Congenital cytomegalovirus (CMV) infection is responsible for long-term morbidity in the form of sensorineural hearing loss and neurodevelopmental impairment in a high proportion of those with symptomatic infection at birth but only around 14% of those without symptoms. In contrast newborns acquiring infection posnatally, generally from maternal breast milk, do not seem to have any associated long-term sequelae. These differing clinical outcomes remain to be explained. In immunocompromised adults, quantitative CMV DNA PCR has given valuable insight into viral dynamics. Such data have informed treatment, particularly in the transplant setting, where CMV disease has been shown to be directly correlated with various measures of viral load in both blood and urine. Pre-emptive therapy has been adopted in many units as a direct result. Newer immunological techniques have also led to the realisation that a large proportion of the human host's cellular immune response is directed towards maintaining CMV in a latent state lifelong following primary acquisition of this virus. No equivalent data exist for the natural history, viral response to treatment or associated immune responses in neonates. This thesis therefore aimed to apply quantitative, virological and immunological techniques already successfully employed in adult patient groups to help define natural history and immune responses along with other measures of treatment response in this neonatal patient group through multicentre collaboration. The results presented give the first calculations of virus dynamics in response to antiviral treatment in this age group. The possibility is raised of suboptimal drug exposure in the clinical setting and data presented on both viral load and CMV-specific immunological responses have potentially important implications for treatment in these infants and for optimal sampling times which will inform future clinical trials and the continued search for biomarkers of treatment efficacy.

Picornaviruses have a single-stranded, positive-sense RNA genome. An internal ribosome entry site (IRES) situated within the 5' untranslated region of the genome mediates cap-independent translation of the picornavirus RNA. In an attempt to further understand the mechanism of IRES-mediated translation, it was decided to investigate the role of several nucleotides within the encephalomyocarditis virus IRES which are absolutely conserved among all cardiovirus and aphthovirus IRES elements and in close proximity to the binding site of the translation initiation factor, eIF4GI. Four nucleotides within the J domain were randomised to generate a pool of mutant IRES elements containing up to 256 different sequences. Analysis of this pool, using a cell selection system to isolate functional IRES elements, revealed that the four nucleotides were essential for IRES activity. Furthermore, it was shown that mutations at these nucleotides severely affected the binding of eIF4GI and eIF4A to the IRES. A clear correlation was seen between the activity of the mutant IRES element and its ability to bind eIF4GI/eIF4A. This strongly suggests that the binding of eIF4GI to the IRES is functionally relevant in vivo. It was also shown that the IRES elements from several different hepatitis C virus genotypes could be used within the cell selection system. This is particularly useful since the analysis of HCV within cells is currently restricted. Finally, the role of residue 20 within the swine vesicular disease virus 2A protease was investigated. This residue is known to affect translation of the picornavirus RNA and virus virulence. To analyse the role of residue 20 to the function of 2A, this residue was substituted for each of the 20 amino acids. This revealed that amino acids substitutions were reasonably well tolerated with the exception of proline.

Single strand (ss) RNA viruses are amongst the most prevalent viral pathogens in nature. A key event in the life cycle of many of these viruses is the packaging of their ssRNA genome into a capsid of defined size and shape. The mechanism by which genome packaging and capsid assembly proceeds is however poorly understood. Increased knowledge of this event is beneficial for novel anti-viral drug design, as well as contributing to our understanding of macromolecular assembly events. This project has explored the role(s) of the RNA genome in the capsid assembly process of the model ssRNA virus, bacteriophage MS2. In vitro capsid reassembly reactions have been carried out using recombinant coat protein and ssRNA transcripts corresponding to different regions of the MS2 genome. These reactions have been assayed by size distribution analysis using native gel shift assays and sedimentation velocity analysis. This has allowed the effects of RNA size, sequence and structure on capsid assembly to be investigated. All the genomic RNAs transcripts, independent of sequence and size, promoted capsid assembly. The efficiency in which they each promote assembly was, however, different. This was shown to be due to the mechanism by which genomic RNA is packaged. It appears that coat proteins bind to RNA causing conformational changes that reduce its volume to that of the capsid interior. This was evident from the observed RNA length dependence on capsid assembly efficiency. Estimates of the hydrodynamic radii of assembly components and the inhibitory effect that ethidium bromide, a compound which stiffens RNA structure, has on capsid formation also supported this hypothesis. The RNA structural transition was investigated using an RNA structure probing assay. The solution structures of the RNA transcripts were compared to the MS2 genome structure within the virion. Lead acetate was used to cause structure-specific cleavages within these RNAs which were then detected by reverse transcription using labelled primers. The results show that the RNA structure is partly conserved in solution and within the virion, implying that the conformational changes during encapsidation involve primarily tertiary structure rearrangement. The data suggest that the MS2 virion RNA has a defined structure within the virion. These results are consistent with cryo-electron microscopy of virions and capsids carried out by other members of the laboratory. One implication of this work is that compounds capable of inhibiting the conformational rearrangements required for virus assembly could serve as potent anti-viral therapeutics. The work presented in this thesis has contributed to our understanding of how ssRNA is packaged into ssRNA virus capsids and, in particular, the roles it plays in capsid assembly.

Bacteria and their viruses (bacteriophage, phage) are the most abundant and diverse taxonomic groups, but ecological and evolutionary research on bacteria-phage interactions has largely focused on studies of simplified communities using a few model organisms. The goal of the thesis is to understand how bacteria and phage interact within natural environments, and how these interactions impact the patterns of phage infectivity and bacterial resistance. Here I investigate the effects of natural environments on the coevolutionary patterns of bacteria (Pseudomonas fluorescens) and phage (SBW25Φ2). In chapter 3 I investigate the effects of nineteen different communities on the coevolutionary interactions of SBW25 and phage, and the degree to which the infectivity of phage to its host, SBW25, changes depending on their local microbial community. Chapter 4 aimed to understand the effects of varying diversities of communities on coevolutionary interactions. In Chapter 5, I looked at how coevolutionary interactions were affected by different communities in different abiotic conditions (pH, temperature and nutrient concentration) and the effect communities had on the ability of SBW25 to adapt to the abiotic conditions. Understanding how biological and physical factors affect coevolutionary interactions in natural environments allows predictions of how phage and bacteria coevolve in natural and unnatural settings.

Due to the severe limitations on the study of gammaherpesviruses presented by their species specificity and restricted growth <i>in vitro</i>, infection of laboratory mice with the naturally occurring virus murine gammaherpesvirus 68 (MHV-68) has become an excellent model system for the study of gammaherpesvirus pathogenesis. This project involved the characterisation of murine herpesvirus 76 (MHV-76), another of the murine herpesviruses isolated at the same time as MHV-68. Molecular analysis revealed that MHV-76 is a deletion mutant of MHV-68, that lacks four genes unique to MHV-68 (M1, M2, M3 and M4) and eight viral tRNA-like genes that are all present at the left end of the MHV-68 genome. Biological characterisation of MHV-76, including the generation and study of rescue viruses which restored the deleted sequence into MHV-76, showed that loss of these genes leads to enhanced clearance of virus from the lungs due to a vigorous inflammatory response. There was also a significant reduction in splenomegaly and the amount of latent virus detectable in the spleen. These results show that these genes at the left end of the genome play a critical role in the viral immune evasion strategy. Detailed studies on one of these unique MHV-68 genes, M2, show that it is a 30kDa protein that has localised homology with the Gab2 and semaphorin protein families. Expression of M2 in mammalian cells reveals that it co-localises to the plasma membrane, as well as being present in diffuse areas of the nucleus and cytoplasm. Generation of a recombinant virus with a disruption in the M2 ORF showed that the M2 gene plays a critical role in the establishment of latency and the regulation of the numbers of latently infected cells in the spleen. The M2 gene is not involved in the genesis of splenomegaly. These studies uncover some of the genetic elements involved in gammaherpesvirus pathogenesis.

The study presented here has extended previous work with the investigation of antigenic, molecular and biological properties of 43 isolates of LI virus collected from around the British Isles. Antigenic analysis using monoclonal antibodies identified two types of naturally occurring escape variants. The amino acid substitutions responsible for the alternative phenotypes were identified and shown to reside within the envelope (E) protein at residues 308 and 311. Investigation of the biological properties of the LI viruses <I>in vitro</I> and <I>in vivo</I> illustrated differences among them, some of which can be associated with genetic determinants. Molecular analysis of the isolates by determining the nucleotide and deduced amino acid sequence of the complete E gene of a representative fragment of the gene has enabled an extended investigation of genetic variation among LI virus isolates. There was a distinct correlation between genetic variation and geographic distribution, with clustering of isolates from particular areas implying the occurrence of microevolution within these regions. Grouping by geography is to be expected for tick-bone viruses with non-migratory hosts which are dispersed only sporadically into new geographic regions. The only exception to this is the Irish viruses which appear to represent two distinct virus populations existing in the same tick population. Phylogenetic analysis of the sequence data implies that the ancestral LI virus was initially introduced into Ireland and at a later date into Great Britain via Wales. The virus was then transported to Scotland from where it was dispersed throughout Scotland, Northern England and Norway. More recently the virus was probably reintroduced into Ireland and also transported to South-West England. The nucleotide substitution rate was estimated for the LI virus isolates included in this study and used to calculate the dates when viral lineages diverged. This analysis implies that LI virus was introduced into the British Isles less than 800 years ago and that the most significant dispersal from Scotland occurred 150-300 years ago.

This thesis uses the alphavirus Semliki Forest virus (SFV) to investigate mechanism(s) of virus induced apoptosis. Cell death in response to SFV infection is apoptotic, although perhaps not completely caspase dependent. During infection membranes of the mitochondria are disrupted, mitochondrial membrane potential is lost and caspases- 3, -8 and -9 are activated. Fas, RNase-L, ISG-12 and PKR are not required for cell death. PKR is however a strong inhibitor of early virus production; relative to wt mouse embryo fibroblasts (MEFs), MEFs with a disruption of the PKR gene produce SFV more rapidly and die more rapidly. Cellular detection of dsRNA can trigger signal transduction cascades resulting in activation of interferons, pro-inflammatory cytokines and pro-apoptotic pathways. Inhibition of these dsRNA induced cascades delays death of SFV infected cells demonstrating that detection of virus RNA replication is one trigger of apoptotic cell death. Synthesis of virus glycoproteins is not required for, but contributes to, cell death. SFV structural proteins build up in the endoplasmic reticulum triggering the unfolded protein response and activating the pro-apoptotic proteins caspase-12 and CHOP. In summary, SFV infection of continuously cultured mammalian cells in vitro leads to caspase activation and apoptotic cell death. The apoptosis is triggered by the detection of dsRNA as well as the virus structural proteins building up in the ER; other factors may also play a role.

Two different types of recombinant viruses each carrying one of two foreign genes, enhanced green fluorescent protein (eGFP) or Cre recombinase, were constructed based on the SFV4 backbone. In the first type of construct the transgene was inserted in the non-structural ORF, between the coding sequences for nsP3 and nsP4, flanked by processing sites recognised by the nsP2 proteinase. In the second type of constructs the 2A sequence from foot-and-mouth disease virus was added to the C-terminus of the foreign gene and this was placed between the capsid and the p62 protein of SFV4 (structural ORF). All recombinant viruses constructed were viable and able to replicate <i>in vitro</i>. eGFP expressing viruses reached titres similar to those of wild-type virus whereas Cre expressing viruses were slightly attenuated. For viruses with the marker gene inserted in the non-structural ORF, western blotting showed that the processing pattern of the non-structural polyprotein was similar to that of SFV4 and verified the expression of both foreign genes. <i>In vivo, </i>following intracerebral inoculation, all viruses caused encephalitis. Viruses expressing the foreign gene as a cleavable component of the structural ORF induced disease slower than SFV4 or viruses carrying the transgene in the replicase ORF. eGFP fluorescence was stronger and occurred later in infection when expressed in the structural ORF than in the replicase ORF. eGFP expression from the replicase ORF marked only recently infected cells; a property useful in pathogenesis studies. eGFP expressing viruses demonstrated the same cell tropism as SFV4 with infection principally of neurons and oligodendrocytes. None of the mice infected intraperitoneally with SFV4 or the recombinant viruses succumbed to infection demonstrating poor neuroinvasiveness. These viruses are likely to be highly valuable for <i>in vivo</i> pathogenesis studies.

I have investigated the range of chromatin motion in living human cells. I have shown that Lac Operator arrays tagged with GFP-LacI are able to move up to 2-3 µm over the period of two hours and on average, 0.2µm per second. These distances are greater than previously reported and similar to motion observed in yeast. I have also determined whether the position of a locus is conserved from one cell cycle to the next by following cells through mitosis. I concluded that although some aspects of positioning were conserved, loci position was established anew each cell cycle. I have investigated the effect of different epigenetic modifications on the mobility of linker histones. I was able to show that while Su(Var)3-9, responsible for tri-methylation of Lysine 9 on histone 3, and MeCP2, a DNA methylation binding protein, have no effect on linker histone mobility, the methylation of DNA does. In humans it is well established that chromosomes are arranged non-randomly within the cell nucleus. I have mapped the radial position of mouse chromosome territories in ES cells to determine if a similar pattern of non-random positioning also exists in the mouse. My results suggest there may be a loose correlation between chromosome size and position within the mouse genome. Furthermore differentiation of mouse ES cells, induced changes in the position of some, but not all, chromosomes, suggesting that gene expression may have a role in chromosome position. To determine directly if nuclear position can regulate gene expression in the mouse I aimed to artificially tether an active gene to the edge of the mouse nucleus. Transfections with Lac Operator containing plasmids were anchored to the nuclear periphery 6-18hrs after transfection and were able to show directly that the nuclear position can have a large effect on gene expression.

The aim of the study was to determine the function of gp 150 and latterly to assess the potential of gap150 as a vaccine antigen to prime and protect inbred mice against MHV-68 infection. For functional studies of gp150 two main strategies were adopted; (i) the production of a recombinant virus in which the gene encoding gp150 is made dysfunctional resulting in a gp150 'knockout' (KO) virus and (ii) generation and use of purified gp150 in cell binding studies to determine if gp150 can bind to cells. Results indicate that gp150-His binds the heterogeneous splenic cell population as a whole i.e. not a particular subset of lymphocytes. This suggests gp150 may interact with a ubiquitous cell surface protein or perhaps a protein specific to leukocytes and could be involved in MHV-68 attachment to these cells. Gene gun nucleic acid immunisation of inbred mice with a plasmid encoding gp150 under the control of a constitutive promoter, alone or in combination with a recombinant vaccinia virus expressing gp150 (VVpg150) was undertaken followed by intranasal challenge with MHV-68. Virus specific antibody appeared earlier in the group that received gp150 DNA plus VVgp150. The groups that received gp150 DNA in conjunction with either VVgp150 or a control vaccinia virus (VVgpt) appeared to have reduced levels of latently infected cells in the spleen day 15 post infection and reduced splenomegaly (a phenomenon of MHV-68 infection) in comparison with control mice. This could indicate that vaccinia virus, in a non-specific manner, boosts the specific immune response to previously administered DNA and in this case was able to limit the level of MHV-68 reaching the spleen. However, this vaccine regime failed to significantly alter the level of infectious virus in the lung or prevent the establishment of latent virus in the spleen.