An insight into selective and potent inhibition of histone deacetylase 8 through induced-fit docking, pharmacophore modeling and QSAR studies

Abstract

Histone deacetylase 8 (HDAC8) has emerged as an important therapeutic target due to its involvement in various cancerous and neurodegenerative disease states. Since pan HDAC inhibition has been linked to various side effects, the need of the hour is to develop inhibitors truly selective for one isoform. This work attempts to explore the structural basis of selective HDAC8 inhibition by docking, pharmacophore and 3 D QSAR studies of 53 highly potent and highly selective triazol-based hydroxamic acid inhibitors. The binding modes of these novel inhibitors have been explored via Glide XP (Extra Precision) and induced-fit docking (IFD) strategies. The IFD poses of highly active and selective inhibitors showed conformational changes in active site residues like Trp141, Phe152 and Phe208, which were further verified by molecular dynamics simulations. A new CH–π interaction, which is atypical of HDAC inhibitors, was also observed in case of some highly selective inhibitors. Two pharmacophore models have been proposed; one highlights the structural basis of potency of these inhibitors and the other focuses on the selectivity. The corresponding QSAR models, obtained from alignment of the inhibitors as per the proposed pharmacophore models, are highly statistically significant. These models highlight the importance of size of the hydrophobic and aromatic groups present in the inhibitors and their contribution to activity of the inhibitors. The ADMET properties of the ligand library have also been analyzed and the predicted descriptors have been correlated with activity using principal components analysis to gain insight into the effect of pharmacokinetic properties on the activity.

Communicated by Ramaswamy H. Sarma

Keywords:

• Histone deacetylase 8
• selective inhibition
• induced-fit docking
• pharmacophore 3D QSAR
• molecular dynamics

Notes

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 QSite, Version 5.5, Schrödinger, LLC, NY2009.

2 LigPrep, Version 2.3, Schrödinger, LLC, NY2009.

3 MacroModel, Version 9.7, Schrödinger, LLC, NY2009.

4 PHASE, Version 3.0, Schrödinger, LLC, NY2009.

5 QikProp, Version 3.2, Schrödinger, LLC, NY2009.

Additional information

Funding

Kriti Kashyap is grateful to University Grants Commission (UGC) for senior research fellowship.