Bicyclic peptides are becoming an increasingly prominent class of potential next-generation therapeutics. Owing to their unique properties, they are capable of bridging the gap between small molecules and antibodies.^{1, 2}
The most frequently used synthetic procedures to generate bicyclic peptides rely on the formation of disulfide or thioether bonds and often require special reagents or catalysts.^{2, 3} We present a substantially different bicyclisation strategy, based on the catalyst-free, biocompatible ‘click’ reaction between cyanopyridine and 1,2-aminothiol, previously developed in the Nitsche group.^{4, 5}
A novel reagent enables the synthesis of unnatural amino acids through its reaction with the nucleophilic side chains of cysteine and lysine-like amino acids.6 These dicyanopyridine-featured amino acids can be generated as Fmoc derivatives for incorporation into standard solid phase peptide synthesis (SPPS). Alternatively, dicyanopyridine amino acids can be synthesised on the solid support during SPPS via the use of inexpensive orthogonally protected amino acids. The bicyclic peptide is subsequently formed within minutes through a spontaneous intramolecular reaction in aqueous solution at physiological pH. The process is orthogonal to all canonical amino acids and fully amenable to automation.
To demonstrate the great potential for drug discovery, we generated a bicyclic inhibitor of the Zika virus protease NS2B-NS3, which is derived from the protease’s recognition sequence. Bicycles can also be further modified post-synthetically, for example, to generate tetracyclic peptides.
This methodology pushes the boundaries of bicyclic peptide synthesis, informs next-generation drug design and is likely to be compatible with genetically encoded peptide libraries.