Abstract: With the aim of developing a method to prepare bovine trypsin by dual-factor activation and subsequent chitosan flocculation, Plackett-Burman design was used to evaluate the effects of 13 process conditions on the activity and extraction rate of trypsin to screen out 4 most significant ones (P＜0.05) including chtosan amount, ratio of bovine pancreas to duodenum and activation pH. Their optimal levels were optimized by the combined use of uniform design and partial least squares regression as follows: chtosan amount of 0.17%, CaCl2 amount of 0.06%, pancreas-duodenum ratio of 8.26:1, and activation pH 5.37, respectively. Under the optimized conditions, the activities of trypsin, lipase and amylopsin and the extraction rate of trypsin were observed to be 3408.71 U/g, 33.11 kU/g, 23.93 kU/g and 13.22%, respectively, which exhibited a relative error of 7.15%, 8.56%, 10.63% and 1.93% when compared with their predicted counterparts. Therefore, the combined use of Plackett-Burman design and uniform design is feasible to optimize the preparation of trypsin by dual-factor activation and subsequent chitosan flocculation.

Key words: Plackett-Burman design, uniform design, partial least squares regression, bovine pancreatin, chitosan, optimization