Anti-viral mechanism of barramundi Mx against betanodavirus involves the inhibition of viral RNA synthesis through the interference of RdRp

Anti-viral mechanism of barramundi Mx against betanodavirus involves the inhibition of viral RNA synthesis through the interference of RdRp

Fish & Shellfish Immunology 28 (2010) 467e475 Contents lists available at ScienceDirect Fish & Shellfish Immunology journal homepage: www.elsevier.com...

1MB Sizes 0 Downloads 32 Views

Fish & Shellfish Immunology 28 (2010) 467e475

Contents lists available at ScienceDirect

Fish & Shellfish Immunology journal homepage: www.elsevier.com/locate/fsi

Anti-viral mechanism of barramundi Mx against betanodavirus involves the inhibition of viral RNA synthesis through the interference of RdRp Yu-Chi Wu a, Yi-Fan Lu b, Shau-Chi Chi a, b, * a b

Institute of Zoology, National Taiwan University, Taipei 10617, Taiwan Department of Life Science, National Taiwan University, Taipei 10617, Taiwan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 4 November 2009 Received in revised form 4 December 2009 Accepted 10 December 2009 Available online 23 December 2009

Nervous necrosis virus (NNV) belongs to the betanodavirus of the Nodaviridae family. It is the causative agent of viral nervous necrosis (VNN) disease, and has inflicted devastating damage on the world-wide aquaculture industry. The fish that survived after the outbreak of VNN become persistently NNV-infected carriers. NNV-persistent infection has been demonstrated in a barramundi brain (BB) cell line, and it involves the type I interferon (IFN) response with the expression of Mx gene. However, little of the defense mechanism in fish cells against NNV infection is understood. In this study, the anti-NNV mechanism of barramundi Mx protein (BMx) was elucidated in cBB cells which were derived from BB cell line after serial treatments by NNV-specific antiserum and then became an NNV-free cell line. After NNV infection of cBB cells, the level of viral RNA-dependent RNA polymerase (RdRp) increased with time over a period of 24 h post-infection (hpi), but decreased when the BMx expression increased 48 and 72 hpi. When the expression of BMx was down-regulated by BMx-specific siRNA, the expression levels of viral RNA, proteins and progeny viral titers were restored. The BMx was found to colocalize with viral RdRp at the perinuclear area 24 hpi and coprecipitate with viral RdRp, indicating that they could bind with each other. Viral RdRp was also revealed to colocalize with lysosomes 48 hpi as the NNV RdRp level started to decline. Therefore, it is suggested that BMx inhibited the viral RNA synthesis by interaction with viral RdRp, and redistributed RdRp to perinuclear area for degradation. Ó 2009 Elsevier Ltd. All rights reserved.

Keywords: Betanodavirus RNA-dependent RNA polymerase Interferon Mx protein Persistent infection

1. Introduction Viral nervous necrosis (VNN) disease, also known as viral encephalopathy and retinopathy (VER), is caused by betanodavirus, inducing mass mortality of cultured marine fish at the larval stage that causes huge economic loss [1,2]. The target organ of the NNV is the nerve system, and the characteristic pathological feature of VNN disease is the vacuolation of the brain and retina [3]. NNV is a non-enveloped icosahedral virus with a diameter of 20e34 nm, containing two positive-sense RNAs without a poly A tail in its genome [4,5]. RNA1 encodes RdRp, and RNA2 encodes the capsid protein. During NNV replication, a subunit RNA derived from RNA1, named RNA3, encodes a B2 protein, which antagonizes the function of cellular RNA interference [6]. The type I interferon (IFN) response is one of the important systems for anti-viral defense. The Mx protein is an IFN-inducible protein with reported anti-viral activity against a number of viruses

* Corresponding author. Institute of Zoology, National Taiwan University, Taipei 10617, Taiwan. Tel.: þ886 2 33662505; fax: þ886 2 23673852. E-mail address: [email protected] (S.-C. Chi). 1050-4648/$ e see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.fsi.2009.12.008

[7e11]. The Mx proteins belong to the superfamily of dynamin-like, large GTPases that associate with intracellular membranes and are involved in a wide range of intracellular transport processes, e.g., endocytosis, intracellular vesicle transport, organelle maturation and cell division [12e15]. It has been reported that Mx proteins exhibit similar biophysical features to that of dynamin, including the propensity to self-assemble into ring-like and helical structures, the ability to tubulate lipids [16e18], and the large GTPases that possess mechanochemical function [12,19]. Importantly, the GTP-binding domain at the N-terminal of the Mx protein is essential for the anti-viral activity [20]. The leucine-zipper motif at the C-terminal is the region for proteineprotein interaction that determines the anti-viral specificity [21,22]. Anti-viral mechanism of human Mx proteins is well studied in mammalian system, but very limited in fish Mx due to the lack of fish-specific antibodies against immune-related proteins or cell organelles. Fish that survive after VNN disease usually become persistently infected carriers [23]. The realization of viral persistent infection mechanism is important for the design of control strategies against NNV infection. The BB cell line, which was derived from the brain tissue of a barramundi (Lates calcarifer) that survived NNV infection, was revealed to exhibit NNV-persistent infection [24].

468

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

An interferon response and the expression of the barramundi Mx gene have been reported in BB cells, and mediate persistent viral infection [25]. Another NNV-free barramundi brain cell line, cBB, was derived from BB cells after sequential treatments with NNVspecific polyclonal antibodies, and was used as a control cell line for studying the mechanism of persistent infection and the anti-viral mechanism of BMx [25]. The expression of the Mx gene can be induced in cBB cells by poly I:C transfection or NNV infection, and is able to down-regulate the proliferation of betanodavirus and birnavirus, but not iridovirus [26]. Therefore, the cBB cell line is an optimal and natural model system to study the innate immune response of fish cells against fish viruses. The objective of this study was to examine the anti-betanodavirus mechanism of BMx, including how BMx functions, which NNV replication step is downregulated by BMx, and which viral protein is BMx's major target. 2. Materials and methods 2.1. Cell lines and viruses The cBB cells [25] were derived from the brain tissue of barramundi, and cultured in Leibovitz's L15 medium supplemented with 10% fetal bovine serum (FBS), and incubated at 28  C. GF-1 cells [27] were derived from the fin tissue of grouper, and maintained in L-15 medium with 5% FBS and incubated at 28  C. The NNV strain B00GD, isolated from NNV-infected barramundi (L. calcarifer) [1], was used in this study. 2.2. Preparation of antibodies Antibodies against BMx, NNV RdRp and capsid protein were prepared in this study. Partial sequence of barramundi Mx gene was amplified from the full length sequence of barramundi Mx gene [26] using primers Mx345-F and Mx345-R (Table 1) and then cloned into the pGEM-T easy vector (Promega). The restriction sites of Bam HI and Hind III were used to insert the partial sequence of barramundi Mx gene into the prokaryotic expression vector pQE30 (QIAGEN), yielding the plasmid pQE-Mx345. The NNV RNA genome was extracted from B00GD, and RNA1 was reverse transcribed into cDNA using the primer RdRp345-R (Table 1). Subsequently, a partial sequence of NNV RNA1 gene was amplified with primers RdRp345-F and RdRp345-R (Table 1) and cloned into pGEM-T easy vector. Using the restriction sites of Bam HI and Kpn I, the partial sequence of NNV RNA1 gene was constructed into the vector pQE30 and termed as pQE-RdRp345. The pQE-Mx345 and pQE-RdRp345 were respectively transformed into the Escherichia coli M15 (QIAGEN).

Table 1 Primers and siRNAs used in this study. Name

Sequence

Mx345-F Mx345-R RdRp345-F RdRp345-R Mx150-F Mx150-R NNV R3 NNV RPCR-F NNV RPCR-R Actin-F Actin-R Mx-1 Mx-2 Mx-3 Mx-4 Non-silencing control siRNA

50 -CCGGATCCATCTTGACCAAGCCTGAT-30 50 -TTAAGCTTTCTCTCCATCTCTGCCTG-30 50 -CCGGATCCGAACCAAAGATGTCTGTC-30 50 -GGTACCGGTACCCTACACTTGAGTGCGACG-30 50 -TGAGGAGAAGGTGCGTCC-30 50 -GCGCCTCCAACACGGAGCTC-30 50 -CGAGTCAACACGGGTGAAGA-30 50 -CAGTCCGACCTCAGTACAC-30 50 -AACACTCCAGCGACACAG-30 50 -CACTCAACCCCAAAGCCAACAGG-30 50 -AAAGTCCAGCGCCACGTAGCACAG-30 r(AGAUGGAGAUGAUCGUUUA)dTdT r(GCUUCAUUGGAUUUCCUAA)dTdT r(GACAAAGAUCGAAGCCAUA)dTdT r(AGCUGAUGUUACACCUUAA)dTdT r(UUCUCCGAACGUGUCACGU)dTdT

The histidine-tagged partial barramundi Mx protein and NNV RdRp were produced in E. coli M15 and purified using Ni-NTA Agarose (QIAGEN). Capsid protein e the sole structural protein of betanodavirus e was prepared by purifying the NNV derived from infected GF-1 cells following our previous method [24]. The polyclonal antibodies (pAb) against BMx, NNV RdRp and capsid protein were prepared by immunization of rabbit and purified by protein A (Millipore). 2.3. Detection of barramundi Mx protein, NNV RdRp and capsid protein by western blot The cBB cells were seeded in 75 cm2 flask (5  106 cells per flask), and then infected with NNV (MOI ¼ 100). At 0, 12, 24, 48, 72 h post-infection (hpi), the cells were washed with PBS, harvested, and lysed with 60 ml of lysis buffer containing 1% NP-40, 50 mM Tris (pH 8.0), 1 mM DTT and 1 protease inhibitor (Roche) for 30 min on ice. The cell lysate was collected after centrifugation at 12 000  g at 4  C for 15 min. The protein concentration of the samples were adjusted to 10 mg, and applied for 10% SDS-polyacrylamide gel electrophoresis (SDS-PAGE). After SDS-PAGE, proteins were electrophoretically transferred to a PVDF membrane (Millipore). The membrane was blocked with 5% skim milk in TBS buffer containing 0.1% Tween 20 for 1 h at room temperature, and incubated with purified antibodies separately against BMx, NNV RdRp and capsid protein. After washing three times with TBS buffer, the membrane was hybridized with peroxidase-labeled goat anti-rabbit IgG (KPL), and the signals were developed by LumiGLO chemiluminescent substrate (KPL) and visualized by autoradiography. The internal control, actin, was reacted with anti-actin mAb (CHEMICON), and then with alkaline-phosphatase (AP)-conjugated goat anti-mouse IgG antibody (KPL). The protein bands were visualized using BCIP/NBT substrate Kit (KPL). 2.4. siRNA for down-regulation of Mx protein expression The sense sequences of the four Mx-specific siRNAs (Mx-1, Mx-2, Mx-3, and Mx-4) and the non-silencing control siRNA (QIAGEN) are listed in Table 1. The four siRNA specific for BMx were designed using the HiPerformance Design Algorithm licensed from Novartis AG, integrated with a stringent in-house homology analysis tool, and synthesized by the QIAGEN company. The siRNAs were labeled with Fluorescein at the 30 -end so that the efficiency of siRNA transfection could be checked. Each siRNA was diluted to a final concentration of 20 mM as a stock, and equal volume of the four Mx-specific siRNAs were mixed as a stock. To transfect the siRNAs, 5 ml of siRNA (1  1010 mol) and 2 ml of lipofectamine 2000 were separately diluted in 100 ml L15 medium without serum and incubated at room temperature for 5 min. The two solutions were mixed and incubated at room temperature for 15 min, and 0.2 ml of the mixture was added into the cBB cells that were pre-seeded in 12-well plates with 0.8 ml of L15e10% FBS. The cBB cells were transfected either with a mixture of four MxsiRNA with the final concentration of 100 nM, or 100 nM nonsilencing control siRNA (N-siRNA) as negative control, and then further transfected with poly I:C (25 ng ml1) 4 h later by adding 12.5 ml of poly I:C (2 mg ml1) pre-mixed with lipofectamine 2000 in L15. The ratio of poly I:C and lipofectamine 2000 was 1 mg:1 ml. After 18 h-incubation, the siRNA-poly I:C-transfected cBB cells were infected with NNV (MOI ¼ 100). After 1 h viral adsorption, the cells were washed with PBS and incubated with fresh culture medium. At 24 hpi, infected cells were washed with PBS and harvested, and the total protein was extracted by 40 ml of lysis buffer for western blot analysis. The total RNA of the cells at 24 hpi was extracted by

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

acid guanidinium thiocyanateephenolechloroform method [28] for real-time RT-PCR analysis of the expression of Mx and progeny NNV RNA2. Furthermore, the cells at 24 hpi were harvested and underwent three frozen/thaw cycles to extract the progeny viruses. The extracted viruses were titrated in GF-1 cells. To avoid the influence of BMx on NNV uncoating, the NNV genome was directly transfected into the siRNA-poly I:C-transfected cBB cells. The total RNA was extracted at 24 h posttransfection, and the expression of NNV RNA2 was examined by real-time RT-PCR. 2.5. Real-time RT-PCR Reverse transcription was carried out by incubating total RNA at for 1 h in 30 ml of 1 reaction buffer containing 0.3 mM oligo 42 (dT)20, 0.3 mM NNV RNA2 reverse primer R3, 0.4 mM dNTP, 11.7 mM DTT, 40 U ribonuclease inhibitor rRNasin (Promega), and 60 U MMLV reverse transcriptase (Promega). Real-time PCR was conducted to determine the expression level of barramundi Mx gene and NNV RNA2. The sequences of primer sets for barramundi Mx (Mx150-F and Mx150-R), NNV RNA2 (RPCR-F and RPCR-R) and actin (Actin-F and Actin-R) are listed in Table 1. Actin was used as the internal control, and the primer sequences of actin were per the published paper [29]. An aliquot (0.5 ml) of the cDNA was added into a real-time PCR mixture with a final volume of 20 ml containing 0.5 mM forward and reverse primers in 1 iQ SYBR Green SuperMix (Bio-Rad). The amplification was carried out in iCycler iQ Realtime PCR Detection System (Bio-Rad) with an initial denaturing step of 94  C for 3 min, followed by 40 cycles of 94  C for 20 s, 60  C for 20 s, 72  C for 20 s, and fluorescence detection at 85  C for 20 s. All samples were analyzed in triplicate. The expression level of Mx gene and NNV RNA2 was normalized with internal control (actin). C

2.6. Immunofluorescence staining To detect the localization of NNV RdRp, capsid protein and BMx, the cBB cells were infected with NNV (MOI ¼ 100) or transfected with poly I:C (1 mg ml1). At 12 and 24 hpi, cells were fixed with Carnoy's solution (CH3COOH: CH3OH ¼ 1:3) for 10 min, and blocked with 5% BSA in PBS for 1 h at room temperature. The cells were incubated with Alexa Fluor 594 (Invitrogen)-labeled anti-Mx pAb, and then with either Alexa Fluor 488 (Invitrogen)-labeled anti-NNV RdRp pAb or Alexa Fluor 488 (Invitrogen)-labeled anti-NNV capsid pAb. Finally, the cell nucleuses were stained with Hoechst 33258 and examined by fluorescence microscopy. To detect the mitochondria and RdRp, the mock-infected or NNV-infected cBB cells were first stained with MitoTracker Red CMXRos (Invetrogen) at 24 hpi. The fixation of cells and the immunostaining of RdRp and nucleuses were the same as described above. To detect the colocalization of lysosome and RdRp, the lysosomes of the mock-infected or NNV-infected cBB cells were stained with the LysoTracker Red DND-99 (Invitrogen) at 24 and 48 hpi, and then fixed with 3.7% formaldehyde. The immunostaining of RdRp and nuclei were the same as described above. The localization of BMx and viral proteins were also examined in cBB cells under the down-regulation of BMx. The cBB cells were pre-transfected with 100 nM Mx-siRNA, or with N-siRNA as negative control. After an 18 h-incubation, cells were infected with NNV (MOI ¼ 100) and fixed with Carnoy's solution at 24 hpi. The immunostaining of BMx, viral proteins and nuclei were the same as described above.

469

post-infection or post-transfection. The three sets of cell lysates were separately resuspended in 500 ml of lysis buffer, containing 50 mM Tris (pH 8), 150 mM NaCl, 1 mM EDTA, 1 protease inhibitor (Roche), and 0.1% NP-40. After centrifugation (10 000  g for 5 min at 4  C), the supernatants were respectively reacted with anti-NNV RdRp pAb, anti-NNV capsid pAb, or anti-Mx pAb at 4  C for 2 h, and then reacted with 40 ml of protein A/G (Santa Cruz) at 4  C for overnight. The protein complexes were washed twice with lysis buffer and resuspended in 30 ml of 2 sample buffer, and the complexes of BMx/RdRp or BMx/capsid protein were examined by western blot. 3. Results 3.1. NNV RdRp level decreased as BMx level increased in NNV-infected cBB cells Mx proteins have been reported to interfere NNV replication in some fish species [26,30,31], consequently, it is important to elucidate if any viral protein might be influenced after the BMx was induced. In this study, the expressions of viral proteins and BMx in the NNV-infected cBB cells at different time points post-infection were examined. After NNV infection, RdRp was detected 12 hpi, and the peak expression was found 24 hpi, and then decreased from 48 to 72 hpi (Fig. 1). In contrast, the amount of capsid protein continuously increased from 12 to 72 hpi, and the expression level of capsid protein was much higher than that of NNV RdRp 12e24 hpi. The expression of BMx was first detected 24 hpi, and the level increased within 48e72 hpi. The NNV RdRp level started dropping as the BMx level began to rise 48e72 hpi, suggesting that the decreased RdRp level might be related to the elevated BMx response. In addition, the molecular weight of RdRp is 110 kDa, of capsid protein is 37 kDa, and of BMx is 71.4 kDa. 3.2. BMx inhibited the replication of NNV To find out at which step(s) during NNV replication the BMx targeted, cBB cells were pre-transfected with Mx-specific siRNA (Mx-siRNA) to knock-down BMx expression; the control cells were transfected with non-silencing control siRNA (N-siRNA). Four hours post siRNA transfection, cBB cells were further transfected with poly I:C to induce BMx expression. The induction of BMx mRNA by poly I:C lasted a period of time, so did the inhibition of BMx mRNA by Mx-siRNA. Hence, we incubated the transfected cells for 18 h to allow the inhibition of Mx-siRNA on the induced BMx mRNA to take place more completely before NNV infection. Moreover, the

2.7. Coimmunoprecipitation assay The cBB and GF-1 cells infected with NNV (MOI ¼ 100) or cBB cells transfected with poly I:C (1 mg ml1) were harvested at 24 h

Fig. 1. Detection of RdRp, capsid protein and BMx in NNV-infected cBB cells at different time course by western blot.

470

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

Fluorescein at the 30 -end of siRNA could be detected in approximate 80% of cells, indicating the siRNA transfection in cBB cells was efficient. Because BMx would not express until 24 hpi (Fig. 1), if cBB cells were solely transfected with siRNA (Mx-siRNA or N-siRNA) and then infected with NNV, the viral replication would take place ahead of the BMx expression, and the levels of NNV replication would be high in both Mx-siRNA- and N-siRNA-transfected cBB cells 24 hpi. However, pre-transfection of siRNA and poly I:C could result in high BMx level in N-siRNA-poly I:C-transfected cBB cells, and low BMx level in Mx-siRNA-poly I:C-transfected cells. After viral infection, the NNV would be immediately interfered by the

pre-existed high or low level of BMx, and the interference on which level of viral proliferation would be observed and compared in advance. In Mx-siRNA-poly I:C-transfected cBB cells, the level of BMx was lower than that in N-siRNA-poly I:C-transfected cBB cells, while both NNV RdRp and capsid protein levels were significantly increased, compared with the control cells (Fig. 2A). The NNV RNA2 was chosen as the representative progeny RNA to examine. In Mx-siRNA-poly I:C-transfected cBB cells, the level of Mx gene expression (Mx mRNA) was about 70% down-regulated compared with that in N-siRNA-poly I:C-transfected cells, and the level of NNV RNA2 was 5-fold higher than that in control cells (Fig. 2B). Set the progeny viral titer in Mx-siRNA-poly I:C-transfected cells 24 hpi as 100%, the multiplication level of viral titer in N-siRNA-poly I:Ctransfected cBB cells turned out 30%, with zero standard deviation (Fig. 2C). The data suggested that BMx inhibited NNV replication through inhibition of viral RNA synthesis, resulting in the reduction of viral protein and titer. 3.3. BMx inhibited the synthesis of NNV RNA When the cBB cells were transfected with Mx- or N-siRNA and poly I:C and then infected with NNV, the down-regulation of NNV RNA (Fig. 2B) might be due to the interference of viral uncoating or direct interference on NNV RNA synthesis. To circumvent the uncoating issue and reconfirm that BMx could down-regulate viral RNA synthesis, cBB cells were directly transfected with NNV genomic RNA1 and RNA2, instead of NNV infection, after the transfection of siRNA and poly I:C. Genomic RNA1 would translate RdRp which was responsible for the synthesis of progeny RNA1 and RNA2. If the function of RdRp were interfered, the level of progeny RNA would decrease. In Fig. 3, the level of progeny RNA2 in the Mx-siRNA-transfected cBB cells 24 h post NNV genomic RNA transfection was approximately 2-fold higher than that in the N-siRNA-transfected cBB cells, reconfirming that BMx could suppress NNV RNA synthesis.

Fig. 2. The impact of BMx on NNV protein translation, RNA synthesis and proliferation of progeny virions at 24 h post-infection. The cBB cells were transfected with MxsiRNA or N-siRNA, and then with poly I:C. After 18 h incubation, the transfected cells were infected with NNV. (A) Viral proteins (RdRp and capsid protein) and BMx were analyzed by western blot at 24 hpi. (B) The expression level of BMx gene and synthesis level of NNV RNA2 were detected by real-time RT-PCR at 24 hpi. The expression level of BMx gene in N-siRNA-transfected cBB cells, and the synthesis level of NNV RNA2 in Mx-siRNA-transfected cBB cells were regarded as 100%. Bars represent standard deviations. **P < 0.01. (C) Progeny virions were titrated at 24 hpi, and the viral titer in Mx-siRNA-transfected cBB cells was regarded as 100%. The results of three repeats were the same, so no standard bar was shown in Fig. 2C.

Fig. 3. NNV RNA synthesis was down-regulated by BMx. The cBB cells were transfected with Mx-siRNA or N-siRNA, and then with poly I:C, and finally with NNV genomic RNA1 and RNA2. Twenty-four hours post-transfection, the synthesis level of NNV RNA2 was examined by real-time RT-PCR. The synthesis level of RNA2 in MxsiRNA-transfected cBB cells was regarded as 100%. Bars represent standard deviations. *P < 0.05.

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

3.4. The localization of BMx, NNV RdRp and capsid protein in NNV-infected cBB cells Several reports have indicated that the sub-cellular localization of Mx protein is critically important for its anti-viral activity and specificity [32e35]. To examine if BMx could interact with NNV proteins, specifically RdRp, and the localization of viral proteins and BMx in cBB cells during viral infection, the cells were immunostained using specific antibodies 12 h and 24 h post NNV infection. In the poly I:C-transfected cBB cells, the expressed BMx appeared as granules at the perinuclear area 24 h post-transfection (Fig. 4A). In NNV-infected cBB cells, RdRp expressed and distributed throughout the cytoplasm 12 hpi when BMx was yet to express (Fig. 4B). As soon as BMx expressed 24 hpi, RdRp became granulelike aggregates and concentrated to the perinuclear area e as BMx showed e and colocalized with BMx (Fig. 4B). In contrast to RdRp, the capsid protein evenly distributed in the cytoplasm 12 hpi

471

through 24 hpi, regardless of the BMx expression 24 hpi (Fig. 4C). The distribution of BMx, RdRp, and capsid protein in NNV-infected cBB cells 48 hpi was the same as 24 hpi (data not shown). When cBB cells was transfected with N-siRNA 24 h before NNV infection, RdRp was found to aggregate at the perinuclear area and colocalize with BMx 24 hpi (Fig. 5A), similar to its distribution observed in solely NNV-infected cBB cells 24 hpi (Fig. 4B). On the contrary, RdRp distributed throughout the cytoplasm when the expression of BMx was down-regulated by Mx-siRNA (Fig. 5A). Although the colocalization of capsid protein and BMx seemed to be found in N-siRNA-transfected and then NNV-infected cBB cells (Fig. 5B), the capsid protein remained dispersed throughout the cytoplasm despite the BMx expression being inhibited by Mx-siRNA or not. The results indicate that BMx could change the distribution of RdRp and colocalize with RdRp at the perinuclear area once the BMx expression takes place, and that RdRp might be the major protein target for BMx interaction.

Fig. 4. The localization of NNV RdRp, capsid protein and BMx in NNV-infected cBB cells. By immunofluorescence staining, the localization of BMx was detected in cBB cells at 24 h post poly I:C transfection (A); while, the colocalization of BMx with RdRp (B) or with capsid protein (C) were examined in the NNV-infected cBB cells at 12 and 24 hpi. Bar ¼ 20 mm.

472

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

Fig. 5. The localization of NNV RdRp and capsid protein in Mx-siRNA- or N-siRNA-transfected cBB cells after NNV infection. The cBB cells were transfected with Mx-siRNA or N-siRNA, and then infected with NNV at 24 h post-transfection. The colocalization of BMx with either (A) RdRp or (B) capsid protein was detected by immunofluorescence staining at 24 h post NNV infection. Bar ¼ 20 mm.

3.5. Coimmunoprecipitation of BMx with NNV RdRp and capsid protein

expressed ahead of BMx in cBB cells after NNV infection, it is reasonable to expect that some, if not all, RdRp had integrated into the mitochondria outer membrane before interacted with BMx, and RdRp might colocalize with mitochondria. If BMx interacted with RdRp and then changed its distribution to perinuclear area, the distribution of RdRp-associated mitochondria might be altered at the same time. The distribution of mitochondria was dispersed in the cytoplasm of mock-infected cBB cells, but turned localized around the nucleus and colocalized with RdRp in the NNV-infected cBB cells 24 hpi (Fig. 7). Furthermore, the colocalization pattern of RdRp/mitochondria was similar to that of RdRp/BMx 24 hpi (Fig. 4B).

The coimmunoprecipitation assay was applied to examine if BMx could directly interact with NNV RdRp or capsid protein. By western blot analysis, RdRp and capsid proteins were both detected in the lysates of NNV-infected cBB and NNV-infected GF-1 cells 24 hpi (Fig. 6A). However, the poly I:C treatment could induce the Mx protein in only cBB cells but not GF-1 cells, so that GF-1 cells could be used as a tool to examine the specificities of antiserum against NNV RdRp, capsid protein, and BMx. The results (lanes 2 and 3 in Fig. 6B and C) revealed that the pAbs against either RdRp or capsid protein could specifically react with their target protein, and would not crossreact with other proteins, including BMx. On the other hand, the result (lanes 2 and 3 in Fig. 6D) confirmed the specificity of anti-BMx pAbs. However, both RdRp and capsid protein were coprecipitated with BMx from NNV-infected cBB cell lysate by the BMx-specific antibodies, indicating that either RdRp or capsid protein could interact with BMx (lane 1 in Fig. 6D). The BMx and capsid protein were coprecipitated with RdRp, from NNV-infected cBB cell lysate, using RdRp-specific pAb, suggesting that a RdRp-BMx-capsid protein complex formed (Fig. 6B). Furthermore, it was confirmed that RdRp and BMx were coprecipitated with capsid protein, using capsid protein-specific pAb (lane 1 in Fig. 6C).

In NNV-infected cBB cells, RdRp localized to the perinuclear area responding to the BMx expression 24 hpi, and the level of RdRp began to drop 48 hpi (Fig. 1). To check if RdRp degraded in the lysosomes 48 hpi, the distribution patterns of lysosome and RdRp were examined. In mock-infected cBB cells, lysosomes distributed as a dot-like aggregates around the nucleus (Fig. 8). In NNV-infected cBB cells, RdRp and lysosomes did not colocalize 24 hpi, rather, they showed up at the perinuclear area, and the colocalization became apparent till 48 hpi (Fig. 8), indicating that the decrease in RdRp level 48 hpi might be related with the colocalization of RdRp with lysosome, and RdRp might be degraded in lysosome.

3.6. Colocalization of RdRp and mitochondria

4. Discussion

The NNV RdRp has been reported to associate with the outer membrane of mitochondria in NNV-infected cells [36]. Since RdRp

To date, only grouper Mx and barramundi Mx in lower vertebrates have reportedly exhibited anti-NNV activity [26,30,31].

3.7. Colocalization of RdRp and lysosomes

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

473

Fig. 6. Coimmunoprecipitation of BMx with either NNV RdRp or capsid protein. The cell lysates from (1) NNV-infected cBB cells, (2) poly I:C-transfected cBB cells, and (3) NNVinfected GF-1 cells, being used as the input for the coimmunoprecipitation, were first analyzed by western blot (A). Viral proteins or BMx from different cell lysates were immunoprecipitated separately by anti-RdRp pAb (B), anti-capsid pAb (C), or anti-Mx pAb (D).

However, the mechanism underlying the anti-NNV activity of the fish Mx protein remains elusive. In the present study, the NNV RdRp protein was found to be the major protein target for barramundi Mx inhibition; BMx was able to down-regulate NNV RNA synthesis by interacting with RdRp; and BMx might directly or indirectly redistribute RdRp to perinuclear area for lysosomal degradation. The Mx protein is an IFN-inducible protein, and some Mx proteins are reported to exhibit anti-viral function. However, the sub-cellular localization of Mx protein is critically important for its anti-viral activity and specificity, and the anti-viral mechanisms of Mx proteins vary in different animal systems. For example, the human MxA, expressed in the cytoplasm, blocks the transport of Thogoto virus nucleocapsids into the nucleus, thereby preventing nuclear viral replication [33]. Moreover, human cytoplasmic MxA protein interferes with the primary transcription of vesicular

stomatitis virus (VSV) in the cytoplasm [32,34], but cannot inhibit the influenza A virus primary transcription that occurs in the nucleus [35]. When a nuclear translocation signal is added to human MxA protein, it turns to translocate to the nucleus and inhibits the transcription of influenza A virus and Thogoto virus, instead of VSV and bunyaviruses [21,22,37]. In the present study, the BMx was detected only in the cytoplasm of poly I:C-transfected cBB cells and not in the nucleus, hence the chance of interfering NNV replication in cytoplasm exists. In NNV-infected cBB cells, the level of RdRp decreased during 48e72 hpi when the level of BMx increased, but the capsid protein continued rising within the same period. It is possible that RdRp is an early protein and capsid is a late protein, so capsid protein level continued to increase when the RdRp level decreased. The decrease of RdRp might also be related with the increased BMx level. In GF-1

Fig. 7. Mitochondria colocalized with RdRp at the perinuclear area in NNV-infected cBB cells at 24 hpi. Mitochondria were stained by MitoTracker Red CMXRos. Then, the cells were fixed by 3.7% formaldehyde and immunostained with Alexa Fluor 488-labeled anti-RdRp pAb and Hoechst 33258. Bar ¼ 20 mm.

474

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

Fig. 8. Viral RdRp colocalized with lysosomes at the perinuclear area in NNV-infected cBB cells at 48 hpi. The lysosomes of NNV-infected cBB cells were stained by LysoTracker Red DND-99 at 24 and 48 hpi. Then, the cells were fixed by 3.7% formaldehyde, and immunostained with Alexa Fluor 488-labeled anti-RdRp pAb and Hoechst 33258. Bar ¼ 20 mm.

cells, no Mx protein would be induced by NNV infection or poly I:C transfection, and the level of RdRp continued rising 48e72 hpi (data not shown); the relationship between BMx and NNV RdRp was examined in the present study. Human MxA protein has been reported to inhibit the transcription of VSV through inhibiting the accumulation of VSV primary transcripts in vivo, and recombinant human MxA protein inhibits VSV transcription in vitro without any interaction with host factors [32]. Moreover, human MxA protein prevents the accumulation of VSV transcripts via inhibiting the activity of viral polymerase rather than destabilizing viral mRNAs [38]. According to our results (Figs. 2 and 3), the synthesis level of NNV RNA was restored in Mx-siRNA-transfected cells, indicating that BMx could inhibit NNV RNA synthesis. Considering the NNV RdRp is responsible for progeny RNA synthesis, the down-regulation of NNV RNA synthesis might be facilitated via the interference of BMx on RdRp. It is worth noting that the difference in levels of NNV RdRp and capsid protein 24 hpi between Figs. 2A and 1. In Fig. 1, the cBB cells were solely infected with NNV, and the viral proteins expressed ahead of BMx. Although the NNV RdRp level decreased when the BMx level increased 48 hpi, the earlier already expressed RdRp had synthesized some progeny RNA1 and RNA2 for the following translation of new RdRp and capsid protein. In Fig. 2A, poly I:C transfection was used to induce BMx expression 18 h ahead of NNV infection; therefore, the expression of BMx in the N-siRNA-poly I:Ctransfected cells was earlier than that of viral protein, and BMx could interact or interfere with RdRp as soon as it expressed; therefore, the levels of progeny RNA (Fig. 2B) and progeny viral proteins (Fig. 2A) were much lower than that observed in the solely NNV-infected cBB cells (Fig. 1). Wu and Chi (2006) demonstrated that in persistently NNVinfected BB cell line, the NNV capsid protein can only be detected in a few cells in each subculture, and the NNV-persistent infection can be re-induced in cBB cells by viral infection with low multiplicity of infection (MOI  1). When cBB cells are treated with the BB cell culture supernatant neutralized by NNV-specific mAbs, the BMx expression is detected in those cells, indicating that the few NNVinfected cells in each subculture can express and release IFN-like cytokine into the culture supernatant to induce IFN response in other non-infected cells. It is unclear how many aspects that IFN

response can protect those non-infected cells from NNV infection or proliferation, but at least, BMx expresses in those non-infected cells before they contact with the NNV particles released from the first NNV-infected cells. Thus, the situation will be similar to the result shown in Fig. 2A: BMx would interact or interfere with viral RdRp once NNV infects and RdRp expresses, then the viral proliferation would be down-regulated or blocked at the early stage of viral infection, resulting in none to very low viral protein detected in those IFN-stimulated cells. In Figs. 4 and 5, the distribution of NNV RdRp was reorganized when BMx expressed, and RdRp colocalized with BMx at the perinuclear area 24 hpi. Moreover, RdRp was also coprecipitated with BMx by BMx-specific antibodies, and BMx was able to coprecipitate with RdRp by RdRp-specific antibodies. It is therefore confirmed that BMx could interact with RdRp. The NNV capsid protein has been reported to interact with recombinant grouper Mx protein in vitro [30] and found to coprecipitate with BMx from NNV-infected cBB cell lysate using BMx-specific antibodies. However, the distribution of NNV capsid protein did not change with BMx as that of NNV RdRp. Because the expression level of capsid protein was much higher than that of BMx in NNV-infected cBB cells 24 hpi (Fig. 6A), it is suggested that only a small fraction of capsid protein would react with BMx, and most capsid protein would not, so that the overall distribution of capsid protein did not change or relocate with BMx. The relationship of BMx and NNV RdRp was focused in this study. What is the relationship between the BMx inhibition of NNV RNA synthesis and the movement of NNV RdRp with BMx to the perinuclear area? The BMx was demonstrated to colocalize with RdRp from 24 hpi on, and the interaction between BMx and RdRp was reconfirmed by coimmunoprecipitation assay. It is also noted that RdRp colocalized with lysosomes since 48 hpi, and the reduction of NNV RdRp in NNV-infected cBB cells started 48 hpi through 72 hpi. However, the level of NNV RdRp at 48e72 hpi could be restored when the lysosomal proteolysis was inhibited by 25 mM NH4Cl (data not shown). Therefore, it is suggested that BMx might directly or indirectly sequester RdRp to lysosomes for degradation. Mitochondria have been reported to be the site for NNV RdRp localization and RNA synthesis [36]. In NNV-infected cells, the expression of viral RdRp was earlier than that of BMx, and some or

Y.-C. Wu et al. / Fish & Shellfish Immunology 28 (2010) 467e475

all RdRp might insert into mitochondria membrane before it interacted with BMx. In Fig. 7, mitochondria distributed throughout the cytoplasm in mock-infected cells, but localized to perinuclear area with RdRp in NNV-infected cells 24 hpi. Because the BMx protein belongs to the superfamily of dynamin-like GTPase that are involved in intracellular membrane trafficking [12], it is likely that BMx might interact with mitochondria-associated RdRp, and transported RdRp along with mitochondria to perinuclear area. On the other hand, if BMx expressed in advance of viral RdRp, for example, in IFN-stimulated and then NNV-infected cBB cells, BMx might interact with RdRp before it inserted into the mitochondria outer membrane, and BMx might sequester RdRp to perinuclear area without involving mitochondria. If NNV RdRp was sequestered by BMx to the perinuclear area, viral RNA synthesis would be interfered. A similar phenomenon has been reported in human MxA protein: The MxA binds with the nucleocapsid protein of La Crosse virus (LACV), forming a complex and colocalizing to the perinuclear area [39]. In summary, the anti-NNV mechanism of BMx involves the inhibition of viral RNA synthesis through the interaction of BMx with RdRp, and probably by sequestering RdRp to lysosomes for degradation. It remains unclear how BMx transports RdRp or the mitochondria-associated RdRp to the perinuclear area, as well as how these complexes internalize into lysosomes. Also, it is unknown if BMx participates in any other anti-viral mechanisms yet to be explored. All these questions will be answered in the future. Acknowledgements The authors appreciate Dr. L. C. Hsu, Dr. B. Wu-Hsieh and Mr. C. J. Cheng for pre-reviewing this manuscript and their precious comments. This work was partially supported by the National Science Council of the Republic of China under the contract No. NSC 96-2313-B-002-042-MY3. References [1] Chi SC, Shieh JR, Lin SJ. Genetic and antigenic analysis of betanodaviruses isolated from aquatic organisms in Taiwan. Dis Aquat Organ 2003;55:221e8. [2] Munday BL, Kwang J, Moody N. Betanodavirus infection of teleost fish: a review. J Fish Dis 2002;25:127e42. [3] Chi SC, Lo CF, Kou GH, Chang PS, Peng SE, Chen SN. Mass mortalities associated viral nervous necrosis disease in two species of hatchery-reared grouper, Epinephelus fuscogutatus and Epinephelus akaara (Temminck & Schlegel). J Fish Dis 1997;20:185e93. [4] Chi SC, Lo BJ, Lin SC. Characterization of Grouper Nervous Necrosis Virus (GNNV). J Fish Dis 2001;24:3e13. [5] Mori K, Nakai T, Muroga K, Arimoto M, Mushiake K, Furusawa I. Properties of a new virus belonging to nodaviridae found in larval striped jack (Pseudocaranx dentex) with nervous necrosis. Virology 1992;187:368e71. [6] Fenner BJ, Thiagarajan R, Chua HK, Kwang J. Betanodavirus B2 is an RNA interference antagonist that facilitates intracellular viral RNA accumulation. J Virol 2006;80:85e94. [7] Haller O, Frese M, Kochs G. Mx proteins: mediators of innate resistance to RNA viruses. Rev Sci Tech 1998;17:220e30. [8] Nakayama M, Nagata K, Kato A, Ishihama A. Interferon-inducible mouse Mx1 protein that confers resistance to influenza virus is GTPase. J Biol Chem 1991;266:21404e8. [9] Schneider-Schaulies S, Schneider-Schaulies J, Schuster A, Bayer M, Pavlovic J, ter Meulen V. Cell type-specific MxA-mediated inhibition of measles virus transcription in human brain cells. J Virol 1994;68:6910e7. [10] Thimme R, Frese M, Kochs G, Haller O. Mx1 but not MxA confers resistance against tick-borne Dhori virus in mice. Virology 1995;211:296e301.

475

[11] Zurcher T, Pavlovic J, Staeheli P. Mouse Mx2 protein inhibits vesicular stomatitis virus but not influenza virus. Virology 1992;187:796e800. [12] Danino D, Hinshaw JE. Dynamin family of mechanoenzymes. Curr Opin Cell Biol 2001;13:454e60. [13] Schmid SL, McNiven MA, De Camilli P. Dynamin and its partners: a progress report. Curr Opin Cell Biol 1998;10:504e12. [14] Sever S, Damke H, Schmid SL. Garrotes, springs, ratchets, and whips: putting dynamin models to the test. Traffic 2000;1:385e92. [15] van der Bliek AM. Functional diversity in the dynamin family. Trends Cell Biol 1999;9:96e102. [16] Accola MA, Huang B, Al Masri A, McNiven MA. The antiviral dynamin family member, MxA, tubulates lipids and localizes to the smooth endoplasmic reticulum. J Biol Chem 2002;277:21829e35. [17] Stowell MH, Marks B, Wigge P, McMahon HT. Nucleotide-dependent conformational changes in dynamin: evidence for a mechanochemical molecular spring. Nat Cell Biol 1999;1:27e32. [18] Sweitzer SM, Hinshaw JE. Dynamin undergoes a GTP-dependent conformational change causing vesiculation. Cell 1998;93:1021e9. [19] McNiven MA, Cao H, Pitts KR, Yoon Y. The dynamin family of mechanoenzymes: pinching in new places. Trends Biochem Sci 2000;25:115e20. [20] Pitossi F, Blank A, Schroder A, Schwarz A, Hussi P, Schwemmle M, et al. A functional GTP-binding motif is necessary for antiviral activity of Mx proteins. J Virol 1993;67:6726e32. [21] Frese M, Kochs G, Meier-Dieter U, Siebler J, Haller O. Human MxA protein inhibits tick-borne Thogoto virus but not Dhori virus. J Virol 1995;69:3904e9. [22] Zurcher T, Pavlovic J, Staeheli P. Mechanism of human MxA protein action: variants with changed antiviral properties. EMBO J 1992;11:1657e61. [23] Johansen R, Amundsen M, Dannevig BH, Sommer AI. Acute and persistent experimental nodavirus infection in spotted wolffish Anarhichas minor. Dis Aquat Organ 2003;57:35e41. [24] Chi SC, Wu YC, Cheng TM. Persistent infection of betanodavirus in a novel cell line derived from the brain tissue of barramundi Lates calcarifer. Dis Aquat Organ 2005;65:91e8. [25] Wu YC, Chi SC. Persistence of betanodavirus in barramundi brain (BB) cell line involves the induction of interferon response. Fish Shellfish Immunol 2006; 21:540e7. [26] Wu YC, Chi SC. Cloning and analysis of antiviral activity of a barramundi (Lates calcarifer) Mx gene. Fish Shellfish Immunol 2007;23:97e108. [27] Chi SC, Hu WW, Lo BJ. Establishment and characterization of a continuous cell line (GF-1) derived from grouper, Epinephelus coioides (Hamilton): a cell line susceptible to grouper nervous necrosis virus (GNNV). J Fish Dis 1999;22: 173e82. [28] Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanateephenolechloroform extraction. Anal Biochem 1987; 162:156e9. [29] Larsen R, Rokenes TP, Robertsen B. Inhibition of infectious pancreatic necrosis virus replication by Atlantic salmon Mx1 protein. J Virol 2004;78:7938e44. [30] Chen YM, Su YL, Shie PS, Huang SL, Yang HL, Chen TY. Grouper Mx confers resistance to nodavirus and interacts with coat protein. Dev Comp Immunol 2008;32:825e36. [31] Lin CH, Christopher John JA, Lin CH, Chang CY. Inhibition of nervous necrosis virus propagation by fish Mx proteins. Biochem Biophys Res Commun 2006; 351:534e9. [32] Staeheli P, Pavlovic J. Inhibition of vesicular stomatitis virus mRNA synthesis by human MxA protein. J Virol 1991;65:4498e501. [33] Kochs G, Haller O. Interferon-induced human MxA GTPase blocks nuclear import of Thogoto virus nucleocapsids. Proc Natl Acad Sci U S A 1999;96:2082e6. [34] Schnorr JJ, Schneider-Schaulies S, Simon-Jodicke A, Pavlovic J, Horisberger MA, ter Meulen V. MxA-dependent inhibition of measles virus glycoprotein synthesis in a stably transfected human monocytic cell line. J Virol 1993;67: 4760e8. [35] Pavlovic J, Haller O, Staeheli P. Human and mouse Mx proteins inhibit different steps of the influenza virus multiplication cycle. J Virol 1992;66:2564e9. [36] Guo YX, Chan SW, Kwang J. Membrane association of greasy grouper nervous necrosis virus protein A and characterization of its mitochondrial localization targeting signal. J Virol 2004;78:6498e508. [37] Turan K, Mibayashi M, Sugiyama K, Saito S, Numajiri A, Nagata K. Nuclear MxA proteins form a complex with influenza virus NP and inhibit the transcription of the engineered influenza virus genome. Nucleic Acids Res 2004;32:643e52. [38] Schwemmle M, Weining KC, Richter MF, Schumacher B, Staeheli P. Vesicular stomatitis virus transcription inhibited by purified MxA protein. Virology 1995;206:545e54. [39] Kochs G, Janzen C, Hohenberg H, Haller O. Antivirally active MxA protein sequesters La Crosse virus nucleocapsid protein into perinuclear complexes. Proc Natl Acad Sci U S A 2002;99:3153e8.