Genetic diversity of porcine reproductive and respiratory syndrome virus and evaluation of three one-step real-time RT-PCR assays in Korea

Background Porcine reproductive and respiratory syndrome virus (PRRSV) has caused huge economic losses in the global swine industry. Frequent genetic variations in this virus cause difficulties in controlling and accurately diagnosing PRRSV. Methods In this study, we investigated the genetic characteristics of PRRSV-1 and PRRSV-2 circulating in Korea from January 2018 to September 2021 and evaluated three one-step real-time reverse transcription polymerase chain reaction (RT-PCR) assays. Results A total of 129 lung samples were collected, consisting of 47 samples for PRRSV-1, 62 samples for PRRSV-2, and 20 PRRSV-negative samples. Nucleotide sequence analysis of open reading frames (ORFs) 5, ORF6, and ORF7 genes from PRRSV samples showed that PRRSV-1 belonged to subgroup A (43/47, 91.49%) and subgroup C (4/47, 8.51%), whereas PRRSV-2 was classified as lineage 1 (25/62, 40.32%), Korean lineage (Kor) C (13/62, 20.97%), Kor B (10/62, 16.13%), lineage 5 (9/62, 14.52%), and Kor A (5/62, 8.06%). Amino acid sequence analysis showed that the neutralizing epitope and T cell epitope of PRRSV-1, and the decoy epitope region and hypervariable regions of PRRSV-2 had evolved under positive selection pressure. In particular, the key amino acid substitutions were found at positions 102 and 104 of glycoprotein 5 (GP5) in some PRRSV-2, and at positions 10 and 70 of membrane protein (M) in most PRRSV-2. In addition, one-step real-time RT-PCR assays, comprising two commercial tests and one test recommended by the World Organization for Animal Health (OIE), were evaluated. Conclusion The results revealed that two of the real-time RT-PCR assays had high sensitivities and specificities, whereas the real-time RT-PCR assay of the OIE had low sensitivity due to mismatches between nucleotides of Korean PRRSVs and forward primers. In this study, we genetically characterized recent PRRSV occurrences and evaluated three one-step real-time RT-PCR assays used in Korea. Supplementary Information The online version contains supplementary material available at 10.1186/s12917-022-03407-0.

The ORF5 sequence of PRRSV has been widely used to study phylogeny, genetic variation, and molecular epidemiology [15]. Many PRRSVs that were genetically and geographically differentiated, were classified into subtypes 1-4 from PRRSV-1 and lineages 1-9 from PRRSV-2 [3,16]. ORF6 encodes the most conserved structural protein of PRRSV. The phylogenetic tree derived from ORF7 resembles the tree derived from the full-length genomes of PRRSV [17]. The conserved regions of ORF6 and ORF7 are often used as target regions for PRRSV detection by nucleotide-based assays [18][19][20][21][22].
ORF5 encodes a highly variable envelope protein, GP5, which plays an important role in viral infectivity and contains immunological domains related to viral neutralization [23,24]. GP5 and M protein, two major envelope proteins, form a disulfide-linked heterodimer or a disulfide-linked multimer that is essential for virion formation [25,26]. GP4, GP5, and M proteins induce neutralizing murine monoclonal antibodies (MAbs). In particular, MAbs recognizing GP5 neutralize PRRSV more effectively than other MAbs [27]. Therefore, GP5 has been considered a major target protein for vaccine design as it is involved in the production of neutralizing antibodies, followed by protection against PRRSV [28]. The non-neutralizing epitope of PRRSV is highly immunodominant and exhibits some features of decoy epitopes, which have been demonstrated to inhibit recognition of neutralizing epitopes in several viral infections [28].
Nucleic acid-based diagnostic methods have been commonly used to diagnose PRRSV owing to their sensitivity, specificity, and relatively rapid test times [29][30][31]. However, RNA viruses, such as PRRSV and swine influenza virus (SIV), have high mutation rates, rapid evolution, and genetic variability, these complicate the development of reliable diagnostic methods [32][33][34][35]. Many studies have shown that genetic differences or mismatches between nucleotides of PRRSV and the primers in molecular-based assays can lead to false results [36][37][38][39]. Therefore, the continuously increasing genetic diversity of PRRSV with the emergence of new strains dictates the need for an accurate diagnosis.
In this study, we investigated the genetic diversity of PRRSVs circulating in Korea through phylogenetic analysis and amino acid analysis from January 2018 to September 2021 and evaluated three one-step real-time reverse transcription polymerase chain reaction (RT-PCR) assays used in Korea.

Clinical sample and detection of PRRSV
A total of 129 lung samples submitted to the Diagnostic Division of the Animal and Plant Quarantine Agency (APQA) to diagnose swine disease, were collected from January 2018 to September 2021. The samples were collected from farms across all Korean provinces and included mainly clinical signs of PRRS such as acute respiratory disease in growing pigs or late-term abortion in sows. All lung tissue samples were homogenized with alpha modification of Eagle's minimum essential medium (EMEM) (Gibco, Grand Island, NY, USA) containing 1% antibiotic (Gibco). Viral RNA was extracted from the supernatant of tissue homogenates using the Keywords: Porcine reproductive and respiratory syndrome virus (PRRSV), Genetic diversity, Evaluation, Diagnostic method RNeasy Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Commercial VDX ® PRRSV HP MP RT-PCR and NA/EU Typing Nested PCR (Median Diagnostics, Gangwon, South Korea) were used for detecting and genotyping PRRSV.
Sequencing and phylogenetic analysis of ORF5, ORF6, and ORF7 of PRRSV ORF5, ORF6, and ORF7 were amplified from PRRSVpositive samples by RT-PCR with specific primer sets ( Table 1). RT-PCR amplification was performed under the following conditions: reverse transcription for 30 min at 50 °C, and termination of reverse transcription for 15 min at 95 °C, followed by 35 cycles of denaturation, annealing, and extension for 30 s at 94 °C, 30 s at 55 °C, 50 s at 72 °C, respectively, and a final extension of 10 min at 72 °C. The PCR products were sequenced by commercial sequencing service company (Macrogen, Daejeon, South Korea). Some ORF5, ORF6, and ORF7 genome sequences among the all sequences obtained in this study were submitted to GenBank under accession number ON892744-ON892781. Reference strains, such as global PRRSV strains, Korean PRRSV strains, and two PRRSV prototype strains (Lelystad and VR2332) obtained from National Center for Biotechnology Information (NCBI), were included in the dataset for phylogenetic analysis. A phylogenetic tree was generated by maximum likelihood analysis using the Kimura two-parameter model (K2P) with MEGA 7.0 (Pennsylvania State University, State College, Pennsylvania, USA), and was evaluated by 1,000 bootstrap replicates. The Markov Chain Monte Carlo (MCMC) algorithms implemented in the BEAST v1.7.5 package was used to estimate the substitution rate per site per years (s/s/y) of the Korean PRRSV strain from 1997 to 2021. The dataset consisted of a total of 238 PRRSV-1 ORF5 sequences and 319 PRRSV-2 ORF5 sequences including Korean PRRSV reference strains available in NCBI and ORF5 sequences obtained in this study. Evolutionary rate was estimated using the relaxed molecular clock model with GTR + Γ4 mixed substitution according to a previous study [40].

Amino acid analysis of GP5 and M protein
Amino acid analysis between lineages and sequence entropy at each codon indicating amino acid diversity was conducted according to a previous study [41]. Briefly, graphical sequence logos for each lineage were generated using the WebLogo tool (http:// weblo go. berke ley. edu/) and sequence entropy was generated using the Shannon Entropy-One tool implemented in the HIV database tool (https:// www. hiv. lanl. gov/). To determine the action of selection pressure on the structural proteins of PRRSV-1 and PRRSV-2, site-by-site selection at single codon sites of each structural protein was estimated using the mixedeffects maximum likelihood model of evolution available at DataMonkey (http:// www. datam onkey. org/) [42,43]. Sites with a p-value ≤ 0.05 were inferred to be positively selected.

One-step real-time RT-PCR assays
Two commercially available certified one-step real-time RT-PCR assays (A and B tests) are the most commonly used for detecting PRRSV in Korea. Test A and B were performed using each kit provided under the same lot number. The one-step real-time RT-PCR (C test) is recommended in the OIE manual of diagnostic tests [44] and is used by a private animal disease diagnostic center. The C test was performed using the QuantiNova Probe RT-PCR kit (QIAGEN). The reaction mix was prepared using 10 μL of 2 × probe RT-PCR Master Mix, 0.2 μL of QN Probe RT-Mix, 3 μL of the final concentration of forward/reverse primer and probe mix according to a previously described protocol [45], 2 μL of RNase-free water, and 5 μL of RNA. Reactions were performed according to the manufacturer's instructions.

Amino acid analysis of GP5 and M protein from PRRSV-1
Previous studies identified one neutralizing epitope at amino acid (aa) 29-35 in GP5, which was reported to be 29 WSFADGN 35 in the Lelystad strain. Additionally, there were T cell epitopes and four B cell epitopes (GP5-I, GP5-II, GP5-III, and M-I) in GP5 and M protein of PRRSV-1 [48][49][50][51]. As shown in Fig. 3, the neutralizing epitope was conserved with a low level of entropy. Subgroup A did not show much variation in the neutralizing epitope region, while only the 20R44-37-1 sample belonging to subgroup C showed variation in position 31 ( 31 F → 31 S) and 35 ( 35 N → 35 S). By contrast, the GP5-III and M-I regions were variable with a high level of entropy, indicating genetic diversity (Fig. 3). A total of 16 codon sites were positively selected in the GP5 and M proteins of PRRSV-1 (Table 2) Fig. 4A, critical amino acid variations in the B cell and T cell epitopes were also found in GP5 of PRRSV-2. In the decoy epitope compared with the VR2332 strain, significant diversity was found with higher amino acid entropy. Interestingly, a specific substitution at position 44 ( 44 N → 44 K) was found in the 21R2-15-1  (Table 2).

Sensitivities and specificities of three one-step real-time RT-PCR assays
The sensitivities of three one-step real-time RT-PCR assays with two reference strains (Lelystad and LMY strain) were estimated to be 10 0 TCID 50 /100μL. In the specificities of all one-step real-time RT-PCR assays determined by testing respiratory disease-causing viruses, such as PCV2, PCV3, CSFV, PPV, and SIV, no cross-reactivity was observed (data not shown).

Clinical evaluation of three one-step real-time RT-PCR assays
Twenty PRRSV-negative clinical samples showed the same results in all one-step real-time RT-PCR assays. As shown in Table 3 Fig. 1B).

Discussion
After 30 years of PRRSV emergence, PRRSV infection remains a critical disease that causes enormous economic losses to the swine industry worldwide. Despite widespread efforts to control and prevent PRRSV infection, the virus has rapidly spread worldwide and has increasing genetic diversity [55]. The evolutionary rate of PRRSV (4.71-9.8 × 10 -2 /sites/year) is the highest among RNA viruses [56], which allows genetic diversity within PRRSV and the emergence of new phenotypes. In this study, a phylogenetic analysis of Korean PRRSV was performed using clinical samples collected   [41]. These changes may support the hypothesis for these epidemic situations such as the importing of breeding pigs and artificial insemination [57,58].  [11,14]. In this study, the lowest nucleotide sequence homology of ORF5 between PRRSV-1 and the Lelystad strain was 79.74%, indicating a decrease of 6.06% over approximately 15 years. The lowest homologies of ORF6 and ORF7 among the different PRRSV-1 isolates were 85.85% and 86.52%, respectively, indicating decreases of 7.35% and 2.28%, respectively. The ORF5 sequence homology between Korean PRRSV-2 and VR2332 was 84.7-99.5% in 2003-2010 and 82.3-99.3% in 2013-2016 [13]. The ORF6 nucleotide sequence identity among the Korean PRRSV-2 strains was 85.5-98.2% [60]. The ORF7 of Korean PRRSV-2 showed a sequence homology of 86.2-100.0% with each other and 88.3-100% with isolates from other geographic regions [61]. In this study, the lowest nucleotide sequence homology of ORF5 between PRRSV-2 and VR2332 was 78.71%, indicating a decrease of 5.99% in over approximately 20 years. The lowest homologies of ORF6 and ORF7 among the different PRRSV-2 isolates is 83.21% and 79.66%, respectively, indicating decreases of 2.29% and 6.54%, respectively. Previous study showed nucleotide substitution rates of 1.46 × 10 -3 for PRRSV-2 viruses and 3.29 × 10 -3 substitutions/site/year for two genotype isolates based on ORF5 sequences data [16,62] and 4.17-9.8 × 10 -2 substitutions/site/year based on ORFs 3-5 sequences of two genotype PRRSVs [56]. Consistent with theses previous investigations, our results also indicate that Korean PRRSV has high substitution rates of 5.8521-7.9748 × 10 -3 for Korean PRRSV-1 and 4.516-5.7664 × 10 -3 for Korean PRRSV-2. Therefore, it is suggested that the mutation rate of PRRSV circulating in Korea has increased over time.
In this study, the neutralizing epitope of PRRSV-1 was found to be conserved, but the GP5-III and M-I regions of PRRSV-1 were variable. Recent studies on the amino acid analysis of GP5 of Korean PRRSV-1 also showed a relatively conserved pattern in the B cell epitope regions, except for GP5-III (aa 165-176) [13,41,63]. GP5 and M of PRRSV-1 are not susceptible to antibody-mediated virus neutralization, in contrast to GP5 of PRRSV-2 is generally considered the main target for virus-neutralizing antibodies [50]. To understand the mechanism of neutralizing antibody against the Korean PRRSV-1, further analysis of the neutralizing antibody-escape mutants of PRRSV in other minor envelope glycoproteins such as GP2, GP3, and GP4 is required. The residue at position 44 of PRRSV-2 GP5 plays a critical role in virus infectivity; position 50 of GP5 and position 8 of the M protein are essential for assembly of PRRSV particles [64,65]. In this study of PRRSV-2, the key residues at position 44 of GP5 and position 8 of M protein were variable in some samples. The key residues at positions 102 and 104 of PRRSV-2 GP5, which determine susceptibility to viral neutralization, were variable [66]. Residues (V102C and G104R) that were identical to those in a neutralizing antibody-escape mutant and with higher amino acid entropy were found in our several samples under positive selective pressure. In addition, amino acid mutations at positions 10 and 70 of PRRSV-2 M protein are related to susceptibility to viral neutralization [53,54]; these features were validated in our samples. The hypervariable regions can modulate the accessibility of neutralizing antibodies to the neutralizing epitope [67]. Recently, variability in the decoy epitope region and hypervariable regions of GP5, as well as mutations in key residues related to neutralizing antibody-escape mutants, have been commonly found in Korean PRRSV-2 amino acid analysis studies [13,41,60]. These variations were also found among the PRRSV-2 samples in this study. Therefore, Korean PRRSV-2 has evolved to genetic variants with resistance to neutralization and may be able to escape neutralization by antibodies that are induced by commercial PRRS modified live vaccines (MLV).
A positive selection signal was detected in the neutralizing epitope region of PRRSV-1. Incidentally, sequence analysis of PRRSV-2 revealed variation and positive selection pressure within the decoy epitope and the neutralizing epitope regions. Recent research demonstrated that vaccination resulted in the emergence of antibodyescaping mutants, in which strong positive selection