A total of 407 complete PRRSV genome sequences from China were downloaded from the GenBank database (Supplementary Table S1). The sequences were aligned using MAFFT (Version 4.470) (14), and conserved regions of the target lineage strains were selected for primer design. Specific primers for different types of strain were designed at the conserved sequences using primer 5.0 software. And the primer melting temperature (Tm), hairpin structures, and complementarity were predicted using Oligo (Version 7) software (15). All primers were synthesized by Beijing Tsingke Biotech Co., Ltd. (Beijing, China). The details of all primers have been listed in Table 1.

For the validation of method specificity, commercially obtained attenuated vaccines of Porcine Epidemic Diarrhea Virus (PEDV, strain AJ1102-R, sourced from Wuhan Keqian Biology Co., Ltd.), Porcine Transmissible Gastroenteritis Virus (TGEV, strain WH-1R, sourced from Wuhan Keqian Biology Co., Ltd.), Porcine Pseudorabies Virus (strain Bartha-K16, acquired from China Animal Husbandry Industry Co., Ltd.), and Porcine Circovirus Type 2 (PCV2, strain WH, supplied by China Animal Husbandry Industry Co., Ltd.) were employed.

Nucleic acids from PRRSV, PEDV, TGEV, PRV, and PCV2 were extracted using the RaPure Viral RNA/DNA Kit (Magen, Guangzhou, China) according to the manufacturer’s instructions. The HiScript® II 1st Strand cDNA Synthesis Kit (Vazyme, Nanjing, China) was then used to synthesize cDNA by reverse transcription. DNA and cDNA products were stored at −80°C.

To construct the recombinant standard plasmid, the target fragments of different PRRSV lineages were amplified by primer pairs 4 to 7 (2× Taq Plus Master Mix II, Vazyme, Nanjing, China). Next, according to the manufacturer’s instructions, the PCR products were purified using the Magen Gel Extraction Kit (Magen, Guangzhou, China). The purified products were cloned into the pMD18-T vector (Takara, Beijing, China) and transformed into DH5α competent cells (Vazyme, Nanjing, China). Then, the bacterial cultures were shaken and grown for 12 h at 37°C, and the final plasmid obtained was named PMD-EU, PMD-US, PMD-HP, and PMD-NL, respectively. The recombinant plasmid was purified using the HiPure Plasmid Micro Kit (Magen, Guangzhou, China) and the concentration was determined using NanoDrop spectrophotometer (Thermo Scientific, United States). Finally, the concentration was converted into a copy number using the following formula:

We further optimized the conditions for multiple qPCR, including annealing temperature and primer concentration. The standard plasmids were 10-fold serial diluted from 1 × 10⁶ copies/μL down to 1 × 10² copies/μL as amplification templates. The reaction system was 20 μL, including 10 μL of 2 × ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China), 2 μL of standard plasmid as a template, different final concentrations of primers, and nuclease-free water. A matrix approach was then used to explore the optimal reaction conditions for this multiple reaction: annealing temperatures of 50, 55, and 60°C; and final primer concentrations of 100 nM-350 nM at five dilutions. Amplification conditions were pre-denaturation at 95°C for 5 min, followed by 40 cycles of 95°C for 10 s, and annealing between 50°C and 60°C for 30s. Fluorescence signals were collected by a LightCycler® 96 thermal cycler Instrument (Roche Applied Science, Penzberg, Germany). The system was optimized by generating the lowest threshold cycle (Ct) and the highest cycle increase (ΔRn) for each specific fluorescent signal.

Based on the results of the amplification, the optimal annealing temperature was selected, and the proportions of primers numbered 4, 5, and 6, 7, were adjusted until a clear double-peak melting curve appeared and corresponded to good amplification efficiency. The primer proportions at this point were considered the optimal primer proportions.

The standard plasmids were continuously diluted tenfold from 1 × 10⁶ copies/μL to 1 × 10² copies/μL. Under optimal reaction conditions and systems, these five concentrations of standard plasmid were used as templates to construct the standard curve. The standard curve was plotted using GraphPad Prism software (version 8).

To evaluate the specificity of multiplex qPCR assay, DNA or cDNA of PEDV, TGEV, PRV and PCV2 extracted from the vaccine strains were used as templates, using standard plasmids PMD-HP, PMD-NL, PMD-EU, and PMD-US as positive controls and ddH₂O as a negative control, and the specificity was verified with qPCR assay under the optimal reaction condition and system. In order to determine the detection limit of the developed detection method in this study, the standard plasmids were continuously diluted tenfold from 10⁶ copies/μL to 10⁰ copies/μL. And qPCR amplification was performed in the optimum reaction conditions and system. Each concentration was tested in triplicate to eliminate differences caused by technical and operational factors.

Finally, in order to assess the reproducibility of the method, the four varieties of standard plasmids (10¹ copies/μL to 10⁶ copies/μL) were used as templates to evaluate intra- and inter-assay reproducibility. For intra-assay reproducibility, each dilution was replicated three times daily under identical conditions. For inter-assay reproducibility, according to MIQE guidelines (16), each dilution was tested in six independent experiments performed by two operators on different days. Coefficients of variation of the Ct values were calculated based on the intra-assay or inter-assay results.

To verify the detection performance of this method, we compared and validated the detection method established in this study with commercial detection kits. Differential diagnosis of PRRSV was performed using the commercial kit VetMAX PRRSV EU&NA 2.0 Kit (Thermo, United States, A35751) according to the manufacturer’s instructions.

For sensitivity comparison experiments, four standard plasmids (PMD-EU, PMD-US, PMD-HP, and PMD-NL) (10¹ copies/μL to 10³ copies/μL) were used as templates for evaluation (n = 24 for each plasmid and concentration, and n = 24 for negative controls). For specificity comparative experiments, nucleic acids from PRRSV, PEDV, TGEV, PRV, and PCV2 were used as templates for evaluation (n = 24 for each virus, and n = 24 for negative controls). For reliability comparison tests, we evaluated the nucleic acids of PRRSV-1 (n = 15), PRRSV-2 (n = 15), HP-PRRSV (n = 15), NL-PRRSV (n = 15) positive samples, and PRRSV negative samples (n = 15) previously stored in our laboratory. It should be noted that because of the low prevalence of PRRSV-1 in mainland China in recent years, we divided the only PRRSV-1 nucleic acid sample into 15 parts to expand the number of positive samples.

Finally, we calculated the agreement rate between our detection method and commercial reagent kits (the total number of true positives and true negatives/the sum of true positives, true negatives, false positives, and false negatives) to demonstrate the detection performance of our constructed method.

Between 2021 and 2023, a total of 316 clinical samples, including 139 tissues samples, 78 blood samples and 99 serum samples, were collected from swine farms in 14 cities in Guangdong Province, China (Figure 1). The detailed information of the samples was listed in Supplementary Table S2. These samples were subsequently analyzed to determine the infection positivity rate of different lineages using the established multiplex qPCR assay.

Next, the complete ORF5 gene sequences of representative positive samples were amplified using primer pairs 1, 2 and 3 from Table 1. The reaction system was 50 μL, containing 25 μL of 2 × KeyPo SE Master Mix (Dye Plus) (Vazyme, Nanjing, China), 2 μL of template cDNA, 10 μM forward and reverse primers with 2 μL each, and the remaining was added to ddH₂O. The PCR reaction for ORF5 gene was executed by pre-denaturation at 94°C for 2 min, followed by 35 cycles of 98°C for 10 s, 55°C (other PRRSV strains) or 45°C (NL-PRRSV) for 30 s, and 68°C for 15 s. The PCR products were then purified using the Magen Gel DNA Recovery Kit (Magen, Guangzhou, China) according the manufacturer’s instructions. The purified PCR products were cloned using a pGM-T cloning kit (Tiangen, Beijing, China), and propagated in DH5α competent cells (Vazyme, Nanjing, China) according to the manufacturer’s instructions. Positive clones were sequenced by Beijing Tsingke Biotech Co., Ltd. The obtained complete ORF5 gene sequences were edited and assembled using DNAstar V7.1 software.

In order to further investigate the genetic characteristics of PRRSV in the surveyed regions of Guangdong Province, a total of 13 complete ORF5 gene sequences of representative PRRSV strains were obtained in this study and uploaded to NCBI GenBank with accession numbers OR539223-OR539235. The complete ORF5 gene sequences of PRRSV were aligned with the relevant reference sequences in GenBank using MAFFT version 7.487 with the parameter L-INS-I. Phylogenetic trees were generated by the neighbor-joining (NJ) method in MEGA X with the bootstrap of 1,000 replicates. Details of the reference sequence were listed in Supplementary Table S3.

In the specificity and repeatability tests, we conducted three tests on each sample, and the final results were expressed as mean ± standard deviation. In the repeatability test, we determined the differences within and between groups by calculating the coefficient of variation (CV). In addition, the data and statistical analyses of PRRSV prevalence were estimated from the ratio of positive samples to the total number of samples analyzed, with a binomial confidence interval of 95%. The data were analyzed using Microsoft Excel 2019 and SPSS version 27.0.

Annealing temperatures of 50, 55, and 60°C, and primer final concentrations of 100–350 nM were selected for the two types of duplex qPCR system optimization. According to the optimization results, the annealing temperature was determined to be 55°C for NL-PRRSV and HP-PRRSV (Supplementary Table S4), 50°C for PRRSV-1 and PRRSV-2 (Supplementary Table S5). The final primer concentrations for NL-PRRSV, HP-PRRSV, PRRSV-1, and PRRSV-2 were 250, 250, 200, and 300 nM, respectively (Table 2).

After optimization, the two types of duplex qPCR enable simultaneous detection of all target nucleic acids. Consequently, the finalized amplification systems for the two types of duplex-qPCRs were established as follows: for NL-PRRSV and HP-PRRSV, 10 μL of 2 × ChamQ Universal SYBR qPCR Master Mix, 250 nM HP-PRRSV primer, and 250 nM NL-PRRSV primer, 2 μL of template, the remaining volume was made up to 20 μL with ddH₂O. For PRRSV-1 and PRRSV-2, 10 μL of 2 × ChamQ Universal SYBR qPCR Master Mix, 200 nM PRRSV-1 primer, and 300 nM PRRSV-2 primer, 2 μL of template, the remaining volume was made up to 20 μL with ddH₂O.

The optimal reaction conditions were pre-denaturation at 95°C for 10 s, followed by 40 cycles of 95°C for 10 s and 50°C (PRRSV-1 and PRRSV-2) or 55°C (NL-PRRSV and HP-PRRSV) for 30 s. The fluorescence signals were captured with a real-time qPCR instrument from Roche for the entire duration of the study.

According to the calculations by oligo 7.0 software, the expected Tm values of the primer pairs used for differential diagnosis in this study were 79°C (PRRSV-1), 84°C (PRRSV-2), 84°C (NL-PRRSV), and 88°C (HP-PRRSV), respectively. These values were used to distinguish PRRSV-1, PRRSV-2, NL-PRRSV, and HP-PRRSV. It is worth noting that in actual testing, there is a deviation between the Tm values and the expected values. The actual Tm values (mean ± SD) were as follows: 79.71 ± 0.23°C (PRRSV-1), 84.43 ± 0.12°C (PRRSV-2), 84.34 ± 0.26°C (NL-PRRSV), and 87.86 ± 0.13°C (HP-PRRSV), respectively.

The recombinant standard plasmid, diluted to 1 × 10⁶ to 1 × 10² copies/μL in a 10-fold gradient, was amplified using the two types of duplex qPCR assay according to the optimal reaction system and reaction procedure. The standard curve was established using obtained Ct values as the y axis coordinates and the logarithm of the plasmid concentration as the x axis coordinates. All standard curves had good correlation coefficients and amplification efficiencies. The slopes, correlation coefficient (R²), and amplification efficiency (Eff%) were as follows: NL-PRRSV (−3.381, R² = 0.9987, and Eff% = 97.59), HP-PRRSV (−3.330, R² = 0.9998, and Eff% = 99.66), PRRSV-1 (−3.377, R² = 0.9997, and Eff% = 97.75), and PRRSV-2 (−3.515, R² = 0.9996, and Eff% = 92.53) (Figure 2).

To assess the sensitivity of our developed qPCR assay method, the two types of duplex qPCR were performed using standard plasmids with concentrations ranging from 1 × 10⁶ to 1 × 10⁰ copies/μL. The results showed that the established detection limit for PRRSV-1 was determined to be 10² copies/μL (Figure 3C), and the detection limits for NL-PRRSV (Figure 3A), HP-PRRSV (Figure 3B), and PRRSV-2 (Figure 3D) were ascertained to be 10¹ copies/μL, suggesting that our duplex qPCR methods have a good sensitivity.

Furthermore, clear distinctions between HP-PRRSV (Figure 3E) and NL-PRRSV (Figure 3F), as well as between PRRSV-1 (Figure 3G) and PRRSV-2 (Figure 3H), were observable based on the derived melting curves. In conclusion, we suggested that both duplex SYBR Green qPCR assay methods demonstrated excellent detection and discrimination capabilities.

To further assess the specificity of the duplex qPCR assay methods, the recombinant standard plasmids PMD-EU, PMD-US, PMD-HP, and PMD-NL were employed as positive controls. DNA or cDNA extracted from PEDV, TGEV, PRV and PCV2 vaccines were used as templates, and ddH₂O was used as a negative control. SYBR Green qPCR was performed using the optimal conditions of amplification. The outcomes demonstrated that this technique specifically identified the nucleic acids of PRRSV-1, PRRSV-2, HP-PRRSV, and NL-PRRSV exclusively in the positive control plasmids (Figure 4), with no amplification detected for the other viral pathogens (Table 3). These results suggest that our identification assays are highly specific.

The repeatability and reproducibility of the constructed duplex SYBR Green qPCR assay were assessed using the recombinant standard plasmids, which were diluted in a 10-fold gradient from 10⁶ copies/μL to 10¹ copies/μL as templates. As shown in Table 4, the results revealed that the coefficient of variation (CV) for both intra-assay and inter-assay replicates of the CT values ranged from 0.02 to 2.12% and 0.02 to 2.28%, respectively. These results indicate that our qPCR assay established in this study has good reproducibility and reliability.

By comparing with commercially available reagent kits, this study further validated the detection performance of the established duplex qPCR method, including sensitivity, specificity, and effectiveness. The results showed that when the nucleic acid content of PRRSV-1 was extremely low (≤10¹ copies/μL), the detection rate was lower than that of commercially available kits (Table 5). However, when the nucleic acid concentration is ≥10² copies/μL, the detection rate of PRRSV-1 is basically consistent with the commercially available kit, with a conformity rate of 97.9–100% (Table 5). For PRRSV-2, HP-PRRSV, and NL-PRRSV, even at a viral nucleic acid concentration of 10¹ copies/μL, the agreement rate between this method and the detection results of commercially available kits is 93.8–95.8% (Table 5). In terms of specificity, our method is basically consistent with the detection results of the kit, with a conformity rate of 95.8–100% (Supplementary Table S6). In addition, in terms of effectiveness, the consistency rate between this differential diagnostic method and the kit is 93.3–100% (Table 6). In summary, the duplex qPCR method established by our research institute for differential diagnosis of PRRSV has high consistency with the detection results of commercially available kits.

Three hundred and sixteen samples collected from different swine farms in Guangdong Province during 2021–2023 were tested using the method established in this study, including serum, blood, and clinical tissue samples (Table 7). Among the 316 samples, for blood samples, the PRRSV-2 positive rate was 11.54% (9/78), and all positive samples belonged to HP-PRRSV (Table 7). For tissue samples, the positive rate for PRRSV-2 was 45.32% (63/139), with HP-PRRSV and NL-PRRSV being 28.78% (40/139) and 2.16% (3/139), respectively (Table 7). On the contrary, positive samples were not detected in all serum samples.

In addition, the PRRSV-1 positive rate was 0% (0/316), and the PRRSV-2 positive rate was 22.78% (72/316) (Table 8). Among all surveyed areas, the prevalence of PRRSV-2 in Yunfu was the highest, reaching 78.43% (40/51), followed by Heyuan (16.25%, 13/80) (Table 8). It is worth noting that our results also indicated a positive rate of 15.51% (49/316) for highly pathogenic PRRSV, while 0.95% (3/316) for NADC30-like PRRSV (Table 8). Moreover, Yunfu is also the city with the most severe HP-PRRSV epidemic, followed by Shanwei (Table 8). However, during this investigation, no co-infection of different strains was found.

To further investigate the genetic characteristics of PRRSV in the surveyed regions of Guangdong Province, representative samples (n = 13) from each region were selected for amplification and sequencing of the complete ORF5 fragment. A total of 13 complete ORF5 gene sequences of PRRSV were successfully obtained, each of which was 603 nt in size, and have been archived in the GenBank database, with accession numbers OR539223 through OR539235.

Phylogenetic analysis revealed that the strains collected from Guangdong belonged to four lineages (1, 3, 5, and 8), with lineage 8 being the dominant strain, particularly sub-lineage 8.3 (Figure 5). Further analysis showed that highly pathogenic PRRSV was mainly prevalent in the Yunfu and Shanwei regions, while classical PRRSV was predominantly observed in the Yangjiang region. Additionally, lineage 3 strains were predominant in the Heyuan region. Notably, PRRSV circulating in the pig population in Maoming exhibited high genetic diversity, belonging to three different lineages: 1, 3, and 5 (Figure 5).

Nucleotide sequence analysis revealed that the HP-PRRSV strains shared 98.7–99.3% complete nucleotide identity with highly pathogenic strain JXA1. However, the classical strain from this study shared 99.2% nucleotide identity with strain CH-1a. In addition, the strains of lineage 1, 3 and 5 shared 90.2–91.7%, 90.9–92.9 and 99.8% nucleotide identity with the reference strains of same lineage, respectively.