The opportunistic bacterial pathogen <i>Serratia marcescens</i> secretes a ‘chitinolytic machinery’ comprising ChiA, ChiB, ChiC, and Cbp21, that is essential for the efficient utilisation of the extracellular polysaccharide chitin. Secretion of these proteins appears to be a two-step process in <i>S. marcescens</i> with initial export to the bacterial periplasm being followed by a final secretion step across the outer membrane. Successful secretion of the chitinolytic machinery ultimately depends on the products of a four-gene operon, <i>chiWXYZ</i>, which resembles a phage lysis cassette. At the outset of this project the structure, function and interrelationships between the ChiWXYZ proteins was largely unknown and thus a major aim of this work was to provide mechanistic insight into this system. This project therefore details the characterisation of key molecular events in the secretion process. The <i>chiW</i> and <i>chiX </i>genes encode putative holin and endolysin proteins, respectively, and these are essential for secretion of chitinases. Using X-ray crystallography, the structure of full-length ChiX was resolved to 1.34 Å, revealing structural similarities to the lysostaphin-type (LAS) family of peptidases. Purified ChiX was shown to possess L-Ala D-Glu endopeptidase activity, establishing the peptidoglycan cross-link in the periplasm as its substrate. ChiX has no obvious signal peptide, however experiments outlined here provided evidence that the holin ChiW was responsible for ChiX passage to the periplasm. Finally, proteomic data sets for strains overproducing a central regulator of the system, ChiR, were analysed in an attempt to discover hitherto unknown components or substrates of the chitinase secretion machinery.

Paracoccus denitrificans, a model Alphaproteobacteria soil denitrifier, can grow solely on nitrate or nitrite as inorganic nitrogen sources using a specialised cytoplasmic assimilatory nitrate/nitrite reducing pathway; Nas. This growth capability is phylogenetically represented throughout heterotrophic and autotrophic bacteria, plants and fungi. Whilst this metabolism has been extensively studied in the latter two, the regulatory mechanisms by which organoheterotrophic bacteria govern this nitrate-dependant metabolism are less understood. The work conducted here primarily investigated genetic regulation of Nas expression in P. denitrificans. In Gram-negative bacteria, transcription of proteins required to import and reduce nitrate/nitrite to ammonium, for nitrogen assimilation, are subject to dual control; promotion in the absence of ammonium by the general nitrogen regulatory system, NtrBC, and nitrate-induced transcriptional anti-termination by the two-component, NasT-NasS complex. Here, a hypothetical gene, nifR3, conserved with the ntr cluster throughout Alphaproteobacteria, was shown to regulate Nas biosynthesis. We report nifR3 encodes a nitrogen-responsive, tRNA-dihydrouridine synthase required for nasABGHC translation. Genomic deletion of nifR3 from P. denitrificans resulted in the lethal loss of nitrate assimilation and severe deficiency of dihydrouridine in tRNA, restored by genetic complementation of nifR3 in trans. Pure NifR3 harboured an FMN cofactor and reversibly catalysed NADH-dependant reduction of uridine, a physiological important post-transcriptional modification. Native band-shift assays using an isolated tRNA fraction of P. denitrificans identified specific targets of NifR3: mature tRNA transcripts encoding PheGAA, LysUUU and TrpCCA. This novel regulatory role of bacterial NifR3 and tRNA-dihydrouridine formation concerning post-transcriptional fine-turning of protein expression will be discussed throughout this thesis, in addition to the function of several other nitrogen-responsive proteins explored here. Separately, we demonstrated that NarJ, the molybdenum-chaperone for biogenesis of respiratory nitrate reductase, NarG, performs an unprecedented wide-spread maturation role of non-Nar nitrate reductases. Here, we found NarJ is solely responsible for fully assembling the functional assimilatory nitrate reductase, NasC, complete with cofactors, even under aerobic conditions.

The main aims of this project were to use a variety of techniques to study various components of the actin cytoskeleton using Caenorhabditis elegans and Schizosaccharomyces pombe, in addition to developing a novel hydrostatic pressure imaging system. The function and localisation of the C. elegans class I myosins: HUM-1 and HUM-5, were first explored using fluorescent protein fusions and phenotype analysis. The impact of a conserved phosphorylation event on myosin Ie function and localisation was then examined in C. elegans and S. pombe HUM-1 and Myo1 respectively. The results show HUM-1 and HUM-5 are non-essential, with HUM-1 expressed in a variety of tissues whereas HUM- 5 was expressed exclusively in the nervous system. Loss of HUM-1 reduces maximal brood size, inducing delayed embryo release. The conserved serine residue is required for the function and localisation of both HUM-1 and Myo1. Mutations of the residue altered HUM-1 localisation in a tissue specific manner and abolished Myo1 membrane association in addition to inducing spore formation. In vitro and in vivo studies were next undertaken to explore the impact of fluorescent protein fusions and temperature sensitive mutations upon the the stability and function of S. pombe tropomyosin. While both Ts mutants, Cdc8-27 and Cdc8-110, had a two-step thermal unfolding transition lower than that of wild-type Cdc8, fluorescent fusions did not impact stability. Amino- terminal fusions however, mimic acetylation whereas carboxyl terminal fusions abolished polymerisation, localisation and functionality. The final part of this project used a methodical approach to develop a hydrostatic imaging system that could be used to follow fluorescently labelled protein dynamics in a live cell context. Preliminary data revealed the application of 100 bar of pressure induced a cell cycle delay in both S. pombe and Candida albicans, however the current limitations of the system make it incompatible with fluorescence microscopy.

The development of ligands for protein surfaces to inhibit protein-protein interactions (PPI)s is challenging, as protein surfaces often lack the clefts and pockets associated with traditionally druggable targets like enzyme active sites. One way in which protein surfaces can be targeted is by the use of protein surface mimetics, whereby a multivalent scaffold is functionalised with many binding groups on its periphery in order to achieve high affinity protein recognition. One such scaffold is a ruthenium(II) tris (bipyridine)s (Ru(II)(bpy)3). The work in this thesis aimed to further develop these Ru(II)(bpy)3 protein surface mimetics; gaining information as to how they interact with proteins, looking at new ways of achieving high affinity protein surface recognition and the development of new applications for these molecules. In Chapter 2 an indepth study of the binding of two Ru(II)(bpy)3 complexes to a model protein, cytochrome c, is presented, looking at the thermodynamic and electrostatic contributions to binding as well as using protein NMR to elucidate the binding site. In Chapter 3 the development of dynamic combinatorial chemistry (DCC) scaffolds based on Ru(II)(bpy)3 complexes and tetraphenyl porphyrins was explored as a potential avenue for new receptor design, enabling the development of biologically compatible DCC systems, prime for protein ligand discovery. Chapter 4 presents another avenue for using the Ru(II)(bpy)3 complexes; using an array approach to discriminate between different protein.

This thesis presents the development of an electrothermally actuated microgripper for the manipulation of cells and other biological particles. The microgripper has been fabricated using a combination of surface and bulk micromachining techniques in a three mask process. All of the fabrication details have been chosen to enable a tri-layer, polymer (SU8) - metal (Au) - polymer (SU8), membrane to be released from the substrate stress free and without the need for sacrificial layers. An actuator design, which completely eliminates the parasitic resistance of the cold arm, is presented. When compared to standard U-shaped actuators, it improves the thermal efficiency threefold. This enables larger displacements at lower voltages and temperatures. The microgripper is demonstrated in three different configurations: normally open mode, normally closed mode, and normally open/closed mode. It has-been modelled using two coupled analytical models - electrothermal and thermomechanical - which have been custom developed for this application. Unlike previously reported models, the electrothermal model presented here includes the heat exchange between hot and cold arms of the actuators that are separated by a small air gap. A detailed electrothermomechanical characterisation of selected devices has permitted the validation of the models (also performed using finite element analysis) and the assessment of device performance. The device testing includes electrical, deflection, and temperature measurements using infrared (IR) thermography, its use in polymeric actuators reported here for the first time. Successful manipulation experiments have been conducted in both air and liquid environments. Manipulation of live cells (mice oocytes) in a standard biomanipulation station has validated the microgripper as a complementary and unique tool for the single cell experiments that are to be conducted by future generations of biologists in the areas of human reproduction and stem cell research.

Heme peroxidises catalyse the H2O2-dependent oxidation of substrates in a two-step process, through formation of two oxy-ferryl intermediates known as Compound I and Compound II. Despite the considerable effort worldwide, important aspects about the reactivity of these enzymes are still to be clarified. Amongst all, the determination of the nature of the Fe-O bond in the oxy-ferryl intermediates, as well as the mechanism by which protons are delivered to the oxy-ferryl species during turnover, are of highest relevance. In this thesis, high resolution crystal structures of both Compound I and Compound II intermediates in two heme peroxidases, cytochrome c peroxidise (CcP) and ascorbate peroxidise (APX), are presented. In order to rule out the photoreduction arising from X-ray exposure during data collection, which causes alteration of ferryl intermediate structures, a multicrystal method has been employed. Results indicate that Compound I, with an Fe-O distance of 1.63 Å for CcP and 1.73 Å for APX, is consistent with an unprotonated oxy-ferryl species (FeIV=O), whereas Compound II, with an Fe-O bond length of 1.83 Å and 1.84 Å for CcP and APX respectively, is consistent with a protonated oxy-ferryl species (FeIV-OH). Also presented in this thesis is the 2.40 Å structure of resting ferric CcP at room temperature obtained, for the first time, by neutron crystallography. This study allowed to establish the location of individual, exchangeable hydrogen atoms thus revealing the protonation states of several key active site residues in the distal (Arg48, Trp51, His52) and proximal (His163, Trp191, Asp235) heme regions. This information was used to revise the reaction mechanism of heme peroxidises and also to infer a possible delivery pathway of protons during turnover. All together, these data not only clarify long-standing inconsistencies on the nature of the oxy-ferryl species, but they also provide new insights into The reaction mechanism of heme peroxidises and provide important information which May apply to other categories of heme enzymes such as the cytochromes P450 and NO synthases.

The Actin Cytoskeleton is essential for Eukaryotic life and is involved in a diverse range of cellular functions. Cyclase Associated Protein (CAP) was first identified in yeast as a regulator of the CYR1 Adenylate Cyclase. Subsequently CAP family members have been identified in every Eukaryotic kingdom and have also been implicated in the regulation of Actin dynamics. It has been proposed that the CAP family promotes the recycling of Actin monomers by cooperating with members of the Profilin and Actin Depolymerising Factor families. This study represents an attempt to investigate the function and developmental role of AtCAP1, an Arabidopsis member of the CAP family. Arabidopsis thaliana is widely used as a model for higher plant development due to its small sequenced genome and the availability of a wide variety of mutants. The elimination of AtCAP1 expression results in a distinct developmental phenotype. Early characteristics include the absence of the root hair collar, reduced root hair initiation and extension. Later onset phenotypes include reduced plant height and a severe reduction in pollen viability. In vivo studies of the CAP-deficient cytoskeleton reveal a distinct loss of fine filamentous Actin and the appearance of dense Actin aggregates. Cell expansion is also significantly reduced. The interaction between AtCAP1 and F-Actin is demonstrated in vitro by a biochemical interaction study and a filament bundling activity is suggested. The multimerisation of AtCAP1 and its interaction with other components of the Actin Cytoskeleton are demonstrated via Yeast Two Hybrid interactions. It is concluded that AtCAP1 is essential for the organisation of the plant cells F-Actin network and that this in turn is required for correct growth and development. It is hypothesised that AtCAP1 function is mediated by regulating the interaction between F-Actin and other Actin-interacting proteins.

The human homologue of the Drosophila Discs Large Tumour suppressor, hDlg, has been subject to various speculations concerning its role in the cell, with studies indicating roles in establishing and maintaining cellular polarity, as well as in controlling cell proliferation. The biochemical mechanism by which it might act in executing either function have, however, remained ambiguous. In this study we show that phosphorylation is a major posttranslational modification of the protein, affecting both location and function. We show that hDlg is phosphorylated both by the MAPKs and the CDKs, and both groups of kinases affect different aspects of the protein's behaviour. Post osmotic shock, the phosphorylation of hDlg by JNK leads to its accumulation in vesicular structures which we identify as endosomes, while its phosphorylation by the p38 MAPK in addition to relocating it to sites of cell-cell contact, also makes it more susceptible to degradation by the HPV18 E6 oncoprotein. Secondly, we show that hDlg is differentially regulated during the cell cycle, with each stage of the cell cycle leading to a different localisation of the protein, including its accumulation at the mitotic spindle in the M phase, as well as at the midbody during cytokinesis. We show also that the protein is phosphorylated by the Cyclin Dependent Kinases (CDK) 1 and 2, in a cell-cycle dependent manner on two sites - serine 158 and serine 442, and that these phosphorylations render the protein more stable and less susceptible to ubiquitination. Finally we show that hDig phosphorylated on serine 158 and on serine 442 is largely nuclear, and that both the HPV18 E6 and the HPV16 E6 oncoproteins, target this nuclear form for degradation. These findings help us understand the processes that regulate hDlg and how these modifications of the protein might contribute to its growth-regulatory function in the cell.

The neuregulins (NRGs) are ligands for receptor tyrosine kinases of the ErbB family. The NRG gene family consists of four members, NRG 1- 4 encoding different isoforms due to alternative splicing. NRG signalling has been implicated in normal development and in the pathology of many diseases. Using immunohistochemical staining of tissue arrays, we detected that NRG 1 a and 1 ~ localised to the cell nuclei of a range of normal and human cancer tissues. The ~3 isoform of NRG 1 localised to two subnuclear compartments: nucleoli and spliceosomes. We tagged NRG 1 ~3 with photoactivateable GFP and demonstrated that the fusion protein re-Iocalised from nucleoli to spliceosomes over a ninety minute period. Using wild type NRG 1 ~3 and its two mutants which localised exclusively to spliceosomes or to nuc1eoH, we explored the possible functions of intranuclear NRG 1 ~3 in each of these compartments separately. We showed using an array capable of detecting 42 receptor tyrosine kinases that wild type NRG 1 and a mutant exclusively local ising to spliceosomes increased phosphorylation and/or expression of the ErbB4 and ErbB2 receptors. Using a transcriptomic analysis the same two constructs induced expression of messenger RNA of Heat Shock Protein 70B' and we confirmed its increased expression at the protein level using western blot analysis. This data supports the hypothesis that intranuclear NRGs could activate receptor signalling and alter gene expression when localised in spliceosomes.

Enabling the entry of therapeutic antibodies into cells for the modulation of intracellular targets has the potential to greatly expand the treatment options for diseases, such as cancer. A class of promising delivery vehicles for therapeutic antibodies is the cell-penetrating peptides (CPP), which are capable of entering cells spontaneously and promote the cellular uptake of conjugated biomolecules, such as antibodies. Demonstration of the CPP-mediated delivery of a functional protein was performed using the well-established example of p27, a cyelin-dependent kinase inhibitor, fused to the Tat delivery domain of the human immunodeficiency virus (HIV). Concomitant with established data, the Tat-p27 fusion protein entered cells, induced G1 cell cycle arrest in MCF-7 cells and induced HepG2 cells to undergo cell scattering and cytoskeletal alterations.