Thesis (Ph. D.)--University of Queensland, 2001. / Includes laminated chart in front pocket. Includes bibliographical references.

Thesis (Ph.D.) - University of Queensland, 2005. / Includes bibliography.

Ribosomal synthesized and post translational modified peptide natural products have attracted a lot of interest in the past decade. Backbone cyclization of the translated linear peptides is generally catalysed by specific enzymes giving them peptidase resistance, thermodynamic stability and various other physiological activities. These features have made backbone cyclic peptide to become an attractive resource for drug discovery. Here, we described the synthesis of linear peptides containing natural and unnatural residues and its biosynthetic mechanism to generate man-made cyclic peptides. In this thesis we used SPPS to make short and medium linear peptide chains, we purified them using HPLC, and analysed them using MS. We incorporated unnatural residues such as homocysteine, homoserine, aminoalanine, propargyl glycine and the substrates were subjected to different enzymatic reaction such as prenylation, heterocyclization and macrocyclization modification reactions to generate small macrocycles (4-6 residues), prenylated linear peptides, and patellamime analogues. The final products were analysed using LC-MS. In our results, we verified that kawaguchipeptin (kgp) gene cluster is responsible for the production of kawaguchipeptins through heterologous expression of the kgp gene cluster in Escherichia coli. The KgpF prenyltransferase was overexpressed and was shown to prenylate C-3 of Trp residues in both linear and cyclic peptides in vitro. We also found out that PatGmac can macrocyclise short peptides (4-6 residues) to generate small macrocyclic peptides. We also tested the flexibility of OscGmac using unnatural amino acid residues such as pseudoprolines and pipecolic acid that can mimic the heterocyle incorporated as the final residue in the natural product. Our results show that OscGmac recognises pseudoprolines before AYD(G) to process a linear peptide.

No description available.

This thesis presents results for the extraction, purification and chemical characterisation and quantification of cyclic peptides. The thesis is divided into six chapters including a general introduction, the materials and methods used, in addition to three chapters with detailed experimental results and a chapter on enzyme chemistry. The first part presents data of a chemically synthesized analogue of a natural product myriastramide C which contains its full structural elucidation and comparison of data obtained to that of the natural product reported in literature. The second part discusses enzyme chemistry, chemical characterisation results of natural, new modified heterocyclic peptides produced in vitro and through chemo enzymatic reactions using genetically modified enzymes. Results presented include mass spectroscopic data and NMR structural characterisation for these peptides. The third section presents data for the quantification of heterocyclic peptides using high pressure liquid chromatography coupled in parallel to electrospray mass spectrometer and inductively coupled plasma (HPLC-ICP-MS), their quantification was achieved using their sulfur content without authentic standards. Naturally occurring cyclic peptide were also quantified using proton nuclei magnetic resonance (qNMR) with a non peptidic ERETIC reference material. In conclusion, this work highlights the possibility of multi-disciplinary science in the production of synthetic and semi-synthetic compounds through the use of enzymes. It is possible to produce new analogs of natural products, hence, providing an avenue for increasing the library of new compounds. Accurate quantification of these compounds is also essential for the acquisition of proper pharmacokinetic data for these new compounds so that unambiguous biological data can be obtained.

Cyclic peptides constitute an attractive class of compounds for drug development, however the numerous problems associated with their synthesis have limited their applicability. The cyclisation step itself is particularly problematic, with solution phase cyclisations being required to be conducted under very high dilution to promote cyclisation over unwanted side reactions such as oligomerisation. In addition, epimerisation, leading to the loss of chiral integrity at the terminal residues is a major concern. Attention is now turning to biochemical cyclisation strategies, such as SICLOPPS and sortase mediated ligation, although these also come with their own inherent disadvantages, for example, in the case of sortase mediated ligation, there is significant “scarring” of the target due to the presence of a four amino acid long recognition sequence. Cyclisation using ribosomally synthesised post translationally modified enzymes is also gaining popularity. One such family of enzymes is the patellamides. PatGmac is capable of performing cyclisations on linear peptide substrates with minimal scarring compared to the aforementioned alternatives and, importantly, with no epimerisation and could constitute a greener and more facile route to cyclic peptides. The work herein details some of the investigations designed to define the range of synthetic utility and test the flexibility of the enzymes. This was done qualitatively, by designing a variety of linear peptide analogues of the natural product, homophymine A, featuring unique structural moieties and evaluating their compatibility with the enzyme. It was also done quantitatively, using an LCMS based semi-quantitative strategy, to assess differences between similar, but different, enzymes and to assess whether there were differences in how different substrates are processed by the same enzyme. In addition to this, a variety of these substrates were also assessed for their proclivity to cyclise under standard chemical conditions for comparison. Lastly, with the increasing appearance cyclic peptides and non-peptidic macrocycles in the libraries of compounds being considered for clinical trials, there is now a growing need for computational modelling of these structures. Herein, a 3D v structure for the natural product callipetin N is proposed for the first time, determined using a combination of computational and NMR techniques.

Thesis (Ph.D.) - University of Queensland, 2006. / Includes bibliography.

Thesis (Ph.D.) - University of Queensland, 2006. / Includes bibliography.

Thesis (Ph.D) - University of Queensland, 2006. / Includes bibliography.

Abstract not available