Bioavailability of lysine in heat-treated foods and feedstuffs

During the processing of foodstuffs, lysine can react with other compounds present to form nutritionally unavailable derivatives, the most common example of which are Maillard products. Maillard products can cause serious problems when determining the available lysine content of processed foods or feedstuffs as they can revert to lysine during amino acid analysis. Several methods have been developed to determine the dietary lysine available for the metabolic processes of animals including animal growth-based assays, reactive lysine chemical methods and digestibility assays. However, growth-based assays are laborious, highly variable and tend to determine utilization rather than availability. Chemically reactive lysine assays do accurately determine the unmodified lysine in a food or feedstuff, but do not determine available lysine as they incorrectly assume that reactive lysine digestion and absorption is 100%. Ileal digestibility assays measure digestible total lysine rather than digestible reactive lysine (available lysine) and so are inaccurate, especially when applied to processed protein sources. This thesis describes the development of a true ileal digestible reactive lysine assay for determining dietary (bio)available lysine. This assay couples the guanidination reaction, for determining reactive lysine, with a true ileal digestibility assay. The resulting apparent digestibility estimate is corrected to a true digestibility value by accounting for the endogenous ileal lysine flow. Selected reaction conditions for the guanidination of lysine in a heated lactose/casein mixture and digesta of rats fed unheated casein and heated lactose/casein was examined. Overall, suitable reaction conditions were 0.6 M O-methylisourea for 7 d in a shaking waterbath at 21 ± 2 °C with an O-methylisourea to lysine ratio of 1000 and a reaction mixture pH of 10.6 for casein and heated lactose/casein and 11.0 for digesta. The accuracy of the guanidination method for determining reactive lysine in a range of “ready-to-eat ” cereal-based breakfast foods and selected feedstuffs was tested by comparison with the reactive lysine content of the same protein sources when determined using the fluorodinitrobenzene method. Overall, there was excellent agreement between the two methods. The accuracy of the newly developed bioassay for determining digestible reactive (available) lysine for predicting lysine deposition was also tested using a heated skim milk powder. The true ileal total and reactive lysine digestibilities were determined for the heated skim milk powder which was then fed to pigs, along with two control diets which were formulated based on either total lysine digestibility or reactive lysine digestibility. All diets were limiting in lysine. The pigs fed the heated skim milk powder deposited the same (P > 0.05) amount of lysine (9.1 g d-1) as the pigs fed the control diet that was formulated based on reactive lysine digestibility (9.1 g d-1) but deposited significantly (P < 0.05) more than the pigs fed the control diet that was formulated based on total lysine digestibility (5.4 g d-1). Consequently for the heated skim milk powder at least, the true ileal digestible reactive lysine assay accurately determined the available lysine content. The new assay demonstrated that for a range of milk protein-based foods, there was little difference between digestible total lysine and digestible reactive lysine for most of the milk products tested. In contrast, for a range of “ready-to-eat” cereal-based breakfast foods, available lysine was 5 – 50% lower than that determined using the traditional assay, which is of concern given that breakfast cereals are perceived to be “healthy” foods. Similarly, the available lysine content of a range of moist and dry commercial cat foods was significantly (P < 0.05) lower (15-55% lower) than previously estimated using the traditional true ileal digestible total lysine assay. The assay was also used to examine the effect of storage for extended periods at elevated temperatures on a hydrolysed-lactose skim milk powder and overall, there was a significant decrease in the available lysine content over time, as much as 60% over 6 mth when the powder was stored at 40 °C. In addition, the decrease in available lysine content of the hydrolysed-lactose skim milk powder was 2 – 5.5 times greater than observed for a normal skim milk powder depending on the storage time and temperature. Overall, foods and feedstuffs that have undergone processing often contain lower amounts of available lysine than thought previously. This new assay not only highlights the inaccuracy of the traditional true ileal digestible total lysine assay as a method for determining available lysine in processed protein sources, but permits the accurate assessment of the available lysine content of processed foods and feedstuffs.