A Novel Motif in the 3′-UTR of PRRSV-2 Is Critical for Viral Multiplication and Contributes to Enhanced Replication Ability of Highly Pathogenic or L1 PRRSV

Highly pathogenic porcine reproductive and respiratory syndrome virus (HP-PRRSV) with enhanced replication capability emerged in China and has become dominant epidemic strain since 2006. Up to now, the replication-regulated genes of PRRSV have not been fully clarified. Here, by swapping the genes or elements between HP-PRRSV and classical PRRSV based on infectious clones, NSP1, NSP2, NSP7, NSP9 and 3′-UTR are found to contribute to the high replication efficiency of HP-PRRSV. Further study revealed that mutations at positions 117th or 119th in the 3′-UTR are significantly related to replication efficiency, and the nucleotide at position 120th is critical for viral rescue. The motif composed by 117–120th nucleotides was quite conservative within each lineage of PRRSV; mutations in the motif of HP-PRRSV and currently epidemic lineage 1 (L1) PRRSV showed higher synthesis ability of viral negative genomic RNA, suggesting that those mutations were beneficial for viral replication. RNA structure analysis revealed that this motif maybe involved into a pseudoknot in the 3′-UTR. The results discovered a novel motif, 117–120th nucleotide in the 3′-UTR, that is critical for replication of PRRSV-2, and mutations in the motif contribute to the enhanced replicative ability of HP-PRRSV or L1 PRRSV. Our findings will help to understand the molecular basis of PRRSV replication and find the potential factors resulting in an epidemic strain of PRRSV.

PRRSV is a member of the Arteriviridae family in the Nidovirales order (including Coronaviridae family, Arteriviridae family and Roniviridae family). Viruses in these families have

Viral Titration and Growth Kinetics
For viral titration, confluent Marc-145 cells seeded in 96-well plates were incubated with 10-fold serially diluted viral suspensions, and the viral titers were measured according to the method of Reed and Muench [44]. Briefly, the confluent Marc-145 cells seeded in 96well plates were incubated with 10-fold serially diluted viral suspensions. After absorption for 1 h at 37 • C, the supernatants were removed, and the fresh DMEM with 2% FBS was added. Following this, the plates were incubated for an additional six days, and viral titers were determined by the presence of visible cytopathic effect (CPE). Finally, the viral titers were calculated according to the method of Reed and Muench.
In order to investigate the replication difference of rescued mutant viruses and their parental viruses (rescued HuN4 or CH-1a), confluent Marc-145 cells or PAMs were incubated with the mutant and parental viruses at a multiplicity of infection (MOI) of 0.01. After inoculation for 1 h at 37 • C, the supernatants were discarded, and cells were washed with PBS three times. Following this, fresh DMEM with 2% FBS was added. At last, cell supernatants were harvested at 0, 24, 48, 72 and 96 h post-infection (hpi). The viral titers of different time points were measured by a microtitration assay using Marc-145 cells in 96-well plates and calculated as 50% tissue culture infective doses (TCID 50 ) per milliliter according to the method of Reed and Muench. Each time point was independently repeated three times.

Indirect Immunofluorescence Assay (IFA)
In order to further characterize chimeric or mutant viruses, Marc-145 cells were infected by those viruses and immobilized at 96 hpi; then, they were confirmed with IFA using a monoclonal antibody directed against the PRRSV N protein as described previously [45].

Viral Plaque Morphology Assays
Parental and mutant viruses were infected with Marc-145 cells in 6-well plates. After seven days, the viral plaques were observed after gentian violet staining. In brief, Marc-145 cells were infected with viruses at a dose of MOI = 0.01. After incubating for 1 h, the supernatants were discarded and washed three times with phosphate-buffered saline (PBS). Then, the pre-prepared DMEM medium containing 2% FBS and 1% low-melting agarose was added to the cells. Subsequently, plates were incubated at 37 • C for 5 to 7 days. Eventually, plaque morphology was observed after paraformaldehyde fixation and crystal violet staining. Ten homologous plaques were selected randomly (avoid selecting large plaques formed by fusion of plaques), for which their diameters were measured by a ruler, and finally analyzed by Graphpad Prism 6 (Graphpad, San Diego, CA, USA).

Luciferase Reporter Assay
3 -UTR has several mutant patterns in different PRRSV lineages or branches. In order to detect the transcription activity of 3 -UTR in different mutant patterns, a PRRSV minigenome system harboring luciferase reporter gene was employed according to a previous study [46]. The PRRSV mini-genome system was a kind gift from Dr. Yan-Dong Tang at the Harbin Veterinary Research Institute. Briefly, 3 -UTR was site-directed mutated according to different lineages and replaced the original 3 -UTR in the PRRSV mini-genome system. Followed by DNA sequencing, the mini-genome system plasmid (0.6 µg) and Renilla luciferase plasmid (0.02 µg) (Promega, Madison, WI, USA) were transfected into Marc-145 cells in 96-well plates. Renilla luciferase activity was used as an internal control for the normalization of luciferase values obtained from cells transfected with the firefly luciferase mini-genome system. At 24 h post-transfection, each well with Marc-145 cells was infected with PRRSV HuN4 at a dose of MOI = 0.1. Twenty-four hours later, the cells were lysed and assayed for firefly luciferase activity using a dual-luciferase reporter assay system (Promega, USA) according to the manufacturer's instruction.

Statistical Analysis
All experiments were performed at least three independent replicates. All data were analyzed by Graphpad Prism 6. The measured values were expressed as the mean with standard deviation (SD). If no virus was detected by plaque assay, the number representing the LOD (limit of detection) was used. Differences were analyzed for statistical significance using two-tailed unpaired t test for two groups or multiple comparison one-way variance (ANOVA) for more than two groups. Differences were considered statistically significant at a value of p < 0.05.

NSP1, NSP2, NSP7, NSP9 and 3 -UTR Are Closely Related to PRRSV Replication
In order to explore the key replication-related viral genes or elements, a series of chimeric viruses by swapping the corresponding genes or elements of HP-PRRSV HuN4 and classical PRRSV CH-1a was constructed based on full-length infectious clones ( Figure 1). The replication efficiency of chimeric viruses was compared with their parental viruses in porcine alveolar macrophages (PAMs). The results showed that HC 1 , HC 2 , HC 7 , HC 9 and HC 3 -UTR had significantly lower viral titers than parental HuN4 ( Figure 2A). Conversely, CH 7 and CH 9 had markedly higher viral titers than the parental CH-1a ( Figure 2B). The replication efficiency of other chimeric viruses was not different from their parental viruses ( Figure S1).    and CH-1a and HC9 and CH9, respectively. (φ, p < 0.05; φφ, p < 0.01; φφφ, p < 0.001). Delta (δ) indicates a significant difference between HuN4 and CH-1a and HC3′-UTR and CH3′-UTR, respectively (δ, p < 0.05; δδδ, p < 0.001).

Mutations in the Motif Enhance the Genomic Synthesis of L1 PRRSV
In order to investigate whether the mutations in the motif of L1 were related to viral replication, a PRRSV mini-genome system harboring the 3 -UTR and double luciferase report system was used to evaluate the synthesis level of viral negative genomic RNA, which is essential for genome replication and synthesis. A series of site-directed mutations was performed according to the nucleotides in the motif of different PRRSV lineages ( Figure 6A). Firefly luciferase activity was normalized with respect to a co-transfected plasmid encoding Renilla luciferase. The luciferase reporter assay showed that the 3 -UTR of HP-PRRSV HuN4 was 1.2 times than that of classical PRRSV CH-1a, with no significant difference ( Figure 6B). All different 3 -UTR of L1 PRRSV exhibited higher replication efficiency than that of classical PRRSV (p < 0.05) ( Figure 6B), suggesting the RNA synthesis level of L1 PRRSV is higher than that of classical PRRSV. was performed according to the nucleotides in the motif of different PRRSV lineages (Figure 6A). Firefly luciferase activity was normalized with respect to a co-transfected plasmid encoding Renilla luciferase. The luciferase reporter assay showed that the 3′-UTR of HP-PRRSV HuN4 was 1.2 times than that of classical PRRSV CH-1a, with no significant difference ( Figure 6B). All different 3′-UTR of L1 PRRSV exhibited higher replication efficiency than that of classical PRRSV (p < 0.05) ( Figure 6B), suggesting the RNA synthesis level of L1 PRRSV is higher than that of classical PRRSV.

The 117-120th Motif Are Loacted into a Pseudoknot of 3′-UTR
3′-UTR secondary structures of representative PRRSVs were predicted in the mFold software ( Figure 7A). Meanwhile, the 117-120th motif was located into a pseudoknot (Figure 7B) in the 3′-UTR, according to previous studies [23,43]. The rescued A117G or A119G mutant viruses showed obviously lower replication ability than that of HuN4, while

The 117-120th Motif Are Loacted into a Pseudoknot of 3 -UTR
3 -UTR secondary structures of representative PRRSVs were predicted in the mFold software ( Figure 7A). Meanwhile, the 117-120th motif was located into a pseudoknot ( Figure 7B) in the 3 -UTR, according to previous studies [23,43]. The rescued A117G or A119G mutant viruses showed obviously lower replication ability than that of HuN4, while G120A mutant viruses were not successfully rescued, indicating that the 117-120th motif in the 3 -UTR is critical for viral replication or rescue. The mutations at positions 117 and 119 did not change the overall conformation of 3 -UTR secondary structure, but the mutation at position 120 changed the overall secondary structure of 3 -UTR ( Figure 3B,C). Nucleotide mutations probably changed the minimum free energy of 3 -UTR secondary structure, which in turn results in alterations in the structure or stability of the pseudoknot.

Discussion
PRRS has caused severe economic losses to the pig industry since it was reported in 1987. The first PRRSV-2 strain, VR-2332 (belonging to L5), was isolated in the US in 1992. In China, the first PRRSV-2 strain, CH-1a (classical PRRSV, belonging to L8), was isolated in 1996. With the gradual evolution of PRRSV, HP-PPRSV (belonging to L8) emerged and

Discussion
PRRS has caused severe economic losses to the pig industry since it was reported in 1987. The first PRRSV-2 strain, VR-2332 (belonging to L5), was isolated in the US in 1992. In China, the first PRRSV-2 strain, CH-1a (classical PRRSV, belonging to L8), was isolated in 1996. With the gradual evolution of PRRSV, HP-PPRSV (belonging to L8) emerged and spread rapidly, and they quickly became the dominant epidemic strains [5]. HP-PRRSV was well documented, and it has higher replicative ability than classical PRRSV [32,33], which helps in understanding why HP-PRRSV spread rapidly. However, the key replicationregulated genes or elements still have not been elucidated.
Our study showed that NSP1, NSP2, NSP7, NSP9 and 3 -UTR are closely related to the replication efficiency of HP-PRRSV. Moreover, previous studies demonstrate that NSP1 has an effect on the synthesis of TNF-α and IFN-β [47,48]; NSP2, NSP3 and NSP5 interact with each other and participate in RTC formation and viral replication [19,49]; NSP7 interacts with NSP9, and NSP9 contributes to the replication ability and pathogenicity of PRRSV [33,50]. Meanwhile, the NSP2, NSP3, NSP5, NSP9 and NSP10 were reported to assemble in viral replication and transcription complex (RTC); then, RTC interacts with 3 -UTR, which initiates the transcription of sgmRNA and the replication of viral genome [19,20]. However, as a critical element of RTC, the function and key regions or sites of 3 -UTR in viral replication are still unclear. Therefore, a follow-up study on 3 -UTR was performed. By swapping the corresponding genes between CH-1a and HuN4, we demonstrated that the 3 -UTR contributes to the enhanced replicative ability of HP-PRRSV. Due to the low replication ability of CH-1a, the chimeric or mutant viruses based on the backbone of CH-1a were not replicated well in Marc-145 or PAM cells. Thus, the site-direct mutations of 3 -UTR were based on the HuN4 strain. Growth kinetics and plaque morphology assays show that the nucleotides at positions 117th and 119th in the 3 -UTR significantly affect viral replication or plaque sizes, suggesting that two nucleotides are critical sites for viral replication. Interestingly, the 120th nucleotide in the 3 -UTR is important for viral rescue, which may be because the 120th nucleotide and those nearby are involved in the formation of a senior RNA structure. Therefore, the 120th and its nearby nucleotides (117th to 119th) consist of a novel motif, 117-120th, which is a key motif related to PRRSV replication.
Alignment of all available entire sequences of PRRSV-2 in 1992-2019 revealed that the 117-120th nucleotides in the 3 -UTR were conservative in individual lineages. For each lineage, there were more or less mutations in the motif when compared with the prototype PRRSV-2, VR-2332 strain. The mutations contribute to the higher replication efficiency of HP-PRRSV HuN4 compared with the CH-1a strain. A highly efficient replication in host cells is one of the important features of dominant epidemic strain. There are many factors that result in HP-PRRSV becoming the dominant epidemic strain, such as increased viral tropism, high level of viral replication, moderate pathogenicity and so on [33,50,51].
In China, L1 PRRSV emerged in 2013, then it spread quickly in main pig-raising provinces since 2015, and L1 PRRSVs mainly include NADC30-like PRRSV and NADC34like PRRSV [6,31]. At present, HP-PRRSV and L1 PRRSVs are the dominant epidemic strains, and the proportion of L1 PRRSVs is higher than that of HP-PRRSV in some provinces [5,6,30]. There is a trend that the L1 PRRSVs surpass HP-PRRSV to become the most dominant epidemic strain in the next few years. In the present study, we found that mutations in the motif of 3 -UTR enhanced the replication of HP-PRRSV or L1 PRRSVs. At present, it is suitable for the most prevalent PRRSV lineages (L1/L3/L4/L5/L6/L7/L8/L9, 96.20%), except the FJ1402-like strain with deletion at positions 118-120th (L1, 3.80%). FJ1402-like strains with deletion at positions 118-120th in 3 -UTR have increased in recent years, and the detailed relationship between deletion and viral replication remains to be further explored.
Pseudoknot in the 3 -UTR of PRRSV can serve as pathogen-associated molecular patterns (PAMPs) and bind to the host cell pattern recognition receptors (PRRs) TLR3 and RIG-I to activate innate immune signaling and produced IFNs [43]. A previous study revealed a hairpin structure in the ORF7 together with the 3 -UTR form a kissing structure, which is essential for PRRSV replication [42]. In Mengovirus, three stem-loop (SL) are formed in its 3 -UTR, in which SL1 is a non-essential region for viral replication, and the absence of SL2 reduces the viral replication capability, and SL3 is necessary for viral replication [52]. In Dengue virus, SLs (A2 and A3) in the D2 region of 3 -UTR are essential for viral replication [53]. Consequently, we speculate that the 117-120th motif plays an important role in PRRSV replication. A pseudoknot is formed in the 3 -UTR according to the previous RNA structure analysis method [23], and the 117-120th motif was located in the pseudoknot.

Conclusions
In summary, our study has revealed that (i) NSP1, NSP2, NSP7, NSP9 and 3 -UTR are closely related to replication efficiency of PRRSV-2; (ii) the nucleotides at positions 117th and 119th in the 3 -UTR are critical for viral replication; (iii) the 117-120th is an important motif in the 3 -UTR of PRRSV-2, and mutations in the motif contribute to the higher replication level of HP-PRRSV or L1 PRRSVs. Our findings discovered the key motif or nucleotide involved in viral replication and would help to understand the mechanism of viral replication. The motif is potentially related to the formation of dominant epidemic of HP-PRRSV strains or L1 PRRSVs, which will require further investigation.