The aim of this thesis is to work towards establishing the molecular basis of the communication between the host nematode <i>Brugia malayi</i> and its host specific <i>Wolbachia</i> strain, <i>w</i>Bma. The genome sequence of <i>w</i>Bma was annotated and compared to the published annotated genome of the <i>Wolbachia</i> found in <i>Drosophila melanogaster </i>(<i>w</i>Mel). Interesting genes discovered in the <i>w</i>Bma genome include a family of seven ankyrin repeat containing (ANK) proteins. ANK proteins are known to be involved in many protein:protein interactions, more commonly in eukaryotes. A comparison of the <i>w</i>Bma ANK proteins with the 23 ANK proteins present in the genome of <i>w</i>Mel shows that two of the <i>w</i>Mel ANK homologues in the <i>w</i>Bma genome are non-functional pseudogenes. <i>w</i>Bma has novel ANK proteins that may be involved in interaction with the host cellular machinery. RT-PCR of the <i>w</i>Bma ANK genes shows that different ANK genes are expressed in different stages of the nematodes’ life cycles. Many interesting <i>Wolbachia</i> novel genes were identified in the genome of <i>w</i>Bma, <i>w</i>Mel and the <i>Wolbachia</i> strain from the filarial nematode that causes river blindness, <i>Onchocerca volvulus</i> (<i>w</i>Ovo). Novel proteins predicted to be located to the bacterial surface or predicted to be secreted are good candidates for proteins that may be involved in communication between <i>Wolbachia </i>and its host. A real-time PCR technique was developed to calculate the number of <i>w</i>Bma per <i>B. malayi </i>nucleus at different stages in the nematodes’ life cycle. gDNA samples prepared from different stages of the life cycle showed that the number of <i>w</i>Bma per nematode nucleus was observed to increase ten-fold during development from microfilaria (MF) to larval fourth stage (L4). Post-L4, while adult male somatic nuclei had approximately six-fold that of L4, the adult female soma is estimated to have approximately 330-fold more. If the <i>Wolbachia</i> were multiplying only to ensure its passage to the next generation of MF, there would be little obvious adaptive advantage in multiplication within the soma. However, if the <i>Wolbachia</i> are supplying some essential metabolic need to the nematode, then their multiplication in the soma will reflect the importance of their role at different lifecycle stages. This data suggests that <i>w</i>Bma are supplying a possible essential role, but that role is yet to be identified. Another way that <i>Wolbachia</i> and its host may be interacting is by gene transfer from the bacteria into the nuclear genome of the host organism. The transfer of genes from prokaryote to eukaryote is still considered an unusual event, but there is evidence that genes have been transferred from organelles of prokaryotic endosymbionts origin, such as chloroplasts and mitochondria to the cell’s nuclear genome. This thesis has identified two <i>w</i>Ovo pseudogenes in an untranscribed region of an <i>O. volvulus </i>nuclear gene.

A deletion analysis of the <I>hsdS </I>gene of <I>Eco</I>KI was initiated to provide information on the roles of the conserved and variable regions. The aim was to correlate phenotype with an analysis of protein products, but attempts to overexpress the <I>hsdS </I>gene of <I>Eco</I>KI in a soluble form were unsuccessful, and the HsdS subunit could not be purified. Another approach to studying the interaction of TRDs with their DNA targets is to compare the amino acid sequences of those TRDs that specify the same DNA target. Areas of sequence which are similar within these TRDs may reflect a similarity of DNA recognition function. From amongst the limited number of TRDs available, there are several sequence alignments of TRDs which specify the same DNA target. A method based upon the Polymerase Chain Reaction (PCR) was developed to amplify new variable regions, which encode TRDs, from wild-type bacteria. In a DNA hybridisation screen of members of the ECOR collection of wild-type <I>Escherichia coli, </I>Barcus <I>et al. </I>(1995) found that almost half contained <I>hsd </I>genes. The conserved regions of the <I>hsd </I>genes were used to design primers that would amplify 5' variable regions of members of a given type I family. Nine IA family and four IB family 5' variable regions were amplified and their DNA sequences determined. The information derived from these sequences illustrates both the evolutionary diversity of the <I>hsdS </I>genes, and the flexible nature of the TRDs as independent target recognising domains. The sequencing of the N-terminal TRD from ECOR17 shares 28% identity with those of <I>Eco</I>KI and <I>Sty</I>SPI.

The <i>trans</i>-acting factors that can interact with the <i>gurken</i> localization signal were investigated. RNA affinity chromatography identified a number of proteins that have a role in the localization and localized translation of a variety of mRNAs in <i>Drosophila</i> and other organisms, in addition to factors previously shown to be important for dorso-anterior <i>gurken</i> localization and translation. These included Hephaestus and CG17838, which were chosen for further study due to the roles of their vertebrate homologues, PTB and SYNCRIP respectively, in neuronal mRNA localization. Both Hephaestus and CG17838 accumulate at the posterior of the oocytes, in a similar manner to localized <i>oskar</i> mRNA, and can associate with <i>gurken</i> and <i>oskar</i>. Furthermore, CG17838 can interact with the known <i>gurken</i> <i>trans</i>-acting factors, Squid and Hrb27C. Analysis of <i>hephaestus</i> mutants suggests that Hephaestus has multiple roles within the <i>Drosophila</i> oocytes including the maintenance of Gurken protein at the dorso-anterior in later stages of oogenesis. I hypothesize that Hephaestus acts to couple <i>grk</i> mRNA localization with localized Gurken processing and secretion. CG17838 is essential for adult viability, and preliminary studies of CG17838 germline clones suggest that CG17838 has a role in <i>gurken</i> localization or localized translation. Furthermore, CG17838 mutant adult flies have a behavioural phenotype indicative of a defect in synaptic transmission. The work is consistent with the concept that core, conserved proteins are involved in the regulation of a number of localized mRNAs. The conservation of a number of the <i>trans</i>-acting factors required for mRNA localization also means that the study of these factors during <i>Drosophila</i> oogenesis can lead to a greater understanding of how the process can go awry in human disease. This may be particularly important for understanding certain disorders of the nervous system, such as Fragile X syndrome and Spinal Muscular Atrophy (SMA).

β-1,4-Mannosyltransferase plays an essential role in the biosynthesis of <i>N</i>-glycans. The application of the enzyme for biocatalysis is hampered by poor expression levels of the corresponding <i>Saccharomyces cerevisiae </i>gene, <i>ALG</i>1 and by very high specificity for phospholipid structures. The aim of the current project has been to investigate ways in which production of the enzyme can be improved and new substrate specificity can be studied. A series of β-1,4-mannosyltransferase fused to either the <i>C</i>-terminus of the green fluorescent protein (GFP) or the <i>C- </i>or <i>N</i>- terminus of the SNAP-tag were produced in <i>Escherichia coli. </i>Alternatively, improved expression levels of the gene fusions and stability of the chimeric proteins were sought by switching to a <i>Pichia pastoris </i>expression system. An <i>N</i>-terminal GFP fusion produced from a bacterial expression vector was the most promising and the corresponding gene fusion was used in directed evolution studies for improved expression levels. Key to these studies was the development of a new assay that could be used in a ‘high-throughput’ fashion. Screening for active enzyme was achieved using hierarchical methods: firstly, identification of colonies exhibiting good fluorescence indicated production of folded protein. Secondly, an affinity assay using GDP-sepharose was developed to isolate fluorescent mutants binding to GDP. A library of <i>ALG</i>1 mutants fused to GFP was generated by error-prone PCR and after screening, a number of promising candidates was isolated and characterised. None of these mutants gave reliably high expression levels than the wild-type gene. It was suggested that the levels of expression of the genes were not stable enough to identify any useful high-expression mutants.

The aim of this project was to develop techniques for the measurement of cytosolic free calcium ([Ca<SUP>2+</SUP>]<SUB>c</SUB>) in <I>Neurospora crassa, </I>using the Ca<SUP>2+</SUP>-sensitive photoprotein aequorin. An <I>in vitro</I> assay of aequorin luminescence was developed for the screen of apoaequorin yield in primary transformants. The assay confirmed that the apoprotein produced by <I>N. crassa </I>was fully functional. The yield of apoaequorin from both transformant series was low, at best 13 <I>fg/</I>μg total soluble protein. Isolation of homokaryotic derivatives from primary transformants improved this level by up to 11-fold, though this was still 32-times lower than that obtained in apoaequorin transformed <I>Nicotiana plumbaginifolia. </I>Possible reasons for poor transgene expression were explored. Southern analysis showed that virtually all transformants contained at least one intact apoaequorin expression cassette. Northern analysis revealed an apoaequorin mRNA which corresponded to the expected transcript size. Apoaequorin mRNA abundance in <I>N. crassa </I>primary transformants was at best approximately 2-fold less than in <I>N. plumbaginifolia, </I>this was not reflected in the protein yield which was almost 400-fold lower. Northern analysis confirmed that the block acting on apoaequorin yield must be at translation or beyond. The codon requirements of apoaequorin and codon usage in a number of apoaequorin transformed species was examined. In general, species which showed a much more compatible codon usage produced more apoaequorin. Apoaequorin was shown to be somewhat unstable in <I>N. crassa </I>with a half life of approximately 45 minutes, which may be a contributory factor to the low yield. Methods were developed whereby the transformed strains were able to report changes in their [Ca<SUP>2+</SUP>]<SUB>c</SUB>. Both apoaequorin and coelenterazine were shown to be non-toxic, and aequorin was regenerated in cells simply by incubation with coelenterazine. Background sources of luminescence were identified and methods for lysing <I>N. crassa</I> cells were developed. Lysis is a prerequisite for the calibration of Ca<SUP>2+</SUP> concentration from aequorin light emission.

Although tremendous progress has been made in many other groups, the forces and factors which affect synonymous codon use in mammals remain something of a mystery. At least some of the differences in codon usage between mammalian genes can be summarised in terms of composition: within any one species some genes have very low G+ C contents (< 30%) and others very high G+ C content (> 90&37), with the majority lying somewhere in between. The very simplicity of this trend and the fact that this composition is correlated to that of introns and isochores suggests that the differences in synoymous codon use may be the result of variation in the pattern of mutation across the genome. This hypothesis is examined by considering the three most likely ways in which the mutation pattern might vary across the genome: (1) temporal changes in the performance of the replicative machinery; (2) variation in the efficiency of DNA repair; and (3) variation in the frequency of gene conversion across the genome. Evidence is found against all these hypotheses. Principally none of them predict the silent substitution rate to be related to G+ C content in the manner which is observed. Furthermore the lack of any discernible difference between the silent site G+ C contents of early and late replicating genes, and the very small parameter range over which DNA repair can generate large differences in synonymous codon use, support the conclusions that replication and repair, respectively, are not responsible for the codon use of mammalian genes. It is therefore suggested that selection might act upon synonymous codon use. However an analysis of codon usage within genes suggests that selection of the type commonly found in other groups, selection upon tRNA interaction, is not operative in mammals. It is tentatively suggested that selection upon mRNA secondary structure might be the responsible agent. Some of the results obtained also have implications for the maintenance of isochores. Since the G+ C contents of isochores and silent sites are correlated, the lack of any distinction with respect to composition between early and late replicating genes suggests that the differences in isochore G+ C content are not cuasd by DNA replication. However it is hypothesised that variation in the frequency of recombination can provide a very elegant explanation of the differences in isochore G+ C contents, and the relationship between gene density and isochore G+ C content.

This thesis describes the optimisation and implementation of a conditional transgene expression system in mouse embryonic stem (ES) cells for the investigation of signalling pathways controlling <i>in-vitro</i> differentiation. This system, the tetracycline-induced conditional gene expression system (tet-system) allows exquisite control over the transgene expression <i>in vitro</i> and <i>in vivo</i>. The system itself is comprised of two components, a transactivator gene, and a transactivator-response element (TRE) which is linked to the transgene to be regulated. The transactivator gene encodes a fusion protein consisting of a modified tetracycline repressor and transcriptional activation domain moieties, and is constitutively expressed. In the presence of doxycyline (dox) the transactivator protein binds to the TRE sequence, which itself consists of an array of operator sequences from the tetracycline operon linked to a minimal promoter region. Transcription of the transgene consequently ensues. This version of the tet-system is referred to as the tet-on system. In order to establish high and homogenous expression of inducible transgenes in Es cells using the tet-on system, the utilisation of different promoters to direct expression of the transcriptor component (rtTA2^s-M2) was investigated, as was the use of different loci at which to target the system components. It was found that the promoter sequence CAG, when linked to the transactivator rtTA2^s-M2 sequence and targeted to the 5’ region of the housekeeping gene <i>hypoxantine ribosyltransferase (hprt)</i>gave rise to homogenous expression of the transactivator. However upon targeting the 3’ region of <i>hprt</i> with the TRE and a linked DsRED2 reporter transgene, addition of dox resulted in heterogenous expression of the transgene. In an alternative strategy, cells expressing the rtTA2^s-M2 transactivator from the endogenous ROSA26 promoter (R26rtTA2^s-M21F cells), were targeted to integrate the TRE and linked DsRED2 reporter to the 5’region of <i>hprt. </i>In cell lines harbouring both components of the tet-system, addition of dox resulted in high and homogenous expression of the DsRED2 transgene.

No description available.

The cohesin complex is primarily composed of two Structural Maintenance of Chromosomes proteins (SMC1 and SMC3), and two non-SMC subunits (rad2l/Sccl and SA1/Scc3). We have demonstrated that the <i>Drosophila </i>rad2l/Sccl homologue, Drad2l, is in a complex with SMC1, SMC3, and SA1/Scc3 and similar to other higher eukaryotes, the majority of Drad21^-dissociates from chromosomes during prophase, although a small pool remains associated with centromeres until anaphase. In my work I have shown that the specific depletion of Drad21, by RNA interference in <i>Drosophila </i>S2 cells, results in a prometaphase delay, premature sister chromatid separation, and the selective destabilisation of SA1/Scc3. In addition, I also observed that the chromosome passenger protein INCENP becomes mis-localised in Drad21 depleted cells. As a collaborative aspect to this work I have been involved in the characterization of a number of Drad21 cleavage mutants. In <i>S. cerevisiae </i>the dissolution of sister chromatids is brought about by cleavage of scclp by a conserved cysteine protease Espl/Separase. Separase cleaves Sccl at a conserved site in budding and fission yeasts, humans and <i>Xenopus. </i>Although three potential cleavage sites have been identified in the <i>Drosophila </i>sequence, only one, gives rise to a phenotype when mutated. As a further project I have been working on a potential chromatin remodelling protein, which when deleted from <i>S. pombe </i>results in segregation defects and sensitivity to microtubule poisons. This protein has also been implicated in the correct targeting of rad2l to centromeric heterochromatin. This protein is conserved from yeast to mammals and I have been specifically characterizing the <i>Drosophila </i>homologue including a number of mutant alleles. Intriguingly, antibodies to this protein show a striking centromeric localization on mitotic chromosomes.

The factors affecting deletion of long DNA palindromes from high copy number cloning vectors are investigated with particular reference to the mode of deletion and it is shown that the deletion of a 571 bp palindrome deleted from pMS7 using 3 bp repeats. A shorter 109 bp palindrome deleted from a related plasmid using 7 bp direct repeats and the mode of deletion is unaffected by the genotype of the host strain. There also appears to be a bias for the deletion of palindromic sequences on the lagging strand of a replication fork. It is also shown that in a wild-type <I>E. coli</I> strain there is inhibition of plasmid multimerization if the plasmids carry long palindromic sequences. It is proposed that the lack of plasmid multimers in this background is a result of the removal of palindromic sequences form the plasmids by the SbcCD protein of <I>E. coli</I>. In an <I>sbcCD </I>strain, plasmid DNA bearing long palindromes is not detected in a monomeric form, instead the DNA is present in multimeric forms, predominantly dimers. The ability to form plasmid multimers in this background may help to stabilise the palindromic sequences. The behaviour of palindromic sequences carried on plasmids is also investigated in <I>recA</I> <I>sbcCD</I> strains with a view to the correct choice of <I>E. coli</I> strain for the cloning of long palindromic sequences. Finally, the influence of an <I>sbcCD</I> mutation on the formation of <I>araB-lacZ</I> cistron fusion is investigated. The SbcCd proteins are thought to have a role in the processing of secondary structures formed by palindromic sequences. The formation of <I>araB-lacZ </I>fusions occurs via a strand transfer complex involving a complex genome rearrangement and secondary structure. Although an <I>sbcCD</I> mutation did not affect the kinetics or sequences specificity of fusion formation it is possible that SbcCD might have a role in processing the strand transfer complex which is not easily detected.