Predictive modelling of inactivation of bovine milk alpha-L-fucosidase in a high-temperature short-time pasteurizer

In response to recent concerns regarding the persistence of food-borne path ogens in cheese and the need to ensure that milk destined for cheesemaking has received the minimal heat-treatment (thermization; 63-65 degrees C for 16 s), the aims of the present study were to examine the bovine milk enzyme alpha-L-fucosidase (FC) as a possible indicator of thermization and to ass ess the use of a reflectance colorimetric procedure for the enzyme assay. S tandard curves in whole and skim milk were prepared to convert the b* colou r units used in reflectance colorimetry to enzyme activity units. The assay procedure was sensitive enough to detect 7.5% raw milk mixed with heated m ilk and showed that the levels of FC activity in whole and skim milk were n ot significantly different (P > 0.05). The enzyme had a broad pH optimum fr om 5.0 to 6.0, with a K-m of 0.6 mM, and a V-max of 61.6 nmol mL(-1) h(-1). Thermal inactivation of FC, determined using a plate heat exchanger, was l inear with temperature using holding tubes of nominal times of 3-60 s. The integrated lethal effect was determined in four independent pasteurizer tri als conducted at intervals of approximately 3 weeks and a linear model for FC was developed. This model was used with the inter-trial variability and @RISK(TM) risk analysis simulation software to demonstrate that a log,, red uction in FC activity of 1.2 can be achieved using conditions corresponding to thermization (65 degrees C for 16 s) with 95% confidence. These results suggest that FC can be used as an indicator of heat-treatment or thermizat ion of milk destined for cheesemaking. (C) 2000 Elsevier Science Ltd. All r ights reserved.