Structurally Decoding the Biosynthetic Logic of Modular Polyketide Synthases

PROJECT SUMMARY/ABSTRACT Microorganisms employ enzymatic assembly lines to produce myriad complex small molecules with medicinal importance (e.g., antibiotics, anticancer agents, and lipid-lowering statins). Each among thousands of naturally occurring assembly lines generates a structurally distinct compound by coordinating multistep biosynthesis across a defined sequence of enzyme active sites. In addition, their modular structures provide a natural framework for product diversification, as assembly lines can recombine into seemingly endless enzyme configurations. While decades of assembly-line research have accurately pinpointed the enzymes involved and their relative timing in the biosynthesis of hundreds of natural products, we cannot yet explain how these pathways are pre-programmed to commit their singular reaction sequences; much less, re-program them to specify new reaction sequences while maintaining catalytic integrity. Understanding how modular biosynthetic pathways are encoded at the level of protein sequence could be harnessed to engineer the biosynthesis of truly unparalleled libraries of user-defined chemical structures (>1050). Such advances have the potential to transform human medicine while simultaneously unlocking sustainable methods for chemical synthesis. To this end, our group has focused on decoding the mechanisms of a highly prevalent and versatile group of enzymatic assembly lines: the polyketide synthases (PKSs). Over the past year, we have made progress towards understanding polyketide antibiotic biosynthesis by a model PKS assembly line, the rifamycin synthase (RIFS). We have expressed and purified as multiple fragments ~40% of the 3.4 MDa RIFS assembly line and reconstituted formation of its natural diketide in vitro. Preliminary structural analysis of RIFS by single-particle cryogenic electron microscopy (cryo-EM) has revealed the architectures of its initial PKS modules, including a defunct dehydrating module, as well as a putative new mode of carrier-protein mediated substrate shuttling. This MIRA proposal presents a 5-year plan to continue structure-function analysis of the RIFS model system while implementing ancillary techniques that can test structural observations both experimentally and computationally. We present applications of site-selective and symmetry-unbiased crosslinking combined with mass spectrometry for (1) conformational probing and (2) cryo-EM sample preparation of PKS modules and bimodules. Building off prior developments, we propose a ‘peripheral non-invasive’ crosslinking method for capturing structures of fleeting substrate- or product-bound module states. In collaboration with Dr. Muyuan Chen, we are using machine-learning approaches to extract information about continuous motion (module dynamics) from cryo-EM data. Finally, we are developing in vitro assays for continuous turnover of truncated assembly lines by exploiting promiscuous thioesterase domains that promote polyketide intermediate off-loading. In summary, the proposed research applies innovative structure-function tools to understand mechanisms of polyketide biosynthesis that can engender the design of artificial biocatalysts for sustainable production of copious medicinal compounds.