350—JOURNAL OF FOOD SCIENCE—Volume 61, No. 2, 1996 Control of Endogenous Enzyme Activity in Fish Muscle by Inhibitors and Hydrostatic Pressure using RSM I.N.A. ASHIE, B.K. SIMPSON, and H.S. RAMASWAMY Table 1—Coded levels used in experimental design Process variable Levels -1 0 1 Hydrostatic pressure (atm) X 1 1000 2000 3000 Crude inhibitor extract (%) X 2 0.1 0.2 0.3 pH X 3 5.5 6.0 6.5 ABSTRACT Effects of combination treatments involving hydrostatic pressure, pH, and crude  2 -macroglobulin on muscle protease activity and viscosity were investigated using response surface methodology. Statistical anal- ysis showed the model was very significant, predicting responses with high accuracy. Proteolytic activities were reduced with increasing pres- sure and inhibitor concentrations while pH elicited a quadratic response. Reducing pH and increasing pressure also reduced viscosity of tissue homogenate, but inhibitor concentration did not affect (p  0.05) this response. The proteases did not show any tendency to regain activity during 3 weeks storage possibly due to a crude inhibitor in the homo- genate and did not seem to have any direct influence on viscosity as shown by a low correlation (r =-0.30). Viscosity progressively in- creased during this period. Key Words: hydrostatic pressure, fish proteases, viscosity,  2 -macro- globulin, response surface INTRODUCTION TEXTURE is the main determinant of quality of fresh and proc- essed seafood products like surimi. With the declining stocks of traditional fish species in the oceans, researchers have been de- veloping strategies to enhance the value of nontraditional fish species for such processed foods. They have also developed strategies to improve the shelf-life of fresh/processed fish. Such studies have shown that the rapid softening of fresh and proc- essed fish tissue may be directly linked to excessive activities of endogenous proteases. Some of the proteases implicated in postharvest fish texture deterioration include the cathepsins, cal- cium-dependent proteases, collagenases, alkaline proteases, di- gestive enzymes, and other fish muscle proteases (Haard et al., 1979; Makinodan et al., 1985; Nip et al., 1985; Sun-Pan et al., 1986; Folco et al., 1988; Lindner et al., 1988; Bremner, 1992; Guizani et al., 1992; Jiang et al., 1992; Yamashita and Kona- gaya, 1992). Effective utilization of fish and other seafood re- sources, therefore, depends on the capacity of any preservation or processing technique to control excessive postmortem activ- ities of such enzymes, while taking cognizance of Food Safety Regulations and consumer perception of the product. The efficacy of a number of naturally occurring proteinase inhibitors in controlling such undesirable enzyme activities has been investigated. Nagahisa et al. (1981) showed that protease activity in Pacific hake muscle tissue was inhibited by several compounds including dried egg white powder and potato ex- tracts. The strength of surimi gels from fish species like Pacific whiting, Atlantic croaker (Micropogon undulatus), and arrow- tooth flounder (Atheresthes stomias) was also increased by ad- dition of bovine serum albumin and egg white (Chang-Lee et al., 1990), beef plasma protein (Hamann et al., 1990; Morrissey et al., 1993), and  2 -macroglobulin (Lorier and Aitken, 1991; Wasson et al., 1992b). Even though some of the effects of hydrostatic pressure on foodstuffs had been known for about a century (Hite, 1899), its commercial exploitation was given further impetus by Hayashi’s The authors are affiliated with the Dept. of Food Science & Ag- ricultural Chemistry, McGill University (Macdonald Campus), 21,111 Lakeshore Rd., Ste Anne de Bellevue, PQ, Canada H9X 3V9. Address inquiries to Dr. B.K. Simpson. (1989) proposal for its possible application in this area. It has inhibited microbial growth (Shimada et al., 1991), caused pro- tein denaturation (Hayakawa et al., 1992), and inactivated var- ious enzymes (Ashie and Simpson, 1995; Ogawa et al., 1990; Ohmori et al., 1991). Such effects have led to its application to tenderize meat (Suzuki et al., 1990), produce fish gels of higher quality than heat-induced gels (Okamoto et al., 1990), process tomato juice (Poretta et al., 1995), and to control enzyme— related seafood texture deterioration (Ashie and Simpson, 1996a). The individual influence of both hydrostatic pressure and in- hibitors has been amply demonstrated. However, investigation of such effects one at a time would be laborious and time con- suming resulting in large quantities of data. Such data would be difficult to interpret and provide no evidence of possible inter- active effects. The mathematical modeling technique of Re- sponse Surface Methodology (RSM) can overcome such limitations. It has been applied to control growth and aflatoxin production by Aspergillus flavus (Ellis et al., 1993), optimize casein extrusion (van de Voort and Stanley, 1984), consistency of tomato juice (Porretta et al., 1995), and optimize surimi gel texture (Chen et al., 1993). Our objective was to determine by RSM, the combined effects and optimum conditions (pH, hy- drostatic pressure, and  2 -macroglobulin) for controlling unde- sirable endogenous enzymes implicated in seafood texture deterioration and to evaluate the effects of treatments on some physicochemical properties of fish tissue. MATERIALS & METHODS CRUDE  2 -MACROGLOBULIN was recovered from blood samples obtained from freshly slaughtered cattle at a local abattoir and fresh bluefish (Po- matomus saltatrix) were purchased from a local market. Pure  2 -mac- roglobulin was purchased from Boehringer Mannheim (Germany) and bovine serum albumin (BSA) was from Sigma Chemical Co. (St. Louis, MO). Experimental design Factors and their corresponding levels were: hydrostatic pressure (1,000–3,000 atm) each applied for 30 min; crude  2 -macroglobulin [0.1–0.3% of tissue homogenate, (w/w)]; and pH (5.5–6.5) (Table 1). The pressure and inhibitor values were based on previous work (Ashie and Simpson, 1995, 1996a) and the pH values were based on levels common in postmortem muscle tissues. To determine simultaneous ef- fects of these factors on proteolysis and viscosity, a 3  3 factorial model (Mullen and Ennis, 1979) was used: 2 2 y =+ X + X + X + X + X 0 1 1 2 2 3 3 11 1 22 2 (1) 2 + X + B XX + XX + XX 33 3 12 1 2 13 1 3 23 2 3 where X 1 ,X 2 , and X 3 correspond to variables, viz; hydrostatic pressure, inhibitor, and pH, respectively, and the  values represent corresponding