Gilvocarcin V (GV) and ravidomycin (RMV) exhibit excellent antitumor activities in the presence of near-UV light at low concentration maintaining a low in vivo cytotoxicity. Although, the exact molecular mechanism for in vivo actions of these antibiotics has yet to be determined, a [2+2] cycloaddition reaction of the vinyl side chain with DNA thymidine residues in addition to the inhibition of topoisomerase II and DNAhistone H3 cross-linking are reported for the GV’s mechanism of action. Such activities have made these molecules interesting candidates for the biosynthetic investigation to generate analogues with improved activity/solubility. Previous biosynthetic studies have suggested that the GV biosynthetic pathway involves a number of synchronously occurring transformations leading to the oxidative C-C bond cleavage and other intriguing biosynthetic reactions, such as the vinyl side chain formation, methylations, Cglycosylation and dehydrogenation. Although gene inactivation results identified many candidate genes whose corresponding enzymes are involved in these biochemical transformations, their exact functional roles and the identity of their natural substrates remained elusive. To provide more insights into these complex biochemical tranfrormations, three specific aims were set up. Specific aim 1 was to clone and characterize the RMV biosynthetic gene cluster. Through the comparison of GV cluster with the RMV cluster, the genes encoding the biosynthesis of sugar and tetracyclic aromatic moieties were identified. RavGT, the sole glycosyltransferase of the RMV cluster has demonstrated to have unprecedented sugar donor substrate flexibility, transferring an amino-pyranose sugar as well as a neutral furanose sugar. Specific aim 2 was to characterize all of the TDP-D-ravidosamine biosynthetic enzymes. The aim also included to a one-pot enzymatic synthetic protocol for the routine production of TDP-D-ravidosamine. Specific aim 3 focussed on a total enzymatic synthesis of defucogilvocarcin M (defucoGM), the polyketide-derived core of GV and RMV. This aim clearly identified the minimal enzymes required to biosynthesize the complex architecture of defucoGM from the simple building blocks acetate and malonate. In addition, the GV-pathway enzyme GilR was fully characterized. Through in vitro studies, GilR was shown to catalyze the dehydrogenation of hemiacetal moiety of the penultimate intermediate pregilvocarcin V to the lactone moiety of GV at the last step.

Natural products continue to be a major chemical lead matter for drug discovery due to their diverse chemical structures and bioactivities. Clinically significant natural products include anti-cancer and anti-infective compounds and while many more of these compounds show promising bioactivity, their clinical relevance is often limited by toxicity or poor solubility. Combinatorial biosynthesis can be employed to modify existing chemical scaffolds towards reducing these limitations. To fully take advantage of these biochemical tools, it is important to understand the biosynthesis and mechanism of action of the molecules. Saccharides in glycosylated natural products provide specific interactions with cellular targets and are often crucial for a compound’s bioactivity. Genetic engineering of sugar pathways can modify glycosylation patterns leading to the diversification of natural products. Saquayamycins (SQN) H and I are cytotoxic angucycline antibiotics containing five deoxyhexoses including the rare amino sugar rednose. Elucidating the biosynthetic pathway of rednose could add to the arsenal of combinatorial biosynthesis tools for drug development. Our research goal of investigating the rednose biosynthetic pathway was pursued through two specific aims: the identification of the Streptomyces sp. KY 40-1 gene cluster involved in the biosynthesis of SQN H and I (sqn) (specific aim 1), and the validation of the proposed L-rednose biosynthetic pathway up to the glycosyl transfer through enzymatic synthesis of NDP-3,6-dideoxy-L-idosamine (specific aim 2). The sqn gene cluster revealed deoxysugar biosynthetic genes that could be used to alter glycosylation patterns to generate novel compounds while the enzymatic synthesis afforded novel genetic engineering tools to generate novel TDP-deoxysugars that could be used to diversify compounds such as aminoglycosides to circumvent resistance mechanisms. The first step to generate TDP-glucosamine enzymatically was accomplished, however later steps were unsuccessful. The aureolic acid mithramycin (MTM) was recently tested in clinical trials for Ewing sarcoma following the discovery of MTM as a potent inhibitor of the oncogenic transcription factor EWS-FLI1 present only in Ewing sarcoma cells It is understood that MTM binds the minor groove of G/C rich DNA as an Mg2+-coordinated dimer disrupting transcription of proto-oncogenes; however, the DNA recognition rules were not completely understood, making further interrogation of MTM’s DNA binding preferences necessary. This research goal of further understanding the mechanism of action for MTM was approached through two specific aims: the investigation of the dimerization of MTM (specific aim 3), and the investigation of MTM’s DNA binding preferences (specific aim 4). This work established that MTM and its biosynthetic precursor premithramycin B (PreMTM B), and several MTM analogues with modified 3-side chains: mithramycin SDK (MTM SDK), mithramycin SA tryptophan (MTM SA-Trp), and mithramycin SA alanine (MTM SA-Ala) dimerize even in the absence of DNA under physiologically relevant conditions. The study also demonstrated that modification of the 3-side chain modulates DNA binding affinity of MTM analogues, established a minimum MTM binding site on DNA, and revealed MTM DNA recognition is driven by direct (sequence) and not indirect (conformation) readout laying the foundation for subsequent research based on the interaction between MTM, DNA, and the oncogenic transcription factor EWS-FLI1 in the rational design of new MTM analogues for the treatment of Ewing sarcoma.