Abstract
Human galectin-3 (hGal-3) is a target protein implicated in various diseases, including fibrosis, and several inhibitors have been identified. In this study, we investigated the quantitative structure-activity relationships (QSAR) that influence hGal-3 binding affinity of 21 inhibitors using four quantum chemical interaction energy terms: electrostatic (ES), charge transfer (CT+mix), exchange (EX), and dispersion interactions (DI). These terms were derived from high-precision fragment molecular orbital (FMO) calculations. Our analysis revealed that the binding affinity could be explained primarily by the CT+mix and DI terms. We further decomposed these terms to analyze the contributions of specific hGal-3 residues and evaluated the correlations between residue contributions and binding affinity. For CT+mix, key interactions with R144, N160, E184, and R186 were identified, while for DI, interactions with R144 and R186 were the most significant. In the case of CT+mix, we found that charge transfer from E184 to the inhibitor stabilized the CT+mix interaction energy, contributing to the binding affinity. DI analysis revealed that parallel stacking between the aromatic ring of the inhibitor and the guanidine moieties of R144 and R186, as well as CH-π interactions with R186, stabilized the DI-induced interaction energy. Based on these findings, we propose a design guideline for high affinity hGal-3 inhibitors: lowering the energy of the unoccupied orbitals in the inhibitor to facilitate charge transfer from E184 and introducing substituents near R144 and R184 to enhance shape complementarity and stabilize dispersion interactions. The FMO-based QSAR approach allows for investigation of the correlation between quantum chemical interactions and binding affinity, offering a new perspective for molecular design.