A novel DAN-based method was used to infer levels of evolutionary constraints in the <i>Drosophila</i> genome by comparing rates of nucleotide substitution in non-coding and putatively neutrally evolving DNA. Introns were found to have a significantly higher rate of substitution than synonymous sites, and, when introns were used as a neutrally evolving standard, constraint in the 500bp of intergenic DNA upstream and downstream of coding regions was found to be about 44%. Selection against mutations in intergenic DNA should therefore make a substantial contribution to the mutational load in <i>Drosophila</i>. Secondly, a fitness- based approach was used to estimate mutational parameters in lines of <i>Caenorhabditis elegans</i> containing large numbers of deleterious homozygous EMS-induced mutations. Replicated inbred sublines were produced for eight mutant lines, and the performance of the sublines, the mutant lines and the wild-type strain was measured for three fitness-related traits. The number of mutations per line was then estimated for each trait by applying a modified version of the Castle-Wright estimator and a maximum likelihood (ML) method. Both the Castle-Wright and the ML analyses suggest that most of the variation among sublines was due to a small number (~1.5-2.5) of large-effect mutations, given that each line is expected to have a large number of mutations, this supports the hypothesis that many have very small (but still deleterious) effects. The average dominance coefficient of mildly deleterious mutations was estimated from a selection of 19 relatively high fitness mutant lines by comparing the performance of heterozygotes and homozygotes to the wild-type for three fitness-related traits (viability, productivity, and relative fitness). There was very little effect of mutations on viability, but for productivity and relative fitness was found to be ~0.1. Combined with the conclusion that most homozygous mutations have very mild effects, this suggests that many newly arising deleterious mutations may have very small heterozygous effects indeed.

A novel method is developed for combining multiple laser scans of microarrays to correct for “signal saturation” and “signal deterioration” effects in the gene expression measurement. A multivariate nonlinear functional regression model with Cauchy distributed errors having additive plus multiplicative scale is proposed as a model for combining multiple scan data. The model has been found to flexibly describe the nonlinear relationship in multiple scan data. The heavy tailed Cauchy distribution with additive plus multiplicative scale provides a basis for objective and robust estimation of gene expression from multiple scan data adjusting for censoring and deterioration bias in the observed intensity. Through combining multiple scans, the model reduces sampling variability in the gene expression estimates. A unified approach for nonparametric location and scale normalisation of log-ratio data is considered. A Generalised Additive Model for Location, Scale and Shape (GAMLSS) is proposed. GAMLSS uses a nonparametric approach for modelling both location and scale of log-ratio data, in contrast to the general tendency of using a parametric transformation, such as arcsinh, for variance stabilisation. Simulation studies demonstrate GAMLSS to be more powerful than the parametric method when a GAMLSS location and scale model, fitted to real data, is assumed correct. GAMLSS has been found to be as powerful as the parametric approach even when the parametric model is appropriate. Finally, we investigate the optimality of different estimation methods for analysing functional regression models. Alternative estimators are available in the literature to deal with the problems of identifiability and consistency. We investigated these estimators in terms of unbiasedness and efficiency for a specific case involving multiple laser scans of microarrays, and found that, in addition to being consistent, named methods are highly efficient and unbiased.

A membrane-DNA complex has been isolated from Sarkosyl lysed mitochondria of Paramecium aurelia. Apart from DNA, which has been demonstrated to be of mitochondrial origin, the complex contains proteins involved in the process of DNA replication. When used as a sole source of enzymes and DNA template, the membrane-DNA complex was capable of synthesizing mitochondrial DNA in vitro. The incorporation of (3H)dTTP was dependent on the presence of endogenous mitochondrial DNA template, magnesium ions and required four deoxyribonucleoside triphosphates for its maximal activity. Actinonycin and ethidium bromide were found to have an inhibitory effect on the incorporation of the label. It was demonstrated that the in vitro DNA synthesis by the membrane-DNA complex involved formation of known replicative intermediates (lariats and dimers) characteristic of the replication of mitochondrial DNA in vivo. It was concluded, therefore, that the membrane-DNA complex represents a mitochondrial replication complex. The complex was further characterized in terms of protein content by SDS polyacrylamide gel electrophoresis. The electrophoretic comparison of complexes isolated from normally grown cells and cells exposed to drugs affecting mitochondrial DNA replication in Paramecium showed some protein variation. Preliminary attempts were made to remove the proteins not involved in the process of DNA replication by salt extraction. Electron microscope studies of the replication complex revealed a novel chromatin-like structure of the mitochondrial genome. Biochemical studies of the putative mitochondrial chromatin showed the presence of five basic proteins which behaved in terms of their acid solubility and electrophoretic mobility similarly to histones extracted from nuclear chromatin. Evidence was obtained for protection of mitochondrial DNA from random digestion by nucleases, thus implying the direct association of DNA with basic proteins. The evolutionary implications of the discovery of chromatin-like structure in mitochondria are discussed.

This thesis documents a new approach for the identification of a small biologically active molecule’s site of interaction, through the rapid synthesis of molecular probes. A marked library approach has been developed whereby a biocompatible marker is attached onto the small molecule’s molecular scaffold. This marker plays no role in the screening process itself, but facilitates the formation of a range of molecular probes from active marked library members. As an example of molecular probe formation, site selective biotinylation will be discussed in the introduction. This marked library concept has been applied to the natural product anisomycin A. Investigations focused on development of a detailed structure activity relationship for anisomycin’s activation of the stress activated protein kinase (SAPK) pathway, along with the synthesis of a number of marked library analogues. The active marked library members were then converted to a range of functional molecular probes utilising the copper(I) catalysed Huisgen cycloaddition as the key coupling step. These molecular probes are being used in the elucidation of anisomycin’s biological target for activation of the SAPK pathway. In a further demonstration of this strategy, a focused library of marked steroids has been synthesised based on the functionalisation of dehydroepiandrosterone B. Directed by the results of preliminary biological screening, a number of marked library members have been converted into fluorescent molecular probes. These probes will be used in future applications to probe the biological action of the dehydroepiandrosterone.

Investigation of a locus, at 11p13, predisposing to Wilms' tumour led to the identification of the Wilms' tumour suppressor gene, (<I>WT1</I>). I used a transgenic approach to examine aspects of both <I>WT1</I> expression and regulation, A 5 kb region (USWT1) of the putative <I>WT1</I>promoter was linked to a lacZ reporter and used to generate several transgenic embryos. The expression pattern driven by this reporter construct was similar to that of the endogenous locus but very weak and susceptible to position effects. To overcome this problem a YAC spanning the human <I>WT1</I> locus was used to drive expression of a lacZ reporter in transgenic mouse embryos. The reporter gene expression driven by this construct mirrors that of the endogenous <I>Wt1 </I>locus except in the sex specific Sertoli and granulosa cells. I used the YAC driven reporter gene expression in these transgenic embryos to carefully study the expression pattern of <I>Wt1</I>. The transgene is expressed in both mesodermal and neuro-ectodermal lineages. It is expressed in mesenchymal cells produced by the proliferating coelomic epithelium; hence <I>Wt1</I> is a marker for this specific cell type. It is also expressed in the interdigtal mesenchyme during limb morphogenesis. This newly identified site of <I>Wt1</I> expression was confirmed by <I>Wt1</I> mRNA <I>in situ</I> analysis and raises a potential role for <I>Wt1</I> in the control of apoptosis. Cells expressing the transgene are present in all the differentiating epaxial musculature of the embryo implying that <I>Wt1</I> may play a role in myogenesis. In the neuro-ectodermal lineage, transgene expression implies that <I>Wt1</I> may play a role in neural patterning. The YAC transgene expression studies presented in this thesis complement further YAC transgenic work we have undertaken. Trangenic animals were made with an unmodified YAC spanning the <I>WT1 </I>locus. This transgene was able to partially rescue the <I>Wt1</I> homozygous null phenotype when crossed onto the <I>Wt1 </I>knockout background.

GCN4 is a transcriptional activatory protein found in <I>S. cerevisiae </I>and is a member of the basic region leucine zipper (bZip) family of DNA-binding proteins. The full length protein is 281 residues in length and consists of an activatory region and a DNA-binding region. The DNA-binding region of the protein consists of two separate regions; a leucine zipper through which he protein dimerises, and a basic region which contacts target DNA. Various studies have shown that it is possible to mutate residues in the basic region and create proteins with changed specificity. The aim of this project was to use λZAPII technology to create a large library of mutants and investigate the effects of single and double mutants in the basic region and linker segment of GCN4. The DNA-binding (bZip) region of GCN4 was extracted from a plasmid containing the entire GCN4 gene and cloned into the <I>lacZ</I> gene of the bacteriophage vector λZAPII. The recombinant vector was infected into a strain of <I>E. coli </I>which allowed the bZip gene to be expressed as a fusion protein with β-galactosidase. <I>In vitro</I> screening experiments were carried out and any 'positive' clones which bound to a DNA target identified. The DNA sequence of each positive clone was then determined. These experiments demonstrated that the λZAPII system was an excellent vehicle for the expression and identification of a DNA library and that identical conditions could be used to screen a large library of mutants. A phagemid containing the bZip gene was excised from λZAPII <I>in vivo</I> and infected into <I>E. coli. </I>These cells were used to overexpress the β-galactosidase-bZip fusion protein. The fusion protein was purified and shown to bind with high affinity to a specific DNA target.

The principal aim of this project was to characterise developmentally important genes located close to the <I>brown</I> gene, using both molecular and genetic techniques. Seven mice with radiation induced mutations at the <I>brown</I> locus were received from B. Cattanach. The phenotype of mice homozygous for one of these deletions, b<SUP>57H</SUP>, revealed the presence of a <I>brown associated fitness</I> of <I>baf</I> gene. These mice were much smaller than their littermates and frequently died soon after birth. Those that did survive were subfertile and commonly had litters of only one or two mice. The <I>baf</I> gene was identified concurrently by E.M. Rinchik in compound heterozygote mice with combinations of different deletions. All viable mice with compound deletions exhibited the <I>baf</I> phenotype indicating the <I>baf</I> gene was situated relatively close to <I>brown</I>. A collaboration with E.M. Rinchik enabled hybrid DNA from 25 of the deletions to be obtained, in which the deleted <I>Mus musculus</I> chromosome was heterozygous with a <I>Mus spretus</I> chromosome. The presence or absence of markers in different deletions can be detected by virtue of polymorphisms between the two species. This enabled microsatellite markers and microdissection clones to be mapped to the region and to further classify deletions within a particular complementation group. PCR primers from the microdissection clones which mapped closest to the <I>brown</I> gene, as well as primers from the <I>brown</I> gene itself, were used to screen two YAC libraries. Eighteen YACs were isolated and will be used to generate a YAC contig of the region which contains the <I>baf</I> gene. This work has improved the physical map at this locus and will allow the isolation of a YAC contig and ultimately the <I>baf</I> gene itself.

The expansion of trinucleotide repeat tracts is the cause of nearly twenty genetic disorders. Almost all these diseases are characterised by anticipation, which means an earlier age of onset and an increased severity of the symptoms from one generation to the next. The mechanisms of trinucleotide repeat expansion are not understood. In the course of this project, I have investigated the instability of a trinucleotide repeat array of 43 copies integrated at the <i>attB </i>site of chromosomes of various <i>Escherichia coli </i>mutants. The trinucleotide repeat tract (CTG)₄₃was integrated into the <i>E. coli </i>chromosome in both possible orientations using an intermediate vector and exploiting site-specific recombination between the <i>attB</i> site of the chromosome and the <i>attP</i> site of the vector. Using this method I have constructed 60 mutation strains of <i>E. coli </i>which contain the trinucleotide repeat tract and are deficient in genes involved in replication, recombination, secondary structure repair or mismatch repair. Techniques for the analysis of the instability of the trinucleotide repeat arrays were developed and used to quantify repeat instability. These included: digestion of chromosomal DNA with a rare-cutting restriction endonuclease and PAGE of the labelled fragments; PCR of the trinucleotide repeat tract, followed by restriction enzyme digestion and PAGE; fluorescent PCR and f-TRAMP (fluorescent trinucleotide amplification which uses just one primer in repeated cycles of linear primer extension): products were separated by capillary electrophoresis and analysed using Gene Scan software. Intensive analyses of different <i>E. coli </i>mutant showed that the trinucleotide repeat arrays integrated into the chromosome are stable. Except in one case, no instability was observed in any mutant deficient in replication, recombination, mismatch repair or secondary structure repair. The only strain, which showed instability, was a <i>mutD</i> mutant (impaired in the proof-reading activity of DNA Polymerase III). Possible explanations for this observation are discussed.

In order to identify new mitotic defects a screen for mutants synthetic lethal with a <i>mad1Δ</i> deletion strain was performed in <i>S. cerevisiae. </i>Mad1p is a component of the spindle assembly checkpoint. Four of the mutants isolated in the screen were novel mutations in the essential <i>sfi1</i> gene. In this study, Sfi1p is shown to be a spindle pole body (SPB) protein that contains a conserved set of 211 amino acid repeats. Its phosphorylation is cell cycle regulated, with most phosphorylation being present in an alpha factor (G0/g1) arrest, the time of spindle pole body duplication. The mutants from the screen all had mutations C-terminal to the conserved 21 amino acid repeats. This C terminal region is not conserved outwith the budding yeast, but within the budding yeasts family is the most conserved region. Deletion of the C terminal region is lethal, and results in mis-localisation of the protein. The <i>sfi1-CT</i> mutants are lethal with a range of spindle checkpoint proteins, indicating that the defect is recognised by the spindle assembly checkpoint. At all temperatures the <i>sfi1-CT</i> mutants showed varying levels of large budded cells delayed in mitosis with spindle pole bodies very close together (0.2 – 0.4 μm) and few normal metaphase (1-1.15 μm) spindles. Examination of these abnormal SPB pairs by electron microscopy revealed many large budded cells containing sets of paired spindle pole bodies still attached by a half bridge, as well as some that had partially separated and had some microtubules between them. In all cases, both spindle pole bodies appear morphologically normal and can nucleate nuclear and cytoplasmic microtubules. This is in contrast to other Sfi1p mutants in the conserved repeats that result in a failure of spindle pole body duplication, and suggests that Sfi1p has multiple functions in both SPB duplication and separation.

This thesis discusses the synthesis and structural analysis of molecules pertinent to current problems in the areas of biosynthesis, structure activity relationships and cell biology. Chapter one investigates the structure of an N-terminal blocked derivative of methionine <SUP>5 </SUP>enkephalin, in DMSO solution, using single and two dimensional nmr techniques. Analysis of the data using standard methods indicates the formation of a type I' beta turn centered around the two glycine residues. Chapter two deals with the development and optimisation of a synthesis of D,L tyrosine which is suitable for the incorporation of isotopic labels to give enriched species such as [2-<SUP>13</SUP>C, <SUP>15</SUP>N]tyrosine which is a potentially useful biosynthesis probe. Possible uses of the probe are discussed and three routes are described. The most successful route is based on a modification of the Strecker reaction. Chapter three describes the synthesis, purification and conformational analysis of a fifteen residue peptide corresponding to the signal sequence of the hemagglutinin protein of the influenza virus A/WSN/33, the portion of the polypeptide responsible for its targeting to the cell surface. Circular dichroism studies of this N-terminal peptide in various solvents indicate that the peptide forms a predominantly β-sheet structure in aqueous solution and in alcoholic solutions of low hydrophobicity, while preferentially adopting a helical structure in more lipophilic solvent systems. The relevance of this to targeting to the cell surface is discussed.