2000 D.Phil Symposium Programme and Abstracts

October 4th

9.00 Introduction  9.15 Stephen Lee  9.30  Sophia Lin 9.45  Gillian McVey  10.00  Catherine Napper  10.15 Jo Owen


11.00  David Manners  11.15 Sai Pathmanathan 11.30  Thil Batuwangala  11.45  Samuel Bouyain  12.00  Migena Bregu  12.15  Mira Aroyo

12.30 - 1.30 LUNCH

1.30  Aileen McGettrick  1.45  Simona Romagnoli  2.00  Sachchidanand  2.15  Robert Sidell  2.30  Nicholas Smith  2.45  Steven Porter

October 5th

9.15  Sujit Dutta  9.30  Henrique Ferreira  9.45  Simon Holton  10.00  Mayumi Kojima  10.15  Jeremy Lack


11.00  Tobias Ulmer  11.15  Kay Wan  11.30  Suet-Mein Tan  11.45  Peter Teriete  12.00  Nathalie Uzcategui Cuello 12.15  Martin Ulmschneider

12.30 - 1.30 LUNCH

1.30  Anna Warren  1.45  James Willcocks  2.00  Navita Rampersaud  2.15  Shaharum Shamsuddin  2.30 Mark Bushell  2.45

Eun-Kyung Kim   3.00  Presentations

The DPhil Symposium is sponsored by MERCK.   Book tokens are awarded for the best presentations.  Drs Margery Ord and Lloyd Stocken generously award one prize and the other is by courtesy of our sponsors.

Presenter: Mira Aroyo

Supervisors:  Dr. F.X.Barre, Professor D.J.Sherratt

The Role of the C-terminus of FtsK in Xer Recombination

In Escherichia coli Xer site-specific recombination acts at the recombination site dif, located in the replication terminus region of the chromosome, to resolve chromosomal dimers to monomers. Recombination at dif is dependent on the initiation of cell division, requires dif to be in its natural position and requires the presence of the C-terminal domain of FtsK. FtsK is a protein acting late in the cell division pathway that exhibits homology to the Bacillus Subtilis DNA transfer protein SpoIIIE. FtsK appears to have three domains. The N-terminal transmembrane domain is essential for cell division and localises the protein to the septum in a FtsZ-dependent manner. The N-terminus is linked to the C-terminus by a proline-glutamine rich region. The C-terminal domain contains a presumptive ATP binding site and is implicated in chromosome segregation and Xer recombination at dif on the chromosome and on a supercoiled plasmid. The nature of the mechanism by which the C-terminus of FtsK acts to activate Xer recombination at dif is yet unknown. It could act directly on the DNA or the two recombinases, or by recruiting other factors.

We have constructed N- and C-terminal truncations of FtsK in order to determine the minimum part of FtsK necessary to induce recombination between two dif sites. The truncations were tagged with a FLAG peptide in order to verify protein expression. The ability of the truncations to complement ftsK- phenotype for plasmid resolution was tested in vivo. The minimum domain necessary was then used to assay the integration of a dif containing plasmid into the chromosomal dif site. The FtsK truncation supported integration even when the dif site was translocated away from the terminus to the lacZ position, indicating that FtsK does not have to be septum located to support recombination. This led us to attempt to assay FtsK’s mechanism of action in vitro and to do that we purified the truncated protein. The truncation used was shown to posses a DNA-independent ATPase activity.


Presenter:  Thil Batuwangala

Supervisor:  Prof. E. Yvonne Jones

Structural studies on receptors from the Drosophila Wingless and Hedgehog signalling pathways

Smoothened and D-Frizzled-2 are segment polarity genes from Drosophila melanogaster. Drosophila has a segmented body plan and the role of segment polarity genes is to pattern the parasegments by maintaining the definition of the anterior and posterior boundaries of each parasegment. Segment polarity genes code proteins that are components of cell signalling, from secreted signalling molecules, cell surface receptors, signal transducers to transcription factors. Smoothened and D-Frizzled-2 proteins are cell surface receptors of the Hedgehog and Wingless signalling pathways respectively. These two pathways act reciprocally to establish an intercellular signalling circuit stabilising the anterior-posterior boundary of the parasegments. Smoothened and D-Frizzled-2 have a conserved cysteine rich domain (CRD). It is predicted that the CRD, a conserved region of ~120 amino acids with ten invariantly spaced cysteines is somehow involved in ligand interactions.
The goals of my project are to use molecular biology and crystallographic techniques, in an attempt to delineate modes of receptor-ligand engagement in wingless and hedgehog signalling. At present the extra-cellular domain of smoothened is being expressed in stable cell lines of SF9 insect cells with expression of recombinant protein at 5-10 mg/l of tissue culture.  Attempts are also being made to express recombinant D-Frizzled-2 using the same expression system in order to determine the structural basis of the functioning of these two receptors and in particular, to understand the role of the CRD.


Presenter:  Samuel Bouyain

Supervisor:  Kurt Drickamer

Selectin/mannose-binding protein chimeras to investigate the selectin extended ligand-binding site

The selectins are carbohydrate-binding proteins involved in recognition between endothelia and leukocytes. Interaction between selectins and their ligands has been investigated by introducing segments of the E-selectin carbohydrate-recognition domain into the homologous rat mannose-binding protein. This approach leads to the identification of selectin regions that are sufficient for ligand binding and structures of these molecules in complex with the tetrasaccharide sialyl Lewis X ligand can be determined. Transfer of three lysine residues from E-selectin is sufficient to create a sialyl Lewis X-binding chimera. In one type of improved chimera, the extra Ca2+-binding site in mannose-binding protein was removed and fluorescence was used to demonstrate that Ca2+ binding is first order. Addition of a second loop of E-selectin makes the chimera more selectin-like but this loop does not assume its correct conformation. In order to create better mimics of E-selectin, two additional mutations have been introduced in and near the loop and are being characterised functionally and structurally. E- and P-selectin also bind galactose 3-sulfate-containing sulfatides. The chimera approach reveals that the three lysine residues are sufficient for sulfatide binding. A similar approach is being used to investigate E- and P-selectin binding to specific glycoprotein ligands.


Presenter:  Migena Bregu

Supervisor:  Dr S.D.Colloms, Professor D.J.Sherratt

The role of accessory sequences in Xer recombination at plasmid site psi

Xer site-specific recombination acts at the psi site of plasmid pSC101 to convert plasmid multimers to monomers, thereby ensuring heritable stability of pSC101. Recombination at psi is preferentially intramolecular and gives products of unique topology. This topological selectivity and specificity is imposed by accessory proteins PepA and ArcA acting at accessory DNA sequences adjacent the core recombination site to form a synaptic complex of a precise geometry. Strand exchange at psi is sequential. XerC catalyses the first strand exchange to make Holliday junction intermediates and XerD catalyses the second strand exchange.

In order to investigate the role of accessory sequences in the formation of the synaptic complex, we have constructed a range of psi reporter plasmids containing different combinations of recombination sites with and without accessory sequences. Our results demonstrated that  two sets of accessory sequences are an absolute requirement for imposing topological selectivity on recombination. Psi reporter plasmids which do not have both sets of accessory sequences, gave rise to products of mixed topology in a PepA independent manner. To ascertain the effect of accessory sequences on the order of strand exchange at psi, we have created hybrid recombination site ëAccDCí by placing accessory sequences next to the XerD binding site, as opposed to the normal psi site ëAccCDí in which accessory sequences are adjacent to the XerC binding site. A plasmid with two ëAccDCí sites gave rise to the same 4-noded catenane produced by plasmids with normal sites. Plasmids containing one normal psi site ëAccCDí psi and one ëAccDCí psi site gave 3-noded knots. The order of strand exchange in these substrates has been determined. Results suggest that XerD catalyses the first strand exchange on the substrate with two hybrid ëAccDCí psi sites, which is the reverse of the normal order of strand exchange at psi.


Presenter:  Mark Bushell



Presenter:  Sujit Dutta

Supervisor: Dr K.A.Watson, Dr. S.J.H. Ashcroft, Prof. L. N. Johnson

Expression and Purification of the First Nucleotide Binding Domain of the Sulphonylurea Receptor

Pancreatic ?-cell ATP-sensitive K+-channels play an important role in insulin secretion. They are composed of heteromultimers of sulphonylurea receptors (SUR) and inwardly rectifying potassium channel subunits (Kir).

SUR forms the regulatory unit of K+-ATP channels and possesses two intracellular nucleotide binding domains, NBD1 and NBD2. The focus of this study has been on NBD1 with the ultimate goal to solve its 3-dimensional structure using X-ray crystallography and perform biochemical studies to gain a better understanding of its role in ATP binding and hydrolysis.

Before attempting to crystallize NBD1, it was necessary to establish a suitable expression system in Escherichia coli that would yield sufficient of amounts soluble protein. A number of expression systems were tested. Fusion of NBD1 to Hexa Histidine and Intein all failed to yield adequate amounts of soluble protein. However, fusion to GST and MBP dramatically increased solubility and has led to production of large amounts of fused product for further biochemical and structural studies.


Presenter:  Henrique Ferreira

Supervisor/co-presenter: Lidia Arciszewska, David Sherratt

Coordinated catalysis in Xer site-specific recombination

The Xer site-specific recombination machine, which functions in the stable inheritance of the Escherichia coli chromosome, consists of a heterotetramer containing two monomers of XerC and two monomers of XerD bound to DNA. This nucleoprotein structure functions in a controlled series of events that lead to strand exchange between two DNA recombination sites. Catalysis is coordinated by each recombinase controlling its partners activity. In an attempt to understand the molecular basis of recombinase machine assembly and subsequent coupling of
recombinase catalysis, we have constructed chimeras of XerC and XerD and studied their properties in partial and complete recombination reactions.  Experiments using Holliday junction (HJ) recombination substrates show that swapping the extreme C-termini of the two proteins has a profound effect on the direction of the HJ resolution. These observations confirm the importance of cyclic "donor-acceptor" interactions between the recombinase molecules that involve the extreme C-termini of both recombinases. Studies with chimeras in which the entire N-terminal domains have been exchanged indicate that interactions involving these regions also have a major impact on the Xer recombination outcome.


Presenter:  Scott Goodall

Supervisors:  Prof. A Watts and Dr D Middleton

Probing the Structure and Dynamics of Acetylcholinesterase Inhibitors in their Binding Site using Solid State Nuclear Magnetic Resonance

Acetylcholinesterase (AChE) is a highly efficient enzyme, the principal role of which is to catalyse the hydrolysis of the neurotransmitter acetylcholine (ACh) to acetic acid and choline. AChE inhibitors are used in chemical warfare, as pesticides and in the treatment of Alzheimer’s Disease, glaucoma and Myasthenia Gravis. Knowledge of the mechanisms through which the enzyme interacts with inhibitors could provide a basis for novel inhibitors with increased potency and selectivity, including species-specific pesticides. In addition, information about the binding site of AChE may help in understanding the recognition of ACh by proteins such as the acetylcholine receptor (AChR), which is similar in the abundance of aromatic residues in the region of the binding pocket.
Solid state nuclear magnetic resonance (SSNMR) is a novel approach to the investigation of AChE and its inhibitors, with the potential to determine the absolute conformation of inhibitors within the binding site and the position of inhibitors with respect to the protein.
The progress made to date towards determination of the conformation of a potent AChE inhibitor when bound to the enzyme will be discussed, including synthesis of isotopically labelled inhibitors and initial NMR experiments on a AChE-inhibitor complex.


Presenter:  Simon Holton

Supervisor:  Dr Jane Endicott

The Cell Cycle and Suc1

Schizosaccharomyces pombe Suc1 is a member of the Cks family of proteins, an ubiquitous family of small proteins essential to the eukaryotic cell cycle.  Cks proteins bind to cyclin-dependent kinases (CDKs) and regulate their activity.  The mechanism of this regulation is not clearly understood.  Structure based sequence alignments of Cks proteins reveal two highly conserved surface regions; a hydrophobic patch that has been shown to bind CDK2, and an 'anion-binding' site consisting of a cluster of positively-charged residues. The function of this 'anion-binding' site is unknown, but has been proposed to mediate Cks-phosphoprotein interactions.

A purification strategy using recombinant protein has been developed to allow the isolation of a ternary complex of Thr160pCDK2-cyclinA-hsCKS1.  Compared to Thr160pCDK2-cyclinA this complex shows an approximate 5 fold enhancement of the phosphorylation of cdc27, a component of the anaphase promoting complex.  Two Suc1 'anion-binding' site mutants have been expressed, purified and crystallised.  X-ray crystal structures of these mutants to resolutions of 2.50Å and 2.90Å show that both are isomorphous with wild type Suc1.  By combining these in vitro effects of Cks1 on CDK  kinase activity with the Suc1 mutant crystallographic data we are investigating the importance of the Cks 'anion binding' site on CDK activity.

Presenter:  Eun-Kyung Kim


Saccaromyces Cerevisiae, a unicellular eukaryote, could survive in the constraints of a dynamic ecosystem through constant adaptation

Saccaromyces Cerevisiae can initiate the developmental programme of Pseudohyphal differenciation and invasive growth in response to starvation for nitrogen. We here present the first demonstration that glucose starvation  induce Pseudohyphal and invasive growth and glucose is essential for low nitrogen induced Pseudohyphal differentiation unlike other carbon sources. This low glucose induced Pseudohyphal formation requires STE MAPK cascade(Ras2p,Ste12p, Ste20p) and cAMP pathway(Ras2p, Flo8p,Sch9p) through SNf3p glucose sensor and the two irrespective pathways converge on flo11 gene promotor,suggesting that they are  the main means of signaling for Pseudohyphal differenciation regardless of the type of limiting nutrients. Diploid invasive response, however, has distinct genetic requirement from diploid surface pseudohyphal differenciation. Ras2p, Gpr1, Sch9p and Snf3p are not required for that. Therefore, we propose that the yeast cells are able to response flexibly to prevailing conditions by influencing the degree of substrate invasion versus surface pseudohyphal differenciation.

Presenter:  Mayumi Kojima

Supervisor:  Dr. R. B. Sim

Regulation of MBL-MASP of the lectin pathway of the complement system

Serum mannose-binding lectin (MBL) is a collectin, which binds to carbohydrates and activates the lectin pathway of complement via its associated serine proteases, MASP1 and MASP2.  These proteases are structurally similar to one another as well as to C1r and C1s of the C1 complex of the classical pathway.  Although the most recently discovered, the lectin pathway is thought to predate the classical and alternative pathways of complement activation, hence the regulation of its first component, the MBL-MASPs complex, is of much interest.

A serpin, C1-inhibitor, is the only physiological regulator of the C1 complex of the classical pathway: it acts as a suicide substrate, forming covalent complexes with activated C1r and C1s, and the resultant C1-inh-C1r-C1s-C1-inh complexes dissociate from C1q.  The target-bound C1q can then recruit more C1r and C1s from plasma or effect other functions such as phagocytosis.

To date, two inhibitors have been implicated in regulation of MASP activities in vivo.  We show here that C1-inh also forms covalent complexes with both MASP1 and MASP2 using Western blotting.  The same complexes were observed when MBL-MASPs mixture was radioiodinated and incubated with serum, indicating the likely in vivo significance of the finding.  Studies were also done to investigate the fate of such complexes, and evidence indicates that C1-inh-MASP complexes also dissociate from MBL.  No significant interaction between MASPs and a plasma protease inhibitor a2-macroglobulin (a2M) was observed, contrary to previous reports.

Presenter:  Jeremy Lack

Supervisor:  A. Kristina Downing,  Iain D. Campbell

The Solution Structure and Surface Properties of the Third Transforming Growth Factor-ß-like Domain (TB3) of Latent Transforming Growth Factor-ß-Binding Protein-1 (LTBP-1)

Transforming growth factor-ß binding protein-1 (LTBP-1) is a critical extracellular matrix protein (ECM) protein.  The third transforming growth factor-ß-like domain (TB3) of LTBP-1 covalently binds to the latent associated pro-peptide (LAP) homodimer.  This protein-protein interaction is mediated by an unusual exchange of an internal disulphide bond of TB3 to form two inter-protein disulphide bridges to LAP.  LAP, the propeptide of transforming growth factor-ß1 (TGF-ß1), non-covalently sequesters TGF-ß1.  The association of LTBP-1 with LAP and TGF-ß1 forms the large latent complex, a critical regulatory step of TGF-ß1, mediating excretion from the cell, localisation to the ECM and release of TGF-ß1 to the cell surface.
The structure of TB3 has been determined using high resolution solution heteronuclear nuclear magnetic resonance (NMR) spectroscopy.  TB modules have 4 disulphide bonds, pairing in a 1-3, 2-6, 4-7, 5-8 pattern.  However, unlike the other known structure of a TB module, TB6 of fribrillin-1, TB3 of LTBP-1 has a two amino acid insertion in its sequence: Phe-50 and Pro-51.  This insertion is only found in TB modules binding LAP.  The structure has revealed that the FP insertion mediates the protein-protein interaction by forcing the solvent exposure of the 2-6 disulphide bond, such that it becomes available to bind the two Cys-33 of the LAP homodimer.
Moreover, the surface properties of TB3 potentially have an important role in the interaction.  TB3 of LTPB-1 has two large surface hydrophobic patches, which are not found on TB6 of fribrillin-1.  These hydrophobic surfaces may be important in the binding of TB3 to LAP.  Another structural difference between TB3 and TB6 is the localisation of five negatively charged residues surrounding the exposed disulphide bond of TB3, creating a large negative electrostatic surface potential which might also have a role in LAP binding.


Presenter:   Stephen Lee

Supervisor:   Prof. D I Stuart

Structural and Functional Analysis of Vitronectin Receptor: avß3 integrin

Integrins are heterodimers made of a and ß subunits.  They are type I transmembrane glycoproteins, consisting of large extracellular domains, followed by classic hydrophobic transmembrane segments and generally short cytoplasmic tails.  The extracellular N-terminal regions of the a and ß subunits associate to form the adhesive ligand binding site, whilst their C-terminal cytoplasmic tails interact with cytoskeletal and intracellular signalling proteins. A fundamental property of integrins is their ability to transmit signals in a bi-directional fashion across the cell membrane.  Response to cytoplasmic signals can regulate the functional activity and surface expression levels of receptors (termed inside-out signalling), whilst binding to extracellular ligands can induce changes in cell shape, gene expression and the activation state of a cell (termed outside-in signalling).  The ability to integrate information between the cytoskeleton and the ECM gave the name to the superfamily, and enables precise co-ordination of adhesion with other cellular events. The goal of this project is to express, purify and to determine the structure of vitronectin receptor (avß3 integrin).  In particular, we are interested in the extracellular domain (ECD) of this receptor molecule where ligand:receptor interactions occur.  Different fragments of this protein will be made and the best candidate chosen for crystallisation trials to elucidate its structure and function.

Presenter:  Sophia Lin

Supervisor:  Dr. Catherine Pears

The role of cell cycle and early developmental gene expression in cell fate decisions in Dictyostelium

In Dictyostelium, cell type-specific differentiation is correlated with the growth conditions in vegetative cells, as well as the position in the cell cycle at the onset of starvation. Cells grown in medium without glucose [G(-)] is found to sort to the prestalk region compared to those cells grown in the glucose-containing medium [G(+)]. Cells which start development in the S/M and early G2 phase exhibit a strong tendency to sort to the stalk, while cells in the middle and late G2 phase tend to become spores. In this study, the analysis of differential expression of early developmental genes under these two conditions was attempted. Northern blot analysis was performed to investigate the cell-cycle dependent regulation of early developmental gene expression during development of synchronised S/M and G2 phase cells, and cells grow under different growth conditions. The discoidin I gene showed an earlier and a higher level of expression in G2 phase cells. Cells expressing discoidin I::lacZ showed greater b-galactosidase enzyme activity in G(-) conditions than in G(+) conditions in vegetative growth, and the increased activity in later developmental stage seen in G(+) cells was not found in G(-) cells. Staining experiments suggested localisation in the spore region. These results demonstrate clear differences in the expression of early developmental genes in populations of cells with tendencies to different cell fates. Ddcdk8, a putative component of the cell-cycle machinery, was further studied in order to gain insight of the regulatory mechanism by which the cell cycle affects the cell fate decision, via differential expression of early genes.

Presenter:  Aileen McGettrick

Supervisor:    Dr. P.A. Handford

Molecular Effects of Calcium Binding Mutations in Fibrillin-1

Fibrillin-1 is a major structural component of 10-12nm extracellular matrix microfibrils and has a striking modular organisation which includes 43 calcium binding epidermal growth factor-like (cbEGF) domains. Mutations in the human fibrillin-1 (FBN-1) gene cause the Marfan syndrome (MFS), an autosomal dominant disease of connective tissue. A significant group of reported FBN-1 mutations are predicted to reduce calcium binding to cbEGF domains by removal of a specific side chain ligand for calcium. These mutations occur in two different protein domain contexts, either in a cbEGF preceded by a transforming growth factor ß1 binding protein-like (TB) domain or in a cbEGF preceded by another cbEGF. In this study different proteases have been used to probe structural changes caused by MFS and protein-engineered calcium binding mutations in the TB6-cbEGF34 region of fibrillin-1. N-terminal sequence analysis of domain pairs digested in the presence and absence of calcium shows that domain context influences the structural consequences of such mutations and in particular increased proteolysis is observed in domain constructs containing tandem repeats of cbEGF domains. Preliminary studies of triple cbEGF domain constructs from this region suggest that the effects of calcium binding mutations are not localised to the mutant domain but may influence proteolytic susceptibility of flanking domains. The structural basis for these observations is under investigation.


Presenter:  Gillian McVey

Supervisor:  Dr A.J. Day

Functional and Structural Studies on Hyaluronan-Binding Proteins

Hyaluronan (HA) is a high molecular weight polysaccharide which is found in all tissues and body fluids of vertebrates, as well as in some bacteria. It has a key role in the formation and the stability of the extracellular matrix (ECM) of cartilage, brain, and skin via its interactions with HA-binding proteins and receptors.  Most HA-binding proteins contain a common structural domain of around 100 amino acids termed a Link module and, collectively, these proteins make up the Link module superfamily which includes the  ECM proteins aggrecan, versican, neurocan and link protein. In cartilage for example, aggrecan and link protein form massive multi-molecular complexes with hyaluronan  which provide the tissue with its load-bearing properties. Link protein is comprised of an N-terminal immunoglobulin (Ig) fold followed by a pair of contiguous Link modules and this configuration of modules forms the G1 domain of aggrecan. Link protein is thought to interact with aggrecan via the Ig fold and both proteins bind to HA via the Link module pair. Likewise, versican and neurocan possess a G1 domain and both are thought to play similar roles in the formation and stability of the ECM of skin and brain respectively.

The purpose of this project is to express regions of HA-binding proteins for use in functional and structural studies. The Link module pairs of versican, aggrecan and link protein have been expressed in a bacterial system and purified to homogeneity. Following refolding, functionally active link protein has been identified using plate assays to detect HA-binding. In addition, a Drosophila expression system is in use for the expression of full length link protein and the G1 domains of aggrecan and versican. Preliminary results indicate that link protein is being expressed. Functional and structural investigations involving these proteins should lead to a greater understanding of HA-binding protein interactions in the ECM.

Presenter:  David Manners

Supervisors:  P. Styles, D.J. Taylor

Combined MRI and 31P MRS facilitate the study of degenerative muscle disease

The three main classes of idiopathic inflammatory myopathy are dermatomyositis (DM), polymyositis (PM) and sporadic inclusion body myositis (IBM). In common with all cases of degenerative muscle disease, the most common finding by MRI are lipomatosis (fatty infiltration of muscle) and oedema-like processes. Both increase intensity in T2 images relative to muscle water signal. To quantitate the biochemical events occurring in disease processes, we have developed a protocol to separate these effects on the basis of tissue chemical shift and relaxation time. This has been combined with a 31P MRS examination, allowing muscle energy metabolism to be monitored at rest and during exercise. The protocol has been used to analyse muscle function and inflammation in 28 patients with idiopathic inflammatory myopathy. Results indicate an abnormality of oxidative phosphorylation in DM and PM (but not IBM) which improves with treatment by anti-inflammatory steroids. T2 relaxation times are also lower in patients on treatment, but remain above control values.

Presenter:  Jeevani Mantotta


Deletion of membrane-bound receptor protein in Rhodobacter spaeroides gives rise to a superswarming phenotype

A gene coding for a classical membrane spanning chemoreceptor (mcpJ) was identifiied in a region of Rhodobacter spaeroides  DNA unlinked to either for the two previously identified operons.  Insertional inactivation and deletion of this gene in a range of backgrounds gave rise to mutants that formed further on semisolid agar plates (swarm plates) in response to 100 µM conentrations of a range of attractants ie. had a 'superswarming' phenotype.

Tethering of the deletion mutants showed normal responses to chemoattractants. However, methanol release studies showed that deleting mcpJ leads to a marked decrease in the amount of methanol released by the cells, suggesting that McpJ mediates a repellent response under normal circumstances.

Presenter:  Catherine Napper

Supervisor:  Maureen Taylor

Ligand Binding By The Macrophage Mannose Receptor

The macrophage mannose receptor is a C-type lectin that recognises carbohydrate structures on pathogenic microorganisms and harmful glycoproteins. The receptor is involved in the innate immune system through its ability to bind to and mediate internalisation of invading pathogens. The goal of my project is to address two questions: 1. Does the mannose receptor enhance uptake and processing of glycosylated antigens and thus have a role in the adapative immune response? 2. What is the conformation of the mannose receptor and how are the domains arranged in order to select particular carbohydrate ligands?
The role of the mannose receptor in antigen processing and presentation is being investigated using a model system developed to compare processing and presentation of glycosylated and non-glycosylated proteins.
An understanding of the conformation of the mannose receptor is needed to visualise how the receptor selects its carbohydrate ligands. The extracellular region of the mannose receptor comprises ten different domains that must be oriented to accommodate a variety of ligands. Hydrodynamic analysis of the extracellular region of the receptor has been performed, using techniques such as analytical ultracentrifugation and analytical gel filtration. Sedimentation and diffusion coefficients obtained from these methods have been used to calculate size and shape parameters for the receptor. Fitting data to computer-generated models reveals that the receptor assumes an extended conformation with some contacts between neighbouring domains.  A U-shape, suggested in some earlier studies, can be ruled out.


Presenter:  Jo Owen

Supervisor:  Iain Campbell, Dave Staunton

Expression and Characterisation of an EGF Pair from Fibulin-1

Fibulin-1 is a member of a new family of extracellular matrix (ECM) proteins. It has suggested functions in formation and stabilisation of elastic fibres and basement membranes due to its ability to bind other ECM proteins such as fibronectin. Fibulin-1 is primarily composed of two recognized sequence motifs; complement repeats and EGF-like domains, with the major part of the protein comprising 9 consecutive EGF-like domains.
The binding site for fibronectin has been mapped to the EGF-like repeats 5&6, which also appear to be responsible for self-association and calcium binding. This pair of EGF-like domains have been cloned, expressed and purified from E. coli and although circular dichroism and initial 1-Dimensional Nuclear Magnetic Resonance (NMR) were not encouraging, the NMR spectra changed upon addition of calcium, indicative of binding and therefore at least partial folding. This research will lead to structural information about the domain pair and studies of its interactions with other ECM proteins.


Presenter:  Saidharshini Pathmanathan

Supervisors:  P. Styles, K.K.Bhakoo

Development of in vitro models of cerebral ischaemia

Cerebral ischaemia is a condition that can arise from a variety of pathologies including stroke, head injury and tumour compression. A stroke occurs due to a sudden decrease in the blood flow to an area of the brain, therefore depriving the region of nutrients and oxygen. This leads to irreversible cell damage and cell death, and will eventually spread to healthy, viable tissue.
The dead tissue, which is no longer salvageable, is termed the ischaemic core. Surrounding this core is a region of partially ischaemic tissue known as the penumbra. This study aims to investigate this metabolically compromised tissue of the penumbra, and the potential to rescue it.
Brain slice and cell superfusion systems have been used in parallel, to mimic brain metabolism that occurs in vivo. The major cell types in the brain (glia and neurons) are studied in isolation to determine the individual cell typeís response to hypoxic-ischaemia.
A threshold of ischaemia where irreversible damage would occur was required in order that the duration of the insult could be manipulated and reperfused to monitor recovery. This would enable us to determine the point of recoverability. In clinical situations, many patients experience a phenomenon known as a transient ischaemic attack (TIA). During the TIA cells are metabolically compromised for a short period of time. Hence the initial threshold of irreversibility was set at 20min.
Having deduced the threshold of recoverability as a function of the duration of the hypoxic insult it is then possible to determine whether this threshold can be extended with metabolic or pharmacological intervention, into a period that would, 'without intervention' lead to cell death.

Presenter: Steven Porter

Supervisor:  Judith Armitage


Rhodobacter sphaeroides has a complex chemosensory system comprising multiple homologues of each of the Escherichia coli chemotaxis proteins. Two histidine protein kinase (CheAs) and six response regulator (5 CheYs and 1 CheB) homologues have been identified, but no CheZ homologues have been discovered. In vitro studies using purified proteins have shown that both CheAs are capable of ATP dependent autophosphorylation. However, the CheAs differ in their abilities to phosphorylate the various response regulators. CheA2 is able to phosphorylate all of the chemotaxis response regulators tested including CheB, whilst CheA1 is only capable of phosphorylating CheY1 and CheY5.
Previous genetic studies have shown that the normal response to attractant addition (increased flagellar rotation rate upon addition of attractant and stop upon removal) is mediated by CheA2 in conjunction with the motor-binding response regulators CheY4 and CheY5. The remaining CheYs facilitate signal termination possibly by acting as phosphate sinks. Deletion of CheA1 does not impair chemotaxis under any of the conditions tested, although this study shows that the signalling pathway mediated by CheA1 results in phosphorylation of a motor-binding response regulator (CheY5) and is therefore capable of influencing flagellar rotation bias. Furthermore, it appears that, unlike the CheA2 signalling pathway, the CheA1 signalling pathway does not utilise CheB-dependent feedback adaptation since CheA1 does not phosphorylate CheB.

Presenter:  Navita Rampersaud

Supervisor:  Prof. Louis Mahadevan

Phosphorylation of nucleoskeleton proteins, lamins A and C concomitant with immediate-early gene induction

In C3H 10T_ mouse fibroblasts diverse stimuli such as growth factors, cytokines and stress result in the induction of immediate-early (IE) genes (e.g. c-fos and c-jun).  IE gene induction is achieved through the activation of the mitogen-activated protein (MAP) kinase pathways (ERK, JNK/SAPK, and p38 MAP kinase) in response to diverse stimuli (EGF, TPA, anisomycin).  Activation of the MAP kinase pathways results in the further activation of downstream kinases (MSK1, RSKs, MAPKAP kinases), phosphorylation of transcription factors (c-Jun and ATF-2) and chromatin-associated proteins (HMG 14, histone H3).  Work in this lab has showed that the phosphorylation of HMG 14 and histone H3 (termed the nucleosomal response) is mediated via alternate pathways and is tightly correlated with IE gene induction.  Furthermore, these phosphorylated proteins have been found associated with nucleosomal particles comprising the c-jun gene.  Another nuclear protein was also found phosphorylated concomitant with IE gene induction, which appeared to have lamin-like properties.  This suggested that post-translational modifications such as phosphorylation on nuclear matrix proteins such as lamins could be involved in the regulation of IE gene expression.  This project aims to verify the identity of the lamin-like protein plus investigate the effects of MAP kinase-activating stimuli on its phosphorylation status.  We report that the nuclear matrix proteins, lamins A and C, but not lamin B, are phosphorylated in response to anisomycin or TPA.  Also, anisomycin- but not TPA-elicited lamin A and C phosphorylation was inhibited by the p38 MAP kinase inhibitor SB 203580, giving the first indication of the involvement of a MAP kinase pathway in lamin phosphorylation during interphase of the cell cycle.  The identification of the lamin kinase is currently under investigation using a selection of MAP kinase-specific inhibitors.


Presenter:  Simona Romagnoli

Supervisor:  Judy P. Armitage

Role of redox sensing in controlling Rhodobacter sphaeroides swimming behaviour

Rhodobacter sphaeroides is a motile, anaerobe facultative bacterium showing taxis to different stimuli. Its sensory system is organised in at least three genetic loci, whose expression is environmentally regulated. Responses to redox stimuli (oxygen, light, DMSO) rely on the relative activity of the different branches of the electron transport chains. Signals generated by membrane bound electron transport complexes converge on the His-kinase encoded by the second chemosensory operon (CheA2). A balanced response is produced according to the current growth conditions, metabolic state and redox poise. In particular taxis towards oxygen seems to be strictly dependent upon the activity of the low affinity terminal oxidase. Recent data have shown that a two components signal transduction system regulating the expression of genes essential for anaerobic metabolism is also involved in taxis to oxygen, both controlling the expression of the second chemosensory system during the transition from aerobic to anaerobic metabolism and interring with the normal pattern of phosphorylation of chemotaxis proteins, linking directly redox sensing to swimming behaviour.


Presenter: Sachchidanand

Supervisors:  Prof. Iain D. Campbell

Mapping the Heparin-Binding site on the 13-14F3 repeat of Fibronectin by NMR

Fibronectin, a multifunctional glycoprotein of the extracellular matrix, plays a major role in cell adhesion. Various studies have revealed that the human 13th and 14th fibronectin type III domains (labelled 13F3 and 14F3) contain a heparin-binding site. Mapping of the heparin binding sites of 13-14F3, 13F3 and 14F3 by NMR chemical shift perturbation and ITC shows that 13F3 provides the dominant heparin binding site and that the residues involved are within the first 29 amino acids of 13F3.  Predictions from earlier biochemical and modelling studies, as well as the X-ray structure of 12-14F3, are tested.  It is shown that the positively charged residues that project into the solvent from the ABE face of the ß-sheet on 13F3 are important but 14F3 does not appear  to contribute significantly to heparin binding.

Presenter:  Shaharum Shamsuddin

Supervisor:  Elena M. Klenova

Biochemical characterization of interaction partners of a transcription factor, CTCF

CTCF is a transcription factor, controlling several important genes, such as c-myc oncogenes, lysozyme and  amyloid protein precursor    (APP). CTCF is also involved in the regulation of insulator function and imprinting. It encodes a 11-Zn finger protein, which is ubiquitously expressed and highly conserved. Identification of proteins interacting with CTCF can help to understand the biological role of CTCF in the cell.  Such proteins were previously isolated in our laboratory. The aim of the present work is to biochemically characterise interactions of some of these proteins with CTCF. By using co-immunoprecipitation assays  we have shown that CTCF interacts in vivo with the RNA Polymerase II Large Subunit (LS Pol II ), transcription factor Yb-1 and tumour suppressor protein BRCA1. To map the  CTCF domains involved in the interactions,  the protein portions corresponding to the three domains of CTCF  ( N-, Zn and C - ) were  generated. We found that YB-1 interacts with the Zn-finger domain of CTCF and this interaction is direct.  The LS Pol II is associated with the C-terminal domain of CTCF and, remarkably, the  dephosphorylated form of CTCF interacts with LS Pol II  stronger than phosphorylated. To determine the amino acids within the CTCF protein which are involved in the interactions the truncated versions of the Zn- and the C- domains have been generated and are now being  analysed.

Presenter:  Jonathan Sharples

Supervisor:  Dr P.J.R Spooner and Prof. A. Watts

Relative Orientation of the ß-Ionone Ring and Polyene Chain for Retinal in Rhodopsin

The mammalian light receptor, rhodopsin is a member of the superfamily of transmembrane G-protein coupled receptors (GPCR) responsible for many cell signal transduction pathways. Defining the detailed structure of retinal within its binding site in rhodopsin is essential for a functional understanding of rhodopsin activation and an understanding of the general mechanism of GPCR activation.

In this study, solid state NMR has been used to characterize the orientation of the ß-ionone ring of retinal with respect to the polyene chain. The rotational resonance NMR method, for determining precise (down to ± 0.1Å) internuclear distances, has been conducted on bovine rhodopsin regenerated with selectively labelled [8,18-13C2]retinal or [8,16/17-13C2]retinal. The carbons 16 and 17 become inequivalent within the protein binding pocket enabling separate distance measurements to be made to both carbons attached to the C1-position of the ring.  NMR measurements appear consistent with a major population (75%) of the 11-cis-retinal adopting a 6-s-cis conformation in the ground state. The work will provide a model for the chromophore in rhodopsin that is refined with both orientational and distance information.

Presenter:  Robert Sidell

Supervisor:  Kieran Clark and Dominique Gaugier

Cardiac Dysfunction in Diabetes

Presenter:  Nicholas Smith


A Novel Statistical Method for Estimating Haplotype Phase from Genotype Data

When considering diploid genotypic data, it is not always possible to directly infer the constituent haplotypes of an individual from its genotype. This is the case when the genotype has two or more heterozygous loci. The ordering of the genotype into haplotypes is called the genotypic phase.
Haplotype data is far more informative than allele data. Many new analytical methods now require haplotype information. These range from inheritance-by-descent gene mapping to coalescence theory for determining the origin and history of populations.
It is sometimes possible to determine haplotypes from genealogical background or by experimental methods such as allele-specific PCR and single-molecule dilution PCR. These methods are costly and time-consuming however. Statistical methods can also be used to estimate the most likely set of haplotypes for a sample of genotypes. This has the advantage that it is fast and cheap and does not require any information about family background.
Currently available programs for estimating haplotype frequencies use the Expectation-Maximization (EM) algorithm. A Baysian approach has been developed which uses the Gibbs sampler, a type of Markov chain, to estimate haplotype frequencies.
In a comparative study using simulated diploid data the Gibbs sampler was found to estimate haplotype frequencies with an accuracy greater than or comparable to that of the EM algorithm. In another simulation the Gibbs sampler was used to estimate disequilibrium constants between pairs of linked biallelic loci when the phase is not known.
Unlike the EM algorithm the Gibbs sampler produces a probability distribution for the set of haplotype frequencies. Thus it can give probabilities for each of the haplotype estimates. This gives the user some indication of the reliability of the estimate and may provide alternative estimates

Presenter:  Suet-Mein Tan

Supervisor:  S. K. Alex Law

Effect of integrin ß2 subunit truncations on LFA-1 (CD11a/CD18) assembly,
surface expression, and function

Leukocyte Function Associated antigen-1(LFA-1, CD11a/CD18, aLß2 integrin) is involved in leukocyte function during immune and inflammatory responses.  We  aimed to determine a minimized ß2 subunit that forms functional LFA-1.   Using a series of truncated ß2 variants, we showed that the subregion  Q23-D300 of the ß2 subunit was sufficient to combine with the aL subunit  intracellularly.  However, only the ß2 variants terminating after Q444  promoted cell surface expression of LFA-1 and the surface expressed LFA-1  variants were constitutively active with respect to ICAM-1 adhesion. Hence  region after Q444 of the ß2 subunit is not required for heterodimer assembly  but contributes to LFA-1 cell surface expression and regulation of function.

Presenter:  Peter Teriete

Supervisor/co-presenters:  Banerji, S., Campbell, I.D., Jackson D.G., Day, A.J.

Solving The Solution Structure Of The Hyaluronan-Binding Domain Of CD44

CD44 , a type I membrane glycoprotein, is a major cell surface receptor for the extracellular matrix and is involved in a variety of functions. There are many isoforms of this protein, found in all major tissues, and diversity is achieved by alternative splicing of the single gene,which encodes CD44. Structural diversity is found in the extracellular portion of the protein, which encompasses the binding site for its main ligand, the glycosaminoglycan hyaluronan (HA). Sequence homology classified this domain as another member of the Link module family. Though, thought to be similar to the known structure of the Link module of TSG-6, previous studies established that additional N- and C-terminal regions are crucial for HA-binding. They are essential for the correct folding of the domain and present additional residues implicated in the actual HA binding. A construct encoding the Link module plus these extra regions were prepared, cloned and expressed in E.coli. 15N and 13C isotope labelled protein was expressed and triple resonance NMR techniques applied to elucidate the solution structure. This work describes the assignment procedures for backbone and sidechain resonances and the collection of distance restraints through NOE assignment.

Presenter:  Tobias Ulmer


Dependence of the Amide 1H Longitudinal Relaxation on Chemical Environment, Solvent Exchange and Oxygen Concentration: Implications for Protein Structure

The 1H longitudinal relaxation of the amide protons of the perdeuterated  2F2 module of  fibronectin has been used to gain inside into the factors  governing 1H longitudinal relaxation itself as well as mapping the  accessibility for molecular oxygen and comparing it to the accessibility of  the solvent water. The method applied demonstrates a general route to  determine the oxygen distribution inside a protein and the potential of detecting water molecules inside the protein is discussed, too.

Presenter:  Nathalie Y. Uzcategui C.

Supervisor:  Dr. EA Gould, Dr. W. James

Tracing The Origin Of Dengue-2 Virus In Venezuela Using Molecular Epidemiological Methods

Dengue fever is a human arthropod-borne disease caused by dengue viruses, which are distinguished into four serotypes. The evolutionary history of the dengue viruses and their involvement with human infections is still not fully understood. There is a well-established zoonotic cycle in which south-east Asian simian primates support the efficient transmission of all four dengue serotypes, implying that this geographical region may be the "home" of the dengue viruses. A remarkable increase in the incidence of Dengue Haemorrhagic Fever/Dengue Shock Syndrome (DHF/DSS) has been noted since the 1950s. The first DHF epidemic in the Americas occurred in Cuba in 1981; subsequently 24 other countries in the region have reported DHF. In a 1997 study, dengue type 2 virus was associated with increased pathogenicity. The strains isolated post-1989 resemble a Thailand strain that is characteristically associated with DHF in Asia. Since 1981, Venezuela has contributed more than 50% of the reported cases of dengue in the Americas. I am collecting, analysing and comparing forest and urban/rural viral isolates of dengue-2 using molecular epidemiological methods. Results thus far indicate that Venezuelan sequences from 1997/98, group phylogenetically with Chinese viruses rather than Thai viruses which were previously identified in the country.

Presenter:  Martin Ulmschneider

Supervisor/co-presenter:  D.P. Tieleman, M.S.P. Sansom

Computer Simulations of the 7 TM a-helices of Bacteriorhodopsin

Bacteriorhodopsin is one of the most intensively studied membrane proteins and has been the subject of many computer simulations and structural studies. There are currently 25 X-ray, NMR and EM structures in the PDB protein database, more than for any other membrane protein. Bacteriorhodopsin therefore represents the ideal test case for simulation methods and protocols. Molecular dynamics simulations of the individual a-helices were carried out in various simulation environments and setups. The effect of different media and system size (octane,  84 & 128 POPC lipids and water), the treatment of the termini (capped & uncapped) and the application of Heavy Hydrogens and Dummy atoms to increase the time step of the simulation were investigated in 10ns molecular dynamics simulations and compared to both previous simulations and the currently available structures. The findings were also compared to simulations with similar setup for the whole protein.

Presenter:  Kay Wan

Supervisor:  Prof.S. Kingsman

A Novel Stategy for the Intracellular Selection of Ribozymes

Although significant developments have been made in the field of hammerhead ribozymes, many obstacles lie in the way of optimising ribozyme function in the complex intracellular environment. This study aims to develop technologies for intracellular selection of ribozymes from a random ribozyme library using a novel strategy based on the properties of suicide genes and on the ability of ribozymes to inhibit gene expression.  This strategy will be exploited to intracellulary select for unknown cofactors which are necessary for the entry and integration of murine leukaemia virus (MLV) into target cells.  To test the feasibility of this strategy, ribozymes targeted specifically against a known cofactor for MLV transduction, namely the amphotropic receptor, Pit2, were designed and a stable cell line expressing anti-pit2 ribozymes has been constructed.  On transduction of the stable cell line with MLV virus containing a functional thymidine kinase (Tk) gene, virus can enter only cells that have been transduced by the MLV-Tk virus, and therefore identifies those cells expressing an active ribozyme.

Presenter:  Anna Warren

Supervisor:  Dr L.S.Cox

Identification of Proliferating Cell Nuclear Antigen (PCNA)-interacting Proteins through a Phage Display Technique

PCNA is an essential DNA replication and repair protein, which functions as a ‘sliding clamp’ on DNA.  The structure of the protein is trimeric, giving it a ring-like structure, which allows PCNA to act as an adaptor molecule between other proteins that have a role involving DNA, and the DNA itself. I have expressed and purified recombinant human PCNA, and verified its trimeric structure by native PAGE, gel filtration and EGS crosslinking.  Interactions of PCNA have been investigated using a phage display technique, screening both a random 12-mer peptide library displayed by filamentous phage, and a human cDNA library displayed by a modified form of T7 phage (T7SELECT-Novagen).  The screen has revealed three peptide motifs (filamentous phage), and three cDNA clones (T7SELECT-expressed proteins) that potentially interact with human trimeric PCNA.  These results and their implications will be discussed.

Presenter:  James Willcocks

Supervisor:  Dr K. Clarke, Prof. C. Ellory

Simultaneous determination of pH and free Mg2+ concentration in erythrocytes using 31P MRS

It is unknown how changes in blood total magnesium affect the levels of free magnesium within red blood cells or in tissue.  Although most cellular magnesium is bound, it is the concentration of free magnesium, [Mg2+]free, in red blood cells that is vital in the regulation of enzyme activity and ion transport.  The presence of other cations, especially H+ and K+, alters Mg2+ binding and may alter [Mg2+]free.  Consequently, we have developed an NMR spectroscopic method to measure [Mg2+]free in blood using the changes in the phosphorus chemical shifts of ATP and 2,3-DPG based on standard 3D titration curves of chemical shift over the pH range of 5.75 to 8.5 and [Mg2+]free range 0 to 5 mM.  The curves were fitted with theoretical equations expressing the observed chemical shift as a function of pH, K+ and [Mg2+]free, based on the appropriate equilibrium constants.  This method has been used to determine intracellular pH and [Mg2+]free in normal and sickle whole blood.  Under simulated arterial conditions, normal blood had a pHex of 7.37 ± 0.03 with an intracellular pH of 7.19 ± 0.02 and [Mg2+]freeof 0.42 ± 0.05 mM (n=18).  However, under the same conditions, sickle blood had a pHex of 7.32 ± 0.03, a pHi of 7.10 ± 0.03 and a [Mg2+]free of 0.32 ± 0.05 mM (n=9).  All were significantly lower than the controls (p<0.005).  This 25% decrease in [Mg2+]f may affect cell volume regulation and be involved in cell sickling.