Multivariate statistical optimization of the ultrasonic-assisted extraction of diamine oxidase (DAO) from pea sprout

Abstract

Pea sprouts can provide a plant-based alternative to the commonly used pork kidney as a source of diamine oxidase (DAO), an enzyme used as a dietary supplement to prevent symptoms of histamine intolerance. This study assessed the influence of process variables on the ultrasonic-assisted extraction of DAO from pea sprouts. DAO activity (DAO_ac), protein content (PC), and their ratio (DAO_sac) were measured on extracts obtained at different times (10, 20, 30, 40 min), temperatures (10, 20, 37 °C), and acoustic pressures (0, 29, 75 kPa). Response surface methodology (RSM) and multivariate regression models were applied to identify the optimal conditions and predictive multivariate models were assessed to develop a process control tool. Ultrasound application significantly enhanced DAO extraction, increasing DAO_ac, PC, and DAO_sac by up to 34.4, 14.9, and 19.4 % compared to conventional methods. The application of a moderate acoustic pressure (29 kPa) maximized yields, while a higher one (75 kPa) reduced extraction time. The moderate optimal temperature identified, 23.0 ± 0.3 °C, represents a clear advantage for the process scaling-up. The Support Vector Machine (SVM) model showed the best predictive performance (adjusted coefficient of determination; R²_adj > 92 % and mean relative error; MRE < 4 %), accurately estimating dependent variables based on process conditions. Therefore, this model could support the implementation of a closed-loop control strategy, dynamically adjusting operational parameters to maintain the extraction process under optimal conditions.