You are here

Insights into Ligand Deconstruction by Computational Solvent Mapping

Today in the Proceedings of the National Academy of Sciences, Boston University in collaboration with Acpharis published critical insight into a long-standing question in the field of Fragment-based drug discovery (FBDD): when existing ligands are “deconstructed” into constituent fragments, often the fragments do not bind in the same place as they did in the larger ligand. This presents a crucial question for those attempting to apply FBDD; will the binding mode of my fragment be conserved when it is grown into a larger ligand? To address this, our team mapped seven proteins where ligand deconstruction studies had been performed. Across all seven targets, a simple rule emerged: when a fragment covers a sufficiently large portion of the top hot spot in mapping, its binding mode is conserved.

The top hot spot forms an anchor site, which dominates the free energy of binding. An optimized fragment binding in this anchor site cannot be pulled away into other nearby hot spots that form weaker sites that are exploited as a fragment is grown into a lead-like molecule. To further test this hypothesis, Boston University followed up the seven retrospective analyses with a prospective study deconstructing an IL-2 inhibitor into two fragments, one predicted to retain its binding mode and one predicted to possibly not retain its binding mode. NMR chemical shift perturbations were consistent with the predictions from mapping, with only the first fragment showing conservation of binding at the same site.

Read more at PNAS