Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Investigation of antifreeze proteins as cryoprotectants for ling cod (Ophiodon elongatus) mince and natural… Liceaga, Andrea 2006

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2006-200162.pdf [ 12MB ]
Metadata
JSON: 831-1.0092920.json
JSON-LD: 831-1.0092920-ld.json
RDF/XML (Pretty): 831-1.0092920-rdf.xml
RDF/JSON: 831-1.0092920-rdf.json
Turtle: 831-1.0092920-turtle.txt
N-Triples: 831-1.0092920-rdf-ntriples.txt
Original Record: 831-1.0092920-source.json
Full Text
831-1.0092920-fulltext.txt
Citation
831-1.0092920.ris

Full Text

INVESTIGATION OF ANTIFREEZE PROTEINS AS CRYOPROTECTANTS FOR LING COD (Ophiodon elongatus) MINCE AND NATURAL ACTOMYOSIN By Andrea Liceaga B.Sc., Institute Tecnologico de Monterrey (ITESM), Mexico, 1995 M . S c , The University of British Columbia, Canada 1999 A THESIS SUBMITTED IN PARTIAL F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Food Science) THE UNIVERSITY OF BRITISH C O L U M B I A October 2006 © Andrea Liceaga, 2006 ABSTRACT Cryoprotectants used commercially to minimize changes in texture and protein properties during freezing and frozen storage of fish typically impart an undesirable sweet taste. Since fish antifreeze proteins (AFP) are known to modify and suppress ice crystal growth, the objective of this study was to evaluate AFP as alternative cryoprotectants for frozen ling cod mince and natural actomyosin (NAM). Mince from ling cod was subjected to freeze-thaw abuse in the absence (control) or presence of AFP (5, 10, 50 or 500 ppm), AFP (50 ppm) with 0.3% phosphates, polyols (4% sucrose + 4% sorbitol or 8% trehalose), or AFP (50 ppm) with polyols (2% sucrose + 2% sorbitol). Freeze-thawed mince with AFP showed higher textural hardness, less salt extractable protein, and higher expressible moisture. These samples also formed a layer of ice crystals, which was not observed in polyol blends or control. Differential scanning calorimetry showed more unbound water in mince with AFP, while Raman spectroscopy indicated increased prevalence of (3-sheet and random coil structures at the expense of a-helix. AFP failed to prevent loss of Ca-ATPase activity in N A M following freeze-thaw abuse. . AFP solutions were evaluated at ambient and subzero (-0.5, -1.8, -4.0°C) temperatures using Raman spectroscopy. AFGP had small peaks near 1620 and 1674 cm"1 attributed to polyproline type-II helix and extended/unordered (^-structures, respectively, and a strong band at 1070 cm"1 assigned to backbone C-C, C-N stretching and carbohydrate vibrations. Sharpening of the amide I band near 1645 cm"1 for AFPI at subzero temperatures showed strengthening of a-helix upon cooling. Strong hydrophobic interactions from aliphatic amino acids were seen at -0.5°C, and hydrogen-bonding and involvement of methyl groups were implicated at subzero ii temperatures. Frequency shifts in the O-Ff stretching band of water were observed in the presence of AFP at subzero temperatures. AFP did not prevent ice recrystallization or protein denaturation in fish mince during freeze-thawing. Conformational changes of AFP were observed at subzero temperatures, especially at -0.5°C. This information could be useful to study future applications of A F P in situations where intense ice crystallization formation will be desired or applications such as chemical adjuvants to cryosurgery. i i i TABLE OF CONTENTS Page Abstract i i Table of Contents iv List of Tables vii List of Figures viii List of Abbreviations x Preface ; xi Acknowledgements xii Dedication xiii 1. Introduction 1 1.1. Hypotheses 4 1.2. Objectives 5 2. Literature Review 6 2.1. Fish deterioration and its control 6 2.2. Fish muscle proteins 7 2.2.1. Structure of fish muscle proteins 9 2.2.1.2. Sarcoplasmic proteins 11 2.2.1.3. Myofibrillar proteins 11 2.2.1.4. Stromal proteins 13 2.3. Factors affecting the quality of frozen fish 15 2.3.1. Factors related to water 17 2.3.2. Factors related to lipids 19 2.3.3. Factors related to T M A O demethylase 19 2.4. Properties that can be used as indicators of protein denaturation 2.4.1. Protein extractability 20 2.4.2. ATPase activity 22 2.4.3 Sulfhydryl and disulfide groups 23 2.4.4. Protein hydrophobicity 25 2.4.5. Protein structure by Raman spectroscopy 26 2.4.5.1. Raman spectroscopy of proteins 30 2.4.5.2. Raman spectroscopy of fish muscle proteins 33 2.5. Antifreeze proteins 34 2.5.1. Antifreeze Glycoprotein 35 2.5.1.1. Antifreeze Glycoproteins 1-5 (AFGP 1-5) 36 2.5.1.2. Antifreeze Glycoproteins 6-8 (AFGP 6-8) 36 2.5.2. Antifreeze peptides/proteins 36 2.5.2.1. Type I AFP 37 2.5.2.2. Type II AFP 38 2.5.2.3. Type III AFP 38 2.5.2.4. Type IV AFP 39 2.5.3. Mechanism of antifreeze action 39 iv 2.5.4. Applications in food science 43 2.5.4.1. Recrystallization inhibition 43 2.5.4.2. Modification of ice crystal size 44 2.5.4.3. Effect on gel formation, measured by Ca 2 +ATPase activity 45 2.5.5. Cryosurgery and cryopreservation 46 2.6. Commercial cryoprotectants 47 3. Materials and Methods 52 3.1. Fish fillets: Soaking or spraying with cryoprotectants 52 3.1.1. Analysis 54 3.1.1.1. Water binding capacity and percent thaw drip 54 3.2. Fish mince treated with AFP compared to commercial cryoprotectants 56 3.2.1. Analysis 58 3.2.1.1. Moisture and crude protein 58 3.2.1.2. Expressible moisture and cook loss 58 3.2.1.3. Texture hardness 59 3.2.1.4. Protein extractability 59 3.2.1.5. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) 60 3.2.1.6. Total sulfhydryl (SH) groups and disulfide (SS) bonds 61 3.2.1.7. Hydrophobicity 62 3.2.1.8. Raman spectroscopy 63 3.3. Natural actomyosin treated with AFP compared to commercial cryoprotectants ... 65 3.3.1. Analysis 67 3.3.1.1. Protein concentration 67 3.3.1.2. Calcium ATPase activity 67 3.3.1.3. Surface hydrophobicity 69 3.4. Fish mince treated with AFP, commercial cryoprotectants or their blends 70 3.4.1. Analysis 71 3.4.1.1. FT-Raman spectroscopy 72 3.4.1.2. Differential scanning calorimetry (DSC) of freeze-thaw mince 73 3.4.1.3. Differential scanning calorimetry (DSC) of fresh mince 74 3.4.1.4. Photomicrographs and video-recording observations 74 3.5. Conformational changes of antifreeze protein solutions at subzero temperatures ... 75 3.5.1. Molecular mass measurements and amino acid analysis 76 3.5.2. Raman spectroscopy 76 3.6. Statistical Analysis 77 4. Results and Discussion 78 4.1. Fish fillets: soaking or spraying with cryoprotectants '. 78 4.2. Fish mince treated with AFP compared to commercial cryoprotectants 81 4.2.1. Expressible moisture (%EM), cook loss (%CL) 81 4.2.2. Texture analysis 87 v 4.2.3. Protein extractability in salt and SDS-PAGE 89 4.2.4. Total sulfhydryl 94 4.2.5. Hydrophobicity 96 4.2.6. Raman spectroscopy 99 4.3. Natural actomyosin treated with AFP compared to commercial cryoprotectants .... 104 4.4. Fish mince treated with AFP, commercial cryoprotectants or their blends 107 4.4.1. Expressible moisture 107 4.4.2. Protein extractability in salt and SDS-PAGE 109 4.4.3. FT-Raman spectroscopy 112 4.4.4. Differential scanning calorimetry (DSC) of freeze-thaw mince 124 4.4.5. Differential scanning calorimetry (DSC) of fresh mince 125 4.4.6. Photomicrographs and video-recording observations •.: 128 4.5. Conformational changes of antifreeze protein solutions at subzero temperatures ... 132 4.5.1. Amino acid composition and mass measurements 132 4.5.2. Raman spectroscopy 132 5. Summary and Conclusions 151 References 157 Appendix I MALDI-TOF mass spectrometry of AFGP 171 Appendix II MALDI-TOF mass spectrometry of AFPI 172 Appendix III Upper and lower 95% confidence limits for Raman band intensities for AFGP and AFPI solutions held at different temperatures 173 Appendix IV Results for moisture and crude protein analyses of ling cod mince 175 Appendix V temperature profile of 3 freeze-thaw cycles 176 vi LIST OF TABLES Page Table 1. Functions and applications of muscle proteins 8 Table 2. Functional properties of muscle proteins in meat and meat products 146 Table 3. Tentative assignment of some bands in the Raman spectrum 32 Table 4. Moisture, water binding capacity and percent drip loss of fish fillets after soaking or spraying 80 Table 5. Percent expressible moisture, cook loss and textural hardness of freeze-thawed mince treated with high AFP concentrations 82 Table 6. Percent expressible moisture, cook loss and textural hardness of freeze-thawed mince treated with low AFP concentrations 83 Table 7. Salt extractable proteins for unfrozen and freeze-thawed mince . samples treated with different concentrations of AFP 92 Table 8. Total SH contents of salt extractable proteins from unfrozen and freeze-thawed mince samples treated with different concentrations of AFP 95 Table 9. Surface hydrophobicity values of salt extractable proteins from unfrozen and freeze-thawed mince samples treated with different concentrations of AFP ... 98 Table 10. Band intensities for unfrozen and freeze-thawed mince with added cryoprotectants 103 Table 11. Ca + 2ATPase activity and Surface hydrophobicity of N A M extracts 106 Table 12. Expressible moisture of unfrozen and freeze-thawed mince treated with AFP, commercial cryoprotectants or their blends 108 Table 13. Salt extractable proteins for unfrozen and freeze-thawed mince treated with AFP, commercial cryoprotectants or their blends 110 Table 14. FT-Raman band intensities for freeze-thawed mince 118 Table 15. Amino acid percentage for AFGP and AFPI 133 Table 16. Raman band intensities in the 400-3400 cm"1 region for AFGP and AFPI solutions held at different temperatures 135 Table 17. Intensity ratio (O-H/C-H) for AFGP and AFPI solutions 150 vii LIST OF FIGURES Page Figure 1. Diagrammatic presentation of the myosin molecule 10 Figure 2. Diagrammatic presentation of the myosin molecule 16 Figure 3. Effect of freezing out of water : 18 Figure 4. Relationships between infrared absorption, Rayleigh scattering and Raman scattering 28 Figure 5. Diagrammatic presentation of a basic Raman spectrometer 29 Figure 6. Inhibition of ice crystal growth following AFP adsorption 41 Figure 7. Phase diagram of a binary system in which the solute forms a glass at low temperatures 51 Figure 8. Overview of working plan for fish fillets soaked or sprayed with cryoprotectants 52 Figure 9. Section of fish fillet used for experiment 3.1 53 Figure 10. Overview of working plan for fish mince treated with antifreeze proteins or commercial cryoprotectants 56 Figure 11. Overview of working plan for natural actomyosin treated with antifreeze proteins and cryoprotectants 65 Figure 12. Overview of working plan for fish mince treated with antifreeze proteins, commercial cryoprotectants and their blends 70 Figure 13. Overview of working plan for analysis of antifreeze proteins solutions held at subzero temperatures 75 Figure 14. Visual observations on fish mince samples treated with high concentrations of AFP and with or without cryoprotectants 84 Figure 15. Visual observations on fish mince samples treated with low concentrations of AFP and with or without cryoprotectants 85 Figure 16. SDS-PAGE profiles of ling cod mince salt extractable proteins treated with and without cryoprotectants 93 viii Figure 17. Raman spectra of unfrozen and freeze thaw mince, treated with AFP and with or without cryoprotectants 102 Figure 18. SDS-PAGE profiles for unfrozen and freeze-thawed mince treatments with blends of AFP and cryoprotectants I l l Figure 19. FT-Raman spectra (400 -1800 cm"1 region) of freeze-thawed mince samples ... 117 Figure 20. FT-Raman spectra showing amide I, C H 2 bending regions of freeze-thawed mince treatments 120 Figure 21. Secondary structure fractions estimated from the amide I region for freeze-thawed mince samples 121 Figure 22. FT-Raman spectra in the C-C, C-N stretch and amide III regions of freeze-thawed mince treatments 122 Figure 23. FT-Raman spectra showing stacked and overlaid spectra of the C-H (2932 cm"1) and O-H (3220 cm"1) stretching regions 123 Figure 24. DSC results showing antifreeze activity of freeze-thawed mince samples 126 Figure 25. DSC results showing antifreeze activity of fresh mince samples 127 Figure 26. Photomicrographs of fish mince with antifreeze proteins 129 Figure 27. Photomicrographs of fish mince with antifreeze proteins cont 130 Figure 28. Raman spectra of AFGP and AFPI solutions held at four temperatures 134 Figure 29. Raman spectra of AFGP and AFPI solutions in the amide I and amide III regions 140 Figure 30. Raman spectra of AFGP and AFPI solutions in the 750-1200cm"1 region .... 142 Figure 31. Raman spectra of AFGP solutions in the 2800-3400 cm"1 region 148 Figure 32. Raman spectra of AFPI solutions in the 2800-3400 cm"1 region 149 ix LIST OF ABBREVIATIONS A F AFP AFGP AFPI AFGP-PO AFPI-PO AFGP-PolPO AFPI-PolPO A O A C Ca 2 +ATPase cm °C C L D L DSC E M FT FT-Raman g H 2 0 J kg L M H C mg mL N A M NIR PO POB SD SDS-PAGE sec SEP SH SS SuSo Treha-PO UF WBC w/w AHA antifreeze activity antifreeze proteins (both AFGP and AFPI) low molecular weight antifreeze glycoprotein; fraction 6-8 (used in this research) type-I antifreeze protein antifreeze glycoprotein with 0.3% phosphates (0.15% tripolyphosphate, 0.15% sodium pyrophosphate; w/w) type-I antifreeze protein with 0.3% phosphates (0.15% tripolyphosphate, 0.15% sodium pyrophosphate; w/w) antifreeze glycoprotein with 0.3% phosphates, 2% sucrose, 2% sorbitol antifreeze protein type I with 0.3% phosphates, 2% sucrose, 2% sorbitol Association of Official Analytical Chemists calcium adenosine triphosphatase centimeter degree Celsius cook loss drip loss differential scanning calorimetry expressible moisture freeze-thawed or freeze-thaw Fourier transform Raman gram water joule kilogram liter myosin heavy chain milligram milliliter natural actomyosin near-infrared 0.3% phosphates (0.15% tripolyphosphate, 0.15% sodium pyrophosphate; w/w) phosphate buffer standard deviation sodium dodecyl sulphate poly acrylamide gel electrophoresis seconds salt extractable protein sulfhydryl disulfide 4% sucrose, 4% sorbitol with no phosphates 8% trehalose with 0.3% phosphates unfrozen (not previously frozen) water binding capacity weight/weight heat of fusion of water x PREFACE Part of this thesis was presented at the 42 n d Annual Meeting of the Society for Cryobiology held in Minneapolis, USA. A copy of the abstract was published in the December issue of Cryobiology, 51(2005): 407. It has also been submitted for publication in Cryobiology with the title "Temperature-induced Changes in Raman spectra of Fish Antifreeze Glycoprotein and Type I Antifreeze Protein". Andrea Liceaga, the thesis author, is the principal author who performed the research, analyzed the data, and prepared the manuscript. Eunice C.Y. Li-Chan, A. Liceaga's supervisor, is the coauthor and provided guidance and reviewing of the above research. Part of this thesis was the subject of an oral presentation entitled "Effect of antifreeze proteins on protein structure of frozen ling cod (Ophiodon elongatus) mince" presented at the 56 t h Annual Meeting of the Pacific Fisheries Technologists (Vancouver, BC). This research will be submitted for publication in the Journal of Agricultural and Food Chemistry. Andrea Liceaga, the thesis author, was the principal presenter/author who performed the research, analyzed the data, and prepared the manuscript. Eunice C.Y. Li-Chan, A. Liceaga's supervisor, is the coauthor who reviewed the presentation and provided guidance in the above research and manuscript. Part of this thesis was the subject of a poster presentation entitled "Physicochemical properties of frozen ling cod (Ophiodon elongatus) as affected by fish antifreeze proteins and other cryoprotective agents" presented at the 2005 Annual Meeting of the Institute of Food Technologists (New Orleans, LA) . This research will be submitted for publication in Food Chemistry. Andrea Liceaga, the thesis author, was the principal presenter/author who performed the research, analyzed the data, and prepared the manuscript. Eunice C.Y. Li-Chan, A. Liceaga's supervisor, is the coauthor who reviewed the presentation and provided guidance in the above research and manuscript. (Andrea Liceaga) (Eunice C.Y. Li-Chan) xi A C K N O W L E D G E M E N T S I have to begin by expressing my immense and most sincere gratitude to my supervisor Dr. Eunice C.Y. Li-Chan, without whom I could not have completed this thesis. Her supervision, dedication, patience and guidance throughout my research will always be remembered. I would also like to express my gratitude to the members of my committee Dr. L . Burtnick, Dr. T. Durance and Dr. B. J. Skura for their valuable advice and encouragement. The lessons and knowledge I have learned as a researcher go far beyond this thesis and I have to thank my thesis supervisor and committee members for that. I would like to thank Dr. L . Creagh, Dr. S. Perry and Dr. G. Meng for their valuable assistance with the DSC, mass spec and Raman analysis, respectively. This research could not be completed without the kind donation of the antifreeze proteins by Dr. G. Fletcher and A/F Protein Canada. A very special thank you to Val Skura, Sherman Yee and Dr. Pedro Aloise for their support and technical assistance throughout my research. To my fellow graduate students, thank you for sharing your experiences with me and for helping to make these years at U B C full of fond memories. A big thank you to the F N H Program Assistant, Miss Tram Nguyen, for not only her help in printing and photocopying this thesis but also for her words of encouragement. I would also like to acknowledge the financial support from the National Sciences and Engineering Research Council (NSERC) of Canada, as well as from the University of British Columbia for the "University Graduate Fellowship" (UGF). I wish to thank my parents and siblings for their support during these years in "school". To my husband James for encouraging me to continue when I felt like giving-up and for supporting and understanding why I had to work in the lab for long hours and helping me with my research despite his allergy to fish. xii This thesis is dedicated to my husband, our families and friends. I dedicate this thesis especially to my grandfather*: "Como te lo prometi Abuelo. Esta tesis es para ti" . xiii 1. INTRODUCTION Fish makes a significant contribution to the diets of many communities around the world. According to the Food and Agriculture Organization (FAO) of the United Nations (2002), more than one billion people worldwide rely on fish as a source of animal protein. In a survey of human fish consumption patterns the FAO also indicated that fresh fish (54%) was the most preferred item, followed by frozen (26%), canned (11%) and cured fish (9%). In recent years, changes in consumer lifestyles have resulted in an increased demand for chilled or frozen products that are conveniently packaged and ready-to-cook or serve (Venugopal, 2006a). Within these new product trends, fish mince offers immense scope for development of diverse products such as surimi-based seafood analogs, sausages, frozen battered and breaded products (Park, 2000; Venugopal, 2006a). Fish mince is the component produced from the separation of the flesh from skin, bones, and fins into a comminuted form. Minces are predominantly prepared from gadoid fish, which include cod, hake, haddock, pollock, and croaker (Claus et al., 1994). Freezing is commonly applied to fish mince in the production of these value-added products. Similarly, high-quality fish are usually filleted, frozen, and eventually sold to consumers (Lian et al., 2000; Santos-Yap, 1996). Freezing and frozen storage is an excellent method of preserving the organoleptic attributes and protein functionality of fish flesh during prolonged periods of time (Careche et al., 1999). Freezing and subsequent cold storage are particularly useful in making seasonal species of fish, like ling cod, herring and mackerel, available all year round. The low temperatures applied 1 during freezing and frozen storage slow down the rate of product quality loss, caused by the action of enzymes and microorganisms (Rodriguez-Herrera et al., 2002). Although freezing initially helps maintain quality, frozen storage may eventually cause fish muscle deterioration indicated by the loss of juiciness and an increase in meat toughness as proteins in the muscle are denatured (Sych et al., 1990). Lanier (2000) defines denaturation as the unfolding of amino acid polypeptide chains. Denaturation exposes more reactive surfaces of a protein, which can lead to protein-protein interactions including those leading to aggregation. In the commercial production of fish mince, surimi and fishery by-products, cryoprotectants are commonly used as a mode of slowing down or preventing fish muscle deterioration. Commercial cryoprotectants currently used require the use of high concentrations (typically 4% sucrose and 4% sorbitol, w/w) in order to impart satisfactory cryoprotection. These concentrations of the commercial cryoprotectants have a tendency to impart a sweet taste to the final product, which is not always the preferred case (Yoon and Lee, 1990). Alternative ingredients are desirable in the context of sensory attributes and issues related to health (e.g. lower calorie or sugar-free products for diabetics). Recent discoveries on prevention of ice crystal formation and recrystallization by antifreeze proteins (AFP) have led to speculations on their possible applications to a variety of food systems (Feeney and Yeh, 1998). AFP have been proposed for use in frozen food products such as ice cream. However, it is known that consumer acceptance of "foreign" proteins being added to foods is not widespread; in other words, consumers do not easily accept having fish components in non-fish food products. 2 To present knowledge there is no literature available on the use of antifreeze proteins in frozen fish fillets and fish mince and there is limited information on the function of antifreeze proteins upon interaction with other food components,""such as muscle proteins. 3 1.1 Hypotheses Spraying and soaking are effective techniques to incorporate solutions of sucrose and sorbitol into ling cod fillets. Antifreeze proteins are as effective as a commercial blend of cryoprotectants (sucrose, sorbitol, phosphates) in maintaining ling cod mince protein functionality and stability upon freeze thaw abuse. Blends of antifreeze proteins with sucrose, sorbitol and phosphates are more effective than the commercial blend alone in decreasing protein functionality changes of ling cod mince upon freeze thaw abuse. Antifreeze proteins are as effective as the commercial blend in stabilizing natural actomyosin extracted from ling cod muscle. Antifreeze glycoprotein and Type I antifreeze protein solutions exhibit different structural and conformational changes when held at subzero temperatures. 4 1.2. Objectives The main objective of this study was to investigate i f antifreeze proteins could prevent fish protein denaturation during freezing and freeze thaw conditions, compared to mince treated with commercial cryoprotectants presently used by fish processors. The specific objectives of this study were to: 1. Establish a technique for spraying and soaking ling cod fillets with sucrose and sorbitol solutions, and to evaluate the potential for applying this approach as a method of incorporation of antifreeze protein solutions into fish fillets. 2. Evaluate the effects of fish antifreeze proteins on protein denaturation and physico-chemical changes of ling cod mince following freeze thaw abuse, and compare these effects to mince treated with sucrose, sorbitol and polyphosphates. 3. Determine the effects of fish antifreeze proteins on Ca-ATPase activity and surface hydrophobicity of ling cod natural actomyosin systems following freeze thaw abuse. 4. Evaluate the effects of blends of antifreeze proteins with commercial cryoprotectants on protein denaturation and physico-chemical changes of ling cod mince following freeze thaw abuse. 5. Analyze the structural and conformational changes of antifreeze glycoprotein and Type I antifreeze protein solutions held at subzero temperatures, using Raman spectroscopy. 5 2. LITERATURE REVIEW 2.1. Fish deterioration and its control Live fish muscle is characterized by being in a relaxed and flexible state. Immediately after death, rigor mortis sets in, making the whole body of the fish contracted and rigid. In addition, the initial biochemical quality of the muscle is prone to rapid changes due to cessation of respiration, breakdown of cellular ATP, autolytic action of proteolytic enzymes on the muscle, oxidation of lipids, and the metabolic activities of microorganisms (Venugopal, 2006b). Freezing and frozen storage are one of the most important long-term preservation techniques for fish, fish mince and value-added fishery products (Rodriguez-Herrera et al., 2002). These techniques can slow down (bio)chemical processes and microbial growth (Rodriguez-Herrera et al., 2002). However, according to Mackie (1993), when muscle tissue is frozen and stored in the frozen state it will inevitably lose some of its functional qualities. The loss of such qualities is usually observed as decreased juiciness and increased toughness. The quality of fish fillets, surimi and minced fish during frozen storage is affected by the storage temperature, storage period, and the type of cryoprotectants used. In order to maintain good quality and prevent such changes, rapid freezing and use of cryoprotectants is required. Cryoprotectants can prevent protein denaturation during frozen storage by increasing the surface tension of water as well as the amount of bound water. By minimizing ice crystal growth and migration of water molecules from the protein, cryoprotectants thus stabilize the protein in its native form during frozen storage (Carpenter and Crowe, 1988; Claus et al., 1994). The most commonly used 6 cryoprotectants are sucrose, sorbitol and polyphosphates; however, the use of sugar and polyols tends to impart a sweet taste to the final product, and this has led to a search for alternative cryoprotectants (Yoon and Lee, 1990; Herrera et al., 2002; Boonsupthip and Lee, 2003). 2.2. Fish muscle proteins Muscle proteins are highly nutritious as they provide our diet with the essential amino acids. They also impart specific properties to meat and meat products, including texture, gelation, water-binding, emulsification, mouthfeel, and physical stability during storage (Table 1). The functional properties exhibited by muscle proteins cannot be easily reproduced by any other food proteins or nonprotein functional ingredients (Xiong, 2004). 7 Table 1. Functions and applications of muscle proteins. Protein Source Function and/or application In situ protein Sarcoplasmic protein Lean tissue Myofibrillar protein Lean tissue Collagen Ingredient protein Myofibrillar protein Collagen Gelatin Surimi Plasma protein Connective tissue Meat trimming; animal by-products Pig, turkey, chicken and fish skins Hydrolyzed collagen from animal skins, hides, and cartilage; cooked pig and poultry skins Fish; mechanically separated and washed meat; meat by-products Beef plasma Water holding; color; fat emulsification Meat binding; water holding; fat emulsification; texture of meat product Texture of meat products Meat binding; water holding; fat emulsification Water holding; juiciness; product yield Water holding; emulsion stabilization; meat binding Seafood analogs; meat binding; water holding; fat emulsification Meat binding; water holding; protease inhibitors for surimi Adapted from Xiong (2004). 8 2.2.1. Structure of fish muscle proteins Fish skeletal muscles can be classified into white and dark muscle. Dark muscle lies along the side of the body next to the skin, and can make up to 30% of the fish muscle, depending on the species (Green-Walker and Pull, 1975). Dark muscle is used by fish for sustained swimming activities, functioning aerobically using lipids for fuel. Lean species of fish such as flounder, hake, sole, cod and pollock have very small amounts of dark muscle (Santos-Yap, 1996). White muscle constitutes the majority of fish muscle; it has minimal myoglobin and restricted blood supply. This muscle is used for anaerobic activities such as short bursts of swimming activities (Pitcher and Hart, 1982). The skeletal muscle is composed of fibers or cells which are arranged in parallel fashion and are held together by connective tissue of collagen and elastin: In fish, the cells are bound together parallel to one another to form segments or myotomes of muscle (Mackie, 1993). Within the muscle cells are myofibrils, which are long thin contractile elements that give the characteristic striated pattern of light (I bands) and dark bands (A bands). The I bands are divided by thin Z lines, and the A bands by a light area (H zone), down the center of which is a darker M line. The sarcomere is the contractile assembly; it lies between two Z lines (Figure 1) and is the repeating unit of the myofibril. The A bands are composed of thick filaments which are formed from an ordered arrangement of 400 myosin molecules, the tail sections are embedded into the thick filament shaft. The main protein of the thin filament is actin, which forms two long helical strands wound around one another, followed by tropomyosin and the troponin complex (Mackie, 1993; Xiong, 2004). 9 Z cDsk A band I band 1 A rS Muscle libet or call 'I i*~ Z Z X Myofilaments # . < v G _ a c t i n Jj "Pe* molecules - — : • '— : • F-ac l i n filament Myosin molecule Figure 1. Fish muscle cells and component myofibrils showing arrangement of thick and thin filaments of the contractile mechanism. Source: Mackie (1993). 10 2.2.1.2. Sarcoplasmic proteins The sarcoplasmic or myogen fraction makes up about 30% of the total muscle proteins. It consists of a large family of proteins that share the property of being soluble in water and dilute salt solution. Sarcoplasmic proteins include myoglobin, other albumins and hundreds of enzymes involved in metabolism (Pearson and Young, 1989). Myoglobin is present in trace amounts in white fish muscle, whereas in dark muscle it can be present in concentrations similar to those of red meat such as beef. Myoglobin is believed to contribute to the oxidative changes within the flesh during postmortem storage of fish (Mackie, 1993). 2.2.1.3. Myofibrillar proteins Myofibrillar proteins constitute the contracting structure of the muscle fibers. They make up about 60% of the total muscle proteins and are soluble in salt solution (Lanier, 2000). These proteins are mainly classified as the structural proteins (myosin and actin), and the regulatory proteins (tropomyosin, troponins C, I and T and actinin) (Suzuki, 1981; Lanier, 2000). Myosin is the most abundant myofibrillar protein in fish muscle. The molecule consists of two heavy chains (200 kDa) associated non-covalently with two pairs of light chains (16 to 28 kDa). Figure 1 illustrates the myosin molecule. The long tail of the molecule consists of two polypeptides in a coiled cc-helix terminating in two globular heads at one end (Mackie, 1993). Associated with the globular heads are four myosin light-chain (LC) molecules. Two types of L C are associated with the myosin heads. One type is referred to as alkaline light chains because they are dissociated from myosin 11 heavy chain (MHC) under alkaline conditions. The second type is known as DTNB light chains because they are dissociated by 5,5'-dithiobis 2-nitrobenzoic acid (Mackie, 1993; McCormick, 1994). In addition to acting as a structural protein, myosin also is an enzyme with ATPase activity. The ATPase activity is dependent on the association of the alkaline light chains with the head region of the M H C . Thus, removal of the alkaline light chain results in loss of myosin ATPase activity. In contrast, the DTNB light chains seem to have a regulatory function mediating the level of ATPase activity by affecting the calcium binding abilities (McCormick, 1994). Although the structure of fish myosin is similar to other mammalian myosin, fish myosin is more unstable than the mammalian. Studies on M g 2 + ATPase of actomyosin and C a 2 + ATPase of myofibrils suggest that the stability of fish myosin is species specific and associated with the temperature of its habitat (Sijo and Prakash, 2005). Myosin is susceptible to proteolytic cleavage. Controlled digestion with trypsin or chymotrypsin cleaves myosin into two components, light meromyosin (LMM) and heavy meromyosin (HMM); the latter contains the globular head region of the molecule. When H M M is treated with papain, it separates into a long tail-like fragment known as the neck portion (S2) and the head portion (SI), whereas trypsin or chymotrypsin cleavage of the myosin rod produces the L M M and the neck portion or S2 (Suzuki, 1981; Mackie, 1993; McCormick, 1994). Actin is the second most abundant myofibrillar protein, followed by tropomyosin and troponin (Mackie, 1993). Actin is located within the thin filaments of the myofibrils, accounting for 15-30% of the total myofibrillar content in the muscle. The thin filament in the muscle consists of two coiled strands of polymerized actin molecules (filamentous 12 or F-actin) that wrap around each other to form a double helix. Globular actin (G-actin) is the monomeric form with a molecular weight of 40 kDa (McCormick, 1994). Tropomyosin and troponin regulate muscle contraction. They are also located within the thin filaments of muscle fibers, accounting for 10% of the myofibrillar protein (Suzuki, 1981). The molecular weight of tropomyosin is 68 kDa. Troponin is formed by three components, namely troponin-T (30 kDa) which combines with tropomyosin, troponin-I (21 kDa) which inhibits ATPase of actomyosin, and troponin-C (19 kDa) which combines with calcium (Suzuki, 1981; Mackie, 1993). Actin and myosin are easily extracted from fish meat with a salt solution as they combine to form a viscous solution of actomyosin filaments. Natural actomyosin (NAM) is a complex composed of myosin, fibrous actin (F-actin), tropomyosin, and the troponins (Asghar et al. 1985). The characteristics of actomyosin and N A M are similar to those of myosin, in which ATPase activity is retained. 2.2.1.4. Stromal proteins Stromal or "connective tissue" proteins make approximately 10% of the total muscle proteins and compose the basic structural elements in the connective tissue of higher animals (Pearson and Young, 1989). These proteins are insoluble in water or dilute salt solution, imparting toughness, shape and protection to the skeletal muscles. The principal proteins in this group are collagen and elastin (Shahidi, 1994). When collagen is heated it can be converted to a water soluble gelatin. At the same time, most of the connective tissue disappears. Elastin is very resistant to moist heat and cooking does not affect elastin in connective tissue (Suzuki, 1981). 13 L M M S2 H M M -SI-ROD (2 coiled a-helices) Figure 2. Diagrammatic presentation of the myosin molecule. LC: light chain, SI: globular subfragment 1, S2: neck subfragment 2, L M M : light meromyosin, HMM: heavy meromyosin, ROD: oc-helical coiled rod. Numbers in parentheses indicate molecular weight in Daltons. Adapted from: Mackie (1993) and McCormick (1994). 14 2.3. Factors affecting the quality of frozen fish Freezing and frozen storage of fish can lead to detrimental changes in functional properties and texture of the fish (Careche and Li-Chan, 1997). For example, after cooking the connective tissue becomes tough, chewy, rubbery, stringy, or fibrous. Gaping refers to a phenomenon in which sheets of connective tissue in fish muscle (myocommata) fail to hold the myotomes together. This is seen particularly in frozen fish since the connective tissue becomes weakened or damaged due to the formation of ice crystals. Gaping in fish fillets has serious economic consequences leading to unacceptability and wastage of otherwise valuable material. Badly gaped fillets cannot be skinned or sold from open display and have only a limited use in certain products like minced meat (Femeena et al., 1999). Gaping is also accompanied by a loss of protein functionality. Xiong (2004) defines functionality of meat proteins as the physical and chemical performances of proteins during processing and storage that affect the texture-related properties of the final meat products. Functional properties are essential for the manufacture of fish value-added products, surimi and surimi-based products (Shenouda, 1980; Sych et al., 1990). The main functional properties in processed meats including fish are gelation, emulsification, and water-holding capacity (Table 2). There are several factors that affect fish protein denaturation during frozen storage. Dehydration and salt concentration, induced by ice crystal formation, combine with lipid oxidation, formation of free fatty acids, and in certain fish species formaldehyde, to produce denaturation of proteins, loss of functionality and texture (Shenouda, 1980). 15 Table 2. Functional properties of muscle proteins in meat and meat products, including fish. Category Property Mode of action Food example Hydration Water binding, holding, and absorbing Protein -water interaction via hydrogen bonds; water entrapment in myofibrillar lattices Fresh meat; pumped/injected meats; marinated and other processed meats Solubility Protein-water interaction . Salted meats; via hydrogen bonds; tumbled/massaged meats protein charge repulsion by the presence of N a + and CI" ions and phosphates Swelling Structure/ Gelation texture Water penetration into myofibrillar lattices Matrix formation by extracted myofibrillar proteins; collagen proti protein interaction Marinated meats; pumped/injected meats Restructured meat rolls and loaves; luncheon meats; gelatin gel foods Cohesion/adhesion/ binding Gels of salt-soluble proteins serve as a binding agent; surimi protein-protein interaction; gelation of collagen protein Restructured meat rolls and loaves; boneless ham; kamaboko; shellfish analogs; gelatin-bound luncheon meats Surface Emulsification Protein adsorption on fat Sausage; frankfurters; particles to reduce surface bologna tension; formation of rigid protein membrane in fat emulsion Adapted from: Xiong (2004). 16 The loss of these functional properties is caused by the unfolding of myofibrillar proteins, which expose nonpolar amino acids allowing them to become available for hydrophobic interactions with similar groups. This in turn leads to protein aggregation, changes in texture, water holding capacity, gelling ability, loss of ATPase activity, total protein solubility, as well as chemical reactions of amino acid residues in proteins with endogenous formaldehyde (Niwa, 1985; Sikorski and Kotakowska, 1994). 2.3.1. Factors related to water Water plays an important role in protein denaturation and functionality during freezing and frozen storage. The structure of fish muscle and the structural changes in proteins can be accompanied by changes in the association between protein and water molecules and in the distribution and mobility of the water in the tissue (Herrero et al., 2005). During freezing, ice crystals form nuclei in the extracellular fluid between muscle cells. In addition, water can diffuse from within the cell to the growing ice crystal, forming structures large enough to penetrate membranes and leading to further damage to membranes. This can also occur with fluctuations in temperature that allow for large ice crystal formation at the expense of smaller ones (Mackie, 1993). The conformation of most native proteins has the hydrophobic side chains buried inside the protein molecule. However, some of these hydrophobic side chains are exposed at the surface of the molecule itself. Matsumoto (1979) suggested that the water molecules arrange themselves around these exposed hydrophobic side chain groups so as to minimize the energy of the interface between water and hydrophobic groups. A network of hydrogen bonds is formed by these water molecules, contributing to the three-17 dimensional structure of the proteins. As shown in Figure 3, as the water molecules freeze out they migrate to form ice crystals resulting in the disruption of the organized H -bonding systems that stabilize protein structure, resulting in the formation of intramolecular cross-linking within the protein molecule, causing the loss of the three-dimensional structure, or between adjacent protein molecules leading to intermolecular protein-protein interactions and consequently aggregation (Matsumoto, 1979; Morrison, 1993; Santos-Yap, 1996; MacDonald et al., 2000). Hydrophobic protein group Figure 3. Proposed effect of freezing out of water on the water mediated hydrophobic-hydrophilic linkage in a protein molecule. Adapted from: Santos-Yap (1996). 18 2.3.2. Factors related to lipids Lipids, in particular oxidized lipids, may affect the hydrogen bonds and hydrophobic interactions in the proteins of frozen fish. It has been suggested that free radicals can react with protein side chains and that carbonyl groups of the oxidized lipids may participate in covalent bonding to form stable protein-lipid aggregates (Sikorski and Kotakowska, 1994). Most of the fat deposits in fish are removed when fish are headed, gutted, and skinned. However, a small percentage of membrane phospholipids are present in fish muscle. These phospholipids are highly unsaturated and are often in contact with muscle heme iron and are therefore sensitive to spoilage by oxidation. This oxidation causes off-flavours and may hasten denaturation of the myofibrillar proteins (Lanier, 2000). Saeed and Howell (1999) showed that fatty fish were more susceptible to toughening during frozen storage, and that there was a transfer of free radicals from fish oil to proteins, followed by aggregation of the proteins and amino acids. In another study involving frozen cod (Gadus morhud), Badii and Howell (2002) concluded that both lipid oxidation products and ice crystal growth were the major factors contributing to toughening of lean fish in frozen storage. 2.3.3. Factors related to TMAO Demethylase A l l marine fish contain trimethylamine oxide (TMAO), a water-soluble nitrogenous compound used by fish for osmo-regulation. T M A O demethylase is an enzyme that is especially prevalent in gadoid (cod-like) species (Claus et al., 1994). During frozen storage, this enzyme degrades T M A O to formaldehyde (FA) together with dimethylamine oxide (DMA) with subsequent cross-linking of F A to muscle proteins 19 (Lanier, 2000). The result of this cross-linking makes fish muscle hold its free water loosely like a sponge (Santos-Yap, 1996). In addition, a large accumulation of D M A and F A in the muscles of frozen fish is generally accompanied by a loss in extractability of myofibrillar proteins, decreased enzymatic activity, as well as changes in hydrophobicity (Sikorski and Kotakowska, 1994). Formaldehyde has been shown to accelerate the formation of high molecular weight polymers in myosin isolated from cod (Ang and Hultin, 1989) and in natural actomyosin from hake (Del Mazo et al., 1994). Careche.and Li-Chan (1997) reported that addition of F A to cod myosin produced changes in the secondary structure of the myosin, causing exposure of the hydrophobic aliphatic groups, eventually leading to the appearance of covalent cross-links. 2.4. Properties that can be used as indicators of protein denaturation in fish muscle during freezing and frozen storage 2.4.1. Protein extractability Protein solubility has been correlated to protein functionality, such as gelation, foaming ability, emulsification, water binding ability and whipping properties, with salt soluble proteins being the major contributors to muscle functionality (Vojdani, 1996). For this reason, changes in protein extractability can be a useful tool in order to evaluate the extent to which protein denaturation and aggregation have occurred during frozen storage or after temperature abuse situations, such as freeze-thawing cycles (Del Mazo et al., 1999; Lian et al., 2000). In frozen storage, actin and myosin become gradually less extractable in salt solutions, forming high molecular weight protein aggregates (Careche and Li-Chan, 20 1997; Careche et al., 1999). These protein aggregates are thought to be stabilized by hydrophobic interactions as well as by disulfide bonds and other crosslinks (Tejada et al., 1996; Careche et al., 1998a). The unextracted residue has a structure resembling the sarcomere, which tends to be more pronounced the longer the frozen storage time. Careche et al. (2002) showed that the unextractable residue from fish muscle stored at -10°C was composed of a network of myosin from the residual sarcomere A-band, in the form of interconnected and/or aggregated thick filaments. The quantity of myofibrillar protein extracted from the muscle by salt solution with 0.45 to 0.60 ionic strength is commonly used as an indication of whether or not denaturation has taken place in myofibrillar proteins (Suzuki, 1981). During freezing and frozen storage myosin and actin become less extractable in salt solutions as different bonds are formed in percentages that change with storage time and temperature, producing protein aggregates. These aggregates will grow in size, so that as frozen storage progresses they can no longer be extracted in 0.6 M NaCl and pass into the fraction not extractable in salt solution (Tejada et al., 1996). Careche et al. (1998a) found that aggregates consisted largely of myosin and actin. However, myosin was the most important constituent, contributing more to the formation of covalent bonds. The authors also found that aggregates formed at -30°C were extracted in solutions such as 2% SDS, which cleave secondary interactions whereas at -20°C the aggregates were not totally extracted in 2% SDS plus 5% (3-mercaptoethanol, indicating a more important role of non-disulfide covalent bonds which could be caused by agents such as formaldehyde or oxidized lipids (Careche et a l , 2002). 21 2.4.2. ATPase activity The catalytic site for adenosine triphosphatase (ATPase) activity and the contractile mechanism are found in the globular head portion of the two.myosin heavy chains. ATPase is capable of hydrolyzing the terminal phosphate group of adenosine triphosphate (ATP) to give adenosine diphosphate (ADP), thus converting chemical to mechanical energy during muscle contraction (Mackie, 1993). The ATPase enzymatic activity of myosin can be used as an index of change in the configuration of the enzyme around the active site. This enzymatic activity can be determined by measuring changes in the amount of inorganic phosphate present in the muscle. The loss of enzymatic activity reflects the extent of freeze damage and alteration of the protein structure in the myosin head region (Suzuki, 1981; Santos-Yap, 1996). During the measurement of ATPase activity, an excess of ATP is added to myosin at a high concentration of potassium chloride (1.5-3.0 mg/mL 0.6 M KCl) and 2 m M magnesium chloride. The reaction is then stopped by adding trichloroacetic acid (TCA) and the amount of inorganic phosphate (Pi) liberated can be measured. Divalent cations exert a strong effect on myosin ATPase. The activity of myosin ATPase is low in the presence of m M concentrations of magnesium ions, but the activity increases almost 100 fold when calcium ions replace magnesium. Ionic strength, temperature and pH also affect the activity of myosin ATPase and actomyosin ATPase reactions (Warris, 2000). Myosin has been shown to lose ATPase more rapidly than actomyosin, indicating that without the protective effect of actin, myosin is more prone to inactivation (Shenouda, 1980; Susuki, 1981). The effect of freezing and frozen storage on fish muscle proteins and ATPase activity is well documented. Del Mazo et al. (1999) found a significant 22 decrease in ATPase activity of N A M extracted from hake fillets within 14 days storage at -20 and -30°C. Ramirez et al. (2000) reported loss of ATPase activity in tilapia myosin stored at -20°C, suggesting that myosin heads could be involved in the fish myosin aggregation. However, the authors also indicate that loss of ATPase activity is not necessarily synonymous with aggregation because it is possible to have no aggregation and still lose 100% of the activity simply by denaturation of the active site. Benjakul and Bauer (2000) also reported a significant decrease in C a 2 + and Mg 2 +-ATPase activities in cod muscle proteins as freeze-thaw cycles increased, indicating denaturation of myosin and a disruption of the actin-myosin complex. This loss of activity was postulated to be due to the tertiary structural changes caused by ice crystals, the increase in ionic strength of the system, and the rearrangement of protein via protein-protein interactions. 2.4.3. Sulfhydryl and disulfide groups Sulfhydryl (SH) groups are considered to be the most reactive functional group in proteins, being easily oxidized to disulfide (SS) groups (Sultanbawa and Li-Chan, 2001). Oxidation of SH groups to SS bonds are thought to be involved, along with hydrophobic interactions, in the denaturation and aggregation of myofibrillar proteins in frozen fish (Tejada et al., 1996; Sikorski and Kotakowska, 1994; Careche et al., 1998a). An intermolecular disulfide bond is formed by oxidation of the reactive (-SH) groups of two cysteine molecules on neighbouring protein chains. The formation or natural occurrence of disulfide bonds between amino acid residues within a protein (intramolecular disulfide bonds) can be converted to protein-protein (intermolecular) disulfide bonds (cross-links) through disulfide interchange (Lanier, 2000). 23 The Ellman's method (Ellman, 1959) is the most widely used method for assaying the content of - S H and SS groups. In this method, 5,5'-dithio-2-nitrobenzoate (DTNB) reacts with SH groups to produce a yellow substance (5-thiobis(2-nitro)benzoic acid; TNB) with a maximum absorbance at 412 nm. The reaction is carried out in the presence of denaturants to expose the buried SH groups (Beveridge et al., 1974). In addition, a quantitative method can be used to determine the total number of SS + SH groups in proteins (Thannhauser et al., 1984). In this method, SS groups of the peptide or protein are first cleaved quantitatively by excess sodium sulfite (pH 9.5) in the presence of 2 M guanidine thiocyanate as the denaturant to make the SS groups accessible. The reaction with sulfite leads to thiosulfonate and a free SH group, with the concentration of the SH group being determined by the reaction with disodium-2-nitro-5-thiobenzoate (NTSB). Protein aggregation can be monitored by the changes in total and reactive SH groups. Ramirez et al. (2000) working with tilapia myosin, reported a decrease in total and reactive SH groups after 5 days frozen storage and confirmed the importance of disulfide bonds in frozen-induced aggregation of fish myosin in solution. Benjakul and Bauer (2000) showed a decrease in total and reactive SH groups in cod muscle proteins upon freeze-thaw cycles as changes in . the actomyosin tertiary structure took place allowing for oxidation of SH groups into disulfides. Sultanbawa and Li-Chan (2001) reported a decrease in total SH and an increase in SS bonds in ling cod N A M without cryoprotectants after freezing, whereas the SH groups and SS bonds remained the same for treatments with cryoprotectants. The authors hypothesized that SS bond formation was a secondary process of aggregation during freeze denaturation in the absence of cryoprotectants. 24 2.4.4. Protein hydrophobicity The formation of intermolecular hydrophobic interactions results from the thermodynamic response of protein surfaces exposed to the water in which they are dispersed or solubilized. The interior of the folded protein chain has a greater density of hydrophobic amino acids. Conversely, the amino acids on the surface of the undenatured (native) protein are largely hydrophilic (Lanier 2000). When the protein denatures (unfolds), the hydrophobic core is exposed leading to protein-protein interactions and aggregates (Tejada et al., 1996; Careche and Li-Chan, 1997; Careche et al., 1998a). A common approach for quantifying protein surface hydrophobicity (S0) is through fluorescent probe methods such as l-anilinonaphthalene-8-sulfonate (ANS) and cis-parinaric acid (CPA) (Nakai et al, 1996; Alizadeh-Pasdar et al., 2004). According to Mackie (1993), an increase in fluorescence could be indicative of a more hydrophobic environment which would occur when a protein undergoes denaturation, with the exposure at the surface of hydrophobic groups which are normally buried in the interior of the molecule. Herrera and Mackie (2004) showed an increase of surface hydrophobicity in fish muscle upon freezing, indicating that freezing and frozen storage modified the structural organization of water, and therefore the network of hydrogen bonds so that buried hydrophobic residues in the native protein became exposed upon denaturation (Ang and Hultin, 1989). Del Mazo et al. (1999) also reported an increase in surface hydrophobicity of N A M extracted from hake fillets stored at -20 and -30°C. In contrast, Benjakul and Bauer (2000) reported an initial increase in surface hydrophobicity for cod muscle proteins subjected to 3 freeze-thaw cycles, followed by a decrease in hydrophobicity when proteins were subjected to 5 freeze thaw cycles. The authors 25 indicated that the decrease in surface hydrophobicity after an increased number of freeze-thaw cycles was probably caused by the association of hydrophobic groups via hydrophobic interactions. Sultanbawa and Li-Chan (2001) also reported a decrease in surface hydrophobicity in frozen ling cod N A M . Possible causes were attributed to the high concentration of N A M in the study (100 mg/mL), which could have produced aggregates held by strong hydrophobic interactions. 2.4.5. Protein structure by Raman spectroscopy Raman spectroscopy is a vibrational spectroscopic technique which has proved to be a useful analytical tool in the investigation of the structure of biological materials. When light is scattered from a molecule most photons are elastically scattered. The scattered photons have the same frequency and wavelength as the incident photons. However, there is a small fraction of light that is scattered at optical frequencies different from, and usually lower than, the frequency of the incident photons. The term "Raman effect" is used to describe the process leading to this inelastic scatter. Conversely, the elastic collision of the photons and sample molecules is known as Rayleigh scattering (Carey, 1982; Tu, 1982). The inelastic collision gives rise to non-resonance and resonance Raman scattering. Non-resonance Raman scattering results when the energy of the incident light is not sufficient to excite the molecule to a higher electronic level. Instead, Raman scattering results in changing the molecule from its initial vibrational state to a different vibrational state (Figure 4). When the energy of the incident light is sufficient to excite the molecule to a higher electronic level, the scattering is called "resonance Raman scattering" (Carey, 1982; Tu, 1982; Li-Chan, 1996). 26 A basic design of a typical Raman spectrophotometer is shown in Figure 5. Intense monochromatic radiation (e.g. argon laser) is focused on the sample. The resulting scattered light is gathered by collection optics and directed to a dispersing system (typically a double or triple monoehromator), which spatially separates the scattered light on the basis of frequency. The Raman spectrum is detected and recorded either sequentially by a single photomultiplier used with a scanning monoehromator, or simultaneously by a multichannel detector. Finally, computers or microprocessors store the data for further analysis (Li-Chan et al., 1994). Raman spectroscopy can be applied to study protein structure in both solid and liquid food systems without the need to extract a purified component (Li-Chan, 1996). Raman spectroscopy is useful in studies of proteins in aqueous solution and does not require chemical modifications of proteins, especially at high protein concentrations (Ogawa et al., 1999). One difficulty encountered with Raman spectroscopy is that the Raman scattering is a relative weak optical effect that requires visible lasers for efficient excitation. In many biomaterials the fluorescence induced by this incident light is much more intense than the Raman scattered light, making the Raman signals difficult to measure giving a lower signal to noise ratio (Susi and Byler, 1988; Marquardt and Wold, 2004). Various techniques have been developed in Raman-related research to enhance its sensitivity and to expand its applications. One such technique is the development of a Fourier transform (FT) Raman spectrometer that uses near-infrared (NIR) excitation (with laser excitation at 1064 nm) since fluorescence is usually negligible at this lower energy excitation wavelength (Liao et al., 2004; Marquardt and Wold, 2004). 27 T Vo v 0 v 0 v 0 V; r\ Vo Vo + Vi ft Excited electronic states A infrared absorption Rayleigh Stokes' scattering I Anti-Stokes' 1 Raman Scattering Ground electronic states Stokes' Resonance Raman Scattering Figure 4. Relationships between infrared absorption, Rayleigh scattering and Raman scattering. v 0 = Frequency of the incident light beam; V; = Av (Raman shift) for Raman scattering; frequency at which incident electromagnetic radiation is absorbed in the infrared region. Adapted from: Li-Chan (1996). 28 collection optics sample Laser ± collimation optics double monochromator PMT 1 Detector (DC; PC) 1 recorder 11. Computer C R T display Figure 5. Diagrammatic presentation of a basic Raman spectrometer like the one used in this study. PMT, photomultiplier tube; DC, direct current amplification; PC, photon counting; CRT, cathode ray tube. Adapted from: Li-Chan et al., 1994. 29 2.4.5.1. Raman spectroscopy of proteins The first application of Raman spectroscopy to study myosin structure was reported by Carew et al. (1975), and later Barret et al. (1978) reported observations of conformational changes in myosin induced by inorganic salts. The Raman spectrum of a protein consists of contributions from various amino acid side chain vibrations as well as vibrations originating from the polypeptide backbone (Carey, 1982; Tu, 1986). Much of the original work of band assignment started with simple amides, such as N-methylacetamide, with the amide I and amide III vibrational modes being predominant in the Raman spectrum. These two modes can be described as predominantly C=0 stretch and mixed C-N stretch, N - H bending vibrations, respectively (Painter, 1984). The vibrational spectral characteristic of the amide (peptide) bond of proteins as observed in the amide I and III bands, are most useful for the investigation of the protein's secondary structure (a-helix, (3-sheet, and random coil). For example, in the amide I region, proteins containing a-helical and (3-sheet structures generally show bands around 1645-1657 cm - 1 and 1665-1680 cm - 1 respectively. Proteins with high content of random coil structure show an amide I band around 1660 cm - 1 . Conversely, the amide III region shows a-helix around 1260-1300 cm - 1 , (3-structures around 1238-1245 cm' 1 and random coil structure near 1250 cm"1 (Painter, 1984; Li-Chan et al., 1994). Raman spectroscopy is also a useful tool to detect the presence of disulfide bonds from an intense band assigned to the disulfide (S-S) stretching vibration in the 500-550 cm - 1 region. Most naturally occurring proteins and peptides that contain S-S bonds show a band near 510 cm"1. This corresponds to the lowest potential energy conformation of the S-S bond, 'gauche-gauche-gauche' conformation of C-C-S-S-C-C. Additional bands 30 at 525 and 540 cm"1 have been assigned to the 'gauche-gauche-trans' and 'trans-gauche-trans' conformations, respectively (Li-Chan et al., 1994). The intensity of the Raman spectral bands assigned to aromatic amino acids can be used to monitor the polarity of the microenvironment or involvement in hydrogen bonding. For example, a decrease in the intensity of a band near 760 cm"1 will indicate the exposure of tryptophan (Trp) residues from a buried, hydrophobic microenvironment to the aqueous solvent. The microenvironment around tyrosine (Tyr) residues is monitored by a pair of bands located at 830 and 850 cm"1 which can indicate if the Tyr residues are exposed, acting as hydrogen bond donor or acceptor; or buried, acting as hydrogen bond donor (Tu, 1986; Li-Chan, 1996). The carboxyl group vibrational bands can be applied to monitor state of ionization of aspartic (Asp) and glutamic (Glu) acids since the ionized group (COO") exhibits a band at 1400-1420 cm"1 while the undissociated form (COOH) exhibits a band at 1700-1750 cm"1. By comparing the intensities in these regions the relative ionization state of the carboxyl groups can be estimated (Tu, 1986). The C - H stretching and bending modes of aliphatic amino acids appear near 2800-3000 cm"1 and 1400-1500 cm"1, respectively (Tu, 1986; Li-Chan, 1996). Changes in band intensity have been applied to monitor hydrophobic interactions between aliphatic groups (Li-Chan et al., 1994). Table 3 lists the tentative assignment of some bands that are useful in the interpretation of protein structure. 31 Table 3. Tentative assignment of some bands in the Raman spectrum used in the interpretation of protein structure. Wavenumber region a(± 2 cm") ' Tentative Assignment , 500-550 S-S stretching of cystine 534, 571 Tip, Aliphatic side chains 620-745 C-S stretch (Met, Cys), Amide IV, Amide V 755 Aliphatic side chains 760 Trp 830, 850 CC ring stretch, and C H 2 residue rock (Tyr) 900 CC residue stretch 938-940 CC residue stretch, C H 3 symmetric stretch; a-helix 1004 CC ring stretch (Phe) 1033 CC ring bend (Phe) 1057 C H 2 twist (Lys, Arg); Ch, CC stretch 1080-1100 C N , CC skeletal stretch (polypeptide backbone) 1127 Isopropyl antisymmetric stretch; C N stretch (backbone) 1150-1180 CH3 antisymmetric rock (aliphatic), C H rock aromatic 1210 Tyr, Phe 1238-1245 Amide III (^-structures) 1250 Amide III (random coil) 1260-1300 Amide III (a-helix, globular fraction) 1295-1304 Amide III (a-helix, tail or fibrous region) 1340 C H bend (residues) 1360 Indole ring (Trp) 1400 COO" symmetric stretch (Asp, Glu) 1425 (sh) Residue vibration (Asp, Glu, Lys) 1454, 1475 C H 3 (antisymmetric), C H 2 , C H bend 1555 Amide II, COO" antisymmetric stretch (Asp), Trp 1609, 1620 (sh) Phe, Tyr ring vibrations 1645-1657 Amide I C=0 stretch, N - H wag (a-helix) 1665-1680 Amide I C=0 stretch, N - H wag (anti-parallel p-sheet) 2878 (sh), 2930-50, 2979 (sh) C H stretch, aliphatic 3074 C H stretch, aromatic 3200 N H stretch (backbone); O H stretch (H 2 0) a sh= shoulder; precision of ± 2 cm"1 based on measured reproducibility of the potassium nitrate standard band at 1050 cm"1. Adapted from: Li-Chan (1996); Careche and Li-Chan (1997). 32 2.4.5.2. Raman spectroscopy of fish muscle proteins The application of Raman spectroscopy to study fish muscle proteins has been widely documented. For example, Careche and Li-Chan (1997) used Raman spectroscopy to study structural changes in cod myosin after modification with formaldehyde or frozen storage, and found that formaldehyde decreased the a-helix content from 95% to 60%. In addition, changes in the vibrational modes of aliphatic residues suggested involvement of hydrophobic interactions after the addition of formaldehyde or frozen storage. Bouraoui et al. (1997) investigated changes in the protein structure of surimi and gels from Pacific whiting. The intensity of 530 cm - 1 band increased considerably after the setting and cooking treatments, indicating the possible involvement of disulfide bonds and/or hydrophobic interactions in gel formation. The amide I region showed a predominantly a-helical structure in raw Pacific whiting surimi that changed to higher anti-parallel [3-sheet and lower a-helix contents in the gels. Ogawa et al. (1999) used Raman spectroscopy to study changes in lemon sole, ling cod, and rock fish actomyosin gels during setting. Conformational changes were observed by a slow unfolding of a-helix and exposure of hydrophobic amino acid residues forming hydrophobic interactions among actomyosin molecules. Sultanbawa and Li-Chan (2001) showed a decrease in the O H stretching region of ling cod N A M and surimi without cryoprotectants after frozen storage, and suggested that the formation of ice crystals could have resulted in dehydration of the protein and aggregation of the myosin heads. More recently, Herrero et al. (2004) studied structural changes in hake muscle proteins during frozen storage. Raman spectroscopy revealed an increase of (3-sheet at the expense of a-helix structure. In addition, the C H stretching band near 2935 cm - 1 increased in intensity, indicating 33 denaturation of the muscle proteins through exposure of aliphatic hydrophobic groups to the solvent. 2.5. Antifreeze proteins Antifreeze proteins (AFP) were discovered in 1957 by Scholander and co-workers when certain fish were identified as having the ability to survive at -1.9°C in sea water. It seemed that the presence of certain macrosolutes could keep the fish from freezing by lowering the freezing temperature below the ocean-freezing temperature (freezing point depression). This cryoprotective effect was achieved without affecting the osmotic pressure that would be caused by colligatively acting substances like salt or sugar, where the degree of freezing-point depression is dependent on the physical properties and concentration of the cryoprotectant (Payne et al., 1994). These newly-discovered macrosolutes were termed "antifreezes". In 1969, DeVries and coworkers reported that the antifreeze in the blood of nototheniid fish was a proteinaceous macromolecule soluble in 10% trichloroacetic acid (DeVries, 1971). By further characterization, it was revealed that this antifreeze comprised a set of glycoproteins that are each made of a repeating tripeptide (Ala-Ala-Thr)n with a disaccharide moiety attached to threonyl residues. Later on, studies revealed that arctic fish also produced antifreezes; however, these were non-glycosylated antifreeze peptides of quite different structure, differing in their amino acid composition and secondary structure (Duman and DeVries, 1976). Antifreeze glycoprotein (AFGP) and the non-glycoprotein antifreeze have a non-colligative mechanism of operation termed adsorption inhibition. Both types of proteins 34 appear to lower the freezing temperature without lowering the melting temperature; the difference between those two temperatures is called thermal hysteresis (Slaughter and Hew, 1981). Complete freeze protection for marine teleosts requires circulating antifreeze concentrations in the range of 10-20 mg/ml plasma. These proteins also seem to alter ice crystal formation. Davies and Hew (1990) reported that in the absence of AFP, ice grows most rapidly along the a-axes to give hexagonal-shaped crystal. It is this growth that is markedly inhibited by AFP as their presence accelerates ice crystal growth along the c-axis, yielding bipyramidal, needle-like ice spicules instead of large ice crystals (DeVries, 1988; Davies and Hew, 1990). 2.5.1. Antifreeze glycoprotein (AFGP) The most studied glycoproteins are from Antarctic fishes, Trematomas borgrevinki and Dissostichus mawsoni, and the northern fish, Boreogadus saida. The chain has a repeating tripeptide unit, (Ala-Ala-Thr) n, to which is linked the disacchari.de N-acetyl-D-galactosamine (al—»3)-galactose through the hydroxyl oxygen of the threonyl residue (DeVries, 1971). Chemical modification of the sugar residues results in the loss of antifreeze activity (Lillford and Holt, 1994). The AFGP has been numbered from 1 to 8, according to increasing electrophoretic mobility and decreasing size. The most plausible model for their structure is one in which the the polypeptide backbone forms a poly-proline II-like left-handed helical structure with three residues per turn (Bush et al., 1984; Lillford and Holt, 1994). This conformation creates a "comb-polymer" with the disaccharide units (hydrophilic) on one side of the helix. It is assumed 35 that these proteins are responsible for binding to the growing ice faces, with the opposite hydrophobic face (methyl groups) further disrupting the ice growth by accretion. 2.5.1.1. Antifreeze glycoprotein fractions 1-5 (AFGP 1-5) have molecular weights in the 30,000-10,000 Da range. The chain has a repeating tripeptide unit, Ala-Ala-Thr, with threonine glycosidically linked to the disaccharide moiety. Two additional alanines are at the C-terminal end (Feeney and Yeh, 1993). 2.5.1.2. Antifreeze glycoprotein fractions 6-8 (AFGP 6-8) are the most prevalent, with molecular weights in the 5,000-2,600 Da range, and have the first alanine in some of the repeats replaced by proline on the C- terminal side of threonine. In Arctic and North Atlantic cods, these glycoproteins also have threonine occasionally replaced by an arginyl residue (Feeney and Yeh, 1993; Fletcher et al., 1982). Their activities are lower than AFGP 1-5, requiring more than 20 times as much protein on a molar basis to achieve the same lowering of freezing temperature (Feeney and Yeh, 1993). 2.5.2. Non-glycosylated antifreeze proteins (AFP) Arctic fish also produce non-glycosylated AFP differing in amino acid composition and secondary structure (Duman and De Vries, 1976). Some of these non-glycosylated AFP are similar to AFGP in having two-thirds of their amino acids as alanine; other amino acids, including charged and hydrophilic ones, make up the other third (Feeney and Yeh, 1993; Davies and Sykes, 1997). These AFP have been further classified into four sub-types as described below. 36 2.5.2.1. Type I AFP Type I AFP (AFPI) of righteye flounders (Pleunectus americanus) and shorthorn sculpin (Myoxocephalus scorpius) is the most extensively characterized AFP. The proteins are heterogenous with at least seven active fractions identified, the major components having an apparent molecular weight of 3,600 Da. Circular dichroism (CD) measurements suggest that it is an oc-helical structure at -1°C (Davies and Hew, 1990; Lillford and Holt, 1994), with the exception of the last unit which adopts a 3io-helix conformation as shown by x-ray crystallography of the winter flounder isoform, HPLC6 (Sicherl and Yang, 1995). The main difference between the flounder and sculpin subtypes is that the former is built up of a clearly defined repeat of 11 amino acids, with -[Thr-(X)2 - N - (X) 7]- as the general sequence (Lillford and Holt, 1994), where X is principally alanine, and N is a polar amino acid (Asp, Glu, Lys, Ser or Thr). The latter is non-repetitive and more amphipathic, with several lysine and arginine side chains projecting from the same face of the helix (Davies and Sykes, 1997). The two subtypes also seem to bind to different planes of ice. The helicity can be attributed to the high alanine content (>60%), which can stabilize the N - and C-terminal cap structures (Lillford and Holt, 1994). The effect of this repeating sequence generates a helical structure with amphiphilic characteristics, where non-polar side chains are on one side of the helix (Davies and Hew, 1990, Laursen et al., 1994). Lillford and Holt (1994) further explain that the structure determined by X-ray crystallography shows a salt bridge between Lys and Glu, which is presumed to contribute to stabilizing the a-helix. 37 2.5.2.2. Type II AFP These AFP are found in sea raven (Hemitripterus americanus), smelt (Osmerus merdox) and herring (Clupea harengus). They contain alanine residues (14.4%) and are also rich in cysteine (or half-cystine). In addition, the sea raven AFP contains Trp, Phe and Tyr (aromatic) residues. Type II AFP are 14,000-24,000 Da homologs of the carbohydrate-recognition domain of C a 2 + dependent lectins. The 14,000-16,000 Da fraction has no a-helical structure but instead, large amounts of (3-sheet and P-turn (Lillford and Holt, 1994). It is believed that these molecules are also amphiphilic with the polar residues on one side of a planar rather than helical configuration. Type II AFP are further classified based on their calcium requirement for activity. The herring and smelt type II AFP seem to be completely dependent on calcium for activity. Calcium is directly involved in binding to ice, in the same way that it mediates sugar binding to C-type lectins. Substitution with other divalent metal ions will decrease the activity of the herring Type II AFP and it will also lead to alterations in ice crystal morphology (Chao et al, 1995). Calcium appears to stabilize the AFP against proteolysis and alter its fluorescence properties (Davies and Sykes, 1997). The sea raven type II AFP is C a 2 + independent and lacks calcium-liganding amino acids. It seems to be more active than its herring/smelt counterparts and produces significantly different ice crystal morphology (Chao et al, 1995). 2.5.2.3. Type III AFP These proteins are commonly found in eel pout (Macrozoanes americanus) of the Zoarcidae family (Mishra and Pattnaik, 1999). Type III AFP is comprised of 61-66 amino 38 acids with no particular bias, except for the absence of alanine and cysteine residues, and molecular weights varying from 6,500 to 7,000 Da (Brown and Sonnichsen, 2002). The secondary structure has two triple-stranded |3-sheets and single helical turn. However, the protein is known to have amphiphilic structures with high binding affinity to the ice-lattice (Lillford and Holt, 1994; Mishra and Pattnaik, 1999; Brown and Sonnichsen, 2002). 2.5.2.4. Type IV AFP Type IV AFP was first isolated from longhorn sculpin (Myoxocephalus octodecimspinosis) by Deng and co-workers in 1997. This newest addition to the AFP family is characterized by being highly a-helical, and has a molecular weight of 12,299 Da and 108 amino acids. Type IV AFP seems to be completely unrelated to the other AFP as it is rich in glutamine and glutamate (26%) and contains only 10% alanine, in contrast with 60% Ala in Type I AFP. Moreover, AFP IV shares 22% sequence identity with the low density lipoprotein receptor-binding domain of human apolipoprotein E3. According to Deng et al. (1997), the four main helices found in AFP IV are amphipathic and can fold into an antiparallel bundle, with the hydrophobic faces buried on the inside of the bundle and the polar-hydrophilic sides facing solvent water. 2.5.3. Mechanism of antifreeze action The general "antifreeze" mechanism appears to involve binding of polymers to small ice crystals resulting in a significant reduction in growth rates. This is one reason why antifreeze proteins are best described as a special class of crystal growth inhibitors 39 (Lilliford and Holt, 1994). On a molar basis they are estimated to be 200-500 times more effective than colligatively functioning compounds (Feeney and Yeh, 1993). The action at the surface of ice crystals in contact with a solution of AFP leads to what is known as thermal hysteresis- lowering of the non-equilibrium freezing point below the melting point. Even in the frozen state, AFP can inhibit the growth of large ice crystals, favoring the formation of small ice crystals instead, in particular when ice approaches the melting temperature and becomes more fluid (recrystallization inhibition). The diversity and complexity in structures for the different types of AFP has made it difficult to establish a specific common mechanism for the AFP ice-binding affinity and specificity. Type I AFP is considered to have the simplest and best characterized structure, and the majority of the antifreeze activity studies have been focused on this protein. The following important features have been deduced from this type of AFP: one, short peptides with high helicity were not active; two, deletion of the first five amino acids had virtually no effect on activity; three, the effects of active peptides on the relative rates of growth of ice crystal axes (a + c) were different (Davies and Hew, 1990; Lillford and Holt, 1994). A single mechanism of action has not been completely established. There are two proposed mechanism models; the first one states hydrogen bonding to ice water molecules as the most important contribution; the second one suggests that en tropic and enthalpic contributions from hydrophobic residues may be important (Davies and Sykes, 1997). Figure 6 shows the first of the proposed mechanisms by which AFP inhibit ice crystal growth, in which inhibition of ice crystal growth occurs upon AFP adsorption. 40 AFP solution Figure 6. Inhibition of ice crystal growth following AFP adsorption. (A) AFP (open circles) in solution and in contact with the ice front (hatched line) at 0°C. (B) The ice-water interface approaching the non-equilibrium freezing point, where the curvature between bound AFP leads to ice growth inhibition. The dotted line represents the overgrowth of a bound AFP. (C) Binding of an AFP from solution to the point of overgrowth in (B). Here the dotted line represents subsequent stabilization of ice-water interface around the newly bound AFP. Adapted from: Knight and Devries (1994) and Fletcher et al. (2001). 41 A study performed by Wierzbicki et al. (1996) attempted to explain and compare the binding mechanisms of winter flounder and shorthorn sculpin type I AFP to different ice planes. The authors state that the ice binding mechanism seems to be based upon protein-crystal surface enantioselective recognition that uses both the a-helical protein backbone matching to the surface topography, and the matching of the side chains of polar/charged residues with specific water molecule positions in the ice surface. Several studies have been performed in an attempt to explain and determine the structural conformation and ice binding mechanisms of AFPs. A n early study by Tomimatsu and Scherer (1976) used Raman spectroscopy to show differences in conformation between the solid, 5% aqueous solution and frozen solution states of high and low molecular weight AFGPs. Differences in conformation and environment of the carbohydrate hydroxyl groups were suggested to be related to the difference in antifreeze activity of high and low molecular weight AFGPs. Based on the structure characterization of type I AFP and AFGP, Davies and Hew (1990) proposed that an amphipathic structure might be required to bind to ice on the hydrophilic side and exclude water on the hydrophobic side. Sicherl and Yang (1995) showed the X-ray crystal diffraction data of a lone a-helical AFP from winter flounder at 4 and at -180°C, and proposed that the flatness of the AFP's ice-binding surface and rigidity of the side chains are critical to the AFP's ice-binding mechanism. However, other studies later indicated that the threonine and asparagine side chains were free to rotate; therefore a unique ice-binding structure in solution was not apparent. In addition, the importance of entropic effects and van der Waals interactions from hydrophobic residues was designated as having a significant role in binding AFP to ice, where some 42 parts of type I AFP remain rigid whereas the polar side chains are quite flexible (Gronwald et al., 1996; Chao et al., 1997). More recent studies also suggest high flexibility for AFGPs, which have a tendency to adopt a three-dimensional fold in the presence of ice (Tsvetkova et al., 2002), whereas the type I AFP from winter flounder has been found to form amyloid-type fibrils at pH 4 and 7 upon freezing and thawing (Graether et al., 2003). 2.5.4. Applications in food science In nature, AFP are present in a wide variety of organisms normally consumed as part of the human diet. Examples include winter flounder, Atlantic cod, blue mussels, carrots, cabbage and Brussel's sprouts (Griffith and Ewart, 1995). Average available consumption of fish AFP in the diet is calculated to be around 1-10 mg/day in the U S A and 50-500 mg/day in Iceland (Crevel et al., 2002). The potential allegenicity of ocean pout (Macrozoacres americanus) AFP has been investigated using the FAO/WHO 2001 decision tree for novel foods. The results indicated that this AFP showed no sequence similarity to known allergens nor was it stable to proteolytic degradation using standardized methods, In addition, AFP tested negative in skin prick tests and IgE-binding in vitro compared to sera from 20 patients with well-documented clinical history of fish allergy (Bindslev-Jensen et al., 2003). Antifreeze proteins have potential to be used in frozen foods as ice crystal regulators due to their ability to depress the freezing point, to inhibit .recrystallization during freezing and thawing, and ability to neutralize the effect of ice nucleators (Mishra and Pattnaik, 1999). The discoveries on prevention of ice crystal formation and 43 recrystallization by antifreeze proteins have led to speculation about their possible application in a variety of food systems (Feeney and Yeh, 1998), as described below. 2.5.4.1. Recrystallization inhibition Recrystallization causes larger ice crystals to grow at the expense of smaller ice crystals, resulting in a reduction of ice crystal number and an increase in the average ice crystal diameter. Large ice crystals can damage membranes and other structures within muscle samples, causing protein denaturation and a reduced water-holding capacity (Payne et a l , 1994). Both AFGP and AFPI can inhibit recrystallization during frozen storage. It has been reported that 0.5 pg/mL of AFGP can give extensive inhibition of recrystallization (Feeney and Yeh, 1993). A U.S. patent by Warren et al. (1992) indicated that 0.1 mg/mL of AFPI could effectively inhibit recrystallization in popsicle and frozen root beer float bar products. In this patent, AFPI was added to the samples, which were then frozen to approximately -80°C and then stored at temperatures between -6° to -8°C. After 1 hr, the control sample had noticeably larger ice crystals, whereas the samples with AFPI did not show any significant crystal growth. 2.5.4.2. Modification of ice crystal size Payne et al. (1994) showed that low concentrations of antifreeze proteins (AFGP and type I AFP) were capable of modifying the size of ice crystals. The AFP were effective at concentrations ranging from 0.1 mg/mL to 1 mg/mL. In the experiment, bovine muscle (Sternomandibularis) samples (3x3x6 mm) were soaked in phosphate 44 buffer saline (pH 7.3) with and without AFGP (0.1 mg/mL) for 4.5 to 6 h at 15°C. Samples were then frozen (-20°C) or chilled (2°C) for 3 days. Upon examination under light-microscopy, the frozen samples with AFP had many small intracellular spaces, which the authors assumed represented ice crystals. The frozen controls had much larger intracellular single spaces. In the same study (Payne et al., 1994), a range of samples (3x3x6 mm) of post rigor ovine muscle (Peroneus longus) were prepared by soaking in 0.9% saline (w/v) containing AFPI or AFGP at various concentrations up to 1 mg/mL. Samples were soaked at 2°C for various times up to 7 days and then held chilled (2°C) or frozen (-20°C) for 5 or 7 days. The samples were then examined by scanning electron microscopy. The observations indicated that with frozen samples, antifreeze proteins reduced the size of ice crystals compared to the large ice crystals formed in the control. The effect depended on the AFP concentration used as well as the period of soaking in the saline solution. These results indicated that a certain level of antifreeze protein must be achieved within the ovine meat sample before ice crystal growth could be inhibited. In both experiments, antifreeze proteins had no effect on chilled (2°C) samples. 2.5.4.3. Effect of AFP on Ca2+ATPase activity of actomyosin Boonsupthip and Lee (2003) studied the effect of Type III antifreeze protein for preservation of gel-forming properties of fish actomyosin under frozen and chilled conditions. The activity of C a 2 + ATPase was used as an indicator of the gel-forming capacity of actomyosin. The AFPIII was used at low (0.05-0.3 g/L) and high (10-100 g/L) concentrations in the actomyosin extract. Sucrose and sorbitol (4% each, w/w) were 45 used for comparison. At 50 and 100 g/L, AFPIII was able to retain higher levels of Ca ATPase activity in the actomyosin compared to the sucrose-sorbitol blend and control. It is important to note that at both low and high AFPIII concentrations, the initial (unfrozen) ATPase activity was twice as high as the control and sucrose-sorbitol blend, making it difficult to assess i f there was a significant influence of AFPIII under frozen conditions. 2.5.5. Cryosurgery and cryopreservation Cryosurgery is a surgical technique that uses cryogenically cooled probes to locally freeze and thereby destroy undesirable tissues (Koushafar et al., 1997). Several papers on cryosurgery have shown that, at concentrations higher than 5 mg/mL, type-I AFP can enhance cellular destruction in cell suspensions of tissues in vitro and in vivo (Koushafar and Rubinsky, 1997; Koushafar et al., 1997; Pham et al., 1999). Conclusions from these studies suggested that introducing antifreeze proteins into tissues prior to freezing them could increase the efficacy and control of tissue destruction by cryosurgery. The in vivo study further showed that 10 mg/mL of type-I AFP injected into mice with subcutaneous prostate tumors, induced a damaging effect attributed to the mechanical interaction, between spicular ice crystals and the cells. However, Larese et al. (1996) suggested that the addition of AFPI type I during cryopreservation may increase the occurrence of "intracellular freezing" at subzero temperatures, which resulted in reduced recovery of cryopreserved cells. Based on these studies and cryosurgery applications, it seems that AFP have dual action depending on factors including the cryopreservation protocol being used, storage temperature, the type and concentration of AFP and the physiological features of the 46 substrate being used. Wang (2000) summarized the properties of antifreeze proteins by indicating that AFP have a "two-fold" interaction with plasma membranes either by -protecting the membrane or causing cytotoxicity. In addition, the interaction of AFP with ice also exhibits a dual behaviour which includes inhibition of recrystallization or inception of ice nucleation. This dual action of AFP at different temperature conditions is summarized as follows, where cooling temperatures can impart protection of cell membrane or toxicity. During freezing, AFP can inhibit ice growth or promote ice nucleation, and finally during thawing AFP can cause inhibition of recrystallization or ice nucleation (Wang, 2000). 2.6. Commercial cryoprotectants Cryoprotectants are compounds that extend the shelf life of frozen foods. The word cryoprotectant is a broad term that includes all compounds that aid in preventing undesirable changes induced in foods or food ingredients by freezing, frozen storage, or thawing. Cryoprotectants may be added either during processing or formulation (Macdonald et al., 2000). Many different compounds can be used to cryoprotect labile proteins during freeze-fhawing, including sugars, amino acids, polyols, methyl amines, carbohydrate polymers, synthetic polymers (e.g. polyethylene glycol, PEG), and some inorganic salts (e.g. potassium phosphate and ammonium sulfate). For most proteins, the cryoprotectant must be at relatively high concentration to confer maximum protection (Carpenter and Crowe, 1988; Macdonald et al., 2000). Typically, low-molecular weight carbohydrates or polyols like sucrose and sorbitol seem to exert their protective effects by a solute exclusion principle, in which the addition of sugars to an aqueous solution of 47 protein results in an unfavorable free-energy change that leads to the stabilizing solute being excluded from the surface and the protein being preferably hydrated (Arakawa and Timasheff, 1982; MacDonald et al., 2000). Another protective effect is thought to involve binding of the cryoprotective molecules with protein molecules at one of the functional groups either by ionic or hydrogen bonds, thus allowing each protein molecule to be coated with the cryoprotectant (Matsumoto, 1979). In addition, by depressing the freezing point with low-molecular weight solutes, the deleterious effects of freeze concentration on proteins at subfreezing storage temperature should be lessened (MacDonald et al., 2000). Sucrose and sorbitol, typically mixed in a 1:1 ratio and added at 8% w/w, are the principal cryoprotectants used commercially to make surimi and formulated seafood products because of low cost, availability, and low tendency to cause Maillard browning (MacDonald and Lanier, 1994). Sugars and sugar alcohols when added to myosin filament suspensions or myofibrils form hydrogen bonds via their hydroxyl groups.with some amino acid residues, including charged amino acids of myosin rods. Sorbitol is known to promote solubilization of myofibrillar proteins at physiological concentrations of NaCl (Mathew and Prakash, 2005). Unfortunately, the rather high level of sucrose and sorbitol required for a cryoprotective effect imparts a sweet taste to the final product. Blends consisting of lower concentrations of sucrose and sorbitol with polyols such as lactitol and Litesse have been proposed as alternatives (Sultanbawa and Li-Chan, 1998) due to their effectiveness for cryoprotection, decreased sweetness and cost. Trehalose (a-D-glucopyranosyl-l,l-a-D-glucopyranoside) is a non-reducing glucose disaccharide naturally found in living organisms such as yeast, as well as some 48 foods such as honey, mushrooms, lobster and shrimp (Osako et al., 2005; Xie and Timasheff 1997; Schiraldi et al., 2002). Trehalose is described to exhibit similar properties as sucrose, and to act as a stabilizer of structure and function of several macromolecules (Sola-Penna and Meyer-Fernandes, 1998). Some theories on the mechanism of action by trehalose are based on the interaction of the disaccharide with biological structures, and on the interaction of trehalose with water and the thermo physical properties of its aqueous solutions (Osako et al., 2005; MacDonald et al., 2000). Trehalose is 40-45% less sweet than sucrose (Jittinandana et al., 2005). Trehalose is hydrolysed by the enzyme trehalase in the human small intestine into two glucose molecules which are absorbed and metabolized. In 2000, trehalose received generally recognized as safe (GRAS) status in the U.S (Pszczola, 2003). In 2005, the Canadian government approved trehalose for use as a nutritive sweetener in human food applications (Health Canada, 2006). Certain high-molecular weight polymers such as polyvinylpyrrolidone, polyethylene glycol, dextran, and maltodexrtrins are good cryoprotective agents because they are sterically excluded from the protein surface by their size, increasing the viscosity (Carpenter and Crowe, 1988). In addition, they raise the glass transition temperature, ensuring that the hydrated system exists as an immobilized, stable glass at conventional cold storage temperatures. With continued lowering of the temperature below the freezing point, the rates of diffusion-limited deteriorative reactions decrease steadily as ice freezes out and solution viscosity increases, lowering diffusivity of the dissolved reactants (MacDonald and Lanier, 1991; MacDonald et al., 2000). Figure 7 shows a schematic state diagram for a binary system in which the solute component does not 49 crystallize. At high concentrations of solute, the glass-transition temperature (Tg) occurs at temperatures above the freezing point. Thus, the mixture cools to form a "candy" glass directly from the liquid state. At solute concentrations below the point Tg' , when the temperature falls below the freezing curve, the solution will exist either as a viscous supersaturated solution in the liquid state or more commonly as a mixture of ice crystals and supersaturated solution. Under these conditions, the system is termed a "rubber", exhibiting high viscosity because of the presence of ice crystals or strong intermolecular solute interactions/entanglements. Protection of proteins is effected when the added solute raises the Tg to a temperature above the storage temperature, thereby ensuring that the system is in a glass state (Mackie, 1993; MacDonald et al., 2000). Polyphosphates at levels ranging from 0.2-0.3% are also commonly added as synergists to the cryoprotective effect of the carbohydrate additives (Sych et al., 1990). Phosphates are added as a mixture of sodium tripolyphosphate or tetrasodium pyrophosphate (Park, 2000). The mechanism of cryoprotection by phosphates is not fully understood. Their function includes acting as metal chelators and/or antioxidants. In addition, water holding capacity and salt solubilization of myofibrillar proteins are known to improve when phosphates are used (Park 2000). 50 [Liquid Solution] J m [Ice + Concentrated Solution] / T g * [Ice + Glass] Weight % Solute Figure 7. Phase diagram of a binary system in which the solute forms a glass at low temperatures. Adapted from: MacDonald et al. (2000). 51 3. MATERIALS AND METHODS 3.1. Fish fillets: Soaking or spraying with cryoprotectants Ling cod fillets Cut into equal pieces (50 g each) Soaked with phosphate buffer (POB) or 4% sucrose, 4% sorbitol and POB unfrozen Sprayed with phosphate buffer (POB) or 4% sucrose, 4% sorbitol and POB Frozen -20 ±2°C and freeze-thawed unfrozen Analyzed for: water binding capacity (WBC) percent drip loss (% DL) Figure 8. Overview of working plan for fish fillets soaked or sprayed with cryoprotectants. The objective of this experiment was to establish a spraying technique and soaking time that could later be adopted as a method of incorporation of antifreeze protein solutions into muscle tissue systems. This experiment was carried out using phosphate buffer and a 4% sucrose, 4% sorbitol blend. AFP were not included in this experiment plan due to the limited supply of AFP and the amounts that would be needed to make enough solution for spraying and or soaking the muscle pieces. 52 Ling cod (Ophiodon elongatus) fish were purchased, within four days after capture, from Albion Fisheries Ltd. (Vancouver, BC, Canada) in July 2002. De-boned and skinned fillets weighing ca. 500 g each were immediately delivered on ice to the laboratory. Upon arrival to the lab, the centre part of each fillet was cut into 8 pieces (ca. 3x3.5x8 cm; each piece weighing approximately 50 g) as shown in Figure 9. Figure 9. Section of fish fillet (shaded area) used for cutting 8 individual pieces of approximately 50 g each. Each piece (ca. 3x3.5x8 cm) weighed approximately 50 g. The fillet pieces were then mixed together and randomly selected for each treatment. The different treatments consisted of soaking or spraying with 0.01 M phosphate buffer (pH 6.8), and soaking or spraying with a solution of phosphate buffer and 4% (w/v) sucrose (Rogers Sugar Ltd., Vancouver, BC) and 4% (w/v) D-sorbitol (Sigma Chemical Co., St. Louis, MO). Spraying was carried out using a commercial sprayer (Spraying Systems Co.) and consisted of spraying each fillet piece 5 times on one side, turning the fillet piece over and spraying 5 times more. Spraying was conducted at 4°C over a stainless steel rack. The soaking procedure was carried out in 5.8 L capacity polyethylene containers. The solution containing phosphate buffer or commercial blend was poured into the container, samples were placed inside, making sure that the solution completely covered them, and allowed to soak for 30 min at 4°C. Treated samples were placed in 0.1 mm thick 6x12" polyethylene bags (Fisherbrand, Fisher Sci., NJ) and 53 frozen at -20 ± 2°C before being submitted to 8 freeze-thaw cycles (Sultanbawa and L i -Chan, 2001). One freeze-thaw cycle consisted of freezing at -20 ± 2°C for 18 hrs and thawing at 4°C for 6 hrs (Appendix V). One sample set was not treated with soaking or spraying solutions and used as control. This temperature abuse is perhaps not representative of conditions likely encountered during normal storage of fish products but may rapidly induce changes that could normally occur during a long period of storage and transpostation (Kim et al., 1986). A separate study was performed in which fish fillets were soaked and sprayed in the same manner as described above. However, in this case, 2-3 drops of blue food colour (Club House, London, ON) were added to the solutions prior to treatment. After soaking or spraying, the samples were transversely cut in order to evaluate the extent of penetration by the solutions. At this point, digital photographs were taken using an Olympus digital camera (Olympus Optical Co., Ltd, Tokyo, Japan). 3.1.1. Analyses After submitting samples to 8 freeze-thaw cycles, samples were kept at 4°C and analyzed for water binding capacity and percent thaw drip. The samples that were not previously frozen (unfrozen) were analyzed on the first day of the experiments and used as comparison for the effect of freeze-thawing on the fish fillets with or without cryoprotectants. 54 3.1.1.1. Water binding capacity and percent drip loss Water binding capacity (WBC) was determined according to the method by Porteous and Wood (1983). Duplicate samples (2.5 g) from each treatment were placed in 50 mL capacity Nalgene™ centrifuge tubes (Nalgene Co., N Y ) to which 5 mL of distilled water were added. The samples were stirred using a glass stirring rod, the mixture was then allowed to equilibrate for 24 hrs at 4 ± 2°C before centrifuging at 1825 x g for 10 min at 4°C using a Sorvall centrifuge, model RC 5B plus (Sorvall Instruments Dupont, CT). The supernatant was decanted and the pellet weighed. W B C was expressed as the gain in weight of the pellet divided by the original weight of the sample. For percent of thaw drip, the weight of each fillet piece was recorded before and after treatment by spraying or soaking prior to freezing. Freeze-thawed samples were removed from their bags and placed on paper towels where they were blot dried and, weighed. Percent drip loss (% DL) was expressed as the percent of sample weight loss following eight freeze-thaw cycles. The analysis was performed in duplicate. 55 3.2. Fish mince treated with antifreeze proteins compared to commercial cryoprotectants Ling cod fillets Minced Mixed with AFGP or AFPI (0.005, 0.01, 0.05, 0.5 mg/g) or commercial blend (4% sucrose, 4% sorbitol and 0.3% phosphates, PO) Unfrozen (UF) 1 Frozen -20 ±2°C and freeze-thawed Analyzed for: moisture, crude protein, expressible moisture, cook loss, texture, salt extractable proteins, SDS-PAGE profiles, T-SH, SS content, hydrophobicity, Raman spectral profiles Figure 10. Overview of working plan for fish mince treated with antifreeze proteins or commercial cryoprotectants. The objective of this study was to evaluate the effects of 'high' (0.05 and 0.5 mg/g = 0.005% and 0.05% w/w) and 'low' (0.005 and 0.01 mg/g = 0.0005% and 0.001% w/w) AFP concentrations on fish mince, and to compare these effects to treatments of mince with commercial cryoprotectants. Concentrations were chosen based on literature values (see literature review) where AFP have shown to inhibit recrystallization during frozen storage of chilled and frozen meat (Payne et al., 1994). 56 Commercially available low molecular weight antifreeze glycoprotein, fraction 6-8 (AFGP) from rock or Greenland (rock) cod (Gadus ogac) and Type I antifreeze protein (AFPI) from winter flounder (Pleuronectes americanus), were kindly donated by Dr. G. Fletcher (A/F Protein Canada Inc., St. John's NF, Canada). The purity of the proteins was > 80% (A/F Protein Canada Inc). The purification process used is described in Wu et al. (2005). For the experiments involving 'low' AFP concentrations, ling cod fish (ca. 6.5 kg) were purchased in January 2003 and for the experiments involving 'high' AFP concentrations, ling cod fish were purchased in April 2003. In both instances, ling cod were purchased within four days after capture, from Albion Fisheries Ltd. (Vancouver, BC, Canada). The following procedures were conducted in a walk-in cold room at 4°C. Fish were deboned, skinned and filleted (4 kg). Fillets were minced using a B E E M -GIGANT Grinder Model TYPE F5-10 (BEEM California Corp., Glendale, CA) with a 4 mm screen. Fifty to 100 g portions were used for each treatment. Commercial blend (4% sucrose, 4% sorbitol w/w; 0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w), AFGP or AFPI, were added to their corresponding mince portion and mixed thoroughly for 3 min using a Kitchen-Aid bowl mixer set on low speed (setting of "1"). The minced samples were packed in 0.1 mm thick 6x12" polyethylene bags (Fisherbrand, 01-812-10C, Fisher Sci., NJ) and stored at -18 + 2°C. One set of unfrozen (not previously frozen) samples from each treatment was analyzed that same day and used as the unfrozen comparison. Frozen samples were submitted to 8 freeze-thaw cycles, as described in section 3.1. 57 3.2.1. Analyses Unfrozen and freeze-thawed samples from each experimental phase were analyzed for moisture, crude protein, expressible moisture, cook loss, texture changes, salt extractable proteins, SDS-PAGE profiles, total sulfhydryl groups, disulfide content, protein surface hydrophobicity and structural changes by Raman spectroscopy. 3.2.1.1. Moisture and crude protein Moisture and crude protein were determined according to the official methods of the A O A C (1998). For moisture, 5 g of sample were dried in a vacuum oven (<50 mm Hg) at 70°C for ca. 18 h, to a constant weight (AOAC 925.09). Crude protein was measured by the combustion method (AOAC 992.15) using a Leco nitrogen analyzer (Leco Corp, Joseph, MI) and 6.25 as the conversion factor. Moisture and crude protein were carried out in triplicate for all treatments. 3.2.1.2. Expressible moisture and cook loss Expressible moisture was determined in triplicate according to the method by Gomez-Guillen et al. (1996). Mince samples (1.5 g) were placed in a centrifuge tube containing a pre-weighed Gilson Pipetman pipet filter. The sample with filter was centrifuged at 4000 x g for 10 min at ambient temperature using a Sorvall centrifuge model RC 5B plus (Sorvall Instruments Dupont, CT). Expressible moisture (EM) was expressed as percent of weight loss in the sample and was measured by weighing the amount of water absorbed by the pipet filter after centrifugation. 58 Cook loss of samples was determined according to Honikel (1998) with some modifications. Triplicate samples (approximately 4 g) from each treatment were placed inside previously weighed plastic cylindrical molds (2.5 cm high and 1.3 cm in diameter). The molds were wrapped with all purpose plastic food wrap (AEP Canada Inc., West Hi l l ON), placed inside a polyethylene bag and cooked in a water bath at 75°C for 15 min. Molds were then inverted over paper towels and allowed to equilibrate at room temperature before weighing. Cook loss was expressed as a percentage of the initial sample weight. 3.2.1.3. Textural hardness Textural hardness was measured using a TA.TX2 texture analyzer (Texture Corp., Scarsdale, N Y / Stable Microsystems, Godalmin, Surrey, UK) , with a 3 mm-diameter stainless steel cylindrical plunger set to a 75 % sample penetration at a test speed of 0.5 mm/sec. Triplicate samples of raw, unfrozen mince were placed in plastic cylindrical molds as described in section 3.2.1.2, and allowed to equilibrate overnight at 4°C. The next day, samples were brought to room temperature (approximately 1 h) before being tested. Textural hardness (in g) was measured using the puncture test (Lian et al., 2000) where hardness was identified as the maximal force required for the first deformation (peak 1) after a 75% penetration. 3.2.1.4. Protein extractability Salt-soluble proteins were extracted in duplicate according to the procedure by Sultanbawa and Li-Chan (1998) and Lian et al. (2000), with some modifications. Fish 59 mince (5 g) in 20 mL of 0.6 M NaCl-0.1 M phosphate buffer (pH 7.5) was homogenized on ice for 1 min at 1800 rpm using a Sorvall Omni mixer (Ivan Sorvall Inc., Norwalk, CT). The homogenate was centrifuged at 10,000 x g for 10 min at 4°C using a Sorvall centrifuge, model RC 5B plus (Sorvall Instruments Dupont, CT). The supernatant was saved and the pellet was submitted to a second extraction as described above. The supernatants from each duplicate sample were combined. Protein concentration was determined in triplicate by the bicinchoninic acid (BCA) protein assay, according to the microplate procedure with bovine serum albumin (BSA) as the standard (Sigma-Aldrich Canada, Oakville, ON). The percent of extractable protein was expressed as extractable protein in the supernatant divided by the crude protein content of each sample. Extracts were frozen at -80°C for further analyses. 3.2.1.5. SDS-polyacrylamide gel electrophoresis (SDS-PAGE) Electrophoretic patterns of salt-extractable myofibrillar proteins before and after freeze-thawing were analyzed by SDS-PAGE (Laemmli, 1970). Supernatants from the salt-soluble extractions (section 3.2.1.4.) were placed in ice and sonicated for 1 min using a Tekmar sonic disruptor (Tekmar Co., Cincinnati, OH) with a microtip probe, a setting of 30, duty cycle of 50 and output control 5. Sonicated samples were diluted to a final concentration of 2 mg/mL using 0.6 M NaCl in 0.1 M phosphate pH 7.5 buffer. To each diluted protein sample (80 uL) were added 20 uL 10% SDS, 2 uL fi-mercaptoethanol (ME) and 1 uL bromophenol blue. Samples were boiled for 20 min and centrifuged at 16,249 x g for 5 min using a Mikro-20 benchtop centrifuge (Hettich Zentrifugen, Tuttlinger, Germany). Sample aliquots (1 uL) were loaded on 10-15% gradient Phastgels 60 along with standard molecular weight markers (Sigma Chemical Co., St. Louis, MO) ranging from 6.5-205 kDa. Electrophoresis was carried out in the PhastSystem™ (Pharmacia L K B Biotechnology, Uppsala, Sweden) at 70 A V h and 15°C with running time of 20 min. Gels were stained for 45 min using Coomassie PhastGel blue R staining solution, and destained in 30% methanol, 10% acetic acid (v/v) for 2 hr. The gels were then incubated for 45 min in a 5% glycerol, 10% acetic acid (v/v) preserving solution before being air dried overnight. 3.2.1.6. Total sulfhydryl (SH) groups and disulfide (SS) bonds Total sulfhydryl content was determined according to the method of Ellman (1959) as modified by Sultanbawa and Li-Chan (2001). Sample supernatants (section 3.2.1.4) were sonicated, as described in section 3.2.1.5 and diluted to a protein concentration of 4 mg/mL using 0.6 M NaCl in 0.1 M phosphate buffer (pH 7.5). To 0.4 mL of diluted protein was added 3.6 mL of 2.5 M guanidine thiocyanate in 85 m M Tris/100 m M glycine/4 m M EDTA buffer (pH 8.0). The 4 mL mixture was divided into four 1 mL aliquots and incubated at room temperature in a 25°C water bath for 30 min. This was followed by adding 25 uL of 0.4% (w/v) Ellman's reagent 5,5'-dithio-bis-2-nitrobenzoic acid, DTNB; Sigma Chemical Co., St. Louis , MO) to each of three of the 1 mL aliquots. The fourth 1 mL aliquot had 25 uL buffer instead of DTNB and was used as protein blank. A reagent blank was prepared with 0.9 mL buffer, 25 pL DTNB and 0.1 mL distilled water. Samples and blanks were incubated for 30 min. The absorbance at 412 nm was measured using an UV-visible spectrophotometer (Unicam, UV2, A T I _ U N I C A M , 61 Cambridge, UK) with a microcuvette of 1 cm path length (VWR Scientific Co., Mississauga, ON). The disulfide bond (SS) content was determined using 2-nitro-5 thiosulfobenzoate (NTSB) reagent according to the methods of Thannhauser et al (1984) and Damodaran (1985), as modified by Sultanbawa and Li-Chan (2001). The NTSB assay solution was prepared from a stock NTSB solution 100 times diluted with freshly made solution of 3 M guanidine thiocyanate, 50 m M glycine, 100 m M sodium sulfite and 3 m M EDTA (pH 9.5). To 200 uL of diluted sample (4 mg/mL), 1.5 mL of freshly prepared NTSB assay solution (pH 9.5) was added. The mixture was incubated in the dark at room temperature for 25 min. Absorbance at 412 nm was measured in triplicate. The reagent blank contained 1.5 mL NTSB assay solution with 200 uL distilled water. Content of SH and SS were calculated as described in Beveridge et al. (1974) using an extinction coefficient of 13600 M"'cnf 1 (Ellman, 1959) and expressed as moles per 105 g protein. 3.2.1.7. Hydrophobicity Protein surface hydrophobicity was determined using l-anilinonaphthalene-8-sulfonic acid (ANS) fluorescent probe according to the method by Careche and Li-Chan (1997). Sample supernatants (section 3.2.1.4.) were sonicated, as described in section 3.2.1.5. Sample supernatants were diluted to protein concentrations of 0.05, 0.10, 0.15, 0.20 and 0.225 mg/mL using 0.6 M NaCl in 0.1 M phosphate buffer (pH 7.5). To each diluted protein solution (4 mL) was added 20 uL of 8 m M ANS in 20 m M phosphate buffer containing 0.6 M NaCl and vortexed immediately. Samples were incubated in the dark for 15 min at room temperature before measuring relative fluorescence intensities 62 (PvFI) starting from the lowest to highest protein concentrations. The RFI was measured at an excitation wavelength of 390 nm, emission wavelength of 470 nm with excitation and emission slits set at 5 nm, using a RF-450 spectrofluorometer (Shimadzu Corp., Kyoto, Japan). The RFI of protein blanks (no ANS) and buffer reagent blank (no protein) were also measured and their values were subtracted from those of the corresponding protein dilutions with ANS. Possible day to day instrumental fluctuations were corrected by measuring the RFI for ANS (20 uL) in methanol (4 mL) and calculating a correction factor by normalizing to a standard value of 30. Surface hydrophobicity was expressed as the initial slope of the net RFI values plotted against protein concentration, calculated by linear regression analysis. Surface hydrophobicity was analyzed in duplicate for treatments with high AFP concentrations, and in triplicate for treatments with low AFP concentrations. 3.2.1.8. Raman spectroscopy Raman spectra were recorded for all samples at 4°C using a JASCO model NR-1100 laser Raman spectrophotometer (Japan Spectroscopic Co., Ltd., Tokyo, Japan) with 488 nm excitation from an argon ion laser cooled with the Coherent Laser Pure heat exchanger system (Coherent Innova 70C series, Coherent Laser group, Santa Clara, CA). Frequency calibration of the instrument was conducted daily using the band at 1050 ±2 cm"1 for 1 M potassium nitrate solution. Each sample was placed in a capillary tube and held at 4.0 ± 0.1 °C in a constant temperature measuring accessory (JASCO model RT-IC, Japan Spectroscopic Co., Ltd., Tokyo, Japan) for Raman measurements. The measurement conditions were: laser power of 200 mW, slit height 4 mm, sample speed 63 120 cm" min" ; data collection every cm" , with 6 scans per sample. The recorded spectra were analyzed using Grams/32® Spectral Notebase™ version 4.14 level II (Galactic Industries Corp., Salem, NH). Spectral data were baseline corrected using the multiple point baseline correction function and smoothed using the maximum entropy smoothing of the Razor analysis program with Gaussian peak shape. Finally, spectral data were also normalized to the intensity of the phenylalanine band at 1004 ± 2 cm"1. 64 3.3. Natural actomyosin with antifreeze proteins compared to commercial cryoprotectants N A M extract 900 mL AFGP (0.2 mg/L NAM)<-AFGP (5 mg/L NAM) -4~ AFPI (0.2 mg/L NAM)<-AFPI (5 mg/L NAM) <-Control (no additives) Commercial blend Unfrozen 50 mL N A M aliquots Freeze thaw 50 mL N A M aliquots Analyzed for: Ca 2 +-ATPase activity and surface hydrophobicity AFGP (0.2 mq/L NAM) AFGP (5 mg/L NAM) AFPI (0.2 mg/L NAM) AFPI (5 mg/L NAM) Control (no additives) Commercial blend Figure 11. Overview of working plan for natural actomyosin treated with antifreeze proteins or commercial cryoprotectants. The objective of this study was to determine the effect of fish antifreeze proteins on Ca-ATPase activity and surface hydrophobicity of ling cod natural actomyosin (NAM) systems following freeze thaw abuse. This study was designed in order to investigate the effects of AFGP and AFPI (0.2 and 5 mg/L N A M ) based on the results reported by 65 Boonsupthip and Lee (2003), who tested the Ca ATPase activity of frozen and chilled tilapia actomyosin containing AFPIII (0.05-0.3 g/L and 10-100 g/L). Natural actomyosin (NAM) was extracted in a walk-in cold room (4°C) according to the method by Ogawa et al. (1995) as modified by Sultanbawa and Li-Chan (2001) with slight modifications. Ling cod fish were purchased, within four days after capture, from Albion Fisheries Ltd. (Vancouver, BC, Canada) in October 2003. Fish were deboned, skinned, filleted and divided into 125 g portions. Each portion was homogenized in a Waring™ commercial blender (model 33BL79, Dynamics Corporation of America, Hartford, CT) for 1 min with 750 mL of cold 0.05 M NaCl in 0.02 M Tris-maleate buffer (pH 7.0). The homogenates were combined and centrifuged at 10,000 x g for 10 min at 4°C using a Sorvall centrifuge, model RC 5B plus (Sorvall Instruments Dupont, CT). The supernatant was decanted, the pellets combined and divided again into 125 g portions. Pellet portions were again homogenized in the blender for 1 min with 750 mL 0.6 M NaCl in 0.02 M Tris-maleate buffer (pH 7.0). The homogenates were combined and centrifuged at 10,000 x g for 10 min at 4°C. The supernatants were combined (approximately 4 L) and filtered through a double layer cotton gauze (grade # 40). The filtered solution was divided into two-2 L portions and each portion was diluted with 15 L of 0.02 M Tris-maleate buffer (pH 6.9) in a 20 L capacity polyethylene tank. Diluted samples were kept at 4°C overnight. The next day, the clear top layer from each tank was decanted. The bottom layer from each tank (total volume 3 L) was centrifuged at 10,000 x g for 10 min at 4°C. The pellets obtained were further concentrated by centrifuging at 15,000 x g for 20 min at 4°C. The resulting viscous pellet (900 mL) was 66 referred to as N A M extract. The protein concentration of the N A M extract was 82 mg/mL as determined by the B C A method. The N A M extract was divided into six 100 mL portions. Each portion was hand-mixed for 1 min using a glass stirring rod, with either the commercial blend, AFGP (fraction 6-8) or AFPI. One portion without additives was used as control. The N A M portions were then divided into approximately 50 mL aliquots, placed into 50 mL Falcon polypropylene tubes (Becton Dickinson and Co., Franklin Lakes, NJ), frozen at -20 + 2°C and submitted to 8 freeze-thaw cycles as described in section 3.1. Another set of six 100 mL portions of N A M extracts were analyzed without freezing and used as unfrozen comparisons. 3.3.1. Analyses N A M extracts (unfrozen and freeze-thawed) were prepared before each analysis as follows: 2.5 g of N A M extract were placed in 50 mL Falcon polypropylene tubes and brought up to a final volume of 25 mL with cold 0.6 M NaCl in 0.02 M Tris-maleate buffer (pH 7.0). Samples were placed in an ice bath and sonicated as described in section 3.2.1.5. Sonicated samples were analyzed for protein concentration, Ca 2 +-ATPase activity and surface hydrophobicity. 3.3.1.1. Protein concentration Protein concentration was determined by the bicinchoninic acid (BCA) protein assay according to the microplate procedure and using bovine serum albumin (BSA) (A-67 4503; Sigma-Aldrich, Oakville, ON, Canada) as the protein standard. Measurements were done in triplicate wells on a single microplate. 3.3.1.2. Calcium ATPase activity The measurement of calcium ATPase activity (Ca2 +-ATPase) was modified from Briskey and Fukazawa (1971). The protein concentration of the N A M extract was adjusted to 2 mg/mL using 0.6 M NaCl in 0.02 M Tris-maleate buffer (pH 7.0). One mL of the diluted extract was added to 8 mL of reaction medium prepared with 0.02 M Tris-maleate buffer (pH 7.0), 0.05 M CaCl 2 and 0.1' M K C l . To trigger the enzymatic reaction, 1 mL of 0.02 M ATP (from equine muscle, Sigma Chemical Co., St. Louis, MO) was added to the mixture. Triplicate 1 mL aliquots were removed at 0, 1,2 and 10 min and added to 1 mL chilled 20% (w/v) trichloroacetic acid (TCA), in disposable borosilicate glass tubes (13 x 100 mm Kimax 51, V W R Scientific, Mississauga, ON). The samples were kept cold on ice. The precipitated proteins were removed after centrifuging at 752 x g for 10 min at 4°C using a Beckman centrifuge (Beckman Instruments, Palo Alto, CA). The concentration of inorganic phosphate in the sample was measured by the method of Briskey and Fukazawa (1971) with modifications. After centrifugation, 1 mL of the supernatant was added to a test tube containing 0.5 mL of 2.5% (w/v) ammonium molybdate (NFD2M0O4 in 5 N sulfuric acid (analytical reagent, B D H Inc., Toronto, ON) and vortexed (Fisher Vortex Genie 2, Scientific Industries Inc., Bohemia NY) . Immediately after, 1 mL of 0.32% (w/v) stannous chloride SnCL; (Fisher Scientific, NJ) was added and again vortexed. After 30 min at room temperature, the developed colour was measured in triplicate by adding 250 uL of the above mix to each well of a 68 microplate. Absorbance was measured at 690 nm using a microplate reader (Labsystems, iEMS reader M F , Helsinki, Finland). Phosphate standards were prepared by dissolving 1.0068 g potassium hydrogen phosphate (NaHiPC^) in distilled water to a final volume of 1 L (200 pg phosphorus/mL). From this stock solution, working standards were prepared ranging from 1 -8 pg phosphorus/mL. A blank was prepared by adding 1 mL 20% T C A to 8 mL of the reaction medium, followed by 1 mL of 0.02 M ATP. A 1 mL aliquot was removed and added to a test tube containing 1 mL chilled 20% T C A . The blanks were centrifuged and treated in the same manner as the samples. Ca 2 +ATPase total activity was defined as the activity at 10 min and expressed as uM Pi at 10 min/mg protein at 25°C. 3.3.1.3. Surface hydrophobicity The surface hydrophobicity of N A M extracts was determined as described in section 3.2.1.7. 69 3.4. Fish mince treated with AFP, commercial cryoprotectants and their blends Ling cod fillets Minced Cryoprotectant blends: • AFP (AFGP or AFPI, 0.05 mg/g) + 0.3% PO • Sucrose, sorbitol (8% sucrose-sorbitol, 1:1 by wt) with no PO • AFP (AFGP or AFPI, 0.05 mg/g) + 0.3% PO + 2% sucrose, 2% sorbitol • Trehalose 8% + 0.3% PO • Control- fish mince only (no cryoprotectants) Unfrozen (UF) Frozen -20 ±2°C and freeze-thawed • Analyzed for: moisture, crude protein, % expressible moisture, salt extractable proteins, SDS-PAGE profiles, Raman and FT-Raman spectral profiles, DSC and photomicrographs Figure 12. Overview of working plan for fish mince treated with antifreeze proteins, commercial cryoprotectants and their blends. In this study, the objective was to study the possibility of a synergistic effect between AFP and a lower concentration of the commercial blend. In addition, a portion of the fish mince was treated with an alternative commercial cryoprotectant, trehalose. Ling cod fish were purchased, within four days after capture, from Albion Fisheries Ltd. (Vancouver, B C , Canada) in February 2004. Fish were deboned, skinned, filleted (2 kg). 70 Fillets were minced using a BEEM-GIGANT Grinder Model T Y P E F5-10 (BEEM California Corp., Glendale, CA) with a 4 mm screen. Mince portions (100 g) were used for each cryoprotectant treatment as follows: AFPI or AFGP (fraction 6-8), each at 0.05 mg/g mince (0.005% w/w), were added to their corresponding mince portion along with 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate, w/w); another two portions were prepared with AFPI or AFGP (each at 0.05 mg/g mince) and a lower concentration of the commercial blend consisting of 2% sucrose, 2% sorbitol and 0.3% phosphates (w/w); one portion was used for the treatment consisting of 4% sucrose + 4% sorbitol with no phosphates (w/w); and one more portion was used for the treatment consisting of 8% trehalose and 0.3% phosphates (w/w). A control treatment, mince with no cryoprotectants, was also used. Each mince treatment was mixed thoroughly for 3 min using a Kitchen-Aid bowl mixer set on low speed (setting of "1"). The minced samples were packed in 0.1 mm thick 6x12" polyethylene bags (Fisherbrand, 01-812-10C, Fisher Sci., NJ) and stored at -18 + 2°C. Sets of unfrozen samples from each treatment were analyzed that same day and used as the unfrozen comparisons. Frozen samples were submitted to 8 freeze-thaw cycles, as described in section 3.1. 3.4.1. Analysis Unfrozen and freeze-thawed samples were analyzed for the following: moisture, crude protein (section 3.2.1.1.), expressible moisture (section 3.2.1.2.), salt extractable proteins (section 3.2.1.4.), SDS-PAGE profiles (section 3.2.1.5.), Raman (section 3.2.1.8.) and FT-Raman spectroscopy, and differential scanning calorimetry (DSC). Following analysis, samples were stored at -80°C. 71 3.4.1.1. FT-Raman spectroscopy Due to increased fluorescence following freeze-thawing of the samples, the visible laser Raman spectra had poor signal to noise ratio. Therefore, the samples were sent for analysis by FT-Raman spectroscopy at the Instituto del Frio- CSIC (Madrid, Spain). Samples were packed in polyethylene bags, covered with frozen ice-packs and placed inside a styrofoam box before being shipped by courier to the laboratory. For the analysis, portions of mince were transferred to N M R glass tubes (5 cm height and 5 mm i.d.; Wilmad Glass Co., Inc., Buena, NJ) to fill ~ l cm height. FT-Raman spectra of samples excited with a 1064 nm N d : Y A G laser line were recorded on a Bruker RFS 100/S FT-spectrometer. The samples, thermostated at 4 °C, were illuminated by laser power at 300 mW. The scattered radiation was collected at 180° to the source, and frequency-dependent scattering of the Raman was corrected by multiplying point by point with (vi a s e r/v) 4 . Reported frequencies are considered accurate to ± 0.5 cm"1, as deduced from frequency standards measured in the spectrometer. Raman spectra were resolved at 4 cm"1 resolution with a liquid nitrogen-cooled Ge detector. Two thousand scans were recorded for each of three replicate samples, resulting in a total of 6000 scans per treatment, and randomizing possible differences in composition. The estimation of the secondary structure composition of the proteins was carried out based on least square analysis of the amide I region using the method of Williams (1983) computed by the Raman Spectral Analysis Package (RSAP) program version 2.1., of Przybycien and Bailey (1989). 72 3.4.1.2. Differential scanning calorimetry (DSC) of freeze-thawed mince Thermal behavior during the thawing of the frozen/freeze-thawed mince samples was evaluated by DSC (model 4207 MC-DSC, Calorimetry Science Corp). Sample treatments (ca. 0.05 g, equivalent mass of water in all samples) were placed in Hastelloy-C ampoules, which were then sealed and accurately weighed. A n empty cell was used as the reference. In the DSC, each sample was cooled to -20°C and held at that temperature for 30 minutes to ensure complete freezing. The sample was then scanned from -20°C to -10°C at l 0 C/min, followed by scanning from -10°C to +5°C at 0.1°C/min, and from +5°C to +25°C at l°C/min. The pre-scan thermostat was set at 10 min for the last two scans. After scanning, the thermogram (heat capacity vs. temperature) showed an endothermic peak of melting for each sample. The endothermic peak was integrated from -10°C to the temperature where the baseline was reestablished in order to measure the heat of fusion {AHA) necessary to melt the ice, which corresponded to the free water in each sample. The heat of fusion of pure water, AHwater, measured in this study was: 328 + 1.72 J/g (mean value + standard deviation of three replicate analyses) which is in close agreement to the known value of 333.05 J/g (Khan et al., 2004). The total amount of water (g H2O) in each sample was determined by moisture analysis (section 3.2.1.1). The antifreeze activity (AF) of the different cryoprotectants was determined using an equation modified from Mitsuiki et al. (1998), and defined as the grams of unfrozen water per gram of total water in the sample, calculated according to equation (1): AF = \-[AHA IAHwater] (1) where AHA = peak area of sample/ mass of water in sample (J /g water in the sample). 73 3.4.1.3. Differential scanning calorimetry (DSC) of fresh mince Fresh ling cod were purchased, within four days after capture, from Albion Fisheries Ltd. (Vancouver, BC, Canada) in April 2005. Fish were deboned, skinned, filleted and prepared into separate formulations as described in section 3.4. The antifreeze activity of the minced samples was analyzed with DSC under the following conditions. Each sample was first submitted to three cooling scans from +20°C to +5°C at 1 °C/min, followed by 30 min holding temperature at +5°C before scanning from +5°C to -10°C at 0.1°C/min, and then from -10°C to -20°C at l°C/min. The samples were then submitted to three heating scans as described in section 3.4.1.2. After scanning, the thermograms were analyzed as described in section 3.4.1.2. 3.4.1.4. Photomicrographs and video-recording observations Changes occurring on the fish mince surface were monitored by capturing images and recording video using a 50x magnification lens attached to a handheld digital USB microscope (The ProScope™ USB Microscope, Bodellin Technologies, OR) connected to a computer (Macintosh i-Mac Power PC). Samples from completed freeze-thaw studies were removed from frozen storage at -80°C for observation using the USB microscope mounted on a camera tripod. Video recording was immediately started and continued for a period of 3 min. During the image capturing, the temperature on the surface of the sample was recorded using an Infrared Thermometer model 39650-04 (Cole Instruments Co., Chicago, IL). Images were taken as snap-shots from the video recordings at "0 min", and 1.5-2.0 min. 74 3.5. Conformational changes of antifreeze protein solutions at subzero temperatures Analysis: Molecular mass, amino acid composition Antifreeze proteins: AFGP, AFPI (O.lg/mL w/v) Solutions (0.1 g/mL w/v) in phosphate buffer (pH 7.5) Held at 22. -0.5.-1.8 and -4.0°C Analysis: Raman spectral profiles Figure 13. Overview of working plan for analysis of antifreeze proteins solutions held at subzero temperatures. The lack of cryoprotectant behaviour of the antifreeze proteins in a muscle (minced) system observed in the studies presented so far in this thesis led to the development of this new investigation. The objective of this study was to examine the conformational changes that take place in AFGP and AFPI solutions when held at subzero temperatures. This objective was established in order to learn more about the behaviour of the AFP when held at temperatures closer to those encountered by cold water fish. The information obtained from this study could help clarify some of the questions regarding the type of interactions involved between AFP and ice crystals, as well as the effect of the AFP's secondary structure on the ice-binding mechanism. Molecular mass measurements, and amino acid analysis were conducted on AFGP 8 and 75 AFPI. In addition, the Raman spectral profiles of AFGP8 and AFPI solutions were analyzed at four different sub-zero temepratures. 3.5.1. Molecular mass measurements and amino acid analysis Molecular mass measurements of the antifreeze proteins were performed in the M S L / L M B Proteomics Core Facility at The University of British Columbia, using an Applied Biosystems Voyager DE-STR M A L D I Time-Of-Flight (TOF) mass spectrometer. Amino acid analysis of the freeze dried AFGP and AFPI was performed by the Amino Acid Analysis Facility HSC Advanced Protein Technology Centre (Dept. of Structural Biology & Biochemistry, Hospital for Sick Children, Toronto, ON). 3.5.2. Raman spectroscopy Raman spectra were recorded for all samples using a JASCO model NR-1100 laser Raman spectrophotometer (section 3.2.1.8.). Low molecular weight antifreeze glycoprotein, fraction 6-8 (AFGP) and type I antifreeze protein (AFPI) (O.lg/mL w/v) were solubilized in pH 7.5 phosphate buffer (3.38 m M NaH 2 P0 4 , 15.5 m M Na 2 HP0 4 ) . Each solution was placed in a capillary tube and held at the pre-set holding temperatures using a constant temperature measuring accessory (JASCO model RT-IC, Japan Spectroscopic Co., Ltd., Tokyo, Japan) for Raman measurements. The measurement conditions were: laser power of 200 mW, slit height 4 mm, sample speed 120 cm"1 min"1; data collection every cm"1, with 6 scans per sample. The solutions were first held at ambient temperature (22°C) for 15 min. Once the spectrum was collected, the temperature in the cooling unit was set to -0.5°C without removing the capillary tube. 76 The solution was allowed to equilibrate for 15-20 min before the spectral data collection at the subzero temperature. The same procedure was performed for the remaining temperatures (-1.8 and -4.0°C), where the capillary tube containing the AFP solution remained in the cooling unit and was allowed to equilibrate for 15-20 min before data collection. Recorded spectra were analyzed using Grams/32® Spectral Notebase™ version 4.14 level II (Galactic Industries Corp., Salem, NH). Spectral data were baseline corrected and smoothed as described in section 3.2.5. Spectral data were normalized to the intensity of the C H 2 bending band at 1450 cm"1 (Li-Chan and Nakai, 1991). 3.6. Statistical Analysis For most experiments, the mean and standard deviation were calculated. Analysis of variance (ANOVA) and Tukey's comparison tests were carried out using the program Minitab Release 13.32 for Windows (Minitab Inc., State College, PA). In the case of Raman spectroscopy, to discriminate between temperature treatments, the standard deviation was assumed from the coefficient of variance of 5% for the method in order to calculate the upper and lower 95% confidence bands for each measurement (Badii and Howell, 2003). 77 4. RESULTS AND DISCUSSION 4.1. Fish fillets: soaking or spraying with cryoprotectants In this study, the possibility of incorporating cryoprotectants into intact fish fillets prior to freezing and frozen storage was evaluated. Due to the high cost and the limited availability of AFP, this experiment was conducted using solutions with 4% sucrose and 4% sorbitol (w/v), phosphate buffer and blue dye. Soaking fillet pieces with phosphate buffer gave a 5% increase in the fillet weight, while soaking with sucrose and sorbitol gave a 3-6% weight increase; spraying with phosphate buffer or sucrose and sorbitol resulted in a 1-2% weight increase on the fillet piece. After soaking the fillet pieces for 30 min, the solution penetrated only 3 mm from the surface of the fillet, as observed by the blue dye. In the case of the sprayed samples, the dye was barely visible. These observations could explain why the results (Table 4) for water binding capacity (WBC) and drip loss for sprayed or soaked samples were not significantly different (P> 0.05) from the control (untreated fillets) after 8 freeze-thaw cycles. In the work performed by Payne et al. (1994), extremely small meat samples (3x3x6 mm) were soaked in AFP solutions (AFGP and type I AFP) at 15°C for a minimum of 4.5 h and up to 7 days prior to freezing at -20°C. The AFP were effective at concentrations ranging from 0.1-1 mg/mL, in decreasing ice crystal size compared to the control. The effect depended on the AFP concentration used as well as the period of soaking in the saline solution, where the highest AFP concentration (1 mg/mL) required a minimum of 48 firs of soaking time while the lower concentrations required longer soaking time. Prolonged soaking periods will not be a practical approach for the food industry, as the costs implicated with maintaining solutions and samples cold for long 78 periods of time prior to freezing seems unlikely. It should also be considered that in order to slow down the onset of microbiological, chemical, and biochemical changes in fish, the soaking practices will have to be conducted at lower temperatures for short periods of time. As seen in this present study, soaking fillet pieces (3x3.5x8 cm) for 30 min was not sufficient time for the cryoprotectant solutions to penetrate the samples; spraying the samples 5 times on each side was also not enough. 79 Table 4. Moisture, water binding capacity (WBC) and percent drip loss of fish fillets after soaking or spraying with sucrose and sorbitol1 or phosphate buffer (POB) 2. Sample % Moisture WBC % Drip loss UF* UF 8-FT4 UF 8-FT control 79.1 1.4 1.3 N D 14.0 (1.0) (0.1) (0.1) (0.5) soaked with 76.3 1.5 1.5 N D 25.0* POB (1.5) (0.1) (0.1) (3.7) sprayed with 79.2 1.5 1.5 N D 16.3 POB (0.6) (0.1) (0.1) (0.2) soaked with 76.2 1.5 1.5 N D 17.2 4% sucrose (1.1) (0.1) (0.1) (4.0) 4% sorbitol sprayed with 79.3 1.5 1.5 N D 14.0 4% sucrose (0.3) (0.1) (0.2) (3.2) 4% sorbitol 1 4% sucrose and 4% sorbitol (w/v) in phosphate buffer 20.1 M phosphate buffer (pH 6.8) 3 U F : unfrozen samples 48-FT: 8 freeze-thaw cycles where one freeze-thaw cycle consisted of freezing for 18 h at -18 ± 2°C and thawing for 6 h at 4°C. Numbers in brackets indicate SD values from three replicates. N D = not determined Values are not significantly different (P> 0.05) between samples. * P= 0.051. Fish used in this experiment were caught in July 2002. 80 4.2. Fish mince treated with AFP compared to commercial cryoprotectants 4.2.1. Expressible moisture (%EM), cook loss (%CL) After freeze-thawing, mince treated with the commercial blend showed significantly lower (P<0.05) expressible moisture (%EM) compared to the control mince and samples treated with AFP at high concentrations (Table 5). However, in terms of cook loss (%CL), both commercial blend and AFGP at 0.05 mg/g mince had significantly lower % C L compared to AFGP and AFPI at 0.5 mg/g mince. In fact, the % C L for these latter two treatments at high AFP concentrations did not differ significantly from the freeze-thawed control (Table 5). The commercial blend was the only formulation that had low % E M after freeze thaw conditions (Tables 5 and 6). Freeze-thawed mince samples treated with AFP (0.01 and 0.005 mg/g) had lower % E M than the control mince; however, these differences were not significantly different (P>0.05). Benjakul and Bauer (2000) reported that the amount of exudates from fish muscle tends to increase with increased number of freeze thawing cycles, and attributed the drip loss to disintegration of membrane structure and increased protein denaturation. The % C L for freeze-thawed AFPI and AFGP (0.01 mg/g) were lower that the control. Nevertheless, these values were still much higher than those observed for the commercial blend. The lower percentage of expressible moisture in the mince treated with the commercial blend is attributed to the carbohydrates' water binding capacity and cryoprotective action that preserves protein functionality resulting in greater water-holding capacity (Jittinandana et al., 2005). 81 Table 5. Percent expressible moisture (% EM), cook loss (% CL) and textural hardness (g) of freeze-thawed1 mince treated with high AFP concentrations and with or without commercial cryoprotectants. Sample2 ' % E M : % CL Textural ' hardness (g) Unfrozen control 24.5a 13.5a 510a (1.0) (3.8) (79) Control 33.9b 29.7b l,817 b (2.0) (10.3) (6) Commercial blend 19.6a 17.3a 528a (3.2) (2.5) (87) AFGP 33.7b 18.3a l,976 b (0.05 mg/g mince) (1.5) (9.8) (283) AFGP 35.3b 25.7b 2,776b (0.5 mg/g mince) (0.6) (1.4) (ND) AFPI 35.7b 29.7b 3,098c (0.5 mg/g mince) (1.5) (3.1) (365) Samples were submitted to 8 freeze-thaw cycles. One freeze-thaw cycle consisted of freezing for 18 h at -18 ± 2°C and thawing for 6 h at 4°C. 2Sample abbreviations are as follows: Control: mince only AFGP: low molecular weight antifreeze glycoprotein, fraction 6-8 AFPI: type-I antifreeze protein Commercial blend: 4% sucrose, 4% sorbitol, 0.15% trypolyphosphate and 0.15% sodium pyrophosphate (w/w) Numbers in brackets indicate SD values from three replicates. ND= not determined Values bearing different superscript letters (a, b, c) within each column are significantly different (P < 0.05). Fish used in this experiment were caught in January 2003. 82 Table 6. Percent expressible moisture (% E M ) , cook loss (% C L ) and textural hardness (g) of unfrozen and freeze-thawed mince treated with low A F P concentrations and with or without commercial cryoprotectants. Sample1 % EM % CE Textural hardness fe) UF2 8-FT3 UF 8-FT UF 8-FT Control 16.9 b x 38.2 c y 10.8 a x 26.5 c y 163 b x 5 3 1 a b y (0.2) (0.4) (1.0) (0.4) (17) (147) Commercial blend 9.0 a x 19.3 a y 3.7 a x 2.8 a x 96 a x 312 a y (1.4) (2.9) (0.4) (1.1) (16) (53) AFGP 16.2 b x 36.2 b c y 4.5 a x 19.4 b y 158 b x 878 c y (0.01 mg/g mince) (0.4) (0.7) (0-2) (3.5) (?) (83) AFGP 16.5 b x 34.7 b c y 6.1™ 33.5 d y 164 b x 556 a b y (0.005 mg/g mince) (1.4) (1.6) (0.5) (0.8) (11) (91) AFPI 13.7 a b x 33.2 b y l l . l a x 21.5 b y 170 b x 897 c y (0.01 mg/g mince) (2-3) (1.2) (7.8) (1.0) (9) (102) AFPI 15.7 b x 34.8 b c y 8.8 a x 29 .9 c d y 156 b x 6 5 2 b c y (0.005 mg/g mince) (4.0) (0.4) (2-8) (0.9) (13) (70) 'Sample abbreviations are listed in Table 5. 2UF: unfrozen samples; 38-FT cycles: 8 freeze-thaw cycles. Numbers in brackets indicate SD values from three replicates. Values bearing different superscript letters (a, b, c, d) within each column are significantly different (P < 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P< 0.05). Fish used in this experiment were caught in April 2003. 83 Following the eight freeze thaw cycles, fish mince treated with AFPI and AFGP (0.5 mg/g mince) had a layer of ice crystals forming on the surface (Figure 14). This observation was less visible in the sample with lower concentration of AFGP (0.05 mg /g mince), and was not visible at all in the commercial blend and control samples. When low AFP concentrations were used, the minced samples did not show a layer of small crystals (Figure 15) as previously seen for the mince with high AFP concentrations. Figure 14. Visual observations on freeze-thawed fish mince treated with 0.05 or 0.5 mg/g AFP, compared to control mince and mince treated with commercial blend. Sample abbreviations are listed in Table 5. 84 A F G P (0.01 mg/g) AFP type I (0.01 mg/g) Figure 15. Visual observations on freeze-thawed fish mince treated with 0.005 or 0.01 mg/g AFP, compared to control mince and mince treated with commercial blend. Sample abbreviations are listed in Table 5. A small number of studies have evaluated water holding capacities of meat samples in the presence of AFP. The effect of AFGP on the loss of water holding capacity was observed in the study performed by Payne and Wilson (1994) using Antarctic and black cod muscle with intrinsic AFP. In their study, the authors found that black cod (Paranotothenia augustata) muscle presented low concentrations of AFGP (<0.01 mg/g muscle) that did not form ice spicules. The black cod muscle had significantly higher drip loss after freeze-thawing. On the contrary, Antarctic cod (Paranotothenia mawsoni) with AFGP concentration of 1.0 mg/g muscle tissue did form ice spicules, and the drip loss did not significantly increase compared to the chilled control. In the present study with ling cod mince, the highest AFP concentration used was 0.5 mg/g and ice spicule formation was present in particular at 0.5 mg/g. However, the results failed to show an improvement of drip loss for mince samples treated with AFP. 85 In other studies, Payne and Young (1995) injected AFP to lambs prior to slaughter and suggested that time is required for the AFP to become incorporated into the muscle tissue, as ice crystals were smallest in the lambs injected with AFGP 24 hours before slaughter, to a final concentration of 0.01 pg/kg. In addition, the best way for AFP to be incorporated into the muscle pieces was by soaking small muscle pieces for extended periods of time. However, as Payne and colleagues (1994) suggested, more cost-effective methods of incorporating AFP into meat need to be developed. In this research, incorporation of AFP was performed by mixing thoroughly for 3 min using an electric mixer. Judging by the ice crystal formation on the mince samples with AFP and absence of visible ice crystals in the control and commercial blend samples, there is clear evidence of an influence of AFP in the minced sample. However, these results did not agree with previous observations reported by Payne and Wilson (1994) on the decrease of drip loss involving high AFP concentrations and ice spicule formation on the fish muscle. 86 4.2.2. Texture analysis Freeze-thawed samples with high AFP concentration exhibited considerable amount of drip loss and a sponge-like texture which would indicate protein denaturation. Samson et al. (1985) explain that gadoid fish tend to exhibit a 'spongy' and 'rubbery' texture upon denaturation of the muscle proteins. This problem seems to be enhanced by mincing practices which could lead to an increase in formaldehyde-induced intermolecular cross-linking between adjacent protein molecules. Hardening and toughening of frozen fish and during frozen storage is attributed to mechanisms that include cell damage caused by ice crystal formation, myofibrillar protein aggregation and denaturation, as well as lipid-protein interactions (Mackie, 1993; Jittinandana et al., 2005). When analyzed for textural hardness, the force required upon 75% penetration was higher for samples treated with AFP at 0.05-0.5 mg/g, followed by the control mince. The only sample that did not exhibit sponge-like texture was the mince containing the commercial blend which required much lower penetration force compared to control and AFP samples (Table 5). The textural hardness of minced samples with AFGP and AFPI (0.01 mg/g), showed significantly higher penetration force values (878 and 897 g, respectively) than the control mince (531 g) after freeze-thawing (Table 6). These observations indicate that in the presence of antifreeze proteins, there was increased textural hardness. The commercial blend exhibited lower textural hardness than the control, but the difference was not significant (P >0.05). In addition, AFGP (0.005 mg/g) was not significantly different from either the control or commercial blend (Table 6). 87 One important observation that should be made is the variability found between Tables 5 and 6 and the different time of catch for the fish used in each of the experiments. Differences in % E M , % C L , textural hardness and proteins extractability (discussed in the next section) could be seen between the two batches for unfrozen fish with no cryoprotectants, as well as fish treated with commercial cryoprotectants. In the present study, higher values for % E M and textural hardness could be seen for ling cod mince caught in the month of January (Table 5) compared to ling cod caught in April of the same year (Table 6). According to Love (1980), fish is subject to seasonal variations which can hinder studies related to their freeze denaturation. Botta et al. (1987) observed seasonal variations in firmness (shear-compression force) of cooked Atlantic cod (Gadus morhud) muscle. In addition, Ingolfsdottir et al. (1998) performed an extensive study on seasonal variations in physicochemical and textural properties of Atlantic cod mince. The researchers found differences between fish caught in winter and summer. For example, hardness and cohesiveness showed a drop in March to May and another drop was observed for cohesiveness during the autumn months. Expressible moisture changed slightly from March to August whereas results from the winter months indicated greater fluctuations. In addition, the authors also reported changes for protein solubility in March and July and protein surface hydrophobicity increased gradually from November to May followed by a sudden drop only to increase again after. More recently, Yuan et al. (2006) found that upon storage in ice, myosin from silver carp (Hypophthalmichthys molitrix) was much more stable in the summer compared to the winter season. 88 4.2.3. Protein extractability in salt and SDS-PAGE profiles Table 7 shows the changes in salt extractability of proteins (%SEP) for mince samples treated with different concentrations of AFP. Freeze-thawed mince samples treated with AFP (0.05-0.5 mg/g) had markedly lower %SEP compared to the unfrozen control mince. When mince samples were treated with lower concentration of AFP (0.005-0.01 mg/g), about 70% of the salt-soluble proteins became insoluble following 8 freeze-thaw cycles. The exception was for the mince with commercial blend which remained similar to its unfrozen counterpart. The significantly lower %SEP for the mince samples indicates that upon freeze thawing as the salt-soluble proteins became denatured, they began forming aggregates that were insoluble in the salt solution. Sych et al. (1990) tested the cryoprotective effects of different materials (carbohydrates, polyols, protein hydrolysates, and hydrocolloids) on Atlantic cod (Gadus morhua) surimi stored at -20°C for several weeks and showed that the carbohydrate/polyol treatment (sucrose and sorbitol), similar to the commercial blend used here, had the greatest stabilizing effect, as measured by %SEP. Changes in protein extractability can be a useful tool in order to evaluate the extent to which protein denaturation and aggregation have occurred during frozen storage or after temperature abuse situations, such as freeze-thawing cycles (Lian et al., 2000). The effect of freezing and frozen storage on fish muscle proteins has been shown to cause protein aggregation along with a general decrease in salt-soluble protein extraction (Benjakul et al., 2003; Badii and Howell, 2002; Lian et al., 2000; Sych et al., 1990). Herrera and Mackie (2004) established what other researchers have expressed, that a decrease in salt solubility is most likely due to changes that take place in the protein 89 microenvironment, and that these changes are in association with the practice of freeze-thawing. Sultanbawa and Li-Chan (1998) indicated that ling cod surimi treated without cryoprotectants had a significant decrease in salt extractable proteins, compared to ling cod mince with polyol blends. Matsumoto (1980) stated that the change in the amount of soluble actomyosin is regarded as the primary criterion of freeze denaturation but cautioned that solubility data do not tell precisely how much protein is denatured nor how much is native; rather, they provide a relative measure of denaturation. The SDS-PAGE profiles of salt extractable proteins from mince treated with AFP at 0.005-0.01 mg/g or commercial blend before freezing and after 8 freeze-thaw cycles are shown in Figure 16A and 16B. The results for the unfrozen samples (lanes 1, 3, 5 in Figures 16A and B) clearly show the M H C band at 205 kDa. In contrast, the intensity of the M H C band receded for the salt extractable proteins from mince samples upon eight freeze-thaw cycles, while the actin band was retained (Lanes 2, 4 and 6, Figure 16A; lanes 4 and 6, Figure 16B). The freeze-thawed commercial blend did not exhibit this pattern and seemed to be the only treatment to retain the presence of the M H C band after freeze thawing (lane 2, Figure 16B). According to Tejada et al. (1996) and Careche et al. (1998a), muscle proteins become gradually inextractable in salt solutions, and depending on species, conditions, and time of storage, these proteins will be extracted to more or less extent in SDS or SDS plus M E , and eventually, a residue nonextractable in these solutions will be obtained. The disappearance of the M H C band and retention of the actin band is consistent with the observations made by Sultanbawa and Li-Chan (1998), in which the SDS-PAGE profiles of salt extractable proteins from ling cod surimi without cryoprotectants did not retain the M H C upon freeze thawing, whereas the SEP of surimi 90 treated with polyol blends did. The reason could be due to freeze-induced denaturation which can result in formation of large myofibrillar aggregates that are not easily extracted in salt, and become too large in size to enter the gel matrix. According to Matsumoto (1980), dissociation of actomyosin into F-actin and myosin occurs immediately after freezing, and it is primarily the myosin component which undergoes aggregation and insolubilization. 91 Table 7. Salt extractable proteins (% SEP) for unfrozen and freeze-thawed mince samples treated with different concentrations of AFP or with and without commercial cryoprotectants. Sample1 % Salt extractable proteins' UF2 8-FT* HighJAFP^concentration Control 60 Commercial blend N D 38a AFGP (0.05 mg/g mince) N D 38a AFGP (0.5 mg/g mince) N D 28 a AFPI (0.5 mg/g mince) N D 34a Low AFP concentration Control 51 J X 30 a y Commercial blend 5 2 ax 53 b x AFGP (0.01 mg/g mince) 58 a x 24 a y AFGP (0.005 mg/g mince) 56 a x 32 a y AFPI (0.01 mg/g mince) 5 4 ax 29 a y AFPI (0.005 mg/g mince) 53 a x 3 2 ay 'Sample abbreviations are listed in Table 5. 2UF: unfrozen samples;38-FT cycles: 8 freeze-thaw cycles. Values represent mean of duplicate samples. Values bearing different superscript letters (a, b) within each column are significantly different (P < 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P < 0.05). N D = not determined Fish used in High AFP experiment were caught in January 2003. Fish used in Low AFP experiment were caught in April 2003. 92 1 2 3 4 5 6 Ac^jj B 2 3 5 6 Figure 16. SDS-PAGE profiles of ling cod mince salt extractable proteins treated with AFP or with and without commercial cryoprotectants. Sample abbreviations are listed in Table 5; UF = unfrozen, FT= after 8 freeze-thaw cycles. A. Lane 1: control mince-UF, 2: control mince-FT, 3: AFGP (0.005 mg/g)-UF, 4: AFGP (0.005 mg/g)-FT, 5: AFPI (0.005 mg/g)-UF, 6: AFPI (0.005 mg/g)-FT. B. Lane 1: commercial blend-UF, 2: commercial blend-FT, 3: AFGP (0.01 mg/g)-UF, 4: AFGP (0.01 mg/g)-FT, 5: AFPI (0.01 mg/g)-UF, 6: AFPI (0.01 mg/g)-FT. 4.2.4. Total sulfhydryl The results for total SH groups in the salt extractable proteins are shown in Table 8. Upon freeze thawing, mince sample formulations with AFGP concentrations of 0.05 and 0.5 mg/g had significantly higher total SH values (P<0.05) than the control, commercial blend and AFPI (0.5 mg/g). However, when lower concentrations of AFP (0.01 and 0.005 mg/g) were used there was a decrease in the total SH content upon freeze thawing compared to the commercial blend (Table 8). Freezing is known to cause protein denaturation and induce the oxidation of SH groups by forming disulfide (SS) bonds (Lin et al., 2005). Similar observations have been made by Herrera and Mackie (2004), Benjakul and Bauer (2000) and Lian et al. (2000) working with farmed rainbow trout fillets, cod muscle proteins and red hake, respectively. Their studies suggested that freeze-thawing causes a decrease in total and surface (reactive) - S H contents as a consequence of the changes in actomyosin tertiary structure, where the masked - S H surface groups become more exposed and oxidized to disulfide bonds. In addition, unfolding and aggregation of proteins may take place at a faster rate in the absence of cryoprotectants, preventing subsequent measurements from being taken as these protein aggregates can mask changes in the number of total and reactive SH groups (Herrera and Mackie, 2004). 94 Table 8. Total SH contents of salt extractable proteins from unfrozen and freeze-thawed mince samples treated with different concentrations of AFP or with and without commercial cryoprotectants. Sample1 :~ Moles/10 g of protein UF2 8-Fr ••' High AFP concenj^jatioii Control 6.2 1.6a (0.9) (0.5) Commercial blend N D 2.0a (0.1) AFGP (0.05 mg/g mince) N D 3.6b (0.3) AFGP (0.5 mg/g mince) N D 3.3b . (0.3) AFPI (0.5 mg/g mince) N D 1.7a ... (°-3> LsOW f\-r.-r, LOilLcrllrllliirtl Control 5.3a 4.3b (0.9) (0.2) Commercial blend 4.3 a 5.5d (0.1) (0.2) AFGP (0.01 mg/g mince) 7.8b 4.3 b c (0.2) (0.3) AFGP (0.005 mg/g mince) 4.9a 4.9 c d (0.2) (0.3) AFPI (0.01 mg/g mince) 5.2a 2.7a (0.1) (0.6) AFPI (0.005 mg/g mince) 4.7a 3.7b (0.1) (0.3) Sample abbreviations are listed in Table 5. 2UF: unfrozen samples;38-FTcycles: 8 freeze-thaw cycles. Numbers in brackets indicate SD values from three replicates. N D = not determined Values bearing different superscript letters (a, b, c, d) within each column are significantly different (p < 0.05). Fish used in High AFP experiment were caught in January 2003. Fish used in Low AFP experiment were caught in April 2003. 95 4.2.5. Hydrophobicity The lowest surface hydrophobicity (S0) values of salt extractable proteins from freeze-thawed mince were seen for the control, commercial blend and AFPI (0.5 mg/g) formulations, while the freeze-thawed AFGP formulations (0.05 and 0.5 mg/g) had higher surface hydrophobicity (Table 9). According to Mackie (1993), an increase in fluorescence could indicate a more hydrophobic environment which would occur when a protein undergoes denaturation with the exposure at the surface of hydrophobic groups normally associated with the interior of the molecule. In the case of freeze-thawed samples treated with low AFP (0.005-0.01 mg/g) concentrations, a decrease in S 0 was also observed (Table 9). Only the freeze-thawed commercial blend had a value (325%"') close to its unfrozen counterpart (354%"'). The rest of the freeze-thawed formulations, including the control, showed lower S 0 values than their unfrozen counterparts. Other studies (Sultanbawa and Li-Chan, 2001; Careche et al., 1998a; Careche and Li-Chan, 1997) also reported a decrease in surface hydrophobicity of the proteins upon freezing and/or frozen storage. Explanations for these results have been suggested to be an extensive denaturation resulting in more interactions between hydrophobic groups through noncovalent interactions; this in turn will cause extensive aggregation which will be only partially disrupted with the aid of sonication and/or dilution. Benjakul et al. (2003) reported that a considerable decrease in free hydrophobic groups could be the result of cross-linking, induced by formaldehyde, between . the methyl groups of hydrophobic amino acids side chains. Careche and Li-Chan (1997) also indicated that in presence of formaldehyde, during frozen storage, more extensive protein denaturation 96 could result in more interactions between hydrophobic groups and formation of covalent bonds. Sultanbawa and Li-Chan (2001), and Herrera and Mackie (2004) found that cryoprotectants such as lactitol and polydextrose were effective in minimizing the exposure of hydrophobic residues on the surface of actomyosin, and also decreased the tendency for proteins to cluster through intermolecular hydrophobic interactions that form aggregates. In this study, the S 0 results indicate that the commercial blend (sucrose, sorbitol and polyphosphates) seemed to have a cryoprotective effect on the muscle proteins as seen by the similar surface hydrophobicity value of the freeze-thawed treatment to its unfrozen counterpart. 97 Table 9. Surface hydrophobicity values (S 0, %"') of salt extractable proteins from unfrozen and freeze-thawed mince samples treated with different concentrations of AFP or with and without commercial cryoprotectants. Sample1 So (%"') UF 8-FT • High AFP concentration Control 326c 243b Commercial blend N D 197a AFGP (0.05 mg/g mince) N D 270 b c AFGP (0.5 mg/g mince) N D 282c AFPI (0.5 mg/g mince) N D 205a Loiv AFP concentration1 _ Control 339" 243 a y (11.6) (21.0) Commercial blend 354 a x 325 b x (12.7) (12.2) AFGP (0.01 mg/g mince) 361 a x 266 a y (16.5) (2.5) AFGP (0.005 mg/g mince) 340 a x 254 a y (13.8) (21.1) AFPI (0.01 mg/g mince) 336 a x 240 a y (17.4) (10.9) AFPI (0.005 mg/g mince) 340 a x 244 a y (14.7) (11.6) Sample abbreviations are listed in Table 5. Surface hydrophobicity values for High AFP concentration represent mean of two replicates. N D = not determined. Surface hydrophobicity values for Low AFP concentration represent mean of three replicates. Numbers in brackets indicate SD values. Values bearing different superscript letters (a, b) within each column are significantly different (P < 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P < 0.05). Fish used in High AFP experiment were caught in January 2003. Fish used in Low AFP experiment were caught in April 2003. 98 4.2.6. Raman spectroscopy The results from Raman spectroscopy are shown in Figure 17 and Table 10. Upon freeze-thawing, the spectral signal to noise ratio decreased drastically for all mince treatments. Due to the decreased quality of the Raman spectra of the freeze-thawed samples, the discussion will be focused on the major bands and regions that were less affected by the low signal to noise ratio. Amide I region The amide I profile can be described in terms of C=0, C-N, in plane N - H , and C-C-N vibrations. In particular, the strongest intensity in this region (1645-1657 cm"1) is attributed to a-helix content (Tu, 1986; Herrero et al., 2004) which can be seen for the unfrozen mince samples in Figure 17A. Upon freeze thawing, the amide I region shifted to a higher frequency. This shift in frequency has been attributed a decrease in the a-helix content as a more random coil structure is generated as it has been reported that frozen storage of fish muscle leads to an increased proportion of (3-sheet and random coil at the expense of a-helix structures (Bouraoui et al., 1997; Careche et al. 1999; Careche et al., 2002). In addition, after freeze thawing, the control and AFGP (0.005 and 0.01 mg/g) mince samples showed lower intensity of the band ca. 1660 cm"1, reflecting the weakening of the a-helix structure. In the case of the freeze-thawed commercial blend and AFPI (0.005 and 0.01 mg/g) mince samples, the amide I band intensity increased (Table 10). CHj bend region Changes in aliphatic residues after frozen storage could be analyzed from the C-H deformations (1450 cm"1). In this study, the control and AFGP (0.005 and 0.01 mg/g) 99 showed a decline in peak intensity upon freeze thawing compared to the commercial blend and AFPI (0.005 and 0.01 mg/g) samples. Studies on hake fillets (Careche et al., 1999; Careche et al., 2002), frozen cod (Badii and Howell, 2004) and ling cod natural actomyosin (Sultanbawa and Li-Chan, 2001) have shown changes, including a decrease in this band's intensity, during frozen storage. These changes have been attributed to hydrophobic interactions of the aliphatic groups. Tyrosine doublet The intensity ratio of tyrosine doublet at 857 and 830 cm"1 is used as a Raman indicator of hydrogen bonding by the tyrosine phenolic group. This ratio decreased after freeze-thawing for the commercial blend, AFGP (0.005 mg/g) and AFPI (0.005 and 0.01 mg/g) mince samples (Table 10). This could be explained by the lower frequency band (830 cm"1) becoming more intense as the phenolic hydroxyl oxygen is ionized or strongly hydrogen bonded to a negative charged acceptor such as a carboxylic ion (Siamwiza et al., 1975; Tu, 1986). The control mince was the only sample to show an increase in ratio upon freeze-thawing, indicative of Tyr residues being relatively exposed on the protein surface which can interact with water molecules as a hydrogen bond donor or acceptor (Tu, 1986). Tryptophan The band intensity at 757 cm"1 is attributed to indole-ring vibrations of tryptophan residues (Tu, 1986; Li-Chan, 1996 and Ogawa et a l , 1999). The control, AFGP (0.005 and 0.01 mg/g) and AFPI (0.005 mg/g) mince samples showed a decrease in the band's intensity upon freeze-thawing (Table 10). A decrease of intensity could be attributed to the exposure of hydrophobic Trp residues arising from a'change in the microenvironment 100 of the Trp side chain due to protein-protein interactions associated with changes generated upon frozen storage. In contrast, there was an increase in the intensity for commercial blend and AFPI (0.001 mg/g) samples. Buried Trp residues tend to exhibit a more intense band than those residues exposed to a polar environment (Tu, 1986; L i -Chan, 1996; Herrero et al., 2004). 101 B Unfrozen Freeze-thawed v ; v ; unfrozen. Freeze-thawed Y E O 0) N. E o 1600 1400 1600 1400 1000 800 Wavenumber cm"1 Y 0 n a. a 4A Control ft A, Commercial blend AFGP (0.005 mg/g) AFGP (0.01 mg/g) yv AFPI (0.005 mg/g) AFPI (0.01 mg/g) 1000 800 Figure 17. Raman spectra of unfrozen and freeze thaw mince, treated with AFP and with or without cryoprotectants. (A) Amide I (1590-1720 cm"1) and C-H bend (1400-1500 cm"1) regions; (B) Tyr (830-860 cm"1) and Trp (757 cm"1) regions. 102 Table 10. Raman band intensities for unfrozen and freeze-thawed mince treated with AFP and with or without cryoprotectants. Raman region (cm ) • r Band Intensity'". UF 8-FT Control 0.9 0.7 Commercial blend 0.8 1-1 Trp AFGP (0.005 mg/g) 0.8 0.7 (757 cm-1) AFGP (0.01 mg/g) 0.8 0.7 AFPI (0.005 mg/g) 0.7 0.3 AFPI (0.01 mg/g) 0.7 1.0 Control 0.8 1.6 Commercial blend 1.0 0.8 Tyr ratio AFGP (0.005 mg/g) 1.5 0.6 (857cm_i/830cm_i) AFGP (0.01 mg/g) 0.9 0.9 AFPI (0.005 mg/g) 1.0 0.7 AFPI (0.01 mg/g) 1.4 1.0 Control 1.3 1.1 C H 2 bend Commercial blend 0.8 1.3 (1450 cm1) AFGP (0.005 mg/g) 1.5 1.0 AFGP (0.01 mg/g) 1.1 0.8 AFPI (0.005 mg/g) 1.0 1.4 AFPI (0.01 mg/g) 0.9 1.4 Control 1.3 0.7"' Amide I Commercial blend 1.1 1.2 (1650 cm"1) AFGP (0.005 mg/g) 1.4 1.4 AFGP (0.01 mg/g) 1.2 0.7 AFPI (0.005 mg/g) 0.8 1.5 AFPI (0.01 mg/g) 1.1 1.5 'Normalized band intensity values (normalized to intensity of Phe at 1004 ± 2 cm"') 103 4.3. Natural actomyosin treated with antifreeze proteins compared to commercial cryoprotectants Calcium ATPase (Ca+ 2ATPase) activity of unfrozen N A M extracts is shown in Table 11. At 10 min, the activity ranged from 0.024 to 0.072 umol Pi/mg protein. It is not clear why there are differences in activity among the unfrozen samples. In the study by Boonsupthip and Lee (2003), N A M extracted from tilapia and N A M treated with sucrose + sorbitol or with different concentrations of AFP III, exhibited marked differences before being submitted to several freeze-thaw cycles. In some treatments, the activity retention was higher than 100%. Carvajal et al. (1999) suggested that these initial differences and excessive enzyme activity could involve unfolding of muscle proteins which is later followed by aggregation, and also the initial interactions of the proteins with different cryoprotectants. Upon freeze-thawing there was no detectable Ca + 2ATPase activity in any of the samples, including the commercial blend and those treated with AFP. These results are evidence of the denaturation of myosin leading to loss of the ATPase enzyme activity as muscle proteins were exposed to freeze-thawing conditions. In contrast, in the study performed by Boonsupthip and Lee (2003), Ca + 2ATPase activity increased with increasing concentration of type-Ill AFP. However, the AFP concentrations used by the researchers were much higher (10xl0 3, 50x103, 100xl0 3 mg/L N A M ) than those used in this study. In addition, based on the results presented by Boonsupthip and Lee (2003), the control (zero freeze-thaw cycles) for N A M had significantly higher Ca + 2ATPase activity compared to samples from the 3, 7 and 10 freeze-thaw cycles. Therefore, it is difficult to 104 assess just how well this type of AFP worked, considering the high concentrations of AFPIII used and the observed differences between the control and treated N A M samples. In terms of surface hydrophobicity, lower S 0 values for freeze-thawed samples were observed compared to their unfrozen counterparts (Table 11). The drop in S 0 was significant (P <0.05) for samples with AFGP and AFPI at 0.2 mg/L N A M . Sultanbawa and Li-Chan (2001) and Careche et al. (1998b) also noticed a sharp decrease in surface hydrophobicity of ling cod and hake N A M , respectively, following freezing which was accompanied by the formation of visually detectable aggregates. This is explained as a consequence of a balance between proteins remaining in solution with more exposed hydrophobicity and aggregates with less exposed hydrophobicity (Careche and Li-Chan, 1997). 105 Table 11. Ca ATPase activity and surface hydrophobicity (S0) of N A M extracts before and after 8 freeze-thaw cycles. Sample Ca-ATPase activity (umol Pi/mg protein at 10 min) UF 8-FT UF 8-FT Control 0.024 N A 1 400 a x 33 r x Commercial blend 0.035 N A 439 a x 303 a x AFGP (0.2 mg/L NAM) 0.072 N A 462 a x 281 a y AFPI (0.2 mg/L NAM) 0.044 N A 502 a x 201 a y . AFGP (5 mg/L NAM) 0.028 N A 380 a x 287 a x AFPI (5 mg/L NAM) 0.031 N A 416 a x 327 a x ; N A = no detectable activity (<0.001 umol Pi/mg protein at 10 min). Ca + /ATPase values represent mean of three replicates. Surface hydrophobicity (S0) values represent mean of two replicates. Values bearing different superscript letters (a, b) within each column are significantly different (P < 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P < 0.05). 106 4.4. Fish mince treated with AFP, commercial cryoprotectants or their blends 4.4.1. Expressible moisture The addition of 2% sucrose, 2% sorbitol and 0.3% phosphates to 0.05 mg/g AFPI (AFPI-PolPO) did not prevent drip loss as seen by the increase in % E M upon freeze-thawing, which was not significantly different (P>0.05) from the control sample. In contrast, % E M for the freeze-thawed AFGP-PolPO blend was similar to the sucrose + sorbitol (SuSo) and 8% trehalose and 0.3% PO (Treha-PO) treatments (Table 12). In addition, the % E M of the mince treated with AFGP-PolPO was better than the control, AFGP-PO and AFPI-PO treatments suggesting a positive effect related to the presence of sucrose, sorbitol and AFGP. This effect was not as evident for the AFPI-PolPO blend, which seemed to show an intermediate result between the control and polyol blends. Mince treated with Treha-PO had higher % E M upon freeze-thawing compared to its unfrozen counterpart; however, this value was not significantly different (P>0.05) from the freeze-thawed sample consisting of the commercial blend without phosphates (SuSo), which had the lowest % E M . The latter result indicated that sucrose and sorbitol are strong cryoprotectants even without the presence of phosphates in the polyol blend. In addition, observations upon freezing and during freeze-thawing showed that mince treatments containing AFP and phosphates (AFGP-PO and AFPI-PO) displayed a thick layer of ice crystals on the surface of the mince. Treatments with AFP containing sucrose, sorbitol and phosphates (AFGP-PolPO and AFPI-PolPO) showed, to a much lesser extent, the same ice crystal layer formation. This ice crystal layer was not observed on the control, SuSo, and Treha-PO treatments. These observations will be discussed in more detail in section 4.4.4. 107 Table 12. Expressible moisture (% EM) for unfrozen and freeze-thawed mince with A F P , commercial cryoprotectants or their blends. Sample1 "~ % EM UF2 8-FT3 Control 18 b x 34 c y (0.4) (1.8) SuSo 17 b x 2 1 a x (0.3) (2.5) AFGP-PO 19 b x 33 c y (0.4) (5.8) AFPI-PO 16 b x 3 2 cy (1.8) (1.4) AFGP-PolPO 12 a x 23 aby (0.2) (1.7) AFPI-PolPO 11 a x 31 bey (2.4) (2.3) Treha-PO 10 a x 22 a y (1.0) (1.5) 'Sample abbreviations as follows: Control - mince with no added cryoprotectants; SuSo = 4% sucrose + 4% sorbitol with no phosphates; AFGP-PO = 0.05 mg/g (0.005%) AFGP, fraction 6-8 with 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w); AFPI-PO = 0.05 mg/g (0.005%) AFPI with 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w); AFGP-PolPO = 0.05 mg/g (0.005%) AFGP, fraction 6-8 with 2% sucrose, 2% sorbitol, and 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w); AFPI-PolPO = 0.05 mg/g (0.005%) AFPI with 2% sucrose, 2% sorbitol, and 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w); Treha-PO = 8% trehalose with 0.3% phosphates (0.15% tripolyphosphate and 0.15% sodium pyrophosphate w/w); 2UF: unfrozen samples;38-FT cycles: 8 freeze-thaw cycles. Numbers in brackets indicate SD values from three replicates. Values bearing different superscript letters (a, b, c) within each column are significantly different (P < 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P < 0.05). 108 4.4.2. Protein extractability in salt and SDS-PAGE The decreased extractability of salt soluble proteins upon freeze-thawing was not prevented by the addition of polyol blends to mince samples with AFP (Table 13). A l l mince treatments had significant lower %SEP upon freeze-thawing (P<0.05). Nevertheless, freeze-thawed SuSo showed significantly higher retention of extractable proteins compared to all other treatments, followed by AFGP-PolPO, AFPI-PolPO and Treha-PO. In the case of the freeze-thawed mince with AFP and phosphates (AFGP-PO and AFPI-PO), protein denaturation occurred, and cryoprotection during freezing and freeze-thawed cycles was not imparted by these blends as seen by significantly lower %SEP values similar to the control (without cryoprotectants). These observations would indicate that rather than an actual synergistic effect between the AFP and polyols, it was the presence of sucrose and sorbitol that was responsible for the cryoprotection in the AFP-polyol blends. The SDS-PAGE profiles of salt extractable proteins from unfrozen mince showed the M H C band at 205 kDa (Figure 18A). Upon freeze-thawing, the M H C band decreased in intensity for the control, AFGP-PO and AFPI-PO treatments (Figure 18B lanes 1, 3 and 4), while for SuSo (Figure 18B lane 2), Treha-PO (Figure 18C lane 3), AFGP-PolPO (Figure 18C, lane 1), and AFPI-PolPO (Figure 18C, lane 2) the intensity of the M H C band did not recede upon freeze-thawing. The AFGP-PolPO treatment also shows a more intense M H C band (Figure 18C, lane 1) compared to AFPI-PolPO. These results are consistent with the differences noted on the % E M , where AFGP-PolPO had lower % E M than the AFPI-PolPO blend. 109 Table 13. Salt extractable proteins (% SEP) from unfrozen and freeze-thawed mince treated with AFP, commercial cryoprotectants or their blends. Sample % Salt extractable ; proteins ;j UF2 8-FT3 Control 65 a x 30 a y SuSo 79a x 56 c y AFGP-PO 70 a x 30 a y AFPI-PO 73 a x 31 a y AFGP-PolPO 73 a x 41 b y AFPI-PolPO 75 a x 44 b y Treha-PO 7 g a x 43 b y 'Sample abbreviations are described in Table 12. 2UF: unfrozen samples; 38-FT cycles: 8 freeze-thaw cycles. Salt extractable values represent mean of two replicates. Values bearing different superscript letters (a, b, c) within each column are significantly different (P< 0.05). Values bearing different superscript letters (x, y) within each row are significantly different (P< 0.05). 110 1 2 3 4 5 6 7 1 2 3 Figure 18. SDS-PAGE profiles for unfrozen (A) and freeze-thawed (B, C) mince treatments with AFP, commercial cryoprotectants or their blends. Sample abbreviations are described in Table 12. A . Lane 1: control mince, 2: SuSo, 3: AFGP-PO, 4: AFPI-PO, 5: AFGP-PolPO, 6: AFPI-PolPO, 7: Treha-PO. B. Lane 1: control mince, 2: SuSo, 3: AFGP-PO, 4: AFPI-PO. C. Lane 1: AFGP-PolPO, 2: AFPI-PolPO, 3: Treha-PO. I l l 4.4.3. FT-Raman Spectroscopy Due to increased fluorescence following freeze-thawing of the samples, the visible laser Raman spectra had poor signal to noise ratios. Therefore, freeze-thawed samples were analyzed by FT-Raman spectroscopy. Figure 19 shows the FT-Raman spectra at the 400-1800 cm"1 wavenumber region after 8 freeze-thaw cycles. Predominant bands can be seen for the tryptophan, backbone C-C, C-N stretch; C-H bend and amide I regions, which have been reported for ling cod proteins (Sultanbawa and Li-Chan, 2001). The intensities for the assigned bands are given in Table 14. Amide I region and CH stretching region The intensity and frequency of the amide I region (1650-1680 cm"1) was lowest for the freeze-thawed mince samples treated with AFGP-PolPO, followed by the control, AFGP-PO and AFPI-PO (Figure 20). In contrast, the highest band intensity was observed for the freeze-thawed SuSo followed by the AFPI-PolPO and Treha-PO mince samples. The strong intensity of this band has been attributed to proteins with high a-helix content (Careche et al., 1999, Krimm and Bandekar, 1986). An estimation of percent secondary structure was made from the amide I band using the RSAP program of Przybycien and Bailey (1989). The results showed the freeze-thawed control having an increase in random coil, while the AFPI-PO mince samples had an increase in P-sheet at the expense of a-helices (Figure 21). Mince treated with SuSo contained the highest proportion of a-helices after freeze-thawing, followed by AFGP-PolPO and Treha-PO. Changes in aliphatic residues after frozen storage can be shown by both the C-H stretching (2940 cm"1) and C-H bending (1450 cm"1) vibrations (Careche and Li-Chan, 112 1997). Upon freeze-thawing, the C-H bending band slightly shifted to a lower frequency of around 1448 + 1 cm"1 for all mince treatments (Figure 20). Sultanbawa and Li-Chan (2001) reported a decrease in frequency for ling cod natural actomyosin with cryoprotectants from 1462 (unfrozen) to 1458 cm"1 after freezing. Changes of this band may result from hydrophobic interactions of the aliphatic residues (Lippert et al., 1976). Amide III region Examination of the amide III region for unfrozen mince showed a clear band at 1267 cm"1, characteristic of a-helix. However, it is important to note that the region of 1267 cm"1 could also be overlapping with the region assigned for (3-structure and random coil (Barret et al., 1978). Helix to random coil or helix to P-sheet transitions are often accompanied by an increase in intensity and broadening around the amide III region (Li-Chan et al., 1994). In this study, changes in the secondary structure of the polypeptide backbone could be seen for the different treatments after freeze-thawing (Figure 22); the 1267 cm"1 peak shifted to 1270 + 1 cm"1 for all the treatments. Mince samples treated with AFGP-PO and Treha-PO had lower intensity compared to the other mince treatments (Table 14); however, these differences were not significant (P >0.05). Tryptophan residues The microenvironment of tryptophan (Trp) residues can be studied from the information displayed in Raman bands located at 760, 880, 1340 or 1363 cm"1 (Tu, 1986 and Li-Chan et al., 1994). In this study, the tryptophan band intensity at 760 was highest for freeze-thawed SuSo followed by AFPI-PolPO and Treha-PO, whereas the control had the lowest intensity (Table 14). The decrease in intensity observed for the control mince could indicate exposure of hydrophobic tryptophan residues upon protein denaturation. 113 As discussed in section 4.2.6, studies on hake fillets and muscle stored at -10 and -30°C (Careche et al., 1999 and Herrero et al., 2004), showed a marked decrease of peak intensity at 759 cm"1, indicating that the exposed tryptophan may play a role in the protein-protein interactions accompanying the detrimental changes generated upon frozen storage at those two temperatures. Tyrosine doublet The intensity ratio for the tyrosine doublet at 857 cm"1 and 830 cm"1 varied from 0.5 to 1.1 for the different mince samples following freeze-thawing (Table 14). According to Li-Chan et al. (1994) arid Tu (1986), i f the intensity I85o/830 tends to be high (0.9-1.45) it could be assumed that the tyrosine residue is exposed and its hydroxyl group can interact with water molecules as a hydrogen bond donor and acceptor. If the ratio is lower (between 0.7 and 1.0), a buried tyrosine residue is assumed which tends to act as a hydrogen bond donor. Finally, i f the ratio is as low as 0.3, strong H-bonding to a negative acceptor is shown. In this study, the freeze-thawed control, AFGP-PO and AFPI-PO treatments had the lowest ratios (0.6, 0.5 and 0.7, respectively), possibly from a change in the microenvironment of tyrosine residues with strong hydrogen bonding from the buried tyrosine. Treatments AFGP-PolPO and AFPI-PolPO had slightly higher, intermediate ratios (0.8). Conversely, SuSo and Treha-PO showed significantly (P <0.05) higher ratios (1.1 and 1.0, respectively), indicating that tyrosine residues were relatively exposed on the protein surface and could be acting as hydrogen bond donors and acceptors. 114 C-H stretching region The C-H stretching region (2800-3000 cm"1) shows hydrophobic groups of amino acids, peptides and proteins, with the strongest C-H band located near 2932 cm"1. SuSo showed significantly higher intensity (P <0.05), followed by the Treha-PO and AFPI-PolPO samples (Table 14). From Figure 23, it can be observed that the samples developed 2 small shoulders at ca. 2878 and 2970 cm"1. Other researchers working with ling cod actomyosin (Sultanbawa and Li-Chan, 2001) have shown similar observations where harsher frozen storage conditions gave a greater evidence of these shoulders. According to Careche et al. (1999), the 2880 cm"1 band is produced by the symmetrical CH2 stretching motion. Herrero et al. (2004) reported development of a shoulder near' 2850 cm"1, which could be interpreted in terms of interaction of lysine or arginine side chains with formaldehyde during frozen storage. O-H stretching region (3100-3600 cm') The band near 3200-3220 cm"1 reflects the O-H stretching vibrations of water (Li-Chan et al., 2002). In this study, the intensity of this band was lowest for the control and highest for the SuSo, followed by the AFGP-PO and Treha-PO samples (Table 14, Figure 23). Loss of water due to ice crystal formation and crystal growth pattern at freezing conditions has been reported by a decrease in the intensity of this band (Careche et al., 1999; Sultanbawa and Li-Chan, 2001). Herrero et al. (2005) also suggested that intensity changes in this region may be attributed to transfer of water to larger spatial domains during frozen storage. The ratio involving the intensity of the O-H stretching (3220 cm"1) band relative to the C-H stretching (2932 cm"1) intensity has been found to be sensitive to differences in freezing and frozen storage conditions, where a decrease 115 in the intensity ratio is correlated to loss of quality during frozen storage (Careche et al., 1999; Herrero et al., 2004). In this study, the ratio was significantly lower (P <0.05) for the control mince. In contrast, AFGP-PO showed the highest ratio (Table 14), which was not significantly different from the remaining samples. The highest ratio for AFGP-PO does not mean that the AFGP had less water losses. Instead, this behaviour could be attributed to the presence of higher proportions of unbound water in this sample. These observations are discussed in the following section, using differential scanning calorimetry. The intensity ratio reamained similar for the SuSo, Treha-PO, AFGP-PolPO and AFPI-PolPO mince samples. 116 Figure 19. FT-Raman spectra (400 -1800 cm"1 region) of freeze-thawed mince samples with AFP, commercial cryoprotectants or their blends. Sample abbreviations are described in Table 12. 117 Table 14. FT-Raman band intensities for freeze-thaw minces with AFP, commercial cryoprotectants or their blends. Raman region (cm?!) Sample 5, Band Intensity1 Control 0.7 (0.2) SuSo 0.9 (0.3) Trp AFGP-PO 0.8(0.1) (760) AFPI-PO 0.8 (0.2) AFGP-PolPO 0.8 (0.2) AFPI-PolPO 0.9 (0.1) Treha-PO 0.8 (0.2) Control 0.6bc (0.0) Tyr I ratio SuSo 1.18 (0.0) (857/830) AFGP-PO 0.5 b c (0.1) AFPI-PO 0.7 b c (0.0) AFGP-PolPO 0.8b(0.1) AFPI-PolPO 0.8b(0.1) Treha-PO 1.0a(0.1) Control L3 (u.2) SuSo 1.4 (0.3) Amide III AFGP-PO 1.2 (0.1) (1267) AFPI-PO 1.3 (0.3) AFGP-PolPO 1.5(0.1) AFPI-PolPO 1.5 (0.1) Treha-PO 1.2 (0.1) Control 4.3 (0.4) SuSo 5.1 (0.9) C H 2 bend AFGP-PO 4.3 (0.1) (1450) AFPI-PO 4.3 (0.4) AFGP-PolPO 4.3 (0.1) AFPI-PolPO 4.9 (0.6) Treha-PO 4.3 (0.3) Control 5.1 (0.5) SuSo 5.8 (0.3) Amide I AFGP-PO 5.2 (0.4) (1656) AFPI-PO 5.2 (0.5) AFGP-PolPO 5.0 (0.3) AFPI-PolPO 5.7 (0.7) Treha-PO 5.5 (0.2) 118 Table 14. FT-Raman band intensities cont... "Raman region (cm"1) Sample Band Intcnsil\ 1 Control 25 K (1.1) SuSo 31a(2.6) C-H AFGP-PO 23c(0.7) (2932) AFPI-PO 23c(1.3) AFGP-PolPO 24 b c ( l . l ) AFPI-PolPO 29 a b(3.5) Treha-PO 29 a b(0.4) Control 21 C(1.M SuSo 33a(4.5) O-H AFGP-PO 32a(2.2) (3220) AFPI-PO 24b c(2.1) AFGP-PolPO 25b c(1.2) AFPI-PolPO 29 a b(0.7) Treha-PO 31a(1.4) Control 0.8a(0.0) SuSo l . l b (0.1) OH/CH ratio AFGP-PO 1.4b(0.1) (3220/2392) AFPI-PO l . l b (0.1) AFGP-PolPO 1.0b (0.1) AFPI-PolPO 1.0b(0.1) Treha-PO l . l b (0.0) 'Normalized band intensity values are average of three spectra (normalized to intensity of Phe at 1004 ± 2 cm"1). Numbers in brackets indicate SD of three replicates. Values, within each Raman region, bearing different letters indicate significant differences (P<0.05). Absence of letters within each Raman region indicates no significant differences. 119 — Control SuSO - - A F G P - P O AFPI-PO 1800 1750 1700 1650 1600 1550 1500 1450 1400 W a v e n u m b e r c m -Figure 20. FT-Raman spectra showing amide I, CFb bending regions of freeze-thawed minces with AFP, commercial cryoprotectants or their blends. Sample abbreviations are described in Table 12. 120 • UF • FT control Suso AFGP- AFPI-PO PO AFGP- AFPI- Treha-PolPO PolPO PO 0.45 0.4 0.35 0.3 $ 0.25 % 0.2 si 0.15 0.1 0.05 0 • UF • FT control Suso AFGP-PO AFGP- AFPI- Treha-PolPO PolPO PO 1.2 1 0.8 § 0 . 6 H •a I 0.4 0.2 0 • UF • FT II trJDi control Suso AFGP- AFPI-PO AFGP- AFPI- Treha-PO PolPO PolPO PO Figure 21. Secondary structure fractions estimated from the amide I band for freeze-thawed minces with AFP, commercial cryoprotectants or their blends. Sample abbreviations are described in Table 1 2 . 1 2 1 1350 1300 1250 1200 1150 1100 1050 1000 Wavenumber cm- 1 Figure 22. FT-Raman spectra in the C-C, C - N stretch and amide III regions o f freeze-thawed mince with A F P , commercial cryoprotectants or their blends. Sample abbreviations are described in Table 12. 122 Control S u S O A F G P - P O A F P I - P O A F G P - P o l P O AFPI -Po lPO Treha-PO 2878 3400 3300 3200 3100 3000 2900 Wavenumber cm-Figure 23. FT-Raman spectra showing the C-H (2932 cm"1) and O-H (3220 cm"1) stretching regions. Sample abbreviations are described in Table 12. 123 4.4.4. Differential scanning calorimetry (DSC) of freeze-thawed mince DSC was used to evaluate the behavior of free and bound water in freeze thawed minced samples with and without AFP. The results indicated that AFGP-PO and AFPI-PO had the lowest antifreeze activity measured by the lower amount of unfrozen or bound water (0.19 and 0.22 g, respectively) compared to the control, SuSo, and Treha-PO treatments (Figure 24). AFP samples containing sucrose, sorbitol and phosphates (AFGP-PolPO and AFPI-PolPO), had similar amounts of unfrozen water compared to the control. The SuSo and Treha-PO treatments had the highest amount of unfrozen water (0.34 g) compared to the other treatments. These results may provide some insight on the visual observations of the ice crystal layer on the surface of the mince samples containing AFP. As mentioned previously, the surface of mince samples treated with AFP (0.05 mg/g) displayed a layer of ice crystals; from the DSC results, it seems that in order for this "external" layer of ice crystals to form, more water was present in the free or unbound form. More free water was therefore being frozen in these samples. Interestingly enough, both AFGP-PolPO and AFPI-PolPO were not significantly different from the control, SuSo and Treha-PO treatments, even though a thin layer of ice crystals was seen for these samples. This observation can be attributed to the water binding properties of the sugars present in these treatment blends. Control, SuSo and Treha-PO treatments did not show this ice crystal layer and, and from the DSC results, contained more unfrozen (bound) water compared to AFGP-PO and AFPI-PO treatments. 124 4.4.5. Differential scanning calorimetry (DSC) of fresh mince As described in the materials and methods section, fresh mince was subject to one full freeze thaw cycle in order to evaluate the antifreeze activity of AFP during the transition of fresh mince samples into a frozen stage. The results of antifreeze activity (g unfrozen water/ g water) for the fresh minces were similar to those observed for mince samples that had been submitted to 8 freeze thaw cycles (Figure 25). AFGP-PO and AFPI-PO had 0.22 and 0.24 g of unfrozen water, respectively. Similar to the results in Figure 24, all the remaining samples had significantly higher (P<0.05) amounts of unfrozen (bound) water. These results clearly indicate that the lack of antifreeze activity by the antifreeze porteins was not a result of previous freeze-thaw conditions, as the antifreeze activity did not improve when fresh samples were submitted to one freeze thaw cycle. 125 0.40 * 0.35 a > I 0.30 0.25 0) TO 0.20 « 0.15 c N o 0.10 3 0 1 0.05 0.00 ab X control I bc I ab ab SuSo AFGP-PO AFPI-PO AFGP-PolPO AFPI-PolPO Treha-PO Figure 24. DSC results showing antifreeze activity (g unfrozen water/g water) of mince samples following 8 freeze-thaw cycles. Columns and error bars represent mean value + standard deviation of four experiments. Columns bearing different superscript letters are significantly different (P < 0.05). . Sample abbreviations are described in Table 12. *Sample calculation: from equation (1), the antifreeze, activity (AF) of 8% trehalose + 0.3% PO (mean peak area = 8.75253, sample weight in DSC = 0.0534 g; water in sample = 0.755 g water/g sample; AHA — 217.1 ± 3.9 J/g water) was 0.34 + 0.01 g of unfrozen FkO/g water. 126 0.5 0.45 | 0.35 £ 0.3 u 13 0.25 c 0.2 N o 0.15 § o.i D) 0.05 0 control 1 a a SuSo AFGP-PO AFPI-PO AFGP-PolPO AFPI-PolPO Treha-PO Figure 25. DSC results showing antifreeze activity (g unfrozen water/g water) of fresh (not previously frozen) minces following one complete freeze-thaw cycle. Columns and error bars represent mean value + standard deviation of four experiments. Columns bearing different superscript letters are significantly different (P < 0.05). Sample abbreviations are described in Table 12. Sample calculation as described in Figure 24. 127 4.4.6. Photomicrographs and video-recording observations The microscopic observations of mince samples with antifreeze proteins and phosphates (AFGP-PO and AFPI-PO) show that when samples were removed from the -80°C storage and exposed to room temperature, formation of an ice crystal layer began to take place as the temperature on the surface of the sample approached -10°C (Figure 26). These observations indicate that the ice crystal formation caused by AFP can only take place at a certain temperature range as the water in the samples begins to thaw, in this case as the temperature of the sample approached -10°C. Figure 27 shows that neither the control nor SuSo treatments displayed this ice crystal formation even after 180 sec (surface temperature also at -10°C), whereas the AFPI-PolPO treatment clearly formed the ice layer within 180 seconds. From the literature, it is known that antifreeze proteins help polar fish survive under hypothermic conditions by inhibiting ice crystal growth, while in other biological systems these same proteins seem to have an opposite effect by promoting ice crystal formation. Larese et al. (1996) proposed that due to AFPI's amphipathic nature, binding of AFP to the ice crystal resulted in the creation of an exterior hydrophobic surface on the ice crystal. This surface could then allow the ice crystal to be in closer proximity to the plasma membrane, making it more favorable for water molecules in the membrane to interact with the ice crystal rather than remain bound to membrane phospholipids, thus facilitating intracellular nucleation. In the study by Larese et al. (1996), the incidence of intracellular ice formation in the presence of AFP (0.5 mg/mL) was highest at -10°C, and dramatically decreased as the temperature approached 0°C. 128 Figure 26. Photomicrographs (50 x magnification) of fish mince taken after removal from -80°C storage. Images represent fish mince with 0.05 mg/g AFGP and 0.3% phosphates (AFGP-PO) at (a) 30 sec; (b) 180 sec, and fish mince with 0.05 mg/g AFPI and 0.3% phosphates (AFPI-PO) at (c) 30 sec; (d) 180 sec. Temperature on the surface of the samples was -35 + 1°C at 30 sec, and -10.0 + 1°C at 180 sec. 129 Figure 27. Photomicrographs (50 x magnification) of fish mince taken after removal from -80°C storage. Images represent fish mince with no cryoprotectants (control) (a) at 30 sec and (b) at 180 sec; (c) SuSo at 30 sec and (d) 180 sec; (e) AFPI-PolPO at 30 sec and (f) at 180 sec. Temperature on the surface of the samples was -35 + 1°C at 30 sec, and -10.0 + 1°C at 180 sec. 130 Other studies have shown that AFP at concentrations above 5 mg/mL can increase cell destruction in suspensions (Koushafar and Rubinsky, 1997), in tissues in vitro (Koushafar et al., 1997), and in tissues in vivo (Pham et al., 1999). A study on rat cardiac explants showed that AFGPs provided no beneficial effect on cardiac viability during hypothermic or frozen preservation; instead, freezing in the presence of the AFGPs greatly increased the heart damage caused by freezing. After freezing storage at -1.4°C, the surface of the frozen AFGP treated hearts appeared grayish white in comparison to the pink to red colour of the frozen hearts not containing AFGPs (Wang et al., 1994). It was proposed that spicule ice formed by AFGPs managed to disrupt the sarcolemma membrane integrity of heart cells, releasing myoglobin from the AFGP treated hearts. A model proposed by Wang (2000) indicates that when the intensity of the AFP-ice complex and membrane interaction reaches a particular level, the AFP-ice complexes aggregate, thereby making the ice-like embryos larger and decreasing the surface free energy. This interaction facilitates the growth of ice and therefore reverses the ice-inhibition effect of AFP. Wang (2000) proposed that in other systems where aggregation does not take place, AFP will inhibit recrystallization and have a beneficial effect of low-temperature preservation. Payne et al. (1994) found that AFP and AFGP at 0.1 and 100 pg/mL were mildly cytotoxic to ram spermatozoa cooled to 5°C. The results were explained on the basis of aggregation-dissociation dynamics of AFP which could be concentration dependent. Dissociated forms (toxic) could exist at low concentrations, and are probably due to the amphiphilic nature of AFP and their interaction with water or cellular membranes. 131 4.5. Conformational changes of antifreeze protein solutions at subzero temperatures 4.5.1. Amino acid composition and mass measurements Table 15 shows that the AFGP (fraction 6-8) was composed primarily of three main amino acids: Ala (57.2%), Thr (24.7%) and Pro (15.6%). AFPI also contained a high percentage of Ala (48.1%) and Thr (9.4%), but contained many other amino acids including Asp (10.7%), Leu (5.7%), Glu (4.9%), Ser (4.3%), Lys (3.1%), Arg (2.6%) and Val (2.4%). MALDI-TOF mass spectrometry showed molecular masses in the range of 1034 to 4623 Da for AFGP (Appendix I), consistent with previous mass spectrometry results reported for AFGP fractions 6-8. Wu et al. (2001) reported that rock cod (Gadus ogac) synthesize antifreeze glycoproteins ranging from 2646 to 24,200 Da with size classes "II," corresponding to AFGP-6 (3685 Da), "II 2" to AFGP-7 (3255-3498 Da) and "II 3" to AFGP-8 (2646 Da). AFGP fractions 3-5 had a molecular mass ranging from 13,686 to 24,196 Da which were not observed in the present study. AFPI showed a molecular mass ranging from 3200 to 3260 Da (Appendix II) in accordance to what has been previously reported for this type of antifreeze protein (Hew et al., 1980; Loewen et al., 1999). 4.5.2. Raman spectroscopy The Raman spectra for AFGP and AFPI in the 400-1800 cm"1 region can be seen in Figure 28. A summary of the peak assignments and changes in relative peak intensities as a function of temperature is given in Table 16, and the results for different regions of the Raman spectra are described below. A summary of the statistical analysis for Table 16 can be found in Appendix III. 132 Table 15. Amino acid percentage for AFGP and AFPI. (mol %) amino acid" AFPI AFGP Asp 10.7 0.6 G l u 4.9 0.3 Ser 4.3 0.3 Gly 2.3 0.3 His 0.7 0.0 A r g 2.6 0.2 Thr 9.4 24.7 A l a 48.1 57.2 Pro 1.8 15.6 Tyr 0.9 0.0 Val 2.4 0.4 Met 0.4 0.0 He 1.2 0.1 Leu 5.7 0.2 Phe 1.3 0.1 Lys 3.1 0.3 aCys and Trp were not determined. 1600 1400 1200 1000 800 600 Wavenumber crrr1 1600 1400 1200 1000 800 600 Wavenumber crrr1 Figure 28. Raman spectra of (A) AFGP and (B) AFPI solutions held at four different temperatures. From top to bottom: 22°C, -0.5°C, -1.8°C and -4.0°C. 134 Table 16. Raman band intensities in the 400-3400 cm"1 region for AFGP and AFPI solutions held at different temperatures. Relative peak intensity (wavenumber ± 2 cm") Peak assignment AFGP AFPI 22°C -0.5°C -1.8°C -4.0°C 22°C -0.5°C -1,8°C -4.0°C NH 3 + torsions of L-alanine and C-C-C bending vibrations 0.8 (463) 0.6 (460) 0.6 (460) 0.8 (463) C0 2 " rocking mode of L-alanine and C-C-C bending vibrations 0.7 (534) 1.3 (530) 1.0 (530) 1.2 (530) C-CH 3 stretch of the polypeptide 0.7 (885) 0.8 (886) 0.6 (889) 0.6 (887) C-N, C-N-C stretching mode of alanine and glutamic acid ' * ' " * * " 1.0 (910) 1.9 (913) (909) T.7 (909) C-C stretch of the proline ring 0.8 (929) 0.7 (929) 0.5 (930) 0.4™ (930) N-acetyl chromophore of the j 1.3 disaccharide J (984) , J 1.1 (980) 1.0 (980) 0.9 (980) C-C, C-N stretching and carbohydrate C-H, COH vibrations 2.2 (1070) 1.9 (1071) 1.8 (1073) 2.3 (1073) 1.0 (1060) 0.8sh (1078) 0.6 (1059) 0.9sh (1078) 1.1 (1063) 1.1sh (1080) 1.3 (1063) 1.5sh (1083) C-C, C H 3 vibrations of alanine 1.4sh (1105) 1.6sh (1105) 1.2sh (1101) 1.7sh (1105) 0.9 (1107) 1.2 (1108) 1.0 (1108) 1.6 (1108) Amide III and carbohydrate COH coordinate 0.8 (1253) 1.0 (1256) » , _ _ _ (1253) 0.9 (1260) 0.5 (1276) 0.8 (1262) 0.7 (1267) 0.8 (1272) Amide I 0.8 (1646) 14 (1645) (1641) 1.6 (1645) Extended structure 0.9 (1674) 0.8 (1674) 0.5 (1674) 0.6 (1672) - 0.7sh (1685) 0.5sh (1686) 0.6sh (1684) Unordered structure 0.9 (1645) 0.7 (1640) 0.4 (1647) 0.7 (1647) Polyproline type II helix 0.8 (1624) 0.7 (1624) 0.5 (1618) 0.6 (1616) 0.9sh (1620) 0.5sh (1620) 0.9sh (1618) C-H stretch, aliphatic 0.9sh (2992) 0.9sh (2993) 0.7sh (2992) 0.7sh (2996) 0.6sh (2993) 0.8sh (2993) 0.7sh (2994) 0.5sh (2993) 2.9 (2946) 3.3 (2946) 2.6 (2945) 2.6 (2945) 2.1 (2945) 2.4 (2946) 2.0 (2946) 2.3 (2946) 1.0sh (2888) 1.0sh (2885) 0.8sh (2891) 0.9sh (2887) 0.9sh (2890) 0.9sh (2888) 0.8sh (2887) 1.0sh (2886) O-H stretching 5.7 (3227) 6.8 (3210) 5.2 (3206) 5.1 (3209) 3.2 (3225) 4.7 (3201) 3.8 (3202) 4.3 (3200) Numbers in parenthesis below the intensity refer to wavenumbers ± 2 cm" ; sh = shoulder. Please see Appendix III for upper and lower confidence limits. 135 Amide I region As described in sections 4.2.6. and 4.4.3., this region is assigned to the intense band falling in the 1600 to 1700 cm"1 range, and involves C=0 stretching and to a lesser degree, C-N stretching and C a - C - N bending, and N - H in-plane bending of peptide groups (Tu, 1986; Krimm and Bandekar, 1986; Herrero et al., 2004). Results for the amide I region of AFGP at 22°C showed a band composed of a series of small peaks in close proximity. The lack of a sharp, clear band at 1650 cm"1, normally attributed to a-helix conformation, is consistent with literature reports showing the lack of or minor presence of a-helix conformation in AFGPs (Bush et al., 1984; Lane et al., 1998; Lane et al., 2000). Instead, the AFGP solution showed small peaks near 1620 cm"1 and at 1674 cm"1, attributed to polyproline type II helix and extended P-structures, respectively (Tomimatsu et al., 1976, Lane et al., 2000, Tsvetkova et al., 2002). The intensity of this atypical band decreased drastically at subzero temperatures, especially as the temperature was lowered to -1.8 and -4.0°C (Figure 29A). Tomimatsu et al. (1976) also noted a large decrease in the amide I intensity of AFGP solutions; however, this intensity increased when rapid freezing was applied. Tsvetkova et al. (2002) reported that the predominant feature of the FTIR-deconvolved amide I band of AFGP was an intense peak at 1642 cm"1, assigned to unordered conformation, and two additional bands at 1618 and 1672 cm"1 (polyproline II and extended (3-structure, respectively); the intensity of the amide I band was lower at -14°C than at 3°C (Tsvetkova et al., 2002). Similar observations could be seen in this study, where the AFGP solutions held at the lowest subzero temperatures (-1.8 and -4.0°C) had the lowest intensity in this region. FTIR spectra of two pentapeptides N A c A A T A A , N A c A A P A A and of AFGP solutions have been 136 reported to show a substantial peak at 1620 cm"1 for the spectra of AFGP and the pentapeptide N A c A A P A A , whereas the peak intensity was much lower at this wavenumber for AFGP 1-5 and the pentapeptide N A c A A T A A which lack proline (Lane et al., 1998). Lane and coworkers (Lane et al., 1998, Lane et al., 2000) described more restricted motions and a restricted conformational space accessible to the proline-containing AFGP compared to the proline-free AFGP 1 -5. Figure 29A also shows a peak ca. 1638 cm"1 for the AFGP solution held at 22°C. This peak has been attributed to the N-acetyl chromophore of the disacccharide moiety (Tomimatsu et al., 1976, Drewes and Rowlen, 1993). The peak was no longer visible at the lower holding temperatures, suggesting involvement of the disaccharide moiety in the antifreeze properties at subzero temperatures. Several studies have indicated the importance of the disaccharide in the activity of the AFGPs, as removal of the disaccharide results in loss of activity. These studies are summarized in Harding et al. (2003), and indicate the requirement for at least some of the hydroxyl groups from the sugar for activity of the protein. In contrast to AFGP, solutions of AFPI showed a band near 1645 cm"1 in the amide I region typical of an a-helix structure, characteristic for this type of antifreeze protein (Feeney and Yeh, 1993; Ananthanarayanan and Hew, 1977). It is important to note that this band was at a lower frequency from the typical 1650 cm"1 band allocated to a-helix in the amide I region. According to Surewicz et al. (1993) for some proteins that are largely a-helical, the major amide I band is shifted to wavenumbers below 1650 cm"1; this shift is caused by unusual amide-solvent interactions in these proteins or from strong hydrogen bonding. When the temperature of the AFPI solution was lowered, a higher 137 peak intensity was observed (Table 16, Figure 29B). Studies on AFPI solutions using N M R and CD have demonstrated that the protein retained its a-helical structure when supercooled; in fact, the protein became more a-helical as the temperature decreased from 25°C to -3.0°C (Gronwald et al., 1996, Graether et a l , 2001). A similar observation can be deduced from our study, where the amide I band acquired a sharper peak as the temperature of the protein solution was lowered. However, it was at -0.5 and -4.0°C that the solution had the highest peak intensities and area, whereas the solution held at -1.8°C seemed to be an intermediate stage in the strengthening of the a-helix structure. Comparing the amide I region of the solution held at 22°C with the subzero solutions shows that as the temperature was lowered, two small shoulders developed on the amide I band (Figures 28 and 29B), which could arise from (3-sheet structures (Dong et al., 1990). The amide I band of the -0.5°C supercooled solution shows one shoulder at 1685 cm"1 and one at 1620 cm"1. When the AFPI solution was cooled from 22°C to -1.8°C, these shoulders decreased in intensity (Table 16). However, cooling of the AFPI solution from 22°C to -4.0°C resulted in more pronounced shoulders with intensities similar to the -0.5°C solution. Graether and colleagues (2003) showed that type I AFP was able to maintain an a-helix conformation following freezing and freeze-thawing; nevertheless it also showed some parallel P-sheet structure. This unique behaviour can be explained by the fact that although AFPI is rich in alanine, which allows it to remain monomeric and a-helical, the alignment of the protein to the ice surface could promote an ordered aggregation of neighbouring molecules into the parallel P-sheet structure (Graether et al., 2003; Laws et al, 2001). In addition, Li-Chan and Qin (1998) suggested that hydrophobic 138 forces dominate the changes, such as aggregation, that result in P-sheet content. The authors indicate that the weaker strength of water hydration to P-sheet than to a-helix structures, may play a role in the changes in water-water, water-protein, and protein-protein interactions that favour aggregate network formation. However, the P-structures in such aggregates may differ from those found in native globular molecules. 139 1700 1600 1500 1400 1300 W a v e n u m b e r c n r 1 Figure 29. Raman spectra of (A) AFGP and (B) AFPI solutions in the amide I and amide III regions. Temperatures are in °C. 140 Amide III region This region, located in the 1230 to 1310 cm"1 range, involves N - H in plane bending and C-N stretching as well as contributions from C a - C stretching and C=0 in plane bending motions arising from the peptide group (Careche et al., 2002). AFGP generally showed a broad band in this region (Figures 28 and 30A). The strong band centered at 1253 cm"1 indicated an extended conformation for AFGP solution at 22°C; lower intensity of this band at -1.8°C could suggest strong hydrogen bonding at this temperature. A strong COH coordinate from the carbohydrate moiety (Tomimatsu et al., 1976) was indicated when the AFGP solutions were held at -0.5 and -4.0°C, as the band at these two temperatures was sharper and centered at a higher frequency (Table 16). In contrast, the AFPI solutions do not show a clear sharp amide III region. The solution held at 22°C showed a distorted band with a peak at 1276 cm"1 and shoulder at 1256 cm"1 (Figures 28 and 30B). Li-Chan et al. (1994) explain that the intensity of the amide III band for a-helix structure is usually weak or moderate, and is based in a region which overlaps with the region for (3-turns. In the present study, when the solutions were held at subzero temperatures, the amide III region shifted to a lower frequency, corroborating the strengthening of the a-helix structure at these temperatures. 1050 to 1150 cm'1 region AFGP solutions showed a strong band centered at 1070 cm"1, corresponding to the bending vibration of C(l)-H and COH in carbohydrates and the backbone C-C, C-N streching vibrations of proteins (Paradkar and Irudayaraj, 2001). This sharp, intense band corroborates, as previously discussed, the involvement of the carbohydrate moiety most likely via hydroxyl groups (Figures 28 and 30A). 141 1200 1100 1000 900 800 Wavenumber crrr1 1200 1100 1000 900 800 Wavenumber crrr1 Figure 30. Raman spectra of (A) AFGP and (B) AFPI solutions in the 750-1200cm region. Temperatures are in °C. 142 The 1070 cm"1 band intensity decreased slightly but not significantly (P >0.05) when the solution was held at -0.5 and -1.8°C, increasing only for the solution kept at -4.0°C (Table 16, Appendix III). The solutions also showed a small shoulder ca. 1105 cm"1. This shoulder increased in intensity for the AFGP solutions held at -0.5 and -4.0°C, and was lowest in both frequency (1101 cm"1) and intensity for the -1.8°C solution. Contrary to the small shoulder seen at 1105 cm"1 for AFGP, the AFPI solutions showed a strong peak at 1107 cm"1 (Figures 28 and 30B), most likely arising from the C-C, CH3 vibrations of alanine (Spiro, 1987). This peak had an intensity of 0.9 for the AFPI solution held at 22°C; at -4.0°C the peak intensity increased significantly (P <0.05), suggesting some involvement of methyl groups believed to be important in the interaction of AFP with ice (Chao, et al., 1997). The differences in the C-C, CH3 vibrations of alanine (at 1105 cm"1) between the AFGP and AFPI solutions, and the fact that this band is represented by a shoulder in AFGP rather than the strong peak seen for AFPI, suggests that although alanine is present in high amounts in both AFP, the behaviour of alanine seems to be different when the two AFP are exposed to subzero temperatures. It is likely that the pattern by which alanine interacts with water and subsequent ice-crystals is different for each type of AFP. Further investigation will be required to determine how the alanine residues in the two proteins are implicated in the water/ice crystal formation. A strong band from backbone C-C, C-N streching vibrations was also seen in the 1030-1090 cm"1 region of the AFPI solutions. The solution held at 22°C showed a major band at 1060 cm"1 and a smaller shoulder at 1078 cm"1. When the AFPI solution was held at -0.5°C, the major band significantly decreased (P <0.05) in intensity (Table 16, Figure 143 30B). The band also increased in intensity (P <0.05) and shifted to 1063 cm"1 for the AFPI solutions held at -1.8 and -4.0°C, while the shoulder also appeared to increase in both intenstiy and frequency. 800 to 1000 cm1 region The AFGP solutions showed a series of bands centered at 885, 929 and 984 cm"1 (Figure 28 and 30A). The band at 984 cm"1 representing the N-acetyl chromophore of the disaccharide (Tomimatsu et al., 1976), was visible at all 4 temperatures; however, the intensity decreased (P <0.05) as the temperature was lowered to -1.8 and -4.0°C, supporting the observations for the 1638 cm"1 peak, which also reveal the involvement of the disaccharide moiety at subzero temperatures. The peak at 929 cm"1 has been attributed to C-C stretch of the proline ring (Marquardt and Wold, 2004). In this study, the intensity of this peak significantly decreased (P <0.05) at the lower temperatures. The 885 cm"1 band has been attributed to C - C H 3 stretch of the polypeptide (Tomimatsu et al., 1976). The AFGP solution held at -0.5°C showed the highest band intensity; this intensity decreased as the solutions were held at -1.8 and -4.0°C (Table 16). AFPI at all four holding temperatures showed a strong band ca. 910 cm"1 (Figures 28 and 30B). The solution held at -0.5°C showed the highest peak intensity compared to the -1.8°C and -4.0°C solutions (Table 16). Nevertheless, the intensity was higher (P <0.05) for all three subzero temperatures than at 22°C. Other studies (Spiro, 1987; Shurvell and Bergin, 1989) have shown a strong single band at 915 cm"1 arising from C-N-C stretching mode of L(+)-glutamic acid, and a strong band ca. 910 cm"1 belonging to the C-N, C-N-C skeletal-stretch mode from alanine. Considering that AFPI has both alanine and glutamic acid, it is possible that this band is reflecting the vibrations of both amino 144 acids, and in the case of the skeletal mode vibrations of alanine, the strengthening of the a-helix at subzero temperatures. 400 to 600 cm'1 region AFGP and AFPI solutions show distinctive bands at 465 cm"1 and 534 cm"1, respectively. Raman spectra of L-alanine at room temperature has shown sharp, intense peaks at 470 cm"1 and 525 cm"1 attributed to the NH3 + torsions and CO2" rocking mode, respectively (Vik et al., 2005). The Raman spectra of the AFPs in this region therefore may be attributed to N - and C-terminal alanine residues of AFGP and AFPI, respectively. AFGP solutions showed a slight but not significant (P >0.05) decrease in intensity of the 465 cm"1 peak when the solution was held at -0.5 and -1.8°C, whereas the -4.0°C solution seemed to have a broader band with similar intensity as the 22°C solution (Figure 28A). The AFPI solutions showed changes in the intensity of the band situated ca. 530 cm"1. A relatively low intensity was observed at 22°C; when the AFPI solutions were held at the subzero temperatures this band became sharper and significantly higher (P <0.05) in intensity (Table 16, Figure 28B). C-H stretching region (2800-3000 cm'1) Hydrophobic groups of proteins, peptides and amino acids can exhibit symmetric CH2 stretching (2842-2859 cm"1), and C-H stretching (2800-3100 cm"1) vibrations in this region. Bands near the 2874-2897 cm"1 region are assigned to C H 3 symmetrical stretching and R3C-H stretching bands of aliphatic amino acids. Vibrations related to C H 3 asymmetrical stretching and a-C-H stretching are located at 2956-2977 cm"1 and 2980-2989 cm"1 respectively, while the =C-H stretching bands of aromatic amino acids can be found near 3061-3068 cm"1 (Howell et al., 1999). In this study, AFGP solutions 145 showed three strong bands centered near 2994, 2946 and 2888 cm"1 (Table 16, Figure 31), which may be arising from side chain C-H stretching vibrations of alanine and threonine, both present in high percentages in this protein. Even though these three sharp bands were present in the spectra collected at all four temperatures, the intensity for the -0.5°C solution increased compared to the AFGP at the other holding temperatures. However, this increase was not significant (P >0.05). AFPI also showed three bands in the 2800-3100 cm"1 region. The bands at 2994 and 2888 cm"1 were not as sharp as those seen in the AFGP solutions. The C-H stretching band (2946 cm"1) from the AFPI solution held at -0.5°C was also the highest in intensity, but not significantly (Figure 32). Increased intensity in the C-H stretch region has been attributed to the hydrophobic side chains from the aliphatic amino acid residues of the protein being exposed to the aqueous environment (Bouraoui et al.,1997; Howell et al., 1999; Herrero et al., 2004). For both types of AFPs, these results could indicate strong hydrophobic interactions from the aliphatic amino acids taking place, in particular at -0.5°C. O-H stretching region (3100-3600 cm'1) The band centered at 3200 cm"1 corresponds to the quasi-crystalline component of the O-H stretching band of water (Li-Chan et al., 2002). This band's intensity increased when the AFGP solution was held at -0.5°C (Figure 31), or when AFPI solutions were held at any of the three subzero temperatures, with the -0.5°C solution being the highest (P <0.05) of the three subzero temperatures (Figure 32, Table 16). An interesting observation was the shift to a lower frequency seen for all subzero solutions from both antifreeze protein solutions at all subzero temperatures. This shift of the O-H stretching vibrations to a lower frequency could be explained by stronger hydrogen bonding as the 146 antifreeze proteins interact with water and begin to modify ice crystal structure. A general rule for hydrogen bonds is that these interactions are reflected in the O-H stretching region by a shift to a lower frequency, whereas free water molecules show a band at a higher frequency (3666 cm"1) (Mathlouthi et al., 1996). Similarly, it has been suggested that the 3250 and 3450 cm"1 bands are associated with O-H vibrations in the hydrogen-bonded network of water molecules, and the 3600 cm"1 band is associated to nonhydrogen bonded O-H vibrations (Walrafen and Chu, 1995). In contrast to the AFP solutions, the buffer alone (without protein) was frozen when held at -0.5°C, and diminution of the water band ca. 3200 cm"1 was observed in the Raman spectrum of the buffer at this temperature. The intensity ratio of 0-H 3 2oo/C-H 2946 bands increased for AFGP and AFPI from 1.9 and 1.5 at 22°C respectively, to 2.0-2.1 and 1.8-1.9, respectively at the three subzero temperatures (Table 17). This ratio has been used for estimating the harshness of freezing/frozen storage conditions on hake muscle (Careche et al., 1999), as changes in the O-H stretching band of FT-Raman spectra of frozen fillets were consistent with different water losses expected due to ice formation and ice crystal growth, and indicated by a low O-H/C-H ratio. In this study, the increase in O-H/C-H ratio of both AFP solutions, as the solutions approached subzero temperatures, may be related to the observation of smaller, spicule-like ice crystal formation at these supercooled temperatures. Perhaps the ice crystal structure influences the nature of the O-H vibrations. 147 -0.5 3050 3000 2950 2900 2850 2800 2750 W a v e n u m b e r c m - 1 3300 3200 3100 " 3 0 0 0 2 9 0 0 2800 Wavenumber cm"1 Figure 31. Raman spectra of A F G P solutions in the 2800-3400 cm"1 region. Temperatures are in °C. 148 3 0 0 0 2 9 5 0 2 9 0 0 2 8 5 0 Wavenumber cm-1 Table 17. Intensity ratio (OH3220/CH2932) for AFGP and AFPI solutions. OH3220/CH2932 Temp°C AFGP VI PI 22~0 ~ ~ 1.9 L 5 ~ -0.5 2.1 1.9 -1.8 2.0 1.8 -4.0 2.0 1.8 150 5. SUMMARY AND CONCLUSIONS The objective of this thesis was to evaluate the possibility of using fish antifreeze proteins as a possible alternative to the commercial cryoprotectants (sucrose, sorbitol and phosphates) commonly used for minced fish and surimi. Antifreeze proteins were selected for this study based on their reported effect on preventing ice crystal growth and recrystallization. These proteins have been shown to lower the freezing point of substances without affecting the melting point. Their unique properties have prompted their evaluation for several applications, including organ transplant, oocytes and embryo preservation and some food applications such as ice cream and frozen or chilled meat. The range of AFP concentrations used in this study was based on those reported in several studies (including food applications) discussed in the literature review, in which the AFP prevented ice crystal growth and/or recrystallization. The possibility of incorporating AFP using spraying or soaking techniques did not prove to be feasible based on the preliminary tests using commercial cryoprotectants. The results showed that soaking or spraying fillets with cryoprotectants did not maintain the water binding capacity of fish fillets and did not decrease the drip losses associated with freeze-thawing cycles, perhaps in part due to the low penetration of the solutions into the samples. The incorporation of different concentrations of AFGP or AFPI (0.005-0.5 mg/g) into ling cod mince did not maintain protein functionality and stability following freeze-thaw abuse, compared to the effect observed for the commercial cryoprotectant. Upon freeze thawing, the amount of expressible moisture and cook loss was significantly higher for untreated (control) mince and mince samples treated with AFP, compared to 151 the commercial cryoprotectant. Myofibrillar proteins also became significantly less extractable in salt solution for untreated (control) mince and mince samples treated with AFP, compared to the commercial cryoprotectant. Blends consisting of antifreeze proteins and commercial cryoprotectants showed some improvement in preventing protein functionality changes of ling cod mince upon freeze-thaw abuse. However, based on the results obtained for expressible moisture and protein extractability in salt, this improvement was likely imparted by the presence of the sucrose and sorbitol rather than the AFP. The treatment consisting of sucrose and sorbitol without phosphates (SuSo) had the lowest drip loss (% expressible moisture) compared to blends of AFP and commercial cryoprotectants, indicating that sucrose and sorbitol alone are able to act as cryoprotectants even without the presence of phosphates. Mince treated with AFP and phosphates resulted in poor protein functionality following freeze-thaw abuse. Rather than a synergistic effect between AFP and phosphates, the results indicate a decrease in cryoprotection from the commercial cryoprotectants in the presence of AFP. This "opposite" effect (lack of cryoprotection) observed by the presence of AFP can be correlated with the massive ice crystal growth visible in mince samples treated with these AFP. Mince samples with varying concentrations of AFP promoted or induced ice crystal formation, creating a visible "frost" on the surface of the mince. Differential scanning calorimetry results showed that mince samples containing AFP had significantly lower amounts of non-freezable (bound) water, compared to the untreated (control) mince and minces treated with sucrose, sorbitol or blends consisting of sucrose, sorbitol, phosphates and AFP. A relationship can be seen between the visual frost formation and the higher amount of freezable (unbound) water in these samples. 152 The application of AFP was also evaluated using natural actomyosin (NAM) extracted from ling cod muscle, instead of mince, in the light of promisisng results obtained by Boonsupthip and Lee (2003) who studied the effect of Type III antifreeze protein (AFPIII) on tilapia actomyosin following freeze-thaw abuse. However, the AFP (AFPI and AFGP) concentrations used in this research (0.02 and 5 mg/L N A M ) failed to prevent the loss of Ca-ATPase activity following 8 freeze-thaw cycles. Boonsupthip and Lee (2003) not only used a different type of AFP (AFPIII), but also worked with much higher concentrations (10xl0 3 , 50x103 and 100xl0 3 mg/L N A M ) than those used in the present research. In addition to exploring potential applications of AFP through studying the effects of AFP on functional properties of frozen food systems such as fish mince and N A M , conformational changes of AFP in solution were studied at three subzero temperatures. This study was established in order to learn more about the behaviour of the AFP when held at temperatures closer to those encountered by cold water fish. The results obtained from this study were able to provide some information on the type of interactions involved between AFP and ice crystals, as well as the effect of the AFP's secondary structure on the ice-binding mechanism. For example, at subzero temperatures, especially at -0.5°C, both antifreeze proteins (AFGP and AFPI) showed marked conformational changes. The amide I region of AFGP had several small peaks near 1620 and 1674 cm"1 attributed to polyproline type-II helix and extended/unordered P-structures, respectively. In contrast, solutions of AFPI showed a band near 1645 cm"1 in the amide I region typical of an a-helix structure, characteristic for this type of antifreeze protein. AFPI also showed strengthening of the a-helix with some P-sheet-like structures at subzero 153 temperatures, indicating that the protein may remain folded and relatively rigid at these low temperatures. The NH3 + torsions of N-terminal alanine in AFGP decreased in intensity at subzero temperatures, whereas the CO2" rocking mode of C-terminal alanine was intensified for AFPI solutions held at subzero temperatures. Finally, increased intensity in the O-H stretching band of water at -0.5°C for AFGP and AFPI, and the shift to a lower frequency at all subzero temperatures indicate possible changes in ice-crystal structure, including strong hydrogen-bonding, and can provide future insight on the affinity of these proteins with water and ice crystals. In the context of the hypotheses stated for this research (section 1.1), the following general conclusions can be made. 1) Different concentrations of antifreeze proteins were significantly less effective than the commercial blend in maintaining ling cod mince protein functionally and stability upon freeze thaw abuse. Blends of antifreeze proteins and commercial cryoprotectants improved some of the protein functionality in mince, but were not as effective as the commercial blend alone in decreasing protein functionality changes of ling cod mince upon freeze thaw abuse. Percent salt soluble protein, SDS-PAGE, chemical and physical analysis showed that untreated mince (control) and mince samples treated with AFP formed protein aggregates, had decreased protein functionality and increased textural hardness upon freeze-thaw abuse. FT-Raman spectral profiles of untreated mince (control) and mince samples treated with AFP showed increased proportion of (3-sheet and random coil structures at the expense of a-helices. 154 2) Antifreeze proteins (0.02 and 0.5 mg/L N A M ) failed to stabilize natural actomyosin extracted from ling cod muscle. Ca-ATPase activity was lost following freeze-thaw abuse. 3) Antifreeze glycoprotein (AFGP) and Type I antifreeze protein (AFPI) solutions clearly exhibited different structural and conformational changes when held at subzero temperatures. In summary, whether this was beneficial or not, it is evident that AFP did have some effect apart from binding to ice crystals and inhibiting crystallization. AFP at concentrations ranging from 0.01 to 0.50 mg/g fish mince caused an opposite effect, compared to the commercial cryoprotectants studied, inducing ice crystal formation and failing to prevent protein denaturation during freeze-thaw abuse. The behavior of these proteins may be explained by literature findings in which AFP • have been shown to induce intracellular ice crystal formation and cell damage. The exact reasons for this dual effect of AFP have not yet been determined. However, it is clear from the results presented in this research that AFP at the concentrations used in this research cause this opposite effect when applied to fish mince, which contradicts some of the previous results reported for meat by Payne et al. (1994). Based on the results from the study, in which the conformational changes of AFP solutions at 0.1 g/mL (w/v) and held at subzero temperatures were evaluated, it is clear that AFP go through conformational changes at temperatures ranging from -0.5 to -4.0°C and in fact, at these temperatures the AFP solutions seemed to indicate an inhibition of ice crystal formation or changes in the ice crystal structure. Unfortunately, storing fish mince at these subzero temperatures is not a 155 common practice and would not be an optimum form of preservation as it would lead to other forms of deterioration (biochemical and microbiological). Nevertheless, it would be useful to consider these subzero temperatures perhaps for other modes of food preservation such as chilled, rather than frozen, meat products. Future work should also explore the application of AFP in other types of food systems, such as dispersed/colloidal systems and/or systems that can benefit from storage at subzero temperatures. In addition, factors such as the rate of cooling and the holding time at subzero temperatures prior to freezing at lower temperatures (i.e. -18°C) should be considered. The information obtained from this research could also be useful in future study applications of AFP in situations where intense ice crystallization formation will be desired or applications such as chemical adjuvants to cryosurgery. Finally, evaluation of other types of antifreeze proteins (types II, III and IV) or a combination (blend) of different types of antifreeze proteins to create 'AFP blends' should be explored. In this case, the application of an experimental design such as response surface methodology (RSM) would be useful for establishing formula optimizations. 156 References Alizadeh-Pasdar, N . , Li-Chan, E., and Nakai, S. 2004. FT-Raman spectroscopy, fluorescent probe, and solvent accessibility study of egg and milk proteins. J. Agric. Food Chem., 52(16): 5277-5283. Ananthanarayanan, V.S., and Hew, C L . 1977. Structural studies on the freezing point-depressing protein of the winter flounder Pseudopleuronectes americanus. Biochem. Biophys. Res. Commun., 74: 685-689. Ang, J.F. and Hultin, H.O. 1989. Denaturation of cod myosin during freezing after modification with formaldehyde. J. Food Sci., 54(5): 814-818. A O A C . 1998. Official Methods of Analysis, 16 t h ed., Association of Official Analytical Chemists; Washington, DC. Arakawa, T., and Timasheff, S.N. 1982. Stabilization of protein structure by sugars. Biochem., 21:6536-6544. Asghar, A. , Morita, J. I., Samejima, K. And Yasui, T. 1985. Functionality of muscle proteins in gelation mechanisms of structured meat products. CRC Critical Reviews in Food Sci. and Nutr., 22: 27-106. Badii, F. and Howell, N . K . 2002. Effect of antioxidants, citrate, and cryoprotectants on protein denaturation and texture of frozen cod (Gadus morhua). J. Agric. Food Chem., 50(7): 2053-2061. Badii, F. and Howell, N . K . 2003. Elucidation of the effect of formaldehyde and lipids on frozen stored cod collagen by FT-Raman spectroscopy and differenctial scanning calorimetry. J. Agric. Food Chem., 51(5): 1440-1446. Barret, T.W., Peticolas, W.L., and Robson, R.C. 1978. Laser-Raman light-scattering observations of conformational changes in myosin induced by inorganic salts. Biophys J., 23: 349-358. Benjakul, S. and Bauer, F. 2000. Physicochemical and enzymatic changes of cod muscle proteins subjected to different freeze-thaw cycles. J. Sci. Food Agric, 80:1143-1150. Benjakul, S., Visessanguan, W., Thongkaew, C , and Tanaka, M . 2003. Comparative study on physicochemical changes of muscle proteins from some tropical fish during frozen storage. Food Res. Int., 36: 787-795. Beveridge, T., Toma, S.J. and Nakai, S. 1974. Determination of SH- and SS-groups in some food proteins using Ellman's reagent. J. Food Sci., 39(1):49-51. 157 Bindslev-Jensen, C , Sten, E., Earl, L .K . , Crevel, R.W.R., Bindslev-Jensen, U . , Hansen, T.K., Stahl Skov, P. and Poulsen, L .K. 2003. Assessment of the potential allergenicity of ice structuring protein type III H P L C 12 using the FAO/WHO 2001 decision tree for novel foods. Food Chem. Toxic, 41: 81-87. Boonsupthip, W. and Lee, T. 2003. Application of antifreeze protein for food preservation: effect of type III antifreeze protein for preservation of gel-forming of frozen and chilled actomyosin. J.Food Sci., 68(5): 1804-1809. Botta, J.R., Bonnell, G., and Squires, B.E. 1987. Effect of method of catching and time and of season on sensory quality of fresh Atlantic cod (Gadus morhua). J. Food Sci., 52(4): 928-931. Bouraoui, M . , Nakai, S. and Li-Chan, E. 1997. In situ investigation of protein structure in Pacific whiting surimi and gels using Raman spectroscopy. Food Res. Int., 30(1): 65-72. Briskey, E.J. and Fukazawa, T. 1971. Myofibrillar proteins of skeletal muscle, Vol . 19. In Chichester, C O . , Stewart, G.F., and Mrak, E . M . (eds). Advances in Food Research. Academic Press Inc., N Y . , p. 279-360. Brown, D.J. and Sonnichsen, F.D. 2002. The structure of fish antifreeze proteins, Ch. 5. In Ewart, K . V . and Hew, C. L. (eds). Fish Antifreeze Proteins. World Scientific Publishing Co., River Edge, NJ., p. 109-138. Bush, C.A., Ralapati, S., Matson S.M., Osuga, D.T., Yeh, Y . and Feeney, R.E. 1984. Conformation of the antifreeze glycoprotein of polar fish as determined from the fully assigned n.m.r. spectrum. Arch. Biochem. Biophys., 28: 386-397. Careche, M . and Li-Chan, E.C.Y. 1997. Structural changes in cod myosin after modification with formaldehyde or frozen storage. J. Food Sci., 62(4): 717-723. Careche, M . , Del Mazo, M.L . , Torrejon, P., and Tejada, M . 1998a. Importance of frozen storage temperature in the type of aggregation of myofibrillar proteins in cod (Gadus morhua) fillets. J. Agric. Food Chem., 46(4): 1539-1546. Careche, M . , Cofrades, S., Carballo, J., and Colmenero Jimenez, F. 1998b. Emulsifying and gelation properties during freezing and frozen storage of hake, pork and chicken actomyosin as affected by addition of formaldehyde. J. Agric. Food Chem., 46(3): 813-819. Careche, M . , Herrero, A . M . , Rodriguez-Casado, A. , Del Mazo, M.L . , and Carmona, P. 1999. Structural changes of hake (Merluccius merluccius L.) fillets: effects of freezing and frozen storage. J. Agric. Food Chem., 47(3): 952-959. 158 Careche, M . , Garcia, M.L . , Herrero, A. , Solas, M.T., and Carmona, P. 2002. Structural properties of aggregates from frozen stored hake muscle proteins. J. Food Sci., 67(8): 2827-2832. Carew, B.E., Asher, I.M., and Stanley, H . E. 1975. Laser Raman spectroscopy- New probe of myosin structure. Science, 188(4191): 933-936. Carey, P.R. 1982. Principles of Raman spectroscopy, Ch.2. In Carey, P.R. (ed). Biochemical Applications of Raman and Resonance Raman Spectroscopies. Academic Press, New York, N Y . , p. 11 -47. Carpenter, J.F. and Crowe, J.H. 1988. The mechanism of cryoprotection of proteins by . solutes. Cryobiol, 25:244-255. Carvajal, P.A., McDonald, G.A., and Lanier, T.C. 1999. Cryostabilization mechanism of fish muscle proteins by maltodextrins. Cryobiol, 38: 16-26. Chao, H. , DeLuca, C. and Davies, P.L. 1995. Mixing antifreeze protein types changes ice crystal morphology without affecting antifreeze activity. FEBSLett., 357: 183-186. Chao, H. , Houston, M.E. , Hodges, R.S., Kay, C M . , Sykes, B.D., Loewen, M . C , Davies, P.L. and Sonnichsen, F.D. 1997. A diminished role for hydrogen bonds in antifreeze proteins binding to ice. Biochem., 36: 14652-14660. Claus, J. R., Colby, J. W., and Flick, G. J. 1994. Processed meats/poultry/seafood, Ch. 5. In Kinsman, D. M . , Kotula, A .W. and Breidenstein, B. C. (eds). Muscle foods: Meat Poultry and Seafood Technology. Chapman & Hi l l , New York, N Y . , p. 147-162. Crevel, R.W.R., Fedyk, J.K., and Spugeon, M.J . 2002. Antifreeze proteins: characteristics, occurrence and human exposure. Food Chem. Toxicol, 40: 899-903. Damodaran, S. 1985. Estimation of disulfide bonds using 2-nitro-S-thiosulfobenzoic acid: Limitations. Anal. Biochem., 145: 200-204. Davies, P.L. and Hew, C.L. 1990. Biochemistry of fish antifreeze proteins. FASEB J., 4: 2460-2468. Davies, P.L. and Sykes, B.D. 1997. Antifreeze proteins. Current Opinion in Struct. Biol, 7: 828-834. Del Mazo, M.L. , Huidobro, A. , Torrejon, P., Tejada, M . , and Careche, M . 1994. Role of formaldehyde in formation of natural actomyosin aggregates in hake during frozen storage. Z. Lebensm. Unters. Forsch., 198: 459-464. 159 Del Mazo, M.L . , Torrejon, P., Careche, M . , and Tejada, M . 1999. Characteristics of the salt-soluble fraction of hake (Merluccius merluccius) fillets stored at -20 and -30°C. J. Agric. Food Chem., 47(4): 1372-1377. Deng, G., Andrews, D. and Laursen, R.A. 1997. Amino acid sequence of a new type of antifreeze protein, from the longhorn sculpin Myoxocephalus octodecimspinosis. FEBSLett., 402: 17-20. DeVries, A . L . 1971. Freezing resistance in some Antarctic fishes. Science, 163: 1073-1075. DeVries, A . L . 1988. The role of antifreeze glycopeptides and peptides in the freezing avoidance in Antarctic fishes. Comp. Biochem. Phys. Part B, 90: 611-621. Dong, A. , P. Huang, and Caughey, S. 1990. Protein secondary structure in water from second-derivative amide I infrared spectra. Biochem., 29: 3303-3308. Drewes, J.A., and Rowlen, K . L . 1993. Evidence for a gamma-turn motif in antifreeze glycopeptides. Biophys. J., 65: 985-991. Duman, J.G. and A . L . DeVries. 1974. Freezing resistance in winter flounder Pseudopleuronectes americanus. Nature, 247: 237-238. Duman, J.G. and A . L . DeVries. 1976. Isolation, characterization and physical properties of protein antifreezes from winter flounder Pseudopleunectus americanus. Comp. Biochem. Physiol. Part B, 54: 375-380. Ellman, G.D. 1959. Tissue sulfhydryl groups. Arch. Biochem. Biophys., 82:70-77. FAO. 2002. State of world fisheries and aquaculture, vol. 95. Food and Agriculture Organization of the United Nations, Rome, Italy. Feeney, R.E. and Yeh, Y . 1993. Antifreeze proteins: Properties, mechanism of action and possible applications. Food Technol., 47(1): 82-88. Feeney, R.E. and Yeh, Y . 1998. Antifreeze proteins: Current status and possible food uses, Trends in Food Sci. and Technol, 9: 102-106. Femeena, H. , Sajan, G , Mukundan, M.K. , and Sherif, P .M. 1999. Role of collagen in gaping of fish fillets. Fishery Technol, 36(1): 40-42. Fletcher, G.L., Hew, C.L., and Joshi, S.B. 1982. Isolation and characterization of antifreeze glycoproteins from the frostfish, Microgadus tomcod. Can. J. Zool, 60: 348-355. 160 Fletcher, G.L., Hew, C.L., and Davies, P.L. 2001. Antifreeze proteins of Teleost fishes. Annu. Rev. Physiol., 63: 359-390. -Gomez-Guillen, M.C. , Borderias, A.J . , and Montero, P. 1996. Rheological properties of gels made from high- and low-quality sardine (Sardine pilchardus) mince with added nonmuscle proteins. J. Agric. Food Chem., 44(3): 746-750. Graether, S.P., Slupsky, C M , Davies, P.L., and Sykes, B.D. 2001. Structure of type I antifreeze protein and mutants in supercooled water. Biophys. J., 81: 1677-1683. Graether, S.P., C M . Slupsky, P.L. Davies, and Sykes, B.D. 2003. Freezing of a fish antifreeze protein results in amyloid fibril formation. Biophys. J., 84: 552-557. Green-Walker, M . and Pull, G. 1975. A survey of red and white muscles of marine fish. J. Fish. Biol, 7: 295-300. Griffith, M . and Vanya Ewart, K . 1995. Antifreeze proteins and their potential use in frozen foods. Biotechnol. Advan., 13(3): 375-402. Gronwald, W., Chao, H. , Reddy, D.V., Davies, P.L., Sykes, B.D., and Sonnichsen, F.D. 1996. N M R characterization of side chain flexibility and backbone structure in the type I antifreeze protein at near freezing temperatures. Biochem., 35: 16698-16704. Harding, M . M . , P.I. Anderberg, and A.D.J . Haymet. 2003. Antifreeze glycoproteins from polar fish. Eur. J. Biochem., 270: 1381-1392. Health Canada. Novel food Information. Retrieved July 22 n d , 2005. http://www.hc-sc.gc.ca/fn-an/gmf-agm/appro/nf-anl 19decdoc_e.html (last modified on 2006-03-27) Herrera, J.R., Pastoriza, L. , and Sampedro, G. 2002. Effects of various cryostabilizers on protein functionality in frozen-stored minced blue whiting muscle: the importance of inhibiting formaldehyde production. Eur. Food Res. Technol, 214: 382-387. Herrera, J.R. and Mackie, I.M. 2004. Cryoprotection of frozen-stored actomyosin of farmed rainbrow trout (Oncorhynchus mykiss) by some sugars and polyols. Food Chem., 84(1): 91-97. Herrrero, A . M . , Carmona, P., Garcia, M.L . , Solas, M.T., and Careche, M . 2005. Ultrastructural changes and structure changes and mobility of myowater in frozen-stored hake (Merluccius merluccius L.) muscle: relationship with functionality and texture. J. Agric. Food Chem., 53(7): 2558-2566. Herrrero, A . M . , Carmona, P., and Careche, M . 2004. Raman spectroscopic study of structural changes in hake (Meluccius merluccius L.) muscle proteins during frozen storage. J. Agric. Food Chem., 52(8): 2147-2153. 161 Hew, C.L., Fletcher, G.L., and Ananthanarayanan, V.S. 1980. Antifreeze proteins from the shorthorn sculpin, Myoxocephalus scorpius: isolation and characterization. Can. J. Biochem., 58: 377-383. Honikel, K.O. 1998. Reference methods for assessment of physical characteristics for meat. Meat Sci., 49(4): 447-457. Howell, N . , Arteaga, G., Nakai, S., and Li-Chan, E.C.Y. 1999, Raman spectra in the C-H stretching region of proteins and amino acids for investigation of hydrophobic interactions. J. Agric. Food Chem., 47(3): 924-933. Ingolfsdottir, S., Stefansson, G., and Kristbergsson, K . 1998. Seasonal variations in physicochemical and textural properties of North Atlantic cod (Gadus morhud) mince. J. Aquatic Food Prod. Technol, 7(3): 39-61. Jittinandana, S., Kenney, P.B., and Slider, S.D. 2005. Cryoprotectants preserve quality of restructured trout products following freeze-thaw cycling. J. Muscle Foods, 16:354-378. Khan, A . A . , Hossain, A. , Hara, K. , Osatomi, K. , Ishihara, T., Osako, K. , and Nozaki, Y . 2004. Concentration dependent effect of enzymatic fish protein hydrolysate on the state of water and denaturation of lizard fish (Saurida wanieso) myofribrills during dehydration. Food Sci. Technol. Res., 10(2): 132-136. Kim, B.Y. , Hamman, D.D., Lanier, T.C., and Wu, M.C . 1986. Effects of freeze-thaw abuse on the viscosity and gel-forming properties of surimi from two species. J. Food Sci., 51(4): 951-956, 1004. Knight, C.A., and DeVries, A . L . 1994. Effects of a polymeric, nonequlibrium 'antifreeze' upon ice growth from water. J. Cryst. Growth, 143: 301-310. Koushafar, H. and Rubinsky, B. 1997. Effect of antifreeze proteins on frozen primary prostatic adenocarcinoma cells. Urol, 49(3): 421-425. Koushafar, H. , Pham, L. , Lee, C. and Rubinsky, B. 1997. Chemical adjucant cryosurgery with antifreeze proteins. J. Surg. Oncol, 66:114-121. Krinim, S. and Bandekar, J. 1986. Vibrational spectroscopy and conformation of peptides, polypeptides and proteins. Adv. Protein Chem., 38: 181-365. Laemmli, U .K. 1970. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227: 680-685. 162 Lane, A . N . , Hays, L . M . , Feeney, R.E., Crowe, L . M . , and Crowe, J.H. 1998. Conformational and dynamic properties of a 14 residue antifreeze glycopeptide from Antarctic cod. Prot. Sci., 7: 1555-1563. Lane, A . N . , Hays, L . M . , Tsvetkova, N . , Feeney, R.E., Crowe, L . M . , and Crowe, J.H. 2000. Comparison of the solution conformation and dynamics of antifreeze glycoproteins from Antarctic fish. Biophys. J., 78: 3195-3207. Lanier, T. 2000. Surimi gelation chemistry. Ch. 9. In Park, J.W. (ed). Surimi and Surimi Seafood. Marcel Dekker, Inc. N . Y . , p. 237-265. Larese, A. , Acker, J., Muldrew, K. , Yang, H. , and McGann, L. 1996. Antifreeze proteins induce intracellular nucleation. Cryo-Letters, 17:175-182. Laursen, R.A., D. Wen, and C A . Knight. 1994. Enantioselective adsorption of the D-and L-forms of an a-helical antifreeze polypeptide in the {202 1} planes of ice. J. Am. Chem. Soc, 116: 12057-12058. Laws, D.D., H.L. Bitter, K . Liu, H.L. Ball, K . Kaneko, H. Wille, F.E. Cohen, S.B. Prusiner, A. Pines, and Wemmer, D.E. 2001. Solid-state N M R studies of the secondary structure of a mutant prion protein fragment of 55 residues that induces neurodegeneration. Proc. Natl. Acad. Sci. USA., 98(20): 11686-11690. Li-Chan, E. 1996. The application of Raman spectroscopy in food science. Trends in Food Sci. Technol, 7: 361-370. Li-Chan, E. and Nakai, S. 1991. Raman spectroscopy study of thermally and/or dithiothreitol induced gelation of lysozyme. J. Agric. Food Chem., 39(7): 1238-1245. Li-Chan, E., Nakai, S., and Hirotsuka, M . 1994. Raman spectroscopy as a probe of protein structure in food systems. In Yada, R.Y. , Jackman, R.L., and Smith, J.L. (eds). Protein Structure-Function Relationships in Foods. Chapman and Hall Inc., N Y . , p. 163-197. Li-Chan, E. and Qin, L. 1998. The application of Raman spectroscopy to the structural analysis of food protein networks. In Sessa, D.J. and Willet, J.L. (eds), Paradigm for Successful Utilization of Renewable Resources. AOCS Press, Champaign, IL., p. 123-139. Li-Chan, E., A . A . Ismail, J. Sedman, and van de Voort, F.R. 2002. Vibrational spectroscopy of food and food products. In Chalmers, J .M. and Griffiths, P.R. (eds), Vol . 5. Handbook of Vibrational Spectroscopy. John Wiley & Sons, Ltd. Chichester, UK. , p. 3644-3646. Lian, P.Z., Lee, C M . and Hufnagel, L. 2000. Physicochemical properties of frozen red hake (Urophycis chuss) mince as affected by cryoprotective ingredients. J. Food Sci., 65(7): 1117-1123. 163 Liao, Y . , Wang, C , Tsenf, C , Chen, H. , Lin, L., and Chen, W. 2004. Compositional and conformational analysis of yam proteins by near infrared Fourier transform Raman specstroscopy. J. Agric. Food Che., 52(26): 8190-8196. Lillford, P.J. and Holt, C.B. 1994. Antifreeze proteins. J. Food Engineering., 22: 474-482. Lin, S.B., Chen, L.C. , and Chen, H.H. 2005. The change of thermal gelation properties of horse mackerel mince led by protein denaturation occurring in frozen storage and consequential air flotation wash. Food Res. Int., 38: 19-27. Lippert, J.L., Tyminski, D. and Desmeules, P.J. 1976. Determination of the secondary structure of proteins by laser Raman spectroscopy. J. Amer. Chem. Soc, 98: 7075-7080. Loewen, M.C. , Chao, H. , Houston, M .E. , Baardsnes, J., Hodges, R.S., Kay, C M . , Sykes, B.D., Sonnichsen, F.D., and Davies, P.L. 1999. Alternative roles for putative ice-binding residues in type I antifreeze protein. Biochem., 38: 4743-4749. Love, M . 1980. Seasonal variation and some alternative approaches to fish biology. Ch. 5. In The Chemical Biology of Fishes. Vol . 2. Academic Press, New York, N Y . , p. 350-387. McCormick, R. J. 1994. Structure and properties of tissues. Ch. 2. In Kinsman, D. M . , Kotula, A . W., and Breidenstein, B. C. (eds). Muscle Foods: Meat, Poultry and Seafood Technology. Chapman & Hall, New York, N Y . , p. 25-62. MacDonald, G.A. and Lanier, T . C 1991. Carbohydrates as cryoprotectants for meats and surimi. Food Technol, 45(3): 150-159. MacDonald, G.A. and Lanier, T.C. 1994. Actomyosin stabilization to freeze-thaw and heat denaturation by lactate salts. Food Sci., 59(1): 101-105. MacDonald, G.A., Lanier, T.C. and Carvajal, P.A. 2000. Stabilization of proteins in surimi, Ch. 5. In Park, J.W. (ed). Surimi and Surimi Seafood. Marcel Dekker, Inc. N Y , p. 91-125. Mackie, I.M. 1993. The effects of freezing on flesh proteins. Food Rev. Int., 9(4): 575-610. Marquardt, B.J. and Wold J.P., 2004. Raman analysis of fish: a potential method for rapid quality screening. Lebensm.-Wiss. u. Technol., 37:1-8. 164 Mathew, S. and Prakash, V . 2005. Polyhydric alcohols mediated inhibition of calcium activated adenosine triphosphatase activity of fish skeletal muscle actomyosin. Int. J. Food Prop., 8: 255-265. Mathlouthi, M . , F. Hutteau, and Angiboust, J.F. 1996. Physicochemical properties and vibrational spectra of small carbohydrates in aqueous solution and the role of water in their sweet taste. Food Chem., 56: 215-221. Matsumoto, J. 1979. Denaturation of fish muscle proteins during frozen storage. In Fennema, O. (ed). Proteins at Low Temperatures. American Chemical Society, Washington, D C , p. 205-224. Matsumoto, J. 1980. Chemical deterioration of muscle proteins during frozen storage. In Whitaker, J.R. and Fujimaki, M . (eds). Chemical Deterioration of Proteins. American Chem. Soc, Washington, D C , p. 95-124. Mishra, V . and Pattnaik, P. 1999. Antifreeze proteins: prospects and perspectives in food sector. Indian Food Industry, 18(4): 238-244. Mitsuiki, M . , Mizuno, A. , Tanimoto, FL, and Motoki, M . 1998. Relationship between the antifreeze activities and the chemical structures of oligo- and poly(glutamic acid)s. J. Agric. Food Chem., 46(3):891-895. Morrison, C.R. 1993. Fish and shellfish, Ch. 8. In Mallet, C P . (ed). Frozen Food Technology. Blackie Academic & Professional, New York, N Y . , p. 197-235. Nakai, S., Li-Chan, E., and Arteaga, G.E. 1996. Measurement of surface hydrophobicity. Ch. 8. In Hall, G . M . (ed). Methods of Testing Protein Functionality. Blackie Academic & Professional, London, UK. , p. 226-259. Niwa, E. 1985. Functional aspect of surimi. In Martin, R.E. and Collete, R.L. (eds). Engineered Seafood Including Surimi. Noyes Data Corporation, Park Ridge, NJ., p. 136-142. Ogawa, M . , Kanamaru, J., Miyashita, H. , Tamiya, T., and Tsuchiya, T. 1995. Alpha-helical structure of fish actomyosin: Changes during setting. J. Food Sci., 60(2): 297-299. Ogawa, M . , Nakamura, S., Horimoto, Y . , An, H. , Tsuchiya, T., and Nakai, S. 1999. Raman spectroscopic study of changes in fish actomyosin during setting. J. Agric. Food Chem., 47(8): 3309-3318. Osako, K. , Hossain, M . A. , Kuwahara, K. , and Nozaki, Y . 2005. Effect of trehalose on the gel-forming ability, state of water and myofibril denaturation of horse mackerel (Trachurus japonicus) surimi during frozen storage. Fisheries Sci., 71: 367-373. 165 Painter, P.C. 1984. The application of Raman spectroscopy to the characterization of food, Ch. 11. In Gruenwedel, D.W. and Whitaker, J.R. (eds). Food Analysis: Principles and Techniques. Vol . 2, Physicochemical Techniques. Marcel Dekker, Inc. New York, N Y . , p. 511-545. Paradkar, M . M . and Irudayaraj, J. 2001. Discrimination and classification of beet and cane inverts in honey by FT-Raman spectroscopy. Food Chem., 76: 231-239. Park, J.W. 2000. Ingredient technology and formulation development, Ch. 12. In Park, J.W. (ed). Surimi and Surimi Seafood. Marcel Dekker, Inc. N Y . , p. 91-125. Payne, S.R. and Young, O A . 1995. Effects of pre-slaughter administration of antifreeze proteins on frozen meat quality. Meat Sci., 41:147-155. Payne, S.R. and Wilson, P.W. 1994. Comparison of the freeze/thaw characteristics of Antarctic Cod (Paranotothenia mawsoni) and Black Cod (Paranotothenia augustata) - Possible effects of antifreeze glycoproteins. J. Muscle Foods, 5:233-244. Payne, S.R., Sandford, D., Harris, A . , and Young, O A . 1994. The effects of antifreeze proteins on chilled and frozen meat. Meat Sci., 37: 429-438. Pearson, A . M . and Young, R. B. 1989. Muscle and Meat Biochemistry. Academic Press, San Diego, CA. , p. 1-33. Pham, L., Dahiya, R. and Rubinsky, B. 1999. An in vivo study of antifreeze protein adjuvant cryosurgery (brief communication). Cryohiol., 38: 169-175. Pitcher, T. and Hart, P. 1982. Fish design plans and fish communities, Ch. 1. In Fisheries Ecology. A V I Publishing Co., Westport CT., p. 9-48. Porteous, J.D. and Wood, D.F. 1983. Water-binding of red meats in sausage formulation. Can. Inst. Food Sci. Technol. J., 16: 212-214. Przybycien, T., and Bailey, J.E. 1989. Structure-function relations in the inorganic salt-induced precipitation of a-chymotrypsin. Biochim. Biophys. Acta, 995:231-245. Pszczola, D.E. 2003. Sweetener + sweetener enhances the equation. Food Technol, 57(11): 48-61. Ramirez, J.A., Martin-Polo, M.O., and Bandman Everett. 2000. Fish myosin aggregation as affected by feezing and initial physical state. J. Food Sci., 65(4): 556-560. Rodriguez-Herrera, J.J., Pastoriza, L., Sampedro, G. 2002. Effects of various cryo-stabilizers on protein functionality in frozen-stored minced blue whiting muscle: The importance of inhibiting formaldehyde production. Eur. Food Res. Technol, 214 (6): 382-387. 166 Saeed, S., and Howell, N.K. 1999. High-performance liquid chromatography and spectroscopic studies on fish oil oxidation products extracted from frozen Atlantic mackerel. J. Am. Oil Chem. Soc, 76: 391-397. Samson, A. , Regenstein, J .M. and Laird, W . M . 1985. Measuring textural changes in frozen minced cod flesh. J. Food Biochem., 9:147-159. Santos-Yap, E . M . 1996. Fish and Seafood, Ch. 4. In: Jeremiah, L .E . (ed). Freezing Effects on Food Quality. Marcel Dekker, Inc., N Y . , p. 109-129. Shahidi, F. 1994. Seafood proteins and preparation of protein concentrates, Ch. 2. In Shahidi, F. and Botta, J. R. (eds). Seafoods: Chemistry, Processing Technology and Quality. Chapman & Hall, New York, N Y . , p. 3-9. Shenouda, S.Y.K. 1980. Theories of protein denaturation during frozen storage of fish flesh. Adv. Food Res., 26: 275-311 Schiraldi, C , Di Lernia, I., and De Rosa, M . 2002. Trehalose production: exploiting novel approaches. Trends in Biotechnol., 20(10): 420-425. Scholander, P.F., Van Dam, L. , Kanwisher, J.W., Hammel, H.T., and Gordon, M.S. 1957. Supercooling and osmoregulation in Arctic fish. J. Cell Comp. Physiol., 49: 5-24. Shurvell, H.F. and Bergin, F.J. 1989. Raman spectra of L(+)-glutamic acid and related compounds. / . Raman Spec, 20: 163-168. Siamwiza, M . N . , Lord, R.C., Chen, M.C. , Takamatsu, T., Harada, I., Matsuura, H. , and Shimanouchi, T. 1975. Interpretation of the doublet at 850 and 830 cm"1 in the Raman spectra of tyrosyl residues in proteins and certain model compounds. Biochem., 22: 4870-4876. Sicherl, F. and Yang, D.S.C.. 1995. Ice binding structure and mechanism of an antifreeze protein from winter flounder. Nature, 375: 427-431. Sijo, M . and Prakash, V . 2005. Polyhydric alcohols mediated inhibition of calcium activated adenosine triphosphatase activity of fish skeletal muscle actomyosin. Int. J. Food Prop., 8:255-265. Sikorski, Z.E. and Kotakowska, A. 1994. Changes in proteins in frozen stored fish, Ch. 8. In Zdzistaw, E., Sun Pann, B., and Shahidi, F. (eds.). Seafood Proteins. Chapman and Hall, New York, NY. , p. 99-112. Slaughter, D. and Hew, C.L. 198L Improvements in the determination of antifreeze protein activity using a freezing point osmometer. Anal. Biochem., 115: 212-218. 167 Sola-Penna, M . and Meyer-Fernandes, J.R. 1998. Stabilization against thermal inactivation promoted by sugars on enzyme structure and function: why is trehalose more effective than other sugars? Arch. Biochem. Biophys., 360(1): 10-14. Spiro, T.G. 1987. Raman spectra and conformations of biological macromolecules. Vol . 1. In Spiro, T.G. (ed). Biological Applications of Raman Spectroscopy. John Wiley & Sons, N Y . , p. 3-45. Sultanbawa, Y . and Li-Chan, E.C.Y. 1998. Cryoprotective effects of sugar polyol blends in ling cod (Ophiodon elongates) surimi during frozen storage. Food Res. Int., 31(2): 87-98. Sultanbawa, Y . and Li-Chan, E .C.Y. 2001. Structural changes in natural actomyosin and surimi from ling cod (Ophiodon elongates) during frozen storage in the absence or presence of cryoprotectants. / . Agric. Food Chem., 49(10): 4716-4725. Surewicz, W.K., H.H. Mantsch, and D. Chapman. 1993. Determination of protein secondary structure by Fourier transform infrared spectroscopy: A critical assessment. Biochem., 32(2): 389-394. Susi, H . and Byler, D . M . 1988. Fourier deconvolution of the amide I Raman band of proteins as related to conformation. Appl. Spec, 42(5): 819-826. Suzuki, T. 1981. Fish and Kri l l Protein Processing Technology. Applied Science Pub., London, UK. , p. 1-28. Sych, J., Lacroix, C , Adambounou, L.T., and Castanigne, F. 1990. Cryoprotective effects of some materials on cod-surimi proteins during frozen storage. / . Food Sci., 55(5): 1222-1227. Tejada, M . , Careche, M . , Torrejon, P., Del Mazo, M.L . , Solas, M.T., Garcia, M . L . , and Barba, C. 1996. Protein extracts and aggregates forming in minced cod (Gadus morhua) during frozen storage. J. Agric Food Chem., 44 (10): 3308-3314. Thannhauser, T.W., Konishi, Y . , and Scheraga, H.A. 1984. Sensitive quantitative analysis of disulfide bonds in polypeptides and proteins. Anal. Biochem., 138: 181-188. Tomimatsu, Y. , Scherer, J.R., Yeh, Y . and Feeney, R.E. 1976. Raman spectra of a solid antifreeze glycoprotein and its liquid and frozen aqueous solutions. J. Biol. Chem., 251: 2290-2297. Tsvetkova, N . , Phillips, B.L. , Krishnan, V . V . , Feeney, R.E., Fink, W.H. 2002. Dynamics of antifreeze glycoproteins in the presence of ice. Biophys. J., 82(1): 464-473. Tu, A.T. 1982. Basic concept and elementary theory. In: Tu, A.T. (ed). Raman spectroscopy in biology: Principles and applications. John Wiley & Sons, N.Y. , p. 1-43. 168 Tu, A.T. 1986. Peptide backbone conformation and microenvironment of protein side chains. In: Clark, J.H., Hester, R.E. (eds). Spectroscopy of Biological Systems. John Wiley & Sons, NY. , p. 47-112. Venugopal, V . 2006a. Availability, consumption pattern, trade, and need for value addition, Ch. 1. In Seafood Processing: Adding Value Through Quick Freezing, Retortable Packaging, and Cook-chilling. CRC Press, Boca Raton, FL. , p. 1-21. Venugopal, V . 2006b. Postharvest quality changes and safety hazards, Ch. 2. In Seafood Processing: Adding Value Through Quick Freezing, Retortable Packaging, and Cook-chilling. CRC Press, Boca Raton, FL. , p. 23-60. Vik, A.F. , Yuzyuk, Y.I. , Barthes, M . , and Sauvajol, J.L. 2005. Low-wavenumber dynamics of L -alanine. J. Raman Spectrosc, 36(8): 749-754. Vojdani, F. 1996. Solubility. In Hall, G .M. (ed). Methods for Testing Protein Functionality. Blackie Academic & Professional, New York, N Y . , p. 11-60. Walrafen, G.E. and Chu, Y . C . 1995. Linearity between structural length and correlated-proton Raman intensity from amorphous ice and supercooled water up to dense supercritical steam. J. Phys. Chem., 99: 11225-11229. Wang, T., Zhu, Q., Yang, X . , Layne, J.R. (JR), and Devries, A . L . 1994. Antifreeze glycoproteins from Antarctic notothenoid fishes fail to protect the rat cardiac explant during hypothermic and freezing preservation. Cryobiol, 31: 185-192. Wang, J.H. 2000. A comprehensive evaluation of the effects and mechanisms of anti-freeze proteins during low-temperature preservation. Cryobiol., 41: 1-9. Warren, G.J., Mueller, G .M. , and McKnown, R.L. 1992. Ice crystal growth suppression polypeptides and method of making. US Patent 5118792. Warns, P.D. 2000. The chemical composition and structure of meat, Ch. 3. In Meat Science: An Introductory Text. C A B I Publishing, New York, N Y . , p. 37-67. . Wierzbicki, A. , Taylor, M.S., Knight, C.A., Madura, J.D., Harrington, J.P. and Sikes, C.S. 1996. Analysis of shorthorn sculpin antifreeze protein stereospecific binding to (2-1 0) faces of ice. Biophys. J., 71: 8-18. Williams, R.W. 1983. Estimation of protein secondary structure from the laser Raman amide I spectrum. J. Mol. Biol, 152: 783-813. Wu, Y. , Banoub, J., Goddard, S.V., Kao, M . H . , and Fletcher, G.L. 2001. Antifreeze glycoproteins: relationship between molecular weight, thermal hysteresis and the inhibition of leakage from liposomes during thermotropic phase transition. Comp. Biochem. Physiol. Part B, 128:265-273. 169 Xie, G. and Timasheff, S.N. 1997. The thermodynamic mechanism of protein stabilization by trehalose. Biophys. Chem., 64: 25-43. Xiong, Y . L . 2004. Muscle proteins. Ch. 5. In Yada, R.Y. (ed). Proteins in Food Processing. Woodhead Publishing Ltd., Boca Raton, FL. , p. 100-122. Yoon, K.S. and Lee, C M . 1990. Cryoprotectant effects in surimi/mince-based extruded products. / . Food Sci., 55(5): 1210-1216. Yuan, C , Kaneniwa, M . , Wang, X . , Chen, Y . , Qu, Y. , Fukuda, Y . , and Konno, K. 2006. Seasonal expression of 2 types of myosin with different thermostability in silver carp muscle (Hypophthalmichthys molitrix). J. Food Sci., 71(1): 39-43. 170 Appendix I. MALDI-TOF mass spectrometry of AFGP. "1 ' 1 • 1 r •— I —•—i—•——l 1 r» 1034.0 2059.4 3084.8 4110.2 5135.6 6161.0 Mass [mfi) Acquired: 14:55:00, March 17, 2005 A F G P FA mar 1 7ANDREA D:\suzanne\AFGP FA mar 17 ANDREA_0003.dat 17.1 Appendix II. MALDI-TOF mass spectrometry of AFPI. 70-60-Mass $m!2'i Acquired: 13:32:00, March 17. 2005 APF 1 refit mar 17 Andrea D:\suzanne\APF 1 refit mar 17 Andrea_0001.dat 172 Appendix III. Raman band intensities in the 400-3400 cm"1 region for AFGP and AFPI solutions held at different temperatures, showing upper and lower 95% confidence limits. Relative peak intensity (wavenumber ± 2 cm") Peak assignment AFGP 22°C -0.5°C -1.8°e -4.0°C NH 3 + torsions of L-alanine and C-C-C bending vibrations 0.78 (463) (0.70, 0.86) . 0.65 (460) (0.59,0.71) 0.65 (460) (0.59, 0.71) 0.79 (463) (0.71, 0.87) C-CH 3 stretch of the polypeptide (0.60, 0.74) 0.79 (886) (0.71, 0.87) 0.57 (889) (0.51, 0.63) 0.59 (887) (0.53, 0.65) C-C stretch of the proline ring 0.76(929) (0.69, 0.83) 0.67 (929) (0.60, 0.74) 0.52 (930) (0.47, 0.57) 0.37(930) (0.33, 0.41) N-acetyl chromophore of the disaccharide 1.30 (984) (1.17,1.43) 1.05 (980) (0.95, 1.15) 1.02 (980) (0.92, 1.12) 0.94 (980) (0.87, 1.07) C-C, C-N stretching and carbohydrate C-H, COH vibrations 2.20(1070) (1.98, 2.42) 1.94(1071) (1.77, 2.15) 1.81 (1073) (1.63, 1.99) 2.30(1073) (2.07, 2.53) C-C, C H 3 vibrations of alanine 1.45sh(1105) (1.31, 1.59) 1.57sh(1105) (1.42, 1.72) 1.24sh (1101) (1.12, 1.36) r 1.69sh(1105) (1.52, 1.86) Amide III and carbohydrate COH coordinate 0.79(1253) (0.71, 0.87) 1.02(1256) (0.92, 1.12) _____ (0.69, 0.85) 0.75(1640) (0.68, 0.82) 0.71 (1624) (0.63, 0.77) 0.67(1253) (0.60, 0.74) 0.53(1674) (0.48, 0.58) 0.44(1647) (0.40, 0.48) 0.47(1618) (0.42, 0.52) 0.88(1260) (0.79, 0.97) ~~0.62(1672) (0.56, 0.68) 0.72(1647) (0.65, 0.79) 0.60(1616) (0.54, 0.66) 0.71sh(2996) (0.64,0.78) 2.60 (2945) (2.35, 2.85) 0.94sh(2887) (0.85, 1.03) Extended structure Unordered structure Polyproline type II helix 0.87(1674) (0.78, 0.96) 0.9(1645) (0.81, 0.99) 0.78(1624) (0.70, 0.86) C-H stretch, aliphatic 0.90sh (2992) (0.81, 0.99) 2.97 (2946) (2.68, 3.26) 1.02sh (2888) (0.92, 1.12) 0.92sh (2993) (0.83, 1.01) 3.30 (2946) (2.98, 3.62) 1 .Osh(2885) (0.90, 1.10) 0.77sh(2992) (0.69, 0.85) 2.61 (2945) (2.35, 2.85) 0.84sh(2891) (0.78, 0.96) O-H stretching 5.70 (3227) (5.14, 6.26) 6.80 (3210) (6.13, 7.47) 5.20 (3206) (4.69, 5.71) 5.10(3209) (4.60, 5.60) ^Numbers in parenthesis next to the intensity refer to wavenumbers ± 2 cm"1; sh = shoulder. To discriminate between treatments, the standard of deviation has been assumed from the coefficient of variation of 5% for the method and upper and lower 95% confidence bands are shown for each measurement in parenthesis below the intensity value (Badii and Howell, 2003). 173 Relative peak intensity (wavenumber ± 2 c m ' ) ' Peak assignment AFPI 22°C -0.5°C -1.8°C -4.0°C C0 2 " rocking mode of L-alanine and C-C-C bending vibrations C-N, C-N-C stretching mode of alanine and glutamic acid 0.66 (534) 1.26 (530) (0.60,0.72) (1.14,1.38) 1.09(910) 1.86(913) (0.98, 1.20) (1.68, 2.04) 1.03 (530) (0.93, 1.13) 1.46 (909) (1.32, 1.60) 1.23 (530) (1.11, 1.35) . 1.74 (909) (1.57, 1.91) C-C, C-N stretching 1.00(1060) 0.60(1059) (0.90,1.10) , (0.54,0.66) 0.86sh(1078) 0.94sh(1078) (0.78, 0.94) | (0.85, 1.03) 1.08 (1063) (0.97, 1.19) 1.07sh (1080) (0.97, 1.17) 1.28 (1063) (1.15, 1.41) 1.54sh (1083) (01.39, 1.69) C-C, CH3 vibrations of alanine 0.93(1107) j 1.20(1108) (0.84, 1.02) : (1.08, 1.32) 1.01 (1108) (0.90, 1.10) 1.60 (1108) (1.44, 1.76) Amide III and 0.49 (1276) 1 0.81 (1262) (0.44, 0.54) (0.73, 0.89) 0.74 (1267) (0.67, 0.81) 0.82 (1272) (0.72, 0.90) Amide I 0.86 (1646) (0.78, 0.94) 1.41 (1645) (1.26, 1.54). 0.66sh (1685) (0.60, 0.72) 0.89sh (1620) (0.80, 0.98) 1.10(1641) (0.99, 1.21) 0.46sh (1686) (0.41,0.51) 0.54sh (1620) (0.49, 0.59) 1.60 (1645) (1.44, 1.76) 0.64sh (1684) (0.59, 0.71) 0.87sh (1618) (0.78, 0.96) C-H stretch, aliphatic O-H stretching 0.64sh (2993) (0.58, 0.70) 2.13 (2945) (1.92, 2.34) 0.98sh (2890) (0.88, 1.08) 3.20 (3225) (2.89, 3.51) 0.77sh (2993) (0.69, 0.85) 2.44 (2946) (2.20, 2.68) 0.98sh (2888) (0.88, 1.08) 4.70 (3201)"* (4.21, 5.13) 0.68sh (2994) (0.61, 0.75) 2.00 (2946) (1.80, 2.20) 0.81 sh (2887) (0.72, 0.88) '3.8T(3202J' (3.49, 4.25) 0.51 sh (2993) (0.46, 0.56) 2.31 (2946) (2.07, 2.53) 1.02sh (2886) (0.92, 1.12) 4.3 (3200) (3.88, 4.72) Numbers in parenthesis next to the intensity refer to wavenumbers ± 2 cm" ; sh = shoulder. To discriminate between treatments, the standard of deviation has been assumed from the coefficient of variation of 5% for the method and upper and lower 95% confidence bands are shown for each measurement in parenthesis below the intensity value (Badii and Howell, 2003). 174 I Appendix IV. Results for moisture and crude protein analyses of ling cod mince. Sample % Moisture1 % Crude protein ; -ti 3(w.b.) i Cntrol 82 (0.3) 16.8 Commercial blend 74 (0.1) 16.7 AFGP (0.05mg/mL) 83 (0.4) 14.6 AFGP (0.5mg/mL) 82 (0.1) 16.6 AFP I (0.5mg/mL) 81 (0.1) 15.6 AFGP (0.005mg/mL) 81 (0.5) 16.9 AFGP (O.OlOmg/mL) 80 (0.7) 17.5 AFP I (0.005mg/mL) 80 (0.3) 17.9 AFP I (O.OlOmg/mL) 80 (0.4) 17.7 SuSo 75 (0.1) 16.7 AFGP-PO 81 (0.2) 16.8 AFPI-PO 81 (0.3) 17.2 AFGP-PolPO 77 (0.1) 16.7 AFPI-PolPO 77 (0.2) 17.5 Treha-PO 75 (0.3) 16.4 'Values represent mean of three replicates + (SD). 2 Values represent mean of two replicates, w.b. = wet basis pH of ling cod fillet (mean of three replicates): 6.38 + 0.03 % ash of ling cod fillet (mean of three replicates): 1.17 + 0.16 175 Appendix V. Temperature profile of 3 freeze-thaw cycles for fish mince without cryoprotectants. • Environment 1 -25 1 Time (minutes) Minced samples were placed in 0.1 mm thick 6x12" polyethylene bags and frozen at -20 ± 2°C before being submitted to 3 freeze-thaw cycles. One freeze-thaw cycle consisted of freezing at -20 ± 2°C for 18 hrs and thawing at 4°C for 6 hrs. 176 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0092920/manifest

Comment

Related Items