Techniques for diagnosing viral diseases of salmonid fish

Cell culture for amplification and the techniques most used for identification of salmonid viruses neutralization, immunofluorescence and, to a lesser extent, immunoperoxldase, complen~ent fixation, agglutination, electron microscopy, immunodifussion or radioimmunoassay may soon be replaced by other techniques such as enzyme immunoassays (immunodot and enzyme-linked immunosorbent assay, ELISA) and hybridization with DNA probes. It is expected that developments in monoclonal antibodies (MAbs) and amplification by the polymerase chain reaction (PCR) will increase sensitivity of enzyme immunoassays and DNA hybridizations, respectively. Some of these new methods should provide detection of the low levels of virus present in adult carriers and perhaps in eggs (although this is more complicated). Other &agnostic methods, such as measurement of virus-specific salmonid immunoglobulins (Igs) by ELISA or stimulation of immunological cellular memory by in vitro CO-culture of salmonid lymphocytes with viral proteins, could also b e further developed. VIRAL DISEASES OF SALMONIDS AND THEIR DIAGNOSIS Infectious pancreatic necrosis (IPN), viral haemorrhagic septicaemia (VHS) and infectious haematopoietic necrosis (IHN) are the major viral diseases causing severe mortalities of farm-reared salmonids. Initially found in North America, IHNV has been detected in Italy and France (Baudin 1987, Bovo et al. 1987) and is now considered a major problem for the European Community (EC 1990). Similarly, VHSV, initially found in Europe, has recently been detected in salmon on the west coast of the United States (Brunson et al. 1989, Hopper 1989). Rapid and sensitive methods of diagnosis are critical if dissemination of the viruses causing these diseases is to be controlled, because no vaccines currently exist for their prevention. Rapid diagnosis is desirable during the acute phase of the disease, whereas highly sensitive diagnosis is necessary to detect viruses during the carrier phase of the disease. Methods for diagnosis of viral diseases can be separated into 2 groups, those measuring the presence of the virus and those detecting the specific response of the host. To detect virus in fish it is necessary first to sample the fish population correctly, second to amplify Addressee for correspondence the content of the sample, and third to identify the virus. To identify the virus we can measure its cytopathic effect on cell culture, proteins, nucleic acids and/or morphology. For most of these methods reagents such as polyclonal or monoclonal antibodies (Abs) and DNA probes are needed, all of which should be complementary to structural components of the virus to be identified. To detect the reaction of the host we can measure either their humoral or their cellular immunological responses. SAMPLING SALMONID POPULATIONS Accurate diagnosis of a viral disease depends to a large extent on the number of viruses per volume of sample. The number of viruses per volume of sample depends on the number of fish with viruses, the number of viruses per fish and the final dilution of the sample before analysis. Maximal levels of viruses reported in fish tissues undergoing natural or experimentally induced viral infections during the acute phase of the disease can vary between 104 and 10' tissue culture infectious dosages (TCIDSO) g-l tissue (de Enkelin & Bearzotti 1981, Mulcahy et al. 1983, Hattori et al. 1984, Kimura et al. 1984, Neukirch 1984a, b, Yoshimizu & Kimura 1985, Mulcahy & Batts 1987, Nishimura et al. 1988, Basurco & Coll 198913). When sampling a populaO Inter-Research 1992 212 Dis. aquat. Org. 13: 21 1-223, 1992 ANTIBODIES TO VIRAL PROTEINS tion of salmonids a statistically significant number of tion, samples can be concentrated to very small final fish must be collected. But it is somewhat difficult to volumes). In addition it can be made highly specific to collect a representative and random sample, especially the virus by select~ng sequence-specific primers (see when there is no mortality (estimated number of virus below). per gram of tissue 5 103). For sampling, fish must be pooled according to age, source, species, strain, and water supply (Amos 1985). The sample size needed to detect at least one carrier fish in populations of a fixed To identify the presence of viral proteins by size, and with an assumed carrier incidence, can be immunological methods specific Abs should be precalculated at the 95 O/ O confidence level from the Poispared. These can be either polyclonal antibodies son probability distribution (Amend & Wedemeyer (PAbs) or monoclonal antibodies (MAbs). Fig. 1 shows 1970). For instance, for a population size of 1000 salthe general schemes followed by most of the monid fishes with a 10 % incidence of carriers, 27 fish immunological methods. must be examined. However, to increase the probabilThere are no significant differences in the protein ity of detecting viruses during an acute phase, one may content among different VHSV isolates (Basurco & Col1 select moribund fishes and process them separately. 1989a), although a n N, protein has been found in To detect viruses during the carrier phase, fry, fingerVHSV nucleocapsids that was not in similar IHNV ling, ovarian fluid, sperm, or visceral samples may be preparations, which could allow for differential diagcombined into 10-fish pools. Visceral homogenates may nosis in some cases (Basurco et al. 1991). Viral protein more effectively identify virus-carriers, but it is not content is a measure of the number of virions; 1 n g of practical to kill a large number of fish. In the pooling of rhabdoviral protein represents 107 virus particles (Hsu samples mentioned above it is assumed that if one fish in & Leong 1985), but because the particle-to-infectivity the pool contains virus in excess of 103 TCIDSo m]-', the ratio is about 1000:1, that will only mean about 104 virus can be detected. The possibility of detecting the infective viruses. Theoretically the immunological virus will be increased if fewer fish per pool are used (for methods to detect these amounts provide an advantage instance in an individual test), or if more than one carrier over the cell culture methods. Immunological methods is present in the pooled sample. measure both viral proteins, which may or may not be Large numbers of individual homogenates can be incorporated into a viral particle, and viral particles made using 96-well plates (300 p1 well-') with 96 flatwhich may or may not be infective. To date however, tipped rods. The samples loaded in the wells are none of the immunological assays has proven to be homogenized simultaneously by rotating the more sensitive than virus isolation by cell culture homogenizer against the flat-bottomed wells. Other techniques. time-saving aspects of this system include ease of samWorkers have used a wide variety of inoculation ple loading and data recording, storage of samples by programs to obtain PAbs in rabbits for VHSV and freezing and pipetting with multiple-pipettes (FrenchIHNV, but the neutralization titres (reciprocal of last Constant & Devonshire 1987). dilution to give neutralization) obtained have generally AMPLIFICATION OF VIRAL CONTENT OF SAMPLES Once sampling has been optimized the viral content of the homogenate requires amplification by inoculation into cell culture (Wolf & Mann 1980, Lannan et al. 1984, Amos 1985). Currently the most sensitive method for detection of IPNV involves CO-cultivation of cell lines with kidney cells or leucocytes from the test fish (Agius et al. 1982, Ahne & Thomsen 1986). However, to avoid inhibition by the tissue extracts, the tissues should be diluted 100-fold or more (Dixon 1987). New developments are expected in this area due to the possibility of amplification of viral genomes by the PCR (polymerase chain reaction) technique. The viral DNA amplification can take place in a few hours and from a larger volume of sample (since by nucleic acid extracA Solid phase B Fig. 1. Scheme of immunological virus detecting methods. The solid phase can be infected cells (immunofluorescence or immunoperoxidase method), nitrocellulose membranes (~mmunodot) or microtltre wells (EUSA). (A) Indlrect methods: (B) sandwich methods. Y-shaped symbols represent specific Abs (PAbs or MAbs); inverted Y-shaped symbols with dots connected to them represent Ab labeled with fluorescence or w t h peroxldase. The specif~c anti-virus Abs could also be labeled to simplify the method Sanz & Coll: Diagnosis of viral diseases of salmonids 213 ranged from 5 102 to lo4 (Habashi et al. 1975, Hill et al. 1981). In contrast, neutralization titres of 106 are usually obtained against IPNV. Antisera containing neutralizing Abs to salmonid viruses, produced by immunizing rabbits, contain antiviral Abs, but also Abs either to the cell line used to obtain the virus or to the fish tissue components. Attempts to absorb out these Abs have not been entirely successful, although some procedures work well, e.g. in vivo adsorption. Alternatively, 2 different cell lines could be used to obtain the virus and to perform the diagnosis, although this does not always