ACE2-Independent Bat Sarbecovirus Entry and Replication in Human and Bat Cells

Abstract

Hundreds of sarbecoviruses have been found in bats, but only a fraction of them have the ability to infect cells using angiotensin-converting enzyme 2 (ACE2), the receptor for SARS-CoV and -2. To date, only ACE2-dependent sarbecoviruses have been isolated from field samples or grown in the laboratory. ACE2-independent sarbecoviruses, comprising the majority of the subgenus, have not been propagated in any type of cell culture, as the factors and conditions needed for their replication are completely unknown. Given the significant zoonotic threat posed by sarbecoviruses, cell culture models and in vitro tools are urgently needed to study the rest of this subgenus. We previously showed that the exogenous protease trypsin could facilitate cell entry of viral-like particles pseudotyped with spike protein from some of the ACE2-independent sarbecoviruses. Here, we tested if these conditions were sufficient to support bona fide viral replication using recombinant bat sarbecoviruses. In the presence of trypsin, some of the spike proteins from clade 2 viruses were capable of supporting bat sarbecovirus infection and replication in human and bat cells. Protease experiments showed a specific viral dependence on high levels of trypsin, as TMPRSS2 and furin had no effect on clade 2 virus entry. These results shed light on how sarbecoviruses transmit and coexist in their natural hosts, provide key insights for future efforts to isolate and grow these viruses from field samples, and further underscore the need for broadly protective, universal coronavirus vaccines. IMPORTANCE Our studies demonstrate that some unexplored sarbecoviruses are capable of replicating in human and bat cells in an ACE2-independent way but need a high trypsin environment. We found that trypsin is not compensated by other known proteases involved in some coronavirus entry. This work provides important information that the trypsin-dependent entry may be a widely employed mechanism for coronaviruses and will help for further understanding the biological features of the less-studied viruses.

Keywords: ACE2-independent; SARS-related coronavirus; sarbecovirus; trypsin-dependent.

FIG 1

Exogenous trypsin mediates sarbecovirus RBD clade 2 replication in human and bat cells. (a) Receptor usage of the different clades of sarbecovirus RBD. (b) Outline of reverse genetics rescue strategies. (c) Viral replication on three different human cell lines was measured by RT-PCR for the RsWIV1 viral backbone (d) 293T or (e) Caco2 cells, which were infected, stained for viral nucleocapsid protein, and measured by fluorescence-activated cell sorter (FACS). (f) Quantification of FACS results. (g) Viral replication, with or without trypsin, on primary bat cell cultures. Error bars represent the corresponding mean ± SEM.

FIG 2

Sarbecovirus RBD clade 2 infection depends on high levels of trypsin. (a) Pseudotyped particles were collected in standard viral growth medium supplemented with 2% fetal bovine serum and used to infect Huh-7 cells with increasing amounts of trypsin. (b) Pseudotyped particles were collected in serum-free formulated viral growth medium and used to infect Huh-7 cells with increasing amounts of trypsin. (c) 293T cells stably expressing human ACE2 were transfected with increasing amounts of TMPRSS2 and infected with serum-free pseudotyped particles. (d) Pseudotypes were generated with a chimeric SARS-CoV-2 spike encoding the Rs4081 RBD and used to infect. (e) BHK cells transfected with human ACE2 or empty vector, or (f) Huh-7 cells, with or without trypsin. Shown are the data from quadruplicate infections. (g to i) Caco2, 293T, or Huh-7 were pretreated with the indicated protease inhibitors and infected with pseudotyped particles, with or without trypsin. Error bars represent the corresponding mean ± SEM.

FIG 3

Purified clade 2 virus RBD binds to human cells. (a) Schematic overview of spike fragments used in binding assays. (b) Expression of spike fragments in HEK 293T/17 cells. (c) Purified spike fragments were incubated with Huh-7 cells at the indicated concentration, and binding was measured by FACS. (d) Heatmap representation of the percentage of cells bound by spike protein from the FACS binding data. (e) Schematic overview of pseudotyped particles bearing different spikes used in binding assays. (f) Purified pseudotyped particles were incubated with Huh-7 cells at the indicated concentration, and binding was measured by FACS. (g) Heatmap representation of the percentage of cells bound by pseudotyped particles from the FACS binding data. (h) Huh-7cells were incubated with 50 μg/mL RBD/S1-Fc proteins or DMEM (mock) at 37°C for 60 min before inoculating with 2-fold serial diluted pseudotyped stocks. Error bars represent the corresponding mean ± SEM. ****, P < 0.0001; **, P = 0.0017; *, P < 0.05.

FIG 4

Known coronavirus receptors do not support Sarbecovirus RBD clade 2 infection. (a) Cells were transfected with the human or bat orthologues of known coronavirus receptors and infected with virus. (b) Cells were incubated with an antibody directed toward ACE2 and subsequently infected with viral pseudotyped. Error bars represent the corresponding mean ± SEM. *, P = 0.0332; ****, P < 0.0001.

FIG 5

Essential role for exogenous protease in CoV entry. (a) Surface-bound host-cell protease mediates clade 1 RBD entry following ACE2 binding. (b) Data from this study exclude the possibility that trypsin may act on the receptor and support the possibility (c) that trypsin cleaves viral spike during entry. (d) Trypsin-dependent, bat-derived betacoronaviruses contain deletions in their RBDs compared to closely related coronaviruses with known receptors. (e) Alphacoronaviruses with known receptors also differ from a subset of closely related but trypsin-dependent, animal viruses with deletions or insertions in their RBDs.