Interferon Inducible Porcine 2′, 5′-oligoadenylate Synthetase-Like Protein Limits Porcine Reproductive and Respiratory Syndrome Virus Infection via the MDA5-Mediated Interferon- Signaling Pathway


 Porcine reproductive and respiratory syndrome virus (PRRSV) is a constant threat to the swine industry worldwide. Currently approved vaccines against PRRSV are losing effectiveness, as new viral strains are often refractory to conventional treatments. Thus, there is an urgent need to find new therapeutic targets to develop novel antiviral drugs. 2′, 5′-oligoadenylate synthetase-like (OASL) protein has antiviral activity, but this has not been demonstrated for PRRSV and the mechanism is not well elucidated. In this study expression of porcine OASL (pOASL) in porcine alveolar macrophages (PAMs) induced by interferon (IFN) -β stimulation and PRRSV infection was examined by real-time polymerase chain reaction (RT-PCR). Exogenous expression and knockdown of pOASL were used to indicate the role of pOASL in the PRRSV replication cycle. The type I IFN signaling pathway was evaluated after pOASL overexpression. Results showed the expression of pOASL in PAMs was significantly increased by IFN-β stimulation or PRRSV infection. pOASL specific small interfering RNA (siRNA) promoted PRRSV replication, whereas exogenous expression of pOASL inhibited infection of PRRSV. The anti-PRRSV activity was lost after knockdown of the Melanoma differentiation-associated protein 5 (MDA5) RNA sensor. Taken together, pOASL inhibits PRRSV infection via the activation of MDA5.

PRRSV strains are mainly subdivided into type I and type II according to their antigenicity, where the majority of China's epidemic strains are type II [7]. Current vaccines provide limited protection against PRRSV, nding new ways to control PRRSV is imperative. The host intrinsic restriction factors usually inhibit virus infection by direct interaction with viral proteins, and they are more promising because host intrinsic restriction factors are less likely to mutant under drug-mediated selective pressure [8,9].
OASL has been shown to inhibit replication of several viruses. For example, Newcastle disease virus replication in goose embryo broblasts is reduced signi cantly by overexpression of the goose OASL [26].
Human OASL has been shown to inhibit some speci c DNA and RNA viruses, such as respiratory syncytial virus, vesicular stomatitis virus, dengue virus, and herpes simplex virus-1 [27,28]. Nevertheless, human OASL does not protect against encephalomyocarditis virus infection [29]. Murine OASL2 strongly inhibits respiratory syncytial virus replication [27], whereas murine OASL1 fails to do so [27]. Instead, murine OASL1 inhibits the production of type I IFN, and OASL1 -/mice are more resistant to infection with encephalomyocarditis virus and herpes simplex virus-1 [30]. Chicken OASL was found to inhibit West Nile virus infection [31]. pOASL has been reported to inhibit the Japanese encephalitis virus infection in PK15 cells, this inhibition is not dependent on the OAS-RNase L pathway [32].
The anti-PRRSV effects of porcine OAS1 (pOAS1) and porcine OAS2 (pOAS2) have been demonstrated [35,36], but those of pOASL and the relationship between pOASL and IFN are not clear. Moreover, pOASL has a different sequence at its C terminus; it remains to be determined whether this feature has a different inhibitory effect on viral replication as compared with other OAS subtypes. Therefore, we evaluated the effect of pOASL on PRRSV replication in vitro and attempted to elucidate the mechanisms underlying its antiviral activity.

Cells and Viruses
Porcine alveolar macrophages (PAMs), isolated from lung lavage samples of seven-week old pigs which were free of PRRSV, pseudorabies virus, porcine circovirus type 2, and classical swine fever virus, were

Small Interfering RNA (siRNA) Synthesis
SiRNAs were used to identify the genes or proteins involved in the antiviral mechanism of pOASL. The nontargeting control siRNA (si-NC), OASL siRNA (si-OASL), RIG-I siRNA (si-RIG-I), RNase L siRNA (si-RNase L), and melanoma differentiation-associated protein 5 (MDA5) siRNA (si-MDA5) were all ordered from GenePharma Co., Ltd. (Suzhou, China). The siRNA sequences were listed in Table 1.   Table 1. The glyceraldehyde-3phosphate dehydrogenase (GAPDH) gene was analyzed as an internal control, and relative changes in the expression of the target genes were calculated by the 2 -△△Ct method [38].

Cell Survival Experiments
The toxicity of pOASL and various siRNAs toward PAMs and CRL-2843-CD163 cells was tested with the Enhanced Cell Counting Kit-8 (Solarbio, Beijing, China).

Virus Titers
Marc-145 cells were used to determine the PRRSV titers in the supernatants. PRRSV titers were expressed as TCID 50 .

Statistical Analyses
All experiments were repeated three times, data were analyzed by Student's t-test. Differences were considered statistically signi cant when values of p < 0.05.
The sample size was su cient for the data analysis using paired two-tailed Student's t-test. For all Statistical analyses, the differences were considered to be statistically signi cant at values of p < 0.05.

pOASL Expression is increased by IFN-β Stimulation and PRRSV Infection
After 6 h of stimulation with 1,000 IU/mL IFN-β, pOASL mRNA expression level in PAMs increased quickly to a peak of 125 times than that in the untreated control cells (Fig. 1A). The pOASL mRNA expression level peak occurred at 12 h post stimulation. This protein level was also tested, showing increased pOASL protein levels (Fig. 1B). This suggests that pOASL is an interferon-stimulated gene (ISG).
The OASL mRNA expression level in the PRRSV-infected PAMs peaked at 36 h post-infection (Fig. 1C), showing a 15-fold increase compared to the untreated cells. This protein level was also tested, showing increased pOASL protein levels (Fig. 1D). This suggests PRRSV infection increases the pOASL expression.

pOASL siRNA Enhances PRRSV Replication
The pOASL siRNA (si-OASL) transfection e ciently reduced the expression of pOASL compared with scrambled siRNA without affecting cell viability ( Fig. 3A to C). After 60 nM si-OASL was transfected into CRL-2843-CD163, PRRSV infected the cells for 24 h, and the results showed that in the presence of si-OASL, the PRRSV genomic mRNA levels were higher than that in cells transfected with si-NC (Fig. 3D).
The TCID 50 results are in line with the mRNA level results (Fig. 3E). This suggests pOASL siRNA enhances PRRSV infection.

Anti-PRRSV Activity is not dependent on RNase L
The si-RNase L (60 nM) was transfected into CRL-2843-CD163 for 48 h, resulting in e cient reduction of RNase L expression ( Fig. 4A and 4B). The pCMV-3xFLAG-7.1-OASL expression plasmid (800 ng) and 60 nM si-RNase L were co-transfected into CRL-2843-CD163 cells, 24 h later, 1.0 MOI PRRSV infected the CRL-2843-CD163 cells for 24 h. In the 800 ng pCMV-3xFLAG-7.1-OASL expression plasmid and 60 nM si-RNase L co-transfected and PRRSV (MOI 1.0)-infected CRL-2843-CD163 cells, there were still signi cant decreases in both the PRRSV genomic mRNA level and viral titers relative to the control group ( Fig. 4C  and 4D). This suggests that anti-PRRSV activity is not dependent on RNase L.

pOASL Increases IFN Responses
Results above have showed that pOASL did not inhibit PRRSV replication via the classical RNase L pathway, so whether anti-PRRSV activity of pOASL is dependent on other pathways needs to further investigated. There are reports revealing that some interferon-stimulated genes (ISGs) have antiviral effects via different mechanisms [43,44]. To investigate the mechanisms, dual-luciferase reporter assays were conducted .The results showed that reporter activities of IFN-β (Fig. 5A), ISRE (Fig. 5B), and NF-κB ( Fig. 5C) were signi cantly increased, indicating that IFN-β, ISRE, and NF-κB pathway were enhanced by pOASL.

IFN Pathway was activated by pOASL
Report revealed that human OASL interacts with human RIG-I and increases IFN signaling pathway. Our results also show that pOASL enhances type I IFN responses. So we speculate that pOASL act its role via RIG-I or MDA5 RNA sensor. Then co-IP assay was carried out to investigate the interaction. In this regard, Flag-tagged pOASL interacted with porcine MDA5 (pMDA5), but not with porcine RIG-I (pRIG-I) (Fig.  6A).Moreover, the RNase A treat the cell lysates, co-IP results found the interaction between them was independent of RNA ( Fig. 6B). Based on the results above, pOASL interacts with pMDA5. Then we speculated pMDA5 pathway mediates the function of pOASL, next mRNA level of IFN-β, myxovirus resistance protein 1 (Mx1) and interferon-stimulated gene 15 (ISG15) in CRL-2843-CD163 cells, which pOASL and pMDA5 were co-expressed, were tested by qRT-PCR. These data indicated that co-expression increased the mRNA levels of IFN-β, Mx1 and ISG15 (Fig. 6C to E). The data suggest that pMDA5mediated IFN pathway was enhanced by pOASL.

Anti-PRRSV Activity is dependent on pMDA5
Si-RIG-I (60 nM) was transfected into CRL-2843-CD163 for 48 h, resulting in e cient reduction of pRIG-I expression (Fig. 7A, B). By contrast, in the 800 ng pCMV-3xFLAG-7.1-OASL expression plasmid and 60 nM si-RIG-I co-transfected and PRRSV (MOI 1.0)-infected CRL-2843-CD163 cells, there were still signi cant decreases in either the PRRSV genomic mRNA expression level or viral titers relative to the levels in the control group (Fig. 7C, D). This suggests that anti-PRRSV activity is not dependent on pRIG-I.
For the case of pMDA5, 60 nM si-MDA5 was transfected into CRL-2843-CD163 cells for 48 h, resulting in e cient reduction of pMDA5 expression (Fig. 7E, F). In the CRL-2843-CD163 cells which co-transfected with 800 ng pCMV-3xFLAG-7.1-OASL and 60 nM si-MDA5 and then infected with PRRSV (MOI 1.0), there were no decreases in both the PRRSV genomic mRNA expression level and viral titers relative to the control group (Fig. 7G, H). This suggests that anti-PRRSV activity is dependent on pMDA5.

Discussion
The pattern recognition receptors (PRRs) are the rst line to defense against invading microorganisms in the  [8,51]. Here, we report that pOASL participates in MDA5-mediated IFN signal pathway. The schematic representation of the signaling pathway is presented in Fig. 8.
Reports have revealed that human OASL interacts with RIG-I and exerts an antiviral effect. Even though it has no enzymatic activity, OASL is usually maintained at low expression levels in cells. When viruses infect human cells, human OASL is notably upregulated by the double-stranded RNA, and IFN [29,52,53].
In the present study, after PRRSV infect the cells, pOASL was induced, whereupon viral replication was inhibited. In stark contrast to our results, Lee demonstrated that murine OASL1 downregulates IFN via IRF7 to impede its expression and therefore aids in viral replication [30]. The discrepancies could be explained that different OASL isoforms might have different regulatory mechanisms in the signaling pathway.
Since pOASL has a nucleotidyltransferase region (data not shown), we surmised that its antiviral activity was dependent on RNase L. Nevertheless, our results indicated that this was not the case; pOASL did not exert its action via the OAS-RNase L pathway, and there may be another critical factor in uencing the antiviral effect. Similarly, one report revealed that pOASL also inhibits Japanese encephalitis virus replication but not through the OAS-RNase L signaling pathway [32]. Thus, our nding for PRRSV is the same as that for Japanese encephalitis virus. Another report showed that pOASL could inhibit replication of classical swine fever virus through the MDA5-dependent pathway [54], and in our study, pOASL also inhibit PRRSV via this pathway.
In line with other studies that have shown the inhibition of PRRSV replication by pOAS1 and pOAS2, our study proves that pOASL inhibits PRRSV replication as well. Besides, pOAS1 and pOAS2 inhibit the replication of Japanese encephalitis virus, whereas pOASL inhibits the replication of classical swine fever virus, thus con rming the antiviral effects of the OAS protein family.
On the other hand, the inhibition of PRRSV replication by pOAS2 is dependent on RNase L [36], whereas inhibition of PRRSV replication by pOASL is not. This phenomenon may be related to the structures difference between pOAS2 and pOASL, which need further experimental veri cation.
A limitation of this study is that testing of other virulent strains was not done. The NADC-30 strains responsible for the most recent epidemics are presumed to follow the same trends. Therefore, future studies should include these viruses. It was not de ned which step pOASL targets in PRRSV replication, we assume that any step could be targeted for inhibition, this also need further investigation in the future.
In conclusion, we demonstrate that pOASL is a new restriction factor which dampens PRRSV infection via MDA5-mediated type I IFN signaling.
Upregulation of pOASL activity boosts host immunity to limit PRRSV infection. Knockout of pOASL increases the PRRSV titer during the virus production. Future investigation of pOASL activity might provide the insight and opportunities needed for the therapeutic developments and improved vaccine candidates.

Conclusions
Porcine OASL inhibits PRRSV replication in vitro through an MDA5-dependent signaling. This may point to future directions regarding new ways to target PRRSV. Further research regarding the regulation of pOASL may provide insight and new antiviral strategies for therapeutic developments.     and expressed as TCID50/mL. All experiments were biologically repeated three times, data represent means ± standard deviations. *P < 0.05; **P < 0.01. Figure 5 pOASL increases IFN responses. HEK293T cells were seeded in 24-well plate; on day one, 200 ng of pIFNβ-Luc (A), ISRE-Luc (B), or NF-κB-Luc (C), 20 ng of pRL-TK, and either the pCMV-3xFLAG-7.1 or pCMV-3xFLAG-7.1-OASL (400 ng) was transfected into the cells for 24h; On day two, cells were treated with 1.5 µg of poly (I: C) for 9 h, the promoter luciferase activity was tested. All experiments were biologically repeated three times, data represent means ± standard deviations. *P < 0.05.