Table 1_High-throughput screening of monoclonal antibodies against carbapenemases using a multiplex protein microarray platform.xlsx

Introduction Carbapenemase-producing bacteria undermine the efficacy of carbapenems, a class of last-resort antibiotics used primarily to treat infections caused by multidrug-resistant Gram-negative pathogens. Carbapenemases are among the most alarming antimicrobial resistance mechanisms because they inactivate all β-lactam antibiotics leaving clinicians with few or no therapeutic options. The genes encoding these enzymes are typically located on mobile genetic elements (MGE), which facilitate rapid horizontal gene transfer among different bacterial species. These MGE’s often additionally carry toxin-antitoxin systems that promote long-term persistence in bacterial populations. Carbapenem-resistant Enterobacteriaceae (CRE) often colonize the gastrointestinal tract without symptoms, serving as silent reservoirs for further dissemination. Infections caused by CRE are associated with high morbidity and mortality and are frequently resistant to multiple drug classes. Given the urgent clinical need for rapid diagnostics, immunochromatographic assays represent a promising and urgently needed approach for economic and available point-of-care detection. However, the development of such assays is often hindered by the time-consuming process of identifying high-affinity antibody pairs. Methods To accelerate this process, we evaluated a protein microarray platform as a high-throughput screening tool to identify optimal monoclonal antibody (mAb) pairs targeting the most clinically relevant carbapenemases. Monoclonal antibodies derived from hybridoma libraries and commercial sources were spotted in triplicates and tested in a single experiment against lysates from reference strains expressing the carbapenemase enzymes KPC, NDM, IMP, VIM, OXA-23/48/58, and MCR-1, an enzyme conferring resistance to colistin. Signal intensities were quantified, and diagnostic performance was assessed across four thresholds. Results A cut-off > 0.2 yielded the best balance, with approximately 61% balanced accuracy and ≥99% specificity. Around ...