Molecular Determinants of Infectious Pancreatic Ne
http://www.100md.com
病菌学杂志 2005年第16期
Center for Biosystems Research, University of Maryland Biotechnology Institute and VA-MD Regional College of Veterinary Medicine, University of Maryland, College Park, Maryland 20742
Section for Pathology, National Veterinary Institute, N-0033 Oslo, Norway
Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, N-0033 Oslo, Norway
ABSTRACT
Infectious pancreatic necrosis viruses (IPNVs) exhibit a wide range of virulence in salmonid species. In previous studies, we have shown that the amino acid residues at positions 217 and 221 in VP2 are implicated in virulence. To pinpoint the molecular determinants of virulence in IPNV, we generated recombinant IPNV strains using the cRNA-based reverse-genetics system. In two virulent strains, residues at positions 217 and 247 were replaced by the corresponding amino acids of a low-virulence strain. The growth characteristics of the recovered chimeric strains in cell culture were similar to the low-virulence strains, and these viruses induced significantly lower mortality in Atlantic salmon fry than the parent strains did in in vivo challenge studies. Furthermore, the virulent strain was serially passaged in CHSE-214 cells 10 times and was completely characterized by nucleotide sequencing. Deduced amino acid sequence analyses revealed a single amino acid substitution of Ala to Thr at position 221 in VP2 of this virus, which became highly attenuated and induced 15% cumulative mortality in Atlantic salmon fry, compared to 68% mortality induced by the virulent parent strain. The attenuated strain grows to higher titers in CHSE cells and can be distinguished antigenically from the wild-type virus by use of a monoclonal antibody. However, the virulent strain passaged 10 times in RTG-2 cells was stable, and it retained its antigenicity and virulence. Our results indicate that residues Thr at position 217 (Thr217) and Ala221 of VP2 are the major determinants of virulence in IPNV of the Sp serotype. Highly virulent isolates possess residues Thr217 and Ala221; moderate- to low-virulence strains have Pro217 and Ala221; and strains containing Thr221 are almost avirulent, irrespective of the residue at position 217.
INTRODUCTION
Infectious pancreatic necrosis virus (IPNV) belongs to the genus Aquabirnavirus and is a prototype virus of the family Birnaviridae (6, 7). The genome of IPNV consists of two segments of double-stranded RNA, which are packaged in a nonenveloped icosahedral shell 60 nm in diameter. Segment A encodes a 106-kDa precursor protein in a single large open reading frame (ORF), which is cotranslationally cleaved by the viral encoded protease (VP4) to generate major capsid protein pVP2 and VP3 (7, 10). The pVP2 (62 kDa) is further cleaved to VP2 (54 kDa) during virus maturation (7, 12). VP2 is the major outer capsid protein, and type-specific neutralizing antibody is produced against this protein (7, 27). It is also believed to be the cell attachment protein (13, 19). VP3 is an internal capsid protein, which binds to virus RNA, forming the ribonucleoprotein core structure (16). Segment A also encodes an arginine-rich minor 17-kDa nonstructural protein (also called VP5) from a small ORF, which precedes and partly overlaps the large ORF (10). This protein has been detected in infected cells; however, it is not essential for virus replication in vitro (22, 36). Recently, it has been shown that this protein is dispensable for viral replication in vivo, and it is not involved in persistent infection or virulence of the virus (32). Segment B encodes a 94-kDa protein, VP1, which functions as the virion associated RNA-dependent RNA polymerase (7, 9). This protein is found both as free polypeptide and covalently linked to the 5' end of the genomic RNA segments (4).
Aquatic birnaviruses exhibit a wide host range and, apart from salmonids, they have been isolated from fish belonging to at least 32 different families, 11 species of mollusks, and 4 species of crustaceans (15). IPNV causes infectious pancreatic necrosis disease in salmonid fish (37). There are two distinct serogroups of IPNV; serogroup A comprises nine serotypes that are pathogenic to fish, whereas serogroup B comprises one serotype, avirulent for fish (15). In an earlier study, Sano and colleagues generated a reassortant virus between virulent and avirulent strains of two different serotypes and demonstrated that virulence of IPNV is associated with segment A and not with segment B, encoding VP1 (29). However, we recently demonstrated that VP1 protein of another birnavirus, infectious bursal disease virus (IBDV) modulates the virulence of IBDV in vivo (21). Virulence variation has been detected not only between serotypes but also within the same serotype (14, 34). Nucleotide sequence analyses reveal that VP2 may be the major determinant of virulence (3, 31, 33). By comparison of the deduced amino acid sequences of various field isolates exhibiting different mortality in Atlantic salmon fry, the putative motifs involved in virulence of IPNV Sp strains have been identified. Virulent strains typically encode a 12-kDa VP5 and have residues Thr, Ala, Thr/Ala, and Tyr/His at positions 217, 221, 247, and 500 of the VP2 gene (31, 33). Recently, to study the role of VP5 protein in virulence of IPNV, we recovered three viruses: one encoding a truncated 12-kDa VP5 (rNVI15), another encoding a full-length 15-kDa VP5 (rNVI15-15K), and one lacking the expression of VP5 (rNVI15-VP5). All three viruses are virulent and cause >80% mortality in Atlantic salmon smolts, suggesting that VP5 is not directly involved in the virulence of IPNV (32).
Earlier studies have demonstrated that virulent strains of IPNV lose their virulence potential after serial passage in cell culture (8, 25). However, the molecular basis for cell adaptation and attenuation is not known because none of these viruses were cloned and characterized by nucleotide sequence analysis. In a recent study, Santi and coworkers observed a substitution of residue at position 221 in VP2 (Ala to Thr) in IPNV field isolates propagated in cell culture, suggesting that this residue may be responsible for virus cell adaptation (31). Since IPNV field isolates exist as quasispecies, it is possible that this mutant arises from selection of one subtype existing as a minor fraction of the virus population in the diseased fish (17). Passage of a cloned virus generated by reverse genetics in cell culture would clearly resolve whether this substitution originates from a mutant in quasispecies population or by a primary mutation generating a variant strain with increased fitness in cultured cells.
In the present study, we investigated the role of residue 217 of VP2 in virulence by experimental challenge of wild Atlantic salmon fry and the role of VP2 residue 221 in virulence (in vivo), cell culture adaptation, and attenuation. Residue 217 was studied by using chimeric viruses generated from the virulent strains rNVI15 and rNVI15-15K and a moderately virulent strain, Sp103. The role of VP1 in virulence was also studied by the creation of a reassortant virus containing the VP1 protein of Sp103 and polyprotein of rNVI15-15K. In addition, rNVI15 and rNVI15-15K were serially passaged in CHSE-214 and RTG-2 cell cultures to evaluate the fate of VP2 residue 221 after cell adaptation. The growth characteristics of recovered viruses were studied using the two cell lines.
MATERIALS AND METHODS
Cells, viruses, and monoclonal antibody (MAb). Chinook salmon embryo cells (CHSE-214; ATCC CRL-1681) were maintained at 15°C in minimal essential medium (MEM) containing Hanks' salts, supplemented with 10% fetal bovine serum (FBS), and used for recovery of recombinant viruses. Rainbow trout gonad cells (RTG-2; ATCC CCL-55) were grown in L-15 medium supplemented with 10% FBS at 15°C and used for propagation of the recombinant viruses. Generation of rNVI15 and rNVI15-15K viruses has been described previously (32). Sp103 is a field strain of IPNV, causing moderate to low rates of mortality (31, 33). The virus was subjected to two rounds of plaque purification on CHSE-214 cells before propagation in RTG-2 cells to obtain a stock. The nucleic acid of the virus was sequenced completely to ensure that no advert mutation occurred due to cell adaptation. The naturally occurring VP5-deficient virus, Sp103-VP5, isolated by plaque purification of Sp103 has been characterized previously (32). Differences in the VP5 gene and the VP2-encoding region between IPNV strains used in this study are summarized in Table 1.
A MAb used in this study was directed against IPNV strain Sp122, having Thr at position 221 in VP2 (33). This MAb 122 was commercially prepared by Immuno-Precise Antibodies, Ltd. A diluted sample (1:20 dilution) of the hybridoma supernatant was directly used for immunoblotting, which recognized both pVP2 and VP2 proteins of IPNV containing a linear epitope of minimum residues spanning positions 220 to 223 (Gly-Thr-Leu-Thr-).
Construction of full-length cDNA clones. All manipulations of DNAs were performed according to standard protocols (28). Construction of full-length cDNA clones of pUC19NVI15A and pUC19NVI15-15K was described previously (32). Each of these clones encodes all the structural proteins (VP2, VP3, and VP4) and VP5 (Fig. 1). The 5' end of Sp103 segment A was cloned by reverse transcription-PCR (RT-PCR) using the primer pairs (A-A5'NC (5'-TAATACGACTCACTATAGGAAAGAGAGTTTCAACG-3') and A-SpKpnR (5'-GGCCATGGAGTGGTACCTTC-3'); the amplified fragment were cloned into pCR2.1 to obtain the plasmid pCRSp103A5'. Plasmids pUC19NVI15-15KVP2 and pUC19NVI15VP2 were prepared by replacing a BstEII-KpnI fragment in plasmid pUC19NVI15-15K and pUC19NVI15, respectively, with the respective BstEII-KpnI fragments derived from plasmids pCRSp103A5' (Fig. 1). Plasmids pUC19NVI15B and pUC19Sp103B, which comprise the full-length segment B of NVI15 and Sp103, respectively, were prepared as described previously (32).
DNA from the above-mentioned plasmids was sequenced by the dideoxy chain termination method using an automated DNA sequencer (Applied Biosystem), and the sequence data were analyzed using PC/Gene (Intelligenetics) software. The integrity of the full-length constructs was tested by an in vitro transcription-translation-coupled reticulocyte lysate system using T7 RNA polymerase (Promega Corp.).
Transcription and transfection of synthetic RNAs. Plasmid pUC19NVI15VP2, pUC19NVI15-15KVP2, and pUC19NVI15-15K were digested with PstI enzyme, whereas pUCNVI15B and pUCSp103B were digested with BglII. Linearized plasmid DNAs were purified as previously described (38). The linerized DNA was used to produce in vitro transcripts with the T7 mMessage mMachine kit (Ambion) according to the manufacturer's instructions. Briefly, approximately 3 μg linearized DNA template was added to the transcription reaction mixture (20 μl) containing 40 mM Tris-HCl (pH 7.9); 10 mM NaCl; 6 mM MgCl2; 2 mM spermidine; 0.5 mM each ATP, CTP, and UTP; 0.1 mM GTP; 0.25 mM cap analog [m7G(5')ppp(5')G], 120 U of RNasin, and 150 U of T7 RNA polymerase. The mixture was then incubated at 37°C for 90 min.
CHSE-214 cells grown to 90% confluence in a T-25 flask were transfected with cRNA of both segments as previously described (38). Briefly, cells were washed once with phosphate-buffered saline (PBS). Three milliliters of OPTI-MEM I (GIBCO/BRL) was added to the monolayer, and the cells were incubated at room temperature for 1 h. Simultaneously, 0.15 ml of OPTI-MEM I was incubated with 12.5 μg of Lipofectin reagent for 45 min in a polystyrene tube at room temperature. Equimolar amounts of RNA transcripts of segments A and B (8 μg each) were resuspended in 0.15 ml of diethyl pyrocarbonate-treated water, added to the OPTI-MEM/Lipofectin mixture, mixed gently, and incubated on ice for 5 min. After the OPTI-MEM I was removed from the monolayers in the T-25 flask and replaced it with a fresh 1.5 ml of OPTI-MEM, and the nucleic acid-containing mixture was added dropwise to the CHSE-214 cells and swirled gently. After 3 h of incubation at room temperature, the mixture was replaced with MEM containing Hanks' salts and 10% FBS (without rinsing the cells). The cultures were incubated at 15°C for 5 days, and the cell supernatant was harvested by being freeze-thawed twice and passaged onto fresh CHSE-214 monolayers. The cytopathic effect (CPE) was usually visualized at 4 days postinfection (p.i.). The flasks were freeze-thawed twice, and the supernatant was kept at –70°C as virus stock used in this work. The nucleotide sequence of the recovered viruses was verified by sequencing the RT-PCR products with the specific primers for both segments A and B, as previously described (33).
Serial passage of recombinant viruses in cell culture. Recombinant viruses were serially passaged in CHSE-214 or RTG-2 cells at a multiplicity of infection (MOI) of 0.001. At different passages, the virus RNA was extracted and used to amplify cDNA fragments of IPNV by RT-PCR. The RT-PCR products were purified and directly sequenced by using an IPNV-specific primer. After nine passages in CHSE-214 or RTG-2 cells, rNVI15 was plaque purified twice before passage in CHSE-214 or RTG-2 cells one more time to make a stock rNVI15C or rNVI15R virus, respectively. The same cell culture adaptation procedure was applied to rNVI15-15K to obtain rNVI15-15KC and rNVI15-15KR viruses. The complete nucleotide sequences of segments A and B of the cell-adapted viruses were determined as described previously (31, 33).
Immunoblot analysis. RTG-2 cells, grown in a T-25 flask, were infected with different IPNVs at an MOI of 1.0. After the CPE was fully developed, the cells were harvested, washed with PBS, and lysed in a sample preparation buffer. An aliquot of the cell lysate was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 5% stacking and 12.5% resolution gels, and proteins were transferred to the nitrocellulose membrane for immunoblot. After electrotransfer, the membrane was blocked in the blocking buffer containing 5% nonfat dry milk in Tris-buffered saline; 150 mM NaCl, 10 mM Tris-Cl, pH 7.5) for 90 min at room temperature and probed with MAb 122, which reacted only with the IPNV strains containing a linear epitope of residues spanning positions 220 to 223 in VP2 (Gly-Thr-Leu-Thr). Detection of the protein was achieved following incubation of the blot with alkaline phosphatase-conjugated goat anti-mouse immunoglobulin G and visualized by naphthol phosphate-fast red color reagents.
Growth curve and plaque assay. To analyze the growth characteristics of IPNVs, confluent CHSE-214 cells or RTG-2 cells (in a 35-mm dish) were infected with the recombinant virus stocks at an MOI of 1.0. Infected cell cultures were removed and stored at –70°C at different time intervals; the supernatants were centrifuged and titrated on CHSE-214 cells by plaque assay. Briefly, the confluent monolayers of CHSE-214 cells, grown in six-well plates, were infected with serially diluted supernatants from virus stock. After a 1-h incubation at 15°C, the cells were washed once by PBS and overlaid with 0.6% SeaPlaque Agrose (Difco) in Eagle MEM containing 5% FBS and 1% L-glutamine. After 3 days of incubation at 15°C, the overlays were removed and the cells were fixed and stained with a solution containing 25% formalin, 10% ethanol, 5% acetic acid, and 1% crystal violet for 5 min at room temperature. After the cells were rinsed with distilled water, the plaques were counted.
Immunofluorescent antibody focus assay. The relative infectivity of rNVI15C and rNVI15 viruses in RTG-2 and CHSE-214 cells was determined by a fluorescent focus assay. RTG-2 and CHSE cells were grown to 90% confluence in 24-well plates (Corning CellBIND). The cells were infected with rNVI15C or rNVI15 virus at MOIs of 0.1, 1.0, and 10.0, respectively. The cells were incubated for 6, 12, 18, and 24 h and then fixed with 4% paraformaldehyde. Fixed monolayers were permeabilized with 0.1% Triton X-100 for 5 min at 4°C before being blocked with 5% bovine serum albumin in PBS for 20 min at room temperature. Then, the cells were incubated with an anti-IPNV polyclonal rabbit antiserum, diluted 1:1,000 in 2.5% bovine serum albumin. After being washed with PBS, the cells were treated with fluorescein-labeled goat anti-rabbit antibody at a 1:100 dilution (Kirkegaard & Perry Laboratories) and examined by fluorescence microscopy (Olympus IX81) at x40 and x100 magnification. The stained monolayers were photographed with a Colorview camera (Soft Imaging System) with the aid of AnalysisB software (Soft Imaging System). For calculation of the average number of foci of infection for rNVI15C and rNVI15, IPNV-positive cells were counted in 10 separate x100 magnification fields of RTG-2 and CHSE-214 cells (infected at an MOI of 1.0) at the 24-h time point. After correcting for the relative virus dilution, the relative infectivity was reported as the CHSE-214/RTG-2 focus-forming ratio.
Fry challenge. Wild Atlantic salmon (Salmo salar L.) at the eyed egg stage were obtained from Haukvik GenBank station for wild salmon, Vinjera, Norway, and transported to the VESO Vikan Research Station, Namsos, Norway, where they were held through hatching and the start of feeding. One week before challenge, 50 fry were humanely killed and weighed for the calculation of average weight (0.2 g). The remaining 7,500 fry were randomly distributed into 30 tanks, each tank containing 250 fry. After 1-week acclimatization, the fry were deprived of feed the day before challenge. The challenge was performed in three parallel, randomly placed tanks for each of the IPNV isolates, in addition to three control tanks. Challenge was carried out by bath exposure at a dose of 105 50% tissue culture infective doses/ml of water, in a total volume of 4 liters per tank. The water was aerated during the challenge. After an exposure period of 3 h, normal flow was resumed. The fry were observed for 32 days, and mortality was recorded on a daily basis. Dead fry were removed each day and frozen at –70°C.
Reisolation and characterization of the virus after challenge. Two to four fish from each group dying during the second or third week after challenge were selected for reisolation of the virus and subsequent sequencing of the viral genome. The fry were homogenized using a stomacher, 1:5 (wt/vol) in PBS. A total of 100 μl of this homogenate was transferred to RLT buffer containing 2-mercaptoethanol, and RNA was isolated with the RNeasy Mini kit in accordance with the supplier's protocol (QIAGEN). Using primers A-A5'NC-2 (5'-GGAAAGAGAGTTTCAACG-3') and A-Sp1689R (5'-AGCCTGTTCTTGAGGGCTC-3'), a 1,689-bp segment covering the VP5 and VP2 coding region was amplified by RT-PCR. The PCR products were directly sequenced, as previously described (32). From the control group and from the fry challenged with rNVI15C, 24 additional dead fry were homogenized 1:10 (wt/vol) in Leibovitz's L-15 medium supplemented with 50 μg ml–1 gentamicin. The tissue homogenates were inoculated onto CHSE-214 cells grown in 24-well tissue culture plates in final dilutions of 1 and 0.1%, and the plates were incubated for 1 week at 15°C. The cell culture medium after the first passage was used to infect new monolayers. The samples were considered negative when no CPE was observed after a 1-week incubation of the second passage.
RESULTS
Construction of chimeric and reassortant viruses. Combined transcripts of pUC19NVI15VP2 plus pUC19NVI15B, pUC19NVI15-15KVP2 plus pUC19NVI15B, and pUC19NVI15-15K plus pUC19Sp103B resulted in the recovery of viral progeny designated rNVI15VP2, rNVI15-15KVP2, and rNVI15-15KVP1, respectively. In the chimeric viruses rNVI15VP2 and rNVI15-15KVP2, residues at positions 217 and 247 in VP2 of rNVI15 and rNVI15-15K were replaced by the corresponding amino acids from Sp103 (Thr to Pro at position 217 and Thr to Ala at position 247, respectively) (Fig. 1 and Table 1). The reassortant virus rNVI15-15KVP1 contains segment A from rNVI15-15K and segment B from Sp103 (Fig. 1 and Table 1). The genomic RNAs of the recovered viruses were analyzed after RT-PCR amplification, and the sequence analysis of the RT-PCR products confirmed the expected mutations in the VP2 and VP1 regions of the chimeric and reassortant viruses. Furthermore, complete nucleotide sequences of segment A of rNVI15VP2 and rNVI15-15KVP2 were determined, which did not exhibit any other unwanted nucleotide substitutions.
Sequence analysis of IPNV strains after serial passage in RTG-2 and CHSE-214 cells. To determine the molecular basis of virus adaptation in cell culture and a possible mechanism for virus attenuation, the recombinant IPNV strain rNVI15 was serially passaged in CHSE-214 cells. At indicated passage, virus RNAs were extracted and amplified by RT-PCR, and their products were directly sequenced. Comparison of the deduced amino acid sequences in the VP2 region of the virus at different passages revealed substitution of a single amino acid residue at position 221 from Ala to Thr due to a point mutation at nucleotide position 779 (G to A) in segment A (Fig. 2 and Table 2). After the fourth passage, about half of the virus population had Thr at position 221, whereas after the ninth passage, all of the virus population had Thr at that position. After two rounds of plaque purification of the ninth-passage virus, we obtained the cell-adapted virus rNVI15C, which had a Thr at position 221 in VP2 (Table 1). Complete nucleotide sequencing of genome segment A and B from rNVI15 and rNVI15C exhibited only one nucleotide change, at position 779, resulting in an Ala Thr substitution at VP2 residue 221, suggesting an important role of this amino acid in cell culture adaptation to CHSE-214 cells.
Similarly, rNVI15 was serially passaged in RTG-2 cells to obtain a virus strain rNVI15R (Table 1). However, passage carried out 10 times in this cell line did not result in any amino acid substitutions, as shown by sequence analyses of rNVI15R genomic segments A and B. To further confirm these findings, another cloned virus rNVI15-15K was also serially passaged in CHSE-214 or RTG-2 cells. This virus gradually acquired the amino acid substitution at position 221 (Ala to Thr) during passage in CHSE-214 cells, but not in RTG-2 cells (Table 2). These results indicate that amino acid substitution at position 221(Ala to Thr) only occurs during virus adaptation to CHSE-214 cells, but not in RTG-2 cells.
To investigate if cell culture adaptation to CHSE-214 cell is similar in all IPNV strains, we serially passaged the recovered chimeric and reassortant virus strains in CHSE-214 and RTG-2 cells. Viral RNAs were extracted after two, four, six, and nine passages, followed by RT-PCR amplification and direct sequencing of the PCR products using a pair of primers specific for the VP2 region. Our results reveal a gradual substitution of a residue from Ala to Thr at position 221, but no other adaptation mutations were detected in this region (Table 2). In rNVI15-15KVP1 virus, the change occurred after only four passages, similar to what was observed for rNVI15 and rNVI15-15K viruses. However, in rNVI15VP2 and rNVI15-15KVP2 viruses, the change could not be detected until the ninth passage; at that time, half of the virus population still had Ala at position 221. Similar results were obtained after serial passage of the field strain Sp103. When the recombinant viruses were serially passaged into RTG-2 cells, not a single amino acid substitution was detected in the VP2 protein. These results indicate that for virus to fully adapt to CHSE-214 cells, Ala at position 221 in VP2 must change to Thr. The adaptation mutation was delayed in viruses encoding Pro at position 217 and Ala at position 247.
Characterization of replication kinetics of recovered viruses in vitro. To determine if there are differences in the growth properties of the parental rNVI15 virus and the cell-adapted viruses, a one-step growth curve study was carried out with both CHSE-214 and RTG-2 cells. Cell-adapted rNVI15C virus caused a CPE in CHSE-214 cells at 18 h p.i., and reached maximal virus production (1 x 108 PFU/ml) at 40 h p.i. (Fig. 3A), whereas rNVI15-infected CHSE-214 cells developed a CPE at 40 h p.i., and the titer was only 3 x 106 PFU/ml. No difference in the replication kinetics of the two strains in RTG-2 cells was detected (Fig. 3B). As expected, rNVI15R exhibited growth kinetics similar to those of rNVI15 in both cell lines.
To investigate whether other VP2 residues or VP1 influence the efficiency of viral replication in vitro, one-step growth curves were made for rNVI15-15KVP2, rNVI15-15KVP1, and Sp103 in CHSE-214 and RTG-2 cells. Figure 3C depicts the growth curve of these viruses in CHSE-214 cells, along with the growth curve of rNVI15-15K at different time points postinfection. The results indicate that Sp103 virus replicated faster and had a titer 1 log higher than that of rNVI15-15K virus. The recombinant virus rNVI15-15KVP2 virus grew to a titer similar to that of the Sp103 virus, whereas rNVI15-15KVP1 virus exhibited replication kinetics similar to those of rNVI15-15K. However, when rNVI15-15KVP2 and rNVI15-15KVP1 viruses were fully adapted to CHSE cells (rNVI15-15KVP2C and rNVI-15KVP1C, respectively), both grew to a titer similar to that of the rNVI15C virus (data not shown). In RTG-2 cells, both rNVI15-15KVP2 and Sp103 viruses grew to titers higher than rNVI15-15K and rNVI15-15KVP1 (Fig. 3D). These results demonstrate that the amino acid substitution at positions 217 (Thr to Pro) and 247 (Thr to Ala) could allow the virus to efficiently replicate in both cell lines, whereas residue substitution in VP1 (Ile25Met and Arg240His) does not affect the viral replication kinetics in vitro.
Plaque phenotype, antigenicity, and fluorescent focus formation of IPNV cell-adapted variants. To determine the role of cell culture adaptation in virus plaque phenotype, rNVI15, rNVI15C, and rNVI15R viruses were subjected to plaque assays of both CHSE-214 and RTG-2 cells and analyzed 3 days postinfection (Fig. 4). Our results indicate that rNVI15C virus produced larger plaques than its parental virus (rNVI15) and rNVI15R, which correlates with the enhanced growth kinetics of this virus in CHSE cells. To confirm that the large-sized plaques are formed due to cell adaptation in CHSE-214 cells, we evaluated the plaque size of rNVI15 virus after different passages in CHSE-214 cells. Our results indicate that after the fourth passage, large-plaque variants begin to emerge and became predominant after six passages (data not shown), which corresponds to the sequence data presented earlier in Fig. 2. However, all the viruses produce similar sized plaques on RTG-2 cells (Fig. 4).
Furthermore, to ascertain that Ala residue at position 221 changes to Thr after cell culture adaptation in CHSE-214 cells, we utilized MAb 122. This MAb, prepared against the Sp122 strain of IPNV, recognizes a linear epitope in VP2 and reacts only with IPNV strains containing a stretch of residues (Gly-Thr-Leu-Thr) spanning positions 220 to 223 (data not shown). Consequently, MAb 122 recognizes the tissue culture-adapted viruses rNVI15C and rNVI15-15KC (Fig. 5, lanes 1 and 2) but fails to react with the wild-type rNVI15 and Sp103 viruses (Fig. 5, lanes 3 and 4), indicating that the adaptation mutation is a part of the linear epitope in VP2.
To determine relative infectivity of rNVI15C and rNVI15 viruses in RTG-2 and CHSE-214 cells, we performed a fluorescent focus assay using equivalent amounts of virus. Our results indicated that rNVI15C virus produced approximately equal numbers of infectious foci in both RTG-2 cells and CHSE-214 cells (Fig. 6A and C). On the other hand, rNVI15 virus infected RTG-2 cells with efficiency similar to that of rNVI15C (Fig. 6B), but the number of infectious foci in CHSE-214 cells was dramatically reduced (Fig. 6D). The relative infectivity ratio of rNVI15C virus in CHSE-214 cells versus RTG-2 cells was approximately 0.8, whereas for rNVI15 virus, it was only 0.1 (Fig. 6E). The reduced infectivity of rNVI15 virus to CHSE-214 cells was consistent at all virus dilutions and time points examined (data not shown). These results clearly indicate that rNVI15C virus preferentially establishes productive infections in CHSE cell cultures.
In vivo study of cell culture adapted and chimeric viruses. To assess the virulence and pathogenic phenotype of the cell-adapted virus and recovered viruses, wild Atlantic salmon fry were infected with IPNV by bath exposure. The challenge results are depicted in Fig. 7. The virulent strains rNVI15 and rNVI15-15K caused 68% (62 to 73%) and 70% (65 to 74%) cumulative mortality. (The numbers given in the parenthesis represent the 95% confidence interval of the mean; see similar results below). For strains rNVI15VP2 and rNVI15-15KVP2, a Thr-to-Pro substitution at position 217 of VP2 resulted in a significant reduction in mortality, ending at 39% (23 to 55%) and 47% (36 to 57%), respectively (Fig. 7A). This is similar to the moderately virulent field strain, Sp103, which caused 36% (31 to 40%) mortality. The results show that the residue at position 217 of VP2 is an important virulence determinant of IPNV serotype Sp strains. However, as shown in Fig. 7B, an Ala-to-Thr substitution at position 221 of VP2 seems to be even more important, which results in loss of virulence. While the wild-type virus rNVI15 and the RTG-2 cell-adapted strain (rNVI15R) induced 70% and 59% (51 to 67%) mortality, respectively, the cell culture-adapted virus (rNVI15C) gave only 15% (13 to 17%) mean cumulative mortality, forming a separate, almost avirulent group, together with the field strain Sp103-VP5 encoding Pro217 and Thr221, and ending at 5% (3 to 8%) mortality. This is not significantly different from the controls, which had 5% (4 to 6%) cumulative mortality. These results clearly indicate that CHSE-214 cell-adapted mutation (Ala221Thr) contributes to the loss of virulence after serial passage in CHSE-214 cells.
VP1, on the other hand, is not involved in the virulence characteristics of IPNV serotype Sp strains. In the fry challenge (Fig. 7C), the reassortant virus containing segment A from rNVI15-15K and segment B from Sp103 caused 60% (51 to 70%) mean cumulative mortality, which is similar to that of rNVI15 and significantly different from that of Sp103.
All viruses were recovered from the challenged fish that were sampled at the peak of mortality, but virus could not be reisolated from the control fish. Viral genomic RNA was extracted directly from diseased fish and amplified with a primer pair encompassing the VP2 and VP5 genes and the VP1 gene for rNVI15-15KVP1. Sequence data confirmed the presence of desired amino acid substitutions in the recombinant viruses, and no unwanted mutations occurred during virus replication in fish.
DISCUSSION
Previous studies of IPNV field strains have indicated that amino acids at positions 217 and 221of the major capsid protein VP2 may be responsible for differences in virulence between strains of the same serotype (3, 31, 33). However, there are additional amino acid substitutions scattered in VP2 and the other viral proteins; consequently, the precise residues involved in virulence were not identified. In this study, we used recombinant viruses generated by the reverse-genetics approach to prove the importance of these two VP2 residues in the virulence of IPNV serotype Sp strains. In addition to being an important determinant of virulence, residue 221 of VP2 is involved in cell culture adaptation of the virus. We found that VP2 residues 217 and 221 are the virulence determinants of IPNV Sp serotype strains, but the two residues seem to influence virulence by discrete mechanisms. Fry challenge studies showed that the viruses can be separated into three groups on the basis of their end point mortality. Viruses rNVI15, rNVI15-15K, rNVI-15KVP1, and rNVI15R, encoding Thr at 217 and Ala at VP2 residue 221, are virulent in nature. Fish started dying 7 to 9 days after challenge, and the mean cumulative mortality ranged from 59% to 70%. Strains rNVI15VP2, rNVI15-15KVP2, and Sp103, encoding Pro at 217 and Ala at 221, caused moderate mortality, ranging from 36% to 47%, and the onset of mortality was delayed 2 to 3 days compared to the virulent strains. The cell culture-adapted virus rNVI15C, and field strain Sp103-VP5, both encoding Thr at residue 221, form a separate, almost avirulent group, irrespective of the amino acid residue at position 217 (Table 1).
To pinpoint the amino acids involved in virulence of IPNV, we generated two chimeric strains, rNVI15VP2 and rNVI15-15KVP2, in which residues 217 (Thr Pro) and 247 (Thr Ala) were changed. Comparison of the deduced amino acid sequences of several IPNV-Sp strains revealed only four amino acid differences at positions 217, 221, 247, and 500 in VP2 protein (Table 1) (31, 33). Out of these four, only two residues, Thr217 and Ala221, are common to all virulent IPNV strains, indicating that VP2 residue 247 (which could be either Thr or Ala) is not directly involved in the virulence characteristics of IPNV. Therefore, reduction in mortality observed for the chimeric strains rNVI15VP2 and rNVI15-15KVP2, compared to that for parent strains rNVI15 and rNVI15-15K, is due to Pro at position 217.
Results from the fry challenge demonstrate that the field strain Sp103 is a moderately virulent strain with a cumulative mortality of 36%. This is significantly higher than in preceding studies, where the same strain was essentially avirulent, giving rise to 5 to 7% mortality (31, 33). Previously, we used commercially bred Atlantic salmon fry, obtained after genetic selection based on several criteria, including disease resistance against IPNV. In the present study, wild Atlantic salmon fry were used that had not been subject to any genetic selection at all; using this group of fish made it possible to separate virus strains at the low end of the virulence spectrum. These findings underline the importance of host-virus interaction in the outcome of virus infection and should be taken into consideration when planning and conducting IPNV challenge studies. An interesting approach would be to challenge different families of fish originating both from wild and commercially bred Atlantic salmon with IPNV strains of different virulence characteristics.
Previous studies have shown that IPNV field isolates tend to lose virulence after serial passage in cell culture (25). However, no attempts have been made to explain the underlying mechanisms. In our study, the major capsid protein VP2 carried the adaptation mutation, which is a substitution of residue 221 from Ala to Thr. There are two lines of evidence indicating that this change is responsible for rapid adaptation of the viruses to CHSE-214 cells. First, the cell-adapted virus rNVI15C replicates much faster and produces larger plaques than the parental strain (rNVI15). Secondly, the recombinant viruses rNVI15VP2, rNVI15-15K, rNVI15-15KVP2, and rNVI15-15KVP1 also acquired this adaptation mutation after passages in CHSE-214 cells.
Similar to rNVI15 virus, adaptation mutation at position 221 was rapidly acquired by rNVI15-15 and rNVI15-15KVP1 viruses. For strains Sp103, rNVI15VP2, and rNVI15-15KVP2, the mutation appeared after nine passages in CHSE-214 cells. The strains encoding Pro at position 217 and Ala at 247 grew to higher titers in both CHSE-214 and RTG-2 cells, compared to strains encoding Thr at 217 and Thr at 247, whereas the adaptive Ala Thr substitution at position 221 enhanced growth only in CHSE-214 cells. It is known that RNA viruses have the ability to quickly adapt to their environment. Alterations at viral entry steps, including receptor recognition, virus attachment and penetration, can affect virus growth and tissue tropism (20). For enveloped virus, mutations acquired during tissue adaptation are often located in the envelope proteins. For instance, it is shown that adaptive mutations in E protein selected by serial passage of Murray Valley encephalitis virus in cell culture affects virus binding to target cells (26). Similarly, mutations in tick-borne encephalitis virus E protein acquired during cell culture adaptation reduce virulence to animals (23). In nonenveloped viruses, adaptive mutations usually occur in outer capsid proteins, functioning as viral cell-binding and receptor recognition proteins. Like other nonenveloped viruses, IPNV is internalized by receptor-mediated endocytosis, and VP2 appears to be the cell attachment protein (7, 13, 19). For IBDV, it was recently shown that cell culture adaptation mutations reside in the VP2 region and facilitate virus entry into cells (2, 21). In regards to IPNV, a possible explanation for the growth characteristics of the various strains and mutants is that residue 217, perhaps in combination with 247, is involved in binding to common receptors present in both CHSE-214 and RTG-2 cells, whereas residue 221 may recognize a specific receptor required for virus entry in CHSE-214 cells but not in RTG cells. Our results from the fluorescent focus assay indicate that Ala Thr substitution at position 221 of the cell culture-adapted strain (rNVI15C) facilitates entry into CHSE-214 cells. Previously, Kuznar and coworkers have shown that IPNV attaches specifically and nonspecifically to CHSE-214 cells, but only specific binding leads to productive infection (19). Recently, Imajoh and colleagues demonstrated that a marine birnavirus binds to a 250-kDa protein in both susceptible and resistant cell lines, suggesting that a common receptor is present in all the cell lines but that a coreceptor is required for the virus to penetrate the cytoplasm by endocytosis in the susceptible cells (18). Further characterization of the cellular receptors for IPNV in different cell lines and crystallization of virus particles is necessary to understand the mechanism of virus-cell interaction in IPNV infection. More importantly, the same Ala-to-Thr substitution at position 221 in VP2 arising after cell culture adaptation was recently established as a molecular determinant for the establishment of persistent IPNV infection (32). This could result from different tissue tropism of the wild-type virus (Ala221) and the mutated virus (Thr221) during acute infection (exocrine pancreatic and liver tissue) and persistent infection (macrophages). Alternatively, the emergence of virus mutants could derive from selective pressure on the viral population by the immune response on epitopes recognized either by neutralizing antibodies or by cytotoxic T lymphocytes. Residue 221 lies within the central variable domain of VP2 containing the major conformational epitopes recognized by neutralizing MAbs (11, 35). Interestingly, the mutant strain rNVI15C can be differentiated from rNVI15 by MAb 122, indicating that this adaptive mutation changes important surface epitopes in the major capsid protein, which leads to the loss of virulence and establishment of persistent infection in fish. Incidentally, the crystal structure of IBDV was recently published, and the sequence alignment of IPNV VP2 with IBDV VP2 reveals that residues at positions 217 and 221 are indeed located in the outmost loops of the P domain, where neutralization-escape mutations occur (5).
Even after nine passages of the recombinant viruses in RTG-2 cells, not a single amino acid substitution was detected in segments A or B. Consequently, rNVI15R maintained its virulence in challenged fry. Other investigators, however, have detected attenuated mutants after passage of IPNV in RTG-2 cells (8). CHSE-214 cell-adapted virus rNVI15C did not produce higher titers than its parental virus in RTG-2 cells, indicating that Ala at position 221 does not enhance virus growth in RTG-2 cells. Blaney and coworkers observed that dengue virus adaptation mutations only occur after serial passage into Vero cells but not in C6/36 cells. Therefore, C6/36 cells were used to propagate the virus to ensure that no adverse cell adaptation mutations were introduced in the viral genome (1). Based on our findings for IPNV adaptation, RTG-2 would be a better cell line for propagating virulent viruses, at least up to 10 passages, to prevent the emergence of cell-adapted mutants.
By making reassortant viruses between two serotypes, VR299 and Ab, Sano and coworkers demonstrated that virulence of IPNV is associated with genomic segment A (29). Using the same approach, they also found that plaque size was dependent on segment A and that it was not associated with virulence (30). In lymphocytic choriomeningitis virus, a single amino acid change in the viral polymerase (residue 1079) is a major determinant of macrophage tropism and virulence (24). Recent studies in our laboratory have shown that VP1 of IBDV affects viral replication kinetics both in vitro and in vivo (21). Therefore, to exclude the possibility that segment B/VP1 is involved in virulence differences among IPNV strains within the same serotype, a reassortant virus was generated in this study. Our results indicate that VP1 is not involved in virulence and adaptation of virus to cell culture.
In conclusion, we have demonstrated that VP2 carries the determinants for IPNV virulence and cell culture adaptation. Virulent strains have Thr217 and Ala221, moderate- to low-virulence strains have Pro217 and Ala221, and strains with Thr221 are almost avirulent. An Ala Thr substitution occurs after passage in CHSE-214 cells but not in RTG-2 cells. The same mutation is selected in persistently infected fish, indicating that this specific change has biological relevance as well. The ability to separate moderate- and/or low-virulence strains from the avirulent strains is dependent on the fish strain used for challenge experiments, underlining the importance the of host-virus interaction in understanding viral infection.
ACKNOWLEDGMENTS
The project was supported by the National Research Initiative of the USDA Cooperative State Research, Education, and Extension Service, grant number 2001-35204-10065 to V.N.V., and a grant from the Norwegian Research Council (number 134136/120) to . E.
We thank Anders N. Haukvik at Haukvik Kraft-Smolt AS and Arnfinn Aunsmo at VESO Trondheim for providing the wild salmon fry; Anne Berit Romstad and Anne Ramstad, study monitors at VESO Vikan; and Gerard H. Edwards for technical assistance.
H.S. and N.S. contributed equally to this work.
REFERENCES
Blaney, J. E., G. G. Manipon, C. Firestone, D. H. Johnson, C. T. Hanson, B. R. Murphy, and S. S. Whitehead. 2003. Mutations which enhance the replication of dengue virus type 4 and an antigenic chimeric dengue virus type 2/4 vaccine candidate in Vero cells. Vaccine 21:4317-4327.
Brandt, M., K. Yao, M Liu, R. A. Heckert, and V. N. Vakharia. 2001. Molecular determinants of virulence, cell tropism, and pathogenic phenotype of infectious bursal disease virus. J. Virol. 75:11974-11982.
Bruslind, L. D., and P. W. Reno. 2000. Virulence comparison of three Buhl-subtype isolates of infectious pancreatic necrosis virus in brook trout fry. J. Aquat. Anim. Health 12:301-315.
Calvert, J. G., E. Nagy, M. Soler, and P. Dobos. 1991. Characterization of the VPg-dsRNA linkage of infectious pancreatic necrosis virus. J. Gen. Virol. 72:2563-2567.
Coulibaly, F., C. Chevalier, I. Gutsche, J. Pous, J. Navaza, S. Bressanelli, B. Delmas, and F. A. Ray. 2005. The birnavirus crystal structure reveals structural relationships among icosahedral viruses. Cell 120:761-772.
Delmas, B., F. S. B. Kibenge, J. C. Leong, E. Mundt, V. N. Vakharia, and J. L. Wu. 2005. Birnaviridae, p. 561-569. In C. M. Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger, and A. L. Ball, (ed.), Virus taxonomy. Academic Press, London, United Kingdom.
Dobos, P. 1995. The molecular biology of infectious pancreatic necrosis virus (IPNV). Annu. Rev. Fish Dis. 5:24-54.
Dorson, M., J. Castric, and C. Torchy. 1978. Infectious pancreatic necrosis virus of salmonids: biological and antigenic features of a pathogenic strain and of a non-pathogenic variant selected in RTG-2 cells. J. Fish. Dis. 1:309-320.
Duncan, R., C. L. Mason, E. Nagy, J. A. Leong, and P. Dobos. 1991. Sequence analysis of infectious pancreatic necrosis virus genome segment B and its encoded VP1 protein: a putative RNA-dependent RNA polymerase lacking the Gly-Asp-Asp motif. Virology 181:541-552.
Duncan, R., E. Nagy, P. J. Krell, and P. Dobos. 1987. Synthesis of infectious pancreatic necrosis virus polyprotein, detection of a virus encoded protease, and fine structure mapping of genome segment A coding regions. J. Virol. 61:3655-3664.
Frost, P., L. S. Hvardstein, B. Lygren, S. Sthl, C. Endresen, and K. E. Christie. 1995. Mapping of neutralization epitopes on infectious pancreatic necrosis viruses. J. Gen. Virol. 76:1165-1172.
Galloux, M., C. Chevalier, C. Henry, J.-C. Huet, B. Da Costa, and B. Delmas. 2004. Peptides resulting from the pVP2 C-terminal processing are present in infectious pancreatic necrosis virus particles. J. Gen. Virol. 85:2231-2236.
Granzow, H., F. Weiland, D. Fichtner, and P. J. Enzmann. 1997. Studies of the ultrastructure and morphogenesis of fish pathogenic viruses grown in cell culture. J. Fish. Dis. 20:1-10.
Hill, B. J. 1982. Infectious pancreatic necrosis virus and its virulence, p. 91-114. In R. J. Roberts (ed.), Microbial diseases of fish. Academic Press, London, United Kingdom.
Hill, B. J., and K. Way. 1995. Serological classification of infectious pancreatic necrosis (IPN) virus and other aquatic birnaviruses. Annu. Rev. Fish Dis. 5:55-77.
Hjalmarsson, A., E. Carlemalm, and E. Everitt. 1999. Infectious pancreatic necrosis virus: identification of a VP3-containing ribonucleoprotein core structure and evidence for O-linked glycosylation of the capsid protein VP2. J. Virol. 73:3484-3490.
Hsu, Y. L., C. C. Chen, and J. L. Wu. 1995. Molecular relationships in infectious pancreatic necrosis virus. Virus Res. 37:239-252.
Imajoh, M., K. Yagyu, and S. Oshima. 2003. Early interaction of marine birnavirus infection in several fish cell lines. J. Gen. Virol. 84:1809-1816.
Kuznar, J., M. Soler, G. Farias, and J. C. Espinoza. 1995. Attachment and entry of infectious pancreatic necrosis virus (IPNV) into CHSE-214 cells. Arch. Virol. 140:1833-1840.
Lee, P., R. Knight, J. M. Smith, J. Wilschut, and D. E. Griffin. 2002. A single mutation in the E2 glycoprotein important for neurovirulence influences binding of Sindbis virus to neuroblastoma cells. J. Virol. 76:6302-6310.
Liu, M., and V. N. Vakharia. 2004. VP1 protein of infectious bursal disease virus modulates the virulence in vivo. Virology 330:62-73.
Magyar, G., and P. Dobos. 1994. Evidence for the detection of the infectious pancreatic necrosis virus polyprotein and the 17-kDa polypeptide in infected cells and of the NS protease in purified virus. Virology 204:580-589.
Mandl, C. W., H. Kroschewski, S. L. Allison, R. Kofler, H. Holzmann, T. Meixner, and F. X. Heinz. 2001. Adaptation of tick-borne encephalitis virus to BHK-21 cells results in the formation of multiple heparan sulfate binding sites in the envelope protein and attenuation in vivo. J. Virol. 75:5627-5637.
Matloubian, M., S. R. Kolhekar, T. Somasundaram, and R. Ahmed. 1993. Molecular determinants of macrophage tropism and viral persistence: importance of single amino acid changes in the polymerase and glycoprotein of lymphocytic choriomeningitis virus. J. Virol. 67:7340-7349.
McAllister, P. E., and W. J. Owens. 1986. Infectious pancreatic necrosis virus: protocol for a standard challenge to brook trout. Trans. Amer. Fish. Soc. 115:466-470.
McMinn, P. C., I. D. Marshall, and L. Dalgarno. 1995. Neurovirulence and neuroinvasiveness of Murray Valley encephalitis virus mutants selected by passage in a monkey kidney cell line. J. Gen. Virol. 76:865-872.
Nicholson, B. L. 1993. Use of monoclonal antibodies in identification and characterization of fish viruses. Annu. Rev. Fish Dis. 3:241-257.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Sano, M., N. Okamoto, H. Fukuda, M. Saneyoshi, and T. Sano. 1992. Virulence of infectious pancreatic necrosis virus is associated with the larger RNA segment (RNA segment A). J. Fish. Dis. 15:283-293.
Sano, M., N. Okamoto, and T. Sano. 1994. Infectious pancreatic necrosis virus plaque size depends on the larger RNA segment and is not linked to virulence. J. Fish. Dis. 17:657-659.
Santi, N., V. N. Vakharia, and . Evensen. 2003. Identification of putative motifs involved in the virulence of infectious pancreatic necrosis virus. Virology 322:31-40.
Santi, N., H. Song, V. N. Vakharia, and . Evensen. 2005. Infectious pancreatic necrosis virus VP5 is dispensable for virulence and persistence. J. Virol. 79:9206-9216.
Shivappa, R. B., H. Song, K. Yao, A. Aas-Eng, . Evensen, and V. N. Vakharia. 2003. Molecular characterization of Sp serotype strains of infectious pancreatic necrosis virus exhibiting differences in virulence. Dis. Aquat. Org. 61:23-32.
Silim, A., M. A. S. Y. Elazhary, and A. Lagace. 1982. Susceptibility of trouts of different species and origins to various isolates of infectious pancreatic necrosis virus. Can. J. Fish Aquat. Sci. 39:1580-1584.
Tarrab, E., L. Berthiaume, S. Grothé, M. O'Connor-McCourt, J. Heppel, and J. Lecomte. 1995. Evidence of a major neutralizable conformational epitope region on VP2 of infectious pancreatic necrosis virus. J. Gen. Virol. 76:551-558.
Weber, S., D. Fichtner, T. C. Mettenleiter, and E. Mundt. 2001. Expression of VP5 of infectious pancreatic necrosis virus strain VR299 is initiated at the second in-frame start codon. J. Gen. Virol. 82:805-812.
Wolf, K. 1988. Fish viruses and fish viral diseases. Canstock Publishing Associates-Cornell University Press, Ithaca, N.Y.
Yao, K., and V. N. Vakharia. 1998. Generation of infectious pancreatic necrosis virus from cloned cDNA. J. Virol. 72:8913-8920.(Haichen Song, Nina Santi,)
Section for Pathology, National Veterinary Institute, N-0033 Oslo, Norway
Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, N-0033 Oslo, Norway
ABSTRACT
Infectious pancreatic necrosis viruses (IPNVs) exhibit a wide range of virulence in salmonid species. In previous studies, we have shown that the amino acid residues at positions 217 and 221 in VP2 are implicated in virulence. To pinpoint the molecular determinants of virulence in IPNV, we generated recombinant IPNV strains using the cRNA-based reverse-genetics system. In two virulent strains, residues at positions 217 and 247 were replaced by the corresponding amino acids of a low-virulence strain. The growth characteristics of the recovered chimeric strains in cell culture were similar to the low-virulence strains, and these viruses induced significantly lower mortality in Atlantic salmon fry than the parent strains did in in vivo challenge studies. Furthermore, the virulent strain was serially passaged in CHSE-214 cells 10 times and was completely characterized by nucleotide sequencing. Deduced amino acid sequence analyses revealed a single amino acid substitution of Ala to Thr at position 221 in VP2 of this virus, which became highly attenuated and induced 15% cumulative mortality in Atlantic salmon fry, compared to 68% mortality induced by the virulent parent strain. The attenuated strain grows to higher titers in CHSE cells and can be distinguished antigenically from the wild-type virus by use of a monoclonal antibody. However, the virulent strain passaged 10 times in RTG-2 cells was stable, and it retained its antigenicity and virulence. Our results indicate that residues Thr at position 217 (Thr217) and Ala221 of VP2 are the major determinants of virulence in IPNV of the Sp serotype. Highly virulent isolates possess residues Thr217 and Ala221; moderate- to low-virulence strains have Pro217 and Ala221; and strains containing Thr221 are almost avirulent, irrespective of the residue at position 217.
INTRODUCTION
Infectious pancreatic necrosis virus (IPNV) belongs to the genus Aquabirnavirus and is a prototype virus of the family Birnaviridae (6, 7). The genome of IPNV consists of two segments of double-stranded RNA, which are packaged in a nonenveloped icosahedral shell 60 nm in diameter. Segment A encodes a 106-kDa precursor protein in a single large open reading frame (ORF), which is cotranslationally cleaved by the viral encoded protease (VP4) to generate major capsid protein pVP2 and VP3 (7, 10). The pVP2 (62 kDa) is further cleaved to VP2 (54 kDa) during virus maturation (7, 12). VP2 is the major outer capsid protein, and type-specific neutralizing antibody is produced against this protein (7, 27). It is also believed to be the cell attachment protein (13, 19). VP3 is an internal capsid protein, which binds to virus RNA, forming the ribonucleoprotein core structure (16). Segment A also encodes an arginine-rich minor 17-kDa nonstructural protein (also called VP5) from a small ORF, which precedes and partly overlaps the large ORF (10). This protein has been detected in infected cells; however, it is not essential for virus replication in vitro (22, 36). Recently, it has been shown that this protein is dispensable for viral replication in vivo, and it is not involved in persistent infection or virulence of the virus (32). Segment B encodes a 94-kDa protein, VP1, which functions as the virion associated RNA-dependent RNA polymerase (7, 9). This protein is found both as free polypeptide and covalently linked to the 5' end of the genomic RNA segments (4).
Aquatic birnaviruses exhibit a wide host range and, apart from salmonids, they have been isolated from fish belonging to at least 32 different families, 11 species of mollusks, and 4 species of crustaceans (15). IPNV causes infectious pancreatic necrosis disease in salmonid fish (37). There are two distinct serogroups of IPNV; serogroup A comprises nine serotypes that are pathogenic to fish, whereas serogroup B comprises one serotype, avirulent for fish (15). In an earlier study, Sano and colleagues generated a reassortant virus between virulent and avirulent strains of two different serotypes and demonstrated that virulence of IPNV is associated with segment A and not with segment B, encoding VP1 (29). However, we recently demonstrated that VP1 protein of another birnavirus, infectious bursal disease virus (IBDV) modulates the virulence of IBDV in vivo (21). Virulence variation has been detected not only between serotypes but also within the same serotype (14, 34). Nucleotide sequence analyses reveal that VP2 may be the major determinant of virulence (3, 31, 33). By comparison of the deduced amino acid sequences of various field isolates exhibiting different mortality in Atlantic salmon fry, the putative motifs involved in virulence of IPNV Sp strains have been identified. Virulent strains typically encode a 12-kDa VP5 and have residues Thr, Ala, Thr/Ala, and Tyr/His at positions 217, 221, 247, and 500 of the VP2 gene (31, 33). Recently, to study the role of VP5 protein in virulence of IPNV, we recovered three viruses: one encoding a truncated 12-kDa VP5 (rNVI15), another encoding a full-length 15-kDa VP5 (rNVI15-15K), and one lacking the expression of VP5 (rNVI15-VP5). All three viruses are virulent and cause >80% mortality in Atlantic salmon smolts, suggesting that VP5 is not directly involved in the virulence of IPNV (32).
Earlier studies have demonstrated that virulent strains of IPNV lose their virulence potential after serial passage in cell culture (8, 25). However, the molecular basis for cell adaptation and attenuation is not known because none of these viruses were cloned and characterized by nucleotide sequence analysis. In a recent study, Santi and coworkers observed a substitution of residue at position 221 in VP2 (Ala to Thr) in IPNV field isolates propagated in cell culture, suggesting that this residue may be responsible for virus cell adaptation (31). Since IPNV field isolates exist as quasispecies, it is possible that this mutant arises from selection of one subtype existing as a minor fraction of the virus population in the diseased fish (17). Passage of a cloned virus generated by reverse genetics in cell culture would clearly resolve whether this substitution originates from a mutant in quasispecies population or by a primary mutation generating a variant strain with increased fitness in cultured cells.
In the present study, we investigated the role of residue 217 of VP2 in virulence by experimental challenge of wild Atlantic salmon fry and the role of VP2 residue 221 in virulence (in vivo), cell culture adaptation, and attenuation. Residue 217 was studied by using chimeric viruses generated from the virulent strains rNVI15 and rNVI15-15K and a moderately virulent strain, Sp103. The role of VP1 in virulence was also studied by the creation of a reassortant virus containing the VP1 protein of Sp103 and polyprotein of rNVI15-15K. In addition, rNVI15 and rNVI15-15K were serially passaged in CHSE-214 and RTG-2 cell cultures to evaluate the fate of VP2 residue 221 after cell adaptation. The growth characteristics of recovered viruses were studied using the two cell lines.
MATERIALS AND METHODS
Cells, viruses, and monoclonal antibody (MAb). Chinook salmon embryo cells (CHSE-214; ATCC CRL-1681) were maintained at 15°C in minimal essential medium (MEM) containing Hanks' salts, supplemented with 10% fetal bovine serum (FBS), and used for recovery of recombinant viruses. Rainbow trout gonad cells (RTG-2; ATCC CCL-55) were grown in L-15 medium supplemented with 10% FBS at 15°C and used for propagation of the recombinant viruses. Generation of rNVI15 and rNVI15-15K viruses has been described previously (32). Sp103 is a field strain of IPNV, causing moderate to low rates of mortality (31, 33). The virus was subjected to two rounds of plaque purification on CHSE-214 cells before propagation in RTG-2 cells to obtain a stock. The nucleic acid of the virus was sequenced completely to ensure that no advert mutation occurred due to cell adaptation. The naturally occurring VP5-deficient virus, Sp103-VP5, isolated by plaque purification of Sp103 has been characterized previously (32). Differences in the VP5 gene and the VP2-encoding region between IPNV strains used in this study are summarized in Table 1.
A MAb used in this study was directed against IPNV strain Sp122, having Thr at position 221 in VP2 (33). This MAb 122 was commercially prepared by Immuno-Precise Antibodies, Ltd. A diluted sample (1:20 dilution) of the hybridoma supernatant was directly used for immunoblotting, which recognized both pVP2 and VP2 proteins of IPNV containing a linear epitope of minimum residues spanning positions 220 to 223 (Gly-Thr-Leu-Thr-).
Construction of full-length cDNA clones. All manipulations of DNAs were performed according to standard protocols (28). Construction of full-length cDNA clones of pUC19NVI15A and pUC19NVI15-15K was described previously (32). Each of these clones encodes all the structural proteins (VP2, VP3, and VP4) and VP5 (Fig. 1). The 5' end of Sp103 segment A was cloned by reverse transcription-PCR (RT-PCR) using the primer pairs (A-A5'NC (5'-TAATACGACTCACTATAGGAAAGAGAGTTTCAACG-3') and A-SpKpnR (5'-GGCCATGGAGTGGTACCTTC-3'); the amplified fragment were cloned into pCR2.1 to obtain the plasmid pCRSp103A5'. Plasmids pUC19NVI15-15KVP2 and pUC19NVI15VP2 were prepared by replacing a BstEII-KpnI fragment in plasmid pUC19NVI15-15K and pUC19NVI15, respectively, with the respective BstEII-KpnI fragments derived from plasmids pCRSp103A5' (Fig. 1). Plasmids pUC19NVI15B and pUC19Sp103B, which comprise the full-length segment B of NVI15 and Sp103, respectively, were prepared as described previously (32).
DNA from the above-mentioned plasmids was sequenced by the dideoxy chain termination method using an automated DNA sequencer (Applied Biosystem), and the sequence data were analyzed using PC/Gene (Intelligenetics) software. The integrity of the full-length constructs was tested by an in vitro transcription-translation-coupled reticulocyte lysate system using T7 RNA polymerase (Promega Corp.).
Transcription and transfection of synthetic RNAs. Plasmid pUC19NVI15VP2, pUC19NVI15-15KVP2, and pUC19NVI15-15K were digested with PstI enzyme, whereas pUCNVI15B and pUCSp103B were digested with BglII. Linearized plasmid DNAs were purified as previously described (38). The linerized DNA was used to produce in vitro transcripts with the T7 mMessage mMachine kit (Ambion) according to the manufacturer's instructions. Briefly, approximately 3 μg linearized DNA template was added to the transcription reaction mixture (20 μl) containing 40 mM Tris-HCl (pH 7.9); 10 mM NaCl; 6 mM MgCl2; 2 mM spermidine; 0.5 mM each ATP, CTP, and UTP; 0.1 mM GTP; 0.25 mM cap analog [m7G(5')ppp(5')G], 120 U of RNasin, and 150 U of T7 RNA polymerase. The mixture was then incubated at 37°C for 90 min.
CHSE-214 cells grown to 90% confluence in a T-25 flask were transfected with cRNA of both segments as previously described (38). Briefly, cells were washed once with phosphate-buffered saline (PBS). Three milliliters of OPTI-MEM I (GIBCO/BRL) was added to the monolayer, and the cells were incubated at room temperature for 1 h. Simultaneously, 0.15 ml of OPTI-MEM I was incubated with 12.5 μg of Lipofectin reagent for 45 min in a polystyrene tube at room temperature. Equimolar amounts of RNA transcripts of segments A and B (8 μg each) were resuspended in 0.15 ml of diethyl pyrocarbonate-treated water, added to the OPTI-MEM/Lipofectin mixture, mixed gently, and incubated on ice for 5 min. After the OPTI-MEM I was removed from the monolayers in the T-25 flask and replaced it with a fresh 1.5 ml of OPTI-MEM, and the nucleic acid-containing mixture was added dropwise to the CHSE-214 cells and swirled gently. After 3 h of incubation at room temperature, the mixture was replaced with MEM containing Hanks' salts and 10% FBS (without rinsing the cells). The cultures were incubated at 15°C for 5 days, and the cell supernatant was harvested by being freeze-thawed twice and passaged onto fresh CHSE-214 monolayers. The cytopathic effect (CPE) was usually visualized at 4 days postinfection (p.i.). The flasks were freeze-thawed twice, and the supernatant was kept at –70°C as virus stock used in this work. The nucleotide sequence of the recovered viruses was verified by sequencing the RT-PCR products with the specific primers for both segments A and B, as previously described (33).
Serial passage of recombinant viruses in cell culture. Recombinant viruses were serially passaged in CHSE-214 or RTG-2 cells at a multiplicity of infection (MOI) of 0.001. At different passages, the virus RNA was extracted and used to amplify cDNA fragments of IPNV by RT-PCR. The RT-PCR products were purified and directly sequenced by using an IPNV-specific primer. After nine passages in CHSE-214 or RTG-2 cells, rNVI15 was plaque purified twice before passage in CHSE-214 or RTG-2 cells one more time to make a stock rNVI15C or rNVI15R virus, respectively. The same cell culture adaptation procedure was applied to rNVI15-15K to obtain rNVI15-15KC and rNVI15-15KR viruses. The complete nucleotide sequences of segments A and B of the cell-adapted viruses were determined as described previously (31, 33).
Immunoblot analysis. RTG-2 cells, grown in a T-25 flask, were infected with different IPNVs at an MOI of 1.0. After the CPE was fully developed, the cells were harvested, washed with PBS, and lysed in a sample preparation buffer. An aliquot of the cell lysate was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis using 5% stacking and 12.5% resolution gels, and proteins were transferred to the nitrocellulose membrane for immunoblot. After electrotransfer, the membrane was blocked in the blocking buffer containing 5% nonfat dry milk in Tris-buffered saline; 150 mM NaCl, 10 mM Tris-Cl, pH 7.5) for 90 min at room temperature and probed with MAb 122, which reacted only with the IPNV strains containing a linear epitope of residues spanning positions 220 to 223 in VP2 (Gly-Thr-Leu-Thr). Detection of the protein was achieved following incubation of the blot with alkaline phosphatase-conjugated goat anti-mouse immunoglobulin G and visualized by naphthol phosphate-fast red color reagents.
Growth curve and plaque assay. To analyze the growth characteristics of IPNVs, confluent CHSE-214 cells or RTG-2 cells (in a 35-mm dish) were infected with the recombinant virus stocks at an MOI of 1.0. Infected cell cultures were removed and stored at –70°C at different time intervals; the supernatants were centrifuged and titrated on CHSE-214 cells by plaque assay. Briefly, the confluent monolayers of CHSE-214 cells, grown in six-well plates, were infected with serially diluted supernatants from virus stock. After a 1-h incubation at 15°C, the cells were washed once by PBS and overlaid with 0.6% SeaPlaque Agrose (Difco) in Eagle MEM containing 5% FBS and 1% L-glutamine. After 3 days of incubation at 15°C, the overlays were removed and the cells were fixed and stained with a solution containing 25% formalin, 10% ethanol, 5% acetic acid, and 1% crystal violet for 5 min at room temperature. After the cells were rinsed with distilled water, the plaques were counted.
Immunofluorescent antibody focus assay. The relative infectivity of rNVI15C and rNVI15 viruses in RTG-2 and CHSE-214 cells was determined by a fluorescent focus assay. RTG-2 and CHSE cells were grown to 90% confluence in 24-well plates (Corning CellBIND). The cells were infected with rNVI15C or rNVI15 virus at MOIs of 0.1, 1.0, and 10.0, respectively. The cells were incubated for 6, 12, 18, and 24 h and then fixed with 4% paraformaldehyde. Fixed monolayers were permeabilized with 0.1% Triton X-100 for 5 min at 4°C before being blocked with 5% bovine serum albumin in PBS for 20 min at room temperature. Then, the cells were incubated with an anti-IPNV polyclonal rabbit antiserum, diluted 1:1,000 in 2.5% bovine serum albumin. After being washed with PBS, the cells were treated with fluorescein-labeled goat anti-rabbit antibody at a 1:100 dilution (Kirkegaard & Perry Laboratories) and examined by fluorescence microscopy (Olympus IX81) at x40 and x100 magnification. The stained monolayers were photographed with a Colorview camera (Soft Imaging System) with the aid of AnalysisB software (Soft Imaging System). For calculation of the average number of foci of infection for rNVI15C and rNVI15, IPNV-positive cells were counted in 10 separate x100 magnification fields of RTG-2 and CHSE-214 cells (infected at an MOI of 1.0) at the 24-h time point. After correcting for the relative virus dilution, the relative infectivity was reported as the CHSE-214/RTG-2 focus-forming ratio.
Fry challenge. Wild Atlantic salmon (Salmo salar L.) at the eyed egg stage were obtained from Haukvik GenBank station for wild salmon, Vinjera, Norway, and transported to the VESO Vikan Research Station, Namsos, Norway, where they were held through hatching and the start of feeding. One week before challenge, 50 fry were humanely killed and weighed for the calculation of average weight (0.2 g). The remaining 7,500 fry were randomly distributed into 30 tanks, each tank containing 250 fry. After 1-week acclimatization, the fry were deprived of feed the day before challenge. The challenge was performed in three parallel, randomly placed tanks for each of the IPNV isolates, in addition to three control tanks. Challenge was carried out by bath exposure at a dose of 105 50% tissue culture infective doses/ml of water, in a total volume of 4 liters per tank. The water was aerated during the challenge. After an exposure period of 3 h, normal flow was resumed. The fry were observed for 32 days, and mortality was recorded on a daily basis. Dead fry were removed each day and frozen at –70°C.
Reisolation and characterization of the virus after challenge. Two to four fish from each group dying during the second or third week after challenge were selected for reisolation of the virus and subsequent sequencing of the viral genome. The fry were homogenized using a stomacher, 1:5 (wt/vol) in PBS. A total of 100 μl of this homogenate was transferred to RLT buffer containing 2-mercaptoethanol, and RNA was isolated with the RNeasy Mini kit in accordance with the supplier's protocol (QIAGEN). Using primers A-A5'NC-2 (5'-GGAAAGAGAGTTTCAACG-3') and A-Sp1689R (5'-AGCCTGTTCTTGAGGGCTC-3'), a 1,689-bp segment covering the VP5 and VP2 coding region was amplified by RT-PCR. The PCR products were directly sequenced, as previously described (32). From the control group and from the fry challenged with rNVI15C, 24 additional dead fry were homogenized 1:10 (wt/vol) in Leibovitz's L-15 medium supplemented with 50 μg ml–1 gentamicin. The tissue homogenates were inoculated onto CHSE-214 cells grown in 24-well tissue culture plates in final dilutions of 1 and 0.1%, and the plates were incubated for 1 week at 15°C. The cell culture medium after the first passage was used to infect new monolayers. The samples were considered negative when no CPE was observed after a 1-week incubation of the second passage.
RESULTS
Construction of chimeric and reassortant viruses. Combined transcripts of pUC19NVI15VP2 plus pUC19NVI15B, pUC19NVI15-15KVP2 plus pUC19NVI15B, and pUC19NVI15-15K plus pUC19Sp103B resulted in the recovery of viral progeny designated rNVI15VP2, rNVI15-15KVP2, and rNVI15-15KVP1, respectively. In the chimeric viruses rNVI15VP2 and rNVI15-15KVP2, residues at positions 217 and 247 in VP2 of rNVI15 and rNVI15-15K were replaced by the corresponding amino acids from Sp103 (Thr to Pro at position 217 and Thr to Ala at position 247, respectively) (Fig. 1 and Table 1). The reassortant virus rNVI15-15KVP1 contains segment A from rNVI15-15K and segment B from Sp103 (Fig. 1 and Table 1). The genomic RNAs of the recovered viruses were analyzed after RT-PCR amplification, and the sequence analysis of the RT-PCR products confirmed the expected mutations in the VP2 and VP1 regions of the chimeric and reassortant viruses. Furthermore, complete nucleotide sequences of segment A of rNVI15VP2 and rNVI15-15KVP2 were determined, which did not exhibit any other unwanted nucleotide substitutions.
Sequence analysis of IPNV strains after serial passage in RTG-2 and CHSE-214 cells. To determine the molecular basis of virus adaptation in cell culture and a possible mechanism for virus attenuation, the recombinant IPNV strain rNVI15 was serially passaged in CHSE-214 cells. At indicated passage, virus RNAs were extracted and amplified by RT-PCR, and their products were directly sequenced. Comparison of the deduced amino acid sequences in the VP2 region of the virus at different passages revealed substitution of a single amino acid residue at position 221 from Ala to Thr due to a point mutation at nucleotide position 779 (G to A) in segment A (Fig. 2 and Table 2). After the fourth passage, about half of the virus population had Thr at position 221, whereas after the ninth passage, all of the virus population had Thr at that position. After two rounds of plaque purification of the ninth-passage virus, we obtained the cell-adapted virus rNVI15C, which had a Thr at position 221 in VP2 (Table 1). Complete nucleotide sequencing of genome segment A and B from rNVI15 and rNVI15C exhibited only one nucleotide change, at position 779, resulting in an Ala Thr substitution at VP2 residue 221, suggesting an important role of this amino acid in cell culture adaptation to CHSE-214 cells.
Similarly, rNVI15 was serially passaged in RTG-2 cells to obtain a virus strain rNVI15R (Table 1). However, passage carried out 10 times in this cell line did not result in any amino acid substitutions, as shown by sequence analyses of rNVI15R genomic segments A and B. To further confirm these findings, another cloned virus rNVI15-15K was also serially passaged in CHSE-214 or RTG-2 cells. This virus gradually acquired the amino acid substitution at position 221 (Ala to Thr) during passage in CHSE-214 cells, but not in RTG-2 cells (Table 2). These results indicate that amino acid substitution at position 221(Ala to Thr) only occurs during virus adaptation to CHSE-214 cells, but not in RTG-2 cells.
To investigate if cell culture adaptation to CHSE-214 cell is similar in all IPNV strains, we serially passaged the recovered chimeric and reassortant virus strains in CHSE-214 and RTG-2 cells. Viral RNAs were extracted after two, four, six, and nine passages, followed by RT-PCR amplification and direct sequencing of the PCR products using a pair of primers specific for the VP2 region. Our results reveal a gradual substitution of a residue from Ala to Thr at position 221, but no other adaptation mutations were detected in this region (Table 2). In rNVI15-15KVP1 virus, the change occurred after only four passages, similar to what was observed for rNVI15 and rNVI15-15K viruses. However, in rNVI15VP2 and rNVI15-15KVP2 viruses, the change could not be detected until the ninth passage; at that time, half of the virus population still had Ala at position 221. Similar results were obtained after serial passage of the field strain Sp103. When the recombinant viruses were serially passaged into RTG-2 cells, not a single amino acid substitution was detected in the VP2 protein. These results indicate that for virus to fully adapt to CHSE-214 cells, Ala at position 221 in VP2 must change to Thr. The adaptation mutation was delayed in viruses encoding Pro at position 217 and Ala at position 247.
Characterization of replication kinetics of recovered viruses in vitro. To determine if there are differences in the growth properties of the parental rNVI15 virus and the cell-adapted viruses, a one-step growth curve study was carried out with both CHSE-214 and RTG-2 cells. Cell-adapted rNVI15C virus caused a CPE in CHSE-214 cells at 18 h p.i., and reached maximal virus production (1 x 108 PFU/ml) at 40 h p.i. (Fig. 3A), whereas rNVI15-infected CHSE-214 cells developed a CPE at 40 h p.i., and the titer was only 3 x 106 PFU/ml. No difference in the replication kinetics of the two strains in RTG-2 cells was detected (Fig. 3B). As expected, rNVI15R exhibited growth kinetics similar to those of rNVI15 in both cell lines.
To investigate whether other VP2 residues or VP1 influence the efficiency of viral replication in vitro, one-step growth curves were made for rNVI15-15KVP2, rNVI15-15KVP1, and Sp103 in CHSE-214 and RTG-2 cells. Figure 3C depicts the growth curve of these viruses in CHSE-214 cells, along with the growth curve of rNVI15-15K at different time points postinfection. The results indicate that Sp103 virus replicated faster and had a titer 1 log higher than that of rNVI15-15K virus. The recombinant virus rNVI15-15KVP2 virus grew to a titer similar to that of the Sp103 virus, whereas rNVI15-15KVP1 virus exhibited replication kinetics similar to those of rNVI15-15K. However, when rNVI15-15KVP2 and rNVI15-15KVP1 viruses were fully adapted to CHSE cells (rNVI15-15KVP2C and rNVI-15KVP1C, respectively), both grew to a titer similar to that of the rNVI15C virus (data not shown). In RTG-2 cells, both rNVI15-15KVP2 and Sp103 viruses grew to titers higher than rNVI15-15K and rNVI15-15KVP1 (Fig. 3D). These results demonstrate that the amino acid substitution at positions 217 (Thr to Pro) and 247 (Thr to Ala) could allow the virus to efficiently replicate in both cell lines, whereas residue substitution in VP1 (Ile25Met and Arg240His) does not affect the viral replication kinetics in vitro.
Plaque phenotype, antigenicity, and fluorescent focus formation of IPNV cell-adapted variants. To determine the role of cell culture adaptation in virus plaque phenotype, rNVI15, rNVI15C, and rNVI15R viruses were subjected to plaque assays of both CHSE-214 and RTG-2 cells and analyzed 3 days postinfection (Fig. 4). Our results indicate that rNVI15C virus produced larger plaques than its parental virus (rNVI15) and rNVI15R, which correlates with the enhanced growth kinetics of this virus in CHSE cells. To confirm that the large-sized plaques are formed due to cell adaptation in CHSE-214 cells, we evaluated the plaque size of rNVI15 virus after different passages in CHSE-214 cells. Our results indicate that after the fourth passage, large-plaque variants begin to emerge and became predominant after six passages (data not shown), which corresponds to the sequence data presented earlier in Fig. 2. However, all the viruses produce similar sized plaques on RTG-2 cells (Fig. 4).
Furthermore, to ascertain that Ala residue at position 221 changes to Thr after cell culture adaptation in CHSE-214 cells, we utilized MAb 122. This MAb, prepared against the Sp122 strain of IPNV, recognizes a linear epitope in VP2 and reacts only with IPNV strains containing a stretch of residues (Gly-Thr-Leu-Thr) spanning positions 220 to 223 (data not shown). Consequently, MAb 122 recognizes the tissue culture-adapted viruses rNVI15C and rNVI15-15KC (Fig. 5, lanes 1 and 2) but fails to react with the wild-type rNVI15 and Sp103 viruses (Fig. 5, lanes 3 and 4), indicating that the adaptation mutation is a part of the linear epitope in VP2.
To determine relative infectivity of rNVI15C and rNVI15 viruses in RTG-2 and CHSE-214 cells, we performed a fluorescent focus assay using equivalent amounts of virus. Our results indicated that rNVI15C virus produced approximately equal numbers of infectious foci in both RTG-2 cells and CHSE-214 cells (Fig. 6A and C). On the other hand, rNVI15 virus infected RTG-2 cells with efficiency similar to that of rNVI15C (Fig. 6B), but the number of infectious foci in CHSE-214 cells was dramatically reduced (Fig. 6D). The relative infectivity ratio of rNVI15C virus in CHSE-214 cells versus RTG-2 cells was approximately 0.8, whereas for rNVI15 virus, it was only 0.1 (Fig. 6E). The reduced infectivity of rNVI15 virus to CHSE-214 cells was consistent at all virus dilutions and time points examined (data not shown). These results clearly indicate that rNVI15C virus preferentially establishes productive infections in CHSE cell cultures.
In vivo study of cell culture adapted and chimeric viruses. To assess the virulence and pathogenic phenotype of the cell-adapted virus and recovered viruses, wild Atlantic salmon fry were infected with IPNV by bath exposure. The challenge results are depicted in Fig. 7. The virulent strains rNVI15 and rNVI15-15K caused 68% (62 to 73%) and 70% (65 to 74%) cumulative mortality. (The numbers given in the parenthesis represent the 95% confidence interval of the mean; see similar results below). For strains rNVI15VP2 and rNVI15-15KVP2, a Thr-to-Pro substitution at position 217 of VP2 resulted in a significant reduction in mortality, ending at 39% (23 to 55%) and 47% (36 to 57%), respectively (Fig. 7A). This is similar to the moderately virulent field strain, Sp103, which caused 36% (31 to 40%) mortality. The results show that the residue at position 217 of VP2 is an important virulence determinant of IPNV serotype Sp strains. However, as shown in Fig. 7B, an Ala-to-Thr substitution at position 221 of VP2 seems to be even more important, which results in loss of virulence. While the wild-type virus rNVI15 and the RTG-2 cell-adapted strain (rNVI15R) induced 70% and 59% (51 to 67%) mortality, respectively, the cell culture-adapted virus (rNVI15C) gave only 15% (13 to 17%) mean cumulative mortality, forming a separate, almost avirulent group, together with the field strain Sp103-VP5 encoding Pro217 and Thr221, and ending at 5% (3 to 8%) mortality. This is not significantly different from the controls, which had 5% (4 to 6%) cumulative mortality. These results clearly indicate that CHSE-214 cell-adapted mutation (Ala221Thr) contributes to the loss of virulence after serial passage in CHSE-214 cells.
VP1, on the other hand, is not involved in the virulence characteristics of IPNV serotype Sp strains. In the fry challenge (Fig. 7C), the reassortant virus containing segment A from rNVI15-15K and segment B from Sp103 caused 60% (51 to 70%) mean cumulative mortality, which is similar to that of rNVI15 and significantly different from that of Sp103.
All viruses were recovered from the challenged fish that were sampled at the peak of mortality, but virus could not be reisolated from the control fish. Viral genomic RNA was extracted directly from diseased fish and amplified with a primer pair encompassing the VP2 and VP5 genes and the VP1 gene for rNVI15-15KVP1. Sequence data confirmed the presence of desired amino acid substitutions in the recombinant viruses, and no unwanted mutations occurred during virus replication in fish.
DISCUSSION
Previous studies of IPNV field strains have indicated that amino acids at positions 217 and 221of the major capsid protein VP2 may be responsible for differences in virulence between strains of the same serotype (3, 31, 33). However, there are additional amino acid substitutions scattered in VP2 and the other viral proteins; consequently, the precise residues involved in virulence were not identified. In this study, we used recombinant viruses generated by the reverse-genetics approach to prove the importance of these two VP2 residues in the virulence of IPNV serotype Sp strains. In addition to being an important determinant of virulence, residue 221 of VP2 is involved in cell culture adaptation of the virus. We found that VP2 residues 217 and 221 are the virulence determinants of IPNV Sp serotype strains, but the two residues seem to influence virulence by discrete mechanisms. Fry challenge studies showed that the viruses can be separated into three groups on the basis of their end point mortality. Viruses rNVI15, rNVI15-15K, rNVI-15KVP1, and rNVI15R, encoding Thr at 217 and Ala at VP2 residue 221, are virulent in nature. Fish started dying 7 to 9 days after challenge, and the mean cumulative mortality ranged from 59% to 70%. Strains rNVI15VP2, rNVI15-15KVP2, and Sp103, encoding Pro at 217 and Ala at 221, caused moderate mortality, ranging from 36% to 47%, and the onset of mortality was delayed 2 to 3 days compared to the virulent strains. The cell culture-adapted virus rNVI15C, and field strain Sp103-VP5, both encoding Thr at residue 221, form a separate, almost avirulent group, irrespective of the amino acid residue at position 217 (Table 1).
To pinpoint the amino acids involved in virulence of IPNV, we generated two chimeric strains, rNVI15VP2 and rNVI15-15KVP2, in which residues 217 (Thr Pro) and 247 (Thr Ala) were changed. Comparison of the deduced amino acid sequences of several IPNV-Sp strains revealed only four amino acid differences at positions 217, 221, 247, and 500 in VP2 protein (Table 1) (31, 33). Out of these four, only two residues, Thr217 and Ala221, are common to all virulent IPNV strains, indicating that VP2 residue 247 (which could be either Thr or Ala) is not directly involved in the virulence characteristics of IPNV. Therefore, reduction in mortality observed for the chimeric strains rNVI15VP2 and rNVI15-15KVP2, compared to that for parent strains rNVI15 and rNVI15-15K, is due to Pro at position 217.
Results from the fry challenge demonstrate that the field strain Sp103 is a moderately virulent strain with a cumulative mortality of 36%. This is significantly higher than in preceding studies, where the same strain was essentially avirulent, giving rise to 5 to 7% mortality (31, 33). Previously, we used commercially bred Atlantic salmon fry, obtained after genetic selection based on several criteria, including disease resistance against IPNV. In the present study, wild Atlantic salmon fry were used that had not been subject to any genetic selection at all; using this group of fish made it possible to separate virus strains at the low end of the virulence spectrum. These findings underline the importance of host-virus interaction in the outcome of virus infection and should be taken into consideration when planning and conducting IPNV challenge studies. An interesting approach would be to challenge different families of fish originating both from wild and commercially bred Atlantic salmon with IPNV strains of different virulence characteristics.
Previous studies have shown that IPNV field isolates tend to lose virulence after serial passage in cell culture (25). However, no attempts have been made to explain the underlying mechanisms. In our study, the major capsid protein VP2 carried the adaptation mutation, which is a substitution of residue 221 from Ala to Thr. There are two lines of evidence indicating that this change is responsible for rapid adaptation of the viruses to CHSE-214 cells. First, the cell-adapted virus rNVI15C replicates much faster and produces larger plaques than the parental strain (rNVI15). Secondly, the recombinant viruses rNVI15VP2, rNVI15-15K, rNVI15-15KVP2, and rNVI15-15KVP1 also acquired this adaptation mutation after passages in CHSE-214 cells.
Similar to rNVI15 virus, adaptation mutation at position 221 was rapidly acquired by rNVI15-15 and rNVI15-15KVP1 viruses. For strains Sp103, rNVI15VP2, and rNVI15-15KVP2, the mutation appeared after nine passages in CHSE-214 cells. The strains encoding Pro at position 217 and Ala at 247 grew to higher titers in both CHSE-214 and RTG-2 cells, compared to strains encoding Thr at 217 and Thr at 247, whereas the adaptive Ala Thr substitution at position 221 enhanced growth only in CHSE-214 cells. It is known that RNA viruses have the ability to quickly adapt to their environment. Alterations at viral entry steps, including receptor recognition, virus attachment and penetration, can affect virus growth and tissue tropism (20). For enveloped virus, mutations acquired during tissue adaptation are often located in the envelope proteins. For instance, it is shown that adaptive mutations in E protein selected by serial passage of Murray Valley encephalitis virus in cell culture affects virus binding to target cells (26). Similarly, mutations in tick-borne encephalitis virus E protein acquired during cell culture adaptation reduce virulence to animals (23). In nonenveloped viruses, adaptive mutations usually occur in outer capsid proteins, functioning as viral cell-binding and receptor recognition proteins. Like other nonenveloped viruses, IPNV is internalized by receptor-mediated endocytosis, and VP2 appears to be the cell attachment protein (7, 13, 19). For IBDV, it was recently shown that cell culture adaptation mutations reside in the VP2 region and facilitate virus entry into cells (2, 21). In regards to IPNV, a possible explanation for the growth characteristics of the various strains and mutants is that residue 217, perhaps in combination with 247, is involved in binding to common receptors present in both CHSE-214 and RTG-2 cells, whereas residue 221 may recognize a specific receptor required for virus entry in CHSE-214 cells but not in RTG cells. Our results from the fluorescent focus assay indicate that Ala Thr substitution at position 221 of the cell culture-adapted strain (rNVI15C) facilitates entry into CHSE-214 cells. Previously, Kuznar and coworkers have shown that IPNV attaches specifically and nonspecifically to CHSE-214 cells, but only specific binding leads to productive infection (19). Recently, Imajoh and colleagues demonstrated that a marine birnavirus binds to a 250-kDa protein in both susceptible and resistant cell lines, suggesting that a common receptor is present in all the cell lines but that a coreceptor is required for the virus to penetrate the cytoplasm by endocytosis in the susceptible cells (18). Further characterization of the cellular receptors for IPNV in different cell lines and crystallization of virus particles is necessary to understand the mechanism of virus-cell interaction in IPNV infection. More importantly, the same Ala-to-Thr substitution at position 221 in VP2 arising after cell culture adaptation was recently established as a molecular determinant for the establishment of persistent IPNV infection (32). This could result from different tissue tropism of the wild-type virus (Ala221) and the mutated virus (Thr221) during acute infection (exocrine pancreatic and liver tissue) and persistent infection (macrophages). Alternatively, the emergence of virus mutants could derive from selective pressure on the viral population by the immune response on epitopes recognized either by neutralizing antibodies or by cytotoxic T lymphocytes. Residue 221 lies within the central variable domain of VP2 containing the major conformational epitopes recognized by neutralizing MAbs (11, 35). Interestingly, the mutant strain rNVI15C can be differentiated from rNVI15 by MAb 122, indicating that this adaptive mutation changes important surface epitopes in the major capsid protein, which leads to the loss of virulence and establishment of persistent infection in fish. Incidentally, the crystal structure of IBDV was recently published, and the sequence alignment of IPNV VP2 with IBDV VP2 reveals that residues at positions 217 and 221 are indeed located in the outmost loops of the P domain, where neutralization-escape mutations occur (5).
Even after nine passages of the recombinant viruses in RTG-2 cells, not a single amino acid substitution was detected in segments A or B. Consequently, rNVI15R maintained its virulence in challenged fry. Other investigators, however, have detected attenuated mutants after passage of IPNV in RTG-2 cells (8). CHSE-214 cell-adapted virus rNVI15C did not produce higher titers than its parental virus in RTG-2 cells, indicating that Ala at position 221 does not enhance virus growth in RTG-2 cells. Blaney and coworkers observed that dengue virus adaptation mutations only occur after serial passage into Vero cells but not in C6/36 cells. Therefore, C6/36 cells were used to propagate the virus to ensure that no adverse cell adaptation mutations were introduced in the viral genome (1). Based on our findings for IPNV adaptation, RTG-2 would be a better cell line for propagating virulent viruses, at least up to 10 passages, to prevent the emergence of cell-adapted mutants.
By making reassortant viruses between two serotypes, VR299 and Ab, Sano and coworkers demonstrated that virulence of IPNV is associated with genomic segment A (29). Using the same approach, they also found that plaque size was dependent on segment A and that it was not associated with virulence (30). In lymphocytic choriomeningitis virus, a single amino acid change in the viral polymerase (residue 1079) is a major determinant of macrophage tropism and virulence (24). Recent studies in our laboratory have shown that VP1 of IBDV affects viral replication kinetics both in vitro and in vivo (21). Therefore, to exclude the possibility that segment B/VP1 is involved in virulence differences among IPNV strains within the same serotype, a reassortant virus was generated in this study. Our results indicate that VP1 is not involved in virulence and adaptation of virus to cell culture.
In conclusion, we have demonstrated that VP2 carries the determinants for IPNV virulence and cell culture adaptation. Virulent strains have Thr217 and Ala221, moderate- to low-virulence strains have Pro217 and Ala221, and strains with Thr221 are almost avirulent. An Ala Thr substitution occurs after passage in CHSE-214 cells but not in RTG-2 cells. The same mutation is selected in persistently infected fish, indicating that this specific change has biological relevance as well. The ability to separate moderate- and/or low-virulence strains from the avirulent strains is dependent on the fish strain used for challenge experiments, underlining the importance the of host-virus interaction in understanding viral infection.
ACKNOWLEDGMENTS
The project was supported by the National Research Initiative of the USDA Cooperative State Research, Education, and Extension Service, grant number 2001-35204-10065 to V.N.V., and a grant from the Norwegian Research Council (number 134136/120) to . E.
We thank Anders N. Haukvik at Haukvik Kraft-Smolt AS and Arnfinn Aunsmo at VESO Trondheim for providing the wild salmon fry; Anne Berit Romstad and Anne Ramstad, study monitors at VESO Vikan; and Gerard H. Edwards for technical assistance.
H.S. and N.S. contributed equally to this work.
REFERENCES
Blaney, J. E., G. G. Manipon, C. Firestone, D. H. Johnson, C. T. Hanson, B. R. Murphy, and S. S. Whitehead. 2003. Mutations which enhance the replication of dengue virus type 4 and an antigenic chimeric dengue virus type 2/4 vaccine candidate in Vero cells. Vaccine 21:4317-4327.
Brandt, M., K. Yao, M Liu, R. A. Heckert, and V. N. Vakharia. 2001. Molecular determinants of virulence, cell tropism, and pathogenic phenotype of infectious bursal disease virus. J. Virol. 75:11974-11982.
Bruslind, L. D., and P. W. Reno. 2000. Virulence comparison of three Buhl-subtype isolates of infectious pancreatic necrosis virus in brook trout fry. J. Aquat. Anim. Health 12:301-315.
Calvert, J. G., E. Nagy, M. Soler, and P. Dobos. 1991. Characterization of the VPg-dsRNA linkage of infectious pancreatic necrosis virus. J. Gen. Virol. 72:2563-2567.
Coulibaly, F., C. Chevalier, I. Gutsche, J. Pous, J. Navaza, S. Bressanelli, B. Delmas, and F. A. Ray. 2005. The birnavirus crystal structure reveals structural relationships among icosahedral viruses. Cell 120:761-772.
Delmas, B., F. S. B. Kibenge, J. C. Leong, E. Mundt, V. N. Vakharia, and J. L. Wu. 2005. Birnaviridae, p. 561-569. In C. M. Fauquet, M. A. Mayo, J. Maniloff, U. Desselberger, and A. L. Ball, (ed.), Virus taxonomy. Academic Press, London, United Kingdom.
Dobos, P. 1995. The molecular biology of infectious pancreatic necrosis virus (IPNV). Annu. Rev. Fish Dis. 5:24-54.
Dorson, M., J. Castric, and C. Torchy. 1978. Infectious pancreatic necrosis virus of salmonids: biological and antigenic features of a pathogenic strain and of a non-pathogenic variant selected in RTG-2 cells. J. Fish. Dis. 1:309-320.
Duncan, R., C. L. Mason, E. Nagy, J. A. Leong, and P. Dobos. 1991. Sequence analysis of infectious pancreatic necrosis virus genome segment B and its encoded VP1 protein: a putative RNA-dependent RNA polymerase lacking the Gly-Asp-Asp motif. Virology 181:541-552.
Duncan, R., E. Nagy, P. J. Krell, and P. Dobos. 1987. Synthesis of infectious pancreatic necrosis virus polyprotein, detection of a virus encoded protease, and fine structure mapping of genome segment A coding regions. J. Virol. 61:3655-3664.
Frost, P., L. S. Hvardstein, B. Lygren, S. Sthl, C. Endresen, and K. E. Christie. 1995. Mapping of neutralization epitopes on infectious pancreatic necrosis viruses. J. Gen. Virol. 76:1165-1172.
Galloux, M., C. Chevalier, C. Henry, J.-C. Huet, B. Da Costa, and B. Delmas. 2004. Peptides resulting from the pVP2 C-terminal processing are present in infectious pancreatic necrosis virus particles. J. Gen. Virol. 85:2231-2236.
Granzow, H., F. Weiland, D. Fichtner, and P. J. Enzmann. 1997. Studies of the ultrastructure and morphogenesis of fish pathogenic viruses grown in cell culture. J. Fish. Dis. 20:1-10.
Hill, B. J. 1982. Infectious pancreatic necrosis virus and its virulence, p. 91-114. In R. J. Roberts (ed.), Microbial diseases of fish. Academic Press, London, United Kingdom.
Hill, B. J., and K. Way. 1995. Serological classification of infectious pancreatic necrosis (IPN) virus and other aquatic birnaviruses. Annu. Rev. Fish Dis. 5:55-77.
Hjalmarsson, A., E. Carlemalm, and E. Everitt. 1999. Infectious pancreatic necrosis virus: identification of a VP3-containing ribonucleoprotein core structure and evidence for O-linked glycosylation of the capsid protein VP2. J. Virol. 73:3484-3490.
Hsu, Y. L., C. C. Chen, and J. L. Wu. 1995. Molecular relationships in infectious pancreatic necrosis virus. Virus Res. 37:239-252.
Imajoh, M., K. Yagyu, and S. Oshima. 2003. Early interaction of marine birnavirus infection in several fish cell lines. J. Gen. Virol. 84:1809-1816.
Kuznar, J., M. Soler, G. Farias, and J. C. Espinoza. 1995. Attachment and entry of infectious pancreatic necrosis virus (IPNV) into CHSE-214 cells. Arch. Virol. 140:1833-1840.
Lee, P., R. Knight, J. M. Smith, J. Wilschut, and D. E. Griffin. 2002. A single mutation in the E2 glycoprotein important for neurovirulence influences binding of Sindbis virus to neuroblastoma cells. J. Virol. 76:6302-6310.
Liu, M., and V. N. Vakharia. 2004. VP1 protein of infectious bursal disease virus modulates the virulence in vivo. Virology 330:62-73.
Magyar, G., and P. Dobos. 1994. Evidence for the detection of the infectious pancreatic necrosis virus polyprotein and the 17-kDa polypeptide in infected cells and of the NS protease in purified virus. Virology 204:580-589.
Mandl, C. W., H. Kroschewski, S. L. Allison, R. Kofler, H. Holzmann, T. Meixner, and F. X. Heinz. 2001. Adaptation of tick-borne encephalitis virus to BHK-21 cells results in the formation of multiple heparan sulfate binding sites in the envelope protein and attenuation in vivo. J. Virol. 75:5627-5637.
Matloubian, M., S. R. Kolhekar, T. Somasundaram, and R. Ahmed. 1993. Molecular determinants of macrophage tropism and viral persistence: importance of single amino acid changes in the polymerase and glycoprotein of lymphocytic choriomeningitis virus. J. Virol. 67:7340-7349.
McAllister, P. E., and W. J. Owens. 1986. Infectious pancreatic necrosis virus: protocol for a standard challenge to brook trout. Trans. Amer. Fish. Soc. 115:466-470.
McMinn, P. C., I. D. Marshall, and L. Dalgarno. 1995. Neurovirulence and neuroinvasiveness of Murray Valley encephalitis virus mutants selected by passage in a monkey kidney cell line. J. Gen. Virol. 76:865-872.
Nicholson, B. L. 1993. Use of monoclonal antibodies in identification and characterization of fish viruses. Annu. Rev. Fish Dis. 3:241-257.
Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
Sano, M., N. Okamoto, H. Fukuda, M. Saneyoshi, and T. Sano. 1992. Virulence of infectious pancreatic necrosis virus is associated with the larger RNA segment (RNA segment A). J. Fish. Dis. 15:283-293.
Sano, M., N. Okamoto, and T. Sano. 1994. Infectious pancreatic necrosis virus plaque size depends on the larger RNA segment and is not linked to virulence. J. Fish. Dis. 17:657-659.
Santi, N., V. N. Vakharia, and . Evensen. 2003. Identification of putative motifs involved in the virulence of infectious pancreatic necrosis virus. Virology 322:31-40.
Santi, N., H. Song, V. N. Vakharia, and . Evensen. 2005. Infectious pancreatic necrosis virus VP5 is dispensable for virulence and persistence. J. Virol. 79:9206-9216.
Shivappa, R. B., H. Song, K. Yao, A. Aas-Eng, . Evensen, and V. N. Vakharia. 2003. Molecular characterization of Sp serotype strains of infectious pancreatic necrosis virus exhibiting differences in virulence. Dis. Aquat. Org. 61:23-32.
Silim, A., M. A. S. Y. Elazhary, and A. Lagace. 1982. Susceptibility of trouts of different species and origins to various isolates of infectious pancreatic necrosis virus. Can. J. Fish Aquat. Sci. 39:1580-1584.
Tarrab, E., L. Berthiaume, S. Grothé, M. O'Connor-McCourt, J. Heppel, and J. Lecomte. 1995. Evidence of a major neutralizable conformational epitope region on VP2 of infectious pancreatic necrosis virus. J. Gen. Virol. 76:551-558.
Weber, S., D. Fichtner, T. C. Mettenleiter, and E. Mundt. 2001. Expression of VP5 of infectious pancreatic necrosis virus strain VR299 is initiated at the second in-frame start codon. J. Gen. Virol. 82:805-812.
Wolf, K. 1988. Fish viruses and fish viral diseases. Canstock Publishing Associates-Cornell University Press, Ithaca, N.Y.
Yao, K., and V. N. Vakharia. 1998. Generation of infectious pancreatic necrosis virus from cloned cDNA. J. Virol. 72:8913-8920.(Haichen Song, Nina Santi,)