A series of canola fractions representing various stages in the protein isolation process were collected to assess the extent of glucosinolate recovery using two ultrafiltration systems. The total glucosinolate level of all fractions was estimated using a diabetic test kit (DTK), One Touch Blood Glucose Meter, to monitor glucose released on hydrolysis of glucosinolates by endogenous myrosinase and results compared to those obtained using gas chromatography (GC). Of the glucosinolates in the meal, approximately 1.9 g/100 g were recovered in the protein isolated from canola meal, regardless of the type of ultrafiltration system. A major reduction in glucosinolates was evident in the ultrafiltration step where, based on GC data, 63% and 39% of the glucosinolates were recovered in the discarded permeate using an Amicon™ stirred cell and Vivaflow™ 200 ultrafiltration, respectively. However, an overestimation of glucose was observed for the DTK results when looking at total glucosinolate recovery, possibly due to the presence of other free glucose in the canola fractions. These findings suggest that low glucosinolate protein isolates can be obtained, in large part due to glucosinolate removal during ultrafiltration of the extracted material. While the DTK method gave good estimates of the glucosinolates in the meal, it cannot be recommended for use with isolation products.
