Study of Class I HDAC-1, -2, and -3 Inhibitors Designed by Bioisosteric Replacement of Zinc Binding Groups and Caps of Traditional Pan Inhibitors: An In Silico Approach

Introduction: Despite the significant contribution of Antiretroviral Therapy (ART) in the management of viral replication and infection, HIV latency still presents a major barrier to complete eradication. Histone Deacetylases (HDACs), particularly Class I HDACs (the isoforms 1, 2 and 3) have been reported to play a pivotal role in maintaining this latency by contributing to transcriptional silencing. Selective HDAC inhibitors (HDACis) that target these isoforms can reactivate HIV reservoirs, encouraging viral growth and its subsequent recognition by the immune system, thus its clearance from the body. This is known as the “shock and kill” hypothesis. Methods: We employed computational methods to design novel HDACis by replacing the Zinc Binding Groups (ZBGs) and caps of known pan-HDAC inhibitors with respective bioisosteric fragments. Ligand design was based on modifying the structures of Vorinostat, Belinostat, Etino-stat, and Givinostat by retaining their linkers while substituting caps and ZBGs. Molecular dock-ing with Biovia (Discovery Studio) was performed to evaluate the binding affinity against the three target proteins (HDAC1, HDAC2, and HDAC3). The top-performing ligands underwent Molecular Dynamics (MD) simulations using GROMACS and binding energy calculations by the MMPBSA method to assess the stability of the complexes. Furthermore, ADMET screening was performed for drug-likeness evaluation and toxicity predictions. Results: Among the sixteen designed ligands, Hdi2 and Hdi10 emerged as the top performers, showing the highest binding affinities. Hdi2 demonstrated exceptionally high scores (-92.20, -80.00, and -74.31 Kcal/mol with HDAC1, HDAC2, and HDAC3, respectively). Hdi10, on the other hand, demonstrated consistent stability across all isoforms. MD simulations revealed high stability for Hdi10 with HDAC2 and HDAC3 and for Hdi2 with HDAC1, as suggested by low values of RMSD, RMSF, and robust hydrogen bonding. MMPBSA analysis revealed strong com-plex stability with up to -51.7 kJ/mol predicted binding energies. ADMET prediction showed negligible toxicity (LD50: 1600mg/kg and 6000mg/kg for Hdi2 and Hdi10, respectively) and zero Lipinski violations. Discussion: These findings indicate the potential of Hdi2 and Hdi10 as selective and stable bind-ers of Class I HDAC isoforms, addressing the limitations of existing pan-HDAC inhibitors ex-plored in HIV latency reversal studies. The observed favorable docking, stability, and safety pro-files highlight the prospects for further exploration of these compounds as more effective and less toxic LRAs. Nevertheless, the results of this study were solely computational predictions and therefore require experimental validation to confirm the biological efficacy and isoform selectiv-ity of the studied ligands. Conclusion: This study identified two promising novel HDAC inhibitors, Hdi2 and Hdi10, for further experimental investigation and optimization as potential LRAs for HIV latency reversal. These findings support the rational design of selective HDACis using computational approaches as efficient and cost-effective methods for the identification of future LRAs.