Molecular characterization and phylogenetic analysis of porcine epidemic diarrhea virus isolates in Eastern China

Introduction: Porcine epidemic diarrhea virus (PEDV) is one of the most common viral pathogens causing swine diarrhea. Methodology: We performed a genetic evolution analysis of the S1 gene of endemic PEDV strains in Eastern China. The S1 genes of 37 PEDVpositive samples were amplified and sequenced, and compared to the standard CV777 strain, 120 nucleotides were found to have mutations. Results: The nucleotide and deduced amino acid homologies between the sequences and those of the CV777 strain were 90%–91% and 88.2%– 90%, respectively, and their homologies to the vaccine strain were 88.6%–89.7% and 86.2%–87.8%, respectively. Genetic evolution and variation analyses indicated that the 37 PEDV strains belonged to genogroup 2-1, while the CV777 strain, vaccine strain, and earlier Chinese strains all belonged to genogroup 1-1. Conclusions: The newly emerged clinical PEDV strains indicate that the PEDV CV777 vaccine currently used in China may not fully protect pigs from infection with recent epidemic strains, and will require the development of new vaccine strains.


Introduction
Diarrhea is one of the most significant diseases currently threatening the pig industry in China.The causes of diarrhea are complicated, and several pathogens can be involved.The common viral pathogens are porcine epidemic diarrhea virus (PEDV), porcine transmissible gastroenteritis virus (TGEV), and porcine rotavirus serotype A (GARV), all of which have similar clinical manifestations.PEDV has the highest infection and incidence rates among these pathogens, and causes acute enteritis and watery diarrhea in pigs.PEDV was first reported in 1978 in Belgium and the UK [1,2].It was isolated from infected pigs for the first time in 1980 in China, and has since become quite common.Since October 2010, many provinces in China have witnessed outbreaks of diarrhea due to PEDV among pig herds [3], and PEDV has caused substantial economic losses.The pig industry in Eastern China has been greatly impacted by porcine diarrhea.PEDV is an RNA virus.The spike (S) protein of PEDV is a glycosylated protein that is involved in viral pathogenesis, which contains distinct S1 and S2 domains.The S protein attaches to host cellular receptors, resulting in viral entry by membrane fusion [4][5][6][7] and stimulating the production of neutralizing antibodies [4][5][6][7][8].Mutations in the PEDV S gene are associated with growth adaptation in vitro and the attenuation of virulence in vivo [9,10].Mutations play an important role in the structure of the S gene, and in determining isolate diversity, the actual epidemiological characteristics of the infection, and the incidence of epidemics causes by mutated strains [11][12][13].S gene sequence analysis is critical for assessing the effectiveness of existing vaccines and for developing additional vaccines.Nucleotide sequence comparison has demonstrated that the S1 domain has a higher mutation rate than the S2 domain and is characterized by higher diversity.In addition, the S1 subunit can bind to specific receptors [14].According to phylogenetic analyses of the S gene, the PEDV strains were divided into two distinct clusters, genogroup 1 (G1; classical) and genogroup 2 (G2; field epidemic or pandemic) [12,15], and S-INDEL strains [16].Most epidemic strains are in the G2 group [15].The present study was performed to analyze S1 gene variations in prevalent PEDV strains and discovered new clinical strains which warrant the development of new vaccine strain.

Sources of samples and viruses
A total of 387 samples were used for the diarrhea virus analysis, including feces and small intestine tissue, were collected from 387 individual diarrheic pigs on 38 pig farms in 4 regions of Eastern China (Zhejiang, Jiangsu, Fujian Provinces, as well as Shanghai) from April 2013 to January 2016.PEDVpositive samples were obtained from the Animal Disease Control Center laboratory of Jinhua Polytechnic.All animal work was approved by Jinhua Polytechnic School Animal Care and Use Committee and the methods were carried out in accordance with the approved guidelines.

Reagents, instruments and primers
PrimeScript One-Step reverse transcription polymerase chain reaction (RT-PCR) Kit Ver.2, RNAiso Plus (total RNA extraction reagent), TaKaRa Ex Taq Hot Start, and DL2000 DNA markers were purchased from Takara Bio Biotechnology (Dalian, China).Agarose were purchased from Sangon Biotech (Shanghai, China).Chloroform, isopropanol, and ethanol were purchased from Shanghai Chemical Reagent (Shanghai, China).The S1000TM Thermal Cycler, Bio-Rad Gel DocTM XR+ UV gel imaging analysis system, and the nucleic acid electrophoresis instrument were purchased from Bio-Rad (Hercules, CA, USA).
Nucleotide sequences of the PCR primers of PEDV-S1 were designed by State Key Laboratory of Agricultural Microbiology in China as follows: Forward: 5′-CCACCATGAAGTCTTTAACTTACTTCTGGT-3′, Reverse: 5′-GAAATTGGCTGTTCATGACTC-3′.Correctly sized PCR products were 1497 bp, and were verified by agarose gel electrophoresis and DNA sequencing.The primers were synthesized by Suzhou Genewiz Bio Technology (Suzhou, China).

Sample pretreatment
The samples were collected from the carcasses of pigs that had died of known or suspected viral diarrhea.Tissue segments of the small intestine (1-2 cm) were placed in homogenate tubes, treated with 250 µL of phosphate-buffered saline, and subjected to highspeed homogenization.This was followed by freezing and thawing three times and centrifugation at 13,400 × g for 5 minutes.Finally, 200 µL of each supernatant was stored at −20°C until analyzed.
Each fecal sample was diluted 10-fold with 10 mM phosphate-buffered saline, vortexed, and centrifuged for 5 minutes at 13,400 × g.The supernatants were stored at −20°C.

RNA extraction
All supplies used for RNA extraction, including deionized water, were treated with 0.1% diethyl pyrocarbonate (DEPC) to avoid RNase contamination.Each liquid sample (200 μL) was combined with 1 mL RNAiso Plus and RNA was extracted according to the manufacturer's instructions.

One-step RT-PCR
The extraction of PEDV template RNA was performed following the PrimeScript One-Step RT-PCR kit instructions.The reaction included 2 μL of PrimeScript One-Step enzyme mix (including PrimeScript reverse transcriptase, TaKaRa Ex Taq Hot Start, and RNase inhibitor), 25 μL of 2× One-Step buffer (including reaction buffer and dNTP mixture (final concentration, 400 μM)), 1 μL each of the forward and reverse primers (10 μM), 5 μL template RNA, and DEPC water to 50 μL.
The thermal cycling profile was as follows: 50°C for 30 minutes; 94°C for 2 minutes; 35 cycles of 94°C for 30 seconds, 54°C for 30 seconds, and 72°C for 1.5 minutes; and 72°C for 10 minutes.The PCR products were electrophoresed in 1.5% agarose gels.

Sequencing and sequence alignment of PCR products
The selection criteria for the studied samples: First of all, they must be PEDV-positive samples by rapid detection method [17] from 387 samples.Secondly, their PEDV S1 gene can be successfully amplified and sequenced.Thirdly, the regional representation was also considered.The PEDV S1-positive PCR products were sent to Suzhou Genewiz Bio Technology for sequencing, sequence alignment, and confirmation of the sequencing results using NCBI BLAST software.Gene sequences were analyzed with DNAMAN (LynnonBiosoft), MegAlign (DNASTAR, Inc.), Vector NTI 11.5 (Invitrogen Corporation), and MEGA (Molecular Evolutionary Genetics Analysis) 6.0 software.

Amplification and analysis of S1 gene sequences
The PEDV-positive rate in the tested samples was 54.5% (211/387) by rapid detection method (Table 1).In this study, according to the selection criteria, 37 strains were finally established.The PEDV S1 genes from 37 selected PEDV-positive samples were amplified by RT-PCR.The amplified S1 sequences were 1494-1497 nucleotides (nt) long; these sequences encoded amino acids (aa) 1-497 and were positioned between nt 20634 and 22131 of the total PEDV S gene sequence.The S1 sequences were compared on the NBCI website, and the target sequence was confirmed.
A similar phenomenon was observed at the amino acid level.The homology of the amino acid sequences was 95.71% as a whole.The 37 endemic strains had highest homologies to the reference strains K14JB01 (Korea) and OhioVBS2 (USA; both 99.8%), whereas homology to the vaccine CV777 strain was lower, at 86.2%-87.8%.
The signal peptide cleavage site and number were predicted to check whether signal peptide has been affected which is relevant to the protein's targeting at virus surface using SignalP 4.1 (CBS Prediction Servers) [19] online signal peptide prediction software.The results showed that the 37 epidemic strains and the CV777 standard strain all contained only one signal peptide.The signal peptide of the PJ-2015-1 strain spanned aa l-26, with the cleavage site at aa 26; the rest of the strains contained signal peptides from aa l-19, with cleavage at aa 19.

Discussion
The major symptom of diarrheic piglets caused by PEDV is watery, gray-yellow, or gray diarrhea, which is sometimes preceded by vomiting.Dehydration in piglets is serious, with death occurring within 2-4 d, and a nearly 100% mortality rate.In a laboratory epidemiological survey, porcine epidemic diarrhea (PED) was much more prevalent from November to January in 2010-2012 [20].The incidence of PEDV in piglets was 10%-100%, and the mortality rate reached 40%-92.58% on some PEDV-exposed pig farms.In the present study, the prevalence of PEDV was as high as 54.5%, and was the main viral pathogen for diarrhea in Eastern China, consistent with reports on PEDV in other provinces in the past five years [21][22][23].
In this study, G2 was divided into two subgroups including G2-1 and G2-2 by evolutionary tree software based on certain algorithms, reflecting the genetic distance and the branch of evolution.G2-2 consisted of two Korean strains, Chinju99 and KNU-0905.The 37 strains were concentrated in the G2-1 subgroup, and all contained mutations, deletions, and insertions (mainly mutations and insertions) compared with the CV777 vaccine strain and older Chinese isolates.PEDV evolutionary strains were prevalent in China, as with the exception of earlier isolated strains such as the CH/S, LJB/03, DX, and JS-2004-2, the 37 strains obtained in this study were, in terms of homology and evolutionary relationships, more similar to strains isolated in recent years in South Korea (K14JB01, KNU-1401, and KNU-1307), Vietnam (VN-KCHY-310113-2013 and HUA-PED47), Thailand (22-53AG0211, 1-55ST0412, and 6-56ST0413), the USA (OhioVBS2), and Canada (2014-022).Thus, PEDV undergoes continuous variation, and the strains isolated after 2013 are different from earlier strains (such as JS-2004-2), consistent with similar reports [24,25].Moreover, strains recently isolated in Asia, the USA [26][27][28][29][30][31], Canada [32], and Europe [16] display high homology.Although the route of transmission is not clear, it has been theorized that contamination of feed or feed ingredients [32,33] or equipment and vehicles used to transport pigs [27] are the main routes of transmission, although further investigation and surveillance are required.However, the homology between the recent endemic strains and the PEDV vaccine strain currently in widespread use is not high, which may impact the protective effect of the vaccine.
The prediction of the N-glycosylation sites of the PEDV S protein has important biological significance [32].A large number of aa mutations have been reported in the S1 region [28].In this study, the S1 proteins of the 37 strains were all predicted to have ten glycosylation sites and 41 amino acid mutations compared with the CV777 vaccine strain which has eleven glycosylation sites.In contrast, the two strains from Vietnam had seven glycosylation sites, six strains (CV777, USA-Iowa107-2013, virulent DR13, attenuated DR13, vaccine strain CV777, and JS2008) had 11, and two strains from South Korea (SM98, 83P-5) had 12.Some amino acid mutations and insertions changed the N-glycosylation sites.Because the S1 domain is the major epitope region, these sites may alter the antigenicity and pathogenicity of the strains, and further studies addressing this are required.However, signal peptide prediction results for the 37 strains were similar to the CV777 strain, indicating that the signal peptide and cleavage site were stable.

Conclusion
Given the increasing incidence of swine PEDV in China, PEDV coinfection with other viral pathogens will further increase the incidence of PED, which occurs widely in China.A diarrhea epidemic occurred despite immunization with attenuated or inactivated diarrhea triple vaccine (PEDV+TGEV+GARV) or bivalent vaccine (PEDV+TGEV) at 43.5% of pig sites [34], indicating that the protective effect of the current PEDV vaccine needs further improvement [35].But the CV777 vaccine is G1 genogroup and samples are G2 genogroup, like Sato et al. [36] show partial protection against G2 genogroup with G1 genogroup vaccine.Because the vaccine protection can depend on vaccine strain and other factors such as coinfection with other virus, inactivation method of virus or vaccine administration way etc.There are large variations in the S1 genes of Eastern China PEDV strains compared with the Chinese vaccine strain CV777, indicating rapid gene mutation in epidemic PEDV strains.More effective surveillance and management strategies need to be developed for porcine diarrhea.Our study also highlights the need to improve current vaccines and simultaneously take steps to enhance the immunity of sows to control the spread of pathogens to piglets.Future work is needed to develop a new vaccine strain different from CV777 based on our present study such as mix of inactivated virus G1+G2 genogroup to enhance the vaccine efficacy against newly emerged PEDV.

Figure 1 .
Figure 1.Analysis and comparison of the nucleotide sequences of the S1 genes of 37 epidemic strains.

Figure 2 .
Figure 2. Phylogenetic tree of the S1 genes of 37 epidemic and reference PEDV strains.The 37 endemic PEDV strains are marked by black stars, Vietnam strains by black circles, Thailand strains by black squares, Chinese local strains isolated in recent years by black triangles, Japanese strains by white squares, South Korea strains by white circles, Taiwan strains by white star, and American and Canadian strains by white triangles.Virus CV777 is the Chinese strain JN599150, attenuated by cell passage.

Figure 3 .
Figure 3. Analysis and comparison of the amino acid sequences of the S1 genes of 37 epidemic strains.