Differentiation of PEDV Classical Attenuated Vaccine Strains from Wild-type Strains using One-Step Real-Time Fluorescent Reverse Transcription PCR Assay Targeting ORF1 Nucleotides Deletion Region

Porcine epidemic diarrhea virus (PEDV) is a pathogen causing serious disease and resulting in severe economic losses in the swine industry. In recent years, although China has adopted a large-scale vaccine immunization strategy, many types of PEDV strains, including classical attenuated vaccine strains, have been discovered in the immunized pig herds. Therefore, monitoring the prevalence of different types of PEDV strains is particularly important for the production of pigs and the safety evaluation of related attenuated vaccines In the study, a one-step real-time uorescent reverse transcription PCR (one-step real-time RT-PCR) assay targeting 24-nucleotide deletion in the ORF1 region of three PEDV classical attenuated vaccine strains (derived from classical strains) was established, which could effectively distinguish PEDV classical attenuated vaccine strains and wild-type strains.


Abstract
Background Porcine epidemic diarrhea virus (PEDV) is a pathogen causing serious disease and resulting in severe economic losses in the swine industry. In recent years, although China has adopted a large-scale vaccine immunization strategy, many types of PEDV strains, including classical attenuated vaccine strains, have been discovered in the immunized pig herds. Therefore, monitoring the prevalence of different types of PEDV strains is particularly important for the production of pigs and the safety evaluation of related attenuated vaccines

Methods
In the study, a one-step real-time uorescent reverse transcription PCR (one-step real-time RT-PCR) assay targeting 24-nucleotide deletion in the ORF1 region of three PEDV classical attenuated vaccine strains (derived from classical strains) was established, which could effectively distinguish PEDV classical attenuated vaccine strains and wild-type strains.

Results
In our study, the RNA detection limits for PEDV wild-type strains and classical attenuated vaccine strains were 3.0×10 3 copies and 3.0×10 2 copies, respectively. This assay was highly speci c for PEDV, with no cross-reactivity for other viruses, causing diarrheal disease. A total of 117 swine fecal samples were analysed by this established real-time RT-PCR assay, indicating that classical attenuated vaccine strains were present in the swine herds in Gansu province, China. Additionally, a pair of primers and two probes of the established assay can be placed in one reaction tube to distinguish PEDV classical attenuated vaccine strains and wild-type strains.

Conclusion
Our results provided an effective and cheap technology platform for clinical rapid identi cation testing and epidemiological investigations of PEDV wild-type strains and classical attenuated vaccine strains Background Porcine epidemic diarrhea virus (PEDV), an enveloped virus belonging to the genus Alpha coronavirus and the family Coronaviridae, can induce acute diarrhea and vomiting in newborn suckling and weaning pigs [1]. In 1971, PED was rst reported among feeder and fattening pigs in England [2]. Since then, the diarrheal disease caused by PEDV has broken out in many regions of the world, including China [3][4][5].
After 2010, the highly pathogenic PEDV variant strains cause the morbidity rate of newborn suckling piglets to reach 100%, with mortality rates of 80-100%, which result in signi cant economic losses in the global swine industry[6-8]. Since the 1990s, many Chinese pig farms have adopted vaccination measures to prevent the spread of the disease [9]. At present, CV777-based live attenuated vaccine and PEDV(CV777) and transmissible gastroenteritis virus (TGEV)-based dual attenuated vaccines have been extensively applied for PED prevention in China [10,11]. There is a short period of carrying attenuated vaccine virus in the orally immunized pig population, and the possibility that the wild-type strains and the attenuated vaccine strains will exist simultaneously in the pig herds [12,13]. In recent study, multiple strains were also found in many immunized pig population in China [4,14,15], and phylogenetic analysis of the whole genome also showed that there are multiple variation sites in PEDV genome [16,17]. The antigenicity of vaccines derived from classical vaccine strains may be altered due to these genetic mutations, resulting in low vaccination e ciency and inability to protect pigs from variant PEDV strains [16]. Those have an impact on accuracy of the epidemiological survey data of the disease to a certain extent.
Porcine epidemic diarrhea virus (PEDV) is a positive-sense, single-stranded virus RNA with a genome of about 28 kb in length [1]. The PEDV genome comprises seven open reading frames in the order of ORF1a, ORF1b, spike (S) gene, accessory gene ORF3, envelope (E) gene, membrane (M) gene and nucleocapsid (N) gene [3,18,19]. The ORF1a and ORF1b region at the 5'end of the genome occupy 2/3 of the entire genome, and the two viral replicase polyproteins (pp1a and pp1ab) encoded by them are cleaved into 16 non-structural proteins (nsp1-nsp16), which are involved in the synthesis and transcription of viral RNA [20]. Among them, the nsp3 is the largest non-structural protein and also a multifunctional protein, which may be closely related to virus replication and transcription [21,22]. A recent variation analysis of 49 PEDV entire-genome sequences in GenBank showed that the N-terminal domain of PEDV nsp3 is a hypervariable region and relatively conservative, and the characteristic mutation sites in this region may potentially serve as markers for the classi cation of PEDV different genotypes [16,23]. Currently, many molecular detection methods targeting N, M, S, ORF3 genes have been developed in the laboratory [24][25][26][27], such as traditional PCR [13,28], nanoparticle-assisted RT-PCR[29], and real-time RT-PCR [12,27,30] methods to detect or distinguish the wild-type strains (classical wild strains and variant strains) and attenuated vaccine strains, these methods is useful for understanding the epidemic status of different strains. In previous studies, researchers rarely used the ORF1 region as the genetic characteristics to distinguish PEDV attenuated vaccine strains and wild-type strains. In present study, we found that Nterminal domain of ORF1 region nsp3 in the three PEDV classical attenuated vaccine strains (PEDV attenuated vaccine strain KC189944, attenuated CV777 and DR13) and ve Vero-cell-adapted isolates (JS2008, SDM, SQ2014, SC1402, HLJBY) bore a 24-nucleotide deletion (see Fig.1 and Table S1 ) compared with 38 PEDV wild-type strains whose sequences were available in GenBank[31]. These three attenuated vaccine strains and ve Vero cell-adapted isolates were arti cially cell-passaged and were not original strains in nature. Unless they are used as live attenuated vaccines for vaccination and spread to the pig population, it is impossible to detect exactly the same strains in the eld. Due to these discoveries, we developed a convenient, cheap, e cient, secure and reliable one-step real-time RT-PCR assay to distinguish PEDV classical attenuated vaccine strains and wild-type strains.

Methods
Viruses and clinical samples PEDV attenuated vaccine strain CV777 and Vero-cell-adapted isolate JS2008 were passaged in Vero E6

Design of primers and probes
According to the presence of a 24-nucleotide deletion ( Fig.1) in the ORF1 regions of three PEDV classical attenuated vaccine strains and ve Vero-cell-adapted isolates, a pair of primers and two probes were designed (synthesized by Sangon Biotech, Shanghai, China) to differentiate PEDV classical attenuated vaccine strains from wild-type strains (Table 1). Additional primers and probes (Table S2)

Generation of RNA standards
The rst strand PEDV classical attenuated vaccine strain CV777 and wild strain DX cDNA were synthesized by reverse transcription with PrimeScript™ rst strand cDNA synthesis kit (Takara, Dalian, China). A PCR fragment of the PEDV ORF1 region was ampli ed using primers F-V/R-V (Table 1) from PEDV classical attenuated vaccine strain CV777 cDNA and named PEDV-V/qRT-PCR. Another PCR fragment of the PEDV ORF1 region was ampli ed using primers F-W/R-W (Table 1)  The one-step real-time RT-PCR assay Real-time RT-PCR assay was carried out using a Bio-Rad CFX Manager (Bio-Rad, USA) instrument. The primers and probes were designed to amplify the PEDV gene ORF1 region in this study or the PEDV S1 domain [12]. The reactions were carried out using the One Step PrimeScript® RT-PCR Kit (Perfect Real Time) (Takara, Dalian, China) with reaction system as follows: 2×One Step RT-PCR Buffer 12.5μL, The one-step RT-PCR assay Eight pairs of primers (Table S2)  Step Buffer 25 μL, each primer 10pmol, 1μl of viral RNA/DNA or 4μL sample RNA/DNA, and ddH 2 O to a total volume of 50 μL in each PCR tube, and cycled as follows: 50℃for 45min, 94℃for 2 min, followed by 35 cycles of 95°C for 5 s, 55°C for 30s, and 72℃for 60s, with a nal extension at 72°C for 10 min.
Speci city and sensitivity analysis of PEDV one-step real-time RT-PCR assay All viral RNA and DNA samples were quantitated using a ND-2000c spectrophotometer (Thermo Fisher Scienti c, Waltham, USA) in our laboratory. Ten nanograms of RNA or DNA extracted from PEDV classical attenuated vaccine strain CV777, Vero-cell-adapted isolate JS2008, PEDV DX, TGEV, PKV, PPV, PDCoV, and PCV-2 in three replicates was used as template for speci city analysis of PEDV one-step real-time RT-PCR assay. PEDV one-step real-time RT-PCR assay was performed using a Bio-Rad CFX Manager instrument (Bio-Rad, USA).
For sensitivity analysis of PEDV one-step real-time RT-PCR assay, ten-fold serial dilutions of both RNA standards were used as template (range: 3.0×10 10 -3.0×10 1 copies). One microliter of each PEDV RNA standard serial dilution (range: 3.0×10 1 -3.0×10 10 copies) was applied to evaluate the dynamic detection range of one-step real-time RT-PCR assay. Each experiment was repeated three times and regression analysis was performed using Bio-Rad CFX Manager (Bio-Rad, USA) to determine detection limits.

Analysis of clinical samples using PEDV real-time RT-PCR and RT-PCR assay
A total of 117 fecal samples were collected from seven pig farms with the background of immunizing with CV777-based monovalent or bivalent attenuated vaccines, six of which from six piglets with oral attenuated vaccine CV777 and showed no clinical symptoms of diarrhea, and the remaining 111 fecal samples were obtained from 111 piglets with clinical status of diarrhea. All fecal samples were used to evaluate the reliability of the established PEDV one-step real-time RT-PCR assay. Another real-time RT-PCR and RT-PCR assays were compared, and all results were shown in Table 3. All positive PCR products were sequenced by TsingKe (XiAn, China).

Results
Speci city, sensitivity, and repeatability analysis of PEDV one-step real time RT-PCR assay For the speci city of PEDV one-step real time RT-PCR assay, only PEDV wild-type strain DX showed a uorescent signal using the primers F3365/R3445 and probe P3390-W (Fig.2.A). only PEDV classical attenuated vaccine strain CV777 or Vero-cell-adapted isolate JS2008 showed uorescent signals using the primers F3365/R3445 and probe P3388-V (Fig.2.B). However, TGEV, PKV, PPV, PDCoV, and PCV-2 showed no uorescent signals in this assay, indicating good speci city (Fig.2). The primers F3365/R3445, probes P3390-W, and P3388-V could be combined in a tube for simultaneously detecting PEDV attenuated vaccine strain CV777 and PEDV wild-type strain DX ,without cross-reactivity for each other (Fig.3).
The detection limits for PEDV wild-type strains and classical attenuated vaccine strains were 3.0 × 10 3 copies/reation (Fig.4) and 3.0 × 10 2 copies/reation, respectively (Fig.5). Using 10-fold serial dilutions of RNA standard as template, each experiment was repeated three times, to establish standard curves for PEDV wild-type strains (Fig.4.B) and classical attenuated vaccine strains (Fig.5.B). As shown in Table 2, the detection limit of the PEDV wild-type strains were 3.0×10 3 RNA copies/reaction for PEDV wild-type strain DX standard RNA, and the detection limit of the PEDV classical attenuated vaccine strains were 3.0×10 2 RNA copies/reaction for PEDV classical attenuated vaccine strain CV777 standard RNA. The epeatability of this established assay was evaluated using two ORF1 region standard RNAs of PEDV classical attenuated vaccine strain CV777 and wild-type strain DX, respectively, with coe cients of variation 0.50-6.8 (Table 2). 117 fecal samples from 117 pigs with suspected PEDV infection were subjected to PEDV one-step realtime RT-PCR (a), real-time RT-PCR (b), and RT-PCR assays (c). PEDV wild-type strains were detected in 83.76, 83.76, and 79.80.% of all samples, respectively, and PEDV classical attenuated vaccine strains were detected in 6.84, 6.84 and 5.98% of all samples, respectively ( Table 3). As shown in Table 3, of the 17 samples that tested negative in the RT-PCR assay(c), 11 samples were PEDV-negative and the remaining 6 samples were PEDV-positive ( ve wild-type strains and one classical attenuated vaccine strain) as detected by the one-step real-time RT-PCR (a) and real-time RT-PCR assays (b). As shown in Fig.6, through the analysis of the established one-step real-time RT-PCR assay, the viral load of 106 PEDVpositive samples was concentrated between 10 3.265 -10 6 copies/reation. Sequencing results for all PEDVpositive samples indicated that the PEDV one-step real time RT-PCR assay (a) had high speci city and sensitivity. In addition, TGEV and PKV were detected in the 11 PEDV-negative samples and 8 PEDVpositive samples (detailed data no shown), indicating that the co-infections of multiple diarrhea viruses have occurred in some pig farms.

Discussion
Despite the current vaccination policy in China, multiple types of PEDV strains still appeared in pig farms, especially the variant strains that began in southern China in 2010, causing huge losses to the pig industry [6,13]. A recent phylogenetic tree study of Chinese PEDV strains and some PEDV representative strains of other countries revealed that the full-length genomic sequences can be divided into two independent subgroups, namely GI (classical strains, GI-a and GI-b) and GII (variant strains, GII-a, and GIIb) [17,32]. The PEDV strains of the GI-b subgroup contained Vero-cell-adapted vaccine strains (PEDV attenuated vaccine KC189944, attenuated CV777 and DR13) and ve Vero-cell-adapted isolates (JS2008, SDM, SQ2014, SC1402, HLJBY) [15,17,[32][33][34][35][36]. By alignment analysis of ORF1 region sequence in the 46 PEDV strains released by GenBank ( Fig.1 and Table S1 ), we found that three classical attenuated vaccine strains (KC189944, attenuated strain CV777 and DR13, GI-b) and Vero-cell-adapted isolates (JS2008, SC1402, SQ2014, HLJBY and SDM, GI-b) have a novel 24-nucleotide deletion in the N-terminal domain of ORF1 region nsp3 compared with 38 wild-type PEDV strains (GI-a and GII). PEDV Vero-celladapted vaccine strains belonging to GI-b subgroup not only have nucleotides variation in the ORF3[16], but there are also 24-nt deletions in the ORF1 region ( Fig.1 and Table S1). Based on these discoveries, a one-step real-time RT-PCR assay was developed to distinguish PEDV classical attenuated vaccine strains and wild-type strains. The 24-nucleotide-deletion pattern in the ORF1 region of PEDV Vero cell-adapted strains could be the marker of adaptation to Vero cell culture, and it is a valuable tool for monitoring of the persistence of classical attenuated vaccine strains and epidemiologic research of PEDV infection in swine herds as well as stability and safety analysis of classical attenuated vaccines. However, there are some situations that need our attention. For example, since the ORF1 sequence of some candidate vaccines has not been reported, especially the vaccine candidates from highly virulent strains (genotype 2a) emerged after 2010, we are not sure whether the ORF1 of these vaccine candidates derived from nonclassical attenuated vaccine strains have the same 24 nucleotides deletion pattern. If this pattern exists in all cell-adapted strains during continuous passages in Vero cells, it can be used as a genotyping marker just like nucleotide deletions of Spike or ORF3 genes in the Vero-cell-adapted strains[37, 38].
Moreover, multiple PEDV strains can co-exist in some co-infection events, and it is possible that the wildtype virus repairs the 24nt deleted region of the classical attenuated vaccine strain. If the 24 nucleotides deletion pattern of the attenuated vaccine strains is repaired, the method we established will not apply. Nevertheless, Commercial vaccines based on classical strains are extensively applied in Chinese pig farms and play a very important role in controlling PEDV infections, while vaccines derived from nonclassical attenuated vaccine candidates are being developed and not yet commercially available. Therefore, our method is safe, accurate, and reliable, and can be applied to the identi cation of classical attenuated vaccine strains and the safety evaluation of attenuated vaccines in pig farms The one-step real-time RT-PCR assay could only detect PEDV, with no cross-reactivity with other enteroviruses. The limits of detection for PEDV wild-type strains and classical attenuated vaccine strains in the one-step real-time RT-PCR assay were 3.0×10 3 copies and 3.0×10 2 copies, respectively. Compared with the other two methods (Table 3), the detection results of 117 fecal samples showed that the our established assay had highly PEDV positive diagnosis agreement with real-time RT-PCR(100%) and RT-PCR assays(95.6%) (Table 3), respectively, indicating that the established one-step real-time RT-PCR assay have high sensitivity, rapidity, and accuracy. The detection results of 117 fecal samples in our established one-step real-time RT-PCR assay, 98 PEDV wild-type strains were detected, including PEDV classical wild strains and variants, 8 PEDV classical attenuated vaccine strains were detected, and TGEV and PKV were detected in the 11 PEDV-negative samples and 8 PEDV-positive samples (data no shown).
The sequencing results of 8 positive samples tested as classical attenuated strains in this study showed that they have the same nucleotide deletion positions as the ORF1 and ORF3 fragments in the classical attenuated vaccine CV777. These results showed that PEDV attenuated vaccines have existed in Gansu province, China and the co-infections of multiple diarrhea viruses have occurred in some pig farms. Therefore, it is necessary to take measures to dynamically monitor the co-infection of other enteric pathogens with PEDV in pig farms. Besides, We did not successfully isolate live virus using Vero-E6 cells in the positive samples tested as PEDV classical attenuated vaccine strains. This may be due to low live virus content or only nucleic acid fragments in the sample.

Conclusion
This one-step real-time RT-PCR assay was developed to distinguish between classical attenuated vaccine strains that were arti cially inoculated and wild-type strains during epidemiological surveillance, and to provide technical support for collecting more accurate epidemiological data of PEDV infection. Since, various PEDV strains may coexist in the same environment, even the same pig, which may cause the vaccine strain to revert to virulence by genetic recombination with other types of PEDV strains. Thus, monitoring the spread of attenuated vaccine strains and changes in virus titer could provide an important reference for evaluating immune protection effect of attenuated vaccine strains and the prevention and control of PEDV. Addtionally, our method only use a pair of primers and two probes that placed in one reaction tube to detect different types of PEDV strains, without cross-reactivity between each other. When distinguishing different types of PEDV strains in a sample, a tube of ampli cation enzyme, a pairs of primers, and a tube of PCR reaction tube can be save, and this also reduced the cost of detecting sample in the commercial application. Moreover, this method allows quantitative calculation of viral load, and has practical value for epidemiological investigations of wild-type strains and classical attenuated vaccine strains. Simultaneously, combined with other methods for detecting PEDV, it can be effective for the genotyping and prevalence of various strains of PEDV. Finally, it also represents an alternative detection tool for preliminary identi cation of clinical samples.  Availability of data and material The data of this study are available from the corresponding author on reasonable request.

Additional le
All data generated or analysed during this study are included in this published article (and its additional le Table S1 and Table S2)   Table S1. The name, accession number, and search website of the PEDV strain in the study.  ORF1 regions alignment results of 38 PEDV wild-types strains and 8 Vero cell-adapted strains whose sequences were available in GenBank, and 24-nucleotide deletions in the ORF1 regions of three PEDV classical attenuated vaccine strains and ve Vero-cell-adapted isolates.

Figure 3
Speci city of primers and probes in a PCR reaction tube: 1-2: PEDV classical attenuated vaccine strain CV777 RNA,Vero-cell-adapted isolate JS2008, wild-type strain DX RNA, two probes, and a pair of primers were placed in a PCR reaction tube ; 3: ddH20, two probes, and a pair of primers.