It is estimated that there are approximately 1030 ocean virioplankton (Suttle 2007; Parsons et al. 2012). A large component of the oceanic viriosphere are the cyanophages, viruses that specifically infect cyanobacteria. Recent advances in genomics has revealed such viruses encode a multitude of genes, often acquired horizontally, that act to redirect metabolism for their own gains (Mann et al. 2003; Lindell et al. 2004a; Millard et al. 2009; Sullivan et al. 2010; Hurwitz et al. 2013; Enav et al. 2014). These genes have been named auxiliary metabolic genes (AMGs). They include multiple subunits of complexes involved with photosynthetic electron transport (PET) and CO2 fixation (Mann et al. 2003; Lindell et al. 2004; Millard et al. 2009; Sullivan et al. 2010; Thompson et al. 2011; Puxty et al. submitted), leading to the hypothesis that cyanophages directly participate in photosynthesis to provide carbon and energy for their own replication. Cyanophages face a dynamically changing light environment during their rather lengthy infection cycles ~12hrs. Therefore, it was hypothesised that changes in light intensity may affect the physiology of phage infection in terms of photosynthesis, CO2 fixation and infection dynamics. During infection of the marine cyanobacterium Synechococcus sp. WH7803 with the well characterised cyanophage S-PM2 I show that decoupling of the photochemical and CO2 fixation reactions of photosynthesis occurs (Chapter 3), which presumably redirects metabolism towards energy generation and away from growth. Moreover, S-PM2 acts to modify the PET which results in improved functioning of PSII at HL. The result is that the lytic cycle is significantly shortened during infection of the Synechococcus host under HL compared with low light (LL) conditions. To understand whether this early lysis is a regulated process, whole transcriptome sequencing of S-PM2 was performed in HL and LL (Chapter 5). This revealed a general increase in expression of all genes in HL but only the cyanophage psbA gene was significantly up-regulated above this background. This AMG encodes a core complex of photosystem II (PSII) of the PET and therefore plays a vital role in supplying energy through photophosphorylation. It is concluded that light poses a metabolic constraint on cyanophage development that requires large amounts of energy for synthesis and assembly of the structural components of the virion. Cyanophages have therefore acquired and evolved coordinated expression of PSII genes to maintain this supply of energy. I further hypothesise that gene expression may pose a significant barrier in the acquisition of AMGs from their host due to incompatible gene regulation. To test this, the phage transcriptome was analysed (Chapter 4) to validate the model of temporal transcriptional regulation in cyanophage S-PM2 as previously proposed by comparison to enterobacteriophage T4. It is shown that the experimental data is largely congruent with the proposed model. This also revealed unpredicted characteristics of the transcriptome, including genome wide transcriptional read-through and antisense expression. It is suggested that this is facilitated by either inefficient transcriptional termination or pervasive transcription initiation and may be a biologically relevant process that allows for moderate expression of recently acquired genes. In addition, genome-wide antisense transcription may act to regulate the inventory or temporal expression of specific mRNAs in these regulatory limited phages. Attempts were therefore made to characterise a previously detected non-coding RNA (ncRNA) antisense to the light regulated S-PM2 psbA gene (Chapter 6). A model is proposed suggesting that the asRNA may act to tweak psbA expression under LL conditions to prevent accumulation of unnecessary PSII proteins. This mechanism has an interesting effect on the rate of splicing of a group I intron encoded by the psbA gene. This study provides an important leap forward in our understanding of the factors that regulate the infection dynamics and therefore ecology of cyanophages. In so doing it also reveals transcriptional constraints and adaptations that go some way to explaining the evolution of cyanophage genomes.

The major reservoirs for most influenza A virus subtypes are wild aquatic birds, especially ducks. However, they are typically resistant to the effects of the infection and usually do not develop clinical disease. In contrast, some influenza viruses cause severe illness or even death in susceptible hosts like chickens and turkeys. Paradoxically, infection of primary duck cells results in rapid cell death, whereas in chicken cells, death occurs less rapidly. Duck cells produce fewer infectious virions in comparison with the longer surviving chicken cells. In order to understand this variation in infectious virus production, chicken and duck embryo fibroblast cells (CEF and DEF) were infected with low pathogenic avian H2N3, and the viruses produced from the two hosts ware characterised. Infectious virus production from chicken cells was significantly greater than that observed from duck cells, from 8–48 hr after infection. Influenza matrix gene and protein expression, analysed by quantitative real time PCR and western blotting of culture supernatants, showed comparable levels between species at 8 and 24 hr post infection. These findings led to investigation of virus budding and morphology following infection of duck and chicken cells with the virus. Differences in morphology of released virions were observed. Budding viruses from duck cells were elongated, while chicken cells produced almost spherical virions. There was a similar clear difference in virus morphology in the duck and chicken culture supernatants. Spherical viruses were observed in chicken supernatants while duck cell supernatants contained pleomorphic virions. No differences between any genes of chicken– and duck–derived viruses were found, suggesting that host cell determinants might be responsible for such variations in virus morphology. DEF cells showed extensive production of filamentous or short filament virions following infection with filamentous (equine H3N8) and non–filamentous (avian H2N3) virus strain, respectively. This was observed even after actin disruption with cytochalasin D (Cyt.D). CEF cells infected with equine H3N8 virus produced extensive filamentous virus, which decreased markedly after disruption of actin with Cyt.D, whereas, following infection with H2N3, spherical virions were observed in the presence or absence of the actin inhibitor. Cells were also transfected with green fluorescent protein – microtubule-associated protein 1A/1B-light chain 3 (GFP–LC3) expression vector and then infected or mock infected with avian H2N3. Short filaments were observed from untransfected and transfected duck cells, while spherical and short filaments were observed from untransfected and transfected chicken cells, respectively. Filamentous virus formation could be enhanced as a result of autophagy which is more marked in duck cells than chicken cells. Further studies such as studying the structure of chicken and duck fibroblast cell membranes, the use of other drugs that inhibit actin in a mechanistically different way, and the role of other cellular proteins in modulating virus morphology should be considered.

Avian influenza viruses are considered to be key contributors to the emergence of human influenza pandemics. While aquatic birds and ducks are the major reservoir for influenza viruses, they are typically resistant to the effects of viral infection, in contrast to the frequently severe disease observed in chickens. In order to understand whether differences in receptors might contribute to this observation, anatomical distribution of influenza virus receptors (sialic acid SAα2,3-Gal and SA α2,6-Gal) in key organs of both species was studied using lectin histochemistry with linkage specific lectins, and receptor binding assays with swine H1N1 (classical A/sw/Iowa/15/30) and avian H2N3 (A/mallard duck/England/7277/06) influenza viruses. Widespread presence of both SAα 2,6-Gal and SAα2,3-Gal receptors were found in all major organs examined in both chickens and ducks. Interestingly, the predominant receptor type in chicken tracheal epithelium (TE) was SAα2,6-Gal whereas SAα2,3-Gal receptors were most abundant in duck TE. Paradoxically, infection of primary cell cultures (duck and chicken lung cells and embryo fibroblasts) with the swine H1N1, the low pathogenicity avian H2N3, and a highly pathogenic H5N1 (A/turkey/England/50-92/91) virus resulted in more extensive and rapid cell death in duck cells than in chicken cells. Infected duck cells displayed morphological features of apoptosis, increased DNA fragmentation and activation of caspase-3/7. Infected duck cells produced comparable levels of viral RNA but less infectious virus than infected chicken cells. Notably, such rapid cell death was not observed in duck cells infected with a contemporary Eurasian lineage H5N1 virus (A/turkey/Turkey/1/05) which has been shown to be fatal to ducks. Gene expression profiling of infected chicken and duck cells, 24hrs post-infection, with a chicken Affymetrix microarray platform revealed differential transcription of many genes between the two avian species. In particular, the array results suggested a possible role of BCoR, HSPA-9, STAT-3, AVEN, BCLAF1, IL-18, IFN-α, and TNF-α genes in mediating the contrasting species phenotypic response to influenza infections. In summary, rapid cell death in duck cells, mediated at least in part by apoptosis, results in reduced infective virus production and may well be an important protective host response of resistant ducks. By contrast, longer surviving infected chicken cells produce much higher infective virus load along with high levels of pro-inflammatory cytokines which could account for the susceptibility of chickens to influenza infections.

Turnip yellows virus (TuYV) is the most common and important virus infecting oilseed rape (Brassica napus) in the UK. It causes reductions in growth and seed yield in oilseed rape. Between 2007 and 2010, the prevalence of TuYV in oilseed rape crops in Lincolnshire, Warwickshire and Yorkshire was determined; incidences of infection ranged from 0 and 100%. The highest levels of infection were detected in Lincolnshire and the lowest in Yorkshire. Highest incidences were recorded during 2009-10 and the lowest in 2008-9. Incidences of TuYV were closely related to the flight activities Myzus persicae vector. Most fields showed slightly aggregated pattern of infection during autumn but spring sampling revealed more random patterns. Phylogenetic analysis of both nucleotide and amino acid sequences of the P0 and P3 genes of TuYY revealed three and two genetic groups of TuYV respectively, infecting oilseed rape in Lincolnshire, Warwickshire and Yorkshire. The P0 gene was more variable than the P3 gene and both were under purifying selection. TuYV populations in the three regions were highly structured with limited gene flow between them. Analysis of molecular variance (AMOVA) indicated 96- 97% of the observed variation was due to the variation between isolates within fields. Three RT-PCR assays were developed to differentiate the three genotypes. They successfully detected and discriminated isolates of the two major genotypes from oilseed rape in Lincolnshire. Twenty seven accessions of a B. napus Diversity Fixed Foundation Set (DFFS) screened for resistance against TuYV infections varied in their susceptibility to the virus. An accession Yudal had partial resistance to some but not all the isolates of the two major genetic groups tested. TuYV caused yield losses of up to 44.7% in a glasshouse experiment. A major QTL for the partial TuYV resistance was detected on chromosome C4 (N14), explaining up to 50.5% of the observed resistance.

Transforming growth factor-beta (TGF-β) drives the development of immature LC from hematopoietic progenitor cells and shapes the cells functions. Here I showed that two LC model cells, MuLC and MDLC, used exchangeably in the research, differ significantly in their phenotype and immune responses. Discrepancies between these models were specifically visible during stimulation with type-I IFN, where MuLC failed to up-regulate ISG levels. Yet both MuLC and MDLC demonstrated low susceptibility to HIV-1 infection, even in the absence of SAMHD-1. This post-entry restriction was conferred by the action of TGF-β on differentiation cells as indicated by our study. Indeed, in the absence of TGF-β supplementation, derived cells showed MDDC phenotype related to high susceptibility of the cells to HIV-1 infection during co-infection with SIV-Vpx. Additionally blocking of the TGF-β signalling, reversed the restrictive phenotype of LC. Importantly this pattern was also confirmed in skin extracted real epidermal LC versus dermal DC, suggesting that SAMHD-1-independent restriction activity operates in TGF-β derived cells. Accordingly to PCR analysis virus replication in LC is interrupted prior to integration, suggesting the role of additional restriction factors at early stages of virus infection or lack of essential viral dependency factors such as dNTPs. Interestingly maturation of MDLC with a synthetic bacterial triacylated lipopeptide or TNF-alpha significantly increased their susceptibility to HIV-1 infection, which may explain why HIV-1 acquisition is increased during co-infection with other STIs. In summary, our study strongly supports the action of SAMHD-1-independent HIV-1 restriction mechanisms in LC. A better understanding of the balance between HIV-1 restriction and propagation from LC to CD4+ T cells may help in the development of new microbicides or vaccines to curb HIV-1 infection at its earliest stages.

The Apicomplexan parasite Toxoplasma gondii is an obligate intracellular parasite. Infection by T .gondii causes the disease toxoplasmosis, which is one of the most prevalent parasitic diseases of animals and humans. It has been 100 years since the first discovery of the parasite in 1908; research on T. gondii has been carried out in many scientific disciplines consistently expanding the understanding of this parasite. In the last ten years, the developments of EST, microarray, genome sequencing and continuing efforts towards genome annotation has centralized the focus of T. gondii research on the understanding of gene expression and gene functions on the genome scale. Equipped with the technical advances in mass spectrometry and bioinformatics, proteomics has become established as an integral component in the post-genomics era by providing first-hand data on the functional products of gene expression. In this study, three complementary proteomic strategies, 1-DE, 2-DE and MudPIT, have been used to characterise the proteome of T. gondii tachyzoites. Protein identifications have been acquired for more than two thousand (2252) unique release 4 genes, representing almost one third (29%) of the predicted proteome of all life cycle stages. Functional predictions for each protein were carried out, which provided valuable insights into the composition of the expressed proteome and their potential biological roles. The T. gondii proteomic data has been integrated into the publically accessible ToxoDB, where 2477 intron-spanning peptides provided supporting evidence for correct splice site annotation of the release 4 genome annotation. The incompleteness of the release 4 genome annotation has been highlighted using peptide evidence, confirming 421 splice sites that are only predicted by alternative gene models. Analysis has also been carried out on the proteomic data in the light of other genome wide expression data. The comparison of the proteome and transcriptome of Toxoplasma and other Apicomplexa parasites has revealed important discrepancies between protein and mRNA expression where interesting candidates have been highlighted for further investigation. A preliminary DIGE study has been developed to characterize protein expression changes in T. gondii grown in the presence or absence of glucose. In conclusion, this study has demonstrated the importance of proteomic applications in understanding gene expression profiles and regulation in T. gondii and highlighted the importance and potential of proteogenomic approaches in genome annotation process. The importance of temporal and quantitative proteomics as well as the future of systems biology has been discussed.

The emergence and spread of insecticide resistance compromises the control of mosquito borne diseases that are responsible for millions of deaths every year in tropical and subtropical areas. Mathematical modelling is a valuable tool that can be used to explore different aspects of the development and management of insecticide resistance. We have used standard population genetics theory and ecological modelling techniques for developing models to evaluate the spread of resistance in the field. We started by developing a methodology to quantify the strength of selection for resistance occurring in nature. We used data from Mexico on the mosquito Aedes aegypti and a maximum likelihood methodology to estimate the selection and dominance coefficients driving the evolution of resistance in the field. We additionally explored the impact of poor data collection, data that combine information from different locations, and the consequences of selection and dominance coefficients varying over the sampling time period. This analysis highlighted factors highly relevant to field work such as the need for frequent surveillance in discrete sentinel sites. The use of insecticidal bed nets represents the primary tool for the prevention of malaria worldwide. It is of extreme importance to maintain their efficacy against mosquitoes, which has been undermined by the development of insecticide resistance. We assed the contribution of a novel design of bed nets in delaying insecticide resistance while at the same time determining the important parameters in driving resistance in an heterogeneous environment. We showed that this new bed net can indeed contribute to the delay of the spread of resistance, but surprisingly could have the reverse effect in specific circumstances. Finally we developed a model for the vector of malaria, that considers the stage-structured nature of the mosquito life cycle and, most importantly, explicitly incorporates insecticide resistance. It can be used to understand the population dynamics of mosquitoes throughout their entire lifecycle while analysing the impact of vector control interventions, alone and in combination, and the spread of insecticide resistance that those interventions induce. We showed that targeting the larval stages has the greatest effect on the adult population followed by targeting non host-seeking female adults. According to our results, low levels of resistance can induce failure of interventions, and the rate of spread of resistance is faster when insecticides target the larval stages.

Human immunodeficiency virus type 1 (HIV-1) is responsible for the acquired immunodeficiency syndrome (AIDS), a leading cause of death worldwide infecting millions of people each year. Despite intensive research in vaccine development, therapies against HIV-1 infection are not curative and the huge genetic variability of HIV-1 challenges drug development. Current animal models for HIV-1 research present important limitations, impairing the progress of in vivo approaches. Macaques require CD8+ depletion or large portions of the genome to be replaced by sequences derived from simian immunodeficiency viruses to progress to AIDS, and the maintenance cost is high. Mice are a cheaper alternative, but need to be 'humanized' and breeding is not possible and knockout experiments are difficult. The development of an HIV-1 clone able to replicate in mice is a challenging proposal. The lack of human co-factors in mice impedes function of the HIV-1 accessory proteins Tat and Rev, hampering HIV-1 replication. The Tat and Rev function can be replaced by constitutive/chimeric promoters, codon-optimized genes and the constitutive transport element (CTE), generating a novel HIV-1 clone able to replicate in mice without disrupting the amino acid sequence of the virus. By minimally manipulating the genomic 'identity' of the virus, we propose the generation of an HIV-1 clone able to replicate in mice to assist in antiviral drug development. My results have determined that murine NIH 3T3 cells are able to generate pseudotyped HIV-1 particles, but they are not infectious. Codon-optimized HIV-1 constructs are efficiently made in human HEK-293T cells in a Tat and Rev independent manner and capable of packaging a competent genome in trans. CSGW (an HIV-1 vector genome) efficiently generates infectious particles in the absence of Tat and Rev in human cells when 4 copies of the CTE are placed preceding the 3’LTR. HIV-1 replication competent genomes lacking tat expression and encoding different promoters are functional during the first cycle of replication when Tat is added in trans. Finally, we developed a replication competent HIV-1 clone lacking tat and rev genes and encoding 4CTEs that could be a future candidate for HIV research. My results shown that the development of an HIV-1 Tat-Rev independent clone could become a candidate for HIV research in a near future, but further investigations are necessary before proposing our model as an alternative yet.

Coinfection with bacteria and viruses is an understudied area of microbiology, despite its potential to modulate pathogen abundance and host survival. We investigated the effect of bacteria on virus infection and developed an in vitro system to study the first step: viral internalization. Our studies show that multiple bacterial species promote the entry of a diverse panel of viruses into lung and gut epithelial cells. Bacteria expressing the toll-like receptor (TLR)5 agonist, flagellin, are most efficient at inducing viral uptake and studies using recombinant flagellin or aflagellate bacterial strains confirm that flagellin has pro-viral activity. Flagellin promotes epithelial cells to support virus entry via TLR5-dependent activation of NF-KB. To extend these observations and study the role of flagellin in the complete viral replicative lifecycle, we studied human immunodeficiency virus (HIV)-1 replication in T cells. Flagellin augments HIV-1 entry and promoter activity and increases the production of extracellular virus. The data presented in this thesis highlight a new role for bacterial flagellin to promote diverse virus infection of epithelial barriers and enhance the spread of HIV-1. This has significant implications for understanding how exposure to multiple pathogens can alter susceptibility to infection and its associated pathogenesis.

Hepatitis C virus (HCV) is a major cause of global morbidity, causing long-term pathologies, including cirrhosis and hepatocellular carcinoma. While hepatocytes are the major site of viral replication, the liver contains multiple non-parenchymal cells that regulate the hepatic microenvironment and may affect HCV infection in vivo. Current understanding of the role of non-parenchymal cells in HCV infection is limited. Therefore, this project aimed to establish co-culture systems that allowed investigations into interactions between hepatocytes and non-parenchymal cells, and how these interactions affected HCV infection. The results showed that in co-culture, activated liver myofibroblasts (aLMFs) negatively regulate HCV entry, replication and spread of infection in a cell contact dependent manner. Soluble factors, including extracellular matrix proteins, and common antiviral pathways did not induce this effect. Instead, we found that aLMFmodulated cell-contact affected hepatocyte membrane receptor dynamics, reducing the mobility of the HCV receptor, CD81, impairing viral entry and replication. In addition, we found that aLMF surface expressed VAP-1 also significantly reduced virus infection independently of receptor modulation. These findings greatly improved our understanding of how the interactions between hepatic cells affect HCV, highlighting the importance of non-parenchymal cells in mediating infection in the liver microenvironment.