Approximately half the world's human population is at risk of infection from mosquito-borne arboviruses. Currently, interactions between the mosquito antiviral response and infecting arboviruses remains poorly understood; deciphering these will be crucial to the development of novel methods to limit replication and transmission that could help control future outbreaks. Previous mammalian studies have shown that the Homo sapiens Small Ubiquitin-related Modifier (SUMO) pathway plays a fundamental role in multiple aspects of cell biology, including the regulation of host cell immunity. However, this pathway and its impact on arbovirus replication remain uncharacterised in mosquito hosts such as Aedes aegypti (Ae. aegypti; Aa). Comparison between the Ae. aegypti and H. sapiens (Hs) SUMOylation pathways demonstrated a high degree of amino acid sequence and structural similarity. The most notable predicted difference is the lack of ability of AaSUMO to form poly-SUMO chains, which have important functions in H. sapiens. Biochemical analysis of the AaSUMOylation pathway identified a conserved function, and confirmed that AaSUMO could not efficiently form poly-SUMO chains, unlike HsSUMO3 (its closest H. sapiens homologue) due to the absence of an internal SUMO conjugation motif. Catalytically inactive mutants revealed the necessity of AaPIAS (Protein inhibitor of activated STAT) to induce the formation of poly-SUMO chains. Confocal microscopy confirmed that AaSUMO protein is expressed in haemocytes, the salivary glands, ovaries, and midgut, all of which are sites of arboviral replication. Q-PCR investigations have also revealed the AaSUMOylation pathway to be ubiquitously expressed. In vitro depletion of the AaSUMOylation pathway led to significantly enhanced levels of Zika, Semliki Forest, and Bunyamwera virus replication, identifying a vital role for AaSUMOylation in the suppression of these arboviruses. Subsequent studies in H. sapiens cells have also identified a significant role for HsPIAS1 in suppressing the replication of Zika, Semliki Forest, and Bunyamwera viruses. Furthermore, depletion of HsSUMO1 significantly enhanced the replication of Bunyamwera virus, indicating that SUMOylation suppresses arbovirus in a virus-dependent manner. Collectively, these data have identified a novel role for the SUMOylation pathway in suppressing arbovirus replication in both the vertebrate and invertebrate species in which arboviruses replicate.

The paramyxovirus family can cause a broad spectrum of diseases from mild febrile illnesses to more severe diseases that may require hospitalisation and can in the most serious cases have fatal outcomes. Understanding the virus infection dynamics is fundamental to the development of novel targets for therapeutic and vaccine development. The advancement of High-throughput sequencing (HTS) has revolutionised biomedical research providing unparalleled opportunities to answer complex questions. In this study we developed a workflow using directional analysis of HTS data to gain a unique opportunity to simultaneously analyse the kinetics of virus transcription and replication for PIV5 strain W3, PIV2, MuV and PIV3. The workflow could be used for the study of all negative strand viruses. The developed workflow was used to investigate a number of characteristics of paramyxovirus transcription including quantification of the transcription gradient, RNA editing resulting in the generation of non-templated mRNAs and the production of read-through mRNAs. Interestingly, the processivity of the RNA polymerase during transcription was shown to remain consistent throughout the infection amongst all of the viruses analysed. Additionally, virus replication and the generation of antigenomes were found to occur at early times post infection. This was surprising, as the current model for virus replication requires sufficient levels of NP to be present in the cytoplasm before the virus can enter replicative mode. These results suggest a revision of this model in which the virus produces local sites of virus transcription and replication in the cytoplasm known as foci and it is the level of NP surrounding the virus genomes at these local sites that dictates the virus ability to enter a replicative mode. PIV5 strain W3 was shown to supress virus gene expression at late times post infection resulting in the establishment of a persistent infection. The developed workflow was used to analyse the infection dynamics of PIV5. There were no changes in the RNA polymerase processivity of transcription that could account for the suppression of protein synthesis. A comparative analysis of PIV5 strains W3 and CPI+ identified a mutation of a serine to a phenylalanine at position 157 of the P protein in CPI+, a phosphorylation site that when phosphorylated by polo-like kinase 1 (PLK-1) was previously shown to play a role in the inhibition of virus RNA synthesis, that abolished the virus ability to supress protein synthesis and establish a persistent infection. This indicates that phosphorylation of serine at position 157 is responsible for the inhibition of virus gene expression and the establishment of persistence.

Arthropod-borne viruses (arboviruses) cause diseases of significant consequence to human and animal health. The aspect of the lifecycle that distinguishes an arbovirus from another viral group, is the requirement for replication in an arthropod vector and vertebrate host. Culicoides midges (order: Diptera; family: Ceratopogonidae) transmit several arboviral diseases of economic importance including bluetongue virus (BTV), a double-stranded RNA virus within the genus Orbivirus (family: Reoviridae). The ability of an arbovirus, such as BTV, to replicate, disseminate and be transmitted to a susceptible host is determined by the interaction between extrinsic factors, such as the titre of ingested virus, and intrinsic factors such as the particular viral and vector genotype. This process is poorly understood. Here, data are presented to address this, describing BTV infection and replication in a model species, Culicoides sonorensis. The percentages of infected cells were objectively determined in insect tissues using automated image classification. BTV infected cells of the posterior midgut and the number of cells infected were viral strain and dose-dependent and correlated with infection rate. Virus replicated to high levels in the compound eyes, fat body and epithelial cells. The brain and other neural tissues were infected at later times tested, coinciding with the expected time of BTV transmission. Viral RNA and antigen were undetectable in the salivary glands and oocytes, but were detected at high prevalence in the mouthparts. These data show, for the first time, that Culicoides-borne arboviruses may exploit an alternative mechanism for transmission to a host than that used by mosquito-borne arboviruses. BTV may be transmitted directly from the mouthparts, without requiring the ability to replicate in the salivary glands.

Order Bunyavirales poses a significant threat to the agricultural industry, animal husbandry and human welfare. In response to viral infections, interferons (IFNs) trigger the expression of numerous interferon stimulated genes (ISGs) whose concerted action can potently inhibit the replication of bunyaviruses. We used a FACS-based method to screen the ability of ~500 unique ISGs from humans or rhesus macaques to inhibit the replication of Bunyamwera virus (BUNV), the prototype of both the Peribunyaviridae family and Bunyavirales order. Candidates possessing antibunyaviral activity were further examined using a panel of divergent bunyaviruses in family Peribunyaviridae, Hantaviridae, Nairoviridae and Phenuiviridae.

As an arbovirus, bluetongue virus (BTV) needs to replicate in its vector, biting midges of the genus Culicoides, in order to be transmitted between susceptible vertebrate hosts, which include wild and domestic ruminants. However, the latest identified serotypes of BTV, namely BTV-25, BTV 26 and BTV-27, have shown an inability to effectively amplify in BTV-susceptible Culicoides sonorensis derived KC cells, or in adult C. sonorensis midges. They also show a ‘non-conventional’ phenotype, specificity for small ruminants, where particularly in goats, longer viraemia, direct contact transmission and asymptomatic infection is observed. This, together with the failure of most earlier serological and RT PCR based assays to detect these ‘novel’, genetically distinct BTV serotypes, may help to explain why these viruses went unnoticed until recently. However, with the improvement of molecular diagnostic assays, increased awareness and surveillance, it appears that the current circulation of these non conventional viruses is more common than initially thought, and their number is expected to increase with reports of additional novel virus strains from Mongolia. The work presented in this thesis is designed to further characterise virus isolate KUW2010/02, a BTV strain belonging to the non-conventional BTV serotype 26 (BTV 26), with particular interest in the viral determinants that restrict vector competence. A reverse genetics system for BTV, established in the Arbovirus Molecular Research Group at the Pirbright Institute, was used to generate specific reassortant viruses in mammalian BSR cells, containing genome segments derived from the BTV-1 reference strain and from BTV-26. The replication of these ‘engineered’ viruses was studied in KC cells, as a model for BTV vector interactions. Four genome segments from BTV-26 were identified that are associated with a restriction of BTV replication in both Culicoides cells (in vitro), and in adult Culicoides (in vivo). These include: genome segment 1 (Seg-1) encoding the viral RNA-dependent-RNA-polymerase VP1; Seg-2, encoding outer capsid protein VP2; Seg-3, encoding the sub-core shell protein VP3; and Seg 7, encoding core surface protein VP7. Further investigations in vitro, revealed an inability of BTV-26, or the reassortant virus containing outer-capsid protein VP2 from BTV-26, to bind to the outer surface of KC cells even though they did bind efficiently to BSR cells. This indicates the absence of a suitable binding receptor for VP2 of BTV-26 on the surface of C. sonorensis cells. The polymerase (VP1) and sub-core (VP3) proteins of BTV-26 were shown to be functional, at least during the initial stages of infection, in a Culicoides cell environment at 27 °C, with detection of low levels of transcription and non structural protein synthesis in these cells. However, a block at an unknown post transcriptional stage(s), restricts the successful completion/amplification of replication and dissemination to other cells, by the BTV viruses containing these genes (Seg 1/VP1, Seg-3/VP3). The core surface protein (VP7) of BTV-26 caused only a partial restriction, detected as a delay in BTV replication in KC cells, with virus titres eventually reaching the same levels as for BTV-1. Mutagenesis studies of VP7 revealed that the lower domain of VP7 of BTV 26 was responsible for this delayed growth phenotype in KC cells. Reassortment (exchange) of segments can occur between BTV-26 and a conventional isolate (BTV-1), confirming the classification of BTV-26 [KUW2010/02] as a member of the species Bluetongue virus. This also indicates the importance of providing a better understanding of these non-conventional viruses (such as BTV-26) since reassortment with co-circulating conventional strains could occur in the field, potentially generating progeny virus strains with novel and unpredictable phenotypes (e.g. with both higher pathogenesis and efficient direct contact transmission). This work excludes adult C. sonorensis midges as a potential vector for BTV 26 strongly suggesting that it is a non vector borne virus. In general, the results obtained shed some light on BTV-Culicoides interactions, a field of research that has not been extensively studied, and demonstrate that these non-conventional viruses are powerful models for the further study of BTV replication in Culicoides systems.

African horse sickness is one of the most deadly infectious diseases of horses. The disease is caused by African horse sickness virus (AHSV), an arbovirus transmitted by culicoides midges. AHSV is classified within the genus Orbivirus, family Reovirdae. The AHSV genome is composed by ten segments of double stranded RNA (dsRNA) encoding seven structural and at least four non-structural (NS) proteins. AHSV shares structural and functional similarities with Bluetongue virus, the ‘prototype’ species of the genus Orbivirus. An alternative open reading frame (ORF), overlapping the main ORF encoding the VP6, has been identified in segment 9 in both AHSV and BTV. This additional ORF encodes the non-structural protein NS4. The BTV NS4 localises in the nucleoli of the infected cells. NS4 is an interferon antagonist and a determinant of virus virulence. In this thesis, I aimed to characterise the AHSV NS4. Unlike the BTV NS4, the AHSV NS4 are relatively variable mong different virus strains. I have divided these proteins into four different subtypes: NS4I, NS4-IIα, -IIβ, and IIγ based upon their sequence similarity and on the presence of N-terminal or C-terminal truncations. In contrast to BTV, all four of these NS4 types localise in the cytoplasm of either transfected or infected cells. In addition, in transient transfection assays all the NS4 types show the ability to hamper gene expression, with NS4-IIβ being the most efficient. In order to further understand the biological significance of the AHSV NS4 we used reverse genetics to generate viruses expressing the four types of NS4 (AHSV-NS4-I, AHSV-NS4-IIα, AHSV-NS4-IIβ, AHSV-NS4-IIγ) and the corresponding NS4 deletion mutants (AHSV-ΔNS4-I etc.). Deletion of NS4 did not affect virus replication kinetics in either KC cells or interferon incompetent cells such as the BSR cell line. Similarly, both AHSV-NS4-IIβ and the corresponding ΔNS4 mutant showed similar replication kinetics in the interferon competent E. Derm cell line and in primary horse endothelial cells. In contrast, AHSV-NS4-I, AHSV-NS4-IIα, and AHSV-NS4-IIγ replicated more efficiently than the corresponding ΔNS4 viruses in these horse cells. Interestingly, the defects in replication of the NS4 viruses were removed after treatment with an inhibitor of the JAK/STAT pathway. Indeed, we observed that primary horse cells infected with the NS4 mutants released higher levels of type I interferon (IFN) than cells infected with the corresponding NS4 expressing viruses. In addition, we found the NS4 to be a determinant of virus virulence in vivo in NIH-Swiss mice infected with the viruses described above. Collectively, the data described in this thesis suggest that the NS4 is one of the proteins used by AHSV to modulate the IFN response.

Chronic HCV infection (CHC) is a significant cause of both liver related and non-liver related morbidity and mortality worldwide. Disease progression through to cirrhosis and hepatocellular carcinoma is highly variable, and once chronicity of infection has been established, the likelihood of spontaneous clearance without antiviral treatment is extremely low. Safe and highly effective oral antiviral therapy is now available for the treatment of CHC, however price and accessibility may limit the global use of these agents. Furthermore, concerns have been raised regarding the incidence of hepatocellular carcinoma in HCV-infected individuals receiving oral antiviral regimens, and there appears to be a ‘point of no return’ beyond which cirrhotic HCV-infected individuals fail to benefit from antiviral treatment. Thus, there remain a number of unanswered questions on the natural history of HCV infection. Ageing of the immune system, or immunosenescence, appears to contribute to poorer clinical and treatment outcomes, however robust, non-invasive, clinically relevant biomarkers are lacking. MicroRNAs (miRNAs) are short, endogenous non-coding RNAs responsible for post-transcriptional control of host gene expression. Specific patterns of miRNA deregulation have been described in the serum, liver tissue and peripheral immune cells of HCV-infected subjects, and it is hypothesised that they may be suitable as both diagnostic and prognostic biomarkers. We interrogated 3 patient cohorts (providing access to local and national clinical data) to identify patient factors associated with disease progression and both spontaneous and treatment-associated clearance of CHC. We found that chronological age and elevated BMI had the strongest association with hepatic cirrhosis. Co-morbid type 2 diabetes mellitus was associated with poor clinical outcomes during antiviral therapy. Spontaneous clearance of CHC occurred rarely (0.36 per 100 person-years follow up), and was associated with female gender, earlier age at infection, low HCV viral load and co-infection with HBV. Current injecting drug use was negatively associated with spontaneous clearance. We also explored the use of miRNAs as biomarkers in these cohorts. We correlated miRNA expression with cellular markers of immunosenescence to identify novel prognostic biomarkers for disease outcomes in CHC. Our findings demonstrated that CHC was associated with a distinct miRNA signature in the serum and peripheral immune cells. Serum miR-21-5p, miR-122-5p and miR-885-5p levels correlated with the expression of previously described biomarkers of ageing, however these miRNAs performed poorly as biomarkers of cirrhosis in CHC. Elevated serum miR-21-5p expression was an independent predictor of virologic relapse following antiviral therapy, together with HCV genotype. MiR-21-5p also appeared to predict the likelihood of an adverse clinical event during treatment. We identified a further microRNA, miR-345-5p, elevated baseline expression of which correlated with negative clinical outcome during treatment, and was associated with the presence of both hepatic and extra-hepatic malignancy. We explored the regulation of miRNA expression in an in vitro model, and found that interferon-stimulated gene expression is necessary for IFN-induced miR-21-5p expression. Finally, we performed pathway analysis for target genes regulated by miRNAs deregulated during CHC, and found that pathways in cancer were highly enriched. Pathway enrichment was similar between HCV-infected cirrhotic subjects and non-cirrhotic, immunosenescent subjects, suggesting that non-cirrhotic individuals with elevated biomarkers of immunosenescence may be at an increased risk of hepatocellular carcinoma and may benefit from enhanced surveillance and prioritisation for antiviral treatment. Overall, the wealth of clinical and molecular data provided the opportunity to explore possibilities for integrating novel biomarkers into clinical decision-making for monitoring liver-related disease in HCV-infected subjects.

Influenza A viruses (IAVs) are significant pathogens of humans and animals whose main natural host is considered to be wild waterfowl. IAVs have jumped the species barrier on multiple occasions, sometimes with devastating consequences. Successful infection and onward transmission (i.e. viral emergence) requires highly specific interactions between virus and host proteins. However, how an avian virus adapts to a mammalian host to establish as a novel pathogen after initial interspecies transmission is not yet clear. It was hypothesized that adaptation of an avian virus to mammals would involve changes in virus-host interactions that would result in more efficient viral replication and counteraction of immune responses. To test this hypothesis this thesis firstly describes the characterization of an equine dermal cell line (E.Derm) for the study of infection with EIVs. A panel of H3N8 AIVs was selected to investigate how equine host barriers affect the replication kinetics of distinct viruses. Finally, the transcriptome of the equine cells was investigated after infection with two evolutionary distinct H3N8 equine influenza viruses (H3N8 EIVs), and treatment with interferon-alpha (IFN-α). H3N8 EIV is an avian-origin virus that emerged in 1960s and has been circulating in horses for over 50 years, thus providing a natural model system to study the interspecies transmission and post-transfer adaptation of an avian influenza virus to a mammalian host. To examine the cellular response to infection, equine dermal cells (E.Derm) were infected with either A/equine/Uruguay/63 or A/equine/Ohio/2003. Total RNA was extracted at 4 and 24 hours post-infection for RNA sequencing and downstream transcriptomics analysis. Mock-infected cells and interferon-treated cells were also included for comparison purposes. RNA-seq data were analysed using CuffDiff2 to identify differentially expressed (DE) genes between samples. Ingenuity Pathway Analysis was used to determine the intracellular pathways in which DE genes were involved. The results showed clear differences on the intracellular pathways affected between the viruses, which were especially evident during the eclipse phase of virus replication. Distinct intracellular pathways were identified as important for EIV adaptation to the horse, which in turn could be employed by other avian influenza viruses to establish in mammals.

Plant viruses are usually linked to plant or crop infection and disease, with many examples of viral infection causing social, environmental and economic damage in plants and crops described worldwide. Therefore, the choice of any particular management strategy which helps to avoid or reduce virus infection and disease in the field, needs to be specifically studied and evaluated according to the etiology of each pathogen, transmission and impact. Field samples were collected from carrot crops during the period 2014-2016 across different regions and countries and were tested using real-time Polymerase Chain Reaction (RT-qPCR). Results revealed high levels of viral infection and showed the elevated presence of previously undescribed viruses such as Carrot torrado virus1 (CaTV1) and Carrot closterovirus 1 (CtCV1). Higher levels of infection found in year-round cultivation fields also indicated that this type of practice, widely used in the UK, might be detrimental and act in favour of virus infection. Further studies identified Cow Parsley and Hogweed as possible virus reservoirs of the tested viruses and identified the aphid species Myzus persicae, as the vector responsible of CaTV1 transmission. Early detection of a pathogen from a field sample relies on the development and availability of a specific, sensitive and rapid diagnostic method. A RT-qPCR diagnostic assay has been developed and validated for the detection of CaTV1 according to the requirements established in the European Plant Protection Organization bulletin. Next Generation Sequencing (NGS) was used to study viral communities in carrots and a new viral sequence from the genus Carlavirus was identified in samples from the UK. By comparison with specific targeted diagnostics, this technique was shown to be a valuable approach for field surveillance and rapid detection of all the viruses present in a sample. The use of more modern and sophisticated techniques such as NGS, has led to the discovery of high numbers of new viruses with unknown impact. As with CaTV1, many viruses have been found to not cause any acute disease in the infected plants. However, in this study some asymptomatic viruses, recently identified using NGS, were found to be more abundant in herbicide resistant than in wild-type sensitive black grass plants. Although no relationship between the presence of these viruses in the plants and the development of herbicide resistance could be established, the effects of asymptomatic infections are discussed.

The recent emergence of SARS and MERS and the discovery of novel coronaviruses in animals and birds suggest that the Coronavirus family encompasses more members that have not yet been identified. This study describes the design and validation of a novel pan-Coronavirus PCR and its application in virus discovery in samples obtained from 813 European rodents and shrews encompassing seven different species. Novel alphacoronaviruses were detected in the species Rattus norvegicus, Microtus agrestis, Sorex araneus and Myodes glareolus. These new viruses, together with the recently described Lucheng Rn coronavirus found in China, form a distinct rodent/shrew-specific clade within the coronavirus phylogeny and genomic analysis suggests a very ancient origin rather than the global spread of these rodent and shrew viruses. These data greatly extend the knowledge of wildlife reservoirs of alphacoronaviruses and provide important insight into their origins. In addition, further virus discovery in two vole samples, using a next-generation sequencing metagenomics approach, revealed the presence of novel viruses belonging to the poxvirus, astrovirus, rotavirus, gammaherpesvirus and influenza virus families. These studies provide important insight into the true extent of virus diversity in wildlife animals.