Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Virology 342 (2005) 13 – 25 www.elsevier.com/locate/yviro Infectious pancreatic necrosis virus induces apoptosis in vitro and in vivo independent of VP5 expression Nina Santi a, Ane Sandtrø b, Hilde Sindre c, Haichen Song d, Jiann-Ruey Hong e, Beate Thu b, Jen-Leih Wu f, Vikram N. Vakharia d, Øystein Evensen b,* a Section for Pathology, National Veterinary Institute, 0033 Oslo, Norway Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, PO Box 8146 Dep., N-0033 Oslo, Norway c Section for Virology, National Veterinary Institute, 0033 Oslo, Norway d Center for Biosystems Research, University of Maryland Biotechnology Institute and VA-MD Regional College of Veterinary Medicine, University of Maryland, College Park, MD 20742, USA e Laboratory of Molecular Virology and Biotechnology, Institute of Biotechnology, National Cheng-Kung University, Tainan 701, Taiwan f Laboratory of Marine Molecular Biology and Biotechnology, Institute of Zoology, Academia Sinica, Taipei 115, Taiwan b Received 28 February 2005; returned to author for revision 25 April 2005; accepted 14 July 2005 Available online 26 August 2005 Abstract Infectious pancreatic necrosis virus (IPNV), the causative agent of a highly infectious disease in salmonid fish, encodes a small nonstructural protein designated VP5. This protein contains Bcl-2 homologous domains and inhibits apoptosis when expressed in cell culture. We have previously reported the generation of three VP5 mutants of IPNV – Sp serotype, using reverse genetics (Santi, N., Song, H., Vakharia, V.N., Evensen, Ø, 2005. Infectious pancreatic necrosis virus VP5 is dispensable for virulence and persistence. J. Virol. 79 (14), 9206 – 9216). The wild-type rNVI15 virus encodes a truncated 12-kDa VP5 protein, rNVI15-15K encodes a full-length 15-kDa VP5, whereas rNVI15-DVP5 is deficient in VP5 expression. In the present report, the role of VP5 in apoptosis was assessed both in vitro and in vivo, using the recombinant IPNV strains. Apoptosis was observed in hepatocytes of Atlantic salmon post-smolts challenged with all three VP5 mutant viruses. Using a double-labeling technique to detect apoptotic cells and IPNV antigens, we found that viral antigen and apoptotic cells codistributed. In addition, numerous double-positive cells were seen. The recombinant viruses also induced apoptosis in infected cell cultures, and the morphology and membrane integrity of infected cells at different time points was similar. In summary, these results indicate that IPNV induces apoptosis in infected cell cultures and in fish, independent of VP5 expression. However, substitutions of putative functionally important amino acids in the BH2 domain of VP5 of IPNV – Sp strains were identified, which might influence the anti-apoptosis effect of the protein, and partly explain the apparent absence of this specific function. D 2005 Elsevier Inc. All rights reserved. Keywords: IPNV; VP5; Apoptosis; Bcl-2 Introduction Infectious pancreatic necrosis virus (IPNV) is the causative agent of an economically important disease in farmed Atlantic salmon (Salmo salar L.), infectious pancreatic necrosis (IPN). Outbreaks are typically seen at fry start feeding, and shortly after sea transfer of smolts. As * Corresponding author. Fax: +47 22597310. E-mail address: [email protected] (Ø. Evensen). 0042-6822/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.virol.2005.07.028 indicated by the name, necrosis of exocrine pancreatic tissue is the most common manifestation of the disease. In addition, severe necrosis has been observed in the liver of experimentally infected Atlantic salmon fry (Taksdal et al., 1997; Santi et al., 2004). Following acute infection, survivors continue to harbor virus in their bodies for long periods, presumably in head kidney macrophages (Reno et al., 1978; Mangunwiryo and Aguis, 1988; Johansen and Sommer, 1995). IPNV is a member of the genus Aquabirnavirus in the family Birnaviridae. The non-enveloped icosahedral capsid 14 N. Santi et al. / Virology 342 (2005) 13 – 25 has a diameter of 60 nm and contains two genome segments (A and B) of double-stranded RNA (Dobos, 1995). The smaller segment B encodes the virus polymerase, VP1 (Duncan et al., 1991). Segment A contains a large open reading frame (ORF) encoding a 107-kDa polyprotein that is processed into the major structural proteins VP2 and VP3 by the viral protease (VP4) (Macdonald and Dobos, 1981; Duncan et al., 1987; Petit et al., 2000). A second ORF, overlapping the N-terminal region of VP2, encodes the small non-structural protein, VP5 (Håvardstein et al., 1990; Magyar and Dobos, 1994). Apoptosis is a genetically controlled process of cell suicide in response to a variety of stimuli. Apoptosis is considered a part of the innate immune response to virus infection, limiting the time and cellular machinery available for viral replication (Everett and McFadden, 1999). Apoptotic cell death in tissues typically involves scattered individual cells, recognized by certain morphological features, such as detachment from neighboring cells, condensation and aggregation of chromatin along the inner surface of the nuclear membrane, and segmentation of the nucleus and the cell into ‘‘apoptotic bodies’’ that are rapidly engulfed by adjacent cells or by macrophages (Kerr et al., 1995). Infectious bursal disease virus (IBDV), another member of the Birnaviridae family, induces apoptosis in peripheral blood lymphocytes and in tissue from infected chickens (Vasconcelos and Lam, 1994; Lam, 1997; Tanimura and Sharma, 1998). Apoptosis precedes necrosis in cultured fish cells infected with IPNV (Hong et al., 1998), but so far attempts to detect apoptosis in IPNV-infected fish have been inconclusive (Eléouët et al., 2001; Imajoh et al., 2005). IBDV also contains a small ORF that encodes a 17kDa non-structural protein called VP5 (Mundt et al., 1995). VP5 of both IBDV and IPNV is dispensable for viral replication in cell culture (Mundt et al., 1997; Weber et al., 2001; Santi et al., 2005) but their function appears to be different. Expression of IBDV-VP5 in cell culture leads to accumulation of the protein in the cell membrane, alteration in membrane permeability, and induction of apoptosis (Lombardo et al., 2000; Yao and Vakharia, 2001). A recombinant IBDV mutant, deficient in VP5 expression, has decreased cytotoxic and apoptotic effect in cell culture and does not induce lesions in the bursa of inoculated chickens (Yao et al., 1998). IPNV-VP5, on the other hand, contains Bcl-2 homologous domains, and stable expression enhances cell viability following exposure to apoptotic inducers (Hong et al., 2002). To date, there are no studies on the role of this protein in IPNV-induced apoptosis using strains deficient in VP5 expression. We recently reported the recovery of three VP5 mutants of IPNV by reverse genetics (Santi et al., 2005). The recombinant rNVI15-DVP5 is a VP5 knockout mutant virus, whereas rNVI15 encodes a truncated 12-kDa VP5, and rNVI15-15K encodes the full-length 15-kDa protein. The replication kinetics of the three viruses in cell culture is similar, and they all cause IPN disease in challenged Atlantic salmon post-smolts, resulting in a cumulative mortality of 86%, 81%, and 81%, respectively (Santi et al., 2005). Since both cellular and viral Bcl-2 homologues have been associated with viral persistence (Levine et al., 1993; Gangappa et al., 2002), we assessed the involvement of VP5 in persistent IPNV infection. The results demonstrated that VP5 is not required for the establishment of a carrier state. Thus, the physiological effect of VP5 expression was not apparent from in vivo studies of VP5 mutant strains. In the present study, we analyze apoptosis induction in tissues during acute IPNV infection in Atlantic salmon post-smolts and evaluate the effect of VP5 expression on the kinetics of IPNV-induced apoptosis in cell culture using recombinant IPNVs of the Sp serotype. Results In vitro infection of cells IPNV-infected cell cultures exhibit cytopathic effects (CPE) as the cells die from virus infection. We wanted to determine if VP5 expression influence morphological changes in IPNV-infected cell cultures. CHSE-214 cells were infected with multiplicity of infection (MOI) 1, using three IPNV-VP5 mutant strains. Morphological changes were observed by phase contrast microscopy at regular intervals, and the monolayers were photographed at 0, 24, 32, and 36 h post-infection (p.i.). CPE was first detected 24 h p.i., affecting single cells or cells in small groups (Fig. 1). A slight increase in the number of affected cells was observed at 32 h p.i. At the 36-h time point, almost full CPE was observed, as most cells had detached from the monolayer. There was no apparent difference in the morphology of monolayers infected with the VP5-deficient strain compared to cells infected with the VP5-expressing strains at any time point throughout the observation period. This experiment was also performed at MOI = 0.1 and MOI = 10, again demonstrating a synchronized development of CPE in CHSE-214 cells infected with the VP5 mutant IPNV strains (data not shown). Alterations in membrane permeability Studies of IBDV-VP5 have demonstrated accumulation of the protein within the plasma membrane, resulting in cell lysis by disruption of the plasma membrane (Lombardo et al., 2000). To determine if IPNV-VP5 expression has any effect on the membrane permeability of infected cells, lactate dehydrogenase (LDH) activity in supernatants of cell infected with rNVI15, rNVI15-15K, and rNVI-DVP5 viruses was measured and compared with mock-infected cells (Fig. 2). For all three recombinant strains, a slight decrease in LDH activity was detected at 16 and 20 h p.i., followed by a steady increase and reaching maximum activity at 55 h. There was no statistical difference in LDH N. Santi et al. / Virology 342 (2005) 13 – 25 15 Fig. 1. Phase contrast micrographs of CHSE-214 cells infected with three recombinant IPNV – Sp VP5 mutants at an MOI = 1.00. Scale bar = 100 Am. Fig. 2. Lactate dehydrogenase (LDH) activity in supernatants from CHSE-214 cells infected with three recombinant IPNV – Sp VP5 mutants at an MOI = 1.00. The data shown are mean of three parallel samples, and error bars indicate SD. The values for control cells and lysed cells are represented by the mean of 24 independent samples; the range of absorbance values was 0.154 – 0.186 for control cells and 0.391 – 0.542 for lysed cells. 16 N. Santi et al. / Virology 342 (2005) 13 – 25 activity between these three viruses at any time point during the observation period. Induction of apoptosis in vitro Hong et al. (1998 and 2002) have previously shown that apoptosis precedes necrosis in IPNV-infected CHSE-214 cells, and that VP5 overexpression enhances host viability and delays induction of internucleosomal DNA cleavage induced by IPNV. When CHSE-214 monolayers were infected with the recombinant strains, the characteristic apoptotic laddering of DNA started to occur at 16 –18 h p.i. for all three viruses (Fig. 3A). The bands were weak at 16, 18, and 20 h, but from 24 h onwards strong DNA laddering was observed (Fig. 3B). Similar results were obtained in another independent experiment (data not shown). Thus, it was not possible to detect delayed induction of internucleosomal cleavage by VP5-expressing strains compared to the VP5-deficient strains. Annexin V binds to phosphatidylserine exposed at the external cell surface of apoptotic cells. Flow cytometric analysis of annexin V/propidium-iodide-stained CHSE-214 cells was performed for quantitative measurement of apoptosis induced by the three recombinant VP5 strains. Results from this experiment indicated a slight antiapoptotic effect due to VP5 expression. The proportion of apoptotic cells was similar at 0, 16, 20, and 32 h time points (Fig. 4A). At the 24-h time point, there were more apoptotic cells in cultures infected by rNVI15-DVP5 (11%) compared to rNVI15 (7%) and rNVI15-15K (4%) (Fig. 4A). The number of dead cells did not rise considerably before the 32-h time point in any of the infected cell cultures (Fig. 4B). We previously reported a substitution of residue 221 in VP2 (Ala to Thr) of IPNV isolates propagated in CHSE-214 cells (Santi et al., 2004). We recently discovered that strains encoding Ala at this position have significantly reduced infectivity to CHSE-214 cells compared to cell-cultureadapted strains (Song et al., 2005). The virus stocks used in this study have the same low passage number in CHSE cells, and nucleotide sequencing has shown that mutants are present at a very low level (data not shown). However, an uneven distribution of these mutants could influence the outcome of the flow cytometric assay, and consequently it was decided to repeat this experiment with rNVI15-15 and rNVI15-DVP5 free of cell-culture-adapted mutants. In this experiment, apoptosis was induced at the same time and at comparable levels in cells infected with the VP5-encoding and the VP5-deficient strain (Fig. 4C). The percentage of dead cells was however significantly higher ( P > 0.05) in cells infected with rNVI15-DVP5 compared to cells infected with rNVI15-15K at 42, 48, and 54 h p.i. (Fig. 4D). A relatively high number of dead cells were observed both for virus infected and mock-infected cells at the 0-h time point, this is probably related to rough handling of the cells during harvest. Induction of apoptosis in vivo To investigate the role of apoptosis in the pathogenesis of IPN, tissue sections of liver and pancreas from Atlantic salmon post-smolts experimentally infected with rNVI15, rNVI15-15K, and rNVI-DVP5 viruses were sampled for histopathological examination. In liver samples harvested 7 days after infection, single cell death of hepatocytes morphologically resembling apoptosis was detected in fish challenged with all three recombinant viruses. After 9 days and onwards, apoptotic cells were found scattered in the liver parenchyma of most fish examined, except for control fish (Fig. 5). Apoptosis was confirmed by in situ labeling of DNA fragmentation using the TUNEL assay. Positive cells were detected in the liver of fish infected with all these viruses, but not in control fish (Fig. 6). Morphological evidence of apoptosis was also detected in the exocrine pancreas of infected cells; however, intensive viral replication followed by secondary necrosis and high enzymatic activity in the exocrine pancreatic cells resulted in high level of background staining in the TUNEL-labeled sections (data not shown). In liver sections, the distribution of viral antigen detected by immunohistochemistry seemed to overlap the distribution of apoptotic cells (Fig. 7). This observation was verified using double staining with TUNEL and IPNV immunohistochemistry. Viral antigen and apoptotic cells codistributed, and numerous double-positive cells were detected, although the darker stain of the TUNEL assay (black color) seemed to mask some of the IPNV immunostaining (red color) (Fig. 8). Sequence alignments of IPNV VP5 Results from the alignment of VP5 protein of rNVI1515K and rNVI15 viruses with cellular and viral Bcl-2 homologous proteins revealed structural differences in the BH2 domain of VP5 compared to the E1S strain used by Hong et al. (2002) (Fig. 9). Recombinant rNVI15 virus has a premature stop codon at nucleotide 427 and encodes a putative 12-kDa VP5, corresponding to a BH2 deletion mutant. Several amino acid substitutions were detected in the BH2 domain of rNVI15-15K-VP5 when compared to IPNV-E1S. In particular, two arginine (Arg) residues (115 plus 121) of rNVI15-15K-VP5 have replaced two tryptophan (Trp) residues (129 plus 135) of the E1S-VP5 BH2 homology domain (Fig. 9). Discussion In this study, we demonstrate for the first time the induction of apoptosis in hepatocytes of IPNV-infected Atlantic salmon post-smolt. In a previous study of IPNVinfected rainbow trout (Oncorhynchus mykiss) fry, apoptotic nuclei were not clearly demonstrated in the pancreas but were widespread in the lamina propria of the intestine N. Santi et al. / Virology 342 (2005) 13 – 25 17 Fig. 3. Agarose gel electrophoresis of DNA isolated from CHSE-214 cells infected with three recombinant IPNV – Sp VP5 mutants at an MOI = 1.00. (A) Cells were collected at 14, 16, 18, 20, and 24 h post-infection. (B) Cells were collected 20, 24, 28, 32, and 36 h post-infection. Lane M: 1 kb+ marker. Lane C: uninfected control cells. (Eléouët et al., 2001). Further, in a previous IPNV challenge study with Atlantic salmon fry, diffuse necrosis of hepatic and exocrine pancreatic tissues was found in diseased fish (Santi et al., 2004). These lesions developed over short time span, involving most of the liver and pancreas within a day or two after onset of mortality. Minor lesions were detected in the gastric glands, and in contrast to the necrosis found in liver and pancreas, the affected cells in gastric glands often displayed the morphological features of apoptotic cells (Santi et al., 2004). Host cell apoptosis is part of the innate immune response to virus infection, serving to limit the progeny production by 18 N. Santi et al. / Virology 342 (2005) 13 – 25 Fig. 4. CHSE-214 cells were mock infected or infected with various recombinant IPNV strains at an MOI = 1.00 and incubated at 15 -C before annexin V/ propidium iodide staining. (A) Percentage of apoptotic cells (Annexin-V-positive/propidium-iodide-negative cells) at different time points in cell cultures infected with IPNV – Sp strains encoding the full-length VP5 protein (rNVI15-15K), a putative 12-kDa VP5 (rNVI15), and a VP5-deficient strain (rNVI15DVP5). (B) Percentage of dead cells (annexin V/propidium iodide double-positive cells) at corresponding time points. (C) Percentage of apoptotic cells at different time points in CHSE-214 cells infected with virus stocks of one VP5-encoding (rNVI15-15K) and one VP5-deficient (rNVI15-DVP5) strain free of cell-culture-adapted mutants. (D) Percentage of dead cells at corresponding time points. In panels A and C, the percentage of apoptotic cells is calculated from the population of viable cells after the necrotic propidium iodide has been gated away. In panels B and D, the percentage of necrotic cells is calculated from the total number of cells. the virus. Double-stranded RNA (dsRNA) formed during viral replication, in addition to being a potent inducer of interferon (IFN), activates protein kinase R (PKR), leading to phosphorylation of the eukaryotic translational initiation factor 2 (eIF2a) and inhibition of protein synthesis. Further, dsRNA also induces apoptosis through PKR activation, and IFN sensitizes cells to this dsRNA-mediated apoptosis (Barber, 2001; Clemens, 2003; Gil and Esteban, 2000; Kibler et al., 1997). Evidence points to the activation of the PKR pathway in IPNV-infected fish cells, as IPNV infection in RTG-2 cells leads to increased eIF2a phosphorylation (Garner et al., 2003). This suggests that apoptosis induction in IPNV-infected cells could be linked to the activation of PKR. However, IPNV induces non-typical apoptotic cell morphology in CHSE-cells, characterized by the formation of small holes in the membrane, associated with loss of integrity, finally leading to post-apoptotic necrosis (Hong et al., 1999). It is conceivable that in the battle between host cell and virus, the cell, in an effort to limit the viral replication, responds by induction of apoptosis. The virus, on the other hand, will aim at producing as many progeny as possible, with a final release of virus particles from the infected cell through lysis/necrosis. The outcome for an IPNV-infected cell might depend upon the host cell’s ability N. Santi et al. / Virology 342 (2005) 13 – 25 19 Fig. 5. HE staining of liver sections from Atlantic salmon post-smolts mock infected (A) or infected with IPNV strains rNVI15-15K (B), rNVI15 (C), or rNVI15-DVP5 (D). Single rounded cells with condensed cytoplasm and chromatin aggregation morphologically resembling apoptotic cells (arrows) is seen in the livers of IPNV-infected fish, but not in control fish. The livers sections shown were selected from representative fish sampled at 20 (A), 10 (B), 14 (C), and 13 (D) days p.i., respectively. to mount an apoptotic response to the virus infection. If true, this may provide an explanation for the differences observed in the liver pathology of IPNV infected fry compared to post-smolts. If the virus has the upper hand, it will replicate at high rate, resulting in production of a high number of progeny, causing cell lysis and necrosis at the end. Rapidly developing diffuse liver necrosis is typically seen in IPNV infected fry. When the balance is in favor of the host cell, Fig. 6. TUNEL assay of liver sections from Atlantic salmon post-smolts mock infected (A) or infected with IPNV strains rNVI15-15K (B), rNVI15 (C), or rNVI15-DVP5 (D). TUNEL-positive cells (black) are seen in the livers of infected fish, but not in control fish. Aggregation of chromatin, characteristic for apoptotic cells, is observed (arrows). The livers sections shown were selected from representative fish sampled at 20 (A), 10 (B), 14 (C), and 13 (D) days p.i., respectively. 20 N. Santi et al. / Virology 342 (2005) 13 – 25 Fig. 7. Immunohistochemistry of liver sections from Atlantic salmon post-smolts mock infected (A) or infected with IPNV strains rNVI15-15K (B), rNVI15 (C), or rNVI15-DVP5 (D). IPNV antigen-positive cells (red) are seen in the livers of infected cells, but not in control fish. Some of the IPNV-positive cells resemble apoptotic cells (arrows). The livers sections shown were selected from representative fish sampled at 20 (A), 10 (B), 14 (C), and 13 (D) days p.i., respectively. it will enter into apoptosis at an early stage of the infection and limit the production and spread of virus progeny. This is in line with the pathological changes observed in post-smolt livers. Furthermore, interferon sensitizes cells to the apoptosis-inducing effect of dsRNA; consequently, the cells localized around the IPNV-infected cell would respond quicker to the viral infection than the cell initially infected. This is in accordance with our observation that many TUNEL-positive cells encircle the IPNV-positive cells in the liver tissues of post-smolts, and Fig. 8. Dual labeling of liver sections from Atlantic salmon post-smolts mock infected (A) or infected with IPNV strains rNVI15-15K (B), rNVI15 (C), or rNVI15-DVP5 (D) by TUNEL assay and IPNV immunohistochemistry. Viral antigen and apoptotic cells appear to co-distribute, and some double-labeled cells are visible (arrows). The livers sections shown were selected from representative fish sampled at 20 (A), 10 (B), 14 (C), and 13 (D) days p.i., respectively. N. Santi et al. / Virology 342 (2005) 13 – 25 21 Fig. 9. Sequence and structural homology of cellular and viral Bcl-2-related proteins. GenBank accession numbers for the protein sequences used are listed below. Homo sapiens Bcl-2 (Human Bcl-2), P10415; Gallus gallus Bcl-2 (Chicken Bcl-2), BAA01978; H. sapiens Bcl-XL (Human Bcl-XL), CAA80661; Danio rerio Bcl-XL (Zebrafish Bcl-XL), AAK81706; H. sapiens Mcl-1 (Human Mcl-1), BC017197; IPNV E1S NS protein, AAK71697; Epstein – Barr virus BHRF1 (EBV Bhrf-1), QQBE4; human adenovirus type 9 E1B 19K (E1B19K), AAD16304; African swine fever virus LMV-5HL (ASFV 5-HL), AAA17034. ( ) Negative charge, (+) positive charge, (l) aliphatic, (t) tiny, (a) aromatic, (c) charged, (s) small, (p) polar, (h) hydrophobic, (b) big. also in line with previous findings in IBDV-infected chickens (Jungmann et al., 2001). In persistently infected fish, IPNV is detected in macrophages (Johansen and Sommer, 1995; Munro et al., 2004). In a recent study of persistently infected rainbow trout, IPNV-positive cells in leucocyte smears from peripheral blood, head kidney, and spleen did not have apoptotic nuclei (Imajoh et al., 2005). Assuming the IPNV-positive cells to be macrophages, a possible explanation for the lack of apoptotic morphology in these cells is that differentiated macrophages are robust, longlived cells and resistant to numerous death stimuli (Mangan et al., 1991; Munn et al., 1995; Perlman et al., 1999). An alternative explanation for this observation is that persistent infection of macrophages is associated with specific inhibition of host-induced apoptosis. Apoptosis was induced in hepatocytes of fish infected with each of these three VP5 mutants. The absence of VP5 expression did not result in increased apoptosis frequency in hepatocytes of fish infected with rNVI15-DVP5, as judged by observation alone. An attempt to quantitatively measure the level of apoptosis induction by the three VP5 variant viruses was not pursued due to large individual variation in the morphological appearances of livers in diseased fish. Various assays were used in the studies of cell cultures infected with the recombinant viruses. The results from the LDH-assay demonstrated that, unlike VP5 of IBDV, IPNVVP5 expression did not have any detectable effect on cell membrane permeability. However, the uptake of virus apparently triggered a transient reduction in membrane permeability, leading to a reduction in LDH leakage from IPNV-infected cells, as compared to control cells at 16 and 20 h p.i. This is an interesting finding, as it indicates changes in the membrane dynamics of the cell caused either by early apoptotic responses in the host cell or by early replication of the virus. We did not observe differences in the morphological appearance of cells infected with the VP5 mutant strains. No delay in the induction of internucleosomal cleavage was detected in cells infected by VP5-expressing strains compared to the VP5-deficient strain, using the agarose gel electrophoresis assay. In the first flow cytometry experiment, Annexin-V-positive cells appeared at an earlier time in cell cultures infected with the VP5-deficient IPNV strains compared to the VP5-expressing strains, indicating a slight anti-apoptotic effect of VP5 expression. However, this was probably due to an unevenly distributed presence of cell-culture-adapted mutants, as no difference could be detected in apoptosis among cells infected with the VP5encoding and VP5-deficient strain in the second experiment. The percentage of necrotic cells was still slightly higher in cell cultures infected with the VP5 knockout mutant at the later time points of the second experiment. The significance of this finding is uncertain. Using the annexin V/propidium iodide flow cytometric assay, the percentage of apoptotic cells did not rise above 20% at any of the time points examined. It is however important to keep in mind that, in contrast to cells expressing IPNV antigen or being necrotic (propidium iodide positive) cells, the apoptotic (annexin V positive/propidium iodide negative) cells will not accumulate in infected cell cultures. Cells are annexin V positive/propidium iodide negative only for a short period, as the duration of the whole event of apoptosis takes no more than a few hours. In the first paper describing the annexin V/propidium iodide method for detection of apoptotic cells, the maximum percentage of apoptotic cells in relation to time elapsed was 24.5% after irradiation and 8% after dexamethasone treatment (Vermes et al., 1995). Parvovirus B19-induced apoptosis of hepatocytes is shown to be infectious dose dependent, where an MOI of 10 leads to 10% apoptotic cells and an MOI of 1000 gives 27% apoptotic cells as measured by annexin V staining (Poole et al., 2004). In summary, no anti-apoptotic effect of VP5 was detected for the IPNV serotype Sp strains studied. This is not in accordance with the previously reported anti-apoptotic 22 N. Santi et al. / Virology 342 (2005) 13 – 25 effect of VP5. Having said this, there are important structural differences in the VP5 encoded by IPNV –Sp strains compared to the E1S strain used in the study by Hong and coworkers (Fig. 9). The BH2 domain of Bcl-2 is required for inhibition of apoptosis and heterodimerization with Bax (Hunter et al., 1996; Yin et al., 1994). Deletion of the BH2 domain alone or both BH2 and BH1 domains causes a loss of the anti-apoptosis function in the IPNV-E1S VP5 (Hong et al., 2002). The putative 12-kDa VP5 encoded by rNVI15 would correspond to a BH2 deletion mutant, which could result in a complete or partial loss of antiapoptotic activity. The carboxy-terminal part of the 15-kDa VP5 encoded by rNVI15-15K is equivalent (in size) to the E1S-VP5. However, in rNVI15-15K-VP5, two arginine (Arg) residues (115 plus 121) have replaced two tryptophan (Trp) residues (129 plus 135) of the E1S-VP5 BH2 homology domain (Fig. 9). The first of these Trp residues corresponds to the functionally important Trp 188 in Bcl-2, and replacement of this residue with alanine (Ala) abrogates Bcl-2’s inhibition of apoptosis (Yin et al., 1994). For Epstein –Barr virus (EBV) Bcl-2 homolog BHRF1, it was shown that substitution of Trp-143 (corresponding to Trp-188 in Bcl-2) with Ala leads to loss of the ability to confer protection against cis-plantin-induced apoptosis (Khanim et al., 1997). Thus, it cannot be excluded that the 15-kDa VP5 of rNVI1515K corresponds to a functional BH2 deletion mutant, resulting in a reduction of or loss of the anti-apoptotic function of E1S VP5. Furthermore, to the extent that the anti-apoptotic activity of VP5 has any impact on the virulence characteristics of IPNV, these observations could explain the identical virulence characteristics of the three VP5 variant viruses, as they may represent functionally identical virus strains. Interestingly, an alignment of the VP5-encoding sequence of twelve marine birnaviruses and three IPNV strains (Zhang and Suzuki, 2004) shows that all isolates encode Trp in VP5, which corresponds to the functionally important Trp 188 in Bcl-2. The data presented indicate a deficit in the anti-apoptotic function of VP5 in the IPNV – Sp strains studied. Further studies are needed to evaluate the importance of Trp residue in the BH2 homologous domain for possible loss of function. strain rNVI15, recovered using reverse genetics (Santi et al., 2005), is identical to the virulent field strain NVI-015 (GenBank accession nos. AY379740 and AY379741). It encodes a putative 12-kDa VP5 due to a premature stop codon at nucleotide position 427. Recombinant rNVI1515K virus encodes the full-length 15-kDa protein, whereas rNVI15-DVP5 contains a mutation in the initiation codon of VP5 gene, which ablates the expression of VP5. Viruses were propagated by inoculation of the second (fish experiment) or third (cell culture experiments) passage supernatant into CHSE-214 cells, followed by incubation at 15 -C. When the cytopathic effect (CPE) was evident, cell culture supernatants were harvested after a brief centrifugation. Virus stocks free of cell-culture-adapted mutants were prepared by inoculation of RTG-2 cells and confirmed by sequencing as previously described (Song et al., 2005). For fish challenge, the infectious titre was determined by end point dilution on CHSE-214 cells grown in 96-well plates. Fifty microliters of 10-fold dilutions (10 1 to 10 8) of cell culture supernatants was used in six parallel wells per dilution. The TCID50 was estimated by the method of Kärber (1931). For in vitro assays, infectious titre was determined by plaque assay. Briefly, monolayers of CHSE214 grown in 6-well plates were inoculated with 10-fold dilutions (10 3 to 10 8) of cell culture supernatants. After 1 h incubation at 15 -C, the inoculum was removed and the cells overlaid with 0.8% SeaPlaque Agarose (BioWhittaker) in L-15 medium containing 5% FBS and 1% l-glutamine. The cells were incubated for 3 days at 15 -C, and then fixed using methanol and acetic acid (v/v 3:1). Overlays were removed and the plaques were counted after staining the monolayers with Giemsa (Sigma). Observation of morphological changes in cell culture CHSE-214 cells were grown in 24-well plates until 80% confluence and infected with the three IPNV strains at a multiplicity of infection (MOI) of 1.00 virus particles per cell. Three parallel wells were examined for each IPNV strain. After incubation at 15 -C, the cellular morphology was observed using phase contrast microscopy (Olympus) at 0, 24, 32, and 36 h. Parallel experiments were performed at MOI = 0.1 and MOI = 10. Cytotoxicity and apoptosis assays Materials and methods Cells and viruses Chinook salmon embryo cells (CHSE-214) were obtained from the American Type Culture Collection (ATCC) and maintained at 20 -C in Leibovitz’s L-15 medium (BioWhittaker) supplemented with 5% fetal bovine serum (FBS) and 25 Ag/ml gentamicin. Rainbow trout gonad cells (RTG-2, ATCC CCL-55) were grown in L-15 medium supplemented with 10% FBS at 20 -C. The recombinant IPNV serotype Sp Leakage of cytoplasmatic contents due to damage of the plasma membrane of infected cells was analyzed using cytotoxicity detection kit (Roche). CHSE-214 cells were seeded onto 96-well plates and grown until 80% confluence and infected with rNVI15, rNVI15-15K, and rNVI-DVP5 viruses at MOI = 1.00. After 0, 8, 16, 20, 24, 28, 32, 36, 40, 48, and 54 h of incubation, the supernatants were collected and the lactate dehydrogenase (LDH) activity was measured according to the manufacturer’s instructions. N. Santi et al. / Virology 342 (2005) 13 – 25 DNA fragmentation assay was performed using apoptotic DNA ladder kit (Roche) in accordance with the manufacturer’s instructions. Briefly, CHSE-214 cells were grown in 6-well plates and infected with rNVI15, rNVI15-15K, or rNVI15-DVP5 viruses at MOI = 1.00. After 14, 16, 18, 20, 24, 28, 32, and 36 h of incubation, both detached and adherent cells were collected, resuspended in binding/lysis buffer, and incubated at room temperature for 10 min. DNA was isolated using a column provided with the kit and eluted in a total volume of 200 Al. After RNase A digestion (25 Ag/ ml) for 20 min at room temperature, DNA was dried and dissolved in 30 Al dH2O. The samples were loaded onto a 1% agarose gel and electrophoresed at 50 V for 2 h, followed by 70 V for 2 h. Gels were stained with ethidium bromide and photographed under UV transillumination. FITC – annexin V/PI double staining The initial experiment was performed with CHSE-214 cells grown in 24-well plates and mock infected or infected with rNVI15, rNVI15-15K, or rNVI15-DVP5 viruses (MOI = 1.00) and incubated at 15 -C for 0, 16, 20, 24, and 32 h. Both adherent and floating cells were collected and washed with PBS before annexin V/propidium iodide (PI) treatment, using the Annexin-V-FLUOS staining kit in accordance with the manufacturer’s instructions (Roche). Samples were analyzed on a FacsCalibur flow cytometer (Becton Dickinson), and the WinMDI 2.8 software (J. Trotter) was used to determine the percentages of viable, apoptotic, and necrotic cells. In short, the PI/annexin V double-positive cells were gated and defined as necrotic cells, and the remaining cell population was defined as viable cells and investigated for the expression of annexin V. The percentage of annexin V binding cells at the different time point was then calculated. In the following experiment with virus stocks free of cell-culture-adapted mutants, CHSE-214 cells grown in 24-well plates were mock infected or infected with rNVI15-15K or rNVI15-DVP5 at MOI = 1. Due to the low infectivity of non-attenuated viruses (Song et al., 2005), cells were harvested after 0, 30, 36, 42, 48, and 54 h p.i. Three parallel wells were sampled for each time point. Cells were further analyzed as described above, except that the staining reagents were supplied from a different manufacturer (annexin V – FITC apoptosis detection kit I, BD Pharmingen). Atlantic salmon post-smolt challenge The challenge study with three VP5 mutant IPNVs was described previously (Santi et al., 2005). The challenge was conducted at VESO Vikan’s research facility in Namsos, Norway, and approximately 2200 Atlantic salmon (S. salar L.) parr (average weight 70 g) going through smoltification were used in the study. Fish were challenged by 1 h immersion with IPN virus at a concentration of 104 TCID50/ml. Fish in three parallel tanks were infected with 23 each of the three recombinant IPNV isolates, while fish in two parallel tanks were given cell culture medium only and kept as controls. One tank in each group was reserved for sampling, from which three fish were collected each day and organ specimens of pancreas and liver were fixed in 10% phosphate-buffered formalin. Mortality was recorded on a daily basis in the remaining tanks. The experiment was terminated at 30 days post-challenge. Immunohistochemistry and in situ detection of apoptotic cells The formalin-fixed tissue samples were embedded in paraffin according to standard procedures. Tissue sections were routinely stained with hematoxylin and eosin (HE) and examined by light microscopy. Cells from liver samples were observed specifically for morphological changes characteristic for apoptosis, such as detachment from neighboring cells, rounding, chromatin condensation, and formation of apoptotic bodies. Selected liver samples with representative changes (from minimum three fish in each group) were examined by immunohistochemistry (IHC) and terminal deoxynucleotidyl transferase mediated dUTP nickend-labeling (TUNEL) assay. These assays were performed both independently and as double labeling on parallel liver sections. IHC was performed as previously described (Santi et al., 2004). Briefly, after blocking using bovine serum albumin (BSA), the slides were incubated with a polyclonal rabbit antiserum specific for IPNV at a dilution of 1:1000 for 30 min at room temperature, followed by the secondary antibody (biotinylated anti-rabbit immunoglobulin diluted 1:300) for 30 min at room temperature. After incubation with Streptavidin – biotin alkaline phosphatase (1:1000 dilution) for 1 h at 37 -C, staining was done with Fast Red substrate (Sigma). Sections were counterstained with hematoxylin and mounted in aqueous mounting medium (Aquamount; BDH Laboratory). The TUNEL assay was performed with the in situ cell death detection kit, POD (Roche). Tissue sections were permeabilized by proteinase K digestion (20 Ag/ml, 10 mM Tris –HCl, pH 7.4) for 20 min at 37 -C. After washing with PBS, endogenous peroxidase activity was blocked by incubation with 3% H2O2 in methanol, before labeling of fragmented cellular DNA with fluorescein-dUTP by the terminal deoxynucleotidyl transferase. Detection was carried out by incubation with antifluorescein antibody conjugated with horseradish peroxidase and diaminobenzidine staining with nickel/cobalt enhancement. Sections were counterstained with hematoxylin. In double labeling, viral antigens were first detected by IHC followed by identification of apoptotic cells using the TUNEL assay. Sequence alignments of IPNV VP5 The predicted amino acid sequences for the BH3, BH1, and BH2 domain in VP5 protein of rNVI15 and rNVI15- 24 N. Santi et al. / Virology 342 (2005) 13 – 25 15K viruses were aligned to the corresponding domains from several cellular Bcl-2 proteins, including human and chicken Bcl-2, human and zebrafish Bcl-XL, and human Mcl-1. They were also aligned to the proposed BH3, BH1, and BH2 domains of viral Bcl-2 homologues proteins, including the IPNV E1S strains, Epstein – Barr virus BHRF1, human adenovirus type 9 E1B-19K, and African swine fever virus LMV-5HL. The alignments were carried out using the ClustalX program (Thompson et al., 1997). Initial computer-assisted alignment of full-length proteins were followed by several rounds of alignment of shorter stretches of these proteins to arrive at an optimized alignment, based on our knowledge of the structure of the Bcl-2 proteins. The final alignment was printed with the aid of the CHROMA software (Goodstadt and Ponting, 2001). Acknowledgments This study was supported by the Norwegian Research Council, grant #134136/120 and #146790/120 to Ø.E.; by the National Research Initiative of the USDA Cooperative State Research, Education and Extension Service, grant # 2001-35204-10065 to V.N.V; and by the Taiwan National Science Council, grant # NSC-92-2313-B-001-011 to J.L.W. The authors wish to thank Dan Torkehagen, Department of Basic Sciences and Aquatic Medicine, Norwegian School of Veterinary Science, for assistance in sampling of the postsmolts after challenge. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.virol.2005. 07.028. References Barber, G.N., 2001. Host defense, viruses and apoptosis. Cell Death Differ. 8, 113 – 126. Clemens, M.J., 2003. Interferons and apoptosis. J. Interferon Cytokine Res. 23, 277 – 292. Dobos, P., 1995. The molecular biology of infectious pancreatic necrosis virus (IPNV). Med. Toxicol. Adverse Drug Exper. 5, 24 – 54. Duncan, R., Nagy, E., Krell, P.J., Dobos, P., 1987. Synthesis of the 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. Duncan, R., Mason, C., Nagy, É., Leong, J.C., Dobos, P., 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. Eléouët, J.F., Druesne, N., Chilmonczyk, S., Monge, D., Dorson, M., Delmas, B., 2001. Comparative study of in-situ cell death induced by the viruses of viral haemorrhagic septicaemia (VHS) and infectious pancreatic necrosis (IPN) in rainbow trout. J. Comp. Pathol. 124, 300 – 307. Everett, H., McFadden, G., 1999. Apoptosis: an innate immune response to virus infection. Trends Microbiol. 7, 160 – 165. Gangappa, S., van Dyk, L.F., Jewett, T.J., Speck, S.H., Virgin, H.W., 2002. Identification of the in vivo role of a viral bcl-2. J. Exp. Med. 195, 931 – 940. Garner, J.N., Joshi, B., Jagus, R., 2003. Characterization of rainbow trout and zebrafish eukaryotic initiation factor 2a and its response to reticulum stress and IPNV infection. Dev. Comp. Immunol. 27, 217 – 231. Gil, J., Esteban, M., 2000. Induction of apoptosis by the dsRNA-dependent protein kinase (PKR): mechanism of action. Apoptosis 5, 107 – 114. Goodstadt, L., Ponting, C.P., 2001. CHROMA: consensus-based colouring of multiple alignments for publications. Bioinformatics 17, 845 – 846. Håvardstein, L.S., Kalland, K.H., Christie, K.E., Endresen, C., 1990. Sequence of the large double-stranded RNA segment of the N1 strain of infectious pancreatic necrosis virus: a comparison with other Birnaviridae. J. Gen. Virol. 69, 1757 – 1775. Hong, J.R., Lin, T.L., Hsu, Y.L., Wu, J.L., 1998. Apoptosis precedes necrosis of fish cell line with infectious pancreatic necrosis virus infection. Virology 250, 76 – 84. Hong, J.R., Lin, T.L., Yang, J.Y., Hsu, Y.L., Wu, J.L., 1999. Dynamics of nontypical apoptotic morphological changes visualized by green fluorescent protein in living cells with infectious pancreatic necrosis virus infection. J. Virol. 73, 5056 – 5063. Hong, J.R., Gong, H.Y., Wu, J.L., 2002. IPNV VP5, a novel anti-apoptosis gene of the Bcl-2 family, regulates Mcl-1 and viral protein expression. Virology 295, 217 – 229. Hunter, J.J., Bond, B.L., Parslow, T.G., 1996. Functional dissection of the human Bcl2 protein: sequence requirements for the inhibition of apoptosis. Mol. Cell. Biol. 16, 877 – 883. Imajoh, M., Hirayama, T., Oshima, S., 2005. Frequent occurrence of apoptosis is not associated with pathogenic infectious pancreatic necrosis virus (IPNV) during persistent infection. Fish Shellfish Immunol. 18, 163 – 177. Johansen, L.H., Sommer, A.I., 1995. Multiplication of infectious pancreatic necrosis virus (IPNV) in head kidney and blood leucocytes isolated from Atlantic salmon, Salmo salar L. J. Fish Dis. 18, 147 – 156. Jungmann, A., Nieper, H., Müller, H., 2001. Apoptosis is induced by infectious bursal disease virus replication in productively infected cells as well as in antigen-negative cells in their vicinity. J. Gen. Virol. 82, 1107 – 1115. Kärber, G., 1931. Beitrag zur kollektiven Behandlung pharmakologischer Reihenversuche. Arch. Exp. Pathol. Pharmakol. 162, 480 – 483. Kerr, J.F.R., Gobé, G.C., Winterford, C.M., Harmon, B.V., 1995. Anatomical methods in cell death. In: Schwartz, L.M., Osborne, B.A. (Eds.), Methods Cell Biol., vol. 46. Academic Press, San Diego, pp. 1 – 27. Khanim, F., Dawson, C., Meseda, C.A., Dawson, J., Mackett, M., Young, L.S., 1997. BHRF1, a viral homologue of the Bcl-2 oncogene, is conserved at both the sequence and functional level in different Epstein – Barr virus isolates. J. Gen. Virol. 78, 2987 – 2999. Kibler, K.V., Shors, T., Perkins, K.B., Zeman, C., Banazak, M.P., Biesterfeldt, J., Langland, J.O., Jacobs, B.L., 1997. Double-stranded RNA is trigger for apoptosis in vaccinia virus-infected cells. J. Virol. 71, 1992 – 2003. Lam, K.M., 1997. Morphological evidence of apoptosis in chickens infected with infectious bursal disease virus. J. Comp. Pathol. 116, 367 – 377. Levine, B., Huang, Q., Isaachs, J.T., Reed, J.C., Griffin, D.E., Hardwick, J.M., 1993. Conversion of lytic to persistent alphavirus infection by the bcl-2 oncogene. Nature 361, 739 – 741. Lombardo, E., Maraver, A., Espinoza, I., Fernández-Arias, A., Rodriguez, J.F., 2000. VP5, the non-structural polypeptide of infectious bursal disease virus, accumulates within the host plasma membrane and induces cell lysis. Virology 277, 345 – 357. Macdonald, R.D., Dobos, P., 1981. Identification of the proteins encoded by each genome segment of infectious pancreatic necrosis virus. Virology 114, 414 – 422. N. Santi et al. / Virology 342 (2005) 13 – 25 Mangan, D.F., Welch, G.R., Wahl, S., 1991. Lipopolysaccharide, tumor necrosis factor-a, and IL-1h prevent programmed cell death (apoptosis) in human peripheral blood monocytes. J. Immunol. 146, 1541 – 1546. Mangunwiryo, H., Aguis, C., 1988. Studies on the carrier state of infectious pancreatic necrosis virus infections in rainbow trout, Salmo gairdneri Richardson. J. Fish Dis. 11, 125 – 132. Magyar, G., Dobos, P., 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. Mundt, E., Beyer, J., Müller, H., 1995. Identification of a novel viral protein in infectious bursal disease virus-infected cells. J. Gen. Virol. 76, 437 – 443. Mundt, E., Köllner, B., Kretzschmar, D., 1997. VP5 of infectious bursal disease virus is not essential for viral replication in cell culture. J. Virol. 71, 5647 – 5651. Munn, D.H., Beall, A.C., Song, D., Wrenn, R.W., Throckmorton, D.C., 1995. Activation-induced apoptosis in human macrophages: developmental regulation of a novel cell death pathway by macrophage colony-stimulating factor and interferon-g. J. Exp. Med. 181, 127 – 136. Munro, E., Gahlawat, S., Ellis, K., 2004. A sensitive non-destructive method for detecting IPNV carrier Atlantic salmon, Salmo salar L., by culture of virus from plastic adherent leucocytes. J. Fish Dis. 27, 129 – 134. Perlman, H., Pagliari, L.J., Georganas, C., Mano, T., Walsh, K., Pope, R.M., 1999. FLICE-inhibitory protein expression during macrophage differentiation confers resistance to Fas-mediated apoptosis. J. Exp. Med. 190, 1679 – 1688. Petit, S., Lejal, N., Huet, J.C., Delmas, B., 2000. Active residues and viral substrate cleavage sites of the protease of the birnavirus infectious pancreatic necrosis virus. J. Virol. 74, 2057 – 2066. Poole, B.D., Karetnyi, Y.V., Naides, S.J., 2004. Parvovirus B-19-induced apoptosis of hepatocytes. J. Virol. 78, 7775 – 7782. Reno, P.W., Darley, S., Savan, M., 1978. Infectious pancreatic necrosis: experimental induction of a carrier state in trout. J. Fish. Res. Board Can. 35, 145 – 1456. Santi, N., Vakharia, V.N., Evensen, Ø., 2004. Identification of putative 25 motifs involved in the virulence of infectious pancreatic necrosis virus. Virology 322, 31 – 40. Santi, N., Song, H., Vakharia, V.N., Evensen, Ø., 2005. Infectious pancreatic necrosis virus VP5 is dispensable for virulence and persistence. J. Virol. 79 (14), 9206 – 9216. Song, H., Santi, N., Evensen, Ø., Vakharia, V.N., 2005. Molecular determinants of infectious pancreatic necrosis virus virulence and cell culture adaptation. J. Virol. 79 (16), 10289 – 10299. Taksdal, T., Stangeland, K., Dannevig, B., 1997. Induction of infectious pancreatic necrosis (IPN) in Atlantic salmon Salmo salar and brook trout Salvelinus fontinalis by bath challenge of fry with infectious pancreatic necrosis virus (IPNV) serotype Sp. Dis. Aquat. Org. 28, 39 – 44. Tanimura, N., Sharma, J.M., 1998. In-situ apoptosis in chickens infected with infectious bursal disease virus. J. Comp. Pathol. 118, 15 – 27. Thompson, J.D., Gibson, T.J, Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25, 4876 – 4882. Vasconcelos, A.C., Lam, K.M., 1994. Apoptosis induced by infectious bursal disease virus. J. Gen. Virol. 75, 1803 – 1806. Vermes, I., Haanen, C., Steffens-Nakken, H., Reutelingsperger, C., 1995. A novel assay for apoptosis flow cytometric detection of phosphatidylserine expression on early apoptotic cells using fluorescein labelled annexin V. J. Immunol. Methods 184, 39 – 51. Weber, S., Fichtner, D., Mettenleiter, T.C., Mundt, E., 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. Yao, K., Vakharia, V.N., 2001. Induction of apoptosis in vitro by the 17-kDa nonstructural protein of infectious bursal disease virus: possible role in viral pathogenesis. Virology 285, 50 – 58. Yao, K., Goodwin, M.A., Vakharia, V.N., 1998. Generation of a mutant infectious bursal disease virus that does not cause bursal lesions. J. Virol. 72, 2647 – 2654. Yin, X.M., Oltvai, Z.N., Korsmeyer, S.J., 1994. BH1 and BH2 domains of Bcl-2 are required for inhibition of apoptosis and heterodimerization with Bax. Nature 369, 321 – 323. Zhang, C.X., Suzuki, S., 2004. Aquabirnaviruses isolated from marine organisms form a distinct genogroup from other aquabirnaviruses. J. Fish Dis. 27, 633 – 643.