The food matrices in the category of dairy, nuts, vegetables, legumes, cereal, cereal products, fruits and confectionery, salt, seeds, tea and color were analyzed for the matrix effect. Most of the food matrices under the category of dairy nuts, vegetables gave over estimation of amandin. Some food matrices in the category of legumes and cereal gave over estimation while other gave underestimations of amandin. The range of amandin detected in dairy and tree nuts varied from 116-198% and 82-300% respectively. Most of the legumes gave overestimation with detection ranging from 120-300% while some gave underestimation 43-90%. All cereal and cereal products gave overestimation 106-183% with the exception of barley, whole-wheat flour and rice wild and raisin bran whole mix. Spices gave lower recoveries (2-80%) with a few exceptions where higher recoveries were observed (range 121-334%). Most vegetables increased the amandin detection ranging 120-400%. Salt (black and white), tea, sugar, cocoa, jaggery, dark chocolate resulted in lower recoveries. All fruit matrices also gave lower recoveries (1.0-50%) with the exception of tropical mix fruit, figs, pineapple, papaya and apricot where the recoveries varied (60-80%). The pattern of inference (over or under estimation) from the food matrices remained similar at all the spike level of amandin (100, 10 and 1mg) tested, but the absolute detection value differed across the spike levels in most of the food matrices. These variances in the detection of amandin from the expected values reflect low levels of cross-reactivity or non-immunological inference in the assay by components of the food. High cross-reactivity was observed in dot blots with polyclonal antibody than with monoclonal for most of the food matrices. Western blots also showed high cross-reactivity with polyclonal antibody that with monoclonal antibody. Amandin could be qualitatively detected in spiked food matrices using western blot, food matrices tested had no effects on immnorecoginition of amandin. The pH of food matrix and type of buffer did affect the amandin recovery, however the extraction ratio (of FM: Buffer) did not affect the detection of amandin. The results suggest that food matrix effects need to be carefully evaluated in allergen detection development assays.
