Porcine reproductive and respiratory syndrome virus WUH3 strain (GenBank accession number: HM853673.2), isolated from the brain of pigs suffering from “high fever” syndrome in China at the end of 2006 and identified as a highly pathogenic North American type PRRSV (type 2), was amplified and titrated as described previously (21). UV-inactivated PRRSV was irradiated 30 cm under short-wave (254 nm) ultraviolet light (40 W) for 1 h. Loss of infectivity was confirmed by the inability of the UV light-exposed viruses to produce cytopathic effect on monolayers of MARC-145 cells. MARC-145 cells were grown and propagated in Dulbecco’s Modified Eagle medium (Invitrogen, Carlsbad, CA, USA) supplemented with 10% heat-inactivated fetal bovine serum, at 37°C in a humidified 5% CO₂ incubator. HEK293T cells were cultured and maintained in RPMI-1640 (Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum. Porcine primary pulmonary macrophages (PAMs) used in this study have been described previously (22).

Mouse monoclonal anti-hemagglutinin (anti-HA), anti-FLAG, and anti-β-actin antibodies were purchased from ABclonal Biotechnology. Anti-FLAG polyclonal antibody (Macgene, China), anti-DHX36 polyclonal antibody (Proteintech, China), anti-TRIF polyclonal antibody (Proteintech, China), anti-MyD88 polyclonal antibody (Proteintech, China), anti-HSP90 polyclonal antibody (Proteintech, China), anti-LaminA + C polyclonal antibody (Proteintech, China), anti-IκBα, anti-NF-κB P65, and anti-phosphor-p-P65 polyclonal antibodies (Cell Signaling) were purchased and used according to the manufacturers’ recommendations. Horseradish peroxidase-conjugated anti-mouse or anti-rabbit IgG antibodies were purchased from Beyotime Institute of Biotechnology (Jiangsu, China). PRRSV N protein monoclonal antibody was produced from hybridoma cells derived from Sp2/0 myeloma cells and spleen cells of BALB/c mice immunized with recombinant N protein from PRRSV strain WUH3.

The FLAG or HA epitope tag was amplified by PCR and cloned into the pCAGGS-MCS vector to generate the pCAGGS-FLAG or pCAGGS-HA plasmid, encoding an N-terminal FLAG or HA tag. Expression plasmids for FLAG-tagged DHX36 were constructed by PCR amplification of the cDNA from PAMs. Plasmids encoding truncated DHX36 were constructed by PCR amplification using the specific primers listed in Table S1 in Supplementary Material. The HA-tagged expression plasmid encoding the PRRSV N protein (pCAGGS-HA-N) used in this study, and its expression was confirmed with HA-tag immunoblotting (23). Reporter plasmids NF-κB reporter plasmid and pRL-TK were described elsewhere (21). All constructs were confirmed by DNA sequencing.

MARC-145 cells grown in 24-well plates were infected with PRRSV or mock-infected at an MOI of 1. Total RNA was isolated at the indicated time points using TRIzol reagent (Invitrogen). Real-time RT-PCR was performed using SYBR Green Real Time PCR Master Mix (Toyobo Biologics, Osaka, Japan) in a LightCycler 480 (Roche Molecular Biochemicals). Individual transcripts in each sample were assayed three times. The PCR conditions were as follows: initial denaturation for 10 min at 95°C, followed by 40 cycles of 15 s at 95°C, 15 s at 58°C, and 40 s at 72°C. The fold change in gene expression relative to normal was calculated using the delta delta cycles to threshold (ΔΔCT) method (24). Primers (Table S2 in Supplementary Material) were designed using the Primer Express software (version 3.0; Applied Biosystems, Carlsbad, CA, USA).

Transient transfection was carried out using Lipofectamine 2000 (Invitrogen). MARC-145 cells were seeded on 24-well plates at a density of 2–4 × 10⁵ cells/well and cultured until the cells reached approximately 70–80% confluence. Cells were then transfected with the indicated plasmids or small-interfering RNA (siRNA) in triplicate. For each transfection, 0.2 µg of the NF-κB reporter plasmid (purchased from Stratagene) along with 0.05 µg pRL-TK and 80 nM siRNA were used. Firefly and Renilla luciferase activities were determined using the dual-luciferase reporter assay system (Promega) according to the manufacturer’s instructions. siRNA sequences used are as follows: si-DHX36, sense 5′-GGAGCCGGAUUUGUAAGCAGUAGAA-3′, antisense 5′-UUCUACUGCUUACAAAUCCGGCUCC-3′; si-MyD88, sense 5′-CUGGUCCAUUGCUAGUGAATT-3′, antisense 5′-UUCACUAGCAAUGGACCAGTT-3′; si-TRIF, sense 5′-GACACCACCUCUCCAAAUATT-3′, antisense 5′-UAUUUGGAGAGGUGGUGUCTT-3′; negative control (NC) siRNA, sense 5′-UUCUCCGAACGUGUCACGUTT-3′, antisense 5′-ACGUGACACGUUCGGAGAATT-3′.

Cytoplasm and nuclear protein extracts from PRRSV-infected PAMs or HEK293T cells after transfection with N protein expression plasmid were prepared with the cytoplasmic and nuclear protein extraction kit (Aidlab Biotechnologies Co., Ltd.) according to the manufacturer’s protocols. Cells cultured in 60-mm dishes were prepared by adding 120 µL 2× lysis buffer A (65 mM Tris-HCl, pH 6.8, 4% sodium dodecyl sulfate (SDS), 3% dl-dithiothreitol, and 40% glycerol). The cell extracts were boiled for 10 min and then resolved with 8–12% SDS-PAGE. The separated proteins were electroblotted (Bio-Rad Transblot Cell System, USA) onto a polyvinylidenedifluoride (PVDF) membrane (Millipore, Billerica, MA, USA) and run for 3–5 h at 40 V on ice. The Western blot was probed with specific antibodies, and the expression of β-actin was detected with an anti-β-actin mouse monoclonal antibody to demonstrate equal protein sample loading.

To investigate the interactions between proteins, HEK293T cells or MARC-145 cells were lysed in immunoprecipitation lysis buffer (RIPA). After the lysate was clarified by centrifugation at 12,000 × g for 10 min, the lysate proteins were incubated overnight at 4°C with the indicated antibodies. Protein A + G agarose beads (30 µl; Beyotime) were then added to each immunoprecipitation reaction for another 6 h. The agarose beads were then washed three times and the captured proteins resolved on 8–12% SDS-PAGE, transferred to PVDF membranes, and analyzed by immunoblotting.

MARC-145 cells seeded on microscope coverslips and placed in 24-well dishes were infected with PRRSV (MOI = 0.5). At 24 hpi, the cells were fixed with 4% paraformaldehyde for 10 min and then permeated with 0.1% Triton X-100 for 10 min at room temperature. After three washes with PBS, the cells were sealed with PBS containing 5% bovine serum albumin for 1 h and then incubated separately with rabbit polyclonal antibody directed against DHX36 (1:200) and mouse monoclonal antibody directed against PRRSV N protein (1:200) for 1 h at 37°C. The cells were then treated with fluorescein-isothiocyanate-labeled goat anti-mouse or Cy3-labeled goat anti-rabbit antibodies (Invitrogen) for 1 h, followed by 4′,6-diamidino-2-phenylindole for 10 min at room temperature. After the samples were washed with PBS, fluorescent images were acquired with a confocal laser scanning microscope (Olympus Fluoview ver. 3.1, Japan).

The statistical analyses (Student’s t test) and generation of graphs were performed with the GraphPad Prism^® 6 software. Data are presented as mean ± standard deviation. *P < 0.05, **P < 0.01.

To better understand the role of DHX36 in PRRSV infection, we first examined the expression levels of endogenous DHX36 after PRRSV infection. MARC-145 cells were mock-infected or infected with PRRSV, and cells were collected at 12, 24, 36, and 48 h post-infection (hpi) for Western blotting analyses. As shown in Figure 1A, elevated endogenous DHX36 protein was first observed at 24 hpi in PRRSV-infected MARC-145 cells and achieved stronger intensity at 36 hpi, while stimulation with UV-inactivated PRRSV failed to alter DHX36 protein expression, indicating that the up-regulation of DHX36 depends on viral replication.

To substantiate that the expression of DHX36 is induced by PRRSV, we then tested the expression of DHX36 in PAMs. To this end, PAMs were mock-infected or infected with PRRSV or UV-inactivated PRRSV and collected at 12 and 36 hpi before subjected to Western blot analyses. As shown in Figure 1B, the expression of DHX36 was increased in PRRSV-infected PAMs. UV-inactivated PRRSV also failed to alter DHX36 protein expression. Collectively, above data support the basal presence of DHX36 in unstimulated cells and demonstrated that DHX36 is modulated by PRRSV infection.

The subcellular localization of DExD/H-box helicase protein in host cell is consistent with its function. DHX36 is a nucleocytoplasmic shuttling protein involved in a wide range of cellular functions both in the cytoplasm and nucleus (3). To explore the subcellular localization of DHX36 after PRRSV infection, MARC-145 cells were inoculated with PRRSV for 24 h, and indirect immunofluorescence assay was performed with a polyclonal antibody against DHX36. The PRRSV-infected cells were monitored by monoclonal antibody against viral N protein. As shown in Figure 2A, endogenous DHX36 protein was distributed diffusely in both cytoplasm and cell nuclei in mock-infected cells; in contrast, even though some of the DHX36 still resided in the nucleus, PRRSV infection resulted in the accumulation of DHX36 in the cytoplasm.

To further reinforce this observation, the nuclear and cytoplasmic fractions were separated from PAMs at 24 h after PRRSV infection, followed by Western blot analyses. As expected, HSP90 and Lamin A + C, the cytoplasmic and nuclear protein markers remained unchanged in their fractions, respectively, eliminating the possibility of fractions cross-contamination and unequal sample loading (Figure 2B). In accord with the results in MARC-145 cells, the increased DHX36 mainly located in cytoplasmic fraction in PRRSV-infected PAMs (Figure 2B). These results suggest a possibility that PRRSV infection might provide potential ligand, which redistributes DHX36 into the cytoplasm.

In cytoplasm, DHX36 plays important roles in sensing poly I:C and viral infection (3, 25). Considering that pigs infected with PRRSV may develop severe interstitial pneumonia, we focused on the relationship between DHX36- and PRRSV-induced pro-inflammatory NF-κB signaling activation. First, we utilized DHX36-specific siRNA to silence DHX36. The knockdown efficiency of the synthesized siRNA-targeting DHX36 (si-DHX36) was evaluated by real-time RT-PCR and Western blotting analyses. Compared with the cells treated with NC siRNA, cells transfected with si-DHX36 significantly decreased DHX36 expression at protein and mRNA levels (Figure 3A). Next, we assessed the impact of DHX36 silencing in response to PRRSV stimulation. Poly I:C was used as a positive control. As expected, when DHX36 was knocked down in MARC-145 cells, the poly I:C-induced activation of NF-κB-promoter-dependent luciferase activity was less pronounced (Figure 3B). Similarly, compared to DHX36-knockdown MARC-145 cells, cells transfected with NC siRNA induced high levels of NF-κB activation following stimulation with PRRSV, highlighting the importance of DHX36 for NF-κB signal activation in response to PRRSV infection (Figure 3B).

To further validate the function of DHX36 in PRRSV-induced activation of NF-κB, we again stimulated NF-κB promoter activity using PRRSV after DHX36 overexpression. Synthetic RNA duplex poly I:C was also included for comparison. Under the same conditions, overexpression of DHX36 significantly augmented the poly I:C- and PRRSV-induced activation of the NF-κB promoter as compared with empty vector-transfected cells (Figure 3C).

Two adaptor molecules, TRIF and MyD88, have been reported to be utilized by DHX36 to mediate downstream signaling (2, 3). To further identify which adaptor is utilized by PRRSV to induce NF-κB activation, we again used siRNA to knock down endogenous MyD88 or TRIF and confirmed the knockdown efficiency by real-time RT-PCR and Western blotting analyses (Figures 4A,B). As shown in Figure 4C, compared with NC siRNA, knockdown of MyD88 significantly reduced PRRSV-induced NF-κB activation, but no appreciable change was detected after knockdown of TRIF. We further validated this observation by Western blotting analysis of IκBα expression level and P65 phosphorylation. To this end, MARC-145 cells were transfected with DHX36-, MyD88-, and TRIF-specific siRNA or control scrambled siRNA, after which the cells were infected with PRRSV, and cell extracts were prepared and separated by SDS-PAGE, followed by immunoblotting. Consistent with the results of the dual-luciferase assay, Western blotting showed that IκBα protein degradation and P65 phosphorylation were strongly impaired in both DHX36- and MyD88-knockdown cells, whereas these parameters barely changed in TRIF-knockdown cells (Figure 4D). These results indicate that MyD88 is the key downstream adaptor of DHX36 responsible for PRRSV-induced NF-κB activation.

NF-κB is the key transcription factor for pro-inflammatory cytokine production following PRRSV infection (26–28). DHX36 is involved in PRRSV-induced NF-κB activation, thus theoretically, it should be associated with PRRSV-induced cytokine production. To test this, MARC-145 cells were transfected with DHX36-specific siRNA to knock down endogenous DHX36 expression, and then infected with PRRSV. At 36 hpi, cells were collected for real-time RT-PCR to analyze the expression of various cytokines. As shown in Figures 5A–D, silencing of endogenous DHX36 led to drastic reduction of IL-6 (Figure 5A), IL-8 (Figure 5B), TNF-α (Figure 5C), and RANTES (Figure 5D) mRNA expression in PRRSV-treated MARC-145 cells. Accordingly, knockdown of MyD88 also resulted in >40% loss in cytokine production (Figures 5A–D), indicating that DHX36 is critical for PRRSV-induced cytokine transcription and that DHX36-driven pro-inflammatory cytokine expression is dependent on MyD88.

Among PRRSV-encoded proteins, the N protein has been identified as the key NF-κB activator (23). To further investigate whether DHX36 is involved in N protein-induced NF-κB signal activation, pCAGGS-HA-N, pCAGGS-FLAG-DHX36, pRL-TK, and an NF-κB reporter plasmid were transiently co-transfected into MARC-145 cells. After 36 h transfection, a dual-luciferase assay was performed to detect NF-κB reporter activation. As shown in Figure 6A, N protein overexpression could initiate the transcription of an NF-κB promoter-driven luciferase construct and overexpression of DHX36 further augmented N protein’s ability to activate NF-κB. We also evaluated the involvement of DHX36 in N protein-induced NF-κB activation by knockdown of endogenous DHX36. As shown in Figure 6B, we observed that loss of DHX36 led to a significant decrease in the ability of N protein to activate luciferase under control of the NF-κB promoter.

To further confirm NF-κB activation, we examined the phosphorylation and nuclear translocation of NF-κB subunit P65 in cells co-expression of N protein and DHX36. As shown in Figure 6C, the amount of phosphorylated P65 and nuclear P65 protein increased, while total P65 was unaltered in cells co-expressing N protein and DHX36, compared to cells expressing only N protein. These results indicate that forced expression of N protein in MARC-145 cells significantly activates NF-κB signaling and this effect involves DHX36.

To test whether PRRSV N protein might function as a ligand of the intracellular sensor DHX36 to initiate NF-κB signaling activation in mammalian cells, we transfected HEK293T cells with FLAG-tagged DHX36 and HA-tagged N protein then examined their interaction. As shown in Figure 7A, DHX36 was detected in anti-HA co-immunoprecipitates from N protein co-transfectants but not from cells co-transfected with the control plasmid, indicating that FLAG-tagged DHX36 forms a complex with HA-tagged N protein in co-immunoprecipitation assays.

Full-length porcine DHX36 cDNA comprises 1012 amino acid residues, containing a helicase core domain and is flanked on either side by N- and C-terminal extensions (Figure 7B). To examine which domain of DHX36 might be involved in N protein binding, five mutants with deletion of different domains of DHX36 were constructed by mutagenesis (Figure 7B). HEK293T cells were co-transfected with various combinations of FLAG-tagged full length or deleted versions of DHX36 and HA-tagged N protein. As shown in Figure 7C, the CO-IP experiments indicated that mutants with deletion of the N-terminal domain of DHX36 were incapable of interacting with PRRSV N protein.

Finally, we examined whether N protein interacted with endogenous DHX36 after PRRSV infection. To address this, MARC-145 cells were mock infected or infected with PRRSV (MOI = 0.5) for 24 h before cells were harvested. The endogenous DHX36 was then precipitated, and immunoblotting was performed with an anti-N monoclonal antibody. As shown in Figure 7D, CO-IP analysis indicated that N protein associated with DHX36 in PRRSV-infected MARC-145 cells. Besides, cytoplasmic fractions contained both DHX36 and N protein in PRRSV-infected cells, which provided evidence of their colocalization (Figures 2A,B). Collectively, the above results indicate that DHX36 recognizes N protein and contributes to both PRRSV- and N protein-induced NF-κB pathway activation.