B-036 Systematic Evaluation of Antibody-Antigen Interactions for Suitability in Immunoassays Using Parallel Microarray ELISA

Abstract

Background A critical step in immunoassay development is the selection of the appropriate capture and detection antibodies. The use of suboptimal antibodies can drastically affect the quality of the final assay. Traditional selection methods involve a direct testing of potential antibodies in the final assay setting; a time-consuming and material-intensive exercise that is often too complex to test all possibilities. Alternatives, such as abstract K-on/K-off determinations via SPR or BLI can be costly and often show limited translatability to the ultimate assay format. Here, we describe a novel systematic approach to rapidly screen antibody-antigen interactions for key assay specifications and test for translatability of the findings. Methods Nanoliter volumes of potential capture antibodies were spotted onto the bottom of ELISA wells, generating a microarray of antibodies in each well (>11,000 spots per plate). For the analysis of capture antibody-antigen interactions, a titration series of biotinylated analyte in negative sample matrix was applied. Binding was visualized using strep-polyHRP, precipitating TMB and high resolution scanning of each ELISA well. Antibody pair performance was investigated using unlabeled analyte and biotinylated detection antibody options. Sensitivity, specificity, dynamic range, binding speed, binding strength and matrix compatibility were investigated using time courses, concentration ranges, dilutions, interferent spikes and pH/salt/detergent challenges. Signals were quantified to calculate and statistically compare each possible interaction. Results Antibody-antigen interactions were investigated via parallel microarray ELISA for two example immunoassays: the HBsAg and cTnI test. For both analytes, a panel of antibodies was screened revealing antibodies with diverse characteristics. Striking differences were observed depending on the sample matrix composition. Without optimization, antibodies could be identified showing dynamic assay ranges of 3-4 log orders. Time course experiments revealed high-affinity antibodies suitable for rapid diagnostic tests. For HBsAg, levels down to 0.01 IU/mL (∼40 pg/mL) could be reproducibly detected, which is well below the minimal requirement of 0.13 IU/mL. Reactivity against multiple virus serotypes could be confirmed. Interestingly, for the cTnI test, a strong synergistic effect was observed when specific monoclonal antibodies were combined in the capture mix, resulting in a 6-fold sensitivity enhancement compared to their separate performances. Finally, we show that the results obtained with the parallel microarray ELISA approach are highly translatable to the final assay format, ranging from bead-based dry-chemistry assays to chemiluminescent immunoassays. Conclusions The parallel microarray ELISA evaluation presents a robust and reliable method for identifying the best antibodies for an immunoassay. It allows for a systematic and quantitative screen of numerous antibody options within 1-2 weeks under real immunoassay conditions with minimal antibody amounts. This method has shown to be highly translatable and is customizable to accommodate required assay specifications.