Experimental Study and Molecular Modeling of Antibody Interactions with Different Fluoroquinolones.

Antibodies against low-molecular-weight compounds exhibit cross-reactivities (CRs) with their structural analogs, varying by orders of magnitude for different substances. This variability limits the informativeness of antibody applications as analytical reagents and for other aims when samples contain several members of the same family, their derivatives, or partial degradation products. Therefore, there is a demand to find some criteria for understanding the relationships between the structural characteristics of antigens of a given chemical class and their immunochemical activity. This study presents an experimental and theoretical investigation of the properties of a monoclonal antibody (MAb) against the S-stereoisomer of gatifloxacin, a member of the widely used (fluoro)quinolone (FQ) family of antibiotics, characterized by high structural diversity. The aim was to determine FQs that form complexes with MAb and suggest a methodology to predict their CRs in silico. For this, the interaction of MAb with 26 FQs was studied using the enzyme-linked immunosorbent assay and presented as CR values to the target antigen. The most pronounced CRs were observed for lomefloxacin, sarafloxacin, and ciprofloxacin. Molecular dynamics (MD) simulations were performed to identify differences in analyte interactions at the MAb antigen-binding site, which determines binding affinity. It has been shown that molecular docking fails to discriminate cross-reactive from non-cross-reactive compounds because FQs have similar cores. Therefore, advanced analysis of MD trajectories was carried out. It allowed for clarification of the dynamic features of analyte-antibody interactions responsible for binding. It was shown by the dynamical network analysis that the sum of betweenness centrality between a node corresponding to the quinolone ring and nodes representing MAb amino acids is higher for cross-reactive haptens. The found regularities can be transferred to other analyte-antibody systems as a binary classifier that discriminates cross-reactive and non-cross-reactive compounds.