Quantitative structure-activity relationship and quantitative structure-pharmacokinetics relationship of 1,4-dihydropyridines and pyridines as multidrug resistance modulators

• Published in: Pharmaceutical research 2005-12, Vol.22 (12), p.1989-1996
• Main Authors: Zhou, Xiao-Fei, Shao, Qingxiang, Coburn, Robert A, Morris, Marilyn E
• Format: Article
• Language: English
• Subjects: Algorithms; Antineoplastic Agents, Phytogenic - metabolism; ATP-Binding Cassette, Sub-Family B, Member 1 - metabolism; Calcium Channel Blockers - pharmacology; Calcium Channels - drug effects; Calcium Channels - metabolism; Chemical Phenomena; Chemistry, Physical; Cytotoxicity; Dihydropyridines - chemical synthesis; Dihydropyridines - pharmacokinetics; Dihydropyridines - pharmacology; Drug Resistance, Multiple - physiology; Glycoproteins; Models, Molecular; Pharmaceuticals; Pharmacokinetics; Pyridines - chemical synthesis; Pyridines - pharmacokinetics; Pyridines - pharmacology; Quantitative Structure-Activity Relationship; Regression analysis; Toxicity; Vinblastine - metabolism
• ISSN: 0724-8741; 1573-904X
• DOI: 10.1007/s11095-005-8112-0

The aim of this study was to develop quantitative structure-activity/pharmacokinetic relationships (QSAR/QSPKR) for a series of synthesized 1,4-dihydropyridines (DHPs) and pyridines as P-glycoprotein (P-gp) inhibitors. Molecular descriptors of test compounds were generated by 3D molecular modeling using SYBYL and KowWin programs. Forward inclusion coupled with multiple linear regression (MLR) was used to derive a QSAR equation for Ca2+ channel binding. A multivariate statistical technique, partial least square (PLS) regression, was applied to derive a QSAR model for P-gp inhibition and QSPKR models. Cross-validation using the "leave-one-out" method was performed to evaluate the predictive performance of models. For Ca2+ channel binding, the MLR equation indicated a good correlation between observed and predicted values (R2 = 0.90), and cross-validation confirmed the predictive ability of the model (Q2 = 0.67). For P-gp reversal, the model obtained by PLS could account for most of the variation in P-gp inhibition (R2 = 0.76) with fair predictive performance (Q2 = 0.62). Nine structurally related 1,4-DHP drugs were used for QSPKR analysis. The models could explain the majority of the variation in clearance (R2 = 0.90), and cross-validation confirmed the prediction ability (Q2 = 0.69). QSAR/QSPKR models were developed, and the QSAR models were capable of identifying synthesized 1,4-DHPs and pyridines with potent P-gp inhibition and reduced Ca2+ channel binding. The QSPKR models provide insight into the contribution of electronic, steric, and lipophilic factors to the clearance of DHPs.