Discovery of a Novel Coronavirus, China Rattus Coronavirus HKU24, from Norway Rats Supports the Murine Origin of Betacoronavirus 1 and Has Implications for the Ancestor of Betacoronavirus Lineage A

Sample collection.

All rodent samples were collected from January 2010 to August 2012 using procedures described previously (5, 14). Samples from southern China were collected from animal markets or restaurants. Samples from Hong Kong were collected from wild and street rodents by the Agriculture, Fisheries and Conservation Department and the Food and Environmental Hygiene Department of the Hong Kong Special Administrative Region (HKSAR), respectively. Alimentary tract samples were placed in viral transport medium containing Earle's balanced salt solution (Invitrogen, NY, USA), 20% glucose, 4.4% NaHCO₃, 5% bovine albumin, 50,000 μg/ml vancomycin, 50,000 μg/ml amikacin, and 10,000 units/ml nystatin, before transportation to the laboratory for RNA extraction. The study was approved by the Committee on the Use of Live Animals for Teaching and Research, The University of Hong Kong, and the Institutional Review Board, The University of Hong Kong/Hospital Authority.

RNA extraction.

Viral RNA was extracted from the samples using a QIAamp viral RNA minikit (Qiagen, Hilden, Germany). The RNA was eluted in 60 μl of buffer AVE and was used as the template for reverse transcription-PCR (RT-PCR).

RT-PCR of the RdRp gene of CoVs using conserved primers and DNA sequencing.

Initial CoV screening was performed by amplifying a 440-bp fragment of the RNA-dependent RNA polymerase (RdRp) gene of CoVs using conserved primers (5′-GGTTGGGACTATCCTAAGTGTGA-3′ and 5′-CCATCATCAGATAGAATCATCATA-3′) designed by the use of multiple-sequence alignments of the nucleotide sequences of available RdRp genes of known CoVs (14, 20). Reverse transcription was performed using a SuperScript III kit (Invitrogen, San Diego, CA, USA). The PCR mixture (25 μl) contained cDNA, PCR buffer (10 mM Tris-HCl, pH 8.3, 50 mM KCl, 2 mM MgCl₂, 0.01% gelatin), 200 μM each deoxynucleoside triphosphate, and 1.0 U Taq polymerase (Applied Biosystems, Foster City, CA, USA). The mixtures were amplified by 60 cycles of 94°C for 1 min, 50°C for 1 min, and 72°C for 1 min and a final extension at 72°C for 10 min in an automated thermal cycler (Applied Biosystems, Foster City, CA, USA). Standard precautions were taken to avoid PCR contamination, and no false-positive result was observed for the negative controls.

PCR products were gel purified using a QIAquick gel extraction kit (Qiagen, Hilden, Germany). Both strands of the PCR products were sequenced twice with an ABI Prism 3700 DNA analyzer (Applied Biosystems, Foster City, CA, USA), using the two PCR primers. The sequences of the PCR products were compared with known sequences of the RdRp genes of CoVs in the GenBank database.

Viral culture.

The three rodent samples positive for ChRCoV HKU24 by RT-PCR were subject to virus isolation in Huh-7.5 (human hepatoma), Vero E6 (African green monkey kidney), HRT-18G (human rectum epithelial), BSC-1 (African green monkey renal epithelial), RK13 (rabbit kidney), MDBK (bovine kidney), NIH 3T3 (mouse embryonic fibroblast), J774 (mouse macrophage), BHK-21 (baby hamster kidney), RK3E (rat kidney), RMC (rat kidney mesangial), RAW 264.7 (mouse macrophage), and primary SD rat lung cells as described previously (48, 49).

Real-time RT-PCR quantitation.

Real-time RT-PCR was performed on rodent samples positive for ChRCoV HKU24 by RT-PCR using previously described procedures (14). Reverse transcription was performed using the SuperScript III kit with random primers (Invitrogen, San Diego, CA, USA). cDNA was amplified in a LightCycler instrument with a FastStart DNA Master SYBR green I mix reagent kit (Roche Diagnostics GmbH, Mannheim, Germany) using specific primers (5′-ACAGGTTCTCCCTTTATAGATGAT-3′ and 5′-TCTCCTGTATAGTAGCAGAAGCAT-3′) targeting the RdRp gene of ChRCoV HKU24 using procedures described previously (14, 50). For quantitation, a reference standard was prepared using the pCRII-TOPO vector (Invitrogen, San Diego, CA, USA) containing the target sequence. Tenfold dilutions equivalent to 3.77 to 3.77 × 10⁹ copies per reaction were prepared to generate concomitant calibration curves. At the end of the assay, PCR products (a 133-bp fragment of RdRp) were subjected to melting curve analysis (65 to 95°C, 0.1°C/s) to confirm the specificity of the assay. The detection limit of this assay was 3.77 copies per reaction.

Complete genome sequencing.

Three complete genomes of ChRCoV HKU24 were amplified and sequenced using the RNA extracted from the original alimentary tract samples as the templates. The RNA was converted to cDNA by a combined random priming and oligo(dT) priming strategy. The cDNA was amplified by degenerate primers designed by multiple-sequence alignments of the genomes of other CoVs for which complete genomes are available, using strategies described in our previous publications (14, 20, 35, 49) and the CoVDB CoV database (51) for sequence retrieval. Additional primers were designed from the results of the first and subsequent rounds of sequencing. These primer sequences are available on request. The 5′ ends of the viral genomes were confirmed by rapid amplification of cDNA ends (RACE) using a 5′/3′ RACE kit (Roche Diagnostics GmbH, Mannheim, Germany). Sequences were assembled and manually edited to produce the final sequences of the viral genomes.

Genome analysis.

The nucleotide sequences of the genomes and the deduced amino acid sequences of the open reading frames (ORFs) were compared to those of other CoVs for which complete genomes are available using the sequences in CoVDB (51). Phylogenetic tree construction was performed using the maximum likelihood method and PhyML software, with bootstrap values being calculated from 100 trees. Protein family analysis was performed using the PFAM and InterProScan tools (52, 53). Prediction of transmembrane domains was performed using the TMHMM server (54). The structure of the ChRCoV HKU24 N-terminal domain (NTD) was predicted using a web-based homology-modeling server, SWISS-MODEL. A BLASTp search against the sequences in the Protein Data Bank (PDB) was performed with the default parameters to find suitable templates for homology modeling. On the basis of the higher sequence identity, the QMEAN Z-score, coverage, and lower E value, the crystal structure of the BCoV NTD (PDB accession number 4H14) was selected as the template. The predicted structure was visualized using the Jmol viewer.

Estimation of divergence dates.

The divergence time was calculated on the basis of complete RdRp and HE gene sequence data using a Bayesian Markov chain Monte Carlo (MCMC) approach implemented in the BEAST (version 1.8.0) package, as described previously (49, 55, 56). One parametric model (Constant Size) and one nonparametric model (Bayesian Skyline) tree priors were used for inference. Analyses were performed under the SRD06 model and by the use of both a strict and a relaxed molecular clock. The MCMC run was 2 × 10⁸ steps long with sampling every 1,000 steps. Convergence was assessed on the basis of the effective sampling size after a 10% burn-in using Tracer (version 1.5) software (55). The mean time of the most recent common ancestor (tMRCA) and the highest posterior density regions at 95% (HPDs) were calculated, and the best-fitting models were selected by use of a Bayes factor and marginal likelihoods implemented in Tracer (56). Bayesian skyline under a relaxed-clock model with an uncorrelated exponential distribution was adopted for making inferences, as Bayes factor analysis for the RdRp and hemagglutinin-esterase (HE) genes indicated that this model fitted the data better than the other models tested. The tree was summarized in a target tree by the Tree Annotator program included in the BEAST package by choosing the tree with the maximum sum of posterior probabilities (maximum clade credibility) after a 10% burn-in.

Cloning and purification of His₆-tagged recombinant ChRCoV HKU24 nucleocapsid protein and spike polypeptide.

To produce fusion plasmids for protein purification, primers 5′-CTAGCTAGCATGTCTCATACGCCA-3′ and 5′-CTAGCTAGCTTATATTTCTGAGCTTCCC-3′ and primers 5′-CTAGCTAGCCAACCAATAGCAGATGTGTA-3′ and 5′-CTAGCTAGCTTATCTCTTGGCTCGCCATGT-3′ were used to amplify the nucleocapsid gene and a partial S1 fragment encoding amino acid residues 317 to 763 of the spike protein of ChRCoV HKU24, respectively, as described previously (31, 49, 57, 58). The sequences, coding for a total of 443 amino acid (aa) and 447 aa residues, respectively, were amplified and cloned into the NheI site of expression vector pET-28b(+) (Merck, KGaA, Darmstadt, Germany) in frame and downstream of the series of six histidine residues. The His₆-tagged recombinant nucleocapsid protein and the spike polypeptide were expressed and purified using Ni-nitrilotriacetic acid affinity chromatography (Qiagen, Hilden, Germany) according to the manufacturer's instructions.

Western blot analysis.

To detect the presence of antibodies against the ChRCoV HKU24 N protein and spike polypeptide in rodent sera and to test for possible cross antigenicity between ChRCoV HKU24 and other βCoVs, 600 ng of purified His₆-tagged recombinant N protein or spike polypeptide of ChRCoV HKU24 was loaded into the well of a sodium dodecyl sulfate (SDS)–10% polyacrylamide gel and subsequently electroblotted onto a nitrocellulose membrane (Bio-Rad, Hercules, CA, USA). The blot was cut into strips, and the strips were incubated separately with 1:2,000, 1:4,000, or 1:8,000 dilutions of sera collected from rodents for which serum samples were available, human sera from two patients with HCoV OC43 infection, sera from two rabbits with rabbit coronavirus (RbCoV) HKU14 infection, and sera from two patients with SARS-CoV infection. The antigen-antibody interaction was detected with 1:4,000 horseradish peroxidase-conjugated anti-rat IgG, anti-human IgG, or anti-rabbit IgG (Zymed) and an ECL fluorescence system (GE Healthcare Life Sciences, Little Chalfont, United Kingdom) as described previously (14, 58).

Nucleotide sequence accession numbers.

The nucleotide sequences of the three genomes of ChRCoV HKU24 have been lodged within the GenBank sequence database under accession no. KM349742 to KM349744.

Identification of a novel CoV from Norway rats in China.

Of 91 alimentary tract samples from rodents in China, RT-PCR for a 440-bp fragment in the RdRp gene of CoVs was positive for a potentially novel CoV in 3 samples from Norway rats (Rattus norvegicus) from a restaurant in Guangzhou (Table 1). None of the 573 alimentary tract samples from rodents in Hong Kong, including those from Norway rats, was positive for CoVs. Sequencing results suggested that the potentially novel virus was most closely related to MHV with ≤85% nucleotide sequence identities and members of the species Betacoronavirus 1, including HCoV OC43, BCoV, equine coronavirus (ECoV) and porcine hemagglutinating encephalomyelitis virus, with ≤84% nucleotide sequence identities. Quantitative RT-PCR showed that the viral load in the positive samples ranged from 1.2 × 10³ to 1.3 × 10⁶ copies/g. Attempts to stably passage ChRCoV HKU24 in cell cultures were unsuccessful, with no cytopathic effect or viral replication being detected.

Genome organization and coding potential of ChRCoV HKU24.

Complete genome sequence data for the three strains of ChRCoV HKU24 were obtained by assembly of the sequences of the RT-PCR products from the RNA directly extracted from the corresponding individual specimens. The three genomes shared >99% nucleotide sequence similarity. Their genome size was 31,234 bases, with the G+C content (40%) being closer to that of murine coronavirus than to that of Betacoronavirus 1 (Table 2). The genome organization was similar to that of other lineage A βCoVs and had the characteristic gene order 5′-replicase ORF1ab, hemagglutinin-esterase (HE), spike (S), envelope (E), membrane (M), nucleocapsid (N)-3′ (Table 2 and Fig. 1). Moreover, additional ORFs coding for nonstructural (NS) proteins NS2a, NS4, NS5, and N2 were found. A putative transcription regulatory sequence (TRS) motif, 5′-CUAAAC-3′, similar to that of αCoVs and the motif 5′-UCUAAAC-3′ in other lineage A βCoVs, was identified at the 3′ end of the leader sequence and preceded each ORF except the ORFs for the NS4, E, and N2 genes (Table 3) (26, 49, 59–61). However, there were base mismatches for HE and NS5, with alternative TRS motifs, 5′-CUGAAC-3′ and 5′-GUAAAC-3′, respectively, being detected.

The coding potential and characteristics of putative nonstructural proteins (nsp's) of ORF1 of ChRCoV HKU24 are shown in Tables 3 and 4. The ORF1 polyprotein possessed 68.6 to 75.0% amino acid sequence identities to the amino acid sequences of the polyproteins of other lineage A βCoVs. It possessed a unique putative cleavage site, G/L, between nsp1 and nsp2, in contrast to the G/V found in the other lineage A βCoVs except HCoV HKU1, which possessed a G/I cleavage site (Table 4 and Fig. 1). Other predicted cleavage sites were mostly conserved between ChRCoV HKU24 and other lineage A βCoVs. However, the lengths of nsp1, nsp2, nsp3, nsp13, nsp15, and nsp16 in ChRCoV HKU24 differed from those of the corresponding nsp's in members of Betacoronavirus 1 and murine coronavirus, as a result of deletions or insertions.

All lineage A βCoVs except HCoV HKU1 possess the NS2a gene between ORF1ab and the HE gene. Unlike RbCoV HKU14, in which the gene for NS2a is broken into several small ORFs (49), ChRCoV HKU24 is predicted to possess a single NS2a protein, as in other lineage A βCoVs. This NS2a protein displayed from 43.7 to 62.0% amino acid sequence identities to the NS2a amino acid sequences of members of Betacoronavirus 1 and 45.7 to 47.3% amino acid sequence identities to the NS2a amino acid sequences of members of murine coronavirus. Although the βCoV-specific NS2 protein has been shown to be nonessential for in vitro viral replication (62), cyclic phosphodiesterase domains have been predicted in the NS2 proteins of some CoVs and toroviruses, and a possible role in viral pathogenicity has been suggested in MHV (63, 64). In contrast to MHV and RCoV, such a domain was not found in ChRCoV HKU24.

Similar to other CoV S proteins, the S protein of ChRCoV HKU24 is predicted to be a type I membrane glycoprotein, with most of the protein (residues 16 to 1302) being exposed on the outside of the virus and with a transmembrane domain (residues 1303 to 1325) occurring at the C terminus (Fig. 2). Two heptad repeats (HRs), important for membrane fusion and viral entry, were located at residues 1045 to 1079 (HR1) and 1253 to 1285 (HR2). The S protein of ChRCoV HKU24 possessed 66.7 to 69.6% amino acid sequence identities to the amino acid sequences of members of Betacoronavirus 1 and 62.4 to 64.3% amino acid sequence identities to the amino acid sequences of members of murine coronavirus. The amino acid sequence identities between the ChRCoV HKU24 NTD and the BCoV and MHV NTDs were 61 and 56%, respectively. BCoV and HCoV OC43 utilize N-acetyl-9-O-acetylneuramic acid as a receptor for the initiation of infection (65, 66). In contrast, MHV utilizes carcinoembryonic antigen-related cell adhesion molecule 1 (CEACAM1) as a receptor, and its receptor-binding domain does not bind sugars (10, 67, 68). Recent structural studies showed that, among the four critical sugar-binding residues in BoV, a Glu → Gly substitution which may explain the reduction in sugar-binding affinity was found in one residue in MHV. In ChRCoV HKU24, a Glu → Ser substitution is found at this position (Fig. 2). Comparison of the amino acid sequences between the S proteins of ChRCoV HKU24 and MHV showed that ChRCoV HKU24 possessed many amino acid substitutions in the region corresponding to the MHV NTD (Fig. 2). In particular, 12 of the 14 important contact residues at the MHV NTD/murine CEACAM1a (mCEACAM1a) interface were not conserved between ChRCoV HKU24 and MHV. Similar to the MHV and BCoV NTDs, the ChRCoV HKU24 NTD is also predicted, using homology modeling, to contain a core structure with a β-sandwich fold like that in human galectins (galactose-binding lectins) (10). Modeling showed that the β-sandwich core structure of ChRCoV HKU24 consists of one six-stranded β sheet and one seven-stranded β sheet that are stacked together through hydrophobic interactions (Fig. 2). In addition, the S protein of ChRCoV HKU24 possessed a unique predicted cleavage site, RAKR, among lineage A βCoVs.

Other predicted domains in the HE, S, NS4, NS5, E, M, and N proteins of ChRCoV HKU24 are summarized in Table 3 and Fig. 1. The NS4 of ChRCoV HKU24 shared 37 to 42% amino acid sequence identity to the NS4 proteins of members of murine coronavirus. In most members of Betacoronavirus 1, NS4 is split into smaller proteins. NS5 of ChRCoV HKU24 is homologous to NS5/NS5a of members of Betacoronavirus 1, having 47.7% to 51.4% amino acid sequence identities, but it has only 39.5% amino acid sequence identity to NS5 of members of MHV. Interestingly, NS5 is not found in the genome of RCoV. The absence of a preceding TRS upstream of the E of ChRCoV HKU24 suggests that the translation of this E protein may be cap independent via an internal ribosomal entry site (IRES), as demonstrated in MHV (69). Similarly, the E proteins of RCoV and HCoV HKU1 were also not preceded by a TRS. This is in contrast to the findings for members of Betacoronavirus 1, which possess a preceding TRS upstream of their E proteins (49, 61). Downstream of the N gene, the 3′ untranslated region contains a predicted bulged stem-loop structure of 69 nucleotides (nt; nucleotide positions 30944 to 31012) that is conserved in βCoVs (70). Overlapping with the bulged stem-loop structure by 5 nt, a conserved pseudoknot structure (nucleotide positions 31008 to 31059) that is important for CoV replication is found. Since nonstructural proteins in CoVs may possess unique functions for replication and virulence (71, 72), further studies are warranted to understand the potential function of the nsp's and NS proteins in ChRCoV HKU24.

Phylogenetic analyses.

Phylogenetic trees constructed using the amino acid sequences of the RdRp, S, and N proteins of ChRCoV HKU24 and other CoVs are shown in Fig. 3, and the corresponding pairwise amino acid sequence identities are shown in Table 2. For all three genes, the three ChRCoV HKU24 strains formed a distinct cluster among lineage A βCoVs, occupying a deep branch at the root of members of the species Betacoronavirus 1 and being the most closely related to members of the species Betacoronavirus 1. Comparison of the amino acid sequences of the seven conserved replicase domains for CoV species demarcation (3), ADP-ribose 1″-phosphatase (ADRP), nsp5 (3C-like protease [3CL^pro]), nsp12 (RdRp), nsp13 (Hel), nsp14 (exonuclease [ExoN]), nsp15 (nidoviral uridylate-specific endoribonuclease [NendoU]), and nsp16 (2′-O-ribose methyltransferase [O-MT]), showed that these domains of ChRCoV HKU24 possessed 69.5 to 81.7%, 82.2 to 86.8%, 88.1 to 92.6%, 88.9 to 94.8%, 80.2 to 88.7%, 70.1 to 79.5%, and 83.8 to 89.7% amino acid identities to those of other lineage A βCoVs, respectively (Table 5). Based on the present results, we propose a novel species, ChRCoV HKU24, to describe this virus under Betacoronavirus lineage A and distinguish it from RCoV.

HE proteins are glycoproteins that mediate reversible attachment to O-acetylated sialic acids by acting as both lectins and receptor-destroying enzymes which aid viral detachment from sugars on infected cells (68, 73). Related HEs have been found in influenza C viruses, toroviruses, and lineage A βCoVs, but not other CoVs. It has been suggested that HEs of lineage A βCoVs have arisen from an influenza C virus-like HE fusion protein, likely as a result of relatively recent lateral gene transfer events (73). Phylogenetic analysis of the HE proteins of lineage A βCoVs, toroviruses, and influenza C viruses showed that they fell into three separate clusters (Fig. 3). The HE of ChRCoV HKU24 also forms a deep branch at the root of all members of the species Betacoronavirus 1 except ECoV and is distinct from members of murine coronavirus. Previous studies have demonstrated the heterogeneity of gene expression of HE proteins among different MHV strains (74). Since the HE of ChRCoV HKU24 is not preceded by a perfectly matched TRS, further studies are required to determine if it is expressed and functional.

Estimation of divergence dates.

Using the uncorrelated relaxed-clock model on complete RdRp gene sequences, the date of tMRCA of ChRCoV HKU24, members of Betacoronavirus 1, and RbCoV HKU14 was estimated to be 1402 (HPDs, 918.05 to 1749.91) (Fig. 4). The date of divergence between HCoV OC43 and BCoV was estimated to be 1897 (HPDs, 1826.15 to 1950.05), consistent with results from previous molecular clock studies (27). Using the uncorrelated relaxed-clock model on complete HE gene sequences, the date of tMRCA of ChRCoV HKU24, members of Betacoronavirus 1, and RbCoV HKU14 was estimated to be 1337 (HPDs, 724.59 to 1776.78) (Fig. 4). The date of divergence between HCoV OC43 and BCoV was estimated to be 1871 (HPDs, 1764.55 to 1944.37). The estimated mean substitution rates of the RdRp and HE data sets were 1.877 × 10⁻⁴ and 4.016 × 10⁻⁴ substitution per site per year, respectively, which are comparable to previous estimates for other lineage A βCoVs (26, 27, 39).

Serological studies.

Western blot analysis using recombinant ChRCoV HKU24 N protein was performed using sera from 144 rodents for which serum samples were available, sera from two patients with HCoV OC43 infection, sera from two rabbits with RbCoV HKU14 infection, and sera from two patients with SARS-CoV infection. Among the sera tested from 74 Norway rats from Guangzhou for which serum samples were available, 60 (81.1%) were positive for antibody against recombinant ChRCoV HKU24 N protein with prominent immunoreactive bands of about 50 kDa (Table 1 and Fig. 5). These 60 positive samples included the 3 serum samples collected from the three Norway rats positive for ChRCoV HKU24 in their alimentary tract samples. In addition, 15 (48.4%) of 31 Norway rats from Hong Kong were also positive for antibody against recombinant ChRCoV HKU24 N protein, although the virus was not detected in alimentary tract samples from these rats. Moreover, 7 (77.8%) of 9 oriental house rats but only 4 (0.13%) of 30 black rats were positive for antibody against recombinant ChRCoV HKU24 N protein. Possible cross antigenicity between ChRCoV HKU24 and other βCoVs, including lineage A and B βCoVs, was found. Sera from two patients with HCoV OC43 infection, sera from two rabbits with RbCoV HKU14 infection, and sera from two patients with SARS-CoV infection were also positive for antibody against recombinant ChRCoV HKU24 N protein by Western blot assay (Fig. 5).

Western blot analysis using recombinant ChRCoV HKU24 spike polypeptide was performed to verify the specificity of antibodies against ChRCoV HKU24 N protein using positive rodent sera and sera from two patients with HCoV OC43 infection, sera from two rabbits with RbCoV HKU14 infection, and sera from two patients with SARS-CoV infection. Among the sera from the 60 Norway rats positive for antibodies against the ChRCoV HKU24 N protein, 21 were positive for antibodies against the ChRCoV HKU24 spike polypeptide with prominent immunoreactive bands of about 50 kDa (Table 1 and Fig. 5). However, serum samples from the three Norway rats positive for ChRCoV HKU24 in their alimentary tract samples were negative for anti-ChRCoV HKU24 spike polypeptide antibody. Of the seven oriental house rats positive for antibodies against the ChRCoV HKU24 N protein, two were positive for antibodies against the ChRCoV HKU24 spike polypeptide. However, serum samples from the 4 black rats and 15 Norway rats from Hong Kong positive for antibodies against the ChRCoV HKU24 N protein were negative for antibodies against the ChRCoV HKU24 spike polypeptide. In contrast to N protein, no cross antigenicity between the ChRCoV HKU24 spike polypeptide and sera positive for antibodies against other βCoVs, including lineage A and B βCoVs, was detected. Sera from two patients with HCoV OC43 infection, sera from two rabbits with RbCoV HKU14 infection, and sera from two patients with SARS-CoV infection were all negative for antibody against recombinant ChRCoV HKU24 spike polypeptide by Western blot assay (Fig. 5).