Evolutionary Affinity and Selectivity Optimization of a Pesticide-Selective Antibody Utilizing a Hapten-Selective Immunoglobulin Repertoire

Abstract

Selectivity and sensitivity are considered as pivotal criteria for the quality of immunochemical assay designs in environmental analysis. They are essentially determined by the variable domains of the implemented antibody. The variable domains of a triazine-selective single-chain Fv (scFv) were genetically engineered by stringent molecular evolution in order to optimize analytical characteristics of the corresponding atrazine immunoassay. Gene variation of the template antibody by sequential shuffling against the variable heavy and light chain repertoire of a triazine-selective immunoglobulin library was enhanced by introducing additional point mutations. Improved scFv variants were selected by phage display employing an atrazine derivative. By this means the paramounting affinity of the initial scFv to sebuthylazine was shifted for the mutant antibodies toward a preferential recognition of the envisaged target analyte atrazine. In addition, the detection limit of the atrazine assay was significantly improved by factor 25 from 5.1 μg/L for the initial template antibody to 0.2 μg/L for the mutant antibodies. The contribution of the engineered antibody variants to the assay improvement is also reflected by a shift of the equilibrium dissociation constant K_D from 1.27 × 10^-8 M of the template antibody to 7.46 × 10^-10 M of the optimized variant. Sequence analysis revealed a bias of amino acid substitutions in the first two complementarity-determining regions (CDR) and the flanking framework regions of both variable chains for the shuffled clones as well as a deletion in the CDR3 of the light chain. Particularly the mutations of the V_L domain turned out to have a decisive impact on the alterations in the analytical performance of the engineered scFv mutants. The application of the mutant antibodies for the atrazine determination of soil samples revealed consistency with HPLC data within the experimental error.