Thesis (Ph. D.)--University of Texas at Austin, 2002. / Vita. Includes bibliographical references (leaves 93-97). Available also in a digital version from Dissertation Abstracts.

Polyketides are a class of small molecules synthesized by a broad spectrum of bacteria, plants, and fungi, and many exhibit powerful bioactive properties. The number of clinically-relevant compounds adapted from polyketide scaffolds is growing, eliciting attempts from synthetic organic chemists to construct polyketide-related compounds in the laboratory from simple chemical building blocks. Unfortunately, the current efficiency by which a skilled artisan can synthesize even small quantities of a polyketide is severely limited by the functional and stereochemical complexity of these compounds. Conceptually, it would be much simpler to genetically reprogram the enzymes responsible for polyketide biosynthesis to produce designer molecules; however, the massive size of polyketide synthase enzymes has hindered efforts towards understanding critical features of their structures and mechanisms. Only very recently has structural information become available for enzymes involved in polyketide biosynthesis, providing an initial glimpse into the inner workings of these subcellular pharmaceutical factories. It will not be possible for mankind to fully realize the potential of engineered polyketide synthases without understanding how their architectures govern the molecules they have evolved to produce. In this work, the structure and mechanism of several enzymes involved in polyketide biosynthesis is investigated. An unprecedented architecture for the ketoreductase-enoylreductase didomain from the second module of the spinosyn polyketide synthase reveals structural divergence from the related mammalian fatty acid synthase, and reconstituted in vitro activity of the enoylreductase domain indicates the isolated enzyme retains activity apart from its parent polyketide synthase module. The dehydratase domain isolated from the tenth module of the rifamycin polyketide synthase, previously hypothesized to only form double bonds with (Z) geometry, was found to have altered stereoselectivity dependent on the carrier handle bound to the substrate. The enoyl-isomerase domain, isolated from the fourteenth module of the bacillaene polyketide synthase, utilizes a catalytic mechanism that relies only on a single active site histidine. A series of ketosynthase domains from trans-acyltransferase polyketide synthases reveal how polyketides bind covalently to the active site of the ketosynthase, and how the flanking subdomain of the ketosynthase is used as an anchor point for the assembly of the polyketide synthase megacomplex.

Not available / text

No description available.

An enzyme has been isolated from the fungus responsible for Dutch Elm Disease, Ophiostoma novo-ulmi, that stereoselectively catalyses the production of S ketoprofen from racemic ethyl ketoprofen. The enzyme has been cloned, over-expressed and mutated by Chiroscience, plc. The main aim, at the outset of this work, was to crystallise this enzyme and determine the three-dimensional crystallographic structure. A mutant form of the enzyme was purified to apparent homogeneity, by two different protocols, and subjected to numerous crystallisation trials. Trials were unsuccessful. Mass spectrometry of the purified sample revealed the presence of two protein species that differed by 832 Da. Analysis of the expression vector and N-terminal sequences of the two protein species revealed that they were both forms of the enzyme that differed by ten amino acids at the N-terminus. A further sample was supplied that contained only one form of the enzyme. This has been purified by the same two purification protocols and subjected to similar crystallisation trials. Needle like crystals have been grown. In order to collect X-ray diffraction data it has been necessary to carry out extensive crystallisation trials with the enzyme. In this respect the work developed into a study of crystallisation techniques and phenomena, which focused on uncoupling the processes of nucleation and growth and generally controlling nucleation. Crystals have been produced that diffract beyond 2.5 A and a native dataset has been collected at the Daresbury synchrotron source. Heavy atom derivatives have so far been nonisomorphous. An immobilised form of the enzyme has been produced by cross linking microcrystals with the bifunctional reagent glutaraldehyde. In this form the enzyme has increased stability towards temperature, pH and organic solvents. Amino acid sequence alignment and chemical analyses have suggested that the enzyme may belong to a superfamily of active serine hydrolases that include penicillin recognising proteins.The vanadium dependent bromoperoxidase from the macroalgae Corallina officinalis has been purified to homogeneity and crystallised. Native and derivative X-ray diffraction datasets have been collected and the structure solved by Multiple Isomorphous Replacement in collaboration with Dr M. Isupov and Dr A. Dalby. The molecule displays 23 point symmetry, which has not been observed in proteins to date. The molecule was crystallised in its dodecameric form, revealing the monomer to be a single domain a-helical protein with a four helix bundle as the main structural motif. The vanadium binding site is located at the end of this four helix bundle. The vanadium binding site of the chloroperoxidase from Cul. inaequalis is also located at the end of a four helix bundle. This illustrates that although there is limited amino acid sequence homology between the two enzymes they are structurally related. Purification

No description available.

No description available.

No description available.

No description available.

No description available.