PDF Outbreak of Viral Hemorrhagic Septicemia (VHS) in Seawater-Farmed Rainbow Trout in Norway

Vol. 85: 93–103, 2009 DISEASES OF AQUATIC ORGANISMS Published June 10 doi: 10.3354/dao02065 Dis Aquat Org...

Vol. 85: 93–103, 2009 DISEASES OF AQUATIC ORGANISMS Published June 10 doi: 10.3354/dao02065 Dis Aquat Org OPEN ACCESS Outbreak of viral haemorrhagic septicaemia (VHS) in seawater-farmed rainbow trout in Norway caused by VHS virus Genotype III O. B. Dale1, I. Ørpetveit1, T. M. Lyngstad1, S. Kahns2, H. F. Skall2, N. J. Olesen2, B. H. Dannevig1,* 1 National Veterinary Institute, PO Box 750 Sentrum, 0106 Oslo, Norway 2 European Community Reference Laboratory for Fish Diseases, National Veterinary Institute, Technical University of Denmark, Hangøvej 2, 8200 Århus N, Denmark ABSTRACT: We describe the finding of a novel viral haemorrhagic septicaemia virus (VHSV) Geno- type III strain that caused disease of both a neurological and septicaemic nature in seawater-farmed rainbow trout Oncorhynchus mykiss in Storfjorden, Norway. In November 2007, an outbreak of VHS associated with slightly elevated mortality was confirmed at a seawater site rearing rainbow trout (90 to 440 g). Within 3 to 4 mo, the disease was recognised in 3 neighbouring sea sites with on- growing rainbow trout. The clinical, gross pathological and histopathological findings were in accor- dance with VHS, and the diagnosis was confirmed by the detection of VHSV in brain and internal tissues by immunohistochemistry, cell culture and reverse transcriptase PCR (RT-PCR). Sequence analysis of the G-gene revealed that the isolated virus clustered with VHSV Genotype III and that the Norwegian isolate represents a unique strain of VHSV. The pathogenicity of the virus strain to rain- bow trout and Atlantic salmon Salmo salar was examined using infection experiments. In immersion trials, the Norwegian isolate produced a cumulative mortality of 70% in rainbow trout, while nearly 100% mortality was obtained after intraperitoneal injection of the virus. For Atlantic salmon, no mortality was observed in immersion trials, whereas 52% mortality was observed after intraperi- toneal injection. The Norwegian isolate thus represents the first VHSV of Genotype III pathogenic to rainbow trout. KEY WORDS: VHS · VHSV · Genotype III · Novirhabdovirus · Rainbow trout · Sea farm Resale or republication not permitted without written consent of the publisher INTRODUCTION The causative agent, VHS virus (VHSV), belongs to the genus Novirhabdovirus of the family Rhabdoviri- Viral haemorrhagic septicaemia (VHS) is a severe dae (Walker et al. 2000). During the last 2 decades, disease historically affecting farmed rainbow trout VHSV has been isolated in wild fish from both fresh- Oncorhynchus mykiss in Europe (Wolf 1988) and is and marine waters (reviewed by Skall et al. 2005) still considered to be one of the most serious viral throughout the temperate part of the northern hemi- diseases in aquaculture. The disease manifestations sphere, and VHS has been recorded in farmed marine of VHS include both a haemorrhagic septicaemia, species such as turbot Scophthalmus maximus (Schlot- and an anaemic and nervous form (Schäperclaus feldt et al. 1991, Ross et al. 1995). In North America and 1979). The infection may persist subclinically in rain- Asia, VHS primarily causes mortality in wild fish (Mey- bow trout once fish are infected, and covert carriers ers et al. 1999, Nishizawa et al. 2002) and a major epi- of cultured or wild fish have been suggested to be demic is ongoing with mass mortality in several wild reservoirs of VHSV (Jørgensen 1982, Enzmann & species in the North American Great Lakes (Groocock Konrad 1985). et al. 2007, Lumsden et al. 2007). There are also reports *Corresponding author. Email: [email protected] © Inter-Research 2009 · www.int-res.com 94 Dis Aquat Org 85: 93–103, 2009 of clinical outbreaks in farmed Japanese flounder capability of adaptation. Therefore, the presence of Paralichthys olivaceus (Isshiki et al. 2001) and reports VHSV in the marine environment is considered to rep- of isolation of VHSV in Atlantic salmon Salmo salar resent a threat to sea-farmed susceptible species. (Traxler et al. 1995, G. Traxler pers. comm., reported in In November 2007, slightly increased mortality and Skall et al. 2005). abnormal swimming patterns were observed in sea- VHSV is enveloped with a single, negative-stranded farmed rainbow trout in western Norway. Based on RNA segment encoding 5 structural proteins and one gross pathological and histopathological examinations, non-structural protein. Phylogenetic studies of the nu- VHS was suspected and the diagnosis verified by cleoprotein (N) and glycoprotein (G) encoding genes of immunohistochemical and virological examinations. In VHSV have identified 4 main genotypes (I to IV) that the present study, the case history and the genetic seem to be more associated with geographical origin characterisation of the causative virus are presented in than with fish species (Snow et al. 1999, 2004, Einer- addition to experimental infection trials with rainbow Jensen et al. 2004, 2005). All freshwater isolates belong trout and Atlantic salmon challenged with isolated to Genotype I or IV whereas seawater isolates are rep- virus. resented in all 4 genotypes. Most VHSV isolated from rainbow trout in European freshwater farms cluster in a subgroup of Genotype I (Ia), while isolates from several MATERIALS AND METHODS marine species, in the Baltic Sea, Kattegat and Skager- rak belong to another subgroup, Ib. Virus isolated from Sampling. For the initial diagnosis, rainbow trout tis- outbreaks of VHS in sea-farmed rainbow trout off the sue samples were collected on 2 occasions and submit- Swedish coast within the seawaters of Kattegat (Nord- ted to the National Veterinary Institute (NVI), Oslo. blom & Norell 2000) were also of Genotype Ib, whereas The first sampling was performed by the local fish the isolates from VHS outbreaks in Finnish sea-reared health service, in connection with an inspection on 15 rainbow trout grouped together in Genotype Id (Raja- November 2007 (7 fish). A follow-up sampling was ini- Halli et al. 2006). This genotype also harbours an old tiated by the Norwegian Food Safety Authority (NFSA) Norwegian isolate (NO-A163.68). The VHSV Genotype on 26 November 2007 (10 fish). At this latter sampling, II has been isolated from herring Clupea harengus, duplicate samples were submitted to the NVI and to sprat Sprattus sprattus and cod Gadus morhua in the the European Community Reference Laboratory (CRL) Baltic Sea, whereas Genotype III has been isolated from for Fish Diseases, Århus, Denmark. Live rainbow trout marine species such as Norway pout Trisopterus es- with behavioural changes were collected from 7 of a markii, haddock Melanogrammus aeglefinus, whiting total of 20 net pens. Samples of gill, heart, liver, pyloric Merlangius merlangus, cod and herring in the North caeca, spleen, kidney and muscle with skin attached Sea and Skagerrak (Skall et al. 2005). VHSV Genotype were collected and transferred to formalin (10% phos- IV seems to be restricted to North America and Asia phate-buffered formalin) and pooled tissue samples and has been isolated from both seawater and freshwa- consisting of kidney, spleen and heart were transferred ter species (Lumsden et al. 2007). to transport medium (Eagle’s Minimum Essential The pathogenicity of the various marine isolates has Medium [EMEM], pH 7.6, supplemented with 10% been evaluated by infection trials, as no suitable newborn bovine serum and 100 µg ml–1 gentamicin). genetic virulence marker has been identified for Brain was only included in the second sampling. Kid- VHSV. Generally, in infection experiments with rain- ney samples were also submitted in RNAlater® in the bow trout VHSV isolated from various marine species first sampling. Subsequent samplings of fish from other may cause no or low mortalities by immersion, while sites were performed in the same manner, but included significant mortality can be obtained following dead fish. All examinations were performed on sam- intraperitoneal injection of the same virus isolates ples from individual fish. (Skall et al. 2004). The Danish, Finnish and Swedish Histopathology and immunohistochemistry. Forma- VHSV isolates from sea-farmed rainbow trout of Geno- lin-fixed tissue samples were processed and embed- type Ia, Id and Ib, respectively, are pathogenic to rain- ded in paraffin wax according to standard procedures. bow trout fingerlings challenged with both immersion Sections (4 to 6 µm) were stained with haematoxylin and intraperitoneal injection (20–70% and 90–100% and eosin (H&E) and examined by light microscopy. mortality, respectively). Sequence analysis of the Immunohistochemical (IHC) examination was per- entire coding regions of the genome indicates that only formed essentially according to Brudeseth et al. 2002 a small number of amino acid residues may be with minor modifications (Taksdal et al. 2007). The involved in the determination of VHSV virulence in monoclonal antibody (MAb) IP5B11 against the VHSV rainbow trout (Betts & Stone 2000). RNA viruses tend nucleoprotein (Lorenzen et al. 1988) was used as the to display very high mutation rates and thereby a high primary antibody and biotinylated rabbit anti-mouse  Dale et al.: VHSV Genotype III in rainbow trout 95 immunoglobulin (Dako) as the secondary antibody ExoSAP-IT® (USB) according to the manufacturer’s before incubation with streptavidin–alkaline–phos- protocol and sequenced using the BigDye® Termina- phatase complex (Amersham Biosciences). The pri- tor v. 3.1 Cycle Sequencing Kit (Applied Biosystems). mary antigen/antibody reaction was visualised by Full length G-gene sequences were amplified, using adding Fast Red salt (Sigma-Aldrich) in Fast Red sub- the GD +/– primer pair primers and sequenced as strate solution. previously described (Einer-Jensen et al. 2004). The Virus isolation and identification. Pooled pieces of cycling conditions were: 50°C for 30 min, 95°C for tissue from each of the sampled fish were homogenized 15 min, followed by 35 cycles of 94°C for 30 s, 55°C for and the samples cleared by low-speed centrifugation 30 s and 68°C for 1 min, and finally 68°C for 7 min. according to routine procedures. The supernatants Nucleotide sequences reported in this paper have were inoculated in 10-fold dilutions onto subconfluent been submitted to the GenBank database and desig- monolayer cell cultures of bluegill Lepomis macro- nated accession numbers EU547740 and AM920657 chirus fry fibroblast (BF-2) (Wolf et al. 1966) and Epithe- correspond to full length and partial sequences of the lioma papulosum cyprini (EPC) (Fijan et al. 1983) in 24 G-gene of the NO-2007-50-385 and NO/650/07 VHSV well tissue culture plates as described in Commision isolates, respectively. Decision 2001/183/EC (Anonymous 2001) and in Lo- Phylogenetic analysis. Multiple sequence alignment renzen et al. (1999). Inoculated cultures were incubated using full length G-gene sequences (1524 nucleotides) at 15°C and inspected regularly with a microscope for was performed using the alignment platform of the CLC the occurrence of cytopathic effect (CPE) at 40× magni- Workbench package (v. 4.0, CLC bio). Publicly available fication. When total CPE was evident, the cell culture sequences from isolates representing all described medium was collected for virus identification by an in- VHSV genotypes were included in the analysis (Einer- direct immunofluorescence test (IFT) or ELISA. Jensen et al. 2004, Elsayed et al. 2006). Duplicate se- Briefly, for IFT, BF-2 and EPC cells grown in 96 well quences were identified manually and excluded from tissue culture plates were inoculated with cell culture the data set. A phylogenetic tree was generated by the supernatants in several dilutions and incubated at Neighbor-Joining distance algorithm (CLC Workbench 15°C for 24 to 48 h. After fixation in 80% acetone, the package, v. 4.0, CLC bio) using the Jukes-Cantor substi- monolayers were incubated with MAb IP5B11 follow- tution model (Jukes & Cantor 1969). The robustness of ing incubation with biotin-labelled goat anti-mouse Ig the tree topology was assessed by using 1000 bootstrap (DAKO) and then a final incubation with fluorescein iterations. Bootstrap values exceeding 70% were consid- isothiocyanate (FITC)-labelled strepatvidin (DAKO). ered to indicate significant relatedness. Appropriate virus positive and negative controls were Infection trials. Infection trials in rainbow trout and included. The cell cultures were examined in an Atlantic salmon were performed to test the patho- inverted epifluorescence microscope. genicity of the Norwegian VHSV strains in these spe- The virus was identified as VHSV by ELISA as cies. The fish were infected by immersion (bath) or by described by Skall et al. 2004. intraperitoneal (i.p.) injection. Reverse trancriptase-PCR (RT-PCR) and sequence Fish: Rainbow trout fingerlings were 10.1 g (SD = analysis. Nucleic acid was extracted from rainbow 3.9 g) in size and were purchased as eyed eggs approx- trout tissue samples with the NucliSens® EasyMAGTM imately 10 mo before infection from a Danish commer- platform for total nucleic acid extraction (BioMérieux) cial fish farm, officially registered free of infectious according to the manufacturer’s instructions. The ini- pancreatic necrosis virus (IPN), infectious haematopoi- tial diagnostic RT-PCR was performed using the QIA- etic necrosis (IHN) and VHS. The fish were hatched GEN OneStep RT-PCR Kit (Qiagen) according to the and grown in the wet laboratory facilities of CRL, manufacturer’s instructions. The primers, V2F (5’-TGG Århus. Atlantic salmon fingerlings were 5.2 g (SD = GAC GAA ACT TTG AGA GG-3’) and V2R (5’-GAT 1.8 g) and were purchased 3 d before infection from CAC AGG GTG GTC AAG G-3’), were designed to a farm breeding progeny of wild caught Atlantic amplify a 323 bp fragment of the G-gene using the salmon for restocking purpose. The salmon fingerlings, Oligo Primer Analysis Software, v. 6.68 (Molecular hatched 9.5 mo before infection, originated from Biology Insights). These primers have been validated parental fish caught in the River Storå in the northwest through the participation of the annual interlaboratory part of Jutland. The parental fish tested negative for profiency tests organized by European Community VHS, IHN, IPN and bacterial kidney disease (BKD). Reference Laboratory (CRL) for Fish Diseases, Århus, Tanks: The tanks contained 8 l of soft freshwater (2.5 Denmark. The cycling conditions were: 50°C for to 4.5°H) and the aerated water temperature was main- 30 min, 95°C for 15 min, followed by 40 cycles of 94°C tained at 9.6°C (SD = 0.6°C). The water used was for 1 min, 55°C for 1 min and 72°C for 1 min, and finally groundwater of drinking quality (not chlorinated), and 72°C for 10 min. The PCR products were purified with new water was added continuously with no recircula- 96 Dis Aquat Org 85: 93–103, 2009 tion. The tanks were kept closed with a transparent lid RESULTS to prevent contamination by aerosols and prevent fish from jumping out. The outbreak Virus isolates: The virus isolate NO-2007-50-385 was passaged once in BF-2 cells and titrated on the The outbreak of VHS was confirmed by NFSA on 27 same cell type according to standard procedures November 2007 at a seawater site rearing rainbow (Lorenzen et al. 1988). trout located in Storfjorden in the county of Møre and Positive and negative controls: The highly pathogenic Romsdal, on the west coast of Norway (see Fig. 4). VHS VHSV strain DK-3592B (Lorenzen et al. 1993) isolated was later confirmed at 3 additional neighbouring sites from a rainbow trout farm was used as a positive control with on-growing rainbow trout located within a dis- by immersion only. As a negative control, EMEM with tance of 5 to 7 km from the primary outbreak site (diag- tris-buffer and 10% newborn calf serum (dilution noses confirmed by NFSA on 29 January 2008, 1 Feb- medium) was likewise used by immersion only. ruary 2008 and 8 February 2008, respectively). All sites Trials: Each group of rainbow trout, both positive were operated by the same company. and negative control fish and test fish, were tested in By November 2007 there were approximately 1.8 triplicate with 31 fish in each tank. The Atlantic salmon million fish at the primary outbreak site weighing from immersed in virus were tested in duplicate with 75 fish 90 to 440g. The fish were kept in 20 net pens. Monthly in each tank. Atlantic salmon injected with virus and mortality at sea site level was reported as 1.7% in the negative control fish were only tested in single November and 1.8% in December 2007 (Aquaculture tanks, with 35 fish in each. Dead fish were removed register 2007, Directorate of Fisheries). daily from each tank. The number of dead fish and Before the outbreak was verified, fish were repeat- clinical symptoms of VHS in fingerlings were recorded edly graded on the primary site and moved to 3 neigh- daily. At the end of the trial, surviving fish were eutha- bouring sites where VHS was later confirmed. The nized with an overdose of benzocaine and counted. rainbow trout were fed commercial dry pelleted feed. The sum of surviving fish and fish that died during the Dead fish (rainbow trout together with cod and saithe trial equalled the number of fish in each tank. The Pollachius virens from other sites in the same fjord) trials were terminated after 28 d. were stored in land-based containers close to the pri- Immersion trial: To obtain a concentration of ap- mary site. proximately 105 TCID50 ml–1 water, virus isolates were An eradication zone was established and fish from added to a vial containing 10 ml dilution medium the primary outbreak site and the 3 neighbouring sites according to their respective titres. Water supply was were removed and disposed of by the end of April stopped while virus dilutions were added to each tank, 2008. resulting in an immersion exposure time of 2 h, after which the continuous water flow was resumed. Intra-peritoneal injection trials: Virus isolates were Clinical signs and gross pathology mixed in vials with 10 ml dilution medium to obtain a dose of approximately 105 TCID50 per fish in an injection In November 2007 the farmer alerted the local fish volume of 50 µl. Fish were anaesthetised by bathing health service as behavioural changes and slightly ele- them in benzocaine solution, injected intraperitoneally vated mortality had been observed in the fish. On with a new needle and syringe for each tank group and inspection ataxic fish swimming in circles or spirals in transferred to tanks supplied with running fresh water. the surface water, occasionally with the appearance of Virus isolation: Fish that died during the trials and cramps, were seen in all 20 net pens. Notably the num- survivors at the end of the trials were sampled for virus ber of ataxic fish could vary from hour to hour. By isolation. Negative and positive control groups were necropsy of 7 ataxic fish with no external lesions, 2 fish also sampled. Fish were frozen at –25°C for later exam- were found to have pale gills, severe anaemia (haema- ination. In each tank the first 3 fish that died after tocrit < 10), pale and grey-orange livers and swollen infection and approximately 10 fish that survived to the dark spleens, and 1 of these 2 fish had extensive haem- end of the trial were individually examined on cell cul- orrhages in the peritoneum covering the pyloric caeca. tures as described in ‘Virus isolation and identification’. Among the other 5 fish, 1 had a very dark red liver; Collection of epidemiological data. The epidemio- otherwise there were no findings except for evidence logical data from the outbreak site were gathered by of lack of feed in the gastrointestinal tract in all 7 fish. the local district offices of the NFSA by means of a pre- On the follow-up sampling of 10 ataxic fish with no vis- designed form together with inspection reports from ible external lesions, the internal findings were limited the NFSA. The form focused on variables of fish health, to swollen spleens and no feed in the gastrointestinal fish stock and management. tract. These observations had not been observed in  Dale et al.: VHSV Genotype III in rainbow trout 97 previous disease investigations at the farm. The later Virological examinations samplings at the 3 affected neighbouring sites in- cluded both live ataxic fish and dead fish. Similar CPE was observed in cell cultures within 2 to 4 d of pathology was found, though some dead fish had more incubation after inoculation with tissue homogenates extensive haemorrhages. Furthermore, VHS could be from 2 out of 7 rainbow trout from the first sampling verified up to time of slaughter in April 2008. and from all 10 fish from the second sampling of the primary outbreak. All samples with CPE tested posi- tive for VHSV in IFT. Histopathology and immunohistochemistry All fish that tested positive in cell cultures also tested positive for VHSV by RT-PCR. Furthermore, sequenc- The main histopathological findings corresponding ing of a 286 bp PCR-product (Isolate NO/650/07, Gen- to the haemorrhagic lesions on necropsy were hepati- Bank Accession number: AM920657) revealed a 98% tis with multifocal, sometimes haemorrhagic, necrosis similarity with the glycoprotein (G) gene of VHSV from in the liver, endocarditis and haematopoietic necrosis herring, isolate DK-4p168 (Einer-Jensen et al. 2005). in kidney and spleen. IHC showed staining of VHSV to The results of the virological examinations from the a variable degree, from sparse staining of endothelial second sampling in November 2007 were confirmed by cells and macrophages in the spleen and kidney in the CRL in Århus. The virus isolated at the CRL was rainbow trout sampled live to extensive staining of designated NO-2007-50-385. necrotic haematopoietic tissues, especially in kidney Similar results were obtained from samplings from and spleen in fish sampled dead (Fig. 1). In the fish the subsequent disease outbreaks at the 3 neighbour- with only nervous signs, changes in the visceral organs ing farms, and the 286 bp sequence of the G-gene from were limited to accumulation of blood pigment (hae- these outbreaks was identical to that of the primary mosiderosis) in the spleen and occasionally hepatitis. outbreak. The main finding in these fish was meningoencephali- tis and the cellular infiltrate of the meninges was mononuclear. In the brain of fish sampled live, small Phylogenetic analysis haemorrhages in conjunction with necrotic foci were found. In dead fish with the same necropsy findings, To determine the phylogenetic relationship of the more severe encephalitis with spongy change, neu- NO-2007-50-385 isolate to other VHSV isolates, the ronal degeneration and gliosis was found. By IHC, full length G-gene was sequenced (GenBank Acces- strong staining was evident in the necrotic lesions. In sion number: EU547740). A phylogenetic analysis was addition, extensive staining of neurons from dead fish performed using 34 reported full length G-gene was observed, but this occurred only occasionally in sequences from isolates representing all known VHSV samples from live fish (Fig. 2). genotypes: Ia to Ie, II, III and IV (Fig. 3). The analysis Fig. 1. Oncorhynchus mykiss. Pathology of haemorrhagic manifestation of VHS caused by VHSV Genotype III in rainbow trout. (a) Kidney, H&E-stained section of necrotic haematopoietic tissue in between intact kidney tubules; some intact haematopoietic tissue is evident in lower left corner. (b) Kidney, immunohistochemistry (IHC) showing virus labelling (red colour) of necrotic haematopoietic tissue in between intact kidney tubules. (c) Spleen, immunohistochemistry showing virus labelling (red colour) of necrosis in spleen. Scale bar = 50 µm (applies to a–c) 98 Dis Aquat Org 85: 93–103, 2009 Fig. 2. Oncorhynchus mykiss. Pathology of nervous manifestation VHS caused by VHSV Genotype III in rainbow trout. (a) H&E- stained section showing extensive inflammation of meninges (scale bar = 50 µm). (b) H&E-stained section showing focal haemor- rhage and gliosis in the vicinity of a large neuron in the brain stem (scale bar = 50 µm). (c) Immunohistochemistry (IHC) of the same area as in (b) showing virus labelling (red colour) of a necrosis (scale bar = 50 µm). (d) IHC of transverse section showing extensive virus nucleoprotein labelling (red colour) (scale bar = 1 µm). (e) IHC of optic lobe showing extensive infection of neurons (red colour) (scale bar = 100 µm). (f) IHC of a section from the cerebellum showing virus labelling of a Purkinje cell (red colour) (scale bar = 10 µm) Fig. 3. Classification of the NO-2007-50-385 isolate as a Genotype III isolate. The phylogenetic tree was generated using 35 unique full length G-gene nucleotide sequences including the sequence of the NO-2007-50-385 isolate using the Neighbor- Joining distance algorithm. Bootstrap values have been obtained from 1000 reassembled data sets. Only bootstrap values above 70% were considered to be significant and are indicated at the nodes. Genotypes (Ia–e, II, III, IV) classified according to Einer- Jensen et al. (2004) are indicated with right square brackets. An arrow indicates the location of the unique Norwegian isolate NO-2007-50-385 with the Genotype III isolates  Dale et al.: VHSV Genotype III in rainbow trout 99 identified the NO-2007-50-385 isolate as a Genotype III isolate. To determine which of the Genotype III isolates is most similar to the NO-2007-50-385 isolate, we compared all Genotype III full length G-gene sequences. The closest relation- ship was observed between NO-2007- 50-385 and the DK-4p168 isolate (from herring, Skagerrak) with 10 nucleotide differences in the G-gene sequences. However, close relationships were also observed with the UK-H17/5/93 (from cod), UK-MLA98/6 PT11 (from Norway pout) and UK-H17/2/95 (from haddock), all from the North Sea (Table 1). The nucleotide sequence of the NO-2007-50- 385 isolate showed least similarity with UK-860/94 isolate (from turbot, western Scotland) and with isolates from fish caught relatively far away from the out- break sites (Fig. 4). Infection trials Moribund infected fish showed clinical Fig. 4. Localisation of Genotype III isolates. Isolate 1 in white box is the Nor- wegian VHS outbreak isolate NO-2007-50-385. Isolates 2 to 5 in grey boxes signs typical of VHS, e.g. exophthalmia, are isolates with 10 to 13 nucleotide differences from NO-2007-50-385. haemorrhages and ascites. Isolates 6, 7 and 9 in black boxes are isolates with ≥18 nucleotide differences In the immersion trial of rainbow trout, the NO-2007-50-385 isolate produced cu- mulative mortalities of 62, 71 and 77%, with an average the positive control rainbow trout had a cumulative of 70% (Fig. 5). By i.p. injection in rainbow trout the cu- mortality of 100% (Fig. 5). No mortality was observed mulative mortality was 100, 97 and 100%. In the im- in the negative control group of Atlantic salmon. Since mersion trial of Atlantic salmon the isolate produced no there are no known VHSV isolates highly pathogenic mortality. However, by i.p. injection in Atlantic salmon to Atlantic salmon, a positive control group for this spe- the cumulative mortality was 52% (Fig. 6). cies could not be included The negative control rainbow trout had a cumulative Some of the infected fish that died during the test mortality of 3% (1 dead fish at the end of the trial) and were examined virologically, and VHSV was isolated Table 1. Differences between the NO-2007-50-385 isolate and other Genotype III VHSV isolates. Genotype III isolate names are according to Einer-Jensen et al. (2004). Numbers of nucleotide differences of full length G-gene sequences relative to NO-2007- 50-385 are shown. Numbers 1 to 9 (left column) and the corresponding International Council for the Exploration of the Sea (ICES) numbers refer to the numbered squares in Fig. 4 Isolate name Place or origin ICES number Nucleotide differences in full length G-gene (1520 bp) 1 NO-2008-50-385 Norway 53F6 – 2 DK-4p168 Skagerrak, Norway 44G0 10 3 UK-H17/5/93 North Sea, E Shetland, UK 49F0 11 4 UK-MLA98/6PT11 North Sea 48F1 12 5 UK-H17/2/95 North Sea, E Shetland, UK 50F1 13 6 IR-F13.02.97 Cape Clear, Ireland 31E0 18 7 DK-4p101 Skagerrak, Norway 43F7 26 8 FR-L59X River Loire, France – 33 9 UK-860/94 Gigha, W Scotland, UK 40E4 47  100 Dis Aquat Org 85: 93–103, 2009 100 fection trial was sequenced and there was 100% identity 90 with the NO-2007-50-385 strain used in the trials. Cumulative mortality (%) 80 70 DISCUSSION 60 In this paper we describe an outbreak of VHS in sea- 50 farmed rainbow trout caused by VSHV belonging to 40 Genotype III. The outbreak occurred on the west coast 30 of Norway in 2007 (Fig. 4) and is the first record of VHS in a salmonid species caused by this genotype. The 20 VHSV Genotype III has hitherto only caused outbreaks 10 of VHS in marine species, i.e. farmed turbot (Ross et al. 0 1995), although it has been isolated from a variety of 0 5 10 15 20 25 30 wild marine, non-salmonid, fish species (Einer-Jensen Number of days after infection et al. 2004, Snow et al. 2004, Skall et al. 2005). Norway Positive control NO-2007-50-385 NO-2007-50-385 Negative control has not experienced outbreaks of VHS in rainbow trout (DK-3592B) immersion i.p. injection (dilution medium) since 1974, which at that time were caused by VHSV of Genotype Id (isolate NO-A163-68 EG46 in Fig. 3). Fig. 5. Oncorhynchus mykiss. Mortality in rainbow trout after The present disease outbreak in Storfjorden, Nor- infection with VHSV NO-2007-50-385 by immersion and in- traperitoneal (i.p.) injection. Negative and positive controls way, started with a slightly elevated mortality and a are by immersion only. Each group was kept in separate tanks widespread ataxia shifting rapidly in extent. This pat- during and after infection tern had a striking resemblance to the descriptions of outbreaks in the early history of VHS. The mortality in from all these. VHSV was isolated from 36% of the sur- VHS outbreaks can be very variable, and a wide range viving rainbow trout after infection by immersion, of disease signs have been attributed to several forms whereas no VHSV was isolated from the surviving of the disease (Schäperclaus 1979). Affected fish corre- Atlantic salmon after immersion. Of 12 surviving sponding to both the acute haemorrhagic form and the Atlantic salmon tested, 3 (25%) carried the virus after chronic nervous and anaemic form were observed. i.p. injection (Table 2). Pathological findings including internal haemorrhages The full length G-gene of VHSV from a single rainbow and necrosis of haematopoietic, but not of excretory trout and single Atlantic salmon that died during the in- kidney tissue, were consistent with the haemorrhagic form of VHS. In ataxic fish without haemorrhagic 100 lesions we found mononuclear meningoencephalitis 90 typical for viral infection. The immunohistochemistry examinations demonstrated the presence of VHSV Cumulative mortality (%) 80 directly in the lesions, thus providing evidence for an 70 etiological role of the isolated virus. 60 In a population of several million fish, as in this farm 50 system, the occurrence of several infections and dis- eases must be expected. Low levels of IPNV were 40 detected in 4 fish, but no other infections were recog- 30 nised (results not shown). As no pathological lesions 20 corresponding to IPN were found, this finding proba- bly reflects the high prevalence of fish in salmonid 10 aquaculture in Norway covertly infected with IPNV. 0 0 5 10 15 20 25 30 During the winter, skin ulcers especially due to Time post-infection (d) Moritella viscosa developed in fish at the on-growing NO-2007-50-385 NO-2007-50-385 Negative control farms (results not shown). As VHS causes anaemia and i.p. injection immersion (dilution medium) leukopenia (Wolf 1988), immunodeficiency causing Fig. 6. Salmo salar. Mortality in Atlantic salmon after infection increased problems with other diseases can also be an with VHSV NO-2007-50-385 by immersion and intraperi- effect of VHS. In the recent VHS outbreaks at Åland, toneal (i.p.) injection. Positive control is not included as no highly pathogenic VHSV isolate is available. Negative control Finland, flexibacterial problems appear to have is by immersion only. Each group was kept in separate tanks increased in the presence of VHS (P. Vennerström, during and after infection EVIRA, pers. comm.).  Dale et al.: VHSV Genotype III in rainbow trout 101 Table 2. Virological examination by cell cultivation followed by identification of virus by ELISA of rainbow trout and Atlantic salmon included in infection trials with VHSV isolate NO-2007-50-385. i.p. = intraperitoneal Group Comments/Observations VHSV positive fish/no. of fish tested Rainbow trout, immersion Fish dying early in infection trial 8/8 Rainbow trout, immersion Surviving trout 29 d after infection 10/28 Atlantic salmon, immersion Surviving salmon 29 d after immersion 0/20 Atlantic salmon, i.p. injection Fish dying early in infection trial 3/3 Atlantic salmon, i.p. injection Surviving salmon 29 d after injection 3/12 VHSV from all the 4 sites described in this report monthly mortality at site level at the time of the pri- were identical in the 286 nucleotides (nt) part of the G- mary outbreak may be regarded as moderate (approx- gene, indicating that the VHSV from all these sites imately 1.7 to 1.8%), whereas the observed mortality originated from a common source. Furthermore, the in rainbow trout bath-infected with the causative virus phylogenetic analysis of the full-length G-gene se- was approximately 70% (Fig. 5) However, due to the quence of isolate NO-2007-50-385 verified that the uncertainty about the reported mortality during the Norwegian VHSV isolate clustered with Genotype III, outbreak, it was difficult to compare the mortality and that the isolate represents a unique isolate not pre- observed in the field with the infection trial. viously described. The source of VHS infection in the Norwegian In experimental infection trials, the Norwegian rainbow trout farms is not known, but due to the VHSV Genotype III isolate induced a mortality in rain- genotype of the causative virus, together with the bow trout of approximately 70% when infected by fact that the highest sequence similarity is observed immersion, which is at the same level as the mortality towards VHSV isolates from northern European in similar infection trials (range, 46 to 89%) using waters, a marine source is most likely. For the out- Genotype Ia VHSV isolates from outbreaks in freshwa- breaks of VHS in sea-farmed rainbow trout in ter rainbow trout farms (Skall et al. 2004). When Sweden (Nordblom & Novell 2000), the isolated infected by i.p. injection, the Norwegian isolate pro- virus (Genotype Ib) was almost identical to isolates duced nearly 100% mortality. In comparison, VHSV found in the adjacent marine environment (Einer- Genotype III isolates from farmed turbot and wild Jensen et al. 2004), indicating a direct transmission marine fish resulted in maximum 3% mortality in rain- of virus from species in the surrounding waters. The bow trout when infected by immersion, and maximum outbreaks of VHS in Finnish rainbow trout farms in 20% when i.p. infected (Skall et al. 2004). Thus, the the Baltic Sea occurred in brackish water. Although Norwegian isolate represents a rainbow trout patho- a definite origin of the outbreaks was not identified, genic strain of VHSV Genotype III not previously because the causative virus was not genetically reported. Sequencing of the full length G-gene of the related to those detected in marine species from re-isolated virus from the experimentally infected rain- this area, it was still suggested that wild fish were bow trout and Atlantic salmon revealed 100% identity the source of the primary infection (Raja-Halli et with the G-gene sequence of the virus isolate NO- al. 2006). 2007-50-385 used for infection. Since VHSV was not Screening of herring and sprat caught at the outlet of detected in the test fish before the infection trials, this Storfjorden resulted in detection of VHSV in 2 out of confirms that the high mortality in the rainbow trout 120 herrings examined. However, these VHSV isolates was caused by the new Norwegian VHSV isolate. A were reported to cluster with VHSV Genotype Ib (A. recent in vitro study shows that a VHSV strain highly Nylund pers. comm.). Farms raising cod and saithe are pathogenic to rainbow trout translocated rapidly located in the same fjord system as the rainbow trout through polarized gill epithelial cell cultures compared farms with VHS. In the case of saithe, wild fish are with a low pathogenic strain (Brudeseth et al. 2008). caught and kept in net pens for further growing. This may explain the differences in mortality between Farmed cod and saithe are both fed raw processed fish infection by i.p. injection and immersion, and indicates of marine origin. All non-affected fish farms in the fjord that translocation through the gills may be the step that (Storfjorden), including both marine and salmonid limits pathogenicity for successful infection in various species, were examined for the presence of VHSV by salmonid species. RT-PCR and cell culture with negative results (results Standardized infection trials are important for evalu- not shown). The latest sampling for the Norwegian ation of the pathogenicity of VHSV. The reported VHS/IHN surveillance program from the outbreak 102 Dis Aquat Org 85: 93–103, 2009 sites was in April 2006 with negative results. Even LITERATURE CITED though VHSV was not detected in the marine farms, Anonymous (2001) 2001/183/EC: commission decision of 22 the prevalence could be lower than that needed for February 2001 laying down the sampling plans and diag- virus detection by examination of 30 fish from each nostic methods for the detection and confirmation of cer- site. According to information from NFSA, the rainbow tain fish diseases and repealing Decision 92/532/EEC. Off trout were fed commercial dried pelleted feed that had J Eur Communities L 67:65–76 been processed in high temperatures and under condi- ➤ Betts AM, Stone DM (2000) Nucleotide sequence analysis of the entire coding regions of virulent and avirulent strains tions that would kill microorganisms. of viral haemorrhagic septicaemia virus. Virus Genes 20: The fish on the primary outbreak site were not 259–262 removed and disposed of before 5 mo after the VHS ➤ Brudeseth BE, Castric J, Evensen O (2002) Studies on patho- diagnosis was confirmed. Within this time, the dis- genesis following single and double infection with viral hemorrhagic septicemia virus and infectious hematopoi- ease was detected on 3 neighbouring rainbow trout- etic necrosis virus in rainbow trout (Oncorhynchus farming sites operated by the same owner. The out- mykiss). Vet Pathol 39:180–189 breaks of VHS in Storfjorden occurred within a ➤ Brudeseth BE, Skall HF, Evensen O (2008) Differences in vir- limited time period. It is therefore difficult to deter- ulence of marine and freshwater isolates of viral hemor- mine whether the virus was introduced by a single rhagic septicemia virus in vivo correlate with in vitro abil- ity to infect gill epithelial cells and macrophages of incidence into the primary outbreak site and there- rainbow trout (Oncorhynchus mykiss). J Virol 82: after spread to the other farms, or whether a parallel 10359–10365 introduction to the other rainbow trout farms in ques- ➤ Einer-Jensen K, Ahrens P, Forsberg R, Lorenzen N (2004) tion had occurred. Evolution of the fish rhabdovirus viral haemorrhagic sep- ticaemia virus. J Gen Virol 85:1167–1179 Whether the virus causing the present outbreak of ➤ Einer-Jensen K, Ahrens P, Lorenzen N (2005) Parallel phylo- VHS was transmitted directly from a possible marine genetic analyses using the N, G or Nv gene from a fixed source or whether an adaptation to rainbow trout group of VHSV isolates reveal the same overall genetic had occurred before disease outbreak is not known. typing. Dis Aquat Org 67:39–45 Although marine VHSV isolates are not, in general, ➤ Elsayed E, Faisal M, Thomas M, Whelan G, Batts W, Winton J (2006) Isolation of viral haemorrhagic septicaemia virus associated with disease in rainbow trout (Skall et al. from muskellunge, Esox masquinongy (Mitchill), in Lake 2004), the present study has demonstrated that VHSV St Clair, Michigan, USA reveals a new sublineage of the of Genotype III can be harmful to farmed rainbow North American genotype. J Fish Dis 29:611–619 trout. In continental European aquaculture, VHS is still Enzmann PJ, Konrad M (1985) Inapparent infections of brown trout with VHS-virus. Bull Eur Assoc Fish Pathol 5:81–83 the most serious fish disease. Mounting evidence sug- Fijan N, Sulimanovic D, Bearzotti M, Muzinic D and others gests that all VHS viruses can develop high patho- (1983) Some properties of the Epithelioma papulosum genicity in many different fish species given time and cyprini (EPC) cell line from carp (Cyprinus carpio). Ann opportunity. The introduction of VHSV Genotype IV to Virol Inst Pasteur 134:207–220 the Great Lakes appears to be a worst case scenario in ➤ Groocock GH, Getchell RG, Wooster GA, Britt KL and others (2007) Detection of viral hemorrhagic septicemia in round this respect. Since VHSV belongs to a group of RNA gobies in New York State (USA) waters of Lake Ontario viruses showing high adaptability to diverse fish spe- and the St. Lawrence River. Dis Aquat Org 76:187–192 cies, the presence of VHSV in Norwegian rainbow ➤ Isshiki T, Nishizawa T, Kobayashi T, Nagano T, Miyazaki T trout aquaculture may represent a threat to the (2001) An outbreak of VHSV (viral hemorrhagic sep- ticemia virus) infection in farmed Japanese flounder Par- Atlantic salmon farming industry. alichthys olivaceus in Japan. Dis Aquat Org 47:87–99 In conclusion, the Norwegian VHSV Genotype III, Jørgensen PEV (1982) Egtved virus: occurrence of inapparent ➤ NO-2007-50-385, is a new contagious strain that infections with virulent virus in free-living rainbow trout, caused outbreak of VHS in both sea-farmed rainbow Salmo gairdneri Richardson, at low temperatures. J Fish trout and in experimentally infected rainbow trout and Dis 5:251–255 Jukes T, Cantor C (1969) Evolution of protein molecules. In: Atlantic salmon fingerlings. The source of infection has Munro HN (ed) Mammalian protein metabolism, Vol III. not been identified, but virus originating from wild or Academic Press, New York, p 21–32 farmed marine species is most likely. ➤ Lorenzen N, Olesen NJ, Jørgensen PEV (1988) Production and characterization of monoclonal antibodies to four Egtved virus structural proteins. Dis Aquat Org 4:35–42 Acknowledgements. L. B. Rønneberg at the fish health ser- ➤ Lorenzen N, Olesen NJ, Jørgensen PEV (1993) Antibody vice Fiske-Liv AS, Ålesund, Norway, is greatly acknowledged response to VHS virus proteins in rainbow trout. Fish for providing information of clinical signs and gross pathology Shellfish Immunol 3:461–473 of the affected fish. The author thanks M. J. Hjortaas for her ➤ Lorenzen E, Carstensen B, Olesen NJ (1999) Inter-laboratory contribution to the design of primers. The skilled technical comparison of cell lines for susceptibility to three 3 assistance of the staff at the Section of Virology and Serology viruses: VHSV, IHNV and IPNV. Dis Aquat Org 37:81–88 and the Section of Fish Health, NVI, Norway, and at CRL ➤ Lumsden JS, Morrison B, Yason C, Russell S and others (2007) Århus, is greatly appreciated. We also thank Dr. H. Sindre for Mortality event in freshwater drum Aplodinotus grun- critically reading of the manuscript. niens from Lake Ontario, Canada, associated with viral  Dale et al.: VHSV Genotype III in rainbow trout 103 haemorrhagic septicemia virus, type IV. Dis Aquat Org European marine and farmed fishes. Dis Aquat Org 58: 76:99–111 99–110 ➤ Meyers TR, Short S, Lipson K (1999) Isolation of the North ➤ Skall HF, Olesen NJ, Mellergaard S (2005) Viral haemor- American strain of viral hemorrhagic septicemia virus rhagic septicaemia virus in marine fish and its implications (VHSV) associated with epizootic mortality in two new host for fish farming: a review. J Fish Dis 28:509–529 species of Alaskan marine fish. Dis Aquat Org 38:81–86 ➤ Snow M, Cunningham CO, Melvin WT, Kurath G (1999) ➤ Nishizawa T, Iida H, Takano R, Isshiki T, Nakajima K, Analysis of the nucleoprotein gene identifies distinct lin- Muroga K (2002) Genetic relatedness among Japanese, eages of viral haemorrhagic septicaemia virus within the American and European isolates of viral hemorrhagic sep- European marine environment. Virus Res 63:35–44 ticemia virus (VHSV) based on partial G and P genes. Dis ➤ Snow M, Bain N, Black J, Taupin V and others (2004) Genetic Aquat Org 48:143–148 population structure of marine viral haemorrhagic septi- Nordblom B, Norell AW (2000) Report on an outbreak of VHS caemia virus (VHSV). Dis Aquat Org 61:11–21 (viral hemorrhagic septicaemia) in farmed fish in Sweden. ➤ Taksdal T, Olsen AB, Bjerkas I, Hjortaas MJ, Dannevig BH, Report for the Standing Veterinary Committee. Swedish Graham DA, McLoughlin MF (2007) Pancreas disease in Board of Agriculture, Department for Animal Production farmed Atlantic salmon, Salmo salar L., and rainbow trout, and Health, Jönköping Oncorhynchus mykiss (Walbaum), in Norway. J Fish Dis ➤ Raja-Halli M, Vehmas TK, Rimaila-Parnanen E, Sainmaa S, 30:545–558 Skall HF, Olesen NJ, Tapiovaara H (2006) Viral haemor- Traxler G, Kieser D, Evelyn TPT (1995) Isolation of North rhagic septicaemia (VHS) outbreaks in Finnish rainbow American strain of VHS virus from farmed Atlantic trout farms. Dis Aquat Org 72:201–211 salmon. In: Margolis L (ed) Aquaculture update no. 72. Ross K, McCarthy U, Huntly PJ, Wood BP and others (1995) A Fisheries and Oceans Canada, Aquaculture Division, outbreak of viral haemorragic septicaemia (VHS) in turbot Pacific Biological Station, Nanaimo (Scophthalmus maximus) in Scotland. Bull Eur Assoc Fish Walker PJ, Benmansour A, Dietzgen R, Fang RX and others Pathol 14:213–214 (2000) Family Rhabdoviridae. In: Van Regenmortel MHV, Schäperclaus W (1979) Fischkrankheiten 4. Auflage. Akade- Fauquet CM, Bishop DHL, Carstens EB and others (eds) mie-Verlag, Berlin Virus taxonomy. Classification and nomenclature of Schlotfeldt HJ, Ahne W, Jørgensen PEV, Glende W (1991) viruses. Academic Press, San Diego, CA, p 563–583 Occurrence of viral haemorrhagic septicaemia in turbot Wolf K (1988) Viral hemorrhagic septicemia. In: Wolf K (ed) (Scophthalmus maximus) — a natural outbreak. Bull Eur Fish viruses and fish viral diseases. Cornell University Assoc Fish Pathol 11:105–107 Press, Ithaca, NY, p 217–249 ➤ Skall HF, Slierendrecht WJ, King JA, Olesen NJ (2004) Exper- ➤ Wolf K, Gravell M, Malsberger RG (1966) Lymphocystis virus: imental infection of rainbow trout Oncorhynchus mykiss isolation and propagation in centrarchid fish cell lines. Sci- with viral haemorrhagic septicaemia virus isolates from ence 151:1004–1005 Editorial responsibility: Mark Crane, Submitted: December 17, 2008; Accepted: April 7, 2009 Geelong, Victoria, Australia Proofs received from author(s): May 29, 2009