Porcine epidemic diarrhea virus (PEDV) is a fatal swine enteric coronavirus that belongs to the subgenus Pedecovirus of the genus Alphacoronavirus within the family Coronaviridae of the order Nidovirales [1, 2]. PEDV is a large, enveloped virus with a non-segmented, positive-sense RNA genome of approximately 28 kb comprising seven canonical coronaviral genes [1, 3]. The virus can be classified into two major genotypes with at least two sub-genotypes: G1 (classical G1a and recombinant G1b) and G2 (local epidemic G2a and global epidemic G2b) [1, 3]. The advent of the highly pathogenic (HP)-G2b PEDV resulted in a pandemic that ruined pig-producing nations in America and Asia during 2013–2014, posing socioeconomic threats to the global swine industry [1, 3]. South Korea also suffered from explosive nationwide outbreaks of HP-G2b PEDV, causing the loss of an estimated one million neonatal piglets during 2013–2014 [3, 4]. Since then, HP-G2b PEDV has continued to cause nationwide epidemics each year and has evolved along with genetic divergence [1]. Consequently, the HP-G2b PEDV strains circulating in South Korea are genetically clustered into eight clades with two subclades on the basis of geographical origin [1].
From December 2023 to January 2024, acute outbreaks of severe diarrheic diseases accompanied by low neonatal mortality (<10%) were reported in multiple unvaccinated farrow-to-finish herds in southern South Korea. Clinical samples (feces or small intestine) collected from diarrheic or dead piglets at seven farms were submitted to our laboratory for diagnosis and subjected to virus-specific RT-PCR assays as described elsewhere [5–7]. All specimens from the seven farms were PEDV-positive, with cycle threshold (Ct) values ranging from 14.4 to 27.2; however, no other viral pathogens causing neonatal diarrhea were identified in these cases. We then determined the entire sequences of the spike (S) genes of the seven isolates detected from each farm as previously published [5–7] and deposited their sequence data in the GenBank database under accession numbers PP441968–441974. Nucleotide (nt) sequencing analysis revealed genetic divergence (different barcode profiles representing a distinct locus from the HP-G2b prototype sequence) among the strains designated as GNU-2389–92 and GNU-2401–3 (Fig. 1), exhibiting 96.4%–98.8% amino acid (aa) homology, and displayed 96.0%–99.4% aa identity to the South Korean HP-G2b strains (Supplementary Table S1). All strains belonged to HP-G2b with the genetic marker, S insertions and deletions (INDELs), compared with the classical G1a prototype CV777 strain [1, 3, 7]. Interestingly, unique deletion (DEL) signatures were identified in the N-terminal domain (NTD; residues 19–233) of S1 from all isolates in comparison with the HP-G2b South Korean prototype strain KNU-141112. The S genes of all PEDV strains determined in this study harbored a novel deletion (DEL) of duad (GV), (VN), or (ST) aa residues at positions 60 and 61 (GNU-2389), 61 and 62 (GNU-2390–2 and GNU-2401–2), or 63 and 64 (GNU-2403), respectively, in their S1 NTD, named S-DEL2 (Fig. 1A). These S-DEL2 patterns in S1 NTD were unprecedented in the genome sequence of other PEDV field strains retrieved from the GenBank database. Due to the DEL of duad (60GV61), (61VN62), or (63ST64) residues in the S1 NTD of the S-DEL2 variants, the size of their S genes was 4,155-nt long, encoding a 1,384-aa protein, which was 6-nt shorter than those of the conventional HP-G2b strains. Further prediction of N-linked glycosylation sites confirmed the loss of one putative N-glycosylation sequon ‘NST’ at N62 on the S1 NTD of all isolates, except for GNU-2389, which is commonly found in classical HP-G2b field viruses (Fig. 1A).
To further uncover the genotypic features of the S-DEL2 variants, their complete genomic sequences were deciphered using the Sanger method as described previously [5–7]. We could not achieve whole genome sequencing (WGS) of four isolates directly from the clinical cases, probably because of high Ct values (>25), representing the low viral load present in the respective PEDV-positive samples. However, the remaining three isolates, GNU-2389, -2391, and -2401, whose full-length genomes were successfully sequenced, and the genomic sequences were deposited at GenBank under accession numbers PP441972–441974. WGS revealed no additional INDELs throughout the entire genome of the fully sequenced S-DEL2 strains, except for the S-coding region (Fig. 1B). The S-DEL2 viruses shared a high level of nt sequence similarity with each other (98.7%–98.8%) and with other domestic HP-G2b strains at the genome level (98.5%–99.6%) (Supplementary Table S2). Detailed information, including the number of nt or aa differences and the percentage of homology between each S-DEL2 variant and the domestic prototype strain KNU-141112, is shown in Supplementary Table S3.
To verify genetic relatedness, we conducted phylogenetic analyses using the S and full-length genome of PEDV determined in this study and those accessible from GenBank, as described previously (Fig. 2) [5–7]. As depicted in Fig. 2A, the S protein-based phylogeny clearly demonstrated four genotypes, viz., G1a, G1b, G2a, and G2b. The South Korean G2b strains were further categorized into six clades (NW, KJ, CK, JH, JD, and NC) and two subclades (CK.1 and CK.2). The novel S-DEL2 variants belonged to the G2b cluster along with the global and domestic HP-G2b strains and were classified into CK (GNU-2402), CK.1 (GNU-2403), CK.2 (GNU-2389), or NC (GNU-2390–2 and GNU-2401). Similarly, whole-genome phylogeny indicated that the S-DEL2 strains were clustered within the same genogroup as the global and domestic G2b strains (Fig. 2B).
Since the national catastrophe of HP-G2b PEDV during 2013–2014, this virus has become endemic in South Korea and has evolved to undergo genetic and pathogenic heterogeneity [1]. Recently, PEDV outbreaks with reduced morbidity and mortality occurred in several commercial swine farms with no previous record of PEDV infection or vaccination. Although we initially considered the concurrent introduction of the genetically identical PEDV strain into those farms, our genetic and phylogenetic analyses revealed distinct differences in the barcode profile and clade categories among the HP-G2b viruses identified in the affected herds. Despite the abovementioned genetic varieties, these viruses demonstrated genetic similarity with the novel S-DEL2 signature at positions 60 and 61, 61 and 62, or 63 and 64 in the S1 NTD. The PEDV S protein mediates viral entry into target cells and thus triggers host immunity to produce neutralizing antibodies [3]. In particular, the S1 NTD contains one neutralizing epitope region, termed NTD/S0 (residues 19–220) (Fig. 1A) [8]. Moreover, various novel G2b variants with small or large INDELs in S1 have been reported in multiple countries, suggesting their association with virulence [9]. Therefore, genetic drift, including INDELs, and shift (i.e., recombination) in the S gene can allow PEDV to evade the host antibody response and/or alter viral pathogenicity [5, 9–11]. The S-DEL2 viruses lost the N62 glycan motif on the NTD/S0 domain of S, resulting from the emergence of S-DEL2 in the S1 NTD, possibly causing the modification of the S protein conformation that may contribute to the antigenic variation of PEDV. In fact, the S-DEL2 variants exhibited a conformational change on residues 51–69 residing on the NTD/S0 region compared with the S 3D structure of KNU-141112, which in turn may confer an immune evasion strategy to the virus (Supplementary Fig. S1). Considering the disease severity in S-DEL2 virus-infected farms, the DEL of duad residues in the S1 NTD may be associated with PEDV virulence in neonatal piglets. The S1 NTD is responsible for sialic acid (SA) binding, which facilitates viral entry by recruiting host SA-containing components to invade the thick mucus barrier close to the enterocytes [12]. More interestingly, the pentad “GENQG” at positions 56–60 in the S1 NTD was proven to be the major residue for the SA binding activity of the PEDV S protein, and infectious clone-derived mutants with the duad (GE) or pentad (GENQG) DEL displayed the attenuation of SA-dependency and virulence [11]. Because the duad “60GV61”, “61VN62”, or “63ST64” residues are located next to or adjacent to the pentad “56GENQG60” in the S1 NTD, it is anticipated that the S-DEL2 variants would weaken the SA-dependent viral entry and diminish the pathogenicity of PEDV. Considering the absence of such field S-DEL2 isolates that can grow in cell culture, the cutting-edge application of reverse genetics (RG) is essential for pursuing loss-of-function research to corroborate the abovementioned aspects. With the availability of the PEDV RG platform in our laboratory [10], our next step is to investigate the precise role of S-DEL2 in the antigenicity, SA dependency, and virulence of PEDV. Nevertheless, our sequence data provide insights into understanding the genetic and pathogenic variation of PEDV field strains. The present study also underscores the need for implementing continuous monitoring and surveillance investigations, followed by swift data sharing and accessibility, to establish PEDV management for preventing future outbreaks/pandemic.