Project Summary Nonribosomal peptides (NRPs) are an abundant class of microbial natural products: small molecules with evolved chemical structures and biological activities that make them an essential source of pharmaceuticals. NRPs have served as leads for antifungal, anticancer, and immunosuppressant drugs, but are particularly important for antibacterial therapeutics, with over 40% of approved antibiotic classes being NRPs or derivatives. The diverse structures and activities of NRPs are generated by a conserved class of proteins known as nonribosomal peptide synthetases (NRPSs), which are composed of multi-domain modules that progressively build NRPs as enzymatic assembly-lines. As the composition and sequence of domains within each module determine which of >500 known monomers will be incorporated into the growing NRP, NRPS engineering has been viewed as a promising approach for producing designer molecules in live cells. However, while ribosomal peptides can be reprogrammed through the genetic code, NRP sequences are determined by the “adenylation (A) domain code”, which describes a series of conserved residues and positions within the amino acyl binding pocket. Unlike the genetic code, the A-domain code is not directly programmable, and attempts to re-write this code and which amino acid is activated by a given domain generally results in poorly active enzymes. Transplantation of binding pockets can occasionally yield domains and modules with desired specificity, but this often fails and still relies on access to characterized protein donors. Directed evolution is a promising approach for re-coding NRPSs, but current methods are slow, manual, and bespoke solutions that rely on specific activities from the resulting natural product or on reactive chemical handles. In this proposal, we will develop a plug-and- play platform for the continuous directed evolution of NRPS activity and substrate specificity. This approach uses promiscuous NRPS components known as epimerization domains and type II thioesterase domains to convert amino acids selected by A domains into (D) isomers. As (D) amino acids (D-AAs) are not present in the cell, they can be selectively detected and quantified by natural D-AA transcription factors, linking NRPS activity and specificity directly to gene expression. In aim 1, we will demonstrate the generation of D-AA signals from NRPS modules and optimize D-AA transcription factors, enabling live cell reporting for NRPS substrate specificity. In aim 2, we will use this system to drive phage-assisted continuous evolution (PACE) of NRPS modules, providing a rapid and adaptable means of evolving highly functional enzymes with desired substrate specificities. This proposal will deliver the first live-cell reporter system for NRPS substrate specificity and enable the rapid evolution of NRPS components with designer specificities. These advances will unlock our ability to program these pharmaceutical assembly-lines for on-demand production of designer molecules in live cells.

Project Narrative Many of the drugs we use are derived from microbial compounds known as nonribosomal peptides, which are produced by modular assembly-line proteins. While these assembly-lines can evolve in nature and enable microbes to make new drug-like chemicals, we haven’t been able to effectively replicate this process in the lab. In this proposal, we will establish a plug-and-play platform to evolve and engineer modules within assembly- lines, allowing us to program the production of new molecules in living bacteria that could be used as probiotic therapies or producers of new drugs.