The continued propagation of antibiotic resistance requires the development of new therapeutics. The lipopeptide antibiotic enduracidin has demonstrated high activity against Gram-positive pathogens including methicillin-resistant Staphylococcus aureus. In addition to a lack of cross-resistance with existing antibiotic classes, enduracidin has no known transferrable resistance mechanism. The development of enduracidin as a human therapeutic is hampered by its poor solubility in plasma. Utilizing chemical and genetic techniques, analogs of enduracidin have been produced and evaluated for biological activity. Making use of the hydroxyphenylglycine (Hpg) biosynthetic pathway, fluorine was incorporated into enduracidin with minimal to no loss of bioactivity. The semisynthetic chemical modification of enduracidin proved to be challenging. The chemical nitration of the Hpg residues was unsuccessful. Modifications to the lipid tail by cleavage at the C2-olefin with ozone and the use of Diels-Alder reagents to react with the lipid tail diene also proved unsuccessful. However, the reduction and dihydroxylation modifications of the lipid tail diene were successful. Introduction of polar hydroxyl groups onto the alkyl tail reduced bioactivity while reduction of the diene had no significant effect. Analysis of the biosynthetic pathways involved in producing the lipid tail and the unusual amino acid enduracididine yielded some insights into the formation of the antibiotic. Through complementation of mutants having disruptions in the biosynthetic gene cluster and crystallographic data, the function of EndR as a cyclase was established. Additionally, the use of 4-hydroxyarginine as an intermediate in enduracididine biosynthesis was demonstrated. The ability of EndQ to function as a transaminase on both 4-hydroxyarginine and 2-ketoenduracididine was also established. The specific functions of EndP and EndQ have not been determined. The introduction of the lipid tail diene by the three enzymes Orf39, Orf44 and Orf45 was confirmed. Orf45 functions as a CoA ligase and a dehydrogenase to introduce the C2 double bond. The functions of Orf39 and Orf44 appear to be the introduction of the C4 double bond and isomerization of the C2 olefin. / Graduation date: 2013

Type II polyketide synthases (PKSs) are responsible for the generation of structurally diverse and clinically important aromatic polyketides. The bacteriostatic agent enterocin (enc) isolated from the marine microbe "Streptomyces maritimus" is derived from a rare benzoate primer unit and contains a unique nonaromatic caged core structure resulting from a Favorskii-like carbon skeleton rearrangement. The apparent diversion between enterocin biosynthesis and all other type II PKS pathways offered the opportunity to discover novel enzymatic strategies that may be exploited to diversify the chemical structures of polyketides. A comprehensive biochemical analysis was performed in order to characterize the key steps in enterocin biosynthesis and finally to reconstitute the whole pathway in vitro using purified recombinant enzymes.A nonribosomal peptide synthetase (NRPS)-like priming mechanism was discovered for the selective activation of a benzoic acid starter unit and its subsequent attachment to the enc PKS to initiate polyketide biosynthesis. This is the first example of a type II PKS that employs an NPRS-like priming mechanism to utilize alternative non- acetate starter units. Secondly, the minimal enc PKS was reconstituted in vitro to give three novel acetate-primed metabolites that had never been identified by heterologous in vivo expression of recombinant enc PKS gene sets. The minimal enc PKS was then merged with the NRPS-like chain initiation module and the resulting multienzyme complex catalyzed the formation of benzoate-primed natural products wailupemycin F and wailupemycin G. Favorskii-like rearrangement of the nascent polyketide chain was replicated in vitro and the flavin-dependent oxygenase EncM was confirmed to be solely responsible for catalyzing this unprecedented rearrangement. Other biosynthetic steps in the late stage of the enc pathway were also replicated in vitro, including the methylation of desmethyl-5-deoxyenterocin to 5-deoxyenterocin and the hydroxylation of 5-deoxyenterocin to enterocin.Finally, the entire enc type II PKS pathway was successfully assembled in vitro using ten recombinant proteins and three commercial enzymes. Five enc-based natural products were generated from benzoic acid and malonyl-coenzyme A. This biochemical investigation on enterocin biosynthesis represents the first complete in vitro reconstitution of a type II PKS system and also provides an alternative strategy to create complex natural products by multienzyme synthesis.