γ-aminobutyric acid type-A (GABAA) receptors mediate inhibitory synaptic transmission in the CNS where neurosteroids and protein kinases are their most potent endogenous modulators. Acting individually, these can either enhance or depress receptor function depending on the type of neurosteroid or kinase present, and the subunit combination of the receptor. However, in vivo, these agents probably act in concert to modulate GABAA receptors and precisely ‘fine-tune’ inhibitory neurotransmission; although how this is achieved remains unclear. Therefore, the relationship between these two modulators, at α1β3γ2L and α4β3δ GABAA receptors, expressed in HEK293 cells, was investigated using whole-cell patch clamp electrophysiology. At α1β3γ2L receptors, the potentiation of GABA responses by tetrahydrodeoxycorticosterone (THDOC) was reduced by PKC inhibition and enhanced by PKC activation, implying a role for this kinase in regulating neurosteroid potentiation. By comparison, neurosteroid potentiation was reduced at α1β3S408A,S409Aγ2L receptors, and was unaltered by PKC inhibitors or activators indicating that phosphorylation of the β3 subunit, by PKC, is important for regulating neurosteroid activity. To determine whether ‘extrasynaptic-type’ GABAA receptors are modulated similarly, experiments were also undertaken with α4β3δ and α4β3S408A,S409Aδ receptors. Neurosteroid potentiation was significantly reduced at both receptor subtypes after treatment with the protein kinase inhibitor, staurosporine. Staurosporine was notably less effective at α4β3S408A,S409Aδ receptors, suggesting that, although β3 subunit phosphorylation may play a role in the regulation of neurosteroid potentiation at α4β3δ receptors, it does not fully account for this modulation. Biochemical experiments on α4 subunits identified a new Ser/Thr phosphorylation site (S443). THDOC-mediated potentiation at α4S443Aβ3S408A,S409Aδ receptors was unaffected by protein kinase inhibition, strongly suggesting that phosphorylation of both the α4 and β3 subunits is required for the regulation of neurosteroid activity at α4β3δ receptors. Furthermore, Western blot analysis for wild-type α1β3γ2L receptors, revealed that THDOC increased phosphorylation of β3S408,S409 implying a ‘reverse’ pathway exists for neurosteroids to modulate the phosphorylation state of the GABAA receptor. Overall, these findings provide an important insight into the regulation of GABAA receptors in vivo, and into the mechanisms by which fine-tuning of GABAergic inhibitory transmission may be achieved by two endogenous neuromodulators.

Cytochrome P450 1A2 metabolizes a wide range of therapeutic drugs, including several used to treat diseases common in sub-Saharan Africa. Variation in the gene (CYP1A2) has been reported to be associated with differential efficacy of therapeutic drugs and adverse drug reactions. To gain a better understanding of the extent of variation in the coding and exonflanking non-coding regions of CYP1A2, 762 chromosomes from members of five ethnic groups (Afar, Amhara, Anuak, Maale and Oromo) distributed in a rough north east to south west transect across Ethiopia were re-sequenced. Substantial variation was observed, much of which was novel. As a consequence, a diagnostic test based on previously known variation cannot predict functional variation in Ethiopians. Evidence of purifying selection acting on CYP1A2 was found and coalescent date estimates of CYP1A2 variants were old, with many pre-dating expansions of anatomically modern human out of Africa. Variants within the transcription factor 7-like 2 gene (TCF7L2), which are associated with an increased risk of type 2 diabetes (T2D), were common in multiple Ethiopian populations. TCF7L2 haplotype distribution varied among groups suggesting that T2D susceptibility may also vary, with most groups likely having a West African TCF7L2 risk for the disease and some having more of a European TCF7L2 risk. Many CYP1A2 and TCF7L2 haplotypes can be of important predictive value in the planning and provision of healthcare. These findings are not only of benefit to native Ethiopians, but are also of increasing importance in the planning of healthcare intervention in the developed world, where growing numbers of individuals with recent Ethiopian descent are living. Comparing data with those from publicly available databases it appears that Ethiopian groups display a very high level of diversity that includes most of the common variation observed elsewhere.

Airway mucus protects and maintains the health of the respiratory tract. Its production is orchestrated by environmental cues, thus inter-individual variation in mucus composition, quantity and rheology is likely to confer differences in disease susceptibility and response, and may also result in environmental specific suitability. Glycoproteins known as mucins are considered to be the major components of mucus. This project is concerned with the large secreted airway mucins that are encoded by MUC5AC and MUC5B, with the overall aim being to study their genetic variation in relation to disease and demography. Using a single base extension genotyping method, this project reports for the first time, significant associations between five dependent allergy related respiratory outcomes, including asthma, and a single nucleotide polymorphism of MUC5AC in a European longitudinal cohort. The cause of these associations could not however be refined and therefore further characterisation of the MUC5AC gene is essential for understanding the relationship between allergic airways and MUC5AC. Variants of the MUC5B gene have also been explored in relation to asthma. Variation of the MUC5B upstream promoter region has been characterised in two asthmatic disease case-control cohorts by Sanger sequencing. Statistically significant associations are reported here between regulatory variants of MUC5B, whereby the ‘high’ expressing promoter haplotype is significantly underrepresented in a sample set of severe asthmatic cases as compared to their controls. To further characterise variation within these genes, the MUC5B promoter has also been sequenced in a sample set of eight African populations and the patterns of regulatory diversity have been examined in relation to population differentiation, geographic demarcation and species conservation profiles. We show here for the first time, a statistically significant overrepresentation of the ‘high’ expressing promoter haplotype in a collection of the Anuak peoples of Ethiopia as compared to four other Ethiopian sample sets of differing Ethnicity.

The interplay of fundamental extracellular factors involved in the control of cells producing type I collagen is still not fully understood. Bone formation by osteoblasts is known to be dependent on a good vascular supply and to be inhibited by hypoxia. This strongly suggests that oxygen carried by the blood acts as a key regulator of connective tissue cell function. To investigate further the role played by vascular oxygen supply in connective tissues, calvarial osteoblasts and dermal fibroblasts were grown under hypoxic (2% oxygen) and normoxic conditions (20% oxygen). Hypoxia caused striking decreases in growth and collagen formation by both cell types, along with increased expression of VEGF mRNA by fibroblasts. A novel assay using osteoblasts isolated from neonatal rat long bones was developed to examine more closely the bone formation process. These osteoblasts were characterised alongside autologous calvarial cells. In vivo, limb bones and flat bones form by distinct developmental processes; unexpectedly, osteoblasts from both sources behaved similarly in culture, with no clear differences in growth, gene expression or bone formation. As part of the characterisation of the long-bone-derived osteoblasts, the direct effects of PTH were examined. The mechanism for the anabolic response of bone in vivo to PTH treatment has yet to be fully elucidated. Unexpectedly, PTH exerted powerful inhibitory effects on cell growth, differentiation and bone formation. Moreover, mRNA expression of sclerostin, a key endogenous inhibitor of bone formation, was downregulated by PTH treatment. These findings indicate that the in vivo anabolic response to PTH involves additional factors that override its direct inhibitory action on osteoblasts. Collectively, the results presented in this thesis strengthen the notion that the production of collagen by connective tissue cells is dependent on oxygen and the vascular supply, and implicate the vasculature in mediating the effects of anabolic hormones on bone.

Protein-protein interactions are central to understanding how cells carry out their wide array of functions and metabolic procedures. Conventional studies on specific protein interactions focus either on details of one-to-one binding interfaces, or on large networks that require a priori knowledge of binding strengths. Moreover, specific protein interactions, occurring within a crowded macromolecular environment, which is precisely the case for interactions in a real cell, are often under-investigated. A macromolecular simulation package, called BioSimz, has been developed to perform Langevin dynamics simulations on multiple protein-protein interactions at atomic resolution, aimed at bridging the gaps between structural, kinetic and crowding studies on protein-protein interactions. Simulations on twenty-seven experimentally determined protein-protein interactions, indicated that the use of contact frequency information of proteins forming specific encounters can guide docking algorithms towards the most likely binding regions. Further evidence from eleven benchmarked protein interactions showed that the association rate constant of a complex, kon, can be estimated, with good agreement to experimental values, based on the retention time of its specific encounter. Performing these simulations with ten types of environmental protein crowders, it suggests, from the change of kon, that macromolecular crowding improves the association kinetics of slower-binding proteins, while it damps the association kinetics of fast, electrostatics-driven protein-protein interactions. It is hypothesised, based on evidence from docking, kinetics and crowding, that the dynamics of specific protein-protein encounters is vitally important in determining their association affinity. There are multiple factors by which encounter dynamics, and subsequently the kon, can be influenced, such as anchor residues, long-range forces, and environmental steering via crowders’ electrostatics and/or volume exclusion. The capacity of emulating these conditions on a common platform not only provides a holistic view of interacting dynamics, but also offers the possibility of evaluating and engineering protein-protein interactions from aspects that have never been opened before.

The invasion of erythrocytes by merozoites is driven by an actomyosin motor assembled below the parasite’s plasma membrane, with the myosin anchored on the inner membrane complex (IMC). The myosin (MyoA) is within a protein complex that is comprised of several proteins including myosin tail domain interacting protein (MTIP) and glideosome associated proteins (GAP) 45 and 50. A ternary complex of MyoA, MTIP and GAP45 is formed and later associates with GAP50. GAP45 is acylated by both myristoyl- and palmitoyl-fatty acids and is phosphorylated. This study has highlighted the GAP45 phosphorylation by calcium dependent protein kinase 1 (CDPK1) in vitro and its possible roles in schizogony. By site directed mutagenesis, substitution of S31, S89, S103 and S156 to alanine decreased the level of GAP45 phosphorylation, with S103A exhibiting a major decrease in 32P incorporation. Phosphorylation on S89 and S103 was studied further in parasites as both residues were among the phospho-sites in phosphopeptides identified in vivo. This study also showed that full length GAP45 labelled internally with GFP (FL-GAP45) is assembled into the motor complex, phosphorylated and transported to the developing IMC in early schizogony, where it accumulates during intracellular development until merozoite release. The C-terminal truncated GFP-GAP45 (N-GAP45; residues 1-29) localised at the plasma membrane instead of the IMC and was not assembled into the motor complex. The N-terminal truncated GFP-GAP45 (C-GAP45; residues 30-205) behaved like FL-GAP45. Modifying serine residues, S89 and S103, in GAP45 with alanine or aspartate had no apparent effect on its assembly into the protein complex or its intracellular location during schizont development and merozoite maturation. A second highly phosphorylated component of the complex (GAP40) was also identified. The early assembly of the motor complex suggests that it has functions in addition to its role in erythrocyte invasion.

The phosphatidylinositol transfer proteins (PITPs) are a family of lipid carrier proteins that bind and transfer phosphatidylinositol (PI) and phosphatidylcholine (PC) between membranes. PITPs are commonly involved in phosphoinositide-requiring processes, including phospholipase C and PI 3-kinase signalling, and membrane trafficking. In this study I focus on the uncharacterised soluble PITP, RdgBβ (PITPNC1). The lipid binding and transfer properties of RdgBβ have scarcely been characterised, and the function of RdgBβ is completely unknown. I uncover that RdgBβ interacts with 14-3-3 through its long, disordered C-terminus. RdgBβ is ubiquitinated and subject to rapid degradation in cells, and binding of 14-3-3 via two phosphorylated residues may serve to protect the protein from protease digestion. Whereas RdgBβ binds 14-3-3 under basal conditions, I deduce that, upon stimulation of cells with phorbol ester, RdgBβ binds the Angiotensin II receptor (AT1R)-associated protein, ATRAP, via its N-terminal PITP domain. Others have shown that ATRAP suppresses Angiotensin II signalling by uncoupling AT1R from G proteins and promoting AT1R internalisation. I find that the RdgBβ-ATRAP interaction is blocked by inhibition of protein kinase C or protein synthesis, and may function to re-localise RdgBβ to the membrane in stimulated cells. Unexpectedly, I find that RdgBβ binds PI and phosphatic acid (PA), rather than PI and PC, and that binding of PA is increased by stimulation of cells with GTPγS. Mass spectrometry is used to analyse the molecular species of PI and PA bound by RdgBβ, and reveals that whereas RdgBβ is non-selective in its binding of PI, it selects short-chain monounsaturated or saturated PA species, likely derived from the hydrolysis of PC by phospholipase D.

The mammalian target of rapamycin (mTOR) regulates cell growth and proliferation in response to nutrients and growth factors. The role of mTOR in the biology of embryonic stem cells has not been extensively characterised. The principle aim of this research was to assess the feasibility of manipulating the mTOR pathway to provide bioprocessing improvements for the expansion and differentiation of embryonic stem cells. In order to achieve this aim, the cellular distribution and localisation of mTOR pathway phosphoproteins was assessed and the effect of pathway inhibition on ES cell proliferation and viability was characterised. The key findings of this study revealed that in mouse ES cells, mTOR pathway proteins were phosphorylated during mitosis. Secondly, inhibition of mTOR in growth factor-stimulated pluripotent stem cells impaired proliferation but not viability. Upon withdrawal of growth factors, cells lost their compact shape and exhibited a spread morphology. Under these conditions mTOR inhibition reduced the viability and proliferation of ‘compact’ cells but not ‘spread’ cells. Thus, during the expansion of ES cells the mTOR pathway may be stimulated. During early differentiation, inhibition of the pathway may improve the purity of the final cell preparation and reduce the prevalence of parent stem cells. In conclusion, the experimental findings discussed here show that temporal manipulation of the mTOR pathway in embryonic stem cells may provide bioprocessing advantages resulting in an increased yield of parent cell populations during expansion and increased purity of differentiated cells. The sensitivity of the mTOR pathway to the bioavailability of nutrients may offer a cost-effective route to optimising cell production in an industrial scale process. Further work to elucidate the role of the mTOR complex 2 in ES cell proliferation may provide an additional level of control to independently modulate cell proliferation and growth during expansion and differentiation.

Eukaryotic transcripts require a number of complex cotranscriptional modifications and processing events before translation to protein. Clp1 and Pcf11 are subunits of cleavage factor IA (CFIA), an essential component of the Saccharomyces cerevisiae pre-mRNA 3’-end processing machinery. The crystal structure of a Clp1-Pcf11 complex was determined previously and revealed the binding of ATP to a highly-conserved P-loop motif and a tight Pcf11-Clp1 interaction facilitated by a number of highly-conserved Pcf11 residues. Nonetheless, the biological function of both Clp1-ATP binding and the Pcf11-Clp1 interaction was not well understood. The work in this thesis combines an in vitro and in vivo investigation of the Clp-ATP and Clp-Pcf11 interactions in an effort to understand the function of these factors in transcription and pre-mRNA 3’-end processing. It is demonstrated that the interaction of ATP and Pcf11 with Clp1 are linked events: Loss of Clp1-ATP binding results in the abrogation of the Pcf11-Clp1 interaction and leads to Clp1 instability in vitro, and similarly, mutations that directly uncouple the Pcf11-Clp1 interaction also disrupt Clp1-ATP binding and cause Clp1 instability in vitro. An in vivo mutational analysis in S. cerevisiae revealed that both Clp1-ATP binding and the Pcf11-Clp1 interaction are essential for yeast survival. Further cell and immunoprecipitation studies demonstrated that one essential function of Clp1 is as a chaperone of Pcf11, and RT-qPCR analysis of mRNA from a sample set of yeast genes points to a role for these proteins in transcription and transcription termination rather than in poly(A) site selection.

From folding to function through compartmentalisation - influence of aminoacid changes and roles of pyridoxal-5'-phosphate on alanine:glyoxylate aminotransferase. Primary hyperoxaluria type 1 (PH1), is a rare inherited disease caused by a deficiency in the liver-specific, pyridoxal-5’-phosphate (PLP)-dependent enzyme alanine:glyoxylate aminotransferase (AGT). AGT is normally localized to the peroxisomes where it catalyses the conversion of the intermediary metabolite glyoxylate to glycine. In a subset of patients with PH1, linked with the Gly170Arg mutation, AGT is mistargeted to mitochondria and patients may respond to pharmacological doses of pyridoxine, a precursor of PLP. Using a model system of stably transformed CHO cells expressing different AGT variants, we have studied four mutations – Gly170Arg, Phe152Ile, Ile244Thr, Gly41Arg - on the background of a common polymorphism, Pro11Leu, characteristic of a minor AGXT allele. All four mutations unexpectedly result in the peroxisome-to-mitochondrion mistargeting of AGT. However, significant differences were found on AGT quaternary status and catalytic activity. These results emphasize the synergistic effect of the Pro11Leu polymorphism and disease-causing mutations and its potential to act as a mitochondrial targeting signal. Varying the concentration of pyridoxine in cells had a biphasic effect on AGT expression, activity and compartmentalization that differed between constructs. Where some mutants had improved function with additional pyridoxine, normal AGT was inhibited at higher levels. The results provide an explanation for the therapeutic effect of pyridoxine in some patients. Overall, PLP was shown to have both chaperone and prosthetic group effects on mutant AGTs. Using a specially developed cell-based assay of indirect glycolate toxicity, we have shown the lower ability of mutant AGTs to detoxify glyoxylate. This is the basis for a functional assay of AGT in mammalian cells which could be used to screen drugs targeted at AGT.