The Identification and Characterization of Two Novel Epitopes on the Nucleocapsid Protein of the Porcine Epidemic Diarrhea Virus

Porcine epidemic diarrhea virus (PEDV) is a highly contagious coronavirus that causes severe diarrhea and death, particularly in neonatal piglets. The nucleocapsid protein (N protein) of PEDV presents strong immunogenicity and contributes to the cross-reactivity between PEDV and TGEV. However, the characterization of epitopes on the PEDV N protein remains largely unknown. Here, two monoclonal antibodies (MAbs) specific to the N protein of a PEDV strain, FJzz1/2011, were generated and screened against a partially overlapping library of 24 GST-fusion N protein-truncated constructs. We confirmed that residues 18–133 (designated NEP-D4) and residues 252–262 (designated NEP-D6) were the epitopes targeted by MAbs PN-D4 and PN-D6, respectively. Sequence analysis revealed that these two epitopes were highly conserved among PEDV strains but were significantly different from other members of the Coronavirinae subfamily. Western blot analysis showed that they could be specifically recognized by PEDV antisera but could not be recognized by TGEV hyperimmune antisera. Indirect immunofluorescence (IFA) assays confirmed no cross-reaction between these two MAbs and TGEV. In addition, the freeze-thaw cycle and protease treatment results indicated that NEP-D4 was intrinsically disordered. All these results suggest that these two novel epitopes and their cognate MAbs could serve as the basis for the development of precise diagnostic assays for PEDV.

Scientific RepoRts | 6:39010 | DOI: 10.1038/srep39010 Fine epitope mapping of PEDV N protein. To identify the antigenic determinants that could be recognized by MAbs PN-D4 and PN-D6, two mutually overlapping GST-fusion-truncated constructs (GST-N1 and GST-N2) that span the complete N protein were expressed for detection by MAbs using ELISA ( Fig. 2b and c). MAb PN-D4 recognized only GST-N1, whereas MAb PN-D6 only recognized GST-N2 (Fig. 2c). To locate the precise positions of the two epitopes, as depicted in Fig. 2a, N1 and N2 were further truncated into two discrete groups of GST-fusion constructs, which were detected by PN-D4 or PN-D6 in ELISA and Western blotting assays. The results in Fig. 3 showed that the minimum antigenic epitopes that MAbs PN-D4 and PN-D6 could recognize were N1-10 (residues 18-133) and N2-10 (residues 252-262), respectively, which were designated NEP-D4 and NEP-D6, respectively.

Immunogenicity analysis of the epitopes.
To determine whether PEDV-infected pigs could produce specific antibodies against NEP-D4 and NEP-D6, we performed Western blotting using anti-PEDV porcine antisera obtained 14 and 21 days after immunization to detect the two GST-fusion peptides. The results showed that both lanes loaded with NEP-D4 or NEP-D6 produced positive bands at the expected molecular masses, indicating that NEP-D4 and NEP-D6 are both immunogenic epitopes of PEDV.
The epitopes are highly conserved and specific. To investigate the conservation of the determined epitopes NEP-D4 and NEP-D6 among different classical and presently circulating PEDV strains, 21 representative strains from GenBank and one strain isolated in our laboratory, HLJ/2011, were selected for sequence alignment using DNASTAR. The result showed that the sequence similarities of NEP-D4 and NEP-D6 among different PEDV strains were 98.3-100% and 81.8-100%, respectively, indicating that NEP-D4 and NEP-D6 were conserved. However, three conspicuous amino acid substitutions, including Asn 123 to Lys in NEP-D4 and Arg 252 to Lys and Ser 255 to Asn in NEP-D6, were observed in several particular strains (Fig. 4a).
To further test whether these mutations changed the antibody-antigen interactions, we used MAbs PN-D4 and PN-D6 to detect the PEDV strain HLJ/2011 by IFA. HLJ/2011 contains all of these mutations and can be stably passaged in Vero cells. The results revealed that HLJ/2011 could be recognized by both PN-D4 and PN-D6 ( Fig. 3g and h), suggesting these three substitutions do not alter the antigenicity of epitopes NEP-D4 and NEP-D6 and that both NEP-D4 and NEP-D6 are highly conserved. In addition, sequence alignment among all the known members within subfamily Coronavirinae revealed that epitope NEP-D4 and NEP-D6 are highly specific to PEDV (Fig. 4b) because the homology of these two regions between PEDV and all other members of Coronavirinae remains very low (< 44%).
Cross-reactivity analysis. To investigate whether PEDV cross-reacts with TGEV on the epitopes NEP-D4 and NEP-D6, we performed IFA against the TGEV strain SH/2012 using MAbs PN-D4 or PN-D6 as the primary antibody. No fluorescent signals were observed ( Fig. 3j and k). Moreover, in Western blotting assays, no bands appeared to indicate the interaction of TGEV hyperimmune antisera and the epitope NEP-D4 and NEP-D6 (Fig. 3e). These results suggested that the epitopes NEP-D4 and NEP-D6 could not cause cross-reactivity between PEDV and TGEV.
Intrinsically disordered regions on epitope NEP-D4. To better understand the effect of NEP-D4 in combination with PN-D4, we used multiple online tools to compute the physical and chemical parameters of NEP-D4. One of these tools, the GeneSilico MetaDisorder server, was used to predict the intrinsic disorder. Three regions presented with the highest propensity for intrinsic disorder, which we designated as PIDR [18][19][20][21][22] , PIDR 107-116 and PIDR 127-133 (Fig. 5a). The prediction suggested that there might be some intrinsically disordered regions on NEP-D4. To verify this prediction, purified GST-tagged NEP-D4 was treated with trypsin and used for detection against PN-D4 or anti-GST antibody by Western blotting. We observed that the trypsin-treated peptide The segments that could be recognized by both PN-D4 and PN-D6 are highlighted in red; the segments that could only be recognized by PN-D4 are highlighted in blue; the segments that could only be recognized by PN-D6 are highlighted in green; and the segments in gray are those that could not be recognized by either PN-D4 or PN-D6. According to (a), the N gene was divided into mutually overlapping N1 and N2 fragments and was directionally cloned into the pGEX-6p-1 vector. (b) Six hours after the addition of 1 mM IPTG, the induced products of GST-N, GST-N1 and GST-N2 were processed and analyzed using SDS-PAGE (12% separating gel and 5% stacking gel). (c) Next, the GST fusion proteins GST-N, GST-N1 and GST-N2 were coated onto ELISA plates (0.5 μ g/well) and were probed with MAbs PN-D4 and PN-D6. After the first round of identification, N1 and N2 were further divided and expressed in two series of fusion proteins, GST-N1-1-GST-N1-11 and GST-N2-1-GST-N2-11 (d,h), which were analyzed by Western blotting and ELISA using PN-D4 (e,g), PN-D6 (i,k), and anti-GST Tag antibody (f,j), respectively. was recognized by anti-GST antibody but not by MAb PN-D4 (Fig. 5b). In addition, we repeated the experiment with the freeze-thaw cycle-treated GST-tagged NEP-D4. The result showed that the treated peptide (up to 5 and 10 cycles) could be recognized by MAb PN-D4 but not by the anti-GST antibody (Fig. 5c). The intrinsically disordered proteins have a high proportion of solvent-exposed residues, making them highly accessible to protease and resistant to cold treatment 65,66 . Taken together, these results suggested that NEP-D4 is likely to be an intrinsically disordered protein.
3D structure modeling of the epitopes. To clarify the spatial locations of NEP-D4 and NEP-D6 on N protein, we generated a 3-dimensional structural model of PEDV N protein using intensive modeling based on nine templates in the Phyre2 server. The prediction result showed that 251 of 441 residues (57%) were modeled at > 90% accuracy; the remaining 191 residues were modeled ab initio. The amino acids of NEP-D6 were located in close proximity and were exposed on the surface of the N protein, suggesting that NEP-D6 is highly likely to be a linear epitope. By contrast, NEP-D4 is much more complex; although most of the amino acids are located on the surface, a small fraction of its residues are buried inside the N protein, suggesting that NEP-D4 is more likely to be a conformational epitope (Fig. 6).

Discussion
As one of the most serious intestinal infectious diseases, PED has posed a substantial threat to the swine industry of many countries over the last few years 12,16,67 . The accurate diagnosis at the early stage of infection plays a key role in PEDV prevention. The N protein of PEDV consists of 441 amino acids and represents an essential agent in the formation of the virus nucleocapsid structure. The abundant expression of N protein in PEDV-infected cells and the high conservation among different PEDV strains make it fairly suitable for use as a target for PEDV diagnosis 30,68 . However, the epitopes on PEDV N protein have been poorly characterized. In the present study, six monoclonal antibodies were generated against PEDV from BALB/c mice immunized with the purified PEDV strain FJzz1/2011. Among these antibodies, MAbs PN-D4 and PN-D6 were demonstrated as N protein-specific; the remaining four MAbs are still being studied (data not shown). Combined with ELISA and Western blotting, peptide scanning was performed to identify the epitopes of MAbs PN-D4 and PN-D6. We confirmed that Leu 18 and Ile 133 were the essential amino acids of the epitope NEP-D4 because the deletion of either of these amino acids stopped the recognition between NEP-D4 and PN-D4. Notably, the absorbance value also largely declined when the peptide was truncated at Leu 18 instead of Ser 17 . Therefore, Ser 17 might have a significant effect on the antigenicity of NEP-D4, although Leu 18 was the only essential amino acid at the N-terminus of this epitope. Likewise, Arg 252 and Lys 262 are crucial amino acids for the binding of epitope NEP-D6 with MAb PN-D6. The results suggested that NEP-D4 (residues 18-133) and NEP-D6 (residues 252-262) are the specific antigenic epitopes of MAbs PN-D4 and PN-D6, respectively.
The epitopes NEP-D4 and NEP-D6 steadily exist on different PEDV strains that are both classical and currently circulating (Fig. 4a). Additionally, the sequence similarities of the epitopes NEP-D4 and NEP-D6 with the representative strains were 98.3-100% and 81.8-100%, respectively. Although three major mutations, Asn 123 to Lys, Arg 252 to Lys, and Ser 255 to Asn, were observed in certain strains, these substitutions do not alter the antigenicity of NEP-D4 and NEP-D6 because PN-D4 and PN-D6 could precisely recognize another PEDV strain, HLJ/2011, which was isolated in our laboratory and contains all three of these mutations ( Fig. 3g and h). In addition, we demonstrated that NEP-D4 and NEP-D6 could be recognized by PEDV pig antisera collected at 14 and 21 days after inoculation with PEDV, indicating that PEDV infection could induce specific host antibodies against NEP-D4 and NEP-D6 ( Fig. 3c and d). These findings suggested that epitopes NEP-D4 and NEP-D6 were highly conserved and immunogenic; both MAbs PN-D4 and PN-D6 and epitopes NEP-D4 and NEP-D6 have the potential for use as effective tools in the serological diagnosis of PEDV during early stages of infection. Thus far, N protein and MAbs against N protein have been widely used for the diagnosis of PEDV. Hou et al. established an ELISA system using recombinant N protein generated in Escherichia coli 69 ; Okda et al. established indirect ELISA and FMIA (Fluorescent microsphere immunoassay) based on recombinant N protein, and blocking ELISA based on N protein-specific MAbs 45 .
Recently, a one-way cross-reactivity between PEDV strains and TGEV hyperimmune pig antisera was clearly observed, and one or more epitopes on the N-terminal region of N protein were hypothesized to be the root cause of this cross-reactivity 45 . Thus, the epitopes of N protein used for PEDV diagnosis should be highly PEDV-specific and the reported immunoassays above should be seriously re-examined. We analyzed the NEP-D4 and NEP-D6 epitope sequences of PEDV and other members within the subfamily Coronavirinae, especially TGEV, PRCV, and PDCoV, all of which can infect newborn piglets. We found that the amino acid sequences of NEP-D4 and NEP-D6 of PEDV differed greatly from their counterparts in other coronaviruses. NEP-D4 shared less than 43.1% homology with TGEV and PRCV of Alphacoronavirus, and only 26.5% homology with PDCoV of Deltacoronavirus. The homology of NEP-D6 between PEDV and the other three viruses (TGEV, PRCV, and PDCoV) remained at zero. We also used IFA to investigate whether NEP-D4 and NEP-D6 could induce cross-reactivity between PEDV and TGEV. The results showed that neither PN-D4 nor PN-D6 recognized TGEV, and the TGEV SH/2012 hyperimmune pig antisera failed to react with NEP-D4 and NEP-D6; these results indicated that NEP-D4 and NEP-D6 were unable to cause a cross-reaction between PEDV and TGEV. Therefore, the MAbs PN-D4 and PN-D6 and their corresponding epitopes NEP-D4 and NEP-D6 have the potential for use in developing an immunoassay that differentiates between PEDV and TGEV.
The epitope NEP-D6 has 11 amino acids and can be recognized by PN-D6 using Western blotting, a method that imposes protein denaturation. This finding indicates that NEP-D6 is a linear epitope, which agreed with the result that NEP-D6 is located on the surface of the predicted 3D structure of N protein ( Fig. 6a and b). However, the epitope NEP-D4 seems much more complicated. Western blotting suggested that NEP-D4 is a linear epitope, which contradicts the previous assumption that the full 116 aa length is too large to be a linear epitope when bound to an antibody. Again, multiple residues are buried inside N protein, and there is a significant spatial distance between the first residue (Leu 18 ) and last residue (Glu 133 ) of NEP-D4 (Fig. 6d,e and f). The NEP-D4 portion of the 3D model is highly reliable because the residues 18-133 were constructed based on five templates (PDB IDs: c3hd4A, d2bxxa1, d2geca1, d1sska and c4ud1B) at nearly 100% confidence, as shown in Fig. 6h. Therefore, NEP-D4 is more likely to be a conformational epitope than a linear epitope. Interestingly, a denatured conformational epitope is unexpectedly recognized by its cognate antibody in Western blotting. Although it is possible that the formation of some disulfide bridges between certain residues may have contributed to this binding effect, this possibility was ruled out by the ability of PN-D4 to react with dithiothreitol (DTT)-treated NEP-D4 in a Western blotting assay (Fig. 5d). Additionally, no disulfide bridges were predicted by the online software PredictProtein (data not shown). We also analyzed the other properties of NEP-D4. The software GeneSilico metadisorder predicted that NEP-D4 is an intrinsically disordered protein, which was further confirmed by two independent experiments: namely, trypsin-treated NEP-D4 could not react with PN-D4, and NEP-D4 underwent freeze-thaw The disorder plot for epitope EP4 was predicted by GeneSilico metadisorder. The x-axis shows the residues from 18 to 133, and the y-axis shows the disordered tendency ranging from 0 to 1. All residues with a disorder probability of more than 0.5 were considered to be disordered. The residues with plotted lines in the light blue region were predicted to be disordered; their counterparts in the bluish white area were considered ordered. (b) Four GST-NEP-D4 samples were treated with Trypsin under the same conditions, and the reactions were stopped at 0 min, 1 min, 5 min and 15 min. Next, the samples were processed and analyzed by Western Blotting using anti-GST antibody and PN-D4 as the primary antibodies. (c) Similarly, four additional GST-NEP-D4 samples were treated with freeze (− 80 °C) and thaw (room temperature) cycles, which were repeated up to 0, 1, 5 and 10, times, respectively. Next, these samples were analyzed by Western blotting using the same antibodies above. (d) To rule out the interference of disulfide bridges, the reduced (1 M DTT-treated) and non-reduced (non DTT-treated) GST-NEP-D4 were analyzed by Western blotting using PN-D4. cycles (up to 5 and 10 times) that were indeed recognized by PN-D4, albeit weakly. As previous findings have shown, the solvent-exposed residues on IDPs make proteins highly susceptible to protease; moreover, IDPs have much higher cold stability than structured proteins 65,66 . Thus, we speculated that folding occurred upon binding, which may have led to the recognition of NEP-D4 by PN-D4. Specifically, when the antibody binds to the first or last residue of the epitope NEP-D4, it may trigger the disordered domain to fold completely to form a certain conformational site, which could, in turn, be recognized by MAb PN-D4. In a previous study, different PEDV and TGEV pig antisera were used in a two-way cross-reactivity examination of different PEDV and TGEV strains, but only a one-way reactivity between TGEV Miller hyperimmune pig antisera and PEDV strains was observed. The complex mechanism remains unclear 47 . However, given that the N-terminal region of the N protein is intrinsically disordered, some disordered epitopes such as NEP-D4, which may be folded under certain conditions, might be the cause of the one-cross reactivity between PEDV and TGEV. In other words, the antibodies generated from PEDF infection cannot induce the folding of the disordered epitope on TGEV, whereas the corresponding epitope on PEDV can fold upon the binding of the antibodies produced from the infection of certain TGEV strains with specific epitopes. These hypotheses merit further investigation.
In conclusion, we first identified two highly conserved and specific epitopes on the N protein of PEDV. While NEP-D6 (residues 252-262) is a linear epitope, NEP-D4 (residues 18-133) is an intrinsically disordered epitope. The recognition between NEP-D4 and its cognate antibody might require folding effects upon antibody binding. Most importantly, epitopes NEP-D4 and NEP-D6 and MAbs PN-D4 and PN-D6 could be used as very effective tools in diagnosing PEDV, even at very early stages of PEDV infection. . The PEDV antisera were previously produced from pigs immunized with strain FJzz1/2011 and were preserved in our laboratory. The antisera used in this study were collected 1 day before and 14 and 21 days after immunization. In addition, the TGEV strain SH/2012, anti-TGEV N protein MAb and its hyperimmune antisera were obtained from SH/2012-immunized pigs and preserved in our laboratory.

Development of MAbs.
The PEDV strain FJzz1/2011 was purified and used as the antigen to immunize mice, and hybridomas secreting MAbs against PEDV were generated. To purify FJzz1/2011, 100 mL of virus suspension was collected and centrifuged at 5,000 × g for 30 min, 8,000 × g for 30 min, and 12,000 × g for 30 min to eliminate macromolecules. The supernatant was transferred to a new tube for ultracentrifugation at 30,000 × g for 1 h. The precipitate was resuspended in 1 mL of phosphate-buffered saline (PBS), followed by sucrose gradient centrifugation at 31,000 × g for 3 h. The purified PEDV suspension was mixed 1:1 with Freund's complete adjuvant. Next, five BALB/c mice were intraperitoneally (ip) injected with the emulsified PEDV particles at a concentration of 50 μ g per mouse. The interval between each injection was two weeks. After the third injection, spleen cells were removed and fused with SP20 cells to form hybridomas under aseptic conditions. The hybridomas were selectively cultured in HAT medium and HT medium. Finally, the positive hybridomas were screened by IFA, where the monolayers of Vero cells were inoculated with the PEDV strain FJzz1/2011 at a multiplicity of infection (MOI) of 0.01. Eighteen hours after infection, the cell lines were fixed with 80% ice-cold ethanol and were used for the selection of hybridomas that could steadily secrete MAbs against PEDV. The purified stable monoclonal hybridomas were then injected into the peritoneal cavity of mice to produce ascites fluid.

Production of recombinant PEDV N protein and N protein truncated constructs. The whole
RNA of FJzz1/2011 was extracted and converted into complementary DNA (cDNA). The entire N gene of PEDV was amplified from the cDNA by PCR using primers (Table 1) containing BamH1 and Xho1 restriction enzyme sites. The following PCR steps were used: initial denaturation at 94 °C for 3 min, followed by 32 cycles of denaturation at 94 °C for 30 seconds, annealing at 56 °C for 30 seconds and elongation at 70 °C for 30 seconds, and a final extension at 70 °C for 10 min. The obtained N gene was processed by restriction endonucleases BamH1 and Xho1 and was cloned into the prokaryotic expression vector pGEX-6p-1. The recombinant plasmid was transformed into the Escherichia coli strain BL21 (Tiangen, Beijing, China) and was expressed at 37 °C for 5 h in the presence of 1 mM isopropyl β -D-thiogalactoside (IPTG) (Sigma-Aldrich, St. Louis, MO, USA). The recombinant PEDV N protein was obtained and analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Using the same method, we obtained a series of recombinant GST fusion N protein-truncated constructs, including GST-N1, GST-N2, GST-N1-1-GST-N1-11, and GST-N2-1-GST-N2-11.
ELISA. The GST fusion N protein was coated onto ELISA plates (Thermo Fisher Scientific, Waltham, MA, USA) at 0.5 μ g/well, followed by overnight incubation at 4 °C. The ELISA plates were blocked with 5% skimmed milk in PBST (PBS plus 0.5% Tween 20) at 37 °C for 2 h. Thereafter, the plates were washed three times with PBST, and ascites fluids raised against PEDV were added as primary antibodies to each well at a dilution of 1:200 at 37 °C for 1 h. The plates were washed again with PBST three times, and HRP conjugated goat anti-mouse IgG H&L (1/10,000 dilution) (Jackson ImmunoResearch, West Grove, PA, USA) was added as a secondary antibody at 37 °C for 1 h. The plates were washed with PBST five times, and 3,3′ ,5,5′ -tetramethylbenzidine (TMB) substrate (Amresco, Cleveland, OH, USA) was added to each well for 10 min at room temperature. Finally, 2 M H 2 SO 4 was added to the plates to stop the reaction, and the absorbance value of each well at 450 nm was read. To map the epitopes of N protein, all other GST fusion proteins (GST-N1, GST-N2, GST-N1-1-GST-N1-11, and GST-N2-1-GST-N2-11) were coated onto ELISA plates and were detected using the MAbs PN-D4 or PN-D6 as the primary antibodies using the same method above.
Western blotting. Western blotting was used to further confirm the minimal antigenic regions targeted by MAbs PN-D4 and PN-D6 and to test whether the identified epitopes on N protein caused the cross-reactivity between PEDV and TGEV. The expressed N protein-truncated constructs were separated by SDS-PAGE on 12% gels and were transferred onto nitrocellulose (NC) membranes. After blocking with 5% skimmed milk for 2 h at room temperature, the membranes were exposed to primary antibodies for 1 h at room temperature. The primary antibodies used in the epitope mapping were MAbs PN-D4 and PN-D6, which were diluted 1:500 (0.5 μ g/ml). The PEDV pig antisera and TGEV pig antisera were diluted 1:2000 to be used as the primary antibodies to detect the cross-reactivity between PEDV and TGEV. After unbound primary antibodies were rinsed, the membranes were immersed into the HRP conjugated goat anti-mouse IgG H&L or goat anti-pig IgG H&L and incubated for 30 min at room temperature. Additionally, the positive GST fusion proteins on the membranes were detected by chemiluminescence.

IFA.
IFA was performed to screen for MAbs specific to the PEDV strain FJzz1/2011 and to detect whether

Bioinformatics analysis.
To further investigate the conservation and specificity of the identified epitopes, we selected another 22 PEDV representative strains ( Table 2) and 32 representative coronavirus strains (Table 3) within the subfamily Coronavirinae for multiple sequence alignment using MegAlign Clustal W method in DNASTAR Lasergene software.
Limited proteolysis and freeze-thaw cycle analyses. The recombinant protein GST-NEP-D4 was purified using Glutathione Sepharose 4B (GE Healthcare, Little Chalfont, UK), and the protein concentration was calculated by the absorbance at 280 nm (NanoDrop, Thermo Fisher Scientific, Waltham, MA, USA). Next, 10 μ g of purified GST-NEP-D4 was removed and digested by 1 μ g of trypsin (Sigma-Aldrich, St. Louis, MO, USA)

Strains Country
Collection Date in 100 mM KCl and 20 mM HEPES (pH 7.4 and 37 °C). The reaction was stopped by the addition of 1/6 volume of 6 × SDS loading buffer and heating to 100 °C for 5 min. Another 10 μ g of purified GST-NEP-D4 was frozen at − 80 °C for an hour, followed by thawing at room temperature. Finally, the samples treated with trypsin and freeze-thaw cycles were analyzed by Western blotting with MAb PN-D4 or an Anti-GST antibody, respectively, using the same methods described above.