Immunoassay for Phenylurea Herbicides: Application of Molecular Modeling and Quantitative Structure–Activity Relationship Analysis on an Antigen–Antibody Interaction Study

Abstract

An indirect competitive enzyme-linked immunosorbent assay (icELISA) for 12 phenylurea herbicides (PUHs) was established with the half-maximum inhibition concentration (IC(50)) of 1.7-920.7 μg L(-1). A method of computer-aided molecular modeling was established in quantitative structure-activity relationship (QSAR) studies to obtain a deeper insight into the PUHs' antibody interactions on how and which molecular properties of the analytes quantitatively affect the antibody recognition. A two-dimensional (2D)-QSAR model based on the Hansch equation and a hologram QSAR (HQSAR) model were constructed, and both showed highly predictive abilities with cross-validation q(2) values of 0.820 and 0.752, respectively. It was revealed that the most important impact factor of the antibody recognition was the PUHs' hydrophobicity (log P), which provided a quadratic correlation to the antibody recognition. Hapten-carrier linking groups were less exposed to antibodies during immunization; thus, groups of the analytes in the same position were generally considered to be less contributive to antibody recognition during immunoassay. But the results of substructure-level analysis showed that these groups played an important role in the antigen-antibody interaction. In addition, the frontier-orbital energy parameter E(LUMO) was also demonstrated as a related determinant for this reaction. In short, the result demonstrated that the hydrophobicity and the lowest unoccupied molecular orbital energy (E(LUMO)) of PUH molecules were mainly responsible for antibody recognition.