Single laboratory method validation for cyanide in beans with insufficient levels of β-glucosidase activity

Methods to determine cyanide in beans, such as the AOAC OMA 915.03 utilise endogenous enzyme linamarase (β-glucosidase) to liberate hydrogen cyanide (HCN). However, several beans are found to contain insufficient levels of β-glucosidase activity to evaluate their levels of cyanide. This implies some beans indicate false negative using official methods. Therefore, we developed a method to analyse cyanide in beans which lack linamarase activity, and a single laboratory validation was performed. This method used 3 U of linamarase and samples were maintained at 38 °C for 24 h to hydrolyse the cyanogenic glucoside in the beans. After hydrolysis, water and antifoam were added, then, this mixture was distilled to trap cyanide in an alkaline solution. The trapped cyanide was then quantified using a 4-pyridinecarboxylic acid-pyrazolone colouring method. To validate this method, non-cyanide containing blank bean (cowpea) was spiked with potassium cyanide (KCN) from 1.0 to 500 mg/kg and linamarin from 5.0 to 25 mg/kg HCN equivalent and analysed. The recovery of cyanide from KCN ranged from 82 to 91%, with relative standard deviation (RSD) of 1.2 to 3.7%. The recovery of cyanide from linamarin ranged from 82 to 83%, with RSD of 6.2 to 8.0%. Lima bean, with naturally containing cyanogenic glucosides, was mixed with non-cyanide containing cowpea to cyanide levels of 5.0 to 40 mg/kg. Standard deviations of ‘within day’ and ‘between days' recoveries using these samples were 2.8 to 4.6% and 3.0 to 7.8% at 5.0 to 40 mg/kg HCN level, respectively. The release of HCN from six beans of Phaseolus lunatus was measured either with and without addition of enzyme. Five of the beans had significant differences of detected HCN with versus without enzyme addition.