Abstact

Linking of fragments in neighboring binding sites is one of the optimization strategies in fragment-based drug discovery, where additive or even more substantial bioactivity improvements can be realized. However, such efforts present a considerable challenge when one fragment binds covalently to the target protein, as small modifications can influence the correct positioning of the covalent warhead toward the targeted nucleophilic residue. Here, we present a case study of fragment linking that yielded single-digit micromolar, covalent inhibitors of the SARS-CoV-2 main protease, starting from fragments that were inactive in the biochemical assay. Using structural information from a recent, high-throughput crystallographic fragment screen, we show that the success of fragment linking in the design of targeted covalent inhibitors is heavily impacted by several factors, including the warhead type, the labeling chemistry, and even subtle changes in the designed linker. Notably, we observe that induced fit effects might override the original fragment orientations in the linked molecule, highlighting the need for reliable structure verification, especially in consecutive rounds of fragment elaboration.