Discovery and genetic analysis of novel coronaviruses in least horseshoe bats in southwestern China

Abstract

To investigate bat coronaviruses (CoVs), we collected 132 rectal swabs and urine samples from five bat species in three countries in southwestern China. Seven CoVs belonging to distinct groups of severe acute respiratory syndrome (SARS)-like CoVs and α-CoVs were detected in samples from least horseshoe bats. Samples from other bat species were negative for these viruses, indicating that the least horseshoe bat represents one of the natural reservoirs and mixers for strains of CoVs and has a pivotal role in the evolution and dissemination of these viruses. The genetic and evolutionary characteristics of these strains were described. Whole-genome sequencing of a new isolate (F46) from a rectal swab from a least horseshoe bat showed that it contained 29 699 nucleotides, excluding the poly (A) tail, with 13 open reading frames (ORFs). Phylogenetic and recombination analyses of F46 provided evidence of natural recombination between bat SARS-like CoVs (Rs3367 and LYRa11) or SARS-CoV (BJ01), suggesting that F46 could be a new recombinant virus from SARS-like CoVs or SARS-CoVs.

Figure 1

Phylogenetic tree of sequences included in this study and representative CoVs based on partial RdRP gene sequence (405 bp). Analyses were conducted using MEGA v.6 with partial RdRP sequences (405 nt) of each isolate taken from GenBank (corresponding to 15 234–15 638 nt of the genome of bat strain Rp3/2004 (DQ071615) using the neighbor-joining algorithm and bootstrap test for inference of phylogeny. Bootstrap values were obtained from 1000 data sets, and only nodes with bootstrap values greater than 70% are presented. Three established CoV genera, Alphacoronavirus, Betacoronavirus, and Gammacoronavirus are marked as α, β and γ, respectively. Four CoV groups in the genus Betacoronavirus are indicated as A, B, C and D. Known human CoVs are in italics. Sequences reported in this study are labeled with triangles and highlighted in bold. Coronavirus, CoV; nucleotide, nt.

Figure 2

Phylogenetic tree based on the whole-genome sequences and RBD region of F46 and representative coronaviruses. (A) Phylogenetic tree based on SARS-CoV S protein amino-acid residues 310–520. Analysis was conducted using MEGA v.6. with a Poisson model. Bootstrap values are determined by 1000 replicates. The key amino-acid residues involved in the interaction with the human ACE2 molecule of identified strains are indicated on the right of the tree. (B) Phylogenetic tree based on the whole genome. Analyses were conducted using MEGA v.6 with the neighbor-joining algorithm and bootstrap test for inference of phylogeny. Bootstrap values were obtained from 1000 data sets, and only nodes with bootstrap values greater than 70% are presented. Three established CoV genera Alphacoronavirus, Betacoronavirus and Gammacoronavirus are marked as α, β, and γ, respectively. Four CoV groups in the genus Betacoronavirus are indicated as A, B, C and D. Angiotensin-converting enzyme 2, ACE2; coronavirus, CoV; receptor-binding domain, RBD; severe acute respiratory syndrome, SARS.

Figure 3

Sequence alignment of CoV S protein RBD. The RBD of S protein from F46, Human/Civet CoVs and other closely related bat SARS-like CoVs were aligned using ClustalW 2.0. The key amino-acid residues involved in the interaction with human ACE2 are indicated by arrows. Angiotensin-converting enzyme 2, ACE2; coronavirus, CoV; open reading frame, ORF; receptor-binding domain, RBD; severe acute respiratory syndrome, SARS.

Figure 4

Recombination analysis of F46 with SARS-CoVs and SL-CoVs. (A) A gene map of LYRa11 is used to position breakpoints. Four breakpoints were detected in the F46 genome with strong P-values (<10^–30) supported by similarity plot and bootscan analyses (B, C). The breakpoints generated recombination of fragment A covering 1–8 and 848 nts (a total of 8 and 848 nt), including the 5′-UTR and partial ORF 1a, and fragment B covering 21 684 to 23 271 nt (a total of 1587 nt), including a partial ORF of the spike gene. Similarity plots (B) and bootscan analyses (C) were conducted with F46 as the query and bat SARS-like CoVs (Rs3367 and LYRa11) and SARS-CoV BJ01 as potential parental sequences. Phylogenetic trees (D, E) were constructed based on the two fragments with a neighbor-joining algorithm and a bootstrap test for inference of phylogeny. Bootstrap values were obtained from 1000 data sets. Coronavirus, CoV; nucleotide, nt; open reading frame, ORF; severe acute respiratory syndrome, SARS; untranslated region, UTR.