The animal study proposal was approved by the Institutional Animal Care and Use Committee (IACUC) of the Yangzhou University Animal Experiments Ethics Committee (permit number: SYXK (Su) IACUC 2012-0029). All experimental methods were conducted in accordance with the relevant guidelines for the Administration of Affairs Concerning Experimental Animals approved by the State Council of the People’s Republic of China.

Four 7-day-old ternaries crossbred piglets (Duroc*Landrace*Yorkshire) were infected with PEDV featured with typical clinical symptoms of porcine epidemic diarrhea. In addition, four normal piglets were selected as the control. The herd of pigs was detected for PCR against PEDV, TGEV, PDCoV, and PoRV. All the animals were raised under the same conditions and humanely sacrificed by intravenous injection of pentobarbital sodium. After the sacrifice, duodenum, jejunum, and ileum samples were snap frozen immediately in liquid nitrogen for the subsequent experiments.

Jejunal contents were collected and diluted in 500 μL phosphate-buffered saline (PBS) buffer, then freeze-thawed repeatedly before the centrifugation to collect the supernatant containing viruses. The supernatants were added to the tubes containing TRIzol reagent (Takara Biotech, Dalian, China) and vortexed according to the manufacturer’s instruction. The cDNA was amplified with PrimeScript RT-PCR kit (Vazyme Biotech, Nanjing, China) to identify the infections of viruses, including PEDV, TGEV, PDCoV and PoRV. The primers are listed in Table 1 .

Duodenum, jejunum, and ileum were fixed in 4% paraformaldehyde solution for 24 hours. Tissues were embedded in paraffin and cut into 5 μm pieces. For hematoxylin-eosin (HE) staining, the paraffin sections were conducted according to the routine procedure. The mounted slides were observed under Olympus iX53 (Tokyo, Japan) light microscope and the photographs were taken using Motic Image Advanced 3.2 Analysis System (Motic China group Co., Ltd., Xiamen, China).

Porcine intestinal epithelial cells (IPEC-J2) were kindly provided by the University of Pennsylvania. Vero kidney cell line (Vero CCL-81) were purchased from ATCC. IPEC-J2 and Vero cells were cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM, Gibco, NY, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, NY, USA) and 100 μg/mL penicillin/streptomycin (Solarbio, Beijing, China) in a humidified atmosphere containing 5% CO₂ at 37°C. Cells were seeded in tissue culture flasks (Costar, Corning, MA, USA).

PEDV strains used in this study were the classic strain CV777, which was propagated and titrated in FBS-free DMEM medium supplemented with 8 μg/mL trypsin (Sigma-Aldrich, MO, USA) in Vero CCL-81 cells. Cells were infected with PEDV at MOI of 0.1 for 2 hours at 37°C (26). After 2 hours, the unattached virus inoculum was removed. Then, the cells were washed three times with cold PBS buffer and maintained in DMEM supplemented with 2% FBS. The infected cells and supernatants were collected after the indicated period and titrated according to TCID₅₀ protocol. The viral titers were calculated using the Spearman-Karber equation. The PEDV genome copy number was calculated by the standard curve equation of CV777 well-established previously (26): y = −3.3354lg(x) + 37.832 (where y represents the number of Ct cycles, x represents the logarithm of the virus copy number based on the base 10).

To deplete TFF1 gene, three different single guide RNAs (sgRNAs) were designed using an CRISPR-Cas9 Design online software. Oligo nucleotides sequences of the sgRNAs were synthesized and cloned into the BbsI digested linearized pGK1.2 (EGFP + Puro^r) vector (Genloci Biotech, Nanjing, China) following the pGK1.2 oligo-cloning protocol. The primers are listed in Table 2 . Colony PCR was performed to validate the sgRNA sequences using the VSP forward primer 5′-CATATGCTTACCGTAACTTGAAAG-3′ and three primer sets: primer set-1: (5’-GCCACAGGATTGAAGCACCA-3’), primer set-2: (5’-GTTGTCTGTGGGGTCATCAA-3’), primer set-3: (5’-GTGGGGTCATCAACGGCCAC-3’) as the reverse primers, respectively. The amplicons were performed by PCR-sequencing (Sangon Biotech, Shanghai, China) and the positive recombinant vectors were named as sgRNA1, sgRNA2 and sgRNA3, respectively. Subsequently, cells pool sequencing verified the efficiency of sgRNAs mixture. The primer used for cloning was as follows: TFF1-sgRNA forward primer (5’-CTTGGCCGGTACACTTTCAG-3’) and reverse primer (5’-CTTCACAGGTCCGTGGTTAGA-3’). IPEC-J2 were transfected with sgRNA1, sgRNA2 and sgRNA3 in 6-well tissue culture plates at a density of 2 × 10⁵/mL. The culture medium with 3 μg/mL puromycin was added to the cells 72 hours later. For another 72 hours, positive monoclonal cells were harvested.

For TFF1 and C/EBPα overexpression, the full-length CDS of TFF1 and C/EBPα were amplified and cloned into a pcDNA3.1 vector (Invitrogen, CA, USA) with T4 DNA ligase for ligation by NheI and HindIII restriction sites, respectively. The recombinant overexpression vectors were defined as TFF1-oe and C/EBPα-oe, respectively. Subsequently, each of the overexpression vectors were transfected into IPEC-J2 using Lipofectamine™ 2000 (Invitrogen, CA, USA) following the manufacturer’s protocol. At 36 hours after transfection, the cells were passaged into the appropriate medium containing 400 µg/mL G418. The screening was conducted over 7 days and the stable cell lines constitute polyclonal pools of cells.

Small interfering RNAs (siRNAs) against C/EBPα mRNA sequences (si-C/EBPα-1, si-C/EBPα-2, and si-C/EBPα-3) of sus scrofa and negative control (si-C/EBPα-NC) were synthesized from GenePharma Corporation (Shanghai, China). The siRNAs and negative control were transfected into IPEC-J2 using Lipofectamine™ 2000 (Invitrogen, CA, USA) following the manufacturer’s instruction.

To assess cell viability, cells were seeded in 96-well plates with 5 × 10³/well and incubated with PEDV at the indicated time. Then 10 μL cell counting kit-8 solution (CCK8, Dojindo Molecular Technologies Inc., Kumamoto, Japan) was added to each well followed by the incubation at 37°C for 2 hours. After the incubation, an absorbance at 450 nm was measured using a Multimode Microplate Reader (Spark™ 10M, Tecan GmbH, Austria). Caspase-3/7 activity was measured using a luminescence-based assay with the luminescent Caspase-Glo 3/7 assay kit (Promega Corporation, MA, USA) according to the manufacturer’s instruction. Each experiment was repeated three times and performed in triplicate.

Cells were seeded in 6-well plates and incubated with the indicated doses of PEDV. The apoptotic cells were detected by Annexin V-FITC/PI apoptosis analysis kit (Solarbio, Beijing, China) according to the manufacturer’s guide. The samples were analyzed by CytoFLEX flow cytometer (Beckman Coulter, CA, USA). For cell cycle detection, TFF1-depletion cells were seeded in tissue culture flask at a density of 2 × 10⁶/mL. Cells were then washed with cold PBS and fixed overnight with cold 70% ethanol. The fixed cells were washed with PBS and stained with 1 mL PI staining reagent (50 mg/mL propidium iodide and 1 mg/mL RNase A in 1 mL of sodium citrate buffer) for 30 min at 37°C in a light-proof manner. Then the cell cycle profiles were analyzed by FACScan flow cytometer (Becton Dickinson, CA, USA). The percentage of cells at G0/G1, S and G2/M phases were calculated using Modifit software.

Cells were seeded on coverslips in 12-well tissue culture plates and infected with PEDV for 2 hours, rinsed, then incubated at 37°C. After 48 hours of PEDV infection, the cells were fixed in 4% paraformaldehyde for 15 min and permeabilized by 0.1% Triton X-100 in PBS for 10 min at RT. The cells were blocked with 5% bovine serum albumin (BSA) in PBS for 30 min at RT and then incubated with the anti-PEDV antibody PEDV N MAb (Medgene, SD-1-5, 1:500) overnight at 4°C. After washing thrice with PBS, the cells were incubated with a fluorochrome-conjugated secondary antibody (HuaBio, HA1015, 1:100) for 30 min at RT. The cell nuclei were stained with DAPI (1 μg/mL) for 5 min. The coverslips were mounted on microscope slides for the visualization under a fluorescence microscope (Olympus, Tokyo, Japan). Images were processed using ImageJ software.

Cells were treated with PEDV at the indicated time as above. RT-qPCR was performed to detect the mRNA expression of inflammatory cytokines (IL-2, IL-6, IL-8, IL-12, IL-α, IL-β, IFN-α, IFN-β, TNF-α) and antiviral genes (Mx1, Mx2, RIG). The primers are listed as shown in Table 3 .

Total RNA was isolated from cells and intestinal samples using the TRIzol reagent (Takara Biotech, Dalian, China) following the manufacturer’s protocol. RT-qPCR amplification was performed using an ABI 7500 Fast Real-Time Quantitative PCR System (Applied Biosystems, Foster City, CA, USA) and relative gene expression was calculated using the 2^−ΔΔCt method (27). The housekeeping gene GAPDH was used as an internal control for each experiment. The experiments were performed at least three times with data presented as means values ± SD. The primers are shown in Table 4 .

Intestinal tissues and cells were lysed in radioimmunoprecipitation assay (RIPA) buffer that contains proteinase inhibitor cocktail on ice for 20 min. The supernatant was gathered, centrifuged at 12,000 g for 20 min, and the protein concentration was determined using the BCA Protein Assay Kit (CWBiotech, Beijing, China). Proteins were separated in SDS-PAGE and transferred to a PVDF membrane (Millipore, MA, USA). The membranes were blocked with 5% skimming milk in Tris-buffered saline-Tween (TBST) for 1 hour, then incubated at 4°C overnight with the primary antibodies [anti-TFF1 (LifeSpan BioScience, ls-c312842, 1:1000), anti-CDK4 (Proteintech Ltd, 11026-1-AP, 1:1000), anti-PCNA (Proteintech Ltd, 10205-2-AP, 1:1000), anti-Bax (HuaBio, ET1603-34, 1:1000), anti-Cleaved Caspase 3 (Cell Signaling Technology, #9664, 1:500), anti-PEDV-N (Youlong-Bio, DA0110,1:1000), anti-β-Actin (CWBIO, CW0096M, 1:1000) and anti-GAPDH (Proteintech Ltd, 10494-1-AP, 1:1000)], respectively. The membranes were rinsed and incubated with the HRP-conjugated secondary antibodies [goat anti-Rabbit IgG antibody (Abcam, ab205718, 1:20000) and anti-Mouse IgG antibody (HuaBio, HA1006, 1:10000)] for 1 hour at RT. The membranes were visualized with an Enhanced Chemiluminescent Detection kit (ThermoFisher Scientific, MA, USA) using the FluorChem FC3 Chemilumilescent system (ProteinSimple Ltd, CA, USA). The relative integrated density was measured and normalized against GAPDH or β-Actin expression. The experiments were performed at least three times with data presented as means values ± SD.

The genomic DNA was extracted from jejunum in diarrheic and normal piglets using TIANamp Genomic DNA Kit (Tiangen Biotech, Beijing, China). The extracted DNA was bisulfite conversed using EZ DNA Methylation-Gold Kit (Zymo Research, CA, USA). The conversion DNA was performed PCR amplification using ZymoTaq Premix, with the primers designed for amplification presented in Table 5 . The amplification system contains 2 μL DNA, 12.5 μL ZYMO Taq Premix, 1 μL forward primer and 1 μL reverse primer (10 pmol/μL) and RNase-free H₂O 8.5 μL. The following PCR reaction was performed to 95°C for 10 min, followed by 40 cycles with each cycle consisting of 95°C for 30 s, and 54°C for 30 s, extension at 72°C for 30 s, and a final 5 min extension at 72°C. The PCR products were purified by TIANquick Midi purification kit (Tiangen Biotech, Beijing, China) and ligated into pMD19-T vector (Takara Biotech, Dalian, China) at 16°C overnight and transformed into Escherichia coli DH-5α competent cells. More than 15 positive clones from each sample were randomly picked up for bisulfite-sequencing. The methylation status at each CpG site was aligned using QUMA online software.

The CpG islands in the promoter region of TFF1 were predicted by MethPrimer website. The parameters were set as follows: CpG islands length greater than 200 bp, GC content greater than 50%, and CpG o/e greater than 0.6. For the reporter-gene assays of methylation on TFF1 promoter, the CpG island fragment of TFF1 promoter (−145 to −478 bp) was ligated to the pGL3-basic vector. The vectors were treated with M.SssI (CpG) methyl transferase (New England Biolabs, Beijing, China) and transfected into IPEC-J2. Luciferase assays were performed 36 hours after transfection. The primer for CpG islands amplification is presented in Table 5 . IPEC-J2 were treated with concentration of 0.1 μM 5’-Aza-2′-deoxycytidine (5’-Aza-2’-dC, Sigma-Aldrich, MO, USA) for 72 hours. Then the cells were harvested for mRNA expression analysis.

ChIP assay was performed using the Pierce Agarose ChIP Kit (ThermoFisher Scientific, MA, USA) following the manufacturer’s instructions. An amount of 80 mg sheared jejunum sample was minced and crosslinked in formaldehyde. After the chromatin digested with micrococcal nuclease, the crude chromatin fragments were immunoprecipitated with the specific anti-C/EBPα antibody (Santa Cruz, sc-365318, 1:500) overnight at 4°C and then incubated with protein-G magnetic beads. Next, the enriched chromatin was purified for ChIP-PCR amplification and the amplified products were examined with 2.0% agarose gel electrophoresis. The primer is presented as follows: forward primer (5’-CGCAGCATCTCTGCTGTGAA-3’) and reverse primer (5’-ACCCTCCCGCTAAGTCAACA-3’).

The probes for electrophoretic mobility shift assay (EMSA) were as follows: the wild-type probe (5′-ACCGCGCTGGCGCAGCATCT-3′) and the mutant-type probe (5′-ACCGCATCAATATGACATCT-3′). Briefly, an oligonucleotide containing the C/EBPα motif and its complementary sequences were synthesized and then annealed to double-stranded structures. The wild-type probe was treated by methyltransferase M.SssI (New England Biolabs, Beijing, China). Total nuclear proteins were extracted from trypsinized monolayers of IPEC-J2 by a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, Nanjing, China). The double-stranded oligonucleotide was 5′ end-labeled with biotin-dUTP. For each binding reaction, 2 μg labeled probes, 2 μg nuclear extract, 4 μg poly (dI–dC), 10 μL binding buffer [20 mM Tris-HCl (pH 7.6), 50 mM KCl, 1 mM DTT, 0.5 mM EDTA, and 10% glycerol] were incubated at 25°C for 30 min. Then the protein–DNA complexes were separated on non-denaturing 6% polyacrylamide gels with precooled 0.5 × TBE buffer. For the super-shift assay, 1 μL anti-C/EBPα antibody (Santa Cruz, sc-365318, 1:500) was added to EMSA reaction mixtures and incubated for 60 min at 4°C prior to the labeled probe addition. Then, the protein–DNA complexes were transferred to nylon membrane at 60 v for 1 hour and autoradiography of the dried gel was detected immediately after UV cross-linking.

The TFF1 promoter sequences were analyzed by Alibaba2 online software to predict the potential transcription factor binding sites (TFBS). According to the core promoter prediction of BDGP online software, the PCR purified products were digested with KpnI and XhoI and ligated to the pGL3-basic luciferase reporter vector. Four plasmids with truncated promoters were constructed by inserting the DNA sequences of the TFF1 promoter ranging from −1200 to −1. The truncated recombinant vectors were defined as pGL3-control (−200 to −1), pGL3-TFF1-p1 (−360 to −1), pGL3-TFF1-p2 (−500 to −1) and pGL3-TFF1-p3 (−1200 to −1).

For the reporter-gene assays, pGL3-TFF1-wt was constructed by inserting the DNA fragments of the TFF1 promoter from −360 to −200 into pGL3-basic vector. The mutant form pGL3-TFF1-mut contains the sequences mutated from GCTGGCGCAG to ATCAATATGA. In addition, the sequences GCTGGCGCAG enrolled in pGL3-TFF1-wt was deleted and named as pGL3-TFF1-del. The IPEC-J2 were co-transfected with C/EBPα-oe and wild type, mutant or deletion forms of TFF1 promoter reporter constructs. The renilla was co-transfected for normalization. After 36 hours post-transfection, luciferase activity was analyzed with a Dual-Luciferase Assay system (Promega Corporation, Madison, USA) on a luminometer according to the manufacturer’s instruction. The relative fluorescence intensity was calculated as Firefly-Luc (Ff)/Renilla-Luc (Rn). All the transfections were performed in sextuplicate and each experiment was repeated at least three times.

Statistical analysis was performed by GraphPad Prism 8.0 software. All statistical details of experiments were included in the figure legends. All experiments were repeated at least three times and the results were shown as mean values ± SD.

The PCR products were amplified to detect PEDV, TGEV, PDCoV, PoRV in the small intestinal samples. The specific PCR products were only amplified in the infected samples with PEDV classic strain CV777 ( Figure 1A ). Our results show that PEDV is the unique infectious agent detected in diarrhea piglets ( Supplementary Figure 1A ). Histological analysis reveals that PEDV infection led to villi atrophy and partial shedding of villous epithelial cells ( Figure 1B ). The TFF1 mRNA expression was shown at higher levels in the porcine duodenum, jejunum, and ileum ( Figure 1C ). To further validate the relationship between TFF1 biological function and PEDV replication, we examined the mRNA expression of the TFF1 in vitro and in vivo, respectively. As shown in Figure 1D , the mRNA expression of TFF1 in jejunum and ileum was significantly up-regulated in the PEDV group compared to that of control (P < 0.01), and TFF1 protein expression was also significantly increased in jejunum (P < 0.01, Figure 1E ; Supplementary Figure 1B ). Furthermore, the mRNA expression of TFF1 was significantly up-regulated at 48 hours post infection of PEDV (P < 0.01, Figure 1F ). These results provide evidence that TFF1 is associated with the host responses toward PEDV infection.

To verify the function of TFF1 during PEDV replication, we constructed TFF1 knockout and overexpression IPEC-J2 cell lines ( Figures 2A–D ; Supplementary Figures 2A–E ). We found that knockout of TFF1 markedly up-regulated the mRNA expression of M gene, whereas TFF1 overexpression down-regulated its expression upon PEDV infection ( Figure 2E ). The knockout also resulted in an enhanced viral particle, while TFF1 overexpression led to a marked reduction of PEDV infectiousness ( Figures 2F, G ). Similarly, TFF1 knockout significantly increased PEDV genome copy number (P < 0.01, Figure 2H ). In addition, the cell culture supernatants were titrated for TCID₅₀ and it reveals that the infectious titers were significantly enhanced in TFF1-knockout cells (P < 0.01, Figure 2I ). Overall, these results suggest that the loss of TFF1 expression exacerbates PEDV replication.

To investigate the effects of TFF1 depletion on the growth and survival of IPEC-J2, we found that TFF1 knockout decreased the viability of IPEC-J2 and promoted the development of cell lesion in a time-dependent manner ( Figures 3A, B ). Similarly, the knockout of TFF1 also resulted in a poor survival of IPEC-J2 as measured by CCK8 and caused the pronounced apoptosis after PEDV infection, as reflected by the activation of caspase3/7 with flow cytometry analysis (P < 0.01, Figures 3C, D ). Moreover, TFF1 knockout led to the significantly decreased - proteins which are important for cell proliferation, growth, and survival, along with an increased levels of Bax and Cleaved-Caspase 3 ( Figure 3E ). Additionally, we found that TFF1 knockout reduced the S-phase but increased the cell growth in the G0/G1 phase ( Figure 3F ). The expression of mRNA and protein for cell cycle genes were significantly down-regulated ( Figures 3G, H ). Furthermore, the mRNA expression of proinflammatory cytokines and antiviral genes were significantly up-regulated after TFF1 knockout ( Figures 3I, J ). These results indicate that TFF1 depletion alters the cell cycle and exacerbates PEDV-induced cell death and inflammatory responses.

The CpG islands (located at −477 to −299 bp) were predicted by MethPrimer ( Figure 4A ; Supplementary Figure 3A ). As shown in Figure 4B , bisulfite-sequencing results reveal that there was different methylation status of 12 CpG sites. Methylation analysis demonstrates that mC-2 loci and mC-6 loci presented hypomethylation status compared to that of the control (P < 0.05, Figure 4C ). Pearson correlation analysis shows that there was a significantly negative correlation between methylation status and TFF1 mRNA expression in the mC-6 loci (R =−0.78, r_0.05 = 0.707, P =0.022, Figure 4D ). According to the predicted core promoter sequences, we constructed four truncated recombinant vectors ( Figure 4E ; Supplementary Figure 3B ). As shown in Figure 4F , luciferase reporter assay shows that the transcriptional activity of pGL3-TFF1-p1 was remarkedly higher than that of pGL3-control (P < 0.01), indicating that the core promoter region of TFF1 was located at −360 to −200 bp. To investigate the epigenetic mechanisms of TFF1 regulation, we first generated a methylated vector DNA substrate with the methyl transferase of M.SssI. Then, we transiently transfected the C/EBPα overexpression vector that failed to activate methylated reporter of TFF1 compared to the unmethylated reporter ( Figure 4G ). However, the mRNA expression of TFF1 was markedly upregulated after 5’-Aza-2’-dC treatment compared to that of control, indicating that DNA methylation inhibition strongly enhanced gene expression of TFF1 ( Figure 4H ). Furthermore, M.SssI or 5’-Aza-2’-dC treatment before PEDV infection considerably regulated PEDV entry in IPEC-J2. As shown in Figure 4I , PEDV invasion was significantly inhibited by 5’-Aza-2’-dC, while the expression of M gene was significantly upregulated with M.SssI treatment. These results suggest that the hypermethylation of TFF1 facilitates PEDV replication.

Having demonstrated the crucial role of TFF1 in IPEC-J2, we next examined whether C/EBPα participated in the transcriptional regulation of TFF1. As shown in Figure 5A , we further predicted the potential transcription factors in 12 CpG sites of TFF1 promoter, in which mC-6 loci was found to be within the binding domain of C/EBPα. We then performed the reporter-gene assays with the promoter of TFF1 and found that it was highly responsive to C/EBPα-mediated transactivation. The mutation or deletion of the putative C/EBPα binding site effectively diminished the C/EBPα-dependent activation ( Figures 5B, C ). Using transient co-transfection and reporter gene assays, we found that overexpression of C/EBPα resulted in transactivation of the TFF1 gene promoter, suggesting that C/EBPα can act as an activator of TFF1 ( Figure 5D ). Finally, a positive DNA fragment was detectable in the products amplified with chromatin fragments that were precipitated using specific anti-C/EBPα antibody ( Figure 5E ). Together, the data indicate that C/EBPα activates TFF1 target promoters via the putative C/EBPα binding site. To investigate the influence of DNA methylation on the nuclear protein binding properties, we performed EMSA using 5′-biotinylated probes corresponding to putative TFF1 sequences including the mC-6 loci. Specifically, two different double-stranded oligonucleotides (wild or mutant type) were treated with or without methyl transferase M.SssI and incubated with nuclear extracts from IPEC-J2. Thereafter, the wild-type probes resulted in significantly increased mobility shift of nuclear proteins compared with the mutant-type and methylated wild-type probes. The shift can be competed by an excess of unlabeled 5′-biotinylated probes which restrain C/EBPα binding to TFF1. Additionally, the specific binding of C/EBPα to the TFF1 was further confirmed using an anti-C/EBPα antibody in a super-shift assay ( Figure 5F ). These results show that methylation at mC-6 loci blocks the C/EBPα binding to TFF1 promoter region.

To examine the regulatory role of C/EBPα in PEDV infection, the cells with PEDV treatment resulted in a similar up-regulation of C/EBPα with that of TFF1 gene ( Figure 6A ). Therefore, siRNAs of C/EBPα were transfected into IPEC-J2 ( Figure 6B and Supplementary Figure 4A ). Interestingly, C/EBPα silencing played an important role in the viral replication, exhibiting that mRNA expression of M gene was significantly up-regulated ( Figure 6C ). Subsequently, C/EBPα overexpression vector was constructed ( Supplementary Figures 4B, C ) and co-transfected with promoter region of TFF1 significantly reduced the invasion rates of PEDV ( Figure 6D ), indicating that the C/EBPα plays an effective role in antiviral target activator on regulating TFF1 expression.