Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The non-specific immune system and innate disease resistance in different strains of teleost fish Balfry, Shannon Kathleen 1997

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

Item Metadata

Download

Media
831-ubc_1997-25013x.pdf [ 7.32MB ]
Metadata
JSON: 831-1.0087945.json
JSON-LD: 831-1.0087945-ld.json
RDF/XML (Pretty): 831-1.0087945-rdf.xml
RDF/JSON: 831-1.0087945-rdf.json
Turtle: 831-1.0087945-turtle.txt
N-Triples: 831-1.0087945-rdf-ntriples.txt
Original Record: 831-1.0087945-source.json
Full Text
831-1.0087945-fulltext.txt
Citation
831-1.0087945.ris

Full Text

THE NON-SPECIFIC IMMUNE SYSTEM AND INNATE DISEASE RESISTANCE IN DIFFERENT STRAINS OF TELEOST FISH by S H A N N O N K A T H L E E N B A L F R Y B . S c , Simon Fraser University, 1984 M . S c , Simon Fraser University, 1991 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T OF T H E R E Q U I R E M E N T S F O R T H E D E G R E E OF D O C T O R OF P H I L O S O P H Y in T H E F A C U L T Y OF G R A D U A T E S T U D I E S (Department of Animal Science) We accept this thesis as conforming to/die required standard T H E U N I V E R S I T Y OF B R I T I S H C O L U M B I A July 1997 © Shannon Kathleen Balfry, 1997 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of /hOlMftt SClr3\S& The University of British Columbia Vancouver, Canada Date TUl>) DE-6 (2/88) A B S T R A C T Evidence for genetic differences in innate disease resistance in fish have been demonstrated, but the mechanisms responsible for these differences are not well understood. This thesis tested the hypothesis that the non-specific immune system plays a significant role in the innate disease resistance of fish. The experimental approach was to compare the activity of the non-specific immune system in different strains of fish, and to investigate correlations between this activity and innate disease resistance Significant strain differences in the activity of the non-specific immune system were observed for chinook salmon {Oncorhynchus tshawytscha), Nile tilapia (Oreochromis niloticus), and coho salmon (O. kisntch). Disease challenges performed on two strains of coho salmon, showed that the disease resistant strain (highest survival) had a more active and sustained internal cellular and humoral non-specific immune response following the challenge. Further coho salmon strain comparisons showed significant differences in innate disease resistance, thought to be associated with the differences in the external non-specific immune system. When compared with other strains, the most disease resistant strain showed higher mucus bactericidal activity, which appeared to be partially due to increased mucus lysozyme activity and hemolytic activity. A bactericidal cationic peptide appeared to be present in the mucus and is reported for the first time in coho salmon. The strain comparisons reported in this thesis demonstrate the genetic variation of innate disease resistance and non-specific immunity in fish. Strain differences were most apparent in the internal non-specific immune system of infected fish, while strain differences in the activity of the external non-specific immune system were seen in healthy uninfected fish. The external and internal non-specific immune system appeared to have a significant role in preventing infections and subsequent disease-related mortality in fish. Increased activity in both the external and internal non-specific immune system was seen in the strains exhibiting the greatest innate disease resistance. TABLE OF CONTENTS Page Abstract ii Table of Contents iv List of Tables vi List of Figures vii Preface ix Acknowledgments x Dedication xi G E N E R A L INTRODUCTION AND OVERVIEW 1 A review of the non-specific immune system in fish 2 The non-specific immune system and genetic variation in disease resistance ... 12 Thesis hypothesis and organization 18 G E N E R A L MATERIALS A N D METHODS Disease challenges 22 Coho salmon strains 23 Measurement of the non-specific immune system 25 Hematology 27 CHAPTER ONE Comparison of chinook salmon {Oncorhynchus tshawytscha) strains for differences in non-specific immunity Introduction 29 Materials and Methods 30 Results and Discussion 32 CHAPTER TWO Strain differences in non-specific immunity of tilapia {Oreochromis niloticus) following challenge with Vibrio,parahaemolylicus Introduction 36 Materials and Methods 37 Results and Discussion 40 CHAPTER THREE Coho salmon {Oncorhynchus kisutch) strain differences in non-specific immunity following a non-lethal challenge with Vibrio angiallarum Introduction... 45 Materials and Methods 46 Results 48 Discussion 50 CHAPTER FOUR Coho salmon (Oncorhynchus kisutch) strain differences in innate disease resistance and activity of the internal humoral and cellular non-specific immune system, following immersion challenges in different doses of Vibrio anguillarum Introduction 59 Materials and Methods 60 Results 63 Discussion 65 CHAPTER FIVE Comparison of coho salmon (Oncorhynchus kisutch) strains for differences in resistance to experimental infections with Vibrio anguillarum Introduction 74 Materials and Methods 75 Results 79 Discussion 81 CHAPTER SIX An examination of the bactericidal activity of fish mucus and serum in different strains of coho salmon {Oncorhynchus kisutch) Introduction 91 Materials and Methods 92 Results 96 Discussion 97 G E N E R A L DISCUSSION .... 108 BIBLIOGRAPHY 115 APPENDIX A. Genetic analysis of lysozyme activity and resistance to vibriosis in farmed chinook salmon, Oncorhynchus tshawytscha (Walbaum) Introduction 128 Materials and Methods 129 Results 131 Discussion 132 APPENDIX B. Observations on lysozyme activity in coho salmon (Oncorhynchus kisutch) Introduction 140 Materials and Methods 140 Results 143 Discussion 144 vi LIST OF T A B L E S Page Table 1. Overview of the genetic variation of non-specific immune components in fish 15 Table 1-1. Comparison of hematological values in three strains of chinook salmon 34 Table 2-1. Means of different leucocyte types in black and red strains of Nile tilapia following injection challenge with Vibrio parahaemolyticus or saline. 43 Table 3-1. Comparison of hematological values in two strains of coho salmon, immediately before, 2 and 16 d following a non-lethal immersion challenge with Vibrio angxtiUarum 55 Table 4-1. Coho salmon strain comparison of hydrogen peroxide production by anterior kidney phagocytes and differential leucocyte numbers, immediately before and following immersion challenges with different doses of Vibrio anguiliarum 69 Table A-1. Results of log-linear analysis of mortality data and A N O V A of lysozyme activity in two different environmental treatment groups consisting of replicated families of farmed chinook salmon 136 Table B - l . Examination of genetic variation of mean lysozyme activity of various tissues, within and between different strains of coho salmon in two different year classes 147 vii LIST OF FIGURES Page Figure 1-1. Comparison of serum bactericidal activity and lysozyme activity in three strains of chinook salmon 35 Figure 2-1. Mean serum lysozyme activity and anterior kidney phagocyte respiratory burst activity in red and black strains of Nile tilapia following injection challenge with Vibrio parahaemolyticus. 44 Figure 3-1. Coho salmon strain comparison of mean plasma, gill, and kidney lysozyme activity, immediately before, 2 and 16 d following a non-lethal immersion challenge with Vibrio anguiUarum 56 Figure 3-2. Coho salmon strain comparison of mean anterior kidney phagocyte respiratory burst activity immediately before, 2 and 16 d following a non-lethal immersion challenge with Vibrio anguillarum 57 Figure 3-3. Coho salmon strain comparison of mean plasma glucose concentrations immediately before and 2 d following a non-lethal immersion challenge with Vibrio anguillarum 58 Figure 4-1. Percent cumulative mortality in two strains of coho salmon following an immersion challenge with Vibrio anguillarum 70 Figure 4-2. Coho salmon strain comparison of mean plasma lysozyme activity, immediately before and following immersion challenges in different doses of Vibrio anguillarum 71 Figure 4-3. Coho salmon strain comparison of mean anterior kidney phagocyte respiratory burst activity, immediately before and following immersion challenges in different doses of Vibrio anguiUarum 72 Figure 4-4. Correlations between mean plasma lysozyme activity and mean percent anterior kidney phagocyte respiratory burst activity, in two strains of coho salmon challenged with different doses of Vibrio anguillarum 73 Figure 5-1. Percent cumulative mortality of five coho strains following injection and immersion challenges with Vibrio anguillarum 87 Figure 5-2. Coho salmon strain comparison of viable Vibrio anguillarum present in the gills and blood immediately following an immersion challenge 88 Figure 5-3. Coho salmon strain comparison of viable Vibrio anguillarum in homo-genized blood following an immersion challenge 89 Figure 5-4. Coho salmon strain comparison of viable Vibrio anguillarum in kidney, viii spleen, and liver, 24 and 48 h following an immersion challenge 90 Figure 6-1. Coho salmon strain comparison of the bactericidal activity of untreated and heat-treated serum, after 1.5 and 7 h incubation with a suspension of Vibrio anguillarum 101 Figure 6-2. Coho salmon strain comparison of the effect of divalent cations and heat on the bactericidal activity of mucus against Vibrio anguillarum 102 Figure 6-3. Coho salmon strain comparison of the effect of divalent cations and heat on the bactericidal activity of serum against Vibrio anguillarum 103 Figure 6-4. Coho salmon strain comparison of the effect of divalent cations and heat on the bactericidal activity of serum against Aeromonas salmonicida 104 Figure 6-5. Coho salmon strain comparison of alternative complement pathway activity, lysozyme activity, and protein concentration, in serum and mucus 105 Figure 6-6. Coho salmon strain comparison of mucus bactericidal activity against defensin-resistant and defensin-susceptible strains of Salmonella typhimurium 106 Figure 6-7. Coho salmon strain comparison of defensin-related bactericidal activity, using gel overlays 107 Figure A - l . Total percent cumulative mortality for 12 full-sib and 6 half-sib families of farmed chinook salmon, following a natural outbreak of vibriosis 137 Figure A-2. Mean plasma lysozyme activity for 12 full-sib and 6 halfrsib families of farmed chinook salmon, three months following a natural outbreak of vibriosis 138 . Figure A-3. Correlation between percent cumulative mortality and mean plasma lysozyme activity of two environmental treatment groups containing replicated chinook salmon families 139 Figure B - l . Correlations between maternal lysozyme activity (serum and kidney) and unfertilized egg lysozyme activity 148 Figure B-2. Coho salmon strain comparison of lysozyme activity in different stages of early development 149 Figure B-3. Correlations between fish weight and serum lysozyme activity in individual coho salmon from two different year classes 150 I X P R E F A C E Chapter Two has been accepted for publication and is in press; Balfry, S.K., Shariff, M . , Iwama, G.K. (1997) Tilapia {Oreochromis niloticus) strain differences in non-specific immunity following challenge with Vibrio parahaemolyticus, Diseases of Aquatic Organisms. All the work was performed by myself. Appendix A has been accepted for publication and is in press; Balfry, S.K., Heath, D.D., Iwama, G.K. (1997) Genetic analysis of lysozyme activity and resistance to vibriosis in farmed chinook salmon, Oncorhynchus tshawytscha (Walbaum), Aquaculture Research. This work was a collaborative effort with Dr. D. Heath. Dr. Heath performed all the data analysis and co-wrote the manuscript. Appendix B is my own unpublished work. Monisha Gough performed the gel electrophoresis work described in Chapter Six. A C K N O W L E D G M E N T S I would like to thank my thesis supervisor, Dr. George Iwama for all his support, guidance and patience provided throughout the course of this work. I am also grateful for the many opportunities given to me (i.e., Malaysia, Hawaii, and financial support to attend all the conferences). I would also like thank my committee members Dr. Trevor Evelyn, Dr. Alec Maule, and Dr. Kim Cheng for their guidance and help with this thesis. I am indebted to all the Iwama-ites who have helped me with this thesis. There are so many to thank, but I have to first thank Ellen Teng for taking such good care of all the fish. Special thanks to my quality control person, courier, and computer guru, Paige Ackerman. To John Morgan, thanks for everything. I could never find a cinnamon bun big enough to express how grateful I am for all your help. I especially have to thank Laura Brown for her friendship and her help with this thesis. Thanks to all the staff and managers from the following DFO hatcheries that were involved with the coho broodstock collection and egg incubation; Capilano, Chehalis, Kitimat, Quinsam, Robertson and Rosewall Creek. Mr. John Ketcheson and Mrs. Gina Prosperi-Porta from the Biological Station have also helped with their advice and endless supply of bacterial isolates. I would like to thank Dr. Shariff and the staff, students, and faculty at the Universiti Pertanian Malaysia for their help and friendship. Thanks also to the following agencies and companies who have provided funding for this thesis; Science Council of British Columbia for the Graduate Research and Engineering Academic ' Training (GREAT) Scholarships and cosponsor Integrated Broodstock Ltd., the Canadian Bacterial Diseases Network, the British Columbia Applied Research Partnership Program and Moore-Clarke Co. (Canada) Ltd., and the Canada-ASEAN centre for the academic travel grant. Finally I would like to thank my husband, Joe Tadey for getting me through another degree with his usual patience, love, and understanding. Thank you Jack, for providing me the opportunity to conduct the truly definitive scientific experiment. And my last thank you goes to all those fish who have laid down their life in the name of science and a Ph.D. xi D E D I C A T I O N This thesis is dedicated to my family: the Balfrys and the Tadeys; those who have left this world, and those who have not yet arrived; human and almost human (Basil). Your love and support have made the dreams of a five year old come true. Thank you ail from the bottom of my heart. 1 GENERAL INTRODUCTION AND OVERVIEW Interest in fish immunology has grown over the last couple of decades, largely due to the world-wide increase in fish farming. As with any other type of agriculture, disease prevention is important to ensure the economic viability of the aquaculture industry. Fish are susceptible to diseases caused by parasites, viruses, bacteria and fungi. There are two possible outcomes once a fish has become infected with a pathogenic organism: (a) the fish will either successfully fight the infection by preventing the growth and colonization of the pathogen, and in doing so survive; or (b) the fish may not be successful in preventing the spread of the infection, and a state of physiological dysfunction (i.e., disease) will develop and the fish will die. The different outcomes (i.e., survival or death) are largely determined by the efficacy of the immune system in preventing the initial infection from occurring, and/or preventing the growth and spread of the pathogen once an infection has started. The immune system of vertebrates, including teleost fish, can be categorized into specific acquired immunity and non-specific innate immunity. Both use cellular and humoral (soluble) mechanisms to provide protection against infections. The specific immune system involves the recognition of specific antigen(s) on a pathogen, thereby providing protection against that pathogen primarily in the form of specific antibodies. Specific disease resistance is therefore affected'by factors that influence antibody production, such as the type and duration of antigen stimulation, age, temperature and stress (Ellis 1988, Tatner 1996). In contrast, the non-specific immune system provides an array of protective mechanisms that are inherently available and provide immediate and permanent protection against a wide variety of pathogens. In general the magnitude of this non-specific protection is consistent, regardless of the type of pathogen or the number of times the pathogen infected the animal. / 2 The cellular and humoral factors of the non-specific immune system provide fish with external and internal protection against infectious agents. These factors can act alone or together with other non-specific or specific immune factors to provide a range of protective mechanisms. The following is a brief summary of the non-specific immune system in fish. For the purposes of this discussion the non-specific immune system is divided into humoral and cellular components. The individual components are reviewed separately, but the reader should be aware that many of these components do not function alone. A R E V I E W OF T H E NON-SPECIFIC I M M U N E S Y S T E M IN FISH 1. Physical barriers The mucus, skin and scales provide a very effective external barrier against invading pathogens. In addition, the continuous shedding of mucus prevents the attachment and subsequent colonization of pathogens, which is the first step in the infection process (see Chapter Five for more discussion on the infection process). The importance of fish mucus in maintaining homeostasis is reviewed by Shephard (1994). Various non-specific immune components have been found to play a role in the external defense system of fish. Peleteiro and Richards (1990) have shown that phagocytic cells, such as macrophages, are present in the epidermis and appear to be capable of crossing the basal lamina. In addition, fish mucus has been found to contain lysozyme, lectins, proteinases, complement, C-reactive protein (CRP), and trypsin-like enzymes (Fletcher and Grant 1968, Harrell et al. 1976, Braun et al. 1990, Alexander and Ingram 1992). Recently, the gel-forming glycoproteins in fish mucus, have been reported to possess antibacterial activity (Marginos et al. 1995). 3 II. Humoral Components Humoral components act in a variety of ways to kill, and/or prevent the growth and spread of pathogens. Many of the components use unique mechanisms to lyse the pathogen. Others act as agglutinins (they aggregate cells) or precipitins (they aggregate molecules). There are also opsonins that bind with the pathogen and in doing so, facilitate its uptake and removal by phagocytic cells. Each component can use a single mechanism or any combination of these and other mechanisms to prevent infections. The following is a brief overview of some of the humoral components involved in the non-specific immune system in fish. (a) Lysozyme. Alexander Fleming discovered lysozyme after his nasal mucus dripped onto a plate of bacteria, and inhibited their growth. He later called this lytic substance lysozyme (Fleming and Allison 1922). The structure and bacteriolytic mechanism of lysozyme has been described in detail (Salton 1957, Chipman and Sharon 1969). Briefly, lysozyme acts as a hydrolase, targeting the |3-( 1 -4) linkages between N-acetylmuramic acid and N-acetylglucosamine in the peptidoglycan layer of bacterial cell walls. Gram-negative bacteria are generally less sensitive to lysozyme because there is an outer membrane covering the peptidoglycan layer. Gram-positive bacteria lack the outer membrane, so the peptidoglycan layer is exposed, thus rendering these bacteria very vulnerable to the bacteriolytic effects of lysozyme. The opposite trend is seen for the sensitivity of bacteria to the lytic effects of complement (discussed below). Gram-negative bacteria are more sensitive to complement than Gram-positive. It is believed that as in mammals, lysozyme and complement act together in fish, to cause the lysis of bacteria (Hjelmeland et al. 1983). Lysozyme is widely distributed in nature. It has been found in various tissues and secretions of humans and animals (Salton 1957). Lysozyme is constitutively expressed, 4 synthesized and secreted by neutrophils, monocytes and macrophages. Increased expression can be induced by certain substances such as lipopolysaccharides and glucocorticoids (Altschmied et al. 1989). Lysozyme turnover in humans is very rapid, with 76% of the total plasma lysozyme eliminated by the kidneys in one hour (Maack and Sigulem 1974). Lysozyme has been found in a variety of freshwater and marine fish species (Lie et al. 1989). Monocytes, macrophages and polymorphonuclear granulocytes are known to synthesize and secrete lysozyme in fish (Murray and Fletcher 1976). Kidney tissue appears to have the greatest concentration of lysozyme activity, likely due to the high concentration of leucocytes in the anterior hematopoeitic portion of the kidney. It is also possible that as in humans, fish kidneys filter and absorb lysozyme and may act as a 'sink' for circulating lysozyme (Maack and Sigulem 1974). Lysozyme has also been detected in many other fish tissues such as spleen, liver, skin, mucus, gills, muscle, ovary and eggs (Takahashi et al. 1986, Lie et al. 1989, Yousif et al. 1991, Takemura and Takano 1995). Unlike mammalian lysozyme, fish lysozyme is an effective bacteriolytic agent against both Gram-positive and Gram-negative fish pathogens (Grinde 1989, Yousif et al. 1994). Mammalian lysozyme is classified as c-type lysozyme, with a molecular weight of approximately 14.5 kDa (Salton 1957). Fish also have c-type lysozyme, which has been further separated into Types I and II. These two types of lysozyme show differences in bactericidal activity, amino acid sequence and isoelectric point (Grinde 1989). Lysozyme activity increases following a stimulation of the immune system (Studnicka et al. 1986, M0yner et al. 1993). The elevation in lysozyme activity has been associated with increases in leucocyte numbers (Fletcher and White 1973), and macrophage activation (Secombes and Fletcher 1992). Changes in lysozyme activity have also been attributed to stress, sex, season, temperature, and degree of sexual maturity (Fletcher and White 1976, Fletcher et al. 1977, Mock and Peters 1990). The genetic variation of lysozyme has also been established and is discussed below. (b) Complement. The bactericidal properties of complement were first described in the 1890s, but it was not until the 1970's that the entire complement system was characterized and accepted (Lepow 1980). Complement refers to a series of interactions that are initiated by two distinct pathways - the alternative complement pathway (ACP) and the classical complement pathway (CCP). The result of complement activation by either pathway is the same, the formation of the lytic membrane attack complex ( M A C ) . The M A C is a complex of complement molecules that polymerize to form a lytic hydrophobic plug that penetrates the bacterial cell membrane to form pores. Lysozyme is thought to play an important role in complement lysis by breaking apart the outer peptidoglycan layer and thus exposing the membrane so the M A C can attach and create pores. The C C P is considered to be a component of the specific immune system because it is activated by an antigen-antibody complex, while the A C P is considered to be part of the non-specific immune system because it is activated by a variety of compounds and surfaces such as the lipopolysaccharide (LPS) found in the cell membranes of Gram-negative bacteria, rabbit erythrocytes, zymosan, and inulin. During the sequence of events that culminates in complement mediated lysis, many by-products (usually peptides) are formed. Some of these substances have important roles in the inflammatory immune response as opsonins, anaphylatoxins, neutrophil and macrophage chemoattractants (Law and Reid 1988). The virucidal, bactericidal, parasiticidal, opsonic and chemoattracting activity offish complement is reviewed by Yano (1996). Complement activity in fish appears to be comparable to that seen in mammals (reviewed by Alexander and Ingram 1992, Sakai 1992). Both A C P and C C P pathways exist, with complement activity detected in the serum and skin mucus (Harrell et al. 1976, Ingram 1987, Alexander and Ingram 1992). The A C P pathway appears to have a greater role in the protection offish, as it is more active in the early stages of infection (Yano 1996) and at low temperatures when antibody production is suppressed (Koppenheffer 1987). Alternative complement pathway bactericidal activity requires M g 2 + , plus factors B (glycoprotein) and D (serine protease) (Law and Reid 1988). Fish complement has been found to be heat sensitive, with complete inactivation of coho salmon (Oncorhynchus kisutch) complement occurring at 45 °C for 20 min (Sakai 1981). Ourth and Bachinski (1987) reported that the bactericidal activity of the A C P may be related to the sialic acid content of the bacterial cell membrane. These investigators found that bactericidal activity was absent against Gram-positive bacteria or virulent Gram-negative bacteria that contained sialic acid in the cell membrane. Sialic acid has been shown to prevent the attachment of factor B in the activation step of the A C P (Law and Reid 1988). L o w A C P activity has been detected in salmonid fish at the alevin and fry stage (approximately 5-6 months after hatching), with activity increasing to higher (adult) levels by one year of age (Sakai 1992). Seasonal variation has been demonstrated for A C P activity (Yano et al. 1984). A C P activity is also affected by the degree of sexual maturation (Raed et al. 1992), as reflected by reduced hemolytic activity in Atlantic salmon with signs of sexual maturation. Genetic variation in complement activity has also been shown and is discussed later. (c) C-reactive Protein. C-reactive protein (CRP) is a normal serum constituent, but is considered an acute-phase protein because its concentration greatly increases following tissue damage or infection (Murai et al. 1990). In the presence of C a 2 + , C R P causes the precipitation and agglutination of a wide variety of carbohydrate and phosphoryl ester containing substances that are present in the pathogen cell wall. It was first reported by Tillet and Francis (1930) and 7 since then it has been detected in both vertebrate and invertebrate animals. In addition, to its agglutinating and precipitating properties, C R P has been shown to activate complement (CCP), macrophages and natural cytotoxic cells (Kilpatrick and Volanakis 1991). C R P has been detected in the serum, eggs, mucus and various tissues of fish (Fletcher and Baldo 1976, Fletcher et al. 1977, Ramos and Smith 1978). Variation in the activity offish C R P has been associated with season (White et al. 1983), sex (Fletcher et al. 1977), and stress (Szalai et al. 1994). (d) Interferon. Interferons (IFNs) are proteins that inhibit viral replication. Three types of IFNs have been identified in mammals (a, p, and y) (Stewart 1980). Fish interferons have been classified into two types, 'a,p' and 'y' , Gravell and Malsberger, 1965) and Graham and Secombes (1988), respectively. While the two I F N types bind to different receptors, the result of the binding has the same effect, i.e., the induction of enzyme systems that prevent the synthesis of viral proteins (Alexander and Ingram 1992). Recently, IFN-y has been shown to be a potent macrophage activating factor (Graham and Secombes 1988). (e) Transferrin. Transferrin is a globular protein capable of reversibly binding iron. The removal of iron by transferrin serves to seriously affect the growth of pathogens, which require iron for normal metabolic function (Putnam 1975). Transferrin has been isolated from the serum of a variety of fish species (reviewed by Yano 1996). A possible role of transferrin in resistance to bacterial kidney disease has been suggested by Suzumoto et al. (1977). However, later conflicting results reports by Pratschner (1978), Winter et al. (1980), and Withler and Evelyn (1990) suggest there is no relationship between disease resistance and transferrin (or the relationship is unclear at this time). (0 Lectin. Lectins are proteins capable of agglutinating and precipitating carbohydrate moieties freely associated or attached to cell surfaces (Alexander and Ingram 1992). In general, two types of lectins are found, the S-type and the C-type (Drickamer 1988). The S-type or sulphydryl-dependent lectins are soluble, noncation dependent, occur intracellular^ or extracellularly, and have specific carbohydrate binding properties. The C-type or calcium-dependent lectins, occur either membrane-bound or extracellularly, and are capable of binding with a variety of carbohydrates. Lectins have been found in the egg, serum and mucus of a variety of fish species (reviewed by Alexander and Ingram 1992, and Yano 1996). It is unclear at this time whether fish lectins are C- or S-type lectins. However, it has been suggested that the S-type lectin may be the dominant lectin in eggs and mucus (Alexander and Ingram 1992). (g) Miscellaneous Humoral Substances. Various lytic substances (hemolysins, chitinase, proteinases), agglutinins/precipitins (serum amyloid P-component, a-precipitin, natural precipitins, natural antibodies, natural hemagglutinins), enzyme inhibitors (serine proteinase inhibitors, cysteine-proteinase inhibitors, metalloproteinase inhibitors, a2-macroglobulin) and pathogen growth inhibitors (caeruloplasmin, metallothionein) have been reported in fish, and have been recently reviewed by Alexander and Ingram (1992) and Yano (1996). Antimicrobial cationic peptides such as defensins, have antibacterial, antiviral, and antifungal properties (Hancock 1997). They are produced in the leucocytes of a variety of animal species (Boman 1995). Antimicrobial peptides have recently been isolated from the skin secretions of the winter flounder, Pleuromctes americanus (Cole et al. 1997), and carp, Cyprinus carpio (Lemaitre et al. 1996). Antibacterial peptides have also been isolated from the Moses sole fish, Pardachirus marmoratus (Oren and Shai 1996). III. Cellular Components The leucocytes of fish have been the subject of several reviews (Ellis 1977, Ainsworth 1992, Hine 1992). Three types of leucocytes are described below with reference to their role in the non-specific immune system. Characteristics of each cell type and the defense mechanisms are discussed. 1. Cell Types (a) Granulocytes. Granulocytes or polymorphonuclear phagocytes include the neutrophils, eosinophils and basophils. The common feature of these three cells is the presence of granules within the cytoplasm, which contain enzymes and other substances that are involved in the immune response against pathogens. Granulocytes are mobile cells capable of moving through the entire animal via the blood and lymphatic system. They are therefore very important in the early stages of an infection. (b) Macrophages/Monocytes. Monocytes are the mobile, immature precursors to the macrophage. In contrast, macrophages are non-mobile, and are found in greatest concentrations in the lymphoid organs (kidney and spleen). Macrophages are mononuclear cells, that have a significant role in the non-specific immune system as highly efficient phagocytes, capable of killing a variety of pathogens. Macrophages also play a role in the specific immune system as antigen-presenting cells. 10 (c) Nonspecific Cytotoxic Cells. Non-specific cytotoxic cells (NCCs) are similar to the mammalian natural killer cells, capable of lysing tumor cell lines and protozoan parasites (Evans and Jaso-Friedmann 1992). They are found in the blood, lymphoid tissues and gut of fish. 2. Phagocytosis Phagocytic cells are responsible for the clearance of foreign substances and senescent blood cells from the body. Monocytes, macrophages, neutrophils, and eosinophils are all phagocytic in fish. Thrombocytes are also thought to be slightly phagocytic (Dannevig et al. 1994). Neutrophils are the most active of all the phagocytes, because they are very mobile and capable of scavenging the entire circulatory system for foreign substances. The killing activity of the neutrophils is, however, not as great as has been reported for macrophages (Secombes and Fletcher 1992). The process of phagocytosis can be divided into three steps (reviewed by MacArthur and Fletcher 1985, and Secombes 1996). First is the attachment of the phagocyte to the pathogen or foreign substance. The relative specificity of the attachment process suggests that surface receptors may be important. Lectin-like receptors, (3-glucan receptors and Fc receptors for antibody have been shown to be present on fish macrophages (Haynes et al. 1988, Saggers and Gould 1989, Engstad and Robertsen 1994). Attachment is generally a passive process, although the opsonization of a pathogen with complement and/or C R P does appear to increase the rate of adherence. The second stage of phagocytosis, is ingestion into the phagocyte. This is an active process involving pseudopodia which surround the substance, to create a phagosome within the cytoplasm of the phagocyte. The final stage of phagocytosis is the actual killing of the pathogen, which utilizes both oxygen-dependent and oxygen-independent processes (Secombes and Fletcher 1992). Pathogens are generally killed intracellularly, although extracellular killing has been 11 observed (Whyte et al. 1989). The oxygen-independent killing mechanism in fish phagocytes is thought to involve enzymes such as lysozyme. In mammals, cationic proteins, and other lysosomal enzymes including cathepsin, P-glucuronidase, and lysosomal acid hydrolases are involved in killing (Speert 1992). The oxygen-dependent killing mechanisms utilize 0 2 through a respiratory burst process, which produces toxic oxygen- and nitrogen-free radicals. These mechanisms have been demonstrated in fish (reviewed by Secombes and Fletcher 1992). The reactive oxygen species (ROS) produced are superoxide anions (0~2), hydrogen peroxide ( H 2 O 2 ) and reactive singlet oxygen (O) . The reactive nitrogen species (RNS) include nitrous oxide (NO). Bacterial virulence factors (catalase and superoxide dismutase), and low temperature have been shown to reduce the respiratory burst activity of macrophages ( Secombes and Fletcher 1992, Hardie et al. 1994). The opsonization of bacteria with complement and antibody have the opposite effect, and can greatly enhance the respiratory burst activity of macrophages (Lamas and Ellis 1994). The phagocytic system is present at a very early age in salmonids, with macrophages capable of phagocytosing carbon particles at 4 d post-hatch (Tatner and Manning 1985). 3. Inflammation The inflammatory response in fish has been described in detail by Finn and Nielson (1977) and more recently by Secombes (1996). The general sequence of events is comparable to that in mammals, and functions to dilute, isolate, destroy and remove the foreign substances or pathogens (Bell 1977). Briefly, the first step in the inflammatory response is vasodilation, which occurs to facilitate the rapid movement of neutrophils to the site of the stimulus. Monocytes and macrophages later join the neutrophils to remove and destroy the pathogen by the process phagocytosis. The next step is the isolation of the pathogen through a healing process (Roberts 1989). If the inflammatory stimulus is not removed by the above acute inflammatory response, then a chronic inflammatory response may occur (Roberts 1989). Lymphocytes and macrophages aggregate at the site, and form a granuloma with later melanization and fibrosis (Roberts 1989). Several of the above listed humoral immune components have very important roles in the inflammatory response in fish. For example, many of the byproducts of complement activation have been shown to be vasoactive, thus promoting the migration of leucocytes to the site of the inflammatory stimulus (Sakai 1992). Eicosaniods such as prostaglandins, leukotrienes, and thromboxanes are produced by fish leucocytes, and act to enhance the inflammatory response (Secombes 1996). 4. Natural Cytotoxicity. Natural cytotoxic cells (NCCs) appear to be important in the protection of fish against neoplasias, viral and parasitic infections (Secombes 1996). There is still much unknown about natural cytotoxicity. The binding of N C C s to the target cells appears to be receptor mediated (Harris et al. 1992). However, the actual cytotoxic mechanism used to lyse the target cells is unclear. Natural cytotoxicity in fish has been shown to vary with temperature (LeMorvan-Rocher et al. 1995), and stress (Evans and Jaso-Friedmann 1994). THE NON-SPECIFIC IMMUNE SYSTEM AND GENETIC VARIATION IN DISEASE RESISTANCE Evidence of a genetic basis for resistance to disease has been demonstrated in a variety of plants (Wallace 1961) and animals (Hutt 1970). There is also strong evidence for a genetic basis to disease resistance in fish (reviewed by Chevassus and Dorson 1990, Fjalestad et al. 1993). 13 Disease resistance has generally been measured as differences in survival between different genetic groups. Survival rates, however, are poorly defined traits under the influence of many factors, such as pathogen virulence, water temperature, and prior exposure to the pathogen, which often results in an inaccurate estimate of the actual genetic contribution to disease resistance (Gavora and Spencer 1983). Genetic differences in survival reflect the complex nature of the immune system, and provide little information on the underlying mechanisms responsible for the observed differences in disease resistance. The non-specific immune system, rather than the specific immune system, was the focus of the present research into disease resistance of fish for several reasons. First the non-specific immune system represents the innate, immediate, permanent form of protection against infections. Environmental effects such as previous pathogen exposure, and water temperature have a significant impact on the specific immune system. Previous pathogen exposure is required for the formation of memory and specific antibodies which are key to the functioning of the specific immune system. L o w water temperature suppresses antibody production (Rijkers 1982) and T-cell responsiveness (Clem et al. 1984), thereby impairing the activity of the specific immune system. Previous pathogen exposure has no impact on the activity of the non-specific immune system. Low water temperatures have however, been shown to decrease the respiratory burst activity of phagocytes (Hardie et al. 1994) and lysozyme activity (Fletcher and White 1976). In addition, the non-specific immune system appears to be the dominant form of protection during the early developmental stages in fish. This can have a tremendous impact on the ability of fish to survive early life-stage infections caused by such pathogens as Saprolegnia spp., and the infectious haematopoietic necrosis virus. Studies on salmonids have indicated that the specific immune system does not mature and become fully functional until the fish are approximately 4 g mean weight (Ellis 1988). The non-specific immune humoral system functions at a much earlier 14 age; for example, fish eggs have been found to contain lysozyme, lectins and hemagglutinins (Ingram 1980, Yousif et al. 1991). The cellular non-specific immune system also becomes functional at a very early age. Tatner and Manning (1985) have shown that the phagocytic system begins to function at 4 d post-hatch and is fully functional by 14 d post-hatch. The non-specific immune system is capable of providing immediate permanent protection against a variety of pathogens. Thus breeding programs that base their selection decisions on the performance of the non-specific immune system, should have fish resistant to a variety of diseases. It is also possible that non-specific factors may be used for the development of transgenic fish with enhanced disease resistance. Recently, there has been increased attention paid to the non-specific immune system in fish and numerous reports have been published on the presence of significant genetic variation in several non-specific immune factors (summarized in Table 1). Reports of genetic variation in many of the non-specific immune factors have provided strong evidence for the role of the non-specific immune system in innate disease resistance. However, a significant correlation between the non-specific factors listed in Table 1, and actual disease resistance (i.e., survival) have been much more difficult to establish. Lysozyme activity represents the best example in the literature of a non-specific immune factor that appears to be significantly correlated with disease resistance. Lund et al. (1995) reported an apparent negative correlation between lysozyme activity and survival of Atlantic salmon (Sa/mo salar) to the bacterial diseases furunculosis {Aeromonas salmonicida), bacterial kidney disease (Renihacterhim salmon'marum) and cold-water vibriosis {Vibrio salmonicida). Roed et al. (1993) have demonstrated a similar negative correlation between lysozyme activity and the survival of Atlantic salmon to vibriosis (V. anguillarum). Fevolden et al. (1992), however, reported conflicting results on the correlation of lysozyme with survival. In comparisons of high and low stress lines of rainbow trout {(). mykiss), Fevolden et al. found a negative correlation of lysozyme with 15 Table 1. Overview of the genetic variation of non-specific immune components in fish. Source of variation detected was classified as between-strain variation (B) or within-strain variation (W). Component Fish species Variation Reference transferrin transferrin transferrin serum lysozyme activity serum lysozyme activity serum lysozyme activity serum lysozyme activity serum lysozyme activity serum lysozyme activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum hemolytic activity serum bactericidal activity serum agglutinins coho coho W B coho & stlhd B rbt B rbt B Atl At l rbt & At l rbt & Atl rbt rbt rbt rbt At l At l At l At l rbt rbt W W W B B B W W W W W W W B Suzumoto et al. 1977 Utter et al. 1970 Winter et al. 1980 Fevolden et al. 1992 Fevolden & R0ed 1993 Lund et al. 1995 Raed et al. 1993 Grinde et al. 1988 Fevolden et al. 1991 Fevolden et al. 1992 Fevolden & Reed 1993 Raed et al. 1990 Hollebecq et al. 1995 R0ed et al. 1992 R0ed et al. 1993 Lund et al. 1995 Marsden et al. 1996 Hollebecq et al. 1995 Cipriano 1983 16 Table 1 (cont'd.) Overview of the genetic variation of non-specific immune components in fish. Source of variation detected was classified as between-strain variation (B) or within-strain variation (W). Component Fish species Variation Reference cytotoxic serum neutralization rbt B Cipriano 1983 mucus precipitins brt B Cipriano & Heartwell 1986 a2-antiplasmin At l W Salte et al. 1993 fibrinogen At l W Salteet al. 1993 anti-protease activity At l W Marsden et al. 1996 coho: coho salmon (Oncorhynchus kisutch) stlhd: steelhead trout (Oncorhynchus mykiss) rbt: rainbow trout (Oncorhynchus mykiss) brt: brown trout (Salmo trutta) Atl : Atlantic salmon (Salmo salar) 17 furunculosis survival and a positive correlation with vibriosis survival. The positive correlation may be attributed to the different pathogen and fish species used by Fevolden et al. (1992), or perhaps to the type of genetic analysis performed (strains compared rather than the more rigorous family analysis). A consistent pattern of correlation between hemolytic activity and disease resistance has been more difficult to establish than for lysozyme. There appears to be a trend for hemolytic activity to be positively correlated with survival (Roed et al. 1993, Hollebecq et al. 1995, Marsden et al. 1996), but there are exceptions. The work published by Fevolden et al. (1992) showed a positive correlation of hemolytic activity with furunculosis survival, and a negative correlation with vibriosis survival. Lund et al. (1995) found no correlation between hemolytic activity and survival to furunculosis, cold-water vibriosis, or bacterial kidney disease. The problem in establishing a clear relationship between hemolytic activity and disease resistance may be related to the actual contribution of hemolytic activity (presumed to be due to alternative complement pathway activity) to the overall protection of fish from disease. Hollebecq et al. (1995) measured both serum hemolytic activity and total serum bactericidal activity, and found the latter measurement of total killing has a stronger positive correlation with survival. Serum lysozyme and hemolytic activity are the dominant factors that have been studied with respect to disease resistance, but other factors that have significant genetic variation have also been shown to be associated with disease resistance. Transferrin has been reported to be correlated with resistance to bacterial kidney disease (Suzumoto et al. 1977, Winter et al. 1980). The ability of serum to inactivate toxic elements of A salmonicida (i.e., cytotoxic serum neutralization) has been shown to be positively correlated with resistance to furunculosis (Cipriano 1983). Cipriano and Heartwell (1986) found the progeny from brown trout selected for high mucus precipitin activity, were more resistant to furunculosis than progeny from low activity parents. Salte et al. (1993) examined a2-antiplasmin and fibrinogen in Atlantic salmon and found 18 them to be positively and negatively correlated, respectively, with survival to furunculosis. Anti-protease activity in Atlantic salmon was also found to be negatively correlated with furunculosis survival (Marsden et al. 1996). The difficulties in establishing a clear relationship between one factor and disease resistance are likely due to the complex nature of the immune system and the presence of confounding exogenous (i.e., environmental) and endogenous (i.e., developmental) effects. The simplest and most practical approach to examine genetic variation is to compare different genetic strains (Price 1985). A strain can be defined as a genetically distinct population, produced from either naturally or artificially segregated breeding (Larkin 1972). Naturally reproducing coho salmon strains are geographically isolated from each other, and presumably have evolved specific adaptations that enable them to survive in that particular habitat. Adaptations become preserved within the strain because mating generally only occurs between those fish that have originated from the same specific geographic location (i.e., fish of the same strain). The different strains are therefore naturally reproductively isolated, which results in the continued maintenance of their unique genotype. Genetic variation between coho salmon strains from British Columbia (B.C.) has been demonstrated by electrophoretic variation (Utter et al. 1970, Wehrhahn and Powell 1987). Recently Beacham et al. (1996) have used minisatellite D N A probes to clearly demonstrate significant genetic variation between coho salmon strains throughout B . C . In addition, they were also able to show that the degree of genetic similarity between strains was related to geographical location. THESIS HYPOTHESIS AND ORGANIZATION 19 Fjalestad et al. (1993) have outlined three ways to measure disease resistance. These are: a) survival rates at a commercial fish farm; b) survival rates following an artificial disease challenge; and c) the measurement of immunological or physiological parameters. The goal of this thesis was to examine disease resistance of fish using all three methods, and in doing so, demonstrate the role of the non-specific immune system in the genetic variability of innate disease resistance. The hypothesis tested throughout this thesis was that non-specific immunity is significant to innate disease resistance in fish. The experimental approach was to compare the non-specific immune system of different strains of fish, in an attempt to describe and identify which factors played a significant role in innate disease resistance. M y interest in this area of research began with the work of Withler and Evelyn (1990) and Beacham and Evelyn (1992a) who found strain differences in innate disease resistance of coho and chinook salmon {(). tshawytscha) from B . C . Their results indicated the importance of the non-specific immune system, but little supporting research had been published at that time. It was accepted at the beginning of the research presented here, that the underlying mechanisms were very complex and likely involved a variety of mechanisms. To minimize environmental effects, the coho salmon strains (1991 and 1992 broodstock) were reared in communal tanks from a very early age (3 g mean weight). In addition, for each study the fish were graded and only similar sized fish (± 1 SD) used to minimize the effects of body size, and possible stress due to social interactions between different sized fish. The marine pathogen V. angiiillamm was chosen as a model pathogen to ensure the freshwater fish would be immunologically naive (with respect to it) and thus enable an accurate measurement of non-specific immunity. 20 The research was initiated with a general examination the non-specific immune system, by comparing different strains of chinook salmon (Chapter One), Nile tilapia Oveochromis nihticus (Chapter Two), and coho salmon (Chapter Three) with regards to the activity of the internal humoral and cellular non-specific immune system.. The results from these studies supported the role of the non-specific immune system in innate disease resistance. The experiments described in Chapter Four were designed to provide a more in-depth examination of the activity of internal humoral and cellular non-specific immune factors following disease challenges in two strains of coho salmon. Challenging the fish with different doses of V. anguillarum and sampling at various times pre- and post-challenge, provided information on strain differences in the magnitude and duration of the immune response to a pathogen. Strain differences in the response of the non-specific immune system appeared to be associated with differences in post-challenge mortality. The focus of Chapters Five and Six, shifted from examining the activity of the internal non-specific immune system, to an examination of the external non-specific immune system. Strain differences in mortality following challenge by immersion and injection demonstrated the role of the external defense system, and infection studies showed possible strain differences in the infection process (Chapter Five). The bactericidal activity of the mucus was then examined in detail (Chapter Six), and strain differences were found that were consistent with previously determined differences in mortality. The work presented in Appendix A , was a collaborative project, examining genetic variation in plasma lysozyme activity and resistance to vibriosis. The importance of this study lies in the ability to detect significant genetic variation in disease resistance following a natural disease outbreak, and the negative correlation of lysozyme with survival. Appendix B examines the results of various studies on the lysozyme activity in different coho salmon strains. Lysozyme 21 activity has become a popular measurement of non-specific immunity and its genetic variation is well established (see above). However, data reported in this study demonstrates the inherent variability in lysozyme activity, with an examination of the importance of maternal contribution, early life stage development and fish size on lysozyme activity. 22 G E N E R A L M A T E R I A L S A N D M E T H O D S DISEASE C H A L L E N G E S Bacteria. Marine fish pathogens V. anguillarum and V. parahaemolylicus, were used to challenge freshwater-reared coho salmon and Nile tilapia, respectively. The fish were reared strictly in freshwater and thus would have had no previous exposure to these marine pathogens. The fish were therefore assumed to be immunologically naive to these pathogens. This would preclude the involvement of specific immune disease resistance mechanisms, and allow a more accurate measurement of the protection provided by the non-specific immune system. The presence of antibodies against V. anguillarum in the serum was examined by measuring agglutinating antibodies titres (Roberson 1990). The titres were always negative indicating that there had been no previous exposure to the pathogen. Immersion Challenge Procedure. Primary isolates (Pacific Biological Station, Nanaimo, B.C., isolate no. R20) of V. anguillarum (Vang) were obtained from coho salmon that had died from being previously inoculated with the pathogen. Eighteen hour Vang cells were harvested from tryptic soy agar (TSA, supplemented with 1.5% NaCl) plates, transferred to sterile peptone-saline (P-S; 0.1/0.85% peptone/NaCl, respectively), and homogenized with a sterile teflon-tipped homogenizer. The concentration of the Vang suspension was estimated from absorbance measurements made at 540 nm (1 OD 5 4o estimated to contain 109 cells/mL). The bacterial suspensions were kept on ice until the challenges were performed. Challenge buckets were setup containing 4 L of P-S. Water was removed from the fish tanks and used to prepare the P-S, in an effort to minimize any stress that may be attributed to differences in water quality or temperature. Air lines were placed into each challenge bath, along with a few drops of sterile antifoam b (Dow-Corning Corp.). An appropriate amount of the bacterial suspension was added to the challenge bath immediately before the fish were added. Lids were placed on the buckets, and the fish left undisturbed for 20 min. After the challenge, the fish were netted out and placed into their tanks. Control or sham challenge groups were similarly challenged except for the addition of the bacterial suspension. The actual concentration of viable Vang used in the challenges was determined from the bacterial suspension that was serially diluted with P-S, dropped (25 pL) in triplicate onto TSA plates, and incubated overnight at room temperature. The next day the colonies were counted and the actual number of Vang cells/mL in the challenge bath was calculated. Following the challenges, dead fish were collected daily and frozen for later necropsy to confirm vibriosis as the cause of death. Kidney material was aseptically taken and streaked onto TSA plates. The fish was assumed to have died of vibriosis if Gram-negative, motile, curved rod shaped bacteria grew without producing pigments. C O H O S A L M O N STRAINS 1991 Coho Strains. Eggs and milt from 9 male and 9 female adult coho salmon were collected on 1 November 1991 from the Kitimat River, Department of Fisheries and Oceans (DFO) hatchery located on the northern coast of B.C. On 12 November 1991, eggs and milt from 7 male and 7 females were collected from the DFO Quinsam River hatchery on the east coast of Vancouver Island, B.C. The gametes were transported on ice to the DFO Rosewall Hatchery on Vancouver Island, where single paired 1:1 matings were performed. The resulting 16 full-sib families (9 Kitimat and 7 Quinsam) were incubated separately in baskets placed in a vertical stack incubator supplied with 10 °C freshwater. Hatched alevins were transported in their incubation 24 baskets on 28 January 1992, to the Aquaculture Facility at the University of British Columbia, (UBC), Vancouver, B.C. The alevins from each family were released into separate compartments in a rearing trough, and fed by hand to satiation twice daily for one month. Approximately 750 fry from each of the families representing the two strains were combined into their respective rearing troughs. The fish were reared in dechlorinated fresh water and fed a commercial diet using automatic feeders. On 16 August 1992 when the fish were a mean weight of 3 g, approximately 1000 fish per strain were fin clipped for strain identification and added in equal proportions into each of 3 200-L oval tanks. Approximately 6 mo later, the fish were moved into 2 1000-L tanks where they were maintained until July 1993. Fish were hand fed a commercial dry diet daily, left at ambient water temperatures (range 2-18 °C) and exposed to natural photoperiod. 1992 Coho Strains. Gametes were collected from ten pairs of returning adult coho salmon from the DFO Kitimat River Hatchery on the northern coast of B.C. on 9 November 1992, from the DFO Capilano River Hatchery on the southern coast of B.C. on 13 November 1992, from the DFO Quinsam River Hatchery on the east coast of Vancouver Island on 16 November 1992, from the DFO Robertson Creek Hatchery on the west coast of Vancouver Island on 17 November 1992, and from the DFO Chehalis River Hatchery on the lower Fraser River drainage on 26 November 1992. The gametes were transported on ice to the DFO Rosewall Creek Hatchery on Vancouver Island, B.C. for the fertilization and incubation of each strain. Ten full sib families were produced from each strain, by 1:1 (maie:female) matings. Each family was incubated separately in baskets placed in a vertical stack incubator supplied with 10 °C fresh water. Hatched alevins were transported on 5 February 1993 (in their family baskets), to the Aquaculture Facility at the UBC, and the baskets were placed into a vertical stack incubator. The development of each strain varied, so the ponding dates ranged from 12 February 1993 - 15 March 1993. Wet 25 weight measurements were made and approximately equal numbers of alevins from 2-3 families per strain were combined and placed into buckets that were floated in freshwater tanks. The fish were reared in dechlorinated fresh water and fed a commercial diet using automatic feeders. Mortality was monitored throughout, to ensure that the mortality in each group of families was evenly distributed. The families from each strain were combined on 21-22 April 1993, so that the fish from each strain were represented by separate, duplicate 200-L tanks. On 19 August 1993, at approximately 3 g mean weight, 500 fish per strain were fin-clipped for identification and combined into a single large tank. The next month the fish were randomly split into 2 1000-L tanks. Fish were maintained in these two communal tanks, fed daily (ad libitum) and mortality monitored. Water temperatures ranged from 3-18 °C, and natural photoperiod was used throughout the rearing. M E A S U R E M E N T OF T H E NON-SPECIFIC I M M U N E S Y S T E M Lysozyme Activity. The lysoplate method used to determine lysozyme activity in serum, plasma, egg, alevin, fry and kidney samples, is described by Osserman and Lawlor (1966) with modifications outlined by Yousif et al. (1994). Briefly, this assay involved preparing agar plates (lysoagar) containing 0.60 mg/mL Micrococcus lysodeikticus (Sigma), 0.02 M NaCl, 0.50% Agarose (Sigma) in phosphate buffer (PB, 0.06 M , pH 6.0). Wells (approximately 3 mm diameter) were punched into the lysoagar which was air dried. Hen egg white lysozyme standards (HEWL, Sigma) were used. The activity of the HEWL standard (under the assay conditions described), using the turbidimetric method described by the supplier of the HEWL (Sigma) with modifications by Grinde (1989). Samples were placed (in triplicate) onto separate plates, along with the HEWL standards and incubated overnight in a moist chamber. The zones of clearance surrounding each well were then measured with calipers, and compared to the H E W L standards using regression analysis. Lysozyme activity of serum and plasma has been found to be the same (Mock and Peters 1990), therefore we were able to pool both results for the correlation testings in Appendix B (lysozyme and fish weight). The lysozyme activity of kidney, gill, whole egg, alevin and fry samples were determined by diluting (1:4 w/v PB) and homogenizing (Polytron homogenizer) the tissues. The tissue homogenates were centrifuged, and the resultant supernatant assayed for lysozyme activity. Comparisons were made between the lysozyme activity of the whole homogenate and the supernatant, and no significant differences were found. Total Bactericidal Activity. The total bactericidal activity of mucus or serum was examined as described below. Lysozyme, alternative pathway complement activity, naturally occurring agglutinins, lysins, precipitins, are all factors with possible roles in the observed bacterial killing. In the assay, serum/mucus was incubated with an equal volume of bacterial suspension (10 3 V. anguillarum cells/mL). After the incubation period, the suspension was serially diluted in phosphate buffered saline (PBS, pH 7.2), and drop inoculated onto tryptic soy agar (TSA) plates supplemented with 1.5% NaCl . The plates were incubated for 24-48 h, and the colonies counted. The survival of bacteria was expressed as a percent of a control (bacterial suspension incubated with tryptic soy broth, T S B or PBS) . The greater the bactericidal activity of the mucus/serum, the fewer bacteria would survive the incubation because of the bactericidal effect of the factors present in the mucus/serum. This method did not provide specific information on the activity of a particular bactericidal factor, but it had the advantage of demonstrating how all the factors can act together to provide 27 protection against the bacterial pathogen. The results of this assay may therefore more closely represent the /'// vivo situation. Phagocyte respiratory burst activity. Respiratory burst activity of phagocytes results in the production of superoxide anion (0"2), which can be simply measured using the nitroblue tetrazolium (NBT) glass-adhered assay described by Anderson (1992). Briefly, anterior kidney tissue homogenates were dropped (approximately 50 uL) into duplicate wells on a cleaned glass slide, and then incubated in a moist chamber for 30 mins. Non-adhered cells were gently washed off with PBS (pH 7.4), leaving adherent phagocytes (monocytes, neutrophils, and macrophages). The N B T (Sigma) solution (0.2% in 0.85% saline) was similarly dropped into each well, and the slide incubated for another 30 mins. After this incubation, a coverslip was placed over the slide and the adherent cells examined microscopically under oil immersion (lOOOx magnification). At least 100 cells from each well (therefore 200 cells per sample) were counted. Those cells that appeared to have the morphological characteristic of monocytes, macrophages, or neutrophils, and had a blue halo, were counted as N B T positive. The blue halo formed was formazan precipitate produced by the reduction of N B T by 0" 2. The assay results were expressed as phagocyte respiratory burst activity (% of cells N B T positive). H E M A T O L O G Y Blood sampling and serum/plasma collection. Blood was collected from small fish (<25 g) by severing the caudal peduncle and collecting drips of blood into a sterile tube. Larger fish were bled from the caudal vessel using a syringe. To minimize the rupture of erythrocytes, the needle was removed from the syringe and the blood gently placed into a sterile tube. If serum rather than 28 plasma was to be collected, non-heparinized syringes were used, and blood was placed in glass rather than plastic tubes. Blood was left at room temperature for approximately 1 h, then 5 °C for 4 h before the serum or plasma was collected by centrifuging the packed blood cells (2000 x g at 5 °C for 3-4 min). Aliquots of serum/plasma were then placed into sterile tubes and frozen at -70 °C for later analysis. Hematocrit measurements. Fresh blood was collected in heparinized capillary tubes and centrifuged (10,000 x g for 5 min) in a microhematocrit centrifuge. Hematocrit values were expressed as the percent packed cell volume (% P C V ) . Differential leucocyte counts. Fresh blood (approx. 25 uL) was smeared onto a clean glass slide, and allowed to air dry. The blood smears were then stained with a modified Wright-Giemsa stain (Difquick ®, Dada Diagnostics P R . , U S A ) according to the manufacturer's instructions. Ten fields were systematically examined from each slide under oil immersion (lOOOx magnification), and the number of lymphocytes, neutrophils, monocytes, and thrombocytes counted. Differential leucocyte ratios were determined by counting the number of erythrocytes in each field (approximately 500 per field), along with the number of each leucocyte type. The number of each leucocyte type was then calculated, and expressed as a ratio of 101 erythrocytes (RBC). The actual number of the different leucocytes present (number per mm" blood) was calculated using the above ratio, and multiplying it with the actual number of erythrocytes (determined from separate erythrocyte counts). 29 C H A P T E R ONE C O M P A R I S O N OF C H I N O O K S A L M O N {Oncorhynchus tshawytscha) STRAINS FOR D I F F E R E N C E S IN NON-SPECIFIC I M M U N I T Y INTRODUCTION Traditional methods to control and prevent disease in cultured fish involve the use of chemicals, antibiotics, and vaccines. These methods are not always effective and can be very costly. In addition, the use of chemotherapeutics has raised concerns about environmental pollution and the development of drug resistant pathogens. A n alternative approach to disease control, is the development of a breeding program designed to selectively produce fish resistant to a wide variety of diseases. Genetic variation in disease resistance is required for such a breeding program. There is increasing evidence that genetic differences in disease resistance exist in fish (reviewed by Chevassus and Dorson 1990, and Fjalestad et al. 1993). However, the majority of published reports on the variation of disease resistance have been based on differences in survival rate, and as a result little information is available on the mechanisms responsible for the observed genetic differences. Beacham and Evelyn (1992a) suggest that the non-specific immune system plays an important role in innate disease resistance. The non-specific immune system provides general protection against all types of pathogens. Physical barriers (i.e., skin and scales), soluble factors (i.e., lysozyme and alternative complement system), and cellular factors (i.e., neutrophils and macrophages) comprise the non-specific immune system. Genetic variation in such non-specific factors as lysozyme activity and complement activity have been reported in Atlantic salmon (Roed et al. 1989, 1992, 1993, Lund et al. 1995). There also appears to be an apparent 30 relationship between the activity of the non-specific immune system (lysozyme activity and alternative complement activity) and disease resistance (as measured by survival rates) (Rjaed et al. 1993). In this study, I compared the non-specific immune system of three strains of chinook salmon with previously identified differences in disease resistance. Beacham and Evelyn (1992a) performed three different disease challenges and found the Kitimat strain consistently had the highest survival rate when compared to the Quinsam and Nitinat strains. The objective of this study was to test the hypothesis that the non-specific immune system plays a role in the enhanced innate disease resistance seen in the Kitimat strain of chinook salmon. Kitimat, Quinsam, and Nitinat chinook strains were compared for differences in serum lysozyme activity, serum bactericidal activity and differential leucocytes counts. M A T E R I A L S A N D M E T H O D S Fish and Sampling. The fish used for this study were the surplus (unchallenged) fish from the disease challenges performed by Beacham and Evelyn (1992a). Details of the mating and rearing of these strains are provided in Beacham and Evelyn (1992a). At the time of sampling, the fish strains had been maintained in a communal tank for approximately one year, and were a mean weight of 100 g. Forty fish from each strain were randomly sampled from the tank. Fish were killed with a lethal dose of 2-phenoxyethanol, weighed and measured. Blood was collected from the caudal vessel (after severing the caudal peduncle with a scalpel), into two hematocrit tubes. The remaining blood was placed in sterile test tubes and chilled. Serum was collected from the chilled blood, aliquoted, and frozen at -80 °C for later analysis of V. anguillarum (Vang) antibody titres, bactericidal activity against Vang, and lysozyme activity. 31 V. anguillarum Antibodies. Serum samples (40 per strain) were thawed on ice and the presence of Vang antibodies determined using the plate agglutination titre method (Roberson 1990). Lysozyme Activity. The lysoplate method was used to determine serum lysozyme activity (described in the General Materials and Methods section). The serum samples (40 per strain) were assayed in triplicate and the mean activity expressed in U/mL. Bactericidal Activity. Ten serum samples from each strain were randomly chosen and assayed in duplicate for bactericidal activity against Vang. The serum samples were thawed on ice, and kept chilled until the bacterial suspension was added. A suspension of virulent Vang (5 x 103 viable cells/mL actual counts) was prepared in peptone-saline (P-S; 0.1%/0.85% peptone/NaCl, respectively) as described in the General Materials and Methods section. Equal volumes of serum and the bacterial suspension were combined, mixed and incubated at room temperature (18-20 °C), on a rotator for 90 min. Controls were similarly prepared by replacing the serum with P-S. After incubation, the samples were placed on ice, serially diluted in P-S, and drop-inoculated onto TSA. Following 24-48 h incubation at room temperature, the Vang colonies were counted, and the mean number of viable cells/mL determined. Bactericidal activity was determined as the percent survival of bacteria relative to the P-S control. Hematology. Methodology described in the General Materials and Methods section was followed for the measurement of differential leucocyte ratios (n=15 per strain) and hematocrits (n=40 per strain). Ten of the fifteen samples examined for differential leucocyte ratios had been previously assayed for bactericidal activity. The additional five samples were selected at random. Data analysis. Analysis of variance tests were used to compare the strains for significant differences. Student-Newman-Keuls multiple comparison tests were then used to identify which strains were different. Data expressed as percentage (hematocrit) and ratios (differential .32 leucocyte ratios) were arcsin square-root transformed, and the lysozyme data were log transformed, before the above analyses were carried out. Correlations between the immune parameters were determined using the Pearson product moment correlation tests. Statistical significance was determined for all tests where p<0.05. RESULTS A N D DISCUSSION There was significant variation between the strains in serum bactericidal activity (p<0.001) (Figure 1 -1 a). The Kitimat strain had the highest serum bactericidal activity, as significantly fewer Vang remained viable following the 90 min incubation period. Antibodies to Vang were not detected in the serum of any of the three chinook strains. It was therefore concluded that the serum killing activity was due to non-specific factors. Lysozyme, complement, transferrin, C-reactive protein and agglutinins are some of the non-specific substances known to have bactericidal/bacteriostatic activities in fish (Alexander and Ingram 1992, Yano 1996). Serum lysozyme activity varied significantly between the three strains (p<0.001) (Figure 1-1 b). The Kitimat and Nitinat strains had significantly higher lysozyme activity when compared with the Quinsam. Kitimat lysozyme activity was higher than the Nitinat, but this difference was not statistically significant (p>0.05) when all three strains were compared for differences. There appeared to be a positive relationship between bactericidal activity and lysozyme activity, but the correlation between the two factors was not significant (p>0.05). Serum complement may contribute to serum bactericidal activity, thereby masking the effects of lysozyme alone. Although fish lysozyme has been shown to possess lytic activity against Gram negative bacteria in the absence of complement (Grinde 1989), other reports suggest that complement is required for fish lysozyme to exert its lytic effect (Hjelmeland et al. 1983). Koppenheffer (1987) reports that serum bactericidal activity is primarily due to the non-specific or 3.1 alternative complement pathway. Complement and other non-specific bacteriolytic substances were not measured in this study, and therefore their role in the observed serum bactericidal activity is unknown. Strain differences in hematocrit and differential leucocyte ratios were not significant, (p>0.05) (Table 1-1). Correlations between serum lysozyme activity and neutrophils were also not significant. It was thought that the strain differences in serum lysozyme activity may be due to differences in the activity or amount of lysozyme produced by the neutrophils. Neutrophils synthesize and secrete lysozyme (Murray and Fletcher 1976), and increases in their number have been associated with elevated serum lysozyme activity (Muona and Soivio 1992). These results indicate that the Kitimat strain of chinook salmon possess an enhanced non-specific immune system when compared with the Nitinat and Quinsam strains. The Kitimat strain of coho had significantly higher serum bactericidal activity than Nitinat and Quinsam coho. In addition, the serum lysozyme activity was significantly higher in the Kitimat coho when compared with the Quinsam strain of coho salmon. These results support the survival data and superior innate disease resistance of the Kitimat coho, as previously reported by Beacham and Evelyn (1992a). Strain variation in innate disease resistance appears to be related to the activity of the non-specific immune system. 34 Table 1-1. Comparison of hematological values (means ± one standard error) between three strains of chinook salmon (Oncorhynchus tshawytscha). No significant differences between the strains were detected by analysis of variance tests (p>0.05). Kitimat strain Nitinat strain Quinsam strain Lymphocytes (No./10 ? RBC) ' 1 12.5 (1.58) 9.00 (1.32) 11.2 (2.09) Neutrophils (No. / lO 3 RBC) ' 1 0.93 (0.25) 0.92 (0.36) 0.97 (0.31) Thrombocytes (No./10 ? R B C ) a 0.45 (0.26) 0.71 (0.39) 0.87 (0.34) Total Leucocytes (No./10 ? R B C ) a 13.9 (1.51) 10.63 (1.24) 13.0 (2.40) Hematocrit (% P C V ) ' 1 50.9 (0.95) 49.0 (1.02) 51.0 (1.21) a n=15 b n=40 35 Chinook Strains: 1 1 Kitimat V77A Nitinat OOPfl Quinsam Figure 1-1. Comparison of non-specific immunity between three strains of chinook salmon (Oncorhynchus tshawytscha). (a) Mean serum bactericidal activity as measured by the percent survival (relative to control) of Vibrio anguillarum (Vang) (n=10 per strain); and (b) mean serum lysozyme activity (n=40 per strain). Different letters refer to significant differences (p<0.05) between strains. Error bars represent one standard error. 36 C H A P T E R T W O STRAIN D I F F E R E N C E S IN NON-SPECIFIC I M M U N I T Y OF T I L A P I A Oreochromis niloticus F O L L O W I N G C H A L L E N G E W I T H Vibrio parahaemolyticus I N T R O D U C T I O N The selection of disease resistant fish strains is one approach to improving the survival of cultured fish. Strain differences in disease resistance have been found for a variety of salmonid diseases (Gjedrem and Aulstad 1974, Bakke et al. 1990, Withler and Evelyn 1990, McGeer et al. 1991, Ibarra et al. 1991). Reports of strain differences in disease resistance of non-salmonid fishes, are lacking in the literature. The carp (Cyprimis carpio) has been the dominant non-salmonid fish used for studying disease resistance (reviewed by Wiegertjes et al. 1996). To date there have been no reports on the genetic variation of disease resistance or non-specific immunity in the tilapia {Oreochromis spp.). Nile tilapia represent one of the world's most important food fishes, in addition to being an important species for aquaculture. Natural selection for enhanced innate, non-specific immunity has been implicated in the evolution of strain differences in disease resistance (Beacham and Evelyn 1992b). The objective of this study was to compare the non-specific immune system of red and black strains of the Nile tilapia. These two strains of royal (Bangkok Dusit Palace, Thailand) Nile tilapia were chosen for a comparison because of their genetic purity (McAndrew and Majumdar 1983). The red colour is an autosomal dominant trait, whereby the development of melanophores is inhibited (McAndrew et al. 1988). The strains of tilapia used in this study were reared under identical environmental conditions; therefore environmental effects are assumed to be negligible. Strain differences in selected components of the non-specific immune system were compared in saline and V. 37 parahaeivolyticus challenged fish. The marine pathogen V. parahaemolyticus was used for the disease challenge, to ensure these freshwater fish would be immunologically naive to the bacteria, and therefore allow a more meaningful measurement of the non-specific immune system. M A T E R I A L S A N D M E T H O D S Fish. Approximately 100 fish from each of the communally-reared red and black strains of royal tilapia were collected from the Malaysian Department of Fisheries fish breeding station in Jitra, Malaysia. Fish were lightly anaesthetized (40 ppm MS222, tricaine methanesulphonate containing 0.5% NaCl) during the transport to the freshwater holding facilities at the Universiti Pertanian Malaysia (UPM) . Upon arrival at U P M a prophylactic antibiotic bath (10 ppm furizolidone) was administered to treat a dermal infection caused by abrasions incurred during transport. The tilapia (135 g mean weight) were placed into a stock tank (29 °C) and acclimated for 2 wks. The fish were fed a commercial diet ad libitum. Bacterial challenge. The challenge bacterium was the Gram-negative pathogen, V. parahaeivolyticus, originally isolated from an infected sea bass (Dicentrarchus labrax) at a cage culture operation in Malaysia ( U P M isolate 7/95). The virulence of the pathogen and challenge dose used for this study were determined from a preliminary LD50 experiment. Challenge inoculum was prepared in sterile phosphate buffered saline (PBS, pH 7.2). After 24 h of growth, V. parahaemolyticus was aseptically removed from tryptic soy agar (TSA) plates supplemented with 1.5% NaCl . The challenge dose was estimated by measuring the absorbance of the bacterial suspension at 700 nm. To confirm the actual challenge dose, aliquots of the challenge dilutions were inoculated onto T S A plates and the colonies counted after a 24 h incubation at 35 °C. The actual challenge dose was then determined as the number of colony forming units (cfu) per mL of PBS 38 Following the acclimation period, the tilapia were removed from the stock tank and randomly distributed into 4 tanks as follows: tank 1:18 red, 18 black; tank 2: 17 red, 17 black; tank 3:18 red, 18 black; tank 4: 17 red, 17 black. The following day, the fish were anaesthetized (MS222, 200 ppm) and challenged by intraperitoneal (ip) injection. Fish from tanks 1 and 2 (infected group), were injected with 7.35 x 108 cfu V. parahaemolyticus (actual dose) in 1 mL sterile PBS. The fish from tanks 3 and 4 (sham group), were injected with 1 mL sterile PBS. Mortalities were collected daily for one week. Each dead fish was necropsied, and kidney and liver tissue cultured onto TSA plates supplemented with 1.5% NaCl. The cause of death was presumed to be V. parahaemolyticus if the culture was found to consist of Gram-negative rods, sensitive to the vibriostat 0/129 (2,4-diamino-6,7-diisopropylpteridine), that produced green colonies when subcultured onto thiosulfate-citrate bile sucrose (TCBS) agar. Sampling. A pre-challenge sample (10 red and 10 black) was taken from the stock tank, one day prior to the challenge. At 18-24 h post-challenge, 7 red and 7 black tilapia were randomly removed from each tank, for a total of 14 fish per strain for each group (infected and sham). The remaining 10 or 11 fish in each of the tanks were left and mortalities monitored for the following week. Blood/serum collection. The sampled fish were euthanized in a lethal dose of MS222, weighed and blood taken from the caudal vessel using a syringe. Blood smears were prepared onto cleaned glass slides for differential leucocyte counts. The remaining blood was placed in test tubes, left for 1 h at room temperature, then 4 h at 4 °C. The tubes were then centrifuged at 2000 x g at 5 °C, for 3-4 min. The serum was collected and frozen at -20 °C for later analysis of lysozyme activity. Assays of non-specific immunity. The respiratory burst activity of anterior kidney phagocytes was determined using a modified nitroblue tetrazolium (NBT) assay (described in the 39 General Materials and Methods section). Briefly, the anterior kidney was aseptically dissected from each fish and placed into 250 uL Leibovitz medium (L- l 5). The kidney was homogenized and 50 uL aliquots were dispensed into four wells on a multiwelled pre-cleaned glass slide. The slides were incubated in a moist chamber at 25 °C for 30 min. Non-adherent cells were gently rinsed off the slide with PBS. NBT (0.2% in 0.85% sterile saline) was dropped (approx. 50 uL) into each well, and the slide was incubated for 1 h in the moist chamber. After incubation, a coverslip was placed over the slide and the slides were examined under oil immersion (lOOOx). A total of 200 adherent phagocytes were counted per sample (50 per well), and the proportion of blue (i.e., active) phagoctyes of the total was calculated. The lysoplate method was used to determine serum lysozyme activity (see General Materials and Methods section). Lysozyme activity was expressed as ug hen egg white lysozyme/mL serum. The serum samples were assayed in triplicate, and the mean used for all statistical analyses. Blood smears prepared from fresh blood were air dried, fixed in 95% methanol for 10 min, and stored at 5 °C. The slides were stained in May-Grunwald (0.25% w/v in methanol) for 10 min, and 15 min at half strength (diluted with deionized distilled water, ddH20). They were then counter-stained in Giemsa (1:10 in ddH20) for 30 min. The slides were rinsed in ddH20 for 5 min and left to air dry. The slides were examined under oil immersion (lOOOx). Approximately 100 leucocytes were counted and classified as lymphocytes, thrombocytes, neutrophils and monocytes. The relative percent of each leucocyte type was calculated. Data analysis. Student /-tests were performed to compare the pre-challenge data with the sham, saline challenge data. Analysis of variance (2-way A N O V A ) tests were used to examine strain (red and black strains) and infection (sham and infected groups) effects (Sokal and Rohlf 1981). Student-Newman-Keuls multiple comparisons tests were used to identify which 40 groups were different. The proportion data (NBT and leucocyte counts) were arcsin square-root transformed and lysozyme data were log transformed, before the analyses. Percent cumulative mortalities for the two strains were compared using Fischer's exact test. Statistical significance was noted for all tests where p<0.05. R E S U L T S A N D DISCUSSION This study documents for the first time that significant strain-related differences in non-specific immunity occur in Nile tilapia. Mortalities began to occur in the V. parahaemolyticus-challenged groups shortly after the post-challenge samples had been taken (at approximately 24 h post-challenge). While there was a two-fold difference in mean mortality rates of the two strains, the percent cumulative mortality for each of the two strains did not differ significantly (p>0.05). The black strain had 28.9% mortality (6 dead/21 total), while the red strain 14.3% mortality (3 dead/21 total). No mortalities occurred in any of the sham challenged groups. The pre-challenge values for lysozyme, phagocyte respiratory burst activity, and leucocyte numbers were not significantly different from the sham (saline challenge) group. The data analyses were then focused on analysis of variance tests examining infection (sham vs. infected groups) and strain (red vs. black) effects. There were significant strain effects on serum lysozyme activity (<0.005) (Figure 2-la). The significant strain effects in lysozyme activity in tilapia, is consistent with the findings of Lund et al. (1995) and Roed et al. (1989, 1993), who showed that there was a genetic basis for variation of lysozyme activity in salmonids. There were no significant effects of infection on lysozyme activity (p>0.05). It is possible that there was no significant infection effect on lysozyme activity because the number of neutrophils did not increase significantly following the challenge (Table 2-1). Neutrophils synthesize and secrete lysozyme 41 (Murray and Fletcher 1976), and increases in serum lysozyme activity have been associated with increases in their numbers (Muona and Soivio 1992). There were no significant differences in the relative numbers of thrombocytes and monocytes, between either strains or challenge treatments. The percent of circulating lymphocytes from both strains, decreased in the infected groups, and showed a significant effect of infection (Table 2-1). Lymphopenia has been reported in fish infected with V. anguillarum, and has been attributed to the migration of lymphocytes to tissues and the destruction of lymphocytes by the bacteria (Harbell et al. 1979, Lamas et al. 1994) There were significant infection (p<0.001) and strain (p<0.005) effects on the respiratory burst activity of the anterior kidney phagocytes (Figure 2-lb). There was however, no significant (p>0.05) interaction between the effects of infection and strain. Both tilapia strains showed a significant increase in phagocyte respiratory burst activity due to the V. parahaemolyticus challenge. Respiratory burst is one of the most important bactericidal mechanisms in fish (Secombes and Fletcher 1992), and may reflect the ability of the fish to protect itself against infection. Similar increases in phagocyte respiratory burst activity occur in coho salmon during the initial stages of infection (Balfry et al. 1994). Despite strain differences in serum lysozyme and phagocyte respiratory burst activity, because mortality rates were so low, it was not possible to determine whether there was a correlation between these non-specific immune parameters and disease resistance (as measured by mortality). Perhaps with a larger sample size, a correlation between these parameters and disease resistance would have been statistically significant. Other researchers have reported a significant correlation between lysozyme and survival (Fevolden et al. 1991, Lund et al. 1995). Further research is needed to compare these and other tilapia strains for differences in non-specific immunity and disease resistance before selection programs could be utilized. The significant strain differences in serum lysozyme activity and phagocyte activity reported here, suggest there may be strain differences in disease resistance in Nile tilapia that could be used to help improve its survival under culture conditions. 43 Table 2-1. Nile tilapia (Oreochromis ni lotions) strain comparison of mean (± one standard error) differential leucocyte types, 18-24 h after intraperitoneal injection with saline (sham group) or 7.38 x 10 s Vibrio parahaemolyticus (infected group). Saline injected (n=14) V. parahaemolyticus injected (n=l4) Red strain Black strain Red strain Black strain % Lymphocytes 36.1 ± 5.57 38.5 ± 4.08 24.9 ± 4.26 * 26.4 ± 3.93* % Thrombocytes 56.0 ± 5.31 53.7 ± 4.12 63.0 ± 5.78 63.6 ± 3.61 % Neutrophils 3.21 ± 1.43 3.76 ± 1.77 6.66 ± 2.08 7.12 ± 1.83 % Monocytes 4.68 ± 2.25 4.02 ± 1.70 5.38 ± 2.09 2.87 ± 1.32 * Significant differences due to infection (p<0.05) as detected by one way A N O V A analysis. 44 " (a) Figure 2-1. Nile tilapia (Oreochromis niloticiis) strain comparison of mean: (a) serum lysozyme activity; and (b) anterior kidney phagocyte respiratory burst activity, 18-24 h after challenge with Vibrio parahaemolylicvs (Vpar) or saline. Different letters refer to significant differences (p<0.05) between means. Error bars represent one standard error. 45 C H A P T E R T H R E E C O H O S A L M O N (Oncorhynchus kisutch) STRAIN D I F F E R E N C E S IN NON-SPECIFIC I M M U N I T Y F O L L O W I N G A N O N - L E T H A L C H A L L E N G E W I T H Vibrio anguillarum I N T R O D U C T I O N Results from previous chapters indicate a role of the non-specific immune system in innate disease resistance. The studies with chinook salmon and Nile tilapia showed significant strain differences in non-specific immune factors, such as serum lysozyme activity, serum bactericidal activity, and phagocyte respiratory burst activity. Strain differences in the activity of the non-specific immune system appear to be related to innate disease resistance, because there was an apparent trend for the most resistant strain to have the highest activity of the non-specific immune factors measured. The following study examines a different fish species for strain differences in innate disease resistance and non-specific immunity. Coho salmon strains were chosen for the following comparison based on previous work published by Withler and Evelyn (1990) suggesting the superior innate disease resistance of Kitimat coho to bacterial kidney disease. There have been no published reports, however, of significant coho strain differences in non-specific immunity. The purpose of this study was therefore to demonstrate strain differences in the non-specific immune system of coho salmon following a disease challenge, and thus provide further evidence to support the hypothesis that the non-specific immune system plays a significant role in innate disease resistance. Determining the role of the non-specific immune system in innate disease resistance may provide useful information for breeding programs designed to selectively produce fish strains with enhanced disease resistance. 46 The Kitimat and Quinsam strains of coho salmon were compared in this study. The marine pathogen V. anguillarum (Vang), was used to challenge these freshwater reared coho salmon which were assumed to be immunologically naive to this pathogen. Various hematological and immunological parameters were measured prior to the disease challenge, and at d2 and d 16 post-challenge. The effects of strain and challenge treatment (i.e., infection) were examined for each parameter measured. M A T E R I A L S A N D M E T H O D S Fish. In November 1991, coho salmon broodstock were collected from the Kitimat River hatchery and the Quinsam River hatchery. Details of the mating and rearing of the two coho strains are provided in the General Material and Methods section. The Kitimat and Quinsam strains were represented by approximately equal numbers of fish from 9 and 7 full-sib families, respectively. Both strains were communally reared from 3 g mean weight (August 1992) until the time of the challenge experiment (November 1992). The water temperature was 7-9 °C during the course of the challenge experiment. Experimental Design and Sampling. Fish in the communal tanks were starved the day before the sampling and disease challenges were performed. On 21 November 1992, 140 fish from each strain (12 g mean weight) were randomly removed from the communal tanks and placed into three adjacent experimental tanks as follows: tanks 1 and 2 were designated Vang challenge tanks and each contained 60 fish per strain, tank 3 was a sham challenge (control) tank containing 20 fish from each strain. Ten fish per strain were sampled from the communal tank on the day of the challenge (dO) to determine the baseline or pre-challenge values for the variables measured. Each tank was then challenged separately as described in the General Materials and Methods section. The challenged fish (Vang and sham) were randomly sampled on d2 and d 16 47 post-challenge (pc). At each sample time, 6 fish per strain were removed from each tank for a total sample size of 12 for the Vang challenged (infected) group and 6 for the sham challenge (uninfected control) group. The sampled fish were euthenized in a lethal dose of MS222 (tricaine methanesulphonate), and the weights and strain (fin clip) were recorded. The tails were severed and blood was collected into two heparinized capillary tubes for hematocrit measurements. The plasma in the hematocrit tubes was collected, and stored at -70 °C for later analysis of lysozyme activity. Drops of whole blood were smeared onto cleaned glass microscope slides to determine differential leucocyte ratios. Erythrocyte numbers (dO, and d2 only), hemoglobin (dO, and d2 only), plasma protein (all samples), and plasma glucose (dO, and d2 only) concentrations were measured as described below. Immediately after each fish was bled, the anterior kidney was aseptically removed and placed in 1 mL sterile, chilled Leibovitz medium (L-15) for the phagocyte respiratory burst activity assay. Disease Challenge. The immersion challenge procedure has been outlined in the General Materials and Methods section. The actual challenge dose was calculated as 1.25 x 108 viable Vang cells/mL. Hematology. Hematocrit values and differential leucocyte ratios were determined as outlined in the General Materials and Methods section. The actual number of each leucocyte type was calculated for the dO and d2 samples using the leucocyte to R B C ratio and the number of erythrocytes (RBCs). A minilab portable instrument was used to determine R B C numbers (dO and d2), hemoglobin (dO and d2), plasma protein (all samples), and plasma glucose (dO and d2) concentrations (Iwama et al. 1995). Plasma Lysozyme Activity. The lysoplate method was used to determine lysozyme activity (see General Materials and Methods section). Lysozyme activity in the plasma, kidney 48 and gill tissue was measured in triplicate, and activity expressed as mean U/mL (plasma) and U/g (kidney and gill). Phagocyte Respiratory Burst Activity. Respiratory burst activity in phagocytes from the anterior kidney was determined by measuring the production of superoxide anion (0 2 ) . The glass adherent nitroblue tetrazolium (NBT) assay was used (described in the General Materials and Methods section). Briefly, the anterior kidney tissue was homogenized in LI 5 medium using the rubber end of the plunger insert from a sterile 1-cc syringe. The NBT assay was performed on the kidney cell suspension, which was run in duplicate for each sample. The percent of NBT positive phagocytes (i.e., phagocytes producing 02") out of the total number of phagocytes was calculated. Data Analyses. Data expressed as percentage (respiratory burst activity and hematocrit) were arcsin square root transformed, the lysozyme data were log transformed, and the blood cell counts were square root transformed (Sokal and Rohlf 1981). Student /-tests were used to compare strains for differences in the pre-challenge (dO) sample. Analysis of variance (2-way ANOVA) tests were used to analyze the d2 and dl6 data, to examine the effects of strain and infection, as well as interactions between strain and infection. Following the A N O V A tests, Student-Newman-Keul multiple comparison tests were used to determine which groups were different. Statistical significance for all tests were determined where p<0.05. R E S U L T S Vibriosis-related mortality was not observed in any of the Vang or sham challenge groups. There were, however, significant strain- and infection-related differences detected for some of the variables that were measured. 49 The Kitimat strain of coho had significantly higher (p<0.05) plasma lysozyme activity at dO (pre-challenge) (Figure 3-1). No strain differences were seen in gill or kidney lysozyme activity at dO. Examination of strain and infection effects at d2 and d 16 post-challenge (using 2 way A N O V A analysis), in the plasma, gill, and kidney reveal significant strain effects in the kidney tissue at d2. At that time, the Kitimat strain had significantly higher kidney lysozyme activity than the Quinsam strain of coho, in both the Vang and sham challenge treatment groups. Similar post-challenge changes were observed in the lysozyme activity of the plasma, gill, and kidney samples from both the Vang and sham challenged groups. The general pattern was for lysozyme activity to increase from dO to d2, then at d 16 decrease (sham challenged groups) or remain unchanged (Vang challenged groups). The challenge treatments (sham or Vang) produced significant effects on the activity of gill lysozyme at d l6 , and kidney lysozyme at d2 and d l 6 post-challenge. For these samples, the sham challenge appeared have the effect of increasing lysozyme activity at d2, and decreasing lysozyme activity at d l 6 post-challenge. There were no significant strain differences in the phagocyte respiratory burst activity at any of the sample times (Figure 3-2). At d l 6 post-challenge, however, the Kitimat strain showed a much higher level of respiratory burst activity in the Vang-challenge group, when compared with the Quinsam strain. The d l 6 data, revealed significant (p<0.05) challenge treatment effects and interaction effects between strain and challenge treatment. This indicates that strain differences depend on the challenge treatment (i.e., the presence of Vang). Phagocyte activity in the Vang challenge groups from each strain increased significantly (p<0.05) from d2 to d l6 , while the activity was unchanged (p>0.05) in the sham challenged groups over the same time period (not shown on Figure 3-2). Plasma glucose concentrations were not significantly different between the two strains of coho prior to the challenge (dO) (Figure 3-3). However, at d2 post-challenge, the concentration 50 of plasma glucose increased significantly in the Quinsam strain, in both challenge treatment groups. Plasma glucose concentrations in the Kitimat strain did not increase from dO to d2, and as such were found to be significantly lower than the Quinsam strain. Plasma protein concentration was the only hematological parameter that showed any significant effect of strain or challenge treatment (Table 3-1). The Kitimat strain of coho salmon had higher plasma protein levels when compared with the Quinsam strain, in both the sham and Vang challenge treatment groups. The presence of Vang, appeared to have a significant effect on plasma protein levels, as both strains showed a significant decrease in plasma protein in the Vang challenge groups when compared with the sham challenge group. Hemoglobin concentration, hematocrit, wbc ratio, wbc counts, and R B C counts did not show any significant strain or challenge treatment differences (p>0.05). The small sample size and large variation associated with these variables, may explain the inability to detect significant differences. DISCUSSION The Vang challenge performed in this study, failed to produce any vibriosis-related mortality, despite a rather high challenge dose of virulent Vang. The reason may be the low water temperature at the time of challenge (7-9 °C). Water temperatures less than 12 °C have been shown to suppress the /'/; vivo growth of Vang (Groberg et al. 1983). If this was the case, the immune system of the fish may have been able to overcome the bacteria, to prevent a fatal f s ^infection from occurring. The lysozyme and phagocyte respiratory burst activity results (i.e., the post-challenge increases) reported in this study, indicate that the non-specific immune system did respond to the initial infection, but because there was no mortality and a minimal physiological post-challenge response, it can be assumed that a diseased state (i.e., vibriosis) did not develop,in the Vang challenged fish. U 51 Lysozyme activity in the different tissues from both strains, showed similar post-challenge changes. Lysozyme activity was greatest in the kidney, followed by the gill and plasma. Lie et al. (1989) compared lysozyme activity in various tissues of rainbow trout and also found the highest activity in the kidney. They attributed the results to the kidney tissue having high concentrations of monocytes, macrophages and polymorphonuclear granulocytes, which synthesize and secrete lysozyme (Murray and Fletcher 1976). The post-challenge change in lysozyme activity was consistent for the different tissues, strains, and challenge treatment groups. The increase in lysozyme activity from dO to d2 may be attributed to stress and infection. The handling and confinement associated with the challenge procedure may have produced a mild stress, which can elevate lysozyme activities (Mock and Peters 1990). It is not clear why the sham groups showed a greater increase from the dO lysozyme levels, than the Vang-challenge groups. The decrease in lysozyme activity observed at d 16 in the sham challenged groups and not in the Vang challenged groups, may be related to the presence of the pathogen (i.e., an infection). The pathogen, and therefore the source of the stimulation, was absent in the sham groups, supporting the possibility that the initial increase in lysozyme activity was related to a stress response. Significant strain differences in lysozyme activity were found in the plasma at dO, and in the kidney tissue at d2 post-challenge. At both times, the lysozyme activity was higher in the Kitimat strain when compared with the Quinsam strain. The Kitimat strain has previously been shown to have higher lysozyme activity than the Quinsam strain (Chapters One and Four). There was a trend for Kitimat fish to have higher lysozyme activity in the other tissues and at other sample times, however, significant strain effects were not detected. The uniformity in the post-challenge changes in gill, kidney and plasma lysozyme activity, suggests that the monocytes, macrophages and polymorphonuclear granulocytes distributed throughout the entire animal, share a common response when stimulated by foreign substances such as bacterial pathogens. The greatest increase in lysozyme activity was seen in the kidney tissue, possibly because there are more lysozyme-producing cells there, than in the plasma or gill tissue. The post-challenge change in phagocyte respiratory burst activity was similar to that observed in lysozyme activity. The d2 increase in activity may have been stress related, while the d 16 levels appear to be related to the presence of the pathogen. Transient increases in phagocyte activity (as measured by superoxide production) have been associated with a confinement stress which is similar to the challenge procedure (Pegg et al. 1995). The decreased respiratory burst activity observed at d l 6 in the sham challenge groups, may have been because the phagocytes were no longer being stimulated (i.e., the fish had recovered from the stress and there was no pathogen present). In contrast, the phagocytes in the Vang challenge groups remained active at d 16, possibly because they continued to be stimulated by Vang. Significant challenge treatment effects at d 16, suggests that the presence of the pathogen was having an impact on the activity of kidney phagocytes. The Quinsam strain had significantly higher plasma glucose levels at d2, while the d2 glucose levels in the Kitimat fish did not change from pre-challenge levels. This suggests that the Quinsam strain may be more sensitive to the stress associated with the challenge procedures. Elevations in plasma glucose are considered to be secondary metabolic effects of stress (Barton and Iwama 1991). After a stressful event, plasma glucose levels have been shown to take up to 72 h to return to baseline (Pickering et al. 1982). Stress susceptibility has been associated with immunosuppression and a decreased ability to resist infections (Maule et al. 1989, Schreck 1996). It is therefore possible that the sensitivity of the Quinsam fish to stress, may be related to the lowered survival rate of this strain (as compared with Kitimat strain) when challenged with 53 different pathogens (Beacham and Evelyn 1992a). Clearly further work to investigate strain differences in stress response, and their effects on disease resistance is needed. The challenge treatment had a significant effect on plasma protein concentrations in both strains at d2 post-challenge, with plasma protein concentrations being lower in the Vang challenged groups. Decreases in plasma protein concentrations may be attributed to an infection (Wedemeyer et al. 1990), which is consistent with the above scenario which proposes the presence of a sublethal vibriosis infection in the Vang challenged groups of fish. The inability to detect significant strain and/or challenge treatment effects on hematocrits, wbc ratios, wbc numbers or R B C numbers, may be related to the inability to create a diseased state in the Vang challenged fish. The increase in lysozyme and phagocyte respiratory burst activity, suggests a stimulation of the non-specific immune system. However, the effectiveness of these factors, combined with the temperature induced suppression of bacterial growth, appear to have resulted in a clearance of the pathogen so a diseased state was not achieved. Fevolden et al. (1992) reported that strain differences in the activity of the non-specific immune system were most evident in the early stages of the infection. A quick and effective non-specific immune system is crucial in preventing an infection, and therefore disease-related mortality. Significant strain differences in the activity of the non-specific immune system were observed in this study, both prior to and during the early stages of an infection. The results indicate that even i f pre-challenge strain differences do not exist in such non-specific immune parameters as lysozyme and respiratory burst activity, it does not mean that these parameters do not have an important role in innate disease resistance. Strain differences in the activity of these variables may become more evident when the fish respond to a stressor or an infection. This would suggest that strain differences in innate disease resistance may be related to genetically based differences in the non-specific immune response to stimuli such as a stressor or an infection. This study shows a tendency for the Kitimat strain of coho salmon to have higher lysozyme and phagocyte respiratory burst activity, when compared with the Quinsam strain of coho. One may therefore predict that the Kitimat coho are likely to be more disease resistant than other coho strains. Table 3-1. Comparison of hematology in two strains of coho salmon (Oncorhynchus kisutch) immediately before and after a non-lethal Vibrio anguillarum (Vang) immersion challenge was performed. * Significant (p<0.05) strain and challenge treatment effects on plasma protein concentration at day 2 post-challenge. Day 0 Pre-challengea Kitimat Quinsam Day 2 Post-challenge Sham challengeb Vang challengec Kitimat Quinsam Kitimat Quinsam Day 16 Post-challenge Sham challengeb Vang challenge0 Kitimat Quinsam Kitimat Quinsam Hematocrit 40.6 43.0 42.3 40.6 43.7 38.3 36.7 36.8 38.7 37.3 (% PCV) (1.7) (1.7) (1.7) (1.2) (1.9) (1.1) (0.8) (0.6) (1.0) (1.4) Plasma Protein 7.02 7.22 7.60* 6.83* 6.77* 6.11* 6.27 6.70 7.05 6.57 (g/100 mL) (0.26) (0.34) (0.33) (0.27) (0.27) (0.13) (0.18) (0.36) (0.14) (0.24) Wbc Ratio 12.8 14.1 8.9 13.8 9.1 7.9 23.4 16.2 13.2 17.1 (\vbc/103rbc) (3.2) (3.0) (4.2) (4.5) (1.9) (1.6) (4.6) (2.6) (2.7) (3.5) Wbc Count 10.7 13.6 8.4 12.4 7.6 7.4 not not not not (x 103/mm3) (2.5) (3.2) (3.6) (44) (2.1) (1.4) assayed assayed assayed assayed Rbc Count 0.84 0.95 1.04 0.91 0.98 0.98 not not not not (xl0 6/mm 3) (0.05) (0.05) (0.15) (0.07) (0.04) (0.06) assayed assayed assayed assayed Hemoglobin 6.69 7.36 7.55 8.05 8.93 8.02 not not not not (g/lOOmL) (0.26) (0.27) (0.16) (0.27) (0.88) (0.13) assayed assayed assayed assayed a n=10 r b n=6 r c n=12 56 0 4 8 12 16 0 4 8 12 16 Days Post-Challenge Days Post-Challenge Figure 3-1. Coho salmon (Oncorhynchus kisutch) strain comparison of the effect of a non-lethal immersion challenge with Vibrio anguillarum (Vang) on the mean lysozyme activity (± sem) in (a) plasma, (b) gill, and (c) kidney tissues. * represents significant differences between strains. # represents significant differences between sham and Vang challenge treatment groups. • represents significant different from the dO pre-challenge value. Significant difference determined where p<0.05. n=10 per strain at dO, n=12 and n=6 per strain for the Vang and sham challenge treatment groups, respectively. 57 O a, 3 o -t—I i— 'cL H> I— <D >^ O o Days post-challenge - e - Sham - Kitimat •R • Sham - Quinsam Vang - Kitimat • Vang - Quinsam Figure 3-2. Coho salmon (Oncorhynchus ki stitch) strain comparison of mean (± sem) percent anterior kidney phagocyte respiratory burst activity following a non-lethal Vibrio anguillarum (Vang) immersion challenge. # represents significant differences between sham and Vang challenge treatment groups. • represents significant different from the dO pre-challenge value: Significant difference determined where p<0.05. n=10 per strain at dO, n=12 and n=6 per strain for the Vang and sham challenge treatment groups, respectively. 58 I I Kitimat V//A Quinsam a T_ a X Sham Vang Post-challenge (d2) 200 175 -^ 150 S E o o 00 a 03 125 100 75 50 Pre-challenge (dO) Figure 3-3. Coho salmon (Oncorhynchus kisulch) strain comparison of mean (± sem) plasma glucose in pre-and post-challenge samples. Fish were immersion challenged with a non-lethal dose of Vibrio anguillarum (Vang). n=12 and n=6 per strain for the Vang and sham challenged groups respectively. n=10 for the pre-challenge group. Different letters represent significant differences between the means of the groups at d2 post-challenge (p<0.05). 59 CHAPTER FOUR COHO SALMON (Oncorhynchus kisutch) STRAIN DIFFERENCES IN INNATE DISEASE RESISTANCE AND ACTIVITY OF THE INTERNAL HUMORAL AND CELLULAR NON-SPECIFIC IMMUNE SYSTEM, FOLLOWING IMMERSION CHALLENGES IN DIFFERENT DOSES OF Vibrio anguillarum INTRODUCTION The previous three chapters have shown significant strain differences in the activity of the non-specific immune system, in chinook salmon, Nile tilapia, and coho salmon. However, a relationship between the activity of the non-specific immune system and actual disease resistance (as measured by strain differences in challenge mortality) was only implied, rather than demonstrated. This was because the challenges produced only low (Chapter Two) or no mortality (Chapter Three), so that strain differences could not be detected. There have been published reports demonstrating the Kitimat strain of coho and chinook are more disease resistant than other strains of these species (Withler and Evelyn 1990, Beacham and Evelyn 1992a), and the results presented thus far indicate that this superior innate disease resistance may be related to a more active non-specific immune system. The goal of this thesis was to produce evidence to support the hypothesis that the non-specific immune system plays a significant role in innate disease resistance, and in doing so, provide information on the underlying mechanisms for the genetic differences in disease resistance. The studies reported in this chapter were designed in an attempt to establish a relationship between innate disease resistance (i.e., mortality differences between strains) and the activity of the non-specific immune system, without relying on other researchers' reports of strain differences 60 in mortality. The experimental approach was to obtain mortality data following a lethal disease challenge, and demonstrate a relationship between strain differences in mortality with differences in the activity of the non-specific immune system. To determine some of the underlying mechanisms of innate disease resistance, the study focused on the internal humoral and cellular non-specific immune system. The same Kitimat and Quinsam coho salmon strains used in Chapter Three were subjected to lethal and non-lethal challenges with the marine pathogen V. anguillarum. Mortality, serum lysozyme activity, phagocyte respiratory burst activity (superoxide and H 2 O 2 production), and differential leucocyte number were examined pre- and post-challenge. M A T E R I A L S A N D M E T H O D S Fish. The origin and rearing history of the Kitimat and Quinsam coho strains used for this study are outlined in the General Materials and Methods section. Briefly, the Kitimat and Quinsam strains were comprised of 9 and 7 full-sib families, respectively. Both strains were reared exclusively in freshwater, and had been together in communal tanks for approximately one year prior to this study. On 15-18 July 1993, the fish were graded and approximately 400 fish from each strain (25 g mean weight ± 1 SD) were transported to the disease facility at UBC and randomly distributed into 16 100-L tanks. One row of 8 tanks contained 22 fish per tank from each strain, and the other row of 8 tanks contained 27 fish per tank from each strain. The fish were acclimated for approximately 2 weeks prior to the challenge. During the period of this experiment, water temperature ranged from 11.2-11.5 °C, and dissolved oxygen 9-10 mg/L. Fish were hand fed a commercial dry diet to satiation daily. Experimental Design and Sampling. The fish were divided into 4 treatment groups: group 1 - sham challenge (control), group 2 - low dose challenge, group 3 - moderate dose 61 challenge, and group 4 - high dose challenge. Each of the groups consisted of 4 tanks: 2 top row tanks (22 fish per strain), and 2 bottom row tanks (27 fish per strain). The fish in the tanks in the bottom rows were used to sample the non-specific immune system, while the top row fish were left undisturbed for the collection of mortality data. The groups (1-4) were challenged on different (subsequent) days, with each tank challenged separately. Samples (n=6 per strain) were randomly collected from the bottom row tanks of each group immediately prior to the challenge (Od), and then 2d, 7d, and 18d post-challenge (pc). The sampled fish were euthenized in a lethal dose of MS222 (tricaine methanesulphonate), and weights and strain (fin clip) recorded. The fish were quickly bled from the caudal vessel using a heparinized syringe. Aliquots of the collected blood were taken to measure various hematological parameters (see below). The remaining blood was transferred to a sterile tube, centrifuged and the resulting plasma stored at -70 °C for later lysozyme activity measurements. Immediately after each fish was bled, the anterior kidney was aseptically removed and placed in sterile, chilled phenol-free Hanks balanced salt solution (HBSS) for the phagocyte function assays. Disease Challenge. The marine fish pathogen V. anguillarum (Vang) was used for the disease challenges. A detailed description on the pathogen and the immersion challenge protocol is provided in the General Materials and Methods section. The treatment groups were challenged with the following doses of Vang (actual concentrations): group 1 (control), no Vang; group 2 (low dose) 5.35 x 10" cells/mL; group 3 (moderate dose), 7.90 x 105 cells/mL; group 4 (high dose), 1.26 x 109 cells/mL. Immediately following the challenge, the fish were placed back into their respective tanks. The fish in the bottom row tanks were sampled at d2, d7 and d 18 post-challenge, while the fish in 62 the top row of tanks were left to collect mortalities. All dead fish collected were confirmed to have died of vibriosis (as described in the General Materials and Methods section). Hematology. The number of leucocytes (lymphocytes, neutrophils, thrombocytes, monocytes) was calculated from the leucocyte to erythrocyte ratio, obtained from the examination of the blood smears (see General Materials and Methods section for description of method used to determined differential leucocyte ratios). The total number of erythrocytes was determined from hemacytometer counts of whole blood. Plasma Lysozyme Activity. The lysoplate method described in the General Materials and Methods section was used to determine lysozyme activity. The plasma samples were assayed in triplicate, and activity was expressed as mean U/mL plasma. Phagocyte Respiratory Burst Activity. The anterior kidney was homogenized with repeated aspirations through a sterile 1-cc tuberculin syringe. The cell suspension was washed twice in phenol-free HBSS and adjusted to a concentration of 106 viable cells/mL using trypan blue exclusion. The suspension was then chilled and used for the measurement of unstimulated phagocyte respiratory burst activity. Production of superoxide (02~) was measured using the glass adherent nitroblue tetrazolium (NBT) assay, described in the General Materials and Methods section. Duplicate samples were examined for each sample, and the proportion of blue (i.e., active) was calculated. The production of hydrogen peroxide (H 2 0 2 ) in the suspension of anterior kidney phagocytes was measured using the peroxidase-dependent oxidation of phenol red by H 2 0 2 (Secombes 1990). A preliminary time course study was carried out to determine the optimal incubation time (45 min). The cell suspension was washed twice in HBSS (1000 rpm at 5 °C, for 20 min), then resuspended in HBSS for a final concentration of approximately 106 cells/mL. One mL of the suspension was placed in a sterile tube along with one mL of the phenol red solution 63 (PRS; 0.02% phenol red, 0.01% horseradish peroxidase in HBSS). The suspension was incubated at room temperature with gentle rotation for 45 min. The reaction was stopped with the addition of 100 pL IN NaOH. Optical density was then measured at 620 nm using a PRS and NaOH blank. A standard curve of absorbance readings and concentration of H 2 O 2 was prepared. Regression analysis of the standard curve allowed the calculation of nmoles H 2 0 2 from the absorbance readings. Data Analyses. The mortality data were arcsin square root transformed and analyzed for strain differences using Chi square analysis. The following data transformations were performed before the statistical analyses were carried out: the proportion data (superoxide production) were arcsin square root transformed, the lysozyme data were log transformed, and the blood cell counts were square root transformed (Sokal and Rohlf 1981). Strain differences for each variable were then determined using Student /-tests at each sample time. Two-way A N O V A ' s were performed to determine the effect of challenge dose on the selected variable. Pearson product moment correlations were used to determine relationship between serum lysozyme and anterior kidney phagocyte superoxide production. For the correlation testing, each strain was examined separately and data from all sample times was combined. Statistical significance for all tests was determined where p<0.05. R E S U L T S There were significant strain differences in percent cumulative mortality in the group receiving the highest challenge dose - group 4 (Figure 4-1). No mortality occurred in the sham, low or moderate dose challenge groups. The Quinsam strain was the least disease resistant with 78% mortality while the Kitimat strain had 63% mortality. Mortalities started on d3 and continued up to and including dl 1 post-challenge. Live fish were sampled at each time. The d2 64 sample included both resistant and heavily infected moribund fish. The d7 sample occurred when the majority of the mortalities had occurred, and likely contained fish that were actively fighting the infection. The d 18 sample was taken from a population of survivors, that were either able to resist the initial infection, or had become infected but recovered. The lysozyme activity data shows that after an increase at d2, the Kitimat strain tended to maintain a higher level of lysozyme activity, than the less resistant Quinsam (Figure 4-2). Lysozyme activity increased significantly from dO to d2 in the low and high dose Vang challenge groups and the sham challenge groups (Kitimat strain only). By d7 the lysozyme activity in the sham challenge group decreased to pre-challenge (dO) levels. At this time, in the low and moderate challenge groups, lysozyme activity decreased in the Quinsam coho, while in the Kitimat coho, lysozyme increased significantly (low dose) or remained elevated (moderate dose). In the high dose challenge group, lysozyme activity in both strains increased up to d7, then decreased by d l8 to levels still significantly greater than those at dO. At d l8 post-challenge, plasma lysozyme activity in Kitimat coho was significantly higher than in the Quinsam coho for all three Vang challenge dose groups. The Kitimat coho had significantly higher phagocyte superoxide production in the moderate group at d7 and d l8 , and in the high dose group at d2 and d7 post-challenge. It is noteworthy that despite an apparent positive dose response in phagocyte superoxide production, the d 18 levels in the three different challenge dose groups were approximately the same. The low and moderate dose groups showed gradual increases up to approx. 10% at d l 8, while the high dose group decreased to approx. 10% at d l8 after peaking at d7. The results from the high dose challenge group, indicate that when compared with the less resistant Quinsam strain of coho, the Kitimat coho responded to the infection with a more rapid and sustained increase in plasma lysozyme activity and kidney phagocyte superoxide production. 65 Positive significant correlations between serum lysozyme activity and anterior kidney phagocyte respiratory burst activity (superoxide production), were detected for both strains in the low (Quinsam: r2=0.41, p<0.01; Kitimat r2=0.34, p<0.05) and in the high dose challenge groups (Quinsam. r2=0.36, p<0.05; Kitimat. r2=0.52, p<0.001) (Figure 4-4). Significant correlations in the control and moderate dose groups were not detected for either strain. Hydrogen peroxide production by the anterior kidney phagocytes showed no clear pattern of dose response or strain differences (Table 4-1). There was great variation in the individual response. However significant strain differences were detected at d l 8 in the sham challenged group. There were no strain differences in the differential leucocytes numbers (Table 4-1), nor in rbc numbers and hematocrit values (data not presented). DISCUSSION The Kitimat strain demonstrated significantly greater disease resistance following the high dose vibriosis challenge. This result was expected due to previous work by Withler and Evelyn (1990), who found that when compared with Robertson Creek coho, the Kitimat coho had higher survival and time to death following challenge with R. salmoninarum, the causative agent of bacterial kidney disease. The superior innate disease resistance of the Kitimat strain of chinook salmon was demonstrated by Beacham and Evelyn (1992a). Beacham and Evelyn (1992a) compared the disease resistance of the Kitimat, Quinsam and Nitinat strains of chinook salmon, and found the Kitimat chinook had the lowest mortality rate, following separate challenges with three different pathogens: V. anguillarum, R. salmoninarum, and A. salmonicida. Significant strain differences in serum lysozyme activity and phagocyte superoxide production, indicate that these factors may be important mechanisms underlying the enhanced disease resistance associated with the Kitimat strain. The levels of both these non-specific immune factors were significantly higher in the Kitimat strain in the later stages of the disease challenges. The high survival of the Kitimat strain may also be related to a quick non-specific immune response, because the phagocyte superoxide production seen at d2 post-challenge, was significantly greater in the Kitimat strain when compared with the less resistant Quinsam strain of coho. It is therefore suggested that the basis for the differences in innate disease resistance between strains may be related to the ability of the fish to react to the initial infection. Increases in lysozyme activity have been found to occur when the immune system is stimulated, i.e., during an infection (Studnicka et al. 1986, Meyner et al. 1993). In addition, increases in fish lysozyme activity have been positively correlated with numbers of monocytes and neutrophils (Fletcher and White 1973), which synthesize and secrete lysozyme (Murray and Fletcher 1976). In this study, the number of leucocytes did increase following the challenges, but there was no significant correlation between their numbers and lysozyme activity. Although it is possible the monocytes and neutrophils synthesized and secreted more lysozyme, the more likely source of the elevated lysozyme was the macrophage. Macrophages are also capable of synthesizing and secreting large amount of lysozyme (Gordon et al. 1974). Macrophages become activated when stimulated by foreign particulates or soluble substances. Activated macrophages produce an array of anti-microbial substances such as superoxide, H 2 0 2 , and lysozyme (Chung and Secombes 1988, Secombes and Fletcher 1992). The suspension of phagocytes used for the superoxide and H 2 0 2 production assays, was assumed to consist largely of macrophages. Significant positive correlations between superoxide production and serum lysozyme in the low and high dose challenge groups suggest the challenge infection activated the macrophages in the anterior kidney. Correlations were not significant in the control group probably because of the 67 limited response of the immune system to the sham challenge. The fish were not exposed to the pathogen and therefore the immune system was not activated. The slight increases in lysozyme and superoxide production seen in the control sham challenge group at d2 post-challenge were also observed in the coho salmon similarly challenged in Chapter Three. The increase may be attributed to a stress response associated with the handling and confinement of the challenge procedure, because it also occurred in the sham challenged group. Temporary increases in serum lysozyme and phagocyte activity have been associated with mildly stressful events (Mock and Peters 1990, Pegg et al. 1995). The stress effects however, appear to be temporary because by d7 post-challenge, the sham group levels returned to baseline (dO). In contrast, in the Vang-challenged groups, the lysozyme and superoxide production levels appeared to reflect the effects of an infection (i.e., activities increased because a pathogen was present to stimulate the immune system). The lysozyme values in the moderate dose challenge group were not as high as those found in the low dose challenge group. The lack of significant correlations between lysozyme and superoxide production in this group were probably due the suppressed lysozyme response. The reason for the suppressed lysozyme response in this group is not known, as the superoxide production was increased after the challenge. The challenge dose appeared to have an effect on the magnitude of the observed elevation in serum lysozyme activity and phagocyte superoxide production. Significant dose effects were, however not detected by two-way A N O V A . H 2 0 2 production by anterior kidney phagocytes did not show any clear pattern of dose effects, and no strain differences were observed. The same phagocyte suspensions demonstrated that respiratory burst activity was initiated with superoxide production. It is possible that catalase 68 or some other H 2 0 2 destroying substance was present in the suspension, and was somehow able to interfere with the measurements of H 2 0 2 production. The results of this study suggest the observed enhanced survival and innate disease resistance in the Kitimat strain, may be related to the activity of the non-specific immune system. The resistant Kitimat coho tended to have higher plasma lysozyme activity in the later stages of the infection, and higher phagocyte superoxide production throughout the infection (especially evident in the high dose challenge group). The rapid and continued elevation in the activity of the non-specific immune system in the disease resistant Kitimat strain may be an important factor in destroying the pathogen and reducing the infection. These results suggest the basis for the difference in innate disease resistance between strains may be related to the ability of the fish to respond to the initial infection. Further work to investigate the process of macrophage activation and the production of anti-microbial substances may provide a clearer picture of the underlying mechanism of strain differences in innate disease resistance. Table 4.1 Examination of pre- and post-challenge values for H 2 0 2 production by anterior kidney phagocytes, and numbers of different leucocytes, in two strains of coho salmon (Oncorhynchus kisutch), challenged with different doses of Vibrio anguillarum. * represents significant differences between strains (p<0.05). Coho Strain Saline sham challenge dO d2 d7 dl8 Low dose challenge dO d2 d7 dl8 Moderate dose challenge dO d2 d7 dl8 High dose challenge dO d2 d7 dl8 H 2 0 2 production Kitimat (nmoles H 2 0 2 ) 3.16 2.96 3.19 3.12* ±0.03 ±0.07 ±0.01 ±0.02 3.18 3.20 3.21 3.13 ±0.01 ±0.01 ±0.01 ±0.08 2.86 3.20 2.78 3.00 ±0.08 ±0.02 ±0.10 ±0.09 3.20 3.17 3.07 3.04 ±0.01 ±0.02 ±0.02 ±0.05 Quinsam 3.13 3.05 3.18 3.18* ±0.06 ±0.06 ±0.01 ±0.02 3.18 3.20 3.17 3.21 ±0.01 ±0.01 ±0.04 ±0.01 2.89 3.20 2.74 3.17 ±0.04 ±0.01 ±0.08 ±0.01 3.22 3.18 3.05 3.06 ±0.01 ±0.02 ±0.04 ±0.06 Lymphocyte Nos. Kitimat (xl0 4/mm 3) 0.80 0.54 1.30 1.66 ±0.24 ±0.11 ±0.21 ±.24 0.90 0.84 1.60 1.58 ±0.35 ±0.17 ±0.35 ±0.34 0.11 1.36 1.36 1.70 ±0.11 ±0.24 ±0.26 ±0.26 1.22 1.21 1.96 1.95 ±0.15 ±0.20 ±0.36 ±0.40 Quinsam 0.60 0.65 0.95 2.10 ±0.14 ±0.26 ±0.20 ±0.41 0.86 1.14 1.46 2.70 ±0.30 ±0.11 ±0.20 ±0.46 0.40 0.97 1.35 2.75 ±0.10 ±0.20 ±0.44 ±0.47 0.60 1.29 2.14 2.06 ±0.22 ±0.17 ±0.37 ±0.39 Thrombocyte Nos. Kitimat (xl0 4/mm 3) 0.28 0.17 0.64 1.09 ±0.18 ±0.05 ±0.14 ±0.27 0.56 0.41 0.56 0.46 ±0.32 ±0.07 ±0.11 ±0.15 0.37 0.17 0.41 0.90 ±0.08 ±0.06 ±0.11 ±0.15 0.82 0.32 0.79 0.26 ±0.21 ±0.09 ±0.22 ±0.08 Quinsam 0.16 0.20 0.27 1.34 ±0.08 ±0.06 ±0.07 ±0.32 0.22 0.30 0.55 0.83 ±0.02 ±0.10 ±0.17 ±0.32 0.40 0.19 0.42 0.64 ±0.20 ±0.06 ±0.11 ±0.35 0.37 0.37 0.51 0.37 ±0.16 ±0.07 ±0.14 ±0.10 Neutrophil Nos. Kitimat (xl0 4/mm 3) 0.00 0.17 0.08 0.02 ±0.00 ±0.06 ±0.04 ±0.02 0.29 0.31 0.16 0.04 ±0.29 ±0.11 ±0.04 ±0.04 0.08 0.15 0.14 0.13 ±0.08 ±0.05 ±0.04 ±0.07 0.07 0.15 0.14 0.07 ±0.07 ±0.04 ±0.04 ±0.05 Quinsam 0.07 0.21 0.19 0.03 ±0.04 ±0.06 ±0.06 ±0.03 0.07 0.39 0.12 0.09 ±0.07 ±0.08 ±0.04 ±0.07 0.20 0.07 0.21 0.00 ±0.10 ±0.04 ±0.06 ±0.00 0.00 0.05 0.16 0.00 ±0.00 ±0.03 ±0.07 ±0.00 Monocyte Nos. Kitimat (xl0 4/mm 3) 0.00 0.02 0.03 0.03 ±0.00 ±0.02 ±0.03 ±0.03 0.00 0.01 0.00 0.00 ±0.00 ±0.01 ±0.00 ±0.00 0.00 0.00 0.02 0.00 ±0.00 ±0.00 ±0.02 ±0.00 0.00 0.02 0.00 0.00 ±0.00 ±0.02 ±0.00 ±0.00 Quinsam 0.00 0.04 0.06 0.02 ±0.00 ±0.03 ±0.03 ±0.02 0.00 0.04 0.02 0.00 ±0.00 ±0.03 ±0.02 ±0.00 0.00 0.02 0.01 0.00 ±0.00 ±0.02 ±0.01 ±0.00 0.00 0.07 0.00 0.00 ±0.00 ±0.04 ±0.00 ±0.00 70 Figure 4-1. Percent cumulative mortality in two strains of coho salmon (Oncorhynchus kisutch) following an immersion challenge with Vibrio anguillarum (high dose group challenged with 1.26 x 109 viable cells/mL). * represents a significant difference (p<0.05) between the two strains when the challenge experiment was terminated (d 14 post-challenge). 71 I f 4 8 12 16 Days post-challenge - e - Kitimat Quinsam i r 4 8 12 16 Days post-challenge 20 Figure 4-2. Coho salmon (Oncorhynchus kisulch) strain comparison of mean (± sem) plasma lysozyme activity following immersion challenges in different concentrations of Vibrio anguillarum (Vang); (a) sham challenge, 0 Vang; (b) low dose challenge, 5.35 x 103 cells/mL; (c) moderate dose challenge, 7.90 x 105 cells/mL; (d) high dose challenge, 1.26 x 109 cells/mL. * represents significant differences (p<0.05) between the two strains at that time. • represents a significant difference from the dO (pre-challenge) level (p<0.05). n=12 per strain. 72 T 0 1 F 4 8 12 16 Days post-challenge - e - Kitimat - B - - Quinsam 4 8 12 16 Days post-challenge 20 Figure 4-3. Coho salmon (Oncorhynchus kisutch) strain comparison of mean (± sem) percent anterior kidney phagocyte respiratory burst activity (measured as superoxide production using the nitroblue tetrazolium or NBT assay), following immersion challenges in different concentrations of Vibrio anguillarum (Vang); (a) sham challenge, 0 Vang; (b) low dose challenge, 5.35 x 103 cells/mL; (c) moderate dose challenge, 7.90 x 10s cells/mL; (d) high dose challenge, 1.26 x 109 cells/mL. * represents significant differences (p<0.05) between the two strains at that time. • represents a significant difference from the dO (pre-challenge) level (p<0.05). n=12 fish per strain. 73 0.0 0.2 0.4 0.6 0.8 0.0 0.2 0.4 0.6 0.8 Arcsin SQRT % resp. burst activity Arcsin SQRT % resp. burst activity • Kitimat o Quinsam Correlations between mean plasma lysozyme activity and mean percent phagocyte respiratory burst activity (measured as superoxide production using the nitroblue tetrazolium or NBT assay), in two strains of coho salmon (Oncorhynchus kisutch) prior to, and following immersion challenges in different doses of Vibrio anguillarum (Vang). Pearson product moment correlations were determined separately for each strain (n=48). (a) sham challenge (0 Vang), Kitimat: r2=0.06, p>0.05, Quinsam: r2="0.01, p>0.05; (b) low dose challenge (5.35 x 103 cells/mL), Kitimat: r2=0.34, p<0.05, Quinsam: r2=0.41, p<0.01; (c) moderate dose challenge (7.90 x 105 cells/mL), Kitimat: r2="0.18, p>0.05, Quinsam: r2="0.17, p>0.05; (d) high dose challenge (1.26 x 109 cells/mL), Kitimat: r2=0.52, p<0.001, Quinsam: r2=0.36, p<0.05. Regression lines: Kitimat solid line, Quinsam dotted line. 74 CHAPTER FIVE COMPARISON OF COHO SALMON (Oncorhynchus kisutch) STRAINS FOR DIFFERENCES IN RESISTANCE TO EXPERIMENTAL INFECTIONS WITH Vibrio anguiUarum INTRODUCTION Strain differences in innate disease resistance are the result of a complex interaction between physical, cellular, and humoral non-specific immune factors. The research presented, thus far, indicates that strains with an internal non-specific immune system that responds quickly to infections, and with a sustained high level of activity, will be better able to survive infections. This chapter explores innate disease resistance from the perspective of the pathogen. Different strains of fish were infected with a pathogen, then the uptake and spread of the pathogen was compared. The objective of the following work was to provide insight into how the non-specific immune system operates as a whole in vivo to prevent infections, and in doing so, relate strain differences in innate disease resistance to differences in the infection process (i.e., the attachment, uptake, and spread of a pathogen). The goal therefore was to provide further evidence to support the hypothesis, that the non-specific immune system plays a significant role in innate disease resistance. Experimental infections produced by injection and immersion, were first performed to determine the route of pathogen entry important for innate disease resistance as measured by strain differences in mortality. Comparisons of strain mortality between the two challenge methods were made to determine whether the external non-specific immune system played a role in innate disease resistance. Pathogens injected directly into the fish evade the external non-75 specific immune system, whereas with the immersion challenge, the pathogen has the difficult task of overcoming various external barriers (skin, scales), humoral factors (mucus lysozyme and complement), and cellular factors (phagocytes). The uptake and spread of a pathogen was examined following immersion challenges, in an attempt to establish a relationship between innate disease resistance and the ability to prevent and inhibit infections. The causative agent of vibriosis, V. anguil/anim, was used as the bacterial pathogen throughout this study. This marine pathogen was chosen, to ensure an immunologically naive response in the freshwater reared coho salmon. Five communally reared coho salmon strains, originating from different geographical locations throughout B.C. were selected for comparison. Two of these strains, Kitimat and Quinsam, were used in previous comparisons (Chapters Three and Four), and differences in innate disease resistance were predicted based on these results. M A T E R I A L S A N D M E T H O D S Fish. Gametes were collected in the fall of 1992, from the Kitimat, Capilano, Quinsam, Robertson, and Chehalis River hatcheries. Details of the matings, and rearing of the fish are provided in the General Materials and Methods section. Briefly, each strain was comprised of equal numbers of fish representing ten full-sib families. The strains were fin-clipped for identification and combined into communal tanks in April of 1993 (3 g mean weight). The fish were reared exclusively in dechlorinated freshwater, that ranged in temperature from 3-16 °C, and dissolved oxygen concentration from 4-10 mg/L. The fish were fed daily ad libitum, and mortalities monitored. Vibrio Challenge Experiment. In October 1994 approximately 160 fish (60 g mean weight ± 1 SD) from each strain were transported to the disease tank facility at UBC and randomly distributed into 16 100-L tanks. The fish were acclimated for 3 weeks, during which • / • - 76 time, immersion heaters were placed into tanks and the water temperature gradually increased to 12 °C. Fish were hand fed a commercial diet, once daily to satiation. The fish were assigned into four groups of four tanks, each containing 10 fish per strain. A virulent isolate of V. anguillarum (Vang) was used for all the challenges described in this chapter. The isolate used and the preparation of the bacterial suspensions followed the protocol outlined in the General Materials and Methods section. The fish in two groups were challenged with two different doses of Vang, using an intraperitoneal (ip) injection challenge method. Fish were lightly anaesthetized in MS222 (tricaine methanesulphonate), and injected with 0.1 mL of the appropriate dose of Vang, which was prepared in phosphate buffered saline (PBS, pH 7.4). The actual doses of Vang injected were 2.28 x 105 cells (low dose injection group) and 2.28 x 106 cells (high dose injection group). The third group was challenged with an immersion challenge (described in the General Materials and Methods section), with 2.23 x 101(1 Vang cells/mL (actual dose). The fourth group represented an unchallenged control. Mortalities were monitored for three weeks post-challenge, and the mean cumulative percent mortality per strain for each challenge was calculated. Each dead fish was confirmed to have died of vibriosis (as outlined in the General Materials and Methods section). V. anguillarum Infection Studies. Three different infection studies were performed to compare the Robertson, Quinsam, and Kitimat coho for differences in the uptake and subsequent distribution of Vang in post-challenge tissue samples. The first study examined the uptake of Vang at the gills and blood immediately following an immersion challenge. The second study repeated the. first study, and included a post-challenge histological examination of the gill and lateral line tissue for the presence of Vang. In the third study, an immersion challenge was performed followed by a time series of blood samples taken from individual fish, to measure changes in the number of Vang in the blood over a period of 48 h post-challenge. In addition, at 77 24 h and 48 h post-challenge fish were killed and the presence of Vang in other organs was examined. Infection Study I: A total of 8 fish from each strain (60 g mean weight) were immersion challenged with an actual dose of 1.3 x 109 Vang cells/mL. Eight different challenges were repeated on three fish at a time (1 fish per strain per challenge). Immediately after each challenge the fish were placed into a lethal dose of MS222 for 1 min, followed by fresh water for approximately 2-5 min. Working very quickly, weights and strain (fin clip) were recorded. Blood (approx. 0.5 mL) was removed from the caudal vessel with a heparinized syringe, and placed in a sterile glass tube and chilled on ice. The tail was then severed to further bleed the fish. The entire left and right gills were dissected out and placed in sterile, iced PBS. The gills were then washed 3 times in PBS, by holding the gill arch with forceps and shaking forcibly. The gills were blotted with tissue paper in between each washing. After the third wash, the filaments were cut away from the gill arch and placed in a prerweighed sterile culture tube. The gill tissue was weighed and a 1:3 (w/v) suspension in sterile chilled tryptic soy broth (TSB) supplemented with 1.5% NaCl was prepared. The gill suspension was homogenized on ice using a sterile teflon tipped homogenizer. The homogenized gill suspension was then aliquotted (3 x 50 pL) and spread onto triplicate tryptic soy agar (TSA) plates. After 24-48 h incubation the number of colonies was counted and the mean number of Vang cells/g gill homogenate was calculated. The blood was similarly spread onto TSA plates and the mean number of Vang cells/mL blood was calculated. Infection Study II: The above protocol was repeated with minor modifications. Briefly, a total of 10 fish per strain (60 g mean weight) were immersion challenged in 1.3 x 108 cells/mL (actual dose). For this trial, histological samples were removed from the skin and gill before any other 78 sampling. Sections from the skin on the left lateral line and the first 2 gill arches on the left side, were removed and placed into Alcian blue fixative (Powell et al. 1992). Blood samples and the remaining gill tissue was removed as described above. The gill was manually washed 3 times, and then washed twice by centrifugation (15 min, 1000 g, at 5 °C). After each wash the supernatant was removed and placed in a sterile culture tube and spread onto TSA plates. A 50% gill homogenate (w/v TSB) was prepared, serially diluted with PBS and drop inoculated onto TSA plates. The blood was also serially diluted and drop inoculated onto TSA plates. The mean number of Vang cells/g gill and Vang/mL blood was calculated for each coho strain. Infection Study III: A total of 8 fish per strain (337 g mean weight) were immersion challenged (8.4 x 107 cells/mL actual dose) as described above. Immediately following the challenge the fish were anaesthetized (MS222) and 0.2 mL blood removed from the caudal vessel using a heparinized syringe. The same fish were then placed back into their tanks (1 fish per strain per tank). The fish were bled again at 2 and 6 h post-challenge. At 24 h post-challenge, half the fish (n=4 per strain) were killed with a lethal dose of MS222 and bled with a syringe (for the blood sample) and by the removal of the caudal peduncle. The kidney, spleen and liver were then aseptically removed, placed in sterile pre-weighed culture tubes and chilled. At 48 h post-challenge the remaining fish were similarly killed and sampled. The blood samples were placed on ice and homogenized for a predetermined time (30 sec), to rupture all the blood cells, and produce a measurement of the total (extracellular and intracellular) number of viable Vang cells in the blood. The homogenized blood was then serially diluted and 25 uL drops spread onto TSA plates. The kidney, spleen and liver samples were homogenized, weighed, and a 50% (w/v TSB) homogenate prepared. The homogenate was serially diluted and drop inoculated onto TSA 79 plates. Blood smears were also prepared at each sample time using 5 uL fresh blood (not homogenized). The smears were heat fixed and stored for later examination. Examination of preserved tissue and blood samples. Gill and lateral line skin samples (taken in study II) were fixed in 2% Alcian blue, embedded, sectioned and mounted (Powell et al. 1992). The slides were deparafinized, stained with fluorescein labeled anti-Vang (Microtek), and examined by fluorescent microscopy. The blood smears (taken in study III) were similarly stained with the fluorescein labeled anti-Vang and examined under a fluorescent microscope. Each smear was examined using a standardized protocol (i.e., 100 fields per smear using the same scanning pattern). Data analyses. Strain differences in the percent cumulative mortality, as well as the number of viable Vang cells in the blood and tissues were examined using one way analysis of variance tests. Subsequent comparisons among the means were made for significant differences using Student-Newman-Keul multiple comparison test or Dunn's comparison test. Student t-tests were used to compare the number of viable Vang cells at 24 h with the number at 48 h post-challenge, in the kidney, spleen, and liver, The percent cumulative mortality data was arcsin square root transformed. Significant differences were determined for the above tests where p<0.05. R E S U L T S Significant strain differences (p<0.05) in total percent cumulative mortality were detected when the fish were challenged by immersion. However no strain differences were detected when the strains were challenged by injection (low injection dose ) (Figure 5-1). The percent ;• ' 80 cumulative mortality for each strain following the low dose injection challenge was; Quinsam -78.9%, Robertson - 83.3%, Chehalis - 82.5%, Capilano - 75.0%, and Kitimat - 79.5%. There were no significant strain differences in the high dose injection challenge group because all five strains experienced 100% mortality. The results from the immersion challenge indicate that the Quinsam strain of coho were the most susceptible to vibriosis with 66.7% cumulative mortality, followed by the Chehalis (59.5%), Capilano (57.5%), Kitimat (46.3%), and Robertson (31.0%) strains. Pairwise comparison tests showed that the Robertson strain had significantly lower mortality when compared to all the other strains. ; Immediately following the immersion challenges performed in studies I and II, the Robertson coho were found to have the least number of viable Vang cells present in the homogenized gill samples. The amount of variation was quite large, but significant strain differences (p<0.05) were detected in study II.. The Robertson strain had significantly fewer Vang cells isolated from the gill tissue than the Quinsam strain (Figure 5-2a). Strain differences were not detected in the number of viable Vang cells found in the blood samples (homogenized or not), taken immediately post-challenge (Figure 5-2b). In comparisons of the resistant Robertson with the susceptible Quinsam strains, there appeared to be a trend whereby the Robertson had a low numbers of cells in the gills, but a high numbers of cells in the blood, while the reverse trend was seen for the Quinsam coho. . Examination of the PBS used to wash the intact gills for viable Vang, revealed large numbers in the first wash, less in the second,, and a negligible number of Vang cells in the third and final wash. The supernatant collected from the washing of the gill homogenates was free of viable Vang cells in the final (third) wash. The coho strains compared in the time course trial in study III showed a similar pattern of change in the number of viable Vang cells isolated from the blood over time (Figure 5-3). There 81 were no strain differences due to the large variation in counts and small sample size (n=4 at 24 h, and 48 h). There was an initial decrease in the number of cells from the number isolated immediately post-challenge (0 h). However, by 24 h the numbers increased, and by 48 h they were greater than (Quinsam and Kitimat) or equal to (Robertson) the numbers detected at 0 h. There appeared to be a trend for the Robertson to have lower number of viable cells in the blood from 24 h onward. The number of viable Vang cells in the liver, kidney, and spleen increased from 24 h to 48 h (Figure 5-4), although again, due to small sample size and large variation in counts this difference was not significant (p>0.05). The increase (from 24 to 48 h) was most apparent in the liver tissue homogenates, although overall, the kidney and spleen had higher number of cells than the liver. There were no strain differences in the number of cells isolated from any of the tissues at either 24 h or 48 h post-challenge. Some strain differences were detected by fluorescent microscopy, in the prevalence of Vang in the blood smears prepared in study III (data not presented): Greater numbers of Vang were observed in the Robertson strain at 0, 2 and 6 h, while more were found in the Quinsam at 24 and 48 h. All strains showed a similar change in the location of Vang cells in the blood, throughout the course of the investigation. At 0 h (i.e., after the 20 min immersion challenge), the Vang cells were found almost exclusively associated with neutrophils. However, by 24 h Vang did not appear to be associated with any blood cell type and were seen extracellularly as either . single cells or grouped into small clumps of cells. It was not possible to quantify the number of Vang cells because of this clumping (with other Vang cells and blood cells). Vang cells were not detected in any of the histological sections of gill and lateral line skin taken at 0 h in study II. DISCUSSION . 82 The appearance of significant strain differences in innate resistance to vibriosis, following the immersion challenge, and not the injection challenge, suggests the presence of genetic variation in the external non-specific immune system. External barriers of skin, scales and mucus plus their associated cellular and humoral defense factors are bypassed when a pathogen is injected. In this study, the ip injection challenge dose was sufficiently high to overwhelm the immune system of all the coho strains compared. The infection therefore could not be controlled, and a disease state was produced which resulted in the death of all (high dose injection) or most (low dose injection) of the fish. It is possible that strain differences may have been observed if a lower dose had been injected, because strain differences in the response of the internal cellular and humoral non-specific immune system have been previously demonstrated in Chapters One, Three, and Four. The strain variation in mortality following the immersion challengcsuggests an important role of external factors in innate disease resistance. Results from this challenge support the data presented in Chapter Four, in which the Quinsam coho were found to be less resistant to Vang infections than the Kitimat coho. This study and the studies described in Chapters Three and Four, were performed on different year classes of coho salmon, yet the results were consistent (i.e., Kitimat coho more resistant than Quinsam coho), thus serving to provide further evidence of a genetic basis to innate disease resistance. Beacham and Evelyn (1992b) compared family mortality rates (within one strain) after challenging coho salmon with R. salmoninarum (the causative agent of bacterial kidney disease, BKD) by immersion and injection, and found no correlation between the two challenge methods. This implies that the amount of variation in experimentally induced disease resistance, within and between strains of fish may be at least partly dependent on the method (and dose) used to challenge the fish. The immersion challenge is a natural method of exposing fish to pathogens, 83 and as a result, natural selection may have had its greatest effects in preventing infections occurring this way. Accordingly, genetic differences in innate disease resistance may be the greatest and therefore more easily detected following a natural immersion challenge. In addition, differences in innate disease resistance would also be more readily detected in situations where the fish are assured to be immunologically naive to the pathogen. In this study, the coho salmon had never been exposed to either sea water or the marine challenge pathogen, Vang. It is therefore believed that the challenge mortality results accurately reflect strain differences in innate disease resistance. Withler and Evelyn (1990) compared the Robertson and Kitimat coho strains for resistance to BKD, and found the Kitimat was more resistant. These results contradict the findings presented in this research, which show the innate disease resistance of the Robertson coho to be greater than that of the Kitimat coho. These contradictory results may be related to the challenge method, because Withler and Evelyn (1990) used the injection challenge. In addition, different bacterial pathogens were used. R. salmoninarum, the causative organism of B K D , can be transmitted vertically (Evelyn et al. 1984) and horizontally (Balfry et al. 1996) in fresh and seawater. It is possible that some of the fish used in their study may have been exposed to R. salmoninarum or been infected at a low (undetectable) level and as such were not immunologically naive to the challenge pathogen. Therefore an accurate assessment of the role of the non-specific immune system, in innate disease resistance may not have been reported. The first step in preventing an infection, and therefore mortality, is to prevent the attachment and subsequent colonization of the host surface by a pathogen. The site of colonization is presumably also the site of entry for the pathogen. Chen and Hanna (1992) have demonstrated the in vitro attachment of Vang to the skin, gills, intestine, and buccal mucosa. Immersion challenges using the ayu (Plecoglossus altivelis) have suggested that the skin is the 84 first site colonized by Vang (Muroga and De La Cruz 1987). Experimentally induced skin contact of fish with Vang (using Vang containing paper), led to the colonization of the contacted area and subsequent mortality (Kanno et al. 1989). The gills have also been shown to be an important site of entry for Vang (Bowers and Alexander 1982, Tatner and Home 1983, Baudin Laurencin and Germon 1987). The relative importance of each site is unclear. Alexander et al. (1981) found the gills rather than the lateral line were the primary sites for antigen entry. Amend and Fender (1976) reported that the lateral line was the site of antigen entry into the fish, although recent work by Ototake et al. (1996) suggest that while the lateral line likely serves as a site of entry for antigens to the skin, the skin is the actual site for antigen entry into the fish. In this study, we were unable to detect any Vang cells following examination of the histological sections of the gill or the lateral line skin. This was probably due to a problem with the technique, rather than an actual absence of the pathogen from these tissues. Vang was assumed to be on the surface of the gills immediately post-challenge, because viable cells were present in the saline collected after the first series of intact gill washes (data not presented). The resistant Robertson coho had significantly less Vang isolated from the gill tissue immediately post-challenge. The isolated Vang was presumed to be located within the gill tissue, because the gills were thoroughly washed to remove any cells that were on the gill surface. The amount of blood within the gill tissue was assumed to be minimal, and equal between the three coho strains. The results indicate that when compared with the other coho strains, the Robertson coho were better able to prevent the pathogen from entering the gill tissue and/or the phagocytes were quicker to remove the invading pathogens from the gill tissue. It is not possible with the results presented here, to determine which explanation is the most accurate. There were no significant strain differences in the number of viable Vang cells isolated from the blood immediately post-challenge, which suggests that the gills were not the only site of entry. 85 Although there were no strain differences in the number" of viable Vang cells in the blood for the first 24 h post-challenge, histological examination of the blood smears indicated that the Robertson blood contained more Vang. The observation of strain differences in the number of Vang cells seen in the blood smears, and the absence of strain differences in the number of viable Vang cells cultured from the blood, may be due to differences in the techniques. The fluorescent stains used to detect Vang in the blood smears do not differentiate live from dead cells, while cultured Vang counts measure only live cells. In addition, the blood used for the viable counts was homogenized (in study III) to disrupt the blood cells and release any Vang that was located intracellularly within phagocytes. The total number of viable Vang cells was therefore measured in the viable counts. The blood smears provided a semi-quantitative (and therefore less accurate) measurement of the number of Vang cells. It was hoped that upon examination of the blood smears, a quantitative ratio of intracellular/extracellular Vang could be determined, but the clumping of Vang prevented any quantification of cell numbers. As the infection progressed, Vang numbers began to increase as detected by both viable cell counts, and from (microscopic) observations of an increase in the number of non-cell associated Vang in the blood smears. The increase from 6 h to 48 h as detected from viable counts, was largest for the less resistant Quinsam coho, and smallest for the most resistant Robertson (not significant due to small sample size and large variation). The number of viable cells in the liver, spleen, and kidney tissue homogenates were not significantly different between the strains. Indicating that the infection had become systemic to the same extent in all strains. I propose the following scenario of the pathogenesis of vibriosis, to explain the observed . strain differences innate disease resistance. Following the immersion challenge, Vang in the.gills of the Robertson coho was removed more quickly than the other strains. The mechanism of uptake into the gill tissue may be via branchial phagocytes which have been shown to be capable of particle uptake (Goldes et al. 1986). The rapid uptake and removal of Vang into circulation seen in the Robertson coho, may explain the elevated number of viable cells found in the blood, and the lower number of Vang in the gills immediately following the immersion challenge. In addition, the Robertson coho may also possess a superior external defense system, which would result in fewer Vang cells entering the gills. The Vang infection appears to have become equally systemic in all strains, as the numbers of cells within the kidney, liver and spleen were not different. However, the decline in the number of Vang in the blood of the Robertson coho, suggests the infection was not progressing as rapidly as in the other coho strains. Chapter Four showed serum lysozyme activity and anterior kidney.phagocyte superoxide production were greater in the resistant Kitimat coho when compared to the susceptible Quinsam coho. In chinook salmon strain comparisons (Chapter One) the serum bactericidal activity against Vang was greater in Kitimat than Quinsam. It can therefore be speculated that the Robertson coho may have a more active internal non-specific immune system, than either Kitimat or Quinsam. In summary, it appears that the Robertson coho may have a greater innate resistance to vibriosis, because of a superior external defense system, and a more active internal non-specific immune system. The combination of the above characteristics may have resulted in the Robertson strain being better equipped to protect itself against infections and thus have higher survival rates because less fish had become diseased. 87 100 Coho Strains: T Robertson • Kitimat • Quinsam —O— Chehalis — A — Capilano i i i r 6 9 12 15 Days post-challenge 18 21 Figure 5-1. Post-challenge cumulative percent mortality for five strains of coho salmon (Oncorhynchus kisutch) challenge with Vibrio anguillarum by (a) intraperitoneal injection (low dose), and (b) immersion. *Robertson had significantly lower mortality than other strains at termination of challenge (d21). 88 5000 Study I Study II Coho Strains: V//A Robertson Kitimat I I Quinsam Study II Study III Figure 5-2. Coho salmon (Oncorhynchus kisutch) strain comparison of viable Vibrio anguillarum (Vang) isolated from (a) homogenized gill and (b) untreated blood (study II) and homogenized blood -(study III), immediately following immersion challenges with Vang. Challenge doses were: study I 1.3 x 109 cells/mL, study II 1.3 x 10* cells/mL, study III 8.4 x 107 cells/mL. Different letters represent significant differences between strains (p<0.05) within each study. n=8 per strain for Studies I and III, n=10 per strain for Study II. . 89 0 10 20 30 40 50 Time post-challenge (h) Figure 5-3. Coho salmon (Oncorhynchus kisutch) strain comparison of viable Vibrio anguillarum (Vang) isolated from homogenized blood following an immersion challenge. No significant strain differences were detected (p>0.05). n=8 for the samples taken at 0, 2, and 6 h post-challenge. n=4 for the samples taken at 24 and 48 h post-challenge. 90 Sample Times: I I 24hpc Y77A 48 h P c Robertson Kitimat Quinsam Figure 5-4. Coho salmon (Oncorhynchus kisutch) strain comparison of viable Vibrio anguillarum (Vang) isolated from; (a) kidney, (b) spleen, and (c) liver homogenates, at 24 (t24) and 48 (t48) h following an immersion challenge with Vang. No significant (p>0.05) strain differences or within strain differences between 24 h and 48 h samples were detected (n=4). 91 CHAPTER SIX AN EXAMINATION OF THE BACTERICIDAL ACTIVITY OF FISH MUCUS AND SERUM IN DIFFERENT STRAINS OF COHO SALMON (Oncorhynchus kisutch) INTRODUCTION The results presented in Chapters Four and Five, have shown that coho salmon strains from B.C., possess differences in innate disease resistance. The external barriers protect fish from a continuous exposure to waterborne pathogens. Research on the genetic variation of this form of protection has, however, received limited attention. Previous work (Chapter Five) suggests the external defense factors may be an important source of protection, as significant strain differences in vibriosis mortality were observed following an immersion challenge, while no strain differences were seen when the fish were challenged by intraperitoneal injection. The mucus layer which covers the skin, gill, and gastrointestinal tract, acts as a physical barrier, and because many biologically active substances have been found in the mucus, it also functions as part of the non-specific immune system (Alexander and Ingram 1992). In this chapter, I describe strain differences in the mechanisms of mucus and serum bactericidal activity. Three strains of coho salmon with established differences in innate disease resistance (see Chapter Five) were compared. The role of the alternative complement pathway in the bactericidal activity was determined by hemolytic assays, and explored further by using chelating agents which selectively remove M g 2 ' and/or Ca 2 + . The requirement of M g 2 + for the alternative pathway and M g 2 ' plus Ca 2 for the classical complement pathway, were used to distinguish the effect of each pathway on the bactericidal activity of serum and mucus. The role of lysozyme in the serum and mucus bactericidal activity was determined, because of the bacteriolytic properties of lysozyme and the previously demonstrated strain differences in serum/plasma lysozyme activity (reviewed in General Introduction, and reported in Chapters One-Four). The source of the observed strain differences in innate disease resistance and mucus bactericidal activity was further investigated for the activity of defensins. Defensins are antimicrobial cationic peptides found in rats, rabbits, guinea pig and human leucocytes (Hancock et al. 1995). There has been no published reports of defensins or other bactericidal cationic peptides in salmonids. 92 M A T E R I A L S A N D M E T H O D S / Fish. Three different coho salmon strains were compared in this study. The Kitimat, Robertson, and Quinsam strains were produced from 1992 broodstock, and were each represented by equal number of fish from 10 full-sib families. The details of the matings, incubation, and rearing of these strains is outlined in the General Materials and Methods section. The strains were reared exclusively in freshwater, and had been together in communal tanks since August 1993, Mucus and serum collection. Fish were lightly anaesthetized (MS222), wiped free of excess water and placed into a plexiglass tube. The top half of.the plexiglass tube had been cut away,.and one end was enclosed and darkened. Each fish was cradled in the tube with its head area covered. To collect the mucus, an anaesthetized fish was placed into the clean dry tube, and held firmly with one hand as it recovered from the anaesthetic. As the fish thrashed around in the tube, copious quantities of mucus were produced and collected onto the sides of the tube. After approximately 2-4 min the fish was removed and killed with a blow to the head. The mucus was scraped off the body of the fish.and off the inside of the tube using a clean dry rubber spatula. The mucus was collected in a sterile test tube and placed on ice. The mucus was then diluted 1:4 v/v with chilled, sterile, deionized, distilled water (ddEbO). Aliquots of the suspension were stored at -70 °C for later analysis. Following the collection of the mucus, the fish were weighed and bled from the caudal vessel with a syringe. Serum was collected as described in the General Materials and Methods section, aliquoted and stored at -70 °C for later analysis. The different strains of coho were sampled at different times for the work presented in this chapter. Mucus bactericidal activity was initially examined from the Robertson, Kitimat and Quinsam coho (10 fish per strain). The mucus from these fish was sampled as described above in December 1994, when the fish were 81 g mean weight. The same strains (10 fish per strain) were sampled again one year later (December 1995) at a mean weight of 239 g. The mucus and serum samples collected at this time (December 1995) were aliquoted, and stored at -70 °C. These same samples were later used to determine strain differences in the cation requirement for bactericidal activity against V. anguiUarum, lysozyme activity, hemolytic activity, protein concentration, and bactericidal activity against cationic peptide-resistant and cationic peptide-sensitive Salmonella lyphimurium. 93 Protein and lysozyme assays. Protein concentrations of the mucus samples were measured in a microplate reader using the bicinchoninic acid procedure (Smith et al. 1985). Prior to the assays performed below, the protein concentrations in the mucus samples were adjusted to a standard protein concentration using ddH 2 0 as a diluent. The protein content of the serum samples was not adjusted. Lysozyme activity of the mucus and serum was measured using protocol described in the General Materials and Methods section. Mucus bactericidal activity against Vibrio anguillarum. The mucus samples collected in December 1994 (n=10 per strain) were thawed on ice, and diluted 1:4 with ddH 20. The mucus samples were divided and half of each sample was heat treated (Sakai 1981). All samples were then chilled until the bacterial suspension was added. The isolate of V. anguillarum (Vang) and the preparation of the bacterial suspension has been previously described in the General Materials and Methods section. A suspension of 5 x 103 viable cells/mL was prepared in sterile, chilled, phosphate buffered saline (PBS; pH 7.4). Equal volumes of mucus and the bacterial suspension were combined, mixed and incubated at 15 °C with gerttle rotation. Controls were included by replacing the mucus with tryptic soy broth (TSB). The mucus and control tubes were sampled after 1.5 and 7 h incubation. At each sample time, the aliquots removed from the tubes were placed on ice, serially diluted in peptone-saline (P-S; 0.1%/0.85% peptone/NaCI, respectively) and drop-inoculated (5 x 25 pL drops per sample) onto tryptic soy agar (TSA) supplemented with 1.5% NaCl. After 24-48 h incubation at room temperature, the Vang colonies were counted, and the mean number of viable cells/mL was determined. Bactericidal activity was determined as the percent survival of bacteria relative to the control. Chelation studies on the bactericidal activity of mucus and serum. The role of divalent cations on the bactericidal activity of mucus and serum was investigated (10 fish per strain sampled December 1995). The protein concentration of the mucus was determined, and all samples standardized to an equal protein concentration. The following chelation treatments were used: gelatin veronal buffer (GVB; veronal-buffered saline containing 0.1% gelatin) supplemented with 0.5 mM Ca 2 ' and 1.0 mM M g 2 ' ( G V B 2 ) , ethylenediamine tetra-acetic acid disodium salt (EDTA; 10 mM), and ethyleneglycol-bis(2-aminoethylene ether) tetra-acetic acid (EGTA; 10 mM) in G V B and 10 mM M g 2 + (EGTA-Mg-GVB) (see Yano 1992 for solution recipes). The bactericidal activity was also determined for heat-treated (45 °C for 30 min) serum and mucus . (Sakai 1981). 94 The bactericidal activity of the mucus and serum was determined in the same manner as described above. The effects of the various treatments on the bactericidal activity of mucus and serum against Vang, and just serum against avirulent A. salmonicida were determined. The preparation of the bacterial suspensions are outlined in the General Materials and Methods section. The samples were thawed on ice, and then 4-50u.L aliquots of each sample were placed into a sterile tube. Three of the tubes were placed on ice while the fourth was heat-treated (as described above), then cooled. Equal volumes (50 uL) of G V B 2 ' , EDTA, and EGTA-Mg-GVB (solutions were chilled), were added to the three .untreated samples, and G V B 2 + was added to the heat-treated (fourth) sample. The samples were mixed and incubated with gentle rotation, at 15 °C for 30 min. After incubation, 20 uL of the bacterial suspension (approx. 104 cells/mL saline; 0.9% NaCl in ddH 20) was added. The samples were mixed and incubated for 2 h. Controls containing 200 uL of the three chelating solutions and 40 uL of the bacterial suspension were combined and incubated as above. The initial concentration of the bacterial suspension in each of the control solutions was determined by drop inoculating onto TSA plates immediately at the start of the incubation. After the 2 h incubation, each sample was diluted in P-S and drop inoculated (in triplicate) onto TSA plates. The number of bacterial colonies were counted after 24-48 h growth. The number of viable Vang cells surviving each treatment was expressed as the mean percent survival of bacteria, relative to the control. Mucus and serum hemolytic activity. The hemolytic assay described by Yano (1992) was used to determine the activity of the alternative complement pathway (10 fish per strain sampled December 1995). Briefly, the activity of the alternative complement pathway (ACH 5 0 ) was measured as the hemolysis of rabbit red blood cells (RaRBC). Mucus samples were standardized for protein concentration and assayed without further dilution. The serum samples were diluted 1:40 in EGTA-Mg-GVB. The samples were incubated at 15 °C for 90 min with occasional shaking. The results were expressed as A C H 5 0 values. One unit of A C H 5 0 was defined as the amount of serum or mucus, capable of lysing 4 x 107 RaRBC at 15 °C in a gelatin veronal buffer containing 10 mM E G T A and 10 mM M g 2 + in a total volume of 0.7 mL. S. typhimurium bactericidal assays. Two strains of Salmonella typhimurhim with established differences in susceptibility to the bactericidal activity of cationic peptides were used (Fields et al. 1989). The cationic peptide-susceptible (Ss; strain C610, MS7953s) and cationic peptide-resistant (Sr; C587, 14028s parent of C610) were obtained from Dr. Hancock, Dept. 95 Microbiology, UBC. The bactericidal activity of the mucus from the Robertson and Quinsam coho strains was determined (10 fish per strain sampled December 1995). Suspensions of the S. typhinrurhim strains were grown overnight at room temperature on TSA plates. The growth was harvested into sterile chilled PBS, mixed and diluted. The concentration of the bacterial suspension was estimated from optical density measurements at 540 nm. The mucus samples were thawed on ice, corrected for protein content, and 100 pL placed into each of two sterile tubes. Aliquots (20 uL) of the Sr and the Ss suspensions (approximately 103 cells/mL) were added to each of the mucus tubes. Controls containing 100 pL TSB were prepared with the addition of each of the S. typhimurium suspensions. The mucus/bacterial suspensions were mixed and incubated at 20 °C with gentle rotation for 2 h. After the incubation, the samples were inoculated in duplicate onto TSA plates, and incubated for 24-48 h. The number of colonies were counted, and the results expressed as the mean percent survival relative to the control. Gel electrophoresis. Mucus samples were examined for cationic bactericidal activity. Mucus samples from Quinsam, Kitimat and Robertson (sampled in December 1995) were run in duplicate, along with hen egg white lysozyme (HEWL, Sigma) and a cecropin/melittin hybrid peptide (CEMA), an antimicrobial cationic peptide standard. The samples were run on an acid-urea gel followed by a gel overlay procedure designed to detect cationic bactericidal activity. Briefly, samples were electrophoresed on a 15% acid urea gel which was incubated in Mueller-Hinton broth containing 0.2 M sodium phosphate buffer pH 7.4 for 1 h. The gel was overlaid with 5 mL of the same broth solution containing 0.6% agar and approximately 105 cells/mL of defensin-susceptible S. typhimurium (Ss). This was followed by another 5 mL of agar. The preparation was incubated overnight at 37 °C and examined for zones of lysis corresponding to the migration site of the C E M A . A second set of mucus samples from Quinsam and Robertson were similarly run in triplicate (acid-urea gel with overlay with Ss) and a slightly different cationic peptide standard, cecropin/melittin hybrid peptide (CEME). Data analyses. The bactericidal activity data (against Vang, A. salmonicida and S. typhimurium) was arcsin square root transformed. The lysozyme activity data was log transformed. Strain differences and treatment differences within each strain, were then examined using one way analyses of variance tests. Comparisons between the groups were then performed using Student-Newman-Keul multiple comparison test. Significant differences were determined where p<0.05. 96 R E S U L T S Significant strain differences (p<0.05) in the bactericidal activity of untreated mucus against Vang were seen (Figure 6-la). The Robertson coho had the greatest bactericidal activity in the mucus, as significantly fewer Vang survived after 1.5 and 7 h incubation, when compared with the other strains. There were no differences in mucus bactericidal activity between the Quinsam and Kitimat strains of coho salmon. There was a tendency for Vang to have higher survival in the heat-treated, rather than the untreated serum. The Robertson coho appeared to have the greatest bactericidal activity in heat-treated mucus, but due to the high amount of variation, the difference was not significant (Figure 6-lb). The effects of the various chelation treatments on the bactericidal activity of mucus from Robertson, Kitimat and Quinsam fish are shown in Figure 6-2. All strains demonstrated a trend for the greatest bactericidal activity in the G V B 2 control, followed by the EGTA-Mg-GVB, EDTA-GVB and heat treatments. Analysis of variance tests comparing the different chelation treatments within the strains, showed significant differences only for the Robertson coho. The mucus bactericidal activity in Robertson coho was significantly greater in the control G V B 2 + treatment, with the chelation and heat treatments significantly decreasing the bactericidal activity. There were, however, no significant differences between these treatments (cation chelation and heat). The effects of the different chelation treatments on the bactericidal activity of the serum showed a similar pattern to that seen for the mucus, i.e., the highest bactericidal activity seen in the G V B 2 + control, followed by the EGTA-Mg-GVB, E D T A - G V B and heat treatments (Figure 6-3). In the serum, however, this trend revealed more significant treatment differences than were seen in the mucus. The chelation of divalent cations had a significant effect on the bactericidal activity for Robertson and Quinsam coho, and the heat treatment had a significant effect on all three strains. The bactericidal activity of serum against A. salmonicida (Figure 6.4) showed a different trend in bacterial survival when compared with Vang. The serum showed the same bactericidal activity with the G V B 2 + control, as it did with the two chelation treatments (EGTA-Mg-GVB and EDTA-GVB) . Bacterial survival was only affected by heat treatment of the serum. Heat treatment of Robertson serum had a significant effect on serum bactericidal activity. The 97 bactericidal activities of the Kitimat and Quinsam strains could not be tested for treatment differences because the data shown represented a single pooled sample. Strain differences within each treatment, for mucus and serum bactericidal activity against Vang, were not statistically significant. Strain differences in serum bactericidal activity against A . sa/monicida within each treatment could not be tested because a single pooled sample was used for Kitimat and Quinsam. It would appear from the results (Figure 6-4) that Robertson had the greatest bactericidal activity against A. salmomcida, as the bacterial survival was much lower for each treatment. There were significant strain differences in mucus hemolytic activity, protein content, and lysozyme activity (Figure 6-5). The Robertson coho had significantly higher mucus hemolytic activity, protein, and lysozyme activity than the Kitimat or Quinsam coho strains (p<0.05). There were no differences in these mucus measurements between the Kitimat and Quinsam strains. Strain differences in these same measurements were not observed when the serum was examined. Mucus bactericidal activity that may be attributed to antimicrobial cationic peptides can be seen in Figure 6-6. Both Robertson and Quinsam showed significant differences in the survival of the resistant and the sensitive strains of S. typhimurium. Differences between the two coho strains were observed in the survival of the sensitive bacteria, whereby the Quinsam showed the greatest bactericidal activity. Mucus bactericidal activity against the sensitive S. typhimurium was not detected in either of the gel overlay procedures (Figure 6-7). The negative results may be attributed to the difficulty encountered in trying to concentrate the mucus enough to enable the detection of bactericidal activity, while maintaining the consistency of mucus required to load the gel apparatus. Differences between the strains were however seen in the first gel preparation, where the Robertson had a band at approximately the same migration site as HEWL. The Kitimat and Quinsam did not show a similar band. DISCUSSION In previous strain comparisons (Chapter Five) the Robertson coho were shown to have the highest survival rate following an immersion challenge with Vang, and have fewer Vang cells in the gill tissues immediately following an immersion challenge. The results from the strain comparisons presented in this chapter, demonstrate that the external mucus from the Robertson coho had the greatest bactericidal activity against the same isolate of Vang that was used to 98 challenge the fish in Chapter Five. These findings suggest that the ability of the Robertson coho to resist Vang infections may be related in part to an effective external defense system. Non-specific immune factors are implicated in this protection because all the coho strains used in this study were freshwater reared and therefore assumed to be immunologically naive to the marine pathogen Vang. This assumption was supported by the negative results from serum agglutinating antibody tests (against Vang) performed on all strains. The non-specific immune factor(s) responsible for the mucus bactericidal activity may be attributed in part, to complement, because there was a decrease in bactericidal activity following heat treatment. The heat treatment used (45 °C for 30 min) has been shown to inactivate coho serum complement (Sakai 1981). Results from the chelations studies, demonstrate a decrease in the bactericidal activity of serum and to a lesser extent, mucus, following heat treatment. The heat-labile factor appears to have a greater contribution to the bactericidal activity of serum than mucus, because there was no difference between the bactericidal activity of heat treated and untreated mucus for the Kitimat and Quinsam coho. Despite the effects of the heat treatments on bactericidal activity, it is unclear from the chelation studies whether complement was the heat-labile factor responsible for the bactericidal activity. The lack of antibodies against Vang should preclude the activation of the classical complement pathway, and therefore one would assume the majority of the bactericidal activity in all coho strains would attributed to the alternative complement pathway. The hemolytic activity found in both serum and mucus indicate that the alternative pathway was indeed active. The observation that the Robertson coho have significantly higher mucus hemolytic activity, is consistent with the mucus bactericidal results presented in Figure 6-1, and the mortality results presented in Chapter Five. The chelation studies however, do not support the above scenario. If the alternative rather than the classical pathway was the major source of bactericidal activity then there should be greater bactericidal activity in the presence of M g 2 + which is required for alternative pathway activation In general there appeared to be no difference whether M g 2 + was present or not, as there was little effect on bactericidal activity in its presence (EGTA-Mg-GVB treatment) or absence (EDTA-GVB treatment). In fact it would appear that Ca 2 + rather than M g 2 + had a greater effect on bactericidal activity, because without it (i.e., the EGTA-Mg-GVB treatment) the bactericidal activity was generally lower. Rjaed et al. (1992) reported similar unusual findings, suggesting that Ca 2 rather than M g 2 ' was more important for the activation of 99 the alternative complement pathway. Other researchers (Leid et al. 1985) have also found that Ca 2 ' can activate the alternative pathway. The serum bactericidal activity against A. salmonicida was not affected by the presence of either C a 2 ' or M g 2 ' . All three strains of fish showed a similar trend with only heat-treatment significantly decreasing bactericidal activity. Strain comparisons suggest that the Robertson coho had greater overall mucus bactericidal activity. Differences in the cation requirements seen between Vang and A. salmonicida are likely due to inherent differences in the composition of the bacterial cell walls. The additional A-layer surrounding A. salmonicida interferes with the lytic effects of complement and other bactericidal substances. In this study, significant strain differences in the bactericidal activity of mucus (Figure 6-1) were seen. Strain differences were also observed in the properties of this bactericidal activity, that may reveal strain differences in the underlying mechanisms of mucus bactericidal activity. For example, Ca 2 + had an apparent effect on the bactericidal activity in both serum and mucus from Robertson coho, while in the Quinsam coho Ca 2 ' had an effect only in serum, and had no effect on either serum or mucus from Kitimat coho. A clear role of alternative complement activity in innate disease resistance could not be established because of the strain differences in the role of Ca 2 ' in bactericidal activity. Hollebecq et al. (1995) compared spontaneous bactericidal and complement activity in rainbow trout, and found bactericidal activity of serum rather than alternative pathway activity to be a more reliable measurement of resistance to furunculosis. There are likely many bactericidal substances present in serum and mucus important for innate disease resistance, thus making it difficult to determine a clear role of any one particular factor. The data presented| in this chapter, shows that mucus lysozyme activity may have a significant effect on innate disease resistance. The disease resistant Robertson coho mucus had significantly higher protein concentration, which may be related to the amount of lysozyme present, because lysozyme activity was also greatest in Robertson coho when compared with the other strains. The Robertson coho had significantly higher mucus lysozyme activity as determined from the Iysoplate method, and also from the presence of a dominant band at the lysozyme migration site in the acid-urea gel. Lysozyme is known to play a significant role in bactericidal activity by lysing the bacterial cell wall. In addition to complement and lysozyme, coho mucus appears to contain bactericidal cationic peptides. Bactericidal assays demonstrate the presence of antimicrobial cationic peptides in mucus. Cationic peptides, known to kill a wide variety of bacteria, fungi, spirochaetes and LOO viruses (Hancock et al. 1995). The bactericidal activity of a group of cationic peptides, cecropins, have been shown to have lytic effects towards eight fish bacterial pathogens including Vang (Kelly et al. 1990). The inability to detect cationic peptide-associated bactericidal activity from the gel overlays may be related to preparation of the mucus: The bactericidal assays used crude mucus, and it is possible that during the lyophilization steps used to concentrate the mucus for the gels, the activity of the cationic peptide was reduced beyond the limits of detection. Strain differences in the cationic peptide-related bactericidal activity were observed. Unexpectedly the disease susceptible Quinsam coho had higher cationic peptide-related bactericidal activity than the disease resistant Robertson coho. In summary, this study demonstrates the existence of many bactericidal factors present in mucus and serum, that may contribute to the overall innate disease resistance of coho salmon. The presence of a bactericidal cationic peptide in mucus from coho salmon was reported for the first time. 101 M 100 -> 3 "2 a 'C a> o ^3 o s 80 -60 40 20 1.5 h incubation 7.0 h incubation Coho Strains: I Robertson Y///\ Quinsam b66® Kitimat 1.5 h incubation 7.0 h incubation Figure 6-1. Coho salmon (Oncorhynchus kisutch) strain comparison of bactericidal activity against Vibrio anguillarum (Vang) after 1.5 and 7 h incubation in; (a) untreated mucus, and (b) heat-treated mucus. Bactericidal activity measured as survival of Vang following 90 min incubation with mucus. Different letters refer to significant strain differences (p<0.05) at that sample time. n=10 per strain. 102 oo c > > 3 in > j s "o 0 s >> ]> o "3 -a 'o •C CD +-» o in 3 o Mucus Treatments: Control (GVB 2 + ) V77X EGTA-Mg-GVB E D T A - G V B Heat ( G V B 2 + ) Figure 6-2. Coho salmon {Oncorhynchus kisutch) strain comparison of the effects of heat treatment and divalent cations on mucus bactericidal activity against Vibrio anguillarum (Vang). Bactericidal activity was measured as survival of Vang following 90 min incubation with mucus, (a) Robertson, (b) Kitimat, and (c) Quinsam strains. Different letters refer to significant differences (p<0.05) between treatment groups. n= 10 per strain. 103 c > o > 'E 3 <D o n3 -O •c o <a B a 100 -80 60 -40 Serum Treatments: I I Control (GVB 2 + ) V77A EGTA-Mg-GVB E D T A - G V B Heat (GVB 2 + ) 100 -Figure 6-3. Coho salmon (Oncorhynchus kisutch) strain comparison of the effects of heat treatment and divalent cations on serum bactericidal activity against Vibrio anguillarum (Vang). Bactericidal activity was measured as survival of Vang following 90 min incubation with serum, (a) Robertson, (b) Kitimat, and (c) Quinsam strains. Different letters refer to significant differences (p<0.05) between treatment groups. n= 10 per strain. 104 Figure 6-4. Coho salmon (Oncorhynchus kisutch) strain comparison of the effects of heat treatment and divalent cations on serum bactericidal activity against Aeromonas salmonicida (Asal). Bactericidal activity measured as survival of Vang following 90 min incubation with serum, (a) Robertson, (b) Kitimat, and (c) Quinsam strains. Kitimat and Quinsam represent data from a single pooled sample (n=4), Robertson (n=10). Different letters refer to significant differences (p<0.05) between treatment groups. 105 Figure 6-5. Coho salmon (Oncorhynchus kisutch) strain comparison of mucus and serum properties: (a) protein concentration; (b) lysozyme activity; and (c) serum hemolytic activity measurements of the alternative complement pathway (ACH50). Different letters refer to significant strain differences (p<0.05). n= 10 per strain. 106 Figure 6-6. r Coho salmon {Oncorhynchus kisulch) strain comparison of mucus bactericidal activity against cationic peptide-resistant and cationic peptide-susceptible Salmonella typhimurium strains. Bactericidal activity was measured as the survival of bacteria following incubation with mucus. Different letters refer to significant differences between groups (p<0.05) 107 (b) R l R2 R3 Q l Q2 Q3 C E M E Figure 6-7. Gel overlay of coho salmon (Oncorhynchus kisutch) mucus samples performed to detect cationic peptide bactericidal activity against a strain of sensitive Salmonella typhimurium. In trial 1 (a) mucus samples from Robertson (R), Kitimat (K), and Quinsam (Q) strains of coho, were compared with a cecropin/melittin cationic hybrid antimicrobial peptide standard ( C E M A ) and hen egg white lysozyme ( L Y Z ) . In trial 2 (b) three mucus samples from Robertson and Quinsam were compared with a second cecropin/melittin cationic hybrid antimicrobial peptide standard ( C E M E ) . No bactericidal activity against the cationic peptide-sensitive bacterium was observed in either trial. 108 G E N E R A L DISCUSSION The hypothesis tested throughout this thesis was that the non-specific immune system plays a significant role in innate disease resistance in fish. The results presented here provide evidence to support this hypothesis. Significant strain differences in the activity of the internal cellular and humoral non-specific immune system, and the external humoral non-specific immune system were detected. These strain differences appeared to be related to innate disease resistance, because the most disease resistant strain tended to have the more active non-specific immune system. Strain differences in challenge survival rates and the activity of the non-specific immune system, suggests a genetic basis to innate disease resistance. As previously discussed, other researchers have also shown genetic variation in the non-specific immune system and disease resistance (reviewed by Chevassus and Dorson 1990, Fjalestad et al. 1993, Wiegertjes et al. 1996). The results presented in this thesis add new information about the role of the non-specific immune system in innate disease resistance, and possible mechanisms of resistance. The first report of strain differences in the activity of the non-specific immune system in Nile tilapia was shown in Chapter Two. Similarly, this thesis represents the first report of strain differences in the activity of the non-specific immune system and an apparent relationship with innate disease resistance for Pacific salmon (Oncorhynchus spp.) (see Chapters One and Three). New insight into possible mechanisms of innate disease resistance have been shown. For example, in Chapter Four there was a clear relationship between the enhanced survival of the Kitimat coho strain, and a sustained, elevated non-specific immune response following experimental infections. Further comparisons of coho salmon strains show that a highly effective 109 external non-specific immune system also appears to contribute to enhanced survival and innate disease resistance (see Chapters Five and Six). The humoral non-specific immune system showed significant strain differences and activities that reflected the ability of the fish to survive infection. In general, the disease resistant strains had the highest mucus lysozyme activity prior to a disease challenge (see Chapter Six), and the greatest increase in serum/plasma lysozyme activity following a disease challenge (see Chapters Three and Four). The bactericidal activity of serum was found to be the greatest in Kitimat chinook (Chapter One), which were shown by Beacham and Evelyn (1992a) to possess enhanced innate disease resistance. Strain differences in mucus bactericidal activity and hemolytic activity, like mucus lysozyme activity, were evident in unchallenged fish, with highest activity detected in the disease resistant strain (Chapter Six). Examination of the serum from the same coho salmon strains in Chapter Six, revealed no strain differences in hemolytic or lysozyme activity. Significant strain differences in the cellular non-specific immune system were detected in Nile tilapia (Chapter Two) and coho salmon (Chapter Three and Four). There also appeared to be a relationship between high post-challenge phagocyte respiratory burst activity and high survival in coho salmon (Chapter Four). The results from the above chapters reveal an interesting trend. Significant strain differences in the activity of the cellular and humoral non-specific immune system, appear to depend on where and when the comparisons are made. Unchallenged and therefore immunologically unstimulated fish, only show strain differences in the activity of the external non-specific immune system. Strain differences in the activity of the internal non-specific immune system generally appear only after the fish have been challenged (i.e., have been immunologically stimulated). Thus it appears that innate disease resistance has evolved in such a way that there is 110 genetic variation in the extent to which the external non-specific immune system is prepared to immediately prevent an infection and subsequent disease. In contrast, the internal non-specific immune system appears to show little genetic variation in the immediate preparedness of the non-specific immune system to fish infections. The non-specific immune system is considered to be an immediate form of protection, but a response is still required for phagocyte activity (e.g. cells migrate to the site of the pathogen invasion, lysozyme secretion from leucocytes, etc.). The response time required for effective protection is still much shorter than would be required for the specific immune system to respond in an individual naive to a particular pathogen. The existence of strain differences in the external, and absence in the internal, non-specific immune system of unchallenged fish, may be explained in terms of adaptive evolution. Fish are continually exposed to waterborne pathogens, therefore the external immune system has evolved mechanisms to prevent infections on a continuous basis. With this evolution of external non-specific immunity, genetic variation may have been established. In contrast, the internal non-specific immune system represents the second line of protection, coming into action after the external protective mechanisms have been evaded. The internal non-specific immune system may experience a more subtle pathogen invasion, as relatively few (compared with the number of pathogens the external non-specific immune system has contact with) pathogens begin to break through the external barriers, and start to stimulate the internal non-specific immune system to increase its protection through activation of its various components. The internal non-specific immune system may have evolved with a requirement for a such a stimulation to elicit and provide full protection against infections. In summary it can be postulated, that the adaptive evolution of the non-specific immune system was such that selective pressure was placed on the external non-specific immune system to provide a constant highly effective protection, while the internal non-specific immune I l l system experienced selective pressure not necessarily to be continually fully protective, but rather to respond quickly and with a high degree of activity once stimulated. The strain differences in non-specific immunity and disease resistance show the presence of genetic variation in these parameters. The initial coho strain comparisons performed on 1991 broodstock of Kitimat and Quinsam strains showed, as expected from the results of Withler and Evelyn (1990) and Beacham and Evelyn (1992a), the Kitimat strain had the highest survival and activity of the internal non-specific immune system (see Chapters Three and Four). The next set of coho strain comparisons was performed on 1992 broodstock of Kitimat, Quinsam, Robertson, Chehalis, and Capilano. The selection of the three new strains was based on geographical location. The comparisons of the activity of the non-specific immune system and innate disease resistance (Chapters Five and Six), showed that while the Kitimat was still more resistant than the Quinsam, the Robertson demonstrated the overall highest innate disease resistance (the Chehalis and Capilano were moderately resistant). The superior resistance of Robertson coho was unexpected based on the results of Withler and Evelyn (1990), who found the Kitimat was more resistant to B K D when compared with the Robertson coho. These contradictory results are discussed in Chapter Five, but can be explained in terms of the nature of the activity of the immune system, and the pathogen used. The ability to detect the same pattern of innate disease resistance in two different year classes of coho salmon (i.e., Kitimat more resistant than Quinsam in 1991 and 1992 broodstock comparisons, and the relatively consistent poor performance of the Quinsam coho), and using a small amount of genetic material (i.e., 7-10 full sib families), provides more evidence for a genetic basis for innate disease resistance. The source of this genetic variation may be due to selective pressures brought about by differences in geographical location of these strains originating rivers. Beacham et al. (1996) have shown significant minisatellite D N A variation between coho salmon strains from B.C., and in general found that strains that are 112 more distant geographically were also more genetically distinct. These authors examined all the coho strains used in this thesis, and interestingly, the Robertson, Kitimat and Quinsam, which were found to be significantly different in innate disease resistance and non-specific immunity, were also different from each other when compared for differences in genetic material (using analysis which was based on allele frequencies). This thesis involved the examination of strain differences in innate disease resistance. An estimation of the actual amount of additive genetic variation of the selected non-specific immune factors could not be determined because the experimental approach used did not permit such calculations. However, other researchers have calculated heritability estimates based on family analyses of such non-specific immune factors as hemolytic activity and lysozyme activity (reviewed by Wiegertjes et al. 1996). In general, these estimates indicate that due to the low heritability and correlation with disease resistance, selection based on these traits would not result in a significant increase in disease resistance. The ultimate goal of most research on genetic variation in disease resistance, is to identify genetic markers that show a significant correlation with enhanced survival. The difficulty in identifying such markers lies in the complex nature of disease resistance, and the current lack of understanding of the mechanisms involved. The main objective of this thesis was modest - to provide some basic information on the role of the non-specific immune system in innate disease resistance. Comparing variation between strains was the simplest approach, and although largely descriptive, I believe the results presented in this thesis have added significantly to our current knowledge on the mechanisms of innate disease resistance' in fish. The key findings in this thesis along with recommendations for.further studies, are summarized below: 113 Healthy, uninfected fish show more genetic variation in the activity of the external non-specific immune system, than the internal non-specific immune system. Genetic variation in the activity of the internal non-specific immune system tends to be more evident following an infection. It is recommended that for all studies examining the genetic variation of the non-specific immune system,.one should include both pre- and post-challenge samples. The challenge dose and method are important when examining genetic variation. This can be related to the above conclusion, as too high of a dose especially if administered by injection can prevent genetic differences from being expressed because the immune system is presumably completely overwhelmed. It is therefore recommended that the more natural immersion challenges be used for all studies on genetic variation in disease resistance. Low sublethal doses as well as the higher lethal doses can be used for immersion challenges while still enabling the detection of significant genetic variation. The choice of pathogen used in disease challenges and immune function assays are important for studying the genetic variation of innate disease resistance. It is absolutely essential to select a pathogen to which the fish are assured to be immunologically naive, because it is necessary to exclude the involvement of the specific immune system. Disease resistance is such a complex process that to accurately study the role of non-specific factors, the confounding effects of the specific immune system must be avoided. The choice of the group(s) of fish is important when performing comparisons of genetic variation in innate disease resistance, and activity of the non-specific immune system. To minimize the confounding effects of environment on the results, the different groups of fish 114 should be reared under identical and preferably communal conditions. The fish should be of similar size, as this thesis has shown significant effects of weight on lysozyme activity. In addition, the age of the fish is important as maternal effects on the activity of lysozyme appear to be significant at very early stages of development, and therefore can have a confounding effect on estimates of genetic variation. The results presented in this thesis underscore the need for further basic research to study in more detail, the mechanisms of the non-specific immune response responsible for genetic variation in innate disease resistance. There are two areas of research that I feel are particularly important. First, is the study of the genetic variation of cytokine expression by leucocytes, and the role in the signalling of the internal non-specific immune system to respond to infections. Results from this thesis indicate that the internal immune system shows little genetic variation until stimulated, and I believe that cytokines may play an important role here. The second area of future research would be to investigate the nature of the bactericidal activity of mucus. This thesis shows a relationships between mucus bactericidal activity and innate disease resistance in which lysozyme and complement appear to have a role. The existence of other non-specific factors such as cationic bactericidal peptides should be explored in more detail. 115 B I B L I O G R A P H Y Ainsworth, A.J. (1992) Fish granulocytes: Morphology, distribution, and function. Ann. Rev. Fish Dis. 2: 123-148. Alexander, J.B., Bowers, A. and Shamshoon, S.M. (1981) Hyperosmotic infiltration of bacteria into trout: Route of entry and the fate of infiltrated bacteria. Develop. Biol. Stand. 49: 441-445. Alexander, J.B^ and Ingram, G.A. (1992) Noncellular non-specific defence mechanisms of fish. Ann. Rev. Fish Dis. 2: 249-279. Altschmied, J., Muller, M . , Baniahmad, A. and Renkawitz, R. (1989) Cooperative interaction of chicken lysozyme enhancer sub-domains partially overlapping with a steroid receptor binding site. Nucleic Acid Res. 17: 4975-4991. Amend, D.F. and Fender, D C . (1976) Uptake of bovine serum albumin by rainbow trout from hyperosmotic infiltration: a model for vaccinating fish. Science. 192: 793-794. Amend, D.F. and Nelson, JR. (1977) Variation in the susceptibility of sockeye salmon Oncorhynchus nerka to infectious haemopoietic necrosis virus. J. Fish Biol. 11: 567-573. Anderson, D P . (1992) In vitro immunization of fish spleen sections and NBT, phagocytic, PFC and antibody assays for monitoring the immune response. In: Techniques in Fish. Immunology, Vol. 2, Stolen, J.S., Fletcher, T.C., Anderson, D.P., Kaattari, S.L. and Rowley, A.F. (Eds.), pp. 79-88, SOS Publications, Fair Haven, N.J. Bakke, T.A., Jansen, P.A. and Hansen, L P . (1990) Differences in the resistance of Atlantic salmon, Salmo salar L., stocks to the monogenean Gyrodatylus salaris Malmberg, 1957. J. Fish Biol. 37: 577-587. Balfry, S.K., Iwama, G.K. and Evelyn, T.P.T. (1994) Components of the non-specific immune system in coho salmon associated with strain differences in innate disease .resistance. Dev. Comp. Immunol. 18 (Suppl 1): p.S82. Balfry, S.K., Albright, L.J. and Evelyn, T.P.T. (1996) Horizontal transfer of Renibacterium salmoninarum among farmed salmonids via a fecal-oral route. Dis. Aquat. Org. 25: 3-69. Barton, B.A. and Iwama, G.K. (1991) Physiological changes in fish from stress in aquaculture with emphasis on the response and effects of corticosteroids. Ann. Rev. Fish Dis. 1: 3-26. Baudin Laurencin, F. and Germon, E. (1987) Experimental infection of rainbow trout, Salmo gairdneri R., by dipping in suspensions of Vibrio anguillarum. ways of bacterial penetrations; influence of temperature and salinity. Aquacul. 67: 203-205. 116 Beacham, T.D. and Evelyn, T.P.T. (1992a) Population and genetic variation in resistance of chinook salmon to vibriosis, furunculosis, and bacterial kidney disease. J. Aquat. Anim. Health 4: 153-167. Beacham, T.D. and Evelyn, T.P.T. (1992b) Genetic variation in disease resistance and growth to chinook, coho, and chum salmon with respect to vibriosis, furunculosis and bacterial kidney disease. Trans. Amer. Fish. Soc. 121: 456-485. Beacham, T.D., Miller, K M . and Withler, R E . (1996) Minisatellite D N A variation and stock identification of coho salmon. J. Fish Biol. 49: 411-429. Bell, G.R. (1977) Aspects of defense mechanisms in salmonids. In: Proceedings from the international symposium on diseases of cultured salmonids. pp. 56-70, April 4-6, 1977, Seattle, Washington. Boman, H.G. (1995) Peptide antibiotics and their role in innate immunity. Ann. Rev. Immunol. 13:61-92. Bowers, A. and Alexander, J.B. (1982) In vitro and in vivo passage of bacteria across restricted, areas and isolated tissues of trout, Salmo trutta L. and S. gairdneri Richardson. J. Fish Dis. 5: 145-151. Braun, R., Arnesen, J.A., Rinne, A. and Hjelmeland, K. (1990) Immunological localization of trypsin in mucus-secreting cell layers of Atlantic salmon, Salmo salar L. J. Fish Dis. 13: 233-238. Chen, D. and Hanna, P.J. (1992) Attachment of Vibrio pathogen to cells of rainbow trout, Oncorhynchus mykiss (Walbaum). J. Fish Dis. 15: 331-337. Chevassus, B. and Dorson, M . (1990) Genetics of resistance to disease in fishes. Aquacul. 85: 83-107. Ching, H.L. and Parker, L. (1989) Experimental exposure of trout and salmon from 12 British Columbian stocks to the myxozoan parasite Ceratomyxa shasta. J. Aquat. Anim. Health 1: 205-208. Chipman, D . M . and Sharon, N . (1969) Mechanism of lysozyme action. Science 165: 454-464. Chung, S. and Secombes, C.J. (1988) Analysis of events occurring within teleost macrophages during the respiratory burst. Dev. Comp. Immunol. 15: 201-208. Cipriano, R.C (1983) Resistance of salmonids to Aeromonas salmonicida. relation between agglutinins and neutralizing activities. Trans. Am. Fish. Soc. 112: 95-99. Cipriano, R.C. and Heartwell, C M . , III. (1986) Susceptibility of salmonids to furunculosis: differences between serum and mucus responses against Aeromonas salmonicida. Trans. Am. Fish. Soc. 115: 83-88. . 117 Clem, L.W., Faulmann, E„ Miller, N.W., Ellsaesser, C.F., Lobb, C.J. and Cuchens, M A . (1984) Temperature-mediated processes in teleost immunity: Differential effects of in vitro and in vivo temperatures on mitogenic responses of channel catfish lymphocytes. Dev. Comp. Immunol. 8: 313-322. Cole, A . M . , Weis, P. and Diamond, G. (1997) Isolation and characterization of pleurocidin, and antimicrobial peptide in the skin secretions of winter flounder. J. Biol. Chem. 272: 120088-12013. Dannevig, B.H. , Lauve, A., Press, C. McL. and Landsverk, T. (1994) Receptor-mediated endocytosis and phagocytosis by rainbow trout head kidney sinusoidal cells. Fish Shell. Immunol. 4: 3-18. Drickamer, K. (1988) Two distinct classes of carbohydrate-recognition domains in animal lectins. J. Biol. Chem. 263: 9557-9560. Ellis, A.E. (1977) The leucocytes offish: a review. J. Fish Biol. 11: 453-491. Ellis, A.E. (1988) Ontogeny of the immune system in teleost fish. In: Fish vaccination. Ellis, A.E. (Ed), pp. 20-31, Academic Press, London. Engstad, R.E. and Robertsen, B. (1994) Specificity of a B-glucan receptor on macrophages from Atlantic salmon (Salmo salar L.) Dev. Comp. Immunol. 18: 397-408. Evans, D.L. and Jaso-Friedmann, L. (1992) Nonspecific cytotoxic cells as effectors of immunity in fish. Ann. Rev. Fish. Dis. 2: 109-121. Evelyn, T.P.T., Ketcheson, J.E. and Prosperi-Porta, L. (1984) Further evidence for the presence of Renibacterium salmoninarum in salmonid eggs and for the failure of povidone-iodine to reduce the intra-ovum infection rate in water-hardened eggs. J. Fish Dis. 7: 173-182. Falconer, D.S. (1981) Introduction to quantitative genetics, 2nd Ed. Longman Inc., N .Y . , 340 pp. Fevolden, S.E., Refstie, T. and Roed, K . H . (1991) Selection for high and low Cortisol stress response in Atlantic salmon (Salmo salar) and rainbow trout (Oncorhynchus mykiss). Aquacul. 95: 53-65. Fevolden, S.E., Refstie, T. and Rjaed, K . H . (1992) Disease resistance in rainbow trout (Oncorhynchus mykiss) selected for stress response. Aquacul. 104: 19-29. Fevolden, S.E. and Raed, K . H . (1993) Cortisol and immune characteristics in rainbow trout (Oncorhynchus mykiss) selected for high or low tolerance to stress. J. Fish Biol. 43: 919-930. Fields, P.I., Groisman, E.A. and Heffron, F. (1989) A Salmonella locus the controls resistance to microbicidal proteins from phagocytic cells. Science 2: 1059-1062. Fienberg, S.E. (1970) The analysis of multidimensional contingency tables. Ecology 51: 419-433. 118 Finn, J.P. and Nielson, N.O. (1971) The inflammatory response of rainbow trout. J. Fish Biol. 3: 461-478. Fjalestad, K.T., Gjedrem, T. and Gjerde, B. (1993) Genetic improvement of disease resistance in fish: an overview. Aquacul. I l l : 65-74. Fleming, A. and Allison, V.D. (1922) Observations on a bacteriolytic substance ("Lysozyme") found in secretions and tissues. Brit. J. Exp. Pathol. 3: 252-260. Fletcher, T.C. and Grant, P.T. (1968) Glycoproteins in the external mucous secretions of the plaice, Pleuronectesplatessa, and other fishes. Proc. Biochem. Soc. 106: 12P. Fletcher, T.C. and White, A. (1973) Lysozyme activity in the plaice (Pleuronectes platessa L.). Experientia 29: 1283-1285. Fletcher, T.C. and Baldo, B.A. (1976) C-reactive protein-like precipitins in lumpsucker (Cyclopterus lurripus L.) gametes. Experientia 32: 1199-1200. Fletcher, T.C. and White, A. (1976) The lysozyme of the plaice Pleuronectes platessa L. Comp. Biochem. Physiol. 55B: 207-210. Fletcher, T . C , White, A. and Baldo, B.A. (1977) C-reactive protein-like precipitin and lysozyme in the lumpsucker Cyclopterus lumpus L. during the breeding season. Comp. Biochem. Physiol. 57B: 353-357. Gavora, J.S. and J.L. Spencer. (1983) Breeding for immune responsiveness and disease resistance. Anim. Blood Groups Biochem. Genet. 14: 159-180. Gjedrem, T. and Aulstad, D. (1974) Differences in resistance to vibrio disease of salmon parr. Aquacul. 3: 51-59. Gjedrem, T. and GJ0en, H .M. (1995) Genetic variation in susceptibility of Atlantic salmon, Salmo salar L., to furunculosis, B K D and cold water vibriosis. Aquacul. Res. 26: 129-134. Goldes, S.A., Ferguson, H.W., Daoust, P.-Y. and Moccia, R.D. (1986) Phagocytosis of the inert suspended clay kaolin by the gills of rainbow trout, Salmo gairdneri Richardson. J. Fish Dis. 9: 147-151. Gordon, S., Todd, J. and Cohn, Z.A. (1974) In vitro synthesis and secretion of lysozyme by mononuclear phagocytes. J. Exp. Med. 139: 1228-1248. Graham, S. and Secombes, C.J. (1988) The production of a macrophage-activating factor from rainbow trout Salmo gairdneri leucocytes. Immunol. 65: 293-297. Gravell, M . and Malsberger, R.G. (1965) A permanent cell line from the fathead minnow (Pimephales prome/as). Ann. N Y . Acad. Sci. 126: 555-565. 119 Grinde, B., Lie, 0 , Poppe, T. and Salte, R. (1988) Species and individual variation in lysozyme activity in fish of interest in aquaculture. Aquacul. 68: 299-304. Grinde, B. (1989) Lysozyme from rainbow trout, Salmo gairdneri Richardson, as an antibacterial agent against fish pathogens. J. Fish Dis. 12: 95-104. Groberg, W.J., J.S. Rohovec and J.L. Fryer. (1983) The effect of water temperature on infection and antibody formation induced by Vibrio anguillarum in juvenile coho salmon (Oncorhynchus kisutch). J. W. Maricult. Soc. 14:240-248. Hancock, R.E.W., Falla, T. and Brown, M . (1995) Cationic bactericidal peptides. Adv. Microb. Physiol. 37: 135-175. Hancock, R.E.W. (1997) Peptide antibiotics. Lancet. 349: 418-422. Harbell, S C . , Hodgins, H.O. and Schiewe, M.H . (1979) Studies on the pathogenesis of vibriosis in coho salmon Oncorhynchus kisutch (Walbaum). J. Fish Dis. 2. 391-404. Hardie, L.J. , Fletcher, T.C. and Secombes, C.J. (1994) Effect of temperature on macrophage activation and the production of macrophage activating factor by rainbow trout (Oncorhynchus mykiss) leucocytes. Dev. Comp. Immunol. 18: 57-66. Harrell, L.W., Etlinger, H . M . and Hodgins, H.O. (1976) Humoral factors important in resistance of salmonid fish to bacterial disease. II. Anti-Vibrio anguillarum activity in mucus and observations on complement. Aquacul. 7: 363-370. Harris, D.T., Kapur, R., Frye, C , Acevedo, A., Camenisch, T., Jaso-Friedmann, L. and Evans, D.L. (1992) A species conserved N K cell antigen receptor is a novel vimentin-like molecule. Dev. Comp. Immunol. 16: 395-403. Haynes, L. , Fuller, L. and McKinney, E.C. (1988) Fc receptor for shark IgM. Dev. Comp. Immunol. 12: 561-571. Heath, D.D., Devlin, R.H., Heath, J.W. and Iwama, G.K. (1994a) Genetic, environmental and interaction effects on the incidence of jacking in Oncorhynchus tshawtscha (chinook salmon). Heredity 72: 146-154. Heath, D.D., Bernier, N.J., Heath, J.W and Iwama, G.K. (1994b) Reply to comment on "Genetic, environmental, and interaction effects on growth and stress response in chinook salmon (Oncorhynchus tshawytscha) fry" by Silverstein. Can. J. Fish. Aqua. Sci. 51: 981-983. Hemmingsen, A.R., Holt, R.A., Ewing, R.D. and Mclntyre, J.D. (1986) Susceptibility of progeny from crosses among three stocks of coho salmon to infection by Ceratomyxa shasta. Trans. Am. Fish. Soc. 115: 492-495. Hine, P .M. (1992) The granulocytes offish. Fish Shell. Immunol. 2: 79-98. 120 Hjelmeland, K., Christie, M . and Raa, J. (1983) Skin mucus protease from rainbow trout, Salmo gairdneri Richardson, and its biological significance. J. Fish Biol. 23: 13-22. Hollebecq, M . - G , Faivre, B., Bourmaud, C. and Michel, C. (1995) Spontaneous bactericidal and complement activities in serum of rainbow trout (Oncorhynchus mykiss) genetically selected for resistance or susceptibility to furunculosis. Fish Shell. Immunol. 5: 407-426. Hutt, F.B. (1970) Genetics of resistance to infection. In: Resistance to infectious disease. Dunlop, R.H and Moon, H.W. (Eds), pp. 1-11, Modern Press, Saskatoon. Ibarra, A . M . , Gall, G.A.E. and Hedrick, R.P. (1991) Susceptibility of two strains of rainbow trout Oncorhynchus mykiss to experimentally induced infections with the myxosporean Cerafomyxa shasta. Dis. Aquat. Org. 10: 191-194. Ingram, G.A. (1980) Substances involved in the natural resistance of fish to infection - a review. J. Fish Biol. 16: 23-60. Ingram, G.A. (1987) Haemolytic activity in the serum of brown trout, Salmo trutta. J. Fish Biol. 31 (Suppl. A): 9-17. Iwama, G.K., Morgan, J.D. and Barton, B.A. (1995) Simple field methods for monitoring stress and general condition of fish. Aquacul. Res. 26: 273-282. Kanis, E., Refstie, T. and Gjedrem, T. (1976) A genetic analysis of egg, alevin and fry mortality in salmon (Salmo salar), sea trout (Salmo trutta), and rainbow trout (Salmo gairdneri). Aquacul. 8: 259-26. Kanno, T , Nakai, T. and Muroga, K. (1989) Mode of transmission of vibriosis among ayu Plecoglossus al five lis. J. Aquat. Anim. Health 1: 2-6. Kelly, D , Wolters, W.R. and Jaynes, J.M. (1990) Effect of lytic peptides on selected fish pathogen. J. Fish Dis. 13: 317-321. Kilpatrick, J.M. and Volanakis, J.E. (1991) Molecular genetics, structure, and function of C-reactive protein. Immunol. Res. 10: 43-53. Kinghorn, B P . (1983) A review of quantitative genetics in fish breeding. Aquacul. 3.1: 283-304. Koppenheffer, T.L. (1987) Serum complement systems of ectothermic vertebrates. Dev. Comp. Immunol. 11: 279-286. Lamas, J. and Ellis, A.E. (1994) Atlantic salmon (Salmo salar) neutrophil responses to Aeromonas salmonicida. Fish Shell. Immunol. 4: 201-219. Lamas, J., Santos, Y , Bruno, D.W., Toranzo, A.E. and Anadon, R. (1994) Non-specific cellular responses of rainbow trout to Vibrio anguillarum and its extracellular products (ECPs). J. Fish Biol. 45: 839-854. 121 Larkin, P A . (1972) The stock concept and management of Pacific salmon. In: The stock concept in Pacific salmon. ( R E . Simon and P.A. Larkin, Eds.), pp-11-15, H R . MacMillan Lecture in Fisheries, University of British Columbia, Vancouver, B.C. Law, S K A . and Reid, K . B . M (1988) Complement. IRL Press, Oxford, 72 pp. Leid, R.W., Coley, SC . , Blanchard, D P . and Perryman, L E . (1985) Equine alternative pathway activation by unsensitized rabbit red blood cells. Vet. Immunol. Immunopath. 9: 71-85. Lemaitre, C , Orange, N . , Saglio, P., Saint, N. , Gagnon, J. and Molle, G. (1996) Characterization and ion channel activities of novel antibacterial proteins from the skin mucosa of carp (Cyprimis carpio). Eur. J. Biochem. 240: 143-149. Le Morvan-Rocher, C. Troutaud, D. and Deschaux, P. (1995) Effects of temperature on carp leucocyte mitogen-induced proliferation and nonspecific proliferation and nonspecific, cytotoxic activity. Dev. Comp. Immunol. 19: 87-95. Lepow, I.H. (1980) Louis Pillemer, prosperdin, and scientific controversy. J. Immunol. 125: 471-478. Lie, 0., Syed, M . and Solbu, H. (1986) Improved agar plate assays of bovine lysozyme and haemolytic complement activity. Acta. Vet. Scan. 27: 23-32. Lie, 0, Evensen, 0, Sorensen, A. and Froysadal, E. (1989) Study of lysozyme activity in some fish species. Dis. Aquat. Org. 6: 1-5. Lund, T , Gjedrem, T., Bentsen, H.B., Eide, D M . , Larsen, H.J.S. and Riaed, K . H . (1995) Genetic variation in immune parameters and associations to survival in Atlantic salmon. J. Fish Biol. 46: 748-758. . Maack, T. and Sigulem, D. (1974) Renal handling of lysozyme. In: Lysozyme. Lysozyme conference, Arden House, 1972. Osserman, E.F., Canfield, R.E., and Beychok, S. (Eds.), pp. 321-333, Academic Press, N .Y. MacArthur, J.I. and Fletcher, T.C. (1985) Phagocytosis in fish. In: Fish immunology. Manning, M.J. and Tatner, M.F. (Eds.), pp. 29-46, Academic Press, London. Marc, A . M . , Quentel, C , Severe, A., Le Bail, P.W. and Boeuf, G. (1995) Changes in some endocrinological and non-specific immunological.parameters during seawater exposure in brown trout. J. Fish Biol. 46: 1065-1081. Marginos, B., Pazos, F., Santos, Y , Romalde, J.L. and Toranzo, A.E. (1995) Response of PasteureUa piscicida and Flexibacter maritimus to skin mucus of marine fish. Dis. Aquat. Org. 21: 103-108. Marsden, M.J. , Freeman, L.C. , Cox, D. and Secombes, C.J. (1996) Non-specific immune responses in families of Atlantic salmon, Salmo sa/ar, exhibiting differential resistance to furunculosis. Aquacul. 146: 1-16. 122 Maule, A G , Tripp, R.A., Kaattari, S.L. and Schreck, C.B. (1989) Stress alters immune function and disease resistance in chinook salmon {Oncorhynchus tshawytscha). J. Endocrinol. 120: 135-142. McAndrew, B.J. and Majumdar, K.C. (1983) Tilapia stock identification using electrophoretic markers. Aquacul. 30: 249-261. McAndrew, B.J., Roubal, F.R., Roberts, R.J., Bullock, A . M . and McEwen, I.M. (1988) The genetics and histology of red, blond and associated colour variants in Oreochromis w'/of/cm. Genetica 76: 127-137. McGeer, J., Baranyi, L. and Iwama, G.K. (1991) Physiological responses to challenge tests in six stocks of coho salmon (Oncorhynchus kistuch).Can. J. Fish. Aquat. Sci. 48: 1761-1771. Mclntyre, J.D. and Amend, D.F. (1978) Heritability of tolerance for infectious hematopoietic necrosis in sockeye salmon (Oncorhynchus nerka). Trans. Am. Fish. Soc. 107: 305-308. Mock, A. and Peters, G. (1990) Lysozyme activity in rainbow trout, Oncorhynchus mykiss (Walbaum), stressed by handling, transport and water pollution. J. Fish Biol. 37: 873-885. Moyner, K., Riaed, K . H , Sevatdal, S. and Heum, M . (1993) Changes in non-specific immune parameters in Atlantic salmon, Salmo salar L., induced by Aeromonas salmonicida infection. Fish. Shell. Immunol. 3: 253-265. Muona, M . and Soivio, A. (1992) Changes in plasma lysozyme and blood leucocyte levels of hatchery-reared Atlantic salmon (Salmo salar L.) and sea trout (Salmo trutta L.) during parr-smolt transformation. Aquacul. 106: 75-87. Murai, T., Kodama, H., Nakai, M . , Mikami, T. and Izawa, H. (1990) Isolation and characterization of rainbow trout C-reactive protein. Dev. Comp. Immunol. 14: 49-58. Muroga, K and De La Cruz, M.C. (1987) Fate and localization of Vibrio anguillarum in tissues of artificially infected ayu (Plecoglossus altivelis). Fish Pathol. 22: 99-103. Murray, C.K. and Fletcher, T.C. (1976) The immunohistochemical localization of lysozyme in plaice (Pleuronectesplatessa L.) tissues. J. Fish Biol. 9: 329-334. Oren, Z. and Shai, Y. (1996) A class of highly potent antibacterial peptides derived from paradaxin, a pore-forming peptide isolated from Moses sole fish Pardachirus marmoratus. Eur. J. Biochem. 237: 303-310. Osserman, E.F. and Lawlor, D.P. (1966) Serum and urinary lysozyme (Muramidase) in monocytic and monomyelocytic leukemia. J. Exp. Med. 124: 921-951. Ototake, M . , Iwama, G.K. and Nakanishi, T. (1996) The uptake of bovine serum albumin by the skin of bath-immunised rainbow trout Oncorhynchus mykiss. Fish Shellfish Immunol. 6: 321-333. 123 Ourth, D.D. and Bachinski, L . M . (1987) Bacterial sialic acid modulates activation of the alternative complement pathway of channel catfish (Ictalurus punctatus). Dev. Comp. Immunol. 11: 551-564. Pegg, J., Balfry, S.K., Gordon, L. , Roome, J.R. and Iwama, G.K. (1995) Stress, immune function and disease resistance in juvenile salmonids. Bull. Aquacul. Assoc. Can. 4: 28-35. Peleteiro, M.C. and Richards, R.H. (1990) Phagocytic cells in the epidermis of rainbow trout, Salmo gairdneri1 Richardson. J. Fish Dis. 13: 225-232. Pickering, A.D., Pottinger, T.G. and Christie, P. (1982) Recovery of the brown trout, Salmo trutta L. , from acute handling stress: a time course study. J. Fish Biol. 20: 229-244. Powell, M.D. , Speare, D.J. and Burka, J.F. (1992) Fixation of mucus on rainbow trout (Oncorhynchus mykiss Walbaum) gills for light and electron microscopy. J. Fish Biol. 41: 813-824. Pratschner, G.A. (1978) The relative resistance of six transferrin phenotypes of coho salmon (Oncorhynchus kisutch) to cytophagosis, furunculosis, and vibriosis. M.Sc. Thesis, University of Washington, Seattle, WA, 71 pp. Price, D.J. (1985) Genetics of susceptibility and resistance to disease in fishes. J. Fish Biol. 26: 509-519. Price, D.J. and Bone, L . M . (1985) Maternal effects and resistance to infection by Ichthyophthirius multifilis mXiphophorus maculatus. In: Fish Immunology. Manning, M.J. and Tatner, M.F. (Eds.), pp. 233-244, Academic Press, London. Putnam, F.W. (1975) Transferrin. In: The plasma proteins. Putnam, F.W. (Ed), Vol. 1, pp. 265-316, Academic Press, London. Ramos, F. and Smith, A C . (1978) The C-reactive protein (CRP) test for the detection of early disease in fishes. Aquacul. 14: 261-266. Rijkers, G.T. (1982) Kinetic of cellular and humoral immune reactions in fish. Dev. Comp. Immunol. Suppl 2: 93-100. Roberson, B.S. (1990) Bacterial agglutination. In: Techniques in fish immunology. Vol. 1, Stolen, J.S., Fletcher, T.D., Anderson, D P . , Roberson, B.S. and van Muiswinkel, W.B. (Eds), pp. 1-86, SOS Publications, Fair Haven, N.J. Roberts, R.J. (1989) The pathophysiology and systematic pathology of teleosts. In: Fish pathology. Roberts, R.J. (Ed.), pp. 55-91, Baillier Tindall, London. Roed, K . H . , Larsen, H.J , Linder, D., and Refstie, T. (1989) The genetic influence on natural immunity in rainbow trout. Anim. Gen. 20 (Suppl. 1): 54. 124 Raed, K .H . , Brun, E., Larsen, H.J. and Refstie, T. (1990) The genetic influence on serum haemolytic activity in rainbow trout. Aquacul. 85: 109-117. Roed, K . H . , Fjalestad, K.T., Larsen, H.J. and Midthjell, L. (1992) Genetic variation in haemolytic activity in Atlantic salmon (Salmo salar L.). J. Fish Biol. 40: 739-750. Roed, K . H . , Fjalestad, K. and Stromshein, A. (1993) Genetic variation in lysozyme activity and spontaneous haemolytic activity in Atlantic salmon (Salmo salar L.). Aquacul. 114: 19-31. Saggers, B.A. and Gould, M.L . (1989) The attachment of microorganisms to macrophages isolated from tilapia Oreochromis spilurus Gunther. J. Fish Biol. 35: 287-294. Sakai, D.K. (1981) Heat inactivation of complements and immune hemolysis reactions in rainbow trout, masu salmon, coho salmon, goldfish and tilapia. Bull. Jap. Soc. Sci. Fish. 47: 565-571. Sakai, D.K. (1992) Repertoire of complement in immunological defense mechanisms offish. Ann. Rev. Fish Dis.,2: 223-247. Salte, R., Gjeen, H .M. , Norberg, K. and Gjedrem, T. (1993) Plasma protein levels as potential markers for resistance to furunculosis. 16: 561-568. Salton, M.R.J. (1957) The properties of lysozyme and its action on microorganisms. Bact. Rev. 21: 82-99. Schreck, C.B. (1996) Immunomodulation: endogenous factors. In. Fish Physiology, Vol. 15, Iwama, G.K. and Nakanishi, T. (Eds), pp. 311-338, Academic Press, N . Y . Secombes, C.J. and Fletcher, T.C. (1992) The role of phagocytes in the protective mechanisms of fish. Ann. Rev. Fish Dis. 2: 53-71. Secombes, C.J. (1990) isolation of salmonid macrophages and analysis of their killing activity. In: Techniques in Fish Immunology, Vol. 1, Stolen, J.S., Fletcher, T . C , Anderson, D P . , Roberson, B.C. and van Muiswinkel, W.B. (Eds.), pp. 137-154, SOS Publications, Fair Haven, N.J. Secombes, C.J. (1996) The nonspecific immune system: Cellular defenses. In: Fish Physiology, Vol. 15, Iwama, G.K. and Nakanishi, T. (Eds), pp. 63-105, Academic Press, N . Y . Shephard, K . L . (1994) Functions for fish mucus. Rev. Fish Biol. Fisheries 4: 401-429. Smith, P.K., Krohn, R.I., Hermanson, G.T., Mallia, A.K. , Gartner, R.H., Provenzano, M.D. , Fujimoto, E.K., Goeke, N . M . , Olson, B.J. and Klenk, D C . (1985) Measurement of protein using bicinchoninic acid. Anal. Biochem. 150: 76-85. Smoker, W W . (1986) Variability of embryo development rate, fry growth, and disease susceptibility in hatchery stocks of chum salmon. Aquacul. 57: 219-226. 125 Sokal, R.R. and Rohlf, F.J. (1981) Biometry, 2nd Ed. W.H. Freeman & Co., San Francisco, U.S.A., 859 pp. Speert, D P . (1992) Macrophages in bacterial infection. In: The macrophage: the natural immune system. Lewis, C.E. (Ed.), pp. 215-263, Oxford University Press, Oxford. Standal, M . and Gjerde, B. (1987) Genetic variation in survival of Atlantic salmon during the sea-rearing period. Aquacul. 66: 197-207. Stewart, W.E. (1980) Interferon.nomenclature. Nature 286: 110. Stunicka, M . , Siwicki, A. and Ryka, B. (1986) Lysozyme level in carp (Cyprinus carpio L ) . Bamidgeh 38: 22-25. Suzumoto, B.K. , Schreck, C.B. and Mclntyre, J.D. (1977) Relative resistances of three transferrin genotypes of coho salmon (Oncorhynchus kisutch) and their hematological responses to bacterial kidney disease: J. Fish. Res. Bd. Can. 34: 1-8. Szalai, A.J., Bly, J.E. and Clem, L.W. (1994) Chelation affects the conformation, liability, and aggregation of channel catfish (Ictaluruspunctatus Rafinesque) phosphorylcholine-reactive (PRP) in response to inflammatory agents, low temperature shock and infection by the fungus Saprolegnia sp. Fish Shell. Immunol. 4: 323-336. Takahashi, Y , Itami, T. and Konegawa, K. (1986) Enzymatic properties of partially purified lysozyme from the skin of carp. Bull. Jap. Soc. Sci. Fish. 52: 1209-1214. Takemura, A. (1993) Changes in an immunoglobulin M (IgM)-like protein during larval stages in tilapia, Oreochromis mossambicus. Aquacul. 115: 233-241. Takemura, A. and Takano, K. (1995) Lysozyme in the ovary of tilapia (Oreochromis mossambicus): its purification and some biological properties. Fish Physiol. Biochem. 14: 415-521. Tatner, M.F. (1996) Natural changes in the immune system of fish. In: Fish Physiology, Vol. 15, Iwama, G.K. and Nakanishi, T. (Eds.), pp. 255-288, Academic Press, N . Y . Tatner, M.F. and Home, M.T. (1983) Factors influencing the uptake of 14C-labelled Vibrio anguillarum vaccine in direct immersion experiments with rainbow trout, Salmo gairdneri Richardson. J. Fish Biol. 22: 585-591. Tatner, M.F. and Manning, M.J. (1985) The ontogenetic development of the reticuloendothelial system in the rainbow trout, Salmo gairdneri Richardson. J. Fish Dis. 8: 35-41. Tillet, W.S. and Francis, T. (1930) Serological reactions in pneumonia with a non-protein somatic fraction of pneumococcus. J. Exp. Med. 52: 561-571. Utter, F .M. , Allendorf, F.W. and Hodgins, H.O. (1970) Genetic variability and relationships in Pacific salmon and related trout based on protein variations. Syst. Zool. 22: 257-270. 126 Wallace, H.A. (1961) Genetics of disease resistance: introductory remarks. Ann. N . Y . Acad. Sci. 91:597-601. Wedemeyer, G.A. and McLeay, D L . (1981) Methods for determining the tolerance of fishes to environmental stressors. In: Stress and Fish. Pickering, A.D. (Ed.), pp. 247-276, Academic Press, London. Wedemeyer, G.A., Barton, B.A. and McLeay, D.J. (1990) Stress and acclimation. In: Methods for Fish Biology. Schreck, C.B and Moyle, P.B. (Eds), pp. 451-489, American Fisheries Society, Bethesda, Maryland. Wehrhahn, C.F. and Powell, R. (1987) Electrophoretic variation, regional differences, and gene flow in the coho salmon (Oncorhynchus kisutch) of southern British Columbia. Can. J. Fish. Aquat. Sci. 44: 822-831. White, A., Fletcher, T.C. and Pepys, M.B. (1983) Serum concentrations of C-reactive protein and serum amyloid P component in plaice (Pleuronectes platessa L.) in relation to season and injected lipopolysaccharide. Comp. Biochem. Physiol. 74B: 453-458. Whyte, S.K., Chappell, L .H . and Secombes, C.J. (1989) Cytotoxic reaction of rainbow trout macrophges for larvae of the eye fluck Dip/ostomum spathaceum (Digenea). J. Fish Biol. 35: 333-345. Wiegertjes, G.F., Stet, R.J.M., Parmentier, H.K. and van Muiswinkel, W.B. (1996) Immunogenetics of disease resistance: a comparative approach. Dev. Comp. Immunol. 20: 365-382. Wilkinson, L. (1990) SYSTAT: The system for statistics. SYSTAT Inc., Evanston, IL , 822 pp. Winter, G.W., Schreck, C.B. and Mclntyre, J.D. (1980) Resistance of different stocks and transferrin genotypes of coho salmon, Oncorhynchus kisutch, and steelhead trout, Salmo gairdneri, to bacterial kidney disease and vibriosis. Fish. Bull. 77: 795-802. Withler, R.E. and T.P.T. Evelyn. (1990) Genetic variation in resistance to bacterial kidney disease within and between two strains of coho salmon from British Columbia. Trans. Am. Fish. Soc. 119: 1003-1009. Withler, R.E., Clarke, W.C., Riddell, B E . and Kreiberg, H. (1987) Genetic variation in freshwater survival and growth of chinook salmon (Oncorhynchus tshawytscha). Aquacul. 64: 85-96. Yano, T. (1992) Assays of hemolytic complement activity. In: Techniques in fish immunology, Vol. 2, Stolen, J.S., Fletcher, T . C , Anderson, D P . , Kaattari, S.L.and Rowley, A.F. (Eds.), pp. 131-141, SOS Publications, Fair Haven, N.J. Yano, T. (1996) The nonspecific immune system: Humoral defense. In: Fish Physiology, Vol. 15, Iwama, G.K. and Nakanishi, T. (Eds.), pp. 106-158, Academic Press, N . Y . 127 Yano, T., Ando, H. and Kakao, M . (1984) Optimum conditions for the assay of hemolytic complement titer of carp and seasonal variation of the titers. J. Fac. Agr. Kyushu Univ. 29:91-101. Yousif, A .N . , Albright, L.J. and Evelyn, T.P.T. (1991) Occurrence of lysozyme in the eggs of coho salmon Oncorhynchus kisutch. Dis. Aquat. Org. 10:45-49. Yousif, A .N . , Albright, L.J. and Evelyn, T.P.T. (1994) In vitro evidence for the antibacterial role of lysozyme in salmonid eggs. Dis. Aquat. Org. 19: 15-19. Zinn, J.L., Johnson, K.A. , Sanders, J:E. and Fryer, J.L. (1977) Susceptibility of salmonid species and hatchery strains of chinook salmon (Oncorhynchus tshawytscha) to infections by Ceratomyxa shasta. J. Fish. Res. Bd. Can. 34: 933-936. 128 APPENDIX A GENETIC ANALYSIS OF LYSOZYME ACTIVITY AND RESISTANCE TO VIBRIOSIS IN FARMED CHINOOK SALMON, Oncorhynchus tshawytscha (WALBAUM) INTRODUCTION Disease related mortalities in farmed salmonids in B.C., represent a significant economic loss to the aquaculture industry. Traditional methods to control and prevent diseases in farmed fish involve the use of chemicals, antibiotics and vaccines. These methods are not always effective and can be very costly. An alternative approach to disease control is the development of a breeding program designed to select fish with higher disease resistance. However, for such a program to be effective, sufficient genetic variation in disease resistance must be present to allow this characteristic to respond to selective breeding. Genetically based differences in disease resistance have been demonstrated for a variety of fishes (see reviews by Fjalestad et al. 1993, and Chevassus and Dorson 1990). Differences within and between stocks of Pacific salmon have been found for resistance to bacterial pathogens (Beacham and Evelyn 1992a, 1992b, McGeer et al. 1991, Withler and Evelyn 1990, Smoker 1986), viral pathogens (Mclntyre and Amend 1978, Amend and Nelson 1977), and parasites (Ching and Parker 1989, Hemmingsen et al. 1986, Zinn et al. 1977). All of these studies report genetic differences for mortality rates following experimentally induced disease infections. Gjedrem and Aulstad (1974) found genetic differences in disease resistance following a natural outbreak of vibriosis in Atlantic salmon, maintained in experimental family tanks. Variation in disease resistance has traditionally been measured using mortality/survival rates. Such traits are not precise measurements and can thus result in inaccurate estimates of the 129 actual genetic contribution to disease resistance (see Gavora and Spencer 1983). Plasma lysozyme activity has been found to be associated with disease resistance (Fevolden et al. 1991) and therefore has potential for use as a an indicator of disease resistance. In addition, plasma lysozyme has been shown to have significant genetic variation in fish (Grinde et al. 1988, Fevolden et al. 1991, Fevolden and R0ed 1993). A natural outbreak of vibriosis on a commercial salmon farm, provided an opportunity to examine the genetic contribution to disease resistance in farmed fish. Mortality rates were compared in two environments with replicated full and half-sib family groups of chinook salmon. Plasma lysozyme was also examined in the family groups to determine its genetic variation. MATERIALS AND METHODS The fish used for this study were domestic chinook salmon, of Robertson Creek stock, mated in 1989 to produce 2 groups of 12 full sib families. Details of the mating and rearing of these groups are given in Heath et al. (1994a). Briefly, eggs from each of 6 females were divided, whereby half the eggs were fertilized by six 2-year-old males, and half by six 3-year-old males. The 2- and 3-year-old males were unrelated; the two year classes were used to maximize the potential genetic variation among sires. The resulting 12 families were further halved into two treatment groups. The larval development of one group (El) was accelerated by incubation in heated water, while the other group (E2) was not accelerated (i.e., ambient temperature). These fish were part of a separate research project investigating the genetic basis for precocious maturation. At approximately 3 g mean weight (6 months post-fertilization), each fish was implanted with a coded wire nose tag to identify the family group. Fish from the two environmental groups were held in separate tanks. E l fish were transferred to seawater grow-out facilities on June 30, 1990 (8.4 g mean weight), and the E2 fish were transferred on July 19, 1990 130 (7.3 g mean weight). The two environmental groups were held in separate netpens at a commercial salmon farm. Throughout the seawater rearing period, mortalities were removed weekly and frozen for later family identification. A disease outbreak occurred at the farm later that summer (Aug. 19-22), and the E l fish experienced elevated mortalities (total mortality « 650 fish). Approximately two months later (Oct. 7-10) a second disease outbreak occurred, this time affecting the E2 fish (total mortality « 500 fish). Following each disease outbreak, the fish were treated for 8 d with antibiotic (0.167 g/kg/day Romet; Syndel Laboratories, Vancouver, B.C.), starting Aug. 23 for E l and Oct. 10 for E2 group. During these disease outbreaks, mortalities were collected daily and their nose tags recovered for family identification. The cause of the mortalities in both groups was determined to be vibriosis based on clinical signs, i.e., skin and fin hemorrhages, liquefied spleen, swollen and hemorrhagic vent. Presumptive identification of the pathogen, V. anguillarum was based on its isolation from the culture of infected tissue on tryptic soy agar (TSA) supplemented with 1.5% NaCl. On Dec. 9 and 10, 1990, approximately 2-4 months after the disease outbreak, 415 E l and 439 E2 fish were randomly sampled from their respective netpens (14 months post-fertilization). The fish were sampled in December to ensure that all effects of the disease outbreak were past: the fish were thus healthy and symptom-free. All fish were killed with an overdose of 2-phenoxyethanol (2.5 mL/L) and blood samples taken from the caudal vasculature using heparinized syringes. Blood was centrifuged, and the.plasma was collected and stored at -80 °C. Nose tags were removed for family identification. Plasma lysozyme activity was determined using 9 fish from each family from both environmental groups. The lysoplate method was used to determine lysozyme activity (described in the General Materials and Methods section). Lysozyme 131 activity was expressed as ug/mL, using hen egg white lysozyme as a reference standard. The plasma samples were assayed in triplicate, and the mean used for all statistical analyses. Data Analyses. Family mortality estimates were based on the identification of the nose tags recovered during the disease outbreaks. The number of surviving fish from each family was calculated as the initial number of fish tagged (500/family) less all mortalities that occurred prior to the disease outbreaks. Outbreak mortality was analyzed using log-linear models because such data consists of discrete counts, rather than continuous variables (Sokal and Rohlf 1981). A hierarchical series of log-linear models was used to test for the effects of rearing in two different environments (i.e., E l and E2), differences between individual male and female parents (sire and dam, respectively), and interactions on outbreak mortality (Wilkinson 1990). Specifically, a model was fitted to test for environment, dam, sire nested within dam, and interaction effects. Initially, all main and interaction terms were included in the model, this "saturated" model fits the observed data exactly, and thus the G statistic is zero (Sokal and Rohlf 1981, Wilkinson 1990). The statistical significance of specific effects of interest were then estimated by removing the term from the model and calculating the significance of the increase in the G statistic (Fienberg 1970, Sokal and Rohlf 1981). A three-way mixed model A N O V A was used to test for dam (random), sire nested within dam (random), environment (fixed), and interaction effects on plasma lysozyme activity. Plasma lysozyme activity measurements were natural logarithm transformed to reduce heteroscadacity (Sokal and Rohlf 1981). Mean family lysozyme activity (log transformed) was regressed against family outbreak mortality (arcsine square-root transformed; Sokal and Rohlf 1981). RESULTS 132 The total percent cumulative outbreak mortality was significantly different between the E1 and E2 groups, 8.1% and 5.7%, respectively (Figure A - l ) . Significant effects of sire nested within dam were detected for mortality (Table A - l ) . The cumulative mortalities for the 12 sired families in the two environmental groups is shown in Figure A - l . There were also significant dam effects on mortality (Table A - l ) ; cumulative mortalities were: daml - 8.0%; dam2 - 6.1%; dam3 - 6.6%; dam4 - 7.7%; dam5 - 4.8%; dam6 - 8.2%. No significant interaction effects between environmental group and any of the sire or dam effects were present. Mean family plasma lysozyme activities for the two environments are shown in Figure A-2. Sire, nested within dam, significantly affected plasma lysozyme activity (Table A - l ) . There was no significant dam effect; however, this is not surprising due to the limitations of the mating design (see Heath et al. 1994b). There was a highly significant effect of environmental group on plasma lysozyme activity (Table A - l ) . There were no significant interactions between environment and any of the sire or dam effects. Family cumulative mortality was found to be significantly positively correlated (p<0.005) with mean family plasma lysozyme (Figure A-3). Approximately 37% of the observed variation in family mortality was explained by variation in plasma lysozyme (r^ =0.37). DISCUSSION In this study, we measured mortality and plasma lysozyme activity in farmed chinook salmon following an outbreak of vibriosis, and found evidence for genetic contributions to disease resistance. While the lysozyme and mortality values differed between the two environmental groups and significant sire and dam effects on vibriosis-related mortality and sire effects on lysozyme activity were detected, no significant genotype-by-environment interactions were found. Although the significant sire effect on lysozyme includes both additive and non-additive genetic components, the magnitude of the sire effect suggests a sizable additive genetic contribution. If the additive genetic component of lysozyme is present, then we would expect significant dam effects on lysozyme; however, no such effect was found. The lack of such an effect in our experiment probably reflects the small number of females included in the breeding design, that is, the power of our design to detect dam effects was low (see Heath et al. 1994b). Our data do not provide conclusive evidence for an additive genetic component to lysozyme activity in chinook salmon; however, a significant additive genetic component to lysozyme activity has been previously reported for other salmonid species (Roed et al. 1989, Rjaed et al. 1993). Our analysis of mortality following the natural outbreak of vibriosis yielded both significant dam and sire effects. The presence of significant dam effects in our design indicates an additive genetic component to disease resistance, if we assume non-genetic maternal effects to be small (see Kinghorn 1983, Falconer 1981). Other studies have also shown a genetic basis to resistance to vibriosis in chinook salmon (Beacham and Evelyn 1992a), Atlantic salmon (Gjedrem and Aulstad 1974, Gjedrem and Gjoen 1995) and chum salmon, O. keta, (Smoker 1986). However, those studies involved experimentally induced infections, not naturally occurring outbreaks. Our results, combined with those reported in the literature, strongly suggest a genetic basis to disease resistance and, to a lesser extent, lysozyme activity in salmonids. Thus, the potential exists for improving the performance and survival of farmed salmon through selective breeding. The outlook for selective breeding is further improved by the lack of significant genotype-by-environment interactions. That is, despite the differences in mortality and lysozyme activity between the E1 and E2 groups, the ranking of the families within each group was not 134 significantly changed, hence selection attempts would not necessarily depend on individual farm conditions. Our data cannot explain the differences in mortality and lysozyme activity between the two environmental groups. One possible explanation of the difference in mortality may be that the E2 group experienced their disease outbreak later in the year when water temperature was lower, and thus the effect of the disease outbreak was less severe (Groberg et al. 1983). Variations in lysozyme have been attributed to changes in season, temperature, diet, and life stage (Fletcher and White 1976). The two environmental groups were sampled at the same time for lysozyme and were fed the same diet, thus perhaps the explanation may lie in the differences in body size of the fish in the two groups. However, although correlations between growth rate and disease resistance have been reported in the literature (Fevolden et al. 1992), neither our data, nor published sources, support a correlation between body size and lysozyme activity in salmonids. The positive phenotypic correlation between the mean family mortality and lysozyme activity indicates a possible genetic association between these two variables. An association between high lysozyme and high mortality has been previously reported in Atlantic salmon and rainbow trout, (Fevolden et al. 1991, Lund et al. 1995). The relationship between mortality and lysozyme activity may be mediated by a stress response. Published research has shown that high lysozyme may be correlated with an enhanced stress response, and that vibriosis.may be stress-induced (Wedemeyer and McLeay 1981, Fevolden et al. 1992, Fevolden and Roed 1993). However, the relationship between stress and lysozyme activity remains unclear since the nature of the lysozyme response to stress depends upon the severity of the stress (Mock and Peters 1990). Although the underlying mechanism responsible for the positive correlation between mortality and lysozyme activity is poorly understood, the relationship does suggest that lysozyme activity may be useful as a marker for disease resistance in cultured stocks of salmonids. In this study, evidence for genetic contributions to vibriosis-related mortality and plasma lysozyme activity was found for replicated full and half-sib families of farmed chinook salmon. The results were interpreted as very promising for the development of vibriosis-resistant strains of chinook salmon by selective breeding, and were the basis for an on-going commercial attempt to breed vibriosis-resistant chinook salmon stocks (J.W. Heath, unpublished data). The nature of the mechanism underlying the observed variation in disease resistance (i.e., enhanced immune function or stress resistance) remains to be determined. Such research is important since the development of disease resistant strains of fish would have a large impact on the viability of salmonid aquaculture. 136 Table A - l . Results of log-linear analysis of mortality data during a natural outbreak of vibriosis and A N O V A of plasma lysozyme activity 2-4 months after the outbreak in replicated full- and half-sib families of farmed chinook salmon [Oncorhynchus tshawytscha) reared in two different environments. The log-linear analysis and the A N O V A were designed to test for dam, sire (nested within dam), environmental group, and interaction effects. A l l interactions involving environmental and parental effects were not significant (p>0.05).. G statistics, F ratios, and significance levels are given (n sp>0.05, *p<° 05, **p < 0 .01, ***p<0 0 0 1 ) -Effect G-statistic / F-ratio Mortality (log-linear model) Dam G = 28.9** Sire (within Dam) G = 94.1 *** Environment G = 25 1*** Lysozyme Activity ( A N O V A ) Dam F 0.33 ns Sire (within Dam) F * * 5.03 Environment F 30.9 137 Figure A - l . Total percent mortality for 12 full-sib and 6 half-sib families of farmed chinook salmon {Oncorhynchus tshawytscha), following a natural outbreak of vibriosis. The families were reared in two different environments: (a) E l (accelerated larval growth); and (b) E2 (non-accelerated larval growth). The paired bars represent progeny from the two sires nested within each dam. See Table A - l for significant differences. 138 700 700 2 3 4 5 Family Group Dam Figure A-2. Plasma lysozyme activity (u.g/mL) for 12 full sib and 6 half-sib families of farmed chinook salmon (Oncorhynchus tshawytscha), 2-4 months after experiencing a natural outbreaks of vibriosis. The families were reared in two different environments, (a) E l (accelerated larval growth); and (b) E2 (non-accelerated larval growth). The paired bars represent progeny from the two sires nested within each dam. Error bars represent one standard error, see Table A - l for significant differences. 0.40 4.50 5.00 5.50 6.00 6.50 Log plasma lysozyme activity (pg/mL) Figure A-3. Correlation between percent cumulative family mortality during an outbreak of vibriosis, and mean family plasma lysozyme activity 2-4 months after the outbreak in full- and half-sib chinook salmon {Oncorhynchus tshawytscha) families reared in two different environments. The regression line shown was statistically significant; r2= 0.37, p<0.005, n=24. APPENDIX B 140 OBSERVATIONS ON LYSOZYME ACTIVITY IN COHO SALMON (Oncorhynchus kisutch) INTRODUCTION Lysozyme isolated from fish has been found to be effective as a bacteriolytic agent against both Gram-positive and Gram-negative fish pathogens (Grinde 1989, Yousif et al. 1994). Lysozyme is therefore an important factor in protecting fish against bacterial pathogens, due to its antibacterial properties and because it is located in areas that are in frequent contact with pathogens (i.e., kidney and skin mucus). The genetic variation of lysozyme has been established (Grinde et al. 1988, R0ed et al. 1989, Lund et al. 1995), as well as a negative correlation between plasma lysozyme activity and survival (Appendix A). It may therefore be possible to selectively breed fish with high lysozyme activity. This would be of interest for fish culturists who are currently seeking alternative approaches to disease control. The following chapter provides some information on the variation of lysozyme activity in coho salmon that may be useful in designing such a breeding program. The maternal contribution to lysozyme activity in eggs and the relationship between fish weight and lysozyme activity is also examined. MATERIALS AND METHODS 1991 coho salmon. These coho were collected in the fall of 1991 from the Kitimat and Quinsam rivers. Details of the matings and rearing are given in the General Materials and ' 141 Methods section. Briefly, eggs and milt from 9 (Kitimat) and 7 (Quinsam) mating pairs were collected and brought to Rosewall Creek Hatchery for fertilization (1:1) and incubation. The whole kidney from each female was removed at the time of the egg take, placed on ice and stored at -50 °C for later analysis of lysozyme activity. Unfertilized eggs (n=20) from each of the females were also taken and similarly stored for later analysis of lysozyme activity. The developing fish from each family were sampled at the eyed egg stage (approximately 150 accumulated thermal units, ATUs), and at the alevin or button-up stage (approx. 750 ATUs). These samples were stored at -50 °C for later analysis of lysozyme activity. Hatched alevins were transported 28 January 1992, to the Aquaculture Facility at the UBC. The families were placed into separate compartments in a rearing trough and fed by hand twice daily for approximately 1 month. At that time approximately 750 fry from each of the families representing the 2 strains were combined into separate rearing troughs. The fish were reared in dechlorinated water and fed a commercial diet using automatic feeders. On 13 May 1992, 70 fry (1 g mean weight) from each strain were randomly sampled and killed. The tails were severed and blood collected from the caudal vessels with a 10 pL capillary tube. The blood was centrifuged and serum collected. The serum samples and 12 whole intact fry from each strain were frozen at -80 °C for later analysis of lysozyme activity. On August 16, 1992 (3 g mean weight) 1000 fish per strain were fin clipped for strain identification and combined in communal tanks. Fish were removed from these tanks for disease challenge experiments in November 1992 (see Chapter Three) and August 1993 (see Chapter Four) at mean weights of 12.7 g and 33.3 g, respectively. For each experiment, plasma lysozyme activity was determined from control pre-challenge samples, and that data used again here. 142 1992 coho salmon. These coho were collected in the fall of 1992 from the Kitimat, Quinsam, Robertson, Chehalis and Capilano rivers. Details of the rriatings and rearing are given in chapter Three. Briefly, eggs and milt from 10 mating pairs were collected and brought to Rosewall Creek Hatchery for fertilization (1:1) and incubation. Serum, whole kidney, and unfertilized eggs (n=20) samples were taken from each female at the time of the egg take, placed on ice and stored at -80°C for later analysis of lysozyme activity. Alevins were transported on 5 February 1993 to the Aquaculture Facility at the UBC. Equal numbers of individual fry from each family were combined in strain tanks. Later the strains (500 fish per strain at 3 g mean weight) were fin clipped for identification and combined in communal tanks. Fish were fed daily, and mortality monitored. Fish were removed from these tanks for different experiments in November 1994 (81.4 g mean weight),.September 1995 (336.6 g mean weight), October 1995 (390.0 g mean weight) and December 1995 (239.0 g mean weight). At each sampling time, plasma or serum was collected and lysozyme activity determined. Lysozyme assay. Lysozyme activity was determined using the lysoplate method described in the General Materials and Methods section. Kidney tissue, whole fry, and alevin samples were homogenized and diluted 1.4 (w/v) using a phosphate buffer (PB; pH 6.0). The homogenates were centrifuged and the supernatant assayed for lysozyme activity. The lysozyme activity of unfertilized and eyed eggs was determined from the egg material that was removed from the individual eggs using a syringe. The egg material was then diluted (1:4 w/v with PB), centrifuged and the supernatant assayed for activity. Lysozyme activity in the plasma and serum samples was measured directly from undiluted, untreated samples. To measure the plasma activity from the 1 g fish, 39 fish were pooled to provide 3 pooled samples, each containing 13 fish. Each sample was assayed in triplicate and the mean lysozyme activity expressed as U/g for kidney, egg, alevin and whole fry samples, and as U/mL for the serum and plasma samples. 143 Statistical analyses. Prior to the statistical analyses, the lysozyme data.was log transformed. Measures of genetic variation within and between the strains was performed by analysis of variance tests, and where only two strains were compared (1991 coho) Student /-tests were used (Sokal and Rohlf 1981). Pearson product moment correlation tests were used to measure the strength of the association between the following variables: maternal kidney lysozyme activity and unfertilized egg lysozyme activity, maternal serum lysozyme activity and unfertilized egg activity, fish weight and serum/plasma lysozyme activity. Statistical significance for all tests was determined where p<0.05. R E S U L T S Significant positive correlations between the lysozyme activity in the maternal kidney and her unfertilized eggs were detected for the 1991 coho (r2=0.65, p<0.01, n=16), but not for the 1992 coho (r2=0.09, p>0.05, n=50) (Figure B- l ) . The 1992 coho showed a significant positive correlation between maternal serum lysozyme activity and unfertilized eggs (r2=0.39, p<0.01, n=50). Correlations between maternal kidney and serum lysozyme activity with unfertilized egg lysozyme activity were generally not significant (p>0.05) when each strain was examined separately (data not presented). Strain differences in lysozyme activity at the various early life stages (Figure B-2), were significant at the unfertilized egg and 1 g fry stages. The Quinsam coho had higher activity as unfertilized eggs, but as the development progressed past the egg stage (i.e., alevin and lg fry), the Kitimat had higher lysozyme activity. There was a highly significant positive correlation (r2=0.63, p<0.0001, n=l 14) between serum/plasma lysozyme activity and fish weight (Figure B-3) when the 1992 coho were examined over a wide range of weights (46-805 g). The correlation was not significant correlation (r2=" 144 0.06, p>0.05, n=l 16) for the 1991 coho, which were sampled over a much smaller range of weights (8.3-57.3 g). Genetic variation in lysozyme activity was examined in both year classes of coho. Variation in lysozyme activity within (i.e., family variation) and between strains at various life stages was statistically significant (Table B- l ) . DISCUSSION Maternal contributions to disease resistance have been established as significant non-additive, maternal effects on survival. The significance of such effects appears to be related to species and life history. For example, maternal effects on survival are highly significant for up to one year in the platy (Xiphophoms macahttus) (Price and Bone 1985), while in salmonids the effects are most important up to the eyed stage (Kanis et al. 1976). Withler et al. (1987) suggests that even in the earliest stages of development, maternal effects become significant only under adverse conditions (i.e., presence of pathogens, poor water quality), due to increased additive genetic variation. The maternal effects on lysozyme activity were not measured in this study, but significant positive correlations between maternal lysozyme activity (kidney and serum) and unfertilized egg lysozyme activity were detected. Strain differences in the lysozyme activity of unfertilized and eyed eggs generally reflected maternal lysozyme activity. However, after the eyed stage the pattern of strain differences in lysozyme activity changed. This change was likely due to a reduction in the maternal effects, associated with the adsorption of yolk material and the development of the lysozyme producing leucocytes in the developing fish. Takemura (1993) reported a similar decrease in maternally derived IgM-like proteins (MLPs) in larval tilapia (0. mossambicus) as the yolk was absorbed. MLPs were found in their lowest concentration when all the yolk was gone, and the larvae beginning to feed independently. After the first feeding, the 145 MLP started to rapidly increase as the larvae began to produce their own MLPs. The coho salmon examined here showed no increase in lysozyme activity until after the eyed stage, which supports the suggestion that like the M L P in tilapia, the coho had used up the maternally derived lysozyme and were independently producing lysozyme. Genetic variation within and between strains of coho were found to be significant at various life stages. However, in light of the preceding discussion, it is believed that strain differences in lysozyme activity are more accurately measured by comparing the genetic variation after the eyed stage, when maternal effects are less important. Measuring strain differences prior . to the eyed stage appears to reflect maternal lysozyme activity. The strain differences in maternal lysozyme activity (and therefore unfertilized eggs) reported in this study, were likely influenced by many unknown, uncontrollable factors (i.e., environment, size, nutrition, disease, and stress) so it was not possible to accurately demonstrate strain differences. Only those lysozyme measurements taken from fish after the eyed stage and reared under controlled conditions, could be used (with confidence) to measure genetic variability. A positive correlation between lysozyme activity and fish weight has been suggested previously by Fevolden et al. (1991) but is reported here for the first time. The wide range of weights obtained from sampling the 1992 coho, provided a large enough sample size to provide significant results. Reports of positive correlations between growth rate and survival (Standal and Gjerde 1987), mortality and plasma lysozyme activity (Appendix A), plasma growth hormone and lysozyme activity (Marc et al. 1995), have only hinted at a positive correlation between lysozyme activity and fish weight. It can therefore be recommended that measurements of lysozyme activity be taken from fish of similar weights. Lysozyme appears to have a significant role in the genetic variation of disease resistance in fish. It has an important role as an antibacterial agent in developing eggs (Yousif et al. 1994). 146 Throughout this thesis, the genetic variation of lysozyme was demonstrated as strain differences. The majority of differences appear to be related to post-challenge effects (i.e., infection related). These results suggest inducible lysozyme changes may have a genetic basis. Lysozyme synthesis and secretion has been associated with the activation of phagocytes (macrophage and neutrophils). While the importance of lysozyme as an antimicrobial agent is controversial, its association with disease resistance is not. It is possible that with further research, lysozyme may be found to be a very significant marker of disease resistance by demonstrating its role as a byproduct of a more microbiocidal mechanism (likely linked to phagocyte activation). This study shows the role of maternal, developmental and size effects on lysozyme activity in fish, and emphasizes the importance of controlling such variables when measuring the genetic variability of lysozyme activity in fish. 147 Table B-l. Examination of the genetic variation of mean lysozyme activity of various tissues, within and between strains of coho salmon {Oncorhynchus kisutch) collected in two different years. The 1991 coho strains were represented by 7 families (n=6 per family), 1992 coho represented by 10 families per strain (n=6 per family). Refer to text for background details on families and strains. Variation determined from analysis of variance tests. Statistical significance determined when p<0.05. Within strain variation: Year class Strain Life Stage p value 1991 Quinsam unfertilized eggs p<0.001 eyed eggs p<0.001 alevins p<0.001 Kitimat unfertilized eggs pO.OOl eyed eggs p<0.001 alevins p<0.001 1992 Quinsam unfertilized eggs p<0.001 Kitimat unfertilized eggs p<0.001 Robertson unfertilized eggs p<0.001 Capilano unfertilized eggs p<0.001 Chehalis unfertilized eggs p<0.001 Between strain variation: Year class Number of strains Life stage p value 1991 2 unfertilized eggs <0.05 2 fry (1 g) <0.05 .1992 •5 5 5 unfertilized eggs maternal kidney maternal serum <0.001 <0.05 <0.05 4.0 3.8 00 > '••2 «J s N O </) j>> >, <D 1 3.6 o _l 00 ° > o 0) • a o >, >, <u c *o 00 o •> o is <D S >^  N O c« >> £ 2 on 00 O -I (a) • • n • y_ • • 1.8 2.0 2.2 2.4 Log lysozyme activity of unfertilized eggs (U/g) Figure B - l . Correlations between maternal lysozyme activity and mean unfertilized egg lysozyme activity coho salmon {Oncorhynchus kisutch) collected in; (a) 1991: r2=0.65, p<0.01, n=16; (b) 1992: r2=0.09, p>0.05, n=50, and (c) 1992: r2=0.39, p<0.01, n=50. Figure B-2. Coho salmon (Oncorhynchus kisulch) strain comparison of lysozyme activity in different stages of early development. * Refers to significant strain differences (p<0.05). 150 s s N O o 1.7 1.6 H 1.5 1.4 H 1.3 1.2 H 1.1 (a) O O o o , o o o o o o# o o oo o oo o o o o 10 Q & 6b u o o o o o o 20 30 40 50 60 '•4—» o N o •s o h-1 150 300 450 Weight (g) 600 750 Figure B-3. Correlations between fish weight and serum lysozyme activity in individual coho salmon {Oncorhynchus kisutch) collected in; (a) 1991: r2="0.06, p>0.05, n=116; and (b) 1992: r2=0 63 p<0.001, n-114. Bibliotheque nationale du Canada Acquisitions and Acquisitions et Bibliographic Services services bibliographiques 395 Wellington Street 395, rue Wellington Ottawa ON K1A0N4 Ottawa ON K1A0N4 Canada Canada Your file Votre reference Our file Notre reference The author has granted a non-exclusive licence allowing the National Library of Canada to reproduce, loan, distribute or sell copies of this thesis in microform, paper or electronic formats. The author retains ownership of the copyright in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur a accorde une licence non exclusive permettant a la Bibliotheque nationale du Canada de reproduire, preter, distribuer ou vendre des copies de cette these sous la forme de rnicrofiche/film, de reproduction sur papier ou sur format electronique. L'auteur conserve la propriete du droit d'auteur qui protege cette these. Ni la these ni des extraits substantias de celle-ci ne doivent etre imprimes ou autrement reproduits sans son autorisation. 0-612-25013-X Canada 

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.831.1-0087945/manifest

Comment

Related Items