Drug discovery efforts face a major challenge when attempting to target large protein surfaces, as small molecules cannot easily bind to them. Peptides show promise for addressing these ‘undruggable’ targets owing to their moderate molecular weight, inherent biocompatibility, and relatively simple synthetic accessibility. However, peptide drug discovery often needs the creation of entirely new peptide libraries for each individual programme, modifications to the peptide topology, and massive drug screening efforts.

The researchers tested their strategy targeting the hepatitis C virus protease NS3a. A specific motif, HXDMT (where X = random amino acid), retained its structure and binding strength to NS3a after peptide bicyclizations using KYL chemistry. One sequence, SQVHCYHGDMTMPIC, showed Kd = 21.0 nM before modification and Kd = 29.8 nM after bicyclization, (where Kd is the dissociation constant, which measures the binding affinity between a ligand and its target), indicating that reshaping doesn’t just preserve the binding ability of some peptides, it can also enhance their stability against protease degradation compared to linear counterparts. This suggests that certain peptides are ‘developable’, capable of withstanding substantial chemical changes. It indicated that discovery efforts contain two classes of peptides: the optimizable peptides that can be reshaped without losing function and the drop out peptides that are unable to tolerate modulations. These two classes are likely to exist in every screen and their ratio will depend on the topology of the library and the target. Derda and colleagues proved that the discovery of chemically and topologically upgraded peptide drugs does not need to start from a naive chemically upgraded peptide library. This offers a ‘shortcut’ for the discovery of new peptide ligands. The ‘later but better’ concept has great potential for other types of genetically encoded libraries, such as mRNA, yeast, and cell surface displayed libraries.