key: cord-0799303-2v6btgbq
authors: Yan, Huan; Jiao, Hengwu; Liu, Qianyun; Zhang, Zhen; Wang, Xin; Guo, Ming; Wang, Bing-Jun; Lan, Ke; Chen, Yu; Zhao, Huabin
title: Many bat species are not potential hosts of SARS-CoV and SARS-CoV-2: Evidence from ACE2 receptor usage
date: 2020-09-10
journal: bioRxiv
DOI: 10.1101/2020.09.08.284737
sha: 8eb115c0c095b8d3b6f21d960342a0df0fa3eb82
doc_id: 799303
cord_uid: 2v6btgbq

Bats are the suggested natural hosts for severe acute respiratory syndrome coronavirus (SARS-CoV) and SARS-CoV-2, the latter of which caused the coronavirus disease 2019 (COVID-19) pandemic. The interaction of viral Spike proteins with their host receptor angiotensin-converting enzyme 2 (ACE2) is a critical determinant of potential hosts and cross-species transmission. Here we use virus-host receptor binding and infection assays to show that ACE2 orthologs from 24, 21, and 16 of 46 phylogenetically diverse bat species – including those in close and distant contact with humans – do not support entry of SARS-CoV, SARS-CoV-2, and both of these coronaviruses, respectively. Furthermore, we used genetic and functional analyses to identify genetic changes in bat ACE2 receptors associated with viral entry restrictions. Our study demonstrates that many – if not most – bat species are not potential hosts of SARS-CoV and SARS-CoV-2, and provides important insights into pandemic control and wildlife conservation.

The unprecedented pandemic of COVID-19, caused by the novel coronavirus SARS-CoV-2, 39 has led to major threats to public health and economic development. It is therefore critically 40 important to identify natural or intermediate hosts of SARS-CoV-2 to prevent further spread of 41 COVID-19 and future emergence of similar diseases. Inferred from sequence similarity of 42 human and bat virus genomes, it was suggested that horseshoe bats (Rhinolophus spp.) might be 43 natural hosts of SARS-CoV and SARS-CoV-2 (1-3). These suggestions have resulted in 44 misguided fears on all bats, and unwarranted attacks on many batsincluding species other than 45 Rhinolophusthereby seriously impacting efforts towards bat conservation (4). Given the 46 remarkable diversity of bats, which includes more than 1400 species across the globe (5), 47 assessing the possibility that diverse bat species act as potential hosts of SARS-CoV and SARS- 48 CoV-2 is urgent and crucial for both controlling outbreaks and protecting populations of wildlife. 49 ACE2 is the host cell receptor of SARS-CoV and SARS-CoV-2, and plays a vital role in 50 mediating viral entry to cause infection (1, 6). The interaction of a virus with its host receptor has 51 been repeatedly demonstrated to serve as a primary determinant of host range (7). Here we test 52 ACE2 orthologs from 46 bat species across the phylogeny, including species occurring in urban 53 and in rural areas, for their ability to support the entry of SARS-CoV and SARS-CoV-2. Hence, 54 this study assesses whether diverse bat species are potential hosts of SARS-CoV or SARS-CoV-55 2. Moreover, by determining the correlation between proximity to humans and probability of 56 being natural hosts of the two viruses, these results provide important insights into pandemic 57 control and wildlife conservation. 

Evolution of ACE2 in bats inhabiting urban or rural areas 61 We collected ACE2 orthologs from 46 bat species across the phylogeny (Figure 1 and   62 Table S1). These species contained 28 species that roost or forage in urban areas in close 63 proximity to humans, and 18 species more restricted to rural areas and hence likely to have 64 minimal contact with humans (Table S2 ). In total, the examined species represent 11 bat families 65 that contain 1345 species, accounting for 96% of all bat species (Table S3) . After aligning the 66 protein sequences of bat ACE2 orthologs, we examined 25 critical residues involved in the 67 binding of the surface spike glycoprotein (S protein) of SARS-CoV-2 ( Figure S1 ) (8) . Genetic 68 variations were observed in nearly all these 25 sites, which may have led to different abilities to 69 support entry of SARS-CoV and SARS-CoV-2 (8) . Furthermore, we detected at least 22 amino 70 acid sites that are putatively under positive selection (Table S4) , indicative of heterogeneous 71 selection pressure across sites. Notably, four of these positively selected sites are located in the 72 binding region of ACE2 to the SARS-CoV-2 S protein (Table S4) .

Interaction between bat ACE2 orthologs and SARS-CoV or SARS-CoV-2 receptor binding 75 domain (RBD) 76 Efficient binding between the S protein and the ACE2 receptor is important for SARS-CoV 77 and SARS-CoV-2 entry. This binding is mainly mediated by the interaction between the critical 78 residues on the RBD and ACE2. To characterize the receptor function of ACE2 orthologs in a 79 range of diverse bat species, we generated a stable cell library consisting of cell lines expressing 80 3 the respective 46 bat ACE2 orthologs through lentiviral transduction of 293T cells lacking ACE2 81 expression (9). All bat ACE2 orthologs were exogenously expressed at a comparable level after 82 puromycin selection, as indicated by Western-blot and immunofluorescence assays detecting the 83 C-terminal 3×Flag tag (Figure 2A-B) . 84 To analyze the interaction, we produced recombinant SARS or SARS-CoV-2 RBD human 85 IgG Fc fusion proteins (RBD-hFC), previously reported to be sufficient to bind human ACE2 86 efficiently (10, 11). The protein binding efficiency was tested on the bat ACE2 cell library 87 through immunofluorescence or flow cytometry targeting the human Fc. As expected, binding 88 was almost undetectable on mock 293T cells, but a strong binding signal was detected on the 89 293T cells expressing human ACE2 ( Figure 2C-D) . Consistent with previous reports (12, 13), 90 SARS-CoV-2 RBD showed higher binding to hACE2 than SARS-CoV, which can also be 91 observed on many bat ACE2 orthologs ( Figure 2C-D) . Previous reports have shown that only a 92 small fraction of ACE2 orthologs from tested mammalian species could not bind with SARS- Table S5 ). Pseudotypes with GFP reporter 117 showed similar results ( Figure S4) . Notably, we found that 24, 21, and 16 of the 46 bat species 118 showed almost no entry for SARS-CoV, SARS-CoV-2, and both of these viruses, respectively 119 (Figures 1 and 3A-B, Table S5 ), suggesting that these species are not likely to be potential hosts 120 of either or both of these coronaviruses. The bat species showing no viral entry include those that 121 occur in urban areas as well as those more restricted to rural areas (Figure 1, Table S1 ), 122 suggesting that there is no correlation between proximity to humans and probability of being 123 natural hosts of SARS-CoV or SARS-CoV-2. Although horseshoe bats were suggested to be 124 4 potential natural hosts of SARS-CoV and SARS-CoV-2 (1-3), only one of the three examined 125 species (Rhinolophus sinicus) supported SARS-CoV entry; this species was suggested to be the 126 potential host of SARS-CoV (3, 15). None of these tested horseshoe bats showed entry for 127 SARS-CoV-2 (Figures 1 and 3) . These results unambiguously indicate that ACE2 receptor 128 usage is species-dependent. 129 The SARS-CoV-2 S protein used here for pseudotyping contains a D614G mutation, which 130 is currently a dominant variation (16). The D614G mutation remarkably improved the in vitro 131 infectivity of SARS-CoV-2, but may not significantly affect the receptor interaction since it is 132 not in the RBD (17). Indeed, we identified a very similar susceptibility profile using an original 133 strain without D614G ( Figure S5) . We further demonstrated that the pseudotyped entry assay 134 mimics the entry of live viruses through a SARS-CoV-2 infection assay ( Figure 3C ). As 135 expected, the profile of SARS-CoV-2 N protein expression is highly consistent with the results 136 from the VSV-dG-based pseudotyped virus entry assay ( Figure 3C) . However, the live virus 137 infection resulted in the phenotype of plaque formation, while the pseudotypes showed evenly 138 distributed single-round infection ( Figure S4 ). 139 When comparing the RBD-hFC binding and pseudotype entry profiles, we found that 140 binding and susceptibility are generally consistent, with a few exceptions. For instance, some 141 species (Bat12, 13, 14) were able to bind to SARS-CoV-2 RBD-hFc efficiently, but cannot 142 support infection of the same virus, indicating that high binding affinity does not guarantee 143 efficient viral entry (Figures 2 and 3) . In contrast, some species (Bat3-8) were defective or less 144 efficient in SARS-CoV RBD-hFc binding, but supported the entry of the same virus to some 145 degree (Figures 2 and 3) . We hypothesize that such minimal binding may be sufficient for viral 146 entry mediated by those ACE2 orthologs; alternatively, additional residues outside the traditional 147 RBD region might be required for efficient interaction. These hypotheses should be tested in the to a lesser degree, SARS-RBD binding (Figures 2 and 3) . We compared their protein sequences 167 and highlighted the residues that may affect RBD interaction. For example, substitutions I27K, 168 5 N31G, and K42E were observed when comparing Bat33 and 34, while Q24L, E30K, K35Q, and 169 G354N were present between Bat38 and 40 ( Figure 4A ). We hypothesized that the discrepancy 170 in binding and infection phenotype is determined by their differences in critical residues for RBD 171 interaction. To test this hypothesis, we designed a residue swap mutagenesis assay to investigate 172 the role of critical residues on RBD binding and virus entry ( Figure 4A ). We generated four 173 swap mutations and corresponding 293T stable cell lines to test whether these substitutions can 174 achieve the gain-of-function and loss-of-function. All bat ACE2 orthologs and related mutants 175 were expressed at a comparable level after lentiviral transduction, as indicated by the 176 immunofluorescence of the carboxyl-terminal (C-terminal) 3×Flag tag ( Figure 4B ). CoV-2. Our sampling covers representative species from 11 bat families, accounting for 96% of 194 all extant bat species, hence providing a broad picture. Moreover, our study included 28 species 195 inhabiting urban areas and 18 species that are not common in cities or do not roost in buildings. 196 Our functional assays demonstrated that there is no correlation between proximity to humans and 197 probability of being natural hosts of SARS-CoV or SARS-CoV-2. Therefore, there is no need to 198 fear the many bat species occurring in cities that are not potential hosts of SARS-CoV and 199 SARS-CoV-2. Species such as horseshoe bats, which are suggested to be potential natural hosts 200 of the two viruses, should also not be feared, as they are less likely to be found in cities. 201 Our results are only partially consistent with a recently published prediction based on 202 sequence similarity, which estimated a binding score between ACE2 and the SARS-CoV-2 S 203 protein for each vertebrate species examined (8) . The predicted binding scores for all 37 bat 204 species fell into low (n=8) and very low (n=29) categories (8) , suggesting that all examined bat 205 species are at low risk for SARS-CoV-2 infection. Our study included 36 of the 37 previously 206 examined bat species (Figure 1 and Table S1 ); 19 of these appeared to support SARS-CoV-2 207 entry by their ACE2 receptors (Figures 1 and 3) , strongly suggesting that these bats are at high Figure 4) . However, these residues are not the genetic 213 6 determinant of New World monkey ACE2 orthologs mediating SARS-CoV-2 entry (7), and 214 many bat ACE2 orthologs carrying residues that were considered unfavorable in the same study 215 (H41 and E42) (7) were fully functional in our study (Figure 4) , further confirming the 216 complexity of ACE2 functionality. 217 We found that closely related species can show strikingly different ACE2 receptor usage. 218 For example, Rhinolophus sinicus can support SARS-CoV entry, whereas its congeneric 219 relatives R. ferrumequinum and R. pearsonii cannot (Figures 1 and 3) , despite the fact that some Our study supports the calls that public education on bat biology will reduce the threat to 236 bats (4, 22). In fact, all bats are potentially safe as long as they are treated with care and respect. 237 We should work collaboratively to combat the pandemic and identify which species are potential 238 hosts, and not fear those species that are not hosts of the virus. Instead, we must respect and care 239 for those species that are potential hosts, and learn about the impact of human activities on their 240 natural habitats, which may lead to zoonotic spillover events. 

ACE2 sequence acquisition and selective pressure analysis 377 We obtained 46 full-length coding sequences of bat ACE2 in this study, of which 32 were 378 taken from a recent study (8) , and 14 were newly extracted from published or recently sequenced 379 genome assemblies (see Table S1 for the sources and accession numbers for the sequences and 380 assemblies). Next, we aligned the deduced ACE2 protein sequences using the MUSCLE program 381 (23) (see Figure S1 for the resulting alignment). The sequence logo was generated with 

Severe acute respiratory syndrome coronavirus-like 303 virus in Chinese horseshoe bats

Tracking changes in SARS-CoV-2 spike: Evidence that D614G increases 309 infectivity of the COVID-19 virus

The D614G mutation of SARS-CoV-2 spike 312 protein enhances viral infectivity and decreases neutralization sensitivity to individual 313 convalescent sera. bioRxiv

Evolutionary arms 315 race between virus and host drives genetic diversity in bat SARS related coronavirus 316 spike genes

Bats and Coronaviruses. 318 Viruses

Possibility for reverse zoonotic transmission of SARS-CoV-2 to free-ranging wildlife: A 325 case study of bats

Susceptibility of ferrets, 329 cats, dogs, and other domesticated animals to SARS-coronavirus 2

 388 HEK293T cells (293T, ATCC, CRL-3216) and VERO-E6 cells (ATCC, CRL-1586) were 389 cultured in Dulbecco's modified Eagle's medium (DMEM; Gibco) supplemented with 10% fetal 390 bovine serum (FBS), 2.0 mM L-Glutamine, 110 mg/L sodium pyruvate, and 4.5 g/L D-glucose. 391 l1-Hybridoma (CRL-2700) secreting a monoclonal antibody targeting against VSV glycoprotein 392 was cultured in Minimum Essential Medium with Earle's salts and 2.0 mM L-Glutamine (MEM; 393 Gibco). All cells were cultured at 37℃ in 5% CO 2 with the regular passage of every 2-3 days. 394 293T stable cell lines overexpressing ACE2 orthologs were maintained in growth medium 395 supplemented with 1μg/ml puromycin. Immunofluorescence assay to evaluate the expression level of ACE2 orthologs 425 The expression levels of ACE2 orthologs were evaluated by the immunofluorescence assay 426 detecting the C-terminal 3×Flag tags. The cells for analysis were seeded in the poly-lysine 427 pretreated 96-well plate at a cell density of 5×10 5 /ml (100 μl/well), and cultured for 24 hrs. Cells 428 were fixed with 4% paraformaldehyde at room temperature for 10 mins, permeablized with 0.2% 429 11 Triton X-100/PBS at room temperature for 10 mins, and blocked with 1% Bovine serum albumin 430 (BSA) at 37℃ for 30 mins. Next they were incubated with the mouse monoclonal antibody