Quantification of Soy-Derived Ingredients in Model Bread and Frankfurter Matrices with an Optimized Liquid...

The detection and quantification of soy protein is important for food allergen management and identifying the presence of undeclared soy proteins. Heat processing and matrix interactions can affect the accuracy of allergen detection methods. The sensitivity of enzyme-linked immunosorbent assay methods can be compromised if protein epitopes are modified during processing. Therefore, a mass spectrometry (MS)-based method was evaluated for the recovery of total soy protein in incurred matrices. MS-based quantification of total soy protein was assessed by using a combination of external and internal standards. The reproducibility of the standard curves was investigated by comparing within-day and among-day variation. Incurred samples were prepared using bread and frankfurters as model food matrices. Several soy-derived ingredients were used to prepare the matrices with varying levels of soy protein (1, 10, 50, or 100 ppm of total soy protein). A pooled standard curve was used to estimate the total soy protein concentration of the incurred food matrices and the percent total protein recovery. The variation of replicate standard curves between days and among all days was not significant. The differences in slopes obtained from replicate standards run on different days were minimal. The most influential factor on the quantitative protein recovery in incurred samples was the effect of the physical matrix structure on protein extraction. The lowest percent protein recoveries, less than 50%, were calculated for uncooked matrices. The cooked matrices had percentage recoveries between 50 and 150% for all total soy protein levels. Other factors, such as type of ingredient, were determined to be not as impactful on recovery. The MS method described in this study was able to provide sensitive detection and accurate quantification of total soy protein from various soy-derived ingredients present in processed food matrices.