Identification of a novel lineage bat SARS-related coronaviruses that use bat ACE2 receptor

Abstract

Severe respiratory disease coronavirus-2 (SARS-CoV-2) has been the most devastating disease COVID-19 in the century. One of the unsolved scientific questions of SARS-CoV-2 is the animal origin of this virus. Bats and pangolins are recognized as the most probable reservoir hosts that harbour highly similar SARS-CoV-2 related viruses (SARSr-CoV-2). This study identified a novel lineage of SARSr-CoVs, including RaTG15 and seven other viruses, from bats at the same location where we found RaTG13 in 2015. Although RaTG15 and the related viruses share 97.2% amino acid sequence identities with SARS-CoV-2 in the conserved ORF1b region, it only shows less than 77.6% nucleotide identity to all known SARSr-CoVs at the genome level, thus forming a distinct lineage in the Sarbecovirus phylogenetic tree. We found that the RaTG15 receptor-binding domain (RBD) can bind to ACE2 from Rhinolophus affinis, Malayan pangolin, and use it as an entry receptor, except for ACE2 from humans. However, it contains a short deletion and has different key residues responsible for ACE2 binding. In addition, we showed that none of the known viruses in bat SARSr-CoV-2 lineage discovered uses human ACE2 as efficiently as the pangolin-derived SARSr-CoV-2 or some viruses in the SARSr-CoV-1 lineage. Therefore, further systematic and longitudinal studies in bats are needed to prevent future spillover events caused by SARSr-CoVs or to understand the origin of SARS-CoV-2 better.

Keywords: ACE2; SARS-related coronavirus; bat; novel lineage; reservoir host.

Figure 1.

Discovery of a novel lineage of bat SARSr-CoVs. (A) Similarity plot analysis based on the full-length genome sequence of bat SARSr-CoV RaTG15. Full-length genome sequences of SARS-CoV-1, SARS-CoV-2, bat SARSr-CoVs, and pangolin CoVs related to SARS-CoV-2 were used as reference sequences. The analysis was performed with the Kimura model, a window size of 1500 base pairs and a step size of 150 base pairs. (B) Phylogenetic tree based on complete genome sequences of betacoronaviruses. The trees were constructed by the Maximum-Likelihood method using the Jukes-Cantor model with bootstrap values determined by 1000 replicates. Bootstraps > 50% are shown. The scale bars represent 0.1 substitutions per nucleotide position. The novel SARSr-CoVs characterized in this study are shown in bold. Ra, Rhinolophus affinis; Rst, Rhinolophus stheno; Rsh, Rhinolophus shameli; Rs, Rhinolophus sinicus; Rac, Rhinolophus acuminatus; Rm, Rhinolophus malayanus; Rc, Rhinolophus cornutus; MHV, murine hepatitis virus.

Figure 2.

Comparison of the receptor-binding domain (RBDs) of SARSr-CoVs. The RBM is shown in pink and the five key residues that contact ACE2 directly are highlighted in green. A comparison of the five critical residues of these SARSr-CoVs is listed in the table. Two deletions in the RBM, aa 444–447 (deletion 1) and aa 473–486 (deletion 2) are indicated by red boxes. GenBank or GISAID entries for each virus can be found in Methods.

Figure 3.

Binding affinity of SARSr-CoV RBDs to ACE2 from human, R.affinis and pangolin. (A-F) Binding of different RBD proteins to human ACE2. (G-L) Binding of different RBD proteins to R.affinis ACE2. (M-R) Binding of different RBD proteins to Malayan pangolin ACE2. (S) Comparison of dissociation constants (KD) between different RBD to human, R.affinis, and Malayan pangolin ACE2. Relative binding is analyzed by comparing with SARS-CoV-2 RBD to human ACE2. (T) Summary of the binding efficiency of different RBD to human, bat or pangolin ACE2. Y, yes; ND, not determined. Evidences for WIV16-CoV, Rc-o0319, RmYN02, and RacCS213 were from previous reports [10,11,22]. The presence of deletion in RBM (related to Figure 2) is indicated. Binding assay of human, R.affinis or pangolin ACE2 to different RBD proteins was measured by Bio-layer interferometry. The parameters of KD value (M), Kon (1/M.s), Koff (1/s) are shown on the upper right side of the picture. Different RBD proteins were immobilized on the sensors and tested for affinity with graded concentrations of human, R. affinis, or pangolin ACE2s. The Y-axis shows the real-time binding response. Values reported representing the global fit to all data. The coefficient of determination (R^2) for these interactions was close to 1.0 (Figure S2B).