Studies of spider silks indicate that they may outperform virtually all synthetic fibres in terms of strength, elasticity and toughness. To date, most silks studied come from only a select few species and likely underrepresent the immense diversity of the clades. Here, protein and mRNA sequence analyses were used to study silk from two types of spider. The first approach used ESI tandem mass spectrometry to sequence peptide fragments of a silk from a tarantula (Mygalomorphae, Theraphosidae), a hitherto neglected family. The results confirm that the common silk types found in araneomorph spiders, Spidroin 1 and Spidroin 2, are also found in mygalomorphs. A putative N-terminal domain that bears a striking similarity to the N-terminus of araneomorph pyriform silk was isolated. If correctly identified, this would be the first ever recorded N-terminal domain for a mygalomorph. The second approach taken was to construct a cDNA library from theraphosid silk glands and adjacent tissue. Sequencing identified a significant number of uniquely truncated rRNAs. These may be the result of specific 'fragile sites' within these transcripts, which would explain the discrete classes of length polymorphisms found. The cDNA library sequences also provided evidence consistent with RNA editing and furthermore identified the presence of both transcribed nuclear pseudogenes and transposable elements. These may reflect past evolutionary horizontal gene transfer events within the spider genome. Similar analysis of next generation sequencing data from the transcriptomes of three Stegodyphus spp. (Araneomorphae) reveal a range of apparent silk types with similarity to major ampullate, minor ampullate and pyriform silks. These were identified by searching for comparative sequence homologies using Microsoft Office Word. No flagelliform silk or recognisable sticky silks were identified, which is consistent with the biology of Stegodyphus species. In addition to studies of silk, previous common conceptions of dimensional morphologies were examined to see if they could adequately sex theraphosid spiders, including the species that was the subject of the silk study already described. An independent samples t-test was conducted to compare morphologies of particular leg hairs and statistical analysis demonstrated that there were significant differences between males and females (t (70) = 9.445, p < .001). This technique may be important in future evolutionary and ecological studies of theraphosids.

An early event in embryo development is the formation of mesoderm, endoderm and ectoderm, known as the primary germ layers. The gene regulatory network (GRN) consisting of the regulatory mechanisms underlying the formation of mesoderm and endoderm (the mesendoderm GRN) has been extensively studied both experimentally and using mathematical models. The Xenopus GRN is complex, with much of this complexity due to large numbers of Mix and Nodal genes. Mice and humans have only single Mix and Nodal genes, meaning that the Xenopus GRN is overly complex compared with higher vertebrates. Urodele amphibians, for example the axolotl, have single Mix and Nodal genes required for mesoderm and endoderm formation giving a model organism for the study of a simplified mesendoderm GRN. We study the axolotl mesendoderm GRN by developing mathematical models that encompass the time evolution of transcription factors in a cell. A detailed investigation reveals that, despite differences in the axolotl mesendoderm GRN compared with the Xenopus, the model can qualitatively reproduce experimental observations. We obtain experimental data to estimate model parameters using a computational algorithm, then test the behaviour of the resulting mathematical model using independent data. We extend mathematical models of the Xenopus mesendoderm GRN to include transcription factors involved in patterning the DV axis. An investigation of this model shows that it can account for the formation of mesoderm, endoderm and anterior mesendoderm forming in regions of the embryo consistent wth experimental data. In the final section of this thesis, we extend a multicellular model of the Xenopus mesendoderm GRN into a grid of cells.

Combined congenic breeding strategy and microarray expression profiling studies from our group previously identified rat glutathione s-transferase mu type-1 (rGstm1) as a functional and positional candidate gene implicated in hypertension. In the previous studies, expression of rGstm1 was lower in the kidney from the hypertensive rat model, stroke-prone spontaneously hypertensive rat (SHRSP0, compared to the normotensive Wistar Kyoto (WKY) and a chromosome 2 congenic strain (SP. WKYG1a2c*). This project aims to characterise Gstm gene family in rat and human tissues. Results from this thesis demonstrate reduced mRNA expression of several members of rGstm gene family, in vascular and renal tissues. Most notably, expression of vascular and renal rGstm1 mRNA and protein, are lower in the SHRSP compared to WKY and SP.WKYG1a2c*. Vascular mRNA expression of rGstm2, rGstm5 and rGstm7 were reduced in the SHRSP males compared to the WKY males and SP.WKYG1a2c* males but not in the females. The expression profile of the human GSTM (hGSTM) gene family in vascular tissue from varicose vein ad saphenous vein have also been characterised in an attempt to identify the true human orthologue of rGstm1. There are five known members in the hGSTM gene family, all of which were expressed in vascular tissues. Results from characterisation of the vascular hGSTM gene family show that four hGSTM members are homologous to rGstm1. These four hGstm genes remain as potential orthologue for rGstm1. Functional roles of vascular and renal rGstm genes have been investigated using two pharmacological intervention studies. Olmesartan (AT1R antagonist) or hydralazine plus hydrochlorothiazide (direct vasodilator and diuretic) have been used to evaluate the oxidative stress status and gene expression of rGstm genes, in the reversal and prevention studies. Both drug treatments improved the BP of SHRSP rats in reversal study and prevented the rise in BP in the prevention study. Olmesartan-treated SHRSP rats demonstrated reduced superoxide (O2-) and hydrogen peroxide (H2O2) production in both reversal and prevention studies. Minor effects by both drug treatments were observed for the rGstm gene family in the prevention but not the reversal study. Both drug treatments did not influence vascular of renal rGstm1 expression in either reversal or prevention studies. Improvement in BP did not improve rGstm1 gene expression. The rGstm1 was not responsive to pharmacological interventions due to strain-dependent genetic abnormalities. Functional polymorphisms in two key enzymes involved in ROS and NO balance were investigated for association with CAD and vascular compliance as single polymorphism and as haplotypes. There was an association between CYBA A-930G with CAD, with the A allele being recessive. There was also an association between NOS3 G894T with CAD, only when G allele was dominant. The T-786C of NOS3 was associated with small artery compliance index (C2), in both CAD and control groups.

The extent of sequence specificity of the Tn3 resolvase catalytic domain was investigated by creating libraries of Tn3 site I variants in which all of the central 16 bp were systematically randomised in overlapping 4 bp blocks and recombination deficient and recombination proficient site I variants were selected using two different independent selection strategies employing an activated Tn3 resolvase mutant NM. A degree of flexibility in the sequences permitted in the central 16 bp of the Tn3 site I was observed especially at the positions 4, 7 and 8, but accumulating changes was found to be in general detrimental to recombination. The data was compared to the naturally occurring site I sequences associated with proteins from the Tn3 resolvase family, and integrated with the available structural information revealing a number of residues in the extended arm region that could account for the sequence selectivity observed. The sequence selectivity of the activated Tn3 resolvase NM catalytic domain was tested in the Z-resolvase context employing a similar but less exhaustive selection strategy using a purified Z-resolvase Z-R(NM). Z-resolvases with sequence selectivity that is different to that of Z-R(NM) were constructed using catalytic domains of activated mutants of Sin and Tn21 resolvases and their in vivo and in vitro properties were tested, highlighting the universality of the Z-resolvase approach and its potential for the future applications. A number of issues concerning the Z-resolvase design such as the optimum length of Z-sites, what is the effect of the Zif268 DNA-binding domain on catalytic activity i.e. is it activating or inhibiting, is symmetry a prerequisite in the design of Z-sites or can a Z-resolvase catalyse recombination on sites with an odd number of bases between Zif268 binding sites i.e. one half-site longer than the other, what is the relative influence of the Z-resolvase linker length, and can Z-resolvase be complemented by resolvase and catalyse recombination on appropriately designed hybrid sites were explored. The sequence selectivity of catalytic domains of Sin and Tn21 resolvases was compared using a combination of a mutant library selection strategy and the Sin-Tn21 resolvase hybrid experiments. An attempt to change the sequence selectivity of Tn3 resolvase catalytic domain into that of Sin resolvase, both in the resolvase and Z-resolvase context by mutating the specific residues, implicated in catalytic domain sequence selectivity was performed. The sequence selectivity of activated Tn3 resolvase catalytic domain was successfully changed into that of Sin resolvase

Loss of adaptive variation arising from population declines and fragmentation is a primary concern in conservation. However, many conservation programmes assess only neutral genetic variation. Whilst assessments of neutral variation are informative about demographic history, inbreeding and genetic structure, they do not provide information on adaptive variation. The Major Histocompatibility Complex (MHC) is a group of genes that has been extensively studied and are known to be important in effective immune responses. Given the threat posed by infectious diseases to wildlife, the MHC is increasingly being assessed in endangered species. African wild dogs (Lycaon pictus, hereafter wild dog) are an endangered canid that has suffered extensive declines in the wild and now persist as small and fragmented populations totalling less than 8,000 individuals. The purpose of this study was to assess how neutral and MHC marker data genetic data can be used to assist conservation of this species. As such, I assessed sequence diversity across ~300bp of mitochondrial DNA, patterns of polymorphism and heterozygosity at 10 neutral microsatellite loci, compared to sequence variation and haplotype diversity at the MHC. Wild dogs were found to be genetically depauperate at the MHC compared to other canids. Patterns of variation indicate a historical loss of variation, followed by more recent diversification. However, it was also shown that evolutionary history contributes to differences in diversity between species. The spatial and temporal structure of MHC diversity was found to be largely correlated with neutral markers, which may suggest that selection is unable to counter strong genetic drift in such small populations. Overall, genetic diversity of both neutral and MHC markers appeared to be largely determined by demographic stability and size of populations. Habitat fragmentation and loss were associated with genetic isolation of wild dog populations, which showed strong structuring. However, the barriers to, or corridors for, dispersal of wild dogs were not always clear. The European captive breeding population was found to have comparable diversity metrics to wild populations, and was found to contain a large proportion of the MHC variation from the Southern African populations from which they were originally sourced. Careful genetic management is now required to correct the severe over- and underrepresentation of some founder lineages in this captive population to reduce inbreeding and loss of genetic variation.

Green fluorescent protein (gfp) and bacterial luminescence (lux) reporter genes have been used to construct a variety of reporter plasmids for Gram positive bacteria with the aim of using these for bacterial localisation and gene expression studies. The native gfp and luxCDABE genes were cloned into a shuttle vector and the resulting plasmids used to transform Listeria monocytogenes. However, the bacterial populations were found to be weakly fluorescent or luminescent compared to E. coli harbouring the same plasmids. When L. monocytogenes expressing gfp were examined by fluorescence microscopy, only a small proportion of the population was seen to fluoresce. This phenomenon was observed regardless of the gfp variant used in the cloning procedure. However, when gfp3 was placed downstream of PxylA, slightly more individual fluorescent cells were observed compared to when gfp3 was expressed from Pxyn, but the majority of the population was still non-fluorescent. Northern blot analysis and subsequent analysis by SDS PAGE and immunoblotting lead to the supposition that translation of gfp was limiting in L. monocytogenes. A variety of factors could potentially lead to poor translation of the protein, for example poor codon usage, the presence of a ribosome stall site, or poor initiation of translation by the ribosomes. These were all investigated in tum to determine why translation of gfp3 was limiting. Modification of the translational initiation region of gfp3, resulted in a homogeneously fluorescent L. monocytogenes population when the modified gene was expressed from PxylA. Individual lux genes, luxA, luxC and luxE were also translationally enhanced in a similar way to gfp3, and reorganised into an operon where the luciferase genes were adjacent to, but separate from the aldehyde genes. This engineered luxABCDE operon was also expressed from PxylA and highly luminescent populations of L. monocytogenes and Staphylococcus aureus obtained. Having optimised translation for expression III Gram positive bacteria, these reporters were used to construct a variety of reporter plasmids that were successfully employed to observe the intracellular invasion and to monitor agr expression in S. aureus.

The genetics of eight small red squirrel (Sciurus vulgaris L. ) populations in northern Belgium is investigated by analysing variation in a section of the mitochondrial control region and five microsatellite loci. The full sequence of the mitochondrial control region in red squirrels is determined and is compared to that of other mammals. The isolation of microsatellite loci is also described. The eight fragment populations are compared with two large Belgian populations and one large population in the Bavarian Forest, Germany. Virtually no variation is found in the control region within any of the Belgian squirrels, although the German population is found to be highly variable. However, the Belgian and German samples show comparable levels of diversity at the microsatellite loci. The lack of variation in the control region of the Belgian squirrels suggests that they have lost variation, due to either selective or demographic pressures. A combination of a bottleneck and metapopulation structuring could lead to reduced diversity levels and explain the observed patterns of variation. The recent effects of habitat fragmentation and population expansion can be seen in the microsatellite data. Three of the fragment populations show evidence of recent bottlenecks or founder events, probably due to the recent colonisation of these areas by squirrels from nearby expanding populations. Estimations of FsT and RsT show that there is some differentiation among the populations, but none of the populations are significantly differentiated from any of the others. There is no correlation between genetic differentiation and geographic distance indicating that migration is influenced by other factors as well as distance. The fragment populations all contain more allelic diversity than would be expected in populations of their size at mutation-drift equilibrium. Migration between the populations appears to be maintaining nuclear variation and counteracting the effects of random genetic drift.

Initially, this thesis investigates patterns of intragenomic codon usage within the genome of the Delta Proteobacterium Bdellovibrio bacteriovorus. Correspondence analyses revealed the primary factor influencing codon usage within this genome to be related to translational selection. The relationship between the degree of codon usage adaptation (as given by the ‘frequency of optimal codons’ statistic) and putative gene expression level was used to look for genes with unusually high or low expression levels in B. bacteriovorus, in comparison to Escherichia coli, in order to gain further insight into the unusual lifestyle of this Delta Proteobacterium. The scope was then broadened to explore intergenomic patterns of codon usage and initially extend a study measuring the strength of selected codon usage bias across bacterial genomes (Sharp et al. 2005). A dataset of 160 fully sequenced bacterial genomes was used and the strength of selected codon usage bias was seen to vary greatly between species. A correlation was observed between (log of) generation time and the strength of selected codon usage bias with fast growing bacteria showing a higher degree of selected codon usage bias than slow growing bacteria. In bacterial species exhibiting significant levels of selected codon usage bias optimal codon choice was examined. It was observed that optimal codon choice is not always conserved across all bacterial genomes under selection but broad trends in optimal codon choice were seen to be associated with particular bacterial clades. In general, optimal codon choice was seen to be linked with differences in mutational biases among the clades, as seen by a correlation between optimal codon choice in particular clades and the G+C content of their genomes. Clades that were A+U rich (Firmicutes, Gamma Proteobacteria main clade) were seen to largely prefer codons of the form NNA/U whilst G+C rich clades (Alpha Proteobacteria, Actinobacteria and Xanthomonas species) showed preference for codons of the form NNG/C in their highly expressed genes. Finally the relationship between optimal codon usage and tRNA abundances was explored. Changes in tRNA abundances were seen to coincide with switches in optimal codon usage. Therefore, switches in codon usage and tRNA abundance are thought to be influenced by changing mutational bias within the genome as reflected by the correlation between optimal codon choice, tRNA gene complements and genomic G+C content.

RNA polymerase III (Pol III) is responsible for transcribing a relatively small but vital set of genes, including 5S rRNA and tRNAs. Pol III transcription has been shown to be upregulated in transformed and cancer cells, suggesting an important role in cell growth and proliferation. Its tight regulation is, therefore, fundamental for cell welfare, and a number of factors have been shown to be implicated in its control. These include the tumour suppressors p53 and Rb, as well as p107 and p130, and the basal transcription factor Dr1. The work in this thesis focused on the role of these repressors in regulating Pol III transcription in human cells.

To induce pluripotency in differentiated cells, it is necessary to remodel the epigenetic marks on the regulatory regions of pluripotency genes to enable their expression. The induction of Nanog expression is crucial for establishing pluripotency. However, the epigenetic mechanisms associated with the reprogramming of Nanog expression are not fully understood. In mammalian chromatin, epigenetic control of gene expression includes DNA methylation and histone modifications. In undifferentiated cells, regulatory regions of the pluripotency genes Nanog and Oct4 are demethylated and enriched with activating histone marks while being relatively depleted of repressive marks. One route to investigate mechanisms of cellular reprogramming is through treatment of cells with oocyte extracts from amphibians, such as Xenopus and axolotl. Previously, our lab demonstrated that a major difference between these extracts is that the axolotl oocyte can reprogram expression of the mammalian Nanog, while Xenopus oocyte cannot. In this study, I used extracts from oocytes of axolotl (AOE) and Xenopus (XOE), and focused on the mechanisms underlying the reversal of epigenetic marks in regulatory regions of the mouse Nanog gene during its reactivation. I demonstrated that AOE remodels the mouse somatic chromatin by increasing the level of 5hmC on both mouse Nanog enhancer and promoter sequences as well as adding the activating histone marks H3K27ac and H3K4me1 specifically to the Nanog enhancer. XOE was unable to induce these modifications. The expression of Nanog ortholog axNanog and histone variant H2A.Z in axolotl oocytes, but not in Xenopus oocyte, is likely to be one of the reasons for the differences. Indeed, I demonstrated the binding of axNanog and H2A.Z on the mouse Nanog gene in response to AOE, but not XOE treatment. Furthermore, my experiments have elucidated the sequence of chromatin remodelling events during oocyte extract reprogramming that begins with H2A.Z deposition at the Nanog enhancer which allows axNanog binding, followed by epigenetic alterations such as 5hmC, H3K27ac. Taken together, this study refines our understanding of the step-wise events necessary for remodelling of somatic cell chromatin and underlies the difference in the reprogramming capacity of different Amphibian oocytes.