An optimized MM/PBSA virtual screening approach applied to an HIV-1 gp41 fusion peptide inhibitor.

VIRus Inhibitory Peptide (VIRIP), a 20 amino acid peptide, binds to the fusion peptide (FP) of human immunodeficiency virus type 1 (HIV-1) gp41 and blocks viral entry. VIRIP derivatives with improved antiviral activity have been developed, and one of those derivatives has recently proven effective and safe in a phase 1/2 clinical trial. Here, molecular dynamics were executed in combination with molecular mechanics/Poisson-Boltzmann surface area (MM/PBSA) free energy calculations to explore the binding interaction between VIRIP derivatives and gp41 FP. A promising correlation between antiviral activity and simulated binding free energy was established thanks to restriction of the flexibility of the peptides, inclusion of configurational entropy calculations, and the use of multiple internal dielectric constants for the MM/PBSA calculations depending on the amino acid sequence. Based on these results, a virtual screening experiment was carried out to design VIRIP analogs with further improved antiretroviral activity. A selection of peptides was tested for inhibitory activity and several VIRIP derivatives were identified with significantly enhanced activity compared to the reference peptides. The results demonstrate that computational modeling strategies using an adapted MM/PBSA methodology improve the accuracy of binding free energy calculations of peptide complexes compared to the classic MM/PBSA protocol. As such, this virtual screening approach generated HIV-1 gp41 FP inhibitors with improved antiviral activity that could be useful for future clinical applications.