Prevalence and genetic diversity of porcine epidemic diarrhea virus in Southwest China during 2020–2022 Authors: Tong Xu, Yuan-Cheng Zhou, Zhe-Yan Liu, Ji-Zhong Zhang, Fang Wu, Dong You, Liang-Peng Ge, Zuo-Hua Liu, Jing Sun, Xiu Zeng, Si-Yuan Lai, Yan-Ru Ai, Jian-Bo Huang, Ling Zhu, Zhi-Wen Xu Published: 2024 ABSTRACT AbstractPorcine epidemic diarrhea (PED), caused by porcine epidemic diarrhea virus (PEDV), has been frequently occurring in the southwestern region of China over the past few years, continuously affecting the development of the swine industry. However, the genetic diversity and prevalence of PEDV strains circulating in the swine population in southwestern China in recent years have not been well studied. To address this gap, a total of 478 clinical samples were collected from 125 pig farms experiencing piglet diarrhea in 18 cities in southwestern China. The detection results revealed that 227 out of 478 samples tested positive for PEDV nucleic acid, with a positivity rate of 47.49%. Complete S gene sequences of 28 PEDV strains were obtained and classified into four subgroups, G1-a subgroup (classical strain), G1-b subgroup (S-INDEL), and two G2 subgroups (G2-a and G2-b), accounting for 17.86% (5/28), 3.57% (1/28), 35.71% (10/28), 42.86% (12/28) of the total sequenced strains, respectively. The coexistence of multiple genotypes indicates the complex genetic background and prevalence of PEDV in southwest China. Amino acid comparisons of the S proteins showed that the 28 PEDV strains sequenced in the study showed different patterns of variation in the epitope domains compared to vaccine strains belonging to different genotypes and contained many unique amino acid mutations compared to the reference strains, which might lead to immune escape of PEDV. The complex epidemiology of PEDV with multiple subgroups co-circulating in Southwest China underscores the importance of selecting appropriate vaccine strains based on locally prevalent strains and the ongoing need for epidemiological surveillance of PEDV. The emergence of new variant strains also highlights the urgency of developing updated vaccines, and effective management practices remain crucial for controlling PED outbreaks in pig farms.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-024-80844-x.Keywords: PEDV, Prevalence, Genetic diversity, Southwest China, Amino acid mutationsSubject terms: Pathogens, Virology Introduction Porcine epidemic diarrhea (PED) is caused by porcine epidemic diarrhea virus (PEDV) and characterized by various clinical manifestations, including diarrhea, vomiting, anorexia, dehydration, and weight loss in piglets1,2. PEDV causes disease in pigs at all stages of growth, with the highest morbidity rate in suckling piglets, resulting in severe economic losses in the European and Asian swine industry3,4. PED cases were initially described in England in the 1971 and subsequently swept throughout most of Europe and Asia5,6. PEDV was first identified in China in 1984 by fluorescently labeled antibody and serum neutralization tests7. Due to the widespread use of inactivated or attenuated bivalent vaccines based on the CV777 strain, PEDV infection was sporadic and regional in China before 20108. The PED pandemic, caused by a new variant with high virulence, initially emerged in southern China in the winter of 2010 and then spread rapidly across the country, with devastating damage to the pig industry9,10. Pigs in all growth stages were not spared and affected by severe vomiting and acute diarrhea, even in vaccinated farms9,11,12. More seriously, the mortality rate of suckling pigs was almost 100%9,11. PEDV is an enveloped, single-stranded, positive-sense RNA virus belonging to the genus Alphacoronavirus of the family Coronaviridae in the order Nidovirales13,14. The S protein of porcine epidemic diarrhea virus (PEDV) is a type I glycoprotein15,16, and plays an important role in virus attachment, receptor binding, and virus-cell membrane fusion during cell entry17–19. In addition, at least five neutralizing epitope regions have currently been identified in the PEDV S protein, including the N-terminal domain S10 region (NTD/S10), the collagenase equivalent domain (COE), the SS2, SS6 and 2C1020,21. Therefore, the PEDV S glycoprotein is a major target for the development of an effective PEDV vaccine and is also essential for understanding the epidemiological status of PEDV in the field3. Since 2010, the new variant PEDVs have caused serious pig diarrhea disease in many provinces of China and has become one of the most important viral enteric diseases. Sun et al. evaluated the genetic evolution and S gene amino acid mutations of nine PEDV variants that emerged between 2020 and 2023 in the central-eastern regions of China, specifically Jiangsu and Anhui provinces22. This research offers crucial data to support local PEDV prevention and control efforts22. Another study, focusing on the prevalence of diarrhea-associated viruses in five southern provinces of China (Guangdong, Guangxi, Fujian, Jiangxi, and Hunan) between 2021 and 2023, revealed that PEDV was the most frequently detected virus, with sample prevalence rates ranging from 47.40 to 52.22% and farm prevalence at 76.06% (162/213)23. Additionally, Zhao et al. reported a high PEDV infection rate of 70.9% (232/327) among diarrheic piglets based on 327 fecal samples collected from four northeastern provinces—Heilongjiang, Jilin, Liaoning, and Inner Mongolia—between January 2015 and March 202224. These findings, along with other recent studies from various provinces in China and other countries, confirm that PEDV remains the predominant pathogen responsible for porcine diarrhea25–28. Considering the sustained and severe damage caused by PED in Southwest China, it is necessary to control and prevent PEDV infection. Nevertheless, little information was currently available about genetic diversity and prevalence of PEDV strains circulating in the swine population of Sichuan Province in recent years. Given the aforementioned premises, the aim of this study was to investigate the prevalence of PEDV in 478 randomly selected clinical samples from southwestern China between 2020 and 2022 using a real-time PCR assay. Moreover, the complete S genes of 28 PEDV strains were sequenced and analyzed to study genetic diversity of the PEDV currently circulating in Southwest China during 2020–2022. Materials and methods Sample collection and viral RNA extraction A total of 478 clinical samples (fecal and intestinal samples) from diarrheal pigs were collected from 125 pig farms in 18 cities (Ganzi, Liangshan, Yaan, Bazhong, Leshan, Yibin, Meishan, Zigong, Neijiang, Chengdu, Ziyang, Suining, Deyang, Guangan, Dazhou, Nanchong, Mianyang, Guangyuan) in the Southwest China during 2020–2022. All 125 farms included in the study experienced diarrhea outbreaks. Farms were not pre-selected based on confirmed PEDV outbreaks, but rather on the occurrence of diarrhea, and samples were subsequently tested for PEDV to determine positivity. Relevant information on animal characteristics, farm type and other relevant variables has been summarized in Supplementary Table S2. All samples were frozen in ice packs and transported to our laboratory for further processing. Each sample was individually prepared and mixed with sterile phosphate-buffered saline (PBS) in a sterile 1.5-mL microcentrifuge tube, frozen and thawed three times, and then centrifuged at 12,000 × g. Then, the supernatant was either used for RNA extraction immediately or stored at −80℃ until use. The serum from the 28 pigs, for which S gene sequences were obtained, was used to assess IgA antibody production. An ELISA kit for PEDV IgA antibodies (Keqian, Wuhan, China) was used to detect IgA levels in the serum of pigs from which the S gene samples were obtained, following the manufacturer’s instructions. According to the manufacturer’s instructions, RNA viral genomes were extracted using the TaKaRa MiniBEST Universal RNA Extraction Kit (Takara, Dalian, China), and then reverse-transcribed into cDNA using the PrimeScript™ II 1st Strand cDNA Synthesis Kit (Takara). PEDV detection and amplification of the complete S genes To detect PEDV in clinical samples, a real-time PCR assay based on SYBR Green І was performed as previously described29. A topographic map of China and its southwest was generated using ArcMap 10.2 (ESRI, Inc.) with WGS84 datum and Universal Transverse Mercator conformal projection. The amplification was performed using a CFX96™ Real-Time PCR Detection System (Bio-Rad, Hercules, CA, USA) with a 20 µl total reaction consisting of 10 µL of 2 × ChamQ Universal SYBR qPCR Master Mix (Vazyme, Nanjing, China), 0.5 µL (10 µM) of each pair of primers, 2 µL of viral cDNA, and 7 µL of ddH2O. The thermal conditions initiated with one cycle at 95 °C for 5 min, followed by 45 cycles at 94 °C for 30 s, 60 °C for 20s and 72 °C for 20 s. The pair of primers for detection was summarized in Table 1. The complete S genes of PEDV strains were sequenced as described previously30, and the four pairs of overlapping primers used to amplify the complete S gene were summarized in Table 1. Briefly, the purified PCR products were cloned using a pMD19-T Vector Cloning Kit (Takara, Dalian, China), and propagated in DH5α competent cells (Takara) according to the manufacturer’s instructions. Three clones of each PCR product were sequenced independently by Sangon Biotech Shanghai Co, Ltd., and all sequencing reactions were performed in duplicate. Sequences with low sequencing quality in one of the four fragments were directly discarded. Table 1List of primer sequences used in this study.Primer nameNucleotide sequence (5’−3’)Product sizePEDV-1 FAAATGGGAAGTCGGCAGA163PEDV-1RGTTTTGTTGTGGCGGTAGPEDV S1FGGTAAGTTGCTAGTGCGTAA1075PEDV S1RGCAGTATGAAGTACAATTGAGCCPEDV S2FGTTTGTAATGGAGCTGCTG1132PEDV S2RACTGTCCAGAATCAGATGTGPEDV S3FACTTGAAGGTGTCACGGACG1079PEDV S3RGCTGTAGAACATCCGTCTGTAPEDV S4FCAGTATTACTCTGGTGTCATG1404PEDV S4RCAACTGTGTCAATCGTGTPEDV-1 F/1R is a pair of primers used for detection and reported by Han et al.29. Four other primer pairs were used to amplify the complete S gene of PEDV as reported by Tian et al.30. List of primer sequences used in this study. PEDV-1 F/1R is a pair of primers used for detection and reported by Han et al.29. Four other primer pairs were used to amplify the complete S gene of PEDV as reported by Tian et al.30. Sequence analysis The complete S genes of PEDV strains acquired in this study were assembled by using DNASTAR Lasergene, and then analyzed with 62 PEDV reference strains deposited in GenBank and described in a previous study31. The information of the 28 identified PEDV strains in this study and reference strains was shown in Supplementary Table S1. Nucleotide and deduced amino acid sequences were aligned using the Clustal W method of MegAlign program. Molecular Evolutionary Genetics Analysis (MEGA) software (version 7.0) was applied to construct phylogenetic tree by the maximum likelihood method with GTR + G + I as the best-fit substitution model and 1,000 bootstrap replicates. Results Prevalence of PEDV in clinical samples collected in Southwest China during 2020–2022 Among the 478 porcine diarrheal samples collected from 18 cities in Sichuan, 227 samples were positive for PEDV with a positive rate 47.49% (227/478). PEDV positivity rates in different cities in different years were summarized in Table 2. The geographical distribution of the samples collected in this study and the positive samples were presented in Fig. 1. To visualize the plains and the surrounding mountainous areas, a topographic map of China and its southwestern region is shown in Supplementary Figure S1. When the data were analysed by year, prevalence of PEDV at sample level was 41.67% (65/156) in 2020, 51.37% (75/146) in 2021 and 49.43% (87/176) in 2022. Clinical samples were collected in 18 cities during 2020–2022, and PEDV was identified in all 18 cities, including Ganzi (14.29%), Liangshan (11.54%), Yaan (56.00%), Bazhong (36.00%), Leshan (42.86%), Yibin (58.33%), Meishan (50.00%), Zigong (30.00%), Neijiang (64.00%), Chengdu (66.67%), Ziyang (50.00%), Suining (45.45%), Deyang (46.15%), Guangan (48.00%), Dazhou (50.00%), Nanchong (73.33%), Mianyang (53.85%), Guyangyuan (46.15%) (Fig. 1). These data indicate that PEDV is the primary pathogen causing diarrhea diseases in Southwest China. As shown in Table 2, the positivity rate of PEDV in diarrheic pigs was highest in 2021 at 51.37% (75/146), followed by 49.43% (87/176) in 2022, and lastly 41.67% (65/156) in 2020. Table 2Information on samples collected in Southwest China between 2020 and 2022 and results of PEDV detection.Cities202020212022TotalGanzi1/6 (16.67%)0/5(0.00%)2/10(20%)3/21(14.29%)Liangshan1/12(8.33%)1/6 (16.67%)1/8(12.50%)3/26(11.54%)Yaan3/6 (50.00%)7/9 (77.78%)4/10(40.00%)14/25(56.00%)Bazhong4/8(50.00%)4/14(28.57)1/3(33.33)9/25(36.00%)Leshan5/15(33.33%)3/8(37.50%)7/12(58.33%)15/35(42.86%)Yibin7/12(58.33%)4/7(57.14%)3/5(60.00%)14/24(58.33%)Meishan4/17(23.53%)12/22(54.55%)12/17(70.59%)28/56(50.00%)Zigong2/7(28.57%)2/3(66.67%)2/10(20.00%)6/20(30.00%)Neijiang7/9(77.78%)2/5(40.00%)7/11(63.64%)16/25(64.00%)Chengdu5/9(55.56%)7/12(58.33%)16/21(76.19%)28/42(66.67%)Ziyang2/5(40.00%)6/10(60.00%)3/7(42.86%)11/22(50.00%)Suining1/3(33.33%)2/5(40.00%)2/3(66.66%)5/11(45.45%)Deyang3/4(75.00%)7/13(53.85%)2/9(22.22%)12/26(46.15%)Guangan5/8(62.50%)2/2(100.00%)5/15(33.33%)12/25(48.00%)Dazhou4/10(40.00%)4/5(80.00%)6/13(46.15%)14/28(50.00%)Nanchong5/8(62.50%)3/4(75.00%)3/3(100.00%)11/15(73.33%)Mianyang4/11(36.36%)5/9(55.56%)5/6(83.33%)14/26(53.85%)Guyangyuan2/6(33.33%)4/7(57.14%)6/13(46.15%)12/26(46.15%)Total65/156(41.67%)75/146(51.37%)87/176(49.43%)227/478(47.49%) Information on samples collected in Southwest China between 2020 and 2022 and results of PEDV detection. Positive rates of PEDV in different cities in Southwest China. The clinical samples in this study were collected from orange-marked cities in southwestern China, and the numbers indicate the PEDV-positive rates in different cities. The map of China and its southwestern regions was downloaded as an open-source SVG file from the NB map website (https://www.nbcharts.com/map/map.php?loginSuccess=true) and further edited using PowerPoint (version office 2021, https://www.microsoft.com/zh-cn/microsoft-365/get-started-with-office-2021). For the purpose of analyzing seasonal variations in PEDV prevalence, the months were grouped into four seasons as follows: winter included December, January, and February; spring comprised March, April, and May; summer encompassed June, July, and August; and autumn included September, October, and November. The positivity rates of PEDV in spring, summer, autumn, and winter were 49.19% (61/124), 14.49% (10/69), 40.00% (38/95), and 62.11% (118/190), respectively (Fig. 2). The positivity rates of PEDV vary across different growth stages of pigs. The highest positivity rate was observed in pre-weaning piglets at 56.32% (205/364), followed by 20% (15/75) in growers and 17.95% (7/39) in finishers (Fig. 2). Moreover, PEDV can be detected in both immune and non-immune pig farms, with distinct positivity rates of 17.72% (28/158) and 62.19% (199/320), respectively (Fig. 2). Detection of PEDV infection in pigs in Southwest China. (A) Total PEDV positivity in diarrheic pigs from 2020 to 2022. (B) PEDV positivity in pigs at different growth stages. (C) PEDV positivity in immunized and non-immunized pigs. (D) PEDV positivity rates in different cities. (E) PEDV positivity rates in different seasons. Seasons were defined as winter (December–February), spring (March–May), summer (June–August), and autumn (September–November). The corresponding positive rates were labeled in the histogram. The samples used for obtaining the S gene sequences were also tested for serum antibodies, and the results are presented in Supplementary Table S2. The antibody positivity rate was lower than the nucleic acid positivity rate. This discrepancy may be due to the fact that many piglets have underdeveloped immune systems and succumb quickly to PEDV infection due to dehydration and malnutrition before generating detectable levels of antibodies. Sequencing and phylogenetic analysis The full S gene sequences from 43 PEDV strains were amplified from diarrhea-affected swine farms in southwest China, and among which 28 were unique sequences and deposited in the NCBI GenBank with accession numbers (PP395567- PP395594). These sequences, ascertained in this research, exhibited nucleotide identities ranging from 93.4 to 99.9% and amino acid (aa) identities between 92.3% and 99.9% among themselves. Moreover, they demonstrated 93.2–99.8% nucleotide similarity and 92.2–99.6% aa similarity when compared with other representative PEDV strains deposited in GenBank. Phylogenetic analysis based on the S gene sequences of the 28 PEDV strains in this study and the other 62 reference strains showed that 90 strains could be classified into five subgroups, including subgroup G1-a (classical strains), subgroup G1-b (S-INDEL strains), subgroup G2-a (variant strains), subgroup G2-b (variant strains) and subgroup G2-c (recombinant strains) (Fig. 3). Among which, 28 PEDV strains obtained in Southwest China were classified into four subgroups, G1-a subgroup (classical strain), G1-b subgroup (S-INDEL) and two G2 subgroups (G2-a and G2-b, variant strain), and, accounting for 17.86% (5/28), 3.57% (1/28), 35.71% (10/28), 42.86% (12/28) of the total analysed strains, respectively (Fig. 3). Phylogenetic analysis of the complete S gene based on 28 PEDV strains identified in this study and 62 reference strains. The phylogenetic tree was constructed using Maximum Likelihood method in the MEGA7.0 program, with GTR + G + I as the best-fit substitution model and 1,000 bootstrap replicates. Sequences obtained in this study are marked with solid black circles. Three vaccine strains (ZJ08, AJ1102 and CV777) prevalent in China are marked with solid triangles. Different colors represent different genotypes. Deduced amino acids sequence analysis of the S protein The deduced amino acid sequences of S genes obtained in this study were analyzed with reference strains from the four subgroups, including classical strains (G1-a subgroup), S-INDEL strains (G1-b subgroup) and variant strains (G2-a and G2-b subgroups). These representative reference strains include two classical strains (CV777-attenuated and ZJ08), one S-INDEL strain (OH851) and three variant strains (AJ1102, ZJ-2011-2 and BJ-2011-1), of which CV777, ZJ08, and AJ1102 are prevalent vaccine strains in China. Among these 34 strains (28 strains in this study and 6 representative reference strains), there were 232 amino acid (aa) mutations in S protein (Supplementary Figure S2). In addition, aa mutations, deletions, and insertions in the five antigenic epitopes of the S protein are summarized in Fig. 4. These five antigenic epitope regions include the S10 (aa positions 20–220), the CO-26 K equivalent (COE) structural domain (aa positions 499–638), the SS2 epitope (aa positions 748–755), the SS6 epitope (aa positions 764–771), and the PEDV antigenic epitope motif (1368 GPRLQPY 1374). The aa positions in the antigenic epitope region were different from those in other studies because aa insertions and deletions in the S proteins of the PEDV strains obtained in this study. The aa insertions and deletions in the antigenic epitope region of the S protein are shown in the red squares of Fig. 4. As depicted in Fig. 4, different subgroups of PEDV strains shared different typical amino acid insertions, deletions and mutations in the S protein. Compared with the classical strains (G1-a) and the S-INDEL strains (G1-b), the variant strains (G2) had four aa insertions at positions (59QGVN62) and numerous aa mutations at positions 27, 28, 29, 30, 55, 56, 57, 58, 64, 68, 69, 70, 84, 86,89 and so on (Fig. 4). Amino acid mutations, deletions and insertions in five antigenic epitope regions between the strains obtained in this study and the reference strains. The three vaccine strains are labeled in red, the non-vaccine reference strain is labeled in blue, and all strains obtained in this study are labeled in black. Different background colors of amino acid sites represent different genotypes. Amino acid sites without background color represent unique amino acid mutations in the strains obtained in this study. Red squares represent deletions and insertions of amino acids in the epitope region of the S protein. This figure was generated using a combination of MEGA (version 7.0, https://www.megasoftware.net/dload_win_gui) and Excel (version office 2021, https://www.microsoft.com/zh-cn/microsoft-365/get-started-with-office-2021). Amino acid sequences were first aligned using MEGA, and the alignment output was exported to an Excel workbook for further editing. In spite of the typical amino acids in the S proteins of the different subgroups, there were some specific amino acid changes in the amino acid sequences of the S proteins of the strains in this study. The four strains (SC2020YB, SC2020ZY, SC2020DY and SC2020BZ) in this study, although belonging to the classical strain (G1-a), contained amino acid mutations (T2K/R, P3S, I5N and L15S) similar to those of the S-INDEL strains (G1-b) and the variant strains (G2-a and G2-b). Notably, the aa mutations (I71L, D74A, S87A, T117H, S118N, S150F) of SC2020ZY clustered in classical strains were consistent with S-INDEL strains when compared to the vaccine strain CV777 in classical strains, which are located in the epitope region NTD/S10. In comparison to the three classical strains (G1-a subgroup) (CV777 attenuated, ZJ08, and SC2021DZ), the other strains in this study contained an aa insertion (157Y/H) in the antigenic epitope region NTD/S10. Two aa deletions at position 163–164 were observed in the variant strain SC2021MY acquired in this study. Many unique amino acid mutations were also observed in the S proteins of the strains obtained in this study compared to the reference strains including the vaccine strains. SC2020LS, SC2020NC, SC2020NJ and SC2021ZG shared a common aa substitution (F10L). Phenylalanine (F) is an aromatic, hydrophobic amino acid, while Leucine (L) is a non-aromatic, hydrophobic amino acid. This substitution results in the loss of the aromatic ring, which could potentially alter the interaction with other aromatic residues or ligands, though it maintains the hydrophobic character, likely preserving the local hydrophobic environment. SC2021GA contains three unique aa substitutions (Y52F, P155S and S527G). Y52F (Tyrosine to Phenylalanine) involves a change from an aromatic, polar amino acid with a hydroxyl group (Y) to an aromatic, non-polar amino acid (F). This substitution removes the hydroxyl group, potentially affecting hydrogen bonding and local polarity, while maintaining the aromatic ring structure. P155S (Proline to Serine) involves a significant structural change, as Proline (P) is a non-polar, cyclic amino acid that introduces kinks in the protein backbone, whereas Serine (S) is a polar amino acid with a hydroxyl group. This change could alter the local protein structure and affect both conformation and molecular interactions. S527G (Serine to Glycine) involves the substitution of a polar amino acid (S) with a small, non-polar amino acid (G), increasing flexibility at this position, which could impact the stability or conformation of the S protein. SC2021BZ contains four unique aa substitutions (S27L, Q59H, S522V and Y608N) within antigenic epitopes (NTD/S10, COE). S27L (Serine to Leucine) involves the replacement of a polar residue (S) with a bulkier, hydrophobic one (L), which could disrupt local interactions and potentially affect overall protein folding or stability. Q59H (Glutamine to Histidine) changes a polar, uncharged amino acid (Q) to a polar, positively charged one (H) at physiological pH, introducing a site for ionic interactions and pH-dependent conformational changes that may influence protein function under different environmental conditions. S522V (Serine to Valine) involves the substitution of a polar residue (S) with a hydrophobic one (V), which could alter the local hydrophobic environment and disrupt hydrogen bonding, affecting the protein’s interaction with surrounding molecules. Y608N (Tyrosine to Asparagine) replaces an aromatic, polar amino acid (Y) with a non-aromatic, polar one (N), eliminating the aromatic character and introducing different hydrogen-bonding potential, which could impact epitope recognition by the immune system. SC2020CD, SC2020LS, SC2020NC, SC2020NJ and SC2021ZG shared a common aa substitution (F541L) within antigenic epitope COE. This substitution, like F10L, results in the loss of an aromatic ring (F to L), which may affect protein interactions, although the hydrophobic nature of the region is likely preserved. Except for the aforementioned amino acid mutations, many other unique amino acid mutations also occurred in the antigenic epitope region NTD/S10 and COE of S proteins of PEDV strains from southwestern China, including T137N, A15V, P155S, T505A, L506M, S527G, S542G, etc. These changes, particularly those involving significant shifts in polarity, charge, or aromaticity, could have important implications for the S protein’s structure, antigenicity, and interactions with host receptors or antibodies. Such mutations may contribute to immune escape mechanisms and could affect the efficacy of vaccines developed from earlier strains32–35. No amino acid mutations were found at the SS2 and 2C10 epitopes in any of the Southwest PEDV strains and in the reference strain. Discussion Since the emergence of the variant strain in October 2010, PEDV has become one of the major causes of piglet diarrhea in the global pig industry, causing significant economic losses to the pig industry worldwide36. Classical vaccine strain CV777 could not provide 100% protection to piglets against the variant strains11. Suitable vaccine strains are essential for vaccine immunization and PED control, and the prevalence and genomic characterization of PEDV are prerequisites for the selection of suitable vaccine strains. Although there have been several reports on the prevalence and genetic characterization of PEDV in Southwest China prior to 201830,37, limited information is available for the last three years. In this study, an epidemiological survey was conducted to investigate the prevalence of PEDV and genomic characterization of the virulent strains in Southwest China in 2020–2022. Of the 478 clinical samples collected in this study, 227 were PEDV-positive with a high positivity rate (47.49%, 227/478), which was higher than the prevalence (35.81%, 231/645) of PEDV in Southwest China from 2014 to 2018 reported in a past study30. This demonstrates that the prevalence of PEDV in the region has increased over recent years. Meanwhile, the high prevalence of PEDs exists globally and in other parts of China as well. For instance, in southern China, the prevalence rates of PEDV were reported to range between 47.40% and 52.22% across different farms23. Similarly, studies in East Central China observed a positivity rate of 18.26%22, while an even higher positivity rate of 70.9% was reported in Northeast China24. This trend is also observed internationally, where PEDV outbreaks in Mexico showed a positivity rate of 78.2%25, and in South Korea, PEDV was consistently reported between 2013 and 2022 with ongoing outbreaks38. These findings demonstrate that PEDV remains a significant pathogen not only in China but also in many other parts of the world, requiring constant vigilance and surveillance to manage its spread. When the data were analysed by year, PEDV positivity rates are relatively stable from 2020 to 2022. When the data were analysed by season, PEDV could be detected in all seasons, and the highest positive rate was 62.11% (118/190) in winter, followed by the positive rate 49.19% (61/124) in spring. PEDV exhibits seasonal patterns of prevalence, with its occurrence often associated with specific seasons, which is consistent with a previous study39. The virus tends to experience increased activity and higher infection rates during colder months, such as winter and early spring. This seasonal prevalence may be influenced by factors such as environmental conditions, pig management practices, and the virus’s stability in different temperatures. During colder seasons, the virus may have increased stability in the environment, and pig populations may experience higher stress levels, potentially making them more susceptible to infection. Additionally, the virus may spread more easily in conditions with lower temperatures and reduced sunlight. Understanding the seasonal patterns of PEDV is crucial for implementing effective control and prevention strategies, as well as optimizing biosecurity measures during periods when the virus is more likely to be prevalent. As shown in Fig. 1, the PEDV positivity rate was higher in the plains of southwest China compared with the surrounding mountainous areas. To visualize the plains and the surrounding mountainous areas, a topographic map of China and its southwestern region is shown in Supplementary Figure S1. The most likely reason for the low prevalence detected in mountainous areas could be the low population density, small pig farming operations and geographical isolation effect. Meanwhile, higher numbers of animals on a farm can facilitate the spread of PEDV, as increased population density generally increases the risk of viral transmission. However, the level of technology on the farm, particularly in terms of environmental control and management practices, plays a crucial role in mitigating this risk. Farms that implement advanced biosecurity measures, effective sanitation, and proper ventilation are able to significantly reduce the likelihood of PED outbreaks, even in large-scale operations. This highlights the importance of combining good management practices with technological solutions to control PEDV, regardless of farm size. The pre-weaning piglets are the most susceptible to PEDV. Due to their underdeveloped immune systems, they rely primarily on passive immunity from the sow for protection. As piglets grow, their susceptibility to PEDV decreases. However, we observed that a few growers and finishers still became infected with PEDV, which is closely related to farm management and environmental factors. In our study, the infection rate among grower pigs was generally low. The few cases of PEDV infection in grower pigs were primarily found in farms with poor hygiene, inadequate management, and overall poor pig health. After we provided recommendations, these farms implemented emergency vaccination, improved cleaning and disinfection protocols, and strengthened management practices. Following these measures, PEDV was brought under control. After this batch was sold, the pens were thoroughly cleaned and disinfected—continuing until multiple areas tested negative for PEDV nucleic acid. Subsequently there were no further major outbreaks in subsequent batches of pigs. Therefore, we recommend immunizing pregnant sows but not post-weaning pigs. Additionally, farms should strengthen cleaning, disinfection, and management measures. To explore the genetic characterization of PEDV in southwest China, 28 unique complete S genes were amplified. Sequence comparison results showed that the S genes of the 28 strains shared 93.2–99.8% nucleotide similarity and 92.2–99.6% aa identity with other representative reference PEDV strains. In order to understand the evolutionary relationship between PEDV strains in Southwest China and other regions, a phylogenetic tree was constructed based on the complete S gene sequences of 28 identified PEDV strains and 62 reference strains. The results showed that 28 PEDV strains could be classified into four subgroups (G1-a, G1-b, G2-a and G2-b), which accounted for 3.57% (1/28), 35.71% (10/28), 42.86% (12/28) and 17.86% (5/28) of the total number of analyzed strains, respectively (Fig. 2). Multiple subgroups of PEDV coexisted in southwestern China, which is consistent with reports from other regions of China and across the globe23–25,30,38. Notably, the study from 2014 to 2018 found three coexisting subgroups in Southwest China: G2-b, G2-c, and an S-INDEL subgroup (G1-c)30. However, the subgroup types and the percentage of different subgroups are different from our study. In Northeast China, the dominant subgroups were G2-b and G2-a, while in southern China, G2-b and recombinant PEDV variants were prevalent23,24. Internationally, similar diversity has been observed, with Mexico reporting G2a, G2b, and S-INDEL strains, and South Korea detecting multiple subgroups, with G2b being the most dominant25,38. While many regions show the coexistence of multiple genotypes, the primary circulating strains vary by location, underscoring the need for region-specific vaccine development and tailored prevention strategies. These findings highlight the complex epidemiology of PEDV and the necessity for ongoing genetic monitoring to inform vaccine updates and improve control measures. Ideally, vaccines should provide broad cross-protection; however, the costs associated with veterinary vaccines and the lengthy development times for broad-spectrum vaccines must be carefully considered. Therefore, the best strategy for swine producers at this time is to conduct real-time monitoring of locally circulating strains and select appropriate vaccine candidates. If the circulating strains are significantly genetically divergent from commercial vaccines, collaboration with vaccine companies to develop region-specific vaccines based on local strains may be warranted. Moreover, alongside vaccination, other factors such as biosecurity measures and management practices are crucial for controlling PED. Vaccine companies should be prepared to update vaccine strains promptly based on national trends while also striving to develop broadly cross-protective vaccines, balancing efficacy with cost-effectiveness. The S protein of PEDV is a major target for the host immune response, making it an important focus for research on vaccine development and understanding the virus’s pathogenicity20,21. Moreover, PEDV S protein contains the most neutralizing epitopes, including the epitope CO-26 K equivalent (COE) domain, SS2, SS6, and 2C1032–34. In recent years, a new neutralizing epitope (S10, mapping to the NTD (20–220 aa) of the S1 subunit) was identified20,30. A total of 102 amino acid mutations were found in these five antigenic epitopes of the S protein in 34 PEDV strains, including 28 strains from Southwest China and 6 representative reference strains (Fig. 3). Significant variation was observed in the S10 antigenic epitope region between different subpopulations, and as reported in previous studies some typical amino acid insertions, deletions and mutations in S proteins can be used as markers to differentiate G1 from G2 strains40,41. These mutations are at the potential N-glycosylation of the PEDV S protein and affect the potential N-glycosylation sites pattern. N-linked glycosylation of viral glycoproteins plays a key role in regulating viral receptor binding and membrane fusion and participates in immune evasion through sugar shielding epitopes35. Thus, amino acid mutations in antigenic epitope regions or N-linked glycosylation sites may affect the pathogenicity, antigenicity, neutralization properties or even cause immune escape of the strain30,32–35. In addition, the lower virulence of the S-INDEL strain compared to the G2 strain and the limited immune cross-protection afforded by classical vaccine strains against the mutant strain13,42–44 may be further evidence in support of this hypothesis. As depicted in Fig. 3, epitopes SS2 and 2C10 are conserved, whereas epitopes COE and SS6 have higher mutation rates, and S-INDEL strains and variant strains possess similar mutation patterns in epitopes COE and SS6, which is in accordance with previous reports31,45–47. Currently, CV777, ZJ08 and AJ1102 are the prevalent vaccine strains in China, with CV777 and ZJ08 belonging to the G1a (classical strains), and AJ1102 belonging to the G2-a (variant strains) (Fig. 2). A large number of amino acid mutations were present on the S proteins of 28 PEDV strains from southwestern China compared with vaccine strains of different genotypes (Fig. 3). Even compared with the vaccine strains of the same genotype, many amino acid mutations were still present in the S protein of the PEDV strain from southwest China (Fig. 3). Many unique amino acid mutations were observed on the S proteins of the PEDV strains obtained in this study compared to representative reference strains (Fig. 3). Notably, several substitutions, such as F10L, Y52F, P155S, and S527G, involve changes in aromaticity, polarity, and hydrophobicity, which are likely to impact the structural integrity and functional properties of the S protein. These amino acid mutations might lead to immune escape and immune failure in PEDV strains30. These hypotheses are supported by clinically observed facts in this study, where PED outbreaks caused by the G2-a strain occurred in CV777(G1-a)-immunized pig farms, and instances of PED caused by the G2-b strain were even observed in farms immunized with AJ1102 (G2-b). These clinical phenomena were shown in Supplementary Table S2. Similar phenomena have also been reported in several recent studies30,39. These data underscore the importance of selecting appropriate vaccine strains and the urgency of developing new vaccine strains. In spite of the amino acid mutations, factors such as poor environment and poor management may have contributed to the observation of PED cases caused by the G2-b strain in farms vaccinated with AJ1102 (G2-b). These findings suggest that biosecurity measures, and environmental hygiene are also crucial factors in preventing PED outbreaks. There are several limitations to the current study. First, the lack of a comprehensive evaluation system for farm environments and management practices restricted our ability to analyze these factors systematically within the dataset. Second, the majority of samples were collected from Sichuan Province, underscoring the need to expand the geographic scope of our sample collection to provide a more representative overview of PEDV prevalence in the southwestern region of China. In future research, our goal is to develop an objective, standardized assessment system for on-farm environments and management practices that will enhance our prevention and control of PEDV and other diseases. Additionally, we will strive to collect samples from a broader range of cities to better reflect the epidemiological trends of PEDV in southwestern China. In conclusion, PEDV was mainly responsible for the diarrheal disease outbreaks in Southwest China during 2020–2022. Moreover, four subgroups, G1-a, G1-b, G2-a and G2-b, were identified in the southwest, with G2-a and G2-b being the dominant genotypes. The different subgroups were evolutionarily distant from each other and showed different patterns of variation in the epitope domain of S proteins, which might account for immunization failures in farms, emphasizing the importance of vaccine strain selection and the urgency of developing vaccines based on new prevalent strains. The present study promotes us to understand the current prevalence and genetic characteristics of PEDV in Southwest China in recent years. Electronic supplementary material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Author contributions Z-WX, S-YL, Y-CZ and J-ZZ contributed to conceptualization. FW and DY contributed to the methodology. TX and Z-YL contributed to software and writing—original draft preparation. L-PG, Z-HL, JS, XZ, J-BH, LZ and Y-RA contributed to validation and supervision. Z-WX and LZ contributed to project administration and funding acquisition. Funding This work was supported by the Sichuan Science and Technology Program Projects (Key R&D Projects) (No. 2023YFN0021), the Key K&D Program of Sichuan Science and Technology Plan (No. 2022YFN0007), the Chongqing Municipal Technology Innovation and Application Development Project (No. cstc2021jscx-dxwt BX0007), the Porcine Major Science and Technology Project of Sichuan Science and Technology Plan (No. 2021ZDZX0010-3). Data availability The datasets generated and analysed during the current study are available in the NCBI GenBank repository under the following accession numbers: PP395567- PP395594. Declarations Competing interests The authors declare no competing interests. Ethics approval All experimental procedures were reviewed and approved by the Sichuan Agriculture University Animal Care and Use Committee (license number SCXK (Sichuan) 2013-0001). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent was obtained from the owners for the participation of their animals in this study. No human studies are presented in the manuscript. No potentially identifiable images or data are presented in this study. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Tong Xu, Yuan-Cheng Zhou and Zhe-Yan Liu contributed equally to this work. Contributor Information Ling Zhu, Email: abtczl72@126.com. Zhi-Wen Xu, Email: abtcxzw@126.com. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Data Availability Statement The datasets generated and analysed during the current study are available in the NCBI GenBank repository under the following accession numbers: PP395567- PP395594.