The Passive Permeability Landscape Around Geometrically Diverse Hexa- and Heptapeptide Macrocycles

Recent advances in DNA and mRNA encoding technologies have enabled the discovery of high-affinity macrocyclic peptides and peptide-like ligands against virtually any protein target of interest. Unfortunately, even the most potent biochemical leads from these screening technologies often have weak cellular activity due to poor absorption. Biasing such libraries towards passive cell permeability in the design phase would facilitate development of leads against intracellular targets. We set out to empirically evaluate the intrinsic permeability of thousands of geometrically diverse hexa- and heptapeptide scaffolds by permuting backbone stereochemistry and N-methylation, and by including peptoid and β-amino acid residues at select positions, with the goals of providing a resource for biasing library-based screening efforts toward passive membrane permeability and studying the effects of the backbone elements introduced on a large number of compounds. Libraries were synthesized via standard split-pool solid phase peptide synthesis, and passive permeability was measured in pools of 150 compounds using a highly multiplexed version of the parallel artificial mem-brane permeability assay (PAMPA) under sink conditions. Compounds were identified using CycLS, a high-resolution mass spectrometry-based method that uses stable isotopes to encode stereochemistry and matches MSMS data to virtual fragment libraries based on the expected macrocyclic products. From the compounds that were identified with high confidence, 823 hexameric and 1330 heptameric scaffolds had PAMPA permeability coefficients greater than 1x10-6 cm/s. The prevalence of high permeability compounds in these two libraries suggests that passive permeability is achievable for hexa- and heptapeptides with highly diverse backbone geometries.