20—JOURNAL OF FOOD SCIENCE—Volume 62, No. 1, 1997 Transglutaminase Effects on Low Temperature Gelation of Fish Protein Sols H.G. LEE, T.C. LANIER, D.D. HAMANN, and J.A. KNOPP ABSTRACT Myosin polymerization and formation of ε-(-glutamyl)lysine linkages were quantified in Alaska pollock surimi gels which contained no ad- ditive (control), or a commercial microbial transglutaminase (MTGase). As preincubation (‘‘setting’’) time at 25°C was increased, the gel strength of control and 0.2% MTGase-added samples increased, with greater increases at higher MTGase levels. SDS-PAGE and HPLC anal- yses showed increasing nondisulfide polymerization and ε-(-gluta- myl)lysine dipeptide content, with increasing setting time and/or added MTGase. Content of ε-(-glutamyl)lysine dipeptide correlated with gel strength (shear stress) and shear modulus at failure (G f ) for these gels. Higher stresses were measured in samples containing 0.2% MTGase than in controls at corresponding levels of ε-(-glutamyl)lysine dipeptide, in- dicating that rate of myosin polymerization may affect ultimate gel strength. Key Words: transglutaminase, surimi, covalent, linkage, crosslink INTRODUCTION SURIMI (water-washed, cryoprotected mince) from some fish species has a unique ability to form translucent, elastic gels be- low 40°C after being comminuted with salt. This is termed ‘‘set- ting.’’ Upon heating at higher temperatures, a stronger, more elastic gel with higher water-holding capacity results, as com- pared to a gel cooked without preincubation (Okada, 1959; Niwa and Nakajima, 1975; Lanier, 1986; Kamath et al., 1992). Un- derstanding the mechanism for this unique setting could enable better control and possible enhancement of surimi gelling prop- erties. It has been reported that transglutaminase (TGase) (R-glutam- inyl-peptide: amine -glutamyltransferase; EC 2. 3. 2. 13) is largely responsible for the ‘‘setting’’ phenomenon. Several re- searchers (Seki et al., 1990; Kimura et al., 1991; Kamath et al., 1992; Kumazawa et al., 1993a) have documented increased pro- tein polymerization during setting and attributed this to a cal- cium-dependent TGase activity endogenous to the muscle protein. TGase catalyzes an acyl-transfer reaction in which the -carboxamide groups of peptide-bound glutaminyl residues are the acyl donors. A variety of primary amines and the lysyl res- idues of proteins could act as acyl acceptors, the latter gener- ating ε-(-glutamyl)lysine crosslinks (Folk, 1983). Ando et al. (1989) isolated microorganisms (Streptoverticil- lium) that produced a TGase which did not require calcium ions for activity. This microbial transglutaminase (MTGase) has been applied experimentally to polymerize  s1 -casein, soy proteins, conalbumin, rabbit myosin, carp myosin, beef myosin and actin, ovomucin and other food proteins (Nonaka et al., 1989, 1994; Muguruma et al., 1990; Tanaka et al., 1990; Kato et al., 1991). MTGase has been shown to be useful in strengthening surimi gels during the setting reaction as well (Seguro et al., 1994). This lower cost MTGase (compared to animal-derived sources) should soon be available commercially. The objective of our study was to investigate the effects of TGase on low temperature gelation of fish protein. Authors Lee, Lanier, and Hamann are with the Dept. of Food Sci- ence, North Carolina State Univ., Raleigh, NC 27695-7624, Author Knopp is with the Dept. of Biochemistry, North Carolina State Univ., Raleigh, NC 27695-7624. MATERIALS & METHODS Materials Frozen surimi prepared from Alaska pollock (Theragra chalco- gramma) was purchased from HFI (Raleigh, NC). Microbial transglu- taminase (MTGase) was supplied by Ajinomoto Co. (Japan). O-phthaladehyde (OPA), synthetic ε-(-glutamyl)lysine, pronase, car- boxypeptidase A, leucine aminopeptidase, and thymol were obtained from Sigma Chemical Co. (St. Louis, MO). All other chemical reagents were analytical grade. Gel preparation Frozen surimi was tempered at room temperature for 1 hr before cub- ing with a knife while frozen. The surimi was then chopped for 15 sec to reduce particle size followed by blending with 2% NaCl and chilled water in a vertical cutter/mixer (Stephan Machinery Corp., Columbus, OH). Water was added to adjust final moisture of all formulas to 78%. In appropriate batches, MTGase was added at this stage. The paste was then chopped at 1750 rpm under vacuum to a final temperature of 5°C and placed in plastic bags and vacuum-sealed to minimize air pockets. Stainless steel tubes (diam = 1.87 cm, length = 17.75 cm) sprayed with a nonstick agent (Pan-Out, Pegler-Sysco) were stuffed with pollock paste using a sausage stuffer (Vogt series 9, Germany). Tubes were capped and randomly assigned to one of two thermal treatments in water baths: 90°C for 20 min, or different setting times at 25°C (optimum setting temperature for Alaska pollock surimi; Kamath et al. 1992), followed by cooking at 90°C for 20 min. After cooking, tubes were immediately removed, placed in an ice-water bath and cooled for 10 min. All gels were removed from the tubes and stored overnight at 4°C in Whirl-Pak bags (Nasco, Fort Atkinson, WI) prior to gel testing. Measurement of microbial transglutaminase effect Different levels of MTGase [0, 0.1, 0.2, 0.3, and 0.4% (w/w)] were incorporated into surimi pastes. Pastes were preincubated in forming tubes for 0 or 1 hr at 25°C prior to cooking at 90°C for 20 min. This temperature was used irrespective of enzyme source (endogenous or mi- crobial). It has been shown that the optimum temperature of the cross- linking reaction was affected more by the substrate stability (i.e., fish species) than the enzyme source (Joseph et al., 1994). Torsion test Torsional analysis was used to determine failure properties of gels. The cut samples (cylindrical specimens; L = 2.87 cm, D = 1.9 cm) were glued to notched styrene disposable disks and ground to dumbbell shapes with minimum diameter 1.0 cm using a specimen shaping ma- chine (Gel Consultants, Inc., Raleigh, NC). The samples were tested at room temperature (25°C) using a Torsion Gelometer (Gel Consultants Inc., Raleigh, NC). The true shear stress (strength) and true shear strain (deformability) at sample failure were calculated with the Gelometer software using appropriate equations for specimen geometry (Diehl et al., 1979). Shear modulus at failure (G f ), a measurement of gel rigidity, was calculated as true shear stress/true shear strain at sample failure. Dynamic small strain test Thermal dynamic testing was conducted using a concentric cylinder apparatus (C25) and a torque bar of 103.33 g-cm on the Bohlin VOR Rheometer (Bohlin Reologi, Edison, NJ). Maximum strain was 0.02 and frequency was 0.05 Hz. Sample was chopped with 2% NaCl and water, adjusted to 78% final moisture, in a Black & Decker Handy Chopper (cat. no. HC-20 Type 2., Black & Decker Inc., Shelton, CT) similar to