Restriction factors are endogenous cellular proteins that block retroviral replication at specific points in the life cycle. Those identified so far include Fvl, Trim5a and TrimCyp. Their characterisation has extended knowledge of retroviral and cellular functions, and has added a new branch to innate immunity. Retroviral susceptibility to Fvl and Trim5a is determined by its capsid, and is manifested in a pre- (Trim5a, TrimCyp) or post- (Fvl) reverse transcription block to replication. Other blocks to replication have been postulated. For example, a novel anti-viral factor, Lv2, is thought to block replication of several primary isolates of HIV-2 in some cell lines. Knowledge of the early steps of virus replication, between entry and nuclear import, is critical to understanding restriction. The intention of the studies described in this thesis was to determine whether alternative routes of virus trafficking might affect susceptibility to Fvl and Trim5a, as well as to the putative Lv2. A system of two receptors was used, Tva800 and Tva950 both permit entry via ASLV envelope protein, but take the virus into the cell by two different endocytic mechanisms. The pathways traversed after binding to Tva800 and Tva950 were investigated and shown not to reroute virions around restriction mediated by Fvl and Trim5a. When virus titration curves were analysed, a distinctive pattern emerged suggesting that entry via Tva800, but not Tva950, requires engagement of more than one receptor- envelope pair. The block to replication caused by the putative factor Lv2 was also analysed. It was concluded that a combination of low surface CD4 expression and poor receptor engagement are the cause of low viral titres in some cell lines, rather than a cellular anti-viral factor per se.

Restriction endonucleases that cleave double stranded DNA molecules to produce discrete DNA fragments with mutually cohesive ends, provide a simple method for the production "in vitro" of genetic rearrangements fusing genes normally unlinked. This thesis describes the construction of two such genetic options and their use in the study of the two processes that govern lysogeny, namely, repression and integration.

Narcissus latent virus (NLV) and Madura mosaic virus (MacMV) are serologically related. However, they have poor serological relationships with other plant viruses with which they have shared characteristics. Coat protein size, particle shape and structure, mode of vector transmission, cytology and serology proved insufficient to classify them. Molecular techniques were employed in order to create tools for the rapid and accurate classification of plant viruses. A carlavirus-specific PCR primer test failed to amplify NLV or MacMV but confirmed that several other viruses belonged to the carlavirus genus. The nucleotide sequences of part of the nuclear inclusion body (NIb) gene, the complete coat protein gene and the 3' untranslated regions of narcissus latent virus (NLV) and Madura mosaic virus (MacMV) were determined. Deduced amino acid sequences for the Nib and coat protein genes revealed that NLV and MacMV are closely related. Comparison of the NIb sequences with other viruses showed that NLV and MacMV have closer affinities with viruses of the Potyviridae than to those of the carlavirus genus with which they were initially classified. It is proposed that NLV and MacMV may form a new genus within the Potyviridae, the Macluraviruses. The viruses associated with narcissus yellow stripe disease were re-evaluated. In order to identify further members for the new genus a second PCR primer was designed to amplify a region shared by the bymoviruses and macluraviruses. Sequence data obtained for the 3'-terminal region of rice necrosis mosaic virus (RNMV) using a fragment amplified by this primer confirmed that RNMV was a member of the bymovirus genus. MacMV and barley mild mosaic bymovirus replicase sequences were used to transform tobacco plants in an attempt to create transgenic resistance.

In this project, the receptor binding and membrane fusion activities of the membrane-anchored glycoprotein haemagglutinin (HA), H3 subtype, of the influenza A virus were studied. Influenza viruses from different hosts can distinguish between a.2-3 and a2-6 linkages that sialic acid (Sia) forms with the penultimate saccharide residue of the receptor. Human viruses prefer a2-6 linkages and avian viruses ct2-3 linkages. The side chain of residue 226 of the HA receptor binding site (RBS) is involved in the specific recognition of those linkages. The HA of the 1968 Hong Kong (HK) pandemic virus (H3 subtype), contained in the recombinant X-31 virus, has a 226-leucine and prefers binding to a2-6 linkages. The L226Q HA of the variant virus X-31/horse serum (X-31/HS), corresponding to a single-site mutant of the X-31 HA having 226-glutamine, prefers a2-3 linkages. To define the molecular interactions of the L226Q HA with the virus receptor, the crystal structures of L226Q HA in complex with the receptor analogues lactoseries tetrasaccharide a (LSTa), a2-3-terminated, and lactoseries tetrasaccharide c (LSTc), a2-6-terminated, were determined. The structures show the saccharide residues Sia and galactose (Gal) of LSTa and only the Sia of LSTc bound on the L226Q HA RBS, which correlates with the higher affinity of L226Q HA for the ct2-3 linkage. However, the L226Q HA binds both the trans and cis configurations of the Sia-Gal glycosidic bond of LSTa, which has never been observed before. These results are discussed in comparison with data from binding assays and the available crystal structures of the X-31 HA, the H3 avian HA of the influenza virus A/duck/Ukraine/63, a potential precursor of the 1968 HK pandemic virus, and other HAs, of different subtypes and hosts of origin, in complex with LSTa and LSTc. Upon endocytosis, at the low endosomal pH, the HA undergoes an irreversible conformational change associated with the fusion of the viral and endosomal membranes, a process by which the virus enters the target cell. To gain further insights into the membrane fusion mechanism catalysed by the influenza HA, the purification and crystallisation trials of three different X-31 HA forms representative of the neutral pH and fusion pH-induced conformations, containing the membrane-interacting segments fusion peptide and transmembrane anchor, were carried out with the future aim of their crystal structure determination. The procedures for the purification and crystallisation trials of the X-31 HA proteins in different conformations are discussed in the context of the general purification and crystallisation of membrane proteins. No promising hint has yet been obtained in the crystallisation trials.

Prion diseases, also known as Transmissible Spongiform Encephalopathies (TSEs), are fatal conditions which affect humans and animals. The hallmark of Transmissible Spongiform Encephalopathies is the accumulation in the brain of PrPSc, which is an abnormal isoform of the cellular protein PrPc. It has been proposed that PrPSc is able to impose its own conformation on PrP , but the molecular mechanism by which PrP is converted into PrPSc is unclear. PrPc and PrPSc share the same primary sequence but they are different in their biochemical properties. PrP is rich in alpha helices, detergent soluble and Protease K sensitive on the other hand, PrPSc is mainly composed of beta sheets, detergent insoluble and Protease K resistant. To better understand PrPSc generation, a genetic screen to identify proteins that preferentially interact with a misfolded version of PrP has been employed. This novel protein-protein interaction system takes place in the yeast cytoplasm where PrP has been reported to adopt a beta sheet-rich conformation characterised by increased Protease K resistance. Initial analysis revealed that one candidate molecule, PrP Interacting Protein 7 (PIP7), encoded a transmembrane protein present on the cell surface, suggesting a potential function as a novel PrP receptor. The interaction between PIP7 and PrP has been confirmed in vitro by GST pull down experiments and in vivo by co- immunoprecipitation studies. A panel of monoclonal antibodies directed against the C- terminus of PIP7, which contains the PrP binding region, has been generated. FACS analysis using these antibodies confirmed the presence of PIP7 on the cell surface of N2a cells, commonly used to study PrPSc propagation, and PIP7 immunohistochemistry has revealed an intense staining in all brain regions including the cortex, hippocampus and cerebellum. Overexpression of PrP in N2a cells resulted in an accumulation of PIP7 surface levels. Furthermore, PIP7 levels appear to be reduced in the brains of PrP knockout mice. The direct correlation between the levels of PIP7 and PrP suggests a potential role for PIP7 as a regulator in prion biology.

The coronavirus genome is a positive-stranded RNA of extraordinary size and complexity. The largest protein encoded by coronaviruses is non-structural protein 3 (nsp3), which contains many domains of unknown function. In this study, a vaccinia virus-based reverse genetics system was used to introduce seven mutations in the C-terminal domain of nsp3, known as the Y domain. Of these, four recombinant MHVs were constructed successfully. However, none of them was rescued, suggesting that the Y domain performs an important function in the MHV replication cycle. The function of nsp3 in replication was then probed by biochemical and structural methods using the temperature-sensitive mutant Brts31 which contains a mutation in nsp3 along with four other temperature-sensitive mutants that contained mutations in each of the other cistrons that function in viral RNA synthesis. Interestingly, Brts31 and several other viruses tested formed smaller and fewer DMVs compared to the parental wild-type virus under conditions where both viruses produced an equal amount of progeny virus. This suggested that the efficiency of producing progeny virus is not closely related to the prevalence or size of DMVs. Two MHV mutants (Brts31 and Brts105) made half as many DMVs as normal. Despite differences in DMV size and number, all mutants replicated as efficiently as wild-type virus. To better understand the relative importance of replicative organelle formation , we carried out competitive fitness experiments using viruses with different DMV phenotypes. None of these viruses was found to be significantly less fit than wildtype, and two were actually fitter in tests in two kinds of cells. This suggests that viruses have evolved to have tremendous plasticity in the ability to form membrane associated replication complexes, and that large and numerous DMVs are not exclusively associated with efficient coronavirus replication.

Eye specific CDP-diacylglycerol synthase is a key regulator in the phototransduction pathway of the fruit fly Drosophila melanogaster. The enzyme is responsible for maintaining regeneration of phosphatidylinositol 4, 5 bisphosphate (PIP2) which is required for normal vision. The Drosophila eye-cds mutants display light dependent retinal degeneration. Mammalian homologues of Drosophila mutants such as eyeless and eyes absent have been associated with mammalian disease. Based on these studies and the discovery of CDS expression in human retina, this gene would appear to be a good candidate for causing a retinal phenotype in mammals. The aim of the work presented here has been to isolate and characterise the murine homologues of eye-cds. To this end both cDNA and genomic clones have been isolated using a combination of traditional molecular biology techniques and a bioinformatic approach. Using the identified clones, the structures of Cdsl and Cds2 have been confirmed. The clones were subsequently used to characterise the genes. In this thesis I have identified the expression pattern of both genes, showing that Cds2 is ubiquitously expressed in both the embryonic and adult mouse whereas Cdsl expression is more specific to the adult eye and brain. In addition, Cdsl has been shown to be expressed only in the rod photoreceptor cells within the retina. Cdsl and Cds2 clones were also used to map the genes to chromosome 5E1 and 2G1 respectively. The remit of this thesis was further expanded to include the potential function of the murine genes. To this end, transgenic constructs of both genes were made for the rescue of the mutant Drosophila phenotype and a construct was prepared as a first step in the generation of a Cds2 null mouse. In summary, the study reported in this thesis has succeeded in isolating and characterising the murine Cdsl and Cds2 genes. The expression pattern of the two genes has been determined both in a large panel of tissues and within specific cell types. The chromosomal location of both genes has been determined and constructs made for use in developing transgenic mice and flies for the study of gene function.

The core antigen (HBcAg) of hepatitis B Virus (HBV) can be expressed in <I>Escherichia coli </I>where it assembles into icosahedral particles of two sizes containing 240 or 180 subunits. These particles could be fractionated by sucrose gradient centrifugation, and light scattering showed their size distribution to be essentially monodisperse. HBcAg is 183 amino acids long and highly enriched in arginine residues in the C-terminal region. Around 50% of these residues are encoded by the rare triplet AGA in <I>E. coli. </I>Supplementation of the level of AGA tRNA in the cell with plasmids expressing the T4 AGA tRNA gene significantly enhanced the yield of HBcAg. SDS polyacrylamide gel electrophoresis of the two kinds of particles showed that around half of their subunits were smaller than the full length HBcAg and varied in size. N-terminal sequence analysis revealed that these smaller species were heterogeneous at the extremely basic C-terminal end. In the virus, the icosahedral nucleocapsids are surrounded by an envelope consisting of cellular lipids and three related surface antigens (L-, M-, and S-HBsAg) which result from alternative translation initiation of a common reading frame. The L-HBsAg is believed to mediate the contact between the envelope and nucleocapsid. The N- and C-termini of this protein were shortened in order to define the minimum stretch of amino acids that contains the exact contact residues. To determine which residues are directly involved in the interaction, single and multiple mutations were generated by site-directed mutagenesis. The resulting mutated proteins were expressed in rabbit reticulocyte lysates, and their ability to interact with HBcAg was examined with an immunoprecipitation assay and a newly established equilibrium binding assay in solution which allows the determination of relative dissociation constants with Scratchard and non-linear regression analyses.

In section A, five papers are presented which examine the evolution of HIV-1 both within and between patients. The first paper presented examines the molecular epidemiology of HIV-1 within Scotland, Northern England, and Ireland (paper I), with attention focused on identifying risk group associated differences within the cohort. This work also provides important background information for the interpretation of molecular data from transmission clusters. The main focus of the work on HIV-1 evolution has been on the transmission of HIV-1, with particular emphasis placed on mother-child transmission. Four papers are presented which examine evolutionary aspects of HIV-1 transmission. The first of these (paper II) examines the viral variants transmitted from mother to child in four mother-child transmission pairs. The second (paper III) analyses similar data from five mother-child transmission pairs, focusing predominantly on viral evolution within the child over the first year of life. The final two papers investigating HIV-1 transmission examine viral variation within two transmission sets. Paper IV examines the vertical transmission of HIV-1 to two infected children born to the same mother at an approximately two year interval, while paper V examines the heterosexual transmission of HIV-1 from a male index to two female contacts and the subsequent vertical transmission of HIV-1 to their two children. The phylogenetic placement of these transmission sets within the Edinburgh cohort is also assessed. In section B, four papers are presented which examine aspects of foraminiferal evolution. The first paper (paper I) focuses on the problems inherent in the amplification of foraminiferal DNA due to the association of large numbers of symbionts, commensals and food particles with each foraminifer. The amplification of foraminiferal sequences for the small subunit ribosomal RNA gene is then described, and the phylogenetic placement of the foraminifera within eukaryote evolution examined (papers II and III). Finally, the phylogenetic relationships within the foraminifera are described (paper V).

This thesis confirms the hypothesis of Low (1973) that many E.coli K-12 strains contain a prophage (the Rac prophage) located a few minutes clockwise of the txp operon on the genetic map. Restriction endonucleases and 32P-labelled phage RNA probes were used to investigate several E.coli K-12 DNAs and hence construct a physical map of this prophage. Some E.coli K-12 strains have lost the entire prophage by a specific deletion. This is consistent with prophage excision by site-specific deletion. X reverse (x-rev)phages are recombination proficient derivatives of phage 1 in which the phage recombination functions have been replaced by analogous functions (RecE) derived from the host chromosome. The data of this thesis support the origin of x-rev phages by recombination between 1 and the Rac prophage following excision of the Rac genome from the E.coli K12 chromosome. E.coli K12 strains which carry 2L mutations express the Rac exonuclease gene (recE) constitutively. Investigation of the DNAs of several such strains showed them to fall into more than one class. In sbcAS strains a large section of the Rac genome (including a hybrid attachment site and probably the prophage repressor gene) is deleted. Several other sbcA-strains carry multiple (and probably tandemly repeated) copies of the prophage genome. Agar phages are characterised by the replacement of the region of the A genome that contains and the late gene promote, P'R, with host-derived DNA that codes for functions analogous to those deleted. This thesis shows the substitutions in Xgsr phages to be derived from a second (and as yet unlocalised) lambdoid prophage (the gsr Prophage) in E.coli K12.