In Malaysia, the prevalence of osteoporosis in women age >45 years is approximately 1 in 4 making it a major public health concern. Osteoporosis is usually asymptomatic in its early stages. Consequently, women who may have osteoporosis remain unidentified. This may lead to unwanted fractures. Fractures are associated with a reduction in quality of life. There is a 3-fold increased risk of death within 5 years in those who fracture. It is therefore imperative to encourage prevention and screening programmes which aid in early detection of osteoporosis. Current research suggests that many individuals with fragility fractures do not undergo appropriate screening and do not engage in preventive health behaviours. Pharmacists can work in collaboration with doctors to screen for osteoporosis, to educate patients on their osteoporosis risk, and to empower patients to take osteoporosis preventive measures. It is with this belief that we conducted this study to determine the effectiveness of a pharmacist osteoporosis screening programme in postmenopausal women. This study design was developed based on the United Kingdom Medical Research Council’s Framework of developing and evaluating complex intervention. Hence, this research project was divided into three phases: phase one was to explore the perceptions of the stakeholders for conducting an osteoporosis screening programme, phase two was to develop tools for the osteoporosis screening programme whilst phase three was to conduct the a feasibility study on the osteoporosis screening programme. Phase one aimed to answer three research questions. The first research question was to explore the barriers and facilitators towards conducting an osteoporosis screening programme. Seven main barriers to the implementation of an osteoporosis screening programme were identified: governmental, organizational and management, work environment, team, task, individual and patient factors. The patient factors were targeted for our intervention. The second research question explored the role of the Malaysian pharmacist in osteoporosis screening. Pharmacists were principally perceived by all participants to be suppliers of medication, although there was some recognition of roles in providing medication advice. Nonetheless, doctors, nurses and policy makers were eager for pharmacists to be more proactive via inter-professional collaboration in osteoporosis screening, prevention advice and disease management. The third research question aimed to explore the components for an acceptable, practical and sustainable osteoporosis screening programme. We systematically identified four intervention (environment restructuring, education, persuasion, enablement) components to develop an acceptable, practical and sustainable osteoporosis programme. The “interventional package” consisted of counselling sessions, osteoporosis risk assessment and bone mineral density. In phase two, the Satisfaction Questionnaire for Osteoporosis Prevention (SQOP) and Osteoporosis Prevention and Awareness Tool (OPAAT) were developed and validated. Both the OPAAT and SQOP were found to be valid and reliable to assess patients’ knowledge of osteoporosis and patients’ satisfaction towards the pharmacist screening programme. Additionally, six osteoporosis risk assessment tools were also validated among Malaysian postmenopausal women. Our results identified that the Osteoporosis Screening tool for Asians (OSTA) was the most suitable risk assessment tool as it had a sensitivity of 81.3% and specificity of 41.0% at an empirical cut-off point of ≤0. A pharmacist-led osteoporosis screening intervention package which consisted of the ‘intervention package’ and collaboration between the doctors and pharmacists was developed and finalized. Phase three was a feasibility study of the developed pharmacist-led osteoporosis screening programme. Based on scientific, process, resources and management assessment the programme was found to be feasible in the Malaysian primary care setting. This was a good start for the implementation of a population-based osteoporosis screening programme in Malaysia as there was currently no such programme available. Future research should involve a randomized controlled trial to assess the effectiveness of the programme.

Pulmonary delivery offers an attractive route for delivering chemotherapeutics, with the benefits of high drug concentrations locally and low side effects systemically. However, fast clearance of small molecules in the lungs and the pulmonary toxicity are the main obstacles in this field. In this thesis, we explored the utility of polyethylene glycol-paclitaxel (PEG-PTX) conjugates to achieve sustained drug release in the lungs. Paclitaxel was linked to 6 kDa and 20 kDa PEG and the conjugates showed good stability and cytotoxicity in vitro. PEG-PTX largely increased the maximum tolerated dose of paclitaxel in mice and significantly enhanced its anti-tumor efficacy following intratracheal instillation in a lung carcinoma mouse model. PEG-PTX 20 kDa presented a prolonged residency and a sustained paclitaxel release in the lungs. This study demonstrated that PEGylation offers a potential delivery system for inhaled chemotherapy with improved anti-tumor efficacy and reduced local toxicity.

Human inflammatory/innate immune responses lie at the core of resistance to infectious disease and determine the nature of pathophysiological reactions of hematopoietic cells like leukaemia and allergy. The crucial step in these events is the ability of immune cells to function properly, which depends on their adaptation to stress caused by pro-inflammatory stimulation. The mechanisms underlying this crucial biochemical dogma, and its role in normal and pathological cross-links between immune cells, are not well understood. This PhD programme was devoted to investigation of these important problems. We found that the inflammatory mediator interleukin 1 beta (IL-1β), derived from human innate immune cells, triggers production of the major hematopoietic growth factor, the stem cell factor (SCF) in MCF-7 human epithelial cells. This process is controlled by the hypoxia-inducible factor 1 (HIF-1) transcription complex, which regulates cellular adaptation to inflammatory/hypoxic stress by promoting angiogenesis and glycolytic degradation of the glucose. Translational mechanism, which is majorly dependent on the mammalian target of rapamycin (mTOR) kinase pathway underlies IL-1β-induced HIF-1 accumulation and also contributes to SCF biosynthesis. The effect is applicable in both in vitro and in vivo systems. Further experiments demonstrated the involvement of this biosynthetic mechanism in the differential control of normal and pathological functions of inflammatory cells including monocytes, basophils and mast cells. Our results also demonstrated possible biochemical mechanisms regarding the cross-talk between inflammation and SCF-dependent blood cancer (leukaemia), which remains a serious medical burden worldwide. Finally, the pathways investigated could be further considered as potential therapeutic targets for pharmacological correction of human inflammatory reactions and treating cancer/leukaemia by classic and principally novel approaches, such as utilisation of gold nanoparticles.

Oral administration is the most common route for administration of drugs. With the growing cost of drug discovery, the development of Quantitative Structure-Activity Relationships (QSAR) as computational methods to predict oral absorption is highly desirable for cost effective reasons. The aim of this research was to develop QSAR models that are highly accurate and interpretable for the prediction of oral absorption. In this investigation the problems addressed were datasets with unbalanced class distributions, feature selection and the effects of solubility and permeability towards oral absorption prediction. Firstly, oral absorption models were obtained by overcoming the problem of unbalanced class distributions in datasets using two techniques, under-sampling of compounds belonging to the majority class and the use of different misclassification costs for different types of misclassifications. Using these methods, models with higher accuracy were produced using regression and linear/non-linear classification techniques. Secondly, the use of several pre-processing feature selection methods in tandem with decision tree classification analysis – including misclassification costs – were found to produce models with better interpretability and higher predictive accuracy. These methods were successful to select the most important molecular descriptors and to overcome the problem of unbalanced classes. Thirdly, the roles of solubility and permeability in oral absorption were also investigated. This involved expansion of oral absorption datasets and collection of in vitro and aqueous solubility data. This work found that the inclusion of predicted and experimental solubility in permeability models can improve model accuracy. However, the impact of solubility on oral absorption prediction was not as influential as expected. Finally, predictive models of permeability and solubility were built to predict a provisional Biopharmaceutic Classification System (BCS) class using two multi-label classification techniques, binary relevance and classifier chain. The classifier chain method was shown to have higher predictive accuracy by using predicted solubility as a molecular descriptor for permeability models, and hence better final provisional BCS prediction. Overall, this research has resulted in predictive and interpretable models that could be useful in a drug discovery context.

The aim of the research is to evaluate the possibility of transferring the Efficient Consumer Response (ECR) model developed in the Fast Moving Consumer Goods (FMCG) sector into the pharmaceutical sector and to propose an adapted model for the German market. The German pharmaceutical market is consolidating distribution channels and demand power is shifting towards pharmacies (Hofmann, 2013a). The manufacturers` aim for differentiation requires being closer to patients and pharmacists. Therefore, they increasingly do business directly with pharmacies (Insight Health, 2013). Wholesalers are caught between the strong supply power of manufacturers and increasing demand power of pharmacies (Hofmann, 2013b). Exploratory research was undertaken using the case study method to consider how the ECR model from FMCG can be adapted for the pharmaceutical wholesale business. A single case study was considered as different wholesalers would not participate due to their competitive market and because I am an employee of the case company (Celesio AG). The study was conducted in the German subsidiary (GEHE Pharma). Semi structured interviews with key account managers from FMCG and pharmaceutical manufacturers, Celesio AG management board, GEHE Pharma management and retail pharmacists were conducted. Additional data were generated linked to participative observation during manufacturer meetings between GEHE Pharma and pharmaceutical manufacturers, as well as from secondary and internal documentary material. Findings suggest that several similarities between the FMCG market and the pharmaceutical market exist. No aspect was found which would not allow implementing ECR principles into the pharmaceutical market in Germany. The model is adapted according to the research findings. The adjusted model considers that the pharmaceutical market shows more complexity in terms of the market actors. In this market three main participants exist: pharmaceutical manufacturers, pharmaceutical wholesalers and retail pharmacists. Whereas in the FMCG market the ECR model incorporates the relationship directly between FMCG manufacturers and grocery retailers; no wholesaler is considered in that model. Therefore, the adapted model needs some adjustments for the pharmaceutical wholesale market, which are presented in the research. Furthermore, the research delivers evidence that the ECR model is not static and can be adjusted in terms of the number of participants, content and different dimensions in the relationship between different stakeholders and can, therefore, also be implemented in other industries. exist: pharmaceutical manufacturers, pharmaceutical wholesalers and retail pharmacists. Whereas in the FMCG market the ECR model incorporates the relationship directly between FMCG manufacturers and grocery retailers; no wholesaler is considered in that model. Therefore, the adapted model needs some adjustments for the pharmaceutical wholesale market, which are presented in the research. Furthermore, the research delivers evidence that the ECR model is not static and can be adjusted in terms of the number of participants, content and different dimensions in the relationship between different stakeholders and can, therefore, also be implemented in other industries.

Shape-based approaches have many potential areas for development in the future for application to in silico pharmacology. Further exploration of the role of molecular shape may lead to better understanding of the substrate specificity of enzymes and the possibility to reduce toxic effects that may be caused by ligands binding to undesired target proteins. Methods exploiting molecular shape for activity and toxicity prediction might have a great influence on the drug discovery process. There are different approaches that might be used for this purpose, e.g. shape fingerprints and shape multipoles. Both methods describe the shape of molecules, discarding any chemical information, using numerical values. Focusing only on shape can lead to identifying novel core structures of molecules, with improved properties. Molecular fingerprints are binary bit strings that encode the structure or shape of compounds; shape is measured indirectly by alignment to a database of standard molecular shapes – the reference shapes. The Shape Database should represent a wide range of possible molecular shapes to produce accurate results. Therefore, this was the main focus of the investigation. The shape multipoles method is a fast computational method to describe the shape of molecules by using only numbers and therefore it requires low storage needs and comparison is performed by simple mathematical operations. To describe the shape, it uses only 13 values (3 quadrupole components and 10 octupole components). The performances of both methods in grouping compounds based on shared biological activity were evaluated using several test sets with slightly better results in case of shape fingerprints. However, the shape multipole approach showed potential in finding differences in shape between enantiomers. Among the possible applications of the shape fingerprints method are solubility prediction (on comparable level as well-established methods) and virtual screening.

Hydrogen sulfide (H2S) is the simplest endogenously produced thiol and has an indispensable role in cardiovascular homeostasis. It has been shown that exogenous H2S supplementation protected the heart against myocardial ischaemia/reperfusion injury through a mechanism which is yet to be defined. In this thesis, it was hypothesised that controlled application of thiol/H2S donors at reperfusion would mitigate acute myocardial infarction. We sought to characterise the cardioprotection and molecular targets of three H2S donors (mesna, GYY4137 (a slow-releasing, non-mitochondrial targeted H2S donor) and AP39 (a mitochondria-targeting H2S donor)). This characterisation was conducted using a broad range of experimental models and techniques including anaesthetised rat model of ischaemia/reperfusion injury, Western blotting and mitochondrial studies using isolated cardiomyocyte mitochondria, namely subsarcolemmal and interfibrillar mitochondria. Mesna did not limit infarct size when it was given pre-ischaemia or at reperfusion. GYY4137 and AP39 significantly limited infarct size when given specifically at the time of reperfusion through different mechanisms. Cardioprotection established by GYY4137 was mediated mainly by triggering of PI3K/Akt/GSK-3β signalling at reperfusion with partial dependency on eNOS activity. Selective mitochondrial delivery of H2S at reperfusion using AP39 had no effect on Akt, eNOS, GSK-3β and ERK1/2. In isolated mitochondria, AP39 inhibited Ca2+-sensitive opening of PTP in subsarcolemmal and interfibrillar mitochondria through attenuation of mitochondrial reactive oxygen species generation. The studies presented in this thesis provided novel mechanistic insights into cardioprotection by H2S. These studies suggest that targeted delivery of H2S represents a novel and effective adjunctive therapy to ameliorate the injurious effects of reperfusion which contribute to acute myocardial infarction.

The research in this thesis describes investigations of (i) mixed polymer HPMC and PEO hydrophilic matrices and their performance in low and high ionic environments and (ii) understanding the internal behaviour of HPMC and PEO mixed systems. It was postulated that using a blend of these polymers might provide advantages over the use of single polymers. A series of ‘realistic’ 30% w/w polymer matrix formulations, containing different weight ratios of HPMC and PEO and a soluble model drug (caffeine), were tested in ionically challenging media, up to 1M sodium chloride (NaCl). Dissolution testing showed how HPMC dominated formulations exhibited accelerated release in high ionic strength media (0.8M NaCl or higher), whereas PEO dominated formulations did not. Power law analyses suggested the release mechanism of matrices in 0.6M NaCl and below were anomalous non-Fickian transport, but case II transport was observed in HPMC dominated matrices at 0.8M NaCl and above. A polymer ratio of 4:6 HPMC:PEO allowed an extended release tablet to be formulated that was resistant to 1M NaCl. In 0.6M NaCl or below, increasing the proportion of HPMC in a mixed HPMC:PEO tablets, increased the duration of extended release. Confocal laser scanning microscopy was used to investigate the structure of the HPMC:PEO matrix hydrated gel layer. The results provided evidence that HPMC and PEO particles swell independently in the gel layer. They remained substantially unmixed during gel layer formation, and each appeared to contribute independently to gel layer structure. Magnetic resonance imaging showed how PEO matrices hydrated more rapidly than HPMC matrices, but PEO matrices completely dissolved after 9 hours. In the case of 4:6 HPMC:PEO and HPMC matrices, a hydrated gel remained. This reflected the behaviour of these matrices in the dissolution tests. Unfortunately, MRI could only be applied in zero salt media, as the dielectric properties of NaCl interfered with the results, and other techniques were required to examine matrix behaviour in high salt media. Texture analysis showed that at low NaCl concentrations, the HPMC gel layer exhibited higher gel strength than PEO, and that by substituting HPMC for PEO increased gel layer strength was obtained. The later stages of gel layer morphology were also investigated by digital optical macroscopy. Images showed greater gel longevity of HPMC and mixed matrices, with evidence for a higher gel strength and less erosion than PEO matrices. Swelling of single polymer particles showed how increasing NaCl concentration significantly inhibited HPMC particle swelling but only had a limited effect on PEO particle swelling. The ability of PEO particles to swell in high salt media may explain the resistance of PEO matrices to high NaCl dissolution media. The miscibility of HPMC and PEO in dilute solution was studied by rheology and phase contrast microscopy. Measurements of storage modulus (G’) at 1% w/v showed how most polymer mixtures showed negative deviations from ideal mixing at all oscillatory frequencies studied (0.1Hz, 1Hz, and 10Hz). This is evidence that these polymers are immiscible in solution. Phase contrast microscopy provided direct optical evidence of phase separation in blended HPMC:PEO solutions (4% w/v). The tendency of these polymers to be immiscible, suggests that they may also be phase separated in the more concentrated environment of the gel layer. Gel layer morphology in binary polymer tablets was investigated directly by confocal microscopy (up to 15 min) and by attenuated total reflectance Fourier transform infrared spectroscopy (ATR-FTIR) up to 3 hours. The confocal microscopy images showed that HPMC and PEO appeared to swell independently during early gel layer. Each polymer appeared to contribute independently to gel layer structure. ATR-FTIR imaging allowed chemical mapping of the three components (water, PEO and HPMC) in the gel layer, providing evidence that each polymer formed individual domains. PEO appears to be more extensively swollen than HPMC and may form the outer part of the gel layer, protecting HPMC from the effect of high ionic media. The work in this thesis suggests that mixed polymer HPMC:PEO matrices may have certain advantages over the use of matrices containing only single polymer. PEO confers resistance to highly ionic media, while HPMC provides a longer drug release than PEO alone. Each polymer appears to contribute separately to the gel layer, but the ability of PEO to swell in highly ionic environments, may allow formation of a diffusion barrier that protects the incorporated HPMC from ionic media, and allows it to contribute to gel layer structure.

Polyphosphate biosynthesis and the stringent response play an important role in the virulence of pathogenic bacteria. Our objective is to validate these pathways as an antimicrobial target and to identify inhibitors of the key enzymes polyphosphate kinase (PPK), (p)ppGpp synthetase I (RelA) and (p)ppGpp synthetase/hydrolase II (SpoT). The role of polyphosphate and (p)ppGpp metabolism in Francisella virulence has been explored with deletion mutants. These exhibited defects for intracellular growth in macrophages and were attenuated in mice, indicating a key role for the FtPPK, FtRelA and FtSpoT in the virulence of Francisella. The development of three in vitro activity assays will enable the discovery of PPK inhibitors. Ion-pairing HPLC analysis has been used to measure substrate kinetics, providing evidence that FtPPK belongs to the PPK2 superfamily with little preference between substrates (KM: ADP - 369 μM; GDP - 624 μM) as displayed by other PPK2 enzymes. 31P NMR spectroscopy has been used to monitor the overall time course of the PPK reaction. To facilitate high-throughput screening, a coupled luminescence based activity assay has been developed in a 96-well plate format. These assays can also be applied to the discovery of inhibitors for FtRelA and FtSpoT. Understanding the structural basis of inhibitor action requires a crystal structure of the target enzyme. For FtPPK, a crystallisation screen has identified conditions for obtaining suitable crystals and data has been collected to 2.1 Å resolution. Future studies will use the high throughput assay to identify PPK inhibitors; NMR and HPLC assays to characterise the mode of action and crystal structures of PPK:inhibitor complexes will identify the precise molecular interactions.

The discovery of drugs that can effectively treat disease and alleviate pain is one of the core challenges facing modern medicine. The tools and techniques of machine learning have perhaps the greatest potential to provide a fast and effcient route toward the fabrication of novel and effective drugs. In particular, modern structured kernel methods have been successfully applied to range of problem domains and have been recently adapted for graph structures making them directly applicable to pharmaceutical drug discovery. Specifically graph structures have a natural fit with molecular data, in that a graph consists of a set of nodes that represent atoms that are connected by bonds. In this thesis we use graph kernels that utilize three different graph representations: molecular, topological pharmacophore and reduced graphs. We introduce a set of novel graph kernels which are based on a measure of the number of finite walks within a graph. To calculate this measure we employ a dynamic programming framework which allows us to extend graph kernels so they can deal with non-tottering, softmatching and allows the inclusion of gaps. In addition we review several graph colouring methods and subsequently incorporate colour into our graph kernels models. These kernels are designed for molecule classification in general, although we show how they can be adapted to other areas in drug discovery. We conduct three sets of experiments and discuss how our augmented graph kernels are designed and adapted for these areas. First, we classify molecules based on their activity in comparison to a biological target. Second, we explore the related problem of lead hopping. Here one set of chemicals is used to predict another that is structurally dissimilar. We discuss the problems that arise due to the fact that some patterns are filtered from the dataset. By analyzing lead hopping we are able to go beyond the typical cross-validation approach and construct a dataset that more accurately reflect real-world tasks. Lastly, we explore methods of integrating information from multiple targets. We test our models as a multi-response problem and later introduce a new approach that employs Kernel Canonical Correlation Analysis (KCCA) to predict the best molecules for an unseen target. Overall, we show that graph kernels achieve good results in classification, lead hopping and multiple target experiments.