Porcine alveolar macrophage cells (PAMs) were collected from PRRSV-negative pigs (6-week-old) as previously reported (Li et al., 2017) and maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS). MARC-145 cells were purchased from China Center for Type Culture Collection (CCTCC, Wuhan, China). All cells were cultured in Dulbecco's minimal essential medium (DMEM) with 10% FBS (FBS, Gibco, Carlsbad, CA, USA). PK-15^CD163 cell lines were established and stored in our lab (Wang et al., 2013). The following isolates of PRRSV were selected for the cultures: PRRSV-2 isolates SD16 (GenBank: JX087437.1), CH-1a (GenBank: AY032626), JXA1 (GenBank: EF112445.1), GD-HD (GenBank: KP793736.1), and VR-2332 (GenBank: EF536003.1), and PRRSV-1 isolates GZ11-G1 (GenBank: KF001144.1) and P073-3 (only partially sequenced and confirmed as a PRRSV-1 isolate, and full sequence is unavailable). Monoclonal mouse antibodies against N protein of PRRSV (6D10), Mab2-5G2, and anti-pCD163 polyclonal antibodies were produced in our lab. Mouse anti-β-actin antibody, 4,6-diamidino−2-phenylindole (DAPI), and blebbistatin (a selective myosin II inhibitor) were purchased from Sigma-Aldrich (St. Louis, MO, USA). HRP-conjugated goat anti-mouse IgG antibodies and Texas Red-affiniPure goat anti-mouse IgG (H+L) were purchased from Jackson ImmunoReseach (West Grove, PA).

The porcine CD163 gene (GenBank ID: JX292263) was amplified and cloned into a pTRIP-CMV-Puro vector. Lentivirus production was performed as previously described (Du and Tikoo, 2010; Li et al., 2017). HEK-293T and BHK-21 cells were transduced with lentivirus pTRIP-CMV-CD163-IRES-puro and screened with puromycin according to the previously reported protocol (Gao et al., 2014; Li et al., 2017), and the new cell lines were termed as HEK-293T^CD163 and BHK-21^CD163, respectively. The susceptibility of HEK-293T^CD163 and BHK-21^CD163 cell lines to PRRSV was determined, and the cell lines thus obtained were referred to as PK-15^CD163 (Wang et al., 2013). Briefly, these cells were inoculated with PRRSV SD16 strain at 1 MOI, and then the cells and cellular supernatant were collected at 48 h post-infection (hpi). The PRRSV replication was determined by Western blot and TCID₅₀, and these cells were also assessed for their susceptibility to other strains of PRRSV (SD16, VR-2332, JXA1, and GD-HD) at 1 MOI using indirect immunofluorescent assay (IFA) at 48 hpi.

Total RNA was extracted from HEK-293T, BHK-21, and MARC-145 cells using TRIzol reagent (Thermo Fisher, USA), respectively. The C-terminal domain of MYH9 (aa 1651–1960) was designated as PRA, and the PRA fragments from other species (human and mouse) and truncated PRA were amplified (primers are listed in Table 1) and cloned into pCold-SUMO vector with the same enzymes SalI and XbaI (Haigene, China). All sequences were confirmed by sequencing analysis (Sangon Biotech, China). These recombinant expression plasmids were transformed into the E.coli strain BL21 (DE3), respectively, and the different PRA proteins were expressed and purified according to a previous study (Li et al., 2018).

PK-15^CD163, BHK-21^CD163, and HEK-293T^CD163 cells were cultured and pretreated with blebbistatin inhibitor at the indicated concentrations for 30 min, and then inoculated with PRRSV SD16 or VR-2332 strain at an MOI of 1, respectively. After removing the medium, the cells were cultured with a 3% FBS DMEM medium containing the same concentrations of blebbistatin. The N protein in mRNA and protein levels of PK-15^CD163, BHK-21^CD163, and HEK-293T^CD163 cells were analyzed by IFA at 36 hpi.

The PRA proteins from other species (human and mouse) and truncated PRA were diluted serially and incubated with PRRSV SD16 at 0.1 MOI. The mixture was maintained at 37°C for 1 h, then PAM cells were added to 6-well plates at 37°C and incubated for 1 h, and later washed with PBS to remove the unbound virus. The mRNA level of the PRRSV-2 type nucleocapsid (N) gene and expression level of protein were evaluated via qPCR and Western blot at 24 hpi. The mRNA level of PRRSV-1 type nucleocapsid (N) gene and the copies of the RNA genome were tested by qPCR according to the protocol described in a previous study (Li et al., 2018).

For antibody blocking assays, PAM and MARC-145 cells were pretreated with anti-MYH9^1676−1791 polyclonal antibody produced in our lab, which was serially diluted to indicated concentrations in RPMI-1640 or DMEM, or with control mouse serum for 1 h at 37°C. Then, the cells were infected with PRRSV without removing the antibodies. The infected cells were measured within 24 hpi by qPCR and IFA.

The cytotoxic effects of blebbistatin, mouse-PRA, human-PRA, monkey-PRA, and MYH9^1676−1791 were evaluated using the Cell Counting Kit-8 (CCK-8) assay (Beyotime, Nanjing, China). Briefly, the cells were cultured on 96-well plates (1 × 10⁵/well) and treated with 100 μl of appropriate medium containing different concentrations of blebbistatin, mouse-PRA, human-PRA, monkey-PRA, and MYH9^1676−1791 at 37°C for 24 h. Then the CCK-8 reagent (10 μl/well) was added and incubated for 2 h according to the manufacturer's instructions. Cell viability was measured at an absorbance of 450 nm, and the data were analyzed by GraphPad Prism.

The total RNA was extracted from PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 using TRIzol reagent (Takara, Japan) to detect the presence of the full-length CD163 gene (Table 1) by running the sample on 1% agarose gel.

The total RNA was extracted from PRRSV-infected cells with TRIzol Reagent. About 1 μg of the total RNA was converted to cDNA using PrimeScript RT Master Mix (Takara, Japan), and then amplified using FastStart Universal SYBR Green Master Mix following the manufacturer's protocol (Roche, Basle, Switzerland). Primer sequences for N protein of PRRSV and GADPH are listed in Table 2. The GADPH was used as an internal control. qPCR reactions were performed using a StepOnePlus® Real-Time PCR System (Applied Biosystems, Foster City, CA, USA), and each assay was run in triplicate. Relative quantification of target genes was calculated using the 2^−ΔΔCt method. Absolute quantification was performed using a template consisting of recombinant plasmid pMD18-T-N containing the PRRSV-1 or PRRSV-2 ORF7 gene as previously described, to quantify the RNA copy number of the PRRSV-1 or PRRSV-2 genome (Mu et al., 2015; Li et al., 2018).

The siRNAs (listed in Table 2) used for targeting MYH9 were obtained from Gene Pharma Co., Ltd. The PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 cells were transfected with these siRNAs according to the protocol of X-tremeGENE siRNA Transfection Reagent, respectively, and non-targeting siRNA served as a negative control. Total RNA of PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 was extracted using TRIzol reagent. The transcription levels of the PRRSV N gene and MYH9 mRNA were normalized against those of GAPDH by the 2^−ΔΔCT threshold cycle (CT) method, and the relative fold change of the PRRSV N gene and MYH9 mRNA were then calculated, respectively. The quantification was performed in StepOne Plus Real-Time PCR System (Applied Biosystems, USA). All the test samples were run in three independent experiments.

Five female BALB/c mice aged 6–8 weeks were injected subcutaneously with ~100 μg of MYH9^1676−1791 protein mixed with complete Freund's adjuvant in a volume of 100 μl. Then, these mice were given booster doses for three times (at 14, 28, and 42 days of initial exposure) subcutaneously with a mixture containing 100 μg of MYH9^1676−1791 protein mixed with incomplete Freund's adjuvant and the same amount of inoculum as used in the primary exposure. Finally, the serum of mice was collected and used to confirm the binding activity with MYH9^1676−1791 protein by IFA and Western blot.

The PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 cells were seeded on coverslips in 12-well tissue culture plates and infected with SD16 or VR-2332 (1 MOI). Cells were fixed with 75% cold ethanol at 36 or 48 hpi. Meanwhile, MARC-145 and PAM cells were plated into 12-well tissue culture plates with coverslips, respectively, and infected with SD16 (0.1 MOI). Then the cells were fixed at 24 hpi as mentioned earlier and blocked with 1% bovine serum albumin (BSA) in PBS. Cells were incubated with the primary antibody (anti-PRRSV N monoclonal antibody, 6D10) and later with secondary antibodies (rhodamine-conjugated goat anti-mouse IgG antibody) (Jackson, West Grove, USA), and were washed three times as described earlier. Finally, the cells were counterstained with DAPI, and cell staining was visualized using Leica microsystems (Leica AF6000, Germany).

The cells were collected and lysed with NP40 lysis buffer (Beyotime, China), and the protein concentrations were determined using a BCA protein assay kit (Thermo, USA). Western blot was performed as described previously (Mu et al., 2015). The membranes were blocked with 5% BSA and then incubated with indicated primary antibodies overnight at 4°C, followed by incubation with HRP-conjugated goat anti-mouse IgG (Jackson Laboratories, West Grove, PA, USA), which act as secondary antibodies. β-actin served as a control. Finally, the proteins were visualized using enhanced chemiluminescence (ECL) reagents (Amersham Biosciences).

Statistical analysis was performed using GraphPad Prism version 5.0 (GraphPad Software, San Diego, CA, USA). Differences among the groups were analyzed by one-way ANOVA followed by Bonferroni post-hoc test or unpaired Student's t-tests. Statistically significant and very significant results were defined as P < 0.05 and P < 0.01.

The HEK-293T and BHK-21 cell lines were infected with packed CD163 lentivirus according to previous studies to obtain the stable expression of pCD163 in the cell lines of human, murine, and swine species (Li et al., 2017), and the positive cell lines were obtained after 4 weeks with puromycin selection. PK-15^CD163 was generated as previously reported in our lab (Wang et al., 2013). To determine the pCD163 expression and PRRSV susceptibility, PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 cells were detected using RT-PCR, Western blot, TCID₅₀, and IFA assays. First, the specific RT-PCR products (3,333 bp) were detected in these recombinant cells, while no PCR products were observed in untransduced PK-15, BHK-21, and HEK-293T cells (Figure 1A). In addition, the pCD163 expression and virus replication in these cells were determined using an anti-pCD163 polyclonal antibody and anti-PRRSV N monoclonal antibody (6D10) by Western blot, and the results suggested that pCD163 mediates PRRSV SD16 infection in PK-15, HEK-293T, and BHK-21 cells (Figure 1B). The titer values of SD16 progeny virus were determined in PK-15^CD163, BHK-21^CD163, and HEK-293T^CD163 cells at 48 hpi (Figure 1C). We further determined whether stably expressing pCD163 cell lines were susceptible to PRRSV SD16, VR-2332, JXA1, and GD-HD infections using IFA. The specificity of CPE was determined using a 6D10 monoclonal antibody. The specific N protein staining was observed in PK-15^CD163, BHK-21^CD163 and HEK-293T^CD163 cells at 48 hpi, while uninfected cells showed negative staining (Figure 1D).

The next step was to determine whether decreased basal levels of MYH9 showed an impact on PRRSV SD16 replication in PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 cells Small interfering RNA (siRNA) duplexes against MYH9 were synthesized by Gene Pharma, and non-targeting siRNA was used as control. The valid siRNA and effective concentration were determined by qPCR (Supplementary Figure 1). The cells were seeded at 60% confluence in 24-well plates before siRNA transfection. Then the cells were transfected with siRNAs for 12 h, followed by infection with SD16 at 1 MOI for 48 h. The results showed that siRNA knockdown led to a >40–50% reduction in mRNA level and protein expression of MYH9, and a significant corresponding reduction in the amount of the ORF7 mRNA and N protein expression in PRRSV SD16-infected cells compared to the negative control (Figure 2).

Blebbistatin is a small molecule inhibitor of non-muscle myosin IIA (MYH9) and was shown to inhibit the MgATPase activity and in vitro motility of non-muscle myosin IIA (Limouze et al., 2004). We next investigated whether blebbistatin could reduce PRRSV infection in PK-15^CD163, HEK-293T^CD163, and BHK-21^CD163 cell lines. These cell lines were pre-inoculated with blebbistatin and HP-PRRSV SD16 strain or classical PRRSV VR-2332 at 1 MOI, respectively, to assess the hypothesis. The specific PRRSV anti-N protein staining was performed to observe PRRSV replication at 36 hpi by IFA, and the infection efficiency was found to be significantly decreased with blebbistatin treatment (Figure 3). Therefore, the results of our study suggest that blebbistatin can block the PRRSV infection in a dose-dependent manner in these cells. Cytotoxicity assays indicated that blebbistatin exhibited little cytotoxicity against these recombinant cell lines. As expected, these results indicated that the MYH9 gene derived from human and mouse species was involved in PRRSV infection similar to the mechanism observed in the swine (Supplementary Figure 2).

A previous study showed that the C-terminal end (aa 1651–1960) of MYH9 (PRA) from PAM cells could block the infection by different PRRSV strains via interaction with GP5 (Li et al., 2018). To assess whether mouse-PRA (mPRA) and human-PRA (hPRA) have a similar inhibitory effect in PRRSV SD16 infection as the swine PRA (sPRA) in vitro. MARC-145 cells served as PRRSV-susceptible cells, which were taken into consideration. The anti-PRRSV activity of the PRA domain obtained from a monkey species (monPRA) need to be tested. We obtained the recombinant PRA from different species using E.coli cells as previously reported (Li et al., 2018) (Supplementary Figure 3). PRRSV SD16 (0.1 MOI) was pre-incubated with indicated PRA protein of different concentrations at 37°C for 1 h, respectively, and PCV2-Cap served as control. Then the PRRSV-susceptible cells were incubated with a mixture of PRRSV and PRA proteins at 37°C for 1 h. Compared with PCV2-Cap protein treatment, PRRSV SD16 infection was blocked by pre-incubation with mPRA, hPRA, or monPRA in a dose-dependent manner, as shown by the reduction of PRRSV N protein mRNA and protein levels in MARC-145 cells (Figure 4A). The anti-PRRSV effect of indicated PRA was confirmed on PAM cells (Figure 4B), and the cytotoxicity assays indicated that PRA obtained from different species exhibited little cytotoxicity in MARC-145 and PAM cells, as expected (Supplementary Figure 4).

The above-mentioned results show that pre-incubation of PRRSV with PRA protein from other species could effectively inhibit PRRSV replication. Based on our previous study, the recombinant swine PRA blocks the internalization of PRRSV by direct interaction with viral GP5 (Li et al., 2018). To further identify the key region of PRA that is required for interaction with PRRSV GP5 protein, we first analyzed the sequence alignment of the amino acids constituting the PRA domain of different species using Lasergene software (Supplementary Figure 5). Based on the sequence analysis, 11 truncated fragments of swine PRA were designed (Figure 5A), and the recombinant truncated proteins without SUMO tag were obtained as previously reported (Supplementary Figure 6). Furthermore, the antiviral activity of these proteins (2.5 μM) was evaluated by pre-incubating MARC-145 cells with PRRSV SD16 (0.1 MOI) as mentioned earlier. The virus blocking assay showed that the MYH9^1676−1791 domain could reduce PRRSV infection (Figures 5B–F), but not by the truncated fragment MYH9^{Δ1676−1791} and control PCV2 Cap protein, which suggested that the amino acid residues (1676–1791) are responsible for the interaction of MYH9 with PRRSV GP5.

Porcine alveolar macrophage cells are the target cells of PRRSV infection in vivo. Based on the cytotoxicity assay (Supplementary Figure 7), we further evaluated the antiviral activity of MYH9^1676−1791 protein against PAMs. As similarly inhibitor effect in MARC-145 cells. PAMs were infected with PRRSV (MOI = 0.1) pretreated with MYH9^1676−1791 protein at various concentrations, which led to dose-dependent inhibition of viral replication at mRNA level (Figure 6A) and protein level (Figure 6B).

Since the sequence alignment of aa 1676-1791 of MYH9 obtained from different species remains conservative, it could block PRRSV infection in MARC-145 and PAM cells. Next, we investigated whether the MYH9^1676−1791 fragment could inhibit other strains of PRRSV in vitro. Therefore, PRRSV-2 strains CH-1a, GD-HD, JXA1, and VR-2332 were pre-incubated with 5 μM of MYH9^1676−1791 at 37°C for 1 h as described earlier. MARC-145 or PAM cells were then infected with the mixture containing MYH9^1676−1791 and PRRSV. Replication of these PRRSV isolates was determined based on the PRRSV N mRNA and protein levels. The result showed that the replication of PRRSV-2 strains was significantly reduced by pre-incubation with MYH9^1676−1791 (Figures 7A,B). In addition, we tested the inhibitory effect of MYH9^1676−1791 on PRRSV-1 isolate GZ11-G1 and P073-3, and the results showed that virus replication levels (Figures 8A,B) and ORF7 gene copies (Figures 8C,D) were significantly decreased on pre-incubation with 5 μM of MYH9^1676−1791.

To test whether the anti-MYH9^1676−1791 serum could block PRRSV infection in susceptible cells, we first proposed that the MYH9 derived from MARC-145 or PAM cells could be recognized by anti-MYH9^1676−1791 serum (Supplementary Figure 8). PAM and MARC-145 cells for antibody blocking assays were pretreated with a specific polyclonal antibody of MYH9^1676−1791, respectively, and then serially diluted to different concentrations in the indicated cell culture medium or with control mouse IgG for 1 h at 37°C. Cells were then infected with PRRSV without removing antibodies for 1 h at 37°C. The PRRSV infection was measured by qPCR at 24 hpi, and the result showed that the polyclonal antibody of MYH9^1676−1791 significantly reduced PRRSV infection (Figures 9A,B). An antibody blocking assay was determined by IFA in these cells with the effective antibody concentration. The PRRSV N protein staining was significantly reduced on treatment with the polyclonal antibody of MYH9^1676−1791 (Figure 9C). These results collectively indicate that aa 1676–1791 of MYH9 plays an important role in PRRSV entry.