Kinesins are motor proteins that have roles in cell division as well as the transport of various organelles and protein complex cargoes within cells. Kinesin-1 is considered the conventional motor protein and consists of three main isoforms, KIF5A, KIF5B and KIF5C. The family of Trafficking Kinesin proteins (TRAKs) bind to the cargo binding domain of kinesin-1 forming a link between the motor protein and their cargoes. The best characterised cargo transported by TRAKs is mitochondria. To understand further the mode of association between kinesin and TRAKs, the aim of this study was to determine the structure of the C-terminal cargo binding domain of kinesin-1. Due to the known binding of TRAK2 with the kinesin-1 family it was hypothesized that TRAKs may stabilize the KIF5A cargo binding domain. Hence, the cDNA encoding the KIF5A cargo binding domain, KIF5A800-1032, and the cDNA encoding the kinesin interacting domain of TRAK2, i.e. TRAK2100-380 were cloned into a bicistronic expression vector to result in epitope-tagged constructs. The expression of both proteins was characterized with respect to yield and solubilisation efficiency. TRAK2100-380 was stabilized by KIF5A800-1032. Affinity chromatography of soluble extracts found that KIF5A800-1032 and TRAK2100-380 did co-purify indicating that they do associate. However, the yield was insufficient for further characterisation. To determine the structure of the cargo binding domain of KIF5A several C-terminal constructs which varied in size and epitope tag location were generated. Bacterial growth and solubilisation conditions were optimized to maximize the yield of the various KIF5A cargo domain recombinant proteins. Each was purified by an appropriate affinity chromatography resin. Conditions were established to optimize the purity, yield, stability and aggregation state. One construct, KIF5A800-951 was isolated to a sufficient yield and did not aggregate. Therefore, KIF5A800-951 is a good candidate for further structural analysis. To refine further the TRAK binding site within the cargo domain of kinesin-1 a series of rationally designed C-terminal truncations of KIF5A and KIF5C were generated and co-expressed with either TRAK1 or TRAK2 in mammalian cells. The binding of kinesin-1 truncations to the TRAKs was determined by co-immunoprecipitation assays followed by quantitative immunoblotting. Deletion of the distal 71 amino acids of KIF5A resulted in an increased co-immunoprecipitation of KIF5A with TRAK2. Three possible TRAK2 binding sites within the KIF5A and KIF5C cargo binding domains were found. Furthermore, differences were found between TRAK1 and TRAK2 in terms of their association sites with KIF5A. Overall these studies yield new insights into kinesin/kinesin adaptor interactions which may impact in the future on a better understanding of neurodegenerative diseases such as hereditary spastic paraplegia and Charcot–Marie–Tooth disease which have both been linked to deficiencies in neuronal transport mechanisms of mitochondria.

Voltage-gated sodium channels (VGSCs) are essential for the initiation and propagation of neuronal impulses. Specifically, these channels initiate the rising phase of action potentials in electrically excitable cells, allowing the conduction of electrical information. There are 9 different sub-types of VGSCs, each found in different tissues, with a minimum 50% identical genetic code.1 Recently, it has been reported that some VGSC sub-types are over-expressed in highly metastatic tumours.2 Their presence allows the up-regulation of some key biochemical processes. In fact, the metastatic potential of the tumours has a positive correlation with the expression of sodium channels.2-6 This thesis focussed on developing a novel positron emission tomography (PET) tracer to target VGSCs. We hypothesised that a PET scan using a VGSC ligand may be able to highlight tumours with high metastatic potential, and assist in the selection of the most appropriate clinical treatment. At present, there is no known clinical PET tracer for imaging VGSCs. Our project involved the adaptation of the 3-(4-substituted-phenoxyphenyl) pyrazoles,7 a known group of VGSC modulators, into novel fluorine-18 labelled PET imaging agents. These were chosen due to their strong affinity for VGSCs and their structural characteristics, which allowed multiple approches to introduce a fluorine-18 label. Several compounds of interest were successfully synthesised. Optimal radiolabelling strategies were identified and methodologies developed, to produce PET agents in good synthesis time, yield and purity. The fluorine-19 standards for these compounds underwent a full VGSC isoform selectivity screen to identify the most potent and selective compounds. The fluorine-18 PET agents were also put through a series of in vitro and in vivo studies including automated and manual electrophysiology, biodistribution, metabolism and autoradiography. Ultimately, we aimed to target specific VGSC sub-types that were highly expressed by specific aggressive tumour types. Following this study, an interesting VGSC PET agent was revealed that warrants further investigation.

The vast majority of protein/peptide drugs are not available orally. Their oral delivery is hindered by intestinal instability and limited permeability. The aim of this project was to gain an understanding of the stability of a large protein, lactase, a small protein, insulin, and a newly discovered peptide throughout the intestinal tract. This was used for the rational design of their oral formulations. Lactase was completely denatured at gastric pH within 10 minutes but was stable with intestinal enzymes. Encapsulation in enteric Eudragit L100 microparticles using a method previously used to encapsulate low molecular weight drugs produced small particles with a high yield and encapsulation efficiency, >90%. They restricted lactase release in acid but did not protect it from denaturation. Porosity and particle morphology investigations using an SEM with a new type of detector revealed surface structures which disappeared upon dispersal in acid and an inner porous structure which may allow acid entry and lactase denaturation. Co-encapsulation of an antacid preserved almost 10% of lactase activity in acid, superior to existing oral lactase supplements. Insulin was not hydrolysed at gastric pH but was immediately and completely digested by gastrointestinal enzymes. To protect it from pepsin insulin was encapsulated in Eudragit L100 microparticles. The particles produced were <100μm with a yield and encapsulation efficiency of >70%. After incubation with pepsin they protected 80% of encapsulated insulin. A small peptide, peptide 1 was gradually degraded in the intestinal fluids. To provide protection and increase its permeability peptide 1 was encapsulated in PLGA nanoparticles but 15% of the encapsulated peptide was immediately released in vitro. Encapsulation of the nanoparticles in Eudragit L100 microparticles successfully prevented any burst release in acid. This should minimise gastric digestion of peptide 1 and concentrate nanoparticle release in the small intestine providing a higher probability of permeation.

That ‘Children are not small adults’ is a commonly quoted adage: nowhere is this more true than in pharmaceutics. When trying to make an “age-appropriate” oral dosage form, a number of patient needs must be met including swallowability, dose-adaptability and acceptability. Acceptability may be enhanced by better tasting, non-gritty medicines: with this in mind this research sought to develop a suspension platform for functionalised multiparticulates, namely for taste-masking. The rheology of the suspending media and its effect on the suspendability of large (>100 μm) placebo particles was investigated before the influence of particle concentration, size and media viscosity of these suspensions on grittiness and acceptability was assessed in two sensory trials containing young adults. It was found that higher concentrations of hydoxypropylmethycellulose were not well tolerated due to their inherent taste and that their acceptability was improved through the addition of flavouring/sweetening agents. Statistical analysis of the results on the refined media and sensory trial showed that particle size and media viscosity had an effect on grittiness, unlike particle concentration. Microparticles of Eudragit® E (a reverse-enteric polymer marketed for taste-masking) containing quinine hydrochloride as a bitter drug were prepared by spray-drying without using organic solvents. Initial experiments resulted in many blockages of the spray dryer which were eventually rectified by increased homogenization and a fractional factorial experimental design employed to screen the influence of different levels of excipients. However, even the optimised process suffered from problems with a low feed solids concentration, low spray rate and low yield. Most particles had an aggregated morphology and the formulations which showed the lowest release in salivary pH were the most aggregated with particle sizes >1 mm. These large particles were not easy to uniformly suspend and would have required a large mass to be administered due to low drug loading which made them unsuitable for use as a uniform platform.

Amorphous solid dispersions are a promising way to improve the dissolution rate and bioavailability of hydrophobic drugs. However, they have a tendency to crystallise to the stable crystalline form during processing or storage, leading to changed physicochemical properties. The study of the stability for amorphous solid dispersions is important. Here, the thermodynamic stability of griseofulvin and indomethacin in polymers were studied using differential scanning calorimetry (DSC), and the effect of moisture on the kinetic stability of indomethacin or griseofulvin solid dispersions with PVP was studied using dynamic vapour sorption – near infrared (DVS - NIR). DSC could estimate the solubility of griseofulvin in HPMCAS accurately but not for the other models. DVS-NIR showed that the water uptake by the solid dispersions of indomethacin and PVP was lower than the values predicted from the sum of the water isotherm for individual components. The reason for this reduction was due to the hydrogen bonds formed between IMC and PVP occupied the water binding sites on PVP. The moisture sorption reduction in the amorphous solid dispersions can stabilise the solid dispersions. A reduced water uptake was also observed for the solid dispersion of 70% GSF and PVP compared to the predicted value. As no hydrogen bonds formed between GSF and PVP, the reason for this reduction was that the water binding sites on PVP were sterically hindered by GSF through the formation of the solid dispersion. A weight loss was absent in the DVS curve for the solid dispersion while crystallisation happened. This was due to the expelled water from amorphous GSF was sorbed by PVP. DVS-NIR offers a great promise in understanding the mechanism of stabilisation of solid dispersions, and therefore may be useful to predict the stability of new API dispersions.

The design of a delivery system that specifically delivers anticancer drug to the tumour site avoiding normal tissues damage has always been a challenge. In this thesis we describe the engineering and biological performance of novel temperature-sensitive liposomes (TSL) that have both a substantial in vivo stability and an efficient content-release by mild hyperthermia (HT). First, we explain the development of novel lipid-peptide hybrids (Lp-Peptide) by anchoring leucine zipper temperature-sensitive peptide within the liposomal lipid bilayer. We characterized this system by studying its physicochemical properties and the interaction of the peptide with the lipid bilayer. Then we examined its potential to retain and trigger the release of the anticancer drug, doxorubicin, in vitro at physiological temperatures and after exposure to mild HT. In addition, the blood kinetics, tumour and other tissues accumulation were explored when we studied the system in vivo. Our data suggested that Lp-Peptide hybrids can increase both immediate and long-term drug accumulation in the tumour. Therefore, we studied their therapeutic activity comparing two different heating protocols to mimic intravascular and interstitial drug release. The last chapter of this thesis explored the opportunities of increasing the therapeutic specificity of TSL by designing anti-MUC-1 targeted vesicles based on the traditional TSL (TTSL) to trigger drug release after specific uptake into cancer cells. The system was evaluated by studying the in vitro cellular binding, uptake and therapeutic efficacy. Taking this system a step further, its biodistribution and therapeutic potential were also examined. Different protocols were applied to explore the effect of HT on the accumulation of targeted TTSL into the tumour and their therapeutic efficacy. In summary, our studies demonstrate the critical factors to consider in the design of clinically relevant TSL and the importance of matching the heating protocol to their physicochemical and pharmacokinetic parameters to maximise therapeutic benefits.

Our work is aimed at developing a synthetic biocompatible gene and siRNA delivery system for the treatment of primary and metastatic tumours. To facilitate delivery of nucleic acid based drugs into the cell, one strategy is to formulate the naked gene with an amine based non-viral gene delivery system via the counterion interaction. The delivery systems including 4arm-PEG-amine, 4arm-PEG-N-2-ethylamine, 8arm-PEG-amine and 8arm-PEG-N-2-ethylamine were synthesised, characterised and complexed with a reporter gene (β-gal plasmid DNA) in phosphate buffer pH 6.0. The resulting complexes were sized and their zeta potential measured (Malvern Zetasizer 3000HS, Malvern Instruments, UK). The complexes were also imaged using transmission electron microscopy and characterised for DNA binding and DNA protection using gel electrophoresis and the ethidium bromide displacement assay. The in vitro transfection efficiency and cell cytotoxicity of the complexes were determined in the A431 and HeLa cells. Additionally, in vivo therapeutic studies in female nude tumour bearing mice were carried out. A promising DNA-polymer complex of 4arm-PEG-N-2-ethylamine produced a complex of 200-300 nm in diameter (polydispersity < 0.6). Complexes had a zeta potential of +19.8 mV (n=3) and were spherical, fibrillar and toroidal in shape. The new gene delivery complex protected DNA from degradation in serum up to 2 hours and was as efficient as poly(ethylenimine) (PEI) in transfecting the A431 cell line, but it was more than 3 orders of magnitude less cytotoxic than PEI. In vivo a gene medicine, comprising the polymer and the tumour necrosis factor alpha gene, was tumouricidal. When complexed with siRNA, the siRNA polymer complex demonstrated a trend of gene silencing activity. A new synthetic gene delivery polymer of 4arm-PEG-N-2-ethylamine has been synthesised. This polymer is biocompatible to cells and is an efficient in vitro and in vivo gene transfer agent.

Methicillin-resistant Staphylococcus aureus (MRSA) is an important human pathogen; it is endemic in hospitals all over the world and can also be found in community settings. The prevention and treatment of MRSA infections has become increasingly difficult due to the ability of these bacteria to develop resistance against all classes of antibiotics. There is an urgent need for alternative ways to treat antibiotic resistant pathogens which do not expose the pathogen to the same selective pressure as antibiotics do. This can be achieved by modification of the bacterial phenotype which will then result in a ‘less fit’ phenotype which may enable the host to successfully remove the pathogen Previous studies have shown that epicatechin gallate (ECg), a component of green tea, is capable of sensitizing MRSA to β-lactam antibiotics. The aim of this study was to develop a zebrafish embryo model for in vivo determination of the effects of ECg on MRSA infections. An increase in zebrafish embryo survival was found after injection of bacteria grown in the presence of ECg. We hypothesised that ECg-grown EMRSA-16 are more susceptible to the innate immune system of the zebrafish embryos in comparison to untreated bacteria. Confocal microscopy revealed modified bacteria within zebrafish granulocytes and four distinct immune cell populations were identified using flow cytometry. An increase in induction of the respiratory burst and a reduction in IL-1β expression determined that bacteria pretreated with ECg are less likely to induce an inflammatory response and are more readily eliminated by the immune system compared to untreated MRSA cells. The results obtained in this study make an important contribution to elucidating the immune-modulatory effects of ECg. They suggest that exposure to ECg reduces the capacity of MRSA to avoid the innate defences of the zebrafish embryo and may therefore enable the modification of the phenotype.

In silico approaches to nucleic acid targeted drug discovery have been used in order to study duplex DNA, in complexes with proteins as well as more unusual form of G-rich DNA folded into higher-order structures termed as G-quadruplexes, in complexes with therapeutic small molecules. The overall aim of this work has been to provide insight into the stability, recognition, energetics of binding and dynamic behavior of these DNAs in complexes with the STAT3βtc homodimer:DNA complex and with therapeutic small molecules in G-quadruplex/pyridostatin and G-quadruplex/fragment complexes by means of combined in silico approaches. The techniques of explicit solvent molecular dynamics (MD) simulations, and subsequent calculations of the free energies of binding, molecular docking, and 3D-pharmacophore modeling have been applied to study STAT3 and G-quadruplex DNA, promising targets for anticancer therapeutic intervention. Analysis of the data obtained from multiple 50-ns MD simulations of the STAT3-DNA complexes has suggested how the transcription factor STAT3 interacts with duplex DNA, the nature of the conformational changes, and ways in which func- tion may be affected. A majority of known pathologic mutations affecting the DNA-biding region of the STAT3 have been found at the protein-DNA interface, and they have been mapped in detail. The STAT3 conformations obtained from these MD simulations have been subsequently used as a basis for a comparative multiple-target molecular docking study with an in-house library of potential STAT3 inhibitors, providing a rational of their binding in the absence of structural data. A novel “dynamic docking” approach (robust platform of numerous MD simulations) has been developed to address the G-quadruplex receptor and ligand flexibility issue, and subsequent conformational change upon binding. The strength of binding at different regions and both sites of the G-quadruplex were then closely examined. An in silico study of a fragment-based approach towards G-quadruplex stabilizing ligands has also been explored, in parallel with experimental studies, to assess whether this could provide a reliable rapid approach to finding hit fragments in the case of the c-MYC promoter quadruplex.

GABAA and glycine receptors are clustered at inhibitory synapses via interactions with the scaffolding protein gephyrin. In turn, gephyrin is translocated to synapses by collybistin, a guanine nucleotide exchange factor (GEF) for Cdc42. Mutations in the human collybistin gene (ARHGEF9) give rise to a range of symptoms including anxiety, epilepsy and intellectual disability. However collybistin knockout mice revealed a selected loss of clustering of distinct GABAA receptor subtypes, whilst glycine receptors clustering remained intact. This suggests that other GEFs might also contribute to gephyrin and inhibitory receptor clustering. This thesis describes the identification and charactertisation of two novel GEFs that were indentified in a yeast two-hybrid screen using a gephyrin bait: IQSEC2 and IQSEC3. Full-length IQSEC2 did not interact with gephyrin in vitro, and is located at excitatory synapses in vivo, so is unlikely to have a role in GABAAR and GlyR clustering. However, four missense mutations in IQSEC2 were found in families with X-linked intellectual disability (XLID) that impair either calmodulin binding to the IQ-like domain (R359C) or ArfGEF activity (R758Q, Q801P and R863W). By contrast, full-length IQSEC3 did interact with gephyrin in vitro and co-localised with gephyrin and inhibitory receptors in vivo, making this ArfGEF a plausible clustering factor. I also show that a gene fusion affecting IQSEC3 may also result in autosomal intellectual disability associated with behavioural defects. Lastly, I examined the interactions between inhibitory GABAAR and GlyR subunits with gephyrin, mapping binding sites for gephyrin on the GABAAR α3 subunit. My analysis revealed that critical determinants of this interaction are located in the motif FNIVGTTYPI in the GABAAR α3 M3-M4 domain and the motif SMDKAFITVL at the N-terminus of the gephyrin E domain. Site-directed mutagenesis of the gephyrin E-domain revealed that GABAAR α3 and GlyR β subunits bind to an overlapping site on the gephyrin E-domain.