The drug-human serum albumin binding interaction was evaluated on a stationary phase immobilized with human serum albumin using a mixture of phosphate buffer (pH 7.0) and acetonitrile modifier as mobile phase. The 33 compounds that have a wide structural and therapeutic diversity were analyzed by performing a large number of experiments. The interaction mechanism was interpreted based on: i) retention characteristics of structurally related compounds, ii) retention modeling, iii) quantitative structure retention relationship (QSRR), and iv) molecular docking. Small structural differences of related compounds (e.g., reflected in different lipophilicity and polarity) have been found to affect their different binding to human serum albumin. It was found that drug retention in HSA column can be successfully described by using the quadratic function. The isocratic (logk(14%)) and extrapolated (b0(LSS)) retention factors showed the highest correlation (r > 0.76) with the constant that defines the binding affinity for human serum albumin (ACD/I-Lab). Therefore, selected chromatographic parameters can demonstrate reliable applicability for rapid screening of drug-plasma protein binding in drug discovery. In QSRR study, the resulting SVM/logk(14%) and MLR/b0(LSS) models display high internal and external predictive power. The constitutional properties (double bonds, aromatic rings, benzyl, allyl, -amino and -sulfur containing functional groups) supported by the charged parts of surface area had a significant impact on human serum albumin-binding affinity, which was also confirmed with molecular docking study. The high structural diversity of the data set provides wide applicability of tested chromatographic conditions and constructed models for defining the pharmacokinetic profile and possible structural modifications that can increase plasma protein binding of newly synthesized, pharmaceutically important compounds.
By Darija Obradović, Milica Radan, Teodora Đikić, Marija Popović Nikolić, Slavica Oljačić, Katarina Nikolić