新規白血病標的メニン-MLL相互作用阻害剤のインシリコ仮想スクリーニングおよび構造ベース創薬アプローチによる同定

Leukemogenesis fundamentally depends on the interaction between Menin and MLL1 fusion proteins, especially in the context of aggressive mixed-lineage leukemia (MLL)-rearranged subtypes. This interaction can be disrupted to facilitate a focused therapeutic strategy. This study aimed to employ a comprehensive in silico drug repurposing strategy to identify new inhibitors of the Menin-MLL1 interaction. A comprehensive library of FDA-approved pharmaceuticals was subjected to a rigorous screening process utilizing virtual methodologies, clustering techniques, and a machine learning-based model (PSICHIC) to forecast binding affinity, subsequently complemented by molecular docking and an assessment of drug-likeness. Compounds that surpassed the antagonist value of the control (0.97076 kcal/mol) were selected for subsequent molecular docking investigations. The leading candidate, compound 52920501, exhibited a remarkable binding energy of - 8.89 kcal/mol in contrast to the control (DS-1594b), alongside advantageous pharmacokinetic characteristics and electronic stability, as validated by Density Functional Theory (DFT). Consistent docking interactions with critical residues (Asn282, Asp285, Asn244) facilitated by 52920501. Molecular dynamics simulations (300 ns) demonstrated that 52,920,501 sustained a stable interaction with Menin, promoted protein compaction, and reduced structural fluctuations. The MM/GBSA analysis conducted over the final 50 ns of the MD simulation revealed that compound 52,920,501 established a stable protein-ligand complex, exhibiting a final ΔTOTAL energy of - 12.83 kcal/mol. This value is comparable to the control, which recorded - 13.41 kcal/mol, yet demonstrates stronger individual interaction components. The inclusion of entropic contributions indicated a binding free energy (ΔG Binding) of - 4.75 kcal/mol for 52920501, in contrast to + 7.05 kcal/mol for the control, highlighting its enhanced binding thermodynamics. The steered molecular dynamics (SMD) simulations provided additional validation of 52920501's robust binding affinity, exhibiting a greater resistance (600 kJ/mol/nm) to external pulling forces in comparison to the control. In summary, compound 52,920,501 presents itself as a promising therapeutic candidate, bolstered by persuasive computational data that underscores its potential as a Menin-MLL1 inhibitor. It warrants further exploration through subsequent in vitro and in vivo studies.