Designing potent inhibitors against the multidrug resistance P-glycoprotein. Issue 19 (15th November 2022)

• Record Type: Journal Article
• Authors: Mahmud, Shafi; Islam, Md. Jahirul; Parves, Md. Rimon; Khan, Md. Arif; Tabussum, Lamiya; Ahmed, Sinthyia; Ali, Md. Ackas; Fakayode, Sayo O.; Halim, Mohammad A.
• Is Part Of: Journal of biomolecular structure & dynamics. Volume 40:Issue 19(2022)
• Page Start: 9403
• Page End: 9415
• Publication Date: 2022-11-15
• Subjects: Multidrug resistance -- P-glycoprotein -- virtual screening -- molecular dynamics -- QSAR -- PCA; Biomolecules -- Periodicals; Molecular structure -- Periodicals; Molecular Biology -- Periodicals; Biomechanics -- Periodicals; 572
• DOI: 10.1080/07391102.2021.1930159
• Languages: English
• ISSNs: 0739-1102

Abstract

Abstract: The multidrug transporter P-glycoprotein is an ATP binding cassette (ABC) exporter responsible for resistance to tumor cells during chemotherapy. This study was designed with computational approaches aimed at identifying the best potent inhibitors of P-glycoprotein. Although many compounds have been suggested to inhibit P-glycoprotein, however, their information on bioavailability, selectivity, ADMET properties, and molecular interactions has not been revealed. Molecular docking, ADMET analysis, molecular dynamics, Principal component analysis (PCA), and binding free energy calculations were performed. Two compounds D1 and D2 showed the best docking score against P-glycoprotein and both compounds have 4-thiazolidinone derivatives containing indolin-3 one moiety are novel anti-tumor compounds. ADMET calculation analysis predicted D1 and D2 to have acceptable pharmacokinetic properties. The MD simulation discloses that D1-P-glycoprotein and D2-P-glycoprotein complexes are in stable conformation as apo-form. Hydrophobic amino acid such as phenylalanine plays significant on the interactions of inhibitors. Principal component analysis shows that both complexes are relatively similar variables as apo-form except planarity and Columbo energy profile. In addition, Quantitative Structural Activity Relationship (QSAR) of the ligand candidates were subjected to the principal component analysis (PCA) for pattern recognition. Partial-least-square (PLS) regression analysis was further utilized to model drug candidates' QSAR for subsequent prediction of the binding energy of validated drug candidates. PCA revealed groupings of the drug candidates based on the similarity or differences in drug candidates QSAR. Moreover, the developed PLS regression accurately predicted the values of the binding energy of drug candidates, with low residual error of prediction. Communicated by Ramaswamy H. Sarma