The KDM5 family consists of four proteins, namely KDM5A, KDM5B, KDM5C and KDM5D. The KDM5 family of demethylase proteins counteract the tri-methylation of H3K4 and can function as transcriptional repressors. They can form a number of different multi-protein complexes that contain other chromatin-modifying factors and co-factors that aid in the recruitment of the protein complex to the genome, whilst also possessing the ability to bind to DNA directly through the ARID domain. Furthermore the KDM5 family has also been implicated in mediating nuclear receptor signalling. The KDM5 family are highly homologous and where co-expressed may act redundantly. A picture is emerging where the KDM5 family may dictate the shift in balance between cellular proliferation and differentiation and are therefore important during development. Although the KDM5 demethylase activity has attracted the greatest attention, it is apparent that this family is an important multi-faceted group of proteins. To study the function of KDM5B two knockout murine lines were generated. The Kdm5b null murine line resulted in early embryonic lethality, whereas a second murine line expressing a KDM5B protein lacking the ARID domain (!ARID) was viable and fertile but displayed a mammary gland phenotype, where terminal end bud development and side branching were delayed at puberty. Since intact oestrogen receptor (ER alpha) signaling is a prerequisite for nulliparous mammary gland development we examined the expression of the progesterone receptor in the adult nulliparous KDM5B !ARID murine mammary gland and found levels to be reduced as compared to the wild-type. Furthermore, co-immunoprecipition of tagged human KDM5B and ER alpha proteins demonstrated an in vivo interaction. These studies suggest that KDM5B is required during embryonic development and may be involved in the ER alpha signaling pathway during nulliparous mammary gland development. Our analysis of the C57BL/6J-Kdm5b !ARID pregnant mammary gland showed a delay in mammary gland morphogenesis and the female could nurse her pups at parturition. Investigation into the molecular mechanisms that may relate to this showed that KDM5B expression is required to positively regulate the PRLR/JAK2/STAT5 signalling pathway associated with mammary gland development during gestation.

Members of the phosphoinositide-specific phospholipase C (PI-PLC) enzyme family play a fundamental role in cell signalling pathways by regulating cytosolic calcium and/or the activity of several protein kinases. This thesis reports the identification, molecular cloning and characterisation of a potential seventh sub-class of the PI-PLC enzyme family, the phospholipase C X-domain containing proteins (PLCXDs), which contain only an X domain in their structure. Comparative sequence analysis has identified at least three PLCXD isoforms in the human and mouse genomes (PLCXDs 1, 2 and 3), and at least four isoforms in the European eel (PLCXDs 1-4). Key amino acid residues responsible for the catalytic properties of PI-PLCs were found to be conserved in human, mouse and eel PLCXDs 1, 2 and 3, but were absent in the sequence of eel PLCXD4. PLCXD isoforms displayed unique tissue-specific expression profiles and some similarities between species. Interestingly, in mouse PLCXD1-3 mRNA were found to be predominantly expressed in the brain, however this is yet to be confirmed in humans. Analysis of in situ hybridisation data in mice revealed each PLCXD to be localised in neurons within different brain regions, highly suggestive of unique roles in brain function. Furthermore, the levels of PLCXD3 protein were reduced by more than 99% in cerebella samples from a mouse model of neurodegeneration (Harlequin mouse) compared to control mice. Human PLCXD1, 2 and 3 were found to increase phosphoinositide turnover when overexpressed in the HeLa cell line, and recombinant PLCXD3, purified to homogeneity from E. coli, was found to interact with various phosphoinositides including PI(4,5)P₂. ³¹P-NMR analysis of PI(4,5)P₂ and PI before and after the addition of PLCXD3 purified from HeLa cells and E. coli revealed no difference in the ³¹P spectra whereas expected chemical shifts were seen following the addition of purified bacterial PI-PLC. Significant formation of inclusion bodies was noted when human PLCXDs 1, 2 and 3 were expressed as recombinant proteins in E. coli. Different strategies aimed at optimising the expression of recombinant PLCXD1, 2 and 3, including the use of different fusion proteins and screening expression in E. coli, mammalian and insect cells had limited success, with the best soluble expression only seen with PLCXD3 in insect cells. Attempts to scale-up the purification of PLCXD3 from insect cells to provide sufficient protein for enzyme assays and crystal screens were unsuccessful. The results presented herein suggest that these novel proteins possess distinct and as yet uncharacterised tissue-specific roles in cell physiology.

Apoptin, the VP3 protein derived from chicken anaemia virus (CAV), induces tumour cell-specific apoptosis and therefore represents a potential anti-cancer therapeutic agent of the future. In human tumour cells, but not in normal cells, Apoptin is phosphorylated and subsequently translocates to the nucleus, which are essential important steps for its cytotoxic activity. Recently, the β isozyme of protein kinase C (PKCβ) was identified as a kinase phosphorylating Apoptin in multiple myeloma cells. However, the exact mechanism and nature of interaction between PKCβ and Apoptin as well as its importance for Apoptin-induced cell death remain to be characterised. This project aimed to further investigate the physical and functional interaction between PKC isoforms, in particular PKCβ, and Apoptin. Recently, the first human Gyrovirus (HGyV) was isolated which shows significant structural and organisational resemblance to CAV and encodes a homologue of CAV-Apoptin. Using a synthetic HGyV-Apoptin construct the subcellular distribution and apoptotic function of this novel Apoptin protein in several human cancer cell lines was analysed and compared to CAV-Apoptin. The results demonstrated a comparable tumour-specific nuclear translocation and cytotoxic effect between the two proteins. A model of colon carcinoma and normal mucosa cell lines was used to investigate the functional link between PKC and Apoptin in other cancer types. Immunoprecipitation and proximity ligation assay studies confirmed the binding of both CAV- and HGyV-Apoptin to PKCβI in HCT116 colorectal cancer cells and identified the N-terminal region of Apoptin to be important for the interaction with PKC. In contrast to HCT116 cells two normal colon mucosa cell lines tested expressed low levels of PKCβI. This differential expression pattern of PKCβI correlated with reduced Apoptin activation in normal cells, as evident by its localisation to the cytoplasm, decreased phosphorylation and lack of cytotoxic activity. Further studies revealed that PKCβI knockdown in HCT116 cells resulted in reduced Apoptin phosphorylation but did not impair Apoptin cytotoxicity. Additionally, overexpression of PKCβI was not sufficient to activate Apoptin in normal cell lines, indicating that other kinases or processes may contribute to the tumour-specific function of Apoptin. Using the FRET-based PKC activity reporter CKAR and fluorescence lifetime imaging microscopy (FLIM) the effect of Apoptin or other stimuli on PKC activity was analysed. Similar to treatment with the phorbol ester TPA, Apoptin expression in HCT116 cancer cells resulted in a significant increase in PKC activity, correlating with the expression and increased phosphorylation levels of Apoptin. In contrast, normal colon mucosa cell lines with low PKCβI expression showed a delayed and reduced induction of PKC activation in response to Apoptin. Overexpression and knockdown studies in combination with FLIM provided some evidence that Apoptin predominantly activates the PKCβI isoform in HCT116 cells. Another tumour-selective cytotoxic agent, TRAIL, was tested on a panel of head and neck cancer cell lines associated with human papillomavirus (HPV) infection. Addition of the proteasome inhibitor Bortezomib sensitised TRAIL-resistant HPV positive head and neck cancer cells to TRAIL. This seemed to involve activation of caspases, the anti-apoptotic protein XIAP as well as stabilisation of functional p53, but the precise mechanism has so far not yet been established. In conclusion, the results of this study propose an important link between Apoptin and PKCβI in cancer cells, involving their interaction and co-localisation, Apoptin-induced activation of PKC and PKCβI-mediated phosphorylation of Apoptin to promote its nuclear translocation and cytotoxic function. In addition, the novel HGyV-Apoptin was shown to function in a similar manner to CAV-Apoptin, providing a further tumourspecific protein for future studies.

An important target protein for the AR-AF1 domain is Transcription Factor IIF (TFIIF). At initiation of transcription, TFIIF recruits additional basal transcription factors, stabilises the transcriptional complex and increases elongation efficiency. Using chromatin immunoprecipitation, it was observed that the AR occupied distinct regions of the prostate-specificantigen enhancer but did not migrate with the elongating transcriptional complex. The major subunit of TFAAF, RAP74, has previously been shown to interact with AR-AF1 by our laboratory and it was observed that AR-AF1 can interact with both terminal regions of RAP74. Now, by selectively disrupting helices that structure the globular RAP74 C-terminal domain it appears that AR-AF1 binds to a groove within this region and specific hydrophobic amino acids are important in this generation. The kinetics of RAP74/AR-AF1 interactions have not been determined using surface plasmon resonance. Interestingly, AR-AF1 interacts differently with N- and C-terminal regions of RAP74 and the overall affinities are in the nanomolar range. The structural properties of AR-AF1 were examined using both fluorescence spectroscopy and gel filtration chromatography. It was found that the transactivation domain is structurally flexible and exists in a conformation that is not random coil or globular suggesting that it may be a molten globule. AR-AF1 interacted weakly with 8-anilinonaphthalene-1-sulfonic acid, a hydrophobic probe used to characterise the molten globule folding state. Gel filtration chromatography indicates that AR-AF1 is ∼65 kDa and has a hydrodynamic radius of ∼36 Co much larger than predictions suggest. Surprisingly, by plotting these properties on "folding curves", AR-AF1 is positioned alongside molten globules.

Copper-containing proteins are involved in a wide range of biological processes mainly via oxidation and reduction reactions. The oxidation state of these proteins can be monitored via Förster resonance energy transfer (FRET) between a covalently attached fluorescent dye and the protein’s redox active centre. Consequently, changes in absorbance upon reduction or oxidation of the protein can be related to changes in the fluorescence intensity and lifetime. This FRET-based approach has been applied to study the catalytic activity of the copper-containing blue nitrite reductase (bNiR) from Alcaligenes xylosoxidans at the single molecule level by means of scanning confocal microscopy combined with fluorescence lifetime imaging (FLIM). bNiR catalyzes the reduction of nitrite to nitric oxide during denitrification. The active centre of bNiR consists of a type 1 (T1) Cu site, which acts as the initial port of entry for electrons, and a type 2 (T2) Cu site, where nitrite reduction occurs. Detailed analysis of single molecules of immobilized, fluorescently labeled bNiR has allowed two populations of molecules to be identified that turn over with different catalytic rates. Previous studies of the catalytic mechanism of copper-containing NiRs distinguished two possible reaction pathways. The single molecule results imply these occur as a consequence of heterogeneity in the enzyme population. Fluorescent labeling of laccases, which catalyze the oxidation of a range of substrates coupled to the four electron reduction of O, with fluorescent dyes was 2 investigated for their potential use in the development of a FRET-based biosensor. A novel expression system for the Trametes versicolor laccase Lcc1 in Schizophyllum commune was developed. The recombinant protein is similar to another native laccase (Laccase A) from T. versicolor and both exhibit significantly higher catalytic efficiency with phenolic compounds than the bacterial small laccase (SLAC) from Streptomyces coelicolor. Regardless, preliminary data indicate involvement of a tyrosyl radical in the catalytic activity of fungal laccases, similar to what is observed in SLAC. bNiR and SLAC are trimers, with each monomer consisting of two cupredoxin- like domains. The structure of the catalytic site and the location of a T1 Cu site are different in these two enzymes. A crystal structure provides detailed insight into why an attempt to introduce the SLAC active site into bNiR was unsuccessful. Attempts to introduce T1 Cu sites into the cupredoxin-like domains of bNiR and SLAC, which normally lack this site, resulted in the introduction of a tetragonal thiolate-containing T2 Cu site.

Metabolomics is a technology which allows us to probe a wide array of interactions between metabolites. These interactions can be revealed by statistical correlations between metabolite levels that may arise via a range of mechanisms. To measure metabolite levels, two main techniques are used: Liquid Chromatography Mass Spectrometry (LC-MS) and Nuclear Magnetic Resonance (NMR). For the measurement of correlation structure high analytical reproducibility of the assays is required. While NMR has previously been shown to be reproducible, LC-MS, has not been similarly assessed. To assess the reproducibility of LC-MS for urinary metabolomics, a multi-laboratory study was devised. We find that the technology is highly reproducible, both within and between laboratories with CVs of < 17%, < 5s drift and under 10% ppm between labs. In LC-MS, ionisation of a single compound can lead to multiple charged species such as isotopologues, adducts etc. These multiple signals have a high mutual correlation and we show that this allows them to be identified with high sensitivity and specificity. The inferred statistical interactions between different metabolites can also be affected by analytical errors. An algorithm was designed to remove statistical metabolite links that could have been caused by the analytical technique. Using this method, a higher confidence can be placed on the remaining interactions, suggesting that they are potential biological interactions. Finally, most biological interactions are dynamic in nature, leading to correlations through time between metabolite levels. To explore these dynamic links, two temporal approaches were developed. These methods are designed to discover temporal correlations between metabolites and to test whether they vary between bio- logical conditions. We successfully demonstrate the methods in both LC-MS and NMR datasets. Overall, this thesis shows that correlation structure in metabolic profiling data is reliable, can be successfully filtered to improve quality and can be interrogated to reveal a new kind of dynamic metabolic biomarker.

Matrix metalloproteinases (MMPs) have been identified as biomarkers for cancer, offering prognostic potential. However, they lack non-invasive detection. This thesis highlights the work carried out on the design, synthesis and development of 19F MRI probes for the molecular imaging of MMP activity. The design of these MMP-activated imaging agents was based on the paramagnetic effects of gadolinium(III) on the NMR/MRI signal of a CF3 group in close proximity to the metal cation. 1st generation probes were prepared, incorporating MMP-2 selective peptide substrates SPAYYTA or SLAYYTA as a linker between gadolinium(III)-DOTA and Asp[(trifluoromethoxy)benzylamide]-OH. Attempts to monitor MMP-2 mediated peptide hydrolysis via HPLC and fluorescent assays proved problematic. The use of 68Ga and its complexation to DOTA provided a highly sensitive assay for the assessment of the peptide compounds. Using this method, assay conditions were optimised to observe on average 20% peptide cleavage over 2 hours with MMP-2, with selectivity observed against MMP-9 and MMP-14. The original gadolinium analogues of these compounds were then analysed by 19F NMR spectroscopy, successfully demonstrating a change in signal upon incubation with MMP-2. Development of the probes was then carried out to improve their clinical potential. Two pathways were explored; increasing compound reactivity through the introduction of spacer groups and increasing 19F sensitivity via increased local concentrations of fluorine, carried out utilising activated cell penetrating peptides (ACPPs). The latter method proved more successful, displaying similar reactivity and selectivity to 1st generation compounds. These ACPP probes are still in the early stages of development, with scope for future investigation. The transferability of 1st generation MMP-2-activated probes was also investigated via the production of a compound of analogous design, incorporating an MMP-14 selective substrate RIGFLR. Testing and development revealed this compound required incorporation of a two amino acid spacer group (SG) in order to demonstrate partial MMP-14 selectivity.

To improve detection of the misuse of erythropoietin (EPO) for performance enhancing purposes, this PhD examined ways to improve recovery and preanalysis concentration of EPO from urine. It also looked at ways to enhance the signal in liquid chromatography tandem mass spectrometry of the acidic glycopeptides from digested EPO, and at distinguishing between recombinant EPO and human urinary EPO based on differences in their glycosylation. Due to a shortage of supply of available analytical standards, model glycoproteins were frequently used in place of endogenous EPO. Immunoextraction with magnetic beads effectively recovered EPO from urine which had been filtered to remove large proteins, but was unsuccessful from unfiltered urine, suggesting more research into the right choice of antibody was needed. The specific and reversible binding of boronic acids to cis-diol groups found in the glycan groups of glycoproteins was investigated as a device for the selective binding of EPO. Attempts were made to functionalise mesoporous silica for use as a column packing material. Although there was evidence that at least one method of functionalisation was successful, the use of this silica to extract glycoproteins and glycopeptides was not. Signal enhancement through the introduction of ‘superchargers’ into LC solvents was investigated. This was effective with small molecules, and also improved detection of sialylated glycopeptides. The results do not fit entirely with current models of how superchargers exert their effect, suggesting they are incomplete. Finally, the cleavage, digestion and derivatisation of N-glycans to identify bisected and non-bisected structures as a way to discriminate between rEPO and huEPO was examined. Samples were analysed using LC-MS and CE-LIF, and although much of the work was carried out using a model glycoprotein, there is some evidence that the approach may be capable of discriminating between artificial and endogenous EPO at the levels found in anti-doping samples.

Burkholderia pseudomallei is a Gram-negative bacterium that is endemic in tropical and sub-tropical regions. B. pseudomallei is an opportunistic pathogen and is the causative agent of the human disease, melioidosis, which accounts for 20% of all septicaemic deaths in Thailand. The phosphorylation via serine/threonine protein kinases (STKs) coordinates a vast array of signal pathways in both eukaryotic and prokaryotic organisms. In bacteria these pathways include: biofilm formation, cell wall biosynthesis, metabolism, sporulation, stress response and virulence. In this study, I identified six genes in the B. pseudomallei K96243 genome, bpsl0220, bpsl0571, bpsl0597, bpsl1828, bpss1584 and bpss2102 encoding putative STKs. The importance of these enzymes in B. pseudomallei was not determined to date, and thus the aim of this study is to Understand the roles of these proteins in B. pseudomallei virulence and regulations of Cellular processes. Four of the genes, bpsl0220, bpsl0597, bpsl1828, and, bpss2102, were successfully cloned and over-expressed the putative STKs in Escherichia coli and the recombinant proteins were purified. Each of the recombinant proteins exhibited autophosphorylation as well as phosphorylation of a general kinase substrate in vitro, thus demonstrating that each of the four proteins has protein kinase activity. Furthermore, I constructed specific mutants in each of the four STK-encoding genes in B. pseudomallei. These mutants were assessed in a variety of in vivo and in vitro, assays. The inactivation of the genes in B. pseudomallei, identified a switch in colony morphotype may be associated with a change in lipopolysaccharide structure and the motility of the bacteria, which warrants further study. One of the mutants, Δbpsl1828 displayed a conditional temperature sensitivity in distilled water, and thus the protein encoded by this gene could potentially contribute to the ability of B. pseudomallei to persist in the tropical and sub-tropical environments. However, future work is required to confirm a role of BPSL1828 in the regulation of environmental stress. A second mutant, Δbpsl0220, was avirulent in the Galleria mellonella model of disease, and introducing the full-length gene into the mutant complemented the phenotype. This result suggests that BPSL0220 may play a role in virulence. However, future work should assess the Δbpsl0220 deletion mutant in the mouse model of melioidosis. If these results confirm the role of BPSL0220 in pathogenicity, subsequent work could investigate if BPSL0220 maybe used as a target for therapeutic treatment of B. pseudomallei.

Histone deacetylases 1 and 2 (HDAC1/2) are highly similar proteins (83% identical) that form the core catalytic components of corepressor complexes that modulate gene expression. Germline deletion of Hdac1 in mice results in early embryonic lethality and conditional deletion of Hdac1 but not Hdac2 causes precocious differentiation in ES cells. Therefore to further investigate the role of HDAC1/2 during the early embryogenesis, we have generated a compound conditional knockout ES cell line Hdac1ko; Hdac2Het in which HDAC1/2 activity is reduced but not entirley lost. Hdac1ko; Hdac2He cells have a significant reduction in total deacetylase activity and disruption of corepressor complex integrity. The prolifration capicity of Hdac1ko; Hdac2He cells is not inhibited, however, upon differentiation they were predisposed to toward the cardiomyocyte lineage. In most cell types, deletion of both Hdac1 and Hdac2 is required to produce a phenotype, suggesting their activity is redundant. To circumvent this functional redundancy, we generated a double conditional knockout (DKO) cells in which both Hdac1 and Hdac2 can be inactivated simultaneously. Loss of HDAC1/2 results in a 60% reduction in total HDAC activity and a loss of cell viability, which is associated with increased abnormal mitotic spindle, chromatin bridges and miconuclei, suggesting that HDAC1/2 are necessary for accurate chromosome segregation. Transcriptome analysis reveals 1,708 differentially expressed genes in DKO cells including a reduction in the expression of the ES cells core pluripotent factors. HDAC1/2 activity can be regulated in vitro through the binding of inositol tetraphosphate (IP4). By rescuing the viability of DKO cells using wt and mutant forms of HDAC1, we demonstrated that mutations that abolish IP4 binding reduce the activity of HDAC1/2 in vivo. We have also shown that treatment of DKO ES cells with RA results in reduces induction of HOX genes, suggesting a positive role of HDAC1/2 in gene activation as well as gene repression.