Extraction of plant-based proteins with ultrasound

Abstract

Background: A marked increase in people eating less meat has led to an increase in the demand for alternative sources such as plant-based proteins. Plant protein is extracted to produce plant-based food products like tofu, which have a high carbon footprint. Ultrasound-aided protein extraction has been explored to make this process more environmentally friendly. This project established a protein extraction protocol from chickpeas, ful medames and soya beans and determined the efficacy of ultrasound technology to increase the yield of protein extraction. Method: A 150 g portion of beans were ground individually into a paste, mixed with dH2O, homogenized, and divided into two portions of 350 mL. Each portion was treated with a bath or probe sonicator. Similarly, 50 g of beans were used to generate 350 mL of bean solutions, which were used as an untreated control. The bean solutions were then filtered and centrifuged to obtain supernatants, which were quantified using Bradford’s assay. The supernatants were freeze-dried and resuspended into 25 mL of dH2O, and then divided into five portions. One portion was used to perform gel electrophoresis to quantify total protein by Coomassie staining. Protein was precipitated out of the remaining supernatants with MgCl2 or CaCl2 or dH2O as a control to measure the loss of protein using Bradford’s assay. Results and discussion: The protein extraction protocol has a significantly higher extraction efficiency for soya bean compared to chickpeas and ful medames in the untreated control when measured as grams of protein per 100 g of bean or percentage yield of total protein reported on the packaging (p <0.001). Similarly, more protein was extracted from soya bean after sonication. No significant difference was found between types of sonication with each bean type. Salt precipitation did not result in a loss in protein after treatment with ultrasound. The total soluble protein increased after ultrasound treatment relative to untreated control when measured by gel electrophoresis and Coomassie staining. Conclusion: This study demonstrates that sonication aids protein extraction. However, the volume of dH2O added to the bean paste was a confounding factor as the solvent capacity of dH2O was reached. Further optimisation of the extraction protocol will aid data normalisation, contributing to this novel field in food science.