Substrate-Based Design of Human Farnesyltransferase Peptide-like Pain Antagonists

Human protein farnesyltransferase is a key enzyme for the lipid modification of a large and important number of proteins, which has been recognized as the promising therapeutic target of pain disorder and other diseases such as inflammation and cancer. In this study, we systematically investigated the binding behavior of existing peptide substrates and antagonists to farnesyltransferase at structural level and revealed that peptide’s C-terminus is primarily responsible for the binding, while exposing N-terminus to solvent. The amino acid property preference profile at each of the six core N-terminal residue positions of a cocrystallized chimera peptide substrate was defined, based on which a combinatorial library that contains more than twenty thousands of peptide-like compounds (PLCs) was generated using sixteen structurally diverse non-natural amino acids as building blocks. Subsequently, a systematic protocol was exhaustively carried out to perform virtual screening against the library, in order to discover those PLCs that match well the property preference profile and simultaneously exhibit high binding potency to farnesyltransferase. Consequently, ninety hits were identified from the library, in which five structurally diverse PLCs with high consensus scores were determined to have potent or moderate affinity to the active site of farnesyltransferase through nonbonded/coordination interactions. These identified PLCs can be considered as promising lead molecular entities to further develop peptidomimetic farnesyltransferase antagonists combating pain, inflammation and cancer.