Design of peptide-based coronavirus inhibitors that target disruption of 3CLpro protease self-association

Abstract

3CLpro is a highly conserved main protease found in coronaviruses, which makes it an attractive target for the development of broad-spectrum coronavirus inhibitors. Its homodimer plays an essential role in coronavirus replication. The SGFRKMAF peptide spanning the N-finger domain of 3CLpro is known to disrupt dimerization, however, its mode of action is not fully understood. We used the conformation of this peptide segment in the observed protein crystal structure as the basis for developing leads for coronavirus dimerization inhibitors. We investigated the interaction of the SGFRKMAF peptide with the 3CLpro monomer in blind docking simulations using a representative ensemble of its conformational states. We identified 8 basins of attraction, distinct regions where the peptide accumulates at the 3CLpro monomer surface. Two binding regions dominate: one, located at the groove between domains II and III of the monomer, and the other at the interface region where dimerization takes place. Peptide binding at these two regions resulted in stable peptide–protein complexes in 90 ns fully solvated molecular dynamics simulations. Using protein–protein docking simulations, we found that peptide binding to at least one of the monomers at the dimer interface region is likely to disrupt 3CLpro dimerization through blocking the “hot spot” residues that predominantly account for dimer stabilization. Peptide binding to the interface region in one of the monomers, and to the groove between domains II and III in the other monomer, leads also to disruption of the native dimer structure. The peptide binding constant at the interface region relative to other regions (K_in/out) was estimated to be ∼0.12 at 310 K, suggesting that at thermodynamic equilibrium the peptide does not solely bind to one of the basins, corroborating the view of a cooperative mechanism between the two basins. In order to optimize the potential of the SGFRKMAF peptide to disrupt 3CLpro dimerization via preferential binding to the interface region, we conducted systematic mutation of the M6 and F8 residues, sites identified as key to blocking hot spot regions in the observed dimer structure. We found that the [M6F, F8S], [M6I, F8Q], [M6Q, F8T] and [M6T, F8I] mutations result in an increase of K_in/out by at least one order of magnitude, with the [M6F, F8S] mutation (the SGFRKFAS peptide) showing the highest value of K_in/out (∼2.53). These mutant peptides are, therefore, candidates for peptide-based lead structures in the development of broad-spectrum coronavirus inhibitors.