Therapeutic proteins are often potent and have rapid onsets of action. Unfortunately protein-based medicines can be immunogenic and have short half-lives. The circulation half-life of many proteins has been improved by the covalent conjugation of poly(ethylene)glycol (PEG) to the protein. For example, PEGylated interferon-2 (PEGASYS® and PEG-INTRON®) has become a first line treatment for hepatitis C. The aim of this thesis was to examine the possibility of using a homobifunctional PEG reagent to make protein dimers. Our group has developed PEGylation reagents that undergo conjugation by bis-alkylation to selectively conjugate either (i) two cysteine thiols from a native disulfide or (ii) two histidine residues in a polyhistidine tag. It was hypothesised that IFN dimers (IFN-PEG-IFN) could be prepared with higher retained activity than monoPEGylated IFN. It was also hypothesised that a heterodimer of IFN and an antibody fragment (Fab) could be made while retaining the activity of both proteins within the heterodimer. His8IFN-PEG-His8IFN and IFN-PEG-IFN homodimers were prepared by site-selective conjugation to the N-terminal 8-polyhistidine tag and to one of the disulfides of IFN respectively. These homodimer conjugates were characterised in terms of purity and in vitro activity. The in vitro cell based assays were optimised to accurately elucidate the specific activities of the IFN conjugates. The His8IFN-PEG20-His8IFN homodimer was found to retain greater activity than IFN-PEG20-IFN. The increased activity was thought to be due to conjugation to the polyhistidine tag, which is distal from the IFN binding surface. It was also found that IFN-PEG10-IFN homodimer retained greater activity than PEG10-IFN, which could be due to the presence of the second IFN molecule in the IFN-PEG10-IFN homodimer. Two different Fabs were used to prepare the IFN-PEG-Fab heterodimers. Conjugation was conducted at one disulfide of IFN and the accessible interchain disulfide of the Fab. One Fab was derived from a polyclonal antibody to albumin (Fabalb) and the rationale for this heterodimer was that a longer lasting form of IFN could be made (IFN-PEG20-Fabalb). The other Fab was derived from bevacizumab (Fabbeva) to give an IFN-PEG20-Fabbeva heterodimer that could, in principle, display antiangiogenic properties. Both heterodimers were evaluated using antiviral and antiproliferative assays to determine the activity of IFN in the conjugate. The IFN-PEG20-Fabalb conjugate displayed a 10-fold reduction in activity compared to IFN-PEG20-Fabbeva. It was thought that Fabalb underwent competitive binding with components of the media. Interestingly, the IFN-PEG20-Fabbeva heterodimer displayed greater activity than PEG20-IFN and IFN-PEG20-IFN in both the antiviral and antiproliferative assays. The binding properties of Fabbeva were determined by SPR. It was observed that the dissociation rate of IFN-PEG20-Fabbeva was similar to Fabbeva and PEG20-Fabbeva. IFN-PEG20-Fabalb was found to have a similar dissociation rate to Fabalb. However, PEG20-Fabalb was found to have a slower dissociation to both IFN-PEG20-Fabalb and Fabalb, this result requires further investigation but was thought to be due to the sample impurity. The association rates of heterodimers were found to similar to the PEG-Fab conjugates but slower than their native Fabs. This data suggests that the novel IFN-PEG20-Fabbeva and IFN-PEG20-Fabalb heterodimer conjugates have retained their binding affinities to their antigens. Overall, it was shown that a homobifunctional bis-alkylating conjugation reagent (e.g. PEG di(bis)sulfone 4) could be used successfully to prepare dimeric protein conjugates. This work highlighted the importance to ensure the homobifunctional conjugation reagent was pure, especially for the preparation of protein heterodimeric conjugates. To develop this work further, it would be important to investigate three broad areas: i) improving the purity of the starting homobifunctional reagents, ii) evaluate the in vivo efficacy of the resulting protein homo-/hetero-dimers and iii) determine the overall potential for efficient scaling of the process to make the desired protein homo-/hetero-dimers.

The thesis contains <I>in vitro</I> and <I>in vivo</I> radioligand binding studies for the N-methyl-D-aspartate (NMDA) receptor, the 5-HT1A receptor and the 5-HT transporter. A novel NMDA antagonist FR115427 ((+)-1-methyl-1-phenyl-1,2,3,4-tetrahydroisoquinoline hydrochloride; (+)FR) was characterised using [<SUP>3</SUP>H]MK-801 binding to rat brain membranes. As with (+)MK-801, the affinity of (+)FR was increased when 10 μM L-glutamate was added in the assay buffer. (+)FR inhibited [<SUP>3</SUP>H]MK-801 binding to rat cortical synaptosomal membranes in the presence of 10 μM L-glutamate with a Ki value of 40 nM although (+)FR was 12-fold less potent than (+)MK-801. These results indicate that (+)FR is a non-competitive NMDA antagonist. [<SUP>3</SUP>H](+)FR synthesised by tritiation of (+)Cl-FR (mono-chlorinated aromatic precursor of (+)FR) was successfully purified by an open column method. <I>In vivo</I> distribution of intravenous injection of [<SUP>3</SUP>H](+)FR in rats was analysed. Intravenous injection of [<SUP>3</SUP>H](+)FR resulted in a rapid accumulation of radioactivity in brain, and the brain regional distribution of radioactivity appeared to reflect the localisation of NMDA receptors. The possibility of using derivatives of citalopram (selective 5-HT uptake inhibitor) as a photoaffinity or SPET ligand for the 5-HT transporter was evaluated using 5-azido-citalopram (5-AC) (photoaffinity ligand) and 5-iodo-citalopram (5-IC) (SPET ligand). 5-AC had high affinity (Ki = 1.65 nM) for [<SUP>3</SUP>H]citalopram binding sites being only 1.8-fold less potent than citalopram itself. In the presence of 5-AC, repeated U.V. irradiation (15W) of rat cortical membranes produced a significant 20% reduction in the Bmax value for [<SUP>3</SUP>H]citalopram binding site. Therefore radiolabelled 5-AC may provide a tool for isolation and characterisation of the 5-HT transporter.

The enzymology of mitomycin C bioactivation was studied <I>in vitro</I> in two solid mouse adenocarcinomas of the colon MAC 16 (high DTD) and MAC 26 (low DTD). Metabolism of mitomycin C by tumour subcellular fractions revealed a novel mitochondrial reductase active under hypoxia in both tumours. DTD and NADPH:cytochrome P-450 reductase activity was confirmed in MAC 16 but not MAC 26. The role of DTD was examined in a tumour homogenate system and was found to play a protective role under hypoxia in MAC 16, predominating over the other enzymes present. MAC 26 showed enhanced hypoxic metabolism due to the presence of the mitochondrial reductase. The studies were extended to a human perspective, using the human colon xenografts HT-29 (high DTD) and BE (low DTD). <I>In vivo</I> studies, again using the MAC tumours, revealed comparable mitomycin C metabolism and antitumour activity in both tumours. Such a result has profound implications on the use of DTD as a target of enzyme-directed bioreactive drug therapy with mitomycin C. The mitomycin C analogue indoloquinone EO9 is a promising new bioreactive drug, although little is known about its metabolism and mechanism of action. The chemical properties of the reactive intermediates of EO9 were studied under controlled conditions via pulse radiolysis, with the aim of proposing a mechanism for its cytotoxicity. Results indicated that whether EO9 undergoes reduction via one-or two-electron reduction, the hydroquinone intermediate, product of two electron reduction via enzymes such as DTD, will predominate and dictate the pattern of cytotoxicity. This suggests a central role of DTD. <I>In vitro</I> metabolism of EO9 by the tumours described above produced a number of metabolites which proved difficult to identify via liquid-chromatography-mass spectroscopy. Unlike mitomycin C and its principle metabolite 2,7-diaminomitosene, no metabolite was found to correlate with metabolic activation under varying oxic and hypoxic conditions in conjunction with DTD activity.

The overall aim of the work detailed in this thesis was to design and analyse sugar-based (trehalose, raffinose and sucrose) carrier systems for the protection of naked plasmid DNA formulations generated by spray drying. Trehalose exists in multiple crystalline anhydrous and hydrated forms and is thought to exist in two different amorphous forms. Reported thermal transitions of trehalose dihydrate vary with environmental conditions and particle size. The first area of this thesis focused on investigating the inter-conversion properties of trehalose dihydrate to observe if an in-depth understanding of its physical properties will provide an insight into its bio-protective properties. Within this area, two standardised forms of α,α-trehalose dihydrate were generated and characterised by performing a series of thermal, spectroscopic and X-ray diffraction techniques. This resulted in the identification of an intermediary anhydrous form. Within the second area of this thesis the use of the fragility parameter m and the strength parameter D as predictors of amorphous stability of generated co-spray dried sucrose-raffinose and sucrose-trehalose samples was investigated. Results showed addition of both raffinose and trehalose improved predicted amorphous stability, with the greatest effect seen at highest additive concentrations. The third area of the thesis was to evaluate the degree of degradation of plasmid DNA spray dried in the presence of amorphous sugars. Spray-drying can be used to develop biopharmaceutical particles for the pulmonary delivery. However, it runs the risk of loss of biological activity, sample instability as well as thermal degradation of the biopharmaceutical. Results showed that plasmid DNA degradation was reduced when co-spray dried in the presence of raffinose and to a greater extent with trehalose. Co-spray drying of plasmid DNA in the presence of sucrose, sucrose-raffinose and sucrose-trehalose formulations offered less protection than trehalose and raffinose. Overall, two key messages are concluded from the work detailed in this thesis. Firstly, the thermal transitions of trehalose dihydrate can be influenced by environmental factors as well as inter-batch variability. This can affect authenticity of polymorphous and amorphous forms identified. Secondly, addition of raffinose and trehalose to amorphous sucrose formulations improved the predicted amorphous stability of the formulations however; these co-spray dried samples offered less protection compared to plasmid DNA co-spray dried in the presence of trehalose and raffinose alone.

The structure and potential function of two components of the type VI secretion system in Serratia marcescens have been investigated by X-ray crystallography. Chapters 1 and 2 describe the background of the type VI secretion system and experimental methods respectively. In chapter 3, the solution of the structure of S. marcescens Lip is presented. This is an essential outer membrane lipoprotein that is conserved among bacterial species. It is part of a membrane-spanning sub-assembly that is thought to anchor the remaining components to the cell envelope. The crystal structure of SmLip was determined using single-wavelength anomalous dispersion methods targeting iodides that were present in the crystallisation solution. The structure was refined to 1.92 Å and displays a fold similar to that of transthyretin; a protein that binds the hormone thyroxine. Despite the sequence identity between the two proteins being low, SmLip can be considered a new member of the transthyretin-like protein family. A role is suggested for loop 1 being involved in protein-protein interactions with other components of the type VI secretion system and therefore a potential drug target.In chapter 4, the structure of S. marcescens Rap1b is described. This protein is thought to be part of a set of periplasmic resistance proteins that protect S. marcescens from succumbing to the effects of its own secreted toxins. The crystal structure of Rap1b was determined using single-wavelength anomalous dispersion methods targeting both iodides from a halide soak and endogenous sulfurs. The structure was refined to 1.88 Å and displays a helical rich fold not previously characterised. This structure is described and is suggested to represent this family of resistance proteins. In addition to determining the structure of components of the type VI secretion system, the assessment of two drug targets have been investigated by biochemical techniques. The need for new antimicrobials and the importance of target selection in the drug discovery process is described in chapter 5. PBP3 from both Pseudomonas aeruginosa and Burkholderia pseudomallei is the focus of chapter 6. This protein is essential in bacteria as it catalyses the cross-linking of peptidoglycan. Peptidoglycan provides bacteria with structural rigidity and strength to withstand pressures that would otherwise rupture the cell. Sequence identity between the two homologues is approximately 40 %, with the catalytic residues conserved. The production of soluble protein and the results of screening for novel chemical fragments using several techniques are presented. The characterisation of P. aeruginosa IMPDH, an essential protein that is present in nearly all living organisms is described in chapter 7. This protein controls the pool of guanine nucleotides within cells. Fragments were identified using bio-layer interterformetry, a label free optical analytical technique that monitors biomolecular interactions in real time. Fragments were assessed for inhibitory effects using enzyme kinetics. The crystal structure of PaIMPDH was determined by molecular replacement and refined to 2.25 Å. This structure displays the catalytic domain with the active site loop fully ordered and in a conformation rarely observed.

Infectious diseases represent a multitude of threats to populations in both the developed and developing world, from the emergence of drug resistant bacteria and new pathogens to the ancient killers of the neglected topical diseases. Yet a common problem unites all infectious diseases, that is the challenge of how do we cost effectively identify new drugs? The arrival of high-throughput low cost sequencing starkly illustrates the nature of the challenge: the genome sequence of any pathogen can now be determined in a few days yet the availability of complete pathogen genomes has not led to the anticipated wave of new therapies. One reason for this failure might be that previous efforts at selecting the best targets from the genome have not taken into account information on the properties of associated small molecule ligands. To improve the exploitation of genomic information in the discovery of drug targets for new anti-infective agents a modular informatics framework is described that enables the large-scale comparative analysis of pathogen and host genomes. Specifically, new methods to predict essential genes, identify druggable domains and predict selectivity are presented, that have advantages over current approaches. The proposed method to predict essentiality is benchmarked against whole genome essentiality datasets and applied in practice to the analysis of a diverse range species including the bacterial pathogen <em>Pseudomonas aeruginosa</em> and eukaryotic parasites <em>Trypanosoma brucei, Trypanosoma cruzi</em>, <em>Leishmania braziliensis, Leishmania infantum, Leishmania major</em> and <em>Schistosoma mansoni</em>. In order to identify druggable and selective targets a domain-based approach to mining genomes for druggable targets is developed. A domain family based approach enables the determination of "binding site signatures" in the primary amino acid sequences which enables the identification and comparison of specific binding modes for both active/orthosteric site and allosteric site ligands. Information in the binding site signatures is used to train and validate a Bayesian model to predict a compounds selectivity between members of a domain family, whether from within a single genome or from multiple species.

Trypanosomatid parasites are the causative agents of neglected tropical diseases for which current therapies are inadequate. As primitive eukaryotic organisms, they also represent a useful model system to investigate fundamental cellular biology while studies of potential drug targets endeavour to develop new drug molecules. Aspects of both of these areas are explored in this thesis. Microtubules are polymers of tubulin and are essential in eukaryotes for cell division, motility and maintenance of cell morphology. Five tubulin-binding cofactors (TBC, named A-E) are proteins implicated in the folding, polymerisation and processing of tubulin, the major component of the trypanosomatid cytoskeleton. At the initiation of this study, there was no structural information available for any trypanosomatid TBC. We therefore sought to investigate these proteins by X-ray crystallography and assess their potential tubulin-interaction capabilities to support the current functional model. The crystal structure of tubulin-binding cofactor A (TBCA) from Leishmania major is presented, determined using diffraction data to 1.9 Å resolution. Prior to tubulin polymerisation, TBCA forms a complex with ß-tubulin in the pathway to aß-tubulin heterodimerisation. It maintains a soluble pool of ß-tubulin and can prevent premature polymerisation. This is a short helical protein, similar in structure to published homologues. The similarities and some distinct local features that may impact on ß-tubulin binding are discussed. In particular, the surface properties of a prominent bend in the helix bundle represents an area that may be capable of interacting with its tubulin partner. Tubulin-binding cofactor C (TBCC) is implicated in stimulating the hydrolysis of GTP bound to ß-tubulin prior to release of the assembly-competent aß-tubulin heterodimer from a supercomplex between TBCC, TBCD, TBCE and both tubulin subunits. Full-length recombinant Trypanosoma brucei and Leishmania major tubulin-binding TBCC were degraded and crystallisation could not be achieved. However, crystals of a truncated TBCC construct were obtained. Despite efforts to optimise crystallisation and diffraction data, the structure was not solved for inclusion in this thesis. Instead, homologous structures were analysed and a potential tubulin interaction site is suggested based on the proposed GTPase-stimulating activity of TBCC and the similarity with the human protein, Retinitis Pigmentosa 2 (RP2), predicted to contain a domain with similar fold. Progress towards the soluble recombinant expression of the other cofactors also lays the foundation for future investigations into trypanosomatid TBC structure and function. Pteridine reductase 1 (PTR1), an enzyme unique to trypanosomatids, is the subject of Part II of this thesis. PTR1 is a broad-spectrum NADPH-dependent reductase, catalysing the two-stage reduction of biopterin to dihydrobiopterin and tetrahydrobiopterin and that of folate to dihydrofolate and tetrahydrofolate. As such, it can provide a bypass mechanism for the reduction of folates, reducing the therapeutic action of traditional antifolate molecules in these organisms. Inhibition of PTR1 is therefore desirable from a drug discovery viewpoint. The crystal structure of Leishmania donovani PTR1 was determined using data extending to 2.5 Å resolution with a view to generating ligand-complex structures and providing a model for inhibitor design. This structure was found to contain a disordered active site, with several loop regions not modelled or relocated. A sulfate molecule from the crystallisation mixture binds in the cofactor phosphate binding-site and the sequential binding of cofactor before substrate or inhibitor can not occur. Although this crystal form was considered unsuitable for further studies, it provides the only structure of PTR1 in the absence of cofactor. With an established crystallisation protocol, Trypanosoma brucei PTR1 then forms the basis of a collaborative investigation of over 100 novel potential inhibitory molecules. Kinetic evaluation, isothermal titration calorimetry (ITC) and co-crystallisation were applied to generate ligand-binding profiles of pyrrolopyrimidine derivatives. Several interesting binding features were identified from the 24 ligand complex structures obtained. These include the discovery of two covalent inhibitors, confirming the reactivity of a non-conserved active site cysteine, and molecules that are able to bind simultaneously at two locations within the active site pocket, exploiting hydrogen-bonding interactions with key catalytic and other nearby residues. The thermodynamic binding profiles of seven inhibitors also provide insight into the enthalpic and entropic contributions to ligand binding. We assessed the suitability of ITC for this system and while a high attrition rate was observed, chemical substitutions were able to enhance the binding entropy. These studies have strengthened our understanding of the structure-activity relationship between PTR1 and inhibitors, offering opportunities to develop new molecules that focus on increasing the potency generated by favourable enthalpy alongside improving the drug-like properties.

In silico modelling approaches are useful since they can minimise experimental costs and once set up can be used to replace animal testing for drugs. As such modelling techniques need to be developed to reduce dependence on animals for validation of pharmacological efficacy. The work within this thesis shows that computational methods can be used to model biological and medical problems effectively. The main aim of this thesis was to investigate Cytochrome P450 enzymes and their effect on drug metabolism through the use of the Hepatic Reductase Null(HRN) mouse. This was done through using a number of computational models and compared with drug data provided by CXR Biosciences. These models ranged from solely ODE (for comparison to experimental data) to multiscale cellular automata and spatial models when analysing the dynamics on the tumour and cellular level. Once these models were developed the parametric sensitivity was derived in order to see whether there were any needless parameters so that the models were streamlined and to test the model`s robustness against error. The novel three-compartment model was developed in order to explain dynamics within the Hepatic Reductase Null mouse was able to explain much of the behaviour in the supplied data. As well as this it was discovered that the transgenic mouse showed reduced speed in metabolism for many of the drugs analysed which meant that different models were necessary.The cellular automaton program presented is applicable to other areas other than the one stated in Chapter 5. For example any area that deals with interactions between tissue media and drugs as in toxicology and drug studies. The cell cycle inside the code deals with tumour cells but this code can be re-parameterised to concentrate on other types of cell including normal cells, hepatic tissue etc. The inclusion of spatial e ects to the deterministic models like the Cytochrome P450 cycle allows for greater realism in predictions of drug passage through the body or across certain tissue media. Due to this it is useful to include both deterministic and spatial modelling with a multiscale approach in models for drug metabolism.

Despite much being known about tubulin, there is a dearth of structural information on proteins that are involved folding, processing and manipulating this important component of the cytoskeleton. Trypanosomatids can be used to investigate these proteins, as they are primitive eukaryotes whose main cytoskeletal component are microtubules. It was envisaged that conserved features discovered would elucidate basic features of eukaryotic biology and specific differences would help to understand the specialisation of these parasites. Therefore several projects were initiated in order to further understanding of several microtubule modifying and manipulation proteins in trypanosomatids. This resulted in two X-ray crystallography structures being produced. One is of the ADP ribosylation factor like 1 (ARL1) and another of the CAP-Gly domain of tubulin binding cofactor B (TBC-B). Enzymes involved in posttranslational modifications of tubulin, tyrosine ligation and deacetylation, were also investigated. These projects did not result in crystal structures, but are discussed in chapters 5 and 6.In chapter 3, the structure of Leishmania major ADP ribosylation factor-like 1 (LmARL1) is described. ARL proteins are a family of small regulatory GTPases that undergo conformational changes upon nucleotide binding. Such changes are implicated in regulating the affinity of ARLs for binding other proteins, lipids or membranes. Much is known about the structure of human and yeast ARL proteins, research stimulated in large part due to their essential nature in vesicular biogenesis and membrane trafficking. There is however, a paucity of structural data on this family of proteins in the Kinetoplastida, despite biological studies implicating them in key events related to vesicular transport and regulation of microtubule dependent processes. The crystal structure of ARL1 from the protozoan parasite L. major in complex with GDP was determined to 2.1 Å resolution. The analysis reveals a high degree of structural conservation with human ADP ribosylation factor 1 (ARF1). Mining L. major genomic data identified putative ARF/ARL family members, which were examined and classified, based on conservation of amino acid sequences combined with what is known about orthologous proteins in Kinetoplastids and in humans, for which a standardised nomenclature has been proposed. This initial classification may guide future studies designed to elucidate biological function of specific family members. Chapter 4 contains the discussion of the two functional domains of Trypanosoma brucei TBC-B, a key protein involved in the correct chaperoning of quasi-folded tubulin to its native polymerisable state. There is currently little known about this system in trypanosomatids and there were no structures of this protein in protists. Although it was possible to crystallise the full length TbTBC-B protein, no diffraction was obtained. Instead the two domains of the protein were crystallised separately. These are the Ubl and CAP-Gly domains solved to 2.35 and 1.6 Å resolution respectively. It was calculated that the missing linker region most likely contains a small region of a-helical content but is mostly unstructured.The general fold of the Ubl domain is highly conserved, despite low sequence identity between this sequence and other TBC-Bs. Due to this low similarity it was not possible to determine which residues are important for complex formation between this protein and other cofactors or with tubulin.The CAP-Gly domain despite having higher sequence and structural fold conservation with other TBC-B CAP-Gly domains contains a major difference in the ß7-ß8 tubulin-binding loop. It is postulated this sequence difference of a glycine in the asparagine position of the highly conserved ‘GKNDG’ sequence may affect this domain’s binding to the a-tubulin C-terminal tail. The peptide binding groove of TBC-B forms crystal contact interactions with the C-terminus of a neighbouring molecule. This allowed for the comparison of this binding interaction with the more typical a-tubulin tail binding seen in other CAP-Gly domains. As this protein and domain both elute as monomers from gel filtration columns and this interaction probably represents a crystallographic artefact.All trypansomatid TBC-B sequences have differences in their tubulin-binding loops and C-terminus that are identical or similar to the a-tubulin tail. In addition to these differences, no other CAP-Gly domains were identified in the trypanosome genome. Although deacetylation is an important modification of tubulin, there are no structures of any tubulin deacetylases and although several histone deacetylase structures are known, none have been reported from protists. Therefore investigations were commenced to identify, clone and obtain recombinant material for crystalisation experiments. A putative tubulin deacetylase form T. brucei was identified and soluble recombinant material could be produced, attempts to crystallise or characterise this protein were prevented by degradation of the protein. At the start of this project, despite tubulin tyrosine ligase (TTL) activity having been identified over thirty years previously, no structures of any TTL enzymes had been determined. It was not possible to produce any material for crystallisation studies from the trypanosomatids therefore other source organisms were investigated. As Sus scrofa TTL had been produced previously in Escherichia coli, it was chosen as a target for crystallisation experiments. However, despite it being possible to obtain folded protein of high purity, no crystals were obtained.

No description available.