Bdellovibrio bacteriovorus HD 100 is a σ-proteobacterium that predates on Gram-negative bacteria. The lifecycle of Bdellovibrio bacteriovorus is complex and regulated in part by the cyclic nucleotide, c-di-GMP. Gene knockouts of diguanylate cyclases reveal discrete phenotypes in Bdellovibrio at different time points of predation. This thesis presents the first structure of the Bdellovibrio diguanylate cyclase, Bd0742 in an inhibitory conformation. Bd0742 mediates invasion but the mechanism of its regulation was unknown. Our structure suggests that an N-terminal tail attached to the forkhead-associated domain of Bd0742 regulates diguanylate cyclase activity via self-binding. We present structural and biochemical evidence demonstrating that the Nterminal tail regulates Bd0742 activity. An active mutant of Bd0742 was produced by introduction of a disulphide bond. We also solved the structure of the glucose-6-phosphate isomerase Bd07 41. Bd07 41 likely catalyses isomerization of glucose-6-phosphate and of fructose-6-phosphate. Finally, this thesis presents a structure of a region of the degenerate diguanylate cyclase, Bd3125. Mutant strains deficient of Bd3125 invade prey more slowly. The Bd3125 structure reveals that a previously cryptic domain is a GAF (cGMP-specific phosphodiesterase, Adenylyl cyclases and FhiA) domain, with a potential role in controlling invasion speed.

An understanding of how viruses interact with their receptors is vital as this step is a major determinant of host susceptibility and disease. Coxsackievirus A9 (CAV9), an enterovirus, harbours an integrin- recognition motif, RGD (Arg-Gly-Asp), in the capsid protein VP1 and although modes of transmission and pathogenesis are still largely unknown, this motif is believed to be primarily responsible for integrins αvβ6 and αvβ3 binding. The conservation of the RGD +1 position in CAV9 and other picornaviruses showed evidence that this is related to viral tropism and infectiousness of the virus. CAV9 has also been reported to interact with the heparan sulphate/heparin class of proteoglycans (HSPG). This thesis describes work designed to improve our understanding of the involvement of a) the RGD motif and more specifically the RGDX position in CAV9 infection, using a large panel of different RGDX variants and a number of cell lines not previously used in CAV9 research b) the significance of possible interactions between CAV9 and HSPG in infection. Several CAV9 variants were tested in a panel of 8 different cell lines. Infection in each cell line was observed to follow either an RGD- dependant or RGD- independent pathway, although the results did not fully correlate with the receptor expression found on the cell lines used. The RGDX position was found to be critical for efficient infection in cells when an RGD- dependent pathway is used. To understand which integrin is likely to be involved in entry, into one of the RGD-dependent cell lines, A549, blocking antibodies against αvβ3 and αvβ6 were used. Neither antibody gave full protection against CAV9, as has been reported previously, suggesting that other integrins might also be used. Two new HSPG-binding CAV9 mutants were discovered, showing that binding to HSPG can be achieved by several mechanisms. Binding to HSPG was found to be significant in some cells, but not others, again illustrating the complexity of interactions between CAV9 and the cell surface. The results obtained have greatly improved our understanding of how CAV9 infects cells. This will be useful in the design of antivirus drugs and also gives a framework for the modification of CAV9 or other RGD containing picornaviruses for specific targeting of cancer cells for oncolytic therapy.

Influenza A viruses (IAVs) represent a significant risk to human and animal health and thus determining the mechanisms of cross-species transmission is critical in understanding viral emergence. H3N8 equine influenza virus (EIV) is a virus of avian origin that emerged in horses in the early 1960s and is still circulating in horses despite the availability of vaccines. Therefore, H3N8 EIV provides unique opportunities to study the underpinning mechanisms of cross-species jumps and adaptation in mammalian hosts. The aim of this project was to determine the role of evolution on EIV adaptation to the horse. To this end, equine and avian cell lines were infected as well as horse tracheal explants with a panel of phylogenetically distinct EIVs and compared their infection phenotypes. Viral replication was quantified and changes in the histology and ciliary activity of infected explants were assessed to compare the phenotype of infection of each virus. Phylogenetically distinct EIVs exhibited different infection phenotypes: while early EIVs grew more efficiently in avian cells, late EIVs grew to higher titres in equine explants. This phenotype was demonstrated to be largely due to the late EIV viral polymerase. Using an in vitro measure of polymerase activity, late EIVs were observed to have more efficient activity in mammalian cells than early EIVs. The Polymerase Acidic (PA) and Nucleoprotein (NP) segments were shown to be the greatest drivers of the mammalian-adapted polymerase phenotype. Including either of these segments from a late EIV in the polymerase complex significantly increased minireplicon activity. Our results suggest that EIV adapted to the horse along its evolutionary history partly by mutations in the PA and NP genes.

Bluetongue virus (BTV) is an arbovirus and the cause of “bluetongue”, a major infectious disease of ruminants. Whilst the BTV structure and replication strategies are well elucidated, less is known on the genetic variability of BTV and the molecular determinants affecting virus-host interactions. In order to investigate the determinants of BTV virulence, in this study, we compared the phenotype and genotype of a highly virulent strain of BTV-8 isolated in the Netherlands a passaged minimally in tissue culture (BTV8L), with a strain passaged extensively in tissue culture (BTV8H). BTV8L was shown to be highly pathogenic in sheep and in a mouse model of bluetongue, while BTV8H was attenuated in both hosts. Full genome sequencing revealed differences in 16 amino acid residues between these two strains. Using reverse genetics, we rescued both viruses, in order two further dissect their biological features. Rescued viruses retained the phenotype of the parental viruses in vivo and in vitro. Reassortants between BTV8L and BTV8H showed that mutations in several segments contributed to attenuation of the high passage virus. The major determinants of BTV8 virulence in IFNAR-/- mice were shown to be located in segments 1, 2, 6 and 10. In vitro studies of selected reassortants showed that through extensive passage in tissue culture BTV8H acquired increased affinity for glycosaminoglycans. This property was conferred by mutations in segment 2 and resulted in increased yields of the virus in vitro and attenuation in vivo. Additionally, BTV8H was unable to replicate in IFN competent primary sheep endothelial cells. Our data showed that multiple segments were involved in decreased efficiency of BTV8H replication in cells in an IFN-induced antiviral state. Moreover, we examined changes in viral population diversity that occured after BTV-8 isolation in insect cells (Culicoides, KC) and after passage in mammalian cells and linked decreased diversity with BTV virulence in vivo. We found, that in general, the number of genetic variants was higher in BTV-8 before cell passaging, or after one passage in KC cells, compared to the number observed after even a single passage in BHK-21 cells. The highest total number of variants was found in virus passaged in KC cells, which suggests that insect vector might serve as an amplifier of quasispecies diversity of BTV. Together, these findings suggest that the virulence of BTV is a multifactorial phenomenon involving many aspects of virus-host interactions and it is not only affected by changes in the viral proteins selected at the consensus level, but also by the genetic variability of the population as a whole.

Paleovirology is the study of viruses over evolutionary timescales. Contemporary paleovirological analyses often rely on sequence data, derived from organism genome assemblies. These sequences are the germline inherited remnants of past viral infection, in the form of endogenous viral elements and the host immune genes that are evolving to combat viruses. Their study has found that viruses have exerted profound influences on host evolution, and highlighted the conflicts between viruses and host immunity. As genome sequencing technology cheapens, the accumulation of genome data increases, furthering the potential for paleovirological insights. However, data on ERVs, EVEs and antiviral gene evolution, are often not captured by automated annotation pipelines. As such, there is scope for investigations and tools that investigate the burgeoning bulk of genome data for virus and and antiviral gene sequence data in the search of paleovirological insight.

Murine cytomegalovirus (MCMV) in its natural host, the mouse, is an excellent model for studying the biology of cytomegalovirus infection. Mostly this model has been used to study gene homologues of human cytomegalovirus (HCMV). Of the predicted 170 MCMV open reading frames (ORFs) only 78 have significant amino acid identity with genes in HCMV. To better understand the biological mechanisms underlying the differences between the viruses, for example their species specificity and immune evasion genes, MCMV unique ORFs need to be examined. Here the role of m29 and m29.1 ORFs in the MCMV (Smith strain), which have no homology with ORFs of any other cytomegalovirus, have been examined. The m29 and m29.1 ORFs are overlapping and encoded on opposite strands of the double-stranded DNA genome. Sequence analysis over this region showed a discrepancy to the published sequence. An additional G (guanine) nucleotide was found at nucleotide position 36,198 that alters the predicted ORFs, m29 being 242 amino acids shorter and m29.1 210 amino acids longer than the predicted sequence. This was confirmed by sequencing the MCMV Birmingham K181 strain, the Birmingham Smith strain and MCMV wild type isolates- N1, K17A and G4. Transcripts from the newly identified m29 and m29.1 ORFs were confirmed by reverse transcriptase PCR (RT-PCR). They were produced at early (3h) and immediate-early (2h) times post-infection respectively as determined by cycloheximide and phosphonoacetic acid treatment but were continuously expressed up to at least 24h post-infection. 5' and 3'-RACE (rapid amplification of cDNA ends) analysis from m29.1 ORF confirmed the production of a ~2.4 kb transcript and a low abundance spliced transcript from which a 123bp intron had been removed. Mutants of ORF m29 and m29.1 have been produced in which ET recombination was used to introduce stop codon mutations within these overlapping ORFs. This was achieved by single base alterations near to the 5` end of each ORF that prevented translation but not transcription of each ORF individually. Linear dsDNAs containing the mutations were introduced into the Smith MCMV BAC replacing an antibiotic cassette that had been inserted into the gene of interest. Mutant viruses, Rc29 and Rc29.1 respectively, were recovered from these mutant BACs by in vitro passage in tissue culture cells. Revertant virus (Rv29.1) was made by a further 2 step process in which the mutant m29.1 ORF was first replaced by the antibiotic cassette and then by the wt ORF. These mutants were characterized both in tissue culture and in immunocompetent BALB/c and immunodeficient SCID mice. Both mutants produced their expected transcripts but Rc29.1 virus produced no corresponding protein as examined by western blot using an antibody produced in rabbits to bacterially expressed protein. Failure to express the m29 ORF in bacteria and failure of a synthetic peptide to generate rabbit antibodies that bound to denatured m29 protein meant that protein expression of the m29 gene in either mutant could not be determined. Mutant virus Rc29 replicated similarly to wild type virus both in tissue culture and in BALB/c mice. Mutant virus Rc29.1 replicated poorly with lower yields, a delay of about 2-3 days in reaching peak titres and an earlier decline compared to wt and revertant (Rv29.1) virus in tissue culture. Rc29.1 virus also showed delayed replication in the salivary glands of BALB/c mice compared to wt and Rv29.1 viruses and in SCID mice peak titres occurred later and mice became sick and had to be humanely killed approximately 8 days later that mice infected with wt virus. These results suggest that m29 and m29.1 ORFs are dispensable for viral replication in vitro in NIH 3T3 cells and in animal hosts. However, the m29.1 ORF is required for optimal viral growth in vitro and in vivo.

1. The growth cycle of SFV in chick embryo fibroblasts, BHK-2l, HEL and 1-929 cells was described, as were the resultant effects on RNA and protein synthesis. The release of haemagglutinin and LDH from infected ,cells was assayed, and shown to parallel the release')of infectious virus. The effect of altering the temperature of incubation on the growth of the virus, and the concommitant change in macromolecular synthesis , were investigated. There was a substantial inhibition of host cell protein synthesis from .3h after:.1nfection ward , under all the conditions investigated. 2. Five proteins could be readily detected in any of the four cell types when they had been infected with SFV for at least .3h. These five proteins were present throughout the growth cycle, whether the temperature o 0 0 of incubation was .30 C, .37 C or .39 C. Five proteins of the same size were also present in chick cells infected with Sindbis virus. Two of the five were the structural proteins, the other three were NVP 95, NVP 72 and NVP 63. 3. On the basis of its amino acid composition, its carbohydrate composition, and its kinetics of labelling, NVP 6.3 appears to be a precursor of the envelope protein. NVP 95 also contains some carbohydrate. 4. Pulse-chase experiments were carried out in infected chick and BHK- 21 cells. The information from these experiments vas supplemented with results from pulse-chase experiments in which the cells had been treated with TPCK,or NaF, or with amino acid analogues, or subjected to the reversal of a temperature jump. From these results, two further high molecular weight proteins became detectable, NVP 165 and ~1TP 127. It was possible to demonstrate that NVP 165, NVP 127 and NVP 95 and NVP 6.3 were precursors to NVP 72 and the structural proteins. Two models were proposed, based on the presence or absence of a second envelope protein. 5. The proteins present in the plasma membrane and endoplasmic reticulum were investigated The only protein contained in the plasma membrane was the envelope protein, while all the proteins normally found in infected cells were present in the endoplasmic reticulum fraction. 6. On the basis of the proteins synthesised on inhibition with cycloheximide and puromycin, and because of the timing or their formation, it was suggested that none or the proteins so far identified was the RNA polymerase. 7. A method for obtaining a partially purified preparation of the RNA polymerase was described. The RNA and protein species associated with the enzyme were analysed. The M band technique vas also utilised to obtain a further fractionation. By a combination of these methods, it was shown that the replicative intermediate is the first RNA species formed, while only NVP72 vas associated with the RNA polymerase. 8. An in vitro assay for the RNA polymerase was described. The product was replicative intermediate and replicative form. It was shown that the assay was not sensitive to antiserum preparations that neutralised the core or envelope proteins. 9. A method for pulse labelling the RNA was described, using the inhibition of macromolecular synthesis caused by glucosamine. It was demonstrated that the replicative intermediate was the first RNA species formed.

Human cytomegalovirus (HCMV) is a human pathogen that can cause severe disease in immunocompromised or immunosupressed individuals and also in newborns infected in utero. Transcription of the viral genome occurs by a process in which three classes of HCMV genes, immediate early, early and late are expressed in a regulated temporal cascade. The HCMV protein pp71, encoded by gene UL82, is located in the tegument of the HCMV virion and is delivered to cells immediately upon infection. This protein has been identified as a transactivator of viral immediate early gene expression. It also stimulates expression from a number of heterologous promoters by a mechanism that is not promoter sequence specific. Protein pp71 has multiple properties; it can increase the infectivity of transfected viral DNA, modulate the cell cycle and interact with the retinoblastoma family of proteins. Within the cell nucleus, pp71 co-localises with the cellular proteins PML and hDaxx at sub-cellular structures named nuclear domain 10 (ND10). The interaction of pp71 with hDaxx is believed to promote the degradation of hDaxx, resulting in relief of repression at the HCMV major immediate early promoter. Protein pp71 has also been reported to have the unusual property of mediating long-term expression of reporter genes cloned into a herpes simplex virus type 1 (HSV 1) vector. This study describes a comparison of pp71 with the non-human UL82 homologues from simian CMV, baboon CMV, rhesus CMV and chimpanzee CMV, named S82, B82, RH82 and Ch82, respectively. Plasmids expressing all of the UL82 homologues as enhanced yellow fluorescent protein (EYFP) or myc-tagged proteins were constructed and analysed for expression by transfection into HFFF2 cells. The EYFP-tagged UL82 homologues all directed b-gal expression in short-term assays, while only pp71 directed both short-term and long-term gene expression. Only myc tagged pp71 was observed to direct gene expression in both the short-term and long-term assays. The EYFP-tagged proteins and myc tagged pp71 and Ch82 were cloned into a mutated HSV-1 vector to produce recombinant viruses. Functional assays in human glioblastoma (U373) cells confirmed that all of the EYFP-tagged and myc tagged non-human UL82 homologues were able to direct short-term expression but only EYFP- and myc-tagged pp71 directed long-term gene expression, confirming results obtained in transfection assays. In agreement with previous reports, pp71 was shown to promote the resumption of gene expression from quiescent HSV-1 genomes. Comparison of the pp71 and Ch82 homologues indicated that pp71 is unique in its ability to do so. No reactivation was observed in cells infected with an HSV-1 recombinant that expressed EYFP-tagged Ch82. In order to establish the region of pp71 responsible for mediating long-term gene expression six plasmids encoding EYFP-tagged hybrid proteins were constructed. The C-terminus, N-terminus and mid-regions of pp71 were substituted for the equivalent Ch82 regions using homologous restriction sites in both coding sequences. All EYFP-tagged hybrids mediated short-term gene expression, while only one protein, with the mid region of pp71 inserted between the C- and N-terminal regions of the Ch82 homologue, appeared to stimulate long-term gene expression. However, levels of expression were significantly lower than that achieved by pp71. A HSV-1 recombinant expressing the hybrid protein was used to confirm results from transfection assays, suggesting that the mid-region of pp71 may be involved in mediating its long-term properties. Given the significantly lower degree of gene expression directed by the hybrid protein in short-term assays it was concluded that alterations to pp71 may result in structural changes that prevent normal function of the protein. Immunofluorescence studies revealed further differences between the non-human UL82 homologues and pp71. In confirmation of previously published findings, in the majority of HFFF2 cells infected with a HSV-1 recombinant expressing EYFP tagged pp71, this protein localised to discrete punctate ND10 foci at all times tested. In cells infected with a HSV-1 recombinant expressing S82 a pattern distinct from that of pp71 was observed. S82 exhibited a punctate/diffuse pattern of fluorescence, which became increasingly diffuse at later times post-infection. The remaining non-human UL82 homologues, despite localising to the discrete punctate foci characteristic of pp71 at early times post infection, all showed nuclear distribution patterns akin to that of S82 at later times, in cells infected with the HSV-1 recombinants expressing EYFP-tagged Ch82, B82 and Rh82. All non-human UL82 homologues, like pp71, co-localised with the endogenous cellular proteins hDaxx and PML at the times tested. Interestingly, however, at later times post-infection, the S82 protein appeared to disperse hDaxx throughout the nucleus, a feature that was not observed with the remaining UL82 homologues. Examination of the hybrid protein observed to stimulate long-term gene expression revealed that, like pp71, it localised to discrete punctate foci, and co-localised with both PML and hDaxx at all times post-infection. In contrast to other published studies, it was not possible to demonstrate pp71-mediated hDaxx degradation, by either pp71 or the non-human UL82 homologues. The work presented in this thesis confirms the previous observation that pp71 directs long-term gene expression, reactivates quiescent genomes and co localises in the nucleus with hDaxx and PML. It also characterises the non-human UL82 homologues of pp71. This study shows that, while each non human UL82 homologue shares some characteristics with pp71, subtle functional differences exist between these proteins.

Hepatitis C virus (HCV) is a major cause of chronic hepatitis worldwide. Present estimates predict that approximately 130 million people are infected with HCV with the majority of all infections progressing to chronicity, ultimately leading to fibrosis, cirrhosis and hepatocellular carcinoma. The virus, which belongs to the family Flaviviridae, has a single-stranded RNA genome of positive polarity that codes for a unique polyprotein of approximately 3000 amino acids. Core protein forms the viral nucleocapsid and is the most highly conserved of all the viral proteins. As well as playing a central role in virion assembly, core has been shown to modulate several cellular processes mainly due to its interaction with several host factors. One such cellular interacting partner of core protein is the DEAD-box RNA helicase DDX3. Although, the binding of these two proteins has been shown by yeast-two hybrid screening and co-immunoprecipitation (co-IP) assays, an authentic interaction is yet to be demonstrated in cells replicating the HCV JFH1 infectious clone. Nevertheless, IF analysis of JFH1-infected cells has shown a proportion of DDX3 is redistributed to distinct cytoplasmic sites where it colocalises with core, implying that these two proteins interact during virus replication. Six JFH1 core alanine mutants (F24A, G27A, I30A, G33A, V34A and Y35A) are known to disrupt this core-DDX3 colocalistion without altering viral genome replication following direct electroporation of viral RNA into cells. All mutants release infectious virus particles, except JFH1G33A. However, substantial reductions in RNA replication levels were observed at early time points following infection of naïve cells with the infectious mutant virus particles. Furthermore, the G33A mutation appears to allow the viral genome some degree of elasticity as during the serial passaging of G33Aelectroporated cells, a progressive increase in infectious virus particles was observed. Based on the diverse phenotypes of these mutants it is difficult to conclude as to which effects (if any) are directly related to the core-DDX3 interaction. In addition, DDX3 has become a topical host factor in HCV research following recent studies demonstrating that its knockdown from target cells causes an overwhelming reduction to HCV replication. However, these studies failed to demonstrate if the core-DDX3 interaction had any functional relevance to this effect. The aim of this study was to clearly define the role of the core-DDX3 interaction in HCV replication. Using highly specific antibodies, a co-IP assay was developed that was sensitive enough to detect an interaction between these two proteins from JFH1 RNAelectroporated cell lysates. Co-IP analysis of the 6 JFH1 core mutants revealed that only one of these (JFH1Y35A) appeared to completely abolish DDX3 binding. Detailed analysis of this mutant revealed its replication properties resembled those of the WT virus following both electroporation of the viral RNA into cells and infection of naïve cells with the virus particles, demonstrating that the core-DDX3 interaction is dispensable for virus replication in cell culture. In addition, DDX3 knockdown experiments revealed that as for JFH1WT, the replication of JFH1Y35A was equally sensitive to the depletion of endogenous DDX3 levels from target cells, indicating that the requirement of DDX3 for HCV replication is unrelated to its interaction with the viral core protein. Interestingly, sucrose gradient studies showed that DDX3 protein sedimented with the extracellular core protein from JFH1WT virus particles. Analysis of the JFH1Y35A virus particles found them to be less dense with a lower content of DDX3, alluding to the exciting possibility of DDX3 being a novel component of the virion. The results of this study indicate that the core-DDX3 interaction is not the mechanism through which HCV utilises this host factor for its replication. However, DDX3 may be incorporated into virus particles through its association with the nucleocapsid. Investigating the latter possibility is the subject of our future work. In a separate study, the replication properties of JFH1G33A and JFH1F24A were fully characterised. These two mutants were chosen for further analysis as they possessed unique properties relating to the virus proliferation in cell culture. Of the original 6 JFH1 core mutants, only JFH1G33A had a defect in secreting infectious virus particles. Upon passaging of cells electroporated with this mutant infectious particle production increased and eventually achieved peak titers similar to WT. The increase in infectivity consistently coincided with the appearance of a second mutation in close proximity to the originally mutated residue 33. Although, similar results were reported previously for this mutant by Dalrymple, (2007), the present study extended this work by confirming that these second site changes were compensatory mutations, which rescued the infectivity of JFH1G33A. In contrast to the assembly defect of JFH1G33A, JFH1F24A was efficient at secreting infectious virus following electroporation, however the infection kinetics of these particles was considerably lower than those derived from JFH1WT. Further analysis, suggested the delayed infection of JFH1F24A virus was likely due to an early event in the HCV lifecycle following entry into the cell and prior to the release of the infectious genome. Identifying the importance of glycine 33 in infectious virus assembly and phenylalanine 24 in virus spread provides the basis for further studies into the role that core plays in virus proliferation in cell culture. In its totality, this study has uncovered important details regarding the role of core, DDX3 and the core-DDX3 interaction in the HCV lifecycle.

Extensive study of the prototypical alphaherpesvirus, Herpes simplex virus type-1 (HSV-1), has revealed much about its biology and structure. The characteristic virion comprises a DNA containing capsid surrounded by consecutive layers of proteinaceous tegument and lipid envelope. Our understanding of the capsid is relatively advanced compared to that of the tegument. The tegument has previously been thought of as an amorphous accumulation of protein filling the space between the capsid and envelope. The tegument has been subdivided into the inner and the outer tegument with the proteins pUS3, pUL36 and pUL37 assigned to the inner tegument. A number of studies have pointed to the essential role of both pUL36 and pUL37 in virion morphogenesis by demonstrating a block on secondary envelopment in the absence of either protein. However, the phenotypes attributable to deletion mutants of UL36 and UL37 varied between HSV-1 and the related alphaherpesvirus, Pseudorabies virus (PrV), and between different PrV mutants. In particular the phenotypes of an HSV-1 mutant unable to express functional pUL36 due to an internal deletion between amino acids 362-1555 (KΔUL36) and PrV mutants lacking UL36 showed considerable variation. Both accumulated DNA filled cytoplasmic nucleocapsids, however, those produced by KΔUL36 formed small aggregates within the cytoplasm in contrast to PrV where the capsids were dispersed. UL36 and UL37 deletion mutants were also reported to retain capsids to the nucleus implying a function in nuclear exit. In order to clarify the seeming disparities between these mutants, novel deletions of both UL36 and UL37 were engineered into HSV-1 17(+). The deletion of UL36 encompassed the entire open reading frame (ORF) while the UL37 deletion removed all but the 3 most C terminal amino acids. Examination by EM revealed that these mutants recapitulate the block on secondary envelopment that is characteristic of inner tegument mutants reported to date, with accumulation of nucleocapsids within the cytoplasm. This examination revealed that the cytoplasmic capsids of the complete UL36 deletion adopted the dispersed phenotype seen in PrV suggesting the phenotype of KΔUL36 is specific to the mutation introduced. Furthermore, this analysis failed to demonstrate a retention of capsids within the nucleus in the absence of either pUL36 of pUL37, indicating that neither has a definitive role in nuclear exit, and the principle function of these proteins is within secondary envelopment. Work presented within this thesis using IF and Immuno-EM methods shows that pUL36 is associated with the cytoplasmic clusters of capsids produced by a mutant lacking pUL37 suggesting that this may contribute to the aggregation of these capsids. Furthermore, these pUL36-decorated cytoplasmic nucleocapsids appear able to recruit pUL48 but are blocked in secondary envelopment, in contrast to PrV lacking UL37 where virions are formed but at much reduced levels. In the absence of pUL36 no tegument is recruited to cytoplasmic nucleocapsids but both minor capsid associated proteins pUL17 and pUL25 are present. Cytoplasmic capsids of mutants lacking pUL36 and pUL37 have been isolated and examined biochemically. Cytoplasmic capsids of KΔUL36 were found to be associated with the N terminal fragment (1-361 amino acids) of pUL36 expressed by this virus, implying that a capsid binding activity is present within this region, which is the first demonstration of such an activity outside of the previously mapped activity within the 62 C terminal amino acids of pUL36. The difference in behaviour between KΔUL36 and the complete UL36 deletion also implies that the aggregation of capsids in KΔUL36 is caused by the presence of this fragment. Assembly of L particles, structures consisting of tegument and envelope but lacking capsids, was unaffected by the deletion of either UL36 or UL37. Biochemical characterisation of purified L particles revealed that deletion of pUL36 prevents the incorporation of pUL37 and deletion of pUL37 leads to the absence of pUL36 despite their expression during infection. This suggests that incorporation of inner tegument to L particles, and envelopment of nucleocapsids, requires pUL36 and pUL37 to be present in a complex. The sole exception to this was the incorporation of the N terminal fragment of pUL36 in KΔUL36 infection into L particles in the absence of pUL37 the mechanism for which is unknown. Little of the role of the inner tegument proteins in initiating infection has been established. Previous studies have hinted at functions in nuclear pore complex binding and genome release, while a growing body of data implicates pUL36 and pUL37 in cytoplasmic transport of nucleocapsids. As both pUL36 and pUL37 are essential in HSV-1 virion morphogenesis a method was required to establish the defects of inner tegument deletion mutants in the initiation of infection. This was provided by infecting cells at a low multiplicity with mutant viruses (0.01pfu/cell) and subsequently fusing the cells into syncytia, in which several nuclei shared a common cytoplasm. The spread of viral DNA replication in these nuclei was tracked by fluorescence in situ hybridisation. In each syncytium the initially infected nucleus produces nucleocapsids that enter the common syncytial cytoplasm and have access to naïve nuclei without needing to undergo secondary envelopment. Using this system the essential role of pUL36 in the initiation of infection was confirmed and it was also shown that pUL37 is dispensable for the initiation of new cycles of viral DNA replication in naïve nuclei. The behaviour of cytoplasmic nucleocapsids within syncytia seems to be equivalent to that of incoming virion derived tegumented nucleocapsids as WT HSV-1 and HSV-1 ΔUL37 infections behaved similarly and were equally affected by treatment with nocodazole suggesting that both are undertaking similar activities. The work presented here has helped to clarify the roles of the inner tegument proteins during virus assembly and the methods developed have extended our ability to assess the contributions of proteins essential for virion formation, in the initiation of infection. The studies presented here reinforce the idea that pUL36 may play multiple roles in the initiation of infection.