Viral Envelope Protein 53R Gene Highly Specific Silencing and Iridovirus Resistance in Fish Cells by AmiRNA

Background Envelope protein 53R was identified from frog Rana grylio virus (RGV), a member of the family Iridoviridae, and it plays an important role in the virus assembly. Although inhibition of iridovirus major capsid protein (MCP) by small hairpin RNAs (shRNAs) has been shown to cause resistance to viral infection in vitro, RNA interference (RNAi) to inhibit aquatic animal virus envelope protein gene product has not been reported. Methodology We devised artificial microRNAs (amiRNAs) that target a viral envelope protein gene RGV 53R. By incorporating sequences encoding amiRNAs specific to 53R of RGV into pre-miRNA155 (pSM155) vectors, which use the backbone of natural miR-155 sequence and could intracellularly express 53R-targeted pre-amiRNAs. The pre-amiRNAs could be processed by the RNase III-like enzyme Dicer into 21–25 nt amiRNAs (amiR-53Rs) in fish cell lines. The levels of 53R expression were analyzed through real-time PCR and RGV virions assembly were observed by electronic microscopy in fish cells transfected with or without amiR-53Rs at 72 h of RGV infection. Conclusion/Significance The results argue that viral envelope protein RGV 53R can be silenced and the virions assembly was deficient by amiR-53R-1, and further identified the first amiRNA of envelope protein gene from iridovirus that was able to cause resistance to virus infection in fish cells. The data demonstrate that the viral infection is efficiently suppressed (58%) by amiR-53R-1 targeting positon 36–57 of RGV 53R. Moreover, electron microscopic observations revealed virion assembly defect or reduced virions assembly capacity was closely correlated to expression of amiR-53R-1. Based on real time PCR of the Mx gene, we found no evidence of activation of IFN by amiR-53R-1.


Introduction
Iridoviruses are one of the significant viral pathogens for aquatic animals that have caused great economic losses in the aquaculture industry worldwide [1]. Rana grylio virus (RGV), a member of iridoviruses, causes systemic infectious iridoviral disease in cultured pig frog (Rana grylio) [2]. Some RGV genes and their functions have been identified, including deoxyuridine triphosphatase (dUTPase), thymidine kinase (TK), 3b-hydroxysteroid dehydrogenase (3b-HSD) and ERV1 (essential for respiration and vegetative growth 1) [3][4][5][6]. Recently, an envelope protein that may play an important role in virus assembly, RGV 53R, was identified and characterized from RGV. RGV 53R has homologues in all sequenced iridoviruses. The structural conserved features shared by iridovirus envelope proteins were also found in RGV 53R. The intracellular distribution and dynamic changes of RGV 53R revealed that it co-localized with the ER components at early stage of post transfection [7]. Therefore, RGV 53R was selected for further characterization to understand the molecular pathogenesis and gene silencing by artificial microRNAs (amiRNAs) in RGV infection.
Discovery of the RNA interference (RNAi) pathway has led to exciting new strategies for developing treatment for viral diseases [8]. To date, siRNAs and miRNAs have been used for silencing expression of the major capsid protein (MCP) encoded by tiger frog virus (TFV) and red seabream iridovirus (RSIV), two iridovirus causing severe disease in aquatic animals [9][10][11]. But the envelope protein of iridovirus has not been demonstrated as a target for RNAi as well as MCP. RNA interference is a process of post-transcriptional sequence specific gene silencing in many eukaryotes. The use of RNAi to inhibit virus, including siRNAs and miRNAs, offers a new approach for controlling viral infections [12,13]. Artificial miRNAs (amiRNAs) expression plasmid vectors such as pre-miRNA 155-designed shRNAs vectors (pSM155) which were designed to target on ORF of endogenous and exogenous genes have been developed and used for RNA interference [14][15][16][17][18]. The plasmid vector pSM155 used in this study to construct plasmids that express amiRNAs was based on the backbone of natural miR-155. And the natural miRNA: miRNA duplex sequence was replaced by the artificial one [19]. The appearance of plasmid-based expression systems that is effective and inexpensive for amiRNAs generation present rational way for the design and expression of 53R targeted amiRNA.

Expression of pSM155-amiR-53Rs
Three pairs of oligonucleotides encoding 53R-specific amiRNAs of RGV (referred as amiR-53R-1, amiR-53R-2 and amiR-53R-3) (Table 1), and a pairs of oligonucleotides corresponding to PB2 gene of avian influenza virus, AIV (referred as amiR-PB2), were annealed and ligated into the pre-miRNA155 (pSM155) vector to create plasmids (pSM155-amiR-53Rs and pSM155-amiR-PB2) capable of producing 53R or PB2 gene encoded pre-amiRNAs in plasmidtransfected cells. The predicted structures of the engineered pre-amiRNAs incorporated into the pSM155 backbone are shown in Fig. 1. When pSM155-amiR-53Rs were transfected into grass carp ovary (GCO) cells, they allowed co-cistronic expression of pre-amiRNAs with GFP in cells under the control of the Pol II human CMV promoter. The co-cistronic expression of the pre-amiRNAs was monitored microscopically under a fluorescence microscope.

Delayed emergence of CPE of RGV in host cells by amiR-53R-1
Cytopathic effect (CPE) was the most intuitive parameter that reflected the viral quantity of virus accumulation [21]. In GCO cells that were infected with RGV after transfected with pSM155-amiR-53R-1, CPE were markedly delayed than that were transfected with pSM155-amiR-PB2, as well as only infected with RGV.
Observation of RGV virions in ultrathin sections of epithelioma papulosum cyprini (EPC) cells by electron microscopy showed that larger numbers of RGV virions were assembled and arranged orderly in the cytoplasm of the host cells only infected with RGV ( Fig. 4-a). In pSM155-amiR-53R-1 transfected EPC cells, several sporadic RGV virions were present in the cytoplasm with irregular arrangement and the number of virions were significantly decreased ( Fig. 4-b).

RGV resistance by amiR-53R-1
It has been shown in above that 53R silencing and RGV inhibition by pSM155-amiR-53R-1 delayed emergence of CPE, or reduced RGV titer and assembled virions in transfected fish cells. So 53R expression was assayed as the RGV inhibition efficiency in RGV infected GCO cells that transiently transfected with pSM155-amiR-53R-1 or pSM155-amiR-PB2 by real-time PCR. The data indicated that the 53R expression was reduced by 58% (FC = 0.4260.05) at 72 h p.i. of RGV in pSM155-amiR-53R-1 transfected cells with compared to the pSM155-amiR-PB2 transfected cells (Fig. 5). No Mx response by amiR-53Rs Mx is a key component of the interferon response. To determinate whether transfection of GCO cells with plasmids expressing 53R-targeted amiR-53Rs and pEGFP-N3-53R resulted in induction of IFN response, mRNA levels of Mx were detected. There were no significant changes in expression of Mx gene in GCO cells that were transfected with different pSM155-amiR-53Rs (Fig. 6). These results indicated that the antiviral activity was not induced by Mx expression but by RNAi. Figure 2. Quantitative analysis of 53R mRNA levels in cells co-transfected with different plasmids. GCO cells co-transfected with pSM155-amiR-53R-1/pEGFP-N3-53R, pSM155-amiR-53R-2/pEGFP-N3-53R and pSM155-amiR-53R-3/pEGFP-N3-53R, respectively. 53R mRNA level from each group was measured by real-time PCR analysis 72 h post transfection. Group co-transfected with pSM155-amiR-PB2/pEGFP-N3-53R was used as negative control. The value of negative control was designated as 1.0 (n = 3). The values represent averages of three independent experiments, with the range indicated (6S.D). *P,0.05 versus control. doi:10.1371/journal.pone.0010308.g002

Discussion
Envelope protein gene 53R has been reported as one of the twenty six core genes existed in all iridoviruses [22]. Present study was the first time that the envelope protein targeted amiRNA was used for silencing an iridovirus efficiently. It revealed that envelope protein was also good target for iridovirus inhibition besides the MCP and offers a new approach for controlling viral infections.  Three amiRNAs targeted to different positions were constructed and used in this study. Two of these, amiR-53R-1 and amiR-53R-3, had the diversity in just one nucleotide in the targeted positions. AmiR-53R-1 targeted sequence was located in positions 36-57 of RGV 53R whereas amiR-53R-3 targeted 37-58. But the inhibition efficiency of these two amiRNAs was distinctly different. AmiR-53R-1 reduced 53R expression in 74% compared to 35% acquired by amiR-53R-3 when co-transfected with pEGFP-N3-53R. It suggested that amiRNA targeted positions of envelope protein was highly specific. And the selection of specific positions in future amiRNA design was very important.
Single or double stranded RNAs could active the immune response of host cells to resist virus infection [23]. So the amiRNAs may active immune response, especially the interferon (IFN)related pathways, to inhibit virus. Poly I:C was used in this study as the positive control. It has been reported that poly I:C was a inducer of IFN in vitro and in vivo [24,25]. Poly I:C also has been used in fish to induce IFN response [26,27]. In the poly I:C treated group of the present study, Mx expression was largely upregulated. So it was considered that Mx gene used in this study was sensitive for IFN. And it was proved in this research that IFN response was not activated apparently by examine the Mx expression, which was a key component in IFN response [28]. The pre-amiRNAs used for RGV 53R inhibition had a mismatch at the +1 position and a bulge at the +11, +12 positions in its stem. It's different from the pre-amiRNAs without a mismatch at the +1 position and a bulge at the +12, +13 positions which was found to activate IFN-related pathways [10]. Different Mx inducing ability may be due to different structures of the pre-amiRNAs. Moreover, RGV titers obtained in the less active amiRNA treated cells were between those from amiR-53R-1 and negative controls. It indicated that the decrease of RGV titers was mostly attributed to the knock down of 53R expression.
Inhibition of viral yield was less at 96 hr p.i. than at the two earlier time points. It may be due to the reduction of pSM155-amiR-53R-1 plasmid in transfected cells. The level of plasmids in cells at 96 hr p.i. was less than at the two earlier time points, probably due to the degradation of plasmids with time. So the inhibition effect was less as the time went by.
On the whole, we reported the first iridovirus envelope protein gene targeted amiRNAs that could mediate iridovirus resistance by the specific gene-silence pathway. This system could be used for the inhibition of iridovirus replication and the future development of strategies to control viral diseases in aquaculture. The most highly inhibition efficiency in our study was 58% (obtained by amiR-53R-1). It is slightly lower than the inhibition efficiency by MCP targeted amiRNAs [10]. Therefore, future work should be done to increase the inhibition efficiency of envelope protein gene targeted amiRNAs.

Cell lines and virus
A strain of frog iridovirus, Rana grylio virus (RGV), and two fish cell lines, grass carp ovaries (GCO) and epithelioma papulosum cyprinid (EPC) cells, were used in these studies (GCO cell and EPC cell were maintained in our laboratory, please see the references) [29,30]. Cell culture, virus propagation and virus titer determination were performed as described previously [31,32].

AmiRNAs device and plasmids construction
Three pairs of oligonucleotides encoding 53R-specific amiRNAs of RGV (referred as amiR-53R-1, amiR-53R-2 and amiR-53R-3) ( Table 1), and a pair of oligonucleotides corresponding to PB2 gene of avian influenza virus, AIV (referred as amiR-PB2), were designed using the BLOCK-iT TM RNAi Designer/miR RNAi (http://rnaidesigner.invitrogen.com). Each oligonucleotide pair (''top strand'' and ''bottom strand'' oligos) was annealed and ligated into the pre-miRNA155 (pSM155) vector to create plasmids (pSM155-amiR-53Rs and pSM155-amiR-PB2) capable of producing 53R and PB2 encoded pre-amiRNAs in plasmid transfected cells. The pSM155 vector was kindly provided by Prof. De-Yin Guo and Dr. On-Sam Sin. It was based on the backbone of natural miR-155 in which the natural miRNA:miRNA duplex sequence was replaced by the artificial one. Briefly, the natural miR-155 sequence was inserted into the pEGFP-N1 vector under a CMV promoter to create pSM155 vector. The oligonucleotides encoding amiRNA was inserted into the BsmBI site of miR-155 to replace the nature miRNA duplex sequence [19]. The 53Rexpressing plasmid (pEGFP-N3-53R) constructed in previous studies [7] was used in co-transfection experiments to express the target 53R.

Transfection
GCO and EPC cells were seeded into 12-well or 6-well cell culture plates using 199 medium containing 5% of FBS 24 h before transfection. Cells were transfected with plasmids using Lipofectamine 2000 (Invitrogen, U.S.A.) following the manufacturer's protocol. The transfection mixtures were removed at 6 h post transfection, and transfected cells were maintained for further processing.
To elucidate antiviral effect of amiR-53R-1 on RGV replication, the expression of 53R gene was monitored in GCO cells transfected with pSM155-amiR-53R-1 and infected with RGV at an MOI of 3 after 24 h of transfection. The pSM155-amiR-PB2 was used for negative control. At 72 h of p.i, cDNA obtained by method as above were used for real-time PCR analysis.
To assess inhibitory effect of amiR-53R-1 on RGV replication in terms of the production of viral particles, GCO cells transfected with pSM155-amiR-53R-1 and infected with RGV at an MOI of 3 after 24 h of transfection. Cell supernatants were collected at 24 h, 48 h, 72 h and 96 h p.i., respectively. The pSM155-amiR-53R-3 was used as comparison and pSM155-amiR-PB2 as negative control. Monolayer of GCO cells seeded in 96-well plates was inoculated with serial 10-fold dilutions of RGV samples for TCID 50 detection.

Electron microscope observation
EPC cells were transfected with pSM155-amiR-53R-1 and infected with RGV at an MOI of 3. Cells were centrifuged at 20006g 72 h p.i. and the pellets were fixed in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH = 7.4) for 1 day, rinsed in 0.1 M phosphate buffer containing 1% osmium tetroxide for 1 h at 4uC, dehydrated in a graded ethanol series and embedded in Spurr's resin. The cell pellets were post-fixed with 2% osmium tetroxide in sodium cacodylate buffer for 1 to 2 h, dehydrated in a graded acetone series, and embedded in LR White acrylic resin. Ultra-thin sections were post-stained with 2% uranyl acetate and 3% lead citrate and observed at 100 kV with a JEOL 1230 transmission electron microscope.

Real-time PCR
RNA extracted by Trizol reagent (Invitrogen, USA) was treated with DNase I and added to the reverse transcription reaction containing 200 U M-MLV reverse transcriptase (promega, USA), 0.5 mM each of dNTP, 16M-MLV reaction buffer, 500 ng of oligo(dT) 15 Primer and 25 U Rnasin (TOYOBO).
Real-time PCR was performed with 2 ml of cDNA in a final volume of 25 m1 containing 11.25 m1 of 2.56RealMasterMix/ 206SYBR solution(TIANGEN BIOTECH), 300 nM of sense primer and antisense primer. PCR was carried out in the 7300 Fast Real-time PCR System (Applied Biosystems, USA) using the following thermal cycling profile: 95uC for 3 min, followed by 40 cycles of amplification (95uC for 20 s, 58uC for 30 s, 68uC for 45 s). The absorption values of the SYBR Green I in each tube were detected at the end of each cycle. The melting curve analysis of PCR products from 55 to 95uC were also performed after PCR amplification. PCR was performed with the 53R forward primer (59-CATCAGAACGGGAGGACAGA-39), the 53R reverse primer (59-CGCCGTGTCGTCCTTGTAG-39), the b-actin forward primer (59-GATGATGAAATTGCCG CACTG-39), the b-actin reverse primer (59-ACCAACCATGACACCCTGATGT-39), the Mx forward primer (59-GCTGGAACGGGAGAAGGGAT-39), and the Mx reverse primer (59-CGCCGT GTCGTCCTTGTAG-39). Fragments were 244 bp (53R), 120 bp (b-actin) and 250 bp (Mx), respectively. The Fold Change (FC = 2 2DDCT ) in 53R and Mx mRNA expression levels were normalized to those of the bactin transcripts measured with the same cDNAs. The relative 53R and Mx mRNA expression values in each group were calculated by the mathematical delta-delta method [33] and all samples were run in triplicate.