Molecular and Thermodynamic Modeling of the Protein-Ligand Interaction. Application to Computer-Assisted Design of Anti-Competitive Inhibitors of Human Histone Deacetylase 2 (HDAC2)

Abstract

Background: Histone deacetylases (HDACs) are promising drug targets for a variety of therapeutic applications. Here we report in silico design and evaluation of novel amine-based hydroxamic acid derivatives (DAHAs), HDAC2 inhibitors with favorable predicted pharmacokinetic profiles. Methods: By using in situ modifications of the crystal structure of suberoylanilide hydroxamic acid SAHA-HDAC2 complex (PDB entry 4LXZ), 3D models of HDAC2-DAHAx complexes were prepared for a training set of 18 DAHAs with experimentally determined inhibitory potencies (half-maximal inhibitory concentrations IC50exp In the search for active conformations of the DAHA1-18, a linear QSAR model was prepared, which correlated computed gas-phase enthalpies of formation (∆∆HMM) of HDAC2-DAHAx complexes with the IC50exp. Further, taking into account the solvent effect and entropy changes upon ligand, binding resulted in a superior QSAR model correlating computed complexation Gibbs free energies (∆∆Gcom). The successive pharmacophore model (PH4) generated from the active conformations of DAHAs served as a virtual screening tool of novel analogs included in a virtual combinatorial library (VCL) of compounds containing hydroxamic acid scaffolds. The PH4 model to identify new DAHA analogs screened the VCL filtered by Lipinski’s rule-of-five. Results: Gas-phase QSAR model: -log10(IC50exp) = pIC50exp -0.2870 x ∆∆HMM +6.5764, R2= 0.83, superior aqueous phase QSAR model: pIC50exp =-0.4005 x ∆∆Gcom + 6.4402, R2=0.93 and PH4 pharmacophore model : pIC50exp = 1.0107 x pIC50pre - 0.0607, R2 = 0.92. The VCL of more than 198 thousand DAHAs was filtered down to 150,713 analogs Lipinski’s rule. The PH4 screening retained 110 new and potent DAHAs with predicted inhibitory potencies pIC50pre up to 520-fold lower than .that of DAHA1 (IC50exp =260nM). Predicted pharmacokinetic profiles of the new analogs were compared to current per oral anti-cancer drugs. Conclusions: This computational approach, which combines molecular modeling, pharmacophore model, analysis of HDAC2-DAHAs interaction energies, in silico screening of VCL of DAHAs, and ADME properties resulted in a set of proposed new HDAC2 inhibitors.