Proposal Summary Bacteria manufacture a diverse range of natural products with pharmaceutical value as potent antibiotics and chemotherapeutic agents, many of which are FDA-approved. Two well biosynthetic systems, the modular polyketide synthases (PKS) and non-ribosomal peptide synthetases (NRPS), are major sources of pharmaceutical natural products. These PKS and NRPS systems are organized into assembly lines of multi- domain modules, where each module extends a biosynthetic intermediate by an acyl unit (PKS) or amino acid (NRPS). PKS and NRPS are generally encoded by gene clusters and synthesize natural products that can be predicted from the gene sequence alone. This makes them very attractive targets for engineering to create new or purposefully altered compounds. Each PKS or NRPS module contains a carrier domain (CP) to tether the pathway intermediate. After the action of a module is complete, the growing chain must be passed to the next module. To ensure pathway fidelity when sequential modules are on different polypeptides, “docking domains” at the polypeptide termini facilitate transfer of the intermediate from the CP of the upstream module to the first catalytic domain of the correct downstream module. A recently discovered cyanobacterium produces three classes of vatiamides, natural products with the ability to kill lung cancer cells. Remarkably, the three vatiamides are synthesized by a branched hybrid PKS/NRPS pathway encoded by a single gene cluster. Intermediate transfer at the branch point is enabled by the natural duplication of a docking domain, allowing the donor module VatM to deliver its product to either VatN, VatQ, or VatS. Here I will test the competing hypotheses that docking domains can be exploited as engineering tools or that CP-enzyme selectivity is critical to intermediate transfer. I will characterize the branch point of the vatiamide pathway by measuring the affinity of VatM for VatN, VatQ, and VatS. Because the VatN, VatQ, and VatS docking domains are identical, any difference in affinity will be due to the composition of the downstream modules, which may be the main obstacle to be engineering. I will test the feasibility of this engineering strategy by creating noncanonical branches in two different pathways. I expect that installation of the same docking domain across multiple modules will facilitate flux through both natural and artificial interfaces. Applying this strategy to existing drug biosynthetic pathways could simultaneously create multiple analogs that could be screened for increased potency and fewer side effects.