12 MG1655 by iterative growth and dilution experiments for approximately 740 generations at pH 4.5. Clones isolated from evolved populations were significantly fitter than the ancestor at pH 4.5. Five of the six evolved strains had acquired an identical mutation in rpoA, and mutations in cytR in addition to other mutations. PCR analysis of the fossil record of the evolved populations showed that the arcA mutations always arose first followed by the rpoA mutations. Investigating the genetic basis of adaptation showed that the mutations in arcA were loss of function in nature and conferred caused an intermediate increase in fitness. Transcriptional analysis showed a global change in their transcriptional signatures with significant upregulation of the arcA regulon. Our study showed that loss of function of ArcA caused an increase in the RpoS activity of the acid evolved strains leading to a general stress resistant phenotype.

Hodgkin lymphoma (HL) is a malignant lymphoma that is diagnosed mostly in young adults, and is the second most common malignancy to affect this age group. This disease is subdivided into two entities with different aetiologies: classical HL (cHL) (~95% of cases) and nodular lymphocyte-predominant HL. In Europe, ~82% of young adults with cHL are non-Epstein-Barr virus associated and epidemiological studies have suggested that a common infectious agent may play a key role in the aetiology of these cases. The molecular biology of HL is not well understood, primarily due to the low number of Hodgkin and Reed-Sternberg (HRS) cells present within these tumours. However, recently developed techniques for the selection and micromanipulation of single HRS cells from tumours, and the development of molecular cytogenetic techniques (i.e. array-comparative genomic hybridisation (CGH)) are overcoming these difficulties. To investigate a potential candidate virus, DNA samples from cHL biopsies were screened for the measles virus (MV) and polyomaviruses (PyV), using immunohistochemistry and highly sensitive PCR assays. Chromosomal imbalances in six well-established cHL-derived cell lines and a cHL case were analysed by array-CGH. To obtain sufficient DNA for array-CGH from the cHL case, single HRS cells were isolated using laser microdissection. DNA was extracted then amplified by degenerate oligonucleotide primer polymerase chain reaction. MV and PyV genomes were not detected within cHL biopsies. Recurrent chromosomal imbalances were confirmed within the cHL-derived cell lines and cHL case, in addition to several novel imbalances. This is the first time that a cHL case has been analysed by array-CGH.

Infection with hepatitis C virus (HCV) is cleared spontaneously in only 20% of cases with the majority of individuals developing a chronic infection. This discrepancy in disease outcome is incompletely understood but current understanding of the immune response to HCV suggests that rapid induction of a broadly neutralising antibody (nAb) response leads to resolution of acute infection. The majority of nAb identified target the envelope glycoproteins, particularly E2, and most appear to inhibit binding of E2 to the cellular receptor CD81. Antibodies targeting other interactions, such as those with the receptor CLDN or the fusion determinant, are underrepresented in the repertoire of anti-HCV antibodies. However, the antibody discovery process may have been biased by the nature of the assays used. Therefore new assays are needed to enable the discovery and characterisation of antibodies in an unbiased manner. To facilitate this, a novel insect cell display library was developed for mapping antibody-binding epitopes. Cells expressing specific E2 mutants provided the necessary proof-of-principle that loss of antibody binding could be detected in this system before a library expressing randomly mutated E2 was developed. Sorting experiments demonstrated that single cells could be isolated and enriched based on loss of antibody binding. Secondly, a method for characterising the immunoglobulin (Ig) genes of HCV infected patients was developed; Ig genes were isolated from small numbers of B cells and the sequences analysed. Finally, a patient cohort was studied with a view to investigating the evolution of the antibody response during early infection. The unreliable nature of the samples prevented such analysis; however a DNA fingerprinting method of testing the origin and relatedness of serum samples was developed. This will improve the reliability of future studies. Together these methods provide a work model for the assessment of samples and the isolation and characterisation of antibodies.

The requirements for prompt diagnosis of foot-and-mouth disease (FMD) during outbreaks, and the need to establish robust laboratory testing capacity within FMD-endemic countries, have motivated the development of point-of-care tests (POCTs) to support current diagnostic strategies. Despite numerous publications detailing the design of platforms and assays for this purpose, the majority have only been evaluated in laboratory settings, using protocols incompatible for use in challenging environments. To address this gap, this thesis describes the development of an end-to-end molecular toolbox for the detection and characterisation of FMD virus (FMDV) RNA in decentralised settings. A critical review and multiway comparison of seven assay formats and 11 sample preparation methods revealed that reverse transcription loop-mediated isothermal amplification (RT-LAMP) and real-time reverse transcription PCR (rRT-PCR) POCT-formats exhibited comparable analytical and diagnostic sensitivity to their laboratory-based equivalents. Additionally, reagent lyophilisation provided a solution for cold chain and storage considerations, whilst not compromising assay performance. Both assays were compatible with simple sample preparation methods, removing the requirement for nucleic acid extraction. For example, dilution of samples in nuclease-free water enabled FMDV RNA to be detected in multiple sample types (epithelial tissue suspensions, serum, oesophageal-pharyngeal fluid and lesion swabs), from as early as one day post infection. Notably, when the robust field-ready protocols were deployed into challenging low-resource laboratory and field-settings within East Africa, POCT results (rRT-PCR = 144; RT-LAMP = 145) were consistent with clinical observations and a reference rRT-PCR, with FMDV detected from acutely infected as well as convalescent cattle. Furthermore, transitioning of East Africa-specific FMDV-typing rRT-PCR assays (for serotypes O, A, Southern African Territories [SAT] 1 and SAT 2) into a multiplex POCT-format enabled rapid identification of FMDV serotype in situ, confirming active outbreaks of both O and A. This thesis also describes the development of GoPrime, a novel real-time PCR (rPCR) primer/probe validation tool. By parameterising GoPrime with experimental data, collected to investigate the effects of primer/probe-template mismatches on cycle threshold and limit of detection, it was possible to quantitatively predict the performance of rPCR assays in silico. The work of this thesis supports the deployment of molecular POCTs into non-specialised, resource-limited and challenging settings for simple, highly sensitive and rapid detection and/or characterisation of FMDV.

Malignant catarrhal fever (MCF) is caused by two closely associated gamma herpes viruses namely alcelaphine herpesvirus 1 (AlHV-1) and ovine herpesvirus 2 (OvHV-2) and characterised with lymphocyte infiltration in non-lymphoid tissues, vasculitis and epithelial damage. The mechanism by which the viruses cause the disease is not fully understood. The hypothesis of this project was that MCF is initiated by aberrant gene expression in endothelium, epithelium and infected T cells of susceptible animals, because they are not the natural hosts for the viruses and the viruses will not have evolved in them. The first goal was to examine whether rabbit epithelium and bovine endothelium can be infected in vitro and in vivo with AlHV-1 using q PCR and, if infected whether viral transcripts could be identified in these tissue cells using q PCR and in situ hybridisation (ISH). The results revealed that endothelium and epithelium can be infected and latent infection can be established in them. This suggests the likelihood of establishing a similar type of infection in vivo. Secondly, the trial to identify latency-associated transcripts using 5-azacitidine treatment on bovine turbinate fibroblast (BT) cells and rabbit large granular lymphocytes (LGLs) was only partially successful. However, pan T antigen was expressed in 5-azacitidine treated but not untreated LGLs cells. This may indicate a function of the drug either directly or through the latency state. Transcriptome analysis in the infected and treated LGLs and BT cells showed that several pathways were affected by 5-aza although a possible latency (low transcript levels) was only seen in the BTs. Transcriptome analysis revealed similar pathways to those described for MCF in the tissues in vivo, and an effect of 5-aza on these. Viral transcripts analysis showed that genes related to productive/lytic cycles were higher than latent ones on day 17 of the in vivo experiment demonstrating that the virus may replicate at this stage of the disease. The attempt to localize the viral transcripts on the rabbit infected tissues using ISH was unsuccessful due to a lack of time.