In 2024, pigs (Duroc) on a farm in Sichuan province experienced abortion and stillbirths of pregnant sows, with abortion rate of ~23% (23/98). The affected 310 piglets exhibited severe clinical respiratory illness, with approximately 90% morbidity (279/310) and 50% mortality (155/310). There are no neighboring pig farms within 10 km and no similar production performance was reported around this pig farm. The sows were vaccinated with the PRRS modified-live virus (MLV) vaccine (Ingelvac® PRRS MLV) before breeding, and the piglets were vaccinated at 21 days after birth. A total of 20 lung samples of the diseased piglets were collected and sent for PRRSV detection.

Tissue samples were homogenized and diluted with RPMI-1640 medium, the supernatant was collected after centrifugation at 10000 × g for 10 min at 4°C to remove residual tissue debris. The remaining lung tissues were stored in a refrigerator at −80°C. The lung samples were processed to detect PRRSV by reverse transcription polymerase chain reaction (RT-PCR). Viral RNA was extracted from the supernatants of each lung homogenate using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). The positive lung samples were used to isolate viruses from Marc-145 cells and immortalized porcine alveolar macrophages (iPAMs). The cells were maintained in RPMI-1640 medium supplemented with 10% fetal bovine serum (VivaCell, Shanghai, China) at 37°C in a humidified 5% CO₂ atmosphere as described previously. Cell cultures were monitored every day for a cytopathic effect (CPE) (11). Indirect immunofluorescence assay (IFA) was performed to identify PRRSV using a specific monoclonal antibody targeting the PRRSV N protein (GeneTex Inc., Irvine, CA, USA). The virus was further purified by plaque assay as previously described (20). The complete genome of the PRRSV isolate was further amplified using 13 primer pairs (Supplementary Table S1). The PCR products were subjected to the commercial service (Sangon, Shanghai, China) using the Sanger sequencing, and the SeqMan program in Lasergene software (DNASTAR, Madison, WI, USA) was used to assemble the full-length genomic sequences. The genomic sequences of the PRRSV isolate were designed as GX2024 and have been submitted to the GenBank database (GenBank Accession No: PV362838).

The genomic sequence of GX2024 was analyzed using MEGA-X (v10.1.8) and compared with PRRSV-2 reference sequences (partial or complete) available in GenBank. Multiple sequence alignment was carried out using the MegAlign program in Lasergene software version 7.1.0 (DNASTAR, USA). The phylogenetic analysis was conducted by using the maximum likelihood (ML) method using MEGA-X with the nucleotide substitution model GTR + G + I. Bootstrap values were calculated for 1,000 replicates of the alignment. Sequences were aligned using the ClustalW multiple alignment algorithm. The genetic lineage classification of PRRSV-2 strains follows the latest classification system described by Yim-im et al. (8).

The recombination pattern of GX2024 was analyzed using the Simplot software (version 3.5.1) and the recombination detection program 4 (RDP4, version 4.96), as previously described (3). Specifically, three representative PRRSV-2 sequences of different lineages, including NADC30 (L1C), RespPRRS MLV (L5), JXA1 (L8E), and GX2024, were aligned using the MEGA-X software. The identification of possible recombination events within the genome of GX2024 were performed using seven methods, including RDP, BOOTSCAN, MAXCHI, CHIMAERA, 3SEQ, GENECONV, and SISCAN, with default settings in RDP4. Recombination events were considered significant (p-value < 1 × 10⁻⁶) when supported by at least five of the seven detection methods.

An animal challenge experiment was performed to evaluate the pathogenicity of the GX2024 isolate. Nine 4-week-old healthy Duroc piglets were purchased from a commercial pig farm with no PRRSV vaccination program. The piglets were confirmed to be negative for PRRSV, African swine fever virus (ASFV), classical swine fever virus (CSFV), porcine epidemic diarrhea virus (PEDV), pseudorabies virus (PRV) or porcine circovirus type 2 (PCV2) by PCR or RT-PCR, and determined to be free of antibodies to PRRSV by a commercial PRRSV antibody ELISA kit (IDEXX HerdChek ELISA, USA). All animals were randomly divided into two groups and maintained in individual biosafety rooms. The piglets in Group 1 (n = 5) were each intranasally (2 mL) and intramuscularly (1 mL) inoculated with 1 × 10^5.58 TCID₅₀/ml GX2024 isolate (the sixth passage in Marc-145) while the remaining piglets in Group 2 (n = 4) were mock infected with uninfected RPMI-1640 medium with the same dose. The experiments were approved by the Animal Ethics Committee of the College of Animal & Veterinary Sciences, Southwest Minzu University (registration protocol SMU-202401058).

After the PRRSV GX2024 inoculation, each piglet’s rectal temperature and clinical symptoms were monitored and scored daily (17). The scoring included the gross clinical, respiratory, and neurological symptom scores. The total score for each piglet represents the sum of the three data points above. Blood was collected from individual pigs periodically at 0, 3, 5, 7, 10, and 14 days post-inoculation (dpi). All the animals were humanely euthanized either at 14 dpi, and carcasses were subjected to harmless treatment. Piglets were weighed weekly after viral infection. In addition, tissue samples, including lungs and nodes were collected. Portions of tissue samples were immersed in 4% paraformaldehyde for histopathological and immunohistochemical (IHC) examinations as previously described (21). The ORF5 gene from the lungs was amplified and sequenced to confirm that it was the original virus.

Viremia and viral loads in tissues were quantified by reverse-transcription quantitative PCR (RT-qPCR) assay using a primer targeting the ORF7 of PRRSV-2 (Supplementary Table S2). A standard curve generated using a serially 10-fold diluted (10⁹ to 10³ copies/μL) plasmid containing the ORF7 was used for absolute quantification. Approximately (0.5 g) of tissue was homogenized for the tissue samples with 1 mL of RPMI-1640 medium. The supernatant was collected after centrifugation at 10000 × g for 10 min at 4°C to remove residual tissue debris (22). For each serum sample, 300 μL supernatant was used for viral RNA extraction using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). A reverse transcription kit was used to transcribe RNAs into cDNAs (TOYOBO, Shanghai, China). Viral RNA loads in serum and tissues were calculated according to the standard curve (23).

Serum samples collected at 0, 3, 5, 7, 10, and 14 dpi were used for PRRSV-specific antibodies using a commercially available ELISA kit (IDEXX HerdChek ELISA, USA). S/P > 0.4 was considered the threshold of serological positivity.

All data were expressed as means ± standard deviations (SD). Statistical significance between the two groups was ascertained by conducting t-tests with GraphPad Prism software version 10 (San Diego, CA, USA), where a p-value less than 0.05 was deemed to indicate statistical significance.

Of the 20 lung samples from the piglets with severe respiratory illness, 19 (95%) were detected PRRSV-positive by RT-PCR. The Sanger sequencing showed the sequences of Nsp2 and ORF5 among the 19 samples were identical. Therefore, GX2024 was determined to be the main causative strain of this outbreak. For viral isolation, iPAMs (immortalized PAMs)and Marc-145 were used to isolate the new PRRSV strain. The GX2024 strain was blindly passaged in iPAMs and Marc-145 cells for six passages. At 60 h after viral inoculation, the infected iPAMs and Marc-145 cells showed marked morphological changes, such as clustering and shedding. No visible CPE was observed in the uninfected cell population (Figure 1a). The viral supernatants from the sixth passage of iPAMs and Marc-145 cells were both tested using RT-PCR to confirm positivity for PRRSV. In addition, PEVD, PCV-2, PRV, and CSFV were all found to be negative by PCR or RT-PCR. We successfully isolated the GX2024 strain using a plaque forming assay and the isolated strain was confirmed by IFA experiments (Figures 1b,c). IFA was used to visualize the presence of the N protein of GX2024 in the infected iPAMs and Marc-145. Fluorescent labeling revealed anti-PRRSV N protein (green) within infected cell populations 48 h post-infection, whereas mock-infected cells showed no detectable signal under identical imaging conditions (Figure 1c). The results indicated that the GX2024 isolate could replicate in iPAM and Marc-145 cells. The TCID₅₀ value of the GX2024 isolate was 1 × 10^5.58/mL using Marc-145 cells.

The complete genomic sequence of GX2024 was 15,018 nucleotides (nt), not including the poly(A) tail structure. The sequences were deposited in the GenBank database under the accession number PV362838. The comparative analysis revealed that GX2024 exhibited sequence similarity levels of 82.4, 87.0, 85.5, and 81.4% when aligned with reference strains VR-2332 (L5A), JXA1 (L8E), NADC30 (L1C), and NADC34 (L1C), respectively (Table 1). Furthermore, the genome of the GX2024 isolate was compared with 10 PRRSV representative strains (VR-2332, RespPRRS MLV, JXA1, JXWn06, NADC30, CHsx1401, FJ1402, NADC34, RFLP 1–4-4, and L1C.5). The ORF1a (Nsp1, Nsp3-8) and ORF1b (Nsp9, NSP11) of GX2024 shared 89.9–100% nt similarity with JXA1-like (L8) strains, which higher than the similarity shared with other strains. Nsp2, NSP10, Nsp12, ORF2a, ORF2b and ORF4-7 of GX2024 shared 83.7–95.9% nt similarity with NADC30-like (L1) strains. The ORF3 shared the highest nt identity (90.1%) with RespPRRS MLV, higher than the similarity shared with its parent strain VR-2332 (89.8%). These findings suggest that the GX2024 isolate may have undergone chimeric recombination events.

In order to determine the genetic relationship between GX2024 and other PRRSV strains, ML phylogenetic trees based on complete genome and ORF5 sequences were generated using MEGA-X software (Figure 2). According to ORF5 genotyping, all PRRSV-2 strains in China belonged to one of the four lineages: L8E (JXA1-/CH-1a-like), L1C (NADC30-/NADC34-like), L5A (VR-2332-like), and L3 (QYYZ-like). The GX2024 isolate was classified in L8E based on its whole genomic sequences (Figure 2a), whereas it belonged to L1 based on ORF5 genotyping (Figure 2b), suggesting that GX2024 may be a recombinant strain. Furthermore, the L1C phylogenetic tree showed that the L1C strains in the USA displayed more abundant genetic diversity than those in China and the GX2024 clustered with L1C.5 strains with RFLP 1–4-4 pattern from the China during 2015–2024 (Figure 2c). These findings indicated that the L1C PRRSVs in China and the USA undergone distinct genetic evolution processes.

Comparison of the NSP2 HV region in GX2024 and other reference strains revealed that the GX2024 isolate shared identical noncontiguous amino acid deletions (111 + 1 + 19-aa), 111-aa deletion at positions 322–432, 1-aa deletion at position 483, and 19-aa deletion at positions 504–522 with the L1 representative strain NADC30, which was reported in the United States in 2008 (Figure 3a). In addition, comparisons of the amino acid analyses of GX2024 GP5 with those of the other L1C strains from China and the United States showed distinct amino acid mutation sites, such as A⁹⁴, F¹²⁰, S¹⁵⁸, and R¹⁹¹, were identified in L1C strains from the USA. Notably, GX2024 isolate exhibited several unique amino acid substitutions, including P¹⁵ → L¹⁵, F²³ → S²³, A²⁶ → V²⁶, and A²⁷ → V²⁷ (Figure 3b).

We examined the sequence alignment between GX2024 and reference PRRSV strains using RDP4 and SimPlot to identify potential recombination events. RDP4 analysis revealed that GX2024 is a natural recombinant virus derived from three parental strains with p-values of ≤ 4.264 × 10⁻⁹, from the results of at least six detection methods (Table 2). The NADC30 was the major parental strain, and the JXA1 and RespPRRS MLV were the minor parental strains. Similarity plotting identified five recombination breakpoints in the GX2024 genome. They were located in ORF1a (nt 1868, 3,565), ORF1b (nt 9,392), ORF3 (nt 12,688), and ORF4 (nt 13,172) (Figure 4a). The five breakpoints separated the genome into six regions, with region A (nt 1–1868) and region B (nt 3,565–9,392) being closely related to the JXA1-like strain (L8), and region C (nt 12,688-13,172) being closely related to the RespPRRS MLV strain (L5). Region D (nt 1869–3,564, 9,393–12,687 and 13,173–3’-UTR) was classified into the NADC30-like cluster (L1) (Figure 4b). Therefore, the naturally recombinant virus GX2024 identified in this study was derived from the recombination of lineage 1 (NADC30-like), lineage 8 (JXA1-like), and the attenuated vaccine strain RespPRRS MLV (lineage 5).