Tulane University (USA)
Peptides and their analogs such as peptide nucleic acids (PNA) are promising tools and therapeutics, but the cell membrane remains a barrier to intracellular targets. Conjugation to classical cell penetrating peptides (CPPs) such as HIV-tat or penetratin facilitates some delivery, however efficiencies are low. Such cell penetrating peptides have potential utility for the cellular delivery of polar compounds, cancer therapy, biosensor design. Yet, despite decades of research on thousands of known examples, useful sequence-structure-function relationships are essentially unknown. Because peptide-membrane interactions within the highly fluid bilayer are dynamic and heterogeneous, accounts of mechanism are necessarily vague and descriptive, with little predictive power. This lack of explicit design principles hinders rational optimization. Here, we use synthetic molecular evolution, iterative peptide library design and orthogonal high-throughput screening, to identify gain-of-function CPPs with dramatically improved ability to deliver cargoes to cells at low concentration. A CPP library containing 8,192 tat/penetratin hybrid peptides coupled to an 18-residue PNA was screened using the Hela pTRE-LucIVS2 splice correction reporter system. The daughter CPPs identified are one to two orders of magnitude more efficient than the parent sequences at delivery of PNA, and also deliver a dye cargo and an anionic peptide cargo. The significant increase in performance following a single iteration of SME demonstrates the power of this approach to peptide sequence optimization.