Quantification of Shrimp in Shrimp-Surimi Mixtures using Urea Gel Isoelectric Focusing TUNG-SHI HUANG, JON S. CHEN, MARTY R. MARSHALL, and CHENG-I WEI ABSTRACT A method of quantitate the percent content by weight of shrimp in surimi-pink shrimp mixtures was developed by comparing peak areas of two Alaska pollock surimi-specific (p1 7.11 and 7.17) and two shrimp-specific (PI 5.46 and 5.52) protein bands on isoelectric focus- ing (IEF) gels. An equation describing the linear relationship between peak areas and protein from the standard samples of surimi, shrimp and surimi-pink shrimp mixtureswas also developed.Using this method, unknown mixtures containing 10.0, 12.4 and 5.7% pink shrimp, re- spectively, were determined in blind studies to contain 8.3, 12.0 and 4.4% shrimp and 85.7, 86.8, and 92.7% surimi, respectively. INTRODUCTION SURIMI is mechanically deboned fish flesh washed with water and mixed with cryoprotectants to prolong frozen shelflife (Lee, 1984, 1986). Due to its light color, bland odor and unique gelling properties, surimi is used as a functional protein ingre- dient in manufacture of a variety of fabricated seafood products such as crab legs, scallops, lobster and shrimp analogs (Lanier, 1986; Regenstein and Lanier, 1986). To formulate the fabri- cated seafood products, surimi is combined with natural shell- fish meat, flavoring agents, salt, water, starch and/or egg white to modify functionality and to enhance flavor and texture. To meet Food & Drug Administration (FDA) guidelines, the fin- ished products must be labeled properly to include information on fish species as the main ingredient, and any other species, such as snow crab meat used as an additional, ingredient. There are problems in enforcing accuratelabeling of specific seafood components in fabricated products, mainly due to lack of reliable analytical methods. Products with claims of 35% crabmeat are widely sold when the use of over 10% crabmeat is known to cause detrimental effects (Regenstein and Lanier, 1986). Thus the establishment of methods to correctly deter- mine content of specific seafood components in fabricated sur- imi products is needed for regulatory purposes and for consumer protection. Isoelectric focusing (IEF) has been extensively used for identification of seafod speciesbecause it provides reliable and reproducible protein patterns for differentiating closely related species (Lundstrom, 1980). Using thin layer polyacrylamide gel IEF, various fish species have been readily identified (Hamilton, 1982). Recently, Wei et al. (1989) modified the IEF urea gel system of An et al. (1989) and greatly improved resolution of protein banding pattern for identification of raw pink, white and rock shrimp species. In this modified system, urea at 9.2M and an ampholyte mixture, containing 80% pH 4-6.5 and 20% pH 3-10 ampholytes, at 6.2% total concentra- tion were added to the gel mixture. Since many shrimp-specific minor bands with pI 5.0-7.5 were well separatedusing their approach, those bands could be useful for species identifica- tion. Our study was carried out to determine whether, using this modified urea gel IEF system, a reliable method could be established to quantitate the weight of a minor component in a surimi mixture, using Alaska pollock surimi-pink shrimp as a model system. Since the protein content in the aqueous su- The authors are with the Food Science & Human Nutrition Dept., Univ. of Florida, Gainesville, FL 32611. Address inquiries to Dr. C-i. Wei. pernatant of the water homogenates of surimi samples was low, another objective was to determine whether use of ultrafiltra- tion to concentrateproteins would affect protein patterns. Also investigated was the effect of two different homogenization procedures on protein patterns. MATERIALS & METHODS Samples and protein extractant Headed pink shrimp (Panaeuscfuol-arum) from Key West, FL were harvcstcd and transportedwithin 48 hr to the Food Science Sr Human Nutrition Dept., Univ. of Florida, Gainesville. They wcrc stored at -33°C. Shrimp were thawed under tap water, pcclcd, and dcvcined. Alaska pollock (Therugm chulcog~~mmu)surimi was manufactured by Alaska Pacific Seafoods (Kodiak, AK) and obtained from the Alaska Fisheries Development Foundation (Anchorage, AK). Surimi was cut into small pieces of about 8Og, put in Whirlpak@bags, and stored at - 33°C until needed. Water containing 0.1 mM phenylmethylsulfonyl fluoride (PMSF), 10 mM EDTA and 0.01% (w/v) sodium azide (to inhibit protcascs and microbial growth) was used to extract proteins from raw samples of surimi, shrimp, and surimi-shrimp mixtures. Effect of different homogenization methods on sample protein patterns Two homogenization procedures were compared for effectiveness in extracting sample proteins for IEF. The first included USC of a food processor (Presto MinnieMax Compact Food Processor, Eau Claire, WI) and a Polytron (Brinkmann Instrument, Westbury, NY), while the second only used the Polytron. In the first study, samples (80g) of surimi only, and surimi mixed with 5 or 10% (w/w) raw pink shrimp at 80g were each mixed with 1.6g (2%, w/w) NaCl and blended at 6°C using the food processor, until a shiny pastewas formed. A paste sample (7g) was removed, mixed with 21 mL aqueous extractant, and subjected to homogenization at room temperature 1 min using the Polytron (setting 9.2). In the second the various surimi samples at 7g were each combined with 0.14g (2%, w/w) NaCl in separatebeakers and chopped into small pieces. They were then combined with aqueous solvent at a ratio of 1:3 (w/v) and homogenized at room tcmpcrature 1 min using the Polytron (setting 9.2). The homogenatesprepared by either method were centrifuged at 26,900 x g (Sorvall RC-SB, rotor head SS-34) for 20 min at 5°C. Supernatantswere collected and the protein concentration determined (Lowry et al., 1951). They were subjected to IEF runs at 50 pg total protein/well. Each cxpcrimcnt was repeated three times. Effect of concentration with an ultrafiltration membrane on protein retention and IEF patterns Samples of surimi, and surimi-pink shrimp mixtures after homog- enization were centrifuged at 26,900 x g for 20 min. Ten mL aliquots of the supernatant of each sample were conccntratcd at room tcmpcr- ature using a Filtron Omegacell equipped with a 10 K nominal mo- lecular weight limit filter membrane (Pharmacia QK Biotechnology Inc., Piscataway, NJ) nitrogen pressure 25 psi. After volumes and protein concentrations of the concentrate and filtrate of each sample were measured, percent loss of protein in the filtrates was determined. The IEF protein banding patterns of the concentrates (50 pg/well) were compared with those of non-conccntratcd supernatants (also at 50 @well). This experiment was repeated six times. Isoelectric focusing A gel mixture containing 4% (w/v) aclylamide, 2% (w/v) Triton X-100 and 9.2 M urea was mixed 5 min at 37°C. Following addition 1206~JOURNAL OF FOOD SCIENCE-Volume 55, No. 5, 1990