The Mechanism of Synchronous Precise Regulation of Two Shrimp White Spot Syndrome Virus Targets by a Viral MicroRNA

MicroRNAs (miRNAs), important factors in animal innate immunity, suppress the expressions of their target genes by binding to target mRNA’s 3′ untranslated regions (3′UTRs). However, the mechanism of synchronous regulation of multiple targets by a single miRNA remains unclear. In this study, the interaction between a white spot syndrome virus (WSSV) miRNA (WSSV-miR-N32) and its two viral targets (wsv459 and wsv322) was characterized in WSSV-infected shrimp. The outcomes indicated that WSSV-encoded miRNA (WSSV-miR-N32) significantly inhibited virus infection by simultaneously targeting wsv459 and wsv322. The silencing of wsv459 or wsv322 by siRNA led to significant decrease of WSSV copies in shrimp, showing that the two viral genes were required for WSSV infection. WSSV-miR-N32 could mediate 5′–3′ exonucleolytic digestion of its target mRNAs, which stopped at the sites of target mRNA 3′UTRs close to the sequence complementary to the miRNA seed sequence. The complementary bases (to the target mRNA sequence) of a miRNA 9th–18th non-seed sequence were essential for the miRNA targeting. Therefore, our findings presented novel insights into the mechanism of miRNA-mediated suppression of target gene expressions, which would be helpful for understanding the roles of miRNAs in innate immunity of invertebrate.

miRNA targets in a laboratory is time consuming and expensive. Nowadays, the functions of many miRNAs have been identified (6)(7)(8)(9). However, most of the existing reports focus on a single target gene of one miRNA (10)(11)(12)(13). In fact, a single miRNA can target several genes by direct miRNA-mRNA interaction. As reported, multiple genes are the targets of miR-1/miR-206 in the progress of C2C12 myoblast differentiation (14). The melanoma cell invasion was suppressed through the downregulation of c-Met and YB1 by MiR-137 (15). Multiple key components in human Hippo pathway, for example, LATS2, β-TrCP, NDR2, and LZTS, were target genes of miR-135b (16). In shrimp, white spot syndrome virus (WSSV) encodes viral miRNAs targeting one or two viral or host genes during virus infection (17,18). Targets of viral miRNAs usually were identified as either viral regulatory proteins or proteins involved in the host immune system. For instance, viral T antigen of SV40 was a target of viral miR-S1. miR-S1 downregulated the viral T antigen to make sure cytotoxic T-cell would not attack the virus during late viral replication (19). Viral gene (HCMV IE1) is downregulated by miR-UL112-1 (20) was shown recently to express two miRNAs in latently infected neurons that are capable of downregulating ICP0 and ICP4 of HSV-1 were downregulated by its own miRNAs (21). Virus-induced apoptosis was inhibited by the downregulation of PUMA by miR-BART5 in EBV-infected cells (22). EBV miR-BHRF1-3 could target host chemokine, CXCL-11 (23). Another example, a few miRNAs of KSHV (miR-K5, miR-K9, and miR-K10) could knock down BCLAF1 (24). In human cancers, there is one-to-multiple relationship between a miRNA and its target genes (25). Although a single miRNA can target several genes, the interaction between a miRNA and its targets is evaluated using one miRNA-one target mRNA assay in all reports at present. Whether a single miRNA can simultaneously target different mRNAs in cells or/and in vivo has not been addressed yet. There is no report on the preference selection of target genes by a miRNA when the miRNA can target multiple genes.
To characterize the interaction between a miRNA and its multiple target mRNAs simultaneously, viruses may be the appropriate models. Viruses are the simplest organisms in the biological community (26,27). A virus possesses a small genome and a short life cycle. The life cycle of a virus is finished in its host cells. Viruses, such as all living beings, have the ability to be genetic, mutated and evolutionary (28)(29)(30). As reported, some viral miRNAs can prevent host defense systems by targeting host genes (18,19). More than half of the viral miRNAs are associated with virus infection. In this context, virus is one of the best models to investigate the miRNA-mRNA interaction.
To explore the interaction between a miRNA and its multiple genes in vivo, WSSV, a double-stranded DNA virus of shrimp, was employed in this study. Our previous studies have revealed that the viral miRNA WSSV-miR-N32 has two viral target genes (17,18,31). The results indicated that WSSV-miR-N32 could simultaneously target its target genes (wsv459 and wsv322). The wsv459 and wsv322 genes, transcribed at the early stage of WSSV infection, played important roles in virus infection. The complementary bases (to the target mRNA) of a miRNA 9th-18th non-seed sequence were required for the miRNA targeting.

resUlTs role of Viral mirna WssV-mir-n32 in Virus infection
To investigate the role of the viral WSSV-miR-N32 in WSSV infection, the expression level of WSSV-miR-N32 in WSSVchallenged shrimp was investigated. WSSV-miR-N32 could be detected by northern blots as early as 2 h postinfection ( Figure 1A), showing that this viral miRNA was transcribed at the very early phase of viral infection.
To evaluate the involvement of WSSV-miR-N32 in virus infection, the WSSV-miR-N32 expression was silenced or overexpressed in shrimp, followed by the examination of WSSV infection. The northern results showed that WSSV-miR-N32 was knocked down by AMO-WSSV-miR-N32 ( Figure 1B). The WSSV-miR-N32 silencing caused major increases of WSSV copies ( Figure 1C) and virus-infected shrimp mortality ( Figure 1D). To overexpress WSSV-miR-N32 in shrimp, the WSSV-miR-N32-mimic and WSSV were co-injected into shrimp. Northern blots showed that WSSV-miR-N32 was significantly overexpressed in shrimp at 12-36 h postinfection compared with the controls (Figure 1E). The WSSV-miR-N32 overexpression resulted in significant decreases of WSSV copies and the survival of WSSV-infected shrimp (Figures 1F,G), indicating that WSSV-miR-N32 could inhibit the WSSV infection in vivo.
Collectively, these data revealed that WSSV-miR-N32 acted as a negative regulator during virus infection.

Underlying Mechanism of WssV-mir-n32 in Virus infection
To explore the underlying mechanism of WSSV-miR-N32 in WSSV infection, the target genes of WSSV-miR-N32 were predicted. The prediction analysis revealed that two WSSV early genes (wsv459 and wsv322) might be the targets of WSSV-miR-N32 (Figure 2A). To evaluate the interaction between WSSV-miR-N32 and its targets, WSSV-miR-N32 mimic and the 3′UTR of target gene were co-transfected into High Five cells ( Figure 2B). The data revealed that the co-transfection of WSSV-miR-N32 and enhanced green fluorescent protein (EGFP)-wsv459-3′UTR or EGFP-wsv322-3′UTR significantly decreased the fluorescence intensity compared with the control, while the fluorescence intensity of cells co-transfected with WSSV-miR-N32 and EGFP-wsv459-3′UTR-mutation, or EGFP-wsv322-3′UTR-mutation did not change (Figures 2B,C). These findings indicated that WSSV-miR-N32 directly targeted wsv459 and wsv322 genes.
When WSSV-miR-N32 was overexpressed in WSSVchallenged shrimp, the expressions of viral wsv459 and wsv322 genes were significantly decreased in vivo compared with the control (Figure 2D). On the other hand, knocking down of WSSV-miR-N32 caused significant increases of wsv459 and wsv322 gene expressions in WSSV-infected shrimp ( Figure 2D). These results indicated that WSSV-miR-N32 inhibited virus infection by directly targeting the viral genes in vivo. The shrimp were infected with WSSV. At different time postinfection, the expression of WSSV-miR-N32 in shrimp hemocytes was detected with Northern blots. The probes used were indicated at the right. U6 was used as a control. (B) The silencing of WSSV-miR-N32 in shrimp. Both WSSV and anti-miRNA oligonucleotide (AMO)-WSSV-miR-N32 or AMO-WSSV-miR-N32-scrambled were co-injected into shrimp. At different time points postinfection, the shrimp hemocytes were collected and subjected to Northern blot analysis. The probes were indicated at the right. U6 was used as a control. (c) The influence of WSSV-miR-N32 silencing on the WSSV copies in shrimp. Quantitative real-time polymerase chain reaction (PCR) was conducted to quantify the virus copies in shrimp treated with WSSV and AMO-WSSV-miR-N32 or AMO-WSSV-miR-N32-scrambled. For each treatment, three shrimp were randomly selected and the mixed RNAs of three shrimp were analyzed by quantitative real-time PCR. (D) The evaluation of shrimp cumulative survival. The treatments were shown on the top. The shrimp mortality was examined at different time after treatment. (e) The overexpression of WSSV-miR-N32 in shrimp. Shrimp were co-injected with WSSV and WSSV-miR-N32-mimic or WSSV-miR-N32-mimic-scrambled, followed by Northern blots to detect the WSSV-miR-N32 expression. (F) The quantification of WSSV copies in shrimp. The virus copies in shrimp were evaluated using quantitative real-time PCR. Three shrimp, selected at random for each treatment, were used for this analysis. The treatments were indicated on the top. (g) The shrimp survival analysis. After the overexpression of WSSV-miR-N32, the shrimp mortality was examined. In all panels, the significant differences between treatments were indicated with asterisks (*p < 0.05; **p < 0.01). The significance of difference between treatments was evaluated with Student's t-test using the data of three independent assays for each treatment.

Functions of Viral wsv459 and wsv322 in Virus infection
To reveal the roles of viral wsv459 and wsv322 genes in virus infection, the time-course expressions of the two genes in the WSSV-challenged shrimp were investigated. Northern blots indicated that the wsv459 and wsv322 mRNAs could be detected at 6 h postinfection (p.i.) onward (Figure 3A), showing that the two genes were both early genes.
To explore the functions of viral genes in virus infection, the expressions of wsv459 and wsv322 genes were silenced in the WSSV-infected shrimp in vivo by sequence-specific siRNAs, respectively. Quantitative real-time PCR data revealed that the Regulation of Two Viral Targets by a miRNA Frontiers in Immunology | www.frontiersin.org November 2017 | Volume 8 | Article 1546 mRNAs of wsv459 and wsv322 genes were depleted compared with the control groups (Figures 3B,C). When mRNAs of wsv459 or wsv322 was knocked down by siRNA, the copies of WSSV were significantly decreased and the WSSV-infected shrimp survivals were increased (Figures 3D,E), indicating that the two early genes of WSSV took great effects on virus infection. These findings indicated that the viral wsv459 and wsv322 genes played positive roles in WSSV infection.

simultaneous interaction between a mirna and its Two Targets
To explore the simultaneous interaction between WSSV-miR-N32 and its two targets (wsv459 and wsv322 genes), WSSV-miR-N32, the synthesized wsv459 mRNA or wsv322 mRNA, and the shrimp Ago1 complex were mixed, followed by the detection of wsv459 mRNA or wsv322 mRNA degradation. Northern blots demonstrated that WSSV-miR-N32 mediated the degradation of wsv459 mRNA or wsv322 mRNA in a miRNA-concentrationdependent manner ( Figure 4A). When WSSV-miR-N32 was incubated with the mixture of wsv459 mRNA, wsv322 mRNA and Ago1 complex, the degraded wsv459 mRNA and wsv322 mRNA were detected ( Figure 4B). The results indicated that WSSV-miR-N32 could simultaneously mediate the degradations of its two target mRNAs in the existence of its two targets.
To investigate the interactions between WSSV-miR-N32 and its two targets wsv459 mRNA and wsv322 mRNA in vivo, the localizations of WSSV-miR-N32, wsv459 mRNA, and wsv322 mRNA were conducted in WSSV-infected shrimp hemocytes. The results showed that WSSV-miR-N32 was co-localized with wsv459 mRNA or wsv322 mRNA in the hemocytes of WSSVinfected shrimp (Figures 4C,D). When WSSV-miR-N32, wsv459 mRNA, and wsv322 mRNA were simultaneously labeled with different probes, the fluorescent signals showed that the three  The influence of WSSV-miR-N32 on viral gene expression in vivo. Shrimp were injected with WSSV and WSSV-miR-N32-mimic or anti-miRNA oligonucleotide (AMO)-WSSV-miR-N32. At 24 h after injection, the mRNA levels of wsv459 and wsv322 in shrimp hemolymph were examined with quantitative real-time polymerase chain reaction. Statistically significant differences between treatments were indicated by asterisks (*p < 0.05; **p < 0.01). molecules were co-localized ( Figure 4E), indicating that WSSV-miR-N32 was interacted with wsv459 mRNA and wsv322 mRNA in vivo. Stated thus, the above results suggested that a viral miRNA (WSSV-miR-N32) could simultaneously target its two targets (wsv459 mRNA and wsv322 mRNA) in vivo.

sites of a mirna non-seed sequence required for mirna Targeting
To further evaluate the sites of a miRNA required for the miRNA binding to its target mRNA, the cleaved fragments of wsv459 mRNA and wsv322 mRNA by WSSV-miR-N32-Ago 1 complex were sequenced. The results of time-course cleavage of 3′UTR showed that the wsv322 mRNA 3′UTR was cleaved, generating the cleaved fragment containing the sequence complementary to the seed sequence of WSSV-miR-N32, while the 5′ end of wsv322 mRNA 3′UTR was degraded ( Figure 5A). The time-course cleavage of the wsv459 mRNA 3′UTR yielded the similar results ( Figure 5B). These data indicated that a miRNA could mediate 5′-3′ exonucleolytic digestion of its target mRNAs and this 5′-3′ exonucleolytic digestion stopped at the sites of target mRNA 3′UTRs close to the sequence complementary to the miRNA seed sequence.
It is well known that the seed sequence of a miRNA plays a decisive role in the interaction between miRNA and its target genes. As reported, except for the miRNA seed sequence, some bases (13th, 16th, 17th, 18th, 20th, 21st, and 22nd) of non-seed sequence may be required for the miRNA targeting of its target gene (6)(7)(8). To explore the sites of a miRNA non-seed sequence required for the miRNA targeting, 13th, 16th, 17th, 18th, 20th, 21st, or 22nd base of WSSV-miR-N32 was mutated, followed by injection of mutated WSSV-miR-N32 into WSSV-infected shrimp. The 13th, 20th, and 22nd bases of WSSV-miR-N32 were Regulation of Two Viral Targets by a miRNA Frontiers in Immunology | www.frontiersin.org November 2017 | Volume 8 | Article 1546 complementary to the wsv459 mRNA sequence and the 18th, 21st, and 22nd bases of WSSV-miR-N32 were complementary to the wsv322 mRNA sequence. The results showed that the 16th and 17th base mutations of WSSV-miR-N32 had no effect on the wsv459 and wsv322 gene expression (Figure 5C), indicating that the base of a miRNA non-seed sequence, which was not complementary to the target mRNA sequence, took no effect on the miRNA targeting. The quantitative real-time PCR data demonstrated that when the 13th G of WSSV-miR-N32 was mutated, compared with the control group, the wsv459 mRNA level increased significantly (wild-type WSSV-miR-N32) (Figure 5D), indicating that the 13th G of WSSV-miR-N32 was required for the miRNA targeting. However, the 20th and 22nd base mutations of WSSV-miR-N32 had no effect on the wsv459 gene expression (Figure 5D). The analysis of interaction between WSSV-miR-N32 and wsv322 mRNA demonstrated that the 18th U of WSSV-miR-N32 was required for the miRNA targeting ( Figure 5E). These data revealed that the complementary bases of a miRNA 9th-18th non-seed sequence, which were complementary to the target mRNA sequences, were required for the miRNA targeting.

DiscUssiOn
Many studies have demonstrated that miRNAs play essential roles in virus-host interactions (17,18,32). They can regulate In all panels, the significant differences between treatments were evaluated with Student's t-test using the results of three independent experiments for each treatment and indicated with asterisks (*p < 0.05; **p < 0.01).  (38). Although there are some studies on multiple targets of one single miRNA, the miRNA-mRNA interactions are conducted using independent experiments of one (miRNA) to one (mRNA) reaction (11,17,39). However, the simultaneous regulation of multiple target genes' expressions by a single miRNA has not been intensively investigated. In this study, the findings showed that a viral miRNA (WSSV-miR-N32) could synchronously regulate its two target genes' expressions in vivo.
The WSSV-miR-N32-guided Ago1 complex simultaneously cleaved mRNAs of two viral target genes (wsv459 and wsv322), and the cleavage of wsv459 mRNA and wsv322 mRNA had no mutual interference. As the first line of immune responses, the innate immunity is fine regulated by miRNAs in animals (17,18,31,38). During virus infection, viral miRNAs can target viral mRNAs. The targeting of viral genes by viral miRNAs leads to virus infection or virus latency, which can be an efficient strategy for virus to resist the innate immunity of host. Therefore, the mechanism of the interaction between a single viral miRNA and its multiple target mRNAs, which was revealed in this study, contributed novel insights into the innate immunity of animals against virus infection.    At present, there is no unified conclusion about the effects of miRNAs in the RISC on their target mRNAs. Some reports show that a miRNA completely complementary or almost completely complementary to its target mRNA can guide the mRNA degradation (32,40,41). However, some studies indicate that the cleavage of target mRNA takes place as long as the seed sequence of a miRNA is complementary to its target mRNA (42,43). It is well known that RISC is a cytoplasmic effector of the miRNA pathway, in which a single stranded miRNA directs it to its target mRNA (44). In the RISC, Ago protein is the core component, which cleaves mRNAs. The findings revealed in this study and previous studies (42,43) show that the 3′ fragment of RISCcleaved mRNA can be detected, indicating that the 5′ fragment of mRNA is degraded. During the cleavage process of mRNA in the RISC, Ago acts as a 5′ phosphomonoester-producing RNA endonuclease (42,43). In this study, the results indicated that a miRNA could direct the 5′-3′ exonucleolytic cleavage of its target mRNA, which stopped at the sites of target mRNA 3′UTR close to the sequence complementary to the miRNA seed sequence, and the complementary bases (to the target mRNA sequence) of a miRNA 9th-18th non-seed sequence were required for the miRNA targeting. In this context, this investigation provided novel insights into the mechanism of miRNA-mRNA interactions in miRNA-induced silencing complex.

shrimp (Marsupenaeus japonicus) culture and Virus infection
Each shrimp (M. japonicus) was cultured in 80-L aquariums at 20-25°C. For each treatment, 20 shrimp were used. To ensure the absence of WSSV before experimental infection, three shrimp, selected at random, were subjected to the detection of WSSV using PCR with WSSV-PCR-primers (forward 5′-TATT GTCTCTCCTGACGTAC-3′ and reserve 5′-CACATTCTTCAC GAGTCTAC-3′). The total DNA was extracted from shrimp using SQ tissue DNA kit (Omega Bio-Tek, USA) according to the manufacturer's manual (17). 100 µL of WSSV (10 5 copies/mL) inoculum/shrimp was injected into each virus-free shrimp (17). Shrimp hemocytes were gathered for later use at different time points postinfection (17,18). All
To overexpress WSSV-miR-N32, the synthesized WSSV-miR-N32-mimic was injected into shrimp. According to the manufacturer's manual, WSSV-miR-N32-mimic (5′-GGGGAA GGUGACGUCCGUUUGA-3′) was synthesized using T7 Kit siRNA Synthesis (TaKaRa, Japan). WSSV-miR-N32-mimicscrambled (5′-GGAAGGGGUGACGUCCGUUUGA-3′) was used as a control. The synthesized miRNAs would be dissolved in Tris buffer (50 mM Tris·HCl (pH 7.5), 100 mM NaCl), then quantified by NanoDrop 2000. WSSV (10 5 copies/mL) and a miRNA mimic (30 nM) were co-injected into shrimp. 12 h later, the miRNA (30 nM) was injected into the same shrimp. PBS and WSSV only (10 5 copies/mL) were used as controls. At various  . (a,B) The miRNA-guided cleavage of target mRNA. White spot syndrome virus (WSSV)-miR-N32 was incubated with the Ago1 complex and 3′ untranslated region (3′UTR) of wsv322 mRNA (a) or 3′UTR of wsv459 mRNA (B) for different time as indicated on the left. Then, the cleaved fragment was sequenced. The sequence complementary to the seed sequence of WSSV-miR-N32 was underlined. (c) The influence of uncomplementary bases of WSSV-miR-N32 non-seed sequence on the expression of its target genes. The mutated or wild-type WSSV-miR-N32 was co-injected with WSSV into shrimp. The shrimp were cultured for 36 h. Then, the shrimp hemocytes were subjected to quantitative real-time polymerase chain reaction (PCR) to detect wsv322 mRNA or wsv459 mRNA. The arrows represented the mutated sites and bases of WSSV-miR-N32. (D,e) The effects of base mutation of WSSV-miR-N32 non-seed sequence on the expression of its target genes. WSSV and wild-type or mutated WSSV-miR-N32 were injected into shrimp. Thirty-six hours later, the wsv322 mRNA level (D) or wsv459 mRNA level (e) in shrimp hemocytes was examined with quantitative real-time PCR. The arrows indicated the mutated sites and bases of WSSV-miR-N32. In all panels, the statistically significant differences between treatments were shown with asterisks (**p < 0.01). times after first injection, shrimp hemolymph was collected for later use.

Detection of WssV copies Using Quantitative (Q) real-time Pcr
Shrimp hemocytes of various treatments were collected at different times postinfection. According to the manufacturer's manual, viral DNA was extracted using a Tissue DNA Kit (Omega Bio-Tek, USA). Then, the copies of WSSV were evaluated using quantitative PCR (qRT-PCR) with specific primers of WSSV (forward 5′-TTGGTTTCAGCCCGAGATT-3′ and reverse 5′-CC TTGGTCAGCCCCTTGA-3′) and specific TaqMan

Mortality of shrimp
The shrimp were cultured with different treatments (20 shrimp per treatment). The cumulative shrimp mortality was examined at different time postinfection (12,24,36,48, and 72 h).

Prediction of Target genes of mirna
To get the target sequences of a miRNA, the genome of WSSV (GenBank accession no: AF332093) was used. The target genes of a miRNA were predicted using three independent bioinformatic softwares including pictar, 1 miRanda, 2 and TargetScan 5.1. 3 The overlapped genes of three algorithms might be the targets of a miNRA.

Fluorescence Plasmid construction
PIZ/EGFP V5-His plasmid (Invitrogen, USA) that contains an EGFP gene was used for the plasmid construction. Primers used as follows: EGFP-wsv459-3′UTR (primers forward 5′-AA CTCTAGAATCTAATTACAGAGTATTCTA-3′ and reverse 5′-AA C C G C C G AC ATG ATC C A AC TG C ATAG TC AC C C G T AT-3′) and EGFP-wsv322-3′UTR (primers forward 5′-AACT CTAGAAATTCTGGCTACCAGAACCCA-3′ and reverse 5′-AA C C G C C G C G T T G C A A G G A A A C A G T G C T C A ATA CCT-3′). Each 3′UTR was linked into downstream of EGFP of the PIZ/EGFP plasmid. SacII and XbaI restriction sites were used. Dpn I-mediated site-directed mutagenesis (New England BioLabs, USA) were used to obtain the mutations. Generally, two incomplete complementary primers were used to mutate the 3′UTR sequence by PCR. The incomplete complementary primers contained two mismatched nucleotides. Dpn I was added into the PCR product. Then the mixture was incubated at 37°C (1 h). After that, DNA was purified and transformed into competent cells DH5α. All the sequences of above vectors were confirmed by DNA sequencing.
high Five cell culture, Transfection, and Fluorescence assays At temperature of 27°C, Trichoplusia ni cell line High Five (Invitrogen, USA) was cultured in Express Five medium (Invitrogen, USA) with 1.8 mM/mL l-glutamine (Invitrogen, USA) were cultured. The cells were co-transfected with a synthesized miRNA mimic (300 nM) and pIZ/EGFP consisting of 3′UTR of wsv459 or wsv322 (6 µg/mL) at about 70% confluence. The miRNA mimic was synthesized by T7 Kit siRNA Synthesis (TaKaRa, Japan). All experiments of transfection were accomplished with Cellfectin (Invitrogen) according to Cellfectin's manual in triplicate. 12 h for after culture, cells were planked with cell density of 2.0 × 10 4 cells/well to 96-well plates. The fluorescence density of High Five was detected using a Flex Station II microplate reader (Molecular Devices, USA) with condition of Ex (490 nm)/Em (510 nm) at 48 h post-transfection. The fluorescence value was revised by subtracting the spontaneous fluorescence of non-EGFP cells.
RNAi assay was carried out by injecting siRNA (30 μg/ shrimp) into shrimp by syringes. First, siRNA (15 μg/shrimp) was injected into virus-free shrimp. Twelve hours later, WSSV (10 5 copies/shrimp) and siRNA (15 µg) were co-injected into the same shrimp. At different time after the last injection, the shrimp hemocytes were harvested. Three samples of shrimp from each treatment, randomly selected, were harvested for experiments. The aforementioned assays were repeated three times for biological repetition.

mirna-Mediated Degradation of Target mrnas
The shrimp Ago1 gene was expressed in E. coli. Subsequently, the purified recombinant Ago1 protein was used to immunize mice to get the Ago1-specific antibody.

GATCAC T A A T A C G A C T C A C T A T A G G G G G A G C T A C G G
AA-3′ and 5′-GATGCATCGAT CGTAGCCGCGCAT-3′), respectively. Then the 3′UTR of wsv459 or wsv322 was synthesized using a T7 Kit (TaKaRa, Japan). The mixture of Ago1 complex, WSSV-miRN-32, wsv459 3′UTR, or/and wsv322 3′UTR was electrophoresed on a 1% agarose gel after incubation. And then the RNAs were transferred to a Hybond-N+ (RPN303B) nylon membrane (GE Amersham, USA). After UV cross-linking, it was incubated in DIG Easy Hyb granule buffer (Roche, Switzerland) for 0.5 h at 42°C for prehybridization and then incubated in the same buffer with DIG-labeled wsv459 probe (5′-AGCGATGCGCGC TAGACTAGTCGATCG-3′) or/and wsv322 probe (5′-GCATC GCATGCTAGCGCGCATGCATC-3′) at 42°C overnight for hybridization. The alkaline phosphatase reaction was carried out by DIG DNA Labeling and Detection Kit II (Roche, Switzerland).

sequencing of the cleaved mrna 3′UTr Fragment
The mRNA 3′UTR of WSSV-miR-N32 target gene was incubated with WSSV-miR-N32 (1,000 nM) and Ago1 complex for different time (0.5, 1, or 2 h). After separation by agarose gel electrophoresis, the cleaved RNA fragment was recovered. Then, the reverse transcription was conducted using PrimeScript ® first strand cDNA synthesis kit with the Random 6 primer (TaKaRa, Shiga, Japan). The double-stranded DNA was synthesized by second strand cDNA synthesis kit (Beyotime Biotechnology, Shanghai, China). Subsequently, the dsDNA was ligated into T vector (TransGen Biotech, Beijing, China) and sequenced.

identification of Bases of a mirna non-seed sequence required for mirna Targeting
To identify the bases of a miRNA non-seed sequence required for miRNA targeting, 13th, 16th, 17th, 18th, 20th, 21st, or 22nd base of WSSV-miRN-32 (5′-GGGGAAGGUGACGUCCGUUUGA-3′) was mutated. The mutated WSSV-miRN-32 was synthesized using a T7 Kit siRNA Synthesis (TaKaRa, Japan) in vitro. The size of mutated miRNAs was detected by agarose gel electrophoresis. The synthesized miRNA was dissolved in Tris buffer (50 mM Tris·HCl (pH 7.5), 100 mM NaCl), then quantified by NanoDrop 2000. Virus-free shrimp were co-injected with WSSV (10 5 copies/mL) and the wild-type or mutated WSSV-miRN-32 (15 µg) (100 µL per shrimp). 12 h after the first injection, another dose of miRNA (15 µg) was injected into the same shrimp. WSSV only group was used as a control group. 36 h after the last treatment, three samples of shrimp from every group, randomly selected, were harvested for experiments. The above experiments were repeated three times for biological repetition.

statistical analysis
All numeric data gathered from above independent treatments were managed by one-way analysis of variation. Student's t-test was used to check the significant differences between different treatments.
aUThOr cOnTriBUTiOns XZ designed the study and reviewed and edited the manuscript. XZ and YH performed the experiments and wrote the paper. All the authors read and approved the manuscript.

FUnDing
This study was supported by National Natural Science Foundation of China (NSFC) (31430089) and National Program on Key Basic Research Project of China (2015CB755903).