Susceptibility and immune responses following ... - Springer Link

Arch Virol (2005) 150: 2195–2216 DOI 10.1007/s00705-005-0588-8

Susceptibility and immune responses following experimental infection of MHC compatible Atlantic salmon (Salmo salar L.) with different infectious salmon anaemia virus isolates S. Mjaaland1 , T. Markussen1 , H. Sindre3 , S. Kjøglum2,5 , B. H. Dannevig3 , S. Larsen4 , and U. Grimholt2 1 Department

of Food Safety and Infection Biology, Norwegian School of Veterinary Science, Oslo, Norway 2 Department of Basic Science and Aquaculture, Norwegian School of Veterinary Science, Oslo, Norway 3 National Veterinary Institute, Oslo, Norway 4 Department of Production Animal Medicine, Norwegian School of Veterinary Science, Epidemiological Section, Oslo, Norway 5Aqua Gen AS, Kyrkseterøra, Norway Received January 25, 2005; accepted May 13, 2005 c Springer-Verlag 2005 Published online July 14, 2005

Summary. Infectious salmon anaemia virus (ISAV) is an aquatic orthomyxovirus causing a multisystemic disease in farmed Atlantic salmon (Salmo salar) where disease development, clinical signs, and histopathology vary to a large extent. Here, an experimental trial was designed to determine the effect of variation in viral genes on virus-host interactions, as measured by disease susceptibility and immune responses. The fish were infected using cohabitant transmission, representing a natural route of infection. Variation caused by host factors was minimized using MHC compatible A. salmon half-siblings as experimental fish. Virus isolates were selected according to HE genotype, as European ISAV isolates can be genotyped according to deletion patterns in their hemagglutinin-esterase (HE) surface glycoprotein, and the course of disease they typically induce, classified as acute versus protracted. The different ISAV isolates induced large variations in death prevalence, ranging from 0–47% in the test-group and 3–75% in the cohabitant fish. The use of MHC compatible experimental fish made it possible to determine the relative contribution of humoral versus cellular response in protection against ISA. Ability to induce a strong proliferative response correlated with survival and virus clearance, while induction of a humoral response was less protective.

2196

S. Mjaaland et al.

Abbreviations: Infectious salmon anaemia (ISA), infectious salmon anaemia virus (ISAV), hemagglutinin-esterase (HE), highly polymorphic region (HPR), major histocompatibility complex (MHC)

Introduction Infectious salmon anaemia virus (ISAV) is an aquatic orthomyxovirus causing a multisystemic disease in farmed Atlantic salmon (Salmo salar L.) [60, 11, 59]. The disease is economically important, as mortalities range from 15–100% in affected farms. The main targets are endothelial [27] and leukocytic cells [14]. As for influenza A and B virus, the ISAV genome consists of eight protein-coding negative single-stranded RNA segments [42, 5]. Field outbreaks of infectious salmon anaemia (ISA) vary considerably in disease development, clinical signs and histopathology. Traditionally, disease outbreaks have been roughly categorized into two main forms: an acute form, with a rapid disease development and high mortality, or a more chronic, protracted disease form, where a slow increase in mortality can be observed during several months. This categorization is, however, a simplification, and transitional forms are often present [44, 4, 38]. The viral genes associated with influenza virus pathogenicity can vary depending on virus strain and host, as virulence is a polygenic trait depending on optimal gene constellations [52, 65]. The crucial role of surface glycoproteins in virus binding and release suggests an important role in virus virulence. In fact, the hemagglutinin (HA) gene of influenza A virus is known to be highly variable, and, at least in avian strains, plays a pivotal role in determining the severity of infection [33, 18, 56]. Similarly, the influenza C virus surface protein HEF (hemagglutinin, esterase, fusion) is known to be a major determinant of cell tropism [24]. Recently, a central role of the second major viral surface protein, neuraminidase, in the virulence of the human influenza A isolate A/WSN/33 was demonstrated [19]. Domains on the nucleoprotein (NP), polymerase (PB2), non-structural (NS1) and matrix (M2) proteins may also be associated with pathogenicity [22, 17, 54, 67]. Recently, the ISAV surface glycoprotein hemagglutinin, coded from genomic segment 6, was identified and described [35, 51]. This is the ISAV genomic segment with the highest variability, although nucleotide and amino acid sequence alignments of the full-length ISAV HA open reading frame shows pair wise nucleotide identities in the range of 94.8–97.2%, and amino acid similarities between 96.5–97.6%, as compared to the Glesvaer strain [41]. In contrast to influenza A HA gene however, where most of the variability is located in distal parts of the molecule, most of the variability within ISAV HA is concentrated in a region predicted to lie immediately outside the viral envelope [35, 51, 9, 41]. This region is characterized by the presence of gaps, rather than single-nucleotide substitutions [41]. The full-length gene was suggested to represent an ancient, avirulent variant, where subsequent variability could be generated through differential deletions as a consequence of strong functional selection, possibly related to a recent or ongoing crossing of a species barrier [41]. The presence

ISAV: virus-host interactions in MHC compatible fish

2197

of such a genotype, termed HPR0, was later confirmed in wild and farmed Atlantic salmon [7, 6]. Until today, it has not been possible to associated disease to this genotype, and all disease associated isolates characterized show one form of a deletion pattern in this highly polymorphic region (HPR). All European ISAV isolates can be genotyped according to deletion patterns in HPR. Most sequences belonging to a given HPR group derives from outbreaks with similar disease development, and cell culture replication and cytopathic effect vary between viruses from different HPR groups [41, 6]. Whether there is a link between the variation in HPR and a virulence pattern is, however, not known. ISAV exhibits, as the influenza viruses, hemagglutination, fusion and receptordestroying activity (RDE). ISAV, as influenza A and B, has two glycosylated surface proteins, however the genes and their functions are organized in a unique way as compared to other orthomyxoviruses. In addition to the 38–43 kDa hemagglutinin protein coded by ISAV genomic segment 6, the genomic segment 5 codes for a 50–53 kDa glycoprotein [16, 51, 35, 5, 12], recently shown to be the viral fusion protein (F) [2]. The ISAV genome sequent 6, also encodes an acetyl esterase exerting receptor destroying enzyme (RDE) activity. The 40–43 kD ISAV surface glycoprotein is thus organized as a hemagglutinin-esterase (HE) protein [12]. Neutralizing antibodies provide the first line defence, and probably the major protection against challenge with homologous influenza A and B virus [3]. Although infections with these viruses elicits antibodies against most influenza proteins [49], the most significant in their protective capacity are the neutralizing anti-HA antibodies [28]. In contrast, ISA-convalescent antiserum achieved after experimental infections with ISA is only partly protective [13]. Increasing evidence suggests that successful clearance of influenza virus infections is dependent on a strong Th1 response [43], characterized by the generation of virus-specific CTL, CD4+ T cells that produce IFN-γ, and virus-specific antibodies of the IgG2a type [37]. The central role of HA in protection against influenza A was recently demonstrated, where immunization with a protein fragment comprising only the globular region of influenza HA, including the receptorbinding pocket, was capable of inducing both a humoral (serum IgG and lung IgA) and a cellular (TH 1) response against the intact virus, as well as significant protection against a viral challenge infection [28]. The virulence of a virus is defined by its comparative capacity to produce disease in a host [61]. In general, host genetic factors are major determinants of susceptibility to infectious diseases in humans [25]. Recently, a strong association between the major histocompatibility complex (MHC) and disease susceptibility/resistance towards ISA was demonstrated in Atlantic salmon, where MHC class I and II alleles, as well as allele combinations provided substantial resistance to an ISAV infection [21]. In humans and mice, both MHC class I and II-restriction pathways participate in clearance of influenza virus [66, 10, 37]. In this study we have investigated virus-host interactions, such as disease susceptibility and immune responses, after experimental infection of Atlantic

2198

S. Mjaaland et al.: ISAV: virus-host interactions in MHC compatible fish

salmon (Salmo salar L.) with genetically different ISAV isolates. The fish were infected using cohabitant transmission, representing a natural route of infection, with virus isolates selected according to HPR genotype and their ability to induce acute versus protracted disease in field outbreaks. Due to the lack of appropriate experimental model systems and tools, it has been impossible to analyse T cell dependent immune reactions in fish. Most work on immune responses in this species has thus been limited to measurement of the humoral response. To avoid a mixed leukocyte reaction and to minimize host factor variation in immune responses we used MHC compatible half-siblings as experimental fish. This enabled us, for the first time, to determine the relative contribution of a humoral versus a cellular response in protection against ISAV. Materials and methods Experimental fish Atlantic salmon was delivered from Aqua Gen AS, Kyrkseterøra, Norway. The base population, collected from 13 Norwegian rivers in 1973, has undergone seven generations of selective breeding. The first selection used growth rate as the only trait, while the following selections included multiple traits such as non-grilse, flesh quality, disease resistance and growth rate. The fish used in the present experiment were hatched in spring 2002. This population has been selected twice for resistance to furunculosis (Aeromonas salmonicida) and once for resistance to infectious salmon anaemia (ISA). The experimental trial was performed at Veso Vikan Akvavet, Namsos, Norway. The test population consisted of 440 fish from each of two half-sibling families (TG-1 and TG-2), all heterozygous for Sasa-UBA∗ 0301/∗ 0201 (MHC class I [20]) and homozygous for Sasa-DAB∗ 0201-DAA∗ 0201 (MHC class II beta-alpha haplotype [57]). After transport from Aqua Gen to Veso Vikan Akvavet, 40 fish from each of the two test fish groups were placed in 11 separate tanks for 3 weeks of acclimatization. The average weight for the total test fish population was 27.6 g; 32.6 g for TG-1 fish and 24.8 g for TG-2. As cohabitant fish, 1320 Atlantic salmon from Aqua Gen (genetic background not defined), with an average weight of 20 g, were used. TG-1 fish were tagged in the fatfin, while TG-2 fish remained untagged. The cohabitant fish were tagged in the right maxille. The fish were kept in 0.6 meter tanks containing 150 l fresh water at temperatures between 10–12 ◦ C. The water-flow was kept at 0.8 l/kg fish per minute and the density maximum at 20 kg/m3 . ISAV isolates Eleven ISAV isolates were selected according to differences in hemagglutinin-esterase (HE) genotype (HPR groups) and ability to induce acute versus protracted disease in field outbreaks (Table 1). Kidney tissues collected from ISA outbreaks were stored at −80 ◦ C until homogenisation and virus isolation. Salmon head kidney cells (SHK-1 cells) were used for isolation of ISAV as described earlier [41]. ISAV was identified by an indirect immunofluorescence test (IIFT), using the anti-ISA monoclonal antibody 3H6F8 as primary antibody [15]. This antibody was a generous gift from Dr. Knut Falk. The isolates (cell culture supernatants from second or third passage of virus) were stored at −80 ◦ C until propagation of virus for use in the experimental trials.

Table 1. Percent accumulated mortality in cohabitant and test-group fish 65 days after challenge with 11 ISAV isolates which where selected according to variation in HE genotype and the course of disease they induce ISAV isolate

TG-1 fish

TG-2 fish

Cohabitant fish

Disease development

PRgroup

ISAV 1

45

47.5

67.5

A

11

30

27.5

75

A

15

22.5

32.5

24

A

7

10

25

67.5

A

4

10

17.5

40

A

14

15

21.7

A

2

0

5

7.5

A

4

0

22.5

P

10

∗VIR 13

#AF 427055

ISAV 2 VIR 1 AF 427043




ISAV 6

7.5

VIR 5 AF 427047

ISAV 7 VIR 11 AF 427053

ISAV 8

35

VIR 12 AF 427054

ISAV 9

7.5

7.5

3.3

P

3

5

5

6.7

P

6

5

2.5

P

8

VIR 7 AF 427049

ISAV 10 Selje AF 427078

ISAV 11

20

VIR 10 AF 427052

Atlantic salmon test group fish (TG-1 and TG-2) were challenged by cohabitant transmission with primary injected (cohabitant) fish. TG-1 and TG-2 represents two Atlantic salmon half-sibling populations with identical MHC class I and II genotypes. The ISAV isolates were selected according to ISAV HE genotype (HPR group) and the type of disease development the isolates induce in the field (A = acute, P = protracted). The isolates are ranked and named according to prevalence of death within the two groups of disease development (ISAV 1; highest prevalence of death in the acute group, ISAV 11; lowest prevalence of death within the protracted group) ∗ The term VIR refers to the terminology used in a previous work [42] # The AF numbers refers to GenBank accession numbers

2200

S. Mjaaland et al. Virus propagation and quantification

SHK-1 cells grown in 75 cm2 flasks were inoculated with a small volume of supernatants at appropriate dilutions from the second or third passages of the different ISAV strains. Cells inoculated with virus were incubated at 15 ◦ C for 4 h, followed by addition of Leibowitz (L-15) medium supplemented with foetal bovine serum (FBS, 5%), glutamin (4 mM), gentamicin (50 µg/ml) and 2-mercaptoethanol (40 µM). The cultures were regularly inspected for development of cytopathic effect (CPE), which usually appeared 1–2 weeks post infection. At full CPE, cell culture medium were collected and cleared by low speed centrifugation. The ISAV 11 isolate deviated from all the others by not developing full CPE, although signs of CPE were seen for this isolate as well. Virus titres were determined by end-point titration in 96-well culture plates with SHK-1 cells using 5-fold serial dilution and 4 parallels per dilution. After one week’s incubation, the cells were fixed and stained for ISAV using IIFT (as described above). The estimation of tissue culture infective doses (TCID50 /ml) was according to K¨arber [36]. For one isolate, ISAV 11, we were not able to estimate any virus titre. Experimental trial A cohabitant transmission model was chosen, representing a natural route of infection. In pre-challenge trials using the Norwegian reference strain Glesvaer 2/90 (isolate ISAV 4), a cumulative mortality of 70% was achieved 27 days after i.p. injection of a virus dose of 103 TCID50 . This infective dose was therefore used for each isolate to infect the cohabitant fish in the experimental trial. All cohabitant fish received an infective dose of 103 TCID50 of the respective isolate in a volume of 100 ul by i.p. injection, with two exceptions. For isolate ISAV 3, the fish received 100 µl of an infective dose of 0.7×102 TCID50 , while for ISAV 11, the fish received 100 µl undiluted supernatant. To ensure adequate levels of viral shedding, 120 cohabitant fish were added to each of the 11 tanks that already containing 40 fish from each of the half-sibling groups. The fish were kept in separate tanks, and feeding, inspection and registration of mortality continued for 9 weeks, when the study was terminated. Sampling and monitoring Day 0-samples from head kidney and liver from 5 fish from each test-group were fixed in formalin, together with imprints from 2 fish from each test-group in each tank. During the trial, a minimum of 10% dead cohabitants and 50% dead test fish per tank were post mortem examined and evaluated for pathological changes consistent with ISA [60, 11]. For verification of ISAV, imprints from up to 2 fish per test group per tank were sampled. A minimum of 10% dead test fish per tank was investigated for bacteriology on blood-agar plates and blood-agar plates added salt. At the end of the trial, samples were collected from the following representative groups: ISAV 1, 3, 4, 5, 7, 8, and 11. Blood samples from 10 fish per test-fish group per infected group were pooled and diluted in L-15 medium added 10% foetal bovine serum (FBS) for use in a leukocyte proliferation assay. Head kidney and heart from 10 fish per test-group per infected group, and 5 fish from each cohabitant group, were collected. The samples were split in two; one part placed in buffered formalin, the other placed in liquid N2 . Serum from the same groups were sampled and kept in liquid N2 . Statistical analysis Fisher’s Exact test was used for comparison of prevalence’s [1]. The prevalence of death was expressed in percent with 95% confidence intervals [1]. Additionally, the proportional hazard

ISAV: virus-host interactions in MHC compatible fish

2201

ratios for both the total material and within the acute and protracted groups were given with 95% confidence intervals [32]. The time until death was graphically expressed using Kaplan and Meier survival curves [48]. Survival analysis, with family as correcting factor, was used for comparison of groups with regard to time until death [48]. Leukocyte proliferation assay Leukocytes from blood samples pooled from 10 fish from each of the test-group fish infected with ISAV 1, 3, 4, 5, 7, 8 and 11 were isolated, as described earlier [8], and monocyte depleted by culture plate adherence. SHK-1 cells, with a MHC class I genotype identical to the experimental fish (SasaUBA∗ 0301/∗ 0201), were seeded in 96-well plates and inoculated with ISAV (multiplicity of infection (MOI) of 0.1) at 70–80% confluence. Three days after inoculation, 4 × 105 leukocytes were added to both ISAV-inoculated and uninfected control wells, and incubated at 15 ◦ C for 2 days, when 100 µl medium was replaced with fresh medium. After 5 days, fresh medium with serum was added (100 µl/well), followed by a 24 h challenge with 0.5 µCi 3 H-thymidine per well. After transfer to new plates, the leukocytes were harvested onto glass fibre filters, and incorporated radioactivity determined, using a liquid scintillation counter. The mean radioactivity, expressed as counts per minute (cpm) of four parallel wells, was calculated. The viability of the leukocytes was ensured by trypan-blue exclusion both prior to and after the proliferation on SHK-1 cells. The percentages of dead cells was low for all groups (