2. Materials and Methods

2.1. Ethics Statement
All procedures involving animal experiment were reviewed, approved and conducted in strict accordance with the Institutional Animal Care and Use Committee (IACUC) of National Taiwan University (Taiwan, Republic of China) with the approval No.: NTU105EL-00160.

2.2. Cells and Viruses
Vero C1008 cells (ATCC No. CRL-1586) were maintained in growth medium containing Dulbecco’s modified Eagle’s medium (DMEM, Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine protein (FBS), 250 ng/mL Amphotericin B, 100 U/mL Penicillin and 100 μg/mL Streptomycin. The recombinant viruses (iPEDVPT-P5 and iPEDVPT-P96), and the chimeric viruses (iPEDVPT-P5-96S and iPEDVPT-P96-5S), were propagated in post-inoculation medium (PI medium) containing DMEM supplemented with 0.3% tryptose phosphate broth (TBP), 0.02% yeast extract, and 10 μg/mL trypsin as described previously [18].

2.3. Generation and Recovery of Recombinant iPEDVPT-P5, iPEDVPT-P96, iPEDVPT-P5-96S and iPEDVPT-P96-5S Viruses
The strategy used to recover iPEDVPT-P96 has been described previously [19]. The approach to constructing a cDNA clone of iPEDVPT-P5 was technically identical to that of the iPEDVPT-P96. However, we split the plasmid B into two fragments because the sequence remained toxic to the One Shot™ TOP10 Chemically Competent E. coli cells (Invitrogen, Carlsbad, USA) despite propagation in LB broth supplemented with 10% SOC medium and being incubated at 30 °C. To generate chimeric viruses carrying heterologous spike (S) genes, namely the iPEDVPT-P5-96S and iPEDVPT-P96-5S, cDNA clones of iPEDVPT-P5 and iPEDVPT-P96, respectively, were used as the backbones. The sequences covering the complete S gene of each virus were exchanged without disruption to the remaining genomic structure. Sequence differences in the S genes of iPEDVPT-P5 and iPEDVPT-P96 viruses are summarized in Table 1. Each plasmid was digested with corresponding type-IIS restriction enzymes as designated in Figure S1, gel-purified, assembled and phenol-chloroform extracted to generate the full-length cDNAs. The cDNAs were then in vitro transcribed to the full-length RNA transcripts using a mMessage mMachine T7 transcription kit (Ambion, Austin, CA, USA) and immediately electroporated into 800 μL of 107 cells/mL Vero cells in RNase-free phosphate buffered saline (PBS) along with 5 μg of PEDV nucleocapsid (N) transcripts. After electroporation, the cells were allowed to recover in growth medium for approximately 16 h and then maintained in PI-medium until cytopathic effects involved over 90% of cell monolayers. The whole flasks were subjected to one freeze-and-thaw cycle and the rescued viruses were passaged once to generate viral stocks (P1). The viral stocks were titrated on Vero cells in 96-well plates to determine the viral titer (see below).

2.4. In Vitro Characterization of Recombinant iPEDVPT-P5, iPEDVPT-P96, iPEDVPT-P5-96S and iPEDVPT-P96-5S Viruses

2.4.1. Immunofluorescence Assay and Syncytia Analysis
Immunofluorescence assay (IFA) was performed to detect PEDV antigens as previously described with modifications [18]. Briefly, Vero cells in 96-well plates (1.75 × 104 cells/well) were infected with the designated P1viruses at a multiplicity of infection (MOI) of 0.005. At 18 h post-infection, cells were fixed with 80% ice-cold acetone, air-dried, and then incubated with an in-house anti-PEDV S antibody, P4B [20], at a dilution of 1:1000 at room temperature (RT) for 1 h (h). After being washed three times with phosphate-buffered saline (PBS), the FITC-conjugated monoclonal goat anti-mouse-IG antibody (BD Pharmingen, San Jose, CA, USA) was applied at a dilution of 1:500 at RT for 1 h. Following the final wash step, the cells were counterstained with mounting medium with 4′,6-diamidino-2-phenylindole (DAPI; Abcam, Cambridge, MA, USA) in the dark for 1 min. Images were visualized and captured using ZOE fluorescent cell imager (Bio-Rad, Hercules, CA, USA). Syncytia analysis were performed concurrently along with IFA by calculating the number of nuclei per syncytium.

2.4.2. Sequence Analysis
Sequence analysis was conducted as described previously [18,19] and two primer pairs (SF-7: ACTCTCGACTGGACATTC and 2R: CAGACTTCGAGACATCTTTG; 5FR-3: ATTAGAGCGATTCTCCATGAC and 5FR-6: TACACACATTGTGGTGCTATTGAG) targeting the C-terminal end of the S gene, which contained both naturally occurred and artificially introduced marker mutations (Figure 1, asterisks and Table 1), as well as the non-structural protein 15 (nsp 15) gene, which contained a naturally occurred mutation (G19470T) were used to verify the identities of the four P1 viral stocks and the recombinant viruses shed in feces per group at the time point of peak fecal viral shedding.

2.4.3. Growth Kinetics, Viral Titration and Plaque Assay
Confluent monolayers of Vero cells were seeded onto six-well plates (5 × 105 cells/well) and infected with each virus at a MOI of 1 and 0.001 at 37 °C for 1 h in triplicates. The cells were then washed twice with Dulbecco’s phosphate-buffered saline (DPBS) and maintained in PI medium. The supernatants at indicated time points were collected and proceeded for viral quantification on Vero cells in 96-well plates using the standard 50% tissue-culture infectious dose (TCID50) assay. In brief, Vero cells in 96-well plates were washed twice with DPBS and then incubated with a 10-fold serially diluted culture supernatant acquired from the aforementioned six-well plate at 37 °C for 1 h. After absorption, the inoculum was removed and replaced with fresh PI-medium following one wash step. The titers were determined at 72 h post-infection using the Reed–Muench method [21].
Plaque assays were performed as previously described [19] to characterize plaque morphologies. Briefly, after absorption of PEDVs at an MOI of 0.0001, confluent monolayers of Vero cells in six-well plates were washed twice with DPBS and then covered with an overlay of pre-warmed PI medium containing 1% agarose. After solidification of the overlays, the plates were incubated at 37 °C for 72 h to produce distinct plaques. The cells were then fixed in 3.16% neutral-buffered formalin for 1 h before removing the semisolid overlays. The plates were stained with 1% crystal violet in 20% ethanol and distilled water for 1 min. Viral plaques were inspected after washing off the crystal violet solution, rinsing the plates with water, and air-drying at room temperature (RT).

2.5. Animal Experiment
Thirty-seven, six-day-old, Large White × Duroc, crossbred, fecal PEDV and TGEV shedding-negative suckling piglets were purchased from a conventional pig farm devoid of G2b PEDV infection history based on the negative result of our long-term surveillance of serum antibody and colostrum against PEDV in this pig farm. These piglets from different sows were fed with artificial milk and were randomly assigned to five groups, acclimated for one day, and then inoculated orally with indicative recombinant viruses at a dose of 2 mL of 0.5 × 102 TCID50/mL or PI-medium, respectively. Clinical signs and weight gain were recorded daily. Fecal consistency was monitored daily and scored visually as: 0 = normal, 1 = loose, 2 = semi-fluid, and 3 = watery as previously described. Calculation of average daily weight gain was only performed on piglets that were not humanly euthanized. The formula used for calculation is as follows: Weight gained/ surviving period.

2.5.1. Quantification of PEDV Fecal Viral Shedding
Methods to quantify fecal PEDV viral shedding has been described previously [19]. Briefly, fecal samples collected from rectal swabs were resuspended in DPBS and then subjected to automated nucleic acid extraction using Cador Pathogen 96 QIAcube HT Kit with QIAcube (Qiagen Inc., Hilden, Germany) according to the manufacturer’s instructions. Complementary DNA was synthesized via reverse transcription using QuantiNova™ Reverse Transcription kit (Qiagen Inc., Hilden, Germany) and proceeded to quantitative real-time PCR analysis using the primer-probe set published previously on a CFX96 Thermal Cycler (Bio-Rad, Hercules, CA, USA). The thermal profile comprised an initial denaturation at 95 °C for 2 min and then 45 cycles of 95 °C for 15 s followed by 60 °C for 15 s. The detection limit of the assay was determined by generating standard curves from serial 10-fold dilutions of known amounts of in vitro transcribed RNA followed by reverse transcription and real-time PCR quantification as described above. The detection limit was calculated as 4.8 log10 RNA copies per mL.

2.5.2. Histopathology and Immunohistochemistry
At three days post-inoculation, three pigs from each virus-treated group and one pig from mock group were humanely euthanized by electrocution followed by exsanguination for histopathological and immunohistochemical assessments, as described previously [22]. Duodenum, jejunum, ileum, cecum, colon, rectum and mesenteric lymph nodes were collected, formalin-fixed, paraffin-embedded, sectioned at 4 μm, and stained routinely with hematoxylin and eosin (H&E) for morphometric analysis by assessing the ratio of villi height to crypt depth blindly by one veterinary pathologist.
Immunohistochemistry was performed to evaluate the distribution of PEDV antigen. Briefly, formalin-fixed paraffin-embedded tissues were sectioned at 4 μm, deparaffined in xylene, rehydrated in serially diluted ethanol, and proceeded to epitope retrieval with the Trilogy antigen retrieval system (Cell Marque, Rocklin, CA, USA). After being washed three times with Tris-buffered saline plus 0.1% Tween 20 (TBST), tissue slides were treated with 3% hydrogen peroxidase (KYB, New Taipei City, Taiwan) and 10% normal goat serum (Dako, Carpinteria, CA, USA) to block the endogenous peroxidase activity and non-specific signals, respectively. For antigen detection, an in-house anti-PEDV N antibody, DE-1, at a dilution of 1:1000 in 10% normal goat serum was applied to the slides for 1 h at RT followed by three times wash with TBST. The first antibodies were then captured using the polyclonal anti-rabbit/mouse immunoglobulin, EnVision-DAB+ system (Agilent Technologies, Santa Clara, CA, USA) at RT for 1 h and color was developed afterward with 3, 3′-diaminobenzidine (DAB) chromogen (Agilent Technologies, Santa Clara, CA, USA). The slides were counterstained with hematoxylin (MUTO, Tokyo, Japan), mounted in Entellan (Merck, Darmstadt, Germany) and cover slipped. Positive signals were visualized under an inverted light microscope (Nikon, Tokyo, Japan).

2.6. Statistical Analysis
All values were expressed as the mean standard ± deviation (SD). Comparison of syncytia size and villous height to crypt depth (VH:CD) ratio were analyzed using statistical software GraphPad Prism 6.0 (GraphPad Prism Inc., San Diego, CA, USA). Variables were compared using the non-parametrical Kruskal–Wallis test; p < 0.05 was considered statistically significant.