Combinatorial biosynthesis of polyketides: Oxygenases in angucycline antibiotic biosynthesis

Nature has traditionally served as an excellent source for many bioactive compounds. Especially the large family of polyketides, which are produced by as wide a variety of organisms as plants, algae, fungi, and bacteria (predominantly Streptomyces species), displays a high degree of structural diversity and therefore a wide range of biological activities. As a result, this family of natural products accounts for many successful drugs in clinical settings, such as erythromycin, amphotericin, and doxorubicin. 'Combinatorial Biosynthesis' is an emerging strategy for the design and generation of novel polyketides with improved activity and/or toxicity profiles, which is based on genetic engineering of biosynthetic pathways by inactivation, recombination, or manipulation of selected biosynthetic genes from one or several different organisms. An important prerequisite for this method is the exact knowledge of the catalytic mechanism and relaxed substrate specificities of the involved enzymes. The gene products with the most profound impact on the activities of the final polyketides are post-PKS tailoring enzymes, particularly glycosyltransferases and oxygenases, which therefore form the most attractive targets for combinatorial biosynthesis approaches. The work presented here focuses on oxygenases involved in the biosynthetic pathways of angucyclines, a polyketide subgroup with an especially intriguing variety of displayed biological properties. Our attention concentrated in particular on the landomycin, urdamycin, and especially jadomycin biosyntheses, the latter showing a rare oxidative rearrangement of the angucycline ring system. Specific Aim 1 addresses the unique amino acid incorporation subsequent to this oxidative ring fission and thus evaluates its importance for the biological activity of the jadomycins. A variety of amino acid analogues were generated, and preliminary information with regard to chemical structures and structure-activity relationships was gained. Specific Aims 2 and 3 were designed to investigate the mechanistic aspects of this cleavage reaction and assess the substrate specificity of the putatively involved oxygenase JadF. In this context, the mechanisms of three consecutive oxygenation steps were identified and the respective enzymes JadF, G, and H were assigned. Furthermore, additional insight into the mechanism of the UrdM oxygenase in the urdamycin pathway was obtained. In Specific Aim 4, several novel landomycins were generated, which allowed the unambiguous assignment of the LndZ5 oxygenase and the LndGT4 glycosyltransferase, providing a more detailed picture of the landomycin biosynthetic pathway. In addition, these compounds were utilized to garner an improved understanding of the structure-activity relationships regarding the potent anticancer activity of the landomycins.