The nsp2 Hypervariable Region of Porcine Reproductive and Respiratory Syndrome Virus Strain JXwn06 Is Associated with Viral Cellular Tropism to Primary Porcine Alveolar Macrophages

The PRRSV nsp2 replicase protein undergoes rapid and broad genetic variations in its middle region in the field, but the underlying significance has remained enigmatic. Here, we demonstrate that the nsp2 hypervariable region not only plays an important regulatory role in maintaining the balance of different viral mRNA species but also regulates PRRSV tropism to primary PAMs. Our results reveal novel functions for PRRSV nsp2 and have important implications for understanding the mechanisms of PRRSV RNA synthesis and cellular tropism.

ABSTRACT

Porcine reproductive and respiratory syndrome virus (PRRSV) poses a major threat to global pork production and has been notorious for its rapid genetic evolution in the field. The nonstructural protein 2 (nsp2) replicase protein represents the fastest evolving region of PRRSV, but the underlying biological significance has remained poorly understood. By deletion mutagenesis, we discovered that the nsp2 hypervariable region plays an important role in controlling the balance of genomic mRNA and a subset of subgenomic mRNAs. More significantly, we revealed an unexpected link of the nsp2 hypervariable region to viral tropism. Specifically, a mutant of the Chinese highly pathogenic PRRSV strain JXwn06 carrying a deletion spanning nsp2 amino acids 323 to 521 (nsp2Δ323–521) in its hypervariable region was found to lose infectivity in primary porcine alveolar macrophages (PAMs), although it could replicate relatively efficiently in the supporting cell line MARC-145. Consequently, this mutant failed to establish an infection in piglets. Further dissection of the viral life cycle revealed that the mutant had a defect (or defects) lying in the steps between virus penetration and negative-stranded RNA synthesis. Taken together, our results reveal novel functions of nsp2 in the PRRSV life cycle and provide important insights into the mechanisms of PRRSV RNA synthesis and cellular tropism.

IMPORTANCE The PRRSV nsp2 replicase protein undergoes rapid and broad genetic variations in its middle region in the field, but the underlying significance has remained enigmatic. Here, we demonstrate that the nsp2 hypervariable region not only plays an important regulatory role in maintaining the balance of different viral mRNA species but also regulates PRRSV tropism to primary PAMs. Our results reveal novel functions for PRRSV nsp2 and have important implications for understanding the mechanisms of PRRSV RNA synthesis and cellular tropism.

INTRODUCTION

The order Nidovirales represents a unique class of positive-stranded RNA viruses (e.g., coronaviruses and arteriviruses) that infect a wide variety of hosts, ranging from invertebrates to humans, and has a huge social and economic impact on our society (1, 2). The family Arteriviridae within this order contains several important veterinary pathogens, including porcine reproductive and respiratory syndrome virus (PRRSV) and equine arteritis virus (EAV) (1, 3). EAV is the prototype of arteriviruses, while PRRSV is the agent that has the most far-reaching impact on global pork production (4, 5). PRRSV mainly causes reproductive failure in sows and respiratory disease in young pigs, and it has been troubling the worldwide swine industry for the past 30 years, leading to staggering economic losses (4–6). It is estimated to cost U.S. pork producers 500 to 600 million dollars per year, with even greater costs in Asia (7–11).

Rapid and broad genetic variations of PRRSV have led to the frequent emergence of many pathogenic strains, including the Chinese highly pathogenic PRRSV (HP-PRRSV), which has been catastrophic to the Asian swine industry since its first outbreak in 2006 (6, 12, 13). The fastest evolving region has been mapped to PRRSV nonstructural protein 2 (nsp2), a replicase protein that has a size of 1,166 amino acids (aa) in the case of HP-PRRSV strain JXwn06 (6, 12, 13). It is a multidomain transmembrane protein that includes an N-terminal papain-like cysteine protease domain (PLP2), a functionally unknown large middle region (300 to 500 aa), a C-terminal transmembrane domain (TMD), and a cytoplasmic tail (CT) (14, 15). In addition, isoforms exist for PRRSV nsp2 that differ mainly in the C terminus, two of which (nsp2TF and nsp2N) are translated by a frameshift mechanism (16, 17). The PLP2 domain contains catalytic sites that are highly conserved among arteriviruses; it possesses both cis- and trans-cleavage activities, which are responsible for nsp2/3 cleavage, and deubiquitinating activity, which likely plays an important role in regulating host innate immunity (14, 18–23). The C-terminal TMD and CT are highly conserved and are required for PRRSV replication (15). The nsp2 middle region is highly heterogeneous and is responsible for size variation among PRRSV strains (24–26). Insertions and most notably deletions, as well as extensive amino acid substitutions, are mostly seen within this region (12, 15, 27–32).

Why does PRRSV nsp2 undergo broad genetic variations in the field? In this study, we initially proposed that the variations might contribute to viral fitness by regulating viral mRNA synthesis, considering that it is a critical component of the viral replication and transcription complex (RTC). Our results from deletion mutagenesis support this notion. Unexpectedly, the mutagenesis study also revealed a link of the nsp2 hypervariable region to PRRSV tropism for primary porcine alveolar macrophages (PAMs). Further studies suggest that nsp2 plays a critical role in the early stages of the PRRSV life cycle.

RESULTS

Identification of nonessential regions within nsp2 of HP-PRRSV strain JXwn06 for replication in MARC-145 cells.

The nsp2 middle region (aa 323 to 800) is highly heterogeneous among PRRSV strains (24–26). We hypothesized that this region serves to regulate the synthesis of both genomic mRNA (gmRNA) and subgenomic mRNAs (sgmRNAs). This idea was tested by deletion mutagenesis in the background of Chinese HP-PRRSV strain JXwn06 (13). To facilitate the mutagenesis process, we modified our original infectious cDNA clone of JXwn06 by putting the full-length genome under the control of the cytomegalovirus (CMV) promoter to bypass the in vitro transcription step for virus rescue (Fig. 1A). With this new tool, a panel of nsp2 deletion mutants were constructed; their viral viability is summarized in Fig. 1B. For each mutant, we used 3 independent clones for virus recovery. Two of the largest deletion mutants that could be rescued in MARC-145 cells were nsp2Δ323–521 and nsp2Δ520–782, which had deletion sizes of 199 aa and 263 aa, respectively. Interestingly, although the nsp2 region of aa 323 to 521 could be deleted as a whole, small deletions within this region (e.g., nsp2Δ323–480 or nsp2Δ433–521) were lethal to the virus, suggesting that this region likely plays a structurally regulatory role.

FIG 1.

Identification of nsp2 nonessential regions for replication of HP-PRRSV strain JXwn06 in MARC-145 cells. (A) Depiction of a DNA-launched system for HP-PRRSV strain JXwn06 under control of the CMV promoter. (B) Deletion mutagenesis of PRRSV nsp2. The deleted regions are shown with dashed lines, with the relative positions indicated in the names of individual mutants. The protease domain (PLP2) and transmembrane (TM) region are shown as black boxes. The viability of each mutant is shown on the right. +, viable; −, nonviable. (C) Multistep growth analysis of PRRSV nsp2 mutants. MARC-145 cells in 24-well plates were infected with the viruses at a MOI of 0.01. At each indicated time point, the total viruses were titrated with the TCID₅₀ assay.

Deletion of nsp2 aa 323 to 521 leads to differential accumulation of viral mRNA species.

We next investigated the mutational effects on viral replication and mRNA synthesis in MARC-145 cells. The viable mutants carrying small deletions (e.g., nsp2Δ522–601 and nsp2Δ602–695) did not exhibit significant growth defects, but large deletions (nsp2Δ323–521 and nsp2Δ520–782) led to severe defects in viral growth, with titers decreasing by ∼1.5 log units (Fig. 1C). In particular, deletion of nsp2 aa 323 to 521 had the greatest impact on PRRSV growth (Fig. 1C). For expression profiling of individual PRRSV mRNA species, we performed Northern blot analysis with a probe targeting open reading frame 7 (ORF7) (Fig. 2A). The investigation focused on the two most defective mutants, namely, nsp2Δ323–521 and nsp2Δ520–782. The mutant nsp2Δ323–450 served as a control, because the deletion region overlapped that of the mutant nsp2Δ323–521, and the infectious-clone-derived wild-type (WT) virus was used as the parental control. Because of the difference in the abundance of sgmRNAs and gmRNA, a given loading amount can lead to overexposure of certain RNA species (e.g., sgmRNA7) but weak exposure of other species (e.g., gmRNA and sgmRNA2 to sgmRNA4). Therefore, we replicated the virus with different loading amounts, to reduce the inaccuracy in statistics. The respective RNA level was first analyzed by the corresponding band density normalized to that of gmRNA (sgmRNA/gmRNA), and the relative abundance of individual sgmRNA species was then calculated as the ratio of the value for the mutant virus to that for the WT virus (Fig. 2B). Among the mutants tested, the virus nsp2Δ323–521 showed the most striking phenotype and exhibited dramatically reduced abundance of a subset of sgmRNAs (e.g., sgmRNA5) (Fig. 2A, lanes 1 and 2, and Fig. 2B), compared to the parental virus (Fig. 2A, lanes 3 and 4). Thus, the mutational effects on the accumulation of mRNA species were not equal; the synthesis of sgmRNA5 was greatly affected and the next was sgmRNA6. Moreover, the relative abundances of sgmRNA5 and sgmRNA6 appear to be reversed (Fig. 2A, lanes 3 and 4). In contrast, we did not observe a significant impact on the levels of sgmRNA2 to sgmRNA4 (Fig. 2A and B). For the mutants nsp2Δ323–450 (Fig. 2A, lanes 5 and 6) and nsp2Δ520–782 (Fig. 2A, lanes 7 to 10), the deletions led to alterations of all sgmRNAs but to different extents, when normalized to the level of gmRNA. To confirm the results described above, we also measured the relative abundance of gmRNA and all sgmRNAs by quantitative PCR (qPCR) with specific primers targeting the leader-body junctions. The results revealed a trend similar to that observed with Northern blot analysis (Fig. 2C).

FIG 2.

Regulation of the balance of gmRNA and sgmRNA species by the nsp2 hypervariable region in MARC-145 cells. (A) Northern blot analysis of PRRSV RNA species. MARC-145 cells were infected with WT virus or the indicated nsp2 deletion mutants at a MOI of 0.1. At 48 hpi, total intracellular RNAs were extracted and analyzed by Northern blotting using a biotin-labeled oligonucleotide complementary to ORF7. For each virus, the viral RNAs were loaded in at least two replicates, with different loading amounts. The lane numbers (lanes 1 to 10) are indicated. (B) Quantitative analysis of sgmRNA levels relative to the gmRNA level. The levels of individual sgmRNAs were normalized to that of gmRNA of the same virus (sgmRNA/gmRNA) in Northern blot analysis by measuring the corresponding band densities, and the relative abundance was then calculated as the mutant virus/WT virus ratio. (C) Relative abundance of sgmRNA species by qPCR. Left, specificity of PCR by gel electrophoresis, with the cellular gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as a control for PCR. Right, relative mRNA accumulation. The levels of individual sgmRNAs were normalized to that of gmRNA of the same virus, and the relative abundance was then calculated as the mutant virus/WT virus ratio. (D) Western blot analysis of viral protein abundance in infected MARC-145 cells and extracellular virions. (E) Intracellular GP5, M, and N levels, relative to GP3 levels. ***, P＜0.001; **, P＜0.01; NS, no significant difference. The error bars indicate standard deviations. The data analyses represent three independent experiments.

Since the mutant nsp2Δ323–521 exhibited a particular phenotype, we focused on this virus and examined the protein expression of glycoprotein 5 (GP5) in particular. We used GP3 as a loading control, because the sgmRNA3 expression was not altered (Fig. 2B and C), and then normalized the GP5 level to the GP3 level (Fig. 2D and E). Consistent with findings at the mRNA level, the expression of GP5 was reduced by about 80% (Fig. 2D, left, and Fig. 2E), as was that of M and N proteins, to a lesser extent (Fig. 2E). The reduced intracellular expression of these major structural proteins might partly account for the reduced virus production in MARC-145 cells. Together, these results provide good genetic evidence regarding the involvement of PRRSV nsp2 in viral mRNA synthesis, and they suggest that the nsp2 middle region contributes to fine tuning of the balance of viral mRNA species.

Deletion of nsp2 aa 323 to 521 abolishes HP-PRRSV tropism to primary PAMs.

The nsp2 mutants were further investigated in primary PAMs, the major target cells in vivo. To our surprise, one of the mutants (nsp2Δ323–521) did not grow (Fig. 3A). This result was subsequently confirmed by multistep growth analysis (Fig. 3B) with a multiplicity of infection (MOI) of 0.01; the mutant nsp2Δ520–782 was used as a control. Deletion of nsp2 aa 520 to 782 slightly impaired the viral growth (Fig. 3B). In contrast, the mutant nsp2Δ323–521 did not grow. The virus titer decreased gradually, and no infectious virions could be detected by the endpoint dilution assay at about 48 h postinfection (hpi). Moreover, this defect was dose independent, as increasing the infective dose (MOIs of 0.1 and 1) did not lead to rescue of the infectivity (Fig. 3B). We also extended the absorption time to 8 h to allow virus attachment and penetration before removal of the inoculum (Fig. 3C). Again, we failed to rescue the infectivity. Thus, these results indicate that PRRSV nsp2 is a critical determinant of viral tropism to primary PAMs and that deletion of nsp2 aa 323 to 521 disrupts a critical function of nsp2 in the PRRSV life cycle.

FIG 3.

Loss of infectivity of the mutant nsp2Δ323–521 in primary PAMs. (A) Titers of WT virus and nsp2 mutants in primary PAMs at 48 hpi (MOI of 0.01). (B) Multistep growth curve of nsp2 deletion mutants. Primary PAMs in 24-well plates were infected with WT virus and nsp2 mutants at the indicated MOI. At different times, total viruses were harvested and titrated in MARC-145 cells by the endpoint dilution assay. (C) Effect of incubation time on virus replication. Primary PAMs were infected at a MOI of 5 at 4°C. At the indicated time points, the inoculum was removed, and the cells were supplemented with new infection medium. At different time points postinfection, total viruses were harvested and titrated in MARC-145 cells by the endpoint dilution assay.

The mutant nsp2Δ323–521 fails to establish infection in a 5-week-old piglet model.

To investigate the replication ability of the mutant nsp2Δ323–521 in vivo, 15 specific-pathogen-free (SPF) pigs (5 weeks of age) were randomly divided into three groups, namely, WT, nsp2Δ323–521, and mock. The nsp2Δ323–521 and WT groups were inoculated via the nasal route with the respective virus at a dose of 2 × 10⁵ times the 50% tissue culture infective dose (TCID₅₀), and the mock-infected group was inoculated with Dulbecco’s modified Eagle medium (DMEM). The results showed that the pigs infected with WT virus quickly developed high fever (Fig. 4B). Viremia emerged at around 3 days postinfection (dpi) (Fig. 4D), and the serum antibodies to PRRSV turned positive around 7 dpi (Fig. 4C). Moreover, all 5 pigs died by 17 dpi (Fig. 4A). In contrast, the piglets infected with nsp2Δ323–521 showed good appetite and did not exhibit any observable clinical symptoms at the end of the experiment (Fig. 4A, B, and E). Also, no viremia was detected (Fig. 4D), and the serum antibodies to PRRSV did not turn positive by the end of the study (Fig. 4C). Upon necropsy, different tissues were collected for detection of PRRSV nucleic acids. Again, all of the reverse transcription (RT)-PCR tests were negative for nsp2Δ323–521 (data not shown), and we did not see obvious lesions for pigs (data not shown). Thus, the inability of the mutant nsp2Δ323–521 to replicate in pigs correlates well with its loss of macrophage tropism in vitro.

FIG 4.

Pathogenicity analysis of the mutant nsp2Δ323–521 in a 5-week-old SPF piglet model. (A) Survival curve for piglets nasally inoculated with JXwn06 WT and nsp2△323–521 viruses in each group (n = 5). (B) Rectal temperature fluctuations. (C) Kinetics of piglet antibodies to PRRSV. An IDEXX HerdChek PRRS 2XR ELISA kit was used to detect PRRSV-specific antibodies, and the antibody level was expressed as a sample value/positive value (S/P) ratio. A ratio of ≥0.4 was regarded as seroconversion. (D) Serum viral loads. The virus titer in the serum of individual piglets was determined in MARC-145 cells by the endpoint dilution assay. ***, P＜0.001, significant differences in viral loads between the JXwn06 WT and nsp2△323–521 virus groups. (E) Average daily gain of piglets.

The mutant nsp2Δ323–521 is defective in viral protein synthesis in infected PAMs.

To dissect the mutational block in the PRRSV life cycle, we examined the synthesis of viral structural and nonstructural proteins during infection. For this purpose, PAMs were infected with WT or nsp2Δ323–521 virus at a MOI of 0.1 (Fig. 5A). At different time points postinfection, the cells were collected, and the expression of intracellular viral proteins was analyzed by Western blotting. For WT virus, nsp2 and N were readily detectable around 12 hpi. In contrast, there was no detectable expression of N and nsp2 for the cells infected with nsp2Δ323–521. Increasing the infection dose to a MOI of 10 did not improve the outcome (Fig. 5B). It should be noted that N protein could be detected at 0 hpi, but this likely represented the input virions, as we could not detect N protein at later time points. We also visualized the infection dynamics by immunofluorescence assay (IFA) (Fig. 5C), in which the cells were infected at a MOI of 0.1. For WT JXwn06, sporadic cells were positive for PRRSV beginning at 6 hpi, and the IFA signals increased gradually as the infection proceeded; at 24 hpi, most of the cells were infected. In contrast, it was extremely uncommon to observe a positive signal for the mutant nsp2Δ323–521. Thus, the defect for the mutant nsp2Δ323–521 lies at the level or upstream of viral mRNA synthesis.

FIG 5.

Protein synthesis defect of the mutant nsp2△323–521 in infected PAMs. (A and B) Primary PAMs were infected with WT and nsp2△323–521 viruses at a MOI of 0.1 (A) or 10 (B). The cells were harvested at the indicated time points postinfection and subjected to Western blot analysis, in which β-actin served as a loading control. (C) PAMs infected at a MOI of 0.1 were fixed at the indicated time points postinfection and subjected to IFA with antibodies to nsp2, nsp9, and N. DAPI was used to stain cell nuclei.

The mutant nsp2Δ323–521 is defective in negative-stranded RNA accumulation in infected PAMs.

We next set out to determine whether viral RNA synthesis was normal in infected PAMs. Real-time PCR was used to quantify the abundance of gmRNA at different time points following virus infection at a MOI of 0.1. WT JXwn06 showed a rapid increase in viral gmRNA, indicating active viral replication (Fig. 6A). In contrast, in the cells infected with mutant nsp2Δ323–521, the gmRNA number decreased gradually with time (Fig. 6A). Moreover, antibodies to double-stranded RNAs (dsRNAs) did not detect positive signals (Fig. 6B). Thus, viral RNA replication does not seem to take place in nsp2Δ323–521-infected PAMs.

FIG 6.

Negative-stranded RNA accumulation defect of the mutant nsp2Δ323–521. (A) Viral genome copy numbers. PAMs were infected with JXwn06 WT and nsp2△323–521 viruses at a MOI of 0.1 and harvested at the indicated times postinfection. The viral genome copy numbers were determined by RT-qPCR. (B) Analysis of viral RNA synthesis with anti-dsRNA antibodies. PAMs infected with WT and nsp2△323–521 viruses at a MOI of 0.1 were fixed and stained with antibodies to nsp2 and dsRNA at 24 hpi. (C) RNAscope analysis of positive- and negative-stranded PRRSV RNA synthesis. PAMs were infected with JXwn06 WT and nsp2△323–521viruses at a MOI of 5. At the indicated time points, the cells were fixed and subjected to RNAscope in situ hybridization analysis. (D) Detection of positive- and negative-stranded RNAs. Positive-stranded and negative-stranded RNAs were detected with specific probes. Nuclear DNA was stained with DAPI. The images were captured with a Leica confocal microscope and processed by using ImageJ.

To investigate the defective stage of viral RNA synthesis, we employed RNAscope in situ hybridization technology. Two different sets of nucleic acid probes were designed to distinguish positive-stranded RNAs from negative-stranded RNAs, and viral RNA synthesis was examined at different time points postinfection (Fig. 6C and D). The RNA probes were PRRSV specific and gave very low background staining of uninfected PAMs (Fig. 6C and D, mock). The positive-stranded RNA probes were able to penetrate the viral particles, and positive signals could be detected for both WT and nsp2Δ323–521 viruses at 0 hpi. Consistent with the stage of virus attachment, the signals mainly surrounded the cell surface at that time point, and the small dots likely represented the individual input virions (Fig. 6C, 0h). For WT virus, as the infection proceeded, the dots moved inside the cell (Fig. 6C, 3h). An intense ring structure, an indication of active RNA synthesis, began to appear in a few cells at 6 hpi and then progressed to most infected cells around 12 hpi. The signals for negative-stranded RNAs were much weaker than those for the positive-stranded RNAs at the same time point (Fig. 6D, 6h and 9h), suggesting a mode of polarized RNA synthesis. The signals became detectable at 6 hpi and intensified around 12 hpi. Moreover, the signals presented as many discrete dot structures, suggesting that the replication sites were not continuous. For the mutant nsp2Δ323–521, it was extremely rare to see positive signals for negative-stranded RNAs during the course of infection (Fig. 6D), suggesting that the mutant might have a defect in negative-stranded RNA synthesis, the first step in viral RNA synthesis. In contrast, we could detect signals for positive-stranded RNAs inside the cells (Fig. 6C). These signals did not intensify, however, but gradually faded away (Fig. 6C, bottom two), suggesting that they originated from the input virions. Thus, viral RNA transcription is defective in nsp2Δ323–521-infected PAMs.

The mutant nsp2Δ323–521 is competent for attachment and internalization.

It is possible that the mutant nsp2Δ323–521 has an entry defect that leads to a block in the efficient release of viral RNA. We first examined the attachment competency by incubating PAMs with either WT or nsp2Δ323–521 virus at 4°C for different times, as indicated (Fig. 7A). The cells were then washed and collected for total RNA extraction, and the amount of the virus attached to the cells was measured by real-time qPCR targeting the coding region of the viral polymerase (nsp9). The results showed that the mutant nsp2Δ323–521 retained an ability to attach to the surface similar to that of the WT virus (Fig. 7A).

FIG 7.

Competence of the mutant nsp2Δ323–521 for attachment and internalization. (A) PAMs infected with either WT or nsp2Δ323–521 virus were incubated at 4°C for different times, as indicated, to allow virus attachment. The cells were then washed, and the amounts of virus attached to the cells were measured by real-time qPCR targeting the coding region of the viral polymerase region (nsp9). NS, no significant difference. The error bars indicate standard errors of the means. The data analyses represent three independent experiments. (B) PAMs were infected with WT or nsp2△323–521 virus at a MOI of 50 at 4°C. After 2 h of incubation, the inoculum was removed, the cells were supplemented with fresh infection medium, and the temperature was shifted to 37°C. At different times postinfection, the cells were fixed and stained with antibodies to GP5 and N. Nuclear DNA was stained with DAPI. The images were captured with a Leica confocal microscope and processed by using ImageJ. (C) Electron microscopic analysis of virus endocytosis was performed at the indicated times postinfection. Bars, 50 nm (top) or 100 nm (bottom). The blue arrow points to PRRSV virions.

We next used confocal microscopy to monitor the dynamics of endocytosis (Fig. 7B). PAMs were infected with either WT or nsp2Δ323–521 virus at a MOI of 50. After incubation at 4°C for 2 h, the inoculum was removed, and the cells were washed three times before being supplemented with fresh culture medium. At different time points postinfection, the cells were fixed and stained with antibodies to either N or GP5. Consistent with the RNAscope detection (Fig. 6C), the mutant nsp2Δ323–521 was clearly competent for attachment and induction of endocytosis (Fig. 7B). At 0 hpi, the signals for structural proteins of mutant nsp2Δ323–521 were mainly on the cell surface, forming a nice ring; by 3 hpi, they moved inside and became diffused in the cytoplasm, indicating active induction of endocytosis. This result was confirmed by electron microscopy, which revealed many endocytosed virions in the intracellular vesicles (Fig. 7C). By 6 hpi, the signals for N and GP5 became strong for the WT virus, indicating the de novo synthesis of large amounts of viral structural proteins. In contrast, the signals for mutant nsp2Δ323–521 did not intensify but rather gradually faded with time (Fig. 7B). These results suggest that the defect for the mutant nsp2Δ323–521 lies downstream of internalization.

The mutant nsp2Δ323–521 may be impaired in the penetration process.

We carried out two different assays to gain insight into the penetration process. In the first assay, we performed the multistep growth curve analysis but with a focus on the eclipse stage of viral replication. The primary PAMs were infected with PRRSV at a MOI of 0.1. After incubation for 2 h at 4°C, the unbound viruses were washed off, and the temperature was then shifted to 37°C to allow virus internalization and penetration. Samples were taken every hour postinfection and then titrated for the presence of intact viral particles. As shown in Fig. 8A, WT PRRSV showed a rapid and steady decrease in virus titers during the first 9 h. This decrease likely reflects the virus penetration process, during which the viral particles lose integrity due to the virus-cell fusion, and some may be attributed to nonspecific degradation in the endosomes. The decrease in virus titer was slower for the mutant nsp2Δ323–521 than for the WT virus, although not statistically significantly, suggesting that the mutant might have a problem with penetration.

FIG 8.

Mutational effect of nsp2 deletion on PRRSV penetration. (A) Entry kinetics of WT and nsp2△323–521 viruses. PAMs were incubated with WT and nsp2Δ323–521 viruses at a MOI of 0.1 at 4°C for 2 h, to allow efficient attachment. The inoculum was then removed, the cells were supplemented with fresh infection medium, and the temperature was shifted to 37°C. At the indicated times postinfection, the cells were collected, and the viruses were titrated on MARC-145 cells by the endpoint dilution assay. (B) Membrane fusion assay. Heat-inactivated or untreated R18-labeled PRRSV virions were used to infect primary PAMs at a MOI of 0.1. After 30 min of incubation at 4°C, the cells were washed, and the temperature was shifted to 37°C to allow virus penetration. The R18 dye dequenching was monitored constantly, and the dequenching rate was calculated as described in Materials and Methods. ***, P＜0.001; NS, no significant difference. The error bars indicate standard deviations. The data analyses represent three independent experiments. (C) Time course analysis of membrane fusion by confocal microscopy. Heat-inactivated or untreated R18-labeled PRRSV virions were used to infect primary PAMs at a MOI of 1. After 30 min of incubation on ice for attachment, the cells were washed, and the temperature was shifted to 37°C to allow virus penetration. R18-labeled PRRSV-cell fusion was monitored at the indicated time points. Nuclear DNA was stained with DAPI. The images were captured with a Leica confocal microscope and processed by using ImageJ.

In the second assay, we took advantage of octadecyl rhodamine B chloride (R18), a small lipophilic dye that has been widely used in studies of virus-cell fusion (33–35). This dye is highly self-quenched at high concentrations but diffuses laterally to host cellular membranes upon virus-cell membrane fusion, leading to an increase in fusion-associated fluorescence (33). We labeled both WT and nsp2Δ323–521 viruses with R18 and purified them by sucrose gradient centrifugation. The R18-labeled viruses were then used to infect primary PAMs at 4°C for 30 min, to allow efficient virus attachment. The inoculum was then removed, and the cells were washed three times before the temperature was shifted to 37°C to initiate the fusion process. The fusion-induced fluorescence intensity was monitored every 15 or 30 min. The R18 dequenching rate was calculated as described in Materials and Methods. As shown in Fig. 8B, the R18 dequenching rate with the mutant nsp2Δ323–521 appeared to be lower than that with the WT virus, indicating likely impaired membrane fusion. Similar results were revealed by the time course confocal microscopy assay (Fig. 8C). This effect is unlikely to be due to the altered abundance of other structural proteins (e.g., GP3 and GP5/M), as they were packaged well into the virion particles, compared to the WT virus (Fig. 2D, right). Because this method measures hemifusion, it is not clear whether the complete fusion step is affected as a result of deletion of the nsp2 hypervariable region. In any case, these data, taken together, suggest that the mutant nsp2Δ323–521 has a defect (or defects) lying in the steps between virus penetration and negative-stranded RNA synthesis.

DISCUSSION

The nsp2 replicase protein possesses multiple functions in the PRRSV life cycle (1, 18, 21–23, 36–39), such as replicase polyprotein maturation and anti-host innate immunity. However, it has remained unclear why the PRRSV nsp2 needs to undergo rapid and broad genetic variations in the field. It has been proposed that the B-cell-epitope-rich middle regions may serve as immune decoys (40–42), but this model does not explain well why large deletions that should have removed certain B-cell epitopes frequently occur within this region. In this report, we reveal that the nsp2 middle region plays an important regulatory role in maintaining the balance of different PRRSV mRNA species. Moreover, we found that the nsp2 middle region is a critical determinant of the PRRSV cellular range to primary PAMs. The relative significance and insight of these findings are discussed below.

Structurally regulatory role for the PRRSV nsp2 middle hypervariable region.

The nsp2 middle region is the most genetically variable spot within the PRRSV genome (12, 27, 43, 44). Deletions, insertions, and substitutions are often seen, and at least 23 deletion patterns have been identified to date (7, 12, 27, 43, 44). Thus, it seems likely that this region does not possess a specific function but rather plays a structurally regulatory role. Our studies here provide further evidence in support of this notion. First, although either nsp2 aa 323 to 521 or aa 520 to 782 could be removed individually, the whole region of nsp2 aa 323 to 782 could not be deleted. Similar results were found in the North American prototype strain VR-2332 (15). Second, the nsp2 region of aa 323 to 521 was dispensable for HP-PRRSV JXwn06 replication in MARC-145 cells, but deletion of nsp2 aa 450 to 521 within the same region was lethal to the virus (Fig. 1B). Third, deletion of the nsp2 hypervariable region differentially modulated accumulation of a subset of viral mRNAs. All of these facts point to a likely scenario in which the genetic variations regulate the overall structure of nsp2, such as protein folding or positioning of the protease domain, TMD, or CT, which may have a profound effect on the nsp2 interactome.

It is also interesting to note that deletion of the same region of nsp2 might have different outcomes for different PRRSV strains. In the case of the North American prototype strain VR-2332, the nsp2 region of aa 443 to 531 was not required for viral replication in cell culture (15). In contrast, deletion of a similar region (aa 433 to 521) was lethal to HP-PRRSV strain JXwn06. A similar lethal effect was observed for HP-PRRSV-derived strain HuN4-F112 (32). In addition, deletion of nsp2 aa 323 to 726 of PRRSV VR-2332 did not affect virus viability (15, 45), whereas deletion of the same region was lethal to the JXwn06 strain (Fig. 1). Since PRRSV nsp2 is highly genetically variable, we speculate that these differential effects may be attributed to nsp2 genetic mutations lying outside the deletion region, which may have a profound influence on the overall structural flexibility of nsp2.

Insight into PRRSV RNA synthesis.

Our findings also provide insight into the process of PRRSV gmRNA and sgmRNA synthesis. Although it is well known that sgmRNA synthesis involves a discontinuous strategy (1) that is guided by specific RNA structures and sequence base pairing, the molecular details of how viral mRNA synthesis is controlled have remained poorly understood. Studies of the arterivirus prototype have revealed that EAV nsp1 coordinates gmRNA and sgmRNA synthesis, mutations within nsp1 subdomains can differentially affect accumulation of viral mRNA species, and the balance of mRNA species is critical for virus production (46). In this report, we provide genetic evidence for the involvement of nsp2 in PRRSV mRNA synthesis. Specifically, deletion of nsp2 aa 323 to 521 significantly affected the accumulation of sgmRNA5. In contrast, the synthesis of sgmRNA2 to sgmRNA4 was not apparently affected (Fig. 2A). Deletion of nsp2 aa 520 to 782 had a different effect. When normalized to the level of gmRNA, the deletion led to alteration of all sgmRNAs but to different extents (Fig. 2B and C). Consequently, these viruses were attenuated in replication in cell culture (Fig. 1C). Thus, our findings provide strong evidence to suggest that PRRSV nsp2 is directly involved in viral mRNA synthesis and the hypervariable region plays a regulatory role in this process. The genetically flexible nature of this region may help regulate the balance of PRRSV mRNA species to accommodate different environments. The fact that the deletion of nsp2 aa 323 to 521 did not cause uniform reduction of all viral mRNA species but rather affected PRRSV RNA in a mRNA-specific manner suggests that the viral RTC is a complex and dynamic system and synthesis of individual mRNAs may be controlled by slightly different mechanisms, despite universal utilization of a discontinuous replication strategy for sgmRNA synthesis. Thus, our findings provide the first glimpse into the process of regulation of PRRSV mRNA synthesis.

PRRSV nsp2 and viral cellular tropism.

It is well known that PRRSV possesses a specific preference for PAMs in vivo, and this biological property has been attributed to the minor structural proteins GP2, GP3, and GP4, which interact with CD163, a scavenger receptor present on the surface of the monocyte/macrophage lineage, for penetration (47–50). It remains unclear, however, whether these minor structural proteins alone are sufficient to confer this tropism. In this study, we provided genetic evidence to suggest that the nsp2 middle region is also a critical determinant of the PRRSV cellular range. This insight was gained through a unique nsp2 deletion mutant (nsp2Δ323–521), which lost infectivity in primary PAMs but replicated relatively well in MARC-145 cells. In addition, we found that the mutant nsp2Δ323–521 could grow in CD163-transfected immortalized PAM cells (3D4/21) (data not shown). We have tried to pinpoint the difference but are discouraged by gene transcriptomic profiling analyses, which showed that there are huge differences between the immortalized PAM cell line and primary PAMs; at least one-half the genes among the thousands are altered in terms of their expression levels or patterns (data not shown). In any case, our studies suggest that the defect is indeed specific to primary PAMs, the most relevant cells for virus replication in vivo. Although the exact mechanism of how the deletion affects nsp2 function remains unclear, dissection of the virus life cycle revealed that the defect lies between virus penetration and negative-stranded RNA synthesis. Moreover, we provided some evidence that this mutant is likely impaired in the penetration step. This idea is supported by the virus entry kinetics assay (Fig. 7A) and the R18 virion labeling assay (Fig. 7B), the latter of which showed that the mutant nsp2Δ323–521 had a reduced fusion rate, compared to the WT virus. The R18 assay generally measures hemifusion, however, and it is not clear whether the subsequent complete fusion is blocked (to be addressed in the future). In addition, this defect is unlikely to be due to the reduced packaging of GP2 to GP4 proteins, as we showed that their incorporation into virions was not affected (Fig. 2D, right). Thus, the impaired phenotype suggests that PRRSV nsp2 is likely involved in the penetration process in a manner that is either direct or indirect. It is possible that PRRSV nsp2 interacts with a currently unknown coreceptor to promote virus penetration (like HIV, which requires coreceptors for fusion [51–53]) or cooperates with minor structural proteins to modulate their function in the fusion process. Consistent with this possibility, PRRSV nsp2 has been shown to be packaged into virion particles (54) and also interacts with the major structural protein GP5 (55). It is also interesting that a recent study showed that nsp2 is linked to limited cross-neutralization ability between different PRRSV strains. Our results here are in agreement with this finding, and further studies are warranted to address the role of nsp2 in early PRRSV infection (56).

The mutant nsp2Δ323–521 was also found to be defective in negative-stranded RNA synthesis. This could be due to the ripple effect of the block in penetration, a defect in viral RTC assembly, or a combination of these events. Thus, it is highly likely that the failure to replicate in primary PAMs is a cumulative effect of multiple defects, ranging from penetration to RNA transcription. We tried to decouple the penetration from the downstream events by directly transfecting viral gmRNA into primary PAMs. Unfortunately, this assay was not successful, due to the extremely low transfection efficiency of the primary cells.

Overall, our studies provide strong genetic evidence that the nsp2 hypervariable region not only plays a critical regulatory role in maintaining the balance of different viral mRNA species but also regulates the PRRSV cell range. These findings represent an important step toward understanding nsp2 functions in the PRRSV life cycle and also provide important insights into PRRSV mRNA synthesis and tissue tropism.

MATERIALS AND METHODS

Ethics statement.

The animal experiments were carried out according to the Chinese Regulations of Laboratory Animals, the Guidelines for the Care of Laboratory Animals (Ministry of Science and Technology of People’s Republic of China), and Laboratory Animal—Requirements of Environment and Housing Facilities (standard no. GB 14925-2010; National Laboratory Animal Standardization Technical Committee). The animal trials in this study were approved by the Laboratory Animal Ethical Committee of China Agricultural University (license no. CAU20160828-2).

Cells, viruses, and commercial antibodies.

MARC-145 cells were grown at 37°C with 5% CO₂ in Gibco DMEM (Invitrogen, Carlsbad, CA, USA) containing 10% fetal bovine serum (FBS) (Invitrogen). The PRRSV JXwn06 strain was used in the study (13). PAMs were prepared from the lung lavage fluid of 5- to 6-week-old healthy piglets free of PRRSV, as described previously (57). Primary PAMs were maintained at 37°C with 5% CO₂ in Gibco RPMI 1640 medium (Invitrogen) supplemented with 10% FBS, 100 mg/ml kanamycin, and 50 U/ml penicillin or were cryopreserved in liquid nitrogen for later use. Mouse anti-β-actin monoclonal antibody (MAb) was purchased from Sigma (product no. A1978). Anti-dsRNA antibody was purchased from Scicons (J2; English and Scientific Consulting Bt., 10010200). Horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG(H+L) and HRP-conjugated goat anti-rabbit IgG(H+L) were purchased from ZSGB-Bio (Beijing, China). Alexa Fluor 488-conjugated goat anti-rabbit IgG(H+L) F(ab′)₂ fragment, Alexa Fluor 568-conjugated goat anti-mouse IgG(H+L) F(ab′)₂ fragment (Invitrogen), and Alexa Fluor 488-conjugated goat anti-mouse IgG(H+L) F(ab′)₂ fragment were all purchased from Molecular Probes (Invitrogen).

PRRSV-specific antibodies.

The mouse MAbs specific to PRRSV nsp1β, nsp2 (E3G11), and nsp9 were prepared in our laboratory and have been reported previously (42, 58). The nsp2 polyclonal antibodies were obtained by immunizing rabbits with immunogen containing the nsp2 protease domain (aa 12 to 323), with a His₆ tag, expressed and purified from Escherichia coli BL21 cells. Mouse anti-GP3 MAb and mouse anti-M MAb (3F7) were kindly provided by Zhijun Tian (Harbin Veterinary Research Institute, Harbin, China) (59, 60). Mouse anti-GP5 MAb was kindly provided by Guangzhi Tong (Shanghai Veterinary Research Institute, Shanghai, China). The mouse anti-N protein MAb was kindly provided by Ping Jiang (Nanjing Agriculture University, Nanjing, China). PRRSV mouse anti-N MAb (MAb SDOW17) was purchased from Rural Technologies (Brookings, SD, USA). The reactivity and specificity of the respective antibodies were tested under the conditions of transfection and infection, and mock-transfected or mock-infected cells served as negative controls. Normal rabbit serum was also used as a negative control. All of the antibodies showed good reactivity and specificity.

Construction of PRRSV full-length cDNA clone of JXwn06.

Low-copy plasmid vector pWSK29 was modified to contain a CMV promoter, multiple cloning site (MCS), hepatitis delta virus (HDV) ribozyme, and SV40 signal. The four PRRSV genomic fragments were precisely assembled stepwise into a modified low-copy-number plasmid vector. Details of the process are as follows. The CMV promoter was amplified from pcDNA3.1 (Invitrogen) by introducing DraIII and SacI restriction sites into the primers (Table 1) and was cloned into pWSK29 digested with DraIII and SacI, leading to generation of the plasmid pWSK29-CMV. Next, a new MCS was designed to contain a restriction enzyme linker (SwaI-XhoI-FseI-BstBI-MluI-AvrII-NheI-AscI-EcoRV-NotI-BsrGI-PacI) to facilitate the assembly of a PRRSV full-length cDNA clone. HDV ribozyme and SV40 signal nucleotides sequences were synthesized by Genscript (Nanjing, China) by introducing SwaI and SacI restriction sites into the primers (Table 1). The MCS-HDV-SV40 sequence was synthesized in vitro and cloned into pWSK29-CMV to generate the plasmid pWSK29-CMV-MCS-HDV-SV40. The four fragments were assembled into this vector in two steps in the order of B-C-D and A. The full-length infectious clone was verified by DNA sequencing, and the final plasmid was named pCMV-JXwn06. A specific restriction site for the enzyme BstBI was introduced into the full-length infectious clones between fragments A and B as a genetic marker that does not exist in the genome of the WT virus, to distinguish between WT virus and rescued virus.

TABLE 1.

Primers used for infectious cDNA clone construction and viral mRNA quantification

Primer namea	Primer sequence and restriction site(s)b	Usage
CMV3.1-F	5ʹ-GGCCCACTACGTGGTTGACATTGATTATTGACTAG-3ʹ (DraIII)	CMV amplification
CMV3.1-R	5ʹ-GCTGGAGCTCGCGGCCGCTTAATTAAGACGGTATTTAAATACTAAACCAGCTCTGCTTATATAGACCTCCCACCGT-3ʹ (SacI, PacI, and SwaI)
MCS-HDV-SV40-F	5ʹ-TAGTATTTAAATACCGTCCTCGAGGC-3ʹ (SwaI)	MCS-HDV-SV40 amplification
MCS-HDV-SV40-R	5ʹ-GCTGGAGCTCGATCCAGACATGATAA-3ʹ (SacI)
A-R	5ʹ-CAGAGCTGGTTTAGTATTTAAATACCGTCATGACGTA-3ʹ (SwaI)	PRRSV fragment A amplification
A-F	5ʹ-CCTCCCCCTGAAGGCTTCGAAATTTGCCTGATCTTTAG-3ʹ (BstBI)
pBlue-A-F	5ʹ-GCGCGATATCATTTAAATACCGTCATGACGTATAGGTGTTGGCTCTATGCCACGGC-3ʹ (EcoRV and SwaI)	PRRSV fragment A shuttle plasmid construction
pBlue-A-R	5ʹ-GCGCGCGGCCGCTTCGAAATTTGCCTGATCTTTAGTCCATTCAGCTGG-3ʹ (NotI and BstBI)
Probe for ORF7-biotin	5ʹ-TGGCTGGCCATTCCCCTTCTTTTTCTTTTGCTGCTTGCCG-3ʹ	Northern blotting
nsp9-F	5ʹ-CCTGCAATTGTCCGCTGGTTTG-3ʹ	Virus titration by PCR
nsp9-R	5ʹ-GACGACAGGCCACCTCTCTTAG-3ʹ
gRNA-F	5ʹ-GTCTCTCCACCCCTTTAACC-3ʹ	gmRNA quantification
gRNA-R	5ʹ-AATGCACGTGGCAACGTCCAC-3ʹ
sgmRNA2-F	5ʹ-CCCTTTAACCATGAAATGGGGT-3ʹ	sgmRNA2 quantification
sgmRNA2-R	5ʹ-GGAGCAAACCAGTCTGATGC-3ʹ
sgmRNA3-F	5ʹ-CCCTTTAACCATGGCTAATAGC-3ʹ	sgmRNA3 quantification
sgmRNA3-R	5ʹ-TTCAAGGATCTCAGCGGCTGC-3ʹ
sgmRNA4-F	5ʹ-CCCTTTAACCATGGCTGCGTC-3ʹ	sgmRNA4 quantification
sgmRNA4-R	5ʹ-CCATGCCTAAGGCAGCTGATG-3ʹ
sgmRNA5-F	5ʹ-CCCTTTAACCATGTTGGGGAAG-3ʹ	sgmRNA5 quantification
sgmRNA5-R	5ʹ-GGAAACAATGTGAGTCAACAC-3ʹ
sgmRNA6-F	5ʹ-CCCTTTAACCATGGGGTCGTC-3ʹ	sgmRNA6 quantification
sgmRNA6-R	5ʹ-GAAGGTAAAAGCACAATTCAG-3ʹ
sgmRNA7-F	5ʹ-CCCTTTAACCATGCCAAATAAC-3ʹ	sgmRNA7 quantification
sgmRNA7-R	5ʹ-GGTAAAGTGATGCCTGACGTC-3ʹ

^aF, forward primer; R, reverse primer.

^bUnderlining indicates enzyme restriction sites.

Recovery of viruses.

Plasmids of the infectious cDNA clones were isolated using a Promega miniprep kit (Promega, Madison, WI). MARC-145 cells were seeded on 6-well plates at a confluence of 70% and were cultured for 24 h before transfection with 2 μg purified plasmid DNA and 6 μl of Lipofectamine LTX reagent (Invitrogen), according to the manufacturer’s protocol. The cytopathic effects (CPEs) were monitored daily, and the rescued viruses were examined by IFA using the PRRSV anti-N MAb (MAb SDOW17; Rural Technologies). The genomes were extracted from third-passage viral cultures of the infected cells for further sequencing verification, as described previously (13).

Deletion mutagenesis of PRRSV JXwn06 nsp2.

Fragment A (bases 1 to 4818) was amplified from pCMV-JXwn06 by introducing EcoRV and NotI sites into the primers (Table 1) and then was cloned into the pBluescript II SK(+) vector (Agilent Technologies); we named the product pBlue-A. To delete JXwn06 nsp2 nucleotides, mutagenesis PCR was performed on pBlue-A by using PfuUltra II Fusion HS DNA polymerase (product no. 600670; Agilent Technologies). After confirmation by sequencing, mutated fragment A was inserted back into the infectious clone backbone to construct the final full-length deletion mutant clone.

Growth kinetics of rescued viruses in MARC-145 cells and in primary PAMs.

MARC-145 cells and primary PAMs were infected with the indicated viruses at the MOI indicated. After 1 to 2 h of incubation at 37°C, MARC-145 cells were first washed with acid buffer (135 mM NaCl, 10 mM KCl, 40 mM citric acid [pH 3.0]) to inactivate the viruses remaining on the cell surfaces and then washed three times with DMEM, while primary PAMs were washed three times with RPMI 1640 medium; the cells were finally overlaid with DMEM and RPMI 1640 medium containing 2% FBS, respectively. At the indicated times postinfection, the medium and cells were harvested, frozen, and thawed. Samples were titrated on MARC-145 cells by using the endpoint dilution assay described previously (13). Each time point was independently tested three times.

Animal experiments.

Fifteen 5-week-old SPF piglets were randomly divided into three groups with 5 pigs each. The groups were raised separately, in different isolated rooms. The piglets were confirmed by PCR to be negative for PRRSV, pseudorabies virus, porcine circovirus type 2, and classic swine fever virus. The PRRSV-specific antibodies were determined with the IDEXX HerdChek PRRS 2XR enzyme-linked immunosorbent assay (ELISA) kit. For the challenge experiment, piglets were intranasally inoculated with JXwn06 or nsp2△323–521 at a dose of 2 × 10⁵ times the TCID₅₀. The mock-infected group was inoculated with 2 ml MARC-145 cell culture supernatant. Serum samples were collected from the challenged pigs at the indicated times to examine viremia using a microtitration infectivity assay, and serological immune responses were determined with the IDEXX HerdChek PRRS 2XR ELISA kit (sample value/positive value ratio). Lung tissue samples were collected at necropsy, fixed with 4% paraformaldehyde solution at room temperature for 48 h, and then processed by routine histopathological procedures. Each sample was stained with hematoxylin and eosin for examination of pathological changes.

Virus titration by qPCR.

Primary PAMs were infected with WT and nsp2 mutant viruses at a MOI of 0.1. After 1 h of incubation, cells were washed three times with RPMI 1640 medium and finally overlaid with RPMI 1640 medium containing 2% FBS. Total RNAs were extracted from infected primary PAMs using TRIzol reagent, according to the manufacturer’s instructions (Invitrogen). Five hundred nanograms of RNA samples were used for cDNA synthesis by using HiScript II Q RT SuperMix for qPCR with genomic DNase (Vazyme, Nanjing, China). The primers used to detect the copy numbers of the PRRSV nsp9 gene have been reported previously (61, 62). Real-time PCR was performed on an ABI7500 real-time PCR system (Applied Biosystems) using SYBR green SYBR Select MasterMix (Applied Biosystems), according to the manufacturer’s instructions. Primers used for the qPCR assay are listed in Table 1.

Analysis of the relative abundance of sgmRNA species by qPCR.

MARC-145 cells were infected with PRRSV strain JXwn06 and mutant strains at a MOI of 0.1. At 48 hpi, the total RNAs were extracted with TRIzol reagent (product no. 15596026; Thermo Fisher). RT was performed using a FastKing RT kit (with genomic DNase) (product no. KR116-02; Tiangen), following the user guide. The cDNA was synthesized from the total cellular RNA by RT using the indicated primers (random primers). The relative qPCR was performed with the Applied Biosystems SYBR Select Master Mix (product no. 4472913; Thermo Fisher), according to the manufacturer’s recommendations. The PCR was performed in a 20-μl reaction mixture containing 0.2 μM gene-specific primer, 10 μl SYBR Select Master Mix, and 2 μl cDNA template. The PCR parameters were as follows: 50°C for 2 min, 95°C for 2 min, and 40 cycles of 95°C for 15 s, 60°C for 15 s, and 72°C for 60 s. To measure total viral RNA, a pair of internal primers in the ORF7 gene was used to amplify all sgmRNAs and gmRNA. To measure gmRNA and each viral sgmRNA, a pair of primers in the corresponding ORF was used for qPCR. The levels of individual sgmRNA species were normalized to the corresponding gmRNA level, and the relative abundance was calculated as the mutant virus/parental virus ratio. The primers used for qPCR are listed in Table 1.

Northern blotting.

MARC-145 cells were infected with rescued virus at a MOI of 0.1. After 1 h of incubation at 37°C, MARC-145 cells were washed three times with DMEM and finally overlaid with DMEM containing 2% FBS. Cells were collected at 48 hpi, and total RNAs were extracted with TRIzol reagent (Invitrogen). The total RNAs were separated by using denaturing formaldehyde agarose gel electrophoresis, followed by downward capillary transfer from the gel to a positively charged nylon membrane (Invitrogen) for 16 h. The membrane was then cross-linked using a UV cross-linker (product no. CL-1000; UVP), prehybridized for 30 min at 42°C, and hybridized with the biotin-labeled probe (Table 1) for 20 h at 65°C in a hybridization oven (product no. HB-1000; UVP). The membrane was washed twice with low-stringency wash solution (product no. AM8673; Ambion) for 5 min at room temperature and then twice with high-stringency wash buffer (product no. AM8674; Ambion) for 15 min at 42°C. Detection of viral RNA signals was carried out using the chemiluminescent nucleic acid detection module, according to the manufacturer’s instructions (product no. 89880; Thermo Scientific), and results were captured with a FluorChem E apparatus (ProteinSimple, Santa Clara, CA, USA).

Virus stock preparation.

For large-scale preparation of virus, MARC-145 monolayers in 75-cm² flasks were infected with JXwn06 and nsp2△323–521 viruses at a MOI of 0.1 and maintained at 37°C in 5% CO₂ until about 70% CPE was reached (about 36 h). Cell supernatants were collected and centrifuged at 5,000 × g for 1 h at 4°C to remove cellular debris. This crude virus preparation was further purified through a 30% sucrose cushion via ultracentrifugation at 26,000 rpm in a SW28 rotor (Beckman Coulter) for 2 h at 4°C. Viral pellets were gently reconstituted in 200 μl of TNE buffer (10 mM Tris-HCl [pH 7.4], 0.1 M NaCl, 1 mM EDTA) per tube and stored in –80°C. The virus stocks were titrated in MARC-145 cells.

Immunofluorescence assay.

Primary PAMs were infected with PRRSV at the indicated MOI. After 2 h of incubation at 4°C or 37°C, PAMs were washed three times with RPMI 1640 medium and overlaid with RPMI 1640 medium containing 2% FBS. At specific times postinfection, the cells were fixed with 3.7% paraformaldehyde for 10 min at room temperature and then washed three times with phosphate-buffered saline (PBS) (5 min each wash). The cells were then permeabilized with 0.1% Triton X-100 with 2% bovine serum albumin (BSA) for 10 min and blocked with 2% BSA in PBS for 30 min at room temperature. Subsequently, the cells were incubated with the specific primary antibodies at room temperature for 1 h and washed three times with PBS (5 min each wash). The cells were then incubated with the appropriate secondary antibodies, including Alexa Fluor 488-conjugated goat anti-rabbit IgG(H+L) F(ab′)₂ fragment (Invitrogen), Alexa Fluor 568-conjugated goat anti-mouse IgG(H+L) F(ab′)₂ fragment (Invitrogen), and Alexa Fluor 488-conjugated goat anti-mouse IgG(H+L) F(ab′)₂ fragment (Invitrogen), for another 1 h at room temperature. Nuclear DNA was stained with 4′,6-diamidino-2-phenylindole (DAPI). The images were captured with a confocal microscope and processed by using ImageJ. The primary antibodies used were as follows: nsp2 polyclonal antibody (1:3,000 dilution), nsp9 MAb (1:2,000), N MAb (1:1,000), GP5 MAb (1:1,000), and mouse anti-dsRNA (1:200) (Scicons). The secondary antibodies were used at a dilution of 1:1,000.

Western blotting.

For Western blotting, the virus- or mock-infected cells were lysed for 30 min at 4°C using NP-40 buffer (50 mM Tris, 150 mM NaCl, 0.5% NP-40, 0.5 mM EDTA) containing 1 mM phenylmethylsulfonyl fluoride (PMSF) and 1 mg/ml protease inhibitor cocktail (product no. p8340; Sigma), with gentle rotation. The total cell lysates were clarified by centrifugation at 12,000 rpm for 30 min. The protein concentrations were determined by using a bicinchoninic acid (BCA) assay kit (product no. 23225; Thermo Scientific). Proteins were fractionated by SDS-PAGE, blotted onto polyvinylidene difluoride (PVDF) membranes (Millipore), and blocked overnight at 4°C with 5% nonfat milk in PBS. The membranes were probed for 8 h at room temperature with specific primary antibodies. After being washed with washing buffer (0.1% Tween 20 in PBS), the membranes were probed for 1 h at room temperature with goat anti-mouse IgG(H+L) or goat anti-rabbit IgG(H+L) secondary antibodies conjugated with HRP (ZSGB-Bio). After being washed with washing buffer (0.1% Tween 20 in PBS), membrane-bound antibodies were detected using enhanced chemiluminescence (ECL) detection reagents (Vigorous, Beijing, China) and a chemiluminescence apparatus (ProteinSimple). The primary antibodies used were as follows: nsp1β MAb (1:5,000 dilution), nsp2 polyclonal antibody (1:2,000), GP3 MAb (1:1,000), GP5 MAb (1:1,000), M MAb (1:1,000), N MAb (1:1,000), and β-actin MAb (1:5,000). HRP-conjugated goat anti-mouse IgG(H+L) (1:10,000) and goat anti-rabbit IgG(H+L) (1:10,000) secondary antibodies were used.

Transmission electron microscopy.

Primary PAMs were infected with PRRSV at a MOI of 50. After 2 h of incubation at 4°C, primary PAMs were washed three times with RPMI 1640 medium and overlaid with RPMI 1640 medium containing 2% FBS. Cells were washed and collected at 0 h and 3 h postinoculation, fixed for 2 h at room temperature with 2.5% glutaraldehyde in 100 mM phosphate buffer, and then postfixed for 2 h at room temperature with 1% osmium tetroxide (Polysciences, Warrington, PA, USA) after three washes in phosphate buffer. The cells were dehydrated in a graded series of ethanol and finally embedded in Eponate 12 resin (Ted Pella, Inc., Redding, CA, USA). After resin polymerization, ultrathin sections were processed using an ultramicrotome (Leica, Wetzlar, Germany) and stained with uranyl acetate (Ted Pella, Inc.) for 1 h at room temperature. Virus particles were observed using a JEOL 1400 electron microscope (JEOL, Tokyo, Japan).

RNA in situ hybridization.

RNA in situ hybridization was performed using the RNAscope multiplex fluorescent detection reagents v2 kit (product no. 323110; Advanced Cell Diagnostics, Hayward, CA), according to the manufacturer’s instructions. A total of 8 double-Z branched pairs were designed to detect the positive-sense genome (product no. 497191; Advanced Cell Diagnostics) by targeting the highly conserved N protein gene (ORF7; nucleotide positions 14,800 to 15,151), whereas a total of 20 double-Z branched pairs, spanning ORF6, ORF7, and the 3′ untranslated region of the PRRSV genome (nucleotide positions 14,286 to 15,273), were used to detect negative-sense viral RNA (product no. 510989; Advanced Cell Diagnostics). Primary PAMs were seeded on Lab-Tek II chamber slides (product no. 154534; Thermo Fisher) and infected with PRRSV at a MOI of 5. After 2 h of incubation at 4°C, primary PAMs were washed three times with RPMI 1640 medium and overlaid with RPMI 1640 medium containing 2% FBS. At the indicated times postinfection, cells were fixed with 10% neutral buffered formalin, followed by dehydration and rehydration of the cells with ethanol and subsequent treatment with hydrogen peroxide (Advanced Cell Diagnostics) and protease III (Advanced Cell Diagnostics). Target probes were hybridized for 2 h at 40°C in a HybEZ oven (Advanced Cell Diagnostics), followed by a cascade of signal amplification and a series of washing procedures. Hybridization signals were detected with a TSA Plus Cyanine 3 kit (product no. NEL744E001KT; PerkinElmer). Nuclear DNA was stained with DAPI. The images were captured with a Leica confocal microscope and processed by using ImageJ.

Virus labeling.

The fluorescent probe R18 (Invitrogen) was inserted into the viral envelope as described previously (33, 34, 63, 64). The total protein concentrations of purified PRRSV virions were quantified with a BCA protein assay kit (Thermo Scientific). To label the viruses with R18 (2 mM in ethanol), 0.5 to 1 mg purified PRRSV (1 mg/ml in TNE buffer [pH 7.4]) was mixed with 5 to 10 μl R18 solution for a final concentration of 20 μM, and the mixture was incubated in the dark for 2 h at room temperature. Nonincorporated R18 was removed by ultracentrifugation at 40,000 rpm for 3 h through a sucrose gradient of 20 to 60% (wt/vol) in a SW 41Ti rotor (Beckman Coulter). The visible virus band was collected at the 40%-60% sucrose interface, the collected virus was subjected to another round of purification via ultracentrifugation at 26,000 rpm for 2 h at 4°C through a 30% sucrose cushion in a SW41 rotor (Beckman Coulter), and 50-μl aliquots were stored at –80°C. All virus preparations were titrated on MARC-145 cells.

Membrane fusion assay.

R18-labeled PRRSV virions were incubated for 30 min at 4°C with PAMs (MOI of 0.1) in 150 mM NaCl, 5 mM HEPES (pH 7.4), followed by three washes of nonadsorbed viruses with RPMI 1640 medium. Mixtures of cells and virus in the cuvettes were then diluted to 2 ml in 150 mM NaCl, 5 mM HEPES (pH 7.4), and R18 fluorescence was continuously monitored with a Hitachi F-7000 spectrofluorometer (excitation wavelength, 560 nm; emission wavelength, 590 nm; slit widths, 5 and 10 nm, for excitation and emission, respectively). The cuvettes were stirred continuously with a magnetic stirrer, and temperature was controlled by a thermostatic circulating water bath at 37°C. Fusion was stopped by the addition of Triton X-100 at 1% (vol/vol [final concentration]) to elute R18 from the membranes, as a measure of 100% relaxation. The percentage of fluorescence dequenching (FDQ) at a given time was calculated according to the following equation: % FDQ = 100(F − F₀/F_t − F₀) (33, 63); F and F₀ are the fluorescence values at the given time and at time zero, respectively, and F_t represents the total fluorescence after the addition of Triton X-100 at the end of the assay.

For monitoring of the membrane fusion by confocal microscopy, primary PAMs seeded on coverslips in 24-well plates were infected with R18-labeled PRRSV, at a MOI of 1, for 30 min on ice. Unbound viruses were washed away, and RPMI 1640 medium supplemented with 2% FBS was added to the cells. Virus fusion was initiated by incubating the cells at 37°C. At the indicated times postinfection, the cells were fixed for 10 min at room temperature with 3.7% paraformaldehyde and then washed three times with PBS (5 min each wash). Nuclear DNA was stained with DAPI. The images were captured with a Leica confocal microscope and processed by using ImageJ.

Statistical analysis.

All statistical analyses were performed with GraphPad Prism v5.0 (GraphPad Software, Inc., La Jolla, CA, USA). Data were expressed as means ± standard deviations. Statistical significance was determined by two-way analysis of variance, and P values of <0.05 were considered statistically significant.