Chemoenzymatic Assembly of Macrolactones: Merging Synthetic Chemistry with Biocatalysis to Afford Diverse Compound Libraries

Abstract

Complex secondary metabolites display a wealth of biological activities and, together with their derivatives, have provided over 60% of new pharmaceutical agents over the past 40 years. Despite their clinical success, limitations in isolation yields as well as the synthetic challenges posed by these scaffolds often hinders the medicinal chemistry efforts necessary to overcome suboptimal pharmacological properties, highlighting the need for alternative methods. A promising strategy for generating libraries of natural product analogues is through the use of biosynthetic enzymes. These systems have long been hypothesized to be capable of producing almost unlimited structural diversity, due to their modular nature, however to date, efforts towards PKS engineering have mostly met with failure. Despite notable successes, decreased product yields and/or failure to produce the desired structures entirely have stymied these efforts. Recent studies focusing on the interrogation of single modules or module domains via biochemical analysis coupled with structural determination have begun to shed light on these complex systems and give us insight into the reasons for the initial failures. The studies presented in this thesis focus on investigations into the structural and mechanistic parameters that govern selectivity in the biosynthetic enzymes of interest from two key natural products, the PKS/NRPS derived cryptophycin family of anticancer agents, and the PKS antibiotic Pikromycin. In the cryptophycin system, synthetic chain elongation intermediates have been coupled with the Crp TE macrocyclizing catalyst to produce a library of heterocyclic unit A analogues. This was met with remarkable success as all the analogues were processed by the TE with similar or greater efficiency than that seen with the native substrate. Current efforts are focused on gaining an NMR structure of this TE with hopes of shedding light on the underlying catalytic mechanism that makes this enzyme so versatile. This facilitated the biological evaluation of these analogues, allowing us to identify one that displays remarkable activity without the presence of an epoxide group that was previously thought to be necessary for maximum efficacy. Utilizing the same strategy in the Pikromycin system, five new pentaketides analogues were generated that could be used with three separate intact PKS modules, the PikAIII-TE and the coupled PikAIII/AIV system. These synthetic intermediates have continued to lend credence to the hypothesis that in PKS systems, the TE tends to be the deciding factor on whether hydrolytic byproducts are formed or macrocycles. Utilizing our biocatalytic platform we have been able to show that the TE can more effectively produce 14 membered macrolactones containing unnatural functionality than 12, leading to the isolation of three new macrolactone products. Altered alkyl chain substituents on these pentaketides substrates have also shed light on the likelihood of a size restriction for the loading of these substrates onto the KS domain of the initial module. The continued investigation of these substrates as well as others continues to build the groundwork for future engineering campaigns aimed at generating more flexible catalysts for the production of novel natural product analogues.