Feline leukaemia virus is a significant pathogen of domestic cats which causes a range of proliferative and non-proliferative haematopoietic disorders. This virus has been extensively studied in the past, however advancements in molecular techniques now allow long-standing controversial topics to be revisited and reanalysed. Although FeLV-A is the only transmittable form of the virus, FeLV-B and –C may arise in infected cats if the initial virus escapes immune clearance and establishes a chronic infection. These studies aimed to investigate previously-unanswered questions regarding FeLV pathogenesis, specifically pertaining to the ability of FeLV-A to evolve into the novel subgroups B and C. These results indicate that strains of FeLV-A possessing residues D83 and D91 in their envelope glycoprotein display increased rates of viral replication, mediated by an enhanced interaction with their cognate receptor, THTR1. Evidence is provided that these viral proteins are also able to bind efficiently to the FeLV-C receptor, FLVCR1, and that these mutations represent the first in a step-wise accumulation of mutations which eventually result in a FeLV-C viral variant emerging within the host. Subsequent studies aimed to elucidate the respective roles of the acquired immune response (neutralising antibodies) and receptor availability in driving this evolutionary process; however a definitive conclusion regarding FeLV-C selection pressures was not reached due to limitations of the model. These studies also describe the first isolation of novel FeLV-B field isolates which present without a FeLV-A co-infection. Characterisation of these strains revealed they possessed recombinant genomes, composed of exogenous LTRs and mostly endogenously-derived env genes. Further investigations into the potential functionality of endogenous FeLV elements within the domestic cat genome revealed numerous intact env genes, the proviruses of which may be restricted from exogenous transmission by their inability to form homodimeric RNA genomes with functional secondary structures. Although this suggestion requires experimental validation, this represents a novel mechanism of endogenous retroviral restriction.

Human cytomegalovirus (HCMV) strain AD169 encodes a single abundant 1.95kb immediate early (IE) mRNA and a single abundant 2.7kb early RNA. The major IE gene (0.756-0.745 map units) was shown in nuclease protection experiments to encode a spliced molecule of 1,736 nucleotides (excluding the poly(A) tail) consisting of four exons of 121, 88, 185 and 1,342 nucleotides. Three introns of 827, 114 and 170 nucleotides were located near the 5' end of the gene. The structural analysis of the major IE gene enabled the amino acid sequence of the major IE polypeptide to be deduced from the DNA sequence. The major early gene, which is contained in both copies of the HCMV long repeat, was found not to be spliced. A translation product of the 2.7kb early RNA has yet to be identified. Reporter genes were used in transient DNA transfection experiments to monitor expression from HCMV and other viral promoters. HCMV infections trans-activated expression from the transfected SV40 early, Rous Sarcoma virus, HSV-1 thymidine kinase (TK), the HCMV major IE and the HCMV major early promoters. Expression from both the HCMV IE and the HSV-1 TK promoters was stimulated much more gradually by HCMV than by HSV-1 infections. Experiments performed using u.v.-irradiated virus and inhibitors of HCMV replication indicated that the transfected IE promoter was stimulated primarily by a de novo synthesised HCMV-encoded gene product(s). When the concentration of the plasmids IEPlcatIEterm and AccHincat transfected into cells was lowered sufficiently, HCMV infection was observed to repress expression from the transfected IE promoter. A sequence in the HCMV major IE gene between -299 and + 69 apparently contains a cis-acting signal which responds to an HCMV-induced repressor. Competitive co-transfection experiments indicated that an HCMV-induced repressor of IE transcription interacts with at least three distinct regions within the IE promoter.

Rotaviruses are the single most important etiological agent of severe diarrhoea in infants and young children in both developed and developing countries. The World Health Organisation has identified the development of a rotavirus vaccine as a priority area for routine childhood immunisation to control rotavirus infections. However, the candidate vaccine strains have not been very successful. The main aim of this project was to map rotavirus virulence to its gene segments. Such studies can help in developing better vaccines for the control of rotavirus induced diarrhoea. A three step approach was undertaken (i) development of an animal model, (ii) construction and characterisation of reassortants between rotavirus strains of different virulence, (iii) mapping virulence to rotavirus gene segments. The mouse model developed revealed that the outcome of rotavirus infection was influenced by viral dose and viral strain as well as by host age and host strain. Homologous murine rotavirus strain was found to be most virulent. Among the heterologous strains studied, the OSU strain was found to be most virulent and UKtc strain the least virulent. The CD- 1 strain of mouse was found to be the most susceptible to virus infection and C57/BL the least susceptible. A very simple and rapid nucleic acid extraction method has been developed that requires only one centrifugation step and circumvents the use of any hazardous organic chemicals, which can be applied to very large numbers of samples saving time and labour. Rotavirus reassortants were constructed in a variety of ways and their genotype determined from relative mobility of their gene segments on polyacrylamide gels and restriction enzyme digestion of PCR amplified products. Twenty two reassortants (2%) were identified out of more than 1100 progeny clones examined and these reassortants belonged to 15 different genotypes. Possible reasons for obtaining this low number of reassortants are discussed. No reassortant could be identified between a murine rotavirus and other heterologous rotavirus strains. Preliminary sequence of VP7 gene of murine rotavirus strains, EDIM and EBR, was found to be different to the published rotavirus sequences including the recently published five murine rotavirus strains. The virulence mapping studies conducted in mice with some of the 22 reassortants obtained in the present study showed that gene 4 of the OSU and UKtc strains was involved in virulence. Segment 5 of OSU strain and segments 5, and 8 of UKtc strain may also be involved in virulence.

The aims of the work presented in this thesis were to use a variety of mutagenesis techniques to investigate the proteins of the HSV-1 capsid. Triplexes are heterotrimers formed by two proteins in a 1:2 stoichiometry. The single-copy protein is called VP19C, and the dimeric protein is VP23. Insertional and deletional mutagenesis was carried out on VP19C and the effects of the mutations on virus growth and capsid assembly were examined. Insertional mutagenesis showed that VP19C can be divided into three regions with respect to their ability to tolerate five amino acid insertions, with two regions of approximately 100 amino acids at the N- and C-terminal regions of the protein being more tolerant of such insertions than a ~350 amino acid central region. The N-terminal ~100 amino acids of the protein, which are particularly insensitive to insertional mutagenesis, correspond to a region that is poorly conserved among herpesviruses. Some, but not all, severely disabled mutants were compromised in their ability to bind VP23 and VP5. Analysis of deletional mutants revealed the presence of an unusual nuclear localisation signal (NLS) near the N-terminus of VP19C. This was mapped to a 33 amino acid region by fusion of specific sequences to a green fluorescent protein (GFP) marker. By replacing the endogenous NLS with that from the simian virus 40 (SV40) large T antigen, we were able to show that the first 45 amino acids of VP19C were not essential for assembly of functional capsids and infectious virus particles. However, removing the first 63 amino acids resulted in the formation of aberrant capsids and prevented virus growth, suggesting that the poorly-conserved N-terminal sequences have some as-yet-unidentified function.

Disruption of gap junctional intercellular communication and/or Connexins (the channel proteins of gap junctions) is frequently reported in malignant cell lines and tumours. Many tumor cells exhibit aberrant gap junctional intercellular communication (GJIC), which can be restored by transfection with Connexin genes. Of the Connexin family, Connexin 43 has attracted the most attention as it is widely expressed in many tissues and Connexin 43 gap junctions correlate with various physiological functions. Connexin 43 is a short-lived protein (half-life of 1-3 h in cultured cells), both lysosomes and proteasomes having been reported to be involved in its degradation. Certain human papillomaviruses (HPV) associated with the development of cancers, especially of the cervix, have been reported to downregulate GJIC in vitro. There is also evidence for reduced gap junctions in cervical dysplasia. The association between HPV and GJIC, and the mechanism and consequence of deregulated GJIC in cervical tumour progression, remains unclear. In HPV-16 associated cervical cancer, the viral oncogene E6 inactivates the tumour suppressor p53, but also has p53- independent functions in tumour progression. One of these may involve interaction with hDlg (the human homologue of Drosophila Discs Large), a tumour suppressor present in epithelial cells at sites of cell-cell contact and which regulates cell polarity and attachment. hDlg contains three PDZ protein-protein interaction domains, the second PDZ domain of which binds E6. Connexin 43 also has a PDZ binding domain in its C-terminus. Previously, it was demonstrated that Connexin 43 relocates from the membrane to the cytoplasm in a novel HPV-16 E6-containing cervical epithelial cell line (named WI2GPXY) that is fully transformed and invasive and deficient in gap-junctional intercellular communication (Aasen T et aI., 2003 Oncogene 22, 7969- 7980). The overall aim of this project was to investigate the relationship of loss of gap junctions to malignant progression by comparing the properties of W12GPXY with those of the non-malignant parental cell line, WI2G, from which W12GPXY was derived. First, microarray was used to identify global differences in RNA expression between the two cell lines, large differences were seen in expression of angiogenesis-related genes and they were confirmed by Real-Time RT-PCR for three genes, IL-8, VEGF and FGF-2. No significant differences were noted for connexin genes but there were differences in MAGUK family members including hDlg. However, protein expression studies by western blot and immunofluorescence staining showed a significant increase (2.9 fold) of HPV-16 E6 in W12GPXY cells, which co-localises with hDlg in the cytoplasm. Connexin 43 also binds hDlg. Treatment ofW12GPXY cells with Leptomycin B to trap E6 in the nucleus or siRNA knockdown of E6 abrogate the relocation and co-location of hDlg and endogenous wild type Connexin 43 and restore Connexin 43 gap junction at points of cell-cell contact. Further, when C33a cells (HPV -negative cervical tumour cells which normally retain large Connexin 43 gap junctions) are transfected with HPV -18 wild type E6, changes in localisation of Connexin 43 and hDlg are consistent with those in W12GPXY cells. However, C33a cells transfected with a mutant E6 lacking the hDlg binding site retain Connexin 43 gap junction plaques. Finally, Connexin 43 associates with hDlg through its PDZ-binding domain and this is required for its relocalisation to the cytoplasm in W12GPXY cells. The results suggest that increased cytoplasmic E6 associated with malignant progression of W12GPXY cells redistributes Connexin 43 from membrane junctions into the cytoplasm by a mechanism involving binding of hDlg to the Connexin 43 C-tenninal tail. The findings have uncovered a new role for hDlg in trafficking of Connexin 43. It also provides a novel mechanism for the loss of gap junctional intercellular communication during malignant progression of squamous epithelial cells. The specific roles played by lysosomes and proteasomes in the degradation of Connexin 43 in W12GPXY cells were also studied. The results suggest the involvement of both proteolytic pathways, although the lysosome seems to be the major compartment for Connexin 43 degradation. Association with HPV-16 E6 and hDlg together with proteasome activity seems to be required for Connexin 43 redirection from the cell membrane and transport into the lysosomal degradation pathway. Taken together, these results suggest that Connexin 43 gap junction intercellular communication was lost from the cell membrane requiring maintenance of E6 and hDlg complexes for proteasomal internalization and consequently transport into lysosomal compartment for degradation in W12GPXY cells.

The diagnosis of mixed genotype hepatitis C virus (HCV) infection is rare and information on incidence in the UK, where genotypes 1a and 3 are the most prevalent, is sparse. Considerable variations in the efficacies of direct-acting antivirals (DAAs) for the HCV genotypes have been documented and the ability of DAAs to treat mixed genotype HCV infections remains unclear, with the possibility that genotype switching may occur. In order to estimate the prevalence of mixed genotype 1a/3 infections in Scotland, a cohort of 512 samples was compiled and then screened using a genotype-specific nested PCR assay. Mixed genotype 1a/3 infections were found in 3.8% of samples tested, with a significantly higher prevalence rate of 6.7% (p<0.05) observed in individuals diagnosed with genotype 3 infections than genotype 1a (0.8%). An analysis of the samples using genotypic-specific qPCR assays found that in two-thirds of samples tested, the minor strain contributed <1% of the total viral load. The potential of deep sequencing methods for the diagnosis of mixed genotype infections was assessed using two pan-genotypic PCR assays compatible with the Illumina MiSeq platform that were developed targeting the E1-E2 and NS5B regions of the virus. The E1-E2 assay detected 75% of the mixed genotype infections, proving to be more sensitive than the NS5B assay which identified only 25% of the mixed infections. Studies of sequence data and linked patient records also identified significantly more neurological disorders in genotype 3 patients. Evidence of distinctive dinucleotide expression within the genotypes was also uncovered. Taken together these findings raise interesting questions about the evolutionary history of the virus and indicate that there is still more to understand about the different genotypes. In an era where clinical medicine is frequently more personalised, the development of diagnostic methods for HCV providing increased patient stratification is increasingly important. This project has shown that sequence-based genotyping methods can be highly discriminatory and informative, and their use should be encouraged in diagnostic laboratories. Mixed genotype infections were challenging to identify and current deep sequencing methods were not as sensitive or cost-effective as Sanger-based approaches in this study. More research is needed to evaluate the clinical prognosis of patients with mixed genotype infection and to develop clinical guidelines on their treatment.

In recent years, a number of newly discovered tick-borne viruses exhibiting a wide spectrum of diseases in humans have been ascribed to the Phlebovirus genus of the Bunyaviridae family. These viruses have a tripartite RNA genome composed of two negative-sense RNA segments (medium and large) and one ambisense segment (small), which encode four structural proteins and one non-structural protein (NSs). The NSs protein is the major virulence factor of bunyaviruses, and acts as an antagonist of a key component of the first line of defence against viral infections: the interferon (IFN) system (Bridgen et al., 2001; Weber et al., 2002). The work presented herein describes the characterization of tick-borne phlebovirus NSs proteins as IFN antagonists. The development of a reverse genetics system for the apathogenic tick-borne Uukuniemi phlebovirus (UUKV) enabled the recovery of infectious UUKV entirely from cDNA. A recombinant UUKV lacking NSs induced higher amounts of IFN in infected cells compared to wild-type UUKV, suggesting a role of NSs in modulating the IFN response. The weak IFN antagonistic activity of UUKV NSs was evident using transient transfection reporter assays in comparison to the NSs protein of either pathogenic Heartland virus (HRTV) or Severe fever with thrombocytopenia syndrome virus (SFTSV). The sensitivity of UUKV, HRTV and SFTSV to exogenous and virus-induced IFN, as well as their growth kinetics in IFN-competent cells were examined. The molecular mechanisms employed by UUKV, HRTV and SFTSV NSs proteins to evade antiviral immunity were investigated using reporter assays, immunofluorescence, and immunoprecipitation studies. Collectively, these assays showed that UUKV NSs was able to weakly inhibit IFN induction but not IFN signalling, through a novel interaction with MAVS (mitochondrial antiviral signalling protein). On the other hand, HRTV and SFTSV NSs proteins potently inhibited IFN induction through an interaction with TBK1, and type I but not type II IFN signalling via an interaction with STAT2. Finally, the development of a minigenome system for HRTV in conjunction with minigenomes developed for UUKV and SFTSV (Brennan et al., 2015) provided preliminary data to assess possible outcomes of tick-borne phlebovirus reassortment. In summary, the results described in this thesis offer insights into how tick-borne phlebovirus pathogenicity may be linked to the capacity of their NSs proteins to block the innate immune system. The data presented also illustrate the plethora of viral immune evasion strategies utilized by emerging phleboviruses, and provide an insight into the possibility of tick-borne phlebovirus reassortment.

The Caliciviridae are a group of small, non-enveloped viruses with a positive sense, single stranded RNA genome. Caliciviruses include the noroviruses, responsible for winter vomiting disease, as well as several important veterinary pathogens. Feline calicivirus (FCV) is an excellent model for studying calicivirus entry, having a known protein receptor and being readily propagated in cell culture. Here we explore calicivirus entry, using FCV. Virus entry is the critical first step of infection and is therefore an important area of study. Both alpha 2-6 linked sialic acid and feline junctional adhesion molecule A (fJAM-A) have been identified as receptors for FCV. The attachment of FCV to fJAM-A, is followed by uptake via clathrin mediated endocytosis. Little is known, however, on the viral escape mechanism leading to delivery of the viral RNA into the cytoplasm. We set out to explore the nature of FCV attachment and uncoating using structural, biochemical and biophysical analyses. By cryogenic electron microscopy we have characterized the virus-receptor interaction at high-resolution. Using electron microscopy and an RNA release assay, we have investigated virion uncoating. Finally, we have explored the importance of receptor glycosylation, and oligomerisation. Our analysis has allowed us to construct an atomic model of the major capsid protein VP1. Upon binding to fJAM-A, FCV undergoes a conformational change (rotation and tilting of the capsomeres). Flexibility in the receptor decorated virion has prevented high-resolution structure analysis of the conformational change or the virus-receptor interaction. We have, however, seen that the structural changes are limited to the capsid spikes. We hypothesised that the conformational change may be a priming step that would prepare the virus for uncoating upon internalisation. We found that upon lowering the pH below 5, receptor decorated virions disassembled, supporting this hypothesis. Disassembly of the virus-receptor complex at low pH presented a tool for estimating the quantity of receptor needed to prime the capsid for uncoating. Cryo-EM studies reveal that FCV bound fJAM-A is monomeric although the receptor was found to be dimeric in solution as previously described for the human and murine homologues. Furthermore, it is hypothesised that this is the form found at tight junctions between cells. We propose that disruption of fJAM-A homodimers may be the mechanism by which induction of viral uptake by endocytosis is triggered. Finally, we have confirmed the presence of an N-linked glycosylation on fJAM-A and show that the removal of this carbohydrate moiety does not affect viral binding in vitro.

Picornaviruses are non-enveloped, positive sense single-stranded RNA viruses which cause many diseases ranging from slight illness to fatal meningitis and encephalitis. There are no vaccines against most picornaviruses so drugs need to be developed. Human parechoviruses (HPeVs) are being increasingly recognised as important human pathogens, but are genetically diverse from other picornaviruses. We therefore need to understand the details of virus replication to improve the opportunities to develop antiviral drugs. Viruses often rearrange the ER, ERGIC and Golgi to give a new membrane structures involved in viral replication and/or assembly. To improve our understanding of HPeV replication, we first studied the secretory pathway compartments and we found in HPeV-infected cells that the Golgi was rearranged and became more concentrated near to the replication complex. ER seemed to disappear almost completely. In terms of HPeV non-structural proteins, 2C had a major effect on the compartments and also co-localised and aggregated lipid droplets. Many viruses such as hepatitis C virus (HCV) and Dengue virus recruit lipid droplets to replication complexes. We found that lipid droplets became larger in HPeV-infected cells, but do not co-localise with replication complexes. To investigate the interaction between 2C and lipid droplets we made several 2C mutants. Mutation of NTPase domain of 2C did not change the interaction with lipid droplets. Instead we found that other domains, including a novel amphipathic helix are important. The results suggest that lipid droplets play a role in HPeV replication and so we investigated the effect of drugs which target lipid droplet formation and lipid homeostasis on HPeV replication. We demonstrated that drugs which target the enzyme DGAT1, which is involved in lipid droplet formation, have a potent effect on HPeV replication. Our results suggest that blocking lipid droplet formation is could be an important strategy for the treatment of HPeV infection.

Noroviruses are members of the Caliciviridae family of positive strand RNA viruses and are one of the leading causes of non-bacterial gastroenteritis in humans. Currently no in vitro cell culture method has been reliably described, which has hindered the progress of norovirus research, leading to little understanding of the replication of these viruses. The purpose of this study was to produce a reliable and sensitive detection assay for noroviruses in oysters, by incorporating a broadly cross-reactive monoclonal antibody. Bivalve shellfish are one of the main transmission routes of noroviruses from the environment to humans, especially when consumed raw. Noroviruses concentrate in the digestive tissue of bivalve shellfish, such as oysters as the shellfish filter feed. In order to develop a detection assay, monoclonal antibodies against the current norovirus surrogate Murine norovirus (MNV, Genogroup V) and prevalent genogroup II noroviruses were produced. Five monoclonal antibodies detected MNV capsid protein and eleven detected GII.4 capsid protein by direct ELISA. Two of the GII.4 monoclonal antibodies showed cross-reactivity, CM160 to Southampton virus (GI) and both CM160 and CM162 detected Desert Shield virus (GII) in ELISA. No broadly cross reactive monoclonal antibody which detected a linear epitope in the capsid protein could be isolated. Alongside development of monoclonal antibodies, chimeric MNV containing a precharacterised human norovirus epitope was created. Monoclonal antibody CM54 had been previously isolated, which detected a linear epitope LEDVRN. This epitope was incorporated into the P2 domain of MNV. LEDVRN is a common epitope to GI capsid proteins but is not present in GII capsid proteins. This was an attempt to side-step the need for MNV/human norovirus cross-reactive monoclonal antibodies, which detected a linear epitope in the P domain of the capsid protein. No viable chimeric virus could be detected after transfection of Raw 264.7 and 293T cells. Detection assays, ELISA, DELFIA and qRT-PCR were compared in the sensitivity of detecting norovirus in oyster digestive tissue, using a Lordsdale (LV) GII.4 monoclonal antibody and GII specific qPCR primers described by Kageyama et al (118). An antigen capture ELISA and DELFIA was developed using hyper-immune rabbit antiserum raised to recombinant LV capsid protein. Both assays were specific for LV in faecal samples. Preliminary experiments indicated that the DELFIA was four fold more sensitive than the ELISA. Ten pacific oysters were analysed, all ten were negative by ELISA, seven positive in the DELFIA and all ten positive by qRT-PCR. This suggests that the majority (70%) of the oysters contained pre-existing norovirus protein or RNA; however these data provide no indication of infectivity. DELFIA demonstrates promise as a more sensitive alternative to the ELISA, while significantly cheaper than qRT-PCR.