As an approach to explore the amide cleavage reaction, we designed, synthesised, fully characterised and investigated different model systems resembling the catalytic zinc sites of peptidases. The models are mononuclear zinc(II) complexes containing the unit (6-NHR-2-pyridylmethyl)amine (R = CO^tBut, H) as a common feature. This unit supplies an N2 coordination motif and an intramolecular amide oxygen capable of binding the metal ion (Zn^…O=C(amide)); such event has been proposed to be crucial for the cleavage of peptide bonds in peptidases. Besides, these models offer different types and numbers of metal binding sites, which affect electronically and sterically the zinc(II) ion as well as its Lewis acidity. Indeed, the effect of the first coordination sphere seems to influence drastically the cleavage of the intramolecular amide bond leading to an overall change in the stability of the amide bond of <i>ca. </i>300-fold. The unit (6-NHR-2-pyridylmethyl)amine also provides amide/amine groups that can hydrogen bond other zinc-bound ligands. Some of these complexes were, therefore, used to explore strategies to induce and manipulate hydrogen bond interactions, and to investigate the factors that influence their strength. These hydrogen bond interactions occur both in solution and in the solid state and their strength was determined by IR spectroscopy and correlated with the X-ray crystal structures. We also report two strategies to position NH groups in the proximity of a zinc(II)-bound amide oxygen N-H^…O=C(amide), and we investigate their effect on the stability of the amide group.

Tryptophan 2,3-dioxygenase (TDO) from <i>Xanthomonas campestris </i>(xcTDO) is a highly specific heme-containing enzyme from a small family of homologous enzymes, which includes indoleamine 2,3-dioxygenase (IDO). TDO contains a histidine residue (histidine 55) in its active site, which hydrogen bonds to the indole nitrogen atom of L-tryptophan, and could function as an active site base. In this study we attempt to resolve the question of whether an active site base is necessary for catalytic activity, to which end, two complementary strategies, based on the active-site structure have been applied. Firstly, active-site mutants were studied, where histidine 55 was replaced by alanine or serine (H55A and H55S). The crystal structures of the H55A and H55S mutant forms were determined to 2.15 Å and 1.90 Å resolution respectively, in binary complexes with L-tryptophan. These structural data, in conjunction with potentiometric and kinetic studies, reveal that histidine 55 is not essential for turnover, but greatly disfavours the mechanistically unproductive binding of L-tryptophan to the oxidized enzyme, allowing control of catalysis. The second strategy utilises 1-methyl tryptophan. Our studies reveal that whilst 1-methyl tryptophan is not a substrate for TDO, and is not an inhibitor of its action, the active site mutants H55A and H55S can deoxygenate 1-methyl tryptophan. These catalytic differences are explained by comparison of the active sites of the enzymes, and support the proposal that a catalytic base is not necessary for enzymatic activity. In addition, these data provide new insight into the future of 1-methyl tryptophan as an inhibitor of IDO activity in cancer treatments.

No description available.

Random mutagenesis libraries were prepared by propagating the <i>Trigonopsis variabilis</i> D-amino acid oxidase (TvDAAO) gene in a mutator strain and by performing error-prone PCR (epPCR) on the <i>Rhodotorula gracilis </i>D-amino acid oxidase (RgDAAO) gene. These libraries were initially screened using a solid-phase for detection of activity towards β-amino acids. To overcome reproducibility problems, a high-throughput liquid-phase screening method in 96-well plate format was developed, evaluated and used to screen the RgDAAO epPCR library. In the absence of any detectable activity towards the selected β-amino acid substrates, an alternative strategy, which minimised the variant library size and increased the range of substrates, was applied. Single site saturation libraries were generated by saturation mutagenesis PCR at the RgDAAO active site residues F58, M213, Y238 and R285. The resulting variant libraries were screened in the liquid-phase for activity towards a range of proteinogenic and non-proteinogenic amino acids. Saturation mutagenesis at the conserved active site residues Y238 and R285 produced mainly deleterious exchanges. However, saturation at the active site residues F58 and M213 produced a range of beneficial and deleterious variants, allowing substrate profiles to be produced for each. Evaluation of these profiles permitted the identification of individual variants. Although no oxidase activity was detected towards β-amino acids, improved activity was observed in both the F58 and M213 saturation mutagenesis libraries towards several proteinogenic and non-proteinogenic substrates. The improved activity detected in the assay for the variants F58L and H58I towards (<i>rac)</i>-tetrahydroisoquinoline-3-carboxylic acid and for F58M towards (<i>R)-</i>pipecolonic acid was confirmed by performing whole cell biotransformations and analysing the progression of the reactions by HPLC.

A group of mutants of the <I>Eco</I>KI R/M-system displaying de novo methylation activity have been isolated (Kelleher <I>et al</I>., 1991). The mutated genes were transferred into an overexpressing plasmid vector. Two of the over-expressed proteins were purified to near homogeneity from clones transformed with the plasmids. Cofactor binding activities of wild-type and the two mutant enzymes were compared by 1,8-anilino-napthalene sulphonic acid fluorescence displacement experiments. A DNA methylation assay based upon the transfer of a tritiated methyl group from the cofactor AdoMet to the substrate DNA was established and used to examine the dependency of the reaction on cofactor, substrate, and enzyme concentration. In addition the stability of the trimeric enzyme at different protein concentrations was followed by HPLC gel filtration. Sequence alignments, secondary structure predictions, and tertiary structural modelling were used to show the similarity of the Type I system <I>Eco</I>KI with methyltransferases from other classes (especially Type II methyltransferases), thereby establishing a structural and suggesting an evolutionary link between the different methyltransferase classes. The information obtained by these comparisons enabled the subsequent modelling of a more refined model of the <I>Eco</I>KI structure. A model is proposed to explain the different activities observed in wild-type and mutant enzymes based on the biochemical and structural data obtained during these investigations.

High-throughput screening is now widely used to isolate specific or improved activities of various enzymes. However, most of the current screening methods are based on in vitro procedures. The aim of this project is the development of an in vivo directed evolution approach to study a novel racemase enzyme. Although in vivo selection in itself is not new, this project emphasises a new approach to the high throughput in vivo concept, selecting an enzyme activity apparrently unintended by nature. The selection, which can screen a library of 10⁶-10⁸ mutants per assay, is complementary to in vitro screening, and applies new genetic tools for strain construction. Since the racemase under investigation is active on commercially important compounds, this method may also be suitable to identify and improve new industrial biocatalysts. The focus of the study is an N-acylamino acid racemase from Amycolatopsis which displays relaxed substrate specificity upon N-acetylated amino acids but no activity towards free amino acids. As such, this enzyme is potentially valuable to use with well established enantioselective acylases for the dynamic kinetic resolution of L- or D-amino acids. However, the racemase displays properties which limit its usefulness as a biocatalyst, such as a low substrate turnover, and inhibition of activity at substrate concentrations exceeding 50mM. The racemase also appears to have evolved serendipitously from a different activity of the enolase superfamily. The in vivo selection developed in this work is based on the construction of specific auxotrophic mutants, which require expression of this racemase activity for growth. This offers a way to explore this very interesting and novel N-acylamino acid racemase activity of an enzyme in the enolase superfamily. The screening of libraries of variants derived from the wild type enzyme was applied, to unveil mutations which improve activity, and demonstrate the value of the in vivo approach. This in vivo selection could therefore potentially help to understand the evolution of new activities in the enolase superfamily and develop a new class of biocatalyst to expand the utility of acylase resolutions by industry.

Serine palmitoyl transferase (SPT) catalyses the first and rate limiting step of sphingolipid biosynthesis; a decarboxylative, Claisen condensation of the amino acid L-serine and the long chain fatty acid palmitoyl CoA to produce the first sphingolipid precursor 3-ketodihydrosphingosine. We propose <i>S. paucimobilis </i>as a model organism for the characterisation of sphingolipid biosynthesis. SPT belongs to the α-oxoamine synthase family which are a small group of pyridoxal 5’-phosphate (PLP)-dependent enzymes. This thesis describes spectroscopic studies of recombinant SPT and reports high resolution crystal structures of the PLP, L-serine (substrate) and L-cycloserine (inhibitor) bound forms of <i>S. paucimobilis </i>SPT (1.3, 1.5 and 1.45 Å respectively). These structures are the first of an SPT from any organism and provide an insight into the mechanisms of catalysis and inhibition that take place at the active site of this important enzyme. Another interesting sphingolipid biosynthesis enzyme is Inositol Phosphorylceramide Synthase (IPCS; also known as Aur1p) which is encoded by the <i>AUR1 </i>gene. Deletion or mutation of this gene in <i>S. cerevisiae </i>was found to be lethal therefore highlighting the importance of this enzyme. Aur1p is unique to fungal cells and catalyses the transfer of phosphatidyl inositol onto phytoceramide to produce inositol phosphophytoceramide (IPC). Since it is unique to fungi, Aur1p provides an attractive target for the development of novel anti-fungal agents however it is an integral membrane protein of the Golgi apparatus with 6 predicted transmembrane domains. This means that to date, no structural information is available for Aur1p. Expression of a recombinant, soluble, affinity-tagged Aur1p now paves the way for full characterisation of this important enzyme.

This thesis describes the application of a number of whole cell systems to the hydroxylation of organic compounds and thus the production of high value synthetic intermediates. Such transformations are currently very difficult, or impossible to achieve using purely synthetic techniques. Biocatalysis on the other hand, offers the potential to achieve this process in a mild, regio- and stereoselective manner. A range of compounds comprised of a carbo- or heterocycle attached <i>via</i> a linker to an aromatic group have been prepared. These compounds were screened as potential substrates with a number of whole cell systems thought to contain cytochromes P-450. Subsequent biotransformation, product isolation and characterisation are described. Investigation of the biohydroxylation of a series of <i>N</i>-carboxybenzylalkylpiperidines by <i>Beauveria bassiana</i> ATCC 7159 and study of the factors important to the selectivity of hydroxylation by the organism is illustrated. Research into the biohydroxylation capabilities of two members of the <i>Rhodococcus </i>genus has also been carried out. <i>Rhodococcus</i> sp. NCIMB 9784 and <i>Rhodococcus rhodochrous </i>NCIMB 9703 were found to be suitable biocatalysts for the hydroxylation of non-natural substrates. Subsequent exploration of the regio- and enantioselectivity of this hydroxylation process is discussed.

No description available.

The type I restriction enzyme <i>Eco</i>KI is an oligomeric enzyme consisting of subunits responsible for DNA target recognition (HsdS), DNA modification (HsdM) and the ATP-dependent translocation and cleavage of unmodified DNA (HsdR). It was shown <i>in vivo</i> and <i>in vitro</i> that some mutations in <i>hsdM</i>, which impair modification activity of <i>Eco</i>KI, result in a restriction-proficient modification-deficient enzyme. The survival of the mutant bacteria is dependent on the presence of functional ClpXP protease; HsdR is degraded by ClpXP in the restriction-proficient modification-deficient mutant. This degradation requires a restriction complex capable of ATP-dependent translocation and occurs before the completion of the restriction pathway. Degradation of the restriction subunit results in a phenomenon referred to as restriction alleviation (RA). If <i>hsdR</i> is present in high copy number, thereby increasing the level of modification-deficient <i>Eco</i>KI, ClpXP is unable to protect the bacterial chromosome from attack by <i>Eco</i>KI. A restriction-proficient modification-deficient mutant retains some residual restriction activity, which is efficient against DNA that enters the bacterium in the double-stranded form but not against DNA that enters in the single-stranded form. The level of HsdR is depleted in the cytoplasmic but not in the membrane fraction under the conditions that lead to restriction alleviation. However, compartmentalisation of <i>Eco</i>KI cannot explain why chromosomal DNA triggers ClpXP-dependent degradation of HsdR whereas foreign DNA does not. ClpXP protects unmodified chromosomal DNA but not foreign DNA from the resident restriction complex.