Experimental and computed nuclear magnetic resonance data and an iterative synthetic strategy have revealed the correct structures of the baulamycins, potentially important antimicrobial compounds, allowing them to be chemically synthesized.
Small-molecule, biologically active natural products continue to be our most rewarding source of, and inspiration for, new medicines1. Sometimes we happen upon such molecules in minute quantities in unique, difficult-to-reach, and often fleeting environments, perhaps never to be discovered again. In these cases, determining the structure of a molecule—including assigning its relative and absolute configurations—is paramount, enabling one to understand its biological activity. Molecules that comprise stereochemically complex acyclic and conformationally flexible carbon chains make such a task extremely challenging2. The baulamycins (A and B) serve as a contemporary example. Isolated in small quantities and shown to have promising antimicrobial activity, the structure of the conformationally flexible molecules was determined largely through J-based configurational analysis3,4, but has been found to be incorrect. Our subsequent campaign to identify the true structures of the baulamycins has revealed a powerful method for the rapid structural elucidation of such molecules. Specifically, the prediction of nuclear magnetic resonance (NMR) parameters through density functional theory—combined with an efficient sequence of boron-based synthetic transformations, which allowed an encoded (labelled) mixture of natural-product diastereomers to be prepared—enabled us rapidly to pinpoint and synthesize the correct structures.