Molecular Diagnosis of Salmonid Diseases [1 ed.] 978-90-481-5974-1, 978-94-017-2315-2

Citation preview

Molecular Diagnosis of Salmonid Diseases

Reviews: Methods and Technologies in Fish Biology and Fisheries VOLUME 3

Series editor:

Jennifer L. Nielsen

u.

S. Geological Survey Biological Resources Division Anchorage, Alaska

The titles published in this series are listed at the end of this volume.

Molecular Diagnosis of Salmonid Diseases

Edited by

Carey O. Cunningham FRS Marine Laboratory, Aberdeen

SPRINGER-SCIENCE+BUSINESS MEDIA, BV.

A C.1. P. Catalogue record for this book is available from the Library of Congress.

ISBN 978-90-481-5974-1 ISBN 978-94-017-2315-2 (eBook) DOI 10.1007/978-94-017-2315-2

Printed on acid-free paper

AII Rights Reserved

© 2002 Springer Science+Business Media Dordrecht Originally published by Kluwer Academic Publishers in 2002 Softcover reprint of the hardcover 1st edition 2002 No part of the material protected by this copyright notice may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage and retrieval system, without written permission from the copyright owner.

General Series Preface Reviews: Methods and Technology in Fish Biology and Fisheries published by Kluwer Academic Publishers is a book series dedicated to the publication of information on advanced, forward-looking methodologies, technologies, or perspectives in fish and fisheries. This series is especially dedicated to relevant topics addressing global, international concern in fish and fisheries. Humans continue to challenge our environments with new technologies and technological applications. The dynamic creativity of our own species often tends to place the greatest burden on our supporting ecosystems. This is especially true for aquatic networks of creeks, lakes, rivers and ocean environments. We also frequently use our conceptual powers to balance conflicting requirements and demands on nature and continue to develop new approaches and tools to provide sustainable resources as well as conserve what we hold most dear on local and global scales. This book series will provide a window into the developing dynamic among humans, aquatic ecosystems (both freshwater and marine), and the organisms that inhabit aquatic environments.

There are many reasons to doubt the increasing social and economic value technology has gained over the last two centuries. Science and technology represent stages in human development. I agree with Ernst Mayer when he said in Toward a New Philosophy of Biology (1988) that "endeavors to solve all scientific problems by pure logic and refined measurements are unproductive, if not totally irrelevant." Living aquatic systems are extremely complex environments that appear alien to most human beings. We only "go there" in a limited capacity and for relatively short periods of time. To reduce these biological systems to simple physio-chemical processes that can be controlled by technology, and made subject to arbitrary management decisions or broad arm-waving policies is to deny their unique nature. That is certainly not what is meant by "Methods and Technology" in this series. Rather, I want to provide a forum for discussions on the living systems themselves and the organisms inhabiting them through new applications of science and technology with special emphasis on aspects of fish and fisheries under-represented in the current literature. Our understanding of aquatic biology in freshwater and marine environments demands a careful and protracted approach ranging broadly from studies of genes, regulatory processes, isolating mechanisms, individual behavior, population structure, biodiversity, to interactive ecosystems. The tools and technologies that allow these investigations change rapidly, always removing old uncertainties and creating new ones. The intent of this series is to monitor that change and document perspectives and developments that mark a fundamental re-evaluation of nature in aquatic habitats and its role in relationship to human society and resource management. Dr. Jennifer L. Nielsen, Series Editor Reviews: Methods and Technology in Fish Biology and Fisheries

us Geological Survey, Biological Resource Division Alaska Biological Science Center Anchorage, Alaska v

CONTENTS Series editor's preface ....................................................................... ix Preface ....................................................................................... ... xi List of contributors ... ..................................................................... xiii Chapter 1 ....................................................................................... 1 MOLECULAR DIAGNOSIS OF INFECTIOUS SALMON ANAEMIA Siri Mjaaland, Espen Rimstad and Carey O. Cunningham Chapter 2 ..................................................................................... 23 DIAGNOSIS AND IDENTIFICATION OF IPNV IN SALMONIDS BY MOLECULAR METHODS Carlos P. Dopazo and Juan L. Barja Chapter 3 ..................................................................................... 49 MOLECULAR DIAGNOSIS OF INFECTIOUS HEMATOPOIETIC NECROSIS VIRUS AND VIRAL HEMORRHAGIC SEPTICEMIA VIRUS James R. Winton and Katja Einer-Jensen Chapter 4 ...................................................................................... 81 MOLECULAR DIAGNOSIS OF AEROMONAS SALMONICIDA INFECTIONS Duncan J. Colquhoun and Carey O. Cunningham Chapter 5 ...................................................................................... 99 THE GENERA FLAVOBACTERIUM AND FLEXIBACTER Joel A. Bader and Clifford E. Starliper Chapter 6 .................................................................................... 141 THE BIOLOGY AND MOLECULAR DETECTION OF PISCIRICKETTSIA SALMONIS Marcia L. House and John L. Fryer Chapter 7 .................................................................................... 157 COMPARISON OF TRADITIONAL AND MOLECULAR METHODS FOR DETECTION OF RENIBACTERIUM SALMONINARUM Ron J. Pascho, Diane G. Elliott and Dorothy M. Chase Chapter 8 ................................................................................... . 211 MOLECULAR APPROACHES FOR THE STUDY AND DIAGNOSIS OF SALMONID STREPTOCOCCOSIS Jesus L. Romalde and Alicia E. Toranzo Chapter 9 ... ................................................................................. 235 GYRODACTYLUS SALARIS MALMBERG, 1957 (PLATYHELMINTHES: MONOGENEA) Carey O. Cunningham

viii

Chapter 10 .................................................................................. 267 MOLECULAR DIAGNOSTICS FOR LOMA SALMONAE AND NUCLEOSPORA SALMONIS (MICROSPORIDIA). Amanda M. V. Brown and Michael L. Kent Chapter 11 .................................................................................. 285 MOLECULAR TOOLS FOR THE DIAGNOSIS OF CERA TOMYXA SHASTA (MYXOZOA) Oswaldo Palenzuela and Jerri L. Bartholomew Chapter 12 .................................................................................. 299 PCR AND IN SITU HYBRIDISATION OF TETRA CAPSULA BRYOSALMONAE (PKX), THE CAUSATIVE AGENT OF PROLIFERATIVE KIDNEY DISEASE David J. Morris and Alexandra Adams Chapter 13 .................................................................................. 315 NUCLEIC ACID - BASED METHODS FOR DETECTION OF MYXOBOLUS CEREBRALIS Karl B. Andree and Dolores B. Antonio Chapter 14 ... ................................................................ , ....... ....... 329 DIAGNOSING KUDOA THYRSITES (MYXOZOA: MYXOSPOREA) INFECTIONS IN FISH Jonathan D. W. Moran and Michael L. Kent

SERIES EDITOR'S PREFACE The early 21 st century marks the coming-of-age of genetics and molecular processes. Recent developments in molecular techniques for human disease diagnosis and control have had significant carry-over effects in the study of fish and fisheries. Over the last 60 years, molecular approaches in disease research and diagnosis have not only set new, rigorous criteria for veterinary and aquaculture applications, they have brought a new wave of thinking about the evolution of disease organisms, infection pathways, the etiological role of disease agents, and diagnostic methodologies. Compared to more traditional methodologies, genetic diagnoses of salmonid diseases provide significant improvements in speed, sensitivity and precision. The worldwide economic value of salmon has placed these fish at the center of this new technological expansion in disease diagnostics and control, especially in light of the extensive development and growth of salmon aquaculture around the world. Nucleotide sequences of the genomes of different disease strains hold important epidemiological information and allow broad investigations of transmission, vectors, reservoirs of infection, and infection processes in both aquaculture and wild fish populations. Molecular techniques greatly increase opportunities for non-lethal sampling, especially in studies of endangered or commercially valuable fishes. Identification of genetic factors in viral agents, bacterial pathogens and intracellular parasites, molecular detection techniques using DNA probes and PCR assays, and genetic studies of disease resistant stocks have enormous financial implications in aquaculture. Important identification of disease pathways through biotic and abiotic links has been demonstrated using genetic and phylogenetic analyses. Many authors in this volume discuss practical considerations in the development and implementation of these new molecular techniques, i.e. cost of equipment, required expertise of diagnostician, use of controls, cross-contamination concerns, the development and sharing of effective genetic protocols, and systematic evaluation of results. Although significant work remains in the development of methods of molecular diagnosis of diseases in fish, questions of application will develop strength and continuity over time as the use of these techniques spread throughout the fisheries scientific community. This volume marks an important event in our effort to disseminate up-to-date information on the development and application of molecular techniques in the diagnosis of salmonid diseases. Anyone with laboratory applications or responsibility for analyses of fish diseases will appreciate the clear vision of the usefulness of these tools provided in this volume. Dr. Jennifer L. Nielsen, Series Editor Reviews: Methods and Technology in Fish Biology and Fisheries US Geological Survey, Biological Resource Division Alaska Biological Science Center Anchorage, Alaska

PREFACE The last 10 years have seen the establishment of molecular technology as a core component of fish disease research and diagnosis. The number of publications describing new methods of detecting and analysing pathogens appears to increase exponentially, with several alternative methods described for detecting many of the pathogens included in this volume. We have now reached a stage where there are many molecular diagnostic methods reported, but information on their suitability in practice or their relative fitness for different purposes is more difficult to obtain. This situation prompted the inception of this book, which aims to redress this imbalance. Chapter authors are those with experience of the practical application of the techniques described. In many cases, the authors of the chapters in this book are also leading names in the development of the techniques they describe. Of course, this does lead to some bias, and many chapters acknowledge that different laboratories will have differing work practices, conventions and preferences. Nevertheless, the experiences of the authors presented here are invaluable to those wishing to establish molecular tests in their own laboratories. Many of the following chapters provide detailed information on the diseases and their causative organism, where this is available. For some diseases, such as proliferative kidney disease caused by Tetracapsula bryosalmonae, the causative organism has only recently been described. In this case, DNA sequences were important in this identification and further information will undoubtedly follow rapidly as a result of the ability to detect and identify this parasite using probes and PCR. Many detailed protocols are also given, and useful information on how best to apply published methods. These valuable details are often difficult to glean from the scientific literature alone and we are fortunate to have gathered so much expertise in these matters in one volume. As can be seen in the topics included in each chapter, some authors consider the use of monoclonal antibodies etc. to fall within the definition of 'molecular techniques', but the book overall emphasises nucleic acidbased methods. Inevitably, there have been omissions. Thanks to a large number of experienced reviewers, the following chapters are themselves comprehensive. However, in a field that develops as fast as this, there will already be improved methods or adaptations for diagnosing some of the diseases included here. Some diseases have not been included as it was impossible to cover them all and maintain the required level of detail in each chapter. Perhaps the most serious lack however, is the lack of critical discussion of the application of molecular diagnostics per se. Many authors have referred to the work of Maura Hiney and others in drawing our attention to the requirement for fastidious evaluation of tests and their ability to fulfil the desired aims. Despite the huge advantages of molecular technology, in some cases it has not fulfilled the promise of extremely fast and sensitive diagnosis and we should remain aware of the limitations as well as the successes of each method for each particular application. Judicious application of molecular diagnosis will ensure the successful development of this field and we can look forward to many exciting developments in the future. Carey O. Cunningham FRS Marine Laboratory. Aberdeen. Scotland

LIST OF CONTRIBUTORS Alexandra Adams Institute of Aquaculture, University of Stirling, Stirling, FK9 4LA, Scotland Karl B. Andree Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California at Davis, One Shields Ave., Davis, CA 95616, USA Dolores B. Antonio Department of Medicine and Epidemiology, School of Veterinary Medicine, University of California at Davis, One Shields Ave., Davis, CA 95616, USA Joel A. Bader United States Department of Agriculture, Agriculture Research Service, Aquatic Animal Health Research Unit, 990 Wire Road, Auburn, Alabama 36830, USA Juan L. Barja Departamento de Microbioloxfa e Parasitoloxfa, Facultade de Bioloxfa, Universidade de Santiago de Compostela, 15706. Spain Jerri L. Bartholomew Oregon State University. Dept. of Microbiology. Nash Hall 220. Corvallis, OR. 973313804 USA AmandaM. V. Brown 6270 University Blvd., Department of Zoology, University of British Columbia, Vancouver, BC V6T lZ4 Canada Dorothy M. Chase Biological Resources Division, Western Fisheries Research Center, 6505 N.E. 65 th Street, Seattle, WA 98115 USA Duncan J. Colquhoun Section for Fish Health, National Veterinary Institute, PO Box 8156 Dep., N-0033 Oslo, Norway Carey O. Cunningham FRS Marine Laboratory, PO Box 101, Victoria Road, Aberdeen AB11 9DB, Scotland Carlos P. Dopazo Departamento de Microbioloxfa e Parasitoloxfa, Instituto de Acuicultura, Universidade de Santiago de Compostela, 15706. Spain Katja Einer-Jensen Danish Veterinary Laboratory, Hang!1lvej 2, DK-8200 Aarhus N, Denmark

XIV

Diane G. Elliott Biological Resources Division, Western Fisheries Research Center, 6505 N.E. 65 th Street, Seattle, W A 98115 USA John L. Fryer Department of Microbiology and the Center for Salmon Disease Research, 220 Nash Hall, Oregon State University, Corvallis, Oregon 97331-3804, USA Marcia L. House Northwest Indian Fisheries Commission, 6730 Martin Way E., Olympia, WA, 985165540, USA Michael L. Kent Center for Salmon Disease Research, Department of Microbiology, 220 Nash Hall, Oregon State University, Corvallis, OR 97331 USA Siri Mjaaland Department of Pharmacology, Microbiology and Food Hygiene, The Norwegian School of Veterinary Science, P.O. Box 8146 Dep., N-0033 Oslo, Norway Jonathan D.W. Moran Microtek International Limited, c/o Institute for Marine Biosciences, National Research Council of Canada, 1411 Oxford Street, Halifax, NS B3H 3Z1, Canada DavidJ .Morris Institute of Aquaculture, University of Stirling, Stirling, FK9 4LA Scotland Oswaldo Palenzuela Institute of Aquaculture "Torre la Sal" (CSIC), 12595 Ribera de Cabanes, Caste1l6n, Spain Ronald J. Pascho Biological Resources Division, Western Fisheries Research Center, 6505 N.E. 65 th Street, Seattle, WA 98115 USA Espen Rimstad Department of Pharmacology, Microbiology and Food Hygiene, The Norwegian School of Veterinary Science, P.O. Box 8146 Dep., N-0033 Oslo, Norway Jesus L. Romalde Departamento de Microbiologfa y Parasitologfa. Facultad de Biologfa. Universidad de Santiago de Compostela. 15706, Santiago de Compos tela, Spain Clifford E. Starliper United States Department of Interior, United States Geological Survey/BRD, National Fish Health Research Laboratory, 1700 Leetown Rd. Kearneysville, WV 25430, USA

xv

Alicia E. Toranzo Departamento de Microbiologia y Parasitologia. Facultad de Biologia. Universidad de Santiago de Compostela. 15706, Santiago de Compostela, Spain James R. Winton Western Fisheries Research Center, 6505 NE 65th Street, Seattle, Washington 98115, USA

Chapter 1 MOLECULAR DIAGNOSIS OF INFECTIOUS SALMON ANAEMIA

Siri Mjaaland', Espen Rimstad' and Carey O. Cunningham 2 / The Nonvegian School of Veterinary Science, Oslo, Nonvay Marine Laboratory, Aberdeen, UK

2FRS

1. DESCRIPTION OF THE PROBLEM Infectious salmon anaemia (ISA) is a viral disease of farmed Atlantic salmon (Salmo salar, Salmonidae). The disease has caused considerable economic loss to the Atlantic salmon farming industry. Initially, ISA was reported to occur in Norway, and for more than ten years the disease was only recognised in this country. Infectious salmon anaemia was recognised by the OlE in 1990 and the disease is listed in the OlE International Aquatic Animal Health Code (Office International des Epizooties (OlE), 1997b) and Diagnostic Manual for Aquatic Animal Diseases (Office International des Epizooties (OlE), 1997a) as one of the "Other significant diseases". In European Union Directive 93/53/EEC the disease is classified on "the green list" as an exotic disease. In 1996/97 the disease condition "haemorrhagic kidney syndrome" in Atlantic salmon reported from Canada (Byrne et aI., 1998) was found to be ISA (Lovely et aI., 1999, Mullins et aI., 1998, Bouchard et aI., 1999), and in 1998 ISA was reported officially from Scotland at the 66 th General Session of the OlE (Rodger et aI., 1998). During 2000 and 2001 outbreaks of ISA have been confirmed on the Faroes and USA (Maine) (Bouchard et aI., 2001). The disease is now officially recognised to occur in these five countries and has been reported from Chile (Kibenge et aI., 2001)

1.1. History, Status and Distribution of ISA Norway. The first recorded outbreak of ISA occurred in Norway in late 1984, in parr from a hatchery that had used seawater as part of the water supply. The cumulative mortality in this hatchery was 80 %. Affected fish were lethargic and severely anaemic. Other pathological signs were ascites, petechiae in the adipose tissue and swimbladder and haemorrhagic liver necrosis (Thorud, 1991). During the following year, ISA was spread to farms receiving smolts from the hatchery with the initial outbreak, and later the disease spread to the neighbouring farms where affected fish had been slaughtered. The incidence of the disease peaked in 1991, with 90 C. 0. Cunningham (ed.), Molecular Diagnosis of Salmonid Diseases, 1-22. © 2002 Kluwer Arademic Publishers.

2

MJAALAND, RIMST AD & CUNNINGHAM

recorded outbreaks (Haste in, 1997). Due to the severity of the disease, ISA became a notifiable disease in Norway in 1988, long before much knowledge existed on the aetiology, epizootiology and pathogenesis of the disease. A control programme was initiated and a radical decrease in affected farms was recorded. ISA has been found in all geographical areas of the Norwegian coast except a small area in the south that has few fish farms (Hastein, 1997). Canada. In Canada, "haemorrhagic kidney syndrome" (HKS) in the New Brunswick area was suspected to be ISA in 1996, but this was ruled out (Byrne et ai., 1998). HKS was confirmed as ISA a year later (Lovely et ai., 1999) and other reports of the disease published secondary to the identification (Mullins et ai., 1998, Bouchard et ai., 1999). In 1997,19 of 20 sites in the area where ISA first occurred were affected. The disease spread outwards during 1998 and the virus was detected in Nova Scotia in 1999 (Hastings et ai., 1999; Ritchie et ai., 2001). Scotland. In Scotland, ISA was first detected in a fish farm in Loch Nevis in 1998. The disease spread by fish movements to the Isle of Skye, Loch Creran, where fish were moved for harvest and processing, and to the Shetland Isles (Stagg et ai., 2001). By September 2000 there had been 11 confirmed cases (Hastings et aI., 1999).

1.2. Symptoms and Signs of the Disease The cumulative mortality during an outbreak of ISA varies considerably from insignificant to moderate although some farms suffer losses that exceed 90 %. Experimental trials with smolts from different farms or genetic backgrounds have been used to show that mortality can vary between 10 to 100 % (Nylund et ai., 1995a). In experimentally infected fish the incubation time is between 10-20 days (Thorud, 1991, Dannevig et aI., 1994, Totland et ai., 1996, Rimstad et ai., 1999). In an affected population of farmed fish, some individuals may harbour the virus for weeks or months before succumbing to ISAV-induced disease (Rimstad et aI., 1999). Prior to an outbreak, a slightly increased mortality is often seen. The disease outbreak itself, with a dramatic rise in mortality, usually develops within 1-3 weeks often following a stress situation. The outbreak is usually restricted to one or two netpens. Classical ISA V-affected fish appear lethargic and may keep close to the walls of the netpen. In terminal stages, diseased fish often sink to the bottom of the cage. The further development of the disease varies, and up to 12 months can pass before clinical ISA has spread to all the netpens in a fish farm (Thorud, 1991, Vagsholm et ai., 1994). Typically, most ISA outbreaks develop during the spring or early summer, in water temperatures between S-ISoC, while a minor peak is reached during the autumn! winter months (Thorud, 1991). Two main forms of ISA outbreaks have been described: an acute form, with a rapid development and high mortality, or a more chronic disease, in which a slow increase in mortality can be observed during several months. The most prominent external signs of the acute form are pale gills, exophthalmus and sometimes haemorrhage in the eye chamber and on the skin. Internally, the principal pathological features are ascites, dark liver (haemorrhagic liver necrosis), swollen spleen and congested intestinal wall as well as petechiae in the adipose tissue and swimbladder (Thorud and Djupvik, 1988, Evensen et ai., 1991, Thorud, 1991). The major histopathological finding is multifocal haemorrhagic liver necrosis that may become confluent (Evensen et ai., 1991, Thorud, 1991). The lesions result in extensive

ISAV

3

congestion of the liver with dilated sinusoids and in later stages, the appearance of blood filled spaces (Evensen et al., 1991). Moribund fish are severely anaemic and haematocrit values below 5 are common. In the chronic form of ISA, the signs are more diffuse and can be difficult to interpret. The liver may appear pale or yellowish and the anaemia may not be as severe as in the acute disease. Furthermore, there is less ascitic fluid, but haemorrhages in the skin and swim bladder and oedema in the scale pockets and swim bladder may be more pronounced than in acutely diseased fish (Evensen et al., 1991). This classification, however, is not absolute and variation in the pathological changes and the severity of these has recently become more evident. The signs exhibited by infected fish range from none to severe, depending on factors such as infective dose, temperature, season, age, immune status, virus strain and pathogenicity. The ability to demonstrate the presence of ISA virus (lSAV) has been the decisive diagnosis of the disease. In some instances the pathological changes are expressed to a greater extent than usual, while other pathological changes are totally or partly lacking. This was emphatically demonstrated in the spring of 19%, when a new syndrome appeared in Atlantic salmon on the Canadian east coast. The major pathological findings were histological, and included renal interstitial haemorrhage with tubular necrosis and casts. The disease therefore became known as haemorrhagic kidney syndrome (HKS) (Byrne et al., 1998). Gross examination of the fish appeared somewhat similar to the reports of ISA from Norway, however, in the New Brunswick fish, gross liver congestion was a rare finding (Byrne et al., 1998, Mullins et al., 1998, O'Halloran et al., 1999). HKS was confirmed as ISA using RTPCR and cell culture followed by an indirect fluorescent antibody test (IFAT) (Lovely et al., 1999). Interestingly, gross hepatic congestion ofISAV infected fish has become a more common finding in New Brunswick (Mullins et al., 1998). Although the pathological changes found to be typical for HKS are not described from the Norwegian ISA epidemic (e.g. Evensen et al., 1991), subsequent archival reviews have revealed that HKS-like changes have been found in a number of tissue samples collected from ISAV outbreaks in Norway (Lovely et al., 1999, T.Poppe & T.HAstein, National Veterinary Institute, Oslo, pers. comm.). The apparent differences in the pathological changes found in HKS may have several possible explanations. The presentation of ISAV infection as HKS may be more prevalent under the environmental conditions peculiar to aquaculture on the east coast of Canada, or due to concurrent infections or strain differences of ISAV, or the stock of Atlantic salmon differing from Norwegian stocks in their susceptibility to infection. (Kibenge et aI., 2000b, Lovely et aI., 1999, Mullins et al., 1998). In Scotland, the combination of the gross pathological findings, histopathology and haematology were themselves indicative of ISA (Office International des Epizooties (OlE), 1997a). The cytopathic effect after inoculation of SHK-l cells, combined with kidney impression smears screened by direct immunofluorescence test using anti-ISAV monoclonal antibody, and RT-PCR, assisted in confirming the first diagnosis ofISA in Scotland (Rodger et al., 1998, Stagg et al., 1999).

4

MJAALAND, RlMST AD & CUNNINGHAM

1.3. The Agent By electron microscopy, ISAV was shown to be an enveloped virus, 100-l30 nm in diameter, replicating by budding from the membrane of infected cells (Hovland et aI., 1994, Dannevig et aI., 1995a, Nylund et aI., 1995b). Endothelial cells seem to be the main target cell for ISAV (Hovland et aI., 1994; Gregory, FRS Marine Lab., pers. comm.). Early attempts to isolate the virus in commercial fish cell lines were unsuccessful, and the virus was first isolated about 10 years after the first outbreak of IS A, using a cell line (SHK-l) established from Atlantic salmon head kidney (Dannevig et aI., 1995a). The morphological, physiochemical and genetic properties of ISA V closely resemble those of the Orthomyxoviridae (Falk et aI., 1997, Mjaaland et aI., 1997, Sandvik et aI., 2000, Rimstad et aI., 2001). Like the influenza A and B viruses, the genome of ISA V consists of eight single stranded RNA segments of negative polarity (Mjaaland et aI., 1997). The sequence of the terminal 8-9 nucleotides at both ends of two of the viral RNA segments are identical and show distinctive sequence homology with the conserved terminal sequences found in the orthomyxoviruses (Sandvik et aI., 2000). Like the influenza viruses, the terminal 21-24 nucleotides seem to form selfcomplementary panhandle structures that are important for transcriptional regulation of viral RNA (Fodor et aI., 1994, Sandvik et aI., 2000). For transcription of ISAV, 818 nucleotide 5'-cap structures cleaved from cellular heteronuclear RNAs (capstealing) are required, and the mRNA is polyadenylated from a signal 13-14 nucleotides downstream of the 5'-end terminus of the vRNA (Sandvik et aI., 2000). Analyses of the nucleotide sequences of genomic segments numbers 2, 3, 4, 6, 7 and 8 of ISA V have been reported (Mjaaland et aI., 1997, Krossey et aI., 1999, Cunningham and Snow, 2000; Rimstad et al. 2001; Ritchie et aI., 2001b; Ritchie et aI., in prep) (Table 1). Neither the nucleotide sequence of genomic segments 3, 4, 6, 7 or 8 nor the putatively polypeptides show significant homologies to known orthomyxovirus sequences (Mjaaland et aI., 1997; Snow & Cunningham, 2001; Rimstad et aI., 2001; Ritchie et aI., 2001b; Ritchie et aI., in prep.). However, the sequence of genomic segment 2 encoding for the PBl part of the viral RNA polymerase shows similarity to RNA polymerases of negative stranded RNA viruses (Krossey et al. 1999). Analysis of the nucleotide sequence of genomic segment 6, encoding the haemagglutinin (accession no. AF220607 and AJ276859) (Rimstad et aI., 2001), segment 3, encoding the nucleoprotein (accession no. AJ276858 and AF306549) (Ritchie et aI., 2001b; Snow & Cunningham, 2001), segment 4 encoding a polymerase (accession no. AF306548) (Ritchie et aI., 2001b), and segment 7, encoding the matrix (Ritchie et a\., in prep.), have been undertaken. The partial sequence of segment 1, encoding polymerase PA and complete sequence of segment 5 encoding the receptor destroying enzyme have also been determined (Snow, FRS Marine Laboratory, pers. comm.).

5

ISAV

Protein ISAV 'flu B First report I('flu equivalent) segment segment Polymerase (PB 1) 1 Snow, unpubl. 1 Polymerase (PB2) Kross0)' et al. (1999) 2 2 Nucleoprotein (NP) Snow & Cunningham (2001) 3 5 Ritchie et al. (2001 b) Polymerase (PA) 4 3 Snow, in prep. Receptor destroying 5 6 enzyme (NA) Haemagglutinin (HA) Rimstad et al. (2001) 4 6 Matrix (M)

7

7

Ritchie et ai. (in prep.)

Non-structural (NS)

8

8

Mjaaland et ai. (1997)

Accession numbers AJ002475 AJ276858 AF306548

AF220607 AJ276859 AF328627 AF361363 AJ306487 AJ306488 Y10404

Table 1. Genome organisation ofISAV

Variation in the virulence and antigenic composition of ISAV is indicated by differences in disease development, clinical signs, and susceptibility to cell cultures (Bouchard et aI., 1999, Kibenge et aI., 2000a, Ritchie et aI., 2001). In a sequence comparison of PCR products representing genome segments 2 and 8 carried out on different Norwegian, Canadian and Scottish virus isolates, it turned out that both segments showed high homology between the Norwegian and Scottish isolates, while the Canadian virus was less similar (Blake et aI., 1999, KrOSS0)' et aI., 1999, Cunningham and Snow, 2000, Inglis et aI., 2000). The origin of the ISAV infection on the East coast of Canada is a contentious issue. The preliminary nucleotide sequence comparison, however, suggested that there are no direct epidemiological connections to the Glesvaer ISA V isolate (Lovely et aI., 1999, Blake et aI., 1999). Within Canadian isolates, it was determined that culture compatibility was reflected by variation in protein profiles (Kibenge et aI., 2000a), with nucleotide variation determined subsequently (Ritchie et aI., 2001). Analysis of the nucleotide sequences of segments 2, 6 and 8 of ISAV isolates from confirmed outbreaks of ISA in Scotland did not reveal direct connections to the Glesvaer isolate and indicated that the outbreaks probably stemmed from a point source (Cunningham and Snow, 2000; Inglis et aI., 2000; Stagg et aI., 2001). However, one isolate, from a farm that was not confirmed with the disease, shows remarkable similarity to the Glesvaer isolate (Inglis et aI., 2000). Other sequences, from wild fish, indicate the existence of several variants of ISAV in Scottish waters (Raynard, 2000, Raynard et aI., in press). Indications of recombination and reassortment in ISAV have recently been found (Nylund, personal communication; Oslo laboratory). It may be the case that only certain types of ISAV, analogous to influenza types, are pathogenic to Atlantic salmon.

6

MJAALAND, RIMST AD & CUNNINGHAM

1.4. Scale of Problems in Wild and Cultured Salmonids In Norway, ISAV has been detected along most of the coast. Outbreaks of ISA have only been described in Atlantic salmon. Epidemiological studies have shown that ISA is mainly transmitted from infected salmonid sources to clean sites through sea water (Iarp and Karlsen, 1997, Vagsholm et aI., 1994). True vertical transmission of ISA is thought to be unlikely although transmission via ovarian fluids may be possible (Totland et aI., 1996, Melville and Griffiths, 1999). The infection spreads slowly. The infection is mainly observed in fish held in seawater or in fish exposed to seawater, but indications of disease outbreaks in fish held in fresh water have also been reported (Nylund and Jakobsen, 1995). ISAV is stable at pH 5.7-9.0 (Falk et aI., 1997). The optimum growth temperature in susceptible fish cell lines is 10-15 D C (Dannevig et aI., 1995a). The virus may survive in seawater for months, while the survival period in fresh water is shorter (Falk et aI., 1997, 0ye & Rimstad, in prep). Under experimental conditions the virus may survive and replicate in sea trout (Sa/mo trutta), rainbow trout (Oncorhynchus mykiss), Atlantic herring (C/upea harengus), and haddock (Melanogrammus aeglfinus), which thus may act as asymptomatic carriers of the virus (Nylund and Jakobsen, 1995, Nylund et aI., 1995c, Nylund et aI., 1997, Rolland and Nylund, 1998a, Griffiths, personal communication). ISAV does not seem to replicate in non-salmonid fish like turbot (Psetta maxima), ballan wrasse (Labrus bergylta), sea bass (Dicentrarchus labrax), or cod (Gadus morhua) (Thorud and Torgersen, 1994, Hastein, 1997). Under experimental conditions, wild Atlantic salmon are susceptible to ISA and show the same clinical symptoms as farmed fish (Nylund et aI., 1995a). ISA has emerged where no connection to existing outbreaks can be found, suggesting that wild fish could act as vectors of the disease. Screening of wild fish in Norway, mainly sea trout but also salmon has not provided evidence that ISAV exists in wild fish. In Scotland, wild fish screening by RT-PCR (Mjaaland et aI., 1997) has also been carried out and evidence for the existence of ISAV in wild fish is increasing (Raynard, 2000, Raynard et aI., in press). In Canada, the virus has been detected in wild plaice Hippoglossoides platessoides (Anne-Margaret MacKinnon, DFO Moncton, personal communication). Under experimental conditions salmon lice can transfer infection between fish (Rolland and Nylund, 1998b). Furthermore, one cannot exclude the possibility that birds can spread infectious material, both within and between different locations.

1.5. "Traditional" Methods of Diagnosis The OlE guidelines (Office International des Epizooties (OlE), 1997a) are used worldwide as a code of practice for confirmation of ISA. Within the European Community, the legal criteria for diagnosis of ISA are contained in Directive 93/531EEC. The OlE Reference Laboratory for ISA in Norway lists necessary requirements to confirm the presence of viral antigen in typical lesions, and to resolve cases with inconclusive disease signs. These are clinical, pathological, histopathological and haematological changes, together with detection of ISAV by means of virus cultivation (Dannevig et aI., 1995a, Dannevig et aI., 1995b) or detection of ISA V by an indirect immunofluorescent antibody test (IF AT) (Falk and Dannevig, 1995, Falk et aI., 1998). The EU Community Reference Laboratory in Denmark has recommended that diagnostic criteria for confirmation should, in

ISAY

7

addition to confinnation of disease by clinical signs and post mortem examination, also include histopathological changes and laboratory evidence of virus e.g. IFAT and/or reverse transcriptase polymerase chain reaction (RT-PCR) (Hastings et aI., 1999).

1.5.1. Diagnosis based on clinical and pathological findings When making a pathological diagnosis the macroscopic, histological and haematological findings should be consistent with the description given for ISA (Byrne et aI., 1998, Evensen et aI., 1991, Hastein, 1997, Mullins et aI., 1998, Rodger et aI., 1998, Thorud and Djupvik, 1988, Thorud, 1991). Confidence in the diagnosis increases if all, or many, of these signs are observed.

1.5.2. Isolation of virus in cell culture Spleen, heart, liver and preferably kidney are suitable organs for virus isolations from clinically infected fish. The SHK-l cell line established from Atlantic salmon head kidney (Dannevig et aI., 1995a, Dannevig et aI., 1995b, Dannevig et aI., 1997) is susceptible to ISAV. Following inoculation with tissue homogenate cytopathic effect (CPE) develops after 10-14 days (or less (Griffiths, pers. comm.» incubation at 15°C, or after serial passages of cell culture medium to new cell cultures. Identification of isolated virus is perfonned with an indirect fluorescent antibody test (IF AT) using anti-ISAV monoclonal antibody as primary antibody (Falk and Dannevig, 1995, Falk et aI., 1998). Propagation of ISAV in other continuous cell lines, such as AS, TO and CHSE-214 cells have been reported (Sommer and Mennen, 1997, Bouchard et aI., 1999, Kibenge et aI., 2000a, Opitz et aI., 2000, Wergeland and Jakobsen, 2001). Further, the ASK cell line from Atlantic salmon also supports growth of the virus (DevoId et aI., 2000). However, the SHK-l cell line is the only one reported to support growth for all known isolates ofISAV (Kibenge et aI., 2000a). 1.5.3. Direct detection of virus in clinical material An indirect fluorescent antibody test (IF AT) using anti-ISAV monoclonal antibody on frozen tissue sections of kidney, heart and liver or on kidney imprints (Falk and Dannevig, 1995, Falk et aI., 1998) has given positive reactions in both experimentally and naturally infected Atlantic salmon (Rimstad et aI., 1999). Cases suspected from pathological signs are verified using IFAT. The criteria for ISA confinnation in Norway, Canada and Scotland are broadly similar. Clinical signs are a requirement in all three countries, as is laboratory evidence of the presence of virus. All three countries also require the presence of appropriate histopathology and in Canada this must be present in a minimum of two fish. More than one positive IF AT test is required in both Scotland and Canada but no exact numbers have been specified in Norway. In Scotland, pools of 5 fish are tested

8

MJAALAND, RIMSTAD & CUNNINGHAM

by RT-PCR or cell culture. In Canada the requirement is a positive result from at least two fish and cell cultures being confirmed by RT-PCR or IFAT. In Norway suspected cell cultures must be validated by IFAT (Hastings et aI., 1999).

2. MOLECULAR DIAGNOSIS OF ISA Most of the diagnostic tools established for ISAV detection have been based on diseased fish containing a high virus load. These diagnostic procedures may not be well suited for detection of virus when the viral load in the individual fish is low, such as early or late in the course of the disease. In Norway, where the requirements for an ISA diagnosis today are clinical, histological and haematological observations in accordance with ISA, or detection of ISA V by IFAT together with suspicious histological and haematological observations, several cases have been identified where post mortem examination and histological observations have been diffuse or atypical or cases where other diseases have been present and complicated the pathological picture. For future diagnosis of ISA it is therefore hoped that detection of virus will have increased importance. Early detection of ISA V is critical to the implementation of effective control and preventive measures, as well as a better understanding of its epizootiology (or epidemiology). There is currently no vaccine or treatment commercially available for ISA. Therefore, the control of the disease relies on reducing the exposure of farmed salmon to the virus. RT-PCR has been shown to be an efficient alternative to virus isolation for the detection of influenza virus in clinical specimens from several mammalian species (Cherian et aI., 1994, Donofrio et aI., 1994), and the method is particularly effective for the identification of influenza virus in samples obtained late in the course of infection (Cherian et aI., 1994). To achieve high sensitivity in RT-PCR of tissue samples, there are several critical steps to consider in the process oftesting, which is shown schematically in Figurel. Influenza virus, especially influenza A, appears in several different variants. Variation in the virulence and antigenic composition of ISAV is indicated by differences in disease development, clinical signs and differences in the susceptibility of cell cultures (Bouchard et aI., 1999, Kibenge et aI., 2000a, Ritchie et aI., 2001). There is evidence that ISA V also undergoes recombination and reassortment (Nykund, personal communication; Mjaaland, unpublished results) events. These may of course affect the applications for which different diagnostic tests are best suited. Various methods for molecular diagnosis of ISAV have been described in the literature. This chapter will examine each step in the process and discuss attempts to optimize each step.

9

ISAV

Sampling and transportation of sample material

l Sample arrival

l Homogenization of sample l +- Backup sample -70°C RNA extraction

• •

RNA quantification Reverse transcription of RNA

l PCR

l Agarose Gel Visualisation

Ｏｬｾ＠

RFLP

sequence

Southern blot & probe

Figure I. Schematic diagram of molecular diagnosis of ISA

2.1. Sampling and Transport Samples of fish tissue taken for virus detection and identification have traditionally been transported in a solution or medium designed to maintain viability of the virus. These solutions have significant disadvantages when molecular methods are to be applied. No substances are present to prevent degradation of RNA or DNA and this degradation can prove significant. Therefore, trials were undertaken to assess the effect of differing sampling and storage conditions on detection of ISAV by RTPCR. Due to logistical constraints, it was only possible to carry out these trials using material from aquarium experiments where salmon had been injected with relatively large doses of ISAV. Therefore the viral loads in these samples were high and may not accurately reflect the situation found in clinical samples obtained from fish farms. Nevertheless, it was possible to qualitatively assess differences between samples treated in different ways. Kidney tissue (approx. O.5g) was placed in 5 ml transport solution (Earl's balanced salt solution Ix, fungizone 5 "glml, polymixin B 200 Vlml and kanamycin 200 "glml (Life Technologies, Paisley, UK», 1ml TRIzol (Life Technologies, Paisley, UK), 1ml

10

MJAALAND, RIMSTAD & CUNNINGHAM

ethanol or Iml RNAlater (Ambion Inc., Austin, USA). Replicate samples were stored at room temperature, 4°C or -20·C for periods of between 24 hours and 1 month prior to homogenisation and RT-PCR. Although, given the high virus titre in each sample, strong positive results were obtained, it could be seen that RNA later and TRIzol produced better results than transport solution. As TRizoI is a more hazardous solution than RNAlater and personnel carrying out sampling may have to transport and handle many tubes containing the sampling solution, RNAlater has been used in preference to TRIzol in Scotland following recommendations from Canada (Griffiths, pers. comm.). This procedure provides excellent preservation of RNA during the sometimes protracted period between sampling and receipt at the laboratory and minimises the risk of sample degradation if there is any unexpected delay in transport.

2.2. RNA Extraction 2.2.1. Homogenization of sample Fresh or thawed tissue samples can either be homogenized in plastic bags containing PBS (to 0.01 g tissue add 0.1 ml DEPC-PBS) (Mjaaland et aI., 1997, Krossey et aI., 1999, Rimstad et aI., 1999) or the tissues can be homogenized directly in a detergent (Devoid et aI., 2000, Opitz et aI., 2000). Homogenization can either be performed using a manual roller (Rimstad et aI., 1999), a Micro Centrifuge Sample Pestle (Opitz et aI., 2000), or mechanically, with beads (Griffiths, pers. comm.). The cell debris can be removed by centrifugation (Opitz et aI., 2000), however this does not seem to be necessary (Mjaaland et aI., 1997, Krossey et aI., 1999, Rimstad et aI., 1999, Devoid et aI., 2000). Tissues placed in RNAlater will soften during storage and sometimes do not require a separate homogenization step.

2.2.2. Extraction of RNA Total RNA is extracted from the homogenates by use of a detergent. The type of detergent used varies; however, the most commonly used is TRIzoI® (Life Technologies) (Mjaaland et aI., 1997, Rimstad et aI., 1999, Kross"'}' et aI., 1999, Devoid et aI., 2000), or Trizol LS reagent (Canadian Life Technologies) (Melville and Griffiths, 1999, Kibenge et aI., 2000a) while Opitz and coworkers (Opitz et aI., 2000) used TriReagent (Sigma). Classical phenol:chloroform extraction has also been used (Blake et aI., 1999, Bouchard et aI., 1999). Additional extractions with the detergent, by adding the detergent to the aqueous phase from the fIrst extraction did not result in improved yield of RNA (Mikalsen et aI., in press). The detergent is used according to the manufacturer's instructions. In short (for TRIzol): I ml of organ suspension is mixed with 0.8 ml TRIzol and incubated at room temperature for 5 min. This is followed by addition of 0.16 ml chloroform, vigorous shaking for 15 sec, incubation at room temperature for 2-3 min, and centrifugation at 12,000 rpm for 15 min at 4°C. The aqueous phase is transferred to a fresh tube with 0.5 ml isopropyl alcohol, incubated at room temperature for 10 min, and centrifuged at 6,500 rpm for 15 min at 4°C. The RNA is then washed with 1 ml 75% ethanol, and centrifuged at 6,500 rpm for 5 min at 4°C. Devoid and coworkers (Devoid et aI.,

ISAY

11

2000) added an additional precipitation step and a double wash of the RNA to reduce the amount of possible inhibitors transferred from the host tissue to the RT-PCR. After drying the RNA is resuspended in DEPC-treated water and stored at -70·C. RNeasy mini-spin columns (Qiagen) were tested as an alternative method of RNA extraction (McBeath, FRS Marine Laboratory, unpublished). After homogenzation, fish tissue or cell culture material is passed through a mini-column, and RNA eluted after the column has been washed to remove impurities. This method was found to be less efficient that TRIzol, with lower RNA concentrations and variable OD26oiOD 28o ratios, but has the advantage that it is a faster method and does not involve handling hazardous chemicals. Qualitative results of trials of various RNA extraction systems are shown in Table 2. • Guanidine thiocyanate!l!henol:chloroform solutions RNAzol (Sigma) TriReagent (Sigma)

Opitz et al. (2000)

• Guanidine thiocyanate! Cesium TriOuoroacetateLiCI-solutions High Pure Isolation kit (Boehringer Mannheim)

Genomic DNA is removed by the use of DNase. Is listed to work for cultivated cells, blood, yeast, and bacteria - not for solid tissue

Quick prep Total RNA (Pharmacia) - no DNase treatment.

Does not require the use of hazardous solvent such as phenol

• Guanidine thiocyanate! ethanol! Silica membrane RNeasy-kit (Qiagen)

No DNase. Does not require the use of hazardous solvent such as phenol. Lower yield than TRIzol.

S.N.A.P. ™ Total RNA Isolation Kit (Invitrogen)

Includes DNase treatment

SV Total RNA Isolation System (Promega)

Includes DNase treatment

Guanidine thiocyanate! Phenol !Chloroform (not commercially available)

Has been tested, but no improvement of OD26oiOD 28o ratio or the results ofRTPCR was achieved

Table 2. RNA extraction systems

12

MJAALAND, RlMSTAD & CUNNINGHAM

2.2.3. RNA quantification The concentration and purity of the RNA are estimated by measuring the optical density at 260 nm (OD 26O) and the ratio OD260/0D280. The ratio OD260/0D280 should be between 1.6-2.0. Accurate quantification ensures that standard amounts of RNA can be added to the reverse transcription reaction, thus ensuring reproducibility.

2.3. Reverse Transcription and peR Extensive trials have been carried out to optimize RT-PCR amplification of ISAV for diagnostic purposes. The main findings we are aware of are presented below. Care must be taken when extrapolating from data such as this in order to set up RTPCR detection of ISAV in other laboratories, as different conditions, particularly those of thermocycler temperatures, time and ramp rates, will affect results. Some empirical testing will still be required to determine optimum conditions for each laboratory's individual equipment and reagents. Nevertheless, the information presented below should still prove useful when establishing PCR detection of ISAV.

2.3.1. Two-step reaction In most of the literature cited, a two-step RT-PCR is applied (Mjaaland et aI., 1997, Rimstad et aI., 1999, Blake et aI., 1999, Opitz et aI., 2000, Bouchard et aI., 1999). The amount of RNA used in cDNA synthesis varies. The effects of different concentrations of RNA (1, 1.5, 2, 2.5, 3 and 5 flg) on amplification were tested (Mikalsen et al. in press). Least amplification was found with I flg RNA and somewhat stronger than the average at 5 flg. No differences were achieved for the other RNA amounts tested. No extra products appeared on the gel after the PCR for any amount of RNA added to the RT reaction. As a standard, between 1.5 and 2.0 flg RNA is recommended (Mjaaland et aI., 1997, Rimstad et aI., 1999, DevoId et aI., 2000). As a potential improvement of the RT-reaction, the RNA was treated with DNase in order to remove genomic DNA. This was not found to be successful and in any case, genomic DNA is not considered to be a problem. No specific annealing of ISA virus primers to genomic DNA has been observed, as non-specific products are seldom seen on the agarose gel (Mikalsen et al. in press). If non-specific amplification is observed, confirmation that one of these originates from amplification ofISA V RNA can be obtained through further tests (see below). The optimal temperature and reverse transcriptase enzyme used for the RTreaction was tested. When the RT-reaction was run at 42°C instead of 37°C, higher sensitivity was achieved, giving 5xlO·5dilutions of standard positive tissue (see below). Further, using AMV reverse transcriptase instead of M-MLV reverse transcriptase at 42°C did not lead to any improvements, rather the opposite (Mikalsen et al. in press). The RT-reaction is run for 1 hour followed by a PCR. This can be further optimized by applying nested PCR (Opitz et aI., 2000). Regarding the PCR, several steps have been tested in the Oslo laboratory in order to improve sensitivity using the primer set from Mjaaland et al (1997). Magnesium chloride concentrations between 0.5-3.0 mM were used. Minor differences in the amount of amplicon were found, but non-specific bands were observed for all

ISAV

13

concentrations except 1.5 mM MgCh. However, this can be expected to vary with the different primer sets that are used. Different cycling conditions were also tested. The results improved by increasing the denaturing temperature from 94°C to 95°C and this should be established as standard. By increasing the annealing temperature from 55 to 60°C no improvement was observed and in fact the sensitivity decreased. The number of cycles was varied in the range 30-40 using a weak positive sample. No product was observed at 30 cycles, the amount was as expected at 35 cycles and somewhat stronger at 40 cycles. When in doubt of the result from a sample when running 35 cycles, 40 cycles can be tried. By end point dilution of RNA no differences were found between 35 and 40 cycles (unpublished results from the Oslo laboratory). These parameters must of course be optimized according to the primer set used.

2.3.2. One-step RT-PCR The one-step RT-PCR procedure seems to provide increased sensitivity compared to the two-step RT-PCR (Sellner and Turbett, 1998, Melville and Griffiths, 1999, Devoid et aI., 2000, Lovely et aI., 1999, Kibenge et aI., 2000a, Mikalsen et al. in press). There are several commercial kits available (Table 3). Of those tested in the Oslo laboratory, the Ready-To-Go RT-PCR Beads (Pharmacia) recommended by Griffiths, with a mixture of M-MLV Reverse Transcriptase (first strand synthesis) and Taq polymerase, RNase inhibitor, buffer and dNTPs, stabilised in beads with a cDNA-synthesis at 42°C, gave the highest sensitivity, as positive result was achieved at 5xlO-6dilution of standard positive tissue sample control (see below). The same was recorded by DevoId et al. (2000). It is easy to use and the results are consistent, and is now the procedure established as the standard in the Oslo laboratory. Several steps were taken in order to improve this procedure: by varying the amount of hexamers in the range from 0.2-2 f.!g, minor, but not significant differences were observed, and 0.2 f.!g is used as a standard. Denaturation of RNA at 55-60T for 10 min before the RT-PCR run made the test less sensitive compared to standard methods. With the Titan™ One-tube RT-PCR (Boehringer Mannheim) (Kibenge et aI., 2000a) higher sensitivity is achieved than with the former standard two-step RTPCR, giving comparable sensitivity to that achieved using M-MLV enzyme at 42°C (positive 5xlO-5-dilution of positive tissue sample) (Mikalsen et al. in press). After addition of DEPC-water and hexamers to the tube, 2 f.!g of RNA is added (on ice). The tubes are placed in the thermocycler when this has reached 30-35°C, and incubated at 42°C for 30 min. The tubes are placed on ice and 1.5 f.!l of each primer (15 pmollf.!1) is added. The tubes are replaced in the thermocycler and incubated at 95°C for 5 min, followed by 35 cycles of 95°C for 30 sec, 55T for 15 sec, 72°C for 30 sec. Finally an elongation step of 72°C for 7 min is added. A disadvantage of one-step RT-PCR is that no separate aliquot of cDNA is available. Therefore, if several analyses are to be undertaken using the same RNA, it is often more efficient and convenient to use a two-step method and use the cDNA for multiple amplification reactions. However, if a large number of samples are to be

14

MJAALAND. RIMSTAD & CUNNINGHAM

tested using a single primer set, one-step methods offer significant savings in time and resources.

Mixture of AMV Reverse Transcriptase (first strand cDNA synthesis) and Expand High Fidelity (Taq DNA polymerase and Pwo DNA polymerase). cDNA synthesis at 50-60 C should avoid problems caused by secondary structure of the RNA and/or mispriming (Opitz et aI., 2000). EZ rTth one-step RT-PCR Used by Melville & Griffiths (1999). system (Perkin Elmer) Enhanced Avian RT-PCR AMV™ reverse transcriptase reaction at 65 C kit claimed to yield more full-length cDNA transcripts. (Sigma) SuperScript One-Step Combination of SuperScript ™ reverse transcriptase RT-PCR System (Gibco) and Taq polymerase. Only target specific primers are used. AmpliTaq Gold RNA PCR kit (Perkin Elmer) Access RT.PCR System IMixture of M-MLV Reverse Transcriptase and Tfl (Promega) iDNA Polymerase. cDNA synthesis at 50-60 C !should avoid problems caused by secondary Istructure of the RNA and/or mispriming.

Titan ™One Tube RTPCRSystem (Boehringer Mannheim)

D

D

TM

D

Table 3. Other one-tube RT-PCR systems

2.3.3. Visualisation of PCR product The PCR product can be visualised by gel electrophoresis. High resolution has been achieved using 3% NuSieve 3:1 agarose (FMC Bioproducts) in 1 x TAE buffer and electrophoresis at 80 V for 75 min (Oslo laboratory). Alternatively, 2% agarose in 1 x TAE SUbjected to 100 V for 60 min in a 7 x 10 em minigel (10 Vcm- I ) has been found to provide good results (Aberdeen laboratory). Appropriate molecular weight markers should be run alongside the PCR products to ensure accurate identification and size estimation. We have found 50 bp ladders, with markers at 150 and 200 bp, to be suitable for this purpose and these are available from several sources. In addition to the expected product, PCR products of approximately 300 bp have occasionally been found (Cunningham and Snow, 2000). Once they had been cloned and sequenced, it was found that these were dimers of the 155bp fragment, resulting from complementarity of 2 nucleotides at the termini of the primers. These additional products did not impair interpretation of the gel results.

15

ISAY

2.3.4. Choice of primers Different sets of primers have been selected for RT-PCR detection of ISAV (Table 4). Devoid et al. (Devoid et aI., 2000) applied three different pairs derived from segment 8 and one from segment 2 and compared the sensitivity. They concluded that one of the pairs from segment 8 (FA-31RA-3) gave the highest sensitivity (Devoid et aI., 2000). The eighth segment of ISAV is conserved enough for the FA-31RA-3 primer set to be used also for diagnosis of ISA in Atlantic salmon from Scotland and the East Coast of Canada (Lovely et aI., 1999, Melville and Griffiths, 1999, Kibenge et aI., 2000a). Others have also selected primers from segment 8 (Blake et aI., 1999, Bouchard et aI., 1999, Mjaaland et aI., 1997, Opitz et al.,2000). Genomic Primer Set!lDent name FA-3 8 RA-3 ＭｾＮ＠

ｾＱＮＡ＠

8 6

Primer sequence

Product

Tm sIZe 1(5' -3') GAAGAGTCAGGATGCCAAGAC --------_..._..._------_._-- 59°C 211 bp GAAGTCGATGAACTGCAGCGA

__. _. ____ ｾｔａｃｇ｟ＡＧｑｍ＠

55°C 157bp GCCAAGTGTAAGTAGCACTCC ATGACTGCACTGACGGACCT ..---.---.-.-------- 50°C 250bp 181\7&250-- f:"::-----ACCTTGTAGATTCCGGACAT ILA2 ISA7F

Reference DevoId et al. (2000) Mjaaland et al. (1997) Inglis unpubl.)

Table 4. Primer sets used for PeR amplification ofISAY

Different primer sets from segment 8, segment 6 and segment 2 were compared (Oslo laboratory). First choice primers are the ones published by Mjaaland et al (1997), also used by others (Lovely et aI., 1999, Rimstad et aI., 1999, Cunningham and Snow, 2000), the second choice is a primer set from genomic segment 6 (accession no. AF 220607 and AJ 276859). The same sensitivity for first and second choice primers was achieved in an end-point dilution test for RNA from ISA virus positive tissue sample. Primers based on sequences from the PB 1 genomic segment gave a lO-fold lower detection level and are therefore not listed as suitable primers for this RT-PCR protocol (unpublished results from the Oslo laboratory). Another set of primers, designed from segment 6, are being assessed for use in multiplex reactions to confIrm the presence of ISAV and analyse a highly variable region of this segment (Inglis, FRS Marine Laboratory, unpublished) (see section 2.3.7.).

2.3.5. Controls Negative controls should include a tissue sample from ISAV negative fIsh or ISAV negative SHK.-I cells running in parallel with the samples to be tested through all steps in the procedure. A negative control solely for the RT-PCR is RNA from ISAV negative fish or SHK.-l cells, while positive control could be a tissue sample from an ISAV positive fIsh or ISAV positive SHK.-I cells. In Scotland, negative

16

MJAALAND, R1MST AD & CUNNINGHAM

controls are included at each of the three steps of RNA extraction, RT and PCR. A positive control solely for the RT-PCR is RNA from ISA V positive fish or SHK-I cells (Mjaaland et aI., 1997, Rimstad et aI., 1999). Care must be taken when using positive controls alongside diagnostic samples. With the use of ISAV positive material there is the risk of contamination between samples during the procedure. In Scotland, it was considered that the consequences of false positive results if such contamination occurred were greater than those of false negative results should the reaction fail in any way. Therefore, no positive control is run alongside clinical samples. The ideal positive control material would produce a PCR product of a different size to that from positive material and yet stiII be amplified using the same primers and conditions as normal. Such a control has recently been developed and is being optimised (Aberdeen laboratory, in prep.) An alternative 'positive' or internal control employs primers that anneal to salmon MHC-I as described by Devoid et al. (2000). This method confirms the integrity of sample RNA and reduces the concern that negative results could be false negatives due to degradation of the sample. Multiplex reactions using both the MHC primers and those of Mjaaland et al. (1997) are undergoing trials in the Aberdeen laboratory and appear promising. If successful, this would provide an efficient method to include a control for RNA integrity with each sample.

2.3.6. Sensitivity To obtain a measurement for the sensitivity of the RT-PCR, a cell culture harvest containing a quantified amount of ISA virus was used. Quantification of the virus was carried out by seeding SHK-I cultures with lO-fold dilutions of the cell culture fluid and subsequently staining with IF AT after I-week incubation at 15°C (Rimstad et aI., 1999, Devoid et aI., 2000). One IFAT unit was termed 1 TCID 50• Each 10-fold dilution of the cell culture fluid was mixed with a tissue suspension from an ISA V negative fish. RNA was extracted, and 2 !1g RNA from each tissue suspension tested by RT-PCR. Both the two-tube RT-PCR and the one-tube ready-to-go RT-PCR (Pharmacia) were tested and sensitivities compared. The detection limit of the onetube RT-PCR was estimated to be 0.08 TCID 50 • There is no standard deviation in this limit, as not enough parallel tests have been run to calculate this. Therefore, for simplicity, the detection limit is concluded to be between 0.1-0.01 TCID50 • A cell culture adapted virus strain was used, and it was added to organ suspension. Under field conditions in which the virus is not cell culture-adapted, the sensitivity could potentially be even further in the favour ofRT-PCR.

2.3.7. Specificity The observation of a single PCR product of the correct size in agarose gels is normally considered sufficient indication of the presence of ISAV. However, in some cases it may be necessary to confirm the identity of the PCR product. This may be the case if multiple fragments are seen on the gel, including one of the size expected for ISA V positive material. Digestion of the PCR product from the method of Mjaaland et al. (1997), with restriction enzyme Sau3A results in two restriction fragments of 84 and 71 bp. This

ISAV

17

method is a rapid and straightforward test to assess the identity of a PCR product. However, the small restriction fragments can be particularly difficult to resolve on an agarose gel, requiring use of acrylamide gel, and if multiple PeR products are present, accurate interpretation of the restriction fragment pattern may be impossible. The greatest possible' detail is obtained from the nucleotide sequence of a PCR product. In the case of a fragment of 155 bp, as produced by the method of Mjaaland et al. (1997), purification and direct sequencing of such a small fragment is problematic and cloning is usually required prior to sequencing. Thus this type of analysis is lengthy, costly and not suitable on a routine basis. A compromise between speed and detail can be achieved through the use of probes. An oligonucleotide probe was designed that will anneal to sequence within the 155 bp PCR product resulting from the method of Mjaaland et al. (1997) (McBeath et al., 2000). PCR products are southern blotted and probed, and positive hybridisation indicates the presence of the specific PCR product. This technique can be particularly useful in cases where mUltiple PCR products are seen and it has been used to confirm the extremely low rate of false positive results obtained with this PeR method. Multiplex PeR is being assessed as a method to confirm specificity of PCR amplification and also to counter some arguments that positive PCR results could indicate the presence of remnants of ISAV and not the intact, viable virus or could be produced from amplification of other, as yet uncharacterised, orthomyxoviruses. A PCR method for amplification of 250 bp of segment 7 has been designed and the primers are being assessed in a multiplex reaction alongside those described by Mjaaland et al. (1997) (Inglis, FRS Marine Laboratory, unpublished). When using RNA derived from ISAV infected cell cultures, the method seemed promising, with no inhibition of either amplification, and the production of two PeR products of different sizes from sample containing ISAV. However, when applied to RNA from experimentally infected fish, inhibition of amplification was observed. Thus it may be better to carry out separate amplifications for each primer set, necessitating the use of two-step RT-PCR protocols. No information is yet available on the possibility of including internal control primers and it is possible that the addition of further primer sets might inhibit some amplifications. Ritchie et al. (2001) report that amplification of a longer product from segment 2 reflects the ability to culture ISAV from disinfectant-treated material whereas PCR using primers FA31RA3 (Devoid et al., 2000) can still achieve amplification from samples which cannot be cultured. This finding reveals the possibility of discriminating 'live' from 'dead' virus and thereby overcoming some of the arguments against use ()f PCR as a diagnostic test, i.e. that PCR products may detect fragments of virus or virus that is non-infectious. However, 0ye & Rimstad (in prep.) report that positive RT-PCR using primers producing amplicons less than 200 bp, was achieved after UVC irradiation of more than 50000 J/m2 while a 3-log reduction of the infectivity of ISAV, suspended in seawater, was achieved by irradiation doses of 51 ± 13 J/m 2 •

18

MJAALAND, RIMST AD & CUNNINGHAM

2.4. Other important considerations in molecular diagnosis of ISAV 2.4.1. Detection of viral RNA shortly after introduction ofISA virus to a tank Several conclusions can be drawn from experimental trials performed in the Oslo laboratory, where fish were either infected as cohabitees of experimentally infected fish or ISAV was added directly to the water: a) There must be a time delay of at least 5 days after the introduction of ISAV infected fish to a salmon farm before RT-peR can be expected to yield conclusive results. b) If the fish in a farm are suspected to have been infected for a reasonable period of time, then hearts and kidneys are the first-choice organs to test. c) If the infection has just been introduced, then the gills should be added to the list of organs to be tested. Obviously, in most cases of samples taken from fish farms, it is impossible to judge when the virus might have been introduced. Nevertheless, this information is helpful for epidemiological investigations and design of experimental infections.

2.4.2. Sampling and transportation of sampled material The influence of sampling and transportation of material on the result of the RTpeR detection of ISAV was evaluated in an experimental trial where fish had been cohabitating with ISAV injected fish. This study concluded that the fITst choice of sampling and transportation is whole fish on ice, express freight. Second choice if the fish is too big to be transported, is sampling of organs in sampling containers on ice and express freight preferably within 24 h. In situations where express transport is not possible or samples must be taken on site, the use of RNAlater is recommended for preservation of material for molecular analyses.

2.4.3. Organisation of diagnostic laboratories The layout and practices in diagnostic laboratories must be carefully considered. Particularly in the case of diseases such as ISAV, where a positive diagnosis can lead to rigorous and harsh control measures such as culling, the implications of inaccurate results are extremely severe. As peR has moved from a research tool to a diagnostic method, an understanding and appreciation of the mechanisms of the reaction and its' risks are even more important for practitioners and when designing facilities used for such testing. The risk of false positive and negative results can be reduced to some extent by the use of confirmatory methods such as probes or internal controls, as discussed in section 2.3.5. and 2.3.7. One of the greatest fears in molecular diagnostics is that contamination may lead to false positive results, with resultant hardship for farmers. The chance of contamination can be minimised by good laboratory organisation and practice. This includes the use of strict aseptic techniques, separation of pre- and post-peR operations, mixing of RT and peR reagents in a 'clean room' where no RNA or DNA is allowed, and very careful

ISAY

19

attention to separation of procedures and equipment, ensuring a one-way transit from RNA extraction, through RT-PCR to gel electrophoresis, with no movement of materials in the opposite direction. In the unfortunate event of contamination occurring, the use of negative controls at every step in testing will reveal both the presence of contamination and the likely point at which it occurred.

2.4. Evaluation of different diagnostic methods Although desirable, it is not always possible to employ all diagnostic methods concurrently. While surveillance programs aim to provide a first-line defence by detecting ISAV infection before a full-blown disease outbreak occurs, a primary ISA diagnosis should ideally include epizootiological, clinical and pathological observations in combination with confirmatory ISAV tests. In a recent study, the efficacy of current diagnostic methods in detecting infectious salmon anaemia (ISA) or ISAV in experimentally infected Atlantic salmon were estimated (Opitz et aI., 2000). The diagnostic methods used were clinical signs of disease, gross pathological and histological observations, indirect fluorescent antibody test (IFAT), virus isolation on SHK-l cells and CHSE cells, and RT-PCR. Since no fish showed typical clinical signs, or pathological lesions associated with ISA, the conclusions were drawn from the results from IFAT, virus isolation and RTPCR. All IFAT positive samples were also positive by virus isolation and RT-PCR, and all fish positive by virus isolation were also positive by RT-PCR. 18 fish found negative by IFAT were positive by RT-PCR; 13 fish negative by virology were positive by RT-PCR; and 10 fish negative by both IFAT and virology were positive by RT-PCR, indicating the sensitivity of the different tests.

2.5. Molecular diagnosis of ISAV - future work The existence of a wide range of virulence within ISAV may explain differences in clinical ISA outbreaks (the disease resistance of the host is of course also of pivotal importance). This may imply that hitherto non-pathogenic orthomyxovirus infection in salmon may exist, resembling the way ISAV infections occur in rainbow and brown trout. If so, the differentiation between pathogenic and non-pathogenic forms of salmon orthomyxovirus would be in great demand. However, such a task would require knowledge of the viral factors that influence the virulence, but currently, this basal knowledge is missing. Studies to elucidate this type of information are largescale and require intensive work, as was the case with investigations of viral haemorrhagic septicaemia virus (VHSV) (Raynard et aI., 1999). The use of quantitative measurement of ISAV could potentially be interesting for special situations, for instance when several pathogens are involved in disease outbreaks and the etiological role played by the individual agents is unknown. Another situation for using quantitative measurement is when testing the effects of different disinfectants. The most likely method to be established for quantitative analysis would be quantitative RT-PCR, like Taqman (Applied Biosystems) and

20

MJAALAND, RlMSTAD & CUNNINGHAM

related procedures. Virus isolation could also be used quantitatively, but is too timeconsuming and cost intensive to be widely used for this purpose. It has been shown that Atlantic salmon surviving ISA outbreaks do have ISAV specific antibodies. This could enable the use of serology for detection of the ISAV infection status of fish populations, and consequently serological methods for detection of ISAV antibodies are being developed (O.B. Dale, K. Falk, National Veterinary Institute, Oslo; R. S. Raynard, FRS Marine Laboratory, Aberdeen, pers. comm.). However, the methodology for detection of specific immune responses in fish is still in its infancy. Much basic knowledge regarding fish immune responses and the ecology of ISAV will be required prior to the general use of serology in surveillance of ISAV induced diseases and it is expected that molecular methods will remain extremely important in diagnosis of this disease.

3. REFERENCES Blake, S., Bouchard, D., Keleher, W., Opitz, M., and Nicholson, B.L. (1999). Genomic relationships of the North American isolate of infectious salmon anemia virus (ISAV) to the Norwegian strain of ISAV. Dis Aquat Org 35,139-144. Bouchard, D., Keleher, W., Opitz, H.M., Blake, S., Edwards, K.C., and Nicholson, B.L. (1999). Isolation of infectious salmon anemia virus (ISAV) from Atlantic salmon in New Brunswick, Canada. Dis Aquat Org 35, 131-137. Bouchard, D. A, Brockway, K., Giray, C., Keleher, W. & Merrill, P. L. (2001) First report of infectious salmon anemia (ISA) in the United States. Bull Eur Ass Fish Patho121: 86-88. Byrne, P.I., MacPhee, D.D., Ostland, V.E., Johnson, G., and Ferguson, H.W. (1998). Haemorrhagic kidney syndrome of Atlantic salmon, Salmo solar L. J Fish Dis 21, 81-91. Cherian, T., Bobo, L., Steinhort; M.C., Karron, R.A., and Yolken, R.H. (1994). Use of PCR-enzyme immunoassay for identification of influenza A virus matrix RNA in clinical samples negative for cultivable virus. J Clin Microbiol 32, 623-628. Cunningham, C.O. and Snow, M. (2000). Genetic analysis of infectious salmon anaemia virus (ISAV) from Scotland. Dis Aquat Org 41, 1-8. Dannevig, B.H., Brudeseth, B.E., Gjl!Jen, T., Rode, M., Wergeland, H.I., Evensen, 0., and Press, C.M. (1997). Characterisation ofa long-term cell line (SHK-l) developed from head kidney of Atlantic salmon (Salmo solar L.). Fish Shellfish Immunol7, 213-226. Dannevig, B.H., Falk, K., and Namork, E. (1995a). Isolation of the causal virus of infectious salmon anaemia (lSA) in a long-term cell line from Atlantic salmon head kidney. J Gen Virol 76, 13531359. Dannevig, B.H., Falk, K., and Press, C.M. (1995b). Propagation of infectious salmon anaemia (ISA) virus in cell culture. Vet Res 26, 438-442. Dannevig, B.H., Falk, K., and Skjerve, E. (1994). Infectivity of internal tissues of Atlantic salmon, Salmo salar L., experimentally infected with the aetiological agent of infectious salmon anaemia (ISA). J Fish Dis 17,613-622. Devoid, M., Krossey, B., Aspehaug, V., and Nylund, A (2000). Use ofRT-PCR for diagnosis of infectious salmon anaemia virus (ISAV) in carrier sea trout Salmo frulla after experimental infection. Dis Aquat Org 40, 9-18. Donofrio, J.C., Coonrod, J.D., and Chambers, T.M. (1994). Diagnosis of equine influenza by the polymerase chain reaction. J Vet Diag Invest 6, 3943. Evensen, 0., Thorud, K.E., and Olsen, Y. (1991). A morphological study of the gross and light microscopic lesions of infectious salmon anaemia in Atlantic salmon (Salmo solar). ResVet Sci 51, 215-222. Falk, K. and Dannevig, B.H. (1995). Demonstration of infectious salmon anaemia (lSA) viral! antigens in cell cultures and tissue sections. Vet Res 26, 499-504. Falk, K., Namork, E., and Dannevig, B.H. (1998). Characterization and applications of a monoclonal antibody against infectious salmon anaemia virus. Dis Aquat Org 34, 77-85. Falk, K., Namork, E., Rimstad, E., Mjaaland, S., and Dannevig, B.H. (1997). Characterisation of infectious salmon anemia virus, an orthomyxo-like virus isolated from Atlantic salmon (Salmon salar L.). J Virol 71, 9016-9023. Fodor, E., Pritlove, D.C., and Brownlee, G.G. (1994). The influenza virus panhandle is involved in the initiation of transcription. J Virol68, 4092-4096.

ISAV

21

Hastings, T., Olivier, G., Cusack, R., Bricknell, I., Nylund, A, Binde, M., Munro, P., and Allan, C. (1999). Infectious salmon anaemia. Bull Eur Ass Fish Pathol19, 286-288. Hovland, T., Nylund, A, Watanabe, K., and Endresen, C. (1994). Observation of infectious salmon anaemia virus in Atlantic salmon, Sa/mo sa/ar L. 10urnal ofFish Diseases 17,291-296. Hastein, T. (1997) Infectious salmon anaemia (ISA). A historical and epidemiological review of the development and spread of the disease in Norwegian fish farms. In: Workshop on infectious salmon anaemia, 26 November 1997, St. Andrews, New Brunswick. Ed.: Hastein, T. Oslo, National Veterinary Institute. ppl-92. Inglis, lA, Bruce, J., and Cunningham, C.O. (2000). Nucleotide sequence variation in isolates of infectious salmon anaemia virus (I SAV) from Scotland and Norway. Dis Aquat Org 43,71-76. Jarp, 1., and Karlsen, E. (1997). Infectious salmon anaemia (ISA) risk factors in sea-cultured Atlantic salmon Salmo salar. Dis Aquat Org 28, 79-86. Kibenge, F.S., Lyaku, J.R., Rainnie, D., and Hanunell, K.L. (20ooa). Growth of infectious salmon anaemia virus in CHSE-214 cells and evidence for phenotypic differences between virus strains. J Gen Virol81, 143-150. Kibenge, F. S. B., Whyte, S. K., Hammell, K. L., Rainnie, D., Kibenge, M. T., and Martin, C. K. (20oob) A dual infection of infectious salmon anaemia (ISA) virus and a togavirus-like virus in ISA of Atlantic salmon Salmo salar in New Brunswick, Canada. Dis Aquat Org 42,11-15. Kibenge, F. S. B., Garate, O. N., Johnson, G., Arriagada, R., Kibenge, M. 1. T. & Wadowska, D. (2001) Isolation and identification of infectious salmon anaemia virus (ISAV) from Coho salmon in Chile. Dis Aquat Org 45: 9-18. Krossoy, B., Hordvik, I., Nilsen, F., Nylund, A, and Endresen, C. (1999). The putative polymerase sequence of infectious salmon anemia virus suggests a new genus within the Orthomyxoviridae. J Virol 73, 2136-2142. Lovely, I.E., Dannevig, B.H., Falk, K., Hutchin, L., MacKiunon, AM., Melville, K.1., Rimstad, E., and Griffiths, S.G. (1999). First identification of infectious salmon anaemia virus in North America with haemorrhagic kidney syndrome. Dis Aquat Org 35, 145-148. McBeath, AJ.A, Burr, K.L-A, and Cunningham, C.O. (2000). Development and use of a DNA probe for confirmation of cDNA from infectious salmon anaemia virus (ISAV) in PCR products. Bull Eur Ass Fish Pathol20, 130-134. Melville, K.1., and Griffiths, S.G. (1999). Absence of vertical transmission of infectious salmon anemia virus (ISAV) from individually infected Atlantic salmon Salmo salar. Dis Aquat Org 38, 231-234. Mikalsen, AB, Teig, A, Heileman, A-L., Mjaaland, S., and Rimstad, E. (in press) Detection of infectious salmon anaemia virus (lSA V), by RT-PCR after cohabitant exposure in Atlantic salmon (Salmo safar L.) Dis Aquat Org. Mjaaland, S., Rimstad, E., Falk, K., and Dannevig, B.H. (1997). Genome characterization of the virus causing infectious salmon anemia in Atlantic salmon (Salmo salar L.): an orthomyxo-Iike virus in a teleost. 1 Virol 71, 7681-7686. Mullins, lE., Groman, D., and Wadowska, D. (1998). Infectious salmon anaemia in salt water Atlantic salmon (Salmo salar L.) in New Brunswick, Canada. Bull Eur Ass Fish Pathol18, 110-114. Nylund, A, and lakobsen, P. (1995). Sea trout as a carrier of infectious salmon anaemia virus. J Fish Bioi 47,174-176. Nylund, A., Kvenseth, AM., and Krossoy, B. (1995a). Susceptibility of wild salmon (Salmo salar L.) to infectious salmon anaemia (ISA). Bull Eur Ass Fish Pathol15, 152-156. Nylund, A, Hovland, T., Watanabe, K., and Endresen, C. (1995b). Presence of infectious salmon anaemia virus (ISAV) in tissues of Atlantic salmon, Salmo salar L., collected during three separate outbreaks of the disease. 1 Fish Dis 18, 135-145. Nylund, A, Alexandersen, S., Rolland, 1.B., and Jakobsen, P. (1995c). Infectious salmon anemia virus (ISAV) in brown trout. J Aquat Animal Health 7, 236-240. Nylund, A, Kvenseth, AM., Krossoy, B., and Hodneland, K. (1997). Replication of the infectious salmon anaemia virus (ISAV) in rainbow trout, Oncorhynchus mykiss (Walbanm). 1 Fish Dis 20, 275-279. O'Halioran, J.L., L'Aventure, J.P., Groman, D.B., and Reid, AM. (1999). Infectious salmon anemia in Atlantic salmon. Can Vet J. 40, 351-352. Office International des Epizooties (OlE) (l997a). Diagnostic Manual for Aquatic Animal Diseases. (Paris: OlE), p. 251. Office International des Epizooties (OlE) (l997b). International Aquatic Animal Health Code. (Paris: OlE), p.192.

22

MJAALAND, RIMSTAD & CUNNINGHAM

Opitz, H.M., Bouchard, D., Anderson, E., Blake, S., Nicholson, B.L., and Keleher, W. (2000). A comparison of methods for the detection of experimentally induced subclinical infectious salmon anaemia in Atlantic salmon. Bull Eur Ass Fish Pathol 20, 12-22. Raynard, R S. (2000) A survey of wild fish in Scotland for evidence of infectious salmon anaemia virusreport for the period May 1998 to December 1999. FRS Marine Laboratory, Aberdeen. Report 05/00. 17pp. Raynard, R S., King, J. A., Snow, M., Munro, A L. S., Olesen, N. 1., and Mortensen, H. F. (1999) Virulence of marine isolates of viral haemorrhagic septicaemia virus in Atlantic salmon and turbot. EAFP 9th International Conference "Diseases of Fish and Shellfish", Rhodes, Greece 1924th September 1999,0-015. Raynard, R. S., Murray, A G., and Gregory, A (in press). Infectious salmon anaemia virus in wild fish from Scotland. Dis Aquat Org Rimstad, E., Falk, K., Mikalsen, AB., and Teig, A (1999). Time course tissue distribution of infectious salmon anaemia virus in experimentally infected Atlantic salmon Safrno safar. Dis Aquat Org 36, 107-112. Rimstad, E., Mjaaland, S., Snow, M., Mikalsen, AB., and Cunningham, C.O. (2001). Characterisation of the infectious salmon anemia virus genomic segment that encodes the putative hemagglutinin. J Virol 75, 5352-5356. Ritchie, R. J., Bardiot, A, Melville, K., Griffiths, S., Cunningham, C. 0., and Snow, M. (in prep.) Characterisation of the genomic segment encoding the putative matrix proteins of Infectious Salmon Anaemia Virus (ISAV). Ritchie, R J., Cook, M., Mellville, K., Simard, N., Cusack, R., and Griffiths, S. (2001a) Identification of infectious salmon anaemia virus in Atlantic salmon from Nova Scotia (Canada): evidence for functional strain differences. Dis Aquat Org 44: 171-178. Ritchie, R 1., Heppell, J., Cook, M. B., Jones, S., and Griffiths, S. G. (2001b) Identification and characterization of segments 3 and 4 of the ISAV genome. Virus Genes 22: 289-297. Rodger, H.D., Turnbull, T., Muir, F., Millar,S., and Richards, RH. (1998). Infectious salmon anaemia (lSA) in the United Kingdom. Bull Eur Ass Fish Pathol 18, 115-116. Rolland, 1.B., and Nylund, A (I 998b). Infectiousness of the organic materials originating in ISA-infected fish and transmission of the disease via salmon lice (Lepeophtheirus safrnonis). Bull Eur Ass Fish PathoI18,173-180. Rolland, 1.B., and Nylund, A (1998a). Sea running brown trout: carrier and transmitter of the infectious salmon anemia virus (ISAV). Bull Eur Ass Fish Pathol 18, 50-55. Sandvik, T., Rimstad, E., and Mjaaland, S. (2000). The viral RNA 3'- and 5'-end structure and mRNA transcription of infectious salmon anaemia virus resemble those of influenza viruses. Arch Virol 145, 1659-1669. Sellner, L.N., and Turbett, G.R (1998). Comparison of three RT-PCR methods. Biotech 25, 230-234. Snow, M., and Cunningham, C. O. (2001) Characterisation of the putative nucleoprotein gene of infectious salmon anaemia virus (lSAV). Virus Res 74,111-118. Sommer, AI., and Mennen, S. (1997). Multiplication and haemadsorbing activity of infectious salmon anaemia virus in the established Atlantic salmon cell line. J Gen Viro178, 1891-1895. Stagg, R, Bruno, D., Cunningham, C., Hastings, T., and Bricknell, I. (1999). Focus on infectious salmon anaemia: Epizootiology and pathology. State Vet J 9,1-5. Stagg, R. M., Bruno, D. W., Cunningham, C. 0., Raynard, R. S., Munro, P. D., Murray, A G., Allan, C. E. T., Smail, D. A, McVicar, A H., and Hastings, T. S. (2001). Epizootiological investigations into an outbreak of infectious salmon anaemia (lSA) in Scotland. FRS Marine Laboratory, Aberdeen Report 13/01: 6Opp. Thorud, K. E. (1991) Infectious salmon anaemia. Transmission trials. Haernatological, clinical, chemical and morphological investigations. Norwegian College of Veterinary Medicine, PhD thesis. Thorud, K.E., and Djupvik, H.O. (1988). Infectious anaemia in Atlantic salmon (Safrno safar L.). Bull Eur Ass Fish Pathol 8, 109-111. Thorud, K.E., and Torgersen, Y. (1994). Regnbueerret og sjoorretkan vrere passive smittebrerere av ILA Norsk Fiskeoppdrett llA, 54-55 (in Norwegian). Totland, G.K., Hjeltnes, B.K., and Flood, P.R (1996). Transmission of infectious salmon anaemia (lSA) through natural secretions and excretions from infected smolts of Atlantic salmon Safrno safar during their presymptomatic phase. Dis Aquat Org 26, 25-31. Vllgsholm, I., Djupvik, H.O., Willumsen, F.V., Tveit, AM., and Tangen, K. (1994). Infectious salmon anaemia (lSA) epidemiology in Norway. Prev Vet Med 19,277-290. Wergeland, H. I., and Jakobsen, R. A (200 1) A salmonid cell line (TO) for production of infectious salmon anaemia virus (ISAV). Dis Aquat Org 44: 183-190. 0ye, A-I(, and Rimstad, E. (In prep.) Inactivation of infectious salmon anaemia virus, viral haemorrhagic septicaemia virus and infectious pancreatic necrosis virus in water using UVC irradiation.

Chapter 2 DIAGNOSIS AND IDENTIFICATION OF IPNV IN SALMONIDS BY MOLECULAR METHODS

Carlos P. Dopazo 1 and Juan L. Barja2 Departamento de Microbioloxia e Parasitoloxia, Instituto de Acuicultural and Facultade de Bioloxid. Universidade de Santiago de Compostela, Spain.

1. ABSTRACT Fifty years ago, Infectious Pancreatic Necrosis Virus (IPNV) was the first virus to be isolated from fish, and hence the origin of the development offish virology. After the initial studies in which farmed salmonids were considered the unique host, the virus was also found in a large number of species of freshwater and marine fishes as well as in crustaceans and molluscs. The economic impact of the disease and the ubiquity of the agent justifies the extensive literature related to the IPN virus including numerous reviews. At present, the complete genomic sequence of the dsRNA has been published for four different strains belonging to three of the serotypes described. Based on these findings, a number of molecular procedures have been developed not only to characterize the virus but also to improve the diagnostic methods, trying to circumvent the necessity for cell cultures in order to detect IPNV directly in infected fish and ova. The present review is focused on the molecular procedures used in the study of IPNV, using methods such as analysis of electropherotypes, RFLPs, nucleic acid hybridization using specific probes, and different RT-PCR protocols. The distinct procedures for extraction of viral RNA, selection of restriction enzymes, sets of primers or probes, as well as the conditions (concentration of virus, minimal time required) to obtain reliable results are examined.

2. INTRODUCTION 2.1. Physicochemical Properties of the Virus. The IPN virus belongs to the genus Aquabirnavirus, within the Family Birnaviridae (van Regenmortel et al. 2000), that includes two other genera (Avibirnavirus and Entomobirnavirus) from birds and insects, respectively. Two 23

c.o. Cunningham (ed.), Molecular Diagnosis (!!'Sallllonid Diseases, 23·48. © 2002 Kluwer Arademic Publishers.

24

OOPAZO & BARJA

species are now officially recognized: Infectious pancreatic necrosis virus (IPNV) (the type strain) and Yellowtail ascites virus. The aquabirnaviruses form a large and antigenically diverse group of viruses. They have been isolated from a variety of aquatic animals, and are frequently not associated with disease. Based on results from cross-neutralization assays, they have been divided in two serogroups (Hill and Way, 1995). Serogroup A contains most of the isolates and those associated with disease in salmonid fish. They have been grouped into nine serotypes. Serogroup B consists of a single serotype (B I) with a smaller number of strains. Apart from these groups, some isolates exist that are nontypable. The proposed serotype nomenclature in Serogroup A is as follows: Al (archetype: WB); A2 (Sp); A3 (Ab); A4 (He); A5 (Te); A6 (C I); A7 (C 2); A8 (C 3); A9 (Jasper). A tenth serotype, AIO (NI), was suggested by some authors (Christie et al., 1988), although it was later demonstrated to be related to the Sp serotype (Melby and Christie, 1994). The IPN virions are approximately 60 nm in diameter, icosahedric shape, nonenveloped and single-shelled with a sedimentation coefficient of 435 S; a buoyant density in CsCI of l.33g1cm3 , and a molecular weight of 55 x 106 daltons, 3.8 x 106 of which correspond to the genome, representing 7% of the mass of the virion. The genome consists of two segments of 14S, double-stranded RNA, resistant to RNase. Segment A is 2.5 X 106 (1.95-2.6) daltons (i.e. 2962-3104 nucleotides) and segment B 2.3 x 106 (1.72-2.3) daltons (2731-2784 nt). The GC content is 54% and the denaturation temperature (Tm) 89°C. The viral RNA is covalently linked, at the 5'end, to a protein (VPg) identical to VPI, which is present in the virion in both free and genome-linked forms. There are no poly(A) tracts at the 3'ends of the RNA segments. (Dobos, 1995) Segment A contains two open reading frames (ORFs) (Figure 1). ORFI encodes a 17 kDa arginine-rich protein and ORF2 encodes a 106 Kd polyprotein which contains (5'to 3') the pre-VP2, NS protease and VP3 polypeptides. These are cleaved by the virus-encoded protease to generate the VP2 and VP3. VP2 represents the dominant (62%) virion protein and is responsible for the production of type-specific neutralizing monoclonal antibodies. The mRNA from segment B is translated to a 94 kDa polypeptide which is the viral RNA- dependent RNA polymerase (VP1). A single cycle of replication takes approximately 18-22 hours, and mRNAs can be detected in infected cells 3-4 hours post infection. The nucleotide sequence of the genome of four different IPNV strains from the serogroup A has been determined: The Sp strain (serotype A2), the Jasper strain (Ja: serotype A9), the Korean strain DRT, but only segment A of the Nl strain I(AIO) (Duncan and Dobos, 1986; HAvarstein et al., 1990; Duncan et al., 1991; Mason, 1992; Chung et al., 1993, 1994). Partial sequences of some other strains can also be found in the Genbank database. This sequence information allowed the construction of the genomic map of IPNV and hence, the selection of primers and probes to develop distinct molecular procedures for the detection and the specific identification of the virus in cell culture and sample tissues.

25

IPNV

.& gmmtA

,sf i

" I,

·h

J

IIXD

---l" 3--."

Rimstad el aI,. 1990 McAllisteret aI .• 1991 ｌＶｰ･ｾｬａＧｴｲ｡＠

VI - - - - - V l

m

et aI., 1994

V

1--0" PI_O"

Pr AI - - - - - - - - - Pr II) PrBI _ _ _ _ PrBl

Blake et ai" 1995

PrDI-PrD2 1IV 1 - -_ _ _ Jw2

OlveU'll et at, 1997. 1998 PrDI _ _ PrD2

Wangetal,,1997 Pr Fl - - - - Pr F)

Willianlset aI., 1999

WEI -

WE2

Cralle et 61., 2000

T,I>1 - - - T>bl

VPI

,

i4i&h¥WW%MM&W,&&&RP'¥ZWM"Wf

V

v¥¥?&5&4Ah&

IOIJ)

Olveil'll et al, 1998

')t

¥b t ' V+LdxL,J4.&M

'%"

10lJ)

A4R - - A4U

Figure J, Position of primer pairs designed for diagnosis of salmonid and non-salmonid IPN viruses by RT-PCR,

* second amplification product in nested PCR.

2.2. The Disease Infectious Pancreatic Necrosis (lPN) is a well-characterized acute disease affecting mainly young hatchery-reared salmon ids. The disease was first described in 1940 in Canada by M'Gonigle (M'Gonigle, 1941), but it was not until 1955 that Wood, Snieszko and Yasutake revealed the infectious nature of the disease and the

26

OOPAZO& BARJA

possible viral origin (Wood et aI., 1955). This was confinned by Wolf, who in 1957 isolated the first strain of IPNV from cultured brook trout (Salvelinus jontinalis), and constituted the first fish virus to be grown in vitro (Wolf, 1988). The strain deposited in the ATCC as the prototype was isolated years after, from diseased rainbow trout fingerlings and received the reference VR-299, today known as West Buxton (WB). lPNV is the most extensively studied fish virus and has been the subject of numerous reviews (Hill, 1982; Wolf, 1988; Bonami and Adams, 1991; Reno, 1999) The new virus was very soon found to cause disease in a variety of salmon ids including members of the genera Salmo, Oncorhynchus and Salvelinus. In 1972, Sonstergard et al. (l972b) reported the isolation of a virus from healthy white suckers (Catostomus commersom) in Canada. This was the first evidence for IPNV infection in a non-salmonid fish species. Since then, numerous viruses which have been classified as IPNV have been isolated from all over the world and in many different species of fish in freshwater, marine and brackish environments, as well as from marine molluscs, crustaceans and other invertebrates (see reviews by Wolf, 1988; Reno, 1999). In Europe, the first known outbreak of IPN occurred in rainbow trout (Oncorhynchus mykiss) in France in 1964 and later, in Denmark, the Sp and Ab strains were identified (Hill & Way, 1995). In Asia, the first isolates ofIPNV were obtained by Sano in Japan in 1971and in 1987 IPNV was isolated in New Zealand, from chinook salmon (0. tshawytscha) returning from the sea, but originating from a salmon stock imported from Japan (Tisdall and Phipps, 1987). Recently, Crane et al. (2000) isolated the virus in Tasmania from farmed Atlantic salmon (Salmo salar) and rainbow trout as well as from different species of wild marine fish. No gross clinical signs were observed but lesions in the pancreatic tissue were revealed in the histopathological study of the salmon. In salmonid fish, the disease causes an acute gastroenteritis and destruction of the pancreas (focal necrosis) in the early stages of life. After that time, around four to six months old, the infected fish generally suffer inapparent infection and have no significant mortalities (Dorson and Torchy, 1981; McAllister, 1988; Reno, 1999). The gross clinical signs in fry are dark pigmentation of the skin, exophthalmia, abdominal swelling, and pale gills. Haemorrhages occur in the caecal mass and mucous material is found in the stomach and intestine. Spiral swimming is frequent. IPN virus has also been associated with cranial haemorrhage and ascites in cultured yellowtail (Seriola quinqeradiata), with nephritis in Japanese eels (Anguilla japonica), and with haematopoietic necrosis in turbot (Scophthalmus maximus) (Sano et aI., 1981; Castric et al., 1987; Reno, 1999). Although lPNV has been isolated from Atlantic salmon, the pathogenic significance of these infections is not always clear (Christie et al. 1988). However, in the last decade, significant mortalities attributed to IPN have occured in the postsmolt period during the first months after transfer to the sea, and constitute a problem for Nordic fish farms. Vagsholm and Djupvik recorded in Norway a cumulative mortality of up to 58% in 1987-88 and 80% in 1991, being very variable among groups, as reported by Jarp et af. (1994). Smail et af. (1992) reported the same occurrence in Scotland since 1987, and suggested that lPNV may be transmitted to the post-smolts from some marine hosts. These results gave rise to the research on a feasible vaccine against IPNV for salmon.

IPNV

27

In salmonids, IPN virus isolates belonging to serotype Sp are noted to be virulent whereas those from the Ab serotype are avirulent. Hill (1982) found the isolates obtained from marine fish (Sp serotype) to be avirulent for rainbow trout although infectious. Similar results were obtained by McAllister et. at (1984) for the Ab isolates from flounder. Regarding the incubation period, the time required for the onset of disease varies with a number of circumstances including temperature, age of fish, species, physiological conditions of fish, and concentration of virus (Dorson and Torchy, 1981). Fish exposed to IPNV by immersion start to die at day 6 to 20, whereas injected fish die after 3 to 10 days. (McAllister and Owens, 1986; Wolf, 1988). The incubation periods are short at water temperatures between 10 and 16°C and mortalities are greatest at 10-13"C, being very low at 4-6°C or above 16"C. Thus an epizootic may persist for up to 4 months at 4-6"C but the onset lasts only 3-6 weeks at 10-16"C. Aquabimaviruses have a high level of resistance to inactivation (Desautels and Mackelvie, 1975; Barja et al. 1983). From the critical interpretation of the appearance of a number of outbreaks, it has become evident that the pathogen is commonly transmitted horizontally by movement of infected fish, and via urine and faeces (Wolf 1988). During epizootic conditions, the hatchery water can contain a concentration of virus of approximately 102 TCID501ml. The virus probably enters into the host by contact with gills, the lateral line, and by ingestion. Vertical transmission was demonstrated by Wolf et al. (1963) from carrier fish at spawning. Moreover, the virus was isolated even from iodophor treated eggs and embryos (Bullock et al 1976). A large proportion of survivors of an epizootic remain as carriers for years (persistently infected), and can become a source of virus spread (see review by Reno, 1999). It has also been shown that poikilotherms, homoiotherms, and fomites can serve as vectors of epidemiological significance. Thus, parasites (Moewus-Kobb, 1965) and piscivorous birds may play a role in the spreading of the virus (Sonstergard and McDermott, 1972a; Peters and Neukirch, 1986; McAllister and Owens, 1993). Improperly disinfected material such as tanks and nets are of particular importance.

3. DETECTION OF IPN VIRUSES Three reasons for diagnosis have been established (Desselberger, 1995): i) when treatment and handling depend on diagnosis, ii) when confirmation of the etiology of a disease provides control tools, and iii) for epidemiological studies. Although this statement has been settled for medical virology, the three reasons clearly apply to fish viruses and hence for IPNV and IPNV-like viruses. First, if a viral disease affects a certain fish stock, unnecessary antibiotic treatment can be avoided if rapidly diagnosed, and in addition fish should be handled with care in order to avoid contamination of healthy stocks. Secondly, if the virus is quickly diagnosed, control measures to avoid dissemination of the disease can be taken, including isolation or even destruction of the contaminated stock, killing of the carriers from a broodstock or rejecting infected imported eggs. Moreover, since at present viral diseases including IPN cannot be effectively cured by therapeutic agents or prevented by vaccines, control of the disease requires access to methodologies that provide rapid, reliable and sensitive diagnosis. Thirdly, the importance of epidemiological studies is indubitable, and the European Union has made mandatory the application of programs

28

DOPAZO & BARJA

of fish health management for inspection of cultured and wild fish. Early diagnosis combined with epizootiological studies provides a tool to control the disease by avoiding the movement of infected stocks between farms, different areas of a country, and different countries.

3.1. Traditional Methods of Diagnosis Traditionally (and currently) the method for detection of aquatic birnaviruses in infected fish requires isolation of the virus in susceptible cell lines (Wolf and Quimby, 1971; Ahne, 1978; Adair and Ferguson, 1981; McAllister and Reyes, 1984; Jiang and Li, 1987; Ledo et al., 1990; Rodriguez et al., 1997). Although widely used in most virology laboratories, virus cultivation is relatively expensive and timeconsuming. Many laboratories still consider it to be the most sensitive method for diagnosis of fish virus, with a theoretical detection limit of 1 viral particle. In fact, it is the first procedure stipulated by several European and American organizations, such as the American Fisheries Society (AFS), the European Union, and the Office International des Epizooties (OlE). Although a number of cell lines derived from several fish species have been reported to be susceptible to aquatic birnaviruses (Kelly et al., 1980; Fryer et al., 1981; Follet and Schmitt, 1990; Lu et al., 1990, 1999; Hedrick et al., 1991; Fernandez et al., 1993; Perez-Prieto et al., 1999), only three are routinely employed for diagnosis, since they are considered to show the highest susceptibility and to a wide range of strains (Kelly et al., 1978; Dopazo, 1991; Lorenzen et al., 1999). Cell lines used for routine purposes are the CHSE-214 (from Chinook salmon embryo), the RTG-2 (rainbow trout gonad) and BF-2 (from bluegill fry). However, the effectiveness of this method obviously depends on the susceptibility of the cell line, which unfortunately relies upon many uncontrollable factors such as the age of the monolayer or even the lineage of the cell line. In this sense, different lineages of a susceptible cell line can show different degrees of susceptibility, as recently reported by McAllister (1997) in an interesting study where the author concluded that the sensitivity of the viral isolation could be seriously compromised by cell-lineage related factors that potentially introduce false-negative results. Once isolated, identification of the virus is required, which is usually performed by serological techniques (Sanz and Coll, 1992). Among them, the neutralization test with specific polyclonal antisera is the most extensively employed because it is relatively easy to apply and to interpret (Lientz and Springer, 1973; Dorson et al., 1979; Dea and Elazhary, 1983; Hedrick et al., 1983a; Ahne and Thomsen, 1986; Lecomte et al., 1992; Kelly and Nielsen, 1993; Wolf, 1988), although it is cumbersome and requires 1 to 4 weeks to obtain confirmatory results. Many other serological techniques have been developed and applied for identification of IPN viruses in salmonid and non-salmonid fishes, using both polyclonal and monoclonal antisera. These include the coagglutination test (Kimura et al., 1984, Taksdal and Thorud, 1999), complement fixation (Finlay and Hill, 1975), fluorescent antibody test (MacDonald, 1980; Swanson and Gillespie, 1981), immunodot-blot (McAllister and Schill, 1986), immunoperoxidase (Nicholson and Henchal, 1978) or the enzymelinked immunosorbent assay (ELISA), which has shown the highest sensitivity and specificity, mainly when using monoclonal antibodies (Nicholson and Caswell, 1982; Hattori et aI., 1984; Chen et al., 1985; Rodak et aI., 1988). Despite the fact that many

IPNV

29

of these methods, mainly the immunoenzymatic-based ones, are widely employed for identification, they are influenced by the same factors that apply to cell culture procedures when applied to the isolated virus. Although these serological techniques have also been employed to detect specific circulating antibodies in infected fish, or the virus in fish tissues (Swanson and Gillespie, 1981; Chen et al., 1985; Ahne et al., 1986; ROdak et al., 1988), they have been deemed insufficiently sensitive, rapid or economically feasible to be used under field conditions (Lee et al., 1994; L6pezLastra et al., 1994; Wang et al., 1995, 1997). In addition, serological analyses of viruses are influenced by other factors such as the method of antisera production and the standardization of the procedure. In fact, a frequent problem involved in the neutralization test is the actual viral concentration used in the assay, as well as the amount of non-infectious virus produced during replication (Wolf and Quimby, 1971; McMichael et al., 1978; Nicholson and Pochebit, 1981). This problem is particularly important when typing new isolates of IPNV, since it is practically impossible to unify the results of different laboratories due to the difficulty in standardising the antigen dose employed. For some authors, standardization of serological techniques would come from the use of monoclonal antibodies (MAbs), a view we are in agreement with. Although many authors have reported the use of MAbs for identification and serotyping of IPNV strains, only a limited number of panels of MAbs have been employed. They have been obtained for different reference strains such as WB (Caswell-Reno et al., 1986; Biering et al., 1997), Sp (Dominguez et al., 1990; Brocchi et al., 1996), Ab (Dominguez et al, 1991), Ja (Tarrab et al., 1995), Nl (Christie et al., 1990; Melby and Christie, 1994; Frost et al., 1995), as well as for some other strains (Chi et al., 1991; Tarrab et al., 1993; Shankar and Yamamoto, 1994). Considering the results reported from all these panels, a great diversity of reaction patterns have been observed, which depend upon the epitope-specificity of the MAbs. This fact is useful for typing viral strains, but could be a cause for mis-detection of new isolates. In spite of the many reports on the use of MAbs for serological classification of IPNV strains found in the literature, to our knowledge only a few have employed them exclusively for detection and identifteation of field samples (Dominguez et al., 1990; Vazquez-Braiias et al., 1994). Moreover, even for serotyping, using MAbs selected for strains from a specific area may not be effective to determine the serotype of strains from a different origin (personal experience, and Novoa et al., 1995). Nevertheless, other authors such as Hill and Way (1995), found the use of MAbs against certain reference serotypes offered a safe method for serotyping new isolates. Although relatively simple, the technology to obtain these types of antibodies is not available in every laboratory, mainly because special skills are required. Those researchers routinely using MAbs for fish viruses do sometimes provide other laboratories with aliquots of the monoclonal antisera, or even with the specific hybridoma However, in order to obtain real standardization of the procedure, a unique panel of MAbs should be commercially available for both detection of any serotype of the virus, and identification of the specific serotype of a new isolate. '

30

DOPAZO & BARJA

3.2. Molecular Techniques for Identifying and Typing IPNV 3.2.1. Overview Scientists generally associate the concept of 'molecular techniques of diagnosis' with DNA-based techniques, mainly nucleic acid hybridization and the polymerase chain reaction (PCR), which were first employed for identification and typing of salmonid bimaviruses in the late 80's (Christie et al., 1988; ROdak et al., 1988). However, other techniques involving detection and characterization of the viral genome or polypeptides must also be considered. Although the aim of the present review is focused on diagnosis of IPN viruses from saImonid fish, an unquestionable fact is that these techniques can and are being applied for diagnosis of this virus from almost any kind of aquatic animal, including fish and shellfish. Molecular diagnostic methods, as other diagnostic procedures, can be applied for detection, identification and/or typing of bimaviruses. A common practice is the use of a molecular technique for identification of a previously isolated virus, which carries the inconveniences associated with isolation in cell culture described above, although avoids those associated with serological identification. Some authors have applied those techniques for detection and identification of the virus in inoculated cell culture long before cytopathic effect (CPE) is visualized, which implies that isolation is not essential (Dopazo et al., 1994). This is particularly useful in the case of strains that replicate poorly under standard culture conditions. Detection and identification of the virus directly in fish tissues has the obvious advantage of avoiding the need for viral isolation. However, it has other associated disadvantages, such as the need for optimization of the extraction of the viral nucleic acid from the tissues, and reducing or eliminating inhibitors that interfere with the molecular technique to be applied (Wiedbrauk and Farkas, 1995). Each molecular method of diagnosis presents its as its own advantages and limitations for particular applications. In the following section, we briefly describe some of the techniques most frequently employed, as well as an overview of their use in molecular diagnosis of salmonid IPN viruses.

3.2.2. Analysis of electropberotypes The separation of the RNA and polypeptide components of the virus have been proven to be important tools for diagnosis and typing of IPNV isolates. Once the nature and characteristics of the genome and polypeptides of IPN virus were confirmed (Dobos, 1976; Dobos et al., 1977), the potential of using these viral components for characterization of new isolates was soon considered. Due to the fact that double-stranded RNA constitutes the viral genome, its high stability makes its extraction from crude virus quite simple. Moreover, since it is bisegmented, a previously isolated virus can be easily identified as IPNV by subjecting the extracted genome to SDS-polyacrylamide gel electrophoresis (SDS-PAGE), and visualization of two fragments in ethidium bromide or silver stained gels. Although Ganga et al. claimed in 1994 to be the first to use this technique for diagnosis, this method has been widely applied in most laboratories, and its potential use to identify isolates of IPNV has previously been reported by other authors (Hedrick et al., 1983a; Hsu et aI., 1989). The use of this technique for identification of salmonid and non-salmonid isolates has been published (Dopazo 1991; Novoa et al., 1993a; Nakajima and

IPNV

31

Sorimachi, 1994; Seo et al., 1998; Jung et al., 1999). Nevertheless, Ganga et al. (1994) reported a new way of using this technique for diagnosis of IPNV, since they were able to detect the viral genome directly from fish tissues when the fish showed clear signs of disease. On the other hand, the electropherotypes (EFTs) of genomic segments and polypeptides, i.e. their mobility in SDS-PAGE, have been widely employed for comparison of different strains of IPNV and for typing new isolates. Thus, although the early studies on comparison of viral proteins did not find differences among typed strains (Underwood et ai., 1977; Chang et ai., 1978) further reports comparing the size of the RNA and protein species of different strains of IPNV (MacDonald and Gower, 1981) led the authors to the conclusion that only 3 serotypes of IPN virus existed, and that each one shows a unique pattern. Other authors compared the composition of RNA and polypeptides of the strains Sp, Ab, and VR-299, and suggested the use of those markers to distinguish between the strains of this virus and to determine the identity of new isolates (Hedrick et aI., 1983 a,b, 1985; Hsu et ai., 1989). For this purpose, early reports suggested that the EFTs of the viral RNA was the most convincing marker to distinguish isolates (MacDonald et ai., 1983), even better than the traditional neutralization test (MacDonald and Gower, 198 I). However, Hedrick et al. (1985) and Hsu et al. (1989) reported diversity among Ab type strains by EFTs of polypeptides, which could not be detected by the genome patterns, and concluded that the analysis of EFTs of polypeptides together with seroneutralization were better markers for comparison and typing. Despite the knowledge of the existence of ten serotypes of IPNV, the 3 traditional serotypes, corresponding to strains originally isolated from salmonid fish, are still employed as reference strains to be compared with any isolate of the virus. In fact, some authors have pointed out the importance of a full characterization of new isolates from a particular area, and its comparison with the reference strains from salmonids, since it can provide some information as to the virulence of the strain (Novoa et al I 993b). This comparison has been extensively reported and, in some cases, a full typing of new strains (from salmonid and non-salmonids) using EFTs of RNA and/or polypeptides has been published (Espinoza et al., 1985; Hsu et ai., 1989, 1993; Novoa et al., 1993 a; b; Nakajima and Sorimachi, 1994; Sohn et al., 1995; Cutrin et al., 2000). Considering the economic and ecological importance of the IPN virus in salmonid fish, characterization of new isolates as part of management programmes is essential in order to establish the degree of diversity of isolates, as well as the geographical relationships. In this sense, determination of genomic variation is simpler and more reliable for classification of viruses than serological techniques, since it is not influenced by uncontrolled factors. Comparison of the EFTs of the viral genomes and polypeptides, however, has been reported to be helpful in determining the identity of individual virus strains (Hedrick et at., 1985; Ganga et al., 1994). Many authors have used this method in epizootiological studies to identity the possible origin of a certain isolate. Thus, in 1985 Espinoza et al. employed the technique to demonstrate that an IPN virus isolated from affected rainbow trout in Chile was introduced from North America. In the 1980's, Hsu et al. (1989) compared RNA and polypeptides of IPNV isolates from Japanese eel and rainbow trout in Taiwan, and observed that most of them were closely related to the Ab type strain, which was widespread in Japan at the time when trout and their eggs were imported to Taiwan from Japanese hatcheries. However, some isolates from trout were of the VR-299 type, and the authors

32

OOPAZO & BARJA

suggested that this serotype could have been introduced from America to Japan and then imported into Taiwan. In a further report, Hsu et af. (1993) compared the protein electropherotypes of trout and eel isolates to the 3 reference strains, and observed that VR-299 had been spread throughout the eel farms in Taiwan. In 1993, Novoa et af. (1993a) reported the first isolation of VR-299 in Spain, although the authors did not suggest introduction of this type from North America, speculating on the possible role of feral animals and reservoirs. More recently, Cutrin et af. (2000) compared the RNA electropherotypes of a large number of strains from epizootics in salmonid and nonsalmonid farms, and from reservoirs, and observed a certain similarity between strains from both origins. A wide range of electrophoretic techniques are known which provide high separation efficiency with a relatively limited amount of equipment (Westermeier, 1997). The procedure involves the extraction of the particular component and a certain degree of purification before subjecting to electrophoresis. The treatment is slightly different depending on the type of viral component subjected to analysis. The process is much shorter and less laborious when studying the genome profile, which makes it a better option than the analysis of the polypeptides for identification of isolated viruses. For this purpose, a single flask of 25 cm 2 of infected cells is required (Hsu et al., 1989), and after extensive CPE is visualized, the crude virus must be subjected to purification. Although some authors have reported a complete purification of the virus in sucrose and/or CsCI gradients (Ganga et af., 1994; Jung et af., 1999) followed by extraction and purification of the dsRNA, in our experience, a partial purification of the virus through a 30% sucrose cushion should be enough for SDS-PAGE (Dopazo, 1991; Cutrin et al., 2000). Since only the visualization of the 2 genomic segments is required for identification, percentages of acrylamide between 5 and 10% (around 0.15% bis-acrylamide) can be chosen for the resolving gel (Hedrick et al., 1983b; Ganga et af., 1994; Jung et aI., 1999; Cutrin et al., 2000). We suggest the use of mini-gels to reduce the running time to 1 to 2 h at around 200 V. Our laboratory has started to employ the GenePhor (Pharmacia) electrophoresis equipment to run gels in 112 h at 500 V, which dramatically reduces the time for diagnosis. The gels sold by the manufacturer must be submerged in a 0.1 % SDS solution prior to subjecting the sample to SDS-PAGE. If the electrophoretic method is to be applied for genotyping of IPNV strains, analysis of the RNA EFTs can also be employed, but if a correlation between genotyping and serotyping is expected, analysis of the EFTs of viral polypeptides is preferred. However, some discrepancies concerning the convenience of RNA or polypeptide EFTs for this purpose can be found in the literature (MacDonald et aJ., 1983; Hedrick et af., 1985; Hsu et af., 1989; Cutrin et al., 2000). Because of this, two variants of the procedure have been described. Some authors employ purification of the virus and direct electrophoresis of the viral proteins in discontinuous SDSpolyacrylamide gels (Hedrick et aI., 1983 a; b; Comps et aI., 1991). In this case, the gels can be easily stained with coomassie blue, although higher sensitivities are obtained with silver stain. Other authors prefer the isotopic labelling of the proteins in infected cell mono layers, which increases the sensitivity of the method, although it can be hazardous, which is an inconvenience. In addition, in order to reduce nonspecific background, the infected mono layers must be subjected to irradiation with V.V. light to reduce labelling of cellular proteins (Hsu et at., 1993), or the labelled proteins must be subjected to immunoprecipitation (Nakajima and Sorimachi, 1994).

IPNV

33

3.2.3. Fingerprinting Other molecular techniques have been employed for classification and typing of IPNV isolates. A method based on RNA fingerprinting has been reported by Hsu et al. (1995) to characterize the genome heterogeneity between isolates. Briefly, 32p _ labelled viral RNA was SUbjected to RNase Tl treatment in the presence of tRNA used as a carrier, and then subjected to two-dimensional gel electrophoresis. The gel is exposed to X ray film, and the pattern of oligonucleotide spots analysed by means of specific computer software. However, to our knowledge, it has never been applied to IPNV by other researchers, perhaps because it is cumbersome and needs special skills, and because other techniques are available which are more standardized and provide the required information at the level of typing.

3.2.4. Nucleic acid sequencing A number of reports have been published on the use of nucleic acid sequencing for genogrouping birnaviruses, and their accession numbers to the Genbank database are listed in Table 1.

Strain

Jasper

Segment A

B Sp

DRT

Nl

A

B A B A

Accession number M 18046 M 58756 U 56907 U 48225* M 58757 D 26526 D 26527 D 00701 *Not lIsted by Regenmortei et al.

Table 1. Genome sequence accession numbers (as in Genbank database) oflPNV, as listed by Regenmortei et al., 2000.

However, to date only Heppell et al. (1993) have published a genogrouping by comparison of sequences of IPNV strains from the 10 serotypes. These authors observed that all strains related to the same serotype were highly homologous at the genomic level, although some serologically different strains showed certain similarities. In a later study, the same authors (Heppell et al., 1995) reported sequence variability at the VP2 region among isolates from the AI, A2, A3, A4 and A7 serotypes, although in this case the aim of this study was not only genogrouping. In a more recent study, Hosono et al. (1996) sequenced a fragment corresponding to the VP2INS junction region from a large number of strains, including most of the reference serotypes, and some isolates from marine fish. The authors first concluded that genogrouping by sequencing should be considered a useful method to group aquatic bimaviruses. Moreover, they reported that the marine bimaviruses were clearly differentiated from the salmonid IPN viruses, and constituted a new genogroup of aquatic birnaviruses. Other authors have employed sequencing of the viral genome for typing new isolates of IPNV (Havarstein et al., 1990; Pryde et al.,

34

OOPAZO & BARJA

1993). However, the technique is time-consuming and relatively expensive, and therefore other methods of typing, such as RFLPs, are preferred.

3.2.5. Nucleic acid hybridization Although poorly efficient in the detection of the viral genome directly in IPNVinfected fish tissues (Dopazo et al., 1994), RNA-DNA dot-blot hybridization is a powerful tool for detecting and identitying IPN viruses in inoculated cell cultures, even before clear CPE is visualized. However, an absolute limit of sensitivity of direct hybridization of 104 to 105 molecules, depending on the size of the target genome, must be taken into consideration (Desselberger, 1995). Nucleic acid hybridization was first used for diagnosis by Christie et al. (1988), who used a oligonucleotide probe (corresponding to the 3' end of segment A) for identification of the Norwegian NI strain by Northern blot hybridization. A 32 p end-labelled oligonucleotide was also employed by Rimstad et al. (1990b) for diagnosis of salmonid isolates by hybridization. This oligonucleotide, corresponding to the NS region, hybridized equally with the 3 traditional reference strains, and was able to detect as little as 10 ng of purified viral RNA. Interestingly, the authors were also able to detect the virus in cells harvested from infected mono layers (with a sensitivity of around 10 7 TCIDsoIml), which represents a clear advantage of the method since it eliminates the need to extract the viral RNA. The procedure was applied to a large number of field samples (88 viral isolates from rainbow trout and Atlantic salmon, obtaining positive results from 87% of the samples). The use of oligonucleotides can result in loss of specificity due to their small size, as suggested by Dopazo et al. (1994). However, for some scientists there is an alternative view that longer probes might hybridize even if there are mismatches within the region of the probe. Nevertheless, to our understanding this problem should be easily solved using high stringency hybridization conditions. In the study by Dopazo et al. (1994),2 radiolabelled cDNA probes were tested for detection of IPN viruses, one corresponding to segment A (WBl, 812 bp) and the second one to segment B (A4, 569 bp). A total of 21 strains were tested, mostly from salmonid fish, and including reference strains from North America and Europe, as well as Spanish isolates. The sensitivity of both probes was 10 to 15 pg of purified viral RNA, corresponding to around 109 viral particles. Both probes showed an undesirable higher specificity for the American strain, and it was extremely high with the A4 probe. Although the method showed a extremely low sensitivity when detecting viral RNA extracted from infected fish tissues, it showed a 100% effectiveness in detection of the virus in infected cell cultures, as early as 8 h postinoculation with American strains, and from 12 to 24 h p.i. with the European and Spanish strains assayed. No significant differences were observed between the salmonid and non-salmonid isolates. The authors suggested that the sensitivity of the method could be increased by improving the extraction of the viral RNA from tissue samples. Other authors reported the use of the hybridization technique using nonisotopic biotinylated probes to detect the virus in infected cell monoloyers from just 1 well from a 24-well plate (Batts and Winton, 1994). Two approaches can be employed. Firstly, extraction of total RNA, including genomic and viral mRNA, is performed. For this purpose, traditional treatments such as proteinase K followed by phenol-chloroform extraction and ethanol precipitation of

IPNV

35

the total RNA (Dopazo et al., 1994), or others as described by several authors (Sambrook et ai., 1989; Rimstad et al., 1990b), can be substituted by the use of commercial kits (those based on phenol and/or guanidine thiocyanate extraction, or resins), with proven effectiveness and reproducible results. Once the RNA is extracted, it must be denatured, which is the most critical step. For this purpose we prefer to boil the RNA to 104°C for 5 min (although routinely 95°C for 5 min is employed by others), immediately transfer to an ice-bath, and add methylmercury hydroxide to a final concentration of 10 J1M. However, other researchers prefer to avoid use of toxic chemicals. The denatured RNA can be loaded onto a positivelycharged nylon membrane by a dot-blot or slot-blot cell. Fixing of the RNA to the membrane is currently performed by baking at 80°C in a vacuum oven, or using UV as a crosslinker. However, some authors have reported the use of a conventional microwave oven for this purpose (Batts and Winton, 1994). The second approach has been reported by Rimstad et al. (1990b), who directly loaded 500 J11 of centrifuged infected tissue culture harvest (with a viral concentration of around 107 TCIDsolml) onto the nylon membrane. This procedure reduces the steps required for the diagnosis, which is of particular importance in intensive surveys involving large numbers of samples. For the probe, isotopic labelling can be used, as described by several authors (Rimstad et aI., 1990b; Dopazo et al., 1994). Although the authors of the present paper have experimented with non-isotopic labelling such as biotin and digoxygenin, the highest levels of sensitivity were obtained with radiolabelling. However, chemiluminescence, which is known to produce strong signals (Desselberger, 1995; Wiedbrauk and Farkas, 1995) should be assayed for IPNV probes. Two types of probes can be employed. Some authors have used oligonucleotide probes (e.g. the 24-bp oligonucleotide DNA probe 5'GAAGGAGATGACATGTGCTACACC-3', reported by Rimstad et al., I 990b), because they are easy to obtain from various companies. Others prefer the use of larger probes in order to reduce the risks of loss of specificity due to the small size of the oligonucleotides (Dopazo et al., 1994). One of these probes, with a size of 372 bp, has been proven to detect most of the IPNV isolates tested (unpublished data). This probe, named IPN-CP, is obtained by reverse transcription followed by PCR (RTPCR) with primers lPN-CPt (5'-TAGTCCCAAACCGAGACCTA-3') and IPN-CP2 (5'-CCTCCGGCTGCGTGTGAC-3'). However, much work is still ongoing in order to standardize its use for routine diagnosis. Recently Biering and Bergh (1996) have reported a different way to apply nucleic acid hybridization technology. In their study, in situ hybridization (ISH) was compared to immunohistochemistry (lHC). Although ISH did not appear to be more sensitive than IRC, background staining was virtually absent, and the labelling was easily localized in the cytoplasm of the infected cells. However, the method was timeconsuming and required 3 days to complete the diagnosis, which could be the reason this method has not been reported in further studies.

3.2.6. RT-peR Until recently, virology research was impossible without growth of viruses in cell culture. However, thanks to PCR, failure to grow virus does not rule out their presence. Due to its extreme sensitivity, since the development of PCR technology in

OOPAZO & BARJA

36

1983 by the Cetus Corporation (Saiki et aI., 1985; Mullis and Faloona, 1987), nucleic acid amplification methods have been used by an increasing number of laboratories. Although it is not yet of extensive use for diagnosis in fish virology, the use of RTPCR for diagnosis of IPN viruses from salmonid and non-salmonid fish is increasing, and several authors have proven it to be a rapid, reliable, sensitive and convenient method of diagnosis (Blake et 01., 1995; Lopez-Lastra et 01., 1994; Wang et 01., 1997; Olveira et 01., 1997, 1998; Williams et 01., 1999). Therefore, it is an important commitment for virologists dedicated to diagnosis to offer PCR among the list of available techniques of their laboratories. Primer sets used for amplification of IPNV nucleic acid are shown in Table 2.

Reference Rimstad et aI., 1990

Name Sequence (5'-3')

Sev,t A

_. L __.... ｟ｾＡ［ａ｢Ｎﾣﾥｃｇｉｧｔ＠ 2

TGATGCCGTTGTTCTCATCAGCTG

-.L____._. . _GAAGGAGATGACAｔＮｇｾＺｴａ＠ McAllister et aI.,

A

1991 LOpez-Lastra et aI., 1994

A

Sizez 486

4

GAGAGATGTGTTTGCCACCATCTC

I

CCAACTGGGTTTGACAAGCC

__. 3103

U---CITCTCATIGACGGGITcOOC·········----··--·- 263

3

JJI. .. _. ._._CGQAAA Ｑ｢Ｎ｟ｾｧｇＡ｜＠ V

Blake etal., 1995

A

TCCAQb..QG.. _. ._. ___ 657 GCGGTCTGCTGGTTGAGCTGG

ﾷｾＱｽＭｗｃｔＥｩ

1180

B 1 GCCGACATCGTCAACTCCAC . .-Pr. ----. .-........---.. .--.. -.. -..---..------.-.........--.::.:,:-::'---- -.--........ 524 Pr B2 GACAGGATCATCTTGGCATA

Olveira et aI., 1997, 1998

Jav!....__Ｎｑｃａｇｾｔ＠ _____._ 607 Jav 2 TGGAGTTCTGGGTCCATCCC

A

B

Wang et aI., 1997

Williams et aI., 1999 Crane et aI., 2000

A

A A

__ CGGAAAｔａｃｇＮｾ＠ ___._._._. __ Pr D2 TGGCTCCTGTCATGGACTGG J7 FL___ｾｑＡ｜ｇＮａｉｍｔ｟＠ Pr F2 GACAGGATCATCTTGGCATA . WBI ｟ｾﾣｍｉｔａｇｑｃ＠ WB2 CGTCTGGTTCAGATTCCACCTGTAGTG . . .Tab Ｑ｟ＮｑｉｇｔｾａｃＹ＠ .-......-.. . Tab2 TGTGCACCACAGGAAAGATGACTC ＮｦＡＺｾｌ＠

274 524 206 421

'Genomic segment. lSize (bp) of the amplified fragment. 3Second amplification in nested peR.

Table 2. Primers used for RT-PCR diagnosis ofIPN virus.

The high sensitivity of this method for detection of IPN virus has been repeatedly proven. In 1990, Rimstad et 01. (1990a), using double-nested PCR, observed a sensitivity of 10 pg (l00 ng after the first step of PCR) for detection of purified RNA

IPNV

37

from salmonid isolates, which strongly improved the sensitivity of the hybridization techniques. In this study, the detection of the polymerization product by hybridization with a biotinylated probe was weaker with the Sp strain than with the VR-299, and better results were obtained when magnetic separation was applied to the synthesized DNA fragment. Detection levels of 1 pg were reported by Lopez-Lastra et af. (1994), in a study where the authors also reported that nested-PCR detected asymptomatic carriers in field samples. Blake et af. (1995) tested 4 primers sets for detection of 9 serotypes of IPNV in samples of kidney and spleen, and observed that 2 of these sets, Pr B and Pr D (amplifying fragments of 524 and 174 bp, respectively), detected all strains with at least the same efficiency as isolation in cell culture, and concluded that this method could be a rapid and reliable substitute for cell culture procedures. Other researchers, however, have employed this technology for identifying viruses isolated in cell culture, or even to confirm the etiology of isolates identified by traditional methods such as neutralization (Rodriguez et ai., 1995, 1997; Olveira et ai., 1998; Alonso et ai., 1999; Jung et ai., 1999). Recently, Wang et ai. (1997) employed a method of RT-PCR in one step, which is easier and more rapid than the traditional 2step procedure. This method showed the same sensitivity as the traditional method (from 15 fg to 15 pg) (Wang et ai., 1995), and was applied not only to infected cells, but also to fish tissues. Finally, McAllister et ai. (1991) and more recently Williams et ai. (1999) demonstrated the potential use of a single RT-PCR assay for simultaneous detection of different viruses. Williams et ai. (1999) reported sensitivities of 100, I. and 32 TCIDsolml for simultaneous detection of lPN, IBN (infectious hematopoietic necrosis) and VHS (viral hemorrhagic septicemia) viruses. This method showed at least the same sensitivity as isolation in cell culture, and shorter times are required for diagnosis in a survey study. The technique can be applied for identification of a previously isolated virus, or even to infected mono layers prior to visualization of CPE. This procedure is widely employed because it ensures detection of infectious IPN viruses. It can also be applied to detect the virus directly in tissues from infected fish. In this way, the procedure accelerates the diagnostic process, can detect the virus even in concentrations below the detection limit of cell culture, and ensures that the virus will be detected even if susceptible cells are not available, or when the sample size is too small for cell culture. However, many scientists do not accept positive PCR results as conclusive diagnosis since it does not ensure detection of presence of infective viruses, but may detect naked non-infective viral nucleic acids (Alonso et ai., 1999; Berthe et ai., 1999; Hiney, 2000). It gets even more complicated when the possibility of prior vaccination against IPNV is considered. Although vaccination against IPN is not common worldwide, it has been applied and has proven effectiveness in reducing problems due to IPN disease in some countries (Leong and Fryer, 1993; Gudding et ai., 1999; Hlistein, 1996). Therefore, even in those cases when the fish farmer can confirm if the fish stock under analysis has, or has not, been subjected to vaccination, a tool to ensure that the virus detected correspond to an infectious virus, and not to a vaccine-derived strain, should be available. Discrimination between infected and vaccinated fish can firstly be compromised if vaccination has been performed with attenuated virus. For instance, unless a live modified vaccine is labelled with a suitable marker, that discrimination can be difficult to attain even in those cases when the virus is isolated in cell culture, as is the case for other viruses (Wolf, 1988). For some viruses, it is suggested that such

38

DOPAZO & BARJA

markers could be a certain epitope which could be differentiated by a specific MAb (World Health Organization, 1991). However, due to the frequency of mutation, the structure of that epitope would be subjected to probable alteration, and thus a given vaccine-specific MAb would no longer recognize the vaccine-derived strain. Nevertheless, in spite of the great amount of work carried out to obtain an effective attenuated vaccine for IPNV, its use is not widespread. Therefore, it is not presently a concern in diagnosis ofIPN viruses. Nowadays, nucleic acid vaccines are emerging as a promising procedure for vaccination of fish against IPNV (Vaughan et aI., 1998). Using this technology, injected DNA can persist in the fish tissues for relatively long periods of time. The use of a molecular test, such as PCR, directly on the tissues could provide a positive result. Therefore, in this case the use of a suitable marker would be advisable in order to differentiate between infected from nucleic acid vaccinated fish. Some authors have developed a method of 'stringent microplate-hybridization of PCR products', which has been successful to differentiate wild from vaccine-derived strains of a certain virus (Inouye and Hondo, 1993; Takeda et al., 1994). Research is encouraged in order to apply this methodology in diagnosis of IPN viruses. As already indicated, one of the main advantages of PCR is the sensitivity of this method, which is theoretically capable of detecting a single target molecule (Clewely, 1989), although when the target is RNA, reverse transcription is the critical step influencing the sensitivity of the PCR (Byrne et aI., 1988). The sensitivity for detection of IPN viruses has repeatedly been proven to be extremely high: from 1 to 10 pg (Rimstad et aI., 1990a; Lopez-Lastra et al., 1994) or even from 15 fg to 15 pg (Wang et aI., 1997). However, this high sensitivity carries serious disadvantages as even minuscule quantities of contaminating nucleic acids can produce false positives. Therefore, optimization of the procedure must be carried out before a virology lab employs RT-PCR for diagnosis. The first aspect to be optimized is the working area. Three separated rooms should be assigned to the PCR-related process (Desselberger, 1995): i) a room for processing fish tissues, ii) a 'clean room' for extraction of RNA and preparation of reagents and reaction mixtures, and iii) a room for the PCR. To avoid carryover of contamination, special care must be taken with other factors: air conditioning (which should be switched off in the 'clean room'), frequent change of gloves, use of positive displacement pipettes or aerosol-resistant pipette tips, opening only one tube at a time during reagent addition, and treating tubes with U.V. light. Some authors have reported the advantage of using single-tube non-interrupted RT-PCR to reduce carryover of amplified DNA sequences (Kwok and Higuchi, 1989; Wang et aI., 1997). One of the most critical steps influencing the effectiveness of RT-PCR is the procedure for extraction of viral RNA. In fact, the aim of the sample processing for PCR must be not only making viral nucleic acids available for amplification, and in a sufficiently high concentration, but also to maximize reduction of inhibitory substances, which is especially important for direct detection in fish tissues. The choice of protocol for extraction is inevitably a compromise between purity of the viral RNA and the time required for extraction of large numbers of samples, and the economic cost. The methods most frequently employed for extraction of IPNV RNA from salmonid tissues or from infected cells have been the guanidinium thiocyanate method as described by Chomczynski and Sachi (1987) (Rodriguez et al., 1995, 1997; Wang et aI., 1997; Alonso et al., 1999), and the traditional proteinase K treatment

IPNV

39

followed by phenol-chloroform extraction and ethanol precipitation (L6pez-Lastra et al., 1994). Although economically affordable, these methods are time consuming and occasionally not reliable due to variability of the quality of the stored reagents. In recent years, there has been considerable interest among different manufacturers in devising methods for preparation of nucleic acids for amplification, reducing the number of steps, the risk of contamination, and the cost. However, the use of these methods for IPN viruses can be found in only few reports, such as that by Williams et al. (1999), who used the TriReagent LS (Sigma Chemical Co.). Our laboratory is involved in an intensive study on optimization of RT-PCR for diagnosis of IPNV. This study includes testing 3 types of extraction methods: the traditional proteinase K method, commercial methods based on guanidinium and phenol extraction (Trizol LS-Reagent, Gibco-BRL), and the one-step resin-based extraction kits (RNeasy, Qiagen; NucleoSpin, Macherey-Nagel; SV Total RNA Isolation System, Promega). The preliminary results (unpublished data) indicate that although expensive, the best results (in terms of sensitivity and reliability) are obtained with the resin-based kits. However, good results are also obtained with the Trizol method, which moreover provides a relatively rapid and low-cost method for processing large quantities of samples. Primer selection must be carefully performed in order to ensure detection of any, or the required, serotype of IPN virus. Many primer pairs for RT-PCR for diagnosis of IPNV can be found in the literature, and a selection is shown in Table 2. Most of them have been chosen for segment A and complementary to the sequences corresponding to the variable VP2 and NS regions. Only primer sets with efficiency proven by the original authors have been chosen (Fig. 1). Among them the set' Jav' is worthy of special mention, having been demonstrated to detect strains from any serotype and origin in infected CHSE-214 cells 12 h pj (Olveira et al., 1998, and unpublished data). Different protocols have been reported for RT-PCR for IPNV, and other protocols and different kits can be found. A single method cannot be prescribed or recommended at present, and each laboratory still requires extensive tests of preselected protocols in order to determine the optimum method and reaction conditions for their purposes. The 2-step regular RT-PCR and the single tube non-interrupted method show similar levels of sensitivity (Blake et at., 1995; Rodriguez et al., 1995, 1997; Wang et al., 1997; Alonso et al., 1999), and this can be dramatically increased (close to 1O,000-fold times) by nested or semi-nested PCR (Rimstad et at., 1990a; L6pez-Lastra et at., 1994). On the other hand, nested PCR also provides an important improvement in specificity, and is a useful tool to reduce risk of false positives since the inner primer set will only polymerize if the outer pair has produced the specific expected DNA fragment. Another way of demonstrating the specificity of the polymerized fragment is by hybridization of the RT-PCR product with a probe known to be specific for IPNV. Other controls must be included in each procedure: i) two negative controls consisting of a non-bimavirus virus, and a reaction mixture without target RNA, and ii) a positive control consisting of a reference strain or a group of strains from different serotypes, to ensure that if an IPN virus is in the sample it will be detected. Finally, another advantage of this method is the ease of interpretation of results using ethidium-bromide stained agarose gels, although better resolution and higher sensitivities can be obtained with silver stained PAGE gels (L6pez-Lastra et al., 1994). Other methods of detection, such as magnetic separation of the synthesized

40

OOPAZO & BARJA

DNA fragment have been reported (Rimstad et al., I 990a), but their use is not widespread perhaps because it is cumbersome.

3.2.7. Restriction fragment length polymorphisms In recent years, the analysis of restriction fragment length polymorphisms (RFLP) has been revealed to be a rapid and apparently accurate way of typing aquatic bimaviruses, and therefore a powerful substitute for EFTs and even serological techniques. In an early study, Heppell et al. (1992) analyzed the RFLPs of strains from the 10 serotypes described by Hill and Way (1995), based on a 359 bp cDNA fragment (amplified by PCR) corresponding to the highly variable NS coding region. They established the existence of 3 major groups approximately corresponding to the traditional serotypes (AI, A2 and A3), and 10 subgroups poorly correlating with serotypes. If special attention is to be paid to the correlation between geno- and serotyping, the RFLP analysis should be applied to cDNA fragments obtained from the hypervariable VP2 coding region (Biering et al., 1997), because it encodes the major outer capsid protein encompassing all neutralization epitopes. This was suggested by Novoa et al. (1995), although they could not find an exact correlation. Better results can be obtained with larger fragments. In a more recent study Lee et al. (1996) performed RFLPs with a 1180 bp fragment representing most of the VP2 region to compare IPNV strains from serotypes AI, A2, A3, A5 and AI0. They found that genogrouping by this method was identical to serological classification with VP2-specific monoclonal antibodies. In addition, Biering et al. (1997), working with the same fragment obtained from strains of a wider range of serotypes (AI, A2, A3, A5, A6, A7 and A9), were able to differentiate by RFLPs all the serotypes tested. However, much work must be conducted including strains from the 10 serotypes of IPNV in order to standardize the assay conditions needed to ensure the correlation between geno- and serotyping, and for correct typing of new isolates. This standardization will make RFLPs a useful tool in epidemiological studies, as well as in identification of these viruses as part offish health management programmes. The procedure can be summarized in 5 steps: i) extraction of viral RNA, ii) cDNA synthesis employing random hexamers as primers in a reverse transcription reaction, iii) PCR using any of the 2 sets of primers shown in Table 3, iv) restriction endonuclease digestion (Table 4), including reference strains from the 10 serotypes, and v) electrophoresis in agarose or acrylamide gels.

Primers (5'-3') up

dw UQ dw

GCT GAC ATC GTG AAC TCC AC GAC AGG ATC ATC TTG GCA TA TGA GAT CCA TTA TGC TTC CAG A GAC AGG ATCATC TTGGCA TAGT

up, upstream; dw, downstream

Table 3. Sets of primers employed for RFLP

Reference Lee et aI., 1996 Biering et aI., 1997

IPNV

Reference

Enzymes

Lee et a!., 1996

Xho I, Bam HI, Eco RI, Pvu II, Mbo I, Ava II, Bst ElI, Dde I, Hae III Pvu II, Ban II, Apa I, Rsa I, Hae III

Biering et al. 1997

41

Table 4. Restriction endonucleases employed for RFLP

Although the procedure is practical for typing large numbers of strains, nevertheless, it might be not useful for typing a new isolate. In this case, our laboratory has designed a variant of the technique. Briefly, a 1180 bp fragment obtained from a viral isolate by RT-PCR is subjected to digestion with Pvu II (Fig 2). Depending on the RFLP observed, a second enzyme is chosen for digestion of a second aliquot of the 1180 bp fragment. This second step leads to a conclusive identification of the isolate tested (Cutrin et al., 1997; Cutrin, 1998).

4. FUTURE WORK AND FUTURE TECHNIQUES It is true that the molecular techniques of diagnosis are becoming a refined tool for detection and identification of viruses, and hence the salmonid birnaviruses, providing a practical way of rapid, sensitive, reliable and safe diagnosis. However, despite the fact that the use ofthose techniques is widespread in human virology, their use in fish virology still needs much research on standardization. The first aspect that concerns scientists is whether the detection of nucleic acid (mainly by RT-PCR) actually means presence of the virus. Research must be conducted in order to determine the maximum time that a naked viral genome can remain detectable in the fish tissues. Meanwhile, standardization of the procedure to detect the virus in infected cell cultures should be considered in order to ensure reproducibility among laboratories. Moreover, different probes and sets of primers from different authors should be compared to choose a group ofthem to be used for standardized routine diagnosis. Finally, many other molecular techniques which increase sensitivity and/or specificity have been reported for diagnosis in clinical virology. These include the 'ligase chain reaction', 'cycling probe technology', 'Q-Beta replicase system', or special methods for detection such as 'signal amplification' or 'sequence based detection' (Widebrauk and Farkas, 1995). The cost of the equipment necessary to carry out some of these (e.g. real-time PCR) is the major drawback for their application in fish diseases laboratories. Much effort and research, however, will be necessary to introduce any of these techniques for the improvement of the diagnosis of salmonid birnaviruses.

ACKNOWLEDGEMENTS The authors wish to thank the Xunta de Galicia for grant XUGA20009B96, as well as the Ministerio de Educaci6n y Ciencia for grants IFD97-0953-C02-02 and MAR99-0637-C02-01 (Scientific Contribution No. 00112001 of the Instituto de Acuicultura).

DOPAZO & BARJA

42

c

B

D

E

1073 _ DJ)

rn

t

'03

no

I

781

t71

m

196 111

-

III

IG

IG

III

IU

77

-

"""Y"" : LnJU.

ｬＡｮＮｙＧ

t

1

DB 1073= rn-

- ... "10

Ｚｾ］ＮＡ＠

..

110_

)!O _

ｾﾧ］＠

ｩｾ］＠ ｬｾ］＠ 131-

,

'OlllO_

196 12+-

196121_

72-

72-

72-

')'1'

!to WI!

1..2 CI

17

Jt1PLP's:

Ul

=-.., 1073= ｾｬＭ

-.,.

ＧＢｾ＠

7ll-

111_ ....... ,v:

ill

Ill)

073=

372-.!!!., ｾ＠

' Ol _

.!:

l!n.Y, us

m f("fl-

'Ol_

ｦｾ］＠

-m-

l!n.Y s: [1lr

"""Y"" :v:.n

"""Y"" : DJlLll

"""Y""""'LU

Ｚ ｾ＠

'Ol_

.,.

--'" ｦｾ］＠ 77-==

310_

ｾ］Ｍ

It94%) when used to diagnose heavily infected fish (i.e., candidates for soft flesh syndrome). When light infections were included in the evaluation, the sensitivity of the test decreased to 79%. This decrease in sensitivity should not be of concern however, as this infection level typically does not cause obvious softening of the musculature. Using this procedure leaves no observable sign of manipulation and permits paired sampling over time for strengthened statistical analyses. On the other hand, this technique is time consuming and labor-intensive.

Figure 5. Atlantic salmon (Salmo salar). Illustration of the hyohyoideus ventralis muscle. Reproduced from Aquaculture 156, St-Hilaire, S., Ribble. c., Whitaker, OJ. and Kent, M.L. Evaluation of a nondestructive diagnostic test for Kudoa thyrsites in farmed Atlantic salmon (Salmo salar). p. 139-144, Copyright 1997, with permission from Elsevier Science and S. St-Hilaire.

Briefly, opercular muscle samples (usually O.l-O.4g) are minced in I ml 0.65% fish saline (6.5 g NaCl in 1000 ml water) and pressed between two Plexiglas plates with the resulting extract collected. A wet mount is prepared by adding one drop of this extract to one drop of fish saline on a glass slide with a coverslip. The wet mount preparation is scanned randomly for 5 min using phase contrast illumination at 320X magnification. When screening multiple samples, all dissection tools and materials must be thoroughly cleaned to prevent inadvertent transfer of spores between samples.

334

MORAN & KENT

2. DIAGNOSIS USING THE REACTION (PCR) ASSAY

POLYMERASE

CHAIN

2.3.General Multiple copies of ribosomal DNA (rDNA) sequence occur within cells, which make it advantageous to use this sequence for developing accurate and sensitive DNA-based diagnostic tests. Hervio et al. (1997) designed the Kudoa-general and K. thyrsites-specific PCR primers based upon the small subunit region of the rDNA (SSU rDNA). This PCR assay has been used to effectively detect all levels of infections in salmonid fishes. It has also been used successfully to detect these infections in non-salmonid species (e.g., tube-snout Aulorhynchus jlavidus and Pacific hake) (Hervio et al. 1997; Shaw et al. 1997) (Fig. 6). When compared with wet mount preparations, the PCR assay showed an increased sensitivity for detecting these infections in Atlantic salmon musculature (92% (22/24) vs. 38% (9/24), respectively) (Hervio et al. 1997). In all likelihood, this wide discrepancy in detection results from the PCR assay's sensitivity to all parasite stages, whereas only the stellate spore stages are consistently detectable in the wet mount preparations. One isolate of putative "K. thyrsites" from netpen-reared Atlantic salmon in Chile could not be amplified using the K. thyrsites-specific PCR assay (Whipps et al. 2001). When investigated further, it was determined that this isolate had a SSU rDNA sequence that was distinct from other confirmed K. thyrsi/es "isolates". Therefore, it is recommended that researchers use both the Kudoa-general primers and the K. thyrsites-specific primers when examining suspect cases from new hosts and/or new geographic regions.

2.4.Sampling (Non-Lethal and Lethal) As discussed previously, non-destructive sampling procedures were used to investigate potential infections in processed salmon without adversely affecting the quality of the final product. This is achieved by sampling the opercular muscle (i.e., m. hyohyoideus ventralis) (Fig. 5). St-Hilaire et al. (l997b) used only wet mount preparations to screen collected muscle samples, but this sampling method could easily be adapted as a non-lethal sampling procedure to collect tissue samples from anesthetized fish for subsequent use in PCR assays. This non-lethal collection may also involve blood samples. However, the effectiveness of screening blood samples for detecting K. thyrsites infections must be investigated further. The presence of K. thyrsites stages in the blood has been confirmed using transmission trials, but it was concluded that these stages were both transient and underwent minimal proliferation, based on the inability to observe the stages in histological preparations (Moran et al. 1999a).

KUDOA THYRSITES

335

30501 1636 1018

517

Figure 6. PCR amplification of DNA using the K. thyrsites-specific PCR assay developed by Hervio et al. (1997). Positive signals, equivalent to 909 base pairs, indicating K. thyrsites infections in Atlantic salmon (Salmo salar), Pacific hake (Merluccius productus), and tube-snout (Aulorhynchus jlavidus), are seen in lanes KtA, KtH, and KtT, respectively. Molecular weight

standard (I kb ladder) is shown as lane I kb. Negative control is represented as lane -ve. Attempts to amplity DNA using the K. thyrsites-specific PCR primers from other Kudoa spp. were unsuccessful, as indicated in lanes Ka (K. amamiensis), Km (K. miniauriculata), and Kp (K. paniformis). No amplification was observed in other fish species infected with the myxozoans Henneguya salminicola (lane Hs) and PKX (lane PKX), the microsporeans Nucleospora salmonis (lane Ns) and Loma salmonae (lane Ls), or the uninfected control fish hosts Atlantic salmon (lane AS) and coho salmon (Oncorhynchus kisutch) (lane CS). Parasite DNA (ISO ng/Ill) was amplified using the primers Ktl8S6f and Ktl8SIr (1.25% agarose gel, ethidium bromide stained). Reproduced from Canadian Journal of Zoology 75 , Hervio, D.M.L., Kent, M.L., Khattra, J., Sakanari, J. , Yokoyama, H. , and Devlin, R.H. Taxonomy of Kudoa species (Myxosporea), using a small-subunit ribosomal sequence, p. 2112-2119, Copyright 1997, with permission from the National Research Council of Canada Research Press.

Moran and Kent (1999) used lethal sampling procedures in both their K. thyrsites seasonality investigations and in their non-salmonid fish species surveys. Abdominal muscle samples were collected post-mortem for immediate analysis using wet mount preparations. This site provides greater muscle samples (4-5 g), which are necessary for detecting low-level infections. However, with both the confirmation of the sensitivity associated with sampling the opercular muscle by St-Hilaire et al. (l997b), and the advent of the highly sensitive PCR assay, such large sample sizes may no longer be necessary. If the refrigerated tissue samples cannot be processed within a reasonable amount of time, it is recommended that the samples be stored at -20°C temporarily. For long term storage, -70°C is recommended to minimize shearing of the DNA, which may affect amplification success.

336

MORAN & KENT

2.3. Template Preparation Digest samples of 25-50 mg of tissue in lysis buffer (10 mM Tris, 10 mM EDTA, pH 8.0, 1% SDS containing 200 J-Ig/ml proteinase K (Gibco BRL 25530-015» at 37°C for at least 4 h (preferably overnight) with agitation. Extract DNA from samples using phenol: chloroform: isoamyl alcohol (25:24:1) (PCI) (Sigma Chemical Co. Cat. No. P-3803) twice, precipitate with cold 100% ethanol (EtOH) and 3M sodium acetate (pH 7.0), and centrifuge at 10 000 rpm. Air dry the DNA pellet and resuspend in 50 J-II of Tris EDTA (TE) buffer (10 mM Tris, 1 mM EDTA, pH 8.0). Store the resuspended DNA at 4°C. Quantity the amount of DNA in each sample using a spectrophotometer and dilute in TE buffer to prepare the PCR template working solution (150 ng/J-II of DNA). In lieu of PCI DNA extraction, various kits are available that are convenient and avoid the usage of PCI. DNA extraction kits, such as those produced by Qiagen (QIAamp DNA Mini Kit Cat. No. 51304), have been used successfully to extract DNA from K. thyrsites-infected Atlantic salmon muscle tissue for use in the PCR assay (unpublished data). When using the QIAamp DNA Mini Kit, either thaw frozen samples (maximum of 25 mg) or rinse in 1X phosphate-buffered saline (PBS, pH 7.2) (if preserved in EtOH) and add lysis buffer (180 J.lI Buffer ATL + 20 J.lI of proteinase K) to each sample. Vortex samples and incubate at 56°C until lysis is complete (typically 3 hours or more). Briefly centrifuge tubes after incubation and add 200 J.lI of Buffer AL to each. Vortex and incubate samples at 70°C for 10 min, followed by a brief centrifugation. Add 200 J.ll of ice-cold 100% EtOH to each sample, vortex for 15 sec, and briefly centrifuge tubes. Apply each sample mixture (maximum of 650 J.l\) to an appropriately labeled QIAamp spin column seated within a 2-ml collection tube. Centrifuge at 8000 rpm for 1 min and transfer spin column to a clean 2-ml collection tube. Add 500 J.ll of Buffer AWI to each spin column and centrifuge at 8000 rpm for 1 min. Transfer each spin column to a clean 2-ml collection tube. Add 500 J.ll of Buffer AW2 to each spin column, centrifuge at 14000 rpm for 3 min, and transfer spin column to clean 1.5-ml microcentrifuge tube. Add 40-200 J.lI of warm Buffer AE to each spin column and incubate at room temperature for 1-5 min. Centrifuge at 8000 rpm for 1 min. Quantify DNA and prepare working solutions (150 ng/lll) for use in the PCR assay.

2.4. PeR-Based Assay (Single-Round) Primers developed by Hervio et al. (1997): K. thyrsites-specific primers: Kt18S6f- 5' -CTCAACCAACTGGCCTCG- 3' Kt18S1r- 5' -CGTCAATTTCTTTAAATTTGG-3'

KUDOA THYRSITES

PCR ( per 50 III reaction): Molecular biology-grade H 20 lOX PCR buffer 2mMdNTPs 50mMMgClz 20 pmolllli Forward primer (Kt18S6t) 20 pmolllli Reverse primer (Kt18S 1r) 5 units/ill Taq polymerase DNA template (150 ng/Ill)

337

34.75 III 5.0 III 5.0 III 1.5 III 1.25 III 1.25 III 0.25 III 1.0 III

PCR thermal cycler program: 95°C for 3 min, followed by 35 cycles (94°C for 1 min, 62°C for 1 min, noc for 1.5 min), and conclude with an extension period ofnoC for 10 min.

2.5. Gel Analysis Agarose gels (1.25% w:v) in IX Tris-acetate (TAE) buffer (40 mM Tris-acetate, 1 mM EDTA) are used to visualize the PCR products. Add appropriate loading buffer to each 10-12 III sample and apply to the wells. Run agarose gel at 100V for 1 hour. Stain gel with ethidium bromide and analyse using ultraviolet (UV) illumination. Positive samples are indicated by a PCR amplicon of909 base pairs (Fig. 6).

3. DIAGNOSIS USING IN SITU HYBRIDIZATION (ISH) 3.1. General A non-radioactive ISH protocol is currently being developed for detecting K. thyrsites in Atlantic salmon musculature, with the ultimate goal of its use as a diagnostic tool for identifying presporulation infections in any fish or invertebrate (Le., potential alternate host) species. This procedure is based upon the ISH protocol of Antonio et al. (1998) that has been used to analyse Myxobolus cerebralis infections in rainbow trout (Oncorhynchus mykiss). The probes used in the M. cerebralis ISH procedure are prepared using SSU rDNA primers designed by Andree et al. (1998). This K. thyrsites ISH protocol employs a cocktail of the four oligonucleotide primers designed by Hervio et al. (1997) (e.g., Kt18S6f, Kt18S1r, KUDlf (5'-CTATCAACTAGTTGGTGA-3'),KUD2r (5'-CAATGTCTGGACCTGGTG-3'), which is 3' -tailed with digoxigenin-Iabeled deoxyuridine triphosphate (DIG-dUTP) (Roche Molecular Cat. No. 1093088). Labeled probes hybridize to parasite DNA present in deparaffinized tissue sections of K. flryrsites-infected Atlantic salmon. By using this procedure, the parasite distribution can be easily visualized throughout all stages and levels of infection. However, we have observed some cross-reaction with the PKX myxosporean in the control tissue sections (unpublished data). As SSU rDNA sequence from additional myxozoan species becomes available, more specific probes should be developed to avoid this cross-reaction. Nevertheless, the following protocol is highly efficient at detecting these infections until such time. Sterile solutions including Buffer 1 (100 mM Tris-HCI, 10 mM NaCI, pH 7.5), Buffer 2 (100 mM Tris-HCl, 100 mM NaCl, 50 mM MgCI2 , pH 9.5), Tris-CaCh (0.2 M Tris-HCI,

338

MORAN & KENT

2.0 mM CaCh, pH 7.2), 20X standard saline citrate (SSC, pH 7.2), and IX PBS must be prepared in advance. The following procedure is performed over 2.5 days.

3.2. Preparation of Tissue Sections Tissue sections are prepared using standard histological techniques (Humason 1979). Both 10% buffered formalin and Davidson's solution are preferred fixatives for fish tissues. The use of Bouin's fixative in ISH should be avoided due to decreased cytological detail (Nuovo and Richart 1989) and the suspected chemical reactivity of picric acid (a component of Bouin's fixative) with the deoxyribose sugars of DNA (Weiss and Chen 1991). After embedding the tissues in paraffin, sections are cut to 5-6 !lm, and adhered to Superfrost Plus pre-treated slides (Fisher Scientific Cat. No. 12-550-15). It is important to avoid handling slides directly to prevent contamination. The tissue sections should be protected from dust by storage in a slide box with dessicant at -20°C for up to several weeks. Prolonged storage may result in decreased signal and/or increased background (Ausubel et al. 1993).

3.3. Preparation of Labeled Probes A cocktail of the four oligonucleotide primers (Le., KUDI f, KUD2r, Kt18S6f, Kt18S I r) was used in preparing the labeled probes for the ISH protocol. A DIG Oligonucleotide tailing kit, sufficient for 25 labeling reactions, is available from Roche Molecular (Cat. No. 1417231) or components may be ordered separately. In a 20 !ll reaction, add 4 !ll of 5X reaction buffer (Roche Molecular Cat. No. 1243276), 4 !ll of CoCIz solution (Roche Molecular Cat. No. 1243306), 1 !ll DIG-dUTP (Roche Molecular Cat. No. 1093088), 1 !ll oligonucleotide cocktail (containing 200 pmol/!ll of each primer), 1 !ll freshly-prepared dA TP (25 !lM) (Roche Molecular Cat. No. 1051440), 1 !ll terminal transferase (Roche Molecular Cat. No. 220582), and 8 !ll DNAse-free water (Sigma Chemical Co. Cat. No. W-4502). After vortexing, the reaction is incubated at 37°C for 15 min and immediately transferred to ice. The reaction is then placed in a spin vacuum for 45 min and a blue precipitate indicates successfully labeled probes. The precipitate is resuspended in 20 !ll of DNAse-free water and stored at -20°C. When stored at this temperature, the labeled probe should remain viable for approximately 6 months.

KUDOA THYRSITES

339

3.4. Treatment of Tissue Sections De-wax and rehydrate tissue sections using the following schedule (at room temperature): Xylene'" (Fisher Scientific Cat. No. X-16-4) - 3x1O min washes 100% EtOH - 2x1O min washes 95% EtOH - 3 min wash SO% EtOH - 3 min wash 70% EtOH - 3 min wash 50% EtOH - 3 min wash Distilled water - 3 min wash Tris-CaCh-3 min wash Air dry tissue sections at room temperature. "'Note: Less toxic alternatives to xylene (e.g., Fisherbrand Hemo-De Clearing Agent. Fisher Scientific Cat. No. 15-1S2-507A) exist and may be considered for dissolving the paraffin wax. However, these alternatives have not been investigated for use in this protocol.

3.S. Digestion and Pre-Hybridization Prepare proteinase K solution at 50 Ilg/ml (add 50 III of 10 mg/ml proteinase K to 10 ml Tris-CaCh buffer). Incubate in a 37°C water bath for 40 min. After incubation, store at room temperature until required. Apply approximately I ml of proteinase K solution to completely cover each tissue section, and transfer slides to hybridization oven at 37°C for 45 min. During this 45 min digestion, prepare pre-hybridization solution (i.e., hybridization solution without the labeled probe). Approximately O.S ml of pre-hybridization solution is required for each slide. Pre-hybridization solution for 12 slides (keep reagents on ice): 5.0 ml deionized formamide (Sigma Chemical Co. Cat. No. F9037) 2.0 ml20X SSC 0.52 ml of 9.5 mg/ml heat-denatured salmon testes DNA (Sigma Chemical Co. Cat. No. D9156) 2.0 ml 50% dextran sulfate (prepared in DNAse-free water) (Sigma Chemical Co. Cat. No. D6001) 0.2 ml 50X Denhardt's solution (Sigma Chemical Co. Cat. No. D2532) Place pre-hybridization solution (9.72-ml volume) on aliquot mixer (Ames Company Model 4651) until required. After 45 min proteinase K digestion, briefly rinse the tissue sections in IX PBS. Wash tissue sections in IX PBS 3xIO min. Air dry the tissue sections at room temperature. Add pre-hybridization solution to tissue sections for 2 to 2.5 hours in a hybridization chamber at room temperature. After incubation, discard the pre-hybridization solution as waste, and briefly wash the tissue sections 3x in 2X SSC. Air dry after washes.

340

MORAN & KENT

3.6. Hybridization Hybridization solution for 12 slides (keep reagents on ice): 5.0 ml deionized formamide 2.0 ml20X SSC 0.52 ml of9.5 mg/ml heat-denatured salmon testes DNA 2.0 ml 50% dextran sulfate 0.2 ml50X Denhardt's solution Add appropriate labeled probes (I9.6J.11) to hybridization solution. Set thermal cycler (e.g., MJ Research PTC-I00) fitted with slide griddle to 100°C. Place slides in slide griddle and add appropriate hybridization solution (0.6 ml) to each slide. After 10-min denaturation, remove slide carefully, and add DNAse-free HybriSlip (Research Products International Corp. Cat. No. 247457) for hybridization at 39°C overnight (maximum 16-17 hours).

3.7. Stringency Washes and Anti-DIG-AP Incubation Rinse tissue sections in warm 2X SSC to remove HybriSlips and wash in 2X SSC for 20 min at 40°C on shaker-incubator. Prepare appropriate volume of blocking buffer (e.g., for 10-12 slides, add 146.55 ml of Buffer 1, 3.0 ml of sheep serum (Sigma Chemical Co. Cat. No. S2263), and 0.45 ml of Triton X-I00 (Fisher Scientific BPI51-100». Prepare antibody dilution buffer (e.g., for 10-12 slides, add 7.9 ml Buffer 1, 80 ",I of sheep serum, and 24 ",I of Triton X-100) without the antidigoxigenin-alkaline phosphatase antibody conjugate (anti-DIG-AP) (Roche Molecular Cat. No. 1093274) and mix gently. Add anti-DIG-AP (e.g., for 10-12 slides, add 8.0 Ill) to antibody dilution buffer approximately 20 min before required and keep on aliquot mixer. Continue with stringency washes: 2X SSC - 20 min wash at 40°C I X SSC - 3xl5 min washes at 40°C O.5X SSC - 10 min wash at room temperature Buffer 1 - 10 min wash at room temperature Add blocking buffer to tissue sections and block for 1 hour at room temperature with mild shaking on orbital shaker. Air dry slides completely (5-10 min) at room temperature. Add antibody dilution buffer containing anti-DIG-AP to each slide. Incubate for 2 hours at I5-20°C. Discard anti-DIG-AP solution and rinse in Buffer 1. Wash in Buffer 1, 2xIO min at room temperature. Wash slides in Buffer 2 for 10 min at room temperature on orbital shaker. Air dry slides completely.

341

KUDOA THYRSITES

3.8. Immunological Detection Approximately 0.8-1 ml is required for each slide (e.g., for 10-12 slides, prepare 10 ml of substrate). Add 45 J.tl ofNBT (Roche Molecular Cat. No. 1383213),35 J.tlof X-phosphate (Roche Molecular Cat. No. 1383221), and 9.92 ml of Buffer 2 to a 15ml centrifuge tube wrapped with foil and place on an aliquot mixer. While in a dark chamber, add I ml of substrate to each slide and incubate for 3 hours at room temperature. Remove substrate and check signal under microscope. If signals are strong, rinse the slides in distilled water and wash 2xl0 min at room temperature. Wash in distilled water overnight at room temperature on orbital shaker, with mild shaking. Counterstain the tissue sections for 3-5 min in 0.05% aqueous Bismarck brown Y (Sigma Chemical Co. Cat. No. B2759). Wash in distilled water 3xlO min each. Air dry completely. Rinse slides in xylene 2xO.5 min each, add Krystalon mounting media (Fisher Scientific Cat. No. SP 15-1 00) and coverslip, and let dry.

3.9. Interpretation of Results Kudoa thyrsites infections are easily discernible using the ISH protocol. Developing parasites are visible as dark blue forms against the lightly stained muscle tissue (Fig. 7).

7 Figure 7. Atlantic salmon (Salmo salar). Somatic muscle tissue section processed using the in situ hybridization protocol described within this chapter (based upon Antonio et al. 1998). This section is from an experimentally infected Atlantic salmon showing positive binding of the digoxigenin-Iabeled probes to the developing parasite stages (previously unpublished). Scale bar = 80 llm.

342

MORAN & KENT

4. DISCUSSION Kudoa infections are somewhat unique in that the concern is not over loss resulting from mortality, but is instead due to the negative impact that these infections have on the quality and marketability of fish products. With regards to K. thyrsites infections in Atlantic salmon, there is significant concern within the international salmon aquaculture industry due to the effect that these infections are having on the market perception of farmed salmon products. Signs of soft flesh syndrome, resulting from heavy infections, are usually not observed until 3-6 days post-harvest. This is typically long after the product has been sent to market, thus leading to claims made against the producers of poor product quality. Aquaculture companies have a considerable investment in each fish and any potential income generated from the final product is effectively "lost" when these claims occur. Currently, the only available option to these companies is to have their employees trained in diagnosing infections microscopically and removing potential cases of soft flesh syndrome prior to product shipment. If performed diligently, the microscopic screening of fish products will prevent the majority of products susceptible to softening from being shipped to market, thus avoiding the adverse market perception. This technique requires skilled training, is fairly labor intensive, and often the determination is subjective. Hervio et a1. (1997) were the first to obtain SSU rDNA sequence from the multivalvulid genus Kudoa and used this data to compare its phylogenetic relationship with previously sequenced bivalvulid myxozoan genera (e.g., Myxidium and Myxobo/us) obtained from GenBank. Hervio et a1. (1997) developed both Kudoageneral and K. thyrsites-specific primers based on the alignment of Kudoa SSU rDNA sequence data. The Kudoa-general primers were successful in amplifyihg SSU rDNA of K. amamiensis from buri, K. miniauriculata from bocaccio (Sebastes paucispinis), K. paniformis from Pacific hake, the three K. thyrsites isolates from Pacific hake, Atlantic salmon, and tube-snout, and K. ciliatae from sand whiting (Si/lago macu/ata). The specificity of the K. thyrsites-specific primers was investigated by attempting to amplify SSU rDNA from the three K. thyrsites isolates, as well as from other myxosporean (e.g., Kudoa ciliatae, Henneguya sa/minico/a, PKX), microsporean (Nuc/eospora sa/monis, Loma sa/monae), and host fish species (Atlantic salmon, coho salmon Oncorhynchus kisutch). Amplification was seen only in those reactions containing DNA from the three K. thyrsites isolates. Using the sequence data, Hervio et a1. (1997) concluded that the three different K. thyrsites isolates collected in British Columbia (BC) waters were, in fact, the same species. In addition, the SSU rDNA sequence of the South African K. thyrsites isolate (from snoek Thyrsites atun) is essentially identical (99.4%) to the original Atlantic salmon isolate collected in BC (Kent 1998). More recently, partial SSU rDNA sequence was obtained from a histozoic myxozoan of netpen-reared Atlantic salmon from Chile that was morphologically identical to K. thyrsites. Using SSU rDNA sequence alignment, this species clearly groups within the genus Kudoa; however, it appears distinct from known Kudoa species (Whipps et a1. 2001). We must therefore consider that the current records of K. thyrsites from around the world and from a number of different fish species may be the result of an assemblage of morphologically indistinguishable species, which must be investigated further using SSU rDNA sequence analyses. For practical research purposes, molecular tests will prove highly useful in investigating the biology of these enigmatic organisms. The information available on

KUDOA THYRSITES

343

the development and transmission of K. thyrsites in fish is incomplete. Moran et al. (1999a) used the K. thyrsites-specific PCR test and determined that there were parasite stages, likely extrasporogonic stages circulating throughout the blood supply, present in all tissues collected. Moran et al. (1999c) confirmed the presence of these blood stages using transmission trials. However, the blood stages were not observed in the corresponding histological sections indicating that the PCR assay greatly increases the ability to detect sparsely occurring stages in light and moderately infected fish. Using sequences of Kudoa reported by Hervio et al. (1997), Yokoyama et al. (2000) developed both a PCR assay and polyclonal antibodies for Kudoa amamiensis. Consistent with our PCR results, they found that their PCR assay detected K. amamiensis infections earlier and at a higher prevalence than traditional methods. The polyclonal antibodies used in the immunofluorescent antibody test (IFAT) reacted to both mature K. amamiensis spores as well as to prespore stages, with no obvious signs of cross-reaction with other myxozoan species (e.g., K. thyrsites, Myxobolus /COi, and M. arcticus) (Yokoyama et al. 2000). The ISH procedure described within this paper uses a cocktail of the two Kudoa-general primers and the two K. thyrsites-specific primers developed by Hervio et al. (1997), which are non-radioactively labeled using digoxigenin. This protocol has yet to be optimized as signs of non-specificity were observed in negative control proliferative kidney disease organism (PKX) infected tissue sections. Using the software program Amplify v1.2 for the Macintosh (Engels 1993), we believe that it is the Kt18S6f and KUD2r primers that are responsible for this slight cross-reactivity. Investigations are currently underway to determine whether labeled probes developed using only the Kt18S1r and KUDlfprimers are sufficient for this ISH protocol. As it currently stands, the ISH may be used as a diagnostic tool for identifying extrasporogonic K. thyrsites infections in Atlantic salmon, with the ultimate goal of it being applied for diagnosing these infections in other fish or invertebrate (i.e., potential alternate host) species. As additional SSU rDNA sequences from other myxozoan species becomes readily available, more specific probes should be developed to avoid this apparent cross-reactivity. Until such time, caution is advised when identifying extrasporogonic stages. This, of course, is not the case when clinical signs associated with the myxozoan infections are diagnostic features (e.g., PKX). However, with the combined use of molecular diagnostic tools such as PCR and ISH, extrasporogonic stages can be confidently identified in tissues other than the final site, thus assisting researchers in producing such valuable in-depth developmentaVpathological studies as those performed for M. cerebralis in rainbow trout (EI-Matbouli et al. 1995). Sakanari (1994) briefly discussed the concerns regarding the requirement for trained personnel, expensive equipment, and the delay in achieving PCR results. These concerns are also inherent with the use of ISH. Presently, these molecular tools cannot address the issue of identifying potential candidates for soft flesh syndrome. These assays are far too sensitive, which if used in this manner would unfortunately exclude from the market valuable products having light infections unlikely to cause this condition. As a result, these techniques are limited to use in research. However, if supported by the industry, non-lethal PCR screening of opercular muscle or blood samples from netpen populations could aid in advising aquaculturalists of impending problems until a dependable soft flesh diagnostic test based upon the intensity of infection is developed.

344

MORAN & KENT

5. FUTURE APPLICATIONS At present, it has not been confirmed whether or not marine myxozoan life cycles mirror those of their freshwater counterparts. With the implication of both annelids and bryozoans as necessary hosts in freshwater myxozoan life cycles, diagnostic molecular probes may be used to screen vast populations of potential alternate hosts from the marine environment. When, and if, the role of invertebrate alternate hosts is confirmed in the K. thyrsites life cycle, the actinospore stages will not resemble the characteristic stellate spore morphology as is seen in the fish hosts. Therefore, molecular techniques will be required to confirm the identity of the alternate stages, as was performed by Anderson et al. (1999), Andree et al. (1997), Antonio et al. (1998; 1999), and Bartholomew et al. (1997). These molecular tools are also being applied to investigate the possible assemblage of Kudoa species that had previously suggested a global distribution of K. thyrsites. It is research areas such as these that molecular diagnostic tools will play such pivotal roles.

ACKNOWLEDGEMENTS The authors would like to thank O.B. Antonio for assistance with the ISH procedure and R.P. Hedrick for logistical support. We would also like to thank OJ. Whitaker for providing K. thyrsites-infected Atlantic salmon. Permission to reproduce figures was provided by H. Yokoyama, S. St-Hilaire, O.M.L. Hervio-Heath, Elsevier Science Publishers, and the National Research Council of Canada Research Press. Funding for JOWM was provided by the Natural Sciences and Engineering Research Council of Canada as a postdoctoral fellowship.

LITERATURE CITED Anderson, c.L., Canning, E.U. and Okamura, B. (1999) 18S rDNA sequences indicate that PKX organism parasitizes Bryozoa. Bull Eur Assoc Fish Pathol 19,94-97. Andree, K.B., Gresoviac, S.J., and Hedrick, RP. (1997) Small subunit ribosomal RNA sequences unite alternate actinosporean and myxosporean stages of Myxobolus cerebralis, the causative agent of whirling disease in salmonid fish. J Eukaryot Microbiol44, 208-215. Andree, K.B., MacConnell, E., and Hedrick, RP. (1998). A nested polymerase chain reaction for the detection of genomic DNA of Myxobolus cerebralis in rainbow trout Oncorhynchus mykiss. Dis Aquat Org 34: 145-154. Antonio, D.B., Andree, KoB., McDowell, T.S. and Hedrick, RP. (1998) Detection of Myxobolus cerebralis in rainbow trout and oligochaete tissues by using a nonradioactive in situ hybridization (ISH) protocoL J Aquat Anim Health 10,338-347. Antonio, 0.8., EI-Matbouli, M. and Hedrick, RP. ([999) Detection of early deve[opmental stages of Myxobolus cerebralis in fish and tubificid oligochaete hosts by in situ hybridization. Parasitol Res 85,942-944. Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.O., Seidman. J.G., Smith, JA, and Struhl, K. (1993) Current protocols in molecular biology. John Wiley & Sons, New York, New York. Barja, J.L. and Toranzo, AE. (1993) Myoliquefaction post-mortem caused by the myxosporean Kudoa thyrsiles in reared Atlantic salmon in Spain. Bull Eur Assoc Fish Pathol 13, 86-88. Bartholomew, J.L., Whipple, M.J., Stevens, D.G. and Fryer, J.L. (1997) The life cycle of Ceratomyxa shasta, a myxosporean parasite of salmon ids, requires a freshwater polychaete as an alternate host. J Parasitol 83. 859-868.

KUDOA THYRSITES

345

Baudin-Laurencin, F. and Bennassr, N. (1993) Post-mortem liquefaction of sea water fiumed brown trout Salmo trulta resulting from Kudoa infections, p. 60. In Diseases of fish and shellfish, Book of Abstracts. Sixth Intl ConfEur Assoc Fish Pathol, Brest, France, 5-10 Sept 1993. EI-Matbouli, M., Hoffinann R.W. and Mandok, C. (1995) Light and electron microscopic observations on the route of the triactinomyxon-sporoplasm of Myxobolus cerebralis from epidermis into rainbow trout cartilage. J Fish Bioi 46, 919-935. Engels, W. R. (1993). Contributing software to the Internet: the Amplify program. Trends in Biochemical Sciences 18,448-450. Hervio, D.M.L., Kent, M.L., Khattra, J., Sakanari, J., Yokoyama, H. and Devlin, R.H. (1997) Taxonomy of Kudoa species (Myxosporea), using a small-subunit ribosomal DNA sequence. Can J Zool 75, 2112-2119. Holliman, A. (1994) Sea trout, Salmo trulta, a new host for the myxosporean Kudoa thyrsites (Gilchrist)? Vet Rec 134,524-525. Humason, G.L. (1979) Animal tissue techniques. W.H. Freeman Co., San Francisco, California. Kabata, Z. and Wbitaker, DJ. (1989) Kudoa thyrsites (Gilchrist, (924) (Myxozoa) in the cardiac muscle of Pacific salmon (Oncorhynchus spp.) and steelhead trout (Salmo gairdneri). Can. 1. Zoo!. 67, 341342. Kent, M.L. (1998) Protozoa and Myxozoa, pp. 49-67. In Kent, M.L. and Poppe, T.T., eds. Diseases of seawater netpen-reared salmonid fishes. Quadra Printers Ltd., Nanaimo, British Columbia. Moran, J.O.w. and Kent, M.L. (1999) Kudoa thyrsites (Myxozoa: Myxosporea) infections in pen-reared Atlantic salmon in the eastern North Pacific Ocean, with a survey of potential nonsalmonid fish reservoir hosts. J AquatAnim Health II, 101-109. Moran, J.D.W., Margolis, L., Webster, J.M. and Kent, M.L. (1999a) Development of Kudoa thyrsites (Myxozoa: Myxosporea) in netpen-reared Atlantic salmon determined by light microscopy and a polymerase chain reaction test. Dis Aquat Org 37, 185-193. Moran, J.D.W., Wbitaker, DJ. and Kent, M.L. (1999b) A review of the myxosporean genus Kudoa Meglitsch, 1947, and its impact on the international aquaculture industry and commercial fisheries. Aquaculture 172, 163-196. Moran, I.D.W., Whitaker, DJ. and Kent, M.L. (1999c). Natural and laboratory transmission of the marine myxozoan parasite Kudoa thyrsites (Gilchrist, 1924) to Atlantic salmon. J Aquat Anim Health II, 110-115. Nuovo, G .L., and Richart, R.M. (1989) Buffered formalin is the superior fixative for the detection of HPV DNA by in situ hybridization analysis. Amer I Pathol 134, 837-842. Palmer, R. (1994) Kudoa - the Irish experience, pp. 18-21. In Conley, D.C., ed. Kudoa Workshop Proceedings. Province of British Columbia, Ministry of Agriculture, Fisheries and Food, Aquaculture Industry Development Report 94-0 I. Sakanari, J. A. (1994) Detecting parasites in fish tissue, p. 11. In Conley, D.C., ed. Kudoa Workshop Proceedings. Province of British Columbia, Ministry of Agriculture, Fisheries and Food, Aquaculture Industry Development Report 94-01. Shaw, R.W., Hervio, D.M.L., Devlin, R.H., and Adamson, M.L. (1997) Infection of Aulorhynchusflavidus (Gill) (Osteichthyes: Gasterosteiformes) by Kudoa thyrsites (Gilchrist) (Myxosporea: Multivalvulida). J Parasitol83, 810-814. St-Hilaire, S., Hill, M., Kent, M.L., Wbitaker, DJ. and Ribble, C. (l997a) A comparative study of muscle texture and intensity of Kudoa thyrsites infection in farm-reared Atlantic salmon Salmo salar on the Pacific coast of Canada. Dis Aquat Org 31, 221-225. St-Hilaire, S., Ribble, C., Wbitaker, DJ. and Kent, M.L. (l997b) Evaluation of a nondestructive diagnostic test for Kudoa thyrsites in farmed Atlantic salmon (Salmo salar). Aquaculture 156, 139-144. Weiss, L.M., and Chen, Y-Y. (1991) Effects of different fIXatives on detection of nucleic acids from paraffm-embedded tissues by in situ hybridization using oligonucleotide probes. J Histochem Cytochem 39, 1237-1242. Wbipps, C.M., Smith, P. and Kent, M.L. (2001) A Kudoa sp. in pen-reared Atlantic salmon (Salmo salar) from Chile. Am Fish SoclFish Health Sec Newslett 29, 5-6. Wbitaker, DJ. and Kent, M.L. (1991) Myxosporean Kudoa thyrsites: a cause of soft flesh in farm-reared Atlantic salmon. J Aquat Anim Health 3, 291-294. Wbitaker, DJ., Kent, M.L. and Margolis, L. (1994) Myxosporean parasites and their potential impact on the aquaculture industry, with emphasis on Kudoa species, pp. 2-7. In Conley, D.C., ed. Kudoa Workshop Proceedings. Province of British Columbia, Ministry of Agriculture, Fisheries and Food, Aquaculture Industry Development Report 94-01. Yokoyama, H., Inoue, D., Sugiyama, A. and Wakabayashi, H. (2000) Polymerase chain reaction and indirect fluorescent antibody technique for the detection of Kudoa amamiensis (Multivalvulida: Myxozoa) in yellowtail Seriolaquinqueradiala. Fish Pathol35, 157-162.

Reviews: Methods and Technologies in Fish Biology and Fisheries 1. 2. 3.

J.R. Sibert and J.L. Nielsen (eds.): Electronic Tagging and Tracking in Marine Fisheries. 2001 ISBN 1-4020-0125-8 D. Symes and J. Phillipson (eds.): Inshore Fisheries Management. 2001 ISBN 1-4020-0128-2 C.O. Cunningham (ed.): Molecular Diagnosis of Salmonid Diseases. 2002 ISBN 1-4020-0506-7

Kluwer Academic Publishers - Dordrecht / Boston / London