Novel Method for Isolation of Porcine Epidemic Diarrhea Virus with the Use of Suspension Vero Cells and Immunogenicity Analysis

In February and December of 2019, two large-scale outbreaks of diarrhea were observed in the same swine farm with 3,000 sows in Shanghai, China. We successfully isolated two porcine epidemic diarrhea virus (PEDV) isolates (strains shxx1902 and shxx1912 in February and December, respectively) from clinical samples in this farm using suspension Vero cells. A third PEDV strain (SH1302) tested positive in another farm of Shanghai, China, in 2013 and was also isolated using suspension Vero cells.

ABSTRACT

In February and December of 2019, two large-scale outbreaks of diarrhea were observed in the same swine farm with 3,000 sows in Shanghai, China. We successfully isolated two porcine epidemic diarrhea virus (PEDV) isolates (strains shxx1902 and shxx1912 in February and December, respectively) from clinical samples in this farm using suspension Vero cells. A third PEDV strain (SH1302) tested positive in another farm of Shanghai, China, in 2013 and was also isolated using suspension Vero cells. The three isolates were better adapted to growth in adherent Vero cells through serial passages in the suspension Vero cells. The three isolated strains were detected positive by an immunofluorescence assay (IFA) and observed through electron microscopy. Phylogenetic analysis of the complete genomic sequence demonstrated that shxx1902 (the 5th passage) and shxx1912 (the 5th passage) clustered with a new GII subgroup (GII-c), which consisted of SINDEL strains from America (e.g., OH851), and their S gene belonged to GII-a. Both strains(the 35th passage) have incurred dramatic changes in their genomes compared with the 5th passage. The 5th and 35th passages of SH1302 belonged to the GI-b genotype. The anti-N protein antibody titer of the strain shxx1902 was elevated to the same level as the vaccine strain (CV777) in mice. The use of the suspension Vero cells to isolate and propagate PEDV provides an effective approach for studies of the epidemiological characteristics and vaccine development of this virus.

INTRODUCTION

Porcine epidemic diarrhea virus (PEDV) (family Coronaviridae family, genus Alphacoronavirus) is an enveloped, single-stranded, positive-sense RNA virus which was about 28 kb. The PEDV genome contains a 5′ untranslated region (5′ UTR) and a 3′ UTR with a polyadenylated tail. The remaining part of the PEDV genome contains open reading frames (ORFs) specifying structural and nonstructural proteins in the following order: spike (S), ORF 3, envelope (E), membrane (M), and nucleoprotein (N) (1, 2). Infected piglets exhibit watery diarrhea, vomiting, dehydration, and significant mortality (approximately 80% to 100%), which has resulted in tremendous economic losses to the swine industry (3, 4). Since December 2010, a large-scale PED outbreak in suckling piglets was observed among swine farms in China, with 80% to 100% morbidity and 50% to 90% mortality (5, 6). However, the propagation of PEDV in cell culture remains challenging. Even although PEDV can be isolated from clinical samples, its titer decreased during serial passages in cell culture (7). This impediment caused the failure of vaccine development. As a result the in vitro culture method of PEDV remains to be researched. In addition, it was reported that PEDV has undergone numerous genetic variations during the process of spreading (8). In our study, three isolated strains were serially passaged, and an analysis of whole-genome differences between parent strains and cell-adapted strains showed multiple nucleotide changes. These frequently occurring mutations suggest that a full understanding of the genomic and epidemiological characteristics is critical in the fight against PEDV epidemics.

Currently, the Chinese classical strains along with the prototype strains (virulent CV777 and DR13) belong to the GI-a subgroup. Cell culture-adapted vaccine strains and other pandemic classical strains are members of GI-b subgroup. Since 2010, most of highly prevalent variant PEDV strains in China belonged to the GII genogroup. The GII genogroup includes three subgroups(GII-a, GII-b, and GII-c). GII-a mainly consists of Chinese strains(AH 2012, HuB1-2017, and HuB7-2017). AJ1102 is used as a vaccine strain in China and belongs to GII-b. A new GII subgroup (GII-c) consists of SINDEL strains from America (e.g., OH851). As demonstrated in clinical practice, the vaccines derived from classical strains are insufficiently protective (5, 9). In February and December of 2019, two large-scale outbreaks of diarrhea were observed in a swine farm with 3,000 sows in Shanghai, China. The morbidity and mortality rates were 10% and 70% to 80%, respectively, among piglets under 1 week old in the delivery room. We isolated two PEDV strains using suspension Vero cells and characterized their whole genome as genotype GII-c. At the same time, a PEDV virus strain of 2013 was also isolated and belonged to genotype GI. The whole sequence of the CV777 vaccine strain (used in the farm before 2017) and AJ1102 vaccine strain (used in the farm since 2017) belonged to GI and GII-b, respectively. This outbreak may have been caused by the dramatic gene mutation of the virus, posing a major challenge to the prevention and control of PED in China. In the present study, the immunogenicity of the isolated virus GII-c genotype shxx1902 was compared with the GI genotype vaccine strain (CV777) in order to develop the vaccine.

MATERIALS AND METHODS

Cells and clinical samples.

This study was conducted accordance with animal welfare guidelines of the World Organization for Animal Health and approved by Shanghai Municipal Commission of Agriculture (permit number 2013 [18]). Permission was obtained from the owners of the domesticated animals sampled. All methods were carried out in accordance with relevant guidelines and regulations. Vero cells (ATCC CCL-81) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Gibco). The clinical samples (small intestine tissues and fecal samples) used in this study were collected from a pig farm in Shanghai, China, where an outbreak of acute diarrhea among piglets occurred in February and December of 2019. The SH1302 sample was collected from a pig farm in Shanghai, China, in February of 2013. These intestinal samples were confirmed positive for PEDV using a commercial real-time RT-PCR (reverse transcriptase PCR; Guangzhou Vipotion Biotechnology Co., Ltd., China). The small intestine tissue was homogenized with serum-free DMEM and then centrifuged at 10,000 × g at 4°C for 10 min. The supernatant was filtered using cellulose acetate with a 0.22-μm pore size (Merck Millipore, Darmstadt, Germany) and used for virus isolation.

Virus isolation.

The three PEDV strains were purified by plaque cloning in Vero cells. Vero cells were seeded into 6-well plates at a density of 2 × 10⁶ viable cells in 2 ml per well in complete DMEM media. On the next day, 10-fold dilutions of PEDV virus inoculum allowing were prepared with approximately 1.5 ml of diluted virus for each plate. A total of 0.5 ml of diluted virus was inoculated per well in duplicate, and plates were incubated for 1 hour at 37°C. Afterward, the medium was removed, and the cells in each well were overlaid with 2 ml of postinoculation medium containing 1.5% agarose. A second overlay was added on day 2 (48 h later) to visualize the plaques. For each 100 ml of second overlay, 3 ml of a 0.33% neutral red solution (Sigma) was added; plaques became visible after approximately 1 to 3 hours and could also be observed the following day, depending on the condition of the cells. Macroscopically visible plaques were picked with a pipette pick, dissolved in 1 ml of DMEM, frozen and thawed three times after receiving neutral red solution, and then inoculated in Vero cells for propagation in preparation for the next round of plaque purification. The plaque purification operation was repeated three times (10).

In this study, suspension Vero cells were used to propagate PEDV. The adherent Vero cells were passaged and cultured in suspension in 20 ml of serum-free maintenance medium (DMEM; Gibco) supplemented with 2 μg/ml trypsin (Gibco), which was added to a 125-ml Erlenmeyer flask (nonpyrogenic polycarbonate; Corning Incorporated) at 37°C, 120 rpm, and 5% CO₂. The cell concentration was adjusted to 2 × 10⁶ cells/ml. A total of 200 μl of the viral inoculum was added to the 125-ml Erlenmeyer flask. The PEDV-infected cells and viral control cells were cultured at 37°C under 5% CO₂. After 72 h, the supernatants were inoculated into the freshly made suspension Vero cells through 35 passages. The supernatants from the 35th passage were inoculated into a T25 flask of adherent Vero cells supplemented with 2 μg/ml trypsin (Gibco) at 37°C under 5% CO₂ conditions. The cytopathic effect (CPE) was monitored daily, and the cells were harvested until the CPE exceeded 80% (11). After one freeze-thaw cycle, the supernatants were collected, packed separately, and stored at –80°C until further use. The virus titer was measured in 96-well plates using a 10-fold serial dilution of the samples. The 50% tissue culture infective dose (TCID₅₀) was expressed as the reciprocal of the highest dilution showing CPE using the Reed-Muench method (12).

Electron microscopy assay.

A total of 50 ml of PEDV-infected suspension Vero cells was frozen and thawed three times. Then, the infected cells were centrifuged at 10,000 rpm for 1 h and filtered with a 0.22-μm pore size filter (Merck Millipore, Germany). Subsequently, polyethylene glycol 8000 (PEG-8000; Solarbio, China) was added to the supernatant with a 10% final concentration. The mixture were stored at 4°C for 7 hours. After centrifugation at 12,000 rpm and 4°C for 2 h, viral particles were resuspended in Tris-buffered saline (TBS) solution in 1 ml and then negatively stained with 3% phosphotungstic acid and examined using a transmission electron microscope (Tecnai G2 Spirit Biotwin) (13).

Indirect immunofluorescence assay (IFA).

After inoculation for 24 h, mock-infected Veros and Vero cells infected with shxx1902, shxx1912, and SH1302 strains in 96-well plates, at a multiplicity of infection (MOI) of 0.1, were fixed with 4% paraformaldehyde at room temperature (20°C to 25°C) for 15 min, permeabilized with 0.2% Triton X-100 in phosphate-buffered saline (PBS) at room temperature for 2 to 5 min, and blocked with 1% bovine serum albumin in PBS at 37°C for 30 min. A 1:32 dilution of mouse anti-PEDV-S protein monoclonal antibody (MAb) S1D12 (Veterinary Medical Research&Development, USA) was incubated at 37°C for 2 hours. Fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse immunoglobulin G (IgG) (Sigma, Beijing, China) was incubated at 37°C for half an hour as the second antibody. The plates were then washed with PBS. Finally, the cells were stained with 4′,6-diamidino-2-phenylindole (DAPI) and visualized for fluorescence under a fluorescence microscope (14).

Viral replication kinetics.

The PEDV shxx1902, shxx1912, and SH1302 strains and CV777 virus strain were separately inoculated into suspension Vero cells in a 125-ml Erlenmeyer flask containing 2 μg/ml trypsin. The viral replication of the 5th and the 35th passages of strains shxx1902, shxx1912, and SH1302 and the CV777 virus strain, both for suspension and adherent Vero cells, were determined. Viral cultures were harvested at 0, 12, 24, 36, 48, 60, 72, and 84 hours postinfection (hpi). The 50% tissue culture infective dose (TCID₅₀) was expressed as the reciprocal of the highest dilution showing CPE of adherent cells by the Reed-Muench method.

Sequence analysis of the PEDV strain.

Based on published primers (8), the PCR products were sequenced by Shanghai Personal Biotechnology Co., Ltd. (Shanghai, China). The 5′ and 3′ ends of the genome of the three PEDV strains were validated using rapid amplification via a cDNA ends (RACE) cDNA amplification kit (TaKaRa, Japan). All fragments were sequenced in both directions in triplicate. The DNAStar software package (DNAStar Inc., Madison, WI, USA) was used to assemble and analyze the sequencing data. The complete genomic sequence and the nucleotide sequence of the S gene of the shxx1902, shxx1912, and SH1302 strains were aligned with sequences of published isolates using MEGA 6 software. Phylogenetic trees were constructed using the neighbor-joining method and supported with a bootstrap test of 1,000 replicates. Similarity plots of the genomes of the PEDV strains identified in the present study were created using the sliding window method, as implemented in the SimPlot, v.3.5.1 package.

Mouse vaccination.

To assess the immunogenicity of the PEDV shxx1902 strain and CV777, mice were inoculated with the PEDV shxx1902 strain and CV777, and the levels of anti-N protein antibodies were evaluated. Viruses were inactivated with 0.2% formaldehyde for 28 h at 37°C and then emulsified 1:1 with an aluminum hydroxide-adjuvanted inactivated vaccine by 15 min of agitation at 200 rpm at 37°C. Twenty-four mice were randomly divided into 3 groups (n = 8) and then inoculated by multipoint subcutaneous injection along the back, with 10⁵ TCID₅₀/mouse of inactivated PEDV shxx1902(P5), CV777, or PBS. Mice were housed in separate cages and provided with adequate feed and water. Mice were boosted at 14 days post-initial inoculation (dpi), with 10⁵ TCID₅₀/mouse of the appropriate inoculum. At 0, 7, 14, 21, 28, 35, 42, and 49 dpi, samples of serum were collected, and indirect enzyme-linked immunosorbent assay (ELISA) was used to detect circulating antibodies. Wells of 96-well flat-bottom microplates were coated with 2 μg/ml of Escherichia coli-expressed N protein for 2 h at 37°C. Wells were then washed three times with PBS and blocked with 5% bovine serum albumin (BSA) for 2 h. Primary antiserum (100 μl of a 1:50 dilution) was aliquoted into the wells and incubated for 2 h at 37°C. Wells were washed again with PBS, and then horseradish peroxidase (HRP)-labeled goat anti-mouse IgG was aliquoted into each well and incubated for 1 h at 37°C. Optical density at 450 nm (OD₄₅₀) values were read after 15 min of color development using tetramethylbenzidine (TMB; Beyotime, China) (13–15).

Statistical analysis.

One-way analysis of variance (ANOVA) was used for the statistical analysis. All statistical analyses were conducted using Prism 8.0 software (GraphPad Software, Inc., La Jolla, CA, USA). A P value less than 0.05 (P < 0.05) was considered statistically significant between the treatment and control groups (**, P < 0.01; ***, P < 0.001).

Data availability.

Genomic sequences of the isolated samples shxx1902 (passage 5 [P5]), shxx1902 (P35), shxx1912 (P5), shxx1912 (P35), SH1302 (P5), and SH1302 (P35) were submitted to GenBank under accession numbers MN841671, MT303066, MT843278, MT843280, MT843279, and MT843277.

RESULTS

Detection of clinical samples.

In our study, a total of 6 intestinal tissues samples and 20 fecal samples were obtained from the diseased piglets from the farm on 12 February 2019. Then, the 10 fecal samples and 5 intestinal tissues samples were obtained from the diseased piglets from the farm on 5 December 2019. The presence of PEDV, transmissible gastroenteritis virus (TGEV), and rotavirus was detected using commercial real-time RT-PCR kits (Suoao Biotech Co., Ltd., Beijing). Out of the 41 diarrheic samples, 40 (97.6%) tested positive alone for PEDV, while all of the samples tested negative for TGEV and rotavirus. These results showed that PEDV played an important role in the etiology of diarrhea. SH1302 detected positive for PEDV was collected from another pig farm in Shanghai, China in 2013.

PEDV propagation in Vero cells.

Suspensions of Vero cell-adapted PEDV(shxx1902, shxx1912, and SH1302) were successfully propagated, and visible CPEs were then observed at the 35th passage of the PEDV-infected adherent cells. Compared with the uninfected Vero cells, the PEDV-infected Vero cells were characterized by clustering and syncytium formation at the initial stage, followed by shrinkage and detachment at 48 hpi. No cytopathic effects occurred in the control wells (Fig. 1). The three virus strains were biologically cloned by three rounds of plaque purification in Vero cells prior to further virus characterization. These results demonstrate that suspension Vero cells are suitable for the PEDV strain to grow.

FIG 1.

Propagation of PEDV strains shxx1902, shxx1912, and SH1302 in Vero cells at 48 hpi.

Electron microscopy.

For transmission electron microscopy, PEDV strains were adsorbed onto freshly discharged 400 mesh carbon Parlodion-coated copper grids (Zhongxingbeirui Co., Ltd., China). The size and morphology of PEDV strains were further examined by negative-staining transmission electron microscopy. As shown in Fig. 2, the virion was circular in shape and 80 to 120 nm in diameter, with surface projections characteristic of coronaviruses.

FIG 2.

Images of PEDV strains shxx1902, shxx1912, and SH1302 particles by electron microscopy, as shown by the arrow. Bar = 100 nm. Magnification: ×50,000.

Immunofluorescence.

Infection of Vero cells with PEDV strains shxx1902, shxx1912, and SH1302 was confirmed by an immunofluorescence (IF) assay (Fig. 3). The three isolated strains were detected using a mouse anti-PEDV S protein monoclonal antibody. IF staining of mock-inoculated Vero cells showed no IF-positive cells. IF-stained cells were visible in the shxx1902-, shxx1912-, and SH1302-inoculated Vero cells.

FIG 3.

Detection of PEDV strains shxx1902, shxx1912, and SH1302 in Vero cells by an immunofluorescence assay at 24 h postinfection; the cells were fixed and incubated with anti-PEDV S monoclonal antibody and DAPI, respectively.

One-step growth curve of the PEDV strains shxx1902, shxx1912, and SH1302 in suspension and adherent Vero cells.

After verification of the 35th generation of PEDV strains shxx1902, shxx1912, and SH1302 by real-time RT-PCR, EM, and IFA, according to the abovementioned methods, a growth curve was generated based on the TCID₅₀ values. The results indicated that the titer of the above three virus strains from the 5th and the 35th passage were almost identical to that of the CV777 vaccine strain. All the virus titers in suspension Vero cells and adherent Vero cells increased from 12 to 72 h and declined at 84 h. The highest viral titer was about 10⁷ TCID₅₀/ml in suspension cell and 10⁶ TCID₅₀/ml in adherent cells (Fig. 4).

FIG 4.

Viral titers of PEDV strains shxx1902, SH1302, shxx1912, and CV777 propagated in suspension Vero cells and adherent Vero cells at the 5th (A) and 35th passage (B).

Analysis of the S gene and whole genome.

Phylogenetic analyses of the whole genomes and S gene revealed that all PEDV strains in this study could be separated into two groups; the shxx1902 (P5) and shxx1912 (P5) strains belonged to GII, which also contained the AJ1102 (GenBank accession no. JX188454) strain, which was the vaccine that is presently used in the farm. However, the two isolated strains and AJ1102 belonged to different subgroups of the same group. The GI group contained the classical PEDV CV777 vaccine strain, as shown in Fig. 5, which contained SH1302 (P5). Through the whole-genome analysis, several strains in China since 2015 (including XM2-4, CH/HNAY/2015, CH-HB2-2018, and JSCZ1601) were shown to be in the same subgroup GII-c with shxx1902 (P5) and shxx1912 (P5). Their S gene belonged to GII-a. However, the whole genome of shxx1902 (the 35th passage) and shxx1912 (the 35th passage) belonged to GII-b and GI-b. Their S gene belonged to GII-a and GI-b. (Fig. 5A and B).

FIG 5.

(A) Phylogenetic analysis of P5 strains and P35 strains (shxx1902, shxx1912, and SH1302) based on the complete genome. P5 strains and P35 strains were indicated by the filled green triangles. Vaccine strains were indicated by the filled green circles (GenBank accession no. KT323979, CV777; JX188454, AJ1102). (B) Phylogenetic analysis of P5 strains and P35 strains (shxx1902, shxx1912, and SH1302) based on the S gene. P5 strains and P35 strains were indicated by filled green triangles. Vaccine strains were indicated by the filled green circles (KT323979, CV777; JX188454, AJ1102).

The sequence analysis showed that the full-length genomic sequence of shxx1902 (P5) and shxx1912 (P5) was 28,051 nucleotides (nt) long, excluding the polyadenylated sequences, and included the following genes: ORF1a (nt 294 to 12602), ORF1b (nt 12602 to 20638), S (nt 20635 to 24795), ORF3 (nt 24795 to 25469), E (nt 25450 to 25680), M (nt 25688 to 26368), and N (nt 26380 to 27705). In addition the full-length genomic sequence of Vero cell-adapted strain shxx1902 (P35) and shxx1912(P35) was 27,974 nt long and included the following genes: ORF1a (nt 294 to 12578), ORF1b (nt 12578 to 20614), S (nt 20611 to 24771), ORF3 (nt 24771 to 25393), E (nt 25374 to 25604), M (nt 25612 to 26292), and N (nt 26304 to 27629). Compared with parent strains, the deletion of 24 nt (8 amino acids [aa]) and 52 nt (133 aa) were situated in ORF1a and ORF3 of the cell-adapted strain. Amino acid sequence identities of the different regions of shxx1902 (P5) and shxx1912 (P5) were 93.4% to 100.0% and 92.1% to 99.4%, respectively, compared with those of P35 (Table 1).

TABLE 1.

Amino acid sequence identities of the different regions of the shxx1902(P5) and shxx1912 (P5) genome compared with that of P35

Strain	Amino acid identity (%) with P35 of gene:
	ORF1a	ORF1b	S	ORF3	E	M	N
shxx1902 (P5)	97.9	100.0	99.7	98.8	93.4	97.3	96.8
shxx1912 (P5)	97.7	99.4	94.4	98.8	92.1	97.3	96.6

The sequence analysis showed that the full-length genomic sequence of SH1302 (P5) included the following genes: ORF1a (nt 282 to 12566), ORF1b (nt 12566 to 20602), S (nt 20599 to 24747), ORF3 (nt 24747 to 25022), E (nt 25353 to 25583), M (nt 25591 to 26271), and N (nt 26283 to 27608). In addition the full-length genomic sequence of Vero cell-adapted strain SH1302 (P35) included the following genes: ORF1a (nt 282 to 12566), ORF1b (nt 12566 to 20602), S (nt 20599 to 24747), ORF3 (nt 24747 to 25022), E (nt 25353 to 25562), M (nt 25570 to 26250), and N (nt 26282 to 27587). Compared with parent strain, the deletion of 21 nt (7 aa) was situated in the envelope protein of the SH1302 cell-adapted strain.

Through the phylogenetic analysis of the full-length genome sequence of PEDV, shxx1902 (P5) and shxx1912 (P5) belonged to GII-c. We performed genome scale similarity comparisons between shxx1902 (P5), shxx1912 (P5), and four other subgroups with SimPlot v.3.5.1, as demonstrated in Fig. 6. This analysis confirmed the chimeric nature of the shxx1902 (P5) and shxx1912 (P5) strains.

FIG 6.

SimPlot analysis for possible recombination events of the shxx1902 (P5) and shxx1912 (P5) genomes compared with those of the other four subgroups. Note that the shxx1902 (P5) strain was set as the query strain; the vertical and horizontal axes indicated the nucleotide similarity percent and nucleotide position (bp) in the graph, respectively.

The immunogenicity of the 5th PEDV shxx1902.

Mice were inoculated two times with aluminum hydroxide-adjuvanted inactivated PEDV shxx1902 (P5), commercial CV777 vaccine strain, or PBS. Because the NP protein is highly abundant in virus-infected cells, it provides an attractive target for the development of antigen-based serological assays (16). Indirect ELISA was used to detect and quantitate circulating antibodies produced against the N protein (Fig. 7). The circulating anti-N antibody was elevated to the same level in shxx1902- and CV777-inoculated mice. Until day 49 postinoculation, anti-N protein antibodies were elevated in both mice. The results showed that the anti-N antibody in shxx1902-inoculated mice was not significantly different from that of CV777-inoculated mice after 21 days postinoculation.

FIG 7.

Porcine epidemic diarrhea virus N protein-specific antibody responses of the two strains PEDV shxx1902 (P5) and CV777 and PBS. Mice at 3 weeks of age were inoculated by multipoint subcutaneous injection with inactivated PEDV shxx1902 (P5) and CV777 and PBS. The serum samples were collected at the indicated time points and were tested by indirect ELISA. Data are shown as the mean OD₄₅₀ ± SD. Differences among the groups were analyzed by one-way ANOVA (**, P < 0.01; ***, P < 0.001).

DISCUSSION

To date, the propagation of PEDV remains a challenge. Many PEDV strains isolated from clinical samples were difficult to culture in Vero cells in recent years (7, 17, 18). In addition, another study successfully isolated and propagated the epidemic PEDV strain using porcine intestinal epithelial cells in vitro; however, this cell line was not available in most labs (19). Virus isolation was also first attempted with 68 clinical samples received at the Iowa State University Veterinary Diagnostic Laboratory which was positive for the U.S. PEDV S-INDEL-variant strain. However, the cell isolation was not successful. Subsequently, this virus strain was isolated using intestine homogenates and cecum contents collected from all 3 pigs inoculated by the virus strain. The research concluded that a low concentration of virus, cytotoxicity of some samples, and variable storage conditions may lead to failure (20). In the present study, we established a promising approach for propagating PEDV using suspension Vero cells. The viral titer of the Vero-adapted PEDV strain shxx1902 (GII-c), shxx1912 (GII-c), and SH1302 (GI) reached 10⁶ to 10⁷ TCID₅₀/ml at passage 35 in adherent cells. The PEDV virus titer usually reached 10⁷ TCID₅₀/ml as the highest value.

The phylogeny of PEDV demonstrates a great deal of diversity in variants, which may lead to limited cross-protection against infection with different strains. The sequence analysis of the whole gene demonstrated that the shxx1902 and shxx1912 isolates belonged to subgroup GII-c, which included PEDV strains isolated in China since 2015. In addition, SH1302 belonged to GI. Since late 2010, highly virulent PEDV GII strains have emerged globally, resulting in heavy losses to the pork industries in numerous countries (14). Phylogenetic studies have revealed that our isolates from 2019 exhibited inconsistent topologies between the spike gene and other genes. In addition, a study by Jiahui Guo et al. confirmed the chimeric nature of the GII-c subgroup (21). This was in accordance with our isolate, which was revealed to belong to GII-c through further analysis with SimPlot v.3.5.1. As a result, detailed analyses of the complete genome sequences of different PEDV strains are essential to understand their relationships.

During extensive passaging of virus in cell culture, most mutations occur and have been correlated with viral attenuation (22). Through Vero cell adaption, there are some variations in the complete genome, which are most similar to the change between the DR13 strain and the attenuated strain. The genome of virulent DR13 is 28,029 nucleotides (nt) long after exclusion of the poly(A) tail, whereas that of attenuated DR13 contains 27,931 nt [excluding the poly(A) tail], which is 95 to 102 nt shorter than those of other sequenced virulent PEDVs (23). In our study, the genome difference between parent strain (P5) and cell-adapted strain (P35) revealed the same change. The three cell-adapted strains (P35) were shorter. Through the adaption process, the whole genome and S gene of GII strains have mutated to a different genotype to adapt to growth in cells. The PEDV S protein is known to play a pivotal role in viral entry and inducing neutralizing antibodies in natural hosts, which makes it a primary target for the development of effective vaccines against PEDV (24, 25). The S gene of the shxx1902 (P5), shxx1912 (P5), and SH1302 (P5) strains was 4,161 nucleotides long, encoding a protein of 1,386 amino acids (aa). The amino acid similarity of the S protein between the CV777 vaccine strain and SH1302 (P5) was 99.4%. The amino acid similarity of the S protein between the CV777 vaccine strain and shxx1902 (P5), shxx1912 (P5), and AJ1102 (P5) was 91.4% to 91.5%. The amino acid similarity of the S protein between AJ1102 (F12), shxx1902 (P5), and shxx1912 (P5) was 96.4%. There was a 4-aa deletion (QGVN) found at position 59 to 62 aa of CV777 and SH1302 (P5) compared with shxx1902 (P5) and shxx1912 (P5). Compared with the CV777 vaccine strain and SH1302 (P5), a 1-aa deletion of D was at position 159 aa of shxx1902 (P5), shxx1912(P5), and AJ1102 (F12). Compared with shxx1902 (P5), a 2-aa deletion was found at position 1388 to 1389 aa of the CV777 vaccine strain, SH1302(P5), and shxx1912(P5), while a 2-aa deletion was found at position 1387 to 1388 aa of AJ1102 (F12). As a result, it will be essential to accumulate a larger pool of data, and the application of reverse genetics will help evaluate PEDV adaptation and pathogenicity.

Effective vaccines can achieve ideal prevention and control of PEDV. The key focus of research was that the main mechanism of protection is mediated by lactogenic immunity in neonatal piglets and that most vaccines for enteric coronaviruses are designed to induce lactogenic immunity by vaccination of the sow (26). In our study, intestine samples were collected from the dead piglets within 1 week after birth. Although these piglets had not been immunized with any vaccines, their mother had been previously immunized with both a live and inactivated PEDV vaccine (AJ1102; GenBank no. JX188454) on days 35 and 20 before birth. Prior to 2017, the CV777 vaccine strain was used on the farm. Prior to the beginning of 2011, PED was generally considered under control or was associated with only mild effects on the swine herds in China. However, PED has subsequently unexpectedly devastated many swine farms, including those in which killed and attenuated vaccines based on CV777 were being used. This suggested that these vaccines were no longer able to confer protection. However, the PEDV outbreak occurred in a pig farm where the new GII genotype vaccine AJ1102 was used. This may have been caused by the biosafety of the farm, the inefficiency of the inactivated vaccine, or evolution of the isolated virus. In our research, both PEDV shxx1902 and CV777 strains induced higher porcine epidemic diarrhea virus N protein-specific antibody responses until day 49 after the first immunization. Therefore, the findings of this study will hopefully facilitate the development of new isolation methods and vaccine strategies, which are expected to be an important step toward improving the prevention of pandemic outbreaks of PEDV.