Iterative Constructing and Screening of DNA-Encoded Libraries Enable the Discovery and Structure Optimization of Macrocyclic Hit Compounds

Authors:

Abstract
Macrocyclic peptides occupy a privileged "middle space" in drug discovery, offering the specificity of biologics combined with the synthetic tractability of small molecules. However, the rapid identification and structural optimization of macrocyclic leads for challenging targets remain formidable tasks. Here, we report a robust DNA-encoded library technology (DELT) platform designed for the synthesis, screening, and iterative optimization of novel dimeric macrocyclic peptides. Initially, a pooled library of approximately 1.2 billion unique members, cyclized via diverse thioether linkages or disulfide formation, was constructed and screened against the oncogenic target MDM2. This first-generation selection yielded compound Comp_1, which incorporates a highly enriched pharmacophoric fragment, F_1, and displays a 𝐾i value of 1.5 𝜇M. Leveraging the synthetic modularity of our platform, we established an iterative optimization campaign by incorporating F_1 into a second-generation asymmetric dimeric library (DEL_2) synthesized with enriched building blocks. A subsequent selection campaign against DEL_2 led to the discovery of highly potent macrocyclic candidates, exemplified by compound Comp_2 with a 𝐾i value of 18 nM—representing an approximate 100-fold affinity enhancement over the first-generation hit. This work not only highlights MDM2-targeting macrocyclic binders but also establishes a highly efficient, generalizable DELT-based iterative workflow for the rapid evolution of peptidic leads.