Molecular insight into the π‐stacking interactions of human ovarian cancer PARP‐1 with its small‐molecule inhibitors and rational design of aromatic amino acid‐rich peptides to target PARP‐1 based on the π‐stacking network

Abstract

AbstractHuman poly(ADP‐ribose) polymerase 1 (PARP‐1) is involved in DNA strand break repair and has been recognized as a promising druggable target for ovarian cancer. The active site of PARP‐1 is π electron‐rich, which contains at least 12 aromatic amino aid residues in its first‐ and second‐shell regions, and can form effective π‐stacking interactions with its natural substrates and inhibitor ligands. In this study, the intermolecular interactions between the 12 PARP‐1 aromatic residues and 16 approved/clinical/preclinical small‐molecule inhibitors were systematically profiled using quantum mechanics/molecular mechanics (QM/MM) analyses and dispersion‐corrected density functional theory (DFT) calculations. It was found that the aromatic residues contribute ~50% affinity to the global binding energy of PARP‐1/inhibitor complexes, in which the four residues (His862, Tyr889, Tyr896 and Tyr907) play a primary role in π‐stacking interactions with most inhibitor ligands. Based on the harvested knowledge, we attempted to perform the rational design of aromatic amino acid‐rich (AAAr) peptides to target PARP‐1 active site by using peptide docking, affinity scoring, and QM/MM analyses. These peptides are only composed of four aromatic amino acids, namely phenylalanine (Phe), tyrosine (Tyr), tryptophan (Trp) and histidine (His), and their π‐stacking contribution is improved to ~60% from the ~50% of small‐molecule inhibitors, although their global binding energy is moderately lower than these inhibitors. Three designed AAAr peptides (WYHY, WYYH and WFYY) with high‐affinity scores were determined as potent PARP‐1 targeted agents; they exhibited high inhibitory activities and are roughly comparable with that of the sophisticated PARP‐1 inhibitor drug Olaparib, suggesting that these designed peptides, as expected, can effectively target PARP‐1 with a satisfactory inhibitory profile. Molecular modeling revealed a number of face‐to‐face, parallel‐displaced, and T‐shaped π‐stacking interactions at PARP‐1/peptide complex interface, which co‐define an intense π‐stacking network and confer both stability and specificity for the complex recognition and association.