Helicases unwind duplex DNA and are required for a multitude of cellular activities. The majority of Superfamily one (SF1) helicases unwind duplex DNA with 3' - 5' directionality and have been studied in depth. However, there is a set of SF1 helicases with 5' - 3' directionality, the mechanism of which is yet to be understood. The initial research of this thesis was to determine the role of the 2B domain of the SF1 3' -5' helicase, Bacillus stearothermophilus PcrA using modified inteins. However, this work, and work attempting to use the same modified inteins to study the Replication Factor C dependent loading of Proliferating Cell Nuclear Antigen onto DNA, proved unsuccessful. The crystal structure of the E.coli RecBCD complex has been solved previously and gives some insights into the 5'- 3' helicase mechanism of the SF1 helicase RecD. Due to stability problems, isolated E.coli RecD protein is an unsuitable target for characterisation therefore RecD homologues from other species were examined. The RecD homologue from Deinococcus radiodurans (drRecD2) was found to be soluble and was chosen for biochemical and structural characterisation. Mutational studies were carried out to investigate the molecular mechanism by which drRecDl unwinds duplex DNA. A molecular pin has been identified and is implicated in splitting open duplex DNA as part of the helicase mechanism. In addition to biochemical characterisation, drRecD2 was crystallised with a 5'-tailed duplex DNA substrate. The crystalisation conditions were optimised leading to diffraction data of 3.5A resolution.

No description available.

The preparation of two compounds designed to mimic the active site of Isopenicillin N Synthase (IPNS) is presented. Both compounds consist of two fragments from the enzyme linked by a lipophilic unit. The first contains a short linker unit namely 11-aminoundecanoic acid and the second contains a longer linker consisting of three of these units. The compounds were prepared by solid phase peptide synthesis and purified using a combination of gel permeation chromatography and Phenyl Sepharose hydrophobic interaction chromatography. These compounds have been examined in a two phase system of <i>n</i>-octane and aqueous buffer, containing the IPNS cofactors and tripeptide substrate. The postulated mimic would involve the lipophilic linker unit residing in the organic phase, thus holding the two peptides in close proximity in the aqueous phase. It was hoped that the addition of the Fe<SUP>2</SUP>+ cofactor might promote the adoption of a structure resembling that of the enzyme active site. Cyclisation of the penicillin precursors ACV and PCV has not been observed under the conditions investigated, however the compound containing the short linker has been observed to bind iron. Work on immunological studies on IPNS fragments has resulted in the preparation of anti-sera to the N-terminal peptide (IPNS(1-21). These anti-sera have been shown to bind to the peptides IPNS(1-21) and Cys-IPNS(1-21) in an ELISA test and have also been used to detect the native enzyme IPNS by the Western blot method.

The enzyme transketolase (TK) [E.C.2.2.1.1] catalyses the stereoselective transfer of a two-carbon ketol unit from a donor substrate such as lithium hydroxypyruvate (LiHPA, i) to an α-hydroxyaldehyde. Research into the use of TK from <I>Escherichia coli </I>as a process catalyst for asymmetric carbon-carbon bond formation has required the development of synthetic routes to novel acceptor substrates. The preparation of novel α-hydroxyaldehydes in enantiomerically pure form from chiral pool α-hydroxy-or α-amino acids, and of racemic α-hydroxy-aldehydes of utility in the synthesis of natural produce analogues, is described. Formerly, biotransformations mediated by TK involved the use of an excess of aldehyde substrate in buffered aqueous solution. An alternative protocol is presented, in which the biotransformations are performed in unbuffered medium; the natural pH change during the biotransformation is offset by use of a pH autotitrator to maintain the solution pH at the process optimum of 7.0. The synthetic utility of the chiral triols produced from TK-mediated biotransformations has been demonstrated through the development of synthetic routes to the natural product nectrisine, ii, and the <I>N-</I>hydroxypyrrolidine sugar analogue iii from the triols iv and v, available <I>via</I> TK-mediated condensation of LiHPA with the appropriate α-hydroxyaldehyde.

Flavocytochrome <I>b</I><SUB>2</SUB> is a L-lactate dehydrogenase found in the intermembrane space of certain yeasts. Its physiological partner is cytochrome <I>c. </I> The enzyme exists as a homotetramer with each subunit consisting of two domains, a flavodehydrogenase domain and a cytochrome domain. The cytochrome domain is homologous with the extensively studied cytochrome <I>b</I><SUB>5</SUB>. The two domains are joined by an inter-domain hinge. A computer model of how a complex may be formed between flavocytochrome <I>b<SUB>2</SUB> </I>and cytochrome <I>c</I> was produced in 1993, and predicted several residues to be important for molecular recognition. In accordance with other simulated models of protein complexes, the negative aspartates and glutamates of flavocytochrome <I>b</I><SUB>2</SUB> were aligned with the positive arginines and lysines of cytochrome <I>c. </I>The haems were found to be parallel, with an iron-iron separation of 25.6Å. An electron-transfer pathway was also proposed between the two haems. This involved the side-chain of isoleucine 50, which is in contact with the flavocytochrome <I>b</I><SUB>2</SUB> haem, the backbone of lysine 51, and the aromatic ring of phenylalanine 52, which was reported to be in van der Waal's contact with the haem of cytochrome <I>c. </I>The model predicted a key residue for complex formation on flavocytochrome <I>b</I><SUB>2</SUB>, glutamate 91. The construction of a mutant-enzyme, with glutamate 91 mutated to a lysine, produced a second-order rate constant for the reduction of cytochrome <I>c</I> of 37.7 μM<SUP>-1</SUP>s<SUP>-1</SUP>, within the value for the wild-type enzyme of 34.8 μM<SUP>-1</SUP>s<SUP>-1</SUP> (pH 7.5, <I>I </I>= 0.1M, 25°C). The haem redox potential for the mutant-enzyme was -13mV, a value again within error of the wild-type enzyme, 017mV. These data, along with data from two other mutant-enzymes, showed that the hypothetical complex was not a realistic model of how cytochrome <I>c</I> binds to flavocytochrome <I>b</I><SUB>2</SUB>. However, with the aid of molecular graphics and the sequence homology conserved within the 'cytochrome <I>b</I><SUB>5</SUB> fold', several residues were postulated to form a binding site for cytochrome <I>c. </I>

A study has been made of the a-oxoglutarate dehydrogenase (a-OGDH) multi-enzyme complex from the bacterium Acinetobacter lwoffi, in order to obtain a more detailed understanding of its molecular, catalytic and, in particular, its regulatory properties. This complex, which catalyses the oxidative decarboxylation of a-oxoglutarate to produce succinyl-CoA is composed of three subunit types (E1, E2 and E3), each possessing a different enzyme activity. Comparison with the analogous complexes from E. coli and mammalian tissues has yielded both differences and similarities. The native A. lwoffi a-OGDH complex was found to have a sedimentation coefficient of 29.9S, corresponding to a molecular weight of 1.82 x 106. Electrophoresis under completely denaturing conditions revealed that the complex is composed of three discrete polypeptide chain types with molecular weights of 55000, 60080 and 80900. Using specific chemical modification it seems likely that the E3 subunit of A. lwoffi a-OGDH is similar to those of the E. coli and mammalian complexes in that they all contain a disulphide bond at the active site, which is important in the reaction mechanism. Consequently, a loss of E3 activity was observed when this S-S bond was reduced. Furthermore, by chemically modifying the E2-bound lipoic acid moiety its involvement in the reaction mechanism was also confirmed. Attempts to dissociate the complex with retention of the individual subunit activities met with only limited success. Studies into the allosteric nature of regulation of the complex yielded some indication as to the mode of action of the nucleotide effectors (NADH and AMP). Using the technique of multiple-inhibition it was shown that NADH acts at a locus other than the a-oxoglutarate binding site. In the presence of protein denaturing agents a partial desensitization to AMP was observed. A possible explanation which is consistent with these results is discussed. Finally, some doubt now exists as to whether the E1 subunit is involved in NADH regulation, since previously reported indirect kinetic evidence that NADH acts on this subunit could not be confirmed directly using a radiochemical assay for the measurement of E1 activity.

I The mode of action of a number of inhibitors of citrate synthase has been studied. The results suggest that ?-oxoglutarate acts as an allosteric inhibitor of the enzyme from Gram negative facultatively anaerobic bacteria, but as an isosteric inhibitor of the enzyme from other sources. Similarly, NADH has been shown to be a powerful allosteric inhibitor of citrate synthases from Gram negative bacteria, but an isosteric inhibitor of the enzyme from Gram positive bacteria and eucaryotes. Other nucleotides have been shown to act only as isosteric inhibitors of citrate synthases from all sources examined. II Techniques have been developed which facilitate the rapid determination of the regulatory properties of a citrate synthase and its molecular size. The striking correlation between these properties and the Gram reaction of bacteria is discussed. It has been proposed that the rapid techniques described here could be of value in bacterial taxonomic studies and for bacteriological identification. III A citrate synthase deficient mutant of Escherichia coli has been isolated using a penicillin enrichment technique. A method has been developed which allows for the direct selection of citrate synthase deficient mutants by virtue of their intrinsic resistance to fluoro- acetate. Two citrate synthase deficient strains of Acinetobacter Iwoffi have been isolated using this method. IV A number of revertants, which have regained citrate synthase activity, were isolated from these citrate synthase deficient strains. A comparative study of the molecular, catalytic and regulatory properties of these enzymes has been carried out and possible structure-function relationships have been discussed. V Using revertant strains of E. coli which produce citrate synthases with regulatory properties different to those of the enzyme from the wild type organism, an attempt has been made to investigate the physiological significance of this altered regulatory behaviour of the enzyme. Revertants which have a citrate synthase which is not inhibited by ?-oxoglutarate (an allosteric inhibitor of the enzyme from E. coli wild type) appear to overproduce and excrete this (or a related) compound under certain growth conditions. Such a finding does suggest that the ?-oxoglutarate inhibition of citrate synthase has a physiological role in the regulation of this enzyme in this organism.

This thesis describes an investigation of the conformation of a small protein, lysozyme from hen egg-white, in aqueous solution which was carried out by means of nuclear magnetic resonance (nmr) spectroscopy. The conformation in the crystalline state had previously been investigated by X-ray diffraction, and a model of this conformation (the X-ray structure) was available. The solution and crystalline states could therefore be compared. Before conformational studies could be undertaken, several problems associated with the nmr spectroscopy of a protein had to be overcome. The spectrum of lysozyme (in this work ¹H resonances were studied) consists of a large number of broad overlapping lines. The individual resonances needed to be resolved and then assigned to specific protons in the molecule. Two different methods were devised to increase the resolution of the spectrum. First, the linewidths of resonances were reduced by a factor of about two by a mathematical manipulation of the spectrum (convolution difference). Secondly, the number of resonances observed in any given spectrum was reduced. This was achieved either by difference spectroscopy (the subtraction of two slightly different spectra) or by use of specific sequences of rf pulses in the Fourier transform nmr experiment. Using these methods, about sixty resonances were completely and separately resolved.

It has been observed that poly(High Internal Phase Emulsion) (polyHIPE) materials can be used as a biocatalysts, via the covalent immobilization of Candida Antarctica Lipase B (CAL-B). Recently, it has been shown that polyHIPEs can be prepared with epoxy functionality, which show potential for the covalent immobilization of enzymes. The aims of our work were, firstly, to produce an open-void glycidyl methacrylate (GMA)-based polyHIPE material. Secondly, these materials were then to be developed for use within a continuous flow set-up. Thirdly, the post-polymerisation of these materials was to be investigated. Finally, these materials were to be used as a support for the covalent immobilization of enzymes. Highly porous, open-void GMA-based polyHIPE materials were accomplished via the photo- initiation, rather than thermal initiation of the continuous phase of the emulsion. The rapid cure of the emulsion effectively ‘locks’ the emulsion morphology, prior to emulsion destabilisation, that is more prominent in the slower thermally initiated HIPEs. Photopolymerised GMA-based polyHIPE materials were further developed for use within a continuous flow-set up. GMA-based polyHIPE materials were functionalized post-polymerisation with tris(2-aminoethyl)amine, morpholine and O,O’-bis(3-aminopropyl)polyethylene glycol. The functionalization of these GMA-based materials was observed via a number of analysis techniques, such as FT-IR spectroscopy, XPS spectroscopy, elemental analysis, Fmoc number determination, 1H HR-MAS NMR spectroscopy, and the covalent attachment of ninhydrin and FITC. Elemental analysis of the morpholine and tris(2-aminoethyl)amine polyHIPE showed that a near quantitative conversion, of 72 and 82 % respectively, was accomplished via the reaction being conducted at reflux for 24 hours. The enzymes, Lipase from Candida Antarctica and Proteinase K from Tritirachium album were immobilized either directly onto the polyHIPE material or via a hydrophilic spacer group, O,O’-bis(3-aminopropyl)polyethylene glycol. CAL was immobilized with a loading of between 5.4 and 7.5 wt. % per g of polyHIPE material.

The use of biotransformations in organic synthesis is a small but growing field. Biotransformations can be used for many steps which prove difficult with traditional chemistry, particularly stereospecific syntheses and chiral resolutions. However, to enable the increasing use of biotransformations in chemical synthesis, a broader range of biocatalysts will be required than those of which we are currently aware. Bioprospecting and genome sequencing are increasing the awareness of Nature's library of biocatalysts. These enzymes need to be characterised for their use as biocatalysts, and where necessary modified to meet needs that may not be catered for in nature. One currently used biocatalyst is transketolase. Transketolase is an enzyme which is produced by most organisms. As a biocatalyst it is most often extracted from spinach, or produced by recombinant Escherichia coli. Transketolase carries out a stereospecific carbon-carbon bond formation, removing a 2-carbon ketol group from a ketose sugar, and adding it to the aldehyde group of the aldose sugar. It is a useful biocatalyst as it has a broad substrate specificity. Significantly it will use hydroxypyruvate as the ketol donor, thus liberating carbon dioxide and driving the reaction in a forward direction. In this thesis the transketolase-like enzyme, 1-Deoxy-D-xylulose 5-Phosphate Synthase (DOX-P Synthase), is identified as another possible biocatalyst. Significantly DOX-P Synthase catalyses the addition of a carbonyl unit, not from hydroxypyruvate, but from pyruvate, a reaction which cannot be catalysed by transketolase. This thesis therefore describes studies that aimed to produce recombinant DOX-P Synthase as a biocatalyst, and to optimise its production. Further studies aimed to characterise the properties of DOX-P Synthase, with the aim of optimising its use as a biocatalyst. Analysis of the substrate range of DOX-P Synthase could then be used to describe the set of reactions for which it would be a useful biocatalyst. Finally, this thesis describes modifications made to the enzyme, carried out by site-directed mutagenesis, and the effects on enzyme properties.