The emergence of novel sparrow deltacoronaviruses in the United States more closely related to porcine deltacoronaviruses than sparrow deltacoronavirus HKU17

Dear Editor,Coronaviruses (CoVs) are single-stranded, positive-sense, enveloped RNA viruses belonging to family Coronaviridae and subfamily Coronavirinae, which consist of four genera: Alphacoronav...

porcine deltacoronavirus (PDCoV), has been identified as one of the major enteric pathogens causing diarrhea in pigs since 2014 4,5 . Both clinical and experimental evidence demonstrated that PDCoV is pathogenic to swine and causes similar lesions in the pig intestine as porcine epidemic diarrhea virus (PEDV) 6,7 . PDCoV strains reported from the United States and other countries had 97.1-99.9% nucleotide identity at the whole-genome level 8 , suggesting that a single genotype is circulating worldwide.
Regarding sparrow deltacoronavirus (SpDCoV), one sparrow CoV strain, HKU17-6124, has been reported thus far 3 . The genomic diversity of sparrow deltacoronaviruses remains unclear. We report the detection and genetic characterization of novel SpDCoV in fecal samples collected from sparrows found on swine farms in the midwestern United States.
In January 2017, multiple batches of porcine oral fluid (pen-based) and floor-picked fecal samples were submitted from a grow-finish stage pig farm in Illinois for routine health monitoring. Unexpectedly, multiple oral fluid samples and one fecal sample were tested positive for PDCoV (the Ct value ranged between 34 and 39) by duplex PEDV and PDCoV real-time RT-PCR (see Technical Appendix). However, no clinical diarrhea or other signs were observed among pigs on the index farm, which prompted us to further investigate the source of positive PCR results. Since the pigs were housed in hoop-style buildings with uncontrolled access for wild bird entry and congregation, fecal samples were also collected from 10 sparrows found inside the pig barn (ISU690-1 to −10) and were submitted to our laboratory for testing in January 2017, along with 8 feed samples (ISU690-11 to −18). Surprisingly, 9/10 fecal samples were positive using PCR targeting the PDCoV N gene, with a Ct of 26.3-37.3, while all 8 feed samples were negative. Four additional sparrow fecal samples (ISU42824-1 to −4) collected in June 2017 from the same farm were tested, and one of the samples (ISU42824-3) was positive by PDCoV N gene-based PCR, with a Ct of 23.6.
Similarly, in an unrelated grow-finish swine farm in Minnesota, two oral fluid samples collected from clinically healthy pigs at the end of October 2017 were tested positive (Ct of 33.3 and 34.4), and one sparrow fecal sample (ISU73347-2) collected in early November 2017 from that barn were tested positive (Ct of 22.6) using a PDCoV N gene-based PCR.
The four SpDCoV strains identified in this study share a slightly higher identity (94.4-94.6% and 96.0-96.4%) to a sparrow CoV strain HKU17-6124 than to PDCoV strains (93.4-93.6% and 95.3-95.8%) in their ORF1ab and N proteins, respectively ( Fig. 1b and Supplementary  Table S2). In contrast, the four SpDCoV strains share a significantly lower identity to sparrow CoV HKU17 (58.2-58.6%) than to three PDCoV strains (82.7-87.7%) in their S protein (Fig. 1b and Supplementary Table S2). In addition, they share a relatively lower identity (90.7-91.5% and 93.5-95.5%) to sparrow CoV HKU17 than the PDCoV strains (94.9-95.8% and 96.1-98.1%) in their E and M proteins, respectively (Supplementary Table S2). These findings were also confirmed by phylogenetic tree analysis of amino-acid sequences in which the four SpDCoV strains clustered together and were closely related to PDCoV and sparrow CoV HKU17 strains in ORF1ab, E, M, and N trees ( Fig. 1c and Supplementary Figure S1). However, the four SpDCoV strains were close to PDCoV but distant from sparrow CoV HKU17 in the S tree (Fig. 1d). In the case of non-structural proteins (NS6, NS7a, and NS7b), the SpDCoV strains share a high amino-acid identity with each other. However, SpDCoV ISU690-4/7 has 76.2% and 79% identity to ISU73347 and ISU42824, respectively, in NS7c, which is much lower than the 97.2% identity shared between ISU73347 and ISU42824 (Supplementary Table S3). An 18-nt deletion in the NS7c gene of ISU690-4 and -7 resulted in a protein that was 6 amino acids shorter than that of ISU73347 and ISU42824 in length (Supplementary Figure S2). Overall, these data further support that the sparrow CoVs identified in our study are novel SpDCoV.
Recombination analysis revealed that PDCoV and sparrow CoV HKU17 are highly variable from SpDCoVs in two regions. The most prominent variable region is located in the S-E-M gene region, and another region is located within the ORF1ab gene (~2675 nt to 3435 nt) (Supplementary Figure S3). However, there is no obvious pattern showing that the emergence of SpDCoV strains identified in the study was through recombination between PDCoV and sparrow CoV HKU17.
Comparison of PDCoV PCR primer and probe sequences (targeting the N gene) with SpDCoV sequences revealed that the probe sequence was exactly identical to the sequence observed in SpDCoV, and only 2 and 0/1 nucleotide mismatches were observed in the forward and reverse primers, respectively (Supplementary Figure S4). Therefore, PDCoV real-time RT-PCR can cross-react with SpDCoV, which complicates the molecular detection of PDCoV. It is plausible that contamination with sparrow feces containing SpDCoVs caused non-specific positive results for PDCoV when the oral fluid and fecal samples  collected from clinical healthy pigs were tested using PEDV/PDCoV duplex RT-PCR. This also raises a concern regarding the specificity of detecting PDCoV from environmental and environmentally contaminated samples. To avoid the cross-reaction issue, further modification of the primers and probe for PDCoV real-time RT-PCR is needed. PDCoV, as a causative agent for pig diarrhea, has been reported in many countries, including the United States, Canada, Japan, South Korea, Thailand, Vietnam, Mainland China, and Laos, indicating that PDCoV is widespread in pig populations. However, all known PDCoVs to date have been relatively conserved and shared 97.1-99.9% nucleotide identity at the whole genome level. In contrast, the SpDCoV strains identified in the present study (ISU690-4/7, ISU42824, and ISU73347) as well as one previously identified sparrow CoV strain, HKU17-6124, exhibited higher genetic diversity (82.1-93.9% nucleotide identity at the whole-genome level). Further studies to understand the genetic diversity of SpDCoV by sequencing more SpDCoV strains are warranted. Another important finding is that, due to the relative conservation between PDCoV and SpDCoV N gene sequences in some regions, the PDCoV N gene-based PCR used in this study can cross-react with SpDCoV. Thus, PDCoV PCR results should be interpreted carefully when pig samples that can be contaminated by bird feces or other excretions are tested. The SpDCoV sequence data reported in this study may help determine the evolutionary relationship of various DCoVs and help to understand the interspecies transmission of DCoVs in future studies.