Search

Mass spectrometry (MS) is employed in many methods for its high sensitivity and specificity, and it is often used in hyphenation with a separation technique such as liquid chromatography (LC). Ion mobility (IM) provides an alternative separation method with higher throughput, much lower solvent use and removes the necessity to buy LC columns, however it is currently predominantly used in academic research groups rather than industry. Effective analytical techniques are required throughout the development and production of any pharmaceutical and mass spectrometry is a central method. This thesis explores the feasibility of using ion mobility coupled with mass spectrometry (IM-MS) for analysis of a range of pharmaceutically relevant molecules, from small molecules such as Thalidomide, to monoclonal antibodies (mAbs) to protein:protein heterocomplexes. It also provides insights to the structures of model and conformationally dynamic proteins and complexes. Many of the IM experiments included in this thesis involve deviation from commercially available IM-MS configurations. Chapter 2 involves doping the pure He drift gas with a chiral modifier to separate enantiomers. The mobility of any ion is dependent on the temperature of the drift gas with which it undergoes many collisions and as such the temperature of the drift gas is also the temperature of the ion of interest. Variable temperature (VT) IM-MS is a technique in which the temperature of the drift gas, and the ions as they move through it, can be deviated away from room temperature, either by heating or cooling. VT-IM-MS is used in Chapters 2-4 for different experiments. In Chapter 2 the temperature of the drift cell is altered in a series of experiments whilst measuring the equilibrium between the analyte ion and chiral modifiers. A van't Hoff analysis is then able to provide thermodynamic parameters for the interactions which provide rationale for observed enantiomeric separation. Experiments on L/D-tryptophan show a difference in both ΔH and ΔS for the 1:1 complex formed by each enantiomer with S-2-butanol. This approach is also rationalise using a predictive computational approach and experimentally extended to other enantiomeric ions and on two other commercially available IM-MS platforms. In Chapter 3, the use of VT IM-MS to probe the stability of protein fold is explored. The drift gas is heated to measure the conformations adopted by thermally unfolded protein structures and cooled which kinetically traps collisionally activated highly elongated structures. This is particularly prominent in ubiquitin and less so for lysozyme due to conformational restriction because of disulfide bonding. For the highly disordered protein alpha synuclein, at 250 K the Collision Cross Section (CCS) distributions are very similar for both collisional activated and non-activated forms, which we attribute to cold denaturation. Chapter 4 uses VT-IM-MS to explore this phenomenon in more detail, by measuring structural rearrangement as a result of cold denaturation of mAbs. Again cold denaturation is observed for three types of mAb at 250 K. These results are significant, this method allows cold denaturation to be measured directly, and surprisingly it is found to occur in the absence of solvent, meaning the structural perturbations occur within the intrinsic fold of the protein. Finally in Chapter 5, the [2Fe-2S]-binding proteins GLRX3 and Anamorsin are analysed using high resolution native-MS and IM-MS and are found to form a 1:1 heterocomplex. Both proteins are shown in IM to be highly disordered, and this disorder is retained on formation of the heterocomplex, resulting in a fuzzy complex. These results are rationalised in context of binding with NUBP1, and along with size exclusion chromatography (SEC), UV-Vis and NMR spectroscopy it is determined that these proteins are essential in providing the [2Fe-2S] clusters and electrons to form [4Fe-4S] clusters on NUBP1. Overall this thesis aims to demonstrate a diverse range of IM-MS experiments, and to relate these to pharmaceutical and structural biology applications.

Many pathogens function by internalising the haem molecules of their host organism and breaking down the porphyrin scaffold to sequester the Fe2+ ion. Typically, this breakdown mechanism is mediated by a haem oxygenase. However, a novel class of reaction has been discovered, which can be performed anaerobically using 'nature's reductant', NADH, and the Yersinia enterocolitica protein, HemS. To study the features of this reaction in more detail, conventional experimental methods were combined with Energy Landscape Theory. Deuterium labelling demonstrated that the reaction was initiated by hydride transfer and stopped-flow spectroscopy showed that the reaction proceeded via a short-lived intermediate. Since no structural information regarding NADH-binding to HemS was available, computational calculations were used to sample the conformational space around possible NADH-protein binding sites and to construct kinetic transition networks. From these networks, pathways showing the unfolding and approach of NADH to haem inside the pocket were determined. These pathways highlighted the roles of various residues, thus allowing for a targeted mutagenesis study. This study, carried out using both computation and laboratory-based experimentation, was especially focussed on a double phenylalanine gate located in the centre of the main cavity. Key insight concerning how this feature regulates the access of NADH to haem was gained. Computational results suggested that the HemS homologues, HmuS, ChuS and ShuS, were also capable of promoting anaerobic haem breakdown, but that catalysis by ChuS and ShuS may be limited by competing functions. Bioinformatics was used to gauge what these possible alternative functions could be, and to place HemS within its wider phylogenetic context. The computational predictions were then tested in the laboratory. The three homologues were all shown to engage in the reductive haem breakdown process but to varying degrees of efficacy. These findings demonstrate that this novel haem breakdown reaction is not unique to HemS, but instead is a feature of a wider class of haemoproteins. A subset of these haemoproteins are known to bind certain DNA promoter regions, suggesting not only that they can catalytically degrade haem, but that they are also involved in transcriptional modulation responding to haem flux. Many of the bacterial species responsible for this class of protein (including those that produce HemS, ChuS and ShuS) are known to specifically target oxygen-depleted regions of the gastrointestinal tract. A deeper understanding of anaerobic haem breakdown processes engaged in by these pathogens could therefore prove useful in the development of future strategies for disease prevention.

Rare monogenic diseases affect around 300 millions worldwide at any given time. Although genetically and clinically heterogenous, rare diseases are frequently caused by pathogenic mutation in a single gene; in particular, missense and in-frame insertion and/or deletion (indel) variants. Differentiating the few truly pathogenic alterations from the background of functionally neutral variants is a challenge which has resulted in a translational bottleneck. Largely due to the absence of segregation and functional data, the majority of reported variants-of-interest remain clinically defined as variants of unknown significance (VUS). Since experimental validation of all variants is impractical, computational prediction tools have been designed to attempt to prioritise variants for downstream analysis. While accurate variant prioritisation can lead to fewer VUS and a higher rate of diagnosis, this has yet to reach clinical standard. The aim of this study was to i) design a protein-specific prediction model to improve missense variant prioritisation and ii) use integrative structural analysis to differentiate pathogenic from benign in-frame indel variants. Using a protein-specific approach to variant analysis, structural and clinical data were combined using machine learning to differentiate pathogenic from benign missense variants in a subset of genes. The resulting prediction model, ProSper (protein-specific variant interpreter), outperformed a number of widely-used or recently developed missense prediction tools for the majority of genes studied. For a group of in-frame indel variants that were classified as VUS in a lab, structural interpretation was used to show variant clustering and accurately differentiate pathogenic variants in a family of genes. Variant prioritisation and interpretation using these approaches in clinically relevant situations can improve rare disease diagnosis and patient management. In addition, structural analysis at the atomic-level can be used to assess a variant's structural impact and provide data on the molecular mechanism of disease. This research adds to the body of evidence showing improved accuracy in variant prediction and prioritisation in groups of genes using a protein-specific approach compared to general predictors. Incorporating integrative structural analysis into the diagnostic framework can accelerate variant interpretation and minimise the rate of VUS leading to correct and timely diagnosis in more patients.

“What I cannot create, I do not understand” – Richard Feynman. While these famous words may stem from a physicist, they appear to have inspired scientists from a wider range of backgrounds: one of the key objectives of synthetic biologists is the design of functional proteins, a field which is said to have recently ‘come of age’ (1). The maquette approach provides a rational way of designing proteins without the complexity of their natural counterparts. The Anderson group have employed this method to produce a variety of four-helix bundle proteins, including C45, a heme-binding peroxidase with impressive thermostability (2). This work focusses on the structural and functional impact of organic cosolvents on C45. The secondary structure was analysed by circular dichroism in concentrations of 10-80% acetonitrile, methanol, ethanol, isopropanol and TFE. A slight increase in helicity could be observed in up to 60% of any of the alcoholic cosolvents. In high (up to 80%) TFE concentrations, this was particularly pronounced, confirming the ability of the cosolvent to stabilise α-helices (3–5). NMR spectroscopy also suggested that TFE may have a stabilising effect on the enzyme. Further, the effect of the cosolvents on the kinetics of the reaction between C45, ABTS and H2O2 was investigated under limiting peroxide concentration. While the general trend observed was a decrease in kinetic activity, in TFE, the activity increased. In the highest tested cosolvent concentration of 80% TFE, the kcat showed a 7-fold increase compared to that in buffer. Total turnover numbers increased even more significantly, by a factor of 34. This was highly unusual. A smaller increase in kinetic activity upon the addition of TFE is not unprecedented (6), however not commonly in concentrations above 30%. This was investigated further using stopped flow spectrophotometry. The reaction did not reach saturation and no kcat or KM for H2O2 could be determined, however an estimate suggests the kcat/KMa/KMb increased 4-fold compared to the reaction in buffer. Further, it was observed that both in buffer and 80% TFE, v0/[E]0 increased significantly with increasing enzyme concentration. This very unusual observation was attributed to different levels of catalytic activity present in the dimer and monomer of C45. The presence of both of those was confirmed by analytical SEC as well as SEC-SAXS. The mechanism of the formation of compound I, the radical intermediate in the catalytic cycle of heme peroxidases, was investigated using stopped-flow spectrophotometry of the reaction between C45 and H2O2. First, the kinetic isotope effect was exploited to confirm which phase of this reaction corresponded to compound I formation. This process was shown to be significantly faster in 80% TFE than buffer. This leads to the conclusion that stabilisation of the radical intermediate compound I plays an important role in the TFE-mediated increase in catalytic activity of C45. This work could provide a route to improving the catalytic activity of small heme peroxidases, which may lay the foundation of novel applications of maquette proteins, especially with the possibilities offered by the rapidly growing research field of enzyme engineering. This is especially relevant in light of the industrial interest in non-aqueous enzymology (7).

Few mathematical models of sequence evolution incorporate parameters describing protein structure, despite its high conservation, essential functional role and the increasing availability of structural data. The primary goal of my PhD project was to create a structurally aware amino acid substitution model in which proteins are represented using an expanded alphabet that relays both amino acid identity and structural information. Each character in this alphabet specifies an amino acid as well as information about the rotamer configuration of its side chain: the discrete geometric pattern of permitted side chain atomic positions, as defined by the dihedral angles between covalently linked atoms. I generated a 55-state “Dayhoff-like” substitution model (RAM55) by assigning rotamer states in 79,558 structures (∼50%of all PDBe entries) and identifying substitutions between closely related sequences. RAM55’s rotamer state exchange patterns clearly show that the evolutionary properties of amino acids depend strongly upon side chain geometry. Exploiting knowledge of these patterns assists in phylogenetic analyses: I show that RAM55 performs as well as, or better than, traditional 20-state models on simulated and empirical data for divergence time estimation, tree inference, side chain configuration prediction and ancestral sequence reconstruction. Further, encoding observed characters in an alignment as ambiguous representations of characters in a larger state-space allows the application of RAM55 to 20-state amino acid data for which structures are not known. Adding structural information to as few as 12.5% of the sequences in an amino acid alignment results in excellent ancestral reconstruction performance compared to a benchmark that considers the full rotamer state information. This strategy significantly expands the applicability of RAM55 to real-world scenarios where structure might only be available for some of the sequences of interest. Thus, not only is rotamer configuration a valuable source of information for phylo-genetic studies, but modelling the concomitant evolution of sequence and structure may have important implications for understanding protein folding and function.

Cytochrome c6A is a homologue of cytochrome c6 found in eukaryotic green algae and higher plants. However it is thought to perform a different function from cytochrome c6. Two current hypotheses exist for this function: that cytochrome c6A acts as a ‘safety valve’, providing an alternative path for electron flow through the photosynthetic electron transport chain and thus alleviating reactive oxygen species production; that cytochrome c6A has a signalling role where it could sense high light stress and convey this signal to the nucleus to affect gene expression. This study aims to provide insight into the ancestry and function of cytochrome c6A. To gain a clearer understanding of the evolutionary history of cytochrome c6A and cyanobacterial homologues cytochrome c6B and c6C, phylogenetic analysis was performed on peptide sequences of these proteins as found in a wide range of photosynthetic organisms. These data were used to refine the model of the evolutionary history of the cytochrome c6 family. This study also found evidence that cytochrome c6B and c6C are in fact orthologues with a similar function. Chlamydomonas reinhardtii is a model organism for eukaryotic photosynthesis, as its unicellularity provides many advantages over land plants. To determine if cytochrome c6A is involved in high light stress in C. reinhardtii, cytochrome c6A knockout and knockdown mutant lines of C. reinhardtii had been previously produced using CrispR-cpf,1 and are characterised in this study. The mutant lines demonstrated potential growth retardation under high or fluctuating light stress, as well as singlet oxygen stress, which was more noticeable under mixotrophic conditions. Chlorophyll fluorescence analysis of the mutant lines established a link between cytochrome c6A and NPQ. As cytochrome c6A is believed to be located in the thylakoid lumen, and initial NPQ is triggered by a change in luminal pH, circular dichroism was used to determine changes in secondary structure of purified cytochrome c6A over a pH range of 2-7. No significant change in structure was observed, but cytochrome c6A did maintain structural integrity even at pH 2. Finally, the transcriptome of the cytochrome c6A knock out line was compared to the background strain under standard and high light conditions through RNAseq. Many photoprotective, motility and CCM genes were differentially regulated under high light stress, but when cytochrome c6A was knocked out these regulations were diminished. Therefore cytochrome c6A has been proposed as a signalling molecule that functions in CCM, NPQ and motility under photosynthetic stress conditions.

As the number of solved protein structures increases, the opportunities for meta-analysis of this dataset increase too. Here we explore two approaches for analysing protein structure, both starting from the three-dimensional co-ordinates of each atom within the structure, which are then abstracted into a more useful form. The first method transforms the protein into a network in which its amino acids are the nodes, and where the edges are generated using a simple proximity test. By applying the Infomap community detection algorithm, we can fragment the protein into highly intra-connected subregions - these subregions are compact and globular, and can be compared with known structural and functional subunits of the protein (also known as domains). By performing this fragmentation process systematically across a large set of proteins, and checking for structurally conserved fragments, we can search for novel candidate domains. This method for automatically decomposing a protein into compact substructures may also be useful in coarse-graining molecular dynamics, analysing the protein’s topology, in de novo protein design, or in fitting electron density maps derived from single particle electron microscopy. The second method calculates a descriptor for each atom of the protein based on its local environment, known as a Smooth Overlap of Atomic Positions (SOAP) descriptor. Using these descriptors we can perform overall comparisons of the subregions identified above. In addition, by comparing the descriptors of a set of proteins known to share common structural or functional features (such as binding of a particular ligand), we can automatically identify the most highly conserved atoms of the set. These atoms may line ligand binding pockets or correspond to allosteric sites, which could inform drug design.

Proteins are the ultimate effectors of biological mechanisms and are involved in every aspect of cellular life. The functional properties of proteins depend on their three-dimensional structure. Indeed, proteins are not rigid entities and internal motions, on a wide range of timescales and distances, are necessary to accomplish a specific function. While certain proteins adopt a compact conformation and undergo small-scale rearrangements, others are more flexible and withstand more dramatic movements. Proteins that exhibit a regulatory role function by binding multiple partners, such as small molecules, DNA, RNA, other proteins. Increased levels of flexibility favour this binding promiscuity. A long-standing goal in molecular biology has been the development of new methods to enable the determination of three-dimensional structures of proteins that exhibit a high level of dynamic complexity. In fact, it is becoming clearer every day that individual structural techniques have several limitations. An integrative structural biology approach provides the tools to decipher the dynamic configuration of proteins by combining information from multiple sources, including biochemical, bioinformatics and biophysical methods. In this thesis, an integrated approach is used to structurally characterize two extremely different nucleic acids binding proteins, the human Staufen1 protein and the archaeal Pyrococcus abyssi MCM complex. In this study, I show that an integrated structural biology approach is not only essential to determine the architecture of small, flexible proteins, which are traditionally difficult targets for conventional approaches, but it is also beneficial to understand the dynamic behavior of large, more compact macromolecular complexes.

Monoclonal antibodies (mAbs) are now the most common protein therapeutics on the market. The therapeutic properties of mAbs originate from their inherent ability to specifically bind to a chosen target and the subsequent stimulation of effector functions as result of mAb glycosylation. Therefore mAbs are able to act as drug delivery systems, inhibitors or activators. Protein therapeutic characterisation possesses inherent analytical difficulties as mAbs are complex structures made up of multiple domains. The higher order structure of mAbs is essential for their specificity for targeted treatments. During the production of mAbs, the bioformulation process can induce structural perturbations that we separate into two main categories: Post-translational modifications (PTMs) and degradation. The most common PTMs are oxidation, deamidation and glycation. Degradation of mAbs covers general loss of structure through aggregation and fragmentation. Due to the wide variation and number of possible modifications to mAbs, monitoring, identification and quantification is difficult. Current analytical methods are dependent on sampling, are often destructive, and require time and expertise. Vibrational spectroscopic methods have been highlighted as possible analytical methods to overcome previous shortcomings. In particular, Raman spectroscopy has the potential to be developed into an in-line analytical tool due to being minimally invasive, non-destructive, information rich and rapid. Raman and FTIR are already utilised for in situ monitoring of pharmaceutical manufacturing, such as small molecules, but have not yet been extended into the quality control of mAbs. In this thesis we use a range of conditions to force the degradation and increase the PTM levels of three of the most common types of mAbs: IgG1, Fab and IgG4. We use circular dichroism (CD), peptide mapping, and size exclusion chromatography (SEC) to quantify and study the structural implications of the PTMs and degradation, and compare these to the information gained from Raman spectroscopy. Bioformulation parameters probed include: agitation, pH, temperature, sugar excipients, chemical stressors and UV light. Primarily, we explore the sensitivity of Raman spectroscopy to detect, quantify and provide a structural insight into the modifications. Additionally, we study the aggregation of mAbs to investigate the structural consequences on the protein, as well as identify spectral markers that could be used to monitor aggregation.

Pseudomonas aeruginosa is a common opportunistic pathogen. Recent work indicates that in many infection scenarios, P. aeruginosa exhibits an exquisite predilection for metabolizing fatty acids to yield acetyl-CoA. In most higher organisms, acetyl-CoA cannot be used for biomass production because the two carbon atoms which enter the TCA cycle are lost as CO2 . However, many bacteria are able to bypass these oxidative decarboxylation steps, allowing them to conserve carbon for gluconeogenesis. They perform this by using the "glyoxylate shunt". Here, isocitrate is cleaved by isocitrate lyase (ICL) to yield succinate and glyoxylate (which, in a subsequent reaction, is combined with a further acetyl-CoA unit to yield the gluconeogenic precursor, malate). However, ICL has to compete with the TCA cycle enzyme, isocitrate dehydrogenase (ICD), for the available isocitrate, and it is the outcome of this "metabolic tussle" which dictates the flux of carbon through the glyoxylate shunt. In E. coli, ICD is inactivated by AceK-dependent phosphorylation, allowing flux through the glyoxylate shunt. However, P. aeruginosa is "wired up" differently because it employs not one, but two highly-expressed isocitrate dehydrogenases (ICD and IDH). For this PhD project, I focused on these three enzymes (ICD, IDH and ICL). I cloned, overexpressed and purified them at high yield to perform a thorough investigation of their kinetics, regulation and more interestingly crystal structures. I found that only one of these (the E. coli-like ICD) is regulated by AceK-mediated phosphorylation. The other, IDH, is allosterically regulated, as is the isocitrate lyase. These findings demonstrate that in P. aeruginosa the rerouting of the carbon flux through the glyoxylate shunt is delicately regulated via allostery mainly. The conditions in which the cells grow and access to either poor or rich carbon sources heavily influence the partitioning of the central metabolism. In P. aeruginosa, the TCA cycle remains more active (than in E. coli for example) even during growth on poor nutrient and this is probably an important aspect to manage oxidative stress accompanying growth. Finally, I have solved the x-ray crystal structures of ICD, IDH and ICL. These are entirely novel structures that have not been defined previously and are new entries to the Protein Data Bank. The structure solving work highlighted very interesting peculiarities to these enzymes when compared with other bacterial pathogens. This emphasizes the growing idea that Pseudomonas aeruginosa is a unique bacterium that cannot be modelled by the well-studied Escherichia coli. All this work crystallizes the knowledge to build up a picture of how flux is likely to be regulated at this "non- canonical" metabolic branchpoint and features new interesting directions for downstream applications such as drug-design.

Amyloidogenic neurodegenerative disorders such as Alzheimer's disease (AD) and Parkinson's disease (PD) are an important health issue. However, the underlying molecular mechanisms of the disease-related protein aggregates, that are present in humans, are only understood partially. I have used and developed biophysical methods to study the structural and biological properties of individual aggregates of Amyloid β peptide and α-Synuclein, proteins whose aggregation is associated with the development of Alzheimer's and Parkinson's disease respectively. I expanded the single aggregate visualisation through enhancement (SAVE) technique, which is a method based on the fluorescent dye Thioflavin T (ThT) that reversibly bind to the aggregates and whose fluorescence increases upon binding. I firstly explored the use of other dyes for these experiments and found that a ThT dimer has higher affinity to α-Synuclein aggregates in vitro. I then applied the SAVE method to the cerebral spinal fluid (CSF) of a cohort of AD patients and control CSF and observed no clear difference in aggregate number. However, these experiments provided insights into how antibodies bind the aggregates in human CSF. I could show, that despite altering the Ca2+ influx into both cells and vesicles, the antibody did not measurably affect the aggregate structure. To study the size specific effects of the Amyloid β 42 (Aβ42) peptide in more detail, I used and optimised gradient ultracentrifugation combined with single aggregate imaging to study the structural properties of the isolated aggregates. This aggregation kinetic independent method allowed me to compare the properties of fluorescently labelled and unlabelled Aβ42 and characterize the size dependent properties of aggregates in a single experiment. Since I could measure the relative concentration of different size aggregates it was also possible to compare the properties of single aggregates of different sizes. I then used biological assays to examine the ability of aggregates to permeabilise membranes resulting in the entry of calcium ions, and their ability to induce TNFα production in microglia cells. Both processes are thought to play key roles in the development of AD. I found that small soluble oligomers are most potent at inducing Ca2+ influx, whereas longer protofilaments are the most potent inducers of TNFα production. My results suggest that the mechanism by which aggregates damage cells changes as aggregation proceeds, as longer aggregates with different structures are formed. Protofilaments with a diameter of 1 nm or less have a structure that could make them particularly potent at causing the signalling of toll-like receptors, providing a molecular basis for their ability to induce TNFα production.

Dinoflagellates hold vast diversities and are major contributors to overall marine primary production. Close relatives to the apicomplexan parasites and ciliates, the dinoflagellates, however, have a shockingly different nuclear biology. The dinoflagellates have massively inflated genomes (2-245 Gbps, up to 80X that of humans), permanently condensed liquid crystalline chromosomes throughout the life cycle, minimal histone proteins expression, and loss of nucleosome-mediated chromosome organisation. These changes co-occur with adoption of a novel nuclear protein termed DVNP (Dinoflagellate/Viral NucleoProtein). DVNP is highly positively charged, nucleus-located, and is found in only dinoflagellates and numerous marine large DNA viruses. I will address two issues in this thesis, 1) what are the properties of the protein DVNP, and can we infer if and how DVNP manages the permanently condensed chromosome state; and 2) how do other proteins accommodate this drastically different organisation of genetic material, more specifically how does transcription work in a cell with permanently condensed chromosomes? To elucidate the location of transcription, specific probe were designed and synthesised to locate RNA polymerase and newly synthesised nascent RNA in relation to the compacted chromosomes, and result showed that both are distributed at the periphery but not the inside of the chromosomes. On the other end, DVNPs are expressed and purified to perform biophysical and biochemical characterisation. In addition, using single-molecule optical tweezers microscopy, DVNP was observed to compact and change the mechanical properties of DNA. Most interestingly, the DVNP/DNA aggregates showed a propensity to travel along the DNA strand en masse. I then endeavoured to solve the NMR structure of DVNP. The results suggest that DVNP plays a central role in chromatin packaging in dinoflagellates. Finally, to marry the results obtained from both experiments, at the end of the thesis I propose a model of a novel nucleosome-independent chromatin management in dinoflagellates and how its transcription could proceed.

Determining protein structure is the central challenge of structural biology, owing to the integral role of proteins in biological systems along with the complexity and diversity of their structures. It is not only an important challenge, but a difficult one; the large size of proteins gives rise to a limitless range of potential structures, while the traditional methods of NMR and crystallography both carry specific limitations. Hence, there is a drive to develop new, orthogonal techniques that could provide additional insight into protein structure. Mass spectrometry (MS) is a ubiquitous analytical technique across all fields of chemistry, and the development of "soft" ionisation techniques has allowed noncovalent interactions to be retained in the gas phase and observed by MS. This in turn has led to the development of "native" MS in which proteins can be analysed and studied in the gas phase with their structures largely intact. Top-down MS of native proteins has been suggested as a potential avenue for studying protein structure. Here, the top-down technique of electron capture dissociation (ECD), which does not disrupt noncovalent interactions, is investigated for its potential to observe protein structure in native MS. Firstly, ECD was used to determine the sites at which a range of metal ions bind to the unstructured protein α-synuclein. These results were analysed through both a traditional top-down workflow and a novel mass defect based analysis. This mass defect analysis demonstrated how metal-containing fragments can detected without determination of their monoisotopic signal and shows potential to be applied in a range of top-down experiments. Secondly, ECD was investigated for its ability to probe the structure of native proteins. Carbonic anhydrase was used as a model protein, and it was observed that ECD can be used to monitor gas-phase unfolding pathways in much greater detail than otherwise available. However, studying multiple structural homologues showed that ECD is not exclusively directed by structure, but rather that availability of positive charge plays an equally important role in directing fragmentation. This was further investigated by producing K→A variants of carbonic anhydrase specifically designed to target ECD-sensitive regions of the protein. Our results demonstrate that minor differences to the primary sequence of a protein can significantly alter the observed ECD fragmentation patterns.

This thesis details the development and application of mass spectrometry (MS) labelling strategies for use in structural biology, and, specifically, how they offer insight into the quaternary structure and dynamics of protein complexes. Chapter 3 first demonstrates the utility of isotopic H/D exchange (HDX) strategies for probing the in solution dynamics and assembly of two protein systems. Firstly, I use solution HDX to probe the light-induced structural change of the FAD-binding protein cryptochrome. Through quantitative analysis, I discover a region at the C-terminus undergoes a long-lived conformational change with high specificity to blue light, suggesting a potential role in the signalling cascade of circadian processes. Secondly, I use HDX-MS to probe the interface of a homo-dimeric ancestor of extant haemoglobin. Differential HDX-MS revealed protected peptide regions consistent with the alpha/beta packing interface in the modernday hetero-tetramer. Studying proteins in the gas-phase by using native MS has the distinct advantage of high resolution separations, manipulations and accurate mass determination, afforded by the exquisite control possible in vacuum. Chapters 4, 5 and 6 describe the implementation of performing gas-phase HDX (gpHDX) inside the spectrometer and how the technique may be used to interrogate protein ion structure with high specificity. I first demonstrate how gpHDX can be used to probe native proteins and show high sensitivity to the surface chemistry of exposed residues. Secondly, I use gpHDX to aid our understanding of how protein complexes asymmetrically dissociate through collisional activation and charge precipitated unfolding. Importantly, I show that the deuterium uptake of a dissociated subunit can be used as a proxy for the intact precursor. Finally I use gpHDX coupled to ion mobility (IM) to reveal how the conformational landscape of an activated protein changes at a global and local level.

Enzymes are powerful biological catalysts that provide rate accelerations of up to 1019 times the rate of the uncatalysed reaction in water by lowering the activation energy of the reaction. Enzymes also have an extraordinary specificity for reacting substrates, which can distinguish between groups with similar structural and chemical properties. Studies of enzyme structure and mechanism have been motivated by the wish to understand how enzymes exhibit such behaviour. Along with the protein tertiary structure, dynamics have also been shown to be important in determining the function of a given protein. Proteins experience motions on various timescales, from several seconds for large scale domain rearrangements to picosecond and femtosecond side- chain rotations and bond vibrations. However, it is not clear how motions on these different timescales are coupled to one other and much effort has been dedicated to understand how complex internal dynamic motions results in such efficient and highly selective enzyme catalysis. In this study we aim to characterise fast timescale dynamics and mechanistic details of phosphoglycerate kinase (PGK) during its catalytic trajectory. PGK (45 kDa) is an enzyme which generates ATP during the first step of glycolysis and it consists of two domains, an N-domain and a C-domain separated by a central -helix. It catalyses the reaction between ADP and 1,3-bisphosphoglycerate, generating ATP and 3-phosphoglycerate as products. We used NMR relaxation experiments to quantify backbone dynamics of human PGK and Geobacillus stearothermophilus PGK along their catalytic trajectories. Our results indicate that interdomain communication is present for both PGK homologues upon binding of either substrate. The backbone dynamics of the ligand bound domain are perturbed and are propagated to the neighbouring domain. Furthermore, we utilised X-ray crystallography experiments to understand the mechanistic details of the active site of PGK when bound to native substrates. Catalytically competent crystals of PGK in a closed conformation have been observed that reveal the details of active site interactions.

P-glycoprotein (P-gp or ABCB1) is a membrane-bound active transporter belonging to the ABC protein superfamily. It is responsible for xenobioIc efflux and also contributes to multidrug resistance in diverse diseases including cancer and epilepsy. P-gp has been increasingly recognised as a potential target for future therapeutics. Although the protein has been studied for decades, understanding of the P-gp transport mechanism is still incomplete. Two P-gp orthologues, mouse (m) and human (h), were therefore expressed in yeasts and purified in the presence of the detergent, n-Dodecyl-β-D- Maltoside (DDM). Purified proteins were examined for aggregation and monodispersity via dynamic light scattering (DLS) and their thermal stability was determined by an assay using a thiol-specific dye (CPM). ATPase activity, measured in a detergent environment, showed that the proteins were active with a basal activity of 60 ± 4 and 35 ± 3 nmol/min/mg for mP-gp and hP-gp, respectively. Crystallisation trials were conducted in the presence of nucleotide. In meso crystallisation using commercial monoolein pre- dispensed plates yielded hexagonal crystal-like objects however they failed to diffract X- rays. P-gp samples were also subjected to cryo-EM where mP-gp in the post-hydrolytic (ADP-bound, vanadate-trapped) state provided the highest resolution dataset that led to a reconstruction of 3D density map at the resolution of 7.9 Å which showed an inward- facing conformation. Rigid-body model fitting unveiled densities that were not accounted for by the fitted model illustrating new features such as bound ADP, extended NBD1- TMD2 linker and alternative allocrite-binding sites. Ultimately, the knowledge of P-gp conformation alteration was enhanced and a refined alternating access mechanism of P- gp was proposed based upon information derived from this study.

Antibodies are soluble proteins produced by the adaptive immune system to bind and counteract invading pathogens. The binding properties of a typical human antibody are determined by the structure of its variable domain, composed of two chains – heavy and light and by the conformation of six loops located on the surface of the variable domain, known as Complementarity Determining Regions (CDRs). In the first chapter, we describe our analysis of the conformational space occupied by five out of six antibody CDRs (L1, L2, L3, H1 and H2) and the development of a novel, length-independent method for grouping these CDRs into structural clusters (canonical forms). We show that using our method we can increase coverage and precision of assigning CDR sequences into clusters. In the next chapter, we describe a method for ranking structural decoys of the CDR-H3 loop. We show that by computationally perturbing CDR-H3 decoys we can improve the performance of existing ranking methods. In the same chapter, we discuss the development of a method for high-throughput assignment of heavy-light chain orientation. The power of the method was demonstrated by assigning orientation to billions of potential Fv sequences. The third Chapter describes the analysis of a large dataset of CDR sequences with the aim of identifying sequence patterns responsible for the loops' structure. Using a neural network methodology, we found several groups of CDR sequences which might be indicative of previously-unseen conformations. In the final results Chapter, we describe how we used the structural knowledge developed throughout the rest of the thesis to create a novel pipeline for computational antibody design. We show that the binders developed using our methodology had similar features to available antibody therapeutics and low predicted propensity to cause an immunogenic response. These results demonstrate the potential for using computational methods for designing high affinity therapeutics with human properties.

The work described in this thesis is an experimental study into the application of Electron Paramagnetic Resonance (EPR) Spectroscopy for the study of biological systems. Using a variety of methods of site-directed spin-labelling (SDSL), this thesis aims to explore long range structure in an assortment of recombinant and native proteins, and complexes thereof. The work described in this thesis covers all aspects of the work, from experimental design, molecular biology and cloning, protein expression and purification, as well as functional characterisation, and finally EPR distance measurements, data analysis and interpretation. Challenges and pitfalls will also be addressed. Chapters 1 and 2 introduce EPR spectroscopy, and its application in the study of long range structure in biological systems. The experimental techniques employed throughout this thesis are also introduced. Chapter 3 details an investigation into the complement C3b:factor H complex. This chapter addresses the challenges associated with the SDSL of cysteine rich proteins. Utilising hidden cysteine residues in native proteins for spin-labelling purposes will also be addressed. Chapter 4 looks at the interactions of the human myosin regulatory light chain (RLC) with cardiac myosin binding protein C (cMyBP-C). Optimisation of expression and purification protocols will be the focus, as well as addressing issues with protein solubility and spin labelling efficiencies. Chapter 5 explores the development of new methods of SDSL, for the specific labelling of cysteine rich proteins. The ability of Escherichia coli to read through the amber stop codon will be exploited for the incorporation of unnatural amino acids for labelling purposes, and novel spin labels, specific for labelling cysteine pairs tested in several model systems. Furthermore, native paramagnetic centres in recombinant proteins will be explored as potential labelling sites.

ATP-dependent chromatin remodelers have been proposed to act sequentially, and to a certain extent non-redundantly, in the priming stages of the DNA Damage Response pathways by establishing chromatin in lesion sites ready to act as a scaffold for repair factors or to be displaced in order to allow DNA repair. Among remodeling factors proposed to play a role in DNA repair is SMARCAD1, a poorly characterized, non-canonical member of the SWR1-like family of SNF2/SWI2 superfamily of ATPases, which has recently been identified as a potential target for ATM/ATR phosphorylation at canonical and non-canonical sites upon DNA damage. The actual mechanism for SMARCAD1 recruitment and involvement in DNA remodeling is still unknown, and unlike most other chromatin remodelers, SMARCAD1 does not contain DNA- or histone-binding domains frequently accompanying such proteins. Instead, in addition to the core ATPase domain, only two CUE domains (a type of helical ubiquitin-binding domain) have been identified. This thesis presents the findings of an investigation intended to structurally characterize SMARCAD1 by dissecting and identifying its domain architecture, and examining the activity and ligand selectivity of its binding domains in the functional context of DNA damage repair. The solution NMR structure of the CUE1 domain is presented, describing a triple helix bundle consistent with other members of the family. Furthermore, a novel SUMO interacting motif was identified and through a combination of NMR titrations and phospho-proteomics analysis, shown to be constitutively phosphorylated which excludes the possibility of DNA damage dependent ATM targeting as the recruitment mechanism for DNA repair. Additionally, it is demonstrated that both CUE domains are poor binders of mono-ubiquitin, however CUE1 specifically mediates the high affinity binary interaction with the transcriptionally repressive master regulator KAP1. This interaction was shown to be independent of post-translational ubiquitylation but rather sustained through direct interaction with the dimeric RBCC domain of KAP1. Finally, mass spectrometry profiling of domain-dependent interactions (based on differential abundance relative to changes due to chemically induced DNA damage) suggests SMARCAD1 may be involved in p53 transcriptional regulation through interactions maintained with CUE1 prior to DNA damage, whereas the SIM domain selectively targets protein interactions upon DNA damage that simultaneously activate p53 transcriptional control and recruit SMARCAD1 to DNA damage repair pathways.

The biophysical properties, biological activity and function of macromolecular systems are highly dependent on their structure. Structure-activity relationships of proteins and their binding partners are critical for drug discovery, biochemical and medical research. While the gas-phase environment might present as an unusual venue from which to explore protein structure, for over the past two decades, nano-electrospray ionization (nESI) coupled to mass spectrometry (MS) has been recognized as having great potential for analysis of protein structure and protein non-covalent complexes. In conjunction with related technique of ion mobility (IM), mass spectrometry (IM-MS) provides insights into protein native-like conformations and any structural changes in may undergo upon ligand binding or alternations induced via physical parameters such as temperature, pressure or solution conditions. As most proteins tend to exist as multiple domains; from the distribution of oligomeric states in the Protein Data Base (PDB) 86% of proteins exist as oligomers; the work presented in this thesis focuses on application of MS techniques to probe the tertiary and quaternary structure of various large and multimeric protein complexes, their dynamics and/or conformational changes. Wherever relevant, the gas-phase studies reported here are complemented by other techniques, such as hydrogen deuterium exchange MS (HDX), molecular modelling (MD) and analytical ultracentrifugation (AUC). Firstly, the dynamics of intact monoclonal antibodies (mAbs) and their fragments are explored with IM-MS. Variations observed in conformational landscapes occupied by two mAb isotypes are rationalized by differences in disulfide linkages and non-covalent interactions between the antibody peptide chains. Moreover, mAb intrinsic flexibility is compared to other multimeric protein complexes in terms of collision cross section distribution span. Secondly, variable temperature MS (VT-MS) and variable temperature IM-MS (IM-MS) are used to probe unfolding and dissociation of four standard multimeric protein complexes (TTR, avidin, conA and SAP) as a function of the of analysis environment temperature. VT-MS is found to allow for decoupling of their melting temperature (Tm) from the protein complex dissociation temperature (TGPD). Whereas, VT-IM-MS is used to investigate structural changes of these protein complexes at elevated temperatures and provide insights into the thermally induced dissociation (TID) mechanism, as well as strength of the non-covalent interactions between subunits. Thirdly, VT-(IM)-MS methodology is applied to study behaviour of three mAbs: IgG1, IgG4 and an engineered IgG4 of increased thermal stability. Such analysis shows to be promising for comparative thermal stability studies for proteins of therapeutic interest. Lastly, the structure of ATP-phosphoribosyltransferase (MtATPPRT), an enzyme catalysing the first step of the biosynthesis of L-histidine in Mycobacterium tuberculosis, is explored. Conformational changes occurring upon feedback allosteric inhibition by L-histidine are probed with MS, IM-MS, HDX-MS and AUC. Reported results serve as the basis for IM-MS/HDX-MS based screening method to be used for screening of a library of novel and promising anti-tuberculosis agents.

The research described in this thesis employs a variety of Electron Paramagnetic Resonance (EPR) techniques for the study of the electronic and structural properties of artificial supramolecular porphyrin systems and of protein complexes of biological relevance. The electron delocalisation in the cationic radical and photoexcited triplet states of linear and cyclic Π-conjugated multiporphyrin arrays was investigated. In the radical cations, information on the extent of delocalisation can be inferred from the measurement of hyperfine couplings, either indirectly from the continuous wave EPR spectrum or directly using pulsed hyperfine EPR techniques. The results of room temperature EPR experiments showed complete delocalisation of the electron on the timescale of the EPR experiments, but frozen solution EPR measurements revealed localisation onto mainly two to three porphyrin units in the larger porphyrin systems. Information on the delocalisation of the triplet state in the same porphyrin systems is contained both in the hyperfine couplings and in the zero-field splitting (ZFS) interaction. The results outlined in this thesis show that the hyperfine couplings provide a much more accurate estimate of the extent of delocalisation. The trends in proton and nitrogen hyperfine couplings with the size of the porphyrin systems indicate uneven spin density distributions over the linear arrays, but complete delocalisation in the cyclic systems. Time-resolved EPR and magnetophotoselection experiments have shown a reorientation of the zero-field splitting tensor between a single porphyrin unit and longer linear arrays, resulting in alignment of the main optical transition moment and the Z axis of the ZFS tensor. Continuous wave and pulsed dipolar EPR techniques were employed for the determination of the structure of two different protein complexes, the homomultimeric twin-arginine translocase A (TatA) protein channel and the ferredoxin-P450 complex of the electron transport chain in Novosphingobium aromaticivorans. The interaction between nitroxide spin labels introduced at different positions of the TatA monomer was investigated in the complex reconstituted in detergent micelles by analysing the dipolar broadening of the EPR spectra and the results of three- and four-pulse Double Electron-Electron Resonance (DEER) measurements. In combination with results from NMR and molecular dynamics calculations, a structure for the channel complex was proposed. The structure of the ferredoxin-cytochrome P450 complex was investigated by orientation-selective DEER between nitroxide labels introduced on the cytochrome P450 protein and the iron-sulfur cluster of the ferredoxin. The distance and orientation information contained in the experimental DEER data was interpreted in terms of a structural model of the protein complex by orientation-selective DEER simulations combined with a modelling approach based on protein-protein docking.

3D proteomics combines chemical cross-linking with mass spectrometry to study the structure of protein assemblies and protein-protein interactions both in vitro and in vivo by providing distance constraints that indicate which residues are in close spatial proximity. I addressed the main bottleneck of this technology: the reliable identification of cross-linked peptides. Reporter ion signatures for cross-linked peptides were developed, by fragmenting model compounds containing two lysine residues joined by a cross-linker backbone or a lysine residue modified with a hydrolysed cross-linker. The reporter ion signatures showed 97% specificity at 90% sensitivity and segregated cross-linked from modified and linear peptides. They decreased the false discovery rate of the identification of cross-linked peptides from 5% to 1% in a large dataset. The signatures permit data sorting during and after mass spectrometry acquisition. The advanced 3D proteomics workflow was applied to study the protein-protein interactions in Mycoplasma pneumoniae cells. In lysates of the bacterium we identified 128 protein-protein interactions (of which 24 are novel) and obtained in vivo topological data on 208 proteins, even for cases where high-resolution structures are not yet available. We showed that our data are in excellent agreement with crystal structures of proteins and complexes where available. We defined a network of ribosomal and RNA polymerase proteins that reveals an intricate link between transcription and translation in bacteria. We demonstrated that the method is suitable for identification of homomultimeric protein complexes by exploiting peptide pairs of identical amino acid sequence. The technology has the potential to provide a complete protein interaction network map after the selective enrichment of cross-lined peptides is achieved. The method was next applied to investigate the structure of condensin and cohesin complexes, which play a crucial role in stabilization of chromosome structure during mitosis. The complexes were purified, cross-linked and their linkage map created. The condensin coiled coil cross-linked on the entire length was modeled. The information was used to direct the analysis of in situ cross-linked condensin in intact chromosomes. I found two high confidence linkages between SMC2 and SMC4 coiled coils and identified H2A as a potential condensin receptor on chromosomes.

Protein evolution is a complex and diverse process, yielding an incredible assortment of biological functions and pathways occurring in the cells of living organisms. The way in which a protein's structure is constrained by its functional role and its notable conservation across even distant evolutionary relationships highlight structure as an important unit when considering the evolutionary dynamics of proteins. This thesis attempts to place the structural landscape of the protein universe within an evolutionary framework. We investigate potential evolutionary histories of protein superfamilies by introducing an age, which estimates when the ancestor of that superfamily first evolved. The range of ages of known protein superfamilies goes right back to those which evolved before the diversification of life into three major superkingdoms. The structures of these proteins are varied but those which have evolved more recently tend to be shorter and have a less elaborate globular packing. Protein structures sit within a complex global landscape of three-dimensional folds and we attempt to model the dynamics of this space using networks of folds. These networks consist of a structurally diverse core of folds with older ages, and neighbouring folds tend to be of similar ages. Moreover, there are a few pivotal folds which appear repeatedly as central in the landscapes, connecting together otherwise disparate portions of the space. Sequence profiles which capture patterns of conservation and variation amongst naturally occurring proteins within a superfamily can be compared to identify distant evolutionary relationships. The power of these profiles to detect such relationships is improved by seeding them with structural alignments. A landscape of evolutionary links crossing between different protein folds is presented.

The 2OG oxygenases comprise a superfamily of ferrous iron dependent dioxygenases with multiple biological roles, including in hypoxia sensing, transcriptional control, and splicing control. It was recently proposed that 2OG oxygenases catalyse the hydroxylation of ribosomal proteins in prokaryotes (ycfD) and in humans (NO66 and MINA53), raising the possibility that 2OG oxygenases also control translation. The work described in this thesis concerned investigations on the biochemical and functional aspects of prokaryotic and mammalian ribosomal protein hydroxylases (ROX) in vitro and in cells. An efficient chromatographic system linked to mass spectrometric analysis (LC-MS) was developed for studying the masses of individual ribosomal proteins (>90% coverage of ribosomal proteome) to ±1 Da accuracy. It was demonstrated that ycfD catalyses the hydroxylation of R81 on L16 in E. coli, in a manner dependent on atmospheric oxygen levels. YcfD deletion results in growth phenotype at low temperatures and in minimal medium, and in decreased global translation rates in minimal medium; ycfD deletion does not affect translational accuracy and ribosome assembly. Furthermore, ycfD-deletion results in increased sensitivity to the antibiotics chloramphenicol and lincomycin. Consistent with a 2OG-oxygenase mediated mechanism of antibiotic resistance, chloramphenicol sensitivity of the E. coli wild-type strain could be increased by inhibiting the activity of ycfD by removing co-factors required for catalytic activity (Fe(II) and O2), and, at least in part, by using a ycfD inhibitor, IOX1, which inhibits ycfD with IC50 of 38 μM in vitro. The therapeutic potential of a post-translational modification mediating antibiotic resistance provides an opportunity for medicinal targeting of ribosome-modifying enzymes, for example ycfD, which may be more ‘druggable’ than the ribosome itself. In co-treatment with an existing antibiotic, such as chloramphenicol, a small molecule inhibitor would achieve a potentiated antibiotic effect. Structural aspects of ROX hydroxylation were pursued by characterising a thermophilic ROX-substrate complex; a ycfD homologue was identified in the thermophilic bacterium Rhodothermus marinus and shown to be a thermophilic 2OG oxygenase ycfDRM, acting on R82 of ribosomal protein L16RM. The activity of ycfDRM in cells was limited at high growth temperature and oxygen solubility was demonstrated as a likely limiting factor of ycfDRM activity, thus identifiying a potential 2OG oxygenase oxygen sensor in prokaryotes. A crystal structure of ycfDRM in complex with L16RM substrate fragment was determined to 3.0 Å resolution. Structural analyses suggested that ycfDRM contains 30% more hydrophobic interactions and 100% more salt-bridge interactions than ycfDEC, suggesting that these interactions are important for thermal stabilisation of ycfDRM. The structures reveal key interactions required for binding of ribosomal proteins. Substantial structural changes were observed in the presence of the substrate fragment, which implies induced-fit binding of the L16RM substrate. The work has informed further structural studies on the evolutionarily related human ROX, NO66 and MINA53, for which substrate structures have been obtained since the completion of the work. The LC-MS analysis of ribosomal proteins was extended to mouse and human cells to demonstrate that the human ROX homologue of ycfD, MINA53, hydroxylates the 60S ribosomal protein rpL27a in cells. It was demonstrated that rpL27a hydroxylation is widespread and found in all mouse organs analysed, as well as in cancer cell lines and in clinical cancer tissues. A partial or complete reduction of rpL27a hydroxylation was observed in a number of clinically identified MINA53 mutations from the COSMIC database of cancer mutations. Structural analysis suggested that mutations occur more frequently at structurally important regions of MINA53, including the βIV-βV insert in the core fold of MINA53. The identification of inhibiting clinical mutations suggests that rpL27a hydroxylation level could be used as a cancer mark, and in the future for selective inhibition by ribosomal antibiotics. The work presented in this thesis demonstrates that it is possible to selectively inhibit modified ribosomes; an inhibitor of unhydroxylated rpL27a could therefore, at least in principle, be active against the sub-set of tumours with inactivating mutation(s) of MINA53, but not normal tissue. Future work should therefore focus on identifying a selective inhibitor of unhydroxylated eukaryotic ribosomes which could be applied for treatment of cancers harbouring deactivating MINA53 mutations. The same approach could be applied to other ribosome modifications (to rRNA, ribosomal proteins, and ribosome-associate factors) that are different in cancer compared to normal cells.

The Fe(II) and 2-oxoglutarate dependent oxygenases (2OG oxygenases) catalyse a broad range of oxidative reactions in various organisms. 2OG oxygenases use 2OG and molecular oxygen to catalyse the oxidation of a variety of substrates including small molecules, fatty acids, nucleic acids and proteins. Several human 2OG oxygenases are implicated in diseases. The fat mass and obesity associated protein (FTO) is linked to obesity, whilst collagen prolyl-4-hydroxylases (CPHs), the procollagen lysyl hydroxylases (PLODs), and the hypoxia inducible factor hydroxylases (PHDs) are linked to cancer. Therefore, structure-based inhibition studies on FTO, CPH and related 2OG oxygenases are of significant therapeutic interest. The obesity-associated FTO is an mRNA N6-methyladenine (m6A) demethylase and a homologue of E. coli alkylated DNA repair protein (AlkB), a DNA repair enzyme that demethylates N1-methyladenine (m1A) and N3-methylcytosine (m3C) bases. Human AlkB homologue 5 (ALKBH5) has a similar substrate profile as FTO, i.e. ALKBH5 is another mRNA m6A demethylase, making it a target for structural and inhibition studies to improve inhibitor selectivity for FTO. The CPH and PLODs catalyse hydroxylation of collagen prolyl- and lysyl-residues; the resultant 4-hydroxyprolyl- and 4-hydroxylysyl-residues are important for the formation of the collagen triple helix and intermolecular crosslinks, respectively. Another 2OG oxygenase, the 2-oxoglutarate and iron-dependent oxygenase domain-containing protein 2 (OGFOD2), is related by sequence similarity to the PLODs but its biological role is unknown. This thesis describes crystallographic, biochemical and inhibition studies on five 2OG oxygenases: AlkB, FTO, ALKBH5, OGFOD2 and CPH. Three AlkB-inhibitor complexes were determined; taken together, these structures serve as proof-of-principle that the AlkB subfamily 2OG oxygenases are amenable to structure-based inhibition studies. Several classes of inhibitors of FTO were then identified and their binding modes were investigated through the determination of 7 crystal structures of FTO-inhibitor complexes. Subsequently, a crystal structure of ALKBH5 was determined, which provided insights into its substrate recognition mechanisms. The ALKBH5 structure also serves as a template for inhibitor design. Preliminary structural and biochemical data were also obtained for OGFOD2 and CPH. Overall, these results contribute to the development of a biophysical understanding of human 2OG oxygenases, and will help to enable the development of selective inhibitors of 2OG oxygenases involved in nucleic acid and collagen modifications.

This thesis examines the dynamic mechanical properties of natural silk fibres, mainly from silkworm species Bombyx mori (B. mori) and spider species Nephila edulis, using dynamic mechanical thermal analysis, DMTA. The aim is not only to provide novel data on mechanical properties of silk, but also to relate these properties to the structure and morphology of silk. A systematic approach is adopted to evaluate the effect of the three principal factors of stress, temperature and hydration on the properties and structure of silk. The methods developed in this work are then used to examine commercially important aspects of the ‘quality’ of silk. I show that the dynamic storage modulus of silks increases with loading stress in the deformation through yield to failure, whereas the conventional engineering tensile modulus decreases significantly post-yield. Analyses of the effects of temperature and thermal history show a number of important effects: (1) the loss peak at -60 °C is found to be associated the protein-water glass transition; (2) the increase in the dynamic storage modulus of native silks between temperature +25 and 100 °C is due simply to water loss; (3) a number of discrete loss peaks from +150 to +220°C are observed and attributed to the glass transition of different states of disordered structure with different intermolecular hydrogen bonding. Excess environmental humidity results in a lower effective glass transition temperature (Tg) for disordered silk fractions. Also, humidity-dynamic mechanical analysis on Nephila edulis spider dragline silks has shown that the glass transition induces a partial supercontraction, called Tg contraction. This new finding leads to the conclusion of two independent mechanisms for supercontraction in spider dragline silks. Study of three commercial B. mori cocoon silk grades and a variety of processed silks or artificial silks shows that lower grade and poorly processed silks display lower Tg values, and often have a greater loss tangent at Tg due to increased disorder. This suggests that processing contributes significantly to the differences in the structural order among natural or unnatural silks. More importantly, dynamic mechanical thermal analysis is proposed to be a potential tool for quality evaluation and control in silk production and processing. In summary, I demonstrate that DMTA is a valuable analytical tool for understanding the structure and properties of silk, and use a systematic approach to understand quantitatively the important mechanical properties of silk in terms of a generic structural framework in silk proteins.

Characterising and understanding factors that affect the rate of molecular evolution in proteins has played a major part in the development of evolutionary theory. The early analyses of amino acid substitutions stimulated the development of the neutral theory of molecular evolution, which later evolved into the nearly neutral theory. More recent work has lead to a better understanding of the role selection plays at the molecular level, but there is still limited understanding of how higher levels of protein organisation affect the way natural selection acts. The investigation of this question is the central aim of this thesis, which is addressed via the analysis of selective pressures in secondary protein structures in insects, mammals and fungi. The analyses for the first two groups were conducted using publically available datasets. To conduct the analyses in fungi, genome sequence data from the fungal genus Microbotryum (sequenced in our laboratory) was assembled and annotated, resulting in the development of a number of bioinformatics tools which are described here. The fungal, insect and mammalian datasets were interrogated with regard to a number of structural features, such as protein secondary structure, position of a site with regard to adaptively evolving sites, hydropathy and solvent-accessibility. These features were correlated with the signals of positive and purifying selection detected using phylogenetic maximum likelihood and Bayesian approaches. I conclude that all of the factors examined can have an effect on the rate of molecular evolution. In particular, disordered and hydrophilic regions of the protein are found to experience fewer physiochemical constraints and contain a higher proportion of adaptively evolving sites. It is also revealed that positively selected residues are ‘clustered’ together spatially, and these trends persist in the three taxa. Finally, I show that this variation in adaptive evolution is a result of both selective events and physiochemical constraint.

ATP-dependent chromatin remodellers represent a class of proteins that restructure chromatin through the action of a conserved helicase-like ATPase domain. Remodellers typically have several accessory binding domains alongside the ATPase. These confer target specificity and most commonly recognise histone post-translational modifications. SMARCAD1 is a ubiquitous chromatin remodeller involved with DNA replication and re- pair. It binds directly to PCNA at the site of DNA replication and recruits co-repressor KAP1 in order to silence newly produced chromatin. In contrast to most other chromatin remodellers, SMARCAD1 does not contain several different types of accessory domains. Only two CUE do- mains have been identified in addition to the SMARCAD1 core ATPase domain. CUE domains are a type of helical ubiquitin-binding domain. This thesis presents the findings of an investigation into the structure and function of the SMARCAD1 double CUE domain. The solution NMR structure is presented with results from NMR binding experiments mapped onto the structure. Each CUE domain was found to be an independent helix bundle connected by a dynamic flexible linker. The N-terminal CUE domain, CUE-1, binds ubiquitin and has an adjacent SUMO (a ubiquitin-like protein) binding motif on a protruding extended helix. The C-terminal CUE domain, CUE-2, has a very similar structure to several published CUE domains but does not bind ubiquitin due to a charged substitution at a highly conserved CUE consensus position. The SMARCAD1 double CUE domain binds KAP1 from nuclear extract and is likely to mediate the interaction between SMARCAD1 and KAP1. SMARCAD1 double CUE domain is not involved with PCNA binding.

The paralogous multi-functional adaptor proteins YAP and TAZ are nuclear effectors of the Hippo pathway, a central regulator of developmental organ size control, tissue homeostasis and tumour suppression. YAP/TAZ target the TEAD transcription factor family to promote cell survival and inhibit apoptosis. TEAD proteins contain a DNAbinding domain and a YAP/TAZ interaction domain. PCR analysis of medaka fish TEAD cDNA revealed the presence of alternative TEAD splice-forms with variations at the C-terminus of the DNA-binding domain. Structural analysis indicated the YAPbinding domain of TEAD proteins is folded and globular. NMR spectroscopy showed that the TEAD binding domain of YAP does not contain secondary structure. YAP and TAZ both contain WW domains, which are small protein-protein interaction modules. Two YAP isoforms are known, YAP1 and YAP2 that contain one and two WW domains, respectively. To date, only a single WW isoform of TAZ has been described. PCR analysis of medaka TAZ cDNA identified both single WW and tandem WW isoforms of TAZ. NMR spectroscopy was used to characterise structural, conformational, and peptide binding features of the tandem WW domains from YAP and TAZ. The YAP WW2 solution structure confirms that the domain has the canonical anti-parallel β-sheet WW fold. WW1 of YAP and both WW domains of TAZ undergo conformational exchange. The region linking the two WW domains is flexible and allows interaction of both WW domains with peptides containing single and dual PPxY binding motifs. In addition to YAP and TAZ, tandem WW domains are also present in the core and upstream Hippo pathway proteins Salvador and Kibra. Both proteins contain one atypical WW domain; the tandem WW domains of these two proteins are unstable. Understanding structure and function of Hippo pathway components could contribute to drug development and will also contribute to knowledge of protein folding and interactions.

Membrane proteins represent about one third of all known vertebrate proteins and over half of the current drug targets. Knowledge of their three-dimensional (3D) structure is worth millions of pounds to the pharmaceutical industry. Yet experimental structure elucidation of membrane proteins is a slow and expensive process. In the absence of experimental data, computational modelling tools can be used to close the gap between the numbers of known protein sequences and structures. However, currently available structure prediction tools were developed with globular soluble proteins in mind and perform poorly on membrane proteins. This thesis describes the development of a modelling approach able to predict accurately the structure of transmembrane domains of proteins. In this thesis we build a template-based modelling framework especially for membrane proteins, which uses membrane protein-specific information to inform the modelling process.Firstly, we develop a tool to accurately determine a given membrane protein structure's orientation within the membrane. We offer an analysis of the preferred substitution patterns within the membrane, as opposed to non-membrane environments, and how these differences influence the structures observed. This information is then used to build a set of tools that produce better sequence alignments of membrane proteins, compared to previously available methods, as well as more accurate predictions of their 3D structures. Each chapter describes one new piece of software or information and uses the tools and knowledge described in previous chapters to build up to a complete accurate model of a transmembrane domain.

We review fold usage on completed genomes in order to explore protein structure evolution and assess the evolutionary relevance of current structural classification systems (SCOP and CATH). We assign folds on a set of 150 completed genomes using fold recognition methods (PSI-BLAST, SUPERFAMILY and Gene3D). The patterns of presence or absence of folds on genomes gives us insights into the relationships between folds and how we have arrived at the set of folds we see today. In particular, we develop a technique to estimate the relative ages of a protein fold based on genomic occurrence patterns in a phylogeny. We find that SCOP's `alpha/beta' class has relatively fewer distinct folds on large genomes, and that folds of this class tend to be older; folds of SCOP's `small protein' class follow opposite trends. Usage patterns show that folds with many copies on a genome are generally old, but that old folds do not necessarily have many copies. In addition, longer domains tend to be older and hydrophobic amino acids have high propensities for older folds whereas, polar - but non-charged - amino acids are associated with younger folds. Generally domains with stabilising features tend to be older. We also show that the reliability of fold recognition methods may be assessed using occurrence patterns. We develop a method, that detects false positives by identifying isolated occurrences in a phylogeny of species, and is able to improve genome wide fold recognition assignment sets. We use a structural fragment library to investigate evolutionary links between protein folds. We show that 'older' folds have relatively more such links than 'younger' folds. This correlation becomes stronger for longer fragment lengths suggesting that such links may reflect evolutionary relatedness.