Development of sandwich and competitive ELISA formats to determine β-conglycinin: Evaluation of their performance to detect soy in processed food

Abstract Two ELISA formats (sandwich and indirect competitive) were developed to quantify β-conglycinin, a major soy allergen. Their performance was evaluated using three model foods incurred with soy proteins. The sandwich format detects the addition of 0.005% and 0.05% soy proteins in pasteurized sausages and baked bread. However, the competitive format detects only 0.1 and 0.5%, respectively. β-conglycinin was not detected in sterilized pâte with any format. An industrial prototype of the sandwich ELISA was in-house validated, showing acceptable results of repeatability, reproducibility and robustness. Model foods spiked with β-conglycinin after processing showed recoveries between 93.3 and 138.7%. However, in model foods incurred with soy proteins before processing the recovery decreased with the increase of the severity of heat treatment applied. The sandwich format could differentiate most of the retail foods with soy declared or not as ingredient. The ELISA format and processing conditions greatly influence the determination of β-conglycinin in food.