The recently sequenced genome of the model organism <i>C. elegans</i> was found to contain a gene that potentially encoded a protein containing a methyl-CpG binding domain and intriguingly, two motifs present in chromatin-associated transcriptional regulators - the PHD-zinc finger and bromodomain. The gene was given the standard name <i>pbm-1</i> for PHD-zinc finger, bromodomain and MBD gene-1. The MBD of pbm-1 was found to be highly homologous to a human EST suggesting evolutionary and functional conservation of this domain. A mouse cDNA, mPBM1 was identified that encoded a protein with the same modular organisation of motifs as the <i>C. elegans</i> pbm-1 protein. The aim of this project was the characterisation of this gene family using a cross-species approach that exploited the advantages particular to the nematode and mammalian systems. The MBDs of pbm-1 and mPBM1 did not associate with methylated sequences in a series of in vitro binding studies. These findings do not rule out the possibility that the MBDs mediate binding to other covalent modifications of DNA or specific sequence targets. In support of the in vitro findings, the mPBM1 protein did not associate with the heavily methylated mouse major satellite when fused to GFP and was found to localise in a diffuse, granular pattern in the nucleus. Furthermore, the MBD of mPBM1 was dispensable for the spatial localisation of the protein. Deletion of a C-terminal element encompassing the PHD-zinc finger and bromodomain regions abolished the wild type localisation in mouse nuclei. The PHD-zinc finger has been identified in a number of chromosomal translocations involved in acute leukaemias suggesting that deregulation of normal expression contributes to the transformation phenotype. The hPBM1 gene was physically mapped to chromosome 2q23 in a region found deleted and rearranged in leukaemias.

Four dominant negative mutants of PRP2 were produced by site-directed mutagenesis of the conserved ATP-binding motif (Motif A) of PRP2. These mutants, when overproduced in a wild-type yeast strain caused a growth inhibition due to inhibition of splicing, leading to an accumulation of pre-mRNA. <I>In vitro</I> analysis of cell-free splicing extracts showed that splicing was inhibited in the presence of the dominant negative proteins. The splicing inhibition could be alleviated if enough wild-type PRP2 was present in the extracts. The dominant negative mutants caused accumulation of the fully assembled but inactive spliceosome. Immunoprecipitation with anti-PRP2 antibodies showed that one mutant protein (PRP2<SUP>GKN</SUP>) was stably associated with the spliceosome. In contrast, another mutant protein (PRP2<SUP>AKT</SUP>), appeared not to be associated with the spliceosome, suggesting that it was sequestering some other essential splicing factor outside the splicing complex. Using an affinity column, two of the dominant negative proteins have been purified. Assays have shown that neither protein can bind poly U RNA and that both show negligible ATPase activity. Both purified proteins can inhibit the activity of a wild-type splicing extract and prevent the formation of the active spliceosome.

Proliferating cell nuclear antigen (PCNA) plays a key role in DNA replication and repair. In this thesis I have determined the crystal structure of <i>S. pombe </i>PCNA on its own and in complex with a peptide derived from the natural inhibitor of PCNA in humans, p21. Using the p21 peptide as a template, chemical libraries were screened for small drug-like molecules that are able to mimic the protein-protein interaction between human PCNA and p21 and tested for binding using various fluorescence-based methods. Binding of selected compounds could not be observed, however, which may have been due to poor solubility of the compounds as well as a lack of assays sensitive enough to pick up on ligands with low affinity. Not all direct PCNA binding proteins bind to the above mentioned pocket and PCNA:protein co-crystal structures of these examples without the PIP-box motif are not available so far, so information on the mode of binding in those cases is not accessible yet. Therefore, I tried to narrow down and characterize the PCNA binding site of Gadd45. This protein does not contain a PIP-motif, but previously was shown using yeast-two hybrid technology to bind directly to PCNA <i>via </i>its C-terminus. <i>In vitro </i>binding of the two full-length proteins could not be confirmed, due to the inherent instability of Gadd45, which also has been reported by others. GST-tagged truncations of the C-terminus of Gadd45 were expressed and purified and binding to PCNA was studied using SPR. These results are at odds with the published binding data and suggest that either the interaction between PCNA and Gadd45 is not direct and needs to mediated by a third factor or Gadd45 binds to PCNA <i>via </i>a different part than the C-termination (as also suggested in the literature).

<i>N</i>-glycosylation of proteins plays an essential role in controlling protein folding; recognition and many other cellular interactions. The pathway in which N-glycans are formed is the <i>Leloir </i>pathway. Of this pathway one core feature is retained through all N-linked glycoproteins, the GlcNac₂Man₃ pentasaccharide core. This moiety if formed on the cytoplasmic face of the endoplasmic reticulum. A number of these enzymes have been identified but a single β-1, 4-mannosyltransferase cloned and expressed, ALG1. Although the enzymes are mannosyltransferases little similarity exists between them, this makes similarity searching to find other enzymes of this pathway impossible. A bioformatic method was employed to build profiles of the known enzymes to identify further enzymes from genome scanning to identify putative transferases of this pathway. A number of potential targets were identified, these ranged from known mannosyltransferases with precise function still undefined to undefined open reading frames. The target genes which most closely fitted the constructed profiles were cloned and expression attempted in a bacterial host. Expression was not observed for a majority of the clones even after modification to remove membrane spanning portions. Expression was observed for YGL047w; YBR070c, YBR095c. Although good expression was achieved, the purified protein showed no activity or affinity for GDP was observed. ALG2, the second known mannosyltransferase of the <i>Leloir</i> pathway, was successfully cloned and expressed. The expressed proteins were shown to have no activity towards the natural substrate (PPGn₂Man₁) or towards GDP. β-1,4-galactosyltransferase activity was also investigated with respect to modified activated sugar donor species. Although modified acceptor species is known to be tolerated investigations indicate that no modification in donor species is possible.

No description available.

Human C4b-binding protein (C4BP) protects host tissue, and those pathogens able to hijack this plasma glycoprotein, from complement-mediated destruction. For example, C4BP localises to apoptotic and necrotic cells, via its affinity for DNA, whereupon this regulator helps to prevent complement activation and subsequent inflammation and tissue damage. The first two CCP modules of the C4BP alpha-chain, plus the four residues connecting them, are necessary and sufficient for binding a bacterial virulence factor, the <i>Streptococcus pyogenes</i> M4 (Arp4) protein. Structure determination of this region of C4BP by NMR reveals two tightly coupled CCP modules in an elongated arrangement. Chemical shift perturbation studies demonstrate that the N-terminal hyper-variable, region of M4 binds to a site including strand 1 of CCP module 2. This interaction is accompanied by an intermodular reorientation within C4BP. The DNA-binding site of C4BP also involves CCP1 and CCP2 of the alpha=chain.  Chemical shift changes locate the binding site for DNA to a groove at the CCP1/2 interface enabling the use of data-driven docking to produce a model of the C4BP12-DNA complex. The work described in this thesis thus provides detailed pictures of interactions whereby a pathogen evades complement and that help protect host tissue during programmed cell death.

The protease thermolysin has been used directly to synthesise dipeptides from Fmoc-protected amino acids on a PEGA₁₉₀₀ solid support. The thermolysin-catalysed solid phase synthesis of longer peptides is reported. Fmoc-protected peptides as long as hexamers (poly-L-leucine) and tetramers (poly-L-phenylalanine and poly-L-tyrosine) were enzymatically synthesised. In each case enzymatic synthesis of peptides resulted in the formation of a library of peptides varying in length, the formation of which was caused by “scrambling” by the enzyme. Due to the reversible nature of the enzyme catalysed peptide bonds, it was believed that the aforementioned solid phase peptide libraries may be exploited to generate dynamic peptide libraries. From the Juliá-Colonna asymmetric epoxidation it is known that chalcone undergoes non-covalent interactions with the amino terminus of poly-L-leucine. Chalcone was therefore employed as a suitable template for the PEGA₁₉₀₀ immobilised poly-L-leucine libraries. However, it was found that the Fmoc-protecting group used in these libraries inhibited binding of chalcone to poly-L-leucine. As an alternative, unprotected poly-L-leucine libraries were generated in solution phase from dileucine using thermolysin. The catalytic activity of poly-L-leucine library members and poly-L-phenylalanine library members in the Juliá-Colonna epoxidation was investigated.  These investigations demonstrated that the peptide amplified by chalcone in the dynamic combinatorial libraries displayed an improved catalytic activity in comparison to other library members. This indicates that dynamic peptide libraries may be exploited as a tool for identifying potential catalysts for the Juliá-Colonna asymmetric epoxidation.

A comprehensive study was undertaken for <i>ab initio </i>protein fold prediction using a fully atomistic protein model and a physicochemical potential. Twenty four EA designs where initially assessed on polyalanine, a prototypical α-helical polypeptide.  Design aspects varied include the encoding alphabet, crossover operator, replacement strategy and selection strategy. By undertaking a comprehensive parameter study, the best performing designs and associated control parameter values were identified for polyalanine. The scaling between the performance and polyalinine size was also identified for these best designs. This initial study was followed by a similar parametric study for met-enkephalin, a five residue polypeptide that has long been used as a <i>de facto </i>standard test case for protein structure prediction algorithms. It was found that the control parameter scalings identified from the polyalinine study were transferable to this real protein, and that the EA is superior to all existing <i>ab initio</i> approaches for met-enkephalin. The best design was finally applied to a series of real proteins ranging in size up to 45 residues to more generally assess the EA’s performance. The thesis is concluded with consideration of the future work required to extend the EA to larger proteins and <i>ab initio </i>structure prediction for non-native environments such as at interfaces, which are of relevance to, for example, biosensors.

No description available.

This PhD project consisted of two parts. The first part was our successful T0100 prediction in CASP4. In this prediction, we produced one of the highest ranked threading alignments through sequence analysis which revealed a, “Cys-staples” pattern formed by putative disulphide bridges between consecutive turns in the parallel β-helix core of different homologues. This pattern was used as an anchoring point in the template-target alignment, and this novel approach motivated the follow-up project which constitutes the second part. The aim of this second part of the project was to apply the aforementioned approach as widely as possible and to produce high-confidence models for all detectable members of the PLL superfamily in GenPept (as retrieved from the NCBI in July 2002). Large-scale detection of PLL proteins was achieved initially with the help of two different third party fold recognition programs, setting the parameters and cut-off values carefully to be stringent in order to minimise false positive predictions. The two resulting datasets were then pooled and clustered. This resulted in twelve families with homologues in PDB, eight families without close homologues in PDB but with some members annotated as pectolytic enzymes, and one new family with no indication of prior classification as PLL. A small fraction of PLL predictions were deemed to be probable false positives, and a few others could not be followed up upon confidently because no homologues could be detected by standard BLAST searches of the public sequence databases. After augmenting the nine families without known structures through standard BLAST searches of SPTrEMBL and careful analysis and editing of automated target-template alignments, all plausible members of the altogether twenty-one families were modelled using an automated modeling procedure.