Detection of soybean Gly m TI in complex foods: Impact of antibody batch variability on indirect ELISA and immunosensor performance

This work aimed to develop a new indirect ELISA for detecting the soybean allergen Gly m TI in processed and complex food matrices and to investigate the influence of antibody batch variability on immunoassay performance. During method development, inconsistencies in Gly m TI detection prompted a comparative evaluation of several batches of anti-Gly m TI IgG using both the newly developed ELISA and an electrochemical immunosensor. The ELISA was optimised in terms of protein extraction, antibody concentration, and linear dynamic range, and successfully applied to model foods simulating biscuits, sausages, and cooked hams, achieving a limit of detection (LOD) of 1,000 mg/kg irrespective of matrix or processing. Given this limited sensitivity, antibody batch variability was further assessed across both ELISA and immunosensor. Antibodies obtained approximately at the same time showed comparable performance, whereas those acquired five years earlier differed substantially. In particular, antibodies of older batches (Ab1/Ab3) outperformed the most recent ones (Ab2/Ab4), exhibiting wider dynamic ranges, lower LOD (0.1 mg/kg), and requiring lower antibody concentrations. These results highlight the significant variability of commercial antibodies and its critical impact on immunoassay sensitivity and reproducibility in food allergen detection. • New ELISA for Gly m TI detection developed and applied to various food formulations • Parameters (protein extraction/antibody concentration/dynamic range) were optimised • ELISA applied to detect 1,000 mg/kg of soybean in biscuits, sausages, cooked-hams • High LOD justified by the different performance of anti-Gly m TI antibody batches • Older antibodies (2019) presented better performances than new antibodies (2024)