Cellular microRNA miR-26a suppresses replication of porcine reproductive and respiratory syndrome virus by activating innate antiviral immunity

Porcine reproductive and respiratory syndrome (PRRS) has caused large economic losses in the swine industry in recent years. Current PRRS vaccines fail to effectively prevent and control this disease. Consequently, there is a need to develop new antiviral strategies. MicroRNAs play critical roles in intricate host-pathogen interaction networks, but the involvement of miRNAs during PRRS virus (PRRSV) infection is not well understood. In this study, pretreatment with miR-26a induced a significant inhibition of PRRSV replication and remission of the cytopathic effect in MARC-145 cells, and this antiviral effect was sustained for at least 120 h. Luciferase reporter analysis showed that the PRRSV genome was not the target of miRNA-26a. Instead, RNA-seq analysis demonstrated that miR-26a significantly up-regulated innate anti-viral responses, including activating the type I interferon (IFN) signaling pathway and promoting the production of IFN-stimulated genes. These findings suggest that delivery of miR-26a may provide a potential strategy for anti-PRRSV therapies.

In the present study, we investigated the exact role of the overexpression miR-26a during PRRSV infection and the miRNAs studied in previous reports 9, 12 were used as controls. We found that miR-26a inhibits PRRSV replication in MARC-145 cells. Furthermore, this phenomenon was not related to miR-26a binding to the PRRSV genome but was the result of up-regulating the innate immune response. Our results are consistent with the emerging notion that miRNAs are broadly involved in viral infection of mammalian cells and suggest that miR-26a may represent a potential therapeutic target for antiviral intervention of PRRSV infection.
It was previously reported that miR-26a suppresses the replication of H1N1 (WSN) IAV 12 . To determine if miR-26a is a non-specific anti-viral agent, the inhibition of VSV, HSV-1 and PRRSV by miR-26a was investigated. Notably, these viral genomes do not contain putative miR-26a binding sites as defined by the MicroInspector program (data not shown). MARC-145 cells were transfected with 1 μ g miR-26a and, 24 h later, infected with IAV (H1N1 or H9N2), VSV, HSV-1, or PRRSV (VR2332 or JXwn06) at a multiplicity of infection (MOI) of 0.1. The viral titers at 24 hpi were determined as the 50% tissue culture infective dose (TCID 50 ); the transfection efficiency was > 20%, as determined using pEGFP-N1 (data not shown). We found that miR-26a decreased the total viral titers of IAV, VSV and HSV-1 by approximately 10-fold (p < 0.05, Fig. 2A-C) and significantly inhibited the replication of PRRSV > 100-fold (p < 0.001, Fig. 2D). These data suggest that miR-26a may act as a broad-spectrum viral antagonist, especially for PRRSV.
Additionally, the level of miR-26a in the transfected cells from 12-120 hpt was validated using real-time PCR. Transfection with plasmids expressing miR-26a resulted in a > 50% increase in miR-26a levels in MARC-145 cells at 12 hpt, which peaked from 24-48 hpt (up to 3-fold) and gradually decreased to normal levels by 96 hpt (Fig. 2E). Next, MTT assays were used to determine if miR-26a affects cell viability. with plasmids encoding miR-26a or control (miR-Ctr). Twenty-four hours post-transfection, the cells were infected with H1N1 and H9N2 IAV (A), VSV (B), HSV-1 (C), VR2332, and JXwn06 PRRSV (D) at a MOI = 1 for 72 h. The viral titers were determined as the TCID 50 . (E) The relative expression of miR-26a in MARC-145 cells transfected with miR-26a or miR-Ctr, as well as in mock-transfected cells, was measured by real-time PCR at the indicated times. (F) The viability of MARC-145 cells mock-transfected or transfected with miR-26a or miR-Ctr was determined by MTT assays up to 120 hpt. These data are expressed as the mean ± SD and compared to the miR-Ctr group (A-D) or mock group (E-F) by paired-sampled t-tests and two-way ANOVA; * P < 0.5, ** P < 0.01, and *** P < 0.001.
Scientific RepoRts | 5:10651 | DOi: 10.1038/srep10651 As shown in Fig. 2F, the viability of MARC-145 cells remained unchanged up to 120 hpt, regardless of the up-regulation of miR-26a expression levels.

Dose-dependent and time-duration inhibition of viral replication by miR-26a. The inhibitory
effect of miR-26a described above was observed at a MOI of 0.1. Thus, we next investigated whether miR-26a displays the same suppression at lower and higher MOIs (0.001, 0.01, 0.1, 1 and 10) of PRRSV infection. MARC-145 cells were first transfected with 1 μ g plasmid (miR-26a or miR-Ctr), followed by PRRSV VR2332 strain infection at 24 hpt. At 72 hpi, the cells were examined for the cytopathic effect (CPE), the culture supernatants were measured by TCID 50 , and cell lysates were analyzed by western blotting (WB). We found that the viral titers were significantly reduced by approximately 1,000-to 10,000-fold by miR-26a (Fig. 3A). Additionally, the levels of N protein were either undetectable at MOI = 0.001-0.1 or remarkably decreased at MOI = 1 and 10 (Fig. 3B). Further, in the presence of miR-26a, no CPE was induced by PRRSV at any MOI tested (Fig. S1).
To examine the duration of viral replication inhibition caused by miR-26a, MARC-145 cells were first transfected with 1 μ g plasmid (miR-26a or miR-Ctr), followed by PRRSV infection (MOI = 0.1) at 24 hpt. At 24, 48, 72, 96, and 120 hpi, the culture supernatants were measured by TCID 50 , and cell lysates were analyzed by WB. As shown in Fig. 3C, compared to the control groups (virus only and miR-Ctr), miR-26a resulted in > 100-to nearly a one million-fold reduction of virus yield during replication (24-120 hpi). Similarly, no CPE occurred until 96 hpi (data not shown). The same inhibition was also observed at the protein level, as measured by N protein in Fig. 3D. These results suggest that the inhibitory effect induced by miR-26a lasts for at least 120 hpi.
To measure the dose-dependent suppression of viral replication, MARC-145 cells seeded in 24-well plates were first individually transfected with the miR-26a or miR-Ctr plasmids in increasing amounts (0, 0.25, 0.5, 0.75, and 1 μ g) and then infected with PRRSV (MOI = 0.1) at 24 hpt. At 48 hpi, the viral titers in the supernatants were measured by TCID 50 , and the cell lysates were analyzed by WB using an anti-N polyclonal antibody. We found that the cells transfected at a higher dose of miR-26a displayed a stronger antiviral effect, both at the virus titer and N protein expression levels ( Fig. 3E-F).
To test if miR-26a could suppress PRRSV in previously infected cells, MARC-145 cells were infected with 0.1 MOI PRRSV, and 0, 4, or 8 h later, these cells were transfected with the miR-26a or miR-Ctr plasmids. At 48 hpi, cells were collected, and the viral titers were measured. As shown in Fig. 3G-H, miR-26a expression in pre-infected cells was also able to effectively inhibit viral replication. The efficiency of virus replication suppression induced by miR-26a at 4 or 8 h after viral infection was not different from that at 0 h after viral infection, and all of the associated viral titers were approximately 100-fold lower than that of the control groups (virus only and miR-Ctr). miR-26a suppresses the replication of PRRSV by impairing the synthesis of viral subgenomic mRNAs, and its inhibitor exacerbates PRRSV infection. We performed parallel assays in MARC-145 cells by transfecting either the miRNA expression vector (miR-26a or miR-Ctr) or miRNA inhibitor (anti-miR-26a or anti-control). At 24 hpt, the cells were inoculated with PRRSV at 0.1 MOI, and at 48 hpi, the culture supernatants and cells were harvested for further analysis (or 72 hpi, only for CPE observation).
Second, at 72 hpi, the infected cells from the control groups (miR-Ctr, anti-Control, and virus only) became reticulated and detached from the monolayer. However, no CPE was observed in the miR-26a and mock groups, while CPE in the anti-miR-26a group was increased relative to the infected control groups (Fig. 4B).
Finally, at 48 hpi, the viral titers of the culture supernatants were measured as the TCID 50 (Fig. 4C), the PRRSV N protein was detected by IFA ( Fig. 4D) or WB ( Fig. 4E) with anti-N polyclonal antibodies, and the N mRNAs were quantified by real-time PCR (Fig. 4F). Our results demonstrate that at the protein, RNA, and viral levels, the replication of PRRSV was significantly inhibited by the over-expression of miR-26a, and PRRSV replication was aggravated by knock-down of miR-26a via miR-26a inhibitor transfection.
The genome of PRRSV is not the target of miR-26a. To determine if miR-26a directly targets the PRRSV genome and/or regulates host genes, we first used two algorithms (PicTar and TargetScan Release 5.1) to predict the PRRSV gene targets of miR-26a. We found that the seed sequence of miR-26a was partially complementary to each of the PRRSV subgenomic sequences. Thus, each PRRSV ORF was inserted between the 3` end of the luciferase gene and the 5' end of the poly(A) tail in the luciferase expression vector pRL-TK (Promega) (Fig. 5A). MARC-145 cells were then co-transfected with miR-26a (or miR-Ctr) and TK-ORFs, and luciferase activity assays were performed. As shown in Fig. 5B, miR-26a did not target the genome of PRRSV. To confirm this result, recombinant plasmids encoding the non-structural and structural proteins of PRRSV (including nsp2, nsp9, Gp2a, E, Gp3, Gp4, Gp5, M and N) with a myc-tag at their N-terminus were co-transfected with either miR-26a or miR-Ctr, and IFA with an anti-myc monoclonal antibody was performed to determine if miR-26a inhibited the production of each protein. Figure 5C shows that ectopic expression of miR-26a did not reduce the fluorescence intensity.

Figure 3. Inhibition of viral replication by miR-26a in MACR-145 cells was dose dependent and long lasting.(A-B)
MARC-145 cells seeded in 24-well plates were first transfected with 1 μ g plasmid (miR-26a or miR-Ctr) or mock transfected, followed by PRRSV VR-2332 infection at the indicated MOI or mock infection at 24 hpt. At 72 hpi, the culture supernatants were measured by TCID 50 (A) and the cell lysates were analyzed with an anti-N polyclonal antibody by WB (B). (C-D) MARC-145 cells were first individually transfected with 1 μ g plasmid expressing miR-26a or miR-Ctr or mock transfected, followed by PRRSV VR-2332 infection (MOI = 0.1) or mock infection at 24 hpt. At 24, 48, 72, 96, and 120 hpi, the culture supernatants were measured by TCID 50 (C) and cell lysates were analyzed by WB (D). (E-F) MARC-145 cells seeded in 24-well plates were first transfected with the miR-26a or miR-Ctr plasmid in increasing amounts (0, 0.25, 0.5, 0.75, and 1 μ g), followed by PRRSV VR-2332 infection (MOI = 0.1) at 24 hpt. At 48 hpi, the supernatants of the cell cultures were harvested and measured by TCID 50 as the viral titers (E) and the cell lysates were analyzed by WB (F). (G-H) MARC-145 cells were first infected with 0.1 MOI PRRSV VR-2332, and 0, 4, or 8 h later, the cells were transfected with plasmid expressing miR-26a or miR-Ctr or mock transfected. At 48 hpi, the culture supernatants were measured by TCID 50 (G) and the cell lysates were analyzed by WB (H). These data are expressed as the mean ± SD and compared to the miR-Ctr group (A, C, G) or "0 μ g" group (E) by paired-sampled t-tests; * P < 0.5, ** P < 0.01, and *** P < 0.001.
To further confirm this relationship, we used real-time RT-PCR to determine the effect of miR-26a on the mRNA levels of PRRSV during the early stage of viral infection (6 hpi at MOI = 10; the primer sequences are available from the corresponding author upon request). When miR-26a was over-expressed, the levels of each mRNA in the subgenome were dramatically decreased (Fig. 5D). This indicates that miR-26a inhibited viral infection by regulating host genes that are involved in the control of PRRSV replication.

miR-26a triggers the INF signaling pathway.
To investigate the antiviral mechanism of miR-26a, RNA-seq was used to determine the changes in cellular transcript levels after miR-26a or miR-Ctr treatment with or without PRRSV infection. The samples were divided into four groups: miR-Ctr, miR-Ctr/V, miR-26a, and miR-26a/V (where "V" stands for PRRSV infection). Sixteen billion tags were obtained from sequencing (four billion for each sample). Using the Cuffdiff program, the numbers of differentially expressed genes (DEGs) between any two groups are shown in Table 1. Of the total transcripts, 156 were significant hits in the miR-26a-treated library, and 215 were significant in the PRRSV-infected plus miR-26a-transfected library. Ninety-five were found in both of the libraries (Fig. 6A). In addition, 150 DEGs  [1][2][3][4][5] were performed as followed: Exp. 1, at 24 hpt, the relative expression of miR-26a in MARC-145 cells was detected by real-time PCR (A); Exp. 2, at 72 hpi, CPE was observed (B), and the culture supernatants were harvested for TCID 50 detection (C); Exp. 3-5, at 48 hpi, the N protein of PRRSV was detected by IFA (D) or WB (E) with an anti-N polyclonal antibody, and the mRNA of N was detected by real-time PCR (F). # indicates the efficiency of infection, expressed as the percent of N-positive cells relative to all cells from three independent experiments. These data are expressed as the mean ± SD and compared to the miR-Ctr group (A, C, D and F) by paired-sampled t-tests; * P < 0.5, ** P < 0.01 and *** P < 0.001.
were found after PRRSV infection in MARC-145 cells, 124 DEGs were significant after virus infection in the miR-26a transfected cells, but only 18 hits were found in both libraries (Fig. 6A).
To uncover functional characteristics of the DEGs among the four groups, we performed an enrichment analysis of Gene Ontology (GO) terms (biological process) and KEGG pathways (Fig. 6B-D). Interestingly, two clusters were associated with distinct functional terms, including the IFN signaling pathway and the antigen presentation pathway (Fig. 6C-D, Table 2). Heat-maps indicate that both the IFN signaling pathway and antigen presentation pathway were triggered by miR-26a and then enhanced by PRRSV infection (Fig. 6C). Some IFN-stimulated genes were significantly up-regulated by miR-26a,    including MX1, IFI44/IFI44L, the OAS family, and the IFIT family. The expression of other antiviral genes, such as RSAD2 and BST2, were also enhanced by miR-26a. Moreover, the expression of several chemokines (CXCL10), cytokines (IL-17A and IL28BP), and complements (CFB) was also increased.

Discussion
miRNAs are crucial post-transcriptional regulators of many biological systems, including mammalian immune systems 18 , which function as host defenses against both RNA and DNA viruses. Together, the host-or virus-encoded miRNAs and their target genes form novel regulatory networks between the host and the virus [8][9][10][11][12][13][14][15] . In the present study, we demonstrated that miR-26a is an effective antagonist against several RNA viruses (including IAV, VSV, and PRRSV) and a DNA virus (HSV-1). Transfection of a miR-26a-expression plasmid before and after PRRSV infection significantly inhibited the replication of PRRSV in cells.
Many recent studies indicate that miR-26a is involved in physiological and pathological processes such as proliferation, megakaryocytopoiesis, innate immunity, neurodegenerative diseases, and tumorigenesis. For instance, miR-26a functionally antagonizes human breast carcinogenesis by targeting MTDH and EZH2 (a component of the Polycomb repressive complex 2) 19 . Kota et al. (2009) report that delivery of miR-26a results in dramatic protection from hepatocellular carcinoma (HCC) by targeting cyclins D2 and E2 20 . Salvatori et al. (2011) find that expression of miR-26a in acute myeloid leukemia cells inhibits cell cycle progression by down-regulating cyclin E2 expression, and they highlight miR-26a as an attractive therapeutic target in leukemia 21 . Through targeting the EZH2 gene and repressing the expression of EZH2, miR-26a dramatically suppresses cell proliferation and colony formation by inducing G(1)-phase cell cycle arrest and inhibits tumorigenesis of nasopharyngeal carcinoma 22 . Further, miR-26a1/2 and miR-26b down-regulate BDNF (brain-derived neurotrophic factor) expression, which is a neurotrophin that plays an essential role in neuronal development and plasticity 23 . In the process of gliomagenesis, miR-26a is a direct regulator of PTEN (phosphatase and tensin homolog), which is involved in the Akt pathway, and highlights dysregulation of Akt signaling as crucial to the development of glioma 24 . Ectopic expression of miR-26a influences cell cycle progression by targeting the bona fide oncogene EZH2, a polycomb protein and global regulator of gene expression that was not previously known to be regulated by miRNAs 25 . Finally, osteogenic differentiation of human adipose tissue-derived stem cells is modulated by miR-26a targeting of the SMAD1 transcription factor 26 .
In this report, we found that a miR-26a-mediated antiviral mechanism is utilized by host cells to defend against PRRSV infection. However, whether the targets of miR-26a (including BDNF, cyclin E2, EZH2, MTDH, PTEN, and SMAD1) are involved the replication of PRRSV remains unknown. Thus, our future work will focus on the identification of targets of miR-26a.
In elucidating the mechanism of miR-26a antiviral activity, luciferase reporter assays and IFA were used to determine if the viral genome is the target of miRNA-26a. The prevailing wisdom is that if the sequences of the miRNA and its mRNA targets are perfectly matched, the target mRNA is degraded. However, repression of translation is the most common effect if the seed sequences of the miRNA and the target mRNA only partially match one another. The subgenomic genes of VR-2332 PRRSV, including the 3` and 5` UTRs, were directionally cloned into the 3`-UTR of the luciferase gene in the pRL-TK vector (Fig. 5A). If miR-26a targeted the genome of PRRSV, luciferase activity would decrease. However, our results demonstrate that ectopic expression of miR-26a did not alter luciferase activity, indicating that miR-26a did not target the PRRSV genome (Fig. 5B). In addition, during the early stage of PRRSV infection (6 hpi), the synthesis of all viral subgenomic mRNAs was impaired by miR-26a, which corresponded to the virus titers measured after miR-26a transfection (Fig. 3). These results indicated that some cellular transcription factor(s) participating in PRRSV transcription and/or replication are regulated by miR-26a (Fig. 5D).
Similar to our results, Li et al. (2015) also prove that miR-26a suppresses PRRSV replication in cells using a synthesized miR-26a mimic instead of a plasmid-based expression system. Furthermore, their luciferase reporters show that miR-26a does not directly target the PRRSV genome but instead affects the expression of type I INF and the IFN-stimulated genes MX1 and ISG15 during PRRSV infection 27 . To systematically study the anti-PRRSV infection mechanism of miR-26a, host cell transcripts were analyzed by RNA-seq after transfecting miR-26a with or without PRRSV infection in MARC-145 cells. The results showed that the MX1 and ISG15 genes are only two of the significant number of factors involved in regulating anti-viral responses, including the RIG-I-and TLR-mediated IFN signaling pathways (Fig. 6C, Table 2). The transcripts of RIG-I, MDA5, IRF7 and TLR3 were also significantly up-regulated by miR-26a (Table 2). In addition, many ISGs were activated by miR-26a, including OASL, OAS1, OSA2, OAS3, MX1 and ISG15, which have high antiviral activity 28 . Although miR-26a activated the RIG-I/MDA5 and TLR3/MYD88 IFN signaling pathways, the expression of IFN-β was not increased in this study, which differs from the results of luciferase reporters 27 . The most probable reason for this phenomenon is that miR-26a may target the 3`-UTR of the IFN-β gene and mediate its mRNA degradation 16 . These results provide further evidence that miR-26a could negatively regulate PRRSV replication by enhancing the innate immune response. Thus, cellular conditions and the transfection efficiency of plasmids expressing miR-26a may influence the antiviral function. Further, some differences in virus titers after miR-26a transfection were observed in Fig. 3A,3C,2D, as well as in the results of WB detection for PRRSV N protein in Fig. 3,4, though the antiviral functions of mi-R26a were significant in those experiments.
Vaccination and drugs are the principal means used to control and treat PRRSV infection. An array of PRRS vaccines have been developed, but none provide sustainable disease control because they suffer from both immune suppression and evasion strategies of the virus, as well as the antigenic heterogeneity of field strains. Effective drugs would be another useful strategy to control this disease. INFs, including type I (IFN-α , IFN-β ) and type III (IFN-λ ), are widely studied for their antiviral effects against PRRSV 17,29 . Further, Wei et al. (2011) explore the inhibitory effects of indigowoad root polysaccharides on PRRSV replication in MARC-145 cells 30 .
Unfortunately, there is currently no effective drug that can be used to treat PRRSV infection. Hence, it is imperative to develop other antiviral strategies to combat PRRSV infection, such as RNA interference (RNAi). Previous research demonstrates that PRRSV replication is inhibited by RNAi induced by recombinant plasmids or adenoviruses expressing shRNAs targeting viral subgenomes [31][32][33][34][35] . Recently, Xiao et al. (2011) were the first to attempt to suppress HP-PRRSV replication in MARC-145 cells through vector-based and lentiviral-mediated artificial microRNAs targeting the GP5 or M protein coding sequences of PRRSV 36 . Guo et al. (2013) report that miR-181 directly impairs PRRSV infection by specifically binding to ORF4 of the viral genomic RNA 8 .
Here, we aimed to investigate if any cellular miRNA(s) could persistently inhibit PRRSV replication. We found that cellular miR-26a significantly inhibits the replication of the highly pathogenic JXwn06 strain, as well as the traditional VR2332 strain, up to 120 hpi in a dose-dependent manner (Fig. 2-3), with evidence of alleviating CPE ( Fig. 4 and Fig. S1), reducing the viral titer in cell supernatants ( Fig. 1-4), and inhibition of viral RNA and protein synthesis (Fig. 3-5). PRRSV can persist in pigs for long periods of time after the initial infection, and persistently infected animals can shed infectious virus for several months 37 . To address whether miR-26a could counteract virus replication in infected cells, MARC-145 cells were transfected with miR-26a 0, 4, and 8 hpi with PRRSV VR2332. Our results show that the viral titers were reduced by approximately 2 logs of TCID 50 in each experimental group transfected with miR-26a ( Fig. 3G-H), suggesting that miR-26a might be applied as an effective anti-PRRSV strategy during PRRS outbreaks.

Plasmid transfection and virus infection.
The plasmids and miRNA inhibitors were transfected into cells using Lipofectamine TM 2000 (Invitrogen) according to the manufacturer's recommendations. The GFP expression plasmid pEGFP-N1 was used as an internal control. Twenty-four hours post-transfection, the cells were washed with PBS three times and infected with virus in infection medium (fresh DMEM with 2% FBS, except for IAV, which required fresh DMEM lacking FBS and containing 2.5 g/ml trypsin) at the indicated MOI. After incubation for 1 h, the cells were washed three times with PBS, and medium was added to the cells.
Indirect immunofluorescence assay (IFA). MARC-145 cells seeded in 24-well plates were transfected with miR-26a and control (miR-Ctr) plasmids for 24 h and then infected with VR2332 PRRSV at MOI = 0.1. At 48 h post-infection, MARC-145 cells were washed with PBS, fixed in 4% paraformaldehyde, and permeabilized with PBST (0.1% Triton X-100 in PBS). Then, they were incubated for 1 h at 37 °C with 4% BSA in PBS followed by incubated with a polyclonal antibody against the N protein (1:500 mouse sera) for 1 h at 37 °C. After washing five times with PBST, the cells were incubated for 1 h at 37 °C with TRITC-conjugated anti-mouse IgG (1:100). After washing five times with PBST, the cells were stained with DAPI (4`,6-diamidino-2-phenylindole) and observed under a Leica confocal microscope.
Western blotting (WB) analysis. WB was performed as previously described 9 . PVDF membranes were probed with a 1:2,000 dilution of mouse anti-N and/or anti-GP5 polyclonal antibody. To normalize protein loading, the PVDF membranes were simultaneously incubated with mouse anti-β actin monoclonal antibody at a dilution of 1:3,000. HRP-conjugated goat anti-mouse IgG (1:5,000) was used as a secondary antibody.
Virus titration. MARC-145 cells were trypsinized and seeded in 96-well plates 24 h before virus infection. Viral supernatants were 10-fold serially diluted with infection medium and added to each well with eight technical replicates of each dilution. One to five days (based on the indicated types of virus) after infection, the TCID 50 was calculated by the Reed-Muench method 42 .

RNA quantification by real-time PCR.
Total RNA was extracted with TRIzol reagent (Invitrogen) following the manufacturer's instructions. Real-time quantitative RT-PCR analysis was performed using a LightCycler (Roche) and SYBR RT-PCR kits (Takara). Primer sequences for the PRRSV protein coding genes and GADPH (internal control) are available from the corresponding author upon request. The relative expression level of the indicated genes was normalized to that of the GADPH internal control using the 2 −ΔΔCt cycle threshold method.
For miRNA analysis, small RNAs including miRNA were isolated using a mirVana™ miRNA Isolation Kit (Applied Biosystems). The expression of miR-26a was analyzed using specific primers and probes (Applied Biosystems). The relative expression levels of the miRNAs were normalized to that of the miR-18 internal control using the 2 −ΔΔCt cycle threshold method.
Dual luciferase reporter assays. Luciferase assays were performed using a dual-luciferase reporter assay system kit (Promega) according to the manufacturer's protocol. Collected cells were washed once with cold PBS, and passive lysis buffer (100 μ l) was then added to the cells. After 10 min, the supernatants were collected by centrifugation at 12,000 g for 30 s, and the relative luciferase expression values were analyzed using a Modulus single-tube multimode reader (Promega). The relative luciferase expression was calculated as the expression of Renilla luciferase (pRL-TK) divided by the expression of firefly luciferase (pGL3-control vector).
RNA-seq analysis. Following transfection with miRNA-26a or miR-Ctr, MARC-145 cells were mock infected or infected with PRRSV VR2332 at MOI = 2 and then harvested 24 hpi. Total cellular RNA was prepared according to the manufacture's protocol using an RNA Isolation Kit (Invitrogen). All of the RNA Integrity Number (RIN) values were > 7.0, and the 28S:18S rRNA ratios were > 1.9, as confirmed using an Agilent Bioanalyzer.
cDNA libraries were constructed from poly(A)-enriched RNA using Illumina kits and sequenced by 2*100 paired-end sequencing on an Illumina HiSeq 2000 instrument. The FASTQ read files for the four samples were used for further data analysis. Data for gene counts were obtained using the Mayo Clinic pipeline and Burrows-Wheeler Alignment (BWA). To analyze the Illumina reads, TopHat and Cufflinks were used to investigate the DEG profiles and changes in transcript abundance in the miR-26a-or miR-Ctr-transfected MARC-145 cells with or without PRRSV infection. Four files of transcriptome data from the miR-Ctr-, miR-Ctr/V-, miR-26a-and miR-26a/V-inoculated groups were aligned to the UCSC Rhesus Macaque genome build in preparation for differential expression analysis. The four files were processed through Cufflinks to assemble the aligned RNA-seq reads into transcripts and estimate the abundances in FPKM of the paired-end reads. The Cufflinks q-value was the false discovery rate (FDR)-adjusted p-value of the uncorrected test statistic; the q-value used in this study was 0.05. The significance status was "yes" when p > q after Benjamini-Hochberg correction for multiple-testing. Cuffmerge was then used to create a single transcript dataset from the multiple reconstructions. Two runs were then conducted using the miR-26a vs. miR-26a/V and the miR-Ctr vs. miR-Ctr/V datasets and the Cuffdiff program to test for differential expression and regulation among the two disease states. Gene annotation of all significant hits was then performed using a MySQL database matching to the Ensembl Sscrofa 9.56 reference genome currently supported by the Integrative Genomics Viewer (Broad Institute). Finally, heat-maps were analyzed using Cluster and TreeView-1.1.6r2-win software.
Statistical analyses. Statistically significant differences between experimental groups were determined using Student's t-test and analysis of variance (ANOVA) with the GraphPad Prism software package (GraphPad Software Inc., La Jolla, CA, USA). P-values < 0.05 were considered statistically significant.