key: cord-0955941-xge6mah1
authors: Tsai, Tsung-Lin; Su, Chen-Chang; Hsieh, Ching-Chi; Lin, Chao-Nan; Chang, Hui-Wen; Lo, Chen-Yu; Lin, Ching-Houng; Wu, Hung-Yi
title: Gene Variations in Cis-Acting Elements between the Taiwan and Prototype Strains of Porcine Epidemic Diarrhea Virus Alter Viral Gene Expression
date: 2018-11-29
journal: Genes (Basel)
DOI: 10.3390/genes9120591
sha: 023b805a96c852619186f55af4bbbdd5f9fa8156
doc_id: 955941
cord_uid: xge6mah1

In 2013, the outbreak of porcine epidemic diarrhea (PED) in Taiwan caused serious economic losses. In this study, we examined whether the variations of the cis-acting elements between the porcine epidemic diarrhea virus (PEDV) Taiwan (TW) strain and the prototype strain CV777 alter gene expression. For this aim, we analyzed the variations of the cis-acting elements in the 5′ and 3′ untranslated regions (UTRs) between the PEDV TW, CV777, and other reference strains. We also determined the previously unidentified transcription regulatory sequence (TRS), a sequence motif required for coronavirus transcription, and found that a nucleotide deletion in the TW strain, in comparison with CV777 strain, immediately downstream of the leader core sequence alters the identity between the leader TRS and the body TRS. Functional analyses using coronavirus defective interfering (DI) RNA revealed that such variations in cis-acting elements for the TW strain compared with the CV777 strain have an influence on the efficiency of gene expression. The current data show for the first time the evolution of PEDV in terms of cis-acting elements and their effects on gene expression, and thus may contribute to our understanding of recent PED outbreaks worldwide.

Porcine epidemic diarrhea virus (PEDV) belongs to the genus Alphacoronavirus of subfamily Coronavirinae, which is in the family Coronaviridae, order Nidovirales [1,2]. The capped PEDV genome of ≈28 kilobases consists of a 5 untranslated region (UTR), open reading frames (ORFs), and a 3 UTR that includes the poly(A) tail [3] . RNA elements in the 5 and 3 untranslated regions (UTRs) that are important for coronaviral gene expression are collectively referred to as cis-acting RNA elements [4] . For the coronavirus, the transcription regulatory sequence (TRS) is a cis-acting RNA element required for subgenomic mRNA (sgmRNA) synthesis that is located downstream of the leader sequence and

Random hexamer primers were used for reverse transcription (RT) with SuperScript III reverse transcriptase (Invitrogen, Carlsbad, CA, USA), and the resulting cDNA was used for PCR with PfuUltra II high-fidelity DNA polymerase (Agilent, Santa Clara, CA, USA) and oligonucleotides PEDV 25(−) and PEDV 450(+) for the 5 UTR and PEDV 27488 (−) and PEDV 3 end (+) for the 3 UTR (Table S1 ). The resulting 50-µL PCR mixture was heated to 94 • C for 2 min and subjected to 34 cycles of 30 s at 94 • C, 30 s at 55 • C, and 30 s at 72 • C. The extreme 5 and 3 termini of the PEDV genome were identified via rapid amplification of cDNA ends (RACE) (Thermo Fisher Scientific) according to the manufacturer's instructions. To amplify each of the sgmRNA for identification of the core sequence, oligonucleotide PEDV 25(−) and a corresponding reverse primer specific for each sgmRNA (Table S1) were used, followed by sequencing analyses. To identify the sgmRNA derived from PEDV 3 UTR, oligonucleotides PEDV 25(−) and PEDV 3 utr (+) were used. The resulting 50-µL PCR mixture was heated to 94 • C for 2 min and subjected to 30 cycles of 30 s at 94 • C, 30 s at 55 • C, and 90 s at 72 • C.

To construct the PEDV TW defective interfering (DI) RNA pTWDI, an overlap PCR mutagenesis procedure was performed, as previously described [32] , but with oligonucleotides T7-PEDV5 UTR(−) and PEDV26696(+), and PEDV TW strain cDNA in the first PCR; oligonucleotides PEDV 26541 (−) and PEDV3 end (+), and PEDV TW strain cDNA in the second PCR; and oligonucleotides T7-PEDV 5 UTR(−) and PEDV 3 end (+), and the products of the first two reactions in a third PCR. The resulting PCR product was then cloned into the TOPO XL vector (Thermo Fisher Scientific). To construct CV777 DI RNA, in which both the 5 and 3 UTR were from PEDV CV777 but the remaining sequences were from PEDV TW, pTWDI was used as a template; the aforementioned overlap PCR mutagenesis procedure was also performed with mutation oligonucleotides listed in Table S1 . Mutants pTWDI-EP, pCV777DI-EP, p∆ AS-TRS, and pS-TRS were similarly constructed, except for the corresponding oligonucleotides used in the first and second reactions, as described in Table S1 .

The PEDV TW strain (GenBank Accession No. KP276252) was obtained from Dr. Hui-Wen Chang (National Taiwan University) [33] and used as the helper virus for the DI RNA replication assay. The PEDV TW strain was maintained in Vero cells, as previously described [34, 35] ; the viral titer was 10 5 TCID50/mL. Vero cells in 35-mm dishes were infected with 200 µL of 10 5 TCID50/mL of PEDVPT-P5. The DNA constructs pTWDI, pCV777DI, p∆A S-TRS, and pS-TRS were linearized with MluI, transcribed in vitro to synthesize RNA transcripts with the mMessage mMachine T7 transcription kit (Thermo Fisher Scientific) according to the manufacturer's instructions and passed through a Biospin 6 column (Bio-Rad, Hercules, CA, USA). After 2 h of infection with the helper virus, 3 µg of RNA transcript was transfected into the cells. To detect replication of PEDV DI RNA and DI RNA-derived sgmRNAs (sgm DI RNA), the supernatant was collected at 48 hour posttransfection (hpt) and then used to infect fresh Vero cells (virus passage 1, VP1). Ten micrograms of TRIzol-extracted total cellular RNA at 48 hpi of VP1 was used for the RT reaction with the reporter oligonucleotide TGEV(+) and SuperScript III reverse transcriptase (Thermo Fisher Scientific). To examine synthesis of DI RNA, SYBR ® green amplification mix (Roche Applied Science, Mannheim, Germany) and primers PEDV476(−) and TGEV(+) were used for qPCR according to the manufacturer's protocol. To analyze the synthesis of sgm DI RNA, primers PEDVL20(−), which binds to the leader sequence, and 5 GD(+), which binds to the reporter gene, were employed to differentiate the DI RNA from the helper virus PEDV. In these experiments, dilutions of plasmids containing the same gene as the detected DI RNA or sgm DI RNA were always run in parallel with the quantitated cDNA for use in standard curves (dilutions ranged from 10 8 to 10 copies of each plasmid). The amount of synthesized RNA was normalized to the levels of internal controls, including helper virus genomic RNA, 18S rRNA, and DI RNA. The reactions were performed with an initial pre-incubation at 95 • C for 5 min, followed by 35 amplification cycles of 95 • C for 15 s and 60 • C for 60 s.

The DNA constructs pTWDI-EP and pCV777DI-EP were linearized with MluI, transcribed in vitro with the mMessage mMachine T7 transcription kit (Thermo Fisher Scientific) according to the manufacturer's instructions, and passed through a Biospin 6 column (Bio-Rad). Vero cells were then transfected with 3 µg of transcript RNA, and proteins in cell lysates were collected at 16 h posttransfection. The protein samples were separated using 12% sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and electrotransferred onto nitrocellulose membranes (GE Healthcare, Chicago, IL, USA). An antibody against enhanced green fluorescent protein (EGFP) was used as the primary antibody; goat anti-mouse IgG conjugated to horseradish peroxidase (HRPO) was the secondary antibody (Jackson Laboratory, Bar Harbor, ME, USA). Detected EGFP was visualized using Western Lightning™ Chemiluminescence Reagent (Perkin Elmer, Waltham, MA, USA) and X-ray film (Kodak, Rochester, NY, USA).

Statistical analysis of the data was performed with Student's unpaired t-test using Prism 6.0 software (GraphPad Software, La Jolla, CA, USA). Values are presented as the mean ± standard deviation (SD) (n = 3); * p < 0.05, ** p < 0.01, and *** p < 0.001. The MegAlign program (DNASTAR, Madison, WI, USA) was used for the alignment of nucleotide and deduced amino acid sequences.

Madhugiria et al. showed that the ≈310 nts of the 5 terminal genome region of Alphacoronavirus form four secondary structures: SL1, SL2, SL4, and SL5 [4] . Accordingly, the sequence representing the 300 nts of the 5 terminal region obtained for PEDV-positive samples (TW strain) were analyzed and compared with that of the PEDV prototype CV777. To understand whether the sequences for the 5 terminal genome region of other PEDV strains are also altered, the reference strains were also selected from GenBank based on the viruses that have been circulating in different countries in recent years (2012-2015) and compared. As shown in Figure 1 , in comparison with CV777, a U-insertion at nt 48; an A deletion at nt 73, which was located immediately downstream of the TRS core sequence (CUAAAC); and a 4-nt deletion (UUCC) from nt 93 to 96 were identified. In addition, the bases at positions 8, 104, 122, and 292 were also altered to a G, C, A, and U, respectively. Of the PEDV reference strains selected from other countries, the 5 -terminal sequence for the TW strain shared 100% nucleotide identity with that of the USA, Germany, and Korea reference strains. The change of the C and A at positions 104 and 122, respectively, also altered the amino acid encoded by the uORF (Figure 1B ,C). Using the Mfold algorithm [36] , four SLs were identified in the 5 terminal genome region of PEDV and the results were in agreement with those found by Madhugiria et al. in Alphacoronaviruses [4] . In comparison with CV777, the change at nt 8 for the TW strain slightly altered the structure of SL1 ( Figure 1C ), and the U-insertion at nt 48 slightly altered the loop but not the stem structure of SL2. The free energy (∆G) of SL1 and SL2 was found to increase from −11.1 and −1.8 kcal/mol for the CV777 strain to −12.5 and −3.9 kcal/mol for the TW strain, respectively. The 4-nt deletion plus 2 alterations in SL4 also slightly affected the overall structure of SL4, and a decrease in ∆G from −27.4 for CV777 strain to −24.3 kcal/mol was calculated for TW strain. Although the U-alteration at nt 292 did not impair the structure of SL5, ∆G decreased from −53.2 to −51.6 kcal/mol. The overall ∆G of the four structures for CV777 and Taiwan strain was found to be −97.3 and −96.2 kcal/mol, respectively. 

Using a sequencing analysis of 3′ UTR, either U or C at nt −49 was found from PEDV-positive samples (TW strain). The PEDV TW strain was therefore divided into two groups: Taiwan group A 

Using a sequencing analysis of 3 UTR, either U or C at nt −49 was found from PEDV-positive samples (TW strain). The PEDV TW strain was therefore divided into two groups: Taiwan group Genes 2018, 9, 591 6 of 16 A (with U at nt −49) and Taiwan group B (with C at nt −49) (Figure 2A ). Regarding 3 UTR, 7-and 8-nt differences between CV777 and TW groups A and B, respectively, were identified ( Figure 2A ). TW strain group A shared 100% nt identity with the USA and Korea strains. The C at nt −49 in the TW group B is unique because it was not identified in any sequenced PEDV genome, as based on data from GenBank; thus, it can be used as a genetic marker to differentiate the TW strain from other PEDV strains. Using the Mfold algorithm [36] , the BSL, PK, and HVR structures were predicted ( Figure 2B ). Of the nts that are different from those of CV777, one nt is located in the BSL structure, and three nts (TW strain group A) and four nts (TW strain group B) are located in the HVR structure. However, these changes do not alter the overall structure of BSL and HVR; the ∆G for both structures was found to be slightly increased in comparison with that for CV777. (with U at nt −49) and Taiwan group B (with C at nt −49) (Figure 2A ). Regarding 3′ UTR, 7-and 8-nt differences between CV777 and TW groups A and B, respectively, were identified ( Figure 2A ). TW strain group A shared 100% nt identity with the USA and Korea strains. The C at nt −49 in the TW group B is unique because it was not identified in any sequenced PEDV genome, as based on data from GenBank; thus, it can be used as a genetic marker to differentiate the TW strain from other PEDV strains. Using the Mfold algorithm [36] , the BSL, PK, and HVR structures were predicted ( Figure 2B ). Of the nts that are different from those of CV777, one nt is located in the BSL structure, and three nts (TW strain group A) and four nts (TW strain group B) are located in the HVR structure. However, these changes do not alter the overall structure of BSL and HVR; the ∆G for both structures was found to be slightly increased in comparison with that for CV777. 

While analyzing the 3 UTR sequences of TW strain, with the primers annealing the 5 leader sequence and 3 UTR, the reverse transcription polymerase chain reaction (RT-PCR) products with a length of ≈1700 base pairs (bp) and ≈250 bp were observed ( Figure 3A , lane 2). After sequencing analysis, the ≈1700 bp RT-PCR product was identified to be N sgmRNA and the ≈250 bp RT-PCR product, which consisted of a leader sequence and a part of the sequence from 3 UTR, was a previously unidentified sgmRNA species. The sgmRNA was derived from the CS-like sequence UUAAAC, which encompassed the stop codon of the N gene ( Figure 3B , upper panel). The potential start codon AUG of the sgmRNA, which was located one nt downstream of the CS-like sequence, was also identified ( Figure 3B , upper panel), and the ORF was predicted to encode a protein of 35 amino acids ( Figure 3B , lower panel). In TGEV, sgmRNA 7 was also derived from a CS at a genome position similar to that in PEDV ( Figure 3C ), and the encoded protein of 78 amino acids has been demonstrated to be involved in the virulence [37] . Such a CS sequence at a position similar to that PEDV was also found in coronavirus 229E and Scotophilus bat coronavirus 512 ( Figure 3C ), though the sgmRNAs derived from both putative CSs have not been experimentally identified. In comparison with CV777, three nts and three amino acids were altered within the ORF ( Figure 3D ,E, respectively). The region downstream of the gene encoding the N protein shows variations among Alphacoronaviruses, which may explain why the conserved counterpart of the Betacoronavirus BSL structure [9, 10] could not be identified as a common structure in Alphacoronaviruses [4] . For PEDV, the BSL structure was identified using the Mfold algorithm [36] , as shown in Figure 2B , yet it was located within the putative ORF, as described above ( Figure 3D ). The function of the protein encoded by the ORF of the sgmRNA and the importance of the predicted BSL structure to PEDV replication remain to be elucidated. 

Coronavirus DI RNA has been intensively employed for gene expression analyses [38] [39] [40] [41] [42] .

To determine whether the observed variations in the 5 and 3 UTRs between TW and CV777 strains alter the efficiency of gene expression, a PEDV DI RNA for the TW strain was constructed; it was designated TW DI ( Figure 4A ). Because the CV777 strain was not available due to regulations in Taiwan, a CV777 DI RNA (designated CV777 DI, Figure 4B ) was constructed via mutagenesis, which had TW DI as a backbone but with the CV777 5 and 3 UTRs. For translation analysis, both DI RNAs were engineered to contain a reporter EGFP gene ( Figure 4A ) designated CV777 DI-EP and TW DI-EP. Vero cells were transfected with 3 µg of in vitro-transcribed RNA, and proteins in cell lysates were collected at 16 h posttransfection. The amount of synthesized EGFP was normalized to the levels of internal controls including β-actin and DI RNA. Note that both the DI RNAs cannot replicate because of the lack of helper virus in transfected cells and the stability of the two DI RNAs was similar at the time of RNA collection (data not shown). As shown in Figure 4C ,D, the translation efficiency of TW DI-EP was ≈2.5-fold better than that of CV777 DI-EP, suggesting that the sequence alterations in the 5 and 3 UTRs of TW DI-EP increased the translation efficiency in comparison with that of strain CV777. Note that, because uORF has been suggested to have an influence on translation, in addition to alterations in sequence and structure at the 5 and 3 UTRs, the amino acid changes for uORF (Figure 1 ) may be a factor leading to the observed difference in translation efficiency. 

Coronavirus DI RNA has been intensively employed for gene expression analyses [38] [39] [40] [41] [42] . To determine whether the observed variations in the 5′ and 3′ UTRs between TW and CV777 strains alter the efficiency of gene expression, a PEDV DI RNA for the TW strain was constructed; it was designated TW DI ( Figure 4A ). Because the CV777 strain was not available due to regulations in Taiwan, a CV777 DI RNA (designated CV777 DI, Figure 4B ) was constructed via mutagenesis, which had TW DI as a backbone but with the CV777 5′ and 3′ UTRs. For translation analysis, both DI RNAs were engineered to contain a reporter EGFP gene ( Figure 4A ) designated CV777 DI-EP and TW DI-EP. Vero cells were transfected with 3 µg of in vitro-transcribed RNA, and proteins in cell lysates were collected at 16 h posttransfection. The amount of synthesized EGFP was normalized to the levels of internal controls including β-actin and DI RNA. Note that both the DI RNAs cannot replicate because of the lack of helper virus in transfected cells and the stability of the two DI RNAs was similar at the time of RNA collection (data not shown). As shown in Figure 4C ,D, the translation efficiency of TW DI-EP was ≈2.5-fold better than that of CV777 DI-EP, suggesting that the sequence alterations in the 5′ and 3′ UTRs of TW DI-EP increased the translation efficiency in comparison with that of strain CV777. Note that, because uORF has been suggested to have an influence on translation, in addition to alterations in sequence and structure at the 5′ and 3′ UTRs, the amino acid changes for uORF (Figure 1 ) may be a factor leading to the observed difference in translation efficiency. Regarding the effect on replication, the replication efficiency for CV777 DI was extremely low when compared with that for TW DI (≈20-fold difference) ( Figure 4E ) after 48 h of VP1. Sequencing analysis revealed that the leader sequence in CV777 DI was, in part, replaced by that of the helper virus TW strain, as a G at nt 8 and a U at nt 48 were identified. Except for the alterations in the leader Regarding the effect on replication, the replication efficiency for CV777 DI was extremely low when compared with that for TW DI (≈20-fold difference) ( Figure 4E ) after 48 h of VP1. Sequencing analysis revealed that the leader sequence in CV777 DI was, in part, replaced by that of the helper virus TW strain, as a G at nt 8 and a U at nt 48 were identified. Except for the alterations in the leader sequence, no other changes in CV777 DI were found. It has been demonstrated that leader-switching occurred with a high frequency in the TRS region between DI RNA and the helper virus [43, 44] , and this may explain the replacement of the nts that occurred in the CV777 DI RNA leader sequence. Because it has been suggested that DI RNA is able to compete and recombine with the helper virus, such a nature may complicate the interpretation of the DI replication assay [45] . In the current study, it was found that the leader sequence of CV777 DI RNA was replaced by that of the helper virus, and consequently the recombination event may affect the quantitative results of the DI RNA replication assay. Whether the replication efficiency between wild-type CV777 and TW strain with full-length genome still shows a dramatic difference remains to be determined. Accordingly, at this point, we can only conclude that the variations in 5 and 3 UTRs between CV777 and TW may influence replication efficiency using the DI RNA system. Further study using PEDV reverse genetic system is required to determine to what extent the variations in 5 and 3 UTRs between the two strains affect replication efficiency.

TRSs are cis-acting elements required for coronavirus sgmRNA synthesis that are located at the 3 end of the leader sequence (TRS-L) and preceded by each gene (TRS-B) . The TRS contains a CS flanked by variable sequences at its 5 (5 TRS) and 3 (3 TRS) ends of the CS. The CS of TRS-L (CS-L) in PEDV CV777 is CUAAAC. The CS of TRS-B (CS-B) for CV777 strain E, M, and N genes has been experimentally determined to be CUAGAC, AUAAAC, and CUAAAC, respectively [46] (Figure 5A , left panel for E, M, and N genes). Although the CS-B for the S and ORF3 genes for CV777 strain was previously assumed to be GUAAAC and CUAGAC, respectively [47] (Figure 5B ,C, respectively, upper left panel), this has yet to be experimentally determined. In comparison with CV777, the CS-L of the TW strain was not altered, i.e., it was CUAAAC ( Figure 1A) . To determine the CS-B for TW, sgmRNA was amplified using RT-PCR followed by sequencing. The sequencing results were analyzed to identify the fusion sites in the genome with the 5 end of the virus, by which the CS-B for the S, ORF3, E, M, and N genes was determined ( Figure 5A, right panel) . The results suggested that the CS-B for E, M, and N was the same for PEDV TW and CV777; however, the CS-B for S and ORF3 for TW was different from the previously assumed CS-Bs for CV777. Instead of the previously assumed GUAAAC for CV777 ( Figure 5B , upper left panel), the CS-B for the TW S gene was CGUAAA ( Figure 5B , upper right panel), which was located one nt upstream of previously assumed GUAAAC. Based on the study by Sola et al. that found that TRS is defined as consisting of the central CS and the four nts immediately 5 (5 TRS) and 3 (3 TRS) of the CS, the overall identity of the S TRS-B with TRS-L for CV777 was slightly lower than that for TW (8 for CV777 and 10 for TW; Figure 5B , upper panel). We speculate the reason why GUAAAC was previously assumed to be the CS-B for the CV777 S gene may be attributed to its higher sequence identity with CS-L in comparison with CGUAAA for TW (5 for CV777 and 3 for TW; Figure 5B , upper panel). However, the higher identity between the 5 TRS-B and 5 TRS-L (1 for CV777 and 4 for TW) and between the 3 TRS-B and 3 TRS-L (2 for CV777 and 3 for TW) may be a more decisive factor for the outcome based on CGUAAA being selected for TW. With regard to ORF3, CS-B CCUUAC for TW ( Figure 5C , upper right panel), which was located 11 nts downstream of the previously assumed CUAGAC for CV777 ( Figure 5C, upper left  panel) , was identified. The same reason as described above for the S gene CS-B may be involved, whereby the previously assumed CUAGAC of ORF3 for CV777 showed higher identity with the CS-L in comparison with that for TW while lower identity between the 5 TRS-B and 5 TRS-L (0 for CV777 and 4 for TW) and between the 3 TRS-B and 3 TRS-L (1 for CV777 and 3 for TW) is observed. Based on the argument above, it is currently assumed that the CS of TRS-B (CS-B) for CV777 strain S and ORF3 genes is CGUAAA ( Figure 5A , left panel and Figure 5B , lower left panel) and CCUUAC ( Figure 5A , left panel and Figure 5C , lower left panel), respectively; however, the exact CS-B of S and ORF3 genes for CV777 remains to be experimentally determined with RNA collected from CV777-infected cells. immediately surrounding the CS-B is also shown. Nucleotide identity between sequences is indicated using an asterisk. The number of nt identity among the 4-nt 3′-TRS is indicated above the 3′-TRS. Postulated polymerase strand-switching during minus-strand synthesis is indicated by an arrow. Because the CS-B for the S and ORF3 genes for CV777 are unknown, the currently assumed CS-B for the two genes is shown in the left panel of (A). The previously (upper left panel) and currently (lower left panel) assumed CS-B for the S gene and ORF3 genes for CV777 is also used for the alignment, as shown in the left panel of (B) and (C), respectively.

According to the current study shown in Figure 1 and previous reports [3, 18, 19] , in comparison with CV777, an A deletion at nt 73 immediately downstream of CS-L was frequently identified for PEDV, including the TW strain. It has been proposed that such a deletion immediately downstream of CS-L may have influence on PEDV sgmRNA synthesis [3, 18, 19] , though further examination to support this hypothesis has been not performed. As the TRS is defined to consist of the central CS and the four nts immediately flanking the 5′ (5′ TRS) and 3′ (3′ TRS) CS [17] and based on the alignment shown in Figure 5A , the one-nt deletion immediately downstream of the CS-L in the TW 

According to the current study shown in Figure 1 and previous reports [3, 18, 19] , in comparison with CV777, an A deletion at nt 73 immediately downstream of CS-L was frequently identified for PEDV, including the TW strain. It has been proposed that such a deletion immediately downstream of CS-L may have influence on PEDV sgmRNA synthesis [3, 18, 19] , though further examination to support this hypothesis has been not performed. As the TRS is defined to consist of the central CS and the four nts immediately flanking the 5 (5 TRS) and 3 (3 TRS) CS [17] and based on the alignment shown in Figure 5A , the one-nt deletion immediately downstream of the CS-L in the TW strain ( Figure 5A , right panel) results in increased identity for the 3 TRS between TRS-B and TRS-L for S, ORF3, E, and M, but not N genes, in comparison with CV777 ( Figure 5A, left panel) . This was also true for comparison of the previously ( Figure 4B ,C, upper left panel) and currently ( Figure 4B ,C, lower left panel) assumed CS-B of S and ORF3 gene for CV777. Because the 3 TRS has a more decisive influence on template-switching during sgmRNA synthesis [16, 17] , the increased identity of the 3 TRS in these genes may enhance sgmRNA synthesis. Furthermore, it has been suggested that if the identity of the 5 TRS and 3 TRS of CS-B with those of CS-L reaches a plateau in G value, template-switching is able to occur for sgmRNA synthesis, even though CS-B exhibits low identity with CS-L [17] . Thus, in comparison with the TW strain, the decreased identity in the CV777 3 TRS may have an influence on the synthesis of S, ORF3, and M sgmRNAs because their CS-Bs also have low identity with CS-L. Accordingly, in comparison with CV777, the increased sequence identity of the 3 TRS between TRS-L and TRS-B of S, ORF3, E, and M (but not N) genes in TW may alter the relative efficiency of sgmRNA synthesis.

To determine whether the altered sequence identity between the leader TRS and body TRS ( Figure 5 ) affected sgmRNA synthesis within the same context of the genome, the TRS-B for the S gene was inserted into TW DI (designated S-TRS; Figure 6A , right panel). In addition, ∆A S-TRS ( Figure 6A , left panel) was also constructed, in which the 3 TRS of the TRS-B for the S gene was mutated to decrease its sequence identity with the 3 TRS of the TRS-L ( Figure 6B ). As shown in Figure 6C , the efficiency of subgenomic DI RNA synthesized from ∆A S-TRS was decreased in comparison with that from S-TRS. The results suggest that the increase in sequence identity between the leader TRS and body TRS leads to enhanced subgenomic mRNA synthesis, which is in agreement with our hypothesis. Indeed, these results were not unexpected based on previously published studies regarding the correlation of TRS sequence identity with sgmRNA synthesis [16, 17, [48] [49] [50] . However, it is the first study on the content of a PEDV genome demonstrating that the increased sequence identity between the leader TRS and body TRS enhances sgmRNA synthesis. The results also support the previously proposed hypothesis [3, 18, 19 ] that the increased 3 TRS sequence identity between TRS-L and TRS-B in the TW strain caused by a deletion immediately downstream of CS-L has influence on PEDV sgmRNA synthesis. In conclusion, the results support our hypothesis that increased 3 TRS sequence identity between TRS-L and TRS-B in the TW strain caused by a deletion immediately downstream of CS-L has an influence on PEDV sgmRNA synthesis.

Overall, we have determined that the cis-acting elements were altered in the PEDV TW strain and the reference strains (except the Thailand strain) in comparison with those of the PEDV prototype strain CV777. Functional analyses suggest that such variations in cis-acting elements for the TW strain compared with the CV777 strain have an influence on the efficiency of gene expression. Therefore, it is speculated that variations in cis-acting elements may also be one of the factors potentially contributing to the pathogenesis. Although the cis-acting elements of the Thailand reference strain are the same as those of the CV777 strain, we still cannot exclude the importance of the cis-acting elements because it is not clear whether the Thailand reference strain is virulent or not based on the insufficient information from GenBank. Accordingly, we also cannot rule out other factors that also play roles in the PED outbreak. Further study is required to determine to what extent the variations in cis-acting elements affect the virulence and pathogenesis using a reverse genetic system. identity between the leader TRS and body TRS enhances sgmRNA synthesis. The results also support the previously proposed hypothesis [3, 18, 19] that the increased 3′ TRS sequence identity between TRS-L and TRS-B in the TW strain caused by a deletion immediately downstream of CS-L has influence on PEDV sgmRNA synthesis. In conclusion, the results support our hypothesis that increased 3′ TRS sequence identity between TRS-L and TRS-B in the TW strain caused by a deletion immediately downstream of CS-L has an influence on PEDV sgmRNA synthesis. 

In this study, we found variations in cis-acting elements in the 5 and 3 UTRs between the PEDV TW and prototype CV777 strains. We also determined the previously unidentified cis-acting element TRS and found that (i) a nucleotide deletion in the TW strain, in comparison with CV777 strain, immediately downstream of the leader core sequence alters the identity between the leader TRS and the body TRS, and (ii) the altered sequence identity has influence on coronavirus subgenomic mRNA synthesis. Functional analyses revealed that such variations in the 5 and 3 UTRs of the TW strain also alter the efficiency of gene expression in comparison with that of the CV777 strain. Because the analyzed cis-acting elements for the TW strain are the same as those of the reference strains for USA and Korea, our results may extend our understanding of PEDV evolution in terms of cis-acting elements and the recent outbreaks of PED worldwide. Further study using reverse genetic system is required to determine whether the altered gene expression is associated with the increased severity of the recent PED outbreaks. In addition, a putative ORF within the 3 UTR and a BSL structure within the putative ORF were also found. The biological significance of the ORF and BSL structure remains to be experimentally demonstrated.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4425/9/12/591/s1, Table S1 : Oligonucleotides used for this study. 

Virus Taxonomy: Classification and Nomenclature of Viruses: Ninth Report of the International Committee on Taxonomy of Viruses

Coronavirus genome structure and replication

Origin, evolution, and genotyping of emergent porcine epidemic diarrhea virus strains in the United States

Coronavirus cis-acting RNA elements

Continuous and discontinuous RNA synthesis in coronaviruses

Reselection of a genomic upstream open reading frame in mouse hepatitis coronavirus 5 -untranslated-region mutants

RNA structure analysis of Alphacoronavirus terminal genome regions

The structure and functions of coronavirus genomic 3 and 5 ends

Characterization of an essential RNA secondary structure in the 3 untranslated region of the murine coronavirus genome

A bulged stem-loop structure in the 3 untranslated region of the genome of the coronavirus mouse hepatitis virus is essential for replication

A phylogenetically conserved hairpin-type 3 untranslated region pseudoknot functions in coronavirus RNA replication

A hypervariable region within the 3 cis-acting element of the murine coronavirus genome is nonessential for RNA synthesis but affects pathogenesis

RNA-RNA and RNA-protein interactions in coronavirus replication and transcription

A new model for coronavirus transcription

A contemporary view of coronavirus transcription

Transcription regulatory sequences and mRNA expression levels in the coronavirus transmissible gastroenteritis virus

Role of nucleotides immediately flanking, the transcription-regulating sequence core in coronavirus subgenomic mRNA synthesis

Full-length genome sequences of porcine epidemic diarrhoea virus strain CV777; Use of NGS to analyse genomic and sub-genomic RNAs

Molecular characterization and phylogenetic analysis of porcine epidemic diarrhea viruses associated with outbreaks of severe diarrhea in piglets in Jiangxi

Experimental infection of pigs with a new porcine enteric coronavirus, CV 777

A new coronavirus-like particle associated with diarrhea in swine

Diseases of Swine

Porcine epidemic diarrhea virus: An emerging and re-emerging epizootic swine virus

US-like strain of porcine epidemic diarrhea virus outbreaks in Taiwan

Porcine aminopeptidase N is a functional receptor for the PEDV coronavirus

Contribution of the porcine aminopeptidase N (CD13) receptor density to porcine epidemic diarrhea virus infection

Identification of the epitope region capable of inducing neutralizing antibodies against the porcine epidemic diarrhea virus

High-level expression of the neutralizing epitope of porcine epidemic diarrhea virus by a tobacco mosaic virus-based vector

Expression of a synthetic neutralizing epitope of porcine epidemic diarrhea virus fused with synthetic B subunit of Escherichia coli heat labile enterotoxin in rice endosperm

Phylogenetic analysis of the spike (S) gene of the new variants of porcine epidemic diarrhoea virus in Taiwan

Isolation and characterization of porcine epidemic diarrhea viruses associated with the 2013 disease outbreak among swine in the United States

The 3 -terminal 55 nucleotides of bovine coronavirus defective interfering RNA harbor cis-acting elements required for both negative-and positive-strand RNA synthesis

Evaluation and comparison of the pathogenicity and host immune responses induced by a G2b Taiwan porcine epidemic diarrhea virus (Strain Pintung 52) and its highly cell-culture passaged strain in conventional 5-week-old pigs

Propagation of the virus of porcine epidemic diarrhea in cell culture

Derivation of attenuated porcine epidemic diarrhea virus (PEDV) as vaccine candidate

Prediction of RNA secondary structure by energy minimization

Coronavirus gene 7 counteracts host defenses and modulates virus virulence

An RNA stem-loop within the bovine coronavirus nsp1 coding region is a cis-acting element in defective interfering RNA replication

A cis-acting function for the coronavirus leader in defective interfering RNA replication

Stem-loop IV in the 5 untranslated region is a cis-acting element in bovine coronavirus defective interfering RNA replication

Host protein interactions with the 3 end of bovine coronavirus RNA and the requirement of the poly(A) tail for coronavirus defective genome replication

5 -proximal hot spot for an inducible positive-to-negative-strand template switch by coronavirus RNA-dependent RNA polymerase

The UCUAAAC promoter motif is not required for high-frequency leader recombination in bovine coronavirus defective interfering RNA

Common RNA replication signals exist among group 2 coronaviruses: Evidence for in vivo recombination between animal and human coronavius molecules

Effect of mutations in the mouse hepatitis virus 3'(+)42 protein binding element on RNA replication

PEDV leader sequence and junction sites

Manipulation of the porcine epidemic diarrhea virus genome using targeted RNA recombination

Downstream sequences influence the choice between a naturally occurring noncanonical and closely positioned upstream canonical heptameric fusion motif during bovine coronavirus subgenomic mRNA synthesis

Genetic manipulation of arterivirus alternative mRNA leader-body junction sites reveals tight regulation of structural protein expression

Regulation of relative abundance of arterivirus subgenomic mRNAs

We thank Jui-Hung Shien (National Chung Hsing University) for many helpful discussions. We also thank the Animal Disease Diagnostic Center of National Chung Hsing University and National Pingtung University of Science and Technology for assisting sample collection.

The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the result.