A multi-locus sequence typing scheme was designed for the Corynebacterium diphtheriae group comprising C. diphtheriae, C. ulcerans and C. pseudotuberculosis. Five MLST genes (atpA, dnaE, fusA, odhA and rpoB) were analysed to resolve the inter-species relationships of the three organisms. No alleles were shared between the species and MLSA analysis indicated that they represent distinct populations with no evidence of recombination between the organisms. One hundred and fifty four C. diphtheriae isolates were analysed by MLST using the same five loci and two additional genes (dnaK and leuA). The data derived was in concordance with previous ribotyping studies and identified two clonal complexes associated with diphtheria outbreaks. No correlation between ST and toxin production or clinical presentation was observed. In contrast to the biovars gravis, mitis and intermedius the atypical belfanti strains were distinct by phylogenetic analysis. Although two C. diphtheriae cultures isolated from a horse clustered with the human strains four isolates obtained from two cats shared no alleles with the other isolates examined. MLST analysis of 69 C. ulcerans isolates using the five MLST genes and the virulence determinant phospholipase D (PLD) revealed that veterinary and human isolates were not genetically distinct. As with the C. diphtheriae population recombination was shown to play an important role in the evolution of the organism. The 73 C. pseudotuberculosis isolates were examined with the five MLST loci, pld and two genes within the Fe acquisition operon; fagCand fagD. The nitrate negative and positive strains were distinguished by MLST but shared ancestry was apparent as the same alleles were identified in two to three of the genes. MLST supported previous studies which indicate that C. pseudotuberculosis biovar ovis is a genetically homogenous species. This is the first MLST scheme designed for the C. diphtheriae group and the first to encompass multiple species within the Corynebacterium genus. MLST provides a useful tool for the discrimination of the three species and will enable the detection of genetic exchange events between and within the organisms.

The debate regarding the contribution of adult stem/progenitor cells during normal growth and beta-cell regeneration is far from being resolved. Therefore, we addressed in two distinct situations the origin of new beta-cells. We exploited a Cre/loxP lineage tracing system to efficiently label beta-cells in double transgenic mice (Z/AP; RIPCreERTAM) to address the origin of new beta-cell during the beta-cell mass expansion in response to one and two pregnancies. Similarly, we examined origin of new beta-cell after diabetic recovery in triple transgenic mouse (Z/AP; RIPCreERTAM; pIns-c-MycERTAM). Finally we evaluated the functionality of regenerated beta-cells after diabetic recovery in the single pIns-c-MycERTAM mouse model by microfluorimetry, in collaboration with Dr P. Squires. We showed that the beta-cell functionality in the pIns-c-MycERTAM line was abnormal. Second, we showed that the human placental alkaline phosphatase label (HPAP) in the double and triple transgenic mice was 1) specific to beta-cells, 2) irreversible and heritable and 3) tamoxifen dose-dependant. Third, the analysis of the proportion of beta-cells labelled for HPAP in one pregnancy, showed that the HPAP labelling index of the non-pregnant animals (0.44±0.05) was greater that in the pregnant group (0.33±0.06),(paired two-tailed t-test, P-value 0.021), indicating a dilution of the label in pregnant animal pancreata. Furthermore the combined results of the mean HPAP labelling index in non-pregnant animals (0.44±0.12) and pregnant animals (0.33±0.09) in one and two pregnancies reinforced our results above by indicating that the difference between the two groups was considered extremely significant (paired, two-sided student t-test, P-value 0.0007). Likewise, we showed that two to three months after the tamoxifen pulse, beta-cells do not fully lose differentiation or transdifferentiate into other lineages of either endocrine or exocrine compartment. In conclusion, we demonstrated for the first time that non-beta-cell progenitors contribute significantly to the increase of the beta-cell mass in response to pregnancy in combination with pre-existing beta-cell self-duplication.

The c-myc oncogene is over-expressed or deregulated in many human cancers. c-myc encodes a transcription factor, the oncoprotein c-Myc (Myc), which acts as a master regulator of genes involved in such diverse cellular processes as replication and growth, loss of differentiation, invasion, and angiogenesis. Myc can also act as its own tumour suppressor by promoting cell death in the form of apoptosis. Thus, for putative cancer cells to arise, apoptosis must be blocked. Conditional MycERTAM transgenic mice allow regulated activation of Myc in distinct cell populations (skin suprabasal keratinocytes and pancreatic islet β-cells) and have highlighted contrasting behaviour between these two adult tissues in vivo: proliferation in the skin, and apoptosis in the pancreas. Given the crucial dependence on tissue location in vivo, we still do not know enough about the key divergence in Myc-regulated genes and proteins under conditions favouring opposing outcomes. To address this, we performed high-throughput transcriptome analysis using oligonucleotide microarrays. The in vivo transcriptional response to deregulated Myc was analysed for skin keratinocytes and laser-captured pancreatic islets following a time-course of MycERTAM activation. Due to the multi-factorial nature of the experimental design, novel statistical tools were developed allowing the use of linear models for inference of changes in gene-expression based on multiple experimental variables. Comparison of the transcriptional response between the two tissues identified potential signalling pathways which may promote apoptosis of β-cells or survival of skin keratinocytes: the DNA damage response pathway, and the Insulin-like growth factor 1 (Igf1) signalling pathway respectively. In addition, a marked change in expression was detected in members of the steroid hormone-regulated Kallikrein serine protease family in suprabasal keratinocytes but not for β-cells. These have been found to play an important role in regulating Igf1/Igf1-receptor ligation through proteolysis of the Igf1 binding proteins, are previously categorised markers for several human cancers, and may indicate a tissue-specific regulatory mechanism for determining ultimate Myc function in vivo.

Transcriptional regulation of genes is fundamental to all living organisms. The spatial, temporal and condition-specific expression levels of genes are in part determined by inherited regulatory codes in non-coding regions of the DNA. A large set of methods have been proposed to detect conserved regions of regulatory DNA by means of sequence alignments. However, it has become clear that some regulatory regions do not show statistically significant alignments even in the presence of functional conservation. Therefore, detecting and characterising elusive regulatory codes remains a challenging problem. In this thesis we develop and validate a novel computational alignment free model for detection of functional similarity of regulatory sequences. We show that our model can detect functional links between pairs of sequences that do not align with a significant score. We apply the model to a) detect enhancers within the same genome that are likely to have similar functions and b) to detect functionally conserved enhancer regions in orthologous genomes. Our method finds regulatory codes that are common to groups of similar enhancers and consistent with previous biological knowledge. The inputs for our model are two sequences that we wish to compare in terms of their functional similarity as well as a set of transcription factor motifs. The mathematical framework of our model is built on two main components: In the first model component, each sequence is mapped to a vector of estimated occupancy levels for all motifs. These vectors are representing which motifs at what multiplicity and specificity are present in each sequence. In the second model component, a statistical approach is established where we first estimate a probability distribution of motif occupancy levels for sequences that function similar to the template sequence. We then compute a statistical similarity score to evaluate if the sequences are more similar to each other than to random background sequences. Two applications of this model are presented: First it is applied to a set of experimentally validated non-alignable enhancers from D. melanogaster. We show that: • Our model can detect statistical links between these enhancers, • Weak binding sites can make a strong contribution to sequence similarity, • Our model treats statistically significant presence and absence of motifs symmetrically. Similarity of sequences, therefore, can be based on a combination of the two. We show examples of motifs making contributions to sequence similarity through their absence. • Using our model, we can create a network of similarities among the fly enhancers. Groups of enhancers in this network show common regulatory codes. One of these regulatory codes is strongly supported by existing experimental data. In the second application of our model we predict functional subregions of a known D. melanogaster enhancer. To achieve this, we first show that the model can detect the orthology of this enhancer between 10 Drosophila species. We then demonstrate how this statistical link can be used to predict functional subregions within this enhancer.

Avian pneumovirus (APV) has remained an important pathogen of domestic fowl since its isolation in the 1970s. A reverse genetics system for APV was developed that affords direct manipulation and analysis of the molecular biology, pathogenicity, and tropism of APV. Using a synthetic minigenome system, the M2-1 protein was found to enhance transcription but not be essential for replication and the APV M2-2 protein was shown to inhibit transcription of a reporter gene. The viral cis-acting sequences were mutated to determine their role in transcription. Initially, a series of mutations originating from vaccine candidates were introduced into the gene end sequence of the LUC gene. The levels of LUC reporter protein expression in the mutants was 40-70% of normal, thus demonstrating a mechanism for reduction of virus immunogenicity as the result of a single point mutation. Heterologous rescue of the APV minigenome was carried out using plasmids expressing the RSV, PVM and hMPV proteins and showed that homologous protein: protein interactions were necessary for minigenome transcription. An APV cloned virus rescue system (Naylor et al., 2004) was used to create APV viruses which contained the gene encoding enhanced green florescent protein (eGFP) either within intact APV, or in mutants lacking the SH and G genes or lacking the SH gene alone. It was demonstrated that the SH and G genes are not essential for APV replication in vitro and in vivo and that the APV genome is capable of accepting insertions of foreign material. Expression of eGFP from the recombinant viruses was investigated in vivo in turkeys at 3 and 5 days post infection. eGFP was found in the sinus tissue of the birds infected with the virus containing the full complement of virus genes in addition to that encoding eGFP.

Several genes, essential for cell growth and division in Escherichia coli, have been mapped to the 76 minute region of the chromosome. DNA sequencing of part of this region revealed three cell division genes (ftsY, ftsE, and ftsX) in a putative operon. A fourth gene (orf4) was also identified that was transcribed in the opposite direction to the putative operon. The genes rpof, f am, dnaM and ftsS have also been mapped to this region, but their location, relative to the putative operon, was unknown. In this study the fam and rpoU genes were independently cloned and shown to be allelic. The dnaM gene was also found to be an allele of rpoH, and the gene was found to lie immediately downstream of the putative cell division operon. The restriction map of the area was extended, and the distance between the putative operon and the nearest known gene clockwise on the chromosome map (pit), was determined. The ftsS gene was found to be an allele of ftsX. Two promoters were identified within the putative operon, one proximal to ftsY and the other proximal to ftsE. A combination of S1 and primer extension mapping, of the mRNA transcripts, identified the transcriptional start sites of these two promoters. A polycistronic message was also identified encoding all three cell division genes, suggesting at least some degree of co-ordinated expression. In conclusion, the transcriptional organisation of the 76 minute, essential gene cluster has been determined, and evidence has been presented. that there is some degree of co-ordinated expression of the component genes.

Since the discovery of cell-free fetal DNA in the circulation of pregnant women fetal-specific DNA biomarkers for non-invasive prenatal diagnosis of fetal aneuploidy have been sought. A model system assessing the DNA methylation of placental DNA and adult peripheral leukocyte DNA has been developed previously to represent fetal and maternal plasma DNA. To use DNA methylation to detect specific DNA molecules it is desirable that cellfree plasma DNA maintains the methylation profile of its tissue source. Using the imprinted gene GNAS1, a test has been developed to assess, for the first time the relative abundance of methylated and unmethylated DNA circulating in plasma. Methylated and unmethylated DNA sequences were found in equal abundance. If this finding is applicable to all plasma DNA sequences, we conclude that the steadystate concentration of DNA in methylated and unmethylated form is a reliable indicator of its input into the circulation. We have also investigated whether the abundances of different single copy gene sequences in cell-free plasma DNA are equal. Using real-time PCR, the relative abundances of six unique genomic DNA sequences in plasma were assessed. We find that plasma DNA from different sequences is not present in equal abundance in normal healthy individuals. The relative abundance of sequences tested differed by as much as 12.3 fold. We present a panel of novel candidate epigenetic biomarkers which have been identified using the model system. Of 366 DNA regions tested 3% were found to be differential. Further characterisation of these candidate epigenetic biomarkers has revealed limitations of the model system. In view of these results, future epigenetic biomarker development should be achieved by a direct assessment of first trimester plasma DNA.

Breast cancer is one of the most common cancers in females in the western world, and despite the advances in diagnosis and treatment it is still associated with significant morbidity. Thus, improvements to existing treatment modalities remain a priority. Understanding the molecular mechanisms controlling tumour growth and its modulation will be key to developing new therapies. In recent years it has been shown that posttranslational modifications (PTMs) of histones and p53 are functionally important in the regulation of cellular processes such as proliferation, differentiation and DNA damage repair. Thus, this study assessed the incidence of histone and p53 PTMs in breast tumours, and investigated how small molecule inhibitors of acetyltransferases can manipulate the levels of these PTMs in tumour cells. Our initial study demonstrated that hypoacetylation of H4K16 is associated with higher grade breast tumours (Elsheikh et al., 2009). Therefore, the expression levels of enzymes that are known to modulate H4K16 acetylation in vivo was assessed using immunohistochemical staining of 880 human breast tumour tissue microarrays. This led to the identification of a cluster of biomarkers (hMOF, H4K16ac, H3K9me3 and SUV39H1) which are significantly associated with patient outcome. We also assessed in tumours the incidence of other potential biomarkers including selected p53 PTMs such as p53K373ac and p53 K386ac. These were also found to be associated with favourable patient outcome in Kaplan-Meier survival analysis. Other potential biomarkers were also assessed such as the histone variant H2A.Z and its hyperacetylated form. H2A.Z correlated with Estrogen Receptor status of the tumours, consistent with a report that the gene encoding this histone variant is estrogen-regulated. In summary, this study has revealed that histone and p53 PTMs in breast tumours are potentially useful biomarkers for the classification of tumour type and as prognostic indicators, for use in conjunction with other clinicopathological indicators, and other well established biomarkers such as estrogen receptor and HER2. In a second aspect of the study, we investigated the effects of the acetyltransferase inhibitors curcumin and garcinol on a breast cancer cell model (MCF-7 cells). Garcinol blocked transcription-related PTMs such as H3K18ac, but surprisingly induced hyperacetylation of H4K16. This was found to be correlated with increased TIP60 expression, and correlated with increased incidence of DNA damage and cell cycle arrest. Other changes in cancer -associated PTMs were also observed, including increased H4K20 trimethylation. Garcinol compounds also reduced colony formation by MCF-7 cells and augmented sensitivity to etoposide. In summary, the data shows that histone and p53 PTMs constitute novel biological prognostic markers in breast cancer, and that targeting the enzymes that regulate these events may provide new avenues to drug therapies. This version does not contain the previously published journal article reproduced in the print thesis.

Muscle atrophy is a feature of many chronic diseases and contributes to both morbidity and mortality, emphasising the importance of understanding the molecular pathways involved. Zinc finger protein 216 (ZNF216) is an atrogene, a gene which is up-regulated during and directly mediates skeletal muscle atrophy, and encodes the ubiquitin (Ub) receptor protein ZNF216. Herein it is demonstrated that ZNF216 mRNA levels increase in the extensor digitorum longus (EDL) in a lipopolysaccharide (LPS)-infusion rat model of muscle atrophy, relative to saline control. However, combined administration of low level dexamethasone (Dex) with LPS, although sparing muscles from atrophy, did not blunt ZNF216 expression which parallels previous observations for the atrogenes muscle atrophy F-box protein (MAFbx) and muscle RING-finger 1 (MuRF1). ZNF216 expression levels were further elevated in biceps femoris muscle in rats dosed with the statin drug simvastatin (in which severe muscle damage and atrophy occurs), relative to control rats. The ZNF216 protein’s Ub-binding ability and its reported association with the 26S proteasome indicates it may shuttle proteins targeted for degradation to the proteasome as part of the atrophy programme. We utilised immobilised recombinant ZNF216 protein and its Ub-binding Znf_A20 domain alone to capture Ub-modified proteins from rat skeletal muscle that may represent ZNF216’s substrates. Bound proteins specifically eluted by deubiquitination were identified via liquid chromatography tandem mass spectrometry (LC-MS/MS) and included adenylate kinase 1 (AK1) and actin, both previously proposed as substrates of MuRF1. However, ion scores for all candidates were below the accepted threshold of significance and immunoblotting failed to validate LC-MS/MS data. This approach also revealed an increase in a low molecular weight Ub-positive protein from EDL muscle after 24hrs of LPS infusion. Retrospective analysis revealed this Ub-positive protein was consistently captured in other experiments and confirmed by protein MS and immunoblotting to represent an unmodified and unanchored (i.e. not attached to a substrate) K48-linked Ub dimer. Subsequent capture of the Ub dimer using the Znf_UBP domain of isopeptidase T (isoT), a Ub-binding domain selective for the free C-terminus of Ub, confirmed the dimer was unanchored and also revealed a ladder of longer endogenous unanchored poly-Ub chains. Optimised affinity capture conditions has afforded the first opportunity to purify longer free poly-Ub chains and perform the initial molecular analyses of endogenous unanchored poly-Ub purified from in vivo sources.

Echinococcus multilocularis has recently been developed as a model for basic and applied studies on trematodes and cestodes. Deep understanding of molecular biology of this parasite is urgently required for efficient and extensive applications such as functional gene probing and anti-helminth drug screening. In our project, we aimed to unveil E. multilocularis miRNA repertoires and the identification of differentially expressed molecules from the mRNA and miRNA transcriptomes using next-generation sequencing technology. Furthermore, one family of E. multilocularis molecules identified, fatty acid-binding proteins that are involved in lipid metabolisms, was phylogenetically and functionally characterized. Our data presented genome-widely developmental expression profiles of E. multilocularis miRNAs. Sixty-five potential miRNAs were predicted with nineteen of them being novel. The majority of differentially expressed miRNAs such as emu-miR-1, 16 and 71 were found to be significantly down-regulated in the primary neoblast-rich cells derived from the vesicles. Comparative analysis revealed that the miR-71/2 cluster was conserved in the platyhelminths. In addition, seven homologues to miRNAs likely linked to planarian neoblasts were present in E. multilocularis with the intact seed regions. Although their expression in E. multilocularis neoblasts remains to be fully certified, these miRNAs are plausible biomarkers for the germline stem cells. For mRNA profiling, we firstly originated an enzymatic approach for the elimination of mitochondrial transcripts to increase the depth of mRNA transcriptomes. In comparison with 2.7% in the control, the frequency of the sequences mapped to the mitochondria was dramatically reduced to 0.04~0.1% in the treated samples without significantly adverse effects on the targets. A number of differentially expressed genes were inferred from E. multilocularis mRNA transcriptomes and FABP-encoding genes were further phylogenetically and functionally characterised. E. multilocularis encoded at least four distinct FABPs and comparative analysis revealed that the current gene set of FABPs may have emerged before speciation of E. multilocularis and E. granulosus. The apparent divergence between Echinococcus and vertebrate FABPs in genomic organizations prompted us to further explore phylogenetically FABP genes across thirty-five invertebrates. The results demonstrated that FABP genes were organized in diverse ways, with a predominant pattern that is commonly present in vertebrates. Moreover, both gene duplication and alternative splicing might be most likely responsible for variety of invertebrate FABP functions. Four E. multilocularis FABPs were developmentally regulated and FABP3 was highly expressed in the vesicles and secreted or released into the hydatid fluid. These proteins appeared to be able to weakly interact with cis-parinaric acid but not a fluorescent fatty acid DAUDA and retinol. In vitro analysis indicated that FABP3 was capable of the induction of cytokine secretion by bone marrow-derived macrophages and dendritic cells probably via Toll-like receptor 2. Additionally, FABP3 was also demonstrated to activate macrophage small reactive molecule production together with a ligand for TLR 2.