Type 1 diabetes (DMT1) is a chronic autoimmune disease that leads to apoptosis and death of pancreatic islet beta cells. These beta cells are influenced by and contribute to a network of gut derived hormones that regulate glucose homeostasis. Our research explores the effect that DMT1 has on this network of gut hormones. A series of clinical research studies was established to collect samples from both healthy participants and individuals with DMT1. Methods to measure gut hormones were explored and refined prior to sample analysis. Results from this analysis revealed that the homeostasis of glucagon, active-ghrelin and peptide YY are affected by DMT1 and that glucagon secretion may be controlled by a glucose independent signal derived from the gut. Important lifestyle markers in DMT1 were identified and included a correlation between leptin and body mass index and gastric inhibitory polypeptide and sedentary behaviour. Histopathological analysis of pancreatic slides taken from donors with DMT1 suggest that the islet area stained for glucagon does not correlate with duration of diabetes, but does correlate with age. Together these suggest that type 1 diabetes does affect other gut hormones involved in glucose homeostasis and this can be influenced by lifestyle factors.

RK2 is a low copy number plasmid responsible for spread and maintenance of important properties (including antibiotic resistance and degradation functions) among bacteria. Gene expression in RK2 is controlled by cooperativity among four repressors (i.e. KorA, KorB, KorC and TrbA) to tightly regulate replication, stable inheritance and conjugative transfer functions. KorB (358 aa) has dual roles as a global regulator and as an active partitioning protein. This study focuses on its role as a global regulatory protein and its interaction with DNA, RNAP and other repressor proteins (e.g KorA and TrbA) of RK2. It is shown for the first time that DNA binding by negatively charged protein KorB (-21) is modulated via a balance of charge in the internal region from aa 235 to 245. KorB binds O\(_B\) and silences the genes around, showing that KorB can spread. TrbA and KorA bound to DNA adjacent to KorB do not block gene silencing by KorB and indeed potentiate its repression, suggesting that KorB can spread past DNA binding proteins and thus that they do not act as road blocks. The fact that KorB E237A, which is defective in silencing, cannot repress at a distance when alone but can do so in presence of TrbA, provides strong evidence of looping. The fact that KorA and TrbA do not potentiate gene silencing by E237A, but do potentiate its repression, indicates strongly that gene silencing is because of spreading instead of looping. Full length KorB is required for distal repression. However, only the region 225-255 aa is critical for proximal repression by KorB. The results suggest a model in which KorB organises DNA loosely over a long region through a wrapping in a way that can accommodate other regulatory proteins. This nucleoprotein complex may also be critical for plasmid partitioning.

Transcription is the first step in gene expression. Thus, RNA polymerase copies information from template DNA to generate the mRNA template for protein production. Transcription is divided into three steps; each step provides a platform for regulation of gene expression. This work contributes to understanding novel aspects of regulatory processes. For example, complex regulation of expression of the nucleoid-associated protein cbpA was found to be dependent not only on sigma factor specificity, but also on binding of the transcription factor Fis upstream of the cbpA start codon. Binding of Fis to the cbpA regulatory region prevents the action of a strong σ70-dependent promoter found within the coding region of a neighbouring gene, yccE. This work carefully dissects the sequence specificity and orientation specificity of DNA sequences found upstream of the cbpA coding region that allows for the binding of Fis. Binding of Fis to the cbpA regulatory region is a redundant process, whereby in the absence of Fis another factor binds to this region. Furthermore, the regulation of cbpA was also found to be dependent on the interplay between the strong σ70-dependent promoter and a weak but convergent σ32-dependent promoter. Finally, horizontally acquired DNA, which is AT-rich in nature, has been known to be regulated by the transcription factor, H-NS. However, the field only has limited knowledge on the mechanisms behind this regulation. It was previously thought that AT-rich genes are subject to canonical regulation by H-NS. However, this work demonstrates the phenomenon of “pseudo-regulation”. Here, H-NS silences activity coming from intragenic promoters, rather than the genuine promoter. This phenomenon is likely widespread, and is demonstrated by this study in a number of AT-rich genes.

The structure, occurrence and biological relevance of G-quadruplex DNA structures has been reviewed, along with a review of several notable G-quadruplex binding compounds published in the literature to date. The synthetic route towards two G-quadruplex DNA binders previously developed within the Hannon group has been modified and improved. Electrospray ionisation mass spectrometry studies have been carried out to evaluate nucleotide binding. The in vitro biological activities of these compounds have been validated against the human ovarian carcinoma cell line A2780 via MTT and comet assays, flow cytometry and inductively coupled plasma mass spectrometry. Both compounds and the corresponding metal-free ligand exhibited higher drug efficiencies than cisplatin against A2780 cells. Both compounds display mild genotoxicity and induce G2/M phase cell cycle arrest. The overall cellular uptake and nuclear localisation demonstrated by both complexes exceeds that of cisplatin. A new class of palladium and platinum(II) complexes have been synthesised from methylthio-substituted terpyridine ligands. In addition to assessing their stability in solution via UV-Vis spectroscopy, initial DNA binding studies with both duplex and quadruplex-forming sequences of DNA have been carried out via circular dichroism and gel electrophoresis. The design and synthesis of alternative ligand systems proffering a range of desirable characteristics to aid future ligand and complex development has been investigated.

During meiosis, homologous chromosomes pair and become connected by formation of the synaptonemal complex. Recombination is initiated by DNA double strand breaks (DSBs), formed by the protein SPO11-1. In Arabidopsis, ~10% of DSBs are repaired as crossovers, which are reciprocal exchanges of DNA between homologues. These physical connections ensure the correct segregation of chromosomes and generate genetic diversity. The remainder are processed as non-crossovers, important for pairing and synapsis. How these processes are integrated and controlled remains poorly understood. Telomeres are thought to play a role in pairing of homologues. The role of telomeres was investigated in Arabidopsis by treatment with colchicine, a microtubule-depolymerising drug, known to disrupt telomere clustering and pairing in rye. A mutant deficient for the telomerase reverse transcriptase TERT was studied, in which telomeres were severely shortened and showed reduced fertility. To track the movement of telomeres during meiosis the telomere binding proteins POT1a and POT1b were chosen for antibody production. Telomeres were found to be dispensable for pairing and synapsis. SPO11-1 RNA interference lines with varying reductions in DSBs were analysed, to investigate how reducing DSBs affects pairing, synapsis, and the crossover/non-crossover decision. Chromosomes showed autonomous crossover control. The synaptonemal-complex was shown to be important in preventing non-homologous interactions.

Following the molecular cloning of the PVM genome, the opportunity to the individual genes and proteins of PVM has arisen. This study investigated nucleocapsid (N) gene and the phosphoprotein (P) gene of PVM and attempted to characterise the polypeptide products expressed from the N and P genes both in vitro in PVM-infected cells. The ability of the PVM N and P proteins to interact with other was also investigated. The nucleotide sequence of the PVM P gene was determined to be 903 nucleotides in length and shown to comprise a long open reading frame capable of encoding the 295 amino acid long P protein and also a smaller second ORF with the potential to express a polypeptide 137 amino acids in length. The PVM P protein shows overall amino acid homology of 35.3%, 35.6% and 28.3% to the P proteins of pneumovirus members HRSV, BRSV and TRTV respectively. The PVM P gene contrasts with the P genes of other pneumovirus genus members which do not possess extensive alternative ORFs. Both the N and P genes of PVM were shown to be capable of directing the synthesis of more than one polypeptide product both in vitro and in PVM-infected BSC1 cells. mRNA transcribed from the PVM P gene long ORF directed the in vitro expression of the 39 kDa P protein and four additional polypeptides. By constructing transcription plasmids that contained 5' terminally truncated P gene cDNA insets, these polypeptides were determined to be expressed by translational initiation on internal P gene initiation codons. Western blot analysis determined that in addition to PVM P protein, two of these in vitro expressed P protein species, with molecular weights of 26 kDa and 23 kDa, were expressed in PVM-infected BSC 1 cells and this observation was supported by the results of anti-P protein monoclonal antibody epitope mapping studies. The ability of the PVM P gene to direct the expression of P protein related polypeptides from internal initiation codons is a feature not yet described for any other pneumovirus member. By immunising rats with a synthetic peptide, antiserum specific for the second polypeptide product (P2) was generated. Western blot analysis using this anti-P2 antiserum identified a species thought to represent P2 in PVM-infected BSC 1 cell material. The ability of the PVM P gene to express a polypeptide from an alternative is a feature common to the P genes of most other morbilliviruses and paramyxoviruses. mRNA transcribed from PVM N gene cDNA was able to direct the in vitro translation of the 43 kDa N protein and also a highly abundant polypeptide with a molecular weight of 24 kDa which was shown to be expressed by way of internal initiation on the fifth N gene AUG codon of the N gene sequence. The 24 kDa N protein related polypeptide was expressed in E. coli, purified, and used to immunise a rabbit for the production of anti-24 kDa polypeptide antiserum. Western blot analysis using this antiserum with PVM-infected BSC1 cells detected the 43 kDa N protein, a highly abundant 30 kDa N protein related species, but not the 24 kDa polypeptide. precise identity of the 30 kDa polypeptide was not determined. Possible mechanisms which could account for the expression of the protein products of the N P genes are discussed. By using a protein blotting technique the interaction that occurs between the N P proteins of PVM was investigated. The P protein binding affinities of in vitro expressed truncated N proteins suggested that many regions of the N protein are co- operatively involved in the binding process, although some regions contributed more than others. The N protein of Sendai virus is believed to bind to the Sendai virus P protein in a similar way. It was also determined that both the amino and the carboxyl- terminal regions of the PVM P protein were found to be essential for binding to N protein. This contrasts with the situation determined for Sendai virus in which 344 P protein amino-terminal amino acids were found to be dispensable for binding N protein.

Genes do not act alone, rather they form part of large interacting networks with certain genes regulating the activity of others. The structure of these networks is of great importance as it can produce emergent behaviour, for instance, oscillations in the expression of network genes or robustness to uctuations. While some networks have been studied in detail, most networks underpinning biological processes have not been fully characterised. Elucidating the structure of these networks is of paramount importance to understand these biological processes. With the advent of whole-genome gene expression measurement technology, a number of statistical methods have been put forward to predict the structure of gene networks from the individual gene measurements. This thesis focuses on the development of Bayesian statistical models for the inference of gene regulatory networks using time-series data. Most models used for network inference rely on the assumption that regulation is linear. This assumption is known to be incorrect and when the interactions are highly non-linear can affect the accuracy of the retrieved network. In order to address this problem we developed an inference model that allows for non-linear interactions and benchmarked the model against a linear interaction model. Next we addressed the problem of how to infer a network when replicate measurements are available. To analyse data with replicates we proposed two models that account for measurement error. The models were compared to the standard way of analysing replicate data, that is, calculating the mean/median of the data and treating it as a noise-free time-series. Following the development of the models we implemented GRENITS, an R/Bioconductor package that integrates the models into a single free package. The package is faster than the previous implementations and is also easier to use. Finally GRENITS was used to fit a network to a whole-genome time-series for the bacterium Streptomyces coelicolor. The accuracy of a sub-network of the inferred network was assessed by comparing gene expression dynamics across datasets collected under different experimental conditions.

Stress responses in plants can be defined as a change that affects the homeostasis of pathways, resulting in a phenotype that may or may not be visible to the human eye, affecting the fitness of the plant. Crosstalk is believed to be the shared components of pathways of networks, and is widespread in plants, as shown by examples of crosstalk between transcriptional regulation pathways, and hormone signalling. Crosstalk between stress responses is believed to exist, particularly crosstalk within the responses to biotic stress, and within the responses to abiotic stress. Certain hormone pathways are known to be involved in the crosstalk between the responses to both biotic and abiotic stresses, and can confer immunity or tolerance of Arabidopsis thaliana to these stresses. Transcriptional regulation has also been identified as an important factor in controlling tolerance and resistance to stresses. In this thesis, networks of regulation mediating the response tomultiple stresses are studied. Firstly, co-regulation was predicted for genes differentially expressed in two or more stresses by development of a novel multi-clustering approach, Wigwams Identifies Genes Working Across Multiple Stresses (Wigwams). This approach finds groups of genes whose expression is correlated within stresses, but also identifies a strong statistical link between subsets of stresses. Wigwams identifies the known co-expression of genes encoding enzymes of metabolic and flavonoid biosynthesis pathways, and predicts novels clusters of co-expressed genes. By hypothesising that by being coexpressed could also infer that the genes are co-regulated, promoter motif analysis and modelling provides information for potential upstream regulators. The context-free regulation of groups of co-expressed genes, or potential regulons, was explored using models generated by modelling techniques, in order to generate a quantitative model of transcriptional regulation during the response to B. cinerea, P. syringae pv. tomato DC3000 and senescence. This model was subsequently validated and extended by experimental techniques, using Yeast 1-Hybrid to investigate the protein-DNA interactions, and also microarrays. Analysis of mutants and plants overexpressing a predicted regulator, Rap2.6L, by gene expression analysis identified a number of potential regulon members as downstream targets. Rap2.6L was identified as an indirect regulator of the transcription factor members of three potential regulons co-expressed in the stresses B. cinerea, P. syringae pv. tomato DC3000 and long day senescence, allowing the confirmation of a predicted gene regulatory network operating in multiple stress responses.

Higher organisms can regulate gene expression through changes in epigenetic marks present on the genome. However, how this regulation takes place in organisms with highly repetitive/complex genomes is not well understood. The acquisition of de novo DNA methylation in plants is mediated by siRNAs through the RNAdirected DNA methylation (RdDM) pathway. The targeted deposition of DNA methylation by this pathway allows for the transcriptional silencing of transposable elements and repeat sequences within the genome, as well as regulating gene expression. In addition, it has been hypothesized that mobile siRNAs may be involved in the epigenetic communication between different seed components. Thus the mobility of non-coding RNAs from extra-embryonic tissues could contribute to epigenetic modifications that could be transmitted to the offspring. The aim of my thesis is to characterise the mechanisms involved in epigenetic gene silencing in maize through the use of a novel transgenic reporter. My work has identified components of the RdDM pathway to be involved in maintenance of gene silencing and show that imprinting and paramutation could be recapitulated using synthetic transgenes. In addition, I developed a novel grafting technique to demonstrate that epigenetic gene silencing could be efficiently transmitted between different seed components. Collectively, this work provides an insight into the complex mechanisms that regulate gene expression in the highly repetitive/complex genome of maize.

Understanding the regulatory DNA sequences are becoming increasingly important in understanding the way plants integrate signalling cues mediated through the actions of the transcription factors (TFs). This thesis presents an interdisciplinary investigations into regulatory elements found in the promoter regions of a model organism Arabidopsis thaliana. The intergenic DNA sequences are studied between sets of orthologous genes in A. thaliana and 3 other related species to uncover hundreds of evolutionary conserved noncoding sequences (CNSs). The CNSs are found to be more skewed towards the annotated transcription start sites (TSSs) and enriched in previously identified transcription factors binding motifs. Furthermore, the nucleosomes are predicted to have strong presence in the uncovered CNS than random intergenic sequences alone. Altogether the evidence presented in the thesis points to the functional nature of the CNSs. Then, the promoters of genes thought to be co-regulated together and transcriptionally active during infection with fungal pathogen Botrytis cinerea are experimentally tested for direct protein-DNA interaction using high-throughput Yeast One-Hybrid (Y1H) library screens against the TFs found in A. thaliana. The resulting predictions were further validated using pairwise Y1H screen to suggest potential common regulation by ORA59, PIF7, ESE1, At4g38900 and ERF14, and uncovering a complex gene regulatory network (GRN) associated with the tested genes. The promoter fragments together with the predictions from the Y1H screens were used in the computational analysis to establish transcription factor specific binding motifs. Some of the newly predicted motifs were mutated and tested again for altered binding of the associated TFs. Furthermore, in planta mutations of the TFs predicted to be interacting with the promoters of the genes in the Y1H screens were found to have significant impact on the susceptibility of A. thaliana to infection with B. cinerea, further informing gene regulatory network active in response to biotic stress.