STABILITY OF FEED ENZYMES IN PHYSIOLOGICAL CONDITIONS ASSAYED BY IN-VITRO METHODS

A series of in vitro incubations were carried out to investigate the s tability of two enzyme preparations in conditions similar to those in the upper gastrointestinal tract of monogastric animals. The two enzym e products, one crude xylanase from Trichoderma longibrachiatum (Multi fekt K) and the other a specifically manufactured feed enzyme (Avizyme SX(R)), were subjected to incubations at low and neutral pH with and without proteolytic enzymes (pepsin and pancreatin). Wheat gluten was employed together with the crude xylanase to investigate its potential as a stabilising agent. Due to the buffering effect of Avizyme SX(R), incubations were carried out with (pH 2.5) and without (pH 3.2) addit ion of either citric or hydrochloric acid. Incubation of the crude xyl anase at low pH followed by incubation at neutral pH resulted in negli gible loss of xylanase activity whereas beta-xylosidase recovery fell to 57 per cent of the initial value (P<0.05). Addition of wheat gluten resulted in full recovery of beta-xylosidase. The recoveries of both beta-glucanase and xylanase were significantly (P<0.05) lower than the initial values after incubation of Avizyme SX(R) in pH 2.5. However, with no pH adjustment (pH 3.2) the recoveries were significantly highe r (P<0.05 for beta-glucanase and P<0.10 for xylanase). The results fro m the pepsin and pancreatin incubations showed similar trends as the o nes of the pH stability experiments. Consequently, gluten addition and no pH adjustment gave the highest enzyme activity recoveries. The res ults suggest that partial enzyme inactivation may occur due to the low pH and proteolytic activities and hence in the GI-tract of monogastri cs. Feeds and feedstuffs can due to their buffering capacity and possi bly by providing substrates for the enzymes markedly reduce the rate o f inactivation. Results from a number of pig and poultry experiments a ppear to support this assumption. In vivo recovery measurements using animal models are needed to substantiate this.