Nat. Chem. 6, 877–884 (2014)

Historically, drug discovery has been driven by screening the products of microbial fermentation for bioactive nonribosomal peptides. Although such serendipity served chemists well in the past, the future lies in more directed approaches. Huang et al. describe how to 'grow' desired synthetic peptides, using off-the-shelf building blocks and amide-mediated ligation. The process uses three types of components—initiators, elongation monomers and terminators—that are mixed in different ratios in small volumes of aqueous, reagent-free solutions. The peptide sequence can be regulated by altering the order of elongation monomer addition. The approach offers several benefits: the peptides can be tested for bioactivity without further purification, and the ease of synthesis eliminates the need to store the products as they can be readily regenerated. The combinatorial nature of the approach makes accessible a vast space of potential compounds. For example, using just 23 components that approximate parts of a known inhibitor of the hepatitis C virus protease, which is recognized as a difficult target owing to its lack of a well-defined active site, the authors generated 6,000 lead molecules. From these, a compound with a half-maximum inhibitory concentration of 1 μM was identified. This approach may have more general applications, however; molecules that have no biological activity but are more relevant to materials science or chemical catalysis might also be constructed by the appropriate choice of elongation monomers.