Uncovering New Hits against Lysine Demethylase (LSD1) using Dual 3D Pharmacophores and Multi-Tier Rational Screening

Herein, we target LSD1 (KDM1A), a Lysine-specific demethylase enzyme responsible for demethylating the nucleosome’s mono- and di-methylated Histone proteins (H3K4/9), leading to epigenetic alterations in gene transcription. Growing evidence supports the involvement of LSD1 in Cancer and neurodegenerative diseases. Owing to the scarcity of clinically approved LSD1 inhibitors, we propose an alternate method to enhance LSD1 inhibitor repertoire by utilizing advanced computational techniques and logical multi-tier screening. In this work, we use a unique strategy via blending Structure-based pharmacophore (pharmacophore-S) and Ligand-based pharmacophore (pharmacophore-L) by employing common feature and Interaction-based pharmacophore mapping techniques to find novel hits against LSD1. The programs embedded with commercially purchased BIOVIA Discovery Studio Academic Research Suite were used for this purpose. The validated pharmacophores (pharmacophore-S and pharmacophore-L) were employed to screen InterBioscreen, COCONUT, and NCI natural compound databases based on fitting values of the candidates. The selected compounds were further screened using a multi-tier drug-likeness filtration based on Lipinski’s rule, Veber’s rules, SMARTS patterns, ADMET properties, and TOPKAT value. The funneled compounds were subjected to the molecular docking approach in the next round, and six compounds were chosen depending on the docking score and binding mode comparison with the reference. Then, MD simulation combined with MM-GBSA free energy binding calculations and Principal component analysis (PCA) were utilized to verify the stability of six compounds bound to LSD1. Overall, this study reveals the utilities of modern-day computational chemistry tools and uncovers four novel compounds as potential inhibitors of LSD1 for further studies.