Design of the influenza virus inhibitors targeting the PA endonuclease using 3D-QSAR modeling, side-chain hopping, and docking.

With the emergence of drug resistance and the structural determination of the PA N-terminal domain (PAN), influenza endonucleases have become an attractive target for antiviral therapies for influenza infection. Here, we combined 3D-QSAR with side-chain hopping and molecular docking to produce novel structures as endonuclease inhibitors. First, a new molecular library was generated with side-chain hopping on an existing template molecule, L-742001, using an in-house fragment library that targets bivalent-cation-binding proteins. Then, the best 3D-QSAR model (AAAHR.500), with q(2)=0.76 and r(2)=0.97 from phase modeling, was constructed from 23 endonuclease inhibitors and validated with 17 test compounds. The AAAHR.500 model was then used to select effective candidates from the new molecular library. Combining 3D-QSAR with docking using Glide and Autodock, 13 compounds were considered the most likely candidate inhibitors. Docking studies showed that the binding modes of these compounds were consistent with the crystal structures of known inhibitors. These compounds could serve as potential endonuclease inhibitors for further biological activity tests.