Gel chromatography indicated that after 

 only 20-min hydrolysis at 34°C most of the sol- 

 uble peptides released from FPC had molecular 

 weights of less than 2,000. Average molecular 

 weights decreased further with longer hydrol- 

 ysis periods. The molecular weights of most 

 of the enzymes constituting Pronase were esti- 

 mated to be in the range of 20,000. These 

 results provided a basis for design of a con- 

 tinuous reactor-ultrafiltration membrane sys- 

 tem. 



Continuous reactor . — The feasibility of 

 using ultrafiltration membranes with a continu- 

 ous reactor in order to retain and reuse enzymes 

 was evaluated. Characterization of commercial- 

 ly available membranes was performed in terms 

 of their rejection efficiencies for the various en- 

 zyme preparations. A polyelectrolyte complex 

 membrane, PM-10 (Amicon Corp., Lexington, 

 Mass.), demonstrated 100% rejection efficiency 

 for Pronase. Three other membranes which 

 were also tested were only partially retentive 

 for Pronase and Monzyme. The PM-10 mem- 

 brane also showed a 23% rejection efficiency 

 for the soluble FPC products of hydrolysis. A 

 less retentive membrane or further hydrolysis 

 to smaller peptides would be required for ex- 

 tended continuous operation. 



Enzymatic solubilization of FPC was tested 

 experimentally in a small continuous reaction 

 system containing an ultrafiltration membrane. 

 A number of problems were encountered. Feed- 

 ing of a homogeneous slurry of FPC to the re- 

 actor at the required low flow rate was a major 

 problem, but this would be less difficult in a larg- 

 er scale system. A more serious problem lies in 

 the unstable nature of proteolytic enzymes. 

 Pronase was found to have an inherent low ther- 

 mal stability and/or self-digestion in the con- 

 tinuous reactor. 



Monzyme, a mixture of B. subtilis neutral 

 and alkaline proteases manufactured by Mon- 

 santo, is relatively stable over a wide range of 

 temperature and pH. Operation with Monzyme 

 at pH 8.8 greatly reduced or eliminated problems 

 of microbial contamination which had been en- 

 countered in the continuous reactor at lower pH. 

 Since it is a relatively inexpensive protease the 

 most economical system for FPC solubilization 

 may or may not include an ultrafiltration mem- 

 brane for enzyme recovery. 



The kinetics of FPC solubilization with 

 Monzyme and with pancreatin have been studied 

 in detail and a continuous reactor system which 

 will provide adequate product for nutritional and 

 other evaluations has been assembled. The 

 batch studies on enzymatic solubilization have 

 been described in a publication by Cheftel et al. 

 (1971), and the continuous system studied will 

 be the subject of a later paper. 



Summary of Experimental Results 



Summary of screening tests. — The results 

 of the enzyme screening tests are listed in Tables 

 1 and 2. Initial tests were at 40°C and pH 7 for 

 a 1-hr period. Slurries containing 5 % by weight 

 of the extracted fish muscle substrate were hy- 

 drolysed with several concentrations of each 

 enzyme and both the remaining insoluble solids 

 (washed and dried) and the dissolved solids in 

 the filtrates were measured. A logarithmic plot 

 of a digestion ratio, DR — ratio of solubilized 

 solids to insoluble solids remaining, versus con- 

 centration of enzyme gave a straight line over 

 the range of digestion studied. For the 1-hr 

 tests the proteolytic enzymes were compared at 

 DR = 0.5 corresponding to solubilization of one- 

 third of the initially insoluble solids. The rel- 

 ative activities of enzymes tested by the 1-hr 

 hydrolysis are listed in Table 1. Ficin has high- 

 est relative activity under these digestion con- 

 ditions. 



The results of 24-hr hydrolysis tests are listed 

 in Table 2. These tests were carried out in 

 0.2m phosphate buffer at pH 6, 7, or 7.5 and a 

 temperature of 40°, 50°, or 65°C, depending on 

 the optimum activity ranges for the individual 

 enzymes. Sodium benzoate, 0.7%, was included 

 as a preservative. The enzymes are compared 

 at DR = 1.5, corresponding to a solubilization 

 of 60% of the initially insoluble solids. 



The more active enzymes as determined by 

 the 24-hr hydrolysis are listed in Table 3 with 

 approximate costs in terms of dollars per pound 

 of bulk enzyme and cents per unit of relative ac- 

 tivity as measured. Pancreatin, pepsin, and 

 papain are relatively economical. In later stud- 

 ies with raw whole hake, autolysis played a 

 major role and differences between added en- 

 zymes were less pronounced. The effectiveness 

 of pepsin relative to other enzjones is lower with 

 whole hake because autolytic activity is reduced 



12 



